Improving the ex vivo expansion of human tumor-reactive CD8 + T cells by targeting toll-like receptors

Abstract

Toll-like receptors (TLRs) are important pattern recognition receptor(s) known to mediate the sensing of invading pathogens and subsequent immune responses. In this study, we investigate whether TLRs could be explored for the preparation of human CD8⁺ T cell products used in adoptive cell therapy (ACT). Following characterization of TLRs expression on human CD8⁺ T cells, we screened TLR-specific agonists for their ability to act in concert with anti-CD3 to stimulate the proliferation of these cells and corroborated the observed co-stimulatory effect by transcriptional profiling analyses. Consequently, we developed an optimal formulation for human CD8⁺ T cell amplification by combining CD3/CD28 antibody, interleukin 7 (IL-7), interleukin 15 (IL-15), and three agonists respectively targeting TLR1/2, TLR2/6, and TLR5. This new formulation performed better in amplifying PD-1+CD8⁺ T cells, a potential repertoire of tumor-reactive CD8⁺ T cells, from tumor patients than the conventional formulation. Importantly, the expanded CD8⁺ T cells showed restored functionality and consequently a robust anti-tumor activity in an in vitro co-culturing system. Together, our study established the utility of TLR agonists in ex vivo expansion of tumor-targeting CD8⁺ T cells, thus providing a new avenue toward a more effective ACT.

Keywords: adoptive cell therapy; agonists; cell proliferation; human PD1+ CD8+ T cells; toll-like receptors; tumor cell killing.

FIGURE 1

Co-stimulation with TLR1/2, TLR2/6, and TLR5 agonists enhances the proliferative response of human blood-derived CD8⁺ T cells to TCR triggering. (A) Expression profiles of surface TLRs in resting CD8⁺ T cells within human blood mononuclear cells (PBMCs). The T cell subpopulations of human PBMCs were identified by staining the cells with antibodies against CD3, CD4, and CD8; the TLR expression was measured by flow cytometry and expressed as the percentage of positively stained cells. Shown is aggregated data of 12 samples. (B and C) Co-stimulatory effects of different TLR agonists on the proliferation of human PBMC-derived CD8⁺ T cells. CD8+T cells were purified from human PBMCs, stained with eFluor 670, and then stimulated with ant-CD3 beads, anti-CD3/CD28 beads, or anti-CD3 beads coupled with the indicated TLRs agonists. Cells were harvested 7 days later and dilution of eFluor 670 was determined by flow cytometry to measure cellular proliferation, with representative flow cytometry plot and data summary (n = 3) shown in (B) and (C), respectively. (D,E) Determination of optimal dosages of TLR agonists required for co-stimulation of ex vivo proliferation of human CD8⁺ T cells. CD8+T cells purified from human PBMCs were stimulated with anti-CD3 beads in presence of various concentration of TLR1/2 agonist, TLR2/6 agonist, or TLR5 agonist after labelling with eFluor 670. Flow cytometry analyses were performed on day 4 and 6 after stimulation to determine the cellular proliferation through assessment of eFluor 670 dilution. Shown is aggregated data of 3 samples.

FIGURE 2

Co-stimulation with a combination of TLR1/2, TLR2/6, and TLR5 agonists enhances the transcriptional induction of proliferative-related genes in human blood-derived CD8⁺ T cells responding to TCR triggering. CD8+T cells purified from human PBMCs were stimulated with anti-CD3 beads in the presence or absence of a combination of TLR1/2 agonist, TLR2/6 agonist, and TLR5 agonist. Cells were harvested 24 h later for extracting total RNA, which was subjected to RNA sequencing. The differentially expressed genes showing >2-fold change between minus and plus TLR agonists groups were clustered based on Gene Ontology (GO) enrichment analysis (A). Shown in (B) is the heatmap of representative proliferation-related genes that were identified to be co-stimulated by TLR agonists. H1 to H3 represent samples from three healthy individuals.

FIGURE 3

TLR agonists is an advantageous supplement benefiting the ex vivo expansion of human blood-derived PD1+ CD8⁺ T cells. (A) Exploration of optimal combination of TLR agonists in facilitating the ex vivo expansion of human blood-derived PD1+ CD8⁺ T cells. PD1+ CD8⁺ T cells isolated from blood samples of health individuals were continuously cultured for 21 days in medium containing anti-CD3/anti-CD28 beads and IL-7/IL-15, either alone or in the presence of different combination of TLR1/2, TLR2/6, and TLR5 agonists. Cell counting was performed at various time points throughout the culture period to generate the cell expansion curve. (B) Expansion curves of PD1+ CD8⁺ T cells derived from blood samples of cancer patients with different tumor types. Among the total of fifteen samples, seven were with lung cancer, four with pancreatic cancer, and the rest four with other types of cancers. Five samples from healthy individuals were also included as controls.

FIGURE 4

Human blood-derived PD1+ CD8⁺ T cells expanded by the new TLR agonists-aided method demonstrated strong effector and anti-tumor capacities. (A) Expression of co-stimulatory molecules (ICOS and 4-BB) and co-inhibitory molecules (CTLA-4 and PD-1) on the expanded PD1+ CD8⁺ T cells, assessed by flow cytometry. Shown is aggregated data of five samples. (B) Assessments of functional capacity of expanded PD1+ CD8⁺ T cells as effector cells. The expanded PD1+ CD8⁺ T cells were stimulated with anti-CD3/anti-CD28 beads in the presence of anti-CD107a, followed by treatment with secretion inhibitor brefeldin A for 5 h, and subsequently harvested for intracellular staining with markers granzyme B, IFN-γ and IL-2 before analyzing by flow cytometry (n = 5). (C) Tumor cell killing activity of expanded PD1+ CD8⁺ T cells. A549 cells were seeded at 1×10⁴ cells per well and incubated for 24 h before adding 1×10⁵ PBMC or the expanded CD8⁺ T cells. The co-culture was monitored for A549 cell index by RTCA system for 45 h. Shown are results of two representative samples.