Box H/ACA snoRNP regulates lipid storage through insulin signaling pathway in Drosophila melanogaster

DGAT, Drosophila melanogaster, box H/ACA snoRNP , insulin 18 pathway, lipid metabolism 19 20 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 2 Text abstracts 21 Lipid homeostasis is essential for organismal physiology, and its 22 disruption contributes to metabolic disorders. Using an unbiased genetic 23 modifier screen in Drosophila, we identified GAR1, a core component of the 24 box H/ACA small nucleolar ribonucleoprotein complex, as a pivotal regulator 25 of systemic lipid storage. We show that the H/ACA snoRNP complex is 26 essential for maintaining lipid droplet morphology in adipose tissue and 27 preventing ectopic fat accumulation. Moreover, null mutants of Gar1 or Dkc1 28 exhibit severe developmental defects, including reduced body size and larval 29 lethality. RNA-seq analysis revealed that Gar1 dysfunction triggered 30 widespread alternative splicing defects, specifically targeting key transcripts 31 within the insulin signaling cascade, including chico, Pi3K92E, sgg, and Lip4. 32 Furthermore, knockdown of Gar1 impaired insulin signaling, as evidenced by 33 the reduced membrane localization of the tGPH fluorescence. Genetic 34 epistasis further positions GAR1 upstream of the lin-28/foxo axis, as knocking 35 down lin-28 or foxo fully rescues the lipometabolic defects in GAR1-deficient 36 animals. These findings reveal a previously unrecognized link between the 37 snoRNP machinery and metabolic process, establishing the box H/ACA 38 complex as a n important coordinator that integrates RNA processing with 39 insulin-mediated nutrient sensing to ensure developmental and lipid 40 homeostasis. 41 42 .CC-BY 4.0 International licensemade available under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 3 Article summary 43 Lipid metabolism is tightly controlled by multiple factors. To find new 44 regulators, the authors performed a genetic screen and identified a small 45 nucleolar protein GAR1 participate in fat storage and larval development. 46 They demonstrated a critical role of box H/ACA snoRNP complex in 47 modulating alternative splicing and balancing insulin cascade. Blocking two 48 insulin-related genes reversed the lipid defects caused by Gar1 loss. These 49 findings revealed the box H/ACA complex integrates RNA processing with 50 insulin-mediated nutrient sensing to ensure developmental and lipid 51 homeostasis, offering a perspective for understanding the metabolic 52 regulation network. 53 54 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 4

55 Lipids serve as the primary energy source in living organisms, and their 56 metabolism significantly influences growth, reproduction, vitality, and disease 57 progression (Kamareddine et al., 2018; Lim et al., 2014; Logan-Garbisch et 58 al., 2014). Lipid droplets (LDs) are organelles specialized in storing neutral 59 lipids, including triglycerides, sterol esters, and esterified ceramides (Chen et 60 al., 2020; Xu et al., 2018) . Pathologically, LDs accumulate in non -adipose 61 tissues such as the liver, heart, kidney, and muscle, a condition termed 62 ectopic fat accumulation (EFA) (Yan et al., 2017). EFA can initiate 63 inflammatory responses that contribute to insulin resistance (Trouwborst et 64 al., 2018), thereby elevating the risk of metabolic syndromes , including 65 cardiovascular diseases, fatty liver disease, and hyperglycemia (Camporez et 66 al., 2015; Ferrara et al., 2019). Moreover, excessive free fatty acids in non -67 adipose cells stimulate the production of reactive oxygen species (ROS) while 68 impairing mitochondrial respiration. This process also induces endoplasmic 69 reticulum (ER) stress responses and increases apoptosis (Ljubkovic et al., 70 2019; Ly et al., 2017). 71 Previous studies have highlighted the crucial roles of insulin and mTOR 72 pathways in maintaining metabolic homeostasis (Caron et al., 2015; Krycer et 73 al., 2020). In humans and other mammals, insulin is produced by pancreatic 74 beta cells, which modulate blood glucose levels, promote cellular proliferation, 75 and stimulate lipid storage (Lin and Smagghe, 2019). In Drosophila, eight 76 insulin-like peptides (ILP1-ILP8) are secreted during nutrient deprivation and 77 bind to the insulin receptor on target cells (Nassel et al., 2015), leading to the 78 phosphorylation of Chico (the IRS homolog) . An active Chico recruit s and 79 phosphorylates PI3K. Phosphorylated PI3K converts PIP2 to PIP3, facilitating 80 AKT recruitment to the plasma membrane where its activity is modulated by 81 PTEN (Zhang and Zhang, 2019). Activated AKT phosphorylates the 82 transcription factor FOXO, preventing its nuclear translocation , and thereby 83 inhibiting the expression of target genes such as Lip4 and 4EBP. The RNA-84 binding protein LIN-28, which is evolutionarily conserved from C. elegans to 85 mammals, regulates the early developmental cell lineage formation (Ambros 86 and Horvitz, 1984). LIN-28A enhances glucose uptake by augmenting insulin-87 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 5 PI3K-mTOR signaling, primarily by derepressing key let-7 targets in the 88 pathway, including IGF1R, INSR, IRS2, PIK3IP1, AKT2, TSC1, and RICTOR 89 (Zhu et al., 2011). In Drosophila, lin-28 deletion mutants exhibit reduced cell 90 size and accelerated larval -pupal transition, whereas its overexpression 91 delays pupation and causes lethality (Gonzalez-Itier et al., 2018). 92 Emerging evidence underscores the importance of non-coding RNAs and 93 their binding proteins in cellular metabolic homeostasis. Small nucleolar 94 RNAs (snoRNAs) range from 60 to 300 nucleotides in length and reside 95 primarily in the nucleolus. They are categorized into three groups based on 96 structure and protein associations: box C/D snoRNAs (snoRD), box H/ACA 97 snoRNAs (snoRA) and orphan snoRNAs (Ojha et al., 2020; Światowy and 98 Jagodzińśki, 2018). SnoRNAs regulates various post-transcriptional 99 processes, including rRNA acetylation, tRNA methylation, rRNA pseudo-100 uridylation, and mRNA alternative splicing (Wang et al., 2025) . Alternative 101 splicing (AS) generates distinct mRNA isoforms in a tissue-specific manner, 102 driving developmental programs, specifying tissue differentiation, and 103 sustaining metabolic homeostasis (Einson et al., 2023; Finegan et al., 2025; 104 Kaminska, 2025; Kanno et al., 2023 ; Zhang et al., 2025 ). For instance, the 105 box C/D snoRNA SNORD88B has been shown to recruit the splicing factors 106 SRSF1 and U2AF1 to regulate the alternative splicing of pre -G3BP1 (Lu et 107 al., 2025). 108 Similar to box C/D snoRNA s, box H/ACA snoRNAs function as small 109 nucleolar ribonucleoprotein (snoRNP) complexes. These complexes 110 comprise core protein components such as DKC1 (dyskerin), NOP10, GAR1, 111 NHP2, and NAF1 (Fatica et al., 2002; McMahon et al., 2015; Watkins et al., 112 1998). NAF1 is present in nascent snoRNPs, whereas GAR1 is exclusively 113 present in mature complexes (Massenet et al., 2017). Mutations in Dkc1, 114 Nop10 and Nhp2 are associated with human dyskeratosis congenita (Heiss 115 et al., 1998; Vulliamy et al., 2008; Walne et al., 2007) . Dkc1 depletion 116 induces cytoskeletal remodeling in human tumor cells (Di Maio et al., 2017) 117 and promotes intestinal stem cell regeneration in Drosophila (Vicidomini et 118 al., 2017). Meanwhile, downregulation of Naf1, Nop10, Dkc1, and Gar1 119 disrupts cyst formation in Drosophila ovaries (Breznak et al., 2023; Morita et 120 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 6 al., 2018). Despite documented roles of snoRNA-binding proteins in diverse 121 biological processes, their specific functions in lipid homeostasis remain 122 largely unknown. 123 Drosophila melanogaster is an excellent model for studying obesity, disease, 124 aging, and development (Heier and Kuhnlein, 2018; Liu and Huang, 2013; 125 Musselman and Kuhnlein, 2018). The previous work demonstrated ectopic 126 lipid accumulation in dSeipin null mutants and flies overexpressing DGAT 127 (midway), with double mutants showing a pronounced synergistic effect . To 128 further explore the regulatory network of lipid metabolism, we conducted a 129 genetic modifier screen in Drosophila to identify the genes involved in lipid 130 droplet deposition and morphology. In this study, we identif ied GAR1, a 131 snoRNA-binding protein , as a critical regulator of lipid homeostasis. 132 Subsequent cellular biolog ical analyses , RNA -seq, and genetic interaction 133 assays revealed a functional link between box H/ACA snoRNA -binding 134 proteins and the insulin signaling pathway in the regulation of lipid metabolism 135 and development. 136

