Natural novel vitamin C derivative, 2-O-β-D-glucopyranosyl-L-ascorbic acid: Resources, biosynthesis, and applications

Abstract

As the only naturally occurring stable derivative of L-ascorbic acid (AA; vitamin C), 2-O-β-D-glucopyranosyl-L-ascorbic acid (AA-2βG) is hydrolyzed in vivo to release active AA. AA-2βG exhibits strong antioxidant and antiphotoaging effects comparable to those of AA, and it plays a key role in maintaining organismal health. Owing to its superior stability and bioavailability, AA-2βG is considered as a promising, longer-lasting natural alternative to conventional vitamin C. It was first identified and is particularly abundant in Lycii Fructus (Gouqizi in Chinese) but has been detected in several crop plants. This review offers a comprehensive overview of recent advances in AA-2βG research, covering key aspects including discovery, structure, natural sources, extraction and detection methods, chemical and in vitro enzymatic synthesis, biosynthetic pathways, as well as applications in health care, skin care, and functional foods. Additionally, we highlight strategies for leveraging plant resources and enhancing AA-2βG biosynthesis, which are expected to accelerate future research and support the sustainable development and utilization of AA-2βG and other high-value natural products.

1. Introduction

L-Ascorbic acid (AA), commonly known as vitamin C [(2R)-2-[(1S)-1,2-dihydroxyethyl]-3,4-dihydroxy-2H-furan-5-one (C₆H₈O₆)] (Fig. 1), plays a crucial role in numerous physiological processes essential for sustaining human life. However, the human body cannot synthesize AA and must obtain it from external sources (Phadke et al., 2022). Despite its importance, AA’s chemically unstable structure makes storage difficult (Paciolla et al., 2019). Therefore, exploring stable derivatives of AA is critical. 2-O-β-D-glucopyranosyl-L-ascorbic acid (AA-2βG) is a naturally occurring derivative of AA derivative with enhanced chemical stability and resistance to oxidative degradation. After ingestion, AA-2βG undergoes hydrolysis to releases AA, showing antioxidant activity while also displaying pharmacological properties distinct from its precursor. Studies have confirmed its antioxidant activity (Starzak et al., 2020), free radical scavenging ability (Chen et al., 2021), melanin formation inhibition (Zhang, Li, Liao, Wang, & Li, 2007), and antihyperglycemic effects (Ma et al., 2017). Nevertheless, further research is required to clarify the mechanisms underlying AA-2βG absorption and digestion (Dong et al., 2024).

Fig. 1.

Chemical structures of L-ascorbic acid and its derivatives. (a) L-Ascorbic acid (AA), (b) 2-O-α-D-Glucopyranosyl-L-ascorbic acid (AA-2αG) and (c) 2-O-β-D-Glucopyranosyl-L-ascorbic acid (AA-2βG).

Since its discovery, studies have examined AA-2βG’s natural resources, chemical synthesis, and biosynthesis. Initially identified in Lycium barbarum L., AA-2βG has subsequently been detected in other plant species, particularly in Lycium and Rubus species. To date, AA-2βG has not achieved industrial-scale production, its biosynthetic pathway remains incompletely understood. Its potential applications extend across health care, skin care, and functional foods. This review provides a comprehensive assessment of the current state of AA-2βG research, emphasizing its chemical characteristics, natural sources, extraction methods, chemical synthesis, biosynthetic pathways, and applications. Additionally, it considers future research directions for AA-2βG, particularly those supporting the development and use of natural products. The primary objectives are to propose promising research avenues and to offer a reference for accelerating AA-2βG development and commercialization.

2. L-Ascorbic acid and its glycosylated derivatives

2.1. AA and 2-O-α-D-glucopyranosyl-L-Ascorbic acid

AA is an essential antioxidant that supports key human physiological functions. These include boosting immune defense, promoting collagen synthesis, protecting against ultraviolet (UV)-induced skin damage, facilitating iron absorption to prevent anemia, and reducing osteoporosis risk (Aghajanian et al., 2015, Boo, 2022, Carr and Maggini, 2017; Gęgotek et al., 2019; Teucher, Olivares, & Cori, 2004). Severe AA deficiency leads to scurvy (Lu, Guo, Sun, Chen, & Liu, 2023). Humans cannot synthesize AA because they lack L-gulonolactone oxidase, the enzyme catalyzing the final step of AA biosynthesis. Therefore, AA must acquired from dietary sources (Phadke et al., 2022). However, the enediol structure of AA is highly unstable, making it prone to oxidation when exposed to light, heat, or oxygen. This instability presents a major challenge for AA preservation (Paciolla et al., 2019). Hence, stabilizing the derivatives of AA is necessary to improve its usability.

