Method for identifying variants in gene products derived from gene constructs used in cell therapy.

Abstract

This invention provides a method for ensuring that gene products used in cell therapy do not carry the risk of reduced efficacy or toxicity due to the production of unintended variants. [Solution] The method includes the step of performing in silico analysis on a gene construct to identify and modify sequences that are likely to cause a variant. The method also includes the step of performing in vivo analysis of the RNA sequence of the product based on the construct. Then, variant detection may be performed based on gap reads from the RNA sequence to determine the variant expression level and variant significance. The method may include the step of repeating the in silico analysis if the identified variant is unacceptable.

Description

Technical Field

[0001] Cross - reference to related applications This application claims the priority of U.S. Provisional Patent Application No. 63 / 217,933, filed on July 2, 2021, the entire content of which is incorporated herein by reference.

[0002] This disclosure relates to methods for detecting and replacing sequences that can cause undesirable variants in gene constructs.

Background Art

[0003] In recent years, with the advancement of medical technology, the use of immunotherapy for treating different types of diseases and disorders, including various forms of cancer, has emerged. Generally, immunotherapy is the treatment of diseases by stimulating or suppressing the immune response. In many cases, modified versions of biological substances, such as the patient's own immune cells, are re - introduced into the patient to initiate and / or augment the immune response.

[0004] For example, engineered immune cells have been demonstrated to have desirable properties in therapeutic treatments, particularly in oncology. Two major types of engineered immune cells contain chimeric antigen receptors (referred to as "CAR" or "CAR - T") and T - cell receptors ("TCR"). These engineered cells are engineered to confer antigen specificity to them while retaining or enhancing their ability to recognize and kill target cells. A chimeric antigen receptor can include, for example, (i) an antigen - specific component ("antigen - binding molecule"), (ii) an extracellular domain, (iii) one or more co - stimulatory domains, and (iv) one or more activation domains. Each domain can be heterologous, i.e., composed of sequences derived from (or corresponding to) different protein chains.

[0005] The introduction of genetic elements into cells using gene constructs (such as viral vectors) is one method of producing cells for applications such as cell therapy. Gene construct production and transduction of cells require multiple biological processes that can introduce heterogeneity into the product. Due to specific requirements for viral vectors, defects in viral packaging, transduction, or transgene transcription can introduce undesirable contaminants that differ from the intended sequence. These "variants" may express unexpected protein sequences, and depending on their frequency and nature, they may reduce efficacy, impair manufacturing, or even increase side effects such as toxicity. To prevent potential program failures, it is crucial to minimize the possibility of variant production during the transgene development phase and cell therapy development.

[0006] What is needed is a systematic method for detecting and identifying variants and potential variant-causing sequences in gene products used in cell therapy applications. Furthermore, improved methods for characterizing variants and / or for risk elimination strategies, such as DNA sequence modifications to remove detected variants, are required to be usable in iterative processes once variants are identified. [Overview of the project]

[0007] In short and generally speaking, this disclosure relates to a system and method for producing gene products used in cell therapy. In one embodiment, the method involves performing in silico analysis on a gene construct to identify and modify the sequence that causes a variant. The method also includes performing an in vivo analysis, which includes a first RNA sequencing step to identify the frequency percentage of a variant, and repeating the in silico analysis if the first RNA sequencing step identifies a variant with a frequency greater than 5% in the gene construct. The in vivo analysis may include a second RNA sequencing step to identify the frequency percentage of the variant, and repeating the in silico analysis if the variant is unacceptable. The method may also include performing variant detection based on gap reads from at least one of the first and second RNA sequencing steps to determine the variant expression level and variant significance. If the variant is determined to be unacceptable, the in silico analysis may be repeated to construct a new gene construct.

[0008] In one embodiment of the disclosed method, in silico analysis may include modifying variants in a gene construct by synonymous codon substitution. Furthermore, in silico analysis may include identifying and removing homologous sequences from the gene construct. In one embodiment, in silico analysis includes identifying identical sequences in the gene construct. Furthermore, in silico analysis may include calculating a matrix of subsection combinations from an input sequence and obtaining a Hamming distance for each combination. The method in one embodiment includes substituting random synonymous codons if the substitution increases the sum of the entire matrix.

[0009] In one embodiment, the in vivo analysis includes RNA sequencing of a product made from a gene construct. In one embodiment, the RNA sequencing is performed in multiple steps to identify high-frequency variants, and then low-frequency variants. Furthermore, the in vivo analysis may include performing an analysis to determine whether the low-frequency variants should be substituted.

[0010] In one embodiment of this method, variant detection includes extracting RNA from a donor sample. Furthermore, if gap recognition alignment is performed, three separate aligners may be used in one embodiment. The p-value for variant significance is calculated using the Wilcox rank-sum test.

[0011] This disclosure also relates to a method for ensuring that gene products used in cell therapy do not carry the risk of reduced efficacy or toxicity due to the production of unintended variants. This method includes performing in silico analysis on a gene construct to identify and modify sequences likely to cause variants. The method also includes performing in vivo analysis of the RNA sequence of the construct-based product. Variant detection may then be performed based on gap reads from the RNA sequence to determine variant expression levels and variant significance. If the identified variant is unacceptable, the method may include repeating the in silico analysis.

[0012] One embodiment of the present disclosure relates to a method for detecting and substituting sequences that may cause undesirable variants in a gene construct. Such a method includes the steps of: performing an in silico analysis of a gene construct to detect the presence of sequences that may cause undesirable variants; substituting the detected sequences that may cause undesirable variants with alternative sequences, wherein the alternative sequences are induced to include synonymous codon substitutions; and performing an in vivo analysis of one or more genes expressed by a gene construct, which includes measuring the frequency percentage of undesirable variants expressed by the gene construct, and the frequency percentage of the undesirable variants is determined at least in part by using a splice-recognizing aligner from the RNA sequencing analysis; and if the frequency percentage of the undesirable variants in the gene product from the in vivo analysis is greater than a predetermined value of an acceptable frequency percentage of the undesirable variants, This includes a process of repeatedly performing a silico analysis step and a substitution step.

[0013] One embodiment of the present disclosure relates to a method for producing a gene product for use in cell therapy. Such a method includes the steps of: performing in silico analysis on a gene construct encoding the gene product to identify and modify sequences that may cause undesirable variants; substituting the detected sequences that may cause undesirable variants with alternative sequences, the alternative sequences being induced to include synonymous codon substitutions; and performing in vivo analysis of one or more genes expressed by the gene construct, which includes the step of performing RNA sequencing analysis on an RNA product transcribed from the gene construct, wherein the frequency percentage of undesirable variants expressed by the gene construct is determined, at least in part, from the RNA sequencing analysis using a splice-recognizing aligner. The method includes a step of determining the frequency percentage of undesirable variants in the gene product from in vivo analysis, and if the frequency percentage of undesirable variants in the gene product from in vivo analysis is greater than a predetermined value of the acceptable frequency percentage of undesirable variants, the in silico analysis step and the substitution step are repeated to produce a new gene construct; and the method includes a step of performing in vivo analysis of one or more genes expressed by the new gene construct, which includes a step of performing RNA sequencing analysis of the RNA product transcribed from the new gene construct, wherein the frequency percentage of undesirable variants is determined at least in part by a step of using a splice recognition aligner from RNA sequencing analysis.

[0014] Other aspects and advantages of this technology will become apparent from the following detailed description, along with the accompanying drawings illustrating the principle of this technology merely as examples. [Brief explanation of the drawing]

[0015] The teachings claimed and / or described herein are further described with respect to exemplary embodiments. These exemplary embodiments are described in detail with reference to the drawings. These embodiments are non-limiting exemplary embodiments, and like reference numerals represent like structures throughout several of the figures of the drawings.

[0016] [Figure 1] Shows an overview of one possibility for creating variants during cell therapy manufacturing.

[0017] [Figure 2] Shows one embodiment of a variant prediction, detection, and removal process.

[0018] [Figure 3] Shows an example of a Sequence Diverger matrix that can be used in the disclosed process.

[0019] [Figure 4] Shows an exemplary process of a Repeat Remover tool that can be used in the disclosed process.

[0020] [Figure 5] Shows an example of a screenshot showing a variant report template that can be used in the disclosed process.

[0021] [Figure 6] Shows an example of a screenshot showing the output from a Repeat Finder and Visualizer tool that identifies the same sequence.

[0022] [Figure 7A] It is a flowchart showing an exemplary process for detecting and replacing sequences that can cause undesirable variants in a gene construct according to one embodiment of the present disclosure.

[0023] [Figure 7B]A flowchart showing an exemplary process for producing a gene therapy product for use in cell therapy according to one embodiment of the present disclosure.

[0024] [Figure 8A] An exemplary code portion that can be used to implement the disclosed method is shown. [Figure 8B-1] An exemplary code portion that can be used to implement the disclosed method is shown. [Figure 8B-2] An exemplary code portion that can be used to implement the disclosed method is shown. [Figure 8C] An exemplary code portion that can be used to implement the disclosed method is shown. **DETAILED DESCRIPTION OF THE INVENTION**

[0025] The present disclosure addresses the need for improved systems and methods for identifying variants in gene constructs and then selecting gene constructs for use in cell therapy. The following disclosure describes a systematic method for the detection and identification of variants and potential variant-causing sequences in gene products used in cell therapy applications. When a variant is identified, risk mitigation strategies such as characterizing the variant and / or performing DNA sequence modifications to remove the detected variant can be used in an iterative process.

[0026] It will be understood that the description herein is merely exemplary and explanatory and is not intended to limit the claimed technology. In this application, the use of the singular includes the plural unless specifically stated otherwise.

[0027] All documents or portions of documents cited in this application, including but not limited to patents, patent applications, articles, books, and papers, are hereby expressly incorporated by reference in their entirety for any purpose. When used in accordance with the present disclosure, the following abbreviations are to be understood to have the following meanings unless otherwise indicated.

[0028] As used herein and in the appended claims, the singular forms "a," "an," and "the" refer to multiple subjects unless the context clearly indicates otherwise.

[0029] As used herein, unless otherwise specified or evident from the context, the term “or” is understood to be inclusive and encompasses both “or” and “and.”

[0030] As used herein, the term "and / or" should be interpreted as a specific disclosure of each of two designated features or components, with or without the other. Accordingly, as used herein in phrases such as "A and / or B," the term "and / or" is intended to include A and B, A or B; A (alone); and B (alone). Similarly, as used in phrases such as "A, B, and / or C," the term "and / or" is intended to include each of the following embodiments: A, B, and C; A, B, or C; A or C; A or B; B or C; A and C; A and B; B and C; A (alone); B (alone); and C (alone).

[0031] As used herein, the terms “for example” and “that is” are used merely as examples and are not intended to be limiting, and should not be construed as referring only to items explicitly listed herein.

[0032] Terms such as "greater than," "at least," and "greater than," for example, "at least one," are limiting. It is not something that will be done, but at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 6 4, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 10 6, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139 It is understood to include 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, or 150, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 2000, 3000, 4000, 5000, or any value greater than the listed value. Any larger number or fraction in between is also included.

[0033] Conversely, the term "less than or equal to" includes each value smaller than the listed value. For example, "100 or fewer nucleotides" includes 100, 99, 98, 97, 96, 95, 94, 93, 92, 91, 90, 89, 88, 87, 86, 85, 84, 83, 82, 81, 80, 79, 78, 77, 76, 75, 74, 73, 72, 71, 70, 69, 68, 67, 66, 65, 64, 63, 62, 61, 60, 59, 58, 57, 56, 55, 54, 5 This includes 3, 52, 51, 50, 49, 48, 47, 46, 45, 44, 43, 42, 41, 40, 39, 38, 37, 36, 35, 34, 33, 32, 31, 30, 29, 28, 27, 26, 25, 24, 23, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1, and 0 nucleotides. Any fewer number or fraction in between is also included.

