Genetic dissection of the Drosophila BLOC-1 complex reveals distinctions in synaptic localization and homeostatic plasticity

Abstract

Neuronal trafficking pathways must operate with high fidelity and speed, adapting to the dynamic demands of synaptic activity to maintain stable functionality. The biogenesis of lysosome-related organelles complex 1 (BLOC-1) is an attractive candidate to stabilize synaptic function during such challenges. BLOC-1 is an evolutionarily conserved protein complex composed of eight subunits involved in vesicle trafficking. In the nervous system, the BLOC-1 is associated with neurodevelopmental diseases and synaptic plasticity. However, the functions of each BLOC-1 component remain enigmatic. Here, we use CRISPR to mutate each Drosophila BLOC-1 gene to investigate roles in synaptic growth, function, and homeostatic plasticity. First, we show that BLOC-1 mutations are viable, with no defects in synaptic growth, morphology, or baseline function. We then demonstrate distinct synaptic localization patterns of BLOC-1 components. Finally, we show that only two of the eight BLOC-1 components, dysbindin and snapin, are necessary for presynaptic homeostatic potentiation. These results indicate separable functions and distinct synaptic localization patterns of BLOC-1 subunits, and a need to reconsider predictions made from biochemical models of BLOC-1.

• The BLOC-1 is an octameric protein complex with unresolved questions, including localization and functions at synapses. New null mutations in all Drosophila BLOC-1 genes were generated using CRISPR/Cas9 gene editing, including the first mutations in 4/8 BLOC-1 genes.

• BLOC-1 components exhibit distinct synaptic localization patterns. BLOC-1 localization does not necessarily predict function in homeostatic plasticity.

• Only two out of eight BLOC-1 subunits are required for homeostatic plasticity at the Drosophila neuromuscular junction.

INTRODUCTION

Efficient membrane trafficking is critical for a variety of biological processes, performing key functions in all cell types. In neurons, membrane trafficking is particularly important given the long distances between cell bodies and synapses and the need to coordinate rapid vesicle recycling during neurotransmission. Hence, many specialized pathways have evolved in neurons to maintain robust, rapid, and reliable membrane trafficking to meet the unique challenges associated with neuronal development, growth, function, and plasticity (Südhof, 2004; Rizzoli, 2014; Kononenko and Haucke, 2015). Ultimately, these pathways are necessary to maintain stable information transfer in the nervous system.

The biogenesis of lysosome-related organelles complex 1 (BLOC-1) is an intriguing candidate to ensure stable neuronal membrane trafficking. In the nervous system, BLOC-1 serves important roles in the trafficking of synaptic vesicles, postsynaptic receptors, and endosomal structures (Di Pietro et al., 2006; Setty et al., 2007; Newell-Litwa et al., 2009; Tang et al., 2009; John Peter et al., 2013; Delevoye et al., 2016; Chen et al., 2017; Monis et al., 2017). Biochemically, the BLOC-1 forms a stable octameric complex composed of Dysbindin, Snapin, Muted, Pallidin, and Blos1-4 (Falcón-Pérez et al., 2002; Ciciotte et al., 2003; Li et al., 2003; Starcevic and Dell'Angelica, 2004; Ghiani and Dell'Angelica, 2011; Mullin et al., 2011), Recent studies indicate the existence of two BLOC-1 subcomplexes, one consisting of Dysbindin, Snapin, Muted, and Blos2, and the other consisting of Pallidin, Blos1, Blos3, and Blos4 (Lee et al., 2012; Wang et al., 2025). However, genetic and cell biological studies have led to conflicting models regarding whether BLOC-1 components necessarily function as a unit or independently, with some phenotypes, such as pigmentation and vesicle trafficking, shared by mutants in BLOC-1 (Gwynn et al., 2004; Starcevic and Dell'Angelica, 2004; Setty et al., 2007; Dickman et al., 2012; Chen et al., 2017), while distinctions are observed in other functions (Zhang et al., 2002; Ciciotte et al., 2003; Dickman and Davis, 2009; Dickman et al., 2012; Larimore et al., 2014; Chen et al., 2017). To date, no study has systematically mutated all eight BLOC-1 components to directly test for shared versus distinct functions of the entire BLOC-1.

