Lactobacillus paracasei fermentation enhances the aroma profile and antidiabetic efficacy of goji berry juice

Abstract

Goji berry (Lycium barbarum L.) is rich in bioactive compounds, and its functional efficacy can be further enhanced through fermentation. This study evaluated the impact of Lactobacillus paracasei-fermentation on the quality attributes of goji berry juice and its in vivo antihyperglycemic efficacy. Fermentation significantly increased total flavonoids (+31.8%) and polysaccharides (+5.4%), enhanced antioxidant activity, and enriched the volatile aroma profile relative to unfermented juice. In a type II diabetic mouse model, four weeks of gavage with fermented goji berry juice alleviated weight loss and polydipsia, reduced fasting blood glucose by 33%, improved glucose tolerance, and corrected dyslipidemia. Histopathological examination revealed partial restoration of liver, kidney, and pancreatic integrity, accompanied by reduced malondialdehyde and elevated superoxide dismutase activity. Moreover, the intervention modulated gut microbiota by increasing the abundance of Akkermansia and Bacteroidetes while suppressing Desulfovibrionaceae, alongside elevated cecal propionic acid and butyric acid. Correlation analyses further revealed that the fermentation-enhanced antioxidants and gut microbiota-derived short-chain fatty acids were significantly correlated with the improvement in key diabetic phenotypes, suggesting a potential mechanism mediated by the microbiota-metabolite axis. These findings provide a scientific basis for developing fermented goji berry products as functional foods or adjuvant therapies for diabetes management.

Introduction

Goji berry (Lycium barbarum L.), a perennial solanaceous shrub widely cultivated in China and parts of Europe, has gained international recognition as a “superfruit” owing to its broad applications in functional foods and herbal medicine^1,2. Goji berry is rich in many natural bioactive components, such as polysaccharides, polyphenols, carotenoids and amino acids³, and it has a variety of pharmacological effects, including hypoglycemia, anti-aging, lowering blood pressure, and so on⁴.

Diabetes mellitus (DM) is a metabolic disorder characterized by impaired insulin secretion or action. Type II DM is the most prevalent form, accounting for approximately 90% of cases worldwide. Uncontrolled DM increases the risk of developing kidney, cardiovascular, and liver diseases, among others, which can have a significant impact on human health and quality of life⁵. Various influences such as genetics, age, unhealthy lifestyle, and obesity are considered risk factors for type II DM⁶. More recently, it is widely acknowledged that dysbiosis of the gut microbiota has emerged as a pivotal contributor to type II DM pathogenesis. Therefore, exploring natural dietary interventions that can modulate gut microbiota while simultaneously improving glucose metabolism is of great significance.

Growing consumer demand for health-promoting, non-alcoholic beverages is driving interest in lactic acid-bacterial (LAB) fermentation, a process that enhances both functionality and sensory quality through microbial biotransformation⁷. Numerous studies have demonstrated that LAB-fermented fruit juices exert antidiabetic effects. For instance, Frediansyah et al. (2021) found that fermentation with Lactobacillus plantarum FNCC 0027 could improve the functional activity of Jamaican cherry juice and enhance the inhibition of diabetes-related enzymes⁸. Barrón-Álvarez et al. (2022) compared the antioxidant activity and effect on blood glucose of fresh Cucurbita ficifolia juice, naturally fermented pumpkin juice, and fermented pumpkin juice by L. plantarum⁹. They observed that both fermented juices retained the functional properties of antioxidant, α-glucosidase inhibition, and hypoglycemic, highlighting the potential of fermentation in functional beverages. Wang et al. (2022) investigated the effects of Lactobacillus fermented apple juice on glucose and lipid metabolism in diabetic mice, finding that it improved the fasting blood glucose and insulin levels of diabetic mice to a certain extent, and reshaped the gut microbiota¹⁰. These findings collectively indicated that fermented fruit and vegetable juices have a regulating effect on blood glucose.

Despite these advances, the integrative effects of L. paracasei-fermented goji berry juice, on both quality attributes and antidiabetic efficacy remain insufficiently elucidated, particularly with regard to the interplay between compositional enhancement, oxidative stress regulation, and gut microbiota-derived short-chain fatty acids (SCFAs). The starter culture L. paracasei 5572 was chosen for its strong probiotic profile, antioxidant capacity, and good performance in plant-based fermentations⁷. Based on these attributes, the present study was designed to systematically evaluate the impact of fermentation on the compositional profile, aroma characteristics, and in vivo antidiabetic activity of goji berry juice. In addition, correlation analysis was further performed to reveal the mechanistic links between juice components, antioxidant activity, gut microbiota, SCFAs, and key diabetic phenotypes. This work provides new insights into the development of FGJ as a functional food and potential adjuvant therapy for type II DM.

Results and discussion

Effect of L. paracasei-fermentation on nutritional components

The physicochemical properties and nutritional compositions of unfermented or fermented goji berry juice (FGJ) are summarized in Table 1. In comparison with unfermented juice, FGJ exhibited a significant increase in lactic acid content (P < 0.05), indicating that L. paracasei efficiently metabolized sugars into lactic acid, thereby demonstrating strong fermentation capacity. The soluble solid content was significantly decreased (P < 0.01), while the polysaccharide content was significantly increased (P < 0.05) after fermentation. This phenomenon may be explained by the fact that, during the early stages of fermentation, lactic acid bacteria hydrolyze sucrose into fructose and glucose at a rate that exceeds sugar consumption¹¹. The soluble protein content remained consistently below 1 mg/mL, which could be attributed to microbial metabolism promoting the binding of soluble proteins with cellulose, lipids, and other components to form water-insoluble substances, and the increased acidity of the juice may also lead to the hydrolysis of soluble proteins into peptides and amino acids¹². Moreover, the total flavonoid content was significantly elevated (P < 0.05) after fermentation, while the total phenolic content remained unchanged.

