BmAbl1 Regulates Silk Protein Synthesis via Glutathione Metabolism in Bombyx mori

Simple Summary Silk yield is the primary economic character in sericulture. During the past 5000 years of domestication, great effort has been made to increase cocoon shell weight (CSW). However, silk yield is a complex trait controlled by quantitative trait loci (QTLs). From these QTLs, BmAbl1 tyrosine kinase, which is located at chromosome 1, is considered to be closely related to CSW. In our study, the functional loss of BmAbl1 led to altered glutathione metabolism, which ultimately affects the amino acid synthesis and metabolic pathway, impairing silk fibroin protein synthesis. We further investigated the role of BmAbl1 in regulating silk fibroin secretion. A new perspective was taken on factors affecting silk fibroin synthesis. This provides a new idea to improve silk fibroin yield. It lays the groundwork for improving silk yield in the future. Abstract Bombyx mori, domesticated from wild silkworms, is an economic insect that feeds on mulberry leaves and produces silk. In the current study, we demonstrated the contribution of BmAbl1 in silk protein synthesis. The inhibition and knockout of BmAbl1 can reduce the larva weight and CSW. The effect on CSW of BmAbl1 is not on the transcriptional level, but on the translational level. RNA-sequencing data suggested that amino acid synthesis and the metabolism process had a great difference between the BmAbl1- and Control strain, particularly glutathione metabolism. An abnormality in glutathione metabolism led to the reduction of free glycine and serine content, which are the main components of fibroin protein. Finally, fibroin protein synthesis has been reduced, including fibroin-heavy chain, fibroin-light chain, and p25 protein. This finding brought to light the role of BmAbl1 in the silk protein synthesis process.


Introduction
Silkworms have been domesticated over the past 5000 years from the wild progenitor. To increase the yield of sericulture, one of the main tasks in domestication is to screen a strain with high cocoon quality. The cocoon contains two main structural proteins, fibroin and sericin. Fibroin is the primary composition that accounts for 70-80% of the weight of the silk fiber. Sericin usually accounts for 20-30% of the weight of the silk fiber. Fibroin consists of the heavy chain, light chain, and P25 [1,2]. The molar ratio between these is generally 6:6:1 [3]. The amino acid composition of the heavy chain occupying most of the fibroin is dominated by Gly (46%), Ala (30%), Ser (12%), Tyr (5.3%), and Val (1.8%) [4].
Traditional breeding employs a cross-breeding method based on genetic theory. In recent years, molecular breeding assisted by QTL mapping has provided new technical support for silkworm selection. Numerous researchers have confirmed that a QTL region on chromosome 1 makes a significant contribution to cocoon shell weight (CSW) [5][6][7][8]. Within this QTL region, the tyrosine protein kinase, BmAbl1, is considered a potential target gene affecting CSW [9,10].
Abl1, a conserved non-receptor tyrosine kinase, plays key roles in various signaling transform pathways [11]. Many receptors send signals and link extracellular stimuli to other signaling pathways via Abl1, such as Notch, Robo, Eph, and others [12][13][14][15]. The signal has been implicated in the regulation of cytoskeletal rearrangement, axon patterning, cell proliferation, migration, and protein secretion [11,16,17]. The Drosophila Abl gene has a great contribution to nervous development, and controls growth cone guidance and synaptogenesis, acting downstream of the Robo receptor [12,18]. In mammals, it is an activity factor of the Ras/Raf/MAPK pathway and PI-3K/Akt pathway regulating the proliferation and differentiation of cells [19]. In B. mori, the overexpression of Ras1 in the posterior silk gland can increase cell size and enhance fibroin production [20,21]. However, whether and how BmAbl1 contributes to the synthesis of silk protein, and how it works, are still unclear.
In the current research, we inhibited the function of BmAbl1 using a blocking agent and gene knockout, respectively. Compared with the Control group, BmAbl1 --suppressed individuals showed lighter larva weight and cocoon weight. The expression of silk protein genes was detected by qRT-PCR, but there was no difference between BmAbl1and the Control group. The RNA-sequencing (RNA-Seq) results indicated that amino acid synthesis and the metabolism pathway were affected, especially glutathione metabolism. An abnormality of glutathione metabolism caused glycine (Gly) and serine (Ser) reduction in BmAbl1 -. Ser and Gly are the major components of fibroin protein. The free amino acid test showed that the content of Gly and Ser in the posterior silk gland (PSG) of BmAbl1was significantly lower than that of the Control group. Correspondingly, the content of fibroin proteins was decreased in the PSG of BmAbl1 -.

Sample Preparation
Nistari, a multivoltine and non-diapausing strain of Bombyx mori, was obtained from the Sericultural Research Institute of Chinese Academy of Agricultural Science. The larvae were reared on sufficient fresh mulberry leaves under a constant 12-h light/12-h dark photoperiod at 25 ± 1 • C and 75% ± 3% relative humidity.

Inhibited BmAbl1 Using Dasatinib
Abelson tyrosine kinase was inhibited by Dasatinib (Adooq Bioscience, Nanjing, China). An amount of 1 mg of Dasatinib was dissolved in 205 mL H 2 O to prepare the stock solution. The stock solution was diluted to 0.6 nm and smeared on the fresh mulberry leaves. First-day third-instar larvae of B. mori were reared with inhibitor-treated leaves until the wandering stage. The Control group larvae were reared with H 2 O-treated leaves. Each group had three biological replicates. Each replicate contained thirty larvae. All larvae were taken for the same treatment.

