Article Figures and data Abstract eLife digest Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract We generated two new genetic tools to efficiently tag genes in Drosophila. The first, Double Header (DH) utilizes intronic MiMIC/CRIMIC insertions to generate artificial exons for GFP mediated protein trapping or T2A-GAL4 gene trapping in vivo based on Cre recombinase to avoid embryo injections. DH significantly increases integration efficiency compared to previous strategies and faithfully reports the expression pattern of genes and proteins. The second technique targets genes lacking coding introns using a two-step cassette exchange. First, we replace the endogenous gene with an excisable compact dominant marker using CRISPR making a null allele. Second, the insertion is replaced with a protein::tag cassette. This sequential manipulation allows the generation of numerous tagged alleles or insertion of other DNA fragments that facilitates multiple downstream applications. Both techniques allow precise gene manipulation and facilitate detection of gene expression, protein localization and assessment of protein function, as well as numerous other applications. https://doi.org/10.7554/eLife.38709.001 eLife digest Organisms have tens of thousands of genes, but finding out exactly what they all do is one of the greatest challenges of modern genetics. To understand a gene’s job, it’s necessary to find out what gene is active in which tissue, where their proteins are located within the cell, and what happens when the sequence of a gene is altered or removed. This multi-step process of ‘annotating’ genes can be challenging in practice. One common approach is to make use of a DNA pattern called a MiMIC and insert it in a specific part of the gene called an intron. A tag for a protein that glows green under the microscope can then be added to a MiMIC to help visualize where and when the protein is being expressed. MiMICs can also be used to integrate a system called T2A-GAL4, which typically creates a severe mutation in the gene and allows to track the timing of when and where the gene is expressed. This helps to discover the role of the gene in cells and tissues. However, a problem with this approach is that when either the protein tag or the T2A-GAL4 system is added, half of the time they point into the wrong direction. This is because each DNA strand is read in one direction only. Now, Li-Kroeger et al. created a so-called ‘Double Header’ system, which includes T2A-GAL4 coding in one direction and the protein tag in the other. Therefore, when the system integrates, there will always be one tag pointing in the correct direction. This makes the system twice as efficient. Not all genes have introns though. To access genes that do not contain introns, Li-Kroeger et al. developed another system, which uses the genome editing tool CRISPR-Cas9 to introduce a different kind of visible marker. Here, the whole gene is typically removed and replaced by a visible marker, which can then be replaced by any DNA, including protein tags and the T2A-GAL4 system. With these approaches, all genes in the fruit fly can now be targeted. The systems perform several tasks, including detecting gene activity and the location of proteins in the cell, and analyzing the role of the protein. The findings will be relevant to researchers interested in fruit fly genetics and cell function. https://doi.org/10.7554/eLife.38709.002 Introduction Comprehensive gene annotation is a central challenge in the post-genomic era. Drosophila melanogaster offers more sophisticated genetic approaches and tools to assess gene function and expression than other multicellular model organisms (Bier et al., 2018; Cox et al., 2017; Germani et al., 2018; Heigwer et al., 2018; Kanca et al., 2017; Kaufman, 2017; Simpson and Looger, 2018). A versatile tool used for functional gene annotation in Drosophila is MiMIC, a Minos-based transposon that integrates a Swappable Integration Cassette (SIC) in the genome (Venken et al., 2011a). MiMIC SICs contain a cassette nested between two attP sites that can be exchanged with any DNA sequence flanked with attB sites through Recombinase Mediated Cassette Exchange (RMCE) by ΦC31 integrase. When a MiMIC is integrated in an intron of a gene flanked on both sides by coding exons (hereafter referred to as a coding intron), the SIC can easily be exchanged with an artificial exon that encodes Splice Acceptor (SA)-(GGS)4 linker-EGFP-FIAsH tag-StrepII tag-TEV protease cleavage site-3XFlag-(GGS)4 linker-Splice Donor (SD) (abbreviated as GFP tag) (Venken et al., 2011a; Nagarkar-Jaiswal et al., 2015a; Nagarkar-Jaiswal et al., 2015b). The GFP-tagged endogenous proteins report the subcellular localization of the gene product and are functional in ~75% of tested genes (Nagarkar-Jaiswal et al., 2015a). Importantly functional GFP-protein traps can be used for multiple assays. These include