Bmovo-1 Regulates Ovary Size in the Silkworm, Bombyx mori

The regulation of antagonistic OVO isoforms is critical for germline formation and differentiation in Drosophila. However, little is known about genes related to ovary development. In this study, we cloned the Bombyx mori ovo gene and investigated its four alternatively spliced isoforms. BmOVO-1, BmOVO-2 and BmOVO-3 all had four C2H2 type zinc fingers, but differed at the N-terminal ends, while BmOVO-4 had a single zinc finger. Bmovo-1, Bmovo-2 and Bmovo-4 showed the highest levels of mRNA in ovaries, while Bmovo-3 was primarily expressed in testes. The mRNA expression pattern suggested that Bmovo expression was related to ovary development. RNAi and transgenic techniques were used to analyze the biological function of Bmovo. The results showed that when the Bmovo gene was downregulated, oviposition number decreased. Upregulation of Bmovo-1 in the gonads of transgenic silkworms increased oviposition number and elevated the trehalose contents of hemolymph and ovaries. We concluded that Bmovo-1 was involved in protein synthesis, contributing to the development of ovaries and oviposition number in silkworms.


Introduction
In Drosophila, isoforms of OVO-B and OVO-A are produced by two major classes of germline ovo transcripts driven by the ovo-A and ovo-B promoters [1,2]. Ovo-B and ovo-A transcripts differ only in their short first exons; OVO-B has 374 fewer residues than OVO-A. Ovo-B mRNA encodes the OVO-B isoform from an AUG initiation codon in exon 2, and the ovo-A transcript encodes the longer OVO-A isoform. The OVO-B and OVO-A isoforms have a common C2H2 zinc-finger domain but different N-termini that include potential effector domains [3]. OVO-B and OVO-A are transcription factors with opposing regulatory activities that are required for female germline survival and oogenesis [4]. Four dominant-negative ovo alleles (ovo D ) strongly suggest that the OVO-A N-terminal region is functionally important [2,5]. All four ovo D mutations introduce novel in-frame AUG codons upstream of the ovo-B initiation codon that the result in ovo-B transcripts that encode slightly truncated ovo-A isoforms. OVO-B expression is high during oogenesis, but OVO-A is expressed very weakly and only in nearly mature follicles [2,4].
OVO-B is a transcriptional activator, and OVO-A is a transcriptional repressor. OVO-B positively regulates the ovarian tumor (otu) promoter and OVO-A represses target promoters [2,4,6,7]. During most of oogenesis, OVO-B is necessary and sufficient for female germline development. OVO-B isoforms supply ovo + function in the female germline and epidermis. OVO-A isoforms have dominant negative activity in both tissues and elevated expression of OVO-A causes maternal-effect lethality. The absence of OVO-A results in maternal-effect sterility, indicating that tight regulation of antagonistic OVO-B and OVO-A isoforms is critical for germline formation and differentiation [4].
The silkworm Bombyx mori is a silk-spinning insect. As a model of Lepidopteran, its genome sequence is established [14,15], and genes related to growth, development, metamorphosis, immunologic response, and fibroin synthesis have been comprehensively studied. However, little is known about genes related to ovary development. In this study, to understand the molecular mechanisms of silkworm nutrient metabolism and ovary development, four alternatively spliced isoforms of ovo genes were characterized. In this study, four alternatively spliced isoforms of Bmovo genes were cloned and investigated, Bmovo-1, Bmovo-2 and Bmovo-4 showed the highest levels of mRNA in ovaries, while Bmovo-3 was primarily expressed in testes. BmOVO-1, BmOVO-2 and BmOVO-3 all had four C2H2 type zinc fingers, but differed at the N-terminal ends, while BmOVO-4 had a single zinc finger. We hypothesized that Bmovo-1 might be involved in regulating protein synthesis. In the study, the transgenic and RNAi techniques were applied to upregulate and downregulate Bmovo-1 in silkworm ovaries to explore the mechanism on ovary size, protein synthesis, nutrition transportation and oviposition number regulated by Bmovo-1.

RNA isolation and cDNA synthesis
Total RNA was isolated from silkworm tissues using total RNA Isolation Kits (TaKaRa, Dalian, China), followed by treatment with DNaseI. cDNA was synthesized by PrimeScript Reverse Transcriptase (TaKaRa), following standard instructions.