Measurement of Tyrosine Kinase Activity
The Protein Tyrosine Kinase (MEIMIAN, Yancheng, China) ELISA Kit was used to measure tyrosine kinase activity. The PSGs of the Treated and Control groups were dissected from the wandering stage of fifth-day fifth-instar larvae. Each group had three biological replicates. Each replicate contained five larvae.
After dissecting the posterior silk gland of the fifth-day fifth-larvae, 100 mg tissues were checked and rapidly frozen with liquid nitrogen. The samples were maintained at 2-8 • C after grinding, and 1 mL PBS (PH7.4) was added. The samples were homogenized by hand, and centrifuged for 20 min at a speed of 2000-3000 rpm to remove the supernatant.
A volume of 40 µL of the sample dilution was added to test the sample well, and then 10 µL of the testing sample was added. HRP-Conjugate reagent (100 µL) was added to each well, except the blank well. It was incubated for 60 min at 37 • C. Then, the liquid Insects 2022, 13, 967 3 of 13 was discarded; it was dried by swing; washing buffer was added to every well; it was stilled for 30 s; and then drained; this was repeated 5 times. Chromogen Solution A (50 µL) and B (50 µL) was added, respectively. It was reacted in the dark for 15 min at 37 • C. The stop solution was added, and, finally, absorbance was read at 450 nm after adding the stop solution within 15 min by SpectraMax Paradigm (Molecular Devices, CA, USA).

cDNA Synthesis and Quantitative Real-Time PCR
To validate RNA-Seq, total RNA was extracted from three biological samples of Treated and Control PSGs using RNAiso Plus (TaKaRa, Dalian, China). For each sample, three pairs of PSGs from the Nistari or BmAbl1were combined. Gel electrophoresis and ultraviolet spectrophotometry were used to determine the integrity and purity of the RNA. One microgram of total RNA from each sample was used to synthesize cDNA using a PrimeScript RT Reagent Kit with gDNA Eraser (Perfect Real Time, TaKaRa)-followed by storage at −20 • C. Real-time qRT-PCR was carried out in a reaction volume of 20 µL, containing 2 µL of template, 10 µL of 2× SYBR Premix EX Taq (TaKaRa), and 0.4 µL of specific primers (10 µM). The PCR amplification efficiency (E) and R [2] of each primer pair were calculated from the slope of a standard curve, which was conducted according to MIQE (Minimum Information for Publication of Quantitative Real-Time PCR Experiments) guidelines. The qRT-PCR primer sequences, which were designed based on the consensus sequence of each alignment based on the assembled RNA-Seq, and their efficiencies, are provided in the Supplementary Material (Table S1). Real-time qRT-PCR was performed with a Roche lightCycler 96 real-time PCR system, using the following conditions: 95 • C for 5 min, followed by 40 cycles of 95 • C for 10 s, 60 • C for 40 s, and dissociation. The GAPDH gene was used as the reference. Three repeated experiments were set up. The expression of the target gene was calculated by the Ct value, and graphics were created based on these data.

Plasmid Construction
The BmAbl1strain was established based on a binary transgenic CRISPR/Cas9 system. An effector plasmid pXL-[RFP-U6-BmAbl1-sgRNA], which expresses sgRNA targeting BmAbl1 under the control of the silkworm small nuclear RNA promoter, U6, was constructed through a series of cloning steps. The primers used in the cloning are included in the Supplementary Material (Table S2).

Silkworm Germline Transformation
Germline transformation was performed on fresh eggs, within 6 h of egg laying. Thirty positive G0 individuals with red fluorescence were selected from 160 eggs injected with the transgenic system. Preblastoderm embryos were injected with transgenic plasmid mixed with the helper plasmid, and incubated in a humidified chamber at 25 • C for 10-12 days until hatching. The G0 moths were sib-mated or backcrossed with WT moths, and the presence of the selection marker gene in G1 progeny was scored during the late embryonic stage using a fluorescence microscope (Leica M165FC). Heterozygous progeny between IE1-Cas9 and pXL-[IE1-DsRed2-U6-Abl1-sgRNA1 + U6-Abl1-sgRNA2] lines were scored for double fluorescence and used in subsequent experiments.
Genomic DNA was extracted from injected embryos using an Ezup Column Animal Genomic DNA Purification Kit (Sangon, Shanghai, China). Mutagenesis in targeted sites was detected by PCR. The primers that were used in PCR are listed in the Supplementary Material (Table S3).

RNA-Sequencing Analysis
There are three biological samples for RNA-Seq. For each sample, three pairs of WSGs were used from the Nistari or BmAbl1at the third-day fifth-instar larvae. Total RNA was extracted from the whole posterior silk gland dissected from third-day fifthinstar larvae. cDNA libraries were then constructed and sequenced using an Illumina HiSeq2500 platform (Illumina, San Diego, CA, USA). The quality of the raw data was assessed using the FastQC and FastX-Toolkit. Adaptor sequences, unknown sequences (N), and low-quality reads were removed from the raw data. The reference genome sequence of B. mori was obtained from SilkDB 3.0 [22]. The clean reads were mapped to the silkworm reference genome using the HISAT2 [23]. The raw counts of each sample were generated by HTSeq [24]. Differential expressed genes were analyzed using the DESeq2 package. p values were adjusted for multiple comparisons using the Benjamini-Hochberg method. A gene with adjusted p ≤ 0.001 and |log2 Fold Change (Log 2 FC)| ≥ 1 was considered a DEG. GO enrichment analysis was performed by ClusterProfile [25], and KEGG enrichment analysis was performed by KOBAS 3.0 [26]. A term with p ≤ 0.05 was considered an enrichment term.

