bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

1

2

3

4 Japonicum infection in

5 Treg-specific USP21 knock-out mice 6 7 Youxiang Zhang1☯, Guina Xu2☯, Baoxin Zhang4, Shan Zhang1, Yangyang Li3, Qing 8 Huang1, Simin Chen1, Fansheng Zeng2, Bin Li3, Zhiqiang Qin5*, Zuping Zhang1* 9 10 11 12 13 1 Department of Pathogen Biology, School of Basic Medical Science, Central South Univesrsity, 14 Changsha 410078, People's Republic of China 15 2 Key Laboratory of Control Technique for Schistosoma and Pathogen Infection for Dongting 16 Lake Region, Yiyang Medical College,Yiyang 413002, China 17 3 Department of Microbiology and Immunology, Shanghai Institute of Immunology, Shanghai 18 Jiao Tong University School of Medicine, Shanghai 200025, China 19 4 Department of Antiepidemic, Armed Police Hospital of Hunan Province, Changsha 410008, 20 Hunan, China 21 5 Key Laboratory of Parasite and Vector Biology, Ministry of Health; National Institute of 22 Parasitic Diseases, Chinese Center for Disease Control and Prevention, Chinese Center for Topical 23 Diseases Research, Shanghai 200025, People's Republic of China 24 25 26 *Correspondence and requests for materials should be addressed to: 27 [email protected]; [email protected] 28 29 30 ☯ These authors contributed equally to this work. 31 32 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

33 Abstract

34 USP21, an E3 de-ubiquitin enzyme playing vital roles in

35 physiological activities, is important for Treg cells to maintain immune

36 homeostasis and control immune tolerance. To understand its diverse

37 functions and potential mechanism is essential for disease development.

38 We, using the USP21 gene-conditional knock-out mice model of

39 Schistosoma Japonicum infection, found more cercariae developed into

40 adults, and more eggs deposited in the liver in KO mice. However,

41 immunohistochemistry showed the degree and the area of egg

42 granuloma and liver fibrosis were both reduced. This suggested

43 knock-out USP21 did affect the immunoregulation between

44 schistosomes and the host. In KO mice the content of IFN-gamma and

45 IL-4, and the expression of anti-SEA IgG and anti-SWAP IgG both

46 increased in the liver, spleen and blood by flow

47 cytometry , while the content of IL-10, lL-17A, IL-23, IL-9 and the

48 expression of USP21 and anti-SEA IgM decreased. This indicated

49 USP21-knockout-Tregs promoted both Th1-type and Th2-type

50 immunity and inhibited other immunities during schistosomes infection,

51 which disordered the host immunity. This study revealed the

52 immunomodulatory of USP21 and preliminarily suggests it might be bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

53 essential to regulate the complex immune network between the host and

54 schistosomes. USP21 provides a new possible target for

55 treatment in the future . 56

57 Author summary

58 Schistosomiasis is a common neglected tropical disease that

59 affects more than 230 million people worldwide. Therefore, the study on

60 the mechanism of immune interaction between schistosomas and the host

61 is not only helpful for the understanding of immune homeostasis, but also

62 helpful for the further development of the treatment of schistosomiasis.

63 Ubiquitin Specific Protease 21(USP21) has been shown to be involved in

64 the regulation of many biological processes, such as maintaining immune

65 homeostasis and regulating cell growth. Here, we observed that the

66 specific deletion of USP21 led to the decrease of mice's ability to resist

67 schistosomes infection and promoted the survival of schistosomes. It

68 was also proved that unstable regulatory T cells would produce

69 polarization phenomenon and promote differentiation to helper T cells,

70 which would lead to disorder of immune response in mice. However, this

71 process reduced the serious immune pathological damage caused by egg

72 granuloma. Our findings reveal that USP21 may be an important bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

73 molecule regulating immune interaction between Schistosoma japonicum

74 and the host. 75

76 Introduction

77 Schistosomiasis is an acute and chronic caused

78 by blood flukes (trematode worms) of the genus Schistosoma. According

79 to the WHO Report 2017, Schistosomiasis transmission was reported

80 from 78 countries, and at least 220.8 million people required preventive

81 treatment in 2017. The estimates of death due to Schistosomiasis

82 varies between 24 072 and 200 000 globally per year. In 2000, WHO

83 estimated the annual death at 200 000 globally[1]. People become

84 infected when larval forms of the parasite–released by freshwater –

85 penetrate the skin during contact with infested water. In the body, the

86 larvae develop into adult schistosomes. Adult worms live in the blood

87 vessels where the females release eggs. Some of the eggs are passed out

88 of the body in the feces or urine to continue the parasite's lifecycle.

89 Others become trapped in body tissues, causing immune reactions and

90 progressive damage to organs[2]

91 The immune systems of infected hosts have several life cycle

92 stages of the parasite that it must confront: penetrating cercariae, bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

93 migrating schistosomula, adult worms and the eggs produced by adult

94 worm pairs. The developing stages of the parasite can express hundreds

95 of antigenic moieties, many of which stimulate strong humoral and

96 cellular immune responses[3, 4]. Some of these responses continue to

97 increase during chronic infection, and others are strongly down-regulated

98 [5-7]. Observations in murine experimental infection models have shown

99 the mechanisms governing the development and regulation of the

100 pathogenic immune response in Schistosomiasis [8-11]. The mechanisms

101 of immunomodulation include IL-10, T regulatory cells, B cells,

102 antibodies, and T cell anergy[12-15]. Besides, several studies have

103 suggested that Th17 cells are involved in the pathogenesis of both S.

104 mansoni and S. japonica [16-18].

105 Treg cells (Tregs) have the immunosuppressive capacity, essential

106 for maintaining immune homeostasis and controlling immune tolerance.

107 It has been shown that inducing of Treg cells can inhibit the development

108 of granuloma pathology [19-24]. FOXP3 is a crucial factor in the

109 phenotype of Treg cells to obtain immunosuppressive activity. Unstable

110 FOXP3 will keep the Treg cell phenotype stable, may bias different T

111 helper cell-like phenotypes in different inflammatory conditions[20, 21,

112 25, 26]. Further, the conditional knock-out mouse model found that

113 USP21 stabilizes the expression of FOXP3 protein by deubiquitination in bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

114 regulatory T cells, thereby regulating the function of Treg cells. A

115 Th1-like sputum cell response, as expressed in cells of the

116 USP21-deficient Treg, preferentially becomes a Th2 cell-like phenotype

117 in a host with a severe Th2-type disorder, whereas, in arthritic conditions,

118 regulatory T cells may lose FOXP3 expression and transform into Th17

119 cell-like cells [23, 27-30]. More importantly, the three (FOXP3 and

120 GATA3, USP21) are closely related to regulatory T cells. In regulatory T

121 cells, FOXP3 binds to the promoter region of USP21 and activates its

122 transcription, while USP21 interacts with GATA3 and deubiquitinates it,

123 inhibiting its degradation by the proteasome to maintain its protein

124 stability.

125 Moreover, GATA3 can regulate the function of T cells in the

126 inflammatory response by stabilizing the expression of FOXP3. Thus, in

127 the regulatory T cells, the above three form a positive feedback pathway.

