Octopamine neuron dependent aggression requires dVGLUT from dual-transmitting neurons

Neuromodulators such as monoamines are often expressed in neurons that also release at least one fast-acting neurotransmitter. The release of a combination of transmitters provides both “classical” and “modulatory” signals that could produce diverse and/or complementary effects in associated circuits. Here, we establish that the majority of Drosophila octopamine (OA) neurons are also glutamatergic and identify the individual contributions of each neurotransmitter on sex-specific behaviors. Males without OA display low levels of aggression and high levels of inter-male courtship. Males deficient for dVGLUT solely in OA-glutamate neurons (OGNs) also exhibit a reduction in aggression, but without a concurrent increase in inter-male courtship. Within OGNs, a portion of VMAT and dVGLUT puncta differ in localization suggesting spatial differences in OA signaling. Our findings establish a previously undetermined role for dVGLUT in OA neurons and suggests that glutamate uncouples aggression from OA-dependent courtship-related behavior. These results indicate that dual neurotransmission can increase the efficacy of individual neurotransmitters while maintaining unique functions within a multi-functional social behavior neuronal network.


Introduction
The classical view of information transfer for many decades was that each neuron released a single neurotransmitter, leading to the 'one neuron, one transmitter' hypothesis [1], formalized by John Eccles as Dale's Principle [2]. Dale himself, however, recognized the possibility that neurons can release more than one molecule [3] and indeed, research from multiple systems and neuronal populations have established that many if not most, neurons release more than one neurotransmitter [4][5][6][7]. Dual neurotransmission has the potential to transform the way we consider the computation and transmission of information by neurons, circuits and networks. Presynaptically, the release of two neurotransmitters could impact information transfer by several mechanisms that are not mutually exclusive including; attenuating signals by modulating presynaptic autoreceptors, transmitting spatially distinct signals by segregating specific vesicle populations to different axon terminals, or conveying similar information through the release of both neurotransmitters from the same synaptic vesicle [8][9][10][11]. In addition, one vesicular neurotransmitter transporter can increase the packaging of the other neurotransmitter into the same synaptic vesicle (SV), a process called vesicular synergy [4,12,13]. At post-synaptic targets, the release of two transmitters can enhance the strength of the same signal and/or convey unique signals through spatially-restricted receptor expression and second messenger cascades [7,14]. While recent studies have provided insight into these phenomena at the cellular level [11,12,15,16], the behavioral relevance of co-transmission in normal as well as pathological conditions is an area of considerable complexity and interest.
The genetic tools of Drosophila provide the ability to genetically dissect the signaling properties of dual transmission on behavioral networks in general and upon the circuits that control aggression in particular. Aggression is a hardwired behavior that has evolved in the framework of defending or obtaining resources [17,18]. Monoamines such as serotonin (5-HT), dopamine (DA), norepinephrine (NE) and octopamine (OA), the invertebrate homologue of NE, have powerful modulatory effects on aggression in systems ranging from insects and crustaceans to humans [19][20][21][22][23]. In humans, aggressive behavior can be expressed at extreme levels and out of context due to medical, neurologic and or psychiatric disorders including depression and schizophrenia [24][25][26]. Pharmacological agents that selectively manipulate monoamine signaling are used to treat anxiety and depression, yet these drugs are often ineffective, and in the case of serotonin/norepinephrine reuptake inhibitors (SNRIs) can induce side effects including increased aggression and impulsivity [25,[27][28][29].
At least two difficulties arise in targeting monoamines to achieve successful outcomes. First, monoamines can be released from synaptic vesicles (SVs) into the presynaptic cleft and by extrasynaptic release from large dense core vesicles (LDCVs) [30][31][32][33]. Thus, monoamines are recognized both as neurotransmitters and as neuromodulators that signal via diffusion [34,35]. The second difficulty is that their effects are likely exerted through interactions with neuropeptides (neuropeptide Y and oxytocin are two examples) and with neurotransmitters including GABA and glutamate [5,14,36,37]. Due in part to recent studies suggesting the expression of vesicular glutamate transporters (VGLUTs) can be altered by psychiatric medications [38][39][40][41] and the importance of dopamine neuron glutamate co-transmission on the schizophrenia resilience phenotype in mice [42], we generated new tools to identify and manipulate glutamate function in monoamine-expressing neurons.
We found that the majority of OA neurons within the Drosophila nervous system also express the vesicular neurotransmitter transporter for glutamate (dVGlut). Functionally, glutamate (GLU) co-expression could convey the same information by promoting the synaptic vesicle packing of OA or GLU may convey distinct information that is separate from the function of OA. In Drosophila, OA synthesis and release are essential for conserved social behaviors; males without OA display low levels of aggression and high levels of inter-male courtship [43][44][45][46][47]. We demonstrate that males deficient for dVGLUT solely in OA-glutamate neurons (OGNs) also exhibit a reduction in aggression, but without a concurrent increase in inter-male courtship. These results indicate both OA and dVGLUT are required in dual-transmitting neurons to promote aggression. However, only OA is required for the suppression of intermale courtship and thus the function of dVGLUT in OGNs is not limited to vesicular synergy.
To ask if the separable effects of OA on courtship circuitry may be attributable to spatially distinct OA signals, we conditionally expressed a new epitope-tagged version of the Drosophila vesicular neurotransmitter transporter for monoamines (V5-tagged VMAT) in OGNs. While the majority of V5-VMAT and dVGLUT expression colocalize, VMAT is detected in distinct puncta without dVGLUT suggesting the possibility of separable signal transmission. Together, these results demonstrate the complex behavior of aggression requires both dVGLUT and OA in dual-transmitting neurons and suggests within monoamine neurons, GLU may provide a therapeutic target to modulate aggression in pathological conditions.

