Metabolic Engineering of Crocin Biosynthesis in Nicotiana Species

Crocins are high-value soluble pigments that are used as colorants and supplements, their presence in nature is extremely limited and, consequently, the high cost of these metabolites hinders their use by other sectors, such as the pharmaceutical and cosmetic industries. The carotenoid cleavage dioxygenase 2L (CsCCD2L) is the key enzyme in the biosynthetic pathway of crocins in Crocus sativus. In this study, CsCCD2L was introduced into Nicotiana tabacum and Nicotiana glauca for the production of crocins. In addition, a chimeric construct containing the Brevundimonas sp. β-carotene hydroxylase (BrCrtZ), the Arabidopsis thaliana ORANGE mutant gene (AtOrMut), and CsCCD2L was also introduced into N. tabacum. Quantitative and qualitative studies on carotenoids and apocarotenoids in the transgenic plants expressing CsCCD2L alone showed higher crocin level accumulation in N. glauca transgenic plants, reaching almost 400 μg/g DW in leaves, while in N. tabacum 36 μg/g DW was obtained. In contrast, N. tabacum plants coexpressing CsCCD2L, BrCrtZ, and AtOrMut accumulated, 3.5-fold compared to N. tabacum plants only expressing CsCCD2L. Crocins with three and four sugar molecules were the main molecular species in both host systems. Our results demonstrate that the production of saffron apocarotenoids is feasible in engineered Nicotiana species and establishes a basis for the development of strategies that may ultimately lead to the commercial exploitation of these valuable pigments for multiple applications.


INTRODUCTION
Carotenoids are an important group of natural and lipid-soluble pigments which humans cannot synthesize, thus acquiring them through diet. In humans, these molecules play multiple physiological and nutritional functions (Fiedor and Burda, 2014). Carotenoids are synthesized by plants, algae, fungi, and bacteria, displaying a yellow to red coloration, depending on the type of carotenoid and the concentration they reach in the different cells. Carotenoids are widely used as food colorants, nutraceuticals, animal feed additives, cosmetic ingredients, and health supplements (Fraser and Bramley, 2004). In all living organisms, carotenoids act as substrates to produce apocarotenoids (Ahrazem et al., 2016a). These metabolites are not simply breakdown products of carotenoids, as their functions include those of signaling and hormonal (Walter et al., 2010;Eroglu and Harrison, 2013;Jia et al., 2017). Apocarotenoids are present as volatile and soluble compounds. Crocins are watersoluble pigments of high commercial value, used mostly in food and to a lesser extent, in the pharmaceutical industries (Ahrazem et al., 2015). Crocins are glycosylated derivatives of the apocarotenoid crocetin and they exhibit a strong coloring capacity. In addition, crocins are powerful free radical quenchers, which is associated with the broad range of their health benefits (Nam et al., 2010;Georgiadou et al., 2012;Christodoulou et al., 2015;Bukhari et al., 2018). Interest in the therapeutic properties of crocins is increasing due to their analgesic and sedative properties (Amin and Hosseinzadeh, 2012), neurological protection, and anticancer activities (Finley and Gao, 2017;Skladnev and Johnstone, 2017). Furthermore, clinical trials indicate that crocins have a positive effect in the treatment of depression and dementia (Lopresti and Drummond, 2014;Mazidi et al., 2016). However, commercial utilization of crocins from saffron is limited due to elevated prices, a consequence of the high labor costs in harvesting and processing flower material (Ahrazem et al., 2015).
Crocus sativus is the main natural source of crocins, which accumulate at high levels during the development of the stigma and confer a dark red coloration (Moraga et al., 2009). Together with the safranal (2,6,6-trimethyl-1,3cyclohexadiene-1-carboxaldehyde), the precursor of picrocrocin (β-D-glucopyranoside of hydroxyl-β-cyclocitral) is responsible for the flavor of the saffron spice (Tarantilis et al., 1995). In addition to Crocus, gardenia (Gardenia jasminoides) fruits are also a commercial source of crocins, but on a much lower scale, as they do not accumulate picrocrocin (Pfister et al., 1996;Moras et al., 2018). Among other plants that produce crocins, but which are not commercially exploited due to the low quantities accumulated, are Buddleja species (Liao et al., 1999). Therefore, only a few plant species can synthesize crocins, and this is because the carotenoid cleavage dioxygenases (CCDs) responsible for the production of crocins are not usually present in plants.
The carotenoid biosynthetic pathway has been modified by genetic engineering in higher plants to increase the general amount of carotenoids and/or to produce specific carotenoids (Farre et al., 2014). In previous studies, CsCCD2L was transiently expressed alone and was sufficient to allow the significant crocins accumulation in tobacco plants (Diretto et al., 2019;Marti et al., 2020;López et al., 2021). In this study, we achieved a stable expression of the CsCCD2L gene alone or in combination with genes involved in the carotenoid biosynthesis and accumulation, like BrCrtZ and AtOrMut, to increase the availability of substrates for CsCCD2L activity. We chose the Nicotiana species as a heterologous host. The high metabolic versatility, non-food status, and high biomass production of the genus Nicotiana make this genus one of the most promising biofactories for the sustainable production of high-value metabolites (Molina-Hidalgo et al., 2021). Metabolic analyses showed the accumulation of these nutritional substances reaching in Nicotiana glauca almost 400 µg/g DW, which indicates that N. glauca is a better system than Nicotiana tabacum to produce crocins because the transgenic N. glauca plants expressing CsCCD2L alone could accumulate much higher amounts of crocins as compared with the transgenic N. tabacum plants coexpressing CsCCD2L, BrCrtZ, and AtOrMut genes.

