Microglia are not protective against cryptococcal meningitis

Abstract

Microglia provide protection against a range of brain infections including bacteria, viruses and parasites, but how these glial cells respond to fungal brain infections is poorly understood. We investigated the role of microglia in the context of cryptococcal meningitis, the most common cause of fungal meningitis in humans. Using a series of transgenic- and chemical-based microglia depletion methods we found that, contrary to their protective role during other infections, loss of microglia did not affect control of Cryptococcus neoformans brain infection which was replicated with several fungal strains. At early time points post-infection, we found that microglia depletion lowered fungal brain burdens, which was related to intracellular residence of C. neoformans within microglia. Further examination of extracellular and intracellular fungal populations revealed that C. neoformans residing in microglia were protected from copper starvation, whereas extracellular yeast upregulated copper transporter CTR4. However, the degree of copper starvation did not equate to fungal survival or abundance of metals within different intracellular niches. Taken together, these data show how tissue-resident myeloid cells may influence fungal phenotype in the brain but do not provide protection against this infection, and instead may act as an early infection reservoir.

Fig. 1. Microglia are hosts to intracellular fungi.

a Frequency of C. neoformans infection within indicated cell types in the brain at day 7 post-infection. Data are pooled from 4 independent experiments (n = 17 mice). b Total number of infected cell types in the brain at day 7 post-infection. Data presented as mean +/- SEM and are pooled from 4 independent experiments (n = 13 mice) and analysed by one-way ANOVA with Bonferroni correction. *Adjusted P-value of <0.0001. c Proportion of indicated cell types within total infected cells at day 7 post-infection. An average value of 4 independent experiments for each cell type is shown. See Fig S1 for gating strategies. d Representative histology of mouse brain at day 6 post-infection, stained with Periodic-Acid Schiff (PAS). White arrows denote intracellular growing yeast, yellow arrows denote extracellular growing yeast. Similar results were observed across 3 brains analysed. e Confocal microscopy images of example tissue lesion with yeast and intracellularly-infected microglia. Mice were infected with GFP-expressing C. neoformans as before, and brains isolated at day 7 post-infection for analysis. Brain sections were stained with DAPI and anti-Iba1 to label microglia. GFP signal was used to identify yeast cells. Similar results were observed across 3 brains analysed. f Ratio of intracellular (grey bar) and extracellular (white bar) yeast counted across 3-5 sections from two mouse brains (at each time point), expressed as proportion of total counted yeast. g Histology of human brain, isolated at autopsy from a patient with HIV-associated cryptococcal meningitis, stained with PAS. White arrows denote intracellular growing yeast, yellow areas denote extracellular growing yeast. Similar results were observed in six sections analysed. Source data are provided as a source data file.

Fig. 2. Depletion of CNS-resident macrophages reduces fungal brain burden.

a Schematic of diphtheria toxin-based depletion of brain-resident macrophages. Cx3cr1-Cre^ER mice were crossed with iDTR mice. Resulting Cre+ (‘wild-type’) and Cre- (‘microglia knock-out’) littermates are treated with tamoxifen to induce Cre expression, and left to rest for 5 weeks to enable turn-over of monocyte-derived macrophages prior to daily treatment with diphtheria toxin to initiate cell depletion before and during intravenous C. neoformans H99 infection. b Total number of microglia in wild-type (n = 11; black circles) and microglia knock-out (n = 7; open circles) brains at day 3 post-infection. Data are pooled from 2 independent experiments and analysed by unpaired two-tailed t-test. ****P < 0.0001. c Fungal burdens in the brain (P = 0.0441) and lung (P = 0.7914) of wild-type (n = 11) and microglia knock-out (n = 7) mice at day 3 or 5 (n = 11 wild-type mice, n = 7 microglia knock-out mice; P = 0.0109) post-infection. Data are pooled from 2 independent experiments and analysed by two-tailed Mann Whitney U-test. Box plots show median with 25%/75% percentiles and maximum and minimum values. *P < 0.05. d Total number of indicated inflammatory cells in the brains of wild-type (n = 11) and microglia knock-out (n = 6) mice at day 3 post-infection. Data are pooled from 2 independent experiments. e Schematic of PLX5622 treatment. Wild-type C57BL/6 mice were fed either control diet or PLX5622 diet for 7 days prior to intravenous infection with C. neoformans H99. Diets were continued throughout infection. f Total number of microglia in untreated (n = 10; black triangle) and PLX5622-treated (n = 10; open triangle) brains at day 3 post-infection. Data are pooled from 2 independent experiments and analysed by unpaired two-tailed t-test. ****P < 0.0001. g Fungal burdens in the brain (P = 0.0186) and lung (P = 0.4876) of untreated (n = 14 brain, n = 5 lung) and PLX5622-treated (n = 14 brain, n = 5 lung) mice at day 3 or day 6 (n = 9 per group; P = 0.1081) post-infection. Data are pooled from 1 (lung) or 3 (brain) independent experiments and analysed by two-tailed Mann Whitney U-test. *P < 0.05. Box plots show median with 25%/75% percentiles and maximum and minimum values. h Total number of indicated inflammatory cells in the brains of untreated (n = 10) and PLX5622-treated (n = 10) mice at day 3 post-infection. Data are pooled from 2 independent experiments and analysed by two-way ANOVA with Bonferroni correction (P = > 0.99 macrophages, P = 0.6948 Ly6C^hi monocytes, P = 0.0456, Ly6C^lo monocytes, P = 0.0001 neutrophils). Mouse icons created with Biorender.com. Source data are provided as a source data file.

