To search for the Wnt-1 homolog, nucleotide sequence of BmWnt-1 (NM_001043850.1) was used as a query to BLAST against a D. punctatus transcriptome (HHL, unpublished data). RACE was used to obtain the full length cDNA of DpWnt-1. The predicted open reading frame (ORF) was subjected to motif search, pattern analysis, and phylogenetic analysis. The MEME online server was used for motif analysis, and parameters were as follows: a minimum width was 6; a maximum width was 12; and a maximum number of motif was 8 (http://meme-suite.org/tools/meme). To understand the phylogenetic relationship of DpWnt-1 with homologs from other animals, a neighbor-joining tree was constructed using MEGA5, http://www.mega-software.net/ (Tamura et al., 2011). The Wnt-1 ORFs included in the analysis are as follows: B. mori (NM_001043850), H. armigera (KJ206240), Amyelois transitella (XM_013345048), P. xylostella (XM_011569928), M. sexta (Z30280), P. xuthus (XM_013325799), Danio rerio (XP_005162280), Fopius arisanus (XM_011300877), Bombus terrestris (XM_003393116), Nasonia vitripennis (XM_001603338), Bactrocera dorsalis (XM_011204079), Drosophila willistoni (XM_002066877), Drosophila melanogaster (NM_078778), Tribolium castaneum (EFA04660), Periplaneta americana (KC311252), Gryllus bimaculatus (BAB19660), Homo sapiens (NP_005421) and Mus musculus (NP_067254).

Total RNA was isolated with Trizol Reagent (Invitrogen, USA) from D. punctatus pupae. Recombinant DNase I-treated (Takara, Japan) RNA was used for cDNA synthesis with the Scientific Revert Aid First Strand cDNA synthesis kit (Thermo, USA). Diluted reverse transcription products were used as templates to amplify DNA fragments. The primer sets used to obtain the DpWnt-1 ORF are listed in Table 1. Template DNA was denatured at 94°C for 2 min, followed by 35 cycles of 94°C for 15 s, 55°C for 30 s and 68°C for 1.5 min. PCR products were cloned into the pCR-Blunt vector for sequencing by ABI 3730 XL sequencer (Applied. Biosystems, USA).

qRT-PCR was performed to analyse the expression profile of DpWnt-1 and 8 Hox genes during the embryonic stage. cDNA samples were prepared from embryos of different developmental stages (day 1–day 8 of wild type) and the first instar larvae of DpWnt-1 mutants. Mastercycler EP realplex (Eppendorf) was used for the qRT-PCR. The primer sets used in qRT-PCR analysis are listed in Table 1. The cycling conditions were as follows: an initial incubation at 95°C for 10 s, 40 cycles of 95°C for 15 s, and 60°C for 30 s according to SYBR Green fluorescent relative quantitative approaches (TaKaRa, Japan). The relative mRNA level of the target genes was calculated using the 2^−ΔΔCt method, in which the target gene expression was normalized to an internal reference, RP32. Three independent replications for each sample were performed.

The Cas9 gene template used in this work was provided by View Solid Biotech (Beijing, China). Cas9 mRNA was synthesized in vitro with the mMESSAGE mMACHINE® T7 kit (Ambion, USA) according to the manufacturer's instructions.

For the in vitro transcription of sgRNA driven by the T7 promoter, target sequences start with GG. With the PAM sequences in consideration, the designed sgRNA sites follow the GGN₁₉GG rule (Wang et al., 2013). We identified two 23 bp sgRNA targeting sites at exon III of DpWnt-1 (Figure 3A). The control sgRNAs were used for targeting the EGFP gene. Two complementary oligonucleotides were annealed and cloned into pJET1.2 (Fermentas, USA). The templates for in vitro transcription were amplified from pJET1.2, and primer sets used in this study are listed in Table 1. sgRNAs were transcribed in vitro with the MAXIscript® T7 kit (Ambion, USA), following the manufacturer's recommended protocol.

Dendrolimus punctatus pupae were originally obtained from Xing'an County of Guilin city, Guangxi province, P.R. China. D. punctatus colonies were provisioned with Masson's pine, and maintained at 27 ± 1°C under a L/D cycle of 16/8 h. Fertilized eggs were collected within 2 h after oviposition, and subjected to microinjection.

