The RNA-binding protein Nab2 regulates levels of the RhoGEF Trio to govern axon and dendrite morphology

Abstract

The Drosophila RNA-binding protein (RBP) Nab2 acts in neurons to regulate neurodevelopment and is orthologous to the human intellectual disability-linked RBP, ZC3H14. Nab2 governs axon projection in mushroom body neurons and limits dendritic arborization of class IV sensory neurons in part by regulating splicing events in ∼150 mRNAs. Analysis of the Sex-lethal (Sxl) mRNA revealed that Nab2 promotes an exon-skipping event and regulates m⁶A methylation on Sxl pre-mRNA by the Mettl3 methyltransferase. Mettl3 heterozygosity broadly rescues Nab2^null phenotypes implying that Nab2 acts through similar mechanisms on other RNAs, including unidentified targets involved in neurodevelopment. Here, we show that Nab2 and Mettl3 regulate the removal of a 5′UTR (untranslated region) intron in the trio pre-mRNA. Trio utilizes two GEF domains to balance Rac and RhoGTPase activity. Intriguingly, an isoform of Trio containing only the RhoGEF domain, GEF2, is depleted in Nab2^null nervous tissue. Expression of Trio-GEF2 rescues projection defects in Nab2^null axons and dendrites, while the GEF1 Rac1-regulatory domain exacerbates these defects, suggesting Nab2-mediated regulation Trio-GEF activities. Collectively, these data indicate that Nab2-regulated processing of trio is critical for balancing Trio-GEF1 and -GEF2 activity and show that Nab2, Mettl3, and Trio function in a common pathway that shapes axon and dendrite morphology.

• Drosophila Nab2, orthologue of the human RBP ZC3H14 lost in inherited intellectual disability, acts through unknown RNA targets to control axon and dendrite morphology.

• This study shows that Nab2 and the Mettl3 methyltransferase guide splicing of trio mRNA, which encodes a conserved GEF-domain protein. Loss of Nab2 is associated with an imbalance in levels of Trio GEF domains in Nab2-deficient neurons. Restoring this balance partially rescues neuronal defects.

• These findings suggest that Nab2 control of Trio levels is required to pattern axon and dendrite growth and suggests that ZC3H14 may play a similar role in the vertebrate brain.

INTRODUCTION

RNA-binding proteins (RBPs) associate with nascent RNA transcripts and govern expression via a multitude of mechanisms, including regulation of splicing, polyadenylation, nuclear export, translation, and stability (Maniatis and Reed, 2002; McKee and Silver, 2007; Corbett, 2018; Corley et al., 2020). These RNA–RBP interactions are particularly important in highly specialized cells such as neurons which require fine-tuned spatiotemporal control of gene expression to ensure proper development and function of the nervous system (Bardoni et al., 2012; Santoro et al., 2012; Conlon and Manley, 2017; Thelen and Kye, 2019; Gebauer et al., 2021). The importance of RBP function in neurons is highlighted by the prevalence of neurodevelopmental diseases that have been linked to defects in RBPs, leading to aberrant processing of RNAs encoding neurodevelopment factors (Cooper et al., 2009; Pak et al., 2011; Bardoni et al., 2012; Darnell and Richter, 2012; Gross et al., 2012; Santoro et al., 2012; Edens et al., 2015; Agrawal et al., 2019). Intriguingly, many of these RBPs are ubiquitously expressed and have roles in relatively common RNA processing mechanisms (Barbe et al., 1996; Franke et al., 1996; Brais, 2003; Lage et al., 2008; Kolb and Kissel, 2011; Pirozzi et al., 2011). Therefore, defining roles for these RBPs in neurons has become key to understanding why they are linked to neurological disease.

One important family of posttranscriptional regulatory proteins consists of polyadenosine binding proteins (Pabs) (Kelly et al., 2010). Conventional Pab family members bind polyadenosine RNA via RNA recognition motifs and modulate a multitude of RNA processing events such as splicing, export, polyadenylation, translation, and stability (Banerjee et al., 2013; Goss and Kleiman, 2013; Wigington et al., 2014). Another less well-studied group of Pabs utilize zinc finger (ZnF) domains to bind specific RNA motifs and modulate downstream processing events (Kelly et al., 2007; Kelly et al., 2010; Kelly et al., 2014). One such ZnF Pab termed zinc finger Cys-Cys-Cys-His-type containing 14 (ZC3H14; also termed MSUT2) is expressed ubiquitously and binds tracts of polyadenosine RNA with high affinity via tandem ZnF domains (Kelly et al., 2007; Leung et al., 2009; Kelly et al., 2010; Kelly et al., 2014; Wheeler et al., 2019). Despite ubiquitous expression, mutations in human ZC3H14 cause a form of inherited nonsyndromic autosomal recessive intellectual disability, which implies a specific requirement for ZC3H14 in the developing brain (Pak et al., 2011; Kelly et al., 2012).

The ZC3H14 protein is evolutionarily conserved among eukaryotes and has been studied in Mus musculus (Zc3h14) (Guthrie et al., 2011; Pak et al., 2011; Soucek et al., 2016), Caenorhabditis elegans (sut-2) (Guthrie et al., 2009), Sacchromyces cerevisiae (Nab2) (Anderson et al., 1993; Green et al., 2002; Hector et al., 2002; Marfatia et al., 2003; Kelly et al., 2007; Kelly et al., 2010; Schmid et al., 2015; Soucek et al., 2016), Sacchromyces pombe (Nab2) (Grenier St-Sauveur et al., 2013), and Drosophila melanogaster (Nab2) (Pak et al., 2011; Kelly et al., 2016; Fasken et al., 2019; Lee et al., 2020; Corgiat et al., 2021; Corgiat et al., 2022; Rounds et al., 2022; Jalloh et al., 2023). These studies have collectively uncovered molecular and neuronal functions for this conserved ZnF Pab. For example, Zc3h14 loss impairs working memory in mice where ZC3H14 protein localizes to synaptosomes in hippocampal neurons and regulates the abundance of synaptic proteins, including CaMK2α (Rha et al., 2017; Jones et al., 2020). Moreover, studies in C. elegans identified SUT-2 as a modulator of Tau-induced toxicity, as loss of sut-2 robustly rescues the toxic consequences of Tau overexpression in worms (Guthrie et al., 2009; Currey et al., 2023), a function of ZC3H14/MSUT2 that extends to mice (McMillan et al., 2021). On the other hand, yeast NAB2 is essential for viability (Anderson et al., 1993) and has critical functions in regulating transcription termination (Alpert et al., 2020), nuclear export (Hector et al., 2002), and transcript stability (Batisse et al., 2009; Schmid et al., 2015; Fasken et al., 2019; Alpert et al., 2020). Moreover, NAB2/Nab2 loss leads to increases in bulk poly(A) tail length in yeast, mice, and flies supporting a conserved function for Nab2 in restricting poly(A) tail length (Green et al., 2002; Hector et al., 2002; Kelly et al., 2010; Kelly et al., 2014). Taken together, these findings suggest that ZC3H14/Nab2 is involved in multiple aspects of posttranscriptional RNA metabolism and that these roles may be particularly significant in neurons.

D. melanogaster is a genetically tractable system in which to define molecular and developmental roles of the ZC3H14 invertebrate homologue, Nab2 (Pak et al., 2011). Our prior studies have determined that Nab2 function is necessary in neurons, as pan-neuronal expression of Drosophila Nab2 or human ZC3H14 is sufficient to rescue viability and locomotor defects associated with zygotic loss of Nab2 (Pak et al., 2011; Kelly et al., 2014). Moreover, Nab2 has a cell-autonomous role in Kenyon cells to pattern axonal projections from these cells into the mushroom bodies (Kelly et al., 2016), a twin neuropil structure that regulates Drosophila associative olfactory learning and memory (Heisenberg, 2003; Kahsai and Zars, 2011; Yagi et al., 2016; Takemura et al., 2017). Biochemical studies show that Nab2 interacts with Fmr1, the fly homolog of Fragile X Syndrome RBP, FMRP (Wan et al., 2000), and that these two RBPs coregulate mushroom body morphology and olfactory memory through a mechanism likely to involve translational repression of shared Nab2-Fmr1 target RNAs (Bienkowski et al., 2017). Beyond the brain, Nab2 limits dendritic branching of class IV dorsal dendritic arborization (ddaC) sensory neurons through a mechanism involving the planar cell polarity (PCP) pathway (Corgiat et al., 2022), suggesting that Nab2 controls RNA targets encoding regulators of the actin cytoskeleton. Our recent work studying the effect of Nab2 loss on the brain transcriptome revealed that Nab2 is required for proper splicing of ∼150 mRNAs (Jalloh and Lancaster et al., 2023). Furthermore, Nab2 limits methylation at the N-6 position of adenosine (m⁶A) on key mRNAs, including the alternatively spliced Sex-lethal (Sxl) transcript (Jalloh and Lancaster et al., 2023). However, Nab2-regulated transcripts encoding factors that guide axon and dendrite morphology have not been identified.

