Neutral sugar, uronic acid, and protein contents

Neutral sugar, uronic acid, and protein contents were determined colorimetrically. For neutral sugar determination, an RRTP solution was mixed with a phenol solution and concentrated sulfuric acid, incubated for color development, and measured at 490 nm using Glc as the standard [DuBois et al., 1956]. Uronic acid content was determined by the m-hydroxydiphenyl method; briefly, an RRTP solution was reacted with a sulfuric acid–borate reagent, followed by the addition of the m-hydroxydiphenyl reagent, and absorbance was recorded at 525 nm using GalA as the standard [Blumenkrantz & Asboe-Hansen, 1973]. Protein content was measured using the Coomassie Brilliant Blue G-250 method (Bradford protein assay), with absorbance recorded at 595 nm and BSA used as the standard [Zor & Selinger, 1996]. Results were calculated from the corresponding calibration curves and expressed as g of standard equivalent per 100 g of dried RRTP sample: g glucose equivalent/100 g for neutral sugars, g galacturonic acid equivalent/100 g for uronic acids, and g BSA equivalent/100 g for proteins. Results were calculated based on calibration curves using Glc for neutral sugars, yielding y = 0.0054 x + + 0.0711 (R² = 0.9987, n = 6); BSA for protein, yielding y = 0.0039 x + + 0.4173 (R² = 0.9985, n = 6); and GalA for uronic acids, yielding y = 0.0285 x + 0.0311 (R² = 0.997, n = 6).

Monosaccharide composition

Monosaccharide profiles were determined by high-performance anion-exchange chromatography (HPAEC) on a Thermo Dionex ICS-6000 system (Thermo Scientific, Waltham, MA, USA) following a validated procedure [Shi et al., 2025]. Each RRTPs sample (6 mg) was hydrolyzed in 2 M trifluoroacetic acid (4 mL) at 121°C for 3 h. Following acid removal, hydrolysates were diluted to 50 mL with distilled water, filtered (0.22 μm), and injected (25 μL) onto a Dionex CarboPac PA20 analytical column (3×150 mm; Thermo Scientific) equipped with a guard cartridge. Separation was performed at 30°C, 0.5 mL/min, using ultrapure water (A), 40 mM NaOH (B), 500 mM sodium acetate (C), and 250 mM NaOH (D) under the following gradient program: 0–20.05 min, 90% A and 10% B; 20.05–53.05 min, 70% A, 20% B and 10% C; 53.05–60.00 min, 50% A and 50% D. Peak identification and compound quantification were carried out using sixteen monosaccharide standards (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) analyzed under identical conditions. The amount of each monosaccharide was calculated from its corresponding external calibration curve and converted to molar amount. Results were expressed as relative molar fractions (mol%), calculated as ni/Σni, where ni represents the molar amount of an individual monosaccharide and Σni represents the total molar amount of all identified monosaccharides in the same sample. For each RRTP sample, the sum of the identified monosaccharide fractions was normalized to 100. Pectic structural parameters were estimated from the relative molar fractions; HG and RG-I were calculated according to Equations (2) and (3), respectively:

Rha/GalA was used to indicate the relative contribution of the RG-I backbone; and (Ara + Gal)/Rha was used to indicate the relative degree of RG-I side-chain branching.

Molecular weight distribution

Molecular-weight distributions were assessed by high-performance gel permeation chromatography (HPGPC) using Agilent 1260 system with a refractive-index detector (Agilent Technologies, Santa Clara, CA, USA) [Shi et al., 2025]. RRTPs were dissolved in a 0.1 M NaNO₃ solution at 5 mg/mL and filtered through 0.45 μm membranes. Samples were separated on a TSK-Gel GMPWXL column (7.8 × 300 mm; Tosoh Bioscience, Tokyo, Japan) at 30°C. A 0.1 M NaNO₃ solution was used as the mobile phase at a flow rate of 0.5 mL/min, and the injection volume was 25 μL. A calibration curve relating retention time to the logarithm of molecular weight (log Mw) was generated using dextran standards (6.3–739 kDa; Agilent Technologies).