[0034] Terms such as "multiple," "at least two," "two or more," and "at least the second" are not limiting, but include at least 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63 ,64,65,66,67,68,69,70,71,72,73,74,75,76,77,78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96,97,98,99,100,101,102,103,104,10 5, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 1 It is understood to include 38, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, or 150, 200, 300, 400, 500, 600, 700, 800, 900, 1000, 2000, 3000, 4000, 5000, or more. Any larger number or fraction in between is also included.

[0035] As used herein, unless otherwise specified or evident from the context, the term “about” means a value or composition within the acceptable margin of error of a particular value or composition as determined by those skilled in the art, which depends to some extent on how such value or composition is measured or determined, i.e., on the limitations of the measuring system. For example, “about” or “approximately” means “about” or “approximately” in the art. In accordance with convention, this may mean within one or two standard deviations or more. "Approximately" or "about" may mean a range of up to 10% (i.e., ±10%). Thus, "approximately" may be understood as being 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.5%, 0.1%, 0.05%, 0.01%, or 0.001% greater or less than the explicitly stated value. For example, approximately 5 mg may encompass any amount between 4.5 mg and 5.5 mg. Furthermore, particularly with respect to biological systems or processes, this term may mean up to an order of magnitude off or up to five times the value. Where a particular value or composition is presented in this disclosure, unless otherwise specified, the meaning of "approximately" or "about" should be assumed to be within the acceptable margin of error for that particular value or composition.

[0036] As described herein, any range of concentration, percentage, ratio, or integer should be understood to include any integer values ​​within the listed ranges, and, where appropriate, fractions thereof (such as one-tenth and one-hundredth of an integer), unless otherwise specified.

[0037] The units, prefixes, and symbols used herein are presented in the format accepted by the Systeme International de Unites (SI). Numerical ranges include the number that defines the range.

[0038] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by those skilled in the art to whom this disclosure relates. For example, Juo, "The Concise Dictionary of Biomedicine and Molecular Biology," 2nd ed., (2001), CRC Academic Press; "The Dictionary of Cell & Molecular Biology," 5th ed., (2013); and "The Oxford Dictionary of Biochemistry and Molecular Biology," Cammack et al. eds., 2nd ed., (2006), Oxford University Press provide those skilled in the art with many of the terms used in this disclosure.

[0039] "Administer" refers to the physical delivery of a drug to a target using any of the various methods and delivery systems known to those skilled in the art. Exemplary routes of administration for the formulations disclosed herein include, for example, intravenous, intramuscular, subcutaneous, intraperitoneal, spinal, or other parenteral routes of administration by injection or infusion. Exemplary routes of administration for the compositions disclosed herein include, for example, intravenous, intramuscular, subcutaneous, intraperitoneal, spinal, or other parenteral routes of administration by injection or infusion. As used herein, the term "parenteral administration" generally means, but is not limited to, methods of administration other than enteral and topical administration by injection, including, intravenous, intramuscular, intra-arterial, subarachnoid, intralymphatic, intrafocal, intracapsular, intraorbital, intracardiac, intradermal, intraperitoneal, transtracheal, subcutaneous, subepidermal, intra-articular, subcapsular, subarachnoid, intraspinal, epidural, and intrasternal injections and infusions, as well as in vivo electroporation. In some embodiments, the formulation is administered via a non-intravenous route, for example, orally. Other non-parenteral routes include topical, epidermal, or mucosal administration routes, such as intranasal, intravaginal, rectal, sublingual, or topical. Administration may also be performed, for example, once, multiple times, and / or over one or more extended periods. In one embodiment, CAR T cell therapy is administered by an "infusion product" containing CAR T cells.

[0040] The term "antibody" (Ab) includes, but is not limited to, glycoprotein immunoglobulins that specifically bind to antigens. Generally, an antibody may comprise at least two heavy (H) chains and two light (L) chains linked together by disulfide bonds, or an antigen-binding molecule thereof. Each H chain comprises a heavy chain variable region (abbreviated herein as VH) and a heavy chain constant region. The heavy chain constant region includes three constant domains, CH1, CH2, and CH3. Each light chain includes a light chain variable region (abbreviated herein as VL) and a light chain constant region. The light chain constant region includes one constant domain CL. The VH and VL regions may be further subdivided into hypervariable regions called “complementarity-determining regions” (CDRs), with more conserved regions called “framework regions” (FRs) interposed between them. Each VH and VL includes three CDRs and four FRs arranged in the order FR1, CDR1, FR2, CDR2, FR3, CDR3, and FR4 from the amino terminus to the carboxy terminus. The variable regions of the heavy and light chains include binding domains that interact with antigens. The constant region of Ab may mediate the binding of immunoglobulins to host tissues or factors, including various cells of the immune system (e.g., effector cells) and the first component of the classical complement system (C1q).

[0041] Antibodies include, for example, monoclonal antibodies, recombinant antibodies, monospecific antibodies, polyspecific antibodies (including bispecific antibodies), human antibodies, modified antibodies, humanized antibodies, chimeric antibodies, immunoglobulins, synthetic antibodies, tetrameric antibodies containing two heavy chains and two light chain molecules, antibody light chain monomers, antibody heavy chain monomers, antibody light chain dimers, antibody heavy chain dimers, antibody light chain-antibody heavy chain pairs, intracellular antibodies, antibody fusions (sometimes referred to as "antibody conjugates" in this specification), heteroconjugate antibodies, single-domain antibodies, monovalent antibodies, single-chain antibodies or single-chain Fv (single-chain Fv, scFv), camelized antibodies, affibodies, Fab Examples include fragments, F(ab')2 fragments, disulfide-bound Fv(sdFv), anti-idiotype (anti-Id) antibodies (e.g., anti-anti-Id antibodies), minibodies, domain antibodies, synthetic antibodies (sometimes referred to herein as “antibody mimetic”), and any of the antigen-binding fragments described above. In some embodiments, the antibodies described herein refer to a population of polyclonal antibodies.

[0042] The terms “antigen-binding molecule,” “antigen-binding portion,” or “antibody fragment” refer to any molecule containing the antigen-binding portion (e.g., CDR) of an antibody from which the molecule originates. An antigen-binding molecule may contain an antigen complementarity-determining region (CDR). Examples of antibody fragments include, but are not limited to, Fab, Fab', F(ab')2, and Fv fragments, dAb, linear antibodies, scFv antibodies, and multispecific antibodies formed from antigen-binding molecules. Peptibodies (i.e., Fc fusion molecules containing peptide-binding domains) are another example of a suitable antigen-binding molecule. In some embodiments, the antigen-binding molecule binds to an antigen on tumor cells. In some embodiments, the antigen-binding molecule binds to an antigen on cells involved in hyperproliferative diseases, or to a viral or bacterial antigen. In some embodiments, the antigen-binding molecule binds to CD19. In further embodiments, the antigen-binding molecule is an antibody fragment that specifically binds to an antigen and contains one or more of its complementarity-determining regions (CDRs). In further embodiments, the antigen-binding molecule is a single-chain variable fragment (scFv). In some embodiments, the antigen-binding molecule includes or consists of an avimer.

[0043] An "antigen" refers to any molecule that can trigger an immune response or be bound by an antibody or antigen-binding molecule. An immune response may involve either antibody production or activation of specific immune cells, or both. Those skilled in the art will readily understand that virtually all proteins or peptides, and any macromolecule, can function as an antigen. Antigens may be expressed endogenously, i.e., by genomic DNA, or by recombination. Antigens may be specific to certain tissues, such as cancer cells, or they may be expressed broadly. Furthermore, larger molecular fragments may act as antigens. In some embodiments, the antigen is a tumor antigen.

[0044] The term "neutralize" refers to an antigen-binding molecule, scFv, antibody, or fragment thereof that binds to a ligand and prevents or reduces the biological effect of that ligand. In some embodiments, antigen-binding molecules, scFvs, antibodies, or fragments thereof directly block the binding site on a ligand, or otherwise alter the ligand's ability to bind through indirect means (such as structural or energy changes within the ligand). In some embodiments, antigen-binding molecules, scFvs, antibodies, or fragments thereof prevent the protein to which they are bound from performing its biological function.

[0045] The term "self" refers to any material derived from the same individual that is later reintroduced. For example, the genetically engineered autologous cell therapy (eACT®) method described herein involves collecting lymphocytes from a patient, which are then engineered to express, for example, a CAR construct, and then administered to the same patient.

[0046] The term "allogeneic" refers to any material that originates from one individual and is then introduced into another individual of the same species, such as allogeneic T cell transplantation.

[0047] The terms "transduction" and "transduced" refer to the process by which foreign DNA is introduced into cells via a viral vector (see Jones et al., "Genetics: principles and analysis," Boston: Jones & Bartlett Publ. (1998)). In some embodiments, the vector is a retroviral vector, DNA vector, RNA vector, adenovirus vector, baculovirus vector, Epstein-Barr virus vector, papovavirus vector, vaccinia virus vector, herpes simplex virus vector, adenovirus-associated vector, lentiviral vector, or any combination thereof.

[0048] "Cancer" refers to a broad group of diseases characterized by the uncontrolled proliferation of abnormal cells in the body. Uncontrolled cell division and proliferation can lead to the formation of malignant tumors, which may invade adjacent tissues and metastasize to distal parts of the body via the lymphatic system or bloodstream. "Cancer" or "cancer tissue" may encompass tumors. In this application, the term cancer is synonymous with malignant tumor. Examples of cancers that may be treated by the methods disclosed herein include, but are not limited to, cancers of the immune system, including lymphoma, leukemia, myeloma, and other leukocyte malignancies. In some embodiments, the methods disclosed herein include, for example, bone cancer, pancreatic cancer, skin cancer, head or neck cancer, skin or intraocular malignant melanoma, uterine cancer, ovarian cancer, rectal cancer, anal cancer, gastric cancer, testicular cancer, uterine cancer, fallopian tube cancer, endometrial cancer, cervical cancer, vaginal cancer, vulvar cancer, [add other solid tumors] multiple myeloma, Hodgkin's disease, non-Hodgkin lymphoma (NHL), mediastinal large B-cell lymphoma (PMBC), diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), transformed follicular lymphoma, splenic marginal zone lymphoma (SMZL), esophageal cancer, small intestine cancer, endocrine system cancer, thyroid cancer, parathyroid cancer, adrenal cancer It can be used to reduce the tumor size of tumors resulting from combinations of the above cancers, including soft tissue sarcomas, cancers of the urethra, cancers of the penis, chronic or acute leukemia, acute myeloid leukemia, chronic myeloid leukemia, acute lymphoblastic leukemia (ALL) (including non-T-cell ALL), chronic lymphocytic leukemia (CLL), solid tumors of childhood, lymphocytic lymphoma, cancers of the bladder, cancers of the kidney or ureter, renal pelvis cancer, central nervous system (CNS) tumors, primary CNS lymphomas, tumor angiogenesis, spinal axis tumors, brainstem gliomas, pituitary adenomas, Kaposi's sarcoma, epidermoid carcinoma, squamous cell carcinoma, T-cell lymphoma, environmentally induced cancers including those induced by asbestos, other B-cell malignancies, and tumors resulting from combinations of the above cancers. In some embodiments, the cancer is multiple myeloma. In some embodiments, the cancer is NHL. Certain cancers may be responsive to chemotherapy or radiotherapy, or they may be resistant to treatment.Treatment-resistant cancer refers to cancer that is not suitable for surgical intervention, either because it is unresponsive to chemotherapy or radiotherapy from the outset, or because it becomes unresponsive over time.