dVGLUT is co-expressed in OA neurons
The co-expression of vesicular neurotransmitter transporters has been primarily used to identify dual-transmitting neurons [48][49][50][51][52]. To examine glutamatergic transmitter expression, we generated a monoclonal dVGLUT antibody and validated its specificity using a new dVGlut allele, dVGlut SS1 . In homozygous dVGlut SS1 progeny, dVGLUT protein is not detectable (S1 Fig, Methods), thus demonstrating the specificity of the dVGLUT antibody. As dVGLUT expression is widespread and mainly found in synaptic terminals (S1 Fig), we used the Gal4-UAS system to identify monoamine neurons that express GLU. In this study, we focused specifically on OA neurons that co-express dVGLUT (OA-glutamate neurons (OGNs)).
Cell bodies of OGNs were visualized by a UAS-dsRed.NLS reporter under control of dVGlut-gal4 (hereafter referred to as dVGlut>dsRed). OGNs were identified by antibodies to tyrosine decarboxylase 2 (TDC2) and tyramine β-hydroxylase (TβH) as OA is synthesized from the amino acid tyrosine via the action of Tdc and Tβh in invertebrates [46]. OGNs from 10 dVGlut>dsRed Tdc2-labeled male brains were quantified by the multi-point ImageJ tool followed by manual verification of each optical section. Within the brain, OA neurons that coexpress glutamate are found in the subesophageal zone (SEZ), the periesophageal neuropils (PENP), the anterior (ASMP) and posterior superior medial protocerebrum (PSMP), and the protocerebral bridge (Fig 1A-1E, S1 Table). Co-expression occurs in each region of interest (Fig 1A-1E [53]. After detecting no reporter expression from a Tβh-gal4 driver, dVGLUT cell body expression in OGNs was detected in brains from tdc2-gal4;UAS-dsRed adults (S4 Fig). In total, this analysis reveals that of the~100 OA neurons in the Drosophila adult nervous system, about 70% express dVGLUT.

dVGLUT is not required for OA neuron identity
To reduce glutamate function solely in OGNs, a UAS-driven inverted repeat transgene targeting dVGlut (UAS-dVGlut-RNAi) was expressed under control of the tdc2-gal4 driver (hereafter tdc2>dVGlut-RNAi) (Fig 2A and 2B). The effectiveness of this UAS-dVGlut-RNAi line has been verified at the transcript level through RT-qPCR ( [12] and S5 Fig) and functionally as the frequency of miniature excitatory postsynaptic potentials (mEPSP) were reduced by this dVGlut RNAi in presynaptic glutamatergic larval motor neurons [12]. As the loss of VGLUT2 in vertebrate dopamine-glutamate dual transmitting neurons impairs survival and

