PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
 
Nat Neurosci. Author manuscript; available in PMC Jul 9, 2012.
Published in final edited form as:
Published online Dec 26, 2010. doi:  10.1038/nn.2716
PMCID: PMC3391700
NIHMSID: NIHMS380793
Autoregulatory and paracrine control of synaptic and behavioral plasticity by octopaminergic signaling
Alex C Koon, James Ashley, Romina Barria, Shamik DasGupta, Ruth Brain, Scott Waddell, Mark J Alkema, and Vivian Budnik
Department of Neurobiology, University of Massachusetts Medical School, Worcester, Massachusetts, USA
Correspondence should be addressed to V.B. (vivian.budnik/at/umassmed.edu)
Adrenergic signaling has important roles in synaptic plasticity and metaplasticity. However, the underlying mechanisms of these functions remain poorly understood. We investigated the role of octopamine, the invertebrate counterpart of adrenaline and noradrenaline, in synaptic and behavioral plasticity in Drosophila. We found that an increase in locomotor speed induced by food deprivation was accompanied by an activity- and octopamine-dependent extension of octopaminergic arbors and that the formation and maintenance of these arbors required electrical activity. Growth of octopaminergic arbors was controlled by a cAMP- and CREB-dependent positive-feedback mechanism that required Octpβ2R octopamine autoreceptors. Notably, this autoregulation was necessary for the locomotor response. In addition, octopamine neurons regulated the expansion of excitatory glutamatergic neuromuscular arbors through Octpβ2Rs on glutamatergic motor neurons. Our results provide a mechanism for global regulation of excitatory synapses, presumably to maintain synaptic and behavioral plasticity in a dynamic range.
Synaptic plasticity is fundamental for an organism’s ability to adapt to a changing environment. Adrenergic receptors and their ligands are key regulators of plasticity. Noradrenaline has been implicated in the retrieval of spatial and contextual memories1, and it enhances LTP by promoting the synaptic delivery of AMPA-type glutamate receptors (GluR)2. Adrenergic signaling has also been implicated in the regulation of plasticity (also called metaplasticity) to reset a homeostatic circuit in response to acute perturbations, thus maintaining the circuit within a dynamic range3. However, the specific mechanisms by which adrenergic signals influence synaptic plasticity are poorly understood.
Octopamine, the invertebrate counterpart of adrenergic ligands, activates receptors that resemble adrenergic receptors4. Octopamine is important for appetitive reinforcement in honeybees5 and flies6,7 and modulates behaviors such as aggression8, egg-laying9, food-seeking10 and sleep11, as well as synaptic functions12.
The Drosophila larval neuromuscular junction (NMJ) is a powerful model system in which to investigate synaptic plasticity. Although glutamate is the primary excitatory neurotransmitter of the NMJ, larval NMJs are also innervated by octopaminergic motor neurons13. Larval NMJs show several forms of synaptic plasticity, such as continuous expansion during larval development to offset a massive increase in muscle size, as part of a homeostatic mechanism to maintain synaptic efficacy14. This process depends on signaling mechanisms such as the bone morphogenetic protein (BMP)15 and Wnt pathways16. Larval NMJs can also respond to changes in the environment such as food availability by rapid increases in synapse strength17,18. In addition, genetic and physiological manipulations that increase presynaptic activity promote synaptic expansion at NMJs19,20. To determine the relevance of octopaminergic innervation of body-wall muscles, we examined octopaminergic terminals during larval foraging behavior. Type II arbors responded to food deprivation by extending new endings. This effect depended on both activity levels and octopamine. Electrical activity at octopaminergic neurons was essential for initial and continued type II innervation of muscles. We uncovered a cAMP and CREB-dependent autoregulatory positive feedback mechanism that regulated the size of type II endings through the activation of Octβ2R autoreceptors. Type II innervation also regulated the plasticity of glutamatergic type I motor neurons through Octβ2Rs expressed in these neurons. Both the autocrine and paracrine mechanisms were required for the adaptive response to starvation.
Locomotor increase associated with type II synaptic change
Larval NMJs respond to acute changes in presynaptic activity by modifications in synaptic structure20. However, the physiological conditions under which this mechanism is used by the intact organism are unknown. Larval foraging behavior is enhanced by food deprivation, which leads to long-lasting enhancement of evoked glutamate release from excitatory type I NMJs17. However, no gross changes in the structure of these endings have been observed17. Most body-wall muscles are co-innervated by at least one additional class of motor neuron, the octopaminergic type II motor neuron13 (Fig. 1a). Octopamine signaling has been implicated in appetitive behaviors and locomotion6,10,21,22. Therefore, to determine whether type II arbors changed structure during starvation, a physiological stimulus that increases locomotor activity, we labeled these arbors by expressing mCD8-GFP using a tyrosine decarboxylase-2 (Tdc2) promoter fused to Gat4 (Tdc2-Gal4; Fig. 1a). We imaged NMJs in intact early third-instar larvae live through the cuticle, deprived the larvae of food for 2 h and then imaged the same NMJs again (Fig. 1).
Figure 1
Figure 1
Food-deprivation increase in larval locomotion is correlated with synaptopod formation at type II arbors, (a) NMJs at muscles 12 and 13 of a third-instar larva expressing mCD8-GFP in type II motor neurons, showing type I and type II endings (arrows), (more ...)
Food-deprived wild-type larvae showed a significant increase in locomotor speed compared to fed controls (Fig. 1c). Notably, type II endings in starved larvae showed dynamic filopodia-like extensions (synaptopods) that extended and retracted with a time course of minutes (Fig. 1b,d and Supplementary Movie 1). Although we also saw synaptopods before food deprivation (‘natural synaptopods’; Fig. 1b), the number of synaptopods was significantly increased upon starvation (Fig. 1b,d). Thus, changes in locomotor activity were accompanied by structural changes at type II endings.
We next investigated whether type II endings were necessary for behavior. We eliminated octopaminergic neurons by expressing the cell-death protein head involution defective (Hid; Supplementary Fig. 1), which substantially reduced locomotor speed and the starvation response (Fig. 1e,f). A similar result has been observed in tyrosine beta-hydroxylase (tbhnMI8) and tdc2R054 mutants, which cannot synthesize octopamine9,23 (Fig. 1e,f). The defects in tbh mutants were specific, as they were rescued by expressing a TBH transgene in octopaminergic neurons (Fig. 1e,f). Thus, the increase in locomotion elicited by food deprivation results in structural changes in octopaminergic endings, and octopamine innervation is necessary for this behavior.
We then investigated whether octopamine was sufficient to increase locomotor activity in the absence of starvation. We expressed channel rhodopsin-2 (ChR2) in octopaminergic neurons and stimulated the neurons with blue light before the locomotion assay. Light-stimulated animals showed a significant increase in locomotor speed and this effect was eliminated in tbh mutants (Fig. 1g). Thus, octopamine neurons are necessary and sufficient to increase locomotion.
