Expression of a Chromoplast-Specific Lycopene β-Cyclase Gene (CYC-B) Is Implicated in Carotenoid Accumulation and Coloration in the Loquat

Carotenoids are the principal pigments in the loquat. Although the metabolic pathway of plant carotenoids has been extensively investigated, few studies have been explored the regulatory mechanisms of loquat carotenoids because knowledge of the loquat genome is incomplete. The chromoplast-specific lycopene β-cyclase gene (CYC-B) could catalyze cyclization of lycopene to β-carotene. In this study, the differential accumulation patterns of loquat with different colors were analyzed and virus-induced gene silencing (VIGS) was utilized in order to verify CYC-B gene function. Using a cloning strategy of homologous genes, a CYC-B gene orthologue was successfully identified from the loquat. At a later stage of maturation, CYC-B gene expression and carotenoids concentrations in the ‘Dawuxing’ variety were higher than in ‘Chuannong 1-5-9′, possibly leading to the difference in pulp coloration of loquat. Interference of CYC-B gene expression in the loquat demonstrated clear visual changes. The green color in negative control fruits became yellow, while TRV2-CYC-B silenced fruits remained green. CYC-B gene expression and total carotenoid content in the pulp decreased by 32.5% and 44.1%, respectively. Furthermore, multiple key genes in the carotenoid metabolic pathway synergistically responded to downregulation of CYC-B gene expression. In summary, we provide direct evidences that CYC-B gene is involved in carotenoid accumulation and coloration in the loquat.


Introduction
Loquat (Eriobotrya japonica Lindl) is a member of the Rosaceae family and is native to the southeast of China [1]. Carotenoids are the principal pigments in loquat and their concentration has been shown to be significantly correlated with fruit coloration [2][3][4]. At present, few studies of the regulatory mechanism of loquat carotenoids have been published due to the lack of information about its genome. With the development of modern molecular biological techniques, an investigation of the genetic mechanisms that regulate the accumulation of carotenoids in the loquat is timely.
Carotenoids are a major class of natural pigments of which more than 800 have been discovered [5,6]. Plant carotenoids are C 40 isoprenoids with polyene chains that may contain up to 15 conjugated double bonds. They are important components of photosynthetic organisms due to their chemical properties [7].
Biomolecules 2019, 9, x FOR PEER REVIEW 3 of 14 possible to obtain direct evidences for its involvement in carotenoid accumulation and coloration of loquat. And the use of the VIGS technique provides a new method for the research of loquat gene function.

Identification of Loquat CYC-B Gene Full-Length cDNA and Phylogenetic Analysis
Loquat fruits were selected to extract total RNA, and first-strand cDNA was synthesized using a PrimeScript™ RT reagent Kit with gDNA Eraser (TaKaRa, Daliang, China), used as the template. A homology-based cloning method was selected because genome information of the loquat is incomplete. Degenerate primers containing the complete coding region (cCYC-B_F: 5′-CCACTTYCCTTTGACCTTCA-3′ and cCYC-B_R: 5′-GTGTTRCATTATTATGCAGCT-3′) were designed by this laboratory. A high-fidelity enzyme was used to amplify the CYC-B gene from 'Dawuxing', with homology alignment verified after sequencing. A phylogenetic analysis of CYC-B proteins from different plants was constructed using MEGA software (v5.2, https://www.megasoftware.net/) with neighbor-joining method (testing of 1000 bootstrap replications).

Plasmid Construction
Construct maps of the TRV1, TRV2, and TRV2-CYC-B plasmids were shown in Figure 2. Primers with EcoR I and BamH I restriction enzyme cutting sites (underlined) were designed in the conserved region of CYC-B g e n e s e q u e n c e ( pCYC-B_F: 5'-CGGGATCCATGGAGTTTGGGTTGATGAA-3' and pCYC-B_R: 5'-CGGAATTCGCCTACTAACAAGCGAAGTT-3'). The fragment of CYC-B gene from 'Zaozhong 6′ was inserted in an antisense orientation before the nopaline synthase (NOS) terminator and after the cauliflower mosaic virus (CaMV) 35S promoter. CYC-B fragment was purified using a SanPrep Column DNA gel extraction kit (Sangon, Shanghai, China). The purified product and TRV2 plasmid were both digested by EcoR I and BamH I restriction enzymes. After ligating with T4 ligase

