Gluconeogenic Enzyme PCK1 Deficiency Is Critical for CHK2 O-Glcnacylation

bioRxiv preprint doi: https://doi.org/10.1101/2020.08.06.240739; this version posted August 6, 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 The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.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-NC-ND 4.0 International license.

1 Gluconeogenic Enzyme PCK1 Deficiency Is Critical for CHK2 O-GlcNAcylation

2 and Hepatocellular Carcinoma Growth upon Glucose Deprivation

4 Jin Xiang1,#, Chang Chen2,#, Rui Liu1,#, Dongmei Gou1,#, Lei Chang3, Haijun Deng1,

5 Qingzhu Gao1, Wanjun Zhang3, Lin Tuo4, Xuanming Pan1, Li Liang1, Jie Xia1, Luyi

6 Huang1, Ailong Huang1,*, Kai Wang1,*, Ni Tang1,*

8 1Key Laboratory of Molecular Biology for Infectious Diseases (Ministry of Education),

9 Institute for Viral Hepatitis, Department of Infectious Diseases, The Second Affiliated

10 Hospital, Chongqing Medical University, Chongqing, China

11 2Institute of Life Sciences, Chongqing Medical University, Chongqing, China

12 3State Key Laboratory of Proteomics, Beijing Proteome Research Center, National

13 Center for Protein Sciences (Beijing), Beijing Institute of Lifeomics, Beijing, China

14 4Sichuan Provincial People’s Hospital, Sichuan, China

16 *CORRESPONDENCE

17 Ni Tang, Kai Wang, Ailong Huang, Key Laboratory of Molecular Biology for Infectious

18 Diseases (Ministry of Education), Institute for Viral Hepatitis, Department of Infectious

19 Diseases, The Second Affiliated Hospital, Chongqing Medical University, Chongqing,

20 China. Phone: 86-23-68486780, Fax: 86-23-68486780, E-mail: [email protected]

21 (N.T.), [email protected] (K.W.), [email protected] (A.L.H.)

22 #These authors contributed equally to this work.

23 1

24 ABSTRACT

25 Elevated hexosamine-biosynthesis pathway (HBP) activity and O-GlcNAcylation are

26 emerging hallmarks of hepatocellular carcinoma (HCC). Inhibiting O-GlcNAcylation

27 could be a promising anti-cancer strategy. Here, we investigate this possibility and

28 demonstrate that deficiency of phosphoenolpyruvate carboxykinase 1 (PCK1), a

29 rate-limiting enzyme in gluconeogenesis, promotes O-GlcNAcylation and hepatoma

30 cell proliferation under low-glucose conditions. PCK1 loss results in oxaloacetate

31 accumulation and AMPK inactivation, promoting uridine

32 diphosphate-N-acetylglucosamine (UDP-GlcNAc) synthesis and CHK2 threonine 378

33 O-GlcNAcylation and counteracting its ubiquitination and degradation.

34 O-GlcNAcylation also promotes CHK2-dependent Rb phosphorylation and HCC cell

35 proliferation. Therefore, blocking HBP-mediated O-GlcNAcylation suppresses tumor

36 progression in liver-specific Pck1-knockout mice. We reveal a link between PCK1

37 depletion and hyper-O-GlcNAcylation that underlies HCC oncogenesis and suggest

38 therapeutic targets for HCC that act by inhibiting O-GlcNAcylation.

40 INTRODUCTION

41 Gaining insight into the fundamental role of metabolic reprogramming in cancer has

42 contributed immensely to our understanding of tumorigenesis and cancer progression

43 1. Nutrient limitations (such as glucose deprivation) in solid tumors may require cancer

44 cells to exhibit the metabolic flexibility required to sustain proliferation and survival 2.

45 Increased aerobic glycolysis (the Warburg effect) is one of main characteristics of

46 cancer cells that supports their intensive growth and proliferation 3. Gluconeogenesis

47 (the pathway opposite to glycolysis) operates during starvation and occurs mainly in

48 the liver, and also plays key roles in metabolic reprogramming, cancer cell plasticity,

49 and tumor growth 4,5. Several key gluconeogenic enzymes, such as

50 phosphoenolpyruvate carboxykinase 1 (PCK1, also known as PEPCK-C),

51 fructose-1,6-bisphosphatase, and glucose-6-phosphatase, were previously found to

52 be dysregulated in several types of cancer, including hepatocellular carcinoma (HCC)

53 6.

55 The cytoplasmic isoform of PCK1, the first rate-limiting enzyme in hepatic

56 gluconeogenesis, catalyzes the conversion of oxaloacetate (OAA) to

57 phosphoenolpyruvate (PEP). The oncogenic or tumor suppressor roles of PCK1 in

58 different types of human cancers are rather controversial. PCK1 has anti-tumorigenic

59 effects in gluconeogenic organs (liver and kidney), but has tumor-promoting effects in

60 cancers arising from non-gluconeogenic organs 4. In colon-derived tumor cells, PCK1

61 is hijacked to participate in truncated gluconeogenesis to meet biosynthetic and

62 anabolic needs 7. However, the underlying mechanism determining its aberrant

63 expression and altered function in multiple types of tumors remains incompletely

64 understood.

66 Recent findings emphasize the role of the hexosamine-biosynthesis pathway (HBP), a

67 sub-branch of glucose metabolism, in carcinogenesis 8–10. The HBP and glycolysis

68 share the first two steps and diverge at fructose-6-phosphate (F6P).

69 Glutamine-fructose-6-phosphate aminotransferase 1 (GFAT1), the rate-limiting

70 enzyme of the HBP, converts F6P and glutamine to glucosamine-6-phosphate and

71 glutamate. Uridine diphosphate N-acetylglucosamine (UDP-GlcNAc), the end

72 products of HBP, is a donor substrate for O-linked β-N-acetylglucosamine (O-GlcNAc)

73 modification (also known as O-GlcNAcylation) 11,12. O-GlcNAc transferase

74 (OGT)-mediated protein O-GlcNAcylation is highly dependent on the intracellular

75 concentration of the donor substrate UDP-GlcNAc, which is proposed to be a nutrient

76 sensor that couples metabolic and signaling pathways 13,14. Increased glucose flux

77 through the HBP and elevated UDP-GlcNAc contribute to hyper-O-GlcNAcylation in

78 cancer cells 15. Previous data suggested that elevated O-GlcNAcylation may serve as

79 a hallmark of cancer 16.

81 Similar to phosphorylation, O-GlcNAcylation is a dynamic post-translational

82 modification that regulates protein subcellular localization, stability, protein–protein

83 interactions, or enzymatic activity according to the nutrient demands of cells 17. OGT

84 and O-GlcNAcase (OGA) are the only enzymes known to be responsible for adding

85 and removing N-acetylglucosamine (GlcNAc) on serine and threonine residues of

86 target proteins 18. Numerous oncogenic factors, including c-MYC, Snail, and β-catenin,

87 are targets of O-GlcNAcylation 19–21. Advances in mass spectrometry (MS)-based

88 methods will facilitate the discovery of novel O-GlcNAc-modified proteins in cancer

89 cells. Therefore, modulating the HBP or O-GlcNAcylation (which regulate oncogenic

90 activation in human cancers) represents a promising anti-cancer strategy that can

91 potentially be used in combination with other treatments 22,23.

93 In this study, we explored the role of the gluconeogenic enzyme PCK1 in regulating

94 the HBP and HCC proliferation under low-glucose conditions. We unravel a molecular

95 mechanism responsible for enhanced UDP-GlcNAc biosynthesis and

96 O-GlcNAcylation induced by PCK1 depletion, and delineate the functional importance

97 of checkpoint kinase 2 (CHK2) O-GlcNAcylation in HCC tumorigenesis. Importantly,

98 our study reveals a novel link between the gluconeogenic enzyme PCK1 and

99 HBP-mediated O-GlcNAc modification, which suggest a therapeutic strategy for

100 treating HCC.

101

102 RESULTS

103 PCK1 Deficiency Increases Global O-GlcNAcylation in Hepatoma Cells under

104 Low-Glucose Conditions and Promotes HCC Proliferation

105 To explore the role of the gluconeogenic enzyme PCK1 in O-GlcNAcylation, we first

106 analyzed the global O-GlcNAcylation levels of hepatoma cells in response to PCK1

107 modulation with various concentrations of glucose (1 to 25 mM, 12 h). In the presence

108 of 5 mM glucose, PCK1 knockout significantly elevated the global O-GlcNAcylation

109 levels (Fig. 1a,b), whereas overexpression of wild-type (WT) PCK1 markedly

110 decreased the global O-GlcNAcylation levels in SK-Hep1, Huh7, and MHCC-97H

111 cells (Fig. 1c and Extended Data Fig. 1a-c). Interestingly, the catalytically inactive

112 G309R mutant of PCK1 24 was unable to reduce the O-GlcNAcylation levels in these

113 hepatoma cells under the same culture conditions (Fig. 1c and Extended Data Fig.

114 1b-c), suggesting that the enzymatic activity of PCK1 may contribute to its role in

115 regulating cellular O-GlcNAcylation levels. Further, 3-mercaptopicolinic acid (3-MPA),

116 a specific inhibitor of PCK1, dose-dependently increased the O-GlcNAcylation levels

117 in PLC/PRF/5 cells (Fig. 1d), whereas dexamethasone (Dex), an activator of

118 gluconeogenesis, lowered the O-GlcNAcylation levels in SK-Hep1 cells (Fig. 1e). In

119 addition, this observation was confirmed via pharmacological or transcriptional

120 inhibition of OGT and OGA (Fig. 1f-i), suggesting PCK1 inhibits OGT-mediated

121 O-GlcNAcylation. Interestingly, PCK1 did not change the mRNA or protein expression

122 levels of OGT, OGA, and GFAT1, the key enzymes involved in regulating

123 O-GlcNAcylation and HBP (Extended Data Fig. 1d-m and Fig. 1d,e). In addition, our

124 cell-proliferation assays indicated that PCK1 suppresses hepatoma cell proliferation,

125 depending on its enzymatic activity and the cellular O-GlcNAcylation levels (Extended

126 Data Fig. 2).

127

128 Additionally, we used an N-nitrosodiethylamine (DEN) and CCl4-induced mouse HCC

129 model to verify the above results in vivo. The O-GlcNAcylation levels in the hepatic

130 tumors of liver-specific Pck1 knockout (LKO) mice were significantly higher than those

131 in WT mice after 12 h of fasting (Fig. 1j,k). Furthermore, using an orthotopic HCC

132 mouse model, we found that PCK1 overexpression or dexamethasone administration

133 decreased the O-GlcNAcylation levels and inhibited HCC growth (Extended Data Fig.

