Anti-ROR antibody constructs

Abstract

To provide anti-ROR antibody constructs, pharmaceutical compositions comprising the constructs, and methods of use thereof.SOLUTION: In a certain embodiment, the ROR antigen is ROR1. In a certain embodiment, the ROR antigen is ROR2. In a certain embodiment, the ROR antigen is ROR1 and ROR2. In a certain embodiment, the ROR antigen is a domain selected from the group consisting of ROR1 Frizzle domain, ROR2 Frizzle domain, ROR1 Ig-like domain, ROR2 Ig-like domain, ROR1 Kringle domain and ROR2 Kringle domain. In a certain embodiment, the ROR antigen includes a human ROR antigen.SELECTED DRAWING: None

Description

[Technical Field]

[0001] 1. CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 62 / 659,635, filed April 18, 2018, the entire disclosure of which is incorporated herein by reference. 2. Sequence Listing

[0002] This application incorporates by reference the Sequence Listing that was submitted herewith as an ASCII text file, said ASCII text file being entitled 14529-001-228_SEQ_LISTING.txt, created on April 15, 2019, and 336,627 bytes in size. [Background technology]

[0003] 3.Background The design and therapeutic use of multispecific antibodies, antibody-derived proteins engineered to recognize multiple targets, is an area of ​​intense research. Multispecific antibodies offer the promise of greater therapeutic control than routinely provided by monospecific monoclonal antibodies. For example, multispecific antibodies can be engineered to provide greater target specificity than monospecific antibodies and reduce the off-target effects associated with many antibody therapies, particularly antibody-based immunotherapies. Multispecific antibodies also offer the promise of therapeutic strategies not possible with monospecific antibodies, such as synergistic targeting of multiple cell receptors, particularly in immunotherapy. One such immunotherapy is the use of bispecific antibodies to recruit and redirect T cells to target and kill specific tumor cell populations through the bispecific engagement of T cell and tumor cell markers. For example, targeting B-cell lymphoma using CD3xCD19 bispecific antibodies, such as the CD3xCD19 BiTE blinatumomab (Blincyto), is described in US Patent Application Publication No. 2006 / 0193852.

[0004] Thus, there is a need for improved multispecific antibodies that specifically bind to distinct cell populations, including tumor cell populations, with improvements including increased affinity or avidity, reduced off-target binding, and / or reduced unintended immune activation.

[0005] Gentile, et al. (Cancer Res; 71(8) April 15, 2011), Rebagay, et al. (Front. Oncol., 18 April 2012), Zhang, et al. (American Journal of Pathology, Vol. 181, No. 6, December 2012), Henry, et al. (Oncotarget, Vol. 6, No. 37 2015), Zhang, et al. (PLoS ONE 7(3): e31127.) and Bainbridge, et al. (PLoS ONE 9(7): e102695.), various tumors can demonstrate cell surface expression of tyrosine protein kinase transmembrane receptor (ROR) antigens. In addition, Balakrishnan et al. (Clin Cancer Res. 2017 Jun 15; 23(12): As described in Gohil et al. (Oncoimmunology. 2017; 3061-3071), ROR expression can be demonstrated to be absent or only limitedly expressed in normal, i.e., non-cancerous, tissues. Thus, ROR antigens can be used as tumor-specific markers in certain tumors. Examples of tumors and cancers with demonstrated ROR expression include Gohil et al. (Oncoimmunology. 2017; 3061-3071), which are incorporated herein in their entirety. 6(7): e1326437.) Cancers include, but are not limited to, pancreatic cancer, ovarian cancer, breast cancer, lung cancer, gastric cancer, melanoma, Ewing's sarcoma, chronic lymphocytic leukemia, mantle cell lymphoma, and B-ALL. Other cancers include, but are not limited to, hematological cancers, prostate cancer, colon cancer, renal cancer, and uterine cancer. The use of ROR multispecific antibodies formatted in various antibody platforms to target tumors is described in Gohil, et al., International Application WO2017 / 053469, International Application WO2014 / 167022, U.S. Patent Application Publication No. 2017 / 0198045, International Application WO2016 / 094873, International Application WO2017 / 127499, and International Application WO2016 / 142768, each of which is incorporated herein by reference in its entirety. Thus, ROR antigen-binding molecules have therapeutic potential in the treatment of cancer. Multispecific ROR-binding molecules that bind to T cell surface antigens in addition to ROR antigens have the potential to provide T cell-redirected killing of ROR-expressing cancer cells. Thus, there is a need for ROR antigen-binding molecules, including multispecific ROR antigen-binding molecules. There is also a need for ROR antigen-binding molecules with improvements, including increased affinity or avidity, reduced off-target binding, and / or reduced unintended immune activation. There is a particular need for multispecific ROR antigen-binding molecules that have improved manufacturability and are easily purified. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Gentile, et al. (Cancer Res; 71(8) April 15, 2011) [Patent Document 2] Rebagay, et al. (Front. Oncol., 18 April 2012) [Patent Document 3] Zhang, et al. (American Journal of Pathology, Vol. 181, No. 6, December 2012) [Non-patent literature]

[0007] [Non-Patent Document 1] International Application No. 2017 / 053469 [Non-patent document 2] International Application No. 2014 / 167022 [Non-patent document 3] US Patent Application Publication No. 2017 / 0198045 Summary of the Invention [Means for solving the problem]

[0008] 4. Overview In a first aspect, an antigen-binding molecule is provided. In all embodiments, the antigen-binding molecule comprises at least one antigen-binding site specific for the ROR antigen, and thus the binding molecule is designated as a ROR antigen-binding molecule.

[0009] Described herein is a ROR antigen-binding molecule comprising: A) CDR1, CDR2, and CDR3 amino acid sequences of a light chain variable region (VL) derived from a tyrosine protein kinase transmembrane receptor (ROR) antigen-binding site, wherein the CDR1, CDR2, and CDR3 VL sequences are selected from Table 6; and B) CDR1, CDR2, and CDR3 amino acid sequences of a heavy chain variable region (VH) derived from a ROR antigen-binding site, wherein the CDR1, CDR2, and CDR3 VH sequences are selected from Table 6. The ROR antigen-binding molecule is a first antigen-binding site specific for (i) ROR1 and ROR2, (ii) ROR1, or (iii) ROR2. In certain embodiments, the ROR antigen-binding molecule further comprises a second antigen-binding site. In certain embodiments, the second antigen-binding site is the same as the first antigen-binding site. In certain embodiments, the second antigen-binding site is specific for a second antigen that is different from the ROR antigen of the first antigen-binding site, hi certain embodiments, the second antigen is the CD3 antigen.

[0010] Described herein is a tyrosine protein kinase transmembrane receptor (ROR) antigen binding molecule comprising first and second polypeptide chains, wherein: (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged, from N-terminus to C-terminus, in an A-B-D-E orientation, and Domain A has a VL amino acid sequence, Domain B has a CH3 amino acid sequence, Domain D has a CH2 amino acid sequence, and Domain E has a constant region domain amino acid sequence; (b) the second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged, from N-terminus to C-terminus, in an F-G orientation, and Domain F has a VH amino acid sequence and Domain G has a CH3 amino acid sequence; and (c) the first and second polypeptides associate via interactions between Domain A and Domain F and between Domain B and Domain G to form the ROR antigen binding molecule, and the interaction between Domain A and Domain F forms a first antigen binding site specific for the ROR antigen.

[0011] In certain embodiments, the ROR antigen is ROR1. In certain embodiments, the ROR antigen is ROR2. In certain embodiments, the ROR antigen is ROR1 and ROR2. In certain embodiments, the ROR antigen is a domain selected from the group consisting of a ROR1 Frizzle domain, a ROR2 Frizzle domain, a ROR1 Ig-like domain, a ROR2 Ig-like domain, a ROR1 Kringle domain, and a ROR2 Kringle domain. In certain embodiments, the ROR antigen comprises a human ROR antigen.

[0012] In certain embodiments, Domain A comprises the CDR1, CDR2, and CDR3 amino acid sequences of a specific light chain variable region (VL) from a specific ROR antigen binding site, wherein the CDR1, CDR2, and CDR3 VL sequence is selected from Table 6, and Domain F comprises the CDR1, CDR2, and CDR3 amino acid sequences of a specific heavy chain variable region (VH) from a specific ROR antigen binding site, wherein the CDR1, CDR2, and CDR3 VH sequence is selected from Table 6. In certain embodiments, the specific ROR antigen binding site is I2A-10, I2A-10 D54E Y55Q, or I2A-27 of Table 6.

[0013] In certain embodiments, Domain A comprises a VL having one or two amino acid mutations compared to the VL sequence of an antibody in Table 6, wherein the one or two amino acid mutations are present in one or more CDR regions in the VL. In certain embodiments, Domain F comprises a VH having one or two amino acid mutations compared to the VH sequence of an antibody in Table 6, wherein the one or two amino acid mutations are present in one or more CDR regions in the VH. In certain embodiments, Domain A comprises a VL having the VL sequence of I2A-10 with one or more mutations in one or more CDR regions. In certain embodiments, Domain A comprises a VL having the VL sequence of I2A-27 with one or more mutations in one or more CDR regions. In certain embodiments, Domain F comprises a VL having the VL sequence of I2A-10 with one or more mutations in one or more CDR regions. In certain embodiments, Domain F comprises a VL having the VL sequence of I2A-27 with one or more mutations in one or more CDR regions.

[0014] In certain embodiments, the amino acid sequences of the B domain and the G domain are identical, where the sequence is an endogenous CH3 sequence.

[0015] In certain embodiments, the amino acid sequences of the B domain and the G domain differ and each independently contain an orthogonal modification to the endogenous CH3 sequence, wherein the B domain interacts with the G domain, and neither the B domain nor the G domain significantly interacts with the CH3 domain lacking the orthogonal modification.

[0016] In certain embodiments, the orthogonal modifications of the B and G domains comprise mutations that create an engineered disulfide bridge between the B and G domains. In certain embodiments, the mutations in the B and G domains that create an engineered disulfide bridge are an S354C mutation in one of the B and G domains and a 349C mutation in the other domain.

[0017] In certain embodiments, the orthogonal modifications of the B and G domains comprise knob-in-hole mutations, ie, a T366W mutation in one of the B and G domains and T366S, L368A, and Y407V mutations in the other domain.

[0018] In certain embodiments, the orthogonal modifications of the B and G domains comprise charge-pair mutations, ie, a T366K mutation in one of the B and G domains and a L351D mutation in the other domain.

[0019] In certain embodiments, the E domain has a CH3 amino acid sequence.

[0020] In certain embodiments, the amino acid sequences of the E domain and the K domain are identical, wherein the sequence is an endogenous CH3 sequence.

[0021] In certain embodiments, the amino acid sequences of the E domain and the K domain are different. In certain embodiments, the different sequences each contain a distinct orthogonal modification to the endogenous CH3 sequence, where the E domain interacts with the K domain, and neither the E domain nor the K domain interacts significantly with the CH3 domain lacking the orthogonal modification.

[0022] In certain embodiments, the orthogonal modification comprises a mutation that creates an engineered disulfide bridge between the E and K domains, hi certain embodiments, the mutation that creates the engineered disulfide bridge is a S354C mutation in one of the E and K domains and a 349C mutation in the other domain.

[0023] In certain embodiments, the orthogonal modifications in the E and K domains comprise knobs-in-hole mutations, hi certain embodiments, the knobs-in-hole mutations are a T366W mutation in one of the E and K domains and T366S, L368A, and Y407V mutations in the other domain.

[0024] In certain embodiments, the orthogonal modifications in the E and K domains comprise charge-pair mutations, hi certain embodiments, the charge-pair mutations are a T366K mutation in one of the E or K domains and a corresponding L351D mutation in the other domain.

[0025] In certain embodiments, the amino acid sequences of the E and K domains are endogenous sequences of two different antibody domains selected to have specific interactions that promote specific association between the first and third polypeptides, hi certain embodiments, the two different amino acid sequences are a CH1 sequence and a CL sequence.

[0026] In certain embodiments, the sequence forming the junction between the A and B domains is IKRTPREP or IKRTVREP.

[0027] In certain embodiments, the sequence forming the junction between the F and G domains is SSASPREP.

[0028] In certain embodiments, at least one CH3 amino acid sequence has a C-terminal tripeptide insertion connecting the CH3 amino acid sequence to the hinge amino acid sequence, wherein the tripeptide insertion is selected from the group consisting of PGK, KSC, and GEC.

[0029] In certain embodiments, the sequences are human sequences.

[0030] In certain embodiments, at least one CH3 amino acid sequence is an IgG sequence. In certain embodiments, the IgG sequence is an IgG1 sequence.

[0031] In certain embodiments, at least one CH3 amino acid sequence has one or more isoallotypic mutations, hi certain embodiments, the isoallotypic mutations are D356E and L358M.

[0032] In certain embodiments, the CL amino acid sequence is a C kappa sequence.

[0033] In certain embodiments, the CH2 sequence has one or more engineered mutations that reduce Fc effector function. In certain embodiments, the one or more engineered mutations are at positions L234, L235, and P329. In certain embodiments, the one or more engineered mutations are L234A, L235A, and P329G. In certain embodiments, the one or more engineered mutations are L234A, L235A, and P329K.

[0034] A tyrosine protein kinase transmembrane receptor (ROR) antigen binding molecule comprising first, second, third, and fourth polypeptide chains, wherein: (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged, from N-terminus to C-terminus, in an A-B-D-E orientation, and Domain A has a VL amino acid sequence, Domain B has a CH3 amino acid sequence, Domain D has a CH2 amino acid sequence, and Domain E has a constant region domain amino acid sequence; (b) the second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged, from N-terminus to C-terminus, in an F-G orientation, and Domain F has a VH amino acid sequence and Domain G has a CH3 amino acid sequence; and (c) the third polypeptide chain comprises Domain H, Domain I, Domain J, and Domain K, wherein the domains are arranged, from N-terminus to C-terminus, in an H-I-J-K orientation, and Domain H has a variable region domain amino acid sequence, Domain I has a constant region domain amino acid sequence, Domain J has a CH2 amino acid sequence, and Domain K has a constant region domain amino acid sequence. (d) the fourth polypeptide chain comprises domain L and domain M, wherein the domains are arranged N-terminally to C-terminally in an LM orientation, and domain L has a variable region domain amino acid sequence, and domain M has a constant region domain amino acid sequence; (e) the first and second polypeptides associate through interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate through interactions between the H and L domains and between the I and M domains; (g) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form a ROR antigen binding molecule, wherein the interaction between the A and F domains forms a first antigen binding site, and the interaction between the H and L domains forms a second antigen binding site, and the first antigen binding site, the second antigen binding site, or the first and second antigen binding sites are specific for a ROR antigen. Also described herein are ROR antigen binding molecules.

[0035] In certain embodiments, the first antigen-binding site is specific for the ROR antigen. In certain embodiments, the second antigen-binding site is specific for the ROR antigen. In certain embodiments, the first and second antigen-binding sites are specific for the ROR antigen.

[0036] In certain embodiments, the ROR antigen is ROR1. In certain embodiments, the ROR antigen is ROR2. In certain embodiments, the ROR antigen is ROR1 and ROR2. In certain embodiments, the ROR antigen is a domain selected from the group consisting of a ROR1 Frizzle domain, a ROR2 Frizzle domain, a ROR1 Ig-like domain, a ROR2 Ig-like domain, a ROR1 Kringle domain, and a ROR2 Kringle domain. In certain embodiments, the ROR antigen comprises a human ROR antigen.

[0037] In certain embodiments, Domain A comprises the CDR1, CDR2, and CDR3 amino acid sequences of a specific light chain variable region (VL) from a specific ROR antigen binding site, wherein the CDR1, CDR2, and CDR3 VL sequence is selected from Table 6, and Domain F comprises the CDR1, CDR2, and CDR3 amino acid sequences of a specific heavy chain variable region (VH) from a specific ROR antigen binding site, wherein the CDR1, CDR2, and CDR3 VH sequence is selected from Table 6. In certain embodiments, the specific ROR antigen binding site is I2A-10, I2A-10 D54E Y55Q, or I2A-27 of Table 6.

[0038] In certain embodiments, Domain A comprises a VL having one or two amino acid mutations compared to the VL sequence of an antibody in Table 6, wherein the one or two amino acid mutations are present in one or more CDR regions in the VL. In certain embodiments, Domain F comprises a VH having one or two amino acid mutations compared to the VH sequence of an antibody in Table 6, wherein the one or two amino acid mutations are present in one or more CDR regions in the VH. In certain embodiments, Domain A comprises a VL having the VL sequence of I2A-10 with one or more mutations in one or more CDR regions. In certain embodiments, Domain A comprises a VL having the VL sequence of I2A-27 with one or more mutations in one or more CDR regions. In certain embodiments, Domain F comprises a VL having the VL sequence of I2A-10 with one or more mutations in one or more CDR regions. In certain embodiments, Domain F comprises a VL having the VL sequence of I2A-27 with one or more mutations in one or more CDR regions.

[0039] In certain embodiments, the second antigen-binding site comprises: A) a specific light chain variable region (VL) amino acid sequence within the third polypeptide chain selected from the group consisting of SEQ ID NO:69 and SEQ ID NO:73; and B) a specific heavy chain variable region (VH) amino acid sequence within the fourth polypeptide chain selected from the group consisting of SEQ ID NO:68, SEQ ID NO:70, SEQ ID NO:71, and SEQ ID NO:72.

[0040] In certain embodiments, the second antigen-binding site comprises: A) a specific light chain variable region (VL) amino acid sequence within the fourth polypeptide chain selected from the group consisting of SEQ ID NO:69 and SEQ ID NO:73; and B) a specific heavy chain variable region (VH) amino acid sequence within the third polypeptide chain selected from the group consisting of SEQ ID NO:68, SEQ ID NO:70, SEQ ID NO:71, and SEQ ID NO:72.

[0041] In certain embodiments, the amino acid sequences of the B domain and the G domain are identical and the sequence is an endogenous CH3 sequence.

[0042] In certain embodiments, the amino acid sequences of the B and G domains are different and each independently contain an orthogonal modification to the endogenous CH3 sequence, wherein the B domain interacts with the G domain and neither the B nor the G domain interacts significantly with the CH3 domain lacking the orthogonal modification.

[0043] In certain embodiments, the orthogonal modifications of the B and G domains comprise mutations that create an engineered disulfide bridge between the B and G domains. In certain embodiments, the mutations in the B and G domains that create an engineered disulfide bridge are a S354C mutation in one of the B and G domains and a 349C mutation in the other domain.

[0044] In certain embodiments, the orthogonal modifications of the B and G domains comprise knobs-in-hole mutations, ie, a T366W mutation in one of the B and G domains and T366S, L368A, and Y407V mutations in the other domain.

[0045] In certain embodiments, the orthogonal modifications of the B and G domains comprise charge-pair mutations, ie, a T366K mutation in one of the B and G domains and a L351D mutation in the other domain.

[0046] In certain embodiments, the E domain has a CH3 amino acid sequence.

[0047] In certain embodiments, the amino acid sequences of the E domain and the K domain are identical and the sequence is an endogenous CH3 sequence.

[0048] In certain embodiments, the amino acid sequences of the E domain and the K domain are different. In certain embodiments, the different sequences each contain a distinct orthogonal modification to the endogenous CH3 sequence, and the E domain interacts with the K domain, and neither the E domain nor the K domain interacts significantly with the CH3 domain lacking the orthogonal modification.

[0049] In certain embodiments, the orthogonal modification comprises a mutation that creates an engineered disulfide bridge between the E and K domains, hi certain embodiments, the mutation that creates the engineered disulfide bridge is a S354C mutation in one of the E and K domains and a 349C mutation in the other domain.

[0050] In certain embodiments, the orthogonal modifications in the E and K domains comprise knobs-in-hole mutations, ie, a T366W mutation in one of the E or K domains and T366S, L368A, and Y407V mutations in the other domain.

[0051] In certain embodiments, the orthogonal modifications in the E and K domains comprise charge-pair mutations, hi certain embodiments, the charge-pair mutations are a T366K mutation in one of the E or K domains and a corresponding L351D mutation in the other domain.

[0052] In certain embodiments, the amino acid sequences of the E and K domains are endogenous sequences of two different antibody domains selected to have specific interactions that promote specific association between the first and third polypeptides, hi certain embodiments, the two different amino acid sequences are a CH1 sequence and a CL sequence.

[0053] In certain embodiments, domain I has a CL sequence and domain M has a CH1 sequence.

[0054] In certain embodiments, domain H has a VL sequence and domain L has a VH sequence.

[0055] In certain embodiments, domain H has a VL amino acid sequence; domain I has a CL amino acid sequence; domain K has a CH3 amino acid sequence; domain L has a VH amino acid sequence; and domain M has a CH1 amino acid sequence.

[0056] In certain embodiments, the ROR antigen binding molecule further comprises a fifth polypeptide chain, wherein (a) the first polypeptide chain further comprises domain N and domain O, wherein the domains are arranged from N-terminus to C-terminus in a N-A-B-D-E orientation, and domain N has a variable region domain amino acid sequence, and domain O has a constant region domain amino acid sequence; (b) the ROR antigen binding molecule further comprises a fifth polypeptide chain comprising domain P and domain Q, wherein the domains are arranged from N-terminus to C-terminus in a P-Q orientation, and domain P has a variable region domain amino acid sequence, and domain Q has a constant region domain amino acid sequence; and (c) the first and fifth polypeptides associate via interactions between the N and P domains and between the O and Q domains to form the ROR antigen binding molecule.

[0057] In certain embodiments, (a) the amino acid sequences of Domain N and Domain A are identical, the amino acid sequence of Domain H is different from the sequences of Domain N and Domain A, the amino acid sequence of Domain O and Domain B is identical, the amino acid sequence of Domain I is different from the sequences of Domain O and Domain B, the amino acid sequences of Domain P and Domain F are identical, the amino acid sequence of Domain L is different from the sequences of Domain P and Domain F, the amino acid sequence of Domain Q and Domain G is identical, and the amino acid sequence of Domain M is different from the sequences of Domain Q and Domain G; (b) the interaction between Domain A and Domain F forms a first antigen-binding site specific for a first antigen, the interaction between Domain H and Domain L forms a second antigen-binding site specific for a second antigen, and the interaction between Domain N and Domain P forms a third antigen-binding site specific for the first antigen. In certain embodiments, the first antigen is an ROR antigen. In certain embodiments, the second antigen is a CD3 antigen.

[0058] In certain embodiments, (a) the amino acid sequences of Domain N, Domain A, and Domain H are different, the amino acid sequences of Domain O, Domain B, and Domain I are different, the amino acid sequences of Domain P, Domain F, and Domain L are different, and the amino acid sequences of Domain Q, Domain G, and Domain M are different; (b) the interaction between Domain A and Domain F forms a first antigen-binding site specific for a first antigen, the interaction between Domain H and Domain L forms a second antigen-binding site specific for a second antigen, Domain N and Domain P form a third antigen-binding site specific for a third antigen, and (c) the first, second, or third antigen is a ROR antigen.

[0059] In certain embodiments, the ROR antigen binding molecule further comprises a sixth polypeptide chain, wherein (a) the third polypeptide chain further comprises domain R and domain S, wherein the domains are arranged from N-terminus to C-terminus in the orientation RSHIJK, where domain R has a variable region domain amino acid sequence, and domain S has a constant domain amino acid sequence; (b) the ROR antigen binding molecule further comprises a sixth polypeptide chain comprising domain T and domain U, wherein the domains are arranged from N-terminus to C-terminus in the orientation TU, where domain T has a variable region domain amino acid sequence, and domain U has a constant domain amino acid sequence; and (c) the third and sixth polypeptides associate via interactions between the R and T domains and between the S and U domains to form the ROR antigen binding molecule.

[0060] In certain embodiments, (a) the amino acid sequences of domains R and A are identical, the amino acid sequence of domain H is different from the sequences of domains R and A, the amino acid sequences of domains S and B are identical, the amino acid sequence of domain I is different from the sequences of domains S and B, the amino acid sequences of domains T and F are identical, the amino acid sequence of domain L is different from the sequences of domains T and F, the amino acid sequences of domains U and G are identical, and the amino acid sequence of domain M is different from the sequences of domains U and G; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen; and the interaction between the R and T domains forms a third antigen-binding site specific for the first antigen. In certain embodiments, the first antigen is an ROR antigen. In certain embodiments, the second antigen is a CD3 antigen.

[0061] In certain embodiments, (a) the amino acid sequences of domains R and H are identical, the amino acid sequence of domain A is different from the sequences of domains R and H, the amino acid sequences of domains S and I are identical, the amino acid sequence of domain B is different from the sequences of domains S and I, the amino acid sequences of domains T and L are identical, the amino acid sequence of domain F is different from the sequences of domains T and L, the amino acid sequences of domains U and M are identical, and the amino acid sequence of domain G is different from the sequences of domains U and M; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and the interaction between the R and T domains forms a third antigen-binding site specific for the second antigen. In certain embodiments, the second antigen is an ROR antigen. In certain embodiments, the first antigen is a CD3 antigen.

[0062] In certain embodiments, (a) the amino acid sequences of domains R, A, and H are different, the amino acid sequences of domains S, B, and I are different, the amino acid sequences of domains T, F, and L are different, and the amino acid sequences of domains U, G, and M are different; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and the interaction between the R and T domains forms a third antigen-binding site specific for a third antigen; and (c) the first, second, or third antigen is an ROR antigen.

[0063] In certain embodiments, the ROR antigen binding molecule further comprises fifth and sixth polypeptide chains, wherein (a) the first polypeptide chain further comprises domain N and domain O, wherein the domains are arranged N-terminally to C-terminally in a NOABDE orientation; (b) the third polypeptide chain further comprises domain R and domain S, wherein the domains are arranged N-terminally to C-terminally in a RSHIJK orientation; (c) the ROR antigen binding molecule further comprises fifth and sixth polypeptide chains, wherein the fifth polypeptide chain comprises domain P and domain Q, wherein the domains are arranged N-terminally to C-terminally in a PQ orientation, and the sixth polypeptide chain comprises domain T and domain U, wherein the domains are arranged N-terminally to C-terminally in a TU orientation; (d) the first and fifth polypeptides associate through interactions between the N and P domains and between the O and Q domains, and the third and sixth polypeptides associate through interactions between the R and T domains and between the S and U domains to form the ROR antigen binding molecule.

[0064] In certain embodiments, (a) the amino acid sequences of Domain N and Domain A are identical, the amino acid sequences of Domain H and Domain R are identical, the amino acid sequences of Domain O and Domain B are identical, the amino acid sequences of Domain I and Domain S are identical, the amino acid sequences of Domain P and Domain F are identical, the amino acid sequences of Domain L and Domain T are identical, the amino acid sequences of Domain Q and Domain G are identical, and the amino acid sequences of Domain M and Domain U are identical; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, Domain N and Domain P form a second antigen-binding site specific for the first antigen, the interaction between the H and L domains forms a third antigen-binding site specific for a second antigen, and the interaction between the R and T domains forms a fourth antigen-binding site specific for the second antigen.

[0065] In certain embodiments, (a) the amino acid sequences of Domain H and Domain A are identical, the amino acid sequences of Domain N and Domain R are identical, the amino acid sequences of Domain I and Domain B are identical, the amino acid sequences of Domain O and Domain S are identical, the amino acid sequences of Domain L and Domain F are identical, the amino acid sequences of Domain P and Domain T are identical, the amino acid sequences of Domain M and Domain G are identical, and the amino acid sequences of Domain Q and Domain U are identical; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, Domain N and Domain P form a second antigen-binding site specific for a second antigen, the interaction between the H and L domains forms a third antigen-binding site specific for the first antigen, and the interaction between the R and T domains forms a fourth antigen-binding site specific for the second antigen.

[0066] In certain embodiments, the sequence forming the junction between the A and B domains is IKRTPREP or IKRTVREP.

[0067] In certain embodiments, the sequence forming the junction between the F and G domains is SSASPREP.

[0068] In certain embodiments, at least one CH3 amino acid sequence has a C-terminal tripeptide insertion connecting the CH3 amino acid sequence to the hinge amino acid sequence, wherein the tripeptide insertion is selected from the group consisting of PGK, KSC, and GEC.

[0069] In certain embodiments, the sequences are human sequences.

[0070] In certain embodiments, at least one CH3 amino acid sequence is an IgG sequence. In certain embodiments, the IgG sequence is an IgG1 sequence.

[0071] In certain embodiments, at least one CH3 amino acid sequence has one or more isoallotypic mutations, hi certain embodiments, the isoallotypic mutations are D356E and L358M.

[0072] In certain embodiments, the CL amino acid sequence is a C kappa sequence.

[0073] In certain embodiments, the CH2 sequence has one or more engineered mutations that reduce Fc effector function. In certain embodiments, the one or more engineered mutations are at positions L234, L235, and P329. In certain embodiments, the one or more engineered mutations are L234A, L235A, and P329G. In certain embodiments, the one or more engineered mutations are L234A, L235A, and P329K.

[0074] Also described herein are tyrosine protein kinase transmembrane receptor (ROR) antigen binding molecules comprising a first antigen-binding site specific for the ROR antigen, wherein the first antigen-binding site comprises: A) specific light chain variable region (VL) CDR1, CDR2, and CDR3 amino acid sequences derived from the specific ROR antigen-binding site, wherein the CDR1, CDR2, and CDR3 VL sequence is selected from Table 6; and B) specific heavy chain variable region (VH) CDR1, CDR2, and CDR3 amino acid sequences derived from the specific ROR antigen-binding site, wherein the CDR1, CDR2, and CDR3 VH sequence is selected from Table 6.

[0075] In certain embodiments, the first antigen-binding site comprises a VL having one or two amino acid mutations compared to the VL sequence of an antibody in Table 6, wherein the one or two amino acid mutations are present in one or more CDR regions in the VL. In certain embodiments, the first antigen-binding site comprises a VH having one or two amino acid mutations compared to the VH sequence of an antibody in Table 6, wherein the one or two amino acid mutations are present in one or more CDR regions in the VH.

[0076] In certain embodiments, the first antigen-binding site is specific for ROR1. In certain embodiments, the first antigen-binding site is specific for ROR2. In certain embodiments, the first antigen-binding site is specific for ROR1 and ROR2. In certain embodiments, the ROR antigen is a domain selected from the group consisting of a ROR1 Frizzle domain, a ROR2 Frizzle domain, a ROR1 Ig-like domain, a ROR2 Ig-like domain, a ROR1 Kringle domain, and a ROR2 Kringle domain. In certain embodiments, the ROR antigen comprises a human ROR antigen.

[0077] In certain embodiments, the ROR antigen-binding molecule further comprises a second antigen-binding site. In certain embodiments, the second antigen-binding site is specific for the ROR antigen. In certain embodiments, the second antigen-binding site is specific for a second antigen different from the ROR antigen. In certain embodiments, the second antigen is the CD3 antigen. In certain embodiments, the antigen-binding site is specific for an epitope of the CD3 antigen. In certain embodiments, the second antigen-binding site comprises A) a specific light chain variable region (VL) amino acid sequence selected from the group consisting of SEQ ID NO: 69 and SEQ ID NO: 73; and B) a specific heavy chain variable region (VH) amino acid sequence selected from the group consisting of SEQ ID NO: 68, SEQ ID NO: 70, SEQ ID NO: 71, and SEQ ID NO: 72. In certain embodiments, the second antigen-binding site comprises A) a specific light chain variable region (VL) amino acid sequence selected from the group consisting of SEQ ID NO:69 and SEQ ID NO:73; and B) a specific heavy chain variable region (VH) amino acid sequence selected from the group consisting of SEQ ID NO:68, SEQ ID NO:70, SEQ ID NO:71, and SEQ ID NO:72.

[0078] In certain embodiments, the ROR antigen binding molecule comprises an antibody format selected from the group consisting of a full-length antibody, a Fab fragment, an Fv, an scFv, a tandem scFv, a diabody, an sc diabody, a DART, a tandAb, and a minibody. In certain embodiments, the ROR antigen binding molecule comprises first and second polypeptide chains, wherein (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged N-terminally to C-terminally in an A-B-D-E orientation, and Domain A has a variable region domain amino acid sequence, and Domain B, Domain D, and Domain E have constant region domain amino acid sequences; (b) the second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged N-terminally to C-terminally in an F-G orientation, and Domain F has a variable region domain amino acid sequence, and Domain G has a constant region domain amino acid sequence; and c) the first and second polypeptides associate via interactions between Domain A and Domain F and between Domain B and Domain G to form the ROR antigen binding molecule, and the interaction between Domain A and Domain F forms a first antigen binding site.

[0079] In certain embodiments, the ROR antigen binding molecule further comprises third and fourth polypeptide chains, wherein (a) the third polypeptide chain comprises domain H, domain I, domain J, and domain K, wherein the domains are arranged N-terminally to C-terminally in a H-I-J-K orientation, and domain H has a variable region domain amino acid sequence, and domains I, J, and K have constant region domain amino acid sequences; (b) the fourth polypeptide chain comprises domain L and domain M, wherein the domains are arranged N-terminally to C-terminally in a L-M orientation, and domain L has a variable region domain amino acid sequence, and domain M has a constant region domain amino acid sequence; (c) the third and fourth polypeptides associate through interactions between the H and L domains and between the I and M domains; and (d) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form the ROR antigen binding molecule, and the interaction between the H and L domains forms a second antigen binding site. In certain embodiments, the first antigen-binding site is specific for the ROR antigen, hi certain embodiments, the second antigen-binding site is specific for CD3.

[0080] In certain embodiments, domain B and domain G have a CH3 amino acid sequence.

[0081] In certain embodiments, the amino acid sequences of the B domain and the G domain are identical and the sequence is an endogenous CH3 sequence.

[0082] In certain embodiments, the amino acid sequences of the B and G domains are different and each independently contain an orthogonal modification to the endogenous CH3 sequence, wherein the B domain interacts with the G domain and neither the B nor the G domain interacts significantly with the CH3 domain lacking the orthogonal modification.