137 Fly stocks and husbandry 138 All the fly stocks were reared on standard corn meal food. For the DGAT 139 modifier screen, the EP strains were provided by Dr. Jiahuai Han, and the 140 RNAi strains were from the NIG stock center. For the genetic interaction with 141 genes of insulin pathway, all the RNAi strains were ordered from Tsinghua fly 142 center and BDSC. The other strains used in this study were listed as following: 143 ppl-Gal4 (Dr. Pierre Leopold ), UAS-DGAT (BDSC, 20167) , Gar1 (BDSC, 144 34013; BDSC, 21775), Dkc1 (VDRC, 109616; VDRC, 34597; BDSC, 36595), 145 Nhp2 (BDSC, 51784), Nop10 (BDSC, 55194), tGPH reporter (BDSC, 8164), 146 lin-28 (Tsinghua, TH01982.N), foxo (BDSC, 32993) and da-Gal4 (Tsinghua). 147 Genetic screen 148 The ppl-Gal4 driver was crossed with UAS-DGAT and designated as 149 ppl>DGAT in this study. The ppl>DGAT females were then mated with males 150 carrying either EP or RNAi strains. Flies were reared at 29° C to maximize 151 GAL4 activity. Wandering third -instar larvae harboring both the ppl>DGAT 152 and the EP or RNAi transgene were dissected, mounted with PBS and 153 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 7 visualized using differential interference contrast (DIC) microscopy on a Zeiss 154 microscope. 155 The list of genes obtained from the genetic screen was submitted to the 156 Metascape online analysis platform (https://www.metascape.org), with the 157 species set to Drosophila melanogaster . Gene Ontology (GO) enrichment 158 analysis was performed using the express analysis mode with default 159 parameters (P-value < 0.01, minimum count ≥ 3, and enrichment factor ≥ 160 1.5). The top 20 significantly enriched GO biological process terms were 161 selected based on enrichment significance, and the enrichment heatmap was 162 generated using the built-in function of Metascape (Zhou et al., 2019). 163 Inverse PCR and sequencing 164 Thirty anesthetized flies were collected into a 1.5 mL tube and froze at -165 80° C. Genomic DNA was extracted and digested by RsaI at 37° C. After 166 digestion, the enzyme was inactivated at 65°C for 20 min, and the digested 167 DNA was self-ligated using T4 DNA ligase to generate circular products. The 168 ligated samples were amplified by PCR with primers LA(f).1 169 (GGGAATTGGGAATTCGTTAA) and LA(r).1 (TAGCGACGTGTTCACTTT 170 GC). The purified PCR products were sequenced with primer LA(f)seq1 171 (CTCTCAACAAGCAAACGTGC). The resulting sequences were blasted 172 against the FlyBase database to identify the affected genes. For additional 173 details, refer to the Drosophila Gene Disruption Project (Bellen et al., 2011). 174 Fluorescent fusion expression vector construction 175 The coding region of Gar1 (GenBank accession number: PX393084) was 176 amplified using primers GAATTCATGGGATTTGGTAAACCTCG and 177 GGTACCCTACCACCGACCCCGACC, and subsequently cloned into the 178 pEasy-T1 vector (TransGen Biotech). The resulting T1 -Gar1 plasmid was 179 then double-digested with EcoRI-KpnI and ligated into the pUAST-attB vector. 180 To construct UAS-Gar1-eGFP, the Gar1 fragment was amplified with 4038 -181 5BglII (AGATCTATGGGATTTGGTAAACCTCG) and 4038 -3nsBamHI 182 (GGATCCCCACCGACCCCGACCACC), digestion with BglII -BamHI, and 183 inserted into the eGFP-N3 vector. The Gar1-eGFP fusion fragment was finally 184 excised via BglII and XbaI digestion and ligated into the pUAST-attB vector. 185 Similarly, the coding regions of other box snoRNP components were 186 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 8 amplified as follows: Dkc1-RA using primers GAATTCTCTCTCCGTCTATT-187 AGTTGATT and AGATCTTATTCCTGAGCTTCGTCTC; Nhp2 using primers 188 GAATTCATGGGCAAAGTGAAAGTAG and GGTACCGACGGGTATGTTT-189 AGTGC; and Nop10 using primers CCTCGAGATGTATCTGATGTACACAA 190 and GGTACCGTAAATGGGCTCCGGCT. 191 Lipid droplet staining 192 To analyze lipid droplet (LD) morphology, wandering third-instar larval fat 193 body and salivary gland were dissected, stained with BODIPY 493/503 194 (Invitrogen; 1 mg/mL, 1:500) or Nile Red (Sigma; 0.5 mg/mL, 1:500) together 195 with DAPI (Sigma; 1 mg/mL, 1:10,000), and imaged using a Leica TCS SP8 196 confocal microscope. The sizes of lipid droplets were quantified using Image-197 Pro Plus 6.0 within a 150 µ m * 150 µ m rectangle from three images per 198 genotype. The 30 largest lipid droplets (LDs) were used to generate violin 199 plots. Statistical significance was determined based on mean values using 200 GraphPad Prism 10. 201 tGPH analysis 202 For the tGPH assay, embryos from da-Gal4 and RNAi mutants were 203 collected within a 4 -hour period and reared at 29° C. Fat body and salivary 204 gland from 68-72 hours larvae were dissected and examined under a Zeiss 205 Axio Imager M2 microscope. In fat body cells, the fluorescent intensity was 206 measured within 100 μm - 100 μm rectangle s from 3-6 figures for each 207 genotype. The cellular segmentation was performed using the online version 208 of Cellpose 4.0. Following this, membrane and cytosolic signals were 209 measured separately using ImageJ (Fiji). The resulting data were plotted and 210 analyzed with GraphPad Prism 10. 211 For gray value measurement in the salivary gland, the intensity was 212 measured within 50 μm - 50 μm rectangles from five figures for each genotype. 213 A line was drawn between the nuclei of two adjacent cells to quantify the 214 fluorescence intensity distribution along this line using ImageJ (Fiji). For the 215 distribution curve, three datasets from different cells per genotype were 216 selected for normalization of both the X -axis (distance) and Y -axis (gray 217 value). Subsequently, the normalized fluorescence distribution profiles were 218 subjected to curve fitting, generating a lowess fit curve for each genotype. 219 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 9 Knockout strategy and larval morphological analysis 220 Two sgRNAs were designed following an established online protocol 221 (Housden et al., 2015; https://www.flyrnai.org/crispr3/web/) and inserted into 222 the pUAST -attB vector using Golden -Gate cloning. The resulting sgRNA-223 expressing flies were subsequently crossed with Act-Cas9 flies. The deletion 224 mutants were identified by PCR amplification followed by sequencing. 225 Heterozygous alleles were balanced using the CyO ·P(ActGFP) JMR1 226 chromosome. 227 Embryos were collected within a 4 -hour interval. After 24 hours, 228 homozygous mutant larvae (lacking GFP fluorescence) were sorted under an 229 Olympus SZX16 stereomicroscope. Morphological observation was carried 230 out using Olympus BX53F or Motic K -400L microscopes. Body length 231 measurements were obtained with Motic Images Advanced 3.2 software, and 232 the resulting data were summarized in column graphs and subjected to 233 statistical analysis using GraphPad Prism 10. 234 Transcriptome analysis 235 Embryos from the w1118 and Gar12 mutants were collected within 4 hours. 236 After 24 hours, wild -type or non-GFP mutant larvae (homozygotes) were 237 transferred to fresh dishes, and rinsed with PBS in order to eliminate 238 contaminants. After drying on filter paper, larvae were immersed in 1mL of 239 Trizol solution (B511311, Sangon, China) for RNA stabilization and extraction. 240 RNA sequencing was performed by Sangon Biotech Co., Ltd. (Shanghai, 241 China). RNA quality was assessed using a NanoPhotometer® 242 spectrophotometer (IMPLEN, CA, USA) and a Qubit® 2.0 Fluorometer 243 (Invitrogen). Sequencing libraries were constructed using the VAHTSTM 244 mRNA-seq V2 Library Prep Kit for Illumina® with 1 μg of total RNA as input. 245 After sequencing, raw data quality was evaluated using FastQC (version 246 0.11.2), and reads were trimmed using Trimmomatic (version 0.36). Clean 247 reads were mapped to the reference genome using HISAT2 (version 2.0) with 248 default parameters. The differential gene expression analysis was performed 249 using DESeq2 (version 1.12.4), with genes satisfying q-value < 0.05 and 250 |FoldChange| ≥ 2 considered significantly differentially expressed. Alternative 251 splicing events were analyzed using rMATS (Shen et al., 2014), and 252 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 10 significant events were filtered with FDR 253 0.1. 254 Q-PCR analysis 255 The late L3 larvae were collected, washed with PBS in order to remove 256 any impurities, dried on filter paper carefully, and subsequently mounted in 257 1mL of Trizol solution (Invitrogen). The RNAs were extracted by Trizol -258 chloroform methods. The first strand cDNAs were obtain using oligo dT primer 259 and M-MLV Reverse Transcriptase (Invitrogen). The specific primers for each 260 gene were designed from online database in NCBI. The real-time quantitative 261 polymerase chain reactions used SYBR green I (ABI). The column chart and 262 statistical analysis were generated using GraphPad Prism 10. 263

264 A modifier screen for genes involved in lipid storage 265 To identify the genes regulating lipid metabolism, we conducted a genetic 266 modifier screen in Drosophila melanogaster by co-expressing DGAT, which 267 encodes a diacylglycerol acyltransferase (Chen et al., 2022) . Compared to 268 the wild type, overexpression of DGAT led to numerous small puncta in the 269 salivary glands (black arrows, Fig. 1 a and 1a' ) and large droplets in the fat 270 bodies of third-instar larvae (Fig. 1g). To investigate the genetic regulators of 271 lipid metabolism, we crossed DGAT-overexpressing flies (ppl>DGAT) with EP 272 or RNAi strains. Genes in RNAi strains were obtained from an online 273 database, whereas those in EP strains were identified using inverse PCR (Fig. 274 S1a and S1b; Table 1 ). We then dissected the fat body and salivary gland 275 tissues from third-instar larval offspring and analyzed LD ’ morphology using 276 DIC microscopy. From a total of 2100 EP and 1800 RNAi strains screened, 277 we identified 109 genes implicated in the regulation of lipid metabolism (Table 278 1). GO enrichment analysis clustered these genes into diverse biological 279 processes, primarily glycerophospholipid metabolism , triglyceride 280 biosynthesis, neural development , chromatin organization , homeosta tic 281 process, and reproduction (Fig. S1c). Based on the morphology changes in 282 the lipid droplets, we categorized the observed phenotypes into seven distinct 283 classes (Table 1). Classes I-V exhibited prominent differences in the salivary 284 glands, whereas classes VI and VII showed alterations primarily in the fat 285 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 11 body. 286 The c lass I double mutants displayed pronounced accumulation of 287 supersized lipid droplets in the salivary gland, resembling those typically 288 found in the fat body (Fig. 1 b and 1b'). This class includes several well -289 characterized lipid metabolism related genes such as Lipin, wun2, CdsA, 290 Desat1, and GluRIIC (Table 1). In contrast, class II mutants showed reduced 291 lipid accumulation in the salivary gland, with few or no cytoplasmic puncta 292 (Fig. 1c and 1c' ). Genes in this class include Rab5, Sema2b, and Nup160 293 (Table 1). Notably, class III double mutants displayed exhibited substantial 294 accumulation of numerous small lipid droplets in salivary gland, along with a 295 moderate increase in lipid droplet size (Fig. 1 d and 1d'). This category 296 included genes encoding fatty acid elongase ( Baldspot) and calcium ion -297 binding protein (rho-4) (Table 1). Interestingly, class IV was characterized by 298 the presence of multiple small cytoplasmic lipid droplets clustered around and 299 enclosing the nucleus, in contrast with the dispersed pattern observed in the 300 control group ( ppl>DGAT). This phenotype was often accompanied by 301 reduced salivary gland cell size (Fig. 1e and 1e'). Associated genes included 302 SMC3, Iswi, and Sbf (Table 1). Moreover, the class V mutants featured large 303 cytoplasmic lipid droplets surrounded by membrane -like circular structures 304 (blue arrows in Fig. 1 f and 1f'). Notably, this group included the lipogenic 305 genes, AGPAT and SREBP (Table 1). 