The Structural Variation Is Associated with the Embryonic Lethality of a Novel Red Egg Mutant Fuyin-lre of Silkworm, Bombyx mori

Bombyx mori presents several types of egg color mutations, all of which have been extensively discussed in sericulture. While the red egg mutation has been previously observed, lethal red-egg mutants have not been reported. In the present work, the red egg mutant Fuyin-lre (Fuyin-lethal red egg) was discovered from the Fuyin germplasm resource of B. mori. This mutant features red-colored eggs and embryonic lethality. Genetic analysis showed that Fuyin-lre follows recessive inheritance, with the red egg gene re governing the egg color, and the embryonic lethality of Fuyin-lre may be caused by mutations of other genes closely linked to re. Digital gene expression (DGE) was employed to compare the transcription profiles of Fuyin and Fuyin-lre eggs after 24 and 48 h of incubation. A total of 48 differentially expressed genes followed the same expression patterns in both groups at both time points (FDR < 0.01 and log 2 Ratio ≥ 1). Further analyses indicated that 8 out of the 48 genes (including re) were closely linked to re. These 8 genes were highly expressed in wild-type Fuyin and the red egg mutant re but showed nearly absent expression in Fuyin-lre. Sequencing of the re gene confirmed that the re gene itself does not induce embryonic lethality, and structure analysis showed that the structural variation of the region where the 8 genes were located may be associated with the embryonic lethality of Fuyin-lre. The present work provides a good foundation for future studies on the mechanism of embryonic lethality and embryonic development in Fuyin-lre.


Introduction
Bombyx mori (domesticated silkworm) of the order Lepidoptera is an agriculturally and economically important insect. Silkworms have been used as experimental insects for studies in genetics, modern molecular biology, and genetic engineering. The genetic mutations of B. mori are an important aspect of sericulture. The Silkworm Genome Database of Southwest University contains over 90% of the world's silkworm mutant resources, and considerable progress has been made in the linkage analysis and development of marker systems for gene positioning [1][2]. Since the Silkworm Genome Research Program was completed [3], various mutated gene resources of B. mori have become important materials for studies on gene functions.
Egg color is an important trait in morphological and genetic studies of B. mori. Egg color mutations may occur in the egg shell or serous membrane. The color of the serous membrane is formed after fertilization. There are many types of color mutants, including white, brown, red, purple, orange, rust-colored eggs [4][5][6], and so on. The important red egg mutants of B. mori may be classified as red (re) [7][8], fourth brown (b-4), orange (ci) [9], and rusty (re/reci/ ci) egg mutants [10]. Osanai-Futahashi et al. [8] performed positional cloning of the red egg gene re in 2012. Unlike the discovery of mutants, the identification of mutated genes in egg mutants has lagged behind. In particular, the traditional identification methods are time-and labor-consuming, and research progress is relatively slow.
Completion of the Silkworm Genome Sequencing Project has enabled studies of mutant genes via molecular markers, genome resequencing, and digital gene expression (DGE) analysis. Researchers from China have completed the draft genome sequence [3] and fine genome map [11] of B. mori as well as the map of genome variation in 40 domestic and wild-type silkworms [12]. The fine genome map of B. mori can reach up to 99.6% coverage. Moreover, 76.7% of the genomic fragments and 82.2% of the genes have been localized to their respective B. mori chromosomes. These advances have greatly accelerated the localization and identification of the mutated genes.
Fuyin-lre is a novel red egg mutant that was first discovered in the B. mori germplasm resources of Fuyin. This mutant differs from previously reported red egg mutants because Fuyin-lre presents embryonic lethality. Fuyin-lre eggs stop developing approximately 48 h after incubation, thereby exhibiting recessive lethal mutation. Thus, significant impacts may be achieved with the maintenance of this variety. In the present study, we analyzed the embryonic development and mode of inheritance of Fuyin-lre and compared differentially expressed genes (DEGs) in wild Fuyin and mutant Fuyin-lre embryos at different developmental stages. The structural variation of the red egg gene re (BGIBMGA003497-1) was found to be the direct reason for the red egg trait in Fuyin-lre. The embryonic lethality of Fuyin-lre may be related to silencing of all 8 genes closely linked to red egg gene re. The present work provides a good reference for future research on the mechanisms of embryonic lethality and development.