Reducing glutamate in OGNs decreases male aggression and inter-male courtship
We and others previously demonstrated OA is required for two distinct social male behaviors; the promotion of aggression, and the inhibition of intermale courtship [43,46,55,56]. To address whether dVGLUT performs a related or separable role in these OA-dependent behaviors, we quantified changes in aggression and intermale courtship. Fights between pairs of tdc2>dVGlut-RNAi males, and transgenic controls were recorded and multiple agonistic parameters quantified including: latency to the first lunge, number of lunges, and number of agonistic wing threats (Fig 2A, [57,58]). As behavioral patterns are scored for 30 minutes after the first lunge, each male pair has the same amount of time to exhibit aggressive events or inter-male courtship ( Fig 2B).
Males with decreased dVGLUT in OGNs neurons exhibited a significant reduction in aggression as measured by lower numbers of lunges and wing threats, and an increase in the latency to initiate aggression (Fig 2D-2F). These aggression deficits are the same as in males that lack OA [43, 46, 47]. Importantly, the locomotor activity of tdc2>dVGlut-RNAi adults during the aggression assay did not differ from dVGlut-RNAi controls (S7A Fig).
Interactions between control male pairings within a fight can include low levels of intermale courtship as measured by unilateral wing extensions (UWE, the courtship song motor pattern). Males without OA exhibit high levels of inter-male courtship[43, 55,56] and previously, we determined the function of three OA-FruM+ neurons is required to suppress intermale courtship [55]. If dVGLUT is only needed to enhance monoamine vesicular packaging and thus modulate OA function, we would expect males with reduced dVGlut levels to display the same behavioral deficits, i.e. high levels of inter-male courtship. However, tdc2>dVGlut-RNAi males did not exhibit inter-male courtship ( Fig 2G). These results suggest; 1) dVGLUT is required in OGNs to promote aggression, and 2) dVGLUT is not required to suppress intermale courtship.

Aggression requires dVGLUT function in OA-GLU brain neurons
In the adult, motor neurons innervating leg and wing muscles express glutamate [59]. Therefore, the observed behavioral deficits in tdc2>dVGlut-RNAi males may reflect impairments at the neuromuscular junction. To address this possibility, we spatially restricted expression of the dVGlut-RNAi transgene to the brain using the teashirt-lexA 8xlexAop2-IVS-Gal80 (hereafter tsh>Gal80) transgenic combination ( Fig 2H). The tsh>Gal80 transgenic combination was effective at blocking Gal4-mediated transcription in the entire VNS including in OGNs that innervate muscles required for courtship and wing threat behaviors (S8 Fig).
With dVGlut function maintained in motor neurons, it was possible all aggressive behaviors would return to control levels. However, latency to initiate aggression remained longer in males with reduced dVGLUT in brain OGNs (tdc2>tsh>Gal80>dVGlut-RNAi) and lunge number remained lower when compared to controls (Fig 2I and 2J). Wing threat numbers were at levels lower than one control ( Fig 2K) which likely reflects the incompleteness of dVGlut RNAi interference. In contrast, providing dVGLUT function in OGN VNS neurons restored intermale courtship to control levels ( Fig 2L). Although total behavioral events by experimental males (lunges, wing threats, intermale courtship) per minute decreased, overall activity did not (S7 Fig) nor did male-female courtship (Fig 3). These results indicating GLU transport in brain OGNs is required to initiate aggression and for the lunge pattern itself may reflect deficits in the detection of male pheromones as we previously described for OA [43]. Specifically, the suppression of intermale courtship requires the function of three OA-FruM + neurons located in the brain [55] and, aggression requires pheromonal information from Gr32a-expressing chemosensory neurons located in the mouth to OA SEZ neurons [43]. The latency to initiate courtship towards a female returned to control levels in males with reduced dVGLUT in brain OGNs. (D) The courtship index was restored to control levels in tdc2-gal4/tsh>Gal80;dVGlut-RNAi males. (E) tdc2-gal4/tsh>Gal80;dVGlut-RNAi males exhibited the same latency to copulation as controls. (F) The copulation success of males with a dVGLUT reduction in brain OGNs was not significantly different from controls. N values for each genotype located on panel A. Finally, males with reduced dVGLUT in brain OGNs (tdc2>tsh>Gal80>dVGlut-RNAi) performed all measured male-female courtship parameters including latency to court, courtship index, latency to copulation and copulation success at levels indistinguishable from controls (Fig 3). Together, these results indicate dVGlut in OGNs is required in males both for aggression and courtship toward a female and at the behavioral level, the functional requirement for dVGLUT in OGN motor neurons vs. central brain neurons is spatially separable.

