Alterations in T and B cell function persist in convalescent COVID-19 patients

Abstract

Background: Emerging studies indicate that some coronavirus disease 2019 (COVID-19) patients suffer from persistent symptoms, including breathlessness and chronic fatigue; however, the long-term immune response in these patients presently remains ill-defined.

Methods: Here, we describe the phenotypic and functional characteristics of B and T cells in hospitalized COVID-19 patients during acute disease and at 3-6 months of convalescence.

Findings: We report that the alterations in B cell subsets observed in acute COVID-19 patients were largely recovered in convalescent patients. In contrast, T cells from convalescent patients displayed continued alterations with persistence of a cytotoxic program evident in CD8⁺ T cells as well as elevated production of type 1 cytokines and interleukin-17 (IL-17). Interestingly, B cells from patients with acute COVID-19 displayed an IL-6/IL-10 cytokine imbalance in response to Toll-like receptor activation, skewed toward a pro-inflammatory phenotype. Whereas the frequency of IL-6⁺ B cells was restored in convalescent patients irrespective of clinical outcome, the recovery of IL-10⁺ B cells was associated with the resolution of lung pathology.

Conclusions: Our data detail lymphocyte alterations in previously hospitalized COVID-19 patients up to 6 months following hospital discharge and identify 3 subgroups of convalescent patients based on distinct lymphocyte phenotypes, with 1 subgroup associated with poorer clinical outcome. We propose that alterations in B and T cell function following hospitalization with COVID-19 could affect longer-term immunity and contribute to some persistent symptoms observed in convalescent COVID-19 patients.

Funding: Provided by UKRI, Lister Institute of Preventative Medicine, the Wellcome Trust, The Kennedy Trust for Rheumatology Research, and 3M Global Giving.

Keywords: B cells; COVID-19; T cells; convalescent patients; long COVID; viral Infection.

Graphical abstract

Figure 1

Alterations in B cell subsets during acute COVID-19 are recovered upon convalescence (A) Cumulative data show ex vivo frequency of CD19⁺ B cells in healthy individuals (n = 38) and COVID-19 patients with mild (n = 24), moderate (n = 26), and severe (n = 12) disease and at convalescence (n = 83). (B) Cumulative data show Ki-67 expression by B cells in healthy individuals (n = 28) and COVID-19 patients with mild (n = 13), moderate (n = 15), and severe (n = 9) disease and at convalescence (n = 75). (C and D) Representative flow cytometry plots and cumulative data show frequencies of naive (CD27⁻IgD⁺), unswitched memory (CD27⁺IgD⁺), switched memory (CD27⁺IgD⁻), and double-negative (CD27⁻IgD⁻) B cells in healthy individuals (n = 38–40) and COVID-19 with mild (n = 22–24), moderate (n = 25–26), and severe (n = 12–13) disease and at convalescence (n = 78–80). (E and F) Representative flow cytometry plots and cumulative data show ex vivo frequency of CD24^hiCD38^hi transitional B cells and CD24^intCD38^int mature B cells in healthy individuals (n = 37) and COVID-19 patients with mild (n = 24), moderate (n = 23), and severe (n = 11) disease and at convalescence (n = 80). (G) tSNE projection of flow cytometry panel visualizing B cell subsets in PBMCs. Representative images for healthy individuals, severe COVID-19 patients, and convalescent patients. Key indicates cell subsets identified on the image. (H) Cumulative data show frequency of CD27^hiCD38^hi plasmablasts in healthy controls (n = 38) and COVID-19 patients with mild (n = 23), moderate (n = 23), and severe (n = 12) disease and at convalescence (n = 81). (I) Graph showing correlation between plasmablasts and IgG⁺ (left), IgA⁺ (center), or IgM⁺ (right) B cell frequencies in acute COVID-19 patients. Graphs show individual patient data, with the bar representing median values. In all graphs, open triangles represent SARS-CoV-2 PCR⁻ patients. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, 1-way ANOVA with Kruskal-Wallis test with Dunn’s post hoc testing for multiple comparisons or Spearman ranked coefficient correlation test. See also Figures S1 and S2.