The combination of Cas9 mRNA (300 ng/μl) and sgRNAs (sgRNA-a and sgRNA-b, 300 ng/μl, respectively), and Cas9 mRNA/sgRNAs (sgRNA-a and sgRNA-b) (500 ng/μl each) were co-injected into preblastoderm embryos. An exogenous gene EGFP and nuclear free water without any sgRNAs or Cas9 mRNA were used as control. These control should have none effect on the embryonic development. Injection was carried out following Tamura et al. (1990) with modification, and injection site was shown in Figure S2. As the egg is oval in shape, we lined up the egg with the micropyle on top and injected compounds to the gonad region. The microinjection was concluded within 6 h. Afterwards, the injected eggs were incubated at 25 ± 1°C in a humidified chamber for 8–10 days until hatch. All hatched larvae were collected and transferred to Masson's pine.

The injected embryos were dissected and checked to calculate the mutation rate and hatching rate on the seventh day of the embryonic stage, and the resultant phenotypes were documented under a multi-function zoom microscope (AZ100, Nikon). The images were recorded with a computer-controlled microscope system. The pictures of DpWnt-1 mutants, including both larvae and pupae, were taken by SLR cameras.

To calculate the efficiency of Cas9/sgRNA-mediated gene alteration in the injected generation, individuals were collected on the eighth day after injection. The DNA fragments surrounding the sgRNA targets were obtained by GBdirect PCR directly from embryos (GBI, China). The primer sets are shown in Table 1. Mutations were confirmed by sequencing.

Proteins from 7 day old embryos were used for the immunoblotting analysis. The primary antibodies, B. mori Anti-Wnt-1 and Anti-β-actin, respectively, were used at 1:1000 dilution. The secondary antibody, anti-rabbit IgG, was diluted at 1:5000. Proteins were extracted and diluted with PBS and quantified using bicinchoninic acid (BCA) protein assay kit (Thermo). A 12.5% SDS-PAGE gel was used to separate the same amount of proteins from both the wild types and mutants. The proteins were then transferred to a polyvinylidene fluoride membrane. Signal visualization was obtained using the ECL Plus Western Blotting detection kit (GE Health-care).

EST sequence of DpWnt-1 (GenBank accession #:KU640201) was initially obtained from D. punctatus transcriptome. The full length cDNAs of DpWnt-1 contained 1182 nucleotides, which encodes 394 amino acids. The nucleotide sequence of DpWnt-1 was rich in cysteine residues-a character of Wnt protein family (Figure S1). Wnt-1 homologs from 18 species shared eight conserved motifs, which located between the N- and C-terminus (Figure 1). Phylogenetic relationship showed that DpWNT-1 clustered with other lepidopterans WNT-1 protein sequences (Figure S2). The expression of DpWnt-1 peaked at the very beginning, declined during the development, and reached the minimum level at the end of embryogenesis (Figure 2), suggesting that DpWnt-1 may play a vital role in D. punctatus during the early embryogenesis.

To functionally characterize DpWnt-1, CRISPR/Cas9 mutagenesis system was introduced into D. punctatus. A total of 240 D. punctatus eggs were co-injected for each concentration of Cas9 mRNA and DpWnt-1 sgRNAs, whereas 120 eggs were injected for the corresponding concentrations for the control EGFP sgRNAs (Table 2). Compared to co-injections of Cas9 protein and DpWnt-1 guide RNAs with those targeting a control gene (EGFP), D. punctatus embryos with an inactive copy of Wnt-1showed a reduced hatching rate (22.5 and 30.5% at a concentration of 500 and 300 ng/μl, respectively), and a range of phenotypic effects (e.g., various body plan defects, absence of tissue differentiation). Among the 120 control eggs injected with EGFP sgRNAs/Cas9 mRNA, 57.5 and 64.2% individuals hatched at a concentration of 500 and 300 ng/μl, respectively. In comparison, 65.8% (79/120) wild type eggs hatched.