Removal of glutamate in OGNs using the B3RT-vGlut conditional allele
The experiments above used two different approaches to reduce neurotransmitter levels, but not eliminate dVGLUT. To completely remove glutamate transporter function in OGNs, a conditional allele of dVGlut, B3RT-dVGlut-LexA (hereafter B3RT-dVGlut), was developed via genome editing. Genome edits to the dVGlut locus included flanking the dVGlut coding exons with B3 recombination target sites (B3RTs) [60] in the same orientation and inserting the coding sequences of the LexA transcription factor immediately downstream of the 3' B3RT ( Fig  4A). With B3RT-dVGlut, glutamate function can be temporally and spatially controlled using Gal4 drivers of interest to express the B3 recombinase that in turn catalyzes the in vivo excision of DNA between the B3RTs (Fig 4B). Two outcomes result after B3 recombinase-mediated excision; 1) a dVGlut null allele is generated solely in the neurons of interest, and 2) a dVGlut-LexA driver is created that allows visualization of glutamatergic neurons when a LexAop reporter is present.
To assess the functionality of dVGlut within the B3RT-dVGlut chromosome pre-and postexcision, the B3RT-dVGlut chromosome was crossed with the null allele, dVGlut SS1 (S1 Fig). In the absence of a Gal4 driver, vGlut SS1 /B3RT-vGlut progeny are fully viable and no LexAopdriven reporter gene expression is detected (Fig 4C). In contrast, when B3 recombinase (UAS-B3) is expressed in the nervous system by the pan-neuronal driver, n-syb-Gal4, dVGLUT expression is eliminated and vGlut SS1 /B3RT-dVGlut;UAS-B3/n-syb-Gal4 progeny are inviable (data not shown). These results establish that the B3RT-dVGlut genome edits preserve dVGLUT function prior to excision, but after excision, as expected with removal of the entire dVGLUT protein-coding sequence, a dVGlut null allele is generated.
To verify the functionality of the B3RT-dVGlut chromosome in Tdc2+ neurons, we crossed tdc2-gal4 with B3RT-dVGlut;UAS-B3. Following B3-mediated excision in Tdc2+ neurons, the resulting dVGlut-lexA driver is active in OGNs demonstrating the dVGlut coding region was removed. The excision of dVGlut and substitution with LexA in the adult nervous system was confirmed by co-localization of nuclear markers ( Fig 4D and 4D'). This result provides additional confirmation the majority of Tdc2+ neurons are glutamatergic. In addition, nuclear reporters were used to confirm the loss of dVGLUT does not obviously alter OGN differentiation (S9 Fig).
To completely remove dVGLUT function, we used the dVGlut SS1 null allele in combination with the B3RT-dVGlut conditional null allele. Due to the requirements for GLU in OA-GLU motor neurons, we crossed the tsh>Gal80 transgenes onto the B3RT-dVGlut chromosome. Males with homozygous null dVGlut mutations in brain OGNs were generated by driving B3 recombinase with tdc2-gal4 (dVGlut SS1 /B3RT-dVGlut tsh>Gal80;UAS-B3/tdc2-gal4). As expected, the complete loss of GLU in brain OGNs reduced male aggression. Specifically, the latency to initiate aggression increased, and lunge numbers decreased (Fig 4E and 4F). Not unexpectedly, the complete elimination of dVGLUT function resulted in aggression deficits significantly worse when compared to the RNAi approach (Fig 4I) including now a reduction in wing threat number ( Fig 4G) which demonstrates an advantage in using the conditional null B3RT-dVGlut allele. Finally, and significantly, the number of inter-male wing extensions did not differ from controls ( Fig 4H) nor from males with a reduction of dVGlut in brain OGNs (Fig 2K). In summation, the dVGlut SS1 /B3RT-dVGlut null combination elegantly and independently validates the aggression phenotypes based on dVGlut RNAi-based reduction, demonstrates the applicability of a powerful new conditional genetic tool, and confirms that dVGLUT function in OGNs is not required to regulate intermale courtship.

