CRISPR/Cas9‐based functional analysis of yellow gene in the diamondback moth, Plutella xylostella

Abstract The diamondback moth, Plutella xylostella (L.), is an economically important pest of cruciferous crops worldwide. This pest is notorious for rapid evolution of the resistance to different classes of insecticides, making it increasingly difficult to control. Genetics‐based control approaches, through manipulation of target genes, have been reported as promising supplements or alternatives to traditional methods of pest management. Here we identified a gene of pigmentation (yellow) in P. xylostella, Pxyellow, which encodes 1674 bp complementary DNA sequence with four exons and three introns. Using the clustered regularly interspersed palindromic repeats (CRISPR)/CRISPR‐associated protein 9 system, we knocked out Pxyellow, targeting two sites in Exon III, to generate 272 chimeric mutants (57% of the CRISPR‐treated individuals) with color‐changed phenotypes of the 1st to 3rd instar larvae, pupae, and adults, indicating that Pxyellow plays an essential role in the body pigmentation of P. xylostella. Fitness analysis revealed no significant difference in the oviposition of adults, the hatchability of eggs, and the weight of pupae between homozygous mutants and wildtypes, suggesting that Pxyellow is not directly involved in regulation of growth, development, or reproduction. This work advances our understanding of the genetic and insect science molecular basis for body pigmentation of P. xylostella, and opens a wide avenue for development of the genetically based pest control techniques using Pxyellow as a screening marker.

damage to cruciferous crops (Furlong et al., 2013). Due to the overuse and misuse of insecticidal chemicals, rapid evolution of resistance to all major classes of pesticides has made DBM increasingly difficult to be effectively controlled. Genetics-based strategies have been proposed as environmentally friendly alternatives to the overuse of insecticides in pest management (Alphey, 2014). Recently, a novel genetic approach of self-sustaining population elimination, clustered regularly interspersed palindromic repeats (CRISPR)-based gene drive system, has been developed in the model insect Drosophila melanogaster  as well as non-Drosophila disease vectors Li et al., 2020), which all showed promising population control results. Although some sex-determination genes have been proposed as potential targets for genetics-based population suppression (Kyrou et al., 2018;Wang et al., 2019), in order to build gene-driven prototypes and assess the driving efficiency in different species, it is desirable to target endogenous phenotypic genes, such as yellow (one of the main melanin synthesis pathway genes), in the first place. In Drosophila, yellow protein was required in producing black melanin, which maintained normal black body pigmentation (Wittkopp et al., 2002). Mutations in the yellow gene were reported to cause a change in the melanin synthesis pattern, turning the coloration from black to yellow (Wittkopp et al., 2002). Similar phenotypes derived from yellow-deficient insects were also observed in Tribolium castaneum (Rylee et al., 2018) and Bombyx mori (Xia et al., 2006;Futahashi et al., 2008). In addition to yellowish body color, the disruption of yellow gene also led to a dehydration-like phenotype during a short developmental stage in Agrotis ipsilon (Chen et al., 2018). However, the regulation of body pigmentation and the possible functions of yellow in DBM remain unclear.
CRISPR/CRISPR-associated protein 9 (Cas9)-induced mutagenesis of target genes have been documented in multiple species of moth insects, including Spodoptera littoralis (Koutroumpa et al., 2016), Spodoptera litura , Helicoverpa armigera Khan et al., 2017), and A. ipsilon (Chen et al., 2018). Since 2016, using the CRISPR/Cas9 approach, several cases of gene manipulation in DBM have been reported by our team (Huang et al., 2016;Peng et al., 2019;Chen et al., 2020) and another research group , providing a relatively mature gene editing platform in this global pest. Therefore, as an efficient genome editing tool, the CRISPR/Cas9 system was utilized in this study to verify the functions of yellow gene in DBM (hereafter Pxyellow).