Structural modification of the AA molecule is a primary strategy to enhance stability. Researchers have synthesized diverse AA derivatives, categorized as lipophilic AA derivatives, glucoside AA derivatives, butenolide derivatives, and conjugates of vitamin C and E (Chen et al., 2021, Choi et al., 2024, Lee and Boo, 2022, Meščić Macan et al., 2019, Fossa Shirata and Maia Campos, 2021). Among these, glucoside AA derivatives are particularly important. By chemically or biologically modifying the hydroxyl groups at AA positions C-2, C-3, C-5, or C-6, derivatives such as 2-O-α-D-glucopyranosyl-L-ascorbic acid (AA-2αG) gain improved stability and bioavailability (Jaber, Jaber, Hamed, & AlKhatib, 2022). However, early AA derivatives, including 5-O-α-D-glucopyranosyl-L-ascorbic acid and 6-O-α-D-glucopyranosyl-L-ascorbic acid had limited practical use due to insufficient stability and inability to release bioactive AA in vivo (Suzuki, 1973, Yamamoto and Muto, 1992, Yamamoto et al., 1992). To date, only two naturally occurring AA derivatives have been identified: 6-O-β-D-glucopyranosyl-L-ascorbic acid (AA-6βG) in Cucurbitaceae and AA-2βG in Lycium barbarum L. and Lycium chinense Mill. (Hancock et al., 2008, Toyoda-Ono et al., 2004). These findings emphasize the need for targeted structural modifications to optimize AA derivatives for functional applications. Balancing structural stability with effective hydrolysis in vivo is key for enhancing their practical utility across diverse applications.

AA-2αG, one of the earliest AA derivatives discovered and applied, shows robust stability and can be hydrolyzed in the digestive tract by α-glucosidase to release bioactive AA, making it an effective supplement (Yamamoto, Muto, Nagata, Nakamura, & Suzuki, 1990). It was first obtained in 1990 through enzymatic glucosylation of AA using rat intestinal and rice α-glucosidases (Yamamoto, Muto, Nagata, Nakamura, & Suzuki, 1990). AA-2αG has a D-glucose moiety attached to the C2 position of AA via an α-1,4-glucoside linkage (Fig. 1) (Mandai, Yoneyama, Sakai, Muto, & Yamamoto, 1992). This structure masks AA’s enediol group, considerably improving its stability and resistance to oxidative degradation (Melo-Guímaro et al., 2022). Furthermore, AA-2αG can be produced cost-effectively on a large scale through enzymatic synthesis, often using low-cost substrates, such as maltose (Shen et al., 2023, Zheng et al., 2023). These features indicate that it is a practical and economically viable AA derivative.

2.2. Discovery of AA-2βG

AA-2βG, chemically known as (R)-5-((S)-1, 2-dihydroxyethyl)-4-(λ1-oxidaneyl)-3-(((2S, 3R, 4S, 5S, 6R)-3, 4, 5-trihydroxy-6-(hydroxymethyl) tetrahydro-2H-pyran-2-yl) oxy) furan-2(5H)-one (C₁₂H₁₈O₁₁) (Fig. 1), is a natural, stable AA derivative first isolated from Lycii Fructus in 2004 by Japanese Suntory Ltd. (Toyoda-Ono et al., 2004). Lycii Fructus are well known for their dual use as food and medicine, especially for their reputed life-extending and antiaging properties. Although antioxidant activity is usually attributed to AA, the low AA content in Lycii Fructus suggests that other bioactive compounds, including AA-2βG, also contribute (Qian, Zhao, Yang, & Huang, 2017). Studies using high-performance liquid chromatography (HPLC)/liquid chromatography–mass spectrometry (LC–MS) and other techniques have quantified AA-2βG in Lycii Fructus at 0.5%–1.4%, further supporting its role in the antioxidant profile of Lycii Fructus (Jaros-Sajda, Budzisz, & Erkiert-Polguj, 2024).

AA-2βG and AA-2αG are isomers that differ only in their glucosyl group configuration: AA-2βG has a β-glucosyl linkage at the C2 position of AA, whereas AA-2αG carries an α-glucosyl group (Fig. 1). This subtle structural difference gives AA-2βG unique biological advantages. Unlike AA-2αG, which is rapidly hydrolyzed by widely distributed α-glucosidases, AA-2βG undergoes slower hydrolysis via β-glucosidase, an enzyme less common in animals, leading to longer systemic retention and reduced gastrointestinal irritation (Toyada-Ono et al., 2005). Although AA-2βG exhibited antioxidant activity comparable to that of AA and AA-2αG in the cell-free 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical scavenging assay (Takebayashi et al., 2008; Zhang et al., 2011), a study using RAW264.7 cells demonstrated its enhanced free radical scavenging activity relative to these two compounds. Mechanistically, this activity is associated with the activation of the Kelch-like ECH-associated protein-1/nuclear factor E2-related factor 2 (Keap1/Nrf2) signaling pathway. Specifically, the β-configuration of AA-2βG (not the α-configuration) promotes Nrf2-DNA binding and favors the activation of antioxidant genes (Wang et al., 2019). Additionally, AA-2βG reduces chemotherapy-induced oxidative damage (Wang et al., 2019). Combined with its natural origin and improved stability, these properties make AA-2βG a promising candidate for pharmacological and nutraceutical use.

