This is the peer reviewed version of the following article:    Sanchez‐Diaz, R., Blanco‐Dominguez, R., Lasarte, S., Tsilingiri, K., Martin‐Gayo,  E., Linillos‐Pradillo, B., . . . Martin, P. (2017). Thymus‐Derived Regulatory T Cell  Development Is Regulated by C‐Type Lectin‐Mediated BIC/MicroRNA 155  Expression. Molecular and Cellular Biology, 37(9), e00341‐00316.  doi:10.1128/MCB.00341‐16    which has been published in final form at: https://doi.org/10.1128/MCB.00341‐16            1 Thymus-derived Treg cell development is regulated by C-type-lectin–1 mediated BIC/miRNA155 expression 2 3 Running Title: C-type lectin regulates tTreg development 4 5 Authors: 6 Raquel Sánchez-Díaza,e, Rafael Blanco-Domingueza, Sandra Lasartea,e, Katerina Tsilingiria, 7 Enrique Martín-Gayob, Beatriz Linillos-Pradilloa Hortensia de la Fuentec,e, Francisco 8 Sánchez-Madridc,e, Rinako Nakagawad, María L. Toribiob, & Pilar Martína,e. 9 10 aSignaling & Inflammation Program, Centro Nacional de Investigaciones Cardiovasculares 11 (CNIC), Madrid, Spain, 28029. 12 bCentro de Biología Molecular Severo Ochoa, Consejo Superior de Investigaciones 13 Científicas, Universidad Autónoma de Madrid, Spain., 28049. 14 cDepartment of Immunology, Hospital de La Princesa, Madrid, Spain, 28006. 15 dImmunity and Cancer Laboratory, The Francis Crick Institute, London, WC2A 3LY, UK 16 eCIBER CV, Instituto de Salud Carlos III, Spain, 28029. 17 Corresponding author. Tel: +34 914531200; Email: pmartinf@cnic.es 18 19 Keywords: Regulatory T cells development, C-type lectin, miRNA-155, Autoimmunity 20 21 22 2 Abstract 23 24 Thymus-derived regulatory T cells (tTregs) are key to prevent autoimmune diseases but the 25 mechanisms involved in their development remain unsolved. Here, we show that the C-type 26 lectin receptor CD69 controls tTreg cell development and peripheral Treg homeostasis 27 through the regulation of BIC/miR-155 and its target, the suppressor of cytokine signaling 28 1 (SOCS-1). Using Foxp3-mRFP/cd69+/- or /cd69-/- reporter mice, shRNA-mediated 29 silencing and miR-155 transfection approaches, we found that CD69 deficiency impaired 30 the signal transducer and activator of transcription 5 (STAT5) pathway in Foxp3+ cells. 31 This results in BIC/miR-155 inhibition, increased SOCS-1 expression and severely 32 impaired tTregs development in embryos, adults and Rag2-/-γc-/- hematopoietic chimeras 33 reconstituted with cd69-/- stem cells. Accordingly, mirn155-/- mice have impaired 34 development of CD69+ tTreg cells and overexpression of miR-155 induced CD69 pathway, 35 suggesting that both molecules might be concomitantly activated, in a positive feedback 36 loop. Moreover, in vitro-inducible CD25+ Tregs (iTregs) development is inhibited in Il2rγ-/-37 /cd69-/- mice. Our data highlight the contribution of CD69 as a non-redundant key regulator 38 of BIC/miR-155-dependent Treg development and homeostasis. 39 40 41 3 Introduction 42 Regulatory T (Treg) cells are a specialized subset of lymphocytes with a dominant role in 43 the prevention of autoimmune diseases (1). Treg subtypes have been classified according to 44 their origin in the thymus, peripheral lymphoid organs or in vitro, and have been 45 extensively characterized; however, the mechanisms that regulate their generation in the 46 thymus remain poorly understood. Understanding how thymus-derived Treg cells (tTregs) 47 (2) become a distinct lineage is crucial for the development of strategies to control immune 48 responses by targeting these cells (3). A central event in tTreg differentiation is the 49 induction of the transcription factor Foxp3 by early signals delivered from the TCR, which 50 results in transcriptional activation and enhanced function of the IL-2 signaling pathway(4). 51 Among other mechanisms, Foxp3 expression is promoted by miR-155 through the 52 inhibition of SOCS1 (suppressor of cytokine signaling 1), enhancing activation and binding 53 of STAT5 (signal transducer and activator of transcription 5) to the Foxp3 promoter and the 54 Foxp3-CNS (conserved non-coding sequence) (5, 6). In a positive feedback loop, Foxp3 55 increases expression of miR-155 by binding to an intronic element of BIC, the gene 56 encoding the miR-155 precursor transcript. Nevertheless, the mechanisms by which 57 miRNAs impact tTreg differentiation and function are not fully elucidated and the data are 58 somewhat contradictory. For example, Dicer, a member of the RNAseIII complex that 59 processes pre-miRNAs into mature miRNAs, plays a key role in tTreg differentiation (7) 60 and function (8); however, lack of Dicer is linked to enhanced miR-155 expression in 61 MRL/lpr mice (9), suggesting that there are Dicer-independent mechanisms for miRNA 62 regulation in Tregs. The Tregs of Lupus-prone mice have an altered phenotype, low levels 63 of Dicer, and a weak suppressive capacity linked to the expression of the C-type lectin 64 4 receptor CD69 (9). Moreover, increased CD69 expression has been detected in activated 65 Dicer-/- TCs, which show defective egress from lymphoid organs (10). In addition, 66 CD4+CD8+ thymocytes include a CD69highTCRhigh Treg cell progenitor subpopulation, 67 indicating that CD69 expression is relevant to tTreg differentiation (11). We hypothesized 68 that CD69, which contributes to the maintenance of immunological tolerance through the 69 regulation of Treg function, makes a substantial contribution to Treg development in the 70 thymus. The C-type lectin CD69 is expressed constitutively by a subpopulation of 71 peripheral Tregs (pTregs) and tTregs (12). Here, we report that CD69 is required for the 72 development of Tregs in the thymus through the promotion of STAT5 phosphorylation and 73 the transcription of BIC/miR-155. FoxP3-mRFP/cd69-/- reporter mice have a significantly 74 below-normal number of tTregs, and Treg differentiation was also impaired in FTOC 75 cultures of cd69-/- embryonic thymuses or wild-type embryonic thymuses treated with anti-76 CD69. Consistently, FoxP3+ tTregs are poorly generated from cd69-/- precursors in mixed 77 bone marrow chimeras. Impairment of STAT5 phosphorylation in FoxP3-mRFP/cd69-/- 78 tTregs leads to enhanced transcription of SOCS-1 and inhibition of miR-155-dependent 79 tTreg development. CD69 thus maintains miR-155-dependent tTreg development through a 80 positive feedback regulatory mechanism, giving rise to a functional pTreg cell subset. Our 81 results strongly support a role for CD69 as a critical receptor in the control of Treg 82 development and homeostasis. 83 84 5 Results 85 CD69 expression is required for development of the tTreg subset 86 To determine whether CD69 is necessary for tTreg development in the thymus, we 87 analyzed CD69 membrane expression in tTregs from cd69+/+, cd69+/-, and cd69-/- 88 littermates bearing a Foxp3-mRFP reporter gene (monomeric red fluorescent protein 89 inserted in the foxp3 locus). In agreement with previous data in non-reporter mice (12), 90 about 30% of tTregs expressing Foxp3-mRFP in wild-type thymus also express CD69 (Fig. 91 1A and B). This percentage is lower in Foxp3-mRFP/cd69+/- heterozygous mice and this 92 subset is absent in Foxp3-mRFP/cd69-/- deficient mice (Fig. 1A and B). The proportions of 93 CD4+ single-positive (CD4SP) thymocytes and the other thymocyte subsets are unaffected 94 in cd69-heterozygous and -deficient reporter littermates (Fig. 1C); but, compared with 95 Foxp3-mRFP/cd69+/+ mice, both genotypes showed a 30% lower cellularity of total and 96 CD4SP thymocytes (Fig. 1D). These results are consistent with previous data showing that 97 the overexpression of CD69 in the thymus increases the levels of SP thymocytes 98 controlling egress to the periphery (13, 14). However, Foxp3-mRFP/cd69-/- and cd69+/- 99 mice showed a marked reduction in the proportion of tTregs compared with cd69+/+ adult 100 reporter mice (Fig. 1E and F), while total tTreg numbers were not altered in the cd69+/- and 101 cd69-/- deficient groups (Fig. 1F), indicating that CD69 could be playing an important role 102 in the regulation of tTreg development masked by thymocyte egress defects in Foxp3-103 mRFP/cd69-/- mice. In addition, we found that cd69-/- adult reporter mice showed also a 104 reduction in the proportion of peripheral Tregs (pTregs) compared with cd69+/+ littermates 105 (Fig. S1A and B). This data is not consistent with the previously reported in non-reporter 106 mice (12). To clarify the differences observed with Foxp3 reporter mice, we performed 107 6 Foxp3 staining in thymus and spleens from Foxp3-mRFP mice. The data indicate that 108 exogenous staining with anti-Foxp3 antibodies differs from the endogenous print of Foxp3-109 mRFP depending on the tissue (Fig. S2), suggesting that the use of anti-Foxp3 antibodies is 110 not always as accurate as the use of reporter genes. In summary, CD69 could be playing a 111 role in both, tTreg development and pTreg homeostasis. 