306 In contrast to the diverse salivary gland phenotypes, changes in the fat 307 body were characterized by alterations in lipid droplet size, with droplets 308 showing either enlargement or reduction. Class VI mutants exhibited enlarged 309 lipid droplets in the fat body and contained genes encoding Lama, Kis, and 310 Lola (Fig. 1 g; Table 1). Class VII mutants, on the other hand, displayed 311 reduced lipid droplet size in the fat body and involved genes such as SERCA, 312 rab39, and FASN1 (Fig. 1 g; Table 1). In summary, this screen uncovered 313 critical genes involved in ectopic fat accumulation, prompting further 314 investigation into their roles in the regulation of lipid metabolism. 315 Subcellular Localization of box H/ACA snoRNP 316 From th is screen, we found that overexpression of Gar1 and DGAT 317 simultaneously led to an increase in EFA in the salivary gland (Fig. S2a; Table 318 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 12 1). GAR1 is a snoRNA binding protein that associates with other proteins and 319 box H/ACA snoRNAs to form the snoRNP complex, thereby regulating 320 multiple cellular processes (Breznak et al., 2023). However, the functional role 321 of the snoRNP complex in lipid metabolism remains unclear. Fluorescence 322 imaging was performed to systematically examine the subcellular localization 323 patterns of these proteins. We amplified and sequenced the coding 324 sequences of Gar1 and found that the cloned sequence contained a 6 -325 nucleotide deletion relative to the FlyBase reference, which resulted in a 326 polypeptide lacking two glycine residues (Fig. S2b). To determine whether 327 cloned GAR1 was functional, we expressed UAS -GAR1-eGFP using the fat 328 body-specific driver ppl-Gal4. As expected, GAR1 -eGFP was exclusively 329 localized to the nucleus and exhibited a punctate distribution in both the larval 330 salivary glands and fat body cells (Fig. 2a). This localization pattern was 331 confirmed by co -staining with phalloidin and DAPI (for nuclei), and is 332 consistent with previous reports in mammalian systems (Chamousset et al., 333 2010; Jacob et al., 2013). Furthermore, da-Gal4-driven overexpression of 334 GAR1 in the Gar12 mutant background fully rescued the mutant phenotype to 335 that of the wild type, demonstrating that GAR1 was successfully cloned (Fig. 336 S2c, S2d, 4k). Based on these findings, we extended analysis to other core 337 H/ACA snoRNPs using similar fluorescent tagging strategies. The sequences 338 of Dkc1-RA, Nop10, and Nhp2 were 100% identical to the reference 339 sequences. Interestingly, DKC1, NHP2, and NOP10 exhibited similar 340 localization, with predominant nuclear localization and faint cytoplasmic 341 signals in fat body cells (Fig. 2 b and 2c). These patterns resembled the 342 localization of CBF5-eYFP (DKC1 -eYFP) in Arabidopsis thaliana 343 (Lermontova et al., 2007). The above findings demonstrate that the four core 344 proteins exhibit conserved localization patterns. Coupled with the initial 345 observation that Gar1 overexpression promotes ectopic lipid accumulation, it 346 suggests that the entire complex plays a critical role in regulating specific 347 physiological processes, such as lipid homeostasis. Nevertheless, the 348 integrated function of the snoRNP s in lipid metabolism remains to be 349 determined, prompting us to investigate their role in Drosophila. 350 Box H/ACA snoRNP regulates lipid homeostasis 351 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 13 To investigate the ir function in vivo , we analyzed changes in lipid 352 deposition in the corresponding knockdown and overexpression flies. Salivary 353 glands were dissected at the wandering third -instar larval stage and stained 354 with BODIPY 493/503. We found that knockdown of Gar1 or Dkc1 reduced 355 the salivary gland cell size and induced numerous small ectopic lipid droplets 356 in the salivary glands (Fig. 3 a, S2a and S3a). In contrast, double mutants 357 combining Gar1 or Dkc1 RNAi with DGAT overexpression showed 358 dramatically reduced salivary gland cell size, along with a mild increase in 359 ectopic lipid accumulation (Fig. 3 a). These findings suggest that both Gar1 360 and Dkc1 are crucial for suppressing EFA formation in the salivary glands. To 361 further explore their role in EFA formation, we overexpressed Gar1 and found 362 that it alone did not produce a significant phenotype compared with the wild 363 type (Fig. 3b). However, co-overexpression of Gar1 and DGAT triggered the 364 formation of large ectopic lipid droplets (Fig. 3 b and S2 a), which was 365 consistent with the results of previous genetic screen. This indicates 366 that while Gar1 overexpression alone does not induce EFA, it strongly 367 enhances EFA formation when DGAT activity is elevated. Collectively, these 368

demonstrate that box H/ACA snoRNP components act as key 369 regulators of ectopic lipid accumulation in Drosophila melanogaster. 370 Next, we investigated whether the dysfunction of box H/ACA snoRNP 371 components affected lipid storage in the fat body. Lipid droplets in larval L3 372 fat bodies were stained with BODIPY 493/503 or Nile red and visualized using 373 confocal microscopy. For each image, a 150 μm * 150 μm area was selected, 374 and the diameters of the 30 largest lipid droplets (Max30 LD s) within each 375 region were measured and compared. The average Max30 LD s’ size in 376 Canton-S flies was 14. 37 μm, whereas it was reduced to 10. 07 μm in 377 ppl>Gar1HMS00979 flies and 10.16 μm in ppl>Gar1GD11196 flies, respectively (Fig. 378 3c and 3d), indicating that Gar1 knockdown decreases lipid droplet size . To 379 determine whether Dkc1 downregulation similarly affects lipid storage, we 380 analyzed LD morphology in the fat bodies of flies expressing ppl-Gal4 driven 381 Dkc1 RNAi (Dkc1kk101240 and Dkc1GL00555). The average Max30 LDs’ sizes 382 were 9.75 μm and 11.13 μm, respectively (Fig. 3 e and 3f), confirming that 383 DKC1 also promotes lipid storage in the adipose tissue. We evaluated the 384 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 14 impact of Nhp2 and Nop10 knockdown on lipid metabolism. As expected, ppl-385 Gal4-driven expression of Nhp2HMC03339 and Nop10HMC03890 exhibited smaller 386 fat body lipid droplets, with average Max30 LD’s sizes of 10.01 μm and 9.66 387 μm, respectively (Fig. 3g and 3h). Collectively, these results demonstrate that 388 box H/ACA snoRNP dysfunction reduces lipid droplet size in the fat body and 389 promotes ectopic lipid droplet accumulation in the salivary gland. The 390 consistent phenotype across multiple components suggests that these four 391 proteins function as integrated complexes to regulate lipid storage, 392 underscoring the critical role of box H/ACA snoRNP in maintaining lipid 393 homeostasis. 394 Dysfunction of box H/ACA snoRNP component resulted in 395 developmental defects 396 In addition to lipid metabolism, we explored the functions of box H/ACA 397 snoRNPs during development. Strikingly, functional impairment of either Gar1 398 or Dkc1 results in severe developmental defects. Wild-type flies typically 399 complete metamorphosis and eclose as adults 10 days after egg laying (AEL) 400 at room temperature, whereas knockdown of Gar1 (Gar1HMS00979 and 401 Gar1GD11196) driven by ppl-Gal4 at 29° C led to pupal arrest and failed eclosion 402 (Fig. 4a), indicating that GAR1 is essential for development. We examined 403 whether Dkc1 affects the developmental process. As expected, ubiquitous 404 knockdown of Dkc1 (Dkc1kk101240and Dkc1GL00555) using da-Gal4 at 29° C 405 resulted in developmental arrest at 4 days AEL and 8 days AEL, respectively 406 (Fig. 4b and 4e). After hatching, the larvae progress through three distinct 407 instar stages (L1, L2, and L3), each characterized by specific morphological 408 features of the mouth hooks and tracheal system, which serve as key 409 developmental markers (Hamid and Mishra, 2021). The L1 larvae exhibited 410 mouth hooks with a single tooth and lacked paired anterior spiracles (Fig. S4a 411 and S4a'). The L2 larvae developed mouth hooks bearing 2 -4 teeth and 412 possessed rounded or club -shaped anterior spiracles (Fig. S4 b and S4b'). 413 The L3 larvae displayed mouth hooks with 9 -12 teeth and branched anterior 414 spiracles (Fig. S4c and S4c'). On day 4 AEL, da>Dkc1kk101240 larvae exhibited 415 a slimmer body size. However, their overall body length was comparable to 416 that of wild -type larvae (Fig. 4 b). They showed obvious developmental 417 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 15 abnormalities: mouth hooks possessed only three teeth (Fig. 4c), and tracheal 418 termini were club-shaped (Fig. 4d), indicating that the mutants were arrested 419 at the L2 stage. By 6 d AEL, wild -type larvae initiated pupariation, whereas 420 da>Dkc1GL00555 flies exhibited typical larval morphology. The mutant larvae 421 had one tooth in the mouth hooks (characteristic of L1) and club -shaped 422 tracheae (characteristic of L2) (Fig. 4e), suggesting developmental arrest at 423 the L1-L2 transition. The above results indicate that knockdown of box H/ACA 424 components leads to developmental defects. 