134 3). Together, these data demonstrate that PCK1 decreases the global

135 O-GlcNAcylation levels in HCC under a low-glucose condition and suppresses

136 hepatoma cell proliferation, both in vitro and in vivo. The enzymatic activity of PCK1 is

137 indispensable for its tumor suppressor role in HCC.

138

139 PCK1 Deficiency Promotes UDP-GlcNAc Biosynthesis via Oxaloacetate

140 Accumulation

141 Next, a metabolomics assay was performed with AdPCK1- and AdGFP-infected

142 SK-Hep1 cells to explore metabolic changes occurring after PCK1 overexpression

143 (PCK1-OE) under a low glucose concentration (5 mM). Principal component analysis

144 showed that PCK1 overexpression dramatically changed the intracellular metabolic

145 profile of SK-Hep1 cells (Extended Data Fig. 4a). Levels of several metabolites of the

146 HBP decreased after PCK1 overexpression, including fructose 6-phosphate, N-acetyl

147 glucosamine 1- phosphate (GlcNAc-1-P), and UDP-GlcNAc (the HBP end product),

148 as shown in Fig. 2a,b and Extended Data Fig. 4b. Our targeted liquid

149 chromatography-tandem MS (LC-MS/MS) results showed that UDP-GlcNAc

150 significantly decreased in PCK1-OE SK-Hep1 cells (Fig. 2c), but increased in

151 PCK1-KO (PKO) cells (Fig. 2d), suggesting that PCK1 may negatively regulate

152 UDP-GlcNAc biosynthesis for O-GlcNAcylation. To investigate how PCK1 modulates

153 UDP-GlcNAc biosynthesis, we performed the pathway-enrichment analysis of

154 metabolite profiles and found that several metabolic pathways were significantly

155 affected, including purine and pyrimidine metabolism which required for uridine

156 triphosphate (UTP) synthesis (Extended Data Fig. 4c). The HBP utilizes glucose,

157 glutamine, acetyl-CoA, and UTP to produce the amino sugar UDP-GlcNAc (Fig. 2e),

158 we assumed that PCK1 may regulate UDP-GlcNAc biosynthesis partially via UTP

159 synthesis. Indeed, metabolomics data showed that levels of several metabolites

160 involved in UTP synthesis including aspartate (Asp, critical metabolite in de novo UTP

161 synthesis) declined upon PCK1 overexpression (Fig. 2f), besides metabolites in

162 glycolysis and the tricarboxylic acid (TCA) cycle (Fig. 2a and Extended Data Fig.

163 4d–g), suggesting that PCK1 may play role in TCA cataplerosis and UTP synthesis

164 which contribute to UDP-GlcNAc biosynthesis.

165

166 PCK1 catalyzes the conversion of OAA (an intermediate of the TCA cycle) to PEP.

167 Consistent with the results of a previous study 7, restoring PCK1 decreased the OAA

168 concentration in SK-Hep1 cells (Fig. 2g). Considering that OAA is converted to Asp by

169 the mitochondrial glutamate-oxaloacetate transaminase (GOT2, also known as

170 aspartate aminotransferase 2; Fig. 2h), we speculated that PCK1 repression may

171 promote UDP-GlcNAc biosynthesis through Asp conservation caused by OAA

172 accumulation. As expected, adding OAA strengthened UDP-GlcNAc biosynthesis (Fig.

173 2i). Accordingly, both OAA and Asp enhanced O-GlcNAcylation in SK-Hep1 cells,

174 while high concentrations of PEP (above 0.5 mM) showed an opposite effect (Fig. 2j).

175 Next, we tested whether GOT2 can contribute to HBP-mediated O-GlcNAcylation in

176 PKO cells. We found that treatment with aminooxyacetic acid (AOA), a specific

177 inhibitor of GOT2, decreased the UDP-GlcNAc levels in PKO cells (Fig. 2k). Moreover,

178 AOA treatment or short-hairpin RNA (shRNA)-mediated GOT2 knockdown reduced

179 the O-GlcNAcylation levels in PKO cells (Fig. 2l,m). In addition, AOA markedly

180 blocked O-GlcNAcylation induced by OAA treatment, but failed to moderate the effect

181 of Asp, indicating that GOT2 plays an essential role in the metabolism of OAA

182 converted to Asp, the HBP, and tumorigenesis (Fig. 2n,o and Extended Data Fig. 4h).

183 Taken together, these data provide strong evidence supporting that OAA

184 accumulation and GOT2-mediated pathway contribute to enhanced UDP-GlcNAc

185 biosynthesis and hyper-O-GlcNAcylation in PKO cells.

186

187 Restoration of PCK1 Suppresses O-GlcNAcylation by Activating the

188 AMPK-GFAT1 Axis

189 The final rate-limiting step of UDP-GlcNAc synthesis involves UTP and GlcNAc-1-P

190 (Fig. 2e). Our metabolomics analysis showed that both UTP and GlcNAc-1-P levels

191 were significantly decreased in PCK1-OE cells (Fig. 2b,f). Since OAA accumulation

192 contributes to the downstream UTP increase and GFAT1 is the rate-limiting enzyme in

193 GlcNAc-1-P synthesis, we then explored whether GFAT1 activity also regulates

194 UDP-GlcNAc production. Previously, we reported that PCK1 activates AMP-activated

195 protein kinase (AMPK) upon glucose deprivation in HCC 25, and other groups showed

196 that AMPK activation reduces O-GlcNAcylation through GFAT1 phosphorylation

197 (which diminished GFAT1 activity) in endothelial cells and cardiac hypertrophy 26,27.

198 We speculated that PCK1 may also inhibit UDP-GlcNAc synthesis through the

199 AMPK-GFAT1 axis. Thus, we tested whether PCK1 can suppress the HBP through

200 the AMPK-GFAT1 axis under low-glucose conditions. As expected, PCK1

201 overexpression promoted the phosphorylation of both AMPK and GFAT1 (Fig. 3a),

202 whereas PCK1-KO downregulated p-AMPK and p-GFAT1 production (Fig. 3b). The

203 AMPK activator metformin partially offset hyper-O-GlcNAcylation and

204 growth-promoting effects mediated by PCK1 depletion (Fig. 3c). However,

205 shRNA-mediated knockdown of AMPK in PCK1-OE cells rescued the inhibitory

206 effects of PCK1 on O-GlcNAcylation (Fig. 3d).

207

208 Furthermore, we investigated whether the inhibition of hepatoma cell proliferation in

209 response to PCK1 depended on the AMPK-GFAT1 axis. We found that metformin

210 suppressed PKO cell proliferation (Fig. 3e,f). In contrast, shRNA against AMPK

211 mRNA (shAMPK) promoted PCK-OE cell proliferation (Fig. 3g,h). In addition, the

212 GFAT1 inhibitor 6-diazo-5-oxo-l-norleucine (DON) reduced UDP-GlcNAc biosynthesis

213 (Fig. 3i), O-GlcNAcylation levels (Fig. 3j), and PKO cell proliferation (Fig. 3e,f). These

214 data indicate that PCK1 suppresses HBP-mediated O-GlcNAcylation and HCC

215 proliferation partially via activation of the AMPK-GFAT1 axis. PCK1 deficiency boosts

216 flux through the HBP and results in an increased availability of UDP-GlcNAc for

217 O-GlcNAcylation. Therefore, both OAA accumulation and the AMPK-GFAT1 axis

218 contributed to hyper-O-GlcNAcylation and PKO cell proliferation upon glucose

219 deprivation (Fig. 3k).

220

221 OGT mediates CHK2 O-GlcNAcylation in PCK1-Deficient Hepatoma Cells

222 To further explore how OGT-mediated protein O-GlcNAcylation facilitates hepatoma

223 cell proliferation in PKO cells, we used immunoprecipitation coupled with tandem MS

224 (IP-MS/MS) to screen for proteins that specifically interact with OGT (Fig. 4a).

225 Flag-tagged OGT was transiently expressed in PKO cells, and subsequent IP-MS

226 identified 617 candidate OGT-binding proteins (Supplemental Table 1). Gene

227 Ontology-based biological-process analysis indicated that several proteins were

228 involved in metabolic processes, apoptotic processes, and cell-cycle progression (Fig.

229 4b). We then focused on CHK2, which is required for checkpoint-mediated cell cycle

230 arrest 28. Interactions between OGT and CHK2 were confirmed by

231 co-immunoprecipitation (co-IP) experiments in HEK293 cells (Fig. 4c,d) and PKO

232 cells (Fig. 4e). Confocal analysis also indicated that OGT and CHK2 co-localized in

233 the nucleus (Fig. 4f). To define the precise region(s) in CHK2 required for this

234 interaction, we expressed full-length HA-tagged OGT in combination with different

235 Flag-tagged fragments of CHK2 in HEK293 cells (Fig. 4g). The C-terminal region of

236 CHK2 (amino acids 69–543) containing kinase domains showed a strong interaction,

237 whereas the N-terminal region (amino acids 1–175) did not interact with OGT (Fig.

238 4h).

239

240 Then, we determined whether CHK2 can be modified via O-GlcNAc.

241 Immunoprecipitated, Flag-tagged CHK2 exhibited a strong O-GlcNAc signal in

242 HKE293 cells (Fig. 4i). Endogenous CHK2 O-GlcNAcylation was confirmed by affinity

243 chromatography in presence of the OGA inhibitor PUGNAc, using succinylated wheat

244 germ agglutinin (sWGA), a modified lectin that specifically binds O-GlcNAc-containing

245 proteins (Fig. 4j). Furthermore, PCK1-OE and the OGT inhibitor ST045849 (ST)

246 decreased CHK2 O-GlcNAcylation (Fig. 4k,l), while PCK1-KO, the OGA inhibitor

247 PUGNAc, or OAA treatment strengthened CHK2 O-GlcNAcylation under low-glucose

248 conditions (Fig. 4l,m).

