CD3 / GD2 bispecific t cell engager

Abstract

This invention relates to a bispecific T cell engager (BiTE) comprising (a) a CD3-specific single-chain variable fragment (scFv) comprising a heavy chain variable region (VH) and a light chain variable region (VL); and (b) a GD2-specific scFv comprising a VH and a VL characterized in that the CD3-specific VL comprising complementarity- determining regions (CDRs) having amino acid sequences of SEQ ID NO: 2, SEQ ID NO: 3, and SEQ ID NO: 4.

Description

[0001] CD3 / GD2 BISPECIFIC T CELL ENGAGER

[0002] FIELD OF INVENTION

[0003] This invention relates to biotechnology, more particularly, to CD3 / GD2 bispecific T cell engager.

[0004] BACKGROUND OF THE INVENTION

[0005] Neuroectodermal tumors arise from abnormalities in cells of neuroectodermal origin including neuroblastoma, ganglioneuroblastoma, and ganglioneuroma. Neuroblastoma is a solid tumor that arises from abnormal development of neural crest cells. It can occur in various locations in the body, such as abdomen, chest, neck, near the spine, and adrenal glands. Neuroblastoma is the most common cancer in infants, accounting for approximately 6% of all childhood cancer and 15% of childhood cancer-related deaths. Patients with low-risk neuroblastoma have a survival rate of over 95%, while those with moderate risk have survival rates ranging from 90% to 95%. However, high-risk patients have a survival rate of less than 50%. Multidisciplinary treatments, including chemotherapy, surgery, radiation therapy, stem cell transplantation, and immunotherapy can increase overall survival rates to approximately 79%. Nevertheless, high-risk neuroblastoma patients still have a survival rate below 50%. Accordingly, neuroblastoma remains a significant health problem in children, and further improvements and developments in treatment methods for high-risk patients are needed.

[0006] Disialoganglioside (GD2) is a cell surface glycosphingolipid with limited expression in normal cells, while being highly expressed in various types of cancer cells originating from neuroectoderm, such as neuroblastoma and melanoma. High levels of GD2 expression are associated with rapid disease progression and lower patient survival rate. Therefore, GD2 is considered a tumor-associated antigen (TAA) and an important target for cancer therapy involving GD2 expression.

[0007] Immunotherapy is a high-potential method for treating various types of cancer, of which bispecific T cell engagers (BiTEs) are one such promising immunotherapeutic treatment specifically designed for cancer therapy. BiTEs are antibodies engineered to bespecific to two antigen sites within a single molecule, wherein one site is specific to an antigen on T cells and the other is specific to an antigen on cancer cells. This construct allows the molecule to bind simultaneously to T cells and cancer cells, thereby bringing the two cell types into close proximity and enabling T cells to kill cancer cells.

[0008] 0NC0IMMUN0L0GY 2016, VOL. 5, NO. 6, el 168557 discloses BiTEs that are specific to CD3 (cluster of differentiation 3) on T cells and GD2 on cancer cells, enabling simultaneous binding to both T cells and cancer cells and thereby allowing T cells to kill cancer cells. As disclosed in the cited document, the molecule comprises (a) a CD3-specific single-chain variable fragment (scFv) comprising a heavy chain variable region (VH) and a light chain variable region (VL), wherein the VL comprises complementaritydetermining regions (CDRs) having the amino acid sequences SASSSVSYMN, DTSKLAS, and QQWSSNPFT; and (b) a GD2-specific scFV comprising a VH and a VL. However, the cited document does not disclose the specificity of said BiTEs to V / 9V62 T cells, which are a subpopulation of T cells capable of stimulating immune responses, including the production of anticancer cytokines and induction of apoptosis in cancer cells.

[0009] CA2956471 discloses BiTEs comprising a CD3-specific scFv comprising a CDR, wherein the CDR comprises the amino acid sequence YGASTRAT. However, the cited document does not disclose the specificity of GD2-BiTEs toward cancer cells, nor does it suggest the application of such molecules for killing GD2-expressing cancer cells.

[0010] Therefore, this invention provides BiTEs that are specific to two antigen sites: CD3 on T cells and GD2 on cancer cells. These molecules can specifically bind to V / 9V62 T cells and to GD2-expressing cancer cells, thereby bringing T cells and cancer cells into close proximity and enhancing the ability of T cells to efficiently target and kill cancer cells. Accordingly, BiTEs have potential applications in the development of immunotherapeutic treatments for cancer patients having GD2-expressing cancer cells.

[0011] SUMMARY OF THE INVENTION

[0012] The present invention relates to a bispecific T cell engager (BiTE) specific to CD3 on T cells and GD2 on cancer cells. This BiTE enhances the efficiency of T cells in killing GD2-expressing cancer cells. The said CD3 / GD2 BiTE comprises(a) a CD3-specific single-chain variable fragment (scFv) comprising a heavy chain variable region (VH) and a light chain variable region (VL); and (b) a GD2-specific scFv comprising a VH and a VL,

[0013] characterized in that the CD3-specific VL comprises complementarity-determining regions (CDRs) having amino acid sequences of SEQ ID NO: 2, SEQ ID NO: 3, and SEQ ID NO: 4.

