ICOS LIGAND VARIANT IMMUNOMODULATING PROTEINS AND USES THEREOF

Abstract

Therapeutic compositions for modulating immune responses in the treatment of cancer and immunological diseases are provided. Provided herein are immunomodulatory proteins, including ICOSL variants, and nucleic acids encoding the proteins, which provide therapeutic utility for a variety of immunological and oncological conditions. Also provided are compositions and methods for producing and using the proteins.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application is a continuation of U.S. Provisional Patent Application No. 62 / 323,608, filed April 15, 2016, entitled "ICOS Ligand Variant Immunomodulatory Proteins and Uses Thereof," U.S. Provisional Patent Application No. 62 / 394,745, filed September 14, 2016, entitled "ICOS Ligand Variant Immunomodulatory Proteins and Uses Thereof," U.S. Provisional Patent Application No. 62 / 410,842, filed October 20, 2016, entitled "ICOS Ligand Variant Immunomodulatory Proteins and Uses Thereof," U.S. Provisional Patent Application No. 62 / 410,842, filed March 16, 2017, entitled "ICOS Ligand Variant Immunomodulatory Proteins and Uses Thereof," U.S. Provisional Patent Application No. This application claims priority to U.S. Provisional Patent Application No. 62 / 472,568, entitled "ICOS Ligand Variant Immunomodulatory Proteins and Uses Thereof," filed March 22, 2017, and U.S. Provisional Patent Application No. 62 / 475,162, entitled "ICOS Ligand Variant Immunomodulatory Proteins and Uses Thereof," filed March 22, 2017, the contents of each of which are incorporated by reference in their entirety.

[0002] INCORPORATION-BY-REFERENCE TO SEQUENCE LISTING This application is filed with a Sequence Listing in electronic format, which is provided as a 979,474 byte file entitled 761612000340SeqList.txt, created on April 13, 2017. The information in the electronic format of the Sequence Listing is incorporated by reference in its entirety.

[0003] Field The present disclosure relates to therapeutic compositions for modulating immune responses in the treatment of cancer and immunological diseases. In some aspects, the present disclosure relates to specific variants of ICOS ligand (ICOSL) that exhibit improved binding affinity to one or both of the cognate binding partner proteins ICOS or CD28. [Background technology]

[0004] background There is growing medical interest in modulating immune responses by intervening in processes occurring at the immune synapse (IS), the region formed between antigen-presenting cells (APCs) or target cells and lymphocytes. Mechanistically, cell surface proteins within the IS can engage in coordinated, often simultaneous, interactions between multiple protein targets and a single protein to which they bind. Interactions at the IS occur in close association with the junction of two cells, and a single protein within this structure can interact (possibly simultaneously) with proteins on both the same cell (cis) and interacting cells (trans). While therapeutic agents capable of modulating the IS are known, improved therapeutic agents are needed. We provide immunomodulatory proteins, including soluble or transmembrane immunomodulatory proteins that can be expressed on cells, that meet this need. Summary of the Invention

[0005] overview In some embodiments, a variant ICOS ligand (ICOSL) polypeptide comprising an IgV domain or a specific binding fragment thereof, an IgC domain or a specific binding fragment thereof, or both, is provided, comprising SEQ ID NO: Based on NO:32, 10, 11, 13, 16, 18, 20, 25, 27, 30, 33, 37, 38, 42, 43, 47, 52, 54, 57, 61, 62, 67, 71, 72, 74, 75, 77, 78, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98, 99, 100, 102, 103, 107, 109, 110, 111, 113, 115, 116, 117, 119, 120, 121, 122, 126, 129, 130, 132, 133, 135, 138, 139, 140, 142, 143, 144 Provided herein are variant ICOSL polypeptides containing one or more amino acid substitutions in unmodified ICOSL or its specific-binding fragments corresponding to positions selected from 4, 146, 148, 151, 152, 153, 154, 155, 156, 158, 161, 164, 166, 168, 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210, 212, 217, 218, 220, 221, 224, 225, or 227. In some embodiments, the unmodified ICOSL is mammalian ICOSL or its specific-binding fragment. In some embodiments, the unmodified ICOSL is human ICOSL or its specific-binding fragment. In some embodiments, the unmodified ICOSL comprises (i) the amino acid sequence set forth in SEQ ID NO:32; (ii) an amino acid sequence having at least 95% sequence identity to SEQ ID NO:32; or (iii) a portion thereof, which comprises an IgV domain or an IgC domain or a specific binding fragment thereof, or both.

[0006] In some embodiments of any one of the above variant ICOSL polypeptides, the specific binding fragment of the IgV domain or IgC domain has a length of at least 50, 60, 70, 80, 90, 100, 110, or more amino acids; or the specific binding fragment of the IgV domain comprises at least 80% of the length of the IgV domain set forth as amino acids 19 to 129 of SEQ ID NO: 5, and / or the specific binding fragment of the IgC domain comprises at least 80% of the length of the IgC domain set forth as amino acids 141 to 227 of SEQ ID NO: 5. In some embodiments, the variant ICOSL polypeptide comprises up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid modifications (optionally, amino acid substitutions, insertions, and / or deletions). In some embodiments, the ICOSL variant comprises an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to SEQ ID NO: 32 or a specific-binding fragment thereof. In some embodiments, any of the variant ICOSL polypeptides provided herein have altered binding to the ectodomain of ICOS, CD28, or CTLA-4 compared to unmodified ICOSL. In some embodiments, any of the variant ICOSL polypeptides have altered binding to the ectodomain of ICOS or CD28 compared to unmodified ICOSL. In some embodiments, the altered binding is an altered binding affinity and / or an altered binding selectivity.

[0007] In some embodiments of any one of the above variant ICOSL polypeptides, the one or more amino acid substitutions are TIFF0007623967000001.tif84164 or a conservative amino acid substitution thereof.

[0008] In some embodiments, the one or more amino acid modifications are: Selected from TIFF0007623967000002.tif216166, TIFF0007623967000003.tif223166, and TIFF0007623967000004.tif33162.

[0009] In some of any such embodiments, the one or more amino acid modifications are at a position corresponding to a position selected from 52, 57, 100, 110, or 198. In some of any such embodiments, the one or more amino acid modifications are at a position corresponding to a position selected from: TIFF0007623967000005.tif4166 or conservative amino acid substitutions thereof. In some embodiments, the variant ICOSL further comprises one or more additional modifications (e.g., any of the modifications described herein). In some of any such embodiments, the variant ICOSL polypeptide comprises: TIFF0007623967000006.tif26161 or conservative amino acid substitutions thereof.

[0010] In some of any such embodiments, the one or more amino acid modifications are: The file is TIFF0007623967000007.tif48164.

[0011] In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL polypeptide comprises an IgV domain, or a specific fragment thereof, and an IgC domain, or a specific fragment thereof. In some embodiments of any one of the above-described variant ICOSL polypeptides, the variant ICOSL polypeptide comprises an amino acid sequence set forth in any of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470, or a specific binding fragment thereof, or comprises an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to any of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470, or a specific binding fragment thereof, and contains one or more amino acid substitutions.

[0012] In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL polypeptide comprises an IgV domain or a specific-binding fragment thereof. In some embodiments, the IgV domain or a specific-binding fragment thereof is the only ICOSL portion of the variant ICOSL polypeptide. In some embodiments, the IgC domain or a specific-binding fragment thereof is the only ICOSL portion of the variant ICOSL polypeptide.

[0013] In some embodiments of any one of the above-described variant ICOSL polypeptides, the variant ICOSL polypeptide comprises an amino acid sequence set forth in any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426, and 434, or a specific binding fragment thereof, that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to any of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426, and 434, or a specific binding fragment thereof, and contains one or more amino acid substitutions.

[0014] In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS, CD28, or CTLA-4 with improved affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS or CD28 with improved affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS and the ectodomain of CD28, each with improved affinity compared to unmodified ICOSL.

[0015] In some of any such embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of CD28 with improved affinity compared to unmodified ICOSL. In some of any such embodiments, the improved affinity for the ectodomain of CD28 is greater than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 30-fold, 40-fold, 50-fold, or 60-fold compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS with improved affinity compared to unmodified ICOSL. In some of any such embodiments, the improved affinity for the ectodomain of ICOS is greater than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 40-fold, 50-fold, 60-fold, or 70-fold compared to unmodified ICOSL.

[0016] In some embodiments of any one of the variant ICOSL polypeptides, the one or more amino acid substitutions correspond to positions selected from 52, 54, or 57. In some embodiments, the one or more amino acid modifications correspond to positions selected from: TIFF0007623967000008.tif4128 or conservative amino acid substitutions thereof. In some of any such embodiments, the one or more amino acid substitution modifications correspond to positions selected from 52 or 57. In some embodiments, the one or more amino acid substitutions are selected from N52H, N52D, N52S, N52K, or N57Y. In some embodiments, any one of the variant ICOSL polypeptides further comprises one more additional amino acid modification (e.g., any of the amino acid modifications described). In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL polypeptide comprises: TIFF0007623967000009.tif33164 or conservative amino acid substitutions thereof.

[0017] In some of any such embodiments, the one or more amino acid modifications are: Selected from TIFF0007623967000010.tif92155TIFF0007623967000011.tif158164.

[0018] In some of any such embodiments, the one or more amino acid modifications are: In some of any such embodiments, the one or more amino acid modifications are selected from among: Selected from TIFF0007623967000013.tif19145.

[0019] In some of any such embodiments, the variant ICOSL polypeptide specifically binds to the CTLA-4 ectodomain with improved affinity compared to unmodified ICOSL, hi some aspects, the improved affinity for the CTLA-4 ectodomain is greater than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 40-fold, 50-fold, 60-fold, or 70-fold compared to unmodified ICOSL.

[0020] In some of any such embodiments, the variant polypeptide specifically binds to the ectodomain of ICOS, CD28, or CTLA4 with improved selectivity compared to unmodified ICOSL. In some embodiments, the improved selectivity comprises a greater ratio between the binding of the variant polypeptide to one cognate binding partner selected from among ICOS, CD28, and CTLA4 and the binding of the variant polypeptide to another of the cognate binding partners compared to the binding of the unmodified ICOSL polypeptide to the one cognate binding partner and the other of the cognate binding partners. In some cases, the ratio is at least or at least about 1.5-fold, 2.0-fold, 3.0-fold, 4.0-fold, 5-fold, 10-fold, 15-fold, 20-fold, 30-fold, 40-fold, 50-fold, or more.

[0021] In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide comprises: TIFF0007623967000014.tif84164, or conservative amino acid substitutions thereof. In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide further comprises one or more amino acid deletions corresponding to position 140 of SEQ ID NO:32.

[0022] In some embodiments of any one of the variant ICOSL polypeptides, the one or more amino acid substitutions are It is selected from TIFF0007623967000015.tif85159.

[0023] In some embodiments of any one of the variant ICOSL polypeptides, the one or more amino acid substitutions are Selected from TIFF0007623967000016.tif84152.

[0024] In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS or CD28 with improved affinity compared to unmodified ICOSL, and specifically binds to the other ectodomain of ICOS or CD28 with reduced affinity compared to unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of ICOS with improved affinity compared to unmodified ICOSL, and specifically binds to the ectodomain of CD28 with reduced affinity compared to unmodified ICOSL. In some embodiments, the one or more amino acid substitutions are Selected from TIFF0007623967000017.tif76163.

[0025] In some embodiments, the variant ICOSL polypeptide specifically binds to the ectodomain of CD28 with improved affinity compared to unmodified ICOSL and specifically binds to the ectodomain of ICOS with reduced affinity compared to unmodified ICOSL, hi some embodiments, the variant ICOSL polypeptide comprises the amino acid substitutions N52S / R75Q / L203P or N30D / K42E / N52S.

[0026] In some embodiments of any one of the variant ICOSL polypeptides, ICOS is human ICOS. In some embodiments, CD28 is human CD28.

[0027] In some embodiments of any one of the variant ICOSL polypeptides, binding is altered (improved or decreased) by more than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 30-fold, 40-fold, or 50-fold compared to unmodified ICOSL.

[0028] In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide is a soluble protein.

[0029] In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide is linked to a multimerization domain. In some embodiments, the variant ICOSL polypeptide is a multimeric polypeptide (optionally a dimeric polypeptide) comprising a first variant ICOSL polypeptide linked to a multimerization domain and a second variant ICOSL polypeptide linked to a multimerization domain. In some embodiments, the first variant ICOSL polypeptide and the second variant ICOSL polypeptide are the same or different. In some embodiments, the multimerization domain is an Fc domain or a variant thereof with reduced effector function. In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide is linked to a moiety that extends the biological half-life of the polypeptide.

[0030] In some embodiments, the variant ICOSL polypeptide is linked to an Fc domain or variant thereof having reduced effector function. In some embodiments, the Fc domain is mammalian (optionally human); or the variant Fc domain comprises one or more amino acid modifications compared to an unmodified mammalian (optionally human) Fc domain. In some embodiments, the Fc domain or variant thereof comprises the amino acid sequence set forth in SEQ ID NO:226 or SEQ ID NO:227, or an amino acid sequence exhibiting at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to SEQ ID NO:226 or SEQ ID NO:227.

[0031] In some embodiments, the Fc domain comprises one or more amino acid modifications selected from among E233P, L234A, L234V, L235A, L235E, G236del, G237A, S267K, R292C, N297G, and V302C, each according to EU numbering. In some aspects, the Fc domain comprises the amino acid modification C220S according to EU numbering. In some embodiments, the Fc domain comprises an amino acid sequence set forth in any of SEQ ID NOs:474, 476, 477, 478, or an amino acid sequence that exhibits at least 85% sequence identity to any of SEQ ID NOs:474, 476, 477, 478 and contains one or more amino acid modifications and / or exhibits reduced effector function.

[0032] In some embodiments, the variant ICOSL polypeptide is indirectly linked to the multimerization domain or Fc via a linker (optionally a G4S linker).

[0033] In some embodiments of any one of the variant ICOSL polypeptides, the variant ICOSL polypeptide is a transmembrane immunomodulatory protein further comprising a transmembrane domain linked to the extracellular domain (ECD) or specific-binding fragment thereof of the variant ICOSL polypeptide. In some embodiments, the transmembrane domain comprises the amino acid sequence set forth as residues 257-277 of SEQ ID NO:5, or a functional variant thereof exhibiting at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to residues 257-277 of SEQ ID NO:5. In some embodiments, the variant ICOSL polypeptide further comprises a cytoplasmic signaling domain linked to the transmembrane domain. In some embodiments, the cytoplasmic signaling domain comprises the amino acid sequence set forth as residues 278-302 of SEQ ID NO:5, or a functional variant thereof that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to residues 278-302 of SEQ ID NO:5.

[0034] In some of any such embodiments, the variant ICOSL polypeptide comprises an amino acid sequence set forth in any of SEQ ID NOs:494-503, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to any of SEQ ID NOs:494-503 and includes one or more amino acid modifications as described.

[0035] In some embodiments of any one of the variant ICOSL polypeptides, the ICOSL variant increases IFN-γ (interferon-γ) expression compared to unmodified ICOSL in an in vitro primary T cell assay. In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL decreases IFN-γ (interferon-γ) expression compared to unmodified ICOSL in an in vitro primary T cell assay. In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL polypeptide is deglycosylated.

[0036] In some embodiments, provided herein are immunomodulatory proteins comprising a variant ICOSL polypeptide according to any one of the provided embodiments linked to a second polypeptide comprising an immunoglobulin superfamily (IgSF) domain. In some embodiments, the IgSF domain has modified affinity and altered binding to one or more of its cognate binding partners compared to the unmodified or wild-type IgSF domain. In some embodiments, the IgSF domain has improved binding to one or more of its cognate binding partners compared to the unmodified or wild-type IgSF domain. In some embodiments, the variant ICOSL polypeptide is a first variant ICOSL polypeptide, and the IgSF domain of the second polypeptide is an IgSF domain derived from a second variant ICOSL polypeptide according to any one of the provided embodiments, wherein the first and second variant ICOSL polypeptides are the same or different. In some embodiments, the variant ICOSL polypeptide is capable of specifically binding to CD28 or ICOS, and the IgSF domain of the second polypeptide is capable of binding to a cognate binding partner other than that to which the variant ICOSL polypeptide specifically binds. In some embodiments, the IgSF domain is derived from a member of the B7 family. In some embodiments, the IgSF domain is a tumor-localizing moiety that binds to a ligand expressed on a tumor, or an inflammation-localizing moiety that binds to a ligand expressed on cells or tissues in an inflammatory environment. In some embodiments, the IgSF domain is a tumor-localizing moiety that binds to a ligand expressed on a tumor. In some embodiments, the ligand is B7H6. In some embodiments, the IgSF domain is derived from NKp30. In some embodiments, the IgSF domain is or comprises an IgV domain. In some embodiments, the variant ICOSL polypeptide is or comprises an IgV domain.

[0037] In some embodiments of any one of the immunomodulatory proteins, the immunomodulatory protein comprises a multimerization domain linked to one or both of a variant ICOSL polypeptide or a second polypeptide comprising an IgSF domain. In some embodiments, the multimerization domain is an Fc domain or a variant thereof with reduced effector function. In some embodiments, the immunomodulatory protein is a dimer. In some embodiments, the immunomodulatory protein is a homodimer. In some cases, the immunomodulatory protein is a heterodimer.

[0038] In some embodiments, provided herein is a conjugate comprising a variant ICOSL polypeptide according to any one of the preceding embodiments or an immunomodulatory polypeptide according to any one of the preceding embodiments linked to a moiety. In some embodiments, the moiety is a targeting moiety that specifically binds to a molecule on the surface of a cell. In some embodiments, the targeting moiety specifically binds to a molecule on the surface of an immune cell. In some embodiments, the immune cell is an antigen-presenting cell or a lymphocyte. In some embodiments, the targeting moiety is a tumor-localizing moiety that binds to a molecule on the surface of a tumor. In some embodiments, the moiety is a protein, peptide, nucleic acid, small molecule, or nanoparticle. In some embodiments, the moiety is an antibody or antigen-binding fragment. In some embodiments, the conjugate is bivalent, tetravalent, hexavalent, or octavalent.

[0039] In some embodiments, provided herein is a nucleic acid molecule encoding a variant ICOSL or an immunomodulatory polypeptide according to any one of the provided embodiments. In some embodiments, the nucleic acid molecule is a synthetic nucleic acid. In some embodiments, the nucleic acid is a cDNA.

[0040] In some embodiments, the present disclosure provides a vector comprising any one of the nucleic acids of the embodiments provided herein.In some embodiments, the vector is an expression vector.In some embodiments, the vector is a mammalian expression vector or a viral vector.

[0041] In some embodiments, provided herein is a cell comprising a vector according to any one of the provided embodiments. In some embodiments, the cell is a mammalian cell. In some embodiments, the cell is a human cell.

[0042] In some embodiments, provided herein is a method for producing a polypeptide or an immunomodulatory protein, the method comprising introducing a nucleic acid molecule according to any one of the provided embodiments or a vector according to any one of the provided embodiments into a host cell under conditions allowing expression of the protein in the cell. In some embodiments, the method further comprises isolating or purifying the variant ICOSL polypeptide or immunomodulatory protein from the cell.

[0043] In some embodiments, provided herein is a method of modifying a cell to express a variant ICOSL polypeptide, the method comprising introducing into a host cell a nucleic acid molecule encoding a variant ICOSL polypeptide or an immunomodulatory polypeptide according to any one of the provided embodiments under conditions in which the cell expresses the polypeptide. In some embodiments, provided herein is a modified cell that expresses a variant ICOSL polypeptide, immunomodulatory polypeptide, nucleic acid molecule, or vector according to any one of the provided embodiments.

[0044] In some embodiments, the variant ICOSL polypeptide or immunomodulatory polypeptide comprises a signal peptide. In some aspects, the variant ICOSL polypeptide or immunomodulatory polypeptide does not contain a transmembrane domain and / or is not expressed on the surface of the cell. In some cases, the variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from the modified cell. In some aspects, the modified cell contains a variant ICOSL polypeptide that contains a transmembrane domain and / or is a transmembrane immunomodulatory protein according to any of the provided embodiments. In some embodiments, the variant ICOSL polypeptide is expressed on the surface of the cell.

[0045] In some embodiments, the modified cells are immune cells. In some embodiments, the immune cells are antigen-presenting cells (APCs) or lymphocytes. In some embodiments, the modified cells are primary cells. In some embodiments, the modified cells are mammalian cells. In some embodiments, the modified cells are human cells. In some embodiments, the lymphocytes are T cells. In some embodiments, the APCs are artificial APCs. In some embodiments, the modified cells further contain a chimeric antigen receptor (CAR) or a modified T cell receptor (TCR).

[0046] Also provided is an infectious agent comprising a nucleic acid molecule encoding a variant ICOSL polypeptide or immunomodulatory polypeptide according to any of the provided embodiments. In some embodiments, the encoded variant ICOSL polypeptide or immunomodulatory polypeptide does not contain a transmembrane domain and / or is not expressed on the surface of the cell expressing it. In some cases, the encoded variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from the cell expressing it. In some cases, the encoded variant ICOSL polypeptide contains a transmembrane domain. In some embodiments, the encoded variant ICOSL polypeptide is expressed on the surface of the cell expressing it.

[0047] In some of any such embodiments, the infectious agent is a bacterium or a virus. In some examples, the virus is an oncolytic virus. In some aspects, the oncolytic virus is an adenovirus, an adeno-associated virus, a herpes virus, a herpes simplex virus, a vesicular stomatitis virus, a reovirus, a Newcastle disease virus, a parvovirus, a measles virus, a vesicular stomatitis virus (VSV), a coxsackievirus, or a vaccinia virus. In some embodiments, the virus specifically targets dendritic cells (DCs) and / or is dendritic cell-tropic. In some embodiments, the virus is a lentiviral vector pseudotyped with a modified Sindbis virus envelope product.

[0048] In some of any such embodiments, the infectious agent further comprises a nucleic acid molecule encoding an additional gene product that can cause death of a target cell or enhance or potentiate an immune response. In some embodiments, the additional gene product is selected from an anti-cancer agent, an anti-metastatic agent, an anti-angiogenic agent, an immunomodulatory molecule, an immune checkpoint inhibitor, an antibody, a cytokine, a growth factor, an antigen, a cytotoxic gene product, a pro-apoptotic gene product, an anti-apoptotic gene product, a cell matrix degrading gene, a gene for tissue regeneration, or a gene for reprogramming human somatic cells to pluripotency.

[0049] In some embodiments, provided herein is a pharmaceutical composition comprising a variant ICOSL polypeptide according to any one of the provided embodiments, an immunomodulatory protein according to any one of the provided embodiments, a conjugate according to any one of the provided embodiments, a modified cell according to any one of the provided embodiments, or an infectious agent according to any one of the provided embodiments. In some embodiments, the pharmaceutical composition further comprises a pharmaceutically acceptable excipient. In some embodiments, the pharmaceutical composition is sterilized.

[0050] In some embodiments, provided herein is an article of manufacture comprising any of the provided pharmaceutical compositions in a vial. In some embodiments, the vial is sealed. In some embodiments, also provided is a kit comprising any of the pharmaceutical compositions and instructions for use. In some aspects, a kit comprising an article of manufacture and instructions for use is provided.

[0051] In some embodiments, provided herein is a method for modulating an immune response in a subject, comprising administering to the subject a pharmaceutical composition according to any one of the provided embodiments. In some embodiments, the method comprises administering modified cells as provided according to any one of the provided embodiments. In some embodiments, the modified cells are autologous to the subject. In some embodiments, the modified cells are allogeneic to the subject. In some embodiments, modulating the immune response treats a disease or condition in the subject.

[0052] In some embodiments, the immune response is enhanced. In some embodiments, an immunomodulatory protein or conjugate comprising a variant ICOSL polypeptide linked to a tumor-localizing moiety is administered to a subject. In some cases, the tumor-localizing moiety is or comprises a binding molecule that recognizes a tumor antigen. In some aspects, the binding molecule comprises an antibody or an antigen-binding fragment thereof, or comprises a wild-type IgSF domain or a variant thereof.

[0053] In some of any such embodiments, a pharmaceutical composition comprising an immunomodulatory protein according to any one of the provided embodiments or a conjugate according to any one of the provided embodiments is administered to the subject. In some embodiments, modified cells comprising a variant ICOSL polypeptide that is a transmembrane immunomodulatory protein are administered to the subject, and / or the modified cells are modified cells according to any one of the above embodiments.

[0054] In some embodiments, an infectious agent encoding a variant ICOSL polypeptide that is a transmembrane immunomodulatory protein is administered to a subject, optionally under conditions where the infectious agent infects tumor cells or immune cells and the transmembrane immunomodulatory protein is expressed on the surface of the infected cells. In some aspects, the transmembrane immunomodulatory protein is any of the provided transmembrane immunomodulatory proteins according to any one of the provided embodiments.

[0055] In some embodiments, the disease or condition is a tumor or cancer, hi some embodiments, the disease or condition is selected from melanoma, lung cancer, bladder cancer, hematological malignancies, liver cancer, brain cancer, kidney cancer, breast cancer, pancreatic cancer, colon cancer, spleen cancer, prostate cancer, testicular cancer, ovarian cancer, uterine cancer, gastric cancer, musculoskeletal cancer, head and neck cancer, gastrointestinal cancer, germ cell cancer, or endocrine and neuroendocrine cancer.

[0056] In some embodiments, the immune response is reduced by the provided method of regulating the immune response. In some embodiments, a soluble variant ICOSL polypeptide or immunomodulatory protein is administered to the subject. In some cases, the soluble polypeptide or immunomodulatory protein is an Fc-fusion protein. In some embodiments, a pharmaceutical composition comprising a variant ICOSL polypeptide according to any one of the provided embodiments or an immunomodulatory protein according to any one of the provided embodiments is administered to the subject.

[0057] In some of any such embodiments, modified cells comprising a secretable variant ICOSL polypeptide are administered to a subject. In some embodiments, modified cells according to any one of the provided embodiments are administered to a subject. In some embodiments, an infectious agent encoding a variant ICOSL polypeptide that is a secretable immunomodulatory protein is administered to a subject, optionally under conditions in which tumor cells or immune cells are infected with the infectious agent and the secretable immunomodulatory protein is secreted from the infected cells.

[0058] In some embodiments, the disease or condition is an inflammatory or autoimmune disease or condition. In some embodiments, the disease or condition is antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis, vasculitis, autoimmune skin disease, transplantation, rheumatic disease, inflammatory gastrointestinal disease, inflammatory eye disease, inflammatory neurological disease, inflammatory lung disease, inflammatory endocrine disease, or autoimmune blood disease. In some embodiments, the disease or condition is selected from inflammatory bowel disease, transplant, Crohn's disease, ulcerative colitis, multiple sclerosis, asthma, rheumatoid arthritis, or psoriasis.

[0059] In some of any such embodiments of the variant ICOSL polypeptide, the amino acid modification is an amino acid substitution, insertion, or deletion. [Brief explanation of the drawings]

[0060] [Figure 1] Figure 1 shows impedance results reflecting the cytotoxic killing activity of cells engineered with an anti-CD19 chimeric antigen receptor (CAR) alone or with exemplary transmembrane immunomodulatory TIPs (CD80-TIP or ICOSL-TIP) or the corresponding CD80 or ICOSL wild-type transmembrane proteins after co-culture with target antigen-expressing cells. Impedance was assessed using an Acea Real-Time Cell Analyzer (RTCA), which measures impedance changes in the culture medium of a 96-well microelectronic plate (E-plate). [Figure 2]Figure 2A shows that primary T cells are effectively transduced with viruses encoding both CAR and TIP proteins. Primary human T cells were activated with anti-CD3 and anti-CD28 beads for 48 hours and then transduced with a lentivirus encoding an anti-CD19 CAR with a BFP reporter and a second lentivirus encoding ICOSL TIP with a GFP reporter. FAC plots show BFP expression on the y-axis and GFP expression on the x-axis, with the percentage of T cells falling into each quadrant displayed. Results show that cultures contain cells transduced with CAR alone (upper left quadrant), TIP alone (lower right quadrant), both viruses (upper right quadrant), and neither virus (lower left). In Figure 2B, TIP expressed on CAR-T cells provides costimulation to CAR-T cells. CAR-T cells, either co-transduced or not co-transduced with TIP, were labeled with Cell-Trace Far Red and incubated with the CD19+ NALM6 cell line to allow CAR association. Proliferation was assessed by the percentage of CAR-expressing cells that diluted out the fluorescent dye. Cells transduced with mutant TIP showed increased proliferation of CAR+ T cells compared with those without TIP or transduced with wild-type ICOSL. Mock-transduced cells, lacking CAR expression, did not proliferate in this assay. [Figure 3A] Figures 3A-3B demonstrate the costimulatory ability of wild-type (WT) or variant ICOSL when co-immobilized with anti-CD3 through cytokine release. 10 nM anti-CD3 was wet-coated with 40 nM (arrow) or 10 nM WT or variant ICOSL onto wells of a 96-well flat-bottom polystyrene tissue culture plate. 100,000 purified CD4+ and CD8+ (pan) T cells were added, and after 72 hours, supernatants were collected for ELISA analysis of cytokine release. Figure 3A shows IFN-γ and Figure 3B shows IL-17 protein levels secreted by pan T cells. Graphs are representative of typical IFN-γ and IL-17 responses from pan T cell costimulation. [Figure 3B] See legend to Figure 3A. [Figure 4A] Figures 4A-4B demonstrate the costimulatory ability of wild-type (WT) or variant ICOSL when co-immobilized with anti-CD3 through proliferation. CFSE-labeled pan T cells were incubated for 72 hours in plates coated with anti-CD3 and ICOSL as previously described. Cells were harvested, washed, stained with fluorescently conjugated anti-CD4 or anti-CD8 antibodies, and analyzed by flow cytometry. Gates and cytometer voltages were set using unstimulated control CFSE-labeled T cells. Proliferation was determined by CFSE dilution from the control. Figure 4A shows the percentage of total proliferating cells (arrows), CD4+ cells (solid bars), and CD8+ cells (hatched bars) T cells after 40 nM ICOSL costimulation. Figure 4B shows the percentage of total pan T cell proliferation after 10 nM ICOSL costimulation. The graph is representative of a typical proliferative response from pan T cell costimulation. [Figure 4B] See legend to Figure 4A. [Figure 5] This figure shows the function of ICOSL vIgD candidates in a human mixed lymphocyte reaction (MLR). ICOSL variants and their mutations are listed on the x-axis along with wild-type ICOSL, negative controls PDL2-Fc and human IgG, and the positive control reference molecule CTLA-Ig belatacept. The line across the graph represents the baseline amount of IFN-γ detected in the supernatant of negative control cultures. For each ICOSL variant candidate or control, three different concentrations were tested, and the arrow indicates the highest concentration of protein in the 40 nM culture. The majority of the ICOSL variant candidates demonstrate superior antagonistic activity at all three concentrations tested compared to belatacept, as reflected by the lower IFN-γ concentrations in these cultures. [Figure 6A]Figure 6 shows the inhibition of B and T cell responses by soluble ICOSL Fc fusion protein in a B cell coculture assay. Figure 6A shows the inhibition of T cell-induced B cell proliferation by soluble ICOSL Fc fusion protein. Purified CD4+ T cells and B cells from a single donor were labeled with CFSE and co-incubated at a 1:1 ratio with or without the indicated ICOSL Fc fusion protein, in the presence or absence of the indicated mitogens. Cells were stimulated with 100 ng / ml Staphylococcal enterotoxin B (SEB), 1 mg / ml pokeweed mitogen (PWM), or both. ICOSL Fc fusion protein was incubated at a final concentration of 40 nM, and the cultures were incubated for 7 days and subjected to FACS analysis. The number of differentiated B cells was determined from the number of cells in the cultures at which their CFSE was diluted. All of the ICOSL Fc fusion proteins tested, except for the wild-type, reduced B cell proliferation. Figures 6B-D show that ICOSL Fc fusion protein inhibited cytokine T cell cytokine production in B-T cocultures. Supernatants from the above cultures were collected on day 7 and analyzed for cytokine content using the LEGENDplex Human T Cell Cytokine Panel (Biolegend). T cell production of IL-5 (Figure 6B), IL-13 (Figure 6C), and IL-21 (Figure 6D) is attenuated by the inclusion of ICOSL Fc fusion protein. [Figure 6B] See legend to Figure 6A. [Figure 6C] See legend to Figure 6A. [Figure 6D] See legend to Figure 6A. [Figure 7A]Figures 7A-7D show various endpoints in a mouse model of graft-versus-host disease (GVHD) in which human PBMC cells were adoptively transferred into immunodeficient NSG mouse hosts. Figure 7A shows the survival curves of treated animals. An aggressive disease course and subsequent death were observed in saline control animals, while similar survival was observed in animals treated with wild-type ICOSL-Fc and the N52H / I143T ICOSL variant. The variant N52H / N57Y / Q100P improved survival to a similar extent as the clinical standard drug belatacept. [Figure 7B] Figure 7B shows a similar trend in weight loss, with ICOSL variant N52H / N57Y / Q100P demonstrating similar weight maintenance as animals treated with belatacept, despite all other groups experiencing rapid weight loss. [Figure 7C] Figure 7C shows clinical scores from observation of the standardized GVHD Disease Activity Index (DAI), again showing lower scores comparable to the clinical standard belatacept in animals treated with ICOSL variant N52H / N57Y / Q100P, while other animal groups experienced higher DAI scores. [Figure 7D] Figure 7D shows flow cytometry measurements of the percentage of CD4 and CD8 cells in blood from experimental animals measured on day 14. The percentage of CD8 cells between experimental groups was nearly identical, but animals treated with ICOSL variants N52H / N57Y / Q100P and belatacept had lower percentages of CD4 cells compared to the other experimental groups. [Figure 8]The localized costimulatory activity conveyed by the indicated variant stack molecule vIgD CL is shown, where C represents the ICOSL costimulatory domain and L represents the NKp30 localization domain. In this assay, target K562 cells expressing the localized surface protein B7-H6 were cultured with human T cells in the presence of anti-CD3, and T cell activation was assessed by IFN-γ levels in the culture supernatant. Inclusion of anti-CD3 alone or without the stacked variant Fc molecule did not induce T cell activation. Similarly, cells cultured with the wild-type localized NKp30 domain alone or the wild-type costimulatory ICOSL domain alone as an Fc fusion protein did not result in T cell activation. Stacked domains containing wild-type versions of both the costimulatory and localization domains induced measurable IFN-γ at the highest concentrations tested, while the variant localization costimulatory stack induced more than two-fold higher IFN-γ levels at the highest concentrations and still induced observed IFN-γ levels as the concentrations were gradually decreased. [Figure 9] This figure summarizes the changes in ear thickness in mice from a standard model of delayed-type hypersensitivity (DTH). PBS-treated animals sensitized with ovalbumin and then challenged in the ear with the same protein exhibit the highest measured levels of ear swelling. Mice treated with the clinical standard drug abatacept have slightly reduced ear swelling after ear challenge. All five ICOSL variant-treated groups demonstrated equal or improved ear swelling reduction compared to abatacept. [Figure 10A] Figures 10A-C show various exemplary configurations of variant IgSF domains (vIgD) conjugated to antibodies (V-Mab). Figure 10A shows various configurations in which vIgD is linked directly or indirectly to the N-terminus and / or C-terminus of the antibody light chain. Figure 10B shows various configurations in which vIgD is linked directly or indirectly to the N-terminus and / or C-terminus of the antibody heavy chain. Figure 10C shows the resulting V-Mab configurations when the light chain of Figure 10A and the heavy chain of Figure 10B are co-expressed in cells. [Figure 10B] See legend to Figure 10A. [Figure 10C] See legend to Figure 10A. [Figure 11A] Figures 11A-11B demonstrate V-Mab specificity for its cognate binding partner. Binding assays were performed on Expi293 cells transiently transfected with DNA for mammalian surface expression of human HER2, CD28, CTLA-4, or ICOS. 200,000 transfected cells were incubated with 100,000 pM to 100 pM of the parental antibody (C1) or various V-Mabs (C2-9). Unbound antibody was removed, bound antibody was detected with fluorescently conjugated anti-human IgG, and cells were analyzed by flow cytometry for MFI and percent positive based on the Fc control. Figure 11A shows V-Mabs binding to HER2 transfectants at levels similar to those of the parental antibodies. Binding to mock-transfected cells is observed with all V-Mabs, but not with WT ICOSL, due to the low level of endogenous HER2 expression in the Expi293 parental cells. Figure 11B shows that binding of the parent IgSF domain (N52H / N57Y / Q100P) to its cognate partner is maintained or increased by V-Mabs (C2, C3, C4, C5, C6, C8, C9). [Figure 11B] See legend to Figure 11A. [Figure 12] We demonstrate the costimulatory and proliferation potential of V-MAbs when co-immobilized with anti-CD3. 10 nM anti-CD3 was wet-coated onto wells of a 96-well flat-bottom polystyrene tissue culture plate along with 30 nM to 3 nM parental antibody, V-MAb, or Fc control. CFSE-labeled pan T cells were added for 72 h. IFN-γ secretion was measured by ELISA, and total T cell proliferation was measured by flow cytometric analysis of CFSE dilution. IFN-γ secretion and proliferation of the IgSF domain (N52H / N57Y / Q100P) were greater than that of WT ICOSL. V-MAbs demonstrate increased cytokine and proliferation costimulatory potential over the parental IgSF. [Figure 13A]Figures 13A-13C show various formats of the provided variant IgSF domain molecules. Figure 13A shows a soluble molecule including: (1) a variant IgSF domain (vIgD) fused to an Fc chain; (2) a stacked molecule containing a first variant IgSF domain (first vIgD) and a second IgSF domain, e.g., a second variant IgSF domain (second vIgD); (3) a tumor-targeted IgSF molecule containing a first variant IgSF domain (vIgD) and an IgSF domain that targets a tumor antigen, e.g., an NkP30 IgSF domain; and (4) a variant IgSF domain (vIgD) linked to an antibody (V-Mab). [Figure 13B] Figure 13B shows a transmembrane immunomodulatory protein (TIP) containing a variant IgSF domain (vIgD), e.g., variant ICOSL, expressed on the surface of a cell. In an exemplary embodiment, the cognate binding partner of the transmembrane-bound vIgD is a costimulatory receptor, e.g., CD28, and the TIP containing vIgD (e.g., ICOSL vIgD) engages the costimulatory receptor such that the TIP induces a positive signal in cells expressing the costimulatory receptor. [Figure 13C] Figure 13C shows a secreted immunomodulatory protein (SIP) in which a variant IgSF domain (vIgD), e.g., variant ICOSL, is secreted from a cell, e.g., a first T cell (e.g., a CAR T cell). In an exemplary embodiment, the cognate binding partner of the secreted vIgD is an activating receptor (e.g., CD28) that can be expressed on a first cell (e.g., a T cell) and / or a second cell (e.g., a T cell; either endogenous or modified, e.g., a CAR T cell). Binding of the SIP to its cognate binding partner blocks signaling through the activating receptor. In all cases, the vIgD can be simply a V domain (IgV), a combination of a V domain (IgV) including the entire extracellular domain (ECD) and a C domain (IgC), or any combination of Ig domains of an IgSF superfamily member. [Figure 14]