chromatin Immunoprecipitation (ChIP) of transcription factors (Nègre et al., 2011), Immunoprecipitation (IP)-Mass Spectroscopy (MS) (David-Morrison et al., 2016; Neumüller et al., 2012; Yoon et al., 2017), rapid conditional removal of gene products (Caussinus et al., 2011; Lee et al., 2018b; Nagarkar-Jaiswal et al., 2015a; Neumüller et al., 2012; Wissel et al., 2016) and sequestration of tagged proteins (Harmansa et al., 2017; Harmansa et al., 2015). Hence, tagging an endogenous gene with GFP enables numerous applications to dissect gene function. The SIC in MiMICs can also be replaced by an artificial exon that encodes SA-T2A-GAL4-polyA signal (abbreviated as T2A-GAL4) (Diao et al., 2015; Gnerer et al., 2015; Lee et al., 2018a). T2A-GAL4 creates a mutant allele by truncating the protein at the insertion site but also expresses GAL4 with the spatial-temporal dynamics of the targeted gene. Hence, T2A-GAL4 facilitates the replacement of the gene of interest with fly or human UAS-cDNAs (Bellen and Yamamoto, 2015; Şentürk and Bellen, 2018; Wangler et al., 2017; Lee et al., 2018a), allowing one to assess putative disease-associated variants and permitting structure-function analysis of the protein of interest. Moreover, these gene-specific GAL4 stocks can be used to drive a variety of UAS constructs to further identify and probe the function of the cells expressing the gene using UAS-Fluorescent proteins or numerous other UAS constructs (Venken et al., 2011b). This is especially useful for genes that are not abundantly expressed, providing a means to amplify the signal, as GAL4 drives overexpression of the UAS transgenes (Diao et al., 2015; Lee et al., 2018a). In summary, MiMIC applications allow the acquisition of valuable data about the function of the gene as well as the cells in which the gene is expressed. Given the usefulness of MiMICs, the Drosophila Gene Disruption Project (GDP) (http://flypush.imgen.bcm.tmc.edu/pscreen) has generated and mapped 17,500 MiMIC insertion stocks (Nagarkar-Jaiswal et al., 2015a; Venken et al., 2011a). This collection includes insertions within introns for ~1860 genes, each of which can be converted to a GFP-tagged protein trap and/or a T2A-GAL4 gene trap (Nagarkar-Jaiswal et al., 2015a; Nagarkar-Jaiswal et al., 2015b; Lee et al., 2018a). However, we needed to develop a complementary strategy to generate resources for genes that do not have a MiMIC randomly inserted within a coding intron. To that end, we recently developed CRIMIC (CRISPR mediated Integration Cassette); a Cas9/CRISPR Homology Directed Repair (HDR) mediated approach that integrates a modified SIC (attP-FRT-SA-T2A-GAL4-polyA-FRT-attP) in a coding intron of choice. This approach greatly expands the number of genes that can be tagged using MiMIC-like technology from 1860 to ~6000 (Lee et al., 2018a) allowing about forty percent of Drosophila protein coding genes to be targeted with SICs. RMCE cassettes can either be injected into embryos as part of a circular plasmid or can be circularized in vivo from an initial insertion locus in the genome through Cre/loxP or Flp/FRT mediated recombination (Diao et al., 2015; Nagarkar-Jaiswal et al., 2015b). Importantly, RMCE cassettes can replace a SIC in either orientation with equal probability due to inverted symmetric attP sequences. Therefore 50% of the insertions are inserted in the opposite orientation of transcription and will not be included in the mature mRNA. Hence, only half of all successful exchange events will result in protein or gene trap lines. Here, we show that by combining GFP-protein traps and T2A-GAL4 gene traps in a single RMCE construct, named Double Header (DH), we significantly increased the number of productive RMCE events for MiMIC/CRIMIC containing genes to generate protein or gene trap alleles. Importantly, we expand the ability to target SICs into genes regardless of the presence of introns to allow access to virtually any gene in the fly genome based on CRISPR/Cas9-mediated HDR. This provides a means to create robust null alleles with simple screening, and to convert the SIC insertion using any DNA, creating scarless modifications to facilitate numerous downstream applications. Results Double Header (DH) improves the tagging rate of MiMIC containing genes SICs in coding introns can be converted into GFP-protein traps or T2A-GAL4 gene traps through RMCE. However, because RMCE of SICs in MiMICs and CRIMICs can occur in either orientation, only one out of two events produces a tag that is incorporated in the gene product. Moreover, each RMCE experiment generates only a protein trap or a gene trap, requiring two independent injections or crosses to generate both reagents. In order to reduce the effort to generate these genetic tools we engineered Double Header (DH), a construct that combines the two key RMCE cassettes to replace intronic MiMICs and CRIMICs: (1) SA-T2A-GAL4-polyA (DHT2A-GAL4) which generates a gene trap that expresses GAL4 in the expression domain of the gene while typically inactivating gene function, in one orientation and (2) SA-GFP-SD (DHGFP) which generates a GFP-protein trap, in the opposite orientation. Hence, insertion of DH via RMCE in a MiMIC in either orientation should result in two valuable reagents. This compound RMCE cassette is flanked by two inverted ϕC31 attB sites in a vector backbone that contains other features including mini-white as shown in Figure 1A. The presence of white in the plasmid backbone allows for a counter selection against the integration of the whole plasmid when incorporated into a white- background, ensuring that only the DNA between the attB sites integrates. Figure 1 with 2 supplements see all Download asset Open asset Double Header optimizes RMCE outcome of MiMICs. (A) Schematics of the Double Header construct and RMCE outcomes. Double Header constructs contain a Splice Acceptor (SA)- super folder GFP-FlAsH-StrepII-TEV-3xFlag (EGFP) – Splice Donor (SD) in one orientation and a SA-T2A-GAL4-polyA in the other orientation. Insertion in the GFP orientation results in GFP protein trap whereas insertion in the T2A-GAL4 orientation results in T2A-GAL4 gene trap. (B) Double Header injection statistics. https://doi.org/10.7554/eLife.38709.003 The artificial exons that are integrated into MiMIC or CRIMIC sites need to be in frame with the preceding Splice Donor (SD) to create a functional tag. Because exon/intron boundaries can occur at any one of three positions in a codon (phase 0,+1,+2), we generated three different DH plasmids (Sequences can be found in Supplementary file 1). Each construct contains the same codon phase for the two modules. Given that the RMCE cassette is about 4.8 kB, or about 1.5 times the size of the single T2A-GAL4 cassette, we anticipated a lower integration rate. We tested the integration efficacy by injecting 30 strains carrying a coding intronic MiMIC insertion. On average,~400 embryos were injected for each MiMIC line together with a plasmid that encodes the ΦC31 integrase. We screened for the loss of yellow+ in the progeny of the injected animals, as MiMICs carry the yellow+ marker (Figure 1—figure supplement 1A). We did not observe the integration of white indicating that a single RMCE event leading to the integration of the entire plasmid is rare. For 16 out of 30 MiMICs, we were able to isolate yellow- flies that carry a DH integration (Figure 1B; Figure 1—figure supplement 1B). We determined the orientation of DH inserts through single fly PCR (Figure 1—figure supplement 2) and determined the orientation of the cassette in 47 out of 72 DH RMCE events. For 6 MiMICs we obtained both orientations and for the other 10 MiMICs we obtained one or the other orientation (Figure 1B; Figure 1—figure supplement 1B). Hence, we generated a total of 22 new reagents, increasing the overall rate of productive RMCE events by injection (Nagarkar-Jaiswal et al., 2015a). Note that the integration efficiency of DH construct by injection is lower than the smaller SICs:~50% versus~66% (Nagarkar-Jaiswal et al., 2015a). However, the number of productive events increases tagging efficacy somewhat as every successful event produces a useful line: 74% versus 66%. In vivo double header mobilization based on transgenes In an effort to avoid embryo injections and to increase integration efficiency of DH, we developed an in vivo RMCE strategy using genetic crosses similar to Trojan exons developed for T2A-GAL4 (Diao et al., 2015). We integrated the same constructs as in Figure 1A, one for each reading frame, in the genome through P-element mediated transformation. These insertions serve as jump-starter constructs because the RMCE cassettes in these transgenes can be excised from their initial landing sites by expressing Cre recombinase in germ line cells (Figure 2). The crossing scheme is outlined in Figure 2-figure supplement 1. Figure 3A. We generated jump-starter insertions in second and third chromosomes for all three possible phases of DH and generated double balanced stocks for subsequent crosses. We tested the efficacy of integration by crosses for DH for third chromosome MiMICs. For 12 out of 13 MiMICs tested we obtained integration of DH. We determined the orientation of DH for 48 out of 102 insertions by PCR. The inconclusive insertions either showed no PCR amplification in one end or both ends of the MiMIC (48/102) or in rare cases conflicting PCR amplification that indicates integration in both orientations (6/102). Interestingly 44/54 inconclusive inserts happened in only two of the MiMICs, indicating locus specific issues. For 6 out of 12 MiMICs we obtained both orientations and for six we obtained one or the other, resulting in 18 tagged genes (Figure 2—figure supplement 1). Hence, the genetic strategy is about twice as efficient (18 constructs for 13 crossed MiMICs