The Ntan1 gene is expressed in perineural glia and neurons of adult Drosophila

The Drosophila Ntan1 gene encodes an N-terminal asparagine amidohydrolase that we show is highly conserved throughout evolution. Protein isoforms share more than 72% of similarity with their human counterparts. At the cellular level, this gene regulates the type of glial cell growth in Drosophila larvae by its different expression levels. The Drosophila Ntan1 gene has 4 transcripts that encode 2 protein isoforms. Here we describe that although this gene is expressed at all developmental stages and adult organs tested (eye, antennae and brain) there are some transcript-dependent specificities. Therefore, both quantitative and qualitative cues could account for gene function. However, widespread developmental stage and organ-dependent expression could be masking cell-specific constraints that can be explored in Drosophila by using Gal4 drivers. We report a new genetic driver within this gene, Mz317-Gal4, that recapitulates the Ntan1 gene expression pattern in adults. It shows specific expression for perineural glia in the olfactory organs but mixed expression with some neurons in the adult brain. Memory and social behavior disturbances in mice and cancer and schizophrenia in humans have been linked to the Ntan1 gene. Therefore, these new tools in Drosophila may contribute to our understanding of the cellular basis of these alterations.

The Ntan1 gene encodes the Drosophila N-terminal asparagine amidohydrolase homolog (NTAN1). It regulates endoreplication rates through the N-end rule pathway 1 . It has been proposed that in Drosophila, similarly to mammalian NTAN1 2 , this enzyme recognizes proteins with an N-terminal asparagine (Asn) and converts it to aspartate (Asp) 1 . After the action of other enzymes, such as ATE1, this site becomes then a potential substrate for ubiquitination and proteasomal degradation 1, 3,4 .
At the cellular level, development of an organism consist on a myriad number of events that lead to the formation of an adult individual. Among these events, genetic regulation of cell number and size, controlled by growth and division 5 , is of extreme importance. In the nervous system, developmental processes need to strictly control and regulate glial and neuronal cell growth in order to achieve a functional organism.
In Drosophila, the surface of the entire nervous system is covered by two thin glial layers: the perineural glia (PNG) and the subperineural glia (SPG) 6 . Both types of glial cells are involved in the formation of the blood-brain barrier 7,8 . During larval growth, while the PNG grows exponentially in number by extensive cell division 9,10 the SPG retains its small number and massively increases the size of its cells individually by both endoreplication and endomitosis becoming then polyploid 11,12 . It has been shown that in the blood-brain barrier of Drosophila the endoreplication of SPG cells is regulated by the Ntan1 gene 1 , previously named Öbek.
In fact, in Drosophila larvae, glial cells present asparagine amidohydrolase activity of Ntan1 1 . The mechanism proposed in these cells for regulating their growth involves different expression levels of the Ntan1 gene. During larval development, in SPG cells high expression levels of Ntan1 limit cell cycle and replication rates controlling ploidy and number of nuclei 1 . On the contrary, PNG cells express moderate levels of Ntan1, allowing then efficient PNG proliferation 1 .
Although this gene has been thoroughly studied in larvae, a detailed description of gene expression that pays attention to its different transcripts both during development and in the adult phase is still lacking. In this article, we tackle this question by analyzing mRNA expression of the different transcripts of the Ntan1 gene during different developmental stages and several adult organs such as the brain, eyes and antennae as representatives of www.nature.com/scientificreports/ central or peripheral nervous systems. Moreover, we show tissue-specific expression of the Ntan1 gene in adults by using available Gal4 lines.