Table 1.

Physicochemical properties and nutritional compositions of FGJ

Physicochemical properties/nutritional compositions	UFGJ	FGJ
pH value	4.63	4.33
Viable count (CFU/mL)	0	2 × 108
Lactic acid content (g/L)	7.05 ± 0.26	7.61 ± 0.06*
Total soluble solid content (°Brix)	13.2 ± 0.14	12.8 ± 0.18**
Total phenolic content (mg/mL)	1.59 ± 0.01	1.68 ± 0.04
Total flavonoid content (mg/mL)	0.22 ± 0.02	0.29 ± 0.01*
Polysaccharide content (mg/mL)	38.34 ± 0.34	40.41 ± 0.22*
Soluble protein content (mg/mL)	0.351 ± 0.04	0.325 ± 0.01

Asterisks indicate significant differences.

UFGJ unfermented goji berry juice, FGJ L. paracasei-fermented goji berry juice.

*:P ≤ 0.05; **P ≤ 0.01.

Effect of L. paracasei-fermentation on antioxidant activity

DPPH, ABTS and hydroxyl radical scavenging activities were measured to evaluate the in vitro antioxidant capacity of FGJ. As shown in Fig. 1A, the DPPH scavenging activity of FGJ increased significantly (P < 0.05) when diluted 15- to 40-fold. Notably, a 15-fold dilution achieved a scavenging rate of 91.22%, marking a significant 4.35% increase compared with the unfermented juice (P < 0.01). Similarly, the ABTS scavenging activity of goji juice improved after fermentation when diluted 20- to 70-fold, with a 20-fold dilution reaching 85.05%, representing a 1.4-fold increase over the unfermented juice (Fig. 1B). In addition, the hydroxyl radical scavenging capacity of the FGJ was significantly higher than that of the unfermented juice at all tested dilution factors (P < 0.05) (Fig. 1C).

Fig. 1. Antioxidant activity of unfermented or fermented goji berry juice.

A DPPH free radical scavenging rate, B ABTS free radical scavenging rate, C Hydroxyl radical scavenging rate. Asterisks indicate significant differences (*P ≤ 0.05, **P ≤ 0.01).

These findings demonstrate that L. paracasei-fermentation significantly enhances the antioxidant capacity of goji berry juice. This improvement may be attributed to fermentation-induced increases in the content of polysaccharides and flavonoids, as well as changes in the water solubility of other bioactive compounds. Additionally, synergistic interactions among inherent compounds such as phenolics and flavonoids may further contribute to the enhanced antioxidant activity^13,14. Future studies should focus on systematically characterizing changes in specific active components and mapping the polyphenol-flavonoid profile to elucidate the exact mechanism of antioxidant activity enhancement.

Changes in volatile aroma compounds

Volatile components of goji berry juice, both before and after fermentation, were analyzed using GC-IMS. A total of 37 volatile compounds were identified (Table S1), primarily including esters, aldehydes, ketones, alcohols, pyrazines, and acids. To visualize the impact of fermentation on these volatile constituents more intuitively, the Gallery Plot plugin was employed to generate an ion mobility spectrometry (IMS) fingerprint (Fig. 2). In this fingerprint, each column represents a single volatile compound, and increasingly intense red coloration signifies higher relative abundance.

Fig. 2.

Fingerprints of volatile compounds in unfermented and fermented goji berry juice.

After fermentation by L. paracasei, the levels of 22 volatile compounds decreased. Among them, 2-methylbutanoic acid, which produces a pungent, spicy, goat cheese-like odor at higher concentrations but a pleasant fruity aroma at lower concentrations¹⁵, was reduced. In addition, elevated aldehyde levels have been demonstrated to contribute to off-flavor formation; notably, in the present study, 2,4-heptadienal and β-cyclohomocitral exhibited a marked reduction. Although earlier research has reported that the high fermentation activity of certain strains may elevate 2,3-butanediol levels¹⁶, the L. paracasei strain used in this study actually lowered its concentration.

In contrast, 15 volatile compounds showed a significant increase in FGJ. For instance, 3-hydroxybutan-2-one and 2,3-pentanedione impart pleasant creamy, caramel-like aromas, while various ester compounds, such as pentyl acetate, ethyl pentanoate, ethyl formate, and diethyl malonate, contribute fruity and floral notes^16,17. The content of these esters rose significantly following fermentation. Previous studies reported that certain esters and ketones possess intrinsic antimicrobial or antioxidant activities^18–20. Although the concentrations in FGJ are typically low and may limit direct in vivo functions, their potential synergistic interactions with primary active substances (e.g., polysaccharides and flavonoids) warrant consideration. Furthermore, the antimicrobial properties of these compounds could potentially influence the structure of the gut microbiota.

Overall, L. paracasei fermentation reduced undesirable volatile compounds while enhancing fruity and floral aromas, thereby improving the overall volatile profile. These enhancements in flavor and composition not only improved the sensory quality of the product but also provided a potential biochemical basis for the subsequent antidiabetic effects observed.

