Mapping and CRISPR homology-directed repair of a recessive white eye mutation in Plodia moths

Summary The pantry moth Plodia interpunctella is a worldwide pest of stored food products and a promising laboratory model system for lepidopteran functional genomics. Here we describe efficient methods for precise genome editing in this insect. A spontaneous recessive white-eyed phenotype maps to a frameshift deletion (c.737delC) in the white gene. CRISPR NHEJ mutagenesis of white replicates this phenotype with high rates of somatic biallelic knockout. G0 individuals with mutant clones on both eyes produced 100% mutant progeny, making white an ideal marker for co-conversion when targeting other genes. CRISPR HDR experiments corrected c.737delC and reverted white eyes to a pigmented state in 37% of G0 mosaic adults. These repaired alleles showed practical rates of germline transmission in backcrosses, demonstrating the potential of the technique for precise genome editing. Plodia offers a promising avenue for research in this taxon because of its lab-ready features, egg injectability, and editability.


INTRODUCTION
Emerging model organisms in functional genetics are opening new opportunities to study gene-to-phenotype relationships across the tree of life. Ideally, in addition to practical life history traits that facilitate their laboratory maintenance, animals that are tailored to genetic investigation will also be suitable for genetic transformation or editing, and harbor recessive phenotypes that can be used as selection markers for longterm line maintenance. The domesticated silkworm Bombyx mori fulfills these criteria and has been a flagship model organism for insect biology, with a long history of use for scientific inquiry, pioneering contributions in genomics, hundreds of mapped mutations, and rapid advances in genome editing (Goldsmith et al., 2005;Ma et al., 2019;Meng et al., 2017;Xu and O'Brochta, 2015). However, the demanding feeding, space, and handling requirements of this large-sized insect limit its widespread adoption by new laboratories interested in the biology of Lepidoptera, i.e., butterflies and moths. Lepidoptera is a large insect order that comprises 160,000 species, accounting for about 10% of all living species named to date (Kristensen et al., 2007;Beccaloni et al., 2012). As it includes important agricultural pests, ecosystem service providers, and model systems in conservation biology, ecology, and evolutionary biology, there is a growing need for developing robust and tractable model systems for functional genomics in this order.
The indianmeal moth, Plodia interpunctella (hereafter Plodia; abbr. Pi), also known as the pantry moth, is a promising emerging model system with existing genome and transcriptome assemblies (Harrison et al., 2012;Roberts et al., 2020;Tang et al., 2017) and life history traits that make it suitable for the long-term maintenance of inbred lines. Much like Tribolium, Plodia is a worldwide pest of stored food products and exhibits convenient laboratory features, including a short life cycle (25 days at 28 C), ease of culture at high density on a low-cost diet (Silhacek and Miller, 1972), inbreeding resistance (Bartlett et al., 2018), and high fecundity (Mbata, 1985). Stock keeping requires minimal human intervention because individuals enter diapause under overcrowding, short photoperiod, or low temperatures (Bell, 1994). In sharp contrast with most other lepidopteran pests, Plodia feeds on a dry diet rather than on fresh plant tissues, and under optimal conditions that avoid mold, there is no requirement for food replacement or waste removal over an entire life cycle, and each stock requires 15 min of work per month for basic maintenance. Egg fertilization and mass-laying are stimulated by CO 2 narcosis, which usually generates dozens of eggs per min for immediate injection, within the time frame of the first cell divisions (Bossin et al., 2007;Dyby and Silhacek, 1997). Unmated females and males can also be isolated before adult emergence for controlled matings and crosses, allowing controlled genetic crossing. Taken together, these properties make Plodia a robust organism for routine germline modification, stock keeping, and isogenic or transgenic line sharing in Lepidoptera.
The white gene encodes a transmembrane protein of the ATP binding cassette (ABC) transporter superfamily (Mackenzie et al., 1999), with conserved functions in the transport of ommochrome and pterin pigment precursors across insects (Figon and Casas, 2019;Grubbs et al., 2015;Tatematsu et al., 2011). Because insects universally use one of these two pigment classes as filtering pigments in their compound eyes (Summers et al., 1982), white loss-of-function mutations generate eye color defects that are easy to screen visually (Ge et al., 2016), and played a foundational role in the early history of genetics (Green, 1996). These features have made white a conventional target for developing targeted mutagenesis methods in insects (Bai et al., 2019;Heu et al., 2020;Li et al., 2020;Liu et al., 2021;Ohde et al., 2018;Sim et al., 2019), though it should be noted that its knockout is recessive-lethal in a subset of tested species (Khan et al., 2017;Reding and Pick, 2020).
Recently, Shirk described an efficient method for targeted mutagenesis using CRISPR procedures that knocked out the white gene in P. interpunctella (Shirk, 2021). Following embryonic injections of Cas9/ sgRNA duplexes targeting the coding position 242 of white, the author obtained G 0 individuals with mosaic white eyes because of nonhomologous end-joining (NHEJ), resulting in biallelic knockouts caused by frameshift mutations in the white exon 1. They then outcrossed Piw-242 G 0 mosaic females to males from the laboratory strain Piw-, which shows autosomal recessive inheritance of a white-eye phenotype. F 1 offspring that inherited both a Piw-242 and Piwwere white-eyed, demonstrating the two types of alleles do not complement, and suggesting Piwis a spontaneous loss-of-function mutation at the white locus itself. In this study, we followed up on these results with a focus on the Piw-spontaneous mutation, mapped this recessive marker to the region of chromosome 10 that includes the white gene, identified a one bp frameshift mutation in the white exon five causing the recessive-white phenotype, and used CRISPR editing with homology-directed repair (HDR) to rescue this mutation to a wild-type state. We also include detailed protocols for the maintenance and genome editing of Plodia stocks. Together, these studies enable Plodia as a tractable lepidopteran model organism for functional genetics via precise genome editing using the CRISPR/Cas9 system.