112 113 Deletion of CD69 inhibits tTreg differentiation in fetal thymus organ cultures 114 To determine if cd69-deficiency leads to a decreased tTreg development, independently of 115 the thymic maturation state or sphingosine 1-phosphate receptor-1 (S1P1)-induced 116 thymocyte egress capacity, we performed a fetal thymus organ culture (FTOC) assay on 117 thymuses from 15-17-day-old mouse embryos (E15-E17), and analyzed total CD4SP 118 thymocytes and tTreg differentiation over 5 days of culture. Compared with cd69+/+ 119 FTOCs, cd69-/- E15-17 FTOCs displayed a marked reduction in the proportion and absolute 120 cell numbers of Foxp3+ tTregs, with insignificant changes in total cell numbers (Fig. 2A 121 and B), indicating that CD69 is required during tTreg differentiation at early stages of 122 development. To confirm these results, we treated E15 FTOCs with an anti-CD69 123 monoclonal antibody (2.2), which downregulates CD69 expression and hence blocks 124 downstream signaling (15) and monitored Treg development over 14 days of culture. 125 Consistent with the cd69-/- FTOC data, throughout the culture period anti-CD69-treated 126 FTOCs showed notably lower proportions and cell numbers of Foxp3+ tTregs than FTOCs 127 treated with isotype control antibody (2.8) (Fig. 2C and D), whereas total FTOC cell 128 numbers were unaltered by either treatment (Fig. 2D). These findings are consistent with 129 previous evidence indicating that immature activated CD69+ thymocytes are the precursors 130 of intrathymic Tregs in humans and mice (11, 16). 131 7 132 Defective tTreg and pTreg generation from cd69-/- progenitors is a cell-autonomous 133 defect 134 To further explore the role of CD69 in tTreg differentiation, we transferred bone marrow 135 (BM) hematopoietic stem cells from Foxp3-mRFP/cd69+/+ or Foxp3-mRFP/cd69-/- 136 littermates into lethally γ-irradiated C57BL/6 recipients (Fig. 3A). Twelve weeks after 137 reconstitution, percentages and numbers of CD4+Foxp3+ Tregs derived from cd69-/- BM 138 precursors were markedly lower in the thymus (Fig. 3A) and blood (Fig. S3) than those 139 derived from cd69+/+ precursors, indicating an impaired Treg regeneration capacity of 140 cd69-/- BM hematopoietic stem cells. Moreover, we analyzed the potential of these 141 precursors to differentiate to tTregs in sublethally irradiated Rag2-/-γc-/- recipients, which 142 lack lymphoid cells (Fig. 3B). Because Rag2-/-γc-/- recipient mice lack NK cells, we 143 depleted donor BM precursors of T cells before transplant to avoid graft versus host disease 144 (17). As before, cd69-/- BM precursors had the lowest tTreg regeneration potential, even 145 though in both systems there were no differences in CD4SP cell numbers between thymus 146 of chimeric mice from cd69+/+ and cd69-/- BM precursors (Fig. 3A and B). These results 147 suggest that the differences observed in the percentage of CD4+Foxp3+ tTregs in the 148 thymus are due to impaired differentiation of this cell subset and not to a defective 149 thymocyte egress (Fig. 3A and B). 150 Finally, to definitely rule out that the differences observed are due to differential egress 151 between cd69+/+ and cd69-/- thymocytes (Fig. 1D), we generated mixed BM chimeric mice 152 by reconstituting sublethally irradiated Rag2-/-γc-/- mice with a 1:1 mixture of wild-type 153 (B6SJL) CD45.1 and cd69-/- CD45.2 BM hematopoietic stem cells from either Foxp3-154 8 reporter (Fig. 3C) or non-reporter mice (Fig. S4A). Thymuses, spleens, lymph nodes and 155 blood were harvested starting from 8 to 10 weeks after transfer. CD4+Foxp3+ Tregs 156 generated from cd69+/+ and cd69-/- precursors were analyzed separately (Fig. 3D, Fig. S4B 157 and Fig. S5). We detected a marked difference in the frequencies of CD4+Foxp3+ tTregs 158 and pTregs originating from the two precursors in both models, with a lower proportion of 159 Tregs derived from cd69-/- CD45.2 CD4+ SP precursors than from cd69+/+ CD45.1 160 precursors (Fig. 3D and E, Fig. S4C and Fig.S5A-C); this occurred even though CD4+ SP 161 cells originating from both precursors in equal proportions in the thymus (Fig. 3E), spleens 162 (Fig. S5A) and lymph nodes (Fig. S5B). These data indicate that cd69+/+ BM hematopoietic 163 stem cells are necessary for the generation of CD4+Foxp3+ tTregs and subsequently pTreg 164 homeostasis. Our data are consistent with the finding that Treg precursors in human thymus 165 form part of the CD69+ thymocyte cell subset (11). 166 167 CD69 deficiency impairs STAT5 signaling and BIC/miR-155-dependent tTreg 168 differentiation 169 To investigate the mechanism of CD69-modulated tTreg development, we examined the 170 Stat5 pathway that stimulates foxp3 promoter, inducing tTreg development (4). Sorted 171 Foxp3-mRFP+-CD69+ and -CD69- Tregs from wild type reporter mice (Fig. S6), were 172 analyzed by intracellular staining and western blot. The analysis showed diminished 173 STAT5 phosphorylation in sorted CD69- tTregs in steady state (Fig. 4A and B), indicating 174 that CD69 expression maintains STAT5 bystander activation of tTregs within the thymus. 175 The analysis of spleen sorted pTregs confirmed diminished STAT5 phosphorylation in 176 secondary lymphoid organs (Fig. S7A). Although we detected no differences in Foxp3 177 activation or expression between CD69 expressing and non-expressing tTregs (Fig. 4C and 178 9 Fig. 1E) or pTregs (12), the transcriptional activation of bic was abrogated in CD69- tTregs, 179 and consequently miR-155 expression was inhibited in those cells (Fig. 4D) and pTregs 180 (Fig. S7B). It has been reported that miR-155 inhibits the expression of suppressor of 181 cytokine signaling 1 (SOCS-1), supporting Foxp3+ tTreg development (6). Importantly, 182 socs-1 gene and protein expression were both upregulated in CD69- tTregs (Fig. 4E and F) 183 and pTregs (Fig. S7B), which had very low levels of miR-155. Moreover, we analyzed 184 STAT5 pathway in sorted tTregs from cd69+/+, cd69+/- and cd69-/- Foxp3-mRFP-reporter 185 mice. STAT5 phosphorylation is partially inhibited in cd69+/- compared to cd69-/- tTregs, 186 that have almost abrogated the pathway (Fig. 4G). Thus, cd69+/- and cd69-/- tTregs have 187 very low levels of miR-155 compared to cd69+/+. Accordingly, socs-1 gene is modestly and 188 strongly upregulated in cd69+/- and cd69-/- tTregs, respectively (Fig. 4H). Our data suggest 189 that the loss of at least one cd69 allele modifies at least in part the expression of the 190 receptor on the membrane (Fig. 1A and B), but is sufficient to prevent fully activation of 191 STAT5 pathway, miR-155 transcription, SOCS-1 inhibition and proper differentiation of 192 tTregs. 193 The overexpression of SOCS-1 regulates STAT5 signaling reducing the proportion of 194 tTregs in cd69-/- mice to levels similar to mirn155-/- mice (6). We analyzed the 195 CD69+/CD69- ratio within tTreg and pTreg cells from mirn155-/- mice (Fig. 5A and C). 