Immediate-Early Promoter-Driven Transgenic Reporter System for Neuroethological Research in a Hemimetabolous Insect

Abstract Genes expressed in response to increased neuronal activity are widely used as activity markers in recent behavioral neuroscience. In the present study, we established transgenic reporter system for whole-brain activity mapping in the two-spotted cricket Gryllus bimaculatus, a hemimetabolous insect used in neuroethology and behavioral ecology. In the cricket brain, a homolog of early growth response-1 (Gryllus egr-B) was rapidly induced as an immediate-early gene (IEG) in response to neuronal hyperexcitability. The upstream genomic fragment of Gryllus egr-B contains potential binding sites for transcription factors regulated by various intracellular signaling pathways, as well as core promoter elements conserved across insect/crustacean egr-B homologs. Using the upstream genomic fragment of Gryllus egr-B, we established an IEG promoter-driven transgenic reporter system in the cricket. In the brain of transgenic crickets, the reporter gene (a nuclear-targeted destabilized EYFP) was induced in response to neuronal hyperexcitability. Inducible expression of reporter protein was detected in almost all neurons after neuronal hyperexcitability. Using our novel reporter system, we successfully detected neuronal activation evoked by feeding in the cricket brain. Our IEG promoter-driven activity reporting system allows us to visualize behaviorally relevant neural circuits at cellular resolution in the cricket brain.


Introduction
Insects are the largest and most diverse group of organisms on Earth. They originated ϳ500 million years ago, at almost the same time as the first terrestrial plants (Misof et al., 2014). Now, they represent ϳ80% of all living organisms and are found in almost all terrestrial and freshwater environments. As a result of their evolution and diversification, insects show a wide variety of behaviors, including locomotion, feeding, molting, diapause, and social behaviors (e.g., agonistic interaction, courtship and mating behavior). Therefore, insects have been widely used for neuroethological studies over the past decades.
Recent insect neuroethology largely relies on the fruit fly Drosophila melanogaster, a versatile model organism with particularly powerful neurogenetic tools [e.g., an abundant collection of transgenic lines (Jenett et al., 2012;Manning et al., 2012), recombination-based genetic systems for targeted gene expression in specific neural circuits (Lee and Luo, 2001;Griffin et al., 2014), optogenetic/thermogenetic tools and genetically encoded calcium indicators (Owald et al., 2015;Riemensperger et al., 2016)], which allow us to manipulate/monitor the activity of specific neural circuits during behavior. In contrast, a relatively small number of researchers use non-Drosophila insects despite the importance of a comparative approach to understanding the general principles and evolution of insect behavior and its underlying neural mechanisms. Without sophisticated genetic tools, researchers have to depend on limited experimental techniques such as electrophysiology, neuroimaging, and behavioral pharmacology. Although these classic techniques are well established, they have some inevitable weaknesses and limitations: for example, neurophysiological techniques are vulnerable to mechanical disturbances caused by animal movements, and pharmacology provides us limited information on the involvement of receptors/signaling cascades in a certain behavior. To facilitate neuroethological studies in non-Drosophila insects, a technical breakthrough is necessary to compensate/overcome these problems. In recent behavioral neuroscience, histologic detection of activityregulated genes has been widely employed for retroactive labeling of behaviorally relevant neural circuits (Clayton, 2000;Guzowski et al., 2005;Maruska et al., 2013;Kawashima et al., 2014). In this study, aiming to introduce such a powerful technique to non-Drosophila insects, we established a novel transgenic reporter system for whole-brain activity mapping in a hemimetabolous insect.
In neurons, the expression of a wide variety of genes is regulated in response to increased neuronal excitation and activation of cellular signaling pathways (Flavell and Greenberg, 2008), which are involved in various cellular functions, including transcription regulation, signaling pathways, metabolism, and synaptic function (Loebrich and Nedivi, 2009;. In terms of expression time course and regulatory mechanisms, activity-regulated genes can be classified into two categories, immediate-early gene (IEGs) and delayed-early genes (DEGs). The neuronal IEGs are rapidly and transiently induced in response to neuronal activation and their expression does not require de novo protein synthesis (Clayton, 2000). Some IEGs encode transcription factors whose protein products, in turn, regulate expression of DEGs. Therefore, transcription factor IEGs are considered as the first wave of genomic response to shape the cellular response toward various physiological/environmental stimuli. In the nervous system, transcription factor IEGs, such as c-fos and egr-1/zif268/NGFI-A, are expressed in response to a wide variety of stimuli and in multiple cell types; therefore, they are widely used for activity mapping in the vertebrate system. In the early studies, researchers directly detected the activityregulated expression of gene products of IEGs to map behaviorally evoked neuronal activity. In the recent studies, the promoter regions of activity-regulated genes were used in combination with optogenetic tools to address the functional importance of behaviorally relevant circuits Minatohara et al., 2016).
So far, activity-regulated gene-based activity mapping has been conducted on limited insect species, such as the honeybee, Drosophila, and silk moth. In the brain of the honeybee, the noncoding RNA kakusei and a homolog of transcription factor c-jun, jun-related antigen (jra), were used to map a neural activity associated with the foraging behavior and waggle dance Kiya and Kubo, 2011) and defensive behaviors (Alaux and Robinson, 2007;Ugajin et al., 2012). In the brain of Drosophila and the silk moth, a nuclear receptor hr38 (NGFI-B/nur77/ NR4A1 homolog) was used for activity mapping (Fujita et al., 2013). In contrast, there has been no report on the activity-regulated genes or neuronal IEGs in basal hemimetabolous insects. Moreover, the activity-regulated promoter has not been characterized in any insect species.
Crickets have been widely used for neuroethology and behavioral ecology for decades (Huber et al., 1989;Horch et al., 2017) with a particular focus on their prominent social behaviors including courtship and aggressive behaviors (Adamo et al., 1995;Pollack, 2000;Hedwig, 2006;Stevenson and Schildberger, 2013). Moreover, modern genetic techniques (e.g., transposon-mediated transgenesis, genome editing techniques) have been introduced to the two-spotted cricket Gryllus bimaculatus over the last decade (Nakamura et al., 2010;Watanabe et al., 2012;Horch et al., 2017). For this reason, we chose G. bimaculatus as a model system to establish a transgenic activity reporting system using an activity-regulated promoter. In the present study, we first identified neuronal IEGs in the cricket to isolate and characterize the activity-regulated promoter in basal insects. We then established a transgenic reporter line for a whole-brain, single cell-resolution activity mapping, which is required for further functional studies of behaviorally relevant neural circuits.

Materials and Methods Animals
A wild-type strain of two-spotted crickets G. bimaculatus DeGeer was inbred for decades in our laboratory (Hokudai WT strain). The Hokudai WT strain was used for identification and initial expression analysis of the candidate IEGs, and promoter analysis. A congenic white-eye strain (Hokudai gwhite strain) was established by five-time backcrossing of gwhite mutant (kindly provided by Prof. Noji in Tokushima Univ., Japan) to the Hokudai WT strain. The Hokudai gwhite strain was used for transgenesis. Crickets were reared in a group on a 14-h:10-h light/dark cycle at 28°C. They were fed a diet of insect food pellet (Oriental Yeast Co.) and water ad libitum. Specimens of the European honeybee Apis mellifera L. were collected from colonies maintained at Tamagawa University. Specimens of the house cricket Acheta domesticus were purchased at a local pet store. Specimens of the katydid Gampsocleis buergeri and migratory locust Locusta migratoria were collected in Tokyo (Japan).

Identification of candidate neuronal IEGs
We amplified cDNAs encoding a full-length protein coding sequence of three candidate IEGs and a partial cDNA of Gryllus hr38 from cDNA libraries derived from the brains of adult crickets prepared in Watanabe et al. (2011).

