Fruit-specific expression of crtB, HpBHY, CrBKT and SlLCYB triggers hyper-production of carotenoids in tomato fruit

Background Tomato is a major source of dietary carotenoids that play an important role in human health. The biosynthesis of carotenoids involved in a number of catalytic steps, among which the phytoene synthase (crtB) is the first committed enzyme for carotenoid biosynthesis, and the lycopene β-cyclase (LCYB) catalyzes the synthesis of β-carotene, β-carotene hydroxylase (BHY) or β-carotene ketolase (BKT) plays a vital role in synthesizing xanthophylls. Although there are some studies about improving the carotenoids of tomato fruit, it is still challenged to engineer tomato landraces with good traits for hyper-production of carotenoids, and the effects of vital carotenogenic genes working together remain unclear. The aim of this study is to improve the production of carotenoids in a tomato landrace by fruit-specific expression of four carotenogenic genes. Results A plant expression cassette containing the tomato E8 promotor to drive the four genes ( crtB , HpBHY , CrBKT , and SlLCYB ) in a polycistronic structure was constructed and introduced into a tomato landrace “ Huang Song ”. Three independent lines were confirmed to be putative transformants, which showed similar phenotypes to wild-type (WT) control excepting for the fruit colors. The transgenic fruit exhibited deep red color due to the accumulation of novel ketocarotenoids including astaxanthin, canthaxanthin, and ketolutein and much more lycopene. The contents of total carotenoids in transgenic fruit were up to 3.1~6.6 mg/g dry weight (DW), 20~44-fold of that in WT fruit which only accumulated low levels of lutein, lycopene and β-carotene. Furthermore, various generations of the transgenic plants (T0, T1, and T2) exhibited similar growing status and carotenoid profiles, indicating that the transformants were genetically stable. Moreover, most of the endogenous carotenogenic genes were up-regulated in transgenic fruits at mature stage,


4
Overexpressing a specific pair of BKT and β-carotene hydroxylase (BHY) triggered the hyperaccumulation of non-native ketocarotenoids in the leave and fruit of a Beta-mutant tomato [19].
Previous studies commonly used a strong constitutive promoter to drive no more than two targeted genes. While the involvement of constitutive promoters commonly resulted in the interference of endogenous metabolic network [20,21]. This problem could be overcome by using a tissue specific promoter, e.g. the tomato E8 promoter [22,23]. The biosynthesis of most metabolites including carotenoids involves in a number of catalytic steps. Thus, to effectively improve the production of a metabolite, it is sometimes necessary to express more than two key enzymes. Multigene transformation is still a challenge of plant genetic engineering, which usually depends on a large construct or co-transformation of multiple constructs [24,25]. 2A peptide sequence of the foot-andmouth disease virus could govern self-processing of polyproteins, which was successfully used in multigene co-transformation [26].
In this study, a 2A-linked polycistronic expression cassette consisting of an endogenous E8 promotor to drive four rate-limiting carotenoid pathway genes encoding for phytoene synthase (PSY/crtB), βcarotene hydroxylase (BHY), β-carotene ketolase (BKT), and Lycopene β-cyclase (LCYB) was constructed and introduced into the tomato landrace "Huang Song". Via driving four genes by one fruit specific promoter, this study alters the carotenoid profiles of a tomato landrace in fruit with not only enhanced production of native carotenoids but also produced the high-value ketocarotenoids (canthaxanthin, astaxanthin, and ketolutein), leading to up to 44-fold increase of total carotenoids and the higher antioxidant activity.

Generation of transgenic tomato lines with increased carotenoids
To improve the production of carotenoids in "Huang Song" tomato fruit, an expression cassette containing the E8 promoter and four rate-limiting genes linked by three 2A sequences was constructed ( Figure 2A) and introduced into the landrace using the method described in materials and methods. Confirmed by PCR and RT-PCR analysis ( Figure 2B, 2C), three independent lines designated as BBBB1, 2, 3 were achieved as putative transformants (T 0 ). E8 promoter is a fruit specific promoter, which directs the gene expression starting from the fruit become ripening [27]. The expression pattern of the exogenous genes coordinated with the E8 expression pattern in that crtB, HpBHY and CrBKT expressed markedly in the ripening stages but not in leaf, flower, and mature fruit ( Figure 2C).
The transformants shared the same phenotype as their mother line ( Figure 2D) except for the color of the fruits ( Figure 2E). In contrast to the yellowish fruit of controls, the transgenic lines bear deep red fruit, indicating that different carotenoid profiles might exist between the fruits.

