Skip to main content
Nucleic Acids Research logoLink to Nucleic Acids Research
. 2024 Oct 1;52(20):12669–12688. doi: 10.1093/nar/gkae810

The TRIM-NHL RNA-binding protein Brain Tumor coordinately regulates expression of the glycolytic pathway and vacuolar ATPase complex

Robert P Connacher 1, Richard T Roden 2, Kai-Lieh Huang 3, Amanda J Korte 4, Saathvika Yeruva 5, Noel Dittbenner 6, Anna J DesMarais 7, Chase A Weidmann 8, Thomas A Randall 9, Jason Williams 10, Traci M Tanaka Hall 11, Eric J Wagner 12, Aaron C Goldstrohm 13,
PMCID: PMC11551770  PMID: 39351871

Abstract

The essential Drosophila RNA-binding protein Brain Tumor (Brat) represses specific genes to control embryogenesis and differentiation of stem cells. In the brain, Brat functions as a tumor suppressor that diminishes neural stem cell proliferation while promoting differentiation. Though important Brat-regulated target mRNAs have been identified in these contexts, the full impact of Brat on gene expression remains to be discovered. Here, we identify the network of Brat-regulated mRNAs by performing RNA sequencing (RNA-seq) following depletion of Brat from cultured cells. We identify 158 mRNAs, with high confidence, that are repressed by Brat. De novo motif analysis identified a functionally enriched RNA motif in the 3′ untranslated regions (UTRs) of Brat-repressed mRNAs that matches the biochemically defined Brat binding site. Integrative data analysis revealed a high-confidence list of Brat-repressed and Brat-bound mRNAs containing 3′UTR Brat binding motifs. Our RNA-seq and reporter assays show that multiple 3′UTR motifs promote the strength of Brat repression, whereas motifs in the 5′UTR are not functional. Strikingly, we find that Brat regulates expression of glycolytic enzymes and the vacuolar ATPase complex, providing new insight into its role as a tumor suppressor and the coordination of metabolism and intracellular pH.

Graphical Abstract

Graphical Abstract.

Graphical Abstract

Introduction

TRIM-NHL proteins regulate stem cell differentiation during neurogenesis across the animal kingdom (1). These proteins are defined by their N-terminal TRIM (tripartite motif) and C-terminal NHL (NCL-1, HT2A and LIN-41) domains (Figure 1A). The Drosophila melanogaster TRIM-NHL protein Brain Tumor (Brat) functions during embryogenesis and stem cell differentiation in the germline and brain (2–8). The extensive research on Brat and molecular tools available for flies make Brat an excellent model for studying the principles of TRIM-NHL function.

Figure 1.

Figure 1.

Identification of Brat-regulated mRNAs by RNA-seq. (A) Diagram of the domain architecture that defines TRIM-NHL proteins. (B) Diagram for Brat protein, following the naming convention used by Loedige et al. (12). (C) Western blot showing efficient depletion of V5-tagged Brat in DL1 cells by RNA interference (RNAi), relative to non-targeting control (NTC) RNAi. RNAi used two independent double-stranded RNAs (dsRNAs #1 or #2), which targeted non-overlapping regions of the brat mRNA. Western blot of tubulin served as a control for equivalent loading of the samples. Size of protein bands in kDa. Depletion of Brat for RNA sequencing (RNA-seq) was conducted in quadruple replicates. (D) Volcano plot showing differential gene expression following depletion of Brat. Genes are represented by log2 fold change (lfc), relative to NTC RNAi and adjusted P-value (Padj). Genes are additionally color-coded by significance-weighted fold change (z-score). (E) Venn diagram showing overlap between genes upregulated (lfc > 0, Padj < 0.05, baseMean ≥ 50) in both RNAi conditions. Significance of overlap determined via Fisher’s exact test (***P< 0.001). (F) Violin plot, with inset boxplot, showing distribution of fold changes for genes upregulated in both RNAi conditions. (G) RNA levels of upregulated genes, relative to 7SK level, measured with reverse transcription-quantitative polymerase chain reaction (RT-qPCR). Mean ± standard deviation values are plotted. Significant differences determined via ANOVA with post-hoc Tukey–Kramer tests (*P< 0.05, **P< 0.01, ***P< 0.001, n = 4). Abbreviations: Dm,Drosophila melanogaster; RING, Really Interesting New Gene; B1 and B2, B-Box zinc fingers 1 and 2; CC, coiled-coil domain; S, serine-rich domain; Q, glutamine-rich domain; H, histidine-rich domain.

Many TRIM-NHL family members, including Brat and human TRIM71, bind specific RNA motifs through their C-terminal NHL domain (1,9–12). TRIM-NHL recognition of these RNA motifs leads to repression manifested in the degradation and/or inhibition of the translation of the bound mRNAs. The absence of the N-terminal RING domain (Figure 1B) reduces the likelihood that Brat functions as a ubiquitin ligase, as suggested for other TRIM-NHL proteins (13). The mechanism by which these proteins mediate mRNA repression and the consequences of this regulation are subject to ongoing research.

Post-transcriptional regulation of gene expression is a well-established role for Brat (1). Extensive genetic analysis has shown that Brat regulates the expression of several key transcription factors across development (1). In the early embryo, Brat contributes to the repression of translation of hunchback mRNA to ensure proper body plan (14) and helps clear maternal mRNAs during the maternal–zygotic transition (3). In larva, Brat ensures differentiation of neural stem cells by repressing zelda and deadpan mRNAs following asymmetric cell division (15,16). In the absence of Brat, the neural stem cells overproliferate, resulting in lethal brain tumors, for which the brat gene was named (2,4,6) Similarly, during oogenesis, Brat represses bone morphogenetic protein (BMP) signaling components to ensure differentiation of germline stem cells (7,8).

The mode by which Brat specifically binds to brat binding site (BBS) motifs in RNA has been extensively characterized through multiple independent approaches. First, two studies using RNA co-immunoprecipitation (RIP) experiments identified hundreds of mRNAs associated with Brat in embryos (3,11). Analysis of the Brat-bound transcripts from both RIP experiments revealed a highly similar enriched sequence motif (3,11). Second, in vitro selection from a randomized RNA library by the Brat NHL domain yielded sequences that matched the RIP motifs. Third, electrophoretic mobility shift assays also demonstrated specific direct binding of the Brat NHL domain to the sequence motif (3,11,15). Fourth, molecular interactions between residues on the NHL surface and RNA ligands were observed in several X-ray crystal structures (PDB 4ZLR, PDB 5EX7) (11,17). Fifth, luciferase reporters containing BBS motif are repressed by Brat in cultured cells (3,11,12). By combining the aforementioned evidence, the BBS motif can be defined as 5′-WUGUUR (W = A/U, R = A/G). This detailed knowledge of Brat RNA-binding specificity has facilitated the identification of new Brat-regulated target mRNAs and supported the prediction of an extensive set of BBS-containing RNAs (18).

The broader impact of Brat-mediated mRNA regulation remains incompletely understood, in part due to the essential nature of Brat. Loss of function of Brat in adult females results in germline defects and lethality in resulting embryos (14,19,20). Zygotic loss of function is lethal during the larval stage due to brain tumor formation (4,21,22), which revealed Brat’s function as a tumor suppressor. Attempts to identify genes dysregulated when Brat is inactivated, depleted or mutated have proven challenging. For example, transcriptomics and proteomics analyses of brain tumors derived from brat mutants used inherently heterogeneous, abnormal tumor cells (23–25). Changes in RNA levels were measured by microarray in brat mutant embryos prior to cellularization (3); however, this approach necessitated the use of the hypomorphic bratfs1 allele rather than a protein null (1–3,11,12,15). To overcome issues with tissue heterogeneity and classical alleles, Brat was specifically reduced in Type II neuroblasts—the cells of origin for brat tumors (5)—via temporally controlled RNAi coupled with cell sorting (15,26,27). As a result, several new Brat-regulated mRNAs were identified. This targeted approach was limited by the difficulty in obtaining sufficient cellular material, which hindered replication and statistical analysis of differential gene expression.

To measure the impact of Brat on the transcriptome, and circumvent the challenges mentioned above, we utilized cultured Drosophila cells. In this rigorously controlled system, Brat can be efficiently and rapidly depleted by RNAi, and the material amounts necessary for deep sequence coverage by RNA-seq are easily achievable. We show that depletion of Brat increases the mRNA levels of several hundred genes, consistent with these genes being targets of Brat-mediated repression. We integrated our results with the published RIP datasets, revealing that the majority of genes upregulated by Brat RNAi encode Brat-bound transcripts. Furthermore, the BBS motif recognized by Brat is significantly enriched in the genes that are upregulated by Brat RNAi. Combined, these results support widespread direct repression of mRNAs by Brat. Our results also show that the location and number of BBS motifs in 3′ untranslated regions (UTRs) confer sensitivity to Brat repression, and we support these conclusions by reporter mRNA assays. Analysis of the functions of genes directly regulated by Brat reveals new roles in controlling glycolytic enzymes and the vacuolar ATPase (V-ATPase) complex, providing new insights into the action of Brat as a tumor suppressor.

Materials and methods

Drosophila stocks

Drosophila stocks used in this study included brat1 (Kyoto DGRC #107544), Df(2L)Sd57 (Kyoto DGRC #101357), Da>Gal4, tub>Gal80ts (a gift of Michael O’Connor) and UAS-Brat (15) (a gift of Jürgen Knoblich). Flies were reared at 25°C. As Brat overexpression is embryonic lethal, all crosses of Gal80ts,Da>Brat flies, and their controls, were maintained at 18°C and shifted to 29°C for 24 h to inactivate Gal80ts.

To characterize the brat1 mutation, genomic DNA was extracted from brat1/Df(2L)Sd57 larvae by homogenizing larvae in Squish Buffer (10 mM Tris–Cl pH 8.0, 25 mM NaCl, 1 mM EDTA). Lysates were treated with 10 μg NEB Proteinase K for 30 min, and the proteinase was inactivated at 95°C for 2 min. The region of the brat coding sequence was amplified from this genomic DNA via PCR using primers RC357 and RC351 (Supplementary File S1), purified and the missense mutation Q782* was identified by Sanger DNA sequencing.

Plasmids and cloning

All plasmids used in this study and the sequences of relevant primers are listed in Supplementary File S1. Reporter genes were based on the vector pAc5 (Invitrogen) as previously described (28). For reporters with 3′UTR BBS motifs, oligos corresponding to 2 × BBS (RC036/037), 4xBBS (RC038/039), 6xBBS (RC040/041) or 8xBBS (5′UTR RC042/043, 3′UTR RC044/045) were inserted between NotI and XhoI sites of pAC5.4 Rluc. Reporters with either wild-type (5′UTR 4xBBS, RC504/505) or mutated BBS (5′UTR 4xMUT, RC506/507) in the 5′UTR were similarly cloned into the KpnI site of pAc5.4 Rluc or pAc5.4 Rluc 8xBBS. The 3′UTRs of VhaAC39-1 (RC726/727), VhaM8.9 (RC728/729), Vha100-2 (RC722/723) and VhaPPA1-1 (RC724/725) were amplified from Drosophila genomic DNA (purified from the DL1 cell line) with the denoted primers and inserted between the NotI and XhoI sites of pAC5.4 Rluc. Fragments encoding the common region of the Tpi and Treh 3′UTRs were chemically synthesized (Twist Bioscience) and inserted into the XhoI–NotI fragment of pAc5.4 Rluc via Gibson Assembly.

For transient expression of Brat, the coding sequence of isoform A (NP_476945.1) (29) was subcloned into pIZ V5 H6 (Thermo Fisher Scientific) such that the C-terminus was fused to V5 and His6 epitope tags. This coding sequence was expressed under the control of the constitutive OpIE-2 promoter and polyadenylation sequence. For induced expression of Brat, the epitope-tagged coding sequence was subcloned into pMT V5 H6, in which the copper-responsive metallothionein promoter stimulates expression.

Guide RNAs (gRNAs) targeting the N-terminus of Brat or Vha100-2 were identified via flyCRISPR’s Optimal Target Finder tool (30). Oligos corresponding to the sequences of brat_gRNA 1 (RC396 and RC397), brat_gRNA 2 (RC398 and RC399) and Vha100-2_gRNA (SY07 and SY08) were inserted into the BspQI-digested pAC sgRNA Cas9.