[0049] As used herein, “antitumor effect” refers to a biological effect that may manifest as a reduction in tumor volume, a decrease in the number of tumor cells, a decrease in tumor cell proliferation, a decrease in the number of metastases, an increase in overall survival or progression-free survival, an extension of life expectancy, or improvement of various physiological symptoms associated with tumor. The antitumor effect may also refer to the prevention of tumor development, for example, by a vaccine.

[0050] As used herein, “cytokine” refers to a non-antibody protein released by a cell in response to contact with a specific antigen, where the cytokine interacts with another cell to mediate a response in that other cell. As used herein, “cytokine” is intended to refer to a protein released by a population of cells that acts on another cell as an intercellular mediator. Cytokines may be endogenously expressed by cells or administered to a subject. Cytokines may be released by immune cells such as macrophages, B cells, T cells, and mast cells to propagate an immune response. Cytokines can induce a variety of responses in recipient cells. Cytokines may include homeostatic cytokines, chemokines, pro-inflammatory cytokines, effectors, and acute-phase proteins. Homeostatic cytokines, such as interleukin (IL) 7 and IL-15, promote the survival and proliferation of immune cells, while pro-inflammatory cytokines can promote inflammatory responses. Examples of homeostatic cytokines include, but are not limited to, IL-2, IL-4, IL-5, IL-7, IL-10, IL-12p40, IL-12p70, IL-15, and interferon (IFN) gamma. Examples of pro-inflammatory cytokines include IL-1a, IL-1b, IL-6, IL-13, IL-17a, tumor necrosis factor (TNF)-alpha, TNF-beta, Fibroblast growth factor (FGF) 2, granulocyte macrophage colony-stimulating factor (GM-CSF), soluble intercellular adhesion molecule 1 (sICAM) -1), soluble vascular adhesion molecule 1 (sVCAM-1), vascular endothelial growth factor (VEGF), VEGF-C, VEGF-D, and placental growth factor (PLGF). However, it is not limited to these. Examples of effectors include, but are not limited to, granzyme A, granzyme B, soluble Fas ligand (sFasL), and perforin. Examples of acute-phase proteins include C-reactive protein (CRP) and serum amyloid A (S Examples include, but are not limited to, AA.

[0051] Chemokines are a type of cytokine that mediates chemotaxis or directional movement of cells. Examples of chemokines include, but are not limited to, IL-8, IL-16, eotaxin, eotaxin-3, macrophage-derived chemokines (MDC or CCL22), monocyte chemotactic protein 1 (MCP-1 or CCL2), MCP-4, macrophage inflammatory protein 1α (MIP-1α, MIP-1a), MIP-1β (MIP-1b), gamma-inducible protein 10 (IP-10), and thymic and activation-regulating chemokines (TARC or CCL17).

[0052] As used herein, “chimeric receptor” refers to a genetically engineered molecule expressed on its surface that is capable of recognizing a specific molecule. Chimeric antigen receptors (CARs) and genetically engineered T cell receptors (TCRs) containing a binding domain capable of interacting with a specific tumor antigen enable T cells to target and kill cancer cells expressing a specific tumor antigen. In one embodiment, T cell therapy is based on T cells engineered to express a chimeric antigen receptor (CAR) or T cell receptor (TCR) comprising (i) an antigen-binding molecule, (ii) a costimulatory domain, and (iii) an activating domain. The costimulatory domain may include an extracellular domain, a transmembrane domain, and an intracellular domain, wherein the extracellular domain may include a hinge domain that can be shortened.

[0053] The “therapeutic effective dose,” “effective dose,” “effective amount,” or “therapeutic effective dose” of a therapeutic agent, such as engineered CAR T cells, is any amount that, when used alone or in combination with another therapeutic agent, protects a subject from the onset of the disease or promotes disease regression, as demonstrated by a reduction in the severity of disease symptoms, an increase in the frequency and duration of disease-free periods, or the prevention of disability or impairment resulting from the disease. Such terms can be used interchangeably. The ability of a therapeutic agent to promote disease regression can be evaluated using various methods known to those skilled in the art, for example, in human subjects during clinical trials, in animal model systems to predict efficacy in humans, or by analyzing the activity of the agent in in vitro assays.

[0054] As used herein, the term “lymphocyte” encompasses natural killer (NK) cells, T cells, and B cells. NK cells are a type of cytotoxic lymphocyte that are the main component of the innate immune system. NK cells reject tumor and virus-infected cells. They act through the process of apoptosis or programmed cell death. They are called “natural killers” because they do not require activation to kill cells. T cells play a major role in cell-mediated immunity (immunity without antibody involvement). Their T cell receptors (TCRs) differentiate from other lymphocyte types. The thymus, a differentiated organ of the immune system, is primarily responsible for the maturation of T cells. There are six types of T cells, namely helper T cells (e.g., CD4+ cells), cytotoxic T cells (TC, cytotoxic T lymphocyte, CTL, T killer cell, cytolytic T cell, CD8+ T cell, or also known as killer T cell), and memory T cells ((i) Stem memory TSCM cells, like naive cells, are CD45RO-, CCR7+, CD45RA+, CD62L+ (L-selectin), CD27+, CD28+, and IL-7Rα+, but also express large amounts of CD95, IL-2Rβ, CXCR3, and LFA-1, and memory cells Cells exhibit numerous functional characteristics specific to each cell type; (ii) Central memory TCM cells express L-selectin and CCR7 and secrete IL-2 but not IFNγ or IL-4; however, (iii) Effector memory TEM cells do not express L-selectin or CCR7 but produce effector cytokines such as IFNγ and IL-4), regulatory T cells (Treg, suppressor T cells, or CD4+CD25+ regulatory T cells), natural killer T cells (NKT), and gamma delta T cells exist. B cells, on the other hand, play a major role in humoral immunity (involving antibodies). B cells produce antibodies and antigens, act as antigen-presenting cells (APCs), and transform into memory B cells after activation by antigen interaction. In mammals, immature B cells are formed in the bone marrow, hence the name.

[0055] The terms “genetically modified” or “modified” refer to methods of altering a cell’s genome, including, but not limited to, deleting coding regions or non-coding regions or parts thereof, or inserting coding regions or parts thereof. In some embodiments, the modified cells are lymphocytes, such as T cells, which may be obtained from either a patient or a donor. The cells may be modified to express exogenous constructs, such as chimeric antigen receptors (CARs) or T cell receptors (TCRs), which are then incorporated into the cell’s genome.

[0056] "Immune response" refers to the action of immune system cells (e.g., T lymphocytes, B lymphocytes, natural killer (NK) cells, macrophages, eosinophils, mast cells, dendritic cells, and neutrophils) and soluble macromolecules (including Ab, cytokines, and complement) produced by either these cells or the liver, resulting in the selective targeting, binding, damage, destruction, and / or elimination from the body of a vertebrate of invading pathogens, pathogen-infected cells or tissues, cancerous or other abnormal cells, or, in the case of autoimmune or pathological inflammation, normal human cells or tissues.

[0057] The term “immunotherapy” refers to the treatment of a person who is suffering from a disease or at risk of suffering from or relapsing from a disease, by means of methods including inducing, enhancing, suppressing, or otherwise modifying the immune response. Examples of immunotherapy include, but are not limited to, T-cell therapy. Examples of T-cell therapy include adoptive T-cell therapy, tumor-infiltrating lymphocyte (TIL) immunotherapy, autologous cell therapy, genetically engineered autologous cell therapy (eACT®), and allogeneic T-cell transplantation. However, those skilled in the art will understand that the pre-treatment methods disclosed herein enhance the efficacy of any transplanted T-cell therapy. Examples of T-cell therapy are described in U.S. Patent Applications Publications 2014 / 0154228 and 2002 / 0006409, U.S. Patents 7,741,465, 6,319,494, and 5,728,388, and International Publication 2008 / 081035. In some embodiments, immunotherapy includes CAR T-cell therapy. In some embodiments, the CAR T-cell therapy product is administered by infusion.

[0058] T cells for immunotherapy may be derived from any source known in the art. For example, T cells can be differentiated in vitro from a hematopoietic stem cell population, or T cells can be obtained from a subject. T cells can be obtained, for example, from peripheral blood mononuclear cells (PBMCs), bone marrow, lymph node tissue, umbilical cord blood, thymic tissue, tissue from an infection site, ascites, pleural fluid, splenic tissue, and tumors. Furthermore, T cells may be derived from one or more T cell lines available in the art. T cells can also be obtained from blood units taken from a subject using various techniques known to those skilled in the art, such as FICOLL® isolation and / or apheresis. Further methods for isolating T cells for T cell therapy are disclosed in U.S. Patent Application Publication 2013 / 0287748, which is incorporated herein by reference in its entirety.

[0059] The term "genetically modified autologous cell therapy" or "eACT®," also known as adoptive cell transfer, is a process in which a patient's own T cells are collected and then genetically modified to recognize and target one or more antigens expressed on the surface of one or more specific tumor cells or malignant tumor cells. T cells can be engineered, for example, to express chimeric antigen receptors (CARs). CAR-positive (+) T cells are engineered to express extracellular single-stranded variable fragments (scFv) specific to a particular tumor antigen, linked to an intracellular signaling region containing at least one costimulatory domain and at least one activating domain. CAR scFv can be designed to target, for example, cells of the B cell lineage, including all normal B cells, and, but not limited to, B cell malignancies, including diffuse large B-cell lymphoma (DLBCL) nonspecific type, primary mediastinal large B-cell lymphoma, high-grade B-cell lymphoma, and DLBCL arising from follicular lymphoma, NHL, CLL, and non-T cell ALL. Exemplary CAR T-cell therapies and constructs are described in U.S. Patent Applications Publications 2013 / 0287748, 2014 / 0227237, 2014 / 0099309, and 2014 / 0050708, which are incorporated in their entirety by reference.

[0060] As used herein, “Patient” includes any human being suffering from cancer (e.g., lymphoma or leukemia). The terms “Subject” and “Patient” are used interchangeably herein.

[0061] As used herein, the term “in vitro cells” refers to any cells cultured ex vivo. In particular, in vitro cells may include T cells. The term “in vivo” means within a patient.

[0062] The terms "peptide," "polypeptide," and "protein" are used interchangeably. A peptide refers to a compound composed of amino acid residues covalently linked by peptide bonds. A protein or peptide contains at least two amino acids, and there is no limit to the maximum number of amino acids that can constitute a protein or peptide sequence. A polypeptide includes any peptide or protein containing two or more amino acids linked to each other by peptide bonds. As used herein, this term refers to both short chains, also commonly called peptides, oligopeptides, and oligomers in the art, and long chains, of which there are many types, commonly called proteins in the art. "Polypeptides" include, for example, biologically active fragments, substantially homologous polypeptides, oligopeptides, homodimers, heterodimers, polypeptide variants, modified polypeptides, derivatives, analogs, and fusion proteins. Polypeptides include natural peptides, recombinant peptides, synthetic peptides, or combinations thereof.

[0063] As used herein, “stimulus” refers to a primary response induced by the binding of a stimulating molecule to its homologous ligand, where the binding mediates a signaling event. “Stimulating molecule” is a molecule on a T cell, such as a T cell receptor (TCR) / CD3 complex that specifically binds to a homologous stimulating ligand present on an antigen-presenting cell. “Stimulating ligand” is a ligand that, when present on an antigen-presenting cell (e.g., APCs, dendritic cells, B cells), specifically binds to a stimulating molecule on a T cell, thereby mediating a primary response by the T cell, including, but not limited to, activation, initiation of an immune response, and proliferation. Examples of stimulating ligands include, but are not limited to, anti-CD3 antibodies, peptide-loaded MHC class I molecules, superagonist anti-CD2 antibodies, and superagonist anti-CD28 antibodies.

