WO2022148736A1

WO2022148736A1 - Vectorization of muc1 t cell engager

Info

Publication number: WO2022148736A1
Application number: PCT/EP2022/050050
Authority: WO
Inventors: Jean-Baptiste Marchand; Christelle REMY
Original assignee: Transgene
Priority date: 2021-01-05
Filing date: 2022-01-04
Publication date: 2022-07-14

Abstract

The present invention is in the field of oncolytic viruses and provides a new poxvirus defective for the ribonucleotide reductase (RR) activity and engineered to express a multi-specific molecule comprising a first binding domain binding specifically to a cell antigen (Ag) present at the surface of a T lymphocyte and a second binding domain binding specifically to an antigen (Ag) present at the surface of a cancer cell or a tumor infiltrating cell. The invention also concerns a composition comprising such a poxvirus as well as the therapeutic use thereof and a method of treatment. It is particularly useful in the fields of virotherapy and immunotherapy for the treatment of cancer.

Description

VECTORIZATION of MUC1 T Cell Engager

FIELD OF THE INVENTION

BACKGROUND OF THE INVENTION

Each year, cancer is diagnosed in more than 12 million subjects worldwide. In industrialized countries, approximately one person out five will die of cancer. Although a vast number of therapies exists, they are often ineffective, especially against malignant and metastatic tumors that establish at a very early stage of the disease. In addition to the traditional cancer treatments usually provide cytotoxicity. New fields have now emerged for some decades with mixed results depending on the patients and the type of cancer.

Oncolytic virotherapy is based on replication-competent viruses which are designed to replicate within cells in division and ultimately cause cell lysis or death (Russell et al., 2012, Nat. Biotechnol. 30(7): 658-70). Several oncolytic viruses have now emerged, all of them having relative advantages and limits making them more appropriate to certain indications (see for example Harrington et al., 2019, Nat. Rev Drug Discov, 18: 689-706 and Sivanandam et al., 2019 Mol Ther 13: 93-106). Numerous preclinical and clinical studies are presently ongoing in various types of cancers to assess the therapeutic potential of oncolytic viruses armed with a variety of therapeutic genes. The first oncolytic virus to be approved by a regulatory agency was a genetically modified adenovirus named H101 (Shanghai Sunway Biotech) that gained approval in 2005 from China's State Food and Drug Administration (SFDA) for the treatment of head and neck cancer. In 2015, the herpes virus talimogen laherparepvec (T VEC or IMLYGIC; Biovex) was approved by FDA for the treatment of advanced melanoma. In addition to its oncolytic effect, this virus encodes an immunostimulatory protein called granulocyte-macrophage colony-stimulating factor (GM- CSF) for recruitment and activation of immune cells (Senzer et al, 2009, J. Clin. Oncol. 27: 5763-71). Vaccinia viruses (VV) have now entered clinical development and several strains armed with different therapeutic genes are currently evaluated (Foloppe et al., 2019 Mol Ther Oncolytics, 14:1-14; doi10.1016/j.omto.2019.03.005; Heo et al., 2013, Nat Med 19: 329-36; Zeh et al., 2015, Mol ther 23: 202-14; Mell et al., 2017, Clin. Cancer. Res 23: 5696-702).

Immunotherapy is also being actively investigated as a potential modality for treating cancers, in an attempt to boost the host’s immune system and to help the body to eradicate abnormal cells. Within immunotherapy there are multiple strategies for achieving anti-tumor responses, one of which is the use of T cell engagers (TCEs). Generally speaking, these molecules are members of the multi-specific antibody family having the ability to bind target antigens located on different cells including T cells. Particularly useful are TCEs with one specificity directed a T cell-specific molecule, usually CD3, while the second specificity recognizes a tumor-associated antigen (see e.g. Huehls et al., 2015, Immunol Cell Biol. 93(3): 290-6). This double specificity allows a TCE to physically link a T cell to a tumor cell, thereby stimulating T cell activation, cytokine production and ultimately killing of the cell expressing the target antigen. Bispecific TCEs can take a variety of formats, which vary in size and complexity. The smallest and simplest formats are fusion proteins comprising single chain variable fragments (scFvs) of two different antibodies on a single peptide chain connected via flexible linkers. Several TCEs have now been developed that target various tumor-associated antigens (TAA). Examples of bi-specific TCE antibodies currently approved or undergoing clinical trials include for example Blinatumomab (Blyncyto®) which targets CD19 for the treatment of non-Hodkin's lymphoma and acute lymphoblastic leukemia and Solitumab which targets EpCAM for treating gastrointestinal and lung cancers (Brischwein et al., 2006, Mol Immunol 43(8): 1129-43).

The transmembrane glycoprotein Mucin 1 (MUC1) is also an attractive target due to the fact that it is overexpressed in a variety of epithelial cancers, and plays a crucial role in cancer progression. Tumor-associated MUC1 differs from the MUC1 expressed in normal cells in various aspects including structural variability in the number of tandem repeat (VNTR), altered glycosylation (the tumor-associated MUC1 contains a preponderance of shorter glycan that unmask epitopes within the VNTR), alternative splicing and function. Antibodies against the tumoral forms of MUC1 were generated including the so-called SM3 Paul et al. (2000, Hum Gene Ther, 10(11): 1417-28) and HMFG2 (Wlkie et al., 2008, J. Immunol. 180 (7): 4901-9). Arming strategies of oncolytic viruses with antibodies and notably bispecific TCEs is now being investigated with the aim to combine virus-induced tumor destruction and TCE- induced anti-tumor T cell responses (see e.g., Yu et al., 2017, Cancer trans Med 3(4): 122- 32; Yu et al., 2014, Mol Ther. 22(1): 102-11 ; WO2016146894, WO2014/138314). MUC1- targeting TCEs have been mentioned into the art (e.g., WO2018/178047, WO2017/167919, WO2016/165632, WO2016/087245). However, clinical data have not been reported yet and MUC1 based therapy presents several obstacles. In particular, tumor-derived MUC1 can impair T cell growth and shield transformed cells from killing by NK and T cells (Agrawal et al., 1998, Nat. Med. 4: 43-9) and the shedding of soluble MUC1 and the steric inhibition by MUC1 may compromise antibody binding on tumor cells. As for most TAA, the specificity of expression of modified or abnormal MUC1 on surface tumor cells is not exclusive and some cells in healthy tissues may express also low density of this abnormal antigen. Therefore, TCE against MUC1 could induce the damage of healthy tissues (so called: “on-target off tumor”) and have a limited tolerance.

Moreover, several factors limiting the clinical efficacy of oncolytic virotherapy have now been identified (Vaha-Koskela et al., 2014, Biomedecines 2(2): 163-94) including among others the incomplete virus penetration and diffusion within the tumors (Karnerva et al., 2005, Gene Ther 12(1) 87-94; Kangasniemi et al., 2006, Clin Cancer Res. 12(10): 3137-44), the limited capacity of oncolytic viruses to reach and infect continuously proliferating cancer cells, the host’s innate antiviral defences (e.g. most oncolytic viruses are sensitive to type I IFN), and neutralizing humoral immunity that may arise after multiple virus administrations. Accumulating evidence, also, suggests the propensity of malignant cells to become resistant to oncolytic virotherapy, thus limiting its potential (Vaha-Koskela et al., 2006, Cancer Res 66: 7185-94).

Therefore, there is still a need to further develop alternative strategies to the existing ones and, in particular those mentioned above, for delivering TCE in situ and provide effective cancer therapies while reducing toxic responses for the subject. The present disclosure provides a solution to this long-felt need.

The present invention is based on the generation of a vaccinia virus expressing a bispecific molecule comprising an anti-CD3 antibody fused to an anti-MUC1 antibody (VV- TCE). Cells infected with the TCE-expressing vaccinia virus construct were shown to express the TCE by Western blot and ELISA analysis. Unexpectedly, the inventors discovered that tandem scFv (TscFv) format is particularly suited for providing high expression levels in supernatants of infected cells and this observation works for both monomeric and dimeric formats. Although all TCE formats bind MUC1 peptides, TscFv secreted in cell culture showed higher efficacy than diabody format in line with the expression data. Interestingly, TCE expression does not impair the viral replication and oncolytic activity. Flow cytometry analysis confirmed that all TCE formats are able to specifically bind CD3 present at the lymphocyte surface and MUC1 present at the surface of two different MUC1-positive tumor cells (HeLa and T147D cells) although intensity of labelling is stronger for TscFv than for diabodies. Importantly, TCE produced by two tumor cell lines infected with TscFv-expressing VV are able to trigger the killing of MUC1 positive cells by CD3-expressing effector immune cells (such as CD8 lymphocyte). This lytic activity is specific since not observed with MUC1- negative cells and is strictly dependent of the presence of CD8 lymphocytes. Therefore, the VV-TCE constructs disclosed herein are particularly useful to treat cancer and other proliferative diseases.

This technical problem is solved by the provision of the embodiments as defined in the claims. Other and further aspects, features and advantages of the present invention will be apparent from the following description of the presently preferred embodiments of the invention. These embodiments are given for the purpose of disclosure.

SUMMARY OF THE INVENTION

In a first aspect, the present invention relates to a poxvirus comprising a nucleic acid molecule encoding a multi-specific molecule comprising at least a first domain binding specifically to a cell antigen present at the surface of a T cell and a second domain binding specifically to an antigen (Ag) present at the surface of a cancer cell or a tumor infiltrating cell, wherein said poxvirus is defective for ribonucleotide reductase (RR) activity. In one embodiment the poxvirus is an oncolytic vaccinia virus preferably selected from the group of vaccinia viruses consisting of Western Reserve (WR), Elstree, Wyeth, Lister, Tian Tan, LIVP and Copenhagen strains. In a preferred embodiment, said vaccinia virus is defective for both thymidine kinase (TK) and ribonucleotide reductase (RR) activities.

In another embodiment said multi-specific molecule is a bispecific molecule comprising a first and a second antigen binding domains having different binding specificities. In preferred embodiments, each of the first and the second antigen binding domains is an antibody, preferably comprising a heavy chain variable region (VH) and a light chain variable region (VL). In other preferred embodiments, the first antigen binding domain in an antibody that specifically binds to CD3 and the second antigen binding domain in an antibody that specifically binds to MUC1. In other preferred embodiments, the multi-specific molecule is selected from the group consisting of diabodies, triabodies, tetrabodies, minibodies, nanobodies and tandem scFv, with a preference for a tandem scFv, a dimeric tandem scFv and a diabody. In some embodiments, the poxvirus comprising the nucleic acid molecule encoding the multi-specific molecule further comprises a nucleic acid molecule encoding one or more polypeptide of interest such as an immunomodulatory polypeptide. In preferred embodiment, said immunomodulatory polypeptide is a flt3L polypeptide. In another aspect, the present invention also relates to a method for producing said poxvirus comprising the steps of a) preparing a producer cell, b) transfecting or infecting the prepared producer cell with the poxvirus, c) culturing the transfected or infected producer cell under suitable conditions so as to allow the production of the virus, d) recovering the produced virus from the culture of said producer cell and optionally e) purifying said recovered virus. In a further aspect, the present invention also relates to a composition comprising a therapeutically effective amount of said poxvirus and a pharmaceutically acceptable vehicle. In preferred embodiments, said composition is formulated for intravenous, subcutaneous, intramuscular or intratumoral administration.

In still a further aspect, the present invention also relates to the poxvirus or the composition for use for treating a cancer. The present invention also provides a method of treating a cancer comprising administering to a subject said poxvirus or composition. In preferred embodiments, the cancer is a MUC1 -positive cancer, preferably selected from the group consisting of lung cancer, breast cancer, prostate cancer, pancreas cancer, gastric cancer, ovary cancer, fallopian tubes cancer, colorectal cancer and kidney cancer.

DETAILED DESCRIPTION

General definitions Several definitions are provided here that will assist in the understanding of the invention. However, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. All references cited herein are incorporated by reference in their entirety.

As used throughout the entire application, the terms "a" and "an" are used in the sense that they mean "at least one", "at least a first", "one or more" or "a plurality" of the referenced components or steps, unless the context clearly dictates otherwise. For example, the term "a cell" includes a plurality of cells, including mixtures thereof. The term “one or more” refers to either one or a number above one (e.g. 2, 3, 4, 5, etc).

The term "and/or" wherever used herein includes the meaning of "and", "or" and "all or any other combination of the elements connected by said term". The term "about" or "approximately" as used herein means within 20%, preferably within 10%, and more preferably within 5% of a given value or range.

The term “for example”, “for instance” or their corresponding abbreviation “e.g.,” means that the specific terms recited are representative examples and embodiments of the disclosure that are not intended to be limited to the specific examples referenced or cited unless explicitly stated otherwise.

As used herein, when used to define products, compositions and methods, the term "comprising" (and any form of comprising, such as "comprise" and "comprises"), "having" (and any form of having, such as "have" and "has"), "including" (and any form of including, such as "includes" and "include") or "containing" (and any form of containing, such as "contains" and "contain") are open-ended and do not exclude additional, unrecited elements or method steps. Thus, a polypeptide "comprises" an amino acid sequence when the amino acid sequence might be part of the final amino acid sequence of the polypeptide. "Consisting of means excluding other components or steps of any essential significance. Thus, a composition consisting of the recited components would not exclude trace contaminants and pharmaceutically acceptable carriers. A polypeptide "consisting of” an amino acid sequence refers to the presence of such an amino acid sequence with eventually only a few additional and non-essential amino acid residues. It is nevertheless preferred that the polypeptide does not contain any amino acids but the recited amino acid sequence. In the present description, the term “comprising” (especially when referring to a specific sequence) may be replaced with consisting of, if required.

Within the context of the present invention, the terms “nucleic acid”, “nucleic acid molecule”, “polynucleotide”, “nucleotide sequence” and "nucleic acid sequence" are used interchangeably and define a polymer of any length of either polydeoxyribonucleotides (DNA) or polyribonucleotides (RNA). They encompass single or double-stranded, linear or circular, natural or synthetic, modified (e.g. genetically modified polynucleotides; optimized polynucleotides), or unmodified polynucleotides, sense or antisense polynucleotides, chimeric mixture (e.g. RNA-DNA hybrids). Exemplary DNA nucleic acids include without limitation, complementary DNA (cDNA), genomic DNA, plasmid DNA, vectors, viral DNA (e.g. viral genomes, viral vectors), oligonucleotides, probes, primers, coding DNA, non-coding DNA, or any fragment thereof etc. Exemplary RNA nucleic acids include, without limitation, messenger RNA (mRNA), precursor messenger RNA (pre-mRNA), coding RNA, non-coding RNA, etc. Nucleic acid sequences described herein may be synthesized by standard methods known in the art, e.g., by use of an automated DNA synthesizer (such as those that are commercially available from Biosearch, Applied Biosystems, etc.) or obtained from a naturally occurring source (e.g. a genome, cDNA, etc.) or an artificial source (such as a commercially available library, a plasmid, etc.) using molecular biology techniques well known in the art (e.g. cloning, PCR, etc).

The term “nucleic acid molecule encoding” refers to a so called “coding sequence” (e.g., a DNA sequence) that is transcribed and translated into a polypeptide in a cell in vitro or in vivo when placed under the control of appropriate regulatory sequences.

“Regulatory sequences” can include nucleotide sequences located upstream (5’ non coding sequences), within, or downstream (3’ non-coding sequences) of a coding sequence, and which influence the transcription, RNA processing or stability, or translation of the associated coding sequence. Regulatory sequences may include promoters, translation leader sequences, introns, polyadenylation recognition sequences, RNA processing sites, effector binding sites and stem-loop structures. The boundaries of the coding sequence are determined by a start codon at the 5’ (amino) terminus and a translation stop codon at the 3’ (carboxyl) terminus. A promoter will usually be located 5’ to the coding sequence and polyadenylation signal and transcription termination sequences will usually be located 3’ to the coding sequence.

The term “polypeptide” “peptide” and “protein” are used interchangeably to refer to a polymer of at least nine amino acid residues bonded via peptide bonds regardless of its size and the presence or not of post-translational components (e.g. glycosylation). No limitation is placed on the maximum number of amino acids comprised in a polypeptide. As a general indication, the term refers to both short polymers (typically having less than 50 amino acid residues also referred as peptides) and to longer polymers (typically 50 amino acid residues or more which may be designated in the art as polypeptide or protein). This term encompasses native polypeptides, modified polypeptides (also designated derivatives, analogs, variants or mutants), polypeptide fragments, polypeptide multimers (e.g. dimers), fusion polypeptides among others. The term also refers to a recombinant polypeptide expressed from a polynucleotide sequence which encodes said polypeptide. Typically, this involves transcription of the encoding nucleic acid into a mRNA sequence and translation thereof by the ribosomal machinery of the cell to which the polynucleotide sequence is delivered. The term “identity” refers to an amino acid to amino acid or nucleotide to nucleotide correspondence between two polypeptide or nucleic acid sequences. The percentage of identity between two sequences is a function of the number of identical positions shared by the sequences, taking into account the number of gaps which need to be introduced for optimal alignment and the length of each gap. Various computer programs and mathematical algorithms are available in the art to determine the percentage of identity between amino acid sequences, such as for example the Blast program available at NCBI or ALIGN in Atlas of Protein Sequence and Structure (Dayhoffed, 1981, Suppl., 3: 482-9). Programs for determining identity between nucleotide sequences are also available in specialized data base (e.g. Genbank, the Wisconsin Sequence Analysis Package, BESTFIT, FASTA and GAP programs). Those skilled in the art can determine appropriate parameters for measuring alignment including any algorithms needed to achieve maximum alignment over the sequences to be compared. For illustrative purposes, “at least 80% identity” means 80% or above (e.g., 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100%) whereas “at least 90%” refers to 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99% or 100% identity and “at least 95%” to 95%, 96%, 97%, 98%, 99% or 100% identity.

As used herein, the term “isolated” refers to a component (e.g. polypeptide, nucleic acid molecule, virus, etc.), that is removed from its natural environment (i.e. separated from at least one other component(s) with which it is naturally associated or found in nature). For example, a nucleotide sequence is isolated when it is separated of sequences normally associated with it in nature (e.g. dissociated from a genome) but it can be associated with heterologous sequences. A synthetic component is isolated by nature.

The terms “virus”, ‘viral particle”, “viral vector” and virion” are used interchangeably and are to be understood broadly as meaning a vehicle comprising at least one element of a wild-type virus genome that may be packaged into a viral particle. The term “virus” as used in the context of poxvirus or any other virus mentioned herein encompasses the viral genome as well as the viral particle (encapsided and/or enveloped genome), e.g., obtained from a wild-type virus or a modified virus. Suitable modifications encompass, e.g., any modification(s) in one or more viral genes as well as chimeric virus comprising fragments from different virus sources. The term "obtained from", “originating” or “originate” and any equivalent thereof is used to identify the original source of a component (e.g. polypeptide, nucleic acid molecule, virus, vector, etc.,) but is not meant to limit the method by which the component is made which can be, for example, by chemical synthesis or recombinant means.

As used herein, the term “host cell” should be understood broadly without any limitation concerning particular organization in tissue, organ, or isolated cells. Such cells may be of a unique type of cells or a group of different types of cells such as cultured cell lines, primary cells and dividing cells. In the context of the invention, the term “host cells” preferably refers to eukaryotic cells such as mammalian (e.g. human or non-human) cells as well as cells capable of producing the virus described herein (e.g., avian cells). This term also includes cells which can be or has been the recipient of the virus as well as progeny of such cells. The term “subject” generally refers to an organism for whom any virus, composition and method described herein is needed or may be beneficial. Typically, the organism is a mammal, particularly a mammal selected from the group consisting of domestic animals (dogs, cats, etc.), farm animals (cows, pigs, sheep, etc.), sport animals (horses, etc), and primates (simian, human, etc). Preferably, the subject is a human who has been diagnosed as having or at risk of having a cancer. The terms “subject” and “patients” may be used interchangeably when referring to a human organism and encompasses male and female that have been diagnosed with a cancer requiring treatment, or suspected of having a cancer, or at risk of developing such a cancer. The subject may be a newborn, an infant, a young adult, an adult or an eldery.

The term “treatment” (and any form of treatment such as “treating”, “treat”, etc.,) as used herein encompasses prophylaxis and/or therapy, eventually in association with conventional therapeutic modalities. Typically, « prophylaxis » refers to prevention, e.g. to prevent, delay the onset or decrease the severity of the first occurrence or relapse of at least one clinical or biochemical symptom usually associated with the targeted pathological condition whereas therapy refers to a pathological condition with the purpose to improve at least one clinical or biochemical symptom (e.g., size of tumor, expression level of associated biomarker(s), stage progression...), to slow down or control the progression of the targeted pathological condition, symptom(s) thereof, in the subject treated in accordance with the modalities described herein.

The term “administering” (or any form of administration such as “administered”, etc.,) as used herein refers to the delivery to a subject of a component such as the virus, composition, method described herein according to the modalities described herein.

The term “combination” or “association” as used herein refers to any arrangement possible of various components (e.g. a poxvirus described herein and one or more substance effective in anticancer therapy). Such a combination encompasses the cases where the individual components are administered to the subject as a single composition (together in the same composition) or separately (i.e. dissociate arrangement), in which case the two or more components may be administered concurrently, sequentially, in an interspersed manner or in any combination of these types of administration. It is appreciated that optimal concentration of each component of the combination can be determined by the artisan skilled in the art.