Fourier transform infrared spectroscopy

Fourier transform infrared (FT-IR) spectra of RRTPs were acquired by the KBr-pellet method on a Spectrum Two FT-IR spectrometer (PerkinElmer, Waltham, MA, USA) as described by Shi et al. [2025]. For each run, RRTP samples (2 mg) were mixed with dry KBr (200 mg), finely ground in an agate mortar, and pressed into translucent disks. A blank KBr pellet was recorded as the background. Spectra were collected from 4,000 to 400 cm⁻¹ at 4 cm⁻¹ resolution to assign functional groups and backbone features.

Thermal stability

Thermal stability of RRTPs was evaluated by thermogravimetric analysis (TGA) on a DTG-60A instrument (Shimadzu, Kyoto, Japan) [Chen et al., 2025]. Approximately 5 mg of the sample was sealed in a high-purity aluminum crucible, with an empty crucible used as the reference. Heating was performed from 25°C to 800°C at 10°C/min under nitrogen (50 mL/min). The thermogravimetric (TG) curve and the derivative thermogravimetric (DTG) curve were calculated separately to resolve mass-loss stages and peak decomposition temperatures.

Structural characterization by X-ray diffraction analysis

X-ray diffraction (XRD) analysis of RRTPs was performed using a D8 Advance diffractometer (Bruker, Leipzig, Germany) equipped with a copper anode producing Cu Kα radiation (λ=1.5406 Å) and operated at 40 kV and 40 mA. Diffraction patterns were collected over 2θ = 5°–60° at a scan speed of 10°/min. Peak deconvolution analysis was performed to resolve the crystalline (A_b) and amorphous (A_a) regions within the polysaccharide samples. The crystallinity index (CrI, %) was calculated according to Equation (4) [Wang et al., 2022b]:

Zeta potential and particle size distribution

Surface charge and hydrodynamic size of RRTPs (2 mg/mL in Milli-Q water) were determined at 25°C by dynamic light scattering on a Nano ZS90 analyzer (Malvern Instruments, Worcestershire, UK), following Shi et al. [2025]. Zeta potential was reported in mV, and particle size was expressed as the intensity-weighted Z-average diameter (nm). To ensure reproducibility, polydispersity indices (PDI) were maintained below 0.40 for all preparations. Measurement parameters were as follows: refractive indices of 1.414 (dispersed phase) and 1.330 (continuous phase), solvent viscosity of 0.8872 cP, and a target signal intensity of 80 counts/s.

α-Glucosidase inhibitory activity

The α-glucosidase inhibitory activity of RRTPs was evaluated using a pNPG-based assay according to Tadera et al. [2006], with appropriate modifications. RRTP solutions were prepared in phosphate-buffered saline (PBS; 100 mM, pH 6.9) at concentrations of 0.5–7.0 mg/mL. α-Glucosidase solution (0.5 U/mL) and pNPG solution (6 mM) were also prepared in the same buffer. For each assay, 100 μL of the RRTP solution was combined with 100 μL of α-glucosidase solution and incubated at 37°C for 10 min. Subsequently, 100 μL of pNPG solution was added, and incubation was continued for 20 min at 37°C. Parallel controls were run in the absence of RRTPs and in the absence of enzyme; acarbose served as the positive control. Reactions were terminated with 1 mL of 1 M Na₂CO₃. Formation of p-nitrophenol was quantified at 405 nm using a microplate reader (Model 1500, Thermo Multiskan SkyHigh, Waltham, USA). Percentage inhibition was calculated according to Equation (7):

