Dengue virus (DENV) is a mosquito-borne flavivirus estimated to cause up to 390 million infections per year of which 100 million cases are symptomatic (1). Dengue fever is caused by infection with any of the four dengue serotypes (DENV 1-4). Understanding the protective T cell response against dengue infection and identifying the associated immunodominant antigens is important for developing an effective, universal dengue vaccine. In this study, high-dimensional profiling in combination with peptide-MHC tetramer screening enabled the identification and in-depth characterization of dengue-specific CD8 T cells in patients with active dengue virus infection. While overall proportions of T cell subsets did not change significantly, many subsets showed increased proliferation and activation during the acute stage of infection. Dengue-specific CD8 T cells expressed a combination of activation and trafficking markers at acute and early-convalescent time points, which clearly distinguished them from other virus-specific T cells assessed in the same patients. During convalescence, dengue-specific CD8+ T cells differentiated into two major cell fates, a CD57+CD127– population of terminally differentiated, senescent memory cells and a CD127+CD57– population of proliferation-capable memory cells. These subsets remained detectable for up to one year after infection. These findings elucidated the impact of natural dengue infection on T cell activation and the generation of a memory T cell response. They also suggest that the intensity of immune activation may affect the composition of immune cells at the early-convalescent stage and the subsequent differentiation fate of dengue-specific T cells, and this may have important implications for the development of dengue vaccines inducing a durable immune response.
Figure 1. Summary figure for the study. High-dimensional mass cytometry with combinatorial peptide-MHC-tetramer multiplexing enables screening of dengue-specific T cell over time during the infection. Dengue infection causes broad immune cell activation and dengue-specific T cells differentiate into two major cell fates.
Peripheral blood mononuclear cells (PBMCs) from dengue patients were collected during different stages of dengue infection: acute (5–9 days post fever onset), post-febrile (14–28 days), and early convalescent (45–78 days). A 27-antibody staining panel comprising antibodies to activation and trafficking T cells markers was employed to profile PBMC by mass cytometry. In parallel, a three-metal combinatorial tetramer approach was used to screen for T cell reactivity against a total of 430 dengue and control antigens (2).
To understand how dengue infection affects various immune cell populations, PBMCs from dengue patients were collected during acute, post-febrile and early convalescent (fig. 2a) stages after infection. Using Uniform Manifold Approximation and Projection (UMAP) analysis to visualize the data (3), nine separate clusters were identified, representing distinct immune subsets such as CD4+ T cells, CD8+ T cells, Vd1+ gd T cells, Vd2+ gd T cells, mucosal-associated invariant T (MAIT) cells, B cells, CXCR5– B cells, plasmablasts, and NK cells (fig. 2b). Analysis at the acute time point showed a large expansion of proliferating Ki67+ immune cell subsets, especially plasmablasts, indicating that dengue infection was causing great expansion of these immune cell subsets (fig. 2b, c).
Figure 2. Dengue infection induces broad changes in immune cells over the course of infection.
PBMCs from dengue patients were collected during acute, post-febrile and early convalescent disease stages (a). UMAP was performed on a combined dataset from 14 samples and the resulting clusters were identified (left b). Representative UMAP plots from one patient in acute stage with Ki67+ cells (pseudo-color) overlaid over Ki67– cells (gray) (right b). Heatmap showing frequency of cells with positive marker expression in each of the identified clusters (c).
Overall, bulk frequencies of CD4+ and CD8+ T cell subsets did not change significantly over time except for Th2 cells, which increased during the recovery period (fig. 3a, c). Proliferation of the majority of CD4+ T cell subsets, and all CD8+ T cell subsets, was observed based on increased expression of Ki67 during the acute stage of dengue infection (fig. 3b, d). In summary, while overall proportions of peripheral blood CD4+, CD8+, and NK subsets did not change significantly, many immune subsets showed increased proliferation and activation during the acute stage of infection.
Figure 3. In-depth analysis of CD4+ and CD8+ T cells shows activation and proliferation during the acute stage of dengue infection. Frequencies of CD4+ and CD8+ T cells expressing activation and trafficking markers at each time point (a, c). Frequencies of Ki67+ cells within each CD4+ T cell subset (b). Frequencies of Ki67+ cells within each CD8+ T cell subset (d).
Using highly-multiplexed peptide-MHC tetramers, T cell reactivity against up to 430 dengue and control antigens was screened while measuring 27 activation and trafficking markers on T cells. An example of positive tetramer staining for NS5 protein on patient T cells is shown in figure 4a. Longitudinal frequencies of dengue -specific CD8+ T cells restricted to 3 HLA alleles are summarized in figure 4b, with the largest breath and number of T cell responses detected against epitopes restricted to the HLA-B*58:01 allele (figure 4b, c).
Figure 4. Peptide-MHC tetramer screening enables accurate characterization of dengue-specific CD8+ T cells. Representative mass cytometry plot of positive tetramer staining (a). Frequency of dengue tetramer+ T cells by HLA type and by epitope at each time point (b). Number of patients that responded to each epitope. Overlapping epitopes are indicated in red and green (c).
Further analysis of dengue-specific T cells over the course of the infection confirmed that most activation and trafficking markers were expressed early during the acute stage and declined thereafter, except for CD57, a marker of replicative senescence, and CD45RA, a marker re-expressed on terminally differentiated T cells, which both increased over time (fig. 5a). In addition, the profile of dengue-specific T cells was compared with EBV- and flu-specific T cells from early-convalescent stage samples in the same patients (fig. 5b). At acute and early-convalescent time points, dengue-specific T cells expressed a combination of activation and trafficking markers, that distinguished them from EBV- and flu-specific T cells.
Figure 5. Disease-specific analysis of tetramer+ CD8+ T cells.
Frequency of dengue tetramer+ T cells that express the respective activation markers at each time point (a). Frequency of early-convalescent dengue, EBV, and flu tetramer+ T cells that express the respective activation markers (b).
Analysis of the trajectory of dengue-specific CD8+ T cell fates indicated that they branched into 2 major subsets, which could be discriminated based on the expression of CD57 and CD127. Following the course of the infection, the frequency of dengue-specific CD57+CD127– CD8+ T cells increased over time, while the CD127+CD57– population peaked faster and persisted at a stable frequency through to early convalescence (fig. 6a). Validation in an independent second cohort demonstrated that these 2 dengue-specific populations remained elevated into the late convalescence stage, up to one year after infection (fig. 6b).
Figure 6. Frequencies of CD57 and CD127 subsets within dengue tetramer+ CD8+ cells over time. Analysis of a first cohort of dengue patients (14 patients from 3 experiments) (a). Analysis of a second cohort of dengue patients (29 patients from 5 experiments) (b).
This analytical strategy may be leveraged to screen T cell reactivity against the entire viral proteome and identify relevant epitopes targeted by the host CD8 T cell response longitudinally during natural infection. The findings from this study elucidate the impact of acute dengue infection on T cell activation, differentiation and the generation of long-term memory T cell responses. They suggest that the quality and quantity of T cell responses during acute infection or vaccination may be useful early indicators of long-lived immune memory, and may guide the design and development of a durable dengue vaccine.
- Bhatt, S., et al. (2013). The global distribution and burden of dengue. Nature 496, 504–507.
- Newell., et al. (2013). Combinatorial tetramer staining and mass cytometry analysis facilitate T-cell epitope mapping and characterization. Nat. Biotechnol. 31, 623–629.
- Becht E., et al. (2018). Dimensionality reduction for visualizing single- cell data using UMAP. Nat Biotechnol 2018.