Analysis of the carotenoids in transgenic tomato fruit
Since the transgenic tomato fruit exhibited different color to controls, we detected the compositions and contents of carotenoids in the fruit using the Ultra-performance liquid chromatography (UPLC) technique.
In contrast to the WT that consisted of mainly lutein, β-carotene and lycopene in fruit, the transgenic tomato accumulated large amounts of lycopene together with some non-native ketocarotenoids (canthaxanthin, ketolutein, astaxanthin) ( Figure 3). Moreover, the contents of each carotenoid from WT and transgenic fruits showed significant difference (Table 1). WT fruit mainly accumulated lutein (56 μg/g DW), lycopene (41 μg/g DW) and β-carotene (38 μg/g DW) with a total carotenoid content of 150 μg/g DW. In contrast, the transgenic plants synthesized much higher amounts of lycopene (2561-6345 μg/g DW) and β-carotene (98-243 μg/g DW) in addition to significant amounts of non-native ketocarotenoids, leading to 21~44-fold increase of total carotenoids compared to the WT. Interestingly, the pool of lutein between WT and transgenic lines (lutein + ketolutein) seems to be constant, but unexpectedly, the transgenic plants only produced small amounts of ketocarotenoids, which might result from the poor conversion from lycopene to β-carotene, as well as canthaxanthin to astaxanthin based on the carotenoid profiles (Table 1).
To investigate genetic stability of the three transgenic lines, we investigated two more generations of the plants and found no difference in plant phenotypes among the samples. As to carotenoid profiles, similar data (Table 1) were found among the different generations though some variations occurred, possibly resulted from the different planting time. Thus, the transgenic lines could inherit stably with normal growth and stable carotenoids feature.

Expression of the carotenogenic genes
To correlate the dynamic changes of carotenoids with the expression of endogenous genes involved in carotenoid biosynthesis, we measured the contents of carotenoids and the transcriptional levels of carotenogenic genes in BBBB3 line at four stages of fruit development and ripening: mature green (MG), breaker (B), pink (P) and over red (OR) stages. Obviously, at MG stage, both transgenic plant and WT shared similar carotenoid profiles (Table 2). However, starting from B stage, the BBBB3 rapidly accumulated lycopene as well as other carotenoids including the ketocarotenoids, leading to a 40-fold increase of total carotenoids as compared with WT ( Table 2). In contrast to other carotenoids which showed steady increases during the fruit development, the α-branch carotenoids (lutein and ketolutein) in BBBB3 fruit seemed almost not change ( Table 2). pyrophosphate synthase (GGPPS2), PSY1, PDS were up-regulated compared to control, and the relative expression of lycopene β-cyclase (CYCB), β-carotene hydroxylase (CrtR-b2) were 3-fold more than control, but the lycopene ε-cyclase (LCY-E) was down-regulated by 84%. In contrast, P450 carotenoid β-and ε-hydroxylases (CYP97A29 and CYP97C11) were up-regulated by 2-fold at B stage.
At P and OR stage, significant increases were observed in 1-deoxy-D-xylulose-5-phosphate synthase (DXS1) that might direct more carbon flux into carotenoids. In contrast, at OR stage, CrtR-b2, CYP97A29 and CYP97C11 were down-regulated by approximately 61% and 65% compared with control ( Table 3).

Transgenic lines showed higher antioxidant activity
To investigate if the enhanced production of carotenoids in transgenic tomato fruits leads to a change of antioxidative activity, we measured the antioxidant activity of BBBB fruits based on DPPH(1,1-diphenyl-2-picrylhydrazyl) radical scavenging activity [28]. All the 3 lines exhibited significantly higher antioxidant activity than that of control, with a tendency of higher levels to higher contents of carotenoids among the transgenic lines ( Figure 4A). In addition, we measured the contents of vitamin C and reduced sugars in the fruit, which showed that transgenic plants produced much more vitamin C. In contrast, the contents of reduced sugars remained unchanged ( Figure 4B).