Effect of FGJ intervention on the apparent state of mice

Mice in the control group maintained glossy fur, exhibited alert behavior, and a healthy mental state (Fig. S1). In contrast, model mice appeared thinner, with sparse, greasy, and yellowish fur, reduced responsiveness, and lethargy. Sharp abdominal contractions were also observed during STZ administration

Water intake in the control group remained stable throughout the experiment, whereas experimental groups exhibited a marked increase beginning in week 8 (P < 0.05). During the intervention period, water intake rose significantly and then stabilized, followed by a slight decrease, indicating that the treatments alleviated polydipsia (Fig. S2A). Similarly, food intake increased after successful modeling and gradually declined following intervention.

Body weight in control mice gradually increased and then stabilized over time. In contrast, all experimental groups experienced continuous weight gain during 1–4 weeks, likely due to high-fat dietary intake, and remained consistently heavier than the control group (Fig. S2B). From week 8 onward, gavage administration moderated the weight loss trend, particularly in the Metformin, HFGJ, and LP groups. Overall, FGJ improved the external condition of diabetic mice, indicating mitigation of type II diabetes-associated symptoms.

Blood glucose indicators in mice

Fasting blood glucose (FBG) is the most commonly used and important indicator for diagnosing disorders of glucose metabolism¹⁰. As shown in Fig. 3A, prior to STZ injection, the FBG in each group was approximately 5 mmol/L, indicating normal levels. By the 8th week, FBG in all experimental groups exceeded 11.1 mmol/L and was significantly higher than that of the control group (P < 0.05), confirming the successful establishment of a type II diabetes model. During the gavage period (weeks 9–12), FBG in the model group remained consistently elevated at approximately 22.5 mmol/L, whereas the metformin group showed a general decline in FBG before reaching a stable phase. In the FGJ groups, FBG increased slowly and then decreased significantly compared with the model group. Notably, FBG declined from 21.11 to 14.16 mmol/L in the HFGJ group and from 19.24 to 16.57 mmol/L in the LFGJ group, suggesting a dose-dependent hypoglycemic effect. The LP and UFGJ groups also showed downward trends, though less pronounced than FGJ (Fig. 3A).

Fig. 3. Detection of blood glucose-related indicators.

A Changes in fasting blood glucose (FBG), B glucose tolerance in mice, C changes in area under the curve (AUC) of blood glucose.

As a standard procedure in diabetes diagnosis, the Oral Glucose Tolerance Test (OGTT) evaluates glycemic control by measuring the body’s response to a glucose load, thereby providing a comprehensive assessment of pancreatic β-cell function and overall glucose homeostasis²¹. As shown in Fig. 3B, after glucose challenge (2 g/kg), blood glucose rose sharply in all experimental groups relative to the control (P < 0.05), peaking at 60 min. At 120 min, none of the intervention groups returned to baseline, indicating that the glucose tolerance ability was significantly reduced, among which, the HFGJ group could restore a certain level of glucose.

The area under the curve (AUC) of blood glucose serves as a crucial indicator for precision glucose management, reflecting overall glucose tolerance (Fig. 3C). The model group exhibited the highest AUC, while the metformin, HFGJ, and LFGJ groups showed significantly reduced values compared with the model group (P < 0.05). Changes in OGTT and AUC values demonstrate that FGJ significantly improved the glucose tolerance of diabetic mice. The observed improvement in glycemic control further suggests that FGJ may exert its effects through metabolic regulation and tissue protection.

Organ index analysis

Changes in organ indices are presented in Table S2. No significant differences were observed in the heart, kidney, or lung indices among the groups (P > 0.05). The liver, which serves as the primary organ for glucose metabolism and plays a key role in regulating blood sugar, exhibited a significantly elevated index in both the model and treatment groups (P < 0.05), indicating pronounced hepatic swelling and impaired function associated with diabetes. However, oral administration of FGJ effectively mitigated these effects compared to the model group.

Similarly, the spleen, the largest immune organ and an essential center for both cellular and humoral immunity²², showed significant enlargement (P < 0.05) in all six experimental groups relative to the control, suggesting diabetes-induced immune dysfunction. In addition, the pancreatic index declined markedly in the model and treatment groups. Collectively, these results demonstrate that FGJ partially alleviated diabetes-associated increases in liver and spleen indices while attenuating the decline in pancreatic index.

Effect of FGJ on serum lipid levels

The liver is the primary organ for glucose and lipid metabolism; reduced hepatic insulin sensitivity leads to aberrant glucose and lipid metabolism and impairs liver function²³. Furthermore, abnormal glucose metabolism can trigger dysregulated lipid metabolism, thereby exacerbating insulin resistance. As shown in Fig. 4A, compared with the control group, triglyceride (TG) and low-density lipoprotein cholesterol (LDL-C) levels in the model group were significantly elevated (P < 0.05), whereas high-density lipoprotein cholesterol (HDL-C) content was markedly reduced. Relative to the model group, TG levels exhibited a statistically significant decline in all treatment groups. Additionally, total cholesterol (TC) decreased in all treatment groups, with significantly lower values observed in the HFGJ and LFGJ groups. Metformin treatment produced the highest HDL-C levels, while LDL-C was significantly lower in the HFGJ, LP and UFGJ groups. Collectively, these findings suggest that FGJ effectively mitigates dyslipidemia in diabetic mice. The observed reduction in FBG and improvement in glucose tolerance were accompanied by favorable changes in organ indices and lipid metabolism, suggesting a systemic interplay between glycemic regulation, hepatic function, and lipid homeostasis.

