Amd1-skp2 cascade regulates hepatocyte proliferation during liver growth and hepatocellular carcinoma development in zebrafish

21 Zebrafish; Liver growth; Hepatocyte proliferation; amd1; skp2 22 23 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 2

24 Rapid hepatocyte proliferation is a common characteristic of liver growth in early liver 25 development and hepatocellular carcinoma (HCC). Clarifying the mechanism how liver growth is 26 regulated in development would benefit studying HCC progress. Although many genes have been 27 reported to be involved in liver growth during early development, the mechanism underlying liver 28 growth is far from being elucidated. Here we found that the liver undergoes rapid growth from 29 2dpf to 5dpf in zebrafish. Comparing the transcriptome for different staged hepatocytes identified 30 813 hepatocyte enriched genes at 3 dpf. Among them, S-adenosylmethionine decarboxylase 31 proenzyme (AMD1) had not been previously reported to be involved in liver development. Our 32 study further confirmed that amd1 was highly enriched in hepatocytes during liver rapid growth 33 and that amd1 mutation inhibited liver growth mainly by repressing hepatocyte proliferation. 34 Mechanistically, amd1 loss of function downregulated the expression of skp2 , and skp2 is required 35 for amd1 to regulate hepatocyte proliferation. Furthermore, the role of Amd1-Skp2 cascade in 36 regulating hepatocyte proliferation was also conserved in a zebrafish HCC model. In conclusion, 37 our study systematically identified some uncharacterized genes possibly being involved in 38 regulating liver growth during zebrafish development, and revealed the role of amd1-skp2 cascade 39 in regulating hepatocyte proliferation during liver growth and HCC progr ession. This work also 40 provided a database which would benefit elucidating the mechanism how hepatocyte proliferation 41 is regulated during liver growth and HCC progress. 42 43 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 3

44 The liver is one of the largest and critical internal organs that fulfils several critical functions, 45 including metabolism, secretion, detoxification, and homeostasis (Li et al., 2021) . Liver 46 organogenesis involves multiple processes including ventral foregut en doderm migration, 47 hepatoblast specification, liver budding, hepatoblast differentiation and the final step of liver 48 growth and morphogenesis (Cheng et al., 2006; Field et al., 2003; Khaliq et al., 2015). In recent 49 decades, the mechanisms by which liver development is regulated have been broadly studied. 50 During this process, many factors, including fibroblast growth factors (FGFs), bone 51 morphogenetic proteins (BMPs), Wnt signalling, hematopoietically expressed homeobox ( Hhex) 52 and ( prospero-related homeodomain protein 1) Prox1, have been reported to be involved in 53 hepatoblast specification and differentiation (Campbell et al., 2021; Goessling and Stainier, 2016; 54 Tachmatzidi et al., 2021). In the later liver growth stage, klf6 (Zhao et al., 2010), Hnf4 α (Zhao et 55 al., 2018), oestrogen (Chaturantabut et al., 2020), hdac3 and Hippo signalling (Cox et al., 2016; 56 Cox et al., 2018; Wu et al., 2022) have been reported to be specifically involved in hepatocyte 57 proliferation (Ober and Lemaigre, 2018). Among all these factors, Yap seems to be most critical 58 for hepatocyte proliferation and liver size control (Avruch et al., 2011; Wu et al., 2022). Although 59 the mechanism underlying hepatocyte proliferation and liver growth has been broadly addressed, 60 the detailed mechanism is far from being elucidated. 61 Recently, bulk RNA sequencing (RNA-seq) or single-cell RNA sequencing (scRNA-seq) have 62 been used to study liver differentiation (Camp et al., 2017; Gao et al., 2022; Segal et al., 2019; 63 Yang et al., 2017), regeneration (Ben-Moshe et al., 2022; Wang et al., 2019) and liver disease 64 (Kimura et al., 2022; Saviano et al., 2020), but no study has screened genes highly expressed in 65 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 4 hepatocytes during liver growth. To identify the genes involved in liver growth, we performed 66 RNA-seq analysis to screen out liver-enriched genes at 3dpf using sorted GFP-labelled 67 hepatocytes from embryos at 3dpf, 7dpf and adult zebrafish livers. Many novel genes, including 68 S-adenosyl methionine decarboxylase proenzyme (AMD1), were identified, and their role in liver 69 development had not been addressed. AMD1 is one of the key enzymes involved in the synthesis 70 of polyamines (Bian et al., 2021; Pegg, 2009). During embryonic development and tissue 71 regeneration, it is associated with embryonic stem cell（ ESC） self-renewal, cell proliferation and 72 cell migration (James et al., 2018; Lim et al., 2018; Zhang et al., 2012; Zhao et al., 2012). In 73 pathological conditions, AMD1 is enriched in multiple cancers, including prostate cancer and 74 neuroblastoma (Evageliou et al., 2016; Zabala-Letona et al., 2017), and AMD1 was also reported 75 to be a potential target for tumour therapy since AMD1 blockade inhibits neuroblastoma 76 progression. In hepatocellular carcinoma (HCC), AMD1 was enriched in human HCC tissues, and 77 AMD1 increased HCC metastasis by stabilizing the interaction of IQGAP1 with FTO (Bian et al., 78 2021). Although the role of AMD1 in ESC self-renewal and tumour progression has been reported, 79 whether AMD1 has a critical role in regulating liver growth has not been addressed. 80 Since Amd1 –/– mouse embryos die between E3.5 and E6.5 days post-coitus (Nishimura et al., 81 2002), the role of AMD1 in liver development could not be evaluated in mouse embryos. To study 82 the role of amd1 in liver development during early embryogenesis, we generated two zebrafish 83 amd1 mutants using the CRISPR/Cas9 method. Our data showed that amd1 loss of function 84 inhibited liver growth. Mechanistically, amd1 loss of function decreased the expression of S-phase 85 kinase-associated protein 2 ( skp2), a key component of the SKP1-cullin 1-F-box (SCF) complex 86 (Wu et al., 2021; Zhang et al., 1995), and the downregulation of skp2 in the amd1 mutant 87 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 5 enhanced the phenotype with decreased hepatocyte proliferation. In addition to the role of amd1 in 88 liver growth, our data also suggested that amd1 was involved in HCC progression and that this 89 role was also partially mediated by skp2. In conclusion, our data revealed some uncharacterized 90 and liver-enriched genes during liver growth and further clarified the role of the amd1-skp2 91 cascade in hepatocyte proliferation, including in liver development and hepatocellular carcinoma 92 (HCC). 93

and materials 94 Ethics statement 95 All experimental methods and protocols were approved by Chengdu Medical College (Sichuan, 96 China). Zebrafish were maintained in accordance with the Guidelines of the Animal Care 97 Committee of Chengdu Medical College. 98 Fish and Fish Maintenance 99 Wild-type (AB), transgenic line Tg(fabp10a:RTTA),Tg(Tetre:EGFP-kras_G12V) (Nguyen et al., 100 2016), Tg(fabp10a:GFP), skp2+/- and amd1+/- line fish were maintained in standard conditions at 101 approximately 28.5 °C. The developmental stages were characterized as previously described 102 (Kimmel et al., 1995). 103 Amd1 mutant construction 104 One sgRNA was designed according to the exon 1 region of the amd1 gene, and the sgRNA was 105 synthesized in vitro as described in the manual (HiScribe™ T7 High Yield RNA Synthesis Kit ，106 NEB, NO. E2040S), the sgRNA and Cas9 protein were mixed and coinjected into the cell 107 cytoplasm at the one-cell stage. Several injected embryos were used to evaluate the mutation 108 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 6 efficiency using sequencing, and the remaining embryos were designated F0 (Founder). F0 adult 109 fish was crossed with wild type to obtain F1 embryos. After evaluating the mutation efficiency 110 using part of F1 embryos, the remaining F1 was maintained and grown to produce F2 embryos. 111 Two types of useful mutations were identified in F2. The targeted sequence of amd1 sgRNA1 was 112 5’- GGAGGTGTGGTTCTCCCGGC-3’. The primers used for amplifying the targeted genomic 113 DNA (used for sequencing) were amd1-check-F: 5 ′ - GTGA TTCCATCCGACGGTTTA -3 ′ 114 andamd1-check-R: 5′ - CTGACATTATCACAGCGTTTCAC -3′ . 