BOXR1030, an anti-GPC3 CAR with exogenous GOT2 expression, shows enhanced T cell metabolism and improved anti-cell line derived tumor xenograft activity

Abstract

Purpose: The solid tumor microenvironment (TME) drives T cell dysfunction and inhibits the effectiveness of immunotherapies such as chimeric antigen receptor-based T cell (CAR T) cells. Early data has shown that modulation of T cell metabolism can improve intratumoral T cell function in preclinical models.

Experimental design: We evaluated GPC3 expression in human normal and tumor tissue specimens. We developed and evaluated BOXR1030, a novel CAR T therapeutic co-expressing glypican-3 (GPC3)-targeted CAR and exogenous glutamic-oxaloacetic transaminase 2 (GOT2) in terms of CAR T cell function both in vitro and in vivo.

Results: Cell surface expression of tumor antigen GPC3 was observed by immunohistochemical staining in tumor biopsies from hepatocellular carcinoma, liposarcoma, squamous lung cancer, and Merkel cell carcinoma patients. Compared to control GPC3 CAR alone, BOXR1030 (GPC3-targeted CAR T cell that co-expressed GOT2) demonstrated superior in vivo efficacy in aggressive solid tumor xenograft models, and showed favorable attributes in vitro including an enhanced cytokine production profile, a less-differentiated T cell phenotype with lower expression of stress and exhaustion markers, an enhanced metabolic profile and increased proliferation in TME-like conditions.

Conclusions: Together, these results demonstrated that co-expression of GOT2 can substantially improve the overall antitumor activity of CAR T cells by inducing broad changes in cellular function and phenotype. These data show that BOXR1030 is an attractive approach to targeting select solid tumors. To this end, BOXR1030 will be explored in the clinic to assess safety, dose-finding, and preliminary efficacy (NCT05120271).

Conflict of interest statement

I have read the journal’s policy and the authors of this manuscript have the following competing interests: TLH is an employee of Takeda Pharmaceuticals, a former employee of Unum Therapeutics, has ownership interest in Unum Therapeutics (now Cogent Biosciences), and has issued patents: PCT/US18/00028 and PCT/US2018/015999-all outside the submitted work. EC is an employee of Catamaran Bio, Inc, a former employee of Unum Therapeutics; has ownership interest in Unum Therapeutics (now Cogent Biosciences) and Catamaran Bio, Inc; and has issued patents: PCT/US2010/0041032A1, PCT/US2019/0153061A1, PCT/US2019/0105348A1, and PCT/US2019/0284298A1-all outside the submitted work. KRW is an employee of SOTIO Biotech Inc., a former employee of Unum Therapeutics; has ownership interest in Unum Therapeutics (now Cogent Biosciences) and has issued patents: PCT/US2012/0282175A1 and PCT/US2010/0028346A1 all outside the submitted work. SM is an employee of SOTIO Biotech Inc., a former employee of Kiniksa Pharmaceuticals. TP is an employee of Novartis and a former employee of Unum therapeutics. TJ is an employee of Novartis, a former employee of Unum Therapeutics, and has ownership interest in Unum Therapeutics (now Cogent Biosciences). TF is an employee of Biogen and a former employee of Unum therapeutics. MG is an employee of Arrakis Therapeutics and a former employee of Unum therapeutics. BS is an employee of Tango Therapeutics, a former employee of Unum Therapeutics, and has issued patents: US9464333, US9388422, US9303250, WO2014055778A2, WO2016069774A1, and WO2017049094A1-all outside the submitted work. LB is an employee of Catamaran Bio, Inc. and a former employee of Unum therapeutics. KEM is an employee of Arrakis Therapeutics, a former employee of Unum Therapeutics, and a consultant for SOTIO Biotech Inc.; has ownership interest in Unum Therapeutics (now Cogent Biosciences), and ownership interests in Baxter International and Takeda Pharmaceutical Co. outside the submitted work; is an inventor multiple patents PCT/US2019/046550, PCT/US2019/050013, PCT/US2019/040346, and PCT/US2019/060287 related to the submitted work; and US 8,252,913, US 8,461,318, US 8,598,327, US 10,144,770, PCT/US2019/044512, PCT/US2018/015999, PCT/US2017/023064, PCT/US2010/023599-outside of the submitted work. SAE is an employee of BlueRock Therapeutics, a former employee of Unum Therapeutics, has ownership interest in Unum Therapeutics (now Cogent Biosciences), and is a co-inventor on patents filed related to this work. GTM is an employee of BlueRock Therapeutics, a former employee of Unum Therapeutics; has ownership interest in Unum Therapeutics (now Cogent Biosciences), and is a named inventor on many CAR-T patents. GJW is an employee of SOTIO Biotech Inc., a former employee of Unum Therapeutics; reports personal fees from Spring Bank Pharmaceuticals, Imaging Endpoints II, MiRanostics Consulting, Gossamer Bio, Paradigm, International Genomics Consortium, Angiex, IBEX Medical Analytics, GLG Council, Guidepoint Global, Genomic Health, Rafael Pharmaceuticals, SPARC-all outside this submitted work; has ownership interest in Unum Therapeutics (now Cogent Biosciences), and ownership interests in MiRanostics Consulting, Exact Sciences, Moderna, Agenus, Aurinia Pharmaceuticals, and Circulogene-outside the submitted work; and has issued patents: PCT/US2008/072787, PCT/US2010/043777, PCT/US2011/020612, and PCT/US2011/037616-all outside the submitted work and is an inventor on a patent filed related to this work. AJS is an employee of SOTIO Biotech Inc.; a former employee of bluebird bio; has ownership interest in bluebird bio, Pfizer, Regeneron, Exelixis, Bristol Myers Squibb, Celgene, and Alkermes-all outside the submitted work. All other authors have no other competing interests related to this work. This does not alter our adherence to PLOS ONE policies on sharing data and materials.