Fig. 4. Effect of different treatments on serum lipid levels and oxidative stress in the liver.

A serum lipid levels, B oxidative stress in the liver.

Results of oxidative stress indices in the liver

Elevated levels of reactive oxygen species (ROS), lipid peroxidation, and malondialdehyde (MDA) are established markers of oxidative stress²⁴. As shown in Fig. 4B, in the diabetic model mice, MDA levels were significantly higher than those in the control group (P < 0.05). Following intervention, particularly in the Metformin and HFGJ groups, MDA levels were markedly reduced compared with the model group (P < 0.05). Superoxide dismutase (SOD) and catalase (CAT) are typically upregulated in response to oxidative stress, and their dysregulation has been implicated in metabolic disorders. Conversely, deficiencies in these enzymes, especially CAT, are associated with an increased risk of obesity, diabetes, and fatty liver²⁵. Glutathione (GSH), a well-known antioxidant that safeguards critical macromolecules from oxidative damage, is often diminished under oxidative stress because of heightened oxidation²⁶.

SOD levels were significantly lower in the model group than in controls (P < 0.05). After intervention, SOD activity increased across all treatment groups, with the most pronounced improvements observed in the HFGJ group. In contrast, the UFGJ group exhibited a significant decrease in CAT levels, while only the LFGJ group showed a marked reduction in GSH-Px activity. These findings suggest that FGJ mitigates oxidative stress to varying extents. The amelioration of oxidative stress by FGJ is concomitant with its enhancement of glycemic control, pointing to a potential mechanistic connection. This coordinated improvement suggests that the reduction of oxidative burden supports glycemic regulation, which in turn further alleviates oxidative stress, thereby protecting pancreatic β-cells and establishing a positive feedback loop for metabolic homeostasis.

Histopathological section observation results

The HE staining results of the liver, kidney, and pancreatic tissue sections are shown in Fig. 5. Long-term consumption of high-fat, high-sugar diets is known to impair liver function, resulting in hepatic injury. In the control group, hepatic architecture remained intact, displaying normal morphology, regularly arranged cells, and uniform distribution, without evident signs of degeneration or necrosis. By contrast, the model group exhibited extensive vacuolar necrosis, hepatocellular swelling, and bright areas containing spherical lipid droplets. After intervention, varying degrees of hepatic improvement were observed. Notably, both the metformin and HFGJ groups showed reduced vacuolar necrosis, suggesting that FGJ exerts a protective and reparative effect on liver tissues in diabetic mice.

Fig. 5.

Histological sections of liver, kidney, and pancreas tissues from mice following different treatments.

Glomerular endothelial cells are essential for maintaining glomerular filtration²⁷. The model group displayed significant renal damage, including glomeruli with obscure boundaries, epithelial cell necrosis, and vacuolation. After treatment, glomerular morphology improved, with size and structure approaching normal, particularly in the metformin and HFGJ groups. In the pancreas, the control group exhibited intact islets and glandular follicular cells, which appeared neatly organized and round or oval in shape. In contrast, the model group exhibited slightly atrophied islets, marked cytoplasmic vacuolar degeneration, and inflammatory cell infiltration around the islets. Following gavage, all treatment groups demonstrated reductions in both inflammatory cells and cytoplasmic vacuolization, although some inflammatory cells remained localized to one side of the islets. Overall, these findings suggest that FGJ attenuates diabetes-induced damage in the liver, kidneys, and pancreas in a dose-dependent manner. The histological improvements are consistent with reductions in oxidative stress and enhanced glycemic control, further supporting the role of FGJ in protecting key metabolic organs.

Effect of FGJ on intestinal short-chain fatty acids

Maintaining interstitial‑fluid pH within its physiological range is critical for preventing insulin‑resistant diabetes²⁸. Probiotic fermentation of fruit- and vegetable-derived juices generates abundant organic acids that act in a prebiotic‑like manner, acidifying the intestinal lumen and thereby suppressing pathogenic microorganisms^29,30. To evaluate this effect, the short‑chain fatty acids (SCFAs) in mouse cecal contents were quantified (Fig. 6A).

Fig. 6. Effect of different treatments on intestinal short-chain fatty acids (SCFAs) and gut microbiota.

A changes in short-chain fatty acids, B OTU petal diagram of gut microbes, C relative abundance at the Phylum level, D relative abundance at the Family level, E relative abundance at the Genus level.

The levels of acetic acid, propionic acid, butyric acid, and valeric acid were all decreased to varying extents in diabetic mice. Compared with the model group, metformin intervention significantly increased the levels of acetic acid, propionic acid, and valeric acid (P < 0.05). Acetate has been reported to interact with receptors in the small intestine and insulin-sensitive tissues, such as the liver, adipose tissue, and skeletal muscle, to improve body weight regulation and insulin secretion³¹. Notably, acetic acid concentrations rose in the LP and UFGJ groups, with the UFGJ intervention showing the most pronounced effect (P < 0.05). Propionate has been shown to inhibit gluconeogenesis in hepatocytes, reduce FBG and improve glucose tolerance in mice³², while butyrate helps prevent diabetes either by acting on its primary receptor or by inhibiting histone deacetylases (HDACs)⁵. Accordingly, the HFGJ group exhibited a marked 22.24% increase in propionic acid content. Likewise, butyric acid levels rose by 7.16% and 7.24% in the HFGJ and UFGJ groups, respectively. In contrast, isobutyric acid and isovaleric acid, as branched-chain SCFAs associated with reduced insulin sensitivity, showed no significant changes relative to the control group.

