bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

1 Fine-tuning of b-catenin in thymic epithelial cells is required for postnatal T cell

2 development

3

4 Sayumi Fujimori*¶#, Izumi Ohigashi*, Hayato Abe**, Makoto M. Taketo†, Yousuke

5 Takahama‡#, Shinji Takada§¶∫#

6

7 *Division of Experimental Immunology, Institute of Advanced Medical Sciences,

8 Tokushima University, Tokushima 770-8503, Japan

9 **Student Laboratory, School of Medicine, Tokushima University, Tokushima 770-

10 8503, Japan

11 †Institute for Advancement of Clinical and Translational Science, Kyoto University

12 Hospital, Kyoto 606-8506, Japan

13 ‡Experimental Immunology Branch, National Cancer Institute, National Institutes of

14 Health, Bethesda, Maryland 20892, USA

15 §Exploratory Research Center on Life and Living Systems, National Institutes of

16 Natural Sciences, Okazaki, Aichi 444-8787, Japan

17 ¶National Institute for Basic Biology, National Institutes of Natural Sciences, Okazaki,

18 Aichi 444-8787, Japan

19 ∫Department of Basic Biology in the School of Life Science, The Graduate University

20 for Advanced Studies (SOKENDAI), Okazaki, Aichi 444-8787, Japan

21 #Correspondence to ,

22 ,

1 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

23 Abstract

24 In the thymus, the thymic epithelium provides a microenvironment essential for the

25 development of functionally competent and self-tolerant T cells. Previous findings

26 showed that modulation of Wnt/b-catenin signaling in thymic epithelial cells (TECs)

27 disrupts embryonic thymus organogenesis. However, the role of b-catenin in TECs for

28 postnatal T cell development remains to be elucidated. Here, we analyzed gain-of-

29 function (GOF) and loss-of-function (LOF) of b-catenin highly specific in TECs. We

30 found that GOF of b-catenin in TECs results in severe thymic dysplasia and T cell

31 deficiency beginning from the embryonic period. By contrast, LOF of b-catenin in

32 TECs reduces the number of cortical TECs and thymocytes modestly and only

33 postnatally. These results indicate that fine-tuning of b-catenin expression within a

34 permissive range is required for TECs to generate an optimal microenvironment to

35 support postnatal T cell development.

36

2 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

37 Introduction

38 The thymus is an organ where lymphocyte progenitors differentiate into functionally

39 competent T cells through interactions primarily with cortical thymic epithelial cells

40 (cTECs) and medullary thymic epithelial cells (mTECs), which construct a three-

41 dimensional epithelial network of the thymus and facilitate the development and

42 selection of developing T cells. To establish a functional thymic microenvironment,

43 thymus organogenesis is tightly regulated by the interaction of thymic epithelial cells

44 (TECs) with neighboring cells, such as mesenchymal cells and hematopoietic cells,

45 through various molecules including morphogens, growth factors, cytokines, and

46 chemokines, during the thymus organogenesis that is initiated around embryonic day 11

47 (E11) in mouse (Blackburn and Manley, 2004; Gordon and Manley, 2011; Takahama et

48 al., 2017).

49 The Wnt/b-catenin signaling pathway has been implicated in thymus development

50 following the discovery of its core components, T-cell factor (TCF)/lymphoid

51 enhancing factor (LEF) family transcription factors, in developing thymocytes

52 (Oosterwegel et al., 1991; Travis et al., 1991). Wnt proteins activate the canonical

53 signaling pathway through interaction with their receptors of the Frizzled family and

54 coreceptors of the LRP5/6 family. This interaction leads to the inhibition of the β-

55 catenin destruction complex, which is composed of adenomatous polyposis coli (APC),

56 axin, casein kinase 1a (CK1a), and glycogen synthase kinase 3b (GSK3b) proteins,

57 thereby resulting in the stabilization of cytosolic β-catenin, its translocation to the

58 nucleus, and enhancement of transcription together with the TCF/LEF family

59 transcription factors (Nusse and Clevers, 2017).

3 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

60 Previous systemic and conditional gene manipulation studies in mouse have

61 demonstrated the importance of Wnt/b-catenin signaling in early thymus development.

62 The systemic deletion of Wnt4, which is abundant in embryonic TECs (Pongracz et al.,

63 2003), reduces the number of TECs as well as thymocytes (Heinonen et al., 2011).

64 Keratin 5 (K5) promoter-driven inhibition of Wnt/b-catenin signaling by the

65 overexpression of Wnt inhibitor Dickkopf1 (DKK1) or by the deficiency of b-catenin

66 reduces the number of TECs including K5+K8+ immature TECs (Osada et al., 2010;

67 Liang et al., 2013). Similarly, a reduction in the number of TECs is observed in

68 perinatal mice in which b-catenin is deleted using Foxn1-Cre, which targets TECs

69 expressing transcription factor Forkhead box N1 (Foxn1) (Swann et al., 2017).

70 However, the Foxn1-Cre-mediated b-catenin deletion causes neonatal lethality due to

71 side effects in the skin epidermis, so that the postnatal effects of b-catenin deficiency in

72 TECs remains unknown. On the other hand, overactivation of Wnt/β-catenin signaling

73 by the deletion of transmembrane protein Kremen 1, which binds to DKK1 and

74 potentiates Wnt signaling inhibition, results in the increase in the number of K5+K8+

75 immature TECs and the disorganization of the TEC network by the decrease in the

76 frequency of mature cTECs and mTECs (Osada et al., 2006). Keratin 14 (K14)

77 promoter-driven loss of APC disorganizes the thymus and reduces its size (Kuraguchi et

78 al., 2006). Overexpression of b-catenin using Foxn1-Cre causes a more severe defect in

79 embryonic thymus development, which is accompanied by the failure of separation of

80 the thymic primordium from the parathyroid primordium (Zuklys et al., 2009; Swann et

81 al., 2017). These results indicate that Wnt/β-catenin signaling is critical for embryonic

82 TEC development. However, the role of Wnt/β-catenin signaling in TECs for thymus

4 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

83 development and function during the postnatal period remains unclear because previous

84 studies relied on mutant mice that exhibit perinatal lethality due to severe effects on

85 various cell types including the skin epidermis.

86 In this study, we assessed the role of β-catenin in TECs by employing β5t-iCre, which

87 was successfully engineered to target TECs in a highly specific manner without side

88 effects in other cells including the skin epithelium (Ohigashi et al., 2013). Thymus-

89 specific proteasome subunit β5t is specifically expressed in cTECs in a Foxn1-

90 dependent manner, and is important for the generation of functionally competent CD8+

91 T cells (Murata et al., 2007; Ohigashi et al., 2021). β5t in postnatal mice is exclusively

92 expressed by cTECs but not mTECs (Ripen et al., 2011; Uddin et al., 2017); however,

93 its transcription is also detectable in embryonic TECs that differentiate into cTECs and

94 mTECs (Ohigashi et al., 2013; Ohigashi et al., 2015). By taking advantage of the highly

95 specific Cre activity only in TECs and not in skin epidermal cells in β5t-iCre mice, here

96 we engineered gain-of-function (GOF) and loss-of-function (LOF) of β-catenin

97 specifically in TECs and addressed the contribution of the Wnt/β-catenin signaling

98 pathway in TECs, focusing on the effects of thymus function on T cell development

99 during the postnatal period.

100

5 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

101 Results

102 Gain-of-function of b-catenin in thymic epithelial cells causes thymic dysplasia

103 To address the potential function of b-catenin in TECs, we generated mice in which

104 GOF of b-catenin was achieved specifically in TECs by intercrossing b5t-iCre mice

105 (Ohigashi et al., 2013) with b-catenin exon 3 floxed mice, in which the conditional

106 deletion of exon 3 led to stabilization of b-catenin and constitutive activation of b-

107 catenin signaling (Harada et al., 1999). Contrary to previous studies demonstrating that

108 the Foxn1-Cre-mediated GOF of b-catenin perturbed the migration of the thymic

109 primordium from the pharyngeal region into the thoracic cavity during embryogenesis

110 (Zuklys et al., 2009; Swann et al., 2017), the b5t-iCre-mediated GOF of b-catenin (b-

111 cat GOF) showed no disruption of the thoracic migration of the embryonic thymus (Fig.

112 1A, B). However, the thymus in these mice showed severe dysplasia by E15.5 (Fig. 1B,

113 C). The b-cat GOF embryos appeared normal in size (Fig. 1B) and exhibited

114 undisturbed limb development (data not shown), indicating that the thymic dysplasia in

115 b-cat GOF mice occurred independent of systemic retardation in embryogenesis.

116 Fluorescence detection of Cre-mediated recombination by further crossing to R26R-

117 tdTomato reporter mice (Madisen et al., 2010) revealed that b5t-iCre-mediated

118 tdTomato-expressing cells were specifically detected in the thymus in the thoracic

119 cavity in both control and b-cat GOF mice at E15.5, further confirming that TECs

120 successfully migrated into the thoracic cavity in b-cat GOF mice (Fig. 1B). Unlike

121 control mice in which tdTomato-labeled cells distributed uniformly throughout the

122 thymus, tdTomato-labeled cells in b-cat GOF mice densely distributed in concentric

123 mid-inner layers, excluding the central core, in the thymus (Fig. 1B). Morphological

6 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

124 alteration of the thymus was apparent in sagittal sections of E15.5 b-cat GOF embryos

125 by hematoxylin and eosin staining and by the detection of keratin 5 (K5) and keratin 8

126 (K8) expression, in which a multi-layered concentric structure of epithelial cells

127 surrounding the central core that mainly consisted of non-epithelial structures sparsely

128 containing erythrocytes was observed (Fig. 1C). b-Catenin was abundant in the

129 concentric layers of the thymus in E15.5 b-cat GOF mice, whereas b5t was undetectable

130 throughout the E15.5 b-cat GOF thymus (Fig. 1C), indicating that GOF of b-catenin in

131 TECs diminishes b5t expression in TECs. Intracellular staining of b-catenin in CD45-

132 EpCAM+ TECs isolated from E15.5 thymus revealed that the intracellular level of b-

133 catenin in TECs from b-cat GOF mice was increased by 70% compared with that from

134 control mice (Fig. 1D). These results indicate that b5t-iCre-mediated GOF of b-catenin

135 specifically in TECs causes severe thymic dysplasia by E15.5 without affecting thoracic

136 migration of the thymic primordium.

