WO2015066262A1 - Methods for preventing toxicity of adoptive cell therapy - Google Patents

Abstract

Methods for preventing or reducing toxicity of adoptive cell therapy using a population of cells deficient in the expression or activity of GM-CSF, GM-CSF receptor or perforin are provided, as are kits for preparing cells deficient in the expression or activity of GM-CSF, GM-CSF receptor or perforin.

Description

METHODS FOR PREVENTING TOXICITY OF ADOPTIVE CELL THERAPY Introduction

[0001] This application claims the benefit of priority of U.S. Provisional Application No. 61/899,431, filed November 4, 2013, the content of which is incorporated herein by reference in its entirety.

[0002] This invention was made with government support under Grant No. R44HL099217 awarded by the National Heart, Lung, and Blood Institute. The government has certain rights in the invention.

Background.

[0003] T cells play important roles in anti-tumor immunity. Genetic modification of T cells with tumor- targeting chimeric antigen receptors (CARs) and adoptive transfer of CAR-modified T cells have emerged as a promising way of treating cancers (Rossig & Brenner (2004) Mol . Ther . 10:5- 18; Ho, et al . (2003) Cancer Cell 3:431-437; Porter, et al . (2011) N. Engl. J. Med. 365:725-733; Kalos, et al . (2011) Sci. Tranal. Med. 3:95ra73; Park, et al . (2011) Trends Biotechnol. 29:550-7) . One advantage of this strategy over other approaches is the ability to expand large number of tumor-specific T cells (>10¹⁰ cells) in a relatively short time (<4 weeks) (Sadelain, et al . (2009) Curr. Opin. Immunol. 21:215-223; Sadelain, et al . (2003) Nat. Rev. Cancer 3:35-45) . In addition, the functional activities of CAR-modified T cells, such as production of Thl cytokines, cytotoxicity and in vivo persistence can be enhanced by integrating the signaling domains of co-stimulatory molecules (such as CD28, 4-1BB and OX40) to CARs (Porter, et al. (2011) supra; Park, et al . (2011) supra) . CARs have been derived from single chain antibody fragment (scFv) against antigens on tumor cells, such as CD19 in B cell lymphoma (Porter, et al . (2011) supra; Carpenito, et al . (2009) Proc. Natl. Acad. Sci . USA 106:3360-65).

[ 0004 ] Natural killer (NK) cells attack tumor and virally- infected cells in the absence of major histocompatibility complex (MHC) restriction, utilizing a combination of signals from activating and inhibitory receptors (Lanier (2005) Annu. Rev. Immunol. 23:225-74) . One group of activating NK cell receptors are natural cytotoxicity receptors (NCRs) , which includes NKp46 (NCR1 and CD335) , NKp44 (NCR2 and CD336) and NKp30 (NCR3 and CD337) (Lanier (2005) supra; Moretta, et al . (2006) Semin. Immunol. 18:151-8; Vivier, et al . (2011) Science 331:44-49). These receptors are expressed on NK cells, some subsets of γδ T cells, and they play important roles in NK-mediated tumor cell-killing (Lanier (2005) supra; Correia, et al . (2011) Blood 118:992-1001). NKp30 is a type I transmembrane protein that contains a single Ig-like extracellular domain followed by a short stalk region connected to a transmembrane segment and intracellular domain (Pende, et al. (1999) J. Exp. Med. 190:1505-16). In mice, NKp30 is not expressed since it is a pseudogene (Hollyoake, et al . (2005) Mol. Biol. Evol . 22 :1661-72) . When expressed on NK cells, NKp30 receptor as a monomer associates with 0ϋ3ζ and FcRy for signal transduction (Pende, et al . (1999) supra; Delahaye, et al . (2011) Nat. Med. 17:700-707). Two cellular ligands for NKp30 receptor have been identified: BAT3 and B7-H6 (Brandt, et al . (2009) J. Exp. Med. 206:1495-1503; Pogge von Strandmann, et al . (2007) Immunity 27:965-74). BA 3 is a nuclear protein, which is involved in the interaction with P53 and induction of apoptosis after stress such as DNA damage (Pogge von Strandmann, et al . (2007) supra; Sasaki, et al . (2007) Genes Dev. 21:848-861). BAT3 can also be released by immature dendritic cells (iDCs) on the surfaces of exosomes to stimulate NK cells (Simhadri, et al . (2008) PLoS One 3:e3377). B7-H6 is a newly identified B7 family member. Unlike BAT3 , B7-H6 is expressed on the surface of tumor cells, but not most normal cells (Brandt, et al . (2009) supra) . Recently, the structure of NKp30 receptor in complex with B7-H6 has been deciphered (Joyce, et al . (2011) Proc . Natl. Acad. Sci . USA 108:6223-6228; Li, et al . (2011) J. Exp. Med. 108:703-14) . NKp30 uses both front and back β-sheets to engage the Ig- like V region of B7-H6 via predominantly hydrophobic interactions (Joyce, et al . (2011) supra; Li, et al . (2011) supra) . NKp30 receptor has been shown to be important in mediating anti -tumor effects in gastrointestinal stromal tumors and lymphoid leukemia (Correia, et al . (2011) supra; Delahaye, et al . (2011) supra). In this respect, CARs based upon NKp30, which contain the CD28 and/or 003ζ signaling domains, have been shown to be effective in adoptive immunotherapy against B7-H6⁺ tumor cells in vivo (Zhang, et al . (2012) J. Immunol. 189:2290-99).

Summary of the Invention

[ 00 05 ] The present invention provides methods for preventing toxicity of adoptive cell therapy by administering to a subject in need of adoptive cell therapy, a population of cells deficient in the expression or activity of GM-CSF or perforin. In some embodiments, toxicity includes a cytokine storm or cytotoxicity toward healthy cells. In other embodiments, the adoptive cell therapy is chimeric antigen receptor-bearing lymphocyte therapy, T cell therapy, natural killer cell therapy, gamma/delta T cell therapy, or natural killer T cell therapy. In accordance with embodiments directed to a population of GM-CSF-deficient cells, said population of cells can be the result a GM-CSF gene deletion or GM-CSF gene disruption; contact with a microRNA, siRNA, shRNA, or antisense molecule that inhibits the expression of GM-CSF; or alternatively, contact with an antibody that blocks that activity of GM-CSF or GM-CSF receptor. In a similar manner, a population of perforin-deficient cells can be the result of a perforin gene deletion or perforin gene disruption; or contact with a microRNA, siRNA, shRNA, or antisense molecule that inhibits the expression of perforin. Kits including reagents for adoptive cell therapy and the preparation of a population of cells deficient in GM-CSF or perforin are also provided.

Brief Description of the Drawings

[0006] Figures 1A-1B provide data showing the changes in serum cytokines 18 hours after infusion of CAR cells from B6 mice (CH) or from GM-CSF-deficient mice. Control T cells that do not cause illness are shown (WT) . The GM-CSF- deficient CAR cells had many reduced cytokines.

[0007] Figures 2A-2B provide data showing the changes in serum cytokines 18 hours after infusion of CAR cells from B6 mice (CH) or from perforin-deficient mice (pfp) . Control T cells that do not cause illness are shown (WT) . The perforin-deficient CAR cells had many reduced cytokines.

[0008] Figure 3 provides data showing the acute weight changes associated with illness. There are changes in the chimeric (CAR) group from B6 mice (CH, wild-type) and the IFNy-deficient mice, but not from the GM-CSF-deficient or perforin-deficient CAR donors. The wild-type control (WT) received control T cells that did not cause illness.

[0009] Figure 4 provides data showing the "health score" of mice treated with control T cells (WT) or chNKG2D T cells from B6 (CH) , IFNy-deficient , GM-CSF-deficien , or perforin-deficient mice. A score of 1 is healthy and a score of 4 is dead. This analysis indicated that when the chNKG2D T cells were from mice deficient in either GM-CSF or perforin, there was no adverse affect on health score.

[0010] Figure 5 illustrates the structure of the LentiCRISPR vector and target sgRNAs (LGH-1, SEQ ID NO : 1 and 2; LGH-2, SEQ ID NO : 3 and 4; LGH-3, SEQ ID NO : 5 and 6; LGH-4, SEQ ID NO : 7 and 8) used to eliminate human GM-CSF. LTR, long-terminal repeat; psi+, packaging signal; RRE, rev response element; cPPT, central polypurine tract; sgRNA, single guide RNA; EFS, elongation factor- la short promoter; SpCas9, Streptococcus pyogenes Cas9; P2A, 2A self - cleaving peptide; Puro, puromycin selection marker; and WPRE , posttranscriptional regulatory element.

Detailed Description of the Invention

[0011] It has now been found that blockade of GM-CSF or perforin prevents toxicity of adoptive cell therapies. Specifically, it has now been shown that when infusing a high dose of perforin-deficient or GM-CSF-deficient chimeric antigen receptor (CAR) T cells, serum cytokine levels are reduced (Figures 1 and 2) and acute weight changes (Figure 3) and toxicity (Figure 4) are not observed when compared to wild-type or interferon-gamma-deficient CAR T cells in mice. Moreover, this result is specific for perforin and GM-CSF and not observed with other effector molecules or cytokines. The present invention can prevent a cytokine storm that may occur when a high dose of activated lymphocytes or CAR-bearing lymphocytes are given to a patient .