Identification of Loquat CYC-B Gene Full-Length cDNA and Phylogenetic Analysis
Loquat fruits were selected to extract total RNA, and first-strand cDNA was synthesized using a PrimeScript™ RT reagent Kit with gDNA Eraser (TaKaRa, Daliang, China), used as the template. A homology-based cloning method was selected because genome information of the loquat is incomplete. Degenerate primers containing the complete coding region (cCYC-B_F: 5 -CCACTTYCCTTTGACCTTCA-3 and cCYC-B_R: 5 -GTGTTRCATTATTATGCAGCT-3 ) were designed by this laboratory. A high-fidelity enzyme was used to amplify the CYC-B gene from 'Dawuxing', with homology alignment verified after sequencing. A phylogenetic analysis of CYC-B proteins from different plants was constructed using MEGA software (v5.2, https: //www.megasoftware.net/) with neighbor-joining method (testing of 1000 bootstrap replications).

Plasmid Construction
Construct maps of the TRV1, TRV2, and TRV2-CYC-B plasmids were shown in Figure 2. Primers with EcoR I and BamH I restriction enzyme cutting sites (underlined) were designed in the conserved region of CYC-B gene sequence (pCYC-B_F: 5 -CGGGATCCATGGAGTTTGGGTTGATGAA-3 and pCYC-B_R: 5 -CGGAATTCGCCTACTAACAAGCGAAGTT-3 ). The fragment of CYC-B gene from 'Zaozhong 6 was inserted in an antisense orientation before the nopaline synthase (NOS) terminator and after the cauliflower mosaic virus (CaMV) 35S promoter. CYC-B fragment was purified using a SanPrep Column DNA gel extraction kit (Sangon, Shanghai, China). The purified product and TRV2 plasmid were both digested by EcoR I and BamH I restriction enzymes. After ligating with T4 ligase overnight, it was transferred into DH5α Escherichia coli (E. coli) competent cells. E. coli was spread using selective media containing kanamycin (Kan, 50 mg/L). After expansion, the quality of E. coli was ascertained by running PCR and agarose gel electrophoresis. A pair of detection primer for TRV2-CYC-B (S1_F: 5 -CGGACGAGTGGACTTAGATTCTG-3 and S1_R: 5 -GCCTATGGCTTCTGTTCATGTG-3 ) were synthesized. TRV2-CYC-B plasmid DNA was extracted using a SanPrep Column Plasmid Mini-Preps kit (Sangon, Shanghai, China), as identified by EcoR I and BamH I digestion then verified by sequencing.
overnight, it was transferred into DH5α Escherichia coli (E.coli) competent cells. E. coli was spread using selective media containing kanamycin (Kan, 50 mg/L). After expansion, the quality of E. coli was ascertained by running PCR and agarose gel electrophoresis. A pair of detection primer for TRV2-CYC-B (S1_F: 5'-CGGACGAGTGGACTTAGATTCTG-3' and S1_R: 5'-GCCTATGGCTTCTGTTCATGTG-3') were synthesized. TRV2-CYC-B plasmid DNA was extracted using a SanPrep Column Plasmid Mini-Preps kit (Sangon, Shanghai, China), as identified by EcoR I and BamH I digestion then verified by sequencing.

Agrobacterium Transformation and Injection
TRV2-CYC-B and TRV1 and TRV2 plasmids were transfected into Agrobacterium tumefaciens GV3101 competent cells by freeze-thaw. The method of Chen et al. [32] and Zhai et al. [31] were used for VIGS. Agrobacterium cells were spread in selective Luria Bertani (LB) solid medium containing Gentamicin (Gen, 40 mg/L), Kan (50 mg/L), and Rifampicin (Rif, 20 mg/L), then cultivated at 28 °C for 48 h-72 h. The growth of the colonies was observed. Recombinant, TRV1, and TRV2 colonies were inoculated with 50 mL of LB liquid medium containing 40 mg/L Gen, 50 mg/L Kan, 20 mg/L Rif, 10 mmol/L 2-morpholinoethanesulfonic acid (MES), and 20 μmol/L acetosyringone (AS), shaken at 28 °C until the OD 600 had risen to 1.2. The presence of the inserts was verified using 1 μL of bacterial solution as a template for PCR. TRV1 detection primers were created as described by Chen et al. [32]. TRV2-CYC-B detection primers were the same as S1 primers. TRV2 detection primers were as follows: TRV2_F: 5'-GCCATTAGCGACATCTAAAT-3' and TRV2_R: 5'-CTAAGTCCACTCGTCCGTAA-3'. The verified cells were collected by centrifugation at 4000 g for 15 min and resuspended in infection buffer (containing MES 10 mmol/L, MgCl2 10 mmol/L, AS 200 μmol/L) twice until the final OD-600 was 1.2. Finally, resuspended cells were gently shaken at 28 °C for 4 h.
Disease-free fruits of uniform size and maturity at the breaker stage were selected as injection materials. In this study, recombinant TRV2-CYC-B and TRV2 were mixed with TRV1 bacterial solution at a ratio of 1:1, respectively, using the empty TRV2 plasmid as a negative control. In addition, groups in which sterile water was injected, or not, were used. Twelve fruits of 'Zaozhong 6′ in each group were selected, using a single fruit as a repeat. A loquat fruit was injected with 300 μL of mixed bacterial solution using sterile syringes. After a week, the pulp of the injection area (no necrosis) was cut into small pieces and immediately frozen in liquid nitrogen, then stored at −80 °C. The primers for viral molecular detection (the TRV1 and TRV2 primers described above) were used