versus 22 constructs for 30 injected MiMICs) as the injection strategy in generating RMCE events and requires significantly less effort. Figure 2 with 1 supplement see all Download asset Open asset Double Header integration through crosses facilitates RMCE. (A) Schematics of the Double Header transgene mobilization in vivo. Double Header transgenes contain loxP sites that can be used to mobilize the RMCE cassette in vivo, without the need for injection. (B) Double header crossing statistics. https://doi.org/10.7554/eLife.38709.006 Figure 3 with 1 supplement see all Download asset Open asset Examples of gene expression patterns obtained by Double Header. Each MiMIC, MI01487, MI05208, MI06794, MI06872, MI08614, MI11741 and MI15073, was converted to either T2A-GAL4 protein traps or GFP protein traps by Double Header insertion. The expression in the larval CNS is shown with either T2A-GAL4 > UAS-mCD8::GFP or GFP-tag (GFP and mCD8::GFP, green). The affected genes are labelled above. Scale bar: 50 µm. https://doi.org/10.7554/eLife.38709.008 Double Header reports the expression pattern of the tagged gene and protein We proceeded to test whether DH functions as expected. We determined the expression patterns of genes tagged in both orientations in third instar larval brain and adult brains for MI01487 (kibra), MI05208 [5-HT2B(5-hydroxytryptamine receptor 2B)], MI06794 [Lgr4(Leucine-rich repeat-containing G protein-coupled receptor 4)], MI06872 (CG34383), MI08614 [Dgk (Diacyl glycerol kinase)], MI11741 (CG12206) and MI15073 (CG9132) (Figure 3). As we selected a few MiMICs that were previously tagged with T2A-GAL4 by the Gene Disruption Project as positive controls (Lee et al., 2018a; Diao et al., 2015) (MI01487 (kibra), MI06794 (Lgr4), MI06872 (CG34383), MI08614 (Dgk), and MI11741 (CG12206), we were able to compare expression patterns obtained by DHT2A-GAL4 to expression patterns obtained by single T2A-GAL4 (http://flypush.imgen.bcm.tmc.edu/pscreen/rmce/). In all cases the expression pattern was very similar to what was previously reported (Figure 3) (Lee et al., 2018a). In addition, tracheal expression of CG12206 is consistent with a previous report (Chandran et al., 2014) and the 5HT2BT2A-GAL4 expression pattern in the adult brain matches an independently generated T2A-GAL4 (Gnerer et al., 2015) (Figure 3—figure supplement 1). In all cases, the DHGFP insertions show consistent patterns of expression in third instar larval brains, albeit at much lower levels than the DHT2A-GAL4 insertions at the same MiMIC site (Figure 3). Note that in adult brains almost no signal of DHGFP was detected, in agreement with previous observations, with the exception of MI15073 which shows ubiquitous expression, (Diao et al., 2015; Lee et al., 2018a) (Figure 3—figure supplement 1). These results indicate that neither the size nor the design of DH alters the functionality or expression patterns of the tagged genes and proteins. As the GFP protein traps should be able to report the subcellular localization of the tagged protein we turned to tissues where subcellular localization and specific cell expression is easily assessed. We therefore dissected and stained egg chambers with anti-GFP. We were easily able to visualize the GFP tagged proteins in the seven protein traps previously examined (Figure 4). The tagged proteins are shown in green and the nuclei are stained with DAPI in red. Kibra is detected in somatic follicle cell cytoplasm, including some migratory border cells. 5HT2B is expressed in both somatic and germline cells including the oocyte. Lgr4 is localized to germline nurse cell nuclei and is enriched in the oocyte anterior-dorsal and ventral cortex. CG34383 is mostly present in follicle cells, especially in their apical domain. Dgk is observed in nurse cell and follicle cell nuclei as well as their cytoplasm. CG12206 is quite enriched in the cytoplasm of centripetal cells and CG9131 is present in both germ cells and follicle cells and enriched in polar cells. In summary, GFP protein tagging with DH can be used to determine the cellular and subcellular localization of tagged proteins. Figure 4 Download asset Open asset Examples of cellular expression patterns and subcellular localization of tagged proteins in egg chambers at stage 9 and 10. Double header GFP protein traps of MIMIC lines shown in Figure 3 were dissected and ovaries were stained with anti-GFP antibody (green) and DAPI (red). Arrowheads indicate features that are referred to in the text; border cells for kibra; nurse cells, follicle cels and oocytes for 5HT2B; GFP is broadly expressed and distributed for Lgr4; note the apical enrichment in follicle cells in CG34383; nuclear and cytoplasmic staining in nurse cells and follicle cells are observed in Dgk; centripedal cells cytoplasm is mostly labeled in CG12206; broad expression and localization with pole cell enrichment in CG9132. Scale bar: 50 