The efficacy of radiotherapy relies upon induction of type i interferon-dependent innate and adaptive immunity

Abstract

The most widely held explanation for the efficacy of local radiotherapy (RT) is based on direct cytotoxicity to cancer cells through the induction of lethal DNA damage. Recent studies have shown that local ablative radiation of established tumors can lead to increased T-cell priming and T-cell-dependent tumor regression, but the underlying mechanism remains unclear. Here, we describe an essential role for type I IFN in local RT-mediated tumor control. We show that ablative RT increases intratumoral production of IFN-β and, more surprisingly, the antitumor effect of RT is abolished in type I IFN nonresponsive hosts. Furthermore, the major target of RT-induced type I IFN is the hematopoietic compartment. RT drastically enhances the cross-priming capacity of tumor-infiltrating dendritic cells (TIDC) from wild-type mice but not type I IFN receptor-deficient mice. The enhanced cross-priming ability of TIDCs after RT was dependent on autocrine production of type I IFNs. By using adenoviral-mediated expression of IFN-β, we show that delivery of exogenous IFN-β into the tumor tissue in the absence of RT is also sufficient to selectively expand antigen-specific T cells leading to complete tumor regression. Our study reveals that local high-dose RT can trigger production of type I IFN that initiates a cascading innate and adaptive immune attack on the tumor.

Figure 1. Local Radiation Therapy increases tumor infiltration by innate immune cells

a) Flow cytometric analysis of CD45⁺ hematopoietic cells infiltrating untreated B16F1 tumors or tumors treated with 25 Gy local RT at 3 and 5 days post-RT. b) Surface staining of infiltrating CD45⁺ cells for the myeloid marker CD11b and DC marker CD11c. Gated percentages represent the proportion of CD11b⁺CD11c⁺ cells in the parental CD45⁺ gate (shown in a). c) Bar graph of data in (a) (*p=0.0432 for No RT vs. RT D+3, p=0.0101 for No RT vs. RT D+5). d) Bar graph of the data in (b) (*p=0.0182, **p=0.0031). Values were calculated by the formula (%CD11c⁺ × %CD45⁺) for each sample individually and then averaged for the three samples in each group. Error bars represent the standard deviation among the three animals in each group (n.s. = not significant).

Figure 2. TIDC Function is altered by local RT

a) C57Bl/6 mice were inoculated with 5×10⁵ B16-SIY tumor cells. 15 day established tumors received 25 Gy of local RT or were left untreated. CD11c⁺ cells were isolated from tumors of 3 mice per group and pooled for analysis as described in the Materials and Methods. Isolated TIDCs cells were cocultured with naïve 2C TCR Tg cells under the indicated conditions and T cell proliferation assessed by tritiated-thymidine incorporation. Error bars represent the standard deviation among three triplicate wells (*p= 0.0167). b) Surface staining of maturation markers expressed by TIDCs from treated and untreated tumors at 72 hours post-RT.

Figure 3. Radiation therapy increases IFN-β within the tumor microenvironment

a). Real-time PCR analysis of IFN-β mRNA levels from whole tumor RNA of established B16F10 tumors (16–20 days) that were untreated or received 20 Gy of local RT. Data shown are from RNA extracted at 6 hr post-RT (*p=0.0123). b). ELISA for IFNβ protein levels at 48 hrs post-RT utilizing whole tumor homogenate (*p=0.0375). c). Time course analysis of IFN-β mRNA levels by RT-PCR at the indicated time points. CD45⁺ and CD45⁻ cells were sorted from established B16F10 tumors treated with 20 Gy of local RT at the indicated time points followed by RNA extraction and RT-PCR. Data shown are representative of three experiments with similar results.

Figure 4. The therapeutic response to RT is dependent on host responsiveness to type I IFNs

a) Tumor growth curves for WT and IFNAR1 KO mice bearing established B16F10 tumors that were either untreated or received local RT. Tumor growth was plotted starting from the initial dose of RT (n=6–9 mice pooled). b) WT mice were lethally irradiated and reconstituted with either WT or IFNAR1 KO bone marrow (BM). Established tumors received ablative local RT (15 Gy) on day 14, day 15, and day 16. (n= 7–8/group, **p=0.0011 for RT groups WT >WT vs. IFNAR KO > WT at day 20 post treatment). c) Same as (b) except, IFNAR1 KO mice were also used as recipients of WT BM. d) B6/RAG1^−/− mice were reconstituted with purified, polyclonal T cells from WT or IFNAR1 KO mice according to procedures detailed in the Materials and Methods. Tumors were treated with 25 Gy of local RT (n=4 mice/group). Data shown are representative of at least two experiments with similar results.

Figure 5. Local ablative RT endows TIDC with T-cell stimulatory capacity in WT mice but fails to generate functional TIDC in IFNAR KO mice

a) CD11c⁺ cells were isolated from tumors (a) and draining lymph nodes (d) of 3 mice/group and pooled for analysis. Isolated CD11c⁺ cells were cocultured with naïve 2C T cells under various conditions and T cell proliferation assessed. Error bars represent the standard deviation among three triplicate wells. b,c) Cytokine concentrations in supernatents from in vitro culture with TIDCs. Levels of IFNγ (b) and TNFα (c) are shown (*p= 0.0192). Data shown are representative of two experiments with similar results. (Unless otherwise indicated ***p<0.001, **p<0.01; individual p values are available where noted above)

Figure 6. Tumor infiltrating myeloid cells produce autocrine IFN-β that is independent of TRIF signaling

a) Gating strategy used to purify infiltrating myeloid cells from tumors for Real-Time PCR analysis of IFNβ expression. Cells were sorted into three populations, (i)CD45⁻, (ii)CD45⁺/CD11b⁺/CD11c⁺, and (iii)CD45⁺/CD11b⁺/CD11c⁻. Expression of Gr-1 is shown for both the CD11c⁺ and CD11c⁻ fractions. b) RT-PCR analysis of IFN-β expression in the three sorted populations depicted in (a) (n.s. p=0.1741). c) Established B16.SIY tumors in WT and TRIF^−/− mice were treated with 20Gy local RT and tumor growth was monitored (n=5).

Figure 7. Ad-IFN-β promotes preferential expansion of tumor Ag-specific cells

a) Established B16-SIY tumors were treated with 2×10¹⁰vp ad-null or ad-IFN-β on day 13 and 15 and tumor growth was monitored (n=4). b,c) WT mice bearing day 12 established B16-SIY tumors were adoptively transferred with a mixture of CFSE labeled 2C and OT-I/Thy1⁺ T cells. b) Representative FACS plot showing the frequency of Ag-specific 2C cells vs. non-specific OT-I T cells. c) Pooled data showing the ratio of 2C to OT-I T cells (n=5–7), (*p=0.0128). d) Specific lysis of SIY-loaded target cells (**p=0.0015). Data shown are representative of two experiments with similar results.