128 Further, conditional knock-out mouse models found that USP21 regulated

129 the expression of FOXP3 protein by deubiquitination in regulatory T

130 cells, thereby regulating the function of Treg cells [26, 31]. In the

131 antiviral reaction, USP21 can bind to and deubiquitinate RIG-I in the

132 cytoplasm to play an immunomodulatory role, or hydrolyze the

133 K27/63-linked polyubiquitin chain on STING to negatively regulate DNA

134 virus-induced type I interferon[32, 33]. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

135 Although regulatory T cells may play an essential role in the

136 immune regulation of schistosomes infection, the molecular mechanisms

137 are not yet precisely defined. Therefore, this study aims to explore the

138 mechanism of action of USP21 in the immune regulation of schistosomes

139 infection by using the USP21 gene knock-out mice model. It will describe

140 the conditional knock-out of USP21 mice infected with related immune

141 molecules in Schistosoma japonicum and further study the molecular

142 mechanism of USP21's immune response to Schistosoma japonicum. 143

144 Results

145 Depletion of USP21 in Treg cells weakens resistance to Schistosoma japonicum in

146 infected mice

147 In order to observe the difference between the KO mice and the

148 WT mice, both were infected with Schistosoma japonicum, and the

149 adults were recovered (according to the steps shown above), the number

150 of females and male were combined, and the total adults were counted.

151 We weighed the liver and spleen of the mice and observed the changes

152 in their color, shape and texture. (Fig 1B and Fig 1D). The number of

153 combined adults and total adults recovered from KO mice was

154 significantly more than that from WT mice (Fig 1A). However, no effect bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

155 on the development of cercariae into male and female adults was found.

156 Effect of USP21 on Schistosoma Japonicum eggs in infectious KO mice

157 A part of the liver was taken for HE staining from the mice 42

158 days after infection in order to observe the pathological changes of egg

159 granuloma. We analyzed the granuloma area and egg number in KO

160 mice and WT mice using appropriate software. The results showed that

161 the inflammatory cells of the egg granuloma of the KO mice were less

162 than that of WT mice, and the color of HE staining was lighter in KO

163 mice (Fig 2B). The area of liver egg granuloma in KO mice was

164 significantly smaller than that in WT mice (Fig 2C), but the number of

165 eggs in KO mice was significantly higher than that in WT mice (Fig

166 2A).

167 Changes of liver Fibrosis in USP21fl/flFOXP3cre mice infected with Schistosoma

168 japonicum

169 The proliferation of collagen fibers was observed by Masson

170 trichromatic staining, which reflected the severity of hepatic fibrosis

171 (Fig 3A). The mRNA expression of α-SMA, collagen-I and collagen-III

172 was detected by RT-PCR and the protein expression of them were

173 analyzed by Western blot. The area of collagen fiber in KO mice was

174 smaller, and the color was lighter than that in WT mice, which was bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

175 statistically significant (Fig 3B). The content of the α-SMA, collagen-I

176 and collagen-III in the KO mice was significantly less than that of the

177 WT mice at mRNA and protein levels (Fig 3C and Fig 3D).

178 Changes of spleen immune in USP21fl/flFOXP3cre mice infected with Schistosoma

179 Japonicum

180 To understand the difference of spleen immune cells between

181 KO mice and WT mice after infection with Schistosoma japonicum, the

182 spleen cells of the two groups, including normal control group, were

183 collected, isolated and then cultured in vitro to detect and analyze the

184 number of Treg cells and the percentage of CD4+CD25+FOXP3high cells

185 in CD4+ cells. The RT-PCR detected the changes of T cell types in the

186 detection of relative mRNA expression of IL-10, IL-17, FOXP3, USP21,

187 and by Multiplex Fluorescent Microsphere Immunoassay in the

188 detection of the content of IFN-gamma, IL-4, IL-10, IL-17A, IL-23,

189 IL-9 in splenocytes. mRNA levels of IL-10, IL-17, FOXP3 and USP21

190 in KO mice were less than that of WT mice (Fig 4C). The Flow

191 Cytometry results were shown in Fig 4A and Fig 4B, and Splenic Cell

192 Culture Cytokine results were showed as Fig 4D.

193 Liver immunity in USP21fl/flFOXP3cre mice infected with Schistosoma Japonicum

194 In order to study whether the immune status of the liver was bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

195 consistent with the condition of the spleen, the study detected the

196 mRNA expression of IL-10, IL-17 and FOXP3 in the liver. We found

197 that the mRNA expression of IL-10, IL-17 and FOXP3 in KO mice was

198 significantly lower than that in WT mice (Fig 5D). The relatively large

199 difference noted might be related to the small proportion of T cells in

200 the liver. The mRNA and protein expression levels (Fig 5A) of USP21

201 in the liver were also lower in KO mice than that in WT mice (Fig 5B

202 and Fig 5C).

203 Specific antibody response in USP21fl/flFOXP3Cre mice infected with Schistosoma

204 japonicum

205 To better understand the specific antibody response of USP21

206 fl/flFOXP3Cre mice to Schistosoma japonicum, Serum samples from

207 different stages were collected to measure the content of anti-SEA and

208 anti-SWAP IgG/IgM antibodies from KO and WT mice in different

209 infectious stages. The results were shown as follows: anti-SEA IgG (Fig

210 6A), anti-SEA IgM (Fig 6B), anti-SWAP IgG (Fig 6C), and anti-SWAP

211 IgM (Fig 6D). bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

212 Serum cytokines in USP21fl/flFOXP3Cre mice infected with Schistosoma

213 japonicum

214 To intensively study on how Treg cells lacking USP21 affect mice

215 resistance to Schistosoma japonicum, we tested the content of

216 IFN-gamma, IL-4, IL-10, IL-17A, IL-23 and IL-9 in serum samples and

217 PBMC from different infectious stages mice, including KO and WT mice.

218 The results were shown as follows: comparison of the changes in the

219 trend of the two groups of cytokines, including IFN-gamma, IL-4, IL-10,

220 IL-17A, IL-23 and IL-9（Fig 7B）, and comparison of cytokines in cell

221 culture in vitro (Fig 7A).

222

223 Discussion

224 USP21 is a member of the ubiquitin-specific proteolytic enzyme

225 (USP) family and plays different biological functions. For example,

226 USP21 regulated the gene expression of hepatocytes during liver

227 regeneration by catalyzing the hydrolysis of ubH2A[34]. In inflammation,

228 USP21 negatively regulated RIPK1 to inhibit the activity in the

229 downstream of TNF-αR1[31], and USP21-mediated deubiquitination of

230 IL-33 promoted the transcription of NF-κB p65[35]. Also, USP21 could

231 bind to and deubiquitinate RIG-I in the cytoplasm to play an bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

232 immunomodulatory role in antiviral reactions[32]. The latest studies have

233 also shown that USP21 could regulate the expression of the cell cycle,

234 proliferation, and craniofacial development related factors by

235 deubiquitinating FOXM1 and Goosecoid (GSC)[36, 37]. We have

236 previously used USP21 knockout mice model to explore their biological

237 role in the growth and development of mice, and the differentiation and

238 growth of lymphocytes and hematopoietic stem cells[38]. It is known that

239 USP21 knock-out elderly mice showed spontaneous T cell activation and

240 , and that the autoimmune system disorder was caused by

241 Th1-like Tregs induced by the unstable expression of FOXP3 in USP21

242 deficient Treg mice model [26].