Introgression of a spontaneous recessive-white marker in a genome strain
We performed crosses to identify a chromosomal location for the recessive white eye marker. We used the ''Dundee'' (UK) inbred strain as a wild-type background (Pi_Dun), to take advantage of its reference genome, available under the name Plodia_dt in NCBI Assembly (Roberts et al., 2020). We used the Piw-inbred strain (Florida, US), isolated in 1986 from a spontaneous mutation at an unknown locus (Shirk, 2021), as a donor for a recessive whiteeye mutation. The wchromosome was introgressed into the Dundee background using two backcrosses (w +/+ x w +/-) separated by five w -/x w +/incrosses ( Figure 1A), resulting in a hybrid strain we dub here ''Foggy Bottom'' (abbr. Pi_Fog). Because crossing-over only occurs in male meiosis in Lepidoptera (Traut et al., 2007), a heterozygous male was used in each cross to foster recombination on the wchromosome to restrict the size of the introgressed region around the mutation, and following this introgression scheme, we inbred Pi_Fog wild-type (WT) and w -/individuals separately to generate two homozygous inbred lines of each eye color phenotypes. Both phenotypic stocks have since been maintained in the lab for >20 generations without decline in viability or fertility. As in the parental Piwstrain, Pi_Fog homozygous individuals for the introgressed w -/mutation differ from the WT state by completely depigmented compound eyes in adults, and a lack of reddish pigmentation in the larval cuticle ( Figures 1B and 1C). The testicular masses, normally clearly visible through the cuticle of male fifth instar larvae, are also depigmented ( Figures 1C and 1D).
Mapping the w-mutation reveals a one bp deletion causing a frameshift of the white gene In order to map the white eye mutation, we sequenced the genome of one Piwindividual, one browneyed Pi_Fog, and four white-eyed Pi_Fog individuals. We aligned the resulting reads, along with the Pi_ Dun genome, to the chromosomal genome assembly of the US strain ''Savannah'' generated by the Ag100Pest consortium (Childers et al., 2021). Differentiation between the white-eyed and brown-eyed individuals as measured with F ST was low except for on a portion of chromosome 10, where a large interval of differentially fixed SNPs were shared between Piwand Pi_Fog w -/individuals, but not with the Pi_ Fog w +/+ individual or with the Pi_Dun individual (Figures 2A and 2B). This introgressed wgenetic interval resulted in a large haplotype that encompasses many genes, including the pigmentation gene candidates, namely cinnabar, white, and scarlet, a situation similar to the mapping of eye color mutants in Bombyx (Tatematsu et al., 2011).  Figure 2C). Then, we examined coding SNPs in alignments of white and brown eyed sequencing reads on the chromosome 10 interval. Along with seven missense mutations ( Figure 2D), we identified the c.737delC mutation in the exon five of white as causing a coding frameshift (p.Pro246HisfsX24) ( Figure 3A). The 2,058 bp coding white sequence comprises 13 exons and encodes a 686 amino acid protein. The frameshift mutation is predicted to cause the loss of key components of White protein (Michalovich et al., 2002), such as all transmembrane domains, and the cytoplasmic Nucleotide Binding Domains that interact with ATP ( Figures 3B and 3C). The high conservation of the amino sequence following the deletion was also illustrated through amino acid alignment among Lepidoptera and holometabolous insect outgroup species, and also highlights the functional importance of the mistranslated region following the p.Pro246HisfsX24 frameshift ( Figure 3D). In summary, the c.737delC mutation is null and likely causal for the recessive white phenotypes derived from the Piwstrain.
CRISPR knockout of white exon five reveals efficient germline editing Recent work described efficient CRISPR somatic mutagenesis of the white gene based on a sgRNA target in exon 1 (Shirk, 2021). Here, we targeted the WT allele of exon five using a sgRNA overlapping with the c.737C residue that is deleted in the wstrain. We aimed to measure the somatic expressivity (rate of iScience Article biallelic G 0 mutants) and germline transmission rates of CRISPR-induced mutations, effectively serving as a proxy for assessing the efficiency of NHEJ-induced modifications.
Out of 755 WT Plodia eggs injected, 143 survived and grew into adulthood, of which 113 moths (79%) revealed eye color phenotypes ( Figure 4 and Table 1). Mosaic phenotypes, including incomplete depigmentation of either or both eyes and unilateral absence of pigmentation, were observed in 41 G 0 crispants (28%). Eye color development in Plodia pupae takes place between the P3-P9 pupal stages (Zimowska et al., 1991), and CRISPR knockout (KO) phenotypes are visible as early as P4. Interestingly, we found that upon freezing and thawing adults for imaging, mutant patches in mosaic eyes acquire a light bronze color because of pigment diffusion from adjacent wild-type clones, likely following the rupture of cell membranes in the ommatidia within 20 min after thawing ( Figure S1). Complete and bilateral absence of pigmentation lacking any sign of pigment diffusion (''full KO'') was observed in 72 moths (51% of surviving adults), suggesting that the injection procedure, within 1 h after laying, can generate biallelic null mutations spanning large fractions of the soma. Together, these data highlight the high efficiency of the CRISPR procedures for generating biallelic somatic knockouts.
To assess the KO efficiency in the germline, G 0 mutants with bilateral white eyes were outcrossed to homozygous recessive mutant moths (w -/-). Fifteen out of 16 attempted crosses were fertile and produced a total of 1,177 white-eyed G 1 progeny (100%), without any brown-eyed progeny (Table 2). This shows that G 0 crispants with little somatic mosaicism, as inferred from bilateral mutant phenotypes, indeed carried 100% loss-of-function alleles in their germline. These data thus indicate the high efficiency of CRISPR iScience Article procedures for generating biallelic knockouts, and that the production of germline mutations can be greatly facilitated by selecting founders in which mosaicism is minimal.