A recent study uncovered multiple Nab2-regulated candidate transcripts with key functions in neurodevelopment (Jalloh and Lancaster et al., 2023). Specifically, this work revealed significant retention of a 5′UTR intron in the trio transcript. Trio is a member of the Dbl homology (DH) family of guanine nucleotide exchange factor (GEF) proteins with well-conserved orthologues in C. elegans and mammals that control F-actin polymerization through the Rac and Rho small GTPases (Bellanger et al., 1998b; Awasaki et al., 2000; Bateman et al., 2000; Newsome et al., 2000; Bateman and Van Vactor, 2001; Briancon-Marjollet et al., 2008; Ba et al., 2016; Pengelly et al., 2016; Katrancha et al., 2017; Backer et al., 2018; Bircher and Koleske, 2021). As a result of these roles, Trio loss affects axon guidance and dendritic branching as well as synaptic transmission and plasticity (Briancon-Marjollet et al., 2008; Iyer et al., 2012; DeGeer et al., 2015; Ba et al., 2016; Katrancha et al., 2019). Notably, Drosophila Trio is enriched in the brain mushroom bodies where it controls axon projection and Trio also regulates arborization of sensory ddaC neurons in the larval peripheral nervous system (PNS) (Awasaki et al., 2000; Iyer et al., 2012; Shivalkar and Giniger, 2012). Moreover, several recent studies have identified loss- and gain-of-function mutations in the human TRIO gene and its paralogue KALRN that lead to genetically dominant forms of intellectual disability and neurodevelopmental disease (Ba et al., 2016; Pengelly et al., 2016; Katrancha et al., 2017; Paskus et al., 2020).

Trio contains two GEF domains, GEF1 and GEF2, that differentially activate Rac1 or RhoA/Rho1 GTPases, respectively (Debant et al., 1996; Bellanger et al., 1998a; Bellanger et al., 1998b). Trio-GEF1 activation of Rac1 regulates motor neuron axon guidance, cell migration, and axon outgrowth (Bateman et al., 2000; Newsome et al., 2000; Briancon-Marjollet et al., 2008; Peng et al., 2010; Song and Giniger, 2011). Comparatively, little is known about the function of Trio-GEF2; however, recent work suggests that it promotes growth cone collapse through RhoA/Rho1 (Backer et al., 2018). Supporting this model of opposing roles for Trio GEF1 and GEF2 function, studies in Drosophila ddaC neurons suggest that Trio promotes dendritic branching via GEF1 and restricts this process via GEF2 (Iyer et al., 2012). Despite insight into Trio GEF specificity for Rac and RhoA/Rho1, how these two opposing Trio activities are modulated within axons and dendrites remains unclear.

Here, we exploit both genetic and molecular approaches to assess the role of Nab2 and the m⁶A machinery in regulating expression of the neuronally enriched protein Trio in the adult fly brain. Consistent with our previous findings that Nab2 limits m⁶A methylation on specific transcripts, reduced levels of either Drosophila m⁶A reader protein—the nuclear reader Yt521-B or the cytoplasmic reader Ythdf—is sufficient to rescue Nab2^null viability and locomotion defects, indicating that m⁶A-mediated changes in RNA nuclear processing and cytoplasmic metabolism contribute to defects in Nab2 mutants. Focusing on the trio mRNA, we find that Nab2 and the m⁶A methyltransferase, Mettl3, each promote an intron-excision event within the 5′UTR of a trio mRNA species encoding only the GEF2 (RhoGEF) domain. Intriguingly, levels of the corresponding Trio-GEF2 protein drop in heads of Nab2^null but not Mettl3^null flies, consistent with a model in which Nab2 modulates both nuclear splicing and cytoplasmic metabolism of the GEF2-only variant of trio mRNA. Critically transgenic expression of Trio-GEF2 rescues axon projection defects in Nab2^null mushroom body neurons and class IV ddaC neurons while Trio-GEF1 has the opposite effect of exacerbating Nab2^null axon projection defects. Together, these data identify Nab2 and Mettl3 as key regulators of trio 5′UTR structure and provide evidence that altered expression of Trio-GEF2 contributes to axon and dendrite defects in Drosophila lacking Nab2.

RESULTS

Loss of m⁶A-reader proteins rescues Nab2^null defects in viability and adult locomotion

Nab2 loss causes severe defects in Drosophila viability, adult locomotion, and lifespan (Pak et al., 2011). Building on the previous finding that Nab2 loss elevates m⁶A methylation on select mRNAs (Jalloh and Lancaster et al., 2023), we hypothesized that some Nab2^null organismal phenotypes could result from ectopic recruitment of m⁶A reader proteins onto affected mRNAs. These m⁶A reader proteins recognize m⁶A-modified adenosines via a YTH domain (Xu et al., 2014) and act downstream of the methyltransferase machinery to bind and regulate the fate of methylated RNAs (Luo and Tong, 2014; Theler et al., 2014; Xu et al., 2014; Xu et al., 2015; Patil et al., 2018). Unlike more complex mammalian systems, Drosophila have a single nuclear m⁶A reader protein, YT-521-B (or Ythdc1) and a single cytoplasmic m⁶A reader protein, Ythdf (Haussmann et al., 2016; Kan et al., 2017).

To assess roles of nuclear Yt521-B and cytoplasmic Ythdf in Nab2 mutant phenotypes, the yt521-B^∆N and ythdf^∆YTH alleles (Lence et al., 2016; Worpenberg et al., 2021) were individually recombined with a zygotic Nab2^null allele (also known as Nab2^ex3; imprecise excision of EP3716) (Pak et al., 2011) and assessed for effects on viability, adult locomotion, and lifespan. Homozygous double mutant yt521-B^∆N/∆N, Nab2^null flies show increase viability compared with Nab2^null flies indicating that the nuclear m⁶A reader is required for the effect of Nab2 loss of viability (Figure 1A). The ythdf^∆YTH, Nab2^null double mutants are inviable; furthermore, heterozygous reduction of cytoplasmic Ythdf (ythdf^∆YTH/+, Nab2^null) does not improve Nab2^null viability (Figure 1A). In contrast, homozygous loss of nuclear Yt521-B has no effect on Nab2^null locomotion, whereas heterozygous reduction of cytoplasmic Ythdf rescues Nab2^null climbing rates by approximately 6-fold as assessed in a negative geotaxis assay (at the 30 s timepoint; Figure 1B). Despite the ability of reader alleles (e.g., yt521-B^∆N/∆N or ythdf^∆YTH/+) to rescue viability or adult locomotion, neither mutant alone rescues Nab2^null lifespan defects (Figure 1C). Together, these genetic rescue data provide evidence that neurological effects of Nab2 loss require nuclear and cytoplasmic m⁶A readers, and suggest that each of these mechanisms may involve different mRNAs.

FIGURE 1:

Loss of m⁶A-reader proteins rescues Nab2^null defects in viability and adult locomotion. (A) Percent of control, Nab2^null, yt521-B^∆N/∆NNab2^null, yhdf^∆YTH/+Nab2^null, yt521-B^∆N/df, and ythdf^∆YTH/+ that eclose as viable adults (calculated as #observed/#expected). At least 100 larvae were analyzed for each biological replicate. (B) Negative geotaxis as a measure of locomotion of age-matched adult flies of indicated genotypes over time in seconds (s) taken to reach the top of a vial. The negative geotaxis was observed for at least 10 animals for each biological replicate. (C) Survival of age matched adult flies of the indicated genotypes over time in days. The survival of at least 10 flies was observed for each biological replicate. At least three biological replicates were tested for each genotype and experiment. Significance values are indicated.

