Elevating body temperature enhances hematopoiesis and neutrophil recovery after total body irradiation in an IL-1-, IL-17-, and G-CSF-dependent manner

Abstract

Neutropenia is a common side effect of cytotoxic chemotherapy and radiation, increasing the risk of infection in these patients. Here we examined the impact of body temperature on neutrophil recovery in the blood and bone marrow after total body irradiation (TBI). Mice were exposed to either 3 or 6 Gy TBI followed by a mild heat treatment that temporarily raised core body temperature to approximately 39.5°C. Neutrophil recovery was then compared with control mice that received either TBI alone heat treatment alone. Mice that received both TBI and heat treatment exhibited a significant increase in the rate of neutrophil recovery in the blood and an increase in the number of marrow hematopoietic stem cells and neutrophil progenitors compared with that seen in mice that received either TBI or heat alone. The combination treatment also increased G-CSF concentrations in the serum, bone marrow, and intestinal tissue and IL-17, IL-1β, and IL-1α concentrations in the intestinal tissue after TBI. Neutralizing G-CSF or inhibiting IL-17 or IL-1 signaling significantly blocked the thermally mediated increase in neutrophil numbers. These findings suggest that a physiologically relevant increase in body temperature can accelerate recovery from neutropenia after TBI through a G-CSF-, IL-17-, and IL-1-dependent mechanism.

Figure 1

Neutrophil numbers in the peripheral blood are increased when TBI is followed by heat treatment. Mice were left untreated (NT), treated with heat alone, with TBI alone, or with TBI followed 2 hours later with mild heating sufficient to raise their body temperature to approximately 39.5°C for 6 hours. The total number of leukocytes (A) and the total number of Ly6G⁺ cells (B-D) were calculated in C57BL/6 (A-B), BALB/c (C), and C3H/HeJ (D) mice treated with 3 Gy TBI with or without heat treatment. Neutrophil numbers were calculated using the percentage of Ly6G⁺ cells and the total leukocyte cell counts. (E) Total number of Ly6G⁺ cells was calculated in C57BL/6 mice treated with 6 Gy TBI with or without heat treatment. Each graph is representative of at least 3 separate experiments. n = 5 mice per group. Statistical analysis comparing TBI alone and TBI followed by heat treatment was performed: *P < .02 (Student t test).

Figure 2

Mild heating increased the number of neutrophils and granulocytic progenitors in the bone marrow after TBI. (A) Bone marrow was isolated from the femur and tibia of C57BL/6 mice that received 3 Gy TBI with or without heat treatment, and then erythrocytes were lysed and total number of bone marrow cells was quantified. (B) The percentage of Ly6G⁺ cells was determined by flow cytometry. Neutrophil numbers were calculated using the percentage of Ly6G⁺ cells and the total bone marrow cell counts. (C-F) C57BL/6 mice were left untreated (NT), treated with TBI alone, or with 3 Gy TBI followed 2 hours later with a heat treatment. At 12 and 48 hours after treatment, bone marrow was isolated from one femur and tibia from each mouse and filtered. Erythrocytes were lysed and total number of cells quantified. Number of HSCs (C), multipotent progenitor (D), granulocyte-macrophage progenitor (E), and pre-granulocyte-macrophage progenitor (F) cells were determined using the percentage of each cell population as determined by flow cytometry and the overall bone marrow cell count. (G-H) CFU assays were performed using 2 × 10⁴ bone marrow cells from mice given either 3 Gy (G) or 6 Gy (H) TBI with or without heat treatment. The bone marrow cells were mixed in methylcellulose with rmSCF, rmIL-3, and rhIL-6. On day 12, colonies were scored on coded plates for unbiased counts. Colony types were identified on a morphologic basis. The number of CFUs per femur was calculated. Each graph is representative of at least 2 separate experiments. n = 3-5 mice per group. †P < .04, compared with untreated mice. *P < .04, TBI alone mice versus TBI followed by heat treatment.

Figure 3

G-CSF is required for the thermally mediated acceleration of neutrophil recovery after TBI. G-CSF concentrations were determined by ELISA in the serum (A-B), intestinal lysates (50 μg; C-D), and bone marrow lysates (40 μg; E-F) of C57BL/6 mice given either 3 Gy (A,C,E) or 6 Gy (B,D,F) TBI. (G-H) Ly6G⁺ cells counts were performed in mice that have been given anti–G-SF neutralizing antibody (G) or isotype control antibody (H) given immediately after 3 Gy TBI and 2 hours after heat treatment. Hatched bars indicate mice that received antibody treatment. Each graph is a representative of at least 3 separate experiments. n = 3-5 mice per group. *P < .03, TBI alone group versus TBI followed by heat. †P < .04, TBI + heat + anti–G-SF group versus the TBI + heat group. n.d. indicates not detected.

Figure 4

Increased neutrophil recovery after heat treatment is dependent on IL-17. C57BL/6 or IL17ra^−/− mice were left untreated (NT), treated with heat treatment alone, TBI alone, or TBI followed 2 hours later with a 6-hour heat treatment. (A-B) Cell lysates from intestine of C57BL/6 mice that received either 3 Gy (A) or 6 Gy (B) TBI with or without heat treatment were collected, and ELISAs were performed using 50 μg lysate per well to determine concentrations of IL-17. (C) Ly6G⁺ cell counts were performed in IL17ra^−/− mice 7 and 14 days after TBI. (D) Serum from IL17ra^−/− mice that received 3 Gy TBI with or without heat treatment was collected, and ELISAs were performed to determine G-CSF concentrations. Each graph is a representative of at least 2 separate experiments. n = 5 mice per group. *P < .01, TBI alone group vs TBI followed by mild heating. n.d. indicates not detected.

Figure 5

IL-1 is important for the thermally mediated increase in neutrophil numbers after TBI. Cell lysates from the intestine of C57BL/6 mice that received either 3 Gy (A,C) or 6 Gy (B,D) TBI with or without heat treatment were collected, and ELISAs were performed using 50 μg lysate per well to determine concentrations of IL-1β (A-B) or IL-1α (C-D). (E) Blocking IL-1 activity with rmIL-1Ra (given immediately before 3 Gy TBI and again every 12 hours for 3 days) reduces the effect of heating. Ly6G+ cell counts were performed on mice that received saline (solid bars) or rmIL-1Ra (hatched bars) in no treatment, TBI alone and TBI + heat groups. Each graph is a representative of at least 2 Ly6G⁺ cell counts were performed on mice that received rmIL-1Ra immediately before 3 Gy TBI separate experiments. n = 5 mice per group. *P < .03, TBI alone group vs TBI followed by heat treatment. †P < .04, TBI + heat group vs TBI + heat + rmIL-1Ra group. n.d. indicates not detected.

Figure 6

Mild heating does not further increase the number of apoptotic cells in the intestine of mice after TBI. C57BL/6 mice were left untreated (NT) or treated with heat alone, 3 Gy TBI alone, or TBI followed 2 hours later with a 6-hour heat treatment. Intestine samples were collected 12 and 24 hours after treatment, and formalin-fixed, paraffin-embedded sections were stained for apoptosis. The number of apoptotic cells per field (when counted at 40×) was quantified at both 12 and 24 hours after treatment. Ten fields were counted per slide. Arrows indicate positive staining. n = 3 mice per group. Images are intestine samples from 12 hours after treatment (original magnification ×40).