243 All stages of the schistosome life cycle inside the mammalian host

244 elicit immune responses, but only the eggs are the target of a vigorous

245 granulomatous inflammatory reaction. The immunopathological damage

246 caused by Schistosoma japonicum is the result of promoting and

247 inhibiting inflammation among different groups of T cells, such as Th1,

248 Th2, Tregs, and Th17. Tregs play an essential immunomodulatory role in

249 reducing its immunopathological reaction[2, 19, 39, 40]. To illustrate the

250 effect and possible mechanism of unstable Tregs on Schistosoma

251 japonicum infection, and to study whether USP21 plays a new role in

252 controlling the stability of Tregs, we used C57BL/6 mice model with the bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

253 conditional knockout USP21 in Tregs, which was infected Schistosoma

254 japonicum and was often used as model high-vs.low-pathology strains

255 [41].

256 Schistosomes infection can be divided into three stages and can

257 present in different clinical symptoms and immune mechanisms: acute

258 infection, active infection, and late chronic infection [42]. Our study

259 found that after Schistosoma japonicum infection, the number of coups

260 and adults recovered from the KO mice was much larger than from the

261 WT mice, and the same trend was noted for the number of the eggs. At

262 the same time, we noted that there was no gender difference in adults

263 between the two groups. It is known that Schistosoma depended on host

264 signals to keep correct development and maturation, such as TGF-α and

265 TNF-α [43]. We speculated that the changes in the immune

266 microenvironment in the host might affect the binding of Schistosoma

267 japonicum. However, there was no apparent difference in the liver and

268 spleen between KO and WT mice with Schistosoma japonicum.

269 Considering that the protection of Schistosoma japonicum infection was

270 mainly based on the clearance of Schistosoma japonicum in the early

271 infection stage, it also seemed that USP21-/- in Tregs might inhibit the

272 lethal abilities of Schistosoma Japonicum. These two theories need to be

273 further evaluated in the future. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

274 Th1 immune response mainly in the early stage of Schistosoma

275 Japonicum infection was to eliminate adults, and Th2 immune response

276 was mainly to anti-eggs in the later stage[44]. Liver egg granuloma was

277 the primary immunopathological reaction, the severity of the symptom of

278 which was usually related to the intensity of infection [42]. We found that

279 the diameter of egg granuloma in KO mice was smaller than that in WT

280 mice at six weeks, and the infiltration of inflammatory-related substances

281 such as neutrophils was less and lighter. The degree of hepatic fibrosis in

282 KO mice was also lower than that in WT mice by RT-PCR, Western blot,

283 and Masson Staining. The immune response of T cells fluctuates at

284 different levels during the infection of Schistosoma japonicum, Th1, and

285 T-reg are co-dominant in transcription level, Th1 and Th2 are

286 co-dominant in protein level [45]. These results suggested that unstable

287 Treg cells may reduce the pathological damage of the liver through some

288 mechanism.

289 Cytokines produced by egg antigen-stimulated liver T cell

290 populations from infected mice are faithful correlates of

291 immunopathology. High-pathology strains sustain a proinflammatory Th1

292 and Th17(IL-17) cell-polarized response alongside the Th2 response [41],

293 whereas an initial short-lived elevation in IFN-γ-producing Th1 cells

294 gives way to a mostly host-protective Th2(IL-4, IL-5, IL-13)-dominated bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

295 environment [46]. We found that the content of IFN- γ and IL-4 in KO

296 mice was higher than that in WT mice, while IL-10, IL-17, IL-23 and

297 IL-9 was lower than that in the WT group. With different durations of

298 infection, the trend of cytokines changes in the serum of two groups was

299 about the same. These results are partly consistent with the previous

300 findings, and once again verify that USP21 knock-out will shift Treg to

301 Th1-like cells[26]. IFN- γ content in the KO group was much higher than

302 that in the WT group, and IFN- γ in the later stage of Schistosoma

303 japonicum infection was maintained stably with IL-4 increasing.

304 However, it was different from those mentioned above that IL-4 content

305 was higher in KO mice than that in WT mice. The unstable Treg might

306 have produced different differentiation effects because of the different

307 microenvironments of the schistosome infection, and series of T-cell

308 activation and long-term stimulation of chronic inflammatory factors

309 could have low reactivity [26, 47, 48]. This coincidence was in line with

310 the weakening effect of other secreted factors in addition to IFN- γ and

311 IL-4. It was also speculated that Th1 and Th2 are the main immunity

312 types of the Schistosomiasis. According to the previous reports, Th17

313 might play an essential role in the formation and growth of the liver

314 immunopathology and the egg granuloma caused by Schistosoma adults,

315 and the development of severe Schistosomiasis was related to the high bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

316 level of IL-17[49, 50]. The results of this study also confirmed that IL-17

317 content in the KO group was less than that in the WT group, and the

318 pathological sections of the liver revealed, as shown in Fig.2B. IL-23 was

319 one of the driving factors to produce IL-17. We also observed that IL-17

320 was related to IL-23, both of which were lower in the KO group than that

321 in the WT group, and both were down-regulated.

322 The liver fibrosis and spleen diseases caused by Schistosoma

323 japonicum were associated with high FOXP3+Tregs in the blood[18].

324 FOXP3 played an important role in regulating the differentiation,

325 development, and function of Tregs, and the three FOXP3, GATA3 and

326 USP21 are closely related to Tregs. To clarify the relation between

327 USP21 and FOXP3+Tregs, our results of CD4+CD25+FOXP3+Tregs were

328 consistent with previous studies after spleen cell isolation despite some

329 differences. The percentage of FOXP3+Tregs in the KO group was lower

330 than that in the WT group, and the percentage of CD4+T cells was higher

331 than that in the WT group. The results indicated that the deletion of

332 USP21 made the expression of FOXP3 unstable and led to Spontaneous

333 activation and increasing of T cells. FOXP3+Tregs were reduced in the

334 KO group because of the easy degradation of unstable FOXP3. Another

335 evidence supported that FOXP3 instability promoted the production of

336 T-like Treg cells. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

337 Soluble egg antigen (SEA) and adult worm antigen (SWA) are the

338 main soluble proteins related to adaptive immune response induced by

339 Schistosoma japonicum infection. Related studies have shown that a high

340 level of anti-Schistosoma IgG was associated with increased

341 susceptibility to parasites. These studies have demonstrated a positive

342 correlation between anti-Schistosoma IgG, especially IgG4 response and

343 severe Schistosomiasis[51, 52]. In this study, the levels of SEA and SWA

344 IgG in USP21 knock-out mice infected with Schistosoma japonicum were

345 higher than those in the WT group while IgM levels were the same or

346 lower. This finding suggested that unstable Treg cells improved the

347 immune response to Schistosoma and reduced the initial immune

348 response and lethality to Schistosoma.

349 Worms live in the host for a long time, and there is growing

350 evidence that they can manipulate the host's immune system through the

351 host's immunomodulatory. The immunomodulatory network is activated

352 after chronic worm infection and can even cross-regulate the

353 non-associated inflammatory response of the host. It has been proven that

354 FOXP3+ and Tregs were essential parts of this regulatory network,

355 especially for the immunosuppression of chronic Schistosomiasis, which

356 was characterized by the independent down-regulation of IL-10 produced

357 by Th2 cytokines [53-55]. Therefore, it is suggested that unstable Tregs bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

358 could also be used by the parasite to avoid the immune system of the

359 host, which may provide a new research basis for the immunoregulation

360 between the host and the parasite, and is of considerable significance to

361 the treatment of the chronic inflammatory disease.