Nab2 and Mettl3 regulate splicing of the trio 5′UTR in the Drosophila head

In light of the effects of Nab2 loss on axon and dendrite development (Kelly et al., 2016; Bienkowski et al., 2017; Rounds et al., 2021; Corgiat et al., 2022), we mined our high-throughput RNA sequencing (RNA-seq) analysis of adult heads from Nab2^null mutants (zygotic null; imprecise excision of EP3761) and isogenic Controls (precise excision of EP3716) (Pak et al., 2011; Jalloh and Lancaster et al., 2023) to identify potential Nab2 target transcripts regulated by m⁶A with functions in neurodevelopment. One transcript identified in this analysis was trio, which encodes a Rho guanine nucleotide exchange factor (RhoGEF) that activates specific downstream Rho family GTPases (Bellanger et al., 1998b; Bircher and Koleske, 2021). There are multiple different variants of the trio transcript, two of which are readily detected in adult fly heads: hereafter referred to as trio Medium (trio M) and trio Long (trio L) (Figure 2A, top panel). Visualization of RNA-seq reads from Nab2^null and Control heads using Integrative Genomics Viewer (IGV) (Robinson et al., 2017) reveals an increase in reads in introns within the 5′UTR of both trio M and trio L in Nab2^null heads relative to Control (Figure 2A). Normal splicing patterns are detected across all other trio intron-exon junctions. Utilizing a publicly available me-RIP-Seq dataset from Drosophila heads (Kan et al., 2021), we bioinformatically identified three m⁶A sites in the trio M 5′UTR (Figure 2A, red lollipops), but none in the trio L 5′UTR. These data suggest that Nab2 is required for removal of 5′UTR introns in trio M and trio L and present the possibility that the removal of the trio M 5′UTR could also involve m⁶A.

FIGURE 2:

Nab2 and Mettl3 regulate splicing of the trio 5′UTR in the Drosophila head. (A) RNA-seq reads across the trio locus in Control and Nab2^null fly heads (Jalloh and Lancaster et al., 2023). Boxed insets highlight the sequencing reads from the trio L (scale bar = 70 nt) and trio M (scale bar = 145 nt) 5′UTR. Red lollipops denote location of mapped m⁶A sites. (B) Diagram of the trio M 5′UTR annotated to show location of color-coded primer pairs. The position of the ATG is also indicated (C) RT-PCR analysis of trio M mRNA from Control, Nab2^null, and Mettl3^null heads. Properly spliced transcript and intron reattaining transcript bands are indicated. Rpl32 serves as a control. qRT-PCR analysis detecting levels of (D) intron retaining or (E) properly spliced trio M transcript from Control, Nab2^null, and Mettl3^null heads, where control is set to 1.0. (F) Diagram of the trio L 5′UTR annotated to show location of color-coded primer pairs. (G) RT-PCR analysis of trio L mRNA from Control, Nab2^null, and Mettl3^null heads. Properly spliced and intron retaining transcript bands are indicated. (H) qRT-PCR analysis detecting levels of intron retaining and (I) properly spliced trio L transcript in Control, Nab2^null, and Mettl3^null heads. RNA from 25 heads was extracted from each genotype for each biological replicate. At least three biological replicates were performed for each experiment. Significance values are indicated.

To experimentally test this prediction, we first analyzed the trio M transcript using reverse transcription PCR (RT-PCR) analysis with primers that detect the trio M 5′UTR intron (exon 1-intron 1 and intron 1-exon 2) (Figure 2B, blue and orange primer pairs). This analysis reveals that the trio M 5′UTR intron-retaining transcript is enriched in adult heads of Nab2^null flies as well as in heads lacking Mettl3, the catalytic subunit of the methyltransferase complex (Figure 2C) with concomitant reduction or loss of properly spliced trio M 5′UTR (exon 1-exon 2) (Figure 2B, red primer pair; see Figure 2C). Primers that detect correctly spliced exon 1-exon 2 trio M transcript (Figure 2B, red primer pair) also amplify a ∼550 bp band in Nab2^null heads that is an aberrantly spliced product corresponding to the trio M pre-mRNA transcript (Figure 2C). This RT-PCR analysis did not detect the ∼4 kb trio M 5′UTR intron-retaining transcript, possibly due to the large size of the expected product. To quantitate these results, we performed reverse transcription quantitative PCR (qRT-PCR) analysis, which confirms a significant increase in the levels of the trio M 5′UTR intron-retaining transcript in both Nab2^null and Mettl3^null heads (Figure 2D). Reciprocal qRT-PCR analysis to quantify the levels of properly spliced trio M confirms reduced levels of the properly spliced transcript in Nab2^null heads and a complete loss of the properly spliced transcript in Mettl3^null heads (Figure 2E).

Shifting the analysis to the trio L 5′UTR intron using primers to detect exon 1-exon 2 confirms the presence of the trio L 5′UTR intron-retaining transcript (exon 1-intron 1-exon 2) in Nab2^null heads and reduced levels of the properly spliced transcript (exon 1-exon 2) (Figure 2F, green primer pair; Figure 2G). In contrast, only properly spliced trio L 5′ UTR is detected in heads of flies lacking Mettl3 (Figure 2G). qRT-PCR analysis confirms increased levels of the trio L 5′UTR intron-retaining transcript in Nab2^null, but not Mettl3^null heads, as compared with Control (Figure 2H). Reciprocally, qRT-PCR analysis using primers that detect levels of properly spliced trio L 5′UTR show reduced transcript levels in Nab2^null heads and no change in Mettl3^null heads compared with Control (Figure 2I). Collectively, these data confirm that splicing of the trio M and trio L 5′UTR introns are both Nab2-dependent, but that only splicing of the trio M 5′UTR and not the trio L 5′UTR is Mettl3 dependent.

Nab2 regulates levels of Trio M in the Drosophila head

The Drosophila Trio L protein is most similar to a form of human Trio protein (Trio A) that is enriched in the human brain and nervous system (Portales-Casamar et al., 2006). As illustrated in Figure 3A, Trio L contains a Sec14 domain, nine spectrin repeats, one Src homology 3 (SH3) domain, and two catalytic GEF domains comprised of tandem DH and pleckstrin homology (PH) domains, referred to as GEF1 and GEF2. The Drosophila Trio M protein corresponds to the C-terminal end of Trio L, including only the SH3 domain and the GEF2 catalytic domain (Figure 3A).

FIGURE 3:

Nab2 regulates Trio M protein level in the Drosophila head. (A) Schematic of the human (H.s.) Trio A and Drosophila melanogaster (D.m.) Trio L and Trio M proteins. Trio contains a Sec14 domain, nine spectrin repeats, one Src homology 3 (SH3) domain, and two catalytic GEF domains (GEF1 and GEF2), comprised of tandem DH and PH domains. Amino acid lengths are indicated for each Trio protein and the amino acids corresponding to Trio GEF1 and GEF2 are indicated in the schematic for Drosophila Trio L. (B) Immunoblotting analysis of Trio L and Trio M protein levels from Control, Nab2^null, and Mettl3^null heads. Lamin serves as a loading Control. Molecular weights in kDa are indicated to the left. (C) Quantification of Trio L (left) and Trio M (right) protein levels in B using Image Lab software. Protein levels are normalized to Control, with the value for Control set to 1.0. Protein from at least 25 heads was extracted from each genotype for each biological replicate. Three biological replicates were performed. Significance values are indicated.

Based on the finding that Nab2 regulates splicing of the trio L and trio M 5′UTR introns and that Mettl3 only regulates splicing of the trio M 5′UTR intron, we tested whether these intron retention events affect levels of Trio L or Trio M proteins in the fly head. Immunoblotting analysis reveals that Trio L and Trio M are the major isoforms of Trio in Control brains, and that Nab2 loss reduces levels of Trio M but has no apparent effect on Trio L protein levels relative to Control (Figure 3B). Densitometry analysis of the Trio L and Trio M protein levels demonstrates that the decrease in the steady-state level of Trio M protein is statistically significant (Figure 3C). Although loss of Mettl3 or Nab2 results in trio M 5′UTR intron retention (Figure 2, C–E), only loss of Nab2 causes a concomitant drop in Trio M protein levels (Figure 3, B and C). These results imply an independent effect of Nab2 on postsplicing metabolism of the trio M 5′UTR intron retaining mRNA, suggesting additional roles for Nab2 in the regulation of trio M translation, trafficking, or turnover.

Trio is altered in the Nab2^null mushroom body

Previous studies demonstrated that Trio is enriched in the mushroom bodies (Awasaki et al., 2000), which are divided into five lobes per hemisphere (α/α’, β/β’, and γ) that project anteriorly from the dorsally located Kenyon cells (Figure 4A) (Heisenberg, 1998; Roman and Davis, 2001). The α/α’ lobes project dorsally, while the β/β’ and γ lobes project medially, toward the central ellipsoid body (EB) (Figure 4A). As demonstrated previously, loss of Nab2 causes defects in α and β lobe structures, specifically loss or thinning of the α lobes and midline fusion of the β lobe structures (Kelly et al., 2016; Corgiat et al., 2022; Rounds et al., 2022) (Figure 4B). To visualize Trio in these structures, we stained Control brains overexpressing membrane tethered GFP in α/β/γ lobes (201Y-Gal4, UAS-mcd8::GFP) with an α-Trio antibody, which recognizes both Trio L and Trio M protein (Awasaki et al., 2000) (Figure 4B). This analysis confirms that Trio is enriched in γ lobes, Kenyon cell bodies, and the calyx as shown by colocalization with GFP, but absent or below the level of detection in α and β lobes (Awasaki et al., 2000) (Figure 4B; Supplemental Figure S1). Intriguingly, Trio accumulates in dysplasic axons near the midline that are not labeled with the 201Y-Gal4 driver (which exclusively drives mcd8::GFP expression in α/β/γ lobes), suggesting these dysplasic axons belong to the β’ lobe (Figure 4B, middle panel, bottom row).