In the peripheral nervous system of Drosophila, such as that involved in olfactory reception, antennal glial cells begin to originate about 16 h after puparium formation (APF) and are closely associated with the development of sensory axons [13][14][15] . In the adult antennae, there are mainly two different types of glia, GH146-type and Mz317type glia. GH146 glial cells correspond to ~ 30% of the glial antennal cells 15 . This glia originates in the brain and migrates to the third antennal segment where it ensheathes the axons of the ORNs which are projected from the antenna to the brain 15 . Mz317 glial cells corresponds to ~ 70% of antennal glial cells. They constitute a peripheral layer to GH146 glia and seem to surround the somas of olfactory sensory neurons 15 . Although Mz317::GFP is considered a marker of perineural and cell body glia labelling most of peripheral glia 16 , nothing is known about the molecular basis of this type of glia. In this report, we demonstrate that the Mz317 driver recapitulates the expression of the Ntan1 gene as well as the other Gal4 drivers located within this gene.

Results
Ntan1 gene is expressed in all developmental stages from embryo to adult Drosophila. Although it was previously shown that NTAN1 was strongly expressed in embryo and larvae of Drosophila in a subset of glial cells in both the PNS and the CNS along with a weakly expression in other tissues 1 , little is known about the Ntan1 gene expression in other stages and adult organs. In order to test the native expression of the Ntan1 gene in different pre-adult developmental stages (embryo, larva and pupa) and some adult organs (third antennal segments, brain and eyes) of Drosophila melanogaster we performed RT-PCR. As described in the material and methods section, primers were designed in order to amplify different size products from gDNA and cDNA (Fig. 1A, Table 4). The expected size for the cDNA of the Ntan1 gene is 685 base pairs (bp), while for the genomic DNA is 901 bp. All products obtained in the wildtype genotype were of the expected sizes (Fig. 1B). These results confirm the expression of the Ntan1 gene in all the developmental stages and adult organs studied.
Gene expression of Ntan1 transcripts at different developmental stages and adult organs. According to Flybase, the Ntan1 gene has 4 transcripts: Ntan1-RA, Ntan1-RB, Ntan1-RC and Ntan1-RD, identical in their coding sequence two-by-two (Ntan1-RA and Ntan1-RB; Ntan1-RC and Ntan1-RD), but differentiable by their 5' and 3' UTR (UnTranslated Region) regions (Fig. 1A). The specific gene expression of each transcript of the Ntan1 gene was analysed in the same developmental stages and organs previously mentioned. Thus being able to determine if there is one or several transcripts specific to a particular stage or organ tested. Amplification of genomic DNA was used as a PCR positive control. The expected cDNA sizes of Ntan1-RA, Ntan1-RB, Ntan1-RC and Ntan1-RD transcripts are 1136 bp, 1336 bp, 1405 bp and 1601 bp, respectively (Fig. 1C, Table 1). Table 1 summarizes the expression data for all transcripts in the different samples studied. Ntan1-RA transcript is expressed in embryo, larva, pupa, brain and third antennal segments but not in the adult eye whereas Ntan1-RB transcript is only expressed in larva and brain. Finally, both Ntan1-RC and Ntan1-RD transcripts are expressed in all developmental stages and organs studied (Table 1, Fig. 1C, Fig. S4).
The Ntan1 gene is highly conserved throughout evolution. Although in Drosophila the Ntan1 gene has four transcripts that differ mainly in their UTR regions, they are translated into two different isoforms. As N-terminal asparagine amidohydrolases have been found in other species, we performed a comparative analysis of the aminoacid sequences of the NTAN1 proteins isoforms of various insects belonging to the Diptera, Hymenoptera and Lepidoptera Orders as well as their corresponding mouse and human orthologs. We infered their evolutionary relationships by constructing a phylogenetic tree that clearly illustrates the evolutionary conservation of the products of the Ntan1 gene ( Fig. 2A). Although the percentage of identity between the Drosophila melanogaster NTAN1-1 isoform and the rest of ortholog isoforms from the species tested ranged from 34 to 87%, the percentage of similarity increases to the range between 61% to 96.5%. In fact, Drosophila and its human ortholog proteins share more than 72% of aminoacid similarity (Fig. 2B). In conclusion, we show that the Ntan1 gene is highly conserved through evolution, not only among insects but also in mammals (Fig. 2). It is generally accepted that high conservation of genes along evolution indicates an essential functional role.