425 To further elucidate the loss of function phenotypes, we generated 426 deletion mutants of Gar1 and Dkc1 using CRISPR-Cas9. For Gar1, the two 427 sgRNAs targeting the first and second exons yielded three deletion mutants 428 (Gar11, Gar12, Gar13), with deleted regions of 839 bp, 455 bp, and 814 bp, 429 respectively (Fig. 4 f and 4g). Since Gar1 is located within an intron of 430 CG34396, we assessed whether its deletion affected CG34396 expression 431 and found no significant difference in mRNA levels between the wild-type and 432 Gar12 mutants (Fig. S3b). Similarly, we generated a Dkc1 null mutant (Dkc11 ) 433 containing a 1-bp insertion and a 5-bp deletion that causes premature 434 translational termination (Fig. 4 h). Similar to the RNAi mutants, the loss of 435 function in both Gar1 and Dkc1 resulted in developmental arrest and larval 436 lethality. On day 2 AEL, the Gar12, Dkc11 and wild type exhibited similar 437 morphology, with one mouth hook tooth and no anterior spiracles (Fig. S5 a–438 S5c'), indicating synchronous development. By 3 d AEL, the wild-type larvae 439 had transitioned to the L2 stage (Fig. S6a and S6a'), whereas both L1 larvae 440 (Fig. S6b–S6c'), and L2 larvae were present in Gar12 and Dkc11 mutants (Fig. 441 S6d–S6e'). On day 4 AEL, wild -type larvae reached the L3 stage (with 11 442 teeth and branched spiracles) (Fig. S7a and S7a'), whereas mutants exhibited 443 only three teeth and club -shaped spiracles (Fig. S7 b–S7c'), confirming 444 developmental delay. By 5 d AEL, mutant showed severe growth impairment, 445 with body lengths of 1.3-1.8 mm compared to 3.8-4.6 mm in control animals 446 (Fig. S8a–S8c). All mutants died between 6 and 11 d AEL , and the lethality 447 could not be improved by 20-Hydroxyecdysone addition (Fig. S8d and S8e). 448 By 8 d AEL, wild -type animals had proceeded to the pupal stage, whereas 449 Gar12 and Dkc11 mutants exhibited developmental arrest at the larval stage 450 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 16 (Fig. 4i and 4l), displaying L2 features such as club-shaped spiracles and four 451 mouth hook teeth (Fig. 4j and 4m). Genetic rescue of Gar12 using da-Gal4 to 452 drive UAS-Gar1 expression restored the normal development, viability, 453 fertility, and body length (Fig. 4k). Similarly, da-Gal4-driven overexpression of 454 Dkc1-mcherry rescued Dkc11 mutants from the wild-type phenotype (Fig. 4n). 455 In summary, these results demonstrate that the box H/ACA snoRNP complex 456 is essential for individual development in Drosophila. 457 Box H/ACA snoRNP regulates lipid homeostasis through insulin 458 signaling pathway 459 To investigate the regulatory role of Gar1 in lipid metabolism and 460 development, we performed the transcriptomic profiling of Gar1 loss-of-461 function mutant (Gar12). RNA was extracted from the L1 larvae of both wild -462 type and Gar12 mutants, and their transcriptomes were compared. This 463 analysis identified 559 differentially expressed transcripts, with 173 464 upregulated and 395 downregulated in the mutant relative to the wild type (Fig. 465 S9a). Subsequent Gene Set Enrichment Analysis (GSEA) revealed the 466 upregulation of processes related to nucleotide stability and the concurrent 467 downregulation of both fatty acid and lipid metabolic processes in Gar12 468 mutants (Fig. 5a). This downregulation is consistent with the observed 469 reduction in lipid storage in the fat body upon Gar1 silencing. Because the 470 snoRNP complex has been reported to be involved in RNA splicing, we 471 analyzed alternative changes between Gar12 homozygous mutants and wild-472 type animals . We detected five types of AS events: skipped exon (SE), 473 mutually exclusive exon (MXE), alternative 5 ’-Splice site (A5SS), alternative 474 3’-Splice site (A3SS) and retained intron (RI). Notably, 366 significant 475 differentially spliced events were identified (Fig. S9 b), including A5SS 476 (14.75%), A3SS (22.40%), MXE (6.83%), RI (22.68%), and SE (33.33%). 477 Among these, splicing factor encoding genes displayed significant AS 478 changes, including Rbfox1 and HnRNP-K (Fig. 5b). Interestingly, we also 479 found key components of the insulin signal showed AS changes in the Gar12 480 mutant flies, including chico, Pi3K92E (dp110), sgg (GSK-3β), and Lip4 s (Fig. 481 5b–5e). Notably, AKHR, an antagonist of insulin signaling, also exhibited AS 482 changes (Fig. 5b). Meanwhile SREBP, which encodes a sterol regulatory 483 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 17 element-binding protein, underwent alternative splicing in an A3SS manner 484 (Fig. 5b and 5 f) These splicing alterations in the insulin pathway genes 485 suggest the potential disruption of insulin signaling in Gar1 mutant. 486 To assess the functional consequences of AS changes in the insulin 487 pathway, we examined the localization of the tGPH reporter, an in vivo 488 reporter for PI3K activity (Britton et al., 2002). In the larval fat body at 68-72 h 489 AEL, we examined the fluorescence intensity and distribution of tGPH 490 between the membrane and cytosol. Knockdown of Dkc1 and Gar1 reduced 491 the overall tGPH fluorescence by 35.96% and 47.07%, respectively (Fig. 6a 492 and 6b). In control flies ( da-Gal4), the tGPH signal was detected in the 493 membrane, cytosol, and nucleus of both the salivary glands and fat bodies(Fig. 494 6a). Notably, Gar1 knockdown specifically caused a more pronounced 495 decrease in membrane-localized tGPH (Fig. 6c). Then we performed a similar 496 analysis of tGPH in salivary gland cells. As expected, the fluorescence 497 intensity was dramatically decreased following the knockdown of Dkc1 or 498 Gar1 (Fig. 6d and 6e). Lines were drawn between the nuclei of adjacent cells 499 to measure the distribution of the fluoresce nce signal (Fig. 6d; black lines). 500 The normalized intensity readings consistently remained at high levels 501 (greater than 80%), with a distinct peak observable at the membrane in both 502 control (da-Gal4) and da>Dkc1 RNAi groups (arrows in Fig . 6f and 6h). In 503 contrast, the Gar1 RNAi sample displayed a flat trough on the membrane with 504 no detectable peak s (arrow in Fig. 6g). These findings indicate that the 505 dysfunction of the box H/ACA snoRNP complex impairs insulin signaling 506 effector activity, revealing a strong association between box H/ACA snoRNP 507 and the insulin pathway. 508 Given the previous findings that impaired box H/ACA snoRNP function 509 disrupted lipid storage, we explored the genetic interactions in lipid deposition 510 between Gar1 and key genes of the insulin pathway. To this end, we 511 employed RNAi strains targeting insulin signaling components and examined 512 the resulting changes in larval salivary gland s and fat bodies. As expected, 513 we found that the knockdown of lin-28 and foxo restored lipid deposition 514 defects in Gar1 mutant. In the fat body, the knockdown of lin-28 (lin-28Th01982.N) 515 and foxo (foxoHMS00793) driven by ppl-Gal4 displayed reduced lipid droplet size 516 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 18 compared to wild type, with average diameter of the 30 largest lipid droplets 517 measuring 12.62 μm and 12.02 μm, respectively (Fig. 7 a and 7b). Notably, 518 foxo knockdown in Gar1HMS00979 mutants restored the lipid droplet size to wild-519 type levels (Fig. 7a and 7b), indicating that FOXO and GAR1 function in the 520 same pathway and complementarily regulate lipid storage in the adipose 521 tissue. Conversely, lin-28 knockdown in the Gar1HMS00979 background was 522 similar to that of lin-28 single mutant, indicating that lin-28 acts downstream 523 of Gar1. We further examined genetic interactions in the salivary gland. 524 Knockdown of lin-28 or foxo alone using ppl-Gal4 alone did not induce ectopic 525 lipid droplet accumulation. Importantly, the knockdown of either gene 526 suppressed ectopic lipid droplet formation in Gar1HMS00979 mutant (Fig. 7 c), 527 confirming that lin-28 and foxo function downstream of Gar1. Taken together, 528 these results demonstrate that box H/ACA snoRNP complexes maintain lipid 529 homeostasis through the insulin pathway, underscoring the tight functional 530 connection between snoRNP and metabolic processes. 531

532 Adipose tissue dysfunction is a major emerging driver of metabolic 533 syndrome, frequently characterized by excessive lipid accumulation in non -534 adipose tissues, such as the liver, muscle, pancreas, and renal sinus 535 (Neeland et al., 2024). In the present study, a genetic screen was designed 536 using the DGAT overexpression background. Using alterations in LD size as 537 the phenotypic readout, we identified some well-known lipogenic and lipolytic 538 genes (Table 1), indicating that this screen was effective. A key advantage of 539 this screen was its capacity to distinguish fat body specific changes (Class VI 540 and VII, Fig. 1g) from ectopic lipid deposition in the salivary glands (Class I, 541 II, and III, Fig. 1 b – 1d). Additionally, DIC is an effective technology for 542 identifying subtle changes in lipid droplets without disrupting their native 543 states. In fact, DIC microscopy enables the tracking of single lipid droplet (LD) 544 dynamics over multi -day timescales (Lyn et al., 2010) . In summary, this 545 screen provides a high throughput platform for analyzing changes in both non-546 adipose and adipose cells in Drosophila. 547 We identified Gar1, which encodes a box H/ACA snoRNA-binding protein 548 that regulates lipid storage. Knockdown of Gar1 function led to dramatic 549 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 19 changes in lipid deposition in both the salivary gland and fat body (Fig. 3a and 550 S2a). The homozygous Gar1 null mutant arrested at the L2 stage (Fig. 4i and 551 4j). This phenotype mimics the delayed larval development caused by Gar1 552 deletion in C. elegans , indicating that GAR1 has a conserved function 553 (Spaulding et al., 2022). GAR1 was initially identified in yeast for its 554 involvement in rRNA processing and interactions with snRNAs (Girard et al., 555 1992) and associates with NHP2, DKC1 , and NOP10 to form the snoRNP 556 complex, regulating protein synthesis, mRNA splicing, and maintenance of 557 genome integrity (Watkins et al., 1998; Henras et al., 1998; Mitchell et al., 558 1999; Pogacic et al., 2000; Kiss et al., 2010). Furthermore, we found that 559 knockdown of Dkc1, Nhp2 or Nop10 led to decreased lipid storage (Fig. 3e–560 3h), and disruption of Dkc1 led to lethality at the L2 stage (Fig. 4b–4e, 4l, 4m), 561 indicating the dysfunction of other H/ACA snoRNA -binding protein 562 phenocopies Gar1 mutant. These results demonstrate that all four proteins 563 function integrally , which may explain the essential roles of snoRNPs in 564 developmental processes and energy homeostasis documented in 565 Drosophila, human cells, and Arabidopsis (Angrisani et al., 2018; Armendariz 566 et al., 2025; Belli et al., 2019; Li et al., 2023; Zeng et al., 2022). For example, 567 NOP56 (component of box C/D snoRNP) and DKC1 were required for root 568 formation and plant height development in Arabidopsis thaliana , with 569 concomitant defects in pseudouridylation and 2' -O-methylation in loss of 570 function mutants (Li et al., 2023). Another study demonstrated that 571 downregulation of Fib, which encodes the box C/D component, attenuated 572 ethanol stress and activation of the Ethanol Stress Response Element 573 (ESRE), and the ESRE was rescued by supplementation with 574 polyunsaturated fatty acids (Armendariz et al., 2025). We hypothesized that 575 snoRNP dysfunction leads to impaired lipid metabolism and developmental 576 defects by mediating key pathways that participate in lipid metabolism and 577 development. 578 To verify this hypothesis, we analyzed alternative splicing changes in 579 Gar12 mutant compared to the wild type and found that AS events occurred 580 in key regulators of the insulin signaling pathway (Fig. 5b–5f). Insulin signaling 581 is modulated at the post -transcriptional level by alternative splicing. For 582 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 20 instance, the insulin receptor has two isoforms, INSR -A and INSR -B, which 583