Phenotypic Measurement of Transgenic Strain
Phenotyped Nistari, BmAbl -, and silkworms were fed with ABL inhibitor. Larval weight was measured from first-day fifth-instar to the wandering stage in groups of 30 individuals with three biological replicates. Thirty individuals were weighed together, and then each individual's weight was averaged. The whole cocoon weight and pupal weight were measured in the same way as above.

Measurement Content of Glutathione (GSH)
A Trace Reducing Glutathione Assay Kit (Jian Cheng, Nanjing, China) was used to measure glutathione (GSH) of the PSG by 5,5 -Dithiobis-(2-nitrobenzoic acid) (DTNB). The samples were extracted from the PSG of the Nistari and BmAbl1strains at the third-day fifth-instar. Tissues (100 mg) were weighed and added with 9-times PBS (pH = 7.4) for ultrasonic crushing. Centrifugation was conducted at 2500 rpm/min for 10 min. A volume of 0.1 mL of supernatant was taken, and 0.1 mL of precipitant was added. It was centrifuged at 3500 rpm/min for 10 min, and the supernatant was tested at 405 nm using SpectraMax Paradigm (Molecular Devices, San Jose, CA, USA). Every repeat had five individuals' PSG. Each set of data was repeated three times.

Measurement Content of Free Amino Acids
The samples for measuring the content of free amino acids were obtained from the PSG of the Nistari and BmAbl1strains at the fifth-day fifth-instar. After the samples were crushed, 1 g of tissue was extracted by adding 50 mL of 0.02 mol/L hydrochloric acid for 30 min. After shaking and filtering, 2 mL of filtrate was added to a centrifuge tube with 2 mL 8% sulfosalicylic acid and left for 15 min. After centrifugation at 10,000× g rpm for 10 min, the supernatant was filtered into a 0.45 µm membrane. After sample pretreatment, we performed quantitative analysis using an amino acid analyzer (LA8080, Hitachi, Japan). The sample was pushed through a sulfonic acid cationic resin separation column, and the amino acid mixture was separated into individual amino acids using different PH buffers. The isolated individual amino acids react with ninhydrin reagents to produce purple or yellow compounds. The absorbance at 570 nm and 440 nm was measured for the quantitative analysis of various amino acids. The reaction temperature was 135 ± 5 • C and the injection volume was 20 µL. Every repeat had five individuals' PSG. Each set of data was repeated three times.

Western Blot
The fibroin heavy chain (Fib-H), fibroin light chain (Fib-L), and P25 protein antibodies (1:4000, ABclonal Technology, Wuhan, China) were used for detecting the expression of Fib-H, Fib-L, and P25 in the PSG of the Nistari and BmAbl1strains by Western blotting. In addition, all primary antibodies were purified from New Zealand White Rabbits. All primary antibodies we used in this article were sponsored by Professor Tan Anjiang [27], Science of Plant Physiology and Ecology, Chinese Academy of Sciences. Total proteins were extracted from the PSG of the Nistari and BmAbl1strains. For each sample, there were three pairs of PSGs from the Nistari or BmAbl1strains. Each 100 mg sample was digested by 1 mL lysis buffer (containing 7 M urea, 2 M thiourea, 40 mg CHAPS, and 200 µL TritonX-100), 10 µL phenylmethanesulfonylfluoride (PMSF), and 0.01 g DTT. The mixture was stewed on ice for 30 min. Protein concentration was determined by a Bradford Protein Assay Kit (Sangon, Shanghai, China). The homogenate was centrifuged at 4 • C and 12,000× g for 10 min. We used 12% SDS-PAGE to separate samples at 80 V for 0.5 h and 120 V for 1 h, and then transferred the proteins to PVDF membranes. The tubulin antibody (1:4000, Bioss ANTIBODIES, Beijing, China) was used to level the protein content. The target protein was combined with the corresponding primary antibody. The secondary antibody was HRP Goat Anti-Rabbit IgG (1:5000, ABclonal Technology, Wuhan, China). The tubulin protein was used as the reference.

Statistical Analysis
The statistical differences among the three biological duplicates were determined with ANOVA and Student's t-test by R. Each set of data was repeated three times. Asterisks were used to indicate significant differences (* p < 0.05; ** p < 0.01; and *** p < 0.001).