[0014] BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 shows binding activity of human scFv CD3 against CD3 epsilon as determined by indirect ELISA screening. Human scFv CD3 was extracted from cell lysates of sixty E.coli clones containing the human scFv gene insert. Human scFv CD3 clones 18, 6, and 85 exhibit the highest CD3 epsilon binding signals, respectively.

[0016] Figure 2 shows fractionated protein expression analyzed by western blot using an anti-E tag antibody. Periplasmic and cytoplasmic proteins of human scFv CD3 clones 6, 18, and 85 are shown. (Lane M represents a molecular weight marker; Lane HB represents protein prepared from E.coli strain HB2151).

[0017] Figure 3 shows binding activity of human scFv CD3 against CD3 epsilon as determined by indirect ELISA. The white bar represents signals from periplasmic proteins, and the gray bar represents signals from cytoplasmic proteins.

[0018] Figure 4 shows an amino acid sequence of human scFv CD3 clone 18. The complementarity-determining regions (CDRs) in the VH and VL regions are underlined with solid lines, while the linker connecting the VH and VL regions is underlined with a dashed line.

[0019] Figure 5 shows (A) a schematic diagram of CD3 / GD2 BiTE in pcDNA3.1. The vector comprises a gene encoding human scFv CD3 clone 18 arranged in a VH-VL orientation, linked to a mouse scFv-GD2 arranged in a VL-VH orientation via a linker having the amino acid sequence of SEQ ID NO: 15. In addition, an interleukin-2 signal peptide (IL-2 ss) is located at the N-terminus, and a polyhistidine tag (Hise) is located at the C-terminus, respectively; and (B) a simplified structural model of CD3 / GD2 BiTE.Figure 6 shows an amino acid sequence of CD3 / GD2 BiTE. The CDRs in the VH and VL regions are underlined with solid lines, while the linker and the polyhistidine (Hise) tag are underlined with dashed lines.

[0020] Figure 7 shows reducing-PAGE and western blot analysis of purified CD3 / GD2 BiTE protein. The molecular weight of CD3 / GD2 BiTE is approximately 58 kDa. (Lane M represents a molecular weight marker).

[0021] Figure 8 shows non-reducing-PAGE analysis of CD3 / GD2 BiTE treated with dithiothreitol (DTT) at different concentrations and stained with InstantBlue.

[0022] Figure 9 shows binding activity of CD3 / GD2 BiTE on (A) CD3-positive Jurkat cells and CD3-negative Raji cells; and (B) GD2-positive SH-SY5Y cells and GD2-negative SK-N-SH cells. Jurkat, Raji, SH-SY5Y, and SK-N-SH cells were incubated with various concentrations of CD3 / GD2 BiTE. Then the binding was determined by flow cytometry using Alexa Flour 488-conjugated 6xHis Tag antibody. The mean percentage of positive cells, which are defined as cells stained with the antibody specific to the polyhistidine tag, is shown as a line graph (left), while the mean fluorescence intensity (MFI) is shown as histograms (right). Experiments were performed in triplicate, with individual results plotted and means ± SD presented. *p < 0.05, **p < 0.01, ***p < 0.001.

[0023] Figure 10 shows immunophenotyping characterization of Vy9V52 T cells by flow cytometry (left). The cell number and percentage of Vy9V52 T cells on day 0 and day 14 are shown in the bar graph (top right). Fold expansion and purity on day 14 are also shown in the bar graph (bottom right).

[0024] Figure 11 shows mean percentage of dead neuroblastoma cells at Vy9V52-T-cell-to-neuroblastoma-cell ratios of approximately 1:1, 5:1, and 10:1 and at CD3 / GD2 BiTE concentrations of approximately 0 to 900 nM. Dead cells were stained with carboxyfluorescein succinimidyl ester (CFSE) and 7- Aminoactinomycin D (7-AAD). The experiments were performed in triplicate, and the results are expressed as means ± SD. *p 0.05.Figure 12 shows mean percentage of dead neuroblastoma cells at Vy9V52-T-cell-to-neuroblastoma-cell ratio of approximately 1:1 and at CD3 / GD2 BiTE concentrations of approximately 0 to 360 nM. Dead cells were stained with CFSE and 7-AAD. The experiments were performed in triplicate, and the results are expressed as means ± SD. *p < 0.05.

[0025] Figure 13 shows three-dimensional (3D) neuroblastoma cell killing effect including (A) fluorescence images of 3D neuroblastoma cells treated with CD3 / GD2 BiTE-armed Vy9V52 T cells; and (B) change in corrected total cell fluorescence (CTCF). Green fluorescent protein (GFP) signals indicate living cells, while ethidium homodimer-1 signals indicate dead cells. The experiments were performed in triplicate, and the results are expressed as means ± SD. *p < 0.05, **p < 0.01.

[0026] DETAILED DESCRIPTION OF THE INVENTION

[0027] The present invention relates to a bispecific T cell engager (BiTE) specific to two antigen sites: CD3 on T cells and GD2 on cancer cells.

[0028] Any embodiments depicted herein shall encompass modification to other aspects of this invention, unless stated otherwise.

[0029] Definition

[0030] This invention contains a Sequence Listing which is presented in XML file format and submitted via the electronic filing system.

[0031] Unless explicitly stated otherwise, technical or scientific terms used herein are defined to the understanding of those having ordinary skills in the art.