[0064] As used herein, “co-stimulatory signal” refers to a signal that, in combination with a primary signal such as TCR / CD3 ligation, elicits an upregulation or downregulation of a T cell response, such as, but not limited to, proliferation and / or major molecules.

[0065] As used herein, “costimulatory ligand” encompasses molecules on antigen-presenting cells that specifically bind to homologous costimulatory molecules on T cells. The binding of costimulatory ligands provides signals that mediate T cell responses, such as proliferation, activation, and differentiation, but is not limited to these. In addition to the primary signals provided by stimulatory molecules, costimulatory ligands induce signals, for example, by binding the T cell receptor (TCR) / CD3 complex to peptide-loaded major histocompatibility complex (MHC) molecules. Co-stimulatory ligands may include, but are not limited to, 3 / TR6, 4-1BB ligand, agonists or antibodies that bind to Toll ligand receptors, B7-1 (CD80), B7-2 (CD86), CD30 ligand, CD40, CD7, CD70, CD83, herpesvirus entry mediator (HVEM), human leukocyte antigen G (HLA-G), ILT4, immunoglobulin-like transcript (ILT)3, inducible costimulatory ligand (ICOS-L), intercellular adhesion molecule (ICAM), ligands that specifically bind to B7-H3, lymphotoxin β receptor, MHC class I chain-related protein A (MICA), MHC class I chain-related protein B (MICB), OX40 ligand, PD-L2, or programmed cell death (PD)L1. In certain embodiments, the co-stimulatory ligand may include, but are not limited to, a co-stimulatory molecule present on T cells, such as ligands that specifically bind to, but are not limited to, 4-1BB, B7-H3, CD2, CD27, CD28, CD30, CD40, CD7, ICOS, CD83, lymphocyte function-associated antigen-1 (LFA-1), natural killer cell receptor C (NKG2C), OX40, PD-1, or an antibody that specifically binds to tumor necrosis factor superfamily member 14 (TNFSF14 or LIGHT).

[0066] "Co-stimulatory molecules" specifically bind to costimulatory ligands, thereby stimulating T cell response. Violent responses, for example, are progeny-binding partners on T cells that mediate proliferation. Co-stimulatory molecules include, but are not limited to, 4-1BB / CD137, B7-H3, BAFFR, BLAME (SLAMF8), BTLA, CD33, CD45, CD100 (SEMA4D), CD103, CD134, CD137, CD154, CD16, CD160 (BY55), CD18, CD19, CD19a, CD2, CD22, CD247, CD27, CD276 (B7-H3), CD28, C D29, CD3 (alpha, beta, delta, epsilon, gamma, zeta), CD30, CD37, CD4, CD4, CD40, CD49a, CD49D, CD49f, CD5, CD64, CD69, CD7, CD80, CD83 ligand, CD84, CD86, CD8α, CD8β, CD9, CD96 (Tactile), CD11a, CD11b, CD11c, CD11d, CDDS, CEACAM1, CRT AM, DAP-10, DNAM1 (CD226), Fcγ receptor, GADS, GITR, HVEM (LIGHTR), IA4, ICAM-1, ICOS, Igα (CD79a), IL2Rβ, IL2Rγ, IL7Rα, integrin, ITGA4, ITGA6, ITGAD, ITGAE, ITGAL, ITGAM, ITGAX, ITGB2, ITGB7, ITGB, KIRDS2, LAT, LFA-1, LIGHT (tumor necrosis factor superfamily member 14; TNFSF14), LTBR, ​​Ly9 (CD229), lymphocyte function-associated antigen-1 (LFA-1 (CD11a / CD18), M Examples include HC class I molecules, NKG2C, NKG2D, NKp30, NKp44, NKp46, NKp80 (KLRF1), OX40, PAG / Cbp, PD-1, PSGL1, SELPLG (CD162), signaling lymphocyte activating molecules, SLAM (SLAMF1;CD150;IPO-3), SLAMF4 (CD244;2B4), SLAMF6 (NTB-A, Ly108), SLAMF7, SLP-76, TNF, TNFr, TNFR2, Toll ligand receptors, TRANCE / RANKL, VLA1, or VLA-6, or fragments, cleaved forms, or combinations thereof.

[0067] The terms “reduce” and “decrease” are used interchangeably herein and mean any change that becomes less than the original. “Reduce” and “decrease” are relative terms and require a comparison between before and after measurements. “Reduce” and “decrease” include complete depletion. Similarly, the term “increase” means any change that becomes higher than the original value. “Increase,” “higher,” and “lower” are relative terms and require a comparison between before and after measurements and / or between reference standards. In some embodiments, the baseline is taken from a general population which may be the general population of patients. In some embodiments, the baseline is derived from an quartile analysis of the general patient population.

[0068] "Treatment" or "to treat" a subject means any type of intervention or process performed on the subject, or administration of an active agent to the subject, for the purpose of reducing, reducing, improving, inhibiting, delaying, or preventing the onset, progression, advancement, severity, or relapse of symptoms, complications, conditions, or biochemical signs associated with the disease. In some embodiments, "treatment" or "to treat" includes partial remission. In other embodiments, "treatment" or "to treat" includes complete remission.

[0069] As used herein, the term “multifunctional T cell” refers to a cell that co-secretes at least two proteins from a pre-specified panel for each cell, in combination with the amount of each protein produced (i.e., the combination of the number of secreted proteins and their intensities). In some embodiments, a single-cell functional profile is determined for each evaluable population of engineered T cells. The profile includes effector (granzyme B, IFN-γ, MIP-1α, perforin, TNF-α, TNF-β), stimulant (GM-CSF, IL-2, IL-5, IL-7, IL-8, IL-9, IL-12, IL-15, IL-21), regulatory (IL-4, IL-10, IL-13, IL-22, TGF-beta 1, sCD137, sCD40L), and chemoattractant (CCL-11, IP-10, MIP Cells can be classified into groups of IL-1β (IL-1β, RANTES) and inflammatory (IL-1β, IL-6, IL-17A, IL-17F, MCP-1, MCP-4). In some embodiments, the functional profile of each cell allows for the calculation of other metrics, including degradation of each sample by cellular polyfunctionality (i.e., which percentage of cells secrete multiple cytokines compared to non-secreting or monofunctional cells) and degradation of the sample by functional groups (i.e., which monofunctional and polyfunctional groups are secreted by cells in the sample and their frequencies).

[0070] As used herein, the terms “quartile” or “quadrant” are statistical terms that describe the division of observations into four defined intervals based on the values ​​of the data and how they compare to the entire set of observations.

[0071] As used herein, the term "Day 0 of the study" is defined as the day on which a subject first receives CAR T cell infusion. The day before Day 0 of the study is Day 1 of the study. Any day after registration and prior to Day 1 of the study is continuous and a negative integer.

[0072] As used herein, the term “objective response” refers to complete response (CR), partial response (PR), or no response. The criteria are based on the revised IWG response criteria for malignant lymphoma.

[0073] As used herein, the term “complete response” refers to the complete recovery of the disease, which becomes undetectable by radioimaging and clinical laboratory evaluation. There are no signs of cancer at a given time.

[0074] As used herein, the term “partial response” refers to a reduction of more than 30% of the tumor without complete recovery. The criteria are based on the revised IWG response criteria for malignant lymphoma, and PR is defined as “a reduction of at least 50% of the sum of the product of diameters (SPD) of up to six main nodules or nodular masses.” These nodes or masses should be selected according to all of the following, and they should be clearly measurable in at least two vertical dimensions. Where possible, they should be from different regions of the body; whenever these sites are involved, they should include diseased areas of the mediastinum and retroperitoneum.

[0075] As used herein, the term "non-response" refers to subjects who have not experienced a complete response (CR) or partial response (PR) after CAR T cell infusion.

[0076] As used herein, the term “sustained response” refers to subjects who maintained a sustained response up to at least one year of follow-up after CAR T cell infusion. Since f / u is no longer available for this cohort, 6-month f / u is used only for Z1 and C3. Nevertheless, the conclusions remain the same.

[0077] As used herein, the term “relapse” refers to a patient who has achieved a complete response (CR) or a partial response (PR) and subsequently experienced disease progression.

[0078] When used herein, the proliferation and persistence of peripheral blood CAR T cells can be monitored by qPCR analysis using CAR-specific primers targeting the scFv portion of the CAR (e.g., the heavy chain of the CD19-binding domain) and its hinge / CD28 transmembrane domain. Alternatively, this can be measured by counting the CAR cell / blood unit volume.

[0079] As used herein, the schedule for blood collection for CAR T cells is as follows: before CAR T cell infusion, day 7, week 2 (day 14), week 4 (day 28), month 3 (day 90), month 6 (day 180), month 12 (day 360), and month 24 (day 720). It is possible.

[0080] As used herein, “CAR T cell peak” is defined as the maximum absolute number of CAR+PBMCs / μL in serum achieved after day 0.

[0081] When used in this specification, "CAR T cell peak time" refers to the time from day 0 to CAR This is defined as the number of days until the peak T cell count is reached.

[0082] As used herein, "Area Under Curve (AUC) of CAR T cell levels from day 0 to day 28" is defined as the area under the curve on a plot of CAR T cell levels against scheduled visits from day 0 to day 28. This AUC represents the total level of CAR T cells over a long period of time.

[0083] When used herein, the schedule for blood collection for cytokines is the day before or the day of conditioning chemotherapy (days 1-5), day 0, day 1, day 3, day 5, day 7, every other day during hospitalization if hospitalized, week 2 (day 14), and week 4 (day 28).

[0084] As used herein, the "baseline" of cytokines is defined as the last value measured before conditioning chemotherapy.

[0085] When used in this specification, the change in magnification from baseline on day X is:

number

[0086] As used herein, “post-baseline cytokine peak” is defined as the maximum level of serum cytokines achieved from baseline (-5 days) to day 28.

[0087] As used herein, the "time to cytokine peak" after CAR T cell injection is defined as the number of days from day 0 to the day on which the cytokine peak is achieved.

[0088] As used herein, the “Area Under Curve (AUC)” of cytokine levels from day -5 to day 28 is defined as the area under the curve on the plot of cytokine levels against scheduled visits from day -5 to day 28. This AUC represents the total level of cytokines over a long period. Assuming that cytokines and CAR+ T cells are measured at specific, separate time points, the trapezoidal rule may be used to estimate the AUC.

[0089] As used herein, the term “negligible effect” and its boundaries and limits will be readily understood by those skilled in the art. As a non-limiting example, those skilled in the art will understand that a negligible effect may mean one or more of the following: an effect that is not statistically significant on the expression of a chimeric antigen receptor; an effect that is not statistically significant on the therapeutic efficacy of a chimeric antigen receptor; an effect that is not statistically significant on the toxicity of a chimeric antigen receptor to a patient; or an effect on the expression and / or efficacy and / or toxicity of a chimeric antigen receptor that does not exceed a predetermined threshold for expression and / or efficacy and / or toxicity of the chimeric antigen receptor.

[0090] It will be understood that chimeric antigen receptors (CAR or CAR-T) are genetically modified receptors, and that T cell receptors (TCRs) may also be genetically modified receptors. These engineered receptors can be readily inserted into immune cells, such as T cells, according to techniques known in the art, and expressed by these immune cells. With CARs, a single receptor can be programmed to both recognize a specific antigen and, upon binding to that antigen, activate immune cells to attack and destroy cells possessing that antigen. If these antigens are present on tumor cells, immune cells expressing CARs can target and kill the tumor cells.