BmAbl1 Inhibited by Dasatinib Reduce Larva Weight and CSW
To investigate the function of BmAbl1, we inhibited its activity using dasatinib. Dasatinib is a dual-specificity tyrosine kinase inhibitor, and is used as the inhibitor of Abl1 in insects. To test whether dasatinib can be used on B. mori, the inhibition efficiency of BmAbl1 activity was measured using the ELISA. The results showed that BmAbl1 activity was significantly inhibited to 41% of the Control ( Figure 1A).
ies (1:4000, ABclonal Technology, Wu Han, China) were used for detecting the expression of Fib-H, Fib-L, and P25 in the PSG of the Nistari and BmAbl1 -strains by Western blotting. In addition, all primary antibodies were purified from New Zealand White Rabbits. All primary antibodies we used in this article were sponsored by Professor Tan Anjiang [27], Science of Plant Physiology and Ecology, Chinese Academy of Sciences. Total proteins were extracted from the PSG of the Nistari and BmAbl1 -strains. For each sample, there were three pairs of PSGs from the Nistari or BmAbl1 -strains. Each 100 mg sample was digested by 1 mL lysis buffer (containing 7 M urea, 2 M thiourea, 40 mg CHAPS, and 200 μL TritonX-100), 10 μL phenylmethanesulfonylfluoride (PMSF), and 0.01 g DTT. The mixture was stewed on ice for 30 min. Protein concentration was determined by a Bradford Protein Assay Kit (Sangon, Shang Hai, China). The homogenate was centrifuged at 4 °C and 12,000 g for 10 min. We used 12% SDS-PAGE to separate samples at 80 V for 0.5 h and 120 V for 1 h, and then transferred the proteins to PVDF membranes. The tubulin antibody (1:4000, Bioss ANTIBODIES, Bei Jing, China) was used to level the protein content. The target protein was combined with the corresponding primary antibody. The secondary antibody was HRP Goat Anti-Rabbit IgG (1:5000, ABclonal Technology, Wu Han, China). The tubulin protein was used as the reference.

Statistical Analysis
The statistical differences among the three biological duplicates were determined with ANOVA and Student's t-test by R. Each set of data was repeated three times. Asterisks were used to indicate significant differences (* p < 0.05; ** p < 0.01; and *** p < 0.001).

BmAbl1 Inhibited by Dasatinib Reduce Larva Weight and CSW
To investigate the function of BmAbl1, we inhibited its activity using dasatinib. Dasatinib is a dual-specificity tyrosine kinase inhibitor, and is used as the inhibitor of Abl1 in insects. To test whether dasatinib can be used on B. mori, the inhibition efficiency of BmAbl1 activity was measured using the ELISA. The results showed that BmAbl1 activity was significantly inhibited to 41% of the Control ( Figure 1A). inhibited BmAbl1 led to larva weight being lighter than Control group during the fifth instar; (C) inhibited BmAbl1 led to cocoon weight, pupa weight, and CSW being lighter than Control group during the fifth instar. Standard deviation was calculated by all biological replicates. * p < 0.01, ** p < 0.005, *** p < 0.001, t-test, n = 30 for Control and Treated, respectively.
We also investigated the difference in larva and cocoon weight between the Treated and Control group. The daily larval weight of the Treated and Control group from the fourthinstar molting stage to the wandering stage showed that the weight of the Treated group was lower than that of the Control at the same time ( Figure 1B). The whole cocoon weight Insects 2022, 13, 967 6 of 13 (WCW) of the Treated decreased by 5.21% and 8.28% in the female and male individuals than that of the Control group, respectively. The CSW of the Treated group in females and males was, on average, lighter by 18.75% and 18.64% of the Control, respectively. The cocoon layer ratio of the Treated group was reduced by 13.33% and 12.27% in the female and male individuals ( Figure 1C). In summary, these results showed that the inhibition of BmAbl1 can reduce larva weight and CSW.

BmAbl1 Knockout Reduces Larva Weight and CSW
To further confirm the effect of BmAbl1 deficiency on CSW, we constructed a binary transgenic CRISPR/Cas9 system to knockout BmAbl1 somatically. Two transgenic silkworm lines were established based on a piggyBac transposon-derived vector. One expresses Cas9 ubiquitously under the control of nanos (nos) promoter (nos-Cas9), and the other line expresses the sgRNA targeting BmAbl1 driven by the silkworm U6 promoter (Figure 2A). After hybridization of the sgRNA and Cas9 strains, 1216 bp-delegation were detected at target sites in offspring ( Figure 2B). Two-hundred and eight bases are missing at the mRNA level, including all ninety-seven bases of exon2, and one-hundred and eleven bases of exon3. Frameshift mutation and the early termination of amino acids have taken place at the protein level (File S1). To exclude the influence of the inserted element, deletion offspring were crossed with Nistari and self-cross. Finally, the homozygous individuals without fluorescence were screened based on PCR and sequencing. The BmAbl1strain showed smaller individuals of the fifth-instar third-day and cocoon size than the normal Nistari strain ( Figure 2C).
We also investigated the difference in larva and cocoon weight between the Treated and Control group. The daily larval weight of the Treated and Control group from the fourth-instar molting stage to the wandering stage showed that the weight of the Treated group was lower than that of the Control at the same time ( Figure 1B). The whole cocoon weight (WCW) of the Treated decreased by 5.21% and 8.28% in the female and male individuals than that of the Control group, respectively. The CSW of the Treated group in females and males was, on average, lighter by 18.75% and 18.64% of the Control, respectively. The cocoon layer ratio of the Treated group was reduced by 13.33% and 12.27% in the female and male individuals ( Figure 1C). In summary, these results showed that the inhibition of BmAbl1 can reduce larva weight and CSW.

