CD8+ T cells induce cachexia during chronic viral infection

Abstract

Cachexia represents a leading cause of morbidity and mortality in various cancers, chronic inflammation and infections. Understanding of the mechanisms that drive cachexia has remained limited, especially for infection-associated cachexia (IAC). In the present paper we describe a model of reversible cachexia in mice with chronic viral infection and identify an essential role for CD8⁺ T cells in IAC. Cytokines linked to cancer-associated cachexia did not contribute to IAC. Instead, virus-specific CD8⁺ T cells caused morphologic and molecular changes in the adipose tissue, which led to depletion of lipid stores. These changes occurred at a time point that preceded the peak of the CD8⁺ T cell response and required T cell-intrinsic type I interferon signaling and antigen-specific priming. Our results link systemic antiviral immune responses to adipose-tissue remodeling and reveal an underappreciated role of CD8⁺ T cells in IAC.

Figure 1:. Infection with LCMV clone 13 leads to transient cachexia.

a) Body weight kinetics for C57BL6/J wild-type mice infected with 2×10⁶ FFU of LCMV-Cl13, in comparison with pair-fed mice (n=5). b-c) Food intake (b) and water intake (c) of LCMV-infected mice in comparison with uninfected controls. (n=10) (p-values: ****<0.0001, two-way ANOVA), Data are representative of three independent experiments for a) and b), c) represents a single experiment. d) Activity, oxygen consumption and respiratory exchange ratio (RER) in LCMV-infected mice compared to uninfected controls (n=10). Data shown represent the average of day 6, 7 and 8 post-infection (p-values: ****<0.0001) from a single experiment. e) Percentage of tissue weight normalized to body weight prior to infection in inguinal, gonadal and interscapular brown adipose tissue (p-values: **0.002, ***0.0006, *0.0172, ***0.0009 unpaired two tailed Student’s t-test), as well as quadriceps, gastrocnemius and soleus muscles (p-values: *0.039, **0.0035, ***0.0001 unpaired two tailed Student’s t-test) at day 7 for pair-fed uninfected mice and day 6 and 8 post-infection for LCMV-infected mice (n=5). f) Body composition as measured in live un-anesthetized LCMV-infected mice using EchoMRI (n=5). e-f) shows representative data of two independent experiments g) Body weight kinetics in LCMV-infected mice supplemented with the equivalent of 1kcal of indicated diet through oral gavage daily between day 4 and 7 post-infection (n=4). (p-values: ***0.0003 two-way ANOVA, Bonferroni correction). Data are representative of three independent experiments. All data shows mean ± s.e.m.

Figure 2:. infection-associated cachexia triggers severe adipose tissue remodeling, and increased lipolysis.

a) Representative MRI cross-sections (left), and longitudinal sections (right). Displayed sections were selected to show iLNs on uninfected controls and LCMV-infected mice at 6 and 8 days post-infection. (Bottom) images shows the fat compartment, after subtracting fat-suppressed acquisitions from the matching non-suppressed images. Images were acquired from a single experiment where (n=3) b) Representative H/E staining of inguinal fat pads from uninfected controls and mice infected at 6 and 8 days post-infection. Images were acquired from a single experiment where (n=3). c) 4–5 images were collected from each inguinal fat pad, at 20X magnification (n=3). The diameter of each adipocyte was measured, and the distribution and median values were evaluation across all conditions. Violin plots represent the density of data points in each condition and the minimum and maximum values. The inner box lots are bound by the upper (75%) and lower (25%) quantile, and the median is represented by an inner horizontal line. Cut-offs were set to include all cells between 10 and 100 μm in diameter (p-values: ****< 0.0001 one-way ANOVA, Bonferroni correction). d) Protein expression of Hormone Sensitive Lipase (HSL), phosphorylated-HSL (pHSL) and Adipose Triglyceride Lipase (ATGL), within the inguinal fat pad of control and infected mice at indicated timepoints (n=3). Data are representative of three independent experiments e) Serum non-esterified fatty acids (NEFA) and triglyceride levels of control and infected mice along 8 days post-infection (n=3). Data are representative of two independent experiments (p-values: **0.002 one-way ANOVA, Bonferroni correction). f) Fold change of inguinal fat pat mRNA expression of scavenger receptor (Cd36) (n=7), Lipoprotein Lipase (Lpl) (n=7) and Diglyceride Acyltransferase 2 (Dgat2) (n=8) as measured by qPCR from infected mice in comparison with pair-fed mice. Data are pooled from two independent experiments (p-values: **0.0021, **0.0038, **0.0069, *0.0279 two-way ANOVA, Bonferroni correction). g) Fold change of Uncoupling protein 1 (Ucp1) mRNA expression of control and infected mice in indicated fat pads (n=6). Data are pooled from two independent experiments (p-values: ***0.0002 unpaired two tailed Student’s t-test). h) Body weight kinetics in adipose tissue-specific knockouts of either ATGL (n=3) or HSL (n=4) compared to flox controls. Data are representative of two and independent experiments. e-i) shows mean ± s.e.m.