196 Consistent with the previous work, mirn155-/- mice display impaired numbers of tTregs and 197 pTregs, as well as an important reduction in the development of CD69+ Tregs, both in 198 thymus (Fig. 5A) and spleens (Fig. 5C). Interestingly, cd69 gene expression was almost 199 abrogated in the thymus of mirn155-/- mice (Fig. 5B), suggesting that cd69 and mirn155 200 could have common regulation pathways. 201 10 In agreement, we found that cd69+/- and cd69-/- thymic precursors are less able to 202 differentiate towards tTregs than CD69-proficient precursors in the same mice (Fig. 1F and 203 Fig. 3E). These data thus strongly suggest that the maintenance of miR-155 expression in 204 tTregs is dependent on CD69-induced STAT5 phosphorylation, reflecting a unique 205 property of CD69 in the development of tTregs. 206 207 Both IL-2Rγ and CD69 signaling is required for the development of in vitro-inducible 208 CD25+ Treg cells 209 To further explore the non-redundant role of CD69 in the development of in vitro-inducible 210 Tregs (iTregs), we analyzed the levels of Foxp3 in the absence of Jak3-STAT5 signaling. 211 We cultured CD4 naïve T cells under Treg-skewed conditions with TGFβ plus IL-2 in the 212 presence of antigen presenting cells. The use of Jak3 chemical inhibitors decreased STAT-5 213 phosphorylation in cd69+/+ iTreg cells to cd69-/- iTreg levels (Fig. 6A), however the 214 percentage of Foxp3-mRFP+ cells is comparable in both genotypes, even high in cd69-/- 215 Tregs cultures and independently of Jak-STAT5 inhibition (Fig.6B), indicating that Jak3-216 STAT5 signaling pathway is not required for Foxp3 expression of inducible Tregs, 217 corroborating previous data in tTregs (Fig. 4C). 218 It has been described that Foxp3 expression is dependent of IL-2Rγc, thus Il2rγ-/- mice had 219 no detectable Foxp3+ cells in thymus or spleen (18). However, the expression of CD25+ 220 Tregs is detectable in thymus and spleen of those mice (18). We aimed to address the role 221 of CD69 in the development of CD25+ iTregs in the absence of IL-2Rγ/Foxp3 signaling 222 pathways. For that purpose, we generated the double knock-out mice Il2rγ-/-/cd69-/-. We 223 analyzed the levels of CD25+ iTreg cells after induction with TGFβ plus IL-2 in the 224 11 presence of Jak3 inhibitors in cells from Il2rγ-/- mice compared to Il2rγ-/-/cd69-/- mice. Jak3 225 inhibition decreased STAT-5 phosphorylation in Il2rγ-/- iTreg cells to the levels of Il2rγ-/-226 /cd69-/- Tregs (Fig. 6C). Interestingly, the differentiation of CD25+ iTreg cells is completely 227 abolished in both, Il2rγ-/-/cd69-/- iTregs and Il2rγ-/- iTreg plus Jak3 inhibitors (Fig. 6D). 228 These data indicate that, in the absence of IL-2Rγ/Foxp3 pathway, CD69-induced Jak3-229 STAT5 activation is pivotal for the development of CD25+ iTreg cells. 230 It has been proposed that miR-155 could regulate different cell type functions depending on 231 the biological context, and miR-155 mediated SOCS-1 repression regulates the competitive 232 fitness of Treg cells (19). We analyzed the expression of mir-155, socs-1, T-bet and Eomes 233 in order to investigate if other miR-155 target genes are affected in iTregs differentiation in 234 the absence of Jak3-STAT5 signaling pathway activation through CD69. We observed 235 diminished expression of miR-155 in cd69-/- compared to cd69+/+ iTreg cells (Fig. S8A), as 236 in ex-vivo CD69- Thymus-derived Tregs (Fig. 4D and G). However, Jak3 inhibition does 237 not contribute to miR-155 inhibition (Fig. S8A), suggesting that other signaling pathways 238 could contribute to miR-155 regulation in iTregs. Moreover, socs-1 expression is strongly 239 induced in cd69-/- iTreg cells compared to cd69+/+ iTregs (Fig. S8B), but not other miR-155 240 target genes as T-bet and Eomes (Fig. S8C). Interestingly, Jak3 inhibits the expression of 241 socs-1, T-bet and Eomes in the absence of CD69 (Fig. S8B and C), supporting the 242 hypothesis that other CD69-dependent mechanisms could be involved in the regulation of 243 those target genes. Altogether, these data suggest that CD69 controls socs-1 expression and 244 Tregs differentiation through miR-155 regulation, although other molecules could be 245 involved in the process. 246 247 12 Expression levels of miR-155 and CD69 are co-regulated in a positive feedback loop 248 CD69 and BIC/miR-155 promoter sequences have two putative STAT5 binding elements 249 upstream of the TATA box and AP-1 element (20) (Fig. S9). Moreover, the transcription 250 factor AP-1, highly induced after TCR stimulation, regulates the activation of both 251 promoters (20, 21), suggesting that both promoters might be concomitantly activated, in a 252 positive feedback loop, by the same TCR/CD3 triggered pathway (Fig. S9). To test this 253 hypothesis, we next investigated whether CD69 downstream signaling regulates miR-155 254 expression in tTregs. Sorted Foxp3+ tTregs from Foxp3-mRFP/cd69+/+ mice, expressing 255 CD69 in steady state, were incubated with anti-CD69 antibody (2.2), which downregulates 256 CD69 membrane expression and dampens its signaling (22) (Fig. 7A). As described above, 257 we observed strong CD69 dampening on the membrane compared with cells incubated with 258 control mouse IgG1 mAb (2.8) (Fig. 7A). qPCR analysis revealed decreased miR-155 259 expression in 2.2-treated CD69+ tTregs (Fig. 7B), to levels comparable to CD69- or cd69-/- 260 tTregs (Fig. 7B and Fig. 4D and H). Moreover, CD69 blockade with 2.2 Abs impairs 261 STAT5 phosphorylation (Fig. 7C) and prevents SOCS1 inhibition (Fig. 7D), meaning that 262 CD69 expression is necessary for miR-155-dependent inhibition of SOCS1 and bona fide 263 formation of tTregs. 264 To verify whether these findings could be extended to human cells, activated CD4+CD25+ 265 PBLs were infected with lentiviruses (LV) carrying different shCD69 sequences (shCD69-1 266 to -3). Endogenous levels of membrane CD69 and hsa-miR-155 were analyzed by FACS 267 and qPCR, respectively (Fig. 8A and B). LV infection of PBLs with three shCD69 268 sequences fully inhibited CD69 expression compared to Mock LV infection (Fig. 8A), 269 inducing loss of hsa-miR-155 transcription (Fig. 8B). Our data indicate that human CD69 270 and hsa-miR-155 are regulated together as in mouse cells. In parallel, we induced the 271 13 expression of CD69 in vitro (Fig. 8C) to corroborate that the STAT5 pathway and hsa-miR-272 155 are activated together with the receptor, whereas SOCS1 is inhibited (Fig. 8D). 273 To test this mechanism functionally, we performed loss and gain of function assays by 274 transfecting human Tregs with anti-hsa-miR-155 or hsa-pre-miR-155. First, we transfected 275 control and anti-CD3 (OKT3)-stimulated CD4+CD25+ human PBLs with anti-hsa-miR-276 155-5p or scrambled anti-miRNA (Fig. 8E). CD69 expression in activated PBLs drops 277 dramatically after inhibition of hsa-miR-155 (Fig. 8F). Moreover, STAT5 activation was 278 reduced and, in agreement, socs1 gene expression was enhanced, indicating that miR-155 279 blockade regulates CD69 signaling pathway. By contrast, overexpression of hsa-miR-155 280 in CD69- Tregs (Fig. 8G) revealed a significant increase in the expression of CD69, STAT5 281 activation and socs1 inhibition (Fig. 8H). Thus, the reciprocal modulation of the C-type 282 lectin and miR-155 in a positive feedback loop could be pivotal to maintain tTregs fitness 283 and pTregs homeostasis. 284 285 286 14 Discussion 287 In this study we have shown that the C-type lectin CD69 plays a key role in the 288 development and homeostasis of Tregs. Using a combined genetic model of Foxp3-reporter 289 and cd69-knockout mice and genetic inhibition approaches, we unequivocally demonstrate 290 that the activation of CD69 pathway promotes STAT5 phosphorylation, BIC/miR-155 291 expression and SOCS-1 inhibition. The role of CD69 as a negative regulator of the immune 292 system has remained a controversial issue during the last years (23). However, very recent 293 studies by independent groups, show that CD69 plays a crucial role in the suppressor 294 function of mice and human Tregs, as well as in the generation of in vitro-induced Treg 295 cells (12, 16, 24-26). Nevertheless, the specific role of the C-type lectin in the development 296 of Tregs in the thymus remains elusive. 