cytotoxic T cells

by Gertrud U. Rey

Long COVID is a chronic manifestation of SARS-CoV-2 infection, and it is most commonly characterized by lingering fatigue, brain fog, memory impairment, and confusion. Although it is unclear how the viral infection leads to long COVID, experts speculate that one or more of the following factors may contribute: an inability to successfully clear virus, a reactivation of latent viruses, a disturbance of the gut microbiome, continuing inflammation, and/or autoimmunity.

Yale University researcher Akiko Iwasaki and colleagues recently explored some of these hypotheses in an attempt to identify diagnostic biomarkers associated with long COVID. The study involved four groups of participants, with the experimental group consisting of both vaccinated and unvaccinated individuals with long COVID. The other three groups served as three separate types of controls:

healthy, vaccinated, uninfected individuals;
healthy, unvaccinated, previously infected individuals without long COVID; and
healthy, vaccinated, previously infected individuals without long COVID.

The investigators obtained blood samples from all participants and analyzed the samples for the presence of specific immune cells. They found that compared to control groups, long COVID participants had lower levels of conventional dendritic cells and memory T helper cells. Conventional dendritic cells typically activate cytotoxic T cells, which in turn kill infected cells. Memory T helper cells are a central element of the adaptive immune response, where they help orchestrate downstream immune functions upon recognition of antigen. Long COVID participants also had increased numbers of “exhausted” T cells, which are no longer functional or capable of eliminating virus. These results suggested that people with long COVID may have insufficient numbers of immune cells able to inactivate virus, thus allowing viruses to linger and continue replicating and circulating. Whether this assumption is true, and whether long COVID participants do indeed have persistently circulating virus, is subject to ongoing analysis.

Previous studies have shown that patients with severe COVID-19 can have increased levels of functional antibodies directed against self antigens that circulate outside of cells (i.e., extracellular autoantibodies), suggesting that SARS-CoV-2 infection can cause autoimmune disorders. To identify a possible connection between autoimmunity and long COVID, the authors screened the collected blood samples for autoantibodies using a technique called Rapid Extracellular Antigen Profiling (REAP). Among other applications, REAP allows one to assess antibody reactivity against a panel of extracellular human proteins known to contribute to well-studied autoimmune disorders. Interestingly, long COVID participants did not have increased levels of autoantibodies compared to control groups, suggesting that the autoantibodies detected in patients with severe COVID-19 may only be present during the acute phase of disease. However, REAP only assesses antibodies directed to extracellular self proteins and does not provide any information regarding intracellular or non-protein-specific autoantibodies. Therefore, one cannot rule out a role for autoimmunity in long COVID.

The authors also used REAP to detect antibodies against various viruses. They found that long COVID participants had antibodies against several other viruses in addition to SARS-CoV-2, including Epstein-Barr virus (EBV), a herpesvirus that is well known for causing infectious mononucleosis (i.e., “mono”), a condition that is often compared to encephalomyelitis/chronic fatigue syndrome (ME/CFS) and long COVID. However, because most adults have antibodies against EBV and there was no significant difference in the percentage of EBV antibody-positive participants between experimental and control groups, it is unlikely that a positive EBV antibody status contributes to the risk of developing long COVID.

The most interesting observation in this study was that compared to control groups, long COVID participants had about 50% lower levels of the steroid hormone cortisol. Because cortisol is a potent anti-inflammatory agent, it is feasible that a shortage of cortisol would lead to persistent inflammation and the long-term tissue damage associated with inflammation. In line with this reasoning, cortisol levels were highest in healthy, vaccinated, uninfected controls (group 1 controls); lower in healthy, unvaccinated, previously infected individuals without persistent symptoms (group 2 controls); and lowest in long COVID participants. Using machine learning methods, the authors determined that cortisol deficiency was the single most significant predictor of a long COVID diagnosis. Low cortisol has also been implicated in ME/CFS, and treatment with hydrocortisone can provide some relief from symptoms.

The authors are careful to note that the small sample size of 215 participants is a considerable limitation of this study. Nevertheless, the results offer some valuable insight that may apply to other chronic conditions. In contrast to public perception, long-term symptoms following an acute viral infection are not unique to COVID-19. Unexplained chronic syndromes involving similar symptoms to long COVID have also been reported after dengue virus, poliovirus, SARS-CoV, Chikungunya virus, West Nile virus, Ross River virus, Coxsackie virus, and influenza virus infections. Because these syndromes are associated with high public health and economic burdens, more work needs to be done to clarify their underlying mechanisms.

