As the COVID-19 pandemic continues to grow, scientists are focusing on understanding how the host immune system responds to the virus in order to better shape public health responses and develop effective vaccines. A new study published on the preprint server medRxiv* in September 2020 reports the T cell memory response to the SARS-CoV-2 spike protein, recombinant nucleocapsid protein, and other pooled peptides derived from convalescent patients.
The current study included 13 and 1 patient who had tested positive for SARS-CoV-2 and SARS-CoV, respectively. The median age was 53, with all participants being within 90 days of the first symptom.
Antibodies in COVID-19
Adaptive immunity is vital for viral clearance and protective immunity. Most earlier studies show the development of specific IgGs against the spike protein and the receptor-binding domain (RBD) of the virus, in addition to neutralizing antibodies within 2-4 weeks of the infection. Some researchers suggest that antibody responses are transient in mild infection, but others claim to have detected ongoing antibody responses for at least 3-4 months. While IgG antibody peaks and then declines slowly, IgA responses are early and short in duration.
T Cell Responses Required for Protective Immunity
Not only is cellular immunity required for virus neutralization, but CD4 and Tfh responses are a crucial component of antibodies with high affinity for the virus and long-lasting protection, making them essential for protective immunity. Earlier studies on SARS-CoV show that T cells specific to the virus last for 11 years or more, while specific antibody responses wane over 3-5 years.
T cells reactive to the N peptide at 17 years from SARS-CoV infection cross-react with the N peptides from the current virus SARS-CoV-2. Most COVID-19 cases show both CD4 and CD8 responses to varying viral antigens in both acute infection and in the convalescent phase, mostly Th1 type. In many cases, severe illness is associated with the most robust immune response, making it a matter of uncertainty as to why immune clearance does not occur in all cases.
Strong CD4+ T Cell Response
The researchers detected CD4 T cells against the S antigen in just over half and three-quarters of donors as shown by the production of IFN-γ and TNF-α, respectively. Against the N protein, the percentage was ~40% and 60% based on these two cytokines, respectively.
Thus, over 90% of convalescent donors had specific CD4 T cell responses against one or more viral antigens at 4-10 weeks from symptom onset, with an IL-2, TNF-α response being predominant. The low CD8 T cell response might be because of the poor recognition of whole protein antigens by these cells.
None of the samples from healthy donors in March 2020 showed T cell responses to any of the viral proteins, but they did in almost every case of influenza antigen exposure. This absence of cross-reactivity is attributed to the insensitivity of the assay.
Multifunctional CD4 T Cells in Influenza
When it came to cytokines, 80% of anti-S CD4 T cells produced only a single cytokine, while in influenza, multiple cytokines were produced. In fact, the frequency of CD4 T cells producing all cytokines was ~9% in influenza vs. ~3% in spike-exposed T cells. The former also had a markedly higher proportion of cells that produced IFN-γ to those secreting TNF-α, at all grades of disease.
Among the latter phenotype, for both S and influenza reactive CD4 cells, most were central memory T cells. The specific recall was seen for all cases exposed to influenza but ~70% and 85% for spike and N proteins, respectively. Though these cells were activated, not all produced cytokine, however.
Overall, therefore, the findings show that T cells directed against SARS-CoV-2 in early convalescence are mostly TCM in phenotype, producing more TNF-α than IFN-γ, indicating an atypical antiviral response unlike that seen in influenza.
Recall Based on Cytokines
The researchers also found that all the peripheral blood mononuclear cells (PBMCs) had anti-N TNF-α responses but only 50% against the spike and 92% against the influenza antigens. Moreover, the TNF-α responses against N were much higher than against S or influenza, while IL-2 was produced only against the latter two.
However, IL-10 was produced in response to all three, the highest being against N. IL-13 was found to be produced against both S and the influenza antigen, but the pro-inflammatory IL-6 only in response to S and N, and not in influenza. In fact, influenza exposure resulted in the highest IFN-γ production compared to either TNF-α or IL-10.
This reflects a Th1 profile, with a constant and high level of IL-10 secretion when exposed to N protein in all the convalescent COVID-19. Both pTfh and T effector cell responses correlate with neutralization titers, and the latter is highest in those who have recovered after the most severe disease.
At all grades of disease severity, however, CD8 T cell responses were significantly lower than for CD4 T cells. The former is the classical T effector cytotoxic cells that kill virus-infected cells. However, the proliferating activated CD4 granzyme+ T cells play a large role in human antiviral defenses and are present in higher proportion in samples from individuals who had moderate to severe disease. The asymptomatic donor also, however, had a high proportion of these cells.
The anti-S IL-2 pTfh cells were present at frequencies corresponding to the frequency of these proliferating spike-reactive IFN-g/granzyme B+ T cells.
The current study used recombinant proteins to assess the T cell response to antigens presented by cells and made use of the fully glycosylated spike protein to simulate the natural presentation of this antigen on the virus.
The overall T cell response to infection with SARS-CoV-2 is strong, with predominating Th1 cell responses. Inflammatory CD4 responses may contribute to the severity of the disease. The researchers also found that in some donors, T follicular helper cells in peripheral blood (pTfh) produce IL-2 and express CCR7+CXCR5+ cells in response to SARS-CoV-2 antigens. The frequency of these cells is strongly related to serum neutralization assays as well as to RBD-IgA, but overall, this response is lower than in influenza.
The CD4 memory T cell response was strong in over 90% of cases of exposure to the influenza H1N1 strain. The reason is not the low dose of the spike protein or the use of the whole influenza virus vs. recombinant spike proteins. However, it could be due to the fact that only two SARS-CoV-2 proteins were used.
Secondly, the predominance of IL-2 and TNF-α over IFN-γ in CD4 T cells responses to this virus contrasts with the IFN-γ-dominant response in influenza. This change in the Th1 profile may result in increased inflammation and impaired viral clearance compared to influenza. Moreover, this pattern is independent of disease severity.
However, specific T cells secreting multiple cytokines are significantly fewer in severe COVID-19 and may mirror T cell exhaustion irrespective of the duration of exposure. The high IL-10 secretion is probably from monocytes or NK cells responding to activated T cells, while IL-6 is from monocytes. The high IL-10 may cause poor antigen presentation and immunosuppression, and this finding may shape future vaccine design.
Most CD4 T cells appeared to be of the central memory cell subtype with a lower ratio of interferon-gamma to TNF-α, compared to the changes in the same donors exposed to influenza. This difference is independent of the severity of the disease.
The SARS-CoV-2 infection resulted in the production of T cells with reduced functional range compared to those found in influenza, perhaps the outcome of T cell exhaustion. N protein exposure resulted in high IL-10 levels, which may have caused immunosuppression. Granzyme B/IFN-γ-producing CD4 and CD8 T cells showed a proliferative response on exposure to pooled peptides, predominantly the former. T follicular helper cell elevations to spike or N protein, and anti-RBD IgA, were correlated with neutralizing activity in serum as well.Source:Web