297 A major issue that has limited this study has been the key role of CD69 in the egress of 298 lymphocytes from lymphoid organs and in particular from the thymus to periphery (13, 14, 299 27-29). Although thymic positive and negative T-cell selection processes are unaffected by 300 CD69 deficiency (30), CD69 controls the egress of mature T cells into the periphery via 301 cortico-medullary blood vessels, through the negative regulation of S1P1 receptors (27, 28), 302 making it not an easy task to study its role in the development of Tregs in the thymus. With 303 the help of Foxp3-reporter mice, we have performed the study of tTregs differentiation in 304 FTOC and in mixed chimeric mice to avoid the effects derived from the different migratory 305 potential of CD69+ and CD69- cells. We demonstrate that the expression of the C-type 306 lectin CD69 is pivotal for tTregs development as they are virtually absent in FTOC cultures 307 from cd69-/- or anti-CD69-treated embryonic thymuses, or in mixed bone marrow chimeras 308 from cd69-/- precursors. In both systems, total numbers of cells within the thymus do not 309 15 change, whereas tTregs proportions originated from CD69- precursors are consistently 310 diminished, demonstrating unequivocally that this effect is not due to a different migratory 311 behavior. 312 We have found that Foxp3+pTregs are also diminished after analysis of spleen and lymph 313 nodes from adult Foxp3-mRFP/cd69-/- reporter compared to cd69+/+ and cd69+/- littermates. 314 In addition, CD69-deficient pTregs have defective suppressive function (12). Thus, defects 315 observed in CD69 deficient precursors affect both, tTreg development and pTreg 316 homeostasis, strongly indicating that CD69 proficient precursors give rise to the CD69+ 317 functionally active pTreg subset. In this regard, two different genetic approaches in mice 318 and a recent study in humans indicate that CD69 expression in pTregs is required to 319 maintain immunological tolerance. CD69-deficiency in mice compromises T-cell induced 320 colitis and the establishment of oral tolerance after antigen challenge in vivo (24), and 321 CD69+ pTregs are essential for the prevention of asthmatic reactions to harmless antigens 322 (12). Furthermore, a subset of CD69+ Tregs in the blood of healthy human donors seems to 323 have a relevant immune-regulatory role (25). 324 The C-type lectin CD69 interacts with Jak3/STAT5 proteins independently of the IL-2 325 pathway, thus inhibiting Th17 responses (31) and controlling the suppressor potential of 326 pTregs (12). STAT5 phosphorylation stimulates foxp3 promoter, inducing tTreg 327 development (4), and Foxp3 binds to an intron within the promoter region of the miR-155 328 host gene bic in Tregs (32). Mirn155-/- and bic-/- mice both have below normal Foxp3+ Treg 329 numbers in thymus and secondary lymphoid organs, indicating an essential role for miR-330 155 in the development of Foxp3+ Tregs (5, 6). We have explored if this pathway could be 331 the responsible for the defects observed in Treg development in cd69-deficient mice, 332 finding a strong inhibition of STAT5 phosphorylation in freshly isolated Foxp3-mRFP+-333 16 CD69- compared to -CD69+ tTregs. Moreover, bic/miR155 transcriptional levels are 334 reduced in Foxp3-mRFP+-CD69- Tregs and consequently, its target SOCS-1 is up-regulated 335 both at mRNA and protein levels. In a mouse model of SOCS-1 overexpression, negative 336 regulation of STAT5 signaling reduces the proportion of Foxp3+ thymocytes to levels 337 similar to those seen in mirn155-/- mice (6). MiR-155 inhibits SOCS1 expression, 338 enhancing Foxp3+ tTreg development (6). Our data demonstrate that CD69 expression 339 enhanced BIC/miR-155 transcription, inhibits SOCS-1 and therefore maintains Tregs 340 differentiation and fitness of Tregs. However, IL-2R signaling also activates Jak/STAT5 341 pathway in Tregs, specifically Foxp3 expression is dependent on IL-2Rγc signaling as 342 Il2rγ-/- mice have no detectable Foxp3+ cells, although a small proportion of CD25+ Tregs 343 are still detectable in these mice (18). Our study shows that the differentiation of CD25+ 344 iTreg cells is inhibited in Il2rγ-/- cultures plus Jak3 inhibitors or Il2rγ-/-/cd69-/- mice, 345 indicating that Jak3-STAT5 signaling pathway activation through CD69 is essential for the 346 development of Tregs. 347 CD69 does not appear as a miR-155 target in the PicTar, Targetscan or miRanda miRNA 348 target prediction databases, and there are no miR-155 target sequences in the CD69 349 3´untranslated region (UTR) (33). However, several studies have shown a correlation 350 between Dicer, a member of the RNAseIII complex that processes pre-miRNAs into mature 351 miRNAs, miR-155 regulation, and CD69 expression. Tregs from MRL/lpr mice are Dicer 352 insufficient, and yet overexpress miR-155 and show increased CD69 expression (9), 353 suggesting that there are Dicer-alternative mechanisms for miRNA regulation. In another 354 study, Dicer-/- TCs showed increased CD69 expression after TCR stimulation and, 355 consequently, defective egress from lymphoid organs (10). As we described for CD69 356 17 above, Dicer plays a key role in tTreg differentiation (7) and Treg function (8). In this 357 regard, CD69 is expressed in lymphocytes early after TCR/CD3 stimulation (34) and its 358 cytoplasmic tail interacts with Jak3/STAT5 molecules (35), triggering this pathway in 359 pTregs (12) and tTregs and therefore inhibiting SOCS-1 transcription and protein 360 expression. Similarly, TCR-induced IL-2 signaling triggers STAT5 signaling and enhances 361 Foxp3-dependent miR-155 expression, limiting SOCS-1 expression and promoting Treg 362 homeostasis (6). Recent data shows that microRNAs could regulate different cell type 363 functions modulating different target genes, depending on the biological context (19). We 364 analyzed the expression of miR-155 and SOCS1 in the absence of Jak3-STAT5 signaling 365 pathway activation through CD69 in the differentiation of iTreg cells. miR-155 expression 366 is inhibited and SOCS-1 is up-regulated in cd69-/- compared to cd69+/+ iTreg cells, however 367 Jak3 inhibition does not contribute to miR-155 dampening suggesting that other 368 microRNAs and/or target genes could be involved. Interestingly, the STAT5 binding 369 elements of the BIC/miR-155 and CD69 promoter sequences are similar, with each 370 containing two putative STAT binding elements upstream of the TATA box and AP-1 371 element (20). Moreover, the transcription factor AP-1, highly induced after TCR 372 stimulation, regulates the activation of both promoters (20, 21). This suggests that both 373 promoters might be concomitantly activated, in a positive feedback loop, by the same 374 TCR/CD3 triggered pathway. 375 Our present study shows that Foxp3-RFP/cd69-/- reporter mice have dramatically reduced 376 tTreg cell population in adult thymus. Moreover, tTregs are unable to develop properly in 377 FTOC cultures from cd69-/- or anti-CD69-treated embryonic thymuses, or in mixed bone 378 marrow chimeras from cd69-/- precursors. The in vitro data confirm that phosphorylation of 379 STAT5 is abrogated in CD69-deficient tTregs and results in inhibition of the BIC/miR-155 380 18 pathway, increased SOCS-1 expression and impaired tTreg development. Our previous 381 studies show that the suppressor function of Tregs is compromised in cd69-deficient mice 382 (12), indicating that CD69 is a key molecule in the development of Foxp3+CD69+ Tregs in 383 the thymus that will give rise to the functionally active subset of Tregs in the periphery. 