115 Bulk RNA Sequencing 116 To compare the transcriptome of hepatocytes in different liver stages, GFP-labelled hepatocytes 117 were sorted using flowcytometry (Moflflo XDP, Beckman) from transgenic line Tg(fabp10:GFP) 118 embryos and adult livers. Approximately 400 hepatocytes were collected for each stage. cDNA 119 libraries were generated from these sorted cells using the Smart-seq2 protocol. RNA sequencing 120 was performed using the PE100 strategy (HiSeq 2500, Illumina). Sequencing data were analysed 121 as previously reported (Liu et al., 2016). To compare the transcriptome of wild-type and amd17-/- 122 embryos at 4 dpf, total RNA was prepared using TRIzol according tothe manual. RNA sequencing 123 and analysis were performed by Novogene Co., Ltd. (Tian Jin, China). 124 Fin-clip and identifying amd1 mutation embryos 125 Since there is no clear morphological phenotype for amd1-/- embryos, to identify amd1 126 homozygotes, the tail fin of zebrafish larvae was cut at 48 hpf, and genomic DNA was 127 individually prepared as following: After anaesthetizing the embryo, the tip of the tail was cut with 128 a scalpel to prepare genomic DNA as previously reported (Wilkinson et al., 2013), and the 129 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 7 embryos were kept for further experiments. Genomic DNA was used to amplify the target region 130 using PCR. Then, we evaluated whether the larvae were homozygotes according to the PCR 131 results. For the wild-type larvae and heterozygotes, the target fragment was obtained, while for the 132 homozygotes, the target fragment was not amplified, no amplicon was detected. The primers used 133 here were as follows: amd1-screen-F: 5’- GTGGTTCTCCCGGCAG-3’, amd1-screen-R: 5 ′ - 134 CTGACA TTA TCACAGCGTTTCAC -3′ . 135 Cas9-sgRNA ribonucleoprotein complex (RNP) preparation and injection, skp2 mutant 136 generation 137 To obtain mosaic mutants in F0 embryos, we optimized the CRISPR-cas9 gene editing process 138 according to a previous report (Wu et al., 2018). In brief, three sgRNAs for skp2 were designed 139 and synthesized in vitro using HiScribe™ T7 High Yield RNA Synthesis Kit(NEB, E2040S). The 140 Cas9-sgRNA ribonucleoprotein complex (RNP) solution was prepared as follows: Cas9 protein 141 (EnGen® Spy Cas9 NLS, NEB, M0646T) 1.3 μ l, 1 M KCl 1.1 μ l, total sgRNA 2500 ng (830ng 142 each), phenol red 0.3 μ l. Finally, RNA-free H 2O was added to 5 μ l total, mixed completely, 143 incubated at 37 °C for 5 minutes, and then placed back into an ice bath for the following injection 144 procedure. The RNPs were injected into the embryo yolk within 15 minutes after fertilization. To 145 examine the mutant efficiency for skp2, 16 injected embryos were used to prepare genomic DNA, 146 and semi-quantitative RT /i4 PCR was performed for evaluation as described in a previous report 147 (Zhu et al., 2019). The primers used are shown in Table S1. To get the stable skp2 mutant lines, the 148

being used to screen amd1 mutant was used to screen skp2 mutant. In our work a 149 frameshift mutant line was obtained. The detailed information was provided in the result section. 150 Chemical treatment 151 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 8 SMIP004 was used to inhibit the function of skp2 as described in a previous report (Li et al., 2019). 152 Two concentrations, 40 μ M and 80 μ M, were selected to inhibit skp2 activity. The chemical 153 SMIP004 was diluted with egg water to the concentration described above. The embryos were 154 incubated with SMIP004 solution (40 μ M or 80 μ M) from 3 hpf to 24 hpf to evaluate which 155 concentration was the proper concentration. Then, the embryos were incubated with SMIP004 156 solution (proper concentration: 80 μ M) from 48 hpf to the stages needed. 157 Plasmid Construction 158 Total RNA was extracted following the manufacturer’s instructions (TRIzol, Ambion, 15596-026). 159 cDNA was prepared using a Revert Aid First Strand cDNA Synthesis Kit (Fermentas, K1622) 160 according to the manufacturer’s instructions. The CDs of amd1 and skp2 were amplified 161 individually using PCR (Prim STAR Max Premix Takara, R045A) and cloned into the PCS 2+ 162 vector (5x In-Fusion HD Enzemy Premix, Takara, 639649). The primers for cloning were as 163 follows: PCS 2+_F: 5 ′ -CTCGAGCCTCTAGAACTA TAGTG-3 ′ , PCS 2+_R: 5 ′164 -TGGTGTTTTCAAAGCAACGATATCG-3 ′ , amd1-pcs2+_F: 5 ′165 -TCTTTTTGCAGGA TCGGAGTCTGTTTGTCTCACGATGG-3 ′ , amd1-pcs2+_R: 5 ′166 -GTTCTAGAGGCTCGACGCTTCTTCATGTCAGAGGATCAG-3 ′ , skp2-pcs2+F:5 ′167 -GCTTTGAAAACACCACAAGTCAGGATGTCAAACGAAAGG-3 ′ , skp2-pcs2+_R: 5 ′168 -GTTCTAGAGGCTCGAGCATTAATGTTTGTAGACGAGTCTGC-3′ . 169 mRNA injection 170 skp2 mRNA was synthesized in vitro using an mMESSAGE Kit (Ambion, AM1340) as the 171 described in the manual. The concentration for skp2 mRNA injection was 40ng/ μ l. skp2 mRNA 172 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 9 was injected at the 1-4 cell stage. 173 RT/i1qPCR 174 RT/i4 qPCR was performed using the Brilliant III Ultra-Fast SYBR Green QPCR Master Mix 175 (Agilent Technologies) and the CFX96 Real-Time System (BIO-RAD) according to the 176 manufacturer’s instructions. The amount of beta-actin was used for normalizer. The primers are 177 listed in Table S1. All experiments were repeated at least 3 times. 178 Whole-mount in situ hybridization and section 179 One color in situ hybridization was performed as described in a previous study (Liu et al., 2019). 180 The previous probes fabp10, prox1, hhex, and fabp2 were used as described in previous reports 181 (Zhang et al., 2022). The CDs of amd1 and skp2 were amplified using PCR and cloned into the 182 vector pcs2 +, then linearized the plasmids and synthesized the individual antisense probe as 183 previously reported (Zhang et al., 2022). Two color in situ hybridization was performed as 184 described in a previous study (Dunn et al., 2022). Specifically, digoxygenin-labeled amd1 probe 185 and fluorescein - labeled uox probe were used in our study. For sectioning, in situ hybridized 186 embryos were re-fixed in 4% paraformaldehyde in PBS, followed by incubation in 15 and 30% sucrose 187 in 0.1% Tween/phosphate buffered saline for 2 hours each. Then, embryos were mounted in 1.5% 188 agarose in 30% sucrose, and balanced in 30% sucrose solution overnight at 4 °C. The mounted embryo 189 was re-mounted in OCT (Sakura), sectioned using a CM1850 cryostat (Leica), 190 Immunostaining 191 The embryos were fixed overnight with PFA (4% in PBS) at 4 °C, washed with PBS (5 min, 3x) 192 and blocked with PBTN (4% BSA, 0.02%NaN 3, in PT) for 2 hours at 4 °C. Then, the primary 193 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 10 antibody against H3p (GTX128116) or Caspase3 (BD 559565) was diluted with PBTN at 1:200 194 and incubated on a shaker at 4 °C overnight. Then, the embryos were washed with PT (0.3% 195 Triton-X-100, in 1X PBS) for at least 20 min 8 times. The secondary antibody, Donkey anti rabit 196 IgG, Texas Red coupled; GeneTex 26800) or Alexa FluorTM 647 （ invitrogen, A21244 ） was 197 diluted with PBTN in 1:500 and added. The embryos were incubated overnight at 4 °C (kept in the 198 dark). Finally, the embryos were washed with PT more than 8 times (30 min each time) and 199 imaged. 200 EDU experiment 201 BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 594 (Beyotime, C0078S) was used for 202 this experiment. 200uM EDU solution containing 2%DMSO and 0.01% phenol red was prepared. 203 Zebrafish embryos were mounted with low melting point agar (0.8%) and the EDU solution was 204 injected pericardially. After injection, the embryos were replaced in egg water at 28.5 ℃ for 40 205 minutes, then fixed them overnight at 4 /i4 with PEM. The fixed embryos were rinsed with PBS for 206 3 times (5 min each time) and treated with pre-cooled acetone at -20 /i4 f o r 4 0 m i n , t h e n w e r e 207 washed by PT for 3 times (20 minutes each time). Next the embryos were incubated with 3%BSA 208 at room temperature for 2 hours and washed with PT 3 times, then according the manual the EDU 209 reaction solution was added and keep the reaction for 30 minutes in dark. After EdU staining 210 reaction the embryos were washed with PT for 3 times, flowing Immunostaining for GFP. 211 TUNEL staining 212 Embryos were fixed in 4% PFA overnight at 4 °C, washed with PBST 3 times (10 minutes each 213 time) and stored in 100% methanol overnight. Then, the embryos were washed 3 times with PBST 214 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 11 3 times, and an In Situ Cell Death Detection Fluorescein kit (Roche11684795910) was applied to 215 examine cell apoptosis according to the manufacturer’s instructions. 216 Microscopy 217 Images of whole-mount in situ hybridized embryos (mounted in 80%-100% glycerol) and section 218 samples were captured at room temperature using an OLYMPUS SZX16. To examine positive 219 pro-apoptotic, proliferating cells in the liver, the Tg(fabp10:EGFP) embryos were fixed in PFA (4% 220 in PBS) overnight and mounted in 1.5% Low Melting-point Agar. Then, the proliferating cells and 221 apoptotic cells were captured at room temperature using an OLYMPUS FLUOVIEW FV1000. 222 Statistical analysis 223 The data were analysed with Novoexpress, ImageJ, statistical software in GraphPad Prism 8 for 224 Windows (GraphPad Software). Quantitative data are presented as the means S.D. Experiments 225 were performed at least three times for each experiment. NS, not significant, “*” p < 0.05, “**” p 226 < 0.01, “***” p < 0.001 and “****” p < 0.0001. 227

