Water-soluble carotenoid: focused on natural carotenoid crocin

Abstract

Carotenoids are natural isoprenoid compounds with diverse health benefits, widely used in food, cosmetics, and pharmaceuticals. However, low bioavailability and chemical instability limit their effect according to their fat-soluble property. Some strategies such as nanoencapsulation, emulsions, complexation, and glycosylation have been explored to enhance carotenoid bioavailability. In addition, there is growing interest in water-soluble carotenoids in nature. This review focuses on recent advancements in improving the water solubility of carotenoids, with special attention to naturally occurring water-soluble carotenoids like crocin. Research progress on the biosynthetic pathways of crocin derived from natural plants is summarized. In addition, heterologous production using genetic and metabolic engineering in plants and microorganisms is discussed, along with its potential applications in bio-industries. Finally, the promising pharmacological properties of crocin, including antioxidant, anti-inflammatory and anticancer effects, are presented. The sustainable production of water-soluble carotenoids through biological synthesis offers a potential for improved absorption and functionality.

Introduction

Carotenoids are a group of natural isoprenoid compounds produced by plants, fungi, algae, and bacteria. They are synthesized through the terpenoid biosynthesis pathway and function as pigments, producing colors that range from red to yellow (Britton and Hornero-Méndez, 1997; Cazzonelli, 2011). Carotenoids also act as scavengers of reactive oxygen species (ROS), providing antioxidant activity that protects cell membranes from oxidative stress (Cvetkovic et al., 2013; Fiedor et al., 2005). In microorganisms, carotenoids enhance membrane rigidity and strengthen mechanical properties (Kirti et al., 2014; Vila et al., 2019). Additionally, some carotenoids serve as precursors to vitamin A (Johnson and Schroeder, 1996; Zhao et al., 2013). In human health, carotenoids exhibit anticancer and antioxidant properties, contributing to reduced cardiovascular disease risk and improved eye health (Krinsky and Johnson, 2005; Mata-Gómez et al., 2014). Due to these diverse functions, carotenoids are widely used in industries such as food, cosmetics, and pharmaceuticals. However, most carotenoids are insoluble in water, leading to low bioavailability and chemical instability (Apanasenko et al., 2015). As a result, research on improving the water solubility of carotenoids is actively being conducted, with particular interest in naturally occurring water-soluble carotenoids and their derivatives.

Many bioactive molecules, including carotenoids, are hydrophobic compounds and have poor water solubility, which poses a major challenge in pharmaceutical and nutraceutical formulations (van Heijst et al., 2024). Specifically, carotenoids offer many health benefits, but their bioaccessibility and bioavailability in food and pharmaceuticals are low. In particular, the bioaccessibility of carotenes without molecular oxygen is significantly lower than that of xanthophylls, which contain molecular oxygen (Yan et al., 2024). Improving the bioaccessibility and bioavailability of carotenoids is expected to offer numerous benefits to various bio-industries, including reducing the risk of diseases related to carotenoid deficiency. Traditional approaches to solubilize hydrophobic compounds, such as carotenoids, include the use of co-solvents and surfactants, along with chemical modifications to increase hydrophilicity. Additionally, efforts are ongoing to discover water-soluble carotenoids beyond the fat-soluble carotenoids identified in nature. Water-soluble carotenoids are expected to offer superior functionality compared to conventional fat-soluble carotenoids according to their improved absorption and bioavailability (Ribeiro et al., 2020). The sustainable and economical production of water-soluble carotenoids with outstanding functionality through biological synthesis, rather than chemical synthesis, represents a commercially promising technology in nutraceutical and pharmaceutical industries.

This article reviews the latest progress on the water-solubilization of fat-soluble carotenoids to increase bioavailability. It provides a description of naturally occurring water-soluble carotenoids, with a particular focus on crocin, including comprehensive information on its biosynthetic pathways in plants. In addition, the research progress of heterologous production of crocin in plants and microorganisms is discussed. Finally, the potential of crocin in bioindustries is highlighted based on its pharmacological properties.

Research on improving the solubility of carotenoids

Carotenoids are compounds with useful functions, but most of them are fat-soluble, making it difficult for them to fully perform their functions in vivo. Their low water solubility makes it difficult to incorporate them into aqueous systems. Fat-soluble carotenoids require fat for absorption in the digestive system. As a result, water-soluble derivatives and various delivery systems such as encapsulation, emulsions, complexation, and glycosylation techniques are being studied (Table 1).

Table 1.

Strategies of water-solubilization of carotenoid

Strategy	Carotenoid	Methods	Results	References
Emulsion	Lycopene	Pea protein, sodium caseinate	– Constantly maintained particle sizes of emulsion after 14 days of cold storage	Ho et al. (2017)
	Astaxanthin	Whey protein isolate, xanthan gum	– Higher stability of WPI-XG stabilized astaxanthin emulsion at low storage temperature (5–37 °C) for 15 days	Boonlao et al. (2020)
Nanoparticle	β-Carotene	Tween 80, gelatin, pectin	– Determination of optimal particle size and β-carotene loss through nanodispersion (Particle size of 155.8 nm and carotenoids loss of 25.3% w/w)	Anarjan et al. (2017)
	Bixin	Poly-ε-caprolactone (PCL), Tween 80	– Increased stability and bioavailability due to nanoencapsulation of bixin	de Sousa Lobato et al. (2013)
			– Physically stable for 119 days at 25 °C
	Crocin	Lipids, cholesterol	– Enhanced apoptotic effect on cancer cells by liposome encapsulation of crocin	Mousavi et al. (2011)
Complexation	Zeaxanthin, lutein	Glycyrrhizic acid (or its disodium salt), arabinogalactan	– Solubility of carotenoid super molecular complex increased approximately 1000-fold	Apanasenko et al. (2015)
	β-Carotene	Humic acid	– Improved solubility of hydrophobic b-carotene by complexation	Martini et al. (2010)
	Astaxanthin	Disuccinate disodium salt	– Upregulation of connexin 43 protein expression and increased gap junctional communication	Hix et al. (2004)
			– Enhanced bioavailability and function of astaxanthin
Glycosylation	Zeaxanthin	Zeaxanthin glucoside	– 3.5 Times more effective quenching singlet oxygen function due to improved solubility of glucoside form compared to natural zeaxanthin	Tatsuzawa et al. (2000)

An emulsion is a mixture of two immiscible liquids, such as oil and water, where one liquid is dispersed in the other in the form of tiny droplets (Mao and Miao, 2015). These droplets are stabilized by an emulsifying agent (such as surfactants, proteins, or phospholipids) that prevent the liquids from separating. Emulsions are commonly used in food, pharmaceuticals, and cosmetics. Lycopene, a fat-soluble compound, poses challenges for delivery in vivo. To address this, oil-in-water (O/W) emulsions stabilized by proteins have been proposed to enhance lycopene’s chemical stability and bioavailability (Ho et al., 2017).

Plant and dairy proteins serve as emulsifiers in these systems, helping to stabilize oil droplets containing lycopene in water (Mao et al., 2010). Lycopene is hydrophobic and easily degrades due to environmental factors such as light, oxygen, and heat, making its application challenging. Plant- and dairy-based proteins form a stabilizing layer around the water droplets, protecting the lycopene and providing several benefits of stability (Mao et al., 2009). This protein layer adsorbs to the surface of the oil droplets, acting as a physical barrier that prevents droplet coalescence (merging), thus maintaining emulsion stability and protecting lycopene from degradation (Cornacchia and Roos, 2011).

Emulsion-based delivery systems have been utilized to enhance the solubility and bioavailability of astaxanthin (Affandi et al., 2011; Ribeiro et al., 2005). Further research has explored the effects of whey protein isolate (WPI) and xanthan gum (XG) on the stability and digestibility of emulsions containing encapsulated astaxanthin (Boonlao et al., 2020). The combination of WPI and XG enhances the physical stability of the emulsion, preventing phase separation and improving astaxanthin storage retention. The WPI-XG stabilized emulsion demonstrated improved stability over a range of storage temperatures (5–37 °C) for 15 days, showing resistance to environmental stressors (Boonlao et al., 2020). Additionally, the encapsulated astaxanthin in WPI-XG emulsions exhibited improved digestibility, enabling allowing for more efficient release and absorption during digestion (Zhao et al., 2024). This enhanced bioavailability could be beneficial in increasing the nutritional and health benefits of astaxanthin in food and supplement products.

Nanodispersion increases the solubility of poorly water-soluble compounds, allowing them to be more easily dispersed in aqueous systems. Nanodispersion also improves the stability and bioavailability of compounds like carotenoids (Tan and Nakajima, 2005). For β-carotene nanodispersion, the interactions between three selected food-grade stabilizers—Tween 80, gelatin, and pectin—were investigated during the formation of carotenoid nanoparticles via a solvent displacement process (Anarjan et al., 2017). This research optimized the three-component stabilizer system to produce β-carotene nanodispersions with minimal particle size and carotenoid loss.

