A METHOD OF GENERATING A LIBRARY OF POPULATIONS OF CELLS COMPRISING A TRANSGENE INTEGRATED INTO THE GENOME AT A

TARGET LOCUS

FIELD OF THE INVENTION

The present invention relates to methods for screening libraries of transgenes, which methods ensure that only a single transgene is inserted into the genome of each cell to avoid false positive or false negative identification of functional transgenes. Cells for use in the present invention include, but are not limited to, T cells, NK cells, or NKT cells. The transgene may encode, for example, a CAR or TOR.

BACKGROUND TO THE INVENTION

Chimeric antigen receptors (CARs) and engineered T cell receptors (TCRs) show great promise as therapeutics in a range of applications, particularly cancer.

Chimeric antigen receptors (CARs) consist of an antigen binding domain, typically a single chain antibody or ligand, fused to a spacer domain, transmembrane domain and intracellular signalling domains derived from coreceptors and Oϋ3z. Given their modular nature, it is possible to generate many iterations of a CAR from a single antigen-binding domain and the functional efficacy of each will vary. At present, there are no rules capable of predicting the efficacy of a CAR and empirical screening of individual CARs is required to identify the most efficacious, which is a costly and time-consuming process.

Methods currently employed to screen CARs for functional efficacy involve the transduction of T-cells with g-retroviral or lentiviral particles produced from a library of plasmids containing CAR encoding sequences. There are several drawbacks to such an approach. 1) It is difficult to control the number of viral integrations into the genome and a proportion of cells will possess more than one integration, giving rise to the possibility of identifying false positives and false negatives during the screening process. 2) Single insertion using retrovirus or lentivirus requires using a low multiplicity of infection (MOI), typically 0.3, which results in a transduction efficiency of approximately 30%, reducing the efficiency of screening and compounding downstream analysis of lead hits from the screen. 3) The expression level of the CAR may vary from cell to cell, depending on the site of integration in the genome, which could result in the failure to identify functional CARs due to insufficient expression at the cell surface.

An alternative to retroviral or lentiviral transduction is the insertion of a CAR coding sequence into the genome by homology-directed repair. This approach typically involves generating a targeted DNA double stranded break in the genome using an endonuclease such as Cas9 (CRISPR-associated 9 - where CRIPSR stands for clustered regularly interspaced short palindromic repeats) (Eyquem, J. et al. Nature 543, 113-117 (2017); Osborn, M. J. et al. Mol. Ther. 24, 570-581 (2016); Roth, T. L. et al. Nature 559, 405^09 (2018)), Cas12a/Cpf1 (Dai, X. et al. Nat. Methods 16, 247-254 (2019)), a transcription activator-like effector nuclease (TALEN) (Poirot, L. et al. Cancer Res. 75, 3853-3864 (2015); Berdien, B., et al. Gene Ther. 21, 539-548 (2014)), zinc finger nuclease (ZFN) (Torikai, H. et al. Blood 119, 5697-5705 (2012)), or homing endonuclease (Osborn, M. J. et al. Mol. Ther. 24, 570-581 (2016); MacLeod, D. T. et al. Mol. Ther. 25, 949-961 (2017)) and providing the cell with a homology- directed repair (HDR) template, which is copied into the genome. Recent published work integrating coding sequences into genome of T cells has focussed on the T cell receptor (TOR) a chain (TRAC), because only a single functional copy of the TOR a chain is usually expressed, due to allelic exclusion of the second copy of the gene, and targeted insertion at this site results in a loss of TCR a chain expression and downregulation of the TCR from the cell surface (Eyquem, J. et al. Nature 543, 113-117 (2017); Roth, T. L. et al. Nature 559, 405- 409 (2018); Berdien, B., et al. Gene Ther. 21, 539-548 (2014); Knipping, F. et al. Methods Clin. Dev. 4, 213-224 (2017)).

SUMMARY OF THE INVENTION

In order to capitalise on the promise of engineered cellular therapies, there remains a need to reliably screen engineered receptors in their intended location, i.e. , expressed by a cell. The present invention relates to methods for screening libraries of transgenes, which methods ensure that only a single transgene is inserted into the genome of each cell to avoid false positive or false negative identification of functional transgenes. Cells for use in the present invention include, but are not limited to, T cells, NK cells, or NKT cells. The transgene may encode, for example, a CAR or TCR.

Targeting of the TRAC locus has distinct advantages over retroviral/lentiviral methods as the site of integration and the expression level of the inserted transgene are constant and the number of insertions is relatively controlled, although there will be a proportion of double positive cells expressing two transgenes due to a failure in allelic exclusion.

In an alternative embodiment, the present invention relates to insertion of transgenes into loci on the X chromosome that do not have a corresponding locus on the Y chromosome. In this embodiment, either male or female cells may be used. Male cells only carry a single X chromosome, which ensures only a single copy of the transgene of interest is integrated into the genome. Female cells carry two X chromosomes but many genes are subject to X chromosome inactivation, where one of the two alleles is silenced to maintain correct gene dosing and ensure that only a single allele is expressed. Genes subject to X chromosome inactivation are therefore suitable for use in this embodiment because only one copy of the transgene of interest will be expressed.

Accordingly, in a first aspect the present invention provides a method of generating a population of cells, wherein each cell comprises a transgene of interest integrated into the genome at a target locus, wherein the transgene of interest is selected from a library of transgenes of interest, the method comprising: i. introducing into the cells an exonuclease system which produces a DNA double-strand break at the target locus; and ii. introducing into the cells a library of exogenous polynucleotides, each comprising a 5' homology arm homologous to a portion of the target locus, a transgene of interest, and a 3' homology arm homologous to a portion of the target locus, wherein the exogenous polynucleotide is inserted into the genome at the target locus by homologous recombination, further wherein the target locus is either:

(a) the T cell receptor alpha constant region (TRAC)] or

(b) a locus on the X chromosome.

In some embodiments, when the target locus is a locus on the X chromosome, only cells from male donors are used.

The cells may be any type of immune cell. In particular, the cells may be T cells, NK cells, NKT cells, monocytes, or macrophages.

The homology arms are used to target the locus of interest and are typically 300-1000 bp in length. These sections of the exogenous polynucleotide contain sequences that are homologous to the target locus, which results in their incorporation via homology directed repair of double-strand breaks in genomic DNA.

