Intracellular Ca²⁺ signaling and store-operated Ca²⁺ entry are required in Drosophila neurons for flight

Abstract

Neuronal Ca²⁺ signals can affect excitability and neural circuit formation. Ca²⁺ signals are modified by Ca²⁺ flux from intracellular stores as well as the extracellular milieu. However, the contribution of intracellular Ca²⁺ stores and their release to neuronal processes is poorly understood. Here, we show by neuron-specific siRNA depletion that activity of the recently identified store-operated channel encoded by dOrai and the endoplasmic reticulum Ca²⁺ store sensor encoded by dSTIM are necessary for normal flight and associated patterns of rhythmic firing of the flight motoneurons of Drosophila melanogaster. Also, dOrai overexpression in flightless mutants for the Drosophila inositol 1,4,5-trisphosphate receptor (InsP₃R) can partially compensate for their loss of flight. Ca²⁺ measurements show that Orai gain-of-function contributes to the quanta of Ca²⁺-release through mutant InsP₃Rs and elevates store-operated Ca²⁺ entry in Drosophila neurons. Our data show that replenishment of intracellular store Ca²⁺ in neurons is required for Drosophila flight.

Several aspects of neuronal function are regulated by ionic calcium (Ca²⁺). Specific attributes of a Ca²⁺“signature” such as amplitude, duration, and frequency of the signal can determine the morphology of a neural circuit by affecting the outcome of cell migration, the direction taken by a growth-cone, dendritic development, and synaptogenesis (1). Ca²⁺ signals also determine the nature and strength of neural connections in a circuit by specifying neurotransmitters and receptors (2). Most of these Ca²⁺ signals have been attributed to the entry of extracellular Ca²⁺ through voltage-operated channels or ionotropic receptors. However, other components of the “Ca²⁺ tool-kit” coupled to Ca²⁺ release from intracellular Ca²⁺ stores are also present in neurons. These molecules include the store-operated Ca²⁺ (SOC) channel, encoded by the Orai gene, identified recently in siRNA screens for molecules that reduce or abolish Ca²⁺ influx from the extracellular milieu after intracellular Ca²⁺ store depletion (3–5). Several reports have confirmed its identity as the pore forming subunit of the Ca²⁺-release activated Ca²⁺ (CRAC) channel (6–8). Activation of this CRAC channel is mediated by the endoplasmic reticulum (ER) resident protein STIM (stromal interaction molecule), also identified in an RNAi screen for molecules that regulate SOC influx (9, 10). STIM senses the drop in ER Ca²⁺ levels, and interacts with Orai by a mechanism which is only just being understood (11). Orai and STIM function in conjunction with the sarco-endoplasmic reticular Ca²⁺-ATPase pump (SERCA) to maintain ER store Ca²⁺ and basal Ca²⁺. The importance of intracellular Ca²⁺ homeostasis and SOC entry (E) in neural circuit formation and in neuronal function and physiology remains to be elucidated.

Here, we report how Orai and STIM mediated Ca²⁺ influx and Ca²⁺ homeostasis in Drosophila neurons contribute to cellular and systemic phenotypes. Reduced SOCE measured in primary neuronal cultures is accompanied by a range of defects in adults, including altered wing posture, increased spontaneous firing, and loss of rhythmic flight patterns. These phenotypes mirror the spontaneous hyperexcitability of flight neuro-muscular junctions and loss of rhythmic flight patterns observed in Drosophila mutants of the inositol 1,4,5-trisphosphate receptor (InsP₃R, itpr, CG1063) gene (12). The InsP₃R is a ligand gated Ca²⁺-channel present on the membranes of intracellular Ca²⁺ stores. It is thought to be critical for various aspects of neuronal function (1, 13). Mutants in the gene coding for the mouse InsP₃R1 are ataxic (14). Cerebellar slices from InsP₃R1 knockout mice show reduced long-term depression, indicating that altered synaptic plasticity of the cognate neural circuits could underlie the observed ataxia (15).