BmAbl1 Knockout Reduces Larva Weight and CSW
To further confirm the effect of BmAbl1 deficiency on CSW, we constructed a binary transgenic CRISPR/Cas9 system to knockout BmAbl1 somatically. Two transgenic silkworm lines were established based on a piggyBac transposon-derived vector. One expresses Cas9 ubiquitously under the control of nanos (nos) promoter (nos-Cas9), and the other line expresses the sgRNA targeting BmAbl1 driven by the silkworm U6 promoter (Figure 2A). After hybridization of the sgRNA and Cas9 strains, 1216 bp-delegation were detected at target sites in offspring ( Figure 2B). Two-hundred and eight bases are missing at the mRNA level, including all ninety-seven bases of exon2, and one-hundred and eleven bases of exon3. Frameshift mutation and the early termination of amino acids have taken place at the protein level (File S1). To exclude the influence of the inserted element, deletion offspring were crossed with Nistari and self-cross. Finally, the homozygous individuals without fluorescence were screened based on PCR and sequencing. The BmAbl1strain showed smaller individuals of the fifth-instar third-day and cocoon size than the normal Nistari strain ( Figure 2C).  One is on the second exon; another is on the third exon; (B) sequencing results show that the fragment has been deleted between two target sites; WT indicates Nistari strain. Dual-1 and dual-2 are BmAbl1 defective strains based on CRISPR/Cas9 method. Nof-1 and nof-2 are BmAbl1 defective strains that removed fluorescence protein from the genome through cross-fertilizing with Nistari and self-fertilizing. (C) Developmental change of bombyx larvae and pupa between WT and BmAbl -. BmAbl1strain showed smaller individual (average 0.73 g) and cocoon size (average 0.96 g) than wide type (average 1.00 g, 1.15 g, respectively) (scale bar, 1 cm.).
We investigated the weight of larvae of BmAbl1and Nistari from the fifth-instar firstday to the sixth day ( Figure 3A). The weights of Nistari were generally higher than BmAbl1during the whole fifth larva stage. The whole cocoon weight (WCW) of male and female BmAbl1was, on average, lighter by 12.77% and 14.29% of Nistari. The CSW of male and female BmAbl1was, on average, lighter by 23.08% and 15.38% of Nistari, respectively ( Figure 3B).
We investigated the weight of larvae of BmAbl1 -and Nistari from the fifth-instar firstday to the sixth day ( Figure 3A). The weights of Nistari were generally higher than BmAbl1 -during the whole fifth larva stage. The whole cocoon weight (WCW) of male and female BmAbl1 -was, on average, lighter by 12.77% and 14.29% of Nistari. The CSW of male and female BmAbl1 -was, on average, lighter by 23.08% and 15.38% of Nistari, respectively ( Figure 3B).

Figure 3.
Knockout BmAbl1 led to a reduction of cocoon shell weight (CSW), but not at the transcription level. (A) Larva weight was decreased during the fifth-instar larva of BmAbl1 -strain compared to Nistari; (B) cocoon weight, pupa weight, and CSW were decreased in BmAbl1 -strain compared to Nistari, both male and female; (C) the expression levels of fibroin heavy-chain, fibroin light-chain, and p25 gene were not different between BmAbl1 -and Nistari; (D) the expression level of sericin 1 was not different between BmAbl1 -and Nistari. Standard deviation was calculated by all biological replicates. * p < 0.01, ** p < 0.005, *** p < 0.001, t-test, n = 30 for Control and n = 30 for Treated; each group had three biological replicates.
In order to explore the reason for CSW reduction on a molecular level, we measured the expression of silk fibroin genes in third-day fifth larva. However, we found that the expression of Fib-H, p25, and sericin-1 was not significantly changed between BmAbl1 -and Nistari ( Figure 3C,D). Only the expression of Fib-L gene had a little change ( Figure 3C, FoldChange < 0.5).

Glutathione Metabolism and Amino Acid Biosynthesis were Affected by Knocking Out BmAbl1 -
To further uncover the influence of knocking out BmAbl1, the transcriptome data of third-day fifth-larval PSG was analyzed and compared between BmAbl1 -and Nistari. The RNA-seq raw data were deposited to NCBI SRA with the accession numbers, SRR21798269 to SRR21798274 (https://www.ncbi.nlm.nih.gov/sra, accessed on 4 October 2022). After quality control, clean reads were mapped to the silkworm genome, which was Figure 3. Knockout BmAbl1 led to a reduction of cocoon shell weight (CSW), but not at the transcription level. (A) Larva weight was decreased during the fifth-instar larva of BmAbl1strain compared to Nistari; (B) cocoon weight, pupa weight, and CSW were decreased in BmAbl1strain compared to Nistari, both male and female; (C) the expression levels of fibroin heavy-chain, fibroin light-chain, and p25 gene were not different between BmAbl1and Nistari; (D) the expression level of sericin 1 was not different between BmAbl1and Nistari. Standard deviation was calculated by all biological replicates. * p < 0.01, ** p < 0.005, *** p < 0.001, t-test, n = 30 for Control and n = 30 for Treated; each group had three biological replicates.
In order to explore the reason for CSW reduction on a molecular level, we measured the expression of silk fibroin genes in third-day fifth larva. However, we found that the expression of Fib-H, p25, and sericin-1 was not significantly changed between BmAbl1and Nistari ( Figure 3C,D). Only the expression of Fib-L gene had a little change ( Figure 3C, FoldChange < 0.5).

