SEARCH

SEARCH BY CITATION

Keywords:

  • GABA ;
  • GABAA receptor;
  • Ggt-1 ;
  • Glutamate;
  • green-light avoidance

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information
Thumbnail image of graphical abstract

Drosophila larvae innately show light avoidance behavior. Compared with robust blue-light avoidance, larvae exhibit relatively weaker green-light responses. In our previous screening for genes involved in larval light avoidance, compared with control w1118 larvae, larvae with γ-glutamyl transpeptidase 1 (Ggt-1) knockdown or Ggt-1 mutation were found to exhibit higher percentage of green-light avoidance which was mediated by Rhodopsin6 (Rh6) photoreceptors. However, their responses to blue light did not change significantly. By adjusting the expression level of Ggt-1 in different tissues, we found that Ggt-1 in malpighian tubules was both necessary and sufficient for green-light avoidance. Our results showed that glutamate levels were lower in Ggt-1 null mutants compared with controls. Feeding Ggt-1 null mutants glutamate can normalize green-light avoidance, indicating that high glutamate concentrations suppressed larval green-light avoidance. However, rather than directly, glutamate affected green-light avoidance indirectly through GABA, the level of which was also lower in Ggt-1 mutants compared with controls. Mutants in glutamate decarboxylase 1, which encodes GABA synthase, and knockdown lines of the GABAA receptor, both exhibit elevated levels of green-light avoidance. Thus, our results elucidate the neurobiological mechanisms mediating green-light avoidance, which was inhibited in wild-type larvae.

We proposed the role of GGT-1 in the inhibition of Drosophila green-light avoidance. GGT-1 in malpighian tubules regulates glutamate levels, and further affects GABA concentrations, which functions as an inhibitory neurotransmitter at the GABAA receptor RDL to suppress responses to green light. These findings elucidate the mechanisms mediating inhibition of green-light avoidance and may explain differential responses to light of different colors.

Abbreviations used
BOs

Bolwig organs

CRY

Cryptochrome

Gad1

glutamic acid decarboxylase 1

GAD

glutamate decarboxylase

Ggt-1

γ-glutamyl transpeptidase 1

GGT

Gamma (γ-) glutamyl transferase

LNs

lateral neurons

LOC

larval optic center

MT

malpighian tubules

PDF

pigment dispersing factor

PRs

photoreceptors

PTTH

prothoracicotropic hormone

RDL

Resistant to Dieldrin

Rh5

Rhodopsin5

Rh6

Rhodopsin6

TNT

tetanus toxin light chain

VGlut

Vesicular glutamate transporter

Light response behavior and circadian clock entrainment require light signals from the environment, mediated by distinct visual pathways (Keene et al. 2011). As in mammals, sensing light is important in Drosophila larvae. Moderate light is perceived by the Bolwig Organs (BOs), which are composed of 12 pairs of photoreceptors (PRs) adjacent to the mouth hooks (Steller et al. 1987). Four PRs express Rhodopsin5 (Rh5), which provides for the sensation of blue light, whereas the other eight PRs express Rhodopsin6 (Rh6) and sense green light (Sprecher and Desplan 2008). Genetic ablation of BOs or Rh5 mutation disrupts rapid light-induced responses at moderate light intensities (Busto et al. 1999; Hassan et al. 2000; Mazzoni et al. 2005). However, entrainment of circadian rhythms is sufficient with either Rh5 or Rh6 (Keene et al. 2011). Class IV multidendritic neurons that tile the larval body wall also drive light avoidance (Xiang et al. 2010). The axons of larval PRs synapse with neurons that express circadian rhythm genes: four pigment dispersing factor (PDF)-expressing lateral neurons (PDF neurons) and a fifth non-PDF lateral neuron (the 5th LN) (Sprecher et al. 2011; Yuan et al. 2011). Light avoidance may be modulated by the combined action of the lateral neurons and circadian related dorsal neurons (Keene et al. 2011; Collins et al. 2012). Two pairs of prothoracicotropic hormone neurons downstream of PDF neurons have been reported to be involved in switching between light avoidances and preferences (Gong et al. 2010; Yamanaka et al. 2013).

γ-glutamyl transferase (GGT) is an enzyme that is evolutionary conserved from bacteria to mammals and plays an important role in antioxidant defense, detoxification, and inflammatory processes (Castellano and Merlino 2012). GGT is mainly located on the external surface of the plasma membrane where it catalyzes and transfers the γ-glutamyl moiety of GSH, and other γ-glutamyl-containing compounds, to substrates such as water, dipeptides, and amino acids (Tate and Meister 1981). Thirteen homologs of human GGT have been identified (at least six are active), and four have been identified to exist, so far, in rats (Pawlak et al. 1988; Potdar et al. 1997; Yamaguchi et al. 2000; Puente and Lopez-Otin 2004). In Drosophila, four genes have been predicted to code the GGT protein according to sequence alignment (Walker et al. 2006). Of these homologs, Ggt-1 encodes the male accessory gland protein, which may result in relatively high levels of free glutamic acid in accessory gland secretions (Walker et al. 2006).

In this study, we report that Ggt-1 knockdown in larvae produced a higher percentage of green-light avoidance compared with controls after RNAi screening. However, following Ggt-1 knockdown, responses to blue light did not change significantly. Further results showed that Ggt-1 impacted green-light avoidance by regulating the level of glutamate, which was further supported by feeding experiments. Moreover, glutamate did not directly function as a neurotransmitter rather it was transformed to GABA to produce this effect. Thus, our results implicate a neural pathway that regulates green-light avoidance, which was suppressed by GABA signals in wild-type larvae.