Agrobacterium Transformation and Injection
TRV2-CYC-B and TRV1 and TRV2 plasmids were transfected into Agrobacterium tumefaciens GV3101 competent cells by freeze-thaw. The method of Chen et al. [32] and Zhai et al. [31] were used for VIGS. Agrobacterium cells were spread in selective Luria Bertani (LB) solid medium containing Gentamicin (Gen, 40 mg/L), Kan (50 mg/L), and Rifampicin (Rif, 20 mg/L), then cultivated at 28 • C for 48 h-72 h. The growth of the colonies was observed. Recombinant, TRV1, and TRV2 colonies were inoculated with 50 mL of LB liquid medium containing 40 mg/L Gen, 50 mg/L Kan, 20 mg/L Rif, 10 mmol/L 2-morpholinoethanesulfonic acid (MES), and 20 µmol/L acetosyringone (AS), shaken at 28 • C until the OD 600 had risen to 1.2. The presence of the inserts was verified using 1 µL of bacterial solution as a template for PCR. TRV1 detection primers were created as described by Chen et al. [32]. TRV2-CYC-B detection primers were the same as S1 primers. TRV2 detection primers were as follows: TRV2_F: 5 -GCCATTAGCGACATCTAAAT-3 and TRV2_R: 5 -CTAAGTCCACTCGTCCGTAA-3 . The verified cells were collected by centrifugation at 4000 g for 15 min and resuspended in infection buffer (containing MES 10 mmol/L, MgCl 2 10 mmol/L, AS 200 µmol/L) twice until the final OD-600 was 1.2. Finally, resuspended cells were gently shaken at 28 • C for 4 h.
Disease-free fruits of uniform size and maturity at the breaker stage were selected as injection materials. In this study, recombinant TRV2-CYC-B and TRV2 were mixed with TRV1 bacterial solution at a ratio of 1:1, respectively, using the empty TRV2 plasmid as a negative control. In addition, groups in which sterile water was injected, or not, were used. Twelve fruits of 'Zaozhong 6 in each group were selected, using a single fruit as a repeat. A loquat fruit was injected with 300 µL of mixed bacterial solution using sterile syringes. After a week, the pulp of the injection area (no necrosis) was cut into small pieces and immediately frozen in liquid nitrogen, then stored at −80 • C. The primers for viral molecular detection (the TRV1 and TRV2 primers described above) were used to amplify samples, in order to verify whether the TRV virus had successfully invaded the loquat fruits.

Quantification of Carotenoid Content
Twelve fruits of two groups (negative control or TRV2-CYCB silenced) were selected to determine total carotenoid content, respectively. Total carotenoid content in loquat samples was conducted using acetone extraction. After extraction using 80% acetone for 24 h, the OD at 470 nm of the supernatant was measured, with 80% acetone used as the control. Every fruit sample was performed in three replicates.

Transcription Level Analysis
Real-time quantitative PCR (qRT-PCR) was performed to quantify gene expression levels in fruits of different developmental stages and the VIGS samples (negative control or TRV2-CYCB silenced). Appropriate gene-specific primers of carotenoid biosynthesis were designed by Fu et al. [4]. A reaction mixture of 20 µL was performed using a CFX96 Real-Time PCR Detection System (Bio-Rad, USA), which included 10 µL SYBR Green I mix (2×), 1 µL gene-specific forward and reverse primers (10 µmol/L), 1.5 µL cDNA template (5 ng/µL), and 6.5 µL sterile water. The cycling procedure was as follows: 95 • C for 30 s, 95 • C for 5 s, 58 • C for 30 s, for 40 cycles. Analysis of the melt curve was performed in accordance with the instructions of the SYBR ® Premix Ex Taq™ II (TaKaRa, Daliang, China) kit. The 2 −∆∆ct values were computed to quantify relative gene expression levels [33]. Every qRT-PCR was performed in three replicates.