Homology-directed repair mediates efficient rescue of the one bp white deletion
To test the potential of precise CRISPR genome editing and HDR in Plodia, we sought to rescue the one bp deletion at c.737delC, based on the assumption that it is the only variant causing the white eye phenotype in the spontaneous w -/strains. We predicted that repaired alleles would be dominant, allowing a rapid quantification of the efficiency of the technique. We used a single-stranded oligodeoxynucleotide (ssODN) as a repair donor template as employed in other organisms (Gratz et al., 2013;Okamoto et al., 2019). The sgRNA and ssODN were designed such that the double-strand break (DSB) site is close to the intended edit ( Figure 5A), with the ssODN made antiparallel to the nontarget (NT) strand. In addition, the ssODN was asymmetric with 40 nt and 33 nt 5 0 and 3 0 homology arms, flanked the intended C insertion at iScience Article c.737delC, and incorporated a silent mutation preventing secondary cutting of repaired alleles. Coinjecting this donor with Cas9:sgRNA heteroduplex targeting the c.737delC allele in w -/homozygotes resulted in efficient dominant rescues of the white-eye phenotype (Tables 3 and 4). Out of 1,014 eggs injected, a total of 120 adult moths emerged, from which 45 moths (38%) exhibited a mosaic phenotype consisting of often large, brown pigmented eye clones on one or both eyes ( Figure 5B). To assess the rate of germline transformation of corrected alleles, we backcrossed G 0 mosaic mutants to the homozygous w -/stock. Because the WT allele is dominant, the proportion of G 1 progeny with WT eye color reflects the proportion of corrected WT alleles in the G 0 germline, which gives an estimate of the germline transformation efficiency using our CRISPR HDR approach. Ten out of 15 attempted crosses produced progenies. Out of these 10 crosses, five produced G 1 WT individuals at ratios of 11.5-77.3% in each progeny. Sanger sequencing followed by chromatogram parsing of the heterozygous signal (Hill et al., 2014) confirmed the presence of a wallele and of an HDR-repaired allele in these samples, with the C insertion and the two synonymous nucleotide changes intended to prevent secondary cuts.

