冈比亚大麻虫在标准的昆虫条件下（26–28°C，75–80％的湿度和12-12-H的光周期）保持。为了获得SOA突变体，我们使用CRISPR-CAS9系统将荧光标记盒（3×P3-Mturquoise2）插入第一个SOA外显子。此外，在荧光标记盒的开头，以及与SOA启动器ATG密码子相对应的位置，包括PHIC31介导的质粒整合的ATTP对接位点，以后允许使用新的SOA副本来挽救突变（请参见下文）。敲入荧光标记盒的设计具有强大的转录终结器和多个终止密码子，可在转录和翻译水平上停止SOA的表达。为此，如前所述，我们在PDSARN Vector36中建立了表达GRNA和修复模板供体质粒。该质粒在AGAP013557 U6启动子的控制下表达了两个GRNA，在SOA中识别目标位点5'-GTCAGCAGCAGCCAGCCTTGATGC-3'和5'-GCATCAAGCTGCTGCTGAC-3'。从SOA基因组序列（每个约1.1 kb长）的同源性区域的5'和3'区域与GRNA靶位点相邻，在该质粒中克隆，侧面是3×P3-Mturquoise标记标记盒。在补充表1中提供了所得基因组插入的序列。将质粒显微注射到大约40-90分钟的gambiae菌株中，该抗体中表达Cas9在YFP标记的Transgene37中的生殖线中表达Cas9。使用尼康SMZ-18双眼显微镜，筛选了蓝色荧光幼虫的存活的注射蚊子的后代，配备了配备有Lumencor Sola Light Light Engine和CFP激发和发射过滤器的蓝色荧光幼虫。回收了几十个mturquoise阳性幼虫， SOA-KI系列是由单个创始人女性建立的。通过PCR和序列验证，敲入合成序列与基因组之间的连接。纯合子和杂合SOA-KI线是由Copas分选的38。为了跟踪几代人的基因型频率的自然动力学，杂合（WT/SOA-KI）线被留下以自然发展> 16代。在每一代人中，对新孵化的新生儿L1幼虫的全部人群进行了COPAS分析，以记录纯合突变体，杂合和WT个体的数量，并通过SOA-KI等位基因中存在的mturquoise Marker的存在和强度得分（WT不荧光）。遗传十字被用来将SOA-KI突变与来自Y Collomosome 39表达GFP的T4性别转基因相结合，从而使COPAS对SOA-KI纯合突变体的全男性或全女性种群进行分类，并控制蚊子幼虫种群，用于生物化学实验中。为了创建SOA-R转基因蚊子线，其中用编码雄性SOA同工型的SOA cDNA序列挽救SOA突变，我们构建了一个质粒，该质粒紧接在全长SOA编码序列之前，其本身就是SV40 3''terminator序列。质粒中包括一个3×p3-dsred荧光标记，作为该SOA救援盒的下游的转基因选择标记（补充表1中提供了救援质粒的序列）。该质粒分别以320和80 ng µl-1的浓度与PHIC31积分酶编码助手质粒质质36共同注射。将整个质粒整合到SOA-KI ATTP位点，将SOA雄性cDNA同工型置于内源性SOA启动子的控制之下。除了CFP之外，还基于DSRED表达选择了转基因蚊子 导致SOA-R转基因线。通过Haut Conseil des Biotechnologies评估了与转基因蚊子的工作，并由Mesri授权（Déclarationd'listirization d'Ogm en MilieuCondinéno。3243no。3243and Synessno。3912）。

　　从培养物出现在第一个pup舞的那一刻开始，通过计算p的外观来评估发育时机。在每个采样时间，将新形成的p子从培养物中取出。

　　在技​​术层2L型笼子中，将小鼠（CD-1菌株）保持在4-5个个体的社会群体中（365×207×140 mm），并具有安全的选择垃圾和筑巢木材，纸张和棉质材料，12-12-H的深光周期，22°C温度和50％±10％的湿度和安全的R04-25 pellets。对于蚊子喂养，根据由Zoletil（42.5 mg kg – 1）的混合物（42.5 mg kg – 1）和rompun（8.5 mg kg – 1）在0.9％NaCl解决方案中，根据法国伦理委员会和法国教育委员会未经教育和研究的协议，在0.9％的NaCl解决方案中处理了雌性CD-1小鼠（> 35 g）。20562–2019050313288887 v.3。我们遵守有关动物使用的所有相关道德法规。

　　Pupae在Trizol中均质（Fisher Scientific，15-596-026）。在添加氯仿并去除水相后，使用制造商的说明手册，将苯酚 - 氯仿相用于DNA。使用La Taq HS聚合酶（Takara，RR042A）进行PCR。PCR产物以1％的三甲酸 - EDTA（TBE）琼脂糖凝胶运行，并使用Chemidoc MP V.3（Bio-Rad）成像。

　　使用Trizol（Fisher Scientific，15-596-026）和Direct-Zol RNA MicroPREP试剂盒（Zymo Research，R2062）提取RNA。对于PUPA样品，与100％乙醇混合后，仅在酚 - 氯仿分离后形成的水相。根据滞留的mRNA准备连接参考指南（2020年6月；文档编号1000000124518 V00），使用Illumina滞留的mRNA准备连接套件进行NGS库制备。对于AG55细胞培养RNA-seq，以1：1,000稀释量制备了100 ng和2μlERCCSpike-Ins（Ambion，44456740）的起始量和2μLERCCSpike-Ins（Ambion，4456740）的文库，并在12个PCR周期中放大。对于PUPA RNA-Seq，以1：100稀释的起始量为1,000 ng和2μLERCCSpike-Ins（Ambion，44456740）制备文库，并在10个PCR周期中放大。在2100生物分析仪（安捷伦技术）上对图书馆进行了高灵敏度DNA，并使用Qubit DSDNA HS分析套件在Qubit 2.0荧光计（Life Technologies）中进行了量化。在NextSeq 500高输出中对汇总样品进行测序，PE的2×73循环加上2×10循环，以读取双重索引。