To determine whether octopaminergic innervation of body-wall muscles alone was sufficient to induce modifications in synaptic physiology, we applied exogenous octopamine to body-wall muscles devoid of central input. Bath application of 10 μM octopamine elicited a 30% increase in the amplitude of excitatory junctional potentials (EJPs) without any change in the amplitude of miniature EJPs (mEJPs; Fig. 1h,j). This was consistent with analysis of tbh mutants, in which EJP amplitude was significantly decreased (Fig. 1i,j). Thus, changes in locomotor activity were accompanied by growth of synaptopods and probably by an increase in synaptic strength.
Synaptopod extension preceded type II synapse formation
We determined the physiological significance of synaptopods (Fig. 2) by examining their dynamics (Fig. 2a and Supplementary Movie 1). Many synaptopods formed varicosity-like structures at their tips, measuring 0.6 ± 0.04 μm in diameter (Fig. 2b and Supplementary Movie 2), smaller than mature type II boutons (1.57 ± 0.05 μm). When new varicosities formed, synaptopod motility halted (Supplementary Movie 2), the varicosity enlarged and sometimes a new motile synaptopod emerged from that varicosity (secondary synaptopod; Fig. 2c and Supplementary Movie 3). Thus, synaptopod formation could be a mechanism for type II arbor extension. We investigated this possibility by examining the same type II NMJs from first- to third-instar larval stage. First-instar type II arbors had synaptopods (Fig. 2d, for example, red arrow; Supplementary Fig. 2a) and synaptopods containing a varicosity at their tips (Fig. 2d, for example, yellow arrowhead in inset; Supplementary Fig. 2a). These structures developed into a completely new or extended a type II branch (Fig. 2d and Supplementary Fig. 2a). To show that newly formed varicosities corresponded to new boutons, we imaged larvae expressing both mCherry and synaptotagmin-1 (Syt1)-GFP as above. Syt1-GFP accumulated in the newly formed varicosities (Fig. 2e). Thus, synaptopod extension is a mechanism for expanding type II arbors, both during an acute increase in locomotor speed and during larval development.
Figure 2
Figure 2
Stepwise development of synaptopods. (a–c) Time lapse imaging of synaptopods in Tdc2>mCD8-GFP larvae showing the extension of synaptopods (a. arrows), the formation of varicosities at the tip of synaptopods (b, arrows) and the formation (more ...)
We also labeled preparations with antibodies to different synaptic markers. FasciclinII (FasII) was present at 100% of synaptopods from the earliest stages (Fig. 2f and Supplementary Fig. 2b), and was maintained throughout subsequent stages of bouton maturation (Fig. 2l,m). By contrast, we found Syt1 in synaptopods only 40% of the time (Fig. 2g,l,m and Supplementary Fig. 2c) but it was always present at the onset of varicosity formation (Fig. 2l,m and Supplementary Fig. 2d), suggesting that vesicles begin to traffic into synaptopods even before the formation of type II varicosities.
We determined the onset of octopamine synthesis using antibodies to TBH (Supplementary Fig. 3a–f) and found that TBH was never observed before the onset of new varicosity formation (Fig. 2h,l,m and Supplementary Fig. 2b–f), which suggests that the accumulation of another type of synaptic vesicle, marked by Syt1, preceded the accumulation of TBH-containing vesicles. We identified active zones using antibodies to Bruchpilot (Brp; Elks/Cast/Erc homolog)24, which was observed in 27% of enlarged varicosities and after the appearance of TBH20,25,26 (Fig. 2i,l,m and Supplementary Fig. 2e). Finally, we observed the MAPI B-related protein Futsch only after a secondary varicosity was formed, but the immunoreactivity was punctate (Fig. 2j,l,m and Supplementary Fig. 2f). We could not determine when postsynaptic GluRs were first present, as the immunoreactivity was very low (Supplementary Fig. 4a). The above data further support the notion that synaptopod extension constitutes a mechanism for the formation of new type II synaptic boutons. Even in intact larvae, many of the synaptopods that formed after starvation developed varicosities (Fig. 2k). These observations also show that the formation of type II boutons follows a precise sequence of synaptic protein addition (Fig. 2l).
Acute activity and octopamine initiate type II outgrowth
The structural changes observed at type II boutons in intact larvae raised the possibility that octopaminergic neurons were activated during food deprivation or increased locomotion, leading to the expansion of type II arbors. We tested this hypothesis by increasing motor neuron activity either with high-K+-induced depolarization or by blue light stimulation of ChR2 expressed in octopaminergic neurons (Fig. 3). We subjected preparations to spaced stimulation, in which 5 cycles of stimulation with high-K+ or blue light were separated by 15-min rest periods20. We imaged identified type II endings before and after stimulation. Stimulated samples showed a significant increase in synaptopod number at type II endings after stimulation (Fig. 3a,c). As in intact larvae, unstimulated preparations also contained natural synaptopods, albeit at a lower frequency (Fig. 3a,b). Thus, similar to the starvation response, synaptopods at type II endings increased in frequency in response to spaced stimulation, and stimulation of octopamine neurons alone was sufficient to elicit this response. This was confirmed by genetically increasing activity at octopaminergic neurons by expressing a dominant-negative Shaker K+ channel subunit (ShDN)27 in type II motor neurons of an eag K+ channel subunit mutant, which resulted in an increase in natural synaptopods (Fig. 3d).
Figure 3
Figure 3
Electrical activity and octopamine regulate the extension of synaptopods. (a,b) Live imaging of synaptopods before and after stimulation with high K+ (a) or octopamine (b) in Tdc2>mCD8-GFP larvae. (c) Net increase in synaptopod number ~2 h after (more ...)
Next, we sought to determine whether octopamine signaling could underlie the effects of activity. The elimination of octopamine in tbh-null mutants resulted in a significant decrease in natural synaptopods (Fig. 3d). By contrast, bath application of octopamine for 15 min to wild-type preparations resulted in a dose-dependent increase in the number of synaptopods, whereas tyramine application had no effect (Fig. 3b,e,f). Induction of synaptopods by octopamine required normal (1.5 mM) Ca2+ levels, as decreasing Ca2+ to 0.1 mM prevented this effect (Fig. 3e, sub-Ca2+).
We determined the relationship between activity and octopamine by stimulating the terminals with activity or octopamine at levels that did not reach the threshold for induction of synaptopods. If the effect of activity was to increase octopamine release, then presenting both subthreshold stimuli together should elicit significant synaptopod formation. The subthreshold stimuli consisted of three cycles of spaced depolarization and application of 10 μM octopamine in 0.1 mM Ca2+, both of which were insufficient to induce synaptopods when presented alone (Fig. 3g). When applied together, however, they increased synaptopods to a level similar to that induced by five cycles of stimulation alone (Fig. 3g). Thus, exogenous octopamine can overcome the effect of insufficient activity for the induction of synaptopods and vice versa, consistent with the notion that synaptopod formation is the result of activity-dependent octopamine release.
Octopamine failed to induce synaptopods in tbh mutants (Fig. 3e). Given that tbh mutants have an accumulation of tyramine9 that might be developmentally deleterious, we also tested tdc2 mutants, which lack tyramine accumulation. In tdc2 mutants the response to octopamine was normal (Fig. 3e), suggesting that in tbh mutants the accumulation of tyramine renders the NMJs insensitive to exogenous octopamine.