Plant Material
Wild-type (Wt) tobacco (N. glauca and N. tabacum cv. SR1) and transgenic plants were grown in pots with soil (Traysubstrat, Klasmann-Deilmann GmbH, Postfach 1250, 49741 Geeste, Germany) in a controlled growth chamber with a 25/20 • C day/night temperature regime, a 12-h photoperiod (mean irradiance 100 µmol m −2 s −1 ) and 60-90% relative humidity. The fully expanded mature leaves (the 5th and 6th leaves) were collected from five Wt and five transgenic plants for each line. All materials were frozen in liquid nitrogen and stored at −80 • C.
After leaf sample collection, Wt and transgenic N. glauca plants were grown in a controlled growth chamber with a 25/20 • C day/night temperature regime, a 12-h photoperiod (mean irradiance 400 µmol m −2 s −1 ) and 60-90% relative humidity. Wt and transgenic tobacco plants were self-pollinated to obtain seeds.
Agrobacterium-mediated transformation of both tobacco species (N. glauca and N. tabacum) was performed according to the leaf disc method of Horsch et al. (1985). For the selection of transformants, the antibiotic hygromycin B at a final concentration of 50 µg/ml was used.

Carotenoid and Apocarotenoid Extraction and Quantification
Metabolite extraction and analysis differed depending on the nature of the metabolites. Polar and apolar metabolites were extracted from 50 to 5 mg of lyophilized leaves, respectively. For the analysis of polar metabolites, leaves were extracted in cold 50% methanol. The soluble fractions were analyzed by a high-performance liquid chromatography-diode array detector (HPLC-DAD) as previously described (Marti et al., 2020). The insoluble fractions (carotenoids) were extracted with 1:2 cold extraction solvents (50:50 methanol:CHCl 3 ), and analyzed by HPLC-DAD as previously described (Marti et al., 2020). Pigments were quantified by integrating peak areas that were converted to concentrations by comparison with the standards and as reported before (Diretto et al., 2019). All the samples were analyzed in triplicate.

mRNA Expression
Total RNA was isolated using RNeasy Plant Mini Kit (Qiagen, Germany), after DNase treatment, first-strand cDNA synthesis was performed using the SuperScript First-Strand Synthesis System (Takara, Otsu, Japan) primed with oligo (dT)18 following the manufacturer's instructions. qRT-PCR was performed with a Fluorescent Quantitative PCR Detector (Apply Biosystems). SYBR Green real-time PCR Master Mix (Promega) was used. The amplified DNA fragments for each gene were confirmed by Sanger sequencing. The actin gene was used as a reference gene. qRT-PCR products were assessed by melting curve and gel electrophoresis to ensure the specificity of the amplification in the reactions. Three technical replicates were carried out for each biological sample. Conditions for qRT-PCR cycling were 95 • C for 3 min, 95 • C for 20 s, 60 • C for 20 s, 40 cycles. The relative expression level of each gene was calculated using the 2 − Cq method. Primers used in the qRT-PCR analysis are listed in Supplementary Table 1.