　　For SOA-KI RNA-seq, the reads were mapped to the ribosomal RNA sequences extracted from the Ensembl AgamP4 genome using the Ensembl AgamP4 annotation (release 48) with STAR (v.2.7.3a) with the following parameters: outFilterMultimapNmax 1000000 outFilterMismatchNoverLmax 0.04 outFilterMismatchNmax 999. Reads mapping将rRNA丢弃，并在下游处理中使用未塑造的读数。对于SOA-R和AG55 RNA-Seq，由于RRNA读取很少，因此对RRNA进行修剪和映射。在所有实验中，读取均使用Ensembl Agamp4注释（第48版）与LNCRNA Annotation40和实验特异性序列（例如SOA-KI或SOA-KI或SOA-KI或SOA-R-R CASSETTE的元素）一起映射到Ensembl Agamp4基因组（第48条）（版本48），以评估AG55实验率的Issportions Informitions;更多信息；使用以下参数使用Star（v.2.7.3a）的Omnibus数据库）：OutFiltermismatchnoverlmax 0.04 Outfiltermismististnmax 999。仅使用唯一的映射读数用于下游分析。使用DeepTools（v.3.1.0）生成了主要比对的覆盖信号轨道（大wig）。使用子读（v.1.6.5）将主要比对分配给特征（v.1.6.5），并与agamp4注释（版本48）与lncRNA annotation40相结合作为参考。使用DESEQ2（V.1.26.0）进行差异表达分析，仅具有FDR的基因< 0.05 were considered as differentially expressed. The visualization of the RNA-seq data of SOA in Anopheles gambiae, A. arabiensis, A. minimus and A. albimanus was obtained using the genome browser tool from Vectorbase (https://vectorbase.org).

　　CUT&Tag was performed as previously described17. In total, 0.4 million cells were used for each reaction. The pupa experiments were performed with flash-frozen tissue samples, which were homogenized in cold PBS and passed through a cell strainer (Corning, 352235). In the initial pupa experiment (WT and SOA-KI male and female pupae), the homogenate was fixed with 0.2% paraformaldehyde (PFA) for 2 min at room temperature. For the SOA-R CUT&Tag, no fixation was applied. The cell culture experiments were all performed on freshly collected cells with a native protocol. The antibodies used are listed in the Supplementary Table 4. We used pA–Tn5 prepared by the IMB Protein Production Core Facility and 15 PCR cycles in the library amplification step. Pooled samples were sequenced on NextSeq 500 High Output, PE for 2×75 cycles plus 2×8 cycles for the dual index read.

　　Reads were trimmed using cutadapt (v.4.0) to remove Illumina adapter sequences and subsequently mapped to the reference genome with bowtie2 (v.2.4.5). For the WT male versus female pupa experiment, we performed an initial analysis to inspect the antibody specificity and therefore removed the multimapping and duplicate reads. We then called peaks using macs2 (v.2.1.2) with the corresponding IgG samples as controls, which identified 139 and 393 filtered peaks in female replicates 1 and 2, respectively, but 1,025, 653, 627 and 808 filtered peaks in males. Because we could not a priori exclude SOA binding to repetitive regions, we then performed a second analysis, in which multimapping and duplicate reads were retained for peak calling using macs2 (v.2.1.2). Note that CUT&Tag fragments can share exact starting and ending positions because the integration sites are affected by DNA accessibility. Therefore, duplicates observed in CUT&Tag are not necessarily a consequence of overamplification by PCR41,42. A greylist was generated on the basis of IgG samples using the R package GreyListChIP (v.1.22.0) and applied for peak filtering in the pupa experiments. This provided 7,742 consensus peaks for downstream analysis with DiffBind (v.3.4) to identify sites that were significantly (FDR < 0.05) differentially bound between samples (results in Supplementary Table 2). Note that the greylist was applied for the pupa datasets and the myb-less experiment in Ag55, whereas no greylist was applied to the long SOA versus empty Ag55 (cell culture) dataset, as this experiment contained almost no background. Background bins instead of library size were used for normalization. Downstream visualization of differentially bound peaks (for example, heatmaps) were generated using deepTools (v.3.5.1). To identify SOA-bound motifs, the sequences of peaks (±200 bp from the summit) with higher binding (FDR < 0.05) in males (pupa) or SOA(1–1265) were extracted using bedtools (v.2.29.2). Peak sequences were then used for motif discovery analysis using MEME-ChIP (MEME v.5.4.1), with the genome sequence as a background. The MEME output was then used in FIMO (v.5.4.1) with default settings and selecting the available metazoan upstream sequences for A. gambiae (AgamP4.34_2019-03-11) or A. aegypti (AaegL3.34_2019-03-11) databases. Overlapping CA motifs identified by FIMO were merged into a single CA motif using ggRanges. For the analysis of repeats, the RepeatMasker annotation was downloaded from https://www.repeatmasker.org/species/anoGam.html, RepeatMasker open-4.0.5-Repeat Library 20140131. Downstream analysis and statistical tests were performed using R studio.

　　ATAC–seq was performed as previously described43 with the following changes. The starting material was flash-frozen pupae. After thawing, whole pupae were homogenized in cold PBS and passed through a cell strainer (Corning, 352235). The cell suspension was counted, and 50,000 cells were used for each reaction. We used 250 ng of Tn5 prepared by the IMB Protein Production Core Facility per reaction and 15 PCR cycles in the library amplification step. Pooled samples were sequenced on NextSeq 500 High Output, PE for 2×75 cycles plus 2×8 cycles for the dual index read.

　　Reads were trimmed using cutadapt (v.4.0) to remove Illumina adapter sequences and subsequently mapped to the reference genome with bowtie2 (v.2.4.5). We excluded multimapping and duplicate reads from downstream analysis. We then called peaks using macs2 (v.2.1.2). Peaks with a length of at least 100 nt were used in downstream analysis with DiffBind (v.3.6.1) to identify sites that were significantly (FDR < 0.05) differentially bound between samples. Coverage signal tracks were generated using deepTools (v.3.5.1). The replicates were merged for visualization in heatmaps by calculating the mean normalized coverage using WiggleTools (v.1.2.8). multiBigwigSummary (Galaxy v.3.5.1.0.0.) was used to calculate the average scores for 20-kb bins on the merged bigwig files visualized in box plots. Heatmaps used to assess the changes in accessibility of SOA bound peaks or genes downregulated in SOA-KI males were generated using deepTools (v.3.5.1).

　　RNA extracted as per the RNA-seq protocol was used for generating cDNA with oligo(dT) as primers. qPCR was performed with FastStart Universal SYBR Green Master (ROX) mix (Roche, 04913850001) in a 7 μl reaction at 300 nM final primer concentration. We used SOA as template and Rp49 as an endogenous control. SOA expressed from the SOA-R cassette was specifically detected with a primer targeting a part of the exogenous SV40 terminator included in the mRNA 3′ UTR. Total SOA mRNA was detected with primers targeting the coding sequence, which enabled comparisons of SOA levels in homozygous SOA-R and WT conditions. Cycling conditions as recommended by the manufacturer were applied. We corrected for primer efficiency using serial dilutions.