A study has shown that nanoencapsulation is an effective method for improving the water solubility of hydrophobic bioactive compounds, such as carotenoids (Couvreur et al., 2002). This technique reduces the particle size to the nanoscale (below 100 nm), thereby increasing the dissolution rate. The smaller particle size provides a larger surface area, facilitating greater interaction with water molecules and enhancing solubility (Paese et al., 2009; Shaikh et al., 2009). For example, nanocapsules are prepared using poly-ε-caprolactone (PCL), a biodegradable polymer known for its stability and biocompatibility (Pohlmann et al., 2013). In this process, bixin is dissolved in an organic solvent along with PCL, which is then added to an aqueous phase containing an emulsifier, such as Tween 80 or Polysorbate 80, to form an oil-in-water emulsion. During organic solvent evaporation, PCL precipitates around the bixin molecules, forming nanocapsules in which bixin is encapsulated within a polymer shell. The surfactant in the aqueous phase stabilizes these nanocapsules, preventing aggregation. The resulting bixin nanocapsules exhibited a monomodal distribution and an average diameter of 195.0 nm. Additionally, these nanocapsules retained their physical stability for 119 days at 25 °C (de Sousa Lobato et al., 2013).

Nanoliposomes are another material that can encapsulate carotenoids and improve their delivery. For instance, liposomal crocin was produced using lipids and cholesterol through a dehydration-rehydration process (Malaekeh-Nikouei and Davies, 2009). To achieve uniform liposome size, the formulation was passed through a polycarbonate membrane at least 11 times. Liposomal encapsulation enables crocin to cross cellular barriers more effectively, improving its cellular and tissue uptake compared to free crocin. Notably, the nanoliposomal form of crocin exhibited a stronger cytotoxic effect than standard crocin. Specifically, the IC₅₀ (the concentration required to kill 50% of cells) was lower for liposomal crocin, indicating higher potency (Mousavi et al., 2011). This research underscores the potential of nanotechnology to enhance the delivery and efficacy of natural anticancer agents like crocin.

Recent research has investigated the water solubilization and stabilization of carotenoids using water-soluble oligosaccharides and polysaccharides. For instance, supramolecular complexes of lutein and zeaxanthin with arabinogalactan, glycyrrhizic acid, and its disodium salt have been studied (Apanasenko et al., 2015). The non-covalent binding of lutein and zeaxanthin with arabinogalactan and glycyrrhizic acid (or its disodium salt) forms water-soluble aggregates with distinct physicochemical properties (Foscan et al., 2019). These aggregates significantly improve the water solubility of carotenoids and reduce their tendency to aggregate. These changes significantly affect their optical and photophysical properties (Polyakov et al., 2010). Lutein and zeaxanthin, when complexed with arabinogalactan (AG) and glycyrrhizic acid (GA) or its disodium salt (sGA), exhibit a marked increase in oxidative stability. The study showed that the solubility of carotenoids increased by more than 1000-fold, highlighting the increased resistance of these carotenoids to oxidative degradation in these complexes. The unique properties of GA and AG complexes were also demonstrated with two other natural carotenoids, β-carotene and canthaxanthin (Polyakov et al., 2013).

In another study, researchers explored the use of humic acids, natural organic substances, to form complexes with β-carotene (Martini et al., 2010). This complexation significantly increased the water solubility of β-carotene, allowing it to disperse more effectively in aqueous systems. β-carotene is prone to degradation when exposed to light, a process known as photodegradation (Hejri et al., 2013). The study found that complexation with humic acid enhanced the photostability of β-carotene by protecting it from light-induced degradation.

To further improve the solubility and bioavailability of astaxanthin, research has focused on creating water-soluble derivatives, particularly disodium disuccinate astaxanthin (Hix et al., 2004). Disodium disuccinate astaxanthin is a chemically modified version of astaxanthin, in which disuccinate groups are attached to the molecule, making it water-soluble (Lauver et al., 2008). The water solubility of this derivative allows it to be more efficiently absorbed and utilized by cells, particularly in tissues where fat-soluble compounds struggle to penetrate. This modification enhances the bioactivity of astaxanthin improving its effectiveness in antioxidant protection and cellular function.

Glycosylation of carotenoids also makes them water-soluble. By adding glycosyl groups, the overall polarity of the molecule increases, making it more compatible with water, a polar solvent (Chen et al., 2021). In addition to improving solubility, glycosylation enhances the stability of carotenoids, particularly in terms of oxidation resistance and long-term storage (Háda et al., 2012). In one study, the viability of Escherichia coli transformants expressing zeaxanthin-glucoside was 1.5 times higher than that of transformants expressing other carotenoids, including zeaxanthin. Additionally, zeaxanthin diglucoside was found to be about 3.5 times more effective at quenching singlet oxygen compared to non-glycosylated zeaxanthin (Tatsuzawa et al., 2000). These findings confirm that glycosylation increases the water solubility of zeaxanthin, thereby enhancing its functionality in aqueous environments in vivo.

Water-soluble carotenoid in nature

As previously mentioned, most carotenoids naturally occur as fat-soluble compounds. However, there are two notable examples of carotenoids that are naturally water-soluble carotenoids (Fig. 1).

Fig. 1.

Structure of the natural water-soluble carotenoids

Norbixin is a natural water-soluble carotenoid derived from bixin, which is extracted from the seeds of Bixa orellana (annatto plant) (Rodrigues et al., 2014). Bixin and norbixin are the primary pigments responsible for the reddish-orange color of annatto seeds and are commonly used as food colorants. Bixin is formed in the seeds of the annatto plant as a derivative of lycopene, a C₄₀ carotenoid (Bouvier et al., 2003). The enzyme carotenoid cleavage dioxygenase (CCD) acts on lycopene, cleaving its structure to generate the apocarotenoid precursor, which eventually forms bixin aldehyde (Cárdenas-Conejo et al., 2015; Rivera-Madrid et al., 2016). This aldehyde is then converted into bixin, which has a carboxylic acid group and a methyl ester group at each end, through the action of dehydrogenase and methyltransferase. When treated with an alkali, the methyl ester groups of bixin are cleaved, resulting in the formation of norbixin, which retains the same basic structure but now has free carboxyl groups at both ends. These carboxyl groups make norbixin more polar, thereby increasing its water solubility. The carboxyl groups can form hydrogen bonds with water molecules, improving norbixin’s solubility. Norbixin is frequently used as a coloring agent in dairy products such as cheddar cheese, yogurt, dairy drinks, and ice cream due to its water-solubility (Lancaster and Lawrence, 1996; Lee et al., 2023). In cheese production, norbixin binds effectively to dairy proteins, enhancing color retention and stability. In ice cream, it can produce various shades, from light yellow in vanilla-flavored products to bright orange in mango or tropical flavors, depending on the concentration. Initially, annatto extract contains mainly bixin, which has low water-solubility. To convert bixin into water-soluble norbixin, manufacturers treat the extract with alkaline agents such as potassium or sodium hydroxide in a process called saponification (Aluko, 2024). This chemical reaction transforms bixin into norbixin, making it suitable for use in water-based applications.

Crocin is also a natural water-soluble carotenoid derived from the dried stigmas of Crocus sativus (commonly known as saffron) (Bhandari, 2015). It is the primary component responsible for saffron’s characteristic color. Recent studies have shown that crocin holds therapeutic potential for various human diseases, including cancer, Alzheimer’s disease (AD), nervous system disorders, and cardiovascular diseases (Finley and Gao, 2017; Godugu et al., 2020; Hosseinzadeh and Nassiri-Asl, 2013). However, saffron, the raw material for crocin, is the most expensive spice due to its labor-intensive harvesting process, low yield, and high cost. As a result, alternative methods for crocin synthesis have been explored. Currently, industrial carotenoid production primarily relies on chemical synthesis (Misawa, 2011). However, chemical synthesis of carotenoids can generate toxic waste, leading to undesirable side effects (Kirti et al., 2014). Moreover, crocin is challenging to synthesize chemically due to its complex structure and multiple chiral centers. Chemical synthesis can also result in the formation of inactive or toxic isomers (Liu et al., 2020a, b). Therefore, engineered plant and microbial production of crocin has emerged as a viable alternative to chemical synthesis. This approach offers a sustainable and cost-effective method for carotenoid production. In the crocin biosynthesis pathway in C. savitus, precursor of crocin, crocetin is synthesized first (Bhandari, 2015). Crocetin is less water-soluble than crocin because it is a dicarboxylic acid composed of hydrophobic conjugated polyenes. Crocetin has two carboxyl groups at either end of its molecule, and in the process of glycosylation, glucose or disaccharide gentiobiose attaches to these carboxyl groups, transforming crocetin into five forms of crocin. According to the position and numbers of these two glycosyls (glucosyl and gentiobiosyl), crocins could be divided into five forms, crocin-I, crocin-II, crocin-III, crocin-IV, and crocin-V (Ding et al., 2018; Koulakiotis et al., 2020). Gentiobiose or glucose are inherently hydrophilic compounds due to their many hydroxyl (–OH) groups. These hydroxyl groups can form hydrogen bonds with water molecules, facilitating the dissolution of the compound in water. Therefore, these glycosylation reactions make carotenoids water-soluble. We focused on crocin and investigated its biosynthesis in plants, the heterologous production of crocin using both plants and microorganisms, and its pharmacological effects.

