We used IL-17A reporter mice to assess IL-17A expression in resting mice and during models of bacterial pneumonia and autoimmune disease. We did not observe constitutive IL-17A expression from cells in naïve mice. However, in appropriate infectious and autoimmune models, we observed IL-17A-expressing CD4+ T cells and greater percentages of IL-17A-expressing γδ T cells, iNKT cells and other CD3ε+ cells. The other CD3ε+ cells were defined by their expression of CD3ε+ and lack of expression of CD4, CD8, γδ TCR and the CD1d-tetramer. It is possible that these cells are type II NKT cells that do not recognize α-galactosylceramide, or T cells that have downregulated the expression of CD4, CD8 or the γδ TCR. In all of these studies, IL-17A production as assessed by the reporter was limited to CD3ε+ T cells and was not expressed by myeloid cell populations. By using novel dual cytokine reporter mice, we demonstrated that IFN-γ-expressing lymphocytes greatly outnumbered IL-17A-expressing lymphocytes in both the acute bacterial and in the more subacute autoimmune challenges.
We were unable to detect expression of the hNGFR reporter in any CD3ε+
cell type in naïve mice, suggesting that IL-17A is not expressed in situ
in resting mice. Segmented filamentous bacteria (SFB), Candidatus arthromitis
, have been shown to induce IL-17A production by lamina propria T cells when introduced into the mouse intestinal flora 
. Cohousing Smart-17A mice with mice obtained from Taconic Farms led to colonization by SFB as determined by PCR, but did not induce hNGFR expression on cells in the lamina propria of the small intestine or colon (data not shown). Our inability to quantitate active IL-17A production from cells in tissues of resting mice contrasts with prior observations made with other strains of IL-17A reporter mice. In the first report, Hirota et al. 
crossed IL-17A-cre mice to Rosa-flox-stop-YFP mice to generate an IL-17A fate-tracking reporter and observed that 11% of lamina propria CD4+
T cells were constitutively YFP-positive in these mice. A study by Esplugues et al. 
using a second IL-17A reporter strain, which utilized an IRES-eGFP knockin strategy more closely akin to the construct used in the Smart-17A mice, reported that 2–4% of lamina propria CD4+
T cells constitutively expressed eGFP. Whether the differences between our results and those using these alternate reporter strains are technical (e.g. permissiveness of the various IRES elements or detection limits of eGFP versus hNGFR) is unclear, and will require direct comparisons of the strains in question.
Here, we crossed our Smart-17A reporter mice to RORγt-GFP reporter mice to ensure our capacity to isolate potential IL-17A-producing cells from the relevant tissues. Although we readily identified populations of RORγt+
cells in multiple organs of naïve mice, these cells did not constitutively express the hNGFR reporter. In line with a previous report 
, we noted substantial populations of RORγt+
γδ T cells in the skin and lungs of resting mice. We also noted smaller percentages of RORγt+
iNKT cells in the lung and other CD3ε+
cells in the lung and skin. Resident dermal RORγt+
γδ T cells have recently been shown to play a role in cutaneous immunosurveillance and to contribute to skin pathology during a mouse model of psoriasis 
. The localization of constitutive RORγt+
innate-like T lymphocytes to the lung is intriguing as this mucosal tissue is relatively sterile and devoid of large numbers of commensal bacteria, a feature that is may contribute to the development of the RORγt+
populations in the intestinal tract 
. Although lung-resident innate-like T cells did not constitutively express IL-17A as assessed by the hNGFR reporter, γδ T cells, iNKT cells and other CD3ε+
cells expressed hNGFR within 24 hr after Klebsiella
infection or 8 hr after administration of IL-1β and/or IL-23. Taken together, these data suggest that RORγt+
innate-like T cells accumulate at multiple epithelial barriers, where they are poised to respond rapidly to compromises in epithelial integrity with IL-17A production and the subsequent recruitment of neutrophils to the injured site.
Earlier reports used ex vivo
restimulation to identify Th17 cells induced during a variety of bacterial and fungal infections, particularly those initiated by mucosal challenges 
. However, as we demonstrate here, ex vivo
restimulation has the potential to reveal IL-17A production by cells that may not be actively producing this cytokine in vivo
(). Although our studies using K. pneumoniae
in Smart-17A mice revealed activation of reporter expression in both innate-like T cells and CD4+
Th17 cells, earlier studies demonstrated increased mortality in mice lacking γδ T cells but no difference in mortality in mice lacking αβT cells 
, suggesting that γδ T cells may be a more important source of IL-17A during this infection. Similar observations were made in experimental Mycobacterium tuberculosis
and in intraperitoneal infection with Escherichia coli
, suggesting a more generalized role for IL-17A production by γδ T cells in immunity to bacterial infections. Our findings corroborate these interpretations.
