Generation of LexA enhancer-trap lines in Drosophila by an international scholastic network

Abstract Conditional gene regulation in Drosophila through binary expression systems like the LexA-LexAop system provides a superb tool for investigating gene and tissue function. To increase the availability of defined LexA enhancer trap insertions, we present molecular, genetic, and tissue expression studies of 301 novel Stan-X LexA enhancer traps derived from mobilization of the index SX4 line. This includes insertions into distinct loci on the X, II, and III chromosomes that were not previously associated with enhancer traps or targeted LexA constructs, an insertion into ptc, and seventeen insertions into natural transposons. A subset of enhancer traps was expressed in CNS neurons known to produce and secrete insulin, an essential regulator of growth, development, and metabolism. Fly lines described here were generated and characterized through studies by students and teachers in an international network of genetics classes at public, independent high schools, and universities serving a diversity of students, including those underrepresented in science. Thus, a unique partnership between secondary schools and university-based programs has produced and characterized novel resources in Drosophila, establishing instructional paradigms devoted to unscripted experimental science.


Introduction
Conditional gene expression systems in Drosophila provide a powerful basis for investigating the function and regulation of genes and cells.Generation of a GAL4-based transactivator to induce expression of target genes fused to upstream activating sequences (UAS) is a widely used binary expression system in Drosophila (Brand and Perrimon 1993;Hayashi et al. 2002;Gohl et al. 2011).Random insertions by transposons encoding GAL4 into the genome ("enhancer trapping"; O'Kane and Gehring 1987) generate strains with endogenous enhancer-directed GAL4 expression.In these enhancer trapping constructs, a weak promoter "reads" the local enhancer landscape and directs the expression of Gal4 according to this regulatory information (Sepp and Auld 1999).Studies of many biological problems benefit from simultaneous manipulation of two or more independent cell populations or genes (reviewed in Rajan and Perrimon (2012) Kim et al. (2021)).In prior studies, parallel use of two binary expression systems allowed important new biological insights, including clonal and lineage analysis (Lai and Lee 2006;Bosch et al. 2015), "tissue epistasis" studies (Yagi et al. 2010;Shim et al. 2013), and discovery of specific cell-cell interactions and contacts (Gordon and Scott 2009;Bosch et al. 2015;Macpherson et al. 2015).These approaches used a second expression system that functions independently of the UAS-Gal4 system, such as the LexA system derived from a bacterial DNA-binding domain (Szüts and Bienz 2000;Lai and Lee 2006;Pfeiffer et al. 2010;Gnerer et al. 2015;Knapp et al. 2015).The fusion of the LexA DNA-binding domain to a transactivator domain generates a protein that regulates expression of transgenes linked to a LexA operator-promoter (LexAop).However, the number and quality of lines expressing a LexA transactivator remain small, compared to the thousands of comparable GAL4-based lines.
To address this resource gap, we previously developed a network of partnerships between a research university (Stanford) and US secondary schools to generate novel LexA-based enhancer trap drivers, in an outreach we called "Stan-X" (Kockel et al. 2016(Kockel et al. , 2019, https://www.stan-x.org/).In these high schools, the enhancer trap and molecular biology experiments whose results are outlined and documented below are integrated into the class schedules as an advanced biology course (colloquially referred to as a "Stan-X course").This class was taught by instructors during the regular school year, with students receiving educational credit.
Here, we describe a significant expansion of this earlier effort into an international scholastic network including Stanford University, and science classes at seventeen independent and public secondary schools and universities in the United States and United Kingdom.To increase our capacity for training the participating high school teachers, we established a teacher training academy, called "Discovery Now".Over 2 weeks each summer, incoming and participating teachers are instructed in underlying principles of genetics and molecular biology, followed by weekly meetings during the subsequent school year.This expanded network of participating high schools successfully produced hundreds of novel LexA-based enhancer-trap lines for the community of science, advancing science instruction paradigms rooted in experimental genetics, molecular, and cell biology.
The annotated primary DNA sequence of SX4 enhancer trap P-element is presented in Supplementary Data File 1.

Construction of SX4 starter strains
The transformation of the P{w[+mC] = LHG]Stan-X[SX4]} P-element vector into the w 1118 fly strain was performed using standard procedures.The SX4 X-linked index transformant was isogenized to the Stan-X background to generate the w 1118 , SX4; iso#32 II ; iso#32 III .SX4 is located at X:19,887,269 in the amnesiac locus (Supplementary Table 1).We noted an Invader natural transposable element (TE) insertion 123 bp upstream of Stan-X[SX4] that is not represented in FlyBase rs6 of the genome, and might be specific to the w 1118 , SX4; iso#32 II ; iso#32 III isogenized background used (see "Fly husbandry and isogenized fly strains" below).
The SX4 X-linked insertion was then mobilized using standard procedures (see below) to the third chromosome balancer TM6B, to create the w 1118 ; TM6B,SX4 orig /ftz,e starter stain for mobilization to the X chromosome (see "P-element mobilization").The SX4 P-element on TM6B is located at 3L:3,250,470 in the gene encoding lncRNA:CR43626.