In a first aspect, the present invention provides a virus comprising a nucleic acid molecule encoding a multi-specific molecule comprising at least a domain binding specifically to a cell antigen present at the surface of a T cell (also called first antigen-binding domain thereafter) and a domain binding specifically to an antigen (Ag) present at the surface of a cancer cell or a tumor infiltrating cell (also called second antigen-binding domain thereafter), wherein said virus is preferably defective for RR activity. Recombinant virus encoding a multi-specific molecule

In some embodiments, the virus of the present invention is preferably selected from the group of adenovirus (Ad), herpes simplex virus (HSV), poxvirus, vesicular stomatitis virus (VSV), parvovirus, myxoma virus (MYXV), Newcastle disease virus (NDV), reovirus, Seneca valley virus (SW) morbillivirus virus, rhabdovirus and Sindbis virus (SINV). In preferred embodiments, the virus is a poxvirus.

As used herein, the term “poxvirus” or “poxviral” refers to any member of the Poxviridae family identified at present time or being identified afterwards or any modified version thereof that is infectious for one or more mammalian (e.g. human cells) or avian cells. The Poxviridae family is a broad family of enveloped DNA viruses containing a double- stranded genome which is subdivided in Entomopoxvirinae infecting insects and Chordopoxvirinae infecting a large range of vertebrates. The latter is subdivided in 18 genera including Avipoxvirus, Capripoxvirus, Leporipoxvirus, Molluscipoxvirus, Orthopoxvirus, Parapoxvirus, Suipoxvirus, Cervidpoxvirus, Yatapoxvirus, etc. Contrary to other viruses, poxviruses remain in the cell cytoplasm for the duration of the infectious cycle, from the time the virus enters the cell until the progeny viruses exit through the plasma membrane. Therefore, all the proteins required for DNA replication and RNA synthesis and maturation are included within the viral genome, including RNA polymerase, enzymes for RNA capping, methylation, polyadenylation and transcription factors.

In some embodiments, the poxvirus of the present invention is engineered from an Orthopoxvirus, with a specific preference for a Vaccinia virus (VV). The VV genome is approximately 200 kb long and encodes approximately 250 genes. For general information, the ends of the genome have a terminal hairpin loop with several inverted terminal repeat sequences which are believed to be involved in genomic replication. In general, the genome is arranged such that conserved genes used for RNA and DNA synthesis, protein processing, virion assembly and structural proteins are located near the central region of the genome whereas the terminal regions of the genome encode genes affecting host range, virulence and interaction with the host immune system. (Gubser et al. (2004) Poxvirus genomes: a phylogenic analysis. J. Gen. Virol. 85:105-117).

In some embodiments, the poxvirus is a vaccinia virus. Any VV strain can be used in the context of the present invention including those cited below and even Modified Virus Ankara (MVA). In some embodiments, the virus of the present invention is oncolytic. Particularly preferred is an oncolytic vaccinia virus preferably selected from the group of vaccinia viruses consisting of Western Reserve (WR), Elstree, Wyeth, Lister, Tian Tan, LIVP and Copenhagen strains. The genomic sequence of many poxviruses and the encoded open reading frames (ORFs) are well known in the art and available in specialized database such as GenBank (see e.g. accession numbers NC_006998 and M35027 for Western Reserve and Copenhagen strains). The gene nomenclature used herein is that of Copenhagen vaccinia strain. It is also used herein for the homologous genes of other Vaccinia virus unless otherwise indicated since gene nomenclature may be different according to the strain but correspondence between Copenhagen and other vaccinia strains are generally available in the literature.

The term “oncolytic” qualifies a virus displaying a preferred propensity to infect and kill dividing cells (e.g. cancer cells) as compared to non-dividing cells (e.g., normal cells). Oncolytic activity may occur through direct cytotoxic activity, e.g., caused by preferential infection, replication in and destruction of the dividing cells or by indirect cytotoxic activity (e.g. cell apoptosis, induction or stimulation of the host’s immune response, which, in addition to destroying existing cancer cells, can establish lasting immunity). Oncolytic activity can be detected by known methods, including, but not limited to, apoptosis assays, such as TUNEL staining, inhibition of cell proliferation, reduction of cell viability following virus infection and/or by detecting a reduction in tumor size before and after treatment.

RR-defective poxvirus In some embodiments, the poxvirus of the invention comprises a genome which has been modified by the man’s hands so as to include one or more modifications in the nucleotide sequence as compared to the wild-type sequence.

In some embodiments, the poxvirus of the present invention comprises one or more genomic modification(s) which result in a poxvirus defective for one or more viral gene product(s). The term “defective” as used herein, denotes the lack of synthesis or the synthesis of a protein unable to ensure the activity of the protein produced under normal conditions by the unmodified viral gene(s). Such a defective character typically results from inactivating mutation(s) within the viral gene sequence or its regulatory elements. Inactivating mutation(s) encompass deletion, mutation and/or substitution of one or more nucleotide(s) (contiguous or not). Such mutation(s) can be made in a number of ways known to those skilled in the art using conventional recombinant techniques. Exemplary modifications are those inactivating one or more viral genes involved in DNA metabolism, virulence (e.g., to reinforce virulence in cancer cells) or IFN pathway (see e.g. Guse et al., 2011 , Expert Opinion Biol. Ther.11 (5):595- 608). Determination if a given poxvirus is defective or not for a viral function is within the reach of the skilled artisan using the information given herein and the general knowledge in the art (e.g. hybridization, PCR techniques, sequencing, enzyme assay, etc). In some embodiments, the poxvirus of the invention is defective for ribonucleotide reductase (RR) activity resulting from inactivating mutation(s) in at least one gene or both genes encoding RR enzyme. The genome of many poxviruses, including vaccinia virus, encodes ribonuclease reductase (RR), an enzyme that catalyses the conversion of ribonucleotides to deoxyribonucleotides, that represents a crucial step for DNA biosynthesis and thus for viral replication. Most mammalian cells (e.g., primary cells) exist in a terminally differentiated state with a low level of ribonucleotide reductase. In contrast, many cancer cells have elevated ribonucleotide reductase levels. Therefore, viruses lacking a functional ribonuclease reductase can survive in cancer cells but are unable to replicate in normal cells. The ribonucleotide reductase enzyme comprises a large subunit (R1) and a small subunit (R2), which are both required for enzymatic activity. The large subunit is encoded by the I4L gene and the small subunit is encoded by the F4L gene. Sequences for the I4L and F4L genes and their location in the poxvirus genome are available in public databases. In the context of the invention, the poxvirus can be modified either in the I4L gene (encoding the R1 large subunit) or in the F4L gene (encoding the R2 small subunit) or in both I4L and F4L genes (see e.g. W02009/065546). Thus, a poxvirus that includes one or more inactivating mutation(s) (e.g., a partial or total deletion of at least one the RR-encoding gene(s) or insertion of a heterologous nucleic acid molecule in the I4L/F4L gene locus) resulting in a non functional RR, may demonstrate increased viral selectivity for cancer cells and an attenuated ability to replicate in normal tissues. In preferred embodiments, the poxvirus described herein is a vaccinia virus (in particular a oncolytic VV) defective for RR in which the I4L gene encoding RR is either partially or totally deleted or inactivated by insertion of a heterologous nucleic acid molecule.

In some embodiments, the poxvirus of the invention may further comprise one or more additional genomic modifications. In preferred embodiments, the poxvirus is further defective for thymidine kinase (TK) activity resulting from inactivating mutation(s) in the TK-encoding gene. TK is encoded by the J2R gene and involved in the synthesis of deoxyribonucleotides. TK is needed for viral replication in normal cells as these cells have generally low concentration of nucleotides whereas it is dispensable in dividing cells which contain high nucleotide concentration. In other words, viruses lacking a functional TK can survive in cancer cells but are unable to undergo replication in normal cells. Thus, a poxvirus that includes one or more inactivating mutation(s) (e.g., a partial or total deletion of at least one the TK-encoding gene or insertion of a heterologous nucleic acid molecule in the J2R gene locus) resulting in a non-functional TK, may demonstrate increased viral selectivity for cancer cells and an attenuated ability to replicate in normal tissues. In particularly preferred embodiments, the poxvirus of the present invention is defective for both TK and RR activities resulting from inactivating mutations in both the J2R and the I4L/F4L loci carried by the viral genome (e.g. as described in W02009/065546 and Foloppe et al. , 2008, Gene Ther., 15: 1361-71). More preferably, the poxvirus described herein is a vaccinia virus (in particular a oncolytic VV) defective for both TK and RR activities in which the RR encoding I4L gene and the TK-encoding J2R gene are independently either partially or totally deleted or inactivated by insertion of a heterologous nucleic acid molecule.

In some embodiments, the poxvirus is further defective for M2 activity resulting from inactivating mutation(s) in the M2-encoding gene ( M2L locus). M2 was reported as a protein retained in endoplasmic reticulum acting as an inhibitor of the NfKb pathway and involved in uncoating of the virus (Liu et al., 2018, J. Virol. 92(7) e02152-17). More recently, M2 protein was assigned a new property that is a capacity of binding to CD80 and CD86 co-stimulatory molecules and thus immunosuppressing immune response mediated by CD80 and/or CD86 pathways (W02020/136232). M2L-defective poxviruses are expected to stimulate or improve immune response, especially the lymphocyte-mediated response, against an antigen (W02020/136235). In some embodiments, the poxvirus described herein is a vaccinia virus (in particular a oncolytic W) defective for TK, RR and M2 activities in which the TK-encoding J2R gene the RR-encoding I4L gene and the M2-encoding M2L gene are independently either partially or totally deleted or inactivated by insertion of a heterologous nucleic acid molecule.

In some embodiments, the poxvirus of the present invention may further comprise one or more additional modifications. For example, it may be defective for one or more additional virus activity. A representative example is a poxvirus further defective for VGF (for vaccinia virus growth factor) activity resulting from inactivation of the VGF-encoding gene. VGF is a secreted protein which is expressed early after cell infection and its function seems important for virus spread in normal cells. Another example is the inactivation of the A56R gene coding for hemagglutinin (Zhang et al., 2007, Cancer Res. 67: 10038-46). Inactivation of interferon modulating gene(s) may also be advantageous (e.g. the B8R or B18R gene) or the caspase- 1 inhibitor B13R gene. Inactivation of the F2L gene may also be envisaged in the context of this invention. F2L locus encodes the viral dUTPase involved in both maintaining the fidelity of DNA replication and providing the precursor for the production of TMP by thymidylate synthase (W02009/065547).

Antigen binding domains In some embodiments, the poxvirus of the present invention comprises a nucleic acid molecule (inserted at a suitable location of the viral genome) encoding a multi-specific molecule. As used herein, the term “multi-specific molecule” refers to a molecule (e.g., a polypeptide) which displays a plurality (i.e. more than one) of binding specificities to different target antigens. In some embodiments, the multi-specific molecule comprises a plurality of antigen binding domains, for example two (bispecific) or three (tri-specific) or four (tetra- specific) binding domains, each specific for a particular target antigen (Ag). In other words, each antigen-binding domain specifically binds to a particular antigen (or epitope thereof) relative to other available molecules (or other antigens). The term "antigen" or "Ag" refers to a molecule that provokes an immune response and this term as used herein also encompasses Ag fragments such as immunogenic domains and epitopes and Ag derivatives as long as such antigens retain their ability to induce an immune response.

In preferred embodiments, the multi-specific molecule is a bispecific molecule comprising two non-identical antigen binding domains having different binding specificities, i.e., a first antigen binding domain having binding specificity for a first antigen (or epitope thereof) and a second antigen binding domain having binding specificity for a second antigen (or epitope thereof). Preferably, the two antigens are present (e.g. expressed) on different types of cells of a given organism although target antigens originating from different species may also be envisaged. For example, the first binding domain may have specificity for an Ag of a particular organism or source and the second binding domain may have specificity for an Ag of a different organism or source. For illustrative purposes, the multi-specific molecule may target a human antigen (e.g. human CD3) and a viral antigen (e.g. oncogenic papillomavirus E6 or E7 antigens).

In addition, the various antigen binding domains of the multi-specific molecule may be obtained or generated from the same or from different species (i.e. a chimeric multi-specific molecule). For illustrative purpose, the first binding domain may originate from a human source (e.g. a human or humanized antibody) while the second binding domain is obtained/generated from an animal source (e.g. a lama antibody).

By “specific” or “specifically”, it is meant that each antigen-binding domain of the multi specific molecule binds preferentially or with higher affinity to the target antigen or binds with greater affinity to the target antigen than to other molecules. Specific binding may refer to non-covalent or covalent preferential binding to the target antigen. In some embodiments, the affinity of each antigen binding domain of the multi-specific molecule for its target antigen to which it specifically binds is characterized by a KD (dissociation constant) of 10⁵M or less (e.g., 10⁶M, 10⁷ M or less, etc). "Affinity" refers to the strength of binding, increased binding affinity being correlated with a lower KD. In some embodiments, the multi-specific molecule of the present disclosure comprises at least a first antigen-binding domain specific for a cell antigen present at the surface of a T cell (first antigen binding domain) and a second antigen-binding domain specific for an antigen present at the surface of a cancer cell or a tumor infiltrating cell (second antigen binding domain). In preferred embodiments, at least one or both of the first and the second antigen-binding domains is/are an antibody. The term “antibody” as used herein encompass full length antibody such as immunoglobulins as well as fragments and derivatives of an antibody provided that such fragments or derivatives retain an ability to bind to or interact specifically with the same antigen (in particular the same epitope) as the parent antibody from which they derive.

The term “antibody” as used herein also refers to monomeric or multimeric (e.g. dimeric) formats as long as such formats retain the ability to bind to or interact specifically with the same antigen (or epitope thereof) as the parent antibody. “Multimeric” refers to the capacity of two or more polypeptide chains to form homo or hetero multimers.

For general purposes, full length antibodies (i.e. immunoglobulins) comprise four polypeptide chains, two heavy (HC or H) chains and two light (LC or L) chains inter-connected by disulfide bonds, that can form multimers. The term "antibody heavy chain" and "antibody light chain," refers respectively to the larger and smaller of the two types of polypeptide chains present in antibody molecules in their naturally occurring conformations. Generally speaking, an antibody heavy chain comprises a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region. The heavy chain constant region comprises three domains, CH1, CH2 and CH3. Each light chain comprises a light chain variable region (abbreviated herein as VL) and a light chain constant region. The light chain constant region comprises one domain (CL). The VH and VL regions usually comprise regions of hypervariability, termed complementarity determining regions (CDRs) interspersed with regions that are more conserved, termed framework regions (FR).

Antibodies generally comprise six complementarity-determining regions CDRs; three in the heavy chain variable (VH) region: HC-CDR1, HC-CDR2 and HC-CDR3, and three in the light chain variable (VL) region: LC-CDR1, LC-CDR2, and LC-CDR3. The six CDRs together define the paratope of the antibody, which is the part of the antibody which binds to the target antigen. In other words, the CDRs confer target binding specificity to the antibody. The V_H and V_L regions comprise framework regions (FRs) either side of each CDR, arranged from amino-terminus to carboxy-terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. However, in some instances (e.g., camelid antibodies) even only three CDRs can have the ability to recognize and bind the target, antigen. The precise amino acid sequence boundaries of a given CDR can be determined using any of a number of well- known schemes, including those described by Kabat et al. (1991, Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD), Chothia et al. (1987, J. Mol. Biol. 196:901-17; Al-Lazikani et al., 1997, JMB 273: 927- 48), and VBASE2 schemes (Retter et al., 2005, Nucl. Acids Res. 33 (1): D671-D674) as well as by the international IMGT (ImMunoGeneTics) information system which uses the IMGT V- DOMAIN numbering rules as described in Lefranc et al. (2003, Dev. Comp. Immunol. 27:55- 77).

The term “antibody” as used herein also refers to a variety of antibody formats including those cited above such as antibody fragments and antibody derivatives as long as they bind the same antigen as the parent antibody. As used herein, an antibody fragment comprises a part but not all the elements present in a full-length antibody. A derivative of an antibody usually has a different amino acid sequence than the parent antibody. In preferred embodiments, such antibody fragments or derivatives comprise the same CDRs as the parent antibody but may differ in the remaining sequences of the variable regions. These antibody fragments and derivatives are generated using conventional techniques known in the art. Preferably, the antibody fragments or derivatives for use herein comprise at least a heavy chain variable region.

In some embodiments, each of the (first and second) antigen-binding domains is an antibody fragment. Suitable antibody fragments for use as an antigen-binding domain include without limitation Fab, Fab’, F(ab’)2, Fd, single-chain Fv (scFv), disulfide-linked Fvs (sdFv), scFab, dAb, single domain antibody fragment (sdAb), single domain antibodies from camelids (also referred to VHH) and minimal recognition units (see for example Holliger and Hudson, 2005, Nature Biotech. 23(9) : 1126-1136; Adair and Lawson, 2005, Drug Design Reviews 2(3), 209-217) as long as they display binding to the relevant target antigen. Such fragments are well known in the art. For illustrative purposes, a Fab fragment contains the variable domain of the light chain (VL) with the constant domain of the light chain (CL) and the variable domain of the heavy chain (VH) with the first constant domain (CH1) of the heavy chain. Fab' fragments differ from Fab fragments by the addition of a few residues at the carboxyl terminus of the heavy chain CH1 domain including one or more cysteines from the antibody hinge region. Fab and F(ab') fragments lack the Fc fragment of intact antibody. A F(ab’)2 is a composed of 2 Fab fragments linked by a disulfide bridge at the hinge region. A scFab fragment comprises VH-CH1 and VL-CL fused in a single polypeptide chain. A Fd fragment consists of the variable region and the first constant domain CH1 of the heavy chain. A Fv fragment consists of the heavy chain and the light chain variable regions of an antibody. A "single-chain Fv" (scFv) comprises a V_H and a V_L of an antibody fused in a single polypeptide chain, where the V_H and a V_L are contiguously linked via a short flexible polypeptide linker which enables the scFv to form the desired structure for target binding as described herein. Unless specified, a scFv may have the V_L and V_H variable regions in either order, e.g., with respect to the N-terminal and C-terminal ends of the polypeptide, the scFv may comprise VL- linker-VH or may comprise VH-linker-VL. Each variable domain may be of any size or amino acid composition and will generally comprise at least one CDR (preferably three CDRs) which is adjacent to or in frame with one or more framework sequences.

The methods for creating and manufacturing such antibody fragments are well known in the art (see for example Verma et al., 1998, J. Immunol. Methods, 216:165-181). For example, they may be derived from full antibody molecules using any suitable standard techniques such as proteolytic digestion or genetically engineering techniques involving the manipulation and expression of nucleic acid sequences encoding the relevant parts of the full antibody molecule. Such nucleic acid sequences are readily available from, e.g., commercial sources, DNA libraries (including, e.g., phage-antibody libraries), or can be synthesized. The nucleic acid sequences may be sequenced and manipulated chemically or by using molecular biology techniques, to arrange one or more variable and/or constant domains into a suitable format, or to introduce codons, create cysteine residues, modify, add or delete amino acids, etc.

In addition to the format (full-length antibody, fragments and/or derivatives thereof as disclosed above), the term “antibody” encompasses a variety of origins, for example, animal origin (e.g., murine, camelid, primate, etc.) or human origin, humanized, synthetic as well as chimeric . A "chimeric" antibody may be engineered to have variable sequences derived from a non-human antibody, such as a rat or a mouse antibody, and constant regions derived from a human antibody. A humanized antibody will comprise variable domains in which all or substantially all of the CDRs correspond to those of a non-human immunoglobulin and the FR regions are those of a human immunoglobulin. Methods of antibody humanization are known in the art (see e.g., US5,530,101; US5,585,089; US5,565,332; US5,693,761; US5,693,762; US6,180,370, EP239400; and WO91/09967). In some embodiments, an antigen-binding domain comprised in the multi-specific molecule of the present disclosure comprises at least one antibody variable region. In preferred embodiments, each of the first and second antigen-binding domains comprises a heavy chain variable region (VH) and preferably a VH region and a light chain variable region (VL). Preferably, the antigen-binding domains for use herein do not comprise all element of an immunoglobulin, notably some or all of the constant domain of an antibody. The VH and the V_L regions may be fused to each other with a peptide linker or by separate polypeptides. In the context of the invention, the V_H region may be N-terminal or C-terminal to the V_L region.

Peptide Linkers

As disclosed herein, the present disclosure contemplates the presence of one or more linker(s), especially to connect the various antigen binding domains comprised in the multi- specific molecule as well as to connect the antibody regions (e.g. V_H and V_L regions) comprised in each antigen binding domains. Linker peptides are known to the skilled person and the present disclosure encompasses any type of linkers whether short (e.g., 2 to 6 amino acids) or long (e.g., 7 to 30 amino acids). Typically, linkers are composed of amino acid residues such as glycine (G), serine (S), threonine (T), asparagine (N), alanine (A) and/or proline (P). When several linkers are to be used within one molecule, one may vary the amino acid sequence or take advantage of the codon degeneracy (e.g., 4 codons possible for G and 6 for S) to decrease the percentage of identity at the nucleic level (e.g. to less than 75%, desirably less than 70%, preferably less than 60%, more preferably less than 50% and even more preferably less than 35%) especially for long linkers of 7 or more amino acids, to reduce the nucleic sequence identity and thus limit undesirable recombination events during the production of the poxvirus of the present invention. It is within the reach of the skilled person to assess the need to include a linker or not between two fused polypeptides.