Crocin biosynthesis pathway in plant

Crocin is naturally biosynthesized in plants such as Crocus sativus (saffron), Gardenia jasminoides (cape jasmine), and Buddleja davidii (lilac) (Diretto et al., 2021; Frusciante et al., 2014; Zhang et al., 2023). In Crocus, Gardenia, and Buddleja spp., crocin is produced through the non-mevalonate (MEP) pathway, C₄₀ carotenoid biosynthesis pathway, and crocin biosynthesis pathway (Fig. 2). First, in the MEP pathway in plastid, which is responsible for the biosynthesis of isoprenoid precursors found in most eukaryotes, pyruvate and glyceraldehyde 3-phosphate (GA3P) are converted into 1-deoxy-D-xylulose 5-phosphate (DXP) by DXP synthase (DXS). Subsequently, through the action of various enzymes involved in the MEP pathway (DXR, CMS, CMK, MCS, HDS, and HDR), C₅ isoprenoid pyrophosphate (IPP) and C₅ dimethylallyl pyrophosphate (DMAPP) are finally synthesized, which serve as precursors for carotenoid biosynthesis.

Fig. 2.

Overall crocin biosynthetic pathways in Crocus sativus, Gardenia jasminoides, and Buddleja davidii. Carotenoid precursors are synthesized via the 2-C-methyl-D-erythritol 4-phosphate (MEP) pathway. Crocin precursors are synthesized via the carotenoid biosynthesis pathway. Enzymes marked in blue and yellow represent those involved in the MEP pathway and carotenoid biosynthesis pathway, respectively. While enzymes in purple, green, and orange are identified from C. sativus, G. jasminoides, and B. davidii, respectively. Key enzymes include DXS (1-deoxy-D-xylulose-5-phosphate synthase), DXR (1-deoxy-D-xylulose 5-phosphate reductoisomerase), MCT (2-C-methyl-D-erythritol 4-phosphate cytidylyltransferase), CMK (4-(cytidine 5′-diphospho)-2-C-methyl-D-erythritol kinase), MDS (2C-methyl-d-erythritol 2, 4-cyclodiphosphate synthase), HDS (4-hydroxy-3-methylbut-2-enyl diphosphate synthase), HDR (4-hydroxy-3-methylbut-2-en-1-yl diphosphate reductase), IPPI (isopentenyl pyrophosphate isomerase), GGPPS (geranylgeranyl diphosphate synthase), PSY (phytoene synthase), PDS (phytoene desaturase), Z-ISO (ζ-carotene isomerase), ZDS (ζ-carotene desaturase), CRTISO (carotenoid isomerase), LCYB (lycopene β-cyclase), BHY (β-carotene hydrolase), CCD (carotenoid cleavage dioxygenase), ALDH (aldehyde dehydrogenase), and UGT (UDP-glucosyltransferase)

Second, the conversion of two C₅ precursors molecules into the C₁₀ geranyl pyrophosphate (GPP) by the enzyme GPP synthase (GPPS) marks the first committed step in the carotenoid biosynthesis pathway. Following this, two molecules of GPP are combined to form C₂₀ geranylgeranyl pyrophosphate (GGPP), a reaction catalyzed by GGPP synthase (GGPPS). The C₄₀ 15-cis-phytoene, which is formed by the combination of two molecules of GGPP by phytoene synthase (PSY), undergoes a series of desaturation and isomerization reactions that generate the chromophores characteristic of carotenoids. In this process, phytoene is converted to ζ-carotene and pro-lycopene through the actions of desaturases (PDS and ZDS) and isomerases (ZISO and CrtISO), without the addition of carbon, forming all-trans lycopene. Lycopene can subsequently convert cyclization and hydroxylation to produce xanthophyll carotenoid compounds. Lycopene is converted into either ε-carotene or β-carotene depending on the cyclization enzyme involved ε-lycopene cyclase (LCYE) or β-lycopene cyclase (LCYB), respectively. These intermediates then lead to the production of the carotenoid lutein and zeaxanthin by hydroxyl enzymes. The key enzymes involved in carotenoid biosynthesis, including GPPS, GGPPS, PSY, PDS, ZDS, ZISO, CRTISO, and LCYB, have been identified in most plants known for carotenoid production.

Third, the crocin biosynthesis pathway begins with the cleavage of the ring structure of zeaxanthin. Zeaxanthin is an important precursor for apocarotenoid biosynthesis and is converted by different carotenoid cleavage dioxygenases (CCDs), leading to various apocarotenoid compounds. Among crocin biosynthesis plants, cleavage of the zeaxanthin molecule at the 7,8 and 7′,8′ double bonds produce one molecule of C₂₀ crocetin dialdehyde and two molecules of 4-hydroxy-2,6,6-trimethyl-1-cyclohexene-1-carboxaldehyde (HTCC) by CsCCD2L in C. sativus and BdCCD4.1/3 in B. davidii. Similarly, GjCCD4a enzyme identified from G. jasminoides exhibits the same catalytic activity as CsCCD2L and BdCCD4.1/3 in cleaving zeaxanthin. However, GjCCD4a catalyzes the production of crocetin dialdehyde largely independent of the substrate, specifically cleaving the 7,8 and 7′,8′ double bonds of lycopene and β-carotene in a manner similar to its cleavage of zeaxanthin (Xu et al., 2020). Crocetin dialdehyde is further converted into C₂₀ crocetin by the action of aldehyde dehydrogenases (ALDH), specifically CsALDH3I1, GjALDH2C3, and BdALDH. They undergo a hydroxylation reaction, adding hydroxyl groups to the aldehyde groups at both ends of crocetin dialdehyde. The final step in crocin biosynthesis is the glycosylation of lipid-soluble crocetin by UDP-glycosyltransferases (UGTs), resulting in the production of water-soluble crocin. Glycosylation reactions catalyzed by UGTs act on the carboxyl and glycosyl groups of crocetins and crocins to form five types of crocins: crocin-I (crocetin di-(β-D-gentiobiosyl) ester), crocin-II (crocetin (β-D-gentiobiosyl) (β-D-glucosyl), crocin-III (crocetin di-(β-D-glucosyl) ester), crocin-IV (crocetin (β-D-gentiobiosyl) ester), and crocin-V (crocetin (β-D-glucosyl) ester) (Ding et al., 2018).

In C. sativus, a distinct UGT has been identified as the key enzyme responsible for crocin biosynthesis in the stigma tissue (Moraga et al., 2004). UGTCs2 was shown to promote the formation of highly glucosylated crocetin esters by using crocetin, crocetin β-D-glucosyl ester, and crocetin β-D-gentiobiosyl ester as substrates, indicating that the enzyme exhibits both glucosylation activities in vitro (Moraga et al., 2004). A variant of UGTCs2, CsUGT74AD1, was found to be responsible for the primary glycosylation of crocetin, producing crocin-V and crocin-III. CsUGT91P3 plays a crucial role in the final steps of crocin biosynthesis by adding multiple glucose molecules to crocin in C. sativus (López-Jimenez et al., 2021). Specifically, CsUGT91P3 can catalyze the addition of three, four, and five glucose molecules to crocin. CsUGT91P3 does not use crocetin directly as a substrate but works in tandem with CsUGT74AD1, which first produces crocins with one or two glucose molecules from crocetin. CsUGT91P3 then further glycosylates these crocins and produces compounds such as crocin-II and crocin-I.

Two UGT, GjUGT75L6 and GjUGT94E5 are known to mediate sequential glucosylation of crocetin to crocin in G. jasminoides, as low expression levels in vivo (Nagatoshi et al., 2012). Among the three UGTs (GjUGT74F8, GjUGT75L6, and GjUGT94E13), it has been suggested that they are capable of glycosylating crocetin and crocins (Xu et al., 2020). Especially, GjUGT74F8 and GjUGT94E13 were highly expressed in G. jasminoides’s fruits, suggesting their involvement in crocin production. GjUGT74F8 was the most highly expressed UGT gene and catalyzed the primary glycosylation of crocetin and produced crocin-V, similar to the reaction catalyzed by C. sativus CsUGT74AD1 (Demurtas et al., 2018). Glycosylation occurs on the remaining carboxyl groups in crocin-IV and crocin-V, leading to the production of crocin-II and crocin-III by GjUGT74F8. In addition, the reversible conversion of crocin-II to crocin-IV and vice versa was observed, suggesting that GjUGT74F8 also possesses hydrolytic activity, capable of removing a single β-D-glucosyl group. GjUGT94E13 can catalyze either the addition of a second glucosyl group to an existing one (secondary glycosylation) or the sequential addition of two glucosyl groups to a carboxyl group (primary and secondary glycosylation). GjUGT94E13 catalyzed the primary and secondary glycosylation of crocetin to form crocin-IV and crocin-I, likely through the sequential addition of β-D-gentiobiosyl esters. Additionally, crocin-II and crocin-III can be completely converted to crocin-I by GjUGT94E13, which adds a second glucosyl moiety to an β-D-glucosyl group through secondary glycosylation.