FIGURE 4:

Trio is altered in the Nab2^null mushroom body. (A) Diagram of the adult Drosophila mushroom body lobes depicting axons of the medially projecting gamma (γ) neurons, the vertical alpha (α) and alpha prime (α’) neurons, the medially projecting beta (β) and beta prime (β’) neurons, and the EB. (B) Immunofluorescence images of Control (w-;UAS-mcd8::GFP/201Y-Gal4;;) and Nab2^null (w-;UAS-mcd8::GFP/201Y-Gal4;Nab2^null;) mushroom bodies driving UAS-mcd8::GFP under the α, β, γ lobe-specific mushroom body 201Y-Gal4 driver. (C) Immunofluorescence images of Control (w-;UAS-mcd8::GFP/C305a-Gal4;;) and Nab2^null (w-;UAS-mcd8::GFP/C305a-Gal4;Nab2^null;) mushroom bodies driving UAS-mcd8::GFP under the α’ and β’ lobe-specific mushroom body C305a-Gal4 driver. False colored panels show fluorescence corresponding to α-GFP (green, mcd8::GFP), α-trio (purple), and merges of the channels. At least 25 brains were analyzed for each genotype. Scale bar = 50 μm.

We extended this analysis to explore how loss of Nab2 impacts Trio localization. Trio is lost in the GFP-positive γ lobes of Nab2^null brains (Figure 4B, bottom row). Given the strong reduction of Trio M protein detected by immunoblotting of Nab2^null heads (see Figure 2B), this result suggests that Trio M may be the primary isoform of Trio present in mushroom body γ lobes.

To explore the localization of Trio specifically in the α’ and β’ lobe structures, we utilized the Gal4-UAS system to overexpress GFP (UAS-mcd8::GFP) using a prime lobe specific Gal4 driver (Cka^C305a-Gal4; exclusively drives mcd8::GFP expression in α’and β’ lobes) (Figure 4C). As reported previously, Trio is enriched in α’ and β’ axons, and is also detected in γ lobe axons and EB of Control brains (Awasaki et al., 2000) (Figure 4C, top row). Nab2^null brains show complete loss or thinning of the α’ lobe axons as well as a distinct defasciculation phenotype in the β’ lobe structures (Figure 4C). These morphological phenotypes are accompanied by Trio loss in Nab2^nullγ lobe axons and EB, and Trio accumulation in the distal portion of β’ lobe axons closest to the brain midline (Figure 4C, middle panel, bottom row).

Expression of Trio GEF2 rescues α/α’ and β’ defects in Nab2^null mushroom bodies

The reduced level of Trio M protein in Nab2^null heads (Figure 3B) raises the possibility that an imbalance in the relative dose of Trio-GEF1 and Trio-GEF2 activities contributes to mushroom body morphology defects observed in Nab2^null flies (Kelly et al., 2016). This model is based on the established role of the Trio protein in patterning of axons in the mushroom body (Awasaki et al., 2000), and predicts that loss of Trio M lowers Trio-GEF2 activity within specific mushroom body lobes.

To test this model, transgenes encoding specifically the Trio-GEF1 (UAS-Trio-GEF1; Figure 3A, AAs 1287–1583 tagged at the 3′end with eight copies of the tag MYC) or Trio-GEF2 domain (UAS-Trio-GEF2; Figure 3A, AAs 1945–2246 tagged at the 3′end with eight copies of the tag MYC) (Awasaki et al., 2000; Newsome et al., 2000), were expressed in all mushroom body lobes of Nab2^null brains using the OK107-Gal4 driver. As previously reported (Kelly et al., 2016; Bienkowski et al., 2017; Corgiat et al., 2021; Corgiat et al., 2022; Rounds et al., 2022) and shown in Figures 5, A–C, Nab2^null mutant brains have highly penetrant defects in the structure of the mushroom body α lobes (missing or thinned) and β lobes (missing, thinned, or midline crossed) as detected by α-FasII staining, which specifically recognizes the α, β and weakly the γ lobes (Crittenden et al., 1998). Transgenic expression of Trio-GEF2 alone has no effect on mushroom body structure in a control background. Intriguingly, transgenic expression of Trio-GEF2 does not rescue Nab2^null β lobe defects, but significantly suppresses Nab2^null α lobe defects (Figure 5, A–C). In contrast, mushroom body expression of Trio-GEF1 in control brains causes complete loss of axonal projection from the Kenyon cells (no lobe structures were detected in any of the brains analyzed; n = 25) and is completely lethal in a Nab2^null background (Figure 5, A–C). Thus, Nab2 loss sensitizes mushroom body axons to the dose of Trio GEF domains such that expression of Trio-GEF2 rescues α lobe axons and expression of extra Trio-GEF1 is lethal to the animal; neither effect is observed in Control brains, indicative of a tight link between Nab2 and Trio GEF dosage in the developing mushroom body.

FIGURE 5:

Expression of Trio GEF2 rescues α/α’ and β’ defects in Nab2^null mushroom bodies. (A) Top: Schematic of the Drosophila mushroom body with lobes stained by α-Fasciclin II (FasII) highlighted. Bottom: Representative max projections of mushroom bodies of indicated genotypes stained by α-FasII. (B) Quantification of frequency of α lobe defects in each indicated genotype. (C) Quantification of frequency of β lobe defects in indicated genotypes. (D) Top: Schematic of the Drosophila mushroom body with lobes overexpressing mcd8::GFP under the C305a-Gal4 driver highlighted. Bottom: Representative max projections of mushroom bodies of indicated genotypes stained with α-GFP. (E) Quantification of frequency of α’ lobe defects in each indicated genotype. (F) Quantification of frequency of β’ lobe defects in indicated genotypes. (G) Quantification of defasciculation phenotype severity in the indicated genotypes. At least 25 brains were analyzed for each genotype. Scale bar = 50 μm.

Given the enrichment of Trio protein in the α’ and β’ lobes (see Figure 4B), we also tested whether transgenic expression of the Trio GEF1 or GEF2 domain could autonomously rescue Nab2^null α’ or β’ lobe morphology as visualized with the Cka^C305a-Gal4 driver (with UAS-mcd8::GFP). As with Trio-GEF1 or Trio-GEF2 expression in the α, β, and γ lobes using the 201Y-Gal4 driver, expression of Trio-GEF2 alone has no effect on α’ and β’ lobe morphology in Control brains but rescues Nab2^null defects in α’ lobe structure and strongly reduces β’ lobe defasciculation (Figure 5, D–G). Expression of Trio-GEF1 in α’ and β’ lobes has the inverse effect of late larval lethality in both Control and Nab2^null animals (Figure 5D). These data are consistent with a model in which Nab2 acts through Trio-GEF2 to guide axon projection and fasciculation in the α’ and β’ lobes.

Expression of Trio GEF2 rescues Nab2^null dendritic arborization defects in class IV ddaC sensory neurons

Nab2 normally restricts branching of sensory dendrites in larval class IV ddaC neurons in body wall neurons (Corgiat et al., 2022). Significantly, Trio-GEF1 promotes and Trio-GEF2 restricts branching of dendrites from these same class IV ddaC neurons (Figure 6, A and B) (Iyer et al., 2012). Thus, we reasoned that an imbalance of Trio-GEF1 and Trio-GEF2 activities in Nab2^null ddaC neurons might contribute to dendritic defects. To test this model, we employed a pickpocket (ppk)-Gal4,UAS-mcd8::GFP system to visualize class IV ddaC cell bodies and dendritic trees. As observed previously, loss of Nab2 (Corgiat et al., 2022) or transgenic expression of Trio-GEF1 (Iyer et al., 2012) individually increases dendritic branch complexity, while transgenic expression of Trio-GEF2 reduces dendritic branch complexity (Iyer et al., 2012) (Figure 6, A and B). Significantly, combining Nab2 loss (Nab2^null) with ddaC-specific expression of Trio-GEF2 rescues overarborization normally observed in Nab2^null larvae (Figure 6, A and B). Intriguingly, transgenic expression of Trio-GEF1 in Nab2^null larvae causes an increase the variance between arborization phenotypes among the neurons analyzed (Figure 6B). Together, these data indicate that loss of Trio-GEF2 in Nab2^null larvae likely contributes to Nab2^null ddaC overarborization defects.

