Cryptococcus neoformans is the predominant causative agent of cryptococcal meningitis in immunocompromised individuals. Conversely, in immunocompromised individuals, C. neoformans establishes a latent pulmonary infection characterized by a paucity of clinical symptoms. Using a mouse inhalation model of latent C. neoformans infection, we previously showed that CD4 T-cells are necessary for preventing lung proliferation. In the current study, we performed unbiased single-cell RNA sequencing (scRNAseq) and found that the CD4 T-cell response was both highly heterogeneous and dichotomous during pulmonary C. neoformans infection, with concomitant expression of genes related to Th1 polarization (Tbx21, Ifng) and immune exhaustion (Ctla4). First, we demonstrated that cells with Th1-like phenotypes are necessary and sufficient to control latent infection via adoptive transfer of T-bet-positive cells into infection-matched CD4-depleted recipient mice. Second, scRNAseq analysis revealed that the subpopulation of effector CD4 T-cells that co-expressed Ctla4 and Gata3 was significantly higher than the subpopulation that co-expressed Ctla4 and Tbx21. Furthermore, our data suggested that CTLA-4 upregulation is beneficial and immunosuppressive against C. neoformans infection, as CTLA-4 blockade promoted fungal proliferation. Thus, we propose a model wherein Th1 control of latent C. neoformans infection is supported by CTLA-4 suppression of Th2 activation.

Cryptococcus neoformans is an opportunistic fungal pathogen that causes cryptococcal meningitis in immunocompromised individuals. In immunocompetent individuals, C. neoformans establishes a latent pulmonary infection characterized by a paucity of clinical symptoms. We previously used a mouse inhalation model of latent C. neoformans infection to show that CD4 T-cells are necessary for preventing lung fungal proliferation. In the current study, we used this same model to perform unbiased single-cell RNA sequencing (scRNAseq) to characterize the phenotype of the CD4 response to pulmonary C. neoformans infection. 

We conducted two separate single-cell RNA sequencing experiments. 1) A preliminary study analyzing total lung CD45+ leukocytes isolated from a C57BL/6 mouse infected with C. neoformans UgCl223 at 14 days post-infection, a mouse infected with UgCl223 at 180 days post-infection, a mouse infected with C. neoformans KN99a at 14 days post-infection, and an uninfected mouse. 2) A robust study analyzing lung CD4+ lymphocytes isolated from three C57BL/6 mice infected with UgCl223 at 14 days post-infection, three mice infected with UgCl223 at 180 days post-infection, and three mice infected with KN99a at 14 days post-infection.

For both studies, lungs were perfused with cold PBS and digested into a single cell homogenate with Collagenase A. For the CD45+ preliminary study, CD45+ cells were isolated via positive selection using the EasySep™ Mouse CD45 Positive Selection Kit (StemCell) per manufacturer’s instructions. For the CD4+ study, lymphocytes were first isolated via a 67%/40% Percoll density gradient, and then CD4 T-cells were isolated via negative selection using the EasySep™ Mouse CD4+ T cell Isolation Kit (StemCell) (see Flow cytometry methods subsection below for a more detailed description).

For both studies, single-cell RNA sequencing libraries were prepared using a Chromium Single Cell Controller (10x Genomics). The CD45+ preliminary study used the Chromium Single Cell 3’ Library and Gel Bead v3.1 dual index kit (10x Genomics). The CD4+ study used the Chromium Single Cell 5’ Library and Gel Bead v2 dual index kit (10x Genomics). Briefly for both studies, cell suspensions were diluted in nuclease-free water to obtain a target cell recovery of 10,000 cells, according to the manufacturer’s instructions. Remaining steps were also carried out according to the manufacturer’s instructions for cDNA amplification and sample index PCR. For both studies, final libraries were sequenced on two NovaSeq 6000 (Illumina) 150-bp paired-end flow cells at the University of Illinois at Urbana-Champaign DNA Services Lab. Libraries were sequenced with read lengths of 28x10x10x150. For both studies, the Mus musculus assembly mm10-2020-A (GRCm38) from NCBI was used with Cell Ranger 7.0.1 for alignment. The ‘cellranger count’ command was then used to call cells and count UMIs for each gene and cell.

For both studies, UMI count matrices were imported into R from the Cell Ranger output. The CD45+ preliminary study was performed with R (version 4.0.4) and Seurat (version 4.3.0.1). The CD4+ study was performed with R (version 4.4.0) and Seurat (version 5.0.3). Genes were discarded if there were fewer than 10 cells with UMIs (unique reads). Counts were then imported into a Seurat object. Percent mitochondrial UMIs (“percent.mt”) were calculated using the thirteen mitochondrial genes in the Cell Ranger reference. The percentage mitochondrial UMIs, along with total UMIs per cell and total features per cell with UMIs, were recorded and are available upon request. Cells with fewer than 200 features with UMIs were removed, as they were assumed to be red blood cells, low quality, or dying cells. Cells with more than 3 median absolute deviations above the median number of UMIs were filtered, given that they were likely to represent doublets. Cells with more than 3 median absolute deviations above the median number of mitochondrial UMIs were discarded, because they were assumed to be dying. UMI counts were normalized via the SCTransform method (Hafemeister & Satija, 2019) using the top 3,000 most variable features and regressing on percentage mitochondrial UMIs and cell cycle scores.

For the CD4+ study, we merged all experimental groups, but opted against performing integration across the different replicates and experimental groups since we did not observe condition-specific clustering in our merged dataset. Normalization, PCA, UMAP, and clustering analysis were performed as described above. To increase the rigor of our study, we identified clusters with elevated Cd4 expression and excluded clusters with low expression of Cd4 and cell clusters with non-T-cell markers (i.e., endothelial, alveolar, fibroblasts). In addition, to reduce artificial cluster separation caused by Trbv gene overrepresentation, all Trbv genes were removed from the variable gene set in this and all subsequent sub-clusterings. Normalization, dimensionality reduction, and clustering analysis were repeated to identify all Cd4 positive cell clusters. Cluster-specific markers were identified using FindAllMarkers(). Clusters were annotated based on the top 10 expressed genes. Of the 10 Cd4 clusters, only cluster 2 had high expression of Cd44, identifying it as the Effector Cd4 cluster.

Cluster 2 (Effector) was isolated from the larger Cd4 positive dataset in order to perform subclustering analysis, which included further normalization, dimensionality reduction, and clustering analysis. We opted to perform integration analysis during the subclustering analysis for better alignment of the Cluster 2 (Effector) subclusters. Differential expression analysis of the Cluster 2 (Effector) subclusters was performed using muscat and limma. A model matrix was constructed to compare conditions pairwise (Lethal 14 DPI versus Latent 14 DPI; Latent 180 DPI versus Latent 14 DPI). Trajectory analysis was conducted using Monocle3 (v1.3.1). Cells were ordered in pseudotime, with the trajectory root manually inferred from Subcluster 1.