Cell culture and luciferase assays

Drosophila d.mel-2 cells (Invitrogen) were cultured at 25°C in Sf900III (Gibco) supplemented with antibiotics (25 U/ml penicillin and 25 μg/ml streptomycin) (Thermo Fisher Scientific). Luciferase assays were performed in white, flat-bottom Nunclon Delta Treated 96-well plates (Fisher). For assays utilizing RNAi, cells were seeded at 5 × 104 per well over 1.2 μg dsRNA and grown for 1 day before transfection. For assays that did not utilize RNAi, cells were seeded at 1 × 105 per well and transfected immediately. In all cases, cells were transfected with pAC5.1 Fluc (0.5 ng), pAC5.4 Rluc plasmids (1 ng), pIZ Brat (50 ng), pIZ vector (Invitrogen) to a total of 100 ng DNA per well, Sf900III media to 4.8 μl and 0.2 μl FuGENE HD (Promega). Six wells per condition were transfected with a master mix. After 2 days, luciferase assays were performed on three wells using the Dual-Luciferase Reporter Assay System (Promega) following the manufacturer’s instructions. Cells from the remaining three wells were pooled, and total protein was harvested for western blot analysis (see below).

Drosophila DL1 cells (S1 cells, obtained from Eric Wagner) (31) were cultured at 25°C in Schneider’s Drosophila Media (SDM, Thermo Fisher) supplemented with 1× antibiotic-antimycotic (Thermo Fisher), 1% (v/v) GlutaMax (Fisher) and 10% (v/v) heat-inactivated fetal bovine serum (Genesee). For luciferase assays, 2 × 106 cells were seeded in six-well plates and transfected with pAC5.1 Fluc (5 ng), pAC5.4 Rluc plasmids (5 ng), pMT Brat (1 μg), pIZ vector/pIZ EGFP to a total of 3 μg DNA, culture media lacking fetal bovine serum to 144 μl and 6 μl FuGene. After 1 day, Brat expression was induced with 500 μM CuSO4. After a further 24 h, cells were resuspended and 100 μl of cells were aliquoted in triplicate into 96-well plates.

Luciferase assays were conducted as described (32). Unless otherwise noted, luciferase assays in d.mel-2 cells were conducted with three biological replicates and three independent experiments. To control for potential variations in transfection efficiency, the luminescence output of the Renilla luciferase reporter in each sample was normalized by dividing by the firefly luciferase output in the same sample. Next, the log2 fold change of each sample in the test conditions was calculated relative to the mean value of the relevant control condition using Equation (1). Specifically, for biological replicate i and experimental replicate j,

graphic file with name M0001.gif (1)

The resulting numeric dependent variable (log2 fold change) and corresponding categorical independent variable (‘condition’) are presented in Supplementary File S2 and fit to a linear model with mixed effects using the R lme4 package. This preserved variation from both biological and experimental sources. Statistical significance of differences between groups was determined via ANOVA (anova() from lmerTest) with a post-hoc Tukey–Kramer test (glht() from multcomp). Alternatively, supporting assays without experimental replicates were fit to a general linear model without mixed effects (using lm() from stats), and differences between groups determined via one-way ANOVA (anova() from stats) with post-hoc Tukey–Kramer (glht() from multcomp). The resulting P-values are two-sided.

Luciferase assays utilizing RNAi of endogenous Brat were conducted with three biological replicates (each well of a 96-well plate). Statistical significance of differences between groups was determined via Welch’s ANOVA (oneway.test() from stats) with post-hoc Games–Howell test (games_howell_test() from rstatix). The resulting P-values are two-sided.

Luciferase assays in DL1 cells were conducted with three biological replicates (each well of a six-well plate) and three technical replicates. The resulting luminescence values were analyzed in the same manner as d.mel-2 assays, except considering technical replicate i and biological replicate j. Statistical significance of differences between groups was determined via Welch’s ANOVA (oneway.test() from stats) with post-hoc Games–Howell test (games_howell_test() from rstatix). The resulting P-values are two-sided.

Western blot analysis

Total protein was collected from transfected and dsRNA-treated cells for western blot analysis. Cells were harvested by centrifugation at 1000 × g for 3 min, and disrupted in 40 μl of lysis buffer [50 mM Tric-Cl, 150 mM NaCl and 0.5% NP-40 supplemented with protease inhibitors (Roche)] using a handheld homogenizer, on ice, for 30 s. Insoluble material was then pelleted at 21 000 × g for 10 min at 4°C. Total protein concentration was determined via DC Protein Assay (Bio-Rad) and western blotting was conducted on the supernatants as described (28). Primary antibodies used in this study are indicated in Supplementary File S1. Chemiluminescent images were obtained using a ChemiDoc Touch Imaging System (Bio-Rad) or Azure 300 (Azure Biosystems) and resulting TIFF files were generated using Image Lab 5.2.1 software (Bio-Rad). Quantitative western blotting confirmed the linear range of detection for Flag::Vha100-2 and Tubulin. Densitometry was conducted via AzureSpot Pro (Azure Biosystems). The line of best fit for FLAG and Tubulin signals was determined via the least-squares method in GraphPad Prism. For analysis of Vha100-2 expression, the FLAG signal was normalized to that of Tubulin (Supplementary File S2) and statistical differences between RNAi conditions were determined via one-way ANOVA with post-hoc Tukey–Kramer test. For measurement of Brat-mediated repression of Flag::Vha100-2, three biological replicates were performed per RNAi condition.

Epitope tagging of endogenous genes

The endogenous brat locus of DL1 cells was amplified from genomic DNA via PCR using primers RC357 and RC359 (Supplementary File S1), purified and any potential SNPs identified via Sanger sequencing. Similarly, the endogenous Vha100-2 locus was amplified from DL1 genomic DNA using primers RC770 and RC771 to identify potential SNPs.

For all CRISPR knock-ins, 3 × 106 DL1 cells were seeded into each well of a six-well plate. Cells were transfected with 1 μg of the relevant pAC sgRNA Cas9, 40 pmol of single-stranded oligo donor template (brat_ssODT or Vha100-2_ssODT), serum-free media to 150 μl and 4 μl FuGENE HD (Promega). After 2 days, 5 μg/ml puromycin was added to the media. Selection for the puromycin-resistant Cas9/gRNA plasmid was maintained, with media changes as necessary, until paired untransfected cells were no longer viable or proliferative. Polyclonal DL1 lines were used for further analysis, as indicated. Monoclonal DL1 lines were generated via limiting dilution. Homozygosity of selected monoclonal lines was verified via PCR amplification of genomic DNA. Samples of polyclonal and monoclonal lines were maintained in liquid nitrogen indefinitely in 90% culture media supplemented with 10% Dimethyl sulfoxide (DMSO).

RNAi depletion and RNA-seq

RNAi

Endogenous Brat was depleted by RNAi using two dsRNAs targeting different regions of the Brat coding sequence, as described previously (33). DsRNA #1 was amplified with 5′-GGATCCTAATACGACTCACTATAGGGCAGATCTTCGACAAGGAGGGACG and 5′-CATACCCACTGGCGCCAGTTGG. dsRNA #2 was amplified with 5′-CAACGAGCTGAACGAGACGCACC and 5′-GGTGTGACTGTTGGTGGTGGCC. In all cases, the T7 promoter (5′-GGATCCTAATACGACTCACTATAGGG) was appended to the 5′ end of all dsRNA primers. dsRNA corresponding to the Escherichia colilacZ gene was used as an NTC. Twenty million V5::Brat DL1 cells were resuspended in 1 ml of serum-free SDM in each well of a six-well plate, and incubated with 20 μg of the indicated dsRNA for approximately 1 h. Afterward, 2 ml of SDM was added, and cells were allowed to grow for 2 days. At this time, cell density ranged from 2.20 to 4.86 × 106 cells/ml with viability of 90–99% as determined via trypan blue staining using a Countess III cell counter (Invitrogen). For CLICK-seq, four biological replicates were prepared for each RNAi condition.

RNA isolation

For each sample, a 2 ml sample of resuspended cells was pelleted at 1000 × g for 3 min, and washed with 1 ml of ice-cold 1× phosphate-buffered saline (PBS). RNA was harvested using a Maxwell RSC with SimplyRNA Tissue kit (Promega) following the manufacturer’s instructions with the following exceptions: 10 μl of DNase I was used per cartridge, total RNA was eluted in 75 μl of water and residual beads were pelleted several times to ensure depletion. The concentration of total RNA was determined via UV absorbance with a Nanodrop One (Thermo). The quality of isolated RNA was determined via Bioanalyzer. All RNA samples produced RIN values greater than 9.8.

CLICK-seq

First, 5 μg total RNA was subjected to poly(A) mRNA isolation using a NEBNext Poly(A) mRNA Magnetic Isolation Module (New England Biolabs). Then, 10 μl of the eluted mRNA was mixed with 1 μl of 100 μM Genomic Adaptor_6N primer and 2 μl of 5 mM AzNTP/dNTP mixture in a 1:35 ratio. The mixture was then heated at 65°C for 5 min and snap-cooled on ice for 3 min. Then, the reverse transcription was carried out using SuperScript III: In brief, 7 μl master mix containing 4 μl 5× Superscript First Strand Buffer, 1 μl 0.1 M DTT, 1 μl RNaseOUT and 1 μl Superscript III was mixed with the RNA sample. Heating was performed as follows: 25°C for 10 min, 50°C for 40 min and 75°C for 15 min, sequentially, in a thermocycler. This was followed by RNase H treatment using 1 U per reaction for 37°C for 30 min and then 80°C for 10 min. The complementary DNA (cDNA) was then purified using Sera-Mag Speedbeads. Speedbeads working solution was made by washing 1 ml beads slurry twice in 1XTE buffer and then resuspending in 50 ml of 1XTE buffer containing 9 g PEG-8000, 1 M NaCl and 0.05% Tween 20. A 1.8× reaction volume of Speedbeads was mixed with cDNA, followed by a 5 min incubation at room temperature. The beads were then collected using a magnetic bead collector, and the supernatant was discarded. Two washes with 80% ethanol were performed while the beads were retained on the magnet. The beads were then dried, and the cDNA was eluted by resuspending the beads in 22 μl 50 mM HEPES (pH 7.2) for 2 min at room temperature. Then, 20 μl of the cDNA was mixed with 11 μl Click Mix and 4 μl of 5 μM UMI-click-adapter (HPLC purified). To activate the reaction, 4 μl of 50 mM vitamin C and 1 μl of Click Catalyst were pre-mixed, added to the cDNA solution and incubated in the dark for 60 min at room temperature.

The ‘Clicked’ cDNA was then purified in the same manner as post-RT purification except eluted in 22 μl of 10 mM Tris–Cl (pH 7.4). Next, half of the eluted cDNA was subjected to PCR amplification with indexing primers i5 and i7. Both primers have a phosphorothioate bond added at the 3′ end to increase stability. Amplification was carried out in the volume of 50 μl including 25 μl 2× OneTaq Master Mix (New England Biolabs), 2 μl of 5 μM i5 and i7 each, cDNA and water. This reaction was subject to the following PCR protocol: 94°C for 4 min, 53°C for 30 s, 68°C for 10 min, 15 cycles of (i) 94°C for 30 s, (ii) 53°C for 30 s and (iii) 68°C for 2 min, followed by 68°C for 5 min. The library was then purified and 200–600 nt products were size-selected with Speedbeads. In brief, 45 μl beads (0.9× volume of PCR reaction) were mixed with the reaction to remove larger fragments. The supernatant was collected and mixed with 10 μl beads (0.2× volume of PCR reaction) to remove smaller fragments and primer dimers. Two washes with 100% ethanol were performed while the beads were on the magnet. Beads were dried before resuspended in 12 μl of 10 mM Tris–Cl (pH 7.4). The library was then collected and subjected to quality control on Agilent 2200 TapeStation and to analysis on Illumina NovaSeq 6000 at Genomics Research Core at the University of Rochester. Sequencing was performed at the NIEHS Epigenomics and DNA Sequencing Core Facility, with the aid of the Integrated Bioinformatics Group. The NovaSeq 6000 NGS system produced >38 000 reads (100 nt single-end) per sample.