3. Natural sources of AA-2βG

3.1. AA-2βG in genus Lycium

AA-2βG was first isolated from mature L. barbarum and L. chinense fruit, with comparable levels observed in these species (Toyoda-Ono et al., 2004). Subsequent studies examined L. barbarum fruit from multiple regions, including Ningxia, Qinghai, Gansu, Hebei, and Inner Mongolia in China, as well as Switzerland and Serbia. Reported AA-2βG content is 5–14 mg/g dry weight (DW) (0.5%–1.4%), except in Serbian L. barbarum, which exhibited much lower levels (Bubloz et al., 2020, Zhou et al., 2021; Ilić et al., 2020; Kosińska-Cagnazzo et al., 2017, Zhou et al., 2021, Tai and Gohda, 2007, Toyoda-Ono et al., 2004, Zhong et al., 2022, Zhu et al., 2022) (Table 1). A hybrid of L. barbarum and L. chinense growing spontaneously in Saxon, Wallis (Switzerland), contained higher AA-2βG content (about 2.4 mg/g DW) compared with local L. barbarum fruits (Kosińska-Cagnazzo et al., 2017). In contrast, no AA-2βG detected in ripe and dried Lycium ruthenicum Murray fruit (black goji berry) (Ilić et al., 2020). Beyond fruit, AA-2βG was also found in the stems and leaves of L. barbarum, with the highest level recorded in the rhizome (about 0.13 mg/g DW) (Bubloz et al., 2020) (Table 1). These findings indicate that AA-2βG occurs in various plant parts, offering insights for extraction and use (Fig. 2).

Table 1.

AA-2βG content across species.

Family	Genus	Species	Organ	Location	AA-2βG content (mg/g)	References
Solanaceae	Lycium	L. barbarum	Fruit	Ningxia, China	*5.500 0–13.300 0	Zhou et al., 2021, Zhou et al., 2021, Tai and Gohda, 2007, Zhong et al., 2022, Zhu et al., 2022
				Qinghai, China	*4.500 0–13.200 0	Zhou et al., 2021, Zhou et al., 2021, Zhu et al., 2022
				Gansu, China	*10.300 0–11.900 0	Zhou et al., 2021, Zhu et al., 2022
				Xinjiang, China	*5.700 0–14.100 0	Zhong et al., 2022, Zhu et al., 2022
				Inner Mongolia, China	*5.000 0	Toyoda-Ono et al., 2004
				Switzerland	*0.400 0–0.700 0	Kosińska-Cagnazzo et al., 2017
				Serbia	*0.500 0–0.600 0	Ilić et al., 2020
			Leave	Switzerland	*0.030 0 ± 0.003 0	Bubloz et al., 2020
			Stem	Switzerland	*0.040 0 ± 0.000 5	Bubloz et al., 2020
			Rhizome	Switzerland	*0.130 0 ± 0.230 0	Bubloz et al., 2020
		L. chinense	Fruit	Hebei, China	*5.000 0	Toyoda-Ono et al., 2004
				Switzerland	*2.800 0 ± 0.100 0	Kosińska-Cagnazzo et al., 2017
		Hybrid of L. barbarum and L. chinense	Fruit	Switzerland	*2.400 0 ± 0.020 0	Kosińska-Cagnazzo et al., 2017
	Physalis	Physalis peruviana L.	Fruit	Colombia	0.099 4 ± 0.006 0	Carole, Isabelle, & Wilfried, 2020
			Calyx	Colombia	0.086 3 ± 0.002 3	Carole, Isabelle, & Wilfried, 2020
	Capsicum	Capsicum annuum L.	Fruit	Vietnam	0.074 5–0.195 0	Carole, Isabelle, & Wilfried, 2020
	Solanum	Solanum xanthocarpum L.	Fruit	Thailand	0.032 7 ± 0.005 5	Carole, Isabelle, & Wilfried, 2020
		Solanum torvum Sw.	Fruit	Thailand	0.167 0 ± 0.027 0	Carole, Isabelle, & Wilfried, 2020
		Solanum lycopersicum var. cerasiforme	Fruit	Italy	0.048 2 ± 0.008 4	Carole, Isabelle, & Wilfried, 2020
Rosaceae	Malus	Malus sylvestris Mill.	Fruit	New Zealand	0.130 0–0.920 0	Richardson et al., 2021
			Leave	New Zealand	0.021 0–0.071 0	Richardson et al., 2021
		Malus x domestica Rehder	Fruit	New Zealand	0.000 0–0.075 0	Richardson et al., 2021
			Leave	New Zealand	0.025 0–0.220 0	Richardson et al., 2021
		Malus baccata (L.) Borkh.	Fruit	New Zealand	0.800 0	Richardson et al., 2021
			Leave	New Zealand	0.060 0	Richardson et al., 2021
	Prunus	Prunus armeniaca L.	Fruit	New Zealand	0.000 6–0.020 0	Richardson et al., 2021
			Leave	New Zealand	0.011 0–0.027 0	Richardson et al., 2021
		Prunus sp.	Fruit	New Zealand	0.001 0	Richardson et al., 2021
			Leave	New Zealand	0.015 0	Richardson et al., 2021
		Prunus persica (L.) Batsch	Fruit	New Zealand	0.001 0–0.001 3	Richardson et al., 2021
			Leave	New Zealand	0.008 0–0.012 0	Richardson et al., 2021
		Prunus cerasus L.	Fruit	New Zealand	0.003 0	Richardson et al., 2021
			Leave	New Zealand	0.008 2	Richardson et al., 2021
		Prunus avium (L.) L.	Fruit	New Zealand	0.001 7	Richardson et al., 2021
			Leave	New Zealand	0.007 7	Richardson et al., 2021
	Rubus	Rubus idaeus L.	Fruit	New Zealand	0.000 1–0.000 9	Richardson et al., 2021
			Leave	New Zealand	0.005 2–0.022 0	Richardson et al., 2021
	Pyrus	Pyrus sp.	Fruit	New Zealand	0.000 8	Richardson et al., 2021
			Leave	New Zealand	0.009 1	Richardson et al., 2021
	Rosa	Rosa rubiginosa L.	Fruit	New Zealand	0.004 9	Richardson et al., 2021
			Leave	New Zealand	0.036 0	Richardson et al., 2021
Ericaceae	Vaccinium	Vaccinium sp.	Fruit	New Zealand	0.000 6–0.001 2	Richardson et al., 2021
			Leave	New Zealand	0.039 0–0.041 0	Richardson et al., 2021
Actinidiaceae	Actinidia	Actinidia arguta (Siebold & Zucc.) Planch. ex Miq.	Fruit	New Zealand	0.002 3–0.002 7	Richardson et al., 2021
			Leave	New Zealand	0.003 4–0.006 1	Richardson et al., 2021