To understand the temporal and spatial nature of intracellular Ca²⁺ signals required during flight circuit development and function, dOrai (CG11430) and dSERCA (encoded by CaP-60A gene, CG3725) function was modulated by genetic means in itpr mutants. This modulation can restore flight to flightless adults, by altering several parameters of intracellular Ca²⁺ homeostasis including SOCE. Our results suggest that components of the central pattern generator (CPG) required for maintenance of normal rhythmic flight in adults have a stringent requirement for SOCE after InsP₃R stimulation.

Results

SOC Entry in Neurons of Drosophila Requires Orai and STIM.

Genes encoding the SOC channel (Orai1) and the store Ca²⁺ sensor (Stim1) are known to maintain intracellular Ca²⁺ store levels ([Ca²⁺]_ER) in stimulated T cells. The replenishment of [Ca²⁺]_ER in T cells is required for their prolonged activation (16). Homologs of mammalian Orai and Stim exist in Drosophila as single genes, and perform similar cellular functions in S2 cells, where their depletion by gene specific double-stranded (ds)RNA leads to abrogation of SOCE (3–5). To investigate SOC channel activity in Drosophila neurons, we reduced levels of dOrai transcripts using dsRNA in primary neuronal cultures derived from larval brains. SOCE was monitored by Ca²⁺ imaging of cultured neurons in Ca²⁺ add-back experiments, after depletion of ER stores with thapsigargin in very low external Ca²⁺ (Fig. 1A). SOCE was significantly reduced in neurons expressing dsRNA for dOrai (UASdOraiRNAi²²¹ denoted as dsdOrai; Fig. 1 B and C). Also, the level of intracellular store Ca²⁺ ([Ca²⁺]_ER) was significantly lower in these cells (Fig. 1D), suggesting that Ca²⁺ entry through Drosophila Orai channels contributes to the maintenance of store Ca²⁺ in neurons. To ascertain that the reduced SOCE observed in cells expressing dOrai dsRNA is gene specific, SOCE was measured in 2 alternate conditions. Double-stranded RNA for the ER Ca²⁺-sensor dSTIM (CG9126), (UASdSTIMRNAi⁰⁷³ denoted as dsdSTIM), and a ligand-gated extracellular Ca²⁺ channel, NMDAR1 (UASdNR1RNAi³³³ denoted as dsdNR1; CG2902) were expressed in all neurons. Normal function of STIM is considered essential for Orai channel activity, whereas SOCE is not predicted to change when levels of a plasma membrane localized ligand-gated Ca²⁺-channel are reduced. Pan-neural expression of dsdStim followed by Ca²⁺ imaging revealed significant reduction of SOCE, [Ca²⁺]_ER (Fig. 1 B–D), and resting cytosolic Ca²⁺ ([Ca²⁺]_i; Fig. 1 E and F). A significantly higher frequency of cells with lower [Ca²⁺]_i were present among the neuronal population with dsdSTIM. However, dsdOrai expression had no effect on [Ca²⁺]_i. The efficacy of the dsRNA strains used was ascertained by semiquantitative RT-PCR, which showed a consistent reduction in the levels of the appropriate transcripts (Fig. S1 A–D). As expected, reduction in the level of dNR1 transcripts did not affect store Ca²⁺ or SOCE (Fig. S1 E and F). These results demonstrate that Ca²⁺ influx, leading to replenishment of ER stores through the STIM–Orai pathway, is conserved in Drosophila neurons. Also, the single STIM-encoding gene in Drosophila appears to regulate both [Ca²⁺]_ER and [Ca²⁺]_i. In mammalian systems, these cellular properties are regulated independently by STIM1 and STIM2, respectively (17).

Fig. 1.