FIGURE 6:

Expression of Trio GEF2 rescues Nab2^null dendritic arborization defects in class IV ddaC sensory neurons. (A) Maximum intensity projections of Drosophila class IV ddaC neurons from Control, Nab2^null, Nab2^null UAS-Trio-GEF2, Nab2^null UAS-Trio-GEF1, UAS-Trio GEF2, and UAS-Trio-GEF1 L3 larvae. Inset black boxes show high magnification views of dendritic arbors. (B) Quantification of total number of dendritic terminals for each genotype. At least eight neurons were measured for each genotype. Significance values are indicated.

Expression of Trio GEF2 rescues Nab2^null defects in viability and locomotion

Given the genetic interactions between the UAS-Trio-GEF1 and UAS-Trio-GEF2 transgenes and Nab2, we assessed whether transgenic expression of these distinct Trio GEF domains in the mushroom body (OK107-Gal4, UAS-Trio-GEF1 or UAS-Trio-GEF2) affects the organismal phenotypes of viability, adult locomotion, and lifespan. Confirming our previous findings (Pak et al., 2011), we detected severe reductions in viability, adult locomotion, and lifespan in Nab2^null flies compared with Control (Figure 7, A–C). As previously noted, expression of Trio-GEF1 in mushroom bodies is lethal at late larval stages (Figure 7A). In contrast, transgenic expression of Trio-GEF2 in mushroom bodies strongly suppresses Nab2^null defects in viability and suppresses defects in adult locomotion but does not significantly alter lifespan (Figure 7, A–C), indicating that a deficit in Trio-GEF2-regulated cytoskeletal dynamics within OK107-expressing brain neurons contributes to developmental and postdevelopmental defects in Drosophila lacking Nab2.

FIGURE 7:

Expression of Trio GEF2 rescues Nab2^null defects in viability and locomotion. (A) Percent of Control, Nab2^null, Nab2^null UAS-Trio-GEF2, UAS-Trio GEF2, or UAS-Trio-GEF1 that eclose as viable adults (calculated as #observed/#expected) using the OK107-Gal4 mushroom body driver. (B) Negative geotaxis of age-matched adult flies of indicated genotypes over time in seconds (s). (C) Survival of age matched adult flies of the indicated genotypes over time in days. At least three biological replicated were tested for each genotype. Significance values are indicated.

DISCUSSION

Here, we identify a role for the Drosophila Nab2 RBP and Mettl3 m⁶A methyltransferase in regulating the trio mRNA, which encodes a conserved RhoGEF protein that is altered in human intellectual disability and regulates axon guidance and dendritic arborization through two GEF domains that individually control the cytoskeletal regulators Rac and RhoA/Rho1 (Debant et al., 1996; Bellanger et al., 1998a; Bellanger et al., 1998b; Bateman and Van Vactor, 2001; Jaffe and Hall, 2005; Briancon-Marjollet et al., 2008; Iyer et al., 2012; Bircher and Koleske, 2021). The data presented provide strong evidence that the trio transcript is a key downstream target of Nab2 in neurons based on an m⁶A- and Nab2-dependent splicing event and identifies specific effects of each Trio GEF domain within axons and dendrites that develop from neurons lacking Nab2. The results of this study combine with our previous work (Corgiat et al., 2022; Jalloh and Lancaster et al., 2023) to support a model in which Nab2 regulates transcripts that encode key regulators of neurodevelopment, including the conserved GEF, Trio. In the broader context, the phenotypic consequences of loss of an RBP result from the collective changes to numerous target transcripts, and defining the mechanistic basis of these phenotypes requires systematic analysis of how individual targets, such as trio, are impacted. In the case of Nab2, evidence now supports both m⁶A-dependent and independent roles in trio mRNA splicing as well as potential effects on downstream processing such as cytoplasmic metabolism of trio mRNA. Taken together, these findings support a model where RBPs such as Nab2/ZC3H14 regulate a collection of target transcripts, potentially through multiple mechanisms, that all contribute to downstream phenotypes.

Previous work illustrating broad rescue of Nab2^null phenotypes by Mettl3 heterozygosity (Jalloh and Lancaster et al., 2023) suggested that other regulators of m⁶A-modified transcripts could also contribute to Nab2^null defects in viability, adult locomotion, and lifespan. Here, we demonstrate that loss of nuclear or cytoplasmic m⁶A reader function rescues some, but not all, organismal phenotypes associated with loss of Nab2 (Figure 1). Homozygous loss of the nuclear m⁶A reader YT521-B suppresses Nab2^null defects in viability, but not locomotion or lifespan, whereas heterozygous loss of the cytoplasmic m⁶A reader Ythdf dominantly suppresses Nab2^null defects in locomotion, but not viability or lifespan. Collectively, these data suggest that mRNA targets of Nab2 responsible for these behavioral phenotypes may be differentially regulated between cell types and in an m⁶A-dependent manner. For instance, Nab2-regulated transcripts encoding proteins that govern Drosophila viability could rely more heavily on nuclear m⁶A regulatory mechanisms, such as splicing or export. On the other hand, Nab2-regulated transcripts encoding proteins that govern Drosophila negative geotaxis could more heavily require cytoplasmic m⁶A regulatory mechanisms, such as translation or stability. Moreover, the inability of YT521-B or Ythdf loss to rescue Nab2^null defects in lifespan suggests that Nab2-regulated transcripts that govern lifespan may not by modified by m⁶A, or that Nab2 plays m⁶A-independent roles in regulating transcripts critical to control lifespan.

Previous studies demonstrated that loss of m⁶A regulatory proteins disrupts axon projection in the Drosophila mushroom body (Kan et al., 2021; Worpenberg et al., 2021). Although Mettl3 heterozygosity broadly rescues Nab2^null behavioral defects (Jalloh and Lancaster et al., 2023), a decrease in Mettl3 does not dominantly suppress mushroom body morphology defects (Supplemental Figure S2). This finding suggests that Mettl3 heterozygosity is insufficient to reduce m⁶A methylation on Nab2-target transcripts to a degree necessary for rescue of axonal projection defects. Given the ability of m⁶A writer and reader alleles to broadly rescue Nab2^null phenotypes, future studies will aim to further define the relationship between m⁶A machinery and Nab2 in relation to regulation of Drosophila mushroom body morphology. Moreover, a genome-wide approach to assess transcriptomic changes in m⁶A in Nab2^null flies will help delineate the neuronal, Nab2-regulated transcripts that exhibit changes in m⁶A methylation.

A number of trio transcript variants exist in the Drosophila brain (Awasaki et al., 2000). Here, we demonstrate that both Nab2 and Mettl3 are required for proper splicing of the 5′UTR intron of trio M. On the other hand, splicing of the 5′UTR intron of trio L is dependent on Nab2 but not Mettl3. These data align with previously published RNA-seq data from Nab2^null Drosophila heads (Jalloh and Lancaster et al., 2023), as well as a publicly available mi-CLIP-seq dataset that mapped m⁶A sites in the trio M 5′UTR intron, but not in the trio L 5′UTR intron (Kan et al., 2021). Notably, our data reveal that disruptions in splicing of the trio 5′UTR due to loss of Nab2 or Mettl3 do not correspond with perturbations in steady-state levels of Trio L and Trio M proteins. These results show that the intron retention event in the 5′UTR of trio M is associated with a significant reduction in the level of Trio M protein in the Nab2^null fly head. Surprisingly, the steady-state level of Trio M protein is unaffected in Mettl3^null heads even though trio 5′UTR intron retention is comparable to the levels observed in Nab2^null heads. These data imply that Nab2 may regulate trio M protein levels in a manner independent of trio M 5′UTR splicing. However, given previously defined roles for Nab2 as a negative regulator of m⁶A methylation (Jalloh and Lancaster et al., 2023), this observation more likely suggests that excess m⁶A on the trio M pre-mRNA upon loss of Nab2 may disrupt subsequent translation, trafficking, or stability. In contrast, retention of the trio L 5′UTR intron that occurs upon loss of Nab2 does not affect the steady-state level of the Trio L protein, suggesting this intron retention event may not disrupt translation. Alternatively, the remaining level of the properly spliced trio L in Nab2^null heads may be sufficient to maintain the steady-state level of Trio L protein. Further studies expanding to additional coregulated Nab2 and Mettl3 targets will be required to define how these factors regulate posttranscriptional events.