Mz317-Gal4 insert is located into the Ntan1 gene and drives gene expression in adult brain and antennal glia. As it was previously shown that the Ntan1 gene was involved in controlling the levels of ploidy in subperineural and perineural glial cells during larval development 1 we wondered if this gene was also being expressed in adult perineural glia. The Mz317-Gal4 line combined with a UAS-GFP reporter line drives the expression of GFP labeling the majority of peripheral glia including most antennal glia 15 but the gene driving this especific mark was unknown. In order to determine the genomic location of the Gal4 insert of the Mz317-Gal4 line we performed the Splinkerette technique 17 . This method allows to determine P-element insertion sites with unknown genomic location. By PCR amplification and posterior sequencing of the 3′ P-element insertion site (Figs. S1 and S2) we were able to conclude that the Mz317-Gal4 line has its Gal4 element inserted into the Ntan1 gene, specifically around 150 bp upstream the second 5′ UTR region.
It was known from previous studies that the Mz317-Gal4 line drives gene expression in antennal perineural glia 15 but in order to further investigate the cellular expression pattern of the Mz317-Gal4 line in different cell types we analyzed GFP expression in the genotype Mz317-Gal4;UAS-mCD8::GFP both in antennal cryosections and brain dissections of Drosophila melanogaster. For this purpose, immunohistochemistry was performed with anti-GFP antibody on antennal cryosections and with anti-GFP and anti-nc82 antibodies on brain samples. The anti-nc82 antibody recognizes the neuropil-specific BRUCHPILOT protein, which allows us to visualize the general structure of the brain 18,19 . Also, to distinguish the specific cell types observed in the www.nature.com/scientificreports/ Mz317-Gal;UAS-mCD8::GFP line, we restricted GFP expression in either neurons or glia by adding gene expression restriction elements such as elav-Gal80 and repo-Gal80, respectively. Genetic crosses were performed to obtain the following genotypes: Mz317-Gal;UAS-mCD8::GFP (GFP expression in all Mz317-type cells), Mz317-Gal4/repo-Gal80;UAS-mCD8::GFP (GFP expression absent in glia due to GAL80 expression driven by repo, a panglial marker), Mz317-Gal4/elav-Gal80;UAS-mCD8::GFP (GFP expression absent in neurons due to GAL80 expression driven by elav, a panneuronal marker). Immunostaining of the UAS-mCD8::GFP line by itself was used as a negative control.
In the antenna, the Mz317-Gal4 line directs GFP expression exclusively in glial cells, as we did not observe any GFP expression in antennal neuronal cells (Fig. 3A). Furthermore, the antennal GFP expression pattern in Mz317-Gal4/elav-Gal80;UAS-mCD8::GFP, that restricts Ntan1-drived expression to non-neuronal cells, matches exactly the GFP expression pattern of the Mz317-Gal;UAS-mCD8::GFP individuals previously described 15 (Fig. 3A). Antennal GFP expression is observed at the base of the sensilla, most likely surrounding the neuronal somas of most olfactory sensory neurons. Likewise, GFP expression is observed around the olfactory nerve. Thus, the expression pattern of the Mz317-Gal4 line corresponds exclusively to perineural antennal glial cells.
As expected, GFP expression is clearly observed as a wrap around the whole brain structure (Fig. 3B). Additionally, we also observe staining in a selected number of non-glial cells in the brain (Fig. 3B). These cells  www.nature.com/scientificreports/ lines with inserts in the Ntan1 gene that showed a similar expression pattern. Two other lines have driver insertions in the Ntan1 gene, Mz97-Gal4 and Ntan1 NP1111 -Gal4. In both cases, the Gal4 elements are inserted in the first intron of the Ntan1 gene ( 1 and https:// flyba se. org/ respectively). Note that the non-transcribed genomic DNA that is located in the middle of both 5'UTRs is considered the first intron according to the Ensembl database. While the Mz97-Gal4 insert is located ~ 100 bp upstream the second transcriptional start site the insertion of the Ntan1 NP1111 -Gal4 is located ~ 1300 bp upstream this region (Fig. 1A).
