Radiation therapy generates platelet-activating factor agonists

Abstract

Pro-oxidative stressors can suppress host immunity due to their ability to generate oxidized lipid agonists of the platelet-activating factor-receptor (PAF-R). As radiation therapy also induces reactive oxygen species, the present studies were designed to define whether ionizing radiation could generate PAF-R agonists and if these lipids could subvert host immunity. We demonstrate that radiation exposure of multiple tumor cell lines in-vitro, tumors in-vivo, and human subjects undergoing radiation therapy for skin tumors all generate PAF-R agonists. Structural characterization of radiation-induced PAF-R agonistic activity revealed PAF and multiple oxidized glycerophosphocholines that are produced non-enzymatically. In a murine melanoma tumor model, irradiation of one tumor augmented the growth of the other (non-treated) tumor in a PAF-R-dependent process blocked by a cyclooxygenase-2 inhibitor. These results indicate a novel pathway by which PAF-R agonists produced as a byproduct of radiation therapy could result in tumor treatment failure, and offer important insights into potential therapeutic strategies that could improve the overall antitumor effectiveness of radiation therapy regimens.

Keywords: antioxidants; cyclooxygenase type 2 enzyme; oxidized glycerophosphocholines; platelet-activating factor; radiation therapy.

Figure 1. Irradiation of melanoma cells generates PAF-R agonists

B16F10 cells were sham-irradiated (e.g. unirradiated) or exposed to ionizing radiation (IR; radiation therapy [RT]). A, B. Examples of intracellular Ca²⁺ mobilization responses in B16F10 cells that were exposed to 5 Gy IR. After 1h, lipids were extracted and PAF-R agonistic activities were measured with Ca²⁺ mobilization responses in FURA-2-labelled PAF-R-expressing KBP (A) or PAF-R-negative KBM (B) cells. KBM cells treated with 1μM endothelin 1 (ET-1) served as positive control. C. Dose-responsiveness of IR-induced PAF-R agonistic activity. B16F10 cells were treated with various IR doses and harvested 1h post-radiation. D. Time course of IR-generated PAF-R agonistic activity. B16F10 cells were irradiated with 10Gy and harvested at various times. Data are mean±SE and expressed as % maximum (1 μM CPAF) intracellular Ca²⁺ response in KBP cells from at least three separate experiments.

Figure 2. Effect of antioxidants on IR-mediated PAF-R agonists generation

B16F10 cells were preincubated with antioxidants, vitamin C (2.5 mM) or N-acetyl cysteine (5 mM) for 1 h following irradiation with 10 Gy.Lipids were extracted and PAF-R agonistic activities were measured with Ca²⁺ mobilization responses in FURA-2-labelled PAF-R-expressing KBP cells. The data are the mean ± SE percentage of peak intracellular calcium response (normalized to 1 μM CPAF). * Denotes statistically significant (P < 0.05) fold changes from vehicle, and antioxidants alone-treated cells.

Figure 3. Ionizing Radiation of B16F10 cells generates PAF-R agonistic activity as measured by IL-8 production in KBP cells

B16F10cells were sham-irradiated or irradiated with 10Gy IR. After 1 h of incubation, lipids were extracted and added to the culture of KBP cells. Following 6 h of incubation, supernatant was removed and IL-8 release was measured by ELISA to test for the PAF-R agonistic activity. In addition, KBP cells were incubated with various doses of PAF or vehicle and used as positive and negative controls. This experiment was performed in triplicate and repeated three times. * Denotes statistically significant (P < 0.05) changes from vehicle or sham-treated cells.

Figure 4. Effects of IR on PAF-R agonists production from several tumor types

A. Murine B16F10, LLC1, EL4 and human fibroblasts were treated with sham or 10Gy IR. B. WT mice were implanted with B16F10, LLC1, EL4 cells (0.5×10⁶) into both hindlimbs. Once tumors reached 10mm size, one tumor was treated with 5Gy IR or etoposide (ETOP; 36mg/kg) or melphalan (MELPH; 15mg/kg) and other tumor with sham or vehicle (100μl PBS). Following 1h incubation (A&B) cells and tumors harvested and lipids extracts were tested for PAF-R agonist activity. Human fibroblasts (in-vitro) and normal skin (in-vivo) treated ± IR or chemotherapeutic agents were used as controls. The data are mean±SE percentage of peak calcium response (normalized to CPAF) from 3 separate experiments (A) and 4-6 separate tumors (B).

Figure 5. Localized irradiation augments the growth of untreated B16F10 melanomas in a PAF-R-dependent manner

A. Groups of 6-7WT and Ptafr−/− (PAFR-KO) mice were implanted with PAF-R deficient B16F10 tumors on both the dorsal hind flanks (day 0). Six days later (and q 2-3days afterwards) left side tumors were sham-irradiated or irradiated with 5Gy of IR. The right side tumors left untreated (shielded). Tumor growth was measured over time and tumor volume was calculated. The data are mean±SE of tumor volume of untreated tumors. B. Tumor volume of untreated tumors at day 15 from sham and IR-treatment. There were significant differences in the growth of RT-treated *(P<0.05) or $ (P <0.1) but not sham-treated tumors (ns= non-significant). This experiment was repeated with additional 6-7 mice/group with similar results.

Figure 6. Role of COX-2 in RT-mediated augmentation of tumor growth

A. Groups of 6-13 WT mice were implanted with PAF-R deficient B16F10 tumors on both the dorsal hind flanks (day 0). Following tumor implantation (day 0) and every 3 days mice were i.p. injected with or without COX-2 inhibitor, SC-236 (200 ng). After 6 days of tumor implantation when mice developed palpable tumors (10mm), left side tumors were irradiated with 5Gy of IR and repeated at every q2-3 days and right side tumors left untreated. Tumor growth was assessed over time and tumor volume was calculated. B. A representative tumor volume data at day 15 (right side) in all the experimental groups. The data depicted are the mean ± SE of tumor volume of untreated tumors over time. Statistical significant differences were noted in the growth of RT-treated *(P<0.05) or $ (P<0.1) but not sham-treated tumors (ns denotes non-significant).