[0012] Given that immune cells lacking functional GM-CSF or perforin result in much less toxicity, a large cell dose can be given without resulting in acute toxicity. Accordingly, the present invention provides methods for preventing or reducing toxicity of adoptive cell therapy by administering to a subject in need of adoptive cell therapy a population of cells deficient in the expression or activity of G -CSF or perforin. In accordance with the present invention, toxicity is intended to include a cytokine storm or cytotoxicity toward healthy cells, which results from administration of immune cells during adoptive cell therapy. As is known in the art, a cytokine storm, cytokine cascade or hypercytokinemia is an immune reaction, which results in hyperrelease of inflammatory mediators, in particular cytokines, in response to stimulation of T cells and macrophages. In one embodiment, cells deficient in the expression or activity of GM-CSF can result in a measurable decrease (e.g., a 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80% or 90% decrease) in the level of cytokines produced in response to adoptive cell therapy. Similarly, cells deficient in the expression or activity of perforin can result in a measurable decrease (e.g., a 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80% or 90% decrease) in the number of healthy cells killed in response to adoptive cell therapy.

[ 0 013 ] Adoptive immunotherapy refers to the autologous, syngeneic or allogeneic transfer of immune cells capable of mediating an immune response. Thus, the present invention has a wide range of applications, including pre-exposure vaccination of individuals with both in vivo- and in vitro- primed T cells, treatment of cancer subjects using tumor- targeted T cell immunotherapy, treatment of bone marrow transplant subjects (for whom opportunistic infections, such as CMV, are problematic and yet amenable to treatment with targeted T cells such as CMV-targeted cytotoxic lymphocytes) , enhancement of conventional vaccination efficacy through T cell adjuvant therapy, treatment of outbreaks of emergent or re-emergent pathogens, etc. The source of the immune cells that are deficient in or lack functional GM-CSF or perforin may be of any kind. In some embodiments, immune cells are obtained from a bank of umbilical cord blood, peripheral blood, human embryonic stem cells, draining lymph nodes, or induced pluripotent stem cells, for example. Immune cells of the invention include, but are not limited to T- lymphocytes (T-cells) , naive T cells (T_N) , memory T cells (for example, central memory T cells (TC_M) , effector memory cells (T_EM) ) _/ natural killer cells, hematopoietic stem cells and/or pluripotent embryonic/induced stem cells capable of giving rise to therapeutically relevant progeny. By way of example, individual immune cells of the invention may be CD4+/CD8-, CD4-/CD8+, CD4-/CD8- or CD4+/CD8+. The immune cells may be a mixed population of CD4+/CD8- and CD4-/CD8+ cells or a population of a single clone. CD4+ immune cells of the invention may produce IL-2, ΙΡΝγ, TNFa and other effector cytokines when co-cultured in vitro with cells expressing the target antigens (for example CD20+ and/or CD 19+ tumor cells) . CD8⁺ immune cells of the invention may lyse antigen- specific target cells when co-cultured in vitro with the target cells. In some embodiments, immune cells may be any one or more of CD45 A⁺ CD62L⁺ naive cells, CD45RO CD62I7 central memory cells, CD62LT effector memory cells or a combination thereof (Berger, et al . (2009) Curr. Opin. Immunol. 21(2)224-232).

[ 0014 ] As indicated, immune cells of the invention are deficient in GM-CSF or perforin expression or activity, or lack functional GM-CSF or perforin proteins. An immune cell deficient in GM-CSF or perforin expression or activity or lacking functional GM-CSF or perforin protein refers to a cell that does not express or expresses significantly reduced levels {e.g., less than 30, 20, 10, or 5% of normal levels) of G -CSP or perforin; a cell in which the activity of GM-CSF or perforin protein has been eliminated or significantly reduced (e.g., less than 30, 20, 10, or 5% of normal activity) ; or a cell in which the expression or activity of the GM-CSF receptor has been eliminated or significantly reduced.

[ 0015 ] As used herein, "Granulocyte Macrophage-Colony Stimulating Factor" (GM-CSF) refers to a small naturally occurring glycoprotein with internal disulfide bonds having a molecular weight of approximately 23 kDa . In humans, it is encoded by a gene located within the cytokine cluster on human chromosome 5. The sequence of the human gene and protein are known and available under GENBANK Accession Nos. M11220 and AAA52578, respectively. See, also, Lee et al . (1985) Proc. Natl. Acad. Sci . USA 82:4360-4364. The protein has an N- terminal signal sequence, and a C-terminal receptor binding domain (Rasko & Gough (1994) In: The Cytokine Handbook, Thomson, et al . Academic Press, New York, pages 349-369) . Its three-dimensional structure is similar to that of the interleukins , although the amino acid sequences are not similar. GM-CSF is produced in response to a number of inflammatory mediators present in the hemopoietic environment and at peripheral sites of inflammation. GM-CSF is able to stimulate the production of neutrophilic granulocytes, macrophages, and mixed granulocyte -macrophage colonies from bone marrow cells and can stimulate the formation of eosinophil colonies from fetal liver progenitor cells. GM-CSF can also stimulate some functional activities in mature granulocytes and macrophages and inhibits apoptosis of granulocytes and macrophages . 2014/063037

[ 0016 ] "Granulocyte macrophage-colony stimulating factor receptor" (GM-CSFR) refers to a membrane bound receptor expressed on cells, which transduces a signal when bound to GM-CSF. The GM-CSF receptor is a member of the haematopoietin receptor superfamily. It is heterodimeric protein composed of an alpha and a beta subunit . The alpha subunit is highly specific for GM-CSF whereas the beta subunit is shared with other cytokine receptors, including IL3 and IL5. This is reflected in a broader tissue distribution of the beta receptor subunit. The alpha subunit, GM-CSFRa, is primarily expressed on myeloid cells and non-haematopoetic cells, such as neutrophils, macrophages, eosinophils, dendritic cells, endothelial cells and respiratory epithelial cells. Full length GM- CSFRa is a 400 amino acid type I membrane glycoprotein that belongs to the type I cytokine receptor family, and is composed of a 22 amino acid signal peptide (positions 1- 22), a 298 amino acid extracellular domain (positions 23- 320) , a transmembrane domain from positions 321-345 and a short 55 amino acid intracellular domain. The signal peptide is cleaved to provide the mature form of GM-CSFRa as a 378 amino acid protein. The sequences of nucleic acids encoding the human GM-CSFR alpha and beta proteins are known and available under GENBA K Accession Nos . M73832 and M59941, respectively. Likewise, the sequences of the human GM-CSFR alpha and beta proteins are known and available under GENBANK Accession Nos. AAA35909 and AAA18171, respectively. See, also, Raines, et al . (1991) Proc . Natl. Acad. Sci. USA 88:8203-8207. GM-CSF is able to bind with relatively low affinity to the a subunit alone (K_d 1-5 nM) but not at all to the β subunit alone. However, the presence of both a and β subunits results in a high affinity ligand-receptor complex (Ka »100 pM) . GM-CSF signaling occurs through its initial binding to the GM- CSFRa chain and then cross - 1 inking with a larger subunit the common β chain to generate the high affinity interaction, which phosphorylates the JAK-STAT pathway.

[ 0017 ] "Perforin" or "pore-forming protein (pfp) " is a cytolytic protein found in the granules of Cytotoxic T lymphocytes (CTLs) and NK cells. Upon degranulation, perforin inserts itself into the target cell's plasma membrane, forming a pore. The sequence of the human gene and protein are known and available under GENBANK Accession Wos. NM_001083116 and NP_001076585 , respectively. The perforin gene has been mapped to chromosome 17 in humans (Shinkai_; et al . (1989) Immunogenetics 30:452-457). It was found that exon 1 encodes an untranslated sequence, and the entire protein is encoded by a portion of exon 2 and all of exon 3, which also contains a 3' untranslated region. The cloning of perforin cDNA encoding human (Lichtenheld & Podack (1989) J^". Immunol. 143:4267-4274) perforin indicates that human perforin is 534 amino acids in length and contains 20 cysteine residues, which are believed to form 10 intra-chain disulphide bonds. The amino terminal 100 residues and the carboxy terminal 150 residues are completely unique to perforin.