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information

Fly stocks

The flies and larvae used in all experiments (except for the glutamate feeding experiments) were raised on a standard medium of corn meal and molasses (Guo et al. 1996) with a 12-h light-dark cycle and under ambient conditions of 25°C and 60% humidity. Larvae of w1118 larvae showed normal phototaxis and were used as controls. FM7a, elav-Gal4, and repo-Gal4 were provided by the Bloomington Drosophila Stock Center (Indiana University, Bloomington, IN, USA). NP3609-Gal4 and NP0136-Gal4 were provided by the Drosophila Genetic Resource Center (Kyoto Institute of Technology, Kyoto, Japan). UAS-VGlutRNAi was provided by Vienna Drosophila RNAi Center (Vienna, Austria). C57-Gal4, C42-Gal4, C507-Gal4 were gifts from Yongqing Zhang's laboratory (Institute of Genetics and Developmental Biology, Beijing, CAS). Actin-Gal4 was a gift from Renjie Jiao's laboratory (Institute of Biophysics, Beijing, CAS). RNAi lines used in screening were gifts from Jianquan Ni's laboratory (Tsinghua University, Beijing, China). Previous descriptions of sevLY3;Rh52;Rh61, UAS-RdlRNAi, and L352F have been provided (Liu et al. 2007; Yamaguchi et al. 2008; Featherstone et al. 2000). The Ggt-1 null mutant fly Ggt-1∆1523 and the transgenic flies UAS-Ggt-1 and UAS-Ggt-1RNAi were generated in our laboratory.

Generation of Ggt-1 null mutants Ggt-1∆1523 by P-element jump-out

To generate a null mutant of Ggt-1, the pGawB P-element in NP0136-Gal4 was imprecisely jumped out. The P-element is a Drosophila transposable element that is widely used to manipulate the genome. First, the insertion site of the pGawB P-element in NP0136-Gal4 was determined by inverse PCR and sequencing to confirm its insertion site in Ggt-1. NP0136-Gal4 genomic DNA was prepared using the QIAamp mini DNA kit (Qiagen, Valencia, CA, USA). The genome DNA was digested with HpaII, and then ligated at 4°C overnight. The following primers were used to amplify the corresponding sequence from the ligated DNA:

iPCR-forward: 5′-CTATCGACGGGACCACCTTA-3′

iPCR-backward: 5′-CTCCACAATTCCGTTGGATT-3′

The PCR products were then sequenced and compared with the Drosophila genome to identify the exact insertion site of the P-element. To generate Ggt-1 null mutants, virgin flies of NP0136-Gal4 were crossed with male flies of ∆2-3 that expresses transposase to mobilize P-elements within the genome. A P-element that is mobilized imprecisely can remove flanking genomic sequences, leading to a local deletion. Approximately 500 chromosomes with independent excision events were balanced over FM7a, and imprecise jump-out lines, Ggt-1∆1523, were finally identified by PCR using the primers.

identify-forward: 5′-CATTCACCTGCTTCCAGATAGA-3′

identify-backward: 5′-CTCCGAAGTGCTATTGTCGTG-3′

Construction of transgenic flies

To increase or decrease the transcriptional level of Ggt-1, UAS-Ggt-1, and UAS-Ggt-1RNAi transgenic flies were constructed. The following PCR primers were used to amplify the corresponding sequence from the genomic DNA of w1118 larvae:

Ggt-1forward: 5′-ATGAATTCATGAGGATTGTGTGGAGTAAGAAGC-3′

Ggt-1backward: 5′-ATAGATCTCTAATGCTGCATCTTGTTGGTTTTG-3′

Ggt-1RNAiforward: 5′-AGCGTAAGGAAGCTGATCCA-3′

Ggt-1RNAibackward: 5′-CCAGGAAGGCAATGTACGTT-3′

The first two primers were used to amplify the full-length cDNA of Ggt-1. EcoRI and BglII sites were added to PCR primers, so that the cDNA can be ligated to the pUAST vector directly to generate pUAST-Ggt-1. For the Ggt-1 RNAi construction, we generally followed the methods of Nagel et al. (2002). The target cDNA sequence of Ggt-1 was amplified by the last two primers. The PCR product was cloned into the pGEMT vector (Promega, Madison, WI, USA) in the appropriate directions to give Ggt-1-T plasmids. The BamHI and KpnI fragment of Ggt-1-T was then cloned into the corresponding site of pHIBS vector (Nagel et al. 2002) to give pHIBS-Ggt-1. The vector, pHIBS-Ggt-1, was then digested by SalI/KpnI and BamHI/EcoRI separately, and consequently subcloned into the XhoI/KpnI and EcoRI/BglII site of pUAST vector successively, to generate a head-to-head orientation separated by the 72-bp-long intron I of the Hairless gene from the pHBIS vector. These pUAST constructs were used to generate transgenic flies through P-element-mediated germline transformation.

