Accelarated immune ageing is associated with COVID-19 disease severity

Abstract

Background: The striking increase in COVID-19 severity in older adults provides a clear example of immunesenescence, the age-related remodelling of the immune system. To better characterise the association between convalescent immunesenescence and acute disease severity, we determined the immune phenotype of COVID-19 survivors and non-infected controls.

Results: We performed detailed immune phenotyping of peripheral blood mononuclear cells isolated from 103 COVID-19 survivors 3-5 months post recovery who were classified as having had severe (n = 56; age 53.12 ± 11.30 years), moderate (n = 32; age 52.28 ± 11.43 years) or mild (n = 15; age 49.67 ± 7.30 years) disease and compared with age and sex-matched healthy adults (n = 59; age 50.49 ± 10.68 years). We assessed a broad range of immune cell phenotypes to generate a composite score, IMM-AGE, to determine the degree of immune senescence. We found increased immunesenescence features in severe COVID-19 survivors compared to controls including: a reduced frequency and number of naïve CD4 and CD8 T cells (p < 0.0001); increased frequency of EMRA CD4 (p < 0.003) and CD8 T cells (p < 0.001); a higher frequency (p < 0.0001) and absolute numbers (p < 0.001) of CD28^-ve CD57^+ve senescent CD4 and CD8 T cells; higher frequency (p < 0.003) and absolute numbers (p < 0.02) of PD-1 expressing exhausted CD8 T cells; a two-fold increase in Th17 polarisation (p < 0.0001); higher frequency of memory B cells (p < 0.001) and increased frequency (p < 0.0001) and numbers (p < 0.001) of CD57^+ve senescent NK cells. As a result, the IMM-AGE score was significantly higher in severe COVID-19 survivors than in controls (p < 0.001). Few differences were seen for those with moderate disease and none for mild disease. Regression analysis revealed the only pre-existing variable influencing the IMM-AGE score was South Asian ethnicity ([Formula: see text] = 0.174, p = 0.043), with a major influence being disease severity ([Formula: see text] = 0.188, p = 0.01).

Conclusions: Our analyses reveal a state of enhanced immune ageing in survivors of severe COVID-19 and suggest this could be related to SARS-Cov-2 infection. Our data support the rationale for trials of anti-immune ageing interventions for improving clinical outcomes in these patients with severe disease.

Fig. 1

CD8 T cell subset distribution post-COVID-19 infection. A Gating strategy used to analyse markers subsets within CD4^+ve and CD8^+ve T cells; naïve (CCR7^+veCD45RA^+ve); central memory (CCR7^+veCD45RA^−ve), effector memory (CCR7^−veCD45RA^−ve) and terminal differentiated effector memory re-expressing RA, EMRA (CCR7^−veCD45RA^+ve) T cells. Comparison of the systemic percentage of: B CD8 T cells; C Naïve CD8 T cells; D Total memory CD8 T cells; E Central memory CD8 T cells; F Effector memory CD8 T cells G EMRA CD8 T cells. PBMCs were isolated from convalescent COVID-19 patients who had mild (n = 15), moderate (n = 29) and severe (n = 55) disease 3–5 months post-infection, and healthy age and sex-matched controls (n = 59). Data represent individual values, mean (centre bar). Statistical analysis by two-sided Mann–Whitney nonparametric test

Fig. 2

CD4 T cell distribution post-COVID-19 infection. Comparison of the systemic percentage of (A) CD4 T cells; B Naïve CD4 T cells; C Total memory CD4 T cells; D Central memory CD4 T cells; E Effector memory CD4 T cells (F) EMRA CD4 T cells. PBMCs were isolated from convalescent COVID-19 patients who had mild (n = 15), moderate (n = 29) and severe (n = 55) disease 3–5 months post-infection, and healthy age and sex-matched controls (n = 59). Data represent individual values, mean (centre bar). Statistical analysis by two-sided Mann–Whitney nonparametric test