Knockouts of the sex-linked gene Blos2 are deleterious
Although we expect the white autosomal recessive marker to be convenient for many genetic manipulation experiments in the future, we also explored the possibility to generate a sex-linked mutation with a phenotypic effect that is convenient to identify. To do this, we used CRISPR NHEJ targeting the gene Biogenesis of lysosome-related organelles complex one subunit 2 (Blos2). Blos2 is located on the lepidopteran Z sex chromosome and mediates the formation of pigment-containing organelles, with viable knockout alleles generating translucent larval epidermis in several moth species (Cheli et al., 2010;Fujii et al., 2010;Lee et al., 2018;Nakade et al., 2014;Zhu et al., 2020). CRISPR mutagenesis induced G 0 mosaic phenotypes in the larval integument, with mutant portions lacking the characteristic pink pigmentation observed in iScience Article WT individuals ( Figure S3A). Surviving pupae showed transient depigmentation clones in the eye before melanization ( Figure S3B), at a stage where eyes are normally brown (Zimowska et al., 1991). This phenotype was not visible past 60% of pupal development (P8 stage) because of the successful deposition of a late dark melanin pigment, suggesting that Blos2 is only necessary for the accumulation of an early eye pigment that is deposited at the mid-pupal stages (20-50% pupal development). The G 0 egg hatching rate of 14% (203 hatchlings out of 1441 egg injections) was comparable to white KO injections (19%). However, we observed a low survival of the Blos2 G 0 mutant larvae, with only 38% (77/203) of larvae reaching the pupal stage, compared to survival rates above 98% in white G 0 mutants or in uninjected stock in identical conditions. G 0 mutants with the largest pupal eye mutant clones failed to hatch as adults, and as a matter of fact, we failed to obtain G 1 phenotypes from several incrosses, possibly because the parents that survived did not carry mutant alleles in their germline. From these observations, we extrapolate that Blos2 null mutations are deleterious in Plodia, and thus that they are not a viable phenotypic marker of the Z chromosome over several generations.  iScience Article color at the larval stage and eye color at the adult stage. In the first effort to knockout eye color genes in Plodia using CRISPR/Cas9 study (Shirk, 2021), the first exon of the white locus was targeted to produce a heritable null mutation. Here, we investigated the genetic basis of a spontaneous eye color mutant phenotype in Plodia by whole-genome sequencing and demonstrated a successful workflow for functional study through precise genome editing. Specifically, CRISPR-mediated NHEJ knockout of white resulted in both biallelic somatic editing and germline transmission at relatively high efficiency. Last, we showed that HDR knock-in with a ssODN donor efficiently corrected a one bp deletion causing recessive white eyes, again with high efficiency. These data, together with streamlined and safe rearing procedures, lay the foundation for further use of Plodia in functional genomics.

