The cells that sample the luminal contents of the airway remain incompletely defined. In the gut, microfold (M) cells have long been known to specialize in the uptake of luminal contents, and they subsequently transfer their endocytosed cargo to the immune cells of gut-associated lymphoid tissue, classically known as Peyeres patches (1-4). In the upper airway, M cells are located adjacent to nasalassociated lymphoid tissues (5-7), but their presence in the normal lung remained uncertain. Then, in 2019, murine inflammatory models were used to definitively demonstrate that lung M cells could be induced and that they were associated with concomitantly induced bronchus-associated lymphoid tissue (8). Given the very recent discovery of airway epithelial M cells, the key signaling pathways governing their differentiation and developmental origins have not been delineated.

Although murine airway inflammation clearly resulted in the differentiation of abundant airway M cells (8), we reasoned that homeostatic M cells might have escaped attention as a result of their scarcity, as was the case with the elusive pulmonary ionocyte (9). Therefore, we reanalyzed our prior single-cell transcriptomic data containing 69,607 murine tracheal epithelial cells (9) and probed for canonical M cell markers. We uncovered a cluster of epithelial cells that we had previously misannotated (Figure E1A in the data supplement). This cluster is characterized by the expression of classical M cell marker genes, including the Spib and Sox8 transcription factors, the chemokines Ccl9 and Ccl20, and Tnfrsf11a (the receptor activator of NF-κB; RANK), which is known to be involved in gut M cell differentiation (10)(Figures 1A and E1B and Table E1). We also identified the bacterial uptake receptor genes Gp2 (4) and a host of genes associated with gut M cells, including Tnfaip2, Marcksl1, and Anxa5 (Figure 1A)(1). We previously miscalled these M cells as immune cells because they had been clustered together with immune cell populations as a result of their shared chemokine expression. Indeed, we identified only 58 M cells among 69,607 airway epithelial cells, making them the single most scarce population in the airway epithelium (0.08%) even when compared with ionocytes, neuroendocrine cells, and tuft cells (see Figure E1C)(9). In the gut, RANK signaling through the noncanonical NF-κB pathway is sufficient to induce M cell differentiation from Lgr5-positive intestinal stem cells (10, 11). Indeed, airway M cells were first identified following RANKL (RANK ligand) induction (6, 8). In the gut, TNF-α-induced canonical NF-κB pathway signaling cooperates with RANKL-mediated noncanonical NF-κB pathway activation to further increase M cell differentiation (12). Reasoning that these pathways of differentiation would be conserved in the gut and airway, we hypothesized that TNF-α might enhance RANKL-induced airway