137

138 Defective thymus development by gain-of-function of b-catenin in thymic epithelial cells

139 To examine how early the thymic abnormality was detectable in the embryos of b-cat

140 GOF mice, we next performed histological analysis of the thymic primordium at early

141 stages of fetal thymus development before E15.5. The expression of Foxn1, a

142 transcription factor that characterizes the thymic epithelium-specified domain of the

143 third pharyngeal pouch (Nehls et al., 1994; Gordon et al., 2001), was detected in the

144 thymic primordium in both b-cat GOF and control embryos at E11.5, indicating that the

145 initiation of the earliest thymus organogenesis is undisturbed in b-cat GOF mice (Fig.

146 2A). At E12.5, when the expression of b5t becomes detectable in the thymus (Ripen et

7 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

147 al., 2011), the colonization of CD45+ lymphocytes in Foxn1-expressing thymic

148 primordium was observed in both b-cat GOF and control embryos (Fig. 2A), further

149 supporting that early thymus organogenesis is undisturbed in b-cat GOF mice.

150 However, the thymus morphology was markedly altered in b-cat GOF embryos at

151 E13.5, namely, the thymic primordium appeared spherical in b-cat GOF embryos and

152 ovoid in control embryos (Fig. 2A). At this stage in b-cat GOF embryos, Foxn1+ TECs

153 and CD45+ lymphocytes accumulated preferentially in the anterior area of the thymic

154 primordium (Fig. 2A). By E15.5, Foxn1 expression became undetectable and the

155 number of CD45+ lymphocytes was markedly reduced in the dysplastic thymus in b-cat

156 GOF embryos (Fig. 2A).

157 Quantitative RT-PCR analysis of CD45-EpCAM+ TECs isolated from E15.5 thymus

158 established that the expression of Wnt/β-catenin target genes, including Axin2, Dkk1,

159 and Msx1, was upregulated in b-cat GOF TECs (Fig. 2B). By contrast, the expression of

160 genes functionally relevant to TECs, such as Foxn1, Pmsb11 (encoding b5t protein),

161 and Ccl25 (encoding chemokine CCL25 protein that attracts lymphocyte progenitors to

162 the thymic primordium (Liu et al., 2006)), was significantly reduced in b-cat GOF

163 TECs, consistent with the results of immunohistochemical analysis. The initial

164 attraction of lymphoid progenitors to the thymic primordium is cooperatively regulated

165 by CCL25 and CCL21 chemokines, the expression of which is dependent on Foxn1 and

166 Glial Cells Missing Transcription Factor 2 (Gcm2), respectively (Liu et al., 2006).

167 Given that the expression of Ccl21a was not altered in b-cat GOF TECs (data not

168 shown), and given that b5t transcription for b5t-iCre-mediated b-catenin GOF is

169 dependent on Foxn1 (Zuklys et al., 2016; Uddin et al., 2017), the overexpression of b-

8 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

170 catenin in TECs primarily affected Foxn1-dependent Ccl25 expression but not Gcm2-

171 dependent Ccl21a expression. Notably, the expression of involucrin (Ivl) and loricrin

172 (Lor) genes, which are expressed by terminally differentiated keratinocytes (Candi et

173 al., 2005) and terminally differentiated subpopulation of mTECs (Yano et al., 2008;

174 Michel et al., 2017), was upregulated in b-cat GOF TECs but not control TECs at E15.5

175 (Fig. 2B). These results suggest that normal differentiation of TECs is disrupted in b-cat

176 GOF embryos; instead, embryonic TECs aberrantly differentiate into cells resembling

177 terminally differentiated keratinocytes.

178 The number of CD45+ thymocytes in E15.5 thymus was significantly reduced in b-cat

179 GOF mice compared with control mice (Fig. 2C). Flow cytometric analysis of CD45+

180 thymocytes indicated that the majority of E15.5 thymocytes in b-cat GOF mice and

181 control mice were equivalently CD4-CD8- double negative (DN) (data not shown).

182 However, the DN thymocytes in b-cat GOF embryos were predominantly arrested at

183 CD44+CD25- DN1 stage (Fig. 2D). Indeed, the numbers of downstream thymocytes at

184 DN2 (CD44+CD25+), DN3 (CD44-CD25+), and DN4 (CD44-CD25-) stages were

185 dramatically reduced in b-cat GOF embryos compared with control embryos, whereas

186 the number of DN1 thymocytes was comparable between b-cat GOF and control

187 embryos (Fig. 2E). These results indicate that embryonic thymus development is

188 severely disturbed by b5t-iCre-mediated GOF of b-catenin specifically in TECs, which

189 in turn results in early arrest of thymocyte development in embryonic thymus.

190

191 Gain-of-function of b-catenin in thymic epithelial cells causes postnatal loss of T cells

192 in secondary lymphoid organs

9 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

193 b-Cat GOF mice grew to adulthood with no gross abnormality in appearance and body

194 weight (Fig. 3A and 3B), whereas the thymic dysplasia persisted throughout the

195 postnatal periods in postnatal b-cat GOF mice (Fig. 3C). Flow cytometric analysis of

196 cells from postnatal lymphoid organs showed that the frequency and the number of

197 ab TCR-expressing T cells in the spleen and the inguinal lymph node (iLN) were

198 significantly reduced in b-cat GOF adult mice (Fig. 4A and 4B), indicating that ab T

199 cells are essentially lost in b-cat GOF mice. We also detected the reduction of gd TCR-

200 expressing T cells in b-cat GOF mice. Vg5+ gd T cells, which are generated in the

201 embryonic thymus and localize to the epidermis to form dendritic epidermal T cells

202 (DETCs) in adult mice (Havran and Allison, 1988), were severely reduced in frequency

203 in the embryonic thymus and the postnatal skin of b-cat GOF mice (Fig. 4C and 4D).

204 Vg4+ and Vg1+ gd T cells, which are generated in the thymus during the perinatal

205 period subsequently to the embryonic generation of Vg5+ gd T cells (Pereira et al.,

206 1995), were similarly reduced in frequency in the iLN and the spleen of b-cat GOF

207 adult mice (Fig. 4E, data not shown). These results indicate that thymus-dependent gd T

208 cells, including Vg5+, Vg4+, and Vg1+ gd T cells, are essentially lost in b-cat GOF

209 mice.

210 Taken together, our results substantiate that GOF of b-catenin in TECs causes severe

211 thymic dysplasia that results in postnatal loss of thymus-derived ab and gd T cells in

212 secondary lymphoid organs.

213

214 Loss-of-function of b-catenin in thymic epithelial cells results in no apparent change in

215 thymus development during embryonic period

10 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

216 To further evaluate the role of b-catenin in TECs, LOF analysis of b-catenin was

217 performed using b-catenin floxed allele targeted by the b5t-iCre-mediated

218 recombination (b-cat LOF). In contrast to b-cat GOF mice, morphological abnormality

219 in the thymus was not apparent in b-cat LOF embryos at E15.5 (Fig. 5A).

220 Immunohistofluorescence analysis confirmed the reduction of b-catenin in the thymus

221 of b-cat LOF mice at E15.5, whereas the expression of Foxn1, K5, and K8 in E15.5

222 thymus was comparable between b-cat LOF mice and littermate control mice (Fig. 5B).

223 Flow cytometric analysis revealed that intracellular b-catenin level was reduced by 77%

224 in CD45-EpCAM+ TECs of b-cat LOF mice at E15.5 (Fig. 5C). However, the number

225 of total thymic cells and the frequency and the number of total TECs and UEA1-Ly51+

226 cTECs were comparable between b-cat LOF and control mice (Fig. 5D, 5E), indicating

227 that b5t-iCre-mediated loss of b-catenin in TECs did not alter embryonic thymus

228 development by E15.5.

229

230 b-Catenin in thymic epithelial cells is required for optimal size of postnatal thymus

231 b-Cat LOF mice grew to adulthood with normal hair coat (Fig. 6A). However, thymus

232 size in b-cat LOF mice was noticeably reduced in comparison to that in control mice

233 during the postnatal period (Fig. 6B), with concomitant reductions in thymus weight but

234 not body weight (Fig. 6C). Flow cytometric analysis revealed that b-catenin expression

235 was effectively reduced by 96% in UEA1-Ly51+ cTECs and by 94% in UEA1+Ly51-

236 mTECs in b-cat LOF mice at 2 wk old (Fig. 7A). Accordingly, quantitative RT-PCR

11 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

237 analysis showed marked reduction of b-catenin mRNA (Ctnnb1) expression in cTECs

238 and mTECs of b-cat LOF mice (Fig. 7B).

239 In the thymus of b-cat LOF mice, the frequency and the number of cTECs were

240 decreased compared with control mice (Fig. 7C and 7D). Axin2 expression in cTECs

241 was significantly reduced in b-cat LOF mice, indicating that Wnt/b-catenin signals were

242 efficiently impaired in cTECs from b-cat LOF mice. However, the expression of Foxn1,

243 IL7, and Cxcl12, which encode functionally important molecules in cTECs, was not

244 altered in cTECs from b-cat LOF mice (Fig. 7B). On the other hand, despite that the

245 expression of b-catenin in mTECs was markedly reduced in b-cat LOF mice, the

246 cellularity of mTECs was not altered in b-cat LOF mice (Fig. 7C and 7D), and the

247 expression of Axin2 was not reduced in mTECs from b-cat LOF mice (Fig. 7B). The

248 expression of functionally relevant molecules in mTECs, such as Aire, Tnfrsf11a, and

249 Ccl21a, was comparable between b-cat LOF and control mTECs (Fig. 7B). In

250 agreement with the results of quantitative RT-PCR analysis, there were no apparent

251 abnormalities in the corticomedullary compartmentalization of the thymus, as evidenced

252 by the detection of b5t+ cTECs in the cortex and CCL21+ mTECs and AIRE+ mTECs

253 in the medulla, in b-cat LOF mice (Fig. 7E). These results indicate that b-catenin in

254 TECs is required for optimizing postnatal thymus size, mainly affecting the number of

255 cTECs, but is dispensable for the development of the normal number of mTECs.