The transgene of interest may be a chimeric antigen receptor (CAR) or T-cell receptor (TCR). Where the transgene of interest is a CAR, the CAR may comprise one or more extracellular ligand-binding domains and one or more intracellular signalling domains. The CAR may comprise an extracellular antigen binding domain, an optional spacer domain, a transmembrane domain, and an intracellular signalling domain.

The exonuclease system may comprise a recombinant meganuclease, a recombinant zinc- finger nuclease (ZFN), a recombinant transcription activator-like effector nuclease (TALEN), a CRISPR/Cas nuclease, or a megaTAL nuclease. Preferably the exonuclease system comprises a CRISPR/Cas nuclease.

The exogenous polynucleotide may further comprise a marker gene or barcode sequence.

Where a locus on the X chromosome is used, the locus may be selected from the group comprising HPRT1 , PGK1, CD40L, IL2RG, CXCR3, CETN2, OFD1 and SLC25A5. Other loci on the X chromosome are also contemplated within the scope of the present invention.

Insertion may be targeted at exons or introns. Where exons are targeted, insertions may be made such that promoters at the site of insertion drive gene expression. Alternatively, the sequence to be inserted may include a separate promoter. Where introns are targeted the sequence to be inserted will typically include a promoter upstream of the transgene of interest. Sequences targeted to an intron may also include a polyadenylation signal sequence downstream of the gene of interest.

In a further aspect, there is provided a method screening a library of CARs, the method comprising: i. generating a population of cells according to the method of the first aspect, where the transgene of interest is a CAR; ii. assessing the resulting population of cells for activity of the transgene of interest.

In another aspect, the present invention provides a method of generating a population of cells, wherein each cell comprises a transgene of interest integrated into the genome at a target locus, wherein the transgene of interest is selected from a library of transgenes of interest, the method comprising: i. introducing into the cells an exonuclease system which produces a DNA double-strand break at the target locus; and ii. introducing into the cells a library of exogenous polynucleotides, each comprising a transgene of interest, wherein the exogenous polynucleotide is inserted into the genome at the target locus by non- homologous end joining (NHEJ), further wherein the target locus is either:

(a) the T cell receptor alpha constant region (TRAC)] or

(b) a locus on the X chromosome.

Embodiments described with respect to the first aspect can also be applied to this further aspect mutatis mutandis.

BRIEF DESCRIPTION OF THE FIGURES

Figure 1 - Insertion of CAR encoding sequence into the TRAC locus A DNA double stranded break is created at the 5’ end of exon 1 of the TRAC locus. Repair of the DNA double stranded break can occur via non-homologous end joining or homolog- directed repair (HDR). HDR can used to insert heterologous DNA sequences into the genome provided there are regions homologous to the target gene flanking it (left and right homology arms). These homology arms are typically 300-1000 bp in length. The sequence encoding the CAR is inserted in-frame with the TCR a chain constant region and expression is facilitated by a self-cleaving peptide sequence placed upstream of the CAR. A stop codon and polyadenylation signal sequence placed downstream of the CAR terminate translation and transcription, respectively.

Figure 2 - Insertion of anti-CD19 and anti-CD22 CARs into the TRAC locus Primary T-cells were nucleofected with sgRNA/Cas9 RNP complexes targeting the TRAC locus and HDR templates encoding an anti-CD19 CAR (FMC63) and an anti-CD22 CAR (LT22, 9A8 or inotuzamab). All CARs were inserted into the TRAC locus with varying efficiency. The reduced efficiency of the anti-CD22 CARs may be due in part to a detection issue related to the biophysical properties of the single chain antibodies. The results show that it is possible to generate distinct single positive populations of anti-CD19 or anti-CD22 CARs from a mixed pool of HD templates (FMC63 and inotuzumab - bottom right flow cytometry plot). It should be noted that a double positive population of anti-CD19 and anti-CD22 CAR positive cells is also present and this might be due to a fault in TCR a chain allelic exclusion. Figure 3 - Diagram of the HDR templates targeting the HPRT1 gene.

A depiction of the HPRT1 gene and below are the proposed HDR templates to insert a CAR encoding sequence into the gene. The HDR templates on the left are for CAR insertion at the 5’ end of the gene (exon 1) to produce either an in-frame insertion or an out of frame insertion that will disrupt expression of the HPRT1 gene product. The out of frame insertion is achieved by placing a stop codon and polyadenylation sequence downstream of the CAR and deleting the initiating methionine from HPRT1. The HDR template on the right makes an in-frame insertion in the HPRT1 gene (the CAR has a stop codon and polyadenylation sequence to facilitate expression).

Figure 4 - The conversion of 6-thioguanine to thioguanine monophosphate.

Diagram showing the conversion of 6-thioguanine to thioguanine monophosphate and subsequent conversion to 2-deoxy-6-thioguanosine triphosphate by kinases and reductase.

Figure 5 - Flow cytometry of cells nucleofected with HDRT targeting the HPRT 1 gene exon 2 SupT 1 cells (which originate from a male donor) were nucleofected with one of four guide RNAs (sgRNAI, 2,3, or 4) and HDR templates comprising either CAT19 CAR or FMC63 CAR.

Figure 6 - Flow cytometry of cells nucleofected with HDRT targeting the HPRT1 C terminus SupT 1 cells (which originate from a male donor) were nucleofected with one of two guide RNAs (92_fwd and 66_rev) and HDR templates comprising either CAT19 CAR or FMC63 CAR.

Figure 1 - Flow cytometry of cells nucleofected with HDRT targeting the CETN2 locus SupT 1 cells (which originate from a male donor) were nucleofected with one of four guide RNAs (59_rev, 60_fwd, 66_fwd, and 76_fwd) and HDR templates comprising either CAT19 CAR or FMC63 CAR.

Figure 8 - Flow cytometry of cells nucleofected with HDRT targeting the IL2RG locus SupT1 cells (which originate from a male donor) were nucleofected with one of three guide RNAs (48_rev, 69_rev, and 80_rev) and HDR templates comprising either CAT19 CAR or FMC63 CAR.