[0032] Unless explicitly stated otherwise, the equipment, apparatus, methods, or chemicals mentioned herein refer to those commonly operated or used by those skilled in the art. Any mention of equipment, apparatus, methods, or chemicals specifically used in this invention will be clearly specified.Unless the context clearly indicates otherwise, the terms “a,” “an,” and “the” in the claims or the specification should be interpreted as “one” as well as “one or more,” “at least one,” and “one or more than one,” and so forth.

[0033] When the term “comprising” is used in the claims or the specification, singular or plural nouns should be interpreted as “one” or “one or more,” “at least one,” and “one or more than one.”

[0034] Throughout this application, the term “about” is used to indicate that any value presented herein may potentially vary or deviate due to variety of factors, such as calculation errors, discrepancies in apparatus or methods, or differences between individual operators implementing the apparatus or methods.

[0035] All compositions and / or processes disclosed and claimed are intended to encompass variations involving actions, operation, modifications, or adjustments to any parameters, provided that they do not significantly deviate from experiments conducted, examples provided, or data presented in this invention, and they achieve similar objectives, utilities and results as described in the present invention, as understood by persons skilled in the art, even if not explicitly mentioned in the claims. Therefore, substitutions or similar modifications to the present invention, including minor changes apparent to those skilled in the art, should be considered to fall within the scope, spirit, and concept of the invention as defined by the appended claims.

[0036] The following specification is not intended to limit the scope of the invention in any manner.

[0037] According to a first embodiment of the invention, this invention relates to a bispecific T cell engager (BiTE) comprising

[0038] (a) a CD3-specific single-chain variable fragment (scFv) comprising a heavy chain variable region (VH) and a light chain variable region (VL); and (b) a GD2-specific scFv comprising a VH and a VL,characterized in that the CD3-specific VL comprises complementarity-determining regions (CDRs) having amino acid sequences of SEQ ID NO: 2, SEQ ID NO: 3, and SEQ ID NO: 4.

[0039] In a preferred embodiment of the invention, the CD3-specific VH comprises CDRs having amino acid sequences of SEQ ID NO: 5, SEQ ID NO: 6, and SEQ ID NO: 7.

[0040] In a preferred embodiment of the invention, the GD2-specific VL comprises CDRs having amino acid sequences of SEQ ID NO: 8, SEQ ID NO: 9, and SEQ ID NO: 10; and the GD2-specific VH comprises CDRs having amino acid sequences of SEQ ID NO: 11, SEQ ID NO: 12, and SEQ ID NO: 13.

[0041] In an embodiment of the invention, the BiTE further comprises a signal peptide for protein secretion.

[0042] In a preferred embodiment of the invention, the signal peptide for protein secretion is an interleukin-2 signal peptide.

[0043] In an embodiment of the invention, the BiTE further comprises a protein purification tag selected from a polyhistidine tag (Hise), a polyarginine tag, and a glutathione S-transferase (GST) tag.

[0044] In a preferred embodiment of the invention, the protein purification tag is the polyhistidine tag (Hise).

[0045] In an embodiment of the invention, the BiTE further comprises a linker selected from amino acid sequences of SEQ ID NO: 14, SEQ ID NO: 15, SEQ ID NO: 16, or a combination thereof.

[0046] In an embodiment of the invention, the BiTE is specific to a CD3 on a T cell and a GD2 on a neuroectodermal tumor.

[0047] In a preferred embodiment of the invention, the T cell is a V / 9V62 T cell.

[0048] In a preferred embodiment of the invention, the neuroectodermal tumor is a neuroblastoma cell.In an embodiment of the invention, the BiTE comprises an amino acid sequence as shown in SEQ ID NO: 1.

[0049] In an embodiment of the invention, this invention also relates to the use of BiTE for preparing a pharmaceutical composition for killing neuroectodermal tumors.

[0050] For a better understanding of the invention, various examples according to this invention will be presented. These examples are provided to illustrate embodiment of this invention. The scope of this invention is defined by the claims.

[0051] These examples are provided to illustrate embodiment of this invention and should not be construed as limiting the scope of the invention.

[0052] The following demonstrates the construction of CD3 / GD2 BiTE according to the invention and their application to enhance the efficiency of T cells in killing GD2-expressing cancer cells, such as neuroblastoma, using a mammalian cell model.

[0053] Cell Culture

[0054] The neuroblastoma cell lines, including SH-SY5Y and SK-N-SH, were obtained from American Type Culture Collection (ATCC) and cultured in high-glucose Dulbecco’s modified eagle medium (DMEM) (Hyclone, Cytiva, USA). A green fluorescent SH-SY5Y (SH-SY5Y GFP) was established through stable transfection and cultured in high-glucose DMEM medium.

[0055] Jurkat cells, a T-cell leukemia cell line, and Raji cells, a B-cell lymphoma cell line, were cultured in RPMI-1640 medium (Hyclone, Cytiva, USA).

[0056] Human embryonic kidney (HEK 293T) cells, used for CD3 / GD2 BiTE construction, were cultured in high-glucose DMEM medium.