[0023] Figure 1 shows an exemplary diagram of the activity of a variant IgSF domain (vIgD) fused to Fc (vIgD-Fc), where vIgD is a variant of the IgSF domain of ICOSL. As shown, soluble ICOSL vIgD interacts with its cognate binding partners, blocking the interaction of CD80 (B7-1) / CD86 (B7-2) or ICOSL with CD28 or ICOS, respectively, thereby blocking costimulation by the CD28 and / or ICOS costimulatory receptors. [Figure 15] 1 shows an exemplary diagram of a stack molecule for localizing a variant IgSF domain (vIgD) to tumor cells. In this format, the stack molecule contains a first variant IgSF domain (first vIgD) and a second IgSF domain (e.g., second vIgD), where the second IgSF domain (e.g., second vIgD) is a tumor-targeting IgSF domain that binds to a tumor antigen. An exemplary tumor-targeting IgSF domain is the IgSF domain of NkP30, which binds to the tumor antigen B7-H6. In this depiction, vIgD is a variant of the IgSF domain of ICOSL. As shown, binding of the tumor-targeting IgSF domain to the surface of a tumor cell localizes the first vIgD on the tumor cell surface, where it can interact with one or more of its cognate binding partners (e.g., CD28 or ICOS) expressed on the surface of adjacent immune cells (e.g., T cells) to stimulate costimulatory receptors. [Figure 16]

[0023] Figures 1A-1C show various exemplary configurations of stacked molecules containing a first variant IgSF domain (first vIgD), e.g., variant ICOSL, and a second IgSF domain, e.g., a second variant IgSF domain (second vIgD). As shown, the first vIgD and second IgSF domain are independently linked, directly or indirectly, to the N- or C-terminus of an Fc subunit. To generate homodimeric Fc molecules, the Fc subunits are matched and capable of forming homodimers upon coexpression of the individual Fc subunits in a cell. To generate heterodimeric Fc molecules, the individual Fc subunits contain mutations (e.g., "knobs-into-holes" mutations in the CH3 domain) that favor heterodimer formation over homodimers when the individual Fc subunits are coexpressed in a cell. [Figure 17] 1 shows an exemplary diagram of the activity of a variant IgSF domain (vIgD) conjugated to an antibody (V-Mab), in which the antibody (e.g., an anti-HER2 antibody) binds to an antigen on the surface of a tumor cell. In this depiction, vIgD is a variant of the IgSF domain of ICOSL. As shown, binding of the antibody to the surface of the tumor cell localizes vIgD on the tumor cell surface, where it can interact with one or more of its cognate binding partners expressed on the surface of adjacent immune cells (e.g., T cells) to activate receptor signaling. In the exemplary embodiment shown, the variant IgSF domain (vIgD) is a variant of the IgSF domain of ICOSL. Binding of ICOSL vIgD to CD28 or ICOS costimulatory receptors provides an agonistic or costimulatory signal. [Figure 18]Figure 1 shows the Nanostring transcriptional signature of primary human T cells incubated with 10 nM anti-CD3 and 40 nM Fc control protein, wild-type ICOSL-Fc, wild-type CD80-Fc, both of these proteins, or variant ICOSL Fc fusion proteins with the indicated mutations. Total RNA from samples was prepared from harvested cells, the RNA was transcribed into Nanostring, and the transcripts of 750 genes in each sample were quantified using a Cancer Immune chip. Altered transcripts include those with levels above or below the diagonal line, including the transcripts indicated. [Figure 19] Transcript levels of exemplary transcripts are shown upon incubation as described in Figure 18 in the presence of various immunomodulatory proteins for the indicated times. [Figure 20A] Figures 20A-B demonstrate VmAb-mediated T cell proliferation when cocultured with HER2-expressing targets. CFSE-labeled pan T cells were activated with K562-derived artificial target cells displaying cell surface anti-CD3 single-chain Fv (OKT3) and HER2 in the presence of VmAbs or control proteins. Proliferation was measured by flow cytometry analysis of CFSE dilution relative to CD4+ (left panel) or CD8+ (right panel) stained T cells. In Figure 20A, K562 cells were titrated and plated with T cells at effector:target (E:T) ratios of 40 to 1280:1. VmAb, parental IgSF domain, or WT ICOSL was added at 1000 pM. In Figure 20B, K562 cells were added to T cells at an E:T ratio of 160:1. VmAb or control proteins were titrated and added at 3000 to 37 pM. [Figure 20B] See legend to Figure 20A. DETAILED DESCRIPTION OF THE INVENTION

[0061] Detailed Description Provided herein are immunomodulatory proteins that are or include variants or mutants of ICOS ligand (ICOSL) or specific-binding fragments thereof that exhibit binding activity to at least one target ligand cognate binding partner (also referred to as a counter-structure protein). In some embodiments, the variant ICOSL polypeptide contains one or more amino acid modifications (e.g., amino acid substitutions, deletions, or additions) compared to the unmodified or wild-type ICOSL polypeptide. In some embodiments, the one or more amino acid modifications (e.g., amino acid substitutions, deletions, or additions) are modifications in the IgSF domain (e.g., IgV) of the unmodified or wild-type ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has altered (e.g., improved or decreased) binding activity or affinity to at least one cognate binding partner (e.g., at least one of ICOS, CD28, or CTLA-4). In some embodiments, the immunomodulatory protein is soluble. In some embodiments, the immunomodulatory protein is a transmembrane immunomodulatory protein that can be expressed on the surface of a cell. Also provided herein, in some embodiments, is one or more other immunomodulatory proteins that are conjugates or fusions containing a variant ICOSL polypeptide provided herein and one or more other moieties or polypeptides.

[0062] In some embodiments, the variant ICOSL polypeptides and immunomodulatory proteins modulate immunological immune responses (e.g., enhance or reduce immune responses). In some embodiments, the variant ICOSL polypeptides and immunomodulatory proteins provided herein can be used to treat diseases or conditions associated with a dysregulated immune response.

[0063] In some embodiments, the provided variant ICOSL polypeptides regulate T cell activation through interaction with costimulatory signaling molecules. Generally, antigen-specific T cell activation requires two distinct signals. The first signal is provided by the interaction of the T cell receptor (TCR) with an antigen bound to a major histocompatibility complex (MHC) present on an antigen-presenting cell (APC). The second signal is costimulatory for TCR engagement and is required to prevent T cell apoptosis or anergy.

[0064] In some embodiments, under normal physiological conditions, T cell-mediated immune response is initiated by antigen recognition by T cell receptor (TCR), and is regulated by the balance of costimulatory signal and inhibitory signal (e.g., immune checkpoint receptor).The immune system relies on immune checkpoint receptor to prevent autoimmunity (i.e., self-tolerance) and protect tissue from excessive damage during immune response (e.g., during attack against pathogenic infection).However, in some cases, these immunoregulatory proteins may be dysregulated in diseases and pathologies, including tumors, as a mechanism to evade the immune system.

[0065] In some embodiments, known T cell costimulatory receptors include CD28, which is a T cell costimulatory receptor for the ligands B7-1 (CD80) and B7-2 (CD86) that coexist on APCs. These same ligands can also bind to the inhibitory T cell receptor CTLA4 (cytotoxic T lymphocyte-associated protein 4) with higher affinity than CD28; binding to CTLA-4 acts to downregulate immune responses. ICOS (inducible costimulator) is another T cell costimulatory receptor that binds to ICOS ligand (ICOSL) on APCs. In some cases, CD28 and CTLA-4 are also known to interact with ICOSL at binding sites that overlap with the binding of ICOSL to the T cell costimulatory receptor ICOS (Yao et al. (2011) Immunity, 34:729-740). Although CD28 and ICOS are related CD28 family activation receptors and share several intracellular signaling motifs, the costimulatory effects of CD28 and ICOS differ. For example, CD28 is expressed on both inactive and activated T cells, and its signaling is important for IL-2 production and subsequent T cell effector function. ICOS is generally not expressed on the surface of T cells until after T cell activation, and signaling via ICOS on activated T cells supports the differentiation of specialized T cell subsets. Thus, in some cases, costimulation by CD28 and ICOS has overlapping and complementary effects.

[0066] In some embodiments, T cell costimulatory receptors CD28 and ICOS have different but complementary roles in regulating immune responses.Enhancing or suppressing the activity of these receptors has clinical significance for the treatment of inflammatory and autoimmune disorders, cancer, and viral infections.However, in some cases, therapies that intervene and change the costimulatory effects of both receptors are constrained by the spatial location requirements and size limitations imposed by the localization of the immune synapse.In some aspects, existing therapeutic agents, including antibody drugs, cannot simultaneously interact with multiple target proteins involved in regulating these interactions.In addition, in some cases, existing therapeutic agents only have the ability to antagonize immune responses, but cannot stimulate immune responses.In addition, the pharmacokinetic differences between drugs that independently target one or the other of these two receptors can make it difficult to appropriately maintain the desired blood concentration of such drug combinations throughout the treatment process.

[0067] In some embodiments, the provided variant ICOSL polypeptides or immunomodulatory proteins modulate (e.g., enhance or reduce) immune activity induced by the costimulatory receptor CD28 or ICOS. Thus, in some embodiments, the provided polypeptides overcome these limitations by providing variant ICOSL (inducible costimulator ligand) with altered (e.g., improved or decreased) binding affinity for both CD28 and ICOS, and in some cases, for CTLA-4, thereby stimulating or antagonizing the complementary effects of receptor costimulation. Also provided are methods of making and using these variant ICOSL.

[0068] All publications (including patents, patent applications, scientific articles, and databases) mentioned herein are incorporated by reference in their entirety for all purposes to the same extent as if each individual publication (including patents, patent applications, scientific articles, and databases) was specifically and individually indicated to be incorporated by reference. To the extent that a definition set forth herein is contrary to or otherwise inconsistent with a definition set forth in a patent, application, published application, or other publication incorporated herein by reference, the definition set forth herein shall take precedence over the definition incorporated herein by reference.

[0069] The section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described.

[0070] I. Definition Unless otherwise defined, all technical terms, notations, and other technical and scientific terms or related terms used herein are intended to have the same meaning as commonly understood by one of ordinary skill in the art to which the claimed subject matter belongs. In some cases, terms having commonly understood meanings are defined herein for clarity and / or ready reference, and the inclusion of such definitions herein should not necessarily be construed as making a significant difference from what is commonly understood in the art.

[0071] Terms used throughout this specification are defined as follows, unless otherwise limited in specific instances. As used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Unless otherwise defined, all technical and scientific terms, acronyms, and abbreviations used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Unless otherwise specified, abbreviations and symbols for chemical and biochemical names are in accordance with IUPAC-IUB nomenclature. Unless otherwise specified, all numerical ranges include not only the values ​​defining the range but also all integer values ​​therebetween.

[0072] The term "affinity-modified (affinity-engineered)," when used in the context of immunoglobulin superfamily domains, refers to a mammalian immunoglobulin superfamily (IgSF) domain having an amino acid sequence that has been altered to improve or decrease (relative to the corresponding wild-type parent or unmodified IgSF domain) the binding affinity or avidity for at least one of its cognate binding partners (or "counterstructures") relative to a parent wild-type or unmodified (i.e., non-affinity-modified) IgSF control domain. Included in this context are affinity-modified ICOSL IgSF domains. In some embodiments, the affinity-modified IgSF domain can contain 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, or more amino acid differences (e.g., amino acid substitutions) in a wild-type or unmodified IgSF domain. The improvement or decrease in binding affinity or avidity can be determined using well-known binding assays, such as flow cytometry. Larsen et al., American Journal of Transplantation, Vol 5: 443-453 (2005). See also Linsley et al., Immunity, 1: 7930801 (1994). An improvement in the binding affinity or avidity of a protein for its cognate binding partner is an improvement that is at least 10% greater than the wild-type IgSF domain control, and in some embodiments, is at least 20%, 30%, 40%, 50%, 100%, 200%, 300%, 500%, 1000%, 5000%, or 10000% greater than the wild-type IgSF domain control value. The reduction in binding affinity or avidity of the protein for at least one of its cognate binding partners is a reduction of 90% or less of the control but 10% or more of the wild-type IgSF domain control value, and in some embodiments, a reduction of 80%, 70%, 60%, 50%, 40%, 30%, or 20% or less of the wild-type IgSF domain control value but 10% or more.Affinity-modified proteins have altered primary amino acid sequences due to substitution, addition, or deletion of amino acid residues. The term "affinity-modified IgSF domain" should not be construed as imposing any requirement on any particular starting composition or method by which the affinity-modified IgSF domain is generated. Thus, the affinity-modified IgSF domains of the present invention are not limited to converting a wild-type IgSF domain into an affinity-modified IgSF domain by any particular affinity-modification process. Affinity-modified IgSF domain polypeptides can be generated, for example, starting from wild-type mammalian IgSF domain sequence information, then modeled in silico for binding to their cognate binding partners, and finally recombinantly or chemically synthesized to generate the subject affinity-modified IgSF domain compositions. As just another example, affinity-modified IgSF domains can be generated by site-directed mutagenesis of wild-type IgSF domains. Thus, affinity-modified IgSF domains represent products, but not necessarily, produced by any given process. A variety of techniques, including recombinant methods, chemical synthesis, or a combination thereof, may be used.

[0073] The term "allogeneic," as used herein, refers to cells or tissues that are removed from one organism and then injected or adoptively transferred into a genetically different organism of the same species. In some embodiments of the invention, the species is murine or human.

[0074] The term "autologous" as used herein refers to cells or tissues that are extracted from the same organism and then injected or adoptively transferred into the organism. Autologous cells or tissues can be modified, for example, by recombinant DNA techniques, so that they are no longer genetically identical to the natural cells or natural tissues extracted from the organism. For example, natural autologous T cells can be genetically modified by recombinant DNA techniques to become autologous modified cells that express transmembrane immunomodulatory proteins and / or chimeric antigen receptors (CARs), which in some cases involves modifying T cells or TILs (tumor-infiltrating lymphocytes). The modified cells are then infused into the patient from whom the natural T cells were isolated. In some embodiments, the organism is a human or a mouse.

[0075] The terms "binding affinity" and "binding avidity" as used herein refer to the specific binding affinity and specific binding avidity of a protein to its counter structure under specific binding conditions, respectively. In biochemical kinetics, avidity refers to the cumulative strength of the affinities of multiple individual non-covalent interactions, such as between ICOSL and its counter structure ICOS and / or CD28. Thus, avidity differs from affinity, which represents the strength of a single interaction. The improved or decreased binding affinity of an affinity-modified ICOSL variant containing an IgSF domain to its counter structure is determined compared to the binding affinity of unmodified ICOSL (e.g., unmodified ICOSL containing a natural or wild-type IgSF domain (e.g., an IgV domain)). Methods for determining binding affinity or avidity are known in the art. See, for example, Larsen et al., American Journal of Transplantation, Vol. 5: 443-453 (2005). In some embodiments, the variant ICOSL of the present invention (i.e., an ICOSL protein containing an affinity-modified IgSF domain) specifically binds to CD28 and / or ICOS with a binding affinity that results in a mean fluorescence intensity (MFI) value that is at least 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% greater than a wild-type ICOSL control, as measured by flow cytometry, e.g., in a binding assay described in Example 6.

[0076] The term "biological half-life" refers to the time required for a substance (e.g., an immunomodulatory polypeptide comprising a variant ICOSL of the present invention) to lose half of its pharmacological or physiological activity or concentration. Biological half-life can be affected by the elimination, excretion, degradation (e.g., enzymatic degradation) of the substance, or absorption and concentration in specific organs or tissues in the body. In some embodiments, biological half-life can be assessed by determining the time required for the plasma concentration of the substance to reach half of its steady-state level ("plasma half-life"). Conjugates that can be used to derivatize the polypeptides of the invention to extend their biological half-life are known in the art and include, but are not limited to, polyethylene glycol (PEG), hydroxyethyl starch (HES), XTEN (extended recombinant peptides; see WO 2013130683), human serum albumin (HSA), bovine serum albumin (BSA), lipids (acylated), and poly-Pro-Ala-Ser (PAS), polyglutamic acid (glutamylated).

[0077] The term "chimeric antigen receptor" or "CAR," as used herein, refers to an artificial (i.e., man-made) transmembrane protein expressed on mammalian cells, comprising at least an ectodomain, a transmembrane domain, and an endodomain. Optionally, the CAR protein includes a "spacer" covalently linking the ectodomain to the transmembrane domain. The spacer is often a polypeptide that links the ectodomain to the transmembrane domain via a peptide bond. CARs are typically expressed on mammalian lymphocytes. In some embodiments, CARs are expressed on mammalian cells such as T cells or tumor-infiltrating lymphocytes (TILs). CARs expressed on T cells are referred to herein as "CAR T cells" or "CAR-T." In some embodiments, CAR-Ts are helper T cells, cytotoxic T cells, natural killer T cells, memory T cells, regulatory T cells, or γδ T cells. When used clinically, for example, in adoptive cell transfer, CAR-Ts with antigen-binding specificity for a patient's tumor are typically engineered to be expressed on natural T cells obtained from the patient. The engineered T cells expressing the CAR are then infused back into the patient. Thus, while CAR-Ts are often autologous, allogeneic CAR-Ts are also within the scope of the present invention. The ectodomain of the CAR comprises an antigen-binding region (e.g., an antibody or its antigen-binding fragment (e.g., scFv)) that specifically binds to a target antigen (e.g., a tumor-specific antigen) under physiological conditions. Upon specific binding, a series of biochemical events (i.e., signal transduction) results in the modulation of the immune activity of the CAR-T. Thus, for example, specific binding of the antigen-binding region of the CAR-T to its target antigen can lead to changes in the immune activity of T cell activity, as reflected by changes in cytotoxicity, proliferation, or cytokine production. In some embodiments, signal transduction upon CAR-T activation is achieved by the CD3 zeta chain ("CD3-z"), which is involved in signal transduction in natural mammalian T cells. CAR-Ts can further comprise multiple signaling domains (e.g., CD28, 41BB, or OX40) that further regulate the immune regulatory response of T cells.CD3-z contains a conserved motif known as an immunoreceptor tyrosine-based activation motif (ITAM) that is involved in T cell receptor signaling.

[0078] The terms "total" or "total," when used in reference to cytokine production induced by the presence of two or more variant ICOSL of the present invention in an in vitro assay, refer to the overall cytokine expression level regardless of the cytokine production induced by each individual variant ICOSL. In some embodiments, the cytokine assayed is IFN-γ, e.g., in the in vitro primary T cell assay described in Example 7.

[0079] The term "cognate binding partner" (used interchangeably with "counterstructure"), with respect to a polypeptide (e.g., an IgSF domain of a variant ICOSL), refers to at least one molecule (typically a native mammalian protein) to which the referenced polypeptide specifically binds under specific binding conditions. In some aspects, a variant ICOSL containing an affinity-modified IgSF domain specifically binds to the corresponding counterstructure of native or wild-type ICOSL with improved or reduced affinity. A species of ligand that is recognized and specifically binds to its cognate receptor under specific binding conditions is an example of a counterstructure or cognate binding partner for that receptor. A "cell surface cognate binding partner" is a cognate binding partner expressed on the surface of a mammalian cell. A "cell surface molecular species" is a cognate binding partner of a ligand of the immune synapse (IS) that is expressed on or by cells (e.g., mammalian cells) that form the IS.

[0080] As used herein, "conjugate," "conjugation," or grammatical variations thereof, refers to the joining or linking of two or more compounds together by any connection or linking method known in the art, resulting in the formation of another compound. It can also refer to a compound produced by joining or linking two or more compounds together. For example, a variant ICOSL polypeptide directly or indirectly linked to one or more chemical moieties or polypeptides is an exemplary conjugate. Such conjugates include fusion proteins, those produced by chemical conjugation, and those produced by any other method.

[0081] The term "competitive binding," as used herein, means that a protein can specifically bind to at least two cognate binding partners, but the specific binding of one cognate binding partner inhibits (e.g., interferes with or prevents) the simultaneous binding of a second cognate binding partner. Thus, in some cases, a protein cannot simultaneously bind to two cognate binding partners. Generally, competitive binders contain the same or overlapping binding sites for specific binding, although this is not a requirement. In some embodiments, competitive binding causes measurable (partial or complete) inhibition of the specific binding of a protein to one of its cognate binding partners due to the specific binding of a second cognate binding partner. Various methods for quantifying competitive binding are known, such as ELISA (enzyme-linked immunosorbent assay).

[0082] The term "conservative amino acid substitution," as used herein, refers to an amino acid substitution in which one amino acid residue is replaced with another amino acid residue having a side chain R group with similar chemical properties (e.g., charge or hydrophobicity). Examples of groups of amino acids with side chains with similar chemical properties include: 1) aliphatic side chains: glycine, alanine, valine, leucine, and isoleucine; 2) aliphatic-hydroxyl side chains: serine and threonine; 3) amide-containing side chains: asparagine and glutamine; 4) aromatic side chains: phenylalanine, tyrosine, and tryptophan; 5) basic side chains: lysine, arginine, and histidine; 6) acidic side chains: aspartic acid and glutamic acid; and 7) sulfur-containing side chains: cysteine ​​and methionine. Conservative amino acids substitution groups are: valine-leucine-isoleucine, phenylalanine-tyrosine, lysine-arginine, alanine-valine, glutamate-aspartate, and asparagine-glutamine.

[0083] The term "corresponding" with respect to a protein position, e.g., a statement that a nucleotide or amino acid position "corresponds to" a nucleotide or amino acid position in a disclosed sequence (e.g., as shown in the sequence listing), refers to a nucleotide or amino acid position identified by alignment with the disclosed sequence based on structural sequence alignment or using a standard alignment algorithm (e.g., the GAP algorithm). For example, corresponding residues can be identified by alignment with a reference sequence having the sequence shown in SEQ ID NO:32 (ECD domain) or the sequence shown in SEQ ID NO:196 (IgV domain) using the structural alignment methods described herein. By aligning the sequences, one skilled in the art can identify corresponding residues, for example, using conserved and identical amino acid residues as criteria.

[0084] The terms "reduce" or "attenuate" or "inhibit," as used herein, refer to a statistically significant decrease. The decrease can be at least a 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% decrease.

[0085] The term "derivative" or "derivatized" refers to the modification of a protein by directly or indirectly covalently attaching the protein to a composition to alter properties such as biological half-life, bioavailability, immunogenicity, solubility, toxicity, potency, or efficacy, while retaining or enhancing its therapeutic benefit. Derivatives of the immunomodulatory polypeptides of the invention are within the scope of the invention and can be produced, for example, by glycosylation, pegylation, lipidation, or Fc fusion.

[0086] As used herein, a domain (typically a sequence of three or more, generally five or seven or more amino acids, e.g., 10-200 amino acid residues) refers to a portion of a molecule (e.g., a protein or encoding nucleic acid) that is structurally and / or functionally distinct from and identifiable with the rest of the molecule. For example, a domain includes a portion of a polypeptide chain that can form an independently folded structure within a protein composed of one or more structural motifs and / or is recognized by a functional activity, such as binding activity. A protein can have one or more distinct domains. For example, a domain can be identified, defined, or distinguished by primary sequence or structural homology to related family members, e.g., homology to a motif. In another example, a domain can be distinguished by its function (e.g., ability to interact with a biomolecule, such as a cognate binding partner). A domain can exhibit an independent biological function or activity, such that it can perform an activity (e.g., binding) independently or fused to another molecule. A domain can be a linear or nonlinear amino acid sequence. Many polypeptides contain multiple domains. Such domains are known and can be identified by those skilled in the art. For purposes of illustration herein, definitions are provided, but it is understood that recognizing a particular domain by name is well within the skill of the art. If necessary, appropriate software can be employed to identify the domains.

[0087] The term "ectodomain," as used herein, refers to the region of a membrane protein (e.g., a transmembrane protein) that is outside the vesicle membrane. Ectodomains often include a binding domain that specifically binds to a ligand or cell surface receptor, e.g., via a binding domain that specifically binds to the ligand or cell surface receptor. The ectodomain of a cellular transmembrane protein is alternatively referred to as the extracellular domain.

[0088] The term "effective amount" or "therapeutically effective amount" refers to an amount and / or concentration of a therapeutic composition (including a protein or cell composition) of the present invention that, when administered ex vivo (by contact with cells from a patient) or in vivo (by administration to a patient), either alone (i.e., as monotherapy) or in combination with an additional therapeutic agent, results in a statistically significant reduction in disease progression, e.g., by ameliorating or eliminating the symptoms and / or pathogenesis of the disease. An effective amount can be an amount that alleviates, reduces, or relieves at least one symptom or biological response or effect associated with a disease or disorder, prevents the progression of the disease or disorder, or improves the patient's physical function. In the case of cell therapy, an effective amount is an effective dose or number of cells administered to a patient via adoptive cell therapy. In some embodiments, the patient is a mammalian patient, e.g., a non-human primate or a human patient.

[0089] The term "endodomain," as used herein, refers to a region found in some membrane proteins (e.g., transmembrane proteins) that extends into the internal space defined by the cell surface membrane. In mammalian cells, the endodomain is the cytoplasmic region of the membrane protein. Within the cell, the endodomain can interact with intracellular components and play a role in signal transduction, and thus, in some cases, can be an intracellular signaling domain. The endodomain of a cellular transmembrane protein is alternatively referred to as a cytoplasmic domain, which, in some cases, can be a cytoplasmic signaling domain.

[0090] The terms "enhanced" or "increased (improved)," as used herein in the context of enhancing the immune activity of mammalian lymphocytes, refer to enhancing one or more activities of the lymphocytes. Enhanced activity can be an enhancement (e.g., by a statistically significant amount) of one or more of cell survival, cell proliferation, cytokine production, or T-cell cytotoxicity. In some embodiments, reference to enhanced immune activity refers to increasing interferon-γ (IFNγ) production (e.g., by a statistically significant amount). In some embodiments, immune activity can be assessed in a mixed lymphocyte reaction (MLR) assay. Methods for performing MLR assays are known in the art. Wang et al., Cancer Immunol Res. 2014 Sep: 2(9):846-56. Other methods for assessing lymphocyte activity are known in the art, including any of the assays described herein. In some embodiments, an enhancement can be an increase (improvement) of at least 10%, 20%, 30%, 40%, 50%, 75%, 100%, 200%, 300%, 400%, or 500% greater than a non-zero control value.

[0091] The term "modified cells" as used herein refers to mammalian cells that have been genetically modified (altered) by human intervention (e.g., recombinant DNA or viral transduction). In some embodiments, the cells are immune cells, such as lymphocytes (e.g., T cells, B cells, NK cells) or antigen-presenting cells (e.g., dendritic cells). The cells may be primary cells derived from a patient or may be cell lines. In some embodiments, the modified cells of the present invention comprise a variant ICOSL of the present invention that has been modified to regulate the immune activity of T cells expressing CD28, ICOS, or CTLA-4, to which the variant ICOSL specifically binds. In some embodiments, the variant ICOSL is a transmembrane immunomodulatory protein (hereinafter referred to as "TIP") that contains an extracellular domain or portion thereof containing an IgV domain linked to a transmembrane domain (e.g., an ICOSL transmembrane domain), and optionally contains an intracellular signaling domain. In some cases, the TIP is configured as a chimeric receptor containing a heterologous cytoplasmic signaling domain or endodomain. In some embodiments, the modified cells are capable of expressing and secreting an immunomodulatory protein as described herein. Some of the modified cells provided further contain a modified T cell receptor (TCR) or a chimeric antigen receptor (CAR).