DISCUSSION
We propose that the depigmentation of white-eyed Pi_Fog w -/makes it ideal for the addition of dominant markers, such as fluorescent proteins driven by the 3xP3 eye and glia promoter that is known to work in Plodia (Bossin et al., 2007). The wild-type Pi_Fog could be used for CRISPR experiments where the white knockout sgRNA is coinjected to select founders with large clones, a co-conversion strategy already used in Drosophila (Ge et al., 2016). Indeed, our knockout data indicated that 100% (N = 15) of G 0 founders with bilateral, non-mosaic white eyes because of biallelic KO, generated 100% offspring (N = 1,177). In other words, white eye phenotypes provide a suitable marker for identifying founders that carry germline edits, and we suggest this co-conversion strategy will be instrumental for establishing stable genome edited lines where it is difficult to screen for phenotypes. The Pi_Dun genome reference is readily accessible online on the NCBI Assembly server under the name Pi_dt (Roberts et al., 2020), facilitating experimental design. Additional considerations in sgRNA design can further facilitate genome editing. First, targeting a sequence overlapping with a restriction site makes genotyping straightforward ( Figure S8). Second, several reports suggest an improvement in Cas9 activity with sgRNA designs terminating in G-3 0 or GG-3' (Doench et al., 2014;Farboud and Meyer, 2015). Finally, in this article, a single experimenter was able to inject hundreds of eggs less than 60 min after egg-laying (AEL) in 2-3 h injection sessions. It is noteworthy that two experimenters, one collecting and aligning eggs, and the other injecting and sealing eggs, can inject eggs within 25 min AEL, which should reduce mosaicism and further increase germline transmission rates if needed. iScience Article Our CRISPR dominant rescue of the recessive white c.737delC mutation revealed efficient HDR-based single-base pair edits, showing ssODN donors can be routinely used to introduce small scarless edits in the Plodia genome. Asymmetric designs using homology arms ranging from 30 nt to 90 nt are generally recommended for such applications (Yoshimi, 2016), and verification of the edits is necessary as the technique can lead to imperfect repairs (Boel, 2018). In this article, we based our donor template design based on previous optimizations to single-base substitution performed in mammalian cell lines (Okamoto et al., 2019), using an ssODN located at or near the 5 0 side of PAM sequence, complementary to the sgRNA strand, and around 75-85 nt in length with 30-40 nt of homology on both sides. Following these findings, our 76 nt repair template was designed to have a 40 nt left homology arm and 33 nt right homology with the nontarget strand ( Figure 5A), a point mutation (C insertion) 4 nt away from the cutting site, and a 2 nt silent mutation (CTC > TTG Leu codon) to prevent secondary cutting. Because the WT (rescued) allele is dominant, we used a backcross to w -/and counted the number of G 1 rescued individuals, as a proxy for estimating the proportion of germline-editing among mosaic founders. Backcrossing 10 of the 37.5% of G 0 individuals in which we could detect somatic repair in the eyes, 24.4% of 496 G 1 progeny showed rescued alleles from precise HDR. Injections for this experiment took place at 15-60 min AEL. As mentioned above, two experimenters injecting within 25 min AEL may limit both somatic and germline mosaicism and boost the editing rates. The dominant rescue assays suggest CRISPR methods enable precise germline editing in Plodia with relatively high efficiency, and we encourage the use and development of similar methods for targeted genetic transformation in this system.

Limitations of the study
Although our study presents an approach to robust CRISPR/Cas9 NHEJ and HDR method in Plodia, we limited our knock-in attempts to a short ssODN template. The impact of longer repair template insertion on knock-in efficiency is yet to be determined. Efficient insertion of larger fragments in insects have been reported using donor dsDNA template, including the use of long donor dsDNA plasmids for HDR (

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:  Traut, W., Sahara, K., and Marec, F. (2007). Sex chromosomes and sex determination in Lepidoptera. Sex. Dev. 1, 332-346.

Materials availability
The Pi_Fog w -/and w +/+ strains are available to other laboratories upon request to the Lead Contact, and are compliant for interstate movement within the continental USA without additional permits (USDA ruling APHIS-2008-0076-0093).

Data and code availability
All short-read data generated in this project are available on the NCBI repository under accession number PRJNA610057. This study did not generate new code. iScience Article 2020), was used as a wild-type strain with a reference genome assembly (NCBI Assembly database: Pi_dt). The Pi_Fog strain consists of an introgression of the spontaneous wmutation (Shirk, 2021), from the Piwstrain (origin: USA, a kind gift of Paul Shirk), into the genomic background of Pi_Dun. For 13 of the 15 generations of crossing ( Figure 1A), pupae were sexed, transferred to individual plastic cups, mixed with an individual from the opposite sex after emergence, supplemented with diet after 3 days, and transferred to a 350 mL rearing container after confirmation of larval eclosure. About 6-8 crosses were performed at each generation, and w +/heterozygous genotypes were inferred from phenotype segregation among offspring. Two generations consisted of random matings within mass-rearing containers. At the end of this crossing scheme, w +/+ and w -/from the resulting hybrid Pi_Fog strain were isolated and both genotypes have been successfully maintained in inbred state for 2 years. More detailed methods for the use of Plodia as a laboratory model system for functional genetics are presented in the supplemental Information (Document S1).