Inhibitory kinetics toward α-glucosidase

The inhibition kinetics of RRTPs against α-glucosidase (0.5 U/mL) were evaluated by measuring enzyme activity at different pNPG concentrations (1–4 mM) in the absence or presence of RRTPs (0, 3, and 5 mg/mL). The inhibition mode and mechanism were analyzed using Lineweaver–Burk double-reciprocal plots within the Michaelis–Menten framework [Lineweaver & Burk, 1934]. The data were analyzed according to the mixed-inhibition equation (Equation (8)), from which the Michaelis constant (K_m) and maximal velocity (V_max) were obtained. The slope and y-intercept of the Lineweaver–Burk plot were determined for each RRTP concentration. The dissociation constant for inhibitor binding to the free enzyme (K_i) was obtained from the absolute value of the x-intercept of the secondary plot of the Lineweaver–Burk slope versus inhibitor concentration (Equation (9)). The dissociation constant for inhibitor binding to the enzyme–substrate complex (K_is) was obtained from the absolute value of the x-intercept of the secondary plot of the Lineweaver–Burk y-intercept versus inhibitor concentration (Equation (10)) [Cornish-Bowden, 1974]. In Equations (8)–(10), v denotes the initial reaction rate, [S] stands for the substrate concentration, and [I] indicates the inhibitor concentration.

α-Amylase inhibitory activity

The α-amylase inhibitory activity of RRTPs was determined using the 3,5-dinitrosalicylic acid (DNS)-based amylase assay described by Bernfeld [1955], with appropriate modifications. Sample solutions were prepared in 0.02 M phosphate buffer (pH 6.9). Equal volumes of the RRTP solution (500 μL) and α-amylase (10 U/mL) were combined and incubated at 37°C for 10 min. Soluble starch (1%, 500 μL) was then added as a substrate, and incubation was continued for 10 min at 37°C. Reactions were terminated by adding the DNS reagent (1 mL) and heating in a boiling water bath for 5 min. After dilution with distilled water (10 mL) and cooling to ambient temperature, absorbance was recorded at 540 nm. Percent inhibition was calculated according to Equation (7). Acarbose was used as the positive control. The IC₅₀ value was defined as the RRTP concentration required to inhibit 50% of α-amylase activity. IC₅₀ values were calculated from dose–response curves by plotting inhibition (%) against RRTP concentration and fitting the data by nonlinear regression.

Monosaccharide composition

The monosaccharide profile of RRTPs was analyzed using HPAEC. Representative chromatograms are presented in Figure S1 (Supplementary Materials), and the relative contents of identified monosaccharides are shown in Table 2. All eight RRTP samples contained the same eight monosaccharides, namely Rha, Fuc, Glc, Man, Xyl, Ara, Gal, and GalA. Despite this qualitative similarity, significant differences in their molar proportions were observed among the samples from different production sites, particularly for Ara, Gal, and GalA (Table 2). RRTP-CJ exhibited the highest GalA content (49.72 mol%), whereas RRTP-QXG and RRTP-SN had the lowest (21.92 and 22.76 mol%, respectively). These trends paralleled the uronic acid content, thereby reinforcing the regional structural heterogeneity of RRTPs. The predominance of GalA, Gal, and Ara – monosaccharides characteristic of pectic polysaccharides – indicated that RRTPs were primarily composed of pectin-type structures. Generally, three main domains are distinguished in the pectin structure: homogalacturonan (HG), rhamnogalacturonan-I (RG-I), and rhamnogalacturonan-II (RG-II). HG consists mainly of linear chains of α-(1→4)-linked GalA residues that may be partially methyl-esterified or acetylated, whereas RG-I contains an alternating Rha–GalA backbone bearing Ara- and Gal-rich side chains [Voragen et al., 2009]. Quantitative analysis revealed that RG-I constituted a substantial portion of the RRTPs, ranging from 46.24 to 76.17 mol% (Table 2). The highest RG-I contents were found in RRTP-QXG (76.17 mol%) and RRTP-SN (67.90 mol%), whereas RRTP-CJ exhibited the lowest value (46.24 mol%).