Silkworm strains
Wild-type Fuyin ( Fig 1A) and its mutant Fuyin-lre with embryonic lethality (Fig 1B) were preserved by the Sericultural and Apicultural Research Institute, Yunnan Academy of Agricultural Sciences. The red egg mutant re (Fig 1C) was obtained from Southwest University. All B. mori individuals were fed with fresh mulberry leaves at a constant temperature of 25 ± 0.5°C and constant humidity of 75%-80%.

Genetic analysis of Fuyin-lre
Normal eggs of the Fuyin-lre ring were conventionally reared until they reached adulthood. Inbreeding of Fuyin-lre and its cross breeding with the red egg mutant re were performed. The segregation of normal and red eggs in F 1 offspring from inbreeding and cross breeding was investigated. Inbreeding was performed, in turn, on the offspring obtained from cross breeding, and subsequent segregation of F 2 was determined.
Anatomy of the embryos, and extracting total RNA extraction and genomic DNA from eggs Eggs laid by Fuyin, Fuyin-lre, and re were separately treated and incubated [13]. Every 24 h after incubation, the eggs were treated with 20% NaOH at 100°C for 3-5s, and then, eggs were blew with a dropper to separate the embryo from eggshell in clean water. The embryonic morphology of these eggs was observed under a microscope to determine the developmental stages of the embryo.
Every 24 h after incubation, 0.5 g of normal eggs laid by Fuyin as well as red eggs laid by Fuyin-lre and re was collected for total RNA extraction using the RNAiso Plus (TaKaRa) kit. The extracted RNA was digested by DNase I (TaKaRa) to remove DNA residues. The concentration and purity of total RNA was determined using a BioMate 3S UV-visible spectrophotometer (Thermo Fisher Scientific), and the quality of total RNA was identified using an Agilent bioanalyzer (Agilent Technologies 2100).The qualified RNA was used in DGE and qRT-PCR.
Normal eggs laid by Fuyin and the red eggs laid by Fuyin-lre and re were selected to extract their genomic DNA according to the literature [14]. Each qualified DNA sample was confirmed by the BioMate 3S UV-visible spectrophotometer before dilution to 100 ng/μL and storage at-20°C.

Illumina sequencing
Total extracted RNA from normal eggs from Fuyin and red eggs from Fuyin-lre at different developmental stages were selected according to the anatomical results of Fuyin-lre embryos. mRNA enrichment was performed using magnetic beads with Oligo (dT), and the RNA concentration of the samples was between 416 and 680 ng/μl. Relevant reagents were added to the mRNA obtained for fragmentation at high temperature. The fragmented mRNA was then used as a template to synthesize cDNA. Magnetic bead purification, end repair, addition of base A to the 3ʹ-end, and addition of adaptors were performed prior to PCR amplification. The library was subsequently constructed. Library quality and production were determined by an Agilent 2100 Bioanalyzer and an ABI StepOnePlus Real-Time PCR System, respectively. Qualified mRNA was sequenced by IlluminaHiSeq 2000. Raw data presented in this publication have been deposited in the NCBI Sequence Read Archive (http://www.ncbi.nlm.nih.gov/sra/) and are accessible through SRA accession number: SRP053552.