L3 dissection and immunohistochemistry (IHC)
Drosophila larva transitions from larval stage L1 to larval stage L3 by intermittent molting.Prior to pupariation, L3 larvae stop feeding and migrate to a pupariation site on the side of a vial.Wandering L3 larvae were bisected and inverted in PBS, and all tissues were fixed in 4% formaldehyde/PBS for 30 min, permeabelized in 0.2% Triton X-100/PBS for 4 hours, and blocked in 3% BSA/PBS for 1 hour.All antibody stainings were performed in 3% BSA/PBS, incubation of primary and secondary antibodies were O/N.PBS was used for all rinses and washes (3× each for primary and secondary antibody incubation steps).Antibodies used were as follows: Chicken anti-RFP 1:2,000 (Rockland,; Goat anti-GFP 1:3,000 (Rockland,; Donkey anti-Goat Alexa Fluor 488 (Life Technologies, A11055); Donkey anti-Chicken Cy3 (Jackson ImmunoResearch,; and Donkey anti-Mouse Alexa Fluor 594 (Life Technologies, A21203).All secondary antibodies were used at 1:500.Tissues were dissected off the cuticle and were mounted in SlowFade Gold mounting medium with DAPI (Life Technologies, S36938).See Kockel et al. (2016) for a detailed protocol.

Epifluorescent microscopy
Microscopy was performed on a Zeiss AxioImager M2 with Zeiss filter sets 49 (DAPI) and 38HE (Alexa Fluor 488) using the extended focus function.Used compound epifluorescent microscopes for high schools with all required lenses, installation services, and optional training sessions are available for sale from MicoOptics (https://www.micro-optics.com/).
The viable X chromosome w 1118 , iso#11 X was isolated by crossing single males w 1118 to y 1 w a FM7c females, and backcrossing single female offspring to corresponding F 0 males.Derived strains were scored for loss of FM7c.The viable and fertile strain w 1118 , iso#11 X was combined with marked chromosomes on II and III to yield w 1118 ,iso#11 X ; L*/CyO; ftz*,e*/TM6,Hu 1 ,Tb 1 .This strain was used to isogenize the X chromosome of the transposase source Δ2-3, and to balance novel insertions of SX4 on II and III (see below).

P-element mobilization from autosome III to X chromosome
The mobilization of a SX4 element located on a third chromosome balancer, TM6B,SX4[orig] to the X chromosome was performed as a pilot experiment in a non-isogenized, mixed background.

Insertion site cloning
We applied an inverse PCR (iPCR) approach (Kockel et al. 2019), to molecularly clone the insertion sites of Stan-X SX4 P-elements.DNA restriction enzymes used are as follows: Sau3AI (NEB R0169) and HpaII (NEB R0171); ligase used: T4 DNA Ligase (NEB M0202); 5′ end cloning: inverse PCR primer "Plac1" CAC CCA AGG CTC TGC TCC CAC AAT and "Plac4" ACT GTG CGT TAG GTC CTG TTC ATT GTT; sequencing primer 5′ end: "SP1" ACA CAA CCT TTC CTC TCA ACA; 3′ end cloning: primer pair "Anna" CGC AAA GCT AAT TCA TGC AGC and "SP1Berta" ACA CAA CCT TTC CTC TCA ACA AAA GTC GAT GTC TCT TGC CGA; and sequencing primer 3′ end: "SP1" ACA CAA CCT TTC CTC TCA ACA.For insertions where the sequence of one end only could be determined by iPCR, we pursued a gene-specific PCR approach (Ballinger and Benzer 1989) using P-element and gene-specific primers.The 5′ end specific P-element primer "Chris" is: GCA CAC AAC CTT TCC TCT CAA C, sequencing primer 5′ end: "Sp1"; 3′ end specific P-element primer "Dove": CCA CGG ACA TGC TAA GGG TTA A, sequencing primer 3′ end: "Dove"; and sequence of gene-specific primers is available upon request.The position and identity of natural TEs, and the insertion of the SX4 element within, were determined by iPCR (Supplementary Table 1) and confirmed with the genome sequence of the host genome w 1118 , SX4; iso32 [II] ; iso#32 [III]  by TE Mapper (Supplementary Table 2).

Generation of sequence logos and position frequency matrices
The construction of the SX4 sequence logo was executed as described (Crooks et al. 2004;Kockel et al. 2019) using http:// weblogo.threeplusone.com/.The input sequence motif data is listed in Supplementary Table 1.The 8 bp genomic insertion site sequence is codirectional to the P-element's direction of insertion (Linheiro and Bergman 2008;Kockel et al. 2019).If P-elements are inserted 5′->3′, the strand of insertion was named + (plus), and unprocessed genomic scaffold sequences as present in FlyBase were used to extract the insertion site sequences.If P-elements are inserted 3′->5′, the strand of insertion is termed − (minus), and the reverse complement of the genomic scaffold sequences was used to extract these insertion site sequences.

Genome sequence data processing and analysis
We used BWA, SAMtools, and freebayes to perform variant calling.Details of the pipeline, along with specific parameters used, are provided in the StanX_tools repository (https://github.com/sanath-2024/StanX_tools).
To use our short-read dataset to find novel, non-reference transposons (Fig. 6 and Supplementary Table 2), we deploy a similar a strategy as Linheiro and Bergman (2012).We used BWA to find reads that align to both, a canonical transposon sequence as well as the FlyBase reference genome.These "split reads" were processed and sorted into groups based on alignment location and orientation.Details are provided in the StanX_tools repository (https://github.com/sanath-2024/StanX_tools).Our TE mapper represents ground-up multithreaded reimplementation in the Rust language, focusing on performance and simplicity.