[0083] In certain embodiments, the orthogonal modifications of the B and G domains comprise mutations that create an engineered disulfide bridge between the B and G domains. In certain embodiments, the mutations in the B and G domains that create an engineered disulfide bridge are a S354C mutation in one of the B and G domains and a 349C mutation in the other domain.

[0084] In certain embodiments, the orthogonal modifications of the B and G domains comprise knobs-in-hole mutations, ie, a T366W mutation in one of the B and G domains and T366S, L368A, and Y407V mutations in the other domain.

[0085] In certain embodiments, the orthogonal modifications of the B and G domains comprise charge-pair mutations, ie, a T366K mutation in one of the B and G domains and a L351D mutation in the other domain.

[0086] In certain embodiments, Domain B and Domain G have an IgM CH2 amino acid sequence or an IgE CH2 amino acid sequence. In certain embodiments, the IgM CH2 amino acid sequence or the IgE CH2 amino acid sequence comprises an orthogonal modification.

[0087] In certain embodiments, domain I has a CL sequence and domain M has a CH1 sequence. In certain embodiments, domain I has a CH1 sequence and domain M has a CL sequence. In certain embodiments, the CH1 and CL sequences each comprise one or more orthogonal modifications, and the domain having the CH1 sequence does not significantly interact with a domain having a CL sequence that lacks the orthogonal modification.

[0088] In certain embodiments, the orthogonal modification comprises a mutation that creates an engineered disulfide bridge between at least one CH1 domain and a CL domain, wherein the mutation is selected from the group consisting of an engineered cysteine ​​at position 138 of the CH1 sequence and position 116 of the CL sequence; an engineered cysteine ​​at position 128 of the CH1 sequence and position 119 of the CL sequence; and an engineered cysteine ​​at position 129 of the CH1 sequence and position 210 of the CL sequence.

[0089] In certain embodiments, the orthogonal modification comprises a mutation that creates an engineered disulfide bridge between at least one CH1 domain and a CL domain, wherein the mutation comprises an engineered cysteine ​​at position 128 of the CH1 sequence and at position 118 of the CL kappa sequence.

[0090] In certain embodiments, the orthogonal modification comprises a mutation that creates an engineered disulfide bridge between at least one CH1 domain and a CL domain, and the mutation is selected from the group consisting of an A141C in the CH1 sequence corresponding to an F118C mutation in the CL sequence; an L128C in the CH1 sequence corresponding to an F118C mutation in the CL sequence; and a P171C mutation in the CH1 sequence corresponding to an S162C mutation in the CL sequence.

[0091] In certain embodiments, the orthogonal modification comprises at least one charge pair mutation between the CH1 domain and the CL domain, wherein the charge pair mutation is selected from the group consisting of: A141L in the CH1 sequence corresponding to an F118S mutation in the CL sequence; A141L in the CH1 sequence corresponding to an F118A mutation in the CL sequence; A141L in the CH1 sequence corresponding to an F118V mutation in the CL sequence; and K147D in the CH1 sequence corresponding to a T129R mutation in the CL sequence.

[0092] In certain embodiments, the orthogonal modification comprises at least one charge pair mutation between the CH1 domain and the CL domain, wherein the charge pair mutation is selected from the group consisting of a G166D in the CH1 sequence corresponding to an N138K mutation in the CL sequence, and a G166K in the CH1 sequence corresponding to an N138D mutation in the CL sequence.

[0093] In certain embodiments, Domain A has a VL amino acid sequence and Domain F has a VH amino acid sequence. In certain embodiments, Domain A has a VH amino acid sequence and Domain F has a VL amino acid sequence.

[0094] In certain embodiments, Domain H has a VL amino acid sequence and Domain L has a VH amino acid sequence. In certain embodiments, Domain H has a VH amino acid sequence and Domain L has a VL amino acid sequence.

[0095] In certain embodiments, domain D and domain J have a CH2 amino acid sequence.

[0096] In certain embodiments, the E domain has a CH3 amino acid sequence.

[0097] In certain embodiments, the amino acid sequences of the E domain and the K domain are identical and the sequence is an endogenous CH3 sequence.

[0098] In certain embodiments, the amino acid sequences of the E domain and the K domain are different. In certain embodiments, the different sequences each contain a distinct orthogonal modification to the endogenous CH3 sequence, wherein the E domain interacts with the K domain, and neither the E domain nor the K domain interacts significantly with the CH3 domain lacking the orthogonal modification.

[0099] In certain embodiments, the orthogonal modification comprises a mutation that creates an engineered disulfide bridge between the E and K domains. In certain embodiments, the mutation that creates an engineered disulfide bridge is a S354C mutation in one of the E and K domains and a 349C mutation in the other. In certain embodiments, the orthogonal modification in the E and K domains comprises a knob-in-hole mutation. In certain embodiments, the knob-in-hole mutation is a T366W mutation in one of the E or K domains and a T366S, L368A, and Y407V mutation in the other. In certain embodiments, the orthogonal modification in the E and K domains comprises a charge-pair mutation. In certain embodiments, the charge-pair mutation is a T366K mutation in one of the E or K domains and a corresponding L351D mutation in the other.

[0100] In certain embodiments, the amino acid sequences of the E and K domains are endogenous sequences of two different antibody domains selected to have specific interactions that promote specific association between the first and third polypeptides, hi certain embodiments, the two different amino acid sequences are a CH1 sequence and a CL sequence.

[0101] In certain embodiments, the ROR antigen-binding molecule further comprises a third antigen-binding site. In certain embodiments, the third antigen-binding site is specific for an ROR antigen. In certain embodiments, the first antigen-binding site and the third antigen-binding site are specific for the same ROR antigen. In certain embodiments, the first antigen-binding site and the third antigen-binding site are specific for different ROR antigens.

[0102] In certain embodiments, the ROR antigen binding molecule comprises a fifth polypeptide chain, wherein (a) the first polypeptide chain further comprises domain N and domain O, wherein the domains are arranged from N-terminus to C-terminus in a NOABDE orientation, where domain N has a variable region domain amino acid sequence, and domain O has a constant region amino acid sequence; (b) the fifth polypeptide chain comprises domain P and domain Q, wherein the domains are arranged from N-terminus to C-terminus in a PQ orientation, where domain P has a variable region domain amino acid sequence, and domain Q has a constant region amino acid sequence; and (c) the first and fifth polypeptides associate through interactions between the N and P domains and between the O and Q domains to form the ROR antigen binding molecule.

[0103] In certain embodiments, (a) the amino acid sequences of Domain N and Domain A are identical, the amino acid sequence of Domain H is different from the sequences of Domain N and Domain A, the amino acid sequence of Domain O and Domain B is identical, the amino acid sequence of Domain I is different from the sequences of Domain O and Domain B, the amino acid sequences of Domain P and Domain F are identical, the amino acid sequence of Domain L is different from the sequences of Domain P and Domain F, the amino acid sequence of Domain Q and Domain G is identical, and the amino acid sequence of Domain M is different from the sequences of Domain Q and Domain G; (b) the interaction between Domain A and Domain F forms a first antigen-binding site specific for a first antigen, the interaction between Domain H and Domain L forms a second antigen-binding site specific for a second antigen, and the interaction between Domain N and Domain P forms a third antigen-binding site specific for the first antigen. In certain embodiments, the first antigen is an ROR antigen. In certain embodiments, the second antigen is a CD3 antigen.

[0104] In certain embodiments, (a) the amino acid sequences of Domain N, Domain A, and Domain H are different, the amino acid sequences of Domain O, Domain B, and Domain I are different, the amino acid sequences of Domain P, Domain F, and Domain L are different, and the amino acid sequences of Domain Q, Domain G, and Domain M are different; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and the interaction between the N and P domains forms a third antigen-binding site specific for a third antigen.

[0105] In certain embodiments, the ROR antigen binding molecule comprises a sixth polypeptide chain, wherein (a) the third polypeptide chain further comprises domain R and domain S, wherein the domains are arranged from N-terminus to C-terminus in the orientation RSHIJK, where domain R has a variable region amino acid sequence, and domain S has a constant domain amino acid sequence; (b) the sixth polypeptide chain comprises domain T and domain U, wherein the domains are arranged from N-terminus to C-terminus in the orientation TU, where domain T has a variable region amino acid sequence, and domain U has a constant domain amino acid sequence; and (c) the third and sixth polypeptides associate via interactions between the R and T domains and between the S and U domains to form the ROR antigen binding molecule.

[0106] In certain embodiments, (a) the amino acid sequences of domains R and A are identical, the amino acid sequence of domain H is different from the sequences of domains R and A, the amino acid sequences of domains S and B are identical, the amino acid sequence of domain I is different from the sequences of domains S and B, the amino acid sequences of domains T and F are identical, the amino acid sequence of domain L is different from the sequences of domains T and F, the amino acid sequences of domains U and G are identical, and the amino acid sequence of domain M is different from the sequences of domains U and G; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen; and the interaction between the R and T domains forms a third antigen-binding site specific for the first antigen. In certain embodiments, the first antigen is an ROR antigen. In certain embodiments, the second antigen is a CD3 antigen.

[0107] In certain embodiments, (a) the amino acid sequences of domains R and H are identical, the amino acid sequence of domain A is different from the sequences of domains R and H, the amino acid sequences of domains S and I are identical, the amino acid sequence of domain B is different from the sequences of domains S and I, the amino acid sequences of domains T and L are identical, the amino acid sequence of domain F is different from the sequences of domains T and L, the amino acid sequences of domains U and M are identical, and the amino acid sequence of domain G is different from the sequences of domains U and M; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and the interaction between the R and T domains forms a third antigen-binding site specific for the second antigen. In certain embodiments, the second antigen is an ROR antigen. In certain embodiments, the first antigen is a CD3 antigen.

[0108] In certain embodiments, (a) the amino acid sequences of Domain R, Domain A, and Domain H are different, the amino acid sequences of Domain S, Domain B, and Domain I are different, the amino acid sequences of Domain T, Domain F, and Domain L are different, and the amino acid sequences of Domain U, Domain G, and Domain M are different; (b) the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and the interaction between the R and T domains forms a third antigen-binding site specific for a third antigen.

[0109] Also described herein is a purified ROR antigen-binding molecule, including any of the ROR antigen-binding molecules described herein. In certain embodiments, the purified ROR antigen-binding molecule is purified by a purification method including a CH1 affinity purification step. In certain embodiments, the purification method is a single-step purification method.

[0110] Also described herein are pharmaceutical compositions comprising any of the ROR antigen binding molecules described herein and a pharmaceutically acceptable diluent.

[0111] Also described herein are methods for treating a subject having cancer, the method comprising administering a therapeutically effective amount of any of the pharmaceutical compositions described herein. In certain aspects, the cancer is selected from the group consisting of pancreatic cancer, ovarian cancer, breast cancer, lung cancer, gastric cancer, melanoma, Ewing's sarcoma, chronic lymphocytic leukemia, mantle cell lymphoma, B-ALL, hematological cancers, prostate cancer, colon cancer, renal cancer, and uterine cancer. 5. Brief description of the drawings [Brief explanation of the drawings]

[0112] [Figure 1] Figure 1 shows the alignment of the CH3-CH3 IgG1 dimer pair paired with CH1-CL. The quaternary structures are aligned with an RMSD of approximately 1.6 Å.

[0113] [Figure 2] FIG. 2 shows a schematic diagram of the construction of the various binding molecules (also referred to as antibody constructs) described herein, along with their respective nomenclature.

[0114] [Figure 3] FIG. 3 shows a higher resolution schematic of the polypeptide chains and their domains, along with their respective nomenclature, for the bivalent 1×1 antibody constructs described herein.

[0115] [Figure 4] FIG. 4 shows the configuration of an exemplary bivalent monospecific construct.

[0116] [Figure 5] Figure 5 shows data from a biolayer interferometry (BLI) experiment described in Example 1, in which bivalent, monospecific binding molecules with the configuration shown in Figure 4 [polypeptide 1: VL-CH3 (knob)-CH2-CH3 / polypeptide 2: VH-CH3 (hole)] were assayed. The antigen-binding site was specific for TNFα. BLI responses from binding molecule immobilization and TNFα binding to the immobilized construct indicate robust, specific, bivalent binding to the antigen. The data are consistent with the molecules having a high percentage of intended pairing between polypeptide 1 and polypeptide 2.

[0117] [Figure 6] FIG. 6 shows the characteristics of an exemplary bivalent 1×1 bispecific binding molecule, “BC1.”

[0118] [Figure 7]Figure 7A shows size-exclusion chromatography (SEC) analysis of "BC1," demonstrating that a single-step CH1 affinity purification step (CaptureSelect™ CH1 affinity resin) yields a single, monodisperse peak of >98% non-aggregated bivalent protein via gel filtration. Figure 7B shows comparative literature data from SEC analysis of CrossMab bivalent antibody constructs [data from Schaefer et al. (Proc Natl Acad Sci USA., 2011 Jul 5;108(27):11187-92)].

[0119] [Figure 8] Figure 8A shows the cation exchange chromatographic elution profile of "BC1" after one-step purification using CaptureSelect™ CH1 affinity resin, showing a single tight peak. Figure 8B shows the cation exchange chromatographic elution profile of "BC1" after purification using standard Protein A purification.

[0120] [Figure 9] FIG. 9 shows a non-reducing SDS-PAGE gel of "BC1" at various stages of purification.

[0121] [Figure 10] Figures 10A and 10B show SDS-PAGE gels of "BC1" after single-step CH1-affinity purification under both non-reducing and reducing conditions (Figure 10A) compared with SDS-PAGE gels of the CrossMab bispecific antibody under non-reducing and reducing conditions published in the reference (Figure 10B).

[0122] [Figure 11-1] Figures 11A and 11B show mass spectrometry analysis of "BC1" demonstrating two distinct heavy chains (Figure 11A) and two distinct light chains (Figure 11B) under reducing conditions. [Figure 11-2] Same as above.

[0123] [Figure 12] FIG. 12 shows mass spectrometry analysis of "BC1" purified under non-reducing conditions, confirming the absence of imperfect pairing after purification.

[0124] [Figure 13] FIG. 13 shows accelerated stability study data demonstrating the stability of "BC1" over 8 weeks at 40° C. compared to two IgG control antibodies.

[0125] [Figure 14] FIG. 14 shows the characteristics of an exemplary bivalent 1×1 bispecific binding molecule “BC6,” which is further described in Example 3.

[0126] [Figure 15] Figure 15A shows size-exclusion chromatography (SEC) analysis of "BC6" after one-step purification using CaptureSelect™ CH1 affinity resin, demonstrating that single-step CH1 affinity purification yields a single monodisperse peak and the absence of non-covalent aggregates. Figure 15B shows an SDS-PAGE gel of "BC6" under non-reducing conditions.

[0127] [Figure 16] FIG. 16 shows the characteristics of an exemplary bivalent, bispecific binding molecule, “BC28,” which is further described in Example 4.

[0128] [Figure 17] FIG. 17 shows SDS-PAGE analysis under non-reducing conditions after single-step CH1 affinity purification of "BC28," "BC29," "BC30," "BC31," and "BC32."

[0129] [Figure 18] FIG. 18 shows the SEC analysis of "BC28" and "BC30," respectively, after one-step purification using CaptureSelect™ CH1 affinity resin.

[0130] [Figure 19] FIG. 19 shows the characteristics of an exemplary bivalent, bispecific binding molecule, “BC44,” which is further described in Example 5.

[0131] [Figure 20] Figures 20A and 20B show size exclusion chromatography (SEC) data for two bivalent binding molecules, "BC15" and "BC16," respectively, under accelerated stability testing conditions. "BC15" and "BC16" have different variable region-CH3 junctions.

[0132] [Figure 21] FIG. 21 shows a schematic diagram of the five polypeptide chains and their domains, along with their respective naming conventions, for the trivalent 2x1 antibody construct described herein, where chain 5 is named "fifth polypeptide chain" in the schematic according to the naming convention.

[0133] [Figure 22] FIG. 22 shows the characteristics of an exemplary trivalent 2×1 bispecific binding molecule “BC1-2×1,” which is further described in Example 7.

[0134] [Figure 23] FIG. 23 shows a non-reducing SDS-PAGE of "BC1" and "BC1-2x1" proteins expressed using the ThermoFisher Expi293 transient transfection system at various stages of purification.

[0135] [Figure 24] Figure 24 compares the avidity of the bivalent 1x1 construct "BC1" with the avidity of the trivalent 2x1 construct "BC1-2x1" using Octet (Pall ForteBio) Biolayer Interference analysis.

[0136] [Figure 25] FIG. 25 shows the salient features of the trivalent 2×1 construct “TB111”.

[0137] [Figure 26] Figure 26 shows a schematic diagram of the five polypeptide chains and their domains, along with their respective naming conventions, for the trivalent 1x2 antibody construct described herein, where chain 5 is named "the sixth polypeptide chain" in the schematic according to the naming convention.

[0138] [Figure 27] FIG. 27 shows the characteristics of an exemplary trivalent 1×2 construct “CTLA4-4×Nivo×CTLA4-4,” which is further described in Example 10.

[0139] [Figure 28] FIG. 28 is an SDS-PAGE gel in which the lanes showing the trivalent 1×2 construct “CTLA4-4×Nivo×CTLA4-4” construct under non-reducing (“−DTT”) and reducing (“+DTT”) conditions are boxed.

[0140] [Figure 29] Figure 29 shows a comparison of antigen binding between two antibodies: the bivalent 1x1 construct "CTLA4-4xOX40-8" and the trivalent 1x2 construct "CTLA4-4xNivoxCTLA4-4." "CTLA4-4xOX40-8" binds monovalently to CTLA4, while "CTLA4-4xNivoxCTLA4-4" binds bivalently to CTLA4.

[0141] [Figure 30] FIG. 30 shows the characteristics of an exemplary trivalent 1×2 trispecific construct “BC28-1×1×1a,” which is further described in Example 11.

[0142] [Figure 31] FIG. 31 shows the size exclusion chromatography of "BC28-1x1x1a" after transient expression and single-step CH1 affinity resin purification, showing a single, clearly defined peak.

[0143] [Figure 32]FIG. 32 shows the results of SDS-PAGE under non-reducing and reducing conditions of bivalent and trivalent constructs after transient expression and one-step purification using CaptureSelect™ CH1 affinity resin, respectively, as further described in Example 12.

[0144] [Figure 33] Figures 33A-33C show Octet binding analysis to three antigens: PD1, antigen "A", and CTLA4. As further described in Example 13, Figure 33A shows binding of "BC1" to PD1 and antigen "A"; Figure 33B shows binding of the bivalent bispecific construct "CTLA4-4xOX40-8" to CTLA4, antigen "A" and PD1; and Figure 33C shows binding of the trivalent trispecific "BC28-1x1x1a" to PD1, antigen "A" and CTLA4.

[0145] [Figure 34] FIG. 34 shows a schematic diagram of the six polypeptide chains and their domains, along with their respective naming conventions, for the particular tetravalent 2×2 construct described herein.

[0146] [Figure 35] FIG. 35 illustrates certain salient features of an exemplary tetravalent 2×2 construct, “BC22-2×2,” which is further described in Example 14.

[0147] [Figure 36] Figure 36 is a non-reducing SDS-PAGE gel comparing the 2x2 tetravalent "BC22-2x2" construct with the 1x2 trivalent "BC12-1x2" and 2x1 trivalent "BC21-2x1" constructs at various stages of purification.

[0148] [Figure 37] FIG. 37 shows the configuration of an exemplary tetravalent 2×2 construct.

[0149] [Figure 38]Figure 38 shows a schematic diagram of the six polypeptide chains and their domains, along with their respective naming conventions, for a particular tetravalent construct described herein, where chain 5 is designated in the schematic as the "seventh polypeptide chain" and chain 6 is designated as the "eighth polypeptide chain" according to the naming conventions.

[0150] [Figure 39] FIG. 39 shows an exemplary configuration of a bispecific tetravalent construct.

[0151] [Figure 40] FIG. 40 shows an exemplary configuration for a trispecific tetravalent construct utilizing a common light chain strategy.

[0152] [Figure 41] Figure 41 shows bispecific antigen engagement by the tetravalent construct outlined in Figure 39, demonstrating that this construct was capable of simultaneous engagement. Biolayer Interference (BLI) responses from B-Body immobilization and TNFα binding to the immobilized construct are consistent with the molecule having a high percentage of intended chain pairing.

[0153] [Figure 42] Figure 42 shows flow cytometry analysis of B-Body binding to cell surface antigens. Shaded signals represent cells without antigen; dotted signals represent cells transiently transfected with surface antigen.

[0154] [Figure 43] FIG. 43 shows an exemplary configuration of a trivalent construct.

[0155] [Figure 44] FIG. 44 shows an exemplary configuration of a trivalent construct.

[0156] [Figure 45]Figure 45 shows the results of SDS-PAGE under non-reducing and reducing conditions of bivalent and trivalent constructs after transient expression and one-step purification using CaptureSelect™ CH1 affinity resin, respectively, as further described in Example 17.

[0157] [Figure 46] FIG. 46 shows the differences in thermal transitions of "BC24jv," "BC26jv," and "BC28jv," measured to assess the pairing stability of the junction variants.

[0158] [Figure 47] Figure 47 shows an Octet (Pall ForteBio) biolayer interference analysis of two-fold serial dilutions (200 to 12.5 nM) used to determine binding affinity to CD3 for the unmutagenized SP34-89 monovalent B-body.

[0159] [Figure 48-1] Figures 48A-48B show Octet (Pall ForteBio) biolayer interference analysis of two-fold serial dilutions (200-12.5 nM) used to determine the binding affinity to ROR1 for two ROR antigen binding site candidates (clone I2-A10 in Figure 48A; clone I2-A27 in Figure 48B). [Figure 48-2] Same as above.

[0160] [Figure 49] Figure 49 shows that the RORxCD3 bispecific 1x1 and 1x2 B-bodies resulted in activation of reporter T cells when mixed with a ROR1-expressing tumor line (HOP-92), but not when mixed with a tumor line that does not express ROR1 (B16).

[0161] [Figure 50]Figure 50 shows that the RORxCD3 bispecific I2A-10 1x2 B-body resulted in cytotoxic T cell-mediated killing when mixed with a ROR1-expressing tumor line (MDA-MD-231), but did not result in cytotoxicity when a CD3 bispecific B-body with irrelevant tumor ABS (e.g., a tumor antigen not expressed on MDA-MD-231) was added to the mixture.

[0162] [Figure 51-1] Figures 51A-51E show that the RORxCD3 bispecific I2A-3 1x1 and 1x2 B-bodies led to reporter T cell activation when mixed with the ROR1-expressing tumor lines HOP-92 (Figure 51A), A549 (Figure 51B), MDA-MD-231 (Figure 51C), JeKo-1 (Figure 51D), and RPMI-8226 (Figure 51E), but not when mixed with a tumor line that does not express ROR1 (B16). [Figure 51-2] Same as above. [Figure 51-3] Same as above. [Figure 51-4] Same as above. [Figure 51-5] Same as above.

[0163] [Figure 52] Figure 52 shows that RORxCD3 bispecific 1x2 B-bodies I2A-1, I2A-3, I2A-10, I2A-14, I2A-22 and I2A-27 conferred cytotoxic T cell-mediated killing when mixed with a ROR1-expressing tumor line (MDA-MD-231), whereas 1x2 B-bodies I2A-16 and I2A-22 did not confer robust cytotoxicity.

[0164] [Figure 53-1] Figure 53A shows published ROR1 expression data for MDA-MD-231 and RPMI-8226 tumor lines. Figures 53B and 53C show that cytotoxic effects correlate with ROR1 in MDA-MD-231 and RPMI-8226 tumor cell lines. [Figure 53-2]Same as above.

[0165] [Figure 54-1] Figures 54A-54F show that I2A-10 and I2A-27 B-bodies activated CD8+ T cells in the PBMC population as determined by quantifying CD25 (Figure 54A), CD69 (Figure 54C), and both CD25 and CD69 (Figure 54E), and activated CD4+ T cells in the PBMC population as determined by quantifying CD25 (Figure 54B), CD69 (Figure 54D), and both CD25 and CD69 (Figure 54F). [Figure 54-2] Same as above. [Figure 54-3] Same as above.

[0166] [Figure 55] FIG. 55 shows that 26% and 36% of candidates I2A-10 (top panel) and I2A-27 (bottom panel), respectively, were internalized after 2 hours of incubation with MDA-MB-231 cells.

[0167] [Figure 56] Figure 56 shows size exclusion chromatography (SEC) analysis demonstrating that a single CH1 affinity purification step yields a single monodisperse peak via gel filtration, with greater than 98% non-aggregated protein for 1x2 B-body candidates I2A-10 (top panel) and I2A-27 (bottom panel).

[0168] [Figure 57] Figure 57A shows a non-reducing SDS-PAGE gel of 1x2 B-body candidates I2A-10 (left panel) and I2A-27 (right panel), demonstrating the major band of the fully assembled construct (high mobility 250 kDa band).

[0169] FIG. 57B shows Bioanalyzer (Agilent) analysis of non-reduced samples of 1×2 B-body candidates I2A-10 and I2A-27 demonstrating the major bands of the fully assembled constructs.

[0170] [Figure 58] Figure 58 shows SDS-PAGE analysis of a bispecific antibody containing standard knobs-and-holes orthogonal mutations introduced into the CH3 domain, confirmed in their native locations within the Fc portion of the bispecific antibody purified using a single-step CH1 affinity purification step (CaptureSelect™ CH1 affinity resin).

[0171] [Figure 59] Figures 59A-59B show Octet (Pall ForteBio) biolayer interference analysis demonstrating FcγRIa binding to trastuzumab (Figure 59A "WT IgG1") and no binding to sFc10 (Figure 59B).

[0172] [Figure 60] Figure 60 shows killing by trastuzumab (Herceptin, "WT-IgG1"), but not by sFc7 or sFc10, in an ADCC assay.

[0173] [Figure 61] Figure 61 shows C1q binding by trastuzumab (Herceptin, "WT-IgG1"), but not by sFc1, sFc7 or sFc10, in a C1q ELISA.

[0174] [Figure 62] Figures 62A-62C show tumor volume monitored for mice implanted with tumor cells, humanized with PBMCs (left solid arrow), and subsequently treated IV (right dashed arrow) with PBS (Figure 62A), 1x2 B-body candidate I2-A10 (Figure 62B), or 1x2 B-body candidate I2-A27 (Figure 62C).

[0175] [Figure 63] Figure 63 shows the tumor volume at the end of the study for each mouse, with the mean and standard deviation for each group shown. The white square in the I2-A27 group was excluded from the analysis due to possible non-humanization by PMBC.

[0176] [Figure 64] Figure 64A shows binding of SP34-89 to Jurkat cells and cynomolgus T cells. Figure 64B shows binding of I2-A27 1x2 B-body™ bispecific antibody to cynomolgus CD3 delta and epsilon heterodimer.

[0177] [Figure 65-1] Figure 65A shows binding of the I2-A27 IgG antibody to ROR1 in a monovalent binding assay. Figure 65B shows minimal binding of the I2-A27 IgG antibody to ROR2. Figure 65C shows binding of the I2-A27 1x2 B-body™ to ROR1 in a monovalent binding assay. Figure 65D shows minimal binding of the I2-A27 1x2 B-body™ to ROR2. Figure 65E shows binding of the I2-A27 1x2 B-body™ to ROR1 in a bivalent binding assay. [Figure 65-2] Same as above. [Figure 65-3] Same as above.

[0178] [Figure 66-1] Figure 66A shows SEC analysis of I2-A27 1x2 B-body™, Figure 66B shows SMAC analysis of I2-A27 1x2 B-body™, and Figure 66C shows HIC analysis of I2-A27 1x2 B-body™. [Figure 66-2] Same as above.

[0179] [Figure 67-1]Figure 67A shows the activity of I2-A27 1x2 B-body™ in a Jurkat assay using MDA-MB-231 (ROR1-expressing) cells. Figure 67B shows the activity of I2-A27 1x2 B-body™ in a Jurkat assay using RPMI-8226 (ROR1- and ROR2-expressing) cells. Figure 67C shows the inactivity of I2-A27 1x2 B-body™ in a Jurkat assay using K562 (ROR2-expressing) cells. Figure 67D shows the inactivity of I2-A27 1x2 B-body™ in a Jurkat assay in the absence of target cell line. [Figure 67-2] Same as above.

[0180] [Figure 68-1] Figure 68A shows CD69 expression by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 68B shows CD69 expression by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing RPMI-8226 cells. Figure 68C shows CD25 expression by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 68D shows CD25 expression by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing RPMI-8226 cells. [Figure 68-2] Same as above.

[0181] [Figure 69] Figure 69A shows LDH release by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 69B shows LDH release by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing RPMI-8226 cells.

[0182] [Figure 70-1]Figure 70A shows granzyme B secretion by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 70B shows granzyme B secretion by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing RPMI-8226 cells. Figure 70C shows TNFα secretion by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 70D shows TNFα secretion by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing RPMI-8226 cells. Figure 70E shows IFNγ release by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 70F shows IFNγ release by various concentrations of I2-A27 1x2 B-body™ and ROR1-expressing RPMI-8226 cells. [Figure 70-2] Same as above. [Figure 70-3] Same as above.

[0183] [Figure 71] Figure 71A shows the activity of I2-A27 1x2 B-body™ samples stored in human serum for 1 week at 4°C or 37°C. Figure 71B shows the inactivity of the samples in a Jurkat assay in the absence of ROR1-expressing cells.

[0184] [Figure 72] Figure 72 shows a stability assay of I2-A27 1x2 B-body™ samples under accelerated conditions.

[0185] [Figure 73] Figure 73 shows a stability assay of I2-A27 1x2 B-body™ samples under real-time conditions.

[0186] [Figure 74] Figure 74 shows an acid stability assay of I2-A27 1x2 B-body™.

[0187] [Figure 75] Figure 75 shows binding of I2-A10 D54E Y55Q 1x2 B-body™ to cynomolgus monkey CD3 delta and epsilon heterodimer.

[0188] [Figure 76-1] Figure 76A shows the binding of the I2-A10 D54E Y55Q IgG antibody to ROR1 in a monovalent binding assay. Figure 76B shows the binding of I2-A10 D54E Y55Q IgG to ROR2 in a monovalent binding assay. Figure 76C shows the binding of the I2-A10 D54E Y55Q 1x2 B-body™ to ROR1 in a monovalent binding assay. Figure 76D shows the binding of the I2-A10 D54E Y55Q 1x2 B-body™ to ROR2 in a monovalent binding assay. Figure 76E shows the binding of the I2-A10 D54E Y55Q 1x2 B-body™ to ROR1 in a bivalent binding assay. Figure 76F shows the binding of the I2-A10 D54E Y55Q 1x2 B-body™ to ROR2 in a bivalent binding assay. [Figure 76-2] Same as above. [Figure 76-3] Same as above.

[0189] [Figure 77] Figure 77 shows binding of I2-A10 D54E Y55Q 1x2 B-body™ to the Ig-like domain of ROR1.

[0190] [Figure 78-1] Figure 78A shows SEC analysis of I2-A10 D54E Y55Q 1x2 B-body™, Figure 78B shows SMAC analysis of I2-A10 D54E Y55Q 1x2 B-body™, and Figure 78C shows HIC analysis of I2-A10 D54E Y55Q 1x2 B-body™. [Figure 78-2] Same as above.

[0191] [Figure 79-1]Figure 79A shows the activity of the I2-A10 D54E Y55Q 1x2 B-body™ in a Jurkat assay using MDA-MB-231 (ROR1-expressing) cells. Figure 79B shows the activity of the I2-A10 D54E Y55Q 1x2 B-body™ in a Jurkat assay using RPMI-8226 (ROR1- and ROR2-expressing) cells. Figure 79C shows the activity of the I2-A10 D54E Y55Q 1x2 B-body™ in K562 (ROR2-expressing) cells. Figure 79D shows the inactivity of the I2-A10 D54E Y55Q 1x2 B-body™ in a Jurkat assay in the absence of the target cell line. [Figure 79-2] Same as above.

[0192] [Figure 80-1] Figure 80A shows CD69 expression by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing MDA-MB-231 cells. Figure 80B shows CD69 expression by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing RPMI-8226 cells. Figure 80C shows CD25 expression by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing MDA-MB-231 cells. Figure 80D shows CD25 expression by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing RPMI-8226 cells. [Figure 80-2] Same as above.

[0193] [Figure 81] Figure 81A shows LDH release by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 81B shows LDH release by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing RPMI-8226 cells.

[0194] [Figure 82-1] Figure 82A shows granzyme B secretion by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing MDA-MB-231 cells. Figure 82B shows granzyme B secretion by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing RPMI-8226 cells. Figure 82C shows TNFα secretion by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing MDA-MB-231 cells. Figure 82D shows TNFα secretion by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing RPMI-8226 cells. Figure 82E shows IFNγ secretion by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1-expressing MDA-MB-231 cells. Figure 82F shows IFNγ secretion by various concentrations of I2-A10 D54E Y55Q 1x2 B-body™ and ROR1- and ROR2-expressing RPMI-8226 cells. [Figure 82-2] Same as above. [Figure 82-3] Same as above.

[0195] [Figure 83] Figure 83A shows the activity of I2-A10 D54E Y55Q 1x2 B-body™ samples stored in human serum for 1 week at 4°C or 37°C. Figure 83B shows the inactivity of the samples in a Jurkat assay in the absence of ROR1-expressing cells.

[0196] [Figure 84] Figure 84 shows the stability assay of I2-A10 D54E Y55Q 1x2 B-body™ samples under accelerated conditions.

[0197] [Figure 85]Figure 85 shows a stability assay of I2-A10 D54E Y55Q 1x2 B-body™ samples under real-time conditions.

[0198] [Figure 86] Figure 86 shows the acid stability assay of I2-A10 D54E Y55Q 1x2 B-body™.