µm. https://doi.org/10.7554/eLife.38709.010 A compact cassette to target a SIC to intron-less genes To tag a gene containing a MiMIC/CRIMIC, the SIC should be integrated within a suitable coding intron, leaving about 50–60% of all Drosophila genes that encode proteins inaccessible. Targeting genes without introns by directly fusing tags in the proper reading frame has been very difficult because HDR is much less efficient than non-homologous end joining (NHEJ) (Gratz et al., 2014). Hence, expanding the range of targetable genes requires precise, seamless genome editing. Gene editing using HDR is well suited to modifying genes without introducing unwanted changes (Bier et al., 2018). HDR repairs double stranded DNA breaks using a donor template that contains two homology regions, each typically about 1000 nucleotides, which flank the desired changes. Recombination on either side of the break replaces the regions with the template, precisely modifying the locus. We therefore developed a novel SIC compatible with HDR (Figure 5;Figure 5—figure supplements 1 and 2) that could be targeted to loci regardless of the presence of introns. Figure 5 with 2 supplements see all Download asset Open asset Schematic of a two-step system for scarless gene editing. (A) In step 1, a cassette containing a dominant marker flanked by nucleotides GG and CC replaces an endogenous locus via Homology Directed Repair (HDR) following double strand breaks produced by Cas9 cleavage (marked by red arrowheads). The removal of the intervening sequence between the Cas9 cut sites alters the sgRNA target sequences (underlined) preventing cleavage of the donor construct or the modified DNA. Screening for the dominant marker facilitates identification of CRISPR gene editing events while the flanking nucleotides GG (boxed inset) and CC create novel Cas9 target sites, allowing subsequent excision. (B) In step two the insert is removed and replaced with any DNA via a second round of HDR with new sgRNA sequences, facilitating the scarless insertion of any DNA sequence desirable. https://doi.org/10.7554/eLife.38709.011 To make a SIC that is HDR compatible, three features are important: (1) a dominant marker for screening that is compact for ease of insertion via HDR (Li et al., 2014), (2) a method to insert the SIC in the desired location, and (3) a strategy to remove the SIC for replacement with the desired end product. To design a compact marker that is compatible with Golden Gate cloning we focused on the yellow gene which has well characterized enhancers (Geyer and Corces, 1987). We identified a 575 nucleotide regulatory region that when fused to the promoter and yellow coding sequence creates a 2.9 kilobase cassette that reliably drives expression only in the wing (Figure 5—figure supplement 2). We refer to this cassette as ywing2+. To enable targeting ywing2+ into precise locations in the genome, we first define a region (or gene) of interest (ROI) flanked by two Cas9 target sites comprised of a 20 nucleotide guide sequence and an NGG PAM (Jinek et al., 2012; Sternberg et al., 2014) (Figure 5A). We then design a HDR donor template with ywing2+ flanked by homology regions. As the donor template removes part of the Cas9 target sequences neither the donor cassette nor the final product are cleaved upon HDR. Injecting the donor template along with sgRNA expression plasmids into embryos carrying a germline-specific source of Cas9, followed by screening offspring for yellow+ wings provides a straightforward method to generate robust null alleles for the gene. Lastly, to make the cassette removable, we flanked the SIC with the nucleotides ‘GG’ and ‘CC’ upstream and downstream of the ywing2+ marker, respectively. Upon insertion, this creates two novel Cas9 target sites that are not present in the endogenous sequence (box inset of Figure 5A), which can be used to remove the inserted cassette for final replacement via a second round of HDR. Finally, to facilitate cloning, the ywing2+ cassette was made compatible with Golden Gate assembly (Engler et al., 2009). We also generated templates for designing replacement HDR constructs containing GFP and mCherry for protein tags or T2A-Gal4 that are compatible with Golden Gate assembly (Figure 5—figure supplement 1). The ywing2+ SIC efficiently replaces genomic loci We tested the efficacy of replacing the coding sequence of 10 loci with ywing2+ (Table 1). Nine out of ten injections led to successful integration of the cassette. We injected an average of ~500 embryos for each gene and recovered 1 to 6 independent founder lines for a total of 22 insertion events or ~2 founder animals/gene. Sanger sequencing confirmed the correct insertion of all but one HDR event, in which the entire plasmid backbone had been integrated. As shown in Table 1, seven insertions are homozygous lethal. To test whether the lethality was specific to the removal of the targeted gene, we