Figure 9 - Flow cytometry of cells nucleofected with HDRT targeting the PGK locus (N- terminus)

PBMCs from two healthy donors (D31 and D34) were treated with Cas9 RNP complex with sgRNA targeting PGK1 N-terminus (PGK1 81 forw) and a repair template for CAT19, FMC63 or both CAT19 and FMC63. CAT19 expression and FMC63 expression were analysed by flow cytometry 7 days post nucleofection.

Figure 10 - Flow cytometry of cells nucleofected with HDRT targeting the PGK locus (C- terminus)

PBMCs from a healthy donor (D31) were treated with Cas9 RNP complex with sgRNA targeting PGK1 C-terminus (PGK1 28 forw) and repair templates for CAT19, FMC63 or both CAT19 and FMC63. CAT19 expression and (a) FMC63 or (b) HA expression were analysed by flow cytometry 7 days post nucleofection.

Figure 11 - Cytometric analysis of co-cultures of Raji cells and genome edited PBMCs Flow cytometry plots showing target cell populations from cytotoxicity assay using CAT19 and FMC63 integrated CAR-T cells after 24 hours. PBMCs from two healthy donors (D31 and D34) were treated with Cas9 RNP complex with sgRNA targeting PGK1 N-terminus (PGK1 81 forw) and a repair template for CAT19, FMC63 or both CAT19 and FMC63. 8 days post nucleofection cells were incubated with Raji WT or Raji CD19 KO target cells at E:T ratios of 1:1 and 1:4 for 24 hours. Cells were stained with CD2 and CD3s to distinguish effector and target populations. Samples with effectors alone or targets alone were used to determine gating.

Figure 12 - Cytometric analysis of co-cultures of Raji cells and genome edited PBMCs PBMCs from two healthy donors (D31 and D34) were treated with Cas9 RNP complex with sgRNA targeting PGK1 N-terminus (PGK1 81 forw) and a repair template for CAT19, FMC63 or both CAT19 and FMC63. 8 days post nucleofection cells were incubated with Raji WT or Raji CD19 KO target cells at E:T ratios of 1:1 and 1:4 for 24 hours. Cells were stained with CD2 and CD3s to distinguish effector and target populations. Samples with effectors alone or targets alone were used to determine gating.

Figure 13 - Percent survival of target cells normalised to the no DNA control PBMCs from two healthy donors (D31 and D34) were treated with Cas9 RNP complex with sgRNA targeting PGK1 N-terminus (PGK1 81 forw) and a repair template for CAT19, FMC63 or both CAT19 and FMC63. 8 days post nucleofection cells were incubated with Raji WT or Raji CD19 KO target cells at E:T ratios of 1 :1 and 1 :4 for (a) 24 and (b) 72 hours. Graphs show % survival of target cells normalized to the No DNA control. Genome edited PBMCs expressing the CAT19 CAR or FMC63 CAR were able to lyse CD19 positive Raji cells. Only limited targeting of CD19 negative Raji cells was observed at the higher 1 :1 ratio (effector: target). Figure 14 - Flow cytometry of cells nucleofected with HDRT targeting the HPRT1 gene exon 2

PBMCs from three healthy donors (D063, D064 and D564) were treated with Cas9 RNP complex with sgRNA targeting HPRT1 Exon2 (HPRT1 Synthego2) and a repair template for CAT19. CAT19 expression was detected by flow cytometry 7 days post nucleofection using a biotinylated anti-idiotype antibody and PE-conjugated streptavidin. Insertion efficiency at exon 2 of the HPRT 1 gene ranged from 6 to 20%.

Figure 15 - Flow cytometry of cells nucleofected with HDRT targeting the HPRT1 gene final exon

PBMCs from three healthy donors (D063, D064 and D564) were treated with Cas9 RNP complex with sgRNA targeting HPRT1 C-terminus (HPRT1 92 forw) and a repair template for CAT19. CAT19 expression was detected by flow cytometry 7 days post nucleofection using a biotinylated anti-idiotype antibody and PE-conjugated streptavidin. Integration efficiency at the final exon of the HPRT 1 gene ranged from 0.6 to 2.0%.

Figure 16 - Flow cytometry of cells nucleofected with HDRT targeting the IL2RG locus PBMCs from three healthy donors (D063, D064 and D564) were treated with Cas9 RNP complex with sgRNA targeting IL2RG N-terminus (IL2RG 69 Rev) and a repair template for CAT19. CAT19 expression was detected by flow cytometry 7 days post nucleofection using a biotinylated anti-idiotype antibody and PE-conjugated streptavidin. Integration efficiency was found to be dependent on the donor and ranged from 10 to 42%.

Figure 17- Insertion of anti-CD19 CAR to the SLC25A5 gene

PBMCs were nucleofected with a HDR template encoding the anti-CD19 (CAT) CAR and RNP complexes formed with one of three sgRNAs targeting the 5’ end of the SLC25A5 gene. Insertion of the anti-CD19 (CAT) CAT was observed when PBMCs were nucleofected with an SLC25A5 targeting sgRNA but not the non-targeting sgRNA. Integration efficiency was up to 46%. These data clearly demonstrated that it was possible to integrate a CAR encoding sequence to the SLC25A5 gene.

DETAILED DESCRIPTION OF THE INVENTION

To circumvent the problems associated with retroviral and lentiviral integration of transgenes into the genome, the present inventors use genome editing to insert transgene encoding sequences either into the TRAC locus or a gene located on the X chromosome. Targeting genes located on the X chromosome ensures that only a single copy of the encoding transgene sequence is expressed and overcomes the problem of cells expressing multiple different copies of the transgene of interest.

Gene editing systems use an exonuclease system to introduce a DNA double strand break at a targeted locus. The exonuclease system may comprise a recombinant meganuclease, a recombinant zinc-finger nuclease (ZFN), a recombinant transcription activator-like effector nuclease (TALEN), a CRISPR/Cas nuclease, or a megaTAL nuclease. A preferred exonuclease system is the CRISPR/Cas nuclease.

The CRISPR/Cas nuclease is used in conjunction with a guide RNA (gRNA) that targets the locus of interest. These guide RNAs may also be termed “single guide RNA” or sgRNA.

