SARS-CoV-2 mutations may evade T cell immunity
A potentially disturbing new study published in December 2020 on the bioRxiv* preprint server suggests that the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) that is causing the current coronavirus disease 2019 (COVID-19) pandemic may be undergoing strategic mutations that affect the host immune capacity to recognize and combat the pathogen via T effector cells.
Broad immune response
The virus elicits a broad spectrum of immune responses, both innate and adaptive. CD8+ cytotoxic T lymphocyte (CTL) responses occur in these patients, in response to the recognition of a number of antigenic epitopes. The CTLs are very important in clearing the infection as they kill the cell infected by the virus.
This action is triggered by the recognition of the viral peptides displayed on the host cell surface, after they are specifically presented by the right human leukocyte antigen (HLA) in order to activate the corresponding CTLs. The HLA group of antigens is produced by the genes encoding the class I major histocompatibility complex (MHCI).
There is much evidence that CTL-control of some RNA viruses prompts the emergence of viral mutations that prevent MHC-1 restricted recognition of viral antigens with subsequent killing by the CTLs.
Study details
The current study aimed to understand the effect of SARS-CoV-2 mutations on viral peptide presentation via MHC-I. They used deep sequencing methods with the viral genome and bioinformatics to analyze the results. The genomes came from viral isolates from 747 patient samples.
The researchers examined 27 CTL epitopes that were shown to be presented by the common subtype HLA-A*02:01, having an allele frequency of 0.29 in Austria, and by the minor subtype HLA-B*40:01 (allele frequency 0.03-0.05 in Austria).
They found around 200 mutations that resulted in amino acid substitutions at CTL epitopes, all present at frequencies of 0.02 or more, in around 230 samples. The frequencies of 33 of them were between 0.1 and 0.5. Nine mutations had become the default allele in 53 samples from different patients. Some of the epitopes overlapped, resulting in 207 different epitopes. Of these, 27 were on anchoring residues, while auxiliary residues were affected by 14 mutations. Both these types of amino acids are essential for MHC-I peptide presentation.
Independent emergence of mutations
Multiple variants were found that had emerged independently in different individuals following infection.
The researchers looked at fixed mutations in over 145,000 sequences retrieved from the Global Initiative for Sharing All Influenza Data (GISAID) database. This global dataset showed mutations in up to 7.34% of epitopes. Each of the 27 CTL epitopes had 10-11,700 non-synonymous mutations, the average being 807. The low-frequency mutations identified in the current analysis were also found in GISAID as fixed mutations.
An alanine to valine mutation in one epitope was found in over 75 of the sequences analyzed. This particular allele was first reported in June 2020, but is now a defining mutation of the 20A.EU1 subclade. This was, however, found to have been present in samples collected from March to April, though at low frequency. This supports the emergence of the same mutation independently in different individuals.
Interestingly, longitudinal sampling from the same patients showed that mutant epitopes arose at a later time period of the infection. This suggests that positive selection pressure due to CTL effector activity was shaping these mutations.
Weaker binding strength for CTL epitopes
Modeling studies to estimate the binding strength of the peptides in the wildtype and mutant viruses to HLA-A*02:01 and HLA-B*40:01 showed weaker peptide binding to MHC-I.
The researchers chose 11 and 17 wildtype and mutant peptides, respectively, that were predicted to have lower binding strength, testing them against recombinant HLA-A*02:01 or HLA-B*40:01 proteins. They found that 9/11 wildtype peptides bound to these HLA antigens, stabilizing their structure by strong binding at physiological temperature.
Of the mutants, however, 11 had reduced binding and stabilizing capacity for MHC-I. MEVTPSGTWL is a peptide that binds only to the minor allele HLA-B*40:01 and not to HLA-A*02:01. Other mutants showed weak or no binding, respectively. One of the predicted CTL epitopes was not bound by wildtype or mutant peptides.
Mutations cause MHC-I breakdown
The researchers then constructed peptide-loaded tetramers of both HLA antigens, for both wildtype and mutant peptides. They found that the latter tetramer type bound with cognate T cells in response to T cell receptor activation at 4°C but not at 37°C. The most likely reason was peptide loss leading to the breakdown of MHC-I structure.
Mutations weaken cellular immune response
They also looked at peptide-specific effector cell responses in peripheral blood mononuclear cells (PBMCs), obtained from COVID-19 patients with either of these alleles. Stimulating HLA-matched PBMCs with these peptides confirmed they were actual T cell epitopes. These virus-stimulated T cells showed interferon-gamma secretion. However, mutant peptides reduced the immune response, with fewer tetramer-positive CTLs and lower IFN-gamma secretion.
What are the implications?
It is hypothesized that the SARS-CoV-2 ORF8 protein reduces the expression of MHC-I molecules on the host cell, but more study is required. The findings of this current study show that effector T cells may give rise to SARS-CoV-2 mutations which escape immune surveillance. “These results imply that mutations found in SARS-CoV-2 isolates promote immune escape from HLA-dependent recognition by CTLs.”
The low frequencies of the non-synonymous mutations showed that they had not become fixed, perhaps because of the short periods of infection with this virus compared to HIV or HCV. Also, the fact that there are many separate HLA profiles within the population affects the spread of the virus, since for each individual, there are different sets of CTL epitopes that are triggered by the infection. Thus, different selection pressures were exerted by these varying subsets, that shape different viral mutational escapes. More work is required to understand how single epitope mutations affect virus control.
Our results highlight the capacity of SARS-CoV-2 to evade adaptive immune responses and provide further evidence for the impact of endogenous CTL responses and their participation in conferring protection in natural and vaccine-induced immunity.”