IL-17-producing γδ T cells switch migratory patterns between resting and activated states

Interleukin 17-producing γδ T (γδT17) cells have unconventional trafficking characteristics, residing in mucocutaneous tissues but also homing into inflamed tissues via circulation. Despite being fundamental to γδT17-driven early protective immunity and exacerbation of autoimmunity and cancer, migratory cues controlling γδT17 cell positioning in barrier tissues and recruitment to inflammatory sites are still unclear. Here we show that γδT17 cells constitutively express chemokine receptors CCR6 and CCR2. While CCR6 recruits resting γδT17 cells to the dermis, CCR2 drives rapid γδT17 cell recruitment to inflamed tissues during autoimmunity, cancer and infection. Downregulation of CCR6 by IRF4 and BATF upon γδT17 activation is required for optimal recruitment of γδT17 cells to inflamed tissue by preventing their sequestration into uninflamed dermis. These findings establish a lymphocyte trafficking model whereby a hierarchy of homing signals is prioritized by dynamic receptor expression to drive both tissue surveillance and rapid recruitment of γδT17 cells to inflammatory lesions.

It is hard to find the single big new finding in this paper. The experiments described show that CNS infiltrating γδT17 cells during EAE were largely CCR6-CCR2+ cells and that these cells appear to be the central in bacterial clearance and immunity to tumours. However it has previously been reported that CCR6 and NK1.1 distinguish IL-17 and IFN-γ producing γδ T cells (Haas et al, Immunity 2012). Furthermore it has been reported that 'CCR6 is required for epidermal trafficking of γδ T cells in an IL-23-induced model of psoriasiform dermatitis' (Mabuchi et al J. Invest. Dermatol, 2013). The authors of the present study use animal models of autoimmunity, infection and cancer and consequently the paper is very disjointed and difficult to see the logic of one experiment to the next. Furthermore there are a significant number of unanswered questions Specific comments 1. What is the frequency of CCR6+CCR2-γδ T cells in adult thymus, iLN and CNS in mice with EAE at onset and peak of disease? Was CCR6 and CCR2 expression assessed on αβ T cells? 2. It would be good to examine the contribution of CCR6+ or CCR2+ γδ T cells in the amplification of Th17 cells using the adoptive transfer model of EAE. 3. The data demonstrates that TCR stimulation downregulates CCR6 expression on γδ T cells; does stimulation with IL-12/IL-18 upregulate CCR6 or CCR2 expression? 4. Markers of T cell activation, such as CD25, CD69, CD44 or CD38 should be assessed in conjunction with downregulation of CCR6 expression. 5. The data suggests that CCR6 regulates the retention of dermal γδT17 cells. It would be nice to assess if CCR2-/-γδ T cells effective/defective in trafficking in models of psoriasis 6. Do γδ T cell subsets have differential expression of CCR6 and CCR2 during disease? 7. The data in Figure 1e and figure 5f are conflicting. 'CCR6 expression was downregulated in proliferated γδ T cells' 'CCR6 downregulation still occurred upon Mitomycin treatment, suggesting that proliferation and CCR6 downregulation were coincident but independent". The authors should include mitomycin C treatment with BrdU incorporation to demonstrate that downregulation of CCR6 in γδ T cells is independent of proliferation and only through activation. 8. The data in the melanoma and EAE models suggest that CCR6 is not required for cellular trafficking, however CCR2 is essential. In the S. pneumonia infection model, it is unclear if CCR6 expression is required for protective γδ T cell responses as only the effects with CCR2-/-γδ T cell transfer were assessed.
Reviewer #2 (Remarks to the Author):gd T This study by McKenzie et al defines a differential role for CCR2 and CCR6 in gdT17 homing to inflamed tissues and the dermis, respectively. Upon activation, the gdT17 cells downregulate CCR6, releasing them to migrate to inflamed sites via CCR2. I found this study most interesting and it clearly represents a robust body of work. The model proposed by the authors may appeal to a broader audience particularly since this type of "release" mechanism appears quite novel.