To identify the coding sequence of Pxyellow, we used one of the yellow genes in B. mori, Bmyellow-y (NP_001037434.1), as a query to blast against our previously published P. xylostella genome sequence (You et al., 2013). Nine putative yellow homologs were found in P. xylostella (gene ID: Px007091, Px007817, Px005439, Px016714, Px011025, Px015683, Px005437, Px005436 and Px010416). The deduced amino acid sequences of these genes contained the conserved domain MRJP (major royal jelly protein), which is the characteristic motif of yellow proteins across different insect species (e.g., B. mori and D. melanogaster), although proteins encoded by Px005436 and Px010416 only comprised partial MRJP domain (Table S1). Phylogenetic analysis showed that these genes were well clustered with other insect homologs, indicating the potentially conserved functions of these genes in different species ( Fig. S1 and Table S2). Based on the well-studied role of Yellow-y protein in promoting melanization in B. mori (Futahashi et al., 2008) and T. castaneum (Arakane et al., 2010), the most likely yellow-y ortholog in DBM (gene ID: Px007091), which showed the lowest E-value by blasting B. mori yellowy against the DBM genome, was identified as Pxyellow and further investigated. This gene was mapped into the region 747 503-755 533 bp in scaffold 25 of the DBM genome. The identified complementary DNA sequence of Pxyellow was 1674 bp, containing four exons, each with 277, 186, 1190 and 21 bp in length, and three introns with 4465, 248, and 1650 bp in length, respectively ( Fig. 1A).
To introduce CRISPR/Cas9-mediated mutagenesis in Pxyellow, two target sites (yellow-sgRNA1 and yellow-sgRNA2) located in Exon III were selected using the Zi-FiT Targeter software (Fig. 1B). The off-target binding capability of two designed single guide RNAs (sgRNAs) was analyzed by blasting target sequences against DBM genome (maximum mismatches = 3). However, no potential off-target sites were found, indicating a high target specificity of these designed sgRNAs.
In order to generate Pxyellow mutants, totally 676 eggs were injected with Pxyellow-sgRNAs and 480 of them hatched, resulting in 71% of hatchability. Additionally, there was no significant difference in hatchability to the negative control (injected with enhanced green fluorescent protein-sgRNA; hatchability = 71.6%) (Table S3). Based on the observation of yellow pigmentation of 1st instar larvae, 272 of the G 0 moths were mutated (mutation rate = 57%) (Table S3). Compared with the wild type individuals (with light-black body and black head capsule), the pigmentation of G 0 newly hatched larvae (Fig. 1C), 1st (Fig. 1D) and 2nd instar (Fig. 1E) larvae turned yellow (especially apparent change in the color of head capsule), while the body color of treated 3rd instar larvae changed to light yellow (Fig. 1F). The body colors of treated 4th instar larvae (Fig. 1G) and early pupae (Fig. 1H) were not observably different from the color of wild types, while the pigmentation of mutants changed from black or dark brown to yellow/tan in the late pupal stage (Fig. 1I). This indicated that Pxyellow was not likely involved in the pigmentation of 4th instar larvae and early pupae, but participated in the melanization of late pupae. A similar result was also obtained from the loss-of-function mutant of Aiyellow-y in A. ipsilon, showing no significant difference in color between the mutants and wild types in early pupae while the initiation of varied pigmentation occurred in late pupal stage (Chen et al., 2018). In addition, the pigmentation of CRISPRtreated adults in our study changed from gray/black to yellow/tan (Fig. 1J). In total, 33 of G 0 adults with mutant phenotypes were randomly selected and sequenced, which revealed various insertions or deletions (indels) at both target sites (representative mutant genotypes are provided in Fig. 1K), suggesting the successful mutagenesis in Pxyellow locus using the CRISPR/Cas9 system.