We also examined synaptic growth by counting the number of type II boutons at the last stage of larval development. Increasing activity through expression of ShDN in octopaminergic neurons of eag mutants led to a significant increase in the number of type II boutons and terminal branches (Fig. 3h and Supplementary Fig. 5a). By contrast, in tbh mutants the number of type II boutons and branches was decreased (Fig. 3h and Supplementary Fig. 4b). These phenolypes were specifically rescued by expressing a tbh transgene in octopaminergic neurons (Fig. 3h and Supplementary Fig. 4b). The defect in tbh mutants was not due to the accumulation of tyramine in these mutants, as a null mutant in tdc2, which lacks both tyramine and octopamine, also showed a decrease in the number of type II boutons (Fig. 3h).
Type II synaptogenesis and maintenance require activity
We also used transgenic approaches to block activity at octopaminergic neurons (Fig. 4). We used constructs encoding ShiDNts, which blocks vesicle recycling at restrictive temperatures28, EKO, a hyperpolarizing Shaker potassium channel29, and Kir2.1, which encodes an inward-rectifying K+ channel that prevents membrane depolarization30. We tested the efficiency of the blockade by examining the ability of adult females to lay eggs, as octopamine function is required for egg laying9. Only expression of Kir2.1 completely blocked egg laying. Strikingly, it also resulted in the complete elimination of type II innervation (Fig. 4a,b). This was not due to a pathfinding or cell death defect because type II motor neuron axons, labeled with mCD8-GFP, were always observed in the segmental nerves (Fig. 4c,d). In 69% of the nerves examined, these axons stalled in the segmental nerve. However, in 31% of the cases axons traveled the entire distance from the CNS and stalled close to the NMJ without innervating the muscles (Fig. 4d). Thus, in the absence of activity, type II endings could not innervate body-wall muscles.
Figure 4
Figure 4
Innervation and maintenance of type II arbors depends on activity. (a,b) NMJs at muscles 12 and 13 in preparations expressing mCD8-GFP in octopamine neurons and double labeled with anti-GFP and anti-HRP antibodies in Tdc2>mCD8-GFP (a) and Tdc2>mCD8-GFP, (more ...)
Although ShiDNts did not completely eliminate octopaminergic function, expressing this transgene and rearing the animals at the restrictive temperature of 29 °C were sufficient to elicit marked abnormalities in the innervation of muscles by type II endings. These included markedly reduced type II arbors (Supplementary Fig. 5b), lack of innervation of muscles by type II arbors (Supplementary Fig. 5d), thinning of type II neurites (Supplementary Fig. 5f,h) and lack of TBH in some type II boutons (Supplementary Fig. 5h).
To determine whether there was a critical period in which activity was required for type II innervation, and whether the lack of innervation was the result of activity-dependent synaptogenesis or degeneration, we determined when type II innervation was established during the larval period and the consequences of blocking activity at these stages. We first observed type II varicosities during the first-instar stage (Fig. 4e,g). Blocking activity eliminated type II boutons at any larval stage (Fig. 4f,h), suggesting that the absence of activity in type II motor neurons prevents synaptogenesis.
We also ubiquitously expressed a temperature-sensitive Gal80, which at 18 °C blocks Gal4-mediated expression31, and simultaneously expressed Kir2.1 in octopaminergic neurons. Larvae were raised at 18 °C, and then switched to 29 °C at different stages to permit expression of Kir2.1. Suppressing the activity of octopaminergic neurons 24 h before the third-instar stage did not elicit any abnormality in type II bouton morphology (Fig. 4i). By contrast, blocking activity starting from the late second-instar stage resulted in breaks in type II arbors (Fig. 4j). This phenotype was most pronounced when activity was blocked from the late first-instar stage (Fig. 4k–m). Thus, activity in type II endings was required for synaptogenesis, whereas prolonged periods of inactivity after innervation led to the degeneration of type II endings.
Octopamine-induced type II growth requires cAMP and dCREB
The finding that octopamine induced the growth of type II arbors suggested that an autoregulatory mechanism controlled the formation of type II endings. Octopamine could be acting on autoreceptors at type II endings, or octopamine might activate a retrograde signal that promotes the growth of type II boutons. Octopamine receptors are G-protein-coupled receptors, which can increase Ca2+ or cAMP4,32. Previous studies suggested that increasing cAMP levels by a mutation in the phosphodisterase Dunce (Dnc) enhanced synaptic growth at all boutons33. Therefore we examined whether manipulating the levels of cAMP in type II motor neurons could influence synaptopod formation in response to octopamine (Fig. 5). Mutations in dnc significantly enhanced the number of naturally occurring synaptopods at type II endings (Fig. 5a,b,d). This phenotype was rescued by expressing Dnc exclusively in the octopaminergic neurons of dnc mutants or by a genomic duplication (Dp) of dnc (Fig. 5d). Thus, octopamine motor neurons contain a cAMP pathway that can promote synaptopod formation.
Figure 5
Figure 5
Synaptopod extension is regulated by the cAMP pathway and requires new protein synthesis and CREB. (a–c) Live imaging of synaptopods (arrows) in Tdc2>mCD8-GFP (a, WT control), dncM14, Tdc2>mCD8-GFP flies (b) and Tdc2>mCD8- (more ...)
These observations were confirmed by using a mutation in rutabaga (rut), which encodes an adenylate cyclase, and thereby decreasing cAMP levels. Flies with the rut2080 mutation had significantly fewer natural synaptopods than wild-type flies and this phenotype was rescued by expressing Rut at octopaminergic neurons (Fig. 5d).
To determine whether cAMP was acutely sufficient for synaptopod induction we elevated cAMP using a photoactivatable adenylate cyclase (PACα)34. We expressed PACα in octopaminergic neurons and stimulated the preparation with blue light using a spaced procedure. There was a significant increase in the number of synaptopods (Fig. 5c,e and Supplementary Movie 4), which showed that acute changes in cAMP were sufficient to induce synaptopods.
To determine whether cAMP was downstream of activity and octopamine in the induction of synaptopods, we applied octopamine to dnc and rut mutants. As described above, dncM14 mutants have more naturally occurring synaptopods than the wild type, whereas rut2080 mutants have fewer. In dnc mutants we expected that the increase in synaptopods by octopamine would be occluded, as natural synaptopods are already saturated in this mutant. In rut mutants, we expected that the decrease in adenylate cyclase activity would render NMJs unresponsive to octopamine. Consistent with these predictions, octopamine failed to increase the number of synaptopods in the two mutants (Fig. 5f) By contrast, increasing cAMP by activating PACα in a tbh mutant background still induced synaptopods (Fig. 5e). Thus the cAMP pathway is probably downstream of octopamine in synaptopod induction.
In long-term plasticity, the cAMP pathway has been associated with the activation of CREB leading to the transcription of genes that are required for the formation of new synapses35. We sought to determine whether the growth of type II endings in response to octopamine release required dCREB function. A dCREB dominant-negative transgene (CREBdn/dCREB2-b) has been shown to block CREB function36; we expressed this transgene in octopamine neurons and found that it suppressed the increase in octopamine-induced synaptopod formation (Fig. 5g). Similarly, the translational inhibitor cycloheximide and the transcriptional inhibitor actinomycin-D completely suppressed octopamine-dependent synaptopod formation (Fig. 5g). Thus, the autoregulatory mechanism that initiates the formation of new type II boutons activates a cAMP cascade that depends on CREB-mediated transcription.