Note: * represents dry weight.

Fig. 2.

Morphological characteristics of Lycium barbarum. (a) Lycium barbarum plants. (b) Leaves, unmatured fruits and mature fruits. (c) Flowers and buds. (d) Dry fruits, known as Lycii Fructus (Gouqizi in Chinese).

Globally, Lycium includes approximately 97 species and six varieties (Miguel, 2022). Some Lycium species, such as L. arabicum from North Africa, the Algerian traditional medicine L. intricatum, and L. europaeum, also exhibit strong antioxidant activity (Affes et al., 2017, Bendjedou et al., 2023, Mejri et al., 2023). This activity may be related to AA-2βG content, although direct confirmation is required. Research has largely focused on L. barbarum, reflecting its pharmaceutical relevance, market demand, and pharmacological importance. L. barbarum is listed as the medicinal herb Lycii Fructus in the Chinese Pharmacopoeia (Chinese Pharmacopoeia Commission, 2020). Lycii Fructus derived from L. barbarum and L. chinense are marketed worldwide for their nutritional value and pharmacological properties (Du et al., 2024), with effects reported historically and in modern research (Chan et al., 2019, Kang et al., 2017, Starzak et al., 2020). These findings highlight the neglect of most Lycium species and the need to explore understudied taxa.

3.2. AA-2βG in other species

Similar to Lycii Fructus, some plant species exhibit strong antioxidant activity despite low endogenous AA content, suggesting the presence of stabilized AA derivatives. AA-2βG has also been detected in plants beyond the Lycium genus (Fig. 3). In 2020, it was first identified via nontargeted LC–MS metabolomics in Malus sylvestris (crab apple) at 13–92 mg/100 g in fresh fruit and 2.1–7.1 mg/100 g in fresh leaves (Richardson et al., 2020). Subsequent studies confirmed AA-2βG in other species at lower concentrations: Prunus armeniaca (apricot, Rosaceae), Actinidia arguta (kiwifruit, Actinidiaceae), and Vaccinium sp. (blueberries, Ericaceae) (Richardson et al., 2021). Traditionally used by the Māori of New Zealand, the leaves of pu̅ha̅ (Sonchus spp.), kawakawa (Macropiper excelsum), and poroporo (Solanum aviculare/laciniatum), as well as immature poroporo fruit and pikopiko fern shoots (Asplenium bulbiferum), also contain AA-2βG, although specific AA-2βG content in these plants was not reported. These results suggest that AA-2βG may be widespread in plants (Table 1).

Fig. 3.

Natural resources of AA-2βG.

For comparison, L. barbarum fruit contain 5–10 mg/g DW AA-2βG, equivalent to 100–200 mg/100 g fresh weight (FW) (assuming a 5:1 fresh-to-dry weight ratio), far higher than levels in crab apples, apricots, kiwifruits, and blueberries. This marked difference highlights L. barbarum as a uniquely rich natural source of AA-2βG.

4. Extraction and determination of AA-2βG

AA-2βG can be obtained via three main approaches: natural extraction, chemical synthesis, and in vitro enzymatic synthesis (Toyoda-Ono et al., 2004; Toyada-Ono et al., 2005; Wojciechowska, Klewicki, Sójka, & Grzelak-Błaszczyk, 2018). Among these, extraction from L. barbarum fruit is the most well-established method owing to the compound’s high stability and abundance in the plant. Although not yet industrialized, laboratory-scale protocols have achieved extraction yields of > 9 mg/g DW (Zhou et al., 2021, Zhou et al., 2021, Yang et al., 2014, Zhong et al., 2022, Zhu et al., 2022).

Most AA-2βG extraction methods use water as a solvent owing to AA-2βG’s polarity and solubility, making water cost-effectiveness and efficient solvent. (Zhou et al., 2021, Zhou et al., 2021, Toyoda-Ono et al., 2004, Yang et al., 2014, Zhong et al., 2022). However, some protocols supplement water with cosolvents at various concentrations to improve extraction efficiency and reduce impurities, e.g., 75% ethanol solution (Zhu, Zhang, Qin, Zhao, & Li, 2022). For extraction optimization, ultrasonic-assisted extraction for 15–30 min enhances efficiency compared with traditional prolonged soaking (Toyoda-Ono et al., 2004). Additionally, heating with cosolvents at > 80°C has demonstrated AA-2βG’s thermal stability across a broad temperature range (Yang, Dong, & Yin, 2014).