Intracellular Ca²⁺ homeostasis in larval neurons is altered on expression of dsdOrai and dsdSTIM. (A) pseudocolor images represent [Ca²⁺]_ER and SOCE in primary cultures of neurons loaded with Fluo-4 from WT larvae and those expressing dsdOrai or dsdSTIM. (Scale bar, 10 μm.) (B) Single cell traces of SOCE by Ca²⁺ add-back after store depletion. Two cells of each genotype are shown. Arrow heads represent peak values of response, which have been plotted as a box chart in C. (C) Box plots of ΔF/F values of SOCE. The bigger boxes represent the data spread, smaller squares represent mean, and the diamonds on either side represent outlier values. (D) Box plot comparison of [Ca²⁺]_ER between neurons of indicated genotypes. (E) Kolmogorov–Smirnov (K-S) plot analyzing the distribution of [Ca²⁺]_i in neurons loaded with Indo-1. The frequency distribution is significantly shifted to the left for cells with dSTIM RNAi (dsdSTIM), indicating a higher frequency of cells with reduced [Ca²⁺]_i (P_K-S<0.05). (F) Box plot representation of [Ca²⁺]_i in neurons with dsdOrai or dsdSTIM. (n ≥ 150 cells; *, P_ANOVA < 0.05; **, P_ANOVA < 0.01).

Reduced SOCE in Drosophila Neurons Causes Flight Defects.

To determine whether reduced SOCE in Drosophila neurons affects neuronal function, motor coordination defects were measured in the appropriate genotypes. No obvious changes were visible in larvae expressing dsRNA for either dOrai or dSTIM. The larvae were viable and pupated normally. However, adult flies with pan-neural expression of dsdOrai and dsdSTIM had defective wing posture with significant loss of flight as seen in the “cylinder drop” test assay (Fig. 2 A and B) (18). Whereas >50% flies with dOrai knockdown were flightless, dSTIM knockdown resulted in a complete loss of flight (Fig. 2B). Expression of dsdOrai and dsdSTIM in glutamatergic neurons, which include the flight motor neurons, reduced flight ability in ≈35% of adult Drosophila, suggesting that the requirement for SOCE in flight extends beyond the glutamatergic domain. This observation is unlikely to be due to a difference in expression levels of the pan-neural and glutamatergic GAL4 strains, because the latter appears to be the stronger driver, as judged by reporter gene expression in pupae and adults (30).

Fig. 2.

RNAi knockdown of dOrai or dSTIM in subsets of neurons gives rise to flight motor defects. Pan-neuronal knockdown of dOrai and dSTIM induces (A) change in wing posture, (B) flight defects (n ≥ 100 flies), (C) higher levels of spontaneous firing (n ≥ 15 flies), and (D) defects in air-puff-induced flight patterns. (E) Representative traces of spontaneous firing activity from the DLMs of the indicated genotypes. Histograms represent mean ± SE; (*, P < 0.05; Student's t test). Both GAL4 control strains were tested and found to be similar to WT. Data shown are for glutamatergic GAL4 flies.

To understand how neuronal store Ca²⁺ and SOCE reduce flight ability, postsynaptic responses from the dorsal longitudinal indirect flight muscles (DLMs) that power flight were measured. Electrophysiological recordings were obtained during tethered flight (initiated in response to an air-puff stimulus) and at rest (Fig. 2 C–E). Nonfliers with pan-neural expression of dsdOrai and dsdSTIM, selected from the cylinder drop test were either unable to initiate rhythmic action potentials in response to an air-puff stimulus or exhibited unsustained (<5 s) and arrhythmic flight patterns (Fig. 2D); the control flies were normal (Fig. S1G). Knockdown by dsdOrai and dsdSTIM in glutamatergic neurons lead to a milder change in flight patterns compared with pan-neural knockdown, consistent with a role for SOCE in nonglutamatergic interneurons in addition to the glutamatergic flight motor neurons. Recordings from resting DLMs of these flies revealed a high arrhythmic spontaneous firing rate of action potentials, the frequency of which was significantly higher than WT (Fig. 2 C and E) or other control flies (Fig. S1H).

Overexpressing dOrai⁺ in Neurons Suppresses Flight Defects in Drosophila itpr Mutants.