Analysis of Fly Cell Atlas IPC and CC cell data
Insulin-producing cell (IPC) and corpora cardiaca (CC) cell nuclei isolation from males and females was conducted in the framework of the Fly Cell Atlas (FCA, Li et al. 2022, https://www.ebi.ac.uk/biostudies/files/E-MTAB-10628/E-MTAB-10628.sdrf.txt).FASTQ sequencing files were aligned to BDGP6 version of the fly whole genome using HISAT2 (Kim et al. 2019).Single cell nuclei RNAseq libraries representing IPCs and CC cells were filtered based on dilp2, dilp3, dilp5, and akh expression, respectively.The location of natural transposons (nTEs) and gene locations in the BDGP6 genome were taken from FlyBase.featureCounts (Liao et al. 2019) was used to assign aligned reads to transposons or genes and to obtain a count matrix for each library.When quantifying counts for nTEs, multi-mapped reads were assigned their full value to each alignment, which gives a theoretical upper bound for how much transcript could exist for a single nTE.When quantifying counts for classes of nTEs or for all TE expression (Supplementary Fig. 2C), multi-mapped reads were assigned a value of 1/x to each alignment, where x is the number of alignments, which estimates the total amount of reads associated with the class of TE or the total number of reads coming from TEs. Count matrices were used as input to Seurat (Hao et al. 2021).Seurat VlnPlot function was used to plot unnormalized counts for gene expression (Supplementary Fig. 2).

Training of Stan-X teachers at the Discover Now Teacher Academy
For incoming Stan-X teachers, the Stan-X Biology Course covering P-element mobilization ("Module 1"), insertion site sequencing by iPCR ("Module 2"), and expression analysis in third instar larvae ("Module 3") was offered as a 2-week training course consisting of a 1-week online (∼3 hours/day) session, followed by a 1 week session of in-person training (8 hours/day) at the Lawrenceville School, NJ, or Stanford University School of Medicine.The 2-week class was offered each year in the summer or winter.The course was staffed by instructors from participating high schools and Stanford University School of Medicine.Application deadlines and other information are detailed online at https:// www.stan-x.org/.Secondary school students spent 9-10 weeks executing the P-element mobilization crosses, mapping and balancing their novel SX4 strains.This was followed by 2-3 weeks for molecular determination of the SX4 insertion site, using iPCR and DNA sequencing using spin column-based genomic DNA preparation.The last weeks of classes are reserved for crosses with reporter strains (w; LexAop2-CD8::GFP), allowing for training in L3 larval dissection and epifluorescent microscopy to describe tissue specific expression patterns of novel SX4 enhancer traps.

High school coursework
Based on performance and recommendation of Stan-X teachers, one to three students were invited to continue studies at Stanford University School of Medicine during summer internships lasting from 2-6 weeks.These studies included further molecular mapping of transposon insertion sites and verification of tissue patterns of enhancer trap expression.Students returning in the fall term helped instructors to run the subsequent iteration of the Stan-X class, and also pursue independent projects.

Generation of starter fly lines for LexA enhancer trap screening
While prior studies mobilizing the X-linked SE1 element successfully generated LexA enhancer trap flies (Kockel et al. 2016(Kockel et al. , 2019)), novel autosomal insertions were recovered at relatively low frequency (<5%) and showed modest expression of LexA.Of note, patterned expression in wing imaginal discs was not recovered, and expression in other tissues was often variegated.To address these limitations, we modified the SE1 element (Methods) to generate the SX4 element (Fig. 1, a and b) and SX4 "starter" fly line.The SX4 P-element carries a LexA::G4 fusion (L = LexA DNA-binding domain; H = Gal4 hinge region; G = Gal4 transcriptional activation domain; together called "LHG", and referred to as "lexA") identical to the SE1 P-element (Kockel et al. 2016).However, the P-element promoter driving lexA in the SE1 element was replaced by the hsp70 promoter.Thus, compared to the original SE1 transposon (Fig. 1a), the SX4 element has multiple modifications (Supplementary Data File 1 and Fig. 1) including (1) removal of attB sequences, (2) replacement of the original P-element promoter with the hsp70 promoter to regulate LexA expression, and (3) placement in the amnesiac locus (amn) at X:19,887,268, a region with 81 reported, independent transgenic insertions, suggesting permissiveness for P-element transposition.
After transformation of the P{w[mC] = LHGStan-X[SX4]} P-element vector, hereafter referred to as "SX4", into the w 1118 recipient strain (Methods), an index SX4 X-linked transformant was isogenized to the iso 113232 genetic background to generate the w 1118 , SX4; iso#32 II , iso#32 III strain (Methods).Prior studies using the SE1 element in a less-defined genetic background observed insertional bias of the SE1 element to a genomic region containing a KP element, a contaminating nonautonomous P-element derivative with an internal deletion encoding a repressor of transposition (Lee et al. 1996, Kockel et al. 2019).Thus, we used whole genome sequencing (Methods) with 76× and 80× coverage for males and females (PRJNA912892), respectively, to confirm the absence of KP elements in the w 1118 , SX4; iso#32 II ; iso#32 III strain.Analysis of 8 bp direct repeat sequences from individual SX4 insertions (n = 281) shows a slight preference of SX4 towards weak palindromic sites (Fig. 1d), as previously reported (Kockel et al. 2019;Linheiro et al. 2012).