Reverse transcription and quantitative PCR

RNA measurements of Brat-regulated genes were additionally conducted via RT-qPCR. For DL1 cells, purified total RNA for RNA-seq was stored at –80°C prior to RT-qPCR. RNA levels following Brat depletion by RNAi were compared to RNA levels of NTC RNAi. The significance of differences between mRNA levels (normalized to 7SK) was determined via ANOVA (anova() from stats) with post-hoc Tukey–Kramer tests (glht() from multcomp). Similar RNA measurements were also conducted via RT-qPCR with RNA isolated from Drosophila larvae. RNA was isolated using a Maxwell RSC with SimplyRNA Tissue kit (Promega) as described above. Larvae of the appropriately indicated genotype were selected, washed in cold 1× PBS and homogenized in Homogenization Buffer (Promega) on ice. For transient ubiquitous expression of Brat, tub>Gal80ts Da>Brat larvae were incubated at 29°C for 24 h. Due to the lethality of ubiquitous Brat expression, two non-melanized wandering third instar larvae composed a single biological replicate. RNA was similarly isolated from these larvae, in triplicate. Purified total RNA was stored at −80°C prior to RT-qPCR. RT-qPCR was conducted as described. The significance of differences between normalized mRNA levels was determined via Welch’s two-sample t-test (t.test() from stats).

All parameters for RT-qPCR, including primer sequences, amplification efficiencies and amplicon sizes, are reported in Supplementary File S1. RT-qPCR was conducted in accordance with MIQE guidelines (Supplementary File S6) (34). Primers were identified using Primer-BLAST to assess specificity (35). LunaScript RT (New England Biolabs) was used to reverse transcribe 1 μg of total RNA, following the manufacturer’s instructions and using 25 ng/μl random hexamer oligonucleotides (IDT) per reaction. As necessary, cDNA was stored at −80°C. Quantitative PCR was performed on an estimated 50 ng of resulting cDNA using Luna qPCR Master Mix (New England Biolabs) in 20 μl total reaction volume with 250 nM oligos. The following qPCR cycles were performed using a CFX96 Real-Time PCR System thermocycler (Bio-Rad): following an initial denaturation at 95°C for 2 min, amplification occurred during 40 cycles of denaturation (95°C for 15 s) and annealing/extension (60°C for 1 min) with fluorescence measurements taken each cycle. The resulting Cq values were analyzed via CFX Manager 3.1 (Bio-Rad) and reported in Supplementary File S2. RNA levels of indicated genes were calculated relative to the level of 7SK using the ΔCT method.

Transcriptome-wide identification of Brat motifs

BBS motifs were identified across the transcriptome using a bespoke Python script, Count-dmel-sites, found at the author’s public GitHub repository (github.com/chaseaw). Briefly, the full sequences of annotated Drosophila 5′UTR, CDS and 3′UTRs (assembly dmel-r6.43) were downloaded from FlyBase and all instances of the BBS consensus motif (k-mers 5′-ATGTTA, ATGTTG, TTGTTA and TTGTTG) were identified (3,11,18). The presence, number, average and maximum number of identified BBS, and positions of identified BBS, are presented in Supplementary File S3. These data were used to bin expressed genes into BBS-containing gene sets.

Analysis of RNA-seq data

Identifying regulated genes

RNA-seq was conducted in accordance with ENCODE guidelines (Supplementary File S5). Reads were aligned to the Drosophila r6.49 reference genome and transcript models using STAR v 20201. The enrichment of polyadenylated RNAs was evident by the low percentage of reads mapped (using bowtie2) to Drosophila rDNA sequences As expected, the expression of genes among biological replicates was highly correlated, with Spearman rho correlation coefficients >0.92. Assessed for a subset of highly expressed genes, empiric coverage correlated with transcript abundance (Spearman rho 0.6488949). The summary statistics and underlying data are presented in Supplementary File S2. Differential gene expression was determined using DESeq2 (36). The complete set of results is reported in Supplementary File S4. Only genes expressed above baseMean ≥ 50 in Brat and NTC RNAi conditions were considered for differential gene expression analysis (7516 genes, excluding brat, hereafter designated ‘expressed’). Only genes with a Benjamini–Hochberg adjusted P-value <0.05 were considered differentially regulated in Brat RNAi relative to NTC RNAi and upregulated if log2 fold change >0 or downregulated if log2 fold change <0. Genes ‘upregulated’ following Brat RNAi were defined as the 158 genes significantly upregulated by both Brat dsRNA RNAi conditions. The correlation of log2 fold change measurements in each RNAi condition was compared via a Spearman rho test. The complete results of all statistical tests are reported in Supplementary File S2.

Previously, genes were identified that were significantly upregulated in bratfs1/− embryos, relative to wild-type embryos, at various time points (3). When comparing those genes to the current Brat RNAi results, only genes expressed in DL1 cells and embryos with unambiguous, updated FlyBase Gene IDs (FBgns) were used for comparisons. Lists of the resulting genes are presented in Supplementary File S4.

Assignment by motifs and binding

Previously, genes bound by Brat were identified using RNA co-immunoprecipitation analyses followed by microarray (so-called RIP-Chip) (3,11); however, the two studies used different metrics for classifying bound versus unbound. To directly compare these results, we applied the same stringent criteria to both RIP-Chip datasets: only genes enriched >1.5-fold relative to the respective negative control RIP with false discovery rate (FDR; or q-value) <0.05 were considered. These genes were converted to the most current FBgn, and only unambiguously assigned FBgn were considered. This resulted in 1155 and 3021 unique FBgn bound by Brat in Laver (3) and Loedige (11), respectively. The resulting lists of genes bound by Brat are presented in Supplementary File S3. Genes upregulated by Brat RNAi were then binned by binding and BBS predictions, as reported in Supplementary File S4. Among all genes expressed in DL1 cells, the significance of overlap between upregulated genes and other subsets (genes with 5′UTR BBS, genes bound in RIPs, etc.) was determined via Fisher’s exact test (37,38) (fisher.test() from stats). Common expressed genes were used to calculate category totals. The resulting P-values are two-sided. Overlap was visualized using the R VennDiagram package.

Similarly, the percentages of Brat-bound mRNAs that contain BBS were identified using Laver (3) and Loedige (11) RIP-Chip datasets. The overlap between the 3106 unique, unambiguously assigned genes bound by Brat in either RIP and genes with predicted 5′UTR (3547), CDS (6457) and 3′UTR (6265) BBS was determined. The significance of this overlap was determined via Fisher’s exact test as described, with all protein-coding genes (13 967) used to calculate category totals.

Analysis of z-scores

The significance-weighted fold change (known as ‘z-score’) was defined as the log2 fold change divided by its standard error. This is equal to the ‘stat’ output of DESeq2 (36) (Supplementary File S4), and equivalent to the ‘Wald statistic’ (39,40). Genes were binned by a categorical independent variable (binding, presence of 3′UTR Brat motifs, etc.), and the numerical z-scores of the two categories were compared via Wilcoxon t-test. Additionally, z-scores for all expressed genes were binned by the number of 3′UTR BBS (as the categories 0, 1 and ≥2 BBS). The significance of differences was determined via Kruskal–Wallis rank sum tests (kruskal.test() from stats) and pairwise Wilcoxon rank sum tests with Holm–Bonferroni corrected P-values (pairwise.wilcox.test() from stats). The resulting P-values are two-sided.

Identification of enriched motifs

Novel enriched motifs were identified among upregulated genes using MEME (41,42). The sequences of 3′UTRs were downloaded from the FlyBase Chado database, and the corresponding FBgn was assigned. The longest 3′UTRs of upregulated genes (or relevant bound subsets) were searched for motifs in a 6–8-nucleotide window with default conditions, except only the given strand was searched and motifs could occur with any number of repetitions. The resulting motifs were visualized using the R ggseqlogo package.

Comparing 3′UTR lengths

Genes bound by Brat were assigned either the longest or shortest 3′UTR sequences (as indicated), and distributions of lengths were compared via the Wilcoxon rank sum test. Additionally, genes were binned by binding in RIPs and the distributions of minimum and maximum 3′UTR motifs were displayed by boxplots.

Gene ontology enrichment

Gene ontology (GO) enrichment was conducted using the Gene Ontology Resource (43,44) with common expressed genes as a reference list. The significance of enrichment was determined via Fisher’s exact test with a Benjamini–Hochberg FDR. The results are presented in Supplementary File S2 and summarized in Table 1.

Table 1.

GO term enrichment of upregulated genes

GO term Genes in reference list Upregulated genes Expected genes Fold enrichment P-value FDR
Glycogen biosynthetic process 6 5 0.13 38.53 1.66e−06 2.20e−04
Gluconeogenesis 6 4 0.13 30.82 3.68e−05 3.43e−03
Glycolytic process 15 9 0.32 27.74 6.77e−10 4.93e−07
Pentose-phosphate shunt 6 3 0.13 23.12 7.12e−04 4.89e−02
Vacuolar acidification 9 4 0.19 20.55 1.19e−04 9.75e−03
Proton transmembrane transport 35 13 0.76 17.17 1.06e−11 2.58e−08
Cellular carbohydrate catabolic process 19 5 0.41 12.17 1.22e−04 9.92e−03
Glucose homeostasis 32 8 0.69 11.56 1.46e−06 2.09e−04
Carbohydrate homeostasis 38 9 0.82 10.95 4.65e−07 8.07e−05
Terminal branching, open tracheal system 26 5 0.56 8.89 4.34e−04 3.04e−02
Aminoglycan metabolic process 29 5 0.63 7.97 6.76e−04 4.69e−02
Proton-transporting ATPase activity, rotational mechanism 16 12 0.35 34.68 1.21e−13 3.17e−10
Imaginal disc growth factor receptor binding 4 3 0.09 34.68 3.06e−04 3.83e−02
Plasma membrane proton-transporting V-type ATPase complex 14 12 0.3 40 3.98e−14 1.70e−11
Vacuolar proton-transporting V-type ATPase,
V1 domain 8 6 0.17 35 2.14e−07 2.74e−05
Vacuolar proton-transporting V-type ATPase, V0 domain 7 5 0.15 33 2.80e−06 2.76e−04
M band 13 5 0.28 18 2.74e−05 1.84e−03

Note: GO terms enriched among genes upregulated following Brat RNAi. GO terms utilized the GO aspects Biological Process, Molecular Function and Cell Component from the Gene Ontology Resource. Given the number of genes for a GO term expressed in DL1 cells (genes in reference list), the expected number of upregulated genes matching this GO term were determined (expected number of genes). Fold enrichment was determined by comparing this expected number to the actual number of upregulated genes in this category. The significance of this enrichment was determined as described in the ‘Materials and methods’ section.

Metabolite profiling

Sample processing

RNAi was conducted in V5::Brat DL1 cells as described above. One set of cells was grown for 3 days in SDM as described. Another set was grown in equivalent media without supplemental GlutaMax for the final 24 h. After 3 days of RNAi, 2 ml of resuspended cells were harvested by centrifugation at 1000 × g for 3 min. Cell pellets were then washed with 750 μl 1× PBS, flash-frozen in liquid nitrogen and stored at −80°C before processing. From the remaining cells, viability was determined and cell lysates were prepared for western blotting as described. RNAi was conducted in four independent experiments for both growth conditions, and metabolite measurements were taken in triplicate.

Frozen cell pellets were resuspended in 100 μl ice-cold water and vortexed thoroughly. A 10 μl aliquot of each sample was used to determine protein concentration. To the remaining sample, 400 μl of ice-cold methanol was added, incubated on ice for 10 min and debris pelleted at 21 300 × g for 15 min at 4°C. Of the resulting supernatant, 100 μl was transferred to v-bottom, deactivated glass vials for mass spectrometry.

Data acquisition

Metabolites were analyzed using a Q-Exactive Plus mass spectrometer (Thermo). Metabolites were separated via HILIC chromatography, similar to the neutral pH ZIC-cHILIC method (45). Briefly, separations were performed using a metal-free ZIC-cHILIC PEEK 2.1 mm × 100 mm, 3 μm particle size, 100 Å pore size analytical column (SeQuant) in conjunction with a metal-free ZIC-cHILIC PEEK 2.1 mm × 20 mm, 3 μm, 100 Å guard column (SeQuant) on a Vanquish ultra-high-performance liquid chromatography (UHPLC) system (Thermo). The mobile phase consisted of solvent A (90% acetonitrile, 10% 5 mM ammonium formate pH 6.7) and solvent B (10% acetonitrile, 90% 5 mM ammonium formate pH 6.7), in a gradient elution as follows: 0% B for 0–2 min, 0–40% B for 2–12 min, 40–100% B for 12–13 min, 100% B for 13–15 min, 100–0% B for 15–15.1 min and 0% B for 15.1–2 min. The flow rate was set to 0.4 ml per min. The column temperature was maintained at 40°C.