[0091] CARs can bind to antigens (such as cell surface antigens) by incorporating antigen-binding molecules that interact with their target antigen. As used herein, “antigen-binding molecule” means any protein that binds to a particular target molecule. Antigen-binding molecules include, but are not limited to, antibodies and immunologically functional fragments, which are their binding sites. Peptibodies (i.e., Fc fusion molecules containing peptide-binding domains) are another example of suitable antigen-binding molecules.

[0092] Preferably, the target molecule may include, but is not limited to, blood-derived cancer-associated antigens. Non-limiting examples of blood-derived cancer-associated antigens include antigens associated with one or more cancers selected from the group consisting of acute myeloid leukemia (AML), chronic myeloid leukemia (CML), chronic myelomonocytic leukemia (CMML), juvenile myelomonocytic leukemia, atypical chronic myeloid leukemia, acute promyelocytic leukemia (APL), acute monoblastic leukemia, acute erythroblastic leukemia, acute megakaryoblastic leukemia, lymphoblastic leukemia, B-lineage acute lymphoblastic leukemia, B-cell chronic lymphocytic leukemia, B-cell non-Hodgkin lymphoma, myelodysplastic syndrome (MDS), myeloproliferative disorders, myeloid neoplasms, granulocytic sarcomas, and blast-cell plasmacytoid dendritic cell neoplasms (BPDCN).

[0093] In some embodiments, the antigen is a tumor-associated surface antigen, e.g., 5T4, alpha-fetoprotein (AFP), B7-1 (CD80), B7-2 (CD86), BCMA, β-human chorionic gonadotropin, CA-125, carcinoembryonic antigen (CEA), CD123, CD133, CD138, CD19, CD20, CD22, CD23, CD24, CD25, CD30, CD33, CD34, CD4, CD40, CD 44, CD56, CD8, CLL-1, c-Met, CMV-specific antigen, CSPG4, CTLA-4, disialoganglioside GD2, ductal epithelial mucin, EBV-specific antigen, EGFR variant III (EGFRvIII), ELF2M, endoglin, ephrin B2, epidermal growth factor receptor (EGFR), epithelial cell adhesion molecule (EpCAM), epithelial tumor antigen, ErbB2 (HER2 / neu), fibroblast-related protein (fap), FLT3, folic acid Binding proteins, GD2, GD3, glioma-associated antigens, sphingoglycolipids, gp36, HBV-specific antigens, HCV-specific antigens, HER1-HER2, HER2-HER3 combinations, HERV-K, high molecular weight melanoma-associated antigens (HMW-MAA), HIV-1 envelope glycoprotein gp41, HPV-specific antigens, human telomerase reverse transcriptase, IGF-1 receptor, IGF-II, IL-11Rα, IL-13R-α2, influenza virus-specific antigens; CD38, insulin growth factor-1 (IGF1), intestinal carboxylesterase, κ chain, LAGA-1α, λ chain, Lassa virus-specific antigens, lectin-reactive AFP, lineage-specific antigens or tissue-specific antigens, e.g., CD3, MAGE, MAGE-A1, major histocompatibility complex (MHC) molecules, major histocompatibility complex (MHC) molecules presenting tumor-specific peptide epitopes, M-CSF, melanoma-associated antigens, mesothelin, MN-CA IX, MUC-1, variant hsp72, variant p53, variant p53, variant ras, neutrophil elastase, NKG2D, Nkp30, NY-ESO-1, p53, PAP, prostase, prostate-specific antigen (PSA), prostate cancer tumor antigen-1 (PCTA-1), prostate-specific antigen protein, PSMA, RAGE-1, ROR1, RU1, RU2 (AS), surface adhesion molecule, survivor and telomerase, TAG-72, extradomain A (EDA) and extradomain B (EDB) of fibronectin, and Teneiss The following are selected from TnC A1 domain (TnC A1), thyroglobulin, tumor stromal antigen, vascular endothelial growth factor receptor-2 (VEGFR2), virus-specific surface antigens, such as HIV-specific antigen (e.g., HIV gp120), and any derivatives or variants of these surface antigens.

[0094] In some embodiments, the target molecule may include a viral infection-associated antigen. The viral infection in this disclosure may be caused by any virus, including, for example, HIV. This list of possible target molecules is not intended to be exclusive.

[0095] The TCRs of this disclosure may, for example, bind to tumor-associated antigens. As used herein, “tumor-associated antigens” include adrenocortical tumors, anal cancers, bladder cancers, bone cancers, brain cancers, breast cancers, carcinoid cancers, carcinomas, cervical cancers, colon cancers, endometrial cancers, esophageal cancers, extrahepatic bile duct cancers, extracranial germ cell cancers, eye cancers, gallbladder cancers, gastric cancers, germ cell tumors, gestational trophoblastic tumors, head and neck cancers, hypopharyngeal cancers, islet cell carcinomas, kidney cancers, colorectal cancers, laryngeal cancers, leukemias, lip and oral cancers, liver cancers, lung cancers, lymphomas, malignant mesotheliomas, Merkel cell carcinomas, mycosis fungoides, spinal cord dysplasia syndromes, and spinal cord proliferation syndromes. This refers to any antigen associated with one or more cancers selected from the group consisting of sexual dysfunction, nasopharyngeal cancer, neuroblastoma, oral cancer, oropharyngeal cancer, osteosarcoma, ovarian epithelial carcinoma, ovarian germ cell carcinoma, pancreatic cancer, sinus and nasal cavity cancer, parathyroid cancer, penile cancer, pituitary cancer, plasmacytoma, prostate cancer, rhabdomyosarcoma, rectal cancer, renal cell carcinoma, transitional cell carcinoma of the renal pelvis and ureter, salivary gland cancer, Sézary syndrome, skin cancer, small intestine cancer, soft tissue sarcoma, gastric cancer, testicular cancer, parathyroid carcinoma, thyroid cancer, urethral cancer, uterine cancer, vaginal cancer, vulvar cancer, and Wilms' tumor.

[0096] In certain embodiments, the present disclosure may be suitable for target molecules against hematological cancers. In some embodiments, the cancer is a cancer of leukocytes. In other embodiments, the cancer is a cancer of plasma cells. In some embodiments, the cancer is leukemia, lymphoma, or myeloma. In certain embodiments, the cancer is acute lymphoblastic leukemia (ALL) (including non-T cell ALL), acute lymphoblastic leukemia (ALL) and hemophagocytic lymphohistiocytosis (HLH), B-cell prelymphoblastic leukemia, B-cell acute lymphoblastic leukemia ("BALL"), blast plasmacytoid dendritic cell neoplasm, Burkitt lymphoma, chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), chronic myeloma Leukemia malformation (CML), chronic or acute granulomatous disease, chronic or acute leukemia, diffuse large B-cell lymphoma, diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL), hairy cell leukemia, hemophagocytic syndrome (macrophage activation syndrome (MAS)), Hodgkin's disease, large cell granuloma, leukocyte adhesion disorder, malignant lymphoproliferative state, MALT lymphoma, mantle cell lymphoma, marginal Lymphoma, monoclonal gammaglobulinemia of unknown significance (MGUS), multiple myeloma, myelodysplasia and myelodysplastic syndromes (MDS), myeloid diseases (AML), including but not limited to acute myeloid leukemia, non-Hodgkin lymphoma (NHL), plasmacytoproliferative disorders (e.g., solitary myeloma, solitary plasmacytoma, extramedullary plasmacytoma, and multiple myeloma), POEMS syndromes (Crow-Fukase syndrome, Takatsuki disease, PEP syndrome), mediastinal large B-cell lymphoma (PMBCL), small cell or large cell follicular lymphoma, splenic marginal zone lymphoma (SMZL), systemic amyloid light chain amyloidosis, T-cell acute lymphoblastic leukemia (TALL), T-cell lymphoma, transformed follicular lymphoma, Waldenström macroglobulinemia, or combinations thereof.

[0097] The TCRs of this disclosure may also bind to viral infection-associated antigens. Viral infection-associated antigens include, for example, antigens associated with any viral infection, including viral infection caused by HIV.

[0098] Various embodiments of this application are described in further detail in the following subsections. Variant creation

[0099] Cell therapy products, including autologous, allogeneic, neoantigen, and other types of products, may express RNA and protein sequences other than the desired or transfected gene(s). The expression of these non-standard sequences (called variants) can occur through at least two different mechanisms.

[0100] The first mechanism, RNA splicing, can produce a variant-expressing product even if transfected with the desired gene sequence, because the product is transcribed into an RNA product and then spliced. RNA splicing can also transfect the product with the variant sequence (at the DNA level) if the product production process involves reverse transcription, because the RNA can be spliced ​​before being reverse transcribed into DNA, and the DNA is then transfected into the cell.

[0101] The second mechanism, homologous recombination, occurs when reverse transcription is part of the manufacturing process. In this phenomenon, an actively transcribing reverse transcriptase "jumps" between two very similar (or identical) sequences in the RNA template, thus skipping intervening sequences and creating a non-standard DNA transcript that can later be transfected into the product. This mechanism is particularly risky for bicistronic CAR products, which may have several identical domains (e.g., co-stimulatory domains) in the two CARs, because it relies on very similar sequences in the template.

[0102] Considering both of these mechanisms, and examining conventional CAR-T cell production procedures using lentiviral vectors, we find several points on which variants may be produced. Firstly, lentiviral vectors can be produced by transfecting HEK293 cells with product and lentiviral-encoding plasmids, as shown in the example in Figure 1. The RNA produced by the HEK293 cells may be spliced ​​before packaging into these vectors. Secondly, T cells are transfected using lentiviral vectors. This involves reverse transcription of the product sequence before integration into the T cell genome, during which homologous recombination may occur. Thirdly, the transfected T cells (here, CAR-T cells) express CARs, and the expressed CAR mRNA may be spliced ​​before translation into its protein.

[0103] Regardless of origin, variants are undesirable. Variants in cell therapy products may have little to no efficacy. Furthermore, variants can cause cell therapy products to recognize different antigens, thus leading to off-target toxicity. The following describes different embodiments of common pipelines or processes for predicting, detecting, and eliminating variants to prevent these consequences from occurring.

[0104] One embodiment of the present disclosure relates to a method for detecting and substituting sequences that may induce undesirable variants in a gene construct. Such a method comprises: performing in silico analysis of a gene construct to detect the presence of sequences that may induce undesirable variants; substituting the detected sequences that may induce undesirable variants with alternative sequences, the alternative sequences being induced to include synonymous codon substitutions; and measuring the frequency percentage of undesirable variants expressed by the gene construct, comprising performing RNA sequencing analysis of RNA products transcribed from the gene construct, wherein the frequency percentage of undesirable variants is determined at least in part by using a splice-recognizing aligner from the RNA sequencing analysis, and the frequency percentage of undesirable variants in the gene products from the in vivo analysis is The method includes repeating the in silico analysis step and the substitution step if the acceptable frequency percentage of undesirable variants is greater than a predetermined value.

[0105] One embodiment of the present disclosure relates to the above method, wherein gap recognition alignment includes using at least two separate aligners.

[0106] One embodiment of the present disclosure relates to any of the above methods, wherein in silico analysis includes detecting at least one of a plurality of homologous sequences and a plurality of identical sequences in a gene construct, wherein at least one of the plurality of homologous sequences and a plurality of identical sequences may cause an undesirable variant in the gene construct, and further includes substituting such detected plurality of homologous sequences and a plurality of identical sequences, including a synonymous codon substitution step.