In order to compare the expression patterns of these Gal4 lines with Mz317-Gal4 we performed immunohistochemistry with anti-GFP antibody in the antenna and anti-GFP and anti-nc82 in the brain in the following genotypes: Mz317-Gal4;UAS-mCD8::GFP, Ntan1 NP1111 -Gal4;UAS-mCD8::GFP and Mz97-Gal4,UAS-Stinger;UAS-mCD8::GFP. In the last genotype, cell nuclei will also be GFP-stained together with membrane-bound GFP (mCD8) due to the additional UAS-Stinger element that is also present in the original MZ97-Gal4 line.
The antennal GFP expression observed was very similar in the three lines tested (Fig. 4A). Both in Mz97-Gal4 and Ntan1 NP1111 -Gal4 we observed glial GFP staining located in the base of the sensilla of the third antennal segment, most likely surrounding the neuronal somas of the olfactory sensory neurons as in Mz317-Gal4 (Figs. 3A and 4A). In the brain, there are slight differences in GFP expression among the different fly lines (Fig. 4B). All brains appear to have a glial wrap around their structure. However, Ntan1 NP1111 -Gal4;UAS-mCD8::GFP and Mz97-Gal4,UAS-Stinger;UAS-mCD8::GFP brains show a lower number of neuronal cells in the superior protocerebrum area, the mushroom body area and in the vicinity of the antennal lobes compared to Mz317-Gal4;UAS-mCD8::GFP (Fig. 4B). In the case of the Mz97-Gal4,UAS-Stinger;UAS-mCD8::GFP line there seems to be a higher expression of GFP, however, this is due to Stinger expression which also marks the cell nuclei. Taken together, these results indicate that the Ntan1 gene is mainly expressed in both antennal and brain perineural glia.
Gene expression of Ntan1 transcripts in different Gal4 lines. After determining the developmental stages and organs in which the different transcripts of Ntan1 were expressed (Fig. 1C, Table 1) we wondered if the slight differences observed in brain staining among the different driver lines tested could be due to qualitative differences in transcript expression in the fly lines tested. For this purpose, we studied the expression of Ntan1 gene in different developmental stages and adult organs of Mz317-Gal4 and Mz97-Gal4,UAS-Stinger lines. In addition, we have determined whether or not there are one or more transcripts specific to a particular stage or adult organ in each fly line. As the Gal4 element of Mz97-Gal4,UAS-Stinger and Ntan1 NP1111 -Gal4 lines is inserted in the same area (first intron), the gene expression analysis of Ntan1 gene has been carried out only in the Mz97-Gal4,UAS-Stinger line. As before, an RT-PCR experiment was performed to check the native expression of the Ntan1 gene in Mz317-Gal4 and Mz97-Gal4,UAS-Stinger lines. Genomic DNA amplification was used as a positive control. The expected size of cDNA of the Ntan1 gene was 685 bp using the primers Ntan1-F GGT GTG CTG CTG CAG GAT GACT and Ntan1-R CGT AGT TAA AGG GTC CAA TGC. In both lines tested, the PCR products obtained had the expected sizes (Fig. 5A). These results confirm the expression of the Ntan1 gene in the different stages and organs studied in Mz317-Gal4 and Mz97-Gal4,UAS-Stinger lines. The expression of each transcript was also evaluated in both lines (Fig. 5B). Although in the wildtype genotype Ntan1-RB transcript is not expressed in embryo and pupa our results show that all transcripts are expressed in embryo, larva and pupa of Mz317-Gal4 line (Table 2). Also, some differences are observed in relation to Mz97-Gal4,UAS-Stinger line and the wildtype genotype. Specifically, there is no expression of Ntan1-RA transcript in embryo of Mz97-Gal4,UAS-Stinger line and additionally, there is no expression of Ntan1-RC transcript in embryo, larva or pupa of Mz97-Gal4,UAS-Stinger line ( Table 2). On the other hand, in adult Drosophila, Ntan1-RA transcript is expressed in the brain and third antennal segments of the Mz317-Gal4 line, while it is only expressed in third antennal segments of Mz97-Gal4,UAS-Stinger. Ntan1-RB transcript is expressed only in the