[0199] [Figure 87] Figure 87 shows the in vivo efficacy of the I2-27 1x2 B-body™ and I2-A10 D54E Y55Q 1x2 B-body™ in reducing tumor volume (mm3) in a mouse model of cancer.

[0200] [Figure 88] Figure 88 shows the in vivo efficacy of multiple administrations of I2-27 1x2 B-body™ and I2-A10 D54E Y55Q 1x2 B-body™ in reducing tumor volume (mm3) in a mouse model of cancer. DETAILED DESCRIPTION OF THE INVENTION

[0201] The drawings depict various embodiments of the present invention for purposes of illustration only. Those skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be utilized without departing from the principles of the present invention as described herein. 6. Detailed Description 6.1.Definition

[0202] Unless otherwise defined, all technical and scientific terms used herein have the meaning commonly understood by one of ordinary skill in the art to which this invention belongs. As used herein, the following terms have the meanings given below.

[0203] "Antigen-binding site" means the region of a ROR-binding molecule that specifically recognizes or binds to a given antigen or epitope.

[0204] "B-body," as used herein and with reference to Figure 3, refers to a binding molecule comprising first and second polypeptide chains, wherein: (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged, N-terminus to C-terminus, in an A-B-D-E orientation, and Domain A has a VL amino acid sequence, Domain B has a CH3 amino acid sequence, Domain D has a CH2 amino acid sequence, and Domain E has a constant region domain amino acid sequence; (b) the second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged, N-terminus to C-terminus, in an F-G orientation, and Domain F has a VH amino acid sequence, and Domain G has a CH3 amino acid sequence; and (c) the first and second polypeptides associate via interactions between Domain A and Domain F and between Domain B and Domain G to form the binding molecule. B-bodies are described in more detail in International Patent Application No. PCT / US2017 / 057268, the entire contents of which are incorporated herein by reference.

[0205] As used herein, the term "treat" or "treatment" refers to both therapeutic treatment and prophylactic or preventative measures, where the objective is to prevent or slow (alleviate) undesirable physiological changes or disorders, such as the progression of multiple sclerosis, arthritis, or cancer. Beneficial or desired clinical results include, but are not limited to, alleviation of symptoms, reduction in the extent of disease, stable (i.e., not worsening) state of disease, delay or slowing of disease progression, amelioration or palliation of the disease state, and remission (partial or complete remission), whether detectable or undetectable. "Treatment" can also mean prolonging survival as compared to expected survival if not receiving treatment. Subjects in need of treatment include those already suffering from a condition or disorder, as well as those prone to have the condition or disorder, or those in whom the condition or disorder is to be prevented.

[0206] "Subject," "individual," "animal," "patient," or "mammal" means any subject for whom diagnosis, prevention, or treatment is desired, particularly a mammalian subject. Mammalian subjects include humans, domestic animals, farm animals, and zoo, sport, or pet animals, such as dogs, cats, guinea pigs, rabbits, rats, mice, horses, cattle, dairy cows, etc.

[0207] The term "sufficient amount" means an amount sufficient to produce a desired effect, for example, an amount sufficient to modulate protein aggregation in a cell.

[0208] A "therapeutically effective amount" is an amount effective for ameliorating symptoms of disease. A therapeutically effective amount can be a "prophylactically effective amount," so that prevention can be considered treatment. 6.2. Other Rules of Interpretation

[0209] Unless otherwise specified, all references to sequences herein are to amino acid sequences.

[0210] Unless otherwise specified, antibody constant region residue numbering is according to the Eu index as set forth in www.imgt.org / IMGTScientificChart / Numbering / Hu_IGHGnber.html#refs (accessed August 22, 2017) and Edelman et al., Proc. Natl. Acad. USA, 63:78-85 (1969), which are incorporated by reference in their entireties, and identifies residues according to their position in the endogenous constant region sequence, regardless of the residue's physical location within the chain of a ROR-binding molecule described herein. An "endogenous sequence" or "native sequence" refers to any sequence, including both nucleic acid and amino acid sequences, derived from an organism, tissue, or cell and which has not been artificially modified or mutated.

[0211] Polypeptide chain numbers (e.g., "first" polypeptide chain, "second" polypeptide chain, etc., or polypeptide "chain 1," "chain 2," etc.) are used herein as unique identifiers for the particular polypeptide chains that form the binding molecule and are not intended to imply any order or content of different polypeptide chains within the binding molecule.

[0212] In this disclosure, the words "comprises," "comprising," "containing," "having," "includes," "including," and linguistic variations thereof have the meaning ascribed to them in U.S. patent law and permit the presence of additional components other than those expressly recited.

[0213] Ranges provided herein are understood to be shorthand for all values ​​within the range, including the recited endpoints. For example, a range of 1 to 50 is understood to include any number, combination of numbers, or subrange from the group consisting of 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, and 50.

[0214] As used herein, the term "or" is understood to be inclusive unless otherwise stated or clear from context. Unless otherwise stated or clear from context, the terms "a," "an," and "the" are understood to be singular or plural.

[0215] Unless otherwise specified or clear from the context, the term "about" as used herein is understood to mean within a range normally accepted in the art, for example, within two standard deviations of the mean. About can be understood to mean within 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.05%, or 0.01% of the indicated value. Unless otherwise clear from the context, all numerical values ​​provided herein are modified by the term about. 6.3.ROR antigen binding molecule

[0216] In a first aspect, an antigen-binding molecule is provided. In all embodiments, the antigen-binding molecule comprises at least a first antigen-binding site specific for the ROR antigen; thus, the binding molecule is designated as a ROR antigen-binding molecule.

[0217] The ROR binding molecules described herein specifically bind to the ROR antigen.

[0218] As used herein, "ROR antigen" refers to a member of the tyrosine protein kinase transmembrane receptor (ROR) family, including members ROR1 and ROR2. In certain embodiments, an ROR-binding molecule has an antigen-binding site that specifically binds only to ROR1. In other embodiments, an ROR-binding molecule has an antigen-binding site that specifically binds only to ROR2. In yet other embodiments, an ROR-binding molecule has an antigen-binding site that is cross-reactive and specifically binds to both ROR1 and ROR2.

[0219] ROR1 and ROR2 proteins typically consist of at least four protein domains: three extracellular domains (Ig-like, FZ, and Kringle domains) and an intracellular protein kinase domain. In some embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to the extracellular portion of the ROR antigen. In certain embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to the Ig-like domain. In other embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to the FZ domain. In still other embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to the Kringle domain. In certain embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to at least a portion of a single ROR domain. In certain embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to at least a portion of two or more ROR domains, such as the junction between the first and second ROR domains. The ROR domain can refer to the ROR1 domain or the ROR2 domain.

[0220] In certain embodiments, the ROR antigen is human. UniProt Accession #Q01973 represents the canonical human ROR1 protein, including its sequence and domain features, and is hereby incorporated by reference in its entirety. SEQ ID NO:94 provides the full-length ROR1 protein sequence. Referring to the full-length sequence from N-terminus to C-terminus, the Ig-like domain is defined as amino acids 42-147, the FZ domain as amino acids 165-299, and the Kringle domain as amino acids 312-391. UniProt Accession #Q01974 represents the canonical human ROR2 protein, including its sequence and domain features, and is hereby incorporated by reference in its entirety. SEQ ID NO:95 provides the full-length ROR2 protein sequence. Referring to the full-length sequence from N-terminus to C-terminus, the Ig-like domain is defined as amino acids 55-145, the FZ domain as amino acids 169-303, and the Kringle domain as amino acids 316-394.

[0221] Gentile, et al. (Cancer Res; 71(8) April 15, 2011), Rebagay, et al. (Front. Oncol., 18 April 2012), Zhang, et al. (American Journal of Pathology, Vol. 181, No. 6, December 2012), Henry, et al. (Oncotarget, Vol. 6, No. As described in further detail in Zhang et al. (PLoS ONE 7(3): e31127), and Bainbridge et al. (PLoS ONE 9(7): e102695), various tumors can demonstrate cell surface expression of ROR antigens. In addition, as described in Balakrishnan et al. (Clin Cancer Res. 2017 Jun 15; 23(12): 3061-3071), which is incorporated herein in its entirety, ROR expression can be absent or demonstrate only limited expression in normal, i.e., non-cancerous, tissues. Thus, ROR antigens can be used as tumor-specific markers in certain tumors. Examples of tumors and cancers with demonstrated ROR expression include, but are not limited to, pancreatic cancer, ovarian cancer, breast cancer, lung cancer, gastric cancer, melanoma, Ewing's sarcoma, chronic lymphocytic leukemia, mantle cell lymphoma, and B-ALL, as described in Gohil et al. (Oncoimmunology. 2017; 6(7): e1326437.), which is incorporated herein in its entirety. Other cancers include, but are not limited to, hematological cancers, prostate cancer, colon cancer, renal cancer, and uterine cancer.

[0222] In various embodiments, the ROR-binding molecule additionally specifically binds to at least one antigen in addition to the ROR antigen.

[0223] In certain embodiments, the ROR-binding molecule is a bispecific, bivalent molecule. In another embodiment, the ROR-binding molecule is a bispecific, trivalent molecule. In certain embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to the ROR antigen and a molecule expressed on the surface of T cells. In certain embodiments, the ROR-binding molecule has an antigen-binding site that specifically binds to the ROR antigen and the protein CD3 expressed on the surface of T cells. Without wishing to be bound by theory, ROR-binding molecules that specifically bind to the ROR antigen and a molecule expressed on the surface of T cells (i.e., CD3) can redirect T cells to ROR-expressing cells (i.e., target cells) and thereby direct T cell-mediated killing (cytotoxicity) of cells expressing the ROR antigen. T cell-mediated killing using bispecific anti-CD3 molecules is described in detail in U.S. Patent Application Publication No. 2006 / 0193852, the entire contents of which are incorporated herein by reference. In some embodiments, the molecule expressed on the surface of T cells is selected from any molecule that can redirect T cells to target cells.

[0224] Referring to Figure 3, in one set of embodiments, a ROR-binding molecule comprises first and second polypeptide chains: (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged from N-terminus to C-terminus in an A-B-D-E orientation, and Domain A has a variable region domain amino acid sequence, and Domain B, Domain D, and Domain E have constant region domain amino acid sequences; (b) the second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged from N-terminus to C-terminus in an F-G orientation, and Domain F has a variable region domain amino acid sequence, and Domain G has a constant region domain amino acid sequence; (c) the third polypeptide chain comprises Domain H, Domain I, Domain J, and Domain K, wherein the domains are arranged from N-terminus to C-terminus in an H-I-J-K orientation, and Domain H has a variable region domain amino acid sequence, and Domain G has a constant region domain amino acid sequence; (d) the fourth polypeptide chain comprises domain L and domain M, wherein the domains are arranged N-terminally to C-terminally in an LM orientation, domain L having a variable region domain amino acid sequence, the fourth polypeptide chain comprising a CH1 domain, and domain M being a CH1 domain or a portion thereof; (e) the first and second polypeptides associate via interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate via interactions between the H and L domains and between the I and M domains; and (g) the first and third polypeptides associate via interactions between the D and J domains and between the E and K domains to form a ROR-binding molecule.

[0225] In one series of embodiments, (a) a first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged from N-terminus to C-terminus in an A-B-D-E orientation, and Domain A has a variable region domain amino acid sequence, and Domain B, Domain D, and Domain E have a constant region domain amino acid sequence; (b) a second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged from N-terminus to C-terminus in an F-G orientation, and Domain F has a variable region domain amino acid sequence, and Domain G has a constant region domain amino acid sequence; (c) a third polypeptide chain comprises Domain H, Domain I, Domain J, and Domain K, wherein the domains are arranged from N-terminus to C-terminus in an H-I-J-K orientation, and the third polypeptide chain comprises a CH1 domain, and Domain I is a CH1 domain or portion thereof. (d) a fourth polypeptide chain comprises domain L and domain M, wherein the domains are arranged N-terminally to C-terminally in an LM orientation, with domain L having a variable region domain amino acid sequence and domain M having a CL amino acid sequence; (e) the first and second polypeptides associate via interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate via interactions between the H and L domains and between the I and M domains; and (g) the first and third polypeptides associate via interactions between the D and J domains and between the E and K domains to form a ROR-binding molecule. 6.3.1. Domain A (variable region)

[0226] In the ROR-binding molecule, Domain A has a variable region domain amino acid sequence. The variable region domain amino acid sequences described herein are antibody variable region domain amino acid sequences, including VL and VH antibody domain sequences. VL and VH sequences are described in further detail below in Sections 6.3.1.1 and 6.3.1.4, respectively. In a preferred embodiment, Domain A has a VL antibody domain sequence and Domain F has a VH antibody domain sequence. 6.3.1.1.VL area

[0227] The VL amino acid sequences useful in the ROR-binding molecules described herein are antibody light chain variable domain sequences. In typical sequences in both the natural antibodies and antibody constructs described herein, a particular VL amino acid sequence associates with a particular VH amino acid sequence to form an antigen-binding site. In various embodiments, the VL amino acid sequence is a mammalian sequence, including a human sequence, a synthetic sequence, or a combination of a human sequence, a non-human mammalian sequence, a mammalian sequence, and / or a synthetic sequence, as described in more detail in Sections 6.3.1.2 and 6.3.1.3 below.

[0228] In various embodiments, the VL amino acid sequence is a variant of a naturally occurring sequence. In certain embodiments, the VL amino acid sequence is a lambda (λ) light chain variable domain sequence. In certain embodiments, the VL amino acid sequence is a kappa (κ) light chain variable domain sequence. In preferred embodiments, the VL amino acid sequence is a kappa (κ) light chain variable domain sequence.

[0229] In the ROR-binding molecules described herein, the C-terminus of Domain A is linked to the N-terminus of Domain B. In certain embodiments, Domain A has a VL amino acid sequence that is mutated at its C-terminus at the junction between Domain A and Domain B, as described in further detail in Section 6.3.19.1 below and Example 6. Complementarity-Determining Regions

[0230] The VL amino acid sequence contains highly variable sequences called "complementarity-determining regions" (CDRs), typically three CDRs (CDR1, CDR2, and CDR3). In various embodiments, the CDRs are mammalian sequences, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human sequences. In preferred embodiments, the CDRs are human sequences. In various embodiments, the CDRs are naturally occurring sequences. In various embodiments, the CDRs are naturally occurring sequences that have been mutated to alter the binding affinity of the antigen-binding site for a particular antigen or epitope. In certain embodiments, naturally occurring CDRs have been mutated in vivo in a host through affinity maturation and somatic hypermutation. In certain embodiments, the CDRs have been mutated in vitro through methods including, but not limited to, PCR mutagenesis and chemical mutagenesis. In various embodiments, the CDRs are synthetic sequences, including, but not limited to, CDRs obtained from random sequence CDR libraries and rationally designed CDR libraries. 6.3.1.3. Framework Regions and CDR Grafting

[0231] The VL amino acid sequence comprises "framework region" (FR) sequences. FRs are generally conserved sequence regions (see Section 6.3.1.2.) that act as scaffolds for the interspersed CDRs, typically in the following configuration (N- to C-terminus): FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4. In various embodiments, the FRs are mammalian sequences, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human sequences. In preferred embodiments, the FRs are human sequences. In various embodiments, the FRs are naturally occurring sequences. In various embodiments, the FRs are synthetic sequences, including, but not limited to, rationally designed sequences.

[0232] In various embodiments, both the FRs and CDRs are derived from the same naturally occurring variable domain sequence. In various embodiments, the FRs and CDRs are derived from different variable domain sequences, where the CDRs are grafted onto the FR scaffold, providing specificity for a particular antigen. In certain embodiments, all of the grafted CDRs are derived from the same naturally occurring variable domain sequence. In certain embodiments, the grafted CDRs are derived from different variable domain sequences. In certain embodiments, the grafted CDRs are synthetic sequences, including, but not limited to, CDRs obtained from random sequence CDR libraries and rationally designed CDR libraries. In certain embodiments, the grafted CDRs and FRs are derived from the same species. In certain embodiments, the grafted CDRs and FRs are derived from different species. In preferred grafted CDR embodiments, the antibody is "humanized," where the grafted CDRs are non-human mammalian sequences, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, and goat sequences, and the FRs are human sequences. Humanized antibodies are discussed in more detail in U.S. Patent No. 6,407,213, which is incorporated herein by reference in its entirety for all that it teaches. In various embodiments, portions or specific sequences of FRs from one species are used to replace portions or specific sequences of FRs from another species. 6.3.1.4.VH area

[0233] The VH amino acid sequence in the ROR-binding molecules described herein is an antibody heavy chain variable domain sequence. In typical antibody configurations, both naturally occurring and the ROR-binding molecules described herein, a particular VH amino acid sequence associates with a particular VL amino acid sequence to form an antigen-binding site. In various embodiments, as described in more detail above in Sections 6.3.1.2 and 6.3.1.3, the VH amino acid sequence is a mammalian sequence, including a human sequence, a synthetic sequence, or a combination of non-human mammalian, mammalian, and / or synthetic sequences. In various embodiments, the VH amino acid sequence is a mutated sequence of a naturally occurring sequence. 6.3.2. Domain B (Constant Region)

[0234] In ROR-binding molecules, Domain B has a constant region domain sequence. The constant region domain amino acid sequences described herein are sequences of antibody constant region domains.

[0235] In various embodiments, the constant region sequence is a mammalian sequence, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human sequences. In preferred embodiments, the constant region sequence is a human sequence. In certain embodiments, the constant region sequence is derived from an antibody light chain. In certain embodiments, the constant region sequence is derived from a lambda or kappa light chain. In certain embodiments, the constant region sequence is derived from an antibody heavy chain. In certain embodiments, the constant region sequence is an antibody heavy chain sequence of an IgA1, IgA2, IgD, IgE, IgG1, IgG2, IgG3, IgG4, or IgM isotype. In certain embodiments, the constant region sequence is derived from an IgG isotype. In preferred embodiments, the constant region sequence is derived from an IgG1 isotype. In certain preferred embodiments, the constant region sequence is a CH3 sequence. CH3 sequences are described in further detail below in Section 6.3.2.1. In other preferred embodiments, the constant region sequence is an orthologous CH2 sequence. Orthologous CH2 sequences are described in further detail below in Section 6.3.2.2.

[0236] In certain embodiments, the constant region sequence has been mutated to include one or more orthogonal mutations. In preferred embodiments, Domain B has a constant region sequence that is a knob-hole (synonymously "knob-in-hole," "KIH") orthogonal mutation, as described in further detail below in Section 6.3.14.2, and a CH3 sequence containing either a S354C or Y349C mutation that forms an engineered disulfide bridge with the CH3 domain containing the orthogonal mutation, as described in further detail below in Section 6.3.14.1. In some preferred embodiments, the knob-hole orthogonal mutation is a T366W mutation. 6.3.2.1.CH3 area

[0237] The CH3 amino acid sequence is the sequence of the C-terminal domain of an antibody heavy chain, as described herein.

[0238] In various embodiments, the CH3 sequence is a mammalian sequence, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human sequences. In preferred embodiments, the CH3 sequence is a human sequence. In certain embodiments, the CH3 sequence is derived from an IgA1, IgA2, IgD, IgE, IgM, IgG1, IgG2, IgG3, or IgG4 isotype, or is a CH4 sequence derived from an IgE or IgM isotype. In certain embodiments, the CH3 sequence is derived from an IgG isotype. In preferred embodiments, the CH3 sequence is derived from an IgG1 isotype.

[0239] In certain embodiments, the CH3 sequence is an endogenous sequence. In certain embodiments, the CH3 sequence is amino acids 224-330 of UniProt Accession No. P01857. In various embodiments, the CH3 sequence is a segment of an endogenous CH3 sequence. In certain embodiments, the CH3 sequence has an endogenous CH3 sequence lacking the N-terminal amino acids G224 and Q225. In certain embodiments, the CH3 sequence has an endogenous CH3 sequence lacking the C-terminal amino acids P328, G329, and K330. In certain embodiments, the CH3 sequence has an endogenous CH3 sequence lacking the N-terminal amino acids G224 and Q225 and the C-terminal amino acids P328, G329, and K330. In preferred embodiments, the ROR-binding molecule has multiple domains with CH3 sequences, where the CH3 sequence can refer to both the full-length endogenous CH3 sequence and a CH3 sequence lacking the N-terminal amino acids, the C-terminal amino acids, or both.

[0240] In certain embodiments, the CH3 sequence is an endogenous sequence with one or more mutations, in certain embodiments, the mutations are one or more orthogonal mutations introduced into the endogenous CH3 sequence to guide the specific pairing of a particular CH3 sequence, as described in more detail below in Sections 6.3.14.1-6.3.14.3.

[0241] In certain embodiments, the CH3 sequence is engineered to reduce the immunogenicity of the antibody by replacing specific amino acids of one allotype with amino acids of another allotype, referred to herein as isoallotypic mutations, as described in more detail in Stickler et al. (Genes Immun. 2011 Apr; 12(3): 213-221), the entire teachings of which are incorporated herein by reference. In certain embodiments, specific amino acids of the G1m1 allotype are replaced. In a preferred embodiment, the isoallotypic mutations D356E and L358M are made in the CH3 sequence.

[0242] In a preferred embodiment, domain B has a human IgG1 CH3 amino acid sequence with the following mutational changes: P343V; Y349C; and the tripeptide insertion 445P, 446G, 447K. In another preferred embodiment, domain B has a human IgG1 CH3 sequence with the following mutational changes: T366K; and the tripeptide insertion 445K, 446S, 447C. In yet another preferred embodiment, domain B has a human IgG1 CH3 sequence with the following mutational changes: Y349C and the tripeptide insertion 445P, 446G, 447K. It has a CH3 sequence.

[0243] In certain embodiments, domain B has a human IgG1 CH3 sequence that differs from it and has a 447C mutation incorporated into the endogenous CH3 sequence.

[0244] In the ROR-binding molecules described herein, the N-terminus of Domain B is linked to the C-terminus of Domain A. In certain embodiments, Domain B has a CH3 amino acid sequence that is mutated at its N-terminus at the junction between Domain A and Domain B, as described in further detail in Section 6.3.19.1 below and Example 6.

[0245] In ROR-binding molecules, the C-terminus of Domain B is linked to the N-terminus of Domain D. In certain embodiments, Domain B has a CH3 amino acid sequence that is extended C-terminally to the junction between Domain B and Domain D, as described in further detail below in Section 6.3.19.3. 6.3.2.2. Orthologous CH2 Regions

[0246] The CH2 amino acid sequences described herein are the sequences of the third domain of an antibody heavy chain, when viewed from N-terminus to C-terminus. CH2 amino acid sequences are generally described in more detail below in Section 6.3.3. In one set of embodiments, the ROR-binding molecule has two or more paired sets of CH2 domains having CH2 sequences, with the first set having a CH2 amino acid sequence from a first isotype and one or more orthologous sets of CH2 amino acid sequences from another isotype. The orthologous CH2 amino acid sequences described herein can interact with the CH2 amino acid sequence from the shared isotype but do not significantly interact with CH2 amino acid sequences from another isotype present in the ROR-binding molecule. In certain embodiments, all of the set of CH2 amino acid sequences are derived from the same species. In preferred embodiments, all of the set of CH2 amino acid sequences are human CH2 amino acid sequences. In other embodiments, the set of CH2 amino acid sequences is derived from a different species. In certain embodiments, the first set of CH2 amino acid sequences is derived from the same isotype as the other non-CH2 domains in the ROR-binding molecule. In certain embodiments, the first set has a CH2 amino acid sequence derived from an IgG isotype, and one or more orthologous sets have CH2 amino acid sequences derived from an IgM or IgE isotype. In certain embodiments, one or more of the set of CH2 amino acid sequences are endogenous CH2 sequences. In other embodiments, one or more of the set of CH2 amino acid sequences are endogenous CH2 sequences with one or more mutations. In certain embodiments, the one or more mutations are orthogonal knob-and-hole mutations, orthogonal charge-pair mutations, or orthogonal hydrophobic mutations. Orthologous CH2 amino acid sequences useful for ROR-binding molecules are described in more detail in International PCT Applications WO2017 / 011342 and WO2017 / 106462, which are incorporated by reference in their entireties. 6.3.3. Domain D (Constant Region)

[0247] In the ROR-binding molecules described herein, Domain D comprises a constant region amino acid sequence, which is described in more detail in Section 6.3.2.

[0248] In a preferred set of embodiments, domain D has a CH2 amino acid sequence. The CH2 amino acid sequence is the CH2 amino acid sequence of the third domain of a native antibody heavy chain, as described herein, from N-terminus to C-terminus. In various embodiments, the CH2 sequence is a mammalian sequence, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human sequences. In preferred embodiments, the CH2 sequence is a human sequence. In certain embodiments, the CH2 sequence is derived from an IgA1, IgA2, IgD, IgE, IgG1, IgG2, IgG3, IgG4, or IgM isotype. In preferred embodiments, the CH2 sequence is derived from an IgG1 isotype.

[0249] In certain embodiments, the CH2 sequence is an endogenous sequence. In particular embodiments, the sequence is amino acids 111-223 of UniProt Accession No. P01857. In preferred embodiments, the CH2 sequence has an N-terminal hinge region peptide connecting the N-terminal variable domain-constant domain segment to the CH2 domain, as discussed in more detail below in Section 6.3.19.3.

[0250] In ROR-binding molecules, the N-terminus of Domain D is linked to the C-terminus of Domain B. In certain embodiments, Domain B has a CH3 amino acid sequence that is extended at the C-terminus at the junction between Domain D and Domain B, as described in further detail below in Section 6.3.19.3. 6.3.4. Domain E (Constant Region)

[0251] In ROR-binding molecules, Domain E comprises a constant region domain amino acid sequence, which is described in more detail in Section 6.3.2.

[0252] In certain embodiments, the constant region sequence is a CH3 sequence. CH3 sequences are described in more detail in Section 6.3.2.1, supra. In certain embodiments, the constant region sequence is mutated to include one or more orthogonal mutations. In preferred embodiments, Domain E has a constant region sequence that is a CH3 sequence containing either a knob-hole (synonymously "knob-in-hole," "KIH") orthogonal mutation, as described in more detail in Section 6.3.14.2, below, as well as a S354C or Y349C mutation that forms an engineered disulfide bridge with the CH3 domain containing the orthogonal mutation, as described in more detail in Section 6.3.14.1, below. In some preferred embodiments, the knob-hole orthogonal mutation is a T366W mutation.

[0253] In certain embodiments, the constant region domain sequence is a CH1 sequence. In certain embodiments, the CH1 amino acid sequence of domain E is the only CH1 amino acid sequence in the ROR-binding molecule. In certain embodiments, the N-terminus of the CH1 domain is linked to the C-terminus of the CH2 domain, as described in further detail in 6.3.19.5 below. In certain embodiments, the constant region sequence is a CL sequence. In certain embodiments, the N-terminus of the CL domain is linked to the C-terminus of the CH2 domain, as described in further detail in 6.3.19.5 below. CH1 and CL sequences are described in further detail in Section 6.3.8.1. 6.3.5. Domain F (variable region)

[0254] In the ROR-binding molecule, Domain F has a variable region domain amino acid sequence. The variable region domain amino acid sequence, as described in further detail in Section 6.3.1, is an antibody variable region domain amino acid sequence, including VL and VH antibody domain sequences. The VL and VH sequences are described in further detail above in Sections 6.3.1.1 and 6.3.1.4, respectively. In a preferred embodiment, Domain F has a VH antibody domain sequence. 6.3.6. Domain G (Constant Region)

[0255] In ROR-binding molecules, domain G comprises a constant region amino acid sequence, which is described in more detail in Section 6.3.2.

[0256] In certain preferred embodiments, the constant region sequence is a CH3 sequence. CH3 sequences are described in further detail below in Section 6.3.2.1. In other preferred embodiments, the constant region sequence is an orthologous CH2 sequence. Orthologous CH2 sequences are described in further detail below in Section 6.3.2.2.

[0257] In certain preferred embodiments, domain G has a human IgG1 CH3 sequence with the following mutational changes: S354C; and a tripeptide insertion 445P, 446G, 447K. In some preferred embodiments, domain G has a human IgG1 CH3 sequence with the following mutational changes: S354C; and a tripeptide insertion 445P, 446G, 447K. In some preferred embodiments, domain G has a human IgG1 CH3 sequence with the following changes: L351D; and a tripeptide insertion 445G, 446E, 447C. 6.3.7. Domain H (variable region)

[0258] In the ROR-binding molecule, Domain L has a variable region domain amino acid sequence. The variable region domain amino acid sequence, as described in further detail in Section 6.3.1, is an antibody variable region domain amino acid sequence, including VL and VH antibody domain sequences. VL and VH sequences are described in further detail above in Sections 6.3.1.1 and 6.3.1.4, respectively. In a preferred embodiment, Domain H has a VL antibody domain sequence. 6.3.8. Domain I (Constant Region)

[0259] In the ROR-binding molecule, Domain I has a constant region domain amino acid sequence. Constant region domain amino acid sequences are described in further detail above in Section 6.3.2. In one set of preferred embodiments of the ROR-binding molecule, Domain I has a CL amino acid sequence. In another set of embodiments, Domain I has a CH1 amino acid sequence. CH1 and CL amino acid sequences are described in further detail in Section 6.3.8.1. CH1 and CL regions

[0260] As used herein, the CH1 amino acid sequence is the sequence of the second domain of an antibody heavy chain, referenced from N-terminus to C-terminus. In certain embodiments, the CH1 sequence is an endogenous sequence. In various embodiments, the CH1 sequence is a mammalian sequence, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human sequences. In preferred embodiments, the CH1 sequence is a human sequence. In certain embodiments, the CH1 sequence is derived from an IgA1, IgA2, IgD, IgE, IgG1, IgG2, IgG3, IgG4, or IgM isotype. In preferred embodiments, the CH1 sequence is derived from an IgG1 isotype. In preferred embodiments, the CH1 sequence is amino acids 1-98 of UniProt Accession Number P01857.

[0261] The CL amino acid sequence useful in the ROR binding molecules described herein is an antibody light chain constant domain sequence.In certain embodiments, the CL sequence is an endogenous sequence.In various embodiments, the CL sequence is a mammalian sequence, including but not limited to mouse, rat, hamster, rabbit, camel, donkey, goat and human sequence.In a preferred embodiment, the CL sequence is a human sequence.

[0262] In certain embodiments, the CL amino acid sequence is a lambda (λ) light chain constant domain sequence. In certain embodiments, the CL amino acid sequence is a human lambda light chain constant domain sequence. In a preferred embodiment, the lambda (λ) light chain sequence is UniProt accession number P0CG04.

[0263] In certain embodiments, the CL amino acid sequence is a kappa (κ) light chain constant domain sequence. In a preferred embodiment, the CL amino acid sequence is a human kappa (κ) light chain constant domain sequence. In a preferred embodiment, the kappa light chain sequence is UniProt Accession No. P01834.

[0264] In certain embodiments, both the CH1 sequence and the CL sequence are endogenous sequences. In certain embodiments, as described in further detail in Section 6.3.8.2 below, the CH1 sequence and the CL sequence separately comprise orthogonal modifications to the endogenous CH1 and CL sequences, respectively. It should be understood that orthogonal mutations in the CH1 sequence do not eliminate the specific binding interaction between the CH1-binding reagent and the CH1 domain. However, in some embodiments, orthogonal mutations may reduce, if not eliminate, the specific binding interaction. The CH1 and CL sequences can also be portions of either endogenous or modified sequences, such that a domain having the CH1 sequence or a portion thereof can associate with a domain having the CH1 sequence or a portion thereof. Furthermore, ROR-binding molecules having a portion of the above-described CH1 sequence can be conjugated to a CH1-binding reagent.

[0265] Without wishing to be bound by theory, the CH1 domain is also unique in that its folding is typically the rate-limiting step in the secretion of IgG (Feige et al. Mol Cell. 2009 Jun 12;34(5):569-79; incorporated herein by reference in its entirety). Thus, purification of ROR-binding molecules based on the rate-limiting component of the CH1-containing polypeptide chain can provide a means of purifying complete complexes from incomplete chains, e.g., purification of complexes having the rate-limiting CH1 domain from complexes having only one or more non-CH1-containing chains.

[0266] As noted, while CH1-limited expression may be beneficial in some aspects, CH1 may limit the overall expression of complete ROR-binding molecules. Thus, in certain embodiments, expression of polypeptide chains containing CH1 sequence(s) is adjusted to improve the efficiency of ROR-binding molecules to form complete complexes. In an illustrative example, the ratio of plasmid vectors constructed to express polypeptide chains containing CH1 sequence(s) can be increased relative to plasmid vectors constructed to express other polypeptide chains. In another illustrative example, when compared to polypeptide chains containing CL sequence(s), polypeptide chains containing CH1 sequence(s) can be the smaller of the two polypeptide chains. In another specific embodiment, expression of polypeptide chains containing CH1 sequence(s) can be adjusted by controlling which polypeptide chains have CH1 sequence(s). For example, engineering an ROR-binding molecule so that a CH1 domain is present in a two-domain polypeptide chain (e.g., a fourth polypeptide chain described herein) instead of the native location of the CH1 sequence in a four-domain polypeptide chain (e.g., a third polypeptide chain described herein) can be used to control the expression of polypeptide chains containing a CH1 sequence(s). However, in other aspects, too high a relative expression level of a CH1-containing chain compared to other chains can result in incomplete complexes with the CH1 chain but without each of the other chains. Thus, in certain embodiments, expression of polypeptide chains containing a CH1 sequence(s) is adjusted to reduce the formation of incomplete complexes that lack the CH1-containing chain and to reduce the formation of incomplete complexes that contain the CH1-containing chain but lack the other chains present in the complete complex. 6.3.8.2. CH1 and CL Orthogonal Modifications

[0267] In certain embodiments, the CH1 and CL sequences independently comprise orthogonal modifications to the endogenous CH1 and CL sequences, respectively. Orthogonal mutations generally are described in more detail below in Sections 6.3.14.1-6.3.14.3.