As illustrated in the examples section, short peptide linkers (SPL) of 2 to 6 amino acid residues (e.g., 2, 3, 4, 5, or 6 amino acid-long) are preferred in the diabody format molecule described herein to join the V_H and V_L domains of the different antigen-binding domains. Representative examples of SPL for use herein include without any limitation GGGGS (SEQ ID NO: 1; also called L3 in the example section), GSGSG (SEQ ID NO: 2), SGSGS (SEQ ID NO: 3), GSTSG (SEQ ID NO: 4) and SGTGS (SEQ ID NO: 5) as well as GS, GSG, GAS, GTS, etc.. GGGGS (SEQ ID NO: 1) SPL is preferred in the context of this invention.

Long and flexible peptide linkers (LPL) may be useful for allowing relative movement of the linked polypeptides. Flexible linkers are known to the skilled person, and several are described, for example in Chen et al. (2013, Adv Drug Deliv Rev 65(10): 1357-69). As illustrated in the examples section, LPL are preferably used to connect the various antigen binding domains comprised in the multi-specific molecule as well as to connect the VH and VL regions in the tandem scFv format described herein. Representative examples of LPL for use herein comprises two or more SPL motifs (2, 3, etc.) with a preference for GGGGSGGGGSGGGGS (as shown in SEQ ID NO: 6; also called L1 in the example section) or GGSGTSGTSGTSGGS SEQ ID NO: 7; also called L2 in the example section).

First antigen-binding domain specific for a cell antigen present at the surface of a T cell

In some embodiments, the first antigen-binding domain is an antibody (e.g. a binding fragment thereof) specific for an antigen present (or expressed) at the surface of a T cell (i.e., a T cell antigen). The term “T cell” or “T lymphocyte” as used herein includes all types of immune cells expressing CD3 at their surface, including T-helper cells (CD4+ T cells), cytotoxic T-cells (CD8+ T cells), T-regulatory cells (Treg) and gamma-delta T cells.

In some embodiments, the first binding domain is an antibody that specifically binds to CD3 (or an epitope thereof) (also referred herein as an anti-CD3 antibody).

Cluster of differentiation 3 (CD3) is a multimeric protein complex composed of four distinct polypeptide chains; epsilon (e), gamma (y), delta (d) and zeta (z), that assemble and function as three pairs of dimers (eg, ed, zz). The CD3 complex serves as a T cell co-receptor that associates noncovalently with the T cell receptor (TCR) (Smith-Garvin et al . , 2009, Ann Rev Immunol 27: 591-619). CD3s, CD3y and CD35 are highly homologous. For general purposes, CD3s is a non-glycosylated polypeptide chain of 20 kDa (UniProt accession no. P07766 and NCBI RefSeq NP_000724.1) and comprises an epitope that is conserved among many species. Both CD3y and CD35 are glycosylated and have a molecular weight of 25-28 kDa and 20 kDa, respectively whereas Oϋ3z (also known as CD247) is a non-glycosylated polypeptide with a molecular weight of 17 kDa that shares no sequence similarity with the other CD3 polypeptide chains.

The term “CD3” as used herein encompasses any native CD3 from any vertebrate source (e.g., mammals such as primates (e.g. humans, simian, etc.), and rodents (e.g. mice and rats)), recombinant CD3, naturally occurring variants of CD3 (e.g., splice variants, allelic variants, CD3 that lacks a transmembrane domain) and CD3 receptor. The term also encompasses "full- length," soluble CD3 or CD3 associated with a cell membrane, unprocessed CD3 as well as any form of CD3 that results from processing in the cell (e.g. monomeric and dimeric CD3). For preclinical testing, it is preferred that the CD3-binding domain specifically binds to the CD3 in the species utilized for the preclinical testing (e.g., mouse CD3 for testing in mice).

In some embodiments, the first antigen-binding domain of the multi-specific molecule described herein is capable of binding specifically one or more CD3 antigen or epitope that is/are present and accessible on the surface of a T cell in vitro or in vivo. In preferred embodiments, the first antigen-binding domain is an antibody or a fragment thereof that specifically recognizes and associates with a CD3 antigen (e.g., epsilon, delta, gamma or zeta) or a dimeric complex of two CD3 polypeptides (e.g., epsilon/delta, epsilon/gamma, and zeta/zeta CD3 dimers) or a CD3-TCR complex polypeptide that is exposed at the extracellular side of the T cell membrane. Alternatively, the first antigen-binding domain is an antibody or a fragment thereof that specifically recognizes and associates with a CD3 antigen expressed on the surface of a cell that normally does not express a CD3 antigen on its surface but has been artificially engineered to express CD3 on its surface (e.g., cells used for assessing the functionality of the CD3 binding domain). Binding of the first antigen-binding domain with the CD3 target may be measurable or detectable with known assays well-known in the art, such as Flow cytometry (cell-based), ELISA and surface plasmon resonance binding assays. Examples of suitable anti-CD3 antibodies are well known in the art and include without limitation muromonab-CD3 (trade name Orthoclone OKT3, marketed by Janssen-Cilag) or fragment thereof. OKT3 recognizes dimeric complex with CD3s. It was the first monoclonal antibody to be approved for clinical use in humans as an immunosuppressive drug given to reduce transplant rejections.

In some embodiments, the first antigen-binding domain of the multi-specific molecule is an anti-CD3 antibody comprising at least one, advantageously, at least two, desirably, at least three, preferably at least four, more preferably at least five of the CDR sequences set forth in Table 1 and, even more preferably, the six CDRs of Table 1..

Table 1: preferred CDRs comprised in the CD3 binding domain

In some embodiments, the anti-CD3 antibody for use as a first antigen-binding domain comprises at least a V_H region comprising a HC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 14; a HC-CDR2 having an amino acid sequence as shown in SEQ ID NO:15 and a HC-CDR3 having an amino acid sequence in SEQ ID NO:16. In some embodiments, the anti-CD3 antibody for use as a first antigen-binding domain comprises a V_L domain comprising a LC-CDR1 having an amino acid sequence as shown in SEQ ID NO:17; a LC-CDR2 having an amino acid sequence as shown in SEQ ID NO:18 and a LC- CDR3 having an amino acid sequence in SEQ ID NO: 19. In preferred embodiments, the anti- CD3 antibody for use as a first antigen-binding domain comprises a V_H domain comprising a HC-CDR1 having an amino acid sequence as shown in SEQ ID NO:14; a HC-CDR2 having an amino acid sequence as shown in SEQ ID NO:15 and a HC-CDR3 having an amino acid sequence in SEQ ID NO: 16 and a V_L domain comprising a LC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 17; a LC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 18 and a LC-CDR3 having an amino acid sequence in SEQ ID NO: 19. Preferably, said antibody that specifically binds CD3 (for use as the first antigen-binding domain) comprises a V_H comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO:20 and/or a V_L comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO: 21. More preferred is an anti-CD3 antibody comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO:22.

Second antigen-binding domain specific for a cell antigen present at the surface of a tumor cell or tumor infiltrating cell

In some embodiments, the second antigen-binding domain is an antibody (e.g., a binding fragment thereof) specific for a cell antigen present or expressed at the surface of a tumor cell or a tumor infiltrating cell (i.e. , called hereinafter a tumor antigen or tumor cell antigen). As used herein, a “tumor antigen” is an antigen which is expressed or over- expressed by a tumor or tumor stroma. A tumor antigen may be any polypeptide, glycoprotein, lipoprotein, glycan, glycolipid, lipid, or fragment thereof which expression may be associated with a cancer. In the context of the invention, the tumor antigen is desirably expressed by the tumor cell itself or by a cell comprised in tumor stroma like fibroblast, MDSC (myeloid-derived suppressor cells), M2 macrophages, Treg, endothelial cells and any cells associated with tumor progression. A tumor antigen is desirably displayed on the external surface (i.e., extracellularly) of the target cells and preferably anchored to the tumor cell membrane.

In some embodiments, the tumor antigen may be a tumor-associated antigen (TAA) whose expression is associated with the development, progression and/or severity of symptoms of a cancer. Such TAAs may be the result of a mutation (e.g., mutated oncogene or mutated tumor suppressor gene), of an abnormal expression by a cancer cell (e.g., oncofetal antigen, overexpressed cellular protein, antigens expressed with abnormal localization or antigens expressed with an abnormal structure for instance in terms of glycosylation or folding) or may be the product of an oncogenic virus. Such tumor-associated antigens are well known in the art (see e.g., Liu et al., 2017, Eur. J. Cancer Care 26(5): DOI10.1111/ecc.12446). Representative examples of tumor antigens include oncofetal antigens such as carcinoembryonic antigen CEA, feto-acinar pancreatic protein (FAPP), alkaline phosphatase placental-like 2 (ALPPL-2), TAG-72; oncoviral antigens such as human papillomavirus (HPV) E6 and E7; overexpressed proteins such as calcium-activated chloride channel 2, cyclin-B1 , 9D7, Ep-CAM, EphA3, Human epidermal growth factor receptor 2 (HER2 also known e.g. as HER2/neu, ERBB2, CD340 and NEU), telomerase, mesothelin, SAP-1 , surviving; cancer-testis antigens: GAGE, MAGE, SAGE, NY-ESO-1; mutated antigens b-catenin, BRCA1/2, CDK4, CML66, Fibronectin, MART-2, p53, Ras, TGF- Rll; and post-translationally altered antigens such as the mucin-1 (MUC1) protein (previously called polymorphic epithelial mucin (PEM) or cancer antigen 15-3). The MUC1 protein is a highly glycosylated mucin (MW> 200 kDa) normally found anchored by a transmembrane region to the apical surface of the mucin-secreting epithelial cells in many types of tissues, including the breast, prostate, lungs, pancreas, stomach, ovaries, fallopian tubes, intestine and kidney (Peat et al., 1992, Cancer Res 52: 1954-60). The extracellular domain includes a 20 amino acid variable number tandem repeat (VNTR) domain, with the number of repeats varying from 20 to 120 in different individuals. These repeats are rich in serine, threonine and proline residues which permits heavy O- glycosylation. In preferred embodiments, the term "MUC1 " refers to tumor-associated MUC1 which is present on cancer cells. Cancer in secretory epithelial cells is often accompanied by excess expression of MUC1 by the tumor cells, a different localization (tumor-associated MUC1 is present apolarly over the whole cell surface in cancer cells contrary to a strictly apical expression for the non-cancerous MUC1) and an aberrant glycosylation (e.g. hypo glycosylated, shortened or immature sugar side chains), revealing new peptide and carbohydrate epitopes (Burchell et al., 1987, Cancer res 47: 5476-82; Acres and Limacher, 2005, Expert Rev Vaccines 4(4): 1-10).

In some embodiments, the second antigen-binding domain is an antibody that specifically binds the MUC1 antigen present (or expressed) at the surface of a tumor cell (also designated herein MUCI-binding domain). In this disclosure, MUC1 refers to MUC1 from any vertebrate source (e.g., mammals such as primates (e.g. humans; human MUC1 is identified by UniProt P15941), non-human primates (e.g. cynomolgus monkeys) and rodents (e.g. mice and rats for preclinical testing) as well as recombinant MUC1 protein, and naturally occurring MUC1 variants. The second antigen-binding domain specific for MUC1 for use herein may bind soluble MUC1, bound MUC1 and/or MUC1 associated with a cell membrane. In some embodiments, the MUCI-binding domain of the multi-specific molecule described herein is capable of binding specifically to a MUC1 antigen or portion that is/are present and accessible on the surface of a MUC1-positive tumor cell in vitro or in vivo. In preferred embodiments, the MUCI-binding domain is an antibody or a fragment thereof that specifically recognizes and associates with a MUC1 antigen that is exposed at the extracellular side of the MUC1-positive tumor cell membrane. Alternatively, the MUCI-binding domain is an antibody or a fragment thereof that specifically recognizes and associates with a MUC1 antigen expressed on the surface of a cell that normally does not express a MUC1 antigen on its surface but has been artificially engineered to express MUC1 on its surface (e.g., cells used for assessing the functionality of the MUCI-binding domain comprised in the multi-specific molecule). Binding of the second antigen-binding domain with the MUC1 target may be measurable or detectable with known assays well-known in the art, such as Flow cytometry (cell-based), ELISA and surface plasmon resonance binding assays.

Anti-MUC1 antibodies are known in the art and the present disclosure includes both full-length (intact) antibody molecules, as well as any binding fragments thereof such as those described above in connection with the term “antibody” that are capable of specific binding to MUC1 and preferably exhibiting sufficient affinity to a tumor-associated MUC1. Examples of suitable anti-MUC1 antibodies are described elsewhere, including the SM3 antibody disclosed in Paul et al. (2000, Hum Gene Ther, 10(11): 1417) and the HMFG2 antibody disclosed in Wilkie et al. (2008, J Immunol 180(7): 4901-9), this latter being preferred.

Preference is given to an anti-MUC1 binding domain comprising any of the CDR sequences set forth in Table 2.

Table 2: preferred CDRs comprised in the MUC1 binding domain

In some embodiments, the antibody that specifically binds the MUC1 antigen according to the present disclosure comprises a V_H comprising a HC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 23; a HC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 24 and a HC-CDR3 having an amino acid sequence in SEQ ID NO: 25 or SEQ ID NO: 26. In some embodiments, the antibody specific for MUC1 according to the present disclosure comprises a V_L comprising a LC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 27; a LC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 28 and a LC-CDR3 having an amino acid sequence in SEQ ID NO: 29. In preferred embodiments, the antibody specific for MUC1 according to the present disclosure comprises a V_H comprising a HC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 23; a HC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 24 and a HC-CDR3 having an amino acid sequence in SEQ ID NO 25 or SEQ ID NO: 26 and a V_L comprising a LC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 27; a LC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 28 and a LC-CDR3 having an amino acid sequence in SEQ ID NO 29. In more preferred embodiments, the MUCI-binding domain comprises a V_H comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO: 30 or SEQ ID NO: 31 (SEQ ID NO: 31 being preferred in the context of the present invention) and a V_L comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO:32. More preferred is an anti- MUC1 antibody comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO: 33. Multi-specific molecule

In some embodiments, the multi-specific molecules of the present disclosure may be constructed from various antibody fragments (such as those described herein) to form specific and/or multimeric antibody formats. In preferred embodiments, the multi-specific molecule is preferably selected from the group consisting of diabodies, triabodies, tetrabodies, minibodies, nanobodies and tandem scFv, etc. Preferred embodiments are directed to a single chain polypeptide comprising a fusion of a V_H and a V_L regions of each antigen binding domain, with peptide linkers to separate them as described hereinafter. In the context of the invention, the V_H and V_L regions of the various antigen binding domains may be situated relative to one another in any suitable arrangement. In particular embodiments, the multi specific molecule may also comprise additional amino acids or peptides, e.g. to facilitate expression, folding, trafficking, processing, purification and/or detection. For example, one or more additional sequences such as linker peptides may be used to connect or separate each entity of the antigen binding domains or to connect or separate each antigen binding domains as described hereinafter.

Signal peptide

Certain embodiments contemplate the presence of a signal peptide (also known as a leader sequence or signal sequence) in the multi-specific molecule to enhance the processing through ER (endoplasmic reticulum) and/or its secretion. Briefly, signal peptides usually comprise 15 to 35 essentially hydrophobic amino acids, are usually inserted at the N-terminus of a polypeptide downstream of the codon for initiation of translation and are then removed by a specific ER-located endopeptidase to give a mature polypeptide. Appropriate signal peptides are known in the art and may be obtained for many cellular or viral proteins that are expressed at the surface of a cell or secreted such as immunoglobulins, tissue plasminogen activator, insulin, rabies glycoprotein, HIV virus envelope glycoprotein or the measles virus F protein or may be synthetic (see e.g. W02008/138649). Signal peptides are recorded in databases such as GenBank, UniProt, Swiss-Prot, TrEMBL, etc., and can be predicted using amino acid sequence analysis tools such as SignalP (Petersen et al., 2011, Nature Methods 8: 785-786) or Signal-BLAST (Frank and Sippl, 2008, Bioinformatics 24: 2172-2176). For guidance, a suitable signal peptide for use in the context of the present invention comprises an amino acid sequence having at least 80%, and preferably at least 90% sequence identity with or consists of (100% identity) the amino acid sequence shown in SEQ ID NO: 10 (MGLGLQWVFFVALLKGVHC).

Detectable moieties In some embodiments, the multi-specific molecule may also comprise one or more detectable moieties. A detectable moiety may be e.g. a fluorescent, luminescent, immuno- detectable, radio, chemical, nucleic acid or enzymatic moiety. Suitable detectable moieties include one or more tag peptide(s). A “tag is typically a short peptide sequence able to be recognized by available antisera or compounds with the goal of facilitating for example, visualisation and/or purification of the tagged protein. Tag peptides can be detected by immunodetection assays using anti-tag antibodies. A vast variety of tag peptides can be used in the context of the invention including, without limitation, PK tag, FLAG octapeptide, MYC tag, HIS tag (usually a stretch of 4 to 10 histidine residues) and e-tag (US 6,686,152). The tag peptide(s) may be independently positioned at the N-terminus of the multi-specific molecule or alternatively at its C-terminus or alternatively internally or at any of these positions when several tags are employed. In a preferred embodiment, the multi specific molecule is equipped with tag peptides introduced at the C-terminus. Such tags are preferably His and Flag tags optionally by a cleavage site (e.g by the thrombin). Examples of suitable tags for tagging the multi-specific molecule described herein comprise the tag peptides which amino acid sequence is disclosed in SEQ ID NO: 11 (HHHHHHDYKDDDDKLVPRGS), SEQ ID NO: 12 (DYKDDDDK) or SEQ ID NO: 13 (GSDYKDDDDKHHHHHH).

Hinges In some embodiments, the multi-specific molecule described herein may comprise a hinge region (which is optional) in order to provide an optimal separation between the antigen binding domains especially in the dimeric tandem scFv format. Such a hinge may act as a flexible LPL allowing the antigen-binding domains to orient in different directions. Hinge regions may be derived from immunoglobulins. Preferably, the hinge region for use herein originates from an immunoglobin D, and more preferably, comprises an amino acid sequence having at least 80%, and preferably at least 90% sequence identity with or consists of (100% identity) the amino acid sequence shown in SEQ ID NO: 8.

Dimerization region In some embodiments, the multi-specific molecule described herein may also comprise a dimerization domain between the antigen-binding domains especially in the dimeric tandem scFv format. Such a dimerization region (which is optional) may allow the association of several multi-specific molecules (e.g. association of two multi-specific molecules to form a dimeric complex). Preferably, the dimerization region for use herein originates from the CH3 region of an lgG1, and more preferably, comprises an amino acid sequence having at least 80%, and preferably at least 90% sequence identity with or consists of (100% identity) the amino acid sequence shown in SEQ ID NO: 9.

Presence of a Fc part

In specific embodiments, the multi-specific molecule described herein may also comprise an immunoglobulin constant region (Fc) or a part thereof. In particular, the Fc part of an antibody comprises a Fc receptor binding site or a sequence having the ability to bind an Fc receptor. The presence of a Fc part may help to optimize the circulating half-life in animals (see, e.g., Wahl et al., 1983, J. Nucl. Med. 24:316). If any, the Fc part is desirably that of a human immunoglobulin consensus sequence and, preferably, from heavy constant CH2 or CH3 regions of an immunoglobin, with a preference for the CH3 region of an antibody and more particularly from an lgGi. Preferably, the multi-specific molecule does not comprise any Fc domain.

In preferred embodiments, the multi-specific molecule disclosed herein is bi-specific and comprises a first binding domain binding specifically to a CD3 antigen (CD3 binding domain) present (or expressed) at the surface of a T cell as described herein and a second binding domain binding specifically to a MUC1 antigen (MUC1 binding domain) present (or expressed) at the surface of a cancer cell as described herein. Desirably, the multi-specific molecule comprises a V_H and V_L regions of an antibody specific for CD3 and V_H and V_L regions specific for MUC1.

In some embodiments, the multi-specific molecule is a tandem scFv. A tandem scFv is typically composed of a single chain polypeptide comprising V_H and V_L regions of one antigen binding domain fused to V_H and V_L regions of another antigen binding domain. In antigen-binding domains, the V_H and V_L domains may be situated relative to one another in any suitable arrangement. The V_H and V_L regions of the two antigen binding domains are preferably connected by a long and flexible peptide linker (LPL) as well as the two scFv. Various arrangements may be envisaged as illustrated below for a bi-specific molecule:

VH (1^st binding domain)-LPI_-VL (1^st binding domain)-l_PI_-VH (2^nd binding domain)-LPL- V_L (2^nd binding domain) ;

VL (1^st binding domain)-l_PI_-VH (1^st binding domain)-l_PI_-VL (2^nd binding domain)-LPL- V_H (2^nd binding domain) ;

V_H (2^nd binding domain)-l_PI_-V_L (2^nd binding domain)-l_PI_-V_H (1^st binding domain)-LPL- V_L (1^st binding domain) ;

V_L (2^nd binding domain)-l_PI_-V_H (2^nd binding domain)-l_PI_-V_L (1^st binding domain)-LPL- V_H (1^st binding domain).