Once a double strand break is introduced, the cell will attempt repair using native homology- directed repair mechanisms. By introducing homology directed repair templates in the form of exogenous polynucleotides carrying suitable homology arms and a transgene of interest, insertions into the locus of interest can be achieved.

TARGETING TRAC

An advantage of targeting the TRAC locus versus insertion at other sites is that the CAR will be expressed at levels comparable to the endogenous TCR a chain and, depending on the target site, can result in disruption of the endogenous gene, loss of expression and downregulation of the TCR from the cell surface. However, insertion at the TRAC locus can result in the expression of two CARs in a population of cells, which would compound the downstream analysis and identification of the CAR T-cells. Example 1 provides data showing that it is possible to target the TRAC locus using Cas9 and insert one or two copies of a CAR into the genome (Figures 1 and 2) by homology-directed repair (HDR). To screen CARs, sequences are cloned into an HDR template and a library generated by amplifying the sequences individually from the plasmid library and combining the resulting PCR products into a single pool. These pooled PCR products provide the cells with templates for homology- directed repair, enabling the insertion of CAR-encoding sequences into their genome. Once the genome edited CAR T-cells have been generated, screening of the CARs can be carried out following the methods described below (screening of CAR T-cells). TARGETING GENES ON THE X CHROMOSOME

The problem of ensuring single copy gene insertion can be overcome by targeting the genes located on the X chromosome. There are several genes located on the X chromosome provided in Table 1, which facilitate the screening process or achieve desirable levels of CAR expression that are known to be efficacious in clinical trials. The genes fall into three categories: house-keeping genes, immune-related, and cytoskeletal proteins (centriolar components). The two preferred genes are the house-keeping genes hypoxanthine-guanine phosphoribosyltransferase 1 ( HPRT1 ) and phosphoglycerate kinase 1 ( PGK1 ).

Table 1.

Targeting insertions to the X chromosome may be carried out in either male or female cells. The use of cells from male donors provides a convenient mechanism for single-copy insertion, since male cells carry only a single X chromosome. Cells from female donors may also be used because of the phenomenon of X chromosome inactivation, in which one X chromosome is made transcriptionally silent. Female cells therefore only express one copy of X-linked genes.

HPRT

An advantage of targeting HPRT is that it facilitates the selection of genome edited CAR T- cells because knockout of HPRT renders the T-cells insensitive to the cytotoxic compound 6- thioguanine, while non-edited cells would remain susceptible to the compound (Liao, S. et al. Nucleic Acids Res. 43 (20) e134 (2015)). HPRT is part of the non-essential purine salvage pathway and catalyses the conversion of hypoxanthine to inosine monophosphate and guanine to guanine monophosphate. The cytotoxic compound 6-thioguanine is converted by HPRT to the nucleotide, which becomes incorporated into DNA and is methylated to form S ⁶ - methylthioguanine (Swann, P. F. et al. Science 273, 1109-1111 (1996); Karran, P. Br. Med. Bull. 79-80, 153-170 (2006)). During DNA replication, S ⁶ -methylthioguanine pairs with either cytosine or thymine. Mispairing of S ⁶ -methylthioguanine with thymine is recognised by the post-replicative mismatch repair pathway and causes apoptosis. Targeting HPRT therefore not only enables the removal of non-edited cells from the pool prior to challenging CAR T-cells with antigen positive target cells (via 6-thioguanine selection), but also result in the production of CAR T-cells resistant to the chemotherapeutic agent 6-thioguanine.

PGK1

The PGK 1 promoter is used in lentiviral and self-inactivating viral plasmids to drive the transcription of inserted transgenes. In clinical trials targeting the pan B-cell marker CD19 for the treatment of paediatric B-cell acute lymphoblastic leukaemia (pALL) and adult B-cell ALL, the PGK1 promoter was used to drive transcription of the CAR (Ghorashian, S.et al. Br. J. Haematol. 169 (4), 463-478 (2015)). As dramatic clinical response rates were observed in the clinical trials, this would suggest that the level of CAR expressed from the PGK1 promoter is sufficient to confer functionality and to drive CAR T-cell persistence. By inserting CAR encoding sequences into the PGK1 gene, without disrupting its expression, the level of CAR expression will be comparable to that from the promoter fragment cloned into self-inactivating g-retroviral and lentiviral constructs.

IL2RG

The lnterleukin-2 receptor gamma subunit (common g chain) is expressed in most lymphocytes and is therefore an attractive target for expression of a CAR, since lymphocytes are the target cell type for these molecules. By inserting CAR encoding sequences into the IL2RG gene, without disrupting its expression, expression of the CAR should be achieved.

CETN2

The centrin-2 protein is a component of the cytoskeleton and is therefore expressed in all cells. By inserting CAR encoding sequences into the CETN2 gene, without disrupting its expression, expression of the CAR should be achieved.

SLC25A5

Solute carrier family 25 member 5 is involved in the exchange of cytoplasmic ADP for ATP across the inner membrane of the mitochondrion. SCREENING OF CELLS

Once a population of cells is created, they can be assessed for activity of the transgene of interest. This may include, for example, assessment of the expression of the transgene. In other cases, the activity of the transgene itself may be assessed. For example, where the transgene is an enzyme, the activity of the enzyme on its substrate can be measured. In the case of CARs, the assessment will typically be based on the ability of the resulting CAR-T cells to kill target cells.

The screening of the CAR T-cells may be carried out in the same manner, regardless of the approached adopted. To facilitate the identification of the CAR sequence, a short 15 base pair (bp) barcode of unique DNA sequence may be incorporated either at the 5’ or the 3’ end of the CAR sequence. The location of the barcode will depend on the approach taken, but the barcode should be positioned at a location ensuring that it can be read by next generation sequencing (NGS), which typically has a limited read length in the order of 250-300 bp.

After generating a pool of CAR T-cells, identification of the most efficacious CARs may be carried out using two different approaches. The fist approach involves the short-term challenge of CAR T-cells with antigen positive target cells and the isolation of responding T-cells, while the second approach requires the long-term culture of CAR T-cells with target cells to determine which CAR provides the most potent signal to the T-cell.

The short-term approach to determining CAR efficacy involves challenging CAR T-cells with antigen positive target cells and isolating responding T-cells that have upregulated CD69 and express this early activation marker on their cell surface. Identification of the CAR expressed by the responding cell can then carried out by extracting genomic DNA from the isolated cells, amplifying across the single chain antibody sequence of the CAR and analysing the PCR products by next generation sequencing (NGS).