from the skipping or inclusion of exon 11, respectively (Li and Huang, 584 2024; Malakar et al., 2016). Additionally, several genes related to glucose and 585 lipid metabolism are regulated by AS, such as LIPIN1, PPARγ, FADS1, and 586 mTOR (Kaminska, 2025; Liu and Klein, 2018). These results confirmed that 587 genes in the insulin pathway were alternatively spliced, suggesting a key role 588 for AS in metabolic regulation. Dkc1 depletion was known to attenuate AKT-589 mTOR signaling through suppressed phosphorylation of 4EBP1, AKT, 590 P70S6K, and GSK -3β (Maiello et al., 2022). Consistent with this, we also 591 observed reduced tGPH signals in both Dkc1-KD and Gar1-KD flies (Fig. 6a–592 6h), indicating a functional link between the insulin signaling pathway and the 593 snoRNP complex. Lin-28 promotes de novo fatty acid synthesis by binding to 594 the SREBP-1/SCAP mRNAs and enhancing SREBP-1 expression (Zhang et 595 al., 2019). Insulin-dependent phosphorylation leads to nuclear exclusion of 596 FOXO, whereas nuclear retention is linked to insulin resistance, reduced 597 lipogenesis, and enhanced lipolysis (Lee and Dong, 2017). In present study, 598 downregulation of Gar1 led to reduced lipid storage and increased EFA levels. 599 These lipid storage defects were suppressed by the knockdown of either lin-600 28 or foxo (Fig. 7a–7c), indicating a functional connection between the box 601 H/ACA complex and the insulin signaling pathway. However, the specific 602 snoRNA effectors and downstream targets of the box H/ACA snoRNP 603 complexes in lipid homeostasis remain unclear. 604 Several studies have highlighted the roles of snoRNAs in lipid 605 metabolism. Box C/D snoRNAs, U32a/U33/U35a, confer resistance to 606 lipotoxicity induced by a high -fat diet (Michel et al., 2011). The C/D snoRNA 607 U60 knockdown resulted in impaired cholesterol trafficking (Brandis et al., 608 2013). H/ACA snoRNA U17 deficiency disrupts the intracellular cholesterol 609 transport in CHO cells (Jinn et al., 2015). Furthermore, mice deficient in U32a, 610 U33, U34, and U35a exhibited improved glucose tolerance, enhanced insulin 611 secretion, reduced ROS production , and better responses to diabetogenic 612 stimuli (Lee et al., 2016). In fact, the adipose tissue is far more than a passive 613 energy storage organ. As an active endocrine entity, it engages in a complex 614 bidirectional dialogue with the brain by releasing upd2, thereby systemically 615 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 21 regulating insulin secretion and whole -body metabolic homeostasis 616 (Ingaramo et al., 2020; Rajan and Perrimon, 2012). Furthermore, insulin 617 secretion is modulated by Rbfox2 via alternative splicing to maintain glucose 618 homeostasis (Moss et al., 2023). Therefore, we hypothesized that snoRNAs, 619 as key epigenetic regulators, may also play a profound role in this 620 sophisticated adipose -brain crosstalk, offering a perspective for 621 understanding the metabolic regulation network. Furthermore, identifying and 622 characterizing specific snoRNAs involved in lipid homeostasis will provide 623 targets for the pathological basis of metabolic syndromes. 624 Data availability statement 625 Fly lines are available upon request. RNA-seq data are available in the 626 NCBI Sequence Read Archive (SRA) database (accession number 627 PRJNA1439316). https://www.ncbi.nlm.nih.gov/sra/PRJNA1439316. The 628 authors affirm that all other data necessary for confirming the conclusions of 629 the article are present within the article, figures, and tables. Supplemental 630

available at GENETICS online. 631 Acknowledgments 632 We thank Dr. Jiahuai Han for providing the EP stocks; Dr. Pierre Leopold 633 for providing the ppl-Gal4 line; Dr. Renjie Jiao and Dr. Haiyang Chen for 634 sgRNA design and transgenic plasmid construction; Dr. Jingyan Zhang for 635 suggestions on improving the figures; and Editage (www.editage.cn) for 636 English language editing. 637 Study funding 638 This work was supported by the National Natural Science Foundation of 639 China [grant number 32260220]; Scientific Research Foundation of Guizhou 640 University [grant number 2014-37]. 641

642 Ambros, V., and Horvitz, H.R. 1984. Heterochronic mutants of the nematode Caenorhabditis 643 elegans. Science 226, 409–416. doi:10.1126/science.6494891. 644 Angrisani, A., Matrone, N., Belli, V., Vicidomini, R., Di Maio, N., Turano, M., Scialo, F., Netti, 645 P.A., Porcellini, A., and Furia, M. 2018. A functional connection between dyskerin and 646 energy metabolism. Redox Biol. 14, 557–565. doi:10.1016/j.redox.2017.11.003. 647 Armendariz, L., Chan, A., Tjahjono, E., Wang, M., and Kirienko, N.V. 2025. Box C/D snoRNPs 648 and MDT -15/MED15 regulate mitochondrial surveillance via fatty acid metabolism. 649 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 22 bioRxiv, 656193. doi:10.1101/2025.05.26.656193. 650 Bellen, H.J., Levis, R.W., He, Y., Carlson, J.W., Evans-Holm, M., Bae, E., Kim, J., Metaxakis, 651 A., Savakis, C., Schulze, K.L., et al. 2011. The Drosophila gene disruption project: 652 progress using transposons with distinctive site specificities. Genetics 188, 731 –743. 653 doi:10.1534/genetics.111.126995. 654 Belli, V., Matrone, N., Sagliocchi, S., Incarnato, R., Conte, A., Pizzo, E., Turano, M., Angrisani, 655 A., and Furia, M. 2019. A dynamic link between H/ACA snoRNP components and 656 cytoplasmic stress granules. Biochim. Biophys. Acta Mol. Cell Res. 1866, 118529. 657 doi:10.1016/j.bbamcr.2019.118529. 658 Brandis, K.A., Gale, S., Jinn, S., Langmade, S.J., Dudley -Rucker, N., Jiang, H., Sidhu, R., 659 Ren, A., Goldberg, A., Schaffer, J.E., et al. 2013. Box C/D small nucleolar RNA (snoRNA) 660 U60 regulates intracellular cholesterol trafficking. J. Biol. Chem. 288, 35703 –35713. 661 doi:10.1074/jbc.M113.488577. 662 Breznak, S.M., Peng, Y., Deng, L., Kotb, N.M., Flamholz, Z., Rapisarda, I.T., Martin, E.T., 663 LaBarge, K.A., Fabris, D., Gavis, E.R., et al. 2023. H/ACA snRNP -dependent ribosome 664 biogenesis regulates translation of polyglutamine proteins. Sci. Adv. 9, eade5492. 665 doi:10.1126/sciadv.ade5492. 666 Britton, J.S., Lockwood, W.K., Li, L., Cohen, S.M., and Edgar, B.A. 2002. Drosophila's 667 Insulin/PI3-Kinase pathway coordinates cellular metabolism with nutritional conditions. 668 Dev. Cell 2, 239–249. doi:10.1016/s1534-5807(02)00117-x. 669 Camporez, J.P., Kanda, S., Petersen, M.C., Jornayvaz, F.R., Samuel, V.T., Bhanot, S., 670 Petersen, K.F., Jurczak, M.J., and Shulman, G.I. 2015. ApoA5 knockdown improves 671 whole-body insulin sensitivity in high -fat-fed mice by reducing ectopic lipid content. J. 672 Lipid Res. 56, 526–536. doi:10.1194/jlr.M054080. 673 Caron, A., Richard, D., and Laplante, M. 2015. The roles of mTOR complexes in lipid 674 metabolism. Annu. Rev. Nutr. 35, 321–348. doi:10.1146/annurev-nutr-071714-034355. 675 Chamousset, D., De Wever, V., Moorhead, G.B., Chen, Y., Boisvert, F. -M., Lamond, A.I., 676 Trinkle-Mulcahy, L., and Matera, A.G. 2010. RRP1B Targets PP1 to mammalian cell 677 nucleoli and is associated with pre -60S ribosomal subunits. Mol. Biol. Cell 21, 4212 –678 4226. doi:10.1091/mbc.E10-04-0287. 679 Chen, F.J., Yin, Y., Chua, B.T., and Li, P. 2020. CIDE family proteins control lipid homeostasis 680 and the development of metabolic diseases. Traffic 21, 94–105. doi:10.1111/tra.12717. 681 Chen, G., Harwood, J.L., Lemieux, M.J., Stone, S.J., and Weselake, R.J. 2022. acyl -682 coA:diacylglycerol acyltransferase: Properties, physiological roles, metabolic 683 engineering and intentional control. Prog. Lipid Res. 88, 101181. 684 doi:10.1016/j.plipres.2022.101181. 685 Di Maio, N., Vicidomini, R., Angrisani, A., Belli, V., Furia, M., and Turano, M. 2017. A new 686 role for human dyskerin in vesicular trafficking. FEBS Open Bio 7, 1453 –1468. 687 doi:10.1002/2211-5463.1230. 688 Einson, J., Glinos, D., Boerwinkle, E., Castaldi, P., Darbar, D., de Andrade, M., Ellinor, P., 689 Fornage, M., Gabriel, S., Germer, S., et al. 2023. Genetic control of mRNA splicing as a 690 potential mechanism for incomplete penetrance of rare coding variants. Genetics 224. 691 doi: 10.1093/genetics/iyad115. 692 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 23 Fatica, A., Dlakic, M., and Tollervey, D. 2002. Naf1 p is a box H/ACA snoRNP assembly 693 factor. RNA 8, 1502–1514. doi: 10.1017/S1355838202022094. 694 Ferrara, D., Montecucco, F., Dallegri, F., and Carbone, F. 2019. Impact of different ectopic 695 fat depots on cardiovascular and metabolic diseases. J. Cell. Physiol. 234, 21630–21641. 696 doi:10.1002/jcp.28821. 697 Finegan, T.M., Deem, K., Tan, R., Lowe, N., Weeks, N., Linhoff, M.W., Wang, X., L ópez-698 Molini, I., Benraiss, A., and Bergstralh, D.T. 2025. Alternative splicing repurposes the 699 Drosophila mitotic regulator Mud for meiotic functions. Genetics 231. doi: 700 10.1093/genetics/iyaf145. 701 French, R.L. 2014. Developmental ethanol exposure leads to dysregulation of lipid 702 metabolism and oxidative stress in Drosophila. G3 (Bethesda) 5, 49 – 59. 