The Rha/GalA and (Ara + Gal)/Rha ratios were further calculated as compositional indicators of the relative contribution of RG-I backbone residues and the abundance of Ara- and Gal-containing side chains, respectively [Dimopoulou et al., 2019]. The Rha/GalA ratios of all RRTPs ranged from 0.09 to 0.25, consistent with the range of 0.05–1.00 commonly reported for RG-I-containing pectins [Schols & Voragen, 1996]. RRTP-QXG had the highest Rha/GalA ratio, whereas RRTP-QXG and RRTP-SN had the highest (Ara+Gal)/Rha ratios. In contrast, RRTP-CJ showed the lowest values for both indices and the highest estimated HG proportion (45.07 mol%), indicating a comparatively greater contribution of GalA-rich, less-branched regions. Together, these results demonstrate clear source-associated differences in monosaccharide composition and pectic domain distribution among RRTPs. RRTP-CJ had an elevated HG content and a relatively linear conformation, while RRTP-QXG exhibited the most extensive RG-I branching. Despite these architectural distinctions, all samples maintained high pectin purity. Such structural heterogeneity is expected to influence the physicochemical and functional properties of the polysaccharides and is therefore relevant to their targeted utilization in food and nutraceutical applications.

Molecular weight distribution

To evaluate source-associated variation, the molecular-weight characteristics of RRTPs from eight producing areas were analyzed by HPGPC. Calibration was performed using a dextran-based standard curve (Figure S2 in Supplementary Materials). Each chromatogram of RRTP separations showed a single dominant, near-Gaussian peak (Figure 1A), suggesting a relatively narrow apparent molecular weight distribution within each RRTP preparation. However, shifts in the retention time of the main peak were observed among samples, indicating significant differences in average molecular weight, which ranged from 77.8 to 181.1 kDa (Table 2). For comparison, the average molecular weights of R. roxburghii fruit polysaccharides obtained by hot-water, saline, alkaline, and citric-acid extraction were previously reported to range from 80.53 to 97.22 kDa [Chen et al., 2025]. The lower molecular weights observed in the present study were comparable to this range, whereas RRTP-LL, RRTP-HZ, RRTP-QXG, and RRTP-SN exhibited higher values. In turn, the average molecular weights observed here were lower than the value of 513.02 kDa reported for water-soluble polysaccharides from R. roxburghii by Wang et al. [2018b]. Therefore, the difference may be attributable to the variation in fruit material, extraction and post-extraction processing, which can influence the molecular-weight distribution of the recovered polysaccharides. Among the analyzed samples, RRTP-ZA and RRTP-CJ displayed the lowest average molecular weights (77.8 kDa and 80.6 kDa, respectively), whereas RRTP-QXG and RRTP-SN exhibited the highest values (179.4 kDa and 181.1 kDa, respectively). The higher average molecular weights of RRTP-QXG and RRTP-SN coincided with their higher Ara and Gal proportions and higher (Gal+Ara)/Rha ratios. This association is structurally plausible because the controlled enzymatic removal of specific pectic domains, including arabinan and galactan side chains, has been shown to progressively decrease the molecular weight of RG-I-rich citrus pectin [Wu et al., 2020]. Overall, these findings demonstrate clear source-associated differences in molecular-weight profiles among RRTPs, further supporting substantial supramolecular heterogeneity among samples from different producing areas.

Figure 1

High-performance gel permeation chromatography (HPGPC) separations of polysaccharides from Rosa roxburghii Tratt. fruits (RRTP) collected in different producing areas of Guizhou Province in China and the corresponding molecular-weight calibration curve (A), Fourier-transform infrared (FT-IR) spectra of RRTPs (B), and crystallinity index (CrI) obtained by X-ray diffraction (XRD) analysis. CrI values not sharing a common letter differ significantly (p<0.05). Mw, molecular weight; RID, refractive-index detector. Sample codes are listed in Table 1.

Fourier transform infrared spectroscopic characterization

Fourier-transform infrared spectroscopy was employed to identify the functional groups and backbone features of the RRTPs. The spectra of all eight samples showed a high degree of overlap (Figure 1B), indicating broadly similar functional-group profiles among the production sources. A broad, intense absorption band at 3,273 cm⁻¹ was attributed to O–H stretching vibrations, while a peak at 2,922 cm⁻¹ corresponded to aliphatic C–H stretching, confirming the polysaccharidic nature of the samples [Chen et al., 2025]. Signals at 1,734 cm⁻¹ and 1,603 cm⁻¹ were assigned to ester carbonyl stretching and asymmetric stretching of ionized carboxylate groups, respectively [Synytsya et al., 2003], which is the evidence for the presence of galacturonic acid residues. Weaker bands near 1,418 cm⁻¹ and 1,370 cm⁻¹ were consistent with C–H bending vibrations. In the range of 1,300–800 cm⁻¹, multiple diagnostic peaks were identified: 1,234 cm⁻¹ and 1,067 cm⁻¹ were attributed to C–O and C–O–C stretching vibrations; 1,012 cm⁻¹ reflected C–O–H bending of pyranose rings [Kačuráková et al., 2000]. A distinctive band at 824 cm⁻¹ indicated α-glycosidic linkages [Barker et al., 1954]. Overall, the similar spectra support the presence of a common pectic-polysaccharide framework in all RRTPs.