Electrospray ionization was used in negative ion mode. Full MS scans were acquired over the m/z range of 80–900 with a resolution of 70 000, an automatic gain control (AGC) target of 3 × 106 ions and a maximum injection time of 200 ms. Additional narrow-range MS scans (with the same parameters) or parallel reaction monitoring (PRM) were performed as segments for each of the metabolites of interest (described in Supplementary File S1). The parameters for PRMs were as follows: a resolution of 17 500, AGC target of 2 × 105 ions, maximum injection time of 100 ms, isolation windows of 2 m/z and a stepped normalized collision energy of 15, 35 and 60. Injections of 8 μl were performed for all metabolites except pyruvate and lactate, for which 1 μl injections were performed to ensure the mass spectrometer response remained in the linear range.

Data processing

Raw data files were processed using the Qual Browser function of the Xcalibur software (Thermo). Metabolites were identified by comparing their actual mass and retention time to metabolite standards that were analyzed using the exact same method and buffers (Supplementary File S1). Relative metabolite abundance was quantified using the area under the curve (AUC) of extracted ion chromatograms (EICs) from the appropriate narrow-range MS1 channel for the exact m/z of the single charged species for lactate, pyruvate and acetyl-CoA. For glucose-6-phosphate, peak areas of EICs of fragment ions (m/z 96.965–96.971) were used. Metabolite levels were normalized to the total protein content of the samples, which was determined using the Protein Assay (Bio-Rad) microplate procedure.

For each metabolite, the resulting numeric dependent variable (protein-normalized AUC) was directly compared to the corresponding categorical independent variable (‘condition’) and fit to a linear model with mixed effects using the R lme4 package. To preserve variation from biological and technical sources, replicate (‘rep’) and mass-spectrometry run (‘Trep’) were considered random effects. Statistical significance of differences between groups was determined via ANOVA (anova() from lmerTest) with a post-hoc Tukey–Kramer test (glht() from multcomp). Relevant comparisons were verified using mixed models that considered only biological variation and simple linear models without random effects. The AUC, normalization factors, final values and all results of these analyses are presented in Supplementary File S2.

Results

Knockdown of brat expression identifies upregulated genes

To measure the impact of Brat on the transcriptome, we performed RNAi to deplete Brat and RNA-seq to measure the resulting changes in transcript levels. As an antibody was not available to confirm the depletion of Brat protein, we first generated embryo-derived DL1 cells (31) in which all copies of the brat gene were tagged with sequences encoding V5 epitope (and His6 tag) at the N-terminus using CRISPR–Cas9 gene editing (Supplementary Figure S1A). As with other TRIM-NHL proteins (Figure 1A), the RNA-binding NHL domain of Brat is at the C-terminus (Figure 1B). The N-terminus was chosen for epitope tagging, as a similar approach did not disrupt in vivo function (15). After clonal isolation, a homozygous V5-Brat DL1 line was verified by PCR genotyping (Supplementary Figure S1B) and western blotting (Figure 1C). To measure changes in the transcriptome upon Brat depletion, we took advantage of the ability of cultured Drosophila cells to process dsRNAs into siRNAs to efficiently and specifically knockdown target gene expression (46,47). To control for unintended off-target effects of individual dsRNAs, Brat was depleted from V5-Brat DL1 cells using two dsRNAs targeting different regions of the brat mRNA, each with four biological replicate samples (Figure 1C). Cells treated with a non-targeting control dsRNA (NTC, corresponding to E. coli LacZ) served as the negative control. To determine transcriptome-wide changes upon Brat depletion, the polyadenylated RNA from these samples was purified and sequenced using CLICK-based RNA-seq (48).

Depletion of Brat significantly (Padj < 0.05) upregulated 222–223 genes, and downregulated 66–88 genes, depending on the Brat dsRNA used (Figure 1D, Supplementary Figure S1C and Supplementary File S4). Based on evidence that Brat functions as a repressor (1), we focused on genes in which the transcript levels were increased following depletion of Brat. Considerations for the 28 genes downregulated in both Brat RNAi conditions (Supplementary Figure S1D) are addressed in the ‘Discussion’ section. Comparison of the two Brat RNAi datasets revealed an intersection of 158 upregulated genes (here-on designated ‘upregulated,’ P-value of overlap <2.2 × 10−16) (Figure 1E and Supplementary File S4). The fold change of these genes correlated well between both RNAi conditions (Spearman rho 0.91) (Supplementary Figure S1E). The fold changes of the upregulated genes span 9–12-fold, with medians of 1.54–1.55 (Figure 1F), consistent with the magnitudes previously observed for several Brat target mRNAs (11,14). We note that a DESeq2 baseMean cutoff of 50 was applied to the datasets to filter genes with very low expression (Supplementary Figure S1F), which had minimal effect on the number of genes with significant differential expression (Supplementary Figure S1G). The log2 fold changes of upregulated genes also correlate well (Spearman rho > 0.8) when log2 fold change thresholds are imposed (Supplementary Figure S1H). We performed RT-qPCR on total cellular RNA to independently confirm significant upregulation of five genes by Brat RNAi (Figure 1G), including subunits of the vacuolar ATPase complex (VhaPPA1-1, Vha100-2 and VhaM8.9), the metabolic enzyme Trehalase (Treh) and an E3 ubiquitin ligase involved in Notch signaling (Su(dx)). Additionally, a comparison of the genes upregulated by Brat RNAi in DL1 cells showed statistically significant overlap with 12–33 genes that were upregulated in bratfs1/Df embryos over the course of MZT (Supplementary Figure S2A) (3), providing support that our model system recapitulates Brat regulation in its native in vivo context.

mRNAs upregulated by Brat knockdown contain 3′UTR BBS

We sought to identify the direct target mRNAs of Brat by integrating our upregulated Brat RNAi dataset with transcripts that were reported to be bound by Brat and that contain the BBS motif. Two published RIP studies identified mRNAs that are bound by Brat in embryos (Supplementary File S3) (3,11), of which 541 genes overlap (P-value < 2.2e−16) (Figure 2A). As 89% (6293/7032) of the genes expressed in DL1 cells are also expressed in the embryo (39), we compared these bound genes to those regulated by Brat. We observed a significant overlap between these Brat-bound transcripts and the genes upregulated by Brat RNAi (Figure 2A and B). Of the upregulated genes, 28% (45/158) are bound in both RIPs and 59% (94/158) are bound in at least one of these RIPs (Figure 2B). To determine whether levels of Brat-bound transcripts generally increase when Brat is depleted, we compared the log2fold change in expression upon Brat depletion between bound and unbound genes. As log2fold change measurements have inherent variability, we weighted the log2fold change by its standard error following (36). Transcripts bound by Brat were significantly (P-value < 2.2e−16) more upregulated by Brat RNAi than unbound transcripts (Figure 2C and D). This functional relationship was observed for both Brat RNAi conditions and the Brat-bound transcripts in either RIP datasets (Supplementary Figure S2B and C). Therefore, the experimental evidence is consistent with Brat-mediated repression of mRNA levels being a consequence of Brat binding.

Figure 2.

Figure 2.

A significant number of Brat-repressed genes are bound by Brat. (A) Venn diagram of genes upregulated by Brat RNAi and those identified in published Brat RNA-immunoprecipitations studies (RIPs). In all cases, Brat-associated genes were limited to those expressed in DL1 cells. (B) Fractions of upregulated genes that are bound by Brat in embryo extracts reported in RIP experiments by Loedige et al. (11) or Laver et al. (3). Genes were additionally considered ‘bound’ by Brat if identified in either RIP, and a smaller subset was identified in both RIPs. Significance of the overlaps between upregulated genes and these bound subsets were determined via Fisher’s exact tests (***P< 0.001). The number of genes per subset is indicated in each column. Kernel density estimate showing distribution of significance-weighted log2 fold changes from (C) NTC versus dsRNA #1 and (D) NTC versus dsRNA #2 for genes either bound in Brat RIPs or that were not bound. Median significance-weighted log2 fold changes for bound (solid) and unbound (dashed) genes displayed as vertical lines. Significance of differences between distributions was determined via a Wilcoxon rank sum test (***P< 0.001). Number of genes per subset in key.

We then assessed the importance of the BBS motif recognized by the Brat NHL domain, 5′-WUGUUR (W = A/U, R = A/G) (3,11) to our observed Brat-dependent regulation. First, we explored the transcriptome-wide distribution of BBS motifs and their locations within transcripts (Supplementary File S3). We observed that BBS motifs are most prevalent in the 3′UTRs of Brat-bound transcripts, compared to 5′UTRs and coding sequences (Figure 3A). We also observed that genes upregulated by Brat RNAi are significantly enriched with 3′UTRs BBS motifs, with 78% (124/158) of genes containing at least one BBS motif in their 3′UTRs (Figure 3B). Among the upregulated Brat-bound mRNAs, 85% (68/80) to 93% (42/45) have 3′UTR BBS motifs (Figure 3C). The number of genes in these subsets is presented in Supplementary Figure S3A. Interestingly, while transcripts bound by Brat were also significantly more likely to contain BBS motifs in their 5′UTRs and coding sequences (Figure 3A), BBS motifs appear in the 5′UTRs and CDS of Brat RNAi upregulated genes no more frequently than expected by chance (Figure 3B). On a transcriptome-wide level, genes with at least one 3′UTR BBS are significantly more upregulated in both Brat RNAi datasets compared to those without motifs (Figure 3D and E), indicating that 3′UTR BBS motifs confer Brat-mediated repression.

Figure 3.

Figure 3.

Brat motifs are enriched in the 3′UTRs of upregulated genes. (A) Fraction of genes bound in either Loedige et al. (11) or Laver et al. (3) Brat RIP experiments with at least one putative Brat binding site (BBS) motif in the 5′UTR, coding sequence (CDS) or 3′UTR. (B) The fraction of upregulated genes with at least one putative BBS motif in each mRNA feature. In panels (A) and (B), significance of the overlap between genes containing 5′UTR, CDS or 3′UTR motifs and upregulated genes was determined via Fisher’s exact test (***P< 0.001). The number of genes per subset is indicated in the columns. (C) Upregulated genes bound by Brat in published RIPs were analyzed for BBS content and location. The percentage of these Brat bound subsets that contain at least one putative Brat motif in each mRNA feature is displayed. Over-representation was determined via Fisher’s exact test (***P< 0.001). The number of genes per subset is reported in Supplementary Figure S3A. Kernel density estimate showing distribution of significance-weighted log2 fold changes from Brat dsRNA #1 (D) and Brat dsRNA #2 (E) for genes that contain 3′UTR BBS motifs versus those that do not. Median values for genes with motifs (solid) and without (dashed) displayed as vertical lines. Significance of differences between the distributions were tested via Wilcoxon rank sum test (***P< 0.001). (F) The 5′-UUGUUD motif was significantly enriched in the 3′UTRs of genes that were upregulated by Brat RNAi (E-value 6.1 × 10−21), identified using the MEME algorithm. (G) The fraction of upregulated genes with at least one putative 5′-UUGUUD motif in each mRNA feature. (H) The fraction of upregulated genes, grouped by binding in published RIP datasets, that contain at least one 5′-UUGUUD motif in the 3′UTR. In panels (A)–(C), (G) and (H), significance of the overlap between genes containing 5′UTR, CDS or 3′UTR motifs and upregulated genes was determined via Fisher’s exact test (***P< 0.001). The number of genes per subset is reported in each column.

