Systemic immunostimulation induces glucocorticoid-mediated thymic involution succeeded by rebound hyperplasia which is impaired in aged recipients

Abstract

The thymus is the central organ involved with T-cell development and the production of naïve T cells. During normal aging, the thymus undergoes marked involution, reducing naïve T-cell output and resulting in a predominance of long-lived memory T cells in the periphery. Outside of aging, systemic stress responses that induce corticosteroids (CS), or other insults such as radiation exposure, induce thymocyte apoptosis, resulting in a transient acute thymic involution with subsequent recovery occurring after cessation of the stimulus. Despite the increasing utilization of immunostimulatory regimens in cancer, effects on the thymus and naïve T cell output have not been well characterized. Using both mouse and human systems, the thymic effects of systemic immunostimulatory regimens, such as high dose IL-2 (HD IL-2) with or without agonistic anti-CD40 mAbs and acute primary viral infection, were investigated. These regimens produced a marked acute thymic involution in mice, which correlated with elevated serum glucocorticoid levels and a diminishment of naïve T cells in the periphery. This effect was transient and followed with a rapid thymic "rebound" effect, in which an even greater quantity of thymocytes was observed compared to controls. Similar results were observed in humans, as patients receiving HD IL-2 treatment for cancer demonstrated significantly increased cortisol levels, accompanied by decreased peripheral blood naïve T cells and reduced T-cell receptor excision circles (TRECs), a marker indicative of recent thymic emigrants. Mice adrenalectomized prior to receiving immunotherapy or viral infection demonstrated protection from this glucocorticoid-mediated thymic involution, despite experiencing a substantially higher inflammatory cytokine response and increased immunopathology. Investigation into the effects of immunostimulation on middle aged (7-12 months) and advance aged (22-24 months) mice, which had already undergone significant thymic involution and had a diminished naïve T cell population in the periphery at baseline, revealed that even further involution was incurred. Thymic rebound hyperplasia, however, only occurred in young and middle-aged recipients, while advance aged not only lacked this rebound hyperplasia, but were entirely absent of any indication of thymic restoration. This coincided with prolonged deficits in naïve T cell numbers in advanced aged recipients, further skewing the already memory dominant T cell pool. These results demonstrate that, in both mice and humans, systemic immunostimulatory cancer therapies, as well as immune challenges like subacute viral infections, have the potential to induce profound, but transient, glucocorticoid-mediated thymic involution and substantially reduced thymic output, resulting in the reduction of peripheral naive T cells. This can then be followed by a marked rebound effect with naïve T cell restoration, events that were shown not to occur in advanced-aged mice.

Keywords: age; glucocoricoids; immune therapy; stress; thymic involution; viral infection.

Figure 1

Strong systemic immunostimulatory regimens result in rapid acute thymic involution in mice. (A) Schema – C57BL/6 mice were given either anti-CD40 (aCD40, 65μg) in combination with rhIL-2 (5x10⁶ IU) (aCD40/IL-2 group) or rat IgG (control group) intraperitoneally on days 0, 4, and 8. Mice were monitored over the course of 33 days, with takedowns occurring at days 2, 12, 19, 26, and 33. (B) Representative flow cytometry staining of CD4⁺ vs CD8⁺ (previously gated on CD3⁺) thymocytes from days 2, 12, and 26. (C) Thymic cellularity of mice on day 2 post injection. (D) Fold change of double positive thymocytes in mice receiving aCD40/IL-2 in comparison to control mice at days 2, 12, 19, and 26. (E, F). Serum levels of IL-6 and IFN-y at day 2 and day 12. (G) Serum corticosterone levels of mice at days 2 and 12 post injection. (H) Percentages of peripheral naïve (CD44^-CD62L⁺ and CD44^-CD62L^-), central memory (CM) (CD44⁺CD62L⁺), or effector memory (EM) (CD44⁺CD62L^-) populations of CD3⁺ over the course of the 33-day experiment. (I–K) Percentage change in peripheral total naïve, central memory, and effector memory CD3⁺ cell counts when compared to baseline controls over the course of the 33-day experiment (D–K) SEM bars, n=3-5 mice per group, representative of 2 experiments. One-way ANOVA with multiple comparisons based on means between groups was used to determine statistical significance; P<0.05*, P<0.01**, P<0.001***, P<0.0001****. (C) SEM bars, n=3-5 mice per group, representative of 2 experiments. Student’s t test was used to determine statistical significance; P<0.05*, P<0.01**, P<0.001***, P<0.0001****. ns, non-significant.

Figure 2

High-dose IL-2 induces apoptosis driven thymic involution. (A) Schema – Mice were administered 5x10⁶ IU of IL-2 or rat IgG at days 0, 4, and 8 intraperitoneally. Mice were monitored over the course of 33 days. Mice were taken down at days 2, 12, 19, 26, and 33. (B) Annexin V representative flow staining on thymuses from mice 24 hours after IL-2 administration (C) Thymic cellularity fold change of HD IL-2 treated mice vs controls over the course of the experiments. (D, F) Pie chart breakdowns of the percentages of naïve, central memory, and effector memory populations in CD8⁺ (D) and CD4⁺ (F) T cell subsets over the course of the experiments. (E, G) Percentages of naïve, effector memory, and central memory populations in CD8⁺ (E) and CD4⁺ (G) T cell subsets over the course of the experiments. (C, E, G) SEM bars, n=3-6 mice per group, representative of 1-2 experiments. One-way ANOVA with multiple comparisons based on means between groups was used to determine statistical significance; P<0.05*, P<0.01**, P<0.001***, P<0.0001****.