228 Amd1 was highly expressed in hepatocytes during rapid liver growth 229 Early studies used microarrays to screen out some liver-enriched genes, and their roles in mouse 230 and zebrafish liver development were proven (Cheng et al., 2006; Jochheim-Richter et al., 2006; 231 Petkov et al., 2004), while many genes regulating liver growth were not identified using this 232 method. Recently, bulk RNA-seq and sc-RNA-seq have been used for the study of liver 233 regeneration and liver disease (Ben-Moshe et al., 2022; Saviano et al., 2020; Wang et al., 2019), 234 but these methods have not been used to screen genes regulating liver growth. We found that 235 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 12 during early zebrafish development, the liver undergoes rapid growth from 2dpf to 5dfp (Fig. 1 A, 236 B) and hypothesized that genes regulating live growth should be highly expressed in hepatocytes 237 during this stage. To identify genes being involved in liver growth, we performed RNA-seq using 238 GFP-labelled hepatocytes sorted at different stages (Fig. 1C): 3dpf, 7dpf and adult zebrafish. The 239

showed that at 3dpf, 1654 genes were enriched in the hepatocytes, which was more than 240 2-fold higher than that in non-hepatocytes (Table S2). Among the top 25 genes of the 1654 genes, 241 16 genes had been reported to be liver-enriched genes in the Zfin database, 3 genes were reported 242 in previously published literature, and 8 genes were not reported previously (Fig. 1 D and Table 243 S3), implying that some new liver-enriched genes were identified. Among all 1654 liver-enriched 244 genes, 262 genes were highly expressed in adult hepatocytes, which was 2-fold higher than that in 245 3dpf hepatocytes (Table. S4). Detailed analysis revealed that among the top 20 genes of the 262 246 genes, only 4 genes were not reported in early literature or in the zfin database (Table S5). In 247 contrast, among the 1654 liver-enriched genes at 3dpf, 813 genes were highly expressed in 3dpf 248 hepatocytes, which was more than 2-fold higher than that in adult hepatocytes (Table S6). To 249 further confirm this result, 4 genes were randomly selected from the 813 genes, and their 250 expression was evaluated in 3dpf hepatocytes and adult hepatocytes using RT /i4 qPCR. The results 251 showed that the expression of all of these genes was higher in 3dpf hepatocytes than in adult 252 hepatocytes (Fig.1 E). These results implied that among the liver-enriched genes at 3dpf, the 253 expression of most of them was higher in 3dpf hepatocytes than in adult hepatocytes, and they 254 could be candidates for liver growth regulation. 255 Since the ratio of liver growth was significantly decreased from 5dpf (Fig. 1A, B), we 256 hypothesized that the expression of candidates regulating liver growth should be down regulated 257 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 13 in hepatocytes at 7dpf compared with that in hepatocytes at 3dpf. Then, we analyzed the 258 expression of 813 genes in hepatocytes at 3dpf and 7dpf. Indeed, the expression of 756 genes was 259 decreased in hepatocytes at 7dpf (Table. S7). This result further indicated that the genes being 260 identified here could be candidates for liver growth control. Next, we performed KEGG analysis 261 to determine the signalling pathways associated with these genes. Interestingly, KEGG analysis 262 showed that 4 genes related to arginine and proline metabolism were highly expressed in 263 hepatocytes at 3dpf, and this result was confirmed by RT /i4 qPCR experiments (Fig. 1F). Among 264 these 4 genes, nos1 has been reported to be involved in liver growth after injury (Cox et al., 2014); 265 amd1, the key enzyme involved in the synthesis of polyamines (Pegg, 2009), has even been 266 reported to be associated with ESC self-renewal and cell proliferation (James et al., 2018; Lim et 267 al., 2018; Zhang et al., 2012; Zhao et al., 2012), but its role in liver development has not been 268 addressed. We examined the role of amd1 during liver growth to confirm that the genes we 269 identified could be candidates for liver growth. 270 The detailed expression pattern of amd1 in early zebrafish development 271 To evaluate the possible role of amd1 in liver growth, we first examined the expression of amd1 in 272 early zebrafish embryos using RT /i4 PCR. The data showed that amd1 was expressed in all staged 273 embryos (Fig. 2A). To evaluate the detailed expression pattern of amd1 during embryogenesis, we 274 prepared amd1 antisense/sense probes to determine its expression from 3 hours post fertilization 275 (hpf) to 4dpf (Fig. 2B). The data showed that amd1 was a maternal factor (Fig. 1 Bb1) and was 276 expressed ubiquitously before 24 hpf (Fig. 2Bb1-b4). After 24 hpf, amd1 was restricted to head 277 and endodermal cells (Fig.2Bb5-b7). Importantly, amd1 was highly expressed in the liver at 3 dpf 278 and 4 dpf (Fig.2Bb6-8; Bb9-b11); this result was consistent with the RNA-seq data. In add ition, 279 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 14 detailed analysis of the liver growth ratio revealed that the liver undergoes the most rapid growth 280 from 3 dpf to 4 dpf (Fig. 1 A, B), from 4 dpf, the ration of liver growth was decreased (Fig. 1B). 281 Therefore, we hypothesized that the amd1 expression level should be higher in hepatocyte in 3 dpf 282 than that in 4 dpf. To evaluate this hypothesis, the relative expression level of amd1 in liver 283 hepatocytes at 3 dpf and 4 dpf was evaluated using sorted GFP-labelled hepatocytes at 3 dpf and 4 284 dpf (Fig. 2C, D). The data showed that the expression of amd1 in hepatocytes in 3 dpf was higher 285 than that in 4 dpf (Fig. 1. D). Here, the correlation between high expression of amd1 and most 286 rapid liver growth at 3 dpf strongly suggested the possible role of amd1 in liver growth. 287 Amd1 mutant generation 288 To explore the role of amd1 in zebrafish liver development, we used the CRISPR /i4 Cas9 method to 289 construct the amd1 mutant line as described in a previous report (Kroll et al., 2021). We designed 290 one sgRNA for the target sequence that localized in exon 1 of amd1 to construct the mutant lines 291 (Fig. S1A). In F1 adult fishes, we screened two useful mutant lines: in mutant 1, the base group 292 “CCCG” was changed to “TT”, and this exchange led to a prestop codon at 168 bp in the CD 293 region (we named it amd1168); in mutant 2, seven bases “CGGCAGG” were deleted, and a prestop 294 codon arose at 141 bp in the CDs region (we named it amd17) (Fig. S1B, C). Since the genome 295 typing of amd17-/- could be performed using PCR, we planed to use the amd17-/- mutant line for 296 most of the future experiments. Comparing the expression of amd1 in mutants and wild-type 297 embryos at 4 dpf suggested that the expression of amd1 was decreased in amd1 7-/- embryos (Fig. 298 S1D). This result indicated that the premature mutant amd1 mRNA was not stable and degraded 299 during embryogenesis (Liu et al., 2019). Comparing the embryonic development between 300 wild-type and amd1 7-/- embryos, no distinct difference was discovered during early development 301 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 15 (Fig. S1E, F), but the homozygotes were not viable and could not grow up to adult. This 302 phenotype is similar to that in Amd1 –/– mouse embryos, in which Amd1 –/– mouse embryos died at 303 the early developmental stage (Nishimura et al., 2002). All these results implied the critical role of 304 amd1 in embryonic development, and our amd1 mutant lines could be used to address the role of 305 amd1 in liver development. 