384 Therefore, we postulated the C-type lectin CD69 as a pivotal molecule for the maintenance 385 of immune homeostasis in health and disease. 386 387 388 389 19 Material and Methods 390 391 Mice. cd69–/– mice were generated in the 129/Sv background as described (31), and 392 backcrossed onto C57BL/6 for at least 12 generations. C57BL/6.Ly5.1 mice (CD45.1+) 393 were purchased from The Jackson Laboratory (B6.SJL-Ptprca Pepcb/BoyJ: 002014). Rag2-/-394 γc-/- (Rag2/Il2rg) mice were provided by the Dr. ML. Toribio´s laboratory (Centro de 395 Biología Molecular, CSIC, Spain) and were intercrossed with C57BL/6 mice to generate 396 the Il2rγ-/- mice that were subsequently intercrossed with cd69-/- mice, to generate the Il2rγ-397 /-/cd69-/- mice. FoxP3-mRFP reporter mice (FIR mice, C57BL/6 background) were 398 generated and provided by the Flavell laboratory (Yale University School of Medicine, 399 New Haven, CT) (36), and were intercrossed with the cd69-/- mice to generate the Foxp3-400 mRFP/cd69+/+ wild-type, Foxp3-mRFP/cd69+/- heterozygous, and Foxp3-mRFP/cd69-/- 401 CD69-deficient littermates. Animals were housed and used in specific pathogen-free (SPF) 402 conditions at the CNIC animal facility. mirn155-/- mice were provided by Dr. R. Nakagawa 403 (The Francis Crick Institute, London). All animal procedures were approved by the ethics 404 committee of the Comunidad Autónoma de Madrid and conducted in accordance with the 405 institutional guidelines that comply with the European Institutes of Health´s; Directive 406 2010/63/EU of the European Parliament and the Council on the Protection of Animals 407 Used for Scientific Purposes (Official Journal of the European Union. Vol. 53:33-79, 408 2010). 409 410 Intracellular staining and FACS. Single-cell suspensions were obtained from adult or 411 fetal thymuses and incubated in FACS buffer (PBS 0.5% BSA, 1μM EDTA, 0.1% NaN3) 412 with fluorochrome-conjugated mouse-specific antibodies against CD4, CD8, CD69, 413 20 CD45.1 and CD45.2. All antibodies were purchased from BD Biosciences. For Foxp3 414 intracellular staining, we used the Foxp3 staining kit (eBioscience). CD69+- and CD69--415 Foxp3-mRFP+ tTreg cells were sorted from Foxp3-mRFP/cd69+/+ thymus using a 416 FACSAria III (BD). For intracellular STAT5 staining, sorted tTregs were fixed with 0.2% 417 paraformaldehyde and permeabilized with 90% methanol, and cells were incubated with 418 anti-Phospho-Stat5 (Tyr694) (Cell Signaling), Alexa Fluor 647 IgG1 Isotype Control and 419 Alexa Fluor 647 anti-phospho-STAT5 (pY694) (Beckton Dickinson). Human PBLs were 420 obtained after Ficoll separation from buffy coats and maintained in RPMI medium 421 supplemented with 10% FCS, 20 mM HEPES, L-glutamine, antibiotics, non-essential 422 aminoacids, sodium pyruvate and -mercaptoethanol. Treated PBLs were incubated with 423 fluorochrome-conjugated human-specific antibodies against CD4, CD25 and CD69 (BD 424 Biosciences) and Foxp3 (Miltenyi Biotec). Cells were analyzed in an LSRFortessaTM flow 425 cytometer (BD) equipped with four lasers (405, 488, 561 and 640 nm), and the data were 426 processed with FlowJo v10.0.4 (Tree Star). 427 428 Fetal Thymus Organ Culture. Uteri were removed female mice at the indicated 429 gestational time points and the embryos were placed in a Petri dish with fresh cold PBS for 430 extraction of thymuses. To place the fetal thymus lobes in culture, we placed 0.8 μm 431 nitrocellulose membrane filters (Millipore) on top of 12-7 mm Gelfoam sponges embedded 432 in pre-warmed IMDM medium (supplemented with 10% FCS, L-glutamine, antibiotics, and 433 β-mercaptoethanol). FTOCs were maintained for 4 to 14 days with medium replaced every 434 3 days. Anti-CD69 monoclonal antibody (2.2) or the isotype control antibody (2.8) was 435 added (50μg/ml) to the culture medium as indicated and replaced every 3 days. At the end 436 21 of the culture period, single-cell suspensions were prepared from the lobes, and cells were 437 counted and analyzed by FACS. 438 439 Western blotting. Lysates of sorted CD69+- and CD69--Foxp3-mRFP+ tTreg cells were 440 prepared in PD buffer (40mM Tris HCl pH 8.0, 0.5M NaCl, 6 mM EDTA, 6 mM EGTA, 441 0.1% NP40) containing protease inhibitor cocktail (Complete Mini, Roche). Proteins (20 442 g) were size-separated on 12% SDS-polyacrylamide gels and transferred onto Trans-Blot 443 nitrocellulose membranes (BioRad). Primary antibodies for immunoblot were as follows: 444 anti-β-actin, anti-SOCS-1 and anti-STAT5 (Santa Cruz); anti-phospho-STAT5 (Cell 445 signaling). Quantitative assessment of protein expression was performed with the Odyssey 446 scanner and analyzed with Image Studio Lite v4.0 western blot analysis software (LI-447 COR). 448 449 In vitro differentiation of Tregs. Inducible Tregs were differentiated from Foxp3-450 mRFP/cd69+/+, Foxp3-mRFP/cd69-/-, Il2rγ-/-/cd69-/- and Il2rγ-/- mice. Naïve CD4 T cells 451 from these mice were isolated and co-cultured 72h with irradiated antigen presenting cells 452 in the presence of plate bound anti-CD3 (2μg/ml) and soluble anti-CD28 (2μg/ml) plus 453 recombinant TGF-β1(10ng/ml) and IL-2 (2ng/ml). The last 9 hours the cells were incubated 454 with or without JAK 3 inhibitor I (CAS 202475-60-3 – Calbiochem) (10μg/ml). For 455 experiments with inhibitor antibodies, after differentiation Treg cells were cultured 4h with 456 anti-2.2 Ab or 2.8 isotype control. 457 458 22 RNA extraction and gene expression analysis. RNA and microRNA were extracted from 459 2- to -6x104 sorted mouse tTregs or 106 human PBLs with the miRNeasy mini kit (Qiagen), 460 followed by DNAse treatment with the Turbo DNAse-free kit (Ambion). For analysis of 461 SOCS-1, Foxp3 and BIC transcripts, reverse transcription was performed using the High 462 Capacity cDNA reverse transcription kit (Applied Biosystems). SOCS-1 and Foxp3 gene 463 expression was analyzed by real-time PCR using SYBR green PCR mix (Applied 464 Biosystems). Mouse and human Gapdh genes were used as the endogenous control. The 465 following primers were used to amplified murine genes: Socs-1, (F) 5´-466 CTGCGGTTCTATTGGGGAC-3´, (R) 5´-AAAAGGCAGTCGAAGGTCTCG-3´; Foxp3, 467 (F) 5´-CACCCAGGAAAGACAG CAACC-3´, (R) 5´-GCAAGAGCTCTTGTCCATTGA-468 3´; cd69, (F) 5´-CCCTTGGGCTGTGTTAATAGTG-3´, (R) 5´-469 AACTTCTCGTACAAGCCTGGG-3´ and Gapdh, (F) 5´-TGAAGCAGGCATCTGAGGG-470 3´, (R) 5´-CGAAGGTGGAAGAGTGGGAG-3´. The following primers were used to 471 amplified human genes: socs-1, (F) 5´-TTTTCGCCCTTAGCGTGAAGA-3´, (R) 5´-472 GAGGCAGTCGAAGCTCTCG-3´, and gapdh (F) 5´-AATGGACTGGTCGTGGAG-3´, 473 (R) 5´-CCCTCCAGGGGATCGTTTG-3´. BIC gene expression was analyzed by real-time 474 PCR using TaqMan Universal PCR Master mix and specific TaqMan probe and primers for 475 bic (assays ID Mm01716204-m1 and ID Hs01374570-m1) (Applied Biosystems). 476 Expression of microRNA was analyzed using TaqMan MicroRNA Reverse Transcription 477 Kit, individual TaqMan MicroRNA Assays for mmu-miR-155-5p (Ref. 002571) and hsa-478 miR-155-5p (Ref. 002287) and TaqMan Universal PCR Master mix (Applied Biosystems). 479 sno135 snRNA (Ref. 001230) was used as the endogenous control. Real-time Quantitative 480 23 PCR analysis was performed with an ABI Prism 7900HT 384 thermal cycler (Applied 481 Biosystems). Relative gene expression was determined using the 2-ΔΔCT method. 