Overall, our findings suggest that FGJ effectively increases the levels of certain SCFAs in diabetic mice, consistent with previous reports on its role in blood glucose regulation³³. These metabolic changes provide a mechanistic bridge to the subsequent alterations in microbial abundance, since SCFAs not only result from but also shape the ecological structure of the gut microbiota.

Sequencing results of the mouse gut microbiota

OTU‑level analysis of the intestinal microbiota

As shown in Fig. 6B, operational taxonomic unit (OTU) clustering revealed clear differences in microbial richness among the seven groups. The control group displayed the greatest number of OTUs, indicating the highest α‑diversity, whereas 226 OTUs constituted a core set shared by all groups. Compared with the diabetic model, administration of FGJ altered OTU counts, suggesting that this intervention modulates the intestinal microbial community structure.

Composition of the gut microbiota at the phylum, family, and genus level

Figure 6C shows the community composition of gut microbiota at the phylum level. The intestinal community was dominated by Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, Verrucomicrobia, Tenericutes, TM7, and Deferribacteres. In the control group, Firmicutes accounted for 59.78% of total sequences but declined significantly in the diabetic model group (P < 0.05). All interventions partially restored Firmicutes levels, although the proportion remained lower in the HFGJ group. Bacteroidetes represented 36.97% in controls and increased significantly in the HFGJ group, rising by 30.57% versus the model (P < 0.05). Proteobacteria comprised only 0.98% in controls yet expanded to 12.23% in the model group (P < 0.05). Metformin, HFGJ, and UFGJ reduced this expansion to 7.96%, 8.79%, and 7.72%, respectively, whereas LFGJ and LP showed less suppression, with levels of 15.72 and 22.26%.

At the family level (Fig. 6D), the predominant taxa included S24-7, Erysipelotrichaceae, Lactobacillaceae, Lachnospiraceae, Desulfovibrionaceae, Rikenellaceae, Ruminococcaceae, Bacteroidaceae, Helicobacteraceae, Verrucomicrobiaceae, and Bifidobacteriaceae. In control mice, Rikenellaceae, Dentinobacteriaceae, and particularly Lactobacillaceae were dominant, accounting for 26.72, 9.12, and 41.65% of total sequences, respectively. All experimental groups exhibited a pronounced decline in Lactobacillaceae (P < 0.05) and a significant rise in Desulfovibrionaceae, a family associated with endotoxin production and inflammation (P < 0.05), indicative of dysbiosis within the intestinal microbiota. Ruminococcaceae, known for promoting nutrient absorption and producing beneficial metabolites such as butyrate³⁴, showed significantly increased abundance following LFGJ and LP interventions compared to the model group. Interestingly, the abundance of Lachnospiraceae, another major butyrate-producing bacterial family critical for gut health, decreased after HFGJ and Metformin intervention. These shifts in butyrate-producer abundance may explain the lack of significant differences in butyrate levels among the three groups.

At the genus level (Fig. 6E), Allobaculum, Lactobacillus, Bacteroides, Oscillospira, Helicobacter, Adlercreutzia, Akkermansia, and Bifidobacterium were most abundant. The beneficial bacteria, such as Lactobacillus and Bifidobacterium, declined sharply in the model group, indicating widespread dysbiosis of the gut microbiota. Intervention with HFGJ and UFGJ effectively attenuated the decline in Lactobacillus, while LFGJ and UFGJ promoted the recovery of Bifidobacterium, suggesting a positive regulatory effect on beneficial bacterial populations. Furthermore, the abundance of Allobaculum, a health-associated genus, was significantly increased in the Metformin and HFGJ groups compared to the model group. In addition, Akermansia is a beneficial gut bacterium that can alleviate intestinal inflammation and regulate immune responses. Following intervention with HFGJ or Metformin, its abundance also significantly increased, indicating an improvement in the gut microbiota.

Shifts in microbial composition and consequent changes in SCFAs are associated with the development of insulin resistance and diabetes. Acetic acid makes up 50%-60% of SCFAs and is primarily synthesized by Bifidobacterium spp. and Lactobacillus spp. via the acetyl-CoA pathway³⁵. Propionate is generated through the succinate pathway by certain Firmicutes (e.g., Ruminococcus, L. paracasei) and Bacteroidetes, while butyrate arises via lysine, glutamate, or 4-aminobutyrate routes employed by multiple Firmicutes^35,36. The present study suggests that FGJ not only increases levels of SCFAs, but also enriches SCFA-producing bacteria, thereby contributing to the restoration of microbial balance and improved metabolic health.

Correlation analysis

To elucidate the potential mechanisms by which FGJ ameliorates type II diabetes, a comprehensive correlation analysis was performed, integrating data on juice composition, glycemic indices, lipid metabolism, oxidative stress markers, gut microbiota, and SCFAs (Fig. 7).

Fig. 7. Correlation heatmap among glycemic indices, lipid profiles, oxidative stress markers, SCFAs, and gut microbiota in diabetic mice.