Statistical Analysis
In this study, IBM SPSS statistical software (v20, https://www.ibm.com/legal/copytrade.shtml) was used to perform Duncan's pairwise comparison with a significance level of 5% (One-way Analysis of Variance, ANOVA, p < 0.05) and Pearson bivariate correlation analysis.

Isolation of Loquat CYC-B Gene
In this study, the 1534 bp-length sequence of CYC-B gene was acquired containing full open reading frame and encoding a protein of 496 amino acids. The CYC-B protein sequence was used to perform a homology blast then conduct additional phylogenetic analysis ( Figure 3). The loquat CYC-B protein had a phylogenetic relationship similar to other Rosaceae plants, of which the closest was Pyrus x bretschneideri and then was Malus domestica. The most distant genetic relationship was Crocus sativus. In addition, CYC-B and CCS proteins from Citrus were highly homologous. These results indicated that the isolated gene from the loquat belonged to the chromoplast-specific lycopene β-cyclase (CYC-B) gene. Full-length cDNA of the CYC-B gene was successfully isolated from the loquat, nominated as EjCYC-B.

Differential Accumulation Patterns of Yellow and White-Fleshed Loquats
The relative expression levels of the CYC-B gene demonstrated a rising trend before the YM stage, which was downregulated at the FM stage except for the pulp of 'Dawuxing' (Figure 4). Expression levels of the CYC-B gene were significantly different between loquats with different colors both in the pulp and peel after the B stage (p < 0.05). In the pulp, CYC-B gene expression of 'Dawuxing' and 'Chuannong 1-5-9′ reached their maximum at the FM and YM stages, respectively.

Differential Accumulation Patterns of Yellow and White-Fleshed Loquats
The relative expression levels of the CYC-B gene demonstrated a rising trend before the YM stage, which was downregulated at the FM stage except for the pulp of 'Dawuxing' (Figure 4). Expression levels of the CYC-B gene were significantly different between loquats with different colors both in the pulp and peel after the B stage (p < 0.05). In the pulp, CYC-B gene expression of 'Dawuxing' and 'Chuannong 1-5-9 reached their maximum at the FM and YM stages, respectively. In the peel, CYC-B gene expression both reached its maximum at the YM stage. At the FM stage, the CYC-B gene expression level of 'Dawuxing' was approximately 3-fold higher than that of 'Chuannong 1-5-9 in the pulp, and approximately twice that of 'Chuannong 1-5-9 in the peel.
In the peel, CYC-B gene expression both reached its maximum at the YM stage. At the FM stage, the CYC-B gene expression level of 'Dawuxing' was approximately 3-fold higher than that of 'Chuannong 1-5-9′ in the pulp, and approximately twice that of 'Chuannong 1-5-9′ in the peel.
The total carotenoid content of loquats with different colors at different developmental stages was detected in our previous study (doi: 10.13417/j.gab.037.004407). The content of total carotenoids of 'Dawuxing' and 'Chuannong 1-5-9′ varied greatly at different maturity stages. Pearson correlation analysis found that there was a significant and positive correlation between carotenoid content and CYC-B gene expression levels in the pulp of 'Dawuxing' at different developmental stages (p < 0.05). This result indicated that the change in CYC-B gene expression level may affect differential accumulation of carotenoids and coloration in the loquat.

Construction and Identification of Silencing Vector
After PCR amplification of the conserved region, a 543-length sequence was obtained. Comparison of homology with the CCS gene sequence of the pear (Pyrus bretschneideri Rehd.) (GenBank accession number: XM_009379544.2) demonstrated that sequence similarity reached 98.34%, indicating that the CYC-B gene fragment from loquat was successfully cloned and linked next to the TRV2 plasmid.
After double digestion of the recombinant plasmid, two distinct target bands appeared ( Figure  5), and a target band corresponding to the length of the cloned fragment of the CYC-B gene obtained. The target gene region of the recombinant plasmid was sequenced, the result identical to that of the previous CYC-B gene cloned fragment. Therefore, one can conclude that the TRV2-CYC-B silencing vector had been successfully constructed.