Drosophila is a powerful model system to uncover fundamental functions of the BLOC-1. First, highly conserved homologues of all eight BLOC-1 genes are encoded in Drosophila (Cheli et al., 2010; Mullin et al., 2015). Second, there is strong evidence that the same biochemical interactions between BLOC-1 subunits observed in mammals are conserved in flies (Cheli et al., 2010). Third, there are a wealth of genetic resources to interrogate the functions of all eight BLOC-1 components, particularly with the recent introduction of specific CRISPR gene editing approaches. One component of the Drosophila BLOC-1, the schizophrenia susceptibility gene dysbindin, was discovered in a forward genetic screen to be necessary for a well-studied form of synaptic plasticity modeled at the glutamatergic fly neuromuscular junction (NMJ) called presynaptic homeostatic potentiation (PHP) (Dickman and Davis, 2009). Although subsequent work found another BLOC-1 component, snapin, to be necessary for PHP, two other components, pallidin and blos1, were found to be dispensable for this process (Dickman et al., 2012; Chen et al., 2017). Intriguingly, Dysbindin and Snapin localize to synaptic vesicle pools (Dickman and Davis, 2009; Dickman et al., 2012), while Pallidin localizes to the neuronal cytoskeleton (Chen et al., 2017), suggesting a model in which localization might predict function of BLOC-1 components.

We have generated new mutant alleles for six of the eight BLOC-1 genes using CRISPR gene editing: dysbindin, snapin, blos2, blos3, blos4, and muted. These mutants were viable and healthy, without major defects in NMJ growth or baseline function. Coupled with previously generated null mutations in the remaining two components (pallidin and blos1), this genetic toolkit provided an opportunity to test the hypothesis that synaptic localization of BLOC-1 components predicts functions. Consistent with this hypothesis, we found distinct localization patterns of BLOC-1 components. However, these distinct localizations did not necessarily correlate with functions in homeostatic plasticity at synapses, further underscoring specialized roles for the BLOC-1.

RESULTS AND DISCUSSION

BLOC-1 subunits are not necessary for synaptic growth or baseline function

We first set out to generate unambiguous null mutations in each of the 4 BLOC-1 genes not yet studied in Drosophila using CRISPR gene editing: blos2, blos3, blos4 and muted. Additionally, although a mutant allele of dysbindin (dysb) has been previously studied (Dickman and Davis, 2009; Shao et al., 2011), this is likely a hypomorphic allele due to transposon insertion into non-coding regions of the dysb locus (LaFave and Sekelsky, 2011; Shao et al., 2011). In addition, no specific mutations in snapin (snap) exist, as it is expressed dicistronically along with an essential gene (Wall et al., 2005; Dickman et al., 2012). Therefore, we also generated new and specific null mutations in dysb and snap as well. Using the same isozygous genetic background, we successfully made null mutations in six BLOC-1 genes: dysb, snap, blos2, blos3, blos4, and muted (Figure 1, A and B). Single-guide RNAs (sgRNA) were designed that targeted early, common exons for each of these genes (Figure 1B, Materials and Methods; Supplemental Table S1). Indel mutations led to frameshift mutations at each locus, resulting in early stop codons in each of the six genes (Figure 1B). Quantitative real-time PCR (qPCR) analysis of the new CRISPR mutants revealed reduced mRNA levels of varying degrees for each of the BLOC-1 alleles (Supplemental Table S1). Although mRNA levels were not substantially reduced in all mutants, these data are in line with successful null mutants generated using identical approaches (Perry et al., 2022; Han et al., 2023). We made two independent null alleles of each gene; each was viable, fertile, and healthy when homozygous, in trans with each other, and in combination with a deficiency (see Materials and Methods and Supplemental Table S1). Thus, there are no essential functions encoded by any of the fly BLOC-1 components.

