Use of Clinical Isolates to Establish Criteria for a Mouse Model of Latent Cryptococcus neoformans Infection

Abstract

The mechanisms of latency in the context of C. neoformans infection remain poorly understood. Two reasons for this gap in knowledge are: 1) the lack of standardized criteria for defining latent cryptococcosis in animal models and 2) limited genetic and immunological tools available for studying host parameters against C. neoformans in non-murine models of persistent infection. In this study, we defined criteria required for latency in C. neoformans infection models and used these criteria to develop a murine model of persistent C. neoformans infection using clinical isolates. We analyzed infections with two clinical C. neoformans strains, UgCl223 and UgCl552, isolated from advanced HIV patients with cryptococcal meningitis. Our data show that the majority of C57BL/6 mice infected with the clinical C. neoformans isolates had persistent, stable infections with low fungal burden, survived beyond 90 days-post infection, exhibited weight gain, had no clinical signs of disease, and had yeast cells contained within pulmonary granulomas with no generalized alveolar inflammation. Infected mice exhibited stable relative frequencies of pulmonary immune cells during the course of the infection. Upon CD4+ T-cell depletion, the CD4^DTR mice had significantly increased lung and brain fungal burden that resulted in lethal infection, indicating that CD4+ T-cells are important for control of the pulmonary infection and to prevent dissemination. Cells expressing the T_bet transcription factor were the predominant activated CD4 T-cell subset in the lungs during the latent infection. These T_bet-expressing T-cells had decreased IFNγ production, which may have implications in the capacity of the cells to orchestrate the pulmonary immune response. Altogether, these results indicate that clinical C. neoformans isolates can establish a persistent controlled infection that meets most criteria for latency; highlighting the utility of this new mouse model system for studies of host immune responses that control C. neoformans infections.

Keywords: Cryptococcus neoformans; T-cells; adaptive immunity; cryptococcal meningitis; cryptococcosis; latent fungal infections; pulmonary granulomas.

Figure 1

UgCl223 and UgCl552 infections had low fungal burdens and minimal mortality. C57BL/6 mice were intranasally infected with C neoformans strains KN99α (triangle), UgCl223 (circle), or UgCl552 (square). (A–C) UgCl223 and UgCl552 infections were characterized by low fungal burdens. (A) Lungs, (B) spleen, and (C) brain were harvested, homogenized, and plated for colony forming units (CFUs) at 0-, 3-, 7-, 14-, 21-, 35-, 50-, and 70-days post-infection. No animals selected for timepoints succumbed to the infection. No organs were collected for CFU analysis for KN99α infection at 35-, 50-, and 70-days post-infection because all mice succumbed to infection prior to these timepoints. Trendlines were obtained using least squares regression with Gaussian distribution. n = 3-4 mice per timepoint per strain. (D) The majority of mice infected with UgCl223 and UgCl552 remained healthy. Mice were monitored for signs of morbidity for 90 days and sacrificed at natural endpoint (20% total weight loss; 1 g/day weight loss for 2 consecutive days; or neurological symptoms including loss of sternal recumbency, partial paralysis, seizure, convulsion, or coma). n = 9-10 mice per strain.

Figure 2

Mice infected with UgCl223 and UgCl552 maintained steady weight gain throughout the course of the infection. C57BL/6 mice were intranasally infected with (A) KN99α, (B) UgCl552, (C) UgCl223. Mouse weights in grams (g) were tracked until 90 days post-infection for UgCl552 and UgCl223, and 24 days post-infection for KN99α. Simple linear regression analysis was performed between mean weight of mice (n = 10) and days post-infection.

Figure 3

UgCl552 infection was contained within pulmonary granulomas. Representative histopathology images of granuloma formation at (A) 0.5x and (B) 20x magnification in lungs of a C57BL/6 mouse at 35 days post-infection with UgCl552. Black arrows point to granulomas. Black stars mark multinucleate giant cells. (C) 2x and (D) 20x magnification of lungs of a C57BL/6 mouse at 14 days post-infection with KN99α illustrating C. neoformans cells and inflammatory infiltrates with PMN’s visible in the insert. Scale bars = 50 μm, unless otherwise noted.