[0092] The term "engineered T cells," as used herein, refers to T cells (e.g., helper T cells, cytotoxic T cells (or cytotoxic T lymphocytes or CTLs), natural killer T cells, regulatory T cells, memory T cells, or γδ T cells) that have been genetically modified (engineered) by human intervention (e.g., recombinant DNA or viral transduction). The engineered T cells comprise a variant ICOSL transmembrane immunomodulatory protein (TIP) of the invention expressed on the T cell, which has been engineered to modulate the immune activity of the engineered T cell itself or of a mammalian cell to which the variant ICOSL expressed on the T cell specifically binds. The terms "engineered T cell receptor" or "engineered TCR" refer to a T cell receptor (TCR) that has been engineered to specifically bind with a desired affinity to a major histocompatibility complex (MHC) / peptide target antigen that is selected, cloned, and / or subsequently introduced into a population of T cells (often used in adoptive immunotherapy). In contrast to engineered TCRs, CARs are engineered to bind to target antigens in an MHC-dependent manner.

[0093] The term "expressed on," as used herein, is used in reference to a protein expressed on the surface of a cell (e.g., a mammalian cell). Thus, the protein is expressed as a membrane protein. In some embodiments, the expressed protein is a transmembrane protein. In some embodiments, the protein is conjugated to a small molecule moiety (e.g., a drug or a detectable label). A protein expressed on the surface of a cell can include a cell surface protein (e.g., a cell surface receptor) expressed on a mammalian cell.

[0094] The term "half-life extending moiety" refers to a portion of a polypeptide fusion or chemical conjugate that extends the half-life of a protein circulating in mammalian serum compared to the half-life of the protein not so conjugated to the moiety. In some embodiments, the half-life is extended by more than 1.2-fold, 1.5-fold, 2.0-fold, 3.0-fold, 4.0-fold, 5.0-fold, or 6.0-fold, or by more than about 1.2-fold, about 1.5-fold, about 2.0-fold, about 3.0-fold, about 4.0-fold, about 5.0-fold, or about 6.0-fold. In some embodiments, the half-life is extended by more than 6 hours, more than 12 hours, more than 24 hours, more than 48 hours, more than 72 hours, more than 96 hours, or more than 1 week after in vivo administration compared to the protein without the half-life extending moiety. Half-life refers to the time required for a protein to lose half of its concentration, amount, or activity. Half-life can be determined, for example, by using ELISA assay or activity assay. Exemplary half-life extending moieties include Fc domain, multimerization domain, polyethylene glycol (PEG), hydroxyethyl starch (HES), XTEN (extended recombinant peptide; see WO 2013130683), human serum albumin (HSA), bovine serum albumin (BSA), lipid (acylation), and poly-Pro-Ala-Ser (PAS), and polyglutamic acid (glutamylation).

[0095] The term "immune synapse," as used herein, refers to the interface between a mammalian cell (e.g., an antigen-presenting cell or tumor cell) that expresses MHC I (major histocompatibility complex) or MHC II, and a mammalian lymphocyte (e.g., an effector T cell or natural killer (NK) cell).

[0096] The Fc (fragment crystallizable) region or domain of an immunoglobulin molecule (also referred to as an Fc polypeptide) primarily corresponds to the constant region of the immunoglobulin heavy chain and is involved in various functions, including antibody effector functions. The Fc domain contains part or all of the hinge domain and the CH2 and CH3 domains of an immunoglobulin molecule. The Fc domain can form a dimer of two polypeptide chains connected by one or more disulfide bonds. In some embodiments, the Fc is a variant Fc with reduced effector function-promoting activity (e.g., 30%, 40%, 50%, 60%, 70%, 80%, 90%, or greater reduction). In some embodiments, references to amino acid substitutions in the Fc region are in accordance with the EU numbering system, unless otherwise specified with reference to a specific SEQ ID NO. EU numbering is known and follows the EU index as reported in the recently updated IMGT Scientific Chart (IMGT®, i.e., international ImMunoGeneTics information system®) http: / / www.imgt.org / IMGTScientificChart / Numbering / Hu_IGHGnber.html (created May 17, 2001, last updated January 10, 2013) and Kabat, EA et al. Sequences of Proteins of Immunological interest. 5th ed. US Department of Health and Human Services, NIH publication No. 91-3242 (1991).

[0097] Immunoglobulin Fc fusions ("Fc fusions"), e.g., immunomodulatory Fc fusion proteins, are molecules comprising one or more polypeptides (or one or more small molecules) operatively linked to the Fc region of an immunoglobulin. An Fc fusion may comprise, for example, an antibody Fc region (which enhances effector function and pharmacokinetics) and a variant ICOSL. The immunoglobulin Fc region may be indirectly or directly linked to one or more variant ICOSL or small molecules (fusion partners). Various linkers are known in the art and can optionally be used to link the Fc to the fusion partner to generate the Fc fusion. Fc fusions of the same species can be dimerized to form Fc fusion homodimers, or non-identical species can be used to form Fc fusion heterodimers. In some embodiments, the Fc is a mammalian Fc, e.g., a murine or human Fc.

[0098] The term "host cell" refers to a cell that can be used to express a protein encoded by a recombinant expression vector. Host cells can be prokaryotic, e.g., Escherichia coli (E. coli), or they can be eukaryotic, e.g., unicellular eukaryotes (e.g., yeast or other fungi), plant cells (e.g., tobacco or tomato plant cells), animal cells (e.g., human cells, monkey cells, hamster cells, rat cells, mouse cells, or insect cells), or hybridomas. Examples of host cells include Chinese hamster ovary (CHO) cells or their derivatives, e.g., Veggie CHO and related cell lines grown in serum-free medium, or the CHO line DX-B11, which is DHFR-deficient. In some embodiments, the host cell is a mammalian cell (e.g., a human cell, monkey cell, hamster cell, rat cell, mouse cell, or insect cell).

[0099] The term "immunoglobulin" (abbreviated "Ig"), as used herein, refers to mammalian immunoglobulin proteins, including any of the five human classes of antibodies: IgA (including subclasses IgA1 and IgA2), IgD, IgE, IgG (including subclasses IgG1, IgG2, IgG3, and IgG4), and IgM. The term also refers to immunoglobulins that are less than full-length, whether wholly or partially synthetic (e.g., recombinantly or chemically synthesized) or naturally produced, such as antigen-binding fragments (Fab), V, and VI. H and V L variable fragments (Fv) containing V linked together in one chain; H and V L Bispecific antibodies include single-chain variable fragments (scFv) containing the V region fragments (Fab', F(ab)2, F(ab')2, dsFv diabodies, Fc, and Fd polypeptide fragments). Homo- and hetero-bispecific bispecific antibodies are included within the scope of the term.

[0100] The term "immunoglobulin superfamily" or "IgSF" as used herein refers to a group of cell surface and soluble proteins involved in cellular recognition, binding, or adhesion processes. Molecules are classified as members of this superfamily based on structural features shared with immunoglobulins (i.e., antibodies); they all possess domains known as immunoglobulin domains or folds. IgSF members include cell surface antigen receptors, coreceptors, and costimulatory molecules of the immune system, molecules involved in antigen presentation to lymphocytes, cell adhesion molecules, certain cytokine receptors, and intracellular muscle proteins. These are typically associated with roles in the immune system. Proteins in the immune synapse are often members of IgSF. IgSF can also be divided into "subfamilies" based on shared characteristics, such as function. Such subfamilies typically consist of 4 to 30 IgSF members.

[0101] The terms "IgSF domain" or "immunoglobulin domain" or "Ig domain" as used herein refer to the structural domain of an IgSF protein. Ig domains are named after immunoglobulin molecules. They contain approximately 70 to 110 amino acids and are classified according to their size and function. Ig domains possess a characteristic Ig fold, a sandwich-like structure formed by two sheets of antiparallel beta strands. Interactions between hydrophobic amino acids within the sandwich and highly conserved disulfide bonds formed between cysteine ​​residues in the B and F strands stabilize the Ig fold. One end of an Ig domain contains a region called the complementarity-determining region, which is important for the specificity of antibodies for their ligands. Ig-like domains can be classified as IgV, IgC1, IgC2, or IgI. Most Ig domains are either variable (IgV) or constant (IgC) domains. IgV domains, with nine β-strands, are generally longer than IgC domains, with seven β-strands. The Ig domains of some members of the IgSF resemble IgV domains in amino acid sequence, yet are similar in size to IgC domains. These are called IgC2 domains, while standard IgC domains are called IgC1 domains. T cell receptor (TCR) chains contain two Ig domains in their extracellular portions: one IgV domain at the N-terminus and one IgC1 domain adjacent to the cell membrane. ICOSL contains two Ig domains: IgV and IgC.

[0102] The term "IgSF species" as used herein refers to a group of IgSF member proteins that have identical or substantially identical primary amino acid sequences. Each mammalian immunoglobulin superfamily (IgSF) member defines a unique identity for all IgSF species belonging to that IgSF member. Thus, each IgSF family member is unique compared to other IgSF family members, and thus each species of a particular IgSF family member is unique compared to other IgSF family member species. Nevertheless, differences between molecules of the same IgSF species can arise due to differences in post-translational modifications such as glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation, and lipidation. Furthermore, minor sequence differences within a single IgSF species due to genetic polymorphism, as well as wild-type truncated forms of an IgSF species due to, for example, proteolytic cleavage, also constitute differences between other forms within a single IgSF species. A "cell surface IgSF species" is an IgSF species expressed on the surface of a cell (generally a mammalian cell).

[0103] The term "immune activity," as used herein in the context of mammalian lymphocytes such as T cells, refers to one or more of cell survival, cell proliferation, cytokine production (e.g., interferon-γ), or T-cytotoxic activity. In some cases, immune activity can refer to the cellular expression of cytokines such as chemokines or interleukins. Assays for determining enhanced or suppressed immune activity include the MLR (mixed lymphocyte reaction) assay, which measures interferon-γ cytokine levels in culture supernatants (Wang et al., Cancer Immunol Res. 2014 Sep: 2(9):846-56), the SEB (staphylococcal enterotoxin B) T-cell stimulation assay (Wang et al., Cancer Immunol Res. 2014 Sep: 2(9):846-56), and the anti-CD3 T-cell stimulation assay (Li and Kurlander, J Transl Med. 2010: 8: 104). Because T cell activation is associated with the secretion of IFN-γ cytokines, detection of IFN-γ levels in culture supernatants from these in vitro human T cell assays can be assayed using commercially available ELISA kits (Wu et al., Immunol Lett 2008 Apr 15; 117(1): 57-62). Induction of an immune response results in enhanced immune activity compared to resting lymphocytes. In some embodiments, immunomodulatory proteins (e.g., variant ICOSL polypeptides containing affinity-modified IgSF domains) as provided herein can increase, or alternatively decrease, IFN-γ (interferon-γ) expression in primary T cell assays compared to wild-type IgSF members or IgSF domain controls. Those skilled in the art will recognize that the format of the primary T cell assay used to determine increased IFN-γ expression differs from the format employed to assay for decreased IFN-γ expression.When assaying for the ability of an immunomodulating protein or affinity-modified IgSF domain of the invention to reduce IFN-γ expression in a primary T cell assay, a mixed lymphocyte reaction (MLR) assay can be used, as described in Example 6. Conveniently, a soluble form of an affinity-modified IgSF domain of the invention can be employed to determine its ability to reduce IFN-γ expression by antagonizing it in an MLR, also as described in Example 6. Alternatively, when assaying for the ability of an immunomodulating protein or affinity-modified IgSF domain of the invention to increase IFN-γ expression in a primary T cell assay, a co-immobilization assay can be used, in which a T cell receptor signal (provided, in some embodiments, by an anti-CD3 antibody) is used in combination with a co-immobilized affinity-modified IgSF domain, such as a variant ICOSL, to determine the ability to increase IFN-γ expression relative to a wild-type IgSF domain control. Methods for assaying the immune activity of modified cells, including assessing the activity of variant ICOSL transmembrane immunomodulatory proteins, are known in the art and include, but are not limited to, the ability to proliferate T cells after antigen stimulation, the ability to sustain T cell proliferation in the absence of restimulation, and anti-cancer activity in appropriate animal models. Assays are also standard. 51 Assays to assess cytotoxicity include Cr release assays (see, e.g., Milone et al., (2009) Molecular Therapy 17: 1453-1464) or flow-based cytotoxicity assays, or impedance-based cytotoxicity assays (Peper et al. (2014) Journal of Immunological Methods, 405:192-198).

[0104] An "immunomodulating polypeptide" is a polypeptide that modulates immune activity. "Modulating" an immune response means either enhancing or suppressing immune activity. An immunomodulating polypeptide can be a single polypeptide chain or a multimer (dimer or higher order multimer) of at least two polypeptide chains covalently linked to each other (e.g., by an interchain disulfide bond). Thus, monomeric, dimeric, and higher order multimeric polypeptides are within the scope of the defined term. Multimeric polypeptides can be homomultimers (of the same polypeptide chain) or heteromultimers (of different polypeptide chains). The immunomodulating polypeptides of the present invention include variant ICOSL.

[0105] The term "increase" as used herein means to increase by a statistically significant amount. The increase can be at least 5%, 10%, 20%, 30%, 40%, 50%, 75%, 100%, or greater than a non-zero control value.

[0106] An "isoform" of ICOSL (inducible costimulator ligand; CD275) is one of multiple naturally occurring ICOSL polypeptides that differ in amino acid sequence. Isoforms can be the product of splice variants of an RNA transcript expressed by a single gene, or the expression product of highly similar but different genes that produce functionally similar proteins, such as can result from gene duplication. As used herein, the term "isoform" of ICOSL also refers to the product of different alleles of the ICOSL gene (e.g., ICOSLG).

[0107] The term "lymphocyte," as used herein, refers to any of three subtypes of white blood cells of the mammalian immune system. These include natural killer cells (NK cells) (which function in cell-mediated cytotoxic innate immunity), T cells (which are involved in cell-mediated cytotoxic adaptive immunity), and B cells (which are involved in humoral antibody-mediated adaptive immunity). T cells include helper T cells, cytotoxic T cells, natural killer T cells, memory T cells, regulatory T cells, or γδ T cells. Also included within the definition of lymphocytes are innate lymphocytes (ILCs).

[0108] The term "mammal" or "patient" specifically includes reference to at least one of a human, chimpanzee, rhesus monkey, cynomolgus monkey, dog, cat, mouse, or rat.

[0109] The term "membrane protein," as used herein, refers to a protein that directly or indirectly attaches (binds) to a lipid bilayer under physiological conditions. The lipid bilayer that forms the membrane can be a biological membrane, such as a cell membrane of a eukaryote (e.g., a mammal), or an artificial (i.e., man-made) membrane, such as a membrane found on a liposome. The membrane protein can be bound to the lipid bilayer by a covalent bond or by a non-covalent interaction (e.g., a hydrophobic interaction or an electrostatic interaction). The membrane protein can be an integral membrane protein or a peripheral membrane protein. A membrane protein that is a peripheral membrane protein is bound to the lipid bilayer by a non-covalent interaction or to an integral membrane protein by a non-covalent interaction. The peripheral membrane protein forms a temporary bond to the lipid bilayer so that the peripheral membrane protein can interact with and / or dissociate from the lipid bilayer under physiological conditions in mammals. In contrast to peripheral membrane proteins, integral membrane proteins form a substantially permanent bond to the lipid bilayer of the membrane, so that they do not dissociate from the lipid bilayer under physiological conditions in mammals.Membrane proteins can form a bond to the membrane through one layer of the lipid bilayer (monotopic) or through both layers of the membrane (polytopic).Integral membrane proteins that interact with only one lipid bilayer are "integral monotopic proteins."Integral membrane proteins that interact with both lipid bilayers are "integral polytopic proteins."Alternatively, they are referred to herein as "transmembrane proteins."

[0110] The terms "modulation" or "modulating," as used herein in the context of immune responses (e.g., mammalian immune responses), refer to any change (e.g., enhancement or reduction) in an existing or potential immune response that occurs as a result of administration of an immunomodulatory polypeptide comprising a variant ICOSL of the present invention or as a result of administration of modified cells expressing an immunomodulatory protein of the present invention (e.g., a variant ICOSL transmembrane immunomodulatory protein). Thus, modulation refers to a change (e.g., enhancement or reduction) in an immune response compared to an immune response that occurs or exists in the absence of administration of an immunomodulatory protein comprising a variant ICOSL or a cell expressing such an immunomodulatory polypeptide. Such modulation includes any induction, activation, suppression, or change in the degree or extent of immune activity of immune cells. Immune cells include B cells, T cells, NK (natural killer) cells, NK T cells, professional antigen-presenting cells (APCs), and non-professional antigen-presenting cells, as well as inflammatory cells (neutrophils, macrophages, monocytes, eosinophils, and basophils). Modulation includes any change made to an existing, developing, or potential immune response, or to the ability to induce, regulate, influence, or respond to an immune response. Modulation includes any change in the expression and / or function of genes, proteins, and / or other molecules in immune cells as part of an immune response. Modulation of an immune response or modulation of immune activity includes, for example, the elimination, deletion, or sequestration of immune cells; the induction or generation of immune cells that can modulate the functional capacity of other cells, such as autoreactive lymphocytes, antigen-presenting cells, or inflammatory cells; the induction of an unresponsive state in immune cells (i.e., anergy); and the enhancement or suppression of immune cell activity or function (including, but not limited to, altering the pattern of proteins expressed by these cells). Examples include alterations in the production and / or secretion of specific classes of molecules, such as cytokines, chemokines, growth factors, transcription factors, kinases, costimulatory molecules, or other cell surface receptors, or any combination of these regulatory events.Modulation can be assessed, for example, by changes in IFN-γ (interferon gamma) expression compared to wild-type ICOSL controls in primary T cell assays (see Zhao and Ji, Exp Cell Res. 2016 Jan1;340(1) 132-138). Modulation can be assessed, for example, by changes in the immune activity of the modified cells compared to cells modified with wild-type ICOSL transmembrane protein, such as changes in the cytotoxic activity of the modified cells or changes in cytokine secretion of the modified cells.

[0111] The term "molecular species" as used herein refers to a population of proteins with identical or substantially identical primary amino acid sequences. Each mammalian immunoglobulin superfamily (IgSF) member defines a collection of identical or substantially identical molecular species. Thus, for example, human ICOSL is an IgSF member, and each human ICOSL molecule is an ICOS molecular species. Differences between molecules of the same molecular species can occur due to differences in post-translational modifications such as glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation, and lipidation. Furthermore, small sequence differences within a single molecular species due to genetic polymorphism, as well as wild-type truncated forms of a single molecular species due to, for example, proteolytic cleavage, also constitute differences between other forms within a single molecular species. A "cell surface molecular species" is a molecular species expressed on the surface of a mammalian cell. Two or more different protein species, each present in only one or the other (but not both) of two mammalian cells forming an IS, are said to be in "cis" or "cis configuration" with respect to each other. Two different protein species, the first of which is present only in the first of two mammalian cells that form an IS and the second of which is present only in the second of two mammalian cells that form an IS, are said to be in "trans" or in a "trans configuration." Two different protein species, each of which is present in both two mammalian cells that form an IS, are in both cis and trans configurations on those cells.

[0112] The term "multimerization domain" refers to an amino acid sequence that promotes stable interaction between a polypeptide molecule and one or more additional polypeptide molecules, each containing complementary multimerization domains, which can be the same or different multimerization domains that form stable multimers with the first domain. Generally, the polypeptide is directly or indirectly connected to the multimerization domain. Exemplary multimerization domains include immunoglobulin sequences or portions thereof, leucine zippers, hydrophobic regions, hydrophilic regions, and compatible protein-protein interaction domains. The multimerization domain can be, for example, an immunoglobulin constant region or domain, such as an Fc domain or portion thereof, derived from IgG (including IgG1, IgG2, IgG3, or IgG4 subtypes), IgA, IgE, IgD, and IgM, and modified forms thereof.

[0113] The terms "nucleic acid" and "polynucleotide" are used interchangeably and refer to a polymer of nucleic acid residues (e.g., deoxyribonucleotides or ribonucleotides) in either single-stranded or double-stranded form. Unless otherwise specified, the terms encompass nucleic acids containing known natural nucleotide analogs, nucleic acids with similar binding properties, and nucleic acids that are metabolized in a manner similar to naturally occurring nucleotides. Unless otherwise specified, a particular nucleic acid sequence also implicitly encompasses not only the sequence explicitly set forth ("reference sequence"), but also conservatively modified variants thereof (e.g., degenerate codon substitutions) and complementary nucleotide sequences. Specifically, degenerate codon substitutions may be achieved by generating sequences in which the third position of one or more selected (or all) codons is substituted with mixed-base and / or deoxyinosine residues. The term nucleic acid or polynucleotide encompasses cDNA or mRNA encoded by a gene.

[0114] The term "non-competitive binding," as used herein, refers to the ability of a protein to simultaneously specifically bind to at least two cognate binding partners. Thus, a protein can simultaneously bind to at least two different cognate binding partners, but the binding interactions do not necessarily have to be of the same duration; in some cases, the protein specifically binds to only one of the cognate binding partners. In some embodiments, binding occurs under specific binding conditions. In some embodiments, simultaneous binding is such that binding of one cognate binding partner does not substantially inhibit simultaneous binding to a second cognate binding partner. In some embodiments, non-competitive binding means that binding of a second cognate binding partner to its binding site on the protein does not displace binding of a first cognate binding partner to its binding site on the protein. Methods for assessing non-competitive binding are well known in the art, such as those described in Perez de La Lastra et al., Immunology, 1999 Apr: 96(4): 663-670. In some cases, in a non-competitive interaction, a first cognate binding partner specifically binds at an interaction site that does not overlap with the interaction site of a second cognate binding partner, such that the binding of the second cognate binding partner does not directly interfere with the binding of the first cognate binding partner. Thus, any effect of the binding of the second cognate binding partner on the binding of the cognate binding partner is via a mechanism other than direct interference with the binding of the first cognate binding partner. For example, in the context of enzyme-substrate interactions, a non-competitive inhibitor binds to a site other than the active site of the enzyme. Non-competitive binding encompasses non-competitive binding interactions in which a second cognate binding partner specifically binds at an interaction site that does not overlap with the binding of the first cognate binding partner, but only binds to the second interaction site when the first interaction site is occupied by the first cognate binding partner.

[0115] The term "pharmaceutical composition" refers to a composition suitable for pharmaceutical use in a mammalian subject, often a human. A pharmaceutical composition typically comprises an effective amount of an active agent (e.g., an immunomodulatory polypeptide comprising a variant ICOSL or a modified cell expressing a variant ICOSL transmembrane immunomodulatory protein) and a carrier, excipient, or diluent. The carrier, excipient, or diluent is typically a pharmaceutically acceptable carrier, excipient, or diluent, respectively.

[0116] The terms "polypeptide" and "protein" are used interchangeably herein and refer to a molecular chain of two or more amino acids linked via peptide bonds. The term does not refer to a specific length of the product. Thus, "peptide" and "oligopeptide" are included within the definition of polypeptide. The term includes post-translational modifications of polypeptides, such as glycosylation, acetylation, phosphorylation, etc. The term also includes molecules containing one or more amino acid analogs or non-standard or unnatural amino acids, which can be synthesized or recombinantly expressed using known protein modification techniques. In addition, proteins may be derivatized.

[0117] The term "primary T cell assay," as used herein, refers to an in vitro assay for measuring interferon-γ ("IFN-γ") expression. Various such primary T cell assays are known in the art, such as the assay described in Example 7. In a preferred embodiment, the assay used is an anti-CD3 co-fixation assay. In this assay, primary T cells are stimulated with immobilized anti-CD3, with or without additional recombinant proteins. Culture supernatants are collected at a certain time point (usually 24-72 hours). In another embodiment, the assay used is an MLR. In this assay, primary T cells are stimulated with allogeneic APCs. Culture supernatants are collected at a certain time point (usually 24-72 hours). Human IFN-γ levels in the culture supernatants are measured by standard ELISA techniques. Commercial kits are available from suppliers, and the assay is performed according to the manufacturer's recommendations.

[0118] The term "purified," as applied to nucleic acids (e.g., nucleic acids encoding immunomodulatory proteins of the invention), generally refers to nucleic acids or polypeptides that are substantially free of other components as determined by analytical techniques well known in the art (e.g., a purified polypeptide or polynucleotide forms a discrete band in an electrophoretic gel, a chromatographic eluate, and / or a medium subjected to density gradient centrifugation). For example, a nucleic acid or polypeptide that gives rise to essentially one band in an electrophoretic gel is "purified." Purified nucleic acids or proteins of the invention are at least about 50% pure, typically at least about 75%, 80%, 85%, 90%, 95%, 96%, 99% or more pure (e.g., on a weight percent or molar basis).

[0119] The term "recombinant" indicates that a substance (e.g., a nucleic acid or polypeptide) has been artificially (i.e., non-naturally) altered by human intervention. The alteration can be performed on a substance in or removed from its natural environment or state. For example, a "recombinant nucleic acid" is one produced by recombining nucleic acids, e.g., during cloning, affinity engineering, DNA shuffling, or other well-known molecular biological procedures. A "recombinant DNA molecule" is composed of segments of DNA joined together by such molecular biological techniques. The term "recombinant protein" or "recombinant polypeptide," as used herein, refers to a protein molecule expressed using a recombinant DNA molecule. A "recombinant host cell" is a cell that contains and / or expresses a recombinant nucleic acid, or a cell that has otherwise been genetically engineered (e.g., by introducing into the cell a nucleic acid molecule encoding a recombinant protein (e.g., a transmembrane immunomodulatory protein provided herein)). Transcriptional control signals in eukaryotes include "promoter" and "enhancer" elements. Promoters and enhancers consist of short arrays of DNA sequences that specifically interact with cellular proteins involved in transcription. Promoter and enhancer elements have been isolated from a variety of eukaryotic sources, including genes in yeast, insect and mammalian cells, and viruses (analogous control elements, i.e., promoters, are also found in prokaryotes). The selection of a particular promoter and enhancer depends on what cell type is to be used to express the protein of interest. The terms "in functional combination," "in functional order," and "operably linked," as used herein, refer to the linking of nucleic acid sequences in a manner or orientation that produces a nucleic acid molecule capable of directing the transcription of a given gene and / or the synthesis of a desired protein molecule.

[0120] The term "recombinant expression vector," as used herein, refers to a DNA molecule containing a desired coding sequence and appropriate nucleic acid sequences necessary for expression of the operably linked coding sequence in a particular host cell. Nucleic acid sequences necessary for expression in prokaryotes include a promoter, optionally an operator sequence, a ribosome binding site, and possibly other sequences. Eukaryotic cells are known to utilize promoters, enhancers, and termination and polyadenylation signals. If desired, a secretory signal peptide sequence may also be encoded by the recombinant expression vector operably linked to the coding sequence of the recombinant protein (e.g., a recombinant fusion protein) so that the expressed fusion protein can be secreted by the recombinant host cell for easier isolation of the fusion protein from the cell. This term includes vectors as autonomously replicating nucleic acid structures as well as vectors that integrate into the genome of a host cell into which they are introduced. Such vectors include viral vectors, such as lentiviral vectors.

[0121] The term "selectivity" refers to the preference for specific binding of a protein or polypeptide of interest to one substrate (e.g., one cognate binding partner) compared to the specific binding of the protein of interest to another substrate (e.g., a different cognate binding partner). Selectivity refers to the binding activity (e.g., binding affinity) (e.g., K) of a protein of interest to a first substrate (e.g., a first cognate binding partner). d1 ) and the binding activity (e.g., binding affinity) (e.g., K d2 ) can be reflected as a ratio.

[0122] The term "sequence identity" as used herein refers to the sequence identity between genes or proteins at the nucleotide or amino acid level, respectively. "Sequence identity" is a measure of identity between proteins at the amino acid level and between nucleic acids at the nucleotide level. Protein sequence identity can be determined by comparing the amino acid sequence at a given position in each sequence when the sequences are aligned. Similarly, nucleic acid sequence identity can be determined by comparing the nucleotide sequence at a given position in each sequence when the sequences are aligned. Methods for aligning sequences for comparison are well known in the art, and include GAP, BESTFIT, BLAST, FASTA, and TFASTA. The BLAST algorithm calculates the percent sequence identity and performs a statistical analysis of the similarity between two sequences. Software for performing BLAST analysis is publicly available through the website of the National Center for Biotechnology Information (NCBI).

[0123] The term "soluble" when used herein in reference to a protein means that the protein is not a membrane protein. Generally, a soluble protein contains only the extracellular domain or a portion thereof of an IgSF family member receptor containing an IgSF domain or its specific binding fragment, but does not contain a transmembrane domain. In some cases, the solubility of a protein can be improved by linking or binding to an Fc domain directly or indirectly via a linker, which in some cases can also improve the stability and / or half-life of the protein. In some aspects, the soluble protein is an Fc fusion protein.

[0124] The term "species," as used herein with respect to a polypeptide or nucleic acid, refers to a group of molecules having identical or substantially identical sequences. Differences between polypeptides of the same species can occur due to differences in post-translational modifications, such as glycosylation, phosphorylation, ubiquitination, nitrosylation, methylation, acetylation, and lipidation. Slightly truncated polypeptide sequences that differ from the full-length species by (or encode) only one, two, or three amino acid residues at the amino or carboxy terminus are considered to be sequences of a single species. Such microheterogeneity is a common feature of manufactured proteins.

[0125] The term "specific binding fragment," as used herein with reference to a full-length wild-type mammalian ICOSL polypeptide or its IgV or IgC domain, refers to a polypeptide that has a subsequence of the IgV and / or IgC domain and specifically binds to mammalian ICOS and / or mammalian CD28 (e.g., human or murine ICOS or CD28) in vitro and / or in vivo. In some embodiments, a specific binding fragment of ICOSL IgV or ICOSL IgC is at least 60%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, or 99% of the sequence length of the full-length wild-type sequence. The sequence of the specific binding fragment can be altered to form a variant ICOSL of the invention.

[0126] The term "specifically binds" as used herein refers to the ability of a protein to bind to a target protein under specific binding conditions such that its affinity or avidity is at least 5 times greater than the average affinity or avidity of the same protein for a population of random peptides or polypeptides of sufficient statistical size, but optionally at least 10, 20, 30, 40, 50, 100, 250, or 500 times greater, or even at least 1000 times greater. A specifically binding protein need not bind only to a single target molecule, but may specifically bind to a non-target molecule (e.g., a paralog or ortholog) due to the similarity in configuration between the target and the non-target. Those skilled in the art will recognize that specific binding to molecules with the same function in different animal species (i.e., orthologs) or to non-target molecules (e.g., paralogs) with substantially similar epitopes to the target molecule is possible and does not impair the specificity of binding determined for a statistically valid population of unique non-targets (e.g., random polypeptides). Therefore, due to cross-reactivity, the polypeptides of the present invention can specifically bind to multiple distinct target molecular species. Solid-phase ELISA immunoassays or Biacore measurements can be used to determine the specific binding between two proteins. Generally, the interaction between two binding proteins is greater than 1×10 -5 Less than M, often 1 × 10 -12 In certain embodiments of the present disclosure, the interaction between two binding proteins has a dissociation constant (Kd) as low as 1×10 -6 M, 1 x 10 -7 M, 1 x 10 -8 M, 1 x 10 -9 M, 1 x 10 -10 M or 1×10 -11 It has a dissociation constant of M.

[0127] With respect to mammalian cells expressing a polypeptide, the term "surface-expressed" or "surface expression" means that the polypeptide is expressed as a membrane protein. In some embodiments, the membrane protein is a transmembrane protein.

[0128] As used herein, "synthetic" refers to a nucleic acid molecule or polypeptide molecule that is produced by recombinant and / or chemical synthesis methods, e.g., with respect to a synthetic nucleic acid molecule or synthetic gene or synthetic peptide.

[0129] The term "targeting moiety," as used herein, refers to a composition that is covalently or non-covalently bound to or physically encapsulates a polypeptide comprising a variant ICOSL of the present invention. A targeting moiety has specific binding affinity for a desired counter structure, such as a cell surface receptor (e.g., the B7 family member PD-L1) or a tumor antigen (e.g., a tumor-specific antigen (TSA) or tumor-associated antigen (TAA), such as B7-H6). Typically, the desired counter structure is localized on a specific tissue or cell type. A targeting moiety can be an antibody, an antigen-binding fragment (Fab), a V-cell, or a mAb. H and V L variable fragments (Fv) containing V linked together in one chain; H and V L and other antibody V region fragments, such as Fab', F(ab)2, F(ab')2, dsFv diabodies, nanobodies, soluble receptors, receptor ligands, affinity matured receptors or ligands, and small molecule (<500 dalton) compositions (e.g., specific binding receptor compositions). Targeting moieties can also be covalently or non-covalently bound (attached) to the lipid membrane of liposomes that encapsulate the polypeptides of the invention.

[0130] As used herein, the term "transmembrane protein" refers to a membrane protein that substantially or completely spans a lipid bilayer, which is found, for example, in biological membranes such as mammalian cells, or in artificial constructs such as liposomes. A transmembrane protein comprises a transmembrane domain ("transmembrane domain") that is integrated into the lipid bilayer and whose integration is thermodynamically stable under physiological conditions. The transmembrane domain can generally be predicted from the amino acid sequence of the transmembrane domain through several commercially available bioinformatics software applications, based on the fact that the hydrophobicity of the transmembrane domain is higher than that of the region of the protein that interacts with the aqueous environment (e.g., cytosol, extracellular fluid). The transmembrane domain is often a hydrophobic α-helix that spans the membrane. A transmembrane protein may span both layers of the lipid bilayer once or multiple times. The transmembrane immunomodulatory proteins provided herein are included in transmembrane proteins. In addition to the transmembrane domain, the transmembrane immunomodulatory proteins of the present invention further comprise an ectodomain, and in some embodiments, an endodomain.