The second, long-term, approach involves repeatedly challenging a pool of CAR T-cells with target cells and sampling the population over time to identify the most frequently observed CARs to determine which ones confer to the T-cells proliferative capacity. The population will narrow over time as the most proliferative T-cells will outcompete the others in the culture. Again, identification of the CARs is carried out by extracting genomic DNA from the T-cells, amplifying across the single chain antibody and barcode sequence and analysing the resulting PCR products by NGS. CHIMERIC ANTIGEN RECEPTOR (CAR)

Classical CARs are chimeric type I trans-membrane proteins which connect an extracellular antigen-recognizing domain (binder) to an intracellular signalling domain (endodomain). The binder is typically a single-chain variable fragment (scFv) derived from a monoclonal antibody (mAb), but it can be based on other formats which comprise an antibody-like antigen binding site or on a ligand for the target antigen. A spacer domain may be necessary to isolate the binder from the membrane and to allow it a suitable orientation. A common spacer domain used is the Fc of lgG1. More compact spacers can suffice e.g. the stalk from CD8a and even just the lgG1 hinge alone, depending on the antigen. A trans-membrane domain anchors the protein in the cell membrane and connects the spacer to the endodomain.

Early CAR designs had endodomains derived from the intracellular parts of either the g chain of the FcsR1 or Oϋ3z. Consequently, these first generation receptors transmitted immunological signal 1, which was sufficient to trigger T-cell killing of cognate target cells but failed to fully activate the T-cell to proliferate and survive. To overcome this limitation, compound endodomains have been constructed: fusion of the intracellular part of a T-cell co stimulatory molecule to that of Oϋ3z results in second generation receptors which can transmit an activating and co-stimulatory signal simultaneously after antigen recognition. The co stimulatory domain most commonly used is that of CD28. This supplies the most potent co stimulatory signal - namely immunological signal 2, which triggers T-cell proliferation. Some receptors have also been described which include TNF receptor family endodomains, such as the closely related 0X40 and 4-1 BB which transmit survival signals. Even more potent third generation CARs have now been described which have endodomains capable of transmitting activation, proliferation and survival signals.

CAR-encoding nucleic acids may be transferred to T cells using, for example, retroviral vectors. In this way, a large number of antigen-specific T cells can be generated for adoptive cell transfer. When the CAR binds the target-antigen, this results in the transmission of an activating signal to the T-cell it is expressed on. Thus, the CAR directs the specificity and cytotoxicity of the T cell towards cells expressing the targeted antigen.

In the context of the present invention, the transgene of interest may comprise a CAR. More particularly, a library of different CARs may be created and then screened using the methods described herein. The various components of the CAR are set out below in more detail below. Each of these components may be varied in the library, either independently or in combination with one another. ANTIGEN BINDING DOMAIN

The antigen-binding domain is the portion of a classical CAR which recognizes antigen.

Numerous antigen-binding domains are known in the art, including those based on the antigen binding site of an antibody, antibody mimetics, and T-cell receptors. For example, the antigen binding domain may comprise: a single-chain variable fragment (scFv) derived from a monoclonal antibody; a wild-type ligand of the target antigen; a peptide with sufficient affinity for the target; a single domain binder such as a camelid; an artificial binder single as a Darpin; or a single-chain derived from a T-cell receptor.

Various tumour associated antigens (TAA) are known, as shown in the following Table 2. The antigen-binding domain used in the present invention may be a domain which is capable of binding a TAA as indicated therein. Table 2

TRANSMEMBRANE DOMAIN

The transmembrane domain is the sequence of a classical CAR that spans the membrane. It may comprise a hydrophobic alpha helix. The transmembrane domain may be derived from CD28, which gives good receptor stability. CAR OR TCR SIGNAL PEPTIDE

The CAR or transgenic TCR for use in the present invention may comprise a signal peptide so that when it is expressed in a cell, such as a T-cell, the nascent protein is directed to the endoplasmic reticulum and subsequently to the cell surface, where it is expressed.

The core of the signal peptide may contain a long stretch of hydrophobic amino acids that has a tendency to form a single alpha-helix. The signal peptide may begin with a short positively charged stretch of amino acids, which helps to enforce proper topology of the polypeptide during translocation. At the end of the signal peptide there is typically a stretch of amino acids that is recognized and cleaved by signal peptidase. Signal peptidase may cleave either during or after completion of translocation to generate a free signal peptide and a mature protein. The free signal peptides are then digested by specific proteases.

SPACER DOMAIN

The receptor may comprise a spacer sequence to connect the antigen-binding domain with the transmembrane domain. A flexible spacer allows the antigen-binding domain to orient in different directions to facilitate binding.

The spacer sequence may, for example, comprise an lgG1 Fc region, an lgG1 hinge or a human CD8 stalk or the mouse CD8 stalk. The spacer may alternatively comprise an alternative linker sequence which has similar length and/or domain spacing properties as an lgG1 Fc region, an lgG1 hinge or a CD8 stalk. A human lgG1 spacer may be altered to remove Fc binding motifs.

INTRACELLULAR SIGNALLING DOMAIN

The intracellular signalling domain is the signal-transmission portion of a classical CAR.

The most commonly used signalling domain component is that of CD3-zeta endodomain, which contains 3 ITAMs. This transmits an activation signal to the T cell after antigen is bound. CD3-zeta may not provide a fully competent activation signal and additional co-stimulatory signalling may be needed. For example, chimeric CD28 and 0X40 can be used with CD3- Zeta to transmit a proliferative / survival signal, or all three can be used together.

The intracellular signalling domain may be or comprise a T cell signalling domain.

The intracellular signalling domain may comprise one or more immunoreceptor tyrosine-based activation motifs (ITAMs). An ITAM is a conserved sequence of four amino acids that is repeated twice in the cytoplasmic tails of certain cell surface proteins of the immune system. The motif contains a tyrosine separated from a leucine or isoleucine by any two other amino acids, giving the signature YxxL/l. Two of these signatures are typically separated by between 6 and 8 amino acids in the tail of the molecule (UCCI_/IC ₍₆ -8 ₎ UCC ).