Sequence analysis of Bmovo
A homology search and multiple alignments were carried out with BLAST (http://www.ncbi.nlm.nih.gov/BLAST) and Clustal W software [16]. A phylogenetic tree was constructed with MEGA5.0 using the neighbor-joining method [17] based on OVO protein sequences from GenBank. The validity of branches was tested by bootstrapping using 1000 replicates. Secondary structure, conserved motifs and protein function were predicted using the Conserved Domain Search Service program (http://www.ncbi. nlm.nih.gov/Structure/cdd/wrpsb.cgi), SMART software (http://smart.embl-heidelberg.de/).

BmOVO-1 antibody preparation
Bmovo-1 PCR product (2.4 kb) was digested with restriction endonuclease of BamHI and XhoI and ligated into the expression vector pET-28a (+) (Novagen, Darmstadt, Germany) to generate the recombinant plasmid pET-28a (+)-Bmovo-1. Fusion proteins were expressed in Escherichia coli strain BL21. Recombinant protein was purified with Ni-NTA agarose (Qiagen, Shanghai, China) and used to immunize Kunming mice (Soochow University, Suzhou, China) by subcutaneous injection. Antibody was confirmed by Western blotting. This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Committee on the Ethics of Animal Experiments of Soochow University of China (Permit Number: 201304326). All the experimental animals were performed under euthanasia, and all efforts were made to minimize suffering.

Immunohistochemistry
Cultured BmN cells derived from silkworm ovaries and cultured in TC-100 medium (Gibco BRL, Rockville, MD, USA) at 27uC were collected and fixed with 4% paraformaldehyde for 15 min, then rinsed with 0.01 M PBST (0.05% of Tween-20 in PBS), and incubated with mouse anti-BmOVO-1 antiserum at 4uC overnight. As a negative control, BmN cells were treated with preimmune antiserum. After rinsing with 0.01 M PBST three times, cells were incubated with FITC-conjugated goat anti-mouse IgG (Tiangen, Beijing, China) at 37uC for 1 h. Excess FITCconjugated goat anti-mouse IgG was removed and cells were stained with DAPI and observed with a fluorescence microscope.
For histochemical observations, gonads were dissected from the larvae of the beginning and the end of the fifth instar larval stage, fixed in 4% paraformaldehyde and embedded in paraffin. Sections were processed with mouse anti-BmOVO-1 antiserum and TRITC-goat anti-mouse IgG (Tiangen, Beijing, China).

Expression patterns of Bmovo genes
To analyze the Bmovo expression profile, cDNAs were synthesized by reverse transcription with 1 mg RNA extracted from tissues using PrimeScript Reverse Transcriptase. Primers for determining relative amounts of alternative spliced isoforms were designed based on cDNA sequences (Table S1). The B. mori housekeeping gene actin A3 was used as an internal control and quantitatively amplified with primers semi-A3-F and semi-A3-R. Q-PCR used a real-time PCR System (Bio-Rad CFX96). Relative expression of Bmovo genes was estimated according to the 2 2DDCt method [18].