[0131] The term "treatment" or "therapy" of a disease or disorder, as used herein, means slowing, halting, or reversing the progression of the disease or disorder, as evidenced by the reduction, arrest, or elimination of any clinical or diagnostic symptoms by administering a therapeutic composition of the invention (e.g., containing an immunomodulatory protein or modified cells), either alone or in combination with another compound as described herein. "Treating" or "treatment" also means reducing the severity of symptoms in acute or chronic diseases or disorders, or reducing the rate of relapse (e.g., as in relapsing or remitting autoimmune disease processes), or reducing inflammation in the case of inflammatory aspects of autoimmune diseases. As used herein in the context of cancer, the terms "treating" cancer or "inhibiting" cancer or "inhibition" of cancer refer to at least one of a statistically significant reduction in tumor growth rate, halting tumor growth, or a reduction in tumor size, mass, metabolic activity, or volume, or a statistically significant improvement in progression-free survival (PFS) or overall survival (OS), as measured by standard criteria such as, but not limited to, Response Evaluation Criteria for Solid Tumors (RECIST). "Preventing" a disease or disorder or "prevention" of a disease or disorder, as used in the context of the present invention, refers to the administration of an immunomodulatory polypeptide or modified cell of the present invention, either alone or in combination with another compound, to prevent the appearance or development of a disease or disorder or some or all of the symptoms of a disease or disorder, or to reduce the likelihood of developing a disease or disorder.

[0132] The term "tumor-specific antigen" or "TSA," as used herein, refers to a counter structure that is primarily present on tumor cells of a mammalian subject but is not generally found on normal cells of the mammalian subject. A tumor-specific antigen need not be present exclusively on tumor cells, but the proportion of cells in a particular mammal that have the tumor-specific antigen or the level of the tumor-specific antigen on the surface of the tumor must be sufficiently high so that it can be targeted with an anti-tumor therapeutic agent (e.g., an immunomodulatory polypeptide of the present invention) and provide protection or treatment for the mammal from the effects of the tumor. In some embodiments, in a random statistical sample of cells from a mammal with a tumor, at least 50% of the cells that display the TSA are cancerous. In other embodiments, at least 60%, 70%, 80%, 85%, 90%, 95%, or 99% of the cells that display the TSA are cancerous.

[0133] The term "variant" (also "modified" or "mutant"), when used with reference to variant ICOSL, refers to ICOSL created by human intervention, e.g., mammalian (e.g., human or murine) ICOSL. Variant ICOSL is a polypeptide having an altered amino acid sequence compared to unmodified or wild-type ICOSL. Variant ICOSL is a polypeptide that differs from the wild-type ICOSL isoform sequence by one or more amino acid substitutions, deletions, additions, or a combination thereof. For purposes herein, variant ICOSL contains at least one affinity-altered domain, whereby one or more of the amino acid differences occur in the IgSF domain (e.g., the IgV domain). A variant ICOSL can contain 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 or more amino acid differences, e.g., amino acid substitutions. A variant ICOSL polypeptide generally exhibits at least 50%, 60%, 70%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to the corresponding wild-type or unmodified ICOSL, e.g., the sequence of SEQ ID NO:5, its mature sequence, or a portion thereof containing the extracellular domain or IgSF domain. In some embodiments, the variant ICOSL polypeptide exhibits at least 50%, 60%, 70%, 80%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or more sequence identity to the corresponding wild-type or unmodified ICOSL, including the sequence set forth in SEQ ID NO: 32 or SEQ ID NO: 196. Both non-naturally occurring and naturally occurring amino acids are included within the scope of permissible substitutions or additions. The variant ICOSL is not limited to any particular production method, including, for example, de novo chemical synthesis, de novo recombinant DNA technology, or a combination thereof.The variant ICOSL of the present invention specifically binds to at least one or more of CD28, ICOS, or CTLA-4 of mammalian species. In some embodiments, the altered amino acid sequence results in a change (i.e., an improvement or decrease) in the binding affinity or avidity to ICOS and / or CD28 compared to wild-type ICOSL protein. The improvement or decrease in binding affinity or avidity can be determined using well-known binding assays such as flow cytometry. Larsen et al., American Journal of Transplantation, Vol. 5: 443-453 (2005). See also Linsley et al., Immunity, 1: 7930801 (1994). The improved binding affinity or avidity of variant ICOSL to ICOS and / or CD28 is at least 5% greater than that of wild-type ICOSL, and in some embodiments, at least 10%, 15%, 20%, 30%, 40%, 50%, or 100% greater than that of the wild-type ICOSL control value. The decreased binding affinity or avidity of ICOSL to ICOS and / or CD28 is at most 95% of the wild-type control value, and in some embodiments, at most 80%, 70%, 60%, 50%, 40%, 30%, 20%, 10%, 5% or less of the binding affinity or avidity of the wild-type ICOS and / or CD28 control value, or to an undetectable level. Variant ICOSL has an altered primary amino acid sequence due to substitution, addition, or deletion of amino acid residues. The term "variant" in the context of variant ICOSL should not be construed as imposing any particular starting composition or method by which the variant ICOSL is made. Variant ICOSL can be generated, for example, starting from wild-type mammalian ICOSL sequence information, then modeled in silico for binding to ICOS and / or CD28, and finally recombinantly or chemically synthesized to generate the variant ICOSL of the present invention.As just another example, variant ICOSL can be generated by site-directed mutagenesis of wild-type ICOSL. Thus, variant ICOSL represents a composition, but not necessarily a product, produced by any given process. A variety of techniques may be employed, including recombinant methods, chemical synthesis, or a combination thereof.

[0134] The terms "wild-type" or "natural" or "native" as used herein in connection with biological material, such as a nucleic acid molecule, a protein (e.g., ICOSL), an IgSF member, a host cell, etc., refer to one that is found in nature and has not been modified by human intervention.

[0135] II. Variant ICOSL Polypeptides Provided herein are variant ICOSL polypeptides that have altered (enhanced or decreased) binding activity or affinity for one or more ICOSL cognate binding partners. In some embodiments, the ICOSL cognate binding partner is CD28, ICOS, or CTLA-4. In some embodiments, the variant ICOSL polypeptide contains one or more amino acid modifications, such as one or more substitutions (alternatively, "mutations" or "exchanges"), deletions, or additions, in the immunoglobulin superfamily (IgSF) domain (IgD) compared to a wild-type or unmodified ICOSL polypeptide, or a portion containing the immunoglobulin superfamily (IgSF) domain of wild-type or unmodified ICOSL, or a specific-binding fragment thereof. Thus, the provided variant ICOSL polypeptides are or include variant IgDs (hereinafter referred to as "vIgD") in which one or more amino acid modifications (e.g., substitutions) are present in IgD.

[0136] In some embodiments, the IgD comprises an IgV domain or an IgC (e.g., IgC2) domain, or a specific-binding fragment of an IgV domain or an IgC (e.g., IgC2) domain, or a combination thereof. In some embodiments, the IgD can be simply an IgV, a combination of IgV and IgC including the entire extracellular domain (ECD), or any combination of the Ig domains of ICOSL. Table 2 provides exemplary residues corresponding to the IgV or IgC regions of ICOSL. In some embodiments, the variant ICOSL polypeptide contains an IgV or IgC domain, or a specific-binding fragment thereof, in which at least one of the amino acid modifications (e.g., substitutions) is in the IgV or IgC domain, or a specific-binding fragment thereof. In some embodiments, the IgV or IgC domain is an affinity-modified IgSF domain due to altered binding activity or affinity.

[0137] In some embodiments, the variant has a modification in another IgSF domain compared to the unmodified ICOSL sequence. In some embodiments, the unmodified ICOSL sequence is wild-type ICOSL. In some embodiments, the unmodified or wild-type ICOSL has the sequence of native ICOSL or its orthologue. In some embodiments, the unmodified ICOSL is or comprises the extracellular domain (ECD) of ICOSL, or a portion thereof, containing one or more IgSF domains (see Table 2). In some embodiments, the extracellular domain of the unmodified or wild-type ICOSL polypeptide comprises an IgV domain and an IgC domain. However, the variant ICOSL polypeptide need not comprise both an IgV domain and an IgC domain. In some embodiments, the variant ICOSL polypeptide comprises or essentially consists of an IgV domain or a specific binding fragment thereof. In some embodiments, the variant ICOSL polypeptide comprises or essentially consists of an IgC domain or a specific binding fragment thereof. In some embodiments, the variant ICOSL is soluble and lacks a transmembrane domain. In some embodiments, the variant ICOSL further comprises a transmembrane domain, and in some cases, a cytoplasmic domain.

[0138] In some embodiments, the wild-type or unmodified ICOSL sequence is a mammalian ICOSL sequence. In some embodiments, the wild-type or unmodified ICOSL sequence can be a mammalian ICOSL sequence, including but not limited to, human, mouse, cynomolgus monkey, or rat. In some embodiments, the wild-type or unmodified ICOSL sequence is human.

[0139] In some embodiments, the wild-type or unmodified ICOSL sequence has (i) the amino acid sequence set forth in SEQ ID NO:5 or a mature form thereof lacking the signal sequence, (ii) an amino acid sequence or a mature form thereof that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to SEQ ID NO:5, or (iii) an IgV domain or IgC domain or a specifically binding fragment thereof of (i) or (ii) is a portion thereof.

[0140] In some embodiments, the wild-type or unmodified ICOSL sequence is or comprises the extracellular domain of ICOSL or a portion thereof. In some embodiments, the unmodified or wild-type ICOSL polypeptide comprises the amino acid sequence set forth in SEQ ID NO:32, or an orthologue thereof. In some cases, the unmodified or wild-type ICOSL polypeptide can (i) comprise the amino acid sequence set forth in SEQ ID NO:32, (ii) comprise an amino acid sequence having at least about 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% sequence identity to SEQ ID NO:32, or (iii) be a specific-binding fragment of the sequence of (i) or (ii) comprising the IgV domain or the IgC domain.

[0141] In some embodiments, a wild-type or unmodified ICOSL polypeptide comprises an IgV domain or an IgC domain, or a specific-binding fragment thereof. In some embodiments, the IgV domain of a wild-type or unmodified ICOSL polypeptide comprises the amino acid sequence set forth in SEQ ID NO:196 (corresponding to amino acid residues 19-129 of SEQ ID NO:5), or an ortholog thereof. For example, the IgV domain of an unmodified or wild-type ICOSL polypeptide can contain (i) the amino acid sequence set forth in SEQ ID NO:196, (ii) an amino acid sequence having at least about 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to SEQ ID NO:196, or (iii) a specific-binding fragment of the amino acid sequence set forth in SEQ ID NO:196 or a specific-binding fragment of the sequence of (i) or (ii). In some embodiments, the wild-type or unmodified IgV domain is capable of binding to one or more ICOSL cognate binding proteins, for example, one or more of CD28, ICOS, or CTLA-4.

[0142] In some embodiments, the IgC domain of a wild-type or unmodified ICOSL polypeptide comprises the amino acid sequence set forth as residues 141-227 of SEQ ID NO:5, or an ortholog thereof. For example, the IgC domain of an unmodified or wild-type ICOSL polypeptide can contain (i) the amino acid sequence set forth as residues 141-227 of SEQ ID NO:5, (ii) an amino acid sequence having at least about 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to residues 141-227 of SEQ ID NO:5, or (iii) (i) or (ii). In some embodiments, the wild-type or unmodified IgV domain is capable of binding to one or more ICOSL cognate binding proteins.

[0143] In some embodiments, the wild-type or unmodified ICOSL polypeptide contains a specific-binding fragment of ICOSL, such as a specific-binding fragment of the IgV domain or the IgC domain. In some embodiments, the specific-binding fragment can bind to CD28, ICOS, and / or CTLA4. The amino acid length of the specific-binding fragment can be at least 50 amino acids, e.g., at least 60, 70, 80, 90, 100, or 110 amino acids. In some embodiments, the specific-binding fragment of the IgV domain contains an amino acid sequence that is at least about 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% of the length of the IgV domain set forth as amino acids 19 to 129 of SEQ ID NO:5. In some embodiments, a specific binding fragment of an IgC domain comprises an amino acid sequence that is at least about 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% of the length of the IgC domain set forth as amino acids 141 to 227 of SEQ ID NO:5.

[0144] In some embodiments, the variant ICOSL polypeptide comprises an ECD domain or a portion thereof comprising one or more affinity-modified IgSF domains. In some embodiments, the variant ICOSL polypeptide comprises an IgV domain or an IgC domain, wherein one or more of the IgSF domain (IgV or IgC) or the specific-binding fragment of the IgV domain or the specific-binding fragment of the IgC domain contains one or more amino acid modifications (e.g., substitutions). In some embodiments, the variant ICOSL polypeptide comprises an IgV domain and an IgC domain, or a specific-binding fragment of the IgV domain and a specific-binding fragment of the IgC domain. In some embodiments, the variant ICOSL polypeptide comprises a full-length IgV domain. In some embodiments, the variant ICOSL polypeptide comprises a full-length IgC domain. In some embodiments, the variant ICOSL polypeptide comprises a specific-binding fragment of the IgV domain. In some embodiments, the variant ICOSL polypeptide comprises a specific-binding fragment of the IgC domain. In some embodiments, the variant ICOSL polypeptide comprises a full-length IgV domain and a full-length IgC domain. In some embodiments, the variant ICOSL polypeptide comprises a full-length IgV domain and a specific binding fragment of the IgC domain. In some embodiments, the variant ICOSL polypeptide comprises a specific binding fragment of the IgV domain and a full-length IgC domain. In some embodiments, the variant ICOSL polypeptide comprises a specific binding fragment of the IgV domain and a specific binding fragment of the IgC domain.

[0145] In any of such embodiments, the one or more amino acid modifications (e.g., substitutions) of the variant ICOSL polypeptide can be located in any one or more of the ICOSL polypeptide domains. For example, in some embodiments, the one or more amino acid substitutions are located in the extracellular domain of the variant ICOSL polypeptide. In some embodiments, the one or more amino acid substitutions are located in the IgV domain or a specific-binding fragment of the IgV domain. In some embodiments, the one or more amino acid modifications (e.g., substitutions) are located in the IgC domain or a specific-binding fragment of the IgC domain.

[0146] Generally, each of the various attributes of the polypeptides (e.g., soluble, secreted, and membrane-bound polypeptides, affinity of ICOSL for CD28, ICOS, and CTLA-4, number of differences per polypeptide chain, number of linked polypeptide chains, number and nature of amino acid modifications per variant ICOSL, etc.) are separately disclosed below. However, as will be apparent to those skilled in the art, any particular polypeptide can include a combination of these independent attributes. It will be understood that references to amino acids, including references to specific sequences set forth as SEQ ID NOs used to describe the domain organization of IgSF domains, are for illustrative purposes and are not intended to limit the scope of the provided embodiments. It will be understood that descriptions of polypeptides and their domains are theoretically derived based on homology analysis and alignment with similar molecules. Thus, the exact locus may vary and is not necessarily the same from protein to protein. Thus, a particular IgSF domain (e.g., a particular IgV or IgC domain) may be longer or shorter by a few amino acids (e.g., 1, 2, 3, or 4).

[0147] Furthermore, the various aspects of the present invention as discussed below are frequently provided within the meanings of defined terms as disclosed above. Therefore, the aspects described in a particular definition should be construed as being incorporated by reference when the defined terms are utilized in the discussion of the various aspects and attributes described herein. Thus, the headings, the order of presentation of the various aspects and embodiments, and the separate disclosure of each independent attribute are not meant to limit the scope of the present disclosure.

[0148] Exemplary Modifications Provided herein are variant ICOSL polypeptides containing at least one affinity-modified IgSF domain (e.g., IgV or IgC) or a specific-binding fragment thereof in the IgSF domain contained in a wild-type or unmodified ICOSL polypeptide, wherein the variant ICOSL polypeptide has altered (improved or decreased) binding activity or affinity for one or more ligands, ICOS, CD28, or CTLA-4, compared to the wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has a binding affinity for CD28, ICOS, and / or CTLA-4 that is different from that of a wild-type or unmodified ICOSL polypeptide control sequence, as determined, for example, by solid-phase ELISA immunoassay, flow cytometry, or Biacore assay. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28, ICOS, and / or CTLA-4. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28, ICOS, and / or CTLA-4 compared to the wild-type or unmodified ICOSL polypeptide. CD28, ICOS and / or CTLA-4 can be mammalian proteins, such as human or murine proteins.

[0149] The binding affinity for each of the cognate binding partners is independent; i.e., in some embodiments, the variant ICOSL polypeptide has improved binding affinity for one, two or three of CD28, ICOS, and / or CTLA-4, and decreased binding affinity for one, two or three of CD28, ICOS, and CTLA-4, compared to wild-type or unmodified ICOSL polypeptide.

[0150] In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide.

[0151] In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and reduced binding affinity for ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and improved binding affinity for ICOS compared to wild-type or unmodified ICOSL polypeptide.

[0152] In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and reduced binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and improved binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide.

[0153] In some embodiments, the variant ICOSL polypeptide has improved binding affinity for ICOS and CTLA-4 compared to wild-type or unmodified ICOS polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for ICOS and reduced binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for ICOS and CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for ICOS and improved binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide.

[0154] In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28, ICOS, and CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and ICOS and reduced binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and CTLA-4 and reduced binding affinity for ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and ICOS and improved binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and ICOS and improved binding affinity for CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has improved binding affinity for CD28 and reduced binding affinity for ICOS and CTLA-4 compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28, CTLA-4, and ICOS compared to wild-type or unmodified ICOSL polypeptide. In some embodiments, the variant ICOSL polypeptide has reduced binding affinity for CD28 and improved binding affinity for ICOS and CTLA-4 compared to wild-type or unmodified ICOSL polypeptide.

[0155] In some embodiments, a variant ICOSL polypeptide having increased or greater binding affinity to CD28, ICOS, and / or CTLA-4 will have an increase in binding affinity of at least about 5%, e.g., at least about 10%, 15%, 20%, 25%, 35%, or 50%, to CD28, ICOS, and / or CTLA-4 compared to a wild-type or unmodified ICOSL polypeptide control. In some embodiments, the increase in binding affinity compared to a wild-type or unmodified ICOSL polypeptide is greater than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 30-fold, 40-fold, or 50-fold. In such examples, the wild-type or unmodified ICOSL polypeptide has the same sequence as the variant ICOSL polypeptide, except that it does not contain one or more amino acid modifications (e.g., substitutions).

[0156] In some embodiments, a variant ICOSL polypeptide having reduced or decreased binding affinity to CD28, ICOS, and / or CTLA-4 will have a decrease in binding affinity of at least 5%, e.g., at least about 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90% or more, to CD28, ICOSL, and / or CTLA-4 compared to a wild-type or unmodified ICOSL polypeptide control. In some embodiments, the decrease in binding affinity compared to a wild-type or unmodified ICOSL polypeptide is more than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 30-fold, 40-fold, or 50-fold. In such examples, the wild-type or unmodified ICOSL polypeptide has the same sequence as the variant ICOSL polypeptide, except that it does not contain one or more amino acid modifications, e.g., substitutions.

[0157] In some embodiments, the equilibrium dissociation constant (K) of any of the foregoing embodiments to CD28, ICOS, and / or CTLA-4 is d ) is 1×10 -5 M, 1 x 10 -6 M, 1 x 10 -7M, 1 x 10 -8 M, 1 x 10 -9 M, 1 x 10 -10 M or 1×10 -11 M, or 1 x 10 -12 It can be less than M.

[0158] In some embodiments, the variant ICOSL polypeptide has increased or greater binding affinity to CD28. In some embodiments, the variant ICOSL polypeptide with increased or greater binding affinity to CD28 will have an increase in binding affinity for CD28 of at least about 25%, e.g., at least about 30%, 40%, 50%, or 60%, compared to a wild-type or unmodified ICOSL polypeptide control. In some embodiments, the variant ICOSL polypeptide with increased or greater binding affinity to CD28 will have an equilibrium dissociation constant (K) for CD28 of less than 200 pM, 300 pM, 400 pM, 500 pM, or 600 pM. d In some embodiments, the variant polypeptide specifically binds to the ectodomain of one of ICOS, CD28, or CTLA4 with improved selectivity compared to unmodified ICOSL. In some embodiments, the improved selectivity is for CD28. In some embodiments, the improved selectivity comprises a greater ratio between the binding of the variant ICOSL polypeptide to one cognate binding partner selected from among ICOS, CD28, and CTLA4 and the binding of the variant ICOSL polypeptide to another of the cognate binding partners compared to the ratio between the binding of the unmodified ICOSL polypeptide to the one cognate binding partner and the binding of the unmodified ICOSL polypeptide to the other of the cognate binding partners. In some embodiments, the ratio is at least or at least about 1.5-fold, 2.0-fold, 3.0-fold, 4.0-fold, 5-fold, 10-fold, 15-fold, 20-fold, 30-fold, 40-fold, 50-fold, or more.

[0159] A wild-type or unmodified ICOSL sequence need not necessarily be used as the starting material to generate the variant ICOSL polypeptides described herein. Therefore, the use of the term "modification," e.g., "substitution," does not imply that the present embodiment is limited to a particular method of producing the variant ICOSL polypeptide. Variant ICOSL polypeptides can be produced, for example, by de novo peptide synthesis and therefore do not necessarily require modification, e.g., "substitution," in the sense of altering a codon to encode the modification, e.g., substitution. This principle also extends to the terms "addition" and "deletion" of amino acid residues, which also do not imply a particular method of production. The means by which variant ICOSL polypeptides are designed or generated are not limited to any particular method. However, in some embodiments, nucleic acids encoding wild-type or unmodified ICOSL are mutagenized from wild-type or unmodified ICOSL genetic material and screened for the desired specific binding affinity and / or induction of IFN-γ expression or other functional activity. In some embodiments, variant ICOSL polypeptides are synthesized de novo using protein or nucleic acid sequences available in a number of publicly available databases, which are then subsequently screened. The National Center for Biotechnology Information provides such information, and its website is publicly accessible via the internet, as is the UniProtKB database, as discussed above.

[0160] Unless otherwise specified, as indicated throughout this disclosure, amino acid substitutions are designated by amino acid position numbers that correspond to the numbering of positions in the unmodified ECD sequence shown in SEQ ID NO:32, or, where applicable, the unmodified IgV sequence shown in SEQ ID NO:196 (containing residues 19-129 of SEQ ID NO:32), as follows: TIFF0007623967000018.tif33160TIFF0007623967000019.tif19160

[0161] It is within the level of ordinary skill in the art to identify the corresponding position of a modification (e.g., amino acid substitution) in an ICOSL polypeptide (including a portion thereof containing its IgSF domain (e.g., IgV)) by alignment with a reference sequence having, for example, SEQ ID NO:32 or SEQ ID NO:196. In descriptions of modifications throughout this disclosure, the amino acid position is shown in the center, the corresponding unmodified (e.g., wild-type) amino acid is listed before the number, and the amino acid substitution of the identified variant is listed after the number. If the modification is a deletion at that position, it is indicated as "del," and if the modification is an insertion at that position, it is indicated as "ins."

[0162] In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) in the wild-type or unmodified ICOSL sequence. The one or more amino acid modifications (e.g., substitutions) can be in the ectodomain (extracellular domain) of the wild-type or unmodified ICOSL sequence. In some embodiments, the one or more amino acid modifications (e.g., substitutions) are in the IgV domain or a specific binding fragment thereof. In some embodiments, the one or more amino acid modifications (e.g., substitutions) are in the IgC domain or a specific binding fragment thereof. In some embodiments of the variant ICOSL polypeptide, some of the one or more amino acid modifications (e.g., substitutions) are in the IgV domain or a specific binding fragment thereof, and some of the one or more amino acid modifications (e.g., substitutions) are in the IgC domain or a specific binding fragment thereof.

[0163] In some embodiments, the variant ICOSL polypeptide has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid modifications (e.g., substitutions). The modifications (e.g., substitutions) can be in the IgV domain or the IgC domain. In some embodiments, the variant ICOSL polypeptide has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid substitutions in the IgV domain or a specific-binding fragment thereof. In some embodiments, the variant ICOSL polypeptide has up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid substitutions in the IgC domain or specific binding fragment thereof. In some embodiments, the variant ICOSL polypeptide has at least about 85%, 86%, 86%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to a wild-type or unmodified ICOSL polypeptide or specific binding fragment thereof (e.g., the amino acid sequence of SEQ ID NO:32 or 196).

[0164] In some embodiments, the variant ICOSL polypeptide is located at positions 10, 11, 13, 16, 18, 20, 25, 27, 30, 33, 37, 42, 43, 47, 52, 54, 57, 61, 62, 67, 71, 72, 74, 77, 78, 75, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98, 99, 100, 102, 103, 107, 109, 110, 111, 113, 115, 116, 117, 119, 120, 121, 122, 126, 129, 130, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183 138, 139, 140, 142, 143, 144, 146, 151, 152, 153, 154, 155, 156, 158, 161, 166, 168, 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210, 212, 217, 218, 220, 221, 224, 225, or 227. In some embodiments, such variant ICOSL polypeptides have altered binding affinity for one or more of CD28, ICOS, and / or CTLA-4 compared to wild-type or unmodified ICOSL polypeptides. For example, in some embodiments, the variant ICOSL polypeptide exhibits improved binding affinity to CD28, ICOS, and / or CTLA-4 compared to the wild-type or unmodified ICOSL polypeptide, hi some embodiments, the variant ICOSL polypeptide exhibits decreased binding affinity to CD28, ICOS, or CTLA-4 compared to the wild-type or unmodified ICOSL polypeptide.

[0165] In some embodiments, the variant ICOSL polypeptide is TIFF0007623967000020.tif77165 or conservative amino acid modifications (e.g., substitutions) thereof. A conservative amino acid modification (e.g., substitution) is any amino acid other than the wild-type or unmodified amino acid that belongs to the same class of amino acids as the substituting amino acid. The amino acid classes are aliphatic (glycine, alanine, valine, leucine, and isoleucine), hydroxyl- or sulfur-containing (serine, cysteine, threonine, and methionine), cyclic (proline), aromatic (phenylalanine, tyrosine, tryptophan), basic (histidine, lysine, and arginine), and acidic / amide (aspartic acid, glutamic acid, asparagine, and glutamine).

[0166] In some embodiments, the variant ICOSL polypeptide is TIFF0007623967000021.tif77165.

[0167] In some embodiments, the one or more amino acid modifications (e.g., substitutions) are TIFF0007623967000022.tif99160TIFF0007623967000023.tif223166TIFF0007623967000024.tif143165.

[0168] In some embodiments, the variant ICOSL polypeptide comprises any of the mutations listed in Table 1. Table 1 also provides exemplary sequences by reference to SEQ ID NOs for the extracellular domain (ECD) or IgV domain of wild-type ICOSL or exemplary variant ICOSL polypeptides. As indicated, the exact locus or residues corresponding to a given domain may vary, depending, for example, on the method used to identify or classify the domain. In addition, in some cases, adjacent amino acids N- and / or C-terminal to a given domain (e.g., IgV) can also be included in the sequence of the variant IgSF polypeptide, for example, to ensure proper folding of the domain when expressed. Therefore, it is understood that the exemplary SEQ ID NOs in Table 1 should not be construed as limiting. For example, a particular domain (e.g., the ECD domain) of a variant ICOSL polypeptide can be several amino acids longer or shorter than the amino acid sequence set forth in the respective SEQ ID NO, e.g., 1 to 10 amino acids, e.g., 1, 2, 3, 4, 5, 6, or 7 amino acids longer or shorter.

[0169] In some embodiments, the variant ICOSL polypeptide comprises any of the mutations listed in Table 1. In some embodiments, the variant ICOSL polypeptide comprises any of the extracellular domain (ECD) sequences listed in Table 1 (i.e., any one of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470). In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, such as at least 96% identity, 97% identity, 98% identity, or 99% identity to any of the extracellular domain (ECD) sequences listed in Table 1 (i.e., any one of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470), and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific binding fragment of any of the extracellular domain (ECD) sequences listed in Table 1 (i.e., any one of SEQ ID NOs: 109-142, 239, 280-325, 364-381, 387-424, 427-433, 435-470) and contains an amino acid modification (e.g., substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises any of the IgV sequences listed in Table 1 (i.e., any one of SEQ ID NOs: 197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426, and 434).In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, such as at least 96% identity, 97% identity, 98% identity, or 99% identity to any of the IgV sequences listed in Table 1 (i.e., any one of SEQ ID NOs:197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426, and 434), and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific binding fragment of any of the IgV sequences listed in Table 1 (i.e., any one of SEQ ID NOs:197-199, 201-208, 210, 212, 240, 326-340, 382-386, 425-426, and 434) and contains an amino acid substitution that is not present in wild-type or unmodified ICOSL. Mutations designated with an "X" indicate that the designated position contains the Q or wild-type residue shown in the corresponding position in SEQ ID NO:32.

[0170] Table 1. Exemplary variant ICOSL polypeptides TIFF0007623967000025.tif218170TIFF0007623967000026.tif218170TIFF00076239670 00027.tif218170TIFF0007623967000028.tif218170TIFF0007623967000029.tif147170

[0171] In some embodiments, the variant ICOSL polypeptide exhibits improved affinity for the ectodomain of CD28 compared to a wild-type or unmodified ICOSL polypeptide comprising, for example, the sequence shown in SEQ ID NO: 32 or 196. In some embodiments, the ICOSL polypeptide exhibits improved affinity for the ectodomain of ICOS compared to a wild-type or unmodified ICOSL comprising, for example, the sequence shown in SEQ ID NO: 32 or 196. In some embodiments, the ICOSL polypeptide exhibits improved affinity for the ectodomain of CD28 and the ectodomain of ICOS compared to a wild-type or unmodified ICOSL comprising, for example, the sequence shown in SEQ ID NO: 32 or 196.

[0172] In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) corresponding to positions 52, 54, or 57. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) selected from N52H, N52D, N52Q, N52S, N52Y, N52K, S54A, S54P, or N57Y, or conservative amino acid modifications (e.g., substitutions) thereof. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) selected from N52H, N52D, N52S, N52K, or N57Y, or conservative amino acid modifications (e.g., substitutions) thereof.

[0173] In some embodiments, the variant ICOSL polypeptide can contain one or more additional amino acid modifications (e.g., substitutions) in addition to an amino acid modification (e.g., substitution) at a position corresponding to position 52, 54, or 57. In some embodiments, the one or more additional amino acid modifications (e.g., substitutions) are at positions 10, 11, 13, 16, 18, 20, 25, 27, 30, 37, 42, 43, 47, 52, 54, 57, 61, 62, 67, 71, 72, 74, 75, 77, 78, 80, 84, 89, 90, 92, 93, 94, 96, 97, 98, 99, 100, 102, 103, 107, 109, 110, 113, 115, 116, 117, 119, 120, In some embodiments, the variant ICOSL is at a position corresponding to: 121, 122, 126, 129, 130, 132, 133, 135, 138, 139, 140, 142, 143, 144, 146, 151, 152, 153, 154, 155, 156, 158, 161, 166, 168, 172, 173, 175, 190, 192, 193, 194, 198, 201, 203, 207, 208, 210, 212, 217, 218, 220, 221, 224, 225, or 227. TIFF0007623967000030.tif84164 or a conservative amino acid substitution thereof.

[0174] In some embodiments of any one of the above variant ICOSL polypeptides, the variant ICOSL polypeptide further comprises one or more amino acid deletions corresponding to position 140 of SEQ ID NO:32.

[0175] In some embodiments, the variant ICOSL polypeptide is TIFF0007623967000031.tif194166TIFF0007623967000032.tif48162.

[0176] In some embodiments, the variant ICOSL polypeptide is TIFF0007623967000033.tif63165.

[0177] In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) selected from N52H / N57Y / Q100R / F172S, N52H / Q100R, N52H / N57Y / Q100R / C198R. TIFF0007623967000034.tif12165.

[0178] In some embodiments, the variant ICOSL polypeptide is TIFF0007623967000035.tif19152.

[0179] In some embodiments, the variant ICOSL polypeptide exhibits improved binding affinity for binding to one of the ectodomains of CD28 or ICOS and decreased binding affinity for binding to the other of the ectodomains of CD28 or ICOS, compared to a wild-type or unmodified ICOSL polypeptide comprising, for example, the sequence set forth in SEQ ID NO:32 or 196.

[0180] In some embodiments, the variant ICOSL polypeptide exhibits improved binding affinity to ICOS and reduced binding affinity to CD28. In some embodiments, the one or more additional amino acid substitutions are at positions corresponding to 52, 57, 80, 100, 130, 152, 161, or 198. ... In some embodiments, the variant ICOSL polypeptide contains one or more amino acid substitutions selected from: TIFF0007623967000037.tif77135.

[0181] In some embodiments, the variant ICOSL polypeptide exhibits improved binding affinity to CD28 and decreased binding affinity to ICOS. In some embodiments, the one or more amino acid substitutions are at positions corresponding to 52, 75, or 203. In some embodiments, the variant ICOSL contains one or more amino acid substitutions selected from N52S, R75Q, L203F, or L203P. In some embodiments, the variant ICOSL polypeptide has the amino acid substitutions N52S / R75Q / L203P.