FIGURE 1:

CRISPR mutagenesis of Drosophila BLOC-1 subunits confirms no significant changes in synaptic growth. (A) Schematic displaying the eight protein subunits that comprise the Drosophila BLOC-1 complex, with subunits required for PHP shown in red, subunits dispensable for PHP in blue, and subunits with undetermined functions in PHP in gray. (B) Schematic of the proteins encoded by the six BLOC-1 genes targeted for mutagenesis using CRISPR/Cas9 gene editing. The sgRNAs used are shown below the region targeted, and the resulting mutation for each allele is shown to the right. Novel CRISPR-generated alleles indicate first amino acid mutated in gene and then number of amino acids until early stop codon. (C) Representative images of muscle 6/7 NMJs in wild type (w¹¹¹⁸), dysb^RS1 (w;;dysb^RS1), snapin^RS1 (w;snapin^RS1), and blos2^RS1 mutants (w;;blos2^RS1) immunostained with antibodies against HRP (neuronal membrane marker) and vGlut (synaptic vesicle marker). Below: representative images of BRP immunostaining (active zone marker) in the indicated genotypes. Quantification of bouton number (D) and BRP density (puncta number/NMJ area) (E) reveals no significant changes to synaptic growth or morphology. Error bars indicate ± SEM. Additional statistical details are shown in extended data Supplemental Table S1.

Previous studies have indicated that some BLOC-1 components may have functions in synaptic development (Ghiani et al., 2010; Wu et al., 2011; Zhou et al., 2012; Mullin et al., 2015). Specifically, synaptic overgrowth was observed in dysb and blos1 mutants (Mullin et al., 2015), while deficits in neurite and dendritic outgrowth were observed in snapin mutants (Wu et al., 2011; Zhou et al., 2012). Therefore, we first examined synaptic growth and morphology in the new BLOC-1 alleles. We immunostained the NMJ with antibodies labeling synaptic markers and quantified synaptic growth, morphology, and active zone density. We observed no significant differences in synaptic growth (bouton number), bouton morphology, or active zone density in any of the BLOC-1 mutants compared with wild-type (Figure 1, C–E; Supplemental Figure S1; Supplemental Table S1). We therefore conclude that dysbindin, snapin, blos2, blos3, blos4, and muted are not required for proper synapse growth or structure at the Drosophila larval NMJ.

Defects in baseline synaptic transmission have been reported in some BLOC-1 mutants (Numakawa et al., 2004; Chen et al., 2008; Dickman and Davis, 2009; Pan et al., 2009; Tang et al., 2009; Cheli et al., 2010). We therefore assessed baseline neurotransmission in the new BLOC-1 mutant alleles using electrophysiology. Specifically, we measured miniature excitatory postsynaptic potential (mEPSP) amplitude, evoked EPSP (eEPSP) amplitude, and calculated quantal content across the mutant alleles in standard conditions (0.4 mM Ca²⁺ saline). We found no significant differences in mEPSP amplitude, eEPSP amplitude or quantal content in any of the mutants compared with wild-type (Figure 2, A–D). Previous studies have found that basal synaptic transmission is reduced in dysb mutants when recorded in lowered extracellular Ca²⁺ (0.2 mM) (Dickman and Davis, 2009). We therefore recorded from all BLOC-1 mutant alleles in this same condition, but found no significant differences in eEPSP amplitude or quantal content in any mutant except for dysb (Supplemental Figure S2). This result is consistent with Dysbindin having a unique role in controlling the Ca²⁺ dependence of transmitter release, as suggested in previous studies (Dickman and Davis, 2009; Dickman et al., 2012). Thus, although dysbindin has a particular role in promoting transmitter release in limiting Ca²⁺ conditions, no other BLOC-1 component is necessary for synaptic growth, structure, or baseline neurotransmission.

FIGURE 2:

Baseline synaptic function is unperturbed in BLOC-1 mutants. (A) Representative mEPSP and EPSP traces from wild-type, dysb^RS1, snapin^RS1, and blos2^RS1 NMJs. Quantification of mEPSP amplitude (B), EPSP amplitude (C), and quantal content (D) reveals no significant defects in baseline neurotransmission in BLOC-1 mutants. Error bars indicate ± SEM. Additional statistical details are shown in extended data Supplemental Table S1.