[0057] All media were supplemented with 10% FBS (Gibco, CA, USA), lOO units / mL penicillin, and 100 pg / mL streptomycin (Gibco, Carlsbad, CA).All cell lines were incubated at approximately 37 °C in approximately 5% CO2 and approximately 70 to 80% humidified atmosphere.

[0058] In vitro expansion of human Vy9V52 T cells

[0059] Peripheral blood mononuclear cells (PBMCs) were isolated from heparinized blood samples using density gradient centrifugation (Lymphoprep, Oslo, Norway). PBMCs were adjusted to a final concentration of approximately 1 x 106cells / mL in complete RPMI 1640 medium, supplemented with 10%FBS, 100 units / mL penicillin, 100 pg / mL streptomycin, 2 mM L-glutamine (Gibco, Carlsbad, CA), 5 pM zoledronic acid (ZOL, ZOLennic, Bangkok, Thailand), and 50 U / mL recombinant human interleukin-2 (PeproTech, Rocky Hill, NJ) in 6-well flat-bottom plates (Costar, Cambridge, USA). Cells were cultured at approximately 37 °C in approximately 5% CO2 for approximately 14 days, with the medium supplement being replaced approximately every 3 days.

[0060] Characterization of human Vy9V52 T cells

[0061] The expression of cell surface markers on Vy9V52 T cells was analyzed using antibody staining. Prior to staining, an Fc block solution (BioLegend, USA) was added to the cells to prevent non-specific Fc receptor binding.

[0062] 7- Aminoactinomycin D (7-AAD) (eBioscience, CA, USA) was used to differentiate live cells from dead cells.

[0063] Antibodies specific for CD45-Pacific Blue clone 2D1, CD3-APC clone OKT3, Vy9-FITC clone B3, and V52-PE clone B3 (all from BioLegend (San Gabriel, CA, USA)) were used to evaluate cell surface markers.

[0064] The analysis was performed using a FACSLyric flow cytometer (BD Biosciences, USA).

[0065] Biopanning analysis of human single-chain variable fragments (scFv) expressed on phage display

[0066] Biopanning was performed using an Ml 3 phage library containing a naive human scFv gene insert in the phage genome and displaying human scFv fragments on the phage surface.Biopanning analysis of human scFv against CD3 was performed using a recombinant CD3 epsilon (Sigma- Aldrich, USA) as a target protein. Recombinant CD3 epsilon, at a concentration of approximately 0.5 pg, was coated onto about 8 ELISA wells (EIA / RIA high protein binding affinity, Costar, USA) and incubated overnight at approximately 4 °C. The pre-coated wells were blocked with a blocking buffer (approximately 2% bovine serum albumin (BSA) in TBS) for approximately 1 h at room temperature, followed by washing with a washing buffer (approximately 0.05% Tween 20 in TBS). Subsequently, the phage-displayed naive human scFv library was added and incubated for approximately 1 h at room temperature. Unbound phages were removed by washing with approximately 300 pL of washing buffer about 10 times. Antigen-bound phages were then eluted with approximately 50 pL of approximately 0.5 M HCl-glycine at a pH of approximately 2.2 and immediately neutralized with approximately 3 pL of approximately 2 M Tris-base. Afterward, approximately 200 pL of log-phase A. coli strain HB2151 was added to the well and incubated for approximately 20 min at approximately 25 °C to allow phage transfection into the A. coli. The phage-transformed A. coli was plated on Luria broth agar (LB agar) containing ampicillin (Bio Basic Inc, Canada) and incubated overnight at approximately 37 °C.

[0067] Ninety E. coli colonies were selected and screened for the insertion of human scFv genes by colony PCR technique. The recombinant E. coli clones containing the human scFv gene were cultured and induced for protein expression using isopropyl B-D-l-thiogalactopyranoside (IPTG) at approximately 30 °C overnight. Total cell lysates from each clone were extracted and tested for the binding activity of human scFv against recombinant CD3 epsilon by indirect ELISA using an HRP-conjugated anti-E tag (GE Healthcare, UK). After ELISA screening, positive clones that demonstrates the highest signals were selected and further analyzed for protein expression, binding activity, and nucleotide sequence.

[0068] Subcellular fractionation

[0069] The recombinant E. coli clones containing the human scFv gene were cultured in approximately 10 mL of LB broth containing ampicillin (Bio Basic Inc, Canada) and induced for protein expression using IPTG at approximately 30 °C overnight. The recombinant E. coli clones were then pelleted by centrifugation at approximately 1,811 xgfor approximately 10 min and washed once with cold PBS. The washed pellet was resuspended gently in approximately 1 mL fractionation buffer (approximately 30 mM Tris-HCl with a pH of approximately 8.0, approximately 20% sucrose, approximately 1 mM Na2EDTA) and incubated at room temperature for approximately 10 min. The cells were centrifuged at approximately 4,472*g for approximately 10 min, and the supernatant was discarded. The pellet was gently resuspended in approximately 266 pL of approximately 5 mM MgSo4 and incubated on ice for approximately 10 min. After centrifugation at approximately 7,558*g for approximately 10 min, the supernatant was collected as the periplasmic fraction. The remaining cell pellet was resuspended with approximately 266 pL of PBS and lysed by sonication. The cell lysate was centrifuged at approximately 7,558*g for approximately 10 min, and the supernatant was collected as the cytoplasmic fraction.