[0268] In certain embodiments, the orthogonal modification in the endogenous CH1 and CL sequences is an engineered disulfide bridge selected from engineered cysteines at positions 138 in the CH1 sequence and 116 in the CL sequence, 128 in the CH1 sequence and 119 in the CL sequence, or 129 in the CH1 sequence and 210 in the CL sequence, as numbered and discussed in more detail in U.S. Patent Nos. 8,053,562 and 9,527,927, each of which is incorporated by reference in its entirety. In a preferred embodiment, the engineered cysteines are at positions 128 in the CH1 sequence and 118 in the CL kappa sequence, as numbered by the Eu index.

[0269] In one set of preferred embodiments, the mutation resulting in a non-inherent cysteine ​​amino acid is an A141C mutation in the CH1 sequence corresponding to an F118C mutation in the CL sequence, or an L128C mutation in the CH1 sequence corresponding to an F118C mutation in the CL sequence, or a P171C mutation in the CH1 sequence corresponding to an S162C mutation in the CL sequence, numbered by the Eu index.

[0270] In various embodiments, the orthogonal mutations in the CL sequence and the CH1 sequence are charge-pair mutations. In certain embodiments, the charge-pair mutation is A141L in the CH1 sequence corresponding to the F118S, F118A, or F118V mutation in the CL sequence numbered by the Eu index, or K147D in the CH1 sequence corresponding to the T129R mutation in the CL sequence, as further described in detail in Bonisch et al. (Protein Engineering, Design & Selection, 2017, pp.1-12), which is incorporated herein by reference in its entirety. In a preferred set of embodiments, the charge-pair mutation is G166D in the CH1 sequence corresponding to the N138K mutation in the CL sequence numbered by the Eu index, or G166K in the CH1 sequence corresponding to the N138D mutation in the CL sequence. Domain J (CH2)

[0271] In ROR-binding molecules, Domain J has a CH2 amino acid sequence, which is described in more detail in Section 6.3.3, supra. In a preferred embodiment, the CH2 amino acid sequence has an N-terminal hinge region connecting Domain J to Domain I, as described in more detail in Section 6.3.19.4, below.

[0272] In ROR-binding molecules, the C-terminus of domain J is linked to the N-terminus of domain K. In certain embodiments, domain J is linked to the N-terminus of domain K with a CH1 amino acid sequence or a CL amino acid sequence, as described in further detail below in Section 6.3.19.5. 6.3.10. Domain K (Constant Region)

[0273] In ROR-binding molecules, Domain K has a constant region domain amino acid sequence. Constant region domain amino acid sequences are described in more detail in Section 6.3.2, supra. In preferred embodiments, Domain K has a constant region sequence that is a CH3 sequence containing either a S354C or Y349C mutation that forms an engineered disulfide bridge with the CH3 domain containing the knob-hole orthogonal mutations, described in more detail in Section 6.3.14.2, below; an isoallotypic mutation, described in more detail in Section 6.3.2.1, above; and an orthogonal mutation, described in more detail in Section 6.3.14.1, below. In some preferred embodiments, the knob-hole orthogonal mutations combined with the isoallotypic mutations are the following mutational changes: D356E, L358M, T366S, L368A, and Y407V.

[0274] In certain embodiments, the constant region domain sequence is a CH1 sequence. In certain embodiments, the CH1 amino acid sequence of domain K is the only CH1 amino acid sequence in the ROR-binding molecule. In certain embodiments, the N-terminus of the CH1 domain is connected to the C-terminus of the CH2 domain, as described in further detail below in 6.3.19.5. In certain embodiments, the constant region sequence is a CL sequence. In certain embodiments, the N-terminus of the CL domain is connected to the C-terminus of the CH2 domain, as described in further detail below in 6.3.19.5. CH1 and CL sequences are described in further detail in Section 6.3.8.1. 6.3.11. Domain L (variable region)

[0275] In the ROR-binding molecule, Domain L has a variable region domain amino acid sequence. The variable region domain amino acid sequence, as described in further detail in Section 6.3.1, is an antibody variable region domain amino acid sequence, including VL and VH antibody domain sequences. VL and VH sequences are described in further detail above in Sections 6.3.1.1 and 6.3.1.4, respectively. In a preferred embodiment, Domain L has a VH antibody domain sequence. 6.3.12. Domain M (Constant Region)

[0276] In the ROR-binding molecule, domain M has a constant region domain amino acid sequence. Constant region domain amino acid sequences are described in further detail in Section 6.3.2, supra. In one set of preferred embodiments of the ROR-binding molecule, domain I has a CH1 amino acid sequence. In another set of preferred embodiments, domain I has a CL amino acid sequence. CH1 and CL amino acid sequences are described in further detail in Section 6.3.8.1. 6.3.13. Pairing of Domains A and F

[0277] In ROR-binding molecules, the domain A VL or VH amino acid sequence and the cognate domain F VL or VH amino acid sequence associate to form an antigen-binding site (ABS). The A:F antigen-binding site (ABS) can specifically bind to an epitope of an antigen. Antigen binding by ABS is described in further detail below in Section 6.3.13.1.

[0278] In various multivalent embodiments, the ABS formed by domains A and F (A:F) is identical in sequence to one or more other ABSs within the ROR-binding molecule and therefore has the same recognition specificity as one or more other sequence-identical ABSs within the ROR-binding molecule.

[0279] In various multivalent embodiments, the A:F ABS is non-identical in sequence to one or more other ABSs in the ROR-binding molecule. In certain embodiments, the A:F ABS has a different recognition specificity than one or more other non-sequence-identical ABSs in the ROR-binding molecule. In certain embodiments, the A:F ABS recognizes an antigen that is different from the antigen recognized by at least one other non-sequence-identical ABS in the ROR-binding molecule. In certain embodiments, the A:F ABS recognizes a different epitope of an antigen that is also recognized by at least one other non-sequence-identical ABS in the ROR-binding molecule. In this embodiment, the ABS formed by domains A and F recognizes an epitope of an antigen, where one or more other ABSs in the ROR-binding molecule recognize the same antigen but not the same epitope. 6.3.13.1. Antigen Binding by ABS

[0280] ABSs and ROR-binding molecules that contain such ABSs are said to "recognize" the epitope (or more generally, the antigen) to which the ABS specifically binds, and the epitope (or more generally, the antigen) is said to be the "recognition specificity" or "binding specificity" of the ABS.

[0281] An ABS is said to bind to its particular antigen or epitope with a particular affinity. As used herein, "affinity" refers to the strength of the non-covalent intermolecular force interaction between one molecule and another. Affinity, or the strength of the interaction, is determined by the dissociation equilibrium constant (K D ) where K D A lower value indicates a stronger interaction between the molecules. D The affinity can be measured by methods well known in the art, including, but not limited to, biolayer interferometry (e.g., Octet / FORTEBIO®), surface plasmon resonance (SPR) technology (e.g., Biacore®), and cell binding assays. For purposes herein, affinity is the dissociation equilibrium constant measured by biolayer interferometry using Octet / FORTEBIO®.

[0282] "Specific binding," as used herein, refers to the affinity between an ABS and its cognate antigen or epitope, where K D The value is 10 -6 M, 10 -7 M, 10 -8 M, 10 -9 M or 10 -10 It is less than M.

[0283] As outlined in Figure 2, the number of ABSs in the ROR-binding molecules described herein defines the "valency" of the ROR-binding molecule. An ROR-binding molecule with a single ABS is "monovalent." An ROR-binding molecule with multiple ABSs is said to be "multivalent." A multivalent ROR-binding molecule with two ABSs is "bivalent." A multivalent ROR-binding molecule with three ABSs is "trivalent." A multivalent ROR-binding molecule with four ABSs is "tetravalent."

[0284] In various multivalent embodiments, the multiple ABSs all have the same recognition specificity. As outlined in Figure 2, such ROR-binding molecules are "monospecific" or "multivalent" binding constructs. In other multivalent embodiments, at least two of the multiple ABSs have different recognition specificities. Such ROR-binding molecules are multivalent and "multispecific." In multivalent embodiments, the ABSs collectively have two recognition specificities, the ROR-binding molecule is "bispecific." In multivalent embodiments, the ABSs collectively have three recognition specificities, the ROR-binding molecule is "trispecific."

[0285] In multivalent embodiments in which the ABSs collectively have multiple recognition specificities for different epitopes present on the same antigen, the ROR-binding molecule is "multiparatopic." In multivalent embodiments in which the ABSs collectively recognize two epitopes on the same antigen, the ROR-binding molecule is "biparatopic."

[0286] In various multivalent embodiments, the multivalency of the ROR-binding molecule improves the avidity of the ROR-binding molecule for a particular target. As used herein, "avidity" refers to the overall strength of interaction between two or more molecules, e.g., multivalent ROR-binding molecules, for a particular target, where avidity is the cumulative strength of the interaction given by the affinities of multiple ABSs. Avidity can be measured by the same methods used to determine affinity, as described above. In certain embodiments, the avidity of the ROR-binding molecule for a particular target is determined by the interaction being a specific binding interaction, where the avidity between the two molecules is greater than or equal to 10. -6 M, 10 -7 M, 10 -8 M, 10 -9 M or 10 -10 K less than M D In certain embodiments, the avidity of an ROR-binding molecule for a particular target is determined by a K Dwhere the affinity of one or more of the individual ABSs is sufficient for specific binding to their respective antigens or epitopes themselves. D In certain embodiments, avidity is the cumulative strength of interaction conferred by the affinity of multiple ABSs for distinct antigens of a common specific target or complex, such as distinct antigens found on individual cells. In certain embodiments, avidity is the cumulative strength of interaction conferred by the affinity of multiple ABSs for distinct epitopes of a common individual antigen. 6.3.14. Pairing of Domains B and G

[0287] In the ROR-binding molecules described herein, a domain B constant region amino acid sequence and a domain G constant region amino acid sequence are associated. The constant region domain amino acid sequences are described in further detail above in Section 6.3.2.

[0288] In one set of preferred embodiments, Domain B and Domain G have a CH3 amino acid sequence. CH3 sequences are described in further detail above in Section 6.3.2.1. In various embodiments, the amino acid sequences of the B and G domains are identical. In some of these embodiments, the sequence is an endogenous CH3 sequence.

[0289] In various embodiments, the amino acid sequences of the B and G domains are distinct and each separately contain an orthogonal modification to the endogenous CH3 sequence, where the B domain interacts with the G domain and neither the B nor the G domain interacts significantly with the CH3 domain lacking the orthogonal modification.

[0290] As used herein, an "orthogonal modification" or, equivalently, an "orthogonal mutation" refers to one or more engineered mutations in the amino acid sequence of an antibody domain that increase the binding affinity of a first domain bearing the orthogonal modification to a second domain bearing the complementary orthogonal modification. In certain embodiments, the orthogonal modification decreases the binding affinity of a domain bearing the orthogonal modification to a domain lacking the complementary orthogonal modification. In certain embodiments, the orthogonal modification is a mutation in the endogenous antibody domain sequence. In various embodiments, the orthogonal modification is a modification of the N- or C-terminus of the endogenous antibody domain sequence, including, but not limited to, the addition or deletion of amino acids. In certain embodiments, the orthogonal modification includes, but is not limited to, engineered disulfide bridges, knob-in-hole mutations, and charge-pair mutations, which are described in further detail in Sections 6.3.14.1-6.3.14.3 below. In certain embodiments, the orthogonal modification includes a combination of orthogonal modifications selected from, but not limited to, engineered disulfide bridges, knob-in-hole mutations, and charge-pair mutations. In certain embodiments, orthogonal modifications can be combined with amino acid substitutions that reduce immunogenicity, such as isoallotypic mutations, which are described in further detail in Section 6.3.2.1, above. 6.3.14.1. Orthogonally Engineered Disulfide Bridges

[0291] In various embodiments, the orthogonal modification comprises a mutation that generates an engineered disulfide bridge between the first and second domains. As described herein, an "engineered disulfide bridge" is a mutation that provides a non-endogenous cysteine ​​amino acid to two or more domains so that a non-native disulfide bond forms when the two or more domains associate. Engineered disulfide bridges are described in more detail in Merchant et al. (Nature Biotech (1998) 16:677-681), the entire contents of which are incorporated herein by reference. In certain embodiments, the engineered disulfide bridge improves the orthogonal association between certain domains. In certain embodiments, the mutation that generates the engineered disulfide bridge is a K392C mutation in one of the first or second CH3 domains and a D399C mutation in the other CH3 domain. In a preferred embodiment, the mutations that generate the engineered disulfide bridges are a S354C mutation in one of the first or second CH3 domains and a Y349C mutation in the other CH3 domain. In another preferred embodiment, the mutations that generate the engineered disulfide bridges are a 447C mutation in both the first and second CH3 domains, provided by a C-terminal extension of the CH3 domains incorporating the KSC tripeptide sequence. Orthogonal Knob-Hole Mutation

[0292] In various embodiments, the orthogonal modification comprises a knob-hole (synonymously, knobs-in-hole) mutation. As described herein, a knob-hole mutation is a mutation that alters the surface conformational characteristics of a first domain such that the first domain preferentially associates with a second domain that has a complementary conformational mutation compared to association with a domain without the complementary conformational mutation. Knob-hole mutations are described in more detail in U.S. Pat. Nos. 5,821,333 and 8,216,805, each of which is incorporated herein by reference in its entirety. In various embodiments, knob-hole mutations are described in more detail in U.S. Pat. Nos. 5,821,333 and 8,216,805, each of which is incorporated herein by reference in its entirety. 16:677-681) are combined with engineered disulfide bridges. In various embodiments, knob-hole mutations, isoallotypic mutations, and engineered disulfide mutations are combined.

[0293] In certain embodiments, the knob-in-hole mutations are a T366Y mutation in the first domain and a Y407T mutation in the second domain. In certain embodiments, the knob-in-hole mutations are F405A in the first domain and T394W in the second domain. In certain embodiments, the knob-in-hole mutations are a T366Y mutation and F405A in the first domain and a T394W and Y407T mutation in the second domain. In certain embodiments, the knob-in-hole mutations are a T366W mutation in the first domain and a Y407A mutation in the second domain. In certain embodiments, the combined knob-in-hole mutation and engineered disulfide mutation are S354C and T366W mutations in the first domain and a Y349C, T366S, L368A, and Y407V mutations in the second domain. In a preferred embodiment, the combined knob-in-hole, isoallotypic, and engineered disulfide mutations are S354C and T366W mutations in the first domain, and Y349C, D356E, L358M, T366S, L368A, and Y407V mutations in the second domain. Orthogonal Charge Pair Mutations

[0294] In various embodiments, orthogonal modification is charge-pair mutation.As used herein, charge-pair mutation is a mutation that affects the charge of the amino acid on the surface of a domain, so that the domain preferentially associates with a second domain that has a complementary charge-pair mutation compared to the association with a domain that does not have a complementary charge-pair mutation.In certain embodiments, charge-pair mutation improves the orthogonal association between specific domains.Charge-pair mutations are described in more detail in U.S. Patent No. 8,592,562, U.S. Patent No. 9,248,182, and U.S. Patent No. 9,358,286, each of which is incorporated herein by reference for all teachings.In certain embodiments, charge-pair mutation improves the stability between specific domains.In a preferred embodiment, charge-pair mutation is a T366K mutation in the first domain and an L351D mutation in another domain. 6.3.15. Pairing of Domains E and K

[0295] In various embodiments, the E domain has a CH3 amino acid sequence.

[0296] In various embodiments, the K domain has a CH3 amino acid sequence.

[0297] In various embodiments, the amino acid sequences of the E and K domains are identical and the sequence is the endogenous CH3 sequence.

[0298] In various embodiments, the sequences of the E and K domains are different. In various embodiments, the different sequences each independently comprise an orthogonal modification to the endogenous CH3 sequence, where the E domain interacts with the K domain and neither the E nor the K domain significantly interacts with the CH3 domain, which lacks the orthogonal modification. In certain embodiments, the orthogonal modifications include, but are not limited to, engineered disulfide bridges, knobs-in-hole mutations, and charge pair mutations, as described in further detail above in Sections 6.3.14.1-6.3.14.3. In certain embodiments, the orthogonal modifications include a combination of orthogonal modifications selected from, but not limited to, engineered disulfide bridges, knobs-in-hole mutations, and charge pair mutations. In certain embodiments, the orthogonal modifications can be combined with amino acid substitutions that reduce immunogenicity, such as isoallotypic mutations. 6.3.16. Pairing of Domains I and M and Domains H and L

[0299] In various embodiments, Domain I has a CL sequence and Domain M has a CH1 sequence. In various embodiments, Domain H has a VL sequence and Domain L has a VH sequence. In a preferred embodiment, Domain H has a VL amino acid sequence, Domain I has a CL amino acid sequence, Domain L has a VH amino acid sequence, and Domain M has a CH1 amino acid sequence. In another preferred embodiment, Domain H has a VL amino acid sequence, Domain I has a CL amino acid sequence, Domain L has a VH amino acid sequence, Domain M has a CH1 amino acid sequence, and Domain K has a CH3 amino acid sequence.

[0300] In various embodiments, the amino acid sequences of the I and M domains each independently comprise an orthogonal modification to their endogenous sequences, where the I domain interacts with the M domain and neither the I nor the M domain interacts significantly with the domain lacking the orthogonal modification. In one set of embodiments, the orthogonal mutations in the I domain are in the CL sequence and the orthogonal mutations in the M domain are in the CH1 sequence. The orthogonal mutations present in the CH1 and CL sequences are described in more detail above in Section 6.3.8.2.

[0301] In various embodiments, the amino acid sequences of the H domain and the L domain each independently contain an orthogonal modification in their endogenous sequences, where the H domain interacts with the L domain, and neither the H domain nor the L domain significantly interacts with the domain lacking the orthogonal modification. In one set of embodiments, the orthogonal mutation in the H domain is in the VL sequence, and the orthogonal mutation in the L domain is in the VH sequence. In certain embodiments, the orthogonal mutation is a charge pair mutation at the VH / VL interface. In a preferred embodiment, the charge pair mutation at the VH / VL interface is a Q39E in VH with a corresponding Q38K in VL, or a Q39K in VH with a corresponding Q38E in VL, as described in more detail in Igawa et al. (Protein Eng. Des. Sel., 2010, vol. 23, 667-677), the entire teaching of which is incorporated herein by reference.

[0302] In certain embodiments, the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, and the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen. In certain embodiments, the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, and the interaction between the H and L domains forms a second antigen-binding site specific for the first antigen. Trivalent ROR-binding molecules

[0303] In another set of embodiments, the ROR-binding molecule has three antigen-binding sites and is therefore referred to as "trivalent."

[0304] Referring to Figure 21, in various trivalent embodiments, the ROR-binding molecule further comprises a fifth polypeptide chain, where (a) the first polypeptide chain further comprises Domain N and Domain O, where the domains are arranged N-terminally to C-terminally in a NOA, B, C, D, E, orientation, where Domain N has a VL amino acid sequence and Domain O has a constant region amino acid sequence; (b) the ROR-binding molecule further comprises a fifth polypeptide chain, where the fifth polypeptide chain comprises Domain P and Domain Q, where the domains are arranged N-terminally to C-terminally in a PQ orientation, where Domain P has a VH amino acid sequence and Domain Q has a constant amino acid sequence; and (c) the first and fifth polypeptides associate via interactions between the N and P domains and between the O and Q domains to form the ROR-binding molecule. As outlined in Figure 2, these trivalent embodiments are referred to as "2x1" trivalent constructs.

[0305] Referring to Figure 26, in a further series of trivalent embodiments, the ROR-binding molecule further comprises a sixth polypeptide chain, wherein (a) the third polypeptide chain further comprises Domain R and Domain S, where the domains are arranged N-terminally to C-terminally in an RSHIJK orientation, with Domain R having a VL amino acid sequence and Domain S having a constant domain amino acid sequence; (b) the ROR-binding molecule further comprises a sixth polypeptide chain, where the sixth polypeptide chain comprises Domain T and Domain U, where the domains are arranged N-terminally to C-terminally in a TU orientation, with Domain T having a VH amino acid sequence and Domain U having a constant domain amino acid sequence; and (c) the third and sixth polypeptides associate via interactions between the R and T domains and between the S and U domains to form the ROR-binding molecule. As outlined in Figure 2, these trivalent embodiments are referred to as "1x2" trivalent constructs.

[0306] In various embodiments, Domain O is linked to Domain A via a peptide linker. In various embodiments, Domain S is linked to Domain H via a peptide linker. In a preferred embodiment, the peptide linker linking Domain O to Domain A or Domain S to Domain H is the 6 amino acid GSGSGS peptide sequence, described in more detail in Section 6.3.19.6. Trivalent 2 x 1 Bispecific Construct [2(AA) x 1(B)]

[0307] With reference to Figure 21, in various embodiments, the amino acid sequences of Domain N and Domain A are identical, the amino acid sequence of Domain H is different from the amino acid sequences of Domain N and A, the amino acid sequences of Domain O and Domain B are identical, the amino acid sequence of Domain I is different from the amino acid sequences of Domain O and B, the amino acid sequences of Domain P and Domain F are identical, the amino acid sequence of Domain L is different from the amino acid sequences of Domain P and F, the amino acid sequences of Domain Q and Domain G are identical, and the amino acid sequence of Domain M is different from the amino acid sequences of Domain Q and G; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and Domain N and Domain P form a third antigen-binding site specific for the first antigen. Trivalent 2 x 1 Bispecific Construct [2(AB) x 1(A)]

[0308] With reference to Figure 21, in various embodiments, the amino acid sequences of Domain N and Domain H are identical, the amino acid sequence of Domain A is different from the amino acid sequences of Domain N and H, the amino acid sequences of Domain O and Domain I are identical, the amino acid sequence of Domain B is different from the amino acid sequences of Domain O and I, the amino acid sequences of Domain P and Domain L are identical, the amino acid sequence of Domain F is different from the amino acid sequences of Domain P and L, the amino acid sequences of Domain Q and Domain M are identical, and the amino acid sequence of Domain G is different from the amino acid sequences of Domain Q and M; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and Domain N and Domain P form a third antigen-binding site specific for the second antigen. Trivalent 2 x 1 Trispecific Construct [2(AB) x 1(C)]

[0309] Referring to Figure 21, in various embodiments, the amino acid sequences of Domain N, Domain A, and Domain H are different, the amino acid sequences of Domain O, Domain B, and Domain I are different, the amino acid sequences of Domain P, Domain F, and Domain L are different, and the amino acid sequences of Domain Q, Domain G, and Domain M are different; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and Domain N and Domain P form a third antigen-binding site specific for a third antigen.

[0310] In certain embodiments, Domain O has a constant region sequence that is a CL from a kappa light chain, and Domain Q has a constant region sequence that is a CH1 from an IgG1 isotype, as described in more detail in Section 6.3.8.1. In preferred embodiments, Domain O and Domain Q have CH3 sequences such that they specifically associate with each other, as discussed in more detail in Section 6.3.14 above. Trivalent 1 x 2 Bispecific Constructs [1(A) x 2(BA)]

[0311] With reference to Figure 26, in various embodiments, the amino acid sequences of domains R and A are identical, the amino acid sequence of domain H is different from the amino acid sequences of domains R and A, the amino acid sequences of domains S and B are identical, the amino acid sequence of domain I is different from the amino acid sequences of domains S and B, the amino acid sequences of domains T and F are identical, the amino acid sequence of domain L is different from the amino acid sequences of domains T and F, the amino acid sequences of domains U and G are identical, and the amino acid sequence of domain M is different from the amino acid sequences of domains U and G; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and domains R and T form a third antigen-binding site specific for the first antigen. Trivalent 1 x 2 Bispecific Construct [1(A) x 2(BB)]

[0312] In various embodiments, the ROR-binding molecule further comprises a second CH1 domain or portion thereof. Referring to Figure 26, in certain embodiments, the amino acid sequences of domains R and H are identical, the amino acid sequence of domain A differs from the amino acid sequences of domains R and H, the amino acid sequences of domains S and I are identical, the amino acid sequence of domain B differs from the amino acid sequences of domains S and I, the amino acid sequences of domains T and L are identical, the amino acid sequence of domain F differs from the amino acid sequences of domains T and L, the amino acid sequences of domains U and M are identical, and the amino acid sequence of domain G differs from the amino acid sequences of domains U and M; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and domains R and T form a third antigen-binding site specific for the second antigen.

[0313] In certain embodiments, the amino acid sequences of domains S and I are CH1 sequences. In certain embodiments, the amino acid sequences of domains U and M are CH1 sequences. Trivalent 1 x 2 Trispecific Construct [1(A) x 2(BC)]

[0314] Referring to Figure 26, in various embodiments, the amino acid sequences of Domain R, Domain A, and Domain H are different, the amino acid sequences of Domain S, Domain B, and Domain I are different, the amino acid sequences of Domain T, Domain F, and Domain L are different, and the amino acid sequences of Domain U, Domain G, and Domain M are different; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and Domain R and Domain T form a third antigen-binding site specific for a third antigen.

[0315] In certain embodiments, Domain S has a constant region sequence that is a CL from a kappa light chain, and Domain U has a constant region sequence that is a CH1 from an IgG1 isotype, as described in more detail above in Section 6.3.8.1. In a preferred embodiment, Domain S and Domain U have CH3 sequences that allow them to specifically associate with each other, as described in more detail above in Section 6.3.14.

[0316] In certain embodiments, the ROR-binding molecule further comprises a second CH1 domain or a portion thereof. In certain embodiments, the amino acid sequences of domains S and I are CH1 sequences. In certain embodiments, the amino acid sequences of domains U and M are CH1 sequences. 6.3.18. Tetravalent 2x2 ROR-binding molecules

[0317] In various embodiments, the ROR-binding molecule has four antigen-binding sites and is therefore referred to as "tetravalent."

[0318] Referring to Figure 34, in a further set of embodiments, the ROR-binding molecule further comprises fifth and sixth polypeptide chains, wherein (a) the first polypeptide chain further comprises domain N and domain O, wherein the domains are arranged from N-terminus to C-terminus in a NOABDE orientation; (b) the third polypeptide chain further comprises domain R and domain S, wherein the domains are arranged from N-terminus to C-terminus in a RSHIJK orientation; (c) the ROR-binding molecule further comprises fifth and sixth polypeptide chains, wherein the fifth polypeptide chain comprises domain P and domain Q, wherein the domains are arranged from N-terminus to C-terminus in a PQ orientation, and the sixth polypeptide chain comprises domain T and domain U, wherein the domains are arranged from N-terminus to C-terminus in a TU orientation; and (d) the first and fifth polypeptides associate via interactions between the N and P domains and between the O and Q domains, and the third and sixth polypeptides associate via interactions between the R and T domains and between the S and U domains to form the ROR-binding molecule.

[0319] In various embodiments, Domain O is linked to Domain A via a peptide linker, and Domain S is linked to Domain H via a peptide linker. In a preferred embodiment, the peptide linker linking Domain O to Domain A and Domain S to Domain H is the 6 amino acid GSGSGS peptide sequence, as described in more detail in Section 6.3.19.6. 6.3.18.1. Tetravalent 2x2 Bispecific Constructs

[0320] Referring to Figure 34, in a series of tetravalent 2x2 bispecific ROR binding molecules, the amino acid sequences of Domain N and Domain A are identical, the amino acid sequences of Domain H and Domain R are identical, the amino acid sequences of Domain O and Domain B are identical, the amino acid sequences of Domain I and Domain S are identical, the amino acid sequences of Domain P and Domain F are identical, the amino acid sequences of Domain L and Domain T are identical, the amino acid sequences of Domain Q and Domain G are identical, and the amino acid sequences of Domain M and Domain U are identical; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, Domain N and Domain P form a second antigen-binding site specific for the first antigen, the interaction between the H and L domains forms a third antigen-binding site specific for a second antigen, and the interaction between the R and T domains forms a fourth antigen-binding site specific for the second antigen.

[0321] Referring to Figure 34, in another series of tetravalent 2x2 bispecific ROR binding molecules, the amino acid sequences of Domain H and Domain A are identical, the amino acid sequences of Domain N and Domain R are identical, the amino acid sequences of Domain I and Domain B are identical, the amino acid sequences of Domain O and Domain S are identical, the amino acid sequences of Domain L and Domain F are identical, the amino acid sequences of Domain P and Domain T are identical, the amino acid sequences of Domain M and Domain G are identical, and the amino acid sequences of Domain Q and Domain U are identical; the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, Domain N and Domain P form a second antigen-binding site specific for a second antigen, the interaction between the H and L domains forms a third antigen-binding site specific for the first antigen, and the interaction between the R and T domains forms a fourth antigen-binding site specific for the second antigen. DOMAIN JUNCTIONS 6.3.19.1. Junction connecting VL and CH3 domains

[0322] In various embodiments, the amino acid sequence forming the junction between the C-terminus of the VL domain and the N-terminus of the CH3 domain is an engineered sequence. In certain embodiments, one or more amino acids are deleted or added at the C-terminus of the VL domain. In certain embodiments, the junction connecting the C-terminus of the VL domain and the N-terminus of the CH3 domain is one of the sequences set forth in Table 2 in Section 6.13.7, below. In certain embodiments, A111 is deleted at the C-terminus of the VL domain. In certain embodiments, one or more amino acids are deleted or added at the N-terminus of the CH3 domain. In certain embodiments, P343 is deleted at the N-terminus of the CH3 domain. In certain embodiments, P343 and R344 are deleted at the N-terminus of the CH3 domain. In certain embodiments, one or more amino acids are deleted or added at both the C-terminus of the VL domain and the N-terminus of the CH3 domain. In certain embodiments, A111 is deleted at the C-terminus of the VL domain, and P343 is deleted at the N-terminus of the CH3 domain. In preferred embodiments, A111 and V110 are deleted at the C-terminus of the VL domain. In another preferred embodiment, A111 and V110 are deleted at the C-terminus of the VL domain, and the N-terminus of the CH3 domain has a P343V mutation. 6.3.19.2. Junction Linking the VH and CH3 Domains

[0323] In various embodiments, the amino acid sequence forming the junction between the C-terminus of the VH domain and the N-terminus of the CH3 domain is an engineered sequence. In certain embodiments, one or more amino acids are deleted or added at the C-terminus of the VH domain. In certain embodiments, the junction connecting the C-terminus of the VH domain and the N-terminus of the CH3 domain is one of the sequences listed in Table 3 in Section 6.13.7 below. In certain embodiments, K117 and G118 are deleted at the C-terminus of the VH domain. In certain embodiments, one or more amino acids are deleted or added at the N-terminus of the CH3 domain. In certain embodiments, P343 is deleted at the N-terminus of the CH3 domain. In certain embodiments, P343 and R344 are deleted at the N-terminus of the CH3 domain. In certain embodiments, P343, R344, and E345 are deleted at the N-terminus of the CH3 domain. In certain embodiments, one or more amino acids are deleted or added at both the C-terminus of the VH domain and the N-terminus of the CH3 domain. In a preferred embodiment, T116, K117, and G118 are deleted at the C-terminus of the VH domain. 6.3.19.3. Junction (hinge) connecting the C-terminus of CH3 to the N-terminus of CH2

[0324] In the ROR-binding molecules described herein, the N-terminus of the CH2 domain has a "hinge" region amino acid sequence. As used herein, a hinge region is a sequence of an antibody heavy chain that connects the N-terminal variable domain-constant domain segment of an antibody to the CH2 domain of the antibody. In addition, the hinge region typically provides both flexibility between the N-terminal variable domain-constant domain segment and the CH2 domain, as well as an amino acid sequence motif that forms disulfide bridges between heavy chains (e.g., the first and third polypeptide chains). As used herein, the hinge region amino acid sequence is SEQ ID NO: 56.

[0325] In various embodiments, the CH3 amino acid sequence is extended at the C-terminus at the junction between the C-terminus of the CH3 domain and the N-terminus of the CH2 domain. In certain embodiments, the CH3 amino acid sequence is extended at the C-terminus at the junction between the C-terminus of the CH3 domain and the hinge region, and then linked to the N-terminus of the CH2 domain. In a preferred embodiment, the CH3 amino acid sequence is extended by inserting a PGK tripeptide sequence followed by the DKTHT motif of the IgG1 hinge region.

[0326] In certain embodiments, the C-terminal extension of a CH3 domain incorporates an amino acid sequence capable of forming a disulfide bond with an orthogonal C-terminal extension of another CH3 domain. In a preferred embodiment, the C-terminal extension of a CH3 domain incorporates a KSC tripeptide sequence followed by a DKTHT motif of an IgG1 hinge region, which in turn forms a disulfide bond with an orthogonal C-terminal extension of another CH3 domain incorporating a GEC motif of a kappa light chain. 6.3.19.4. Junction (hinge) connecting the C-terminus of CL and the N-terminus of CH2

[0327] In various embodiments, the CL amino acid sequence is linked via its C-terminus to the hinge region, which is then linked to the N-terminus of the CH2 domain. Hinge region sequences are described in more detail in Section 6.3.19.3 above. In a preferred embodiment, the hinge region amino acid sequence is SEQ ID NO:56. 6.3.19.5. Junction Linking the C-Terminus of CH2 to the Constant Region Domain

[0328] In various embodiments, the CH2 amino acid sequence is linked via its C-terminus to the N-terminus of the constant region domain. Constant regions are described in more detail in Section 6.3.4 above. In a preferred embodiment, the CH2 sequence is linked to the CH3 sequence via its endogenous sequence. In other embodiments, the CH2 sequence is linked to a CH1 or CL sequence. Examples discussing linking a CH2 sequence to a CH1 or CL sequence are described in more detail in U.S. Patent No. 8,242,247, the entire contents of which are incorporated herein. 6.3.19.6. Junctions Linking Domain O to Domain A or Domain S to Domain H in Trivalent and Tetravalent Molecules

[0329] In various embodiments, the heavy chains of the antibody (e.g., the first and third polypeptide chains) are extended at their N-terminus to include additional domains that provide additional ABS. With reference to Figures 21, 26, and 34, in certain embodiments, the C-terminus of the constant region domain amino acid sequence of Domain O and / or Domain S is linked to the N-terminus of the variable region domain amino acid sequence of Domain A and / or Domain H, respectively. In some preferred embodiments, the constant region domain is a CH3 amino acid sequence and the variable region domain is a VL amino acid sequence. In some preferred embodiments, the constant region domain is a CL amino acid sequence and the variable region domain is a VL amino acid sequence. In certain embodiments, the constant region domain is linked to the variable region domain via a peptide linker. In a preferred embodiment, the peptide linker is the 6-amino acid GSGSGS peptide sequence.