[0107] One embodiment of the present disclosure relates to any of the above methods, wherein the in silico analysis further comprises calculating a matrix of subsection combinations from a gene construct and obtaining a Hamming distance for each of the subsection combinations.

[0108] One embodiment of the present disclosure relates to any of the above methods, wherein the in silico analysis further comprises substituting a plurality of random synonymous codons in a gene construct with a plurality of alternative sequences, the plurality of alternative sequences increasing the sum of the entire matrix.

[0109] One embodiment of the present disclosure relates to any of the above methods, wherein the gene construct includes a sequence encoding a chimeric antigen receptor.

[0110] One embodiment of the present disclosure relates to any of the methods described above, wherein a predetermined value of an acceptable frequency percentage of an undesirable variant is determined based on whether the undesirable variant is associated with at least one of the following: whether the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface; whether the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor; and whether the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor.

[0111] One embodiment of the present disclosure relates to any of the above methods, wherein if the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface, a predetermined value for the acceptable frequency percentage of the undesirable variant is 0.1%, and if the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor, a predetermined value for the acceptable frequency percentage of the undesirable variant is 0.01%.

[0112] One embodiment of the present disclosure relates to any of the above methods, wherein if an undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor, the repetition of the in silico analysis step and the substitution step are not performed.

[0113] One embodiment of the present disclosure relates to any of the above methods, further comprising the steps of identifying and removing a subpopulation of high-frequency variants, identifying a subpopulation of low-frequency variants, and further comprising the step of performing an in vivo analysis to determine whether the subpopulation of low-frequency variants should be replaced.

[0114] In one embodiment of the present disclosure, if an undesirable variant is detected and none of the above conditions or requirements are met, the in silico analysis step and the sequence removal step are repeated in an attempt to eliminate the undesirable variant and / or the undesirable variant is a potential hazard. The above methods may be further characterized in additional trials to assess the risks to the individuals.

[0115] One embodiment of the present disclosure relates to a method for producing a gene product for use in cell therapy. Such a method includes the steps of: performing in silico analysis on a gene construct encoding a gene product to identify and modify sequences that may cause undesirable variants; substituting the detected sequences that may cause undesirable variants with alternative sequences, wherein the alternative sequences are induced to include synonymous codon substitutions; and performing in vivo analysis of one or more genes expressed by the gene construct, which includes the step of performing RNA sequencing analysis on an RNA product transcribed from the gene construct, wherein the frequency percentage of undesirable variants expressed by the gene construct is at least partially determined by using a splice recognition aligner from the RNA sequencing analysis. The process includes: a step of determining, if the frequency percentage of undesirable variants in the gene product from in vivo analysis is greater than a predetermined value of the acceptable frequency percentage of undesirable variants, a step of repeating the in silico analysis step and the substitution step to produce a new gene construct; and a step of performing in vivo analysis of one or more genes expressed by the new gene construct, which includes a step of performing RNA sequencing analysis of the RNA product transcribed from the new gene construct, wherein the frequency percentage of undesirable variants is determined, at least in part, by a step of using a splice recognition aligner from RNA sequencing analysis.

[0116] One embodiment of the present disclosure relates to the above method, wherein gap recognition alignment includes the step of using at least two separate aligners.

[0117] One embodiment of the present disclosure relates to any of the above methods, wherein in silico analysis comprises the steps of detecting at least one of a plurality of homologous sequences and a plurality of identical sequences in a gene construct, wherein at least one of the plurality of homologous sequences and a plurality of identical sequences may cause an undesirable variant in the gene construct, and further comprising the steps of substituting such detected plurality of homologous sequences and a plurality of identical sequences, including a synonymous codon substitution step.

[0118] One embodiment of the present disclosure relates to any of the above methods, wherein the in silico analysis further comprises the steps of: calculating a matrix of subsection combinations from a gene construct; and obtaining a Hamming distance for each of the subsection combinations.

[0119] One embodiment of the present disclosure relates to any of the above methods, wherein the in silico analysis further comprises the step of substituting a plurality of random synonymous codons in a gene construct with a plurality of alternative sequences, the plurality of alternative sequences increasing the sum of the entire matrix.

[0120] One embodiment of the present disclosure relates to any of the above methods, wherein the gene construct includes a sequence encoding a chimeric antigen receptor.

[0121] One embodiment of the present disclosure relates to any of the methods described above, wherein a predetermined value of an acceptable frequency percentage of an undesirable variant is determined based on whether the undesirable variant is associated with at least one of the following: whether the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface; whether the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor; and whether the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor.

[0122] One embodiment of the present disclosure relates to any of the above methods, wherein if the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface, a predetermined value for the acceptable frequency percentage of the undesirable variant is 0.1%, and if the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor, a predetermined value for the acceptable frequency percentage of the undesirable variant is 0.01%.

[0123] One embodiment of the present disclosure relates to any of the above methods, wherein if an undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor, the repetition of the in silico analysis step and the substitution step are not performed.

[0124] One embodiment of the present disclosure relates to any of the above methods, further comprising the steps of identifying and removing a subpopulation of high-frequency variants, identifying a subpopulation of low-frequency variants, and further comprising the step of performing an in vivo analysis to determine whether the subpopulation of low-frequency variants should be replaced.

[0125] One embodiment of the present disclosure relates to any of the above methods, wherein if an undesirable variant is detected and does not satisfy any of the above conditions or requirements, the in silico analysis step and the sequence removal step are repeated in an attempt to eliminate the undesirable variant, and / or the undesirable variant is further characterized in additional tests to assess the risk to potential patients.

[0126] One embodiment of the present disclosure relates to a method for reducing the risk when a gene product used in cell therapy has a risk of reduced efficacy or toxicity due to the production of undesirable variants. Such a method includes the steps of: performing in silico analysis of a gene construct encoding a gene product to detect the presence of sequences that may cause undesirable variants; substituting the detected sequences that may cause undesirable variants with alternative sequences, wherein the alternative sequences are induced to include synonymous codon substitutions; measuring the frequency percentage of undesirable variants expressed by the gene construct, which includes performing in vivo analysis of one or more genes expressed by the gene construct, including performing RNA sequencing analysis of an RNA product transcribed from the gene construct, wherein the frequency percentage of undesirable variants is determined at least in part by using a splice-recognizing aligner from the RNA sequencing analysis; and repeating the in silico analysis and substitution steps if the frequency percentage of undesirable variants in the gene product from the in vivo analysis is greater than a predetermined value of an acceptable frequency percentage of the undesirable variants.

[0127] Variant prediction, detection, and elimination One embodiment discloses a general-purpose pipeline for processing variants. This pipeline relies on both an algorithm for identifying and removing potential variants that cause sequencing, and a sequencing and alignment method for testing for the presence of variants. The algorithm and sequencing-based portions of this pipeline are described in more detail in subsequent sections.

[0128] Figure 2 shows an overview of one embodiment of the pipeline. This pipeline uses loop logic to make efforts to identify and remove variants that cause sequencing in silico, followed by physical testing and further in silico work if variants are found. Since variant detection is performed in multiple sequencing steps, little effort is spent on construct sequences that can be easily identified as variant-producing.

[0129] In this embodiment, the pipeline is configured to be flexible. The frequency cutoff shown in Figure 2 is an example, and alternative methods (discussed in the sequencing section) can be used to determine whether a variant needs to be addressed. Similarly, it is possible to increase / decrease the number of donors during the sequencing process, or to add / remove additional sequencing steps (e.g., Rentvirus sequencing) to maximize efficiency or to explore variants in more depth.

[0130] Referring to Figure 2, in one embodiment, prior to any physical work, the planned construct sequence is subjected to an algorithm designed to identify splice sites and highly homologous sequences (which can cause homologous recombination). If any of these are identified, other algorithms are used to remove them via synonymous codon substitution (the construct protein sequence remains unchanged). Once this initial in silico screening and modification is complete, a small test batch of the product using a single donor is prepared, and the RNA sequence of this batch is used to identify any high-frequency (e.g., greater than approximately 5%) variants. If high-frequency variants are found, the in silico screening and modification process is repeated, guided by the knowledge of the identified variants, before attempting to prepare the small batch and re-sequence the RNA. If no high-frequency variants are found, a larger batch of the test product using 5-10 donors is prepared, and the RNA sequence of this batch is used to identify any low-frequency (less than 5%) variants. If any of these variants are found, the sequence and expression levels can be analyzed in silico to determine whether the variant poses a safety / efficacy risk. As an example of this evaluation, variants resulting in a CAR with a spliced ​​co-stimulatory domain, occurring in only 1% of the product, are unlikely to be problematic, as they likely result in only a very small decrease in efficacy. In contrast, splicing in the ScFv domain of the CAR, even at a low percentage, can lead to off-target binding and therefore toxicity. If safety / efficacy risks cannot be ruled out, products expressing these specific variants can be constructed and physically tested. If these tests suggest or cannot be performed safety / efficacy issues, the construct sequence can be redesigned in silico and the entire process restarted. Otherwise, if no low-frequency variants are found or confirmed not to pose a safety / efficacy risk, the construct sequence can be cleared for further development.

[0131] Splice site prediction and removal algorithm During the in silico process of this pipeline, potential donor and receptor splice sites can be identified in one embodiment using the SpliceAI-Jaganathan et al., Cell 2019 (https: / / github.com / Illumina / SpliceAI) and MaxEntScan-Yeo and Burge, J Comput Biol 2003 (http: / / hollywood.mit.edu / burgelab / maxent / Xmaxentscan_scoreseq.html) algorithms. In other embodiments, other algorithms may be used.

[0132] In one embodiment, if a site in the construct is identified by any of these algorithms, the site may be modified using synonymous codon substitution until the variant is no longer detected or the likelihood of its predicted splicing is significantly reduced. This step keeps the construct protein sequence invariant. Similarly, if sequencing has already been performed and a variant has been detected, these algorithms can be used to determine whether the variant is likely due to splicing and to reduce the likelihood of the variant occurring using synonymous codon substitution.

[0133] Homology identification and removal algorithms Since homologous recombination is driven by very similar sequences in a construct, this disclosure describes algorithms and tools to assist in spotting and removing these sequences. Similar to splice site removal algorithms, these algorithms are performed in the initial in silico step and can be re-executed after the sequencing step if the sequencing indicates homologous recombination.

[0134] To identify sequences likely to cause homologous recombination, one embodiment includes a Repeat Finder tool. This tool analyzes and displays the construct sequence and draws an arch connecting all pairs of identical sequences longer than a given (user-defined) length. Alternatively, the tool can connect sequence pairs having a given size and a given level of similarity, as determined by a user-defined Levenshtein distance. Figure 6 shows an example screenshot illustrating the output display from the Repeat Finder tool. The thickness of the arch depends on the length of the identical sequences it connects. By searching for arches clustered together or groups of thick arches, the user can easily identify highly similar sequences, even if they are not identical.

[0135] To reduce the similarity between all subsections of a given construct, one embodiment includes a Sequence Diverger tool. This tool takes a construct sequence of length n (base pairs) and user-selected subsection size k (also base pairs). The tool then creates a matrix that is n-k+1 in both dimensions. Each position (x,y) in this matrix corresponds to a pair of subsections that begin with bases x and y of the construct sequence. The value of this position is the Hamming distance between these two subsections. An example of the matrix is ​​shown in Figure 3. With this design, the sum of the matrix effectively describes the similarity of the construct sequence to itself. A construct sequence with many very similar subsections will have a smaller sum than one where most of the subsections are different from each other. This means that reducing the similarity between all subsections of the construct, and therefore the likelihood of homologous recombination events, is a problem of increasing this matrix sum. Sequence Diverger achieves this by performing random synonymous codon substitutions across the entire construct sequence and retaining only those that increase the matrix sum. A user-specified number of permutations are performed before the algorithm finishes and the corrected sequence is returned. The user can track the matrix sum with respect to the number of permutations attempted, and thus get an intuitive sense of how many steps are required.