Data analysis
Original images produced by the sequencer were converted into raw data by Base Calling software. Low-quality and adaptor sequences were removed to obtain clean reads for subsequent analysis. Reference genome sequences of B. mori were downloaded from SilkDB (http:// silkworm.swu.edu.cn/silkdb/), and clean reads were aligned with reference genome sequences using Tophat software [15]. Gene expression was calculated by the RPKM (reads per kb per million reads) algorithm [16]. The expression levels of genes in different samples were compared to screen differentially expressed genes (DEGs). The sequencing-based method reported by Audic et al. [17] was adopted to analyze DEGs at two time points. The P-value of the difference test was corrected by multiple hypothesis testing, and the threshold of P-value was determined by controlling the false discovery rate (FDR). FDR is a technique used to determine the threshold of P-value during multiple comparison and analysis. The smaller the FDR, the larger the fold change and the more significant the expression difference. Genes with FDR 0.01 [18] and log 2 ratios 1 were considered DEGs. The DEG sequences were BLAST-searched, mapped, annotated, and analyzed using Kyoto Encyclopedia of Genes and Genomes (KEGG) metabolism pathways and Blast2GO software (version 2.7.2) according to the Blast2GO Tutorial [19].

qRT-PCR analysis
The criteria for DEGs screening were based on the results of DGE analysis. qRT-PCR detection was performed on screened genes to verify the reliability of DGE analysis. Based on the embryonic anatomy of Fuyin-lre, 2 μg of the total RNA of normal eggs by Fuyin, red eggs by Fuyinlre, and red eggs by re were used to synthesize first-strand cDNA with a PrimeScript reverse transcriptase kit (TaKaRa) in a 20 μL reaction system. The cDNA products were diluted by fivefold with ddH 2 O. Each 20 μL qRT-PCR reaction system consisted of 9 μL of SYBR Premix Ex Tag (Roche, 2×), 0.8 μL of the specific primers, 1 μL of cDNA template, and 9.2 μL of ddH 2 O; here, three replicates were produced. qRT-PCR was performed in a StepOnePlus realtime PCR system (Applied Biosystems) using default parameters. The housekeeping B. mori actin 3 gene (GenBank ID: NM_001126254) [20] was used as a reference to eliminate bias among samples, and qRT-PCR results were converted and calculated via the 2 -ΔΔCt method [21]. The primers used for qRT-PCR are shown in S1 Table.

Identification of mutation site
To investigate the mutation site, several pairs of primers were designed for PCR amplification among Fuyin, Fuyin-lre, and re according to the results of DGE and qRT-PCR. The reference genomic sequence was obtained by blasting against the silkworm genome database (http:// www.silkdb.org/silkdb/). Bmactin 3 gene was used as a reference to eliminate difference among samples. PCR reactions were carried out under the following conditions: 1 cycle of 94°C for 5 min, 31 cycles of 94°C for 30 s, 55-58°C for 30 s and 72°C for 1 min, with a final extension step of 5 min at 72°C. About 6 μL of amplified DNA products were then separated on 1.0% agarose gels, and visualized with UV transilluminator. The primers used for PCR amplification are shown in S2 Table.

Results and Discussion
Inheritance rule of Fuyin-lre Single batch rearing and inbreeding was performed on normal eggs of Fuyin-lre to determine the inheritance rule of the mutants. For each generation of inbreeding, normal and mutant egg batches were present. After nine generations of inbreeding, four batches with the red egg mutation were randomly selected to count the number of normal and red eggs. Results showed that the ratio of normal eggs to red eggs was 3:1. The hatching rate of normal eggs exceeded 90%, whereas that of the red eggs was 0% ( Table 1). Results of inbreeding showed that Fuyin-lre is controlled by a recessive gene. Red egg individuals in the batch showed the genotype lre/lre, whereas normal individuals showed two genotypes (+/+ and +/lre). The +/+:+/lre:lre/lre ratio was 1:2:1.
To determine whether or not the re gene controls the red egg trait of Fuyin-lre, normal individuals of Fuyin-lre (+/+ or +/lre) were hybridized with the red egg mutant re. The F 1 generation of the reciprocal cross showed normal and red egg batches. All of the eggs in the normal egg batch were normal. The ratio of normal egg to red egg numbers in the red egg batch was 1:1, and the eggs of both colors could be normally hatched. The normal and red eggs of the F 1 generation in the red egg batch were subsequently collected and separately reared. And then, three mating patterns were observed: normal × normal (N×N), normal × red (N×R), and red × red (R×R). Results indicated that red eggs appear in all batches of N×N, with a normal egg to red egg number ratio of 3:1. Red eggs also appeared in all batches of N×R with a segregation ratio of 1:1. All eggs in R×R batches were red. All eggs of the three different phenotypes were incubated by pickling, and all of the eggs in the batches showed segregation ratios of 3:1 or 1:1 could be hatched normally. The hatching rate of red eggs in the all-red-egg batches was 72%-74%. The ratio of the number of hatched eggs to that of un-hatched eggs satisfied the law of segregation ( Table 2). The results described above indicate that the gene controlling the red egg trait of Fuyin-lre is re. As all of the red eggs in batches of Fuyin-lre could not be hatched, the genes related to embryonic lethality are probably the genes very closely linked to re.