TRV-Mediated CYC-B Gene Silencing Decreased Carotenoid Accumulation of the Loquat
After infection by Agrobacterium, obvious phenotypic changes of the loquat fruits were observed. It was found that groups that were not injected or had been injected with sterile water and the negative control changed as normal from green to yellow, while the TRV2-CYC-B silenced fruits  The total carotenoid content of loquats with different colors at different developmental stages was detected in our previous study (doi: 10.13417/j.gab.037.004407). The content of total carotenoids of 'Dawuxing' and 'Chuannong 1-5-9 varied greatly at different maturity stages. Pearson correlation analysis found that there was a significant and positive correlation between carotenoid content and CYC-B gene expression levels in the pulp of 'Dawuxing' at different developmental stages (p < 0.05). This result indicated that the change in CYC-B gene expression level may affect differential accumulation of carotenoids and coloration in the loquat.

Construction and Identification of Silencing Vector
After PCR amplification of the conserved region, a 543-length sequence was obtained. Comparison of homology with the CCS gene sequence of the pear (Pyrus bretschneideri Rehd.) (GenBank accession number: XM_009379544.2) demonstrated that sequence similarity reached 98.34%, indicating that the CYC-B gene fragment from loquat was successfully cloned and linked next to the TRV2 plasmid.
After double digestion of the recombinant plasmid, two distinct target bands appeared ( Figure 5), and a target band corresponding to the length of the cloned fragment of the CYC-B gene obtained. The target gene region of the recombinant plasmid was sequenced, the result identical to that of the previous CYC-B gene cloned fragment. Therefore, one can conclude that the TRV2-CYC-B silencing vector had been successfully constructed. The relative expression levels of the CYC-B gene demonstrated a rising trend before the YM stage, which was downregulated at the FM stage except for the pulp of 'Dawuxing' (Figure 4). Expression levels of the CYC-B gene were significantly different between loquats with different colors both in the pulp and peel after the B stage (p < 0.05). In the pulp, CYC-B gene expression of 'Dawuxing' and 'Chuannong 1-5-9′ reached their maximum at the FM and YM stages, respectively. In the peel, CYC-B gene expression both reached its maximum at the YM stage. At the FM stage, the CYC-B gene expression level of 'Dawuxing' was approximately 3-fold higher than that of 'Chuannong 1-5-9′ in the pulp, and approximately twice that of 'Chuannong 1-5-9′ in the peel.
The total carotenoid content of loquats with different colors at different developmental stages was detected in our previous study (doi: 10.13417/j.gab.037.004407). The content of total carotenoids of 'Dawuxing' and 'Chuannong 1-5-9′ varied greatly at different maturity stages. Pearson correlation analysis found that there was a significant and positive correlation between carotenoid content and CYC-B gene expression levels in the pulp of 'Dawuxing' at different developmental stages (p < 0.05). This result indicated that the change in CYC-B gene expression level may affect differential accumulation of carotenoids and coloration in the loquat.

Construction and Identification of Silencing Vector
After PCR amplification of the conserved region, a 543-length sequence was obtained. Comparison of homology with the CCS gene sequence of the pear (Pyrus bretschneideri Rehd.) (GenBank accession number: XM_009379544.2) demonstrated that sequence similarity reached 98.34%, indicating that the CYC-B gene fragment from loquat was successfully cloned and linked next to the TRV2 plasmid.
After double digestion of the recombinant plasmid, two distinct target bands appeared ( Figure  5), and a target band corresponding to the length of the cloned fragment of the CYC-B gene obtained. The target gene region of the recombinant plasmid was sequenced, the result identical to that of the previous CYC-B gene cloned fragment. Therefore, one can conclude that the TRV2-CYC-B silencing vector had been successfully constructed.