As perturbations in the cAMP pathway altered synaptopod formation, and mutations that prevented an increase in synaptopod formation in response to octopamine also caused behavioral defects, we predicted that dnc and rut mutants would show similar defects. Although dncM14 mutants showed decreased locomotor speed and rut mutants had normal locomotor speed (Fig. 5h), the response to starvation was blocked in both mutants, similar to tbh mutants (Fig. 5i). The behavioral defect in rut mutants was completely rescued by expression of a Rut transgene in octopaminergic neurons (Fig. 5i). Thus, normal cAMP levels in octopaminergic neurons are required for the response to starvation and, at least in the case of Rut, the starvation response can be separated from a defect in basal locomotion.
Octβ2R autoreceptors mediate the autoregulatory mechanism
Presynaptic octopamine autoreceptors are plausible candidates for mediating autoregulation of synaptic structure. Four octopamine receptors have been identified in the Drosophila genome—OAMB, Octβ1R/OA2, Octβ2R and Octβ3R. OAMB receptors are homologous to mammalian α-adrenergic receptors and can increase levels of Ca2+ or cAMP4,32. Octβ1R, Octβ2R and Octβ3R receptors have been less studied, but share similarities with mammalian β-adrenergic receptors and are thought to increase cAMP. We interfered with the function of OAMB and Octβ2R. In the oamb584 genetic null allele37, the number of natural synaptopods, the induction of synaptopods by octopamine and the number of type II boutons were normal (Fig. 6a–c).
Figure 6
Figure 6
Presynaptic Octβ2R autoreceptors, but not OAMB receptors, regulate the growth of type II arbors. (a) Number of natural synaptopods in the indicated genotypes (n = 175, 12, 13, 25, 14); WT control is Tdc2>mCD8-GFP. (b) Number of type II (more ...)
To block the function of Octβ2R we used a hypomorphic allele11 as well as RNA interference (UAS-Octβ2R-RNAi). We also verified the expression of Octβ2R in the nervous system and body-wall muscles (either pre- or postsynaptically) and the effectiveness of the RNAi transgene by RT-PCR (Supplementary Fig. 6a,b). In octβ2R mutants the number of natural synaptopods was significantly decreased (Fig. 6a), which indicates that Octβ2R is the likely mediator of the cAMP-dependent autoregulatory mechanism that controls the growth of type II endings. Accordingly, octβ2R mutants did not show increased synaptopods in response to octopamine (Fig. 6c). This effect was cell autonomous, as expressing Octβ2R-RNAi in octopaminergic neurons alone was sufficient to decrease the number of natural synaptopods (Fig. 6a) and to suppress the ability of octopamine to induce an increase in synaptopods (Fig. 6c). In addition, either the octβ2R mutation or expression of Octβ2R-RNAi in octopamine neurons resulted in a significant decrease in the number of type II boutons (Fig. 6b). We used C380-Gal4 to express Octβ2R-RNAi in both type I and type II motor neurons, which also caused a significant reduction in the number of type II boutons (Fig. 6b). By contrast, the expression of Octβ2R-RNAi in muscles using C57-Gal4 had no effect. These observations identify Octβ2R receptors as the likely autoreceptor that regulates the growth of type II boutons.
The octβ2R mutant showed a reduction in evoked RJP amplitude similar to that seen in tbh mutants (FJP amplitude, 18.3 ± 1 mV in wild type versus 12.1 ± 1 mV in octβ2R mutants; n = 7, P < 0.001), which suggests that the removal of either octopamine or its receptors decreases synaptic strength. As expected, bath application of octopamine to octβ2R mutants did not change EJP amplitude in mutants as it did in the wild type (ratio EJP amplitude is 0.95 upon 10 μM octopamine application versus 1.33 in wild type; n = 5, P < 0.005). Thus, Octβ2R receptors are probably responsible for the octopamine-induced changes in synaptic strength.
We predicted that removing the receptor would also eliminate the starvation response. Flies with mutations in octβ2R failed to respond to starvation by increasing locomotor speed (Fig. 6d). This detect was also observed when Octβ2R was downregulated either in octopamine neurons or in both type I and type II motor neurons, but not in muscles (Fig. 6d).
Type II endings regulate type I synaptic bouton outgrowth
Excitatory transmission at the larval NMJ is mediated by the release of glutamate from type 1 NMJs. Like type II arbors, type I arbors expand continuously throughout larval development, in strong correlation to muscle size38. We considered the possibility that type II innervation might regulate this form of plasticity at type I boutons. To test this possibility, we first eliminated type II boutons by expressing Hid in octopaminergic neurons. The absence of type II innervation led to a substantial reduction in the number of type I boutons (Fig. 7a and Supplementary Fig. 6c). We obtained similar results when we expressed Kir2.1 or in tbh mutants (Fig. 7a). The reduction in the number of type I boutons in tbh mutants was restored by expressing a tbh transgene in octopamine neurons (Fig. 7a). These results suggest that type II innervation regulates the plasticity of type I endings and therefore that it is involved in a form of metaplasticity.
Figure 7
Figure 7
Type II motor neurons regulate the growth of type I arbors. (a) Number of type I boutons at muscle 6 and 7 (A3) in third-instar larvae after eliminating either type II motor neurons or octopamine production from type II motor neurons (n = 12, 17, 21, (more ...)
To further characterize the influence of type II endings on type I arbors, we examined octβ2R mutants, and found a reduction in the number of type I boutons (Fig. 7b). A potential mechanism by which type II arbors might regulate the growth of type I endings is through the presence of octopamine receptors at type I boutons. Therefore, we used the C380-Gal4-driver to downregulate Octβ2R in both type I and type II motor neurons or the BG439-Gal4 driver to express Gal4 in type I motor neurons but not in type II motor neurons (Supplementary Fig. 7). Both manipulations resulted in a significant decrease in the number of type I boutons (Fig. 7b). By contrast, the downregulation of Octβ2R in octopamine neurons alone did not result in a significant decrease in the number of type I boutons. These results suggest that Octβ2R is required in type I motor neurons for normal expansion.
Adrenergic signaling is involved in the regulation of synaptic plasticity39,40. However, the precise mechanisms of this regulation are poorly understood. We show that octopamine regulates behavioral and synaptic plasticity through an autoregulatory mechanism that promotes the growth of type II innervation and in turn the expansion of excitatory glutamatergic arbors. This process seems to be associated with physiological stimuli that lead to increased locomotion. We propose that food deprivation elicits the release of octopamine by type II terminals. Octopamine binds to Octβ2R receptors and thereby increases cAMP, which activates CREB-dependent regulation of transcription and leads to new type II synaptic growth (Supplementary Fig. 8a). This autoregulatory mechanism might control the amount of octopamine released by type II arbors. In turn, octopamine release stimulates the growth of type I arbors through Octβ2R at type I motor neurons. This mechanism would regulate, in a global fashion, excitatory transmission at the NMJ (Supplementary Fig. 8a).