SOCE activation through Orai and STIM in vivo requires a signal for depletion of intracellular Ca²⁺ stores. Based on the similar phenotypes of Drosophila itpr mutants (12), we hypothesized that, in the context of flight, intracellular Ca²⁺ store depletion probably occurs through InsP₃ signaling and the InsP₃R. Therefore, the effect of overexpressing dOrai in the genetic background of itpr mutants was tested. For this purpose, UASdOrai⁺ (expressing WT dOrai) transgenic strains were generated and expressed in selected neuronal subdomains. These subdomains include the glutamatergic domain tested above, aminergic domain, and the Drosophila insulin-like peptide 2 producing neurons (Dilp2 neurons). Expression of a UASitpr⁺ (expressing WT InsP₃R) transgene in the aminergic domain has been previously demonstrated to rescue itpr mutant phenotypes (12, 19), and more recently a similar rescue has been observed by UASitpr⁺ expression in Dilp2 neurons (N. Agrawal and G. Hasan, personal communication). Expression of 2 copies of UASdOrai⁺ in Dilp2 neurons and aminergic domain could partially suppress the altered wing posture of itpr^ka1091/ug3 (hereafter referred to as itpr^ku; Fig. S2A). Although flight ability was not restored, there was initiation of flight patterns on air-puff delivery, normally completely lacking in itpr^ku animals (Fig. 3A). Also, spontaneous hyperactivity of the DLMs in itpr^ku was suppressed to a significant extent by expressing UASdOrai⁺ either ubiquitously (hsGAL4-leaky or hsGAL4^L at 25 °C) (12); or in the aminergic, Dilp2, and glutamatergic subneuronal domains (Fig. 3B; Fig. S2C).

Fig. 3.

Overexpression of dOrai⁺ in neuronal subsets partially suppresses flight defects in an itpr mutant. (A) Flies overexpressing dOrai⁺ in aminergic, Dilp2, and glutamatergic neurons or ubiquitously under a HS promoter and subjected to HS initiate unsustained (<5 s) flight patterns in response to air puff. (B) Overexpression of dOrai⁺ (2 copies) either ubiquitously (by a leaky HS GAL4 at 25 °C) or in the indicated subneuronal domains suppresses the elevated spontaneous firing of itpr^ku flies. (**, P < 0.01; Student's t test; n ≥ 20 flies).

To ascertain whether dOrai⁺ expression is required during flight circuit formation in pupae and/or during acute flight in adults, ubiquitous expression of dOrai⁺ in UASdOrai⁺/hsGAL4^L; itpr^ku organisms was up-regulated by a heat shock (HS) either in 24-h pupae or in 1-day-old adults. In both conditions, a significant number of flies could initiate flight in response to an air puff. Thus, levels of dOrai⁺ can modulate flight circuit activity both during its development and in adult function (Fig. 3A; Fig. S2B). However, the flight patterns obtained were not sustained and appeared arrhythmic (Fig. 3A), indicating that although dOrai⁺ overexpression can suppress the flight defects and associated physiology of itpr mutants to a significant extent, it is insufficient to regain complete flight (see Discussion).

Pan-Neuronal Expression of dOrai⁺ Restores Intracellular Calcium Homeostasis in itpr^ku.

To understand the cellular basis of dOrai⁺ suppression of itpr mutant phenotypes, SOCE and [Ca²⁺]_ER were measured in primary neurons from itpr^ku larval brains. SOC influx was greatly diminished (Fig. 4 A and B), whereas [Ca²⁺]_ER was significantly elevated in neurons derived from itpr^ku larvae grown at 25 °C, as compared with the WT. The mean [Ca²⁺]_ER in itpr mutant appeared twice as much as WT (Fig. 4B). The percentage of cells with detectable SOC was ≈3–5%, as compared with 70–80% in WT. Pan-neural overexpression of dOrai⁺ in itpr^ku neurons restored detectable SOCE to 70% of itpr^ku neurons. Also, [Ca²⁺]_ER went back to WT levels, indicating that reduced SOCE and elevated [Ca²⁺]_ER in itpr^ku are linked homeostatic processes. Overexpression of dOrai⁺ in WT neurons did not effect SOCE and [Ca²⁺]_ER, although it did elevate [Ca²⁺]_i to ≈1 μM in WT and itpr^ku backgrounds (Fig. 4 C and D).

Fig. 4.