For purification, chromatographic separation is widely applied to isolate AA-2βG from crude extracts. Furthermore, AA-2βG can be purified via adsorption using specific adsorbent materials, e.g., wood fibers effectively adsorb AA-2βG, with AA-2βG content in these fibers reaching 15.68%, and strong alkaline anion exchange resins absorb up to 14.07% AA-2βG (Zhou, Yin, & Tang, 2018). Overall, natural extraction remains the most practical method for obtaining AA-2βG, although future advancements in in vitro enzymatic or chemical synthesis may enable scalable alternatives.

As mentioned, AA-2βG is highly soluble in water and typically detectable via HPLC or ultra-HPLC. Owing to its high polarity, AA-2βG is difficult to separate on standard C₁₈ columns, necessitating columns with stronger polar fixation (Zhou et al., 2021, Zhou et al., 2021, Yang et al., 2014, Zhong et al., 2022, Zhu et al., 2022). The choice of mobile phase requires the establishment of appropriate concentration gradients for separation and detection, depending on the column properties and test solution solvent. AA-2βG exhibits UV absorption, and detection is commonly performed with a UV detector at 260 nm. Gradient elution is generally favored for complex matrices, although isocratic elution may be sufficient for simpler samples. For example, in 2021, the Ningxia Chemical Analysis and Testing Association created a group standard using a C₁₈ column (250 mm × 4.6 mm) with an isocratic elution consisting of 20% methanol, 20 mmol/L phosphoric acid, and 5 mmol/L tetrabutylammonium bromide (20:40:40 volume percentage). Low-concentration samples, such as plant leaves, require optimized gradients or advanced detectors, e.g., tandem mass spectrometry. For example, liquid chromatography–high-resolution accurate mass spectrometry has been applied to measure low AA-2βG levels in blueberry and kiwifruit leaves (about 0.5 mg/100 g FW) (Richardson et al., 2021). Thus, method optimization is required depending on sample type and available instrumentation.

5. Synthesis of AA-2βG

5.1. Chemical synthesis of AA-2βG

AA-2βG can also be synthesized chemically, although large-scale industrial production has not yet been achieved. The main challenges is the low reactivity of the C2 hydroxyl group in AA, which forms intramolecular hydrogen bonds with the carbonyl group. To data, three chemical synthesis methods have been reported for AA-2βG, relying on protective group strategies and suitable glycosyl donors (Gazivoda et al., 2005, Gazivoda et al., 2006, Gazivoda et al., 2007 (Fig. 4). Following its discovery in 2004, AA-2βG was chemically synthesized to confirm its structure using 5,6-O-isopropylidene-L-ascorbic acid (1) as the starting material and 2,3,4,6-tetra-O-acetyl-β-D-glucopyranosyl 2,2,2-trichloroacetate (3) as the glycosyl donor. Subsequent removal of the protecting groups produced 2-O-(2,3,4,6-tetra-O-acetyl-β-D-glucopyranosyl)-L-ascorbic acid (5), and further deprotection yielded the target compound AA-2βG. This four-step reaction, involving two chromatographic purifications, achieved only a 13% yield, restricting its use to small-scale laboratory synthesis rather than mass production (Toyoda-Ono et al., 2004). In 2017, researchers improved the method using the commercially available compound (1) as the raw material and acetobromo-α-D-glucose (6) as the glycosyl donor, which reacted to generate compound (5). The reaction subsequently produced AA-2βG, and by modifying reagents and conditions, the overall yield increased to 53% (Ma et al., 2017). In 2023, researchers described a method more suitable for industrial-scale synthesis of AA-2βG (He et al., 2023). In this process, AA is used as a raw material to obtain compound (1), and the 3-position hydroxyl group of this compound is protected with tert-butyldiphenylsilane (TBDPS). Using compound (6) as the glycosyl donor, one-step deprotection of the TBDPS and isopropylidene protecting groups is then performed to generate compound (5). The approach yielded AA-2βG at 31.4% over five steps and required only one column chromatographic purification, reducing purification demands compared with earlier methods. For industrial production of AA-2βG, feasibility may depend more on reducing purification steps and costs than on maximizing reaction yield.

Fig. 4.

Chemical synthesis of AA-2βG. (1) 5,6-O-Isopropyridene-L-ascorbic acid. (2) 3-O-Benzyl-5,6-O-isopropylidene-L-ascorbic acid. (3) 2,3,4,6-Tetra-O-acetyl-β-D-glucopyranosyl 2,2,2-trichloroacetate (4) 2-O-(2,3,4,6-Tetra-O-acetyl-β-D-glucopyranosyl)-3-O-benzyl-5,6-O-isopropylidene-L-ascorbic acid. (5) 2-O-(2,3,4,6-Tetra-O-acetyl-β-D-glucopyranosyl)-L-ascorbic acid. (6) Acetobromo-α-D-glucose. (7) 3-O-Tert-butyldiphenylsilyl-5,6-O-isopropylidene-L-ascorbic acid. (8) 2-O-(2,3,4,6-Tetra-O-acetyl-β-D-glucopyranosyl)-3-O-tert-butyldiphenylsilyl-5,6-O-isopropylidene-L-ascorbic acid. Δ: heating at 120–130 °C without a solvent; TBDPS: tert-butyldiphenylsilyl; DMSO: dimethyl sulfoxide; TBDPSCl: tert-butyldiphenylchlorosilane; DIPEA: N,N-diisopropylethylamine; THF: tetrahydrofuran; DMAP: dimethylaminopyridine; DCM: dichloromethane; TFA: trifluoroacetic acid; AA-2βG: 2-O-β-D-glucopyranosyl-L-ascorbic acid.

