Long-Term, Sex-Specific Effects of GCRsim and Gamma Irradiation on the Brains, Hearts, and Kidneys of Mice with Alzheimer's Disease Mutations

Abstract

Understanding the hazards of space radiation is imperative as astronauts begin voyaging on missions with increasing distances from Earth's protective shield. Previous studies investigating the acute or long-term effects of specific ions comprising space radiation have revealed threats to organs generally considered radioresistant, like the brain, and have shown males to be more vulnerable than their female counterparts. However, astronauts will be exposed to a combination of ions that may result in additive effects differing from those of any one particle species. To better understand this nuance, we irradiated 4-month-old male and female, wild-type and Alzheimer's-like mice with 0, 0.5, or 0.75 Gy galactic cosmic ray simulation (GCRsim) or 0, 0.75, or 2 Gy gamma radiation (wild-type only). At 11 months, mice underwent brain and heart MRIs or behavioral tests, after which they were euthanized to assess amyloid-beta pathology, heart and kidney gene expression and fibrosis, and plasma cytokines. Although there were no changes in amyloid-beta pathology, we observed many differences in brain MRIs and behavior, including opposite effects of GCRsim on motor coordination in male and female transgenic mice. Additionally, several genes demonstrated persistent changes in the heart and kidney. Overall, we found sex- and genotype-specific, long-term effects of GCRsim and gamma radiation on the brain, heart, and kidney.

Keywords: Alzheimer’s disease; CNS; GCRsim; gamma; neurodegeneration; radiation; space radiation.

Figure 1

(A) Study timeline: At 4 months of age APPswe/PS1dE9 Tg and WT mice were transported to and from Brookhaven National Laboratory for GCRsim (0, 0.5, or 0.75 Gy) or gamma (0, 0.75, or 2 Gy) irradiation (n = 7–15 mice/group). A select group of mice underwent pre-irradiation brain and cardiac MRI scans at 3.5 months of age and 7 months post-irradiation. These mice received follow-up scans. Mice not involved in MRI studies participated in behavioral testing instead. At 12 months of age, mice were euthanized, and tissues were harvested for immunohistochemistry, histology, gene expression analyses, and ELISAs. Created with BioRender.com. (B) Survival curve of mice included in GCRsim analyses. (C) Survival curve of mice receiving gamma irradiation. Sham-irradiated controls are included in (A). Survival curve analyses in (B,C) utilized the log-rank test. (D,E) Differences in body weight in GCRsim (D) and gamma (E) mice at 12 months compared to 4 months of age. Body weight analyses were performed using repeated measures 2-way ANOVAs with Šídák’s multiple comparisons corrections. *: p < 0.05, **: p < 0.01, ****: p < 0.0001.

Figure 2

Differences in outcomes from behavioral tests (n = 4–11 mice/group). (A–D) Differences in the percentage of total distance (A,B) and time (C,D) spent in the novel arm of the spatial novelty Y maze (SNYM). (E,F) No significant differences were found in the context test. (G,H) Depressive-like behavioral changes were assessed using the tail suspension test. (I–L) Differences in the percentage of time (I,J) spent in the open arm of the elevated plus maze (EPM) as well as the number of entries (K,L). (M,N) Radiation and baseline sex effects on the distance traveled in the center of the open field. (O,P) The pre-pulse inhibition test (PPI) was used to assess sensorimotor gating and revealed several differences at a pre-pulse of 82 decibels (dB). (Q,R) Startle reactivity was compared at multiple white noise tones and differences found at each tone are indicated. Data were analyzed via 2-way ANOVAs followed up with 1-way ANOVAs with Tukey comparisons and planned, unpaired, 2-tailed t-tests between sham groups as necessary. Non-parametric data were analyzed using the Kruskal–Wallis test with Dunn’s multiple comparisons correction and/or Mann–Whitney U tests. Dotted lines indicated males and females were analyzed separately. #: p < 0.1, *: p < 0.05, **: p < 0.01, ***: p < 0.001.

