Systemic metabolic changes in acute and chronic lymphocytic choriomeningitis virus infection

Abstract

Objective: Viral infection of cells leads to metabolic changes, but how viral infection changes whole-body and tissue metabolism in vivo has not been comprehensively studied. In particular, it is unknown how metabolism might be differentially affected by an acute infection that the immune system can successfully clear compared to a chronic persistent infection.

Methods: Here we used metabolomics and isotope tracing to identify metabolic changes in mice infected with acute or chronic forms of lymphocytic choriomeningitis virus (LCMV) for three or eight days.

Results: Both types of infection alter metabolite levels in blood and tissues, including itaconate and thymidine. However, we observed more dramatic metabolite changes in the blood and tissues of mice with persisting LCMV infection compared to those infected with the acute viral strain. Isotope tracing revealed that the contribution of both glucose and glutamine to the tricarboxylic acid (TCA) cycle increase in the spleen, liver, and kidneys of mice infected with chronic LCMV, while acute LCMV only increases the contribution of glutamine to the TCA cycle in the spleen. We found that whole-body turnover of both glutamine and thymidine increase during acute and chronic infection, whereas whole-body glucose turnover surprisingly does not change. Activated T cells in vitro produce thymidine and virus-specific T cells ex vivo have increased thymidine levels, nominating T lymphocytes as the source of thymidine in LCMV infection.

Conclusions: In sum, we provide comprehensive measurements of whole-body and tissue metabolism in acute and chronic viral infection, and identify altered thymidine metabolism as a marker of viral infection.

Keywords: Immunometabolism; Isotope tracing; Metabolomics; Tissue metabolism; Whole-body metabolism.

Figure 1

Chronic LCMV infection changes serum metabolite levels more dramatically than acute infection. (A–B) Serum metabolite changes in day 8 acute Armstrong strain LCMV infection (A) or chronic Clone 13 strain (B) compared to uninfected; n = 4 uninfected for each, n = 6 for Armstrong infected, n = 5 for Clone 13. Two-tailed t-tests in (A–B) performed on log₂ transformed metabolite intensities. (C–D) Metabolic pathways enriched among significantly increased metabolites in serum of (C) Armstrong infected mice or (D) Clone 13 infected mice compared to uninfected. (E–F) Serum metabolite levels of itaconate (E) and thymidine (F) in LCMV infected mice. Includes multiple experiments, all infected sera normalized to uninfected sera from the same experiment; n = 27 uninfected, n = 6 acute day 3, n = 6 chronic day 3, n = 18 acute day 8, n = 11 chronic day 8; two-tailed t-tests in performed on log₂ fold changes from uninfected. (G–H) All (G) and selected (H) serum metabolite changes in chronic versus acute day 8 infection, ratios of fold changes compared to uninfected shown. (I) In which strain of LCMV did serum metabolites exhibit a greater absolute-value fold-change compared to uninfected. For (G–I), n = 18 acute day 8, n = 8 chronic day 8, two-tailed t-tests done comparing log₂ fold changes from uninfected.

Figure 2

Acute LCMV infection alters tissue metabolites. (A) Principal component analysis of metabolite levels in spleen, liver, kidney, and intestine of mice on day 8 of acute Armstrong strain LCMV infection or uninfected mice. Metabolites are log transformed, row-mean normalized, and standard deviation set to 1. (B) Heatmap with hierarchical clustering of metabolite levels, same data as (A). Metabolites are log transformed and row-mean normalized. (C–F) Tissue metabolites in (C) spleen, (D) liver, (E) kidney, (F) small intestine of mice on day 8 of acute LCMV infection compared to uninfected. Selected metabolites significantly changed from uninfected tissue are labeled. In all panels, n = 4 mice uninfected, n = 6 infected. Two-tailed t-tests in (C–F) performed on log₂ transformed metabolite intensities.

Figure 3

Chronic LCMV infection changes tissue metabolites more than acute infection. Tissue metabolites in (A) spleen or (B) liver of mice on day 8 of chronic infection compared to uninfected; n = 4 uninfected, n = 5 infected. Two-tailed t-tests in (A–B) performed on log₂ transformed metabolite intensities. (C–F) Levels of (cC spleen or (D) liver thymidine, and (E) spleen and (F) liver itaconate in uninfected mice or mice on day 8 of LCMV infection. n = 8 uninfected, n = 6 acute LCMV, n = 5 chronic LCMV; infected values normalized to uninfected mice from the same experiment. (G–H) Spleen (G) and liver (H) metabolite changes from uninfected on day 8 of chronic or acute infection. (I–J) Selected spleen (I) and liver (J) metabolite changes on day 8 of acute or chronic infection, relative to uninfected serum. (K) In which strain of LCMV did spleen or liver metabolites exhibit a greater absolute-value fold-change compared to uninfected tissues. For (G–K), n = 15 acute day 8, n = 8 chronic day 8. All two-tailed t-tests performed on log₂ fold changes from uninfected.