BLOC-1 components localize to distinct presynaptic structures

Past research has shown that Dysbindin and Snapin colocalize with synaptic vesicle markers, while Pallidin shows a distinct localization pattern, resembling that of axonal microtubule markers (Dickman and Davis, 2009; Dickman et al., 2012; Chen et al., 2017). Interestingly, biochemical studies have found two stable BLOC-1 subcomplexes consisting of Dysbindin, Snapin, Blos2, and Muted in one, and Blos1, Blos3, Blos4, and Pallidin in the other (Lee et al., 2012; Wang et al., 2025). Because two components of the first subcomplex (Dysb and Snap) both localize to synaptic vesicle pools and are required for homeostatic plasticity (Dickman and Davis, 2009; Dickman et al., 2012), while a member of the other subcomplex, Pallidin, exhibits distinct localization and is dispensable for plasticity (Chen et al., 2017), we hypothesized that synaptic localization might correlate with distinct BLOC-1 functions. Importantly, this hypothesis predicts that Blos2 should both colocalize with synaptic vesicle pools and be required for homeostatic plasticity.

Therefore, we generated and expressed a UAS-blos2::smFP-3xFlag transgene to determine Blos2 localization. For comparison, we also generated a new UAS-dysb::3xFlag transgene. As expected, neuronal expression of UAS-dysb::3xFlag showed clear colocalization with the synaptic vesicle marker vGlut (Figure 3, A and B). In contrast, staining with an anti-Pallidin antibody showed distinct localization, resembling the microtubule marker Futsch (Figure 3, A and B), as previously reported (Chen et al., 2017). Finally, neuronal expression of UAS-blos2::3xFlag was found to colocalize with the synaptic vesicle marker vGlut (Figure 3, A and B). Interestingly, anti-Pallidin staining in other BLOC-1 mutants revealed normal Pallidin expression, suggesting that Pallidin stability or localization does not require the presence of other individual subunits (Supplemental Figure S3). A tagged Blos1 transgene was recently developed by FlyORF (UAS-blos1::3xHA, FlyORF #F000608), and we tested NMJ staining of this transgene following motor neuron expression. However, we were unable to detect any signal at the NMJ. Thus, three components of one identified subcomplex of BLOC-1, Dysbindin, Snapin, and Blos2, localize to synaptic vesicles, distinct from the other BLOC-1 subcomplex, which may localize to axonal cytoskeletal structures.

FIGURE 3:

BLOC-1 components exhibit distinct localizations at presynaptic terminals. (A) Representative images of either wild-type muscle 4 NMJs or NMJs expressing transgenes encoding tagged dysbindin (UAS-dysb::3xflag) or blos2 (UAS-blos2::smFP) under the control of the motor neuron driver OK6-GAL4. NMJs were immunostained with anti-HRP, -vGlut or -futsch (neuronal microtubule marker), and -Pallidin, -Flag or -GFP. (B) Representative images of terminal boutons stained with the same antibodies in A. Note, the characteristic “doughnut” patterns of synaptic vesicles (vGlut), which colocalize with Dysb::Flag as expected, and, interestingly, with Blos2::smFP. In contrast, Pallidin exhibits a distinct localization, somewhat overlapping with axonal microtubules (futsch).

Dysbindin and snapin are uniquely required for presynaptic homeostatic plasticity

In our final set of experiments, we tested the hypothesis that BLOC-1 components that localize to synaptic vesicle pools are necessary for the expression of PHP. PHP is an evolutionarily conserved form of homeostatic plasticity, where acute pharmacological or chronic genetic challenges that diminish postsynaptic glutamate receptor function lead to enhanced presynaptic neurotransmitter release that maintains stable synaptic strength (Davis and Müller, 2015; Frank et al., 2020; Goel and Dickman, 2021). Both dysb and snap were previously found to function in presynaptic motor neurons for the expression of PHP (Dickman and Davis, 2009; Dickman et al., 2012), while two other BLOC-1 components, pallidin and blos1, were dispensable for this process (Chen et al., 2017). Given both the biochemical and synaptic localization distinctions between dysb, snap, and blos2 on one side, and pallidin and blos1 on the other, we tested both acute and chronic forms of PHP to determine whether the remaining 4 BLOC-1 components were necessary for PHP. We predicted that of the remaining components, blos2 and perhaps muted might also be required for PHP expression.