Figures

Fig 1. GPC3 target expression in human normal and tumor tissues by IHC.

(A) Immunohistochemical staining of normal healthy TMA for GPC3 expression. Representative images (20x magnification) of different organs with different levels of GPC3 are shown (B) IHC staining of tissues from hepatocellular carcinoma (HCC), squamous cell lung cancer (SCC), Merkel cell carcinoma (MCC) and liposarcoma patients for GPC3 expression. Representative images (20x magnification) show samples with different levels of GPC3 (and H-scores).

Fig 2. Co-expression of GPC3-targeted CAR with exogenous GOT2 in BOXR1030 T cells.

(A) Depiction of expression construct and domains for the GPC3 CAR and BOXR1030 used in studies. (B) Representative surface expression of the anti-GPC3 scFv following transduction, measured on day 9 by flow cytometry from a single donor. (C) Summarized CAR expression by flow cytometry for n = 5 healthy donors. (D) mRNA copy number of endogenous and exogenous (codon optimized) GOT2 measured by qPCR (n = 3 donors). (E) The frequency of CD45RA+ CD27+, CD45RA+ CD27-, CD45RA- CD27+, and CD45RA- CD27- T cells in the indicated subsets at baseline (n = 11 healthy donors; no stimulation). P-values were determined by paired t-test. Data represented as mean+/- standard deviation (SD).

Fig 3. AST enzymatic activity in BOXR1030 CAR T cells.

(A) Schematic describing the role of GOT2 in mitochondrial conversion of glutamate to aKG and OAA to aspartate. Image created with biorender.com. (B) Intracellular aspartate levels were measured using a plate-based colorimetric assay comparing BOXR1030 and control CAR T cells under non-stimulated conditions. (C) BOXR1030 and control CAR T cells were activated for 8 hours with GPC3+ Hep3B target cells and evaluated for AST activity. Data represented as mean +/- SD of two technical replicates.

Fig 4. In vitro BOXR1030 activity.

(A) Expression of GPC3 target antigen on target cell lines were measured by flow cytometry and represented as median fluorescence intensity (MFI). (B) Histogram plots of GPC3 MFI for selected cell lines. (C-F) T cell cytokine release of BOXR1030 and control CAR T cells was measured by ELISA following co-culture with target cell lines for 24 hours. Results for IFN-g (C) and IL-2 (D) are the averages across 3 donors, and error bars indicate SD. Results for TNF-a (E) and IL-17A (F) are for a single donor, and error bars indicate SD between technical replicates. (G) Target cell cytotoxicity was measured with a luciferase-based assay, and percent cytotoxicity was normalized to samples treated with untransduced T cells following 24 hours of co-culture (n = 3 donors, error bars indicate SD). (H) T cell proliferation was evaluated after a 7 day incubation with target cell lines. CAR+ T cell counts were measured by flow cytometry. (n = 3 donors, error bars indicate SD). (I) Intracellular cytokine levels were measured by flow cytometry in CD4+ and CD8+ T cells (n = 10, error bars indicate SEM).

Fig 5. In vitro activity of BOXR1030 under conditions simulating the solid tumor microenvironment.

BOXR1030 or control CAR T cells were repeat stimulated with GPC3+ target cell lines on day 0 and on day 3 in hypoxic conditions (A) or low glucose conditions (B). (A and B) The gMFI of CellTrace Violet (CTV) was measured for BOXR1030 and control CAR stimulated with target cells on day 7 in the indicated culture conditions and proliferation was plotted as 1/gMFI. (n = 5 for JHH7 stimulated conditions and n = 11 for Hep3B stimulated conditions, statistical analysis was performed using a 2-tailed paired t-test, and p-values

Fig 6. In vivo anti-tumor activity of BOXR1030 T cells.

(A) Plot on the right shows Hep3B tumor-bearing model (mean tumor volume 108.7±34.1mm3) NSG mice were treated with two weekly doses of 1 x 106 CAR+ control or BOXR1030 T cells each (total dose of 2 x 106 CAR+ cells) (dosing days indicated by arrows) and tumor volumes were measured over the course of 110 days. Plot on the left shows data for individual mice and plot on the right shows mean data. Mean tumor volume plots are discontinued when less than 50% of group is remaining. (B) Plot on the right shows JHH7 tumor-bearing mice (mean tumor volume 49.8±7.2mm3) were treated with two weekly doses of 5 x 106 CAR+ control or BOXR1030 T cells each (total dose of 10 x 106 CAR+ cells) (dosing days indicated by arrows) and tumor volumes were measured out to 50 days. Plot on the left shows data for individual mice and plot on the right shows mean data. Mean tumor volume plots are discontinued when less than 50% of group is remaining. (C) CAR+ T cells were measured in peripheral blood of Hep3G tumor bearing mice on days 15, 25, 40 and 60 post T cell treatment and data are reported as counts per ul of blood. (D) Percent PD1+TIM3+CD4+ and PD1+TIM3+CD8+ tumor infiltrating T cells were measured by FACS on days 7 and 14 following T cell administration. (E) Biodistribution of BOXR1030 T cells was measured in mouse tissues by qPCR at days 15 and 45 post treatment. Data are reported as BOXR1030 copies / 100ng DNA.