[0070] The recombinant E. coli without the human scFv gene was used to prepare the total cell lysate, periplasmic fraction, and cytoplasmic fraction as a negative protein control.

[0071] Indirect ELISA

[0072] Microtiter plates (Greiner CELLSTAR, Merck, Germany) were coated with approximately 0.5 pg of recombinant CD3 epsilon and incubated overnight at approximately 4 °C. BSA (Calbiochem, Merck, Germany) at approximately 0.5 pg was used as an irrelevant antigen. The coated wells were washed with approximately 200 pL of PBST and blocked with approximately 200 pL of approximately 5% skim milk in PBS for approximately 1 h at approximately 37 °C. After washing for about 3 times, approximately 500 pg / mL of human scFv CD3 from total cell lysate, periplasmic fraction, or cytoplasmic fraction was added to the wells and incubated at approximately 37 °C for approximately 1 h. Unbound proteins were removed by washing about 3 times with washing buffer. HRP-conjugated anti-E tag (GE Healthcare, UK) was added to the wells to detect scFvs bound to recombinant CD3 epsilon. The wells were incubated at approximately 37 °C for approximately 1 h and then washed about 3 times with washing buffer. Subsequently, approximately 50 pL of 2,2’-Azino-bis(3-ethylbenzothiazoline-6-sulfonic acid (ABTS) substrate (Vector Laboratories, USA) was added. The absorbance at approximately 405 nm was measured using a microplate reader (BioTek Synergy HTX,USA). The recombinant E. coli without the human scFv gene was used to prepare total cell lysate, periplasmic fraction, or cytoplasmic fraction as a negative protein control.

[0073] The absorbance of human scFv CD3 against CD3 epsilon was analyzed by subtracting the absorbance of the negative protein in CD3 epsilon, followed by the absorbance of human scFv CD3 in BSA-coated well, respectively.

[0074] Construction of CD3 / GD2 BiTE

[0075] The sequence of human scFv against CD3 is the human scFv-CD3 clone 18, derived from phage display selection. The sequence of mouse scFv against GD2 (mouse scFv-GD2), kindly provided by Professor Malcolm K. Brenner (Baylor College of Medicine, Texas), were obtained. The gene encoding CD3 / GD2 BiTE consisting of human scFv-CD3 linked with mouse scFv-GD2, was designed and optimized for codon usage in the mammalian cell expression system. The mammalian cell expression vector pcDNA3.1(+) containing CD3 / GD2 BiTE (pcDNA3.1-CD3 / GD2 BiTE) was constructed by Genscript Biotech Corporation (NJ, USA).

[0076] and purification of CD3 / GD2 BiTE

[0077] The expression vector pcDNA3.1-CD3 / GD2 BiTE was transfected into the HEK 293T cell line using Lipofectamine 3000 reagent (Invitrogen, Thermo Fisher Scientific, USA). After transfection, the cells were cultured for approximately 72 h, and the culture medium was collected to purify the CD3 / GD2 BiTE using Ni-NTA agarose (Protino, Macherey-Nagel, Germany). The purified CD3 / GD2 BiTE protein was further characterized by SDS-PAGE and western blot analysis.

[0078] Reducing-PAGE and western blotting

[0079] Protein analysis was conducted using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) with a Mini-PROTEAN II system (Bio-Rad, CA, USA). Proteins were separated on approximately 15% SDS-PAGE gels under denaturing conditions, and the bands were stained with InstantBlue™ (Sigma-Aldrich, USA). Then separated protein bands were transferred onto a nitrocellulose membrane using a Bio-Rad Mini Trans-Blot cell. The nitrocellulose membrane was incubated in a blocking buffer (approximately 5% non-fat dry milk in TBST buffer: approximately 10 mM Tris-HCl with a pH ofapproximately 8.0, approximately 150 mM NaCl, and approximately 0.05% Tween-20) for approximately 1 h at room temperature.

[0080] To detect human scFv CD3 protein, the membrane was incubated overnight at approximately 4 °C with mouse anti-E-Tag monoclonal antibody (GE Healthcare, UK). After washing for about 3 times with TBST, AP-conjugated goat anti-mouse immunoglobulin (DAKO, Denmark) was added and incubated for approximately 1 h at room temperature. Protein bands were visualized on the nitrocellulose membrane using a substrate (2,6-dichloroindophenol (DCIP)) (Sigma Chemical Co., USA).

[0081] To detect the CD3 / GD3 BiTE protein, the membrane was incubated overnight at approximately 4 °C with anti -histidine tag antibody clone H8 (Sigma-Aldrich, USA). After washing with TBST about 10 min for 2 times, the membrane was incubated for approximately 2 h at room temperature with HRP-conjugated anti-mouse IgA + IgG + IgM (H+L) antibody (KPL, SeraCare, MA, USA). The specific protein bands were visualized using enhanced chemiluminescence (ECL) (Bio-Rad, CA, USA).

[0082] Non-reducing-PAGE

[0083] Non-reducing PAGE was performed to study the natural structure of the CD3 / GD2 BiTE. Approximately 10 pg of CD3 / GD2 BiTE were incubated with the reducing agent, dithiothreitol (DTT) (Affymetrix, Santa Clara, CA, USA), at various concentrations at approximately 37 °C for approximately 30 min. The reduced protein was then subjected to approximately 12% SDS-PAGE under native conditions and stained with InstantBlue™ (Sigma- Aldrich, USA).