Reducing Bmovo in silkworms with RNAi
Bmovo-specific ovo-dsRNA-164 (sense: 59-GCA CUG AUG GUG ACG UAU Utt-39, antisense: 59-AAU ACG UCA CCA UCA GUG Ctt-39) was synthesized by Gima Corporation (Shanghai, China). A total of 40 silkworm larvae were injected with 2 mg ovo-dsRNA-164 per larva at the first day of the third instar stage and reared on fresh mulberry at 25uC. Expression of Bmovo in gonads of injected larvae at the first day of the fifth instar stage was estimated by Q-PCR. Negative control larvae were injected with random siRNA-NC (sense strand: 59-UUC UCC GAA CGU GUC ACG Utt-39; antisense strand: 59-ACG UGA CAC GUU CGG AGA Att-39). Blank controls were not injected and were raised normally.
Sperm-mediated gene transfer was performed as described by Zhou [21]. The transgenic vector piggyBac-A3GFP-IeNeo-VlgBmovo-1 at 1 mg/mL was mixed at a ratio of 1:1 with a helper pigA3 vector containing the piggyBac transposase sequence under the control of the A3 promoter. The mixture was injected into the copulatory pouch of virgin moths (Strain Haoyue). Moths were free to copulate and oviposit. After spawning, embryos were incubated at 25uC at relative humidity 85-90% for 10 days until hatching. Larvae were reared on mulberry leaves at room temperature.
Newly hatched larvae were reared on mulberry leaves coated with 10 mg/mL G418 until approximately 10% remained. Surviving silkworms were observed under a stereomicroscope (Olympus SZX12, Tokyo, Japan) and fluorescent silkworms were fed normal mulberry leaves. Exogenous genes (EGFP and NEO) were identified with PCR from G2-generation transgenic silkworms using primers (DEGFP-1/DEGFP-2 and DENEO-1/ DENEO-2). Genomic DNA extracted from fluorescent G2generation moths was denatured by boiling and dotted onto nylon membranes. DNA hybridization was carried out with a DIG-labeled gfp probe. Membrane washing and signal detection used a DIG DNA labeling and detection kit (Roche, Mannheim, Germany), according to the manufacturer's instructions. Negative controls were genomic DNA from a normal silkworm; positive controls were plasmid pigA3GFP.

Determination of trehalose content and trehalase activity assay
Trehalose contents of hemolymph and ovaries were estimated by the method of Ge et al. [22]. Trehalase activity in hemolymph was determined by the method of Yamashita et al. [23] and enzyme activities were expressed as mmol/ml hemolymph?h.

Statistics
All data are presented as mean 6 standard deviation (SD). The percentages of data were transformed into arcsine squareroot data for analysis. Statistical differences were evaluated using Student's ttest for unpaired samples.

Cloning and characterization of the silkworm Bmovo gene
To characterize the Bmovo gene, predicted gene sequences were obtained in silico based on DmOVO protein sequences. Three PCR products of 2.4, 0.8 and 0.5 kb were amplified from testes and ovary cDNA. Sequences from ovary and testes cDNA samples were same. The longest sequence was 2421 bp (GenBank accession No. GU477588) encoding a protein (BmOVO-1) with 806 amino acids. Two shorter sequences of 781 bp (GenBank accession No. HQ831344) and 498 bp (GenBank accession No. JQ665224) encoded BmOVO-3 with 248 (35-781 nt) and BmOVO-4 with 165 amino acids ( Figure 1A). To detect alternatively spliced isoforms of Bmovo, primers Bmsovo-BH/ Bmsovo-M2 and Bmsovo-M/Bmsovo-HD were used and specific 1.1 kb and 0.65 kb fragments were amplified from gonadal cDNA. Sequences of the products overlapped for an isoform of 1731 bp (GenBank accession No. HQ831345), encoding a protein (BmOVO-2) of 576 amino acids.
The deduced amino acid sequence of BmOVO-1 was aligned with the sequences of the other 5 isoforms. BmOVO-1 shared 52.2% similarity with OVO of Apis mellifera (XP_624482), 44.6% similarity with Drosophila virilis (XP_002057918), 47.9% similarity with Pediculus humanus corporis (XP_002423789), 50.8% similarity with Solenopsis invicta (EFZ10998), and 52.9% similarity with Tribolium castaneum (XP_974881). Although the sequence identities of OVO proteins from different species were not high, C2H2 zinc-finger domains were conserved (data not shown), suggesting that the DNA-binding sequences for OVO proteins were similar in different species.
Phylogenetic analysis by the neighbor-joining method suggested that ovo genes from different insect species clustered, while vertebrate genes formed another group. In insects, subgroups were the fruit fly ovo genes and ovo genes from silkworms and other insects. Although both fruit flies and mosquitoes belong to Diptera, their ovo genes were located on different branches ( Figure S1).