　　RT–PCR was conducted using a oneStep Reverse Transcription-PCR kit (Qiagen, 210212) according to the user manual. In this kit, the reaction mixture contains all of the reagents required for both RT and PCR. For each reaction, 2 ng of RNA was used with primers for SOA binding to exons 2 and 3 (rt15 + rt16, Supplementary Table 5). Hence, RT is primed in a gene-specific fashion from the primer in exon 3. S7 was used as a loading control (rt01 + rt02). A total of 33 PCR cycles were used for SOA, 27 cycles for S7. The PCR products were separated on a 2% TBE agarose gel and imaged using ChemiDoc MP V3 (Bio-Rrad). Uncropped gel pictures are provided in Supplementary Fig. 1.

　　The expression cassettes for Ag55 cells were cloned into a pFastBac Dual backbone (Thermo Fisher, 10712024) used for baculovirus generation. Plasmids were generated by Gibson assembly and restriction cloning (details can be provided upon request). The EF1a promoter (approximately 1 kb upstream of the TSS of AGAP007405) was amplified from genomic DNA with primers s047 and s048 (Supplementary Table 5) using LA Taq polymerase (Takara, RR002A). The coding sequence of SOA was amplified from cDNA generated from an adult male RNA sample. Primstar GXL (Takara, R050A) was used to amplify the coding sequence from the start codon to the end, excluding the stop codon. The vector expressing SOA(1–229) was cloned from the vector with full-length SOA coding sequence, as was the vector expressing SOA(112–1265) (myb-less). All constructs contain a C-terminal 2×HA tag followed by a T2A cleavage site and eGFP, which enables assessment of the infection rate.

　　pFastBac vectors with expression cassettes were transposed into the baculoviral genome using chemically competent DH10Bac cells (Thermo Fisher Scientific) according to the manufacturer’s protocol. Preparation of the baculoviral genome, transfection/P0 virus generation and P1 virus amplification were performed as described in the Bac-to-Bac manual (Thermo Fisher Scientific), with the exception of using Cellfectin® II transfection reagent and Sf-900 III serum-free medium (Thermo Fisher Scientific).

　　Ag55 cells provided by M. Adang were cultured in Leibovitz L15 medium with 10% FBS (Gibco, 10270-10,6 lot: 2260092) and 1× penicillin–streptomycin (Gibco, 15140122) at 27 °C, 80% humidity. Ag55 cells were authenticated by RNA-seq. Cells were tested every 6 months for mycoplasma (Mycoalert PLUS Mycoplasma Detection kit, Lonza LT07-701). All tests were negative. For the CUT&Tag experiment, 2 million cells were seeded in a 6-well plate. After 16 h, 600 μl of baculovirus in Sf-900 III serum-free medium was added to the cells. For the RNA-seq experiment, 0.75 million cells were seeded per each well of a 24-well plate. After 16 h, 200 μl of baculovirus in Sf-900 III serum-free medium was added. In both experiments, after 6 h the medium was changed to fresh L15. For the western blotting, 20 million cells were seeded in a 10-cm dish and infected with 6 ml of baculovirus on the next day and the baculovirus was not removed. Cells were collected for further processing 48 h after the addition of the baculovirus.

　　Cells were collected and washed with PBS. The cell pellet was resuspended in hypotonic lysis buffer (25 mM HEPES, pH 7.6, 10 mM NaCl, 5 mM MgCl2, 0.1 mM EDTA and 1× protease inhibitor cocktail) and incubated on ice for 15 min. Next, NP-40 was added to a final concentration of 0.1% and the cells were vortexed for 30 s. The nuclei were pelleted and washed with sucrose buffer (25 mM HEPES, pH 7.6, 2 mM MgCl2, 3 mM CaCl2, 0.3 M sucrose and 1× protease inhibitor cocktail). The nuclear pellet was then resuspended in HMG-K400 buffer (25 mM HEPES, pH 7.6, 2.5 mM MgCl2, 10% glycerol, 0.2% Tween, 400 mM KCl and 1× protease inhibitor cocktail) and rotated for 30 min at 4 °C. After centrifugation, the supernatant was either used directly for western blotting or for IP with the HA antibody. IP was performed by incubating 0.160 mg of nuclear soluble protein extract with 2 μl of HA antibody overnight. The bound SOA–antibody complexes were captured using Protein G dynabeads (1 h at 4 °C) followed by 3 washes in HMGT-K400 buffer. IPs were eluted by incubation in 2× LDS buffer with 200 mM DTT (37 °C, 10 min). For the SOA antibody IP, chromatin extracts from Ag55 cells infected with male SOA(1–1265), female SOA(1–229) or empty baculovirus control, which are all tagged with a C-terminal 2×HA epitope, were prepared. Cells were fixed in 0.1% PFA and nuclei prepared by using a previously published Nexson protocol44. The chromatin was sheared by sonication and diluted into the final IP buffer (0.05% SDS, 125 mM NaCl, 10 mM Tris (pH 8), 1 mM EDTA). Next, 5% of the input was removed and the remaining material was incubated with SOA antibody overnight. The bound SOA–antibody complexes were captured using Protein G dynabeads (1 h at 4 °C) followed by 3 washes in RIPA (25 mM HEPES pH 7.6, 150 mM NaCl, 1 mM EDTA, 1% Triton-X 100, 0.1% SDS, 0.1% DOC and protease inhibitors), 1 wash in LiCl buffer (250 mM LiCl, 10 mM Tris-HCl, 1 mM EDTA, 0.5% NP-40 and 0.5% DOC) and 2 washes in TE buffer. IPs were boiled in 1× Laemmli buffer (95 °C, 10 min).

　　Proteins were separated by 4–12% NuPAGE gradient gels in 1× MOPS buffer. Gels were transferred to a 0.45 µm PVDF membrane in Tris-glycine transfer buffer with 10% methanol (16 h at 60 mA). Membranes were blocked for 1 h in 5% milk in PBS–0.2% Tween, then incubated with primary antibodies (Supplementary Table 4) overnight at 4 °C. For SOA antibody, 5% horse serum was used as a blocking agent. Secondary HRP-coupled antibodies were used at 1:5,000 dilution for 1 h. Blots were developed using Lumi-Light Western Blotting substrate (Roche, 12015200001) and/or SuperSignal West Femto (Thermo Fisher, 34094) and imaged on a ChemiDoc MP V3 (Bio-Rad). Uncropped western blots are provided in Supplementary Fig. 1.