306 Liver growth was repressed in amd1 mutants 307 To observe the role of amd1 in liver development easily, we crossed amd17+/- with the 308 Tg(fabp10:GFP) transgenic line to obtain amd1 heterozygotes with the Tg(fabp10:GFP) 309 transgenic background. Then, these fish were incrossed to obtain amd1 homozygote embryos, and 310 the liver phenotype was evaluated. The data showed that from 3 dpf to 5 dpf, the liver in amd17-/- 311 embryos was smaller than that in controls (Fig.3Aa1-a6, B). We also examined the liver size 312 between amd1 mutants and controls in non-transgenic embryos, the liver size was also smaller in 313 amd17-/- embryos at 3 dpf (Fig. 3C- F). Being similar, we also observed that the liver in amd1168-/- 314 embryos was smaller than that in control (Fig. S2A-D). To further examine whether amd1 was 315 involved in liver specification during early embryogenesis, the early markers prox1 and hhex were 316 evaluated at 48 hpf (Shin et al., 2007). The data showed that the expression of prox1 and hhex was 317 intact in amd1 mutants in the early stage (Fig. S3A, B). These results demonstrated that amd1 loss 318 of function only inhibited liver growth while not disturbed hepatocyte specification. Since amd1 319 was also expressed in the gut at 3 dpf and 4 dpf (Fig. 2Bb7-b11), we examined whether gut 320 growth was repressed in amd1 mutants. The results showed that the gut marker fabp2 (Parmar and 321 Wright, 2013) was downregulated in amd1 mutants (Fig. S4 A, B), implying that amd1 also plays 322 a critical role in gut development. 323 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 16 Hepatocyte proliferation was inhibited in amd1 mutant embryos 324 During liver development, both hepatocyte proliferation delay and apoptosis increase were 325 reported to give rise to a smaller liver (Chen et al., 2005; Chu and Sadler, 2009; Ma et al., 2019). 326 To observe the underlying reason why the size of liver was reduced in amd1 mutants, we 327 examined hepatocyte proliferation and apoptosis in amd1 7-/- embryos with a Tg(fapb10a:GFP) 328 transgenic background. H3P immunostaining showed that hepatocyte proliferation was decreased 329 in amd17-/- embryos (Fig. 3G, H), but TUNEL experiment and Caspase3 immunostaining showed 330 that the apoptosis of hepatocytes was not increased significantly in amd1 7-/- embryos (Fig. S5B-E). 331 To further confirm the role of amd1 in regulating proliferation, the proliferating related markers 332 were examined using RT-qPCR and the data demonstrated that the expression of cdk1 , cdk4, chk1 333 and mcm5 was downregulated in amd1 mutants (Fig. S5A). This result above further demonstrated 334 that amd1 is required for hepatocyte proliferation during liver growth, which is consistent with 335 that being reported in early literature, in which AMD1 is required for EST self-renewal and cell 336 proliferation (James et al., 2018; Zhang et al., 2012). 337 Skp2 expression was downregulated in amd17-/- embryos. 338 Early research showed that AMD1 played multiple roles during normal embryo development and 339 disease, including the regulation of polyamine synthesis and gene transcription (Bian et al., 2021; 340 James et al., 2018; Patel et al., 2018; Zabala-Letona et al., 2017). Mechanistically, AMD was 341 reported to lie downstream of c-Myc, C/EBP/ β and mTORC1 (Snezhkina et al., 2016; 342 Zabala-Letona et al., 2017), and it also lies upstream of MINDY1(James et al., 2018). Therefore, 343 the mechanism by which amd1 regulates hepatocyte proliferation during liver growth is 344 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 17 complicated. To elucidate the detailed role of amd1 in liver growth, we analysed the gene 345 expression in amd1 7-/- embryos at 4 dpf using RNA-seq. The results showed that in amd17-/- 346 embryos, 94 genes were upregulated, and 104 genes were downregulated (Fig. 4A). Among the 347 downregulated genes, KEGG analysis showed that four genes were in the mTOR signalling 348 pathway, including skp2 (Fig. 4B C). Detailed evaluation using RT /i4 qPCR revealed that both skp2 349 and the skp2 downstream gene rho were downregulated (Fig. 4D). In situ experiments further 350 showed that skp2 was also downregulated in amd17-/- embryos on 3 dpf (Fig. 4E), especially in the 351 liver and gut (Fig. 4 E). In addition, we further examined the detailed expression pattern of skp2 in 352 early embryonic development using in situ experiments. The data showed skp2 was a maternally 353 expressed gene (Fig. 4Ff1, f2), it was expressed ubiquitously before 24 hpf (Fig. 4Ff3- f5) and 354 then restricted to the liver, gut, eyes and boundary of the hindbrain and midbrain at 3 dpf (Fig. 355 4Ff6). The enriched expression of skp2 in liver and gut at 3 dpf further implied the role of skp2 in 356 liver growth. In conclusion, all the results above suggested the possibility that skp2 is required for 357 amd1 to regulate liver growth during embryonic development. 358 Skp2 plays a critical role in liver development 359 Skp2, a key component of the SKP1-cullin 1-F-box (SCF) complex (Wu et al., 2021; Zhang et al., 360 1995), largely functions as an oncoprotein (Cai et al., 2020). Previous work showed that Skp2 is 361 involved in cell proliferation, migration, invasion and metastasis in some malignant tumours (Cai 362 et al., 2020; Wang et al., 2012). In HCC, Skp2 was also involved in HCC metastasis (Chen et al., 363 2021; Zhang et al., 2017).Th ese studies above implied the possibility that skp2 is essential for 364 amd1 to regulate liver development in zebrafish. To evaluate this hypothesis, first we generated 365 skp2 mosaic mutation embryos using the CRISPR/Cas9 method (Fig. 5A and Fig. S6A)(Kroll et 366 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 18 al., 2021) and examined liver development in Tg(fabp10:GFP) transgenic lines. After the exon2 of 367 skp2 was edited in mosaic mutants (Fig. S6A-C), the early embryonic development was delayed 368 (Fig. S6D) and liver growth (Fig. S6E) was repressed. Next we generated the stable skp2 mutant 369 line to confirm the role of skp2 in liver growth. We screened out a frameshift mutant line for skp2: 370 in this mutant 80 nucleotides between “AAAAAGCGG” (labeled with green) and 371 “ACAAGAAGG” (labeled with yellow) were deleted, but “AGTA” was inserted between these 372 two sequences (Fig. 5A-C). This mutantion gave rise to a frameshift and there is no stop codon at 373 the terminal region of skp2 CDs. RT-qPCR further showed that the level of skp2 was decreased in 374 skp2 mutant embryos at 3 dpf (Fig. 5D). Further, we found that the skp2 homozygotes did not 375 display distinct external phenotype at early developmental stage (Fig. S7) but were not viable and 376 could not grow up to adulthood. The data also showed that the liver size was decreased (Fig. 5Ee2, 377 e4, e5 and Fig. S8A-D), as well the hepatocytes proliferation was decreased (Fig. 5F), 378 demonstrating the critical role of skp2 in liver growth. Finally, we examined whether 379 over-expressing skp2 by injection of skp2 mRNA could restore the liver development in amd1 380 mutant. The data showed that, overexpression of skp2 partially increased the liver size in control 381 embryos (Fig. 5Gg1, g2) and restored the liver size in amd1 mutants (Fig. 5Gg3, g4). In 382 conclusion, all these data suggested that skp2 is required liver development, it is also essential for 383 amd1 to regulate liver development. 384 Skp2 is required for amd1 to regulated liver development at liver rapid growth stage 385 Since skp2 was ubiquitously expressed at early stage, we further evaluated whether skp2 is also 386 involved in early liver development. The data showed that the expression of prox1 and hhex was 387 decreased in skp2 mutants at 48 hpf (Fig. S9A, B), implying skp2 is also involved in early liver 388 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 19 development. This result gave rise to a possibility that the liver phenotype is a secondary result of 389 early developmental defects in skp2 mutants. To observe the direct role of skp2 at liver growth 390 stage, we used the skp2 inhibitor SMIP004 (Li et al., 2019) to block the activity of skp2 when the 391 liver undergoes rapid growth (Fig. 6A). During concentration titration, we found that treatment 392 with 40 µM SMIP004 did not lead to gastrulation defects or smaller livers (Fig. 6A-C), while 393 treatment with 80 µM SMIP004 decreased the activity of Skp2, displaying decreased expression 394 of rho and ahcy (Fig. 6 E), the two downstream genes of skp2 (Cai et al., 2020). Then, we used 80 395 µM SMIP004 to treat embryos from 2 dpf to 4 dpf and analysed liver size (Fig. 6A, F). The data 396 showed that skp2 inhibition did not lead to distinc embryonic phenotype (Fig. 6D) but resulted in 397 smaller liver at 4 dpf in Tg(fabp10:GFP) transgenic embryos and wild-type embryos (Fig. 6Gg2, 398 Hh2). This result confirmed that skp2 played a vital role during liver growth. In addition, skp2 399 inhibition decreased hepatocyte proliferation (Fig.6Ii2, i5), confirming the role of skp2 in liver 400 growth. To further evaluate whether skp2 is required for amd1 to regulate liver growth, we 401 determined whether treatment with SMIP004 gave rise to a much smaller liver and much less 402 proliferating hepatocyte in amd17-/- embryos. The data showed that SMIP004 treatment further 403 reduced the size of the liver in amd1 7-/- embryos (Fig. 6Gg4, Hh4), and the H3P labelled cells was 404 decreased (Fig. 6Ii8). In conclusion, all the data above showed that skp2 is reqiured for amd1 to 405 regulate liver growth at liver rapid growth stage. 