Generating novel LexA enhancer-trap lines
To generate LexA-based enhancer trap fly lines, we mobilized the X-linked SX4 P-element to isogenic autosomes iso#32 II ; iso#32 III or the third chromosome SX4 insertion TM6BSX4 orig to the X chromosome (Methods).To facilitate the mobilization of the nonautonomous SX4 P-element, the SX4 donor strain was crossed into a genetic background expressing the activated P-element transposase variant Δ2-3 (Robertson et al. 1988).The transposase source E. S. Kim et al. | 5 was subsequently removed through outcross in the next generation to stabilize the transposition event, generating LexA P-element enhancer-trap lines (Methods; Supplementary Table 1; O'Kane and Gehring 1987).These novel enhancer traps report interactions of its relatively weak hsp70 promoter with the local enhancer environment of the insertion site by LexA expression, displaying spatial and temporal expression specificity (O'Kane and Gehring 1987).We used a similar experimental logic (Methods) to mobilize the SX4 located on III (TM6B,SX4 orig ), and isolate LexA insertions in the X-chromosome (Supplementary Table 1).

Characterization of Stan-X P-element insertion sites
We next used iPCR-based molecular methods to map the chromosomal insertion position of the Stan-X P-elements to the molecular coordinates of the genomic scaffold (Fig. 2 and Supplementary Table 1).The 301 novel insertions of this study were distributed across autosomes II and III, and their chromosomal arms (2L, 70 insertions; 2R, 69 insertions; 3L, 67 insertions; 3R, 79 insertions).We also isolated 11 insertions on the X-chromosome in a pilot P-element mobilization screen using the TM6B,SX4 orig as a starter line.At 19 loci, multiple P-element insertions (ranging from 2-4) mapped within 1 kb of previously derived lines, supporting the unique value of the individual LexA enhancer traps within the collection.In summary, we identified insertions at 268 unique loci, including one intergenic region (Supplementary Table 1); and all lines were submitted to a fly stock repository (Bloomington, IN).
Natural transposons (natural TEs) constitute a significant portion of repetitive DNA in the genome and represent ∼6% of sequenced euchromatin (Kapitonov and Jurka 2003;Kaminker et al. 2002).For the first time, we report 17 out of 301 (5.6%) SX4 lexA enhancer trap insertions into natural transposons, conforming with the expected frequency.This allows ingress to the somatic spatio-temporal expression pattern, if present at all, associated with the repetitive elements at these loci (see below).All tagged natural transposons are present in multiple copies throughout the genome, and represent repetitive DNA.Out of 17 SX4 enhancer trap insertions into natural TEs, 11 were unambiguously mapped to a specific site within a single copy of a natural TE.Of these 11 natural TEs tagged by SX4, seven are present in re-lease6 of the Drosophila genome (R6, https://flybase.org/;1360{} 1206, Invader1{}757, Opus{}1033, Juan{}1190, F{}1209, mdg3{}1215, and Invader4{}1371).Four of 11 natural TEs tagged by SX4 are not represented in R6 of FlyBase and are specific to our iso 113232 background (2× 1360, Copia, HMS Beagle, see below, Methods), illuminating differences between the genomes of iso 113232 and the FlyBase reference strain iso 1 .Four independent SX4 insertions mapped to a cluster of natural TEs (Juan, F, 1360, mdg3: Supplementary Table 1) on the pericentromeric-euchromatin boundary at 4.4-5.0Mbp of chromosomal arm 3R.Six of 17 SX4 insertions into natural TEs could be assigned to a specific clade of natural TEs (insertions into Doc, 2× Opus, 1731, 1360, Rt1a), but could not be unambiguously mapped to a specific location within the Drosophila genome due to sequence read limitations during insertion site cloning by iPCR (Supplementary Table 1).
Of the total of 301 novel lexA insertions presented here, the majority (295) were unambiguously mapped by DNA sequencing (295/301, 98%) and integrated into gene elements, including promoters, and the known first exon or intron of transcription units, similar to results from prior work (Bellen et al. 2011).Of the 295 mapped insertions presented here, we observed an even distribution of insertional direction by the SX4 P-element into genomic DNA.Using the 5′ and 3′ ends of the SX4 P-element as coordinates, we found that 149/301 insertions were oriented 5′ to 3′, and 146/ 301 insertions were oriented 3′ to 5′.In six cases, we were unable to determine the direction of P-element insertion.As detailed above, 17/301 (5.6%) insertions of SX4 were in a natural TE (Supplementary Table 1).
The analysis of LexA lines already present within ±1 kb of SX4 insertions sites revealed two loci at which three LexA enhancer traps were previously generated; escargot (esg) and SNF4γ, which Fig. 2. Map of novel Stan-X lexA enhancer trap insertions across chromosomal arms of X, II, and III.Chromosome arms are drawn to scale, and the enhancer trap positions are designated by their molecular coordinates.Scale below is in megabases (Mb).P-element insertions indicated below the chromosomal scaffold are oriented 3′ to 5′, and insertions above are oriented 5′ to 3′ relative to the reference sequence release 6 in FlyBase.Multiple insertions of identical orientation near identical genes are unified as single entry, and are separately listed in Supplementary Table 1.Insertions into natural transposable elements (natural TEs) are indicated in blue.
encodes the AMPK subunit gamma.These loci are known hotspots for P-element insertion, and the current study identifies three additional independent insertions into esg, all in the promoter region of that gene.Three SX4 insertions integrated in loci previously tagged twice with LexA insertions (CG33298, α-Est10, lncRNA:CR43626), and thirty-seven SX4 insertions of this study mapped within ±1 kb of the insertion site for one prior LexA enhancer trap (Supplementary Table 1).In summary, our approach generated multiple novel LexA-based autosome and sex chromosome enhancer traps.