I have some minor comments on this study: 1. 3. Whilst the importance of human gdT17 cells is less clear, it is an obvious question that may be worth discussing. Eg do any human gdT cells express these chemokine receptors and are they modulated by stimulation?
Reviewer #3 (Remarks to the Author):gd17, cytokine Authors Duncan R.McKenzie et al. describe "IL-17-produicing γδ T cells switch migaroatory pattern between resting and activated states". Authors found that "naturally occurring" IL-17A-eYFP+γδ T cells coexpress CCR2 and CCR6 in thymus, spleen and LN, irrespective of Vγ4 or Vγ6 expression, and the CCR6-CCR2+ IL-17A-eYFP+γδ T cells were increased in non-lymphoid organs such as , lung , PP and si LPL of naive mice and in the LN and CNS after induction of inflammation such as EAE, Melanoma and S.pneumoniae infection. CCR6 are down regulated in Brdu+ IL-17A-eYFP+γδ T cells during inflammation or after in vitro stimulation, presumably via BATF and IRF4 induction. This is wellperformed paper and their findings are interesting and informative for localization of IL-17-produicing γδ T cells in inflamed tissues. My comments are as follows: General comment. Notable finding in this paper are in vivo significance of down regulation of CCR6 activated γδ T cells and its molecular mechanism. However, authors show only difference in Fig 6, in which CCR6 transduced IL-17A-eYFP+γδ T cells migrated away from tumor. Authors also should do this experiment in EAE and S. penumoniae infection. For molecular mechanism, author show only descriptive data that BATFKO or IRF-4 KO IL-17A-eYFP+γδ T cells fail to downregulate CCR6 only partly after in vitro stimulation with IL-23 and IL1berta.
Specific comments #1. Figure 2: Score of EAE may be ameliorated in CCR2-/-mice. Tumor size may be enlarged in CCR2-/-mice. Authors show the progress in diseases in CCR2-/-mice There are only small difference in migration in inflamed sites between activated CCR2-/-and CCR2+/+ IL-17A-eYFP+γδ T cells. FACS profile in EAE should be shown in transfer experiments.
#2. Figure 5. Author show only descriptive data that BATFKO or IRF-4 KO IL-17A-eYFP+γδ T cells fail to dow-nregulate CCR6 after in vitro stimulation with IL-23 and IL1b. BATF is basic leucine zipper protein that dimeraizes with JUN proteins to the transcriptional activity of functionally important gens in lymphocytes for functional maturation and migration of effector T cells. How BATF or IRF4 downregulate CCR6? Author conclude that down-regulation of CCR6 occurs independent to cell proliferation. Dose Mytomycin C inhibit cell proliferation sufficiently? Do BATFKO or IRF-4KO γδ T cells express IL23R, IL1R, or IL-17A normally? BATF regulates localization of αβ T cells in intestinal niches via directing expression of CD103 and CCR9 (J.E.M. 210:475-489, 2010). BALT KO mice may show an abnormal localization of γδ T cells. CCR6 promoter/enhancer assay may clarify the direct effect of these molecules on cytokines-induced down-regulation. #2 Figure 4: What is the role of CCR6 on IL-17A-eYFP+γδ T cells except migration to dermis? Is there any difference in localization of CCR6-/-γδ T cells in other organs.

REVIEWERS' COMMENTS:
Reviewer #1 (Remarks to the Author): The authors have addressed all my comments either by arguing them or with additional data. While I don't accept all the arguments, the manuscript is improved and represents a nice body of work.
Reviewer #2 (Remarks to the Author): I thank the authors for their rebuttal. I am satisfied with this response and have no further comments.
Reviewer #3 (Remarks to the Author): Authors responded satisfactorily to the reviewer's comments and suggestions.

Reviewer #1 (Remarks to the Author):gd, EAE, infectious and tumor model
It is hard to find the single big new finding in this paper.
