A flagellin-derived toll-like receptor 5 agonist stimulates cytotoxic lymphocyte-mediated tumor immunity

Abstract

Toll-like receptor (TLR) mediated recognition of pathogen associated molecular patterns allows the immune system to rapidly respond to a pathogenic insult. The "danger context" elicited by TLR agonists allows an initially non-immunogenic antigen to become immunogenic. This ability to alter environment is highly relevant in tumor immunity, since it is inherently difficult for the immune system to recognize host-derived tumors as immunogenic. However, immune cells may have encountered certain TLR ligands associated with tumor development, yet the endogenous stimulation is typically not sufficient to induce spontaneous tumor rejection. Of special interest are TLR5 agonists, because there are no endogenous ligands that bind TLR5. CBLB502 is a pharmacologically optimized TLR5 agonist derived from Salmonella enterica flagellin. We examined the effect of CBLB502 on tumor immunity using two syngeneic lymphoma models, both of which do not express TLR5, and thus do not directly respond to CBLB502. Upon challenge with the T-cell lymphoma RMAS, CBLB502 treatment after tumor inoculation protects C57BL/6 mice from death caused by tumor growth. This protective effect is both natural killer (NK) cell- and perforin-dependent. In addition, CBLB502 stimulates clearance of the B-cell lymphoma A20 in BALB/c mice in a CD8(+) T cell-dependent fashion. Analysis on the cellular level via ImageStream flow cytometry reveals that CD11b(+) and CD11c(+) cells, but neither NK nor T cells, directly respond to CBLB502 as determined by NFκB nuclear translocation. Our findings demonstrate that CBLB502 stimulates a robust antitumor response by directly activating TLR5-expressing accessory immune cells, which in turn activate cytotoxic lymphocytes.

Conflict of interest statement

Competing Interests: The authors have read the journal’s policy and have the following conflicts: Lyudmila Burdelya and Andrei Gudkov are paid consultants of Cleveland BioLabs, Inc. and Andrei Gudkov is a shareholder of Cleveland BioLabs, Inc. The other authors have no conflict of interest. This does not alter the authors’ adherence to all the PLOS ONE policies on sharing data and material.

Figures

Figure 1. CBLB502 enhances the ability of splenocytes to control tumor growth in vitro.

(A) SCCVII (a tumor line known to express TLR5 mRNA), A20 parental, luciferase-expressing, and RMAS parental and luciferase-expressing tumor cells were analyzed for TLR5 mRNA expression via RT-PCR. GAPDH was used as an internal loading control. (B) 2000 luciferase-expressing RMAS or A20 cells were cultured alone in 1 ml of complete media or co-cultured with 6×106 C57BL/6 or BALB/c splenocytes, respectively. At 0 hr, the cells were treated with 10 µl (10 µg/mL) CBLB502 or media diluent. Tumor burden was measured by bioluminescence imaging after 72 hrs. Two tailed t-tests were used to determine significance (*P<0.05, **P<0.01). Shown are representative data from three independent experiments with similar results.

Figure 2. CBLB502 stimulates NK cell- and perforin-dependent tumor immunity.

(A) Kaplan-Meier survival curve of WT C57BL/6 mice injected intravenously (IV) RMAS cells. Mice were treated with subcutaneous (SC) injections of either CBLB502 or PBS as described in Materials and Methods (PBS n = 8, CBLB502 n = 12). (B) Kaplan-Meier survival curve of WT C57BL/6 mice that were non-depleted or given either anti-CD8α antibody or anti-asialo GM1 antibody three days prior to IV injection with RMAS cells. PBS (n = 8), non-depleted CBLB502 (n = 12), NK depleted PBS (n = 8), NK depleted CBLB502 (n = 8), CD8 depleted PBS (n = 8), CD8 depleted CBLB502 (n = 8). (C) Peripheral blood was collected via eye bleeding 2 weeks after depletion with anti-asialo GM1 or anti-CD8α and the percentage of NK1.1+CD3− and CD3+CD8+ T cells, respectively, were analyzed (n = 8 mice/group). (D) Kaplan-Meier survival curve of WT CBLB502 (n = 27), WT PBS (n = 27), Prf1−/− CBLB502 (n = 15), Prf1−/− PBS (n = 14). All asterisk (*) represent statistical significance as determined by Log-rank (Mantel-Cox) test versus non-depleted PBS control group (*P<0.05). All data shown are representative of one of at least three individual experiments, or data combined from multiple experiments.

Figure 3. CBLB502 treatment stimulates CD8+ T cell-dependent tumor immunity.

(A) Tumor burden measured by bioluminescence imaging in BALB/c mice that were either non-depleted or given anti-CD8α antibody 3 days prior to IV inoculation with A20-luciferase expressing cells. Four different groups CBLB502 (n = 8), PBS (n = 8), CD8 depletion CBLB502 (n = 7) and CD8 depletion PBS (n = 8) were treated with PBS or CBLB502. (B) Kaplan-Meier survival curve of BALB/c mice from the experiment described in panel A. Data are combined from 2 of 3 individual experiments. (C) To confirm depletions of CD8+ T cells, peripheral blood was harvested via eye bleed day 16 after depletion and the percentage of CD3+CD8+ T cells in peripheral blood leukocytes was analyzed (n = 3–4 mice/group). Two tailed t-tests were performed to evaluate significance (**P<0.01). Shown as mean ± SD. Statistical analysis of survival was performed by Log-rank (Mantel-Cox) test (*P<0.05) with all groups being compared to PBS-treated non-depleted cohort.

Figure 4. CD11b+ CD11c− and CD11b+ CD11c+ accessory cells, but not NK or T lymphocytes, directly respond to CBLB502 treatment.

Nuclear translocation of the NFκB p65 subunit was used to evaluate TLR5 expression and functionality. Splenocytes were treated with 10 ng/mL TNF-α, 100 ng/mL LPS, 100 ng/mL CBLB502 or PBS for 1 hr. (A) Splenocytes were stained for CD4, CD8, or NK1.1 along with CD3. After cell surface staining, cells were permeabilized and stained intracellularly for NFκB p65. Target populations were then gated as CD3+CD4+, CD3+CD8+ and CD3−NK1.1+ and similarity score between DAPI and the FITC-labeled NFκB was measured. (B) Splenocytes were stained for CD11b, CD11c and NFκB as described in panel A. Target populations were then gated as CD11b+CD11c− or CD11b+CD11c+ to quantitate the similarity score. (C) Representative images at 40× magnification of NK1.1+CD3− cells and CD11b+ CD11c− cells treated with PBS, TNF-α, LPS, or CBLB502. Two tailed t-tests were used to determine significance versus PBS-treated control samples (*P<0.05). Data are representative from one of least three experiments.

Figure 5. CD80 and CD86 are up-regulated in vivo after CBLB502 treatment in a TLR5 dependent manner.

WT or TLR5−/− C57BL/6 mice were injected s.c. with 1 µg CBLB502, 1 µg flagellin, or equivalent volume of PBS. 24 h post-injection splenocytes were harvested and stained for flow cytometry with CD11b, CD11c, CD80 and CD86 antibodies. For LPS injections, 10 µg LPS was injected intraperitoneally and splenocytes were harvested 6 hours later and stained with the aforementioned antibodies immediately. (A) CD80 expression on CD11b+CD11c− and CD11b+CD11c+ cells from WT and TLR5−/− mice after indicated treatments. (B) CD86 expression on these same cell subsets. Two tailed t-tests were used to determine significance versus PBS-treated controls (**P<0.01, n = 3–6 mice per group). Data are representative from one of least three experiments.