　　The untagged SOA fragments were generated from His6–GST-3C–SOA expression vectors and used for electrophoretic mobility shift assay (EMSA), size-exclusion chromatography coupled to multi-angle light scattering (SEC–MALS) and antibody generation. His6–GST-3C–SOA fragments (1–122, 1–229 and 1–331) were expressed from pET vectors in Escherichia coli (BL21 DE3 codon+) overnight at 18 °C using 1 mM IPTG in LB medium. Cells were lysed in lysis buffer (50 mM Tris-Cl pH 8.0, 800 mM NaCl, 1 mM EDTA, 1 mM DTT, 5% glycerol and EDTA-free complete protease inhibitor cocktail) using a Branson Sonifier 450 and cleared by centrifugation (40,000g, 30 min at 4 °C). Additional 250 mM NaCl was added to the cleared lysates and a PEI-based precipitation of nucleic acids (0.2% w/v polyethylenimine, 40 kDa, pH 7.4) for 5 min at 4 °C was performed, followed by a second round of centrifugation (4,000g, 4 °C, 15 min). Recombinant proteins were affinity-purified from cleared lysates using a NGC Quest Plus FPLC system (Bio-Rad) and a GSTrap HP 5 ml column (Cytiva) following the manufacturer’s protocols. Proteins were digested with 3C protease (1:100 w/w) overnight at 4 °C during dialysis in 50 mM Tris-Cl pH 8.0, 800 mM NaCl, 1 mM DTT and 5% glycerol to cleave off the His6–GST tag. Digested proteins were re-run over the GSTrap HP 5 ml column to absorb out the His6–GST, concentrated using Amicon 15 ml spin concentrators (Merck Millipore) and subjected to gel filtration (Superdex 200 16/60 pg in 25 mM Na-HEPES, 800 mM NaCl, 1 mM DTT and 10% glycerol, pH 7.4). Peak fractions containing the recombinant proteins after gel filtration were pooled, and protein concentration was determined by using absorbance spectroscopy and the respective extinction coefficient at 280 nm before aliquots were flash-frozen in liquid nitrogen and stored at −80 °C. The His6–MBP-tagged SOA fragments and His6–MBP control were used in EMSA and fluorescence polarization (FP) experiments. His6–MBP-tagged SOA fragments and His6–MBP control were expressed from a pET vector in E. coli (BL21-CodonPlus(DE3)-RIL, Agilent) using LB medium and overnight incubation with 0.5 mM IPTG at 18 °C. Cells were lysed in lysis buffer (30 mM Tris-Cl, 500 mM NaCl, 10 mM imidazole, 0.5 mM TCEP, complete protease inhibitors, 2 mM MgCl2 and 150 U ml–1 benzonase, pH 8.0) using a high-pressure homogenizer (constant systems CF1 at 1.9 kBar). The lysate was cleared by centrifugation (40,000g, 4 °C, 30 min) and loaded onto a HisTrap FF 5 ml column (Cytiva) using a NGC Quest Plus FPLC system (Bio-Rad). The column was washed with buffer A (30 mM Tris-Cl, 500 mM NaCl and 10 mM imidazole, pH 8.0), followed by a second wash with buffer A containing 1 M NaCl and a third wash with buffer A containing 25 mM imidazole. Recombinant proteins were eluted by applying a linear gradient of 25–500 mM imidazole (pH 8.0) in buffer A over 15 column volumes. Peak elution fractions were pooled and concentrated using an Amicon 15 ml spin concentrator with 10 kDa cut-off (Merck Millipore). Concentrated proteins were applied to a gel filtration column (Superdex 200 16/60 pg, Cytiva, in 10 mM Na-HEPES pH 7.4, 150 mM NaCl, 1 mM TCEP and 5% glycerol). Peak fractions containing recombinant proteins were pooled and concentrated to 200 µM using an Amicon 15 ml spin concentrator with 10 kDa cut-off. Aliquots of the recombinant proteins were snap-frozen in liquid nitrogen and stored at −80 °C. The recombinant proteins were analysed by SDS–PAGE and visualized by Coomassie staining.

　　Tagless SOA(1–122) was re-buffered in PBS using a PD-10 column (Cytiva) for immunization. Immunization was carried out by Eurogentec using their polyclonal 28-day speedy programme. For epitope purification of the SOA antibody from the serum, 2 ml sulfolink resin (Thermo Fisher Scientific) was covalently conjugated with 3 mg tagless SOA(1–122) according to the manufacturer’s protocol. Next, 10 ml final bleed was incubated with the SOA(1–122)-conjugated sulfolink resin at 4 °C overnight while rotating. After incubation, the resin was washed with PBS containing 0.1% Triton X-100, followed by PBS in a gravity-flow poly-prep column (Bio-Rad). Elution was performed using low pH (100 mM glycine-Cl and 150 mM NaCl, pH 2.3) followed by immediate neutralization of elution fractions with Tris-Cl pH 8.0. The eluted antibody was re-buffered using a PD-10 column (PBS, 0.05% NaN3 and 10% glycerol) and concentrated to 1 mg ml–1 using an Amicon spin-concentrator before flash-freezing in liquid nitrogen and storage at −80 °C.

　　To validate the specificity of the SOA antibody described in this study, we performed western blotting comparing female Ag55 cells ectopically expressing full-length SOA(1–1265), SOA lacking the myb-domain epitope or an empty control. The SOA constructs additionally contained a C-terminal HA-tag. This revealed a specific band present in only full-length, but not the two control conditions (Extended Data Fig. 5a), and two nonspecific bands present in all conditions. Note that we were unable to detect endogenous SOA proteins by western blotting from Ag55 cells or from male/female tissues, which is probably due to the low abundance of the SOA protein. We conducted IP experiments with HA antibody or SOA antibody and detected the captured proteins by western blotting with the other antibody (SOA antibody for HA-IP and HA antibody for SOA-IP, respectively; Extended Data Fig. 5b,c). The specific SOA band detected in the input was also enriched by IP. Furthermore, SOA antibody could not recognize a SOA version lacking the myb domain (amino acids 1–112, the epitope used to raise the antibody), whereas the SOA(1–229) fragment (female isoform) could be successfully detected. We also conducted IP experiments with SOA antibody versus IgG control from male pupal extracts. The bound proteins in this endogenous setup were then identified in an unbiased fashion by mass spectrometry (MS) (Extended Data Fig. 5d,e and Supplementary Table 1). SOA was the only protein not detected in the control and displayed by far the highest enrichment relative to the few contaminants, both in terms of the number of identified unique peptides identified (n = 12, 11, 13 and 12 for the 4 replicates) as well as the intensity. We also validated the specificity of the antibody by CUT&Tag and IF using the SOA-KI loss-of-function mutants as a control. In both cases, the detected signals and peaks vanished (Fig. 4a,b), which directly supports specificity. Last, the CUT&Tag experiment from Ag55 cells expressing HA-tagged SOA(1–1265) was performed in parallel with SOA and HA-tag antibodies. The two profiles (HA antibody, SOA antibody) produced similar profiles (data not shown).