As an independent, unbiased approach, we performed de novo motif enrichment analysis and found the motif 5′-UUGUUD (E-value 6.1e−21) enriched in the 3′UTR of transcripts upregulated by Brat RNAi (Figure 3F). The full results from this motif enrichment are presented in Supplementary File S2. This functionally defined Brat motif is nearly identical to the BBS defined by Brat:RNA binding studies (3,11,18). For example, the BBS motifs enriched in Brat-bound mRNAs from the two Brat RIP datasets are nearly identical to our Brat motif (Supplementary Figure S3B). Additionally, the biochemically defined BBS motifs derived from the published Brat RNACompete assays closely match our functionally defined Brat motif (Supplementary Figure S3C) (3,11). As with the 5′-WUGUUR BBS, this 5′-UUGUUD motif is significantly enriched in the 3′UTRs of upregulated genes (Figure 3G) with 87% (82/94) to 98% (44/45) upregulated genes containing the functional Brat motif in the 3′UTR (Figure 3H). Together, these results strongly support the functional role of Brat binding specifically to BBS sites in the 3′UTR of target mRNAs to elicit repression.

The magnitude of Brat repression is proportional to the number of 3′UTR BBS motifs

We further explored the functional relationship of BBS motifs to Brat binding and repression of mRNAs. We observed that mRNAs bound by Brat (3,11) have significantly longer 3′UTRs (P-value < 2.2e−16), whether the longest (Supplementary Figure S3D) or shortest (Supplementary Figure S3E) 3′UTR isoforms are analyzed. The expression of genes regulated by Brat (median RPKM 37.2) is significantly higher (P-value < 8.45e−11) than all other genes (median RPKM 17.2) (Supplementary Figure S3F). Moreover, the Brat-bound mRNAs generally have more BBS motifs in their 3′UTRs (minimum 2.1 ± 1.5, maximum 3.1 ± 2.8) than in unbound transcripts (minimum 1.5 ± 0.9, maximum 2.0 ± 1.8) (Figure 4A and B, and Supplementary File S3). Genes that Brat RNAi upregulates generally have more 3′UTR BBS motifs (minimum 2.4 ± 1.6, maximum 2.9 ± 2.0) than unaffected genes (minimum 1.7 ± 1.2, maximum 2.4 ± 2.4) (Figure 4C and D, and Supplementary File S3). Additionally, genes with more 3′UTR BBS have higher significance-weighted log2 fold changes (Supplementary Figure S4A–C). A complete summary of these relationships, with statistics, is presented in Supplementary File S2. These results suggested that the increased number of BBS motifs promotes binding and repression by Brat.

Figure 4.

Figure 4.

The number of 3′UTR BBS motifs sensitizes mRNAs to repression by Brat. (A, B) Genes were divided by whether they were bound by Brat in either RIP dataset or unbound. The percentage of genes with the indicated number of 3′UTR BBS motifs are displayed as histograms. Both the minimum (A) and maximum (B) number of BBS motifs per gene were considered for genes with multiple mRNA isoforms. Similarly, histograms display the number of 3′UTR motifs among genes upregulated following Brat RNAi and unregulated, when the minimum (C) and maximum (D) number of BBS motifs per gene are considered. In panels (A)–(D), the number of genes per subset are indicated in the key. The number of genes in each category is presented in Supplementary File S3. (E) Renilla luciferase reporters (Rluc) with various numbers of 5′-UUGUUG BBS motifs in the 3′UTR were co-transfected with a control firefly luciferase, and either empty vector (EV) or V5-tagged Brat cDNA. (F) The log2 fold change in reporter expression caused by Brat for each reporter, relative to EV, is plotted for three independent experiments, each with three biological replicates. Biological replicates within the same experimental replicate are denoted (squares, circles and triangles). Mean ± standard deviation values are plotted. Statistically significant differences were determined via ANOVA and post-hoc Tukey–Kramer test, using a mixed linear model (n = 9, ***P < 0.001, n.s.: non-significant).

We directly tested the functional relationship of BBS motif number using a Renilla luciferase (Rluc) reporter (Figure 4E) by inserting an increasing number of BBSs into a minimal 3′UTR (i.e. 0–8 copies of 5′-UUGUUG). These Rluc BBS reporters were co-transfected into Drosophila d.mel-2 cells with either a Brat-V5 expression vector or, as a negative control, empty expression vector. A firefly luciferase gene was co-transfected to normalize for potential variation in transfection efficiency. While overexpressed Brat does not repress reporters with 0–2 BBS motifs, those with 4, 6 and 8 BBS motifs were significantly repressed. The magnitude of repression increased proportionally with the number of BBS motifs (Figure 4F). The opposite effect was observed for these reporters when endogenous Brat was depleted via RNAi: repression was alleviated in proportion with the number of BBS motifs (Supplementary Figure S4D). Thus, both transcriptomic analysis and reporter mRNA data support the conclusion that increasing BBS motifs promote Brat repression.

5′UTR BBS motifs do not confer repression by Brat

Our transcriptome-wide analysis of BBS motifs found that many transcripts have BBS motifs in their 5′UTRs (Supplementary Figure S5A–C). Additionally, we observed 1344 Brat bound (Figure 3A) and 56 upregulated (Figure 3B) mRNAs have 5′UTR BBS motifs, though they were not statistically over-represented in either dataset. We hypothesized that these sites might be functional, bolstered by the reported ability of the Caenorhabditis elegans TRIM-NHL protein Lin-41 to repress translation when bound to a 5′UTR (49). We therefore tested whether BBS motifs in the 5′UTR of a Renilla luciferase reporter could confer repression by Brat. The results show that reporters with four wild-type BBS motifs in the 5′UTR (Figure 5A) were not significantly repressed by Brat-V5 relative to an RNA-binding defective Brat mutant (N933A) (11) or transfection with an empty vector (Figure 5B). Nor was repression evident compared to reporters lacking BBS motifs or mutated BBS motifs in the 5′UTR (Figure 5B). Furthermore, while reporters with 3′UTR motifs were readily repressed by Brat, but not the N933A mutant, this repression was not enhanced by adding of 5′UTR BBS motifs (Supplementary Figure S5D). From these data, we conclude that BBS motifs are not functional in the context of the 5′UTR.

Figure 5.

Figure 5.

Brat does not repress through the 5′UTR. (A) Four 5′-UUGUUG BBS motifs, or mutated (5′-UCC UUG) versions, were placed in the 5′UTR of Renilla luciferase reporters. (B) Luciferase assays testing the activity of Brat, or a mutant that does not bind RNA in vitro (N933A) on these reporters. For all reporters, this activity was compared to an empty vector (EV) negative control. At the bottom, western blot detection of V5-tagged wild type and mutant Brat proteins. Tubulin served as a loading control. Sizes of bands are in kDa. Mean ± standard deviation values are plotted, with replicate measurements (n = 3) within the same experimental replicate (n = 3) denoted (squares, circles and triangles). Significance of differences between conditions was determined via ANOVA with post-hoc Tukey–Kramer Test (***P < 0.001, n.s.: non-significant).

Brat regulates genes encoding components of the glycolysis pathway and vacuolar ATPase complex

To derive insights into Brat’s regulatory network, we analyzed the functions of Brat-repressed genes by performing GO enrichment analysis (Table 1). The results reveal significant enrichment of two cellular processes. The first is central carbon metabolism. Multiple glycolytic genes are upregulated when Brat is depleted (Figure 6A), many of which contain 3′UTR BBS motifs and were shown to be bound by Brat in RIP experiments (Supplementary Figure S6A, Supplementary File S3). While many of these genes are shared by both glycolytic and gluconeogenic pathways (e.g. Pgi, Ald1, Gapdh1, Tpi, Pgk and Eno), some—such as HexA and PyK—are energy-investing steps unique to glycolysis (Figure 6B). In flies, the primary hemolymph carbohydrate is trehalose, which is converted to glucose by Treh (50). The upregulation of HexA, PyK and Treh indicates that depletion of Brat would increase glucose utilization and glycolytic flux. In addition, enzymes involved in glycogen biosynthesis (Pgm1, CG9475, AGBE, GygandGlyS) and the pentose phosphate pathway (Taldo, Pgd, Rpiand Pgm1) are also upregulated by Brat RNAi.

Figure 6.

Figure 6.

Brat represses genes in the glycolysis pathway. (A) Log2 fold change of expression of the indicated glycolytic enzymes significantly (Padj > 0.05, baseMean ≥ 50) upregulated following RNAi of Brat, as measured by RNA-seq (n = 4). (B) Conversion of glucose to pyruvate via glycolysis (blue arrows) and conversion of oxaloacetate to glucose via gluconeogenesis (red arrows) in Drosophila. Additional metabolites such as trehalose, starch and ɑ-d-glucose also enter this pathway (black arrows). Finally, pyruvate is converted to acetyl-CoA and enters the TCA cycle (dashed arrow). Enzymes that catalyze each step of the pathway, that are expressed in DL1 cells, are represented by gene symbols. Genes significantly upregulated following Brat RNAi (discorectangle) and genes expressed but not significantly upregulated (rectangle). Note enzymes required for two steps of gluconeogenesis are not expressed in DL1 cells (dashed border). Created in BioRender. R. Connacher, 2024, BioRender.com/j19h070. Adapted from ‘Glycolysis and Glycolytic Enzymes’, by BioRender.com (2023). Retrieved from https://app.biorender.com/biorender-templates. Fly-specific genes from KEGG pathway dme00010. (C) Western blot verifying the depletion of endogenous, tagged V5::Brat from DL1 cells grown in normal media (+GlutaMax) and glutamine-starved media (−GlutaMax). Glucose-6-phosphate abundance in cells grown in normal media (D) or glutamine-starved media (E). Replicate measurements (n = 3) within the same experimental replicate (n = 4) are denoted (squares, circles, upward triangles and downward triangles). (F) Renilla luciferase reporters containing 3′UTRs of glycolytic enzymes. (G) These reporters were co-transfected with exogenous Brat under the control of an inducible promoter, and compared to corresponding empty vector (EV) samples. Mean ± standard deviation values are plotted, with replicate measurements (n = 3) within the same experimental replicate (n = 3) denoted (squares, circles, upward triangles). In panels (E)–(G), statistically significant differences were determined via ANOVA and Tukey–Kramer tests for post-hoc comparisons, utilized a mixed linear model (***P< 0.001, n.s.: non-significant).

Next, we sought to measure the effect of Brat on central carbon metabolism by measuring levels of several glycolytic metabolites in cells treated with Brat RNAi, relative to negative control RNAi. Western blotting confirmed the thorough depletion of Brat (Figure 6C). As the supplemental glutamine (GlutaMax) in standard culture media can enter the tricarboxylic acid (TCA) cycle as an alternative carbon source (51), depletion of Brat was conducted in parallel with cells lacking this additive. To measure relevant differences in metabolite abundance, while accounting for technical variation, RNAi was repeated in four independent experiments using the same batches of media. Each metabolite was then measured by mass-spectrometry in triplicate. The results show that depletion of Brat significantly increased the amount of glucose-6-phosphate in cells grown in standard culture media (Figure 6D) and glutamine-deprived cells (Figure 6E). This buildup of glucose-6-phosphate is consistent with increased expression of enzymes that generate glucose-6-phosphate via phosphorylation (Hex-A) or isomerization (Pgm1, Pgi and Treh) in response to Brat RNAi, but not the next stage of glycolysis (Pfk). We did not detect significant changes in other downstream metabolites in either culture conditions (Supplementary Figure S6B–G), except for a modest increase in pyruvate and lactate among glutamine-starved cells (Supplementary Figure S6E). Taken together, these results provide the first evidence that Brat regulates the glycolytic pathway.

To further demonstrate that Brat represses the expression of glycolytic genes, we appended the 3′UTRs of two Brat-responsive genes, Tpi and Treh, to Renilla luciferase reporters (Figure 6F). In this context, each reporter is transcribed by the actin promoter. Overexpressed Brat robustly repressed both 3′UTR reporters, relative to the empty vector negative control (Figure 6G), whereas Brat had no effect on the control reporter with a 3′UTR that lacked BBS motifs. These results support the conclusion that Brat represses the expression of these glycolytic genes through their BBS-containing 3′UTRs.