[0330] In various embodiments, the light chains (e.g., the second and fourth polypeptide chains) of the antibody are extended at their N-terminus to include additional variable domain-constant domain segments of the antibody. In certain embodiments, the constant region domain is a CH1 amino acid sequence and the variable region domain is a VH amino acid sequence. 6.4. Specific Bivalent ROR-Binding Molecules

[0331] In a further embodiment, bivalent ROR binding molecules are provided.

[0332] Referring to Figure 3, in one set of embodiments, the ROR-binding molecule comprises first, second, third, and fourth polypeptide chains, wherein (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged, from N-terminus to C-terminus, in an A-B-D-E orientation, and Domain A has a VL amino acid sequence, Domain B has a CH3 amino acid sequence, Domain D has a CH2 amino acid sequence, and Domain E has a constant region domain amino acid sequence; (b) the second polypeptide chain comprises Domain F and Domain G, wherein the domains are arranged, from N-terminus to C-terminus, in an F-G orientation, and Domain F has a VH amino acid sequence and Domain G has a CH3 amino acid sequence; and (c) the third polypeptide chain comprises Domain H, Domain I, Domain J, and Domain K, wherein the domains are arranged, from N-terminus to C-terminus, in an H-I-J-K orientation, and Domain H has a variable region domain amino acid sequence. (d) a fourth polypeptide chain comprises domain L and domain M, wherein the domains are arranged N-terminally to C-terminally in an LM orientation, with domain L having a variable region domain amino acid sequence and domain M having a constant region domain amino acid sequence; (e) the first and second polypeptides associate through interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate through interactions between the H and L domains and between the I and M domains; and (g) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form a ROR-binding molecule.

[0333] In a preferred embodiment, domain E has a CH3 amino acid sequence; domain H has a VL amino acid sequence; domain I has a CL amino acid sequence, domain K has a CH3 amino acid sequence; domain L has a VH amino acid sequence; and domain M has a CH1 amino acid sequence.

[0334] In certain embodiments, the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, and the interaction between the H and L domains forms a second antigen-binding site specific for a second antigen, and the ROR-binding molecule is a bispecific, bivalent ROR-binding molecule. In certain embodiments, the interaction between the A and F domains forms a first antigen-binding site specific for a first antigen, and the interaction between the H and L domains forms a second antigen-binding site specific for the first antigen, and the ROR-binding molecule is a monospecific, bivalent ROR-binding molecule. 6.4.1. Bivalent, bispecific B-Body "BC1"

[0335] 3 and 6, in one set of embodiments, the ROR-binding molecule has first, second, third, and fourth polypeptide chains, wherein (a) the first polypeptide chain comprises domain A, domain B, domain D, and domain E, wherein the domains are arranged N-terminus to C-terminus in an A-B-D-E orientation, and domain A has a first VL amino acid sequence; domain B has a human IgG1 CH3 amino acid sequence with a T366K mutation and a C-terminal extension incorporating a KSC tripeptide sequence followed by a DKTHT motif of the IgG1 hinge region; domain D has a human IgG1 CH2 amino acid sequence; and domain E has a human IgG1 CH2 amino acid sequence with a S354C and T366W mutation. (b) a second polypeptide chain having domain F and domain G, wherein the domains are arranged, from N-terminus to C-terminus, in an FG orientation, where domain F has a first VH amino acid sequence and domain G has a human IgG1 CH3 amino acid sequence with an L351D mutation and a C-terminal extension incorporating a GEC amino acid disulfide motif; (c) a third polypeptide chain having domain H, domain I, domain J, and domain K, wherein the domains are arranged, from N-terminus to C-terminus, in an H-I-J-K orientation, where domain H has a second VL amino acid sequence, domain I has a human CL kappa amino acid sequence, domain J has a human IgG1 CH2 amino acid sequence, and domain K has a human IgG1 CH2 amino acid sequence with Y349C, D356E, L358M, T366S, L368A, and Y407V mutations. (d) a fourth polypeptide chain has a domain L and a domain M, wherein the domains are arranged N-terminally to C-terminally in an LM orientation, and domain L has a second VH amino acid sequence and domain M has a human IgG1 CH1 amino acid sequence; (e) the first and second polypeptides associate via interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate via interactions between the H and L domains and between the I and M domains;(g) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form a ROR-binding molecule; (h) domain A and domain F form a first antigen-binding site specific for a first antigen; (i) domain H and domain L form a second antigen-binding site specific for a second antigen;

[0336] In a preferred embodiment, the first polypeptide chain has the sequence of SEQ ID NO:8, the second polypeptide chain has the sequence of SEQ ID NO:9, the third polypeptide chain has the sequence of SEQ ID NO:10, and the fourth polypeptide chain has the sequence of SEQ ID NO:11. 6.4.2. Bivalent, bispecific B-Body "BC6"

[0337] 3 and 14, in one set of embodiments, the ROR-binding molecule has first, second, third, and fourth polypeptide chains, wherein (a) the first polypeptide chain comprises domain A, domain B, domain D, and domain E, wherein the domains are arranged from N-terminus to C-terminus in an A-B-D-E orientation, and domain A has a first VL amino acid sequence, domain B has a human IgG1 CH3 amino acid sequence with a KSC tripeptide sequence followed by a C-terminal extension incorporating a DKTHT motif of an IgG1 hinge region, domain D has a human IgG1 CH2 amino acid sequence, and domain E has human IgG1 CH3 amino acids with S354C and T366W mutations; and (b) the second polypeptide chain comprises domain F and domain G, wherein the domains are arranged from N-terminus to C-terminus in an F-G orientation, and domain F has a first VH amino acid sequence, and domain G has a human IgG1 CH3 amino acid sequence with a C-terminal extension incorporating a GEC amino acid disulfide motif. (c) a third polypeptide chain having a domain H, a domain I, a domain J, and a domain K, wherein the domains are arranged from N-terminus to C-terminus in a H-I-J-K orientation, and wherein domain H has a second VL amino acid sequence, domain I has a human CL kappa amino acid sequence, domain J has a human IgG1 CH2 amino acid sequence, and domain K has a human IgG1 CH2 amino acid sequence with Y349C, D356E, L358M, T366S, L368A, and Y407V mutations. (d) a fourth polypeptide chain has a domain L and a domain M, wherein the domains are arranged N-terminus to C-terminus in an LM orientation, and domain L has a second VH amino acid sequence and domain M has a human IgG1 amino acid sequence; (e) the first and second polypeptides associate through interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate through interactions between the H and L domains and between the I and M domains; (g) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form a ROR-binding molecule;(h) Domain A and Domain F form a first antigen-binding site specific for a first antigen; (i) Domain H and Domain L form a second antigen-binding site specific for a second antigen; 6.4.3. Bivalent, bispecific B-Body "BC28"

[0338] 3 and 16, in one set of embodiments, the ROR-binding molecule has first, second, third, and fourth polypeptide chains, wherein (a) the first polypeptide chain comprises domain A, domain B, domain D, and domain E, wherein the domains are arranged from N-terminus to C-terminus in an A-B-D-E orientation, and domain A has a first VL amino acid sequence, domain B has a human IgG1 CH3 amino acid sequence with a Y349C mutation and a PGK tripeptide sequence followed by a C-terminal extension incorporating a DKTHT motif of an IgG1 hinge region, domain D has a human IgG1 CH2 amino acid sequence, and domain E has human IgG1 CH3 amino acid sequences with S354C and T366W mutations; and (b) the second polypeptide chain comprises domain F and domain G, wherein the domains are arranged from N-terminus to C-terminus in an F-G orientation, and domain F has a first VH amino acid sequence, and domain G has a human IgG1 CH3 amino acid sequence with a S354C mutation and a C-terminal extension incorporating a PGK tripeptide sequence. (c) a third polypeptide chain having a domain H, a domain I, a domain J, and a domain K, wherein the domains are arranged from N-terminus to C-terminus in a H-I-J-K orientation, and wherein domain H has a second VL amino acid sequence, domain I has a human CL kappa amino acid sequence, and domain J has a human IgG1 (d) a fourth polypeptide chain having a domain L and a domain M, wherein the domains are arranged N-terminus to C-terminus in an LM orientation, and wherein domain L has a second VH amino acid sequence and domain M has a human IgG1 CH2 amino acid sequence, and K has a human IgG1 CH3 amino acid sequence with Y349C, D356E, L358M, T366S, L368A, and Y407V; (e) the first and second polypeptides associate through interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate through interactions between the H and L domains and between the I and M domains; (g) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form an ROR-binding molecule; (h) domain A and domain F form a first antigen-binding site specific for a first antigen; and (i) domain H and domain L form a second antigen-binding site specific for a second antigen.

[0339] In a preferred embodiment, the first polypeptide chain has the sequence of SEQ ID NO:24, the second polypeptide chain has the sequence of SEQ ID NO:25, the third polypeptide chain has the sequence of SEQ ID NO:10, and the fourth polypeptide chain has the sequence of SEQ ID NO:11. 6.4.4. Bivalent, bispecific B-Body "BC44"

[0340] 3 and 19, in one set of embodiments, the ROR-binding molecule has first, second, third, and fourth polypeptide chains, wherein (a) the first polypeptide chain comprises Domain A, Domain B, Domain D, and Domain E, wherein the domains are arranged N-terminus to C-terminus in an A-B-D-E orientation, and Domain A has a first VL amino acid sequence; Domain B has a human IgG1 CH3 amino acid sequence with a Y349C mutation, a P343V mutation, and a C-terminal extension incorporating a PGK tripeptide sequence followed by a DKTHT motif of the IgG1 hinge region; Domain D has a human IgG1 CH2 amino acid sequence; and Domain E has a human IgG1 CH2 amino acid sequence with a S354C mutation and a T366W mutation. (b) a second polypeptide chain having domain F and domain G, wherein the domains are arranged, from N-terminus to C-terminus, in an FG orientation, where domain F has a first VH amino acid sequence and domain G has a human IgG1 CH3 amino acid sequence with a S354C mutation and a C-terminal extension incorporating a PGK tripeptide sequence; (c) a third polypeptide chain having domain H, domain I, domain J, and domain K, wherein the domains are arranged, from N-terminus to C-terminus, in an H-I-J-K orientation, where domain H has a second VL amino acid sequence, domain I has a human CL kappa amino acid sequence, domain J has a human IgG1 CH2 amino acid sequence, and domain K has Y349C, T366S, L368A, and Y407V. (d) a fourth polypeptide chain has a domain L and a domain M, wherein the domains are arranged N-terminus to C-terminus in an LM orientation, and domain L has a second VH amino acid sequence and domain M has a human IgG1 amino acid sequence; (e) the first and second polypeptides associate through interactions between the A and F domains and between the B and G domains; (f) the third and fourth polypeptides associate through interactions between the H and L domains and between the I and M domains; (g) the first and third polypeptides associate through interactions between the D and J domains and between the E and K domains to form a ROR-binding molecule;(h) Domain A and Domain F form a first antigen-binding site specific for a first antigen; (i) Domain H and Domain L form a second antigen-binding site specific for a second antigen;

[0341] In a preferred embodiment, the first polypeptide chain has the sequence of SEQ ID NO: 32, the second polypeptide chain has the sequence of SEQ ID NO: 25, the third polypeptide chain has the sequence of SEQ ID NO: 10, and the fourth polypeptide chain has the sequence of SEQ ID NO: 11. 6.5. Specific Trivalent ROR-Binding Molecules 6.5.1. Trivalent 1x2 bispecific B-Body "BC28-1x2"

[0342] With reference to Section 6.4.3 and Figure 26, in one set of embodiments, the ROR-binding molecule further comprises a sixth polypeptide chain, wherein (a) the third polypeptide chain further comprises Domain R and Domain S, wherein the domains are arranged, from N-terminus to C-terminus, in a RSHIJK orientation, where Domain R has a first VL amino acid sequence and Domain S has a human IgG1 CH3 amino acid sequence with a Y349C mutation and a C-terminal extension incorporating a PGK tripeptide sequence followed by a GSGSGS linker peptide linking Domain S to Domain H; and (b) the ROR-binding molecule further comprises a sixth polypeptide chain, wherein the sixth polypeptide chain comprises Domain T and Domain U, wherein the domains are arranged, from N-terminus to C-terminus, in a TU orientation, where Domain T has the first VH amino acid sequence and Domain U has a S354C mutation and a C-terminal extension incorporating a PGK tripeptide sequence. (c) the third and sixth polypeptides associate through interactions between the R and T domains and between the S and U domains to form an ROR-binding molecule; and (d) the R and T domains form a third antigen-binding site specific for the first antigen.

[0343] In a preferred embodiment, the first polypeptide chain has the sequence of SEQ ID NO:24, the second polypeptide chain has the sequence of SEQ ID NO:25, the third polypeptide chain has the sequence of SEQ ID NO:37, the fourth polypeptide chain has the sequence of SEQ ID NO:11, and the sixth polypeptide chain has the sequence of SEQ ID NO:25. 6.5.2. Trivalent 1x2 trispecific B-Body "BC28-1x1x1a"

[0344] With reference to Section 6.4.3 and Figures 26 and 30, in one set of embodiments, the ROR-binding molecule further comprises a sixth polypeptide chain, wherein (a) the third polypeptide chain further comprises Domain R and Domain S, wherein the domains are arranged, from N-terminus to C-terminus, in a RSHIJK orientation, where Domain R has a third VL amino acid sequence, and Domain S has a human IgG1 CH3 amino acid sequence with a T366K mutation and a C-terminal extension incorporating a KSC tripeptide sequence followed by a GSGSGS linker peptide linking Domain S to Domain H; and (b) the ROR-binding molecule further comprises a sixth polypeptide chain, wherein the sixth polypeptide chain comprises Domain T and Domain U, wherein the domains are arranged, from N-terminus to C-terminus, in a TU orientation, where Domain T has a third VH amino acid sequence, and Domain U has a L351D mutation and a C-terminal extension incorporating a GEC amino acid disulfide motif. (c) the third and sixth polypeptides associate through interactions between the R and T domains and between the S and U domains to form an ROR-binding molecule; and (d) the R and T domains form a third antigen-binding site specific for a third antigen.

[0345] In a preferred embodiment, the first polypeptide chain has the sequence of SEQ ID NO:24, the second polypeptide chain has the sequence of SEQ ID NO:25, the third polypeptide chain has the sequence of SEQ ID NO:45, the fourth polypeptide chain has the sequence of SEQ ID NO:11, and the sixth polypeptide chain has the sequence of SEQ ID NO:53. 6.6. Other ROR-binding molecular platforms

[0346] The various antibody platforms described above are not limiting. The antigen binding site described herein, including specific CDR subsets, can be formatted into any binding molecule platform, including but not limited to, full-length antibodies, Fab fragments, Fvs, scFvs, tandem scFvs, diabodies, scdiabodies, DARTs, tandAbs, minibodies, camelid VHHs, and other antibody fragments or formats known to those skilled in the art. Exemplary antibody and antibody fragment formats are described in detail in Brinkmann et al. (MABS, 2017, Vol. 9, No. 2, 182-212), which is incorporated herein by reference for all its teachings. 6.7. Antigen specificity

[0347] In addition to ROR antigens, other antigens to which the ROR-binding molecules described herein can specifically bind can be selected from a wide variety of molecular targets. For example, the antigen binding site(s) may be E-Cad, CLDN7, FGFR2b, N-Cad, Cad-11, FGFR2c, ERBB2, ERBB3, FGFR1, FOLR1, IGF-Ira, GLP1R, PDGFRa, PDGFRb, EPHB6, ABCG2, CXCR4, CXCR7, integrin-avb3, SPARC, VCAM, ICAM, annexin, ROR1, ROR2, TNFα, CD137, angiopoietin 2, angiopoietin 3, BAFF, beta amyloid, C5, CA-125, CD147, CD125, CD147, CD152, CD19, CD20, CD22, CD23, CD24, CD25, CD274, CD28, CD3, CD30, CD33, CD37, CD4, CD 40, CD44, CD44v4, CD44v6, CD44v7, CD50, CD51, CD52, CEA, CSF1R, CTLA-2, DLL4, EGFR, EPCAM, HER3, GD2 ganglioside, GDF-8, Her2 / neu, CD2221, IL-17A, IL-12, IL-23, IL-13, IL-6, IL-23, integrin, CD11a, MUC1, Notch, TAG-72, TGFβ, TRAIL-R2, VEGF-A, VEGFR-1, VEGFR2, VEGFc, hematopoietin (four-helix bundle), EPO (erythropoietin), IL-2 (T-cell growth factor), IL-3 (multi-colony CSF), IL-4 (BCGF-1, BSF-1), IL-5 (BCGF-2), IL-6 Interferons include IL-4 (IFN-β2, BSF-2, BCDF), IL-7, IL-8, IL-9, IL-11, IL-13 (P600), G-CSF, IL-15 (T-cell growth factor), GM-CSF (granulocyte-macrophage colony-stimulating factor), OSM (OM, oncostatin M), and LIF (leukemia inhibitory factor); interferons include IFN-γ, IFN-α, and IFN-β; and immunoglobulin superfamily members such as B7.1 (CD80) and B7.2 (B70, CD86), etc.); TNF family members (such as TNF-α (cachectin), TNF-β (lymphotoxin, LT, LT-α), LT-β, ​​Fas, CD27, CD30, and 4-1BBL); and those not classified as part of a particular family (such as TGF-β, IL-1α, IL-1β, IL-1RA, IL-10 (cytokine synthesis inhibitory factor F), IL-12 (NK cell stimulator), MIF, IL-16, IL-17 (mCTLA-8), and / or IL-18 (IGIF, interferon-γ inducing factor)); in bispecific antibody embodiments, the antibody may bind, for example, to two of these targets. Furthermore, the Fc portion of the heavy chain of the antibody may be used to target Fc receptor-expressing cells, such as using the Fc portion of an IgE antibody to target mast cells and basophils.

[0348] In addition to ROR antigens, other antigens to which the ROR-binding molecules described herein can specifically bind can be selected, including TNFR1 (also known as CD120a and TNFRSF1A), TNFR2 (also known as CD120b and TNFRSF1B), TNFRSF3 (also known as LTβR), TNFRSF4 (also known as OX40 and CD134), TNFRSF5 (also known as CD40), TNFRSF6 (also known as FAS and CD95), TNFRSF7 (also known as TNFRSF9), TNFRSF8 (also known as TNFRSF1A), TNFRSF9 (also known as TNFRSF1B), TNFRSF10 (also known as TNFRSF1C), TNFRSF11 (also known as TNFRSF1D), TNFRSF12 (also known as TNFRSF1F), TNFRSF13 (also known as TNFRSF1G), TNFRSF15 (also known as TNFRSF1H), TNFRSF16 (also known as TNFRSF1H), TNFRSF17 (also known as TNFRSF1H), TNFRSF18 (also known as TNFRSF1F), TNFRSF19 (also known as TNFRSF1H), TNFRSF19 ...F), TNFRSF19 (also known as TNFRSF1 FRSF6B (also known as DCR3), TNFRSF7 (also known as CD27), TNFRSF8 (also known as CD30), TNFRSF9 (also known as 4-1BB), TNFRSF10A (also known as TRAILR1, DR4, and CD26), TNFRSF10B (also known as TRAILR2, DR5, and CD262), TNFRSF10C (also known as TRAILR3, DCR1, CD263), TNFRSF10D (TRAILR4, DCR2, and CD26 4), TNFRSF11A (also known as RANK and CD265), TNFRSF11B (also known as OPG), TNFRSF12A (also known as FN14, TWEAKR, and CD266), TNFRSF13B (also known as TACI and CD267), TNFRSF13C (also known as BAFFR, BR3, and CD268), TNFRSF14 (also known as HVEM and CD270), TNFRSF16 (NGFR, p75NTR, and CD27 1), or TNFRSF17 (also known as BCMA and CD269), TNFRSF18 (also known as GITR and CD357), TNFRSF19 (also known as TROY, TAJ, and TRADE), TNFRSF21 (also known as CD358), TNFRSF25 (also known as Apo-3, TRAMP, LARD, or WS-1), EDA2R (also known as XEDAR).

[0349] In addition to ROR antigens, other antigens to which the ROR-binding molecules described herein can specifically bind can be selected from immuno-oncology targets, including, but not limited to, checkpoint inhibitor targets such as PD1, PDL1, CTLA-4, PDL2, B7-H3, B7-H4, BTLA, TIM3, GAL9, LAG3, VISTA, KIR, 2B4, BY55, and CGEN-15049. 6.8. Further Modifications

[0350] In a further set of embodiments, the ROR-binding molecule has additional modifications. 6.8.1. Antibody-Drug Conjugates

[0351] In various embodiments, the ROR-binding molecule is conjugated to a therapeutic agent (e.g., a drug) to form a ROR-binding molecule-drug conjugate. Therapeutic agents include, but are not limited to, chemotherapeutic agents, imaging agents (e.g., radioisotopes), immunomodulatory agents (e.g., cytokines, chemokines, or checkpoint inhibitors), and toxins (e.g., cytotoxic agents). In certain embodiments, the therapeutic agent is linked to the ROR-binding molecule via a linker peptide, as discussed in more detail in Section 6.8.3 below.

[0352] Methods for preparing antibody-drug conjugates (ADCs) that may be adapted for conjugating drugs to the ROR-binding molecules disclosed herein are described, for example, in U.S. Pat. No. 8,624,003 (the Pott method), U.S. Pat. No. 8,163,888 (the one-step method), U.S. Pat. No. 5,208,020 (the two-step method), U.S. Pat. No. 8,337,856, U.S. Pat. No. 5,773,001, U.S. Pat. No. 7,829,531, U.S. Pat. No. 5,208,020, U.S. Pat. No. 7,745,394, WO2017 / 136623, WO2017 / 015502, WO2017 / 015496, WO2017 / 015495, WO2004 / 01 0957, WO2005 / 077090, WO2005 / 082023, WO2006 / 065533, WO2007 / 030642, WO2007 / 103288, WO20 13 / 173337, WO2015 / 057699, WO2015 / 095755, WO2015 / 123679, WO2015 / 157286, WO2017 / 165851, WO2009 / 073445, WO2010 / 068759, WO2010 / 138719, WO2012 / 171020, WO2014 / 008375, WO2014 / 09 3394, WO2014 / 093640, WO2014 / 160360, WO2015 / 054659, WO2015 / 195925, WO2017 / 160754, Storz (MAbs. 2015 Nov-Dec; 7(6): 989-1009), Lambert et al. (Adv Ther, 2017 34: 1015), Diamantis et al. (British Journal of Cancer, 2016, 114, 362-367), Carrico et al. Chem Biol., 2007. 3: 321-2), We et al. (Proc Natl Acad Sci USA, 2009. 106: 3000-5), Rabuka et al. (Curr Opin Chem Biol., 2011 14: 790-6), Hudak et al. (Angew Chem Int Ed Engl., 2012: 4161-5), Rabuka et al. (Nat Protoc., 2012 7:1052-67), Agarwal et al. (Proc Natl Acad Sci USA., 2013, 110: 46-51), Agarwal et al. (Bioconjugate Chem., 2013, 24: 846-851), Barfield et al. (Drug Dev. and D., 2014, 14:34-41), Drake et al. (Bioconjugate Chem., 2014, 25:1331-41), Liang et al. (J Am Chem Soc., 2014, 136:10850-3), Drake et al. (Curr Opin Chem Biol., 2015, 28:174-80) and York et. al. (BMC Biotechnology, 2016, 16(1):23). 6.8.2. Further Bond Substructures

[0353] In various embodiments, the ROR-binding molecule has a modification that includes one or more additional binding moieties. In certain embodiments, the binding moieties are antibody fragments or antibody formats, including, but not limited to, full-length antibodies, Fab fragments, Fvs, scFvs, tandem scFvs, diabodies, sc diabodies, DARTs, tandAbs, minibodies, camelid VHHs, and other antibody fragments or formats known to those skilled in the art. Exemplary antibody and antibody fragment formats are described in detail in Brinkmann et al. (MABS, 2017, Vol. 9, No. 2, 182-212), the entire teachings of which are incorporated herein by reference.

[0354] In certain embodiments, one or more additional binding moieties are linked to the C-terminus of the first or third polypeptide chain. In certain embodiments, one or more additional binding moieties are linked to the C-terminus of both the first and third polypeptide chains. In certain embodiments, one or more additional binding moieties are linked to the C-terminus of both the first and third polypeptide chains. In certain embodiments, individual portions of the one or more additional binding moieties are linked separately to the C-terminus of the first and third polypeptide chains to form functional binding moieties.

[0355] In certain embodiments, one or more additional binding moieties are linked to the N-terminus of any of the polypeptide chains (e.g., the first, second, third, fourth, fifth, or sixth polypeptide chain). In certain embodiments, individual portions of the additional binding moieties are separately linked to the N-terminus of different polypeptide chains to form functional binding moieties.

[0356] In certain embodiments, one or more additional binding moieties are specific for different antigens or epitopes of the ABS within the ROR-binding molecule. In certain embodiments, one or more additional binding moieties are specific for the same antigen or epitope of the ABS within the ROR-binding molecule. In certain embodiments, where the modification is two or more additional binding moieties, the additional binding moieties are specific for the same antigen or epitope. In certain embodiments, where the modification is two or more additional binding moieties, the additional binding moieties are specific for different antigens or epitopes.

[0357] In certain embodiments, one or more additional binding moieties are linked to the ROR-binding molecule using in vitro methods, including, but not limited to, reactive chemistry and affinity tag systems, as discussed in more detail below in Section 6.8.3. In certain embodiments, one or more additional binding moieties are linked to the ROR-binding molecule via Fc-mediated binding (e.g., Protein A / G). In certain embodiments, one or more additional binding moieties are linked to the ROR-binding molecule using recombinant DNA techniques, such as encoding the nucleotide sequences of fusion products of the ROR-binding molecule and the additional binding moieties in the same expression vector (e.g., a plasmid). 6.8.3. Functional Groups / Reactive Groups

[0358] In various embodiments, the ROR-binding molecules have modifications that include functional or chemically reactive groups that can be used in downstream processes, such as the attachment of additional moieties (e.g., drug conjugates and additional binding moieties, discussed in more detail above in Sections 6.8.1 and 6.8.2) and downstream purification processes.

[0359] In certain embodiments, the modification is a chemically reactive group, including, but not limited to, reactive thiols (e.g., maleimide-based reactive groups), reactive amines (e.g., N-hydroxysuccinimide-based reactive groups), "click chemistry" groups (e.g., reactive alkyne groups), and aldehydes bearing formylglycine (FGly). In certain embodiments, the modification is a functional group, including, but not limited to, affinity peptide sequences (e.g., HA, HIS, FLAG, GST, MBP, and the Strep system). In certain embodiments, the functional group or chemically reactive group comprises a cleavable peptide sequence. In certain embodiments, the cleavable peptide is cleaved by means including, but not limited to, photocleavage, chemical cleavage, protease cleavage, reducing conditions, and pH conditions. In certain embodiments, protease cleavage is performed by an intracellular protease. In certain embodiments, protease cleavage is performed by an extracellular or membrane-associated protease. ADC therapy employing protease cleavage is described in detail in Choi et al. (Theranostics, 2012; 2(2): 156-178. 6.8.4. Reduced Effector Function

[0360] In certain embodiments, the ROR-binding molecule has one or more engineered mutations in the amino acid sequence of the antibody domain that reduce effector functions naturally associated with antibody binding, including, but not limited to, cellular functions resulting from Fc receptor binding to the Fc portion of an antibody, such as antibody-dependent cellular cytotoxicity (ADCC, also known as antibody-dependent cell-mediated cytotoxicity), complement fixation (e.g., C1q binding), antibody-dependent cell-mediated phagocytosis (ADCP), and opsonization. Engineered mutations that reduce effector function are described in more detail in U.S. Patent Application Publication No. 2017 / 0137530, Armour, et al. (Eur. J. Immunol. 29(8) (1999) 2613-2624), Shields, et al. (J. Biol. Chem. 276(9) (2001) 6591-6604), and Oganesyan, et al. (Acta Cristallographica D64 (2008) 700-704), each of which is incorporated by reference in its entirety.

[0361] In certain embodiments, the ROR-binding molecule has one or more engineered mutations in the amino acid sequence of the antibody domain that reduce binding of the Fc portion of the ROR-binding molecule to an FcR receptor. In some embodiments, the FcR receptor is an FcRγ receptor. In certain embodiments, the FcR receptor is an FcγRIIa and / or FcγRIIIA receptor.

[0362] In certain embodiments, the one or more engineered mutations that reduce effector function are mutations in the CH2 domain of the antibody. In various embodiments, the one or more engineered mutations are at positions L234 and L235 in the CH2 domain. In certain embodiments, the one or more engineered mutations are L234A and L235A in the CH2 domain. In other embodiments, the one or more engineered mutations are at positions L234, L235, and P329 in the CH2 domain. In certain embodiments, the one or more engineered mutations are L234A, L235A, and P329G in the CH2 domain. In a preferred embodiment, the one or more engineered mutations are L234A, L235A, and P329K in the CH2 domain. 6.9. Purification method

[0363] Provided herein are methods for purifying ROR-binding molecules that include the B-body platform.

[0364] In one series of embodiments, the method includes: i) contacting a sample containing an ROR-binding molecule with a CH1-binding reagent, wherein the ROR-binding molecule comprises at least a first, second, third, and fourth polypeptide chain associated in a complex, the complex comprising at least one CH1 domain or portion thereof, the number of CH1 domains in the complex being at least one less than the valency of the complex, and the contacting is performed under conditions sufficient for the CH1-binding reagent to bind to the CH1 domain or portion thereof; and ii) purifying the complex from one or more incomplete complexes, wherein the incomplete complex does not comprise the first, second, third, and fourth polypeptide chains.

[0365] In a typical naturally occurring antibody, two heavy chains assemble, each having a CH1 domain as the second domain, numbered N- to C-terminus. Thus, a typical antibody has two CH1 domains. CH1 domains are described in more detail in Section 6.3.8.1. In various ROR-binding molecules described herein, the CH1 domain typically found in a protein is replaced with another domain, effectively reducing the number of CH1 domains in the protein. In a non-limiting illustrative example, the CH1 domain of a typical antibody can be replaced with a CH3 domain to generate an antigen-binding protein having only a single CH1 domain.

[0366] ROR-binding molecules can also refer to molecules based on engineered antibody architecture such that they no longer possess typical antibody architecture. For example, antibodies can be extended at their N- or C-termini to increase the valency of the antigen-binding protein (described in more detail in Section 6.3.13.1), and in certain instances, the number of CH1 domains is also increased beyond the typical two CH1 domains. Such molecules can also have one or more of their CH1 domains replaced so that the number of CH1 domains in the protein is at least one less than the valency of the antigen-binding protein. In some embodiments, the number of CH1 domains replaced by other domains generates an ROR-binding molecule with only a single CH1 domain. In other embodiments, the number of CH1 domains replaced by other domains is two or more, but generates an ROR-binding molecule with at least one less CH1 domain than the valency of the antigen-binding protein. In certain embodiments, when an ROR-binding molecule has two or more CH1 domains, the multiple CH1 domains can all be present on the same polypeptide chain. In other specific embodiments, when an ROR-binding molecule has two or more CH1 domains, the multiple CH1 domains can be a single CH1 domain in multiple copies of the same polypeptide chain present in the intact complex. 6.9.1. CH1 Coupling Reagent

[0367] In an exemplary, non-limiting method for purifying ROR-binding molecules, a sample containing the ROR-binding molecule is contacted with a CH1-binding reagent. The CH1-binding reagents described herein can be any molecule that specifically binds to a CH1 epitope. Various CH1 sequences that provide CH1 epitopes are described in more detail in Section 6.3.8.1, and specific binding is described in more detail in Section 6.3.13.1.

[0368] In some embodiments, the CH1-binding reagent is derived from an immunoglobulin protein and has an antigen-binding site (ABS) that specifically binds to a CH1 epitope. In certain embodiments, the CH1-binding reagent is an antibody, also referred to as an "anti-CH1 antibody." The anti-CH1 antibody can be derived from various species. In certain embodiments, the anti-CH1 antibody is a mammalian antibody, including, but not limited to, mouse, rat, hamster, rabbit, camel, donkey, goat, and human antibodies. In certain embodiments, the anti-CH1 antibody is a single-domain antibody. The single-domain antibodies described herein have a single variable domain that forms an ABS and specifically binds to a CH1 epitope. Exemplary single-domain antibodies include, but are not limited to, heavy chain antibodies derived from camels and sharks, which are described in more detail in International Application WO 2009 / 011572, the entire teachings of which are incorporated herein by reference. In a preferred embodiment, the anti-CH1 antibody is an antibody derived from a camel (also referred to as a "camelid antibody"). Exemplary camelid antibodies include, but are not limited to, human IgG-CH1 CaptureSelect™ (ThermoFisher, #194320010) and human IgA-CH1 (ThermoFisher, #194311010). In some embodiments, the anti-CH1 antibody is a monoclonal antibody. Monoclonal antibodies are typically produced from cultured antibody-producing cell lines. In other embodiments, the anti-CH1 antibody is a polyclonal antibody, i.e., a population of different anti-CH1 antibodies, each recognizing a CH1 epitope. Polyclonal antibodies are typically produced by collecting antibody-containing sera from animals immunized with an antigen of interest or a fragment thereof, in this case CH1.