482 483 Chimeric mice. Eight-twelve-week-old Rag2-/- γc-/- recipient mice were irradiated with one 484 split dose of 6.5 Gy γ-radiation, whereas C57BL/6 recipients were irradiated with two split 485 6.5 Gy doses. The mice were i.v. injected with bone marrow cells from Foxp3-486 mRFP/cd69+/+ or Foxp3-mRFP/cd69-/- littermates. In mixed chimeras, irradiated Rag2-/- γc-487 /- recipients were transplanted with a mixture of CD45.1 cd69+/+ or CD45.2 cd69-/- bone 488 marrow precursors from non-reporter or reporter Foxp3-mRFP+, at a ratio of 1:1. After at 489 least 10 weeks, the contribution of the different donor bone marrow precursors to the tTreg 490 subset was determined by FACS. 491 492 Transient transfection. PBLs (106) cells were transiently transfected for 4 hours with 50 493 pM of anti-miR-155 (AM12601 Ambion) by Lipotransfectin (Niborlab) according to the 494 manufacturer´s instructions. As a negative control, random anti-miR sequence control 495 (AM1701 negative control #1 Ambion) was included in the assay. Transfected cells were 496 stimulated with plate-bound anti-CD3 antibody (OKT3; 3 ug/mL) for 24 hours. When 497 indicated, PBLs (0.5x106) cells were transfected for 7h with 50 pM of Pre-miR-hsa- miR-498 155 or Pre-miR negative control (Ambion) by Lipofectamine RNA iMAX (Invitrogen). 499 Transfected cells were stimulated with PMA during 4 h with 50ng/ml phorbol myristate 500 acetate (PMA) and 750ng/ml ionomycin (P+I). After stimulation, the levels of CD69 and 501 phospho-Stat5 were monitored by flow cytometry and transcriptional levels of hsa-miR-502 155 and socs1 were monitored by qPCR. 503 24 504 Plasmids. The pLKO lentiviral plasmids containing shCD69 sequences were from Sigma 505 Aldrich (TRCN0000057693; TRCN0000057694; TRCN0000057695) and the pLKO 506 lentiviral control plasmid is a pLKO empty vector from Sigma Aldrich (Ref. SHC001). The 507 shCD69 sequences used were as follows: SHCD69-1 (5´-3´): 508 CCGGGCATGGAATGTGAGAAGAATTCTCGAGAATTCTTCTCACATTCCATGCTT509 TTTG; SHCD69-2 (5´-3´): 510 CCGGAGGCCAATACACATTCTCAATCTCGAGATTGAGAATGTGTATTGGCCTTT511 TTTG; SHCD69-3 (5´-3´): 512 CCGGGTGGTCAAATGGCAAAGAATTCTCGAGAATTCTTTGCCATTTGACCACTT513 TTTG. 514 515 LV production, titration, and infection. HEK-293 cells were cultured in DMEM 516 containing 10% FBS (Sigma Aldrich) and L-glutamine plus antibiotics. HEK-293 were 517 transiently transfected by the calcium phosphate method with 3 HIV-derived plasmids and 518 the VSV pseudotyped LV system (provided by F. Sánchez-Madrid, Hospital de la Princesa, 519 Spain) to obtain LV expressing the shCD69 sequences. The supernatant containing LV 520 particles was collected 48 hours after removal of the calcium phosphate precipitate and 521 ultracentrifugated for 2 hours (Optima L-100 XP Ultracentrifuge Beckman). LVs were 522 collected by adding cold PBS and were titrated by qPCR. PBLs isolated from healthy 523 donors were infected with LV particles (MOI=10) for 5 hours. Subsequently, virus-524 containing medium was replaced with fresh complete RPMI medium supplemented with 525 10% FBS. After 12 hours, infected cells were selected with puromycin for 48 hours. 526 Selected cells were stimulated with plate-bound anti-CD3 antibody (OKT3; 3 μg/mL) for 527 25 24 hours. After stimulation, the levels of CD69 were monitored by flow cytometry and 528 levels of miRNA 155 were monitored by Taqman qPCR. 529 530 Statistical analysis. Experiments were performed according to a randomized complete 531 block design (treatments and different time points have been taken into account) or a fully 532 randomized design. To determine significant differences, P values were calculated by 533 Student’s t test as appropriate, and differences were considered significant values at 534 P<0.05. Means of more than two experimental groups were compared by 1-way ANOVA. 535 To account for multiple comparisons, the Tukey was used to compared selected pairs of 536 means, and the Bo; ;nferroni post-test was used to compare all pairs of means. All statistical 537 analyses were carried out with Prism v5 (GraphPad Software). Each experiment was 538 repeated at least three times, unless otherwise indicated in the figure legends. 539 540 541 542 26 Acknowledgements 543 We thank S. Bartlett for editorial assistance and Prof. Richard A. Flavell (Yale University, 544 New Haven, CT) for kindly provided the foxp3-mRFP reporter mice. This study was 545 funded by grant from the Spanish Ministry of Economy and Competitiveness (SAF2013-546 44857-R to M.L.T); grant INDISNET 01592006 from the Comunidad de Madrid to P.M. 547 and FSM; grant from Instituto de Salud Carlos III (PI-FIS-2016-9488 to P.M) and CIBER 548 Cardiovascular (CIBER CV) to. F.S-M and P.M.; Fundació La Marató TV3 (20152330 31) 549 to P.M. and F.S-M. R.S.D. was funded by a predoctoral fellowship from the Comunidad de 550 Madrid; S.L. was funded by a contract from the RETICS Enfermedades Cardiovasculares 551 (Instituto de Salud Carlos III); K.T. is co-funded by the EU Marie Curie Program 552 (COFUND CNIC IPP). The CNIC is supported by the Ministry of Economy, Industry and 553 Competitiveness (MINECO) and the Pro CNIC Foundation, and is a Severo Ochoa Center 554 of Excellence (MINECO award SEV-2015-0505). 555 556 Authors contributions 557 R.S-D. and R.B-D. performed research and analyzed the data; S.L., K.T., H. dlF., B.L-P. 558 and E.M-G. performed research; R. N. contributed with mirn155-/- mice; F.S-M. and 559 M.L.T. designed research and analyzed the data.; P.M. designed research, collect and 560 analyzed the data and wrote the paper. 561 Competing financial interests 562 The authors declare no that they have no competing interest. 563 564 565 27 References 566 1. Campbell, D. J., and M. A. Koch. 2011. Phenotypical and functional specialization 567 of FOXP3+ regulatory T cells. Nat Rev Immunol 11: 119-130. 568 2. Abbas, A. K., C. Benoist, J. A. Bluestone, D. J. Campbell, S. Ghosh, S. Hori, S. 569 Jiang, V. K. Kuchroo, D. Mathis, M. G. Roncarolo, A. Rudensky, S. Sakaguchi, E. 570 M. Shevach, D. A. Vignali, and S. F. Ziegler. 2013. Regulatory T cells: 571 recommendations to simplify the nomenclature. Nat Immunol 14: 307-308. 572 3. Josefowicz, S. Z., L. F. Lu, and A. Y. Rudensky. 2012. Regulatory T Cells: 573 Mechanisms of Differentiation and Function. Annu Rev Immunol 30: 531-564. 574 4. Burchill, M. A., J. Yang, K. B. Vang, J. J. Moon, H. H. Chu, C. W. Lio, A. L. 575 Vegoe, C. S. Hsieh, M. K. Jenkins, and M. A. Farrar. 2008. Linked T cell receptor 576 and cytokine signaling govern the development of the regulatory T cell repertoire. 577 Immunity 28: 112-121. 578 5. Kohlhaas, S., O. A. Garden, C. Scudamore, M. Turner, K. Okkenhaug, and E. 579 Vigorito. 2009. Cutting edge: the Foxp3 target miR-155 contributes to the 580 development of regulatory T cells. J Immunol 182: 2578-2582. 581 6. Lu, L. F., T. H. Thai, D. P. Calado, A. Chaudhry, M. Kubo, K. Tanaka, G. B. Loeb, 582 H. Lee, A. Yoshimura, K. Rajewsky, and A. Y. Rudensky. 2009. Foxp3-dependent 583 microRNA155 confers competitive fitness to regulatory T cells by targeting SOCS1 584 protein. Immunity 30: 80-91. 585 7. Cobb, B. S., A. Hertweck, J. Smith, E. O'Connor, D. Graf, T. Cook, S. T. Smale, S. 586 Sakaguchi, F. J. Livesey, A. G. Fisher, and M. Merkenschlager. 2006. A role for 587 Dicer in immune regulation. J Exp Med 203: 2519-2527. 588 8. Liston, A., L. F. Lu, D. O'Carroll, A. Tarakhovsky, and A. Y. Rudensky. 2008. 589 Dicer-dependent microRNA pathway safeguards regulatory T cell function. J Exp 590 Med 205: 1993-2004. 591 9. Divekar, A. A., S. Dubey, P. R. Gangalum, and R. R. Singh. 2011. Dicer 592 insufficiency and microRNA-155 overexpression in lupus regulatory T cells: an 593 apparent paradox in the setting of an inflammatory milieu. J Immunol 186: 924-930. 594 10. Zhang, N., and M. J. Bevan. 2010. Dicer controls CD8+ T-cell activation, 595 migration, and survival. Proc Natl Acad Sci U S A 107: 21629-21634. 596 11. Martin-Gayo, E., E. Sierra-Filardi, A. L. Corbi, and M. L. Toribio. 2010. 597 Plasmacytoid dendritic cells resident in human thymus drive natural Treg cell 598 development. Blood 115: 5366-5375. 599 12. Cortes, J. R., R. Sanchez-Diaz, E. R. Bovolenta, O. Barreiro, S. Lasarte, A. 600 Matesanz-Marin, M. L. Toribio, F. Sanchez-Madrid, and P. Martin. 2014. 601 Maintenance of immune tolerance by Foxp3+ regulatory T cells requires CD69 602 expression. J Autoimmun 55: 51-62. 