Structural characterization

XRD analysis was conducted to qualitatively and semi-quantitatively evaluate the solid-state organization of RRTPs across amorphous, semicrystalline, and crystalline regions. The diffraction profiles of the eight RRTP samples are shown in Figure S3 in Supplementary Materials. All samples exhibited broad halos centered around ~22° (2θ), without sharp Bragg reflections, indicating a predominantly amorphous microstructure. A minor exception was observed for RRTP-QXG, which displayed weak and broadened features near ~12° and ~22° (2θ), suggestive of localized short-range ordering within an otherwise disordered matrix. Although the positions of these diffuse features, and hence the associated d-spacings, were largely conserved in all samples, variations in peak widths and intensities caused crystallinity indices (CrI) to vary from 11.7% to 32.1% (Figure 1C). The highest CrI was recorded for RRTP-QXG. This relatively high CrI may reflect more regular packing of some polysaccharide chains, resulting in localized short-range ordered domains within the predominantly amorphous matrix. Similar broad XRD profiles, characterized by diffuse diffraction features at approximately 20–23° and the absence of prominent sharp reflections, have previously been reported for polysaccharides extracted from R. roxburghii fruit [Chen et al., 2025; Zhang et al., 2024].

Particle size and zeta-potential

Particle size and zeta-potential distributions of the RRTP samples were shown in Figure S4 in Supplementary Materials, while average values are summarized in Table 2. All samples exhibited unimodal yet relatively broad particle size distributions, indicating compositional homogeneity within each preparation while also suggesting a general tendency toward aggregation. RRTP-ZA exhibited the largest particle diameter (705 nm), followed by RRTP-CJ (613 nm) and RRTP-SN (554 nm). RRTP-LL, RRTP-PZ, RRTP-SC, and RRTP-QXG showed statistically comparable particle sizes, whereas RRTP-HZ had the smallest diameter (362 nm). Notably, RRTP-ZA and RRTP-CJ displayed relatively large hydrodynamic diameters despite their low average molecular weights (77.8 and 80.6 kDa, respectively). This pattern may be associated with differences in the organization of their pectic domains. Both samples had comparatively low RG-I contents (57.41 and 46.24 mol%, respectively) and low (Gal+Ara)/Rha ratios (8.81 and 7.94, respectively), indicating a relatively limited contribution of Ara- and Gal-rich neutral side chains. The composition and length of RG-I side chains have been shown to influence pectin hydration, while enzymatic removal of arabinose residues can enhance intermolecular association and aggregation [Kaczmarska et al., 2024; Larsen et al., 2011]. Therefore, the less extensive neutral side chains of RRTP-ZA and RRTP-CJ may provide weaker hydration and steric shielding, allowing closer intermolecular contacts and the formation of larger hydrodynamic assemblies. In contrast, RRTP-HZ, RRTP-QXG, and RRTP-SN exhibited higher estimated RG-I contents and/or (Gal+Ara)/Rha ratios, suggesting more extensive neutral side-chain structures that may enhance hydration and steric separation between polysaccharide chains. This structural feature may partly account for their smaller particle dimensions relative to their apparent molecular weights.