249

250 Next, we sought to map the O-GlcNAcylation site(s) on CHK2. Flag-tagged CHK2 was

251 purified from PKO cells and analyzed by MS. As shown in Fig. 4n, threonine 378

252 (T378) was the main O-GlcNAcylation site on CHK2. We then generated site-specific

253 point mutants of CHK2. Mutating T378 to Ala (T378A) largely reduced the O-GlcNAc

254 signal compared with WT CHK2 and the T383A mutant control (Fig. 4o), indicating

255 that T378 is the major CHK2 site carrying the O-GlcNAc modification. The potentially

256 O-GlcNAcylated residue T378 and the surrounding amino acids are highly conserved

257 among vertebrates (Fig. 4p), indicating that it serves an evolutionarily conserved role

258 regulating the CHK2 protein. Taken together, these data indicate that CHK2 interacts

259 with and can be O-GlcNAcylated by OGT in PKO cells.

260

261 O-GlcNAcylation on T378 Stabilizes CHK2 and Promotes Hepatoma Cell

262 Proliferation

263 To examine the effect of O-GlcNAcylation on CHK2 under the low-glucose conditions,

264 Flag-tagged WT and T378A CHK2 were overexpressed with HA-tagged ubiquitin

265 (HA-Ub) in PCK1-KO and parental PLC/PRF/5 cells. WT CHK2 ubiquitination was

266 alleviated in PKO cells compared with parental cells, whereas the T378A mutation or

267 OGT inhibitor ST045849 enhanced CHK2 ubiquitination (Fig. 5a). We then performed

268 a series of cycloheximide-chase experiments to assess the half-life of these proteins.

269 Endogenous CHK2 was more stable, with half-life of over 24h, in PLC/PRF/5 cell

270 treated with PUGNAc, indicating CHK2 O-GlcNAcylation may enhance its stability

271 (Extended Data Fig. 5a,b). In comparison with the parental cells, the CHK2 half-life

272 was prolonged in PKO cells, but the T378A mutation or ST045849 treatment reduced

273 CHK2 half-life from 24 h to 12 h (Fig. 5b-e). In addition, overexpression of WT PCK1

274 promoted CHK2 ubiquitination and degradation (Extended Data Fig. 5c-g). As

275 expected, the G309R PCK1 mutation did not affect the half-life of CHK2 (Extended

276 Data Fig. 5d-g). These results suggested that O-GlcNAc modification of T378

277 stabilizes CHK2 by preventing its ubiquitination and degradation in PCK1-deficient

278 hepatoma cells.

279

280 Given that CHK2 dimerization is essential for its activation 29, we next detected CHK2

281 dimerization in cells co-transfected with vectors encoding Flag-CHK2 and Myc-CHK2.

282 Our results indicated that PKO cells displayed strengthened CHK2 dimer formation,

283 whereas ST045849 treatment weakened this association (Extended Data Fig. 5h). A

284 similar result was observed by crosslinking analysis (Extended Data Fig. 5i),

285 suggesting O-GlcNAcylation of CHK2 may promote its dimerization. Interestingly,

286 T378, the autophosphorylation site of CHK2, is located in the dimerization interface 29.

287 We then performed dimeric CHK2 homology modeling, followed by molecular

288 dynamic (MD) simulation. Our model disclosed that the O-GlcNAcylated residue T378

289 interacts with the amino acids VSLK of another CHK2 kinase domain. The

290 acetylglucosamine group occupies a cavity which locates in the edge of interaction

291 interface and forms three hydrogen bonds with the backbone of VSLK motif, thus

292 might strengthen the stability of the CHK2 dimer (Extended Data Fig. 5j-l). Since

293 dimerization promotes CHK2 activation and phosphorylates its downstream targets,

294 such as retinoblastoma (Rb) in HCC 30, we subsequently checked the phosphorylation

295 of CHK2 substrates and downstream signaling in response to PCK1 expression.

296 Overexpressing WT PCK1 decreased the p-Rb and p-CDK2 levels, but increased the

297 p27 levels (Fig. 5f). In contrast, PCK1 KO or OAA treatment reversed the regulatory

298 effects of these molecules (Fig. 5g,h). Notably, the OGT inhibitor ST045849 or the

299 GFAT1 inhibitor DON partially offset the regulatory effects mediated by PCK1

300 deficiency (Fig. 5g,i). These data indicated that O-GlcNAcylation promotes CHK2

301 dimerization and subsequently enhances downstream Rb phosphorylation.

302

303 To further test whether the loss of CHK2 O-GlcNAcylation affects its downstream

304 signaling and hepatoma cell proliferation, we transiently overexpressed WT CHK2 or

305 the T378A mutant in CHK2-KO cells. WT CHK2 restored the phosphorylation of Rb

306 and CDK2 and promoted HCC proliferation, whereas the CHK2 T378A mutant failed

307 to exert this stimulatory role on tumorigenesis (Fig. 5j and Extended Data Fig. 6a,b),

308 indicating that T378 O-GlcNAcylation plays an essential role in CHK2 activation. The

309 T378A point mutation (which eliminated the O-GlcNAc modification) decreased the

310 capacity of CHK2 to phosphorylate Rb. In agreement, the OGA inhibitor PUGNAc

311 enhanced the ability of WT CHK2 to phosphorylate Rb and CDK2, but not that of the

312 CHK2 T378A mutant (Fig. 5j). Accordingly, PCK1 KO or OAA treatment promoted WT

313 CHK2 activation (Fig. 5k,l). Finally, we explored whether PCK1 deficiency-induced

314 malignancy may rely upon CHK2 O-GlcNAcylation. As expected, CHK2 depletion

315 suppressed PKO cell proliferation, which was rescued by re-expressing WT CHK2,

316 but not the T378 mutant (Extended Data Fig. 6c-f). These data suggest that CHK2

317 T378 O-GlcNAcylation conferred a growth advantage for the PKO cells. Collectively,

318 these findings indicate that the O-GlcNAcylation of residue T378 stabilizes CHK2 and

319 activates its downstream targets such as Rb, thus promoting PCK1-deficient

320 hepatoma cell proliferation in vitro.

321

322 Targeting HBP-Meditated O-GlcNAcylation Suppresses DEN/CCl4-Induced

323 Hepatocarcinogenesis in Vivo

324 Next, we used the DEN/CCl4-induced mouse model of liver cancer to further verify our

325 results in vivo. Based on our in vitro data, we proposed that blocking the HBP with an

326 inhibitor of GOT2 (AOA) or GFAT1 (DON) could suppress the growth of HCC by

327 reducing O-GlcNAcylation. LKO mice (Fig. 6a,b) were generated as described

31 328 previously . The mice were treated with DEN/CCl4 to induce hepatocarcinoma,

329 which was followed by administering AOA or DON (twice a week) for 16 weeks (Fig

330 6c). The LKO mice exhibited accelerated liver tumorigenesis with increased tumor

331 masses and nodules, and higher serum levels of aspartate aminotransferase (AST)

332 and alanine aminotransferase (Fig. 6d-h). Our MS data showed that the OAA and

333 UDP-GlcNAc levels were higher in the liver tumors of LKO mice (Fig. 6i,j), indicating

334 that PCK1 deficiency promotes the HBP in vivo. In contrast, mice treated with AOA or

335 DON exhibited slower tumor growth and a reduced number of tumor nodules,

336 compared with those of untreated LKO mice (Fig. 6d-f). Furthermore, AOA or DON

337 treatment also decreased UDP-GlcNAc and O-GlcNAcylation levels in LKO mice (Fig.

338 6j-l). These data suggested that hyper-O-GlcNAcylation conferred a growth

339 advantage for tumor cells in vivo. Consistent with our in vitro data, the levels of total

340 CHK2, O-GlcNAcylated CHK2, and p-Rb were significantly enhanced in the liver

341 tumors of LKO mice, which was partially reversed by administering AOA or DON (Fig.

342 6l). In summary, these findings suggested that PCK1 depletion increases

343 susceptibility to DEN/CCl4-induced carcinogenesis and promotes

344 hepatocarcinogenesis via enhanced CHK2 O-GlcNAcylation; thus, blocking

345 HBP-meditated O-GlcNAcylation suppresses HCC in LKO mice.

346

347 PCK1 Deficiency Strengthens CHK2 O-GlcNAcylation in Primary Human HCC

348 Finally, we investigated PCK1 expression and global O-GlcNAcylation in 40 paired

349 human primary HCC tissues and tumor-adjacent normal tissues. As shown by our

350 immunohistochemistry (IHC) and immunoblot results, PCK1 was down-regulated in

351 most HCC tissues (Fig. 7a-c and Extended Data Fig. 7), and deficient PCK1

352 expression was significantly associated with a larger tumor size and accelerated

353 proliferation (Fig. 7a,d, Pearson correlation’s coefficient (r) = -0.3935, p = 0.0160;

354 Supplemental Table 2). In addition, downregulated PCK1 expression was significantly

355 associated with poor tumor differentiation and prognosis (Supplemental Table 2).

356 Moreover, the global O-GlcNAcylation was significantly higher in HCC tissues than in

357 adjacent normal tissues (Fig. 7b,e). Consistently, the p-AMPK and p-GFAT1 levels

358 were reduced in HCC tissues (Fig. 7b). We also observed a negative correlation

359 between PCK1 protein-expression levels and O-GlcNAcylation levels in HCC (Fig. 7f,

360 r = -0.3565, p = 0.0240). In addition, the sWGA pull-down assay showed that

361 enhanced CHK2 O-GlcNAcylation was associated with PCK1 down-regulation (Fig.

362 7g,h). Consistent with our in vitro data, we observed a strong negative correlation

363 between p-Rb levels and PCK1 expression (Fig. 7i, r = -0.5347, p = 0.0270). In

364 conclusion, this clinical validation supports the finding that PCK1 repression

365 strengthens CHK2 O-GlcNAcylation and promotes tumor growth in human primary

366 HCC.

367

368 DISCUSSION

369 DISCUSSION

370 Emerging evidence has demonstrated that protein O-GlcNAcylation plays key role in

371 tumorigenesis. Increased glucose flux through the HBP elevates UDP-GlcNAc, which

372 enhances cellular O-GlcNAcylation. However, cancer cells are frequently faced with

373 limited nutrients due to an insufficient and inappropriate vascular supply and rapid

374 nutrient consumption 2. Previous metabolomics data demonstrated that the glucose

375 concentrations in tumor tissues are generally lower than those in non-transformed

376 tissues 32. The mechanisms underlying tumor growth during periods of metabolic

377 stress through enhanced HBP activity and O-GlcNAcylation have not been fully

378 elucidated. It remains unknown whether gluconeogenesis contributes to maintaining

379 HBP-mediated O-GlcNAcylation in cancer cells under low nutrient conditions 4. Here,

380 we present the first evidence that deficiency of the gluconeogenic enzyme PCK1

381 promotes cellular O-GlcNAcylation and tumorigenesis in HCC (Fig. 7j). Moreover, we

382 identify that CHK2 O-GlcNAcylation at T378 maintains its stability and oncogenic

383 activity in hepatoma cells. Therefore, our study provides a link between PCK1

384 repression and hyper-O-GlcNAcylation underlying HCC oncogenesis.