We first assessed both acute and chronic PHP in the newly generated dysb and snap null alleles. Application of the irreversible glutamate receptor antagonist philanthotoxin-433 (PhTx) or genetic loss of the glutamate receptor subunit GluRIIA reduced mEPSP amplitude by ∼50% compared with baseline values in wild type, dysb, and snap, as expected (Figure 4, A–D). In wild-type, EPSP amplitudes are maintained due to a homeostatic enhancement in the number of synaptic vesicles released (quantal content); this almost 200% increase in quantal content offsets the ∼50% decrease in glutamate receptor sensitivity, demonstrating robust PHP expression. In contrast, in both dysb and snap mutants, EPSP amplitudes are proportionately reduced with mEPSPs, and no increase in quantal content is observed (Figure 4, A–D). Hence, PHP is blocked in both new dysb and snap alleles, as expected. Next, we applied PhTx or combined genetic loss of GluRIIA to the remaining BLOC-1 mutants. Although mEPSPs were reduced in all genotypes, robust increases in quantal content were also observed in each mutant, demonstrating these genes are not required for acute or chronic forms of PHP (Figure 4, A–D). Indeed, counter to our prediction, robust PHP expression was observed in the remaining 4 BLOC-1 mutants, including blos2, as well as blos1 and pallidin in previous studies (Chen et al., 2017). Thus, dysbindin and snapin have unique roles in mediating PHP expression, where localization to synaptic vesicles does not predict functions in this form of plasticity. We propose a model summarizing these findings in Figure 5.

FIGURE 4:

Snapin and dysbindin are uniquely required for both acute and chronic forms of PHP. (A) Schematics and representative mEPSP and EPSP traces from wild type, dysb, snapin, and blos2 synapses at baseline (gray line) and after PhTx application (black line). (B) Quantification of mEPSP amplitudes and quantal content after PhTx application normalized to baseline values in the indicated genotypes. dysbindin and snapin are unique in their necessity for the enhancement of presynaptic neurotransmitter release (quantal content) characteristic of PHP expression. (C and D) Similar traces and quantification as A and B but after chronic PHP challenge (genetic loss of GluRIIA). Note that like acute PHP, dysbindin and snapin are uniquely required for chronic PHP expression. Error bars indicate ± SEM. Additional statistical details are shown in extended data Supplemental Table S1. ***p < 0.001 and ****p < 0.0001.

FIGURE 5:

Differential localization and functions of BLOC-1 components. (A) Schematic showing known and putative localization of BLOC-1 components. Dysbindin, Snapin, and Blos2 colocalize with synaptic vesicle synaptic vesicle markers, while Pallidin localizes to distinct regions, similar to axonal cytoskeletal markers. The four components with undetermined localization are indicated in gray. Note that two biochemical subcomplexes of BLOC-1 components are known, implying that Muted may localize with synaptic vesicles, while the others may localize with Pallidin. (B) Schematic showing that Dysbindin and Snapin are uniquely required for PHP, while other subunits are dispensable for this process.

Unitary versus distinct BLOC-1 functions

Several lines of evidence now clearly establish that definitions of a unitary “BLOC-1″ complex have limited utility in understanding the in vivo functions of this complex. Support for a unitary BLOC-1 first emerged from biochemical studies, in which a stable octameric complex was isolated (Falcón-Pérez et al., 2002). Subsequent genetic and cell biology work appeared to largely support the idea of a unitary BLOC-1, in which loss of any of the components might phenocopy each other or lead to the degradation of other components (Falcón-Pérez et al., 2002; Moriyama and Bonifacino, 2002; Starcevic and Dell'Angelica, 2004; Di Pietro et al., 2006; Gautam et al., 2006; Feng et al., 2008; Salazar et al., 2009; Cullinane et al., 2012; Chen et al., 2017). Indeed, studies in the yeast BLOC-1 found that localization of the complex to endosomes is dependent upon the presence of every subunit (John Peter et al., 2013). In the nervous system, at least three BLOC-1 components—Dysb, Pallidin, and Blos1—are necessary to maintain the synaptic vesicle pool during periods of high intensity stimulation, pointing to a potentially common shared function of the complex (Chen et al., 2017).