256

257 b-Catenin in thymic epithelial cells affects postnatal thymocyte number

258 We finally examined thymocytes in postnatal b-cat LOF mice. The number of

259 thymocytes was reduced in b-cat LOF mice compared with control at 2 wk, 4 wk, and 8

12 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

260 wk of age (Fig. 8A). The frequency of DN, DP, CD4+CD8-, and CD4-CD8+

261 thymocytes, as well as the frequency of TCRb-d+ thymocytes, was not altered in b-cat

262 LOF mice at 8 wk (Fig. 8B). As a result of the reduction in the number of total

263 thymocytes, the number of thymocyte subsets defined by CD4 and CD8, as well as the

264 number of TCRb-d+ thymocytes, was reduced in b-cat LOF mice, although the

265 reduction in the number of CD4+CD8- thymocytes in these mice was not statistically

266 significant (p = 0.069) (Fig. 8C). On the other hand, in contrast to the significant

267 reduction in thymocyte numbers, the number of ab T cells and gd T cells was not

268 significantly altered in the secondary lymphoid organs, including the spleen and the iLN

269 (Fig. 8D and 8E, data not shown).

270 Our results indicate that b-catenin in TECs is not essential for the generation of

271 functional TECs that support T cell development. However, the loss of b-catenin in

272 TECs results in the reduction in the number of cTECs, which leads to the reduction in

273 the number of thymocytes during the postnatal period.

274

13 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

275 Discussion

276 The important role played by Wnt/b-catenin signaling in thymus organogenesis has

277 been well appreciated. However, the role of Wnt/b-catenin signaling in TECs with

278 regard to postnatal immune system development remains unclear due to the difficulty of

279 performing an in vivo analysis without generating extrathymic side effects. Most

280 typically, the conventional “TEC-specific” Foxn1-Cre-mediated genetic manipulation

281 of Wnt/b-catenin signaling molecules causes neonatal lethality of the animals due to

282 side effects in the skin epidermis. In the present study, we employed the recently

283 devised b5t-Cre, which enables highly efficient and specific genetic manipulation in

284 TECs without generating side effects in the skin epidermis or other cells in the body. By

285 analyzing b5t-Cre-mediated deficiency in b-catenin highly specific in TECs, we found

286 that the TEC-specific loss of β-catenin causes postnatal reduction in the number of

287 cTECs, which leads to the reduction in the number of thymocytes. Nonetheless, the

288 TEC-specific loss of β-catenin did not cause an arrest in the development of the thymus,

289 including the corticomedullary architecture, and the subsequent generation of T cells,

290 indicating that β-catenin in TECs is dispensable for the development of cTECs and

291 mTECs as well as of postnatal T cells. In contrast, our results also showed that b5t-Cre-

292 mediated constitutive activation of b-catenin in TECs causes dysplasia in thymus

293 organogenesis and loss of thymic T cell development. These results suggest that fine-

294 tuning of β-catenin-mediated signaling activity in TECs is required for optimal

295 development and maintenance of functional thymus. Thus, b-catenin-mediated

296 signaling in TECs needs to be maintained at an intermediate level to prevent thymic

297 dysplasia induced by its excessive signals, whereas no or extremely low β-catenin

14 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

298 signals cause hypoplasia of the postnatal thymus. The need for fine-tuning of b-catenin

299 during the development of the central nervous system and other organs, such as bone,

300 heart, intestine, liver, lung, mammary gland, pancreas, reproductive organs, skin, and

301 tooth, was reported (Grigoryan et al., 2008). Our LOF and GOF experiments in TECs

302 established that the thymus is included in tissues that require fine-tuning of b-catenin

303 during development, and revealed that the fine-tuning of b-catenin in TECs is required

304 for supporting postnatal T cell development in the thymus.

305 Previous conditional gene targeting in TECs was performed using Cre-deleter mouse

306 lines with any one of the promoter sequences for Foxn1, K5, and K14. However,

307 extrathymic side effects were induced in epithelial cells other than TECs in the Cre-

308 deleter mouse lines. For example, Foxn1-Cre-mediated genetic alteration was detectable

309 in the skin in addition to TECs (Zuklys et al., 2009), and mice rendered β-catenin-

310 deficient using Foxn1-Cre died within a few days of birth due to defects in skin

311 development (Swann et al., 2017). In contrast, the b5t-locus knock-in Cre-deleter mice

312 used in the present study enabled highly efficient and specific genetic manipulation in

313 TECs without any side effects in other organs including the skin (Ohigashi et al., 2013).

314 Indeed, our b-cat LOF and GOF mice using the TEC-specific b5t-iCre grew postnatally

315 without perinatal lethality or skin defects. Interesting, both of our b-cat LOF and GOF

316 mice exhibited thymic phenotypes different from previously observed phenotypes (Fig.

317 9). The b5t-iCre-mediated GOF of b-catenin caused dysfunction of the embryonic

318 thymus, in line with previous b-cat GOF studies using Foxn1-Cre (Zuklys et al., 2009;

319 Swann et al., 2017). However, unlike those previous studies, defective migration of the

320 thymus into the thoracic cavity during the embryogenesis was not observed in our b5t-

15 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

321 iCre-mediated b-cat GOF mice. Therefore, our study clarified that the b-cat GOF-

322 mediated thymic dysfunction is not due to impaired migration of the thymic primordium

323 from the pharyngeal region to the chest cavity. Instead, the difference in migration

324 phenotype may be due to the different onset of Cre activity in TECs between different

325 Cre lines. The Foxn1-mediated direct transcriptional regulation of b5t and a slightly

326 later expression of b5t than Foxn1 (Ripen et al., 2011; Uddin et al., 2017) may result in

327 Foxn1-Cre-mediated, but not b5t-Cre-mediated, b-cat GOF interference with thymic

328 migration toward the thoracic cavity.

329 More prominently, unlike Foxn1-Cre-mediated β-cat LOF mice, b5t-Cre-mediated b-

330 cat LOF mice did not die perinatally but grew to adulthood, which enabled the analysis

331 of postnatal T cell development in TEC-specific β-catenin-deficient thymus. Our results

332 demonstrated that the loss of b-catenin in TECs impacts neither thymic architecture nor

333 the expression of functionally relevant molecules in cTECs and mTECs. Rather, we

334 found that the b5t-iCre-mediated LOF of b-catenin reduces the number of postnatal

335 cTECs. We further found that the number of postnatal thymocytes in the b-cat LOF

336 thymus is reduced by half compared with that in control thymus, although the T cell

337 development is not disturbed. Thus, our analysis of b5t-Cre-mediated b-cat LOF mice

338 revealed that β-catenin signaling in TECs is dispensable for TEC development and T

339 cell development, but is required for the provision of cTECs, which support postnatal T

340 cell generation.

341 Our results indicated that b5t-iCre-mediated b-catenin-deficient mice exhibit postnatal

342 reduction in the number of cTECs but not mTECs and that their expression of Axin2 is

343 reduced only in cTECs but not in mTECs, implying the differential responsiveness to b-

16 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

344 catenin signaling between cTECs and mTECs. It is further interesting to note that the

345 reduced number of cTECs but not mTECs is associated with an equivalent reduction in

346 the number of immature thymocytes in the thymic cortex and the number of mature

347 thymocytes in the thymic medulla, suggesting that the number of mature thymocytes

348 generated in the thymus is correlated with the number of cTECs rather than the number

349 of mTECs.

350 Our results showed that GOF of b-catenin in TECs results in the decrease in the

351 expression of TEC-specific molecules and the increase in the expression of terminal

352 differentiated keratinocytes, in agreement with previous results using Foxn1-Cre-

353 mediated b-cat GOF mice (Zuklys et al., 2009). It is possible that TECs retain the

354 potential to transdifferentiate into epidermal keratinocytes in the presence of an

355 excessive amount of b-catenin, as a similar transdifferentiation of epithelial cells upon

356 b-catenin overexpression was reported in other tissues, such as mammary gland and

357 prostate (Miyoshi et al., 2002; Bierie et al., 2003). It is noted that the terminally

358 differentiated mTEC subpopulations, including post-Aire mTECs and Hassall’s

359 corpuscles, express involucrin (Yano et al., 2008; Nishikawa et al., 2010; White et al.,

360 2010) and loricrin (Michel et al., 2017). Therefore, it is further possible that the

361 alteration of b-catenin-mediated signals in mature mTECs promotes the terminal

362 differentiation of the mTECs or the deviation to give rise to Hassall’s corpuscles.

363 Cellular interactions between TECs and neighboring cells, such as mesenchymal cells

364 (Shinohara and Honjo, 1997; Jenkinson et al., 2003; Jenkinson et al., 2007; Itoi et al.,

365 2007), endothelial cells (Bryson et al., 2013; Wertheimer et al., 2018), and thymocytes

366 (Hollander et al., 1995; Klug et al., 2002), are important for thymus development. Wnts

367 and Wnt signaling molecules are expressed in TECs and neighboring cells in various

17 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

368 amounts (Balciunaite et al., 2002; Pongracz et al., 2003; Osada et al., 2006; Heinonen et

369 al., 2011; Ki et al., 2014; Brunk et al., 2015, Nitta et al., 2020) and at various timings

370 (Osada et al., 2006; Kvell et al., 2010, Varecza et al., 2011; Griffith et al., 2012; Ki et

371 al., 2014). The presence of a variety of Wnts and Wnt signaling molecules in

372 developing TECs may trigger the fine-tuning of β-catenin signaling activity to establish

373 the development of diverse TEC subpopulations.