The major finding is outlined in the title: γδT17 cells switch their trafficking pattern between resting and activated states. This is entirely new knowledge and establishes a new paradigm in T cell trafficking, whereby γδT17 co-express inflammatory (CCR2) and homeostatic (CCR6) chemokine receptors and respond to inflammatory stimulus by losing the homeostatic migratory receptor enabling them to be rapidly recruited to inflamed sites. In terms of T cell homing, this is a novel mode of control over migration, which we are convinced will be of significant interest to the field. To quote the two other expert reviewers who critiqued this study: "this study is most interesting and it clearly represents a robust body of work. The model proposed by the authors may appeal to a broader audience particularly since this type of "release" mechanism appears quite novel" (reviewer 2); and "This is well-performed paper and their findings are interesting and informative for localization of IL-17-produicing γδ T cells in inflamed tissues" (reviewer 3). To reiterate, this broad finding encompasses the following specific novel findings: • γδT17 cells switch off CCR6 expression during activation (via IRF4/BATF signaling) • CCR2, but not CCR6, is the universal inflammatory homing receptor for γδT17 cells • downregulation of CCR6 enhances γδT17 cell homing to inflamed tissue by disrupting their homeostatic circuit through uninflamed skin The experiments described show that CNS infiltrating γδT17 cells during EAE were largely CCR6-CCR2+ cells and that these cells appear to be the central in bacterial clearance and immunity to tumours. However it has previously been reported that CCR6 and NK1.1 distinguish IL-17 and IFN-γ producing γδ T cells (Haas et al, Immunity 2012). Furthermore it has been reported that 'CCR6 is required for epidermal trafficking of γδ T cells in an IL-23-induced model of psoriasiform dermatitis' (Mabuchi et al J. Invest. Dermatol, 2013).
The reviewer seems to be suggesting a lack of novelty in our findings with these remarks, but may be largely missing the major point of our study. We make no claim to be the first to describe CCR6 as a marker of γδT17 cells or to be the first to show that CCR6 regulates γδT17 cell trafficking in some way. Our data substantially build upon the report by Haas et al. to provide entirely novel information that changes the concept of how γδT17 cells regulate their migration. We show that γδT17 cells express CCR2 along with CCR6 at rest, but during activation (either in vitro or in vivo) we report that CCR6 expression is lost (Fig 1d- , which may further explain these differences due to possible tissue-and model-specific factors. The submitted manuscript highlighted the discrepancies between the Mabuchi report and our data, but we have now incorporated more discussion of this point (lines 341-345).
The authors of the present study use animal models of autoimmunity, infection and cancer and consequently the paper is very disjointed and difficult to see the logic of one experiment to the next.
We believe that the use of animal models of autoimmunity, cancer and infection to investigate the broad concepts of γδT17 cell trafficking is a strength of the paper, as it allows us to extending our findings from model-specific phenomena to broad underlying principles of γδT17 cell biology. We consider that the selection of these models is logical: • EAE is initially used to examine activation of γδT17 cells, as it is a wellcharacterised model in this regard, and extends our recently published findings with Th17 cells in this model (Kara et al. 2015) • we then use EAE and cancer to study γδT17 cell trafficking in scenarios where local tissue-resident γδT17 cells are absent, so that the analysis focuses on migrated cells • subsequently, we extend the finding that CCR2 recruits γδT17 cells to these scenarios to an infectious setting that involves a local resident population, examining γδT17 trafficking-dependent protection in S. pneumoniae infection • examining γδT17 cell trafficking into uninflamed skin enables us to study their homeostatic trafficking patterns as the skin is a major homeostatic reservoir of these cells • finally, we return to inflammation to study why CCR6 expression is lost upon γδT17 cell activation -we examined trafficking of Ccr6 tg γδT17 cells during cancer in the submitted manuscript, but as requested by reviewer 3, have extended these experiments to include EAE and S. pneumoniae upon revision to more broadly apply our findings (Fig 6e-f) We have modified the results section to more clearly link one experiment to the next (lines 162-163, 220-222).