5.2. In vitro enzymatic synthesis of AA-2βG

In vitro enzymatic synthesis of AA-2βG has been explored using cellulases from fungi, such as Aspergillus niger, Trichoderma viride, and Trichoderma reesei. Cellulases form a complex enzyme system, with most members belonging to the glycoside hydrolase family, which can exploit their transglycosylation activity to synthesize glycosylated derivatives (Wang & Yu, 2022). In a 2005 pioneering study, AA-2βG was synthesized using AA as the substrate, cellobiose as the glycosyl donor, and cellulases from several different fungal species as catalysts (Toyada-Ono et al., 2005), with the cellulase derived from T. viride exhibiting the highest catalytic efficiency. However, the reaction not only yielded low amounts of AA-2βG but also produced large quantities of AA-6βG. Although in vitro enzymatic synthesis represents a more environmentally friendly alternative to chemical synthesis, its scalability remains constrained by low yields and byproduct formation.

5.3. Biosynthetic pathway of AA-2βG

Although the biosynthetic pathway of AA is well established, the enzyme catalyzing AA glycosylation has not been identified (Smirnoff & Wheeler, 2024) (Fig. 5). Four identified AA biosynthetic pathways have been described in plants: the D-Galacturonate pathway, the D-mannose/L-galactose (D-Man/L-Gal) pathway, the L-Gulonate pathway, the Myo-inositol pathway (Gilbert et al., 2009, Li et al., 2019, Tyapkina et al., 2019, Zhang et al., 2016) (Fig. 5). Among these, the D-Man/L-Gal pathway is predominant in plants. In Lycii Fructus, AA biosynthesis mainly occurs through D-Man/L-Gal pathway, in which galactose dehydrogenase (GalDH) and L-galactose-1,4-lactone dehydrogenase (GLDH) are the key enzymes. Particularly, GLDH catalyzes the final step of the L-galactose pathway, converting L-galactono-1,4-lactone into AA, whereas GalDH catalyzes the conversion of L-galactose into L-galactono-1,4-lactone (Yin et al., 2024). Phosphomannose mutase (PMM) and GDP-D-mannose epimerase (GME) are key regulatory genes in the L-Gal pathway, crucial roles in AA biosynthesis. Dehydroascorbic acid reductase (DHAR) is recognized as the main positive regulatory gene of total AA content in the ascorbate–glutathione cycle (Yin et al., 2024). Despite these genes being validated in other plant species, their functions in Lycii Fructus are not supported by in vivo or in vitro evidence (Fenech et al., 2019, Wheeler et al., 1998). Several enzymes and genes regulate AA levels but none directly control AA-2βG.

Fig. 5.

Biosynthetic pathway of AA-2βG. ME: Methylesterase; GMP: GDP-D-mannose pyrophosphorylase; GME: GDP-D-mannose 3′5′ epimerase; GGP: GDP-L-galactose-phosphorylase; GPP: L-galactose-1-P phosphatase; GLDH: L-galactose dehydrogenase; GLDH: L-galactono-1,4-lactone dehydrogenase; PDE: phosphodiesterases; UGD: UDP-Glucose dehydrogenase; UGE: UDP-glucose epimerase; GKE: Glucuronic acid kinase; GRE: Glucuronate reductase; MIOX: Myo-inositol oxygenase. AA-2βG: 2-O-β-D-glucopyranosyl-L-ascorbic acid.

Early attempts to detect glycosyltransferase activity in Lycii Fructus extracts showed no detectable activity. In Arabidopsis thaliana, overexpression of UGT87A2 increased AA-2βG levels by 2–5 fold, suggesting that AA-2βG is produced by UGT87A2 and may be upregulated under environmental stress (Paul & von, 2010). However, this study did not include in vitro functional validation of UGT87A2. A metabolome–transcriptome study later identified ten UDP-glycosyltransferases and three β-glucosidases that potentially participate in catalyzing AA-2βG formation (Huang et al., 2025). Whether AA-2βG biosynthesis occurs through direct glycosylation of AA or via intermediate steps remains unresolved.

6. Health benefits and applications of AA-2βG

6.1. Applications of AA-2βG in health care

AA-2βG exhibits multiple therapeutic properties, including ocular protection, anticancer activity, diabetes management, and anti-inflammatory effects (Fig. 6). Notably, it more strongly augments antioxidant capacity in oxidative-damaged lenses compared with AA, helping preserve lens transparency (Huang, Zhang, Zhang, Zhou, & Niu, 2012), and it selectively induces apoptosis in cervical cancer (HeLa) cells by stabilizing p53 while downregulating prosurvival proteins (Zhang et al., 2011). Additionally, AA-2βG inhibits α-glucosidase, highlighting its potential for glycemic control (Ma et al., 2017), and, alleviates dextran sulfate sodium-induced inflammatory bowel disease (IBD). It also promotes short-chain fatty acids production and regulates gut microbiota composition (Huang et al., 2019).

Fig. 6.

