Herein, we have demonstrated that the increased capacity of local antigen experienced TRM to persist in the epidermis when levels of TGFβ are limited is mediated by increased expression of TGFβRIII. We also show that local antigen experienced TRM have increased proliferative capacity during repeated antigen recalls. In addition, the increased proliferative capacity was directly correlated with the strength of TCR stimulation during TRM development. Finally, we found that local antigen experienced TRM appear more transcriptionally related to fully differentiated TRM. Taken together, these data support a model in which TCR engagement by cognate antigen in the skin is a required final step in TRM differentiation resulting in their increased fitness exemplified by increased proliferative capacity and the ability to persist in the epidermis when active TGFβ is limited.
We propose that the augmented fitness of local antigen experienced TRM represents a mechanism to enrich for high avidity TCR clones in the epidermis. Skin inflammation recruits TEFF into the skin, some of which develop into TRM. In the absence of competition with pre-existing TRM, TRM form comparably in the presence or absence of cognate antigen. Thus, we found equivalent numbers of bystander TRM at the DNFB and local antigen experienced TRM at the VV sites with both TCR transgenic and endogenous T cells. In contrast, when new TEFF are recruited into sites with pre-existing TRM, there is clonal competition for limited amounts of active TGFβ, resulting in enrichment of fitter, local antigen experienced TRM (Hirai et al., 2021). We now find that this enrichment likely results from 2 different competitive advantages. First, antigen encounter in the skin results in increased expression of TGFβRIII, which increases TGFβ avidity for the signaling by the TGFβ receptor. Since TGFβ signaling is required for epidermal persistence, this would provide an advantage for local antigen experienced TRM over bystanders. Second, local antigen experienced TRM have increased proliferation when re-encountering antigen in the epidermis. Following repeated challenges which would be expected outside of SPF conditions, the combination of improved expansion and persistence would work together to enrich for high avidity clones, thereby shaping the epidermal CD8+ T cell memory pool. Recently, it has been observed that TRM can contribute significantly to the pool of circulating memory cells (Steinert et al., 2015; Beura et al., 2018b; Fonseca et al., 2020; Wijeyesinghe et al., 2021; Behr et al., 2020). Thus, mechanisms augmenting epidermal TRM fitness that shape the pool of epidermal TRM may also affect the pool of systemic memory cells and represent an example of extra-thymic clonal section.
When TRM were challenged in a primary antigen recall response, we noted that local antigen experienced TRM expanded to a greater extent than bystanders. This expansion resulted from increased in-situ proliferation with minimal contribution from newly recruited TEFF, consistent with prior reports (Park et al., 2018; Beura et al., 2018a; Çuburu et al., 2012). Interestingly, a similar phenomenon occurred following a second encounter with antigen. This indicates that an encounter with peptide at a late time point after TRM differentiation (>50 days) is insufficient to convert bystander TRM into local antigen experienced TRM. Thus, there appears to be a window during TRM development when TCR engagement can allow for full differentiation. We also observed after a single recall response that TRM contracted to an elevated baseline, suggesting an increase in the epidermal niche. We speculate this may result from a reduced T cell intrinsic requirement for survival and/or homeostatic proliferation factors, such as IL-7 or IL-15 or increased expression of these factors by keratinocytes (Richmond et al., 2018; Adachi et al., 2015). Altered sensitivity or availability of TGFβ is unlikely to explain the increased niche size, as this would be predicted to vary between local antigen experienced and bystander.
Transcriptional analysis of TRM isolated from the small intestine have revealed intra-organ heterogeneity, with unique transcriptional populations arising early during TRM development (Milner et al., 2020; Kurd et al., 2020; Fitz Patrick et al., 2021). This aligns well with our identification of 6 distinct transcriptional clusters of epidermal TRM. Cluster 3 appears to represent fully differentiated TRM based on comparison with other TRM datasets. In addition, cluster 3 cells more highly expressed the activation and proliferation-associated genes Junb, Fos and Dusp1 as well as Nr4a1. Increased basal expression of the AP-1 family members Junb and Fos could contribute to the enhanced proliferation of antigen-experienced epidermal TRM during a recall response. Intriguingly, memory CD8 T cells lacking the transcription factor Zbtb20 manifest elevated expression of AP-1 family members and mount more robust antitumor responses (Hao et al., 2024). The Nr4a1 gene encodes for Nur77, which is induced by TCR signaling and its expression correlates with peptide avidity. Notably, Nur77 is required for TRM formation in the liver (Mackay et al., 2013; Mackay et al., 2015; Aluwihare et al., 2009; Jennings et al., 2020; Boddupalli et al., 2016). Interestingly, cells in cluster 3 only accounted for 27% of TRM that had the opportunity to encounter their cognate antigen in the VV-treated flank. We speculate that not all clones at the VV site fully develop into fitter TRM due to lower TCR avidity or specificity to viral antigens only expressed early during infection, which would be absent once the clones arrived into skin.
In sum, TCR signaling during TRM differentiation represents a previously unappreciated final step in TRM differentiation. This results in fitter TRM with a lower requirement for TGFβ transactivation due to increased expression of TGFβRIII and enhanced proliferation in response to peptide stimulation. Moreover, the differing responses to altered peptide ligands indicate that the degree of fitness depends on TCR signal strength. Thus, polyclonal TRM likely develop into a spectrum of bystander to local antigen experienced cells based on TCR avidity. Though we have focused entirely on epidermal T cells, we suspect that these mechanisms may play a role in other epithelial tissues where residency is also dependent upon TGFβ. Additionally, we have solely investigated memory CD8+ T cells after acute inflammation; the role of ongoing TCR-engagement during chronic antigen encounter remains unexplored.