　　The desired amount of protein was diluted into 10 μl of 1× EMSA buffer (20 mM HEPES-KOH (pH 7.5), 100 mM KCl and 0.05% NP-40). GST or MBP was used as a negative control. The protein amounts were 100 fmol (1×) to 12.5 pmol (125-fold excess over DNA). Next, 100 fmol of the DNA probe (601-sequence, 147 bp45 or X-chromosome promoter sequences bound by SOA, 300 bp; Supplementary Table 1) was added, incubated at room temperature for 30 min and subjected to gel electrophoresis (1.6% TBE agarose). DNA was stained with SYBR Safe and detected using a Typhoon FLA9500 gel scanner. The experiment was repeated three times with similar results. Uncropped gel pictures are provided in Supplementary Fig. 1.

　　SEC–MALS measurements were performed at 25 °C in 25 mM HEPES (pH 7.5), 500 mM NaCl and 1 mM DTT as the column buffer using a GE Healthcare Superdex 200 10/300 Increase column on an Agilent 1260 HPLC at a flow rate of 0.5 ml min–1. Loading concentrations were 200 µM for the SOA(1–112) and SOA(1–229) fragments and 11 µM for the SOA(1–331) fragment. Elution was monitored using an Agilent multi-wavelength absorbance detector (data collected at 280 and 260 nm), a Wyatt Heleos II 8+ multi-angle light scattering detector and a Wyatt Optilab differential refractive index detector. The column was equilibrated overnight in the running buffer to obtain stable baseline signals from the detectors before data collection. Inter-detector delay volumes, band-broadening corrections and light-scattering detector normalization were calibrated using an injection of 2 mg ml–1 BSA solution (Thermo Pierce) and standard protocols in ASTRA 8. Weight-averaged molar mass (Mw), elution concentration and mass distributions of the samples were calculated using ASTRA 8 software (Wyatt Technology).

　　To generate dsDNA oligonucleotide substrates, Cy5-labelled ssDNA 20-mers were annealed with reverse-complement 20-mer oligonucleotides at 50 µM in TE buffer by heating to 90 °C for 1 min and subsequent incubation on ice (all oligonucleotides synthesized and HPLC-purified by Integrated DNA Technologies, sequences in Supplementary Table 1). Using a 384-well plate (Corning, low-volume, polystyrene, black), Cy5-labelled ssDNA and dsDNA oligonucleotide substrates (5 nM) were incubated with varying concentrations of His6–MBP-tagged SOA fragments or with a His6–MBP control in a total volume of 20 µl FP buffer (10 mM Na-HEPES pH 7.4, 150 mM NaCl, 1 mM TCEP, 0.1 g l–1 BSA, 5% glycerol and 0.05% Triton X-100). After 10 min of incubation at 20 °C, FP of the Cy5-labelled oligonucleotides were analysed on a Tecan Spark 20M plate reader at 20 °C (excitation wavelength of 625 nm; emission wavelength of 665 nm; gain of 120; flashes of 15; integration time of 40 µs). Normalized FP values were calculated by subtracting the FP value of each oligonucleotide-only measurement from all conditions that contained variable amounts of the respective recombinant protein. The normalized FP values from three independent experiments, including standard deviations, were plotted using GraphPad Prism 8. EC50 values, which serve as a proxy for the binding constant (Kd), were determined by applying a four parameter [agonist] versus response fit with variable slope in GraphPad Prism 8 if applicable.

　　Approximately 0.2 ml (dry volume) of sex-separated pupae were homogenized for each replicate in 0.5 ml of cytoplasm isolation buffer (Cell Signaling Technologies, 9038S) using a handheld homogenizer. After 5 min of incubation on ice, the homogenate was cleaned by spinning through a cell strainer (Corning, 352235) on a FACS tube (500g for 5 min). Cell fractionation of nuclei was continued according to the manual using a Cell Fractionation kit (Cell Signaling Technologies, 9038S). The nuclei were resuspended in 0.125 ml of NIB (250 mM NaCl, 50 mM HEPES, pH 7.6, 0.1% IGEPAL, 10 mM MgCl2, 10% glycerol and protease inhibitors complete, Roche). For the antibody validation experiment, NIB contained 600 mM NaCl. This was sonicated using a Bioruptor Plus, 5 cycles on/off (high), 30 s each followed by 5 min of centrifugation at 12,000g. The supernatant was quantified using Bradford reagent (Avantor PanReac AppliChem, A6932.0250) and 0.4 mg nuclear protein extract used per replicate with n = 5 males and n = 5 female extracts used in total. For the antibody validation experiment, n = 4 male replicates were used for each condition (SOA antibody, IgG control). Per IP and replicate, 20 µl of Protein G dynabeads (Thermo Fisher, 10004D) were washed 2× with NIB, then incubated with 4 µl of SOA antibody (rabbit polyclonal, clone 87) in 40 µl NIB for 45 min on a wheel. This was washed 2× with NIB and resuspended in 40 µl of NIB, which was then added to the nuclear extracts and incubated for 30 min at 4 °C on a wheel. Unbound proteins were removed by three washing steps with 200 µl NIB. Bound proteins eluted by heating beads in 30 µl 1×LDS buffer (Thermo Fisher Scientific) supplemented with 100 mM DTT for 10 min at 70 °C and 1,400 r.p.m. in a thermomixer (Eppendorf). Proteins were subsequently run on a 4–12% NOVEX NuPage gel (Thermo Fisher Scientific) for 8 min at 180 V in 1× MOPS buffer (Thermo Fisher Scientific). Proteins were fixed and stained with 0.25% Coomassie Blue G-250 (Roth) in 10% acetic acid (Sigma)–43% ethanol (Roth). The gel lane was minced and destained with a 50% ethanol–50 mM ammonium bicarbonate (ABC) pH 8.0 solution. Proteins were reduced in 10 mM DTT–50 mM ABC pH 8.0 for 1 h at 56 °C and then alkylated with 50 mM iodoacetamide–50 mM ABC pH 9.0 for 45 min at room temperature in the dark. Proteins were digested with mass-spectrometry-grade trypsin (Sigma) overnight at 37 °C. Peptides were extracted from the gel using twice a mixture of 30% acetonitrile (VWR) and 50 mM ABC pH 8.0 solution followed by two times with pure acetonitrile, which was ultimately evaporated in a concentrator (Eppendorf) and loaded on an activated self-made C18 mesh (AffiniSep) StageTips46.