In B. davidii, crocetin glucosylation is catalyzed by BdUGT74BC1 and BdUGT74BC2, both of which belong to the family 1 plant GTs (Diretto et al., 2021). They catalyze primary glycosylation of crocetin as crocin-III and crocin-V. BdUGT74BC1, similar to CsUGT74AD1 and GjUGT74F8, is involved only in primary glucosylation, adding a β-D-glucosyl group to crocetin or crocin-V. BdUGT74BC2 can produce crocin-IV by glycosylating the β-D-gentiobiosyl group, but this reaction can only use crocetin as a substrate, not crocin. The activity of BdUGT74BC1 and BdUGT74BC2 is not possible to produce crocins with three or more glucose residue units. Although crocin-I and crocin-II have been identified in B. davidii, the genes responsible for the sequential addition or secondary glucosylation of the β-D-glucosyl group in crocin-IV and crocin-III remain unclear. A member from the UGT94 family is also involved in the crocin glucosylation in B. davidii (BdUGT94AA3) like G. jasminoides (Nagatoshi et al., 2012; Xu et al., 2020). Furthermore, the activity of BdUGT94AA3 was tested on crocin-II, enabling the formation of crocin-I.

In this manner, different UGT types are classified based on their substrate specificity and glycosylation sites, and these vary across plants that biosynthesize crocin (Liu et al., 2020a, b). Above UGTs can be categorized into three types. UGT-I (UGTCs2/CsUGT74AD1 in C. sativus; GjUGT75L6/GjUGT74F8 in G. jasminoides; BdUGT74BC1 in B. davidii) is responsible for the primary glycosylation of crocetin, producing crocin-III and crocin-V. UGT-II (CsUGT91P3 in C. sativus; GjUGT94E5 in G. jasminoides; BdUGT94AA3 in B. davidii) is involved in the secondary glycosylation of the glycosyl group, forming one or two gentiobiosyl groups. UGT-III (GjUGT94E13 in G. jasminoides; BdUGT74BC2 in B. davidii) performs both primary and secondary glycosylations. With further research on related genes, these UGTs can serve as valuable genetic resources for identifying biosynthetic pathways and facilitating heterologous production.

Production of crocin in heterologous plant systems

C. sativus is the primary natural source of crocin, which accumulates at high levels during pistil development (Moraga et al., 2009). G. jasminoides fruits are also a commercial source of crocin. However, they do not accumulate picrocrocin which is a compound produced during crocin biosynthesis (Moras et al., 2018; Pfister et al., 1996). Additionally, Buddleja species also produce crocin, but they are not commercially viable due to low accumulation levels (Liao et al., 1999). Consequently, research has focused on developing heterologous production systems in different plant species to enable the commercial production of crocin. The heterologous production of crocin is most performed in engineered Nicotiana species. Nicotiana species grow naturally in many regions worldwide and have been used as plant factories for the high-yield production of a wide range of metabolites (Gómez-Gómez et al., 2023). For heterologous production of crocin, it is essential to introduce CCDs, which is not commonly present in most plants (Ahrazem et al., 2022a).

Recently, many plants have been successfully transformed to produce crocin (Table 2). Viral vector and plasmid-based systems are the most effective methods for producing crocin in Nicotiana species. The first attempt to produce and accumulate crocin in plants that do not naturally produce these compounds was achieved using a viral vector derived from tobacco etch virus (TEV, genus Potyvirus, family Potyviridae). A virus-driven expression vector (pGTEVΔNIb) was used to insert the CsCCD2L and BdCCD4.1 genes into N. benthamiana for the production of crocin (Martí et al., 2020). This approach resulted in crocin accumulation of 2.18 mg/g dry weight (DW) in N. benthamiana leaves. Further optimization, involving the introduction of PaCrtB (phytoene synthase from Pantoea ananais) and CsBCH2 (β-carotene hydroxylase 2 from C. sativus), increased the crocin yield to 3.493 mg/g DW. Similarly, stable crocin production in N. glauca was achieved by constructing the pUPD2-BdCCD4.1 vector, which placed the BdCCD4.1 gene under the control of the CaMV35S promoter (Huang et al., 2022). Transgenic N. glauca lines produced crocin at levels of 321.6 µg/g DW in petals and 302.7 µg/g DW in leaves, demonstrating that the engineered species could serve as a viable crocin production platform. Crocin biosynthesis in N. glauca and N. tabacum was further advanced through the construction of two binary vectors (Ahrazem et al., 2022a). One vector carried CsCCD2L under the control of the CaMV35S promoter, while the second vector carried CsCCD2L, BrCrtZ (β-carotene hydroxylase from Brevundimonas sp.), and AtOrMut (a mutant gene from Arabidopsis thaliana). In this study, engineered N. glauca produced up to 400 µg/g DW of crocin in its leaves, and co-expression of BrCrtZ and AtOrMut enhanced crocin production in N. tabacum to 136 μg/g DW.

Table 2.

Heterologous crocin biosynthesis in plants

Host	Enzymesa	Sources	Production	Yield	References
Nicotiana benthamiana	CrtB	Pantoea ananais (CrtB)	Crocin	3.493 mg/g DW in leaves	Martí et al. (2020)
	BCH	Crocus sativus (BCH2, CCD2L)
	CCD	Buddleja davidii (CCD4.1/3)
Nicotiana glauca	CCD	Buddleja davidii (CCD4.1)	Crocin	321.6 µg/g DW in petals	Huang et al. (2022)
				302.7 µg/g DW in leaves
Nicotiana glauca	CrtZ	Brevundimonas sp. (CrtZ)	Crocin	400 µg/g DW in leaves	Ahrazem et al. (2022a)
	AtOrMut	Arabidopsis thaliana (AtOrMut)
	CCD	Crocus sativus (CCD2L)
Nicotiana tabacum	CrtZ	Brevundimonas sp. (CrtZ)	Crocin	136 µg/g DW in leaves	Ahrazem et al. (2022a)
	AtOr	Arabidopsis thaliana (AtOrMut)
	CCD	Crocus sativus (CCD2L)
Nicotiana benthamiana	ALDH	Gardenia jasminoides (CCD4a, ALDH2C3, UGT74F8, UGT94E13)	Crocin	1.058 mg/g DW in tobacco	Xie et al. (2023)
	CCD
	UGT
Nicotiana benthamiana	CrtB	Bacterial strain (CrtB)	Crocin	1.61 mg/g DW in leaves (CCD4a)	Zheng et al. (2022)
	BCH	Oryza sativa (BCH)		0.67 mg/g DW in leaves (CCD2L)
	CCD	Crocus sativus (CCD2L)
		Gardenia jasminoides (CCD4a)
Nicotiana benthamiana	CCD	Crocus sativus (CCD2L)	Crocin	883.7 µg/g DW in leaves	Demurtas et al. (2023)
Nicotiana benthamiana	CCD	Bixa orellana (CCD4-3)	Crocin	0.039 mg/g DW in leaves	Frusciante et al. (2022)
Nicotiana tabacum	CCD	Bixa orellana (CCD4-3)	Crocin	0.059 mg/g DW in leaves	Frusciante et al. (2022)
Solanum lycopersicum	CCD	Bixa orellana (CCD4-3)	Crocin	110 µg/g DW in fruits	Frusciante et al. (2022)
	UGT	Crocus sativus (CCD2L, UGT74AD1)
Solanum lycopersicum	CCD	Crocus sativus (CCD2L, UGT74AD1)	Crocin	14.48 mg/g DW in fruits	Ahrazem et al. (2022b)
	UGT
Solanum tuberosum	CCD	Crocus sativus (CCD2L, UGT74AD1)	Crocin	360 µg/g in tubers	Gómez-Gómez et al. (2022)
	UGT
Solanum tuberosum	CCD	Crocus sativus (CCD2L, UGT74AD1)	Crocin	3.648 mg/g DW in tubers	Gómez-Gómez et al. (2024)
	UGT
Citrus paradisi	CrtB	Bacterial strain (CrtB)	Crocin	0.022 mg/g in fruits	Zheng et al. (2022)
	BCH	Oryza sativa (BCH)
	CCD	Crocus sativus (CCD2L)
		Gardenia jasminoides (CCD4a)