[ 0018 ] Cells deficient in or lacking functional GM-CSF or perforin can be prepared using any conventional method. In some embodiments, GM-CSF or perforin expression is inhibited or blocked by, e.g., gene deletion, gene disruption, si NA, shRNA or antisense approaches. In other embodiments, GM-CSF or perforin activity is inhibited or blocked by, e.g., a GM-CSF or perforin antagonist or antibody; or in the case of GM-CSF, a GM-CSF receptor antagonist or antibody. [ 0019 ] In certain embodiments, the expression of endogenous GM-CSF or perforin is blocked by genetically modifying the immune cell. Although in some cases homologous recombination is used, in particular cases non-homologous end joining is used to edit the genome. Any suitable protocol to modify the genome of a particular immune cell is useful, although in specific embodiments gene modification is achieved using an engineered nuclease such as a zinc finger nuclease (ZFP) , TALE-nuclease (TALEN) , or CRISPR/Cas nuclease. Engineered nuclease technology is based on the engineering of naturally occurring DNA-binding proteins. For example, engineering of homing endonucleases with tailored DNA-binding specificities has been described, (see, Chames, et al . (2005) Nucleic Acids Res. 33(20) :el78; Arnould, et al . (2006) J. Mol . Biol. 355:443-458). In addition, engineering of ZFPs has also been described. See, e.g., US 6,534,261; US 6,607,882; US 6,824,978; US 6,979,539; US 6,933,1 13; US 7,163,824; and US 7,013,219.

[ 0020 ] In addition, ZFPs and TALEs have been fused to nuclease domains to create ZFNs and TALENs ; functional entities that are able to recognize their intended nucleic acid target through their engineered (ZFP or TALE) DNA binding domains and cause the DNA to be cut near the ZFP or TALE DNA binding site via the nuclease activity. See, e.g., Kim et al. (1996) Proc. Natl. Acad. Sci . USA 93 (3):1156-1 160. More recently, ZFNs have been used for genome modification in a variety of organisms. See, for example, US 2003/0232410; US 2005/0208489; US 2005/0026157; US 2005/0064474; US 2006/0188987; US 2006/0063231; and WO 07/014275.

[0021] The present invention may involve any nuclease of interest. Non-limiting examples of nucleases include meganucleases, TALENs and zinc finger nucleases. The nuclease may include heterologous DNA-binding and cleavage domains {e.g., zinc finger nucleases; TALENs ; meganuclease DNA-binding domains with heterologous cleavage domains) or, alternatively, the DNA-binding domain of a naturally- occurring nuclease may be altered to bind to a selected target site (e.g., a meganuclease that has been engineered to bind to site different than the cognate binding site) .

[0022] In certain embodiments, the nuclease is a meganuclease (homing endonuclease) . Naturally-occurring meganucleases recognize 15-40 base-pair cleavage sites and are commonly grouped into families: the LAGLIDADG family, the GIY-YIG family, the His-Cyst box family, the HNH family, the PD-(D/E)xK family and the Vsr-like family. Exemplary homing endonucleases include I-Scel, I-Ceul, PI- Sce, I-SceW, I-Csml, I-Panl, I-SceII, I-Ppol, I-SceIII, I- Crel, I-Tevl, I-TevII and I-TevIII, the recognition sequences of which are known in the art. See, US 5,420,032; US 6,833,252; Belfort, et al . (1997) Nucleic Acids Res. 25:3379-3388; Dujon, et al . (1989) Gene 82:115-1 18; Perler, et^■ al . (1994) Nucleic Acids Res. 22, 1125-1127; Jasin (1996) Trends Genet. 12:224-228; Gimble, et al . (1996) J. Mol . Biol. 263:163-180; Argast, et al . (1998) J. Mol. Biol. 280:345-353.

[ 0023 ] ZFNs include a zinc finger protein that has been engineered to bind to a target site in a gene of choice and cleavage domain or a cleavage half -domain. As described above, zinc finger binding domains can be engineered to bind to a sequence of choice. See, for example, Beerli, et al . (2002) Nature Biotechnol . 20:135-141; Pabo, et al .

(2001) Ann. Rev. Biochem. 70:313-340; Isalan et al . (2001) Nature Biotechnol. 19 -.656-660 ; Segal, et al . (2001) Curr. Opin. Biotechnol. 12:632-637; Choo, et al . (2000) Curr. Opin. Struct. Biol. 10:411-416. An engineered zinc finger binding domain can have a novel binding specificity, compared to a naturally-occurring zinc finger protein. Engineering methods include, but are not limited to, rational design and various types of selection. Rational design includes, for example, using databases including triplet (or quadruplet) nucleotide sequences and individual zinc finger amino acid sequences, in which each triplet or quadruplet nucleotide sequence is associated with one or more amino acid sequences of zinc fingers which bind the particular triplet or quadruplet sequence. See, for example, US 6,453,242 and US 6,534,261.

[ 0024 ] Exemplary selection methods including phage display and two-hybrid systems are disclosed in US 5,789,538; US 5,925,523; US 6,007,988; US 6,013,453; US 6,410,248; US 6,140,466; US 6,200,759; and US 6,242,568; as well as WO 98/37186; WO 98/53057; WO 00/27878; WO 01/88197 and GB 2,338,237. In addition, enhancement of binding specificity for zinc finger binding domains has been described, for example, in WO 02/077227.

[ 0025 ] In addition, zinc finger domains and/or multi- fingered zinc finger proteins may be linked together using any suitable linker sequences. See, e.g., US 6,479,626; US 6,903,185; and US 7,153,949 for exemplary linker sequences 6 or more amino acids in length.

[ 0026 ] The CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats) /Cas (CRISPR Associated) nuclease system is an engineered nuclease system based on a bacterial system that can be used for genome engineering. It is based on part of the adaptive immune response of many bacteria and archea . When a virus or plasmid invades a bacterium, segments of the invader's DNA are converted into CRISPR RNAs (crRNA) by the 'immune' response. This crRNA then associates, through a region of partial complementarity, with another type of RNA called tracrRNA to guide the Cas9 nuclease to a region homologous to the crRNA in the target DNA called a "protospacer . " Cas9 cleaves the DNA to generate blunt ends at the double- stranded break at sites specified by a 20-nucleotide guide sequence contained within the crRNA transcript. Cas9 requires both the crRNA and the tracrRNA for site-specific DNA recognition and cleavage. This system has now been engineered such that the crRNA and tracrRNA can be combined into one molecule (the "single guide RNA"), and the crRNA equivalent portion of the single guide RNA can be engineered to guide the Cas9 nuclease to target any desired sequence (see, Jinek et al . (2012) Science 337:816-821; Jinek et al . (2013) eLife 2:e00471; Segal (2013) eLife 2:e00563). Thus, the CRISPR/Cas system can be engineered to create a double-stranded break at a desired target in a genome, and repair of the double-stranded break can be influenced by the use of repair inhibitors to cause an increase in error prone repair.

[ 0027 ] In some embodiments, the DNA binding domain is an engineered domain from a TAL effector similar to those derived from the plant pathogens Xanthomonas (see Boch, et al . (2009) Science 326:1509-1512; Moscou & Bogdanove (2009) Science 326:1501) and Ralstonia (see Heuer, et al (2007) Appl. Environ. Microbiol. 73 (13) : 4379-4384 ; WO

2010/079430) .

[ 0028 ] Nucleases (e.g., ZFNs or TALENs) can be screened for activity prior to use, for example in a yeast -based chromosomal system as described in WO 2009/042163 and WO 2009/0068164. Nuclease expression constructs can be readily designed using methods known in the art. See, e.g., US 2003/0232410; US 2005/0208489; US 2005/0026157; US 2005/0064474; US 2006/0188987; US 2006/0063231 and WO 07/014275. Expression of the nuclease may be under the control of a constitutive promoter or an inducible promoter, for example the galactokinase promoter which is activated (de- repressed) in the presence of raffinose and/or galactose and repressed in presence of glucose.

[ 002 9 ] The engineered nucleases (e.g., ZFNs, TALENs , CRISPR/Cas) or polynucleotides encoding the same may be delivered to a target cell by any suitable means including, for example, by injection of ZFN, TALEN or CRISPR/Cas proteins or by use of ZFN, TALEN or CRISPR/Cas encoding mRNA. Methods of delivering proteins such as zinc finger proteins are described, for example, in US 6,453,242; US 6,503,717; US 6,534,261; US 6,599,692; US 6,607,882; US 6,689,558; US 6,824,978; US 6,933,113; US 6,979,539; US 7,013,219; and US 7,163,824. ZFNs, TALENs, and CRISPR/Cas may also be delivered using vectors containing sequences encoding one or more of the ZFNs, TALENs, and CRISPR/Cas protein (s) . Any vector systems may be used including, but not limited to, plasmid vectors, retroviral vectors, lentiviral vectors, adenovirus vectors, poxvirus vectors, herpesvirus vectors and adeno-associated virus vectors, etc. See, e.g., US 6,534,261; US 6,607,882; US 6,824,978; US 6,933,113; US 6,979,539; US 7,013,219; and US 7,163,824. Furthermore, it will be apparent that any of these vectors may include one or more zinc finger or TALEN protein- encoding sequences. Thus, when one or more ZFNs, TALENs or CRISPR/Cas proteins are introduced into the cell, the sequences encoding the ZFNs, TALENs or CRISPR/Cas proteins may be carried on the same vector or on different vectors. When multiple vectors are used, each vector may include a sequence encoding one or multiple ZFPs, TALENs or CRISPR/Cas systems. [ 0030 ] Conventional viral and non-viral based gene transfer methods can be used to introduce nucleic acids encoding engineered ZFPs, TALENs or CRISPR/Cas systems into immune cells in vitro. Non-viral vector delivery systems include DNA plasmids, naked nucleic acid, and nucleic acid complexed with a delivery vehicle such as a liposome or poloxamer, lipofection or electroporation . Viral vector delivery systems include DNA and RNA viruses, which have either episomal or integrated genomes after delivery to the cell. For a review of such delivery systems, see, e.g., Anderson (1992) Science 256:808-813; Nabel & Feigner (1993) TIBTECH 11:211-217; Mitani & Caskey (1993) TIBTECH 11:162- 166; Dillon (1993) TIBTECH 11:167-175; Miller (1992) Nature 357:455-460; Van Brunt (1988) Biotechnology 6(10):1149- 1154; and Yu, et al . (1994) Gene Therapy 1:13-26. In embodiments, a lentivirus is used.