703 doi: 10.1534/g3.114.015040. 704 Girard, J.P., Lehtonen, H., Caizergues-Ferrer, M., Amalric, F., Tollervey, D., and Lapeyre, B. 705 1992. GAR1 is an essential small nucleolar RNP protein required for pre -rRNA 706 processing in yeast. EMBO J. 11, 673–682. doi: 10.1002/j.1460-2075.1992.tb05099.x. 707 Gonzalez-Itier, S., Contreras, E.G., Larrain, J., Glavic, A., and Faunes, F. 2018. A role for 708 Lin-28 in growth and metamorphosis in Drosophila melanogaster. Mech. Dev. 154, 107–709 115. doi:10.1016/j.mod.2018.06.002. 710 Hamid, R., and Mishra, R.K., 2021. Experiments with Drosophila for biology courses, first ed, 711 Indian Academy of Sciences, Bengaluru. 712 Heier, C., and Kuhnlein, R.P. 2018. Triacylglycerol metabolism in Drosophila melanogaster. 713 Genetics 210, 1163–1184. doi:10.1534/genetics.118.301583. 714 Heiss, N.S., Knight, S.W., Vulliamy, T.J., Klauck, S.M., Wiemann, S., Mason, P.J., Poustka, 715 A., and Dokal, I. 1998. X-linked dyskeratosis congenita is caused by mutations in a highly 716 conserved gene with putative nucleolar functions. Nat. Genet. 19, 32 –38. 717 doi:10.1038/ng0598-32. 718 Henras, A., Henry, Y., Bousquet -Antonelli, C., Noaillac -Depeyre, J., Gelugne, J.P., and 719 Caizergues-Ferrer, M. 1998. Nhp2p and Nop10p are essential for the function of H/ACA 720 snoRNPs. EMBO J. 17, 7078–7090. doi:10.1093/emboj/17.23.7078. 721 Housden, B.E., Valvezan, A.J., Kelley, C., Sopko, R., Hu, Y., Roesel, C., Lin, S., Buckner, 722 M., Tao, R., Yilmazel, B., et al. 2015. Identification of potential drug targets for tuberous 723 sclerosis complex by synthetic screens combining CRISPR-based knockouts with RNAi. 724 Sci. Signal. 8, rs9. doi:10.1126/scisignal.aab3729. 725 Ingaramo, M.C., Sá nchez, J.A., Perrimon, N., and Dekanty, A. 2020. Fat body p53 regulates 726 systemic insulin signaling and autophagy under nutrient stress via Drosophila upd2 727 repression. Cell Rep. 33, 108321. doi:10.1016/j.celrep.2020.108321. 728 Jacob, M.D., Audas, T.E., Uniacke, J., Trinkle -Mulcahy, L., Lee, S., and Matera, A.G. 2013. 729 Environmental cues induce a long noncoding RNA -dependent remodeling of the 730 nucleolus. Mol. Biol. Cell 24, 2943–2953. doi:10.1091/mbc.E13-04-0223. 731 Jinn, S., Brandis, K.A., Ren, A., Chacko, A., Dudley -Rucker, N., Gale, S.E., Sidhu, R., 732 Fujiwara, H., Jiang, H., Olsen, B.N., et al. 2015. snoRNA U17 regulates cellular 733 cholesterol trafficking. Cell Metab. 21, 855–867. doi:10.1016/j.cmet.2015.04.010. 734 Kamareddine, L., Robins, W.P., Berkey, C.D., Mekalanos, J.J., and Watnick, P.I. 2018. The 735 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 24 Drosophila immune deficiency pathway modulates enteroendocrine function and host 736 metabolism. Cell Metab. 28, 449–462. doi:10.1016/j.cmet.2018.05.026. 737 Kaminska, D. 2025. Alternative splicing regulation in metabolic disorders. Obes. Rev. 26, 738 e13950. doi:10.1111/obr.13950. 739 Kanno, T., Chiou, P., Wu, M.T., Lin, W.D., Matzke, A., and Matzke, M. 2023. A GFP splicing 740 reporter in a coilin mutant background reveals links between alternative splicing, siRNAs, 741 and coilin function in Arabidopsis thaliana. G3 (Bethesda) 13. doi: 742 10.1093/g3journal/jkad175. 743 Kiss, T., Fayet-Lebaron, E., and Jady, B.E. 2010. Box H/ACA small ribonucleoproteins. Mol. 744 Cell 37, 597–606. doi:10.1016/j.molcel.2010.01.032. 745 Krycer, J.R., Quek, L.-E., Francis, D., Zadoorian, A., Weiss, F.C., Cooke, K.C., Nelson, M.E., 746 Diaz-Vegas, A., Humphrey, S.J., Scalzo, R., et al. 2020. Insulin signaling requires 747 glucose to promote lipid anabolism in adipocytes. J. Biol. Chem. 295, 13250 –13266. 748 doi:10.1074/jbc.RA120.014907. 749 Lee, J., Harris, A.N., Holley, C.L., Mahadevan, J., Pyles, K.D., Lavagnino, Z., Scherrer, D.E., 750 Fujiwara, H., Sidhu, R., Zhang, J., et al. 2016. Rpl13a small nucleolar RNAs regulate 751 systemic glucose metabolism. J. Clin. Invest. 126, 4616–4625. doi:10.1172/JCI88069. 752 Lee, S., and Dong, H.H. 2017. FoxO integration of insulin signaling with glucose and lipid 753 metabolism. J. Endocrinol. 233, R67–R79. doi:10.1530/JOE-17-0002. 754 Lermontova, I., Schubert, V., Bö rnke, F., Macas, J., and Schubert, I. 2007. Arabidopsis CBF5 755 interacts with the H/ACA snoRNP assembly factor NAF1. Plant Mol. Biol. 65, 615 –626. 756 doi:10.1007/s11103-007-9226-z. 757 Li, J., and Huang, G. 2024. Insulin receptor alternative splicing in breast and prostate cancer. 758 Cancer Cell Int. 24, 62. doi:10.1186/s12935-024-03252-1. 759 Li, X., Yan, Z., Zhang, M., Wang, J., Xin, P., Cheng, S., Kou, L., Zhang, X., Wu, S., Chu, J., 760 et al. 2023. SnoRNP is essential for thermospermine -mediated development in 761 Arabidopsis thaliana. Sci. China Life Sci. 66, 2–11. doi:10.1007/s11427-022-2235-4. 762 Lim, W.L.F., Martins, I.J., and Martins, R.N. 2014. The Involvement of lipids in Alzheimer's 763 disease. J. Genet. Genomics 41, 261–274. doi:10.1016/j.jgg.2014.04.003. 764 Liu, X., and Klein, P.S. 2018. Glycogen synthase kinase -3 and alternative splicing. Wiley 765 Interdiscip. Rev. RNA 9, e1501. doi:10.1002/wrna.1501. 766 Lin, X., and Smagghe, G. 2019. Roles of the insulin signaling pathway in insect development 767 and organ growth. Peptides 122, 169923. doi:10.1016/j.peptides.2018.02.001. 768 Liu, Z., and Huang, X. 2013. Lipid metabolism in Drosophila: development and disease. Acta 769 Biochim. Biophys. Sin. (Shanghai) 45, 44–50. doi:10.1093/abbs/gms105. 770 Ljubkovic, M., Gressette, M., Bulat, C., Cavar, M., Bakovic, D., Fabijanic, D., Grkovic, I., 771 Lemaire, C., and Marinovic, J. 2019. Disturbed fatty acid oxidation, endoplasmic 772 reticulum stress, and apoptosis in left ventricle of patients with type 2 diabetes. Diabetes 773 68, 1924–1933. doi:10.2337/db19-0423. 774 Lu, B., Li, Q., Wang, H., Cheng, J., Zhang, Y., Chen, X., Li, B., and Zhao, Y. 2025. 775 SNORD88B promotes proliferation and invasion of trophoblast by regulating G3BP1 m6 776 A modification and alternative splicing. Cell. Mol. Life Sci. 82, 225. doi:10.1007/s00018-777 025-05758-x. 778 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 25 Ly, L.D., Xu, S., Choi, S.-K., Ha, C.-M., Thoudam, T., Cha, S.-K., Wiederkehr, A., Wollheim, 779 C.B., Lee, I. -K., and Park, K. -S. 2017. Oxidative stress and calcium dysregulation by 780 palmitate in type 2 diabetes. Exp. Mol. Med. 49, e291. doi:10.1038/emm.2016.157. 781 Lyn, R.K., Kennedy, D.C., Stolow, A., Ridsdale, A., and Pezacki, J.P. 2010. Dynamics of lipid 782 droplets induced by the hepatitis C virus core protein. Biochem. Biophys. Res. Commun. 783 399, 518–524. doi:10.1016/j.bbrc.2010.07.101. 784 Maiello, D., Varone, M., Vicidomini, R., Belli, V., De Rosa, M., Dama, P., Furia, M., and 785 Turano, M. 2022. Dyskerin downregulation can induce ER stress and promote 786 autophagy via AKT -mTOR signaling deregulation. Biomedicines 10, 787 1092. doi:10.3390/biomedicines10051092. 788 Malakar, P., Chartarifsky, L., Hija, A., Leibowitz, G., Glaser, B., Dor, Y., and Karni, R. 2016. 789 Insulin receptor alternative splicing is regulated by insulin signaling and modulates beta 790 cell survival. Sci. Rep. 6, 31222. doi:10.1038/srep31222. 791 Massenet, S., Bertrand, E., and Verheggen, C. 2017. Assembly and trafficking of box C/D 792 and H/ACA snoRNPs. RNA Biol. 14, 680–692. doi:10.1080/15476286.2016.1243646. 793 McMahon, M., Contreras, A., and Ruggero, D. 2015. Small RNAs with big implications: new 794 insights into H/ACA snoRNA function and their role in human disease. Wiley Interdiscip. 795 Rev. RNA 6, 173–189. doi:10.1002/wrna.1266. 796 Michel, C.I., Holley, C.L., Scruggs, B.S., Sidhu, R., Brookheart, R.T., Listenberger, L.L., 797 Behlke, M.A., Ory, D.S., and Schaffer, J.E. 2011. Small nucleolar RNAs U32a, U33, and 798 U35a are critical mediators of metabolic stress. Cell Metab. 14, 33 –44. 799 doi:10.1016/j.cmet.2011.04.009. 800 Mitchell, J.R., Wood, E., and Collins, K. 1999. A telomerase component is defective in the 801 human disease dyskeratosis congenita. Nature 402, 551–555. doi: 10.1038/990141. 802 Morita, S., Ota, R., and Kobayashi, S. 2018. Downregulation of NHP2 promotes proper cyst 803 formation in Drosophila ovary. Dev. Growth Differ. 60, 248–259. doi: 10.1111/dgd.12539. 804 Moss, N.D., Wells, K.L., Theis, A., Kim, Y.K., Spigelman, A.F., Liu, X., MacDonald, P.E., and 805 Sussel, L. 2023. Modulation of insulin secretion by RBFOX2 -mediated alternative 806 splicing. Nat. Commun. 14, 7732. doi:10.1038/s41467-023-43605-4. 807 Musselman, L.P., and Kuhnlein, R.P. 2018. Drosophila as a model to study obesity and 808 metabolic disease. J. Exp. Biol. 221. doi:10.1242/jeb.163881. 809 Nassel, D.R., Liu, Y., and Luo, J. 2015. Insulin/IGF signaling and its regulation in Drosophila. 810 Gen. Comp. Endocrinol. 221, 255–266. doi:10.1016/j.ygcen.2014.11.021. 811 Neeland, I.J., Lim, S., Tchernof, A., Gastaldelli, A., Rangaswami, J., Ndumele, C.E., Powell-812 Wiley, T.M., and Despré s, J.P. 2024. Metabolic syndrome. Nat. Rev. Dis. Primers 10, 77. 813 doi:10.1038/s41572-024-00563-5. 814 Ojha, S., Malla, S., and Lyons, S.M. 2020. snoRNPs: Functions in ribosome biogenesis. 815 Biomolecules 10, 783. doi:10.3390/biom10050783. 816 Pogacic, V., Dragon, F., and Filipowicz, W. 2000. Human H/ACA small nucleolar RNPs and 817 telomerase share evolutionarily conserved proteins NHP2 and NOP10. Mol. Cell. Biol. 818 20, 9028–9040. doi:10.1128/MCB.20.23.9028-9040.2000. 819 Rajan, A., and Perrimon, N. 2012. Drosophila cytokine unpaired 2 regulates physiological 820 homeostasis by remotely controlling insulin secretion. Cell 151, 123 –137. 821 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 26 doi:10.1016/j.cell.2012.08.019. 822 Shen, S., Park, J.W., Lu, Z.X., Lin, L., Henry, M.D., Wu, Y.N., Zhou, Q., and Xing, Y. 2014. 823 rMATS: robust and flexible detection of differential alternative splicing from replicate 824 RNA-Seq data. Proc. Natl. Acad. Sci. U. S. A. 111, E5593 –5601. 825 doi:10.1073/pnas.1419161111. 826 Spaulding, E.L., Feidler, A.M., Cook, L.A., and Updike, D.L. 2022. RG/RGG repeats in the C. 827 elegans homologs of Nucleolin and GAR1 contribute to sub-nucleolar phase separation. 828 Nat. Commun. 13, 6585. doi:10.1038/s41467-022-34225-5. 829 Światowy, W., and Jagodzińśki, P.P. 2018. Molecules derived from tRNA and snoRNA: 830 Entering the degradome pool. Biomed. Pharmacother. 