Glutathione Metabolism and Amino Acid Biosynthesis Were Affected by Knocking Out BmAbl1 -
To further uncover the influence of knocking out BmAbl1, the transcriptome data of third-day fifth-larval PSG was analyzed and compared between BmAbl1and Nistari. The RNA-seq raw data were deposited to NCBI SRA with the accession numbers, SRR21798269 to SRR21798274 (https://www.ncbi.nlm.nih.gov/sra, accessed on 4 October 2022). After quality control, clean reads were mapped to the silkworm genome, which was obtained from SilkDB 3.0. In total, 657 differentially expressed genes (DEGs) were identified between BmAbl1and Nistari using the DESeq2 package (Table S4). Compared to Nistari, 264 genes were up-regulated and 393 genes were down-regulated in BmAbl1 -( Figure 4A).
We employed the related functions of DEGs using enrichment analysis (Table S5). GO enrichment analysis showed that glutathione and tyrosine metabolic processes were significantly affected in BmAbl1 -( Figure 4B). The molecular functions of DEGs were primarily involved in glycosyltransferase activity, cation binding, phospholipid binding, glutathione transferase activity, and pentosyltransferase activity.
KEGG enrichment analysis indicated that amino acid biosynthesis and metabolism pathways had a significant difference between BmAbl1and Nistari, especially glutathione metabolism; glycine, serine, and threonine metabolism; phenylalanine metabolism; and tyrosine metabolism ( Figure 4C, Table S6). The DEGs in tyrosine metabolism and phenylala- nine metabolism pathways were mainly up-regulated in BmAbl1 -, but were down-regulated in other pathways.
Insects 2022, 13, x FOR PEER REVIEW 8 of 13 obtained from SilkDB 3.0. In total, 657 differentially expressed genes (DEGs) were identified between BmAbl1 -and Nistari using the DESeq2 package (Table S4). Compared to Nistari, 264 genes were up-regulated and 393 genes were down-regulated in BmAbl1 - (Figure 4A). We employed the related functions of DEGs using enrichment analysis (Table S5). GO enrichment analysis showed that glutathione and tyrosine metabolic processes were significantly affected in BmAbl1 -( Figure 4B). The molecular functions of DEGs were primarily involved in glycosyltransferase activity, cation binding, phospholipid binding, glutathione transferase activity, and pentosyltransferase activity.
KEGG enrichment analysis indicated that amino acid biosynthesis and metabolism pathways had a significant difference between BmAbl1 -and Nistari, especially glutathione metabolism; glycine, serine, and threonine metabolism; phenylalanine metabolism; and tyrosine metabolism ( Figure 4C, Table S6). The DEGs in tyrosine metabolism and phenylalanine metabolism pathways were mainly up-regulated in BmAbl1 -, but were down-regulated in other pathways.
Both GO and KEGG enrichment analyses implicated that the functional defect of BmAbl1 inhibited the glutathione metabolism process. Glutathione, as an antioxidant, widely exists in animals, plants, and fungi. It can be resolved into glutamate, cysteine, and glycine. This process involves glutathione transferase (EC:2. 5  It is particularly noteworthy that glutathione transferase Delta2 (BMSK0003598), one of the subunits of glutathione transferase (EC:2.5.1.18) in B. mori, is highly expressed in the glutathione metabolism process of Nistari, but is extremely suppressed in the BmAbl1strain. This result implied that glutathione metabolism was slowed, and less Gly was produced in the silk gland. To detect whether it specifically expressed in the silk gland, we compared the expression level of glutathione transferase Delta2 in different organs between the Nistari and BmAbl1strain. We found that glutathione transferase Delta2 had the second-highest expression level in the posterior silk gland of Nistari. The deletion of BmAbl1 resulted in a significant decrease in glutathione transferase Delta2 expression in posterior silk glands, but no change or increase in other tissues ( Figure 5B). We also observed that the expression of pyrimidodiazepine synthase (BMSK0006405, EC:1.5.4.1) Insects 2022, 13, 967 9 of 13 and glucose-6-phosphate dehydrogenase (BMSK0001869, EC:1.1.1.49) was significantly inhibited in BmAbl1 -, and it will slow the conversion between reduced glutathione (GSH) and oxidized glutathione (GSSG).
Insects 2022, 13, x FOR PEER REVIEW 9 of 13 BMSK0003984. We observed that the expression of these genes was significantly downregulated in BmAbl1 -, except BMSK0003439. The qRT-PCR results demonstrated the RNAseq results ( Figure 5A). It is particularly noteworthy that glutathione transferase Delta2 (BMSK0003598), one of the subunits of glutathione transferase (EC:2.5.1.18) in B. mori, is highly expressed in the glutathione metabolism process of Nistari, but is extremely suppressed in the BmAbl1strain. This result implied that glutathione metabolism was slowed, and less Gly was produced in the silk gland. To detect whether it specifically expressed in the silk gland, we compared the expression level of glutathione transferase Delta2 in different organs between the Nistari and BmAbl1 -strain. We found that glutathione transferase Delta2 had the second-highest expression level in the posterior silk gland of Nistari. The deletion of BmAbl1 resulted in a significant decrease in glutathione transferase Delta2 expression in posterior silk glands, but no change or increase in other tissues ( Figure 5B). We also observed that the expression of pyrimidodiazepine synthase (BMSK0006405, EC:1.5.4.1) and glucose-6-phosphate dehydrogenase (BMSK0001869, EC:1.1.1.49) was significantly inhibited in BmAbl1 -, and it will slow the conversion between reduced glutathione (GSH) and oxidized glutathione (GSSG).
The reduction of Gly content might further cause the reduction of Ser and threonine (Thr) content. KEGG enrichment results also showed that the glycine, serine, and threonine metabolism pathways changed in BmAbl1 -. We verified the expression of DEGs in the glycine, serine, and threonine metabolism pathways using qRT-PCR. The change of threonine ammonia-lyase (BMSK0014693, EC:4.3.1.19), related to serine metabolism, had a very low expression level and could not be detected. This suggested that the changes in Ser content might not be caused by threonine ammonia-lyase (File S3).
According to pathway enrichment analysis, a series of enzymes affecting the amino acid synthesis and metabolism were also identified from DEGs. We performed qRT-PCR to verify the expression of these DEGs. The results were consistent with RNA-seq data. Compared to Nistari, the expression of fumarylacetoacetase in BmAbl1 -(BMSK0001189, The reduction of Gly content might further cause the reduction of Ser and threonine (Thr) content. KEGG enrichment results also showed that the glycine, serine, and threonine metabolism pathways changed in BmAbl1 -. We verified the expression of DEGs in the glycine, serine, and threonine metabolism pathways using qRT-PCR. The change of threonine ammonia-lyase (BMSK0014693, EC:4.3.1.19), related to serine metabolism, had a very low expression level and could not be detected. This suggested that the changes in Ser content might not be caused by threonine ammonia-lyase (File S3).