385

386 Since the liver is the major site of gluconeogenesis during fasting, the role of

387 gluconeogenesis in HCC has begun to draw more attention recently. During glucose

388 starvation, cancer cells redistribute gluconeogenic intermediates to downstream

389 pathways to facilitate their proliferation 4. PCK1, the rate-limiting enzyme of

390 gluconeogenesis, is downregulated in HCC 33,34,25. PCK1 mediates not only

391 gluconeogenesis, but also glyceroneogenesis and TCA cataplerosis 35. In this study,

392 we found that PCK1 silencing promoted the HBP and UDP-GlcNAc biosynthesis, thus

393 enhancing cellular O-GlcNAcylation under a low-glucose condition (5 mM glucose).

394 Interestingly, previous findings showed that the maximum rate of gluconeogenesis is

395 approached at glucose concentrations under 5 mM, whereas high glucose levels

396 inhibit gluconeogenesis 36,37. Consistent with this observation, we did not detect any

397 significant change in O-GlcNAcylation levels under high-glucose conditions (25 mM

398 glucose) upon PCK1 depletion or overexpression, indicating that PCK1 regulated

399 HBP flux and protein O-GlcNAcylation, depending on glucose availability.

400

401 O-GlcNAcylation depends on OGT/OGA levels, and the HBP produces the donor

402 substrate UDP-GlcNAc. Since PCK1 did not affect the expression levels of OGT and

403 OGA, we studied the effect of PCK1 on HBP-mediated UDP-GlcNAc biosynthesis. As

404 a nutrient sensor, UDP-GlcNAc levels are dependent on glucose, amino acid, fatty

405 acid, and nucleotide availability 38. Here, we revealed a dual role for PCK1 in

406 regulating UDP-GlcNAc biosynthesis through OAA accumulation and the

407 AMPK-GFAT1 axis, under low-glucose conditions. On the one hand, the OAA levels

408 decreased upon PCK1 overexpression in hepatoma cells, but accumulated in the liver

409 tumors of LKO mice. LKO mice are unable to remove oxaloacetate from the TCA

410 cycle 39. OAA is converted to Asp by GOT2, a key enzyme that plays a role in the TCA

411 cycle and amino acid metabolism 40. Accordingly, amino acids synthesized from the

412 TCA cycle (including Asp) were elevated in the blood of PCK1 KO mice 41.

413 Pharmacological or transcriptional inhibition of GOT2 suppressed

414 hyper-O-GlcNAcylation induced by PCK1 deficiency or additional OAA treatment in

415 vitro. Moreover, treatment with the GOT2 inhibitor, AOA reduced UDP-GlcNAc and

416 O-GlcNAcylation levels in LKO mice, which may represent an important therapeutic

417 perspective for HCC treatment. These findings implied that the cataplerotic function of

418 PCK1 and the GOT2-mediated pathway are involved in regulating UDP-GlcNAc

419 biosynthesis. Whether OAA-derived Asp is incorporated into UDP-GlcNAc via

420 pyrimidine synthesis requires further study, based on stable isotope tracing.

421

422 On the other hand, our previous work showed that enforced PCK1 expression leads to

423 energy reduction and activates AMPK upon glucose deprivation in HCC 25. GFAT1 is

424 an AMPK substrate, so we hypothesized that PCK1 may suppress HBP-mediated

425 O-GlcNAcylation through the AMPK-GFAT1 axis. In support of this hypothesis, we

426 found that PCK1 regulated GFAT1 phosphorylation and O-GlcNAcylation levels in an

427 AMPK-dependent manner, under glucose restrictions. As a rate-limiting enzyme of

428 HBP, GFAT1 phosphorylation at Ser243 by AMPK diminishes its enzymatic activity 42,

429 which decreases UDP-GlcNAc biosynthesis and O-GlcNAcylation levels. Interestingly,

430 the AMPK-GFAT1 axis was also reported to regulate protein O-GlcNAcylation in

431 angiogenesis and cardiac hypertrophy, indicating that this signal axis plays multiple

432 physiological roles 26,27. In addition, AMPK directly phosphorylated residue Thr444 of

433 OGT, the enzyme responsible for O-GlcNAcylation 43. Therefore, PCK1 may regulate

434 the HBP and O-GlcNAcylation through different mechanisms.

435

436 It is known that O-GlcNAcylation is crucial for cell-cycle regulation and DNA-damage

437 responses 44. Several proteins that regulate the growth and proliferation of tumor cells

438 can be O-GlcNAcylated. For example, the G1/S checkpoint protein Rb is heavily

439 O-GlcNAcylated during the G1 phase 45. By characterizing the role of

440 O-GlcNAcylation upon PCK1 deficiency, we uncovered CHK2 as novel target of OGT.

441 CHK2, a cell-cycle checkpoint kinase, play key roles in DNA-damage responses and

442 cell-cycle progression. CHK2 is expressed in the nucleus in a subset of HCC and

443 correlates with HCC progression 30. Several post-translational modifications, including

444 phosphorylation, ubiquitination, and acetylation have been reported to be critical for

445 CHK2 function 46–48. Herein, we identified Thr378 as a key O-GlcNAcylation site on

446 CHK2 using LC-MS/MS. Importantly, the loss of O-GlcNAcylation by the T378A

447 mutation increased CHK2 ubiquitination, thus promoted its degradation. Moreover, we

448 found that O-GlcNAcylation promoted CHK2 dimerization and activation, therefore

449 enhancing Rb phosphorylation in PCK1-deficient hepatoma cells. Further structural

450 analysis of O-GlcNAcylated CHK2 may help us to understand how O-GlcNAcylation

451 contributes to CHK2 activation.

452

453 Histopathology had revealed increased O-GlcNAcylation levels in HCC tumor tissues

454 49. Our findings not only provide an underlying mechanism whereby disrupted

455 gluconeogenesis may activate the HBP and increase the availability of UDP-GlcNAc

456 for O-GlcNAcylation under nutrient limitations, but also provides potential therapeutic

457 targets for HCC. Preclinical evaluation of DON and AOA through the inhibition of

458 glutamine metabolism has provided promising results for acute myeloid leukemia 22,

459 high MYC-expressing atypical teratoid/rhabdoid tumors 50, and breast cancer 51. Here,

460 we showed that both DON and AOA inhibited the growth of HCC in vitro and in vivo,

461 largely by blocking HBP-mediated O-GlcNAcylation. These data suggest that DON

462 and AOA inhibit cell growth through a novel mechanism and provide a strong rationale

463 for further clinical drug development, particularly for PCK1-deficient HCC.

464

465 In summary, we uncovered a link between gluconeogenesis disruption and

466 O-GlcNAcylation upon glucose deprivation in HCC. We demonstrated that PCK1

467 deficiency can promote HBP-mediated UDP-GlcNAc biosynthesis through OAA

468 accumulation and the AMPK-GFAT1 axis. Moreover, the OGT-mediated

469 O-GlcNAcylation of CHK2 on Thr378 stabilizes CHK2 and promote its oncogenic

470 activity in HCC. The results of this study expands our understanding of PCK1 in

471 hepatic carcinogenesis and indicates the potential of targeting HBP-mediated

472 O-GlcNAcylation for HCC therapy.

473

474 METHODS

475 Cell Culture and Treatment

476 PLC/PRF/5, SK-Hep1, Huh7, MHCC-97H, and HEK293 cells were cultured in DMEM

477 (HyClone, Logan, UT, USA) supplemented with 10% FBS (Gibco, Rockville, MD,

478 USA), 100 U/ml penicillin, and 100 mg/ml streptomycin (HyClone) at 37°C in 5% CO2.

479 For low-glucose treatment, cells were briefly washed with PBS (DINGGUO, BF-0011)

480 and then maintained in glucose-free medium (Gibco, 11966025) supplemented with

481 10% FBS and glucose at various concentrations for 12 h. In addition, 3-MPA, Dex, ST,

482 TG, AOA, DON, or metformin was added to the medium, as indicated.

483

484 Animal Studies

485 AlbCre(+/-), Pck1(flox/flox) (LKO) mice were generated from crosses between

486 AlbCre(+/-) mice (Model Animal Research Center of Nanjing University, Nanjing, China)

487 and Pck1(flox/flox) mice with a C57BL/6 background (Mutant Mouse Resource &

488 Research Centers, MMRRC:011950-UNC), as described previously 31, and AlbCre(-/-),

489 Pck1(flox/flox) (WT) mice were used as a control (n = 6/group). HCC was induced in mice

490 by combined treatment with DEN (75 mg/kg) and CCl4 (2 ml/kg, twice per week for 12

52 491 weeks), as described previously . At 16 weeks after DEN/CCl4 treatment, the LKO

492 mice were administered an intraperitoneal injection of 5 mg/kg

493 AOA-hemihydrochloride (Sigma) or 1 mg/kg DON (Sigma) twice a week for 16 weeks.

494 At 32 weeks, the mice were sacrificed after fasting for 12 h, and liver tissues with

495 tumors were collected for examination. Mouse serum ALT and AST were detected

496 using an automatic biochemical analyzer (Catalyst One, IDEXX, USA).