374 We think that the β-catenin-mediated regulation of TEC development is dependent on

375 TCF/LEF-mediated canonical Wnt signaling but not on the adherent function of β-

376 catenin, which plays a role in the interaction between E-cadherin and a-catenin at

377 adherens junctions (Kemler, 1993). It was previously shown that E-cadherin deficiency

378 induced by Foxn1-Cre did not impact the number of TECs and was less effective than

379 b-catenin deficiency in newborn mice (Swann et al., 2017). In addition, Foxn1-Cre-

380 mediated E-cadherin deficient adult mice exhibited no defect in the thymus (Swann et

381 al., 2017), in contrast to b5t-iCre-mediated b-catenin deficient mice, further suggesting

382 that β-catenin-dependent signaling in TECs is likely independent of the adherent

383 function of β-catenin.

384 The thymic microenvironment provided by TECs is essential for the development of

385 functionally competent and self-tolerant T cells. Our in vivo genetic study of b-catenin

386 specifically manipulated in TECs demonstrates that fine-tuning of b-catenin-mediated

387 signaling in TECs is essential to maintain TEC function integrity for T cell production

388 during embryogenesis and the postnatal period. Future studies of the function of Wnts

389 and Wnt signaling components in TECs and neighboring cells should provide further

390 insight into the mechanism for fine-tuning β-catenin signaling in TECs to support T cell

391 development in the thymus.

18 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

392 Materials and Methods

393 Mice

394 C57BL/6 mice were obtained from SLC Japan. b5t-iCre (Ohigashi et al., 2013), b-

395 catenin floxed (b-catfl) (Brault et al., 2001), b-catenin exon 3 floxed (b-catex3fl)

396 (Harada et al., 1999), and Rosa26-CAG-loxP-stop-loxP-tdTomato (R26R-tdTomato)

397 (Madisen et al., 2010) used in this study have been described previously and genotyped

398 accordingly. To generate conditional targeting mice, we used male mice harboring b5t-

399 iCre allele for the mating to avoid potential germline recombination rarely observed in

400 the offspring of female mice harboring b5t-iCre allele. TEC-specific b-catenin gain-of-

401 function (b-cat GOF) (b5tiCre/+; b-catex3fl/+) mice were generated by intercrossing b-

402 catex3fl/ex3fl females with b5tiCre/+ male mice. For the lineage tracing of cells,

403 C57BL/6 females and b-catex3fl/ex3fl females were crossed with

404 b5tiCre/+; R26RtdTomato/tdTomato males to produce b5tiCre/+; R26RtdTomato/+

405 (control) and b5tiCre/+; b-catex3fl/+; R26RtdTomato/+ (b-cat GOF) mice, respectively.

406 TEC-specific b-catenin loss-of-function (b-cat LOF) (b5tiCre/+; b-catfl/fl) mice were

407 generated by intercrossing b-catfl/fl females with b5tiCre/+; b-catfl/+males.

408 Embryos were staged on the basis of standard morphological criteria; plug date was

409 considered to be E0.5. Postnatal mice were analyzed at indicated ages up to 11 weeks

410 old in an age-matched manner. Mice were housed in a 12-hour light-dark cycle in

411 climate-controlled, pathogen-free barrier facilities. All mouse experiments were

412 performed with consent from the Animal Experimentation Committee of the University

413 of Tokushima (T2019-62).

19 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

414

415 Immunohistochemistry

416 Embryos and thymus tissues were fixed overnight in 4% PFA/PBS, cryoprotected in

417 30% sucrose, embedded in optimal cutting temperature compound (Sakura Finetek),

418 and sectioned at 10 µm thickness. Immunohistochemistry was performed on the

419 cryosections using the following primary antibodies: rabbit anti-Foxn1 antiserum (Itoi

420 et al., 2006, 1:100), rabbit anti-b5t antibody (Murata et al., 2007, 1:200), rat anti-

421 CCL21/6Ckine antibody (R&D systems, 1:10), eFluor 660 conjugated anti-Aire

422 antibody (e-Bioscience, 1:20), and biotinylated anti-CD45 antibody (BioLegend,

423 1:100). Immunohistochemistry using monoclonal mouse anti-b-catenin (BD

424 Transduction Laboratories, 1:200), chicken anti-Keratin 5 antibody (BioLegend, 1:500),

425 and mouse anti-Keratin 8 antibody (BioLegend, 1:500) was performed after heat-

426 induced antigen retrieval. For detection of antibodies, AlexaFluor-conjugated secondary

427 antibodies (Invitrogen) were used at 1:500 dilution as necessary. Nuclear

428 counterstaining was performed using TO-PRO3 (Thermo Fisher Scientific, 1:1000).

429 Fluorescent images were obtained and analyzed with a TCS SP8 (Leica) confocal laser

430 scanning microscope.

431

432 Flow cytometry and cell sorting

433 For flow cytometric analysis of TECs, minced thymuses from fetuses and postnatal

434 mice were digested with 0.5 and 1 unit/ml Liberase TM (Roche) in the presence of

435 0.01% DNase I (Roche), respectively. Single-cell suspensions were stained for the

436 expression of EpCAM (BioLegend, clone G8.8), CD45 (BioLegend, clone 30-F11), and

437 Ly51 (BioLegend, clone 6C3) and for the reactivity with UEA-1 (Vector Laboratories).

20 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

438 For the intracellular staining of b-catenin in TECs, surface-stained cells were fixed in

439 2% paraformaldehyde and permeabilized with 0.05% saponin, and stained with

440 monoclonal mouse anti-b-catenin against aa. 571-781 (BD Transduction Laboratories,

441 Clone 14/Beta-Catenin, 1:100) or mouse IgG1 isotype control (BD Pharmingnen, clone

442 MOPC-21, 1:200), followed by Alexa Fluor Plus 488-conjugated anti mouse IgG

443 antibody (Invitrogen, 1:800).

444 For the isolation of TECs, CD45- cells were enriched with magnetic-bead-conjugated

445 anti-CD45 antibody (Miltenyi Biotec) before multicolor staining for flow cytometric

446 cell sorting.

447 For flow cytometric analysis of thymocytes, spleen cells, and lymph node cells, cells

448 were multicolor stained for CD4 (eBioscience, clone RM4-5), CD8a (Invitrogen, clone

449 53-6.7), CD25 (BioLegend, clone PC61), CD44 (eBioscience, clone IM7), TCRb

450 (BioLegend, clone H57-597), TCRd (BioLegend, clone GL3), Vg5 (BD Pharmingen,

451 clone 536), Vg4 (BD Pharmingen, clone UC3-10A6), and Vg1 (BioLegend, clone 2.11).

452 For the analysis of DETCs, epidermal cells were prepared as described previously (Liu

453 et al., 2006). Briefly, dorsal skin of 8-week-old mice was placed dermal side down in

454 0.25% Trypsin-EDTA in PBS at 37°C for 1 h and epidermal sheet was peeled off from

455 dermis. Minced epidermal sheets were mechanically dissociated to make a single-cell

456 suspension in 1xPBS containing 2% FCS. Cells were filtered and cell surface staining

457 was performed using antibodies specific for CD3e (eBioscience, clone145-2C11) and

458 Vg5 (BD Pharmingen, clone 536).

459

460 Quantitative real-time PCR analysis

21 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

461 Total RNA was isolated using an RNeasy Plus Micro Kit (Qiagen) according to the

462 manufacturer’s instructions. Total RNA was reverse-transcribed (RT) with PrimeScript

463 Reverse Transcriptase (TaKaRa) to produce first-strand cDNA. Quantitative real-time

464 polymerase chain reaction (PCR) was performed using TB Green Premix Taq II

465 (TaKaRa) and a StepOnePlus Real-Time PCR System (Applied Biosystems). Values

466 were calculated by using the comparative C(t) method and normalized to mouse Gapdh

467 expression. Most of the primer sets amplified products across exon/intron boundaries

468 and the PCR products were confirmed by gel electrophoresis and melting curves.

469 Sequences are available by request.

470

471 Statistical analysis

472 Statistical analysis was performed using GraphPad Prism 7 software. Statistical

473 significance was evaluated using two-tailed unpaired Student’s t-test with Welch’s

474 correction for unequal variances. All values are expressed as means and SEMs, unless

475 otherwise specified. The n numbers are indicated in figure legends.

476

477 Acknowledgements

478 We thank Takashi Amagai and Manami Itoi for anti-Foxn1 antibody, Shigeo Murata

479 for anti-b5t antibody, and Hitomi Kyuma and Yukiko Ito for technical support. We also

480 thank all members of Takada Laboratory and Takahama Laboratory for helpful

481 discussion.

482

22 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

483 References

484 Balciunaite, G., Keller, M. P., Balciunaite, E., Piali, L., Zuklys, S., Mathieu, Y. D., Gill,

485 J., Boyd, R., Sussman, D. J., & Hollander, G. A. (2002). Wnt glycoproteins regulate the

486 expression of FoxN1, the gene defective in nude mice. Nat Immunol, 3(11), 1102-1108.

487 doi:10.1038/ni850

488

489 Bierie, B., Nozawa, M., Renou, J. P., Shillingford, J. M., Morgan, F., Oka, T., Taketo,

490 M. M., Cardiff, R. D., Miyoshi, K., Wagner, K. U., Robinson, G. W., & Hennighausen,

491 L. (2003). Activation of beta-catenin in prostate epithelium induces hyperplasias and

492 squamous transdifferentiation. Oncogene, 22(25), 3875-3887.

493 doi:10.1038/sj.onc.1206426

494

495 Blackburn, C. C., & Manley, N. R. (2004). Developing a new paradigm for thymus

496 organogenesis. Nat Rev Immunol, 4(4), 278-289. doi:10.1038/nri1331

497

498 Brault, V., Moore, R., Kutsch, S., Ishibashi, M., Rowitch, D. H., McMahon, A. P.,

499 Sommer, L., Boussadia, O., & Kemler, R. (2001). Inactivation of the beta-catenin gene

500 by Wnt1-Cre-mediated deletion results in dramatic brain malformation and failure of

501 craniofacial development. Development, 128(8), 1253-1264.