ITAMs are important for signal transduction in immune cells. Hence, they are found in the tails of important cell signalling molecules such as the CD3 and z-chains of the T cell receptor complex, the CD79 alpha and beta chains of the B cell receptor complex, and certain Fc receptors. The tyrosine residues within these motifs become phosphorylated following interaction of the receptor molecules with their ligands and form docking sites for other proteins involved in the signalling pathways of the cell.

The intracellular signalling domain component may comprise, consist essentially of, or consist of the Oϋ3-z endodomain, which contains three ITAMs. Classically, the Oϋ3-z endodomain transmits an activation signal to the T cell after antigen is bound.

The intracellular signalling domain may comprise additional co-stimulatory signalling. For example, 4-1 BB (also known as CD137) can be used with Oϋ3-z, or CD28 and 0X40 can be used with Oϋ3-z to transmit a proliferative / survival signal.

The CAR may have the general format: antigen-binding domain-TCR element.

As used herein “TCR element” means a domain or portion thereof of a component of the TCR receptor complex. The TCR element may comprise (e.g. have) an extracellular domain and/or a transmembrane domain and/or an intracellular domain e.g. intracellular signalling domain of a TCR element.

The TCR element may selected from TCR alpha chain, TCR beta chain, a CD3 epsilon chain, a CD3 gamma chain, a CD3 delta chain, CD3 epsilon chain.

TRANSGENIC T-CELL RECEPTOR (TCR)

The T-cell receptor (TCR) is a molecule found on the surface of T cells which is responsible for recognizing fragments of antigen as peptides bound to major histocompatibility complex (MHC) molecules.

The TCR is a heterodimer composed of two different protein chains. In humans, in 95% of T cells the TCR consists of an alpha (a) chain and a beta (b) chain (encoded by TRA and TRB, respectively), whereas in 5% of T cells the TCR consists of gamma and delta (g/d) chains (encoded by TRG and TRD, respectively).

When the TCR engages with antigenic peptide and MHC (peptide/M HC), the T lymphocyte is activated through signal transduction.

In contrast to conventional antibody-directed target antigens, antigens recognized by the TCR can include the entire array of potential intracellular proteins, which are processed and delivered to the cell surface as a peptide/MHC complex.

It is possible to engineer cells to express heterologous (i.e. non-native) TCR molecules by artificially introducing the TRA and TRB genes; or TRG and TRD genes into the cell using a vector. For example the genes for engineered TCRs may be reintroduced into autologous T cells and transferred back into patients for T cell adoptive therapies. Such ‘heterologous’ TCRs may also be referred to herein as ‘transgenic TCRs’.

The transgenic TCR for use in the present invention may recognise a tumour associated antigen (TAA) when fragments of the antigen are complexed with major histocompatibility complex (MHC) molecules on the surface of another cell.

The transgenic TCR for use in the present invention may recognise a TAA listed in Table 2.

In the context of the present invention, a library of transgenic TCRs may be created. This library may then be screened using the methods described herein.

DEFINITIONS

As used herein, the terms “polynucleotide”, “nucleotide”, and “nucleic acid” are intended to be synonymous with each other.

It will be understood by a skilled person that numerous different polynucleotides and nucleic acids can encode the same polypeptide as a result of the degeneracy of the genetic code. In addition, it is to be understood that skilled persons may, using routine techniques, make nucleotide substitutions that do not affect the polypeptide sequence encoded by the polynucleotides described herein to reflect the codon usage of any particular host organism in which the polypeptides are to be expressed. Suitably, the polynucleotides of the present invention are codon optimised to enable expression in a mammalian cell, in particular an immune effector cell as described herein. Nucleic acids according to the invention may comprise DNA or RNA. They may be single- stranded or double-stranded. They may also be polynucleotides which include within them synthetic or modified nucleotides. A number of different types of modification to oligonucleotides are known in the art. These include methylphosphonate and phosphorothioate backbones, addition of acridine or polylysine chains at the 3' and/or 5' ends of the molecule. For the purposes of the use as described herein, it is to be understood that the polynucleotides may be modified by any method available in the art. Such modifications may be carried out in order to enhance the in vivo activity or life span of polynucleotides of interest.

The terms “variant”, “homologue” or “derivative” in relation to a nucleotide sequence or amino acid sequence includes any substitution of, variation of, modification of, replacement of, deletion of or addition of one (or more) nucleic acid(s) from or to the sequence.

This disclosure is not limited by the exemplary methods and materials disclosed herein, and any methods and materials similar or equivalent to those described herein can be used in the practice or testing of embodiments of this disclosure. Numeric ranges are inclusive of the numbers defining the range. Unless otherwise indicated, any nucleic acid sequences are written left to right in 5' to 3' orientation; amino acid sequences are written left to right in amino to carboxy orientation, respectively.

Where a range of values is provided, it is understood that each intervening value, to the tenth of the unit of the lower limit unless the context clearly dictates otherwise, between the upper and lower limits of that range is also specifically disclosed. Each smaller range between any stated value or intervening value in a stated range and any other stated or intervening value in that stated range is encompassed within this disclosure. The upper and lower limits of these smaller ranges may independently be included or excluded in the range, and each range where either, neither or both limits are included in the smaller ranges is also encompassed within this disclosure, subject to any specifically excluded limit in the stated range. Where the stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in this disclosure.

It must be noted that as used herein and in the appended claims, the singular forms "a", "an", and "the" include plural referents unless the context clearly dictates otherwise.

The terms "comprising", "comprises" and "comprised of as used herein are synonymous with "including", "includes" or "containing", "contains", and are inclusive or open-ended and do not exclude additional, non-recited members, elements or method steps. The terms "comprising", "comprises" and "comprised of also include the term "consisting of.

The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. Nothing herein is to be construed as an admission that such publications constitute prior art to the claims appended hereto.

The invention will now be further described by way of Examples, which are meant to serve to assist one of ordinary skill in the art in carrying out the invention and are not intended in any way to limit the scope of the invention.

EXAMPLES PREPARATION Reconstitution of sgRN A

CRISPR guide RNAs are supplied by Synthego. CRISPR guide RNAs are stored at -20°C unreconstituted and -80°C once reconstituted in re-suspension buffer.