TRV-Mediated CYC-B Gene Silencing Decreased Carotenoid Accumulation of the Loquat
After infection by Agrobacterium, obvious phenotypic changes of the loquat fruits were observed. It was found that groups that were not injected or had been injected with sterile water and the negative control changed as normal from green to yellow, while the TRV2-CYC-B silenced fruits remained green ( Figure 6A). Agarose gel electrophoresis detection found that TRV virus molecules were detected in the TRV2-CYC-B silenced group and negative control ( Figure 6B). In addition, no TRV virus molecules were detected in groups not injected and those in which sterile water was injected. The results indicated that the TRV virus had successfully invaded the pulps of the loquat.
CYC-B gene expression in the pulp of TRV2-CYC-B silenced and negative control fruits was quantified by qRT-PCR. Compared with the negative control, CYC-B gene expression in the silenced fruits decreased significantly, to 64.8% of the negative control, silencing efficiency reaching 35.2% ( Figure 6C). Total carotenoid content in the pulp of silenced fruits was 55.9% of the negative control, reduced by 44.1% ( Figure 6D). Pearson correlation analysis found that expression levels of the CYC-B gene were significantly and positively correlated with total carotenoid content (p < 0.01). The results demonstrated that TRV-mediated CYC-B gene silencing resulted in a significant decrease in carotenoid accumulation in loquat pulp and interfered loquat coloration. The results also indicated that the VIGS system had been successfully constructed in the loquat. remained green ( Figure 6A). Agarose gel electrophoresis detection found that TRV virus molecules were detected in the TRV2-CYC-B silenced group and negative control ( Figure 6B). In addition, no TRV virus molecules were detected in groups not injected and those in which sterile water was injected. The results indicated that the TRV virus had successfully invaded the pulps of the loquat. CYC-B gene expression in the pulp of TRV2-CYC-B silenced and negative control fruits was quantified by qRT-PCR. Compared with the negative control, CYC-B gene expression in the silenced fruits decreased significantly, to 64.8% of the negative control, silencing efficiency reaching 35.2% ( Figure 6C). Total carotenoid content in the pulp of silenced fruits was 55.9% of the negative control, reduced by 44.1% ( Figure 6D). Pearson correlation analysis found that expression levels of the CYC-B gene were significantly and positively correlated with total carotenoid content (p < 0.01). The results demonstrated that TRV-mediated CYC-B gene silencing resulted in a significant decrease in carotenoid accumulation in loquat pulp and interfered loquat coloration. The results also indicated that the VIGS system had been successfully constructed in the loquat.  (1), injected with sterile water (2), negative control (3) and TRV2-CYC-B silenced (4), respectively. M represented DL2000 Marker (TaKaRa, Daliang, China). qRT-PCR (C) and total carotenoid content (D) analysis in the pulp of negative control and silenced fruits. Different letters indicate significant differences between negative control and silenced fruits (One-way ANOVA, p < 0.05).

Expression Analysis of Carotenoid Metabolism Related Genes after TRV-CYC-B Silencing
mRNA levels of 12 genes connected to carotenoid metabolism were analyzed in the pulp of TRV2-CYC-B silenced and negative control fruits (Figure 7). In linear carotenoid synthesis, there were no significant changes in the expression of PDS and ZDS genes, while the expression levels of PSY and CRTISO genes decreased significantly. In carotenoid cyclization branches, expression levels of LCYB, LCYE, ZEP, VDE, and CCD genes decreased significantly, while expression of the ECH gene increased significantly and that of the BCH gene did not change significantly. Additionally, qRT-PCR (C) and total carotenoid content (D) analysis in the pulp of negative control and silenced fruits. Different letters indicate significant differences between negative control and silenced fruits (One-way ANOVA, p < 0.05).

Expression Analysis of Carotenoid Metabolism Related Genes after TRV-CYC-B Silencing
mRNA levels of 12 genes connected to carotenoid metabolism were analyzed in the pulp of TRV2-CYC-B silenced and negative control fruits (Figure 7). In linear carotenoid synthesis, there were no significant changes in the expression of PDS and ZDS genes, while the expression levels of PSY and CRTISO genes decreased significantly. In carotenoid cyclization branches, expression levels of LCYB, LCYE, ZEP, VDE, and CCD genes decreased significantly, while expression of the ECH gene increased significantly and that of the BCH gene did not change significantly. Additionally, PAP (plastid lipid-associated protein) mRNA levels also decreased significantly. To summarize, the expression of multiple structural genes involved in carotenoid accumulation changed after silencing the CYC-B gene.
Biomolecules 2019, 9, x FOR PEER REVIEW 9 of 14 PAP (plastid lipid-associated protein) mRNA levels also decreased significantly. To summarize, the expression of multiple structural genes involved in carotenoid accumulation changed after silencing the CYC-B gene.