502

503 Brunk, F., Augustin, I., Meister, M., Boutros, M., & Kyewski, B. (2015). Thymic

504 epithelial cells are a nonredundant source of Wnt ligands for thymus development. J

505 Immunol, 195(11), 5261-5271. doi:10.4049/jimmunol.1501265

506

23 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

507 Bryson, J. L., Griffith, A. V., Hughes, B., 3rd, Saito, F., Takahama, Y., Richie, E. R., &

508 Manley, N. R. (2013). Cell-autonomous defects in thymic epithelial cells disrupt

509 endothelial-perivascular cell interactions in the mouse thymus. PLoS One, 8(6), e65196.

510 doi:10.1371/journal.pone.0065196

511

512 Candi, E., Schmidt, R., & Melino, G. (2005). The cornified envelope: a model of cell

513 death in the skin. Nat Rev Mol Cell Biol, 6(4), 328-340. doi:10.1038/nrm1619

514

515 Gordon, J., Bennett, A. R., Blackburn, C. C., & Manley, N. R. (2001). Gcm2 and Foxn1

516 mark early parathyroid- and thymus-specific domains in the developing third

517 pharyngeal pouch. Mech Dev, 103(1-2), 141-143. doi:10.1016/s0925-4773(01)00333-1

518

519 Gordon, J., & Manley, N. R. (2011). Mechanisms of thymus organogenesis and

520 morphogenesis. Development, 138(18), 3865-3878. doi:10.1242/dev.059998

521

522 Griffith, A. V., Fallahi, M., Venables, T., & Petrie, H. T. (2012). Persistent degenerative

523 changes in thymic organ function revealed by an inducible model of organ regrowth.

524 Aging Cell, 11(1), 169-177. doi:10.1111/j.1474-9726.2011.00773.x

525

526 Grigoryan, T., Wend, P., Klaus, A., & Birchmeier, W. (2008). Deciphering the function

527 of canonical Wnt signals in development and disease: conditional loss- and gain-of-

528 function mutations of beta-catenin in mice. Genes Dev, 22(17), 2308-2341.

529 doi:10.1101/gad.1686208

530

24 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

531 Harada, N., Tamai, Y., Ishikawa, T., Sauer, B., Takaku, K., Oshima, M., & Taketo, M.

532 M. (1999). Intestinal polyposis in mice with a dominant stable mutation of the beta-

533 catenin gene. EMBO J, 18(21), 5931-5942. doi:10.1093/emboj/18.21.5931

534

535 Havran, W. L., & Allison, J. P. (1988). Developmentally ordered appearance of

536 thymocytes expressing different T-cell antigen receptors. Nature, 335(6189), 443-445.

537 doi:10.1038/335443a0

538

539 Heinonen, K. M., Vanegas, J. R., Brochu, S., Shan, J., Vainio, S. J., & Perreault, C.

540 (2011). Wnt4 regulates thymic cellularity through the expansion of thymic epithelial

541 cells and early thymic progenitors. Blood, 118(19), 5163-5173. doi:10.1182/blood-

542 2011-04-350553

543

544 Hollander, G. A., Wang, B., Nichogiannopoulou, A., Platenburg, P. P., van Ewijk, W.,

545 Burakoff, S. J., Gutierrez-Ramos, J. C., & Terhorst, C. (1995). Developmental control

546 point in induction of thymic cortex regulated by a subpopulation of prothymocytes.

547 Nature, 373(6512), 350-353. doi:10.1038/373350a0

548

549 Itoi, M., Tsukamoto, N., & Amagai, T. (2006). Expression of Dll4 and CCL25 in

550 Foxn1-negative epithelial cells in the post-natal thymus. Int Immunol, 19(2), 127-132.

551 doi:10.1093/intimm/dxl129

552

25 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

553 Itoi, M., Tsukamoto, N., Yoshida, H., & Amagai, T. (2007). Mesenchymal cells are

554 required for functional development of thymic epithelial cells. Int Immunol, 19(8), 953-

555 964. doi:10.1093/intimm/dxm060

556

557 Jenkinson, W. E., Jenkinson, E. J., & Anderson, G. (2003). Differential requirement for

558 mesenchyme in the proliferation and maturation of thymic epithelial progenitors. J Exp

559 Med, 198(2), 325-332. doi:10.1084/jem.20022135

560

561 Jenkinson, W. E., Rossi, S. W., Parnell, S. M., Jenkinson, E. J., & Anderson, G. (2007).

562 PDGFRalpha-expressing mesenchyme regulates thymus growth and the availability of

563 intrathymic niches. Blood, 109(3), 954-960. doi:10.1182/blood-2006-05-023143

564

565 Kemler, R. (1993). From cadherins to catenins: cytoplasmic protein interactions and

566 regulation of cell adhesion. Trends Genet, 9(9), 317-321. doi:10.1016/0168-

567 9525(93)90250-l

568

569 Ki, S., Park, D., Selden, H. J., Seita, J., Chung, H., Kim, J., Iyer, V. R., & Ehrlich, L. I.

570 R. (2014). Global transcriptional profiling reveals distinct functions of thymic stromal

571 subsets and age-related changes during thymic involution. Cell Rep, 9(1), 402-415.

572 doi:10.1016/j.celrep.2014.08.070

573

574 Klug, D. B., Carter, C., Gimenez-Conti, I. B., & Richie, E. R. (2002). Cutting edge:

575 thymocyte-independent and thymocyte-dependent phases of epithelial patterning in the

576 fetal thymus. J Immunol, 169(6), 2842-2845. doi:10.4049/jimmunol.169.6.2842

26 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

577

578 Kuraguchi, M., Wang, X. P., Bronson, R. T., Rothenberg, R., Ohene-Baah, N. Y., Lund,

579 J. J., Kucherlapati, M., Maas, R. L., & Kucherlapati, R. (2006). Adenomatous polyposis

580 coli (APC) is required for normal development of skin and thymus. PLoS Genet, 2(9),

581 e146. doi:10.1371/journal.pgen.0020146

582

583 Kvell, K., Varecza, Z., Bartis, D., Hesse, S., Parnell, S., Anderson, G., Jenkinson, E. J.,

584 & Pongracz, J. E. (2010). Wnt4 and LAP2alpha as pacemakers of thymic epithelial

585 senescence. PLoS One, 5(5), e10701. doi:10.1371/journal.pone.0010701

586

587 Liang, C. C., You, L. R., Yen, J. J., Liao, N. S., Yang-Yen, H. F., & Chen, C. M.

588 (2013). Thymic epithelial beta-catenin is required for adult thymic homeostasis and

589 function. Immunol Cell Biol, 91(8), 511-523. doi:10.1038/icb.2013.34

590

591 Liu, C., Saito, F., Liu, Z., Lei, Y., Uehara, S., Love, P., Lipp, M., Kondo, S., Manley,

592 N., & Takahama, Y. (2006). Coordination between CCR7- and CCR9-mediated

593 chemokine signals in prevascular fetal thymus colonization. Blood, 108(8), 2531-2539.

594 doi:10.1182/blood-2006-05-024190

595

596 Madisen, L., Zwingman, T. A., Sunkin, S. M., Oh, S. W., Zariwala, H. A., Gu, H., Ng,

597 L. L., Palmiter, R. D., Hawrylycz, M. J., Jones, A. R., Lein, E. S., & Zeng, H. (2010). A

598 robust and high-throughput Cre reporting and characterization system for the whole

599 mouse brain. Nat Neurosci, 13(1), 133-140. doi:10.1038/nn.2467

600

27 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

601 Michel, C., Miller, C. N., Kuchler, R., Brors, B., Anderson, M. S., Kyewski, B., &

602 Pinto, S. (2017). Revisiting the road map of medullary thymic epithelial cell

603 differentiation. J Immunol, 199(10), 3488-3503. doi:10.4049/jimmunol.1700203

604

605 Miyoshi, K., Shillingford, J. M., Le Provost, F., Gounari, F., Bronson, R., von Boehmer,

606 H., Taketo, M. M., Cardiff, R. D., Hennighausen, L., & Khazaie, K. (2002). Activation

607 of beta -catenin signaling in differentiated mammary secretory cells induces

608 transdifferentiation into epidermis and squamous metaplasias. Proc Natl Acad Sci U S

609 A, 99(1), 219-224. doi:10.1073/pnas.012414099

610

611 Murata, S., Sasaki, K., Kishimoto, T., Niwa, S., Hayashi, H., Takahama, Y., & Tanaka,

612 K. (2007). Regulation of CD8+ T cell development by thymus-specific proteasomes.

613 Science, 316(5829), 1349-1353. doi:10.1126/science.1141915

614

615 Nehls, M., Pfeifer, D., Schorpp, M., Hedrich, H., & Boehm, T. (1994). New member of

616 the winged-helix protein family disrupted in mouse and rat nude mutations. Nature,

617 372(6501), 103-107. doi:10.1038/372103a0

618

619 Nishikawa, Y., Hirota, F., Yano, M., Kitajima, H., Miyazaki, J., Kawamoto, H., Mouri,

620 Y., & Matsumoto, M. (2010). Biphasic Aire expression in early embryos and in

621 medullary thymic epithelial cells before end-stage terminal differentiation. J Exp Med,

622 207(5), 963-971. doi:10.1084/jem.20092144

623

28 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

624 Nitta, T., Tsutsumi, M., Nitta, S., Muro, R., Suzuki, E. C., Nakano, K., Tomofuji, Y.,

625 Sawa, S., Okamura, T., Penninger, J. M., & Takayanagi, H. (2020). Fibroblasts as a

626 source of self-antigens for central immune tolerance. Nat Immunol, 21(10), 1172-1180.

627 doi:10.1038/s41590-020-0756-8

628

629 Nusse, R., & Clevers, H. (2017). Wnt/beta-catenin signaling, disease, and emerging

630 therapeutic modalities. Cell, 169(6), 985-999. doi:10.1016/j.cell.2017.05.016

631

632 Ohigashi, I., Frantzeskakis, M., Jacques, A., Fujimori, S., Ushio, A., Yamashita, F.,

633 Ishimaru, N., Yin, D., Cam, M., Kelly, M. C., Awasthi, P., Takada, K., & Takahama, Y.

634 (2021). The thymoproteasome hardwires the TCR repertoire of CD8+ T cells in the

635 cortex independent of negative selection. J Exp Med, 218(4).