To prepare a final concentration of 80 mM

METHOD

Preparation of RNA/HDR Templates

Homology-directed repair templates (HDRT) were generated by PCR using suitable primers. Templates were purified by SPRI according to standard methods.

All templates were normalised, using P3 to 1 pg/mI. 4mI of template (single integrations) and 4mI of each repair template (double integrations) was added to each well of a sterile 96-well, V-bottom plate first, followed by addition of the appropriate sgRNA on top of the HDR template.

Preparation of cells for nucleofection

Transfection buffer was prepared by combining the following for each sample: SupT1-Cas9 cells (AU54111) were washed twice with an equal volume of PBS by centrifuging at 400g for 5 minutes to remove culture medium and RNases present in the serum that would rapidly degrade the gRNA duplexes. After 2x washes, cells were re-suspended in 21 pl_ of transfection buffer per sample to be nucleofected (e.g. 1x106 SupT1-Cas9 cells per nucleofection).

The total volume of cells was added to the sgRNA/HDR template and the mixture was transferred to the electroporation cuvette/plate. Cells were nucleofected using the CM- 150 pulse code. 80mI of prewarmed RPMI was added to each cuvette after electroporation and cells were left to recover in the incubator for 15 minutes.

100mI of cells were transferred to a 96 well plate containing 100mI of 60mM HDR Enhancer and incubated at 37°C overnight. Cells were centrifuged at 400g for 5 minutes and resuspend in RPMI. Transfer cells to a 24 well plate containing 900mI of RPMI. After 7 days cells were stained by flow cytometry with the antibodies in the table below to check the % integration.

Insertion at the TRAC locus

We have targeted the TRAC locus using sgRNA/Cas9 ribonucleic acid protein (RNP) complexes to introduce a DNA double strand break (Figure 1) and provided primary T cells with two HDR templates, encoding either an anti-CD19 or anti-CD22 CAR, and measured the frequency of insertion of each template into the target site. The data showed that only one CAR was typically inserted at the TRAC locus with distinct single positive populations, but double insertions also occurred with a CAR double positive population Figure 2). In the example shown, the double positive population represented 19% of the genome edited population, while the remaining genome edited cells represented the two single positive populations.

A single anti-CD19 CAR (FMC63) and three different anti-CD22 CARs were chosen for integration into the TRAC locus. The anti-CD22 CARs were: clone LT22, clone 9A8 and Inotuzumab. The CARs were inserted into the TRAC locus either singly or in combination (Figure 2). The integration efficiency of the anti-CD19 CAR alone was approximately 30%, but this decreased by nearly half when simultaneously integrated into the genome with an anti- CD22 CAR. Integration of the anti-CD22 CARs appeared to be less efficient and ranged from 3-15% although these differences may be due in part to a detection issue related to the biophysical properties of the anti-CD22 single chain antibodies fused to the CARs. Nevertheless, the results show that is clearly possible to generate distinct single populations of either anti-CD19 or anti-CD22 CAR positive T-cells from a mixed pool of HDR templates. It should be noted however that there is a population of anti-CD19 and anti-CD22 CAR double positive T-cells, probably arising from a failure of allelic exclusion of the TRAC loci.

Insertion at the HPRT1 gene

Insertion of CARs into the HPRT1 gene should result in the constitutive expression of the CAR. Several HDR templates were constructed to test the possibility of inserting CAR encoding sequences into the HPRT1 gene. The first targets the 5’ end of the gene, specifically exon 2, with the CAR sequence placed in-frame with HPRT 1 or out of frame with the intention of disrupting the HPRT1 gene to prevent expression of a functional protein (Error! Reference source not found. Figure 3). The second targets the 3’ end of the HPRT1 gene and will be placed in-frame with the gene with the intention of retaining enzymatic activity.

By inserting the CAR at the 5’ end of the HPRT1 with a stop codon and polyadenylation signal sequence downstream of the CAR encoding sequence it is possible to disrupt expression of HPRT1. This will render modified T-cells insensitive to the chemotherapeutic drug 6- thioguanine, which is converted to 6-thioguanine monophosphate by HPRT1 and ultimately to 2-deoxy-6-thioguanosine triphosphate (Figure 4). Incorporation of 6-thioguanine into DNA and its methylation to produce S6-methylthioguanine can cause mis-pairing with thymine during DNA replication. Mis-pairing of 6-thioguanine and thymine is detected by post-replicative mismatch repair and causes apoptosis. Using this approach, it should be possible to eliminate unmodified T-cells from the population prior to carrying out downstream applications such as cytotoxicity assays.

Figure 5 shows the results of nucleofecting SupT 1 cells (which originate from a male donor) with one of four guide RNAs (sgRNAI , 2,3, or 4) targeting exon 2 of the HPRT1 gene. Cells were also nucleofected with HDR templates comprising either CAT19 CAR or FMC63 CAR.

Figure 6 shows the results of nucleofecting SupT 1 cells (which originate from a male donor) with one of two guide RNAs (92_fwd and 66_rev) targeting the C terminus of the HPRT 1 gene. Cells were also nucleofected with HDR templates comprising either CAT19 CAR or FMC63 CAR.

Since SupT1 cells have the karyotype XXYY, some double insertions are seen when both HDR templates are used. However, expression from the desired locus using a mixture of templates has been achieved. Insertion at the CETN2 gene

Insertion of CARs into the CETN2 gene should result in the constitutive expression of the CAR. HDR templates have been constructed to test the possibility of inserting CAR encoding sequences into the CETN2 gene.

Figure 7 shows the results of nucleofecting SupT 1 cells (which originate from a male donor) with one of four guide RNAs (59_rev, 60_fwd, 66_fwd, and 76_fwd) targeting the CETN2 gene and HDR templates comprising either CAT19 CAR or FMC63 CAR.

Since SupT1 cells have the karyotype XXYY, some double insertions are seen when both HDR templates are used. However, expression from the desired locus using a mixture of templates has been achieved.

Insertion at the IL2RG gene

Insertion of CARs into the IL2RG gene should result in the constitutive expression of the CAR in lymphocytes. HDR templates have been constructed to test the possibility of inserting CAR encoding sequences into the IL2RG gene.