Abnormal Glutathione Metabolism Reduced Glycine and Serine Content in PSG of BmAbl1 -
RNA-seq and qRT-PCR results implicated that diminished glutathione metabolism might cause the accumulation of glutathione, and decrease free Gly and Ser content (Files S2 and S3). To confirm our hypothesis, we first compared the glutathione content in the PSG of fifth-day fifth-instar larvae between BmAbl1and Nistari. The ELISA experiment results showed that the glutathione content in BmAbl1 -(avg = 78.50 µmol/mg) significantly increased to 1.49 times more than Nistari (avg. = 52.55 µmol/mg) ( Figure 6A, p = 0.0002).
This result demonstrated that abnormal glutathione metabolism caused by knocking out BmAbl1led to glutathione accumulation.

Abnormal Glutathione Metabolism Reduced Glycine and Serine Content in PSG of BmAbl1 -
RNA-seq and qRT-PCR results implicated that diminished glutathione metabolism might cause the accumulation of glutathione, and decrease free Gly and Ser content (Files S2 and S3). To confirm our hypothesis, we first compared the glutathione content in the PSG of fifth-day fifth-instar larvae between BmAbl1 -and Nistari. The ELISA experiment results showed that the glutathione content in BmAbl1 -(avg = 78.50 μmol/mg) significantly increased to 1.49 times more than Nistari (avg. = 52.55 μmol/mg) ( Figure 6A, p = 0.0002). This result demonstrated that abnormal glutathione metabolism caused by knocking out BmAbl1 -led to glutathione accumulation. Fibroin is composed of Fib-H, Fib-L, and p25, with a 6:6:1 molar ratio. The amino acid composition of Fib-H is Gly (46%), Ala (30%), Ser (12%), Tyr (5.3%), and Val (1.8%) [4].
The key enzymes (such as EC:2.5.1.18, EC:3.4.19.13, and EC:2.3.2.2) in the glutathione decomposition pathway were significantly down-regulated in BmAbl -. Abnormalities of the glutathione metabolism pathway may induce the decrease of free Gly and Ser (File S2 and S3). To further confirm our hypothesis, we next measured free amino acid contents in the WSG of fifth-day fifth larvae. The results showed that Gly and Ser had a significant downregulation in the BmAbl1 -strain, but Ala had no difference ( Figure 6B).

Discussion
The key enzymes (such as EC:2.5.1.18, EC:3.4.19.13, and EC:2.3.2.2) in the glutathione decomposition pathway were significantly down-regulated in BmAbl -. Abnormalities of the glutathione metabolism pathway may induce the decrease of free Gly and Ser (Files S2 and S3). To further confirm our hypothesis, we next measured free amino acid contents in the WSG of fifth-day fifth larvae. The results showed that Gly and Ser had a significant down-regulation in the BmAbl1strain, but Ala had no difference ( Figure 6B).
We further compared three major fibroin proteins content in the PSG of fifth-instar larvae between BmAbl1and Nistari. The western blot results showed that Fib-H, Fib-L, and P25 protein were all down-regulated compared to Nistari ( Figure 6C).