Figure 3

Differences in outcomes from non-cognitive tests (n = 4–11 mice/group). (A–D) The total distance traveled and vertical counts during the open field test were compared across groups. (E–H) Strength and endurance were assessed using the grip strength and wire hang test, respectively. (I–L) The latency to fall (I,J) and percent improvement (K,L) on the rotarod test determined differences in motor coordination. Data were analyzed via 2-way ANOVAs followed up with 1-way ANOVAs with Tukey comparisons and planned, unpaired, 2-tailed t-tests between sham groups as necessary. Non-parametric data were analyzed using the Kruskal–Wallis test with Dunn’s multiple comparisons correction and/or Mann–Whitney U tests. Dotted lines indicate that males and females were analyzed separately. #: p < 0.1, *: p < 0.05, **: p < 0.01.

Figure 4

(A,B) Insoluble Aβx-40 (A) and Aβx-42 (B) levels were quantified using an MSD 4G8 Aβ-triplex ELISA. (C,D) Thio S staining (C) and quantification (D) of Aβ in the hippocampus. (E,F) 3A1 immunohistochemical staining (E) and quantification (F) of hippocampal beta-amyloid. Data were analyzed via 2-way ANOVAs followed up by 1-way ANOVAs with Tukey comparisons and planned (n = 7–15 mice/group), 2-tailed t-tests between sham groups as necessary. Non-parametric data were analyzed using the Kruskal–Wallis test with Dunn’s multiple comparisons correction and/or Mann–Whitney U tests. #: p < 0.1, ***: p < 0.001.

Figure 5

(A,B) Analyses of paired pre- and post-irradiation lateral ventricle volumes in GCRsim (A) and gamma (B) irradiated mice using T1 and T2 brain MRI scans (n = 3–4 mice/group). (C–F) Analyses similar to (A,B), however, with hippocampal (C,D) and cortex (E,F) volumes. Repeated measures 2-way ANOVAs with Šídák’s multiple comparisons corrections were used for analyses. #: p < 0.1, *: p < 0.05, **: p < 0.01, ***: p < 0.001.

Figure 6

MRI analyses of the heart (n = 3–4 mice/group). (A,B) Analyses of paired pre- and post-irradiation left ventricular mass index in GCRsim (A) and gamma (B) irradiated mice using short-axis FLASH MRI scans. (C–H) Analyses similar to (A,B), however, with ejection fraction (C,D), stroke volume index (E,F), and left ventricular wall thickness (G,H). Left ventricular wall thickness was calculated by averaging the measurements of the posterior, anterior, septal, and lateral walls. Left ventricular mass was calculated by multiplying the left ventricular wall thickness by 1.05 g/cm³, the relative density of myocardium and the conversion coefficient. The difference in volumes between left ventricular systole and diastole was used for stroke volume, which, when divided by the end diastolic volume and multiplied by 100, gave the ejection fraction. Repeated measures 2-way ANOVAs with Šídák’s multiple comparisons corrections were used for analyses. #: p < 0.1, *: p < 0.05, **: p < 0.01.

Figure 7

(A) Overview of the differentially expressed genes in heart and kidney on GCRsim-treated mice. All bars are referenced to APP/PS1 sham controls. Values between 0 and 1 denote negative fold changes. (B) Expression of VLCAD gene in the heart tissue of GCRsim-treated WT and APP/PS1 animals, compared to sham-irradiated controls. (C) Relative expression of Casp3 and GLUT4 in heart tissue of high-dose GCRsim-treated APP/PS1 animals, compared to sham controls. ns: no significance, *: p < 0.05, **: p < 0.01, ***: p < 0.001, ****: p < 0.0001.

Figure 8

Plasma levels of 7 inflammation-related cytokines measured via MSD ELISA, (A,B) IFN-γ, (C,D) IL-2, (E,F) IL-5, (G,H) IL-6, (I,J) IL-10, (K,L) KC-GRO, and (M,N) TNF-α (n = 7–15 mice/group). GCRsim affected levels of IL-2, IL-5, IL-6, and IL-10 in males. KC/GRO was both the only cytokine altered in females and the only one impacted by gamma irradiation. Data were analyzed via 2-way ANOVAs followed up with 1-way ANOVAs with Tukey comparisons and planned, unpaired, 2-tailed t-tests between sham groups as necessary. Non-parametric data were analyzed using the Kruskal–Wallis test with Dunn’s multiple comparisons correction and/or Mann–Whitney U tests. #: p < 0.1, *: p < 0.05, **: p < 0.01.