In preferred embodiments, the multi specific molecule is a tandem scFv antibody comprising the V_H and V_L regions of an antibody that binds specifically to CD3 such as the ones described herein (SEQ ID NO: 20 and 21), the V_H and V_L regions of an antibody that binds specifically to MUC1 such as the ones described herein (SEQ ID NO: 31 and 32) and the long peptide linker LPL is the (GGGGS)3 linker described in SEQ ID NO: 6. Such a protein can be linked at its N-terminus to an initiator Met and, optionally to a signal peptide (such as the one described in SEQ ID NO: 10). It may also be linked at its C-terminus to tag peptides (such as the one described in SEQ ID NO: 13).

In even more preferred embodiments, the multi-specific molecule comprises an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ ID NO: 34, SEQ ID NO: 35 or SEQ ID NO: 36. As illustrated in Table 4, SEQ ID NO: 34 provides an example of a suitable tandem scFv described herein (HMFG2/OKT3 tandem scFv expressed by the exemplified pTG19274) whereas in SEQ ID NO: 35 the tandem scFv antibody comprises a signal peptide at its N- terminus (with an initiator Met; e.g. as defined in SEQ ID NO: 10). The polypeptide described in SEQ ID NO: 36 also comprises a signal peptide at its N-terminus (with an initiator Met; e.g. as defined in SEQ ID NO: 10) and, further tag peptides at its C-terminus (e.g. as defined in SEQ ID NO: 13). It is within the reach of the skilled person to modify the signal and/or tag peptides (replacement, deletion, etc.).

In some embodiments the multi-specific molecule is a dimeric tandem scFv. A dimeric tandem scFv is typically composed of a single chain polypeptide comprising V_H and V_L regions of one antigen binding domain fused to V_H and V_L regions of another antigen binding domain and further comprising a dimerization domain preferably located in between the antigen binding domains. Various arrangements may be envisaged as illustrated below for a bi-specific molecule:

V_H (1^st binding domain)-LPI_-V_L (1^st binding domain)- dimerization domain- SPL-V_H (2^nd binding domain)-LPL- V_L (2^nd binding domain);

V_L (1^st binding domain)-l_PI_-V_H (1^st binding domain)- dimerization domain- SPL-V_L (2^nd binding domain)-LPL- V_H (2^nd binding domain);

V_H (2^nd binding domain)-l_PI_-V_L (2^nd binding domain)- dimerization domain - SPL-V_H (1^st binding domain)-LPL- V_L (1^st binding domain);

• V_L (2^nd binding domain)-l_PI_-V_H (2^nd binding domain)- dimerization domain - SPL-V_L (1^st binding domain)-LPL- V_H (1^st binding domain).

Dimeric tandem scFv may also comprise a hinge as described hereinafter, and the following formats might be considered: · V_H (1^st binding domain)-l_PI_-V_L (1^st binding domain)-hinge- dimerization domain-SPL-V_H (2^nd binding domain)-LPL- V_L (2^nd binding domain);

V_L (1^st binding domain)-l_PI_-V_H (1^st binding domain)- hinge- dimerization domain-SPL-V_L (2^nd binding domain)-LPL- V_H (2^nd binding domain); • V_H (2^nd binding domain)-LPI_-V_L (2^nd binding domain)- hinge- dimerization domain -SPL-V_H (1^st binding domain)-LPL- V_L (1^st binding domain);

• V_L (2^nd binding domain)-l_PI_-V_H (2^nd binding domain)- hinge- dimerization domain -SPL-V_L (1^st binding domain)-LPL- V_H (1^st binding domain). In preferred embodiments, the multi specific molecule is a dimeric tandem scFv antibody comprising the V_H and V_L regions of an antibody that binds specifically to CD3 such as the ones described herein (SEQ ID NO: 20 and 21), the V_H and V_L regions of an antibody that binds specifically to MUC1 such as the ones described herein (SEQ ID NO: 31 and 32), the short linker (SPL) is the GGGGS linker described in SEQ I D NO: 1 , the long peptide linker LPL is the (GGGGS)3 linker described in SEQ ID NO: 6 and, further, the CH3lgG1 dimerization domain shown in SEQ ID NO: 9) and, optionally, the hinge IgD region shown in SEQ ID NO: 8. Such a protein can be linked at its N-terminus to an initiator Met and, optionally to a signal peptide (such as the one described in SEQ ID NO: 10). It may also be linked at its C-terminus to tag peptides (such as the one described in SEQ ID NO: 13). In even more preferred embodiments, the multi-specific molecule comprises an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ ID NO: 37, SEQ ID NO: 38 or SEQ ID NO: 39. As illustrated in Table 4, SEQ ID NO: 37 provides an example of a suitable dimeric tandem scFv described herein (HMFG2-/OKT3 dimeric tandem scFv expressed by the exemplified pTG19280) whereas in SEQ ID NO: 38 the dimeric tandem scFv antibody comprises a signal peptide at its N-terminus (with an initiator Met; e.g. as defined in SEQ ID NO: 10). The polypeptide described in SEQ ID NO: 39 also comprises a signal peptide at its N-terminus (with an initiator Met; e.g. as defined in SEQ ID NO: 10) and, further tag peptides at its C- terminus (e.g. as defined in SEQ ID NO: 13). It is within the reach of the skilled person to modify the signal and/or tag peptides (replacement, deletion, etc.).

In some embodiments, the multi-specific molecule is a single chain diabody. A single chain diabody is typically composed of a single chain polypeptide comprising a fusion of V_H and V_L regions of each antigen binding domain interspersed one another (i.e. the VH region of one antigen binding domain fused to the VL region of another antigen binding domain and vice versa) that form two linked scFv upon folding of the polypeptide on itself. The V_H and V_L regions are preferably connected by a short peptide linker (SPL; the linker is too short to form intrachain pairing with the adjacent V_H and V_L domains) and each V_H and V_L entity is connected to the other one with a long and flexible peptide linker (LPL). The present disclosure encompasses various formats such as the ones illustrated below for a bi specific molecule:

• VH (2^nd binding domain)-SPL-VL (1^st binding domain)-LPL-VH (1^st binding domain)-SPL- V_L (2^nd binding domain); • V_L (2^nd binding domain)-SPI_-V_H (1^st binding domain)-LPI_-V_L (1^st binding domain)-SPL- V_H (2^nd binding domain);

• V_H (1^st binding domain)-SPL-Vi_ (2^nd binding domain)-l_PI_-V_H (2^nd binding domain)-SPL- V_L (1^st binding domain); · V_L (1^st binding domain)-SPI_-V_H (2^nd binding domain)-l_PI_-V_L (2^nd binding domain)-SPL- V_H (1^st binding domain).

In preferred embodiments, the multi specific molecule is a single chain antibody comprising the V_H and V_L regions of an antibody that binds specifically to CD3 such as the ones described herein (SEQ ID NO: 20 and 21), the V_H and V_L regions of an antibody that binds specifically to MUC1 such as the ones described herein (SEQ ID NO: 31 and 32), the short linker (SPL) is the GGGGS linker described in SEQ I D NO: 1 and the long signal peptide LPL is the (GGGGS)3 linker described in SEQ ID NO: 6. Such a protein can be linked at its N-terminus to an initiator Met and, optionally to a signal peptide (such as the one described in SEQ ID NO: 10). It may also be linked at its C-terminus to a tag peptide (such as the one described in SEQ ID NO: 13).

In even more preferred embodiments, the multi-specific molecule comprises an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ I D NO: 40, SEQ I D NO: 41 or SEQ I D NO: 42. As illustrated in Table 4, SEQ ID NO: 40 provides an example of a suitable single chain diabody described herein (e.g., hscDb3-muc1 expressed by the exemplified pTG19277) whereas in SEQ ID NO: 41 the single chain antibody comprises a signal peptide at its N- terminus (with an initiator Met; e.g. as defined in SEQ ID NO: 10). The polypeptide described in SEQ ID NO: 42 also comprises a signal peptide at its N-terminus (with an initiator Met; e.g. as defined in SEQ ID NO: 10) and, further a tag peptide at its C-terminus (e.g. as defined in SEQ ID NO: 13). It is within the reach of the skilled person to modify the signal and/or tag peptides (replacement, deletion, etc.).

Many modifications useful in construction of disclosed multi-specific molecules and methods for making them are known in the art. The multi-specific molecules described herein may be readily engineered by standard molecular biology techniques or automatized synthesis techniques using the information described herein and the general knowledge of those of ordinary skill in the art. For illustrative purposes, they may be engineered, e.g., by independently generating the nucleic acid molecules encoding the antigen binding domains having the required specificities, which may be linked in any convenient and appropriate combination to generate such multi-specific (e.g., bi-specific) molecules.

Nucleic acid molecules and expression In some embodiments, the present invention provides a nucleic acid molecule encoding the multi-specific (e.g., bi-specific) molecules described herein as well as a poxvirus comprising such a nucleic acid molecule.

Such a nucleic acid molecule can be optimized for providing high level expression in a particular host cell or subject. It has been indeed observed that, the codon usage patterns of organisms are highly non-random and the use of codons may be markedly different between different hosts. Typically, codon optimization is performed by replacing one or more "native" codon corresponding to a codon infrequently used in the host organism by one or more codon encoding the same amino acid which is more frequently used. It is not necessary to replace all native codons corresponding to infrequently used codons since increased expression can be achieved even with partial replacement.

Further to optimization of the codon usage, expression in the host cell or subject can further be improved through additional modifications of the nucleic acid molecule. For example, various modifications may be envisaged so as to prevent clustering of rare, non- optimal codons being present in concentrated areas and/or to suppress or modify "negative" sequence elements which are expected to negatively influence expression levels. Such negative sequence elements include without limitation the regions having very high (>80%) or very low (<30%) GC content; AT-rich or GC-rich sequence stretches; unstable direct or inverted repeat sequences; R A secondary structures; and/or internal cryptic regulatory elements such as internal TATA-boxes, chi-sites, ribosome entry sites, and/or splicing donor/acceptor sites.

Preferred embodiments encompass nucleic acid molecules comprising a nucleotide sequence encoding a multi-specific molecule described herein (more specifically a multi specific molecule comprising an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ ID NO: 34-42) and more preferably:

A nucleic acid molecule comprising a nucleotide sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO: 43 (nucleotide sequence encoding HMFG2/OKT3 tandem scFv expressed by pTG19274 with SP and tags) or a part thereof of at least 500 nucleotides, notably the part starting at position 1 to position 1545 (with SP-encoding sequence in italic and without tag) or the part starting at position 58 to position 1545 (without SS and tags). It is within the reach of the skilled person to modify the sequence encoding the signal and/or tag peptides (replacement, deletion, etc.) and to add for proper translation an ATG initiator and a STOP codon if necessary. A nucleic acid molecule comprising a nucleotide sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO:44 (nucleotide sequence encoding hscDb3-muc1 diabody expressed by the exemplified pTG19277 with SP and tags) or a part thereof of at least 500 nucleotides, notably the part starting at position 1 to position 1485 (with SP-encoding sequence in italic and without tag) or the part starting at position 58 to position 1485 (without SS and tags). It is within the reach of the skilled person to modify the sequence encoding the signal and/or tag peptides (replacement, deletion, etc.) and to add for proper translation an ATG initiator and a STOP codon if necessary. A nucleic acid molecule comprising a nucleotide sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO: 45 (nucleotide sequence encoding HMFG2-/OKT3 dimeric tandem scFv expressed by the exemplified pTG 19280 with SP and tags) or a part thereof of at least 500 nucleotides, notably the part starting at position 1 to position 2010 (with SP-encoding sequence in italic and without tag) or the part starting at position 58 to position 2010 (without SS and tags). It is within the reach of the skilled person to modify the sequence encoding the signal and/or tag peptides (replacement, deletion, etc.) and to add for proper translation an ATG initiator and a STOP codon if necessary. In accordance with the present invention, the nucleic acid molecule encoding the multi-specific molecule is operably linked to suitable regulatory elements for its expression in a host cell or subject. As used herein, “operably linked” means that the elements being linked are arranged so that they function in concert for their intended purposes. For example, a promoter is operably linked to a nucleic acid molecule if the promoter effects transcription from the transcription initiation to the terminator of said nucleic acid molecule in a permissive host cell.

It will be appreciated by those skilled in the art that the choice of the regulatory sequences can depend on such factors as the nucleic acid itself, the virus into which it is inserted, the host cell or subject, the level of expression desired, etc. The promoter is of special importance. In the context of the invention, it can be constitutive directing expression of the nucleic acid molecule in many types of host cells or specific to certain host cells or regulated in response to specific events or exogenous factors (e.g. by temperature, nutrient additive, hormone, etc.,) or according to the phase of a viral cycle (e.g. late or early). One may also use promoters that are repressed during the production step in response to specific events or exogenous factors, in order to optimize virus production and circumvent potential toxicity of the expressed polypeptide(s).

VV promoters Poxvirus genes and their associated promoters are classified as either early, or late or combinations thereof, such as early/late. Transcription of the early genes proceeds shortly after fusion of the virion to a cell membrane and entry of the virus particle into the cytoplasm. The early genes encode proteins that stimulate proliferation of neighbouring cells, protect against the host immune system, replicate the viral genome, and transcribe the intermediate class of genes. Expression of the late genes begins after DNA replication. Late or early/late promoters are preferred for driving expression of the multi-specific molecule by the poxvirus described herein

Representative examples of suitable promoters for use herein include without limitation endogenous vaccinia promoters such as those of the following loci 7.5K, H5R, TK, B2R, B8R, C11R, F11L (Orubu et al., 2012, PloS One 7(6)e40167), A14L, A35R and K1L as well as synthetic promoters such as those described in Chakrabarti et al. (1997, Biotechniques 23: 1094-7); Hammond et al. (1997, J. Virol Methods 66: 135-8; and Kumar and Boyle (1990, Virology 179: 151-8); Erbs et al. (2008, Cancer Gene Ther. 15(1): 18-28; p11 K7.5).

Those skilled in the art will appreciate that the regulatory elements controlling the expression of the nucleic acid molecule(s) inserted into the poxviral genome may further comprise additional elements for proper initiation, regulation and/or termination of transcription (e.g. polyA transcription termination sequences), mRNA transport (e.g. nuclear localization signal sequences), processing (e.g. splicing signals), and stability (e.g. introns and non-coding 5' and 3' sequences), translation (e.g. an initiator Met, tripartite leader sequences, IRES ribosome binding sites, signal peptides, etc.).

In some embodiments, expression of the nucleic acid molecule described herein is driven by the pH5R promoter, and more preferably by a promoter having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO: 46.

Generation of the poxvirus encoding the multi-specific molecule (recombinant poxvirus) The nucleic acid molecule encoding the multi-specific molecule described herein can be inserted at any location of the poxviral genome and insertion can be performed by routine molecular biology well known in the art. Various sites of insertion may be considered, e.g. in a non-essential viral gene, in an intergenic region, or in a non-coding portion of the poxvirus genome. J2R and/or I4L locus is particularly appropriate in the context of the invention. As described above, upon insertion of the nucleic acid molecule into the poxvirus genome, the viral locus at the insertion site may be deleted at least partially, e.g. resulting in suppressed expression of the viral gene product encoded by the entirely or partially deleted locus and a defective virus for said virus function.

In some embodiments, the nucleic acid molecule and its regulatory elements is inserted within the J2R locus of the poxvirus genome.

The present invention also provides a method for generating the poxvirus described herein, and particularly a oncolytic vaccinia virus, by homologous recombination between a transfer plasmid comprising the nucleic acid molecule encoding the multi-specific molecule (with its regulatory elements) flanked in 5’ and 3’ with viral sequences respectively present upstream and downstream the insertion site and a virus genome. The present invention also encompasses such a plasmid. In one embodiment, said method comprise a step of generating said transfer plasmid (e.g. by conventional molecular biology methods) and a step of introducing said transfer plasmid into a suitable host cell, notably together with a poxvirus genome comprising the flanking sequence present in the transfer plasmid. Preferably, the transfer plasmid is introduced into the host cell by transfection and the virus by infection.

The size of each flanking viral sequence may vary. It is usually at least 100bp and at most 1500 bp, with a preference for approximately 150 to 800bp on each side of the recombinant nucleic acid, advantageously from 180 to 600bp, preferably from 200 to 550bp and more preferably from 250 to 500bp. In certain embodiments, identification of the recombinant poxvirus may be facilitated by the use of a selection and/or a detectable gene. In preferred embodiments, the transfer plasmid further comprises a selection marker with a specific preference for the GPT gene (encoding a guanine phosphoribosyl transferase) permitting growth in a selective medium (e.g. in the presence of mycophenolic acid, xanthine and hypoxanthine) or a detectable gene encoding a detectable gene product such as GFP, e-GFP or mCherry. In addition, the use of an endonuclease capable of providing a double-stranded break in said selection or detectable gene may also be considered. Said endonuclease may be in the form of a protein or expressed by an expression vector.

Homologous recombination permitting to generate the recombinant poxvirus is preferably carried out in appropriate host cells (e.g. HeLa or CEF cells).

In some embodiments, the poxvirus comprising the nucleic acid molecule encoding the multi-specific molecule described herein further comprises a nucleic acid molecule encoding one or more polypeptide of interest For example, the poxvirus may comprise nucleic acid molecule(s) encoding a polypeptide for reducing growth or proliferation of infected cells, or a polypeptide for rendering infected cells sensitive to treatment with a prodrug agent, or a polypeptide for disrupting tumor structure (e.g. enzymes for digesting tumour matrix) to facilitate immune cell infiltration and/or an immunomodulatory polypeptide(s).

In some embodiments, the poxvirus encoding the multi-specific molecule described herein additionally comprises a nucleic acid molecule encoding an immunomodulatory polypeptide. Immunomodulatory polypeptide(s) for use herein are preferably selected to facilitate the immune response to a cancer in a subject, in particular the cell-mediated immune response. In particular embodiments, the immunomodulatory polypeptide is selected from the group consisting of i) an agonist of an effector immune response promoting activation, recruitment, proliferation, activity and/or survival of effector immune cells (e.g., a cytokine (e.g. IL-2, IL-7, IL-12, IL-15, IL18, etc), a chemokine (MIP-1a or RANTES), an agonist antibody or a ligand for a costimulatory receptor (e.g. 4-1 BB, 0X40, CD28, CD27, ICOS, FLT3, CD30 orGITR)); ii) an antagonist of an immune checkpoint inhibitor (e.g., an antagonist antibody or ligand of an immune checkpoint inhibitor such as PD1, PD-L1, PD-L2, CTLA-4, LAG-3, TIM-3, TIGIT, VISTA, etc.,) and iii) an antagonist of an immunosuppressive response (e.g. an antagonist of a cytokine/chemokine promoting activation, recruitment, proliferation, activity and/or survival of regulatory T cells (Tregs) and/or myeloid- derived suppressor cells (MDSCs)).

In a preferred embodiment, the immunomodulatory polypeptide is Flt3 ligand (Flt3L) with a specific preference for a Flt3L polypeptide comprising an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO: 47 or the part thereof starting at amino acid residue 21 (Ser). It is within the reach of the skilled person to modify the sequence encoding the signal peptide (replacement, deletion, etc.) and to add for proper translation an ATG initiator. In some embodiments, FLt3L-encoding nucleic acid molecule comprises the nucleotide sequence shown in SEQ ID NO: 48) and its expression is driven by the pH5.R promoter (SEQ ID NO: 46). It In the context of the invention, the FLt3L-encoding nucleic acid molecule is preferably inserted within the I4L locus of the poxvirus genome.

Production of the recombinant poxvirus Typically, the poxvirus of the invention is produced into a suitable host cell line using conventional techniques including culturing the transfected or infected host cell under suitable conditions so as to allow the production and recovery of infectious poxviral particles.

Therefore, in another aspect, the present invention relates to a method for producing the poxvirus described herein. In some embodiments, said method comprises the steps of a) preparing a producer cell, b) transfecting or infecting the prepared producer cell with the poxvirus, c) culturing the transfected or infected producer cell under suitable conditions so as to allow the production of the virus (e.g. infectious poxviral particles), d) recovering the produced virus from the culture of said producer cell and optionally e) purifying said recovered virus.

In one embodiment, the producer cell is a mammalian (e.g. human or non-human) cell selected from the group consisting of HeLa cells (e.g. ATCC-CRM-CCL-2™ or ATCC-CCL- 2.2™), HER96, PER-C6 (Fallaux et al., 1998, Human Gene Ther. 9: 1909-17) and hamster cell lines such as BHK-21 (ATCC CCL-10) or an avian cell such as one of those described in W02005/042728, W02006/108846, W02008/129058, WO2010/130756, WO2012/001075, etc) as well as a primary chicken embryo fibroblast (CEF) prepared from chicken embryos obtained from fertilized eggs. Preference is given to HeLa cells and CEF. Producer cells are preferably cultured in an appropriate medium which can, if needed, be supplemented with serum and/or suitable growth factor(s) or not (e.g. a chemically defined medium preferably free from animal-or human-derived products). An appropriate medium may be easily selected by those skilled in the art depending on the producer cells. Such media are commercially available. Producer cells are preferably cultured at a temperature comprised between +30°C and +38°C (more preferably at approximately 37°C) for between 1 and 8 days before infection. If needed, several passages of 1 to 8 days may be made in order to increase the total number of cells.