Light avoidance assay

The light avoidance assay was performed according to the methods provided by Xiang et al. (2010). Briefly, each time five early to middle 3rd instar larvae (72–96 h after egg laying) were washed out of the food medium. They were isolated to an agar plate and left to acclimatize for 5 min, at 24 ± 0.5°C. Then, a larva was moved to the center of another agar plate. Shortly after it began to crawl linearly, a light spot (1.5 cm in diameter) of a light-emitting diode (MacroAmbition, Beijing, China) (at peak wavelengths of 460 nm for blue or 520 nm for green) was applied from above at just in front of the larval head. Avoidance responses were then recorded if the larva changed its direction before its whole body length was under the light spot, otherwise they would be regarded as non-responders. The larvae were tested one by one and each larva was tested once. Those larvae that were picked out from food at one time were treated as one group. Fifty to eighty larvae were tested for each stock.

Real-time quantitative PCR (qPCR)

qPCR was carried out to quantify the RNA levels of Ggt-1. Total RNA from 3rd instar larvae was isolated and reverse transcription was performed using SuperScript III (Invitrogen, Carlsbad, CA, USA). qPCR was conducted by Chromoe 4 (Bio-Rad Laboratories, Hercules, CA, USA). Relative differences in specific gene expression levels were quantified as 2∆Ct (∆Ct is the Ct value of the gene of interest subtracted from the Ct value of actin). The primers used for qPCR were:

qPCR-Ggt-1 forward: 5′-GTTGACCAGGAAGGCAATGT-3′

qPCR-Ggt-1 backward: 5′-GAGCTATGAGCAGGAGG-3′

qPCR-actin forward: 5′-CAGGCGGTGCTTTCTCTC TA-3′

qPCR-actin backward: 5′-AGCTGTAACCGCGCTCAGTA-3′

Western blotting

Western blotting was carried out to quantify the protein levels of γ-glutamyl transferase-1 (GGT-1). Polyclonal antibodies against GGT-1 were generated by injecting the polypeptide containing the 564–579 amino acid sequence of the protein into rabbits (Biomed, Beijing, China). The antisera were then purified and the polyclonal antibody obtained (Biomed). Proteins from the 3rd instar larvae were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (12% polyacrylamide) and electrophoretically transferred to polyvinylidene difluoride membranes. The membranes were blocked in Phosphate buffered salts (PBST), which contained 5% non-fat milk powder and 0.1% Tween 20, for 1 h at room temperature (24°C), and then incubated with primary antibodies (anti-GGT-1, 1 : 50) overnight at 4°C. After three 3-min washes with PBST, the membranes were incubated with goat anti-rabbit horseradish peroxidase conjugated IgG (1 : 500, Zhongshanjinqiao, Beijing, China) for 1 h at room temperature (24°C), and then subjected to three additional PBST washes. Proteins were visualized by enhanced chemiluminescence (Pierce, Rockford, IL, USA).

Glutamate and GABA determination

To determine the levels of glutamate and GABA in whole body or larval brain, liquid chromatography-electrospray tandem mass spectrometry was used. Concentrations of glutamate and GABA were measured according to Zhu et al. (2011). Briefly, one 3rd instar larvae or 10 brains of 3rd instar larvae were homogenized in 150 μL ice-cold 0.5 M formic acid immediately after dissection and centrifuged at 13 000 g at 4°C for 30 min. The standards were spiked in the brain homogenates to prepare the calibrating solutions of glutamate and GABA (25, 50, 125, 250, 500, 1250 ng/mL). The isotope-labeled neurotransmitter was used as the internal standard and the concentration was kept consistent at 500 ng/mL.

The chromatographic separation was performed on an Agilent UHPLC 1290 series system (Agilent Technologies, Santa Clara, CA, USA). The sample was separated on an C18 column (3.0 μm i.d., 100 mm × 2.1 mm) (Advanced Chromatography Technologies, Aberdeen, Scotland). Mobile phases ‘A’ and ‘B’ were composed of 0.1% formic acid in water and 0.1% formic acid in acetonitrile, respectively. The gradient elution profile was chosen as follows: 0–2 min, isocratic gradient 1.0% (B); 2–6 min, linear gradient 1.0–90.0% (B); 6–10 min, isocratic gradient 90.0% (B). A pre-equilibration period of 5 min was used between each run. The flow rate was 0.2 mL/min. The column temperature was 25°C, and the injection volume was 8 μL.

Eluates were detected using an Agilent QQQ-MS/MS (6460A) (Agilent Technologies) equipped with an electrospray ionization ion source in positive ion mode. The following conditions were optimized: drying gas, nitrogen (12 L/min, 350°C); capillary voltage, 4000 V; scan mode, selective reaction monitoring. The detected ion pairs, the acquired fragmentor and the collision energy were tuned with the aid of Agilent optimization software (B02.01: Agilent Technologies). The mass spectrometry calibration was performed with the autofeature of Agilent Mass Hunter Chemstation software (version B01.03: Agilent Technologies). Agilent Mass Hunter workstation software version B.01.00 was used for data acquisition and processing.

GGT activity

To test the effects of Ggt-1 null mutant and Ggt-1 over-expressional lines, enzyme activity of GGT-1 was measured according to Ikeda and Taniguchi (2001). Briefly, in the reaction mix, γ-glutamyl-p-nitroanilide was used as a donor and glycylglycine was used as an acceptor. Thirty 3rd instar larvae were homogenized in 300 μL Drosophila Ringer's solution, and centrifuged at 13 000 g for 10 min. Then, 10 μL of supernatant was added to a 1 mL reaction mix in a 96-well plate and incubated at 37°C for 60 min in the dark. Optical density values were measured (at 410 nm) before incubation and after 60-min incubation using the microplate reader TECAN infinit M200 (Tecan Group, Männedorf, Switzerland). The relative enzyme activity was normalized to the value of control w1118 larvae.