Applications of AA-2βG. IBD: inflammatory bowel disease.

AA-2βG supports health maintenance through regulatory mechanisms (Fig. 6). It restructures the intestinal microbiota and strengthens tight junctions, reducing systemic inflammation and neuroinflammatory responses (Dong et al., 2020). Its also prevents high-fat refined diet–induced cognitive deficits, and its superior free radical scavenging activity and ability to maintain glutathione homeostasis further reinforce systemic benefits. Notably, the free radical scavenging activity of AA-2βG is stronger than that of AA and AA-2αG. It downregulates Keap1 expression while upregulating Nrf2 and heme oxygenase-1 (HO-1) expression. Moreover, its β-configuration enhances Nrf2–DNA binding affinity. In mouse macrophage RAW264.7 cells, AA-2βG protects against hydrogen peroxide–induced cell death, reduces oxidative stress, and more effectively maintains the ratio of cellular glutathione (GSH) to oxidized glutathione (GSSG). The structural features of AA-2βG, particularly its glycosylation likely improves stability and bioactivity, facilitating interaction with cellular signaling pathways linked to oxidative stress (Wang et al., 2019).

Among the signaling pathways associated with AA-2βG’s antioxidant properties, the insulin/insulin-like growth factor-1 signaling (IIS) pathway is central, with the transcription factors DAF-16, HSF-1, and SIR-2.1 being key regulators. AA-2βG also modulates gut microbiota, which in turn influences systemic oxidative stress and inflammatory responses (Fang et al., 2025). As a potential functional food ingredient, AA-2βG may reduce neuroinflammation associated with Western-style diets (Dong et al., 2024; Wang et al., 2019). Its antioxidant effects arise from a multifaceted network involving IIS pathway regulation, gut microbiota modulation, and enhanced structural stability that supports redox reactions. Given these mechanisms, AA-2βG has promise as a natural health care product for reducing oxidative stress and promoting general health.

6.2. Applications of AA-2βG in skin care

AA is widely used in skin care products for reducing wrinkles, improving elasticity, preventing melasma and photoaging (Boo, 2022, Correia and Magina, 2023). However, AA’s instability and rapid degradation limit its effectiveness in skin care formulations. To overcome this, stable AA derivatives, such as magnesium ascorbyl phosphate (MAP), have been developed for cosmetics (Caritá et al., 2020). Other AA derivatives, including AA-2αG and 3-O-ethyl-L-ascorbic acid (i.e., vitamin C ethy) are applied in skin care for collagen stimulation and UV protection (Boo, 2022, Chen et al., 2021). Regarding tyrosinase activity inhibition, AA-2βG is twofold more effective than AA, reducing melanin synthesis and hyperpigmentation (Zhang, Li, Liao, Wang, & Li, 2007). AA-2βG’s stability and natural origin position it as a promising antimelasma and antioxidant agent in formulations, although no commercial products currently use it directly (Fig. 6).

Commercially, Lycii Fructus extracts are already incorporated in cosmetics. The International Nomenclature Cosmetic Ingredient and the Inventory of Existing Cosmetic Ingredients in China published by the Chinese National Medical Products Administration list L. chinense fruit water, L. chinense fruit extract, etc. (Yue, Li, Zhang, Jiang, & Chen, 2025). Lycium extracts contain multiple various nutrients, and their skin care effects likely result from the combined action of various components. Importantly, they are considered mild and safe. AA derivatives are already established in cosmetics; for example, AA-2αG was approved as an active ingredient in quasi-drugs in the 1990 s (Maeda, 2022). However, as AA-2βG has not yet been extracted and standardized for cosmetic use, safety and chemical specifications have not been established. Although no current skin care products contain AA-2βG, its natural origin aligns with the “clean beauty” trend. Combined with strong efficacy data, it represents a promising candidate for antiphotoaging, antioxidant, and whitening products. Future research should aim to optimize AA-2βG extraction or synthesis methods and establish safety evaluations to accelerate its commercialization in cosmetics.

6.3. Applications of AA-2βG in functional food

AA-2βG’s thermal stability supports its use as a natural antioxidant in functional foods (Fig. 6). In Lycii Fructus–enriched extrudates, it retains activity under high-temperature processing, enhancing antioxidant capacity by 20-fold (Kosińska-Cagnazzo et al., 2017, Ménabréaz et al., 2021). Incorporating Lycii Fructus into beer at different brewing stages can improve flavor and antioxidant levels. Consumers favor beer with Lycii Fructus introduced early in brewing, as this preserves bioactivity, sensory quality, and key compounds, such as AA-2βG (Ducruet et al., 2017).

AA-2βG demonstrates multifaceted applications, with stability, bioactivity, and natural origin positioning it as a superior alternative to AA. Thus further commercialization efforts are warranted across sectors. Nevertheless, its readiness for market use in skin care and food is limited by challenges in industrial-scale production, safety validation, and sustainability. Ongoing research on bioprocess optimization, safety testing, and scalable manufacturing will be essential to translate laboratory findings into commercial applications.