Discussion
Carotenoids are the principal pigments in loquat, regulated by multiple factors. Since the discovery of the chromoplast-specific lycopene β-cyclase (CYC-B) gene by Ronen et al. [22], studies of the CYC-B gene become more widespread, finding that over-expression of CYC-B promoted formation of β-carotene in the tomato. Therefore, it is clear that the CYC-B gene plays an important role in the regulation of the accumulation of carotenoids. In this study, a CYC-B gene orthologue was successfully identified from the loquat using a homology-based cloning method, a simple, fast, and efficient method that uses highly conserved sequences to design degenerate primers to clone target genes. This has been used for carrots [34], Phalaenopsis Hybrida [35], and tobacco [36], etc.
In this study, the transcription levels of the CYC-B gene exhibited significant upregulation at the majority of developmental stages of yellow and white-fleshed loquats, as previously reported by Hadjipieri et al. [37]. At the FM stage, the expression of CYC-B gene was downregulated except in the pulp of 'Dawuxing'. However, total carotenoid content increased at the FM stage in our previous study. We speculate that in addition to the CYC-B gene, other genes cooperate to regulate total Figure 7. qRT-PCR analysis of genes related to carotenoid metabolic pathways after silencing. Plastids are the sites for the biosynthesis and accumulation of plant carotenoids. PAP, plastid lipid-associated protein; PSY, phytoene synthase; PDS, phytoene desaturase; ZDS, ζ-carotene desaturase; ZISO, 15-cis-ζ-carotene isomerase; CRTISO, carotene isomerase; LCYE, lycopene ε-cyclase; LCYB, lycopene β-cyclase; CYC-B, chromoplast-specific lycopene β-cyclase; BCH, β-carotene hydroxylase; ECH, ε-carotene hydroxylase; ZEP, zeaxanthin epoxidase; VDE, violaxanthin de-epoxidase; NXY, neoxanthin synthase; CCD, carotenoid cleavage dioxygenases. Different letters indicate significant differences between negative control and silenced fruits (One-way ANOVA, p < 0.05).