108, 36 –42. 831 doi:10.1016/j.biopha.2018.09.017. 832 Tian, Y., Bi, J., Shui, G., Liu, Z., Xiang, Y., Liu, Y., Wenk, M.R., Yang, H., and Huang, X. 2011. 833 Tissue-autonomous function of Drosophila seipin in preventing ectopic lipid droplet 834 formation. PLoS Genet. 7, e1001364. doi:10.1371/journal.pgen.1001364. 835 Trouwborst, I., Bowser, S.M., Goossens, G.H., and Blaak, E.E. 2018. Ectopic Fat 836 Accumulation in Distinct Insulin Resistant Phenotypes; Targets for Personalized 837 Nutritional Interventions. Front. Nutr. 5, 77. doi:10.3389/fnut.2018.00077. 838 Vicidomini, R., Petrizzo, A., di Giovanni, A., Cassese, L., Lombardi, A.A., Pragliola, C., and 839 Furia, M. 2017. Drosophila dyskerin is required for somatic stem cell homeostasis. Sci. 840 Rep. 7, 347. doi:10.1038/s41598-017-00446-8. 841 Vulliamy, T., Beswick, R., Kirwan, M., Marrone, A., Digweed, M., Walne, A., and Dokal, I. 842 2008. Mutations in the telomerase component NHP2 cause the premature ageing 843 syndrome dyskeratosis congenita. Proc. Natl. Acad. Sci. U. S. A. 105, 8073 –8078. 844 doi:10.1073/pnas.0800042105. 845 Walne, A.J., Vulliamy, T., Marrone, A., Beswick, R., Kirwan, M., Masunari, Y., Al -Qurashi, 846 F.H., Aljurf, M., and Dokal, I. 2007. Genetic heterogeneity in autosomal recessive 847 dyskeratosis congenita with one subtype due to mutations in the telomerase-associated 848 protein NOP10. Hum. Mol. Genet. 16, 1619–1629. doi:10.1093/hmg/ddm111. 849 Wang, Y., Fu, M., Zheng, Z., Feng, J., and Zhang, C. 2025. Small Nucleolar RNAs: Biological 850 Functions and Diseases. MedComm (2020) 6, e70257.doi: 10.1002/mco2.70257. 851 Watkins, N.J., Gottschalk, A., Neubauer, G., Kastner, B., Fabrizio, P., Mann, M., and 852 Luhrmann, R. 1998. Cbf5p, a potential pseudouridine synthase, and Nhp2p, a putative 853 RNA-binding protein, are present together with Gar1p in all H BOX/ACA-motif snoRNPs 854 and constitute a common bipartite structure. RNA 4, 1549 –1568. 855 doi:10.1017/s1355838298980761. 856 Xu, S., Zhang, X., and Liu, P. 2018. Lipid droplet proteins and metabolic diseases. Biochim. 857 Biophys. Acta Mol. Basis Dis. 1864, 1968–1983. doi:10.1016/j.bbadis.2017.07.019. 858 Yan, Y., Wang, H., Hu, M.L., Jiang, L.F., Wang, Y., Liu, P.S., Liang, X.H., Liu, J.Y., Li, C.Q., 859 Lindstrom-Battle, A., et al. 2017. HDAC6 Suppresses age -dependent ectopic fat 860 accumulation by maintaining the proteostasis of PLIN2 in Drosophila. Dev. Cell 43, 99–861 111. doi:10.1016/j.devcel.2017.09.001. 862 Zeng, H., Pan, J., Hu, C., Yang, J., Li, J., Tan, T., Zheng, M., Shen, Y., Yang, T., Deng, Y., 863 et al. 2022. SNHG25 facilitates SNORA50C accumulation to stabilize HDAC1 in 864 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 27 neuroblastoma cells. Cell Death Dis. 13, 597. doi:10.1038/s41419-022-05040-z. 865 Zhang, M., and Zhang, X. 2019. The role of PI3K/AKT/FOXO signaling in psoriasis. Arch. 866 Dermatol. Res. 311, 83–91. doi:10.1007/s00403-018-1879-8. 867 Zhang, T., Li, K., Xiao, Y., Tan, X., Tan, X., Cui, M., and Rong, Y.S. 2025. Angel wing: a class 868 of alternative splicing regulators in animals. Nucleic Acids Res. 53,18. 869 doi:10.1093/nar/gkaf972. 870 Zhang, Y., Li, C., Hu, C., Wu, Q., Cai, Y., Xing, S., Lu, H., Wang, L., Huang, D., Sun, L., et 871 al. 2019. Lin28 enhances de novo fatty acid synthesis to promote cancer progression via 872 SREBP-1. EMBO Rep. 20, e48115. doi:10.15252/embr.201948115. 873 Zhou, Y., Zhou, B., Pache, L., Chang, M., Khodabakhshi, A. H., Tanaseichuk, O., Benner, C., 874 & Chanda, S. K. 2019. Metascape provides a biologist-oriented resource for the analysis 875 of systems-level datasets. Nat. Commun. 10, 1523. doi:10.1038/s41467-019-09234-6. 876 Zhu, H., Shyh -Chang, N., Segrè , Ayellet V., Shinoda, G., Shah, Samar P., Einhorn, 877 William S., Takeuchi, A., Engreitz, Jesse M., Hagan, John P., Kharas, Michael G., et al. 878 2011. The Lin28/let -7 axis regulates glucose metabolism. Cell 147, 81 –94. 879 doi:10.1016/j.cell.2011.08.033. 880 Table 881 Table 1. Genes related to ectopic lipid accumulation identified in the screen 882 Group Gene Difference Function or features Source Class I CG31809 +++ very-long-chain 3-oxoacyl-CoA reductase EP PyK +++ pyruvate kinase EP DIP-ε +++ immunoglobulin subtype EP l (3)72Dn +++ maturation of SSU-rRNA from tricistronic rRNA transcript EP Lipin +++ phosphatidate phosphatase EP CdsA +++ phosphatidate cytidylyltransferase RNAi wun2 ++ lipid phosphate phosphatase EP pes ++ scavenger receptor EP mbl ++ alternative splicing, transcript localization and miRNA and circRNA biogenesis EP Rpn2 ++ subunit of the 26S proteasome EP tou ++ transcription factor, chromatin remodeling and nervous system development EP Rim ++ GTPase, Rab3 interacting EP GstS1 ++ glutathione S transferase EP Xrp1 ++ genome stability maintenance EP ics ++ Ras suppressor 1 RNAi ND-B18 ++ NADH dehydrogenase RNAi ct ++ transcriptional factor RNAi G6pd ++ glucose-6-phosphate dehydrogenase RNAi GXIVsPLA2 ++ phospholipase RNAi .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 28 yata ++ ATP binding RNAi Pss ++ phosphatidylserine synthase RNAi mEFTu1 ++ mitochondrial translation elongation factor Tu 1 RNAi Sr-CII ++ scavenger receptor RNAi Rae1 ++ WD40-repeat β propeller protein RNAi cid ++ centromere-specific histone H3 variant RNAi GluRIIC + glutamate receptor EP Hsp70Bbb + heat shock protein binding EP CG12464 + NADH-ubiquinone oxidoreductase flavoprotein 3 EP Ppa + F box proteins EP CG12054 + zinc finger and chromatin remodeling-associated protein EP CG14253 + unknown EP Desat1 + stearoyl-CoA 9-desaturase EP CG8468 + MFS transporter EP CG3036 + MFS transporter EP Gar1 + snoRNA binding EP CG43108 + unknown EP cnc + transcription factor EP CG13982 + unknown EP Pld + phospholipase D EP Alg3 + alpha-1,3- mannosyltransferase EP CG11961 + metalloexopeptidase EP PlexA + Semaphorin ligands RNAi Toll-3 + Toll/interleukin-1 receptor RNAi Tig + integrin binding RNAi flamingo + cadherin RNAi Rpt1 + proteasome ATPase RNAi Topors + DNA topoisomerase binding RNAi CG17029 + inositol monophosphate 1- phosphatase RNAi wds + histone acetyltransferase RNAi X11Lβ + beta-amyloid binding RNAi DCAF12 + DDB1-CUL4-X-box E3 ubiquitin- protein ligase RNAi Cdk1 + protein serine/threonine kinase RNAi CG5938 + cysteine rich hydrophobic domain RNAi ato + DNA-binding transcription factor RNAi Zip99C + iron transporter RNAi CG7878 + ATP-dependent RNA helicase RNAi CG7974 + mRNA splicing RNAi trc + protein serine/threonine kinase RNAi CG8878 + protein serine/threonine kinase RNAi .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 29 upSET + histone H3K4me3 reader activity RNAi lid + histone demethylase RNAi BcDNA:LD19168 + nuclear receptor corepressor RNAi Smr + transcription corepressor RNAi Argk2 + creatine kinase RNAi Muc68Ca + extracellular matrix structural constituent RNAi Mayo + G-protein coupled receptor RNAi Class II Sema2b ++ Semaphorin EP Hrb27C ++ mRNA binding RNAi peb ++ DNA binding transcription factor RNAi rb ++ cargo adaptor RNAi 5-HT7 ++ G-protein coupled serotonin receptor RNAi gish + protein serine/threonine kinase EP Nup160 + component of the nuclear pore complex RNAi CG6295 + triacylglycerol lipase RNAi mael + DNA binding RNAi Prosα7 + proteinase RNAi sqh + myosin heavy chain binding RNAi brm + transcription coactivator RNAi yip7 + serine-type endopeptidase RNAi Pino + unknown RNAi Rab5 + GTPase RNAi mon2 + ARF guanyl-nucleotide exchange factor RNAi Class III Baldspot ++ fatty acid elongase EP Chmp1 ++ vacuolar protein sorting-associated RNAi ND51 ++ NADH dehydrogenase RNAi rho-4 ++ serine-type endopeptidase activity RNAi Class IV CG14982 ++ centrosome-associated, FAM110 EP Sbf + Rab guanyl-nucleotide exchange factor RNAi eIF4A3 + RNA binding; ATP binding RNAi SMC3 + chromatin binding RNAi Iswi + DNA-dependent ATPase RNAi Class V AGPAT ++ 1-acylglycerol-3-phosphate O- acyltransferase EP SREBP + DNA binding transcription activator RNAi Class VI lama +++ phospholipase B-like EP lola +++ DNA binding transcription factor EP kis +++ DNA-dependent ATPase RNAi CSN3 +++ proteasome component (PCI) domain RNAi .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 30 Note: "+" indicates a slight change of lipid storage or LDs' morphology in the 883 salivary gland or fat body, "++ "indicates significant change, "+++" indicates dramatic 884 change. 885 886 REPTOR ++ DNA binding transcription factor EP CG42524 ++ unknown EP Pacs + phosphofurin acidic cluster sorting protein 1 EP mir-8 + mRNA 3'-UTR binding EP MCU + calcium channel EP GPAT1 + glycerol-3-phosphate O- acyltransferase EP CG10465 + BTB domain RNAi Class VIII SERCA +++ calcium-transporting ATPase RNAi FASN1 +++ fatty acid synthase RNAi Dro16-0 ++ ecdysteroid kinase-like RNAi Rab39 ++ GTPase RNAi SkpA ++ ubiquitin ligase complex scaffold RNAi .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 31 Figure legends and alt text 887 888 Fig. 1 . The double mutants based on ppl>DGAT exhibit seven distinct 889 morphological types of lipid droplets. a, a') DGAT overexpression induced ectopic 890 lipid droplet accumulation . b–f) The changes in ectopic fat accumulation in larval 891 salivary gland. b'–b') The schematic diagrams of b–f, respectively. g) The changes 892 of lipid droplets in larval fat body, and ppl>DGAT was used as a control. Scale bar: 893 50 μm. 894 ALT TEXT : Graphs depict seven different patterns of lipid deposition changes 895 isolated from the genetic screen based on overexpression of a triglyceride synthesis 896 enzyme, with subfigures labelled from a to g. b to f represents prominent differences 897 in the salivary glands, which is enhanced or diminished lipid accumulation in the 898 salivary gland. g shows alterations increased or decreased sizes of lipid droplets in 899 the fat body. 900 901 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 32 902 Fig. 2. The localization of box H/ACA snoRNA binding proteins. a) GAR1-903 eGFP fusion proteins showed punctate distributions in the nuclei of the salivary gland 904 and fat body. The top panel: fat body, the bottom panel: salivary gland. Green (GAR1-905 eGFP), Cyan (Phalloidin-FITC), Magenta (DAPI). b) DKC1 was co -localized with 906 NHP2. Cyan (DKC1-mCherry), Green (NHP2-eGFP), Magenta (DAPI). c) DKC1 was 907 co-localized with Nop10. Cyan (DKC1-mCherry), Green (Nop10 -eGFP), Magenta 908 (DAPI). Scale bar: 50 μm. 909 ALT TEXT : Graphs illustrate the subcellular localization of four core box H/ACA 910 snoRNA binding proteins. a) shows GAR1 localized in the nucleus with a punctate 911 pattern. (b) and (c) depict DKC1 highly expressed in the nucleus without puncta, 912 colocalized with NOP10 and NHP2. 