[0136] As shown in the example in Figure 3, for a given sequence (AACGAACG) and a given subsection size (4), the Hamming distance (HD) is calculated for all possible subsection pairs. The matrix sum describes how similar different subsections of the sequence are to each other, and therefore maximizing this sum should reduce the likelihood of homologous recombination.

[0137] In another embodiment, the Repeat Remover tool can be used to help prevent the construct sequence from containing identical subsections that could cause homologous recombination. This tool takes the construct sequence and a user-selected subsection size k (base pairs). In one embodiment, the Repeat Remover tool then undergoes a process in which each subsection of size k, starting from construct position 1, is compared with all other subsections in the construct. If identical subsections are found, a random synonymous codon substitution in that subsection is used to eliminate the similarity. This process is repeated in cycles until the entire construct is scanned without any identical subsections being found, or until a user-specified number of cycles is reached. An example of this is shown in Figure 4. Synonymous codon substitution is used to eliminate a pair of identical sequences and introduce another pair. Since this can be done in multiple cycles, multiple iterations may be necessary.

[0138] As shown in Figure 4, the Repeat Remover tool compares all subsections of a user-selected size (e.g., 6–10 base pairs) to each other to determine if they are identical. If identical pairs are found, they are removed using synonymous codon substitution. In this embodiment, this continues until all identical pairs are removed or until a user-defined number of cycles is reached.

[0139] While the Sequence Diverger and Repeat Remover tools share the same objective—removing highly similar sequences that could cause homologous recombination—they employ different approaches and potentially different results. The Repeat Remover focuses on eliminating identical sequences, and may do so at the expense of creating very similar (but not identical) sequences. In contrast, the Sequence Diverger focuses on reducing sequence similarity overall, and may do so at the expense of leaving a small number of identical sequences in place. In one embodiment, the in silico process can be performed using both tools in either order, or using only one, depending on what works best for the specific construct being developed. In other embodiments, neither tool may be used in the process; instead, manual removal of very similar sequences identified via the Repeat Remover (or Repeat Visualizer) may be used.

[0140] These embodiments are flexible in that the tools described above can be used individually or in combination with each other, and additional tools or features can be added. For example, all three tools can be configured to ignore very similar / identical sequences that are within a given distance from each other. This is appropriate because there are some observations that homologous recombination events require the occurrence of a minimum distance between very similar sequences. If certain regions of the construct are known to be very sensitive to codon usage frequency, it is also possible to configure the Sequence Diverger and Repeat Remover tools so that these regions do not change. Finally, the Sequence Diverger tool can be modified to accept or reject synonymous codon substitutions based on simulated annealing logic, rather than only accepting those that increase the matrix sum. This can help the tool find the global maximum more effectively.

[0141] In certain embodiments, the Repeat Finder, Sequence Diverger, and Repeat Remover tools may be individual computer modules. In other embodiments, these three tools may be programmed into a single module or computer.

[0142] Variant detection based on sequencing There is some flexibility in how RNA sequencing can be performed to identify variants. The sequencing should ideally be of sufficient quality and depth to allow for the observation of variants. As an example, in one embodiment, a CAR-T product can be sequenced using a HiSeq 2500 sequencing lane, with a depth of approximately 300 million reads per sample and 150 bp long paired-end reads.

[0143] Following sequencing, the reads should be aligned to the construct using one or more splice recognition aligners. This step offers flexibility based on precise alignment parameters and which aligner is used. This embodiment found good results using three different aligners (STAR, HISAT2, and TopHat2) simultaneously. This is because it helps identify and ignore errors in any single aligner (see below).

[0144] In one embodiment, after alignment, all reads with gap alignment (which represent variants) are extracted, and the resulting product protein sequences are translated. The percentage of each variant (protein sequence different from the intended sequence) in the sample is calculated using the following formula.

number

[0145] Following the allocation of lead and sample percentages to each variant, the final step in this embodiment is to determine which variants require further investigation. This step should allow for flexibility, considering whether it is an initial small screening to identify highly expressed variants or a larger screening to identify all variants (see Figure 2).

[0146] One option is to establish a cutoff based on the number of reads. For example, a sample may be considered positive for a variant if it has five or more reads supporting that variant. Another option would be to consider whether the percentage of samples supporting the variant (as shown in Figure 2) exceeds a certain cutoff. In both cases, if samples from multiple donors are used, a second cutoff based on multiple donors may be established. For example, only variants that are positive in samples from a given number or percentage of donors are considered. In another embodiment, it is possible to use a statistical test, such as the Wilcox rank-sum test, where each donor sample is considered an independent experiment, to determine whether the proportion of donors positive for a given variant is significantly different from zero. Finally, if multiple aligners are used, cutoffs based on them can be used to address inaccuracies in any one pipeline. For example, a sample may be considered positive for a variant only if it has five or more reads supporting that variant in at least two of the three aligners. Alternatively, the variant may only be considered if it returns a positive Wilcox rank-sum test across multiple donors with at least two pipelines.

[0147] Following statistical analysis of variants, visual inspection of variant-supporting and normal leads may be performed using a program such as Integrated Genomics Viewer (IGV). Some obvious variants may be the result of alignment errors that can be found through visual inspection. In certain embodiments of the process, in addition to performing the tests described above, the process may include isolating all leads in a sample that support a particular variant and placing them in a single BAM file for easy inspection.

[0148] RNA-seq profiling and bioinformatics analysis In another embodiment, RNA sequence profiling and bioinformatics analysis include quality control checks using FastQC (version 0.11.7), etc. Using FastQC's default parameters, the gene constructs should pass quality control. Profiling and analysis may also include an alignment step. To maximize splice event detection, three different splice recognition aligners—STAR (version 2.7.3a) (PMID: 23104886), HISAT2 (version 2.1.0) (PMID: 31375807), and TopHat2 (version 2.1.0) (PMID: 23618408)—are used to align reads to a construct array. A custom alignment index corresponding to the construct array can be generated for each tool. In one embodiment, the alignment may be performed on the Seven Bridges platform, but it may also be performed on other computing platforms.

[0149] In one embodiment, profiling and analysis may include STAR alignment. A reference index for STAR alignment can be created using the genomeGenerate command, where the genomeSAindexNbases parameter is set to 5 and all other parameters are set to their default values. In one embodiment, alignment in STAR can be performed using all default parameters. In one embodiment, the following command can be used to create the index and perform alignment with the STAR tool. STAR Index Command: STAR--runMode genomeGenerate--genomeDir. / genomeDir--runThreadN 32--genomeSAindexNbases 5-genomeFastaFiles construct_name.fa--limitGenomeGenerateRAM 60000000000 STAR Alignment Command: STAR--runThreadN 32--readFilesCommand zcat--genomeDir. / genomeDir--limitBAMsortRAM 0--outSAMtype BAM Unsorted--readFilesIn R1.fastq.gz R2.fastq.gz

[0150] Please understand that other commands may be used to create indexes and perform alignment.

[0151] Furthermore, profiling and analysis may include HISAT2 alignment. In one embodiment, a reference index for HISAT2 alignment is created using the hisat2-build command of HISAT2 version 2.0.1. HISAT2 alignment can be performed with the -no-softclip-no-unal option enabled and the -pen-cansplice and -pen-noncansplice parameters set to 0. All other parameters may be set to their defaults, and the reads are then sorted using Sambamba (version 0.6.6) (PMID:25697820). In one embodiment, the following commands can be used to create the index and perform alignment with the HISAT2 tool. HISAT2 index command: hisat2-build-p1 construct_name.fa index / construct_name_HISAT2-2.0.1 HISAT2 alignment command: hisat2-met-file metrics.txt--no-softclip--no-unal-p 20--pen-cansplice 0--pen-noncansplice 0-x. / index_files_path-1 R1_001.fastq.gz-2 R2_001.fastq.gz-S / dev / stdout

[0152] To create an index and perform alignment with the HISAT2 tool, other tools Please understand that a mand may be used.

[0153] In one embodiment, the profiling and analysis steps may include TopHat2 alignment. A reference index for TopHat2 alignment can be created using the BowTie2-build command (version 2.2.6) (PMID:21154709), etc. In one embodiment, TopHat2 alignment can be performed using all default parameters. The following commands can be used to create the index and perform alignment with the TopHat2 tool. BowTie2 - Build Commands: bowtie2-build-f construct.fa. / construct_name TopHat2 Alignment: tophat2-num-threads 1-output-dir. / tophat_out-no-coverage-search. / construct_name. R1_001.fastq.gz R2_001.fastq.gz

[0154] In one embodiment, the analysis step includes processing the reads. In this embodiment, the aligned reads from each alignment method may be further processed on the Seven Bridges platform. First, all non-gap reads may be removed using SAMtools (version 1.6) (PMID:19505943), view functions, etc. Then, the remaining gap reads in BAM file format may be converted to SAM format using SAMtools (version 1.9). Next, the nucleotide sequences from each gap read may be translated into the corresponding amino acid sequences using an R (version 3.6.2) (https: / / www.R-project.org / ) script (translate_and_group.R), etc. This script may also be used to calculate the number of reads supporting each unique gap event. In one embodiment, gap reads with overhangs of less than 10 base pairs may be removed. In one embodiment, this script used the Seqinr (version 3.6.1) (ISBN: 978-3-540-35305-8, https: / / cran.r-project.org / web / packages / seqinr / index.html) library package to translate gap DNA sequences into amino acid sequences. SAMtools command to remove non-gap reads: samtools view-h / path / to / input_bam.ext|awk'{if($0~ / ^@ / || $6~ / N / ){print$0}}' |samtools view-Sb->input_bam_gapped.bam SAMtools command to convert BAM to SAM file format: samtools view--output-fmt SAM-o hits_gapped.sam hits_gapped.bam

[0155] In one embodiment, the final step of the bioinformatics analysis may be performed on an Amazon Web Services (AWS) Virtual Private Cloud (VPC) using an EC2 instance running an R (version 3.6.2) script (app.R). This script imported the output file (translate_and_group.R) obtained from the s script, along with the BAM output file from the alignment, to calculate the occurrence rate of each gap event. The formula used for this calculation is as follows:

number

[0156] In one embodiment, the app.R script relies on the SAMtools (version 1.10) depth function to calculate the coverage of the construct and generates a BAM file (using a call to the SAMtools view function) with reads from all unique gap events for visualization purposes. Following this analysis, gap events can be considered variants if the event passes the following filtering criteria. The p-value (nonparametric one-sample Wilcoxon rank-sum test) was less than 0.01 in two of the three methods, and this cross-validation was used to minimize method-specific artifacts.

[0157] The null hypothesis and alternative hypothesis are as follows: H0:μ=0 H a :μ≠0

[0158] In one embodiment, a conservative threshold of p-value < 0.01 can be selected to minimize the false positive rate due to a large number of false presumptive variants and possibly sequencing artifacts supported by fewer than five reads in a single donor.

[0159] Report generation In one embodiment, a variant detection method may generate and display a report for each variant including the following information shown in Figure 5: 1) the expected sequence of the protein product, 2) the variant ID or number, 3) a diagram or schematic annotating the change in the characteristics of the protein product (i.e., CAR), 4) the frequency of alignment for each aligner used (e.g., Tophat, HISAT, STAR), 5) the p-value of the statistical significance of the detection (if available), and 6) visualization of the reads aligned to the construct.