7. Conclusion and perspectives

AA-2βG, a unique plant-derived AA derivative, combines the antioxidant activity of AA with markedly greater stability, making it a promising candidate for future research and applications. Although AA-2βG is widely distributed across 14 genera in seven plant families, its abundance varies substantially. Lycium barbarum fruit contain the highest reported levels of AA-2βG (up to 14% DW), establishing them as the primary source for extraction. Within Lycium, only L. chinense, L. barbarum, and L. ruthenicum have been analyzed for AA-2βG content, leaving the majority of the genus unexplored, despite reported antioxidant potential. Beyond Lycium, AA-2βG occurs at trace levels (<80 mg/100 g FW) in crab apple, kiwifruit, and blueberries, highlighting their roles as supplementary natural resources. Although AA-2βG is currently extracted from L. barbarum, its broad phylogenetic distribution emphasizes opportunities for diversifying production through biotechnological or agro-engineering approaches.

The species-specific accumulation of AA-2βG in L. barbarum and L. chinense positions it as potential biomarker for quality control in Lycii Fructus. Current pharmacopoeial standards, those in the Chinese Pharmacopoeia, require quantification of polysaccharides and betaine. However, betaine is widely distributed across plants and lacks species specificity compared with AA-2βG. Similarly, although the Korean Pharmacopoeia sets a betaine content for Lycii Fructus (≥ 0.5% DW) consistent with the Chinese standard, and the Japanese Pharmacopoeia does not stipulate any content requirement, none include AA-2βG. Given its pharmacological activity and species specificity, AA-2βG should be evaluated as a more precise indicator of Lycii Fructus quality (Dobrijević et al., 2023).

Despite advances, industrial-scale production of AA-2βG remains challenging. In contrast, substantial progress has been made in the industrialization of AA-2αG. In Australia, safety and chemical standards for AA-2αG have been established, including classification as a high-purity compound (> 98%). The major catalytic enzymes employed to enhance AA-2αG production include sucrose phosphorylase (SPase), cyclodextrin glucosyltransferase (CGTase), and α-glucosidase (AGase) (Shen et al., 2025). Under optimized conditions, yields of AA-2αG can reach 113 g/L (Gan et al., 2023). Ongoing research is demonstrating that enzymatic catalysis and genetic engineering are key to improving production efficiency and product stability (Gan et al., 2023, Gudiminchi et al., 2016, Liu et al., 2013, Shen et al., 2025, Zhang et al., 2024, Zhou et al., 2022). In contrast to AA-2αG, the industrial-scale production of AA-2βG faces several bottlenecks limiting its efficient and cost-effective manufacture: natural extraction is hindered by low native content, poor extraction efficiency, and complex purification; chemical synthesis is constrained by lengthy purification and high costs; and in vitro enzymatic synthesis suffers from limited scalability due to low yields and undesirable byproducts. Among these approaches, biosynthesis represents the most promising strategy for scalable production (Jiang et al., 2025, Liu et al., 2024, Xu et al., 2024b, Zhu et al., 2025).

Elucidating the complete biosynthetic pathway of AA-2βG is a central research prioritiey. Its defining feature, i.e., the β-glucosyl linkage at the C2 position of AA, likely results from catalysis by UDP-glycosyltransferases (UGTs) or β-glucosidases (BGLUs) (Huang et al., 2025), although this step remains unconfirmed. Expanding genomic resources for Lycium species, including L. barbarum, L. ferocissimum, and L. ruthenicum, now enable genome-wide screening and functional characterization of the candidate UGTs and BGLUs involved in AA-2βG biosynthesis (Cao et al., 2021, Shah et al., 2022, Xu et al., 2024a). Moreover, integrating metabolomic and transcriptomic datasets across tissues, developmental stages, and genotypes with varying AA-2βG content will help identify and validate the enzymes responsible for biosynthesis. Additionally, artificial intelligence (AI)-driven approaches represent a state-of-the-art strategy. The use of deep learning to train predictive models can accelerate the prediction of the key enzyme in AA-2βG biosynthesis, thereby improving its biosynthetic efficiency (Reinhardt et al., 2025, Zheng et al., 2022). Once this pathway is elucidated, optimizing biosynthesis in vivo or via in vitro fermentation will become feasible, accelerating industrial-scale production of AA-2βG and advancing its commercialization. Prior studies on the regulatory mechanisms of AA and the development of gene editing systems in L. barbarum provide a foundation for increasing AA-2βG accumulation (Ai et al., 2023, Feng et al., 2015, Zhang et al., 2020). Future efforts should prioritize molecular breeding and metabolic engineering strategies in Lycii Fructus plants, leveraging genomic resources and CRISPR-based technologies to enhance AA-2βG production.

In conclusion, AA-2βG stands out as a naturally derived AA derivative with superior stability, consumer trust, and strong potential for application in health care, skin care, and functional foods. Achieving a full understanding of its biosynthetic pathways will provide a basis for sustainable production and commercialization.

CRediT authorship contribution statement

Mengyue Wang: Conceptualization, Writing – review & editing, Visualization. Haotian Wu: Writing – Original Draft, Visualization. Li Xiang: Conceptualization, Writing – review & editing, Funding acquisition. Ranran Gao: Supervision, Writing – review & editing. Qinggang Yin: Writing – review & editing. Yang Chu: Writing – review & editing. Lan Wu: Writing – review & editing. Yanyan Su: Supervision, Writing - review & editing. Gangqiang Dong: Conceptualization, Writing – review & editing. Yuhua Shi: Conceptualization, Writing – review & editing. Li Xiang: Conceptualization, Writing – review & editing, Funding acquisition.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.