Increases in larval locomotion, type II motor neuron activity or exogenous octopamine resulted in the extension of synaptopods. With the demonstration that synaptopod extension constitutes a mechanism for the formation of type II boutons, these results suggest that the above events control the growth of octopaminergic endings in an acute manner. Analysis of mutations in octopamine receptors and components of the cAMP cascade revealed an autoregulatory mechanism that controls this growth. First, expression of Octβ2R in type II motor neurons was required for type II synaptic growth. Second, altering cAMP levels by mutations in dnc or rut modified this response in a manner consistent with positive regulation by cAMP. This regulation was cell autonomous in octopaminergic motor neurons, as the defects in synaptopod formation and type II synaptic growth were also elicited or rescued by transgene expression in octopaminergic motor neurons alone, in a chronic or acute manner. The presence of auto-octopamine receptors had been suggested in locusts41, although the identity of the proposed autoreceptor was not known. However, it was proposed that the locust octopamine autoreceptors served to inhibit octopamine release. By contrast, our experiments are consistent with a positive feedback mechanism that enhances synaptic growth. Autoregulatory mechanisms that control the amount of neuromodulator release have been previously demonstrated for neuromodulators such as dopamine42.
As in other forms of synaptic plasticity, including late LTP and long-term memory35, the autoregulatory mechanism required the function of CREB and new protein synthesis. This finding underscores the universality of mechanisms by which the nervous system modifies the efficacy of connections in a long-lasting manner. Octopamine receptor activation leading to CREB signaling has also been demonstrated in Caenorhabditis elegans10.
Our studies showed that this pathway regulated the structure of octopaminergic arbors in an autoregulatory fashion, and that this influenced the growth of type I excitatory arbors. The presence of a positive feedback that controls the growth of modulatory inputs in an acute manner provides a mechanism by which animal experience can modify circuitry and thus by which animals can adapt to a changing environment.
Activity was absolutely required for innervation of body-wall muscles by type II arbors, as reduced activity perturbed type II synaptogenesis. This is in contrast to the widely held view that although activity is important for the refinement of connections, it is not required for initial synaptogenesis43. Part of this view arises from the examination of arbors that mediate classical neurotransmission43. By contrast, the dependence of modulatory terminal growth on activity has been less studied. Our studies provide compelling evidence that octopamine has an influence on bouton outgrowth in octopaminergic type II and type I motor neurons. Studies of type I bouton outgrowth have identified local factors that influence the development of pre- or postsynaptic compartments, including Wnts and BMPs15,16. We suggest that octopamine release by type II arbors might mediate a more global regulation of outgrowth.
At the Drosophila larval NMJ, glutamatergic type Ib motor neurons innervate each muscle in an approximately 1:1 manner44 (Supplementary Fig. 8b, type Ib). In addition, two glutamatergic type Is motor neurons innervate the entire ventral or dorsal muscle field within each hemisegment44,43 (Supplementary Fig. 8b, type Is). By contrast, the three octopaminergic neurons per segment innervate most of the body-wall muscles in a bilateral fashion13 (Supplementary Fig. 8b, type II). The layout of this innervation suggests that type II synapses might establish global regulation of the plasticity of type I arbors. This might serve as a mechanism for setting excitability levels in the entire body wall, and thereby keep synaptic function in a dynamic range. Similarly, studies in mammalian systems have shown that adrenergic signaling can affect plasticity at glutamatergic synapses, either through changes in ionotropic GluR localization2 or through regulation of metabotropic GluR, which affects the ability of a synapse to become potentiated depending on its history3. Octopamine might regulate the ability of type I NMJs to trigger muscle contraction by long-term regulation of type I synaptic growth.
Two previous studies at the Drosophila larval NMJ have shown that octopamine enhances synaptic transmission46,47. However, another study reported that octopamine might inhibit glutamatergic transmission in first-inslar larvae48. Our studies suggest that blocking activity or interfering with octopamine signaling in type II neurons leads to a decrease in type I synaptic outgrowth, consistent with the idea that octopamine release is a positive regulator of type I transmission. We suggest that in the short term, octopamine enhances synaptic strength, as observed in our electrophysiology experiments, leading to the observed increase in crawling behavior after starvation. This would be consistent with studies showing that increases in locomotor speed induced by food deprivation led to an enhancement of synaptic efficacy17.
Octopamine is a potent modulator of invertebrate behavior and is secreted during starvation in invertebrates10,22. Nevertheless, its function at the synaptic level is poorly understood. Our study shows that octopamine can influence synapses at the structural level through the activation of Octβ2R autoreceptors in octopamine neurons and through the presence of these receptors in type I motor neurons.
An important question is whether octopamine is simply involved in locomotion and the lack of starvation response in mutants that cannot synthesize octopamine is an indirect effect of defective locomotion. It is not possible to answer this question in tbh mutants, as basal locomotion was reduced in these mutants. However, our experiments revealed conditions in which changes in basal activity could be genetically separated from changes in the starvation response. One such case is rut mutants, which have normal locomotion but lack the starvation response. This effect seemed to be due to the function of Rut in octopamine neurons, as the defective starvation response was completely rescued by expressing a Rut transgene in octopamine neurons. A second, albeit less clear observation regards octβ2R mutants. Although baseline locomotion was much less altered in these mutants than in tbh mutants, these animals still could not mount a starvation response (Fig. 6d,e). Thus, it is likely that octopamine neurons are involved not only in locomotion, but also in the response to starvation.
Octopamine is also required for appetitive memory in adult fruit flies7. Notably, the appetitive memory procedure requires starvation before the assay, and tbh mutants cannot learn in this procedure. Octopamine has been proposed to mediate the reinforcing effects of sugar in appetitive memory formation5,7. Our studies raise the possibility that this mechanism might involve structural changes at synaptic sites.
Although our studies focused on structural changes at type II NMJs, many of our manipulations affected all octopamine neurons, as Tdc2-Gal4 drives Gal4 in all octopaminergic neurons. Thus, our studies cannot rule out an influence from other octopaminergic neurons, besides motor neurons, in the changes observed and in the behavior. However, the finding that the manipulations resulted in specific changes in type II NMJ terminals and that octopamine modulates synaptic strength at the NMJ argues that at least some of the effects are likely to be due to the peripheral octopamine innervation.
In summary, our studies reveal important mechanisms by which activity regulates the ability of motor neurons to scale the release of regulatory signals, which is important for the adaptation of the organism to the environment. In addition, they show a mechanism by which excitatory synapses are regulated in a global manner, presumably to maintain synaptic plasticity in a dynamic range.