Reducing GLU by EAAT1 overexpression recapitulates the decrease in aggression
At this point, GLU function within OGNs has been altered by reducing glutamate transport into synaptic vesicles. Whether the aggression phenotypes of OGN dVGLUT mutant males are due to deficits in the concentration of GLU into synaptic vesicles, the packaging of OA, or a reduction of released GLU is not clear. After release, glutamate is rapidly removed from synapses by excitatory amino acid transporters (EAATs) [61,62]. Therefore, to reduce GLU signaling after release, we increased expression of the only high-affinity glutamate transporter in Drosophila, EAAT1 ( Fig 5A) [63,64].
EAAT1 is expressed in glia throughout the nervous system [64]. By examining 2-10 individual EAAT1-GFP clones in~40 brains, we determined OGN neuronal cell bodies and arborizations are consistently enmeshed by EAAT1-expressing glia (Fig 5B and 5C). To reduce glutamate signaling after release, EAAT1 expression was increased via a transgene (EAAT1--gal4;UAS-EAAT1). While a loss of EAAT1 impairs larval movement [65], overexpression of EAAT1 has been used in adult long-term memory formation assays which requires locomotion [66]. Similar to the dVGLUT loss-of-function results above, the aggressive behavior of males with reduced GLU signaling by EAAT1 overexpression (EAAT1-gal4;UAS-EAAT1) was altered in two parameters: the latency to initiate lunging increased and lunge number decreased (Fig 5D and 5E). Locomotor activity during the aggression assay did not differ ( Fig  5F). Although future experiments will be needed to determine if the promotion of aggression requires dVGLUT packaging of OA in synaptic vesicles and OGN glutamate signaling to downstream targets, results from this section support the hypothesis that OGN-mediated aggression requires GLU.

OA and Glu signal to a shared aggression-promoting circuit
If Glu and OA convey signals to separable aggression-promoting circuits, a loss of both neurotransmitters would reduce aggression greater than the loss of either alone (Fig 6A). If, however, Glu and OA signal to a shared circuit or circuits that converge, a loss of both transmitters would reduce aggression to the same levels as the loss of one alone. To address this question, we incorporated the previously described null allele Tβh nM18 [67] and generated Tβh nM18 ; tdc2>dVGlut-RNAi males. Additive deficits did not occur when males without OA and dVGLUT in OGNs were compared to males lacking only OA (Fig 6B-6D) indicating that both signals, at least partially, converge onto a shared aggression-promoting pathway.
Here, increased levels of inter-male courtship due to the absence of OA supersedes or relieves the lack of UWE due to a reduction in dVGlut function (Fig 2). At this point, it is possible the UWE phenotype occurs via OA-modulated circuitry that involves other neurotransmitters [56] or the actions of OA occur at spatially distinct locations.