Embryonic development of Fuyin-lre
To determine the last stage of development of Fuyin-lre embryos, red eggs were dissected with the method mentioned in the Materials and Methods. The Fuyin-lre embryos strongly adhered to the egg shell, thereby rendering dissection difficult. Intact embryos were difficult to obtain, and a large amount of adherent substances were present; thus, only a rough contour of the embryo was visualized (Fig 2). The embryonic body of the aborted embryos was fairly long and showed an enlarged head and tail folds. Depressions on the head fold, abdomen, and tail were only indistinctly visualized. No longitudinal groove was observed in the first and the second segments. This morphology resembles that of the normal embryo at approximately 48 h of incubation. Therefore, we hypothesized that Fuyin-lre embryos stop developing after approximately 48 h of incubation.

DGE profiles
The embryonic anatomy of Fuyin-lre showed that the embryos stop developing after approximately 48 h of incubation. Therefore, the eggs of wild-type Fuyin and its mutant Fuyin-lre obtained after 24 and 48 h of incubation were selected as materials to construct four DGE tag libraries. High-throughput sequencing was performed with HiSeq 2000 (Illumina) to analyze differences in transcriptional levels in Fuyin and Fuyin-lre embryos at 24 and 48 h.
Sequencing produced 11.7-12.3 million raw reads for each sample (S1 Fig). After removal of adaptor-containing reads, reads with an N proportion larger than 10%, and low-quality reads (the number of bases with Q 5 accounted for over 50% of raw reads), the number of clean reads accounted for over 99.5% of the raw reads of each example; these values far exceed the lowest standards of sequencing (clean reads/raw reads 90%, N 2%, adaptor 2%). Sequencing saturation analysis can be used to determine whether or not the reads meet the required standard. In this study, sequencing results showed that the number of detected genes increases very slowly when the number of reads is approximately 3.0 M (S2 Fig). These results confirm the high reliability of the sequencing results.

DEGs analysis
The DEGs of Fuyin and Fuyin-lre obtained after 24 and 48 h were detected using the RPKM algorithm [16] and by referring to the detection method of Audic et al. [17]. Here, FDR 0.001 and log 2 ratios 1 were selected for DEG screening. The results identified 82 DEGs in Fuyin and Fuyin-lre at 24 h; of these genes, 30 genes were upregulated and 52 genes were downregulated. At 48 h, 237 DEGs in Fuyin and Fuyin-lre were detected, among which 99 genes were upregulated and 138 genes were downregulated. DEGs with the same expression trends at the two time points (simultaneously upregulated or downregulated) were analyzed to confirm screening accuracy. A final total of 48 DEGs were selected for subsequent analysis; here, 17  genes were upregulated and 31 genes, including the red egg gene re (BGIBMGA003497-1), were downregulated (S3 Table). Genetic analysis indicated that genes related to the embryonic lethality of Fuyin-lre may be genes closely linked to re. Thus, the 48 DEGs were aligned with the reference genome (http:// www.silkdb.org/silkdb/), and 8 genes (including re, Table 3), namely, BGIBMGA003497-1 (re), BGIBMGA003697, BGIBMGA003698, BGIBMGA003699, BGIBMGA003496, BGIBMGA003700, BGIBMGA003701, and BGIBMGA003495, which were closely linked to re were found. BGIBMGA003495-TA is the gene farthest from re, with a distance of approximately 250 kb (Fig 3). Alignment with the reference genome revealed 9 genes in this region. Besides the 8 DEGs previously mentioned, another gene BGIBMGA003696 was relevant, although this gene was not included in the DGE results. According to DGE analysis of two genes (BGIBMGA003497-2 and BGIBMGA003702) flanking this region, no obvious differences in the expression of the two genes in Fuyin and Fuyin-lre could be observed.