Red and blue colors represent positive and negative correlations, respectively. Asterisks indicate statistical significance (*P ≤ 0.05, **P ≤ 0.005, ***P ≤ 0.001).

The in vitro antioxidant capacity of the juice, assessed via DPPH, ABTS and hydroxyl radical scavenging assays, exhibited a strong positive correlation (P < 0.001) with its total flavonoid and polysaccharide content. This confirms that the improvement in antioxidant activity following fermentation is directly linked to the elevation of these bioactive compounds, which are well recognized for their radical scavenging abilities.

Notably, the enhanced physicochemical properties of FGJ were subsequently linked to in vivo metabolic improvements. The antioxidant activity showed a significant negative correlation with key diabetic markers, including FBG and AUC. Furthermore, a strong negative correlation was observed between the flavonoid content and serum levels of TG. These results suggest that the intake of FGJ, rich in antioxidants, directly contributes to the alleviation of hyperglycemia, dyslipidemia, and systemic oxidative stress in diabetic mice.

The most insightful findings emerged from the analysis linking the gut microbiome to host metabolism. Propionic acid, a major SCFA that was significantly elevated in the HFGJ group, exhibited a strong negative correlation with FBG and AUC, underscoring their crucial role in improving glucose homeostasis and lipid metabolism. The beneficial effects of SCFAs were further supported by the positive correlation with specific gut bacteria abundance. For instance, the abundances of Lactobacillus and Akkermansia were positively correlated with propionic acid. Akkermansia, a mucin-degrading bacterium renowned for its positive role in metabolic health³⁷, also showed a significant negative correlation with FBG and MDA. On the other hand, the abundance of Desulfovibrio, a sulfate-reducing bacterium often associated with gut inflammation³⁸, was positively correlated with isobutyric acid, acetic acid and isovaleric acid.

These findings integrate the compositional, biochemical, and microbial outcomes into a coherent framework, whereby FGJ exerts antidiabetic efficacy through a cascade of metabolic improvements, including lowering blood glucose, ameliorating dyslipidemia, reducing oxidative stress, and reshaping gut microbiota to enhance SCFA production. This multi-level correlation underscores the interplay between host metabolism and microbial ecology in mediating the health benefits of fermented functional foods.

In summary, fermentation of goji berry juice with L. paracasei significantly improved its nutritional and functional properties, as evidenced by elevated levels of flavonoids and polysaccharides, enhanced antioxidant activity, and a refined volatile aroma profile. These compositional and sensory improvements provided a biochemical foundation for subsequent physiological benefits. In a type II diabetic‑mouse model, high‑dose fermented juice lowered FBG by 33%, improved glucose tolerance, and corrected dyslipidemia. It also attenuated oxidative stress and partially ameliorated liver, kidney, and pancreatic histopathology, while reducing diabetes‑related weight loss and polydipsia. Moreover, the intervention reshaped gut microbiota composition, elevated concentrations of SCFAs, and thereby supported glycemic control. Correlation analysis uncovered coherent linkages among the enhanced bioactivity, metabolic parameters, and gut microbiota composition, supporting that the antidiabetic effects of FGJ are mediated through a gut microbiota-metabolism axis. Collectively, these results highlight FGJ as a promising functional food candidate for managing hyperglycemia and associated metabolic disorders, warranting further mechanistic studies and product‑development efforts. However, the current characterization methods for the quality attributes of goji berry juice are not comprehensive. Future studies will focus on the characteristics of different fermentation strains and employ more advanced analytical techniques to thoroughly investigate the evolution of its components during the fermentation process.

Methods

Materials and reagents

The goji berry used in the experiment was Ningqi 10. Lactobacillus paracasei 5572, characterized by its strong probiotic functions, potent antioxidant capacity, and excellent adaptability in plant-based food fermentation⁷, was provided by the Key Laboratory of Food Microbial Application Technology and Safety Control of Ningxia University. High-fat feed (XTHF 60) was produced by Jiangsu Synergy Pharmaceutical Bioengineering Co. MRS Broth was obtained from Beijing Coolaber Science& Technology Co., Ltd. Streptozotocin (STZ), Catalase (CAT) Activity Assay Kit, and Malondialdehyde (MDA) Content Assay Kit were purchased from Beijing Solarbio Science & Technology Co., Ltd. Glutathione peroxidase (GSH-Px) assay kit was purchased from Nanjing Jiancheng Bioengineering Institute. Superoxide Dismutase (SOD) Activity Assay Kit was obtained from Beijing Boxbio Science & Technology Co., Ltd.

Preparation of fermented goji berry juice (FGJ) samples

Preparation of bacterial suspension: L. paracasei was inoculated into MRS medium and incubated at a constant temperature of 37°C for 12 h. The bacterial solution was subjected to freeze centrifugation, and the bacterial cells were resuspended and diluted using sterile saline until the concentration reached 10⁹ CFU/mL (OD₆₀₀ = 1.5).

FGJ were prepared were prepared as follows. After filtration, the pH of the goji berry juice was adjusted to 4.5. The juice was then sterilized at 65 °C for 30 min, cooled, and inoculated with a prepared bacterial suspension (5% v/v). Fermentation was carried out at 37 °C for 12 h.

Determination of nutritional components in FGJ

The total phenolic content was determined using the Folin-Ciocalteau method¹⁵, while the total flavonoid content was measured using the colorimetric method³⁹. The polysaccharide and soluble protein contents were quantified using the phenol-sulfuric acid method and the Coomassie Blue Staining method, respectively.