Spatial segregation of VMAT and dVGLUT within OGN
To compare localization of the two transporters within OGNs, we generated a conditionally expressible epitope-tagged version of VMAT, RSRT>STOP>RSRT-6XV5-VMAT, via genome editing. RSRT>STOP>RSRT-6XV5-VMAT has two insertions: 1) a STOP cassette between VMAT coding exons 5 and 6 and, 2) six in-frame tandem copies of a V5 epitope tag within exon 8 which is common to both VMAT-A and VMAT-B isoforms (Fig 7A). The effectiveness of the STOP cassette is confirmed by the lack of V5 expression prior to STOP cassette excision by Gal4-driven R recombinase (S11 Fig) and the effectiveness of the epitope multimerization strategy has also been determined [68]. The conditionality of the RSRT>STOP>RSRT-6XV5-VMAT allele permits visualization of VMAT in subsets of neurons at expression levels driven by the endogenous promoter.
To focus on transporter distribution within OGNs, we expressed RSRT>STOP>RSRT-6XV5-VMAT under control of the split Gal4 combination of tdc2-Gal4-AD and dVGlut-Gal4-DBD (tdc2-dVGlut-gal4) which drives expression in OGNs (Fig 7B, S6C-S6F Fig). V5-VMAT was visualized in tdc2-dVGlut-gal4; V5-VMAT UAS-R by an antibody to V5 and dVGLUT using mAb dVGLUT (S10 Fig). Fig 7C illustrates that as expected, a large fraction of the V5-VMAT puncta in the AL or SEZ (S11 Fig) either co-localize with dVGLUT or are in close proximity (arrowheads). High resolution images in Fig 7D and 7H, however, reveal V5-VMAT puncta without dVGLUT (arrows). As OA can be found in SVs as well as LDCVs [69,70], we incorporated a synaptic marker (UAS-Synaptotagmin (Syt):HA) and re-examined V5-VMAT and dVGLUT expression in the AL and SEZ (Fig 7F, S11D Fig). We found V5-VMAT puncta that either co-localize or are in close proximity to Syt:HA and dVGLUT (Fig 7F-7J, S11D-S11H Fig). While the behavioral significance of potential OA synaptic release on aggression circuitry remains to be determined, previous work has demonstrated amine-dependent behaviors can be altered by shifting the balance of OA release from SVs to LDCVs [70]. In addition, as mentioned above, we have previously shown that three OA-FruM + neurons are required to suppress intermale courtship and recent work has identified a small subset of OA receptor OAMB-expressing neurons that when silenced, decrease aggression and increase intermale courtship [56]. The SEZ areas of V5-VMAT and dVGLUT puncta highlighted in Figs 7 and 8 are consistent with projections made by OA-FruM + neurons which are also OGNs (S12 Fig) raising the possibility of distinct OA and GLU inputs to key downstream targets.
Due to the large number of tdc2-dVGlut-gal4 neurons, we repeated the experiment using the OA-specific MB113C-split-gal4 to drive V5-VMAT in~2 OGNs (Fig 8A and 8B) [71]. Fig  8C illustrates that as expected, many V5-VMAT puncta in the SEZ either co-localize with dVGLUT or are in close proximity (arrowheads). High resolution images in Fig 8D and 8H, however, indicate small, but distinct regions that contain V5-VMAT puncta without dVGLUT (arrows). Within the areas of dVGLUT and V5-VMAT possible colocalization, this level of analysis does not indicate whether the two transporters segregate into adjacent but distinct puncta, nor are questions of transporter colocalization on the same vesicles addressed. Nevertheless, our results demonstrate that within OGNs, V5-VMAT and dVGLUT puncta can differ in localization suggesting the aggression vs. intermale courtship phenotype differences may be due to spatial differences in signaling by glutamate and octopamine.