The dOrai⁺ overexpression in itpr^ku neurons restores intracellular Ca²⁺ homeostasis. (A) SOCE measurements in the indicated genotypes (*, P_ANOVA < 0.05, compared with itpr^ku; and **, P < 0.01, compared with WT). (B) [Ca²⁺]_ER in the indicated genotypes (**, P_ANOVA <0.01, compared with WT). (C) K-S plot for [Ca²⁺]_i in neurons of the indicated genotypes (P_K-S < 0.05 for genotypes expressing dOrai⁺ compared with WT). (D) Box plot representation of [Ca²⁺]_i (*, P_ANOVA < 0.05; **, P_ANOVA < 0.01; n ≥ 170 cells).

The significance of deranged SOCE and [Ca²⁺]_ER in itpr^ku neurons was determined by measuring these parameters in cells of itpr^ku derived from second instar larvae maintained at 17.5 °C. The itpr^ku is a cold-sensitive allelic combination, and is lethal during the third instar larval stage at 17.5 °C (19). Interestingly, SOCE and [Ca²⁺]_ER in these conditions were similar to WT neurons grown under identical conditions at 17.5 °C (Fig. S3 A and B). These data suggest that itpr^ku organisms up-regulate store Ca²⁺ at 25 °C as a compensatory mechanism to allow for survival at that temperature, and that reduced SOCE may be a result of elevated store Ca²⁺. The observation that return of [Ca²⁺]_ER and SOCE to normal by dOrai⁺ overexpression at 25 °C is insufficient for restoration of complete flight in itpr^ku suggests that additional aspects of intracellular Ca²⁺ signaling are essential for flight in these organisms.

Restoration of Flight in itpr^ku by Dominant Alleles of dOrai and dSERCA.

To investigate the additional properties of intracellular Ca²⁺ signaling required for flight, genetic interactions between itpr and dOrai were further probed. For this purpose, mutant alleles with P-inserts in the dOrai gene were obtained. The 2 alleles obtained, and referred to as dOrai¹ and dOrai², both contain an EP{gy2} construct (enhancer P-element) (20) at a distance of 13 bps from each other in the 5′ UTR of the dOrai gene (Fig. S4). The 2 dOrai alleles were initially tested for their interaction with itpr^ku by measuring viability at 17.5 °C. Introduction of a single copy of either dOrai mutant allele could suppress cold-sensitive lethality of itpr^ku (Fig. S5D). A single copy of either dOrai allele also suppressed the wing posture defect of itpr^ku grown at 25 °C to a significant extent (Fig. S5A), suggesting that both dOrai¹ and dOrai² are hypermorphs. Subsequent observations support this conclusion further. The presence of a single copy of either dOrai allele in the background of itpr^ku restored flight initiation in response to an air puff (Fig. 5A; Fig. S5B), and suppressed hyperactivity of flight neuromuscular junctions (NMJs) (Fig. 5D; Fig. S5C). As expected for a hypermorph, the dOrai^2/+ mutant allele can also partially suppress flight-related defects and reduced SOCE and [Ca²⁺]_ER arising from pan-neuronal expression of dsdSTIM (Figs. 5 A, D, and E, and 6 D and E; Fig. S5C). Store Ca²⁺ and SOCE in neurons heterozygous for dOrai^2/+ are not significantly different from WT (Fig. S5E).

Fig. 5.

Suppression of flight and related physiological defects by dominant mutants of dOrai and dSERCA. (A) Air-puff-induced flight patterns in the indicated genotypes. (B) Flight defects in itpr^ku are suppressed by the presence of both Kum¹⁷⁰ and dOrai² or dOrai¹, but not with dOrai mutants or Kum^170/+ on their own. (C) Snapshots taken within the first 5 s of air-puff-induced flight initiation in (i) itpr^ku; (ii) Kum¹⁷⁰/dOrai²; itpr^ku (Movie S1). (D) Spontaneous hyperactivity in DLMs of indicated genotypes; n ≥ 15. (E) Wing posture defects induced by dsdSTIM are suppressed by dOrai² (50%) or Kum¹⁷⁰ (10%). Histograms represent mean ± SE; (*, P < 0.05; **, P < 0.01, compared with itpr^ku; Student's t test).

Fig. 6.