Statistical methods

Larval green-light avoidance is expressed as the percentage of larvae that turned away from their former direction when the green light was applied. The two tailed Fisher's exact test was used to assess significance between avoidance probability and zero, and differences between avoidance probabilities. To calculate SEM and add error bars to the figures, the larvae picked out from food at one time were treated as a group to generate a single value of avoidance probability, and for each stock, the larvae tested in the behavioral assay are divided into 10–16 groups.

The Student's t-test was used to assess significance between Ggt-1 transcriptional levels, GGT activity, and levels of neurotransmitters. Ggt-1 transcriptional levels, GGT activity, and levels of neurotransmitters are expressed as the mean ± SEM.

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information

Green-light avoidance is found in larval stages of Drosophila

Previous experiments have shown that blue light predominantly contributes to larval avoidance responses to light, whereas green light negligibly impacts larval light avoidance (Keene et al. 2011; Xiang et al. 2010). To confirm these findings, we examined early middle 3rd instar larval photoavoidance in a light avoidance assay (Fig. 1a, details see Materials and Methods). Our findings showed that w1118 larvae, which showed normal phototaxis in a previous study (Gong et al. 2010), responded to both blue and green light, although their responses to blue light were significantly higher compared with green light at all light intensities (= 0.0059 for 0.02 Wm−2, p = 0.0039 for 0.14 Wm−2, p = 0.0004 for 0.35 Wm−2, p = 0.00002 for 1.5 Wm−2) (Fig. 1b). At the higher blue-light intensity of 1.5 Wm−2, w1118 larvae avoided this light significantly more often than at the lower intensity of 0.14 Wm−2 (Fig. 1b, p = 0.0014), whereas larval avoidance of green light did not differ significantly at each of the two tested light intensities (Fig. 1b, for 0.02 Wm−2 and 0.14 Wm−2, p = 0.4229; for 0.02 Wm−2 and 0.35 Wm−2, p = 0.7014; for 0.02 Wm−2 and 1.5 Wm−2, p = 0.4433; for 0.14 Wm−2 and 0.35 Wm−2, p = 0.2976; for 0.14 Wm−2 and 1.5 Wm−2, p = 0.1488; for 0.35 Wm−2 and 1.5 Wm−2, p = 0.8405).

image

Figure 1. Ggt-1 knockdown enhances larval green-light avoidance. (a) Schematic of the larval light avoidance paradigm. The light spot is indicated by the white solid circle. (b) Percentage of w1118 larvae avoiding blue and green light, ranging from 0.02 Wm−2 to 1.5 Wm−2. Larvae avoided blue light more frequently at a higher intensity of 1.5 Wm−2 compared with lower intensity at 0.14 Wm−2. However, they did not respond differently to green light at either of the two tested intensities. At the same light intensities, larvae showed a higher probability of avoidance in blue light compared with green light. (c) Ggt-1 knockdown driven by actin-Gal4 significantly reduced protein expression levels of Gamma (γ-) glutamyl transferase (GGT)-1 compared with parents and control w1118 larvae. (d) Ggt-1 knockdown driven by actin-Gal4 strongly elevated green-light avoidance compared with parents and control w1118 larvae, whereas blue-light avoidance was not affected. Experiments were carried out at 0.14 Wm−2. *< 0.05, **< 0.01, n.s., not significant. Fifty to seventy larvae were tested for each condition in (b) and (d).

Download figure to PowerPoint

To investigate the mechanism underlying green-light avoidance, we used a Gal4/UAS system. Gal4 is a yeast transcription activator protein, which consists of a DNA-binding domain and an activation domain. It can specifically bind to a DNA fragment called UAS (upstream activating sequence) to activate gene transcription. In Drosophila, the Gal4 gene is placed after a native gene promoter, and the UAS controls expression of a target gene, so that the target gene can be expressed in specific cells. We first used actin-Gal4, which drove overall expression, to drive RNA interference of different genes and then tested larval green or blue-light avoidance. Although most lines manifested relatively low green-light avoidance, one line with Ggt-1 knockdown (actin>Ggt-1RNAi) demonstrated significantly enhanced green-light avoidance compared with parental control lines (Fig. 1d, p = 0.0150 for actin-Gal4/+; p = 0.0313 for UAS-Ggt-1RNAi) and w1118 larvae (Fig. 1d, p = 0.0313). However, this knockdown line exhibited similar blue-light avoidance compared with the parental control lines (Fig. 1d, p = 1.0000, for actin-Gal4/+; p = 0.6820, for UAS-Ggt-1RNAi) and w1118 larvae (Fig. 1d, p = 1.0000). The protein level of GGT-1 was significantly reduced compared with parental control lines and w1118 larvae, when Ggt-1 RNAi was driven by actin-Gal4 (Fig. 1c).

These results show that the avoidance of green light by larval flies was relatively weak compared with blue-light avoidance. However, knockdown of Ggt-1 enhanced larval avoidance of green light.