　　Peptides were separated on a 25 cm self-packed column (New Objective) with 75 µm inner diameter filled with ReproSil-Pur 120 C18-AQ (Dr. Maisch). The EASY-nLC 1000 (Thermo) column was mounted onto a Q Exactive Plus mass spectrometer (Thermo), and peptides were eluted from the column in an optimized 90 min gradient from 2 to 40% acetonitrile–0.1% formic acid solution at a flow rate of 200 nl min−1. The mass spectrometer was operated in a data-dependent acquisition mode with one MS full scan and up to ten MS/MS scans using HCD fragmentation. MS raw data were searched against Anopheles_gambiae.AgamP4.pep.all (15,125 entries) with the Andromeda search engine47 of the MaxQuant software suite (v.1.6.5.0)48. Cys-carbamidomethylation was set as fixed modification and Met-oxidation and protein N-acetylation were considered as variable modifications. Match between run option was activated. Before further processing, protein groups marked with reverse, only identified by site or with fewer than two peptides (one of them unique) were removed.

　　In our initial IF stainings, tissues were dissected and then fixed in 4% formaldehyde in PEM (0.1 M PIPES (pH 6.9), 1 mM EGTA and 1 mM MgCl2) for 20 min and washed three times with PBS. Samples were blocked for 1 h rocking with freshly prepared 0.5% BSA, 0.3% Triton X-100 in 1×PBS solution. The samples were washed with Basilicata-blocking (BB) buffer (0.5% BSA in PBS–0.2% Tween (Sigma Aldrich, P1379)), followed by overnight incubation with primary antibody (anti-SOA, rabbit polyclonal, 1:300 in BB). Samples were washed three times in BB and then stained with a secondary antibody (Alexa fluorophore-labelled goat anti-rabbit, ThermoFisher, A21430, 1:400 in BB). Samples were thoroughly washed with BB, then with 1×PBS–0.2% Tween. For the embryo staining, 19 h AEL-stage embryos were placed in small baskets (Falcon 40 µm cell strainers, 352340) and dechorionated in bleach (4.8% chlorine) for 1–2 min with visual monitoring of chorion dissolution under a binocular microscope. As soon as chorion disappeared, they were rinsed with PBS followed by fixation in PBS, 4% PFA and 0.1% Triton X-100 for 20 min at room temperature. They were then rinsed 3 times with PBS and then stored in methanol at −20 °C. Before IF staining, the black endochorion was then manually peeled off with a needle under a binocular microscope using a Petri dish with a double-sided tape with embryos submerged in 100% methanol. The peeled embryos were transferred using a 1.5 ml pipette into a 1.5 ml Eppendorf tube containing PBS. Blocking and antibody incubations were performed as for the dissected tissues. During the course of the project, we realized that lower PFA concentrations significantly improved the signal-to-noise of the SOA staining; therefore we changed the fixation step in our protocol to 1% PFA for 15 min. We also noted that prolonged incubation with primary antibody (60–72 h) improved signal-to-noise; for embryos prolonged incubation was crucial to obtain SOA staining. For the RNAseA experiment, midguts were dissected in PBS and then rinsed 2× with CSK buffer (10 mM PIPES-KOH, pH 7.0, 100 mM NaCl, 300 mM sucrose and 3 mM MgCl2), then incubated for 10 min in CSK, 0.5% Triton X-100 and 1 mg ml–1 RNaseA (or control). The midguts were then rinsed 2× in CSK buffer. For each condition, 2 midguts (2 replicates) were then put in 0.15 ml TRIzol for RNA isolation to check the effectiveness of the RNase treatment versus control. Meanwhile, the remaining midguts were fixed with 1% PFA in PEM for 15 min at room temperature and stained as per the standard conditions described above. For actinomycin D treatment, the tissues were dissected and put into 0.5 ml of L15 tissue culture medium, 10% FBS and penicillin–streptomycin. Actinomycin D was added to a final concentration of 5 μg ml–1 to half of the samples, the other half was left untreated (control), and both conditions were incubated for 1 h at 26 °C in a tissue culture incubator. The tissues were then fixed in PEM and 1% PFA for 15 min at room temperature and the staining was conducted as described above. As a positive control, we co-stained for phosphorylated RNA Pol2, which has been previously described to increase after actinomycin D treatment49.

　　Fourth instar larva were immobilized on ice for 15–20 min, then they were placed in a drop of 75 mM KCl and the head and abdomen was cut off with an ultrafine dissection scissor and discarded. The thorax was placed in a fresh drop of 75 mM KCl on a glass microscopy slide and the gut and tissues attached to it were gently pulled out with forceps and discarded. The remaining thorax piece containing the imaginal discs and salivary glands was gently opened and placed in a fresh drop of fixative (25% acetic acid, 1% methanol-free PFA in H2O). Imaginal discs and salivary glands immediately turn white and are now easy to spot. They were dissected in approximately 5–7 min under a binocular microscope, attempting to completely remove the fat and cuticle. After 7–8 min, the fixative was removed and a fresh drop of PBS–0.1% Tween containing 1:1,000 of DAPI solution was added. A coverslip was put on the dissected discs and salivary glands and excess solution carefully removed with a Kimtech wipe. The coverslip was gently tapped with the rubber of a pencil while observing squashing under a fluorescent microscope. When spreading was sufficient, the slide was put in liquid nitrogen and the coverslip was flicked off with a razor blade. The slide was then placed in PBS and stored at 4 °C until staining. For the RNA FISH experiment, all solutions described above additionally contained RNasin Ribonuclease inhibitor (Promega N2511) at 1:1,000 dilution.