Figure 4

Lymphoid cells were the predominant immune cell in the lungs of UgCl223-infected mice. C57BL/6 mice were intranasally infected with UgCl223 or KN99α (lethal infection). For UgCl223, single cell lung suspensions were generated at 14- and 100-days post-infection. For KN99α, single cell lung suspensions were generated at 14-days post-infection. (A) Leukocyte, (B) lymphocyte, and (C) myeloid cell populations were quantified using flow cytometry. L = KN99α lethal infection at 14-days post-infection; U = uninfected control; 14 = UgCl223 infection at 14-days post-infection; 100 = UgCl223 infection at 100-days post-infection. The gating strategy is shown in Supplementary Figure 1. In brief, cells were identified as the following: Eosinophils = CD11b+CD24+SiglecF+, Dendritic Cells = CD11c+CD11b+/-CD24+MHCII+CD24-, Macrophages = CD64+CD24-, Monocytes = CD11b+Ly6C+/-MHCII-, T-cells = CD11b-CD11c-CD24-MHCII-, B-cells = CD11b-CD11c-MHCII+CD24^int, NK cells = CD11b+CD64-MHCII-Ly6C-. n = 3 mice per group. *p < 0.05, **p < 0.001 by one-way ANOVA with Bonferroni correction.

Figure 5

The adaptive immune response to UgCl223 infection within the lungs was dominated by CD4 T-cells. C57BL/6 mice were intranasally infected with UgCl223. Single cell lung suspensions were generated at 0, 7-, 14-, and 100-days post-infection. (A) Total CD4+ and CD8+ T-cell counts, (B) Total CD4+ and CD8+ T-cell proportion out of total CD3+ T-cells, (C) CD4+CD44+ and CD8+CD44+ activated T-cell counts, and (D) CD4+CD44+ and CD8+CD44+ activated T-cell proportion out of total CD3+ T-cells were quantified using flow cytometry. The gating strategy of the flow cytometric plots was doublet exclusion, exclusion of B220+CD11c+CD11b+F4/80+NK1.1+/dead cells/CD45 IV-labelled cells, gating on CD3+ T-cells, gating on CD4+ and CD8+ T-cells, and then gating on CD4+CD44+ and CD8+CD44+ T-cells. n = 3 mice per group. *p < 0.05, **p < 0.001, ***p < 0.0001, ****p < 0.00001 by two-way ANOVA. with Bonferroni correction. Statistical significance was noted between CD4+ vs. CD8+ groups, and CD4+CD44+ vs. CD8+CD44+ groups, but is not indicated on the graph.

Figure 6

Diphtheria toxin (DT) treatment ablated CD4+ T-cells in CD4^DTR mice and caused lethal UgCl223 infections. (A) Scheme showing generation of CD4^DTR mice and CD4+ T-cell ablation. A cross of CD4-cre mice to iDTR mice generate mice that contain CD4 T-cells sensitive to diphtheria toxin (DT). Filled triangles = loxP sites; arrows = transcriptional activity; open oval = promoter. (B) Experimental timeline showing CD4^DTR and C57BL/6 mice were intranasally infected with UgCl223 and treated with DT starting at 28 days post-infection. Mice received a booster dose of DT every 4 days to maintain CD4 depletion. (C) Lungs and (D) brain were analyzed for CFUs at 0-, 7-, 14-, 21-, 28-, 35- and 49-days post-infection. # = no CFUs were determined for timepoints prior to 28-days post-infection for DT-treated CD4^DTR mice. Black arrows indicate start of DT treatment. n = 3 mice per timepoint per group. (E) Mice were monitored for signs of morbidity and sacrificed at natural endpoint (20% total weight loss, 1g/day weight loss for 2 consecutive days, or neurological symptoms including loss of sternal recumbency, partial paralysis, seizure, convulsion, or coma). n = 4-5 mice per group. For CFU analysis, **p < 0.001, ****p < 0.00001 by two-way ANOVA. For survival kinetics, p=0.0075 by log-rank test.