In step b), producer cells are infected by the poxvirus under appropriate conditions using an appropriate multiplicity of infection (MOI) to permit productive infection of producer cells. For illustrative purposes, an appropriate MOI ranges from 10³ to 20, with a specific preference for a MOI comprises from 0.01 to 5 and more preferably 0.03 to 1. Infection step is carried out in a medium which may be the same as or different from the medium used for the culture of producer cells.

In step c), infected producer cells are then cultured under appropriate conditions well known to those skilled in the art until progeny poxvirus (e.g. infectious virus particles) is produced. Culture of infected producer cells is also preferably performed in a medium which may be the same as or different from the medium/media used for culture of producer cells and/or for infection step, at a temperature between +32°C and +37°C, for 1 to 5 days.

In step d), the poxvirus produced in step c) is collected from the culture supernatant and/or the producer cells. Recovery from producer cells may require a step allowing the disruption of the producer cell membrane to allow the liberation of the virus. The disruption of the producer cell membrane can be induced by various techniques well known to those skilled in the art, including but not limited to freeze/thaw, hypotonic lysis, sonication, microfluidization, high shear (also called high speed) homogenization or high-pressure homogenization. The recovered poxvirus may then be at least partially purified before being distributed in doses and used according to the present invention. A vast number of purification steps and methods is available in the art, including e.g. clarification, enzymatic treatment (e.g. endonuclease, protease, etc), chromatographic and filtration steps. Appropriate methods are described in the art (see e.g. W02007/147528; W02008/138533, W02009/100521, WO2010/130753, WO2013/022764).

In some embodiments, the present invention also provides a cell infected with the poxvirus described herein. Compositions and kits of parts

The present invention also provides a composition comprising a therapeutically effective amount of the poxvirus (e.g., oncolytic VV) described herein and a pharmaceutically acceptable vehicle. Such a composition may be administered once or several times and via the same or different routes. A “therapeutically effective amount” corresponds to the amount of poxvirus that is sufficient for producing an improvement of the clinical status. An improvement of the clinical status can be easily assessed by any relevant clinical measurement typically used by physicians or other skilled healthcare staff. Such a therapeutically effective amount may vary as a function of various parameters, in particular the mode of administration; the disease state; the age and weight of the subject; the ability of the subject to respond to the treatment; kind of concurrent treatment; the frequency of treatment; and/or the need for prevention or therapy. When prophylactic use is concerned, the composition of the invention is administered at a dose sufficient to prevent or to delay the onset and/or establishment and/or relapse of the targeted disease (e.g. cancer), especially in a subject at risk. For “therapeutic” use, the composition is administered to a subject diagnosed as having the targeted disease (e.g. cancer) with the goal of treating the disease, eventually in association with one or more conventional therapeutic modalities. In particular, a therapeutically effective amount could be that amount necessary to cause an observable improvement of the clinical status over the baseline status or over the expected status if not treated or treated with an irrelevant virus (not expressing a multi-specific molecule described herein), e.g. reduction in the tumor number; reduction in the tumor size, reduction in the number or extend of metastasis, increase in the length of remission, stabilization (i.e. not worsening) of the state of disease, delay or slowing of disease progression or severity, amelioration or palliation of the disease state, prolonged survival, better response to the standard treatment, improvement of quality of life, reduced mortality, etc. For example, techniques routinely used in laboratories (e.g. flow cytometry, histology, medical imaging) may be used to perform tumor surveillance. A therapeutically effective amount could also be the amount necessary to cause the development of an effective non-specific (innate) and/or specific (adaptative) immune response. Typically, development of an immune response, in particular T cell response, can be evaluated in vitro, in suitable animal models or using biological samples collected from the subject (ELISA, flow cytometry, histology, etc). One may also use various available antibodies so as to identify different immune cell populations involved in anti-tumor response that are present in the treated subjects, such as cytotoxic T cells, activated cytotoxic T cells, natural killer cells and activated natural killer cells.

In a preferred embodiment, the composition is formulated in individual doses, each dose containing from about 10³ to 10¹² vp (viral particles), iu (infectious unit) or pfu (plaque forming units) of the poxvirus depending on the quantitative technique used. The quantity of virus present in a sample can be determined by routine titration techniques, e.g. by counting the number of plaques following infection of permissive cells (e.g. HeLa cells) to obtain a plaque forming units (pfu) titer, by measuring the A260 absorbance (vp titers), or still by quantitative immunofluorescence, e.g. using anti-virus antibodies (iu titers). Further refinement of the calculations necessary to adapt the appropriate dosage for a subject or a group of subjects may be routinely made by a practitioner, in the light of the relevant circumstances. As a general guidance, individual doses which are suitable for the poxvirus composition comprise from approximately 10³ to approximately 10¹² pfu, advantageously from approximately 10⁴ pfu to approximately 10¹¹ pfu, preferably from approximately 10⁵ pfu to approximately 10¹° pfu; and more preferably from approximately 10⁶ pfu to approximately 10⁹ pfu and notably individual doses of approximately 10⁶, 5x10⁶, 10⁷, 5x10⁷, 10⁸, 5x10⁸ or 10⁹ pfu are particularly preferred. The term “pharmaceutically acceptable vehicle” is intended to include any and all carriers, solvents, diluents, excipients, adjuvants, dispersion media, coatings, antibacterial and antifungal agents, absorption agents and the like compatible with administration in mammals and in particular human subjects. Non-limiting examples of pharmaceutically acceptable vehicles include water, NaCI, normal saline solutions, lactated Ringer's, saccharide solution (e.g. glucose, trehalose, saccharose, dextrose, etc) alcohols, oils, gelatins, carbohydrates such as lactose, amylose or starch, fatty acid esters, hydroxymethycellulose, and the like as well as other aqueous physiologically balanced salt solutions may be used (see for example the most current edition of Remington: The Science and Practice of Pharmacy, A. Gennaro, Lippincott, Williams&Wilkins). In one embodiment, the composition is formulated appropriately to ensure the stability of the poxvirus (e.g., Vaccinia virus) under the conditions of manufacture and long-term storage (i.e. for at least 6 months, with a preference for at least two years) at freezing (e.g. between -70°C and -10°C), refrigerated (e.g. 4°C) or ambient (e.g. 20-25°C) temperature. Such formulations generally include a liquid carrier such as aqueous solutions.

Advantageously, the composition is suitably buffered for human use, preferably at physiological or slightly basic pH (e.g. from approximately pH 7 to approximately pH 9 with a specific preference for a pH comprised between 7 and 8 and more particularly close to 7.5). Suitable buffers include without limitation TRIS (tris(hydroxymethyl)methylamine), TRIS-HCI (tris(hydroxymethyl)methylamine-HCI), HEPES (4-2-hydroxyethyl-1- piperazineethanesulfonic acid), phosphate buffer (e.g. PBS), ACES (N-(2-Acetamido)- aminoethanesulfonic acid), PIPES (Piperazine-N,N’-bis(2-ethanesulfonic acid)), MOPSO (3- (N-Morpholino)-2-hydroxypropanesulfonic acid), MOPS (3-(N-morpholino)propanesulfonic acid), TES (2-{[tris(hydroxymethyl)methyl]amino}ethanesulfonic acid), DIPSO (3-[bis(2- hydroxyethyl)amino]-2-hydroxypropane-1-sulfonic acid), MOBS (4-(N- morpholino)butanesulfonic acid), TAPSO (3-[N-Tris(hydroxymethyl)methylamino]-2- hydroxypropanesulfonic Acid), HEPPSO (4-(2-Hydroxyethyl)-piperazine-1-(2-hydroxy)- propanesulfonic acid), POPSO (2-hydroxy-3-[4-(2-hydroxy-3-sulfopropyl)piperazin-1- yl]propane-1- sulfonic acid), TEA (triethanolamine), EPPS (N-(2-Hydroxyethyl)-piperazine-N’- 3-propanesulfonic acid), and TRICINE (N-[Tris(hydroxymethyl)-methyl]-glycine). Preferably, said buffer is selected from TRIS-HCI, TRIS, Tricine, HEPES and phosphate buffer comprising a mixture of Na₂HP0₄ and KH2PO4 or a mixture of Na₂HP0₄ and NaH₂P0₄. Said buffer (in particular those mentioned above and notably TRIS-HCI) is preferably present in a concentration of 10 to 50 mM.

It might be beneficial to also include in the composition a monovalent salt so as to ensure an appropriate osmotic pressure. Said monovalent salt may notably be selected from NaCI and KCI, preferably said monovalent salt is NaCI, preferably in a concentration of 10 to 500 mM.

The composition may also be formulated so as to include a cryoprotectant for protecting the poxvirus at low storage temperature. Suitable cryoprotectants include without limitation sucrose (or saccharose), trehalose, maltose, lactose, mannitol, sorbitol and glycerol, preferably in a concentration of 0.5 to 20% (weight in g/volume in L, referred to as w/v). For example, sucrose is preferably present in a concentration of 5 to 15% (w/v).

The poxvirus composition, and especially liquid composition thereof, may further comprise a pharmaceutically acceptable chelating agent for improving stability. The pharmaceutically acceptable chelating agent may notably be selected from ethylenediaminetetraacetic acid (EDTA), 1,2-bis(o-aminophenoxy)ethane-N,N,N',N'- tetraacetic acid (BAPTA), ethylene glycol tetraacetic acid (EGTA), dimercaptosuccinic acid (DMSA), diethylene triamine pentaacetic acid (DTPA), and 2,3-Dimercapto-1- propanesulfonic acid (DMPS). The pharmaceutically acceptable chelating agent is preferably present in a concentration of at least 50 mM with a specific preference for a concentration of 50 to 1000 mM. Preferably, said pharmaceutically acceptable chelating agent is EDTA present in a concentration close to 150 pM.

Additional compounds may further be present to increase stability of the poxvirus composition. Such additional compounds include, without limitation, C2-C3 alcohol (desirably in a concentration of 0.05 to 5% (volume/volume or v/v)), sodium glutamate (desirably in a concentration lower than 10 mM), non-ionic surfactant (US7,456,009, US2007-0161085) such as Tween 80 (also known as polysorbate 80) at low concentration below 0.1%. Divalent salts such as MgCh or CaCh have been found to induce stabilization of various biological products in the liquid state (see Evans et al. 2004, J Pharm Sci. 93:2458-75 and US7,456,009). Amino acids, in particular histidine, arginine and/or methionine, have been found to induce stabilization of various viruses in the liquid state (see WO2016/087457). The presence of high molecular weight polymers such as dextran or polyvinylpyrrolidone (PVP) is particularly suited for freeze-dried compositions obtained by a process involving vacuum drying and freeze-drying and the presence of these polymers assists in the formation of the cake during freeze-drying (see e.g. WO2014/053571).

In accordance with the present invention, the composition can also be adapted to the mode of administration to ensure proper distribution or delayed release in vivo. Biodegradable and biocompatible polymers can be used, such as ethylene vinyl acetate, polyanhydrides, polyglycolic acid, collagen, polyorthoesters, polylactic acid and polyethylene glycol (see e.g. J. R. Robinson in “Sustained and Controlled Release Drug Delivery Systems”, ed., Marcel Dekker, Inc., New York, 1978; W001/23001; W02006/093924; W02009/053937). For illustrative purposes, Tris-buffered formulations (Tris-HCI pH8) comprising saccharose 5 % (w/v), sodium glutamate 10 mM, and NaCI 50 mM are adapted to the preservation of the composition described herein from -20°C to 5°C.

Administration The actual amount of composition to administer, and rate and time-course of administration(s) will depend on the nature and severity of the disease being treated. Prescription of treatment is within the responsibility of general practitioners, and typically takes account of the condition to be treated, the condition of the individual patient, the site of delivery, the method of administration and other factors known to practitioners. Any of the conventional administration routes is applicable in the context of the invention including parenteral, topical or mucosal routes. Parenteral routes are intended for administration as an injection or infusion and encompass systemic as well as local routes. Parenteral injection types that may be used to administer the poxvirus composition include intravenous (into a vein), intravascular (into a blood vessel), intra-arterial (into an artery such as hepatic artery), intradermal (into the dermis), subcutaneous (under the skin), intramuscular (into muscle), intraperitoneal (into the peritoneum) and intratumoral (into a tumor or its close vicinity) and also scarification. Administration can be in the form of a single bolus dose, multiple doses or may be, for example, by a continuous perfusion pump. Mucosal administrations include without limitation oral/alimentary, intranasal, intratracheal, intrapulmonary, intravaginal or intra-rectal route. Topical administration can also be performed using transdermal means (e.g. patch and the like). Preferably, the poxvirus composition is formulated for intravenous, subcutaneous, intramuscular or intratumoral administration.

Administrations may use conventional syringes and needles (e.g. Quadrafuse injection needles) or any compound or device available in the art capable of facilitating or improving delivery of the poxvirus in the subject (e.g. electroporation for facilitating intramuscular administration). An alternative is the use of a needleless injection device (e.g. Biojector TM device). Transdermal patches may also be envisaged.

In some embodiments, the virus of the present invention may be administered in such a way as to minimize the subject’s anti-virus responses (e.g. neutralization by anti-virus antibodies). For example, the virus may be coated in nanoparticles to maximise delivery to the subject.

The composition described herein is suitable for a single administration or multiple administrations. It is also possible to proceed via sequential cycles of administrations that are repeated after a rest period. Multiple doses may be separated by a predetermined time interval, which may be from 24h to about six months (e.g. 24h, 48h, 72h, etc.,), preferably from about a week to about a month (e.g. every week, 2 weeks, 3 weeks or every month, etc). Intervals can also be irregular. The doses can vary for each administration within the range described above. By way of example, 2 to 10 doses may be administered first weekly or every 2 or 3 weeks followed by 2 to 15 administrations at longer intervals (e.g. 1-6 months). One or more, or each, of the dose administrations may be accompanied by simultaneous or sequential administration of another therapeutic agent.

Methods of use of the poxvirus and composition of the invention

Aspects of the present invention are concerned with the virus or the composition disclosed herein for use for treating a cancer in a subject in need thereof. Accordingly, the present invention provides a method of treating a cancer comprising administering to a subject a poxvirus comprising a nucleic acid encoding a multi-specific molecule comprising: (a) a first antigen-binding domain specific for T cell surface molecule, and (b) a second antigen-binding domain specific for a cancer cell antigen. The present invention also provides a poxvirus comprising a nucleic acid encoding a multi-specific molecule as described herein for use for treating a cancer in a subject in need thereof as well as such a poxvirus for use in the manufacture of a medicament for treating a cancer in a subject in need thereof.

The present invention also provides a method for lysing or killing tumor cells expressing at their surface the targeted cancer antigen specifically recognized by the second binding domain by immune cells expressing at their surface the targeted antigen specifically recognized by the first binding domain comprising administering to a subject a poxvirus comprising a nucleic acid encoding a multi-specific molecule comprising: (a) a first antigen binding domain specific for T cell surface molecule, and (b) a second antigen-binding domain specific for a cancer cell antigen.

In preferred embodiments, said poxvirus is a RR-defective oncolytic vaccinia virus, said first antigen-binding domain specifically binds CD3 and/or said second antigen-binding domain specifically binds tumor-associated MUC1 antigen. In some embodiments, said poxvirus further comprises a nucleic acid molecule encoding an immunomodulatory polypeptide such as a Flt3L.

As used herein, the term “cancer” may be used interchangeably with any of the terms “tumor”, “malignancy”, “neoplasm” and encompasses any disease or pathological condition resulting from uncontrolled cell growth and spread. These terms are meant to include any type of tissue, organ or cell, any stage of malignancy (e.g. from a prelesion to stage IV). Typically, tumors, especially malignant tumors, show partial or complete lack of structural organization and functional coordination as compared to normal tissue and generally show a propensity to invade surrounding tissues (spreading) and/or metastasize to farther sites. The present invention encompasses solid tumors and blood born tumors as well as primary and metastatic tumors. A “tumor cell”, “cancer cell” or “neoplastic cell” can be used interchangeably to refer to a cell that divides at an abnormal (i.e. increased) rate but the term also encompasses cells present in the tumor stroma such as those cited above.

In some embodiments, the subject is a patient having a cancer, i.e. exhibiting symptoms of cancer. Preferably, the patient displays a cancer symptom and/or a cancer diagnostic marker. Such a cancer symptom and/or a cancer diagnostic marker can be measured and/or assessed and/or quantified by a person skilled in the art of medicine. In some embodiments, the cancer to be treated in accordance with the present invention is a solid tumor. Representative examples of cancers that may be treated using the composition and methods of the invention include, without limitation, bone cancer, gastrointestinal cancer, liver cancer (e.g. hepatocarcinoma), pancreatic cancer, gastric cancer, colorectal cancer, esophageal cancer, bile duct carcinoma, oro-pharyngeal cancer, laryngeal cancer, lung cancer (e.g. non-small cell lung cancer), skin cancer, squamous cell cancer, melanoma, uterine cancer, cervical cancer, endometrial carcinoma, vulvar cancer, ovarian cancer, breast cancer (e.g. metastatic breast cancer), prostate cancer (e.g. hormone refractory prostate adenocarcinoma), bladder cancer, kidney cancer (e.g. clear cell carcinoma), cancer of the head or neck, etc.

In some embodiments, the cancer to be treated is a MUCI-positive cancer and, preferably, a cancer overexpressing MUC1 at the cell surface. In the context of the present invention, “MUC1 positivity” or “MUC1 overexpression” may be used interchangeably to mean a higher level of MUC1 expression compared to equivalent non-cancerous cells/non-tumor tissue. The non cancerous tissue used for such a comparison (reference tissue) may be collected from the patient himself or from one or several healthy individuals. MUC1 positivity may be determined by any suitable means relying on gene expression or protein expression. Gene expression can be determined e.g. by detection of mRNA encoding the MUC1 antigen, for example by quantitative real-time PCR (qRT-PCR) or RNA sequencing. Protein expression can be determined e.g. by detection of the MUC1 antigen, for example by antibody-based methods, by western blot, immunohistochemistry, immunocytochemistry, flow cytometry, or ELISA, etc.

Preferred embodiments are directed to methods and use relying on a composition comprising a Vaccinia virus (e.g., of Copenhagen strain) which is defective for TK and RR activities encoding a multi-specific molecule with anti-CD3 and anti-MUC1 specificities as described herein (e.g. any of the polypeptides shown in SEQ ID NO: 34-42) optionally encoding a FLT3 ligand (such as the amino acid sequence shown in SEQ ID NO: 47) for use for treating a subject with a MUCI-positive cancer. Preferably, said MUC1-positive cancer is selected from the group consisting of lung cancer (e.g. non-small cell lung cancer), breast cancer, prostate cancer, pancreas cancer, gastric cancer, ovary cancer, fallopian tubes cancer, colorectal cancer and kidney cancer.

Typically, the administration of the composition described herein provides a therapeutic benefit to the treated subject which can be evidenced by an observable improvement of the clinical status over the baseline status or over the expected status if not treated, or treated with an irrelevant virus (not expressing the multi-specific molecule described herein) or treated with the standard of care only. An improvement of the clinical status can be easily assessed by any relevant clinical measurement typically used by physicians or other skilled healthcare staff. In the context of the invention, the therapeutic benefit can be transient (for one or a couple of months after cessation of administration) or sustained (for several months or years). As the natural course of clinical status which may vary considerably from a subject to another, it is not required that the therapeutic benefit be observed in each subject treated but in a significant number of subjects (e.g. statistically significant differences between two groups can be determined by any statistical test known in the art, such as a Tukey parametric test, the Kruskal-Wallis test the U test according to Mann and Whitney, the Student’s t-test, the Wilcoxon test, etc).

For instance, a therapeutic benefit in a subject afflicted with a cancer can be evidenced, e.g., by a reduction in the tumor number, a reduction of the tumor size, a reduction in the number or extent of metastases, an increase in the length of remission, a stabilization (i.e. not worsening) of the state of disease, a decrease of the rate of disease progression or its severity, a prolonged survival, a better response to the standard treatment, an amelioration of the disease’s surrogate markers, an improvement of quality of life, a reduced mortality, and/or prevention of the disease’s recurrence, etc.

An improvement of the clinical status can be easily assessed by any relevant clinical measurement typically used by physicians or other skilled healthcare staff. For example, techniques routinely used in laboratories such as blood tests, analysis of biological fluids and biopsies (e.g. by flow cytometry, histology, histology, immunoassays, quantitative PCR assays, transcriptomic analysis) as well as medical imaging techniques to perform tumor surveillance. Such measurements are routine in the art in medical laboratories and hospitals and numerous kits (e.g. immunoassays, quantitative PCR assays, RNA sequencing) are available commercially. They can be performed before the administration (baseline) and at various time points during treatment and after cessation of the treatment.

In another embodiment, the poxvirus or composition described herein is for use for enhancing an anti-tumoral adaptative immune response or for enhancing or prolonging an antitumor response. The present invention also provides a method or the use of the poxvirus described herein or composition thereof for stimulating or improving an immune response in the treated subject, said method or use comprising administering the composition to a subject in need thereof, in an amount sufficient according to the modalities described herein so as to stimulate or improve the subject’s immunity. The stimulated or improved immune response can be specific (i.e. directed to epitopes/antigens) and/or non-specific (innate), humoral and/or cellular, notably a CD4+ or CD8+-mediated T cell response. The ability of the composition described herein to stimulate or improve an immune response can be evaluated either in vitro (e.g. using biological samples collected from the subject) or in vivo using a variety of direct or indirect assays which are standard in the art (see for example Coligan et al. , 1992 and 1994, Current Protocols in Immunology; ed J Wiley & Sons Inc, National Institute of Health or subsequent editions). Those cited above in connection with the antigenic nature of a polypeptide are also appropriate.