　　The slides were incubated in a coplin jar containing PBS and 0.4% Triton X-100 for 30 min at room temperature on an orbital shaker set at 220 r.p.m. The slides were rinsed 2× with PBS and 0.1% Tween. The slides were then incubated on the orbital shaker with blocking buffer (PBS, 0.1% Tween, 0.2% BSA and 5% horse serum; filtered) for 30–60 min at room temperature. The slides were placed in a wet chamber, and incubation with primary antibody in blocking buffer (0.25 ml solution, slide covered with Parafilm) was conducted overnight at 4 °C. The slides were washed in a coplin jar on the orbital shaker 3× in PBS and 0.2% Tween. Secondary antibodies were incubated for 1–2 h in a wet chamber at room temperature (0.25 ml of solution, slide covered with Parafilm). The slides were washed in a coplin jar on the orbital shaker 2× in PBS and 0.2% Tween followed by a 15 min incubation with PBS, 0.1% Tween and DAPI (1:1,000) in a wet chamber as for the antibodies. The slides were rinsed with PBS and then mounted with Prolong Gold.

　　Polytene squashes were prepared as described above. RNA FISH was performed according to the manufacturer’s protocol for IF followed by smFISH, referred to as the sequential protocol. PBS was prepared from a 5× sterile PBS solution with DEPC water and 1 μl RNAseIn per 50 ml of 1× buffer was added. Slides with squashes were briefly rinsed 2× in PBS, 0.1% Tween and RNAseIn for 10 min and 1× with PBS. Primary antibody in PBS incubation was performed 60–72 h at 4°C in a humidified chamber. Excess antibody was washed out 3× with PBS followed by secondary antibody incubation in PBS for at least 3 h. Unbound secondary was washed out 2× in PBS and the slide was then crosslinked in 4% PFA–PBS for 10 min at room temperature. Excess of fixative was removed using PBS washes and then the smFISH protocol was started using 1× wash buffer A (SMF-WA1-60-BS, LGC Biosearch Technologies) supplemented with 10% formamide. This was followed by hybridization in Stellaris RNA FISH hybridization buffer (SMF-WA1-60-BS, LGC Biosearch Technologies) supplemented with 10% with formamide containing 125 nM probe mix targeting the introns of the X-linked gene act5c (AGAP000651, sequences in Supplementary Table 1), which was incubated overnight in a humidified chamber at 37 °C. Excess probe was removed by two washes with wash buffer A, 30 min each at 37 °C, followed by a brief wash in wash buffer B (SMF-WB1-20-BS, LGC Biosearch Technologies). Slides were mounted in Vectashield vibrance with DAPI (H-1800, Vector Laboratories) and imaged after 1 h using Visiscope Microscope, ×63 water objective.

　　The protocol was based on the spatial CUT&Tag50 with the following modifications. pA–Tn5 produced by the IMB Protein Production Core Facility was loaded with pre-annealed oligonucleotides Tn5MErev, Tn5ME-A-ATTO488 and Tn5ME-B-ATTO488. Adult male midguts were dissected, fixed with 0.2% PFA in PEM buffer with RNAseIN (1:1,000) at room temperature for 5 min. The fixation step was quenched with 2.5 M glycine (1:20). After quenching, the midguts were washed 2 times with the CUT&Tag wash buffer (20 mM HEPES pH 7.6, 150 mM NaCl, 0.5 mM spermidine and 1× protease inhibitor cocktail) and rinsed briefly with RNAse-free water. The midguts were then incubated for 5 min at room temperature in permeabilization buffer (0.1% NP40 and 0.05% digitonin in wash buffer) and washed once with the NP40–digitonin wash buffer (0.01% NP40 and 0.05% digitonin in wash buffer). Subsequently, the midguts were incubated overnight with the SOA antibody (1:100 dilution) at 4 °C on a Nutator in the antibody buffer (2 mM EDTA and 0.1% BSA in NP40–digitonin wash buffer). The next day, the midguts were rinsed once with NP40–digitonin wash buffer, then incubated on the Nutator for 1 h at room temperature with the secondary antibody (1:100 dilution of F(ab′)2-goat anti-rabbit IgG (H+L) cross-adsorbed secondary antibody, Alexa Fluor-555; 555A21430 ThermoFisher) in the same buffer. This was followed by a rinse with the NP40–digitonin wash buffer. Next, the pA–Tn5 complex pre-loaded with fluorescently labelled oligonucleotides was added into Dig-300 buffer (20 mM HEPES pH 7.6, 300 mM NaCl, 0.5 mM spermidine, 0.05% digitonin and 1× protease inhibitor cocktail) at a final concentration of 31 nM and incubated for 1 h at room temperature on the Nutator. After a 5-min wash with the Dig-300 buffer, the midguts were incubated in tagmentation buffer (10 mM MgCl2 in Dig-300 buffer) for 1 h at 37 °C. The tagmentation step was stopped by adding EDTA to final concentration of 40 mM and incubating for 5 min on the Nutator. The midguts were finally washed with 1× NEBuffer 3.1 and then stained with DAPI.

　　Slides were mounted using ProLong Gold Antifade mountant with DAPI (P36935, Thermo Fisher Scientific), unless otherwise stated, and imaged using a fluorescence spinning disc confocal microscope, VisiScope 5 Elements (Visitron Systems), which is based on a Ti-2E (Nikon) stand and equipped with a spinning disc unit (CSU-W1, 50 μm pinhole; Yokogawa). The set-up was controlled using VisiView 5.0 software, and images were acquired with a ×100/1.49 NA oil-immersion objective (CFI Apo SR TIRF ×100, Nikon) or ×60/1.2 NA water-immersion (CFI Plan Apo VC60x WI) and a sCMOS camera (BSI; Photometrics). 3D stacks of images were recorded for each sample. Confocal imaging was performed using a Stellaris 8 Falcon (Leica Microsystems) confocal system equipped with white light laser. Images (1,552 × 1,552 pixel format, 0.93 pixel size) were acquired using a HC PL APO CS2 ×63/1.40 NA oil-immersion lens, and fluorescence was detected using a detector HyD S for DAPI (emission band 427–460 nm), HyD X for Alexa488 (500–545 nm) and HyD R for Alexa555 (560–730 nm). Tissue images were acquired through 87 slices at 200-nm step intervals using a line accumulation of 3 times. 3D view of the z-stacks and image processing were obtained using Imaris software (v.9.9.1). The IF stainings were replicated in at least four independent experiments.