[ 003 1] Alternatively, the expression of GM-CSF or perforin in immune cells is eliminated or reduced post- transcriptionally, e.g., with an antisense, microRNA, siRNA or shRNA molecule. In certain embodiments, the expression of GM-CSF or. perforin is reduced by at least 70, 80, 90, 95 or 99% as compared to the expression of GM-CSF or perforin in immune cells not contacted with the above-referenced molecule .

[ 0032 ] As used herein, the term "antisense" refers to a nucleotide sequence that is complementary to a nucleic acid encoding GM-CSF or perforin, e.g., complementary to the coding strand of the double-stranded cDNA molecule or complementary to the mRNA sequence encoding GM-CSF or perforin. The antisense nucleic acid may be complementary to an entire GM-CSF or perforin coding strand, or to only a portion thereof. Alternatively, the antisense molecule is antisense to a "non-coding region" of the coding strand (e.g., the 5' and 3' untranslated regions) . For example, the antisense molecule can be complementary to the region surrounding the translation start site of GM-CSF or perforin mRNA and can be, for example, about 7, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, or more nucleotides in length. An antisense molecule can be constructed using chemical synthesis and enzymatic ligation reactions using procedures known in the art. For example, an antisense molecule can be chemically synthesized using naturally occurring nucleotides or variously modified nucleotides designed to increase the biological stability of the molecule or to increase the physical stability of the duplex formed between the antisense and sense nucleic acids, e.g., phosphorothioate derivatives and acridine substituted nucleotides can be used. The molecule also can be produced biologically using an expression vector. In accordance with this embodiment, the antisense molecule can be produced in vitro and subsequently contacted with an immune cell, or the immune cell can be transformed or transduced with the expression vector such that the antisense molecule is produced in vivo.

[ 0033 ] In a further embodiment of the present invention, GM-CSF or perforin short interfering nucleic acid molecules

(siRNA) that inhibit the expression of the target gene can also be used to reduce the level of target gene expression. The term "short interfering RNA" or "siRNA" as used herein, refers to any nucleic acid molecule capable of inhibiting or down-regulating GM-CSF or perforin gene expression, for example by mediating RNA interference ("RNAi") or gene silencing in a sequence-specific manner. Chemical modifications can also be applied to any siRNA sequence of the present invention. For example, the siRNA can be a double-stranded molecule including self-complementary sense 3037

and antisense regions. The siRNA can be assembled from two separate oligonucleotides, where one strand is the sense strand and the other is the antisense strand, wherein the antisense and sense strands are self -complementary (i.e. each strand includes a nucleotide sequence that is complementary to nucleotide sequence in the other strand; such as where the antisense strand and sense strand form a duplex or double stranded structure, for example, wherein the double stranded region is about 19 base pairs) . Alternatively, the siRNA is assembled from a single oligonucleotide, where the self -complementary sense and antisense regions of the siRNA are linked by means of a nucleic acid based or non-nucleic acid-based linker. In yet another embodiment, the siRNA can be a circular single- stranded polynucleotide having two or more loop structures and a stem including self-complementary sense and antisense regions, wherein the circular polynucleotide can be processed either in vivo or in vitro to generate an active siRNA molecule capable of mediating RNAi . As used herein, siRNA molecules need not be limited to those molecules containing only RNA, but further encompasses molecules including chemically-modified nucleotides or those in combination with non-nucleotides . In certain preferred embodiments, the siRNA molecule of the invention lacks 2'- hydroxy (2 ' -OH) containing nucleotides. Such siRNA molecules can, however, have an attached linker or linkers or other attached or associated groups, moieties, or chains containing one or more nucleotides with 2¹ -OH groups. Optionally, siRNA molecules of the invention can include ribonucleotides at about 5, 10, 20, 30, 40, or 50% of the nucleotide positions. The modified siRNA molecules of the invention can also be referred to as short interfering modified oligonucleotides "siMON." As used herein, the term 2014/063037

si NA is preferably meant to be equivalent to other terms used to describe nucleic acid molecules that are capable of mediating sequence specific RNAi , for example double- stranded RNA (dsRNA) , micro-RNA (miRNA) , short hairpin RNA (shRNA) , short interfering oligonucleotide, short interfering nucleic acid, short interfering modified oligonucleotide, chemically-modified siRNA, post- transcriptional gene silencing RNA (ptgsRNA) , translational silencing, and others. In addition, as used herein, the term RNAi is preferably meant to be equivalent to other terms used to describe sequence specific RNA interference, such as post -transcriptional gene silencing, or epigenetics. For example, siRNA molecules of the invention can be used to epigenetically silence genes at the post- transcriptional level or the pre-transcriptional level. In a non-limiting example, epigenetic regulation of GM-CSF or perforin gene expression by siRNA molecules of the invention can result from siRNA-mediated modification of the chromatin structure to alter GM-CSF or perforin gene expression. siRNA molecules are known in the art and available from commercial sources such as Santa Cruz Biotech. By way of illustration, the siRNA molecule 5'- AAGCCCACCCAGAGAAGTGTT-3 ' (SEQ ID NO:l) has been shown to decrease the expression of perforin (Zheng, et al . (2007) Blood 109:2049-2057).

[ 0 034 ] In yet another embodiment, the antisense molecule of the present invention is an oi-anomeric nucleic acid molecule. An a-anoraeric nucleic acid molecule forms specific double-stranded hybrids with complementary RNA in which, contrary to the usual a-units, the strands run parallel to each other (Gaultier, et al . (1987) Nucleic Acids, Res. 15:6625-6641). The antisense nucleic acid molecule can also include a 2 ' -o-methylribonucleotide 14 063037

(Inoue, et al . (1987) Nucleic Acids Res. 15:6131-6148) or a chimeric RNA-DNA analogue (Inoue, et al . (1987) FEBS Lett. 215 : 327-330) .

[ 0035 ] In still another embodiment, an antisense molecule of the invention is a ribozyme . A ribozyme having specificity for GM-CSF- or perforin-encoding nucleic acid molecules can include one or more sequences complementary to the nucleotide sequence of GM-CSF or perforin cDNA disclosed herein, and a sequence having known catalytic sequence responsible for mRNA cleavage (see US 5,093,246; Haselhoff & Gerlach (1988) Nature 334:585-591). For example, a derivative of a Tetrahymena L-19 IVS RNA can be constructed in which the nucleotide sequence of the active site is complementary to the nucleotide sequence to be cleaved in a GM-CSF- or perforin-encoding mRNA (see, e.g., US 4,987,071 or US 5,116,742). Alternatively, GM-CSF or perforin mRNA can be used to select a catalytic RNA having a specific ribonuclease activity from a pool of RNA molecules .

[ 003 6 ] In a further embodiment, GM-CSF or perforin expression can be inhibited by targeting nucleotide sequences complementary to the regulatory region of the GM- CSF or perforin (e.g., a GM-CSF or perforin promoter and/or enhancers) to form triple helical structures that prevent transcription of the GM-CSF or perforin gene in target cells (see generally, Helene (1991) Anticancer Drug Des. 6 (6) .-569-84; Helene, et al . (1992) Ann. N.Y. Acad. Sci . 660:27-36; Maher (1992) Bioassays 14 (12) : 807-15) . The potential sequences that can be targeted for triple helix formation can be increased by creating a so-called "switchback" nucleic acid molecule. Switchback molecules are synthesized in an alternating 5' -3', 3' -5' manner, such that they base pair with first one strand of a duplex and then the other, eliminating the necessity for a sizeable stretch of either purines or pyrimidines to be present on one strand of a duplex.

[ 0037 ] The antisense molecules may also be modified at the base moiety, sugar moiety or phosphate backbone to improve, e.g., the stability, hybridization, or solubility of the molecule. For example, the deoxyribose phosphate backbone of the nucleic acid molecule can be modified to generate peptide nucleic acids (see Hyrup, et al . (1996) Bioorg. Med. Chem. 4:5-23) . As used herein, the terms "peptide nucleic acid" or "PNA" refers to a nucleic acid mimic, e.g., a DNA mimic, in which the deoxyribose phosphate backbone is replaced by a pseudopeptide backbone and only the four natural nucleobases are retained. The neutral backbone of a PNA can allow for specific hybridization to DNA and RNA under conditions of low ionic strength. The synthesis of PNA oligomers can be performed using standard solid phase peptide synthesis protocols as described in Hyrup et al . (1996) supra; Perry-0 ' Keefe , et al . (1996) Proc. Natl. Acad. Sci. USA 93:14670-675.