[0084] Binding assay of CD3 / GD2 BiTE to target antigens

[0085] The binding activity of CD3 / GD2 BiTE to target molecules, including CD3 and GD2 antigens, was evaluated by flow cytometry.

[0086] GD2-positive SH-SY5Y cells, GD2-negative SK-N-SH cells, CD3-positive Jurkat cells, and CD3-negative Raji cells, at a cell concentration of approximately 1 x 105cells / mL, were incubated with various CD3 / GD2 BiTE concentrations at approximately 4 °C for approximately 1 h. After washing with approximately 1 mL of PBS, cells were stained with Hise Tag antibody -Alexa 488 (Thermo Fisher Scientific, USA) and incubatedat approximately 4 °C for approximately 1 h. Cells were then washed twice and resuspended in approximately 100 pL of PBS before analyzed by flow cytometry (FACSCanto II, BD Biosciences, USA).

[0087] In vitro cytotoxicity assay

[0088] SH-SY5Y cells were used as target neuroblastoma cells of the cytotoxicity assay. Briefly, approximately 5 x 104cells / mL of SH-SY5Y cells were stained with CellTrace™ carboxyfluorescein succinimidyl ester (CFSE) (Thermo Fisher Scientific, USA) and added to 96-well flat-bottom plates (Costar, Cambridge, USA) to incubate overnight. Next, the Vy9V52 T cells were pre-incubated with CD3 / GD2 BiTE for approximately 30 min, then co-cultured with CFSE-stained SH-SY5Y cells at T-cell-to-SH-SY5Y-cell ratios of approximately 1:1, 1:5, and 1:10 for approximately 24 h. After that, SH-SY5Y cells were stained with 7-AAD (Biolegend, United Kingdom), indicating dead cells, and such cells were analyzed by flow cytometry (FACSCanto II, BD Biosciences, USA). The percentage of dead neuroblastoma cells was calculated using the following formula:

[0089] dead neuroblastoma cells (CSFE+, 7 — AAD+) % Dead neuroblastoma cells = [ - - - — - - — - - -1 x 100 total neuroblastoma cells (CSFE+, 7 — AAD— )

[0090] Three-dimensional (3D) neuroblastoma killing assay

[0091] To create 3D neuroblastoma cells, approximately 2 x 104cells / mL of SH-SY5Y GFP cells were cultured in a medium with approximately 2.5% Matrigel matrix (Corning Inc., NY, USA). Cells were cultured in ultra-low attachment polystyrene plate (Coming Inc., NY, USA), centrifuged at approximately 1000xg at approximately 4 °C for approximately 10 min, and incubated for approximately 72 h.

[0092] The Vy9V52 T cells were stained with CellTrace™ Violet (Invitrogen, Thermo Fisher Scientific, USA) and incubated with CD3 / GD2 BiTE before being co-cultured with 3D neuroblastoma cells at T-cell-to-SH-SY5Y-GFP-cell ratio of approximately 1:1. After approximately 48 h of co-culture, an ethidium homodimer-1 (Invitrogen, Thermo Fisher Scientific, USA) staining was performed to determine dead cells. The 3D neuroblastoma cell morphology was captured using Operetta CLS High-Content Analysis System (PerkinElmer, USA). The corrected total cell fluorescence (CTCF) of the 3Dneuroblastoma cells was analyzed using ImageJ software and calculated using the following formula:

[0093] CTCF = Integrated Density — (Selected cell area x MFI on background readings')

[0094] Statistical analysis

[0095] The experiments were performed in triplicate, and the results are expressed as the means ± SD. For individual comparisons, statistical analysis was performed using oneway ANOVA. Data with a statistical value of P < 0.05 was considered statistically significant.

[0096] Results

[0097] Human scFv CD3 screening

[0098] Human scFv against CD3 epsilon from an Ml 3 phage library displaying human scFv fragments was screened by phage panning technology with recombinant CD3 epsilon. CD3 epsilon-bound phages were eluted and transformed into E. coll strain HB2151. Ninety colonies that grew on LB agar containing ampicillin were randomly selected for human scFv gene insertion analysis by colony PCR. Sixty out of ninety colonies contained human scFv gene insertion.

[0099] These sixty clones were cultured and induced for human scFv protein expression. Subsequencetly, scFv was extracted from the total cell lysate, and their binding activity against CD3 epsilon was determined using indirect ELISA. The human scFv CD3 clones 18, 6, and 85 exhibited the highest signals, respectively (Figure 1). These three clones were selected for the protein expression analysis. Periplasmic and cytoplasmic proteins were isolated using fractionation buffer and then analyzed by western blot and indirect ELISA.

[0100] The western blot analysis (Figure 2) revealed protein bands of scFv CD3 clones 6, 18, and 85 with a molecular weight of approximately 28 kDa. In contrast, no bands were observed in fractionated proteins prepared from recombinant A. coli HB2151. The indirect ELISA results (Figure 3) showed that human scFv CD3 clone 18 exhibited the highest binding signal in both periplasmic and cytoplasmic fractions.The amino acid sequence of human scFv CD3 clone 18 indicating VH-VL orientation and CDR regions was shown in Figure 4.

