Editorial: Pulmonary resident memory CD8 T cells: here today, gone tomorrow.

Infection with influenza A virus can result in the establishment of crossreactive protective immunity against subsequent exposure to influenza A viruses of a distinct serotype. This phenomenon, termed heterosubtypic immunity, is conferred by the adaptive arm of the immune system and is predominantly attributed to T cell detection of shared viral protein epitopes. Paradoxically, heterosubtypic immunity to influenza virus wanes over time [1], despite the fact that long-lived, influenza-specific memory T cells persist within secondary lymphoid organs (Fig. 1A). For some time, memory T cells have been characterized as TCM, which recirculate among blood, lymph, and secondary lymphoid organs, or TEM, which were all thought to recirculate among blood, lymph, and nonlymphoid tissues. Recently, a nonrecirculating subset of memory T cells that resides permanently in nonlymphoid tissues has been described in several organs, including the skin, intestine, kidney, brain, lung, and female reproductive tract [2]. This subset is most often referred to as TRM cells, which have been shown to protect against viral infections in the skin [3, 4] and lung [5] and precipitate local inflammatory cascades that recruit circulating T cells to sites of infection [6]. It has been noted that TRM cells often express the C-type lectin CD69, as well as the E 7 integrin heterodimer (which is most often identified by staining cells with antibodies specific for E, otherwise known as CD103) [2]. CD69 may be associated with retention of TRM, as expression antagonizes S1P responsiveness, and S1P promotes egress of lymphocytes into the circulation. E 7 may also contribute directly to the local maintenance of TRM by anchoring T lymphocytes to epithelial cells through interactions with E-cadherin. However, many critical questions remain: what induces CD69 and CD103 expression among T cells in different tissue compartments? What is the longevity of TRM within different locations? How are TRM established within a particular location? Under what conditions do TRM contribute to protection? In this issue of the Journal of Leukocyte Biology, Wu and colleagues [7] help address these questions by revisiting an old mystery: the enigma that heterosubtypic immunity to respiratory influenza challenge is not long-lasting. Heterosubtypic immunity was examined using two serologically distinct recombinant strains of influenza virus that express chicken ovalbumin (WSN-OVAI and X31-OVA), as well as transgenic CD8 T cells (OT-I) expressing a TCR that recognizes the OVA-derived SIINFEKL peptide when it is presented by H-2K-bearing cells. One month after challenging mice with WSN-OVAI via the respiratory route, but not the i.p. route, OT-I cells were found in the epithelial layer lining the large lung airways. This demonstrated that the site of primary challenge impacted the establishment of T cell memory in the lung. Additionally, many memory T cells in the airway epithelium expressed CD103, consistent with the phenotype of TRM. Further analysis revealed that mice challenged only a month earlier with WSN-OVAI exhibited much more rapid control upon a subsequent heterosubtypic challenge with X31-OVA. In fact, a significant reduction in X31-OVA titers preceded the recruitment of CD8 T cells from outside of the lung. This implies that TRM within the lung, rather than recirculating memory T cells, were most responsible for viral control. This interpretation was supported by the observation that CD103 CD8 T cells in the lung epithelium dissipated within 7 months, and it was their loss that correlated with the waning of heterosubtypic immunity. However, many questions remain. Why is local infection important for the establishment of TRM within the lung epithelium? Is this a result of programming of a homing phenotype during priming, or is the infectious milieu of an infected lung driving recruitment? More importantly, what is regulating the short-term maintenance of local TRM and the subsequent attrition? Interestingly, TRM do not appear to wane significantly in the skin and intestinal mucosa [3, 4, 8], suggesting that the lung is unique in this regard. Why is the lung different, teleologically and mechanistically? One possible difference between tissues is in the regulation of CD103 itself. Many studies implicate a role for TGFin driving CD103 expression. TGFis constitutively expressed in the small intestinal mucosa, where CD103 is maintained on virtually all memory CD8 T cells within the epithelium, and interfering with TGFsignaling or CD103 expression results in a gradual loss of intestinal intraepithelial memory CD8 T cells [9–11]. In contrast, CD103 is expressed by only a minority of CD8 T cells in the lung, and many of these cells are lost over time. Perhaps TGFis also the driver of CD103 expression in the respiratory tract, but available