However, it is now clear that despite the possibility of some shared functions, BLOC-1 components exhibit distinct cellular localization and synaptic functions. First, biochemical analysis of BLOC-1 identified two stable subcomplexes within the larger complex, one consisting of Pallidin, Blos1, Blos3, and Blos4 and the other consisting of Dysbindin, Snapin, Blos2, and Muted (Lee et al., 2012; Wang et al., 2025). The existence of these two subcomplexes may enable separable functions between these groups of subunits. Second, at least three BLOC-1 components localize to synaptic vesicle pools—Dysb, Snap, and Blos2, which parallel insights from mammalian studies, where Snapin was found to associate with the SNARE machinery mediating synaptic vesicle release (Ilardi et al., 1999; Chheda et al., 2001; Pan et al., 2009). In contrast, Pallidin localizes to the axonal cytoskeleton, which may also reflect insights from other systems in which pallidin contributes to the differentiation of developing smooth muscle cells, the organization of the actin cytoskeleton, and even the regulation of the metastatic cancer cells (Parast and Otey, 2000; Mykkänen et al., 2001; Falcón-Pérez et al., 2002; Jin et al., 2010; Najm and El-Sibai, 2014). Second, only two of the eight BLOC-1 components, Dysbindin and Snapin, are required for PHP, and several other BLOC-1 components exhibit potentially unique functions (Dickman and Davis, 2009; Dickman et al., 2012). For example, Blos2 may act to regulate Notch signaling and gene transcription (Sun et al., 2008; Zhou et al., 2016). Meanwhile, mice without blos1 or snapin are embryonic lethal, suggesting essential functions for these two subunits in mammalian embryonic development (Tian et al., 2005; Zhang et al., 2014). Indeed, there are even differences regarding what side of the synapse BLOC-1 components function, with some involved in presynaptic membrane trafficking, and others in postsynaptic trafficking of receptor complexes (Ji et al., 2009; Tang et al., 2009; Cai et al., 2010; Marley and von Zastrow, 2010; Zhang et al., 2014; Niwa et al., 2017). In fact, Blos1, Blos2, and Snapin are components of an entirely separate protein complex called the BLOC-one-related complex (BORC) (Pu et al., 2015). This separate complex consists of, in addition to the BLOC-1 proteins, four additional proteins unrelated to BLOC-1. The BORC complex associates with lysosomes and promotes the intracellular movement of lysosomes (Pu et al., 2015). Thus, in many cases, BLOC-1 components are promiscuous and can dynamically associate with a variety of complexes.

There are, however, a few caveats that should be discussed. First, it is possible that the BLOC-1 does indeed function as a unitary complex, but that there is redundancy between components. We only tested mutations in individual mutants, and it is possible that loss of several components together would reveal functions in viability, fertility, pigmentation, or plasticity. Double or even triple knockout (KO) of BLOC-1 genes in mice, however, did not reveal decreased viability or further depigmentation compared with single KO animals, but the additional mutations were associated with a greater number of lung abnormalities characteristic of lung disease (Gautam et al., 2006). Second, our localization analysis relied on overexpression of tagged subunits. Although these can rescue PHP defects following loss of dysb or snap (Dickman and Davis, 2009; Dickman et al., 2012; Chen et al., 2017), it is possible the tags lead to localization patterns that do not reflect endogenous structures. Furthermore, we only performed fixed tissue staining; it is known that delicate intracellular membranous networks, which BLOC-1 may be a component of, may dissolve during fixation, leading to apparent differences in localization. Live imaging of BLOC-1 components in future studies may help to control for such potential fixation artifacts.