^aEnzyme abbreviations are as follows: CrtB, phytoene synthase; BCH, β-carotene hydroxylase; CCD, carotenoid cleavage dioxygenase; CrtZ, β-carotene hydroxylase; AtOr, carotenoid accumulation gene from Arabidopsis thaliana; AtOrMut, mutant gene derived from AtOr; ALDH, aldehyde dehydrogenase; UGT, UDP-glycosyltransferase; DW, dried weight

Multiple gene introduction and comparative studies have explored the effects of different genes on crocin production. The insertion of multiple crocin biosynthesis genes (GjCCD4a and GjALDH2C3) and two UDP-glucosyltransferase genes (GjUGT74F8 and GjUGT94E13) into N. benthamiana was explored (Xie et al., 2023). This approach used two double-gene expression vectors linked by 2A peptides and ultimately combined these into a multigene expression vector named AU-CU. In the T1 generation, the highest crocin content reached 1.058 mg/g DW, with crocin-I and crocin-II comprising up to 99% of the total crocin component. A comparative study by Zheng et al. (2022) introduced two different CCD genes, CsCCD2L and GjCCD4a, into N. benthamiana alongside TpCrtB (phytoene synthase fused with plastid transit peptide) and OsBCH (β-carotene hydroxylase from Oryza sativa). After 6 days of agroinfiltration, GjCCD4a produced higher crocin levels (1.61 mg/g DW) compared to CsCCD2L (0.67 mg/g DW), demonstrating the efficiency of GjCCD4a in crocin production.

In a notable study, Demurtas et al. (2023) improved crocin yields using genome editing techniques to accumulate high levels of zeaxanthin. By employing a viral vector to express CsCCD2L, they achieved efficient crocin biosynthesis. The highest crocin production in genome-edited N. benthamiana plants reached 883.7 µg/g DW using a TEV-based expression system, which significantly outperformed traditional agroinfiltration methods.

Interestingly, Bixa orellana (known for producing bixin, a fat-soluble carotenoid) was also explored for crocin production. Although B. orellana does not naturally produce crocin, Frusciante et al. (2022) showed that BoCCD4-3, when combined with a transit peptide, produced 0.039 mg/g DW of crocin in N. benthamiana leaves and 0.059 mg/g DW in N. tabacum. Overall research demonstrates that Nicotiana species, which do not have natural crocin biosynthesis pathways, can be engineered to produce crocin.

A variety of plant species, such as Solanum lycopersicum (tomato), Solanum tuberosum (potato), and citrus, are attractive as hosts for the heterologous production of crocin. Tomato is optimal platform for crocin production due to its high lycopene content, a precursor of zeaxanthin, which is essential for crocin biosynthesis. Frusciante et al. (2022) demonstrates the feasibility of using tomatoes as a biotechnological platform for crocin production. Transformation of tomato was performed using pBI121 vector constructs containing BoCCD4-3 fused with the CsCCD2L transit peptide, driven by the 35S promoter. The second vector including CsUGT74AD1 also under the control of the 35S promoter. Transgenic fruits with an orange phenotype exhibited a maximum crocin productivity of 110 µg/g DW, with 81% of the total production consisting of crocin with more than 3 glucose moieties, and approximately 18% accounting for other crocin less than 3 glucose moieties.

The combination of promoters has affected crocin biosynthesis gene expression and crocin production. Ahrazem et al. (2022b) constructed plasmids based on pUPD vector with CsCCD2L, CsUGT74AD1, and CsUGT709G1. Three different fruit-specific promoters (pE8, p2A11, and p35S) were used to express the saffron biosynthetic genes in tomatoes. The constructs were designed to express in the order of CsCCD2L, CsUGT2, and CsUGT709G1. Under 4 different promoter combinations (O1, p35S-p35S-pE8; O2, p2A11-p35S-pE8; O3, p35S-p35S-p35S; O4, pE8-p35S-p2A11), finally 4 binary vectors were created. The engineered tomatoes accumulated up to 14.48 mg/g DW of crocin. Especially, the extract from transgenic tomatoes (more than 10 μg/ml of crocin contents) significantly delayed paralysis in Caenorhabditis elegans CL4176 induced by Aβ1–42 expression representing human model of AD. The transgenic extract performed as well as the Ginkgo biloba extract and showed better results compared to the wild-type tomato extract. This study demonstrates that the potential of high-valuable hydrophilic apocarotenoid crocin have benefits for human health. To maximize crocin production, the tomato variety (Tomaffron) transformed with the optimal saffron cassette O1 was used in subsequent studies (Lobato-Gómez et al., 2024). Xantomato was transformed with the CsCCD2L gene to enhance the production of saffron apocarotenoids. The optimal Xantoffron, developed through heterozygous hybridization of two transgenic tomatoes, achieved 15 times more crocin production than the existing Tomaffron variety.

A strategy was devised to produce crocin using potatoes, a genus closely related to tomatoes. Gómez-Gómez et al. (2022) suggested the production of crocin in a heterologous plant system by generating stable transgenic lines, expressing the CsCCD2L, CsUGT74AD1, and CsUGT709G1 genes under the control of the patatin promoter. The coding sequences of CsCCD2L, CsUGT74AD1, and CsUGT709G1 were synthesized and cloned into the level 0 vector pUPD2. The final binary construct, labeled O6 = pDGB32 [pPATCsCCD2LT35S-pNos-Hyg-T35S-pPATUGT74AD1T35S-pPATUGT709G1T35S-], was assembled with a tuber-specific patatin promoter to drive the expression of the three genes. The engineered potato tubers accumulated 360 µg/g DW of crocin in the optimal lines. In a follow-up study on this, Gómez-Gómez et al. (2024) engineered potato lines Group Phureja 01H15, which naturally accumulate high levels of zeaxanthin. They are a highly effective platform for producing saffron apocarotenoids. In the engineered potato lines, crocin accumulation reached 3.648 mg/g DW, up to tenfold higher than those obtained in the previous study with same genetic tool.

In a detailed study of Zheng et al. (2022), white-colored callus from Citrus paradisi which naturally contains few of carotenoid levels, was genetically engineered to produce various lines expressing combinations of carotenoid and apocarotenoid biosynthetic genes (tpCrtB, OsBCH, GjCCD4a, and CsCCD2L). The first step in this Agrobacterium-mediated serial transformation involved introducing the bacterial PSY gene fused with a tpCrtB, resulting in an orange-red callus according to producing β-carotene and α-carotene. In the second step, the OsBCH gene was introduced to remodel the xanthophyll content, creating PSY/BCH yellow callus lines, which accumulated substantial levels of β,β-xanthophylls while containing lower levels of cyclic carotenoids. Subsequently, transformed calli containing the GjCCD4a/BCH/PSY combination achieved crocin production levels of up to 0.022 mg/g DW.

Prospects for crocin production using engineered microbial host

The engineered plant-based crocin biosynthesis provided a basis of prokaryotic chassis for heterologous production. With advancements in synthetic biology, the biosynthesis of crocin from simple carbon sources in fermentation facilities using engineered microorganisms that heterologously express target genes or enzymes has emerged as a promising complement to traditional sources (Wang et al., 2019a, b). Furthermore, the biosynthesis of carotenoids is integrated with metabolic engineering, wherein microorganisms are equipped with a new metabolic pathway that enables them to produce functional compounds from intermediates of their endogenous pathways (Varghese et al., 2023). Recent discoveries in microbial genome data have uncovered new gene pathways, greatly enhancing our understanding of carotenoid production in non-carotenogenic microorganisms (Lee et al., 2024). Currently, crocin production primarily relies on complex extraction and purification from saffron stigmas. This process yields a product with low purity and requires large-scale cultivation of saffron crocus, which can negatively impact the natural environment. The limited availability of crocin has been the primary factor restricting its further application. Therefore, development of more economical and environmentally friendly production methods is highly desirable. Microbial fermentation requires minimal production space and is not dependent on factors such as season or soil composition (Mussagy et al., 2019).

Due to extensive research and development, E. coli has become the best-characterized organism on Earth and one of the most important organisms in industry (Pontrelli et al., 2018). Its rapid growth, simple culture requirements, metabolic flexibility, and vast biochemical and physiological knowledge, along with the numerous tools for genetic and genomic engineering, have made E. coli one of the most proper host for metabolic engineering and synthetic biology. In fact, several studies have explored the production of crocin and its precursor, crocetin, using E. coli. Yeast, particularly Saccharomyces cerevisiae, has long been a preferred model organism for the mass production of recombinant proteins and bio-active metabolites derived from plants. S. cerevisiae is generally regarded as safe and benefits from well-established genetic manipulation tools (Meadows et al., 2016). Its native mevalonate pathway, which supports high-throughput isoprenoid production, including β-carotene, lutein, and zeaxanthin, makes it an ideal host for the heterologous synthesis of plant-derived compounds. Additionally, its robustness in diverse cultivation environments enables efficient large-scale fermentation (Guo et al., 2024). According to the existing studies, many microbes have been successfully transformed to synthesize crocetin or crocins in E. coli and S. cerevisiae (Table 3).

