Lymphoma B-cell responsiveness to CpG-DNA depends on the tumor microenvironment

Abstract

Background: Toll-like receptor (TLR) agonists have important properties that can be exploited for immunotherapy against tumors. Locally injected immunostimulatory oligodeoxynucleotides containing CpG motifs (CpG-ODNs), which are TLR9 agonists, have shown promise in cancer models. Several studies have demonstrated that these motifs have immunologic effects similar to those of bacterial DNA and can stimulate monocytes, macrophages, dendritic, and B cells, which then produce several proinflammatory cytokines. However, these CpG-ODNs appear to produce opposite effects on tumor B cells.

Methods: In this study, we investigated the direct effects of a murine class B CpG (1826) ODNs on lymphoma B cells in vitro and in vivo, using mouse models of non-Hodgkin B lymphomas developing in immunoprivileged sites, specifically the brain and the eye, and in subcutaneous sites.

Results: In vitro, CpG-ODNs produced antiproliferative and proapoptotic effects on lymphoma B cells. In vivo, it had an antitumor effect when injected into tumors in murine models of subcutaneous lymphoma (SCL) and primary cerebral lymphoma (PCL). However, its intravitreal administration into a primary intraocular lymphoma (PIOL) mouse model did not produce an antitumor effect. In vitro experiments using supernatant from mouse PIOL samples demonstrated that the PIOL molecular microenvironment inhibits the antiproliferative effect of CpG-ODNs on lymphoma B-cells.

Conclusions: Responsiveness to CpG stimulation differs in subcutaneous, cerebral, and ocular tumors, according to the tumoral and molecular microenvironment, and this should be considered for further therapeutic approaches.

Figure 1

CpG inhibits cell proliferation and induces apoptotic death of A20.IIA lymphoma cells in vitro. (A) Flow cytometric analysis of TLR9 expression by A20.IIA cells after anti-TLR9 Ab staining (filled histogram), overlaid with isotype control (gray line). (B) CpG inhibits the proliferation of A20.IIA cells in vitro. 10⁴ A20.IIA cells were stimulated for 72 hours with various concentrations of CpG or control ODNs ranging from 0.0003 to 30μg/mL or with medium alone. The incorporation of the [³H] thymidine was measured by a scintillation counter. *P < 0.05; **P < 0.01. The data shown are representative of 1 of 3 experiments. (C) CpG induces apoptotic cell death of A20.IIA cell line. Cells were incubated for 72 hours with CpG or control ODNs at 3 and 30 μg/mL, or medium alone. The percentage of AnnV/PI positive cells was determined by flow cytometric analysis. ***P < 0.001.

Figure 2

CpG-ODNs decrease the burden of subcutaneous and cerebral tumors but fail to induce PIOL regression. The 2-tumor-site SCL model: (A) Representative bioluminescence images of SCL treated with control ODNs (upper panel) and CpG-ODNs (lower panel). The mice were injected with 5x10⁶ A20.IIA-GFP-luc2 cells. Treatment was injected in situ when the tumor reached 0.5 to 0.7 cm in diameter. (B) Flow cytometric analysis of GFP⁺ CD19⁺ tumor cells, 7 days after the end of CpG-ODN administration in right tumors compared to left (untreated) tumors. PCL lymphoma model: (C) Representative bioluminescence images of PCL mice treated with control ODNs (upper panel) and CpG-ODNs (both lower panels), showing 2 different profiles of responsiveness to CpG motifs. The mice were injected with 5x10⁴ A20.IIA-GFP-luc2 cells and treated one week after tumor inoculation. (D) The percentage of CD19⁺ GFP⁺ tumor cells, as determined by flow cytometry, in the brain of mice treated with CpG at 60 μg/2μL, in comparison with PBS 1X (Control)-injected mice (n = 5 per group). PIOL lymphoma model: (E) Representative bioluminescence images of PIOL mice treated with control ODNs (upper panel) and CpG-ODNs (lower panel). The mice were injected with 10⁴ A20.IIA-GFP-luc2 cells. CpG-ODN treatment was administered on day 0 intravitreously. (F) Flow cytometric analysis of the percentage of GFP⁺CD19⁺ tumor cells in PIOL-inoculated right eyes (n = 14 per group). A week after tumor injection, right eyes of mice were treated intravitreally by 20 μg/2 μL of mice injected with CpG-ODNs or PBS 1X (Control). Results obtained from 2 independent experiments were pooled. Statistical test: Mann–Whitney; NS: not significant.

Figure 3

PIOL supernatant counteracts in vitro antiproliferative effect of CpG-ODNs on A20.IIA malignant B cells. 10⁴ cells were stimulated for 72 hours with various concentrations of CpG or control ODNs in concentrations ranging from 0.003 to 60 μg/mL or with medium alone and with the presence of supernatant from (A) PBS 1X injected eyes (PIE), (B) SCL, (C) PCL, or (D) PIOL. The incorporation of the [³H] thymidine was measured by a scintillation counter. *P < 0.05; **P < 0.01. The data shown are representative results from 1 of 3 experiments.

Figure 4

Soluble molecule present in PIOL but not in normal ocular microenvironment is able to abrogate in vitro effect of CpG-ODNs in a dose-dependent manner. 10⁴ cells were stimulated for 72 hours with CpG or control ODNs at 30 μg/mL and in the presence of several diluted doses of control supernatant (PIE) or PIOL supernatant (1X, 1/20, 1/35, 1/50, 1/75, 1/100, 1/200, 1/500). The incorporation of the [³H] thymidine was measured by a scintillation counter. *P < 0.05; **P < 0.01.

Figure 5

The PIOL microenvironment did not modify TLR9 expression or internalization of CpG-ODNs by tumor cells. (A) 10⁴ A20.IIA cells were incubated with PIOL or PIE supernatant. 3 days later, cytometric analysis was performed of TLR9 expression by cells incubated with PIOL supernatant, overlaid with isotype control and compared to TLR9 expression by cells incubated with PIE supernatant or medium alone. (B) 10⁴ A20.IIA cells were incubated for 24 hours with medium alone or with PIOL or PIE supernatant and in the presence or absence of FITC-labeled CpG-ODNs at 3 μg/mL. FITC expression by A20.IIA tumor cells was analyzed by flow cytometry.