Ggt-1 is necessary for green-light avoidance in larva

The Ggt-1 gene is located on the X chromosome (Fig. 2a). To determine if Ggt-1 indeed inhibits green-light avoidance, we generated a mutant line Ggt-1∆1523 by P-element imprecise jump out. We used NP0136-Gal4, which has a P-element insertion in Ggt-1 gene, to make a local deletion. A P-element is a transposable element in Drosophila. When a P-element is mobilized from the original insertion site, a double-strand break is formed. If degradation of the ends of the break happens before repair, a local deletion will be generated (Hummel and Klämbt 2008). After sequencing, we found that Ggt-1∆1523 had a 1523 bp deletion in Ggt-1 (Fig. 2a).

image

Figure 2. Knockdown of Ggt-1 enhances green-light avoidance in middle 3rd instar larvae. (a) Schematic of wild-type and mutant loci of the Ggt-1 gene. Exons for the gene encoding Ggt-1 are represented by a solid box and introns are represented by a line. Open boxes represent UTR regions. Deleted fragments are represented by brackets. (b) Ggt-1∆1523 totally abolished Gamma (γ-) glutamyl transferase (GGT)-1 protein expression, whereas control w1118 larvae showed normal protein levels. (c) Ggt-1∆1523 showed reduced Ggt-1 transcription compared with control w1118 larvae, n = 9. (d) Ggt-1∆1523 had strongly reduced GGT activity compared with control w1118 larvae, leaving only about 20% activity. n = 9. (e) Ggt-1∆1523 had elevated green-light avoidance compared with control w1118 larvae, whereas its blue-light avoidance was similar to controls. All behavioral experiments were carried out at 0.14 Wm−2. Fifty to sixty larvae were tested for each condition. *< 0.05, ***< 0.001, n.s., not significant.

Download figure to PowerPoint

First, we validated the mutant. In Ggt-1∆1523, the mRNA level of Ggt-1 was significantly (p = 7.03 × 10−6) reduced compared with w1118 larvae (Fig. 2c) and no signal was detected for the Ggt-1∆1523 protein (Fig. 2b), suggesting that it was a null mutant. We next examined the enzyme activity of γ-glutamyl transferase (GGT) in Ggt-1 mutant larvae. In Ggt-1∆1523, enzyme activity was significantly (p = 4.39 × 10−7) lower (~20% remaining) compared with w1118 larvae (Fig. 2d), indicating that GGT-1 was the most abundant functional GGT in Drosophila larvae.

Then, we tested the mutant in the light avoidance assay. Ggt-1∆1523 larvae showed significantly (p = 0.0231) increased green-light avoidance compared with w1118 larvae (Fig. 2e). However, no significant differences were observed in blue-light avoidance between Ggt-1∆1523 larvae and control w1118 larvae (Fig. 2e, p = 0.4233).

These results demonstrate that mutation of Ggt-1 significantly and specifically increased green-light avoidance.

Ggt-1 is sufficient to deregulate green-light avoidance in larval stages

We over-expressed Ggt-1 on a mutant background to detect whether Ggt-1 is sufficient to deregulate larval green-light avoidance.

We constructed a UAS-Ggt-1 line, using the full-length Ggt-1 cDNA. When this line was driven by an overall expressing Gal4, actin-Gal4, mRNA expression of Ggt-1 was dramatically (= 4.98 × 10−7) increased (Fig. 3b) in qPCR assay. Similarly, dense protein expression of Ggt-1 was detected (Fig. 3a), indicating that this over-expressing system worked at both the mRNA and protein level. Moreover, GGT activity was significantly (= 1.46 × 10−6) increased in larvae over-expressing Ggt-1 compared with w1118 larvae (Fig. 3c), providing further evidence that the over-expressed enzyme was functional.

image

Figure 3. Malpighian tubules (MT) is specific for the enhancement of green-light avoidance. (a) Over-expression of Gamma (γ-) glutamyl transferase (GGT)-1 driven by actin-Gal4 strongly increased GGT-1 protein levels in middle 3rd instar larvae compared with parents and control w1118 larvae. (b) Over-expressing larvae had elevated transcriptional level of Ggt-1 compared with parents and control w1118 larvae, n = 9. (c) Over-expression of normally functioning GGT-1 dramatically increased GGT activity in over-expressing larvae compared with parents and control w1118 larvae, n = 9. (d) Ggt-1 knockdown driven by C42-Gal4 strongly increased green-light avoidance compared with control w1118 larvae. Ggt-1 knockdown driven by C507-Gal4 also increased green-light avoidance compared with control w1118 larvae. Fifty to seventy larvae were tested for each condition. (e) Over-expression of Ggt-1 driven by actin-Gal4 restored green-light avoidance compared with control w1118 larvae. Over-expression of Ggt-1 driven by C42-Gal4 and C507-Gal4 also restored green-light avoidance to w1118 levels. Fifty to seventy larvae were tested for each condition. (f) C42-Gal4 driven expression in MT and salivary glands. C507-Gal4 labeled MT, guts, and salivary gland. Arrows indicate the two pairs of epithelial tubes of MT. Scale bar = 200 μm. All behavioral experiments were carried out with green light at 0.14 Wm−2. *< 0.05, ***< 0.001, n.s., not significant.

Download figure to PowerPoint

We first used actin-Gal4 to over-express Ggt-1 in all cells in the Ggt-1∆1523 background to rescue the green-light avoidance behavior. The larval green-light avoidance probability was brought down to a level as low as control w1118 larvae (Fig. 3e, p = 0.8353). This finding demonstrates that Ggt-1 was sufficient to inhibit green-light avoidance.