Table 3.

Heterologous crocin and crocetin biosynthesis in microorganisms

Host	Enzymesa	Sources	Production	Yield	References
Escherichia coli	CCD	Crocus sativus (CCD2L)	Crocetin	4.42 mg/L in 100 ml shake-flask for 48 h	Wang et al. (2019a, b)
	ALDH	Neurospora crassa (ALDH8)
	GT	Bacillus subtilis (YjiC, YdhE, YojK)
Escherichia coli	UGT	Gardenia jasminoides (UGT74F8, UGT94E13)	Crocin	60.8 mg/L in in vivo production	Pu et al. (2020)
Escherichia coli	CCD	Crocus sativus (CCD2L)	Crocetin	34.77 mg/L of crocetin and 15.78 mg/L of crocin in 5-L bioreactor for 51 h	Lee et al. (2024)
	ALDH	Synechococcus elongatus (ALDH7942)	Crocin
	UGT	Nicotiana sylyestris (UGT)
		Catharanthus roseus (UGT3)
Escherichia coli	CCD	Gardenia jasminoides (mutant CCD4a, ALDH2C3, UGT94E13)	Crocin	8.43 mg/L and 11.5 mg/L of crocin in shake-flask for 84 h	Li et al. (2024)
	ALDH
	UGT
Saccharomyces cerevisiae	CrtZ	Crocus sativus (CrtZ, CCD2L)	Crocetin	6278 μg/L in 5-L fed-batch bioreactor for 124 h	Chai et al. (2017)
	CCD	Synthetic ALD
	ALD
Saccharomyces cerevisiae	CrtZ	Pantoea stewartii (CrtZ)	Crocetin	12.43 mg/L in 5L-fed-batch bioreactor in 160 h	Song et al. (2020)
	CCD	Crocus sativus (CCD2L)
Saccharomyces cerevisiae	CCD	Crocus sativus (CCD2L, codon-optimized ALDH3)	Crocetin	62.79 μg/g DCW	Tan et al. (2019)
	ALDH
Saccharomyces cerevisiae	CrtZ	Pantoea ananatis (CrtZ)	Crocetin	139.67 μg/g DCW in 250 ml shake-flask in 192 h	Liu et al. (2020a, b)
	CCD	Crocus sativus (CCD2L, ALDH)
	ALDH

^aEnzyme abbreviations are as follows: CCD, carotenoid cleavage dioxygenase; ALD or ALDH, aldehyde dehydrogenase; GT, glycosyltransferase; UGT, UDP-glycosyltransferase; CrtZ, β-carotene hydroxylase; DCW, dried cell weight

Wang et al. (2019a, b) first constructed a heterologous crocetin and crocin biosynthesis pathway in E. coli. The crocin biosynthesis pathway genes, including CsCCD2L, GjUGT94E5, GjUGT75L6, and various ALDHs, were assembled into plasmids and transformed into zeaxanthin-producing strains YL4 and YL5, which were specifically engineered for high carotenoid production. A CRISPR-Cas9 approach was also employed to integrate these genes into the E. coli chromosome for stable expression. By optimizing gene expression using various regulatory elements, crocetin production reached 4.42 mg/L in strain YL4 (pCsCCD2L-GjUGT94E5-GjUGT75L6, pTrc-NcALDH8 from Neurospora crassa). However, it was discovered that GjUGT94E5 and GjUGT75L6 could not glucosylate crocetin in E. coli. Therefore, glycosyltransferase (GTs) from Bacillus subtilis (YjiC, YdhE, and YojK) were introduced. GTs from B. subtilis were led to the production of crocin and crocin-V was confirmed by LC–ESI–MS analysis in strain YL4 (pCsCCD2L-UGT94E5-UGT75L6, pTrc-ALD8, pET28a-YjiC-YdhE-YojK). This study represents the first successful construction of an E. coli cell factory capable of producing crocin-V, marking an important study towards microbial production of saffron apocarotenoids.

Pu et al. (2020) followed this by focusing on maximizing the activity of plant-derived UGTs in E. coli, which enabled the in vivo production of five different crocins using crocetin as a substrate. In this study, two UGT genes (GjUGT74F8 and GjUGT94E13) were inserted into a plasmid and transformed into E. coli, along with overexpressed UDP-glucose biosynthetic genes (pLSB208-pgm-galU) for facilitating crocin biosynthesis. The engineered strain B8pg (harboring pLSB208-pgm-galU and pET32a-UGT74F8) produced 45.0 mg/L of crocin-III and 15.8 mg/L of crocin-V, achieving a crocetin to crocin conversion efficiency of 66.1%. This study highlights the superior specificity and conversion efficiency of plant UGTs compared to microbial UGTs for catalyzing crocetin glycosylation.

Further advancements were made by Lee et al. (2024), who achieved the complete synthesis of crocetin and four crocin derivatives in E. coli. Using the genetically modified MGI strain of E. coli, which was engineered for enhanced biosynthesis of IPP through the overexpression of dxs, dxr, idi, and ispA genes, which was a platform strain in previous research (Moon et al., 2020). Synthetic module expression techniques were employed including the use of specific promoters, terminators, and multiple gene copies to further enhance production. The final strain Z1pCA7942(p)pNC contained CsCCD2L (from C. sativus), ALDH7942 (from Synechocystis elongatus), NsUGT (from Nicotiana sylyestris), and CaUGT3 (from Catharanthus roseus) genes, producing 34.77 mg/L of crocetin, 6.29 mg/L of crocin-V, 5.29 mg/L of crocin-III, 1.48 mg/L of crocin-II, and 2.72 mg/L of crocin-I. These yields were achieved through batch fermentation using glycerol as the primary carbon source, presenting a more sustainable alternative to plant-based extraction.

In a more recent study, Li et al. (2024) focused on improving crocin production by enhancing the enzyme activity of CCD which is the rate-limiting enzyme in crocin biosynthesis. Using a semi-rational design approach combined with site-directed mutagenesis, various truncated versions of GjCCD4a were constructed and tested for catalytic activity. The truncated version N2 (deletion of the 4–99 bp region) exhibited the highest catalytic efficiency. The initial set of mutations was introduced in these residues, such as F86L and S348A, which improved enzyme activity. Iterative rounds of mutagenesis focused on these residues to enhance catalytic efficiency. The mutant GjCCD4a was fused with an MBP (maltose-binding protein) tag to improve the solubility of the protein and the constructed N2-F86L-S348A-mbp (N212m) showed a 25.08-fold increase in catalytic efficiency compared to the wild-type GjCCD4a in the E. coli. Additionally, different gene order within operons (mutant GjCCD4a and GjALDH2C3) significantly influence protein expression and enzyme activity. Similar to Pu et al. (2020), the pSLB208-Eupgm-T7-EugalU and GjUGT74F8 plasmid was introduced into E. coli harboring mutant GjCCD4a to enhance native UDP-glucose biosynthesis and constructed E57 strain. The final engineered strain E57 produced crocin III and V with a total yield of 11.50 mg/L, while E579 (E57 harboring GjUGT94E13) produced five types of crocins with a total of 8.43 mg/L at the shake-flask level.

Chai et al. (2017) successfully established the crocetin biosynthetic pathway in previously engineered S. cerevisiae (SyBE_Sc0014CY06), which was capable of producing β-carotene (Xie et al., 2014). The S. cerevisiae strain was engineered using centromeric plasmid vectors such as pRS416 and pRS425, with the genetic module expression cassettes under galactose-regulated promoters (GAL1, GAL7, and GAL10). The final constructed strain, SyBE_Sc0123C053 (P_GAL1-Ps_CrtZ-T_HIS5, pRS426-T_HIS5-P_GAL10-CCD2L-T_TEF2-P_GAL7- Syn_ALD-T_PGI1), achieved a final crocetin titer of 6,278 μg/L during fed-batch fermentation in a 5-L bioreactor, which was significantly high crocetin titer reported in eukaryotic cells.

Building of this, Song et al. (2020) further enhanced crocetin production in S. cerevisiae through tuning key enzymes coupled with precursor engineering. A systematic optimization was conducted to enhance crocetin overproduction in yeast. By blocking key genes in the acetyl-CoA pathway, such as CIT2 and MLS1, they improved precursor supply. They also fused PsCrtZ from Pantoea stewartii and CsCCD2L using a GGGS linker to enhance enzyme efficiency. As a result, they achieved a significant increase in crocetin production through multiple steps. The final production yield reached 12.43 ± 0.62 mg/L in a 5-L fed-batch bioreactor. The significance of the study lies in its systematic optimization approach, which integrated precursor engineering, enzyme tuning, and process optimization to achieve high-level crocetin production in S. cerevisiae.