Several BLOC-1 components have garnered attention in recent years due to associations with various diseases. Originally, Hermansky–Pudlak syndrome, a disease in humans associated with reduced pigmentation in the skin, hair, and eyes and bleeding diathesis, was found to result from loss of the BLOC-1 components blos3, blos4, pallidin, dysbindin, and muted (Introne et al., 1993; Zhang et al., 2002; Ciciotte et al., 2003; Li et al., 2003; Wei, 2006; Setty et al., 2007; Cullinane et al., 2012). Some patients also experience progressive lung disease and forms of immune deficiencies (Wei, 2006). Subsequently, genome-wide association studies identified dysbindin (DTNBP1) as a putative schizophrenia susceptibility gene (Straub et al., 2002; Mullin et al., 2015), with muted and potentially snapin also suggested to be associated with this disease (Straub et al., 2005; Ryder and Faundez, 2009; Li et al., 2023). More recently, deficits in the sleep/wake cycle have been observed in pallidin mutant mice (Lee et al., 2018). Given the seemingly disparate nature of diseases and phenotypes associated with BLOC-1, it is tempting to speculate that the distinctions we observed in Drosophila BLOC-1 localization and functions may be reflected in disease association and etiology.

MATERIALS AND METHODS

Request a protocol through Bio-protocol

Fly stocks

Drosophila stocks were raised at 25°C on standard molasses food. The w¹¹¹⁸ strain is used as the wild-type control unless otherwise noted as this is the genetic background in which all genotypes are bred. All experiments were performed on Drosophila third-instar larvae of both sexes. Details of all fly stocks used, including their sources, are listed in Supplemental Table S2.

Molecular biology

BLOC-1 mutants were generated using a CRISPR/Cas9 gene editing strategy as described (Kikuma et al., 2017; Perry et al., 2022). gRNA sequences for each respective mutant are listed in the Supplemental Table. gRNAs were cloned into the pU63 vector (#112811, Addgene) with intervening tRNA (F + E) sequences for expressing multiple gRNAs (Poe et al., 2019). Constructs were sent to BestGene (Chino Hill, CA) for targeted insertion into attp2 or attp40 docking sites. Two or three sgRNAs were used to target each of the six genes. sgRNA sequences are detailed in Supplemental Table S2. gRNA flies were crossed to nos-Cas9 (#54591, BDSC) to induce active germline CRISPR mutagenesis in male flies, and 20 independent lines were screened by PCR for mutations. This identified two independent indel mutations for each gRNA that shifted the open reading frame, leading to early stop codons. The alleles selected for additional analysis had the earliest stop codons (Figure 1B) and are predicted to be null mutants. Importantly, both the gRNA and nos-Cas9 chromosomes were only present in the germline for one generation before outcrossing to the isozygous w¹¹¹⁸ line; this ensured that neither of the transgenes nor their genetic backgrounds were present in the final mutant alleles.

To generate the UAS-dysb::3xflag or UAS-blos2::smFP transgenes, the full-length dysb or blos2 cDNA was cloned into the pACU2 vector (31223; Addgene, Cambridge, MA) with in frame C-terminal 3xflag or spaghetti monster FLAG (10xFLAGsmFP) tags using standard approaches. Constructs were sequence verified and injected into the w¹¹¹⁸ strain by BestGene (Chino Hill, CA) using the VK18 insertion site on the second chromosome (Venken et al., 2006). We also tested the UAS-blos1::3xHA transgene developed by FlyORF (F000608) using the pGW-HA.attB vector (Bischof et al., 2013), but were unable to detect any obvious signal at the NMJ despite testing multiple (Bouin's solution, methanol, and 4% paraformaldehyde). We therefore conclude that the Blos1::3xHA transgenic cassette does not successfully express at the Drosophila NMJ.