Whole-genome resequencing
Adult individuals with Piw-, Pi_Fog w -/-, and Pi_Fog w +/+ genotypes were isolated for whole-genome resequencing. For each individual, the head and thorax were placed in a 1.5 mL microtube, lyzed with a matching pestle in 180 mL Buffer ATL (Qiagen), incubated with 400 mg Proteinase K at 56 C for 20 min, and supplemented with 12 mL of RNAse Cocktail (Thermo Fisher Scientific) for 12 0 at room temperature. DNA was extracted with the DNeasy Blood and Tissue kit (Qiagen), eluted using 70 mL of Buffer AE (Qiagen, 10 mM Tris-Cl, 0.5 mM EDTA, pH 9.0), and quantified using a Qubit 3 fluorometer and the HS dsDNA kit (Thermo Fisher Scientific), yielding between 0.7 and 4.2 mg of DNA per individual. A minimum of 1 mg per individual was used for preparation of five barcoded samples, which were sequenced in a large multiplexed library on a NovaSeq S1 150 bp PE run (Illumina). Raw reads are available on the NCBI SRA under BioProject accession number PRJNA610057.

Association mapping of the w-allele
Reads were aligned to the Plodia Savannah strain genome assembly using BWA-MEM with default parameters (Childers et al., 2021;Li, 2013). Variant calling was performed in GATK using tools Haploty-peCaller and GenotypeGVCFs with default parameters, with the output filtered for just bi-allelic SNPs using Samtools (Li et al., 2009;McKenna et al., 2010). Population statistics were plotted using code from Simon H Martin (2021), with F ST calculated in 100 kb windows at 10 kb sliding intervals. Chromosome genotype plots were generated using the R package GenotypePlot (Whiting, 2021). Variant call files (VCFs) for the genes white, cinnabar and scarlet were examined in IGV.

In-silico protein analysis
The reference genome sequence (NCBI Assembly Plodia_dt: the ''Dovetail'' version of the WT -Dundee strain Pi_Dun) was used to identify a complete coding sequence, and was translated into a 685 amino acid residues. White protein sequence. Protein family domains were visualized using profile-hidden Markov modeling, HMMER web server 2.41.1 (Potter et al., 2018) following a search against the Drosophila melanogaster Reference Proteome in UniProt. To visualize the tertiary and quaternary structure of the White protein, we generated the 3D model of the protein using SWISS-MODEL (Waterhouse et al., 2018). The protein was categorized as ABCG2 and was fitted into the available 6hbu.1.A human ABCG2 mutant protein template (Manolaridis et al., 2018). The generated template offered 0.87 coverage with tertiary and quaternary structure quality estimate scores (GMQE and QSQE) of 0.62 and 0.47.

Detailed rearing procedure
Plodia is well suited for laboratory culture, and routine stock maintenance involves a single human intervention per generation. Briefly, adult moths are left to mate in a mass-rearing container and anesthetized with CO 2 , which induces oviposition and allows the collection of large quantities of eggs ( Figures S4A-S4G). As fertilization occurs at oviposition (Lum et al., 1981), this generates synchronized embryos that can then be used for inoculation of a new stock ( Figures S4H-S4J), or for microinjection ( Figures S4K and S4L).
A wheat bran artificial diet suitable for optimal larval growth was adapted from a previously published protocol (Silhacek and Miller, 1972). Following mixing of the ingredients (Table S1), the diet can be kept ll

OPEN ACCESS
Larval containers must be completely insect proof, tolerate freezing temperatures for devitalization, and resist accidental falls without opening. We recommend food storage containers with tight-locking lids and thick walls (e.g. LocknLock Rectangular, 350 mL). Vents must be added to allow air circulation: a 2 cm diameter hole is perforated in the lid, and 100 3 100 copper wire mesh is used to cover the hole on the inside of the lid, with a soldering iron used to melt the mesh into the plastic without ( Figure S5).
Plodia adults of the Pi_Fog strain emerge about 25 days after fertilization when reared at 28 C. Moths containers are then left one to two days before egg collection to ensure fertile matings. To pass the stock, the container is saturated with CO 2 gas until all adults are anesthetized, then the Plodia adults are quickly transferred into a mason jar covered with a steel 40 3 40 mesh disc under the lid (oviposition jar, Figure S6). CO 2 narcosis induces oviposition, resulting in a rapid accumulation of fertilized eggs within 2 h. Silanization of the inside of the container (e.g. Rain-X glass treatment) can help reduce the sticking of eggs to the jar, but it is usually not required. For inoculation of a new generation, eggs are transferred to a stainless steel cup (which avoids static charging), by inversion and tapping of the eggs through the mesh lid of the oviposition jar. Eggs are weighed on a weighing boat, which is then placed inside a rearing container with compressed diet. For 350 mL containers, this is dones at a ratio of 10-12 mg of eggs for 50 g of diet. A piece of filter paper is taped over the mesh vent ( Figure S4A) to prevent dessication. The stock can then be left unattended for the next 25 days (peak of adult emergence), although we recommend periodic monitoring of