brain of both lines, whereas Ntan1-RC transcript is expressed in the brain, third antennal segments and eyes of the Mz317-Gal4 line and third antennal segments and eyes of the Mz97-Gal4,UAS-Stinger line. Finally, Ntan1-RD transcript is expressed in all three organs studied of both lines similarly to the wildtype genotype (Table 3). To sum up, the main gene expression differences in adult Drosophila between the wildtype genotype and the Gal4 lines studied are seen in the expression of Ntan1-RA and Ntan1-RC transcript of Mz97-Gal4,UAS-Stinger in the brain (Table 3). We observed no expression of the transcript Ntan1-RB in the embryo and in the pupa of the wildtype genotype (Fig. 1C, Table 1), whereas there is expression of this transcript in the Gal4 lines studied (Fig. 5B, Table 2). We performed an additional PCR using the G6PD housekeeping gene for the embryo and brain of the Mz97-Gal4 fly line, where we did not observe any expression of Ntan1-RA and Ntan1-RC (Fig. S3). With this data we can infer that even slight differences in the insertion sites of P(GawB) driver elements change gene expression of the Ntan1 transcripts that might account for slight differences also in GFP expression drived by the lines tested.

Discussion
In this article, we offer a detailed gene expression study of the Ntan1 gene, previously characterized as encoding an N-terminal asparagine amidase in Drosophila 1 as well as in other species of invertebrates and vertebrates 2 . At the cellular level its role has been related to the control of cell cycle and replication rates controlling ploidy and number of nuclei in some cellular types, especially in glia. In this cell type moderate expression levels associate to perineural glial cells (PNG) while high expression levels appear in subperineural glial cells during larval development in Drosophila 1 . In summary, the quantitative expression of the gene rather than just its presence or absence appears to be at the basis of its function deduced from studies of the entire gene. However, the gene encodes several transcripts and the importance of qualitative differences in expression depending on the developmental stage and cell type cannot be excluded beforehand.
In this report we study the expression of the Ntan1 gene and each of its 4 transcripts − RA to − RD during different developmental stages and in some adult organs to see if there is a specific expression pattern for any of  www.nature.com/scientificreports/ them. It should be reminded that the differences in the sequence of the transcripts mainly affect their regulatory regions rather than the coding sequence that is identical between − RA and − RB (isoform 1) and between -RC and -RD (isoform 2), respectively. The results indicate that although there is expression of the Ntan1 gene in all stages and organs tested there are some specificities in the transcript expression (Fig. 1B,C, Table 1). Looking at the different developmental stages (embryo, larva and pupa), − RA, − RC and − RD transcripts appear always, only the − RB transcript shows expression in the larva but does not express in embryo and pupa. In the adult stage the Ntan1-RB transcript is possibly specific to the brain, as it does not appear in neither the third antennal However, it is important to highlight that although the coding sequence of the Ntan1-RB transcript is identical to the sequence of Ntan1-RA transcript (Fig. 1A) they differ in their regulatory regions, 5′ UTR and 3′ UTR. Thus, these differences in transcript expression might account for the different expression levels of the Ntan1 gene that drive different cell growth mechanisms in PNG cells and the SPG cells 1 . In vertebrates a similar behavior has been reported for this gene, as variable expression of an approximately 1.4-kb transcript was reported in all 8 mouse tissues examined, with lowest expression in testis. However, a transcript of about 1.1 kb was prominently expressed in testis only 2 .