Functional and therapeutic properties

In some embodiments, the poxvirus, composition or method described herein can be used to provide one or more functional or therapeutic properties such as the ones described hereinafter with a specific preference for · Ability to cause killing of tumor cells or tumor infiltrating cells expressing at their surface the targeted antigen specifically recognized by the second binding domain by immune cells expressing at their surface the targeted antigen specifically recognized by the first binding domain; and/or

• Ability to prime a lymphocyte-mediated immune response (especially against a cancer antigen); and/or

• Ability to induce the shutdown of the tumor vasculature; and/or

• Ability to induce infiltration of immune effector cells in the tumor; and/or

• Ability to kill cells infected with the poxvirus described herein; and/or

• Ability to stimulate or improve an anti-tumoral response; and/or · Ability to improve the therapeutic efficacy with respect to the non-vectorized multi specific molecule administered as such to the same type of subject or group of subjects; and/or

• Ability to reduce the toxicity with respect to the non-vectorized multi-specific molecule administered as such to the same type of subject or group of subjects.

In particular and compared to an irrelevant poxvirus or a non-vectorized multi-specific molecule (protein composition), the poxvirus (e.g. oncolytic vaccinia virus) or composition described herein comprising a nucleic acid molecule encoding a multi-specific molecule comprising at least a first binding domain binding specifically to a CD3 antigen present at the surface of a T lymphocyte and a second binding domain bonding specifically to a MUC1 antigen present at the surface of a cancer cell or a tumor infiltrating cell is also useful for at least one of the following purposes in a treated subject or a group of treated subjects

• Ability to bind immobilized recombinant targeted antigen, in particular, ability to bind CD3; and/or Ability to bind immobilized recombinant targeted antigen, in particular, ability to bind MUC1 peptide; and/or

Ability to bind the immune cells expressing at their surface the targeted antigen specifically recognized by the first binding domain, in particular, ability to bind CD3-positive immune cells; and/or

Ability to bind the cells expressing at their surface the targeted antigen specifically recognized by the second binding domain, in particular, ability to bind the cells having MUC1 antigen expressed at their surface; and/or

Ability to kill cancer cells, in particular MUC1 -positive cancer cells in presence of CD3 positive effectors cells; and/or

Ability to inhibit tumor growth, in particular MUCI-positive tumors in presence of CD3 positive effectors cells; and/or

Ability to improve the killing of tumor cells or tumor infiltrating cells, in particular MUCI-positive tumor cells as compared to a conventional virus which does not express a multi-specific molecule as described herein; and/or

• Ability to reduce the toxicity with respect to the non-vectorized multi-specific molecule administered (e.g. aCD3-aMUC1 TCE composition) as such to the same type of subject or group of subjects.

The ability of the virus of the present disclosure to express a multi-specific molecule having one or more of the functional or therapeutic properties may be assessed by the skilled person using conventional techniques available in the art. For instance, binding to the targeted antigens or cells may be assessed in vitro as described herein using techniques such as flow cytometry, ELISA, etc. The ability of the virus of the present disclosure to cause killing of cancer cells may be assessed in vitro, e.g., by estimating the number or the viability of appropriate cells after incubation with culture medium of cells infected by the TCE-armed virus as compared to with culture medium of cells infected by an unarmed virus (see e.g, the killing assay described in the Example section). It may also be evaluated in vivo in appropriate tumor models by following tumor growth or animal survival for a period of time after treatment as compared to no treatment.

Combination therapy

In some embodiments, the poxvirus, composition or methods of the invention are used as stand-alone therapy. In other embodiments, they can be used or carried out in conjunction with one or more additional therapies, in particular standard of care therapy(ies) that are appropriate for the type of cancer afflicting the treated subject. Standard-of-care therapies for different types of cancer are well known by the person skilled in the art and usually disclosed in Cancer Network and clinical practice guidelines. Such one or more additional therapy(ies) are selected from the group consisting of surgery, radiotherapy, chemotherapy, cryotherapy, hormone therapy, immunotherapy, gene therapy, photodynamic therapy and transplantation, etc. Such additional anticancer therapy/ies is/are administered to the subject in accordance with standard practice before, after, concurrently, sequentially or in an interspersed manner with the poxvirus or composition described herein. Concurrent administrations of two or more therapies do not require that the agents be administered at the same time or by the same route, as long as there is an overlap in the time period during which the composition and additional anti-cancer therapy are exerting their therapeutic effect. Concurrent administrations include administering the poxvirus composition within the same day (e.g. 0.5, 1, 2, 4, hours) as the other therapeutic agent. Although any order is contemplated by the present invention, it is preferred that the poxvirus composition be administered to the subject before the other therapeutic agent. In specific embodiments, the poxvirus or composition described herein may be used in conjunction with surgery. For example, the composition may be administered after partial or total surgical resection of a tumor (e.g. by local application within the excised zone, for example).

In other embodiments, the poxvirus or composition described herein can be used in association with radiotherapy. Those skilled in the art can readily formulate appropriate radiation therapy protocols and parameters (see for example Perez and Brady, 1992, Principles and Practice of Radiation Oncology, 2nd Ed. JB Lippincott Co; using appropriate adaptations and modifications as will be readily apparent to those skilled in the field). The types of radiation that may be used notably in cancer treatment are well known in the art and include electron beams, high-energy photons from a linear accelerator or from radioactive sources such as cobalt or cesium, protons, and neutrons. Dosage ranges for radioisotopes vary widely, and depend on the half-life of the isotope, the strength and type of radiation emitted, and the uptake by the neoplastic cells. Regular X-rays doses for prolonged periods of time (3 to 6 weeks), or high single doses are contemplated by the present invention. In certain embodiments, the poxvirus or composition described herein may be used in conjunction with chemotherapy. Chemotherapy refers to treatment of a cancer with a small molecule. Representative examples of suitable chemotherapy agents currently available for treating cancer include, without limitation, alkylating agents, topoisomerase I inhibitors, topoisomerase II inhibitors, platinum derivatives, inhibitors of tyrosine kinase receptors, cyclophosphamides, antimetabolites, DNA damaging agents and antimitotic agents. Representative examples of suitable chemotherapy agents currently available for treating infectious diseases include among other antibiotics, antimetabolites, antimitotics and antiviral drugs (e.g. interferon alpha). The chemotherapy may be administered by one or more routes of administration, e.g., intravenous subcutaneous, intradermal or intratumoral injection or by oral route. The chemotherapy may be administered according to a treatment regime. The treatment regime may be a pre-determined timetable, plan, scheme or schedule of chemotherapy administration which may be prepared by a medical practitioner.

In further embodiments, the poxvirus or composition described herein may be used in conjunction with immunotherapeutics such as anti-neoplastic antibodies as well as siRNA and antisense polynucleotides. In a preferred embodiment, the poxvirus or composition described herein is used or the method described herein is carried out in combination with one or more antibody molecule(s) that specifically binds to a check point inhibitor. Representative examples of such antibody molecule that specifically binds to a check point inhibitor include without limitation antibody against CTLA4, PD1, PDL1 , PDL2, Tim3, 0X40 among many others. In preferred embodiments, the methods and use described herein further comprise a step of administering a checkpoint inhibitor, notably an antagonistic antibody directed to PD1 or its ligand PDL1 or PDL2. Antagonist anti PD1, anti PD-L1 and anti PDL2 antibodies are available in the art from various providers such as Merck, Sigma Aldrich, AstraZeneca and Abeam and some have been FDA approved or under advanced late clinical development. Representative examples of anti PD1 antibodies usable in the present disclosure are Nivolumab (Opdivo©) and Pembrolizumab (Keytruda©). Representative examples of anti PDL1 antibodies usable in the present disclosure are e.g., BMS-936559 (under development by Bristol Myer Squibb also known as MDX-1105; WO2013/173223), atezolizumab (under development by Roche; also known as TECENTRIQ®; US8,217,149), durvalumab (AstraZeneca; also known as EVIFINZI™; WO2011/066389), MPDL3280A (under development by Genentech/Roche) as well as Avelumab (developed by Merck and Pfizer under trade name Bavencio; WO2013/079174).

In still further embodiments, the poxvirus or composition described herein may be used in conjunction with adjuvant such as TLR (Toll Like Receptor). Representative examples of suitable adjuvants include, without limitation, TLR3 ligands (Claudepierre et al., 2014, J. Virol. 88(10): 5242-55), TLR9 ligands (e.g. Fend et al., 2014, Cancer Immunol. Res. 2, 1163- 74; Carpentier et al., 2003, Frontiers in Bioscience 8, e115-127; Carpentier et al., 2006, Neuro-Oncology 8(1): 60-6; EP 1 162 982; US 7,700,569 and US 7,108,844).

The present invention also provides a kit of parts comprising a poxvirus described herein or composition thereof. In some embodiments the kit may have at least one container having a predetermined quantity of the poxvirus of the present invention. Other features, objects, and advantages of the invention will be apparent from the description and drawings and from the claims. The following examples are incorporated to illustrate preferred embodiments of the invention. However, in light of the present disclosure, those skilled in the art should appreciate that changes can be made in the specific embodiments that are disclosed without departing from the spirit and scope of the invention.

BRIEF DESCRIPTION OF THE DRAWINGS In the examples below, reference is made to the following figures:

Figure 1 : Expression of the different TCE formats in supernatants of infected and transfected cells evaluated by immunoblot

DF1 cells were infected with a vaccinia virus and transfected with shuttle plasmids encoding different formats of FLAG-tagged MUC1 TCE under the control of the pH5R promoter. After 48 hours cell supernatants were harvested and were analyzed by immunoblot after an electrophoresis under either non-reducing (left blot) or reducing (right blot) conditions and using an anti-FLAG-HRP conjugated antibody.

Figure 2: Expression of the different TCE formats in supernatants of infected and transfected cells evaluated by ELISA using MUC1 peptide to capture the TCE.

DF1 cells were infected with a vaccinia virus and transfected with shuttle plasmids encoding different formats of FLAG-tagged MUC1 TCE under the control of the pH5R promoter. After 48 hours cell supernatants were harvested and were analyzed by ELISA using coated MUC1 peptide as capture reagent. Antigen-antibody complexes were detected using an anti-FLAG-HRP conjugated antibody.

Figure 3: Expression evaluated by immunoblot of the different TCE formats in supernatants of DF1 infected by three recombinant vaccinia viruses

DF1 were infected at MOI 0.05 in triplicate with COPTG19274, COPTG19277 or COPTG19280 encoding different formats of TCE recognizing MUC1. VVTG18058 (same virus backbone but without any transgene) was used as negative control. Cell supernatants were harvested after 2 days post-infection and were analyzed by Western Blot after an electrophoresis under either non-reducing (left blot) or reducing (right blot) conditions and using an anti-FLAG -HRP conjugated antibody.

Figure 4: Expression evaluated by ELISA on MUC1 peptide of the different TCE formats in supernatants of DF1 infected by different recombinant vaccinia viruses DF1 were infected at MOI 0.05 in triplicate with COPTG19274, COPTG19277 or COPTG19280 encoding different formats of TCE recognizing MUC1. VVTG18058 (same virus backbone but without any transgene) was used as negative control. After 48 hours cell supernatants were harvested and were analyzed by ELISA using coated MUC1 peptide as capture reagent. Antigen-antibody complexes were detected using an anti-FLAG-HRP conjugated antibody.

Figure 5: Replication of COPTG19274, COPTG19280, COPTG19365 and WTG17137 in two human tumor cell lines HCT116 (A), LoVo (B) and in CEF (C) Tumor cells (HCT116, LoVo) or CEF were infected in triplicate with either

COPTG19274, COPTG19280, COPTG19365 or VVTG17137 at MOI 10³ or 10² respectively. Viral concentration was estimated by qPCR at different time post-infection (i.e. 1h, 24h, 48h and 72h). Figure 6: Oncolytic activities of COPTG19274, COPTG19280, COPTG19365 and

WTG17137 in two human tumor cell lines HCT116 and LoVo

HCT116 (A) or LoVo (B) cells were infected in triplicate with either COPTG19274, COPTG19280, COPTG19365 or VVTG17137 (as benchmark) at the different MOI indicated on the graph ranging from 10² to 10⁵. A negative control corresponding of uninfected cells was also plated (Mock infected cells). Cell viability was measured 5 days after infection and reported as percentage of viability with respect to non-infected cells (i.e. mock representing 100% viability).

Figure 7: Binding to CD3+ human lymphocytes of TCE produced in the supernatants of infected COPTG19274, COPTG19277, COPTG19280, or COPTG19365 infected CEF.

CEF were infected at MOI 0.05 in triplicate with COPTG19274, COPTG19277 COPTG19280 or COPTG19365 encoding different formats of TCE recognizing MUC1. VVTG18058 (VV empty: same virus backbone but without any transgene) was used as negative control. Cell supernatants were harvested after 2 days post-infection and centrifuged and 0.1 pm filtered to remove all the virus present. Supernatants were then 10-fold concentrated using Vivaspin concentrator (5 kDa cut off). These concentrated supernatants were used undiluted, 2 or 4-fold diluted to label human PBMC purified from healthy donors. The binding of TCE to T lymphocytes was detected by using an anti-FLAG (PE anti- DYKDDDDR, Biolegend 637310) conjugated antibody. A monoclonal anti-CD3 antibody was used as positive control. The cells were analyzed using MACSQuant (Miltenyi) flow cytometer and reported as the percentage of CD3 positive cells. Figure 8: Binding to MUC1+ HeLa cells of TCE produced by infected COPTG19274, COPTG19277, COPTG19280, or COPTG19365 infected CEF

Same samples as those described in figure 7 were analyzed for their ability to bind MUC1 displayed by HeLa cells. HeLa cells were labeled with supernatants at the three dilutions. The binding of TCE to HeLa cells was detected by using an anti-FLAG (PE anti- DYKDDDDR, Biolegend 637310) conjugated antibody. A monoclonal anti-MUC1 antibody (H23) was used as positive control. The cells were analyzed using MACSQuant (Miltenyi) flow cytometer. The results are presented either as the percentage of positive cells (A) or as the intensity of labeling (i.e. median of fluorescence intensity, B).

Figure 9: Binding to MUC1+ T47D cells of TCE produced by infected COPTG19274, COPTG19277, COPTG19280, or COPTG19365 infected CEF

Same samples as those described in figure 7 were analyzed for their ability to bind MUC1 displayed by T47D cells. Cells were labeled with supernatants at the three dilutions. The binding of TCE to T47D cells was detected by using an anti-FLAG (PE anti-DYKDDDDR, Biolegend 637310) conjugated antibody. A monoclonal anti-MUC1 antibody (H23) was used as positive control. The cells were analyzed using MACSQuant (Miltenyi) flow cytometer. The results are presented either as the percentage of positive cells (A) or as the intensity of labeling (i.e. median of fluorescence intensity, B).

Figure 10: In vitro HeLa killing assay of TCE produced by COPTG19274, or COPTG19280 -infected LoVo or HCT116 cells

LoVo or HCT116 human tumor cells were infected at MOI 0.05 in triplicate with COPTG19274 or COPTG19280 encoding different formats of TCE recognizing MUC1. VVTG18058 (same virus backbone but without any transgene) and mock infected cells were used as negative controls. Cell supernatants were harvested after 24 hours post-infection, centrifuged and 0.1 pm filtered to remove the virus. Either 5,10 or 20 pL of these treated supernatants were added to HeLa cells in parallel of purified human CD8 lymphocytes (10 CD8 for 1 HeLa cell) purified from PBMC of healthy donors (two donors 1 and 2 represented in (A) and (B)). The mixture was incubated 20H00 and then the cells were washed to remove dead cells and CD8 lymphocytes. The quantity of viable HeLa cells on the plate was measured using Cell Titer 96 kit (Promega G5421) following provider’s recommendations. The percentage of viable cells was reported (HeLa cells incubated with mock supernatants was used as reference: i.e.: 100% of viability). (C) represents the control with MUC1-negative HCT 116 cells and (D) the control without any PBMC.

Figure 11 : Number of cell clusters per field of observation in the different samples of the in vitro HeLa killing assay Four to five pictures per experimental conditions were taken as described in figures 11. The number of clusters of Hel_a/CD8 cells were counted using image analysis software. The mean and standard deviation of number of clusters of Hel_a/CD8 cells per field was reported.

Figure 12: In vitro tumor cells killing assay of TCE produced by COPTG19672- infected A549 cells in the presence of purified CD8 lymphocytes

A549 human tumor cells were infected at MOI 0.01 in triplicate with COPTG19672 encoding untagged TCE recognizing MUC1 (triangle). VVTG18058-infected (square) and mock-treated (circle) cells were used as negative controls. Cell supernatants were harvested after 72 hours post-infection, centrifuged and 0.1 pm filtered to remove the virus. Either 0.1, 1, 5 or 20 pL of these treated supernatants were added to HeLa (A), OVCAR3 (B) or MIA PaCa-2 (C) cells in parallel of purified human CD8 lymphocytes (10 CD8 for 1 tumor cell) purified from PBMC of healthy donors. The mixture was incubated 20H00 and then the cells were washed to remove dead cells and CD8 lymphocytes. The quantity of viable tumor cells on the plate was measured using Cell Titer 96 kit (Promega G5421) following provider’s recommendations. The percentage of viable cells was reported (tumor cells incubated with mock supernatants was used as reference of 100% of viability). Figure 13: In vitro tumor killing assay of TCE produced by COPTG19672- infected A549 cells in the presence of PBMC

A549 human tumor cells were infected at MOI 0.01 in triplicate with COPTG19672 encoding untagged TCE recognizing MUC1. VVTG18058 and mock infected cells were used as negative controls. Cell supernatants were harvested after 72 hours post-infection, centrifuged and 0.1 pm filtered to remove the virus. Two batches of COPTG19672-infected A549 cells supernatants (1 or 2L) were tested. Twenty pL of these treated supernatants were added to MIA-PaCa-2 (A) or HeLa (B and C) cells in parallel of PBMC from two different healthy donors (B and C) (10 PBMC for 1 HeLa cell) or purified human CD8 lymphocytes (D) (10 CD8 for 1 HeLa cell). The mixture was incubated 20H00 and then the cells were washed to remove dead cells and lymphocytes. The quantity of viable tumor cells on the plates was measured using Cell Titer 96 kit (Promega G5421) following provider’s recommendations. The percentage of viable cells was reported (tumor cells incubated with mock supernatants represent 100% of viability). Figures 14-17: In vitro activation of T lymphocytes by TCE produced by

COPTG19672-infected A549 cells in the tumor killing assay

A549 human tumor cells were infected at MOI 0.01 in triplicate with COPTG19672 encoding untagged TCE recognizing MUC1. VVTG18058 and mock infected cells were used as negative controls. Cell supernatants were harvested after 72 hours post-infection, centrifuged and 0.1 pm filtered to remove the virus. Twenty pL of these treated supernatants were added to HeLa cells in parallel of purified PBMC from four different healthy donors (10 PBMC for 1 HeLa cell; Don 1 to 4). The mixture was incubated 20H00 and then the cell culture medium containing the dead tumor cells and the lymphocytes was recovered. T Lymphocytes were stained using anti-CD3 (BioLegend 300406), anti-CD8 (Miltenyi 130-113-154), anti CD25 (BD 555432) and anti-CD69 (Miltenyi 130-112-610) fluorescently conjugated antibodies. Cell populations were analyzed using a MACSGuantIO (Miltenyi) flow cytometer. The percentage of CD4+CD25+ (Figure 14), CD4+CD69+ (Figure 15) CD8+CD25+ (Figure 16) and CD8+CD69+ (Figure 17), was determined using Kaluza software (Beckman Coulter). Live CD3 CD8 double positive cells were gated in a dot plot and from this gate the CD3+CD8+CD25+ or CD3+CD8+CD69+. CD4+ cells were determined as live CD3+CD8-. In each Figure, (a), (b), (c) and (d) represent the results obtained with the mixture of VV supernatant, HeLa and PBMC obtained from an individual donor and (e), (f), (g) and (h) are controls omitting HeLa cells.

EXAMPLES

Example 1 : Generation of plasmid vectors expressing anti-hMUC1 TCE and screening of the different constructions by infection/transfection

A series of plasmid constructs were generated to evaluate the most effective antibody format(s). All genes encoding the different constructions were produced by synthetic ways. These genes were further subcloned into shuttle plasmid using regular molecular biology techniques. Three antibody formats were tested, respectively tandem scFv, diabodies and dimeric tandem scFv using VH and VL regions of the anti-CD3 OKT3 (available at Orthoclone) and V_H and V_L regions of the anti-MUC1 antibody SM3 (described by Paul et al., 2000, Human Gene Ther 10(11 )-1417-28) or HMFG2 (described by Wilkie et al., 2008, J. Immunol 180(7): 4901-9). Table 3 provides a summary of the plasmid constructions illustrated herein and their main structural features.