Figure 4

LCMV infection increases glutamine and thymidine whole-body turnover. (A) Schematic of whole-body metabolite turnover measurement using isotope-labeled metabolite infusion. (B) Whole-body turnover of glucose determined by [U–¹³C] glucose infusion. (5 experiments pooled, overall n = 19 uninfected, n = 5 acute LCMV day 3, n = 3 chronic LCMV day 3, n = 11 acute LCMV day 8, n = 11 chronic LCMV day 8.) (C) Whole-body turnover of glutamine determined by [U–¹³C] glutamine infusion. (3 experiments pooled, overall n = 14 uninfected, n = 3 acute LCMV day 3, n = 4 chronic LCMV day 3, n = 6 acute LCMV day 8, n = 6 chronic LCMV day 8.) (D) Whole-body turnover of lactate determined by [U–¹³C] lactate infusion. (4 experiments pooled, overall n = 10 uninfected, n = 6 acute LCMV day 3, n = 3 chronic LCMV day 3, n = 6 acute LCMV day 8, n = 3 chronic LCMV day 8.) (E) Whole-body turnover of thymidine determined by [¹⁵N₂] thymidine infusion. (2 experiments pooled, overall n = 8 uninfected, n = 4 acute LCMV day 3, n = 4 chronic LCMV day 3, n = 4 acute LCMV day 8, n = 9 chronic LCMV day 8). All t-tests are Student’s two-tailed t-tests between uninfected vs other groups, acute d3 vs chronic d3, or acute d8 vs chronic d8; all tests not significant if not shown.

Figure 5

Glucose contribution to tissue TCA cycle increases in chronic LCMV infection. (A) Schematic of measurement of nutrient contribution to TCA metabolites using labeled nutrient infusion. (B) Average carbon labeling of malate TCA intermediate in spleen from [U–¹³C] glucose infusion in LCMV-infected and uninfected mice. (n = 14 uninfected, n = 5 acute LCMV day 3, n = 3 chronic LCMV day 3, n = 6 acute LCMV day 8, n = 8 chronic LCMV day 8.) (C) Average labeling of malate in liver from [U–¹³C] glucose infusion. (n = 14 uninfected, n = 5 acute LCMV day 3, n = 3 chronic LCMV day 3, n = 6 acute LCMV day 8, n = 8 chronic LCMV day 8.) (D) Average labeling of malate in kidney from [U–¹³C] glucose infusion. (n = 9 uninfected, n = 5 acute LCMV day 3, n = 3 chronic LCMV day 3, n = 6 acute LCMV day 8, n = 3 chronic LCMV day 8.) (E) Average labeling of malate in quadriceps from [U–¹³C] glucose infusion. (n = 7 uninfected, n = 3 chronic LCMV day 3, n = 8 chronic LCMV day 8.) (F) Average labeling of malate in spleen from [U–¹³C] glutamine infusion. (G) Average labeling of malate in liver from [U–¹³C] glutamine infusion. (H) Average labeling of malate in kidney from [U–¹³C] glutamine infusion. For f-h, n = 12 uninfected, n = 3 acute LCMV day 3, n = 4 chronic LCMV day 3, n = 3 acute LCMV day 8, n = 6 chronic LCMV day 8. (I) Average labeling of malate in quadriceps muscle from [U–¹³C] glutamine infusion; n = 2 uninfected, n = 3 chronic LCMV day 8. All t-tests are Student’s two-tailed t-tests between uninfected vs other groups, acute d3 vs chronic d3, or acute d8 vs chronic d8; all tests not significant if not shown.

Figure 6

T lymphocytes may be a major source of thymidine. (A) Schematic of thymidine metabolism pathway. (B) Labeling of spleen thymidine from [U–¹³C] glucose infusion. (5 experiments pooled, overall n = 13 uninfected, n = 5 acute LCMV day 3, n = 3 chronic LCMV day 3, n = 6 acute LCMV day 8, n = 7 chronic LCMV day 8.) T-tests are Student’s two-tailed t-tests between uninfected vs other groups, acute d3 vs chronic d3, or acute d8 vs chronic d8; all tests not significant if not shown. (C) Fractional composition of CD45+ spleen cells in n = 6 uninfected, n = 6 day 8 acute, and n = 5 day 8 chronic LCMV infected mice measured by flow cytometry. (D) Activation status (CD44+ fraction) of transferred LCMV-specific P14 CD8 T cells in the spleen in n = 5 uninfected, n = 6 day 8 acute, and n = 5 day 8 chronic LCMV infected mice measured by flow cytometry. (E) Thymidine levels in tissues of healthy mice, n = 3 mice. (F) Serum thymidine in mice with NOTCH-1 induced T-cell acute lymphocytic leukemia compared to healthy controls, n = 5 for each group. (G) Production of thymidine by mouse T cells upon activation, n = 4 biological replicates per group. (H) Thymidine level in CD8+ T cells, either CD8+ cells from uninfected mice or P14 virus-specific CD8+ T cells isolated from LCMV-infected mice on day 8, n = 11 uninfected and chronic d8, n = 12 acute d8. All t-tests are Student’s two-tailed t-tests, tests in (F–G) are performed on log₂ transformed metabolite intensities.