[0182] In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) in unmodified ICOSL or a specific binding fragment thereof corresponding to positions 16, 30, 42, 52, 57, 90, 100, 102, 110, 115, 120, 122, 138, 143, 152, 156, 172, 194, 198, 203, 221, or 224 based on the numbering of SEQ ID NO: 32. TIFF0007623967000038.tif19165. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) in unmodified ICOSL or a specific binding fragment thereof corresponding to positions 115, 172, or 198 based on the numbering of SEQ ID NO: 32. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) selected from H115R, F172S, or C198R. In some embodiments, the one or more amino acid modifications (e.g., substitutions) are selected from TIFF0007623967000039.tif121166. In some embodiments, the variant ICOSL polypeptide potentially exhibits enhanced protein solubility or enhanced protein expression ("solubility variants") compared to wild-type or unmodified ICOSL polypeptide.

[0183] In some embodiments, the variant ICOSL polypeptide comprises any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 435-470. In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, e.g., at least 96% identity, 97% identity, 98% identity, or 99% identity, to any of the extracellular domains (ECDs) set forth in SEQ ID NOs: 435-470, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific-binding fragment of any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 435-470, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL.

[0184] In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) in unmodified ICOSL or a specific binding fragment thereof corresponding to positions 52, 57, 100, 138, 198, or 203 based on the numbering of SEQ ID NO: 32. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) selected from N52H, N52Y, N57Y, Q100R, Q100P, F138L, C198R, or L203P. In some embodiments, the one or more amino acid modifications (e.g., substitutions) are selected from: The file is TIFF0007623967000040.tif12137.

[0185] In some embodiments, the variant ICOSL polypeptide comprises any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 427-433. In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, e.g., at least 96% identity, 97% identity, 98% identity, or 99% identity, to any of the extracellular domains (ECDs) set forth in SEQ ID NOs: 427-433, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific-binding fragment of any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 427-433, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises the IgV sequence set forth in SEQ ID NO: 434. In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, e.g., at least 96% identity, 97% identity, 98% identity, or 99% identity, to the IgV sequence set forth in SEQ ID NO: 434, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific-binding fragment of the IgV sequence set forth in SEQ ID NO: 434, and contains an amino acid substitution that is not present in wild-type or unmodified ICOSL.

[0186] In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) in unmodified ICOSL or a specific binding fragment thereof corresponding to positions 52, 84, 91, 119, 155, 168, 207 based on the numbering of SEQ ID NO: 32. In some embodiments, the variant ICOSL polypeptide has one or more amino acid modifications (e.g., substitutions) selected from A91S, N52H, N52Q, N84Q, N119Q, N155H, N155Q, N168Q, N207Q. In some embodiments, the one or more amino acid modifications (e.g., substitutions) are selected from: TIFF0007623967000041.tif85165. In some embodiments, the variant ICOSL polypeptide potentially exhibits reduced glycosylation compared to a wild-type or unmodified ICOSL polypeptide.

[0187] In some embodiments, the variant ICOSL polypeptide comprises any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 387-424, 427-433, 435-470. In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, e.g., at least 96% identity, 97% identity, 98% identity, or 99% identity to any of the extracellular domains (ECDs) set forth in SEQ ID NOs: 387-424, 427-433, 435-470, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific-binding fragment of any of the extracellular domain (ECD) sequences set forth in SEQ ID NOs: 387-424, 427-433, 435-470, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises any of the IgV sequences set forth in SEQ ID NOs: 425-426. In some embodiments, the variant ICOSL polypeptide comprises a polypeptide sequence that exhibits at least 90% identity, at least 91% identity, at least 92% identity, at least 93% identity, at least 94% identity, at least 95% identity, e.g., at least 96% identity, 97% identity, 98% identity, or 99% identity to any of the IgV sequences set forth in SEQ ID NOs: 425-426, and contains an amino acid modification (e.g., a substitution) that is not present in wild-type or unmodified ICOSL. In some embodiments, the variant ICOSL polypeptide comprises a specific binding fragment of any of the IgV sequences set forth in SEQ ID NOs:425-426 and contains amino acid substitutions that are not present in wild-type or unmodified ICOSL.

[0188] III. Variant Polypeptide Formats Immunomodulating polypeptides comprising variant ICOSL provided herein containing vIgD can be formatted in a variety of ways, for example, as soluble proteins, membrane-bound proteins, secreted proteins, conjugates or fusions, or for expression by infectious agents. In some embodiments, a particular format can be selected for a desired therapeutic application. In some cases, immunomodulating polypeptides comprising variant ICOSL polypeptides are provided in a format that antagonizes or blocks the activity of their cognate binding partner (e.g., CD28). In some embodiments, antagonism against CD28 can be useful for treating inflammation or autoimmunity. In some cases, immunomodulating polypeptides comprising variant ICOSL polypeptides are provided in a format that agonizes or stimulates the activity of their cognate binding partner (e.g., CD28). In some embodiments, agonism against CD28 can be useful for treating oncological indications. Those skilled in the art can easily determine the activity of a particular format, for example, to antagonize or agonize one or more specific cognate binding partners. Exemplary methods for assessing such activity are provided herein, including in the Examples.

[0189] In some aspects, immunomodulatory proteins comprising the vIgD of ICOSL are provided, which are soluble, e.g., fused to an Fc chain. In some aspects, one or more additional IgSF domains (e.g., one or more additional vIgDs) can be linked to the vIgD of ICOSL as provided herein (hereinafter referred to as "stacked" or "stacked" immunomodulatory proteins). In some embodiments, the modular format of the provided immunomodulatory proteins provides flexibility for modifying or generating immunomodulatory proteins to regulate the activity of multiple counter structures (multiple cognate binding partners). In some embodiments, such "stacked" molecules can be provided in a soluble format, and in some cases may be provided as membrane-bound or secreted proteins. In some embodiments, the variant ICOSL immunomodulatory protein is provided as a conjugate containing vIgD of ICOSL linked directly or indirectly to a targeting agent or moiety (e.g., an antibody or other binding molecule) that specifically binds to a ligand (e.g., an antigen), e.g., to target or localize vIgD to a particular environment or cell when administered to a subject. In some embodiments, the targeting agent, e.g., an antibody or other binding molecule, binds to a tumor antigen, thereby localizing the vIgD-containing variant ICOSL to the tumor microenvironment, e.g., modulating the activity of tumor-infiltrating lymphocytes (TILs) specific to the tumor microenvironment.

[0190] In some embodiments, the provided immunomodulatory proteins are expressed in cells and provided as part of engineered cell therapy (ECT). In some embodiments, the variant ICOSL polypeptide is expressed in a membrane-bound form in cells, such as immune cells (e.g., T cells or antigen-presenting cells), thereby providing a transmembrane immunomodulatory protein (hereinafter also referred to as "TIP"). In some embodiments, depending on the cognate binding partner recognized by TIP, engineered cells expressing TIP can activate the cognate binding partner by providing either a positive or negative costimulatory signal to other engineered cells and / or endogenous T cells. In some aspects, the variant ICOSL polypeptide is expressed in a secreted form in cells, such as immune cells (e.g., T cells or antigen-presenting cells), thereby producing a secreted or soluble form of the variant ICOSL polypeptide (hereinafter also referred to as "SIP"), for example, when the cells are administered to a subject. In some aspects, the SIP can antagonize its cognate binding partner in the environment in which it is secreted (e.g., the tumor microenvironment). In some embodiments, the variant ICOSL polypeptide is expressed in an infectious agent (e.g., a viral or bacterial agent) that is capable of infecting a cell, e.g., an immune cell (e.g., a T cell or an antigen-presenting cell) or tumor in vivo upon administration to a subject for delivery or expression of the variant polypeptide into the cell as a TIP or SIP.

[0191] In some embodiments, a soluble immunomodulatory polypeptide, such as a variant ICOSL containing vIgD, can be encapsulated in a liposome, which can itself be conjugated to any one or any combination of the provided conjugates (e.g., targeting moieties). In some embodiments, the soluble or membrane-bound immunomodulatory polypeptide of the present invention is deglycosylated. In more specific embodiments, the variant ICOSL sequence is deglycosylated. In even more specific embodiments, the IgV and / or IgC (e.g., IgC2) domains of the variant ICOSL are deglycosylated.

[0192] Non-limiting examples of the formats provided are set forth in Figures 13A-13C and further described below.

[0193] A. Soluble Proteins In some embodiments, the immunomodulatory protein containing the variant ICOSL polypeptide is a soluble protein. Those skilled in the art will recognize that cell surface proteins typically have an intracellular domain, a transmembrane domain, and an extracellular domain (ECD), and that the extracellular domain or its immunologically active sequence can be used to produce a soluble form of such a protein. Thus, in some embodiments, the immunomodulatory protein containing the variant ICOSL polypeptide lacks the transmembrane domain or a portion of the transmembrane domain. In some embodiments, the immunomodulatory protein containing the variant ICOSL polypeptide lacks the intracellular (cytoplasmic) domain or a portion of the intracellular domain. In some embodiments, the immunomodulatory protein containing the variant ICOSL polypeptide contains only the vIgD portion containing the ECD domain or a portion thereof containing the IgV domain and / or IgC (e.g., IgC2) domain or a specific binding fragment thereof containing an amino acid modification.

[0194] In some embodiments, the immunomodulatory polypeptide comprising a variant ICOSL can comprise one or more variant ICOSL polypeptides of the present invention. In some embodiments, the polypeptide of the present invention comprises exactly one, two, three, four, or five variant ICOSL sequences. In some embodiments, at least two of the variant ICOSL sequences are the same variant ICOSL sequence.

[0195] In some embodiments, the immunomodulatory polypeptides provided comprise two or more vIgD sequences of ICOSL. Multiple variant ICOSL polypeptides within a polypeptide chain can be identical (i.e., homologous) or non-identical (i.e., heterologous) variant ICOSL sequences. In addition to single polypeptide chain embodiments, in some embodiments, two, three, four, or more polypeptides of the present invention may be covalently or non-covalently linked to each other. Thus, monomeric, dimeric, and higher (e.g., 3-, 4-, 5-, or higher) multimeric proteins are provided herein. In some embodiments, for example, exactly two polypeptides of the present invention can be covalently or non-covalently linked to each other to form a dimer. In some embodiments, the linkage is via an interchain cysteine ​​disulfide bond. Compositions comprising two or more polypeptides of the present invention can be compositions of polypeptides of the same or substantially the same species (e.g., homodimers) or polypeptides of different species (e.g., heterodimers). A composition having multiple linked polypeptides of the invention may have one or more identical or non-identical variant ICOSL polypeptides of the invention in each polypeptide chain, as described above.

[0196] In some embodiments, the immunomodulatory protein comprises a variant ICOSL polypeptide linked to an immunoglobulin Fc (resulting in an "immunomodulatory Fc fusion," e.g., an "ICOSL-Fc variant fusion," also referred to as an ICOSL vIgD-Fc fusion). In some embodiments, the variant ICOSL polypeptide is linked at the N-terminus of the Fc. In some embodiments, the variant ICOSL polypeptide is linked at the C-terminus of the Fc. In some embodiments, two or more (the same or different) ICOSL variant polypeptides are independently linked to the N-terminus and C-terminus.

[0197] In some embodiments, the Fc is a murine Fc or a human Fc. In some embodiments, the Fc is a mammalian or human IgG1, IgG2, IgG3, or IgG4 Fc region. In some embodiments, the Fc is derived from an IgG1, e.g., a human IgG1. In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO:226, or an amino acid sequence exhibiting at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to SEQ ID NO:226.

[0198] In some embodiments, the Fc region contains additional modifications that alter (e.g., reduce) one or more of its normal functions. Generally, the Fc region is involved in effector functions, such as complement-dependent cytotoxicity (CDC) and antibody-dependent cellular cytotoxicity (ADCC), in addition to antigen binding, which is the primary function of immunoglobulins. Additionally, the FcRn sequence present in the Fc region plays a role in regulating serum IgG levels by extending in vivo half-life through conjugation to the in vivo FcRn receptor. In some embodiments, such functions can be reduced or modified in the Fc for use with the provided Fc fusion proteins.

[0199] In some embodiments, one or more amino acid modifications can be introduced into the Fc region of the ICOSL-Fc variant fusions provided herein, thereby generating an Fc region variant. In some embodiments, the Fc region variant has reduced effector function. There are many examples of changes or mutations in the Fc sequence that can alter effector function. For example, WO 00 / 42072, WO 2006019447, WO 2012125850, WO 2015 / 107026, US 2016 / 0017041, and Shields et al. J Biol. Chem. 9(2): 6591-6604 (2001) describe exemplary Fc variants with improved or reduced binding to FcR. The contents of these publications are specifically incorporated herein by reference.

[0200] In some embodiments, the provided variant ICOSL-Fc fusions comprise an Fc region that exhibits reduced effector function, making them desirable candidates for applications in which the in vivo half-life of the ICOSL-Fc variant fusion is important but certain effector functions (e.g., CDC and ADCC) are unnecessary or deleterious. In vitro and / or in vivo cytotoxicity assays can be performed to confirm reduced / depleted CDC and / or ADCC activity. For example, Fc receptor (FcR) binding assays can be performed to confirm that the ICOSL-Fc variant fusion lacks FcγR binding (and thus likely lacks ADCC activity) but retains FcRn binding ability. NK cells, the primary cells for mediating ADCC, express only FcγRIII, whereas monocytes express FcγRI, FcγRII, and FcγRIII. FcR expression on hematopoietic cells is summarized in Table 3 on page 464 of Ravetch and Kinet, Annu. Rev. Immunol. 9:457-492 (1991). Non-limiting examples of in vitro assays to assess ADCC activity of a molecule of interest are described in U.S. Pat. No. 5,500,362 (see, e.g., Hellstrom, I. et al. Proc. Nat'l Acad. Sci. USA 83:7059-7063 (1986)) and Hellstrom, I et al., Proc. Nat'l Acad. Sci. USA 82:1499-1502 (1985); U.S. Pat. No. 5,821,337 (see, Bruggemann, M. et al., J. Exp. Med. 166:1351-1361 (1987)). Alternatively, non-radioactive assay methods may be employed (see, e.g., ACTI™ Non-Radiotoxicity Assay for Flow Cytometry (CellTechnology, Inc. Mountain View, Calif.); and CytoTox 96™ Non-Radiotoxicity Assay (Promega, Madison, Wis.)). Useful effector cells for such assays include peripheral blood mononuclear cells (PBMC) and natural killer (NK) cells.Alternatively, or in addition, the ADCC activity of the molecule of interest may be assessed in vivo, for example, in an animal model such as that disclosed in Clynes et al. Proc. Nat'l Acad. Sci. USA 95:652-656 (1998). A C1q binding assay may also be performed to confirm that the ICOSL-Fc variant fusion is unable to bind to C1q and therefore lacks CDC activity. See, for example, the C1q and C3c binding ELISAs in WO 2006 / 029879 and WO 2005 / 100402. To assess complement activation, a CDC assay may be performed (see, e.g., Gazzano-Santoro et al., J. Immunol. Methods 202:163 (1996); Cragg, MS et al., Blood 101:1045-1052 (2003); and Cragg, MS and MJ Glennie, Blood 103:2738-2743 (2004)). FcRn binding and in vivo clearance / half-life determinations can also be performed using methods known in the art (see, e.g., Petkova, SB et al., Int'l. Immunol. 18(12):1759-1769 (2006)).

[0201] ICOSL-Fc variant fusions with reduced effector function include those with substitutions at one or more of residues 238, 265, 269, 270, 297, 327, and 329 in the Fc region according to EU numbering (U.S. Patent No. 6,737,056). Such Fc variants include Fc variants with substitutions at two or more of amino acid positions 265, 269, 270, 297, and 327 according to EU numbering, including the so-called "DANA" Fc variant in which residues 265 and 297 are substituted with alanine (U.S. Patent No. 7,332,581).

[0202] In some embodiments, the Fc region of the ICOSL-Fc variant fusion has an Fc region in which any one or more of the amino acids at positions 234, 235, 236, 237, 238, 239, 270, 297, 298, 325, and 329 (as indicated by EU numbering) have been substituted with a different amino acid compared to the native Fc region. Such Fc region modifications are not limited to the above modifications and may include, for example, deglycosylated chains as described in Current Opinion in Biotechnology (2009) 20 (6), 685-691. TIFF0007623967000042.tif33165; modifications such as G236R / L328R, L235G / G236R, N325A / L328R, and N325LL328R described in WO 2008 / 092117; amino acid insertions at positions 233, 234, 235, and 237 (denoted by EU numbering); and modifications at sites described in WO 2000 / 042072.

[0203] Certain Fc variants with improved or reduced binding to FcRs have been described (see, e.g., U.S. Patent No. 6,737,056; WO 2004 / 056312, WO 2006019447, and Shields et al., J. Biol. Chem. 9(2): 6591-6604 (2001)).

[0204] In some embodiments, ICOSL-Fc variant fusions are provided that include a variant Fc region comprising one or more amino acid substitutions that extend half-life and / or improve binding to the neonatal Fc receptor (FcRn). Antibodies with extended half-life and improved binding to FcRn are described in US2005 / 0014934A1 (Hinton et al.) or WO 2015107026. These antibodies include an Fc region having one or more substitutions therein that improve binding of the Fc region to FcRn. Such Fc variants include those having a substitution at one or more of the following Fc region residues according to EU numbering: 238, 256, 265, 272, 286, 303, 305, 307, 311, 312, 317, 340, 356, 360, 362, 376, 378, 380, 382, ​​413, 424, or 434, e.g., a substitution at residue 434 of the Fc region (U.S. Patent No. 7,371,826).

[0205] In some embodiments, the Fc region of the ICOSL-Fc variant fusion comprises one or more amino acid substitutions E356D and M358L. In some embodiments, the Fc region of the ICOSL-Fc variant fusion comprises one or more amino acid substitutions C220S, C226S, C229S. In some embodiments, the Fc region of the ICOSL variant fusion comprises one or more amino acid substitutions R292C and V302C. See also Duncan & Winter, Nature 322:738-40 (1988); U.S. Patent No. 5,648,260; U.S. Patent No. 5,624,821; and WO 94 / 29351 for other examples of Fc region variants.

[0206] In some embodiments, modifications are made in the Fc region that result in reduced C1q binding and / or complement-dependent cytotoxicity (CDC), e.g., as described in U.S. Pat. No. 6,194,551, WO 99 / 51642, and Idusogie et al., J. Immunol. 164: 4178-4184 (2000).

[0207] In some embodiments, an ICOSL-Fc variant fusion is provided comprising a variant Fc region comprising one or more amino acid modifications, wherein the variant Fc region is derived from an IgG1, e.g., human IgG1. In some embodiments, the variant Fc region is derived from the amino acid sequence set forth in SEQ ID NO:226. In some embodiments, the Fc contains at least one amino acid substitution that is N82G according to the numbering of SEQ ID NO:226 (corresponding to N297G according to EU numbering). In some embodiments, the Fc further contains at least one amino acid substitution that is R77C or V87C according to the numbering of SEQ ID NO:226 (corresponding to R292C or V302C according to EU numbering). In some embodiments, the variant Fc region further comprises a C5S (corresponding to C220S according to EU numbering) amino acid modification according to the numbering of SEQ ID NO:226. For example, in some embodiments, the variant Fc region comprises the following amino acid modifications: N82G and one or more of the amino acid modifications C5S, R77C, or V87C relative to SEQ ID NO:226.

[0208] In some embodiments, an ICOSL-Fc variant fusion is provided comprising a variant Fc region, wherein the variant Fc comprises an amino acid sequence set forth in any of SEQ ID NOs: 474, 476, 477, 478 or 507, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or higher sequence identity to any of SEQ ID NOs: 474, 476, 477, 478 or 507.

[0209] In some embodiments, the Fc is derived from IgG2, e.g., human IgG2. In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO:227, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or higher sequence identity to SEQ ID NO:227.

[0210] In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO: 505, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or higher sequence identity to SEQ ID NO: 505. In some embodiments, the IgG4 Fc is a stabilized Fc in which the CH3 domain of human IgG4 is replaced with the CH3 domain of human IgG1 and exhibits inhibited aggregate formation, an antibody in which the CH3 and CH2 domains of human IgG4 are replaced with the CH3 and CH2 domains of human IgG1, respectively, or an antibody in which arginine at position 409 of human IgG4 as defined by the EU index proposed by Kabat et al. is replaced with lysine and exhibits inhibited aggregate formation (see, e.g., U.S. Patent No. 8,911,726). In some embodiments, the Fc is an IgG4 containing the S228P mutation, which has been shown to prevent recombination between therapeutic antibodies and endogenous IgG4 by Fab arm exchange (see, e.g., Labrijin et al. (2009) Nat. Biotechnol., 27 (8) 767-71). In some embodiments, the Fc comprises the amino acid sequence set forth in SEQ ID NO:506, or an amino acid sequence exhibiting at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or higher sequence identity to SEQ ID NO:506.

[0211] In some embodiments, the variant ICOSL polypeptide is directly linked to the Fc sequence. In some embodiments, the variant ICOSL polypeptide is indirectly linked to the Fc sequence, for example, via a linker. In some embodiments, one or more "peptide linkers" link the variant ICOSL polypeptide and the Fc domain. In some embodiments, the peptide linker can be a single amino acid residue or longer. In some embodiments, the peptide linker has at least one amino acid residue but is no longer than 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 amino acid residue. In some embodiments, the linker is (in the single-letter amino acid code) GGGGS ("4GS") or a multimer of a 4GS linker, for example, two, three, four, or five repeats of the 4GS linker.

[0212] In some embodiments, the variant ICOSL-Fc fusion protein is a dimer formed by two variant ICOSL Fc polypeptides linked to the Fc domain. In some specific embodiments, ICOSL-Fc variant fusion polypeptides of the same or substantially the same species (allowing for three or fewer N- or C-terminal amino acid sequence differences) are dimerized to produce a homodimer. In some embodiments, the dimer is a homodimer in which the two variant ICOSL Fc polypeptides are the same. Alternatively, ICOSL-Fc variant fusion polypeptides of different species can be dimerized to produce a heterodimer. Thus, in some embodiments, the dimer is a heterodimer in which the two variant ICOSL Fc polypeptides are different.

[0213] Also provided is a nucleic acid molecule encoding a variant ICOSL-Fc fusion protein. In some embodiments, the nucleic acid molecule encoding the variant ICOSL-Fc fusion protein is inserted into an appropriate expression vector to produce the Fc fusion protein. The resulting variant ICOSL-Fc fusion protein can be expressed in a host cell transformed with the vector, where the interchain disulfide bonds formed between the Fc portions cause assembly between the Fc domains to produce a dimer, such as a bivalent variant ICOSL-Fc fusion protein.

[0214] The resulting Fc fusion protein can be easily purified by affinity chromatography on a Protein A or Protein G column. For heterodimer formation, additional purification steps may be required. For example, when cells are transformed with two nucleic acids encoding different variant ICOSL polypeptides, the variant ICOSL molecules bearing the Fc domain are also expressed as disulfide-linked homodimers, so heterodimer formation must be achieved biochemically. Therefore, homodimer formation can be reduced under conditions that favor the disruption of interchain disulfides but do not affect the interchain disulfides. In some cases, different variant ICOSL Fc monomers are mixed in equimolar amounts and oxidized to form a mixture of homodimers and heterodimers. The components of this mixture are separated by chromatographic techniques. Alternatively, the formation of this type of heterodimer can be biased by genetically modifying and expressing Fc fusion molecules containing variant ICOSL polypeptides using the knob-into-hole method described below.

[0215] B. Stacked molecules with additional IgSF domains In some embodiments, the immunomodulatory protein can contain any of the variant ICOSL polypeptides provided herein linked directly or indirectly to one or more other immunoglobulin superfamily (IgSF) domains ("stacked" immunomodulatory protein constructs, also referred to as "type II" immunomodulatory proteins). In some aspects, this can create unique multidomain immunomodulatory proteins that provide multitargeted modulation of the immune synapse by binding to two or more, e.g., three or more, cognate binding partners.

[0216] In some embodiments, the immunomodulatory protein comprises a combination ("non-wild-type combination") and / or arrangement ("non-wild-type arrangement" or "non-wild-type permutation") of a variant ICOSL domain with one or more other affinity-modified and / or affinity-unmodified IgSF domain sequences of another IgSF family member (e.g., a mammalian IgSF family member) that is not found in a wild-type IgSF family member. In some embodiments, the immunomodulatory protein contains two, three, four, five, or six immunoglobulin superfamily (IgSF) domains, wherein at least one of the IgSF domains is a variant ICOSL IgSF domain (vIgD of ICOSL) according to the description provided.

[0217] In some embodiments, the sequence of the additional IgSF domain can be a modified IgSF domain containing one or more amino acid modifications (e.g., substitutions) compared to a wild-type or unmodified IgSF domain. In some embodiments, the IgSF domain can be affinity-unmodified (e.g., wild-type) or affinity-modified. In some embodiments, the unmodified or wild-type IgSF domain can be from mouse, rat, cynomolgus monkey, or human origin, or a combination thereof. In some embodiments, the additional IgSF domain can be an IgSF domain of an IgSF family member shown in Table 2. In some embodiments, the additional IgSF domain can be an affinity-modified IgSF domain containing one or more amino acid modifications (e.g., substitutions) compared to an IgSF domain contained in an IgSF family member shown in Table 2.

[0218] In some embodiments, the additional IgSF domain is an affinity-modified or unmodified IgSF domain contained in an IgSF family member of a family selected from the following: the signal-regulatory receptor (SIRP) family, the triggering receptor-like expressed on myeloid cells (TREML) family, the carcinoembryonic antigen-related cell adhesion molecule (CEACAM) family, the sialic acid-binding Ig-like lectin (SIGLEC) family, the butyrophilin family, the B7 family, the CD28 family, the V-set and immunoglobulin domain-containing (VSIG) family, the V-set transmembrane domain (VSTM) family, the major histocompatibility complex (MHC) family, the signaling lymphocyte activation molecule (SLAM) family, the leukocyte immunoglobulin-like receptor (LIR), the nectin (Nec) family, the nectin-like (NECL) family, the poliovirus receptor-related (PVR) family, the natural cytotoxicity-inducing receptor (NCR) family, the T-cell immunoglobulin and mucin (TIM) family, or the killer cell immunoglobulin-like receptor (KIR) family. In some embodiments, the additional IgSF domains are independently selected from the group consisting of CD80 (B7-1), CD86 (B7-2), CD274 (PD-L1, B7-H1), PDCD1LG2 (PD-L2, CD273), ICOSLG (B7RP1, CD275, ICOSL, B7-H2), CD276 (B7-H3), VTCN1 (B7-H4), CD28, CTLA4, PDCD1 (PD-1), ICOS, BTLA (CD272), CD4, CD8A (CD8α), CD8B (CD8β), LAG3, HAVCR2 (TIM-3), CEACAM1, TIGIT, PVR (CD155), PVRL2 (CD112), CD226, CD2, CD160, CD200, CD200R1 (CD200R), and NC R3 (NKp30).

[0219] The first column of Table 2 provides the name and, in some cases, several possible aliases for that particular IgSF member. The second column provides the protein identifier from the UniProtKB database, a publicly accessible database via the Internet at uniprot.org, and in some cases, provides the GenBank number. The Universal Protein Resource (UniProt) is a comprehensive resource for protein sequence and annotation data. The UniProt database includes the UniProt Knowledgebase (UniProtKB). UniProt is a collaboration between the European Bioinformatics Institute (EMBL-EBI), the SIB Swiss Institute of Bioinformatics, and the Protein Information Resource (PIR), and is primarily supported by grants from the US National Institutes of Health (NIH). GenBank is the NIH's gene sequence database, a collection of all publicly available, annotated DNA sequences (Nucleic Acids Research, 2013 Jan;41(D1):D36-42). Column 3 provides the region in which the indicated IgSF domain is located. The region is identified as the range encompassing the residues that the domain defines. Column 3 also indicates the IgSF domain class of the identified IgSF region. Column 4 provides the region in which the indicated additional domains are located (signal peptide, S; extracellular domain, E; transmembrane domain, T; cytoplasmic domain, C). It should be understood that the domain descriptions may vary depending on the method used to identify or classify the domains and may be independently identified from different sources. The descriptions of residues corresponding to domains in Table 2 are exemplary only and may be several amino acids longer or shorter (e.g., 1, 2, 3, or 4).Column 5 indicates some of the IgSF members listed (ie, some of their cell surface cognate binding partners).

[0220] Table 2. IgSF members of the present disclosure TIFF0007623967000043.tif178167TIFF0007623967000044.tif192167TIFF00076239670 00045.tif199167TIFF0007623967000046.tif214167TIFF0007623967000047.tif136167

[0221] In some embodiments, the provided immunomodulatory proteins, in addition to containing a variant ICOSL polypeptide, also contain at least two, three, four, five, or six additional immunoglobulin superfamily (IgSF) domains, e.g., the IgD domain of an IgSF family member set forth in Table 2. In some embodiments, the provided immunomodulatory proteins contain at least one additional IgSF domain (e.g., a second IgSF domain), wherein the at least one additional or second IgSF domain is a wild-type or unmodified IgSF domain contained in the amino acid sequence set forth in any of SEQ ID NOs: 1-27 and 341, or a specific-binding fragment thereof. In some embodiments, the wild-type or unmodified IgSF domain is an IgV domain or an IgC domain, e.g., an IgC1 or IgC2 domain.

[0222] In some embodiments, the provided immunomodulatory proteins, in addition to containing a variant ICOSL polypeptide, also contain at least one additional IgSF domain (e.g., a second IgSF domain) that is vIgD that contains one or more amino acid modifications (e.g., substitutions, deletions, or mutations) compared to an IgSF domain within a wild-type or unmodified IgSF domain, such as an IgSF domain within an IgSF family member set forth in Table 2. In some embodiments, the additional or second affinity-modified IgSF domain comprises at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or more sequence identity to the wild-type or unmodified IgSF domain, or a specific-binding fragment thereof, contained in the amino acid sequence set forth in any of SEQ ID NOs: 1-27 and 341. In some embodiments, the wild-type or unmodified IgSF domain is an IgV domain or an IgC domain, e.g., an IgC1 or IgC2 domain. In some embodiments, the additional or second IgSF domain is an affinity-modified IgV domain or an IgC domain.

[0223] In some embodiments, the provided immunomodulatory proteins contain at least one additional or second IgSF domain that is vIgD that contains one or more amino acid substitutions compared to a wild-type or unmodified IgSF domain other than ICOSL (e.g., IgV).

[0224] In some embodiments, the additional or second IgSF domain contains one or more amino acid substitutions compared to an IgSF domain within a wild-type or unmodified IgSF domain, e.g., an IgSF domain within an IgSF family member set forth in Table 2. In some embodiments, the additional or second affinity-modified IgSF domain comprises at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or more sequence identity to a wild-type or unmodified IgSF domain, or a specific-binding fragment thereof, contained in the amino acid sequence set forth in any of SEQ ID NOs: 1-27. In some embodiments, the wild-type or unmodified IgSF domain is an IgV domain or an IgC domain, e.g., an IgC1 or IgC2 domain. In some embodiments, the additional or second IgSF domain is an affinity-modified IgV domain or an IgC domain. Tables 3-5 provide exemplary polypeptides containing one or more affinity-engineered IgSF domains that can be used in the stack constructs provided herein.

[0225] In some embodiments, one or more additional IgSF domains (e.g., a second IgSF domain) are IgSF domains of another IgSF family member (e.g., IgV) that bind to or recognize a tumor antigen. In such embodiments, the IgSF family member serves as a tumor-localizing moiety, thereby bringing the vIgD of ICOSL into close proximity with immune cells within the tumor microenvironment. In some embodiments, the additional IgSF domain (e.g., a second IgSF domain) is an IgSF domain of NkP30 that binds to or recognizes B7-H6 expressed on tumor cells. In some embodiments, at least one additional (e.g., a second) IgSF domain (e.g., NkP30) is a vIgD containing one or more amino acid modifications (e.g., substitutions, deletions, or additions). In some embodiments, the one or more amino acid modifications improve binding affinity and / or selectivity for B7-H6, e.g., by at least or at least about 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 30-fold, 40-fold, or 50-fold, compared to an unmodified IgSF domain (e.g., NkP30).

[0226] Table 3: Exemplary variant CD80 polypeptides TIFF0007623967000048.tif152168TIFF0007623967000049.tif220168TIFF0007623967000050.tif220168TIFF0007623967000051.tif83168

[0227] Table 4: Exemplary variant NKp30 polypeptides TIFF0007623967000052.tif76170

[0228] Table 5. Exemplary variant CD86 polypeptides TIFF0007623967000053.tif61170

[0229] The number of such affinity-unmodified or affinity-modified IgSF domains (whether in non-wild-type combinations or non-wild-type arrangements) present in a "stacked" immunomodulatory protein construct is at least two, three, four, or five, and in some embodiments exactly two, three, four, or five IgSF domains (whereby the determination of the number of affinity-modified IgSF domains ignores any non-specific binding split sequences thereof and / or split sequences thereof that are substantially immunologically inactive).

[0230] In some embodiments of the stacked immunomodulatory proteins provided herein, the number of IgSF domains is at least two, and wherein the number of affinity-modified IgSF domains and the number of affinity-unmodified IgSF domains are each independently at least 0, 1, 2, 3, 4, 5, or 6. Thus, the number of affinity-modified IgSF domains and the number of affinity-unmodified IgSF domains, respectively (affinity-modified IgSF domains:affinity-unmodified IgSF domains), can be exactly or at least 2:0 (affinity-modified:wild-type), 0:2, 2:1, 1:2, 2:2, 2:3, 3:2, 2:4, 4:2, 1:1, 1:3, 3:1, 1:4, 4:1, 1:5, or 5:1.