[ 0038 ] The antisense and short interfering RNA molecules of the invention can be directly introduced into an immune cell or generated in situ such that they hybridize with or bind to cellular mRNA and/or genomic DNA encoding GM-CSF or perforin to thereby inhibit expression of said GM-CSF or perforin, e.g., by inhibiting transcription and/or translation. The molecules can be delivered to cells using vectors, or by viral mechanisms (such as retroviral or adenoviral infection delivery) . To achieve sufficient intracellular concentrations of the molecules, vector constructs in which the molecule is placed under the control of an appropriate promoter. [0039] The immune cells of the invention can be modified with, e.g., siRNA, miRNA, shRNA, ribozyme or antisense, by ex vivo treatment with the appropriate molecule to effect inhibition, reduction or blockade of GM-CSF or perforin expression. By way of illustration, immune cells such as primary human T cells (e.g., from human peripheral blood mononuclear cells (PBMC) , bone marrow, or umbilical cord blood) are obtained and a viral or non-viral based nucleic acid transfer method (e.g., as described herein) is used to introduce the siRNA, miRNA, shRNA, ribozyme or antisense into immune cells in vitro. Stably transfected immune cells can be selected by measuring the expression of GM-CSF or perforin. Immune cells with a significantly reduced level of GM-CSF or perforin as compared to a normal immune cell are considered deficient in GM-CSF or perforin.

[0040] As an alternative to gene disruption, gene deletion or post-transcriptional gene silencing, the present invention also includes the use of GM-CSF or perforin antagonists. An antagonist of GM-CSF or perforin can inhibit one or more of the activities of the naturally occurring form of the polypeptide by, for example, competitively inhibiting GM-CSF- or perforin-mediated activity. As used herein, the term "biological activity" with reference to perforin refers to the cytolytic activity of a perforin polypeptide; that is, its ability to bind to a target cell membrane and polymerize into pore-like transmembrane channels leading to cell lysis. The activity of perforin also includes the capacity to synergize with other toxins such as granule toxins and other molecules to induce apoptosis . The target cell can be any cell that is capable of being lysed by native perforin. The term "biological activity" with reference to GM-CSF refers to the stimulatory activity of GM-CSF toward bone marrow cells U 2014/063037

and fetal liver progenitor cells, as well as its functional activities in mature granulocytes and macrophages.

[ 0041] The biological activity of perforin can be assessed by the skilled artisan by any number of means known in the art including, but not limited to, the measurement of target cell lysis, the delivery of granzyme B molecules into the target cell, the measurement of target cell membrane disruption (such as by changes in ion transport) , the induction of apoptosis in the target cell, the modification of vesicular trafficking and the general assessment of target cell death. The target cell may be a red blood cell (RBC) and hence a common means of measuring perforin activity is by a RBC lysis test. It may also be any nucleated cell. Likewise, the biological activity of GM-CSF can be assessed using conventional methods such as measuring bone marrow cell production of, e.g., neutrophilic granulocytes or macrophages.

[ 0042 ] Perforin activity can be inhibited by perforin antagonists such as small organic molecules, polypeptides or antagonistic antibodies. For example, benzylidene-2- thioxoimidazolidinone compounds have been shown to selectively inhibit the activity of perforin. See EP 2515903. Furthermore, soluble forms of a perforin molecule capable of binding in competition with endogenous perforin may be used. Such competitors include fragments of the perforin polypeptide that are able to bind native perforin to inhibit its biological activity, but have no inherent perforin activity of their own. See, e.g., WO 2005/083098. A perforin antagonist may also include antibodies or antigen-binding fragments thereof (including, for example, polyclonal, monoclonal, humanized, anti- idiotypic , chimeric or single chain antibodies, and Fab, F(ab')₂ and Fab expression library fragments, scFV molecules, and epitope- 4 063037

binding fragments thereof) , or perforin fragments or other small molecules that bind to a native perforin polypeptide and inhibit the biological activity of perforin.

[ 0043 ] GM-CSF activity can be inhibited by GM-CSF antagonists such as antibodies specifically binding to GM- CSF, peptides, or small organic molecules specific for GM- CSF. Also within the meaning of the term GM-CSF antagonist are antibodies specifically binding to the GM-CSF receptor or small organic molecules specific for the GM-CSF receptor. The term GM-CSF antagonists also refers to non- antibody scaffold molecules, such as fibronectin scaffolds, ankyrins, maxybodies/avimers , protein A-derived molecules, anticalins, affilins, protein epitope mimetics (PEMs) or the like. See, US 6,818,418; US 7,115,396; US 5,770,380; Beste, et al . (1999) Proc . Natl. Acad. Sci. USA 96(5):1898- 1903; and Murali, et al . (2003) Cell Mol . Biol. 49 (2) :209- 216) .

[ 0044 ] GM-CSF antagonists selectively interfere with the induction of signaling by the GM-CSF receptor by causing a reduction in the binding of GM-CSF to the receptor. Such antagonists may include antibodies that bind the GM-CSF receptor a or β subunit, antibodies that bind to GM-CSF, GM-CSF analogs such as E21R, and other proteins or small molecules that compete for binding of GM-CSF to its receptor or inhibit signaling that normally results from the binding of the ligand to the receptor. An example of a neutralizing anti -GM-CSF antibody is the E10 antibody described in Li, et al . (2006) Proc. Natl. Acad. Sci. USA 103 (10) : 3557-3562. E10 is an IgG class antibody that has an 870 pM binding affinity for GM-CSF. The antibody is specific for binding to human GM-CSF as shown in an ELISA assay, and shows strong neutralizing activity as assessed with a TF1 cell proliferation assay. An additional exemplary neutralizing anti -GM-CSF antibody is the MT203- antibody described by Krinner, et al . (2007) Mol . Immunol. 44:916-25. MT203 is an IgGl class antibody that binds GM- CSF with picomolar affinity. The antibody shows potent inhibitory activity as assessed by TF-1 cell proliferation assay and its ability to block IL-8 production in U937 cells. Additional GM-CSF antibodies are described, e.g., in WO 2006/122797. MOR04357 (Morphosys, Martinsried/Planegg, Germany; Steidl, et al . (2008) Mol. Immunol. 46 (1) : 135-44) may also be used as the GM-CSF antagonist.

[ 0045 ] Other proteins that may interfere with the productive interaction of GM-CSF with its receptor include mutant GM-CSF proteins, GM-CSF peptides and secreted proteins including at least part of the extracellular portion of one or both of the GM-CSF receptor chains that bind to GM-CSF and compete with binding to cell -surface receptor. For example, a soluble GM-CSF receptor antagonist can be prepared by fusing the coding region of the sGM- CSFR with the CH2-CH3 regions of murine IgG2a. An exemplary soluble GM-CSF receptor is described by Raines, et al. (1991) Proc . Natl. Acad. Sci . USA 88:8203. An example of a GM-CSFRoi-Fc fusion protein is provided, e.g., in Brown, et al . (1995) Blood 85:1488. In addition, a peptide corresponding to residues 17-31 of GM-CSF (the A helix of GM-CSF) has been shown to inhibit high affinity receptor binding, while a second peptide corresponding to residues 54-78 of GM-CSF (the B and C helices of GM-CSF) has been shown to inhibit low affinity receptor binding (VonFeldt, et al (1995) Peptide Res . 8:20-27, 30-32).

[ 0046 ] Other GM-CSF antagonists include GM-CSF mutants. For example, GM-CSF having a mutation of amino acid residue 21 of GM-CSF to Arginine or Lysine (E21R or E21K) described by Hercus, et al . (1994) Proc. Natl. Acad. Sci USA 91:5838, has been shown to have in vivo activity in preventing dissemination of GM-CSF-dependent leukemia cells in mouse xenograft models (Iversen, et al . (1997) Blood 90:4910).

[ 0047 ] In some embodiments, the GM-CSF antagonist may be a peptide. For example, A GM-CSF peptide antagonist may be a peptide designed to structurally mimic the positions of specific residues on the B and C helices of human GM-CSF that are implicated in receptor binding and bioactivity

(e.g., Monfardini, et al . (199S) J. Biol. Chew. 271:2966- 2971) .

[ 0048 ] Any antibody specific for GM-CSF receptor may be used with the present invention. Such GM-CSF antagonists include antibodies to the GM-CSF receptor alpha chain or beta chain. The GM-CSF receptor antibody for use in the invention may be to the alpha chain. An anti-GM-CSF receptor antibody employed in the invention can be in any antibody format as explained herein, e.g., intact, chimeric, monoclonal, polyclonal, antibody fragment, humanized, humaneered, and the like. Examples of anti-GM- CSF receptor antibodies, e.g., neutralizing, high-affinity antibodies, suitable for use in the invention are known

(see, e.g., US 5,747,032; Nicola, et al . (1993) Blood 82 : 1724) .