Figure 2

Acute alterations in CD4⁺ T cells and persistent alterations in CD8⁺ T cells during COVID-19 (A) Representative FACS plots showing CD45RA and CCR7 staining on CD4⁺ (gated CD3⁺CD8⁻) and CD8⁺ (gated CD3⁺CD4⁻) T cells. (B and C) Graphs showing frequencies of (B) CD8⁺ and (C) CD4⁺ T cells that have a naive (CD45RA⁺CCR7⁺) and TEMRA (CD45RA⁺CCR7⁻) phenotype in healthy individuals (n = 44) and COVID-19 patients with mild (n = 18–19), moderate (n = 18), and severe (n = 8) disease and at up to 6 months of convalescence (n = 83). (D) Graphs showing frequencies of CD8⁺ and CD4⁺ T cells that stain positive for Ki-67 in healthy individuals (n = 28–30), and COVID-19 patients with mild (n = 14), moderate (n = 11–13), and severe (n = 9) disease and at convalescence (n = 81). (E–G) Graphs showing frequencies of (E) CD8⁺Perforin⁺ cells, (F) CD8⁺GranzymeB⁺ cells, and (G) CD8⁺CD107a⁺ cells and CD8⁺GranzymeB⁺ Ki-67⁺ cells in healthy individuals (n = 29–37), and COVID-19 patients with mild (n = 12–17), moderate (n = 12–15), and severe (n = 7–9) disease and at convalescence (n = 81–83). (H and I) Graphs track frequencies of (H) Perforin⁺ and GranzymeB⁺ and (I) Ki-67⁺, CD107a⁺, and GranzymeB⁺Ki-67⁺CD8⁺ T cells in the same COVID-19 patient at acute (gray circles) and convalescent (maroon circles) time points (n = 14). (J) Graph shows frequencies of Tregs within CD4⁺ T cells of healthy individuals (n = 20) and COVID-19 patients with mild (n = 10), moderate (n = 12), and severe (n = 8) disease and at convalescence (n = 82). (K) Graph shows frequencies of Tfh within CD4⁺ T cells of healthy individuals (n = 34) and COVID-19 patients with mild (n = 12), moderate (n = 15), and severe (n = 7) disease and at convalescence (n = 83). (L) Graph shows frequencies of Tfh in individual acute COVID-19 patients with mild (n = 4), moderate (n = 5), and severe (n = 3) disease at their first and last time points of hospitalization. (M) Graph tracks frequency of Tfh CD4⁺ T cells in the same COVID-19 patient at acute (gray circles) and convalescent (maroon circles) time points (n = 14). Graphs show individual patient data, with the bar representing median values. In all graphs, open triangles represent SARS-CoV-2 PCR⁻ patients. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, 1-way ANOVA with Kruskal-Wallis test with Dunn’s post hoc testing for multiple comparisons (for B–G and K) or Wilcoxon matched-pairs signed rank test (I and M). See also Figures S1 and S2.

Figure 3

Changes in cytokine production by lymphocytes during acute and convalescent COVID-19 (A–C) Graphs showing frequencies of CD4⁺ T cells that stain positive for (A) IL-10, (B) IL-17, and (C) IFNγ and TNF-α following 3-h stimulation with PMA and ionomycin in healthy individuals (n = 25–30), acute COVID-19 patients (n = 29–33), and convalescent COVID-19 patients with normal (n = 55–57) or abnormal chest X-ray findings (n = 25–26). (D) Graphs showing frequencies of CD8⁺ T cells that stain positive for IFNγ and TNF-α following 3-h stimulation with PMA and ionomycin in healthy individuals (n = 28), acute COVID-19 patients (n = 24–31), and convalescent COVID-19 patients with normal (n = 54–57) or abnormal chest X-ray findings (n = 21–24). (E and F) Graphs show frequencies of (E) CD4⁺ and (F) CD8⁺ T cells that stain positive for IFNγ and TNF-α following 3-h stimulation with PMA and ionomycin in convalescent COVID-19 patients who initially presented with mild (n = 13–14), moderate (n = 25–28), and severe (n = 34–41) disease. (G and H) Graphs track frequencies of (G) CD4⁺ and (H) CD8⁺ T cells that stain positive for IFNγ and TNF-α in the same COVID-19 patient at acute (gray circles) and convalescent (maroon circles) time points (n = 14). (I) Graphs showing frequencies of CD19⁺ B cells positive for IL-10, IL-6, and TNF-α following 48-h stimulation with CpGB in healthy individuals (n = 22–27), acute COVID-19 patients (n = 22–32), and convalescent COVID-19 patients with normal (n = 52–54) or abnormal chest X-ray findings (n = 24–27). (J) Graphs track frequencies of CD19⁺ B cells that stain positive for IL-10, IL-6, and TNF-α in the same COVID-19 patient at acute (gray circles) and convalescent (maroon circles) time points (n = 11–14). Graphs show individual patient data, with the bar representing median values. In all graphs, open triangles represent SARS-CoV-2 PCR⁻ patients. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, 1-way ANOVA with Kruskal-Wallis test with Dunn’s post hoc testing for multiple comparisons, except for graphs showing CD4⁺TNF-α⁺ and CD8⁺IFNγ⁺ T cells in (E) and (F), where 1-way ANOVA with Holm-Sidak post hoc test was used, or Wilcoxon matched-pairs signed rank test (G, H, and J). See also Figures S3 and S4.

Figure 4

Distinct immune profiles emerge in previously hospitalized convalescent COVID-19 patients (A) Heatmap of indicated immune parameters by row. Each column represents an individual convalescent COVID-19 patient. The patients were clustered using one minus Pearson correlation hierarchical clustering. Significance was determined by 2-way ANOVA, followed by a Tukey’s multiple comparison test. Asterisk next to lymphocyte characteristic indicates a significant difference between patient groups. Dominant immune characteristics of each group are indicated at the bottom of the heatmap. Black and white squares indicate patients displaying a normal (white) or abnormal (black) chest X-ray at follow-up. (B–G) Graphs show patient characteristics and clinical details of convalescent COVID-19 patients in each of the 3 immune groups identified, specifically: (B) age; (C) BMI; (D) sex; (E) severity of acute COVID-19 (with 1 being mild, 2 moderate and 3 severe); (F) length, in days, of hospitalization for acute COVID-19; and (G) time, in days, from hospital discharge to follow-up of convalescent patients. Graphs show individual patient data, with the bar representing median values. ∗p < 0.05, 1-way ANOVA with Kruskal-Wallis test with Dunn’s post hoc testing for multiple comparisons.