406 Amd1-skp2 cascade plays a critical role during hepatocyte proliferation in a zebrafish HCC 407 model 408 Both hepatocytes in developing liver and HCC were characterized by rapid cell proliferation 409 (Chaturantabut et al., 2019; Perugorria et al., 2019). The role of amd1 in regulating liver growth 410 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 20 implied that amd1 plays a vital role during hepatocyte proliferation in zebrafish HCC progression. 411 To evaluate this hypothesis, we generated an HCC model in zebrafish larvae as described in 412 previous report ((Yang et al., 2019) and Fig. S10A). In the HCC model, liver-specific 413 overexpression of Kras G12V gave rise to a much larger liver compared with control embryos (Fig. 414 S10Bb3, b6), and the expression of 5 HCC markers was greatly upregulated (Fig. S10C), 415 indicating the HCC model was generated successfully. Next, we sorted GFP-labelled hepatocytes 416 from normal liver and HCC model, then analysed the expression of amd1 (Fig. 7A, B). The data 417 showed that amd1 was upregulated in HCC cells (Fig.7B). Further data showed that amd1 loss of 418 function decreased the size of liver in the HCC model (Fig. 7Cc3, D), and hepatocyte proliferation 419 was also decreased (Fig. 7Ee5, F). These results demonstrated the role of amd1 in hepatocyte 420 proliferation was conserved in zebrafish HCC model. 421 To evaluate whether skp2 mediates amd1 to regulate hepatocyte proliferation in the zebrafish HCC 422 model, first the expression of skp2 was examined. The data showed that the expression of skp2 423 was also significantly upregulated in hepatocytes in the zebrafish HCC model (Fig.7B). Next, we 424 determined whether skp2 was required for zebrafish HCC and whether the liver size was much 425 smaller in amd1 7-/- embryos after blocking skp2 activity. Indeed, the size of the liver was much 426 smaller in amd17-/- embryos treated with SMIP004 from 48hpf to 4dpf in zebrafish HCC model 427 (Fig. 7Cc4, D), and hepatocyte proliferation was also decreased in amd17-/- embryos treated with 428 SMIP004 (Fig. 7Ee8, F). These results demonstrated the possibility that skp2 also is required for 429 amd1 to regulate hepatocyte proliferation in a zebrafish HCC model. 430

431 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 21 Hepatocyte proliferation is one of the critical events during liver organogenesis and HCC 432 progression. Even though many genes have been reported to be involved in this process (Cox et al., 433 2016; Cox et al., 2018; Wu et al., 2022; Yang et al., 2019), the underlying mechanism is far from 434 being elucidated. Here, we screened 813 genes that were highly expr essed in hepatocytes at 3 dpf. 435 Meanwhile, we selected one of them, amd1, to evaluate whether it plays a crucial role during 436 normal liver growth and HCC. We identified the role of amd1 in hepatocyte proliferation during 437 normal liver development and HCC progression and proved that skp2 mediated, at least partially, 438 amd1 to regulate liver growth during liver development and HCC progression. 439 For the genes highly expressed in hepatocytes at 3 dpf, some of them, such as Lats1, Rhbg and 440 Npas2, have been reported to be associated with liver growth or hepatocyte survival (Leibing et al., 441 2018; Yi et al., 2016; Yuan et al., 2017), demonstrating that the genes identified in our research 442 could be the subjects for further studies. Noticeably, in our screening work, some genes were 443 highly expressed in hepatocytes at 3 dpf but not specifically enriched in hepatocytes, possibly 444 these kinds of genes could also be candidates for liver growth regulation. To this point, the 445 evidence was that hdac3, Id2a and oestrogen have been reported to be involved in liver growth 446 (Chaturantabut et al., 2020; Farooq et al., 2008; Khaliq et al., 2015)[2, 10,], but they are not 447 specifically enriched in hepatocytes at 3 dpf. Of course, our RNA-seq work did not identify all the 448 genes regulating liver growth. For example, Yap has been reported to regulate liver growth (Cox et 449 al., 2016; Cox et al., 2018), but it was not found in our gene list. Therefore, more work is needed 450 to further evaluate the role of the genes being identified in liver growth control. 451 Amd1, one of the genes we identified, was previously reported to be associated with ESC 452 self-renewal, cell proliferation and cell migration (James et al., 2018; Lim et al., 2018; Zhang et al., 453 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 22 2012; Zhao et al., 2012), while its detailed role in liver development was not addr essed. Our 454 current data demonstrated that it is crucial for liver growth. When investigating the mechanism 455 underlying, we identified skp2 as one of the key downstream genes that regulate hepatocyte 456 proliferation during liver growth. Even in early mouse embryos, Myc functions as the critical 457 mediator for Amd1 to control ESC self-renewal (Zhang et al., 2012; Zhao et al., 2012), while 458 during zebrafish liver outgrowth, we did not find that the expression of myc was significantly 459 downregulated in amd1 mutants. In contrast, skp2 seems to mediate the role of amd1 in regulating 460 hepatocyte proliferation. These results are consistent with early reports in mice, in which c-Myc 461 was reported to be dispensable for normal liver growth during the postnatal period (Baena et al., 462 2005; Sanders et al., 2012). In addition to the role of amd1 in liver growth in zebrafish 463 development, in a previous study, Amd1 was reported to stabilize the interaction of IQGAP1 with 464 FTO, which upregulating the expression of nanog, kif4 and sox2 to accelerate HCC progression 465 (Bian et al., 2021). Here, we further identified that the Amd1-Skp2 cascade played a vital role in 466 hepatocyte proliferation in a zebrafish HCC model. In our zebrafish HCC model, the RNA-seq 467 data and RT /i4 qPCR showed that both amd1 and skp2 were significantly increased (Fig. 7B), and 468 the liver size was reduced when the role of amd1 or skp2 was blocked (Fig. 7C-F). These data 469 implied that the mechanism by which amd1 regulates hepatocyte proliferation is complicated. 470 In conclusion, our work identified some uncharacterized genes enriched in hepatocytes at 3 dpf. 471 These genes may be candidates for regulating hepatocyte proliferation in normal liver 472 development and HCC. As one of these genes, amd1 was identified to play a crucial role in 473 regulating hepatocyte proliferation via skp2 in normal liver outgrowth and HCC progression. 474 List of abbreviations 475 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 23 HCC: hepatocellular carcinoma; AMD1: S-adenosylmethionine decarboxylase proenzyme; BMPs: 476 bone morphogenetic proteins; RNA-seq: bulk RNA sequencing; scRNA-seq: single-cell RNA 477 sequencing; ESC: embryonic stem cell; skp2 : S-phase kinase-associated protein 2; RNP: 478 ribonucleoprotein complex; hpf: hours post fer tilization; dpf: days post fertilization. 479 Funding 480 This work was supported by the National Natural Science Foundation of China (No. 32070805), 481 the Science and Technology Department of Sichuan Province (2021ZYD0074) and Disciplinary 482 Construction Innovation Team Foundation of Chengdu Medical College: CMC-XK-2102. 483 Informed Consent Statement 484 Not applicable. 485 Data Availability Statement 486 All the data was in the manuscript and supplementary materials. 487

488 We would like to thank Dr Chi Liu to help sorting GFP labelled hepatocytes, Xiaojun Yang to 489 read and gave critical comments. We also would like to thank the members working in our fish 490 facility to help take care of all the fish lines in this study. 491 Conflicts of Interest 492 The authors declare no conflict of interest. The funders had no role in the design of the study, in the 493 writing of the manuscript, or in the decision to publish the manuscript. 494 Author contributions: 495 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 24 Conceptualization: SZH, ZHG; Methodology: SZH, KZ, ZHG ； Investigation: KZ, XYW, BYC, 496 YD, YL; Visualization: SZH, KZ, YD, ZHG； Funding acquisition: SZH ； Project administration: 497 SZH; Supervision: SZH; Writing – original draft: SZH, ZHG, KZ,ML; Writing – review & 498 editing: SZH, ZHG, XDL, ML. 499 Ethics approval and consent to participate 500 All experimental methods and protocols were approved by Chengdu Medical College (Sichuan, 501 China). Zebrafish were maintained in accordance with the Guidelines of the Animal Care 502 Committee of Chengdu Medical College. 503 504 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 25