Tissue expression patterns of LexA from SX4 insertions
Larval growth from stage 1 (L1) to stage 3 (L3) is facilitated by intermittent exoskeletal shedding, resulting in wandering third instar larvae prior to pupariation.To verify enhancer trapping by the SX4 P-element, we intercrossed novel insertion lines with flies harboring a "reporter" transgene encoding LexAop linked to a cDNA encoding a membrane-GFP (LexAop2-CD8::GFP; Pfeiffer et al. 2010), then confirmed membrane-associated GFP expression in tissues dissected from L3 larvae (Methods, Figs.3-5, Supplementary Table 1, Kockel et al. 2016).We analyzed wandering 3rd instar larvae of bitransgenic offspring (hereafter referred to as SX4>lexAop-GFP) after immunohistochemistry (IHC) staining for GFP, and simultaneous counter-staining for cell nuclei (DAPI).Images from selected SX4>lexAop-GFP tissues were collected, and tissue expression catalogued (Supplementary Table 1).Within the collection, we detected GFP expression in multiple tissues of L3 larva, including neuronal cell types in the central nervous system (CNS), ventral nerve cord (VNC), and peripheral nervous system, imaginal discs, and a wide range of other somatic tissues like fat body, Malpighian tubules, trachea, and ring gland (Supplementary Table 1).
patched (ptc) has a well-studied transcriptional expression pattern (Phillips et al. 1990, Supplementary Table 1, Fig. 3), and ptc-Gal4 is among the most-referenced Gal4 lines in the Drosophila literature (https://flybase.org/GAL4/freq_used_drivers/).To test if a lexA enhancer trap in ptc reproduced the known expression pattern, we analyzed line SX4S8-9, located near the transcriptional start of ptc (ptc-lexA SX4S8-9 ) in 2L at 44D1.Analysis of third instar wing discs revealed a LexA expression domain from ptc-lexA SX4S8-9 along the anterior-posterior boundary, as described for ptc RNA whole mount in situ hybridization of wing imaginal discs (Phillips et al. 1990).In addition, we observed expression in the CNS and VNC, enteric neurons, and putative hub cells and spermatocytes of the larval testis, consistent with prior reports (Fig. 3, Li et al. 2022).In summary, LexA expression from the SX4S8-9 enhancer-trap element inserted into the ptc locus reproduced several features of ptc expression across diverse somatic and germ line tissues and cell types.
To address if distinct lexA enhancer trap insertions produce distinct expression patterns, we surveyed expression of LexA by imaging the membrane-tagged GFP reporter in L3 larval brains and associated tissues like the ring gland of 6 independent SX4 insertions.For example, LexAop-CD8::GFP expression was directed by LexA from an insertion in CG9426 (SX4Aq854), bsf/Ntf-2r (SX4Hb22-1), lola (Sx4Lw221A), and vnc (SX4Pr4).We observed distinct patterns of cell labeling in the CNS, VNC, and ring gland (Fig. 4, a-f), in accordance with SX4 enhancer trap insertions located in distinct loci of the Drosophila genome give rise to distinct expression patterns.To facilitate accessibility of all molecular and imaging data, including supplementary images (Supplementary Table 1), we uploaded these to a database (https://stanx.stanford.edu),searchable by expression pattern, cytology, and specific genes.

Identification of SX4 lines that express LexA in insulin-secreting neurons
Systemic insulin in Drosophila emanates from a paired cluster of neurons in the pars intercerebralis comprised of 12-14 insulin-producing cells (IPCs: Fig. 5, a-f).IPCs express genes encoding Drosophila insulin-like peptides (Ilp's), including ilp-2, ilp-3, and ilp-5 (Brogiolo et al. 2001;Rulifson et al. 2002;Li et al. 2022).Prior enhancer trap studies identified homogeneous LexA expression in these insulin + cells, suggesting shared regulatory features within individual IPCs.However, we noted heterogeneous expression of the SX4Et7 enhancer trap, with expression of the LexAop::GFP reporter only in a subset of 1-2 IPCs (green, Fig. 5a") within the cluster of Ilp2 + IPCs (red, Fig. 5a').To investigate the possibility of heterogeneous genetic regulation in IPCs, we screened 87 lines and identified 16 additional lines with LexA activity in IPCs, identified by expression of the LexAop::GFP reporter in Ilp2 + IPCs (Fig. 5 and Supplementary Table 1).
We selected six genes trapped by an SX4 insertion (Fig. 5: SX4Et7 in B52; SX4Et21 in kis; SX4Et23 in Afd1; SX4Et38 in Hel89B; SX4Pr5 in Star; and SX4Lw22e in myd) that were confirmed to drive expression in IPCs and cross-referenced expression of that gene using IPC gene expression data from the Fly Cell Atlas (FCA, Li et al. 2022).The IPC FCA data is based on single-nuclei RNAseq (snRNAseq) of fluorescence activated cell sorting-sorted IPC nuclei labeled by ilp2-Gal4; UAS-2xunc84::GFP (Li et al. 2022).Single-nuclei libraries from IPCs had robust expression of ilp-2 confirming their IPC identity (Supplementary Fig. 2).In total, snRNAseq data from 473 IPC nuclei (232 male, 241 female) were correlated to the IPC expression of genes tagged by the selected enhancer traps, including all six cases detailed in Fig. 5.For example, kis was expressed in 362/473 nuclei, while Adf1 was expressed in 41/473 nuclei.In summary, snRNAseq data of IPCs confirms the expression of genes tagged by SX4 LexA enhancer traps in IPCs.