362 In conclusion, the significant heterogeneity of the host's immune

363 response to schistosome infection, involving different APC, T-cell

364 subsets and cytokines, led to different immunopathology levels. Our

365 study still found that unstable Tregs in mice infected with Schistosoma

366 japonicum might not only benefit the host from the excessive

367 inflammatory response but also regulate the living microenvironment of

368 parasites, both of which achieved a new balance. There were still some

369 limitations in this study because many factors, such as the subsequent

370 survival status of mice, were not considered. However, this paper could

371 provide a theoretical basis to study the regulation of USP21 further on

372 different immune cell groups induced by Schistosoma Japonicum, and

373 provide a new possible target for schistosomiasis treatment

374 (anti-Schistosoma japonicum) in the future.

375 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

376 Materials and Methods

377 Ethics statement

378 We carried out all experiments in strict accordance with

379 the Laboratory Animal Regulation (1988.11.1) and made every effort to

380 minimize the suffering. The IACUC of National Institution of Parasitic

381 Diseases of Chinese Center for Diseases Control and Prevention and

382 Control approved all animal procedures for the use of experimental

383 (License No: NJMU 07-0137).

384 Experimental animals, parasites, and the establishment of infection models

385 Two kinds of female mice, FOXP3Cre (WT) and

386 USP21fl/flFOXP3Cre (KO), were provided by Shanghai Institute of

387 Bioscience, Chinese Academy of Sciences. Schistosoma japonicum

388 cercariae escaped from naturally infected snails from the National

389 Institute of Parasitic Diseases of Chinese Center for Diseases Control and

390 Prevention (Shanghai, China). Each of the two kinds mice was divided

391 into a normal control group (NC/Uninfected) (n =2) and infection group

392 (INF) (n =10  2) for 6-8 weeks under specific pathogen-free conditions,

393 then the mice of INF group were infected with 25  1 cercariae via

394 abdomen and killed all at 6-7weeks after infection, including mice of bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

395 Uninfected group. Serum, stored at -80℃ and used for ELISA, was

396 collected from mice blood at different times, respectively before

397 challenge infection, 20 days and 30days after infection, and 42 days after

398 challenge infection with eyeball enucleation. The animal experiments

399 were carried out randomly by blind strategies, strictly per the Laboratory

400 Animal Regulation, and we made every effort to minimize pain. The

401 model of mice infected with the Schistosoma japonicum was established

402 by repeating three times.

403 Collection of Schistosoma japonicum worms and weight of the spleen and liver of

404 the infected mice

405 The mice, 42 days after challenge infection, were killed by the

406 cervical dislocated method. Worms were collected from the portal vein

407 system by cardiac infusion of saline, and the number of male and female

408 adult worms and hugs was determined. The spleen and liver of the mice

409 were weighed and photographed.

410 HE and Masson staining

411 The same site of each infected mouse was fixed in 4%

412 paraformaldehyde and embedded in the paraffin block. We stained a

413 slice (5μm thick) with hematoxylin-eosin and performed the analysis with

414 a microscope (magnification ×100). To evaluate the pathological changes bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

415 and the number of eggs among groups, the area of liver single egg

416 granuloma (n ≥ 5 per mouse) was measured by software, and the average

417 number of eggs was counted by a random selection of visual field (n ≥ 5

418 per mouse) at magnification × 40. Similarly, the same site of the liver was

419 stained with Masson. In order to evaluate the degree of hepatic fibrosis

420 between the two groups, we measured and analyzed the ratio of collagen

421 area (n ≥ 5 per mouse) using appropriate software.

422 Isolation of Peripheral Blood lymphocytes (PBMC) and Culture in vitro

423 0.5-1ml blood was taken from mice 42 days after infection by

424 eyeball enucleation, serum and plasma were separated, and serum was

425 stored at -80℃ to be studied. We mixed the plasma with sterilized PBS

426 in the proportion of 1:1, and the mixture was gently reversed back and

427 forth for about ten times and gently added the same volume of

428 lymphocyte isolation to form stratification. After centrifugation, the study

429 team took the middle layer to add red blood cell lysate to decompose red

430 cells, centrifuge again to get PBMC. The PBMC were resuspended with a

431 complete culture medium containing 5 μg /ml RMPI1640. The cell

432 density under the microscope was adjusted to 2×105-106/ml, then divided

433 into culture plate or dish, and cultured at 37℃、5% CO2 for 72 h under

434 sterile conditions. We then stored the culture supernatant after bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

435 centrifugation at -80℃.

436 Cell Isolation of Spleen and Culture in vitro

437 The mice, 42 days after infection, were killed by the cervical

438 dislocated method to isolate the spleen. The spleen passed through a

439 100-μm cell strainer to get tissue homogenate and was added red blood

440 cell lysate to decompose red cells repeatedly until there were no red blood

441 cells. Then a complete culture medium containing 5 μg /ml RMPI1640

442 was added to resuspend cells. The cell density under the microscope was

443 adjusted to 2×105-106/ml, then divided into culture plate or dish, and

444 cultured at 37℃、5%CO2 for 72h under sterile conditions. The culture

445 fluids centrifuged to remove supernatant were stored at-80℃ , The

446 single-cell suspension was detected by flow cytometry.

447 Flow Cytometry

448 For cell-surface marker and Treg cells analysis, CD4+ T cells were

449 isolated from 5×105-106/ml single-cell suspension using the CD4 (L3T4)

450 MicroBeads, which was permeabilized in PBS with 2% fetal bovine

451 serum, and then were stained on the surface. The cell membrane was

452 destroyed, and the nuclear transcription factor FOXP3 was stained

453 according to the instructions of the manufacturer. Compensation was

454 performed with BD LSRFortessa (BD Biosciences). The data were bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

455 acquired and analyzed with FlowJo (Tree Star) software.

456 RNA extraction and RT-PCR

457 We extracted the total RNA from 30 mg liver and spleen with

458 Trizol, based on the manual. The quality and quantity of RNA were then

459 evaluated by ultra-micro-nucleic acid protease analyzer. Total RNA (1μg)

460 was reverse transcribed with RevertAid First Strand cDNA Synthesis Kit

461 (K1622, Thermo Fisher Scientific). The expression of USP21, IL-10,

462 IL-17, FOXP3 in liver and spleen, and α-SMA, collagenⅠ, collagen Ⅲ in

463 the liver were detected by iTaq™ Universal SYBR® Green super mixture

464 (1725121, bio-rad) and fluorescence quantitative PCR 7500 system

465 (applied biosystems, USA), and quantitative analysis was carried out by

466 2-ΔΔCt method with glyceraldehyde-3-phosphate dehydrogenase (GAPDH)