[0101] Based on protein expression and the highest binding activity against CD3 epsilon, the human scFv CD3 clone 18 was selected to construct the CD3 / GD2 BiTE.

[0102] Vector for CD3 / GD2 BiTE construction

[0103] The mammalian cell expression vector, pcDNA3.1(+), containing the genes encoding CD3 / GD2 BiTE (pcDNA3.1-CD3 / GD2 BiTE), was constructed as illustrated in Figure 5 and 6.

[0104] The human scFv CD3 clone 18, VH-VL orientated, and the mouse scFv GD2, VL-VH orientated, were synthesized and linked by a linker (SEQ ID NO: 15).

[0105] An interleukin-2 signal peptide was fused to the N-terminus to facilitate the secretion of the CD3 / GD2 BiTE

[0106] A polyhistidine tag (Hise) was added to the C-terminus for detection and purification.

[0107] CD3 / GD2 BiTE construction

[0108] HEK 293T cells were transfected with pcDNA3.1-CD3 / GD2 BiTE and cultured for approximately 72 h. After post-transfection, the cell culture medium containing the CD3 / GD3 BiTE was collected for separation and purification using Ni-NTA agarose.

[0109] Protein production and purity were confirmed by SDS-PAGE and western blot analysis. A single band of approximately 58 kDa was observed on reducing SDS-PAGE stained with InstantBlue (Figure 7).

[0110] Western blot analysis using an anti-His tag antibody confirmed the production and purification of the CD3 / GD2 BiTE, showing approximately 58 kDa band on the blot (Figure 7).

[0111] Additionally, protein dimerization was assessed by treating the CD3 / GD2 BiTE protein with DTT before running non-reducing PAGE. As shown in Figure 8, the single band was observed for the protein treated with approximately 1 mM DTT, indicating thatthe CD3 / GD2 BiTE remained monomeric. However, a complete reduction occurred at approximately 10 and 50 mM of DTT, resulting in approximately 58 kDa band on nonreducing PAGE.

[0112] The production yield of CD3 / GD2 BiTE was approximately 7.2 mg / L of culture medium, demonstrating that the CD3 / GD2 BiTE could be efficiently produced in mammalian cells with high yield and purity.

[0113] Binding activity of CD3 / GD2 BiTE to target antigens

[0114] The binding activity of CD3 / GD2 BiTE to target molecules, including CD3 and GD2, was evaluated by flow cytometry.

[0115] CD3 -positive cells (Jurkat cells), CD3 -negative cells (Raji cells), GD2-positive cells (SH-SY5Y cells), and GD2-negative cells (SK-N-SH cells) were incubated with CD3 / GD2 BiTE at various concentrations, followed by staining with Alexa 488-conjugated 6xHis Tag antibody to measure the percentage of positive cells (cells stained with antibodies).

[0116] When Jurkat cells were incubated with CD3 / GD2 BiTE at concentrations of approximately 18, 180, and 900 nM, the percentage of positive cells was 14.82%, 58.85%, and 86.98%, respectively. The mean fluorescence intensity was elevated with higher CD3 / GD2 BiTE concentrations (Figure 9A).

[0117] Similarly, the binding interaction of CD3 / GD2 BiTE with SH-SY5Y cells had a comparable pattern to that observed with Jurkat cells (Figure 9B).

[0118] In contrast, the percentage of positive cells for Raji and SK-N-SH cells remained below 10% at all tested CD3 / GD2 BiTE concentrations. In conclusion, CD3 / GD2 BiTE demonstrated strong and specific binding to its targets, the CD3 molecule on Jurkat cells and the GD2 molecule on SH-SY5Y cells, in a dose-dependent manner.

[0119] In vitro cytotoxicity to neuroblastoma cells

[0120] Vy9V52 T cell populations were expanded in vitro and characterized using flow cytometry. As shown in Figure 10, approximately 3.39±2.67% of yST cells were present in initial PBMCs at a concentration of approximately IxlO6cells / mL on day 0. After 14days of expanded culture, the Vy9V52 T cells increased to approximately 88.56±10.32%, with a fold expansion of approximately 103.29±50.76. The purity of the expanded Vy9V52 T cells was approximately 82.32±8.92%. These findings confirmed that Vy9V52 T cells were activated and proliferated effectively using ZOL and IL-2.

[0121] To evaluate whether the CD3 / GD2 BiTE enhances in vitro cytotoxicity of Vy9V52 T cells, the cytotoxicity assay was tested at Vy9V52-T-cell-to-neuroblastoma-cell ratios of approximately 1:1, 5:1, and 10:1 and CD3 / GD2 BiTE concentrations corresponding to 50% and 100% of binding activity. As shown in Figure 11, the mean percentage of dead neuroblastoma cells at Vy9V52-T-cell-to-neuroblastoma-cell ratios of 1:1, 5:1, and 10:1 was approximately 21.65 ± 1.20%, 40.40 ± 4.15%, and 44.58 ± 5.43%, respectively.