Given that Nab2 regulates splicing and downstream processing of numerous neuronally-enriched mRNAs (Jalloh and Lancaster et al., 2023), morphological defects observed in Nab2^null mushroom bodies are likely due to collective processing defects that affect multiple RNAs critical for axon development. Our results support a model where loss of Trio M, and therefore GEF2 function, contributes to morphological defects in mushroom bodies of Nab2^null flies. Previous studies demonstrated that Trio is a critical regulator of mushroom body morphology and Trio is enriched in the α’, β’, and γ lobes, but is virtually absent in the α and β lobes (Awasaki et al., 2000). We confirm these findings and further demonstrate that upon loss of Nab2, Trio levels are depleted in the γ lobes. Despite the established functions of Trio in regulating γ lobe formation (Awasaki et al., 2000), γ lobe defects are not detected upon loss of Nab2 (Kelly et al., 2016). Given that Trio M is the only isoform of Trio depleted in Nab2^null brains (Figure 3B) and γ lobes show no defects (Figure 5A), this finding suggests that Trio L, and therefore GEF1 function, is responsible for patterning γ lobe axons in the developing brain.

Studies of axon pathfinding mechanisms in the mushroom body demonstrate that the α’ and β’ lobes guide development of the α and β lobes (Fushima and Tsujimura, 2007). Given that Trio is enriched in the α’/β’/γ lobes of the mushroom body, our data suggest that loss of Trio M, and therefore GEF2 levels, in α’/β’ lobes contributes to Nab2^null α/β lobe defects. We demonstrate that transgenic expression of Trio-GEF2 in α’/β’ lobes (C305a-Gal4) of Nab2^null mushroom bodies rescues α’ lobe defects and β’ lobe defasciculation phenotypes in a cell autonomous manner (Figure 5). Moreover, expression of Trio-GEF2 in all mushroom body lobes (OK107-Gal4) rescues Nab2^null α lobe defects; however, whether this rescue occurs in a cell autonomous manner remains unknown (Figure 5). Collectively, these data demonstrate that Trio-GEF2 rescue is limited to some mushroom body lobes (α, α’, β’) and not others (β), implying that Trio-GEF2 is required for projection of only some Nab2^null axons. In this regard, misprojection defects in Nab2^null β axons are not rescued by multiple genetic manipulations that rescue α axon defects (e.g., by transgenic expression of Trio-GEF2 [this study] or by single copy alleles of fmr1, Atx2, or PCP components [Kelly et al., 2016; Bienkowski et al., 2017; Rounds et al., 2021; Corgiat et al., 2022]). These findings imply specific roles for Nab2 in these two types of Kenyon cell projections and suggest that Nab2 regulates different mRNAs to govern development of distinct mushroom body lobes.

Previous studies established that the Trio-GEF1 domain acts primarily through activation of Rac1 to promote axon outgrowth and pathfinding, while Trio-GEF2 acts primarily through RhoA/Rho1 to restrict neurite outgrowth (Leeuwen et al., 1997; Iyer et al., 2012; Bircher and Koleske, 2021). Given that loss of Nab2 disrupts the ratio of GEF1 and GEF2 in Drosophila heads by decreasing the level of Trio M but not Trio L, we hypothesized that expression of the GEF1 effector, Rac1, in mushroom body neurons would exacerbate Nab2^null phenotypic and morphological defects, whereas expression of the GEF2 effector, RhoA/Rho1, would rescue these same defects. Intriguingly, we observed that expression of either Rac1 or RhoA/Rho1 in the absence of Nab2 in Trio-enriched mushroom body neurons is lethal (Supplemental Figure S3). These data suggest that further expression of Rac1 in Nab2^null flies in which GEF1 levels, and therefore likely Rac1 activation, dominates is detrimental to nervous system development and the lethality induced by expression of RhoA/Rho1 upon loss of Nab2, indicates that Trio-GEF2 may act via other unknown effectors to govern mushroom body development. On the other hand, alternative RhoGEFs may compensate for the loss of Trio M protein in the mushroom body thereby leading to the lethality of Nab2^null animals transgenically expressing Rac1 or RhoA/Rho1. In sum, further studies are required to elucidate how loss of Nab2 alters Rac1 and RhoA/Rho1 activity in the Drosophila brain.

Nab2 and Trio have established roles in sculpting dendritic arborization of class IV ddaC neurons in the Drosophila PNS (Iyer et al., 2012; Corgiat et al., 2022). Here, we demonstrate that transgenic expression of the Trio-GEF2 domain rescues overarborization defects in Nab2^null class IV ddaC neurons. In line with previous studies (Iyer et al., 2012), we validate that transgenic expression of Trio-GEF1 in class IV ddaC neurons causes dramatic overarborization defects, while transgenic expression of Trio-GEF2 causes underarborization defects compared with control animals. Interestingly, we also demonstrate that expression of Trio-GEF1 in class IV ddaC neurons of Nab2^null flies results in a wide range of arborization phenotypes. Very few of these animals survive to the wandering third instar larval stage and no animals survive to adulthood. Given these observations, disruption of Nab2-regulated mRNAs in these neurons as well as overactivation of Rac1 by GEF1 may severely disrupt ddaC development such that arborization defects are highly variable from animal-to-animal. Overall, these data support a role for Trio M, and therefore GEF2 loss, in contributing to the established overarborization defects in Nab2^null class IV ddaC neurons (Corgiat et al., 2022).

In aggregate, these data reveal a role for Nab2 and Mettl3 in regulating splicing and protein levels of the RhoGEF Trio to support proper nervous system development. Genetic interactions between the m⁶A machinery and Nab2 support a role for Nab2 in the regulation of m⁶A methylation. We show for the first time that loss of Trio M, and therefore GEF2 levels, in Nab2^null flies contributes to several Nab2^null defects, including neuronal defects, such as mushroom body morphology and class IV ddaC arborization. Moreover, we demonstrate that transgenic expression of Trio-GEF2 broadly rescues Nab2^null viability and adult locomotion. This regulatory relationship between Nab2 and Trio-GEF2 could be cell autonomous, but further experiments are required determine the nature of this interaction. Given that mutations in human ZC3H14 and TRIO are both linked to intellectual disabilities (Pak et al., 2011; Ba et al., 2016), dysregulation of Trio function in neurons is one potential mechanism to explain axonal and dendritic phenotypes observed in Nab2^null Drosophila (Pak et al., 2011; Kelly et al., 2016; Bienkowski et al., 2017; Corgiat et al., 2021; Corgiat et al., 2022; Jalloh and Lancaster et al., 2023) and Zc3h14 mutant mice (Jones et al., 2020), as well as the cognitive defect observed in human patients lacking ZC3H14 (Pak et al., 2011).

MATERIALS AND METHODS

Request a protocol through Bio-protocol.