603 13. Feng, C., K. J. Woodside, B. A. Vance, D. El-Khoury, M. Canelles, J. Lee, R. 604 Gress, B. J. Fowlkes, E. W. Shores, and P. E. Love. 2002. A potential role for CD69 605 in thymocyte emigration. Int Immunol 14: 535-544. 606 14. Nakayama, T., D. J. Kasprowicz, M. Yamashita, L. A. Schubert, G. Gillard, M. 607 Kimura, A. Didierlaurent, H. Koseki, and S. F. Ziegler. 2002. The generation of 608 mature, single-positive thymocytes in vivo is dysregulated by CD69 blockade or 609 overexpression. J Immunol 168: 87-94. 610 28 15. Lamana, A., D. Sancho, A. Cruz-Adalia, G. M. del Hoyo, A. M. Herrera, M. Feria, 611 F. Diaz-Gonzalez, M. Gomez, and F. Sanchez-Madrid. 2006. The role of CD69 in 612 acute neutrophil-mediated inflammation. Eur J Immunol 36: 2632-2638. 613 16. Wirnsberger, G., F. Mair, and L. Klein. 2009. Regulatory T cell differentiation of 614 thymocytes does not require a dedicated antigen-presenting cell but is under T cell-615 intrinsic developmental control. Proc Natl Acad Sci U S A 106: 10278-10283. 616 17. Noval Rivas, M., M. Hazzan, K. Weatherly, F. Gaudray, I. Salmon, and M. Y. 617 Braun. 2010. NK cell regulation of CD4 T cell-mediated graft-versus-host disease. J 618 Immunol 184: 6790-6798. 619 18. Fontenot, J. D., J. P. Rasmussen, M. A. Gavin, and A. Y. Rudensky. 2005. A 620 function for interleukin 2 in Foxp3-expressing regulatory T cells. Nat Immunol 6: 621 1142-1151. 622 19. Lu, L. F., G. Gasteiger, I. S. Yu, A. Chaudhry, J. P. Hsin, Y. Lu, P. D. Bos, L. L. 623 Lin, C. L. Zawislak, S. Cho, J. C. Sun, C. S. Leslie, S. W. Lin, and A. Y. Rudensky. 624 2015. A Single miRNA-mRNA Interaction Affects the Immune Response in a 625 Context- and Cell-Type-Specific Manner. Immunity 43: 52-64. 626 20. Yin, Q., X. Wang, J. McBride, C. Fewell, and E. Flemington. 2008. B-cell receptor 627 activation induces BIC/miR-155 expression through a conserved AP-1 element. J 628 Biol Chem 283: 2654-2662. 629 21. Castellanos, M. C., C. Munoz, M. C. Montoya, E. Lara-Pezzi, M. Lopez-Cabrera, 630 and M. O. de Landazuri. 1997. Expression of the leukocyte early activation antigen 631 CD69 is regulated by the transcription factor AP-1. J Immunol 159: 5463-5473. 632 22. Esplugues, E., D. Sancho, J. Vega-Ramos, C. Martinez, U. Syrbe, A. Hamann, P. 633 Engel, F. Sanchez-Madrid, and P. Lauzurica. 2003. Enhanced antitumor immunity 634 in mice deficient in CD69. J Exp Med 197: 1093-1106. 635 23. Gonzalez-Amaro, R., J. R. Cortes, F. Sanchez-Madrid, and P. Martin. 2013. Is 636 CD69 an effective brake to control inflammatory diseases? Trends in molecular 637 medicine 19: 625-632. 638 24. Radulovic, K., C. Manta, V. Rossini, K. Holzmann, H. A. Kestler, U. M. Wegenka, 639 T. Nakayama, and J. H. Niess. 2012. CD69 regulates type I IFN-induced tolerogenic 640 signals to mucosal CD4 T cells that attenuate their colitogenic potential. J Immunol 641 188: 2001-2013. 642 25. Vitales-Noyola, M., L. Doniz-Padilla, C. Alvarez-Quiroga, A. Monsivais-Urenda, 643 H. Portillo-Salazar, and R. Gonzalez-Amaro. 2015. Quantitative and functional 644 analysis of CD69(+) NKG2D(+) T regulatory cells in healthy subjects. Human 645 immunology 76: 511-518. 646 26. Lin, C. R., T. W. Wei, H. Y. Tsai, Y. T. Wu, P. Y. Wu, and S. T. Chen. 2015. 647 Glycosylation-dependent interaction between CD69 and S100A8/S100A9 complex 648 is required for regulatory T-cell differentiation. FASEB journal : official publication 649 of the Federation of American Societies for Experimental Biology. 650 27. Matloubian, M., C. G. Lo, G. Cinamon, M. J. Lesneski, Y. Xu, V. Brinkmann, M. 651 L. Allende, R. L. Proia, and J. G. Cyster. 2004. Lymphocyte egress from thymus 652 and peripheral lymphoid organs is dependent on S1P receptor 1. Nature 427: 355-653 360. 654 28. Shiow, L. R., D. B. Rosen, N. Brdickova, Y. Xu, J. An, L. L. Lanier, J. G. Cyster, 655 and M. Matloubian. 2006. CD69 acts downstream of interferon-alpha/beta to inhibit 656 S1P1 and lymphocyte egress from lymphoid organs. Nature 440: 540-544. 657 29 29. Weinreich, M. A., and K. A. Hogquist. 2008. Thymic emigration: when and how T 658 cells leave home. J Immunol 181: 2265-2270. 659 30. Lauzurica, P., D. Sancho, M. Torres, B. Albella, M. Marazuela, T. Merino, J. A. 660 Bueren, A. C. Martinez, and F. Sanchez-Madrid. 2000. Phenotypic and functional 661 characteristics of hematopoietic cell lineages in CD69-deficient mice. Blood 95: 662 2312-2320. 663 31. Martin, P., M. Gomez, A. Lamana, A. Cruz-Adalia, M. Ramirez-Huesca, M. A. 664 Ursa, M. Yanez-Mo, and F. Sanchez-Madrid. 2010. CD69 association with 665 Jak3/Stat5 proteins regulates Th17 cell differentiation. Mol Cell Biol 30: 4877-666 4889. 667 32. Zheng, Y., S. Z. Josefowicz, A. Kas, T. T. Chu, M. A. Gavin, and A. Y. Rudensky. 668 2007. Genome-wide analysis of Foxp3 target genes in developing and mature 669 regulatory T cells. Nature 445: 936-940. 670 33. Ziegler, S. F., S. D. Levin, L. Johnson, N. G. Copeland, D. J. Gilbert, N. A. Jenkins, 671 E. Baker, G. R. Sutherland, A. L. Feldhaus, and F. Ramsdell. 1994. The mouse 672 CD69 gene. Structure, expression, and mapping to the NK gene complex. J 673 Immunol 152: 1228-1236. 674 34. Testi, R., J. H. Phillips, and L. L. Lanier. 1989. T cell activation via Leu-23 (CD69). 675 J Immunol 143: 1123-1128. 676 35. Martin, P., and F. Sanchez-Madrid. 2011. CD69: an unexpected regulator of TH17 677 cell-driven inflammatory responses. Sci Signal 4: pe14. 678 36. Wan, Y. Y., and R. A. Flavell. 2005. Identifying Foxp3-expressing suppressor T 679 cells with a bicistronic reporter. Proc Natl Acad Sci U S A 102: 5126-5131. 680 681 682 683 30 Figure legends 684 685 Figure 1. CD69 expression is required for thymus-derived Treg cell homeostasis in 686 adult mice. (A) Density plots show CD69 expression in CD4+CD8-Foxp3+ gated 687 thymocytes from 8-12-week-old Foxp3-mRFP/cd69+/+ (wild type), Foxp3-mRFP/cd69+/- 688 (heterozygous), and Foxp3-mRFP/cd69-/- (deficient) reporter littermates. Numbers indicate 689 the proportions (%) of gated cells. (B) The bar chart shows the percentage (± S.D.) of 690 CD69+ (black) and CD69- (white) tTregs within the thymus of the indicated reporter mice. 691 (C) Flow cytometry analysis of thymocyte subsets in 8-10-week-old reporter littermates. 692 The percentages of thymus-derived T cell subsets are shown. (D) Cellularity of the thymus 693 (left) and total number of CD4 SP cells (right) in reporter littermates. (E) Analysis of 694 endogenous Foxp3 expression in tTregs in the thymuses of reporter littermates. (F) 695 Percentages (left) and total cell number (right) of gated CD4+CD8-Foxp3+ tTregs in adult 696 reporter littermates. Data are from at least 7 litters with 3 to 12 littermates each. A total of 697 16 Foxp3-mRFP/cd69+/+ (wild type), 11 Foxp3-mRFP/cd69+/- (heterozygous), and 12 698 Foxp3-mRFP/cd69-/- (deficient) mice were analyzed. Error bars show S.D. Data were 699 evaluated by ANOVA followed by Bonferroni’s multiple comparison test: * P < 0.05, ** P 700 < 0.01, *** P < 0.001. 701 702 Figure 2. tTregs differentiation in fetal thymus organ culture requires CD69 703 expression. (A) Representative density plots of 5 days FTOC from Cd69+/+ and Cd69-/- 704 embryo in the C57BL/6 background. Embryonic thymuses were removed from 15 to 17 705 days old embryos and the percentages of tTregs development in the lobes were analysed by 706 FACS. (B) Cellularity of foetal thymus lobes (left) and total cell number of CD4+Foxp3+ 707 31 (right) from Cd69+/+ and Cd69-/- embryos. (C) FTOCs from wild type 17 days old embryos 708 (E17) were maintained up to 14 days in culture in the presence of anti-CD69 monoclonal 709 antibody (2.2) or the isotype control antibody (2.8). Density plots shows the percentage of 710 tTregs on days 4, 11 and 14 after culture. (D) Cellularity of foetal thymus lobes (left) and 711 total number of CD4+Foxp3+ cells (right) en each condition. A total of 31 and 36 embryos 712 from five Cd69+/+ and four Cd69-/- females respectively, were analysed. The 2 lobes from 713 each fetal thymus were analysed separately. Error bars show S.D. Values are calculated 714 relative to data for Cd69+/+ control lobes from four independent FTOC assays. * P < 0.05, 715 ** P < 0.01, *** P < 0.001 (Student’s t-test). 