636 doi:10.1084/jem.20201904

637

638 Ohigashi, I., Zuklys, S., Sakata, M., Mayer, C. E., Hamazaki, Y., Minato, N., Hollander,

639 G. A., & Takahama, Y. (2015). Adult thymic medullary epithelium is maintained and

640 regenerated by lineage-restricted cells rather than bipotent progenitors. Cell Rep, 13(7),

641 1432-1443. doi:10.1016/j.celrep.2015.10.012

642

643 Ohigashi, I., Zuklys, S., Sakata, M., Mayer, C. E., Zhanybekova, S., Murata, S., Tanaka,

644 K., Hollander, G. A., & Takahama, Y. (2013). Aire-expressing thymic medullary

645 epithelial cells originate from beta5t-expressing progenitor cells. Proc Natl Acad Sci U

646 S A, 110(24), 9885-9890. doi:10.1073/pnas.1301799110

647

29 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

648 Oosterwegel, M., van de Wetering, M., Dooijes, D., Klomp, L., Winoto, A.,

649 Georgopoulos, K., Meijlink, F., & Clevers, H. (1991). Cloning of murine TCF-1, a T

650 cell-specific transcription factor interacting with functional motifs in the CD3-epsilon

651 and T cell receptor alpha enhancers. J Exp Med, 173(5), 1133-1142.

652 doi:10.1084/jem.173.5.1133

653

654 Osada, M., Ito, E., Fermin, H. A., Vazquez-Cintron, E., Venkatesh, T., Friedel, R. H., &

655 Pezzano, M. (2006). The Wnt signaling antagonist Kremen1 is required for

656 development of thymic architecture. Clin Dev Immunol, 13(2-4), 299-319.

657 doi:10.1080/17402520600935097

658

659 Osada, M., Jardine, L., Misir, R., Andl, T., Millar, S. E., & Pezzano, M. (2010). DKK1

660 mediated inhibition of Wnt signaling in postnatal mice leads to loss of TEC progenitors

661 and thymic degeneration. PLoS One, 5(2), e9062. doi:10.1371/journal.pone.0009062

662

663 Pereira, P., Gerber, D., Huang, S. Y., & Tonegawa, S. (1995). Ontogenic development

664 and tissue distribution of V gamma 1-expressing gamma/delta T lymphocytes in normal

665 mice. J Exp Med, 182(6), 1921-1930. doi:10.1084/jem.182.6.1921

666

667 Pongracz, J., Hare, K., Harman, B., Anderson, G., & Jenkinson, E. J. (2003). Thymic

668 epithelial cells provide WNT signals to developing thymocytes. Eur J Immunol, 33(7),

669 1949-1956. doi:10.1002/eji.200323564

670

30 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

671 Ripen, A. M., Nitta, T., Murata, S., Tanaka, K., & Takahama, Y. (2011). Ontogeny of

672 thymic cortical epithelial cells expressing the thymoproteasome subunit beta5t. Eur J

673 Immunol, 41(5), 1278-1287. doi:10.1002/eji.201041375

674

675 Shinohara, T., & Honjo, T. (1997). Studies in vitro on the mechanism of the

676 epithelial/mesenchymal interaction in the early fetal thymus. Eur J Immunol, 27(2),

677 522-529. doi:10.1002/eji.1830270225

678

679 Swann, J. B., Happe, C., & Boehm, T. (2017). Elevated levels of Wnt signaling disrupt

680 thymus morphogenesis and function. Sci Rep, 7(1), 785. doi:10.1038/s41598-017-

681 00842-0

682

683 Takahama, Y., Ohigashi, I., Baik, S., & Anderson, G. (2017). Generation of diversity in

684 thymic epithelial cells. Nat Rev Immunol, 17(5), 295-305. doi:10.1038/nri.2017.12

685

686 Travis, A., Amsterdam, A., Belanger, C., & Grosschedl, R. (1991). LEF-1, a gene

687 encoding a lymphoid-specific protein with an HMG domain, regulates T-cell receptor

688 alpha enhancer function. Genes Dev, 5(5), 880-894. doi:10.1101/gad.5.5.880

689

690 Uddin, M. M., Ohigashi, I., Motosugi, R., Nakayama, T., Sakata, M., Hamazaki, J.,

691 Nishito, Y., Rode, I., Tanaka, K., Takemoto, T., Murata, S., & Takahama, Y. (2017).

692 Foxn1-beta5t transcriptional axis controls CD8(+) T-cell production in the thymus. Nat

693 Commun, 8, 14419. doi:10.1038/ncomms14419

694

31 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

695 Varecza, Z., Kvell, K., Talaber, G., Miskei, G., Csongei, V., Bartis, D., Anderson, G.,

696 Jenkinson, E. J., & Pongracz, J. E. (2011). Multiple suppression pathways of canonical

697 Wnt signalling control thymic epithelial senescence. Mech Ageing Dev, 132(5), 249-

698 256. doi:10.1016/j.mad.2011.04.007

699

700 Wertheimer, T., Velardi, E., Tsai, J., Cooper, K., Xiao, S., Kloss, C. C., Ottmuller, K. J.,

701 Mokhtari, Z., Brede, C., deRoos, P., Kinsella, S., Palikuqi, B., Ginsberg, M., Young, L.

702 F., Kreines, F., Lieberman, S. R., Lazrak, A., Guo, P., Malard, F., Smith, O. M., Shono,

703 Y., Jenq, R. R., Hanash, A. M., Nolan, D. J., Butler, J. M., Beilhack, A., Manley, N. R.,

704 Rafii, S., Dudakov, J. A., & van den Brink, M. R. M. (2018). Production of BMP4 by

705 endothelial cells is crucial for endogenous thymic regeneration. Sci Immunol, 3(19).

706 doi:10.1126/sciimmunol.aal2736

707

708 White, A. J., Nakamura, K., Jenkinson, W. E., Saini, M., Sinclair, C., Seddon, B.,

709 Narendran, P., Pfeffer, K., Nitta, T., Takahama, Y., Caamano, J. H., Lane, P. J.,

710 Jenkinson, E. J., & Anderson, G. (2010). Lymphotoxin signals from positively selected

711 thymocytes regulate the terminal differentiation of medullary thymic epithelial cells. J

712 Immunol, 185(8), 4769-4776. doi:10.4049/jimmunol.1002151

713

714 Yano, M., Kuroda, N., Han, H., Meguro-Horike, M., Nishikawa, Y., Kiyonari, H.,

715 Maemura, K., Yanagawa, Y., Obata, K., Takahashi, S., Ikawa, T., Satoh, R., Kawamoto,

716 H., Mouri, Y., & Matsumoto, M. (2008). Aire controls the differentiation program of

717 thymic epithelial cells in the medulla for the establishment of self-tolerance. J Exp Med,

718 205(12), 2827-2838. doi:10.1084/jem.20080046

32 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

719

720 Zuklys, S., Gill, J., Keller, M. P., Hauri-Hohl, M., Zhanybekova, S., Balciunaite, G.,

721 Na, K. J., Jeker, L. T., Hafen, K., Tsukamoto, N., Amagai, T., Taketo, M. M., Krenger,

722 W., & Hollander, G. A. (2009). Stabilized beta-catenin in thymic epithelial cells blocks

723 thymus development and function. J Immunol, 182(5), 2997-3007.

724 doi:10.4049/jimmunol.0713723

725

726 Zuklys, S., Handel, A., Zhanybekova, S., Govani, F., Keller, M., Maio, S., Mayer, C. E.,

727 Teh, H. Y., Hafen, K., Gallone, G., Barthlott, T., Ponting, C. P., & Hollander, G. A.

728 (2016). Foxn1 regulates key target genes essential for T cell development in postnatal

729 thymic epithelial cells. Nat Immunol, 17(10), 1206-1215. doi:10.1038/ni.3537

730

33 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

731 Figure legends

732 Figure 1. Thymic dysplasia caused by overexpression of b-catenin in TECs. (A)

733 Thoracic cavity of control mice and b-cat GOF mice at E15.5. Dotted lines show the

734 outline of the thymic primordium. In many cases, blood clots were observed in the

735 central core of the thymic primordium in b-cat GOF mice. Representative data from

736 three independent experiments are shown. TH: thymus, H: Heart, Bars: 1 mm. (B)

737 Labeling of b5t-iCre activated cell progenies with tdTomato fluorescence in the thymus

738 of control mice and b-cat GOF mice at E15.5. Mouse sections were nuclear

739 counterstained with TO-PRO3. Bottom panels are magnifications of white-boxed areas

740 in the top panels. Representative data from three independent experiments are shown.

741 Bars: indicated in figures. (C) Hematoxylin and eosin staining (top) and

742 immunofluorescence staining for K5 and K8 (middle) and b-catenin and b5t (bottom)

743 on sagittal sections of thymic primordium in control mice and b-cat GOF mice at E15.5.

744 Representative results from three independent experiments are shown. Bars: 100 µm.

745 (D) Intracellular staining of b-catenin in CD45-EpCAM+ TECs isolated from control

746 mice and b-cat GOF mice at E15.5. Histograms show b-catenin expression in control

747 TECs (blue line) and b-cat GOF TECs (red line). Shaded area and black line represent

748 the fluorescence in the absence of anti-b-catenin antibody in control TECs and b-cat

749 GOF TECs, respectively. Plots on the right show net median fluorescence intensity

750 (MFI) values (means and SEMs, n = 4-5). Numbers in parentheses indicate percentage

751 of control value. *p < 0.05.

