Lung-Resident Memory CD8⁺ T Cells in Human Influenza: How Innate Are They?

Although the lower respiratory tract is the major site of influenza virus infection and the ensuing immunopathology, respiratory immune responses are understudied; such studies are needed to fully understand the mechanisms underlying recovery from respiratory diseases. In this issue of the Journal, Paterson and colleagues (pp. 826–841) used the human experimental influenza virus infection model to define cellular and molecular features of memory CD8⁺ T cells in BAL in comparison to CD8⁺ T cells found in the blood (1). Using peptide-MHC class I tetramers, influenza-specific CD8⁺ T-cell responses were identified and longitudinally tracked in both the blood and BAL before infection and on Days 7 and 28 after infection. Subsequent in-depth analysis of the total CD8⁺ T-cell population by RNA sequencing revealed interesting innate-like transcriptional signatures in BAL after influenza virus infection, providing novel insights into our current understanding of human CD8⁺ tissue-resident memory T cells. These intriguing findings were subsequently supported by confocal imaging of endobronchial tissue biopsies, showing infiltrating CD8⁺ T cells coexpressing innate-like markers during influenza virus infection. With their study, Paterson and colleagues changed the way we think about lung-resident memory CD8⁺ T cells in influenza and possibly also other respiratory diseases.

Experimental infection of healthy volunteers with influenza viruses was first pioneered by McMichael and colleagues in 1983 (2). Paterson and colleagues showed that preexisting circulating influenza-specific CD8⁺ T cells strongly correlated with reduced viral load, which is reminiscent of McMichael’s seminal findings. Since then, both circulating influenza-specific CD4⁺ and CD8⁺ T cells depict strong correlates of protection in influenza disease (3–5). However, the current study delves deeper into the human respiratory tract using bronchoscopy to unravel direct effector mechanisms of protection at the site of infection, which the authors first pioneered using the respiratory syncytial virus (RSV) human infection model (6). This body of work is truly novel and uses extremely rare longitudinal BAL samples, not to mention the massive costs and ethics approvals of conducting experimental virus infection studies in humans.

Very few studies have identified lung-resident influenza-specific CD8⁺ tissue-resident memory T cells postmortem using ex vivo peptide-MHC class I tetramer staining (7–9). These studies included the immunodominant HLA-A2–restricted M1_58–66 (A2/M1₅₈) epitope, which is also used in the current manuscript together with the well-characterized A1/NP_44–52 epitope (10). Prominent BAL and blood tetramer⁺ CD8⁺ T-cell populations displayed similar kinetics, with A2/M1₅₈ and A1/NP_44–52 tetramer⁺ CD8⁺ T cells being minimally detected preinfection and peaking on Day 7 during infection before contracting on Day 28 at convalescence. However, the BAL tetramer⁺ CD8⁺ T cells were of much higher frequencies (up to 4–5% of CD8⁺ T cells) compared with <1% in the blood during infection, similar to the authors’ previous RSV infection study (6) and the postmortem influenza studies (7–9). Nguyen and colleagues also observed low influenza tetramer⁺ CD4⁺ and CD8⁺ T-cell frequencies of <1% in the peripheral blood, characterized by highly activated HLA-DR, CD38, PD-1, or CD71 phenotypes, in patients hospitalized with influenza disease (11). The authors hypothesized that larger numbers could be migrating to the respiratory tract during acute infection. As such, Paterson and colleagues elegantly provide evidence of enrichment and trafficking of activated influenza-specific tetramer⁺ CD8⁺ T cells to the BAL during acute influenza virus infection.

Although the authors performed tetramer staining in a small number of the available HLA-A1⁺ and HLA-A2⁺ BAL samples, which they diligently acknowledge in the article, the quality of tetramer staining is very convincing. In future studies, it would be enticing to include additional tetramers encompassing a range of influenza class I epitopes spanning several different common HLAs (11) to understand their immunodominance hierarchy and perhaps class II epitopes to examine the role of CD4⁺ T cells at the site of infection. However, the technical requirements for generating peptide-MHC tetramers or the high costs of purchasing these reagents and HLA typing of the participants remain considerable factors here.

The study paved new avenues for exploring antigen-specific T cells in the respiratory specimens. The authors used a bulk RNA sequencing methodology and cell-sorting strategy to analyze bulk CD8⁺ T-cell populations from 12 blood and BAL samples at three different time points. Unexpectedly, innate-like genes normally related to monocytes and DCs represented unique features of the lung-resident CD8⁺ T cells. Taking into account all the caveats of obtaining and assaying the precious BAL samples, it would be a tremendous leap forward if single-cell RNA sequencing could be performed in future studies, as the authors allude to in their discussion, but coupled with antibody-barcoded technology (i.e., barcoded anti-CD69 and anti-CD103 or barcoded streptavidin-labeled tetramers) to further define these CD8⁺ T-cell subsets and surrounding BAL cell populations at a greater resolution. Perhaps then we could carefully delineate the types of T cells comprising the spectrum of innate-like and more adaptive-like CD8⁺ T-cell subsets, which can include innate T lymphocytes such as mucosal-associated invariant T (MAIT) cells and natural killer T cells. The role of innate T lymphocytes during influenza disease was previously suggested, with higher numbers of functional MAIT and T cells in the blood being associated with reduced disease severity in hospitalized patients with influenza (11). Another idealistic approach could be the use of highly sophisticated spatial transcriptomics (12) with endobronchial tissue samples. Given how unique these respiratory samples are and the quality of the experimental design, it is easy to get carried away with an endless wish list of elaborate technologies to use in future experiments.

Overall, Paterson and colleagues not only paved the way for conducting valuable human experimental virus infection studies to closely monitor clinical disease and viral load but also defined the immune mechanisms of protection by directly interrogating immune cells at the site of infection. Until now, we have only been able to scratch the surface and describe influenza-specific CD8⁺ T-cell responses in the blood as correlates of protection, but this article gives absolute insights into the parallel kinetics, phenotype, activation, and trafficking status of influenza-specific CD8⁺ T cells in the blood versus the respiratory tract during influenza virus infection. Without a doubt, this experimental virus infection model will be paramount to assessing future universal influenza vaccine and therapeutic candidates.