[0122] Upon increasing the CD3 / GD2 BiTE concentrations, the percentage of dead neuroblastoma cells had increased at all ratios, with significant enhancement in cytotoxicity effect observed only at Vy9V52-T-cell-to-neuroblastoma-cell ratio of approximately 1:1. Specifically, the percentage increased from approximately 21.65±1.20% to 30.87±2.90% (P < 0.005) with approximately 90 nM CD3 / GD2 BiTE. However, at Vy9V52-T-cell-to-neuroblastoma-cell ratios of approximately 5:1 and 10:1, the addition of CD3 / GD2 BiTE did not significantly enhance cytotoxicity.

[0123] The effect of CD3 / GD2 BiTE concentration was evaluated. As shown in Figure 12, at Vy9V52-T-cell-to-neuroblastoma-cell ratio of approximately 1:1, the percentage of neuroblastoma cell death increased in dose-dependent manner of CD3 / GD2 BiTE. Significant cytotoxic effects were observed at CD3 / GD2 BiTE concentrations of approximately 180 and 360 nM, with approximately 38.10±4.20% and 42.30±4.45% neuroblastoma cell death, respectively. This indicated that the combination of CD3 / GD2 BiTE enhanced the cytotoxicity approximately 1.57-fold and 1.75-fold higher compared to that of without CD3 / GD2 BiTE (mock).

[0124] Notably, cytotoxicity of CD3 / GD2 BiTE at approximately 360 nM with Vy9V52-T-cell-to-neuroblastoma-cell ratio of approximately 1:1 (Figure 12) was comparable to that of without CD3 / GD2 BiTE (mock) at Vy9V52-T-cell-to-neuroblastoma-cell ratios of approximately 5:1 and 10:1 (Figure 11), demonstrating that CD3 / GD2 BiTE stimulated approximately 5 to 10 times less killing of neuroblastoma cells by Vy9V52 T cells.Cytotoxicity against 3D neuroblastoma cells

[0125] The killing ability of the combination of CD3 / GD2 BiTE and Vy9V52 T cells was evaluated using a 3D neuroblastoma cell model of SH-SY5Y GFP cells. 3D neuroblastoma cells and Vy9V52 T cells armed with CD3 / GD2 BiTE were co-cultured for approximately 48 h at T-cell-to-neuroblastoma-cell ratio of approximately 1:1. Dead cells were stained with ethidium homodimer-1. As a result, the GFP signal indicated that neuroblastoma cells were slightly decreased when treated in the absence of CD3 / GD2 BiTE, whereas the GFP signal was significantly decreased at approximately 360 nM of CD3 / GD2 BiTE (p < 0.05) (Figures 13A and 13B).

[0126] Meanwhile, the ethidium homodimer-1 signal, indicating dead cells, significantly increased in a dose-dependent manner of CD3 / GD2 BiTE, ranging from approximately 1.75 to 2.02-fold. Notably, approximately 180 nM of CD3 / GD2 BiTE exhibited the highest CTCF fold change at 2.20-fold (p < 0.01) (Figure 13A and 13C).

[0127] These results indicated that the combination of CD3 / GD2 BiTE and Vy9V52 T cells effectively demonstrated killing capability against GD2-expressing cancer cells such as neuroectodermal tumors.

[0128] BEST MODE OF THE INVENTION

[0129] Best mode of the invention is as provided in the description of the invention.

Claims

CLAIMS1. A bispecific T cell engager (BiTE) comprising(a) a CD3-specific single-chain variable fragment (scFv) comprising a heavy chain variable region (VH) and a light chain variable region (VL); and (b) a GD2-specific scFv comprising a VH and a VL,characterized in that the CD3-specific VL comprises complementarity-determining regions (CDRs) having amino acid sequences of SEQ ID NO: 2, SEQ ID NO: 3, and SEQ ID NO: 4.

2. The BiTE according to claim 1, wherein the CD3 -specific VH comprises CDRs having amino acid sequences of SEQ ID NO: 5, SEQ ID NO: 6, and SEQ ID NO: 7.

3. The BiTE according to claim 1, wherein the GD2-specific VL comprises CDRs having amino acid sequences of SEQ ID NO: 8, SEQ ID NO: 9, and SEQ ID NO: 10; and the GD2-specific VH comprises CDRs having amino acid sequences of SEQ ID NO: 11, SEQ ID NO: 12, and SEQ ID NO: 13.

4. The BiTE according to claim 1, wherein said molecule further comprises a signal peptide for protein secretion.

5. The BiTE according to claim 1, wherein said molecule further comprises a protein purification tag selected from a polyhistidine tag (Hise), a polyarginine tag, and a glutathione S-transferase (GST) tag.

6. The BiTE according to claim 1, wherein said molecule further comprises a linker selected from amino acid sequences of SEQ ID NO: 14, SEQ ID NO: 15, SEQ ID NO: 16, or a combination thereof.

7. The BiTE according to claim 1, wherein said molecule is specific to a CD3 on a T cell and a GD2 on a neuroectodermal tumor.

8. The BiTE according to claim 7, wherein the neuroectodermal tumor is a neuroblastoma cell.

9. The BiTE according to any one of claims 1 to 7, wherein said molecule comprises an amino acid sequence as shown in SEQ ID NO: 1.

10. The use of BiTE according to any one of claims 1 to 9 for preparation of a pharmaceutical composition for killing neuroectodermal tumors.

Images