[0160] Those skilled in the art will understand that the subject matter can be embodied in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments described herein should be considered illustrative in all respects and not limiting to the subject matter described herein. [Examples]

[0161] The following embodiments disclose an exemplary table of potential software packages that may be used in RNA sequencing profiling and bioinformatics analysis, along with exemplary software code portions that may be used in the disclosed processes. [Examples]

[0162] This embodiment provides an example of a method 101 for detecting sequence substitutions that may cause undesirable variants in a gene construct. As shown in Figure 7A, method 101 includes at least the following steps: Method 101 includes step 103 of performing in silico analysis of the gene construct to detect the presence of sequences that may cause undesirable variants. Method 101 also includes step 105 of replacing the detected sequences that may cause undesirable variants with alternative sequences. In step 105, the alternative sequences are induced via the use of synonymous codon substitutions. Method 101 includes step 107 of measuring the frequency percentage of undesirable variants expressed by the gene construct. Step 107 includes performing in vivo analysis of one or more genes expressed by the gene construct by performing RNA sequencing analysis of RNA products transcribed from the gene construct, and the frequency percentage of undesirable variants is at least partially derived from the splice analysis of the RNA sequencing analysis. This is determined by using a recognition aligner. Method 101 also includes repeating the in silico analysis step 103 and the substitution step 105 if the frequency percentage of undesirable variants in the gene product from in vivo analysis 107 is greater than a predetermined value of the acceptable frequency percentage of undesirable variants.

[0163] In method 101, step 107 requires the use of at least two distinct aligners. Method 101 is also preferably used for gene constructs encoding chimeric antigen receptors.

[0164] In Method 101, the predetermined acceptable frequency percentage of an undesirable variant is 0.1% if the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface, and the predetermined acceptable frequency percentage of an undesirable variant is 0.01% if the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor.

[0165] In Method 101, if the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor, the repetition of steps 103 and 105 of the in silico analysis is not performed. [Examples]

[0166] This embodiment provides an example of a method 201 for producing a gene product used in cell therapy. As shown in Figure 7B, method 201 includes at least the following steps: Method 201 includes performing in silico analysis on a gene construct encoding the gene product to identify and modify sequences that may produce undesirable variants 203. Method 201 also includes step 205 of replacing the detected sequences that may produce undesirable variants with alternative sequences. In step 205, the alternative sequences are induced by using synonymous codon substitutions. Method 201 includes step 207 of measuring the frequency percentage of undesirable variants expressed by the gene construct. Step 207 includes performing in vivo analysis of one or more genes expressed by the gene construct by performing RNA sequencing analysis on RNA products transcribed from the gene construct. In step 207, the frequency percentage of undesirable variants is determined at least in part by using a splice recognition aligner from the RNA sequencing analysis. Method 201 includes repeating in silico step 203 and substitution step 205 to construct a new gene construct if the frequency percentage of undesirable variants in the gene product from in vivo analysis is greater than a predetermined value of the acceptable frequency percentage of undesirable variants. Method 201 also includes step 209 to measure the frequency percentage of undesirable variants expressed by the new gene construct. Step 209 includes performing in vivo analysis of one or more genes expressed by the new gene construct by performing RNA sequencing analysis of RNA products transcribed from the new gene construct, and the frequency percentage of undesirable variants is determined at least in part by using a splice recognition aligner from the RNA sequencing analysis.

[0167] In method 201, step 207 requires the use of at least two distinct aligners. Method 201 is also preferably used for gene constructs encoding chimeric antigen receptors.

[0168] In Method 201, if an undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface, the predetermined value for the acceptable frequency percentage of the undesirable variant is 0.1%, and if the undesirable variant is related to a change in the binding domain of the chimeric antigen receptor, the predetermined value for the acceptable frequency percentage of the undesirable variant is It is 0.01%.

[0169] In Method 201, if the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor, the repetition of steps 203 and 205 of the in silico analysis is not performed. [Examples]

[0170] [Table 1] [Examples]

[0171] This embodiment provides an exemplary code portion that may be used to carry out the disclosed method. In the exemplary code portion shown in Figure 8A, x is the percentage coverage, R is the number of reads supporting the gap, n is the number of bases in the open reading frame (ORF), and d is the read depth at each position in the ORF (see lines 25-45 of the app.R script). [Examples]

[0172] This embodiment provides exemplary code portions that may be used to carry out the disclosed method. In the exemplary code portion shown in Figure 8B, the app.R script relies on the SAMtools (version 1.10) depth function to calculate the coverage of the construct and generates a BAM file (using calls to the SAMtools view function) with reads from all specific gap events for visualization purposes (see pages 48-82 of app.R). [Examples]

[0173] This embodiment provides an exemplary code portion that can be used to carry out the disclosed method. In the exemplary code portion shown in Figure 8C, a 0A conservative threshold of p-value < 0.01 was selected to minimize the false positive rate due to a large number of false-estimation variants and possibly sequencing artifacts, supported by fewer than 5 reads in a single donor. (See lines 131-136 of the app.R script).

[0174] All publications, patents, patent applications, and other documents cited herein are incorporated herein by reference in whole for all purposes to the same extent as each individual publication, patent, patent application, or other document is individually indicated to be incorporated herein by reference for all purposes.

[0175] While various specific embodiments / models have been illustrated and described, it will be understood that various modifications can be made without departing from the spirit and scope of this disclosure.

Claims

1. A method for detecting and substituting sequences that may cause undesirable variants in a gene construct, The steps include performing in silico analysis of the gene construct to detect the presence of the sequence that may cause the undesirable variant, A step of replacing the detected sequence that may cause the undesirable variant with an alternative sequence, wherein the alternative sequence is induced to include a synonymous codon substitution. A step of measuring the frequency percentage of the undesirable variant expressed by the gene construct, comprising a step of performing RNA sequencing analysis of an RNA product transcribed from the gene construct, and a step of performing in vivo analysis of one or more genes expressed by the gene construct, wherein the frequency percentage of the undesirable variant is determined at least partially from the RNA sequencing analysis using a splice recognition aligner, A method comprising: repeating the in silico analysis step and the substitution step if the frequency percentage of the undesirable variant in the gene product from the in vivo analysis is greater than a predetermined value of the acceptable frequency percentage of the undesirable variant.

2. The method according to claim 1, wherein the gap recognition alignment includes the step of using at least two separate aligners.

3. The aforementioned in silico analysis, A step of detecting at least one of a plurality of homologous sequences and a plurality of identical sequences within the gene construct, wherein the at least one of the plurality of homologous sequences and the plurality of identical sequences can cause an undesirable variant in the gene construct. The method according to claim 1 or 2, further comprising a step of substituting the detected plurality of homologous sequences and plurality of identical sequences, including a synonymous codon substitution step.

4. The method according to any one of claims 1 to 3, wherein the in silico analysis further comprises the steps of calculating a matrix of subsection combinations from the gene construct and obtaining a Hamming distance for each of the subsection combinations.

5. The method according to any one of claims 1 to 4, wherein the in silico analysis further comprises the step of substituting a plurality of random synonymous codons in the gene construct with a plurality of alternative sequences, wherein the plurality of alternative sequences increase the sum of the entire matrix.

6. The method according to any one of claims 1 to 5, wherein the gene construct comprises a sequence encoding a chimeric antigen receptor.

7. The method according to any one of claims 1 to 6, wherein the predetermined value of an acceptable frequency percentage of an undesirable variant is determined based on whether the undesirable variant is associated with at least one of the following: whether the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface; whether the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor; and whether the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor.

8. If the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface, the predetermined value of the acceptable frequency percentage of the undesirable variant The method according to any one of claims 1 to 7, wherein the predetermined value of the acceptable frequency percentage of the undesirable variant is 0.01%, and if the undesirable variant is related to a change in the binding domain of the chimeric antigen receptor, the predetermined value of the undesirable variant is 0.01%.

9. The method according to any one of claims 1 to 7, wherein if the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor, the in silico analysis step and the step of repeating the substitution step are not performed.

10. The process further includes identifying and removing subpopulations of high-frequency variants, and identifying subpopulations of low-frequency variants. The method according to any one of claims 1 to 9, further comprising the step of performing an analysis to determine whether the subpopulation of the low-frequency variants should be replaced.

11. A method for producing gene products used in cell therapy, The process involves performing in silico analysis on a gene construct encoding the gene product in order to identify and modify sequences that may cause undesirable variants. A step of replacing the detected sequence that may cause the undesirable variant with an alternative sequence, wherein the alternative sequence is induced to include a synonymous codon substitution. A step of measuring the frequency percentage of the undesirable variant expressed by the gene construct, comprising a step of performing RNA sequencing analysis of an RNA product transcribed from the gene construct, and a step of performing in vivo analysis of one or more genes expressed by the gene construct, wherein the frequency percentage of the undesirable variant is determined at least partially from the RNA sequencing analysis using a splice recognition aligner, If the frequency percentage of the undesirable variant in the gene product obtained from the in vivo analysis is greater than a predetermined value for the acceptable frequency percentage of the undesirable variant, the in silico analysis step and the substitution step are repeated to produce a new gene construct. A method comprising the steps of: measuring the frequency percentage of the undesirable variant expressed by the new gene construct, the step of performing an in vivo analysis of one or more genes expressed by the new gene construct, the step of performing an RNA sequencing analysis of an RNA product transcribed from the new gene construct, wherein the frequency percentage of the undesirable variant is determined at least in part from the RNA sequencing analysis using a splice recognition aligner.

12. The method according to claim 11, wherein the gap recognition alignment includes the step of using at least two separate aligners.

13. The aforementioned in silico analysis, A step of detecting at least one of a plurality of homologous sequences and a plurality of identical sequences within the gene construct, wherein the at least one of the plurality of homologous sequences and the plurality of identical sequences can cause an undesirable variant in the gene construct. The method according to claim 11 or 12, further comprising a step of substituting the detected plurality of homologous sequences and plurality of identical sequences, including a synonymous codon substitution step.

14. The method according to any one of claims 11 to 13, wherein the in silico analysis further comprises the steps of calculating a matrix of subsection combinations from the gene construct and obtaining a Hamming distance for each of the subsection combinations.

15. The method according to any one of claims 11 to 14, wherein the in silico analysis further comprises the step of substituting a plurality of random synonymous codons in the gene construct with a plurality of alternative sequences, wherein the plurality of alternative sequences increase the sum of the entire matrix.

16. The method according to any one of claims 11 to 15, wherein the gene construct comprises a sequence encoding a chimeric antigen receptor.

17. The method according to any one of claims 11 to 16, wherein the predetermined value of an acceptable frequency percentage of an undesirable variant is determined based on whether the undesirable variant is associated with at least one of the following: whether the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface; whether the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor; and whether the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor.

18. The method according to any one of claims 11 to 17, wherein if the undesirable variant adversely affects the transport of the chimeric antigen receptor to the cell surface, the predetermined value of the acceptable frequency percentage of the undesirable variant is 0.1%, and if the undesirable variant is associated with a change in the binding domain of the chimeric antigen receptor, the predetermined value of the acceptable frequency percentage of the undesirable variant is 0.01%.

19. The method according to any one of claims 11 to 17, wherein if the undesirable variant has been previously characterized as causing a negligible effect on the expression or function of the chimeric antigen receptor, the in silico analysis step and the step of repeating the substitution step are not performed.

20. The process further includes identifying and removing subpopulations of high-frequency variants, and identifying subpopulations of low-frequency variants. The method according to any one of claims 11 to 19, further comprising the step of performing an analysis to determine whether the subpopulation of the low-frequency variants should be replaced.

Images