Antioxidant activity analysis

DPPH radical scavenging activity

DPPH radical scavenging activity was determined according to a published method with some modifications⁴⁰. Firstly, a 0.2 mmol/L DPPH• solution was prepared using anhydrous ethanol. Both unfermented and fermented goji berry juices were diluted to different concentrations. Then, 2 mL of each diluted sample was mixed with an equal volume of the above DPPH• solution. The mixture was kept in the dark for 30 min, and the absorbance at 517 nm was measured. The DPPH• radical scavenging activity was calculated using Eq. 1.

$$\mathrm{DPPH}\mathrm{RSA}(\%)=[1-(({\mathrm{A}}_2-{\mathrm{A}}_3)/{\mathrm{A}}_1)]\times 100$$ 1

Where A₁ is the absorbance of the DPPH• solution mixed with anhydrous ethanol; A₂ represents the absorbance of the DPPH• solution mixed with the sample after reaction; A₃ is the absorbance of the sample solution mixed with anhydrous ethanol.

ABTS radical scavenging activity

The ABTS solution was prepared by mixing equal volumes of ABTS (7.4 mmol/L) and K₂S₂O₈ (2.6 mmol/L). The mixture was kept in the dark at room temperature for 12 h to generate free radicals. Subsequently, the solution was diluted with anhydrous ethanol until the absorbance at 734 nm reached 0.70 ± 0.02, resulting in a 7 mmol/L ABTS•+ solution. Both unfermented and fermented goji berry juices were diluted to different concentrations, mixed thoroughly, and incubated in the dark at room temperature for 10 min⁴¹. The absorbance at 734 nm was measured, and the ABTS radical scavenging activity was calculated by Eq. 2.

$$\mathrm{ABTS}\mathrm{radical}\mathrm{scavenging}\mathrm{activity}(\%)=[1-(({\mathrm{A}}_2-{\mathrm{A}}_3)/{\mathrm{A}}_1)]\times 100$$ 2

Where A₁ represents the absorbance of the ABTS•+ mixed solution without the sample; A₂ is the absorbance of the ABTS•+ mixed solution with the sample; A₃ represents the absorbance of the sample solution mixed with ultrapure water.

Hydroxyl radical scavenging activity

The reaction mixture for the hydroxyl radical scavenging assay was prepared by sequentially combining 1 mL of each of the following solutions: 9 mmol/L ethanol-salicylic acid, 9 mmol/L ferrous sulfate, 8.8 mmol/L hydrogen peroxide (30%, v/v), and the sample solution. The mixture was incubated at 37 °C for 30 min, and the absorbance was measured at 510 nm⁷. The hydroxyl radical scavenging rate was calculated according to Eq. 3.

$$\mathrm{Hydroxyl}\mathrm{radical}\mathrm{scavenging}\mathrm{activity}(\%)=[1-(({\mathrm{A}}_2-{\mathrm{A}}_3)/{\mathrm{A}}_1)]\times 100$$ 3

Where A₁ is the absorbance of the hydroxyl radical-generating mixture with ultrapure water, A₂ is the absorbance of the mixture with the sample, and A₃ is the absorbance of the sample solution in ultrapure water.

Analysis of volatile aroma compounds

The changes in volatile compounds in goji berry juice before and after fermentation were analyzed using GC-IMS. 1 mL of the sample was incubated at 40 °C for 20 min. The injection needle temperature was set at 85 °C, and the injection volume was 500 μL. The separation was performed on an MXT-WAX column (30 m × 0.53 mm × 1 μm) maintained at 80 °C. High-purity N₂ was used as the carrier gas with a flow rate program as follows: 0–2 min, maintained at 2 mL/min; increased from 2 mL/min to 10 mL/min; then increased to 100 mL/min; and finally maintained at 100 mL/min for 20–40 min. The IMS drift tube temperature was set at 45 °C, and the drift gas was N₂ with a flow rate of 150 mL/min.

Establishment of the diabetic mouse model and grouping

SPF-grade ICR male mice (5 weeks old) were obtained from the Experimental Animal Center at Ningxia Medical University. The mice were acclimated for 4 days and then randomly assigned to a control group (n = 10) or an experimental group (n = 60). The control group was fed a standard diet, whereas the experimental group received a high-sugar and high-fat diet. Body weight was measured at regular intervals. After 5 weeks of feeding, the mice were fasted for 12 h without water and then intraperitoneally injected once daily with 0.1 mL of a 20 mg/kg STZ solution for 10 consecutive days. The control group received an equivalent volume of sterile citric acid-sodium citrate buffer. Mice with FBG level ≥11.1 mmol/L were considered successfully modeled for type II DM⁴².

After successful model establishment, the experimental group received 4 weeks of gavage administration with the test samples, during which the animals had free access to food. In total, seven groups were included in the study (Table S3): the control group (Control), the model control group (Model), the metformin group (Metformin), the high-dose fermented goji berry juice group (HFGJ), the low-dose fermented goji berry juice group (LFGJ), the unfermented goji berry juice group (UFGJ), and the bacterial suspension group (LP). Subsequent evaluations of blood glucose, lipid profiles, oxidative stress, histopathology, and gut microbiota were all designed to elucidate the mechanisms underlying the improvement of diabetes.