913 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 33 914 Fig. 3. Dysfunction of Gar1 or Dkc1 exhibited ectopic lipid accumulation 915 in the salivary gland and reduced lipid droplet sized in the fat body . a) 916 Knockdown of Gar1 or Dkc1 led to ectopic lipid deposition. b) Overexpression of 917 Gar1 enhanced ectopic fat accumulation in the salivary gland at the wandering stage. 918 c) Knockdown of Gar1 by ppl-Gal4 resulted in smaller lipid droplets in the fat body at 919 the wandering stage. e, g) Reduced lipid droplet sizes were also observed in 920 knockdown of other snoRNP-encoding genes. Genotypes are indicated in the figure. 921 All RNAi lines were driven by UAS-GAL4 system, and the ‘ppl>’ was an abbreviation 922 for ppl>Gal4. Lipid droplets were stained with Bodipy 493/503 (a, b, e) or Nile red (c, 923 g). Scale bar: 50 μm. d, f, h) LD’ diameters in control and mutants. Violin plots show 924 Max30 LDs from three figures in each genotype (n = 90 LDs per genotype). Box plots 925 inside violins show median and quartiles. Statistical analysis was determined by one-926 way ANOVA using mean values, followed by Dunnett's multiple comparisons test. d) 927 The LD’ diameters in control and Gar1 mutants (CS vs. ppl>Gar1HMS00979, df = 6, 928 adjusted p=0.0003; CS vs. ppl>Gar1GD11196, df = 6, adjusted p = 0.0004). f) The LD’ 929 diameters in control and Dkc1 mutants (ppl-Gal4 vs ppl>Dkc1kk101240, df = 6, adjusted 930 p = 0.0075; ppl-Gal4 vs ppl>Dkc1GL00555, df = 6, adjusted p = 0. 0368) . h) The LD’ 931 diameters in control, Nhp2 mutant and Nop10 mutant (ppl-Gal4 vs ppl>Nhp2HMC03339, 932 df = 6, adjusted p = 0.0051; ppl-Gal4 vs ppl>Nop10HMC03890, df = 6, adjusted p = 933 0.0035). ***, p < 0.001; **, p < 0.01; *, p < 0.05. 934 ALT TEXT : Graphs depict that dysfunction of snoRNP display abnormal lipid 935 deposition. a and b show lipid droplet staining in the salivary gland es. c), e) and g) 936 show reduced lipid droplet size in the fat bodies following knockdown of all four box 937 snoRNP components, with corresponding statistical analyses presented in d), f) and 938 h), respectively. 939 940 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 34 941 Fig. 4 . Loss of Gar1 or Dkc1 function causes developmental delay and 942 arrest at the larval stage. a) Knockdown of Gar1 results in pupal lethality when 943 raised at 29°C driven by ppl-Gal4. b–e) da-Gal4-driven knockdown of Dkc1 at 29°C 944 caused larval lethality by 4 days AEL (b–d) and 6 days AEL (e). f) Genomic locus of 945 Gar1 and its deficiency alleles. The Gar1 gene is located within the intron of 946 CG34396. Deleted regions are indicated by dashed lines; sgRNA target sites are 947 marked by orange rectangles. g) PCR of Gar1 mutants and wild -type controls. 948 Amplification was performed using primers shown as blue boxes in ( f). M, DNA 949 marker. h) Schematic diagram of Dkc11 mutant. i) Gar1 loss of function mutants 950 exhibited developmental arrest on 8 d AEL. j) Mouth hook and spiracle morphology 951 of Gar12 mutant on 8 d AEL. k) The morphology of wild-type and UAS-Gar1 rescued 952 flies at 3 d post -eclosion, raised at 25°C. l) Dkc11 mutant exhibited developmental 953 arrest on 8 d AEL. m) Mouth hook and spiracle morphology of Dkc11 mutant on 8 d 954 AEL. n) The morphology of wild -type and UAS-Dkc1-mCherry rescued flies at 5 d 955 post-eclosion, raised at 25°C. Scale bars were indicated in the relative figures. 956 ALT TEXT : Graphs depict that dysfunction of snoRNP disrupt larval 957 developmental process, with subfigures labelled from a to n. (a–e), (i, j) and (l, m) 958 show Gar1 and Dkc1 mutants died at 1st and 2nd larval stage, as characterized by the 959 morphological changes in the mouth hook and tracheal . f–h) illustrate the deletion 960 sites in the relative mutants. k and n) illustrate the mutants could be restored to the 961 adult stage by overexpressing the wild-type copy of GAR1 or DKC1. 962 963 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 35 964 Fig. 5 . Gar1 regulates lipid metabolism and alternative splicing. a) 965 Positively enriched ( enrich distribution > 0) and negatively enriched (enrich 966 distribution > 0) GO terms identified by Gene Set Enrichment Analysis in 967 Gar1² mutants compared to w1118 controls. b) The distribution of alternative splicing 968 events in Gar1² mutants relative to w1118 controls. Significant s plicing events, 969 including skipped exon (SE), mutually exclusive exon (MXE), alternative 5 ’-Splice 970 site (A5SS), alternative 3’-Splice site (A3SS) and retained intron (RI) were selected 971 with a false discovery rate 0.1 . 972 The highlighted circles indicated the splicing related and insulin related genes. c–f) 973 Sashimi plots depicting splicing patterns of chico, PI3K92E, Gsk3β ( sgg), 974 and SREBP. The inclusion levels of specific exons or alternative splice sites are 975 indicated, revealing isoform switches in the relative genes. 976 ALT TEXT : Graphs depicting that Gar1 regulated lipid metabolism and insulin 977 pathway at both transcriptional and post -transcriptional level s, with subfigures 978 labelled from a to f . a) illustrate significantly upregulated and downregulated GO 979 processes. b) Volcano plot display distribution of splicing events, with notable genes 980 labeled. c-f) represent alternative splicing changes upon loss of Gar1. 981 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 36 982 Fig. 6. Decreased insulin signaling activity in H/ACA snoRNP dysfunction. 983 a) Subcellular localization of the tGPH reporter (green) in third-instar larval fat body. 984 Nuclei are stained with DAPI (Magenta). Scale bar: 50 μm. b) A significant reduction 985 in tGPH fluorescence intensity was observed in both Gar1 RNAi and Dkc1 RNAi fat 986 bodies. Statistical analysis was determined by one -way ANOVA, followed by 987 Dunnett's multiple comparisons test (da-Gal4.tGPH/+ vs. da.tGPH, Gar1GD11196, df = 988 15, adjusted p = 0.0021 ; da-Gal4.tGPH/+ vs. da .tGPH, Dkc1GD10940, df = 15, p = 989 0.0002). c) Alteration in tGPH fluorescence distribution in Gar1 RNAi mutant fat 990 bodies. Statistical analysis was determined by Brown-Forsythe ANOVA, followed by 991 Dunnett's T3 multiple comparisons test (da-Gal4.tGPH/+ vs. da.tGPH, Gar1GD11196, 992 df = 9.409, adjusted p <0.0001; da-Gal4.tGPH/+ vs. da.tGPH, Dkc1GD10940, df = 7.716, 993 adjusted p =0.7921 ). d) tGPH localization in third-instar larval salivary glands of Gar1 994 and Dkc1 RNAi mutants. Green (tGPH), Magenta (DAPI). Scale bar: 50 μm. e) The 995 fluorescence intensity of tGPH in Gar1 RNAi and Dkc1 RNAi salivary glands . 996 Statistical analysis was determined by one -way ANOVA, followed by Dunnett's 997 multiple comparisons test (da-Gal4.tGPH/+ vs. da .tGPH, Gar1GD11196, df = 12, 998 adjusted p = 0.0001; da-Gal4.tGPH/+ vs. da.tGPH, Dkc1GD10940, df = 12, adjusted p 999 = 0.0011). f, g, and h) Normalized fluorescence distribution in salivary gland lines. In 1000 this assay, homozygous da-Gal4.tGPH / da-Gal4.tGPH crossed to w1118, Gar1GD11196 1001 or Dkc1GD10940. The offsprings were raised in the 25 °C after synchronization .The 1002 genotypes are shown in the figures. ****, p < 0.0001;***, p < 0.001; **, p < 0.01; ns, 1003 not significant. 1004 ALT TEXT: Graph depicting the insulin activity decreased upon loss of Gar1, 1005 illustrated with the intensity measurement and distribution analysis of PI3K indicator. 1006 a and b) show reduced lipid droplet size in the fat bodies following knockdown of all 1007 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint 37 four box snoRNP components, with corresponding statistical analyses presented in 1008 d), f) and h), respectively. 1009 1010 Fig. 7. Gar1 acts upstream of the insulin signaling pathway genetically. a) 1011 Bodipy 493/503 staining of lipid droplets in the fat body of wandering -stage larvae 1012 with the indicated genotypes. Green: Bodipy 493/503 (LDs); Magenta: DAPI (nuclei). 1013 Scale bar: 50 μm. b) Size distribution of LDs shown in ( a). Statistical analysis was 1014 performed by two -way ANOVA. LD’ diameters in control and mutants. Violin plots 1015 show Max30 LDs from three figures in each genotype (n = 90 LDs per genotype). 1016 Box plots inside violins show median and quartiles. Statistical analysis was 1017 determined by one -way ANOVA using mean values, followed by Tukey's multiple 1018 comparisons test (ppl-Gal4 vs ppl-Gar1HMS00979, df = 12, adjusted p = 0.0001; ppl-1019 Gal4 vs ppl>lin-28TH01982.N, df = 12, adjusted p = 0.0119; ppl-Gal4 vs ppl> Gar1HMS00979 1020 / lin-28TH01982.N, df = 12, adjusted p = 0.0089; ppl-Gal4 vs. ppl>foxoHMS00793, df = 12, 1021 adjusted p = 0.0029; ppl-Gar1HMS00979 vs ppl>lin-28TH01982.N df = 12, adjusted p = 1022 0.0979; ppl-Gar1HMS00979 vs ppl> Gar1HMS00979 / lin-28TH01982.N, df = 12, adjusted p = 1023 0.1296; ppl>lin-28TH01982.N vs ppl> Gar1HMS00979 / lin-28TH01982.N, df = 12, adjusted 1024 p >0.9999; ppl-Gar1HMS00979 vs ppl> Gar1HMS00979 / foxoHMS00793, df = 12, adjusted p = 1025 0.0003; ppl>foxoHMS00793 vs ppl> Gar1HMS00979 / foxoHMS00793, df = 12, adjusted p = 1026 0.0071; ppl-Gal4 vs. ppl> Gar1HMS00979 / foxoHMS00793, df = 12, adjusted p =0.9931). 1027 ***, p < 0.0001; **, p < 0.01; **, p < 0.05; ns, not significant. c) Knockdown of lin-28 1028 or foxo suppresses the ectopic lipid accumulation in the Gar1HMS00979 mutants. Green: 1029 Bodipy 493/503 (LDs). Scale bar: 50 μm. 1030 ALT TEXT : Graph depicting the genetic interaction between insulin related 1031 genes and Gar1, illustrated with the morphology analysis (a, c) and statistical 1032 visualization of lipid deposition (b). 1033 .CC-BY 4.0 International licensemade available under a (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 The copyright holder for this preprintthis version posted April 1, 2026. ; https://doi.org/10.64898/2026.03.30.715344doi: bioRxiv preprint