attempted rescue of the lethality with a genomic duplication of the locus for four genes (Nmnat (Nicotinamide mononucleotide adenylyltransferase), CG13390, Med27 (Mediator complex subunit 27), and CG11679, and tested for failure to complement molecularly defined deletions for two genes (ubiquilin and Nmnat) (Zhai et al., 2006). In all cases, lethality mapped to the targeted locus showing that no second-site lethal mutations were induced in these lines. For the gene almondex (amx), the ywing2+ insertion produced flies that were female sterile. Female sterility was previously observed for amx and a genomic fragment previously reported to rescue female sterility likewise rescued this phenotype in amxΔCDS,ywing2+(Cohorts for Heart and Aging Research in Genomic Epidemiology consortium et al., 2016). For the gene Stub1 (STIP1 homology and U-box containing protein 1), four positive lines were recovered; two are homozygous lethal while two are viable and fertile. Sanger sequencing confirmed the correct insertion of the cassette in all four lines, suggesting that the gene is not essential. Hence, the lethality is either caused by off-target cleavage events or the presence of a floating lethal mutation in the original strain. Thus while off-target cleavage may have occurred, this evidence suggests that it is not common, in agreement with what we have observed when we use CRIMIC (Lee et al., 2018a). In summary, we created null alleles for nine genes and show that the ywing2+ knock in cassette inserted precisely based on Sanger sequencing. Table 1 Summary statistics for cassette knock-in experiments https://doi.org/10.7554/eLife.38709.014 ConstructGenotype injected:No. independent positive lines obtainedLethalityRescue of lethality/failure to complementywing2+ ΔNmnaty1 M{nos-Cas9.P}ZH-2A w*6lethalGenomic Fragment (Zhai et al., 2006)/NmnatΔ4790–1ywing2+ ΔStub1y1 M{nos-Cas9.P}ZH-2A w*4Viable/Fertile*NDywing2+ ΔUbqnyw iso#6; +/+; attP2(y-){nos-Cas9}2lethalFails to complement Ubqn1ywing2+ ΔItp-r83Ay1 M{nos-Cas9.P}ZH-2A w*1lethalNDywing2+ ΔCG18769y1 M{nos-Cas9.P}ZH-2A w*2lethalNDywing2+ ΔCG13390y1 M{nos-Cas9.P}ZH-2A w*2lethalRescued by Genomic Fragment (this study)ywing2+ ΔMed27yw iso#6; +/+; attP2(y-){nos-Cas9}1lethalRescued by Genomic Fragment (this study)ywing2+ ΔCG11679yw iso#6; +/+; attP2(y-){nos-Cas9}2lethalRescued by genomic duplication BSC Dp(1:3) 304Ywing2+ Δrhoy1 M{nos-Cas9.P}ZH-2A w*0NANAywing2+ Δamxyw iso#6; +/+; attP2(y-){nos-Cas9}2Female sterileRescued by Genomic Fragment *two of four lines Step two: removal of ywing2+ allows ‘scarless’ modification of endogenous loci The ywing2+ cassette is designed to introduce two new gRNA target sites upon replacing the endogenous locus. These newly introduced gRNA target sites can now be used for the removal of the cassette via CRISPR/Cas9 mediated HDR and replacement with the desired DNA sequence. We attempted to incorporate protein tags for five genes (Table 2). We successfully incorporated tags for Nmnat, Stub1, CG11679 and Med27 but failed for amx. We tagged Nmnat and Stub1 with GFP, and CG11679 and Med27 with Flag tags (see Mat. and Meth.). Internally GFP-tagged Nmnat (Nmnat::GFP::Nmnat) and C-terminally Flag tagged CG11679 (CG11679::Flag) reverted the lethality of the ywing2+ knock in allele and hence produced functional proteins. However, the C-terminal Flag-tagged Med27 (Med27::Flag) is recessive pupal lethal, similar to the ywing2+ knock in allele, suggesting that the C-terminal Flag tag disrupts protein function. Because the loss of Stub1 (Table 1) does not result in an overt phenotype, we cannot determine if Stub1::GFP is functional but Sanger sequencing showed that the replacement of ywing2+ with Stub1::GFP happened precisely. Taken together, the data indicate that Cas9 mediated cassette replacement occurred correctly for four out of five genes. Table 2 Summary statistics for cassette swapping experiments https://doi.org/10.7554/eLife.38709.015 ConstructInjected genotype:No. embryos injectedNo. fertile adultsNo vials with y- flies% of y- flies confirmed positiveNmnat:GFP:Nmnat wt #1y1 M{nos-Cas9.P}ZH-2A w*;+; ywing2+ ΔNmnat/TM6B514746%Nmnat:GFP:Nmnat wt #2y1 M{nos-Cas9.P}ZH-2A w*;+; ywing2+ ΔNmnat/TM6B60716521%Nmnat:GFP:NmnatW129G #1y1 M{nos-Cas9.P}ZH-2A w*;+; ywing2+ ΔNmnat/TM6B6530--Nmnat:GFP:NmnatW129G #2y1 M{nos-Cas9.P}ZH-2A w*;3 KB NMNAT GRC; ywing2+ ΔNmnat41831355%Nmnat:GFP:NmnatΔ251…257y1 M{nos-Cas9.P}ZH-2A w*;3 KB NMNAT GRC; ywing2+ ΔNmnat496291124%Nmnat:GFP:NmnatC344S, C345Sy1 M{nos-Cas9.P}ZH-2A w*;3 KB NMNAT GRC; ywing2+ ΔNmnat386301214%Stub1:GFPy1 M{nos-Cas9.P}ZH-2A
Gastric papillary adenocarcinoma (GPA), a well-differentiated gastric adenocarcinoma, is associated with a worse prognosis compared to other differentiated gastric adenocarcinomas. Therefore, there is an urgent need to characterize its endoscopic manifestations for guiding biopsy site selection and achieving accurate diagnosis.