Discussion
Carotenoids are the principal pigments in loquat, regulated by multiple factors. Since the discovery of the chromoplast-specific lycopene β-cyclase (CYC-B) gene by Ronen et al. [22], studies of the CYC-B gene become more widespread, finding that over-expression of CYC-B promoted formation of β-carotene in the tomato. Therefore, it is clear that the CYC-B gene plays an important role in the regulation of the accumulation of carotenoids. In this study, a CYC-B gene orthologue was successfully identified from the loquat using a homology-based cloning method, a simple, fast, and efficient method that uses highly conserved sequences to design degenerate primers to clone target genes. This has been used for carrots [34], Phalaenopsis Hybrida [35], and tobacco [36], etc.
In this study, the transcription levels of the CYC-B gene exhibited significant upregulation at the majority of developmental stages of yellow and white-fleshed loquats, as previously reported by Hadjipieri et al. [37]. At the FM stage, the expression of CYC-B gene was downregulated except in the pulp of 'Dawuxing'. However, total carotenoid content increased at the FM stage in our previous study.
We speculate that in addition to the CYC-B gene, other genes cooperate to regulate total carotenoid biosynthesis at the FM stage. Previous research has found that PSY, CYCB, and BCH genes have a synergistic effect on carotenoid accumulation [38]. In our previous study, the content of total carotenoids of 'Dawuxing' and 'Chuannong 1-5-9 varied greatly, first decreasing and then increasing. During loquat fruit maturation, the chloroplasts disintegrated and began to form chromoplasts, resulting in that chloroplast carotenoids decreased and chromoplast carotenoids increased. Fu et al. [4] have verified this through observation of the structure of loquat plastids. After the D stage, loquat fruits began to accumulate a large quantity of chromoplast carotenoids except for the pulp of 'Chuannong 1-5-9 . In the pulp, carotenoid content of 'Dawuxing' was approximately 6-fold higher than that of 'Chuannong 1-5-9 at the FM stage. This result indicates that the pulp of 'chuannong 1-5-9 could not accumulate carotenoids effectively because of the failure to develop normal chromoplasts, exhibiting a white color, causing the difference in loquat pulp coloration. Fu et al. [4] also found the mature flesh of 'Luoyangqing' cultivars appeared red-orange due to the accumulation of abundant carotenoids and far fewer carotenoids accumulated in the flesh of white-fleshed cultivars.
After injection with Agrobacterium, the in vitro loquat fruits changed color from the breaker stage, fruit phenotype demonstrating obvious changes. It was found that there was shrinkage in the infected region of the silenced fruits, which we speculate is caused by the pressure difference of the infected region caused by enlarged interstitial spaces after excessive bacterial liquid injection. Following CYC-B gene silencing, the infected part of the silenced fruits retained a green phenotype, suggesting a delay in fruit maturity and interfering loquat coloration. In this study, decreased expression levels of the CYC-B gene were found in the pulp, and as a result of VIGS, led to additional reduction in carotenoid content, indicating that the CYC-B gene could positively regulate carotenoid accumulation in the loquat. So, we further confirmed that the CYC-B gene was closely related to loquat coloration, as found in other studies [23,24]. In addition, we found CYC-B gene silencing efficiency was 35.2%, while total carotenoid content reduced by 44.1%, indicating that the CYC-B gene played a key role in the regulation of carotenoid synthesis. Furthermore, this laboratory constructed the TRV-PSY silencing vector and found the silencing efficiency was 55.6%, total carotenoid content reducing by 46.8%. Additional experiments were conducted to explore why a lower silencing efficiency of CYC-B gene also caused a large drop in carotenoid concentration. After interfering with CYC-B expression, apart from the PDS, ZDS, and BCH genes, eight genes in the carotenoid metabolism pathway synergistically responded to downregulation of CYC-B gene expression, and the expression levels of seven genes (PSY, CRTISO, LCYB, LCYE, ZEP, VDE, CCD) were downregulated, except for the ECH gene. We speculate that suppression of multiple key structural genes may lead to a large reduction in carotenoids. Although ECH gene expression levels were upregulated, metabolized substrate synthesis of ECH gene was downregulated, which would reduce the synthesis and accumulation of carotenoids from the source. Moreover, downregulation of PAP gene expression might affect the development of plasmid structure and disturb the accumulation of carotenoids. Leitner-Dagan et al. [39] also found that inhibition of CHRC (a cucumber PAP) expression by RNAi resulted in a 30% decrease in carotenoid content in tomato flowers.
In this study, we found that the downregulation of the CYC-B gene after silencing was not complete. These factors influencing this are likely to be a selection of silencing vector [40], temperature [41,42], length and location of inserted fragments, and the similarity with target gene sequences [43]. VIGS was first successfully applied to Nicotiana benthamiana, with a modified tobacco mosaic virus (TMV) vector used to silence the PDS gene of N. benthamiana [44]. A variety of VIGS vectors have been developed based on a TMV vector for different hosts. The virus vectors suitable for different host plants affect systematic silencing in plants, thus resulting in different silencing efficiency. Previous research has found that temperature plays a key role in a plant's silencing phenotype [41,42]. Studies have found that a lower temperature is conducive to maintenance of gene silencing for longer periods of time. In addition, a high similarity between inserted fragments and target gene sequences, with an inserted fragment of 200-1300 bp, located in the center of the cDNA sequence, and avoiding the silencing ends of the sequence also promote good silencing efficiency when using VIGS [43]. VIGS is an approach for the transient induction of RNA silencing in plants and has a subsequent symptom recovery stage. Thus, timely harvesting after infection is conducive to detection of high silencing efficiency. In this study, poor silencing efficiency may be due to inappropriate harvesting time.
The biosynthesis of plant carotenoids is a complex process, regulated not only by structural genes [9]. Moreover, cauliflower OR transgene regulates the accumulation of carotenoids by controlling differentiation and formation of chromoplasts [45]. Zhou et al. [46] discovered that the OR gene regulates carotenoid biosynthesis by controlling PSY gene expression, revealing a novel regulatory mechanism of carotenoids. Phytochrome interacting factors (PIFs) are a central factor in light-mediated reactions [47], which regulate PSY gene expression and carotenoid accumulation by binding to G-Box element on PSY gene promoter, and transducing light signals [48]. Environmental factors and chemical substances also affect carotenoid biosynthesis. While different factors have differential effects on carotenoid biosynthesis, they are possibly jointly involved in the carotenoid metabolic process, whose regulation mechanism requires additional elucidation. At present, genomic information of the loquat is incomplete. VIGS, a technique for rapid identification of gene function, is helpful for exploration of metabolic mechanisms of loquat.

Conclusions
TRV-mediated VIGS was used to cause effective silencing of the CYC-B gene, finding that the CYC-B gene positively regulated carotenoid accumulation and affects coloration in the loquat. Moreover, there was a multiple-gene synergistic response to the downregulation of CYC-B gene expression in the carotenoid metabolic pathway. These results directly indicate that the CYC-B gene plays a vital role in carotenoid accumulation and coloration in the loquat. Clarifying CYC-B gene function is of great significance for breeding new loquat varieties.