Partial cDNA cloning of Gryllus fra
Partial cDNA of Gryllus fra was amplified using genespecific primers (GSPs) designed on the basis of G. bimaculatus expressed sequence tag (EST) clones corresponding to fra (GenBank IDs AK282142 and AK254978).

Partial cDNA cloning of Gryllus jra
Partial cDNA of Gryllus jra was amplified using GSPs designed on the basis of a G. bimaculatus EST clone corresponding to Gryllus jra (GenBank ID AK263019).

Partial cDNA cloning of Gryllus egr
Partial cDNA of Gryllus egr was amplified using degenerate primers. Degenerate primers were designed on the basis of conserved amino acid sequences (GVQLAEY, TSKGHEI, FQCRICMR and HAKVHLK) among insect Egr homologs. Sequences of the degenerate primers are listed in Table 1.

Partial cDNA cloning of Gryllus hr38
Partial cDNA of Gryllus hr38 was amplified using degenerate primers. Degenerate primers were designed on the basis of conserved amino acid sequences (NRCQFC and RDDQELL) among insect hr38 homologs. Sequences of the degenerate primers are listed in Table 1.

Full-length cDNA cloning of Gryllus fra, jra, and egr-B
First, nucleotide sequences flanking the translational initiation and termination sites of each gene were ob-tained by 5= and 3= rapid amplification of cDNA ends (RACEs) using the FirstChoice RLM-RACE kit (Ambion). RACE PCR was conducted using GSPs designed on the basis of the nucleotide sequences of partial cDNAs of each gene. Complementary DNAs containing full-length ORF of the genes were amplified using GSPs designed at the 5= and 3= untranslated region of the genes. Primers used to amplify cDNAs containing full-length ORF are listed in Table 1. All PCRs were performed using the Q5 High-Fidelity DNA polymerase (New England Biolabs). Amplified cDNA fragments were cloned into the pGEM-T Easy vector (Promega), and their nucleotide sequences were determined.

Sequence comparison and structural analysis of the proteins
The deduced amino acid sequences of Gryllus fra, jra, egr, and hr38 were aligned with those of the corresponding parts of homologous genes of other species using the MAFFT (Katoh and Standley, 2013; RRID: SCR_011811) or MUSCLE algorithms (Edgar, 2004;RRID: SCR_011812) and refined by manual inspection on the Geneious program (ver. 9) created by Biomatters (available from http://www.geneious.com/; RRID: SCR_010519). The leucine zipper domain of Gryllus Fra and Jra proteins, and the zinc finger motifs of Gryllus Egr protein were predicted by the SMART program (Schultz et al., 1998;Letunic et al., 2012; available from http://smart.embl-heidelberg.de; RRID: SCR_005026).

Induction of neuronal hyperexcitability
Adult male crickets 1 week after the imaginal molt were used for pharmacological experiment. Crickets were individually isolated in a 100-ml beaker (ø4.5 cm) for 3 days without food and water before pharmacological treatment. To induce neuronal hyperexcitability, 5 mM picrotoxin (PTX; Sigma-Aldrich; CAS: 124-87-8) dissolved in the cricket physiological saline (140 mM NaCl, 10 mM KCl, 1.6 mM CaCl 2 , 2 mM MgCl 2 , 44 mM glucose, 2 mM TES, pH 7.2) containing 5% dimethyl sulfoxide (DMSO) was injected. The cricket physiological saline containing 5% DMSO was injected as a vehicle control. To block de novo protein synthesis, 20 mM cycloheximide (Sigma-Aldrich; CAS: 66-81-9) dissolved in a cricket physiological saline was injected 1 h before PTX injection. 3 l of each solution was injected into the head capsule using a 27gauge needle attached to a 10-l microsyringe (Hamilton 701 LT Syringe; Sigma-Aldrich).

RNA extraction and reverse transcription for the initial expression analysis
Crickets were anesthetized on ice, and the brain (the supraesophageal and subesophageal ganglia without the optic lobes) was dissected in ice-cold saline, immediately chilled in liquid nitrogen, and stored at -80°C until use. Total RNA was extracted from a single cricket brain using the PureLink RNA Mini kit (Life Technologies) according to the manufacturer's instruction. RNA samples were reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Life Technologies). 0.5 g of total RNA was reverse transcribed in a 20-l reaction according to the manufacturer's instruction.

RNA extraction and reverse transcription for the other RT-qPCR expression analyses
Total RNA was extracted from a single cricket brain using the TRIzol reagent (Life Technologies). Genomic DNA contamination was digested with DNase I (1 U in 10 l reaction; TaKaRa) at 37°C for 1 h, followed by incubation at 75°C for 10 min to inactivate the enzyme. RNA samples were reverse transcribed using the High-Capacity cDNA Reverse Transcription Kit (Life Technologies). 1.5 l of total RNA was reverse transcribed in a 10-l reaction according to the manufacturer's instruction.

Quantitative PCR
Quantitative PCR analysis was performed using the KAPA SYBR FAST qPCR Kit (Kapa Biosystems) and the Eco Real-Time PCR System (Illumina). 0.5 l of each reverse transcription product was added to a 10 l qPCR reaction. Quantitative PCR reaction was performed at 50°C for 2 min and 95°C for 5 min followed by 40 cycles of 95°C for 10 s and 60°C for 30 s each. Sequences of the primers used for qPCR are listed in Table 1.

Determination of the transcription start sites (TSSs) of Gryllus and Apis egr-B homologs
We performed the 5= RACE using the FirstChoice RLM-RACE kit (Ambion) to clone the 5= ends of Gryllus and Apis egr-B homologs. RNA processing reactions were con-  Figure 1. Expression of candidate neuronal IEGs in the brain of cycloheximide pretreated crickets 30 min after PTX injection. A, PTX-induced neuronal hyperexcitability in the cricket. Crickets show seizure-like behavior ϳ2 min after PTX injection. 1 h before PTX/vehicle injection, 20 mM cycloheximide was injected to block de novo protein synthesis. B-G, Expression of (B) Gryllus fra total transcript, (C) fra-A isoform, (D) fra-B isoform, (E) jra, (F) egr-B, and (G) hr38 in the brains of cycloheximide pretreated crickets 30 min after injection of vehicle (5% DMSO in saline) or PTX. Expression levels of each target gene were normalized with that of Gryllus ef1␣ gene ( Fig. 1-1). RT-qPCR analyses were performed on eight biological replicates. Box plots indicate the 25th to 75th percentile ranges and central values. Error bars indicate the 5th to 95th percentile ranges. The "ϩ" denotes the mean. Asterisks denote statistical significance ‫,ء(‬ p Ͻ 0.05). See Table 3 for the details of statistical analysis. See Figs. 1-2, 1-3, 1-4, and 1-5 for the structures of the encoded proteins of candidate neuronal IEGs. ducted according to the manufacturer's instruction. Reverse transcription was conducted using the Superscript III reverse transcriptase (Life Technologies) according to the manufacturer's instruction. 5= RACE PCRs were conducted using the Q5 High-Fidelity DNA polymerase (New England Biolabs). Complementary DNAs were inserted into the pGEM-T Easy vector (Promega) by standard TA cloning procedure using the T4 DNA ligase or the Gibson assembly technique using the Gibson Assembly Master Mix (New England Biolabs). Complementary DNA fragments for the Gibson assembly cloning were amplified with primers with 5= extensions (5=-GCC GCG GGA ATT CGA TT-3= was attached to the 5= end of the 5= RACE inner primer; 5=-CCG CGA ATT CAC TAG TGA TT-3= was attached to the 5= end of the gene-specific reverse primer). More than 20 cDNA clones were sequenced to determine the TSSs.