Discussion
Since the silkworm was domesticated as an economic insect 5000 years ago, the cocoon quantity is always an important trait in artificial screening. However, a series of factors can affect the cocoon quantity of a silk strain, such as the development of the silk gland [21], the number of silk gland cells [28], the capabilities of DNA replication [29], and protein synthesis [20,30]. On the molecular level, energy supply, the content of free amino acid, mRNA content of fibroin, the number of ribosomes, and the coordination of the regulatory system are all affecting factors of silk protein synthesis in the silk gland of fifth-instar larvae. Many efforts have been made to find out the major genes which are strongly related to silk yield [5][6][7][8]. Chromosome 1 is widely confirmed to contribute to cocoon shell weight in previous work. BmAbl1 is one of the potential targets in candidate regions. A recent report suggested that BmAbl1 had a strong correlation with cocoon shell weight [9,10]. However, how BmAbl1 affects the synthesis of silk protein is still unknown.
In the current study, our results suggested that the loss-of-function of BmAbl1 had little effect on the expression of fibroin and sericin1 genes, but the synthesis of fibroin protein was reduced significantly. RNA-Seq data suggested that the glutathione metabolism pathway was suppressed through the down-regulation of glutathione-S-transferase (GST) and gamma-glutamyltransferase (GGT) in BmAbl1 -. The free amino acid measurement results showed that the inhibition of glutathione metabolism led to the reduction of free Gly and Ser, which are the major components of fibroin protein.
Some questions should be answered to address how defective BmAbl1 down-regulated the expression of GST and GGT (Figure 7). The first question is which transcription factor (TF) regulates by BmAbl1. A potential TF is BMSK0005992, a homologous of cnc in D. melanogaster and Nrf2 in mice. It has been demonstrated that Nrf2 is an essential activator in the expression of GST family genes in mice [31,32]. ChIP-sequencing data also supported that GstD2 and GstE6 were the targets of CNC in D. melanogaster [33]. Misra et al. have demonstrated that CNC causes changes in the expression of enzymes, including 36 different P450, 17 GSTs [34]. This is consistent with our observation in RNA-seq data. The next question is how Abl1 activates the transcription factor activity of CNC. It has been known that c-Abl phosphorylated protein kinase C (PKC) can trigger Nrf2 nuclear translocation [35,36]. A previous study suggested that c-Abl is activated by PKC and c-Abl phosphorylates, and thereby further activates PKC [35]. The phosphorylation of Nrf2 by PKC promoted its dissociation from Keap1, and translocated into the nucleus to activate ARE-mediated gene expression [32,36]. However, this potential pathway needs to be further verified in B. mori.
weight in previous work. BmAbl1 is one of the potential targets in candidate regions. A recent report suggested that BmAbl1 had a strong correlation with cocoon shell weight [9,10]. However, how BmAbl1 affects the synthesis of silk protein is still unknown.
In the current study, our results suggested that the loss-of-function of BmAbl1 had little effect on the expression of fibroin and sericin1 genes, but the synthesis of fibroin protein was reduced significantly. RNA-Seq data suggested that the glutathione metabolism pathway was suppressed through the down-regulation of glutathione-S-transferase (GST) and gamma-glutamyltransferase (GGT) in BmAbl1 -. The free amino acid measurement results showed that the inhibition of glutathione metabolism led to the reduction of free Gly and Ser, which are the major components of fibroin protein.
Some questions should be answered to address how defective BmAbl1 down-regulated the expression of GST and GGT (Figure 7). The first question is which transcription factor (TF) regulates by BmAbl1. A potential TF is BMSK0005992, a homologous of cnc in D. melanogaster and Nrf2 in mice. It has been demonstrated that Nrf2 is an essential activator in the expression of GST family genes in mice [31] , [32]. ChIP-sequencing data also supported that GstD2 and GstE6 were the targets of CNC in D. melanogaster [33]. Misra et al. have demonstrated that CNC causes changes in the expression of enzymes, including 36 different P450, 17 GSTs [34]. This is consistent with our observation in RNA-seq data. The next question is how Abl1 activates the transcription factor activity of CNC. It has been known that c-Abl phosphorylated protein kinase C (PKC) can trigger Nrf2 nuclear translocation [35,36]. A previous study suggested that c-Abl is activated by PKC and c-Abl phosphorylates, and thereby further activates PKC [35]. The phosphorylation of Nrf2 by PKC promoted its dissociation from Keap1, and translocated into the nucleus to activate ARE-mediated gene expression [32,36]. However, this potential pathway needs to be further verified in B. mori. Figure 7. Potential pathway of GST activated by Abl1 to degrade GSH into Gly. "?" indicates the inferred pathway from the literature. Figure 7. Potential pathway of GST activated by Abl1 to degrade GSH into Gly. "?" indicates the inferred pathway from the literature.

Conclusions
In this work, we confirmed the contribution in the synthesis of silk protein using chemical inhibitor and knockout BmAbl1 methods. The functional defect in BmAbl1 can reduce the weight of the larva and the cocoon. Our results suggested that the loss of function of BmAbl1 impaired the synthesis of fibroin protein via glutathione metabolism. The inhibition of glutathione metabolism reduced the content of free Gly and Ser in the BmAbl1strain. As the major component of fibroin protein, the reduction of free Gly and Ser content further affected silk protein production.

Data Availability Statement:
The data presented in this study are available in supplementary material. The RNA-seq raw data were deposited to NCBI SRA with the accession numbers, SRR21798269 to SRR21798274 (https://www.ncbi.nlm.nih.gov/sra, accessed on 4 October 2022).

Conflicts of Interest:
The authors declare no conflict of interest.