All constructions described in Table 3 are under the control of the vaccinia virus pH5R promoter (SEQ ID NO: 46) and contain at their N-terminus a signal peptide (MGLGLQWVFFVALLKGVHC; SEQ ID NO: 10) and at C-terminus the FLAG and HIS tags (GSDYKDDDDKHHHHHH; SEQ ID NO: 13) for detection and purification (except pTG19672 that does not contain any tag). All these constructions were made in shuttle plasmid pTG19161 for insertion in the TK locus of the VV genome.

Co-infection/transfection in DF1 cells (ATCC CRL-12203, chicken fibroblasts spontaneously transformed) was carried out with the goal of selecting the most effective TCE construction(s) to be vectorized in VV genome. Briefly, DF1 cells were cultured in 6-well culture plate before being infected at MOI 1 by an empty vaccinia virus without any transgene (VVTG18058). pTG15839 (encoding GFP) was used as a negative control. After 2 hours, 1 pg of the different plasmids described above complexed with 4.5 pL of lipofectamine 2000 (Invitrogen ref 11668-027) in opti-MEM culture medium was added to each well. The plates were then incubated 48H at 37 °C and 5 % CO2. The culture supernatants were then centrifuged and filtrated on 0.2 pm filters to remove most of the virus particles and cellular debris.

Cell supernatants were analyzed by Western Blot (WB) after an electrophoresis under either non-reducing or reducing conditions. Blots were developed using an anti-FLAG- HRP (Sigma, ref A8592) conjugated antibody. The results illustrated in Figure 1 indicate that the WB profile under reducing condition

(right blot) of the TCE produced by infected DF1 migrated at a level close to their calculated masses (i.e. , 54 kDa for tandem scFv (pTG19273 and pTG19274) and diabodies (pTG19275, PTG19276, pTG 19277 and pTG19278) and 72 kDa for dimerictandem scFv (pTG19280)). However, the intensity of signal is much stronger for the three tandem scFv constructs (both monomeric and dimeric) than for the four diabodies constructs, indicating that tandem scFv are expressed at a higher level than diabody constructs.

Moreover, under non-reducing conditions (left blot) the profile of expression was similar to the one observed under reducing conditions except that the dimeric tandem scFV (pTG19280) appeared as disulfide crosslinked multimers (mostly dimers). This result was not expected as there are no natural inter-chain disulfide bonds in the dimeric construct. These crosslinked disulfide molecules might form by oxidation of the unpaired cysteine that is present in the VH of OKT3. As expected, no polypeptide was highlighted when transfection was performed with pTG 15839 negative control plasmid.

The amount of secreted TCE was roughly estimated by ELISA in supernatants of infected/transfected DF1 cells. ELISA allowed to measure semi-quantitively the amount of polypeptide able to bind one of the two targets (i.e. MUC1) produced in cell supernatants. pTG15839 (encoding GFP) was used as a negative control.

Briefly, microplates were coated by an overnight incubation at 4 °C with 100 pL per well of a 24-mer MUC1 tandem repeat peptide (TAPPAHGVTSAPDTRPAPGSTAPP, ProteoGenix; SEQ ID NO: 49) at 3 pg/mL. After incubation, the coating solution was discarded, blocking solution was added and plates were incubated for 1 to 2 hours at room temperature (RT) before being washed. Samples (culture supernatants) were added to the wells and 2-fold serially diluted in blocking buffer. The plates were then incubated for 2 hours at 37°C before being washed. Anti-Flag-HRP conjugated antibody diluted in blocking solution was added to each well and plates were incubated 1 hour at 37°C before being washed. After incubation with TMB (3,3',5,5'-Tetramethylbenzidine: HRP’s subtrate) solution 30 min at RT in darkness, 2 M of H2SO4 (stop solution) was added to each well to stop the enzymatic reaction. Absorbances were read at 450 nm on microplate reader. Absorbances were plotted versus supernatant dilutions. As illustrated in Figure 2, TCE able to bind MUC1 peptide were detected for 5 out of the 7 constructions tested. Indeed, no significative signal was observed for the two diabodies containing SM3 molecule. Thus, HMFG2 containing constructs displayed a higher binding to MUC1 peptide than SM3 containing constructs whatever the format (either tandem scFv or diabody) whereas immunoblot did not shown any strong difference in term of level of expression. Together, these results indicate that HMFG2 antibody is a better MUC1 binder than SM3.

Moreover, the results show that pTG19280 (dimeric HMFG2 tandem scFv) displayed the highest signal followed by pTG19274 (HMFG2 tandem scFv) and pTG19277 (HMFG2 diabody starting with HMFG2’s VH region at the N terminus of the TCE).

Example 2 : Generation of TCE-armed vaccinia viruses

Three recombinant tk- rr- deleted Copenhagen viruses containing at their J2R locus the expression cassette obtained from pTG19274, pTG19277 or pTG19280 were generated (viruses COPTG19274, COPTG19277 and COPTG19280 respectively). They also contained a mCherry expression cassette in the I4L locus. A fourth virus (COPTG 19365) was generated encoding the pTG19274 expression cassette in J2R locus and the human FLT3 ligand under the pH5R promoter in the I4L locus. Another virus named COPTG19672 was also generated that corresponds to the same construction as COPTG19274 except that the tags were omitted and that it contains no expression cassette in the I4L deleted locus. This virus is therefore compatible with a potential clinical development.

Generation of COPTG19374. COPTG19277 and COPTG1928Q The vaccinia virus transfer plasmid, pTG19161, was designed to allow insertion of nucleotide sequences by homologous recombination in J2R locus of the vaccinia virus genome. It originates from the plasmid pUC18 into which were cloned the flanking sequences (BRG and BRD) surrounding the J2R ) locus. It also contains the pH5R promoter (SEQ ID NO: 46). Synthetic fragments coding for the tandem scFv (htscFv2; SEQ ID NO: 43), the diabody (hscDb3; SEQ ID NO: 44) and the dimeric tandem scFv (HMFG2-H-CH3-OKT3; SEQ ID NO: 45) preceded by the signal peptide (e.g., SEQ ID NO: 10) and followed by the Tags (SEQ ID NO: 13) were generated by synthetic way. The coding sequences were optimized for human codon usage, a Kozak sequence (ACC) was added before the ATG start codon and a transcriptional terminator (TTTTTNT) was added after the stop codon. Moreover, some patterns into the open reading frames were excluded: TTTTTNT, GGGGG), CCCCC which are deleterious for expression in poxvirus.

As described above, the htscFv2, hscDb3, and HMFG2-H-CH3-OKT3 fragments were inserted in pTG19161 restricted with Pstl and EcoRI by homologous recombination, giving rise to pTG19274, pTG19277 and pTG19280, respectively.

COPTG19274, COPTG19277 and COPTG19280 were generated on chicken embryo fibroblast (CEF) by homologous recombination for insertion in J2R locus and by using COPTG19156 as parental virus and the transfer plasmids pTG19274, pTG19277 and pTG19280. COPTG19156 contains the expression cassette of a mCherry at the I4L locus and the expression cassette of the GFP gene in its J2R locus. CEF were isolated from 12 day-old embryonated Specific pathogen free (SPF) eggs (Charles River). The embryos were mechanically dilacerated, solubilized in a Tryple Select solution (Invitrogen) and dissociated cells cultured in MBE (Eagle Based Medium; Gibco) supplemented with 5% FCS (Gibco) and 2 mM L-glutamine. The homologous recombination between the transfer plasmids pTG19274, pTG19277, pTG19280 and the parental COPTG19156 enables the generation of recombinant vaccinia viruses which have lost their GFP expression cassette and gained the expression cassettes and the selection was performed by isolation of red fluorescent plaques. The viral stocks of COPTG19274, COPTG19277 and COPTG19280 were amplified on CEFs in two F175 flasks to generate appropriate stocks of viruses which can be aliquoted and stored at -20°C until use. Viral stocks were titrated on CEF or Vero cells and infectious titers were expressed in pfu/mL. For illustrative purposes, the produced viral stock of COPTG19274 titrated 7.6x10⁶ pfu/mL. These stocks were analyzed by PCR to verify the integrity of the expression cassette and recombination arms using appropriate primer pairs. The stock was also analyzed by sequencing of expression cassette. Alignment of sequencing results showed 100% homology with the theoretical expected sequence. If needed, viral preparations were purified using conventional techniques (e.g. as described in W02007/147528). Briefly, the crude harvest containing infected cells and culture supernatants was recovered 72h post infection and stored at -20°C until use. After thawing, this suspension was homogenized in order to release viral particles. Large cellular debris were then eliminated by depth filtration. The clarified viral suspension was subsequently concentrated and diafiltered with the formulation buffer by using tangential flow filtration and size hollow fiber microfiltration filters. Finally, the purified virus was further concentrated using the same tangential flow filtration system, aliquoted and stored at -80°C until use. Generation of COPTG 19365

The vaccinia virus transfer plasmid, pTG19334, was designed to allow insertion of nucleotide sequences by homologous recombination in I4L locus of the vaccinia virus genome. It originates from the plasmid pUC18 into which were cloned the flanking sequences (BRG and BRD) surrounding the I4L locus. This plasmid contains also the pH5R promoter (SEQ ID NO: 46).

A synthetic fragment named “FLT3L-Hu” of 599 bp containing the human FLT3-ligand encoding sequence (SEQ ID NO: 48) was generated by synthetic way. The coding sequences were optimized for human codon usage, a Kozak sequence (ACC) was added before the ATG start codon and a transcriptional terminator (TTTTTNT) was added after the stop codon. Moreover, some patterns into the open reading frames were excluded: TTTTTNT, GGGGG, CCCCC which are deleterious for expression in poxvirus. The FLT3L-encoding fragment was inserted in pTG19334 restricted with Pvull by homologous recombination, giving rise to pTG19365 with the FLT3L gene under the control of pH5R promoter.

COPTG19365 was generated by homologous recombination between the transfer plasmid pTG19365 and the virus COPTG19274, which contains the tandem scFv expression cassette at the J2R locus and a mCherry expression cassette at the I4L locus. The homologous recombination enables the generation of recombinant vaccinia virus which has lost its mCherry expression cassette and gained the FLT3-L expression cassette and the selection was performed by isolation of non -fluorescent plaques. The process of generation of viral stocks in CEF was described above.

Generation of COPTG 19672 The plasmid pTG19274, coding for the tandem scFv, was modified in order to remove the C-terminal Tags in the expression cassette, giving rise to pTG19672.

COPTG19672 was generated on chicken embryo fibroblast (CEF) by homologous recombination for insertion in J2R locus and by using COPTG19104 as parental virus and the transfer plasmids pTG19672. COPTG19104 contains the expression cassette of the mCherry at the J2R locus and is deleted in I4L locus. The homologous recombination enables the generation of recombinant vaccinia virus which has lost its mCherry expression cassette and gained the tandem scFv expression cassette and the selection was performed by isolation of non -fluorescent plaques. The process of generation of viral stocks in CEF was described above.

Example 3: Level of expression and MUC1 binding of TCE expressed by COPTG19274, COPTG19277 and COPTG1928Q.

Virus-mediated expression of the different TCE formats described in Example 2 was evaluated in supernatants of DF1 infected with COPTG19274, COPTG19277, and COPTG19280 by immunoblot. A non-relevant Vaccinia Virus (VVTG18058) was used as negative control. VVTG18058 is a vaccinia virus (Copenhagen strain) deleted in J2R and I4L loci encoding no transgene. DF1 cells were infected at MOI 0.05 with COPTG19274, COPTG19277 or COPTG19280 viruses in triplicate. Cell supernatants were harvested after 48 hours of infection and were analyzed by WB after an electrophoresis under either non reducing (left blot) or reducing (right blot) conditions. Blots were developed using an anti- FLAG-HRP (Sigma, ref A8592) conjugated antibody.

The results illustrated in Figure 3 confirmed those obtained by the co infection/transfection method. More precisely: - Under reducing conditions, the TCE produced by infected DF1 migrated at a level close to their calculated masses (i.e., 54 kDa for tandem scFv and diabodies and 72 kDa for dimeric tandem scFv).

- the intensity of signal is much stronger for the 2 tandem scFv constructs

(monomeric and dimeric) than for the diabody construct (pTG19277) confirming that tandem scFv are expressed at a higher level than diabody constructs.

- Under non-reducing conditions the profile the dimeric tandem scFV

(COPTG19280) appeared as disulfide crosslinked multimers (mostly dimers).

The ability to bind MUC1 peptide of the TCE secreted in supernatants of infected DF1 cells was evaluated by ELISA, as described above. VVTG18058 was used as a negative control. As illustrated in Figure 4, the three constructs tested produced a TCE able to bind MUC1 peptide. However, the results confirm those of Western blot and co infection/transfection experiments. More specifically, the aCD3-aMUC1 TCE produced in the supernatants of cells infected with COPTG19280 (dimeric HMFG2 tandem scFv) and COPTG19274 (HMFG2 tandem scFv) displayed the highest signals. On the other hand, aCD3-HMFG2 diabody produced in the supernatants of cells infected with COPTG19277 had a MUC1 binding signal at least 10-fold lower than the TCEs of the two other viruses probably because of a lower expression level (observed by immunoblot).

Example 4: Replication assays and oncolytic activities of Vaccinia viruses expressing MUC1 TCE

Replication of COPTG 19274, COPTG 19280 and COPTG19365 Replication of TCE-expressing viruses was evaluated on CEF (cells of production) isolated from 11 or 12 day-old embryonated SPF eggs (Charles Rivers) and on two human tumor cell lines LoVo (ATCC^® CCL-229™) and HCT116 (ATCC^® CCL-247™). VVTG17137 was used as a benchmark of virus replication. VVTG17137 is a vaccinia virus (Copenhagen strain) deleted in J2R and I4L loci encoding the suicide gene FCU1 (described in W02009/065546). CEF and tumor cells were prepared in suspension and infected at MOI of 10² and 10³ respectively (three wells per cells and per time point). After different times of incubation (24, 48 and 72h), viral concentration was measured by qPCR. As illustrated in Figure 5, the replication of the three TCE-armed viruses was similar to the one displayed by VVTG17137 in (B) the human colorectal adenocarcinoma LoVo cell line (ATCC® CCL- 229™), (A) the human colorectal adenocarcinoma HCT 116 cell line (ATCC® CCL-247™) and (C) CEF demonstrating that expression of tandem scFv TCE (COPTG19274), dimeric tandem scFv TCE (COPTG 19280) or tandem scFv TCE and FLT3L (COPTG19365) did not impair viral replication. Oncolytic activities of COPTG 19274. COPTG 19280 and COPTG19365

Oncolytic activity is representative of the lytic activity of the tested viral samples on tumor cells. It was assessed by quantification of cell viability after 5 days of incubation on the two human colorectal adenocarcinoma cell lines LoVo (ATCC® CCL-229™) and HCT116 (ATCC® CCL-247™). COPTG19274, COPTG19280 and COPTG19365 oncolytic activities were compared to the one of VVTG17137 as benchmark. A negative control corresponding of uninfected cells was also plated (Mock infected cells).

Cells were prepared, distributed in Eppendorf tubes (1.2x10⁶ cells/tube) before being infected with the virus at a MOI of 10⁵ to 10² and incubated 30 min at 37°C. Appropriate complete medium was added to Eppendorf tube and an aliquot of this suspension was added in each well (in triplicate) in 6-well plate containing 2 mL of appropriate complete medium. Plates were incubated at 37°C with 5% CO2 for 5 days and cell viability was determined on Vi-Cell counter. Results were expressed as a percentage of the cell viability with respect to mock infected cells (representing 100% viability). The results show that TCE-expressing COPTG19274, COPTG19280 and COPTG19365 displayed the same oncolytic activity as the one displayed by the VVTG17137 benchmark virus in the two tumor cell lines assessed, respectively HCT 116 (Figure 6A) and LoVo (Figure 6B) demonstrating that expression of the different formats of TCE (COPTG19274 and COPTG19280) eventually co-expressed with FLT3L (COPTG19365) did not impair oncolytic activity of the recombinant vaccinia virus. Interestingly, for the four viruses, a substantial oncolytic activity was reached at a weaker MOI (10⁴) in HCT 116 cells than in LoVo cells (1 O³), confirming the resistance nature of the LoVo cells. Example 5: Functionality of the TCE produced by the recombinant vaccinia viruses

Interaction of VV-produced TCE with CD3-positive human lymphocytes CEF were infected at MOI 0.05 with one of the following viruses COPTG19274, COPTG19277, COPTG19280 and COPTG19365 and infected cells were cultured for 48H. VVTG18058 (same viral backbone but without transgene) was used as negative control. The supernatants were then harvested, centrifuged and 0.1pm filtered to remove all the viral particles (as described in Example 1). The supernatants were concentrated 10-fold using a concentrator (Vivaspin) with a 5 kDa cut off. 50 pL of either the undiluted or 2 or 4-fold diluted concentrated supernatants, were then added to human PBMC. The binding of the VV- produced TCE to lymphocytes T was detected by using an anti-FLAG (Phycoerythrin (PE) labeled anti-DYKDDDDR, Biolegend 637310) conjugated antibody. Anti-CD3 monoclonal antibody (PerCPVio700 anti-CD3, Miltenyi 130-097-582) was used as positive control. The percentage of CD3 positive cells in PBMC was assessed by flow cytometry (MACSQuant; Miltenyi). As shown in Figure 7, the supernatants of the cells infected by the four viruses tested

(COPTG19274, COPTG19277, COPTG19280 and COPTG19365) were all able to label the T lymphocytes in human PBMC and in a same extent as the benchmark anti-CD3 antibody. This labeling is specific as it was not observed with supernatants of either uninfected CEF or CEF-infected with a non-armed virus (VVTG18058: VV empty). Note that about the same percentage of labeled cells was observed for the three dilutions tested for each supernatant.

Interaction of VV-produced TCE with MU C1 -positive human lymphocytes The concentrated supernatants obtained from CEF cells infected with COPTG19274, COPTG 19277, COPTG 19280, COPTG 19365 or VVTG 18058 (negative control) were harvested and treated as described above. 50 pL of either the undiluted or of 2 or 4-fold diluted concentrated supernatants, were then added to MUCI-positive human tumor cells, HeLa (Figure 8) and T47D (Figure 9), respectively. The binding to the MUCI-positive cells was detected using an anti-FLAG (PE anti-DYKDDDDR, Biolegend 637310) conjugated antibody. H23 anti-MUC1 antibody was used as positive control and the percentage of labelled cells was assessed by flow cytometry using MACSQuant (Miltenyi) flow cytometer. The results are presented either as the percentage of labelled cells (A) or as the intensity of labelling (i.e. median of fluorescence intensity) (B). As illustrated in Figures 8A and 9A, the supernatants of the four viruses tested

(COPTG19274, COPTG19277, COPTG19280 and COPTG19365) were all able to label the two human tumor MUCI-positive cells HeLa (Figure 8A) and T47D (Figure 9A) and in the same extent as the benchmark anti-MUC1 antibody. This labeling is specific as it was not observed with supernatants of uninfected CEF or CEF-infected with a non-armed virus (VVTG 18058). About the same percentage of labeled cells was observed for the three dilutions tested for each supernatant.

The intensity of labeling was also measured for each construct and for the three dilutions used (Figures 8B and 9B). This measurement showed that supernatants collected from COPTG19274 and COPTG19280 infected cells producing tandem scFv format (mono or dimeric) displayed a stronger labeling than the one provided by supernatants collected from COPTG19277 producing diabody TCE.

Example 6: Killing of MUC1+ human tumor cells by TCE produced by recombinant vaccinia viruses infected cells The ability of the VV-produced TCE to lyse tumor cells was tested on two human colorectal cell lines LoVo and HCT116. Each cell line was infected at MOI 0.05 with COPTG19274 or COPTG19280. VVTG 18058- infected cells and uninfected (mock) cells were used as negative controls. The supernatants were then harvested 24H post infection and treated as described above (centrifuged and 0,1 pm filtered) to remove all virus particles. In parallel of this process, human CD8 lymphocytes were purified from two different healthy donors using StraightFromTM Buffy Coat CD8 Microbead Kit (Miltenyi order N°130- 114-978) following provider’s recommendations.

The virus-free supernatants containing the VV-produced TCE (obtained from COPTG19274 or COPTG19280-infected cells), or the negative control supernatants (uninfected or VVTG18058-infected supernatants), were then added at different volumes (5, 10 and 20 pL) to either HeLa (MUC1⁺) or HCT116 (MUCT) cells in presence, or not, of purified human CD8 lymphocytes. The ratio CD8:HeLa cells was 10:1. After 24H of incubation, the culture medium was removed and the adherent cells (HeLa or HCT116) washed to eliminate both non-adherent dead cells and CD8 lymphocytes. The number of viable cells on the plate was then measured using Cell Titer 96 kit (Promega G5421 kit) and following the provider’s recommendations. The percentage of viable cells was calculated as following: lOOxODsample/ODmock and reported for each volume of supernatant added. Wells where 20 mI_ of TCE containing supernatants were added but where the CD8 lymphocytes were omitted was used as negative control.