Furthermore there are a significant number of unanswered questions Specific comments 1. What is the frequency of CCR6+CCR2-γδ T cells in adult thymus, iLN and CNS in mice with EAE at onset and peak of disease? Was CCR6 and CCR2 expression assessed on αβ T cells?
As is evident from Fig 1a, c-d, f in the submitted manuscript, CCR2 -γδT17 cells are negligible in thymus, iLN, CNS and all tissues, and this does not change during EAE.
We have modified the text to emphasize this observation more strongly (lines 119-120). CCR6 and CCR2 in αβ T cells was entire the focus of our recent paper (Kara et al. 2015, Nat Commun, ref #18), which provided the basis for the present study as was clearly written in the manuscript (lines 65-69, 81-85).
2. It would be good to examine the contribution of CCR6+ or CCR2+ γδ T cells in the amplification of Th17 cells using the adoptive transfer model of EAE.
Please note that we are not studying the migration of γδ T cells generally. Rather, we are studying the IL-17-producing subset (γδT17 cells). Only this subset expresses CCR6 and/or CCR2 (Fig 1b, Fig S2a), therefore γδT17 cells would be the only relevant cells for such an experiment. However, as is a key message of the paper, γδT17 cells are either CCR6 + CCR2 + (resting) or CCR6 -CCR2 + (activated). Therefore, it is not possible to separate CCR6 + from CCR2 + γδT17 cells as proposed by the reviewer, so the experiment is not possible.
The data demonstrates that TCR stimulation downregulates CCR6 expression on γδ T cells; does stimulation with IL-12/IL-18 upregulate CCR6 or CCR2 expression?
We have conducted this experiment and thank the reviewer for the suggestion. We have included the data in the revised manuscript ( Fig S2b -  3. Markers of T cell activation, such as CD25, CD69, CD44 or CD38 should be assessed in conjunction with downregulation of CCR6 expression. We have conducted this experiment. As is shown below and now included in the manuscript (Fig S2c), IL-23 and IL-1β-stimulation drove upregulation of activation markers CD25, CD44 and CD69 in γδT17 cells, concomitantly with CCR6 downregulation. Marker CD38 did not change substantially upon stimulation (although its relevance to γδT17 cell biology is unknown). This result further supports our conclusion that activation drives repression of CCR6 expression in γδT17 cells.  The only γδ T cell subset to express CCR6 and/or CCR2 is the γδT17 subset. Figure  1b in the submitted manuscript demonstrated that non-IL-17-producing γδ T cells do not migrate to either CCL20 or CCL2. We now include data below and in the manuscript (Fig S2a, lines 91-92, 108-110) demonstrating that while CD44 hi CD27γδ T cells (i.e. γδT17 cells) are CCR6 + CCR2 + at rest and become CCR6 -CCR2 + during EAE, other γδ T cell subsets (CD44 hi CD27 + and CD44 lo CD27 + ) are CCR6 -CCR2at rest and remain so during EAE. Thus, CCR6/CCR2 expression is restricted to IL-17producing γδ T cells at homeostasis and during inflammation.

Figure S2 a) Representative flow cytometry of CCR6/CCR2 expression by subsets of γδ T cells distinguished by CD44 and CD27 expression in sLNs of WT mice either naïve or at EAE onset (n=3/group).
6. The data in Figure 1e and figure 5f are conflicting. 'CCR6 expression was downregulated in proliferated γδ T cells' 'CCR6 downregulation still occurred upon Mitomycin treatment, suggesting that proliferation and CCR6 downregulation were coincident but independent". These data are not at all conflicting; rather, they are complementary. Fig 1e  demonstrates that BrdU + γδT17 cells during EAE are largely CCR6 -, suggesting that proliferated (hence activated) γδT17 cells lose CCR6 expression. To then separate activation and proliferation (a product of activation), Figure 5f used Mitomycin C treatment to block proliferation in stimulated γδT17 cells. As CCR6 was downregulated upon stimulation even when proliferation was blocked, we conclude that activation induces both proliferation and loss of CCR6 in γδT17 cells (entirely consistent with Figure 1e), but that proliferation itself is not the driver of CCR6 downregulation, rather it is coincident with the phenomenon.