The second process enriched in Brat-repressed genes is proton transport and vacuolar acidification, owing to the regulation of the V-ATPase complex (Table 1). The V-ATPase complex converts the chemical energy of ATP hydrolysis to physical, rotational energy in order to transport hydrogen ions across a lipid bilayer (52). The V-ATPase structure consists of the ATPase V1 domain connected to the membrane-embedded V0 domain via peripheral and central stalks (Figure 7A) (53). V-ATPases acidify lysosomes and late endosomes, and maintain cytosolic pH in specific cell types (52). In our RNA-seq data, we find that DL1 cells express 18 genes, including several paralogs, encoding the 15 subunits of the complex (Figure S7A) (54). Depletion of Brat significantly increased the mRNA levels of 14 of the 18 expressed V-ATPase subunits (Figure 7B). Remarkably, our results reveal that Brat represses genes corresponding to all subunits of the V-ATPase complex, except F (Vha14-1). All of the V-ATPase genes upregulated by Brat RNAi contain 3′UTR BBS motifs (Supplementary Figure S7A). Moreover, 10 of the V-ATPase gene mRNAs were identified as Brat-bound in published RIP experiments (Supplementary Figure S7A, Supplementary File S4) with Vha55, Vha36-1, VhaSFD, VhaPPA1-1, Vha100-2, VhaM9.7-2/b, VhaAC45 and VhaM8.9 mRNAs detected in both RIP datasets. Thus, a preponderance of evidence supports the direct repression of V-ATPase mRNAs by Brat.

Figure 7.

Figure 7.

Brat represses V-ATPase complex subunits. (A) Structure of the human V-ATPase complex, PDB #ID 6WM2. The ATPase hexamer of subunits A and B and proton-transporting subunits c and c″ are surrounded by a scaffold of subunits G, E, C, H, a and e. Rotation is transferred between the ATPase and proton-transporting components by the central stalk of subunits D and d and F. Additional subunits include S1 and M8.9. (B) Log2 fold change (± standard error) of V-ATPase subunits significantly (Padj > 0.05, baseMean ≥ 50) upregulated following RNAi of Brat, as measured by RNA-seq (n = 4). (C) RNA levels of V-ATPase subunits and the metabolic gene Treh, relative to 7SK level, in heterozygous and brat mutant larvae determined via RT-qPCR. (D) Similarly, RNA levels of V-ATPase subunits were determined in larvae in which ubiquitous expression of Brat was induced (Da>Brat), compared to larvae only containing the Da-Gal4 driver (Da>). For all RT-qPCR, mean ± standard deviation values are plotted. Significant differences determined via Welch’s t-test (*P< 0.05, **P< 0.01, ***P< 0.001 and n = 4). (E) Endogenous Brat was depleted via RNAi, and the level of tagged Vha100-2 detected via Western blot. (F) Quantitation of band intensity via densitometry. Mean ± standard deviation values are plotted. Significance of differences were determined via ANOVA with a post-hoc Tukey–Kramer test, using a linear model (**P< 0.001 and n = 3). In all blots, sizes of bands are displayed in kDa.

Building on our observations in cells, we then measured the effect of Brat on V-ATPase and Trehalase mRNA levels in larvae using RT-qPCR (Figure 7C). First, we analyzed the strong loss of function brat1 allele, which contains a premature termination codon within the NHL domain at residue Q872, resulting in a protein product incapable of binding to RNA (Supplementary Figure S7B) (4). Compared to heterozygous animals, hemizygous brat1 larvae exhibited significant increases in the mRNA levels from genes encoding V-ATPase subunits Vha100-2 and VhaM8.9 and the Trehalase mRNA (Figure 7C). To further analyze Brat-mediated repression in vivo, we employed a temperature-sensitive Gal80 and ubiquitous Da-Gal4 driver to overexpress a Brat transgene under the control of a UAS promoter (15). When shifted to the permissive temperature, the induced Brat protein significantly reduced the mRNA levels of V-ATPase subunits Vha100-2 and VhaM8.9 compared to the driver-only control (Figure 7D). These results verify that Brat represses V-ATPase and Trehalase expression in vivo.

We also measured Brat repression of protein expression from an endogenous V-ATPase subunit. To do so, the N-terminus of the Vha100-2 gene was tagged with Flag epitope via CRISPR–Cas9 gene editing in DL1 cells. RNAi of Brat caused a 2-fold increase in Flag-Vha100-2 protein, as measured by quantitative western blot (Figure 7E and F) within the linear response range of the assay (Supplementary Figure S7C and D).

Finally, as with glycolytic genes, we appended the 3′UTRs of four V-ATPase subunits—VhaAC39-1, VhaM8.9, Vha100-2 and VhaPPA1-1—to luciferase reporters (Figure 8A and B). Overexpression of Brat robustly repressed both V-ATPase 3′UTR reporters relative to empty vector negative control. Taken together, our data provide evidence that Brat coordinately and directly represses the expression of the V-ATPase complex in cultured cells and Drosophila larvae.

Figure 8.

Figure 8.

The 3′UTRs of V-ATPase subunits confer sensitivity to Brat (A) and (B) Renilla luciferase reporters containing 3′UTRs of V-ATPase subunits VhaAC39-1 and VhaM8.9 (A) or Vha100-2 and VhaPPA1-1 (B). These reporters were co-transfected with exogenous Brat under the control of an inducible promoter, and compared to corresponding empty vector samples. Mean ± standard deviation values are plotted, with replicate measurements (n = 3) within the same experimental replicate (n = 4) denoted (squares, circles and triangles). Statistically significant differences were determined via ANOVA and Tukey–Kramer tests for post-hoc comparisons, utilized a mixed linear model (***P< 0.001, n.s.: non-significant).

Discussion

The results of this study expand the paradigm of Brat-mediated post-transcriptional repression on a transcriptome-wide level and identify new functional categories of target genes. We rigorously demonstrated that Brat significantly reduces mRNA expression levels of 158 genes, bolstered by two RNAi conditions and multiple replicates. Among these, 59% (94/158) were previously shown to be bound by Brat (1,11) and 78% (124/158) contain 3′UTR BBS motifs, consistent with these genes being direct Brat-regulated target mRNAs. More stringently, 27% (42/158) are ‘high confidence’ direct targets that are repressed by Brat, have BBS motifs and were shown to be bound by Brat in both RIPs. Our results expand the repertoire of Brat-regulated genes by at least 108 new targets, in addition to 50 genes previously implicated as Brat targets (3,15). We further corroborated Brat-mediated repression of several new target mRNAs in larvae.

We observed that repression of gene expression is significantly associated with the presence of 3′UTR BBS motifs in the transcriptome data, which we experimentally validated using reporter gene assays. The enrichment of 5′-UUGUUD motifs in 3′UTRs of repressed RNAs complements the 5′-WUGUUR motifs previously demonstrated by multiple binding studies (3,11). Collectively, the data support a conservative definition of the BBS as 5′-WUGUUD, which incorporates both binding and regulation. Furthermore, the sensitivity of Brat-mediated repression depends on the number of 3′UTR motifs, as indicated by transcriptome-wide analysis and demonstrated by luciferase reporter assays. While the NHL domain is capable of binding RNA as a monomer (11), this may reflect the necessity of multiple Brat molecules to efficiently occupy and repress RNAs. This possibility is particularly intriguing, as Brat has been reported to form dimers through the coiled-coil domain (55). These broad observations are also supported by documented Brat-repressed genes, hunchback and deadpan, which have 3′UTR BBS motifs necessary for their regulation (1,9,11,15).

Our study provides strong evidence that Brat reduces mRNA levels; however, we cannot exclude that Brat may reduce the translation of select target mRNAs in addition to promoting mRNA decay. In the future, global analysis of the impact of Brat on translation will be necessary. Our data do not support a significant effect of Brat acting through 5′UTR motifs, both on a transcriptome level or in direct tests using reporter genes. Interestingly, this contrasts with the C. elegans TRIM-NHL LIN-41, which can inhibit the translation of mRNAs when bound to 5′UTR motifs (49). Finally, while our results demonstrate that Brat represses mRNAs, we cannot rule out the possibility that Brat additionally affects non-coding RNAs.

Notably, several previously reported Brat target genes were not differentially expressed in DL1 cells in response to Brat RNAi. In the case of hunchback and deadpan (14,15), their mRNA levels are very low, falling below the expression cutoff. Whereas myc and BMP signaling components Med, shn and Mad were shown to be repressed by Brat in stem cells (7,8,15,56), they were not differentially expressed in DL1 cells. We speculate that additional regulatory factors may supersede the effect of Brat on these mRNAs in DL1 cells. Alternatively, differential 3′ end processing of these transcripts may alter their responsiveness.

In our data, RNAi of Brat also led to the downregulation of a small number of genes (28, excluding Brat itself, Supplementary Figure S1G). Currently, no evidence supports Brat’s direct effect on the stabilization of mRNAs. No significant overlap of these downregulated genes was detected with the Brat-bound transcripts. Moreover, few of the genes that were downregulated by Brat RNAi have 3′UTR BBS motifs. Only one transcript fits the criteria for a direct BBS-containing, Brat-bound and regulated target: sulfateless, which encodes an enzyme involved in heparin biosynthesis. As no phenotypic connection between Brat and this gene has been documented, the potential biological relevance is unclear. More than likely, these downregulated genes may be secondary, indirect effects of Brat RNAi.

Our results in DL1 cells, with supporting evidence in larvae, demonstrate that Brat represses mRNAs encoding 14 of the 15 V-ATPase subunits. This coordinate regulation of the V-ATPase is a prime example of the post-transcriptional regulon theory for RBP function (57). Such control may help ensure stoichiometric expression of V-ATPase subunits and modulate V-ATPase levels in response to developmental or environmental cues. Repression by Brat reduces their expression levels and, therefore, is anticipated to reduce the abundance of V-ATPase complexes in endosome and lysosome membranes. A potential consequence is that the acidification of these compartments would be diminished. As intracellular pH and its regulation play crucial roles in a variety of biochemical and physiological processes, future research should explore the potential impact of Brat on pH-mediated processes (52).

Similar coordinated regulation of V-ATPase expression—at the transcriptional level—has been previously reported. The Drosophila TFEB homolog, the transcription factor Mitf, activates many V-ATPase subunits by recognizing CLEAR/M-box motifs in the promoters of these genes (58,59). This coordinates the expression, and subsequent activity, of the V-ATPase complex (59). In addition to this established mode of transcriptional coordination, we propose that Brat coordinates the mRNA levels of V-ATPase subunits post-transcriptionally through canonical BBS motifs in the 3′UTRs of V-ATPase genes. A similar paradigm has been demonstrated for miR-1 (60). A question for future research is how this opposing activating and repressing regulation affects V-ATPase activity, which may differ in specific tissues or developmental processes.

Intriguingly, genetic analysis previously linked the V-ATPase complex to Brat in the context of brain tumors. Depletion of V-ATPase subunits in larval neural stem cells suppressed the formation of brain tumors caused by loss of Brat function (61). At that time, this result was interpreted to reflect attenuation of Notch signaling, which is normally dampened by Brat in differentiating neural stem cells by a multi-prong mechanism [reviewed in (1,62)]. Instead, our new results provide a direct functional connection between Brat and V-ATPase. Brat normally represses V-ATPase levels, whereas loss of Brat function causes neural stem cell proliferation, promoted by increased V-ATPase levels, at the expense of differentiation. Together, these results support a novel layer of regulation by which Brat ensures proper differentiation of neural stem cells.

Our data reveal a second Brat regulon: Brat represses the expression of multiple glycolytic enzymes through their 3′UTRs (Figure 5). While several Brat-repressed genes are involved in both glycolysis and gluconeogenesis, our observation that Brat also represses energy-committing enzymes HexA and PyK suggests a targeted effect on glycolysis. Consistent with this observation, we found that glucose-6-phosphate levels are increased upon depletion of Brat. These results have important implications for Brat’s tumor suppressor function. Tumors rely on increased glycolysis rather than aerobic respiration for survival and proliferation (63,64). Our new results provide a direct functional connection wherein Brat normally represses the expression of glycolytic enzymes. This implies that the observed metabolic changes following loss of Brat function, that contribute to tumor formation, may be causative rather than merely correlative. Research into the metabolic changes during cancer progression is ongoing. While targeted RNAi of Brat in Type II neuroblasts increases the levels of glycolytic and fermentation enzymes (27), more thorough analysis shows Brat-derived tumors convert glucose to lactate while preferentially utilizing glutamine in the TCA cycle (26). Finally, an additional functional connection of Brat to the control of glycolysis occurs during metamorphosis. Ecdysone signaling switches Drosophila neural stem cells from primarily utilizing glycolysis to oxidative phosphorylation at the onset of metamorphosis (65), at which ecdysone also induces Brat expression (66–68).