Isolation of the upstream regions of orthopteran egr-B homologs
To isolate the putative core promoter region, genomic DNA of four orthopteran insects were extracted using the Wizard Genomic DNA Purification Kit (Promega). Based on the high-level sequence similarity in the core promoter region of egr homologs in basal insects, we designed primers to amplify the putative core promoter region of orthopteran egr-B homologs. PCRs were performed using a pair of the forward and reverse primers listed in Table 1. Then, the upstream regions of orthopteran egr-B homologs were obtained by using inverse PCR. The procedure for inverse PCR was modified from Watanabe and Aonuma (2014). Genomic DNAs of three orthopteran insects were digested with BfaI, BglII, DpnII, EaeI, HhaI, HindIII, MseI, MspI, NlaIII, Taq␣I, XapI or XceI. 10 ng of restriction fragments were circularized using the T4 DNA ligase (Thermo Fisher Scientific) in 10 l reaction. Primers were designed on the basis of the nucleotide sequence of the putative core promoter of egr-B homologs. Inverse PCRs were repeatedly performed to obtain Ͼ1.5-kbp genomic DNA fragments flanking the first exon of egr-B homologs. All PCRs were conducted using the Q5 High-Fidelity DNA Polymerase (New England Biolabs). PCR products were cloned into the pGEM-T Easy vector (Promega) and their sequences were determined.

Database search and sequence comparison of stimulus-regulated transcription factors among vertebrates and insects
To understand the structural features of the upstream sequence of insect/crustacean egr-B homologs, we searched potential binding sites for stimulus-regulated transcription factors (TFs) conserved among vertebrates, insects, and crustaceans. There have been few reports on the DNA-binding specificities of stimulus-regulated TFs in insects and crustaceans, so we examined sequence conservation in the DNA-binding domains of stimulusregulated TFs between mouse and insects, to assess whether each candidate TF in insects might have DNAbinding properties similar to vertebrate homologs. We analyzed components of the stimulus-regulated transcription factor complexes which are known to involved in regulating stimulus-regulated gene expression (Dolmet-sch et al., 1997;Hu et al., 2001;Tullai et al., 2007) as follows: AP-1, CREB, C/EBP, Egr, NFAT, MEF2, and SRF, which include Fos-family proteins, Junfamily proteins, ATF/CREB family proteins, Maf-family proteins, C/EBPs, Egr-family proteins, NFATs, MEF2, and SRF. We first retrieved mouse homologs of stimulusregulated TFs from the GenBank protein sequence database. Then, BLASTp searches were conducted in search for homologs of stimulus-regulated TFs in Drosophila melanogaster, Tribolium castaneum, Apis mellifera, Acyrthosiphon pisum, and Zootermopsis nevadensis. We also performed tBLASTn searches on the Gryllus firmus transcriptome shotgun assembly (TSA) database to find cDNA sequence of the cricket homologs of the target genes. The amino acid sequences of the DNA binding domains of stimulus-regulated TFs of mouse and insects were aligned using the MAFFT or MUSCLE algorithms on the Geneious program (ver. 9). GenBank IDs of stimulusregulated TFs are listed in Table 3-1.

Structural analysis of the potential gene regulatory region of egr genes
Potential transcriptional regulatory sequences in the upstream sequences of insect/crustacean egr-B homologs were predicted using the LASAGNA-Search 2.0 program ; available from http:// biogrid-lasagna.engr.uconn.edu/lasagna_search/; RRID: SCR_010883). Potential binding sites for stimulus-regulated transcription factors common to vertebrates and insects were searched. Phylogenetic footprinting analysis of the upstream regions of the polyneopteran egr-B homologs was conducted using the mVISTA program (Frazer et al., 2004; available from http://genome.lbl.gov/ vista/mvista/submit.shtml) with default setting. GenBank IDs of genomic sequences are listed in Table 2.

Construction of piggyBac donor plasmid
A piggyBac donor plasmid pXL-BacII[3xP3-mCherry] {Gbegr-EYFPnls:PEST-2xARE} was constructed to examine promoter activity of Gryllus egr-B in vivo. This plasmid contains the expression cassettes for an IEG reporter (Gbegr-EYFPnls:PEST-2xARE) and a visible selection marker (3xP3-mCherry). The IEG reporter cassette consists of (1) a 2.2-kb upstream sequence of Gryllus egr-B coding sequence; (2) a coding sequence of nucleartargeted EYFP C-terminally fused to a PEST domain of mouse ornithine decarboxylase (EYFPnls:PEST); (3) two repeats of synthetic AU-rich elements (2xARE); and (4) a SV40 3= untranslated region (UTR). The expression cassettes of the IEG reporter and 3xP3-mCherry were separated by the gypsy insulator sequence. The donor plasmid was constructed through the following procedures.

Gbegr-EYFPnls:PEST-2xARE expression cassette
A ϳ2.2-kb genomic fragment upstream to Gryllus egr-B was amplified by PCR from Gryllus bimaculatus genomic DNA purified using the Wizard Genomic DNA Purification Kit (Promega). A nucleotide sequence encoding the nuclear localization signal of the SV40 Large T-antigen (PKKKRKV) was added to the 3= end of the coding sequence of EYFP derived from the pBSII-ITR1.1k-EYFP plasmid by PCR. The coding sequence of the PEST domain of mouse ornithine decarboxylase (residues 416 -461 of NP_038642.2) was synthesized by Eurofins Genomics. Two repeats of synthetic AU-rich elements (5=-TTT ATT TAT TTA TTT ATT TA-3=) were added to the 5= end of the SV40 3= UTR derived from the pBSII-ITR1.1k-EYFP plasmid by PCR. The DNA fragments of the promoter, protein coding sequence, and 3= UTR were assembled in order by restriction enzyme digestion and ligation, or by using the Gibson assembly technique.
All PCRs were conducted using the Q5 High-Fidelity DNA Polymerase (New England Biolabs). Gibson assembly was performed using the Gibson Assembly Master Mix (New England Biolabs). After construction, the plasmid was amplified in Escherichia coli strain DH5ɑ and purified using the PureLink HiPure Plasmid Kit (Life technologies).

piggyBac transposon-mediated transgenesis In vitro synthesis of piggyBac transposase mRNA
To construct the expression plasmid for piggyBac transposase mRNA (pTD1-piggyBac), the coding sequence of piggyBac transposase derived from the pBSII-IFP2-orf plasmid was inserted into the pTD1 expression vector (Shimadzu). The expression cassette of the piggy-Bac transposase in the pTD1-piggyBac was amplified by PCR with the Q5 High-Fidelity DNA Polymerase (New England Biolabs) with following primers (forward 5=-GCA GAT TGT ACT GAG AGT G-3= and reverse 5=-CAG GAA ACA GCT ATG AC-3=). The PCR fragment was subsequently used as templates for in vitro transcription using the T7 mScript Standard mRNA Production System (CELLSCRIPT). Then, the 5= cap and poly(A) tail were added to the transcribed RNA using the kit. The piggyBac mRNA was purified by phenol-chloroform extraction and concentrated by ammonium acetate precipitation.

Microinjection
The piggyBac donor plasmid (1 g/l) and piggyBac mRNA (1 g/l) were dissolved in nuclease-free water (Qiagen) to make an injection solution for germline transformation. The injection solution was injected into fertilized eggs of the Hokudai gwhite strain. Eggs were laid in a wet cotton or paper towel and collected within 30 min of egg laying. They were rinsed with 70% ethanol, aligned on the hand-made injection chamber, shortly air-dried, and covered with mineral oil (Nacalai Tesque). Micropipettes were pulled on a micropipette puller (PA-81; Narishige) using thin-walled filament glass capillary (GD-1.2; Narishige). The capillaries were backfilled with the injection solution using the GELoader tip (Eppendorf) and connected to the IM-6 microinjector (Narishige) filled with mineral oil. Injection solution was injected into the dorsal posterior part of the eggs. Eggs were transferred onto the wet cotton in the glass Petri dish immediately after injection. Petri dishes were incubated at 28°C in the moist chamber until hatching.