Bmovo expression patterns in silkworms
To investigate Bmovo gene expression patterns, 8 tissues were collected from fifth instar larvae on the third day. The highest expression of Bmovo-1 was in ovaries and was 3 times the expression in testes. Bmovo-1 was also highly expressed in the head and silk glands. The expression profile of Bmovo-2 was similar to the profile for Bmovo-1 in different tissues with the highest relative expression in ovaries. In contrast, expression of Bmovo-3 was 7.5 times higher than in the ovary. The highest expression of Bmovo-3 was in testes, then silk glands and hemocytes. Bmovo-4 was expressed at a lower level in different tissues in third-day of fifth instar silkworms; expression of Bmovo-4 was higher in silk glands than in ovaries (Figure 2A). Testes expression was highest for Bmovo-1 followed by Bmovo-2, Bmovo-3, and Bmovo-4. Expression of Bmovo-1 was 20 times higher in testes than Bmovo-2 or Bmovo-3. Ovarian expression was highest for Bmovo-1 followed by Bmovo-2, Bmovo-4, and Bmovo-3. Expression of Bmovo-1 was 11 times higher than expression of Bmovo-2, 450 times higher than Bmovo-4 and 708 times higher than Bmovo-3 ( Figure 2B).
The four alternatively spliced isoforms of Bmovo were expressed differently at different developmental stages ( Figure 2B). In testes, transcription of Bmovo-1 was highest at the fourth day of the fourth instar stage followed by the seventh day of the pupal stage. Transcriptional of Bmovo-1 at the fourth day of the fourth instar stage was 3 times higher than at the third and seventh days of the fifth instar stage. Compared to the seventh day of the fifth instar stage, transcription of Bmovo-2 was 13.96 times higher at the fourth day of the fourth instar stage and 11.38 times higher at the seventh day of the pupal stage. Compared to the seventh day of the fifth instar stage, transcription was 2 times higher at the third day of the fifth instar stage. Transcription of Bmovo-4 was the highest at the seventh day of the fifth instar stage and was 1715 times higher than that at the seventh day of the fifth instar stage, 3.5 times higher than the fourth day of the fourth instar stage and 4.4 times higher than the seventh day of the pupal stage. In the ovary, transcription of Bmovo-1 was the highest at the fourth day of the fourth instar stage, and was 2.2 times higher than at the third day of the fifth instar stage, 1.5 times higher than at the seventh day of the fifth instar stage, and 2.0 times higher than at the seventh day of the pupal stage. Transcription of Bmovo-2 was the highest at the seventh day of the pupal stage and was 19 times higher than at the fourth day of the fourth instar stage, 38 times higher than at the third day of the fifth instar stage and 96.8 times higher than at the seventh day of the fifth instar stage. Transcription of Bmovo-3 was the highest at the fourth day of the fourth instar stage and was 1432.2 times higher than at the third day of the fifth instar stage, 1883.5 times higher than at the seventh day of the fifth instar stage and 0.4 times higher than at the seventh day of the pupal stage. Transcription of Bmovo-4 was the highest at the fourth day of the fourth instar stage and was lower level at the third day of the fifth instar stage, the seventh day of the fifth instar stage and the seventh day of the pupal stage.

BmOVO isoform analysis and immunohistochemistry
To examine the protein products from the Bmovo gene, total protein extracted from the testes and ovaries were subjected to SDS-PAGE and Western blotting. Specific protein bands of 89 kDa and 63 kDa representing BmOVO-1 and BmOVO-2, respectively, were detectable in both ovary and testis samples. A specific protein band of 18 kDa representing BmOVO-4 was detected only in the ovary samples and a specific protein band of   27 kDa representing BmOVO-3 was found only in testis samples ( Figure 3A).
Immunofluorescence showed that BmOVO was expressed in follicular cells, nurse cells, primary oocytes and spermatids at different stages ( Figure 3B). BmOVO in BmN cells was distributed in both the nucleus and cytoplasm, but mainly in nucleus ( Figure 3C).

Suppression of Bmovo expression in silkworms with
Bmovo-siRNA Expression of the Bmovo gene was reduced by injection of Bmovo-siRNA into silkworms at the third and fourth instar stage. In testes, Bmovo-1 transcription decreased 14% and Bmovo-2 transcription decreased 12% compared to the negative control. In ovaries of larvae in the first day of the fifth instar stage, Bmovo-1 transcription decreased 24% and Bmovo-2 transcription decreased 19% compared to the negative control. No difference was observed between the blank control and the negative control.

