Psychosocial Stress and Age Influence Depression and Anxiety-Related Behavior, Drive Tumor Inflammatory Cytokines and Accelerate Prostate Cancer Growth in Mice

Abstract

Prostate cancer (PCa) prevalence is higher in older men and poorer coping with psychosocial stressors effect prognosis. Yet, interactions between age, stress and PCa progression are underexplored. Therefore, we characterized the effects of age and isolation combined with restraint (2 h/day) for 14 days post-tumor inoculation on behavior, tumor growth and host defense in the immunocompetent, orthotopic RM-9 murine PCa model. All mice were tumor inoculated. Isolation/restraint increased sympathetic and hypothalamic-pituitary-adrenal cortical activation, based on elevated serum 3-methoxy-4-hydroxyphenylglycol/norepinephrine ratios and corticosterone levels, respectively. Elevated zero maze testing revealed age-related differences in naïve C57Bl/6 mice, and increased anxiety-like behavior in tumor-bearing mice. In open field testing, old stressed mice were less active throughout the 30-min test than young non-stressed and stressed, and old non-stressed mice, suggesting greater anxiety in old stressed mice. Old (18 month) mice demonstrated more depression-like behavior than young mice with tail suspension testing, without effects of isolation/restraint stress. Old mice developed larger tumors, despite similar tumor expression of tumor vascular endothelial growth factor or transforming growth factor-beta1 across age. Tumor chemokine/cytokine expression, commonly prognostic for poorer outcomes, were uniquely age- and stress-dependent, underscoring the need for PCa research in old animals. Macrophages predominated in RM-9 tumors. Macrophages, and CD4⁺ and CD4⁺FoxP3⁺ T-cell tumor infiltration were greater in young mice than in old mice. Stress increased macrophage infiltration in old mice. Conversely, stress reduced intratumoral CD4⁺ and CD4⁺FoxP3⁺ T-cell numbers in young mice. CD8⁺ T-cell infiltration was similar across treatment groups. Our findings support that age- and psychological stress interacts to affect PCa outcomes by interfering with neural-immune mechanisms and affecting behavioral responses.

Keywords: IL-9/IL-17 balance; aging; anxiety/depression-related behavior; psychosocial stress; tumor immunity.

Graphical Abstract

Figure 1

Experimental Design. Experimental timeline of events is illustrated, including mice arrival (triangle), accommodation to vivarium conditions, pre- and post-treatment behavioral testing (yellow/orange bars, respectively), tumor inoculation (diamond), collection of blood by retroorbital bleeding (↓), and tissue collection for endpoint assessments (circle).

Figure 2

Stress Most Affected Body and Tumor Weights in Old Mice and Increased Young and Old Spleen Weight. Mean body weights ± SEM in g (n=16-25 per group) did not differ across time post-RM9 tumor inoculation between age-matched groups (A) or after collapsing data across age (B). (C) However, there was a greater mean change (Δ) in body weight (p<0.001) from baseline to study endpoint in old stressed mice compared with old non-stressed or young stressed mice. (D) Mean spleen weights (mg) were stress-dependently increased in young and old mice (p<0.05). (E) Tumor weights (mg) were greater (p<0.01) in old than in young stressed mice. *p < 0.05, **p < 0.01, ***p < 0.01.

Figure 3

Greater Anxiety-like Behavior in Young Mice in Zero Maze, and Post-treatment than at Baseline, but No Difference in Time Immobile on Tail Suspension Test. Behavioral effects of restraint stress and isolation in young and old tumor-bearing C57BL/6 mice. (A) Young mice spent significantly more time in the dark post-tumor inoculation compared with old mice (p<0.0001). Stress did not significantly alter the time spent in the dark for either young or old mice. Young mice spent significantly more time in the dark compared with old mice overall (p<0.0001). (B) When groups were collapsed, all mice spent significantly more time in the dark post-treatment compared with baseline (p<0.0001). Data are expressed as the mean % of time spent in the dark ± S.E.M. (C) Old non-stressed mice were significantly less active throughout the test compared with old stressed mice (p<0.05) and with young mice regardless of being stressed or non-stressed (p<0.05). Data are expressed as the mean % spent moving ± S.E.M. for each treatment group for each 3 min increment over the 30 min time period for the OFT (n=17-26 mice per group). (D) There were no group differences in percent time immobile on TST. *p < 0.05, ***p < 0.001.

Figure 4

(A) Increased HPA and SNS Activation in Stressed Old Mice. Mean circulating corticosterone levels (ng/ml ± S.E.M) from young and old non-stressed and stressed mice at baseline and D15 post-tumor inoculation. Old stressed mice had a significant increase in corticosterone levels post-treatment than at baseline. (B) After data are collapsed across age, Stressed mice had significantly higher corticosterone levels than Non-stressed mice (**p < 0.02). Data are expressed as corticosterone in ng/ml ± S.E.M. *Daily restraint stress began1 day after tumor inoculation. (C) Corticosterone expressed in ng/ml was higher (*p < 0.05) post-treatment compared with baseline levels in stressed compared with non-stressed mice. (D, E) Tumor expression of NE and EPI (ng/mg) were similar in all treatment groups. (F, G) However, in old mice, prostate dopamine, a precursor in NE synthesis, was lower (**p < 0.01) in stressed than non-stressed mice, and prostate MHPG, a metabolite of NE degradation, was elevated in old stressed mice compared with old non-stressed or young stressed mice (**p < 0.01 or *p < 0.05, respectively). (H) Likewise, MHPG/NE ratios, used to estimate NE turnover, was also higher in prostate tumors from old stressed mice than from old non-stressed or young stressed mice (**p < 0.01 or *p < 0.05, respectively). *p < 0.05, **p < 0.01.