497

498 For the orthotopic implantation model, BALB/c nude mice (n = 6/group) were

499 randomly divided into 5 groups. For each nude mouse, AdGFP-, AdPCK1-, AdG309R-,

500 or mock-infected MHCC97H cells (1 × 105) were suspended in a 50-µl PBS/Matrigel

501 (356234, BD Biosciences) mixture (1:1 ratio, v/v) and then implanted into the left liver

502 lobe. At 14 days after injection, the Dex group was intraperitoneally injected with 5

503 mg/kg Dex per day for 14 days. At 4 weeks after implantation, all mice were sacrificed

504 after fasting for 12 h. All animal procedures were performed according to protocols

505 approved by the Institutional Animal Care and Use Committee at the Laboratory

506 Animal Center of Chongqing Medical University. All procedures were also approved

507 by the Research Ethics Committee of Chongqing Medical University (reference

508 number: 2017012).

509

510 Clinical Specimens

511 HCC samples and paired, adjacent normal liver tissues were obtained from the

512 Second Affiliated Hospital of Chongqing Medical University between 2015 and 2018,

513 with approval from the Institutional Review Board of Chongqing Medical University.

514 Clinical samples were collected from patients after obtaining informed consent in

515 accordance with a protocol approved by the Second Affiliated Hospital of Chongqing

516 Medical University (Chongqing, China).

517

518 Plasmid Constructs

519 For recombinant plasmid construction, DNA fragments (PCK1 NM_002591, CHEK2

520 NM_007194.4, ΔN truncated CHK2 69-543aa, ΔC truncated CHK2 1-221aa, and OGT

521 NM_181672.2) were amplified by PCR and separately cloned into the pSEB-3Flag,

522 pBudCE4.1-3HA, or pBudCE4.1-5Myc vector. PCK1 G309R, CHK2 T378A, and

523 CHK2 T383A were constructed by overlapping PCR. To construct plasmids

524 expressing shRNAs, two pairs of oligonucleotides encoding shRNAs targeting OGT,

525 OGA, AMPK, or GOT2 (Supplemental Table 3) were cloned into the pLL3.7 vector

526 (from Prof. Bing Sun, Shanghai Institute of Biochemistry and Cell Biology, Chinese

527 Academy of Sciences, China).

528

529 Adenovirus Production

530 The full-length cDNA fragment of PCK1 (NM_002591) or G309R (PCK1 mutation

531 925G>A) was inserted into the pAdTrack-TO4 vector (from Dr. Tong-Chuan He,

532 University of Chicago, USA). Recombinant adenoviral, AdPCK1 and AdG309R, were

533 generated using the AdEasy system as described previously 25. The adenoviral

534 AdGFP, expressing only GFP, was used as a control.

535

536 CRISPR/Cas9-mediated Knockout Cells

537 PCK1- or CHEK2-knockout cells (PKO or CKO cells) were established using the

538 CRISPR-Cas9 system (from Prof. Ding Xue, the School of Life Sciences, Tsinghua

539 University, Beijing, China), as described previously 25. Single-cell HCC clones stably

540 expressing single guide RNA (sgRNA) sequences were propagated and validated by

541 immunoblotting and DNA sequencing. The sequences of all oligonucleotides used to

542 generate the knockout cell lines are listed in Supplemental Table 3.

543

544 Cell Lysis and Western Blotting

545 Protein lysates from cells or liver samples were extracted using cell lysis buffer

546 (Beyotime Biotechnology, Jiangsu, China) containing 1 mM phenylmethanesulfonyl

547 fluoride (Beyotime). Equal volumes of protein samples were separated by 10%

548 SDS/PAGE and electro-transferred to PVDF membranes (Millipore, Billerica, MA,

549 USA). The immunoblots were probed with the indicated antibodies. Proteins bands

550 were visualized with Chemiluminescent Substrate (Bio-Rad, USA) and exposed to

551 CareStream Film (Kodak, USA). Quantification of bands in western blots was

552 performed using Image-Pro Plus software.

553

554 RNA Extraction, Reverse Transcription PCR, and Quantitative Real-Time PCR

555 Total RNA was extracted from cells using the TRIzol™ reagent (Invitrogen, Carlsbad,

556 CA, USA) and reverse-transcribed into cDNA using the PrimeScript™ RT Reagent Kit

557 (RR047A, TaKaRa, Japan). Quantitative real-time PCR analysis of target

558 mRNA-expression levels was achieved using SYBR Green qPCR Master Mix

559 (Bio-Rad, Hercules, CA, USA) and the CFX Connect Real-time PCR Detection

560 System (Bio-Rad). The specific primers used are shown in Supplemental Table 3. All

561 samples were analyzed in triplicate for each experiment. The relative gene-expression

562 levels were normalized to β-actin and calculated using the 2–ΔΔCt method.

563

564 Cell-Proliferation Assay

565 Cells were re-seeded in 96-well plates at a density of 1 × 103 cells/well and cultured in

566 low-glucose medium containing 5 mM glucose for 5 days. The plates were scanned

567 and phase-contrast images were acquired in real time every 24 h post-treatment.

568 Quantified time-lapse curves were generated using IncuCyte ZOOM software (ESCO,

569 Changqi South Street, Singapore). For clone formation capacity, cells were seeded

570 into 6-well plate at 800/well. After 7-10 days on 5 mM glucose condition, the cells were

571 fixed with paraformaldehyde and then incubated in crystal violet. The clones were

572 photographed and the number of clones was counted by Image-Pro Plus software,

573 version 6.0 (Media Cybernetics, Inc., Bethesda, MD, USA).

574

575 Metabolites Detection and Analysis

576 Cells were washed twice with pre-cooled physiological saline, and metabolites were

577 extracted with 400 μL cold methanol/acetonitrile (1:1, v/v) to remove the protein. The

578 mixture was centrifuged for 20 min (14,000 × g, 4°C). The supernatant was dried in a

579 vacuum centrifuge. For LC-MS analysis, the samples were re-dissolved in 100 μL

580 acetonitrile/water (1:1, v/v) solvent. Mouse liver tumor tissues (60 mg) were extracted

581 with 1 mL cold 90% methanol. The lysates were homogenized twice using an MP

582 homogenizer (24 × 2, 6.0 M/S, 60 s). The homogenates were sonicated on ice. After

583 centrifugation at 14,000 × g for 20 min, the supernatant was dried and re-dissolved in

584 100 μL acetonitrile/water (1:1, v/v) solvent.

585

586 Untargeted metabolomics profiling of PCK1-overexpressing cells was performed

587 using ultra-high-performance liquid chromatography (Agilent 1290 Infinity LC) coupled

588 to with quadrupole time-of-flight mass spectrometry (UHPLC-QTOF/MS) at Shanghai

589 Applied Protein Technology Co., Ltd 53. UDP-GlcNAc levels were quantified by

590 performing targeted LC-MS/MS analysis (ACQUITY UPLC I CLASS, Xevo G2-S QTof).

591 TCA-derived metabolites were detected by UHPLC, using an Agilent 1290 Infinity LC

592 column coupled to 5500 QTRAP system (AB SCIEX) at Shanghai Applied Protein

593 Technology Co., Ltd.

594

595 Immunoprecipitation

596 PKO or SK-Hep1 cells were transfected for 48 h with a fusion vector expressing

597 Flag-tagged fusion protein (OGT-Flag, CHK2-Flag, T378A-Flag, or T383A-Flag). Cells

598 were treated with 50 μm PUGNAc for 12 h and resuspended in lysis buffer (50 mM

599 Tris HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, and 1% Triton X-100) containing

600 protease (Roche) and phosphatase (Beyotime Biotechnology) inhibitor cocktails.

601 Pre-cleared lysates were incubated overnight with an anti-FLAG M2 affinity gel

602 (Sigma, A2220) at 4°C.

603

604 For co-IP analysis, HEK293 cells were co-transfected with vectors expressing

605 OGT-HA and either CHK2-Flag, ΔC-Flag, or ΔN-Flag. PKO cells were co-transfected

606 with the CHK2-Flag and CHK2-Myc vectors. Pre-cleared cell lysates were incubated

607 overnight with an anti-Flag or anti-HA antibody, and coupled with 40 μl protein A/G

608 agarose beads. Immunoprecipitated complexes were eluted and subjected to

609 immunoblotting using the indicated antibodies.

610

611 sWGA Pull-Down Assay

612 Hepatic cells and liver tissues were lysed in a buffer containing 125 mM NaCl, 50 mM

613 Tris (pH 7.4), 5 mM EDTA, and 0.1% NP-40. The lysate was denatured in

614 glycoprotein-denaturing buffer and digested with PNGase (NEB P0704S) to remove

615 N-linked glycoproteins. Pre-cleared lysates were incubated overnight with

616 sWGA-conjugated agarose beads (Vector Laboratories, Burlingame, CA).

617 Precipitated complexes were eluted and immunoblotted with the indicated antibodies.

618

619 IP-MS assay

620 PKO cells transfected with Flag-tagged OGT before cultured in 5 mM glucose were

621 collected and incubated overnight at 4°C with ANTI-FLAG M2 affinity gel (Sigma,

622 A2220). Immunoprecipitated complexes were eluted and stained with Coomassie blue.

623 Excised protein gel bands were sent to Shanghai Applied Protein Technology Co., Ltd.

624 to identify OGT-binding proteins. After performing the reduction reaction and alkylation

625 treatment, the samples were incubated with trypsin (mass ratio, 1:50). The tryptic

626 products were desalted, lyophilized, and re-dissolved in 0.1% formic acid solution. MS

627 analysis was performed using a Q Exactive MS instrument (Thermo Fisher) coupled

628 to an Easy-nLC 1000 LC instrument (Thermo Fisher) in Shanghai Applied Protein

629 Technology Co., Ltd. The raw MS data were analyzed using Mascot 2.2 software to

630 identify immunoprecipitating proteins. Further, pathway-enrichment analysis was

631 performed using the KEGG database (http://kobas.cbi.pku.edu.cn/), p values < 0.05.

632

633 Immunofluorescence Assays

634 CHK2 and OGT staining was performed with anti-Flag (mouse 1:200) and anti-HA

635 (rabbit 1:800) antibodies, using PKO cells co-transfected with CHK2-Flag and

636 OGT-HA vectors. Specific signals were visualized using secondary antibodies (goat

637 anti-rabbit IgG/TRITC or goat anti-mouse IgG/FITC). For nuclear staining, cells were

638 treated with 1 μg/mL DAPI (10236276001, Roche Diagnostics GmbH, Mannheim,

639 Germany). Stained sections were detected by a laser-scanning confocal microscope

640 (Leica TCS SP8, Solms, Germany).

