A novel anti-B7-H3 chimeric antigen receptor from a single-chain antibody library for immunotherapy of solid cancers

Abstract

B7-H3 (CD276) has emerged as a target for cancer immunotherapy by virtue of consistent expression in many malignancies, relative absence from healthy tissues, and an emerging role as a driver of tumor immune inhibition. Recent studies have reported B7-H3 to be a suitable target for chimeric antigen receptor-modified T cell (CAR-T) therapy using CARs constructed from established anti-B7-H3 antibodies converted into single-chain Fv format (scFv). We constructed and screened binders in an scFv library to generate a new anti-B7-H3 CAR-T with favorable properties. This allowed access to numerous specificities ready formatted for CAR evaluation. Selected anti-human B7-H3 scFvs were readily cloned into CAR-T and evaluated for anti-tumor reactivity in cytotoxicity, cytokine, and proliferation assays. Two binders with divergent complementarity determining regions were found to show optimal antigen-specific cytotoxicity and cytokine secretion. One binder in second-generation CD28-CD3ζ CAR format induced sustained in vitro proliferation on repeat antigen challenge. The lead candidate CAR-T also demonstrated in vivo activity in a resistant neuroblastoma model. An empirical approach to B7-H3 CAR-T discovery through screening of novel scFv sequences in CAR-T format has led to the identification of a new construct with sustained proliferative capacity warranting further evaluation.

Keywords: B7-H3; CAR-T cell; neuroblastoma; pediatric cancer; phage display; solid tumor.

Graphical abstract

Figure 1

Anti-B7-H3 scFvs were identified that showed binding against plate bound and cell bound B7-H3 (A) Bacterial clones demonstrating an anti-B7-H3 response in screening were regrown and retested in triplicate. Following induction of scFv-myc production, bacterial supernatant was tested in an ELISA against recombinant B7-H3 or PBS as a negative control. A commercial anti-B7-H3 and the serum from the immunized mouse were used as positive controls and a secondary only (anti-myc) as a negative control (mean and SD, n = 3). (B) The structure of the scFv-myc and scFv-Fc proteins. (C) Jurkat cells were transduced with one of three isoforms of B7-H3: human 4IgB7-H3, human 2IgB7-H3, and artificial T-B7-H3. (D) The binding of scFv-Fc against different cell bound isoforms of B7-H3. Representative 1 of 2.

Figure 2

The anti-B7-H3 CAR-T cell shows T cell effector functions similar to anti-GD2 and anti-CD19 CAR-T cells (A) The second-generation CAR design used in this study incorporating the CD8 H/Tm and the CD28-CD3ζ endodomains. (B) Cr⁵¹ cytotoxicity assay of lead CAR-T cells against isogenic B7-H3 ± cell lines and LAN-1 cells. Comparison made with anti-GD2 CAR-T cells against isogenic GD2 ± cell lines and LAN-1 (mean and SD, n = 3; ns p ≥ 0.05, ∗∗∗∗p < 0.0001). (C) B7-H3 and GD2 antigen density of cell lines used in the study measured using a Quantbrite antigen quantification kit (BD Bioscience). (D) CAR-T cells were cultured with LAN-1, Kelly, or no antigen stimulus for 18 h. IL-2 and IFN-γ were measured in supernatant by ELISA (mean and SD, n = 3–5; ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001). (E) Antigen-specific cytokine response at the end of the tumor rechallenge assay at day 9 (24 h after second stimulation) and day 26 (24 h after the fourth stimulation). Mean and range of data are from two donors. Each sample analyzed once in duplicate.

Figure 3

TE9-28ζ CAR T cells show superior cytokine production compared with TE9-41BBζ CAR T cells (A) Diagram of the second-generation CD28-CD3ζ and 4-1BB-CD3ζ CAR constructs used in this study. (B) Transduction efficiency of TE9-28ζ (mean and range, n = 6) or TE9- 41BBζ (mean and range, n = 3). (C) CAR-T cells or untransduced cells were cultured with cells containing antigen targets (LAN-1 or Kelly) or no antigen targets (unstimulated) for 18 h. Cells were stained for CD107a, CD69, and CD25 (mean and range, n = 3; ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗∗p < 0.0001). (D) CAR-T cells or untransduced cells were cultured with LAN-1 or Kelly target cells or no antigen stimulus for 18 h. Cells were pelleted and supernatant used in ELISA compared with standard values of IL-2 or IFN-γ (mean and range, TE9-28ζ, untransduced, n = 6. TE9-BBζ, n = 3; ∗p ≤ 0.05, ∗∗∗∗p < 0.0001). (E) CARs and untransduced T cells were cultured with LAN-1 or Kelly targets, or US for 7 days then re-stimulated with fresh antigen targets and incubated for a further 24 h. IFN-γ and IL-2 production were measured using ELISA (mean and range, n = 3; ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001).