The authors should include mitomycin C treatment with BrdU incorporation to demonstrate that downregulation of CCR6 in γδ T cells is independent of proliferation and only through activation.
The submitted manuscript used a cell proliferation dye 'eFluor 670' for this purpose in Figure 5f. As is evident from the flow cytometry plot, unstimulated γδT17 cells were CCR6 + and did not divide. Stimulated γδT17 cells proliferated (i.e. diluted the dye) and lost CCR6 expression. However, stimulated cells treated with mitomycin C lost CCR6 but did not divide. Therefore we had already fulfilled this question, but with a more sensitive measure of proliferation than BrdU incorporation. We have amended the text to more clearly highlight these data (lines 257-262).
8. The data in the melanoma and EAE models suggest that CCR6 is not required for cellular trafficking, however CCR2 is essential. In the S. pneumonia infection model, it is unclear if CCR6 expression is required for protective γδ T cell responses as only the effects with CCR2-/-γδ T cell transfer were assessed.
Experiments with S. pneumoniae utilised transfers of in vitro activated and expanded γδT17 cells. As we clearly demonstrate, activated γδT17 cells lack CCR6 expression (Fig 1). Accordingly, γδT17 cells expanded in vitro, and used in these experiments, lack functional CCR6 expression as expected (Fig 6b/c -empty virus-transduced cells). Thus, as the cells do not express CCR6, it is evident that in these experiments, CCR6 is not required for γδT17 cell trafficking and/or protection.

Reviewer #2 (Remarks to the Author):gd T
This study by McKenzie et al defines a differential role for CCR2 and CCR6 in gdT17 homing to inflamed tissues and the dermis, respectively. Upon activation, the gdT17 cells downregulate CCR6, releasing them to migrate to inflamed sites via CCR2. I found this study most interesting and it clearly represents a robust body of work. The model proposed by the authors may appeal to a broader audience particularly since this type of "release" mechanism appears quite novel.
I have some minor comments on this study: 1. Fig 2 -Is there an effect of CCR6, CCR2 and CCR2.CCR6 KO on tumor growth/EAE scores?
We have included data demonstrating the effect of Ccr6, Ccr2 or Ccr2 Ccr6 deficiency upon EAE scores and tumor weights at d7 post-challenge ( Fig S3). As there are clear differences, we highlight the necessity of our co-transfer experiments to assess cell-intrinsic recruitment of γδT17 cells to these inflammatory sites (lines 146-148).
The use of these gating strategies has been previously published and we have clarified these approaches in the results section (lines [93][94][95][194][195]. In short: • CD3-bright is a phenotype used to distinguish Vγ6 + (CD3-bright) γδT17 cells from Vγ4 + ('conventional' CD3 staining) γδT17 cells as is published (Paget et al. 2015 Immunol Cell Biol, ref #25). We use this as a complementary strategy to Vγ4 staining to show that Vγ4 + and Vγ6 + γδT17 cells share trafficking characteristics. • CD3 lo is a phenotype used to distinguish dermal γδ T cells (CD3 lo TCR-γδ lo i.e. γδT lo ) from contaminating dendritic epidermal T cells (DETCs, CD3 hi TCRγδ hi ) in dermal cell preparations (Gray et al. 2011 J. Immunol, ref #22). We only refer to γδT lo cells when describing dermal γδ T cells, which are essentially all γδT17 cells (Fig S5a) 3. Whilst the importance of human gdT17 cells is less clear, it is an obvious question that may be worth discussing. Eg do any human gdT cells express these chemokine receptors and are they modulated by stimulation?
We agree and have now included a discussion of this as yet unexplored area in the revised manuscript (lines 373-384).