Electrophysiology

All dissections and sharp-electrode current clamp recordings were performed as described (Han et al., 2023) using modified hemolymph-like saline (HL-3) containing: 70 mM NaCl, 5 mM KCl, 10 mM MgCl₂, 10 mM NaHCO₃, 115 mM Sucrose, 5 mM Trehelose, 5 mM HEPES, and 0.5 mM CaCl₂, pH 7.2 from cells with initial resting potentials lower than −60 mV, and input resistances >5 MΩ. Recordings were performed on an Olympus BX61 WI microscope using a 40x/0.80 NA water-dipping objective and acquired using an Axoclamp 900A amplifier, Digidata 1440A acquisition system and pClamp 10.5 software (Molecular Devices). mEPSPs were recorded in the absence of any stimulation and low pass filtered at 1 kHz. All recordings were made on abdominal muscles 6 or 7, segments A2 or A3 of third-instar larvae. mEPSPs were recorded for 60 s and analyzed using MiniAnalysis (Synaptosoft) and Excel (Microsoft) software. The average mEPSP amplitude for each NMJ was obtained from ∼100 events in each recording. To acutely block postsynaptic receptors, larvae were incubated with or without philanthotoxin-433 (PhTx; 20 µM, Sigma) in HL-3 for 12 min.

Immunocytochemistry

Third-instar larvae were dissected in ice-cold 0 Ca²⁺ HL-3 and immunostained as described (Li et al., 2021). In brief, larvae were either fixed in Bouin's fixative for 5 min (Sigma, HT10132-1L), 100% ice-cold methanol for 5 min, or 4% paraformaldehyde (PFA) for 10 min. Larvae were then washed with PBS containing 0.1% Triton X-100 (PBST) for 30 min, blocked with 5% normal donkey serum followed by overnight incubation in primary antibodies at 4°C. Preparations were then washed 3x in PBST, incubated in secondary antibodies for 2 h, washed 3x in PBST, and equilibrated in 70% glycerol. Before imaging, samples were mounted in VectaShield (Vector Laboratories). Details of all antibodies, their source, dilution used, and references are listed in Supplemental Table S2.

Confocal imaging and analysis

Samples were imaged using a Nikon A1R Resonant Scanning Confocal microscope equipped with NIS Elements software and a 100x APO 1.4NA oil immersion objective using separate channels with four laser lines (405, 488, 561, and 647 nm) as described (Perry et al., 2017). Maximum intensity projections were utilized for quantitative image analysis using the general analysis toolkit of NIS Elements software. To quantify neuronal membrane area, a mask was created around the Horse Radish Peroxidase (HRP) channel. To calculate active zone density, only BRP puncta within this HRP mask were counted and normalized to neuronal membrane surface area. All measurements based on confocal images were taken from NMJs acquired from at least six different animals.

qPCR analysis

qPCR methods were performed as described previously (Dickman et al., 2012; Li et al. 2021). Total RNA was extracted from 20 to 40 adult fly heads per sample and isolated using TRizol (Thermo Fisher Scientific). Reverse transcription was performed (Maxima H Minus RT; Thermo Fisher Scientific) using Invitrogen Oligo(dT)12-18 Primer (Thermo Fisher Scientific) and 2 µg total RNA. cDNA from the room temperature reaction was used as a template in a 10 µl PCR reaction (Applied Biosystems Powerup SYBR Green, Thermo Fisher Scientific). This 10 µl reaction was done in triplicate with Gapdh2 used as an internal control. In addition, three replicates of nontemplate control were performed for each set of primers. PCR was performed using the BIO-RAD C1000 Thermal Cycler, and amplification plots were analyzed by BIO-RAD CFX Manager Software. The relative quantity of amplified cDNA for each gene was calculated using the ΔΔC_t method and normalized to Gapdh2 expression in each sample (Supplemental Table S1).

Statistical analysis

Data were analyzed using GraphPad Prism (version 7.0), MiniAnalysis (Synaptosoft), or Microsoft Excel software (version 16.22). Sample values were tested for normality using the D'Agostino & Pearson omnibus normality test that determined that the assumption of normality of the sample distribution was not violated. Data were then compared using a one-way ANOVA and tested for significance using a Tukey's multiple comparison test. In all figures, error bars indicate ±SEM, with the following statistical significance: *p<0.05, **p<0.01, ***p<0.001, and ****p<0.0001; ns, not significant. Additional statistics and sample number values (n) for all experiments are summarized in Supplemental Table S1.

Abbreviations used:

(BLOC-1)

biogenesis of lysosome-related organelles complex-1

(NMJ)

neuromuscular junction

(PHP)

presynaptic homeostatic potentiation