Figure 5

Stress Promoted Tumor Proliferation and Apoptosis, but Age-dependent Infiltration of Immune Cell Subsets and Vascularization Support Age-dependent Anti-tumor Defense Mechanisms. Quantitation of immunostaining for proliferation (A), apoptosis (B), angiogenesis (C), F4/80⁺ macrophages, and CD8⁺, CD4⁺, and CD4⁺FoxP⁺ cells (D–G, respectively) in young (open bar) and old (gray bar) non-stressed and stressed mice D14 after orthotopic prostate tumor inoculation. (A) Stress increased prostate tumor expression of Ki67 in young and old mice compared with non-stress age-matched mice. (B) Conversely, CD95 expression was lower in RM-9 tumors from stressed than in non-stressed young and old mice. (C) CD31 immunoreactivity was higher in stressed than non-stressed old mice, but in young mice expression was greater in non-stressed than the stressed group. (D) F4/80⁺ tumor macrophages were fewer (*p < 0.05) in old than young non-stressed mice, but were higher in stressed than non-stressed old mice. (E) CD8⁺ Tc cell numbers were similar in non-stressed mice, but were higher in old than young in stressed mice. Moreover, more (p<0.05) CD8⁺ cells were present in old than young stressed mice. (F) CD4⁺, including those expressing FoxP3⁺ (G), were lower in old than young non-stress mice, and in young stressed than non-stressed mice. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 6

Growth Factors and Chemokines Expression in Prostate Tumors. (A, B) VEGF and TGF-β were highly expressed in RM-9 tumors, regardless of age or treatment group. The gray bar indicates the range of VEGF (A) and TGF-β (B) expression in the murine prostate gland in the absence of RM9 tumors (based on collapsed data from an n of 8 per age group of normal prostate samples; no age-related differences were identified in these control samples). Mean tumor expression of VEGF and TGF-β expression were higher than in basal prostate levels, but no age- or stress-related differences were uncovered. (C) In young mice, RM-9 tumor G-CSF levels were higher (*p < 0.05) in stressed than non-stressed mice. (D) In contrast, tumor GM-CSF in old mice was greater (*p < 0.05) in stressed than non-stressed mice. (E) CCL3/MIP-1α was one of the most highly expressed chemokine that was quantified, and one or two chemokines where effects of both age and stress were identified. In young mice, stress reduced (*p < 0.05) tumor CCL3/MIP-1α compared with levels in non-stress mice. In tumors from stress mice, CCL3/MIP-1α was higher (**p < 0.01) in old than young mice. (F) An increase (*p < 0.05) in tumor CCL4/MIP-1β levels was observed in old compared with young stressed mice. (G) In the stressed groups, CCL2/MCP-1 was lower (**p < 0.01) in old than young mice. (H) Only effects of age were observed for CXCL2/MIP-2 such that levels were higher in old than young mice regardless of stressed group (***p < 0.001). (I) Tumor CCL5/RANTES concentrations were reduced in old compared with young stressed mice, but in non-stressed mice, levels were higher in old than in young mice.

Figure 7

Expression of Interleukins in Prostate Tumors. (A) IL-1β expression was lower (p<0.05) in old than in young tumors, regardless of the stress-related group. (B, C) There was an effect of stress on IFN-γ and IL-6 tumor expression in old mice such that stress mice had elevated tumor expression than non-stressed mice (***p < 0.001). (D) Tumors from old stressed mice expressed higher levels of IL-9 than young stressed mice, but in non-stressed mice, tumor IL-9 levels were lower in old than young mice. (E) Tumor levels of IL-12p40 were low in non-stressed mice, regardless of age. In both young and old mice stress significantly increased (**p < 0.01) tumor IL-12(p40) expression compared with non-stress age-matched mice (F) Overall, IL-12p70 expression in prostate tumors was lower than IL-12p40 shown in (E), with low levels in old mice regardless of stress group. Still, in young and old mice, psychosocial stress reduced (**p < 0.01) tumor IL-12p70 expression. (G) Prostate tumor IL-17 levels were low in all treatment groups, but a stress-related increase (*p < 0.05) in tumor levels of IL-17 was observed in old compared with young mice.

Figure 8

Stress- and Age-Mediated Changes in Prostate Cytokine Expression. A summary of the changes in tumor chemokine/cytokine expression in young (left) and old (right) stressed mice is represented as a “teeter-totter’ with an increase or decrease in expression represented as an elevated or depressed bars, respectively. CCL, C-C Motif Chemokine Ligand; CXCL, (C-X-C motif) ligand 1; G-CSF, Granulocyte Colony-stimulating Factor; GM-CSF, Granulocyte/Monocyte Colony-stimulating Factor; IFN, interferon; IL, interleukin; MCP, Monocyte Chemoattractant Protein-1; MIP, Macrophage Inflammatory Protein; RANTES, Regulated upon Activation, Normal T Cell Expressed and Presumably Secreted; TGF, Transforming Growth Factor; Th, T-helper; VEGF, Vascular Endothelial Growth Factor (VEGF).