Fly strains
Flies were reared in standard Drosophila medium at 25 °C except where indicated. Animals used in RNAi experiments were reared at 29 °C to increase knockdown efficiency. The following stocks were used: the wild-type strain Canton-S (CS), Tdc2-Gal4 (Bloomington Stock Center), UAS-Hid49, tbhnM18 (ref. 9), tdc2RG54 (ref. 23), UAS-Kir2.1 (ref. 30), tubP-Gal80ts (Bloomington), UAS-ShDN27, eag1 (Bloomington), UAS-mCD8-GFP (Bloomington), ydncM14cvf(Bloomington)y yw dncMLf36a (Bloomington), UAS-Dnc (remobilized to the second chromosome)50, rut2080 (Bloomington), UAS-Rut (Bloomington), UAS-dCRF.B2-b (Bloomington), oamb584 Pbac{WH}Octβ2R[f05679] (Bloomington), Dp(1;2)51b (duplication of dnc; Bloomington), UAS-Syt1-GFP (Bloomington), C380 (Budnik), BG439 (V. Budnik, unpublished), UAS-Dicer-2 (Bloomington), UAS-Octβ2R-RNAi (8486 and 104524; Vienna Drosophila RNAi Center) and UAS-PACα (see below).
Generation of PACα flies
Euglena gracilis PACα cDNA was provided by M. Watanabe. A 3,104-bp EagI fragment encompassing the full-length PACα cDNA with 5′ leader sequence was ligated into the EagI site of the Drosophila transformation vector pUAST. The insert was verified to be in the appropriate orientation by PCR and end sequencing, and transformed into flies by germline transformation.
Generation of TBH antibodies
TBH antibodies were raised against amino acids 277–670 of TBH, which is the following peptide sequence: QETTYWCHVQR LEGNLRRRHHIVQFEPLIRTPGIVHHMEVFHCEAGEHEEIPLYNGDCEQLPPRAKICSKVMVLWAMGAGTFTYPPEAGLPIGGPGFNPYVRLEVHFNNPEKQSGLVDNSGFRIKMSKTLRQYDAAVMELGLEYTDKMAIPPGQTAFPLSGYCVADCTRAAIPATGIIIFGSQLHTHLRGVRVLTRHFRGEQELREVNRDDYYSNHFQRMRTLHYKPRVLPGDALVTTCYYNTKDDKTAALGGFSISDEMCVNYIHYYPATKLEVCKSSVSEETLENYFIYMKRTEHQHGVHLNGARSSNYRSIEWTQPRIDQLYTMYMQEPLSMQCNRSDGTRFEGRSSWEGVAATPVQIRIPIHRKLCPNYNPLWLKPLEKGDCDLLGECIY The specificity of the antibody was demonstrated by the lack of immunorcactivity in null tbhnM18 mutants (Supplementary Fig. 1c,f).
Immunocytochemistry
Larval body-wall muscles were dissected and fixed for 15 min in 4% paraformaldehyde. For TBH immunocytochemistry samples were fixed in Bouin’s fixative. Antibodies and their concentrations were: anti-TBH 1:400, anti-mCD8a 1:100 (Invitrogen), anti-HRP-Texas Red 1:200 (Jackson ImmunoResearch), anti-FasII 1:2 (Developmental Studies Hybridoma Bank, DSHB), anti-Syt1 1:100 (gift from T. Littleton), anti-Bruchpilol 1:100 (nc82; DSHB), anti-Futsch 1:50 (22C10, DSHB), anti-GluRIIA 1:10 (8B4D2; DSHB). Secondary antibodies conjugated to FITC, Texas Red (Jackson) or Alexa 647 (Invitrogen) were used at a concentration of 1:200. Imaging of fixed preparations was as described20.
Animal rearing conditions for synaptopod analysis
All animals used in syn-aptopod analysis carried a copy of Tdc2-Gal4 and a copy of UAS-CD8:GFP. Egg collection was done in standard 25-mm diameter cornmeal/agar/molasses food vials at 25 °C with −60% humidity, and larvae were kept at low density. Wandering late third-instar larvae were used for experiments.
Stimulation procedures and live imaging of dissected preparations
Synaptopods were imaged from live preparations as described20. Both the spaced high K+ depolarization procedure and the ChR2 stimulation procedure were as described20. Briefly, the high K+ procedure consisted of 5 incubation cycles with 90 mM K+-containing saline, each lasting 2 min and separated by 15-min resting intervals. For the ChR2 stimulation procedure, animals were placed inside a drop of HL3 saline (~300 μl) containing 1.5 mM Ca2+. The procedure consisted of five cycles of blue light stimulation from four 491-nm 1-W LEDs placed 1.2 meters away from the animals. Each cycle consisted of a repeating 5-min procedure of 2 s on and 3 s off, separated by 15-min resting intervals. Saline was exchanged once during each resting interval. Animals were imaged 15 min after stimulation or subjected to the crawling assay. For octopamine stimulation, larvae were dissected in HL3 saline containing 0.1 mM Ca2+ and preparations gently glued onto a custom-made glass imaging chamber using surgical glue. Then, identified NMJs were imaged on an Improvision spinning disc confocal microscope (PerkinElmer) with a C9100-13 Hamamatsu cooled EM-CCD camera and using a 40× 1.2 NA objective, with a 2.4× optical zoom. After imaging for less than 30 min, animals were partially unglued to allow muscles to contract freely, and 10 μM octopamine in HL3 containing 1.5 mM Ca2+ was then applied for 15 min followed by 5 × 15-min washes with 0.1 mM Ca2+ HL3 saline before imaging again. In some experiments actinomycin D (Sigma) 5 μM in HL3, and cycloheximide (Sigma) 100 μM in IIL3 were applied throughout octopamine incubation and washes. The final concentration of DMSO in these solutions was approximately 0.03%. For PACα experiments, NMJs were imaged as above, then stimulated with a broad spectrum blue light dental gun placed ~2 cm away. Stimulation consisted of 5 × 5-s light exposures, each separated by 2 min rest.
Live imaging of intact (undissected) larvae
For live-imaging through the cuticle of intact larvae, animals were anesthetized using Sevoflurane (Baxter) and identified NMJs (muscle 9, A4) were imaged for not longer than 30 min. Larvae were allowed to recover for 1.5 h on food plates and then used for food deprivation experiments (see below) before imaging. Larvae that did not recover in 1.5 h after the first imaging session or after starvation were discarded. For non-starved controls, larvae were placed in food between imaging sessions.
Crawling assay
Synchronized larvae were grown at 25 °C in 28.5-mm diameter standard food vials at low density until mid third-instar larval stage. After washing with water, individual larvae were loaded onto a 24 × 24-cm 3% agar plates and allowed a pre-run of 25 s on the agar before recordings were made. Larval tracks were then recorded manually for 1 min on transparency paper over the plate lid, and the distances crawled were measured using ImageJ. All behavioral experiments were carried out in a 25 °C, 60% humidity behavioral room, n represents one animal. Percentage increase in locomotor speed of individual animals was calculated by: (locomotor speed of individual animals after starvation minus mean locomotor speed before starvation) × 100 divided by the mean locomotor speed before starvation. Mean percentage and s.e.m. were calculated.
Starvation assay
Larvae were maintained in food or food-free moisturized 35-mm Petri dishes for 2 h and then either assayed for synaptopod formation or subjected to the crawling assay.
Quantification of boutons and synaptopod number
The number of type I boutons was obtained at muscles 6 and 7 of abdominal segment A3, and the number of type II boutons was measured at muscle 12 in A3. For muscle area measurements the muscle length and width were measured using an ocular scale bar. Measurements of synaptopod number were from muscle 12 (A4) in dissected preparations and those from intact larvae were from muscle 9 (A4). Numbers of synaptopods in the histograms represent the total number of synaptopods per 100 μm of each arbor. Synaptopods were defined as such if they measured 0.5 μm or more. n represents number of NMJs. At most two NMJs (segment A3) were quantified in each animal.