Genetics
Embryos injected with the piggyBac donor plasmid and piggyBac mRNA were raised to adult and backcrossed to Hokudai gwhite strain. The F 1 embryos exhibiting mCherry fluorescence in the compound eyes were selected. Each transgenic cricket was repeatedly backcrossed to the Hokudai gwhite strain to generate heterozygous lines.

Genotyping
To establish IEG promoter-driven reporter lines, heterozygous transgenic lines were selected according to the inducible expression of EYFPnls:PEST protein after PTX treatment. 6 h after PTX injection, the brains were dissected for EYFP immunohistochemistry according to the experimental procedure described below. Then, the genomic flanking sequence of piggyBac insertion was determined by inverse PCR. To establish homozygous lines, each heterozygous line was inbred, and homozygous transgenic crickets were selected by PCR-based genotyping.

Inverse PCR
First, the genomic regions flanking the piggyBac insertions were amplified by inverse PCR. Genomic DNA of the heterozygous transgenic crickets was extracted using the Wizard Genomic DNA Purification Kit (Promega). Genomic DNA was digested with DpnII, HindIII, or MspI. 10 ng of restriction fragments were circularized using the T4 DNA ligase (Promega) in 10 l reaction. PCRs were performed to amplify the 5= and 3= flanking regions of the piggyBac insertion using the Q5 High-Fidelity DNA Polymerase (New England Biolabs). Following primers were used for amplification of the 5= flanking region of the insertion (primers for 1st PCR: 3=-GCT CCA AGC GGC GAC TGA GAT GTC C-3= and 5=-GCT TGT CAA TGC GGT AAG TGT CAC TG-3=; primers for nested PCR: 5=-GAC GGA TTC GCG CTA TTT AGA AAG AGA G-3= and 5=-CGG TAA GTG TCA CTG ATT TTG AAC TAT AAC G-3=). The 3= flanking region of the insertion was amplified by inverse PCR using primers designed on the basis of the nucleotide sequence of the 5= flanking region of the insertion. PCR products were cloned into the pGEM-T Easy vector (Promega) and their sequences were determined.

PCR-based genotyping
Heterozygous transgenic lines were inbred to generate homozygous transgenic animals. Each adult transgenic cricket in the inbred colonies was isolated after the imaginal molt, and homozygous animals were selected by PCR-based genotyping. Genomic DNA was isolated from the wings of newly emerged transgenic crickets using the modified HotSHOT method (Meeker et al., 2007). The tip of a hind wing (ϳ2 mg) was dissected from crickets and homogenized in liquid nitrogen. Samples were incubated at 95°C for 30 -60 min in 50 l of 50 mM NaOH. 5 l of 1 M Tris-HCl (pH 7.5) was added to each tube for neutralization. After brief centrifugation, supernatants were diluted 16 times with TE buffer (pH 8.0). PCR was conducted using the Q5 High-Fidelity DNA Polymerase (New England Biolabs) in a 10-l reaction. 0.5 l of genomic DNA solutions were added to the reactions. For genotyping, forward and reverse primers were designed at the 5= and 3= genomic regions flanking the piggyBac insertion (forward primer [line19_fw primer], 5=-CAC ATT CAC ACA TAT CCG CAG TTC-3=; reverse primer [line19_rv primer], 5=-CGT TCT TCA ATT TCA TTT TTC TCC TC-3=), respectively (Fig. 5C). PCR products were run through 1.5% agarose gel in TBE buffer and visualized by ethidium bromide. We selected homozygous transgenic animals as follows: a 358-bp genomic fragment was amplified if the cricket had no transgene or one copy of transgene (heterozygous crickets), whereas no PCR product was amplified from homozygous crickets.

Generation of anti-Gryllus Tdc2 polyclonal antibody
Complementary DNA cloning for tyrosine decarboxylase (tdc) genes The full-length cDNA of two tdc genes (Gryllus tdc1 and tdc2) was obtained as follows. A partial cDNA of Gryllus tdc1 was amplified using GSPs designed on the basis of a G. bimaculatus EST clone corresponding to Gryllus tdc1 (GenBank ID; AK258423). A partial cDNA of Gryllus tdc2 was amplified using degenerate primers designed on the basis of conserved amino acid sequences (IEYADSFNT and FFVRMVSDP) among the insect Tdc2 proteins. Then, 5= and 3= RACEs were performed as described above to determine translational initiation and termination sites of the genes. Finally, full-length cDNAs of the genes were amplified, and their nucleotide sequences were determined. Sequences of the primers are listed in Table 1.

Recombinant protein expression and antibody production
A cDNA fragment encoding amino acids 470 -683 of Gryllus Tdc2 protein (GenBank ID: BAO52000.1) was inserted between the BamHI and EcoRI sites of the pGEX-6P-1 vector (GE Health care Japan). The antigen peptide was produced as a glutathione S-transferase (GST) fusion protein in the Rossetta 2(DE3)pLysS cells (Millipore). GSTtagged protein was bound to the Glutathione HiCap Matrix (Qiagen), and antigen peptide was cleaved using the PreScission protease (GE Health Care Japan). The anti-Gryllus Tdc2 polyclonal antibody was generated in the guinea pig, and IgG fraction was purified from antiserum with Protein A (Frontier Institute; Ishikari-shi). Antibody specificity was checked by Western blot. An intensely stained band of ϳ200 kDa was detected in the lane of the central brain, which matches the calculated molecular mass of Gryllus Tdc2 protein (data not shown). Anti-Gryllus Tdc2 polyclonal antibody (1.4 mg IgG/ml) was stored in 50% glycerol at -20°C.

Whole-mount immunohistochemistry
Whole-mount fluorescent immunohistochemistry was performed according to Gonzalez-Bellido and Wardill (2012) with modifications. Crickets were anesthetized on ice, and the brain (the supraesophageal and subesophageal ganglia without the optic lobes) was dissected in ice-cold saline. The brains were fixed with 4% paraformaldehyde (PFA) in phosphate-buffered saline (PBS; P-5493, Sigma-Aldrich) at 4°C overnight. When we conducted anti-octopamine immunohistochemistry, brains were fixed with 1% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4; Electron Microscopy Sciences) with 1% sodium metabisulfite (SMB) for 2 h on ice, and with 4% PFA in PBS at 4°C overnight. The samples were treated with 0.5% sodium borohydride in Tris-buffered saline (TBS; 0.05 M Tris-HCl, 77 mM NaCl, pH 7.4) containing 0.45% SMB for 1 h.
Fluorescent images were captured using the Olympus FLUOVIEW FV1000 confocal laser scanning microscope (Olympus) or the Zeiss LSM 7 DUO laser scanning microscope (Zeiss). Microscopy parameters were adjusted to the brains of naïve crickets to prevent saturated pixels in the final images. For each brain, we captured a stack of optical sections using a 10ϫ objective lens. Image processing was conducted using Fiji software (Schindelin et al., 2012;Schneider et al., 2012;RRID: SCR_002285). To visualize intensity of reporter protein expression, the EYFP immunoreactivity was displayed with "Red Hot" or "Magenta Hot" look-up tables.

Behavioral experiments
Adult male crickets 1 week after the imaginal molt were used for behavioral experiments. Each cricket was individually isolated in a 100-ml beaker (ø4.5 cm) for 3 days without food and water before behavioral experiment.

Gustatory stimulation
Each cricket was fed a drop of 0.5 M sucrose solution (5 l) three times with 5-min interval. 1 h after feeding, crickets were collected for RT-qPCR expression analysis. 6 h after feeding, crickets were collected for immunohistochemistry.