[0231] In some embodiments of the stacked immunomodulatory proteins, at least two of the non-affinity-modified and / or affinity-modified IgSF domains are identical IgSF domains.

[0232] In some embodiments, the stacked immunomodulatory proteins provided herein comprise at least two affinity-modified and / or affinity-unmodified IgSF domains derived from a single IgSF member but in a non-wild-type arrangement (or "permutation"). One example of a non-wild-type arrangement or permutation is an immunomodulatory protein of the invention comprising affinity-modified and / or affinity-unmodified IgSF domain sequences of a non-wild-type series compared to those found in wild-type ICOSL, which IgSF domain sequence serves as the source of the variant IgSF domain as provided herein. Thus, in one example, the immunomodulatory protein can comprise an IgV proximal to the transmembrane domain and an IgC distal to the transmembrane domain, regardless of whether the affinity-unmodified and / or affinity-modified form is present. It is also within the scope of the provided subject matter that both non-wild-type combinations and non-wild-type arrangements of affinity-unmodified and / or affinity-modified IgSF domains are present in the immunomodulatory proteins provided herein.

[0233] In some embodiments of stacked immunomodulatory proteins, the affinity-unmodified and / or affinity-modified IgSF domains are non-identical (i.e., different) IgSF domains. Non-identical affinity-modified IgSF domains specifically bind to different cognate binding partners under specific binding conditions and are "non-identical" regardless of whether the wild-type or unmodified IgSF domains to which they are modified are the same. Thus, for example, a non-wild-type combination of at least two non-identical IgSF domains in an immunomodulatory protein can include at least one IgSF domain sequence that is unique and originates from one ICOSL family member, and at least one second IgSF domain sequence that is unique and originates from another IgSF family member that is not ICOSL, where the IgSF domains of the immunomodulatory protein are in their affinity-unmodified and / or affinity-modified forms. However, in alternative embodiments, the two non-identical IgSF domains originate from the same IgSF domain sequence, but at least one has been affinity-modified so that it specifically binds to a different cognate binding partner.

[0234] The multiple affinity-unmodified and / or affinity-modified IgSF domains within a stacked immunomodulatory protein polypeptide chain need not be directly covalently linked to one another. In some embodiments, an intervening stretch of one or more amino acid residues indirectly covalently links the affinity-unmodified and / or affinity-modified IgSF domains to one another. Linkages can be via N- to C-terminal residues.

[0235] In some embodiments, the linkage can be via the side chain of an amino acid residue that is not located at the N-terminus or C-terminus of the non-affinity-modified and / or affinity-modified IgSF domain. Thus, the linkage can be via a terminal or internal amino acid residue or a combination thereof.

[0236] In some embodiments, two or more IgSF domains, including the vIgD of ICOSL and one or more additional IgSF domains (e.g., a second variant IgSF domain) from another IgSF family member, are linked by a covalent or non-covalent bond. In some embodiments, the two or more IgSF domains are linked directly or indirectly, for example, via a linker. In some embodiments, an intervening range of one or more amino acid residues indirectly covalently links the IgSF domains to each other. Linkage can be via residues from the N-terminus to the C-terminus. In some embodiments, linkage can be via the side chain of an amino acid residue that is not located at the N-terminus or C-terminus of the IgSF domain. Thus, linkage can be via terminal or internal amino acid residues, or a combination thereof.

[0237] In some embodiments, one or more "peptide linkers" connect the vIgD of ICOSL to an additional IgSF domain (e.g., a second variant IgSF domain). In some embodiments, the peptide linker can be a single amino acid residue or longer. In some embodiments, the peptide linker has at least one amino acid residue but is no longer than 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 amino acid residue. In some embodiments, the linker is (in the single-letter amino acid code): GGGGS ("4GS") or a multimer of 4GS linkers, e.g., 2, 3, 4, or 5 repeats of the 4GS linker. In some embodiments, the peptide linker is (GGGGS)2 or (GGGGS)3. In some embodiments, the linker can also include a series of alanine residues, alone or in addition to another peptide linker (e.g., a 4GS linker or multimer thereof). In some embodiments, the number of alanine residues in each series is 2, 3, 4, 5, or 6 alanines.

[0238] In some embodiments, the affinity-unmodified and / or affinity-modified IgSF domains are linked by a "wild-type peptide linker" inserted at the N-terminus and / or C-terminus of the first and / or second affinity-unmodified and / or affinity-modified IgSF domain. In some embodiments, there is a leading peptide linker inserted at the N-terminus of the first IgSF domain and / or a first trailing sequence inserted at the C-terminus of the first affinity-unmodified and / or affinity-modified IgSF domain. In some embodiments, there is a second leading peptide linker inserted at the N-terminus of the second IgSF domain and / or a second trailing sequence inserted at the C-terminus of the second affinity-unmodified and / or affinity-modified IgSF domain. When the first and second affinity-unmodified and / or affinity-modified IgSF domains are derived from the same parent protein and are connected in the same orientation, the wild-type peptide linkers between the first and second affinity-unmodified and / or affinity-modified IgSF domains do not overlap. For example, when the first trailing wild-type peptide linker and the second leading wild-type peptide linker are the same, the type II immunomodulatory protein does not include either the first trailing wild-type peptide linker or the second leading wild-type peptide linker.

[0239] In some embodiments, the type II immunomodulatory protein comprises a first leading wild-type peptide linker inserted at the N-terminus of a first affinity-unmodified and / or affinity-modified IgSF domain, wherein the first leading wild-type peptide linker comprises at least five (e.g., at least about any of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more) contiguous amino acids from an intervening sequence in the wild-type protein from which the first affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately preceding domain (e.g., signal peptide or IgSF domain). In some embodiments, the first leading wild-type peptide linker comprises the entire intervening sequence in the wild-type protein from which the first affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately preceding domain (e.g., signal peptide or IgSF domain).

[0240] In some embodiments, the type II immunomodulatory protein further comprises a first trailing wild-type peptide linker inserted at the C-terminus of the first affinity-unmodified and / or affinity-modified IgSF domain, wherein the first trailing wild-type peptide linker comprises at least five (e.g., at least about any of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more) contiguous amino acids from an intervening sequence in the wild-type protein from which the first affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately following domain (e.g., the IgSF domain or the transmembrane domain). In some embodiments, the first trailing wild-type peptide linker comprises the entire intervening sequence in the wild-type protein from which the first affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately following domain (e.g., the IgSF domain or the transmembrane domain).

[0241] In some embodiments, the type II immunomodulatory protein further comprises a second leading wild-type peptide linker inserted at the N-terminus of the second affinity-unmodified and / or affinity-modified IgSF domain, wherein the second leading wild-type peptide linker comprises at least five (e.g., at least about any of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more) contiguous amino acids from an intervening sequence in the wild-type protein from which the second affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately preceding domain (e.g., signal peptide or IgSF domain). In some embodiments, the second leading wild-type peptide linker comprises the entire intervening sequence in the wild-type protein from which the second affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately preceding domain (e.g., signal peptide or IgSF domain).

[0242] In some embodiments, the type II immunomodulatory protein further comprises a second trailing wild-type peptide linker inserted at the C-terminus of the second affinity-unmodified and / or affinity-modified IgSF domain, wherein the second trailing wild-type peptide linker comprises at least five (e.g., at least about any of 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, or more) contiguous amino acids from an intervening sequence in the wild-type protein from which the second affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately following domain (e.g., IgSF domain or transmembrane domain). In some embodiments, the second trailing wild-type peptide linker comprises the entire intervening sequence in the wild-type protein from which the second affinity-unmodified and / or affinity-modified IgSF domain is derived, between the parent IgSF domain and the immediately following domain (e.g., IgSF domain or transmembrane domain).

[0243] Exemplary leading and trailing sequences for a type II protein containing a CD80 IgSF domain are set forth in SEQ ID NO:231 and SEQ ID NO:232. Exemplary leading and trailing sequences for a type II protein containing an ICOSL IgSF domain are set forth in SEQ ID NO:233 and 234. Exemplary leading and trailing sequences for a type II protein containing a CD86 IgSF domain are set forth in any of SEQ ID NO:236-238. An exemplary wild-type linker sequence for a type II protein containing an NKp30 IgSF domain is set forth in SEQ ID NO:235.

[0244] In some embodiments, two or more IgSF domains, including the vIgD of ICOSL and one or more additional IgSF domains (e.g., a second variant IgSF domain) from another IgSF family member, are linked or bound to Fc to form a dimeric multidomain stack immunomodulatory protein. In some embodiments, the variant ICOSL polypeptide and the second IgSF domain are independently linked, directly or indirectly, to the N- or C-terminus of the Fc subunit. In some embodiments, the variant ICOSL polypeptide and the second IgSF domain are linked, directly or indirectly, and one of the variant ICOSL or the second IgSF domain is also linked, directly or indirectly, to the N- or C-terminus of the Fc subunit. In some embodiments, the link to the Fc is via a peptide linker (e.g., a peptide linker as described above). In some embodiments, the link between the variant ICOSL and the second IgSF domain is via a peptide linker (e.g., a peptide linker as described above). In some embodiments, the vIgD, one or more additional IgSF domains, and the Fc domain of ICOSL can be linked together in any of a number of configurations as shown in Figure 16. Exemplary configurations are described in the Examples.

[0245] In some embodiments, the stacked immunomodulatory protein is a dimer formed by two stacked immunomodulatory Fc fusion polypeptides. Also provided are nucleic acid molecules encoding any of the stacked immunomodulatory proteins. In some embodiments, the dimeric multidomain stacked immunomodulatory protein can be produced in cells by expression, or in some cases, co-expression, of the stacked immunomodulatory Fc fusion polypeptides described above, following the production of the dimeric Fc fusion protein.

[0246] In some embodiments, the dimeric multidomain stack immunomodulatory protein is bivalent for each Fc subunit, monovalent for each subunit, or bivalent for one subunit and tetravalent for the other.

[0247] In some embodiments, the dimeric multidomain stacked immunomodulatory protein is a homodimeric multidomain stacked Fc protein. In some embodiments, the dimeric multidomain stacked immunomodulatory protein comprises a first stacked immunomodulatory Fc fusion polypeptide and a second stacked immunomodulatory Fc fusion polypeptide, wherein the first and second polypeptides are the same. In some embodiments, the Fc portion of the polypeptide can be any Fc, as described above.

[0248] In some embodiments, the multidomain stack molecule is a heterodimer comprising two different Fc polypeptides, at least one of which is an Fc polypeptide containing at least one variant ICOSL polypeptide and / or at least one second IgSF domain (e.g., a second variant IgSF domain). In some embodiments, the multidomain stack molecule contains a first Fc polypeptide containing a variant ICOSL and a second IgSF domain and a second Fc polypeptide containing a variant ICOSL and a second IgSF domain. In some embodiments, the multidomain stack molecule contains a first Fc polypeptide containing a variant ICOSL polypeptide and a second IgSF domain and a second Fc polypeptide that is not linked to either the variant ICOSL polypeptide or the second IgSF domain.

[0249] In some embodiments, the multidomain stack molecule contains a first Fc polypeptide containing one, two, three, four, or more variant ICOSL polypeptides and / or one, two, three, four, or more second IgSF domains, such that the total number of IgSF domains in the first stacked Fc polypeptide is two, three, four, five, six, or more. In one example of such an embodiment, the second stacked Fc polypeptide contains one, two, three, four, or more variant ICOSL polypeptides and / or one, two, three, four, or more second IgSF domains, such that the total number of IgSF domains in the second stacked Fc polypeptide is two, three, four, five, six, or more. In another example of such an embodiment, the second Fc polypeptide is not linked to any of the variant ICOSL polypeptides or second IgSF domains.

[0250] In some embodiments, the heterodimeric stack molecule contains a first stack immunomodulatory Fc fusion polypeptide and a second stack immunomodulatory Fc fusion polypeptide, wherein the first and second polypeptides are different. In some embodiments, the heterodimeric stack molecule contains a first Fc subunit containing a first variant ICOSL polypeptide and / or a second IgSF domain (e.g., a second variant IgSF domain), and a second Fc subunit containing the other of the first variant ICOSL polypeptide or the second IgSF domain. In some embodiments, the heterodimeric stack molecule contains a first stack immunomodulatory Fc fusion polypeptide and a second stack immunomodulatory Fc fusion polypeptide, wherein the first and second polypeptides are different. In some embodiments, the heterodimeric stack molecule contains a first Fc subunit containing a first variant ICOSL polypeptide and / or a second IgSF domain (e.g., a second variant IgSF domain), and a second Fc subunit containing both the first variant ICOSL polypeptide and a second IgSF domain (e.g., a second variant IgSF domain) but in a different orientation or configuration than the first Fc subunit.

[0251] In some embodiments, the Fc domain of one or both of the first and second stacked immunomodulatory Fc fusion polypeptides comprises a modification (e.g., a substitution) such that the interface of the Fc molecules is altered to facilitate and / or promote heterodimerization. In some embodiments, the modification comprises the introduction of a protrusion (knob) into the first Fc polypeptide and a cavity (hole) into the second Fc polypeptide, thereby facilitating complexation of the first and second Fc-containing polypeptides by allowing the protrusion to be positioned within the cavity. The amino acids targeted for replacement and / or modification to create the protrusion or cavity within the polypeptide are typically interface amino acids that interact or contact one or more amino acids at the interface of the second polypeptide.

[0252] In some embodiments, the first polypeptide modified to contain a protrusion (hole) amino acid comprises a replacement of a natural or original amino acid with an amino acid having at least one side chain that protrudes from the interface of the first polypeptide and can therefore be positioned in a compensatory cavity (hole) in the adjacent interface of the second polypeptide. In most cases, the replacement amino acid has a larger side chain volume than the original amino acid residue. Those skilled in the art will know how to determine and / or evaluate the properties of an amino acid residue to identify an ideal replacement amino acid to create the protrusion. In some embodiments, the replacement residue for forming the protrusion is a naturally occurring amino acid residue, such as arginine (R), phenylalanine (F), tyrosine (Y), or tryptophan (W). In some examples, the original residue identified for replacement is an amino acid residue with a small side chain, such as alanine, asparagine, aspartic acid, glycine, serine, threonine, or valine.

[0253] In some embodiments, the second polypeptide to be modified to contain a cavity (hole) comprises a natural or original amino acid replaced with an amino acid having at least one side chain that is recessed from the interface of the second polypeptide and can accommodate a corresponding protrusion from the interface of the first polypeptide. In most cases, the replacement amino acid has a smaller side chain volume than the original amino acid residue. Those skilled in the art know how to determine and / or evaluate the characteristics of amino acid residues to identify ideal replacement residues for cavity formation. Generally, replacement residues for cavity formation are naturally occurring amino acids, including, for example, alanine (A), serine (S), threonine (T) and valine (V). In some examples, the original amino acid identified for replacement is an amino acid with a large side chain, such as tyrosine, arginine, phenylalanine, or tryptophan.

[0254] The CH3 interface of human IgG1, for example, involves 16 residues on each domain located on four antiparallel β-strands, buried 1090 Å from each surface (see, e.g., Deisenhofer et al. (1981) Biochemistry, 20:2361-2370; Miller et al., (1990) J Mol. Biol., 216, 965-973; Ridgway et al., (1996) Prot. Engin., 9: 617-621; U.S. Pat. No. 5,731,168). The modification of CH3 domain to create protrusion or cavity is described in, for example, U.S. Patent No. 5,731,168; International Patent Application WO 98 / 50431 and WO 2005 / 063816; and Ridgway et al., (1996) Prot. Engin., 9: 617-621.In some examples, the modification of CH3 domain to create protrusion or cavity is typically targeted to the residue located on the two central antiparallel β-strands.The purpose is to minimize the risk that the protrusion created may be accommodated by protruding into the surrounding solvent, rather than being accommodated by the compensatory cavity in the partner CH3 domain.

[0255] In some embodiments, the heterodimeric molecule contains a T366W mutation in the CH3 domain of the "knob chain" and a T366S, L368A, or Y407V mutation in the CH3 domain of the "hole chain." In some cases, an additional interchain disulfide bridge between the CH3 domains (Merchant, AM, et al., Nature Biotech. 16 (1998) 677-681) can also be used, for example, by introducing a Y349C mutation in the CH3 domain of the "knob" or "hole" chain and an E356C or S354C mutation in the CH3 domain of the other chain. In some embodiments, the heterodimeric molecule contains an S354C or T366W mutation in one of the two CH3 domains and a Y349C, T366S, L368A, or Y407V mutation in the other of the two CH3 domains. In some embodiments, the heterodimeric molecule comprises an E356C, T366W mutation in one of the two CH3 domains and an Y349C, T366S, L368A, Y407V mutation in the other of the two CH3 domains. In some embodiments, the heterodimeric molecule comprises a Y349C, T366W mutation in one of the two CH3 domains and an E356C, T366S, L368A, Y407V mutation in the other of the two CH3 domains. In some embodiments, the heterodimeric molecule comprises a Y349C, T366W mutation in one of the two CH3 domains and an S354C, T366S, L368A, Y407V mutation in the other of the two CH3 domains. Other examples of knob-in-hole technology are known in the art, for example, as described in EP 1 870 459 A1.

[0256] In some embodiments, the Fc subunit of the heterodimeric molecule can additionally contain one or more other Fc mutations, such as any of those described above, hi some embodiments, the heterodimeric molecule contains an Fc subunit with a mutation that reduces effector function.

[0257] In some embodiments, an Fc variant containing a CH3 protrusion / cavity modification can be connected anywhere in the stacked immunomodulatory polypeptide, but typically via its N- or C-terminus, to the N- or C-terminus of a first and / or second stacked immunomodulatory polypeptide, for example, to form a fusion polypeptide. The linkage can be direct or indirect via a linker. Typically, the knob and hole molecules are produced by coexpression of a first stacked immunomodulatory polypeptide linked to an Fc variant containing a CH3 protrusion modification and a second stacked immunomodulatory polypeptide linked to an Fc variant containing a CH3 cavity modification.

[0258] C. Conjugates and Fusions of Variant Polypeptides and Immunomodulatory Proteins In some embodiments, the variant polypeptides provided herein, which are immunomodulatory proteins comprising a variant of the IgSF family Ig domain (vIgD), can be conjugated or fused, directly or indirectly, to a moiety, e.g., an effector moiety, e.g., another protein, to form a conjugate ("IgSF conjugate"). In some embodiments, the binding can be covalent or non-covalent (e.g., via a biotin-streptavidin non-covalent interaction). In some embodiments of an ICOSL-Fc variant fusion, any one or combination of any two or more of the aforementioned conjugates can be attached to the Fc or variant CD80 polypeptide, or both.

[0259] In some embodiments, the moiety can be a targeting moiety, a small molecule drug (a non-polypeptide drug of less than 500 Dalton molar mass), a toxin, a cytostatic agent, a cytotoxic agent, an immunosuppressant, a radioactive substance suitable for diagnostic purposes, a radioactive metal ion for therapeutic purposes, a prodrug-activating enzyme, an agent that extends biological half-life, or a diagnostic or detectable agent.

[0260] In some embodiments, the effector moiety is a therapeutic agent that provides cytotoxicity, cytostaticity, or otherwise some therapeutic benefit, e.g., a cancer therapeutic agent. In some embodiments, the effector moiety is a targeting moiety or agent, e.g., an agent that targets a cell surface antigen, e.g., an antigen on the surface of a tumor cell. In some embodiments, the effector moiety is a label that can generate a detectable signal either directly or indirectly. In some embodiments, the effector moiety is a toxin. In some embodiments, the effector moiety is a protein, peptide, nucleic acid, small molecule, or nanoparticle.

[0261] In some embodiments, one, two, three, four, five, or more effector moieties, which may be the same or different, are conjugated, linked, or fused to a variant polypeptide or protein to form an IgSF conjugate. In some embodiments, such effector moieties can be attached to the variant polypeptide or immunomodulatory protein using various molecular biological or chemical conjugation and linking methods known in the art and described below. In some embodiments, the effector moiety can be linked or conjugated to the variant polypeptide or immunomodulatory protein using a linker, such as a peptide linker, a cleavable linker, a non-cleavable linker, or a linker that facilitates the conjugation reaction.

[0262] In some embodiments, the IgSF conjugate comprises the following components: (protein or polypeptide), (L) q and (effector part) mwherein the protein or polypeptide is any of the described variant polypeptides or immunomodulatory proteins capable of binding to one or more cognate counterstructure ligands as described; L is a linker for joining the protein or polypeptide to the moiety; m is at least 1; q is 0 or greater; and the resulting IgSF conjugate binds to one or more counterstructure ligands. In certain embodiments, m is 1 to 4 and q is 0 to 8.

[0263] In some embodiments, an IgSF conjugate is provided that includes a variant polypeptide or immunomodulatory protein provided herein conjugated with a targeting agent that binds to a cell surface molecule, e.g., for targeted delivery of the variant polypeptide or immunomodulatory protein to a specific cell. In some embodiments, the targeting agent is a molecule that has the ability to localize and bind to a molecule present on normal cells / tissues and / or tumor cells / tumors in a subject. In other words, the IgSF conjugate including the targeting agent can bind (directly or indirectly) to a ligand present on cells, e.g., tumor cells. Targeting agents of the present invention that are contemplated for use include antibodies, polypeptides, peptides, aptamers, other ligands, or any combination thereof, that can bind to a component of a target cell or molecule.

[0264] In some embodiments, the targeting agent can bind to tumor cells or to the vicinity of tumor cells (e.g., tumor vasculature or tumor microenvironment) after administration to a subject. The targeting agent can bind to a receptor or ligand on the surface of cancer cells. In another aspect of the present invention, a targeting agent is selected that is specific to non-cancerous cells or tissues. For example, the targeting agent can be specific to a molecule normally present on a particular cell or tissue. Furthermore, in some embodiments, the same molecule can be present on normal cells and cancer cells. Various cellular components and molecules are known. For example, if the targeting agent is specific to EGFR, the resulting IgSF conjugate can target both EGFR-expressing cancer cells and EGFR-expressing normal skin epithelial cells. Therefore, in some embodiments, the IgSF conjugates of the present invention can act by two separate mechanisms (targeting cancer cells and non-cancerous cells).

[0265] In various aspects of the invention disclosed herein, the IgSF conjugates of the invention comprise a targeting agent capable of binding to / targeting a cellular component, such as a tumor antigen, a bacterial antigen, a viral antigen, a mycoplasmal antigen, a fungal antigen, a prion antigen, or an antigen from a parasite. In some aspects, the cellular component, antigen, or molecule can each be used to refer to the desired target of the targeting agent. For example, in various embodiments, the targeting agent is selected from the group consisting of epidermal growth factor receptor (EGFR, ErbB-1, HER1), ErbB-2 (HER2 / neu), ErbB-3 / HER3, ErbB-4 / HER4, EGFR ligand family; insulin-like growth factor receptor (IGFR) family, IGF-binding proteins (IGFBPs), IGFR ligand family; platelet-derived growth factor receptor (PDGFR) family, PDGFR ligand family; fibroblast growth factor receptor (FGFR) family, FGFR ligand family, vascular endothelial growth factor receptor (VEGF) family, and VEGF-binding proteins. long-acting factor receptor (VEGFR) family, VEGF family; HGF receptor family; TRK receptor family; ephrin (EPH) receptor family; AXL receptor family; leukocyte tyrosine kinase (LTK) receptor family; TIE receptor family, angiopoietin 1, 2; receptor tyrosine kinase-like orphan receptor (ROR) receptor family, e.g., ROR1; CD171 (L1CAM); B7-H6 (NCR3LG1); PD-L1, tumor glycosylated antigens, e.g., sTn or Tn, e.g., sTn of MUC1Ag; LHR (LHCGR); phosphatidylserine, discoidin domain receptor (DDR) family; RET receptor family; KLG receptor family; RYK receptor family; MuSK receptor family; transforming growth factor-α (TGF-α) receptor, TGF-β; cytokine receptors, class I (hematopoietin family) and class II (interferon / IL-10 family) receptors, tumor necrosis factor (TNF) receptor superfamily (TNFRSF), death receptor family; cancer-testis (CT) antigen, lineage-specific antigen, differentiation antigen, alpha-actinin-4, ARTCl, breakpoint cluster region-Abelson (Bcr-abl) fusion product, B-RAF, caspase-5 (CASP-5), caspase-8 (CASP-8), β-catenin (CTNNBl), cell division cycle 27 (CDC27), cyclin-dependent kinase 4 (CDK4), CDKN2A, COA-I, dek-can fusion protein, EFTUD-2, elongation factor 2 (ELF2), Ets variant gene 6 / acute myeloid leukemia 1 gene ETS (ETC6-AML1) fusion protein, fibronectin (FN), e.g., fibronectin ectodomain A (EDA), GPNMB, low-density lipid receptor / GDP-L-fucose:β-D-galactose 2-α-L-fucosyltransferase (LDLR / FUT) fusion protein, HLA-A2 arginine-isoleucine exchange at residue 170 of the α-helix of the α2 domain in the HLA-A2 gene (HLA-A*201-R170I), HLA-A*201-R170I, ...1, heat shock protein 70-2 mutant (HSP70-2M), K1AA0205, MART2, melanoma ubiquitous mutant 1, 2, 3 (MUM-I, 2, 3), prostatic acid phosphatase (PAP), neo-PAP, myosin class I, NFYC, OGT, OS-9, pml-RAR alpha fusion protein, PRDX5, PTPRK, K-ras (KRAS2), N-ras (NRAS), HRAS, RBAF600, SIRT2, SNRPDl, SYT-SSXl or -SSX2 fusion protein, triosephosphate isoform Melase, BAGE, BAGK-1, BAGE-2,3,4,5, GAGE-1,2,3,4,5,6,7,8, GnT-V (abnormal N-acetylglucosaminyltransferase V, MGAT5), HERV-K-MEL, KK-LC, KM-HN-I, LAGE, LAGE-I, CTL-recognized antigen on melanoma (CAMEL), MAGE-Al (MAGE-I), MAGE-A2, MAGE-A3, MAGE-A4, MAGE-A5, MAGE-A6, MAGE-A8, MAGE-A9, MAGE-AlO, MAGE-AIl, MAGE-A12, MAGE-3, MAGE-Bl, MAGE-B2, MAGE-B5, MAGE-B6, MAGE-Cl, MAGE-C2, Mucin 1 (MUCl), MART-1 / Melan-A (M LANA), gplOO, gplOO / Pmell7(SILV), tyrosinase (TYR), TRP-I, HAGE, NA-88, NY-ESO-I, NY-ESO-l / LAGE-2, SAGE, Spl 7, SSX-1, 2, 3, 4, TRP2-INT2, carcinoembryonic antigen (CEA), kallikrein 4, mammaglobin-A, OAl, prostate-specific antigen (PSA), TRP-1 / gp75, TRP-2, adipophilin, interferon-inducible protein 2 (AIM-2) absent in melanoma, BING-4, CPSF, cyclin Dl, epithelial cell adhesion molecule (Ep-CAM), EphA3, fibroblast growth factor-5 (FGF-5), glycoprotein 250 (gp250), EGFR (ERBB1), HER-2 / neu (ERBB2), interleukin-13 receptor α2 chain (IL13Rα2), IL-6 receptor, intestinal carboxylesterase (iCE), alpha-fetoprotein (AFP), M-CSF, mdm-2, MUCl, p53 (TP53), PBF, PRAME, PSMA, RAGE-I, RNF43, RU2AS , SOXlO, STEAPl, survivin (BIRC5), human telomerase reverse transcriptase (hTERT), telomerase, Wilms tumor gene (WTl), SYCPl, BRDT, SPANX, XAGE, ADAM2, PAGE-5, LIPl, CTAGE-I, CSAGE, MMAl, CAGE, BORIS, HOM-TES-85, AF15ql4, HCA661, LDHC, MORC, SGY-I, SPOl 1, TPXl, NY-SAR-35, FTHL17, NXF2, TDRDl, TEX15, FATE, TPTE, immunoglobulin idiotype, Bence-Jones protein, estrogen receptor (ER), androgen receptor (AR), CD40, CD30, CD20, CD19, CD33, cancer antigen 72-4 (CA 72-4), cancer antigen 15-3 (CA 15-3), cancer antigen 27-29 (CA 27-29), cancer antigen 125 (CA 125), cancer antigen 19-9 (CA19-9), beta-human chorionic gonadotropin, beta-2 microglobulin, squamous cell carcinoma antigen, neurospecific enolase, heat shock protein gp96, GM2, sargramostim, CTLA-4, 707 alanine proline (707-AP), adenocarcinoma antigen recognized by T cells 4 (ART-4), carcinoembryonic antigen peptide-1 (CAP-I), calcium activated chloride channel-2 (CLCA2), cyclophilin B (Cyp-B), human signet ring tumor-2 (HST-2), human papillomavirus (HPV) proteins (HPV-E6, HPV-E7, major or minor capsid antigen, etc.), Epstein-Barr virus (EBV) proteins (EBV latent membrane protein-LMP1, LMP2; etc.), hepatitis B or C virus proteins, and HIV proteins.

[0266] In some embodiments, the IgSF conjugate, through its targeting agent, binds to tumor cells, tumor vasculature, or cellular components of the tumor microenvironment, thereby promoting killing of the targeted cells through modulation of the immune response (e.g., by activating costimulatory molecules or inhibiting negative regulatory molecules of immune cell activation), inhibition of survival signals (e.g., growth factor or cytokine or hormone receptor antagonists), activation of death signals, and / or immune-mediated cytotoxicity (e.g., by antibody-dependent cellular cytotoxicity). Such IgSF conjugates can function by several mechanisms to suppress, reduce, or eliminate tumor cells, e.g., to facilitate delivery of a conjugated effector moiety to the tumor target, e.g., by receptor-mediated endocytosis of the IgSF conjugate; or such conjugates can recruit, bind, and / or activate immune cells (e.g., NK cells, monocytes / macrophages, dendritic cells, T cells, B cells). Moreover, in some cases, one or more of the aforementioned pathways may be affected by administration of one or more IgSF conjugates of the invention.

[0267] In some embodiments, the targeting agent causes the IgSF conjugate to localize to, e.g., bind to, tumor cells, tumor vasculature, or cellular components of the tumor microenvironment, thereby modulating cells of the immune response near the tumor. In some embodiments, the targeting agent facilitates delivery of the conjugated IgSF (e.g., vIgD) to tumor targets, e.g., to interact with its cognate binding partner and alter signaling in immune cells (e.g., NK cells, monocytes / macrophages, dendritic cells, T cells, B cells) that bear the cognate binding partner. In some embodiments, the localized delivery agonizes or stimulates costimulatory receptors.

[0268] In some embodiments, the targeting agent is an immunoglobulin. As used herein, the term "immunoglobulin" includes natural or artificial monovalent or polyvalent antibodies, including, but not limited to, polyclonal, monoclonal, multispecific, human, humanized, or chimeric antibodies, single-chain antibodies, Fab fragments, F(ab') fragments, fragments produced by an Fab expression library, single-chain Fv (scFv); anti-idiotypic (anti-Id) antibodies (including, for example, anti-Id antibodies against the antibodies of the present invention), and epitope-binding fragments of any of the above. As used herein, the term "antibody" refers to immunoglobulin molecules and immunologically active portions of immunoglobulin molecules, for example, molecules containing an antigen-binding site that immunospecifically binds to an antigen. The immunoglobulin molecules of the present invention can be of any type (e.g., IgG, IgE, IgM, IgD, IgA, and IgY), class (e.g., IgG1, IgG2, IgG3, IgG4, IgA1, and IgA2), or subclass of immunoglobulin molecules.

[0269] In some embodiments, the IgSF conjugate, through its antibody targeting moiety, binds to tumor cells, tumor vasculature, or cellular components of the tumor microenvironment, thereby promoting apoptosis of the targeted cells through modulation of the immune response (e.g., by activating costimulatory molecules or inhibiting negative regulatory molecules of immune cell activation), inhibition of survival signals (e.g., growth factor or cytokine or hormone receptor antagonists), activation of death signals, and / or immune-mediated cytotoxicity (e.g., by antibody-dependent cellular cytotoxicity). Such IgSF conjugates can function through several mechanisms to suppress, reduce, or eliminate tumor cells, e.g., to facilitate delivery of a conjugated effector moiety to the tumor target, e.g., through receptor-mediated endocytosis of the IgSF conjugate; or such conjugates can recruit, bind, and / or activate immune cells (e.g., NK cells, monocytes / macrophages, dendritic cells, T cells, B cells).

[0270] In some embodiments, the IgSF conjugate, through its antibody targeting moiety, binds to tumor cells, tumor vasculature, or cellular components of the tumor microenvironment, thereby modulating the immune response (e.g., by activating costimulatory molecules or inhibiting negative regulatory molecules of immune cell activation). In some embodiments, such conjugates are capable of recognizing, binding to, and / or modulating (e.g., inhibiting or activating) immune cells (e.g., NK cells, monocytes / macrophages, dendritic cells, T cells, B cells).