Introduction of CsCCD2L in the Nicotiana glauca and Nicotiana tabacum Genomes
We set out to produce crocins in two different host tobacco species N. tabacum and N. glauca. For this purpose, two binary vectors were constructed using the Goldenbraid strategy (Sarrion-Perdigones et al., 2013. The first construct contained only CsCCD2L transgene under the control of the cauliflower mosaic virus (CaMV) 35S promoter together with the hygromycin phosphotransferase selectable marker (hpt) gene (Supplementary Figure 1). The second construct carried three genes: CsCCD2L, BrCrtZ, and AtOrMut is driven by the CaMV35S promoter, the tobacco polyubiquitin Ubi.U4, and the Arabidopsis AtUBQ10 promoter, respectively, together with the hygromycin gene as a selection marker (Supplementary Figure 1).
To produce crocins in stably transgenic plants, two tobacco species (N. tabacum and N. glauca) were transformed with CsCCD2L under the control of the CaMV35S promoter (Supplementary Figure 1). Seventy-four putative transgenic N. tabacum plants (T0) were obtained. Four out of these seventyfour plants showed a bleaching phenotype (Supplementary Figure 2). All plants were screened for apocarotenoid analysis in green leaves by HPLC, and all these transgenic plants accumulated crocins. Wt and all the crocin accumulating transgenic N. tabacum T0 plants were fertile and self-pollinated to obtain T1 seeds. N. tabacum T1 seeds were plated on hygromycin B (at a final concentration of 50 µg/ml) selection MS medium. All the lines showed resistance/susceptible segregation under hygromycin selection. T1 and T2 transgenic plants were grown in a greenhouse as described in section "Materials and Methods" (mean irradiance 100 µmol m −2 s −1 ) and exhibited a normal phenotype (Figure 2) and were fertile. Based on the segregation patterns of hygromycin-resistant sensitive seedlings, 29 out of the 55 lines were single locus for the introduced transgene. Analysis of genomic DNA confirmed the presence of the intact CsCCD2L cassette.
Ten putative transgenic N. glauca plants (T0) were obtained and screened by HPLC-DAD analysis. All of these 10 transgenic T0 lines produced crocins in leaves. After leaf sample collection, Frontiers in Plant Science | www.frontiersin.org Wt, and transgenic N. glauca T0 plants were grown in a controlled growth chamber with a 25/20 • C day/night temperature regime, a 12-h photoperiod (mean irradiance 400 µmol m −2 s −1 ) and 60-90% relative humidity. Wt and all the crocin accumulating transgenic N. glauca T0 plants were fertile and were selfpollinated to obtain T1 seeds, which took approximately one and a half years in our controlled growth chamber. The transgenic N. glauca T1 seeds were plated on hygromycin B (at a final concentration of 50 µg/ml) selection MS medium. All the lines showed resistance/susceptible segregation under hygromycin selection. T1 transgenic plants were grown in a greenhouse as described in section "Materials and Methods" (mean irradiance 100 µmol m −2 s −1 ) and exhibited a faint yellow leaf phenotype (Figure 2). Based on the segregation of hygromycin-resistant sensitive seedlings, seven out of the ten lines were deduced to be single locus transgene insertion lines. Analysis of genomic DNA confirmed the presence of the intact CsCCD2L cassette similar to N. tabacum.