Figure 3:. Type I IFN and CD8 T cells play critical roles in inducing infection-associated cachexia.

a) Serum cytokine concentration of LCMV-infected wild-type mice (n=4). Data are representative of a single experiment. b) Body weight of LCMV-infected mice, either genetically ablated from IFNγ or TNF (n=4), or treated with anti-IFNγ or anti-TNF depleting antibodies (n=5), or a combination of both (n=9). Data are representative of a single experiment for genetic knockouts, two independent experiments for single antibody treatment, and pooled data from two independent experiments for the double depletion of IFNγ and TNF. c) Mice treated with anti-IL-6 depleting antibodies (n=5). d) Ifnar1^−/− mice (n=4) (p-values: ****< 0.0001 two-way ANOVA) e) Mice treated with anti-CD4 depleting antibodies (n=4). f) Cd8^−/− mice and mice treated with anti-CD8 depleting antibodies (n=4) (p-values: *0.027, *0.046 two-way ANOVA). c-f) are representative of two independent experiments. Data shows mean ± s.e.m.

Figure 4:. CD8 T cells modulate adipose tissue lipid metabolism in a type I IFN-dependent manner.

Analysis of Ifnar1^fl/flCD4^cre/+ mice compared to Ifnar1^fl/fl controls. a) Body weight kinetics of LCMV-infected mice (n=4) (p-values: ****< 0.0001 two-way ANOVA). Data are representative of four independent experiments and show mean ± s.e.m. b) Body composition as measured using EchoMRI in live un-anesthetized LCMV-infected mice (n=4) (p-values: ****< 0.0001 two-way ANOVA). Data are representative of a single experiment and show mean ± s.e.m. c-h) RNA-seq analysis of inguinal fat pads of LCMV-infected and uninfected Ifnar1^fl/flCD4^cre/+ mice compared to Ifnar1^fl/fl controls at 6 days post-infection (n=3). Data shows principal component analysis (PCA) (c), number of modulated genes (d), and significantly modulated genes (<0.05 Adjusted p-value, log fold change > 0.7 or < −0.7) in the limma implementation of 2×2 factorial interaction model. Top five enriched ClueGo pathways for up- and down-regulated genes (e). Enrichment of top 10 modulated metabolic pathways (f). Heatmap of the top modulated genes involved in lipid biosynthesis, uptake, or breakdown and release (g). RNA expression of Cd36, Lpl, and Dgat2 (h) (p-values: **** < 0.0001, *0.015 two-way ANOVA, Bonferroni correction). i) Serum triglyceride and non-esterified fatty acid levels (n=4) (p-values: **0.0027, ***0.0004 two-way ANOVA). j) Circulating levels of cortisol and corticosterone, as measured by ELISA (n=4) (p-values: **0.0010, *0.0185, ***0.0001, ***0.0004 two-way ANOVA, Bonferroni correction). h-j) data are representative of two independent experiments, and show mean ± s.e.m.

Figure 5:. CD8 T cells trigger cachexia during the early stage of T cell priming and antigen recognition.

a) Immunohistochemistry co-staining with DAPI, LCMV_NP (FITC), and CD8⁺ T cells (Alexa Fluor 647) in inguinal fat pads of LCMV-infected mice ay 6 and 8 days compared to uninfected mice (n=3). b) LCMV-NP⁺ and CD8⁺ staining from 4–5 images per inguinal fat pad (n=3), as quantified using cell profiler. (p-values: *0,0138, ****<0,0001, ***0,0001 one-way ANOVA, Bonferroni correction) a-b) data are representative of a single experiment. c) Fold change in circulating CD8⁺ T cells on 6 and 8 days post infection relative to uninfected control after daily gavage administration of either FTY720 (0.3mg/kg) or water (n=5). (p-values: **** < 0.0001, ***0.0004, *0.0143 two-way ANOVA, Bonferroni correction). d) Body weight kinetics of infected FTY720-treated and water-treated compared to uninfected controls (n=5). Data are representative of two independent experiments. e) Body weight kinetics of Prf1^−/− mice compared to wild-type controls (n=4). Data are representative of a single experiment. f) Body weight kinetics in WT+Cd8^−/− and OT-I Rag1^−/−+Cd8^−/− chimeric mice infected with LCMV (n=6), data are pooled from two independent experiments. b-f) data shows mean ± s.e.m.