Resources table

Reagent or resource	Source	Identifier
Antibodies
Mouse monoclonal 9.4A anti-Trio	University of Iowa Developmental Studies Hybridoma Bank	RRID:AB_528494
Mouse monoclonal ADL101 anti-Lamin	University of Iowa Developmental Studies Hybridoma Bank	RRID:AB_528332
Mouse monoclonal 1D4 anti-Fasciclin II	University of Iowa Developmental Studies Hybridoma Bank	RRID: AB_528235
Rabbit Polyclonal Anti-Green Fluorescent Protein (GFP)	Thermo Fisher Scientific	Catalogue no. A-11122
Alexa Fluor 488 AffiniPure Polyclonal Goat Anti-rabbit IgG	Jackson ImmunoResearch Laboratories	RRID: AB_2338046
Cy3 AffiniPure Polyclonal Goat Anti-Mouse IgG	Jackson ImmunoResearch Laboratories	RRID: AB_2338690
Chemicals, peptides, and recombinant proteins
TRIzol reagent	Invitrogen	Catalogue no. 15596018
1-Bromo-3-chloropropane	Scientific Laboratory Supplies	Catalogue no. B9673
2-propanol/isopropanol	Thermo Fisher Scientific	Catalogue no. A416-1
Ethanol 200 proof	Thermo Fisher Scientific	Catalogue no. 04-355-233
Diethyl pyrocarbonate (DEPC)	Sigma-Aldrich	Catalogue no. D5758
Agarose LE, Quick Dissolve	Genesee Scientific	Catalogue no. 20-102QD
Red Safe	iNtRON Biotechnology	Catalogue no. 21141
NaCl	Sigma-Aldrich	Catalogue no. S7653
KCl	Sigma-Aldrich	Catalogue no. P3911
ZnCl2	Sigma-Aldrich	Catalogue no. 208086
Polysorbate 20 (Tween 20)	Fisher Bioreagents	Catalogue no. BP-337-500
Triton X-100	Sigma-Aldrich	Catalogue no. T8787
Tris Base Ultrapure	USBiological Life Sciences	Catalogue no. T8600
Sodium Dodecyl Sulfate	Fisher Bioreagents	Catalogue no.YBP166500
Sodium deoxycholate	Thermo Fisher Scientific	Catalogue no. 89905
NP-40	Thermo Fisher Scientific	Catalogue no. 85124
Dithiothreitol (DTT)	Thermo Fisher Scientific	Catalogue no. R0861
Bromophenol Blue	Thermo Fisher Scientific Chemicals	Catalogue no. A18469-18
Glycerol	USBiological Life Sciences	Catalogue no. G8145
RNaseOUT	Invitrogen	Catalogue no. 10777019
p-Coumaric acid	Sigma-Aldrich	Catalogue no. C9008
Luminol	Sigma-Aldrich	Catalogue no.A8511
Acetic Acid, Glacial	Thermo Fisher Scientific	Catalogue no. A38-212
EDTA	USBiological Life Sciences	Catalogue no. E2210
Paraformaldehyde	Electron Microscopy Sciences	Catalogue no. 15713
VECTASHEILD mounting medium	Vector Laboratories Inc.	REF#H-1000
Normal Goat Serum	Jackson ImmunoResearch Laboratories	RRID:AB_2336990
Diethyl Ether Anhydrous (stabilized with BHT)	Tokyo Chemical Industry (TCI) America	Catalogue no. D3497
Halocarbon oil 27	Sigma-Aldrich	Catalogue no. H8773
Critical commercial assays
DNase I, Amplification Grade	Invitrogen	Catalogue no. 18068015
M-MLV Reverse Transcriptase	Invitrogen	Catalogue no. 28025013
Taq DNA Polymerase (1000U)	Qiagen	Catalogue no. 201205
QuantiTect SYBR Green PCR Kit	Qiagen	Catalogue no. 204145
MicroAmp Fast Optical 96-Well Reaction Plate with Barcode, 0.1 ml	Applied Biosystems	Catalogue no. 4346906
MicroAmp Optical Adhesive Film	Applied Biosystems	Catalogue no. 4311971
4–20% Mini-PROTEAN TGX Stain-Free™ Protein Gels	Bio-Rad	Catalogue no. 4568093
Nitrocellulose Membrane, 0.2 µm	Bio-Rad	Catalogue no. 1620112
iScript Reverse Transcriptase Supermix	Bio-Rad	Catalogue no. 1708841
Phusion High Fidelity PCR	Thermo Fisher Scientific	Catalogue no. F530S
Pierce BCA Protein Assay Kit	Thermo Fisher Scientific	Catalogue no. 23225
SuperFrost Plus slides	Thermo Fisher Scientific	Catalogue no. 12-550-15
Il90p-Experimental models: Organisms/Strains
D. melanogaster: w-;;Nab2pex41 (Control);	(Pak et al., 2011)	N/A
D. melanogaster: w-;;Nab2ex3 (Nab2null); (Zygotic loss of Nab2)	(Pak et al., 2011)	N/A
D. melanogaster: w-;;Mettl3null;	(Lence et al., 2016)	N/A
D. melanogaster:YT521-B∆N	(Lence et al., 2016)	N/A
D. melanogaster: Ythdf∆YTH (Ythdf0)	(Worpenberg et al., 2021)	N/A
D. melanogaster:w1118;;Df(3R)BSC655 (ythdf deficiency)	Bloomington Drosophila Stock Company	BDSC:26507
D. melanogaster: y1w67c23;UAS-mcd8::GFP, 201Y-Gal4;;	Bloomington Drosophila Stock Company	BDSC:64296
D. melanogaster:w*;Ckac305a-Gal4	Bloomington Drosophila Stock Company	BDSC:30829
D. melanogaster: y1w*;UAS-mcd8::GFP;;	Bloomington Drosophila Stock Company	BDSC:5137
D. melanogaster: w*;;;OK107-Gal4	Bloomington Drosophila Stock Company	BDSC:854
D. melanogaster:y1w*:UAS-trio-GEF1-myc(Trio-GEF1 domain; AAs 1287-1583 tagged at the 3′end with eight copies of the tag MYC)	Bloomington Drosophila Stock Company	BDSC:9133
D. melanogaster: y1w*:UAS-trio-GEF2-myc(Trio-GEF2 domain; AAs 1945-2246 tagged at the 3′end with eight copies of the tag MYC)	Bloomington Drosophila Stock Company	BDSC:9134
D. melanogaster:w*;;ppk-Gal4	Bloomington Drosophila Stock Company	BSDC:32079
D. melanogaster:w*;UAS-Rac1	Bloomington Drosophila Stock Company	BSDC:6293
D. melanogaster:w*;UAS-Rho1	Bloomington Drosophila Stock Company	BSDC:58819
Oligonucleotides
RT-PCR trio L Forward: AACAAAACAGAGAGCGCCC	This study	N/A
RT-PCR trio L Reverse: GATGGGCACTGCAGCATAA	This study	N/A
qRT-PCR trio L Intron Retention Forward: 5′-TTAGCCCGCGTCAAGTC-3′	This study	N/A
qRT-PCR trio L Intron Retention Reverse: 5′-CTGCTTGTGCCACCAAAT-3′	This study	N/A
qRT-PCR trio L Properly Spliced Forward: 5′-GTTGTGTTGACAAAAGAGTG-3′	This study	N/A
qRT-PCR trio L Properly Spliced Reverse: 5′-GATGGGCACTGCAGCATAA-3′	This study	N/A
RT-PCR trio M Intron Retention (1) Forward: 5′-CAGCAGTCTCTTCTTCACTAA-3′	This study	N/A
RT-PCR trio M Intron Retention (1) Reverse: 5′-ACTCGGATTGTTGTTTCACTTT-3′	This study	N/A
RT-PCR trio M Intron Retention (2) Forward: 5′-GACTGCGCAAACATAGCATTA-3′	This study	N/A
RT-PCR trio M Intron Retention (2) Reverse: 5′-ATCCGCTCGTTGAGAAACT-3′	This study	N/A
RT-PCR trio M Properly Spliced Forward: 5′-CGCTAAAGAGGAGCGCAATA-3′	This study	N/A
RT-PCR trio M Properly Spliced Reverse: 5′-CTTCTTAACACTCTTCATGATTCG-3′	This study	N/A
qRT-PCR trio M Intron Retention Forward: 5′-TTGAGTGAACCCGCTAAAG-3′	This study	N/A
qRT-PCR trio M Intron Retention Reverse: 5′-CTTTGGAGTGCTTGTTCTTATC-3′	This study	N/A
qRT-PCR trio M Properly Spliced Forward: CGTCCATAAATTGAGTCGGAGAAC	This study	N/A
qRT-PCR trio M Properly Spliced Reverse: 5′-CTTCTTAACACTCTTCATGATTCG-3′	This study	N/A
RT-PCR and qRT-PCR rpl32 Forward: AAGATGACCATCCGCCCAGCATAC	This study	N/A
RT-PCR and qRT-PCR rpl32 Reverse: ACGCACTCTGTTGTCGATACCCTT	This study	N/A
Software and algorithms
Fiji/ImageJ	(Schindelin et al., 2012; Rueden et al., 2017)	https://imagej.net/
Integrated Genome Viewer (IGV)	(Robinson et al., 2011)	https://igv.org/
Image Lab	Bio-Rad	https://image-lab-4-0.software.informer.com/
GraphPad (Prism)
Other
ChemiDoc	Bio-Rad	Catalogue no. 12003153
Humidified Incubators	Shel Lab	SRI20PF
#5 Dumont Fine Forceps	Ted Pella Inc.	Prod#5622
Motorized Pestle	Argos Technologies	Catalogue no. A0001
NanodropOne	Thermo Fisher Scientific	N/A

Resource availability

Lead contact.

Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Ken Moberg (kmoberg@emory.edu).

Materials availability

The D. melanogaster lines generated in this study are available by contacting the Lead Contact.

Data and code availability

This study did not generate any dataset or codes.

Experimental model and subject details

Flies (D. melanogaster) were raised on standard cornmeal agar medium and maintained in an incubator set at 25°C with a 12-h light/dark cycle. Crosses were reared under the same conditions, and standard medium was supplemented with dry yeast. The GAL4-UAS binary transgenic system was used to express transgenes of interest. Details of genotypes can be found in the Key Resources Table. One- to 5-day-old flies were used for experiments in this study. An equal number of males and females was used for all experiments.

Method details

D. melanogaster stocks and genetics.

D. melanogaster stocks were raised on standard cornmeal agar and maintained in humidified incubators (SRI20PF, Shel Lab) at 25°C with 12-h light/dark cycles. Crosses were reared under the same conditions and supplemented with dry yeast. The strains used in this study are described in the Key Resources Table.

Viability and lifespan analysis.