641

642 CHK2 O-GlcNAcylation Site Mapping

643 MS was performed to identify CHK2 O-GlcNAcylation sites, as described previously 54.

644 Briefly, CHK2 was immunoprecipitated from PKO cells transfected with CHK2-Flag

645 and subjected to 10% SDS-PAGE for Coomassie blue staining. The band

646 corresponding to CHK2 was excised from the gel and digested overnight with trypsin.

647 Online LC-MS/MS system consisting of an Easy-nLC system and an Orbitrap Fusion

648 Lumos Tribrid MS instrument (Thermo Scientific, Germany) equipped with a

649 nanoelectrospray ion source were performed. The data from the Raw MS files were

650 analyzed against the UniProt database with MaxQuant software (version 1.5.2.8). The

651 first-search peptide tolerance was 20 ppm and the main-search peptide tolerance was

652 6 ppm. The MS/MS tolerance was 0.02 Da. The peptide-spectrum match and the

653 false-discovery level was set to 1%. Matches between runs were determined and the

654 minimum score for modified peptides was set to 40.

655

656 CHK2 Protein-Stability Assay

657 PCK1-OE SK-Hep1 cells or PKO cells transfected with CHK2 WT or mutant plasmids

658 were treated with 40 μM cycloheximide and harvested after 0, 12, 24, or 36 h. The

659 cells were lysed, and CHK2 was detected using an anti-Flag antibody (Sigma).

660

661 CHK2-Oligomerization Assay

662 The PKO cells were incubated with HEPES buffer (40 mM HEPES pH 7.5, 150 mM

663 NaCl, 0.1% NP-40) supplemented with protease (Roche) and phosphatase (Beyotime)

664 inhibitor cocktails for 15 min and centrifuged at 14,000 × g for 20 min at 4°C.

665 Whole-cell extracts were cross-linked with 0.025% glutaraldehyde for 5 min at 37°C

666 and then quenched with 50 mM Tris-HCl (pH 8.0). The samples were then separated

667 on a 4–20% precast gel and analyzed by western blotting using the indicated

668 antibodies.

669

670 Molecular Dynamic Simulation

671 Missing loops and side chains of the dimeric CHK2 structure were complemented

672 using SWISS-model with a previously reported CHK2 dimer structure as template

673 (PDB code: 3I6U). The O-GlcNAcylation was added onto the predicted structure using

674 PyMOL 2.3 software. Molecular dynamic simulation was conducted using GROMACS

675 2019 package 55. Topologies were generated by ACPYPE with Amber99sb force field

676 parameters 56. After an initial energy minimization, 10 ns of unrestrained NPT MD

677 simulation was performed under constant temperature of 300 K and constant

678 pressure of 1 atm.

679

680 Immunohistochemical (IHC) Assay

681 Human or mouse liver tissues were fixed in 4% paraformaldehyde for at least 24 h

682 and embedded in paraffin, according to standard procedures. The sections were

683 incubated overnight at 4°C with primary antibodies against PCK1 (1:500), p-GFAT

684 (1:200), O-GlcNAc (1:100), or Ki67 (1:100). Subsequently, the slides were incubated

685 with secondary anti-rabbit or anti-mouse IgG (ZSGB-BIO, Beijing, China) and

686 visualized using 3, 3′-diaminobenzidine (ZSGB-BIO). The stained slides were

687 scanned with a Pannoramic Scan 250 Flash or MIDI system, and images were

688 acquired using Pannoramic Viewer software, version 1.15.2 (3DHistech, Budapest,

689 Hungary). The mean staining intensity was analyzed using Image-Pro Plus software.

690

691 Quantification and Statistical Analysis

692 Integral optical density (IOD) values were measured using Image-Pro Plus software

693 (version 6.0) to determine the intensity of protein expression. The IOD was calculated

694 as the mean density × the area. The relative protein-expression levels were

695 normalized to an internal reference control, such as β-actin. All statistical analyses

696 were performed using GraphPad Prism 7 (GraphPad Software Inc.). Data are

697 represented as mean ± SD. Student’s t-test was used to compare two groups.

698 One-way ANOVA followed by Tukey’s test was used to compare more than two

699 groups. Pearson’ correlation coefficient (r) was used to test linear correlations.

700 Statistical significance was defined as p values < 0.05. *P < 0.05, **P < 0.01, ***P <

701 0.001.

702

703 References

704 1. Ward, P. S. & Thompson, C. B. Metabolic Reprogramming: A Cancer Hallmark Even Warburg Did Not

705 Anticipate. Cancer Cell 21, 297–308 (2012).

706 2. Boroughs, L. K. & DeBerardinis, R. J. Metabolic pathways promoting cancer cell survival and growth.

707 Nat. Cell Biol. 17, 351–359 (2015).

708 3. Lunt, S. Y. & Vander Heiden, M. G. Aerobic Glycolysis: Meeting the Metabolic Requirements of Cell

709 Proliferation. Annu. Rev. Cell Dev. Biol. 27, 441–464 (2011).

710 4. Grasmann, G., Smolle, E., Olschewski, H. & Leithner, K. Gluconeogenesis in cancer cells – Repurposing

711 of a starvation-induced metabolic pathway? Biochim. Biophys. Acta BBA - Rev. Cancer 1872, 24–36

712 (2019).

713 5. Wang, Z. & Dong, C. Gluconeogenesis in Cancer: Function and Regulation of PEPCK, FBPase, and

714 G6Pase. Trends Cancer 5, 30–45 (2019).

715 6. Nwosu, Z. C. et al. Identification of the Consistently Altered Metabolic Targets in Human

716 Hepatocellular Carcinoma. Cell. Mol. Gastroenterol. Hepatol. 4, 303-323.e1 (2017).

717 7. Montal, E. D. et al. PEPCK Coordinates the Regulation of Central Carbon Metabolism to Promote

718 Cancer Cell Growth. Mol. Cell 60, 571–583 (2015).

719 8. Hanover, J. A., Krause, M. W. & Love, D. C. The hexosamine signaling pathway: O-GlcNAc cycling in

720 feast or famine. Biochim. Biophys. Acta BBA - Gen. Subj. 1800, 80–95 (2010).

721 9. Ma, Z. & Vosseller, K. O-GlcNAc in cancer biology. Amino Acids 45, 719–733 (2013).

722 10. Hanover, J. A., Chen, W. & Bond, M. R. O-GlcNAc in cancer: An Oncometabolism-fueled vicious cycle. J.

723 Bioenerg. Biomembr. 50, 155–173 (2018).

724 11. Torres, C. R. & Hart, G. W. Topography and polypeptide distribution of terminal N-acetylglucosamine

725 residues on the surfaces of intact lymphocytes. Evidence for O-linked GlcNAc. J. Biol. Chem. 259,

726 3308–3317 (1984).

727 12. Hart, G. W., Housley, M. P. & Slawson, C. Cycling of O-linked β-N-acetylglucosamine on

728 nucleocytoplasmic proteins. Nature 446, 1017–1022 (2007).

729 13. Zachara, N. E. & Hart, G. W. O-GlcNAc a sensor of cellular state: the role of nucleocytoplasmic

730 glycosylation in modulating cellular function in response to nutrition and stress. Biochim. Biophys.

731 Acta BBA - Gen. Subj. 1673, 13–28 (2004).

732 14. Ferrer, C. M., Sodi, V. L. & Reginato, M. J. O-GlcNAcylation in Cancer Biology: Linking Metabolism and

733 Signaling. J. Mol. Biol. 428, 3282–3294 (2016).

734 15. Jóźwiak, P., Forma, E., BryÅ›, M. & KrzeÅ›lak, A. O-GlcNAcylation and Metabolic Reprograming in

735 Cancer. Front. Endocrinol. 5, (2014).

736 16. Fardini, Y., Dehennaut, V., Lefebvre, T. & Issad, T. O-GlcNAcylation: A New Cancer Hallmark? Front.

737 Endocrinol. 4, (2013).

738 17. Nagel, A. K. & Ball, L. E. Intracellular Protein O-GlcNAc Modification Integrates Nutrient Status with

739 Transcriptional and Metabolic Regulation. in Advances in Cancer Research vol. 126 137–166 (Elsevier,

740 2015).

741 18. Love, D. C. & Hanover, J. A. The Hexosamine Signaling Pathway: Deciphering the ‘O-GlcNAc Code’. Sci.

742 Signal. 2005, re13–re13 (2005).

743 19. Chou, T.-Y., Hart, G. W. & Dang, C. V. c-Myc Is Glycosylated at Threonine 58, a Known Phosphorylation

744 Site and a Mutational Hot Spot in Lymphomas. J. Biol. Chem. 270, 18961–18965 (1995).

745 20. Park, S. Y. et al. Snail1 is stabilized by O‐GlcNAc modification in hyperglycaemic condition. EMBO J. 29,

746 3787–3796 (2010).

747 21. Sayat, R., Leber, B., Grubac, V., Wiltshire, L. & Persad, S. O-GlcNAc-glycosylation of β-catenin regulates

748 its nuclear localization and transcriptional activity. Exp. Cell Res. 314, 2774–2787 (2008).

749 22. Ricciardiello, F. et al. Inhibition of the Hexosamine Biosynthetic Pathway by targeting PGM3 causes

750 breast cancer growth arrest and apoptosis. Cell Death Dis. 9, 377 (2018).

751 23. Sharma, N. S. et al. Targeting tumor-intrinsic hexosamine biosynthesis sensitizes pancreatic cancer to

752 anti-PD1 therapy. J. Clin. Invest. 130, (2019).

753 24. Vieira, P. et al. Novel homozygous PCK1 mutation causing cytosolic phosphoenolpyruvate

754 carboxykinase deficiency presenting as childhood hypoglycemia, an abnormal pattern of urine

755 metabolites and liver dysfunction. Mol. Genet. Metab. 120, 337–341 (2017).

756 25. Tuo, L. et al. PCK1 negatively regulates cell cycle progression and hepatoma cell proliferation via the

757 AMPK/p27Kip1 axis. J. Exp. Clin. Cancer Res. 38, 50 (2019).

758 26. Zibrova, D. et al. GFAT1 phosphorylation by AMPK promotes VEGF-induced angiogenesis. Biochem. J.

759 474, 983–1001 (2017).

760 27. Gélinas, R. et al. AMPK activation counteracts cardiac hypertrophy by reducing O-GlcNAcylation. Nat.