Agonistic behavior
A pair of weight-matched crickets (weight range: 0.6 -0.8 g) were introduced into a round glass arena (ø12 cm) for agonistic interaction. We observed their behavior for 5 min, and winner and loser were determined according to following criteria: the winner sings an aggressive song and chases the loser, while the loser flees from the winner. In some cases, aggressive behavior did not escalate and dominance hierarchy was not established in 5 min. After 5-min interaction, each cricket was re-isolated in a 100-ml beaker. 1 h after the start of interaction, crickets were collected for RT-qPCR expression analysis.

Gryllus egr homolog is expressed as a neuronal IEG in the cricket brain
Although the activity-regulated genes or neuronal IEGs have not been identified in any hemimetabolous insects including the cricket, recent genome-wide analyses of activity-regulated genes in vertebrates (Nedivi et al., 1993;Matsuo et al., 2000;Wada et al., 2006;Spiegel et al., 2014) and several insect species (Guan et al., 2005;Ugajin et al., 2018) highlighted evolutionarily conserved transcription factor IEGs. On the basis of previous reports, we selected four transcription factor genes as our candidates for activity-regulated genes in the cricket as follows: fos-related antigen (fra; c-fos homolog), jun-related antigen (jra; c-jun homolog), early growth response (egr; egr-1/zif268/NGFI-A homolog), and hormone receptor 38 (hr38; NGFI-B/nur77/NR4A1 homolog).
First, we cloned cDNAs encoding four candidate activity-regulated genes from the cricket brain cDNA library. We isolated cDNAs encoding two transcript variants of Gryllus fra (fra-A and -B isoforms; LC215243 and LC215244, respectively), Gryllus jra (LC215245), Gryllus egr (LC215246), and Gryllus hr38 (LC341255; see Figs. 1-2, 1-3, 1-4, and 1-5). In Drosophila, the stripe gene (Drosophila egr homolog) has two alternative promoters to express alternative transcripts (stripe-A and -B; Frommer et al., 1996). The use of alternative promoters was also known in the honeybee. In the honeybee brain, variant I and II of AmEgr (equivalent to egr/stripe-A of other insects) are transiently expressed during metamorphosis, whereas the other isoform, variant III (equivalent to egr/ stripe-B), is induced in an activity-dependent manner (Ugajin et al., 2016). Although we did not obtain cDNA clones of egr-A in the cricket, we concluded that the egr gene expressed in the cricket brain is a homolog of egr-B found in other insects based on the similar structural ( Fig.  1-4C ) and expression characteristics.
Next, we examined whether the expressions of the candidate activity-regulated genes are rapidly induced in the cricket brain in an activity-dependent manner. In this study, the ␥-aminobutyric acid A (GABA A ) receptor blocker picrotoxin (PTX) was used to cause neuronal hyperexcitability (Saffen et al., 1988;Ugajin et al., 2013). In the cricket, injection of 3 l of 5 mM PTX caused immediate locomotor hyperactivity and, in turn, seizurelike behavior (e.g., abdominal contraction and shaking of appendages). The effect of PTX persisted at least 30 min, and most crickets recovered from the seizure within 90 min after injection. In addition, to clarify whether the candidate genes are induced as neuronal IEGs, 20 mM cy-cloheximide was injected 1 h before PTX injection to block de novo protein synthesis (Matsumoto et al., 2003). Quantitative RT-PCR analysis revealed that, 30 min after PTX injection, the expression levels of Gryllus fra-B and egr-B were significantly elevated in the brains of PTX-injected crickets (ϳ1.5-fold up-regulation) versus those of vehicle-injected crickets (Fig. 1). These data demonstrated that, in the cricket brain, Gryllus fra-B and egr-B were rapidly induced as neuronal IEGs in response to neuronal hyperexcitability.

2D
Adult male crickets 1 week after the imaginal molt were isolated for 3 days. Cricket were fed with 0.5 M sucrose solution 3 times with 5 min intervals. Brains were dissected for RNA extraction before feeding (naïve), or 1 h after feeding (n ϭ 8 in each group).

2E
Adult male crickets 1 week after the imaginal molt were isolated for 3 days. Weight-matched crickets were introduced into an arena to interact for 5 min, then re-isolated. Brains were dissected for RNA extraction before interaction (naïve), or 1 h after start of interaction (n ϭ 8 in each group).

5D
Adult male crickets 1 week after the imaginal molt were isolated for 3 days. Then, brains were dissected for RNA extraction (n ϭ 8 in each group). 1-5D Same as in Fig. 2A Next, we examined expression time course of activityregulated genes in the cricket brain after PTX injection, especially focused on Gryllus egr-B (Fig. 2). In the brains of cycloheximide pretreated crickets, the expression of Gryllus egr-B reached a maximum 60 min after PTX injection [ϳ4.5-fold up-regulation relative to both the naïve animals and the control (0 min after PTX injection)] and decreased to near baseline level by 120 min after PTX injection ( Fig. 2A). To ensure that the activity-dependent transcriptional activation of Gryllus egr-B is composed only of an immediate-early component, we compared the expression kinetics of Gryllus egr-B after PTX injection with or without the administration of cycloheximide ( Fig.  2A, B). The expression kinetics of Gryllus egr-B was not affected by blockade of de novo protein synthesis (twoway ANOVA, effect of pretreatment: F(1,98) ϭ 0.9604, p ϭ 0.3295; effect of time: F(6,98) ϭ 9.034, p Ͻ 0.0001; interaction: F(6,98) ϭ 0.3634, p ϭ 0.9004), indicating Gryllus egr-B was induced as an IEG in the cricket brain after PTX injection. Next, we examined when PTX-induced Gryllus egr-B is actively transcribed by examining the expression time course of Gryllus egr-B premature mRNA (pre-mRNA). In the brains of cycloheximide-pretreated crickets, the expression of Gryllus egr-B pre-mRNA significantly increased within 15 min after PTX injection [ϳ7fold up-regulation relative to both the naïve animals and the control (0 min after PTX injection)], was sustained at a high level for Ͼ1 h, and then decreased to near baseline level by 120 min after injection (Fig. 2C). These data indicate that, after PTX injection, transcription of Gryllus egr-B is rapidly induced and sustained at a high level while its corresponding mRNA is elevated in the brain. Finally, we tested behaviorally evoked expression of Gryllus egr-B. The expression of Gryllus egr-B in the brain was significantly increased 1 h after feeding of sucrose solution and agonistic interaction (Fig. 2D, E). These data indicated that Gryllus egr-B is induced as a neuronal IEG in response to strong, sustained, widespread neuronal hyperexcitability caused by PTX, as well as behaviorally evoked neuronal activation in the cricket brain.
Although we could not detect significant increase of Gryllus fra-A, jra, and hr38 30 min after PTX injection in the initial expression analysis (Figs. 1A, C, D), these genes are slowly up-regulated in the brain under the influence of cycloheximide after PTX injection (Figs. 1-5 and 2-1). The expression of Gryllus fra-A, fra-B, and hr38 were elevated 1 h after feeding of sucrose solution, whereas those of Gryllus fra-B and hr38 were elevated 1 h after agonistic interaction. These data indicate that all candidate activityregulated genes were expressed as neuronal IEGs in the brain with gene-specific expression kinetics and regulation mechanisms.