[ 0049 ] Neutralizing antibodies and other GM-CSF antagonists may be identified using any number of assays that assess GM-CSF function. For example, cell -based assays for GM-CSF receptor signaling, such as assays which determine the rate of proliferation of a GM-CSF-dependent cell line in response to a limiting amount of GM-CSF, are conveniently used. The human TF-1 cell line is suitable for use in such an assay. In some embodiments, the neutralizing antibodies of the invention inhibit GM-CSF-stimulated TF-1 cell proliferation by at least 50% when a GM-CSF concentration is used which stimulates 90% maximal TF-1 cell proliferation. In other embodiments, the neutralizing antibodies inhibit GM-CSF stimulated proliferation by at least 90%. Thus, typically, a neutralizing antibody, or other GM-CSF antagonist for use in the invention, has an EC₅₀ of less than 10 nM. Additional assays suitable for use in identifying neutralizing antibodies suitable for use with the present invention will be well known to persons of skill in the art.

[ 0050 ] Methods of preparing polyclonal antibodies are known to the skilled artisan. Polyclonal antibodies can be raised in a mammal by one or more injections of an immunizing agent and, if desired, an adjuvant. The immunizing agent includes perforin, GM-CSF or GM-CSF receptor protein, e.g., a human perforin, GM-CSF or GM-CSF receptor protein, or fragment thereof. In some embodiment, an antibody for use in the invention is purified from human plasma. In such embodiments, the antibody is typically a polyclonal antibody that is isolated from other antibodies present in human plasma. Such an isolation procedure can be performed, e.g., using known techniques, such as affinity chromatography .

[ 0051] In some embodiments, an antibody of the invention is a monoclonal antibody. Monoclonal antibodies may be prepared using hybridoma methods, such as those described by Kohler & Milstein (1975) Nature 256:495. In a hybridoma method, a mouse, hamster, or other appropriate host animal, is typically immunized with an immunizing agent, such as human perforin, GM-CSF or GM-CSF receptor protein, to elicit lymphocytes that produce or are capable of producing antibodies that will specifically bind to the immunizing agent. Alternatively, the lymphocytes may be immunized in vitro. The immunizing agent preferably includes human perforin, GM-CSF or GM-CSF receptor protein, fragments thereof, or fusion protein thereof.

[ 0052 ] Human monoclonal antibodies can be produced using various techniques known in the art, including phage display libraries (Hoogenboom & Winter (1991) J. Mol. Biol. 227:381; Marks, et al . (1991) J. Mol. Biol. 222:581). Similarly, human antibodies can be made by introducing human immunoglobulin loci into transgenic animals, e.g., mice in which the endogenous immunoglobulin genes have been partially or completely inactivated. Upon challenge, human antibody production is observed, which closely resembles that seen in humans in all respects, including gene rearrangement, assembly, and antibody repertoire. This approach is described, e.g., in US 5,545,807; US 5,545,806; US 5,569,825; US 5,625,126; US 5,633,425; US 5,661,016.

[ 0053 ] In some embodiments the antibodies of the invention are chimeric or humanized monoclonal antibodies. Humanized forms of antibodies are chimeric immunoglobulins in which residues from a complementary determining region (CDR) of human antibody are replaced by residues from a CDR of a non-human species such as mouse, rat or rabbit having the desired specificity, affinity and capacity.

[ 0054 ] An antibody that is employed in the invention can be in any format. For example, in some embodiments, the antibody can be a complete antibody including a constant region, e.g., a human constant region, or can be a fragment or derivative of a complete antibody, e.g. , an Fd, a Fab, Fab', F(ab')₂, a scFv, an Fv fragment, or a single domain antibody, such as a nanobody or a camel id antibody. Such antibodies may additionally be recombinantly engineered by methods well known to persons of skill in the art. As noted above, such antibodies can be produced using known techniques . [ 0055 ] In some embodiments, the antibody is a humaneered antibody. A humaneered antibody is an engineered human antibody having a binding specificity of a reference antibody, obtained by joining a DNA sequence encoding a binding specificity determinant (BSD) from the CDR3 region of the heavy chain of the reference antibody to human VH segment sequence and a light chain CDR3BSD from the reference antibody to a human VL segment sequence. Methods for humaneering are provided in US 2005/0255552 and US 2006/0134098. Methods for signal-less secretion of antibody fragments from E. coli are described in US 2007/0020685.

[ 0056 ] In some embodiments, the antibodies suitable for use with the present invention have a high affinity binding for human perforin, GM-CSF or GM-CSF receptor protein. High affinity binding between an antibody and an antigen exists if the dissociation constant (K_D) of the antibody is <about 10 nM, typically <1 nM, and preferably <100 pM. In some embodiments, the antibody has a dissociation rate of about 10^"4 per second or better.

[ 0057 ] A variety of methods can be used to determine the binding affinity of an antibody for its target antigen such as surface plasmon resonance assays, saturation assays, or immunoassays such as ELISA or RIA, as are well known to persons of skill in the art. An exemplary method for determining binding affinity is by surface plasmon resonance analysis on a BIACORE 2000 instrument (Biacore AB, Freiburg, Germany) using CM5 sensor chips, as described by Krinner, et al . (2007) Mol . Immunol. 44 (5) : 916-25.

[0058 ] Immune cells deficient in perforin or GM-CSF expression or activity are of use in reducing the toxicity associated with a number of adoptive immunotherapies including, but not limited to, chimeric antigen receptor- bearing lymphocyte therapy, T cell therapy, natural killer cell therapy, gamma/delta T cell therapy, or natural killer T cell therapy, in particular in the treatment of cancer, an infectious disease or other conditions related to immune-dysfunction, e.g. inflammatory diseases, neuronal disorders, diabetes, and cardiovascular disorders.

[ 0059 ] As used herein, chimeric antigen receptor (CAR) - bearing lymphocyte therapy includes the use of CAR-modified T cells to recognize tumor cells via binding of the CAR to its tumor-associated antigen, independent of T cell receptor-MHC/peptide interactions. As a result T cells are activated and can efficiently eliminate tumor cells. "Chimeric antigen receptor" or "CAR" as used herein refers to engineered receptors, which graft an antigen specificity onto cells (for example T cells such as naive T cells, central memory T cells, effector memory T cells or combination thereof) . CARs are also known as artificial T- cell receptors, chimeric T-cell receptors or chimeric immunoreceptors . Antigens specific for cancer which may be targeted by the CARs (for example bispecific CARs) include, but are not limited to, any one or more of 4-IBB, 5T4, adenocarcinoma antigen, alpha- fetoprotein, BAFF, B-lymphoma cell, C242 antigen, CA-125, carbonic anhydrase 9 (CA-IX) , C-MET, CCR4, CD152, CD19, CD20, CD200, CD22, CD221, CD23 (IgE receptor), CD28, CD30 (TNFRSF8) , CD33, CD4 , CD40, CD51, CD52, CD56, CD74, CD80, CEA, CNT0888, CTLA-4, DR5 , EGFR, EpCAM, CD3 , FAP, fibronectin extra domain-B, folate receptor 1, GD2 , GD3 ganglioside, glycoprotein 75, GPNMB, HER2/neu, HGF, human scatter factor receptor kinase, IGF-I receptor, IGF-I, IgGl , LI-CAM, IL-13, IL-6, insulin-like growth factor I receptor, integrin Οίδβΐ, integrin νβ3, MORAb-009, MS4A1, MUC1, N-glycolylneuraminic acid, NPC-1C, PDGF-Ra, PDL192, phosphatidylserine, prostatic carcinoma cells, RANKL, RON, ROR1 , SCH 900105, SDC1 , SLAMF7, TAG-72, tenascin C, TGF-β, TRAIL-R1, TRAIL- R2 , tumor antigen CTAA16.88, VEGF-A, VEGFR-1, VEGFR2 or vimentin. Examples of CARs which target the above antigens include but are not limited to bispecific CARs, bispecific CARs co-expressed with EGFRt, bispecific CARs co-expressed with EGFRt and CCR, bispecific CARs co-expressed with EGFRt and DHFR (for example mutant DHFR) or bispecific CARs co-expressed with EGFRt and CDR and DHFR (for example mutant DHFR) . In some embodiments, the bispecific chimeric antigen receptors target and bind at least two different antigens. Examples of pairings of at least two antigens bound by the bispecific CARs of the invention include but are not limited to CD19 and CD20, CD19 and CD22, CD20 and LI -CAM, LI -CAM and GD2 , EGFR and LI -CAM, EGFR and C-MET, EGFR and HER2, C-MET and HER2 and EGFR and R0R1.