The two Brat regulons discovered in this study, V-ATPase and glycolysis, are functionally interconnected. In flies and mammals, the V-ATPase complex has been directly associated with mTOR signaling (69). In mammals, Aldolase bridges the V-ATPase complex with central carbon metabolism [reviewed in (70)]. Our results indicate that Brat may mediate that coordination. We also observed that Brat directly represses Trehalase and Aldolase 1 expression. The former agrees with a previous study, wherein RNAi depletion of Brat in neural stem cells increased Treh mRNA level (15). Trehalase breaks down the disaccharide trehalose—the major circulating form of glucose in insects (50)—into glucose. As trehalose is utilized by larvae (71) increased Trehalase could promote glycolytic flux in brat-derived tumors.

The role of Brat as a tumor suppressor during neural stem cell differentiation is well established (2,4–6), including its ability to repress key transcription factors (15,16,29). The Brat-mediated repression of the glycolytic pathway and the V-ATPase complex may provide additional layers of tumor suppression. The multi-pronged repression of these pathways may be necessary for fine-tuning regulation in other contexts, such as during oogenesis and embryonic development (3,7,8).

Brat is part of a growing list of TRIM-NHL proteins that function as post-transcriptional repressors, including mammalian TRIM71 and C. elegans LIN-41 (1,49,72,73). In addition to similarities in structure and architecture, these proteins function similarly in different organisms. In mouse embryonic stem cells, TRIM71 largely alters RNA abundance, rather than translation, by recognizing 3′UTR motifs (72). In C. elegans, LIN-41 can alter RNA abundance through 3′UTR motifs, but also controls translation through 5′UTR motifs (49). The precise molecular mechanisms and cofactors involved in mRNA regulation by these RBPs remain to be fully elucidated, and we anticipate that Brat will continue to serve as an archetype. Continued research is essential to discover the mechanisms and regulatory networks relevant to the essential roles of TRIM-NHLs in development and how their dysfunction contributes to diseases, including cancer and congenital hydrocephalus (1,74).

Supplementary Material

gkae810_Supplemental_Files

Acknowledgements

We thank Katherine McKenney and Elise Dunshee for their insightful comments and suggestions throughout this study. We thank the staff of the NIEHS Epigenomics and DNA Sequencing Core Facility and the NIEHS Integrative Bioinformatics Support Group for support of the RNA sequencing and analysis. We thank Michael O’Connor for supporting fly genetics and kindly providing several fly stocks used in this publication. We thank Joseph Duchamp of Indiana University of Pennsylvania for advice regarding statistical analysis of the luciferase assays. We thank Fred Lih for advice regarding liquid chromatography conditions for the LC–MS experiments. Portions of the graphical abstract (PDB 1QF7, 6WM2) were created in BioRender (R. Connacher, 2024, BioRender.com/z92c743).

Notes

Present address: Noel Dittbenner, Department of Genetics, Cell Biology, and Development, University of Minnesota, Minneapolis, MN 55455, USA.

Present address: Richard T. Roden, Alix School of Medicine, Mayo Clinic, Rochester, MN 55905, USA.

Contributor Information

Robert P Connacher, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Richard T Roden, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Kai-Lieh Huang, Department of Biochemistry and Biophysics, University of Rochester Medical Center, 575 Elmwood Avenue, Rochester, NY 14642, USA.

Amanda J Korte, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Saathvika Yeruva, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Noel Dittbenner, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Anna J DesMarais, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Chase A Weidmann, Department of Biological Chemistry, Center for RNA Biomedicine, University of Michigan Medical School, 1150 West Medical Center Drive, Ann Arbor, MI 48109, USA.

Thomas A Randall, Integrative Bioinformatics Support Group, National Institute of Environmental Health Sciences (NIEHS), 111 TW Alexander Drive, Research Triangle Park, NC 27709, USA.

Jason Williams, Epigenetics & Stem Cell Biology Laboratory, National Institute of Environmental Health Sciences (NIEHS), 111 TW Alexander Drive, Research Triangle Park, NC 27709, USA.

Traci M Tanaka Hall, Epigenetics & Stem Cell Biology Laboratory, National Institute of Environmental Health Sciences (NIEHS), 111 TW Alexander Drive, Research Triangle Park, NC 27709, USA.

Eric J Wagner, Department of Biochemistry and Biophysics, University of Rochester Medical Center, 575 Elmwood Avenue, Rochester, NY 14642, USA.

Aaron C Goldstrohm, Department of Biochemistry, Molecular Biology, and Biophysics, University of Minnesota, 1214A 6-155 Jackson Hall, 321 Church Street S.E., Minneapolis, MN 55455, USA.

Data availability

The RNA-seq data generated in this study have been deposited to the National Institutes of Health BioProject database under the BioProject ID PRJNA1048880. Differential expression data are reported here in Supplementary File S4.

Supplementary data

Supplementary Data are available at NAR Online.

Funding

National Institute of General Medical Sciences, National Institutes of Health [R01 GM145835 and R01 GM105707 to A.C.G.]; Intramural Research Program of the National Institutes of Health, National Institute of Environmental Health Sciences [ZIA-ES050165 to T.M.T.H., ZIC ES103005 to J.G.W.]; Edith Walters Jones and Robert Jones Fellowship (to R.C.); National Cancer Institute [K22 CA262349 to C.A.W.]; University of Michigan Center for RNA Biomedicine (to C.A.W.); Rogel Cancer Center [P30 CA046592 to C.A.W.]; University of Rochester (to E.J.W). Funding for open access charge: National Institutes of General Medical Sciences, National Institutes of Health [R01 GM145835 to A.C.G].

Conflict of interest statement. None declared.