[0369] In some embodiments, the CH1 binding reagent is a molecule that is not derived from an immunoglobulin protein. Examples of such molecules include, but are not limited to, aptamers, peptoids, and affibodies, as described in detail by Perret and Boschetti (Biochimie, February 2018, Vol. 145:98-112). 6.9.2. Solid support

[0370] In an exemplary, non-limiting method for purifying ROR-binding molecules, the CH1-binding reagent can be attached to a solid support in various embodiments of the present invention. A solid support, as described herein, refers to a material to which other entities, such as a CH1-binding reagent, can be attached or immobilized. Solid supports, also referred to as "carriers," are described in more detail in International Application WO2009 / 011572.

[0371] In certain embodiments, the solid support comprises beads or nanoparticles. Examples of beads and nanoparticles include, but are not limited to, agarose beads, polystyrene beads, magnetic nanoparticles (e.g., Dynabeads™, ThermoFisher), polymers (e.g., dextran), synthetic polymers (e.g., Sepharose™), or any other material suitable for attachment of a CH1 binding reagent. In certain embodiments, the solid support is modified to allow attachment of a CH1 binding reagent. Examples of solid support modifications include, but are not limited to, chemical modifications that form covalent bonds with proteins (e.g., activated aldehyde groups) and modifications that specifically pair with the cognate modification of a CH1 binding reagent (e.g., "click chemistry" modifications such as biotin-streptavidin pairing, disulfide linkages, polyhistidine-nickel, or azido-alkynyl pairing).

[0372] In certain embodiments, prior to contacting the CH1-binding reagent with the ROR-binding molecule, the CH1-binding reagent is attached to a solid support, also referred to herein as an "anti-CH1 resin." In some embodiments, the anti-CH1 resin is dispersed in a solution. In other embodiments, the anti-CH1 resin is "packed" into a column. The anti-CH1 resin is then contacted with the ROR-binding molecule, and the CH1-binding reagent specifically binds to the ROR-binding molecule.

[0373] In other embodiments, after the CH1-binding reagent contacts the ROR-binding molecule, the CH1-binding reagent is attached to a solid support. As a non-limiting example, a CH1-binding reagent having a biotin modification can be contacted with the ROR-binding molecule, and then the CH1-binding reagent / ROR-binding molecule mixture can be contacted with a streptavidin-modified solid support to attach the CH1-binding reagent, including the CH1-binding reagent specifically bound to the ROR-binding molecule, to the solid support.

[0374] In methods in which the CH1-binding reagent is attached to a solid support, in various embodiments, the bound ROR-binding molecule is released or "eluted" from the solid support to form an eluate containing the ROR-binding molecule. In some embodiments, the bound ROR-binding molecule is released by reversing the paired modification (e.g., reduction of a disulfide linkage), adding a reagent to compete with and remove the ROR-binding molecule (e.g., addition of imidazole to compete with polyhistidine for binding to nickel), cleaving and removing the ROR-binding molecule (e.g., the modification may include a cleavable moiety), or otherwise interfering with specific binding of the CH1-binding reagent to the ROR-binding molecule. Methods for interfering with specific binding include, but are not limited to, contacting the ROR-binding molecule bound to the CH1-binding reagent with a low pH solution. In a preferred embodiment, the low pH solution comprises 0.1 M acetic acid, pH 4.0. In other embodiments, the bound ROR-binding molecule can be contacted with a range of low pH solutions, i.e., a "gradient." 6.9.3. Further Purification

[0375] In some embodiments of the exemplary, non-limiting method, a single iteration of the method using contacting the ROR-binding molecule with a CH1-binding reagent followed by elution of the ROR-binding molecule is used to purify the ROR-binding molecule from one or more incomplete complexes. In certain embodiments, no other purification steps are performed. In other embodiments, one or more additional purification steps are performed to further purify the ROR-binding molecule from one or more incomplete complexes. The one or more additional purification steps include, but are not limited to, purifying the ROR-binding molecule based on other protein characteristics, such as size (e.g., size exclusion chromatography), charge (e.g., ion exchange chromatography), or hydrophobicity (e.g., hydrophobic interaction chromatography). In a preferred embodiment, additional cation exchange chromatography is performed. Furthermore, the ROR-binding molecule can be further purified by repeating the contacting of the ROR-binding molecule with the CH1-binding reagent described above and modifying the CH1 purification method between iterations, for example, using an elution step for the first iteration and gradient elution for subsequent elutions. 6.9.4. Complex Assembly and Purity

[0376] In an embodiment of the present invention, at least four distinct polypeptide chains associate together to form a complete complex, e.g., an ROR-binding molecule. However, incomplete complexes may also be formed that do not contain at least four distinct polypeptide chains. For example, an incomplete complex may be formed that has only one, two, or three of the polypeptide chains. In other examples, an incomplete complex may contain four or more polypeptide chains but not at least four distinct polypeptide chains, e.g., the incomplete complex is improperly associated with two or more copies of the distinct polypeptide chains. The methods of the present invention purify a complex, e.g., a fully assembled ROR-binding molecule, from an incomplete complex.

[0377] Methods for evaluating the effectiveness and efficiency of purification steps are well known to those skilled in the art, including, but not limited to, SDS-PAGE analysis, ion-exchange chromatography, size-exclusion chromatography, and mass spectrometry. Purity can also be evaluated according to various criteria. Exemplary criteria include, but are not limited to, 1) evaluating the percentage of total protein in the eluate provided by fully assembled ROR-binding molecules; 2) evaluating the fold enrichment or percent increase of a method for purifying a desired product, e.g., comparing the total protein provided by fully assembled ROR-binding molecules in the eluate with that in the starting sample; and 3) evaluating the percentage or percent reduction of undesired products, such as the incomplete complexes described above, including determining the percentage or percent reduction of specific undesired products (e.g., unassociated single polypeptide chains, dimers of any combination of polypeptide chains, or trimers of any combination of polypeptide chains). Purity can be evaluated after any combination of the methods described herein. For example, purity can be assessed after a single iteration using an anti-CH1 binding reagent, as described herein, or after an additional purification step, as described in more detail in Section 6.9.3. The effectiveness and efficiency of the purification step can also be used to compare the described method using an anti-CH1 binding reagent to other purification methods known to those skilled in the art, such as Protein A purification. 6.10. Manufacturing method

[0378] The ROR-binding molecules described herein can be readily produced by expression using standard cell-free translation, transient transfection, and stable transfection approaches currently used for antibody production. In certain embodiments, Expi293 cells (ThermoFisher) can be used to produce ROR-binding molecules using ThermoFisher protocols and reagents such as ExpiFectamine, or other reagents known to those skilled in the art, such as polyethyleneimine, as described in detail in Fang et al. (Biological Procedures Online, 2017, 19:11), which is incorporated herein by reference for all its teachings.

[0379] As further described in the Examples below, expressed proteins can be easily separated from unwanted proteins and protein complexes using a CH1 affinity resin, such as CaptureSelect CH1 resin, and protocols provided by ThermoFisher. Other purification strategies include, but are not limited to, the use of Protein A, Protein G, or Protein A / G reagents. Ion exchange chromatography, as routinely used in the art, can be used to effect further purification. Pharmaceutical Compositions

[0380] In another aspect, there is provided a pharmaceutical composition comprising a ROR-binding molecule described herein and a pharmaceutically acceptable carrier or diluent. In typical embodiments, the pharmaceutical composition is sterile.

[0381] In various embodiments, the pharmaceutical composition comprises the ROR-binding molecule at a concentration of 0.1 mg / ml to 100 mg / ml. In specific embodiments, the pharmaceutical composition comprises the ROR-binding molecule at a concentration of 0.5 mg / ml, 1 mg / ml, 1.5 mg / ml, 2 mg / ml, 2.5 mg / ml, 5 mg / ml, 7.5 mg / ml, or 10 mg / ml. In some embodiments, the pharmaceutical composition comprises the ROR-binding molecule at a concentration greater than 10 mg / ml. In certain embodiments, the ROR-binding molecule is present at a concentration of 20 mg / ml, 25 mg / ml, 30 mg / ml, 35 mg / ml, 40 mg / ml, 45 mg / ml, or even 50 mg / ml or higher. In certain embodiments, the ROR-binding molecule is present at a concentration greater than 50 mg / ml.

[0382] In various embodiments, pharmaceutical compositions are described in more detail in U.S. Pat. Nos. 8,961,964, 8,945,865, 8,420,081, 6,685,940, 6,171,586, 8,821,865, 9,216,219, U.S. Patent Application No. 10 / 813,483, WO2014 / 066468, WO2011 / 104381, and WO2016 / 180941, each of which is incorporated herein in its entirety. 6.12. Treatment Method

[0383] In another aspect, a method of treatment is provided, comprising administering to a subject a ROR-binding molecule (e.g., an antibody) described herein in an amount effective to treat the subject. Such ROR antigen-binding molecules are useful in treating ROR-expressing cancers, including cancers that express the ROR1 antigen, cancers that express the ROR2 antigen, and / or cancers that express both the ROR1 and ROR2 antigens.

[0384] In some embodiments, the antibodies of the present disclosure may be used to treat a variety of cancers, including bladder, blood (myelocytic leukemia [acute and chronic], acute lymphocytic leukemia, chronic lymphocytic leukemia, myeloproliferative disorders, multiple myeloma, myelodysplastic syndromes), bone, bone marrow, brain (astrocytoma, medulloblastoma, glioma, ependymoma, germ cell tumor [pinealoma], glioblastoma multiforme, oligodendroglioma, schwannoma, retinoblastoma, congenital tumors), breast, colon, esophagus (squamous cell carcinoma, adenocarcinoma, leiomyosarcoma, lymphoma, ), gastrointestinal tract, gums, head, kidney (adenocarcinoma, Wilms' tumor [nephroblastoma], lymphoma, leukemia, renal cell carcinoma), liver, lung, nasopharynx, neck, ovary, prostate (adenocarcinoma, sarcoma, castration-resistant prostate cancer), skin, stomach (carcinoma, lymphoma, leiomyosarcoma), testicle (seminoma, teratoma, embryonal carcinoma, teratocarcinoma, choriocarcinoma, sarcoma, stromal cell carcinoma, fibroma, fibroadenoma, adenomatous tumor, lipoma), tongue, or uterus. In some embodiments, the cancer is selected from the group consisting of neoplasms, malignancies, carcinomas, undifferentiated carcinomas, giant cell spindle cell carcinomas, small cell carcinomas, papillary carcinomas, squamous cell carcinomas, lymphomas, basal cell carcinomas, pilomatrix carcinomas, transitional cell carcinomas, papillary transitional cell carcinomas, adenocarcinomas, malignant gastrinomas, cholangiocarcinomas, hepatocellular carcinomas, mixed hepatocellular and cholangiocarcinomas, trabecular adenocarcinomas, adenoid cystic carcinomas, adenocarcinomas of adenomatous polyposis, adenomatous familial polyposis coli, solid tumors, carcinoid malignancies, bronchiolo-alveolar adenocarcinomas, papillary adenocarcinomas, chromophobe carcinomas, acidophil carcinomas, acidophil adenocarcinomas, basophilomas, clear cell adenocarcinomas, granular cell carcinomas, follicular adenocarcinomas, papillary and follicular adenocarcinomas, nonencapsulating sclerosing carcinomas, carcinoma); adrenocortical carcinoma; endometrial carcinoma; skin adnexal carcinoma; apocrine gland carcinoma; sebaceous gland carcinoma; earwax gland carcinoma; mucoepidermoid carcinoma; cystadenocarcinoma; papillary cystadenocarcinoma; papillary serous cystadenocarcinoma; mucinous cystadenocarcinoma; mucinous adenocarcinoma; signet ring cell carcinoma; duct carcinoma infiltration; medullary carcinoma; lobular carcinoma; inflammatory tumor; Paget's disease of the breast; acinic cell carcinoma; adenosquamous carcinoma; adenocarcinoma with squamous metaplasia; malignant thymoma; malignant ovarian stromal tumor; malignant thecoma; malignant granular cell tumor; malignant androgen cell tumor; Sertoli cell carcinoma; malignant Leydig cell tumor; malignant lipid cell tumor; malignant paraganglioma; malignant extramammary paraganglioma; pheochromocytoma; glomerular sarcoma; malignant melanoma;Amelanotic melanoma; superficial spreading melanoma; malignant melanoma of giant pigmented nevus (malig melanoma); epithelioid cell melanoma; malignant blue nevus; sarcomas (angiosarcoma, fibrosarcoma, rhabdomyosarcoma, liposarcoma); fibrosarcoma; malignant fibrous histiocytoma; myxosarcoma; liposarcoma; leiomyosarcoma; rhabdomyosarcoma; embryonal rhabdomyosarcoma; alveolar rhabdomyosarcoma; stromal sarcoma; malignant mixed tumor; mixed Müllerian tumor; nephroblastoma; hepatoblastoma; carcinosarcoma; malignant mesenchymoma; malignant Brenner tumor; malignant phyllodes tumor; synovial sarcoma; malignant mesothelioma; dysgerminoma; embryonal carcinoma; malignant teratoma; malignant ovarian goiter; choriocarcinoma; malignant mesononephroma; angiosarcoma; malignant hemangioendothelioma; Kaposi's sarcoma; Malignant hemangiopericytoma; Lymphangiosarcoma; Osteosarcoma; Parosteal osteosarcoma; Chondrosarcoma; Malignant chondroblastoma; Mesenchymal chondrosarcoma; Giant cell tumor of bone; Ewing's sarcoma; Malignant odontogenic tumor; Ameloblastic odontosarcoma; Malignant ameloblastoma; Ameloblastic fibrosarcoma; Malignant pinealoma; Chordoma; Malignant glioma; Ependymoma; Astrocytoma; Protoplasmic astrocytoma; Fibrous astrocytoma; Astroblastoma; Glioblastoma; Oligodendroglioma; Primitive neuroectodermal; Cerebellar sarcoma; Ganglion; Neuroblastoma; Retinoblastoma; Olfactory neurogenic tumor; Malignant meningioma; Neurofibrosarcoma; Malignant schwannoma; Malignant Granular cell tumor; Malignant lymphoma (reticulum cell sarcoma); Hodgkin's disease; Hodgkin's; Paragranuloma; Malignant small lymphocytic lymphoma; Malignant diffuse large cell lymphoma; Malignant follicular lymphoma; Mycosis fungoides; Other specified non-Hodgkin's lymphoma; Malignant histiocytosis; Multiple myeloma; Mast cell sarcoma; Immunoproliferative small intestinal disease; Leukemia; Lymphocytic leukemia; Plasma cell leukemia; Erythroleukemia; Lymphosarcoma cell leukemia; Myeloid leukemia; Basophilic leukemia; Eosinophilic leukemia; Monocytic leukemia; Mast cell leukemia; Megakaryoblastic leukemia; Granulocytic sarcoma; Hairy cell leukemia; Mucinous tumors; rhabdomyoma; fibroma; squamous cell carcinoma of the head and neck; laryngeal and hypopharyngeal cancer; nasal cavity and paranasal sinus cancer; nasopharyngeal cancer; salivary gland cancer; oral cavity; oropharyngeal cancer; bronchogenic carcinoma (squamous, undifferentiated small cell, undifferentiated large cell, adenocarcinoma, non-small cell lung cancer); alveolar (bronchiolar) carcinoma; bronchial adenoma; chondroitinous hamartoma; colorectal cancer; gastrointestinal stromal tumor; carcinoid; Turcot's syndrome; gastric cancer; adenocarcinoma of the esophagogastric junction; pancreatic (ductal adenocarcinoma, insulinoma, glucagonoma, gastrinoma, carcinoid tumor, vipoma);Small intestine (adenocarcinoma, lymphoma, carcinoid tumor, Kaposi's sarcoma, leiomyoma, hemangioma, lipoma, neurofibroma, fibroma); large intestine (adenocarcinoma, tubular adenoma, villous adenoma, hamartoma, leiomyoma); metastatic breast cancer; ductal carcinoma in situ; invasive ductal carcinoma; tubular carcinoma; mucinous carcinoma; lobular carcinoma in situ; triple-negative breast cancer; bladder and urethra (squamous cell carcinoma, transitional cell carcinoma, adenocarcinoma, urothelial carcinoma); clear cell carcinoma; hepatoma (hepatocellular carcinoma); angiosarcoma; hepatocellular adenoma; hemangioma; osteogenic sarcoma (osteosarcoma); malignant fibrous histiocytoma; malignant giant cell tumor; chordoma; osteochondroma (osteocartilaginous exostosis) exostoses); benign chondroma; chondromyxofibroma; osteoid; giant cell tumor; medullary thyroid carcinoma; differentiated thyroid carcinoma; papillary thyroid carcinoma; follicular thyroid carcinoma; Hurthle cell carcinoma; anaplastic thyroid carcinoma; skull (osteoma, hemangioma, granuloma, xanthomas, osteitis deformans); meninges (meningiomas, meningiosarcomas, gliomatosis); spinal cord (neurofibromas, meningiomas, gliomas, sarcomas); uterus (clear cell); cervix (cervical cancer, pre-neoplastic cervical dysplasia); ovary (ovarian cancer [serous cystadenocarcinoma, mucinous cystadenocarcinoma, unclassified carcinoma], granulosa theca cell tumor, Sertoli-Leydig cell tumor, dysgerminoma, malignant teratoma); vulva (squamous cell carcinoma, carcinoma in situ, adenocarcinoma, fibrosarcoma, melanoma); vagina (clear cell carcinoma, squamous cell carcinoma); botryoid sarcoma (embryonal rhabdomyosarcoma); fallopian tube (carcinoma); non-Hodgkin's lymphoma [malignant lymphoma]; Kaposi's sarcoma; dysplastic nevus lentigo (moles dysplastic nevi); hemangioma; dermatofibroma; keloid; psoriasis; neuroblastoma; adrenocortical carcinoma; pheochromocytoma; paraganglioma; Merkel cell carcinoma; pancreatic neuroendocrine and carcinoid tumors; neuroendocrine tumors; carcinoid tumors; pancreatic cancer; gastroesophageal; renal clear cell carcinoma; and primary peritoneal cancer.

[0385] The antibodies of the present disclosure can be administered to a subject, either per se or in the form of a pharmaceutical composition, for example, to treat cancer, autoimmunity, transplant rejection, immune responses after trauma, graft-versus-host disease, ischemia, stroke, and infectious diseases (e.g., by targeting viral antigens such as gp120 of HIV).

[0386] In another embodiment, the ROR-binding molecules (e.g., antibodies) described herein can be used in combination with one or more additional therapies to treat a subject with cancer. Additional therapies that can be used in combination with the ROR antigen-binding molecules (e.g., antibodies) described herein include, but are not limited to, (i) surgery; (ii) radiation therapy; (iii) endocrine therapy; (iv) immunotherapy (including adjuvant therapy and cell therapy, such as CAR T-cell therapy); and (v) chemotherapy, including cytotoxic and chemotherapeutic agents.

[0387] Any therapeutic agent active against cancer can be used in combination with the ROR antigen-binding molecules (e.g., antibodies) provided herein. Examples of such agents for cancer treatment can be found, for example, at https: / / www.cancer.gov / about-cancer / treatment / drugs (last accessed January 22, 2019) and in Cancer Principles and Practice of Oncology by VT Devita and S. Hellman (editors), 11 th edition (2018), Lippincott These can be found in public sources such as Williams & Wilkins Publishers. A person of ordinary skill in the art would be able to discern which combinations of agents would be useful based on the particular characteristics of the drugs and the type of cancer involved.

[0388] In certain embodiments, the additional treatment is radiation therapy, including, for example, gamma irradiation, neutron beam radiation therapy, electron beam radiation therapy, proton radiation therapy, brachytherapy, and systemic radioisotopes. Radiation therapy can include irradiation or the associated administration of radiopharmaceuticals. The radiation source can be either external or internal to the subject being treated (radiation treatment can be, for example, in the form of external beam radiation therapy (EBRT) or brachytherapy (BT)). Exemplary radioactive elements include, for example, radium, cesium-137, iridium-192, americium-241, gold-198, cobalt-57, copper-67, technetium-99, iodine-123, iodine-131, and indium-111.

[0389] In certain embodiments, the additional treatment is immunotherapy. Immunotherapy (also called biological response modifier therapy, biological therapy, biotherapy, immunotherapy, or biological therapy) is a treatment that uses parts of the immune system to fight disease. Immunotherapy can help the immune system recognize cancer cells or enhance the response to cancer cells. Immunotherapy includes active and passive immunotherapy. Active immunotherapy, including immunotherapeutic agents, stimulates the body's own immune system (e.g., vaccines), while passive immunotherapy, including immunotherapeutic agents, generally uses immune system components (e.g., antibodies) produced outside the body, antibodies conjugated with drugs, toxins, or radionuclides, and targeted therapeutic agents.

[0390] Exemplary immunotherapeutic agents include immune checkpoint inhibitors. In some embodiments, the immune checkpoint inhibitors used in the treatment methods can completely or partially reduce, inhibit, interfere with, or modulate one or more checkpoint proteins that regulate T cell activation or function. Many checkpoint proteins are known, such as CTLA-4 and its ligands CD80 and CD86; and PD-1 and its ligands PD-L1 and PD-L2 (Pardoll, Nature Reviews Cancer, 2012, 12, 252-264). Immune checkpoint inhibitors include or are derived from antibodies.

[0391] In certain embodiments, the checkpoint inhibitor is an OX40 (CD134) agonist. In some embodiments, the checkpoint inhibitor is an anti-OX40 antibody. In some embodiments, the anti-OX40 antibody is anti-OX-40. In some embodiments, the anti-OX40 antibody is MEDI6469.

[0392] In certain embodiments, the checkpoint inhibitor is a CD40 agonist. In some embodiments, the checkpoint inhibitor is an anti-CD40 antibody. In some embodiments, the anti-CD40 antibody is CF-870,893.

[0393] In certain embodiments, the checkpoint inhibitor is a CTLA-4 inhibitor. In some embodiments, the CTLA-4 inhibitor is an anti-CTLA-4 antibody. Examples of anti-CTLA4 antibodies include, but are not limited to, those described in U.S. Patent Nos. 5,811,097; 5,811,097; 5,855,887; 6,051,227; 6,207,157; ​​6,682,736; 6,984,720; and 7,605,238. In some embodiments, the anti-CTLA-4 antibody is tremelimumab (also known as ticilimumab or CP-675,206). In some embodiments, the anti-CTLA-4 antibody is ipilimumab (also known as MDX-010 or MDX-101). Ipilimumab is a fully human monoclonal IgG antibody that binds to CTLA-4. Ipilimumab is commercially available under the trade name Yervoy™.

[0394] In certain embodiments, the checkpoint inhibitor is a PD-1 / PD-L1 inhibitor. Examples of PD-1 / PD-L1 inhibitors include, but are not limited to, those described in U.S. Patent Nos. 7,488,802; 7,943,743; 8,008,449; 8,168,757; and 8,217,149, and PCT Patent Application Publication Nos. WO2003042402, WO2008156712, WO2010089411, WO2010036959, WO2011066342, WO2011159877, WO2011082400, and WO2011161699.

[0395] In certain embodiments, the checkpoint inhibitor is a PD-1 inhibitor. In some embodiments, the PD-1 inhibitor is an anti-PD-1 antibody. In some embodiments, the anti-PD-1 antibody is BGB-A317, nivolumab (also known as ONO-4538, BMS-936558, or MDX1106), or pembrolizumab (also known as MK-3475, SCH900475, or lambrolizumab). In some embodiments, the anti-PD-1 antibody is nivolumab. Nivolumab is a human IgG4 anti-PD-1 monoclonal antibody and is commercially available under the trade name Opdivo™. In some embodiments, the anti-PD-1 antibody is pembrolizumab. Pembrolizumab is a humanized monoclonal IgG4 antibody and is commercially available under the trade name Keytruda™. In some embodiments, the anti-PD-1 antibody is the humanized antibody CT-011. CT-011 administered alone failed to demonstrate a response in the treatment of acute myeloid leukemia (AML) at relapse. In some embodiments, the anti-PD-1 antibody is the fusion protein AMP-224. In some embodiments, the PD-1 antibody is BGB-A317. BGB-A317 is a monoclonal antibody specifically engineered to eliminate its ability to bind to Fc gamma receptor I, and has a unique binding signature for PD-1 with high affinity and excellent target specificity.

[0396] In certain embodiments, the checkpoint inhibitor is a PD-L1 inhibitor. In one embodiment, the PD-L1 inhibitor is an anti-PD-L1 antibody. In some embodiments, the anti-PD-L1 antibody is MEDI4736 (durvalumab). In some embodiments, the anti-PD-L1 antibody is BMS-936559 (also known as MDX-1105-01). In some embodiments, the PD-L1 inhibitor is atezolizumab (also known as MPDL3280A and Tecentriq®).

[0397] In certain embodiments, the checkpoint inhibitor is a PD-L2 inhibitor. In some embodiments, the PD-L2 inhibitor is an anti-PD-L2 antibody. In some embodiments, the anti-PD-L2 antibody is rHIgM12B7A.

[0398] In certain embodiments, the checkpoint inhibitor is a lymphocyte activation gene-3 (LAG-3) inhibitor. In some embodiments, the LAG-3 inhibitor is IMP321, a soluble Ig fusion protein (Brignone et al., J. Immunol., 2007, 179, 4202-4211). In some embodiments, the LAG-3 inhibitor is BMS-986016.

[0399] In certain embodiments, the checkpoint inhibitor is a B7 inhibitor. In some embodiments, the B7 inhibitor is a B7-H3 inhibitor or a B7-H4 inhibitor. In some embodiments, the B7-H3 inhibitor is anti-B7-H3 antibody MGA271 (Loo et al., Clin. Cancer Res., 2012, 3834).

[0400] In certain embodiments, the checkpoint inhibitor is a TIM3 (T-cell immunoglobulin domain and mucin domain 3) inhibitor (Fourcade et al., J. Exp. Med., 2010, 207, 2175-86; Sakuishi et al., J. Exp. Med., 2010, 207, 2187-94).

[0401] In certain embodiments, the checkpoint inhibitor is a GITR agonist. In some embodiments, the checkpoint inhibitor is an anti-GITR antibody. In some embodiments, the anti-GITR antibody is TRX518.

[0402] In certain embodiments, the checkpoint inhibitor is a CD137 agonist. In some embodiments, the checkpoint inhibitor is an anti-CD137 antibody. In some embodiments, the anti-CD137 antibody is urelumab. In some embodiments, the anti-CD137 antibody is PF-05082566.

[0403] In certain embodiments, the checkpoint inhibitor is recombinant human interleukin-15 (rhIL-15).

[0404] In certain embodiments, the checkpoint inhibitor is an IDO inhibitor. In some embodiments, the IDO inhibitor is INCB024360. In some embodiments, the IDO inhibitor is indoximod.

[0405] Other exemplary immunotherapies include adjuvant therapies that include immunotherapeutic agents such as cytokines, chemokines, interferons, interleukins, or lymphokines. Examples include cytokines such as granulocyte-macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), macrophage inflammatory protein (MIP)-1-alpha, interleukins (including IL-1, IL-2, IL-4, IL-6, IL-7, IL-12, IL-15, IL-18, IL-21, and IL-27), tumor necrosis factors (including TNF-alpha), and interferons (including IFN-alpha, IFN-beta, and IFN-gamma); aluminum hydroxide (alum); bacillus Calmette-Guerin (BCG); keyhole limpet hemocyanin (KLH); incomplete Freund's adjuvant (IFA); QS-21; DETOX; levamisole; and dinitrophenyl phosphate (DNP), as well as combinations thereof, such as combinations of interleukins, e.g., IL-2, with other cytokines, such as IFN-alpha.

[0406] Other exemplary immunotherapies include cell therapy, e.g., a population of immune cells, such as leukocytes (nucleated white blood cells), that comprise (e.g., express) a receptor that binds to an antigen of interest. Leukocytes of the present disclosure can be, for example, neutrophils, eosinophils, basophils, lymphocytes, or monocytes. In some embodiments, the leukocytes are lymphocytes. Examples of lymphocytes include T cells, B cells, natural killer (NK) cells, or NKT cells. In some embodiments, the T cells are CD4+ Th (T helper) cells, CD8+ cytotoxic T cells, γδ T cells, or regulatory (suppressor) T cells. In some embodiments, the immune cells are dendritic cells. In some embodiments, the cell therapy is CAR-T cell therapy. In some embodiments, the bispecific CAR is composed of two distinct antigen recognition domains present in tandem in a single transgenic receptor (referred to as TanCAR; see, e.g., Grada Z et al. Molecular Therapy Nucleic Acids 2013; 2:e105, which is incorporated herein by reference in its entirety). Thus, in some embodiments, the method comprises delivering to the tumor a combination comprising a ROR antigen binding molecule (e.g., an antibody) and an immunotherapeutic agent, wherein the immunotherapeutic agent is an engineered nucleic acid encoding an antigen, or an engineered nucleic acid that induces expression of an autoantigen, and delivering to the tumor an immune cell expressing a bispecific CAR that binds to two antigens, wherein one of the two antigens is encoded by the engineered nucleic acid.

[0407] Other exemplary immunotherapies include immunotherapeutic agents, such as cancer vaccines, that can be used to induce an immune response in a subject against a cancer antigen. An exemplary method involves administering to a subject an RNA vaccine comprising at least one RNA polynucleotide having an open reading frame encoding at least one antigenic polypeptide or immunogenic fragment thereof, either in the same composition or in separate compositions administered simultaneously or sequentially, in combination with the administration of an ROR antigen-binding molecule (e.g., an antibody), thereby inducing an immune response in the subject specific to the antigenic polypeptide or immunogenic fragment thereof, wherein the anti-antigenic polypeptide antibody titer in the subject is increased after vaccination compared to the anti-antigenic polypeptide antibody titer in a subject vaccinated with a prophylactically effective dose of a traditional vaccine against cancer.

[0408] In certain embodiments, the additional treatment comprises chemotherapy, such as one or more cytotoxic agents or one or more chemotherapeutic agents. Cytotoxic agents can inhibit or prevent cell function and / or cause cell death or destruction. Cytotoxic agents include radioisotopes (e.g., At 211 , I 131 , I 125 , Y 90 ,Re 186 ,Re 188 , Sm 153 , Bi 212 , P 32 , Pb 212 , and radioactive isotopes of Lu); chemotherapeutic agents; growth inhibitory agents; enzymes, such as nucleases and fragments thereof; and toxins, such as small molecule or enzymatically active toxins of bacterial, fungal, plant or animal origin, including fragments and / or variants thereof.

[0409] In certain embodiments, the additional treatment comprises one or more chemotherapeutic agents. Chemotherapeutic agents include compounds useful in the treatment of cancer. Chemotherapeutic agents include (i) antihormonal agents that act to regulate or inhibit hormone action in tumors, such as antiestrogens and selective estrogen receptor modulators; (ii) aromatase inhibitors that inhibit the enzyme aromatase, which regulates estrogen production in the adrenal glands; (iii) antiandrogens; (iv) protein kinase inhibitors; (v) lipid kinase inhibitors; (vi) antisense oligonucleotides, including those that inhibit the expression of genes in signal transduction pathways involved in abnormal cell proliferation; and (viii) vaccines, such as gene therapy vaccines. Chemotherapeutic agents can also include antibodies.

[0410] Exemplary kinase inhibitors include erlotinib (Tarceva®), gefitinib (Iressa®), dasatinib (Sprycel®), nilotinib (Tasigna®), crizotinib (Xalkori®), ruxolitinib (Jakafi®), vemurafenib (Zelboraf®), vandetanib (Caprelsa®), pazopanib (Votrie®), and others. nt (registered trademark), afatinib, alisertib, amuvatinib, axitinib, baricitinib, bosutinib, brivanib, canertinib, cabozantinib (Cabometyx (registered trademark)), cediranib, ceritinib, crenolanib, dabrafenib, dacomitinib, danusertib, dovitinib, foretinib, ganetespib, ibrutinib ib, idelalisib, imatinib, iniparib, lapatinib, lenvatinib, linifanib, linsitinib, masitinib, momelotinib, motesanib, neratinib, nintedanib, niraparib, oprozomib, olaparib, palbociclib, pictilisib, pirfenidone, ponatinib, quizartinib, regorafenib, rigosertib, rucaparib, saracatinib tinib), sridegib, sorafenib, sunitinib, tandutinib, tasocitinib, telatinib, tivantinib, tivozanib, tofacitinib, trametinib, veliparib, vismodegib, volasertib, cobimetinib (Cotellic®), XL-147, XL-765, XL-499, XL-880, and others.In some embodiments, an ROR antigen binding molecule (e.g., an antibody) can be used in combination with an HSP90 inhibitor (e.g., XL888), a liver X receptor (LXR) modulator, a retinoid-related orphan receptor gamma (RORy) modulator, a CK1 inhibitor, a CK1-α inhibitor, a Wnt pathway inhibitor (e.g., SST-215), or a mineralocorticoid receptor inhibitor (e.g., esaxerenone or XL-550) for the treatment of cancer.