467 as control. The PCR cycling conditions were as follows: 40 cycles of

468 95◦C for 30s; 95◦C for 5s; 60◦C for 34s; and then 95◦C for 15s; 60◦C

469 for 60s; 95◦C for 15s; and 60◦C for 60s. All primers were blasted to

470 NCBI to ensure their specificity. All primers were showed as Table 1

471

472

473

474

475

476 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

477 478 Table 1 Primers used for mRNA analysis 479

480

481 Genes Primer Sequence(5’_3’) 482 GAPDH Forward primer 483 ACTCCACTCACGGCAAATTC Reverse primer 484 TCTCCATGGTGGTGAAGACA USP21 Forward primer 485 ACCCAGGAAAGACAGCAACC 486 Reverse primer CTCGAAGACCTTCTCACAACC A 487 FOXP3 Forward primer GTGATTTTAATAAGCTCCAAG 488 ACCA Reverse primer GATCATCATGTATGCTTCTATG 489 CAG 490 IL-10 Forward primer TATCCCTCTGTGATCTGGGAA G 491 Reverse primer ATCTTCTCGACCCTGAAAGTG 492 A IL-17 Forward primer CACAGCCCTGGTGTGCGACAA 493 T 494 Reverse primer TTGCTCTGGGCTTCATCCCCCA α-SMA Forward primer TCCTGCGCCTAATGTCCACCG 495 A 496 Reverse primer AAGCGACTGTTGCCTTCGCCTC Collagen Ⅰ Forward primer TCCTGGTGGCAAGGGTGATCG 497 T 498 Reverse primer TGGAGCACCAGAAGGACCAGC A 499 Collagen Ⅲ Forward primer GCTCACCACACACTGCTTCT 500 Reverse primer GGATTCACAGCTTCACAGGA 501

502 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

503 Western Blotting

504 30-50mg liver from each mouse was lysed on ice for 3 to 4 hours

505 by the cold ripa lytic buffer(1-1.2ml) containing cocktail protease

506 inhibitor (B14001, bimake). After centrifuging 15min at 14000g/min, the

507 supernatant was moved to another tube to detected the total protein

508 concentration by BCA protein detection kit (P0010, Biyuntian). The

509 protein was boiled for 3 min and loaded onto 10% polyacrylamide gel,

510 electrophoresed, and transferred to PVDF membrane (MilliporeMA,

511 USA). The proteins were separated by SDS–gel electrophoresis. Protein

512 bands detected by incubation for12-16h at 4˚C with polyclonal primary

513 antibodies to GAPDH（ sc-32233, Santa Cruz） , USP21（ ab171028,

514 abcam）, α-SMA（#19245T, Cell Signaling Technology）, Collagen Ⅰ

515 （BA0325, Boster）, Collagen Ⅲ（sc-271249, Santa Cruz）followed

516 by blotting with secondary antibody HRP-Goat anti-Rabbit IgG

517 （ab97080, Abcam) , HRP-Rabbit anti-Mouse IgG（ab6728, Abcam）.

518 Immune complexes were visualized with WesternBrightTMECL substrate

519 (K-12045-D10, Advansta) and the luminescent signal recorded in the

520 chemiluminescence imaging system (Bio-Rad, USA) to detect the

521 expression of proteins. Quantitative analysis of the expression of USP21,

522 α-SMA, CollagenⅠ, Collagen Ⅲ was carried out by using ImageJ (NIH,

523 Bethesda, USA). bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

524 ELISA

525 The ELISA assay has been described previously[56]. Briefly, for

526 the S.j-SEA-, and S.j-SWAP-ELISAs, all recombinant antigens were

527 diluted to a final concentration of 1 μg/mL with coating buffer; for the

528 S.j-SEA-and S.j-SWAP-ELISAs, 50ng of each antigen was mixed per

529 well at 4˚C overnight with 100 μL added per well. After blocking by

530 blocking buffer (1% BAS in PBST) at 37˚C for one hour. Then serum

531 samples diluted at 1:250 with blocking buffer were added (100 μL/well)

532 and incubated for one hour at 37˚C. HRP-Goat anti-Mouse IgM

533 （ab97230, abcam）and HRP-Rabbit anti-Mouse IgG（ab6728, abcam）

534 were used as the secondary antibody (1:20,000, 100 μL/well), and

535 samples were incubated for one hour at 37˚C. Streptavidin-HRP (BD

536 Pharmingen, CA, USA) (1:10,000) was then applied to each well (100

537 μL/well). PBST washes were applied five times after each step, 2 min

538 between each wash. Reactions were developed using TMB as substrate

539 (100 μL/well) for 5 min and stopped using 2 M sodium hydroxide (50

540 μL/well). Optical density (OD) values were read at 450 nm using a

541 microplate reader, and all tests were run in duplicate on each test plate.

542 Cytokine detection (Multiplex Fluorescent Microsphere Immunoassay)

543 The mouse serum and cell culture samples were prepared

544 according to the manufacturer's instructions. 50µl pre-mixed beads were bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

545 added to each of 96 wells and then were incubated with the samples.

546 Detecting antibodies (IFN-γ, IL-4, IL-10, IL-17A, IL-23, IL-9) were

547 added after washing, followed by adding SA-PE after washing again. The

548 96-wells clean-dried plate was put into Bio-Plex 200 to detect. The

549 standard curve was fitted by a five-parameter non-linear regression

550 method, and the concentration was calculated. The results included the

551 label, the median of the fluorescence intensity and the concentration.

552 Statistical analyses

553 Statistical comparisons were performed with Prism 7.0 (GraphPad

554 Prism.) and SPSS18.0, using t-test for comparisons of two datasets, and

555 ANOVA for multiple comparisons. All the data were set for ͞x±S.D and

556 tests were considered statistically significant at p≤0.05. 557

558 Acknowledgment

559 We thank Dr. Adiele Onyeze for thorough reviews that greatly

560 improved the manuscript. 561 562 References 563 1. Ross AG, Bartley PB, Sleigh AC, Olds GR, Li Y, Williams GM, et al. 564 (2002) Schistosomiasis. N Engl J Med 346:1212-20. pmid: 11961151 565 2. Wilson MS, Mentink-Kane MM, Pesce JT, Ramalingam TR, Thompson R, Wynn 566 TA (2007) Immunopathology of schistosomiasis. Immunology and Cell 567 Biology 85:148-54. pmid: WOS:000245102500011 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