To build a homozygous mutant line for further investigation, the G 0 mutant adults were first crossed with wild type adults in pairs to generate the G 1 generation. Thirty-five G 1 individuals randomly collected from 10 G 0 parents were sequenced and 11 of them showed mutations (inheritance efficiency = 31.4% [11/35]). In addition, mutant G 1 s were outcrossed with wild types in pairs for producing G 2 generation, 12 of which were sequenced to confirm their genotypes. Note that both male and female heterozygotes in G 1 and G 2 generations showed wild type-like pigmentation instead of yellowish body color observed in G 0 , indicating that the Pxyellow mutation generated here was recessive and Pxyellow is not located in sex chromosomes. G 2 individuals hosting the same mutant type (a 16 bp insertion linked to a frame-shift mutation; Fig. 1K) were pairwise inbred to generate G 3 s, of which the homozygous mutants showing abnormally yellow pigmentation were maintained as a Pxyellow knockout strain ( Fig. 2A-D). The phenotype of G 3 homozygous mutants were mostly consistent with the G 0 mosaics, although the body color of some G 0 mosaics retained patchy wild type-like dark traits while G 3 ho-mozygotes showed fully yellow pigmentation (Fig. 2D). This observable mutant phenotype in G 0 s was probably linked to the long development time of Lepidoptera, giving more chance for more G 0 cells to mutate during embryo development. It was noted that the egg color of homozygous mutants at the later embryonic stage was light yellow instead of dark gray observed in their wild type counterparts (Fig. 2A).
To explore whether Pxyellow deficiency resulted in any fitness cost in DBM, a series of growth, development and reproduction tests were conducted using both the Pxyellow mutant line (group A) and wild type control (group B) with 30 pairs of adults set up in each group. Due to oviposition failure in some replicates (which might be caused by individual fertility variation naturally existing in the DBM population), only 27 pairs of group A and 28 pairs of group B were kept for subsequent fitness analysis. The results showed no significant difference between the Pxyellow-deficient group and the wild type group in oviposition, hatchability or pupal weight (Fig. 2E). These findings suggested that the mutagenesis of Pxyellow induced by CRISPR/Cas9 conferred the phenotypic change in body pigmentation without affecting the growth, development or reproduction of DBM. This is consistent with previous research in A. ipsilon where deficiency in Aiyellow-y did not obviously affect the moth growth (Chen et al., 2018). However, it has been reported that yellow family genes comprised rather diverse gene functions. For example, yellow-g and yellow-g2, participated in the development of egg desiccation resistance in Aedes Albopictus (Noh et al., 2020). Although the egg hatchability was not affected in the Pxyellow knockout line, further investigation may be needed to confirm whether it played other roles in DBM embryonic development.
Due to the ease of screening mutant phenotypes, Pxyellow can be used as a germline transformation marker for constructing transgenic DBM, providing a useful and measurable tool in genetically based pest control prototypes (i.e., CRISPR/based gene drive systems). Based on our result that disruption of Pxyellow likely had no undesirable impact on insect fitness, drivers (e.g., Cas9/sgRNA expressing cassette) can be integrated into yellow locus to build viable transgenic lines, followed by cage/field assays to test the spread of transgenics in populations . It is noted that previous reports in D. melanogaster showed changes in male mating behavior and the consequent reduction in male-specific mating success due to yellow null-mutation (Massey et al., 2019). This could be an obstacle in assessing homing efficiency since the transgenics may retain mating disabilities when paired with wild types. Al-though no observable mating defect was seen in our Pxyellow mutant line, a mating competition assay may be required in the future to evaluate the potential ability of mutant lines in transmitting the transgenics into natural populations.
This is the first report of a phenotypic gene, yellow, in DBM with CRISPR/Cas9-mediated loss-of-function analysis. In summary, Pxyellow played a critical role in the pigmentation patterns in DBM, and the Pxyellowdeficient phenotype could be easily observed through the majority of developmental stages.

Supporting Information
Additional supporting information may be found online in the Supporting Information section at the end of the article.

Fig. S1
Phylogenetic tree of yellow gene families based on the alignment of their amino acid sequences from six insect species.
Supporting Materials and methods. Table S1 Putative yellow orthologs in Plutella xylostella.
Table S2 GenBank information of yellow sequences used for construction of the phylogenetic tree (Fig. S1). Table S3 Mutagenesis mediated by clustered regularly interspersed palindromic repeats (CRISPR)/CRISPRassociated protein 9 targeted Pxyellow.