Introduction of CsCCD2L, BrCrtZ, and AtOrMut Mutant Genes in the Nicotiana tabacum Genome
The flower formation of N. glauca depends very much on the season, even in a greenhouse. The plant prefers long-day conditions and high intensity of sunlight. Thus, it may take up to 1 year after seedling transfer to the soil before the first flowers set seeds . Therefore, we have chosen the faster growing N. tabacum as a model to improve crocin biosynthesis in leaves by transformation with one binary vector carrying CsCCD2L, BrCrtZ, and AtOrMut mutant genes driven by three different promoters (CaMV35S promoter, N. tabacum Ubi.U4 promoter, and Arabidopsis ubiquitin 10 promoters, respectively) (Supplementary Figure 1). Seventy-five putative transgenic N. tabacum T0 plants were generated and screened by crocin analysis in leaves. Thirty-two transgenic plants accumulated crocins in leaves. All the crocin accumulating transgenic N. tabacum T0 plants were fertile and self-pollinated to obtain T1 seeds. T1 seeds were plated on a selection MS medium containing hygromycin B. All 32 lines showed resistance/susceptible segregation for hygromycin selection. T1 and T2 transgenic plants were grown in a greenhouse (mean irradiance 100 µmol m −2 s −1 ) and exhibited a yellow leaf phenotype (Figure 2) and were fertile. Based on the segregation of hygromycin-resistant and hygromycin-sensitive tobacco seedlings on selection MS medium supplemented hygromycin B, 15 out of the 32 lines were deduced to be single-locus transgene insertion lines. The genomic DNA was further verified using PCR to check for the presence of CsCCD2L, BrCrtZ, and AtOrMut mutant transgenes.

Crocin Accumulation in Transgenic Nicotiana glauca and Nicotiana tabacum
Polar aqueous extracts from seven T1 N. glauca and six T2 N. tabacum hygromycin resistant plants were subjected to extraction and analysis for the de novo production of crocins. Identification and quantification of the amounts of crocins were conducted by HPLC-DAD using the method described by  (Figures 3A,B) and N. tabacum (Figures 4A,B) the chromatograms of the transgenic lines showed several peaks with maximum absorbance from 433 to 440 nm. Further analyses of the retentions times and the spectra data compared with the standards for crocins led to the identification of crocins with different degrees of glycosylation, ranging from one to four glucose molecules (Figures 3B, 4B). Such apocarotenoids were absent in the Wt extracts.
To evaluate whether the introduction of AtOrMut and BrCrtZ genes in addition to CsCCD2L can boost the level of crocins in N. tabacum, polar extracts from seven T2 homozygous lines derived from single-locus transgene insertion lines were analyzed to determine the amount of crocins (Figure 5). All the lines accumulated higher levels of crocins as compared with those lines from N. tabacum expressing CsCCD2L alone. Interestingly, N. glauca transgenic plants accumulated far higher levels of crocins. Line Nt24 was the top line expressing the three transgenes in N. tabacum, accumulating crocins to 136 µg/g DW.

Carotenoid Profiles in Transgenic
Nicotiana glauca and Nicotiana tabacum Lines accumulating higher levels of crocins were selected for more in-depth analyses. Carotenoids were extracted from four N. tabacum lines expressing the three transgenes (CsCCD2L, BrCrtZ, and AtOrMut), three N. glauca lines, and Wt controls and analyzed by HPLC-DAD (Figures 6, 7 and Supplementary  Tables 2, 3). In general, the carotenoid content of all transgenic lines was much lower compared to their corresponding Wt controls. The most prominent differences in the carotenoid composition of the analyzed leaves were for neoxanthin, β-carotene, and lutein. Neoxanthin, β-carotene, and lutein levels were substantially decreased in the leaves of transgenic plants compared with Wt ( Figures 6C, 7C and Supplementary  Tables 2, 3). Changes in the levels of chlorophyll were more pronounced in transgenic plants of N. glauca than N. tabacum (Figures 6B, 7B).

Expression Analysis of Transgenes and Selected Endogenous Carotenogenic Genes
The expression levels of transgenes and selected endogenous carotenoid biosynthetic genes (PSY1, PSY2, LCYB, and BCH) in leaves were monitored by RT-qPCR (Figures 8A,B). The expression of CsCCD2L in transgenic N. tabacum plants was much higher than those of BrCrtZ or AtOrMut genes, implying that the activity of the CaMV35S promoter for CsCCD2L is much stronger than those of tobacco polyubiquitin Ubi.U4 and the Arabidopsis AtUBQ10 promoters for BrCrtZ and AtOrMut genes, respectively. Endogenous carotenoid biosynthetic genes, including phytoene synthase (PSY1 and PSY2) genes, lycopene β-cyclase (LCYB) and β-carotene hydroxylase (BCH) genes, were downregulated in all the transgenic plants as compared with the Wt control (Figures 8A,B).