Figure 3

High-dose (HD) IL-2 therapy results in significant decrease in thymic output, concurrent T-cell activation, and increased glucocorticoids humans. (A) Schema – patients with metastatic melanoma or renal cell carcinoma received 6x10⁵ IU of IL-2 every 8 hours for 14 doses. Blood samples were collected from patients at days 0, 2, 8, and 28. (B, C) Percentage of naïve (CD45RA⁺CD45RO⁺, 6H) and memory (CD45RA^-CD45RO⁺, 6I) of CD3⁺ T cells from peripheral blood of patients at days 0, 2, and 8. (D) Quantification of TREC⁺ cell numbers in the peripheral blood of patients at baseline vs week 4. (E, F) Serum cortisol and sTNFR levels of patients at days 0, 2, 8, and 28 days. (G): TREC frequency fold change of individual patients at day 28 when compared to homeostatic baselines before treatment. (B, C, E, F): SEM bars, n=5-9 patients per group, representative of samples from 1 clinical trial. One-way ANOVA with multiple comparisons based on means between groups was used to determine statistical significance in (B, C) while samples were paired for ANOVA analysis across baseline, D2, D8, and 4w in (F, G); P<0.05*, P<0.01**, P<0.001***. (D): Paired t test was used to determine statistical significance; P<0.05*. ns, non-significant.

Figure 4

Abrogation of the corticosterone response via adrenalectomy can protect the thymus from immunotherapy or viral infection associated thymic involution. (A) Schema – C57BL/6 mice either had a SHAM or adrenalectomy (adx.) surgery performed several weeks prior to the experiment. Adx and SHAM mice were either inoculated with 1x10⁴ PFU of MCMV (controls received 0.2 mL of RPMI), or were injected with aCD40 (65μg) in combination with rhIL-2 (5x10⁶ IU) (controls received 0.2 mL of PBS). Mice were assessed at 36 hrs, DPI 2, and DPI 4. (B, C) Thymic cellularity and serum corticosterone of SHAM and adx mice, 2 days after aCD40/IL-2 treatment. (D, E) Thymic cellularity and serum corticosterone of SHAM and adx mice, 2 days after aCD40/IL-2 treatment 36 hours after MCMV inoculation. (F–I) Serum cytokine levels of TNF (F), IL-6 (G), IL-1β (H), and IFN-y (I) 4 days after MCMV infection. (B–I) SEM bars, n=3-8 mice per group, representative of 2-3 combined experiments. One-way ANOVA with multiple comparisons based on means between groups was used to determine statistical significance; P<0.05*, P<0.01**, P<0.0001****. ns, non-significant.

Figure 5

Adrenalectomized mice are protected from corticosterone mediated thymic apoptosis that occurs during an acute systemic immune challenge. (A) RNAseq heatmap of apoptosis related genes and their differential expression from the extracted RNA of thymocytes from SHAM control and SHAM MCMV infected (36 hours post infection) mice. (B) GSEA of apoptosis related genes in the heatmap of (A) with SHAM infected vs SHAM control as the basis of enrichment (C–F) Normalized reads of the genes Bcl2 (C) , Casp8 (D) , Bax (E) , and FasL (F) between SHAM control and SHAM infected mice 48 hours post infection. (G) Heatmap of apoptosis related gene differentials between SHAM infected and adx infected mice at 36 hours post infection. (H) GSEA of apoptosis related genes in the heatmap of 5G, with adx infected vs SHAM infected as the basis of enrichment. (I–L) Comparison of normalized reads of the genes Bcl-2 (I) , Casp3 (J) , Card10 (K) , and Bid (L) between SHAM and adx infected mice. (C–F, I–L) SEM bars, n=4 mice per group. Student’s t test used to determine statistical significance; P<0.05*, P<0.01**.

Figure 6

Age impairs thymic recovery and prolongs memory expansion in mice infected with MCMV. (A) Schema; young (2-5mo), middle aged (7-12mo), and aged (22-27mo) mice were infected with low dose MCMV (1x10⁴ PFU) and assessed at days 3 (blood draw only), 6, 9, and 21 and 25. (B) Serum corticosterone levels of young, middle aged, and aged infected mice at DPI 3. (C–F) Thymic cellularity of control and infected young, middle aged, and aged mice at DPI 6, 9, and 21 through 25. (G–I) Percentages of naïve (CD44^-CD62L^+/-) CD3 T cells in the spleens of young, middle aged, and aged mice controls and infected mice assessed at DPI 9 and 21-25 (J) Representative flow cytometry of the spleens of infected young, middle aged, and aged mice at days 9 and 21-25, as well as control mice for each age group. (K) Pie chart representation of the naïve, central memory, and effector memory ratios of young, middle aged, and aged control and infected mice at days 9 and 21-25. SEM bars, n=3-15 mice, representative of 2-4 experiments. One-way ANOVA with multiple comparisons based on means between groups was used to determine statistical significance in all panels; P<0.05*, P<0.01**, P<0.001***, P<0.0001****.