Statistical analysis
For comparisons between more than two sample groups an analysis of variance (ANOVA) with Tukey post-hoc test was performed. For pair-wise comparisons a Student t-test was used. Numbers in histograms represent mean ± s.e.m. Unless otherwise noted, sample number (n) represents one synaptic arbor for anatomical measurements, or one animal for behavioral analyses.
Genotype abbreviations
Type II driver control, Tdc2/+; Type I+II driver control, C380-Gal4/+; muscle driver control, C57-Gal4/+; [transgene]-type II, Tdc2-Gal4>[transgene];[transgene]-type I+II, C380-Gal4>[transgene];[transgene]-muscle, C57-Gal4>[transgene] unless otherwise indicated; Dcr, UAS-Dicer-2.
Electrophysiology
Membrane potential recordings were performed on dissected third-instar larvae as described20. Briefly, larvae were dissected in a custom magnetic chamber in 0.3 mM Ca2+HL3 saline, and the A3 segmental nerves carefully cut close to the ventral ganglion. The chamber was then moved to the recording setup where it was perfused with 0.5 mM Ca2+ HL3. Muscle 6 in segment A3 was impaled with a 15–20-MΩ electrode, and voltage recordings were collected with an Axoclamp2A amplifier (Molecular Devices), using Heka Pulse software (Heka). Only samples with resting membrane potentials between −60 and −63 mV were considered. For EJP recordings the segmental nerve was stimulated with a suction electrode at 0.3 Hz, with a stimulus of 0.3 ms and sufficient voltage to evoke responses from both type I boutons. Four minutes of both mEJP and EJP data were recorded for each sample. Data was analyzed using Minianalysis software (Synaptosoft), and statistical analysis was done with Origin software (OriginLab). Bath application of octopamine was performed by changing the perfusion solution from 0.5 mM Ca2+ HL3 to 0.5 mM Ca2+ HL3 containing 10 μM octopamine. Samples were allowed to equilibrate for 2 min in the new saline, and then evoked and spontaneous events were recorded again. Statistical analysis was done as above.
RT-PCR
Total mRNA was extracted from dissected larval body wall muscles or larval brains using a combination of Trizol (Invitrogen) and the RNeasy kit (Qiagen). Synthesis of cDNA for +RT reactions was performed using the Superscript III kit (Invitrogen), where −RT reactions lacked reverse transcriptase. The +RT and −RT reactions were then diluted and used for PCR using the forward primer CATGCTGATGCACCGACCATC and the reverse primer CACTCCTCGCAGGTCATGGAG. These primers were specifically designed to recognize all known splice variants of octβ2R, and across two exon-intron junctions to avoid false signals from any contaminating genomic DNA. For semi-quantitative RT-PCR, we determined the linear range of Octβ2R amplification for the amount of starting cDNA (50 ng) and number of PCR cycles for wild type, and then amplified cDNA made from RNA isolated from wild type and C380>UAS-Octβ2R-RNAi.
Quantification of branch-points
Type II arbors were examined in an epifluorescence microscope and each branch bifurcation was counted as a single branch point. For these quantifications n is the number of total type II arbors (segment A3) quantified for each genotype.
Supplementary Material
supplemental materials
Acknowledgments
We thank M. Yoshihara, S. Speese, Y. Fuentes-Medel and C. Korkut for comments on the manuscript, C. Brewer for assistance with data analysis and the UMass Amherst antibody facility for production of the TBH antibody. This work was supported by US National Institutes of Health grants R01 MH0700000 to V.B., MH09883 to S.W., MH081982 to S.W. and GM084491 to M.J.A. M.J.A. was also supported by the Bill & Melinda Gates Foundation.
Footnotes
Note: Supplementary information is available on the Nature Neuroscience website.
AUTHOR CONTRIBUTIONS
A.C.K. designed and performed most experiments and contributed to manuscript writing; J.A. contributed to tool development, electrophysiology, experimental design and manuscript writing; R.B. performed RT-PCR and some immunocytochemistry; S.W., S.D. and R.B. generated, characterized and validated PACα function; M.J.A. helped with the design of the TBH antibody; and V.B. directed the project and wrote the manuscript in collaboration with A.C.K. and J.A.
COMPETING FINANCIAL INTERESTS
The authors declare no competing financial interests.
Reprints and permissions information is available online at http://www.nature.com/reprintsandpermissions/.
1. Murchison CF, et al. A distinct role for norepinephrine in memory retrieval. Cell. 2004;117:131–143. [PubMed]
2. Hu H, et al. Emotion enhances learning via norepinephrine regulation of AMPA-receptor trafficking. Cell. 2007;131:160–173. [PubMed]
3. Kuzmiski JB, Pittman QJ, Bains JS. Metaplasticity of hypothalamic synapses following in vivo challenge. Neuron. 2009;62:839–849. [PubMed]
4. Balfanz S, Strunker T, Frings S, Baumann A. A family of octopamine [corrected] receptors that specifically induce cyclic AMP production or Ca2+ release in Drosophila melanogaster. J Neurochem. 2005;93:440–451. [PubMed]
5. Hammer M, Menzel R. Multiple sites of associative odor learning as revealed by local brain microinjections of octopamine in honeybees. Learn Mem. 1998;5:146–156. [PubMed]
6. Schroll C, et al. Light-induced activation of distinct modulatory neurons triggers appetitive or aversive learning in Drosophila larvae. Curr Biol. 2006;16:1741–1747. [PubMed]
7. Schwaerzel M, et al. Dopamine and octopamine differentiate between aversive and appetitive olfactory memories in Drosophila. J Neurosci. 2003;23:10495–10502. [PubMed]
8. Zhou C, Rao Y. A subset of octopaminergic neurons are important for Drosophila aggression. Nat Neurosci. 2008;11:1059–1067. [PubMed]
9. Monastirioti M, Linn CE, Jr, White K. Characterization of Drosophila tyramine beta-hydroxylase gene and isolation of mutant flies lacking octopamine. J Neurosci. 1996;16:3900–3911. [PubMed]
10. Suo S, Kimura Y, Van Tol HH. Starvation induces cAMP response element-binding protein-dependent gene expression through octopamine-Gq signaling in Caenorhabditis elegans. J Neurosci. 2006;26:10082–10090. [PubMed]
11. Crocker A, Shahidullah M, Levitan IB, Sehgal A. Identification of a neural circuit that underlies the effects of octopamine on sleep:wake behavior. Neuron. 2010;65:670–681. [PMC free article] [PubMed]
12. Breen CA, Atwood HL. Octopamine—a neurohormone with presynaptic activity-dependent effects at crayfish neuromuscular junctions. Nature. 1983;303:716–718. [PubMed]
13. Monastirioti M, et al. Octopamine immunoreactivity in the fruit fly Drosophila melanogaster. J Comp Neurol. 1995;356:275–287. [PubMed]
14. Schuster CM, Davis GW, Fetter RO, Goodman CS. Genetic dissection of structural and functional components of synaptic plasticity. II. Fasciclin II controls presynaptic structural plasticity. Neuron. 1996;17:655–667. [PubMed]
15. Marqués G, Zhang B. Retrograde signaling that regulates synaptic development and function at the Drosophila neuromuscular junction. Int Rev Neurobiol. 2006;75:267–285. [PubMed]
16. Korkut C, Budnik V. WNTs tune up the neuromuscular junction. Nat Rev Neurosci. 2009;10:627–634. [PubMed]
17. Steinert JR, et al. Experience-dependent formation and recruitment of large vesicles from reserve pool. Neuron. 2006;50:723–733. [PubMed]
18. Sigrist SJ, et al. Experience-dependent strengthening of Drosophila neuromuscular junctions. J Neurosci. 2003;23:6546–6556. [PubMed]
19. Budnik V, Zhong Y, Wu CF. Morphological plasticity of motor axons in Drosophila mutants with altered excitability. J Neurosci. 1990;10:3754–3768. [PubMed]
20. Ataman B, et al. Rapid activity-dependent modifications in synaptic structure and function require bidirectional wnt signaling. Neuron. 2008;57:705–718. [PMC free article] [PubMed]
21. Saraswati S, Fox LE, Soil DR, Wu CF. Tyramine and octopamine have opposite effects on the locomotion of Drosophila larvae. J Neurobiol. 2004;58:425–441. [PubMed]
22. Davenport AP, Evans PD. Changes in haemolymph octopamine levels associated with food deprivation in the locust, Schistocerca gregaria. Physiol Entomol. 1984;9:269–274.
