Ex Vivo Lung Perfusion Provides New Insights into Human Lung-Resident Immune Cell Localization and Functional Interactions

Our immune system functions from the level of the whole organism down to the complex assortment of cells permanently or transiently residing within individual tissues. Peripheral tissues are constantly exposed to pathogens, requiring a specialized set of immune cells to recognize these threats. Tissue-resident memory T (T_RM) cells, a noncirculating population of effector memory T cells, are considered to be a major player for protection from infection and cancer and participate in autoimmunity, allergy, and organ transplant rejection (1, 2). How T_RM cells are recruited, maintained, and activated or reactivated are critical concepts in the development of successful preventative and therapeutic measures. Ongoing generation and maintenance of T_RM cells in the lung is of particular interest, as the harsh environment of the lung may not allow for extended persistence of T-cell populations (3).

Much of what we currently understand about T_RM cells comes from the use of animal models (2). The ability to perform techniques like intravascular staining in mouse models has transformed our understanding of this cell population, particularly in highly vascularized tissues such as the lungs (2, 4). Although the use of animal models is instrumental to our understanding of immune responses, there are nonetheless substantial differences between human and animal immune systems, particularly rodent models (5). Nonhuman primate and porcine animal models, although more representative of humans, are limited by cost and availability of facilities (6, 7). Numerous two- and three-dimensional models such as air–liquid interfaces (8) and organoids (9) have been established with human cells; however, they cannot model complex interactions between cells in the context of the whole organism. In this issue of the Journal, Snyder and colleagues (pp. 1230–1244) use ex vivo lung perfusion (EVLP) of human lungs in a novel way to explore interactions between T_RM cells and lung-resident macrophages (M_LRs) within a whole organ (10).

EVLP has been used in transplantation after a clinical trial reported positive outcomes in 2011 (11). Since this time, EVLP has been used in translational research as it closely models the in vivo environment of the lung. For example, EVLP has been used in the pretransplantation treatment of massive pulmonary embolism and reduction of the antimicrobial burden, and to test chemotherapeutic and other anticancer drugs on lesions (reviewed in Reference 12). Much of the research using EVLP is aimed at whole lung processes, and by and large, has a goal of improving lung function in the broad context of transplantation and disease. There is little evidence for the use of EVLP to study how individual immune cell populations function in relationship to each other. In a recent article, Dumigan and colleagues performed EVLP experiments with porcine lungs, demonstrating that the EVLP model results in tissue pathology that resembles that seen in vivo, and further used EVLP to better understand the role of macrophage polarization in the immune response and pathology associated with Klebsiella pneumoniae infection (13), demonstrating the potential of EVLP for this purpose.

The results presented by Snyder and colleagues (10) support the innovative use of EVLP for the study of complex human immune cell interactions and provide unexpected findings about tissue-resident immune cell populations in human lungs. To demonstrate the utility of EVLP for this purpose, the group first used ex vivo CD45 staining to distinguish immune cells in circulation from those in the parenchyma, which they dubbed “protected” cells. As expected, the phenotype of protected CD4⁺ and CD8⁺ tissue cells include the expression of CD103 and CD69. For myeloid origin cells, protected M_LRs closely resembled alveolar macrophages, expressing CD64 and CD206. These protected T_RM cells and M_LRs were colocalized in the airways.

Interestingly, they found two major distinct protected T_RM cell populations: all T_RM cells were CD45⁺CD14⁻CD206⁻CD4/CD8⁺CD69⁺, and either PD1^hi or PD1^lo, with different expression of the transcriptional factors HOBIT and TBET. Particularly in the CD8⁺ T_RM cell compartment, PD1^hiHOBIT^lo (PD1^hi) and HOBIT/TBET^hiPD1^lo (HOBIT^hi) were functionally distinct subsets, preferentially producing higher effector cytokines or higher baseline granzyme B, respectively. In previous reports, CD103⁺ T_RM cells in human lung did not express TBET and eomesodermin (14, 15), HOBIT and Blimp-1 were not to be expressed by CD103⁺ lung T_RM cells compared with their expression by circulating T_EM cells in mice (16). Moreover, tumor-infiltrating CD8⁺ CD103⁺ T cells have increased expression of TBET (14), suggesting that human lung T_RM cells have a capacity to adapt to the microenvironment. It was also reported CD4⁺ cells and IFN-γ are required for generating CD8⁺ lung T_RM cells in mice (17). Given that both CD4⁺ and CD8⁺ have a unique transcriptional profile in human subjects with EVLP compared with animal models and other sampling methods with human subjects, further experimentation with CD4⁺ and CD8⁺ interactions is required.

Lung-resident myeloid cells are believed to participate in the development and activation of T_RM cells based on several animal models. For example, lung-resident dendritic cells (DCs) that present antigen and express TGF-β are believed to be crucial to the development of lung T_RM cells in mice (18). In the nonhuman primate model, interstitial macrophages differ from alveolar macrophages in that they have high turnover in the lung and increased expression of TNF-α in response to IFN-γ plus LPS treatment (19). Interestingly, Snyder and colleagues (10) found that human lung parenchyma was enriched CD64⁺CD206⁺ (alveolar) macrophages compared with DCs. Furthermore, M_LR could induce cytokine response from CD4⁺ or CD8⁺ PD1^hi cells, suggesting potential differences between the murine and EVLP human model. The minor population of DCs could also have the potential to influence T_RM cell function, however, and further study of the T_RM cell, M_LR, and DC interaction in the development and maintenance of T_RM cells is necessary.

Although these results provide exciting evidence for the use of EVLP of human lungs for immunological studies, there are nonetheless caveats that should be considered. EVLP provides an ability to look at complex cell interactions within the context of the whole lung, however, it is not feasible to evaluate long-terms kinetics in infection or disease models. Although these studies would be limited to those cells already present in the lung, the ability to obtain lungs from subjects with existing disease conditions could allow for short-term studies of cellular interactions and pathogenesis early in infection. There are also potential concerns with EVLP-induced inflammation or altered cellular metabolism that should be considered. Finally, it remains to be seen whether EVLP will be translatable to a larger number of laboratory groups.

Despite these limitations, EVLP should be considered as a promising model by which to study numerous human lung-resident cell populations in situ, as well as to specifically isolate tissue-resident populations of these cells for additional phenotypic and functional analyses.