Evidence of RNA and LexA expression from natural TEs in somatic cells
Robust repression of natural TE sequences in both somatic and germ line cells has been reported (Czech et al. 2018;van den Beek et al. 2018).We therefore addressed if somatic expression of lexA derived from SX4 enhancer traps integrated into natural TEs was detectable.We analyzed the L3 brain expression patterns of four independent lines harboring a SX4 element integrated into distinct natural TEs: two different 1360 elements (tapped by SX4Ch7 and SX4Aq839, respectively), Copia (SX4Et51), and HMS Beagle (SX4Et8) (Fig. 6, Supplementary Tables 1 and 2; Methods).Analysis of LexAop-CD8::GFP expression driven by these lines revealed a clearly detectable presence of lexA activity in L3 brains, as well as distinct expression patterns of the SX4 lines from each other (Fig. 6).This includes SX4Et8, which was expressed in IPCs (Fig. 6d, see below), and SX4Ch7 and SX4Aq839, independent insertions into 1360 elements that are positioned at different genomic locations (Fig. 6, a and b).These findings suggest that the genomic location, rather than the TE itself, influences LexA expression of the SX4 insertion (Treiber and Waddell 2020).
To independently confirm somatic expression of natural TEs, we analyzed snRNAseq data from IPCs and CC cells for amplicons  Treiber andWaddell 2020, Li et al. 2022).This analysis revealed RNAseq amplicons aligning to natural TEs in snRNAseq libraries derived from IPCs and CC cells.On average, ∼5% and ∼2% of the overall RNA content of CC cells and IPC snRNAseq libraries, respectively, map to natural TE sequences (Supplementary Table 1 and Supplementary Fig. 2).We stratified this expression to two "classes" of natural TE: class II DNA cut-and-paste TEs (1360) and class I retrotransposons (Invader1, Opus, Juan, F, mdg3, and Invader4; Supplementary Fig. 2) and found both classes of natural TEs expressed in CCs and IPCs.The presence of multiple TE copies precluded unambiguous determination of expression levels derived from a single specific natural TE insertion.In summary, the RNA expression from natural TEs as determined by snRNAseq suggests somatic RNA expression derived from natural TEs in CC cells and IPCs (Tsurumi and Li 2019, Copley and Shorter 2023, Lindehell et al. 2023).The specific expression patterns in L3 brains of SX4 enhancer traps inserted into natural TEs support the hypothesis of a local mechanism of derepression in this early life stage.

An international scholastic network to generate resources for Drosophila genetics
In our prior studies, we produced and characterized novel fly enhancer-trap lines through an interscholastic partnership of secondary school and university-based researchers in the United States (Kockel et al. 2016(Kockel et al. , 2019;;Chang et al. 2022).This involved development of curricula permitting flexible scheduling of three laboratory-based "modules".These are comprised of fly intercrosses and P-element mobilization (module 1), molecular biology and enhancer-trap mapping (module 2), and immunofluorescencebased microscopy of dissected larval tissues to confirm LexA activity in specific tissues (module 3: Methods).In some schools, this occurred in year-long courses (Supplementary Fig. 1A), while in other schools, specific elements like mapping of P-element genomic insertions, or genome sequencing, were achieved by shorter classes (Kockel et al. 2019;Methods).This curricular model integrated and enhanced the longitudinal quality of genetic experiments performed across years at specific schools (flies generated in one course could be characterized the following semester by another set of students).This prior work also demonstrated effective, productive collaborations of students and instructors across institutions (for example, flies generated in one school could be shared with another school that performed molecular mapping studies).Over a 10-year span (2012-2022), this network of collaborating schools has expanded to 17 schools (Fig. 7).
To meet growth of the Stan-X network and demand for teacher training, the Teacher Academy "Discovery Now" was instituted in 2018.Incoming teachers receive a 2-week intensive training, one week online, one week in person (Supplementary Fig. 1B).This course prepares new instructors to implement the Stan-X research curriculum of molecular biology and genetics, and provides grounding in essential course logistics like equipment acquisition.This summer training for instructors is provided annually (www.Stan-X.org).
Here, we assessed if the curriculum of Drosophila-based genetics, molecular biology, genomics, and tissue analysis framing original, high-quality research could be adopted at additional secondary schools and universities, including abroad.As indicated by the data and resources detailed here, our studies show that research at secondary schools and universities in the United States and United Kingdom fostered production and sharing of data and fly strains, and achievement of student learning goals.
In this study, two hundred ninety-four high school students, thirty-two high school teachers, and staff supervised Stan-X classes    In addition to curricular development at these schools, these interscholastic partners benefitted from structured interactions with network leaders at Stanford, Lawrenceville, and Exeter that included weekly research teleconferences with course instructors and classes during the school year.In addition, there were university-based summer internships for students (n = 21) or instructors (n = 6), and development of annual student-led conferences with participants from multiple schools for presenting data and curricular innovations.In turn, university collaborators made regular visits to secondary school classes during the school year (Methods).
There were also multiple positive outcomes for students and teachers at partnering schools that were unanticipated.These included (1) emergence of student course alumni as instructors at their home institution or another Stan-X partner site (n = 6), (2) interscholastic collaboration and data development through sharing of Stan-X fly strains and other resources, (3) regular video conferencing and in-person multi-institutional student symposia organized independently by Stan-X instructors, (4) additional professional development opportunities for adult teachers, including presentation of pedagogy at professional meetings, promotion, and travel to other Stan-X partner schools, (5) development of new courses founded on Clustered Regularly Interspaced Short Palindromic Repeats and other approaches to generate novel LexA or LexAop strains (Chang et al. 2022;Wendler et al. 2022) or fly genomics (see Kockel et al. 2019), ( 6) development of Stan-X summer school courses at Harvard, Oxford, Lawrenceville, and Exeter (Supplementary Fig. 1B, https://stan-x.org),and (7) philanthropic funding for science curricular innovation and infrastructure modification to Stan-X partners.Thus, an intercontinental consortium of students and instructors at secondary schools and university-based programs have formed a unique research network actively generating novel fly strains suitable for investigations by the community of science.