Screening and identification of transgenic silkworms overexpressing Bmovo-1
A transgenic vector was mixed with helper plasmid and injected into copulatory pouches of six copulated female moths. Eggs were incubated at 25uC. Fluorescence was observed in some eggs, indicating that vector was introduced into the eggs. Transgenic silkworms exhibited neomycin resistance if the neo gene was successfully introduced into the genome and expressed. Larvae were reared on mulberry leaves coated with 10 mg/mL G418 from the second instar stage. Obvious developmental differences and deaths were observed among silkworms in response to G418. Fluorescence was observed in some surviving larvae ( Figure 4A), pupae ( Figure 4B) and moths ( Figure 4C).
Genomic DNA of copulated moths was used amplify the egfp gene with primers DEGFP-1 and DEGFP-2. A 0.72 kb specific band with the egfp gene was amplified from fluorescent moths. G0 generation fluorescence moths were mated for a G1 generation. A G2 generation was generated by mutual crossing between fluorescent moths of generation G1. G3 and G4 generations were obtained in the same way. Fluorescent G2 silkworms were identified by PCR amplification using primer pairs DEGFP-1/ DEGFP-2 and DENEO-1/DENEO-2. Specific bands for egfp (0.72 kb) and neo (0.8 kb) were detected in samples from fluorescent silkworms ( Figure 4D). PCR products were purified and sequenced confirming egfp and neo genes in G2 moths and indicating that fluorescent G2 silkworms were transgenic. To verify that target genes were stably inherited in descendants of transgenic silkworms (silkworm +Bmovo-1 ), genomic DNA from G3 and G4 generation moths were amplified with primer pairs Neo-1/Neo-2 to detect products of neo gene ( Figure 4E). These results indicated that heterogeneous genes were stably inherited in transgenic silkworms. Dot hybridization showed that a egfp probe hybridized with genomic DNA from fluorescent G2-G5 moths, indicating that the fluorescent silkworms were transgenic ( Figure 4F).
To verify Bmovo-1 overexpression in silkworm +Bmovo-1 , transcription of Bmovo-1 in the gonads of G5 generation silkworms at the third day of the fifth instar stage was estimated by Q-PCR. Transcription of Bmovo-1 in silkworm +Bmovo-1 was 2.265 higher in ovaries and 1.829 times higher in testes than in wild type silkworms. Egg productivities per moth were 423 for the RNAi group, 505 for the negative control group and 506 for the normal group. Although egg productivity per moth in RNAi group was decreased by 16.40% compared with the normal group, disparity was not significant among the RNAi, negative control and normal groups for the rate of laid eggs. The rate of fertilized eggs was only slightly lower in the RNAi group than in the negative control and normal groups. However, the rates of hatched eggs from RNAi (R) 6 RNAi (=) and RNAi (R) 6normal (=) were significantly lower than that from normal (R) 6 RNAi (=); no differences in hatched egg rates were observed from normal (R) 6 RNAi (=) ( Table 1). Investigations of cocoon quantity from RNAi-injected silkworms showed increases of 10.47% for pupal weight and 11.16% for cocoon shell weight for females, and 0.96% and 10.79% for males compared to the blank control group (Table 2). In male silkworms in which the Bmovo-1 gene was upregulated in silkworm +Bmovo-1 , pupal weight decreased by 11.28% and cocoon shell weight decreased by 22.04% in female silkworm +Bmovo-1 and decreased by 11.37% and 17.09% in G5 generation males (Table 3). In the G6 generation, pupal weight and cocoon shell weight were also decreased. The oviposition number of G6 generation silkworm +Bmovo-1 increased 13.37% compared to wild type silkworms ( Table 4). Investigation of gonad development showed that ovary weight in silkworm +Bmovo-1 (G7 generation) worms was increased compared to wild type, but testes of silkworm +Bmovo-1 were less heavy than wild type ( Figure 5A and 5B).
To understand nutrient resorption in different tissues, trehalose contents of hemolymph and ovaries at the fifth day of the fifth instar stage of silkworm +Bmovo-1 were estimated. Trehalose was 338.78 mg/100 ml in the hemolymph of wild type silkworms and 377.29 mg/100 ml in G7 generation silkworm +Bmovo-1 . Trehalose was 3.06 mg/100 mg in ovaries of wild type silkworms and 3.39 mg/100 mg in G7 generation silkworm +Bmovo-1 . Compared to wild type, trehalose was increased by 11.37% in the hemolymph and 10.56% in the ovaries of silkworm +Bmovo-1 . Moreover, trehalase activity in hemolymph was also determined, but no significant difference was found between silkworm +Bmovo-1 and wild type (data not shown).