752

34 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

753 Figure 2. Defective thymus development in b-cat GOF embryos. (A)

754 Immunofluorescence staining for CD45 and Foxn1 on sagittal sections of thymic

755 primordium in control mice and b-cat GOF mice at E11.5-E15.5. The sections were

756 nuclear counterstained with TO-PRO3. Anterior-posterior (A-P) and dorsal-ventral (D-

757 V) orientations of the images are indicated. Representative data from three independent

758 experiments are shown. Bars: 100 µm. (B) Quantitative RT-PCR analysis of mRNA

759 expression levels (means and SEMs, n = 3-4) of indicated genes relative to Gapdh

760 levels in CD45-EpCAM+ TECs isolated from the thymus of control mice and b-cat

761 GOF mice at E15.5. (C) The numbers of CD45+ thymocytes were analyzed by flow

762 cytometry. Plots show the numbers (means and SEMs, n = 3-4) of CD45+ thymocytes

763 in the thymus of control mice and b-cat GOF mice at E15.5. (D) Flow cytometric

764 analysis of double negative (DN) thymocytes from control mice and b-cat GOF mice at

765 E15.5. Shown are profiles of CD44 and CD25 expression. Numbers in dot plots indicate

766 frequency of cells within indicated area. (E) Cell numbers (means and SEMs, n = 3-4)

767 of indicated DN thymocyte subpopulations from control mice and b-cat GOF mice at

768 E15.5 are plotted. *p < 0.05; **p < 0.01; *** p < 0.001; N.S., not significant.

769

770 Figure 3. Thymic dysplasia in postnatal b-cat GOF mice. (A) Appearance of control

771 mice and b-cat GOF mice at 8 wk. Bar: 1 cm (B) Body weight of control mice and b-cat

772 GOF mice at 8 wk (means and SEMs, n = 4-6). N.S., not significant. (C) Appearance of

773 the thymus in control mice and b-cat GOF mice at postnatal stages (P1 to 8 wk). Bar: 1

774 mm.

775

35 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

776 Figure 4. Loss of ab T cells and gd T cells in secondary lymphoid organs by

777 overexpression of b-catenin in TECs. (A) Flow cytometric analysis of splenocytes from

778 control mice and b-cat GOF mice at 11 wk. Shown are representative dot plot profiles

779 of CD3 and TCRb expression (left) and CD3 and TCRd expression (right) in PI- viable

780 cells. Numbers in dot plots indicate frequency of cells within indicated area. (B) Cell

781 numbers (means and SEMs, n = 4-6) of indicated subpopulations in the spleen (left) and

782 the iLN (right) from control and b-cat GOF mice at 8 wk are plotted. (C) Shown are

783 representative dot plots of TCRd and Vg5 expression in CD45+ cells in the thymus from

784 control mice and b-cat GOF mice at E15.5. Frequency of TCRd+ Vg5+ cells is plotted

785 (mean and SEMs; n = 4-6). (D) Flow cytometric analysis of DETCs in the skin

786 epidermis from control mice and b-cat GOF mice at 8 wk. Shown are representative dot

787 plot profiles of CD3 and Vg5 expression in epidermal cells. Numbers in dot plots

788 indicate frequency of cells within indicated area. Frequency of CD3+Vg5+ cells in

789 epidermal cells is plotted (means and SEMs; n = 3). (E) Histograms for Vg4 and Vg1

790 expression in TCRb-d+ cells in the iLN from control mice and b-cat GOF mice at 8 wk

791 are shown. Frequency of Vg4+ cells and Vg1+ cells is plotted (means and SEMs; n = 3).

792 *p < 0.05; ** p < 0.01; *** p < 0.001; N.S., not significant.

793

794 Figure 5. No apparent change in thymus development in b-cat LOF embryos. (A)

795 Thoracic cavity of control mice and b-cat LOF mice at E15.5. Dotted lines show the

796 outline of the thymic primordium. Representative data from three independent

797 experiments are shown. TH: thymus, H: Heart, Bars: 1 mm. (B) Immunofluorescence

36 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

798 staining for b-catenin and Foxn1 (left) or K5 and K8 (right) on sagittal sections of the

799 thymus from control mice and b-cat LOF mice at E15.5. Shown are merged images

800 with nuclear counterstaining (TO-PRO3) (top) and images obtained in each channel

801 (middle, bottom). Representative data from three independent experiments are shown.

802 Bars: 100 µm. (C) Intracellular staining of b-catenin in CD45-EpCAM+ TECs from

803 control mice and b-cat LOF mice at E15.5. Histograms show b-catenin expression in

804 control TECs (blue line) and b-cat LOF TECs (red line). Shaded area and black line

805 represent the fluorescence in the absence of anti-b-catenin antibody in control TECs and

806 b-cat LOF TECs, respectively. Plots show net MFI values for b-catenin (means and

807 SEMs, n = 4). Numbers in parentheses indicate percentage of control value. (D) Flow

808 cytometric analysis of enzyme-digested thymic cells from indicated mice at E15.5.

809 Shown are profiles of EpCAM and CD45 expression in PI- viable cells (left) and UEA1

810 reactivity and Ly51 expression in CD45-EpCAM+ cells (right). Numbers in dot plots

811 indicate frequency of cells within indicated area. (E) Plots show the number of total

812 thymic cells (left) and the frequency and the number of total TECs (middle) and cTECs

813 (right) from control mice and b-cat LOF mice at E15.5 (means and SEMs, n = 4). **p <

814 0.01; N.S., not significant.

815

816 Figure 6. Reduction of thymus size in postnatal b-cat LOF mice. (A) Appearance of

817 control mice and b-cat LOF mice at 8 wk. Bar: 1 cm (B) Appearance of the thymus

818 from control mice and b-cat LOF mice at postnatal stages (2 wk to 8 wk).

819 Representative data from at least three independent experiments are shown. Bar: 1 mm

820 (C) Bars show body weight (top) and thymus weight (bottom) at 2 wk (left, n = 3), 4 wk

37 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

821 (middle, n = 4), and 8 wk (right, n = 5) in control mice and b-cat LOF mice (mean and

822 SEMs). *** p < 0.001; N.S., not significant.

823

824 Figure 7. Characteristics of TECs in postnatal b-cat LOF mice. (A) Intracellular staining

825 of b-catenin in UEA1-Ly51+ cTECs (left) and UEA1+Ly51- mTECs (right) from

826 control mice and b-cat LOF mice at 2 wk. Histograms show b-catenin expression in

827 cTECs and mTECs from control mice (blue line) and b-cat LOF mice (red line). Shaded

828 area and black line represent the fluorescence in the absence of anti-b-catenin antibody

829 in control TECs and b-cat LOF TECs, respectively. Plots show net MFI values for b-

830 catenin in cTECs and mTECs (means and SEMs, n = 3). Numbers in parentheses

831 indicate percentage of control value. (B) Quantitative RT-PCR analysis of mRNA

832 expression levels (means and SEMs, n = 5) of indicated genes relative to Gapdh levels

833 in UEA1-Ly51+ cTECs (top) and UEA1+Ly51- mTECs (bottom) in the thymus of

834 control mice and b-cat LOF mice at 2 wk. (C) Flow cytometric analysis of enzyme-

835 digested thymic cells from control mice and b-cat LOF mice at 2 wk. Shown are

836 representative profiles of EpCAM and CD45 expression in PI- viable cells (left) and

837 UEA1 reactivity and Ly51 expression in CD45-EpCAM+ viable cells (right). Numbers

838 in dot plots indicate frequency of cells within indicated area. (D) Plots show the number

839 (means and SEMs, n = 6) of cTECs and mTECs in the thymus from control mice and b-

840 cat LOF mice at 2wk. (E) Immunofluorescence analysis of b5t (green), CCL21 (red),

841 and Aire (cyan) on transverse sections of thymus from control mice and b-cat LOF mice

842 at 2 wk. Representative data from three independent experiments are shown. Bars: 100

38 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

843 µm. Ctrl: Control, LOF: b-cat LOF, *p < 0.05; ** p < 0.01; *** p < 0.001; N.S., not

844 significant.

845

846 Figure 8. Reduced total thymocyte production in b-cat LOF mice during postnatal

847 development. (A) Plots show the number (means and SEMs, n = 4-6) of total

848 thymocytes in the thymus of indicated mice at 2 wk (left), 4 wk (middle), and 8 wk

849 (right). (B) Flow cytometric analysis of thymocytes from control mice and b-cat LOF

850 mice at 8 wk. Shown are representative dot plot profiles of TCRb and TCRd expression

851 and CD4 and CD8 expression in PI- viable cells. Numbers in dot plots indicate

852 frequency of cells within indicated area. Plots show the frequency of indicated

853 thymocyte subpopulations (means and SEMs, n = 5). (C) Cell numbers (means and

854 SEMs, n = 5) of indicated subpopulations in the thymus from control mice and b-cat

855 LOF mice at 8 wk. (D) Flow cytometric analysis of lymphocytes in the iLN from

856 control mice and b-cat LOF mice at 8 wk. Shown are representative dot plot profiles of

857 CD4 and CD8 expression in TCRb+d- viable cells. Numbers in dot plots indicate

858 frequency of cells within indicated area. (E) Cell numbers (means and SEMs, n = 4-5)

859 of indicated subpopulations in the iLN from control mice and b-cat LOF mice at 8 wk.

860 ** p < 0.01; *** p < 0.001; N.S., not significant.

861

862 Figure 9. Phenotypic differences between conditional b-cat LOF and GOF mice during

863 postnatal period. GOF of b-catenin using Foxn1-Cre and b5t-iCre causes severe defect

864 in TEC development and arrest in T cell development at DN1, whereas Foxn1-Cre-

865 mediated GOF of b-catenin only results in the failure of migration of thymic

39 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

866 primordium into the thoracic cavity. LOF of b-catenin using Fonx1-Cre results in

867 perinatal lethality, whereas LOF of b-catenin using b5t-iCre does not lead to reduced

868 viability throughout development. b5t-iCre-mediated LOF of b-catenin results in

869 reduction in thymus size and the number of cTECs and total thymocyte production

870 during the postnatal period.

40 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.1.