568 3. Yazdanbakhsh M, Sacks DL (2010) Why does immunity to parasites take 569 so long to develop? Nat Rev Immunol 10:80-1. pmid:WOS:000273876200001 570 4. Allen JE, Maizels RM (2011) Diversity and dialogue in immunity to 571 helminths. Nat Rev Immunol 11:375-88. pmid: 21610741 572 5. Nausch N, Bourke CD, Appleby LJ, Rujeni N, Lantz O, Trottein F, et 573 al. (2012) Proportions of CD4+ memory T cells are altered in 574 individuals chronically infected with Schistosoma haematobium. Sci 575 Rep 2:472. pmid: 22737405 576 6. McSorley HJ, Maizels RM (2012) Helminth infections and host immune 577 regulation. Clin Microbiol Rev 25:585-608. pmid: 23034321 578 7. Yang Q, Qu J, Jin C, Feng Y, Xie S, Zhu J, et al. (2019) Schistosoma 579 japonicum Infection Promotes the Response of Tfh Cells Through 580 Down-Regulation of Caspase-3-Mediating Apoptosis. Front Immunol 581 10:2154. pmid: 31572373 582 8. Johnson PT, Hoverman JT (2012) Parasite diversity and coinfection 583 determine pathogen infection success and host fitness. Proc Natl Acad 584 Sci U S A 109:9006-11. pmid: 22615371 585 9. Riner DK, Ferragine CE, Maynard SK, Davies SJ (2013) Regulation of 586 innate responses during pre-patent schistosome infection provides an 587 immune environment permissive for parasite development. PLoS Pathog 588 9:e1003708. pmid: 24130499 589 10. Zhan T, Ma H, Jiang S, Zhong Z, Wang X, Li C, et al. (2019) 590 Interleukin-9 blockage reduces early hepatic granuloma formation and 591 fibrosis during Schistosoma japonicum infection in mice. Immunology 592 158:296-303. pmid: 31436861 593 11. Liu J, Zhu L, Wang J, Qiu L, Chen Y, Davis RE, et al. (2019) Schistosoma 594 japonicum extracellular vesicle miRNA cargo regulates host 595 macrophage functions facilitating . PLoS Pathog 596 15:e1007817. pmid: 31163079 597 12. Colley DG, Secor WE (2014) Immunology of human schistosomiasis. 598 Parasite Immunol 36:347-57. pmid: 25142505 599 13. Burns LA, Maroof A, Marshall D, Steel KJA, Lalnunhlimi S, Cole S, et 600 al. (2019) Presence, function, and regulation of IL-17F-expressing 601 human CD4(+) T cells. Eur J Immunol. pmid: 31850514 602 14. Razawy W, Asmawidjaja PS, Mus AM, Salioska N, Davelaar N, Kops N, et 603 al. (2019) CD4(+) CCR6(+) T cells, but not gammadelta T cells, are 604 important for the IL-23R-dependent progression of antigen-induced 605 inflammatory arthritis in mice. Eur J Immunol. pmid: 31778214 606 15. Chen X, Yang X, Li Y, Zhu J, Zhou S, Xu Z, et al. (2014) Follicular 607 helper T cells promote liver pathology in mice during Schistosoma 608 japonicum infection. PLoS Pathog 10:e1004097. pmid: 24788758 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

609 16. Zaretsky AG, Taylor JJ, King IL, Marshall FA, Mohrs M, Pearce EJ (2009) 610 T follicular helper cells differentiate from Th2 cells in response 611 to helminth antigens. J Exp Med 206:991-9. pmid: WOS:000266010000006 612 17. Zhang YM, Wang YJ, Jiang YY, Pan W, Liu H, Yin JH, et al. (2016) T 613 follicular helper cells in patients with acute schistosomiasis. 614 Parasite Vector 9. pmid: WOS:000377338600001 615 18. Romano A, Hou X, Sertorio M, Dessein H, Cabantous S, Oliveira P, et 616 al. (2016) FOXP3+ Regulatory T Cells in Hepatic Fibrosis and 617 Splenomegaly Caused by Schistosoma japonicum: The Spleen May Be a 618 Major Source of Tregs in Subjects with Splenomegaly. PLoS Negl Trop 619 Dis 10:e0004306. pmid: 26731721 620 19. Khattri R, Cox T, Yasayko SA, Ramsdell F (2003) An essential role for 621 Scurfin in CD4(+)CD25(+) T regulatory cells. Nat Immunol 4:337-42. 622 pmid: WOS:000181987700012 623 20. Fontenot JD, Gavin MA, Rudensky AY (2003) Foxp3 programs the 624 development and function of CD4+CD25+ regulatory T cells. Nat Immunol 625 4:330-6. pmid: 12612578 626 21. Hori S, Nomura T, Sakaguchi S (2003) Control of regulatory T cell 627 development by the transcription factor Foxp3. Science 299:1057-61. 628 pmid: 12522256 629 22. Taylor MD, van der Werf N, Maizels RM (2012) T cells in helminth 630 infection: the regulators and the regulated. Trends Immunol 33:181-9. 631 pmid: 22398370 632 23. Turner JD, Jenkins GR, Hogg KG, Aynsley SA, Paveley RA, Cook PC, et 633 al. (2011) CD4+CD25+ regulatory cells contribute to the regulation 634 of colonic Th2 granulomatous pathology caused by schistosome 635 infection. PLoS Negl Trop Dis 5:e1269. pmid: 21858239 636 24. Zhou X, Bailey-Bucktrout SL, Jeker LT, Penaranda C, 637 Martinez-Llordella M, Ashby M, et al. (2009) Instability of the 638 transcription factor Foxp3 leads to the generation of pathogenic 639 memory T cells in vivo. Nat Immunol 10:1000-7. pmid: 19633673 640 25. Samstein RM, Arvey A, Josefowicz SZ, Peng X, Reynolds A, Sandstrom 641 R, et al. (2012) Foxp3 exploits a pre-existent enhancer landscape for 642 regulatory T cell lineage specification. Cell 151:153-66. pmid: 643 23021222 644 26. Li Y, Lu Y, Wang S, Han Z, Zhu F, Ni Y, et al. (2016) USP21 prevents 645 the generation of T-helper-1-like Treg cells. Nat Commun 7:13559. 646 pmid: 27857073 647 27. Wan YY, Flavell RA (2007) Regulatory T-cell functions are subverted 648 and converted owing to attenuated Foxp3 expression. Nature 445:766-70. 649 pmid: 17220876 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

650 28. Levescot A, Nelson-Maney N, Morris A, Bouyer RG, Lee P, Nigrovic P 651 (2017) Autoimmune Arthritis in IL-1 Receptor Antagonist-Deficient 652 Mice Is Associated with a Pathogenic Conversion of Foxp3+Regulatory 653 T Cells into Th17 Cells. Arthritis Rheumatol 69. pmid: 654 WOS:000411824106216 655 29. Dominguez-Villar M, Baecher-Allan CM, Hafler DA (2011) 656 Identification of T helper type 1-like, Foxp3(+) regulatory T cells 657 in human autoimmune disease. Nat Med 17:673-5. pmid: 658 WOS:000291308800026 659 30. Bailey-Bucktrout SL, Martinez-Llordella M, Zhou X, Anthony B, 660 Rosenthal W, Luche H, et al. (2013) Self-antigen-driven activation 661 induces instability of regulatory T cells during an inflammatory 662 autoimmune response. Immunity 39:949-62. pmid: 24238343 663 31. Zhang J, Chen C, Hou XX, Gao YY, Lin F, Yang J, et al. (2013) 664 Identification of the E3 Deubiquitinase Ubiquitin-specific Peptidase 665 21 (USP21) as a Positive Regulator of the Transcription Factor GATA3. 666 J Biol Chem 288:9373-82. pmid: WOS:000316862200054 667 32. Fan YH, Mao RF, Yu Y, Liu SF, Shi ZC, Cheng J, et al. (2014) USP21 668 negatively regulates antiviral response by acting as a RIG-I 669 deubiquitinase. J Exp Med 211:313-28. pmid: WOS:000331281600011 670 33. Chen Y, Wang L, Jin J, Luan Y, Chen C, Li Y, et al. (2017) p38 inhibition 671 provides anti-DNA virus immunity by regulation of USP21 672 phosphorylation and STING activation. J Exp Med 214:991-1010. pmid: 673 28254948 674 34. Nakagawa T, Kajitani T, Togo S, Masuko N, Ohdan H, Hishikawa Y, et 675 al. (2008) Deubiquitylation of histone H2A activates transcriptional 676 initiation via trans-histone cross-talk with H3K4 di- and 677 trimethylation. Genes Dev 22:37-49. pmid: 18172164 678 35. Tao LQ, Chen C, Song HH, Piccioni M, Shi G, Li B (2014) 679 Deubiquitination and stabilization of IL-33 by USP21. Int J Clin Exp 680 Patho 7:4930-7. pmid: WOS:000345120900036 681 36. Arceci A, Bonacci T, Wang X, Stewart K, Damrauer JS, Hoadley KA, et 682 al. (2019) FOXM1 Deubiquitination by USP21 Regulates Cell Cycle 683 Progression and Paclitaxel Sensitivity in Basal-like Breast Cancer. 684 Cell Rep 26:3076-86 e6. pmid: 30865895 685 37. Liu FW, Fu Q, Li YP, Zhang K, Tang MY, Jiang W, et al. (2019) USP21 686 modulates Goosecoid function through deubiquitination. Bioscience 687 Rep 39. pmid: WOS:000475404500001 688 38. Pannu J, Belle JI, Forster M, Duerr CU, Shen SY, Kane L, et al. (2015) 689 Ubiquitin Specific Protease 21 Is Dispensable for Normal Development, 690 Hematopoiesis and Lymphocyte Differentiation. Plos One 10. pmid: 691 WOS:000350682600061 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