761 Commun. 9, 374 (2018).

762 28. Pilié, P. G., Tang, C., Mills, G. B. & Yap, T. A. State-of-the-art strategies for targeting the DNA damage

763 response in cancer. Nat. Rev. Clin. Oncol. 16, 81–104 (2019).

764 29. Cai, Z., Chehab, N. H. & Pavletich, N. P. Structure and Activation Mechanism of the CHK2 DNA Damage

765 Checkpoint Kinase. Mol. Cell 35, 818–829 (2009).

766 30. Carloni, V. et al. CHK2 overexpression and mislocalisation within mitotic structures enhances

767 chromosomal instability and hepatocellular carcinoma progression. Gut 67, 348–361 (2018).

768 31. She , P. et al. Phosphoenolpyruvate Carboxykinase Is Necessary for the Integration of Hepatic Energy

769 Metabolism. Mol. Cell. Biol. 20, 6508–6517 (2000).

770 32. Birsoy, K. et al. Metabolic determinants of cancer cell sensitivity to glucose limitation and biguanides.

771 Nature 508, 108–112 (2014).

772 33. Ma, R. et al. Switch of glycolysis to gluconeogenesis by dexamethasone for treatment of

773 hepatocarcinoma. Nat. Commun. 4, 2508 (2013).

774 34. Liu, M.-X. et al. Metabolic reprogramming by PCK1 promotes TCA cataplerosis, oxidative stress and

775 apoptosis in liver cancer cells and suppresses hepatocellular carcinoma. Oncogene 37, 1637–1653

776 (2018).

777 35. Beale, E., Hammer, R., Antoine, B. & Forest, C. Disregulated glyceroneogenesis: PCK1 as a candidate

778 diabetes and obesity gene. Trends Endocrinol. Metab. 15, 129–135 (2004).

779 36. Rigoulet, M. & Leverve, X. M. Stimulation by glucose of gluconeogenesis in hepatocytes isolated from

780 starved rats. Biochem J 245, 8 (1987).

781 37. Ruan, H.-B. et al. O-GlcNAc Transferase/Host Cell Factor C1 Complex Regulates Gluconeogenesis by

782 Modulating PGC-1α Stability. Cell Metab. 16, 226–237 (2012).

783 38. Yang, X. & Qian, K. Protein O-GlcNAcylation: emerging mechanisms and functions. Nat. Rev. Mol. Cell

784 Biol. 18, 452–465 (2017).

785 39. Beale, E. G., Harvey, B. J. & Forest, C. PCK1 and PCK2 as candidate diabetes and obesity genes. Cell

786 Biochem. Biophys. 48, 89–95 (2007).

787 40. Altman, B. J., Stine, Z. E. & Dang, C. V. From Krebs to clinic: glutamine metabolism to cancer therapy.

788 Nat. Rev. Cancer 16, 619–634 (2016).

789 41. Hakimi, P. et al. Phosphoenolpyruvate carboxykinase and the critical role of cataplerosis in the control

790 of hepatic metabolism. Nutr Metab. 2, 33 (2005).

791 42. Eguchi, S. et al. AMP-activated protein kinase phosphorylates glutamine : fructose-6-phosphate

792 amidotransferase 1 at Ser243 to modulate its enzymatic activity. Genes Cells 14, 179–189 (2009).

793 43. Bullen, J. W. et al. Cross-talk between Two Essential Nutrient-sensitive Enzymes: O-GlcNAc

794 TRANSFERASE (OGT) AND AMP-ACTIVATED PROTEIN KINASE (AMPK). J. Biol. Chem. 289, 10592–10606

795 (2014).

796 44. Liu, C. & Li, J. O-GlcNAc: A Sweetheart of the Cell Cycle and DNA Damage Response. Front. Endocrinol.

797 9, 415 (2018).

798 45. Wells, L., Slawson, C. & Hart, G. W. The E2F-1 associated retinoblastoma-susceptibility gene product is

799 modified by O-GlcNAc. Amino Acids 40, 877–883 (2011).

800 46. Ahn, J. Y., Schwarz, J. K., Piwnica-Worms, H. & Canman, C. E. Threonine 68 phosphorylation by ataxia

801 telangiectasia mutated is required for efficient activation of Chk2 in response to ionizing radiation.

802 Cancer Res. 60, 5934–5936 (2000).

803 47. García-Limones, C. et al. CHK2 stability is regulated by the E3 ubiquitin ligase SIAH2. Oncogene 35,

804 4289–4301 (2016).

805 48. Zhang, W. et al. SIRT1 modulates cell cycle progression by regulating CHK2

806 acetylation−phosphorylation. Cell Death Differ. 27, 482-496 (2020)

807 49. Satriano, L., Lewinska, M., Rodrigues, P. M., Banales, J. M. & Andersen, J. B. Metabolic rearrangements

808 in primary liver cancers: cause and consequences. Nat. Rev. Gastroenterol. Hepatol. 16, 748-766

809 (2019).

810 50. Wang, S. Z. et al. Unbiased Metabolic Profiling Predicts Sensitivity of High MYC-Expressing Atypical

811 Teratoid/Rhabdoid Tumors to Glutamine Inhibition with 6-Diazo-5-Oxo-L-Norleucine. Clin. Cancer Res.

812 25, 5925–5936 (2019).

813 51. Korangath, P. et al. Targeting Glutamine Metabolism in Breast Cancer with Aminooxyacetate. Clin.

814 Cancer Res. 21, 3263–3273 (2015).

815 52. Yang, F. et al. GSTZ1-1 Deficiency Activates NRF2/IGF1R Axis in HCC via Accumulation of

816 Oncometabolite Succinylacetone. EMBO J. 38, e101964 (2019).

817 53. Lu, L. et al. HJC0152 suppresses human non–small‐cell lung cancer by inhibiting STAT3 and modulating

818 metabolism. Cell Prolif. 53, e12777 (2020).

819 54. Peng, C. et al. Regulation of the Hippo-YAP Pathway by Glucose Sensor O-GlcNAcylation. Mol. Cell 68,

820 591-604.e5 (2017).

821 55. Abraham, M. J. et al. GROMACS: High performance molecular simulations through multi-level

822 parallelism from laptops to supercomputers. SoftwareX 1–2, 19–25 (2015).

823 56. Sousa da Silva, A. W. & Vranken, W. F. ACPYPE - AnteChamber PYthon Parser interfacE. BMC Res. Notes

824 5, 367 (2012).

825

826

827 ACKNOWLEDGMENTS

828 We would like to thank Dr. T.-C He (University of Chicago, USA) for providing the

829 pAdEasy system and critical reading of the manuscript. We are grateful to Prof. Ding

830 Xue (Tsinghua University, China) for supplying the CRISPR/Cas9 system. We also

831 thank Prof. Bing Sun (Shanghai Institute of Biochemistry and Cell Biology, China) for

832 providing the pLL3.7 vector. We thank Zhimin Lu (Zhejiang University, China) for

833 suggestions and critical reading of the manuscript. This work was supported by the

834 China National Natural Science Foundation (grant no. 81872270), the Natural

835 Science Foundation Project of Chongqing (cstc2018jcyjAX0254,

836 cstc2019jcyj-msxmX0587), the Major National S&T program (2017ZX10202203-004),

837 and the Science and Technology Research Program of Chongqing Municipal

838 Education Commission (KJQN201900429).

839

840 AUTHOR CONTRIBUTIONS

841 NT, AH, and KW conceived and designed the study. J Xiang, CC, RL and DG

842 performed most experiments and analyzed the data. LC and WZ performed CHK2

843 O-GlcNAcylation site mapping. HD and LT helped with data analysis. LL generated

844 CHK2 mutants. QG, XP, and J Xia assisted with mice experiments. LH performed

845 molecular dynamic simulation. J Xiang, KW, and NT wrote the manuscript with all

846 authors providing feedback.

847 DECLARATION OF INTERESTS

848 The authors declare no competing interests.

849 40

850 FIGURES AND FIGURE LEGENDS

851

852 Fig. 1 PCK1 Deficiency Enhances Protein O-GlcNAcylation. a,b, Immunoblotting

853 analysis of global O-GlcNAcylation levels in PCK1-knockout PLC/PRF/5 cells (PKO

854 cells) treated with medium containing different levels of glucose for 12 h (a). 41

855 Densitometric analysis was performed with Image-pro plus software (b). c,

856 Representative western blot analysis of the indicated proteins in SK-Hep1 cells

857 overexpressing green fluorescent protein (GFP; control cells), WT PCK1, or an

858 enzymatically deficient mutant (PCK1 G309R) after incubation in medium containing 5

859 mM glucose for 12 h. Mock-treated cells served as a blank control. d,e, Immunoblots

860 of the indicated proteins in PLC/PRF/5 cells treated with 3-MPA for 12 h (d) and

861 SK-Hep1 cells treated with Dex for 7 days (e). f,g, OGT and protein O-GlcNAcylation

862 levels in PKO cells. Cells were transfected with shRNA targeting OGT mRNA or a

863 scrambled control shRNA (shcon) for 48 h (f), or treated with 50 μM ST045849 (ST)

864 for 12 h (g). h,i, Immunoblotting analysis of SK-Hep1 cells. PCK1-expressing cells

865 were transfected with an OGA shRNA1/2 plasmid for 48 h (h), or treated with 25 μM

866 Thiamet G (TG) for 12 h (i). j,k, Immunoblotting (j) and densitometric analysis (k) of

867 liver tumors from DEN/CCl4-induced WT and LKO mice after fasting for 12 h. Data are

868 represented as mean ± standard deviation (SD; n = 6 experiments). ***p < 0.001,

869 Student’s t-test.

870

871

872 Fig. 2 PCK1 Knockout Promotes UDP-GlcNAc Synthesis Partially Through

873 Oxaloacetate Accumulation. a,b, Heatmap of metabolites (a) and fold-changes in

874 intermediate metabolites of the HBP (b). c,d, Relative UDP-GlcNAc levels were

875 measured by LC-MS in PCK1-OE SK-Hep1 cells (c) and PKO cells (d). e, Schematic 43

876 representation of the HBP. Glucose intake feed into the HBP that produces

877 UDP-GlcNAc. N-Acetylglucosamine-1-phosphate (GlcNAc1P) and UTP, terminal

878 metabolites of the HBP and pyrimidine synthesis, represent the final rate-limiting steps

879 of UDP-GlcNAc synthesis. f, Fold-changes in the intermediate metabolites of uridine

880 synthesis. g, Relative OAA levels, as measured by LC-MS in PCK1 -OE SK-Hep1

881 cells. h, OAA is converted to Asp by the mitochondrial GOT2 enzyme, which can be

882 prevented by the GOT2 inhibitor, AOA. i, Relative levels of UDP-GlcNAc, as

883 measured by LC-MS in PKO cells treated with 1 mM OAA. j, Protein O-GlcNAcylation

884 levels in SK-Hep1 cells cultured for 12 h in medium containing 5 mM glucose and PEP

885 (left), OAA (middle), or Asp (right). k, Relative UDP-GlcNAc levels were measured by

886 LC-MS in PKO cells treated with 20 μM AOA. l,m, Protein O-GlcNAcylation levels in

887 PKO-cells treated with 20 μM AOA for 12 h (l) or transfected with a GOT2 shRNA1/2

888 plasmid for 48 h (m). n, Immunoblots of SK-Hep1 lysates treated for 12 h with OAA (1

889 mM), Asp (1 mM), or AOA (20 μM), as indicated. o, Proliferation ability of SK-Hep1

890 cells treated as described in (N). Data are represented mean ± SD (n ≥ 3 experiments).