Figure 4

Second-generation CAR T cells with a CD28 H/Tm show superior cytokine production and proliferation in the presence of low antigen expression (A) A schematic of the second-generation TE9-CAR T cells with the CD8 and CD28 H/Tm. (B) Cytokine production by T cells transduced with second-generation CARs containing either a CD8 H/Tm or a CD28 H/Tm or untransduced. T cells were incubated overnight with different concentrations of plate bound B7-H3 and the supernatant analyzed for cytokine production (mean with range, n = 3; ∗∗p ≤ 0.01, ∗∗∗∗p < 0.0001). (C) The transduction efficiency of TE9-CD8 St and TE9 CD 28 H/Tm (mean with range, n = 6). (D) TE9 CAR T cells with either a CD28 H/Tm or CD8 H/Tm were incubated with LAN-1, Kelly, K562, or no antigen stimulus for 18 h. Supernatant was used to quantify cytokine production using ELISA (mean and range, n = 6; ∗∗p ≤ 0.01, ∗∗∗∗p < 0.0001). (E) After 7 days of co-culture, T cells were re-stimulated with fresh target cells or no antigen targets. Supernatant from 7-day co-cultures was used to quantify cytokine production using ELISA (mean and range, n = 4; ∗∗∗∗p < 0.0001). (F) TE9 CAR T cells with either a CD28 H/Tm or CD8 H/Tm were stained with CSFE and incubated with LAN-1, Kelly, K562, or no antigen stimulus for 7 days. Histograms show dilution of CSFE due to proliferation against different targets (representative 1 of 3). The ΔMFI (median fluorescence intensity) was calculated as the difference between the MFI of the test condition compared with the unstimulated untransduced control (mean and range, n = 3; ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001). (G) CAR T cells with the CD8 and CD28 H/Tm domains and untransduced T cells were evaluated for cytotoxicity against LAN-1 target cells in a chromium release assay (mean and SD, n = 3).

Figure 5

TE9 is comparable with two previously described anti-B7-H3 scFvs in in vitro models (A) CD28-CD3ζ CAR T cells were produced with the anti-B7-H3 scFvs MGA271 and 376.96. (B) Cytotoxicity of anti-B7-H3 CAR T cells or untransduced T cells measured using a chromium release assay against LAN-1 target cells (mean with SD, n = 3; ∗∗∗∗p < 0.0001). (C) CAR T cells or untransduced T cells were cultured with LAN-1, Kelly, or no target cells (unstimulated) for 18 h. ELISA of cell supernatant was used to measure IFN-γ and IL-2 production (mean with SD, n = 6; ∗p ≤ 0.05, ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001). (D) T cells or untransduced T cells were cultured with LAN-1, Kelly, or no target cells (unstimulated) for 18 h. Cells were stained for T cell activation markers CD25, CD69, and the degranulation marker CD107a as a proxy for cytotoxicity. The MFI was measured using flow cytometry (mean with SD, n = 3; ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001). (E) CAR T cells or untransduced T cells were stained with CSFE and cultured with LAN-1, Kelly, or no target cells for 7 days. At the end of 7 days, the MFI was measured using flow cytometry. The histogram is representative 1 of 3. The bar chart shows the change in MFI from untransduced, unstimulated cells, which is taken as 0 (mean with SD, n = 3; ∗p ≤ 0.05, ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001). (F) CAR T cells or untransduced T cells were cultured with LAN-1, Kelly, or no target cells for 7 days then re-stimulated with fresh antigen targets. Twenty-four hours after re-stimulation, cell supernatant was collected for quantification of IFN-γ and IL-2 production using ELISA (mean with SD, n = 3; ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001).

Figure 6

TE9-28ζ shows superior expansion and cytokine production in long-term assays compared with GD2-28ζ CAR T cells were transduced with TE9-28ζ (TE9), TE9-28-ILR2ζ (TE9-ILR2), or GD2-28ζ (GD2). (A) A schematic of the CAR T cells used in this study. (B) CAR T cells or untransduced cells were cultured with either LAN-1, Kelly, or no antigen stimulus. Each week, cells were given a fresh antigen stimulus, cultured for a further 24 h then analyzed. The production of IFN-γ and IL-2 was determined using ELISA after each antigen stimulus (mean and range, n = 4; ∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001). (C) The proliferation as measured by the fold change of CD3+ cells measured using flow cytometry. The significance is shown between cell numbers on day 28 (mean with range, n = 4; ∗∗p ≤ 0.01, ∗∗∗p ≤ 0.001, ∗∗∗∗p < 0.0001).

Figure 7

In vivo testing of TE9-28ζ, TE9-BBζ, and ⍺GD2-28ζ Mice treated with TE9-28ζ show increased survival and reduced tumor growth compared with other groups. (A) Experiment plan. (B) IVIS imaging. (C) Survival curve showing percentage of overall survival: analyzed and found to be significant using log-rank (Mantle-Cox) test (∗∗p ≤ 0.01). (D) Region of interest (ROI) measurements taken at weekly IVIS imaging (photons/second/cm²/sr). (E) Main graph: largest tumor diameter (mm) measured with digital calipers. Insert: representation of tumor size at day 50 (mean, n = 6; ∗p ≤ 0.05, ∗∗p ≤ 0.01). (F) Antigen expression after treatment as illustrated by the MFI of fluorophores used to stain antigens in tumor samples (mean and range, mice in untransduced, and anti-GD2-28ζ groups, n = 6. Mice in TE9-28ζ group n = 5).