Discussion
Here, we introduced a novel lex-A enhancer trap construct in a unique isogenic background (iso 113232 ) and used this element to generate more than 300 novel LexA enhancer trap insertions through scholastic courses at seventeen institutions.We characterized gene expression of a substantial fraction of these insertions in 3rd instar larval organs or tissues like, CNS, VNC, and gut, with a special emphasis on IPC expression in the larval CNS.Future studies could address similarities and differences of independent derived enhancer-trap lines with a similar insertion site (e.g. a region of 1 kb).We also generated a SX4 P-element on the third chromosome TM6B balancer (TM6B, SX4 orig ) that was successfully mobilized for selection of X-linked enhancer traps.Analysis of the SX4 insertion site sequence, insertion directions, and genomic insertion sites associated with the 5′ end of transcription units revealed a similar profile compared to the first-generation SE1 lexA enhancer trap (Kockel et al. 2019).Thus, the new SX4 LexA enhancer-trap element has multiple hallmarks of a P-element insertion vector.In summary, the new SX4 "starter" lines represent an advance over the prior SE1 LexA enhancer-trap line (Kockel et al. 2016).
Prior work revealed the presence of KP elements as an impediment for P-element mobilization experiments (Kockel et al. 2019) reflecting: (1) bias for replacement of the KP element by the enhancer trap P-element, (2) KP mobilization by Δ2-3 transposase giving rise to uncontrolled genetic heterogeneity, and (3) dominant-negative effect of KP on Δ2-3 transposase, lowering the overall transposition rate per male germ line.Here, we used the genetic background iso 113232 , where we confirmed the absence of an autosomal KP element by genomic sequencing.In this genetic background, as predicted, we observed increased transposition frequency and more randomized insertional distribution of our SX4 starter P-element across the chromosomes.These features have also improved workflows in participating school courses.
Each natural transposon family is present in multiple identical, or nearly-identical copies of DNA sequence per genome; thus, unambiguous mapping of insertions within these repetitive sequences requires special strategies.Here, we report the index, successful mapping of SX4 P-element insertions into natural transposons.We successfully mapped 11/17 SX4 insertions in natural TE sequences, as well as the position of the SX4-tagged natural TE within the Drosophila genome.Of note, 4/11 natural TEs that were tagged by SX4 and positionally identified are not represented in the current release 6 of the Drosophila genome.Therefore, these represent unique natural TEs copies specific to the iso 113232 background.
Intact natural TEs encode transposase enzyme (class II, cut-and-paste TEs), or other factors mediating replication and insertion (class I, RNA transposons; McCullers and Steiniger 2017).To prevent mobilization and genome instability, the transcription of natural TEs is thought to be repressed in somatic and germ line tissues (Senti and Brennecke 2010;Czech et al. 2018;van den Beek et al. 2018).However, SX4 enhancer traps tagging natural TEs showed clear LexA expression in a variety of somatic cells, including L3 neurons, indicating the accessibility of transcriptional machinery to the genomic locus harboring the TE.This is consistent with observations of active transcription of natural TEs in the Drosophila adult brain (Treiber andWaddell 2020, Lindehell et al. 2023), increased natural TE expression in aged flies (Tsurumi andLi 2020, Yang et al. 2022), and our analysis of TE expression by snRNAseq in IPCs and CC cells.This suggests that somatic transcription and, possibly, transposition of natural TEs might occur in somatic tissue in vivo (Siudeja et al. 2021;Yang et al. 2022, Copley andShorter 2023).
Experimental biology benefits from temporal-or cell typespecific control of gene expression, exemplified by the binary expression strategies pioneered in the Drosophila GAL4-UAS system (Brand and Perrimon 1993).Intersectional approaches, like simultaneous use of the LexA-LexAop and GAL4-UAS systems, have also greatly enhanced experimental and interpretive power in fly biology, particularly studies of neuroscience and intercellular communication (Simpson 2016;Dolan et al. 2017;Martín and Alcorta 2017).Thus, new LexA enhancer-trap lines presented here significantly expand the arsenal of available LexA expression tools (Pfeiffer et al. 2010;Kockel et al. 2016).Prior studies have demonstrated that P-element insertion in flies is nonrandom (O'Hare and Rubin 1983;Berg and Spradling 1991;Bellen et al. 2011), with a strong bias for transposition to the 5′ end of genes (Spradling et al. 1995).
Here and in prior work, we have found a similar preference with SX4 P-element transposition; 89% of unique insertions were located in the promoter or 5′ UTR regions of genes.Molecular characterization and studies of LexAop-regulated GFP reporter genes indicate that the enhancer traps described here are distinct, with LexA expressed in multiple tissues, including the CNS, VNC, fat body, and muscle.These enhancer-trap lines were submitted to the Bloomington Stock Center to enhance resource sharing.
The resources and outcomes described here significantly extend and develop the interscholastic partnership in experiment-based science pedagogy described in our prior studies (Kockel et al. 2016(Kockel et al. , 2019)), which previously involved Stanford University researchers and biology classes at four US secondary schools.A scholastic network, called Stan-X, now links university researchers with secondary school and undergraduate students and teachers around the world.The Stan-X network used P-element mobilization in Drosophila melanogaster to generate LexA enhancer-trap lines reported here.Curricula based on fruit fly genetics, developmental and cell biology, and molecular biology provided a practical framework for offering authentic research experiences for new scientists detailed previously (Kockel et al. 2016(Kockel et al. , 2019;;Redfield 2012).Important research and educational goals, including a keen sense of "ownership" of problems (Hatfull et al. 2006) and discovery, were achieved because the outcomes from experiments were "unscripted".In addition, work permitted students and instructors to create tangible connections of their experimental outcomes (data, new fly strains) to a global science community.Data and tools from this international scholastic network demonstrated how university research laboratories can collaborate with community partners, including with resource-challenged schools serving youth under-represented in science, to innovate experiment-based STEM curricula and experiential learning that permit discovery, the sine qua non of science.
Indices of practical outcomes from our work include steady requests for LexA enhancer-trap lines (currently >450 Stan-X lines) from the Drosophila Bloomington Stock Center, and Stan-X fly strain use has been cited in 24 publications since 2016 (e.g. Lee et al. 1996;Cohen et al. 2018;Zhou et al. 2020;Ribeiro et al. 2022).Our interscholastic partnerships and classroom-based research have expanded to include high schools and universities on multiple continents (Fig. 7).The secondary schools encompass a spectrum of public, charter, independent and "high needs" schools, with day or boarding students.Three Stan-X partners are in public high schools serving ethnically and economically diverse urban communities (Lowell, San Francisco; Pritzker, Chicago; and Commack, Long Island, NY), while the remainder are independent secondary schools or private universities.This experience demonstrates the feasibility and challenges of expanding the Stan-X model to public schools, which have unique resource challenges.Stan-X programs have instructed seven hundred fifty-two students since 2012, 67% female.At independent schools, 55% of Stan-X students were female (n = 562); at public high schools, 70% were female (n = 190).These findings suggest that curriculum-based experimental science programs like Stan-X could help address persistent genderbased disparities in science, though this possibility requires further study with case controls.Similar to the experience of others (DrosAfrica 2020), we have found that the Stan-X curriculum can also be used abroad to foster Drosophila-based pedagogy.Additional work outside the scope of this study is also needed to assess the longitudinal impact of programs like Stan-X on ethnic or socio-economic disparities in the scientific workforce.
In summary, this experience demonstrates the feasibility of developing productive global partnerships between schools to foster experience-based science instruction with a powerful experimental genetic organism.The thriving partnerships described here form a dynamic network of instructors, students, classes, and school leaders that have produced useful science, and enhanced the personal and professional growth and development of its participants.