Tan et al. (2019) combined transcriptomic analysis with metabolic engineering to provide deeper insights into apo-carotenoid biosynthesis in C. sativus for microbial production of crocetin. They identified a total of 61,202 unigenes, 28 regulators, and 32 putative carotenogenic genes including one aldehyde dehydrogenase (CsALDH3) that oxidizes crocetin dialdehyde into crocetin. Additionally, 15 candidate genes were predicted to be closely associated with safranal and crocin production with one aldehyde dehydrogenase (CsALDH3). That was validated to oxidize crocetin dialdehyde into crocetin, leading to the creation of a crocetin-producing yeast strain. When CsALDH3 was overexpressed in S. cerevisiae, the engineered strain produced 62.79 μg/g dry cell weight (DCW) of crocetin, demonstrating the effectiveness of combining transcriptomic data with metabolic engineering for crocetin biosynthesis.

In another study, Liu et al. (2020a, b) constructed the stable and temperature-responsive engineered S. cerevisiae strain using CRISPR-Cas9 for crocetin biosynthesis. By integrating the crocetin biosynthesis genes into the yeast chromosome and employing a temperature-dependent GAL4M9 mutant, crocetin production was optimized at 24 °C. The strain construction involved knocking out GAL4 and GAL80, replacing them with GAL4M9. Subsequently, integrating the crocetin biosynthesis pathway genes directly into the yeast chromosome to create a stable, multi-copy system. More than two chromosomal copies of CCD2L and ALDH resulted in high level of crocetin production compared to one-copy reference strain. The final strain (TL009), which contained three chromosomal copies of CCD2L and ALDH, achieved a crocetin production level of 139.67 ± 2.24 µg/g DCW, with a zeaxanthin conversion yield of 77% under the optimal temperature conditions. Through these various engineering approaches, S. cerevisiae has been established as a promising microbial cell factory for crocetin biosynthesis. With ongoing advancements, The crocin biosynthesis in yeast may soon be achieved. Future study can be predicted to achieve the crocin biosynthesis to introduce additional genes into this crocetin-producing S. cerevisiae.

Pharmacological properties of crocin

Crocin is gaining attention as potential therapeutic agents in the treatment and prevention of various diseases. Many of these compounds have been investigated in experimental studies and clinical trials, showing promising results. The pharmacological properties of crocin that have been identified are summarized in Table 4.

Table 4.

Pharmacological effects of crocin

Disease	Research designs	Dose	Functions	References
Tumor/cancer	Crocin’s apoptosis activity in MCF-7 cells	up to 200 μg/mL	– Apoptosis induced by upregulating pro-apoptotic protein Bax	Lu et al. (2015)
			– Downregulating anti-apoptotic protein Bcl-2
Vascular dementia	Cognitive impairment in a rat model of chronic cerebral hypoperfusion	25 mg/kg	– Improvement of cognitive and memory impairment	Hosseinzadeh et al. (2012)
Atherosclerosis	Quail fed a hyperlipidemic diet	25, 50, 100 mg/kg/day	– Reduction of oxidized-LDL level—Inhibition of smooth muscle cell proliferation	He et al. (2005)
Obesity/type 2 diabetes	Experimental obesity models and type 2 diabetes models	150 mg/kg	– Proven anti-obesity effect	Hazman et al. (2016)
			– Reducing fasting blood sugar, serum cholesterol, triglycerides, and LDL levels
Type 2 diabetes	Evaluation of the hypoglycemic effect of crocin in a rat model of streptozotocin-induced type 2 diabetes	50, 100 mg/kg	– A significant reduction in serum glucose levels and advanced glycation end products	Shirali et al. (2013)
			– Improved lipid levels in diabetic rats
			– Improved insulin sensitivity
Alzheimer’s disease	Investigation of crocin’s effects in rats with AD	40 mg/kg	– Improved cognitive function	Lin et al. (2019)
			– Reduced neuronal cell death
Alzheimer’s disease	Rats with AD induced by trimethyltin chloride	25, 50 mg/kg	– Reduction of Aβ40 (amyloid-beta peptide associated with AD)	Sadoughi (2019)
			– Increase of anti-apoptotic marker Bcl-2
Anti-inflammatory effects	Confirmation of anti-inflammatory ability in inflammation-induced microglial cells	5–40 μM	– Reducing the levels of inflammatory cytokines (TNF-α, IL-1β) and NO production	Nam et al. (2010)
			– Inhibition of the nuclear factor-kappa B pathway and mitogen-activated protein kinases

Non-provitamin A carotenoids are considered a safer alternative due to their reduced toxicity, while still providing protective benefits. Their anticancer mechanisms include modulating carcinogen metabolism, promoting cell differentiation, inhibiting uncontrolled cell growth, modulating the immune response, offering antioxidant protection, inducing apoptosis, and improving intercellular communication (Hoshyar and Mollaei, 2017). These effects have drawn increasing research attention to crocin, a major component of saffron stigmas, which has demonstrated significant antitumor effects in various animal and cell models. This demonstrates potential role of crocin in cancer prevention and treatment (Khorasanchi et al., 2018).

Since dysregulation or inhibition of apoptosis is a hallmark of cancer development inducing apoptosis is considered one of the most effective anticancer strategies. In a study, crocin was found to induce apoptosis in MCF-7 cells, a human breast cancer cell line (Lu et al., 2015). Crocin activates caspases, particularly caspase-8, which plays a critical role in initiating apoptosis in cancer cells. This activation leads to the release of cytochrome c from mitochondria, a key step in the intrinsic apoptotic pathway (Hosseini et al., 2022). Furthermore, crocin upregulates pro-apoptotic proteins such as Bax and downregulates anti-apoptotic proteins like Bcl-2, tipping the balance towards cell death. These findings suggest that crocin holds promise as a potential preventive and therapeutic agent for breast cancer (Lu et al., 2015).

Vascular dementia is widely recognized as one of the most prevalent forms of dementia after AD. Research indicated that its prevalence sharply increased after the age of 65 (Murray et al., 2007). Individuals with vascular dementia often exhibit cognitive decline triggered by cerebrovascular conditions, including widespread cerebral arteriosclerosis and infarctions (He et al., 2008). Studies have shown that vascular dementia is associated with ischemic, hypoxic, or hemorrhagic damage in brain regions critical for cognitive function and memory (Chui et al., 2000; DeCarli, 2004). Increased free radical production, along with a reduction in antioxidant defenses, can lead to neuronal damage, and a deficiency in antioxidants may contribute to the onset of vascular dementia.

According to Ochiai et al. (2004), crocin was found to be more effective than α-tocopherol in reducing lipid peroxidation within cell membranes and restoring intracellular superoxide dismutase activity in PC-12 cells at the same concentration. In a study by Hosseinzadeh et al. (2012), the effects of saffron extract and crocin on vascular cognitive impairment were investigated using a rat model of chronic cerebral hypoperfusion. This model, which mimics the brain damage and behavioral deficits similar to vascular dementia, served as the foundation for the research (Guang and Du, 2006). A Morris water maze experiment was used to assess whether crocin improved cognitive impairment in the rat model. The results showed a significant improvement in escape latency time, which decreased from 24.64 s in the control group to 8.77 s in the crocin-treated group (25 mg/kg). Additionally, the crocin-treated group reduced the distance traveled to reach the platform from 772 cm in the control group to 251 cm. The percentage of time spent in the target quadrant also increased from 24.16% in the control group to 34.25% in the crocin group (25 mg/kg) (Hosseinzadeh et al., 2012). These findings suggest that crocin enhances spatial cognitive function after chronic cerebral hypoperfusion, likely due to its antioxidant properties.

Atherosclerosis is a condition in which fats, cholesterol, and other substances accumulate in and on the artery walls, forming plaque (Björkegren and Lusis, 2022). This buildup can narrow the arteries and restrict blood flow. If the plaque ruptures, it can lead to blood clots, potentially causing heart attacks, strokes, or other serious cardiovascular issues. According to the study by He et al. (2005), crocin significantly reduced the development of atherosclerotic plaques in quails fed a high-cholesterol diet. Notably, crocin was shown to lower oxidized-LDL cholesterol levels and inhibit the proliferation of smooth muscle cells, both contributing to plaque formation in arteries. Crocin reduces oxidative stress that plays an important role in the development of atherosclerosis, thereby protecting vascular endothelial cells (Vahed et al., 2024). Additionally, crocin helps regulate lipid metabolism and supports the reverse cholesterol transport pathway, whereby excess cholesterol is removed from tissues and transported back to the liver for excretion (He et al., 2005). These combined effects help slow or prevent the progression of atherosclerosis, making crocin a valuable compound for cardiovascular health.

