Epitope-specific airway-resident CD4+ T cell dynamics during experimental human RSV infection

Abstract

BACKGROUNDRespiratory syncytial virus (RSV) is an important cause of acute pulmonary disease and one of the last remaining major infections of childhood for which there is no vaccine. CD4+ T cells play a key role in antiviral immunity, but they have been little studied in the human lung.METHODSHealthy adult volunteers were inoculated i.n. with RSV A Memphis 37. CD4+ T cells in blood and the lower airway were analyzed by flow cytometry and immunohistochemistry. Bronchial soluble mediators were measured using quantitative PCR and MesoScale Discovery. Epitope mapping was performed by IFN-γ ELISpot screening, confirmed by in vitro MHC binding.RESULTSActivated CD4+ T cell frequencies in bronchoalveolar lavage correlated strongly with local C-X-C motif chemokine 10 levels. Thirty-nine epitopes were identified, predominantly toward the 3' end of the viral genome. Five novel MHC II tetramers were made using an immunodominant EFYQSTCSAVSKGYL (F-EFY) epitope restricted to HLA-DR4, -DR9, and -DR11 (combined allelic frequency: 15% in Europeans) and G-DDF restricted to HLA-DPA1*01:03/DPB1*02:01 and -DPA1*01:03/DPB1*04:01 (allelic frequency: 55%). Tetramer labeling revealed enrichment of resident memory CD4+ T (Trm) cells in the lower airway; these Trm cells displayed progressive differentiation, downregulation of costimulatory molecules, and elevated CXCR3 expression as infection evolved.CONCLUSIONSHuman infection challenge provides a unique opportunity to study the breadth of specificity and dynamics of RSV-specific T-cell responses in the target organ, allowing the precise investigation of Trm recognizing novel viral antigens over time. The new tools that we describe enable precise tracking of RSV-specific CD4+ cells, potentially accelerating the development of effective vaccines.TRIAL REGISTRATIONClinicalTrials.gov NCT02755948.FUNDINGMedical Research Council, Wellcome Trust, National Institute for Health Research.

Keywords: Adaptive immunity; Immunology; Infectious disease; T cells.

Conflict of interest statement

Conflict of interest: The authors have declared that no conflict of interest exists.

Figures

Figure 1. Flow diagram outlining study design and participating subjects.

(A) Healthy adult volunteers (n = 49) were enrolled and inoculated with RSV M37 for polyclonal CD4+ T cell analysis and epitope discovery. (B) A second cohort (n = 8) was enrolled for tetramer analysis of RSV-specific CD4+ T cells.

Figure 2. Enrichment of activated and regulatory CD4+ T cells in the lower airway during RSV infection.

(A) Whole blood (n = 49) and BAL (n = 24) samples were stained with anti-CD3, -CD4, -CD8, -CD38, and –Ki-67 for analysis by flow cytometry. Plots are gated on CD3+CD4+ lymphocytes. One representative infected subject is shown for blood (upper panels) and BAL (lower panels). Median and individual data points of Ki-67+CD38+CD4+ T cells in the (B) blood and (C) BAL of infected (PCR+, red) or uninfected (PCR–, blue) volunteers are shown. Tests of the 5 a priori hypotheses were conducted by Wilcoxon’s signed-rank test with Bonferroni-adjusted α levels of 0.01 (**P < 0.001). (D) Frequencies of Ki-67+CD38+ cells on day 10 after infection are compared between paired blood and BAL samples in infected individuals (n = 12). Tests of the 5 a priori hypotheses were conducted by Wilcoxon’s signed-rank test with Bonferroni-adjusted α levels of 0.01 with no statistically significant differences seen. (E) Whole blood and BAL samples were stained with anti-CD3, -CD4, -FoxP3, and -CD25. One representative infected BAL sample is shown gated on CD3+CD4+ lymphocytes. (F) Mean and individual data points of FoxP3+CD25+CD4+ T cells in the blood and BAL of infected (PCR+, red circles) or uninfected (PCR–, blue squares) volunteers are shown. P values for Wilcoxon’s signed-rank (intragroup) and Mann-Whitney tests (intergroup) are shown. *P < 0.05.

Figure 3. Activated CD4+ T cell frequencies in the lower airway correlate strongly with CXCL10 expression.

(A) CD4+ cells (brown) were identified in bronchial biopsies (n = 12) by immunohistochemistry and enumerated in infected individuals. Scale bars: 20 μm. Individual data points are presented as number of positive cells per square millimeter of subepithelium or per 0.1 mm2 of epithelium. *P < 0.05 by Wilcoxon’s signed-rank test. (B) Differential cytokine and chemokine gene expression in bronchial brushings 7 days after infection compared with preinfection was analyzed by RT2 Profiler qPCR array. Greater than 2-fold upregulated (red) and downregulated (green) genes are shown. (C) Mean ± SEM gene expression levels of CXCL9, CXCL10, and CXCL11 are shown in infected (red) and uninfected (green) individuals. (D) Spearman’s correlations between bronchial mucosal lining fluid CXCL10 or CXCL11 concentrations and the frequency of Ki-67+CD38+CD4+ T cells in BAL are shown.