505 Avruch, J., Zhou, D., Fitamant, J. and Bardeesy, N. (2011). Mst1/2 signalling to Yap: gatekeeper 506 for liver size and tumour development. Br J Cancer 104, 24-32. 507 Baena, E., Gandarillas, A., Vallespinos, M., Zanet, J., Bachs, O., Redondo, C., Fabregat, I., 508 Martinez, A. C. and de Alboran, I. M. (2005). c-Myc regulates cell size and ploidy but is not 509 essential for postnatal proliferation in liver. Proc Natl Acad Sci U S A 102, 7286-91. 510 Ben-Moshe, S., Veg, T., Manco, R., Dan, S., Papinutti, D., Lifshitz, A., Kolodziejczyk, A. A., 511 Bahar Halpern, K., Elinav, E. and Itzkovitz, S. (2022). The spatiotemporal program of zonal liver 512 regeneration following acute injury. Cell Stem Cell 29, 973-989 e10. 513 Bian, X., Shi, D., Xing, K., Zhou, H., Lu, L., Yu, D. and Wu, W. (2021). AMD1 upregulates 514 hepatocellular carcinoma cells stemness by FTO mediated mRNA demethylation. Clin Transl Med 11, 515 e352. 516 Cai, Z., Moten, A., Peng, D., Hsu, C. C., Pan, B. S., Manne, R., Li, H. Y. and Lin, H. K. (2020). 517 The Skp2 Pathway: A Critical Target for Cancer Therapy. Semin Cancer Biol 67, 16-33. 518 Camp, J. G., Sekine, K., Gerber, T., Loeffler-Wirth, H., Binder, H., Gac, M., Kanton, S., 519 Kageyama, J., Damm, G., Seehofer, D. et al. (2017). Multilineage communication regulates human 520 liver bud development from pluripotency. Nature 546, 533-538. 521 Campbell, S. A., Stephan, T. L., Lotto, J., Cullum, R., Drissler, S. and Hoodless, P. A. (2021). 522 Signalling pathways and transcriptional regulators orchestrating liver development and cancer. 523 Development 148. 524 Chaturantabut, S., Shwartz, A., Evason, K. J., Cox, A. G., Labella, K., Schepers, A. G., Yang, S., 525 Acuna, M., Houvras, Y., Mancio-Silva, L. et al. (2019). Estrogen Activation of G-Protein-Coupled 526 Estrogen Receptor 1 Regulates Phosphoinositide 3-Kinase and mTOR Signaling to Promote Liver 527 Growth in Zebrafish and Proliferation of Human Hepatocytes. Gastroenterology 156, 1788-1804 e13. 528 Chaturantabut, S., Shwartz, A., Garnaas, M. K., LaBella, K., Li, C. C., Carroll, K. J., Cutting, C. 529 C., Budrow, N., Palaria, A., Gorelick, D. A. et al. (2020). Estrogen Acts Through Estrogen Receptor 530 2b to Regulate Hepatobiliary Fate During Vertebrate Development. Hepatology 72, 1786-1799. 531 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 26 Chen, J., Ruan, H., Ng, S. M., Gao, C., Soo, H. M., Wu, W., Zhang, Z., Wen, Z., Lane, D. P. and 532 Peng, J. (2005). Loss of function of def selectively up-regulates Delta113p53 expression to arrest 533 expansion growth of digestive organs in zebrafish. Genes Dev 19, 2900-11. 534 Chen, J., Xuan, Z., Song, W., Han, W., Chen, H., Du, Y., Xie, H., Zhao, Y., Zheng, S. and Song, P. 535 (2021). EAG1 enhances hepatocellular carcinoma proliferation by modulating SKP2 and metastasis 536 through pseudopod formation. Oncogene 40, 163-176. 537 Cheng, W., Guo, L., Zhang, Z., Soo, H. M., Wen, C., Wu, W. and Peng, J. (2006). HNF factors 538 form a network to regulate liver-enriched genes in zebrafish. Dev Biol 294, 482-96. 539 Chu, J. and Sadler, K. C. (2009). New school in liver development: lessons from zebrafish. 540 Hepatology 50, 1656-63. 541 Cox, A. G., Hwang, K. L., Brown, K. K., Evason, K., Beltz, S., Tsomides, A., O'Connor, K., Galli, 542 G. G., Yimlamai, D., Chhangawala, S. et al. (2016). Yap reprograms glutamine metabolism to 543 increase nucleotide biosynthesis and enable liver growth. Nat Cell Biol 18, 886-896. 544 Cox, A. G., Saunders, D. C., Kelsey, P. B., Jr., Conway, A. A., Tesmenitsky, Y., Marchini, J. F., 545 Brown, K. K., Stamler, J. S., Colagiovanni, D. B., Rosenthal, G. J. et al. (2014). S-nitrosothiol 546 signaling regulates liver development and improves outcome following toxic liver injury. Cell Rep 6, 547 56-69. 548 Cox, A. G., Tsomides, A., Yimlamai, D., Hwang, K. L., Miesfeld, J., Galli, G. G., Fowl, B. H., Fort, 549 M., Ma, K. Y., Sullivan, M. R. et al. (2018). Yap regulates glucose utilization and sustains nucleotide 550 synthesis to enable organ growth. EMBO J 37. 551 Dunn, K., Vashisht, A. and Hammond-Weinberger, D. R. (2022). Comparative in situ hybridization 552 protocols in zebrafish. Biotechniques 73, 123-130. 553 Evageliou, N. F., Haber, M., Vu, A., Laetsch, T. W., Murray, J., Gamble, L. D., Cheng, N. C., Liu, 554 K., Reese, M., Corrigan, K. A. et al. (2016). Polyamine Antagonist Therapies Inhibit Neuroblastoma 555 Initiation and Progression. Clin Cancer Res 22, 4391-404. 556 Farooq, M., Sulochana, K. N., Pan, X., To, J., Sheng, D., Gong, Z. and Ge, R. (2008). Histone 557 deacetylase 3 (hdac3) is specifically required for liver development in zebrafish. Dev Biol 317, 336-53. 558 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 27 Field, H. A., Ober, E. A., Roeser, T. and Stainier, D. Y. (2003). Formation of the digestive system in 559 zebrafish. I. Liver morphogenesis. Dev Biol 253, 279-90. 560 Gao, Y., Jin, Q., Gao, C., Chen, Y., Sun, Z., Guo, G. and Peng, J. (2022). Unraveling Differential 561 Transcriptomes and Cell Types in Zebrafish Larvae Intestine and Liver. Cells 11. 562 Goessling, W. and Stainier, D. Y. (2016). Endoderm specification and liver development. Methods 563 Cell Biol 134, 463-83. 564 James, C., Zhao, T. Y., Rahim, A., Saxena, P., Muthalif, N. A., Uemura, T., Tsuneyoshi, N., Ong, 565 S., Igarashi, K., Lim, C. Y. et al. (2018). MINDY1 Is a Downstream Target of the Polyamines and 566 Promotes Embryonic Stem Cell Self-Renewal. Stem Cells 36, 1170-1178. 567 Jochheim-Richter, A., Rudrich, U., Koczan, D., Hillemann, T., Tewes, S., Petry, M., Kispert, A., 568 Sharma, A. D., Attaran, F., Manns, M. P. et al. (2006). Gene expression analysis identifies novel 569 genes participating in early murine liver development and adult liver regeneration. Differentiation 74, 570 167-73. 571 Khaliq, M., Choi, T. Y., So, J. and Shin, D. (2015). Id2a is required for hepatic outgrowth during 572 liver development in zebrafish. Mech Dev 138 Pt 3, 399-414. 573 Kimmel, C. B., Ballard, W. W., Kimmel, S. R., Ullmann, B. and Schilling, T. F. (1995). Stages of 574 embryonic development of the zebrafish. Dev Dyn 203, 253-310. 575 Kimura, Y., Koyama, Y., Taura, K., Kudoh, A., Echizen, K., Nakamura, D., Li, X., Nam, N. H., 576 Uemoto, Y., Nishio, T. et al. (2022). Characterization and role of collagen gene expressing hepatic 577 cells following partial hepatectomy in mice. Hepatology. 578 Kroll, F., Powell, G. T., Ghosh, M., Gestri, G., Antinucci, P., Hearn, T. J., Tunbak, H., Lim, S., 579 Dennis, H. W., Fernandez, J. M. et al. (2021). A simple and effective F0 knockout method for rapid 580 screening of behaviour and other complex phenotypes. Elife 10. 581 Leibing, T., Geraud, C., Augustin, I., Boutros, M., Augustin, H. G., Okun, J. G., Langhans, C. D., 582 Zierow, J., Wohlfeil, S. A., Olsavszky, V. et al. (2018). Angiocrine Wnt signaling controls liver 583 growth and metabolic maturation in mice. Hepatology 68, 707-722. 584 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 28 Li, C., Du, L., Ren, Y., Liu, X., Jiao, Q., Cui, D., Wen, M., Wang, C., Wei, G., Wang, Y. et al. 585 (2019). SKP2 promotes breast cancer tumorigenesis and radiation tolerance through PDCD4 586 ubiquitination. J Exp Clin Cancer Res 38, 76. 587 Li, X., Song, G., Zhao, Y., Ren, J., Li, Q. and Cui, Z. (2021). Functions of SMC2 in the 588 Development of Zebrafish Liver. Biomedicines 9. 589 Lim, H. K., Rahim, A. B., Leo, V. I., Das, S., Lim, T. C., Uemura, T., Igarashi, K., Common, J. 590 and Vardy, L. A. (2018). Polyamine Regulator AMD1 Promotes Cell Migration in Epidermal Wound 591 Healing. J Invest Dermatol 138, 2653-2665. 592 Liu, C., Wu, C., Yang, Q., Gao, J., Li, L., Yang, D. and Luo, L. (2016). Macrophages Mediate the 593 Repair of Brain Vascular Rupture through Direct Physical Adhesion and Mechanical Traction. 594 Immunity 44, 1162-76. 595 Liu, J., Zhu, C., Ning, G., Yang, L., Cao, Y., Huang, S. and Wang, Q. (2019). Chemokine signaling 596 links cell-cycle progression and cilia formation for left-right symmetry breaking. PLoS Biol 17, 597 e3000203. 598 Ma, Z., Zhu, P., Shi, H., Guo, L., Zhang, Q., Chen, Y., Chen, S., Zhang, Z., Peng, J. and Chen, J. 599 (2019). PTC-bearing mRNA elicits a genetic compensation response via Upf3a and COMPASS 600 components. Nature 568, 259-263. 