Most individuals with type 2 diabetes tend to excess body fat of obesity, indicating a strong connection between the two conditions (Golay and Ybarra, 2005). This relationship is evident because both obesity and type 2 diabetes are characterized by insulin resistance (Kahn et al., 2006). As adipose tissue increases due to obesity, more adipokines are released, exacerbating insulin resistance in peripheral tissues and the pancreas (Bastard et al., 2000). The accumulation of fat in non-adipose tissues, known as lipotoxicity, leads to oxidative stress in these tissues, further aggravating insulin resistance (Hussain et al., 2010). Over time, insulin resistance causes elevated blood glucose levels (hyperglycemia), and oxidative stress from glucotoxicity and lipotoxicity damages pancreatic β-cells. This β-cell dysfunction, often caused by inflammation associated with excess adipose tissue, contributes to the eventual development of type 2 diabetes.

In a study by Hazman et al. (2016), rodents were induced to develop obesity and type 2 diabetes through a high-fat diet or other diabetogenic agents like streptozotocin. Crocin not only improved glucose and lipid metabolism by significantly reducing fasting blood sugar levels, serum cholesterol, triglycerides, and LDL levels, but also showed anti-obesity effects. Additionally, crocin improved insulin sensitivity through its antioxidant and anti-inflammatory mechanisms (Shirali et al., 2013). This study highlights crocin’s potential as a therapeutic agent for managing obesity and type 2 diabetes. It achieves this by improving glucose and lipid metabolism, reducing oxidative stress, and enhancing insulin sensitivity (Hazman et al., 2016).

AD is a progressive neurodegenerative disorder and the most common cause of dementia, primarily affecting older adults (Castellani et al., 2010). It is characterized by a gradual decline in memory, thinking, reasoning, and behavioral abilities due to the buildup of amyloid plaques and neurofibrillary tangles in the brain (Wang et al., 2019a, b). These pathological changes lead to the death of brain cells, resulting in cognitive decline and memory impairment, which eventually leads to the inability to perform daily functions. While the exact causes of AD are not fully understood, it is believed that a combination of genetic, environmental, and lifestyle factors contribute to its development.

A study investigating the effects of crocin on apoptotic and inflammatory markers, brain-derived neurotrophic factor (BDNF), protein markers, and Aβ40 (beta-amyloid) was conducted using a rat model of AD induced by trimethyltin chloride (TMT) (Sadoughi, 2019). TMT is a neurotoxin known to induce hippocampal neurodegeneration, mimicking AD-like symptoms (Geloso et al., 2011). The results revealed several significant effects of crocin. Crocin reduced cell death by suppressing apoptosis-related markers. Specifically, it decreased the expression of pro-apoptotic markers such as caspase-3 while increasing the anti-apoptotic marker Bcl-2. These results are supported by the fact that, when a similar AD mouse model was used, a significant increase in the number of apoptotic neurons in the hippocampus and prefrontal cortex of the mice was confirmed through TUNEL staining. In addition, a significant increase in neuronal survival in crocin-treated AD mice was confirmed by western blot of apoptosis-related proteins such as Bax, caspase-3, and Bcl-2 (Lin et al., 2019). Crocin also enhanced the production of BDNF, a molecule essential for neuron survival, plasticity, and cognitive function, which is typically diminished in AD. Additionally, crocin treatment decreased Aβ40 levels, a form of amyloid-beta peptide associated with AD, in the hippocampal regions, suggesting a potential role in reducing amyloid plaque formation. Finally, crocin increased neuronal density in the CA1, CA2, and CA3 regions of the hippocampus (Sadoughi, 2019), indicating that crocin may protect neurons from the degenerative effects of AD, thereby preserving cognitive functions associated with these regions.

Inflammation is the body’s natural response to harmful stimuli such as pathogens, damaged cells, toxic compounds, or other irritants. It is a crucial part of the immune system’s defense mechanism to eliminate the initial cause of cell injury, clearing out necrotic cells and tissues, and initiating tissue repair (Kaur et al., 2010). One study investigated how crocin could modulate the inflammatory response in microglial cells to mitigate damage. Microglial cells play a key role in the brain’s immune defense; however, when overactivated, they release neurotoxic molecules that contribute to neurodegenerative diseases like AD (Frank-Cannon et al., 2009). Activated microglial cells promote neuronal injury through the release of proinflammatory factors such as tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-6, nitric oxide (NO), and ROS (Graeber and Streit, 2010). In the study by Nam et al. (2010), microglial cells were cultured and treated with lipopolysaccharide to induce inflammation, a common method to mimic neuroinflammation.

When microglial cells were exposed to crocin, it was shown to reduce the levels of inflammatory cytokines (TNF-α and IL-1β) and NO production, indicating its ability to suppress the activation of pro-inflammatory pathways. Crocin likely inhibits the nuclear factor-kappa B pathway, a key regulator of inflammation, as well as mitogen-activated protein kinases, both of which play significant roles in the inflammatory response of microglial cells (Nam et al., 2010). Furthermore, crocin exhibits antioxidant properties by reducing oxidative stress through the lowering of ROS levels (Assimopoulou et al., 2005). This antioxidant effect contributes to its anti-inflammatory action, as oxidative stress is closely linked to neuroinflammation (Reuter et al., 2010).

Future perspectives

The low water solubility of fat-soluble carotenoids presents a significant challenge in maximizing their bioavailability and functionality in various applications. Traditional fat-soluble carotenoids often face difficulties in being absorbed effectively in the human body, limiting their therapeutic potential. Therefore, extensive research efforts have focused on enhancing the solubility of these compounds. Some strategies such as encapsulation, emulsification, complexation, and glycosylation have been employed to increase the water solubility of carotenoids. These strategies have shown promise in improving the bioaccessibility and stability of carotenoids, leading to enhanced absorption and effectiveness. However, challenges remain about achieving sustainable and economically viable methods of producing water-soluble derivatives.

The increasing demand for water-soluble carotenoids highlights their superior functionality, not only in terms of absorption and bioavailability, but also in their broad potential for commercial applications. Water-soluble carotenoids are attractive compounds due to their diverse pharmacological properties including antioxidant, anti-inflammatory, and anticancer effects. Crocin, a water-soluble carotenoid derived from C. sativus, G. jasminoides, and B. daviddi, offers some advantage in terms of its solubility, making it a more effective candidate for therapeutic and industrial applications compared to its fat-soluble counterparts.

Crocin is biosynthesized through a multi-step process that involves the glycosylation of crocetin. Glycosylation reaction transforms crocetin into crocin, enabling it to become water-soluble well. This biosynthetic pathway has been well-characterized in plants such as C. sativus, G. jasminoides, and B. daviddi, and key enzymes involved in the pathway have been identified, facilitating further research into optimizing crocin production.

Despite the success in characterizing crocin biosynthesis in plants, the large-scale production of crocin from natural sources like saffron is impractical due to high costs and limited availability. To address these limitations, heterologous production systems using engineered microorganisms and plants have been explored as a more sustainable and economical alternative. Advances in metabolic engineering and genetic engineering have enabled the introduction of crocin biosynthetic pathways into heterologous hosts, such as Nicotiana sp., E. coli, and S. cerevisiae. These engineered systems have shown promise in producing crocin at higher yields and lower costs, offering a commercially viable approach to crocin production. However, challenges remain in optimizing the heterologous production of crocin to achieve industrial demands. Current microbial production systems are still limited in yield and scalability, and further research is needed to enhance the efficiency of these systems. Additionally, the bioactivity of heterologously produced crocin must be thoroughly evaluated to ensure that it matches the efficacy of naturally derived crocin.

The pharmacological properties of crocin further emphasize its potential as a valuable compound in therapeutic applications. Crocin has been shown to exhibit potent antioxidant properties, reducing oxidative stress and protecting cells from damage. It also demonstrates anti-inflammatory and anticancer effects, making it a promising candidate for the treatment of various diseases including cancer, cardiovascular conditions, and neurodegenerative disorders. The ability of crocin modulate key biological pathways including apoptosis and inflammation, underscores its therapeutic potential.

In conclusion, while significant progress has been made in improving the water solubility of carotenoids and developing sustainable production methods for crocin, ongoing research is essential to overcome existing limitations. The continued development of heterologous production systems, along with a comprehensive understanding of crocin’s pharmacological properties, will be able to apply of crocin in bioindustries successfully.

Author contributions

Y.L. and C.Y.H.: Writing original draft, visualization, and investigation. E.-S.C.: Reviewing and editing manuscript. M.-J.S.: Conceptualization, supervision, funding acquisition, reviewing and editing manuscript.

Funding

This work was supported by Incheon National University Research Grant in 2022.

Declarations

Conflict of interest

The authors declare that there is no conflict of interest.

Ethical approval

Not applicable.

Consent for publication

All authors read and approved the final manuscript and potential publication.