23. Hardie SL, Zhang JX, Hirsh J. Trace amines differentially regulate adult locomotor activity, cocaine sensitivity, and female fertility in Drosophila melanogaster. Dev Neurobiol. 2007;67:1396–1405. [PubMed]
24. Kittel RJ, et al. Bruchpilot promotes active zone assembly, Ca2+ channel clustering, and vesicle release. Science. 2006;312:1051–1054. [PubMed]
25. Rasse TM, et al. Glutamate receptor dynamics organizing synapse formation In vivo. Nat Neurosci. 2005;8:898–905. [PubMed]
26. Fouquet W, et al. Maturation of active zone assembly by Drosophila Bruchpilot. J Cell Biol. 2009;186:129–145. [PMC free article] [PubMed]
27. Mosca TJ, Carrillo RA, White BH, Keshishian H. Dissection of synaptic excitability phenotypes by using a dominant-negative Shaker K* channel subunit. Proc Natl Acad Sci USA. 2005;102:3477–3482. [PubMed]
28. Kitamoto T. Conditional modification of behavior in Drosophila by targeted expression of a temperature-sensitive shibire allele in defined neurons. J Neurobiol. 2001;47:81–92. [PubMed]
29. White BH, et al. Targeted attenuation of electrical activity in Drosophila using a genetically modified K(+) channel. Neuron. 2001;31:699–711. [PubMed]
30. Paradis S, Sweeney ST, Davis GW. Homeostatic control of presynaptic release is triggered by postsynaptic membrane depolarization. Neuron. 2001;30:737–749. [PubMed]
31. McGuire SE, Mao Z, Davis RL. Spatiotemporal gene expression targeting with the TARGET and gene-switch systems in Drosophila. Sci STKE. 2004;2004:pl6. [PubMed]
32. Han KA, Millar NS, Davis RL. A novel octopamine receptor with preferential expression in Drosophila mushroom bodies. J Neurosci. 1998;18:3650–3658. [PubMed]
33. Zhong Y, Budnik V, Wu CF. Synaptic plasticity in Drosophila memory and hyperexcitable mutants: role of cAMP cascade. J Neurosci. 1992;12:644–651. [PubMed]
34. Schröder-Lang S, et al. Fast manipulation of cellular cAMP level by light in vivo. Nat Methods. 2007;4:39–42. [PubMed]
35. Benito E, Barco A. CREB’s control of intrinsic and synaptic plasticity: implications for CREB-dependent memory models. Trends Neurosci. 2010;33:230–240. [PubMed]
36. Perazzona B, Isabel G, Preat T, Davis RL. The role of cAMP response element-binding protein in Drosophila long-term memory. J Neurosci. 2004;24:8823–8828. [PubMed]
37. Lee HG, et al. Octopamine receptor OAMB is required for ovulation in Drosophila melanogaster. Dev Biol. 2003;264:179–190. [PubMed]
38. Schuster CM, Davis GW, Fetter RD, Goodman CS. Genetic dissection of structural and functional components of synaptic plasticity. I. Fasciclin II controls synaptic stabilization and growth. Neuron. 1996;17:641–654. [PubMed]
39. O’Dell TJ, Connor SA, Gelinas JN, Nguyen PV. Viagra for your synapses: enhancement of hippocampal long-term potentiation by activation of beta-adrenergic receptors. Cell Signal. 2010;22:728–736. [PMC free article] [PubMed]
40. Roeder T. Tyramine and octopamine: ruling behavior and metabolism. Annu Rev Entomol. 2005;50:447–477. [PubMed]
41. Howell KM, Evans PD. The characterization of presynaptic octopamine receptors modulating octopamine release from an identified neurone in the locust. J Exp Biol. 1998;201:2053–2060. [PubMed]
42. Floresco SB, et al. Afferent modulation of dopamine neuron firing differentially regulates tonic and phasic dopamine transmission. Nat Neurosci. 2003;6:968–973. [PubMed]
43. Aamodt SM, Constantine-Paton M. The role of neural activity in synaptic development and its implications for adult brain function. Adv Neurol. 1999;79:133–144. [PubMed]
44. Landgraf M, Bossing T, Technau GM, Bate M. The origin, location, and projections of the embryonic abdominal motorneurons of Drosophila. J Neurosci. 1997;17:9642–9655. [PubMed]
45. Takizawa E, Komatsu A, Tsujimura H. Identification of common excitatory motoneurons in Drosophila melanogaster larvae. Zoolog Sci. 2007;24:504–513. [PubMed]
46. Nagaya Y, Kutsukake M, Chigusa SI, Komatsu A. A trace amine, tyramine, functions as a neuromodulator in Drosophila melanogaster. Neurosci Lett. 2002;329:324–328. [PubMed]
47. Kutsukake M, Komatsu A, Yamamoto D, Ishiwa-Chigusa S. A tyramine receptor gene mutation causes a defective olfactory behavior in Drosophila melanogaster. Gene. 2000;245:31–42. [PubMed]
48. Nishikawa K, Kidokoro Y. Octopamine inhibits synaptic transmission at the larval neuromuscular junction in Drosophila melanogaster. Brain Res. 1999;837:67–74. [PubMed]
49. McNabb SL, et al. Disruption of a behavioral sequence by targeted death of peptidergic neurons in Drosophila. Neuron. 1997;19:813–823. [PubMed]
50. Cheung US, Shayan AJ, Boulianne GL, Atwood HL. Drosophila larval neuromuscular junction’s responses to reduction of cAMP in the nervous system. J Neurobiol. 1999;40:1–13. [PubMed]