Effects of dOrai and dSERCA mutants on different aspects of intracellular Ca²⁺ release. (A) Changes in stimulated Ca²⁺ release through InsP₃R (measured as ΔF/F) (**, P_ANOVA < 0.01, compared with WT; *, P_ANOVA < 0.05, compared with itpr^ku; n ≥ 150 cells). (B) Effect of Kum¹⁷⁰ and dOrai² on perdurance of InsP₃R mediated Ca²⁺-release signals; n ≥ 40 cells, with similar peak response times. (C) Single cell traces of SOCE by Ca²⁺ add-back after store depletion. (D) SOCE measured in cultured neurons of indicated genotypes (**, P_ANOVA < 0.01, compared with WT). SOCE in dOrai^2/+; itpr^ku is significantly higher than itpr^ku (*, P_ANOVA < 0.05), and is normal in cells of Kum¹⁷⁰/dOrai²; itpr^ku (**, P_ANOVA < 0.01, compared with itpr^ku). Heterozygous dOrai^2/+ partially restores SOCE in dsdSTIM expressing neurons (**, P_ANOVA < 0.01). (E) [Ca²⁺]_ER measurements (**, P_ANOVA < 0.01; *, P_ANOVA < 0.05 in Kum^170/+ genotypes, compared with WT); dOrai^2/+ restores [Ca²⁺]_ER in itpr^ku double mutants (*, P_ANOVA < 0.05). Presence of dOrai² restores [Ca²⁺]_ER in neurons expressing dsdSTIM (**, P_ANOVA < 0.01; n ≥ 170 cells).

The partial suppression of itpr^ku phenotypes by hypermorphic dOrai mutant alleles is reminiscent of the recently demonstrated interaction between a dominant mutant allele of dSERCA (21) called Ca-P60A^Kum170 (referred to here as Kum¹⁷⁰) and itpr mutants. This allele has been shown to delay cytoplasmic Ca²⁺ clearance after neuronal depolarization (22). Therefore, we tested the effect of introducing Kum¹⁷⁰ in dOrai^1or2/+; itpr^ku organisms. Flies of the genotype dOrai²/Kum¹⁷⁰; itpr^ku exhibited normal wings (Fig. S5A) and normal levels of spontaneous electrical activity in DLM recordings, consistent with the previously demonstrated dominant effect of Kum¹⁷⁰ (22). Strikingly, flight ability was restored in a significant number of these triple mutant flies. This observation is in contrast to the complete loss of flight ability in itpr mutants and itpr, dOrai or itpr, dSERCA double mutant combinations; >60% of dOrai¹/Kum¹⁷⁰; itpr^ku adults and ≈50% of dOrai²/Kum¹⁷⁰; itpr^ku adults passed as “fliers” in the cylinder drop test assay (Fig. 5B). Air-puff delivery elicited sustainable rhythmic flight patterns similar to WT in a high proportion of these flies (Fig. 5A; Fig. S5B, and Movie S1). Thus, down-regulating SERCA function restores or compensates for the additional intracellular Ca²⁺ signaling deficits required for free flight, which are lacking in dOrai^1or2/+; itpr^ku organisms. The nature of these Ca²⁺ signals was investigated next.

Ca²⁺ Release Through InsP₃ Receptor and SOCE Together Contribute to Maintenance of Flight.

Ca²⁺ release through the InsP₃R was measured by stimulating neurons ectopically expressing the Drosophila muscarinic acetylcholine receptor (mAChR) with increasing concentrations of the agonist carbachol (23). For the WT InsP₃R Ca²⁺, release increased as a function of carbachol concentration (Fig. S3D); it was greatly attenuated in itpr^ku (Fig. 6A; Fig. S3 D–F). Expression of mAChR transcripts, as determined by semiquantitative RT-PCR, was similar in mutant and WT (Fig. S3C).