All experimental protocols were approved by the Animal Care and Use Committee of Ningxia Medical University (Approval No. IACUC-NYLAC-2023-229) and were conducted in strict compliance with the European Directive 2010/63/EU on animal experimentation, ensuring ethical integrity and scientific rigor.

Observation of the apparent state

The body weight values of the mice were measured and recorded every 7 d. Since the modeling, the apparent status was always observed, such as hair condition, dietary condition, water intake condition, and urination condition.

Measurement of blood glucose-related indicators

FBG determination

Samples were gavaged for each group, and FBG was monitored at a fixed time each week thereafter. To ensure the accuracy of the test results, the mice were allowed to drink normal water, and blood was collected from the tails after 12 h of fasting for the test.

Oral glucose tolerance test (OGTT):

Initial blood glucose at 0 min was measured and mice were given 20% glucose solution by gavage at a dosage of 2 g/kg; blood glucose values were measured at 30, 60, 90, and 120 min, and the area under the glucose curve (AUC) was calculated according to Eq. 4⁴³.

$$\mathrm{AUC}(\mathrm{hmmol}/\mathrm{L})=(0.5\mathrm{A}+\mathrm{B}+\mathrm{C}+0.5\mathrm{D})/2$$ 4

Where A, B, C and D were blood glucose values at 0, 30, 60 and 120 min, respectively.

Collection and processing of mice samples

At the end of the sample gavage, fresh feces were collected from mice. Each group was placed in separate cages, and their fresh feces were collected into sterilized and dried EP tubes, placed on ice and quickly transferred to a −80 °C fridge. Mice were anesthetized with 100 mg/kg ketamine and 20 mg/kg xylazine by intraperitoneal injection. Upon loss of pedal withdrawal reflex, mice were killed by cervical dislocation immediately after collection of blood by eyeball enucleation. Blood samples were collected and left at room temperature to naturally stratify, centrifuged at 4 °C for 15 min, and the upper layer of serum was collected and transferred to a −80 °C refrigerator for storage. Following aseptic dissection, the heart, liver, spleen, lungs, kidneys, and pancreas were gently rinsed in sterile saline, and the excess water was removed from the filter paper and weighed to calculate the organ index. After rinsing the livers, kidneys, and pancreas in precooled saline at 4 °C, they were put into a fixed tube for fixation and subsequent histopathological studies.

Indicator measurement

The organs were rinsed with sterile saline and weighed, and the ratio of the weight of the viscera to the body weight of the mice was indexed for each organ. The serum levels of Lipid IV were determined according to the kit instructions. Liver homogenates were prepared according to the instructions of each kit, and oxidative stress indices of mouse liver homogenates were determined.

Pathological observations on tissue sections

Mouse liver, kidney, and pancreas tissues were collected and gently rinsed with sterile saline. Excess moisture was removed using filter paper, and the tissues were fixed in 4% paraformaldehyde for histopathological analysis. After paraffin embedding, the samples were dewaxed, rehydrated, and stained with hematoxylin-eosin (HE). The sections were then dehydrated, mounted, and examined under a light microscope. Photomicrographs were taken to document pathological changes, with particular attention to areas displaying specific lesions.

Analysis of short-chain fatty acids and gut microbiota

Short-chain fatty acids (SCFAs) in mouse cecal contents were quantified using GC-MS, following the procedure described by ref. ⁴⁴. Pure standards of acetic, propionic, butyric, isobutyric, valeric, and isovaleric acids were purchased from Sigma-Aldrich (St. Louis, MO, USA) and dissolved in water to prepare a 100 mg/mL stock solution. This stock solution was then diluted to generate a series of working standard solutions. An internal standard (4-methylpentanoic acid) was prepared at 375 μg/mL in diethyl ether. The stock solution was stored at −20 °C, and working solutions were used immediately after preparation. Mass spectrometry conditions were adopted according to the literature method⁴⁵.

Fecal DNA was first extracted and quantified using a NanoDrop spectrophotometer. The quality of the extracted DNA was assessed by 1.2% agarose gel electrophoresis. Target fragments were then PCR-amplified to obtain sufficient DNA for subsequent sequencing analysis. The PCR products were purified using magnetic beads to remove impurities, thereby improving DNA purity. These purified products were subsequently subjected to fluorescence-based quantification to accurately determine DNA concentration, ensuring a reliable input for sequencing. Finally, DNA sequencing was performed on the high-throughput Illumina MiSeq platform, providing a large volume of high-quality data to analyze the diversity of the gut microbiota.

Statistical analysis

All experiments were performed in triplicate, and the data were expressed as means ± standard deviation. SPSS 23.0 software was used to analyze the data statistically, combining one-way analysis of variance (ANOVA) and multiple comparisons to calculate the significance of the data. Origin 2024 was applied to graph the data and perform correlation analysis.

Author contributions

J.X. Conceptualization, data curation, formal analysis, investigation, visualization, and writing-original draft, Y.Q. Investigation, methodology, software, and visualization, X.W. Investigation, validation, and methodology, W.D. Software, visualization, and writing—review and editing, H.W. Visualization, resources, H.L. Conceptualization, project administration, supervision, and writing—review and editing, and H.F. Conceptualization, funding acquisition, project administration, resources, writing—review and editing.

Data availability

The authors declare that all relevant data supporting this study are included in the paper. Raw data will be made available by the corresponding authors upon request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41538-025-00683-6.