[0411] Kinase inhibitors include tyrosine kinase inhibitors such as EGFR inhibitors; small molecule HER2 tyrosine kinase inhibitors such as mubritonib (TAK165, Takeda); CP-724.714 (Axon Medchem BV, an oral selective inhibitor of ErbB2 receptor tyrosine kinase); dual HER inhibitors such as EKB-569 (available from Wyeth), which preferentially binds to EGFR but inhibits both HER2 and EGFR overexpressing cells; lapatinib (GSK572016; available from Glaxo-SmithKline), an oral HER2 and EGFR tyrosine kinase inhibitor; PKI-166 (available from Novartis); pan-HER inhibitors such as canertinib (CI-1033; Pharmacia), which inhibit Raf-1 signaling; and ISIS inhibitors, which inhibit Raf-1 signaling. Raf-1 inhibitors, such as the antisense agent ISIS-5132 available from Pharmaceuticals; non-HER-targeted TK inhibitors, such as imatinib mesylate (GLEEVEC®, available from GlaxoSmithKline); multi-targeted tyrosine kinase inhibitors, such as sunitinib (SUTENT®, available from Pfizer); VEGF receptor tyrosine kinase inhibitors, such as vatalanib (PTK787 / ZK222584, available from Novartis / Schering AG); the MAPK extracellular-regulated kinase 1 inhibitor CI-1040 (available from Pharmacia); PD quinazolines such as 153035, 4-(3-chloroanilino)quinazoline; pyridopyrimidines; pyrimidopyrimidines; pyrrolopyrimidines such as CGP 59326, CGP 60261, and CGP 62706; pyrazolopyrimidine, 4-(phenylamino)-7H-pyrrolo[2,3-d]pyrimidine; curcumin (diferuloylmethane, 4,5-bis(4-fluoroanilino)phthalimide); tyrphostins containing a nitrothiophene moiety; antisense molecules (e.g., antisense molecules that bind to HER-encoding nucleic acids); quinoxalines (U.S. Pat. No. 5,804,396); tryphostin (U.S. Pat. No. 5,804,396); Affinitac (ISIS 3521; Isis / Lilly); PKI166 (Novartis); semaxinib (Pfizer); INC-1C11 (Imclone), rapamycin (sirolimus, RAPAMUNE®); or those described in any of the following patent publications: U.S. Pat. No. 5,804,396; WO1999 / 09016 (American Cyanamid); WO1998 / 43960 (American Cyanamid); WO1997 / 38983 (Warner Lambert); WO1999 / 06378 (Warner Lambert); WO1999 / 06396 (Warner Lambert); WO1996 / 30347 (Pfizer, Inc); WO1996 / 33978 (Zeneca); WO1996 / 3397 (Zeneca) and WO1996 / 33980 (Zeneca).

[0412] Combination treatment with a ROR antigen binding molecule (e.g., an antibody) provided herein and an additional therapy, such as a therapeutic agent, can be simultaneous, separate, or sequential, in any order. For combined, simultaneous administration of a therapeutic agent, such as a ROR antigen binding-providing molecule, and another therapeutic agent, such as an immunotherapeutic or chemotherapeutic agent, the therapeutic agents can be administered as a single composition or separate compositions, as appropriate. 6.13. [Example]

[0413] The following examples are offered by way of illustration and not by way of limitation. 6.13.1. Method

[0414] Non-limiting illustrative methods for the purification of various antigen binding proteins and their use in various assays are described in more detail below. 6.13.1.1.Expi293 Expression

[0415] The various antigen-binding proteins being tested were expressed using the Expi293 transient transfection system according to the manufacturer's instructions. Briefly, four plasmids encoding the four individual chains were mixed in a 1:1:1:1 mass ratio unless otherwise noted and transfected into Expi293 cells using the ExpiFectamine 293 transfection kit. Cells were cultured at 37°C, 8% CO2, 100% humidity, with shaking at 125 rpm. Transfected cells were fed once 16-18 hours after transfection. Cells were harvested on day 5 by centrifugation at 2000g for 10 minutes. The supernatant was collected for affinity chromatography purification. 6.13.1.2. Protein A and Anti-CH1 Purification

[0416] Clarified supernatants containing various antigen-binding proteins were separated using either Protein A (ProtA) or anti-CH1 resin on an AKTA Purifier FPLC. In cases where a head-to-head comparison was performed, the supernatants containing various antigen-binding proteins were split into two equal samples. For ProtA purification, a 1 mL Protein A column (GE Healthcare) was equilibrated with PBS (5 mM sodium potassium phosphate pH 7.4, 150 mM sodium chloride). The sample was loaded onto the column at 5 mL / min. The sample was eluted using 0.1 M acetic acid pH 4.0. Elution was monitored by absorbance at 280 nm, and the elution peaks were pooled for analysis. For anti-CH1 purification, a 1 mL CaptureSelect™ XL column (ThermoFisher) was equilibrated with PBS. The sample was loaded onto the column at 5 mL / min. The sample was eluted using 0.1 M acetic acid pH 4.0. Elution was monitored by absorbance at 280 nm and elution peaks were pooled for analysis. 6.13.1.3.SDS-Page analysis

[0417] Samples containing various isolated antigen-binding proteins were analyzed by reducing and non-reducing SDS-PAGE for the presence of complete and incomplete products and overall purity. 2 μg of each sample was added to 15 μL SDS loading buffer. Reduced samples were incubated at 75°C for 10 minutes in the presence of 10 mM reducing agent. Non-reduced samples were incubated at 95°C for 5 minutes without reducing agent. Reduced and non-reduced samples were loaded onto a 4-15% gradient TGX gel (BioRad) in running buffer and run at 250 volts for 30 minutes. After completion of the run, the gel was washed with DI water and stained using GelCode Blue Safe Protein Stain (ThermoFisher). Prior to analysis, the gel was destained with DI water. Densitometric analysis of scanned images of the destained gel was performed using standard image analysis software to calculate the relative abundance of bands in each sample. IEX Chromatography

[0418] Samples containing various isolated antigen-binding proteins were analyzed by cation exchange chromatography for the ratio of intact to incomplete products and impurities. The clarified supernatant was analyzed using a 5 ml MonoS (GE Lifesciences) column in an AKTA Purifier FPLC. The MonoS column was equilibrated with buffer A, 10 mM MES, pH 6.0. The sample was loaded onto the column at 2 ml / min. The sample was eluted using a 0-30% gradient of buffer B (10 mM MES, pH 6.0, 1 M sodium chloride) over 6 CV. Elution was monitored by absorbance at 280 nm, and sample purity was calculated by peak integration to identify the abundance of the monomeric and contaminant peaks. The monomeric and contaminant peaks were pooled separately for analysis by SDS-PAGE as described above. 6.13.1.5. Analytical SEC Chromatography

[0419] Samples containing various isolated antigen-binding proteins were analyzed by analytical size-exclusion chromatography for the ratio of monomer to high molecular weight products and impurities. The clarified supernatant was analyzed using an industry-standard TSK G3000SWxl column (Tosoh Bioscience) on an Agilent 1100 HPLC. The TSK column was equilibrated with PBS. 25 μL of each 1 mg / mL sample was loaded onto the column at 1 ml / min. The samples were eluted using an isocratic flow of 1.5 CV of PBS. Elution was monitored by absorbance at 280 nm, and the eluted peaks were analyzed by peak integration. 6.13.1.6.Mass spectrometry

[0420] Samples containing various isolated antigen-binding proteins were analyzed by mass spectrometry to confirm the exact species by molecular weight. All analyses were performed by a third-party laboratory. Briefly, samples were treated with a cocktail of enzymes to remove glycosylation. Both samples were run in a reducing format to specifically identify each chain by molecular weight. All samples were run under non-reducing conditions to identify the molecular weight of all complexes in the sample. Mass spectral analysis was used to identify the number of unique products based on molecular weight. 6.13.1.7. Antibody Discovery by Phage Display

[0421] Phage display of the human Fab library was performed using standard protocols. Biotinylated extracellular domains of human ROR1 and ROR2 proteins were purchased from Acro Biosystems and biotinylated with EZ-Link NHS Biotin (Thermo Scientific catalog number 20217). Phage clones were screened for their ability to bind to the extracellular domains of ROR1 (Acro catalog number RO1-H522y) and ROR2 (Acro catalog number RO2-H52E5) by phage ELISA using standard protocols. Briefly, a Fab-formatted phage library was constructed using an expression vector (also called a phagemid) capable of replication and expression in phage. Both heavy and light chains were encoded in the same expression vector, with the heavy chain fused to a truncated variant of the phage coat protein pIII. The light and heavy chain-pIII fusions are expressed as separate polypeptides and assemble in the bacterial periplasm, where redox potential allows disulfide bond formation to form phage-displayed antibodies containing candidate ABS.

[0422] Libraries were generated using sequences derived from a specific human heavy chain variable domain (VH3-23) and a specific human light chain variable domain (Vk-1). The light chain variable domains in the screened libraries were generated by introducing diversity into the VL CDR3 (L3), while the light chain VL CDR1 (L1) and CDR2 (L2) remained human germline sequences. For the screened libraries, all three CDRs of the VH domain were diversified to match the positional amino acid frequencies of the CDRs found in the human antibody repertoire by length. The phage display heavy chain (SEQ ID NO: 74) and light chain (SEQ ID NO: 75) scaffolds used in the libraries are listed below, where lowercase "x" represents CDR amino acids that were varied to generate the library, and bold italics represent CDR sequences that were constant.

[0423] As described in detail by Kunkel, TA (PNAS January 1, 1985. 82 (2) 488-492), the entire contents of which are incorporated herein by reference, diversity was generated by Kunkel mutagenesis using primers to introduce diversity into the VL CDR3 and VH CDR1 (H1), CDR2 (H2), and CDR3 (H3) to mimic the diversity found in natural antibody repertoires. Briefly, single-stranded DNA was prepared from isolated phages using standard procedures, and Kunkel mutagenesis was performed. The chemically synthesized DNA was then electroporated into TG1 cells, followed by recovery. The recovered cells were subcultured and infected with M13K07 helper phage to produce a phage library.

[0424] Phage panning was performed using standard procedures. Briefly, the first round of phage panning involved the addition of approximately 5 x 10 phage from the prepared library in a volume of 1 mL in PBST-2% BSA. 12The assay was performed with targets immobilized on streptavidin magnetic beads exposed to phages. After 1 hour of incubation, the phages bound to the beads were separated from the supernatant using a magnetic stand. The beads were washed three times to remove non-specifically bound phages and then subjected to OD 600 Approximately 0.6% of ER2738 cells (5 mL) were added. After 20 minutes, infected cells were subcultured in 25 mL 2xYT + ampicillin and M13K07 helper phage and grown overnight at 37°C with vigorous shaking. The following day, phage were prepared using standard procedures by PEG precipitation. Preclearance of specific phage on SAV-coated beads was performed prior to panning. A second round of panning was performed using a KingFisher magnetic bead handler with 100 nM bead-immobilized antigen using standard procedures. A total of three to four rounds of phage panning were performed to enrich for phages displaying Fabs specific for the target antigen. Target-specific enrichment was confirmed using polyclonal and monoclonal phage ELISA. DNA sequencing was used to identify isolated Fab clones containing candidate Fabs.

[0425] To measure binding affinity in the ROR binder discovery campaign, the VL and VH domains identified in the above-described phage screen were formatted into a bivalent, monospecific, native human full-length IgG1 architecture and immobilized on a biosensor in an Octet (Pall ForteBio) biolayer interferometer. Next, the extracellular domains of ROR1 (Acro Cat. No. RO1-H522y) and ROR2 (Acro Cat. No. RO2-H52E5), as well as individual ROR1 domains, were immobilized on a biosensor in an Octet (Pall ForteBio) biolayer interferometer. Soluble ROR antigens containing Frizzled (Acro Catalog No. RO1-H5222), Ig-like (Acro Catalog No. RO1-H5221), and Kringle (Acro Catalog No. RO1-H5223) domains were added to the system and binding was measured.

[0426] For experiments performed using the B-body format, the VL variable regions of individual clones were formatted into domains A and / or H of the bivalent 1x1 B-body "BC1" scaffold, and the VH regions were formatted into domains F and / or L thereof, as shown below and with reference to Figure 3.

[0427] "BC1" Scaffolding: First polypeptide chain (SEQ ID NO: 78) Domain A = Antigen 1 B-body domain A / H scaffold (SEQ ID NO: 76) Domain B = CH3 (T366K; 445K, 446S, 447C tripeptide insertion) Domain D=CH2 Domain E=CH3(T366W, S354C) Second polypeptide chain (SEQ ID NO:79): Domain F = Antigen 1 B-body domain F / L scaffold (SEQ ID NO: 77) Domain G = CH3 (L351D; 445G, 446E, 447C tripeptide insertion) Third polypeptide chain (SEQ ID NO:80): Domain H = Antigen 2 B-body domain A / H scaffold (SEQ ID NO: 76) Domain I = CL (kappa) Domain J=CH2 Domain K = CH3 (Y349C, D356E, L358M, T366S, L368A, Y407V) Fourth polypeptide chain (SEQ ID NO:81): Domain L = Antigen 2 B-body domain F / L scaffold (SEQ ID NO: 77) Domain M=CH1.

[0428] For the ROR ABS candidate formatted into a bivalent, bispecific 1x1 format with anti-CD3 SP34-89, Domain H has the amino acid sequence of SEQ ID NO: 69, Domain L has the amino acid sequence of SEQ ID NO: 68, while Domain A has the candidate ROR VL sequence and Domain F has the candidate ROR VH sequence.

[0429] For the BC1 1x2 format, the variable domains were formatted into a 1(A) x 2(BA) format as described in Section 6.3.17.4 with reference to Figure 26. Figure 26 presents a schematic diagram of the five polypeptide chains and their domains according to the respective naming conventions for the trivalent 1x2 antibody constructs described herein, where chain 5 is designated the "sixth polypeptide chain" in the schematic. Unless otherwise specified, the ROR antigen binding site (ABS) is a bivalent binder ("A" specificity) and CD3 is a monovalent binder ("B" specificity). The SP34-89 1x2 chain 3 scaffold has the sequence of SEQ ID NO: 82, and the junction between domain S and domain H is a 10-amino acid linker with the sequence TASSGGSSSG (SEQ ID NO: 83), unless otherwise noted. Polypeptide chain 2 and chain 5 are identical in a 1(A) x 2(BA) format (see, e.g., Figure 26; according to the naming convention, chain 5 is designated the "sixth polypeptide chain" in this schematic). NFκB GFP Jurkat T Cell Stimulation Assay

[0430] The NFκB / Jurkat / GFP transcriptional reporter cell line was purchased from System Biosciences (Cat# TR850-1). The anti-CD28 antibody used for costimulation was purchased from BD Pharmingen (Cat# 555725). Solution C background suppression dye was purchased from Life Technologies (K1037). Briefly, in a 96-well black-walled, clear-bottom plate, Jurkat cells (effector cells, E) were mixed with tumor cells (T) at an E:T ratio of 2:1 to 4:1 in the presence of a dilution series of B-body™ antibodies and 1 μg / mL of anti-CD28 antibody. The plate was incubated at 37°C / 5% CO2 for 6 hours, after which a 6x solution of Solution C background suppressor was added to the plate, and GFP fluorescence was read on a plate reader. EC50 values, which refer to the concentration of antibody producing a half-maximal response, were determined from the dilution series. 6.13.1.9. Primary T cell cytotoxicity assay

[0431] Cells (T) expressing target tumor antigens and effector cells (E) were mixed at E:T ratios ranging from 3:1 to 10:1. Effector cells used included PBMCs or isolated cytotoxic CD8+ T cells. Candidate redirecting T cell antibodies were added to the cells in a dilution series. Controls included a medium-only control, a tumor cell-only control, and an untreated E:T cell control. The mixed cells and control conditions were incubated at 37°C / 5% CO2 for 40–50 h. The Cytotoxicity Detection Kit Plus (LDH) was purchased from Sigma (Cat. 4744934001) and was performed according to the manufacturer's instructions. Briefly, lysis solution added to tumor cells served as the 100% cytotoxicity control, and untreated E:T cells served as the 0% cytotoxicity control. The level of lactate dehydrogenase (LDH) in each sample was determined by absorbance at 490 nm and normalized to the 100% and 0% controls. EC50 values, which refer to the concentration of antibody that produces a half-maximal response, were determined from the dilution series. 6.13.2. Example 1 Bivalent Monospecific and Bivalent Bispecific Constructs

[0432] A bivalent monospecific B-Body recognizing TNFα was constructed using standard molecular biology techniques with the following configuration: VL(certolizumab)-CH3(knob)-CH2-CH3 / VH(certolizumab)-CH3(hole). In this construct: First Polypeptide Chain (SEQ ID NO: 1) Domain A = VL (certolizumab) Domain B = CH3 (IgG1) (knob: S354C + T366W) Domain D=CH2 (IgG1) Domain E=CH3(IgG1) Second polypeptide chain (SEQ ID NO: 2) Domain F = VH (certolizumab) Domain G = CH3 (IgG1) (hole: Y349C, T366S, L368A, Y407V) Third polypeptide chain: Identical to the first polypeptide chain Fourth polypeptide chain: Identical to the second polypeptide chain.

[0433] References to domains and polypeptide chains follow Figure 3. The overall organization of the construct is shown in Figure 4. The sequence of the first polypeptide chain, having domain A, abbreviated and identified as "(VL)", is provided in SEQ ID NO: 1. The sequence of the second polypeptide chain, having domain F, abbreviated and identified as "(VH)", is provided in SEQ ID NO: 2.

[0434] The full-length construct was expressed in an E. coli cell-free protein synthesis expression system for approximately 18 hours at 26°C with gentle agitation. After expression, the cell-free extract was centrifuged to pellet insoluble material, and the supernatant was diluted 2-fold with 10x kinetic buffer (Forte Bio) and used as a sample for biolayer interference.

[0435] Biotinylated TNFα was immobilized on a streptavidin sensor and subjected to a wave-shift response of approximately 1.5 nm. After establishing a baseline with 10x kinetic buffer, the sensor was immersed in the antibody construct analyte solution. The construct gave a response of approximately 3 nm, comparable to the conventional IgG format of certolizumab, demonstrating the ability of bivalent monospecific constructs to assemble into functional full-length antibodies. The results are shown in Figure 5.

[0436] We have also constructed a bivalent, bispecific antibody with the following domain organization: First polypeptide chain: VL-CH3-CH2-CH3 (knob) Second polypeptide chain: VH-CH3 Third polypeptide chain: VL-CL-CH2-CH3 (hole) Fourth polypeptide chain VH-CH1.

[0437] The sequences (excluding the variable region sequences) are shown in SEQ ID NO: 3 (first polypeptide chain), SEQ ID NO: 4 (second polypeptide chain), SEQ ID NO: 5 (third polypeptide chain), and SEQ ID NO: 6 (fourth polypeptide chain), respectively. 6.13.3. Example 2 Bivalent, bispecific B-Body "BC1"

[0438] We constructed a bivalent, bispecific construct, designated "BC1," specific for PD1 and a second antigen, "Antigen A." The salient features of the "BC1" construct are shown in FIG. 6.

[0439] In more detail, with reference to the domains and polypeptide chains according to Figure 3 and modifications from the native sequence indicated in brackets, the construction was as follows: First Polypeptide Chain (SEQ ID NO: 8) Domain A = VL ("Antigen A") Domain B = CH3 (T366K; 445K, 446S, 447C tripeptide insertion) Domain D=CH2 Domain E=CH3(T366W, S354C) Second polypeptide chain (SEQ ID NO: 9): Domain F = VH ("Antigen A") Domain G = CH3 (L351D; 445G, 446E, 447C tripeptide insertion) Third polypeptide chain (SEQ ID NO: 10): Domain H=VL("Nivo") Domain I = CL (kappa) Domain J=CH2 Domain K = CH3 (Y349C, D356E, L358M, T366S, L368A, Y407V) Fourth polypeptide chain (SEQ ID NO: 11): Domain L=VH("Nivo") Domain M=CH1.

[0440] The A domain (SEQ ID NO: 12) and F domain (SEQ ID NO: 16) form an antigen-binding site (A:F) specific for "antigen A." The H domain has a VH sequence derived from nivolumab, and the L domain has a VL sequence derived from nivolumab; H and L associate to form an antigen-binding site (H:L) specific for human PD1.

[0441] The B domain (SEQ ID NO: 13) has the sequence of human IgG1 CH3 with several mutations, namely, T366K, 445K, 446S, and 447C insertions. The T366K mutation is the charge-pair cognate of the L351D residue in domain G. The "447C" residue in domain B is derived from a C-terminal KSC tripeptide insertion.

[0442] Domain D (SEQ ID NO: 14) has the sequence of human IgG1 CH2.

[0443] Domain E (SEQ ID NO: 15) has the sequence of human IgG1 CH3 with the mutations T366W and S354C. 366W is a "knob" mutation. 354C introduces a cysteine ​​that can form a disulfide bond with the cognate 349C mutation in domain K.

[0444] Domain G (SEQ ID NO: 17) has the sequence of human IgG1 CH3 with the following mutations: L351D, and a 445G, 446E, 447C tripeptide insertion. The L351D mutation introduces the charge pair cognate to the domain B T366K mutation. The "447C" residue in domain G is derived from the C-terminal GEC tripeptide insertion.

[0445] Domain I (SEQ ID NO: 19) has the sequence of the human C kappa light chain (CK).

[0446] Domain J [SEQ ID NO: 20] has the sequence of the human IgG1 CH2 domain and is identical to the sequence of domain D.

[0447] Domain K [SEQ ID NO: 21] has the sequence of human IgG1 CH3 with the following changes: Y349C, D356E, L358M, T366S, L368A, Y407V. The 349C mutation introduces a cysteine ​​that can form a disulfide bond with the cognate 354C mutation in domain E. 356E and L358M introduce isoallotypic amino acids that reduce immunogenicity. 366S, 368A, and 407V are "hole" mutations.

[0448] Domain M [SEQ ID NO: 23] has the sequence of the human IgG1 CH1 region.

[0449] "BC1" can be readily expressed at high levels using mammalian expression, at concentrations exceeding 100 μg / ml.

[0450] We have found that the bivalent, bispecific "BC1" protein can be easily purified in a single step using the CH1-specific CaptureSelect™ affinity resin from ThermoFisher.

[0451] As shown in Figure 7A, SEC analysis demonstrates that a single CH1 affinity purification step yields a single monodisperse peak that is >98% monomeric via gel filtration. Figure 7B shows comparative literature data for SEC analysis of CrossMab bivalent antibody constructs.

[0452] Figure 8A shows the cation exchange chromatography elution profile of "BC1" after one-step purification using CaptureSelect™ CH1 affinity resin, showing a single tight peak. Figure 8B shows the cation exchange chromatography elution profile of "BC1" after purification using standard Protein A purification, showing an additional elution peak consistent with the co-purification of an incompletely assembled product.

[0453] Figure 9 shows an SDS-PAGE gel under non-reducing conditions. As seen in lane 3, single-step purification of "BC1" using CH1 affinity resin yields a nearly homogeneous single band. Lane 4 shows a subsequent minimal additional purification with a cation exchange polishing step. Lane 7 shows, by comparison, mild purification using standard Protein A purification, and lanes 8-10 show further purification of the Protein A purified material using cation exchange chromatography.

[0454] FIG. 10 shows an SDS-PAGE gel of "BC1" after single-step CH1-affinity purification under both non-reducing and reducing conditions (Panel A) compared with an SDS-PAGE gel of a CrossMab bispecific antibody under non-reducing and reducing conditions published in the literature (Panel B).

[0455] Figure 11 shows a mass spectrometry analysis of "BC1" demonstrating two distinct heavy chains (Figure 11A) and two distinct light chains (Figure 11B) under reducing conditions. The mass spectrometry data in Figure 12 confirms the absence of mispairing after purification.

[0456] Accelerated stability studies were performed to evaluate the long-term stability of the "BC1" B-Body design. Purified B-Body was concentrated to 8.6 mg / ml in PBS buffer and incubated at 40°C. Structural integrity was measured weekly using analytical size-exclusion chromatography (SEC) with a Shodex KW-803 column. Structural integrity was determined by measuring the percentage of intact monomer (% monomer) in relation to aggregate formation. Data are shown in Figure 13. IgG Control 1 is a positive control with good stability characteristics. IgG Control 2 is a negative control known to aggregate under the incubation conditions. The "BC1" B-Body was incubated for 8 weeks without any loss of structural integrity, as determined by analytical SEC.

[0457] Furthermore, the TM of the bivalent construct was approximately 72°C, and the inventors determined that "BC1" has high thermostability.

[0458] Table 1 compares "BC1" with CrossMab in key developability characteristics. [Table 1] *Data from Schaefer et al. (Proc Natl Acad Sci USA. 2011 Jul 5;108(27):11187-92) 6.13.4. Example 3 Bivalent, bispecific B-Body "BC6"

[0459] We constructed a bivalent, bispecific B-Body designated "BC6." "BC6" is similar to "BC1," but is identical to BC1 except that it retains the wild-type residues at residue 366 in domain B and residue 351 in domain G. Thus, "BC6" lacks the charge pair cognates T366K and L351D, which were designed to facilitate correct pairing in domains B and G of "BC1." The salient features of the "BC6" construct are shown in Figure 14.

[0460] Despite the absence of charge-pair residues present in "BC1," single-step purification of "BC6" using CH1 affinity resin was found to result in a highly homogeneous sample. Figure 15A shows SEC analysis of "BC6" after one-step purification using CaptureSelect™ CH1 affinity resin. The data demonstrate that single-step CH1 affinity purification yields a single, monodisperse peak, similar to that observed for "BC1," demonstrating that the disulfide bonds between polypeptide chains 1 and 2 and between polypeptide chains 3 and 4 remain intact. The chromatogram also demonstrates the absence of noncovalent aggregates.

[0461] Figure 15B shows an SDS-PAGE gel under non-reducing conditions, where lane 1 is the first lot of "BC6" loaded after single-step CH1 affinity purification, lane 2 is the second lot of "BC6" loaded after single-step CH1 affinity purification, and lanes 3 and 4 demonstrate the further purification that can be achieved using ion exchange chromatography following CH1 affinity purification. 6.13.5. Example 4 Bivalent, bispecific B-bodies "BC28", "BC29", "BC30", and "BC31"

[0462] We constructed bivalent 1x1 bispecific B-Body constructs "BC28," "BC29," "BC30," and "BC31" with engineered disulfides within the CH3 interface of domains B and G as alternative S-S linkages to the C-terminal disulfides present in "BC1" and "BC6." Literature has shown that disulfide bonds at the CH3 interface are insufficient to exert orthogonality in the context of Fc CH3 domains. The overall organization of these B-Body constructs is outlined in Figure 16, and salient features of "BC28" are shown below: Polypeptide chain 1: "BC28" chain 1 (SEQ ID NO: 24) Domain A = VL (antigen "A") Domain B = CH3 (Y349C; 445P, 446G, 447K insertions) Domain D=CH2 Domain E=CH3(S354C, T366W) Polypeptide chain 2: "BC28" chain 2 (SEQ ID NO: 25) Domain F = VH (antigen "A") Domain G = CH3 (S354C; 445P, 446G, 447K insertions) Polypeptide chain 3: "BC1" chain 3 (SEQ ID NO: 10) Domain H=VL("Nivo") Domain I = CL (kappa) Domain J=CH2 Domain K = CH3 (Y349C, D356E, L358M, T366S, L368A, Y407V) Polypeptide chain 4: "BC1" chain 4 (SEQ ID NO: 11) Domain L=VH("Nivo") Domain M=CH1.

[0463] The "BC28" A:F antigen-binding site is specific for "antigen A." The "BC28" H:L antigen-binding site is specific for PD1 (nivolumab sequence). "BC28" domain B has the following changes compared to wild-type CH3: Y349C; 445P, 446G, 447K insertions. "BC28" domain E has the following changes compared to wild-type CH3: S354C, T366W. "BC28" domain G has the following changes compared to wild-type: S354C; 445P, 446G, 447K insertions.

[0464] Thus, "BC28" has an engineered cysteine ​​at residue 349C in domain B and an engineered cysteine ​​at residue 354C in domain G ("349C-354C").

[0465] "BC29" has engineered cysteines at residue 351C in domain B and 351C in domain G ("351C-351C"). "BC30" has engineered cysteines at residue 354C in domain B and 349C in domain G ("354C-349C"). BC31 has engineered cysteines at residue 394C and 394C in domain G ("394C-394C"). BC32 has engineered cysteines at residue 407C in domain B and 407C in domain G ("407C-407C").

[0466] Figure 17 shows SDS-PAGE analysis under non-reducing conditions after one-step purification using CaptureSelect™ CH1 affinity resin. Lanes 1 and 3 show high levels of expression and substantial homogeneity of intact "BC28" (lane 1) and "BC30" (lane 3). Lane 2 shows oligomerization of BC29. Lanes 4 and 5 show poor expression of BC31 and BC32, respectively, indicating insufficient ligation of BC32. Another construct, BC9 (a construct reported by Genentech with introduced cysteines at residues 392 in domain B and 399 in domain G ("392C-399C"), forming a disulfide pair) showed oligomerization on SDS PAGE (data not shown).

[0467] Figure 18 shows SEC analysis of "BC28" and "BC30" after one-step purification using CaptureSelect™ CH1 affinity resin. Furthermore, we demonstrated that "BC28" can be easily purified using a single-step purification using Protein A resin (results not shown). 6.13.6. Example 5 Bivalent, bispecific B-Body "BC44"

[0468] FIG. 19 shows the general organization of our currently preferred bivalent, bispecific 1×1 construct, bivalent, bispecific 1×1 B-Body “BC44.” First polypeptide chain (“BC44” chain 1) (SEQ ID NO: 32) Domain A = VL (antigen "A") Domain B = CH3 (P343V; Y349C; 445P, 446G, 447K insertions) Domain E=CH2 Domain E=CH3(S354C, T366W) Second polypeptide chain (= "BC28" polypeptide chain 2) (SEQ ID NO: 25) Domain F = VH (antigen "A") Domain G = CH3 (S354C; 445P, 446G, 447K insertions) Third polypeptide chain (= "BC1" polypeptide chain 3) (SEQ ID NO: 10) Domain H=VL("Nivo") Domain I = CL (kappa) Domain J=CH2 Domain K = CH3 (Y349C, D356E, L358M, T366S, L368A, Y407V) Fourth polypeptide chain (= "BC1" polypeptide chain 4) (SEQ ID NO: 11) Domain L=VH("Nivo") Domain M=CH1 6.13.7. Example 6 Variable - CH3 junction operation

[0469] We generated a series of variants in which the VL-CH3 junction between domains A and B and the VH-CH3 junction between domains F and G were mutated, and evaluated the expression levels, assembly, and stability of bivalent 1x1 B-Body constructs. While many solutions are possible, we chose to use only residues naturally found within the VL, VH, and CH3 domains to reduce the introduction of T cell epitopes. Structural evaluation of the domain configuration further narrowed down the preferred sequence combinations. Tables 2 and 3 below show the junctions for several junction variants based on "BC1" and other bivalent constructs. [Table 2] [Table 3]

[0470] Figure 20 shows size exclusion chromatography of "BC15" and "BC16" samples at the indicated weeks in the accelerated stability testing protocol at 40° C. "BC15" remained stable; "BC16" was found to become unstable over time. 6.13.8. Example 7 Trivalent 2x1 bispecific B-Body construct ("BC1-2x1")

[0471] We constructed a trivalent 2x1 bispecific B-Body, "BC1-2x1," based on "BC1." The salient features of the construct are shown in Figure 22.

[0472] More specifically, using reference to the domains and polypeptide chains summarized in Figure 21, First Polypeptide Chain Domain N = VL ("Antigen A") Domain O=CH3(T366K, 447C) Domain A = VL ("Antigen A") Domain B = CH3 (T366K, 447C) Domain D=CH2 Domain E = CH3 (Knob, 354C) Fifth polypeptide chain (= "BC1" chain 2) Domain P = VH ("Antigen A") Domain Q = CH3 (L351D, 447C) Second polypeptide chain (= "BC1" chain 2) Domain F = VH ("Antigen A") Domain G = CH3 (L351D, 447C) Third polypeptide chain (= "BC1" chain 3) Domain H=VL("Nivo") Domain I = CL (kappa) Domain J=CH2 Domain K=CH3 (Hole, 349C) Fourth polypeptide chain (= "BC1" chain 4) Domain L=VH("Nivo") Domain M=CH1

[0473] Figure 23 shows a non-reducing SDS-PAGE of proteins expressed using the ThermoFisher Expi293 transient transfection system.

[0474] Lane 1 shows the eluate of the trivalent 2x1 "BC1-2x1" protein after one-step purification using CaptureSelect™ CH1 affinity resin. Lane 2 shows the lower molecular weight, faster migrating, bivalent "BC1" protein after one-step purification using CaptureSelect™ CH1 affinity resin. Lanes 3-5 show the purification of "BC1-2x1" using Protein A. Lanes 6 and 7 show the purification of "BC1-2x1" using CH1 affinity resin.

[0475] Figure 24 compares the avidity of the bivalent "BC1" construct with that of the trivalent 2x1 "BC1-2x1" construct using Octet (Pall ForteBio) analysis. Biotinylated antigen "A" is immobilized on a surface over which the antibody construct is passed for binding analysis. 6.13.9. Example 8 Trivalent 2x1 trispecific B-Body construct ("TB111")

[0476] The inventors have designed a trivalent 2x1 trispecific molecule "TB111" with the structure outlined in Figure 25. With reference to the domain naming convention shown in Figure 21, TB111 has the following structure ("Ada" indicates the V region derived from adalimumab): Polypeptide chain 1 Domain N:VH("Ada") Domain O: CH3 (T366K, 394C) Domain A: VL ("Antigen A") Domain B: CH3 (T366K, 349C) Domain D:CH2 Domain E: CH3 (Knob, 354C) Polypeptide chain 5 Domain P:VL("Ada") Domain Q: CH3 (L351D, 394C) Polypeptide chain 2 Domain F: VH ("Antigen A") Domain G: CH3 (L351D, 351C) Polypeptide chain 3 Domain H:VL ("Nivo") Domain I: CL (kappa) Domain J:CH2 Domain K: CH3 (hole, 349C) Polypeptide chain 4 (= "BC1" chain 4) Domain L:VH ("Nivo") Domain M:CH1 This construct was not expressed. 6.13.10. Example 9 Trivalent 1x2 bispecific construct ("BC28-1x2")

[0477] Applicants constructed a trivalent 1x2 bispecific B-body with the following domain and chain structure, with reference to the domain and chain nomenclature depicted in Figure 26: First polypeptide chain (= "BC28" chain 1) (SEQ ID NO: 24) Domain A = VL (antigen "A") Domain B = CH3 (Y349C; 445P, 446G, 447K insertions) Domain D=CH2 Domain E=CH3(S354C, T366W) Second polypeptide chain (= "BC28" chain 2) (SEQ ID NO: 25) Domain F = VH (antigen "A") Domain G = CH3 (S354C; 445P, 446G, 447K insertions) Third polypeptide chain (SEQ ID NO: 37) Domain R = VL (antigen "A") Domain S = CH3 (Y349C; 445P, 446G, 447K insertions) Linker=GSGSGS Domain H=VL("Nivo") Domain I=CL Domain J=CH2 Domain K = CH3 (Y349C, D356E, L358M, T366S, L368A, Y407V) Fourth polypeptide chain (= "BC1" chain 4) (SEQ ID NO: 11): Domain L=VH("Nivo") Domain M=CH1 Sixth polypeptide chain (= "BC28" chain 2) (SEQ ID NO: 25) Domain T = VH (antigen "A") Domain U = CH3 (S354C; 445P, 446G, 447K insertions).