Reviewer #3 (Remarks to the Author):gd17, cytokine
Authors Duncan R.McKenzie et al. describe "IL-17-produicing γδ T cells switch migaroatory pattern between resting and activated states". Authors found that "naturally occurring" IL-17A-eYFP+γδ T cells coexpress CCR2 and CCR6 in thymus, spleen and LN, irrespective of Vγ4 or Vγ6 expression, and the CCR6-CCR2+ IL-17A-eYFP+γδ T cells were increased in non-lymphoid organs such as , lung , PP and si LPL of naive mice and in the LN and CNS after induction of inflammation such as EAE, Melanoma and S.pneumoniae infection. CCR6 are down regulated in Brdu+ IL-17A-eYFP+γδ T cells during inflammation or after in vitro stimulation, presumably via BATF and IRF4 induction. This is well-performed paper and their findings are interesting and informative for localization of IL-17-produicing γδ T cells in inflamed tissues. My comments are as follows: General comment. Notable finding in this paper are in vivo significance of down regulation of CCR6 activated γδ T cells and its molecular mechanism. However, authors show only difference in Fig 6,  Replies to these general comments are included below.
Specific comments #1. Figure 2: Score of EAE may be ameliorated in CCR2-/-mice. Tumor size may be enlarged in CCR2-/-mice. Authors show the progress in diseases in CCR2-/-mice There are only small difference in migration in inflamed sites between activated CCR2-/-and CCR2+/+ IL-17A-eYFP+γδ T cells. FACS profile in EAE should be shown in transfer experiments.
We have included EAE scores and tumor sizes ( Fig S3a) in Ccr6, Ccr2 and Ccr2 Ccr6 deficient mice (also see point 1, reviewer 2). As CNS inflammation is reduced, while tumor size increased, in Ccr2 deficient mice, we highlight the importance and novelty of our co-transfer experiments in Fig 2i- j (lines 146-148). We now show the flow cytometry for EAE transfer experiments (Fig 2j).
#2. Figure 5. Author show only descriptive data that BATFKO or IRF-4 KO IL-17A-eYFP+γδ T cells fail to dow-nregulate CCR6 after in vitro stimulation with IL-23 and IL1b. BATF is basic leucine zipper protein that dimeraizes with JUN proteins to the transcriptional activity of functionally important gens in lymphocytes for functional maturation and migration of effector T cells. How BATF or IRF4 down-regulate CCR6?
We now include data from our own and others' ChIP-seq datasets identifying binding of both IRF4 and BATF to a conserved site in the Ccr6 promoter in two murine T cell subsets (Fig S7b, lines 254-257). Therefore it is likely that IRF4 and BATF cooperate (as is reported, Ciofani et al. 2012 Cell) to repress Ccr6 expression in γδT17 cells. We feel that further investigation of this point surpasses the scope of the manuscript, which concerns migration rather than molecular interactions. We believe that the 'descriptive' data included in the original submission, i.e. that IRF4 and BATF are required for CCR6 downregulation, are the most relevant for the biological function of γδT17 cells and thus the message of the paper. Author conclude that down-regulation of CCR6 occurs independent to cell proliferation. Dose Mytomycin C inhibit cell proliferation sufficiently?
The submitted manuscript includes the use of cell proliferation eFluor670 dye to demonstrate that mitomycin C treatment completely blocked γδT17 cell proliferation upon stimulation (Fig 5f). We reiterate this point in the text (line 259-261).