By attending to the two characteristics, transcript and isoform type, qualitative and quantitative determination seems to play an important role in gene function. In Drosophila there are some organs such as the compound eyes that express only one Ntan1 isoform, while in the other cases analyzed both are expressed, but including either one or both transcripts. However, it should not be forgotten that both the different developmental stages studied and the adult organs include a variable number of cell types, so that the data obtained as a whole may be masking specificities at the cellular level. Some data in this direction was reported in mouse, where the Ntan1 expression decreased following differentiation in C2C12 mouse myoblasts, but not in MEL mouse leukemia cells 2 .
Detailed studies at the cellular level are difficult to address in small organisms by RNA expression analysis when the gene studied shows widespread expression in different cell types and quantitative issues play an important role. In such cases, immunohistochemical analysis with the gene product does not usually give consistent results either. In Drosophila, thanks to the development of numerous genetic tools such as the Gal4/UAS system and its variants, this analysis can be done by selective labeling with reporter genes such as GFP whose expression is directed with drivers that recapitulate the expression pattern of the gene under study 23,24 . In many cases, this is obtained by insertions within the gene. These tools can even serve to follow gene expression in different cell types throughout development like in the G-TRACE method 25,26 , as well as to direct the generation of functional mutants restricted to some cell types. However, the validity and significance of the data obtained depends on the accuracy and reliability with which these drivers recreate the expression of the genes of interest.
In this sense, we have located a new driver within the Ntan1 gene (Mz317-Gal4 line) and we have studied the expression of the gene in this line compared to wildtype flies. Moreover, we have described the expression patterns in glia and neurons of other driver lines that had already been described by their insertion within this gene (Mz97-Gal4 and Ntan1 NP1111 ), in order to characterize how they recapitulate the expression of the original gene and can be used for further studies.
In this work we report for the first time that the Mz317-Gal4 line 16 has an insert within the Ntan1 gene. According to Sen et al. 15 Mz317::GFP is a marker for perineural glia that labels the majority of peripheral glia. We characterized the expression of Ntan1 in the Mz317-Gal4 line both at the molecular and cellular level. RT-PCR Table 2     www.nature.com/scientificreports/ analysis of the adult organs in the Mz317-Gal4 line recapitulates the same expression pattern that the wildtype individuals. At the cellular level, immunohistochemistry in the third antennal segment and brain of Mz317-Gal4;UAS-mCD8::GFP individuals reveals the cell types marked in both structures. Although we detected specific expression of perineural glia in the third antennal segment, in agreement with what has been previously reported 15 , the labeling in the brain includes mostly glial cells and a reduced number of neurons.
We also compared Ntan1 gene expression in wildtype and Mz317-Gal4 individuals with that obtained with other drivers previously mapped to the Ntan1 gene, both at the molecular and cellular levels. The transcript Ntan1-RB is not expressed neither in the embryo nor in the pupa of the wildtype genotype whereas there is expression of this transcript in the Gal4 lines studied. In adults, the expression pattern of the Mz317-Gal4 line matches that of wildtype individuals while the Mz97-Gal4,UAS-Stinger line shows no expression of the -RA or -RC transcript in the brain. In a previous work, it was indicated that the Mz97-Gal4 insertion might have led to a mutant Ntan1 allele 1 . In homozygous Mz97-Gal4 flies there is still expression of various transcripts of the Ntan1 gene, although quantitative differences in gene expression might not be detected by the methodology used. Our data suggest that a mutant phenotype may be generated if it depends on the quantitative expression of either or both Ntan1 isoforms in the brain.