De novo truncations in Interferon Regulatory Factor 2 Binding Protein Like ( IRF2BPL ) lead to severe childhood-onset neurodegenerative disorders. To determine how loss of IRF2BPL causes neural dysfunction, we examined its function in Drosophila and zebrafish. Overexpression of either IRF2BPL or Pits , the Drosophila ortholog, represses Wnt transcription in flies. In contrast, neuronal depletion of Pits leads to increased wingless ( wg ) levels in the brain and is associated with axonal loss, whereas inhibition of Wg signaling is neuroprotective. Moreover, increased neuronal expression of wg in flies is sufficient to cause age-dependent axonal loss, similar to reduction of Pits. Loss of irf2bpl in zebrafish also causes neurological defects with an associated increase in wnt1 transcription and downstream signaling. WNT1 is also increased in patient-derived astrocytes, and pharmacological inhibition of Wnt suppresses the neurological phenotypes. Last, IRF2BPL and the Wnt antagonist, CKIα, physically and genetically interact, showing that IRF2BPL and CkIα antagonize Wnt transcription and signaling.
Neurotransmission is a tightly regulated Ca2+-dependent process. Upon Ca2+ influx, Synaptotagmin1 (Syt1) promotes fusion of synaptic vesicles (SVs) with the plasma membrane. This requires regulation at multiple levels, but the role of metabolites in SV release is unclear. Here, we uncover a role for isocitrate dehydrogenase 3a (idh3a), a Krebs cycle enzyme, in neurotransmission. Loss of idh3a leads to a reduction of the metabolite, alpha-ketoglutarate (αKG), causing defects in synaptic transmission similar to the loss of syt1. Supplementing idh3a flies with αKG suppresses these defects through an ATP or neurotransmitter-independent mechanism. Indeed, αKG, but not glutamate, enhances Syt1-dependent fusion in a reconstitution assay. αKG promotes interaction between the C2-domains of Syt1 and phospholipids. The data reveal conserved metabolic regulation of synaptic transmission via αKG. Our studies provide a synaptic role for αKG, a metabolite that has been proposed as a treatment for aging and neurodegenerative disorders.
Infantile neuroaxonal dystrophy (INAD) is caused by recessive variants in PLA2G6 and is a lethal pediatric neurodegenerative disorder. Loss of the Drosophila homolog of PLA2G6, leads to ceramide accumulation, lysosome expansion, and mitochondrial defects. Here, we report that retromer function, ceramide metabolism, the endolysosomal pathway, and mitochondrial morphology are affected in INAD patient-derived neurons. We show that in INAD mouse models, the same features are affected in Purkinje cells, arguing that the neuropathological mechanisms are evolutionary conserved and that these features can be used as biomarkers. We tested 20 drugs that target these pathways and found that Ambroxol, Desipramine, Azoramide, and Genistein alleviate neurodegenerative phenotypes in INAD flies and INAD patient-derived neural progenitor cells. We also develop an AAV-based gene therapy approach that delays neurodegeneration and prolongs lifespan in an INAD mouse model.