[0271] Antibody targeting moieties of the present invention include antibody fragments, including, but not limited to, Fab, Fab', and F(ab'), Fd, single-chain Fv (scFv), single-chain antibodies, disulfide-linked Fv (sdFv), and fragments comprising either the VL or VH domain. Antigen-binding antibody fragments, including single-chain antibodies, can comprise the variable region alone or in combination with all or a portion of the following: hinge region, CH1, CH2, and CH3 domains. Also included in the present invention are antigen-binding fragments that also comprise any combination of the variable region and hinge region, CH1, CH2, and CH3 domains. Also included in the present invention are Fc fragments, antigen-Fc fusion proteins, and Fc-targeting moiety conjugates or fusion products (Fc-peptides, Fc-aptamers). The antibody targeting moieties of the present invention can be derived from any animal origin, including birds and mammals. In one aspect, the antibody targeting moiety is human, murine (e.g., mouse and rat), donkey, sheep, rabbit, goat, guinea pig, camel, horse, or chicken. Furthermore, such antibodies may be humanized versions of animal antibodies. The antibody targeting moieties of the present invention may be monospecific, bispecific, trispecific, or of higher order multispecificity.

[0272] In various embodiments, the antibody / targeting moiety recruits, binds, and / or activates immune cells (e.g., NK cells, monocytes / macrophages, dendritic cells) via interactions between Fc (in the antibody) and Fc receptors (on the immune cells) and via a conjugated variant polypeptide or immunomodulatory protein provided herein. In some embodiments, the antibody / targeting moiety recognizes or binds to tumor material via a conjugated variant polypeptide or immunomodulatory protein provided herein, localizing to tumor cells and facilitating modulation of immune cells in the vicinity of the tumor.

[0273] Examples of antibodies that can be incorporated into the IgSF conjugate include, but are not limited to, antibodies such as cetuximab (IMC-C225; Erbitux®), trastuzumab (Herceptin®), rituximab (Rituxan®; MabThera®), bevacizumab (Avastin®), alemtuzumab (Campath®; Campath-1H®; Mabcampath®), panitumumab (ABX-EGF; Vectibix®), ranibizumab (Lucentis®), ibritumomab, ibritumomab tiuxetan (Zevalin®), tositumomab, iodine I 131 Tositumomab (BEXXAR®), catumaxomab (Removab®), gemtuzumab, gemtuzumab ozogamicin (Mylotarg®), abatacept (CTLA4-Ig; Orencia®), belatacept (L104EA29YIg; LEA29Y; LEA), ipilimumab (MDX-010; MDX-101), tremelimumab (ticilimumab; CP-675,206), PRS-010, PRS-050, aflibercept (VEGF Trap, AVE005), volociximab (M200), F200, MORAb-009, SS1P (CAT-5001), cixutumumab (IMC-A12), matuzumab (EMD72000), nimotuzumab (h-R3), zalutumumab (HuMax-EGFR), necitumumab IMC-11F8, mAb806 / ch806, Sym004, mAb-425, Panorex@(17-1A) (murine monoclonal antibody); Panorex@(17-1A) (chimeric murine monoclonal antibody); IDEC-Y2B8 (murine, anti-CD2O MAb); BEC2 (anti-idiotypic MAb, mimicking the GD epitope) (containing BCG); Oncolym (Lym-1 monoclonal antibody); SMART MI95 Ab, humanized 13'I LYM-I (Oncolym), Ovarex (B43.13, anti-idiotypic murine MAb; MDX-210 (humanized anti-HER-2 bispecific antibody); 3622W94 MAb binding to EGP40 (17-1A) pancarcinoma antigen on adenocarcinoma; anti-VEGF, Zenapax (SMART anti-Tac (IL-2 receptor); SMART MI95 Ab, humanized Ab, humanized); MDX-210 (humanized anti-HER-2 bispecific antibody); MDX-447 (humanized anti-EGF receptor bispecific antibody); NovoMAb-G2 (pancarcinoma-specific Ab); TNT (chimeric MAb against histone antigen); TNT (chimeric MAb against histone antigen); Gliomab-H (Monoclon s-humanized Ab); GNI-250 These include Mab; EMD-72000 (chimeric EGF antagonist); Lymphoside (humanized LL2 antibody); and MDX-260 bispecific, targeting GD-2, ANA Ab, SMART ID10 Ab, SMART ABL 364 Ab, or ImmuRAIT-CEA. As exemplified in the foregoing list, it is routine to produce antibodies against specific target epitopes.

[0274] In some embodiments, the antibody targeting moiety is a full-length antibody or antigen-binding fragment thereof containing an Fc domain. In some embodiments, the variant polypeptide or immunomodulatory protein is conjugated to the Fc portion of the antibody targeting moiety, for example, by conjugation to the N-terminus of the Fc portion of the antibody.

[0275] In some embodiments, vIgD is linked, directly or indirectly, to the N- or C-terminus of the antibody light and / or heavy chain. In some embodiments, the linkage can be via a peptide linker, such as any of those described above. Various configurations can be constructed. Figures 10A-10C show exemplary configurations. In some embodiments, the antibody conjugate can be produced by coexpression of the antibody heavy and light chains in a cell.

[0276] In one aspect of the present invention, the targeting agent is an aptamer molecule. For example, in some embodiments, the aptamer is composed of a nucleic acid that functions as a targeting agent. In various embodiments, the IgSF conjugates of the present invention comprise an aptamer specific to a molecule on tumor cells, tumor vasculature, and / or tumor microenvironment. In some embodiments, the aptamer itself can comprise a biologically active sequence in addition to a targeting module (sequence), where the biologically active sequence can induce an immune response against the target cell. In other words, such an aptamer molecule is a dual-use agent. In some embodiments, the IgSF conjugates of the present invention comprise an aptamer conjugated to an antibody, where the aptamer and the antibody are specific for binding to different molecules on tumor cells, tumor vasculature, tumor microenvironment, and / or immune cells.

[0277] The term "aptamer" includes DNA, RNA, or peptides selected based on their specific binding properties to a particular molecule. For example, aptamers can be selected to bind to specific genes or gene products in tumor cells, tumor vasculature, tumor microenvironment, and / or immune cells as disclosed herein, where selection is performed by methods known in the art and familiar to those skilled in the art.

[0278] In some aspects of the present invention, the targeting agent is a peptide.For example, the variant polypeptide or immunomodulatory protein provided herein can be conjugated to a peptide that can bind to components of cancer or tumor cells.Therefore, such IgSF conjugates of the present invention include peptide targeting agents that bind to cellular components of tumor cells, tumor vasculature, and / or components of the tumor microenvironment.In some embodiments, the targeting agent peptide can be an integrin antagonist or agonist.Integrins containing alpha and beta subunits include many types well known to those skilled in the art.

[0279] In one embodiment, the targeting substance is Vvβ3. Integrin Vvβ3 is expressed on a wide variety of cells and has been shown to mediate several biologically relevant processes, including osteoclast adhesion to bone matrix, vascular smooth muscle cell migration, and angiogenesis. Suitable targeting molecules for integrins include RGD peptides or peptidomimetics and non-RGD peptides or peptidomimetics (see, for example, U.S. Patent Nos. 5,767,071 and 5,780,426) for other integrins, such as V4.βi (VLA-4), V4-P7 (see, for example, U.S. Patent No. 6,365,619; Chang et al., Bioorganic & Medicinal Chem Lett, 12:159-163 (2002); Lin et al., Bioorganic & Medicinal Chem Lett, 12:133-136 (2002)).

[0280] In some embodiments, an IgSF conjugate is provided, comprising a variant polypeptide or immunomodulatory protein provided herein conjugated to a therapeutic agent. In some embodiments, the therapeutic agent includes, for example, daunomycin, doxorubicin, methotrexate, and vindesine (Rowland et al., Cancer Immunol. Immunother. 21:183-187, 1986). In some embodiments, the therapeutic agent has intracellular activity. In some embodiments, the IgSF conjugate is internalized, and the therapeutic agent is a cytotoxin that blocks cellular protein synthesis, leading to cell death therein. In some embodiments, the therapeutic agent is a cytotoxin, including a polypeptide with ribosome-inactivating activity, including, for example, gelonin, bouganin, saporin, ricin, ricin A chain, bryodin, diphtheria toxin, restrictocin, Pseudomonas exotoxin A, and variants thereof. In some embodiments, when the therapeutic agent is a cytotoxin comprising a polypeptide with ribosome-inactivating activity, the IgSF conjugate must be internalized upon binding to the target cell in order for the protein to be cytotoxic to the cell.

[0281] In some embodiments, an IgSF conjugate is provided that comprises a variant polypeptide or an immunomodulatory protein provided herein conjugated to a toxin. In some embodiments, the toxin is, for example, a bacterial toxin, e.g., diphtheria toxin, a plant toxin, e.g., ricin, a small molecule toxin, e.g., geldanamycin (Mandler et al., J. Nat. Cancer Inst. 92 (19):1573-1581 (2000); Mandler et al., Bioorganic & Med. Chem. Letters 10:1025-1028 (2000); Mandler et al., Bioconjugate Chem. 13:786-791 (2002)), maytansinoids (EP 1391213; Liu et al., Proc. Natl. Acad. Sci. USA 93:8618-8623 (1996)), and calicheamicin (Lode et al., Cancer Res. 58:2928 (1998); Hinman et al., Cancer Res. 53:3336-3342 (1993)). Toxins can exert their cytotoxic and cytostatic effects by mechanisms including tubulin binding, DNA binding, or topoisomerase inhibition.

[0282] In some embodiments, IgSF conjugates are provided, which comprise the variant polypeptide or immunomodulatory protein provided herein conjugated with a label that can indirectly or directly generate detectable signals.These IgSF conjugates can be used for research or diagnostic purposes, for example, for in vivo cancer detection.Preferably, the label can generate detectable signals either directly or indirectly.For example, the label can be a radiopaque or radioisotope, such as 3H, 14C, 32P, 35S, 123I, 125I, 131I; a fluorescent (fluorophore) or chemiluminescent (chromophore) compound, such as fluorescein isothiocyanate, rhodamine, or luciferin; an enzyme, such as alkaline phosphatase, β-galactosidase, or horseradish peroxidase; an imaging agent; or a metal ion. In some embodiments, the label is a radioactive atom, such as 99Tc or 123I, for scintigraphy studies, or a spin label, such as zirconium-89, iodine-123, iodine-131, indium-111, fluorine-19, carbon-13, nitrogen-15, oxygen-17, gadolinium, manganese, or iron, for nuclear magnetic resonance (NMR) imaging (also known as magnetic resonance imaging, MRI). Zirconium-89 may be complexed to various metal chelators and conjugated to antibodies, for example, for PET imaging (WO 2011 / 056983). In some embodiments, the IgSF conjugate is indirectly detectable. For example, the IgSF conjugate can be detected using a secondary antibody specific for the IgSF conjugate and containing a detectable label.

[0283] IgSF conjugates may be prepared using any method known in the art, see, for example, WO 2009 / 067800, WO 2011 / 133886, and U.S. Patent Application Publication No. 2014322129, which are incorporated by reference in their entireties.

[0284] The variant polypeptide or immunomodulatory protein of an IgSF conjugate can be "conjugated" to the effector moiety by any means that allows the variant polypeptide or immunomodulatory protein to associate with or link to the effector moiety. For example, the variant polypeptide or immunomodulatory protein of an IgSF conjugate can be conjugated to the effector moiety by chemical or recombinant means. Chemical means for preparing fusions or conjugates are known in the art and can be used to prepare IgSF conjugates. The method used to conjugate the variant polypeptide or immunomodulatory protein to the effector moiety must be capable of connecting the variant polypeptide or immunomodulatory protein to the effector moiety without interfering with the ability of the variant polypeptide or immunomodulatory protein to bind to its one or more counterstructure ligands.

[0285] The IgSF-conjugated variant polypeptide or immunomodulatory protein may be indirectly linked to an effector moiety. For example, the IgSF-conjugated variant polypeptide or immunomodulatory protein may be directly linked to a liposome containing one of several effector moieties. Alternatively, the effector moiety and / or the variant polypeptide or immunomodulatory protein may be bound to a solid surface.

[0286] In some embodiments, the variant polypeptide or immunomodulatory protein and the effector moiety of the IgSF conjugate are both proteins, and they can be conjugated using techniques well known in the art. There are hundreds of crosslinkers available that can conjugate two proteins (see, for example, "Chemistry of Protein Conjugation and Crosslinking," 1991, Shans Wong, CRC Press, Ann Arbor). Crosslinkers are generally selected based on the reactive functional groups available on or inserted into the variant polypeptide or immunomodulatory protein and / or effector moiety. In addition, if there are no reactive groups, photoactivatable crosslinkers can be used. In certain cases, it may be desirable to include a spacer between the variant polypeptide or immunomodulatory protein and the effector moiety. Cross-linking agents known in the art include homobifunctional agents: glutaraldehyde, dimethyladipimidate, and bis(diazobenzidine), and heterobifunctional agents: m-maleimidobenzoyl-N-hydroxysuccinimide and sulfo-m-maleimidobenzoyl-N-hydroxysuccinimide.

[0287] In some embodiments, the variant polypeptide or immunomodulatory protein of the IgSF conjugate may be modified at specific residues for chemical conjugation of the effector moiety. Specific residues known in the art for use in chemical conjugation of molecules include lysine and cysteine. The crosslinker is selected based on the reactive functional groups inserted on the variant polypeptide or immunomodulatory protein and the reactive functional groups available on the effector moiety.

[0288] IgSF conjugates can also be prepared using recombinant DNA technology. In such cases, the DNA sequence encoding the variant polypeptide or immunomodulatory protein is fused to the DNA sequence encoding the effector portion to obtain a chimeric DNA molecule. The chimeric DNA sequence is transfected into host cells that express the fusion protein. The fusion protein can be recovered from the cell culture and purified using techniques known in the art.

[0289] Examples of conjugating a labeling effector moiety to a variant polypeptide or immunomodulatory protein include those described in Hunter, et al., Nature 144:945 (1962); David, et al., Biochemistry 13:1014 (1974); Pain, et al., J. Immunol. Meth. 40:219 (1981); Nygren, J. Histochem. and Cytochem. 30:407 (1982); Wensel and Meares, Radioimmunoimaging And Radioimmunotherapy, Elsevier, NY (1983); and Colcher et al., "Use Of Monoclonal Antibodies As Radiopharmaceuticals For The Localization Of Human Carcinoma Xenografts In Athymic Mice", Meth. Enzymol., 121:802-16 (1986).

[0290] Radiolabels or other labels may be incorporated into the conjugates by known methods. For example, peptides may be biosynthesized or synthesized by chemical amino acid synthesis using suitable amino acid precursors, e.g., containing fluorine-19 instead of hydrogen. Labels such as 99Tc or 123I, 186Re, 188Re, and 111In can be attached via cysteine ​​residues in the peptide. Yttrium-90 can be attached via lysine residues. Iodine-123 can be incorporated using the IODOGEN method (Fraker et al., Biochem. Biophys. Res. Commun. 80:49-57 (1978)). "Monoclonal Antibodies in Immunoscintigraphy" (Chatal, CRC Press 1989) describes other methods in detail.

[0291] Conjugates of variant polypeptides or immunomodulatory proteins and cytotoxic agents can be prepared using a wide variety of bifunctional protein coupling agents, such as N-succinimidyl-3-(2-pyridyldithio)propionate (SPDP), succinimidyl-4-(N-maleimidomethyl)cyclohexane-1-carboxylate (SMCC), iminothiolane (IT), bifunctional derivatives of imidoesters (e.g., dimethyl adipimidate HCl), activated esters ( For example, radioactive nucleotides may be prepared using disuccinimidyl suberate, aldehydes (e.g., glutaraldehyde), bis-azido compounds (e.g., bis(p-azidobenzoyl)hexanediamine), bis-diazonium derivatives (e.g., bis-(p-diazoniumbenzoyl)-ethylenediamine), diisocyanates (e.g., toluene 2,6-diisocyanate), and bis-active fluorine compounds (e.g., 1,5-difluoro-2,4-dinitrobenzene). For example, ricin immunotoxins can be prepared as described in Vitetta et al., Science 238:1098 (1987). Carbon-14 labeled 1-isothiocyanatobenzyl-3-methyldiethylenetriaminepentaacetic acid (MX-DTPA) is an exemplary chelating agent for conjugating radionucleotides to antibodies. See, e.g., WO 94 / 11026. The linker may be a "cleavable linker" that facilitates the release of the cytotoxic drug in cells. For example, an acid-labile linker, a peptidase-sensitive linker, a photolabile linker, a dimethyl linker, or a disulfide-containing linker (Chari et al., Cancer Research 52:127-131 (1992); U.S. Patent No. 5,208,020) may be used.

[0292] The IgSF conjugates of the present invention expressly contemplate, but are not limited to, drug conjugates prepared with cross-linker reagents: BMPS, EMCS, GMBS, HBVS, LC-SMCC, MBS, MPBH, SBAP, SIA, SIAB, SMCC, SMPB, SMPH, sulfo-EMCS, sulfo-GMBS, sulfo-KMUS, sulfo-MBS, sulfo-SIAB, sulfo-SMCC, and sulfo-SMPB, and SVSB (succinimidyl-(4-vinylsulfone)benzoate), available commercially (e.g., from Pierce Biotechnology, Inc., Rockford, IL, USA). See pages 467-498 of the Applications Handbook and Catalog 2003-2004.

[0293] D. Transmembrane and Secreted Immunomodulatory Proteins and Modified Cells Provided herein are modified cells (alternatively, "modified cells") that express an immunomodulatory variant ICOSL polypeptide. In some embodiments, the expressed immunomodulatory variant ICOSL polypeptide is a transmembrane protein and is expressed on the surface. In some embodiments, the expressed immunomodulatory variant ICOSL polypeptide is expressed in and secreted from the cell.

[0294] 1. Transmembrane immunomodulatory proteins In some embodiments, the immunomodulatory polypeptide comprising a variant ICOSL can be a membrane-bound protein. As described in more detail below, the immunomodulatory polypeptide can be a transmembrane immunomodulatory polypeptide comprising a variant ICOSL, containing an ectodomain containing at least one affinity-modified IgSF domain (IgV or IgC), a transmembrane domain, and optionally, a cytoplasmic domain. In some embodiments, the transmembrane immunomodulatory protein can be expressed on the surface of an immune cell, e.g., a mammalian cell, including the surface of a lymphocyte (e.g., T cell or NK cell) or an antigen-presenting cell. In some embodiments, the transmembrane immunomodulatory protein is expressed on the surface of a mammalian T cell, including a T cell such as a helper T cell, a cytotoxic T cell (or a cytotoxic T lymphocyte or CTL), a natural killer T cell, a regulatory T cell, a memory T cell, or a γδ T cell. In some embodiments, the mammalian cell is an antigen-presenting cell (APC). Typically, but not exclusively, the ectodomain (or alternatively, "extracellular domain") comprises one or more amino acid modifications (e.g., amino acid substitutions) of a variant ICOSL of the invention. Thus, for example, in some embodiments, a transmembrane protein comprises an ectodomain comprising one or more amino acid substitutions of a variant ICOSL of the invention.

[0295] In some embodiments, the engineered cells express a variant ICOSL polypeptide that is a transmembrane immunomodulatory polypeptide (TIP), which can be a membrane protein such as a transmembrane protein. In typical embodiments, the ectodomain of the membrane protein comprises the extracellular domain of the variant ICOSL provided herein, or its IgSF domain, containing one or more amino acid substitutions in at least one IgSF domain as described. The transmembrane immunomodulatory proteins provided herein further contain a transmembrane domain linked to the ectodomain. In some embodiments, the transmembrane domain provides the encoded protein for cell surface expression on the cell. In some embodiments, the transmembrane domain is directly linked to the ectodomain. In some embodiments, the transmembrane domain is indirectly linked to the ectodomain via one or more linkers or spacers. In some embodiments, the transmembrane domain contains predominantly hydrophobic amino acid residues, such as leucine and valine.

[0296] In some embodiments, a full-length transmembrane anchor domain can be used to ensure that TIP is expressed on the surface of engineered cells, such as engineered T cells. Conveniently, this can be derived from the particular native protein (e.g., ICOSL or other native IgSF protein) for which affinity is being engineered, and can be simply fused to the sequence of the first membrane-proximal domain in the same manner as a native IgSF protein (e.g., ICOSL). In some embodiments, the transmembrane immunomodulatory protein comprises the transmembrane domain of a corresponding wild-type or unmodified IgSF member (e.g., the transmembrane domain contained in the amino acid sequence shown in SEQ ID NO:5) (Table 2). In some embodiments, the membrane-bound form comprises the transmembrane domain of a corresponding wild-type or unmodified polypeptide, e.g., corresponding to residues 257-277 of SEQ ID NO:5.

[0297] In some embodiments, the transmembrane domain is a non-natural transmembrane domain that is not the transmembrane domain of native ICOSL. In some embodiments, the transmembrane domain is derived from a transmembrane domain from another non-ICOSL family member polypeptide that is a membrane-bound or transmembrane protein. In some embodiments, a transmembrane anchor domain from another protein on T cells can be used. In some embodiments, the transmembrane domain is derived from CD8. In some embodiments, the transmembrane domain can further contain the extracellular portion of CD8 that serves as a spacer domain. Exemplary CD8-derived transmembrane domains are set forth in SEQ ID NO: 246 or 483, or portions thereof containing the CD8 transmembrane domain. In some embodiments, the transmembrane domain is a synthetic transmembrane domain.

[0298] In some embodiments, the transmembrane immunomodulatory protein further comprises an endodomain, such as a cytoplasmic signaling domain, linked to the transmembrane domain. In some embodiments, the cytoplasmic signaling domain induces cell signaling. In some embodiments, the endodomain of the transmembrane immunomodulatory protein comprises the cytoplasmic domain of the corresponding wild-type or unmodified polypeptide, such as the cytoplasmic domain contained in the amino acid sequence set forth in SEQ ID NO:5 (see Table 2).

[0299] In some embodiments, the provided transmembrane immunomodulatory protein is or comprises a variant ICOSL. The provided transmembrane immunomodulatory protein comprises an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to SEQ ID NO:257, and contains an ectodomain and a transmembrane domain that includes at least one affinity-modified ICOSL IgSF domain as described. In some embodiments, the transmembrane immunomodulatory protein contains any one or more amino acid substitutions in the IgSF domain (e.g., IgV domain) as described, including any shown in Table 1. In some embodiments, the transmembrane immunomodulatory protein can further comprise a cytoplasmic domain as described. In some embodiments, the transmembrane immunomodulatory protein can further contain a signal peptide. In some embodiments, the signal peptide is the native signal peptide of a wild-type IgSF member, for example, as contained in the amino acid sequence set forth in SEQ ID NO:5 (see, eg, Table 2).

[0300] In some embodiments, the amino acid substitution In some embodiments, a transmembrane immunomodulatory protein is provided, comprising: TIFF0007623967000054.tif11164. In some embodiments, the transmembrane immunomodulatory protein is provided as set forth in SEQ ID NO:257, but with an amino substitution TIFF0007623967000055.tif11164 at the corresponding positions in SEQ ID NO: 257, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO: 257 and amino acid substitutions. TIFF0007623967000056.tif11164, being or comprising variant ICOSL.

[0301] In some embodiments, the amino acid sequence set forth in SEQ ID NO:496 or 497 (each containing the amino acid substitution N52D), SEQ ID NO:498 or 499 (each containing the amino acid substitutions N52H / N57Y / Q100P), SEQ ID NO:500 or 501 (each containing the amino acid substitutions E16V / N52H / N57Y / Q100R / V110D / H115R / Y152C / K156M / C198R) or SEQ ID NO:502 or 503 (each containing the amino acid substitutions N52H / N57Y / Q100R), or SEQ ID NO: Provided are transmembrane immunomodulatory proteins comprising an amino acid sequence that comprises at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to any of SEQ ID NOs:495-503 and contains the indicated amino acid substitutions. In some embodiments, when expressed in a modified cell, such transmembrane immunomodulatory proteins are expressed on the surface of the cell.

[0302] Also provided are nucleic acid molecules encoding such transmembrane immunomodulatory proteins. In some embodiments, the nucleic acid molecule encoding the transmembrane immunomodulatory protein comprises a nucleotide sequence encoding an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% sequence identity to SEQ ID NO: 257, and contains an ectodomain comprising at least one affinity-modified IgSF domain as described, a transmembrane domain, and optionally, a cytoplasmic domain. In some embodiments, the nucleic acid molecule can further comprise a sequence of nucleotides encoding a signal peptide. In some embodiments, the signal peptide is the native signal peptide of the corresponding wild-type IgSF member (see, for example, Table 2).

[0303] An exemplary transmembrane immunomodulatory protein is ICOSL TIP, which comprises i) the amino acid sequence set forth in SEQ ID NO:383, or ii) an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:243 and includes an affinity-modified domain or amino acid substitutions therein contained in SEQ ID NO:243. Also provided are i) the nucleotide sequence set forth in SEQ ID NO:244, ii) a sequence encoding a TIP that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% sequence identity to SEQ ID NO:244 and includes the affinity-engineered domain of SEQ ID NO:243, or iii) a sequence of i) or ii) with degenerate codons.

[0304] In some embodiments, a CAR-associated transmembrane immunomodulatory protein is provided, wherein the endodomain of the transmembrane immunomodulatory protein comprises a cytoplasmic signaling domain that includes at least one ITAM (immunoreceptor tyrosine-based activation motif)-containing signaling domain. ITAM is a conserved motif found in many protein signaling domains involved in immune cell signaling, including the CD3-zeta chain ("CD3-z"), which is involved in T cell receptor signaling. In some embodiments, the endodomain comprises a CD3-zeta signaling domain. In some embodiments, the CD3-zeta signaling domain comprises the amino acid sequence set forth in SEQ ID NO:243, or an amino acid sequence that is at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% similar to SEQ ID NO:247 and retains T cell signaling activity. In some embodiments, the endodomain of the CAR-associated transmembrane immunomodulatory protein can further comprise a costimulatory signaling domain that further modulates the immunomodulatory response of T cells. In some embodiments, the costimulatory signaling domain is CD28, ICOS, 41BB, or OX40. In some embodiments, the costimulatory signaling domain is derived from CD28 or 4-1BB and comprises an amino acid sequence set forth in any of SEQ ID NOs:484-487, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% similarity to SEQ ID NOs:484-487 and retains T cell costimulatory signaling activity. In some embodiments, the provided CAR-associated transmembrane immunomodulatory proteins have the characteristic of CARs that stimulate T cell signaling through binding of the affinity-engineered IgSF domain to a cognate binding partner or counter-structure. In some embodiments, specific binding of the affinity-engineered IgSF domain to its counter-structure can lead to altered immune activity of T cell activity, as reflected by changes in cytotoxicity, proliferation, or cytokine production.

[0305] In some embodiments, the transmembrane immunomodulatory protein does not contain an endodomain capable of mediating cytoplasmic signal transduction. In some embodiments, the transmembrane immunomodulatory protein lacks the signal transduction mechanism of the wild-type or unmodified polypeptide and therefore does not induce cell signal transduction by itself. In some embodiments, the transmembrane immunomodulatory protein lacks the intracellular (cytoplasmic) domain of the corresponding wild-type or unmodified polypeptide, or lacks a portion of the intracellular domain, for example, the cytoplasmic signal transduction domain contained in the amino acid sequence set forth in SEQ ID NO:5 (see Table 2). In some embodiments, the transmembrane immunomodulatory protein does not contain an ITIM (immunoreceptor tyrosine-based inhibitory motif), which is contained in certain inhibitory receptors, including, for example, inhibitory receptors of the IgSF family (e.g., PD-1 or TIGIT). Thus, in some embodiments, the transmembrane immunomodulatory protein contains only an ectodomain and a transmembrane domain (e.g., any of those described).

[0306] 2. Secreted immunomodulatory proteins and modified cells In some embodiments, an ICOSL variant immunomodulatory polypeptide containing any one or more of the amino acid mutations described herein is secretable, for example, when expressed in a cell. Such a variant ICOSL immunomodulatory protein does not contain a transmembrane domain. In some embodiments, the variant ICOSL immunomodulatory protein is not conjugated to a half-life extending moiety (e.g., an Fc domain or a multimerization domain). In some embodiments, the variant ICOSL immunomodulatory protein contains a signal peptide, such as an antibody signal peptide or other efficient signal sequence, that directs the domain outside the cell. When an immunomodulatory protein contains a signal peptide and is expressed by a modified cell, the signal peptide causes the immunomodulatory protein to be secreted by the modified cell. Generally, the signal peptide, or a portion of the signal peptide, is cleaved from the immunomodulatory protein upon secretion. The immunomodulatory protein can be encoded by a nucleic acid (which can be part of an expression vector). In some embodiments, the immunomodulatory protein is expressed and secreted by a cell (e.g., an immune cell, e.g., a primary immune cell).

[0307] Thus, in some embodiments, a variant ICOSL immunomodulatory protein is provided, further comprising a signal peptide. In some embodiments, provided herein is a nucleic acid molecule encoding a variant ICOSL immunomodulatory protein operably linked to a secretory sequence encoding a signal peptide.

[0308] A signal peptide is a sequence at the N-terminus of an immunomodulatory protein that signals secretion of the immunomodulatory protein from a cell. In some embodiments, the signal peptide is about 5 to about 40 amino acids in length (e.g., about 5 to about 7, about 7 to about 10, about 10 to about 15, about 15 to about 20, about 20 to about 25, about 25 to about 30, about 30 to about 35, or about 35 to about 40 amino acids).

[0309] In some embodiments, the signal peptide is a native signal peptide from the corresponding wild-type ICOSL (see Table 6). In some embodiments, the signal peptide is a non-native signal peptide. For example, in some embodiments, the non-native signal peptide is a variant native signal peptide from the corresponding wild-type ICOSL and can include one or more (e.g., 2, 3, 4, 5, 6, 7, 8, 9, or 10 or more) substitutions, insertions, or deletions. In some embodiments, the non-native signal peptide is a signal peptide of a family member from the same IgSF family as the wild-type IgSF family member, or a variant thereof. In some embodiments, the non-native signal peptide is a signal peptide of an IgSF family member from an IgSF family different from the wild-type IgSF family member, or a variant thereof. In some embodiments, the signal peptide is a signal peptide or variant thereof from a non-IgSF protein family, such as a signal peptide from an immunoglobulin (e.g., IgG heavy chain or IgGκ light chain), a cytokine (e.g., interleukin-2 (IL-2), or CD33), a serum albumin protein (e.g., HSA or albumin), a human azurocidin preprotein signal sequence, luciferase, a trypsinogen (e.g., chymotrypsinogen or trypsinogen), or other signal peptide capable of efficiently secreting proteins from cells. Exemplary signal peptides include any listed in Table 6.

[0310] Table 6. Exemplary signal peptides TIFF0007623967000057.tif94170

[0311] In some embodiments of the secreted variant ICOSL immunomodulatory protein, the immunomodulatory protein includes a signal peptide when expressed, and the signal peptide (or a portion thereof) is cleaved from the immunomodulatory protein upon secretion.

[0312] In some embodiments, the engineered cells express a variant ICOSL polypeptide that is secreted from the cells. In some embodiments, such a variant ICOSL polypeptide is encoded by a nucleic acid molecule encoding an immunomodulatory protein under the operable control of a signal sequence for secretion. In some embodiments, the encoded immunomodulatory protein is secreted when expressed from the cells. In some embodiments, the immunomodulatory protein encoded by the nucleic acid molecule does not include a transmembrane domain. In some embodiments, the immunomodulatory protein encoded by the nucleic acid molecule does not include a half-life extending moiety (e.g., an Fc domain or a multimerization domain). In some embodiments, the immunomodulatory protein encoded by the nucleic acid molecule includes a signal peptide. In some embodiments, the nucleic acids of the invention further include a nucleotide sequence encoding a secretory or signal peptide operably linked to the nucleic acid encoding the immunomodulatory protein, thereby enabling secretion of the immunomodulatory protein.

[0313] 3. Cells and Cell Modifications Provided herein are modified cells that express any of the provided immunomodulatory polypeptides. In some embodiments, the modified cells express any of the provided transmembrane immunomodulatory polypeptides on their surface. In some embodiments, the modified cells express the immunomodulatory protein and are capable of secreting it from the cell under conditions suitable for protein secretion, or are capable of secreting it. In some embodiments, the immunomodulatory protein is expressed on lymphocytes, such as tumor-infiltrating lymphocytes (TILs), T cells, or NK cells, or on myeloid cells. In some embodiments, the modified cells are antigen-presenting cells (APCs). In some embodiments, the modified cells are modified mammalian T cells or modified mammalian antigen-presenting cells (APCs). In some embodiments, the modified T cells or APCs are human or murine cells.

[0314] In some embodiments, the modified T cells include, but are not limited to, helper T cells, cytotoxic T cells (alternatively, cytotoxic T lymphocytes or CTLs), natural killer T cells, regulatory T cells, memory T cells, or γδ T cells. + or CD8 + In addition to MHC signals, modified T cells also require costimulatory signals, which in some embodiments are provided by the variant ICOSL transmembrane immunomodulatory polypeptide, which is expressed in a membrane-bound form as discussed above.

[0315] In some embodiments, modified APCs include, for example, APCs that express MHC II, such as macrophages, B cells, and dendritic cells, as well as aAPCs, including both cellular and acellular (e.g., biodegradable polymeric microparticles) artificial APCs (aAPCs). Artificial APCs (aAPCs) are synthetic versions of APCs that can act in a similar manner to APCs in that they present antigens to T cells and activate them. Antigen presentation is carried out by MHC (class I or class II). In some embodiments, in modified APCs such as aAPCs, the antigen loaded on MHC is, in some embodiments, a tumor-specific antigen or tumor-associated antigen. The antigen loaded on MHC is recognized by the T cell receptor (TCR) of T cells, which in some cases can express ICOS, CD28, or other molecules recognized by the variant ICOSL polypeptide provided herein. Materials that can be used to modify aAPCs include poly(glycolic acid), poly(lactic-co-glycolic acid), iron oxide, liposomes, lipid bilayers, sepharose, and polystyrene.