A C Control b-cat GOF Control b-cat GOF

TH TH HE

H H

E15.5 PRO3

B - Control b-cat GOF TO / K8

E15.5 / K5 PRO3 - TO / 5t b / cat - b

1 mm

D Control without b-catenin Ab b5t b-catenin GOF Control with b-catenin Ab b5t b-catenin GOF

TECs 10 (100%) (170%)

8 100 m * µ 6 tdTomato/TO-PRO3

(x1000) 4 catenin Net MFI Net catenin - 2 b

0 b-catenin Ctrl GOF bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.2. A B Control b-cat GOF A Axin2 Dkk1 Msx1 0.10 0.025 ** 0.05 ** V D ) *

Ct 0.08 0.020 0.04 D P 0.06 0.015 0.03 E11.5 CD45 0.04 0.010 0.02 Foxn1

in TECs (2TECs in 0.02 0.005 0.01

TO-PRO3 expression Gene

0.00 0.000 0.00 Ctrl GOF Ctrl GOF Ctrl GOF

Foxn1 Psmb11 Ccl25 E12.5 0.20 0.8 1.5 )

Ct 0.15 0.6 D 1.0

0.10 0.4

0.5 0.05 0.2 in TECs (2TECs in

Gene expression expression Gene *** ** ** E13.5 0.00 0.0 0.0 Ctrl GOF Ctrl GOF Ctrl GOF

Ivl Lor * 0.0015 * 0.006 Control

) b-cat GOF Ct

D 0.0010 0.004 E15.5

0.0005 0.002 in TECs (2TECs in Gene expression expression Gene 0.0000 0.000 Ctrl GOF Ctrl GOF

C D E Control

) Control b-cat GOF b-cat GOF 5 ) 9 20 59 6 5 2.0 3 1.5 2 1.0

thymocytes (x10 thymocytes N.S. 1 *** 0.5 + 12 59 27 8 ** * *** 0 0 Ctrl GOF (x10 cells of Number CD44 CD45 CD25 DN1 DN2 DN3 DN4 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.3.

A C Control b-cat GOF Control b-cat GOF

P1

8 wk

2 wk

B 8 wk

25 N.S. 20 4 wk 15

10

5

Body weight (g) weight Body 0 Ctrl GOF

8 wk bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.4.

A B Control 8 wk b-cat GOF ) ) 8 PI- PI- Spleen 6 iLN 2.0 6 1 23 24 1 N.S. 1.5 4 N.S. 1.0 2 0.5

Control 0.0 0

Number of cells (x10 cells of Number Ctrl GOF Ctrl GOF 75 1 75 0 (x10 cells of Number ) ) 6 7 0 0 0 0

3 5

4 T cells (x10 cells T

T cells (x10 cells T 2 3 ab cat GOF cat ab - 2 b 1 ** 1 100 0 99 1 * 0 0 CD3 CD3 Ctrl GOF Ctrl GOF Number of of Number

TCRb TCRd of Number

C D E E15.5 Thymus 8 wk Skin 8 wk iLN

CD45+ PI- TCRb-d+ TCRb-d+

1.6 2.9 28 24 Control Control Control Control

0.4 0 3 1 cat GOF cat GOF cat GOF cat GOF cat - - - - b b b b d CD3 TCR Vg5 Vg5 Vg4 Vg1

+ TCRd+Vg5+ cells CD3+Vg5+ cells Vg4 cells Vg1+ cells 2.0 4 50 50

cells 40 40 1.5 3 + 30 30 1.0 2 cells T cells T *** gd 20 gd 20 0.5 1 ** ** ** 10 10 0.0 0 % of 0 % of 0 % of CD45

Ctrl GOF cells Epidermal of % Ctrl GOF Ctrl GOF Ctrl GOF bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.5. A B Control b-cat LOF Control b-cat LOF Control b-cat LOF E15.5 E15.5 PRO3

TH TH - PRO3 - TO / TO H /

H K8 / Foxn1 /

E15.5 K5 cat - b C Control without

b-catenin Ab cat b5t b-catenin LOF - K5 Control with b b5t b-catenin LOF b-catenin Ab

TECs 2.5 (100%) (23%)

2.0

1.5 K8 Foxn1 1.0

(x1000) ** 0.5 catenin Net MFI Net catenin - b 0.0 b-catenin Ctrl LOF

D E TECs cTECs 10 100 - - + E15.5 thymus PI CD45 EpCAM (%) (%) + 8 80 N.S.

N.S. + 1 2 4.7 Control 6 60 Ly51 -

b-cat LOF EpCAM - 4 40

2 20 UEA1 Control CD45 0 0 28 69 Ctrl LOF Ctrl LOF

0.5 25000 15000 )

2 3 6 5.9 0.4 N.S. cells 20000 N.S. cells +

N.S. + 10000 0.3 15000 Ly51

0.2 10000 - EpCAM cat LOF LOF cat -

- 5000 b 0.1 5000 Total cells (x10 cells Total UEA1

35 60 CD45 0.0 0 0 UEA1 EpCAM Ctrl LOF Ctrl LOF Ctrl LOF CD45 Ly51 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.6.

A B Control b-cat LOF Control b-cat LOF

2 wk

8 wk 4 wk

8 wk

C 2 wk 4 wk 8 wk 25 25 25 N.S. 20 20 N.S. 20 15 15 15 10 N.S. 10 10 5 5 5

0 0 (g) weight Body 0 Body weight (g) weight Body (g) weight Body Ctrl LOF Ctrl LOF Ctrl LOF

100 100 100

80 80 80 60 60 *** 60 *** 40 *** 40 40 20 20 20

0 0 0 Thymus weight (mg) weight Thymus Ctrl LOF (mg) weight Thymus Ctrl LOF (mg) weight Thymus Ctrl LOF bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.7. (100%) (4%) (100%) (6%) cTECs mTECs 2.0 A 6

1.5 Control without 4 b5t b-catenin LOF b-catenin Ab 1.0 Control with 2 b5t b-catenin LOF b-catenin Ab 0.5 catenin Net MFI Net catenin catenin Net MFI Net catenin * * - - in cTECs (x1000) cTECs in b b in mTECs(x1000) in 0 0.0 b-catenin Ctrl LOF b-catenin Ctrl LOF B 2 wk cTECs Ctnnb1 Axin2 Foxn1 Il7 Cxcl12 Control b-cat LOF 0.25 0.005 0.25 0.15 N.S. 2.5 ) N.S. N.S.

Ct 0.20 0.004 0.20 2.0 D ** 0.10 0.15 0.003 0.15 1.5

0.10 0.002 0.10 1.0 ** 0.05

in cTECs (2cTECs in 0.05 0.001 0.05 0.5 Gene expression Gene

0.00 0.000 0.00 0.00 0.0 Ctrl LOF Ctrl LOF Ctrl LOF Ctrl LOF Ctrl LOF

2 wk mTECs

Ctnnb1 Axin2 Aire Tnfrsf11a Ccl21a

0.25 0.005 0.8 0.20 1.0 ) N.S.

Ct N.S.

D 0.20 0.004 0.8 0.6 N.S. 0.15 N.S. 0.15 0.003 0.6 0.4 0.10 0.10 0.002 0.4

0.2 0.05 0.05 0.001 0.2 in mTECs(2 in

Gene expression Gene ***

0.00 0.000 0.0 0.00 0.0 Ctrl LOF Ctrl LOF Ctrl LOF Ctrl LOF Ctrl LOF

C D cTECs mTECs 20000 80000 PI- CD45-EpCAM+

cells 15000 cells 60000 N.S. - 0.2 72 4 + 10000 40000 Ly51 Ly51 - +

5000 * 20000 Control UEA1 UEA1 0 0 11 13 Ctrl LOF Ctrl LOF

3 0.3 84 E Control b-cat LOF cat LOF LOF cat - b Aire / 5t b 8 5 UEA1 EpCAM CCL21 CD45 Ly51 / 5t b bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.8.

A ) ) )

8 2 wk 8 4 wk 8 8 wk

1.5 2.0 1.5

1.5 1.0 1.0 *** *** 1.0 ** 0.5 0.5 0.5

0.0 0.0 0.0 Ctrl LOF Ctrl LOF Ctrl LOF Number of thymocytes (x10 thymocytes of Number (x10 thymocytes of Number (x10 thymocytes of Number

B C Control Control

b-cat LOF ) b-cat LOF N.S. 7 8 wk Thymus 100 15

80 10 PI- PI- *** 60 5 0 0 5 88 10 N.S. 1.0

5 N.S. N.S. 0.5 N.S. ** *** 88 thymocytes of % 0 0.0 Control Number of cells (x10 cells of Number DN DP CD4 CD8 DN DP CD4 CD8 12 3 4 SP SP SP SP 0 0 5 89 - + - +

TCRb d ) TCRb d 5

0.25 2.0

88 1.5 cat LOF LOF cat

- 0.20 N.S.

b *** 1.0 d 12 2 4 0.15 0.5 CD4 TCR 0.10 0.0 TCRb CD8 thymocytes of % Ctrl LOF Ctrl LOF Number of cells (x10 cells of Number D E

8 wk iLN 8 wk iLN Control TCRb+d- b-cat LOF 49 1 Total cells TCRb+d- TCRb-d+ ) ) ) 6 6 6 2.5

8 2.0 10 Control N.S. N.S. N.S. 8 6 1.5 N.S. 1 49 6 4 1.0 55 1 4 2 0.5 2

0 0.0 0 Number of cells (x10 cells of Number (x10 cells of Number Number of cells (x10 cells of Number Ctrl LOF Ctrl LOF Ctrl LOF Ctrl LOF CD4SP CD8SP cat LOF LOF cat - b 1 43 CD4 CD8 bioRxiv preprint doi: https://doi.org/10.1101/2021.05.02.442353; this version posted May 2, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder. This article is a US Government work. It is not subject to copyright under 17 USC 105 and is also made available for use under a CC0 license.

Fig.9.

Normal b-catenin GOF b-catenin LOF

Control Foxn1-Cre b5t-Cre Foxn1-Cre b5t-Cre

Thymus

Perinatal lethality

Heart

Undisturbed Undisturbed Migration failure of thymus thoracic migration thoracic migration

Thymic dysplasia Thymic dysplasia Small thymus size

Defect in Defect in Reduction in TEC development TEC development the number of cTECs

Arrest in T cell Arrest in T cell Reduction in development at DN1 development at DN1 thymocyte production