DISCUSSION
Crocins are mainly known for being responsible for the color of the saffron spice. In addition to their pigment capacity, crocins show bioactive properties with several therapeutic and pharmacological applications (Mykhailenko et al., 2019). The saffron apocarotenoids had a global market size valued at 374.6 million USD in 2020 and is expected to reach USD 721.5 million by 2028, being the medical application segment the one projected to be the fastest-growing segment. 1 Between 120,000 and 200,000 flowers are needed to produce 1 kg of dried saffron stigma threads, which equates to 370-470 h of work. Consequently, the process is very labor-intensive and risky since it is highly dependent on environmental conditions, leading to high costs.
To overcome these limitations for obtaining saffron apocarotenoids while considering the advances of metabolic engineering, its progression into synthetic biology, and the elucidation of the saffron pathway, we were able to engineer plants from both N. glauca and N. tabacum expressing CsCCD2L. Several efforts have been made by the scientific community to transfer the synthesis of apocarotenoids from saffron to other hosts, looking for the low-cost production of these rare metabolites. It is known that the use of the saffron CsCCD2L enzyme resulted in a maximum accumulation of 1.22, 15.70, and 4.42 mg/l of crocetin in Saccharomyces cerevisiae (Chai et al., 2017;Tan et al., 2019) and Escherichia coli (Wang et al., 2019). In a subsequent study aimed to confirm, in planta, the role of a novel UGTs in picrocrocin biosynthesis (Diretto et al., 2019), Nicotiana benthamiana leaves were transiently transformed, via A. tumefaciens, with CsCCD2L, alone or in combination with UGT709G1, which led to the production of 30.5 µg/g DW of crocins, with a glycosylation degree ranging from 1 to 4 (crocin 1-4). In addition, Marti et al. (2020), using a virus-driven expression of CsCCD2L in adult N. benthamiana plants demonstrated that CsCCD2L expression alone is sufficient for significant crocin accumulation in transiently transformed tobacco (up to 2.18 mg/g DW) (Marti et al., 2020).
The expression of CsCCD2L in N. glauca and N. tabacum allowed the accumulation of notable amounts of up to 400 and 36 µg/g DW of crocins, respectively. The underlying reason for the accumulated crocin variation observed among the analyzed N. glauca and N. tabacum lines could be due to the higher transcript levels of CsCCD2L N. glauca lines, and as well to the higher levels of zeaxanthin present in the leaves of N. glauca plants. The amount of crocins obtained is much higher in lines from N. glauca than in those from N. tabacum and the ones previously described using an A. tumefaciens-mediated transient expression of the CsCCD2L in N. benthamiana leaves, alone or in combination with UGT709G1 (30.5 µg/g DW of crocins). However, this accumulation is lower than the one reported in leaves of adult N. benthamiana using a recombinant virus that expressed CsCCD2L (2.18 mg/g DW of crocins) (Marti et al., 2020). However, crocin production in N. gluaca provides be more effective compared to the data obtained for the production of other carotenoids in this plant; expression of the β-carotene ketolase (CrtO) gene led to a total ketocarotenoid concentration in leaves of 136.6 (young) or 156.1 (older) µg/g dry weight and in petals of 165 µg/g dry weight Mortimer et al., 2017).
The profiles of crocins were different from those reported in saffron stigma, where trans-crocin 4, followed by transcrocin 3 were the major crocins detected. In contrast, in our transgenic plants, trans-crocin-3 was more abundant than trans-crocin-4. In N. benthamiana leaves expressing transiently only CsCCD2L, the major crocin species were the trans-crocin 3, and trans-crocin with two glucose molecules (Supplementary Figure 3). These data indicated that different endogenous UGTs belonging to Nicotiana species recognize crocetin and crocins and can transfer different glucose molecules by changing the qualitative pattern. However, other factors, such as the presence of hydrolase activities that affect the stability of the different crocins synthesized in tobacco cells, and the transport of these crocins to the vacuole and their stability therein, could be influencing the different profiles observed.