Figure 8 shows results of nucleofecting SupT 1 cells (which originate from a male donor) with one of three guide RNAs (48_rev, 69_rev, and 80_rev) targeting the IL2RG gene and HDR templates comprising either CAT19 CAR or FMC63 CAR.

Since SupT1 cells have the karyotype XXYY, some double insertions are seen when both HDR templates are used. However, expression from the desired locus using a mixture of templates has been achieved.

Insertion at the PGK1 gene

HDR templates were designed to insert the CAT19 CAR or HA-tagged FMC63 CAR in-frame at the 5’ and the 3’ end of the PGK1 gene. The structure of these templates was similar to those described for HPRT1 (Figure 3).

Peripheral blood mononuclear cells (PBMCs) were nucleofected with HDR templates encoding the CAT19 CAR and/or HA-tagged FMC63 CAR and staining carried out with anti- CAT19 CAR idiotype and anti-HA epitope tag antibodies. The stained PBMCs were analysed by flow cytometry. Figure 9 and Figure 10 show the flow cytometric analysis of PBMCs edited to introduce the CAT19 CAR or HA-tagged FMC63 CAR coding sequences by homology-directed repair to the 5’ or 3’ end of the PGK1 gene. Integration efficiencies of between 13% and 15% were obtained, depending on the target site, with insertion at the 5’ end being more efficient than at the 3’ end of the PGK1 gene. The CAT19 CAR was successfully inserted at the 5’ and 3’ end of the PGK1 gene, with a slightly higher efficiency at the 5’ end compared with the 3’ end. As integration of the CAR encoding sequences appeared to be more efficient at the 5’ end of the gene, the two CAR encoding sequences were introduced simultaneously to cells to generate a heterogeneous pool of cells expressing either the CAT19 CAR or the HA- tagged FMC63 CAR (Figure 9). Flow cytometry confirmed the successful integration of only one CAR per cell, either CAT19 CAR or HA-tagged FMC63 CAR, as only single positive populations were present and there was no double positive population. These results validated the targeting of genes located on the X-chromosome as a means of controlling copy number.

To determine if the insertion of CAR encoding sequences to the PGK1 gene resulted in expression levels of the CAR on the cell surface high enough to elicit the cytolysis of target cells, co-culture experiments were set up using the gene edited cells and CD19 positive and CD19 negative (knockout) Raji cells. The genome edited PBMCs and Raji cells were co cultured at two different effectortarget cell ratios, 1 :1 and 1:4, for 24 and 72 hours.

Figure 11 and Figure 12 show flow cytometric analysis of co-cultures of Raji cells and genome edited PBMCs, from two independent donors, stained with antibodies recognising the T cell-specific antigens CD2 and CD3s at 24 hours (Figure 11) and 72 hours (Figure 12). Staining with this combination of antibodies enabled the clear distinction of the different populations, which were enumerated.

Figure 13 shows the percent survival of target cells normalised to the no DNA control. The data clearly show that the genome edited PBMCs expressing the CAT19 CAR or FMC63 CAR were able to lyse CD19 positive Raji cells, whereas only limited targeting of CD19 negative Raji cells was observed at the higher 1:1 ratio (effector: target). The heterogeneous population of edited PBMCs, nucleofected with both the CAT19 and FMC63 CARs, was able to lyse CD19 positive Raji cells with similar efficiency as the individual populations nucleofected with either the CAT 19 CAR or the FMC63 CAR, indicating that there was no loss in cytolytic activity when PBMCs were engineered using a heterogeneous pool of HDR templates. Targeting HPRT1 gene in PBMCs: insertion of CAT19 CAR

PBMCs from three donors were nucleofected with RNP complexes targeting exon 2 or the final exon of HPRT1 gene and HDR templates encoding the CAT19 CAR. The edited PBMCs were cultured for 7 days before staining with the anti-CAT19 idiotype antibody and analysing them by flow cytometry.

Figure 14 and Figure 15 show the flow cytometric data of the edited PBMCs. Depending on the donor, the insertion efficiency at exon 2 of the HPRT1 gene ranged from 6 to 20% (Figure 14). In contrast, the integration efficiency at the final exon of the HPRT1 gene was considerably lower and ranged from 0.6 to 2.0% (Figure 15). These data indicated that is was possible to integrate CAR encoding sequences by HDR to two independent sites in the HPRT 1 gene with integration at exon 2 being the most efficient.

Insertion at the IL-2RG: insertion of the CAT 19 CAR

To determine if it was possible to integrate CAR encoding sequences to the IL-2RG gene of PBMCs, nucleofections were carried using RNP complexes targeting the first exon of the IL- 2RG gene and the CAT19 CAR HDR template.

PBMCs edited to express the CAT19 CAR were stained with anti-CAT19 CAR idiotype antibody and analysed by flow cytometry. Figure 16 shows the flow cytometric data and the integration efficiency was found to be dependent on the donor and ranged from 10 to 42%. The results of this experiment clearly indicated that it was possible to integrate CAR encoding sequences to the IL-2RG gene and this is viable target site for the use in screening libraries of CARs to determine which one has the optimal architecture eliciting potent cytolytic activity and cytokine secretion.

Insertion at the SLC25A5: insertion of the CAT19 CAR

To determine if it was possible to integrate CAR encoding sequences to the SLC25A5 gene of PBMCs, nucleofections were carried using RNP complexes targeting the 5’ end of the SLC25A5 gene and the CAT19 CAR HDR template.

PBMCs edited to express the CAT19 CAR were stained with anti-CAT19 CAR idiotype antibody and analysed by flow cytometry. Figure 17 shows the flow cytometric data. Insertion of the anti-CD19 (CAT) CAT was observed when PBMCs were nucleofected with an SLC25A5 targeting sgRNA but not the non-targeting sgRNA. Integration efficiency was up to 46%. The results of this experiment clearly indicated that it was possible to integrate CAR encoding sequences to the SLC25A5 gene and this is viable target site for the use in screening libraries of CARs to determine which one has the optimal architecture eliciting potent cytolytic activity and cytokine secretion.

All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described methods and system of the invention will be apparent to those skilled in the art without departing from the scope and spirit of the invention. Although the invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention which are obvious to those skilled in molecular biology or related fields are intended to be within the scope of the following claims.