The results demonstrate a dose-effect killing on HeLa cells of supernatants harvested from HCT116 and LoVo infected by COPTG19274 or COPTG 19280, but not by VVTG 18058. This dose-effect killing was observed for CD8 purified from two different donors (Figures 10A and 10B). In marked contrast, cell survival remained at 100% when COPTG19274 or COPTG19280 supernatants are added to the MUCI-negative HCT116 cells (Figure 10C) or if the CD8 lymphocytes were omitted in the wells (Figure 10D). As expected, 100% cell viability was also observed with supernatants of VVTG 18058- infected cells whatever the cell targets, HeLa and HCT116 respectively (Figures 10A-D).

Cells were also observed with microscope just before their washes and pictures were taken of different fields at the 100-fold magnification (10x objective lens). Clusters of cells (HeLa and CD8) were clearly observed in wells where HeLa cells were put in presence of TCE produced by either COPTG19274 or COPTG19280. These clusters were counted on 4- 5 different fields for each condition and the results are illustrated in Figure 11. The picture analysis allowed to quantify the cell clusters and confirmed that those were observable only when TCE from either COPTG19274 or COPTG19280 were present. Mock or VVTG18058- treated cells did not produce any clusters.

Together these results demonstrate that the TCE produced by two tumor cells lines infected by COPTG19274 or COPTG19280 are functional, i.e. able to trigger the killing of MUC1+ cells (i.e. HeLa) by effector cells (i.e. CD8 lymphocytes). Moreover, this killing is specific as MUC1 negative cells are spared and is strictly dependent of the presence of CD8 lymphocytes. Moreover, the ability of killing MUC1+ tumor cells by the TCE expressed in infected tumor cells was assessed using the vaccinia virus COPTG19672 (described in Example 2) expressing the untagged aMUC1-aCD3 TCE molecule. In this experiment dose-response curve was generated using different volumes of culture supernatant of A549 COPTG19672- infected cells on three MUC1+ tumor cells, respectively HeLa (A) (human cervix cells adenocarcinoma; ATCC® CCL-2™), OVCAR3 (B) (human ovary adenocarcinoma; ATCC® HTB-161™) and MIA PaCa-2 (C) (human pancreas carcinoma; ATCC CRL-1420™). Specifically, A549 human tumor cells were infected at MOI 0.01 in triplicate with COPTG19672. VVTG18058 and mock infected cells were used as negative controls. Cell supernatants were harvested after 72 hours post-infection, centrifuged and 0.1 pm filtered to fully remove the virus. Either 0.1 , 1 , 5 or 20 pL of these treated supernatants were added to HeLa, MIA PaCa-2 or OVCAR3 cells in parallel of purified human CD8 lymphocytes (10 CD8 for 1 tumor cell) purified from PBMC of healthy donors. The mixture was incubated 20H00 and then the cells were washed to remove dead cells and CD8 lymphocytes. The quantity of viable tumor cells on the plate was measured using Cell Titer 96 kit (Promega G5421) following provider’s recommendations. The percentage of viable cells was reported (tumor cells incubated with mock supernatants was used as reference: i.e. : 100% of viability). The results displayed Figure 12 A-C demonstrated that supernatant harvested from Mock-treated and VVTG18058-infected cells do not affect the viability of the MUCI-positive tumor cells whereas the TCE produced by COPTG19672 decrease cell viability (below 50% from 5mI_ of cell supernatant) in the three cases.

The same killing experiment was carried out on HeLa and MIA PaCa-2 MUC1+ tumor cells, except that PBMC were used instead of purified CD8 lymphocytes. A control with purified CD8 lymphocytes (D) was also included. Specifically, A549 human tumor cells were infected at MOI 0.01 in triplicate with COPTG19672 encoding the untagged TCE recognizing MUC1. VVTG18058 and mock infected cells were used as negative controls. Cell supernatants were harvested after 72 hours post-infection, centrifuged and 0.1 pm filtered to remove the virus. Twenty mI_ of these treated supernatants were added to MIA-PaCa-2 (A) or HeLa (B and C) cells in parallel of PBMC from two different healthy donors (10 PBMC for 1 HeLa cell) (A, B and C) or purified human CD8 lymphocytes (D) (10 CD8 for 1 HeLa cell). The mixture was incubated 20H00 and then the cells were washed to remove dead cells and lymphocytes. The quantity of viable tumor cells on the plates was measured using Cell Titer 96 kit (Promega G5421) following provider’s recommendations. The percentage of viable cells was reported (tumor cells incubated with mock supernatants represent 100% of viability). The results displayed Figure 13 A-D demonstrated that supernatant harvested from Mock- treated and VVTG18058-infected cells do not affect the viability of the MUC1-positive tumor cells whereas the TCE produced by COPTG19672 decrease viability of both HeLa and MIA PaCa-2 cells. PBMC-mediated cell effect is less stringent than with purified CD8 lymphocytes due to the fact that proportion of CD8 in PBMC is less important (typically PBMCs count approximately 70% of CD4 and CD8 lymphocytes in average).

Lymphocyte activation was studied by measurement of CD25 and CD69, two well- known activation markers of lymphocytes, on surface of both CD4 and CD8 present in PBMC and in presence or absence of culture supernatants containing COPTG19672’s TCE produced by infected A549 tumor cells. A549 human tumor cells were infected at MOI 0.01 in triplicate with COPTG19672 encoding untagged TCE recognizing MUC1. VVTG18058 and mock infected cells were used as negative controls as well as a non-treated control (co). Cell supernatants were harvested after 72 hours post-infection, centrifuged and 0.1 pm filtered to remove the virus. Twenty pL of these treated supernatants were added to HeLa cells in parallel of purified PBMC from four different healthy donors (10 PBMC for 1 HeLa cell). The mixture was incubated 20H00 and then the cell culture medium containing the dead tumor cells and the lymphocytes was recovered. T Lymphocytes were stained using anti-CD3 (BioLegend 300406), anti-CD8 (Miltenyi 130-113-154) anti CD25 (BD 555432) and anti-CD69 (Miltenyi 130-112-610) fluorescently conjugated antibodies. Cell populations were analyzed using a MACSQuantIO (Miltenyi) flow cytometer. The percentage of CD4+CD25+ (Figure 14), CD4+CD69+ (Figure 15) CD8+CD25+ (Figure 16) and CD8+CD69+ (Figure 17) was determined using Kaluza software (Beckman Coulter). Live CD3 CD8 double positive cells were gated in a dot plot and from this gate the CD3+CD8+CD25+ or CD3+CD8+CD69+. CD4+ cells were determined as live CD3+CD8-. In each of Figures 14 to Figure 17, (a) to (d) illustrate HeLa cells treated with the vaccinia (COPTG19672 or VVTG18058 control) or mock supernatants with purified PBMC from four different healthy donors (“Don 1 to 4) whereas (e) to (h) represent the same experiment without tumor cells (the vaccinia or mock supernatants with purified PBMC without HeLa targets).

Clearly, the supernatant containing TCE encoded by COPTG19672 activated both CD4 (Figures 14 and 15) and CD8 (Figures 16 and 17) T lymphocytes present in PBMC preparations from several healthy donors only in presence of tumor cells. Indeed, the percentage of CD25+ and CD69+ surface markers is significantly higher when HeLa cells are treated with COPTG19672 supernatants and PBMC than upon treatment with VVTG18058 supernatant or mock treatment or no treatment (“co” slots) whatever the PBMC donor (panels (a) to (d)). Moreover, this lymphocytes activation is strictly dependent of the engagement of TCE on tumor cells as shown on panels (e) to (h) of Figures 14-17 where tumor cells were omitted in the assay and where no activation of T lymphocytes by TCE was observed. Therefore, the destruction of tumor cells by either purified PBMC from four healthy donors is coming along with an activation of both CD4 and CD8 Lymphocytes as demonstrated on Figures 14-17.

Sequence listing

Table 4: Sequences used in the examples

REFERENCES

Acres and Limacher, 2005, Expert Rev Vaccines 4(4): 1-10 Adair and Lawson, 2005, Drug Design Reviews 2(3), 209-217 Agrawal et al., 1998, Nat. Med. 4: 43-9 Al-Lazikani et al., 1997, JMB 273: 927-48 Brischwein et al., 2006, Mol Immunol 43(8): 1129-43 Burchell et al., 1987, Cancer res 47: 5476-82 Carpentier et al., 2003, Frontiers in Bioscience 8, e115-127 Carpentier et al., 2006, Neuro-Oncology 8(1): 60-6 Chakrabarti et al., 1997, Biotechniques 23: 1094-7 Chen et al., 2013, Adv Drug Deliv Rev 65(10): 1357-69 Chothia et al.,1987, J. Mol. Biol. 196:901-17 Claudepierre et al., 2014, J. Virol. 88(10): 5242-55 Erbs et al., 2008, Cancer Gene Ther. 15(1): 18-28 Evans et al., 2004, J Pharm Sci. 93:2458-75

Fallaux et al., 1998, Human Gene Ther. 9: 1909-17 Fend et al., 2014, Cancer Immunol. Res. 2, 1163-74 Foloppe et al., 2008, Gene Ther., 15: 1361-71

Foloppe et al., 2019 Mol Ther Oncolytics, 14:1-14; doi10.1016/j.omto.2019.03.005 Frank and Sippl, 2008, Bioinformatics 24: 2172-2176

Gubser et al., 2004, Poxvirus genomes: a phylogenic analysis. J. Gen. Virol. 85:105-117 Guse et al., 2011, Expert Opinion Biol. Ther.11(5):595-608 Hammond et al., 1997, J. Virol Methods 66: 135-8 Harrington et al., 2019, Nat. Rev Drug Discov, 18: 689-706 Heo et al., 2013, Nat Med 19: 329-36

Holliger and Hudson, 2005, Nature Biotech. 23(9) : 1126-1136 Huehls et al., 2015, Immunol Cell Biol. 93(3): 290-6

Kabat et al., 1991, Sequences of Proteins of Immunological Interest, 5th Ed. Public Health Service, National Institutes of Health, Bethesda, MD Kangasniemi et al., 2006, Clin Cancer Res. 12(10): 3137-44

Karnerva et al., 2005, Gene Ther 12(1) 87-94

Kumar and Boyle, 1990, Virology 179: 151-8 Lefranc et al. , 2003, Dev. Comp. Immunol. 27:55-77 Liu et al., 2017, Eur. J. Cancer Care 26(5): DO110.1111/ecc.12446 Liu et al., 2018, J. Virol. 92(7) e02152-17 Mell et al., 2017, Clin. Cancer. Res 23: 5696-702 Orubu et al., 2012, PloS One 7(6)e40167 Paul et al., 2000, Hum Gene Ther, 10(11): 1417-28 Peat et al., 1992, Cancer Res 52: 1954-60

Perez and Brady, 1992, Principles and Practice of Radiation Oncology, 2nd Ed. JB Lippincott Co Petersen et al., 2011, Nature Methods 8: 785-786

Remington: The Science and Practice of Pharmacy, A. Gennaro, Lippincott, Williams&Wilkins

Retter et al., 2005, Nucl. Acids Res. 33 (1): D671-D674

J. R. Robinson in “Sustained and Controlled Release Drug Delivery Systems”, ed., Marcel Dekker, Inc., New York, 1978

Russell et al. , 2012, Nat. Biotechnol. 30(7): 658-70

Senzer et al, 2009, J. Clin. Oncol. 27: 5763-71

Sivanandam et al., 2019 Mol Ther 13: 93-106

Smith-Garvin et al., 2009, Ann Rev Immunol 27: 591-619 Vaha-Koskela et al., 2006, Cancer Res 66: 7185-94

Vaha-Koskela et al., 2014, Biomedecines 2(2): 163-94

Verma et al., 1998, J. Immunol. Methods, 216:165-181

Wahl et al., 1983, J. Nucl. Med. 24:316

Wilkie et al., 2008, J. Immunol. 180 (7): 4901-9 Yu et al., 2014, Mol Ther. 22(1): 102-11

Yu et al. , 2017, Cancer trans Med 3(4): 122-32

Zhang et al., 2007, Cancer Res. 67: 10038-46

Zeh et al., 2015, Mol ther 23: 202-14 EP239400 EP 1 162 982 W09 1/09967 W001/23001 W02005/042728 W02006/093924 W02006/108846 W02007/147528 W02008/129058 W02008/138533 W02008/138649 W02009/053937 W02009/065546 W02009/065547 W02009/100521 WO2010/130753 WO2010/130756 WO2011/066389 WO2012/001075 WO2013/022764 WO2013/079174 WO2013/173223 WO2014/053571 WO2014/138314 WO2016/087245 WO2016/087457 WO2016/146894 WO2016/165632 WO2017/167919 WO2018/178047 W02020/136232 W02020/136235 US2007-0161085 US 5,530,101 US 5,565,332 US 5,585,089 US 5,693,761 US 5,693,762 US 6,180,370 US 6,686,152 US 7,108,844 US 7,456,009 US 7,700,569 US 8,217,149

Claims

1. A poxvirus comprising a nucleic acid molecule encoding a multi-specific molecule comprising at least a first domain binding specifically to a cell antigen present at the surface of a T cell and a second domain binding specifically to an antigen (Ag) present at the surface of a cancer cell or tumor infiltrating cell, wherein said poxvirus is defective for ribonucleotide reductase (RR) activity.

2. The poxvirus according to claim 1 , wherein said poxvirus is an oncolytic vaccinia virus preferably selected from the group of vaccinia viruses consisting of Western Reserve

(WR), Elstree, Wyeth, Lister, Tian Tan, LIVP and Copenhagen strains.

3. The poxvirus according to claim 2, wherein said vaccinia virus is defective for both TK and RR activities.

4. The poxvirus according to anyone of claims 1 to 3, wherein said multi-specific molecule is a bispecific molecule comprising two non-identical antigen binding domains having different binding specificities.

5. The poxvirus according to claim 4, wherein each of the first and the second antigen binding domains is an antibody.

6. The poxvirus according to claim 5, wherein each of the first and second antigen- binding domain is an antibody fragment independently selected from the group consisting of Fab, Fab’, F(ab’)2, Fd, single-chain Fv (scFv), disulfide-linked Fvs (sdFv), scFab, dAb, single domain antibody fragment (sdAb), single domain antibodies from camelids and minimal recognition units.

7. The poxvirus according to claim 6, wherein each of the first and second antigen- binding domains comprises a heavy chain variable region (VH) and preferably a VH region and a light chain variable region (VL).

8. The poxvirus according to anyone of claims 4 to 7, wherein the first binding domain is an antibody that specifically binds to CD3.

9. The poxvirus according to claim 8, wherein said antibody that specifically binds to CD3 comprises at least a V_H region comprising a HC-CDR1 having an amino acid sequence as shown in SEQ ID NO:14; a HC-CDR2 having an amino acid sequence as shown in SEQ ID NO:15 and a HC-CDR3 having an amino acid sequence in SEQ ID NO:16 and/ora V_L domain comprising a LC-CDR1 having an amino acid sequence as shown in SEQ ID NO:17; a LC-CDR2 having an amino acid sequence as shown in SEQ ID NO:18 and a LC-CDR3 having an amino acid sequence in SEQ ID NO: 19.

10. The poxvirus according to claim 9, wherein said antibody that specifically binds to CD3 comprises a V_H comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO:20 and/or a VL comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO: 21 .

11. The poxvirus according to claim 10, wherein said antibody that specifically binds to CD3 comprises an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO:22.

12. The poxvirus according to anyone of claims 4 to 11 , wherein the second binding domain is an antibody specific for a cell antigen present at the surface of a tumor cell or a tumor infiltrating cell.

13. The poxvirus according to claim 12, wherein the second antigen-binding domain is an antibody that specifically binds the MUC1 antigen present at the surface of a tumor cell.

14. The poxvirus according to claim 13, wherein said antibody that specifically binds the MUC1 antigen comprises a VH comprising a HC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 23; a HC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 24 and a HC-CDR3 having an amino acid sequence in SEQ ID NO: 25 or SEQ ID NO: 26 and/or a VL comprising a LC-CDR1 having an amino acid sequence as shown in SEQ ID NO: 27; a LC-CDR2 having an amino acid sequence as shown in SEQ ID NO: 28 and a LC-CDR3 having an amino acid sequence in SEQ ID NO: 29.

15. The poxvirus according to claim 14, wherein said antibody that specifically binds the MUC1 antigen comprises a VH comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO: 30 or SEQ ID NO: 31 and a VL comprising an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO:32.

16. The poxvirus according to claim 15, wherein said antibody that specifically binds the MUC1 antigen comprises an amino acid sequence having at least 80%, advantageously at least 90% or preferably at least 95% or even 100% sequence identity with SEQ ID NO: 33.

17. The poxvirus according to anyone of claims 4 to 16, wherein the multi-specific molecule is selected from the group consisting of diabodies, triabodies, tetrabodies, minibodies, nanobodies and tandem scFv.

18. The poxvirus according to claim 17, wherein said multi-specific molecule is a tandem scFv.

19. The poxvirus according to claim 18, wherein said multi-specific molecule comprises an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ ID NO: 34, SEQ ID NO: 35 or SEQ ID NO: 36.

20. The poxvirus according to claim 17, wherein said multi-specific molecule is a dimeric tandem scFv.

21. The poxvirus according to claim 20, wherein said multi-specific molecule comprises an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ ID NO: 37, SEQ ID NO: 38 or SEQ ID NO: 39.

22. The poxvirus according to claim 17, wherein said multi-specific molecule is a single chain diabody.

23. The poxvirus according to claim 22, wherein said multi-specific molecule comprises an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with any of SEQ ID NO: 40, SEQ ID NO: 41 or SEQ ID NO: 42.

24. The poxvirus according to anyone of claims 1 to 23, wherein said nucleic acid molecule is selected from the group of nucleotide sequences having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity

(i) with SEQ ID NO: 43 or a part thereof starting at position 1 to position 1545 ;

(ii) with SEQ ID NO: 44 or a part thereof starting at position 1 to position 1485 ; and

(iii) with SEQ ID NO: 45 or a part thereof starting at position 1 to position 2010 .

25. The poxvirus according to claim 24, wherein expression of said nucleic acid molecule is driven by the pH5R promoter, and more preferably by a promoter having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO: 46.

26. The poxvirus according to anyone of claims 1 to 25, wherein said poxvirus comprising the nucleic acid molecule encoding the multi-specific molecule further comprises a nucleic acid molecule encoding one or more polypeptide of interest such as an immunomodulatory polypeptide.

27. The poxvirus according to claim 26, wherein said immunomodulatory polypeptide is Flt3 ligand (Flt3L) and, preferably, a Flt3L polypeptide comprising an amino acid sequence having at least 80%, advantageously at least 90%, preferably at least 95% or even 100% sequence identity with SEQ ID NO: 47 or the part thereof starting at Ser residue in position 21.

28. A method for producing the poxvirus according to anyone of claims 1 to 27, comprising the steps of a) preparing a producer cell, b) transfecting or infecting the prepared producer cell with the poxvirus, c) culturing the transfected or infected producer cell under suitable conditions so as to allow the production of the virus, d) recovering the produced virus from the culture of said producer cell and optionally e) purifying said recovered virus.

29. The method according to claim 28, wherein said producer cell is HeLa or CEF cells.

30. A composition comprising a therapeutically effective amount of the poxvirus according to anyone of claims 1 to 29 and a pharmaceutically acceptable vehicle.

31. The composition according to claim 30 comprising from approximately 10³ to approximately 10¹² pfu, advantageously from approximately 10⁴ pfu to approximately 10¹¹ pfu, preferably from approximately 10⁵ pfu to approximately 10¹° pfu; and more preferably from approximately 10⁶ pfu to approximately 10⁹pfu of said poxvirus.

32. The composition according to claim 31 , wherein said composition is formulated for intravenous, subcutaneous, intramuscular or intratumoral administration.

33. The poxvirus according to anyone of claims 1 to 29 or the composition according to anyone of claims 30 to 32 for use for treating a cancer in a subject in need thereof.

34. A method of treating a cancer comprising administering to a subject a poxvirus according to anyone of claims 1 to 29 or the composition according to anyone of claims 30 to 32.

35. The poxvirus or the composition for use according to claim 33 or the method of treating according to claim 34, wherein said poxvirus is a RR-defective oncolytic vaccinia virus, said first antigen-binding domain specifically binds CD3 and/or said second antigen binding domain specifically binds tumor-associated MUC1 antigen and optionally wherein said poxvirus further comprises a nucleic acid molecule encoding an immunomodulatory polypeptide such as a Flt3L.

36. The poxvirus or the composition for use or the method of treating according to claim

35, wherein said cancer to be treated is a MUCI-positive cancer.

37. The poxvirus or the composition for use or the method of treating according to claim

36, wherein said MUCI-positive cancer is selected from the group consisting of lung cancer, breast cancer, prostate cancer, pancreas cancer, gastric cancer, ovary cancer, fallopian tubes cancer, colorectal cancer and kidney cancer.

38. The poxvirus or the composition for use according to anyone of claims 33 to 37 or the method of treating according to anyone of claims 34 to 37, wherein said use or method result in one or more of the following properties:

• Ability to kill cancer cells, in particular MUCI-positive cancer cells in presence of CD3 positive effectors cells; and/or

• Ability to inhibit tumor growth, in particular MUCI-positive tumors in presence of CD3 positive effectors cells; and/or • Ability to improve the killing of tumor cells, in particular MUCI-positive tumor cells as compared to a conventional virus which does not express a multi-specific molecule as described herein; and/or