We have analysed resting γδT17 cells from WT, Irf4 -/and Batf -/mice. We thank the reviewer for the important suggestion to determine whether γδT17 cells in these mice express normal amounts of IL-23R and IL-1R1 as this could impact upon their activation. As is shown below and now included in the manuscript (Fig S7a, lines 251-253), Irf4 -/and Batf -/-γδT17 cells express similar levels of both receptors to WT cells. Therefore we maintain our conclusion that reduced CCR6 repression and proliferation of these cells in response to IL-23 and IL-1β stimulation is due to acute IRF4 and BATF-mediated signaling, and not due to IRF4 and BATF-driven surface expression of the appropriate stimulatory receptors. The submitted manuscript demonstrated that γδT17 cells are less frequent in Irf4 -/-mice ( Fig S6c, formerly Fig S4c). We show below that γδT17 cells in all strains express similar amounts of IL-17A on a per cell basis (based upon gMFI). However, this does not affect our analysis of CCR6 downregulation in vitro, as all such analyses (Fig 5e) were pre-gated on IL-17 + γδ T cells. Therefore differential frequency of IL-17 + cells does not change our conclusion that IRF4 and BATF drive CCR6 downregulation upon activation. While this may be the case, we are investigating with the role of CCR6 and CCR2 in γδT17 cell biology. We consider that pursuing the role of BATF in γδ T cell localization regarding CCR9 is beyond the scope of this manuscript.

IL
CCR6 promoter/enhancer assay may clarify the direct effect of these molecules on cytokines-induced down-regulation. This is addressed above.
#3 Figure 6: This is one of the most important figures in this paper, showing in vivo a significance of down regulation of CCR6 on γδ T cells for preventing sequestration into uninflamed dermis from inflamed tissues. However, authors show only minor difference in Fig 6, in which CCR6-transduced IL-17A-eYFP+γδ T cells migrated away from tumor. Author should show that the CCR6-transduced IL-17A-eYFP+γδ T cells γδ T cell in the dermis Authors also should do this experiment in EAE and S. penumoniae infection and the γδ T cell in the dermis The experiment in the submitted manuscript demonstrates that while Ccr6-tg γδT17 cells were deficient at homing to B16 melanomas, they instead homed to uninflamed dermis (Fig 6d). As suggested, we have now performed this experiment in EAE and S. pneumoniae as requested, and obtained complementary results: Ccr6-tg γδT17 cells were less able to home to the inflamed CNS (EAE) and nasal passage (S. pneumoniae infection) than control-transduced γδT17 cells. Moreover, Ccr6-tg γδT17 cells preferentially homed instead to uninflamed dermis during S. pneumoniae infection. Note that it was not possible to obtain sufficient uninflamed skin during EAE to assess dermal homing in this model (due to subcutaneous CFA injection in both hind flanks precluding the harvest of back skin). These experiments are included in Fig 6e-f (lines 291-296). These new data support our previous conclusion: maintained CCR6 expression diverts activated γδT17 cells into uninflamed skin and thus away from inflamed tissue. We believe that these data significantly strengthen the manuscript and we thank the reviewer for this suggestion. e-f) In vitro-expanded γδT17 cells from WT or F 1 mice were transduced with empty pMIG or pMIG-Ccr6, respectively. Equal numbers of mixed GFP + cells were transferred i.v. into Ly5.1 mice either e) 24 hr post-infection with S. pneumoniae (n=4) or f) at EAE onset (n=3) and organs were analysed 48 hr later. Ratio of recovered WT to F 1 γδT17 cells within transduced (GFP + ) and untransduced (GFP -) populations, normalized to input values. e-f) Paired two-tailed Student's ttest.
Minor comments #1 Introduction: many references are misscited. Authors should cite more original papers.
We have thoroughly checked all references and believe that all are correctly cited. The sole review cited in the original manuscript has been supplemented with a number of original primary papers in the revised manuscript (line 192).
#2 Figure 4: What is the role of CCR6 on IL-17A-eYFP+γδ T cells except migration to dermis? Is there any difference in localization of CCR6-/-γδ T cells in other organs.
Of all tissues examined, the only to demonstrate a reduction in γδT17 cells in Ccr6 -/mice was the dermis (Fig 4a and Fig S5b, formerly Fig S3b). We found a greater number of γδT17 cells in the peritoneal cavity of Ccr6 -/mice therefore it is likely that γδT17 cells unable to enter the dermis pool in this location instead.