Figure 4. CD69+ resident memory CD4+ T cells in BAL exhibit advanced differentiation.

Whole blood/PBMCs (n = 10) and BAL (n = 5) from individuals infected with RSV were costained with anti-CD3, -CD4, and phenotypic markers and then analyzed by flow cytometry. (A) CD69 and CD103 as canonical markers of resident memory T cells are shown in blood and BAL from infected volunteers. Mean ± SEM frequencies are shown. (B) Memory markers CD45RA and CCR7, (C) costimulatory markers CD27 and CD28, (D) homing markers CCR5 and CD62L, and (E) cytotoxicity markers perforin and granzyme B are shown in blood and BAL. In CD69+ (Trm) and CD69– (non-Trm) subsets from BAL, frequencies of (F) CCR5 and CD62L, (G) CD45RA and CCR7, and (H) CD27 and CD28-expressing CD4+ T cell are shown. P values for paired t test are shown. **P < 0.01.

Figure 5. RSV-specific CD4+ T cells primarily express 1 or 2 Th1 cytokines.

(A) PBMCs from RSV-challenged subjects (n = 18) were assayed by IFN-γ ELISpot stimulated with whole RSV. Median and individual data points are shown with P values for 2-tailed Mann-Whitney test comparing infected and uninfected groups, and 2-tailed Wilcoxon’s matched-pairs tests comparing time points. ***P < 0.001. (B) Median and IQR of T helper cell subsets expressing different cytokines are shown. (C–E) PBMCs from infected subjects (n = 10) on day 10 after infection were cultured in vitro with live RSV or media for 24 hours and Brefeldin A added 4 hours before staining with fixable viability dye and with antibodies against CD3, CD4, and (C) IFN-γ, IL-2, and TNF; (D) IL-4, IL-5; or (E) IL-17A and IL-17F. Representative FACS plots from a single participant are shown.

Figure 6. Immunodominant CD4+ T cell epitopes are found in the surface F and G proteins.

Fresh PBMCs from individuals infected with RSV (n = 10) were assayed by IFN-γ ELISpot using overlapping peptides covering the RSV proteome. (A) ELISpot responses to peptide pools on days 0, 10, and 28 after infection are arranged according to the originating protein. (B) Total ELISpot responses to peptide pools on days 0, 10, and 28 after infection are shown. Median and individual counts are shown with P values for 2-tailed Wilcoxon’s matched-pairs tests comparing time points. (C) Median and IQR ELISpot responses to each peptide pool are shown in RSV genome order. (D) Median and IQR ELISpot responses to G-DDF (n = 3) and (E) F-EFY (n = 3) peptides are shown. P values are for 2-tailed Wilcoxon’s matched-pairs tests. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 7. Immunodominant epitopes from F and G proteins induce proliferation and cytokine production in vitro.

(A) CFSE-stained PBMCs from uninfected HLA-matched donors were cultured with F-EFY, G-DDF, or DMSO and supplemented with IL-2 and IL-7. (B) CFSE dilution was analyzed by flow cytometry (n = 9). PBMCs from 6 representative donors expressing differing HLA alleles are shown. Blue histograms show cultures with peptide epitopes and red show DMSO negative controls. Median and individual data points of frequencies of CFSElo CD4+ T cells after stimulation are plotted. (C) PBMCs (n = 9) were restimulated after 10 days with peptide epitopes and assayed by intracellular staining for IFN-γ and flow cytometry. Medians and individual data points of IFN-γ+CD4+ T cell frequencies are plotted. Two-tailed Mann-Whitney test was used to compare RSV- and DMSO-only responses (**P < 0.01).

Figure 8. Epitope-specific CD4+ T cells preferentially accumulate and express CXCR3 in the airway during acute RSV infection.

PBMCs from DPA1*01:03/DPB1*04:01–expressing individuals infected with RSV (n = 5) before infection and on day 10 were stained using DPB1*04:01/G-DDF tetramer and analyzed by flow cytometry. (A) FACS plots from 1 representative donor are shown, gated on CD3+ lymphocytes. (B) Cumulative data from blood (n = 11) and BAL (n = 7) are shown with P values for 2-tailed Mann-Whitney test comparing infected and uninfected groups, and 2-tailed Wilcoxon’s matched-pairs tests comparing time points. *P < 0.05, **P < 0.01, ***P < 0.001. PBMCs and BAL cells were costained with phenotypic markers of (C) T cell resident memory (CD69 and CD103), (D) proliferation/activation (Ki-67 and CD38), (E) memory subsets (CD45RA and CCR7), (F) costimulation (CD27 and CD28), homing receptors (G) CCR5 and CD62L and (H) CXCR3 and CCR4. FACS plots from 1 representative donor are shown, gated on CD3+CD4+ T lymphocytes. Red dots represent tetramer+ cells, and gray contours show total CD4+ T cells. The median frequencies of each subset are shown. Where not indicated with asterisks, differences were not statistically significant.