It is well-known that BCH overexpression resulted in increased zeaxanthin and xanthophyll contents, both in microbes and plants (Lagarde et al., 2000;Du et al., 2010;Arango et al., 2014). Therefore, overexpression of BCH could enhance the levels of zeaxanthin as the preferred substrate of CsCCD2L (Frusciante et al., 2014). On the other hand, the Orange gene has proven to play an important role in carotenoid accumulation by activating chromoplast differentiation in non-green tissues (Lu et al., 2006), although its expression in green tissues exerts no effect on carotenoid levels (Yuan et al., 2015), we decided to determine whether its introduction could increase the carotenoid content by promoting the formation of carotenoid-sequestering structures. In this context, BrCrtZ and AtOrMut were introduced in N. tabacum to boost the crocin content, since the accumulation of these metabolites was 10-fold lower than N. glauca. The introduction of the two transgenes together with CsCCD2L has most likely led to an increase in the carotenoid pool used as a substrate by the enzyme and, consequently, in a 3.5fold increase in crocin compared to the sole expression of CsCCD2L.
By introducing CsCCD2L alone or in combination with BrCrtZ and AtOrMut, a decreased level of the major and other minor carotenoids normally present in N. glauca and N. tabacum was detected. Among all the carotenoids detected, lutein was one of the carotenoids showing the higher decrease, likely due to the ability of CsCCD2L to act over the β-ring of this molecule as reported in previous in vitro studies where lutein can act as a substrate of this cleavage activity (Frusciante et al., 2014). Zeaxanthin, the crocin precursor, is not a major carotenoid in Nicotiana plant leaves, although the production of the crocins in N. glauca and N. tabacum suggests that the expression of CsCCD2L can drag the leaf metabolic flux toward the production of these apocarotenoids. The level of β-carotene was also lower among the N. glauca transgenic plants, the same phenomenon has been shown in N. tabacum plants transformed with the three genes suggesting a hydroxylation of β-carotene to zeaxanthin under the influence of CsCCD2L in N. glauca and the combination of the exogenous BrCrtZ and CsCCD2L in N. tabacum. Gotz et al. (2002), previously described the increase of zeaxanthin in tobacco plants (N. tabacum L. cv. Samsun) by transformation with a heterologous carotenoid gene encoding β-carotene hydroxylase (CrtZ) from Erwinia uredovora under constitutive promoter control. The increase of zeaxanthin in transgenic plants is due to the catalytic activity of the additional β-carotene hydroxylase rather than being primarily caused by an elevation in the conversion of violaxanthin to zeaxanthin in high light (Gotz et al., 2002).
It has been reported that the expression of an appropriate CsCCD2L in N. benthamiana is sufficient to activate the apocarotenoid pathway in this plant (Diretto et al., 2019;Marti et al., 2020). It seems that both N. glauca and N. tabacum followed the same pattern as N. benthamiana. The crocins produced in this report, using metabolically engineered N. glauca (400 µg/g DW) and N. tabacum (136 µg/g DW), are lower than those obtained in N. benthamiana (2 mg/g DW) using a virus-driven system expressing transient CsCCD2L. However, the stable expression of CsCCD2L can compensate for the low apocarotenoid yields leading to higher apocarotenoid productivities by reducing costs and labor. Further increase of N. glauca and N. tabacum content in these apocarotenoids can also be achieved by combining different types of promoters and the expression of other carotenogenic genes to increase the content of the precursors of crocins, zeaxanthin, and lutein. Thus, an approach of focused metabolic engineering intended toward the production of crocins will establish N. glauca as a potential industrial host for the production of these apocarotenoids. N. glauca is not only a better system for crocin biosynthesis and accumulation than N. tabacum but also a nicotine-free species and perennial plant from which the leafy biomass can be harvested repeatedly for crocin production. , wild-type (Wt) and selected transgenic lines for crocin accumulation (Ng lines expressing CsCCD2L gene; Nt lines coexpressing CsCCD2L, BrCrtZ, and AtOrMut genes). RT-qPCR was used to quantify gene expression levels in three biological replicates per sample. Error bars represent the SD. The asterisk above the bars indicates P-values (Student's t-test) statistically significant (P < 0.05).

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author/s.

AUTHOR CONTRIBUTIONS
CZ, OA, PC, and LG-G conceived, designed the research, and wrote the manuscript. CZ, OA, XH, AR-M, TC, and LG-G conducted the experiments. CZ, OA, XH, AR-M, TC, PC, and LG-G analyzed the data. All authors read and approved the final version of the manuscript.