qRT-PCR
To verify the reliability of DGE analysis and compare differences in the expression levels of 9 DEGs (including BGIBMGA003696) between re and BGIBMGA003495 in Fuyin, Fuyin-lre, and re mutants, qRT-PCR was performed. All 8 genes except BGIBMGA003696 were highly expressed in Fuyin and re eggs, but the expression of these genes was nearly absent in Fuyin-lre (Table 3 and Table 4). However, gene BGIBMGA003696 was absent in all of three traits. This result is consistent with the DGE results, which are considered to be of very high reliability.
The results described above show that re mutation did not influence the expression of other linked genes in the re mutant. Thus, embryonic development proceeded normally except for the fact that the eggs were red-colored. The embryonic lethality of Fuyin-lre may be attributed  to silencing of all genes closely linked to re. The region where the 9 genes are located may have undergone significant structural changes, thereby causing abnormal expression of all the genes in this region. Consequently, the Fuyin-lre embryos did not develop normally.
Structural analysis of re and its neighboring sequences The hypothesis of structural variation was tested by PCR amplification in the region where the 9 genes described above are located using the Fuyin, Fuyin-lre, and re genomes as templates.
Results indicated that several pairs of primers could produce good amplification effects when the Fuyin and re genome was used as a template (Fig 4). However, when the Fuyin-lre genome was used as a template, the sequences could not be amplified by PCR. Sequencing results showed that mutation of the re gene in the re mutant occurs differently from the previously reported mutation in exon 9 of the re gene [8]. In this study, a frameshift mutation occurred because of base deletion in exon 1 and the presence of an additional base in exon 8 of the re gene ( Fig 5). The stop codon that terminates protein translation appeared at the front end of the ORF. Thus, the re gene could be transcribed but not translated into the corresponding proteins. This phenomenon could explain the red color of re eggs. As demonstrated above, complete functional loss of the re gene only produced the red color of the mutant but did not cause embryonic lethality. Given the structural variation of Fuyin-lre, the re gene could not be transcribed normally (Table 4), which is consistent with the DGE and qRT-PCR results. Therefore, the hypothesis of structural variation in the re gene and its neighboring sequences in Fuyin-lre mutants was confirmed. The genes between BGIBMGA003497-1 and BGIBMGA003495 (BGIBMGA003696 excluded) were expressed in normal eggs and, except for the re gene, have unknown functions (Table 3). Therefore, the specific genes connected with embryonic development are difficult to determine. The specific mode of mutation and key genes related to embryonic development must be identified through further study.

Conclusion
The Fuyin-lre mutant is a newly discovered red egg mutant with embryonic lethality. The mutant produced red-colored eggs and failed to develop normally because of embryonic lethality. Genetic analysis of Fuyin-lre revealed that the Fuyin-lre mutant is recessive and that the recessive homozygote presents embryonic lethality. DGE analysis and qRT-PCR were performed to compare DEGs in Fuyin and Fuyin-lre; 9 genes closely linked to re (including BGIBMGA003696) demonstrated structural variation. We confirmed that re mutation does not directly cause embryonic lethality. None of the red eggs of Fuyin-lre individuals developed normally, which suggests that genes causing the embryonic lethality of Fuyin-lre are closely linked to the re gene. This result is consistent with the hypothesized structural variation of the 9 genes closely linked to re in Fuyin-lre. The results of the present work may help guide future investigations on the underlying mechanisms of Fuyin-lre mutation and B. mori embryonic development.
Supporting Information