There are numerous advantages to administering a vaccine by the intranasal route: it is simple, painless, and does not require special training. The last time I discussed SARS-CoV-2 vaccine candidates for potential intranasal delivery, the idea was still fairly new. However, multiple new studies are now showing that delivery of a SARS-CoV-2 vaccine into the nose induces immune responses that are superior to those observed after delivery into the muscle.

The authors of one such recent study compared the efficacy of a single dose of two different adenovirus-based SARS-CoV-2 vaccine candidates after intramuscular injection to that after administration by the intranasal route. Both vaccines encoded three viral antigens: 1) the full-length S1 domain of the SARS-CoV-2 spike protein, which contains the region that binds to the host cell receptor ACE2; 2) the full-length nucleocapsid protein, which is essential for binding and packaging the viral genome; and 3) a truncated version of the RNA-dependent RNA polymerase (RdRp), the enzyme responsible for copying the viral genome during infection. The three antigens were inserted into either one of two adenovirus vectors: a virus of human origin to produce “Tri:HuAd” or a virus of chimpanzee origin to produce “Tri:ChAd.” Most currently approved SARS-CoV-2 vaccines only encode the full-length spike protein, but not nucleocapsid or RdRp. These two additional genes were likely included in the present study because nucleocapsid is the most abundant viral protein in infected cells, and because RdRp is well conserved among coronaviruses, thus potentially leading to broader immune responses.

To assess the antibody responses induced by either vaccine candidate, the authors immunized mice with a single dose of Tri:HuAd or Tri:ChAd into the nose or into the muscle and collected serum and lung wash samples 2, 4 and 8 weeks later. The collected samples were then tested for the presence of IgG antibodies, which provide the majority of antibody-based immunity. Analysis of sera collected at 4 weeks revealed that mice immunized intranasally with either vaccine had significantly higher levels of spike protein-specific IgG than did mice that were immunized intramuscularly, and these high levels were sustained through 8 weeks post-vaccination. However, only lung wash samples from mice immunized with either vaccine intranasally contained detectable levels of spike protein-specific IgG at the 4 week time point, suggesting that only intranasal immunization led to IgG production in the airways. Interestingly, the airways of Tri:ChAd recipients harbored almost twice as much spike-specific IgG as those of Tri:HuAd recipients, a difference that was not observed in the serum, suggesting that the vector of chimpanzee origin induced better local immune responses when administered via the nose. Additionally, only the lung wash samples from intranasal Tri:ChAd recipients contained any observable anti-spike IgA antibodies, which are prevalent in mucosal surfaces and represent a first line of defense against invasion by inhaled and ingested pathogens. These results suggested that both vaccines, but Tri:ChAd in particular, induced better immune responses when they were delivered intranasally.

Lung wash samples were also analyzed for vaccine antigen-specific T cells by detecting characteristic cell surface proteins on these T cells, and for specific cytokines produced by the T cells in response to their exposure to peptides encoding each of the three vaccine-encoded antigens: S1, nucleocapsid, or RdRp. This analysis revealed that when the vaccine was injected into the muscle, neither Tri:HuAd nor Tri:ChAd induced detectable airway antigen-specific cytotoxic T cells, which are capable of killing virus-infected cells. In contrast, both vaccines induced significant numbers of cytotoxic T cells in the airways when they were delivered intranasally. Although the detected T cells were specific for all three antigens, S1-specific cells were more abundant, and they were also multifunctional in that they produced more than one cytokine characteristic of cytotoxicity. However, in the context of intranasal vaccination, the Tri:ChAd vaccine induced higher total numbers of cytotoxic T cells than the Tri:HuAd vaccine, suggesting that a single dose of Tri:ChAd delivered intranasally is highly effective at inducing vaccine antigen-specific cytotoxic T cells in the respiratory tract.