Figure 7

CD8 T-cells were not required to control lung fungal burden during UgCl223 infection. (A) Experimental timeline showing C57BL/6 mice were intranasally infected with UgCl223 and treated with 15 mg/kg CD8 monoclonal antibody (2.43) starting at 28 days post-infection for 3 consecutive days and then weekly with booster injections. (B) Representative flow cytometric plots showing CD4+CD8- and CD4-CD8+ T-cells isolated from the lungs of infected mice at 1-week post-CD8 depletion (left), 2-weeks post-CD8 depletion (middle), and 3-weeks post-CD8 depletion (right). The gating strategy of the flow cytometric plots was doublet exclusion, gating on live cells, gating on CD3+/TCRβ+ cells, then gating on CD4+CD8- T-cells and CD4-CD8+ T-cells. Frequency of CD3+TCRβ+CD4-CD8+ T-cells was determined by calculating the percentage of CD3+TCRβ+CD4-CD8+ T-cells out of the grandparent gate (i.e., live singlet lymphocytes). (C) Lungs and (D) brain were analyzed for CFUs at 28-, 35-, and 49-days post-infection. n = 4-5 mice per timepoint per group. ***p < 0.0001, ****p < 0.00001 by two-way ANOVA.

Figure 8

Tbet+ T_H1 cells were the predominant CD4 T-cell subset in the lungs during the majority of the UgCl223 infection. (A) Flow cytometric analysis showing the proportions of CD4+CD44+Tbet+FoxP3- (T_H1), CD4+CD44+FoxP3+ (T_REG), and CD4+CD44+GATA3+FoxP3- (T_H2) cells over time at 0-, 7-, 21-, 56-, and 77-days post-infection. n=3 mice per timepoint. Statistical analysis performed by 2-way ANOVA with Bonferroni correction. † denotes P<0.05 for Tbet+ vs. GATA3+Tbet- and GATA3+Tbet- vs. FoxP3+GATA3-Tbet-. ‡ denotes P< 0.0001 for Tbet+ vs. GATA3+Tbet-; P < 0.05 for Tbet+ vs. FoxP3+GATA3-Tbet-; and P < 0.001 for GATA3+Tbet- vs. FoxP3+GATA3-Tbet-. (B) Flow cytometric analysis showing the proportions of CD4+CD44+Tbet+GATA3-FoxP3- and CD4+CD44+Tbet+Gata3+FoxP3- cells over time at 0-, 7-, 21-, 56-, and 77- days post-infection. ****p < 0.0001 by 2-way ANOVA with Bonferroni correction.

Figure 9

Pulmonary CD4 T-cells isolated from UgCl223 infected mice had diminished IFNγ production upon restimulation. (A) The proportion of IFNγ+ CD4+CD44+ cells in the total CD4+CD44+ cell population (%). PMA/ionomycin-stimulated CD4+CD44+ (left) and unstimulated CD4+CD44+ (right) cells were isolated from single cell suspensions of lungs harvested from C57BL/6 mice that were intranasally infected with UgCl223 (n=3) at 21-days post-infection versus mice that were uninfected (n=3) and analyzed by flow cytometry. (B) The proportion of IFNg+Tbet+CD4+CD44+ cells in the total Tbet+CD4+CD44+ cell population (%). Following cell sorting, Tbet-zsGreen+ cells from UgCl223-infected Tbet-zsGreen FoxP3-RFP mice (n=3) at 56-days post-infection and from uninfected mice (n=3) were stimulated with PMA/ionomycin (left) or were untreated (right), and then analyzed by flow cytometry. *p < 0.05 by two-tailed t-test.