Discussion
In this study, we manipulated a high-yielding but low carotenoid-containing tomato landrace by the fruit-specific expression of four rate-limiting enzymes which triggered the great enhanced production of native carotenoids as well as significant amounts of ketocarotenoids, leading to up to 44-fold increase of total carotenoids in the transgenic fruit as compared with the mother line, then the enhanced antioxidant activity.
Generally, the constitutive promoter CaMV35S was used to drive gene expression in a variety of plant tissues and developmental stage [29], however, its constitutive characteristics might interfere endogenous metabolic network, normal plant development, and fruit ripening [20,21,30]. In addition, other unexpected results, e.g., susceptibility to environmental influences, ununiform accumulation of target product, were found for transgenic plants using the 35S promoter [31,32]. In this regard, tissue specific promoters could overcome these problems [13,33]. Using the fruit specific promoter 11 E8 to drive four rate-limiting step genes, we generated stable tomato lines that showed normal growth and accumulated ~ 6 mg/g lycopene together with significant amounts of extra canthaxanthin, ketolutein, and astaxanthin, leading to 3-fold increase of total carotenoids in the transgenic fruit as compared with that of common cultivated lines [34]. Furthermore, the involvement of E8 promoter might guarantee the stable growth, development, and carotenoid profiles of the transgenic plants when their progenies were planted in different seasons though small variation of carotenoid contents existed ( Table 1). The small variation of carotenoids among different generations of the transgenic plants could result from such environmental factors as temperature and light that had been shown to influence carotenoid production in tomato fruit [35][36][37], since T 1 plants were planted in winter-spring seasons, whereas T 2 in spring-summer seasons.
The expression of transgenes led to enhanced production of endogenous carotenoids and ketocarotenoids production during fruit ripening ( Table 2). This coincided with the upregulation of most of the endogenous carotenogenic genes (GGPPS, PSY1, PDS, CYCB) at mature stage and sustained upregulation of DXS1 during late stages of fruit ripening (Table 3). Coincidently, ultrastructure of the tomato fruit cells revealed that transgenic cells consisted of much more lycopene crystals and plastoglobules than controls (Supplement Fig. 1). This result supports that besides the synthetic pathway, carotenoid storage also plays a role in carotenoid production.
It is unexpected that the fruit-specific expression of the four transgenes led to the accumulation of lycopene up to 6 mg/g but much less production of ketocarotenoids in the fruit. This could result from uneven translation of the four genes linked by the 2A sequence. 2A-polycistronic system is a useful tool for multigene transformation [38][39][40][41], but the cleavage efficiency and relative expression of the linked genes may not be consistent and therefore could result in unexpected transgene function [42,43]. Possibly, the upstream position of crtB resulted in more efficient conversion of GGPP into phytoene which could be further converted to lycopene by the endogenous downstream genes.
It seemed that the conversion of lycopene to β-carotene remained to be a limiting step in the transgenic lines because lycopene accounted for ~ 90% of total carotenoids. And interestingly, there were no violaxanthin or zeaxanthin in BBBB lines except at MG stage (Table 2), which could result from the low expression of both endogenous hydroxylase (CrtR-b2, CYP97A29 and CYP97C11) ( Table 3) and transgenic HpBHY. Thus, the low production of ketocarotenoids could result from the poor availability of their precursors (β-carotene and zeaxanthin) and poor conversion from canthaxanthin to astaxanthin.
Carotenoid accumulation depends on not only the biosynthetic and metabolic rates but also the sequestration capacity [44,45]. The proportion of esterified astaxanthin are correlate to the final content as esterified carotenoids are more stable and easier to store in plastids [46,47]. Previously we had used 35S promoter to drive overexpression of HpBHY and CrBKT in a Beta mutant of tomato which accumulated much higher amounts of astaxanthin with esterified astaxanthin accounting for 81.3% of total astaxanthin [19]. In contrast, only 22.91% of total astaxanthin was astaxanthin-esters in this study (Table 1). Two main factors differed between the previous and present studies: the promoters and tomato lines used. Thus, to further enhance ketocarotenoid production, stronger fruit specific promoters and β-carotene-rich tomato lines have to be used. Furthermore, the activity of lycopene β-cyclase and β-carotene hydroxylase need to be enhanced. One alternative choice is to cross the transgenic lines with a Beta mutant line which can provide a high level of β-carotene and possibly higher activity of astaxanthin esterification, since similar approaches have been shown to greatly increase ketocarotenoids in tomato fruit [30,31].

Conclusions
In summary, we have successfully manipulated a tomato landrace for high-yield production of carotenoids including the high-value ketocarotenoids specifically in the fruit. The transformants can serve as starting materials for further enhancement of astaxanthin production in tomato. Our study provides insights into metabolic engineering of tomato for enhanced production of value-added carotenoids.  [48]. We harvested nine fruits divided into three groups as three replication samples at each stage, all the samples were stored at -80 ℃ refrigerator until further use.

Extraction and analysis of carotenoids
Carotenoids were extracted and analyzed as described by Huang [19] and Ye [49]. Pigments were identified on the basis of absorption spectra and retention times relative to standard compounds.
Pigments were finally quantified by integrating peak areas that were converted to concentrations by comparison with authentic standards purchased from Sigma-Aldrich.

Gene expression analysis
Total RNA was isolated from leaf, flower, and fruit using Plant RNA Kit (Omega

Analysis of antioxidant activity, reducing sugar and vitamin C
Fruit antioxidant activity analysis was conducted by DPPH radical scavenging activity [28]. Comparing with the standard curve constructed from Trolox, we quantified the antioxidant activity as Trolox equivalent antioxidant capacity (TEAC) in μmol Trolox equivalent per gram of fresh fruit. Total ascorbic acid (vitamin C) content was determined by the 2, 4-dinitrophenylhydrazine method [50].
The total ascorbic acid content was calculated by comparison with ascorbic acid standard curve and absorbance. Reduced sugars concentration was determined using dinitrosalicylic acid assay according to Miller [51].

Statistical analysis
At least three biological replications and two measurement replicates were applied to all the experiments. Data were shown as means ± SD (Standard Deviation) and analyzed using IBM SPSS Statistics 25 (IBM, Armonk, NY, USA). The significant differences between transgenic lines and WT were examined by Student's t-test.

Supplementary Information
Supplementary data associated with this article can be found in the online version. Supplementary

Ethics approval and consent to participate
Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in this published article and its additional files. All plant materials are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This study was financially supported by a grant from Yunnan high talents program which carried out by the Government of Yunnan Province, China. The funding body played no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.