Silkworm ovaries have different Bmovo transcripts and isoforms
In B. mori, the ovo gene region produces at least 4 transcripts and is different from the D. melanogaster shavenbaby (svb)-ovo gene region. The open reading frame (ORF) of the Bmovo gene consists of 5 exons. Bmovo-1 mRNA encodes only the BmOVO-1 isoform from the first AUG initiation codon in exon 1. BmOVO-3 and BmOVO-4 isoforms initiate from the second AUG initiation codon in exon 1 and the initiation codon for BmOVO-2 isoform is the first AUG in exon 3b. D. melanogaster shavenbaby (svb)-ovo has 5 annotated transcripts and 5 annotated polypeptides of 975, 1351, 1222, 1208 and 1351 amino acids. Bmovo has at least 4 transcripts and 4 predicted polypeptides of 806, 576, 248 and 165  D. melanogaster Svb/OVO isoforms contain four C2H2 zincfinger motifs at the C-termini but have different N-termini that include potential effector domains. OVO-B and OVO-A are transcription factors with opposing regulatory activities [8]. The activation region contains a glycine-rich region (61% charged residues), two acidic regions (pI = 3.4, 3.7) and an extensive glutamine/histidine-rich region; the repression region contains a charged basic region (57% charged residues, pI = 11.9) and a serine-rich domain (46% charged residues) [4]. In B. mori, all BmOVOs have a common C2H2 zinc-finger domain, suggesting that BmOVO is a transcription factor that binds to promoters. BmOVO-1, BmOVO-2 and BmOVO-3 have four common C2H2 zinc-finger domains and BmOVO-4 has one. In Drosophila, OVO-B and OVO-A likely compete for similar binding sites. This might result in cross-regulation of target genes by the opposing activities of OVO-B and OVO-A [4]. The four BmOVO isoforms might also compete for binding sites and crossregulate expression of potential target genes.
BmOVOs differed at N-termini where the potential effector domains were located. BmOVO-1 had 6 acidic regions (A1-A6) and 2 basic regions (B1-B2); BmOVO-2 had 2 acidic regions (A4 and A5) and 2 basic regions (B1-B2) and BmOVO-3 had only an acidic region (A6). No effectors were found in BmOVO-4, suggesting that the difference in the conserved domain might result in a functional difference among the different BmOVO isoforms. The distribution and number of acidic and basic regions at the N-termini of BmOVO sequences differed from DmOVO, suggesting differences in function between BmOVO and DmOVO. In Drosophila, OVO-B and OVO-A are required for different developmental events; Dmovo is abundant in the ovary of adults and at the embryonic development stage of 00-02 hours [24]. Expression of GAL-tagged OVO proteins showed that overall OVO expression was strong during oogenesis, while OVO-A was expressed weakly and only in nearly mature follicles [2]. Consistent with this differential expression pattern, the germlinespecific expression of ovo in females correlates with its function in oogenesis. This expression, however, is also seen in males that do not require ovo [2,4]. Expression pattern of Bmovo gene and its protein in silkworms We found that in contrast to expression of Dmovo, Bmovo was expressed in several tissues of the fifth instar larvae of silkworms. The highest amounts of Bmovo-1 and Bmovo-2 mRNA were in ovaries. The highest amounts of Bmovo-3 were in testes. Bmovo-4 was mainly expressed in silk glands and ovaries, while Dmovo-B expression in larvae is restricted to gonads in both sexes [2]. These results suggested that BmOVO functions were not identical to DmOVO. Transcriptions of Bmovo-1, Bmovo-3 and Bmovo-4 were controlled by the same promoter but the transcripts had different AUG initiation codons, suggesting that differences in mRNA abundance of Bmovo-1, Bmovo-3 and Bmovo-4 might result from alternative splicing. In contrast to the expression pattern of Bmovo-1, Bmovo-3 and Bmovo-4, the expression pattern of Bmovo-2 might be determined by its own promoter. Western blotting showed that BmOVO-1, BmOVO-2 and BmOVO-4 were mainly expressed in ovaries, while BmOVO-3 was mainly expressed in testes. However, whether gonad development is influenced by the interaction of the 4 alternatively spliced isoforms is unclear.