601 Nguyen, A. T., Koh, V., Spitsbergen, J. M. and Gong, Z. (2016). Development of a conditional liver 602 tumor model by mifepristone-inducible Cre recombination to control oncogenic kras V12 expression in 603 transgenic zebrafish. Sci Rep 6, 19559. 604 Nishimura, K., Nakatsu, F., Kashiwagi, K., Ohno, H., Saito, T. and Igarashi, K. (2002). Essential 605 role of S-adenosylmethionine decarboxylase in mouse embryonic development. Genes Cells 7, 41-7. 606 Ober, E. A. and Lemaigre, F. P. (2018). Development of the liver: Insights into organ and tissue 607 morphogenesis. J Hepatol 68, 1049-1062. 608 Parmar, M. B. and Wright, J. M. (2013). Comparative evolutionary genomics of medaka and 609 three-spined stickleback fabp2a and fabp2b genes with fabp2 of zebrafish. Genome 56, 27-37. 610 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 29 Patel, M. M., Volkov, O. A., Leija, C., Lemoff, A. and Phillips, M. A. (2018). A dual regulatory 611 circuit consisting of S-adenosylmethionine decarboxylase protein and its reaction product controls 612 expression of the paralogous activator prozyme in Trypanosoma brucei. PLoS Pathog 14, e1007404. 613 Pegg, A. E. (2009). S-Adenosylmethionine decarboxylase. Essays Biochem 46, 25-45. 614 Perugorria, M. J., Olaizola, P., Labiano, I., Esparza-Baquer, A., Marzioni, M., Marin, J. J. G., 615 Bujanda, L. and Banales, J. M. (2019). Wnt-beta-catenin signalling in liver development, health and 616 disease. Nat Rev Gastroenterol Hepatol 16, 121-136. 617 Petkov, P. M., Zavadil, J., Goetz, D., Chu, T., Carver, R., Rogler, C. E., Bottinger, E. P., Shafritz, 618 D. A. and Dabeva, M. D. (2004). Gene expression pattern in hepatic stem/progenitor cells during rat 619 fetal development using complementary DNA microarrays. Hepatology 39, 617-27. 620 Sanders, J. A., Schorl, C., Patel, A., Sedivy, J. M. and Gruppuso, P. A. (2012). Postnatal liver 621 growth and regeneration are independent of c-myc in a mouse model of conditional hepatic c-myc 622 deletion. BMC Physiol 12, 1. 623 Saviano, A., Henderson, N. C. and Baumert, T. F. (2020). Single-cell genomics and spatial 624 transcriptomics: Discovery of novel cell states and cellular interactions in liver physiology and disease 625 biology. J Hepatol 73, 1219-1230. 626 Segal, J. M., Kent, D., Wesche, D. J., Ng, S. S., Serra, M., Oules, B., Kar, G., Emerton, G., 627 Blackford, S. J. I., Darmanis, S. et al. (2019). Single cell analysis of human foetal liver captures the 628 transcriptional profile of hepatobiliary hybrid progenitors. Nat Commun 10, 3350. 629 Shin, D., Shin, C. H., Tucker, J., Ober, E. A., Rentzsch, F., Poss, K. D., Hammerschmidt, M., 630 Mullins, M. C. and Stainier, D. Y. (2007). Bmp and Fgf signaling are essential for liver specification 631 in zebrafish. Development 134, 2041-50. 632 Snezhkina, A. V., Krasnov, G. S., Lipatova, A. V., Sadritdinova, A. F., Kardymon, O. L., 633 Fedorova, M. S., Melnikova, N. V., Stepanov, O. A., Zaretsky, A. R., Kaprin, A. D. et al. (2016). 634 The Dysregulation of Polyamine Metabolism in Colorectal Cancer Is Associated with Overexpression 635 of c-Myc and C/EBPbeta rather than Enterotoxigenic Bacteroides fragilis Infection. Oxid Med Cell 636 Longev 2016, 2353560. 637 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 30 Tachmatzidi, E. C., Galanopoulou, O. and Talianidis, I. (2021). Transcription Control of Liver 638 Development. Cells 10. 639 Wang, G., Chan, C. H., Gao, Y. and Lin, H. K. (2012). Novel roles of Skp2 E3 ligase in cellular 640 senescence, cancer progression, and metastasis. Chin J Cancer 31, 169-77. 641 Wang, S., Zhang, C., Hasson, D., Desai, A., SenBanerjee, S., Magnani, E., Ukomadu, C., 642 Lujambio, A., Bernstein, E. and Sadler, K. C. (2019). Epigenetic Compensation Promotes Liver 643 Regeneration. Dev Cell 50, 43-56 e6. 644 Wilkinson, R. N., Elworthy, S., Ingham, P. W. and van Eeden, F. J. (2013). A method for 645 high-throughput PCR-based genotyping of larval zebrafish tail biopsies. Biotechniques 55, 314-6. 646 Wu, B. K., Mei, S. C., Chen, E. H., Zheng, Y. and Pan, D. (2022). YAP induces an oncogenic 647 transcriptional program through TET1-mediated epigenetic remodeling in liver growth and 648 tumorigenesis. Nat Genet 54, 1202-1213. 649 Wu, R. S., Lam, II, Clay, H., Duong, D. N., Deo, R. C. and Coughlin, S. R. (2018). A Rapid Method 650 for Directed Gene Knockout for Screening in G0 Zebrafish. Dev Cell 46, 112-125 e4. 651 Wu, T., Gu, X. and Cui, H. (2021). Emerging Roles of SKP2 in Cancer Drug Resistance. Cells 10. 652 Yang, L., Wang, W. H., Qiu, W. L., Guo, Z., Bi, E. and Xu, C. R. (2017). A single-cell 653 transcriptomic analysis reveals precise pathways and regulatory mechanisms underlying hepatoblast 654 differentiation. Hepatology 66, 1387-1401. 655 Yang, Q., Yan, C., Wang, X. and Gong, Z. (2019). Leptin induces muscle wasting in a zebrafish 656 kras-driven hepatocellular carcinoma (HCC) model. Dis Model Mech 12. 657 Yi, J., Lu, L., Yanger, K., Wang, W., Sohn, B. H., Stanger, B. Z., Zhang, M., Martin, J. F., Ajani, 658 J. A., Chen, J. et al. (2016). Large tumor suppressor homologs 1 and 2 regulate mouse liver progenitor 659 cell proliferation and maturation through antagonism of the coactivators YAP and TAZ. Hepatology 64, 660 1757-1772. 661 Yuan, P., Li, J., Zhou, F., Huang, Q., Zhang, J., Guo, X., Lyu, Z., Zhang, H. and Xing, J. (2017). 662 NPAS2 promotes cell survival of hepatocellular carcinoma by transactivating CDC25A. Cell Death Dis 663 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 31 8, e2704. 664 Zabala-Letona, A., Arruabarrena-Aristorena, A., Martin-Martin, N., Fernandez-Ruiz, S., 665 Sutherland, J. D., Clasquin, M., Tomas-Cortazar, J., Jimenez, J., Torres, I., Quang, P. et al. 666 (2017). mTORC1-dependent AMD1 regulation sustains polyamine metabolism in prostate cancer. 667 Nature 547, 109-113. 668 Zhang, D., Zhao, T., Ang, H. S., Chong, P., Saiki, R., Igarashi, K., Yang, H. and Vardy, L. A. 669 (2012). AMD1 is essential for ESC self-renewal and is translationally down-regulated on 670 differentiation to neural precursor cells. Genes Dev 26, 461-73. 671 Zhang, H., Kobayashi, R., Galaktionov, K. and Beach, D. (1995). p19Skp1 and p45Skp2 are 672 essential elements of the cyclin A-CDK2 S phase kinase. Cell 82, 915-25. 673 Zhang, S., Chen, Q., Liu, Q., Li, Y., Sun, X., Hong, L., Ji, S., Liu, C., Geng, J., Zhang, W. et al. 674 (2017). Hippo Signaling Suppresses Cell Ploidy and Tumorigenesis through Skp2. Cancer Cell 31, 675 669-684 e7. 676 Zhang, Y., Xia, J., Liu, M., Chen, B., Yang, M., Yu, X., Ou, Y., Li, S., Liu, X., Feng, Y. et al. 677 (2022). Mcm5 Represses Endodermal Migration through Cxcr4a-itgb1b Cascade Instead of Cell Cycle 678 Control. Biomolecules 12. 679 Zhao, J., Adams, A., Roberts, B., O'Neil, M., Vittal, A., Schmitt, T., Kumer, S., Cox, J., Li, Z., 680 Weinman, S. A. et al. (2018). Protein arginine methyl transferase 1- and Jumonji C domain-containing 681 protein 6-dependent arginine methylation regulate hepatocyte nuclear factor 4 alpha expression and 682 hepatocyte proliferation in mice. Hepatology 67, 1109-1126. 683 Zhao, T., Goh, K. J., Ng, H. H. and Vardy, L. A. (2012). A role for polyamine regulators in ESC 684 self-renewal. Cell Cycle 11, 4517-23. 685 Zhao, X., Monson, C., Gao, C., Gouon-Evans, V., Matsumoto, N., Sadler, K. C. and Friedman, S. 686 L. (2010). Klf6/copeb is required for hepatic outgrowth in zebrafish and for hepatocyte specification in 687 mouse ES cells. Dev Biol 344, 79-93. 688 Zhu, C., Guo, Z., Zhang, Y., Liu, M., Chen, B., Cao, K., Wu, Y., Yang, M., Yin, W., Zhao, H. et al. 689 (2019). Aplnra/b Sequentially Regulate Organ Left-Right Patterning via Distinct Mechanisms. Int J 690 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 32 Biol Sci 15, 1225-1239. 691 692 693 694 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 33 Figure Legends 695 Figure 1. liver enriched gene screening and analysis 696 (A-B) Liver size comparing from 2dpf to 7dpf using in situ hybridization and living 697 Tg(fabp10:GFP) transgenic embryos. Lateral view. (C) Schedule for sorting different 698 staged-hepatocytes and bulk RNA sequencing (RNA-seq). (D) In the top 25 genes of 699 liver-encriched genes on 3dpf, 56%, and 12% of them was reported to be enriched in embryonic 700 liver in zfin database or being reported in early literatures. 