Discussion
Addressing the functional complexities of ''one neuron, multiple transmitters" is critical to understanding how neuron communication, circuit computation, and behavior can be regulated by a single neuron. Over many decades, significant progress has been made elucidating the functional properties of neurons co-expressing neuropeptides and small molecule neurotransmitters, where the neuropeptide acts as a co-transmitter and modulates the action of the neurotransmitter [5,6,72]. Only recently have studies begun to examine the functional significance of co-transmission by a fast-acting neurotransmitter and a slow-acting monoamine.
In this study, we demonstrated that OA neurons express dVGLUT and utilized a new genetic tool to remove dVGLUT in OA-glutamate neurons. Quantifying changes in the complex social behaviors of aggression and courtship revealed that dVGLUT in brain OGNs is required to promote aggressive behavior and a specific behavioral pattern, the lunge. In contrast, males deficient for dVGLUT function do not exhibit an increase in inter-male courtship. These results establish a previously undetermined role for dVGLUT in OA neurons located in the adult brain and reveal glutamate uncouples aggression from inter-male courtship. It has been suggested that classical neurotransmitters and monoamines present in the same neuron modulate each other's packaging into synaptic vesicles or after release via autoreceptors [9,49,[73][74][75]. For example, a reduction of dVGLUT in DA-glutamate neurons resulted in decreased AMPH-stimulated hyperlocomotion in Drosophila and mice suggesting a key function of dVGLUT is the mediation of vesicular DA content [12,49,76]. In this study, the independent behavioral changes suggests enhancing the packaging of OA into vesicles is not the sole function of dVGLUT co-expression and suggests differences in signaling by OA from OGNs on courtship-related circuitry.
Co-transmission can generate distinct circuit-level effects via multiple mechanisms. One mechanism includes spatial segregation; the release of two neurotransmitters or a neurotransmitter and monoamine from a single neuron occurring at different axon terminals or presynaptic zones. Recent studies examining this possible mechanism have described; (i) the release of GLU and DA from different synaptic vesicles in midbrain dopamine neurons [15,77] and (ii) the presence of VMAT and VGLUT microdomains in a subset of rodent mesoaccumbens DA neurons [78]. In this study, we expressed a new conditionally expressed epitope-tagged version of VMAT in OGNs and visualized endogenous dVGLUT via antibody labeling. Within OGNs, the colocalization of VMAT and dVGLUT puncta was not complete suggesting the observed behavioral phenotype differences may be due to spatial differences in OA signaling.
A second mechanism by which co-transmission may generate unique functional properties relies on activating distinct postsynaptic receptors. In Drosophila, recent work has identified a small population of male-specific neurons that express the alpha-like adrenergic receptor, OAMB, as aggression-promoting circuit-level neuronal targets of OA modulation independent of any effect on arousal [56] and separately knockdown of the Rdl GABAa receptor in a specific doublesex+ population stimulated male aggression [79]. Future experiments identifying downstream targets that express both glutamate and octopamine receptors would be informative, as well as using additional split-Gal4 lines to determine if segregation of transporters is a hallmark of the majority of OGNs. Finally, a third possible mechanism is Glu may be coreleased from OGNs and act on autoreceptors to regulate presynaptic OA release (reviewed in [75]).
Deciphering the signaling complexity that allows neural networks to integrate external stimuli with internal states to generate context-appropriate social behavior is a challenging endeavor. Neuromodulators including monoamines are released to signal changes in an animal's environment and positively or negatively reinforce network output. In invertebrates, a role for OA in responding to external chemosensory cues as well as promoting aggression has been well-established [43,47,56,[80][81][82][83]. In terms of identifying specific aggression circuitcomponents that utilize OA, previous results determined OA neurons directly receive malespecific pheromone information [43] and the aSP2 neurons serve as a hub through which OA can bias output from a multi-functional social behavior network towards aggression [56]. The ability of OA to bias behavioral decisions based on positive and negative reinforcement was also recently described for food odors [84]. In vertebrates, it has been proposed that DA-GLU cotransmission in the NAc medial shell might facilitate behavioral switching [85]. Our finding that the majority of OA neurons are glutamatergic, suggests that the complex social behavior of aggression may rely on small subsets of neurons that both signal the rapid temporal coding of critical external stimuli as well as the frequency coding of such stimuli resulting in the enhancement of this behavioral network. One implication of our finding regarding the separable OA-dependent inhibition of inter-male courtship is the possibility of identifying specific synapses or axon terminals that when activated gate two different behavioral outcomes. A second implication is that aggressive behavior in other systems may be modified by targeting GLU function in monoamine neurons.
Finally, monoamine-expressing neurons play key roles in human behavior including aggression and illnesses that have an aggressive component such as depression, addiction, anxiety, and Alzheimer's [86,87]. While progress is being made in addressing the functional complexities of dual transmission, the possible pathological implications of glutamate co-release by monoamine neurons remains virtually unknown. Analyzing the synaptic vesicle and release properties of monoamine-glutamate neurons could offer new possibilities for therapeutic interventions aimed at controlling out-of-context aggression.

Drosophila husbandry and stocks
All flies were reared on standard cornmeal-based fly food. Unless noted otherwise, during developmental and post-eclosion, flies were raised at 25˚C,~50% humidity and a 12:12hr light-dark cycle (1400±200 lx white fluorescent light) in humidity and temperature-controlled incubators. A list of stocks can be found in S1 Data.