Embryo microinjections
To anesthetize and induce synchronized egg-laying, adult Plodia moths in a rearing container are exposed to CO 2 , then transferred to the oviposition jar. The jar is tapped upside down to remove older eggs. After 10-12 min of oviposition at room temperature, the eggs are collected in a metal container ( Figure S7A). Using a wet paintbrush, the eggs are picked up and arranged on a 1 3 3 cm parafilm in a cell culture dish ( Figure S7B) such that the micropyle is on the opposite side of the intended injection site ( Figure S7C), adding a droplet of water as adhesive on the bottom, if necessary. Using a microcapillary needle, a droplet of injection mix is pushed by air pressure into the posterior side of the egg ( Figure S7D). After injection, it is important to seal the site of injection by tapping onto the top of the egg a fine copper wire dipped in thin glue ( Figure S7E). The time from egg collection to egg sealing is kept under 60 min. The total injected eggs are then counted and incubated at 28 C. For the first 24-48 h, injected eggs are kept in the rearing container in the presence of a slightly damp tissue wipe, and with tape over the copper mesh hole. This procedure provides humidity saturation that prevents the eggs from drying out. After 24-48 h, the tape is removed to allow ventilation, and the wet tissue wipe is removed from the container, preventing molding. To prevent the escape of larvae outside of the injection dish, it is recommended to add a small amount of artificial diet (one or two flakes per dish) when removing the wipe tissue, as this will attract hatchlings and facilitates counting and transfer in further procedures.

CRISPR NHEJ knock-outs
Syncytial embryos of Pi_Fog w +/+ were injected with Cas9 and short guide RNA (sgRNA, Pi_wh5_KO_sg or Pi_BLOS2_KO_sg) complexes at 500:250 ng/mL concentration, within 15-60 min after egg laying (AEL). Pupae were then individually isolated from cardboard ''hotels'' for phenotyping under dissection microscope and outcrossing ( Figure S6). Selected white exon 5 mutants lacking any pigmentation in both eyes were crossed with the w -/strain to determine the germline transmission of NHEJ-induced null mutations.

CRISPR HDR rescue
To investigate whether the insertion of cytosine at position 737 in white exon 5 is necessary to restore the brown eye phenotype in the Pi_Fog w -/strain, eggs were injected with Cas9, Pi_wh5_HR_sg RNA, and Pi_s-sODN_wh5_76NT single-stranded oligo donor (ssODN) complex at 500, 250, and 1000 ng/mL concentration, respectively. All eggs were injected within 15-60 min AEL. iScience Article cocoon and lined up on a double-sided tape that has been dabbed with a Kimwipe to reduce stickiness. The courtship behavior of male and female Plodia (Grant et al., 1975) consists of a stereotypical sequence initiated with the pheromone call of females that show an upward-bending of the abdomen to expose a posterior gland. If males are sexually mature (in our experience, at least 24 h after emergence), they often immediately respond to this call and perform a rapid fanning of their wings to spread wing pheromones while walking towards the female, followed by copulation. Successful matings are easily visualized at the bench, often seconds after mixing of virgin males and females in the same container. If mating does not successfully occur, it can be elicited by placing the container in the dark, or it may indicate that the individuals have emerged too recently.

Genotyping assays
For Restriction fragment-length polymorphism (RFLP), one leg of Plodia was submerged into 0.5 mL DNARelease Additive and 20 mL Dilution Buffer, heated at 98 C for 2 min, and 1 mL of this lysate was used for PCR in a 15 mL volume (95 C for 5 min, followed by 30 cycles with 95 C for 15 s, 56 C for 15 s, 68 C for 60 s; and final elongation at 68 C for 1 min) using 7.5 mL Taq 2x Master Mix and 0.2 mL of each primer (10 mM stock). Alternatively, a pupa can be pricked between dorsal thoracic segments with a tungsten needle for sampling $0.2 mL of hemolymph, which then is mixed with 10 mL of the DNArelease/Dilution buffer solution. A volume of 2 mL of this DNA extract can then be used for PCR amplification.
For the genotyping of c.737delC the 40 mg of product amplified by Pi_wh11-12_F/R primers was digested for 1 h at 37 C with 1 U of NlaIV (New England Biolabs). For restriction analysis, 5 mL of the restriction product and 0.5 mL O'GeneRuler 100 bp DNA ladder (Thermo Fisher Scientific) were loaded onto a 2% agarose gel with 0.5 X TBE buffer and electrophoresed at 50 V ( Figure S9).
For Sanger sequencing, 40 ng of PCR product was purified with 1 U ExoI and 1 U of rSAP (New England Biolabs) in a 6.5 mL reaction volume. The reaction was incubated for 15 min at 37 C followed by enzyme inactivation for 15 min at 80 C. Then, 25 pmol of Pi_wh11-12_R primer was added to 3.2 mL of purified PCR product in 15 mL total volume per reaction.