Abstract The Interferon Regulatory Factor 2 Binding Protein Like ( IRF2BPL ) gene encodes a member of the IRF2BP family of transcriptional regulators. Currently the biological function of this gene is obscure, and the gene has not been associated with a Mendelian disease. Here we describe seven individuals affected with neurological symptoms who carry damaging heterozygous variants in IRF2BPL. Five cases carrying nonsense variants in IRF2BPL resulting in a premature stop codon display severe neurodevelopmental regression, hypotonia, progressive ataxia, seizures, and a lack of coordination. Two additional individuals, both with missense variants, display global developmental delay and seizures and a relatively milder phenotype than those with nonsense alleles. The bioinformatics signature for IRF2BPL based on population genomics is consistent with a gene that is intolerant to variation. We show that the IRF2BPL ortholog in the fruit fly, called pits ( protein interacting with Ttk69 and Sin3A ), is broadly expressed including the nervous system. Complete loss of pits is lethal early in development, whereas partial knock-down with RNA interference in neurons leads to neurodegeneration, revealing requirement for this gene in proper neuronal function and maintenance. The nonsense variants in IRF2BPL identified in patients behave as severe loss-of-function alleles in this model organism, while ectopic expression of the missense variants leads to a range of phenotypes. Taken together, IRF2BPL and pits are required in the nervous system in humans and flies, and their loss leads to a range of neurological phenotypes in both species.
Abstract Cdk8 in Drosophila is the orthologue of vertebrate CDK8 and CDK19 . These proteins have been shown to modulate transcriptional control by RNA polymerase II. We found that neuronal loss of Cdk8 severely reduces fly lifespan and causes bang sensitivity. Remarkably, these defects can be rescued by expression of human CDK19, found in the cytoplasm of neurons, suggesting a non-nuclear function of CDK19/Cdk8. Here we show that Cdk8 plays a critical role in the cytoplasm, with its loss causing elongated mitochondria in both muscles and neurons. We find that endogenous GFP-tagged Cdk8 can be found in both the cytoplasm and nucleus. We show that Cdk8 promotes the phosphorylation of Drp1 at S616, a protein required for mitochondrial fission. Interestingly, Pink1, a mitochondrial kinase implicated in Parkinson’s disease, also phosphorylates Drp1 at the same residue. Indeed, overexpression of Cdk8 significantly suppresses the phenotypes observed in flies with low levels of Pink1, including elevated levels of ROS, mitochondrial dysmorphology, and behavioral defects. In summary, we propose that Pink1 and Cdk8 perform similar functions to promote Drp1-mediated fission.
One of the most fundamental changes in cell morphology is the ingression of a plasma membrane furrow. The Drosophila embryo undergoes several cycles of rapid furrow ingression during early development that culminates in the formation of an epithelial sheet. Previous studies have demonstrated the requirement for intracellular trafficking pathways in furrow ingression; however, the pathways that link compartmental behaviors with cortical furrow ingression events have remained unclear. Here, we show that Rab8 has striking dynamic behaviors in vivo. As furrows ingress, cytoplasmic Rab8 puncta are depleted and Rab8 accumulates at the plasma membrane in a location that coincides with known regions of directed membrane addition. We additionally use CRISPR/Cas9 technology to N-terminally tag Rab8, which is then used to address both endogenous localization and function. Endogenous Rab8 displays partial coincidence with Rab11 and the Golgi, and this colocalization is enriched during the fast phase of cellularization. When Rab8 function is disrupted, furrow formation in the early embryo is completely abolished. We also demonstrate that Rab8 behaviors require the function of the exocyst complex subunit Sec5 as well as the recycling endosome Rab11. Active, GTP-locked Rab8 is primarily associated with dynamic membrane compartments and the plasma membrane, while GDP-locked Rab8 forms large cytoplasmic aggregates. These studies suggest a model in which active Rab8 populations direct furrow ingression by guiding the targeted delivery of cytoplasmic membrane stores to the cell surface through exocyst tethering complex interactions.
Abstract In most eukaryotic cells fatty acid synthesis occurs in the cytoplasm as well as in mitochondria. However, the relative contribution of mitochondrial fatty acid synthesis (mtFAS) to the cellular lipidome of metazoans is ill-defined. Hence, we studied the function of the fly Mitochondria enoyl CoA reductase (Mecr), the enzyme required for the last step of mtFAS. Loss of mecr causes lethality while neuronal loss leads to progressive neurological defects. We observe an elevated level of ceramides, a defect in Fe-S cluster biogenesis and increased iron levels in mecr mutants. Reducing the levels of either iron or ceramide suppresses the neurodegenerative phenotypes indicating that increased ceramides and iron metabolism are interrelated and play an important role in the pathogenesis. Mutations in human MECR cause pediatric-onset neurodegeneration and patient-derived fibroblasts display similar elevated ceramide levels and impaired iron homeostasis. In summary, this study shows an as-yet-unidentified role of mecr/MECR in ceramide and iron metabolism providing a mechanistic link between mtFAS and neurodegeneration.