Next, carbachol-stimulated Ca²⁺ release in itpr^ku was measured in the presence of dOrai² and Kum¹⁷⁰ double and triple mutant combinations. Kum¹⁷⁰ had no direct effect on Ca²⁺-release through the InsP₃R on carbachol stimulation. The presence of dOrai² in either dOrai^2/+; itpr^ku or in dOrai²/Kum¹⁷⁰; itpr^ku organisms restored carbachol-stimulated Ca²⁺ release to WT levels (Fig. 6A; Fig. S3F). However, this restoration is clearly not the only factor in flight maintenance, because dOrai^2/+; itpr^ku organisms are flightless. Therefore, we measured additional parameters that are likely to contribute to the flight rescue in triple mutants. These measurements include perdurance of the carbachol-stimulated Ca²⁺ peak, SOCE, [Ca²⁺]_ER, and [Ca²⁺]_i.

The presence of a single copy of Kum¹⁷⁰ delayed Ca²⁺ sequestration after carbachol-stimulated release, and led to greater perdurance of the Ca²⁺ peak; this effect of Kum¹⁷⁰ was also present in cells derived from dOrai²/Kum¹⁷⁰; itpr^ku organisms (Fig. 6B; Fig. S3F). Consistent with the known function of SERCA, Kum¹⁷⁰ had a dominant effect and reduced levels of store Ca²⁺ in all genotypes tested including Kum^170/+; itpr^ku and dOrai²/Kum¹⁷⁰; itpr^ku (Fig. 6E). Concurrent with the lower store, SOCE was greatly elevated in Kum¹⁷⁰ heterozygotes (Fig. 6D). Significantly, SOCE was normal in neurons derived from dOrai²/Kum¹⁷⁰; itpr^ku larvae, as compared with itpr^ku, dOrai^2/+; itpr^ku and Kum^170/+; itpr^ku (Fig. 6 C and D). Thus, the combined effect of Orai² and Kum¹⁷⁰ on itpr^ku is to restore near WT levels of InsP₃-stimulated Ca²⁺-release, followed by a broader curve of Ca²⁺ persistence and normal SOCE. In dOrai^2/+; itpr^ku organisms SOCE improved over itpr^ku, but remained low as compared with WT, similar to the observation with pan-neuronal expression of dOrai⁺ in itpr^ku (Figs. 4B and 6 C and D). Importantly, in the triple mutants, [Ca²⁺]_ER remained low (Fig. 6E), indicating that steady store Ca²⁺ levels do not effect flight directly, but perhaps contribute to driving the higher level of SOCE observed. Larval neurons heterozygous for dOrai² or Kum^170/+ had elevated levels of basal cytosolic Ca²⁺ with or without itpr^ku in the background (Fig. S5 F and G). Higher [Ca²⁺]_i is unlikely to contribute directly to flight rescue, because itpr mutants with high [Ca²⁺]_i also exhibit flight defects.

Discussion

We have shown that SOC entry through the Orai/STIM pathway and the rate of clearance of cytoplasmic Ca²⁺ by SERCA together shape intracellular Ca²⁺ response curves in Drosophila larval neurons. The phenotypic changes associated with altering Orai/STIM function on their own and in itpr mutant combinations suggest that these Ca²⁺ dynamics are conserved through development among neurons in pupae and adults. The development and function of the flight circuit appears most sensitive to these cellular Ca²⁺ dynamics, changes in which have a profound effect on its physiological and behavioral outputs. Direct measurements of Ca²⁺ in flight circuit neurons are necessary in future to understand why these cells are more sensitive to changes in intracellular Ca²⁺ signaling. Other circuits such as those required for walking, climbing and jumping remain unaffected. Possible effects of altering intracellular Ca²⁺ homeostasis on higher order neural functions have yet to be determined.

The flow of information in a neural circuit goes through multiple steps within and between cells. Suppression experiments, such as the ones described here, present a powerful genetic tool for understanding the mechanisms underlying both the formation of such circuits and their function. The correlation observed between adult phenotypes and Ca²⁺ dynamics in populations of larval neurons from the various genotypes supports the following conclusions. Out-spread wings, higher spontaneous firing, and initiation of rhythmic firing on air-puff delivery in itpr^ku are suppressed by either increasing the quanta (through hypermorphic alleles of dOrai and by dOrai⁺ overexpression) or by increasing the perdurance (through mutant Kum¹⁷⁰) of the intracellular Ca²⁺ signal (Fig. S6 Center). Flight ability and maintenance of flight patterns requires SOCE in addition to increased quanta and perdurance of the Ca²⁺ signals, suggesting that SOCE in neurons contributes to recurring Ca²⁺ signals essential for flight maintenance (Fig. S6 Right).