716 717 Figure 3. CD69+ hematopoietic stem cells are more prone to develop tTregs after 718 reconstitution. (A) Eight to twelve-week-old C57BL/6 (A) or Rag2-/- γc-/- (B) recipient 719 mice received two or one split doses of 6,5 Gy γ-radiation respectively and were i.v. 720 injected with bone marrow cells from Foxp3-mRFP/cd69+/+ or Foxp3-mRFP/cd69-/- 721 littermates. (C) In mixed chimeras, irradiated Rag2-/-γc-/- recipients were transplanted with a 722 mixture of CD45.1-Foxp3-mRFP/cd69+/+ or CD45.2-Foxp3-mRFP/cd69-/- bone marrow 723 precursors at a ratio of 1:1. After at least 10 wks (D), the contribution of the different donor 724 bone marrow precursors to tTreg cells development and CD69 expression in tTregs were 725 determined by FACS (A-D). (E) Percentages of gated CD4+ SP cells and CD4+CD8-Foxp3+ 726 tTregs within CD45.1 or CD45.2 donors in the thymus. All data are representative of at 727 least 3 independent experiments with at least 3 recipient mice per group or 6 recipient mice 728 for mixed chimeras. Error bars show S.D. * P < 0.05, ** P < 0.01, *** P < 0.001 (Student’s 729 t-test). 730 32 731 Figure 4. Expression of miR-155 and target proteins in CD69+ deficient and proficient 732 Treg cells. 733 (A) Left, representative histogram showing the levels of STAT5 phosphorylation analyzed 734 by FACS in CD69+ or CD69- sorted tTregs cells. Right, the levels of STAT5 735 phosphorylation shown as the fold difference compared with isotype control-treated cells. 736 Lines link measurements of CD69+ and CD69- tTregs from the same mouse. (B) 737 Representative WB showing the levels of STAT5 phosphorylation in tTregs sorted as in A. 738 Phosphorylation levels are normalized to STAT5 and β−actin total protein levels. q-PCR 739 analysis of the relative expression of Foxp3 (C), BIC promoter, mmu-miR-155 (D) and 740 socs-1 (E) in CD69+ and CD69- tTregs. Expression was normalized to the levels in CD69+ 741 tTregs. (F) Representative WB of SOCS1 protein expression in CD69+ and CD69- 742 Foxp3mRFP+ sorted tTreg cells. SOCS1 levels are normalized to mean β−actin levels from 743 of at least 4 independent sortings. (G) Left, representative histogram showing the levels of 744 STAT5 phosphorylation in sorted tTregs cells from cd69+/+, cd69+/- and cd69-/- Foxp3-745 reporter mice. Right, quantification of STAT5 phosphorylation levels shown as Geometric 746 mean fluorescence intensity. (H) mmumiR-155 and socs1 transcriptional levels analyzed by 747 qPCR in tTregs cells from cd69+/+, cd69+/- and cd69-/- Foxp3-reporter mice. All data are 748 derived from at least 5 independent sortings/experiments (3 animals per sorting). Data were 749 analyzed by t-test (A-E) except for WB analyses, for which representative gels are shown. 750 Error bars show S.D. ** P < 0.01, *** P < 0.001 (Student’s t-test). (G-H) Data were 751 analyzed by ANOVA followed by Bonferroni’s multiple comparison test: * P < 0.05, ** P 752 < 0.01, *** P < 0.001. 753 33 754 Figure 5. CD69+ Treg development was impaired in the thymus and spleen of 755 mirn155-/- mice 756 Density plots show CD4+CD8-Foxp3+ tTregs and CD69 expression in gated CD4+CD8-757 Foxp3+ gated thymocytes (A) or splenocytes (C), from wild type (WT) or mirn155-/- mice. 758 Numbers indicate the proportions (%) of gated cells. The bar chart shows total cell number 759 (upper) of gated CD4+CD8-Foxp3+ tTregs and the percentage (± S.D.) of CD69+ (black) 760 and CD69- (white) tTregs (lower) within the thymus from WT or mirn155-/- mice. (B) cd69 761 relative expression in thymocytes from WT or mirn155-/- mice analysed by qPCR. All data 762 are derived from 5 mice WT and 3 mirn155-/-. Data were analyzed by t-Test. Error bars 763 show S.D. * P < 0.05 (Student’s t-test). 764 765 Figure 6. CD69 expression rescues iTreg differentiation in the absence of IL-766 2Rγc/Foxp3 signaling pathway. (A) Naïve CD4+ T cells from Foxp3-mRFP/cd69+/+ or 767 Foxp3-mRFP/cd69-/- littermates were cultured for 72 hours under Treg-skewed conditions 768 and treated with a chemical Jak3 inhibitor or an equal concentration of DMSO for the last 9 769 hours. The percentages of Phospho-STAT5+ cells and the levels of STAT5 phosphorylation 770 analyzed by FACS and compared to and isotype Ab are shown. (B) Quantification of 771 reporter Foxp3-mRFP+ cells treated as in A. (C) Naïve CD4+ T cells from Il2rγ-/-/cd69-/- 772 and Il2rγ-/- mice were cultured as in A and the percentages of Phospho-STAT5+ cells and 773 the levels of STAT5 phosphorylation were analyzed by FACS. (D) Quantification of 774 CD25+ Treg cells were analyzed by FACS. Data are from two independent experiment (n=3 775 34 from each genotype). Error bars show S.D. Data were evaluated by ANOVA followed by 776 Bonferroni’s multiple comparison test: * P < 0.05, ** P < 0.01, *** P < 0.001. 777 778 Figure 7. CD69 downstream signaling regulates miR-155, STAT5 and socs1 779 expression in Tregs. (A) Left, representative plots of CD69 expression in sorted mouse 780 Foxp3-mRFP/cd69+/+ tTregs cells treated with anti-CD69 2.2 or 2.8 isotype control. Right, 781 CD69 expression after Ab treatment analyzed by FACS. Bars correspond to the mean ± 782 S.D. of one representative experiment of four. (B) qPCR analysis of mmu-miR-155 783 expression in sorted CD69+ or CD69- Foxp3-mRFP+ tTregs after Ab treatment. Results are 784 normalized by snoRNA135 expression and the expression was relative to 2.8-treated 785 CD69+ cells. (C) Left, representative histogram showing the levels of STAT5 786 phosphorylation in iTreg cells from cd69+/+ or cd69-/- reporter mice treated with anti-CD69 787 2.2 or 2.8 isotype control. Right, quantification of STAT5 phosphorylation levels shown as 788 Geometric mean fluorescence intensity. (D) socs1 transcriptional levels were analyzed by 789 qPCR. Data from A-B are derived from 3 independent sortings/experiments (3 animals per 790 sorting) and iTregs differentiated from at least 4 mice per group (C-D). Data was analyzed 791 by 1-way ANOVA and Bonferroni’s post-test (B). CD69 expression after Ab treatment was 792 analyzed by t test (A). * P < 0.05, ** P < 0.01, *** P < 0.001 (Student’s t-test). 793 794 Figure 8. Co-regulation of CD69 and miR-155 expression in human Tregs. (A) Left, 795 Representative histograms of CD69 expression after LV infection with 3 different shCD69 796 sequences (1-3) or a sh control sequence, stimulated or not with human anti-CD3 Abs 797 (OKT3 clone). Right, CD69 fold induction relative to non-stimulated cells. (B) q-PCR 798 analysis of hsa-miR-155 expression in human CD4+ T cells after LV infection. (C) Human 799 35 PBLs were stimulated or not with PMA/Iono for 4 hours and the percentage of CD69+ cells 800 and phospho-Stat5 (D) were analyzed by FACS. (D) hsa-miR-155 and human socs1 gene 801 expression were analyzed by qPCR. (E) Human PBLs were transfected with anti-hsa-miR-802 155-5p or anti-miRNA Scramble and hsa-miR-155 expression was analyzed by qPCR. (F) 803 Representative histograms and quantification of CD69 expression, STAT5 phosphorylation 804 and human socs-1 transcription in CD4+ PBLs treated as in (E). (G) Human PBLs were 805 transfected with hsa-pre-miR-155-5p or pre-miRNA-control and hsa-miR-155 expression 806 was analyzed by qPCR. (H) CD69 expression, STAT5 phosphorylation and human socs-1 807 transcription in CD4+ PBLs treated as in (G). Results from miRNA qPCRs are normalized 808 to snoRNA135 expression. All data are mean ± S.D. of at least 3 independent donors from 809 a total of ten donors. Data was analyzed by 1-way ANOVA and Bonferroni’s post-test or 810 by t-test. * P < 0.05, ** P < 0.01, *** P < 0.001. 811 812