The cysts of 64 primary spermatocysts emerge before the beginning of the third instar stage, then produce larger cysts of secondary spermatocysts from the fourth instar to the fifth instar stages, entering the second meiotic division that results in sperm bundles with mature sperm at the end of the fifth instar stage [25]. Q-PCR showed that transcription of the four Bmovo isoforms in testes changed with development. Levels of Bmovo-1, Bmovo-2 and Bmovo-3 at the fourth day of the fourth instar stage were higher than levels at the seventh day of the fifth instar stage. The highest transcription of Bmovo-4 was at the seventh day of the fifth instar stage, and the lowest transcription was at the third day of the fifth instar stage. Ovo expression is not required in Drosophila males [2,4]. Therefore, the function of Bmovo in spermatogenesis of silkworms and Drosophila is worth further study.
Bmovo impacting silkworm economic traits and production traits and trehalose metabolism Tight regulation of antagonistic OVO-B and OVO-A isoforms is critical for germline formation and differentiation in Drosophila [4]. In B. mori, after Bmovo-1 and Bmovo-2 genes were silenced, the egg productivity per moth decreased by 16.40% compared with the control, suggesting that both Bmovo-1 and Bmovo-2 were required for oogenesis.
Although the rate of fertilized eggs was not different among the RNAi, negative control and normal groups, the rates of hatched eggs of both the RNAi (R) 6 RNAi (=) group and RNAi (R) 6 normal (=) group were significantly decreased compared with the control ( Table 1), suggesting that repressing expression of Bmovo-1 and Bmovo-2 had little effect on spermatogenesis but influenced the development of embryogenesis and embryos. An investigation of cocoon quantity showed that pupal and cocoon shell weights in females increased and pupal weight in males did not change. However, cocoon shell weight increased in the RNAi group, indicating that silk protein synthesis was elevated by repressing Bmovo-1 and Bmovo-2.
The weight of both pupae and cocoon shells in silkworm +Bmovo-1 decreased compared with the control, while oviposition number increased by 13.37%. These data indicated that oogenesis and silk protein synthesis were influenced by regulating Bmovo-1 expres- Bmovo-1 Regulates Ovary Size PLOS ONE | www.plosone.org sion. Moreover, the gonadal weight was affected by overexpressing Bmovo-1 in silkworms. Ovary weight in silkworm +Bmovo-1 increased while testis weight decreased slightly, indicating that the function of BmOVO-1 was similar to the function of DmOVO-B. Genetic and molecular data indicated that Dmovo acted upstream of otu + and ultimately Sxl + . OVO-B positively regulated otu transcription in the female germline, suggesting that part of the function of OVO-B was upregulating Otu production. Maternal OVO-A was required for the progeny germline; OVO-A was extremely toxic when produced during early oogenesis. OVO-A repressed target genes expression is known to be part of the genetic hierarchy, including ovo [4]. BmOVO-1 acted similarly to OVO-B in Drosophila and might be a transcriptional activator that positively regulated otu and sxl transcription in the female silkworm germline. Overexpression of Bmovo-1 in silkworm gonads decreased fib-L transcription, reducing cocoon shell weight. Germline-specific expression of ovo in Drosophila females correlates with function in oogenesis. This expression, however, is also observed in males that do not require ovo [2,4]. In silkworm +Bmovo-1 , testis weight decreased slightly; nevertheless, the transcript levels from genes associated with spermatogenesis (aly, achi) were not significantly changed, suggesting that BmOVO-1 had no effect on spermatogenesis. Trehalose is a non-reducing disaccharide comprising two glucose molecules. It is the main haemolymph (blood) sugar with high concentrations presented in insects, which is hydrolysed by trehalase (E.C. 3.2.1.28) [26]. In silkworm +Bmovo-1 , the trehalose content in the hemolymph and ovary were respectively elevated by 11.37% and 10.56%. In general, the ovary absorbs the nutrient from the hemolymph for egg development. Bmovo-1 overexpression in silkworm ovary may promote the transportation of lipid and/or sugar, which are conducive to ovary development and the excessive consumption of nutrient for ovary development altered nutrient partitioning and finally deters the silk protein synthesis ( Figure 6).