692 39. Lacorcia M, Prazeres da Costa CU (2018) Maternal Schistosomiasis: 693 Immunomodulatory Effects With Lasting Impact on Allergy and Vaccine 694 Responses. Front Immunol 9:2960. pmid: 30619318 695 40. Tang CL, Zhang RH, Liu ZM, Jin H, He L (2019) Effect of regulatory 696 T cells on the efficacy of the fatty acid-binding protein vaccine 697 against Schistosoma japonicum. Parasitol Res 118:559-66. pmid: 698 30607606 699 41. Larkin BM, Smith PM, Ponichtera HE, Shainheit MG, Rutitzky LI, 700 Stadecker MJ (2012) Induction and regulation of pathogenic Th17 cell 701 responses in schistosomiasis. Semin Immunopathol 34:873-88. pmid: 702 23096253 703 42. McManus DP, Dunne DW, Sacko M, Utzinger J, Vennervald BJ, Zhou XN (2018) 704 Schistosomiasis. Nat Rev Dis Primers 4:13. pmid: 30093684 705 43. Amiri P, Locksley RM, Parslow TG, Sadick M, Rector E, Ritter D, et 706 al. (1992) Tumour necrosis factor alpha restores granulomas and 707 induces parasite egg-laying in schistosome-infected SCID mice. 708 Nature 356:604-7. pmid: 1560843 709 44. Kalantari P, Bunnell SC, Stadecker MJ (2019) The C-type Lectin 710 Receptor-Driven, Th17 Cell-Mediated Severe Pathology in 711 Schistosomiasis: Not All Immune Responses to Helminth Parasites Are 712 Th2 Dominated. Front Immunol 10:26. pmid: 30761125 713 45. Farwa A, He C, Xia L, Zhou H (2018) Immune modulation of Th1, Th2, 714 and T-reg transcriptional factors differing from cytokine levels in 715 Schistosoma japonicum infection. Parasitol Res 117:115-26. pmid: 716 29188369 717 46. Pearce EJ, Caspar P, Grzych JM, Lewis FA, Sher A (1991) Downregulation 718 of Th1 cytokine production accompanies induction of Th2 responses by 719 a parasitic helminth, . J Exp Med 173:159-66. pmid: 720 1824635 721 47. Stadecker MJ, Asahi H, Finger E, Hernandez HJ, Rutitzky LI, Sun J (2004) 722 The immunobiology of Th1 polarization in high-pathology 723 schistosomiasis. Immunol Rev 201:168-79. pmid: 15361240 724 48. El-Ahwany E, Bauiomy IR, Nagy F, Zalat R, Mahmoud O, Zada S (2012) 725 T regulatory cell responses to immunization with a soluble egg antigen 726 in Schistosoma mansoni-infected mice. Korean J Parasitol 50:29-35. 727 pmid: 22451731 728 49. Rutitzky LI, Stadecker MJ (2006) CD4 T cells producing 729 pro-inflammatory interleukin-17 mediate high pathology in 730 schistosomiasis. Mem Inst Oswaldo Cruz 101 Suppl 1:327-30. pmid: 731 17308791 732 50. Wen XY, He L, Chi Y, Zhou S, Hoellwarth J, Zhang C, et al. (2011) 733 Dynamics of Th17 Cells and Their Role in Schistosoma japonicum bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

734 Infection in C57BL/6 Mice. Plos Neglect Trop D 5. pmid: 735 WOS:000298134000027 736 51. Yang Q, Qiu HN, Xie HY, Qi YW, Cha HF, Qu JL, et al. (2017) A Schistosoma 737 japonicum Infection Promotes the Expansion of Myeloid-Derived 738 Suppressor Cells by Activating the JAK/STAT3 Pathway. Journal of 739 Immunology 198:4716-27. pmid: WOS:000405271300018 740 52. Negrao-Correa D, Fittipaldi JF, Lambertucci JR, Teixeira MM, Antunes 741 CMD, Carneiro M (2014) Association of Schistosoma mansoni-Specific 742 IgG and IgE Antibody Production and Clinical Schistosomiasis Status 743 in a Rural Area of Minas Gerais, Brazil. Plos One 9. pmid: 744 WOS:000336971300043 745 53. Wynn TA, Cheever AW, Williams ME, Hieny S, Caspar P, Kuhn R, et al. 746 (1998) IL-10 regulates liver pathology in acute murine 747 Schistosomiasis mansoni but is not required for immune 748 down-modulation of chronic disease. J Immunol 160:4473-80. pmid: 749 9574553 750 54. Maizels RM, Balic A, Gomez-Escobar N, Nair M, Taylor MD, Allen JE (2004) 751 Helminth parasites - masters of regulation. Immunological Reviews 752 201:89-116. pmid: WOS:000223819600008 753 55. Baumgart M, Tompkins F, Leng J, Hesse M (2006) Naturally occurring 754 CD4+Foxp3+ regulatory T cells are an essential, IL-10-independent 755 part of the immunoregulatory network in Schistosoma mansoni 756 egg-induced inflammation. J Immunol 176:5374-87. Epub 2006/04/20. 757 pmid: 16622005 758 56. Cai P, Weerakoon KG, Mu Y, Olveda DU, Piao X, Liu S, et al. (2017) 759 A Parallel Comparison of Antigen Candidates for Development of an 760 Optimized Serological Diagnosis of Schistosomiasis Japonica in the 761 Philippines. EBioMedicine 24:237-46. Epub 2017/09/26. pmid: 28943229 762 763 764 765

766 Figure Legends 767 FIGURE 1. Depletion of USP21 in Treg cells weakens resistance to 768 Schistosoma japonicum in infected mice 769 (A-E) FOXP3Cre(WT) and USP21fl/flFOXP3Cre(KO) mice(n=26 ±2/group)

770 were infected with 252 cercariaes through the abdomen, and were

771 sacrificed at 6-7 weeks after infection. Liver and spleen tissues were bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

772 collected and the adults were recovered. (A) Comparision of the number

773 of females, males , hugs and adults recovered from WT and KO infected

774 INF groups (n=mean ±SD), *p=0.01in the total number of adults，and

775 **p=0.036 in total number of hugs.(B) Representative images of the liver

776 of WT and KO, including NC and INF group. (C) Liver weight of WT

777 and KO including NC and INF group(weight=mean ±SD), p＞0.05 was

778 considered no statistically significant between the normal control group

779 and the infected group. .(D) Representative images of the spleen of WT

780 and KO, including NC and INF group. (E) Spleen weight of WT and KO

781 including NC and INF group,(weight=mean ±SD), **p=0.007 between the

782 normal control group , p＞0.05 represented no statistical significance in

783 the infected group.