[ 0 06 0 ] Antigens specific for inflammatory diseases which may be targeted by the CARs of the invention include but are not limited to any one or more of AOC3 (VAP-1) , CAM- 3001, CCL11 (eotaxin-1) , CD125, CD147 (basigin) , CD154 (CD40L) , CD2, CD20, CD23 (IgE receptor), CD25 (a chain of IL-2 receptor), CD3 , CD4 , CD5 , IF -θί , IFN-γ, IgE, IgE Fc region, IL-1, IL-12, IL-23, IL-13, IL-17, IL-17A, IL-22, IL-4, IL-5, IL-5, IL-6, IL-6 receptor, integrin a4 , integrin 4β7 LFA-1 (CDlla) , MEDI-528, myostatin, OX-40, rhuMAb β7, scleroscin, SOST, TGFfil , TNF- or VEGF-A. Antigens specific for neuronal disorders which may be targeted by CARs include but are not limited to any one or more of beta amyloid or MABT5102A. Antigens specific for diabetes which may be targeted by CARs include but are not limited to any one or more of L-Ιβ or CD3. Antigens specific for cardiovascular diseases which may be targeted by CARs include but are not limited to any one or more of 3037

C5 , cardiac myosin, CD41 (integrin alpha-lib), fibrin II, beta chain, ITGB2 (CD18) and sphingosine-l-phosphate .

[ 0061] Antigens specific for infectious diseases which may be targeted by CARs include but are not limited to any one or more of anthrax toxin, CCR5 , CD4 , clumping factor A, cytomegalovirus, cytomegalovirus glycoprotein B, endotoxin, Escherichia coli, hepatitis B surface antigen, hepatitis B virus, HIV-1, Hsp90, Influenza A hemagglutinin, lipoteichoic acid, Pseudomonas aeruginosa, rabies virus glycoprotein, respiratory syncytial virus and TNF-a.

[ 0062 ] Further examples of target antigens include but are not limited to surface proteins found on cancer cells in a specific or amplified fashion (e.g. , the IL- 14 receptor, CD19, CD20 and CD40 for B-cell lymphoma, the Lewis Y and CEA antigens for a variety of carcinomas, the Tag72 antigen for breast and colorectal cancer, EGF-R for lung cancer, folate binding protein and the HER-2 protein which is often amplified in human breast and ovarian carcinomas) , or viral proteins (e.g., gpl20 and gp41 envelope proteins of HIV, envelope proteins from the Hepatitis B and C viruses, the glycoprotein B and other envelope glycoproteins of human cytomegalovirus, the envelope proteins from oncoviruses such as Kaposi's sarcoma-associated Herpes virus). Other potential targets of the CARs include CD4 , where the ligand is the HIV gpl20 envelope glycoprotein, and other viral receptors, for example ICAM, which is the receptor for the human rhinovirus, and the related receptor molecule for poliovirus. Additional targets of CARs include antigens involved in B-cell associated diseases.

[0063 ] CARs may also include a co-stimulatory domain. This domain may enhance cell proliferation, cell survival and development of memory cells. Co-stimulatory domains include, for example, members of the TNFR super family, CD28, CD137 (4-1BB), CD134 (OX40), DaplO, CD27, CD2 , CD5 , ICAM-1, LFA-1, Lck, TNFR-1, TNFR-II, Fas, CD30, CD40 or combinations thereof. If a CAR includes more than one co- stimulatory domain, these domains may be arranged in tandem, optionally separated by a linker.

[ 0 064 ] Nucleic acids encoding CARs can be introduced into immune cells as naked DNA or in a suitable vector. In addition to CAR- encoding nucleic acids, the naked DNA or vector can also include nucleic acids encoding one or more engineered nucleases for genetic modification of the immune cells (as described herein) or nucleic acids encoding, e.g., antisense RNA, siRNA, shRNA, or miRNA to block post- transcriptional expression of GM-CSF, G -CSF receptor, or perforin. Methods of stably transfecting immune cells by electroporation using naked DNA are known in the art. See, e.g., US 6,410,319. Naked DNA generally refers to the DNA encoding a chimeric receptor contained in a plasmid expression vector in proper orientation for expression. Advantageously, the use of naked DNA reduces the time required to produce immune cells expressing the chimeric receptor .

[ 0 065 ] Alternatively, a viral vector (e.g., a retroviral vector, adenoviral vector, adeno-associated viral vector, or lentiviral vector) can be used to nucleic acids into immune cells. Suitable vectors include, e.g., non- replicating vectors. A large number of vectors are known which are based on viruses, where the copy number of the virus maintained in the cell is low enough to maintain the viability of the cell. Illustrative vectors include the pFB-neo vectors (STRATAGENE^®) disclosed herein as well as vectors based on HIV, SV40, EBV, HSV or BPV.

[ 0 06 6 ] Once it is established that the immune cells are deficient in perforin or GM-CSF expression or activity and optionally transfected or transduced with a CAR, the immune cells are introduced or administered to the subject to treat cancer, an infectious disease or other conditions related to immune-dysfunction, e.g., inflammatory diseases, neuronal disorders, diabetes, and cardiovascular disorders. To facilitate administration, the immune cells according to the invention can be made into a pharmaceutical composition or made implant -appropriate for administration in vivo, with appropriate carriers or diluents, which further can be pharmaceutically acceptable. The means of making such a composition or an implant have been described in the art (see, for instance, Remington's Pharmaceutical Sciences, 16th Ed., Mack, ed. (1980)). Where appropriate, the immune cells can be formulated into a preparation in semisolid or liquid form, such as a capsule, solution, injection, inhalant, or aerosol, in the usual ways for their respective route of administration. Means known in the art can be utilized to prevent or minimize release and absorption of the composition until it reaches the target tissue or organ, or to ensure timed-orelease of the composition. Desirably, however, a pharmaceutically acceptable form is employed which does not ineffectuate the immune cells. Thus, desirably the immune cells can be made into a pharmaceutical composition containing a balanced salt solution, preferably Hanks' balanced salt solution, or normal saline.

[ 0 067 ] A pharmaceutical composition of the present invention can be used alone or in combination with other well-established agents useful for treating, e.g., cancer or an infectious disease. Whether delivered alone or in combination with other agents, the pharmaceutical composition of the present invention can be delivered via various routes and to various sites in a mammalian, particularly human, body to achieve a particular effect. One skilled in the art will recognize that, although more than one route can be used for administration, a particular route can provide a more immediate and more effective reaction than another route. For example, intradermal delivery may be advantageously used over inhalation for the treatment of melanoma. Local- or systemic delivery can be accomplished by administration comprising application or instillation of the formulation into body cavities, inhalation or insufflation of an aerosol, or by parenteral introduction, comprising intramuscular, intravenous, intraportal, intrahepatic, peritoneal, subcutaneous, or intradermal administration.

[ 006 8 ] A composition of the present invention can be provided in unit dosage form wherein each dosage unit, e.g., an injection, contains a predetermined amount of the composition, alone or in appropriate combination with other active agents. The term unit dosage form as used herein refers to physically discrete units suitable as unitary dosages for human and animal subjects, each unit containing a predetermined quantity of the composition of the present invention, alone or in combination with other active agents, calculated in an amount sufficient to produce the desired effect, in association with a pharmaceutically acceptable diluent, carrier, or vehicle, where appropriate. The specifications for the novel unit dosage forms of the present invention depend on the particular pharmacodynamics associated with the pharmaceutical composition in the particular subject.

[ 0069 ] Desirably an effective amount or sufficient number of immune cells is present in the composition and introduced into a subject such that long-term, specific response is established to, e.g., reduce the size of a tumor, eliminate tumor growth or regrowth, or eliminate an infection than would otherwise result in the absence of such treatment. Desirably, the amount of immune cells introduced into the subject causes a measurable decrease in the signs or symptoms of the disease being treated when compared to otherwise same conditions wherein the immune cells are not present.

[ 0070 ] The amount of immune cells administered should take into account the route of administration and should be such that a sufficient number of the immune cells will be introduced so as to achieve the desired therapeutic response. Furthermore, the amounts of each active agent included in the compositions described herein (e.g., the amount per each cell to be contacted or the amount per certain body weight) can vary in different applications. In general, the concentration of immune cells desirably should be sufficient to provide in the subject being treated at least from about 1x10^s to about lxlO⁹ immune cells, even more desirably, from about 1x10⁷ to about 5x10⁸ immune cells, although any suitable amount can be utilized either above, e.g., greater than 5xl0⁸ cells, or below, e.g., less than lxlO⁷ cells. The dosing schedule can be based on well- established cell-based therapies (see, e.g., Topalian and Rosenberg (1987) Acta Haematol. 78 Suppl 1:75-6; US 4,690,915) or an alternate continuous infusion strategy can be employed.

[ 0071 ] These values provide general guidance of the range of immune cells to be utilized by the practitioner upon optimizing the methods of the present invention for practice of the invention. The recitation herein of such ranges by no means precludes the use of a higher or lower amount of a component, as might be warranted in a particular application. For example, the actual dose and TU 2014/063037

schedule can vary depending on whether the compositions are administered in combination with other pharmaceutical compositions, or depending on interindividual differences in pharmacokinetics, drug disposition, and metabolism. One skilled in the art readily can make any necessary adjustments in accordance with the exigencies of the particular situation.