Fig. 3. Specific microglia depletion reduces brain fungal burden.

a Sall1-Cre^ER mice were bred with Rosa26^Ai14 animals and treated with tamoxifen to activate Cre and dTomato expression. Microglia and meningeal macrophage expression of dTomato is shown for representative Cre+ and Cre- littermates. b Frequency of dTomato expression in the indicated cell populations in the brain in 4 Cre+ littermates. Data are from 1 experiment and analysed by one-way ANOVA comparing to microglia (P = 0.0098 macrophages, P = 0.0035 Ly6C^hi, P = 0.0021 Ly6C^lo, P = 0.0047 neutrophils, P = 0.005 meningeal). **P < 0.01. c Frequency of dTomato expression in microglia and brain macrophages in uninfected (n = 4) and infected (n = 4 mice at days 0 and 3, n = 3 mice at day 6) Cre+ mice. d Total number of microglia (P = 0.0018) and meningeal macrophages (P = 0.5682) in wild-type (n = 4; black square) and microglia-depleted (n = 3; open square) brains at day 3 post-infection. Data are representative of 2 independent experiments and analysed by unpaired two-tailed t-test. **P < 0.01. e Total number of indicated inflammatory cells in the brains of wild-type (n = 4) and microglia-depleted (n = 3) brains at day 3 post-infection. Data are representative of 2 independent experiments. f Fungal burdens in the brain (P = 0.0442) and lung (P = 0.7346) of wild-type (n = 4) and microglia-depleted (n = 3) brains at day 3 post-infection. Data are representative of 2 independent experiments and analysed by unpaired two-tailed t-test. Data shown as mean +/− SEM. *P < 0.05. g Fungal brain burdens in wild-type (Sall1^CreERCsf1r^flox; Cre-negative littermates) and microglia-deficient (Sall1^CreERCsf1r^flox; Cre-positive littermates) mice at day 3 post-infection with C. deneoformans B3501 (n = 10 wild-type, n = 7 microglia-deficient) and C. neoformans Zc15 (n = 6 wild-type, n = 8 microglia-deficient). Data pooled from two independent experiments. Box plots show median with 25%/75% percentiles and maximum and minimum values. Source data are provided as a source data file.

Fig. 4. Loss of microglia does not affect brain infection with C. neoformans rdi1∆.

a Representative histology from brain of mice infected with wild-type C. neoformans H99 or rdi1Δ C. neoformans at day 3 post-infection, stained with PAS. White boxes in top images denote area of accompanying enlarged image. White arrows denote intracellular fungi, yellow areas denote extracellular fungi. Similar results were observed across 2 brains analysed. b Number of intracellular (grey bar) and extracellular (white bar) yeast counted across 3-5 sections from two mouse brains (for each strain). c Brain fungal burdens (median value denoted by bar) in untreated (black triangle) or PLX5622-treated (open triangle) mice at day 3 post-infection, infected with either wild-type C. neoformans H99 (n = 11 wild-type, n = 9 microglia-depleted) or rdi1Δ C. neoformans (n = 11 wild-type, n = 11 microglia-depleted). Data pooled from 2 independent experiments and analysed by one-way ANOVA with Bonferroni correction. ****P < 0.0001. ns = not significant. Source data are provided as a source data file.

Fig. 5. C. neoformans is protected from copper starvation when associated with microglia.

a Example flow cytometry plots of ^pCTR4GFP C. neoformans grown in copper-deficient YNB media with and without copper supplementation for 18 hours. mCherry is under the control of the ACT1 promotor and acts as a housekeeper, GFP expression is controlled by the CTR4 promotor. b Gating strategy for analysing pCTR4-GFP yeast cells in the brains of mice at day 7 post-infection. Yeast were first gated on using the mCherry marker, then split into extracellular (CD45−) and intracellular (CD45+ ) groups. Intracellular yeast were further gated to specifically analyse microglia (Ly6G^-Ly6C^-CD45^intCX3CR1^hi). c Frequency (P = 0.0439) of GFP+ cells and median fluorescent intensity (MFI; P = 0.0379) of ^pCTR4GFP in yeast cells that were extracellular or intracellular within microglia. Each point represents an individual mouse. Data are pooled from two independent experiments and analysed by paired two-tailed t-test. *P < 0.05. d Ratio (P = 0.0439) between GFP and mCherry in in yeast cells that were extracellular or intracellular within microglia. In panels c and d, each point represents an individual mouse. Data are pooled from two independent experiments and analysed by paired two-tailed t-test. *P < 0.05. e Frequency of GFP+ cells and GFP/mCherry ratio in ^pCTR4GFP C. neoformans grown in copper-deficient YNB media supplemented with indicated concentrations of copper. Yeast cells were analysed by flow cytometry after 18 hours growth. Each point represents a technical replicate. Data are representative of 3 independent experiments and shown as mean +/- SEM. Source data are provided as a source data file.