Somewhat unexpected were the small percentages of IL-17A-expressing CD4+ T cells recovered from lymph nodes and spinal cords and cerebellums during the peak of EAE disease, a canonical experimental model for Th17 induction. Despite the relatively small percentage of IL-17A reporter-positive CD4+ T cells, the large numbers of CD4+ T cells in the inflammatory milieu was such that CD4+ T cells still comprised a sizeable portion of total IL-17A-producing cells, making up 15% of the total in the LN and 42% in the CNS. The vast majority of effector CD4+ T cells, however, were Th1 cells as assessed by expression of the Great IFN-γ reporter allele and, based on our in vitro studies of the decay of the reporter on Th17 cells, are unlikely to have secreted IL-17A over the prior 1–2 days. Our findings seem to contrast with those obtained using an IL-17A-cre/ ROSA-flox-stop-YFP fate-tracking reporter mouse (14). In that study, the authors observed that over half of the CD4+ and γδ T cells were YFP+ in the spinal cords of mice during the peak of EAE. However, cells positive for the IL-17A-cre/Rosa-flox-stop-YFP reporter were those that expressed the Cre recombinase from an activated IL-17A locus at any time during their development, whereas the Smart-17A reporter specifically marks cells that have recently expressed or are currently expressing IL-17A. Thus, differences in construction of these various reporter mice may account for the differences in the experimental results.
Recent reports have raised the possibility that Th17 cells represent an unstable transient phenotype rather than a fully differentiated Th subset akin to that of Th1 and Th2 cells. During the development of EAE, IL-17A+
and IL-17A/IFN-γ double-positive CD4+
T cells have been observed in the lymph nodes and spinal cords of immunized mice 
. Adoptive transfer of encephalitogenic Th17 cells purified from an IL-17F-cre BAC transgenic mouse into RAG-2−/−
or wild-type recipients revealed that a portion of these cells began to secrete IFN-γ during the progression of EAE 
. Similarly, studies using the IL-17A-cre/Rosa-flox-stop-YFP mice revealed that only half of the reporter-positive fate-marked CD4+
T cells were positive for intracellular IL-17A in the spinal cords during the peak of EAE 
, providing further evidence for a switch from an IL-17A-producing to an IFN-γ-producing phenotype. However, all of these studies used ex vivo
restimulation to assess IL-17A and IFN-γ production from isolated spinal cord cells. Our study, which relied solely on direct ex vivo
detection of cytokine production using knockin reporter mice, provides additional evidence for the marked predominance of IFN-γ production by CD4+
T cells during each of the clinical stages of EAE. We saw only low numbers of IL-17A+
cells as compared to IFN-γ+
cells, and approximately half of the Th17 cells concordantly expressed IFN-γ. Taken together, these findings suggest that activation of the IL-17A locus may be important in the early differentiation of pathogenic CD4+
T cells in EAE, but that actual production of IL-17A might not be a major mechanism driving the neurological manifestations of the disease. Indeed, recent reports have demonstrated critical contributions by GM-CSF rather than IL-17 in the pathogenesis of EAE 
A consistent observation was the pronounced segregation of IL-17A-expressing and IFN-γ-expressing cells within the innate-like T cell populations. This segregation occurred in both the infectious and autoimmune models, raising questions about the mechanisms of regulation that account for this exclusionary pattern of cytokine production. IFN-γ and IL-17 have been shown to limit differentiation of Th17 and Th1 cells, respectively, and this mutual inhibition could be operating in these innate populations as well 
. It has also been suggested that antigen-naive γδ T cells predominately produce IL-17A when activated, whereas antigen-experienced γδ T cells produce IFN-γ 
, which could potentially explain the differences in cytokine-secreting populations seen in our models. Further characterization of the cells expressing these cytokines is needed to address this question more fully.
Cytokine reporter mice allow for the functional marking of cells during the course of an inflammatory challenge and have provided essential insights into the coordination of the immune response in vivo. The Smart-17A mice and Great mice used in this study permitted the detection of IL-17A-producing and IFN-γ-producing cells in situ without restimulation. Using these mice, we demonstrated that innate-like T cells, particularly γδ T cells, comprised major cell populations poised for acute IL-17A production. During both K. pneumoniae infection and EAE, models previously demonstrated to induce potent IL-17 expression, we show that the numbers of IL-17A-producing cells were far fewer than the numbers of IFN-γ-producing cells in the same tissues, suggesting different levels of regulation of these two inflammatory cytokines in vivo. Although IL-17A production has been elicited from a number of different cell types using restimulation, our reporter system suggests that cytokine production is limited to T cells in the models we studied. These cytokine reporter mice will be valuable tools for future studies investigating the full contributions of IL-17A-expressing cells to vertebrate immunity.