891 *p < 0.05, **p < 0.01, ***p < 0.001, Student’s t-test (two groups) or one-way analysis of

892 variance (ANOVA) followed by Tukey’s test (more than two groups).

893

894 Fig. 3 PCK1 Activates AMPKThr172/GFAT1Ser243 Phosphorylation and Inhibits

895 UDP-GlcNAc Biosynthesis. a,b, Representative immunoblots showing AMPKThr172

896 and GFAT1Ser243 phosphorylation in PCK1-OE cells (a) and PKO cells (b). c,d,

897 Immunoblot analysis. PKO cells were treated with metformin (Met, 2 mM) for 12 h (c) 45

898 or transfected with an AMPK shRNA1/2 plasmid (d). e-h, Hepatoma cell growth

899 curves and colony formation capacity. PKO cells and PCK1-OE cells were treated as

900 indicated. i, Relative levels of UDP-GlcNAc in PKO cells treated for 24 h with the

901 GFAT1 inhibitor DON (20 μM), as measured by LC-MS. j, PKO cells were treated with

902 20μM DON for 24 h. k, Working model whereby PCK1 ablation promotes

903 UDP-GlcNAc biosynthesis, O-GlcNAcylation, and proliferation of HCC cells through

904 increased oxaloacetate accumulation and activation of the AMPK-GFAT axis. Data are

905 represented as mean ± SD (n ≥ 3 experiments). *p < 0.05, **p < 0.01, ***p < 0.001, as

906 determined using Student’s t-test (two groups) or one-way ANOVA, followed by

907 Tukey’s test (more than two groups).

908

909

910 Fig. 4 PCK1 Deficiency Promotes CHK2 O-GlcNAcylation at T378. a,b,

911 IP-LC-MS/MS analysis of O-GlcNAc-modified proteins. (a) Flowchart describing the

912 processes used for IP-LC-MS/MS analysis. (b) Kyoto Encyclopedia of Genes and

913 Genomes-based analysis of significantly enriched pathways represented by proteins 47

914 that bound to Flag-tagged OGT. PKO cells were transfected with OGT-Flag, and cell

915 lysates were immunoprecipitated using anti-Flag agarose beads. c,d, Co-IP of

916 OGT-HA and CHK2-Flag was examined using an anti-HA antibody (c) or an anti-Flag

917 antibody (d). e, Co-IP of endogenous OGT and CHK2 in PKO cells. f, Subcellular

918 co-localization of OGT and CHK2 in PKO cells was determined by

919 immunofluorescence staining. g, Schematic representation of the CHK2 constructs.

920 WT CHK2 contains three domains, including a SQ/TQ cluster domain, a

921 Forkhead-associated (FHA) domain, and a kinase domain. Truncation mutants of

922 CHK2, comprising amino acids (aa) 69–543 or 1–221, were designated as ΔN and ΔC,

923 respectively. h, Interactions between OGT and full-length (aa 1–543), the ΔN

924 truncation mutant (aa 69–543), or the ΔC truncation mutant (aa 1–221) in HEK293

925 cells were determined by Co-IP. i, CHK2 IP with anti-Flag M2 agarose beads in

926 HEK293 cells transfected with a Flag-CHK2-expression construct or a vector control.

927 k, PKO cells were treated with 50 μM PUGNAc or 50 μM ST for 24 h, incubated in 5

928 mM glucose, and followed by a sWGA pull-down assay. Western blot was determined

929 by anti-CHK2. k-m, Cell lysates of PCK1-OE cells (k), PKO cells (l), or SK-Hep1 cells

930 treated with 1mM PEP or OAA (m) were immunoprecipitated with anti-Flag agarose

931 beads and immunoblotted, as indicated. n, LC-MS analysis of CHK2-Flag identified

932 residue T378 as the CHK2 O-GlcNAcylation site, which corresponded to the

933 O-GlcNAcylated CHK2 peptide ILGETSLMR. o, HEK293 cells were transfected with

934 vectors encoding Flag-tagged versions of WT CHK2, T378A CHK2, or T383A CHK2.

935 Cell lysates were purified using anti-Flag M2 agarose beads and probed with an

936 anti-Flag or anti-O-GlcNAc antibody. Quantitative analysis of the

937 O-GlcNAcylation/Flag (IP) band was performed using Image-pro plus software. p,

938 Cross-species sequence alignment of CHK2.

939

940

941 Fig. 5 O-GlcNAcylation at T378 Stabilizes CHK2 and Activates Its Downstream

942 Targets. a, CHK2 ubiquitination in PKO cells in the presence of HA-tagged ubiquitin

943 (Ub-HA). b-e, Half-life and quantitative analysis of Flag-tagged WT CHK2 (b,c) and

944 T378A mutant CHK2 (d,e) in PKO cells. Cells were treated with 40 μM cycloheximide 50

945 (CHX) for the indicated time, and CHK2 levels was analyzed by immunoblotting. f-i,

946 Representative immunoblots of CHK2, p-Rb, p-CDK2, and p27 expression in

947 PCK1-OE cells (f), PKO cells (g,i), or SK-Hep1 cells following the indicated

948 treatments (h). j,l, CKO cells (CHK2-knockout SK-Hep1 cells) were transfected with

949 vectors CHK2-Flag or T378A-Flag, followed by treatment with 50 μM PUGNAc for 24

950 h (j) or 1 mM OAA for 12 h (l). Cells were lysed and analyzed by western blotting. k,

951 PCK1/CHK2 double-knockout PLC/PRF/5 cells (PKO/CKO cells) were transfected

952 with a CHK2-Flag or T378A-Flag expression vector, followed by Immunoblotting.

953

954

955 Fig. 6 O-GlcNAcylation Promotes DEN/CCl4-Induced Hepatocellular

956 Carcinogenesis in PCK1-Knockout Mice. a, Reproductive strategy for generating

957 AlbCre(-/-), Pck1(flox/flox) (WT), and AlbCre(+/-), Pck1(flox/flox) (liver-specific knockout, LKO)

958 mice. b, PCK1 protein expression in WT and LKO mouse organs involving the heart, 52

959 liver, spleen, and kidney were confirmed by immunoblotting. c, Schematic

960 representation of the experimental procedures used with WT and LKO mice. Mice

961 were injected intraperitoneally with 75 mg/kg DEN or 4% CCl4 (every 3 days) as

962 indicated, followed by combined administration of 5 mg/kg AOA or 1 mg/kg DON

963 (twice per week) for 16 weeks. Control mice were provided a normal diet (ND). d-f,

964 Gross appearances (d) and hematoxylin and eosin staining (e) of liver samples with

965 tumors, and the numbers of tumor nodules (f). n = 6/group. The yellow dotted-line

966 circles represent tumors. g,h, AST (g) and ALT (h) levels in mouse serum samples (n

967 = 6/group). i, Glycolysis-metabolite proﬁles, derived from liver tumors of WT or LKO

968 mice, were determined by performing LC-MS/MS metabolomics assays. j, Relative

969 UDP-GlcNAc levels (n = 6/group). k,l, The indicated proteins in liver tumors were

970 assessed by immunohistochemical labeling (k, scale bars: 100 μm) and western

971 blotting (l).

972

973

974 Fig. 7 PCK1 Deficiency Strengthens Protein O-GlcNAcylation and Correlates

975 with Human HCC Growth. a, IHC staining of PCK1, O-GlcNAcylation, and Ki67 in

976 clinical HCC samples. Scale bars: 100 μm. b, Representative human HCC samples

977 were indicated by immunoblots. c, Relative PCK1 protein-expression levels were 54

978 compared between non-tumor tissues (NT) and tumors (T) from 40 patients with HCC

979 (see also Figure S6). Relative protein-expression levels were normalized to those in

980 NT samples. d, The correlation between HCC tumor sizes (n = 40) and PCK1

981 expression. e, Relative O-GlcNAc levels of proteins in samples from 40 patients (see

982 also Figure S6). f, Correlation analysis between PCK1 expression and O-GlcNAc

983 levels (n = 40). g,h, Analysis of CHK2 O-GlcNAcylation in HCC tumors by performing

984 sWGA pull-down assays (g). CHK2 O-GlcNAcylation levels were quantified (h). i,

985 Correlation analysis between PCK1 and p-Rb expression in tumor tissues from 20

986 patients with HCC. Data are represented as mean ± SD. P values were derived from

987 Pearson’s correlation coefficient (r). j, Molecular Model for the role of PCK1 Deficiency

988 in regulating CHK2 O-GlcNAcylation and HCC Growth upon low glucose. In a

989 low-glucose microenvironment, PCK1 ablation promotes oxaloacetate accumulation

990 and GFAT1 activation to increase UDP-GlcNAc synthesis through the

991 hexosamine-biosynthesis pathway. Increased O-GlcNAc modification enhances

992 Thr378 O-GlcNAcylation in CHK2, which leads to its dimerization and Rb

993 phosphorylation, and HCC cell proliferation. Inhibitors AOA and DON suppress HCC

994 growth, indicating a unique potential for targeting O-GlcNAc signaling in the treatment

995 of HCC.