References

  • 1. Connacher R.P., Goldstrohm A.C.. Molecular and biological functions of TRIM-NHL RNA-binding proteins. Wiley Interdiscip. Rev. RNA. 2021; 12:e1620. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2. Lee C.Y., Wilkinson B.D., Siegrist S.E., Wharton R.P., Doe C.Q.. Brat is a Miranda cargo protein that promotes neuronal differentiation and inhibits neuroblast self-renewal. Dev. Cell. 2006; 10:441–449. [DOI] [PubMed] [Google Scholar]
  • 3. Laver J.D., Li X., Ray D., Cook K.B., Hahn N.A., Nabeel-Shah S., Kekis M., Luo H., Marsolais A.J., Fung K.Y.et al.. Brain tumor is a sequence-specific RNA-binding protein that directs maternal mRNA clearance during the Drosophila maternal-to-zygotic transition. Genome Biol. 2015; 16:94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Arama E., Dickman D., Kimchie Z., Shearn A., Lev Z.. Mutations in the beta-propeller domain of the Drosophila brain tumor (brat) protein induce neoplasm in the larval brain. Oncogene. 2000; 19:3706–3716. [DOI] [PubMed] [Google Scholar]
  • 5. Bowman S.K., Rolland V., Betschinger J., Kinsey K.A., Emery G., Knoblich J.A.. The tumor suppressors Brat and Numb regulate transit-amplifying neuroblast lineages in Drosophila. Dev. Cell. 2008; 14:535–546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6. Betschinger J., Mechtler K., Knoblich J.A.. Asymmetric segregation of the tumor suppressor brat regulates self-renewal in Drosophila neural stem cells. Cell. 2006; 124:1241–1253. [DOI] [PubMed] [Google Scholar]
  • 7. Harris R.E., Pargett M., Sutcliffe C., Umulis D., Ashe H.L.. Brat promotes stem cell differentiation via control of a bistable switch that restricts BMP signaling. Dev. Cell. 2011; 20:72–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8. Newton F.G., Harris R.E., Sutcliffe C., Ashe H.L.. Coordinate post-transcriptional repression of Dpp-dependent transcription factors attenuates signal range during development. Development. 2015; 142:3362–3373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Kumari P., Aeschimann F., Gaidatzis D., Keusch J.J., Ghosh P., Neagu A., Pachulska-Wieczorek K., Bujnicki J.M., Gut H., Grosshans H.et al.. Evolutionary plasticity of the NHL domain underlies distinct solutions to RNA recognition. Nat. Commun. 2018; 9:1549. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10. Loedige I., Gaidatzis D., Sack R., Meister G., Filipowicz W.. The mammalian TRIM-NHL protein TRIM71/LIN-41 is a repressor of mRNA function. Nucleic Acids Res. 2013; 41:518–532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11. Loedige I., Jakob L., Treiber T., Ray D., Stotz M., Treiber N., Hennig J., Cook K.B., Morris Q., Hughes T.R.et al.. The crystal structure of the NHL domain in complex with RNA reveals the molecular basis of Drosophila brain-tumor-mediated gene regulation. Cell Rep. 2015; 13:1206–1220. [DOI] [PubMed] [Google Scholar]
  • 12. Loedige I., Stotz M., Qamar S., Kramer K., Hennig J., Schubert T., Loffler P., Langst G., Merkl R., Urlaub H.et al.. The NHL domain of BRAT is an RNA-binding domain that directly contacts the hunchback mRNA for regulation. Genes Dev. 2014; 28:749–764. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13. Tocchini C., Ciosk R.. TRIM-NHL proteins in development and disease. Semin. Cell Dev. Biol. 2015; 47–48:52–59. [DOI] [PubMed] [Google Scholar]
  • 14. Sonoda J., Wharton R.P.. Drosophila Brain Tumor is a translational repressor. Genes Dev. 2001; 15:762–773. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15. Reichardt I., Bonnay F., Steinmann V., Loedige I., Burkard T.R., Meister G., Knoblich J.A.. The tumor suppressor brat controls neuronal stem cell lineages by inhibiting Deadpan and Zelda. EMBO Rep. 2018; 19:102–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16. Komori H., Golden K.L., Kobayashi T., Kageyama R., Lee C.Y.. Multilayered gene control drives timely exit from the stem cell state in uncommitted progenitors during Drosophila asymmetric neural stem cell division. Genes Dev. 2018; 32:1550–1561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17. Wang Y., Yu Z., Wang M., Liu C.P., Yang N., Xu R.M.. 5EX7: crystal structure of Brat NHL domain in complex with an 8-nt hunchback mRNA. 2015; 10.2210/pdb5EX7/pdb. [DOI]
  • 18. Arvola R.M., Weidmann C.A., Tanaka Hall T.M., Goldstrohm A.C.. Combinatorial control of messenger RNAs by Pumilio, Nanos and Brain Tumor proteins. RNA Biol. 2017; 14:1445–1456. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19. Schupbach T., Wieschaus E.. Female sterile mutations on the second chromosome of Drosophila melanogaster. II. Mutations blocking oogenesis or altering egg morphology. Genetics. 1991; 129:1119–1136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. Wright T.R., Beermann W., Marsh J.L., Bishop C.P., Steward R., Black B.C., Tomsett A.D., Wright E.Y.. The genetics of dopa decarboxylase in Drosophila melanogaster. IV. The genetics and cytology of the 37B10-37D1 region. Chromosoma. 1981; 83:45–58. [DOI] [PubMed] [Google Scholar]
  • 21. Kurzik-Dumke U., Phannavong B., Gundacker D., Gateff E.. Genetic, cytogenetic and developmental analysis of the Drosophila melanogaster tumor suppressor gene lethal(2)tumorous imaginal discs (1(2)tid). Differentiation. 1992; 51:91–104. [DOI] [PubMed] [Google Scholar]
  • 22. Woodhouse E., Hersperger E., Shearn A.. Growth, metastasis, and invasiveness of Drosophila tumors caused by mutations in specific tumor suppressor genes. Dev. Genes Evol. 1998; 207:542–550. [DOI] [PubMed] [Google Scholar]
  • 23. Loop T., Leemans R., Stiefel U., Hermida L., Egger B., Xie F., Primig M., Certa U., Fischbach K.F., Reichert H.et al.. Transcriptional signature of an adult brain tumor in Drosophila. BMC Genomics. 2004; 5:24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24. Juschke C., Dohnal I., Pichler P., Harzer H., Swart R., Ammerer G., Mechtler K., Knoblich J.A.. Transcriptome and proteome quantification of a tumor model provides novel insights into post-transcriptional gene regulation. Genome Biol. 2013; 14:r133. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25. Janic A., Mendizabal L., Llamazares S., Rossell D., Gonzalez C.. Ectopic expression of germline genes drives malignant brain tumor growth in Drosophila. Science. 2010; 330:1824–1827. [DOI] [PubMed] [Google Scholar]
  • 26. Bonnay F., Veloso A., Steinmann V., Köcher T., Abdusselamoglu M.D., Bajaj S., Rivelles E., Landskron L., Esterbauer H., Zinzen R.P.et al.. Oxidative metabolism drives immortalization of neural stem cells during tumorigenesis. Cell. 2020; 182:1490–1507. [DOI] [PubMed] [Google Scholar]
  • 27. Landskron L., Steinmann V., Bonnay F., Burkard T.R., Steinmann J., Reichardt I., Harzer H., Laurenson A.-S., Reichert H., Knoblich J.A.. The asymmetrically segregating lncRNA cherub is required for transforming stem cells into malignant cells. eLife. 2018; 7:e31347. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28. Arvola R.M., Chang C.-T., Buytendorp J.P., Levdansky Y., Valkov E., Freddolino P.L., Goldstrohm A.C.. Unique repression domains of Pumilio utilize deadenylation and decapping factors to accelerate destruction of target mRNAs. Nucleic Acids Res. 2020; 48:1843–1871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29. Komori H., Xiao Q., McCartney B.M., Lee C.Y.. Brain tumor specifies intermediate progenitor cell identity by attenuating beta-catenin/armadillo activity. Development. 2014; 141:51–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30. Gratz S.J., Ukken F.P., Rubinstein C.D., Thiede G., Donohue L.K., Cummings A.M., O’Connor-Giles K.M. Highly specific and efficient CRISPR/Cas9-catalyzed homology-directed repair in Drosophila. Genetics. 2014; 196:961–971. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31. Schneider I. Cell lines derived from late embryonic stages of Drosophila melanogaster. J. Embryol. Exp. Morphol. 1972; 27:353–365. [PubMed] [Google Scholar]
  • 32. Van Etten J., Schagat T.L., Goldstrohm A.C.. A guide to design and optimization of reporter assays for 3′ untranslated region mediated regulation of mammalian messenger RNAs. Methods. 2013; 63:110–118. [DOI] [PubMed] [Google Scholar]
  • 33. Weidmann C.A., Raynard N.A., Blewett N.H., Van Etten J., Goldstrohm A.C.. The RNA binding domain of Pumilio antagonizes poly-adenosine binding protein and accelerates deadenylation. RNA. 2014; 20:1298–1319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34. Bustin S.A., Benes V., Garson J.A., Hellemans J., Huggett J., Kubista M., Mueller R., Nolan T., Pfaffl M.W., Shipley G.L.et al.. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 2009; 55:611–622. [DOI] [PubMed] [Google Scholar]
  • 35. Ye J., Coulouris G., Zaretskaya I., Cutcutache I., Rozen S., Madden T.L.. Primer-BLAST: a tool to design target-specific primers for polymerase chain reaction. BMC Bioinformatics. 2012; 13:134. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36. Love M.I., Huber W., Anders S.. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014; 15:550. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37. Maleki F., Ovens K., Hogan D.J., Kusalik A.J.. Gene set analysis: challenges, opportunities, and future research. Front. Genet. 2020; 11:654. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38. van Belle G., Fisher L.D., Heagerty P.J., Lumley T.. Biostatistics: A Methodology for the Health Sciences. 2004; 2nd edn.Hoboken, NJ: John Wiley and Sons, Inc. [Google Scholar]
  • 39. Haugen R.J., Barnier C., Elrod N.D., Luo H., Jensen M.K., Ji P., Smibert C.A., Lipshitz H.D., Wagner E.J., Freddolino P.L.et al.. Regulation of the Drosophila transcriptome by Pumilio and CCR4-NOT deadenylase complex. RNA. 2024; 30:866–890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40. Wolfe M.B., Schagat T.L., Paulsen M.T., Magnuson B., Ljungman M., Park D., Zhang C., Campbell Z.T., Goldstrohm A.C., Freddolino P.L.. Principles of mRNA control by human PUM proteins elucidated from multimodal experiments and integrative data analysis. RNA. 2020; 26:1680–1703. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41. Bailey T.L., Johnson J., Grant C.E., Noble W.S.. The MEME Suite. Nucleic Acids Res. 2015; 43:W39–W49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42. Bailey T.L., Elkan C.. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc. Int. Conf. Intell. Syst. Mol. Biol. 1994; 2:28–36. [PubMed] [Google Scholar]
  • 43. Ashburner M., Ball C.A., Blake J.A., Botstein D., Butler H., Cherry J.M., Davis A.P., Dolinski K., Dwight S.S., Eppig J.T.et al.. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat. Genet. 2000; 25:25–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44. Ontology Consortium G., Aleksander S.A., Balhoff J., Carbon S., Cherry J.M., Drabkin H.J., Ebert D., Feuermann M., Gaudet P., Harris N.L.et al.. The Gene Ontology knowledgebase in 2023. Genetics. 2023; 224:iyad031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45. Hosseinkhani F., Huang L., Dubbelman A.-C., Guled F., Harms A.C., Hankemeier T.. Systematic evaluation of HILIC stationary phases for global metabolomics of human plasma. Metabolites. 2022; 12:165. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46. Ulvila J., Parikka M., Kleino A., Sormunen R., Ezekowitz R.A., Kocks C., Rämet M.. Double-stranded RNA is internalized by scavenger receptor-mediated endocytosis in Drosophila S2 cells. J. Biol. Chem. 2006; 281:14370–14375. [DOI] [PubMed] [Google Scholar]
  • 47. Saleh M.-C., van Rij R.P., Hekele A., Gillis A., Foley E., O’Farrell P.H., Andino R.. The endocytic pathway mediates cell entry of dsRNA to induce RNAi silencing. Nat. Cell Biol. 2006; 8:793–802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48. Routh A., Head S.R., Ordoukhanian P., Johnson J.E.. ClickSeq: fragmentation-free next-generation sequencing via click ligation of adaptors to stochastically terminated 3′-azido cDNAs. J. Mol. Biol. 2015; 427:2610–2616. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49. Aeschimann F., Kumari P., Bartake H., Gaidatzis D., Xu L., Ciosk R., Grosshans H.. LIN41 post-transcriptionally silences mRNAs by two distinct and position-dependent mechanisms. Mol. Cell. 2017; 65:476–489. [DOI] [PubMed] [Google Scholar]
  • 50. Shukla E., Thorat L.J., Nath B.B., Gaikwad S.M.. Insect trehalase: physiological significance and potential applications. Glycobiology. 2015; 25:357–367. [DOI] [PubMed] [Google Scholar]
  • 51. Yoo H.C., Yu Y.C., Sung Y., Han J.M.. Glutamine reliance in cell metabolism. Exp. Mol. Med. 2020; 52:1496–1516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52. Collins M.P., Forgac M.. Regulation and function of V-ATPases in physiology and disease. Biochim. Biophys. Acta Biomembr. 2020; 1862:183341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53. Cipriano D.J., Wang Y., Bond S., Hinton A., Jefferies K.C., Qi J., Forgac M.. Structure and regulation of the vacuolar ATPases. Biochim. Biophys. Acta. 2008; 1777:599–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54. Allan A.K., Du J., Davies S.A., Dow J.A.T.. Genome-wide survey of V-ATPase genes in Drosophila reveals a conserved renal phenotype for lethal alleles. Physiol. Genomics. 2005; 22:128–138. [DOI] [PubMed] [Google Scholar]
  • 55. Liu C., Shan Z., Diao J., Wen W., Wang W.. Crystal structure of the coiled-coil domain of Drosophila TRIM protein brat. Proteins. 2019; 87:706–710. [DOI] [PubMed] [Google Scholar]
  • 56. Song Y., Lu B.. Regulation of cell growth by Notch signaling and its differential requirement in normal vs. tumor-forming stem cells in Drosophila. Genes Dev. 2011; 25:2644–2658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57. Keene J.D. RNA regulons: coordination of post-transcriptional events. Nat. Rev. Genet. 2007; 8:533–543. [DOI] [PubMed] [Google Scholar]
  • 58. Bouché V., Espinosa A.P., Leone L., Sardiello M., Ballabio A., Botas J.. Drosophila Mitf regulates the V-ATPase and the lysosomal-autophagic pathway. Autophagy. 2016; 12:484–498. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59. Zhang T., Zhou Q., Ogmundsdottir M.H., Möller K., Siddaway R., Larue L., Hsing M., Kong S.W., Goding C.R., Palsson A.et al.. Mitf is a master regulator of the v-ATPase, forming a control module for cellular homeostasis with v-ATPase and TORC1. J. Cell Sci. 2015; 128:2938–2950. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60. Gutiérrez-Pérez P., Santillán E.M., Lendl T., Wang J., Schrempf A., Steinacker T.L., Asparuhova M., Brandstetter M., Haselbach D., Cochella L.. miR-1 sustains muscle physiology by controlling V-ATPase complex assembly. Sci. Adv. 2021; 7:eabh1434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61. Wissel S., Harzer H., Bonnay F., Burkard T.R., Neumuller R.A., Knoblich J.A.. Time-resolved transcriptomics in neural stem cells identifies a v-ATPase/notch regulatory loop. J. Cell Biol. 2018; 217:3285–3300. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62. Janssens D.H., Lee C.Y.. It takes two to tango, a dance between the cells of origin and cancer stem cells in the Drosophila larval brain. Semin. Cell Dev. Biol. 2014; 28:63–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63. Warburg O. The metabolism of carcinoma cells. J. Cancer Res. 1925; 9:148–163. [Google Scholar]
  • 64. Jiang H., Kimura T., Hai H., Yamamura R., Sonoshita M.. Drosophila as a toolkit to tackle cancer and its metabolism. Front. Oncol. 2022; 12:98275. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65. Homem C.C.F., Steinmann V., Burkard T.R., Jais A., Esterbauer H., Knoblich J.A.. Ecdysone and mediator change energy metabolism to terminate proliferation in Drosophila neural stem cells. Cell. 2014; 158:874–888. [DOI] [PubMed] [Google Scholar]
  • 66. Beckstead R.B., Lam G., Thummel C.S.. The genomic response to 20-hydroxyecdysone at the onset of Drosophila metamorphosis. Genome Biol. 2005; 6:R99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67. Stoiber M., Celniker S., Cherbas L., Brown B., Cherbas P.. Diverse hormone response networks in 41 independent Drosophila cell lines. G3. 2016; 6:683–694. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68. Uyehara C.M., McKay D.J.. Direct and widespread role for the nuclear receptor EcR in mediating the response to ecdysone in Drosophila. Proc. Natl. Acad. Sci. U.S.A. 2019; 116:9893–9902. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69. Zoncu R., Bar-Peled L., Efeyan A., Wang S., Sancak Y., Sabatini D.M.. mTORC1 senses lysosomal amino acids through an inside-out mechanism that requires the vacuolar H+-ATPase. Science. 2011; 334:678–683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70. Hayek S.R., Rane H.S., Parra K.J.. Reciprocal regulation of V-ATPase and glycolytic pathway elements in health and disease. Front. Physiol. 2019; 10:127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 71. Matsuda H., Yamada T., Yoshida M., Nishimura T.. Flies without trehalose. J. Biol. Chem. 2015; 290:1244–1255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 72. Welte T., Tuck A.C., Papasaikas P., Carl S.H., Flemr M., Knuckles P., Rankova A., Buhler M., Grosshans H.. The RNA hairpin binder TRIM71 modulates alternative splicing by repressing MBNL1. Genes Dev. 2019; 33:1221–1235. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 73. Liu Q., Novak M.K., Pepin R.M., Maschhoff K.R., Worner K., Chen X., Zhang S., Hu W.. A congenital hydrocephalus-causing mutation in Trim71 induces stem cell defects via inhibiting Lsd1 mRNA translation. EMBO Rep. 2023; 24:e55843. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 74. Furey C.G., Choi J., Jin S.C., Zeng X., Timberlake A.T., Nelson-Williams C., Mansuri M.S., Lu Q., Duran D., Panchagnula S.et al.. De novo mutation in genes regulating neural stem cell fate in human congenital hydrocephalus. Neuron. 2018; 99:302–314. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

gkae810_Supplemental_Files

Data Availability Statement

The RNA-seq data generated in this study have been deposited to the National Institutes of Health BioProject database under the BioProject ID PRJNA1048880. Differential expression data are reported here in Supplementary File S4.


Articles from Nucleic Acids Research are provided here courtesy of Oxford University Press

RESOURCES