Data availability
All Stan-X SX4 derivatives and associated data are available at the Bloomington Stock Center.The genome sequence data of w[1118], SX4; iso#32 [II]; iso#32[III] is available on SRA https://www.ncbi.nlm.nih.gov/sra/PRJNA912892, or accession number PRJNA912892.All molecular and image data are additionally available at https:// stanx.stanford.edu.Course manuals, scaffolding problem sets, and sample course daily and weekly schedules are available on request.
Supplemental material available at G3 online.

Fig. 1 .
Fig. 1. wt P-element and lexA enhancer traps.a) wt P-element described in O'Hare and Rubin (1983).b) SE1 lexA enhancer trap used in Kockel et al. (2019).c) SX4 lexA enhancer trap used in this study.The SX4 enhancer trap encodes a lexA DNA-binding domain fused to the hinge and transactivation domain of Gal4, driven by the hsp70 promoter.The enhancer trap is marked by the white selectable eye color marker.See Supplementary Data File 1 for annotated sequence of SX4.d) Sequence logo (see Methods) derived from 281 independent 8 bp direct repeat sequences caused by SX4 insertion.

Fig. 7 .
Fig. 7.The Stan-X program.a) Timeline of school recruitment into the Stan-X program.Recruitment takes place during the school year, followed by training of the teachers in the Discover Now Teacher Academy in the summer (Supplementary Fig. 1).See text for details.
All three Stan-X Biology Course modules are taught at Phillips Exeter Academy, NH; Commack High School, Dalton School, and Chapin School in NY; Pritzker College Prep and Latin School of Chicago, both in Chicago, IL; The Lawrenceville School, NJ;