Fig. 3

CD8 T cell senescence and exhaustion post-COVID-19 Comparison of systemic percentage of (A) CD28^−ve CD57^+ve senescent CD8 T cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. B Frequency of KLRG1^+ve senescent CD8 T cells in healthy age and sex-matched controls (n = 51) and mild (n = 15), moderate (n = 24) and severe (n = 46) COVID-19 survivors 3–5 months post-infection. C percentage and (D) absolute numbers of PD1^+ve exhausted CD8 T cells in healthy age and sex-matched controls (n = 33) and severe (n = 38) COVID-19 survivors 3–5 months post-infection. Statistical analysis by two-sided Mann–Whitney non-parametric test. If not indicated p-valueue is not significant

Fig. 4

The impact of COVID-19 on Regulatory T cells and Th17 cells. A Comparison of systemic percentage of Foxp3^+ve CD4 T cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3 months post-infection. B Representative flow cytometry plot showing Foxp3^+ve regulatory T cells in a healthy control and severe convalescent COVID-19 patients. C Comparison of systemic percentage of RORγt^+ve CD4 T cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. D Representative flow cytometry plot showing RORγt^+ve Th17 cells in a healthy control and severe convalescent COVID-19 patient. E Th17/T_reg ratio in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. Statistical analysis by two-sided Mann–Whitney non-parametric test. If not indicated, p value is not significant

Fig. 5

B cell subset distribution post-COVID-19. A Gating strategy used to analyse subsets within CD19^+ve B cells; naïve (CD27^−ve); memory (CD27^+ve), regulatory B cells (CD38^hiCD24^hi) and plasma cells (CD24^−veCD38^+ve) B cells. B Comparison of systemic percentage of total B cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. C Absolute numbers of B cells in healthy age and sex matched controls (n = 39), moderate (n = 14) and severe (n = 46) COVID-19 convalescent patients. Comparison of systemic percentage of (D) memory B cells, (E) Plasma cells, (F) regulatory B cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. Statistical analysis by two-sided Mann–Whitney non-parametric test. If not indicated, p-value is not significant

Fig. 6

NK cells in severe COVID-19 convalescent patients. A Comparison of the systemic percentage of total NK cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. B Absolute numbers of NK cells in healthy age and sex matched controls (n = 39), moderate (n = 14) and severe (n = 46) COVID-19 convalescent patients. C Comparison of the systemic percentage of CD56^dim cytotoxic NK cells in healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection. D, E Comparison of the systemic percentage of senescent NK cells in CD56 dim NK cell pool healthy age and sex-matched controls (n = 59) and mild (n = 15), moderate (n = 29) and severe (n = 55) COVID-19 survivors 3–5 months post-infection (F) Granzyme B expression levels in NK cells in healthy age and sex-matched controls (n = 29) and severe (n = 31) COVID-19 convalescent patients. Statistical analysis by two-sided Mann–Whitney non-parametric test. If not indicated, p value is not significant

Fig. 7

Immunological ageing score (IMM-AGE) and transcriptome signatures in severe COVID-19 convalescent patients. A IMM-AGE scores calculated by the pseudotime algorithm²³ in healthy age and sex-matched controls (n = 39) and mild (n = 15), moderate (n = 33) and severe (n = 42) COVID-19 survivors 3–5 months post-infection. Statistical analysis by two-sided Mann–Whitney non-parametric test. If not indicated, p value is not significant. B (B) A heatmap showing the relative expression levels of a selection of significantly differentially expressed genes between the healthy control and severe COVID-19 groups. The gene IDs can be seen on the X axis. The figure legend colour corresponds to the relative expression levels of a given gene within a group. C An map plot showing the relationships between the pathways associated with the set of significantly differentially expressed genes between healthy control participants and survivors of severe covid-19 infection. Node size denotes the number of genes associated with a specific pathway, with increasing size reflecting a greater number, and colour reflects the adjust p-valuealue