[0316] In some embodiments, the immunomodulatory proteins provided herein, e.g., transmembrane immunomodulatory proteins or secreted immunomodulatory proteins, are co-expressed or engineered in cells expressing an antigen-binding receptor, e.g., a recombinant receptor, e.g., a chimeric antigen receptor (CAR) or a T cell receptor (TCR). In some embodiments, the engineered cells, e.g., engineered T cells, recognize a desired antigen associated with cancer, inflammatory and autoimmune disorders, or viral infection. In specific embodiments, the antigen-binding receptor contains an antigen-binding moiety that specifically binds to a tumor-specific antigen or a tumor-associated antigen. In some embodiments, the engineered T cells are CAR (chimeric antigen receptor) T cells that contain an antigen-binding domain (e.g., scFv) that specifically binds to an antigen, e.g., a tumor-specific antigen or a tumor-associated antigen. In some embodiments, the antigen-binding domain (e.g., scFv) is specific for CD19. An exemplary CAR is an anti-CD19 CAR, e.g., a CAR containing an anti-CD19 scFv as set forth in SEQ ID NO: 482 or SEQ ID NO: 245. In some embodiments, the TIP protein is expressed in modified T cell receptor cells or modified chimeric antigen receptor cells. In such embodiments, the modified cells co-express TIP and a CAR or TCR.

[0317] In some embodiments, the CAR further contains an intracellular signaling domain (endodomain) comprising a spacer or hinge, a transmembrane domain, and an ITAM signaling domain (e.g., a CD3ζ signaling domain). In some embodiments, the CAR further comprises a costimulatory signaling domain. In some embodiments, the spacer or hinge is located between the antigen-binding domain and the transmembrane domain, e.g., between the antigen-binding domain and the plasma membrane when expressed on a cell. In some embodiments, the spacer or hinge is derived from an IgG subclass (e.g., IgG1 and IgG4), IgD, or CD8 (see, e.g., Qin et al. (2017) J. Hematol. Oncol., 10:68). In some embodiments, the spacer or hinge is derived from IgG1.

[0318] In some embodiments, the spacer and transmembrane domains are hinge and transmembrane domains derived from CD8, e.g., as set forth in SEQ ID NO:246 or 483, or in an amino acid sequence exhibiting at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or higher sequence identity to SEQ ID NO:246 or 483. In some embodiments, the endodomain comprises a CD3-zeta signaling domain. In some embodiments, the CD3-zeta signaling domain comprises the amino acid sequence set forth in SEQ ID NO:243, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, or 99% identity to SEQ ID NO:247 and retains T cell signaling activity. In some embodiments, the endodomain of the CAR-associated transmembrane immunomodulatory protein can further comprise a costimulatory signaling domain that further modulates the immunomodulatory response of T cells. In some embodiments, the costimulatory signaling domain is CD28, ICOS, 41BB, or OX40. In some embodiments, the costimulatory signaling domain is derived from CD28 or 4-1BB and comprises an amino acid sequence set forth in any of SEQ ID NOs:484-487, or an amino acid sequence that exhibits at least 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% identity to SEQ ID NOs:484-487 and retains T cell costimulatory signaling activity.

[0319] In another embodiment, the modified T cell possesses a TCR comprising a recombinant or modified TCR. In some embodiments, the TCR can be a natural TCR. Those skilled in the art will recognize that natural mammalian T cell receptors generally comprise alpha and beta chains (or gamma and delta chains) involved in antigen-specific recognition and binding. In some embodiments, the TCR is a modified TCR that has been modified. In some embodiments, the TCR of the modified T cell specifically binds to a tumor-associated or tumor-specific antigen presented by an APC.

[0320] In some embodiments, immunomodulatory polypeptides, such as transmembrane immunomodulatory polypeptides or secreted immunomodulatory polypeptides, can be incorporated into engineered cells, such as engineered T cells or engineered APCs, by a variety of strategies, such as those used in recombinant host cells. A variety of methods for introducing DNA constructs into primary T cells are known in the art. In some embodiments, viral transduction or plasmid electroporation is used. In typical embodiments, the nucleic acid molecule encoding the immunomodulatory protein or the expression vector contains a signal peptide that localizes the expressed transmembrane immunomodulatory protein to the cell membrane or for secretion. In some embodiments, the nucleic acid encoding the transmembrane immunomodulatory protein of the present invention is subcloned into a viral vector, such as a retroviral vector, that allows expression in host mammalian cells. The expression vector can be introduced into mammalian host cells, and under host cell culture conditions, the immunomodulatory protein is expressed on the surface or secreted.

[0321] In an illustrative example, primary T cells can be purified ex vivo (CD4 cells or CD8 cells, or both) and stimulated with an activation protocol consisting of beads coated with various TCR / CD28 agonists, such as anti-CD3 / anti-CD28. After a two- or three-day activation process, recombinant expression vectors containing immunomodulatory polypeptides can be stably introduced into primary T cells using lentiviral or retroviral transduction protocols or plasmid electroporation strategies standard in the art. Cells can be monitored for immunomodulatory polypeptide expression, for example, by flow cytometry using anti-epitope tags or antibodies that cross-react with the native parent molecule and polypeptide (including variant ICOSL). T cells expressing immunomodulatory polypeptides can be enriched by sorting with anti-epitope tag antibodies or enriched for high or low expression, depending on the application.

[0322] By expressing immunomodulatory polypeptides, engineered T cells can be assayed for appropriate function by a variety of means. Coexpression of the engineered CAR or TCR can be verified to show that this portion of the engineered T cells was largely unaffected by the expression of the immunomodulatory protein. Once verified, the functionality of the engineered T cells can be assessed using standard in vitro cytotoxicity, proliferation, or cytokine assays (e.g., IFN-γ expression). Exemplary standard endpoints are the rate of lysis of a tumor line, the proliferation of engineered T cells, or IFN-γ protein expression in the culture supernatant. Engineered constructs that result in a statistically significant increase in tumor line lysis, increased proliferation of engineered T cells, or increased IFN-γ expression relative to control constructs can be selected. Additionally, unmodified cells, such as natural primary or endogenous T cells, can also be incorporated into the same in vitro assay to measure the ability of the immunomodulatory polypeptide constructs expressed on engineered cells, such as engineered T cells, to modulate the activity of bystander natural T cells (including, in some cases, activating and generating effector functions). Increased expression of activation markers, such as CD69, CD44, or CD62L, can also be monitored on endogenous T cells, and increased proliferation and / or cytokine production can indicate the desired activity of the immunomodulatory protein expressed on the modified T cells.

[0323] In some embodiments, similar assays can be used to compare the function of engineered T cells containing CAR or TCR alone with those containing CAR or TCR and TIP constructs. Typically, these in vitro assays are performed by plating various ratios of engineered T cells and "tumor" cell lines containing cognate CAR or TCR antigens in culture. Standard endpoints are tumor line lysis rate, engineered T cell proliferation, or IFN-γ production in the culture supernatant. Engineered immunomodulatory proteins that resulted in statistically significant increases in tumor line lysis, engineered T cell proliferation, or IFN-γ production relative to the same TCR or CAR construct alone can be selected. Engineered human T cells can be analyzed in immunodeficient mice, such as the NSG strain, which lacks T, NK, and B cells. Engineered human T cells in which CAR or TCR bind to a target counter structure on the xenograft and coexpress the affinity-engineered IgSF domain of TIP can be adoptively transferred in vivo at different cell numbers and ratios compared to the xenograft. For example, CD19 containing a luciferase / GFP vector can be used. + The engraftment of leukemia tumor lines can be monitored by bioluminescence or ex vivo by flow cytometry. In one common embodiment, xenografts are introduced into murine models, followed by the introduction of modified T cells several days later. Modified T cells containing immunomodulatory proteins can be assayed for increased survival, tumor clearance, or the proliferation number of modified T cells compared to modified T cells containing CAR or TCR alone. Similar to in vitro assays, endogenous natural (i.e., unmodified) human T cells can also be co-adoptively transferred to predict the success of epitope spreading in the population, which leads to better survival or tumor clearance.

[0324] E. Infectious Agents Expressing Variant Polypeptides and Immunomodulatory Proteins Also provided are infectious agents containing nucleic acids encoding any of the variant polypeptides described herein, including secreted or transmembrane immunomodulatory proteins, e.g., ICOSL vIgD polypeptides. In some embodiments, such infectious agents are capable of delivering nucleic acids encoding variant immunomodulatory polypeptides described herein, e.g., ICOSL vIgD polypeptides, to target cells of a subject, e.g., immune cells and / or antigen-presenting cells (APCs) or tumor cells of the subject. Also provided are nucleic acids contained in such infectious agents and / or nucleic acids, e.g., vectors and / or plasmids, for generating or modifying such infectious agents, as well as compositions containing such infectious agents.

[0325] In some embodiments, the infectious agent is a microorganism or pathogen. In some embodiments, the infectious agent is a virus or bacteria. In some embodiments, the infectious agent is a virus. In some embodiments, the infectious agent is a bacteria. In some embodiments, such an infectious agent can deliver a nucleic acid sequence encoding any of the variant polypeptides, including secreted or transmembrane immunomodulatory proteins described herein, such as an ICOSL vIgD polypeptide. Thus, in some embodiments, cells of a subject infected with or contacted by an infectious agent can express or secrete the variant immunomodulatory polypeptide on the cell surface. In some embodiments, the infectious agent can also deliver nucleic acids encoding one or more other therapeutic agents or other therapeutic agents to cells and / or the environment within the subject. In some embodiments, other therapeutic agents that can be delivered by the infectious agent include cytokines or other immunomodulatory molecules.

[0326] In some embodiments, the infectious agent (e.g., a virus or bacterium) contains a nucleic acid sequence encoding any of the variant polypeptides (e.g., ICOSL vIgD polypeptides) comprising secreted or transmembrane immunomodulatory proteins described herein, and upon contact and / or infection of cells of a subject, the cells express the variant polypeptide (e.g., ICOSL vIgD polypeptides) comprising secreted or transmembrane immunomodulatory proteins encoded by the nucleic acid sequence contained in the infectious agent. In some embodiments, the infectious agent can be administered to a subject. In some embodiments, the infectious agent can be introduced into cells from a subject ex vivo.

[0327] In some embodiments, variant polypeptides comprising transmembrane immunomodulatory proteins, such as ICOSL vIgD polypeptides, expressed by cells infected with an infectious agent are transmembrane proteins and expressed on the surface. In some embodiments, variant polypeptides comprising secreted immunomodulatory proteins, such as ICOSL vIgD polypeptides, expressed by cells infected with an infectious agent are expressed and secreted from the cells. The transmembrane immunomodulatory proteins or secreted immunomodulatory proteins can be any of those described herein.

[0328] In some embodiments, the cells of interest targeted by the infectious agent include tumor cells, immune cells, and / or antigen-presenting cells (APCs). In some embodiments, the infectious agent targets cells in the tumor microenvironment (TME). In some embodiments, the infectious agent delivers a nucleic acid encoding a variant polypeptide comprising a secreted or transmembrane immunomodulatory protein, e.g., an ICOSL vIgD polypeptide, to an appropriate cell (e.g., an APC, e.g., a cell that presents a peptide / MHC complex on its cell surface, e.g., a dendritic cell) or tissue (e.g., lymphoid tissue), thereby achieving a desired effect, e.g., inducing and / or enhancing immunomodulatory and / or specific cell-mediated immune responses, e.g., CD4 and / or CD8 T cell responses, which may include cytotoxic T cell (CTL) responses. In some embodiments, the infectious agent targets an APC, e.g., a dendritic cell (DC). In some embodiments, the nucleic acid molecule delivered by the infectious agent described herein comprises appropriate nucleic acid sequences, e.g., regulatory elements, e.g., a promoter, necessary for expression in a particular target cell of an operably linked coding sequence encoding a variant immunomodulatory polypeptide.

[0329] In some embodiments, an infectious agent containing a nucleic acid sequence encoding an immunomodulatory polypeptide can also contain nucleic acid sequences encoding one or more additional gene products, e.g., cytokines, prodrug-converting enzymes, cytotoxins, and / or detectable gene products. For example, in some embodiments, the infectious agent is an oncolytic virus, and the virus can include nucleic acid sequences encoding additional therapeutic gene products (e.g., Kirn et al., (2009) Nat Rev Cancer 9:64-71; Garcia-Aragoncillo et al., (2010) Curr Opin Mol Ther 12:403-411; U.S. Patent Nos. 7,588,767, 7,588,771, 7,662,398, and 7,754,221, and U.S. Patent Application Publication Nos. 2007 / 0202572, 2007 / 0212727, 2010 / 0062016, 2009 / 0098529, 2009 / 0053244, See Nos. 2009 / 0155287, 2009 / 0117034, 2010 / 0233078, 2009 / 0162288, 2010 / 0196325, 2009 / 0136917, and 2011 / 0064650. In some embodiments, the additional gene product is capable of causing death of a target cell (e.g., a tumor cell). The gene product may be a therapeutic gene product capable of enhancing, strengthening, or modulating an immune response (e.g., a cytokine). Exemplary gene products also include anti-cancer agents, anti-metastatic agents, anti-angiogenic agents, immunomodulatory molecules, immune checkpoint inhibitors, antibodies, cytokines, growth factors, antigens, cytotoxic gene products, pro-apoptotic gene products, anti-apoptotic gene products, cell matrix degrading genes, genes for tissue regeneration, and genes for reprogramming human somatic cells to pluripotency, as well as other genes described herein or known to those skilled in the art. In some embodiments, the additional gene product is granulocyte-macrophage colony-stimulating factor (GM-CSF).

[0330] 1. Viruses In some embodiments, the infectious agent is a virus. In some embodiments, the infectious agent is an oncolytic virus or a virus that targets specific cells, such as immune cells. In some embodiments, the infectious agent targets tumor cells and / or cancer cells of the subject. In some embodiments, the infectious agent targets immune cells or antigen-presenting cells (APCs).

[0331] In some embodiments, the infectious agent is an oncolytic virus. An oncolytic virus is a virus that accumulates in tumor cells and replicates in tumor cells. Replication in tumor cells and optional delivery of nucleic acids encoding variant ICOSL polypeptides or immunomodulatory polypeptides described herein can lyse tumor cells, shrink tumors, and eliminate them. Oncolytic viruses can have a wide range of host and cell types. For example, oncolytic viruses accumulate in immune-privileged cells or immune-privileged tissues, including tumors and / or metastases, and also wound tissues and cells, thus allowing delivery and expression of nucleic acids encoding variant immunomodulatory polypeptides described herein in a wide range of cell types. Oncolytic viruses can also replicate specifically in tumor cells, resulting in tumor cell lysis and efficient tumor regression.

[0332] Exemplary oncolytic viruses include adenovirus, adeno-associated virus, herpes virus, herpes simplex virus, vesicular oral virus, reovirus, Newcastle disease virus, parvovirus, measles virus, vesicular stomatitis virus (VSV), coxsackievirus, and vaccinia virus. In some embodiments, oncolytic viruses can specifically colonize solid tumors without infecting other organs and can be used as infectious agents to deliver nucleic acids encoding the variant immunomodulatory polypeptides described herein to such solid tumors.

[0333] Oncolytic viruses for use in delivering nucleic acids encoding the variant ICOSL polypeptides or immunomodulatory polypeptides described herein can be any known to those skilled in the art, for example, vesicular stomatitis virus (VSV). virus) (see, e.g., U.S. Pat. Nos. 7,731,974, 7,153,510, 6,653,103, and U.S. Patent Application Publication Nos. 2010 / 0178684, 2010 / 0172877, 2010 / 0113567, 2007 / 0098743, 20050260601, 20050220818, and European Patent Nos. 1385466, 1606411, and 1520175); herpes simplex virus (see, e.g., U.S. Pat. Nos. 7,897,146, 7,731,952, 7,5 Nos. 50,296, 7,537,924, 6,723,316, and 6,428,968, and U.S. Patent Application Publication Nos. 2014 / 0154216, 2011 / 0177032, 2011 / 0158948, 2010 / 0092515, 2009 / 0274728, 2009 / 0285860, 2009 / 0215147, 2009 / 0010889, 2007 / 0110720, 2006 / 0039894, 2004 / 0009604, and 2004 / 0063094, and International Patent Publication WO 2007 / 052029, WO 1999 / 038955); retroviruses (see, e.g., U.S. Patent Nos. 6,689,871, 6,635,472, 5,851,529, 5,716,826, 5,716,613 and U.S. Patent Application Publication No. 20110212530); vaccinia viruses (see, e.g., U.S. Patent No. 2016 / 0339066), and adeno-associated viruses (see, e.g., U.S. Patent Nos. (See Patent Nos. 8,007,780, 7,968,340, 7,943,374, 7,906,111, 7,927,585, 7,811,814, 7,662,627, 7,241,447, 7,238,526, 7,172,893, 7,033,826, 7,001,765, 6,897,045, and 6,632,670).

[0334] Oncolytic viruses also include viruses that have been genetically modified to attenuate their pathogenicity, improve their safety profile, enhance their tumor specificity, and are equipped with additional genes, such as cytotoxins, cytokines, prodrug-converting enzymes, that also improve the overall efficacy of the virus (see, e.g., Kirn et al., (2009) Nat Rev Cancer 9:64-71; Garcia-Aragoncillo et al., (2010) Curr Opin Mol Ther 12:403-411; U.S. Patent Nos. 7,588,767, 7,588,771, 7,662,398, and 7,754,221, and U.S. Patent Application Publication Nos. 2007 / 0202572, 2007 / 0212727, 2010 / 0062016, 2009 / 00 (See, e.g., Nos. 98529, 2009 / 0053244, 2009 / 0155287, 2009 / 0117034, 2010 / 0233078, 2009 / 0162288, 2010 / 0196325, 2009 / 0136917, and 2011 / 0064650). In some embodiments, the oncolytic virus can be a virus that has been modified to selectively replicate in cancerous cells and is therefore oncolytic. For example, the oncolytic virus is an adenovirus that has been modified to have an altered tropism for tumor therapy and a...

Claims

(i) an extracellular domain (ECD) of ICOS ligand (ICOSL), having an amino acid sequence as set forth in SEQ ID NO:32, with one or two amino acid substitutions selected from N52H, N57Y, or Q100R; (ii) a portion of the ECD comprising an IgV domain having the amino acid sequence set forth in SEQ ID NO:196, with one or two amino acid substitutions selected from N52H, N57Y, or Q100R; (iii) one, two or three amino acid substitutions selected from N52H, N57Y, or Q100R, and M10V, M10I, V11E, S13G, E16V, A20V, S25G, F27S, F27C, N30D, Q37R, R61S, L74Q, D77G, N84Q, K92R, H94E, L96I, L98F, S99G, L102R, V107I, S109G, S109N, V110D, V110N, V110A, H115R, N119Q, F120I, S121G, V122A, S126T, H129P, S1 or the extracellular domain (ECD) of ICOS ligand (ICOSL), having the amino acid sequence as set forth in SEQ ID NO:32, with up to eleven amino acid substitutions consisting of one or more amino acid substitutions selected from the group consisting of 32F, E135K, F138L, C140del, S142F, I143V, I143T, N144D, Y146C, Y152C, W153R, I154F, K156M, L161P, L166Q, F172S, L173S, S192G, V193M, N194D, C198R, L203P, V210A, S212G, D217V, I218T, I218N, R221G, R221I, I224V, and T225A; or (iv) one, two or three amino acid substitutions selected from N52H, N57Y, or Q100R, as well as M10V, M10I, V11E, S13G, E16V, A20V, S25G, F27S, F27C, N30D, Q37R, R61S, L74Q, D77G, N84Q, K92R, H94E, L96I, L98F, S99G, L102R, V107I, S109G, S109N, V110D, V110N, V110A, H115R, N119Q, F120I, S121G, V122A, S126T, H129P, S1 32F, E135K, F138L, C140del, S142F, I143V, I143T, N144D, Y146C, Y152C, W153R, I154F, K156M, L161P, L166Q, F172S, L173S, S192G, V193M, N194D, C198R, L203P, V210A, S212G, D217V, I218T, I218N, R221G, R221I, I224V, and T225A, The variant ICOSL polypeptide specifically binds to the ectodomain of human CD28 with improved affinity compared to the binding of unmodified ICOSL to the ectodomain of human CD28.

2. The variant ICOSL polypeptide of claim 1, wherein the improved affinity for the ectodomain of human CD28 is greater than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 30-fold, 40-fold, 50-fold, or 60-fold improved compared to the unmodified ICOSL.

3. The variant ICOSL is N52Y / N57Y / F138L / L203P, N52H / N57Y / Q100P, N52H / C198R, N52H / C198R / T225A, N52H / K92R, N52H / S99G, N57Y / Q100P, N52Y / N57Y / Y152C, N52Y / N57Y / H129P / C198R, N52H / L161P / C198R, N52H / I143T, N52H / Q100R, N52H / N57Y / R61S / Q100R / V110D / L173S、 N52H / N57Y / Q100R / V122A、 E16V / N52H / N57Y / Q100R / V110D / H115R / Y152C / K156M / C198R、 Q37R / N52H / N57Y / Q100R / V110N / S142F / C198R / D217V / R221G、 N52H / N57Y / Q100R / V110D / C198R、 V11E / N30D / N52H / N57Y / H94E / L96I / L98F / N194D / V210A / I218T、 N52H / N57Y / H94E / L96I / F120I / S126T / W153R / I218N、 N52H / N57Y / Q100R / V110D / F172S / C198R、 S25G / F27C / N52H / N57Y / Q100R / V110D / E135K / L173S / C198R、 N52H / N57Y / V110A / C198R / R221I、 M10I / S13G / N52H / N57Y / D77G / V110A / H129P / I143V / F172S / V193M / C198R、 N52H / N57Y / Q100R / V110D / N144D / F172S / C198R、 N52S / H94E / L98F / Q100R、 N52H / N57Y / Q100R / V110D / C198R / S212G、 N52H / N57Y / Q100R / C198R、 N52H / N57Y / Q100R / L102R / V110D / H115R / C198R、 N52H / C140del / T225A、 N52H / N57Y / L74Q / V110D / S192G、 N52H / S121G / C198R、 N52H / N57Y / Q100R / V107I / V110D / I154F / C198R / R221G、 N52H / N84Q / N119Q、 N57Y / Q100R / C198R、 N57Y / F138L / L203P、 Q100R / F138L、 N52H / N57Y / Q100R / H115R / F172S / C198R、 N52H / N57Y / Q100R / H115R / I143V / F172S / C198R、 N52H / N57Y / Q100R / L102R / H115R / F172S / C198R、 N52H / N57Y / Q100R / H115R / F172S / N194D、 N52H / N57Y / H115R / F172S / C198R、 N52H / N57Y / Q100R / H115R / C198R、 N52H / N57Y / H115R, N52H / N57Y / Q100R / H115R, N52H / N57Y / Q100R / H115R / F172S / I224V, N52H / N57Y / Q100R / H115R / F172S, N52H / N57Y / Q100R / F172S, N52H / Q100R / H115R / I143T / F172S, N52H / N57Y / Q100P / H115R / F172S, N52Y / N57Y / Q100P / F172S, E16V / N52H / N57Y / Q100R / V110D / H115R / C198R, E16V / N52H / N57Y / Q100R / V110D / H115R / Y152C / K156M / F172S / C198R, N52H / N57Y / Q100P / C198R, N52H / N57Y / Q100P H115R / F172S / C198R, N52H / N57Y / Q100P / F172S / C198R, N52H / N57Y / Q100P / H115R, N52H / N57Y / Q100P / H115R / C198R, N52H / Q100R / C198R, N52H / Q100R / H115R / F172S, N52H / Q100R / H115R / F172S / C198R, or N52H / N57Y / Q100R / F172S / C198R 3. The variant ICOSL polypeptide of claim 1 or 2, comprising an amino acid substitution selected from among:

4. comprising an amino acid sequence as set forth in any one of SEQ ID NOs: 110, 112, 113, 115, 117-119, 121, 122, 126-128, 135, 285, 289, 290, 300-302, 308, 310, 315-318, 321, 322, 364, 365, 367, 372, 375, 376, 381, 414, 427, 430-432, 435, 437-439, 441-453, 461-467, 469, 470; A variant ICOSL polypeptide according to any one of claims 1 to 3.

5. 2. The variant ICOSL polypeptide of claim 1, which specifically binds to the ectodomain of human ICOS with improved affinity compared to the binding of unmodified ICOSL to the ectodomain of human ICOS.

6. The variant ICOSL polypeptide of claim 5, wherein the improved affinity for the ectodomain of human ICOS is greater than 1.2-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, 10-fold, 20-fold, 40-fold, 50-fold, 60-fold, or 70-fold improved compared to the unmodified ICOSL.

7. The variant ICOSL polypeptide of any one of claims 1 to 6, which is a soluble protein.

8. 8. The variant ICOSL polypeptide according to any one of claims 1 to 7, which is linked to a multimerisation domain.

9. The variant ICOSL polypeptide of claim 8, wherein the multimerization domain is an Fc domain or a variant thereof having reduced effector function.

10. A transmembrane domain linked to the extracellular domain (ECD) of said variant ICOSL, or a portion thereof, including an IgV domain having the amino acid sequence shown in SEQ ID NO:

196.

10. The variant ICOSL polypeptide of any one of claims 1 to 9, which is a transmembrane immunomodulatory protein further comprising:

11. The variant ICOSL polypeptide of claim 10, further comprising a cytoplasmic signaling domain linked to the transmembrane domain.

12. A variant ICOSL-Fc fusion protein comprising a variant ICOSL polypeptide according to any one of claims 1 to 11 and an Fc domain.

13. A conjugate comprising a variant ICOSL polypeptide according to any one of claims 1 to 11 or a variant ICOSL-Fc fusion protein according to claim 12 linked to a moiety.

14. 14. The conjugate of claim 13, wherein the moiety is a targeting moiety that specifically binds to a molecule on the surface of a cell, optionally the cell being an immune cell or a tumor cell.

15. 15. The conjugate of claim 13 or claim 14, wherein the moiety is an antibody or an antigen-binding fragment.

16. The conjugate of any one of claims 13 to 15, which is a fusion protein.

17. A nucleic acid molecule encoding a variant ICOSL polypeptide according to any one of claims 1 to 11 or a variant ICOSL-Fc fusion protein according to claim 12.

18. A vector comprising the nucleic acid molecule of claim 17.

19. 19. A cell comprising the nucleic acid of claim 17 or the vector of claim 18.

20. A method for producing a variant ICOSL polypeptide or an immunomodulatory protein, comprising the step of introducing a nucleic acid molecule described in claim 17 or a vector described in claim 18 into a host cell under conditions such that the protein is expressed in the host cell.

21. The method of claim 20, further comprising the step of isolating or purifying the variant ICOSL polypeptide or immunomodulatory protein from the cells.

22. A method for modifying a cell to express a variant ICOSL polypeptide, comprising the step of introducing into a host cell a nucleic acid molecule encoding a variant ICOSL polypeptide according to any one of claims 1 to 11 or a variant ICOSL-Fc fusion protein according to claim 12 under conditions in which the polypeptide is expressed in the host cell.

23. A modified cell comprising a variant ICOSL polypeptide according to any one of claims 1 to 11, a variant ICOSL-Fc fusion protein according to claim 12, a nucleic acid molecule according to claim 17, or a vector according to claim 18.

24. The modified cell of claim 23 , wherein the variant ICOSL polypeptide or immunomodulatory polypeptide does not contain a transmembrane domain and / or is not expressed on the surface of the cell.

25. The modified cell of claim 23 or claim 24, wherein the variant ICOSL polypeptide or immunomodulatory polypeptide is secreted from the modified cell.

26. 24. The modified cell of claim 23, comprising a variant ICOSL polypeptide that comprises a transmembrane domain and / or is a transmembrane immunomodulatory protein of claim 10 or claim 11.

27. The modified cell of claim 26, wherein the variant ICOSL polypeptide is expressed on the surface of the cell.

28. The modified cell of any one of claims 23 to 27, wherein the cell is an immune cell.

29. The modified cell of any one of claims 23 to 28, further comprising a chimeric antigen receptor (CAR) or a modified T cell receptor.

30. 13. An infectious agent comprising a nucleic acid molecule encoding a variant ICOSL polypeptide according to any one of claims 1 to 11 or a variant ICOSL-Fc fusion protein according to claim 12.

31. 31. The infectious agent of claim 30, which is a bacterium or a virus.

32. A pharmaceutical composition comprising a variant ICOSL polypeptide according to any one of claims 1 to 11, a variant ICOSL-Fc fusion protein according to claim 12, a conjugate according to any one of claims 13 to 16, a modified cell according to any one of claims 23 to 29, or an infectious agent according to claim 30 or claim 31, and a pharma- ceutical acceptable excipient.

33. 33. An article of manufacture comprising the pharmaceutical composition of claim 32 in a vial.

34. 34. A kit comprising the pharmaceutical composition of claim 32, the article of manufacture of claim 33, and instructions for use.

35. 33. Use of the pharmaceutical composition of claim 32 in the manufacture of a medicament for use in modulating an immune response in a mammalian subject.

36. 33. The pharmaceutical composition of claim 32 for use in modulating an immune response in a mammalian subject.

37. 37. The use of claim 35 or the pharmaceutical composition for use of claim 36, wherein modulating the immune response treats a disease or condition in a subject.

38. 38. The use of claim 35 or claim 37 or the pharmaceutical composition for the use of claim 36, wherein the immune response is reduced.

39. 39. The use or pharmaceutical composition for use according to any one of claims 35 to 38, wherein said disease or condition is an inflammatory or autoimmune disease or condition.

40. 40. The use or pharmaceutical composition for use according to any one of claims 35 to 39, wherein the disease or condition is selected from the group consisting of antineutrophil cytoplasmic antibody (ANCA) associated vasculitis, vasculitis, autoimmune skin diseases, transplantation, rheumatic diseases, inflammatory gastrointestinal diseases, inflammatory eye diseases, inflammatory neurological diseases, inflammatory lung diseases, inflammatory endocrine diseases, and autoimmune blood diseases.

41. The disease or condition is Addison's disease, allergy, alopecia areata, Alzheimer's disease, antineutrophil cytoplasmic antibody (ANCA) associated vasculitis, ankylosing spondylitis, antiphospholipid syndrome (Hughes syndrome), asthma, atherosclerosis, rheumatoid arthritis, autoimmune hemolytic anemia, autoimmune hepatitis, autoimmune inner ear disease, autoimmune lymphoproliferative syndrome, autoimmune myocarditis, autoimmune oophoritis, autoimmune orchitis, azoospermia, Behcet's disease, Berger's disease, bullous pemphigoid, cardiomyopathy, cardiovascular disease, celiac sprue / celiac disease, chronic fatigue immune dysregulation syndrome (CFIDS), chronic idiopathic polyneuropathy, chronic inflammatory demyelinating polyradiculoneuropathy (CIPD), chronic relapsing polyneuropathy (Guillain-Barre syndrome), chalazioni syndrome, bronchitis ... Schizophrenia, Chronic Obstructive Pulmonary Disease (COPD), CREST syndrome, Crohn's disease, Dermatitis, Dermatitis Herpetiformis, Dermatomyositis, Diabetes, Discoid lupus, Eczema, Epidermolysis bullosa acquisita, Essential mixed cryoglobulinemia, Evans syndrome, Exophthalmos, Fibromyalgia, Goodpasture's syndrome, Graves' disease, Hashimoto's thyroiditis, Idiopathic pulmonary fibrosis, Idiopathic thrombocytopenic purpura (ITP), IgA nephropathy, Immunoproliferative diseases or disorders, Inflammatory bowel disease (IBD), Interstitial lung disease, Juvenile arthritis, Juvenile idiopathic arthritis (JIA), Kawasaki disease, Lambert-Eaton myasthenic syndrome, Lichen planus, Lupus nephritis, Lymphocytic hypophysitis lypophisitis), Meniere's disease, Miller Fisher syndrome / acute disseminated encephalomyeloradiculopathy, mixed connective tissue disease, multiple sclerosis (MS), muscular rheumatism, myalgic encephalomyelitis (ME), myasthenia gravis, eye inflammation, pemphigus foliaceus, pemphigus vulgaris, pernicious anaemia, polyarteritis nodosa, polychondritis, polyglandular syndrome40. The use or pharmaceutical composition for use according to any one of claims 35 to 39, wherein the inflammatory bowel disease is selected from the group consisting of: rheumatic syndromes (Whittaker syndrome), polymyalgia rheumatica, polymyositis, primary agammaglobulinemia, primary biliary cirrhosis / autoimmune cholangiopathy, psoriasis, psoriatic arthritis, Raynaud's phenomenon, Reiter's syndrome / reactive arthritis, restenosis, rheumatic fever, rheumatic diseases, sarcoidosis, Schmidt's syndrome, scleroderma, Sjogren's syndrome, stiff man syndrome, systemic lupus erythematosus (SLE), systemic scleroderma, Takayasu's arteritis, temporal arteritis / giant cell arteritis, thyroiditis, type 1 diabetes, ulcerative colitis, uveitis, vasculitis, vitiligo, interstitial bowel disease and Wegener's granulomatosis.

Images