Interactions between lung macrophage cells and SARS-CoV-2 can increase the expression of the host-derived peptide MHCII on macrophage surfaces, allowing these cells to become a critical component of “trained innate immunity.” This altered innate immunity results from a functional modification of innate immune cells that leads to an antibody-independent temporary immune memory that is characterized by an accelerated response to a second challenge with the same pathogen. Neither Tri:HuAd nor Tri:ChAd induced detectable MHCII expression on macrophages in the lungs of mice when delivered intramuscularly. In contrast, both vaccines induced increased levels of MHCII-expressing lung macrophages 8 weeks after immunization when delivered intranasally, with a higher number of these cells in the lungs of Tri:ChAd recipients, suggesting that intranasal immunization with Tri:ChAd induces strong trained innate immunity.

The authors next tested the ability of the vaccines to protect mice from SARS-CoV-2 infection by intentionally infecting immunized mice with a lethal dose of SARS-CoV-2 at 4 weeks after immunization, and monitoring them for clinical signs of illness in terms of weight loss. Unvaccinated control animals and 80% of animals immunized intramuscularly with either vaccine were not protected from infection or disease and died by 5 days post-infection. In contrast, animals immunized with Tri:HuAd showed a temporary and slight weight loss, but rebounded over the course of two weeks, while animals immunized with Tri:ChAd did not lose weight at all for the duration of the two-week time course. This result correlated with reduced viral load and reduced lung pathology at 14 days post-infection in mice immunized intranasally compared to unvaccinated controls and intramuscular vaccine recipients, with Tri:ChAd recipients exhibiting the most reduced viral burden and lung pathology.

Having determined that intranasal administration of Tri:ChAd induced the best immune responses, all subsequent experiments were carried out using this vaccination strategy. To assess the individual contributions of B cells and T cells in the vaccine-induced immune response, the authors immunized two groups of mice: one group that was genetically deficient in B cells and a second group of genetically wild type mice that were depleted of T cells by receiving injections of antibodies against T cell-characteristic cell surface molecules. Because T cells are needed for B cell stimulation and production of antibodies, the antibody injections were done on day 25 after immunization to ensure that vaccine-induced B cell functions and antibody production remained intact in the T cell-depleted mice. At 28 days after vaccination, both groups of animals were infected with SARS-CoV-2 and monitored for weight loss. Unvaccinated mice that were either B cell-deficient or T cell-depleted and control unvaccinated mice with normal B and T cell functions were subject to drastic weight loss. In stark contrast, neither B cell-deficient nor T cell-depleted mice that had been intranasally immunized with Tri:ChAd showed any weight loss throughout the 2-week experimental time course, despite the fact that both groups of animals had significantly elevated viral burden in the lung. This result suggested that Tri:ChAd was immunogenic enough to compensate for the lack of B or T cells and protect the mice from clinical disease after SARS-CoV-2 infection.

The authors next wanted to analyze the role of trained innate immunity in the absence of B and T cell immunity, so they immunized two groups of mice. The first group consisted of B cell-deficient animals that were depleted of T cells shortly after vaccination. Because T cells induce vaccine-mediated trained innate immunity early after vaccination, this strategy produced mice that lacked B cells, T cells, and trained innate immunity. The second group consisted of B cell-deficient animals that were depleted of T cells starting at day 25 post-vaccination to allow for induction of trained innate immunity and thus produce mice that lacked B cells and T cells, but possessed intact trained innate immunity. At 28 days after vaccination, both groups of animals were infected with SARS-CoV-2 and monitored for weight loss and lung pathology. Unvaccinated control mice with functional B cells, T cells, and trained innate immunity and unvaccinated mice lacking B cells, T cells, and trained innate immunity showed severe weight loss and severe lung pathology. In contrast, vaccinated mice lacking B cells and T cells but possessing trained innate immunity appeared healthy with no weight loss and no pathology in the lungs, suggesting that the immune protection observed in animals lacking B cells, T cells, or both, was likely mediated by trained innate immunity.

The study demonstrated similar results after infection of mice with alpha and beta variants of SARS-CoV-2; however, it would be interesting to see if these vaccines have similar efficacy against the later variants, including delta and omicron. Although these results are promising, it is important to remember that mice can be poor predictors of human disease outcomes. Nevertheless, I look forward to seeing the results of the clinical trial that is currently underway to compare the relative efficacies of Tri:HuAd and Tri:ChAd following intranasal delivery in humans.

[For a more detailed discussion of this paper, please check out TWiV 867.]

Why do Some People Develop Long COVID?

An Intranasal SARS-CoV-2 Vaccine Candidate