32% of them is uncharacterized. (E) 4 701 genes were randomly selected from the genes being enriched and highly expressed in hepatocytes 702 on 3dpf. The expression level of them was higher in hepatocytes on 3dpf than that in adult 703 hepatocytes: the expression level in adult hepatocytes of egfra,chmp4c, cyp17a1and ptgis is 704 69.7% ,8.3%, 5.8% and 3.8% of that in hepatocytes on 3dpf embryos, respectively. (F) The 705 relative expression of gatm(0.63%), nos2b(9.7%), nos1(1.0%) and amd1(3.0%) in 706 3dpf-hepatocytes is much higher than that in adult-hepatocytes. Values are reported as mean ± 707 SEM. “***” P < 0.001, “****” P < 0.0001. Scale bars, 50μ m. 708 709 Figure 2. The expression pattern of amd1 in early embryonic development 710 (A, B) The expression of amd1 in early embryonic development. (A) PCR amplification for amd1 711 in embryos at 8-cell stage, shield stage, bud stage, 24hpf, 48hpf, 72hpf and 96hpf. (B) In situ 712 hybridization staining for amd1 sense probe (Bb1’-b7’) and antisense probe (Bb1-b8). Especially 713 from 2dpf to 4dpf, amd1 was expressed highly in endoderm cells (Bb5-b8, red arrow showed), 714 including in liver (Bb6-b8, yellow arrow and dashed box showed). (b10-b11) Double staining for 715 uox (fast red) and amd1(blue) at 4dpf. Uox and amd1 was colocalized in liver, amd1 was also 716 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 34 expressed in gut. (C, D) Comparing the expression of amd1 in hepatocytes on 3dpf and 4dpf using 717 RT-qPCR. Sorting the GFP lableled hepatocytes on 3dpf and 4dpf (C). The data also showed the 718 ratio of hepatocyte in embryos on 3dpf (27.2%) is largely lower than that on 4dpf (47.8%) （ C） . 719 The expression level of amd1 in hepatocytes on 4dpf is 8.9% of that in hepatocytes on 3dpf （ D） . 720 Values are reported as mean ± SEM. “****” P < 0.0001. Scale bars, 200μ m. 721 722 Figure 3. liver growth and hepatocyte proliferation were repressed in amd17-/- liver 723 (A, B) Comparing the liver size in amd17-/- embryos and control embryos using Tg(fabp10:GFP) 724 transgenic line. From 2dpf to 5dpf, the liver size is smaller in amd17-/- embryos (Aa1-a6). The size 725 of liver in amd17-/- embryos was 45.7% (n=16, p<0.0001 ), 52.2% (n=18, p<0.0001) and 49.3% 726 (n=14, p<0.0001) of that in control embryos on 3dpf, 4dpf and 5dpf, respectively. (C, D) 727 Comparing the liver size in amd17-/- embryos and control embryos using fabp10staining (Cc1-c6). 728 The size of liver in amd1 7-/- embryos was 62.4% (n=15, p<0.0001) , 61.0% (n=13, p<0.0001) and 729 46.3% (n=19, p<0.0001) of that in control embryos on 3dpf, 4dpf and 5dpf, respectively.(E, F) 730 Comparing the liver size using uox staining (E). The size of liver in amd1 7-/- embryos was 53.6% 731 (n=12, p<0.01) ,45.5% (n=15, p<0.0001) and 46.0% (n=14, p<0.0001) of that in control embryos 732 on 3dpf, 4dpf and 5dpf, respectively (F). (G, H) H 3P staining for amd17-/-embryos and control 733 embryos. On 4.5dpf, the hepatocytes stainned with H3P in amd17-/-embryos were decreased than 734 that incontrol embryos G). In control embryos, approximate 7.1 hepatocytes were stained with 735 H3P (n=10); in amd17-/-embryos, approximate 2.6 hepatocyteswere stained with H 3P (n=10, P< 736 0.0001) (B). Values are reported as mean ± SEM. NS, not significant, “****” P < 0.0001, Scale 737 bars, 200μ m. 738 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 35 739 Figure 4. skp2 was downregulated in amd17-/-embryos 740 (A-C) RNA sequencing data analyisis. In amd17-/- embryos, 94 genes were up-regulated, 104 genes 741 were down-regulated, there is no significant difference for the expresion of 28439 genes between 742 cotrols and amd17-/- embryos (A). In the down-regulated genes, some of them were belonged to 743 mTOR signaling (B). skp2 was significantly downregulated in amd17-/- embryos (C). (D) RT-qPCR 744 experiments showed that skp2 (0.545 folds to control, p=0.0011) and skp2 downstream gene rho 745 (0.626 folds to control, p=0.0036) was downregulated in amd17-/-embryos. (E) In situ experiments 746 showed that skp2 was downregulated in liver and gut in amd17-/-embryos on 4dpf. (F) The 747 expression of skp2 was examined at 4-cell stage (f1), 128-cell stage (f2), shield stage (f3), bud 748 stage (f4), 24hpf (f5) and 3dpf (f6). On 3dpf, skp2 was enriched in eyes, pancreas, liver and gut 749 (f6). Values are reported as mean ± SEM. “**” P < 0.01. Scale bars,100μ m. 750 751 Figure 5. skp2 is required for liver development and lies downstream of amd1 752 (A) Three targets of skp2 sgRNA in exon2 of skp2 gene. (B) The sequencing results of skp2 wild 753 type (up sequence) and skp2 +/- embryos (down sequence). (C) 80 nucleotides between 754 “AAAAAGCGG” (green line labelled) and “ACAAGAAGG” (yellow line labelled) were deleted, 755 but “AGTA” were added in these two sequences. This mutantion gave rise to a frame-shift and 756 there is no stop codon at the terminal region of skp2 CDs. (D) The expression of skp2 in skp2-/- 757 mutants was 55.2% of that in wild type controls on 4dpf. Values are reported as mean ± SEM. “**” 758 P < 0.01. (E) Liver size was compared in controls and skp2 mutants using Tg(fabp10:GFP) 759 transgenic line and wild type line embryos. The liver size in skp2 mutants (e2, e4) is smaller that 760 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 36 in controls (e1, e3). Statistical analysis was did for the liver size in skp2 mutants and controls (e5). 761 Values are reported as mean ± SEM. “****” P < 0.0001. (F) Cell proliferation was examined using 762 Edu staining on 4.5dpf. The number of hepatocytes staining with Edu in skp2 mutants is smaller 763 than that in controls. Values are reported as mean ± SEM. “**” P < 0.01. (G) skp2 overexpression 764 increased the size of liver. On 3.5 dpf, in wild type embryos injection of skp2 mRNA increased the 765 size of liver (86.6%, n=18); meanwhile the phenotype “smaller liver” in amd17-/- embryos was 766 rescued in 86.3% of embryos by injecting skp2 mRNA (n=11). “*” P < 0.05, “***” P < 0.001, 767 “****” P < 0.0001, Scale bars, 100μ m. 768 769 Figure 6. skp2 is required for liver growth and mediates amd1 regulates liver development 770 (A) The live phenotype after downregulatio of skp2 using mosaic knock out on 3.5dpf.Most of 771 embryos (88.1%, n=12, p< 0.0001) displayed smaller liver in embryos injected with Cas9/ skp2 772 sgRNAs.(B)Schedule for inhibiting skp2 activity using SMIP004 treatment. (C) After inhibiting 773 skp2 activity, 92.6% of embryos (Cc2, n=10, p< 0.0001) displayed smaller liver than that in 774 controls (Cc1, n=11); in amd17-/- embryos, skp2 inhibition (Cc4, 100%, n=13, p< 0.0001) made 775 the liver much smaller that controls (Cc3, n=12). (D) In situ experiment also showed that, 100% of 776 embryos (Dd2, n=11) displayed smaller liver than that in controls (Dd1, n=11). Inhibiting skp2 777 activity made most of amd17-/- embryos (Dd4, 91.6%, n=11, p< 0.0001) displayed much smaller 778 liver that controls (Dd3, n=10). (E) Cell proliferation evaluation using H 3P staining. After 779 treatment with SMIP004, H 3P staining hepatocytes were decreased comparing with control 780 (Control, n=5; SMIP treatment, n=6, p=0.0004). SMIP004 treatment further decre ased the number 781 of hepatocytes staining with H3p (n=6, p=0.171). (F) skp2 overexpression increased the size of 782 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint 37 liver. On 3.5 dpf, in wild type embryos injection of skp2 mRNA increased the size of liver (86.6%, 783 n=18); meanwhile the phenotype smaller liver in amd17-/- embryos was rescued in 86.3% of 784 embryos by injecting skp2 mRNA (n=11, p=< 0.0001). Values are reported as mean ± SEM. “*” P 785 < 0.05, “***” P < 0.001, “****” P < 0.0001, Scale bars,100μ m. 786 787 Figure 7. amd1-skp2 cascade is reqiured for hepatocyte proliferation in zebrafish HCC 788 model 789 (A) The schedule for examing the expression of amd1 and skp2 in hepatocytes for normal liver 790 and HCC model. (B) Comparing with that in normal liver, the expression of amd1 (5.39 folds to 791 normal liver, p=0.0162) and skp2 (7.93 folds to normal liver, p=0.0035) was increased 792 significantly. (C, D) dox induced overexpression of Kras increased the liver growth (Cc2, n=18; D, 793 n=12 , p=< 0.0001), amd1 loss of function decreased the size of liver in HCC model (Cc3, n=15; 794 D, n=9, p=< 0.0001), simultaniously inhibiting skp2 and amd1 made the liver much smaller in 795 HCC model (Cc4, n=16; D, n=11 , p=0.0012). (E, F) proliferating hepatocytes staining with H3p. 796 Comparing with control, amd1 loss of function decreased the number of H3p staining hepatocytes 797 (Ee5; F, n=6, p=< 0.0001), simultaniously inhibiting skp2 and amd1 furtherdecreased the number 798 of H3p staining hepatocytes (Ee8; F, n=6, p=0.0005). Values are reported as mean ± SEM. “*” P < 799 0.05, “**” P < 0.01, “***” P < 0.001, “****” P < 0.0001, Scale bars,100μ m. 800 801 802 was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (whichthis version posted May 26, 2024. ; https://doi.org/10.1101/2024.05.25.595867doi: bioRxiv preprint