Cis-regulatory elements for stimulus-induced gene expression are enriched in the upstream genomic regions of insect/crustacean egr-B homologs
To construct an IEG promoter-driven reporter system, we determined the nucleotide sequences of the gene regulatory region of Gryllus egr-B. By using RNA ligasemediated 5= RACE and inverse PCR, we determined the transcription start site (TSS) and obtained a ϳ1.6-kbp upstream genomic fragment of Gryllus egr-B. Sequence comparison of the upstream regions of egr-B homologs revealed that the core promoter region (ϳ200 bases flanking the TSS) of Gryllus egr-B showed a high level of sequence similarity with the corresponding part of most of egr-B homologs of basal insects (hemimetabolous insects ϩ Coleoptera [Tribolium castaneum]) and the crustacean  Table 3 for the details of statistical analysis. See Figs. 2-1, 2-2, and 1-5 for the expression characteristics of other candidate neuronal IEGs.
Daphnia pulex (Fig. 3A; See Table 2). The core promoter region of basal insect/crustacean egr-B contained cisregulatory elements regulating stimulus-dependent gene expression (two CREs and SRE), as well as sequence elements for core promoter function such as a CCAATbox, initiator element (Inr; consensus: YCATTC), and a downstream promoter element (DPE; consensus: AGTYY; see Fig. 3A, B). Additionally, we found that two more Coding region A * ****** *******  The phylogenetic relationship of insect/crustacean species is indicated as a phylogram tree. AP-1, activator protein 1; CREB, cAMP response element-binding protein; C/EBP, CCAAT-enhancer-binding protein; MEF2, myocyte enhancer factor 2; NF-AT, nuclear factor of activated T-cells; SRF, serum response factor. See Table 2 for the details of genomic sequences used for promoter analysis. See Fig. 3-1 and Table 3-1 for the structural conservations of the transcription factors used for the binding site prediction. sequence motifs, a GAGA motif and a GC-rich motif, which are structurally related to the promoter elements associated with RNA polymerase II (Pol II) stalling in Drosophila (Hendrix et al., 2008), are conserved in the upstream sequences of egr-B homologs of most hemimetabolous insects (Fig. 3A, B). The GAGA motif (consensus: GRGAGGGRVGGAGAGS) is conserved in the upstream sequences of polyneopteran egr-B homologs positioned at ϳ80 bp upstream to the TSS of Gryllus egr-B, but lacking in those of egr-B homolog of the other taxa (i.e., Paraneoptera, Coleoptera, and crustacea). The GC-rich motif (consensus: GCGCSSSGGCGCGC) is conserved in the upstream sequences of egr-B homolog of basal insects positioned at ϳ30 bp downstream to the TSS of Gryllus egr-B. Next, we expanded our analysis to the distal promoter region by predicting potential cis-regulatory motifs (Fig.  3C, see also Fig. 3-1). Generally, the upstream regions of insect/crustacean egr-B homologs were highly enriched with potential binding sites for transcription factors activated by various intracellular signaling pathways (e.g., AP-1, C/EBP, CREB, SRF, and NF-AT), which are known to regulate activity-dependent gene expression in vertebrates (Sealy et al., 1997;Hogan et al., 2003;Benito et al., 2011;Nonaka et al., 2014). In addition, phylogenetic footprinting analysis of the upstream region of egr-B homologs of polyneopteran insects revealed two conserved cis-regulatory modules (CRMs; CRM -800 and CRM -400 in Fig. 4A), which consist of proximal, conserved potential binding sites for stimulus-regulated transcription factors (Fig. 4B). These data indicate that the ϳ1.6-kbp upstream genomic fragment of Gryllus egr-B contains cis-regulatory elements for stimulus-regulated transcription factors, some of which constitute the core promoter and CRMs conserved across insect/arthropod species.

Construction of the IEG promoter-driven transgenic reporter system
An IEG promoter-driven reporter system was constructed using the upstream genomic fragment of Gryllus egr-B (Fig. 5A, B). The IEG reporter cassette consisted of a nuclear-targeted destabilized EYFP (EYFPnls:PEST; Li et al., 1998) driven by a ϳ2.2-kb genomic DNA fragment containing the ϳ1.6-kbp upstream sequence and 5= UTR of Gryllus egr-B. Two repeats of a synthetic AU-rich element were inserted between the coding sequence of EYFPnls:PEST and the SV40 3= UTR to shorten mRNA half-life (Pham et al., 2008). The IEG reporter cassette along with a 3xP3-mCherry expression cassette was integrated into the cricket genome using the piggyBac transposon. Successful transgenic crickets were selected according to mCherry fluorescence in the compound eyes, and reporter expression in the brain was checked by EYFP immunoreactivity after PTX treatment. In the Drosophila brain, Masuyama et al. (2012) reported that the inducible expression of EGFP protein under the control of activity-dependent synthetic transcription factor, CaLexA, became detectable ϳ4 h after stimulation. In the present study, we collected brains of the IEG reporter line 6 h after PTX injection to detect activity-dependent expression of reporter protein. Three of 37 mCherry-expressing transgenic lines showed EYFP immunoreactivity in the brain after PTX injection. One transgenic line with a low background EYFP expression (line #19; Fig. 5C) was selected, and a homozygous transgenic line was established as an "IEG reporter line" for subsequent analyses. The IEG reporter line develops normally, and we did not observe any behavioral abnormality.
We tested whether the expression of IEG reporter mimics innate IEG expression in the brains of the IEG reporter line. First, we compared baseline expression levels of the reporter gene (EYFPnls:PEST) and Gryllus egr-B. Quantitative RT-PCR expression analysis revealed that, in the brains of naïve animals, EYFPnls:PEST showed a higher baseline expression level than Gryllus egr-B (Fig. 5D). In addition, Time course analysis of PTX-induced expression of EYFPnls:PEST and Gryllus egr-B revealed differences in the expression kinetics of the genes after PTX injection. The expression of EYFPnls:PEST reached a maximum 60 -90 min after PTX injection (ϳ1.8-fold upregulation relative to the naïve animals), and slowly decreased to near baseline by 120 min after injection (Fig. 5E). On the other hand, the expression of Gryllus egr-B reached a maximum 90 min after PTX injection (ϳ6.5-fold up-regulation relative to the naïve animals) and decreased to the baseline level by 120 min after PTX injection (Fig. 5E). Correlation analysis revealed that the expression level of the EYFPnls:PEST was strongly correlated with that of Gryllus egr-B in the brain of the IEG reporter line (Pearson's r ϭ 0.8269, p Ͻ 0.0001; Fig. 5F). With these data, we concluded that the neuronal activity-driven expression of the reporter gene mimics that of Gryllus egr-B in the brain of the IEG reporter line.