　　Our results indicated that the SOA+ allele speeds up male development by about 4 h. To investigate the evolutionary implications of such a progression of development, we used the standard one-locus-two-alleles model of viability selection, with different viabilities in males and females22. In this model, the relative viability of the three genotypes SOA–/SOA–, SOA+/SOA– and SOA+/SOA+ is 1, 1 + hm  × sm and 1 + sm, respectively, in males and 1, 1 – hf × sf and 1 – sf, respectively, in females. Here sm is the selection differential in favour of the SOA+ allele in males, whereas sf is the selection differential against SOA+ in females. The factors hm and hf denote the degree of dominance of the SOA+ allele. Throughout, we assumed that SOA+ is dominant in males (hm = 1) and recessive in females (hf = 0) based on the general finding that selectively favoured alleles tend to be dominant in each sex51. However, we also considered other dominance values, and they led to the same conclusion (persistence of the SOA+ allele at considerable frequencies for a wide range of selection coefficients) as long as hm >0。

　　我们对SM的估计是基于这样的理由：较短的发育时间有利于成年的生存。根据专门针对蚊子蚊子生命周期量身定制的人群模型52，男性的每日生存概率为0.9。因此，将开发加速4小时（相当于一天的六分之一），因此对应于0.95/6/0.9 = 1.0177的生存益处。因此，我们假设SOA+的男性的发育进展转化为选择系数SM = 0.0177。由于这是一个粗略的估计值，有时使用了不同的生存概率53，我们还考虑了SM的其他值，范围从0.005到0.05。我们还考虑了女性的选择系数SF，范围从0到0.05。在图5H中，SF在5,000代中设置为零，这对应于替代剪接（去除女性中SOA+的负适应性效应）的假设已经发展。

　　DNA和蛋白质序列是从vectorbase检索的。使用Clustal Omega创建蛋白质和DNA比对。SOA域的成对百分比相似性是在Jalview（v.2.11.2.3）中获得的。用espript可视化对齐。使用Vectorbase的Biomart工具获得了1：1直系同源物的列表。SOA基因座，其在其他物种中的同义区域及其瘫痪分析是从vectorbase获得的。使用Mega软件（V.7.0）进行系统发育和进化距离计算。使用Adobe Illustrator和Adobe Photoshop（2021版）组装数字。

　　使用了以下资源：CASADAPT（https：//github.com/marcelm/cutadapt）；Bowtie2（https://github.com/benlangmead/bowtie2）;MacS2（https://github.com/macs3-project/macs）;wiggletools（https://github.com/ensembl/wiggletools）;meme（https://meme-suite.org/meme/）;gviz（https://bioconductor.org/packages/release/bioc/html/gviz.html）;Star（https://github.com/alexdobin/star）;diffbind（https://bioconductor.org/packages/diffbind/）;DeepTools2（https://deeptools.readthedocs.io/en/latest/）;IGV（https://software.baradinstitute.org/software/igv/）;R（https://www.r-project.org）;deseq2（http://bioconductor.org/packages/deseq2/）;vectorbase（https://vectorbase.org/vectorbase/app）;Clustal Omega（https://www.ebi.ac.uk/tools/msa/clustalo/）;espript（https://espript.ibcp.fr/espript/espript/）;核定位信号预测（https://nls-mapper.aib.keio.acio.ac.jp/cgi-bin/nls_mapper_form.cgi）;iupred2（https://iupred2a.elte.hu/）;Cys2HIS2锌指蛋白（http://zf.princeton.edu/）的DNA结合位点预测指标。

　　所有统计数据均使用R Studio计算。在小提琴图中，中心线代表中位数，小提琴的形状代表基础数据的分布。对于所有小提琴图，使用双面Wilcoxon秩和测试获得了P值（扩展数据图7J，3D，2G和10H，K），并在图4D和扩展数据图中进行了额外的Bonferroni校正。7d9e和10j，l。在盒子图中，将盒子分为两个部分的线代表中位数，框底部和顶部边缘代表四分位数（IQRS; 0.25th至0.75th Quartile（Q1 – q3）），晶须代表Q1-- 1.5×IQR（底部），Q3+1.5+1.5+1.5+1.5×IQR（TOP）。条形图代表均值代表重复的数据点的平均值。在FDR低于0.05时，结果被认为是显着的。NA，未进行分析。对于所有饼图，P值是通过单方面的Fisher的精确测试来获得X染色体的过度代表的。为此，我们将SOA的峰值与在所有染色体臂上分布的峰数量相等（切割和TAG，图2C，3F和5E），或分析在上调和下调组中X-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-C-CH-CLING基因在所有基因上与所有Chromosos and arm and arns and chromos arms-sse-se-se-se-se-se rna-se相比。在扩展的数据图8b中，将X染色体和常染色体上含CA重复的启动子的过度代表与所有X连锁和常染色体基因进行了比较。为了评分图4F和扩展数据中的发育延迟图9L，通过对数秩检验进行分层数据（Mantel – Haenszel检验）获得P值。在图5G中，通过双面t检验获得了Benjamini-Hochberg校正的P值，并在基因型之间进行成对比较。图形传说中提供了更多详细信息。进一步的数据， 提供补充表1-3。复制免疫染色的结果与以下结果相似：图1G和扩展数据图5G实验（WT雄性，女性）进行了7次，每个实验至少从每个性别的n = 5个成年人（生物学重复）中解剖的组织；扩展数据图。5H和6A实验（聚南汉，幼虫组织）进行了3次，每次至少有2张幻灯片，为此，每张载玻片包含至少n = 4个幼虫（生物学重复）的组织。扩展数据图5i，J（胚胎）进行了两次，每个数据至少n = 30个胚胎（生物学重复）；图2a实验（SOA IF和CO-FISH）进行了2次，每张2个幻灯片。每个载玻片都包含至少n = 4个成年人的组织（每个实验8个生物学重复）；扩展数据图6B实验（切割和参见）是在n = 1成人（生物学重复）中进行的一次组织进行的。扩展数据图7H实验（RNase A）进行了2次，每个实验至少从至少n = 5个成年人（生物学重复）中解剖的组织；扩展数据图7i实验（放线菌素D）是用至少n = 5个成年人（生物学重复）解剖的组织一次进行的。图4B实验（SOA-KI）进行了2次，每个实验至少从每个基因型的n = 5个成年人中解剖（生物学重复）；扩展数据图10b实验（SOA-R IF和Co-Fish）一次用2个载玻片进行一次，每张载玻片包含至少n = 4个幼虫（生物重复）解剖的组织；图5C和扩展数据图10C实验（SOA-R）进行了2次，每个实验至少是从每个性别的n = 5个成年人中解剖的（生物学重复）。

　　有关研究设计的更多信息可在与本文有关的自然投资组合报告摘要中获得。