Aggression assays
Male pupae were isolated and aged individually in 16 x 100mm borosilicate glass tubes containing 1.5ml of standard food medium as previously described [88]. A dab of white or blue acrylic paint was applied to the thorax of two-day old males under CO 2 anesthesia for identification purposes. Flies were returned to their respective isolation tubes for a period of at least 24 hours to allow recovery. For aggression testing, pairs of 3-5 day old, socially naïve adult males were placed in 12-well polystyrene plates (VWR #82050-930) as described previously [43]. All assays were run at 25˚C and~45-50% humidity levels.

Scoring and statistics
All aggression was assayed within first two hours of lights ON time (Zeitgeber hours 0-2) and scored manually using iMovie version 8.0.6. Total number of lunges, wing threats, and unilateral wing extensions were scored for a period of 30 minutes after the first lunge according to the criteria established previously [43,88]. The time between the aspiration of the flies into the chamber and the first lunge was used for calculating the latency to lunge. Male-male courtship was the number of unilateral wing extensions (singing) followed by abdomen bends or repeated wing extensions. All graphs were generated with Graphpad Prism and Adobe Illustrator CS6. For data that did not meet parametric assumptions, Kruskal-Wallis Test with Dunn's multiple comparison was used unless otherwise specified. A standard unpaired t-test was performed in the case of only two comparisons and a modified chi-square test to compare copulation success.

Activity levels
Activity levels were measured by tracking the flies in each assay using the OpenCV module in the Python programming language to analyze the video and then output XY-coordinate and distance data. The distance traveled was calculated for each fly by determining the starting location followed by the second location after a 250-ms time interval and then taking the sum of the distance traveled in each interval. To calculate pixels moved per second, the distance data was divided by the total time spent tracking.

Construction of UAS-B3
B3 recombinase derived from pJFRC157 [60] was PCR amplified using primers designed to add the syn21 translational enhancer sequence [94] and remove the PEST domain. The verified PCR product was cloned into pENTR (Invtrogen) and subsequently transferred to pBID20xUAS, a derivative of the pBID vector [95] with 20 copies of the UAS binding sequence. Injection of UAS-B3 was performed by Genetivision into landing site VK31.
To assess the functionality of dVGlut on the B3RT-dVGlut chromosome pre-and post-excision, the B3RT-dVGlut chromosome was crossed with the homozygous lethal dVGlut null allele, dVGlut SS1 in the presence and absence of the pan-neuronal driver n-syb-Gal4. In the absence of a Gal4 driver, dVGlut SS1 /B3RT-dVGlut progeny are fully viable and no LexAopdriven reporter gene expression is detected (Fig 2). When B3 recombinase (UAS-B3) is expressed in the nervous system by n-syb-Gal4, dVGlut SS1 /B3RT-dVGlut;UAS-B3/n-syb-Gal4 progeny are inviable, therefore after excision, as expected with removal of the entire dVGlut protein-coding sequence, a dVGlut null allele results.

Generation of dVGlut SS1
The dVGlut SS1 allele was generated by CRISPR/Cas9 genome editing with the same guide RNAs used to generate B3RT-dVGlut LexA. dVGlut SS1 was identified based on failed complementation with the existing dVGlut 2 allele [97]. Sequencing of PCR products from this allele indicated a deletion of 2442bp that includes dVGlut amino acids 53-523. Genomic DNA sequence at the breakpoints of the dVGlut SS1 allele are indicated with the deleted region in bold: GGACCAGGCGGCGGCCACGC. . . . . .AACCTCCGGCCGAGGAGCAA.
The R and B3 recombinases from yeast recognize sequence-specific recombination target sites, RSRTs and B3RTs, respectively [60]. These recombinases are highly efficient and highly specific as they exhibit virtually no cross-reactivity with each other's recombinase target sites. When pairs of recombinase target sites are in the same orientation, as is the case for both B3RT-vGlut-LexA and RSRT-STOP-RSRT-6XV5-vMAT, the recombinases catalyze excision of the intervening DNA and leave behind a single recombinase target site.

dVGlut antibody
Drosophila anti-dVGLUT mouse monoclonal antibodies (10D6G) were generated (Life Technologies Europe) using the C-terminal peptide sequence TQGQMPSYDPQGYQQQ of dVGLUT coupled to KLH.