IEG promoter-driven reporter system can be used for whole-brain activity mapping
First, we asked whether our reporter system can be used for whole-brain activity mapping. The brains of the IEG reporter line were collected 6 h after PTX or vehicle injection, and the distribution of the reporter protein (EYFPnls:PEST) was examined by whole-mount fluorescent immunohistochemistry (Fig. 6). In the brain of vehicle-injected animals (n ϭ 7), intense EYFP immunoreactivity was only detected in cell clusters located in the lateral parts of the supraesophageal ganglion (white arrowheads in Fig. 6A, C). The distribution of the EYFPnls: PEST protein in the brain of naïve animals was similar to that in the vehicle-injected animals (data not shown). Six hours after PTX injection, EYFP immunoreactivity was detected throughout the ganglia (Fig. 6B, D; n ϭ 6) and was restricted to the nucleus. In addition, we did not observe EYFP immunoreactivity in the marginal glia (the cells located at the margins of the neuropil) or the epithelial glia after PTX injection, suggesting that the reporter protein is only induced in the neurons (Movies 1 and 2). To examine the expression time course of the reporter protein, we conducted Western blot analysis using several anti-GFP antibodies. However, we could not detect bands corresponding to EYFPnls:PEST before and after PTX injection. We speculated that restricted localization of the reporter protein in the nuclei might facilitate detection of the reporter protein in immunohistochemistry.
Next, we tested whether our reporting system can label behaviorally evoked neural activation under the physiological conditions. As our RT-qPCR analysis revealed that feeding of sucrose solution induced expression of neuronal IEGs in the cricket brain, we examined the expression of the reporter protein induced by feeding of sucrose solution in the brain of the IEG reporter line. In the insect brain, the octopaminergic/tyraminergic neurons are activated by feeding of sucrose solution or presentation of "reward" in the context of learning and memory studies. In the honeybee, an identified octopaminergic neuron, VUMmx1, in the subesophageal ganglion responds to sucrose solution applied to the antenna and proboscis (Hammer, 1993). In Drosophila, the octopaminergic/tyraminergic neurons in the subesophageal ganglion control larval appetite     Figure 4. Phylogenetic footprinting revealed conserved cis-regulatory modules in the upstream regions of polyneopteran egr-B homologs. A, mVISTA plot of the upstream regions of polyneopteran egr-B homologs based on MLAGAN alignment using the upstream region of Gryllus egr-B as a reference sequence. Positions of potential transcription factor binding sites in the upstream region of Gryllus egr-B are indicated by arrowheads (Fig. 3B). The horizontal and vertical axes of the plot represent the position in the sequences and the percentage identity, respectively. Two conserved cis-regulatory modules (CRMs; CRM -800 and CRM -400 ) and the conserved core promoter region are shaded blue and red on the plot, respectively. B, Nucleotide sequence alignments of two conserved CRMs (CRM -800 and CRM -400 ) found in the upstream region of polyneopteran egr-B homologs. The conserved bases are marked with asterisks under the alignment. Cis-regulatory elements conserved among most of the sequences are indicated above the alignment. Black bars under the alignments indicate sequence motifs conserved across species where no transcription factor is assigned. AP-1, binding site for activator protein 1; AP-4, binding site for activating enhancer binding protein 4; ATF2, binding site for activating transcription factor 2; CDP/Cut, binding site for CCAAT-displacement protein/cut homeobox; C/EBP, binding site for C/EBP; CRE, cAMP-responsive element; SRE, serum response element. See Table 2 for the details of genomic sequences used for promoter analysis.  Selcho et al., 2014). Furthermore, in the cricket, the pharmacological blockade of ␣-adrenergic-like octopamine receptor impaired appetitive learning, indicating that the octopaminergic system is involved in reward signaling (Unoki et al., 2005(Unoki et al., , 2006Matsumoto et al., 2015). These allow us to hypothesize that the feeding of sucrose solution would activate the octopaminergic/tyraminergic system in the subesophageal ganglion of the cricket. According to this idea, we asked whether the feeding of sucrose solution induces the reporter protein expression in the octopaminergic/tyraminergic neurons in the subesophageal ganglion, especially focused on the dorsal unpaired median (DUM) neurons (Pflüger and Stevenson, 2005).
To label the octopaminergic/tyraminergic neurons in the cricket brain, we first generated anti-Gryllus tyrosine decarboxylase 2 (Tdc2) antibody. In Drosophila, Tdc2 is responsible for the production of tyramine (a precursor of octopamine) in the nervous tissues (Cole et al., 2005;see Fig. 7-1), and the tdc2-Gal4 driver is used for targeted gene expression in the octopaminergic/tyraminergic neurons in Drosophila. Our anti-Gryllus Tdc2 antibody successfully immunolabeled the octopaminergic/tyraminergic DUM neurons in the subesophageal ganglion of the Hokudai WT strain (Fig. 7A, B; see also Fig. 7-2), which is confirmed by double immunostaining with the anti-octopamine antibody (Fig. 7C). Then, we examined whether reporter protein expression becomes detectable in the nuclei of the DUM neurons in response to the feeding of sucrose solution. In the subesophageal ganglia of the naïve IEG reporter line, the DUM neurons did not show nuclear-localized staining of EYFP immunoreactivity (n ϭ 4; Fig. 7D, upper columns). As we expected, 6 h after feeding of sucrose solution, the nuclear-localized reporter protein expression was detected in several neurons in the DUM1 and DUM2 clusters (n ϭ 4; Fig. 7D, lower columns). These data demonstrate that our IEG promoter-driven reporter system can visualize neuronal activation caused by continued (baseline expression level). Box plots indicate the 25th to 75th percentile ranges and central values. Error bars indicate the 5th to 95th percentile ranges. The "ϩ" denotes the mean. Asterisks donate statistical significance to the control (naïve animals; ‫,ء‬ p Ͻ 0.05; ‫,ءء‬ p Ͻ 0.01; ‫,ءءء‬ p Ͻ 0.001; ‫,ءءءء‬ p Ͻ 0.0001). F, Correlation plot between the expression levels of EYFPnls:PEST and Gryllus egr-B in the brains of the IEG reporter line. The data from PTX-injected crickets (n ϭ 75; black circles), vehicle pre-injected crickets (n ϭ 8; gray circles), and naïve crickets (n ϭ 8; white circles) were plotted. See Table 3 for the details of statistical analysis. PTX-induced neuronal hyperexcitability, as well as behaviorally evoked neuronal activation.