At the cellular level, the expression patterns of the different Gal-4 lines mapped to the Ntan1 gene are very similar. In the third antennal segment, expression is restricted to PNG glia, whereas in the brain staining shows as a glial wrap with the exception that Ntan1 NP1111 -Gal4;UAS-mCD8::GFP and Mz97-Gal4,UAS-Stinger;UAS-mCD8::GFP show a lower number of neurons than Mz317-Gal4;UAS-mCD8::GFP.
The differences in gene expression among the different lines, both at the molecular and cellular level, relate to the different positions of the Gal4 insertion that may lead to mutations in some cases. This indicates the importance of performing previous expression studies if we want to use Gal4 insertion lines as tools for the study of gene expression as well as to generate cell type-specific mutations. However, the difficulties in targeting a certain cell type can be solved in some cases with additional elements like the Gal80 inserts.
Summarizing, we consider that the Ntan1 gene could be easily used as a marker and/or driver of just perineural glia, by restricting neuronal expression with a Gal80 insertion. Therefore, with this work we show that the Gal4 lines associated to the Ntan1 gene in Drosophila will add new genetic tools for studying the regulation and function of glia.
Glia is essential in the development and homeostasis of the nervous system due to its interaction with neurons and its modulatory role in neuronal activation 27 . In mammals, glia can constitute between 33 and 66% of the brain mass 28,29 . However, in Drosophila, glia makes up approximately 10% of the cells in the nervous system 6 . The low percentage of glial cells relative to other species and the similarity of its nervous system to vertebrates 27,30,31 , make Drosophila melanogaster an excellent model organism to investigate the biological mechanisms behind glial regulation of the nervous system. Finally, it has been shown that the Ntan1 gene performs numerous functions in vertebrates and defects in the gene can have serious effects at the level of the organism. Thus, alterations in the Ntan1 gene expression have been associated with different types of memory and social behavior 32,33 , sporadic childhood cancer 34 and risk of schizophrenia 35 . Therefore, studies in Drosophila using all the new tools available achieve a greater projection in determining the cellular basis of these phenotypes, which is necessary for their possible reversion. HsNTAN1-2: NP_001257695.1). All sequences were aligned with MUSCLE using default parameters. The alignment was used to build a Neighbor Joining tree with evolucionary distances computed using the Jones, Taylor and Thorton matrix-based method 36 , 1000 bootstrap replicates and gaps handled with pairwise deletion. Phylogenetic analyses were conducted in MEGA11 37 . Immunostaining. Immunofluorescence on antennal cryosections were performed as described in 38  www.nature.com/scientificreports/ Ligase, New England BioLabs, Ipswich, USA) to double stranded splinkerette oligonucleotide 17 for 2 h at room temperature in a total volume of 50 µl. Then, ligated DNA was used as a template to amplify both the 3′ and 5′ P-element insertion sites using primers 3′SPLNK#1 (CAC TCA GAC TCA ATA CGA CAC) and 3′SPLNK#2 (GGA TGT CTC TTG CCGAC) and primers 5′SPLNK#1-GAWB (TGG GAG AGT AGC GAC ACT CC) and 5′SPLNK#2-GAWB (GAG CTT TTT AAG TCG GCA AAT ATC G) respectively. After amplification, 20 μl of the amplification products were analysed by gel electrophoresis on a 2% agarose gel with a DNA ladder according to the amplified DNA size of the sample (Fig. S1). The gel band obtained from the 3′ P-element insertion site of Mz317-Gal4 was extracted, purified and Sanger sequenced using the 3′SPLNK-SEQ (CGG GAC CAC CTT ATG) primer by the Servicios Científico-Técnicos from the University of Oviedo using standard methods (Fig. S2). The sequence generated is available in the NCBI repository (GenBank OP158015).

Data availability
All data supporting the results in this work are available from the corresponding author on request. The Mz317-Gal4 3′ insertion site sequence generated is available in the NCBI repository with the accession number GenBank OP158015 (https:// www. ncbi. nlm. nih. gov/ genba nk/). Not-cropped original images of all the gel electrophoresis are provided in the Supplemental Data.