CRISPR/Cas9 system induced mutations in the pine moth with high efficiency. Eighty percentage (8 of 10) of the dissected embryos had mutations at the target sites, and the overall mutagenesis frequency was 32.9% in the injected generation at a higher dose (500 ng/μl). Similarly, at a lower dosage (300 ng/μl), 70% (7 of 10) of the dissected embryos had mutations at the target sites and the overall mutagenesis frequency was ~17.5% (Table 2). The genotypes of the wild types and DpWnt-1 mutants were confirmed by both sequencing and Western blotting analysis (Figures 3B,C). All examined DpWnt-1 mutants, including embryos and larvae, had alterations at the target sites that led to at least five type of deletions (Figure 3D). The deletion occurred at target sites individually, simultaneously, or was absent from both sites.

Knocking out DpWnt-1 has great impact on eggs development. Most of eggs showed abdominal segments distortion and only some of them could hatch and develop into pupae, of which none reached the adult stage. When injected with 500 ng/μl of Cas9 mRNA and DpWnt-1 sgRNA, 22.9% of the embryos showed abnormal anterior-posterior (A-P) axis and abdominal segmentation phenotypes, 7.5% showed defective legs, and 2.5% showed head malformations. In contrast, when the injection concentration is 300 ng/μl, 9.5% of embryos showed abnormal A-P axis and abdominal segmentation phenotypes, 6.3% showed defective legs, and 1.7% showed head malformations. As a control, 240 eggs were co-injected with EGFP-sgRNA/Cas9 mRNA. A total of 146 eggs (60.8%) hatched, and no morphological changes were observed (Table 2).

Understanding the function of Wnt-1 is critical for exploring its potential role in pest management. In this study, we cloned and characterized DpWnt-1 homolog and identified one Wnt-1 gene in D. punctatus, DpWnt-1. The motif and phylogenetic analyses confirmed that DpWnt-1 is most closely related to BmWnt-1 (Dhawan and Gopinathan, 2003).

In Drosophila, with long germ embryos, Wnt-1 expression was first detected in the whole segments of the blastoderm during cellularization (Baker, 1987; Vorwald-Denholtz and De Robertis, 2011). In Tribolium, with short-germ embryos, Wnt-1 was initially detected in the blastoderm stage, expressed sequentially from anterior to posterior with the germ band elongation and at the ventral portion of each segment during the late embryonic stage (Nagy and Carroll, 1994). In short/intermediate germ embryos, Wnt-1 was detected in a broad median of the germ disk and then retracted posteriorly within segmentation process (Nakao, 2010). The expression pattern of DpWnt-1 during the embryonic stage showed the same trend with that of Bombyx (Zhang et al., 2015). BmWnt-1 was present in a maternal gradient and might play a role during the blastoderm formation (Nakao, 2010; Zhang et al., 2015). We hypothesized that D. punctatus may have a short/intermediate germ band, in which segmentation proceeds consecutively from anterior to posterior and show visible anterior and posterior segments after gastrulation.

In this study, embryonic injection of a mixture of sgRNAs/Cas9 mRNA successfully induced mutations in DpWnt-1, demonstrating that CRISPR/Cas9-mediated genome editing can specifically and efficiently induce gene alterations in D. punctatus. Besides D. punctatus, CRISPR/Cas9 system has also been exploited in seven other Lepidoptera species, including B. mori, S. litura, S. littoralis, P. xylostella, P. xuthus, H. armigera, and D. plexippus, to manipulate genes associated with development (embryogenesis), pigmentation, metamorphosis, resistance mechanism, and adult mating (Wang et al., 2013, 2016; Li et al., 2015; Bi et al., 2016; Huang et al., 2016; Koutroumpa et al., 2016; Markert et al., 2016; Zhu et al., 2016). Moreover, the frequency of mutation is dose dependent. Knocking out DpWnt-1 led to a high embryonic mortality (~70%), and none of the DpWnt-1 mutants could developed from larva to adult, suggesting that DpWnt-1 is a potential candidate for conditional lethal gene.

Although CRISPR/Cas9 system is clearly applicable in Dendrolimus, additional experiments are needed to fully established this genome editing technology in this major forest pest. In situ hybridization study of DpWnt-1 not only will validate genome editing results at the translational level, but also provide the spatial expression pattern, and the potential Hox targets. Also, without genome information, we could not pinpoint the off-target effects, which is a routine problem for this technology. With other genomic resources (Yang et al., 2016), the potential off-target effects can be predicted.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.