Biosafety measures for wild-type stock
Plodia pantry moths are non-infectious, non-parasitic, non-vector, and non-exotic insects that are found on all continents except Antarctica, and generally considered as native pests of stored food products. They are notorious for their infestation potential in kitchens and pantries, and can proliferate within a laboratory building if baseline containment measures are not taken. We prescribe important recommendations for working with this insect at the wild-type state.
d Transfers between primary containers. The entire lifecycle is maintained in escape-proof containers, so the risk of egg/diet spills or flying moth escapes is limited to moments where these containers are opened. The transfer of narcosed adults to an oviposition jar should be quick, over a tray ( Figures 4B and 4C), and under a fumehood. Lab surfaces must be wiped with ethanol if potentially contaminated with eggs. The transfer of adults isolated in cups or flasks must be done inside a secondary container, such as a mesh cage.
Secondary containment. Plodia cultures should be in a dedicated incubator or in a clean insectary with limited access and no direct escape routes. ''Pantry moth'' pheromone traps are highly effective for detecting an early infestation, and may help to catch accidentally released males before they mate. Indoor ultraviolet insect traps can also be installed as preventive measures in the insectary and adjacent rooms.
Lab environment. Food must be prohibited in the laboratory areas surrounding the insectary. In addition, Plodia can infest the wax of laboratory bee cultures.
Disposal and cleaning. Stringent measures to devitalize all the material for disposal in order to avoid a laboratory infestation. All containers, individuals (including excess eggs), cultures and waste (wipes) are sealed in a plastic bag and frozen at À20 C for at least 72 h in order to achieve complete disinfestation, a conservative procedure based on published measures of an LT 95 estimate of 7 h at this temperature (Johnson, 2007). Dilute bleach (5%) treatment for 10 min dissolves egg chorions and is an effective chemical decontaminant for surfaces and equipment that can not be frozen. iScience Article Personal protective equipment. Disposable gloves are recommended to avoid contaminating cultures with mold. Lab coats are recommended to avoid the export of egg material outside of the lab. Most importantly, experimenters must use respiratory protection (fumehood, respiratory filter) when opening or cleaning mass-rearing containers, due to the allergen potential of moth scale volatiles (Binder et al., 2001).

Arthropod containment level 2 measures for genetically modified stock
The generation and maintenance of transgenic or genome edited arthropod stock is prone to formal regulations that are country or institution specific. Published guidelines about Arthropod Containment Levels 1 and 2 (ACL-1 and ACL-2) protocols (American Committee of Medical Entomology-American Society of Tropical Medicine and Hygiene, 2019) provide an advisory framework for safe genetic modification in Plodia. While the ACL-2 standard was designed for arthropods of public health concern, it is flexible and overall a good target for responsible research conduct, but may be insufficient if the laboratory modifications present a risk of fitness advantage or genetic contamination in the wild (e.g. insecticide resistance, gene drives). In addition to the aforementioned baseline recommendations with wild-type Plodia stock, we advise the following when working with genetically modified Plodia.
Disposal. All material must be treated as Biohazardous waste (similar to transgenic Drosophila) following ACL-2 protocols.
Facilities. Simple ACL-2 barriers to flying and crawling insects, for instance using a dedicated insectary or incubator room with self-closing doors, and custom ventilation meshes, should be accommodated whenever possible. More expensive facilities with double-door vestibules, or negative air pressure, are not needed for Plodia, since accidentally released adult moths should be extremely rare with the aforementioned primary containment procedures.
Personel. ACL-2 measures have increased stringency on access and training compared to ACL-1, which are important to reduce the risk of accidental releases due to negligence.

QUANTIFICATION AND STATISTICAL ANALYSIS
Our study did not include statistical analysis or quantification.