Viability at 25°C was measured by assessing eclosion rates of 100 wandering L3 larvae collected for each genotype, and then reared in a single vial. Hatching was recorded for 5–6 d. At least three independent biological replicates per genotype were tested for significance and calculated using group analysis on GraphPad (Prism). Lifespan was assessed at 25°C as described previously (Morton et al., 2020). In brief, newly eclosed flies were collected, placed in vials (10 flies per vial), and then transferred to fresh vials weekly, or as needed. Survivorship was scored daily. At least three independent biological replicates were tested for each genotype, and significance was calculated using group analysis on GraphPad (Prism).

Locomotion assays.

Negative geotaxis was tested as previously described (Morton et al., 2020). Briefly, newly eclosed flies (day 0) were collected, and kept in vials for 2–5 d. Cohorts of 10 age-matched flies were transferred to a 25 ml graduated cylinder for analysis. Flies in graduated cylinders were tapped to bring flies to the bottom of the vial and the rate at which the flies traveled to the top of the vial (25 ml mark) was measured at 5, 10, 15, and 30 s). At least three biological replicates per genotype were analyzed and significance was calculated using grouped analysis on GraphPad (Prism).

Drosophila decapitation.

CO₂-anesthetized flies were collected and frozen at −80°C for approximately five minutes. Frozen flies were then placed on a metal plate over dry ice. Gently, #5 Dumont fine forceps (Ted Pella, Inc.) were placed between the Drosophila head and thorax to remove the head from the remainder of the body. Heads were carefully placed in Eppendorf tubes, on ice, for subsequent processing.

RNA isolation for RT-PCR and real-time qPCR.

Total RNA was isolated from ∼25 adult fly heads using the TRIzol (Invitrogen) method. Briefly, Drosophila heads were homogenized in 0.1 ml TRIzol using a motorized pestle (Argos Technologies) on ice. TRIzol was added to samples to bring to a total volume of 0.5 ml and 0.1 ml of 1-Bromo-3-chloropropane (Scientific Laboratory Supplies) was added. Samples were vortexed on high speed for 10 s and incubated at room temperature for 15 min. Next, samples were centrifuged for 15 min at 13,000 × g at 4°C. The top, aqueous layer was removed and placed into a clean Eppendorf tube. An equal volume of 2-propanol (∼250 μL) was added. Samples were inverted 10 times and incubated at room temperature for 10 min. Next, samples were centrifuged for 15 min at 13,000 × g at 4°C. The supernatant was removed, and 0.5 ml of 75% ethanol was added. Samples were centrifuged a final time for 15 min at 13,000 × g at 4°C. The supernatant was removed, and the samples were allowed to air dry until the remaining ethanol evaporated (∼5 min). The pellet was resuspended in 10–20 μL of DEPC water. RNA concentration and purity was assessed using a Spectrophotometer (Thermo Fisher Scientific). Total RNA (1 μg) was treated with DNaseI (Qiagen). cDNA was generated using M-MLV Reverse Transcriptase (Invitrogen). Qiagen Taq polymerase (Qiagen) was used for PCR amplification of target transcripts and products were resolved and imaged on 1–2% agarose gels (Chemi-Doc). qRT-PCR reactions were carried out in technical triplicate with QuantiTect SYBR Green Master Mix using Applied Biosystems StepOne Plus real-time machine (ABI). Results were analyzed using ∆∆CT method, normalized to loading Control (e.g., rpl32), and plotted as relative levels normalized to Control. At least three independent biological replicates were performed for each genotype, and significance was calculated using group analysis on GraphPad (Prism). Primers used for all PCR reactions are listed in the Key Resources Table.

Immunoblotting.

For analysis of Trio protein levels, ∼25 adult fly heads (1–5 d old) were decapitated and collected on dry ice. Protein lysates were prepared by homogenizing heads in 0.5 ml of RIPA-2 Buffer (50 mM Tris-HCl, pH 8; 150 mM NaCl; 0.5% sodium deoxycholate; 1% Igepal CA-630 0.1% SDS) supplemented with protease inhibitors (1 mM PMSF; Pierce Protease Inhibitors; Thermo Fisher Scientific) and 1% SDS. Samples were sonicated 3 × 10 s with 1 min on ice between repetitions, and then centrifuged at 13,000 × g for 15 min at 4°C. Protein lysate concentration was determined by Pierce BCA Protein Assay Kit (Life Technologies). Head lysate protein samples (40–60 μg) in reducing sample buffer (50 mM Tris-HCl, pH 6.8; 100 mM DTT; 2% SDS; 0.1% Bromophenol Blue; 10% glycerol) were resolved on 4–20% Criterion TGX Stain-Free Precast Polyacrylamide Gels (Bio-Rad), transferred to nitrocellulose membranes (Bio-Rad), and incubated for 1 h in blocking buffer (5% nonfat dry milk in 0.1% TBS-Tween) followed by overnight incubation with anti-Trio mAb (1:1000; DHSB #9.4A) diluted in blocking buffer. Primary antibody was detected using species-specific horseradish peroxidase conjugated secondary antibody (Jackson ImmunoResearch) with enhanced chemiluminescence (ECL, Sigma). Densitometry analysis was performed using Image Lab software (Bio-Rad). At least three independent biological replicates were performed for each genotype, and significance was calculated using group analysis on GraphPad (Prism).

Drosophila brain dissection, immunohistochemistry, visualization, and statistical analysis.

For Drosophila mushroom body morphology imaging, brains were dissected using #5 Dumont fine forceps (Ted Pella, Inc.) in PBS supplemented with 0.1% Triton X-100 (0.1% PBS-T). The proboscis was removed to provide a forceps grip point, and the remaining cuticle and trachea were peeled away from the brain. Brains were submerged in 1X PBS on ice until all brains were dissected. Dissected brains were fixed in 4% paraformaldehyde for 30 min and then permeabilized in 0.3% PBS-Triton X-100 (0.3% PBS-T) for 30 min, on ice. Brains were carefully transferred to 0.5 ml Eppendorf tubes in 0.1% PBS-T. For both primary and secondary antibody incubations, brains were left rocking at 4°C for 24–72 h (see list of dilutions and incubation times below) in 0.1% PBS-T supplemented with normal goat serum (Jackson ImmunoResearch) at a 1:20 dilution. Immunostained brains were mounted on SuperFrost Plus slides in Vectasheild (Vector Laboratories) using a coverslip bridge. Brains were imaged on a Nikon A1R confocal microscope. Approximately 25 brains were analyzed for each genotype, for each experiment. Maximum intensity projections were generated using Fiji ImageJ software.

• Antibody: Mouse monoclonal 9.4A anti-Trio, Incubation time: 48–72 h, Dilution: 1:50

• Antibody: Mouse monoclonal 1D4 anti-Fasciclin II, Incubation time: 48–72 h, Dilution: 1:50

• Antibody: Rabbit Polyclonal Anti-Green Fluorescent Protein (GFP), Incubation time: Overnight-24 h, Dilution: 0.125:100

• Antibody: Alexa Fluor 488 AffiniPure Polyclonal Goat Anti-rabbit IgG, Incubation time: Overnight-24 h, Dilution: 1:100

• Antibody: Cy3 AffiniPure Polyclonal Goat Anti-Mouse IgG, Incubation time: Overnight-24 h, Dilution: 1:100

Drosophila neuron live imaging confocal microscopy, neuronal reconstruction, data analysis, and statistical analysis.

Live imaging of class IV ddaC neurons was performed as described previously (Iyer et al., 2012). Briefly, ∼10 wondering third instar ppk-Gal4,UAS-mcd8::GFP-labeled larvae were mounted in 1:5 (vol/vol) diethyl ether: halocarbon oil under an imaging bridge of 22 × 22 mm glass coverslips topped with a 22 × 55 mm glass coverslip. The ddaC images were captured on a Nikon A1R inverted Confocal microscope. Maximum intensity projections were generated using Fiji ImageJ software. Quantitative morphological data were compiled using the Simple Neurite tracer (SNT) plugin for Fiji (Ferreira et al., 2014; Arshadi et al., 2021). Batch processing was completed using a custom Fiji macro and Rstudio script created and gifted by Dr. Atit A. Patel (Dr. Dan Cox Lab, Georgia State University) (Lottes et al., 2024) and the resulting data were exported to Excel (Microsoft).

Statistical analysis.

Group analysis on biological triplicate experiments was performed using one-way or two-way ANOVA (Tukey’s multiple-comparison test) on GraphPad (Prism). Sample sizes (n) and p-values are denoted in text, figures, and/or figure legends and indicated by asterisks (e.g., *, p < 0.05).

Abbreviations used:

GEF

guanine nucleotide exchange factor.

Nab2

nuclear polyadenosine RNA-binding protein 2

Pab

polyadenosine binding protein

RBP

RNA-binding protein

ZC3H14

zinc finger Cys-Cys-Cys His type containing 14.