The signals that trigger InsP₃ generation in Drosophila neurons and the nature of the downstream cellular response remain to be investigated. Previous work has shown that rescue of flight and related physiological phenotypes in itpr mutants require UASitpr⁺ expression in early to midpupal stages, indicating the InsP₃R activity is necessary during development of the flight circuit (12). Due to perdurance of the InsP₃R, its requirement in adults was not established. We now find that a basal level of dOrai⁺ expression through development followed by ubiquitous overexpression in adults can help initiate flight in itpr^ku, indicating a requirement for SOCE in adult neurons that probably constitute the CPG for flight. The precise neuronal circuit and neurons in the flight CPG are under investigation (24). Aminergic, glutamatergic, and insulin producing neurons could assist in development and/or directly constitute the circuit. Similar patterns of neuronal activity in the flight circuit of itpr mutants, either by generating different combinations of Ca²⁺ fluxes (as shown here), or by UASitpr⁺ expression in nonoverlapping neuronal domains (N. Agrawal and G. Hasan, personal communication) supports the idea that different aspects of neuronal activity can compensate for each other to maintain constant network output.

Precisely how hypermorphic dOrai alleles modify itpr^ku function to increase the quanta of Ca²⁺ release remains to be investigated. The ability of itpr^ku to maintain elevated [Ca²⁺]_ER at 25 °C suggests a possible interaction between this heteroallelic combination and Orai/STIM. The mutated residue in itpr^ka1091 (Gly to Ser at 1891) lies in the modulatory domain, whereas in itpr^ug3, it lies in the ligand binding domain (Ser to Phe at 224); both residues are conserved in mammalian InsP₃R isoforms (25). The mutant residues could directly affect InsP₃R interactions with a store Ca²⁺ regulating molecule like STIM (26). Recent reports also demonstrate the formation of macromolecular assemblies of InsP₃R, SERCA, and SOC channels, suggesting specific functional interactions between them (27).

Last, our results suggest new ways of treating diseases where altered intracellular Ca²⁺ signaling or homeostasis has been suggested as a causative agent. Perhaps, the best documented of these diseases are spino-cerebellar ataxia 15, which arises by heterozygosity of the mammalian IP₃R1 gene (28), severe combined immunodeficiency due to a mutation in Orai1 (3), and Darier's disease from a mutation in SERCA2 (29). Based on the underlying changes in intracellular Ca²⁺ properties in these genetic diseases, our study suggests ways of deciding appropriate combination of drugs that might target the causative gene products and their functionally interacting partners.

Materials and Methods

Drosophila melanogaster Strains.

The WT Drosophila strain used throughout is Canton-S. UASdOrai (S. Ziegenhorn), UASmAChR (from Dm mAChR cDNA clone) (23), Ca-P60A^Kum170ts (21), dOrai¹¹⁰⁴² and dOrai²⁰¹¹⁹ (referred as dOrai¹ and dOrai², respectively), Pan-neuronal GAL4 (Elav^C155), and Ubiquitous GAL4 (hsp70^L, Heat shock^Leaky) from Bloomington Stock Center. UASRNAi strains (VDRC). Glutamatergic GAL4 (OK371) (30), aminergic (Ddc) (31), GAL4 expressing in ILP2 producing neurons is Dilp2GAL4 (32) .

Flight Assay and Electrophysiology.

Flight tests, recording techniques, and data analysis have been published previously (12), and are described in detail in SI Materials and Methods.

Primary Neuronal Cultures, Calcium Imaging, and Data Analysis.

Primary neuronal cultures were generated according to previously published protocols (22, 33). Calcium imaging for Fluo-4 was performed as described previously (2). Measurements for [Ca²⁺]_i were performed using the ratiometric dye Indo-1 (Invitrogen Technologies). Changes in fluorescence were quantified using the ImagePro plus software, V1.33. Detailed protocols are included in SI Materials and Methods.