At the Checkpoint: Lung CD8⁺ T Cells, Respiratory Viruses, and Chronic Obstructive Pulmonary Disease

CD8⁺ T cells possess frightening potential to harm us while protecting against viruses and other pathogens. Given how rapidly activated CD8⁺ T cells divide, if they were not subject to tight physiological control, their metabolic needs alone could sap us as quickly as any malignancy. Fortunately, once viruses are defeated, most effector CD8⁺ T cells are rapidly eliminated by Fas-induced apoptosis. Even earlier, the potentially destructive power of their proinflammatory cytokine production is regulated by complex feedback inhibition. But some viral infections cannot be defeated, and then these same inhibitory pathways become part of elaborate schemes of molecular détente with the pathogen. Because CD8⁺ T cells respond not to intact organisms, but to small peptides presented in the context of class I major histocompatibility molecules, one result can be impaired ability to combat related, and even distinct, respiratory viruses. In extreme cases, notably chronic HIV infection, but also in some persistent nonviral infections and cancers, such counterinflammatory measures culminate in CD8⁺ T-cell exhaustion (1).

These principles are relevant to a tantalizing clinical puzzle: Why are patients with chronic obstructive pulmonary disease (COPD) plagued by enhanced susceptibility to viral lower respiratory tract infections (VLRTI)? Evidence of such susceptibility includes the prominent role for rhinoviruses, a largely innocuous nuisance in healthy subjects, as triggers of acute exacerbations (2), as well as exaggerated lethality after influenza infection in COPD. In both cases, emerging evidence at least partially implicates bacterial superinfection (3, 4). Such opportunists exploit virally damaged epithelia and specific regulatory mechanisms that dampen inflammation after viral eradication. Nonetheless, could acquired (or perhaps minor genetic) defects in mucosal immunity also make patients with COPD simply less able to eliminate incipient VLRTIs? Smoking definitely decimates lung innate immune defenses (5), including antiviral recognition (6). Yet COPD can clearly progress, despite decades of smoking cessation. At the molecular level, how is prolonged susceptibility to VLRTI sustained?

In this issue of the Journal, McKendry and colleagues (pp. 642–651) provide a novel answer, by showing that lung CD8⁺ T cells from patients with COPD respond defectively to H3N2 influenza virus in vitro (7). They identify the culprit to be a specific inhibitory cell surface receptor, programmed cell death protein 1 (PD-1), also known as CD279. PD-1 is a member of the CD28 family, part of the immunoglobulin domain superfamily that figures so prominently in communication between immune cells. Unlike CD28, but similar to CTLA4, PD-1 possesses immunosuppressive signaling motifs. Hence, binding of its ligands PD-L1 and PD-L2, which are expressed by multiple cell types, arrests CD8⁺ T cell activities. If PD-1 sounds familiar, it should, as it is the chief target of checkpoint inhibition, the recent transformative breakthrough in tumor immunotherapy (8). By blocking PD-1 (or related molecules), these agents potently unleash selective tumor cell killing.

This innovative study (7) has multiple strengths, including its application of cutting-edge immunological concepts and techniques. It employed a model devised by the authors (9), in which strips of tissue from human lungs removed for clinical indications are virally infected, permitting elegant in vitro dissection of the pulmonary immune response. Results show that, relative to smokers without COPD, H3N2 infection more greatly up-regulated PD-1 expression on lung CD8⁺ T cells of COPD subjects. Conversely, CD107a expression in response to virus (a measure of degranulation, and thus a surrogate for killing of infected cells) was less robust in COPD. Evidence was also found for antiviral responses by lung CD4⁺ T cells, an unexpected finding predicted from animal models but poorly documented in humans.

Several findings of McKendry and colleagues merit discussion regarding superficially disparate results from other investigative groups also using surgical specimens. The depressed antiviral capacity of lung CD8⁺ T cells the researchers found in COPD might be seen to contradict evidence that this cell type has an exaggerated capacity to produce proinflammatory cytokines on polyclonal or antigenic stimulation (10–12). Importantly, however, McKendry and colleagues also found greater inducible IFN-γ release by lung CD8⁺ T cells in COPD (7). Similarly, by demonstrating retained expression of mRNA transcripts for the T-cell immunoglobulin domain and mucin domain 3 molecule, they verified that lung CD8⁺ T cells in COPD did not meet one classic definition of T-cell exhaustion. That result complements the finding that in some subjects with advanced COPD, lung CD4⁺ T cells are highly resistant to polyclonal activation, although similarly negative for multiple molecules characteristic of anergy or exhaustion (13). These two examples suggest that immunodeficiency localized to the lungs may prove as common in advanced COPD as evidence of autoimmunity.

These three phenomenon (impaired host defense, exaggerated inflammation, and loss of self-tolerance) are quite possibly interrelated. COPD progression appears to be characterized by increasingly dysfunctional responses to multiple classes of microbes that are handled by healthy hosts with little or no inflammation. The quandary is, Why? Certain microbes, including influenza viruses and Pseudomonas aeruginosa, are unquestionably uniquely capable of devastating damage. Thus, one possibility is that COPD progression depends primarily on the acquisition of specific “bad bugs.” Alternatively, lung destruction in COPD might be driven by loss of tolerance to the local microbial community itself, as in inflammatory bowel disease. These two processes, each potential outcomes of impaired immunity, might contribute to different COPD phenotypes or predominate at different times in individual patients with COPD. By exposing normally hidden self-antigens, either could lead to autoimmunity, especially in emphysema (14).

McKendry and colleagues have produced a significant contribution to the understanding of COPD immunopathogenesis, a topic of immense global importance. Checkpoint inhibition is revolutionizing cancer treatment. Could the same principles improve antiviral responses in COPD, ideally without inducing autoimmune mayhem (15)? The significant monetary cost of available checkpoint inhibitors, mostly humanized monoclonal antibodies, currently makes this approach impractical in COPD. However, development of small-molecule inhibitors might change that situation.

Finally, the molecular basis for the exaggerated PD-1 expression in COPD remains to be determined. Some possibilities include epigenetic changes intrinsic to lung CD8⁺ T cells, lack of appropriate help from equally defective lung CD4⁺ T cells, and inappropriate signals from monocytes that matured into dendritic cells in a toxic environment. The field is open for yet more imaginative suggestions.