[0072] The immune cells of the invention can be used in adoptive cell therapies for the treatment of a variety of cancers . Forms of cancer that may be treated in accordance with this invention include squamous cell carcinoma, leukemia, acute lymphocytic leukemia, acute lymphoblastic leukemia, B-cell lymphoma, T-cell lymphoma, Hodgkins lymphoma, non-Hodgkins lymphoma, hairy cell lymphoma, Burketts lymphoma, acute or chronic myelogenous leukemias, promyelocytic leukemia, fibrosarcoma, rhabdomyoscarcoma; melanoma, seminoma, teratocarcinoma, neuroblastoma, glioma, astrocytoma, neuroblastoma, glioma, schwannomas; fibrosarcoma, rhabdomyoscaroma , osteosarcoma, melanoma, xeroderma pigmentosum, keratoacanthoma, seminoma, thyroid follicular cancer, and teratocarcinoma. Immune cells also can be useful in the treatment of other carcinomas of the bladder, breast, colon, kidney, liver, lung, ovary, prostate, pancreas, stomach, cervix, thyroid or skin. Immune cells also may be useful in treatment of other hematopoietic tumors of lymphoid lineage, other hematopoietic tumors of myeloid lineage, other tumors of mesenchymal origin, other tumors of the central or peripheral nervous system, and/or other tumors of mesenchymal origin. Advantageously, the immune cells of the invention also may be useful in reducing cancer progression in prostate cancer cells, melanoma cells (e.g., cutaneous melanoma cells, ocular melanoma cells, and/or lymph node- associated melanoma cells) , breast cancer cells, colon cancer cells, and lung cancer cells. The immune cells of the invention can be used to treat both tumorigenic and non- tumorigenic cancers {e.g., non-tumor- forming hematopoietic cancers) . The immune cells of the invention are particularly useful in the treatment of epithelial cancers (e.g., carcinomas) and/or colorectal cancers, breast cancers, lung cancers, vaginal cancers, cervical cancers, and/or squamous cell carcinomas (e.g., of the head and neck) . Additional potential targets include sarcomas and lymphomas. Additional advantageous targets include solid tumors and/or disseminated tumors (e.g., myeloid and lymphoid tumors, which can be acute or chronic) .

[0073] In addition to cancer treatment, the immune cells of the invention are also of use in adoptive cell therapies for the treatment of an infectious disease caused by bacteria, protozoa, fungi or viruses. In particular embodiments, a viral infection is treated. Viral infections that can be treated include, but are not limited to, hepatitis type A, hepatitis type B, hepatitis type C, influenza, varicella, adenovirus, herpes simplex type I (HSV-1) , herpes simplex type 2 (HSV-2), rinderpest, rhinovirus, echovirus, rotavirus, respiratory syncytial virus, papilloma virus, papilloma virus, cytomegalovirus (CMV, e.g., HCMV) , echinovirus, arbovirus, huntavirus, coxsackie virus, mumps virus, measles virus, rubella virus, polio virus, and/or human immunodeficiency virus type I or type 2 (HIV-1, HIV-2) .

[0074] To facilitate preparation of the immune cells with reduced toxicity, the present invention also includes kits for adoptive cell therapy. A kit of this invention can include an engineered nuclease for GM-CSF or perforin gene deletion or gene disruption; one or more microRNAs, siRNAs, shRNAs , or antisense molecule that inhibit the expression of GM-CSF or perforin; or a GM-CSF, GM-CSF receptor, or perforin antagonist. In some embodiments, the kit includes isolated immune cells including, but are not limited to T- lymphocytes, naive T cells, memory T cells, effector memory cells, natural killer cells, hematopoietic stem cells or pluripotent embryonic/ induced stem cells capable of giving rise to therapeutically relevant progeny. In other embodiments, the kit further includes nucleic acids encoding a CAR. The above-referenced reagents are typically provided, either individually or together, in a container and can be in liquid (ready-to-use) or lyophilized forms. Moreover, the kit can include a compatible pharmaceutical carrier as well as instructions for preparing immune cells with reduced toxicity and administering the same to a subject in need of treatment.

[0075] The following non-limiting examples are provided to further illustrate the present invention.

Example 1: LentiCRISPR Vectors to Eliminate Human GM-CSF

[0076] Target guide sequences (20 nt) with 3' protospacer adjacent motif (PAM) sequence (NGG) were selected according to established methods (Shalem, et al . (2014) Science 343:83-7) . The double - stranded oligonucleotides (Table 1), with BsmBI ends, were phosphorylated and cloned into the BsmBI site in vector lentiCRISPR as depicted in Figure 5. sgRNA vectors were sequence verified, then used to co- transfect 293T cells with packaging plasmids pCMV-VSVg and pCMV-dR8.91 to make lentivirus stocks. U937 and THP-1 cells were transduced with lentiviruses expressing the huGM-CSF target sgRNAs LGH-1, LGH-2, LGH-3 and LGH-4. TABLE 1

[ 0077 ] Day 1 post-transduction cells were transferred to RPMI media containing puromycin (2.0 g/mL) , and maintained in puromycin media for 10 days. THP-1 cells expressing LGH- 1 and LGH-3, and U937 cells expressing LGH-2 and LGH-4 were subjected to single-cell dilution to isolate single clones. Four clones for each construct were expanded in puromycin media before harvesting and isolating genomic DNA. Genomic regions containing the LGH-1 and LGH-2 recognition site in GM-CSF Exon 1 and the LGM-3 and LGH-4 region in Exon 3 were PCR-amplified and the -500 bp amplicons were cloned into vector pGE T. Sequencing of amplicons (from two different single cell colonies per group) revealed the insertions or deletions (indels) introduced by the CRISPR machinery.

[ 0 07 8 ] Of three sequences for LGH-1 in THP-1 , three deletions were detected. Of four sequences for LGH-2 in U937, four different indels were detected. Of two sequences for LGH-3 in THP-1, two contained different indels. Of the one sequence for LGH-4 in U937, this sequence contained an indel . Thus, all sgRNAs used were able to lead to alterations in the human GM-CSF gene. See Table 2. TABLE 2

sgRNA target sites are underlined.

Claims

What is claimed is :

1. A method for preventing toxicity of adoptive cell therapy comprising administering to a subject in need of adoptive cell therapy, a population of cells deficient in the expression or activity of GM-CSF thereby preventing toxicity of the adoptive cell therapy.

2. The method of claim 1, wherein toxicity comprises a cytokine storm.

3. The method of claim 1, wherein the adoptive cell therapy comprises chimeric antigen receptor-bearing lymphocyte therapy, T cell therapy, natural killer cell therapy, gamma/delta T cell therapy, or natural killer T cell therapy.

4. The method of claim 1, wherein the population of cells comprise a GM-CSF gene deletion or GM-CSF gene disruption .

5. The method of claim 1, wherein the population of cells have been contacted with a microR A, siRNA, shRNA, or antisense molecule that inhibits the expression of GM-CSF.

6. The method of claim 1, wherein the population of cells have been contacted with an antibody that blocks that activity of GM-CSF or GM-CSF receptor.

7. A method for preventing toxicity of adoptive cell therapy comprising administering to a subject in need of adoptive cell therapy, a population of cells deficient in the expression or activity of perforin thereby preventing toxicity of the adoptive cell therapy.

8. The method of claim 7, wherein the toxicity is cytotoxicity toward healthy cells.

9. The method of claim 7, wherein the adoptive cell therapy comprises chimeric antigen receptor-bearing lymphocyte therapy, T cell therapy, natural killer cell therapy, gamma/delta T cell therapy, or natural killer T cell therapy.

10. The method of claim 7, wherein the population of cells comprise a perforin gene deletion or perforin gene disruption .

11. The method of claim 7, wherein the population of cells have been contacted with a microRNA, siRNA, shRNA, or antisense molecule that inhibits the expression of perforin .

12. A kit comprising

(a) an engineered nuclease for GM-CSF gene deletion or gene disruption;

(b) a microRNA, siRNA, shRNA, or antisense molecule that inhibits the expression of GM-CSF;

(c) a GM-CSF or GM-CSF receptor antagonist; or

(d) a combination of (a) , (b) and (c) .

13. The kit of claim 12, further comprising an isolated population of T-lymphocytes , naive T cells, memory T cells, effector memory cells, natural killer cells, hematopoietic stem cells or pluripotent embryonic/induced stem cells.

14. The kit of claim 12, further comprising a nucleic acid molecule encoding a chimeric antigen receptor.

15. A kit comprising

(a) an engineered nuclease for perforin gene deletion or gene disruption;

(b) a microRNA, siRNA, shRNA, or antisense molecule that inhibits the expression of perforin; or

(c) a combination of (a) and (b) .

16. The kit of claim 15, further comprising an isolated population of T- lymphocytes , naive T cells, memory T cells, effector memory cells, natural killer cells, hematopoietic stem cells or pluripotent embryonic/induced stem cells.

17. The kit of claim 15, further comprising a nucleic acid molecule encoding a chimeric antigen receptor.