Then, different Gal4 lines were used to drive over-expression in specific tissues. C42-Gal4 and C507-Gal4, which specifically drive over-expression in malpighian tubules (MT), both rescued green-light avoidance to the same level as in w1118 larvae (Fig. 3e, p = 0.7061 and 0.6943, respectively). However, elav-Gal4 and repo-Gal4, which drive over-expression in neurons and glial cells, respectively, did not rescue green-light avoidance (Figure S1b, p = 0.0189 and 0.0441, respectively), indicating that GGT-1 was not functional in the CNS. Gut-Gal4 (NP3609-Gal4) was also used to rescue the enhanced green-light avoidance phenotype of Ggt-1∆1523 larvae, and it also was not functional (Figure S1b, p = 0.0405).

Ggt-1 RNAi was also over-expressed in different tissues driven by different Gal4 lines. As a result, Ggt-1 knockdown in MT driven by C42-Gal4 and C507-Gal4 (Fig. 3f) significantly (p = 0.0441 and 0.0034, respectively) enhanced green-light avoidance behavior compared with w1118 larvae (Fig. 3d). However, Ggt-1 RNAi over-expression in the CNS, driven by elav-Gal4 or repo-Gal4, respectively, did not significantly (p = 1.0000 and 0.4140, respectively) elevate green-light avoidance (Figure S1a). RNAi driven by Gal4 stocks expressed in other tissues, such as the guts (NP3609-Gal4) or muscle (C57-Gal4) (Figure S1c), also did not elevate larval green-light avoidance (p = 0.6795 and 0.3785, respectively) (Figure S1a).

These results demonstrated that GGT-1 in MT was both necessary and sufficient for suppressing larval green-light avoidance.

Green light stimulates light avoidance via the Rh6 pathway

To elucidate which neurons mediate green-light avoidance, we first asked if photosensitive cells were involved. Known photosensitive cells in Drosophila larvae include Rh5-expressing PRs and Rh6-expressing PRs in BOs, Class IV multidendritic neurons on the body wall, and Cryptochrome (CRY) neurons in the brain (Emery et al. 1998, 2000; Xiang et al. 2010). We used ppk-Gal4 and CRY-Gal4 to specifically drive gene over-expression in Class IV multidendritic neurons and CRY neurons, respectively. Blocking Class IV multidendritic neurons or CRY neurons by tetanus toxin light chain (TNTG) did not change the level of larval responses to green light (Fig. 4, p = 0.6709 for ppk > TNTG;= 1.0000 for Cry TNTG). However, green-light avoidance was greatly lost in rh6-mutant larvae (= 0.0054), whereas rh5-mutant larvae responded to green light normally (Fig. 4, p = 1.0000). Therefore, these results indicated that Rh6 mediates green-light avoidance. Furthermore, Ggt-1 and Rh6 double mutants also showed very little green-light avoidance, which was not significantly different compared with Rh6 mutant (Fig. 4, p = 1.0000). This result indicated that elevated green-light avoidance was also mediated by Rh6.

image

Figure 4. Rhodopsin6 (Rh6) mutants exhibit low green-light avoidance. rh6 mutant and Ggt-1;rh6 double-mutant larvae showed very low green-light avoidance compared with control w1118 larvae. rh5 mutant larvae showed similar green-light avoidance to control w1118 larvae. Genetically blocking CRY neurons or Class IV multidendritic neurons did not affect green-light avoidance. Fifty to seventy larvae were tested for each condition. All behavioral experiments were carried out with green light at 0.14 Wm−2. **< 0.01, n.s., no significance.

Download figure to PowerPoint

Ggt-1 affects larval green-light avoidance via the regulation of glutamate

Drosophila larvae have relatively high levels of hemolymph glutamate compared with mammals and humans (McDonald 1975; Pierce et al. 1999; Augustin et al. 2007; Piyankarage et al. 2008). Because MT has a large surface area exposed to hemolymph fluid and glutamate is produced via GGT-mediated hydrolysis (Tate and Meister 1981), it is reasonable to assume that GGT expression in MT affects total glutamate concentrations in hemolymph. We therefore examined glutamate levels in w1118 and Ggt-1∆1523 larvae. At the whole-body level, glutamate level was significantly (= 0.0089) decreased in Ggt-1 null mutants (Fig. 5a). Moreover, in the brain, glutamate levels were also significantly (= 0.0286) decreased (Fig. 5a), which indicates that GGT-1 affects the level of glutamate in the brain.

image

Figure 5. Gamma (γ-) glutamyl transferase (GGT)-1 affects larval green-light avoidance via regulation of glutamate. (a) Ggt-1∆1523 showed lower glutamate levels in both the whole body and brain compared with control w1118 larvae. GABA levels were also decreased in the brains of Ggt-1∆1523 larvae. n = 9. (b) Ggt-1∆1523 fed glutamic acid showed a tendency toward decreased green-light avoidance with increasing dosages of glutamic acid in their food. With concentrations of more than 20 μM, larvae showed significantly decreased green-light avoidance compared with Ggt-1∆1523 larvae fed normal food. w1118 larvae fed glutamic acid ranging from 1 to 200 μM showed normal green-light avoidance compared with w1118 larvae fed normal food. Fifty to eighty larvae were tested for each condition. All behavioral experiments were carried out with green light at 0.14 Wm−2. *< 0.05, **< 0.01.