[0478] Similar to the H:L linked antigen binding site, the A:F antigen binding site is specific for "antigen A." The R:T antigen binding site is specific for PD. The specificity of this construct is therefore antigen "A" x (PD1 - antigen "A"). 6.13.11. Example 10 Trivalent 1x2 bispecific construct ("CTLA4-4xNivoxCTLA4-4")

[0479] We constructed a trivalent 1x2 bispecific molecule ("CTLA4-4xNivoxCTLA4-4") with the general structure outlined in Figure 27. The domain nomenclature is shown in Figure 26.

[0480] FIG. 28 is an SDS-PAGE gel in which the lanes showing the "CTLA4-4 x Nivo x CTLA4-4" construct under non-reducing and reducing conditions are boxed.

[0481] Figure 29 compares the antigen binding of two antibodies, "CTLA4-4xOX40-8" and "CTLA4-4xNivoxCTLA4-4." "CTLA4-4xOX40-8" binds monovalently to CTLA4, while "CTLA4-4xNivoxCTLA4-4" binds bivalently to CTLA4. 6.13.12. Example 11 Trivalent 1x2 trispecific construct "BC28-1x1x1a"

[0482] We have constructed a trivalent 1x2 trispecific molecule with the general structure outlined in Figure 30. With reference to the domain and chain nomenclature shown in Figure 26, First polypeptide chain (="BC28" chain 1) [SEQ ID NO: 24] Domain A = VL (antigen "A") Domain B = CH3 (Y349C; 445P, 446G, 447K insertions) Domain D=CH2 Domain E=CH3(S354C, T366W) Second polypeptide chain (= "BC28" chain 2) (SEQ ID NO: 25) Domain F = VH (antigen "A") Domain G = CH3 (S354C; 445P, 446G, 447K insertions) Third Polypeptide Chain (SEQ ID NO: 45) Domain R=VL(CTLA4-4) Domain S = CH3 (T366K; 445K, 446S, 447C insertions) Linker=GSGSGS Domain H=VL("Nivo") Domain I=CL Domain J=CH2 Domain K = CH3 (Y349C, D356E, L358M, T366S, L368A, Y407V) Fourth polypeptide chain (= "BC1" chain 4) (SEQ ID NO: 11) Domain L=VH("Nivo") Domain M=CH1. Sixth polypeptide chain (=hCTLA4-4 chain 2) (SEQ ID NO: 53) Domain T=VH(CTLA4) Domain U = CH3 (insertion of L351D, 445G, 446E, 447C)

[0483] The antigen binding sites of this trispecific construct were as follows: The antigen-binding site A:F was specific for “antigen A”; The antigen-binding site H:L was specific for PD1 (nivolumab sequence); The antigen binding site R:T was specific for CTLA4.

[0484] FIG. 31 shows size exclusion chromatography of "BC28-1x1x1a" after transient expression and one-step purification using CaptureSelect™ CH1 affinity resin, showing a single, clearly defined peak. 6.13.13. Example 12 SDS-PAGE analysis of bivalent and trivalent constructs

[0485] FIG. 32 shows SDS-PAGE gels under non-reducing and reducing conditions of various constructs after transient expression and one-step purification using CaptureSelect™ CH1 affinity resin, respectively.

[0486] Lanes 1 (non-reducing conditions) and 2 (reducing conditions, +DTT) are the bivalent 1x1 bispecific construct "BC1." Lanes 3 (non-reducing) and 4 (reducing) are the trivalent bispecific 2x1 construct "BC1-2x1" (see Example 7). Lanes 5 (non-reducing) and 6 (reducing) are the trivalent 1x2 bispecific construct "CTLA4-4xNivoxCTLA4-4" (see Example 10). Lanes 7 (non-reducing) and 8 (reducing) are the trivalent 1x2 trispecific "BC28-1x1x1a" construct described in Example 11.

[0487] SDS-PAGE gels demonstrate the complete assembly of each construct, and for each construct the major band on the non-reducing gel is shown at the expected molecular weight. 6.13.14. Example 13 Binding analysis

[0488] Figure 33 shows Octet binding analysis to three antigens: PD1, antigen "A," and CTLA-4. In each case, the antigen is immobilized and the B-body is the analyte. For reference, the 1x1 bispecifics "BC1" and "CTLA4-4xOX40-8" are also compared, demonstrating that the 1x1 B-body specifically binds only to antigens for which the antigen-binding site is selected.

[0489] Figure 33A shows that "BC1" binds to PD1 and antigen "A" but not to CTLA4. Figure 33B shows that the bivalent, bispecific 1x1 construct "CTLA4-4xOX40-8" binds to CTLA4 but not to antigen "A" or PD1. Figure 33C shows that the trivalent, trispecific 1x2 construct "BC28-1x1x1a" binds to PD1, antigen "A" and CTLA4. 6.13.15. Example 14 Tetravalent constructs

[0490] Figure 35 shows the overall organization of the 2x2 tetravalent bispecific construct "BC22-2x2." The 2x2 tetravalent bispecific was constructed on the "BC1" scaffold by duplicating each variable domain-constant domain segment. The domain nomenclature is outlined in Figure 34.

[0491] Figure 36 is an SDS-PAGE gel. Lanes 7-9 show the "BC22-2x2" tetravalent construct after one-step purification using CaptureSelect™ CH1 affinity resin ("CH1 eluate") and after further ion-exchange chromatography purification (lane 8, "pk1 after IEX"; lane 9, "pk2 after IEX"). Lanes 1-3 are the trivalent 2x1 construct "BC21-2x1" after CH1 affinity purification (lane 1), and lanes 2 and 3 are the trivalent 2x1 construct "BC21-2x1" after further ion-exchange chromatography. Lanes 4-6 are the 1x2 trivalent construct "BC12-1x2."

[0492] Figure 37 shows the overall organization of the 2x2 tetravalent construct.

[0493] Figures 39 and 40 are schematic diagrams of tetravalent constructs with alternative configurations. The domain nomenclature is shown in Figure 38. 6.13.16. Example 15 Bispecific antigen engagement by B-bodies

[0494] A tetravalent, bispecific 2x2 B-Body "B-Body-IgG 2x2" was constructed. More specifically, using the domain and polypeptide chain nomenclature references summarized in Figure 38, First Polypeptide Chain Domain A = VL (certolizumab) Domain B = CH3 (IgG1, knob) Domain D=CH2 (IgG1) Domain E=CH3(IgG1) Domain W = VH (antigen "A") Domain X=CH1(IgG1) a third polypeptide chain (identical to the first polypeptide chain) Domain H=VL (certolizumab) Domain I = CH3 (IgG1, knob) Domain J=CH2 (IgG1) Domain K=CH3 (IgG1) Domain WW = VH (antigen "A") Domain XX=CH1(IgG1) Second polypeptide chain Domain F = VH (certolizumab) Domain G = CH3 (IgG1, whole) a fourth polypeptide chain (identical to the third polypeptide chain) Domain F = VH (certolizumab) Domain G = CH3 (IgG1, whole) Seventh polypeptide chain Domain Y = VH ("Antigen A") Domain Z=CL Kappa Eighth polypeptide chain (identical to the seventh polypeptide chain) Domain YY = VH ("Antigen A") Domain ZZ=CL Kappa.

[0495] This was cloned and expressed as described in Example 1. Here, the BLI experiment consisted of immobilizing biotinylated antigen "A" onto a streptavidin sensor, followed by establishing a baseline using 10x kinetic buffer. The sensor was then immersed in cell-free expressed "B-Body-IgG 2x2," and a new baseline was then established. Finally, the sensor was immersed in 100 nM kinetic buffer, where a second binding event was observed. Exposure to TNFα confirmed bispecific binding of both antigens by a single "B-Body-IgG 2x2" construct. The results are shown in Figure 41. 6.13.17. Example 16 Antigen-specific cell binding of "BB-IgG 2x2"

[0496] Expi-293 cells were mock transfected or transiently transfected with antigen "B" using an Expi-293 transfection kit (Life Technologies). 48 hours after transfection, Expi-293 cells were harvested and fixed in 4% paraformaldehyde for 15 minutes at room temperature. Cells were washed twice in PBS. 200,000 antigen B or mock-transfected Expi-293 cells were placed in 100 μL of PBS in a V-bottom 96-well plate. Cells were incubated with "B-Body-IgG 2x2" at a concentration of 3 μg / mL for 1.5 hours at room temperature. Cells were centrifuged at 300 × G for 7 minutes, washed with PBS, and incubated with 100 μL of FITC-labeled goat anti-human secondary antibody at a concentration of 8 μg / mL for 1 hour at room temperature. The cells were centrifuged at 300×G for 7 minutes, washed with PBS, and cell binding was confirmed by flow cytometry using a Guava easyCyte. The results are shown in Figure 42. 6.13.18. Example 17 SDS-PAGE analysis of bivalent and trivalent constructs

[0497] Figure 45 shows SDS-PAGE gels under non-reducing and reducing conditions of various constructs after transient expression and one-step purification using CaptureSelect™ CH1 affinity resin, respectively.

[0498] Lanes 1 (non-reducing conditions) and 2 (reducing conditions, +DTT) are the bivalent 1x1 bispecific construct "BC1." Lanes 3 (non-reducing) and 4 (reducing) are the bivalent 1x1 bispecific construct "BC28" (see Example 4). Lanes 5 (non-reducing) and 6 (reducing) are the bivalent 1x1 bispecific construct "BC44" (see Example 5). Lanes 7 (non-reducing) and 8 (reducing) are the trivalent 1x2 bispecific "BC28-1x2" construct (see Example 9). Lanes 9 (non-reducing) and 10 (reducing) are the trivalent 1x2 trispecific "BC28-1x1x1a" construct described in Example 11.

[0499] SDS-PAGE gels demonstrate the complete assembly of each construct, and for each construct the major band on the non-reducing gel is shown at the expected molecular weight. 6.13.19. Example 18 Stability analysis of variable-CH3 junction operation

[0500] The pairing stability between various junction variant combinations was evaluated. Differential scanning fluorimetry was performed to determine the melting temperatures of various junction variant pairings between the VL-CH3 polypeptide (domains A and B) of chain 1 and the VH-CH3 polypeptide (domains F and G) of chain 2. The junction variants "BC6jv," "BC28jv," "BC30jv," "BC44jv," and "BC45jv" (which have the corresponding junction sequences of "BC6," "BC28," "BC30," "BC44," and "BC45," respectively, as seen in Tables 2 and 3 above) have Tms in the range of 76-77°C, demonstrating increased pairing stability (see Table 4). Figure 46 shows the differences in the thermal transitions of "BC24jv," "BC26jv," and "BC28jv," demonstrating that "BC28jv" has the greatest stability of the three. The x-axis of the figure is temperature and the y-axis is the change in fluorescence divided by the change in temperature (-dFluor / dTemp). Experiments were performed as described by Niesen et al. (Nature Protocols, (2007) 2, 2212-2221), which is incorporated herein by reference for all its teachings. [Table 4] 6.13.20. Example 19 ROR×CD3 candidate binding molecule

[0501] As described below, various ROR×CD3 antibodies were constructed and tested. CD3-binding arm

[0502] A series of CD3-binding arm variants based on a humanized version of the SP34 anti-CD3 antibody (SP34-89, SEQ ID NOS: 68 and 69) were engineered by point mutations in either the VH or VL amino acid sequences (SEQ ID NOS: 70-73). Various VH and VL sequences were paired together as described in Table 5. [Table 5]

[0503] The VL and VH variants were cloned into one arm of a 1x1 BC1 B-body, while the other arm contained an irrelevant antigen-binding site. Figure 47 demonstrates the binding affinity of the non-mutagenized SP34-89 monovalent B-body as determined by Octet (Pall ForteBio) biolayer interferometry analysis. Using two-fold serial dilutions of the construct (200 to 12.5 nM), a binding affinity of 23 nM for SP34-89 was determined (k on =3×10 5 M -1 s -1 , k off =7.1×10 -3 s -1 ), which matches the affinity for other SP34 variants in the literature. The kinetic affinity also matched the equilibrium binding affinity. 6.13.20.2.ROR Binding Arm

[0504] Chemically synthesized Fab phage libraries with diversity introduced into the Fab CDRs were screened against ROR antigens using a monoclonal phage ELISA format, whereby plate-immobilized ROR variants were evaluated for phage binding as described above. Phage clones expressing Fabs that recognized the antigen were sequenced. The first screening campaign for binding to ROR1 identified an antigen-binding site (ABS) clone designated "I2A" in Table 6, and the second screening campaign for binding to ROR2 identified an ABS clone designated "I2C" in Table 6. [Table 6-1] [Table 6-2] [Table 6-3] "I2A", screening against ROR1; "I2C", screening against ROR2; * indicates unsequenced; bold indicates potential isomerization sites (see below).

[0505] The VH and VL sequences described above were formatted into a bivalent, monospecific, native human, full-length IgG1 architecture. Figures 48A-48B demonstrate Octet (Pall ForteBio) biolayer interferometry analysis of two-fold serial dilutions (200-12.5 nM) against two ROR-binding candidates (Figure 48A, clone I2-A10; Figure 48B, clone I2-A27). Binding affinities and kinetics were determined from the serial dilutions and are shown in Table 7. [Table 7]

[0506] The ROR-binding candidates formatted into a bivalent, monospecific, native human, full-length IgG1 architecture were further characterized for binding to ROR1 and / or ROR2. Table 8 presents candidates that specifically bound only to ROR1, candidates that specifically bound only to ROR2, and candidates that were cross-reactive to both ROR1 and ROR2. The ROR-binding candidates were also characterized for their binding to specific ROR1 domains. Table 9 presents candidates that specifically bound to the Frizzled, Ig-like, and Kringle domains. [Table 8] [Table 9]

[0507] Selected ROR-binding candidates were further analyzed for sequence motifs that may adversely affect antibody properties relevant to clinical development, such as stability, variability, and immunogenicity. Computational analysis was performed according to Kumar and Singh (Developability of biotherapeutics: computational approaches. Boca Raton: CRC Press, Taylor & Francis Group, 2016). The analysis results, presented in Table 10, demonstrate that a limited number of deleterious sequence motifs are present in the included clones, illustrating their potential for further clinical development. [Table 10] *The predicted T cell epitopes found in Herceptin are present in these molecules 6.13.21. Example 20 In vitro efficacy of ROR×CD3 bispecific B-body

[0508] Candidate ROR and CD3 antigen binding sites were formatted into B-body BC1 1x1 and 1x2 formats and tested in a range of tumor efficacy models. 6.13.21.1. Bispecific B-body format comparison

[0509] The ROR antigen binding site (ABS) candidate I2A-3 and the CD3 ABS candidate SP34-89 were formatted into bispecific B-body "BC1" 1x1 and 1x2 formats. Referring to Figures 3 and 26, the ROR ABS candidate forms A:F and R:T binding sites, while the CD3 ABS candidate forms an H:L binding site. Referring to Figure 26, two 1x2 formats were constructed with either a 10-amino acid junction or a 16-amino acid junction between the S and H domains. The different constructs were tested in the NFκB GFP Jurkat T cell stimulation assay described herein. Briefly, reporter T cells (effector cells) were mixed with either the non-small cell lung cancer target tumor cell line HOP-92, which expresses the ROR1 antigen, or the B16 melanoma target tumor cell line, which does not express ROR1. A dilution series of the different B-body constructs was then incubated with the effector:target mixture.

[0510] As shown in Figure 49 and presented in Table 11, the RORxCD3 bispecific 1x1 and 1x2 B-bodies resulted in reporter T cell activation when mixed with a ROR1-expressing tumor line (HOP-92), whereas no activation was observed when the RORxCD3 bispecific 1x1 and 1x2 B-bodies were mixed with a tumor line (B16) that does not express ROR1. Additionally, the 1x2 B-body format, bearing bivalent specificity for ROR1, was more potent than the 1x1 B-body format, and variation in junction length resulted in minimal differences in potency. [Table 11] 6.13.21.2. CD3 Binding Alone Does Not Activate T Cells

[0511] The ROR Ab candidate I2A-10 and the CD3 Ab candidate SP34-89 were formatted into a bispecific B-body "BC1" 1x2 format ("I2-A10 1x2"). In a separate construct, a control arm against a tumor antigen other than ROR1 was also formatted into a bispecific B-body "BC1" 1x2 format along with the CD3 Ab candidate SP34-89 ("Neg Ctrl"). Referring to Figures 3 and 26, the ROR candidate Ab and the "Neg Ctrl" arm Ab form A:F and R:T binding sites, while the CD3 Ab candidate forms an H:L binding site. The different constructs were tested in a T cell cytotoxicity assay. Briefly, isolated CD8+ T cells (effector cells) were mixed with the triple-negative breast cancer tumor cell line MDA-MD-231 (target cells) expressing the ROR1 antigen. A dilution series of the different B-body constructs was then incubated with the effector:target mixture.

[0512] As shown in Figure 50, the RORxCD3 trivalent bispecific 1x2 B-body resulted in cytotoxic T cell-mediated killing when mixed with a ROR1-expressing tumor line (MDA-MD-231), but did not result in cytotoxicity when a CD3 bispecific B-body with irrelevant tumor ABS (e.g., a tumor antigen not expressed in MDA-MD-231) was added to the mixture. RORxCD3 Bispecific B-Body Efficacy in Multiple Tumor Models

[0513] The ROR1 Ab candidate I2A-3 and CD3 Ab candidate SP34-89 were formatted into the bispecific B-body "BC1" 1x1 and 1x2 formats described above, with the 1x2 format having a 10-amino acid junction between the S and H domains. The different constructs were tested in the NFκB GFP Jurkat T cell stimulation assay described herein. Briefly, reporter T cells (effector cells) were mixed with ROR1-expressing tumor cell lines: HOP-92 (non-small cell lung cancer 1), A549 (non-small cell lung cancer 2), MDA-MD-231 (triple-negative breast cancer), JeKo-1 (mantle cell lymphoma), and RPMI-8226 (multiple myeloma). The constructs were also mixed with the B16 melanoma tumor cell line (target cells), which does not express ROR1. A dilution series of the different B-body constructs was then incubated with the effector:target mixture.

[0514] As shown in Figures 51A-51E and presented in Table 12, the RORxCD3 bispecific 1x1 and 1x2 B-bodies resulted in reporter T cell activation when mixed with the ROR1-expressing tumor lines, HOP-92 (Figure 51A), A549 (Figure 51B), MDA-MD-231 (Figure 51C), JeKo-1 (Figure 51D), and RPMI-8226 (Figure 51E), but no activation occurred when these constructs were mixed with a tumor line (B16) that does not express ROR1. The 1x2 B-body format, bearing bivalent specificity for ROR1, was more potent than the 1x1 B-body format in all tumor lines tested. Thus, the RORxCD3 bispecific 1x1 and 1x2 B-bodies demonstrated efficacy across a range of tumor models. [Table 12] 6.13.21.4. ROR ABS Candidate Screening for T Cell Activation

[0515] ROR Ab candidates I2A-1, I2A-3, I2A-10, I2A-14, I2A-16, I2A-20, I2A-22, and I2A-27, along with CD3 Ab candidate SP34-89, were formatted into the bispecific B-body "BC1" 1x2 format described above, which had a 10-amino acid junction between the S and H domains. The different constructs were tested in the T cell cytotoxicity assay described above. Briefly, isolated CD8+ T cells (effector cells) were mixed with the triple-negative breast cancer tumor cell line MDA-MD-231 (target cells), which expresses the ROR1 antigen, at an E:T ratio of 6:1. Next, dilution series of the different B-body constructs were incubated with the effector:target mixture.

[0516] As shown in Figure 52 and presented in Table 13, the RORxCD3 bispecific 1x2 B-body I2A-1, I2A-3, I2A-10, I2A-14, I2A-22, and I2A-27 conferred cytotoxic T cell-mediated killing when mixed with a ROR1-expressing tumor line (MDA-MD-231), whereas 1x2 B-bodies I2A-16 and I2A-20 did not confer potent cytotoxicity. In addition, I2A-22 conferred an anomalous dose-response curve. Thus, ROR Ab candidates I2A-1, I2A-3, I2A-10, I2A-14, and I2A-17 conferred effective T cell-mediated killing. [Table 13] 6.13.21.5. Cytotoxic Killing Correlates with ROR1 Expression

[0517] The ROR1 Ab candidate I2A-10 and the CD3 Ab candidate SP34-89 were formatted into the bispecific B-body "BC1" 1x2 format described above, which had a 10-amino acid junction between the S and H domains. The I2A-10 candidate was tested in a T cell cytotoxicity assay as described. Briefly, isolated CD8+ T cells (effector cells) were mixed with ROR1-expressing tumor cell lines, MDA-MD-231 (triple-negative breast cancer) and RPMI-8226 (multiple myeloma), at an E:T ratio of 4:1. A dilution series of the candidate was incubated with the effector:target mixture.

[0518] Figure 53A illustrates published ROR1 expression data for the MDA-MD-231 and RPMI-8226 tumor lines. Figures 53B and 53C demonstrate that the cytotoxic efficacy observed in our experiments correlates with ROR1 in the MDA-MD-231 and RPMI-8226 tumor cell lines. 6.13.21.6. Primary T cell activation by ROR ABS candidates I2A-10 and I2A-27

[0519] The ROR ABS candidates I2A-10 and I2A-27, along with the CD3 ABS candidate SP34-89, were formatted into the bispecific B-body "BC1" 1x2 format described above, which had a 10 amino acid junction between the S and H domains. The amino acid sequences of the four polypeptide chains that make up the I2A-10 BC1 1x2 B-body are listed in SEQ ID NOs: 84, 85, 87, and 88. The amino acid sequences of the four polypeptide chains that make up the I2A-27 BC1 1x2 are listed in SEQ ID NOs: 89, 90, 92, and 93.

[0520] The different constructs were tested in a T cell activation assay quantified by flow cytometry. Briefly, isolated peripheral blood mononuclear cells (PBMCs, effector cells) were mixed with the triple-negative breast cancer tumor cell line MDA-MD-231 and pancreatic cancer tumor cell line PANC1 (target cells), which express the ROR1 antigen, at an E:T ratio of 7:1. Next, dilutions of the different B-body constructs were incubated with the effector:target mixture and allowed to incubate together for 44 hours. Cells were then stained for T cell markers CD3, CD4, and CD8, as well as activation markers CD25 and CD69, and analyzed by flow cytometry.

[0521] As shown in Figures 54A-F, I2A-10 and I2A-27 B-bodies activated CD8+ T cells in PBMC populations as determined by expression of CD25 (Figure 54A), CD69 (Figure 54C), and both CD25 and CD69 (Figure 54E), and activated CD4+ T cells in PBMC populations as determined by expression of CD25 (Figure 54B), CD69 (Figure 54D), and both CD25 and CD69 (Figure 54F). Thus, ROR Ab candidates can activate primary T cells. 6.13.21.7. Internalization of ROR ABS Candidates I2A-10 and I2A-27

[0522] ROR Ab candidates I2A-10 and I2A-27 were formatted, along with CD3 Ab candidate SP34-89, into the bispecific B-body "BC1" 1x2 format described above, which had a 10-amino acid junction between the S and H domains. The different constructs were tested for internalization by tumor cell lines, as quantified by flow cytometry. Briefly, MDA-MB-231 cells were incubated with I2-A10, I2-A27, or an isotype control for 2 hours at 37°C or 4°C. After 2 hours, labeled secondary antibodies were added for 30 minutes at 4°C and then analyzed by flow cytometry. Percent internalization was calculated based on mean fluorescence intensity (MFI) by normalizing between the isotype control (0%) and the 4°C control (100%).

[0523] As shown in Figure 55, 26% and 36% of candidates I2A-10 (upper panel) and I2A-27 (lower panel), respectively, were internalized after 2 hours of incubation with MDA-MB-231 cells. Internalization by tumor cells enables various antibody-drug conjugate strategies to kill tumor cells expressing ROR antigens. 6.13.22. Example 21 Single-step purification of ROR×CD3 bispecific B-body

[0524] ROR ABS candidates I2A-10 and I2A-27 were formatted with CD3 ABS candidate SP34-89 into the bispecific B-body "BC1" trivalent 1x2 format described above, which had a 10 amino acid junction between the S and H domains. Constructs were purified using a one-step purification using CaptureSelect™ CH1 affinity resin.

[0525] Figure 56 shows size exclusion chromatography (SEC) analysis demonstrating that the single-step CH1 affinity purification step yielded a single monodisperse peak by gel filtration, with >98% unaggregated protein for 1x2 B-body candidates I2A-10 (top panel) and I2A-27 (bottom panel).

[0526] Figure 57A shows a non-reducing SDS-PAGE gel of 1x2 B-body candidates I2A-10 (left panel) and I2A-27 (right panel), demonstrating the major band of the fully assembled construct (high-migration 250 kDa band). Figure 57B shows a Bioanalyzer (Agilent) analysis of non-reduced samples for 1x2 B-body candidates I2A-10 and I2A-27, demonstrating the major band of the fully assembled construct.

[0527] The anti-CH1 purification efficiency of bispecific antibodies was also examined for ROR-binding molecules with only standard knob-and-hole orthogonal mutations introduced into the CH3 domain found in its native position within the Fc portion of the bispecific antibody, with no other domain modifications. Thus, the two antibodies tested, KL27-6 and KL27-7, each contained two CH1 domains, one in each arm of the antibody. As described in more detail in Section 6.13.1, each bispecific antibody was expressed, purified from unwanted protein products on an anti-CH1 column, and run on an SDS-PAGE gel. As shown in Figure 58, there was a significant band at 75 kDa representing an incomplete bispecific antibody, which, with reference to Figure 3, could be interpreted as a complex containing only (i) the first and second, or (ii) the third and fourth polypeptide chains. Thus, using anti-CH1 to purify a complete bispecific molecule with a CH1 domain in each arm resulted in background contamination from incomplete antibody complexes. 6.13.23. Example 22 Fc mutations that reduce effector function

[0528] A series of engineered Fc variants were generated in the monoclonal IgG1 antibody trastuzumab (Herceptin, "WT-IgG1"), which have mutations at positions L234, L235, and P329 in the CH2 domain. The specific mutations of the variants tested are listed in Table 14 below and include sFc1 (PALALA), sFc7 (PGLALA), and sFc10 (PKLALA). All variants showed similar stability as determined by melting temperature (Table 14, TM1 and TM2).

[0529] WT-IgG1 and Fc variants were immobilized on Octet biosensors, and soluble FcγRIa was added to the system to determine binding. Figures 59A-59B show Octet (Pall ForteBio) biolayer interferometry analysis demonstrating FcγRIa binding to trastuzumab (Figure 59A "WT IgG1") but not to sFc10 (Figure 59B). After addition of FcγRIa, an increase in signal was observed for trastuzumab, but no observable increase in signal was detected for sFc10, demonstrating that FcγRIa no longer binds to antibodies with the engineered mutations. A binding summary of the tested variants is presented in Table 14. Additionally, all variants retained strong binding to HER2 (not shown). [Table 14]

[0530] WT-IgG1 and Fc variants were tested in an antibody-dependent cellular cytotoxicity (ADCC) assay as another measure of FcγR binding, particularly FcγRIIIa. As shown in Figure 60, trastuzumab (Herceptin, "WT-IgG1") demonstrated killing, whereas neither sFc7 nor sFc10 produced detectable levels of killing. WT-IgG1 and Fc variants were also tested for complement component C1q binding by ELISA. As shown in Figure 61, trastuzumab (Herceptin, "WT-IgG1") demonstrated C1q binding, whereas neither sFc1, sFc7, nor sFc10 produced detectable C1q binding. Thus, the results demonstrate that the tested Fc variants have reduced levels of Fc effector function. 6.13.24. Example 23 In vivo efficacy of ROR×CD3 bispecific B-body

[0531] In vivo efficacy was measured using xenograft studies in humanized mice. 6 MDA-MB-231 tumor cells were engrafted subcutaneously into the hind flank of NOD Scid gamma (NSG) mice (Jackson Labs) and allowed to grow to approximately 120–150 mm. Next, 1 × 10 cells from a single donor were cultured in the same 3x10 ... 7 Mice were humanized by intravenous (IV) injection of human PBMCs. Humanized NSG mice were randomized into three groups of eight mice. Three days after PBMC engraftment, they were IV-dosed with PBS, 0.5 mg / Kg of ROR ABS candidate I2-A10 ("I2-A10") formatted into a 1x2 B-body architecture with CD3 ABS candidate SP34-89, or 0.5 mg / Kg of ROR ABS candidate I2-A27 ("I2-A27") formatted into a 1x2 B-body architecture with CD3 ABS candidate SP34-89. Dosing was continued twice weekly for three weeks. Animals were monitored for tumor growth, body weight, and overall health. After study completion, animals were sacrificed for analysis. Flow cytometry analysis was performed using standard techniques to determine the humanization status of NSG mice.

[0532] Tumors are collected and analyzed using standard immunohistochemistry techniques to monitor human T cell infiltration into the tumor in response to treatment. IHC is performed using standard techniques. Briefly, FFPE samples are deparaffinized and rehydrated by baking at 60°C and placing in 100% xylene solution, then rehydrated with an ethanol series (100% ethanol, 95% ethanol, 70% ethanol, 50% ethanol, PBS). Antigen retrieval is performed by incubating slides in 10 mM Na citrate + 0.05% Tween® 20, pH 6.0 buffer at 95°C for 10 minutes. The AbCam Mouse on Mouse IHC Kit (Cat. Ab127055) is used according to the manufacturer's instructions for staining. Briefly, endogenous peroxidase activity is blocked with hydrogen peroxide solution, and endogenous nonspecific interactions are blocked with Rodent Block solution. Slides are incubated with primary antibody according to the manufacturer's recommendations (typically 2 hours at room temperature or overnight at 4°C) and then stained with the mouse on mouse HRP polymer from the AbCam kit. DAB chromagen staining and hemotoxylin counterstaining are used for visualization. 6.13.24.1.RORxCD3 Trivalent Bispecific B-Body Results in Tumor Growth Reduction

[0533] As shown in Figures 62A-62C, tumor volume was monitored in mice engrafted with tumor cells, humanized with PBMCs (solid arrows on the left), and then subsequently treated IV with PBS (Figure 62A), 1x2 B-body candidate I2-A10 (Figure 62B), or 1x2 B-body candidate I2-A27 (Figure 62C) (dashed arrows on the right). Figure 63 shows the tumor volume at the conclusion of the study for each mouse, with the mean and standard deviation for each group shown. The open square in the I2-A27 group was removed from the analysis due to likely non-humanization by PBMCs. The results demonstrate that I2-A27 treatment resulted in a significant reduction in tumor growth of approximately 40%.

[0534] The humanized status of the NSG mice in the study was confirmed by flow cytometry analysis. As shown in Table 15, all mice analyzed were successfully humanized with human lymphocytes, as determined by staining for human CD45 (%hu CD45+). Sample 03_B10(b) was not analyzed. The humanized lymphocyte populations in all mice analyzed also contained human T cells, including CD4- and CD8-positive T cells, as determined by staining for human CD3 (%CD3 gated on huCD45+ lymphocytes, and %CD4+ / CD8+ gated on human CD3+ T cells).

[0535] Immunohistochemistry will be performed on tumor samples to confirm human T cell infiltration into the tumor in response to treatment. Table 15: Humanization of NSG mice [Table 15] 6.13.25. Example 24 ROR1×CD3 bispecific B-body(TM) efficacy

[0536] Additional ROR antigen binding molecules containing the six CDRs of I2A-27 shown in Table 6 were prepared and experiments were performed to test their efficacy as trivalent bispecific constructs in the ROR1xCD3 bispecific B-body™ format. Generation of Additional ROR1xCD3 1x2 B-body™ Bispecific Antibodies Based on 6.13.25.1.I2-A27

[0537] ROR ABS candidate I2A-27, along with CD3 ABS candidate SP34-89, was formatted into the bispecific B-body "BC1" 1x2 format described above, which had a 10 amino acid junction between the S and H domains (see, e.g., Figure 26 for a schematic diagram of the domains). The amino acid sequences of the five polypeptide chains that form the exemplary I2A-27 BC1 1x2 trivalent bispecific ROR antigen binding molecule are listed as SEQ ID NO:96 (chain 1), SEQ ID NO:97 (chain 2), SEQ ID NO:98 (chain 3), SEQ ID NO:99 (chain 4), and SEQ ID NO:97 (chain 5).

[0538] In the following description in this Example, the term "I2-A27 1x2 B-body™" or "I2-A27" or "A27" refers to this ROR1xCD3 1x2 B-body™ having strand 1 of SEQ ID NO:96, strand 2 of SEQ ID NO:97, strand 3 of SEQ ID NO:98, strand 4 of SEQ ID NO:99 and strand 5 of SEQ ID NO:97. 6.13.25.2. ROR1xCD3 1x2 B-body™ Bispecific Antibody Binding to CD3 and ROR1

[0539] As described above in Example 19, SP34-89 exhibited a mono...

Claims

[Claim 1] The invention as set forth in the drawings.

Images