Fig. 6. Fungal heterogeneity in CTR4 expression correlates with IFNg stimulation of host cells.

a Expression of pCTR4GFP by yeast was assessed by flow cytometry in the indicated cell types as in other experiments. Each point represents an individual mouse analysed (n = 8 mice total). Data is pooled from three independent experiments and analysed by one-way ANOVA (***P = 0.0003, *P = 0.043 extracellular vs monocytes, *P = 0.134 extracellular vs neutrophils, *P = 0.0327 microglia vs macrophages). Representative FACS plots for microglia and macrophages within the same mouse are shown as an example. b Frequency of CXCL9+ microglia and macrophages in the infected brain at day 7 post-infection. Data pooled from three independent experiments (n = 9 mice) and analysed by unpaired two-tailed t-test. Box plots show median with 25%/75% percentiles and maximum and minimum values. *P < 0.0172. c Gating strategy used to compare ^pCTR4GFP expression by intracellular yeast within untreated or IFNγ-treated microglia. Cells were first gated to exclude free yeast, doublets and dead cells. Yeast bound to the surface of microglia but not internalised were removed from analysis using a calcofluor white (CFW) counter stain for the fungal cell wall. d Frequency of intracellular ^pCTR4GFP+ yeast in untreated and IFNγ-treated microglia after 2 hours of infection. Bars represent mean data from 4 independent experiments, points show technical replicates from one representative experiment. Data are analysed by unpaired two-tailed t-test on pooled means (P = 0.0266). e Ratio between mCherry and ^pCTR4GFP expression by intracellular yeast in untreated and IFNγ-treated microglia after 2 hours of infection. Each point represents mean value from 2 or 3 technical replicates tested in 4 independent experiments. Data are analysed by unpaired two-tailed t-test (P = 0.0055). **P < 0.01. f BV2 cells were infected and stained as above prior to analysis with an imaging flow cytometer. Example images of untreated BV2 cells (left) and IFNγ-treated BV2 cells (right) are shown alongside their respected uninfected control (top row). Data was quantified by measuring the frequency of cells within the ^pCTR4GFP^high gate, set within GFP + BV2 cells that were singlets and in focus. Data pooled from two independent experiments and analysed by two-tailed unpaired t-test (P = 0.0059). **P < 0.01. Source data are provided as a source data file.

Fig. 7. C. neoformans CTR4 expression does not correlate with immune cell killing or metal abundance.

a Abundance of indicated metals in FACS-purified cell populations measured by ICP-MS. Total concentrations were divided by the number of cells used as input to generate the abundance (parts per billion, ppb) per cell. Data pooled from two independent sorts (iron, magnesium) or a single sort (zinc). b Total copper concentration in whole mouse brains at indicated time points post-infection. Each point represents an individual mouse (n = 6 mice at day 0, n = 5 mice at day 3, n = 7 mice at days 6 and 9). Data pooled from two independent experiments. c Ex vivo fungal killing assay. Mice were infected with GFP-expressing C. neoformans and GFP+ myeloid cells were FACS-purified at day 7 post-infection. Infected cells (n = 1000 cells in experiment 1, n = 2000 cells in experiment 2) were lysed in water prior to plating onto YPD agar plates. Colonies were counted to calculate the percentage viability and frequency of fungal killing. Data pooled from two independent experiments, shown as mean +/- SEM and analysed by one-way ANOVA (P = 0.0378 macrophages vs neutrophils, P = 0.0199 monocytes vs neutrophils). *P < 0.05. d Brain fungal burdens at day 3 post-infection from Sall1^CreERCsf1r^flox mice infected intravenously with 5×10⁵ ∆ctr1/∆ctr4 C. neoformans (n = 10 wild-type mice, n = 6 microglia-depleted mice, P = 0.0047) or 2×10⁴ CTR4^OE (n = 5 wild-type mice, n = 8 microglia depleted mice, P = 0.2396). Wild-type (grey bars) refers to Cre-negative littermate controls, microglia-depleted (open bars) refers to Cre+ littermates. Box plots show median with 25%/75% percentiles and maximum and minimum values. Data pooled from two independent experiments and analysed by two-tailed Mann Whitney U-test. **P < 0.01. Source data are provided as a source data file.