Download figure to PowerPoint

To test whether glutamate impacts larval green-light avoidance behavior, glutamate concentrations ranging from 1 to 200 μM were added to the food of Ggt-1∆1523 larvae. Overall, compared with larvae fed with normal food, the percentage of green-light avoidance was significantly decreased with increasing concentrations of glutamate (starting from 20 μM, = 0.0308) (Fig. 5b). Larval green-light avoidance gradually decreased to the levels of w1118 larvae with glutamate concentrations exceeding 20 μM (Fig. 5b). Green-light avoidance was not significantly changed in w1118 larvae-fed glutamate, although a tendency toward decline was apparent (Fig. 5b).

Taken together, these results indicated that the hydrolytic effects of Ggt-1 were a source of glutamate, which can spread to the brain, and Ggt-1 affected green-light avoidance via the regulation of glutamate.

GABA directly inhibits green-light avoidance in larvae via GABAA receptors

How does glutamate inhibit green-light avoidance? There exists at least three possibilities. First, glutamate may function as an excitatory neurotransmitter to excite inhibitory interneurons, which suppresses circuits underlying green-light avoidance. Second, glutamate may function as an inhibitory neurotransmitter to directly suppress circuits underlying green-light avoidance. Third, glutamate may be used to generate GABA through catalysis by glutamate decarboxylase (GAD), which then suppresses circuits underlying green-light avoidance.

To test the first two hypotheses, we induced knockdown of Vesicular glutamate transporter (VGlut). VGlut is the only vesicular glutamate transporter in the Drosophila genome (Daniels et al. 2006). Because it regulates the loading of glutamate into synaptic vesicles within the synaptic terminal, knocking down VGlut will decrease the efficiency of glutamatergic neurotransmission. These results show that green-light avoidance was not affected compared with control w1118 larvae (Fig. 6a, p = 0.8316). To test the third hypothesis, we first measured the concentrations of GABA in the brains of Ggt-1 mutants. Our results show that GABA concentration was significantly (Fig. 6a, p = 0.00587) reduced in the larval brains of Ggt-1∆1523 compared with w1118 larvae. Then, the glutamic acid decarboxylase 1 mutant L352F was examined in the larval light avoidance assay. In L352F, lower GAD activity and higher glutamate level have been observed (Featherstone et al. 2000, 2002). In our assay, elevated green-light avoidance was observed (Fig. 6b, p = 0.0253), whereas blue-light avoidance was not significantly different from that of control w1118 larvae (Fig. 6b, p = 0.8452). Taken together, these results indicated that it was GABA, not glutamate, that directly controls green-light avoidance.

image

Figure 6. GABA inhibits green-light avoidance via Resistant to Dieldrin (RDL). (a) Knockdown of VGlut by RNAi did not change the probability of green-light avoidance. (b) glutamic acid decarboxylase 1 (Gad1) mutant L352F showed elevated green-light avoidance compared with control w1118 larvae, whereas blue-light avoidance was not significantly affected. (c) Green-light avoidance of Rdl RNAi driven by elav-Gal4 was strongly elevated compared with parents and control w1118 larvae. All behavioral experiments were carried out at 0.14 Wm−2. *< 0.05, n.s., not significant. Fifty to seventy larvae were tested for each condition.

Download figure to PowerPoint

We next explored the function of the GABAA receptor, Resistant to Dieldrin. Green-light responses were significantly increased in Rdl RNAi driven by elav-Gal4 compared with the parental lines, elav-Gal4/+ (= 0.0157), UAS-RdlRNAi/+ (= 0.0380) and control w1118 larvae (p = 0.0231) (Fig. 6c).

Overall, these results indicate that GABA acts directly to modulate green-light avoidance via the GABAA receptor, Resistant to Dieldrin.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information

In vertebrates and invertebrates, light sensing is important for detecting foods, mates, predators, and prey, or simply for circadian entrainment, pupil reflexes, and phototaxis. Drosophila share many key characteristics with mammals. For example, they use glutamate as their main excitatory neurotransmitter and GABA as their main inhibitory neurotransmitter. Light sensing in Drosophila also shares many key characteristics with mammalian light sensing. Like in mammals, photoreceptors used to sense light in Drosophila are also composed of a vitamin A-based chromophore and opsins, which have seven-transmembrane structures (Yau and Hardie 2009). The human eye perceives equal absolute irradiance green and blue as brighter green (Rovamo et al. 1996). Similarly, Drosophila perceives blue and green as having a different brightness (Zhou et al. 2012). Moreover, in Drosophila larval light avoidance, blue light dominates light avoidance, whereas green light produces very weak avoidance. Here, we reveal the molecular basis of differences between green- and blue-light avoidance in Drosophila larvae.

In previous studies, Rh6 mutation or ablation of Rh6 PRs did not affect light avoidance, but Rh5 mutation or ablation of Rh6 PRs diminished light avoidance (Keene et al. 2011; Kane et al. 2013). These results indicate that green-light-sensitive Rh6 PRs are not required for light avoidance. Experiments by Xiang et al. (Xiang et al. 2010) have shown that larvae have weak green-light avoidance. In our experiments, we also observed larval responses to green light, and in green-light-sensitive Rh6 mutants, such responses were almost eliminated.

Using RNAi techniques to screen genes involved in light avoidance behavior, we found that knocking down Ggt-1 enhances larval green-light avoidance, as attested by Ggt-1 null mutants. In mammals, GGT is most abundant in the kidneys (Tate and Meister 1981). In Drosophila, we also found that expression of Ggt-1 in MT (organs analogous to the kidneys in mammals) affects larval green-light avoidance. As in mammals, GGT can hydrolyze GSH to generate glutamate. In Drosophila, we have observed that GGT-1 regulates the levels of glutamate in the brain.