Discussion
In the present study, we established an IEG promoterdriven transgenic reporter system for whole-brain activity mapping in the cricket. In the cricket brain, the neuronal IEGs, Gryllus fra-B, egr-B, and hr38, were up-regulated within 1 h after PTX-induced neuronal hyperexcitability, and after feeding of sucrose solution and agonistic interaction. We found that the region ϳ1.6 kbp upstream of Gryllus egr-B is sufficient to drive reporter gene expression in response to neuronal hyperexcitability. Inducible reporter protein expression was broadly detected in the brain of the IEG reporter line. Finally, we detected feedingevoked neuronal activations in the cricket brain with our reporter system. Our study also revealed that the core promoter region of egr-B homologs show a high level of structural conservation in basal insects and crustaceans.
The egr-1/zif268/NGFI-A gene was first identified as an IEG induced by a variety of extracellular stimuli (e.g., serum, mitogens, and growth factors) in various cell types (Sukhatme et al., 1987;Milbrandt, 1987), and was later shown to be induced after depolarization in neurons (Sukhatme et al., 1988). Then, it became widely used as a marker for neuronal activation in various vertebrate species (Brennan et al., 1992(Brennan et al., , 1999Mello and Clayton, 1994;Hoke et al., 2004;Burmeister and Fernald, 2005). In the cricket brain, Gryllus egr-B was rapidly induced as an IEG in response to neuronal activation. Neuronal activity-regulated expression of egr homologs was also reported in the brain of Drosophila (Guan et al., 2005; and the honeybee (Ugajin et al., 2013). These data support the idea that the egr-1/zif268/NGFI-A homologs are evolutionarily conserved neuronal IEGs between vertebrates and insects.
In the present study, using the promoter region of Gryllus egr-B and the nuclear-targeted destabilized EYFP as a reporter, we established a retroactive, whole-mount, single-cell-resolution activity mapping system to highlight behaviorally relevant neuronal circuits in the cricket brain. Transgenic reporter systems for neuronal activity mapping have been constructed in the mouse and Drosophila by using innate activity-dependent promoters (Smeyne et al., 1992;Wang et al., 2006;Reijmers et al., 2007;Barth, 2007;, as well as synthetic promoters (Kawashima et al., 2013;Fujita et al., 2013;Sørensen et al., 2016) or a synthetic activity-regulated transcription factor (Masuyama et al., 2012). In rodents, the activity-dependent promoters were used to express the channelrhodopsin gene in the behaviorally relevant neurons to investigate the formation of memory engrams (Garner et al., 2012;Ramirez et al., 2013). In our future research, we will combine our IEG promoterbased gene expression system with optogenetic/chemogenetic tools or genetically encoded calcium indicators to uncover the key neural circuits involving cricket social behaviors. Moreover, recent advances in genetic engineering techniques (transposon-based germline transgenesis and CRISPR/Cas-based knock-in; O'Brochta and Atkinson, 1996;Atkinson et al., 2001;Kimura, 2001;Handler, 2002;Sun et al., 2017) allow us to introduce the IEG promoter-based reporter system into non-Drosophila insects classically used in insect neuroethology.
When neuronal IEG mRNAs are induced in the vertebrate nervous system, expression reaches its peak around 30 min after stimulation, and then rapidly returns to baseline by 60 min (Zangenehpour and Chaudhuri, 2002, Mello and Clayton, 1994, Burmeister et al., 2005. On the contrary, in the cricket brain, the expression levels of neuronal IEGs peaked more slowly (60 -90 min after PTX injection; Figs. 2, 1-5, and 2-1). This is consistent with previous studies in other insect species Ugajin et al., 2012Ugajin et al., , 2016Ugajin et al., , 2018Fujita et al., 2013). These data indicate that, in general, mRNAs of insect neuronal IEGs show slower expression kinetics than those in the vertebrate systems. The expression dynamics of inducible gene mRNA is determined depending on the rates of transcription and mRNA decay (Shalem et al., 2008). In the cricket brain, transcription of Gryllus egr-B was rapidly initiated on stimulation, and promoter activity was sustained at a high level for Ͼ1 h after stimulation. In the vertebrate system, neuronal IEGs are categorized into two subgroups: rapid and delayed IEGs (Saha and Dudek, 2013). The expression of delayed IEGs requires several time-consuming steps to recruit Pol II to the promoter for the initiation of transcription (e.g., signal transduction, posttranslational modifications of transcription factors, and chromatin remodeling), whereas transcription of rapid IEGs is initiated by stalled Pol II. In insects, Pol II stallingdependent rapid transcription initiation was found in Drosophila heat shock response genes (Teves and Henikoff, 2013), where inducible gene expression becomes detectable within 5 min after stimulation (Vazquez et al., 1993). Rapid initiation of Gryllus egr-B transcription suggests that this gene is expressed as a rapid IEG in the cricket brain. This is also supported by the existence of promoter elements for Pol II stalling in the core promoter region of Gryllus egr-B, which is discussed in detail in the following section. The level of Gryllus egr-B mRNA is decreased when its promoter activity is down-regulated, indicating that Gryllus egr-B mRNA undergoes rapid degradation. Our expression analysis also revealed that another neuronal IEG, Gryllus hr38, has a slower induction kinetics with exceptionally long mRNA half-life when compared to the case of Gryllus egr-B ( Fig. 1-5), suggesting that Gryllus hr38 is expressed as a delayed IEG in the cricket brain. As mentioned above, the core promoter region of Gryllus egr-B contained promoter elements associated with Pol II stalling, which are highly conserved in the upstream region of basal insects/crustacean egr-B homologs. Pol II stalling was first reported on Drosophila heat shock genes (Gilmour and Lis, 1986). Now, Pol II stalling is considered as a widespread mechanism of transcription regulation for rapid and precise control of gene expression in metazoans (Core and Lis, 2008;Nechaev et al., 2010;Levine, 2011). Genome-wide survey in Drosophila (Hendrix et al., 2008) revealed that ϳ20% of stalled promoters contain an arrangement of core promoter elements, namely, GAGA motif, Inr, GC-rich motif, and/or DPE. Our promoter analysis revealed that all of the core promoter elements found in Drosophila stalled promoters are contained in the core promoter region of Gryllus egr-B and basal insect egr-B homologs. The GAGA motif is found in the Drosophila hsp70 promoter (Shopland et al., 1995;Wilkins and Lis, 1997) and other stalled promoters (Lee et al., 2008;Hendrix et al., 2008). In Drosophila, a GAGA transcription factor, GAF, occupies the GAGA motif in the stalled promoter and recruits transcription coregulators, such as the nucleosome remodeling factor and the negative elongation factor, to the core promoter region (Fuda et al., 2015;Tsai et al., 2016). In Drosophila, the Inr and DPE are enriched by ϳ2-fold in the stalled promoters relative to the core promoter region of constitutively expressed genes, respectively (Hendrix et al., 2008). These elements are highly conserved in the core promoter of basal insect egr-B homologs, as are some sequence homologies of Drosophila stalled promoters (Hendrix et al., 2008). In the present study, we found a GC-rich motif between the TSS and DPE of the core promoter region of basal insect egr-B homologs. In Drosophila embryo, ϳ25% of stalled promoters contain a GC-rich region between ϩ1 and ϩ50, which could promote Pol II stalling by attenuating the movement of the Pol II complex (Hendrix et al., 2008). To our surprise, the putative core promoter region of Daphnia egr-B homolog shares the similar arrangement of cisregulatory and core promoter elements found in the upstream sequence of basal insect egr-B homologs (Fig.  3A). This finding indicates that ancestral arthropods acquired a prototypic egr-B core promoter before the diversification of insects from crustaceans (ϳ500 million years ago; Misof et al., 2014). To our knowledge, the core promoter region of insect/crustacean egr-B homologs is the most ancestral structurally conserved activityregulated promoter.
In addition to the structural analysis of the core promoter, in the present study we investigated the spatial distribution of cis-regulatory elements in the upstream region of insect egr-B homologs. In vertebrates, structural and functional analyses of activity-regulated promoters/ enhancers revealed transcription factor binding sites essential for inducible gene expression (West et al., 2002;Benito and Barco, 2015). For example, Kawashima et al.
(2009) identified a synaptic activity-responsive CRM at ϳ7 kb upstream of the TSS of the mouse Arc gene, which consists of closely localized binding sites for CREB, MEF2, and SRF. The promoter region of the c-fos gene contains binding sites for SRF and CREB, which play crucial roles in activity-regulated gene expression (Ghosh et al., 1994). The promoter region of the down syndrome critical region isoform 4 gene contains several binding sites for NF-AT, which are required for depolarizationdependent expression of the gene (Vihma et al., 2016). Although biochemical properties of stimulus-regulated transcription factors have not been characterized in insects, our exploratory in silico analysis showed structural conservation in the DNA-binding domains of stimulusregulated transcription factors among insects and vertebrates. This suggests that insect stimulus-regulated transcription factors share DNA-binding properties similar to their vertebrate counterparts (Table 3-1 and Fig. 3-1). Structural analyses of the promoter region of egr-B homologs first revealed that the putative core promoter region of basal insect/crustacean egr-B homologs contains two conserved CREs and one SRF. Also, potential binding sites for stimulus-regulated transcription factors are enriched in the distal upstream regions of insect/ crustacean egr-B homologs (Fig. 3C). Comparative structural analysis of the upstream sequences of polyneopteran egr-B homologs revealed two conserved CRMs, which contain binding sites for stimulus-regulated transcription factors, such as AP-1, CREB, SRF, and C/EBP. CRMs integrate the multiple transcription factors and their associated cofactors at a specific timing to perform elaborate and accurate regulation (Jeziorska et al., 2009). In fact, when we stimulated an individual intracellular signaling pathway, the expression of Gryllus egr-B was not induced (Fig. 2-2). Although the reason is unclear, it is possible that synergic activation of multiple signaling pathways and their downstream transcription factors is necessary to induce expression of Gryllus egr-B in the cricket brain. Further studies are necessary to understand the upstream signal transduction pathways involved in transcriptional activation of insect egr-B homologs.
In the vertebrate nervous system, egr family genes are involved in long-term plasticity and the formation and consolidation of long-term memory (Poirier et al., 2008). Genome-wide analysis for the direct target of Egr-1 protein in PC12 neurons revealed that a majority of Egr-1 target genes are down-regulated when egr-1 is overexpressed, suggesting that the Egr-1 protein principally functions as a transcription suppressor (James et al., 2005). However, other studies revealed that the activityinduced Egr-1 and Egr-3 proteins up-regulate the expression of GABA A receptor subunits in the hippocampal neurons to regulate homeostatic excitatory/inhibitory balance (Roberts et al., 2005;Mo et al., 2015). These controversial findings suggest that the target gene and transcriptional regulation of the Egr family proteins are determined according to cell type and/or in genedependent manners. In the insect nervous system, the function of the egr homolog, as well as its downstream target genes, is poorly understood. Further studies are needed to address the biological significance of the activity-regulated expression of the egr homolog in the insect brain.