784 785 FIGURE 2. Effect of USP21 on Schistosoma Japonicum eggs in 786 infectious KO mice 787 (A-C) The liver was taken for HE staining and the number of

788 eggs was counted under the microscope on the 42nd day of the

789 schistosomiasis infected mouse(n=10/group). (A) The statistical

790 graph of liver eggs in WT-INF group and KO-INF group (the data

791 expressed as mean ±SD, *p=0.043). (B) Representative images of

792 liver H&E staining in the WT-INF and KO-INF groups (original

793 magnification: ×100, insertion: ×40).The arrows indicated the bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

794 granuloma. (C) Comparison of liver egg granuloma area between

795 WT-INF and KO-INFgroup. The data was expressed as mean

796 ±SD,*p=0.022.

797 798 FIGURE 3. Changes of liver Fibrosis in USP21fl/flFOXP3cre mice 799 infected with Schistosoma japonicum 800 (A-E) The liver was taken and masson stained on the 42nd day of

801 the schistosoma infected mouse(n=10/goup).（A）Representative

802 images of collagen deposition in WT-INF and KO-INF mice

803 (original magnification: ×100) were obtained. The blue areas

804 were collagen granules.（ B） Comparison of the percentage of

805 hepatic fibrosis area between WT-INF and KO-INF group. The

806 data was expressed as mean ±SD,*p=0.039.（C）The mRNA level

807 comparison of α-SMA, collagen-Ⅰ and collagen-Ⅲ in WT and KO

808 including NC and INF group , the data representation for mean±SD. The

809 mRNA level comparison of α-SMA, collagen-Ⅰ and collagen-Ⅲ

810 between infectious groups, the data expresed as mean±SD. In

811 infection groupα -SMA:**p<0.001; collagen-Ⅰ :**p=0.008,

812 collagen-Ⅲ :**p=0.006. (D） Representative images of SMA,

813 collagen-Ⅰ and collagen-Ⅲ protein expression by western blot in WT

814 and KO INF including NC and INF group（E）Comparison of protein

815 expression between WT and KO including NC and INF group , the data bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

816 expressed as mean ±SD. In infection groupα -SMA :

817 *p=0.034 ,collagen-Ⅰ:**p=0.006 ,collagen - Ⅲ:*p=0.03.

818 819 FIGURE 4. Changes of spleen immune in USP21fl/flFOXP3cre mice 820 infected with Schistosoma Japonicum 821 (A-D)Spleen was collected and splenic cells were cultured in

822 vitro on the 42nd day in schistosomiasis infected

823 mice(n=6/group).（ A） CD4+CD25+FOXP3high cells were determined

824 by flow cytometry in WT and KO mice infected with schistosoma

825 japonicum after 42 days, including NC and INF group, .（ B） The

826 comparison of the portion of CD4+CD25+FOXP3high cells in CD4+T cells

827 in INF of WT and KO, including NC and INF group. The data was

828 expressed as mean ±SD, infection groups:**p=0.007.（C）The mRNA

829 levels of FOXP3, IL-10, IL-17 and USP21 in WT and KO including NC

830 and INF group, were compared, and the data was expressed as mean ±SD.

831 Between infection group FOXP3 **p=0.001,IL-10 **p<0.001,IL-17

832 **p<0.001, USP21 **p<0.001.（D）The contents of IFN-gamma, IL-4,

833 IL-10, IL-17A, IL-23 and IL-9 in cultured spleen cells of WT and KO

834 including NC and INF group were compared, and the data was expressed

835 as mean ±SD. In NC group IFN-gamma **p=0.002， IL-4 *p=0.002;

836 between infection groups IFN-gamma *p=0.013， IL-4 **p=0.005，

837 IL-10**p=0.004. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

838 839 FIGURE 5. Liver immunity in USP21fl/flFOXP3cre mice infected with 840 Schistosoma Japonicum 841 （A） Representative maps of USP21 protein expression were

842 determined by Western blot in WT and KO including NC and INF

843 group(n=6/group).（B）Protein expression of WT and KO, including

844 NC and INF group(n=6/group) , was compared. The data was expressed

845 as mean ±SD, and the comparison between the infection groups

846 was**p<0.001.（C）Comparison of USP21 mRNA levels between WT

847 and KO, including NC and INF group(n=6/group), was presented as

848 mean ±SD, **p=0.001 between the infection groups.（D）The

849 mRNA levels of FOXP3,IL-10, and IL-17 in WT and KO, including NC

850 group and INF group(n=6/group), were compared, and the data was

851 expressed as mean ±SD, between the infection groups IL-10

852 **p=0.003 , IL-17 *p=0.026 853

854 FIGURE 6. Specific antibody response in USP21fl/flFOXP3Cre mice

855 infected with Schistosoma japonicum

856 （A-D）The serum from WT and KO including NC and INF group(n=26

857 ±2/group) at different stages (uninfected, day 20, day 30, day 42). （A）

858 The changes of anti-SEA IgG content were compared, and the data was

859 expressed as mean ±SD. Between infection group*p=0.01 on day

860 20, **p＜0.001 on day 30, and**p＜0.001 on day 42.（B） bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

861 Comparison of anti-SEA IgM content was made, the data was expressed

862 as mean ±SD, and **p=0.002 on day 42 between the infection

863 groups.（C）Comparison of anti-SWAP IgG contents was taken to be

864 expressed as mean ±SD, between infection groups**p＜0.001 on day

865 20, **p=0.003 on day 30, and **p＜0.001 on day 42.（D）The

866 content of anti- SWAP IgM was compared, and the data was

867 expressed as mean ±SD. Between infection groups **p＜0.001 on day

868 20, **p=0.004 on day 30,**p＜0.001 on day 42.

869

870 FIGURE 7. Serum cytokines in USP21fl/flFOXP3Cre mice infected 871 with Schistosoma japonicum

872 （A）The contents of IFN-gamma, IL-4, IL-10, IL-17A, IL-23 and IL-9

873 in isolated peripheral blood lymphocyte cultures on day 42 of WT and

874 KO, including NC and INF group(n=10/group), were compared, and the

875 data was expressed as mean ±SD, IL-4 **p=0.008,IL-23

876 *p=0.035between the infection groups.（B）Contents of serum

877 IFN-gamma, IL-4, IL-10,IL-17A, IL-23 and IL-9 was compared in INF of

878 WT and KO infection groups（n=26±2/group）at different periods

879 (uninfected, 20th day of infection, 30th day of infection and 42nd day of

880 infection), and the data was expressed as mean ±SD. IFN-γ *p=0.024

881 on day 20 and **p＜0.001 on day 42; IL-10 **p＜0.001 on day 20

882 and *p=0.017 on day 30; IL-17A *p=0.026 on day 20, *p＜0.001 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

883 on day 30 and **p＜0.001 on day 42;IL-23 **p＜0.001 on day

884 30 ; IL-9 **p＜0.001 on day 42. 885 886 bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license. bioRxiv preprint doi: https://doi.org/10.1101/2020.03.09.983502; this version posted March 9, 2020. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.