GGT plays an important role in maintaining the homeostasis of GSH through γ-glutamyl cycle in mammals (Zhang et al. 2005) and it is the most abundant non-protein thiol in cells, which has multiple critical biological functions, such as protecting cells from oxidative or other forms of stresses (Dickinson and Forman 2002). In our experiments, increased green-light avoidance is not the consequences of stresses because Ggt-1 null mutants do not show a significant difference in the percentage of light avoidance for blue light compared with control larvae. We also tested Ggt-1 null mutants in the light-dark choice assay, and they showed similar light avoidance to w1118 larvae (Figure S2).

In mammals, GGT can hydrolyze GSH, providing a physiological reservoir of glutamate (Elce and Broxmeyer 1976; Koga et al. 2011), the principal excitatory neurotransmitter of the CNS. Our findings that glutamate levels are decreased in Ggt-1 null mutant indicate that hydrolytic effect of GGT-1 is also a source of glutamate in Drosophila larvae. Because of the large surface of MT bathing the hemolymph and the location of GGT in the outer membrane, it is reasonable to assume that glutamate generated by GGT proteins in MT cells can be directly released and diffused in the hemolymph. A recent study shows that hormones produced by prothoracicotropic hormone neurons can be secreted into and spread in hemolymph and functions at BOs and class IV dendritic arborization neurons to affect larval light avoidance (Yamanaka et al. 2013), which indicates that hemolymph is a medium to spread neuromodulators. This can also explain our results that glutamate generated by MT might spread in hemolymph and be transported to the brain to impact green-light avoidance.

Glutamate is the main excitatory neurotransmitter in mammals. To test whether glutamate functions directly as a neurotransmitter in larval green-light avoidance, Vglut was knocked down to decrease glutamatergic neurotransmission. VGlut is the homolog of Vglut in mammals and only one single vesicular glutamate transporter exists in the Drosophila genome (Daniels et al. 2006). It takes charge in loading glutamate into synaptic vesicles within the synaptic terminal and thus it is important for synaptic output. Our findings show that decreasing glutamatergic neurotransmission does not release the inhibited green-light avoidance in larvae, indicating that glutamate may not function directly in the light avoidance circuit as a neurotransmitter. As in mammals, glutamate can be transformed to GABA by catalytic effects of GAD in Drosophila. Focusing on the possible role of GABA, our findings reveal that GABA levels in the brain indeed decreases in Ggt-1 mutants. Both glutamic acid decarboxylase 1 mutants and GABAA receptor Rdl knockdown lines can elevate green-light avoidance in larvae. Hamasaka et al. (2005) have shown that GABA signals entered the larval optic center, where the axons of BOs project and can be traced back to the lateral midbrain. These observations suggest that GABA signals may be released in this area to inhibit green-light avoidance. Therefore, we hypothesize that, in wild-type larvae, green light signals from BOs project to larval optic center, and are inhibited by GABA signals. More study is needed to test this hypothesis.

Although the biological significance of the inhibition of responses to green light in wild-type larvae is unknown, we speculate that it may be related to larval foraging behavior for food. Green light has a longer wavelength compared with blue light, and therefore reaches deeper layers of food after the food is diluted by larvae. Blue light, however, is easily scattered on the surface of food because of its shorter wavelength. When larvae crawl on the surface of food, they are driven by blue light to dig into the food. After digging into the deeper layers of food, where green light is more abundant, the need to generate robust avoidance responses is no longer required. This phenomenon may be important for preventing larvae from digging too deeply into food.

Our results reveal that green-light avoidance is inhibited by GABA signals. Since our results indicate that green-light avoidance and blue-light avoidance may involve different neural circuits, future studies could further explore the neural circuits generating green-light avoidance in the brain and how the signals from green light are inhibited by GABAergic neurons to better understand light avoidance behavior.

Acknowledgments and conflict of interest disclosure

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information

We are grateful to Yongqing Zhang (Institute of Genetics and Developmental Biology, CAS) and Justin Blau (Department of Biology, New York University) for providing the flies. We thank Haiyun Gong, Yanchao Cai, Jing Liu, and Xudong Zhao of the IBP core facility center for technical assistance. We also thank the Bloomington Drosophila Stock Centers, the Vienna Drosophila RNAi Center, Drosophila Genetic Resource Center at Kyoto Institute of Technology, and Center of Biomedical Analysis at Tsinghua University for fly stock. This work was supported by the National Natural Sciences Foundation of China (31030037), the Ministry of Science and Technology of China (2012CB825504), and the Strategic Priority Research Program B of the Chinese Academy of Sciences (XDB02040200), the External Cooperation Program of BIC, Chinese Academy of Sciences (GJHZ201302).

All experiments were conducted in compliance with the ARRIVE guidelines. The authors have no conflict of interest to declare.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. Acknowledgments and conflict of interest disclosure
  7. References
  8. Supporting Information
FilenameFormatSizeDescription
jnc12735-sup-0001-FigS1-S2.pdfPDF document218K

Figure S1. CNS and other tissues are not required for the enhancement of green-light avoidance.

Figure S2. Ggt-1 null mutants show normal light preferences in a light-dark choice assay.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.