Fig. 3: The frequency of circulating functional CXCR3⁺ T_FH cells in convalescent individuals who had severe and non-severe COVID-19 is positively correlated with neutralization titres.

a–d, The frequency of circulating total T_FH cells (a), CXCR3⁺ T_FH cells (b) and CXCR3⁻ T_FH cells (c), as well as the ratio of CXCR3⁺/CXCR3⁻ T_FH cells (d) in healthy controls (HC) (n = 17), severe groups (n = 10) and non-severe groups (n = 16). The gating strategy for the T_FH cells (PD-1⁺CXCR5⁺CD4⁺ T cells) is presented in Extended Data Fig. 8a, and a comparison of T_FH-cell subsets between healthy controls (n = 17) and individuals who recovered from COVID-19 (n = 26) is shown in Extended Data Fig. 8b. NS, not significant. e–h, Correlation analysis of the neutralization titres of antibodies and the frequencies of total T_FH cells (e), CXCR3⁺ T_FH cells (f) and CXCR3⁻ T_FH cells (g), as well as the neutralization titres of antibodies and the ratio of CXCR3⁺/CXCR3⁻ T_FH cells (h) in individuals who recovered from COVID-19 (n = 26). Neutralization titres were log₁₀-transformed. i–l, Spike-specific circulating T_FH-cell response after antigen stimulation. Spike-specific CD154⁺ T_FH cells (n = 5) (i) and OX40⁺CD25⁺ T_FH cells (n = 6) (j), and intracellular IL-21 (n = 5) (k) and IFNγ (n = 6) (l) were measured using flow cytometry after stimulation. For a–d, data are median ± IQR (25–75%). One-way analysis of variance was used to compare the difference between multiple groups, and Tukey’s multiple-comparisons test was used to compare differences within the groups. For e–h, Spearman’s rank correlation coefficient was used to describe the association between the frequency of T_FH-cell subsets and the neutralization titres. For i–l, paired t-tests were used to analyse the difference in the frequency of spike-specific T_FH-cell subsets from COVID-19-convalescent individuals and healthy controls after stimulation. P < 0.05 was considered to be a two-tailed significant difference.

Source data