Zeta-potential measurements showed net negative surface charges across all RRTPs, ranging from −12.5 mV (RRTP-QXG) to −20.8 mV (RRTP-CJ) (Table 2 and Figure S4B in Supplementary Material). RRTP-CJ exhibited the most negative zeta potential, whereas RRTP-QXG and RRTP-SN showed the lowest absolute values and did not differ significantly (p≥0.05) from each other. The negative charge of these polysaccharides was mainly attributable to the dissociation of carboxyl groups in GalA-containing pectic domains. Consistent with its relatively high uronic acid content, RRTP-CJ possessed a greater negative surface charge, indicating stronger electrostatic repulsion between dispersed polysaccharide particles [Lutz et al., 2009]. In addition to uronic acid content, differences in the accessibility and ionization of carboxyl groups, as well as the molecular conformation of the polysaccharides, may also contribute to the variation in zeta potential among the RRTP samples. An absolute zeta potential of approximately 30 mV is commonly regarded as an empirical boundary for dispersions stabilized predominantly by electrostatic repulsion [Uskoković, 2012]. The absolute zeta-potential values of all RRTP samples were below this level, indicating relatively limited electrostatic stabilization and a tendency toward intermolecular association in aqueous solution. Among the samples, the more negative zeta potential of RRTP-CJ may nevertheless provide comparatively stronger electrostatic repulsion and better dispersion stability.

Thermal properties

Thermogravimetric analysis reflects the thermal stability of RRTPs, which is an important parameter for evaluating their suitability as functional food ingredients. The TG and DTG curves are presented in Figure 2. Heating from 25 to 800°C revealed three distinct stages of thermal degradation. Stage I corresponded to the evaporation of free and bound water and accounted for 11.9% to 15.6% of mass loss. The highest moisture losses were observed for RRTP-SN (15.6%), RRTP-PZ (15.2%), and RRTP-HZ (15.1%), while RRTP-CJ (11.9%) and RRTP-QXG (12.0%) exhibited the lowest losses, reflecting differences in water-binding capacity. Stage II represented the principal decomposition phase, characterized by cleavage of glycosidic linkages and degradation of polysaccharide side chains [Aburto et al., 2015]. RRTP-SC, RRTP-PZ, and RRTP-QXG showed the lowest peak decomposition rates (−6.12%, −5.62%, and −5.55%/min, respectively), indicating greater thermal resilience than the other samples. Stage III involved the breakdown of more recalcitrant structures, including C–O bond cleavage and terminal disruption of the glycosidic backbone [Aburto et al., 2015]. RRTP-ZA and RRTP-CJ exhibited the greatest mass losses during this phase, consistent with lower structural resistance under high thermal stress. Among all samples, RRTP-CJ demonstrated the lowest overall thermal stability. Despite exhibiting the highest onset temperature for stage II degradation (188.2°C), it incurred the largest stage II mass loss (67.3%), the greatest cumulative loss (84.1%), and retained the lowest residual mass (15.9%) at 800°C. In contrast, RRTP-QXG exhibited superior thermal resistance, with a relatively lower peak decomposition rate, the smallest total mass loss (73.5%), and the highest residual mass at terminal temperature. These thermal behaviors were closely associated with structural characteristics. A higher molecular weight, a high RG-I content, and greater side-chain complexity, reflected by elevated (Gal+Ara)/Rha ratios, have been reported to enhance thermal stability in pectic polysaccharides [Deng et al., 2020]. RRTP-QXG, which exhibited the highest values of these parameters, showed the greatest thermal stability, whereas RRTP-CJ, with comparatively lower values, was the least stable. Collectively, these findings indicate that the thermal performance of RRTPs is governed primarily by molecular architecture, especially molecular weight distribution and branching complexity. Understanding these structure–property relationships is important for evaluating RRTPs from different production areas and for guiding their targeted utilization in food systems requiring specific thermal tolerance.

Figure 2

Thermogravimetric (TG) and derivative thermogravimetric (DTG) curves of polysaccharides from Rosa. roxburghii Tratt. fruits (RRTP) collected in different producing areas of Guizhou Province in China including Sinan, RRTP-SN (A); Hezhang, RRTP-HZ (B); Longli, RRTP-LL (C); Panzhou, RRTP-PZ (D); Shuicheng, RRTP-SC (E); Qixingguan, RRTP-QXG (F); Congjiang, RRTP-CJ (G), and Zheng’an, RRTP-ZA (H).

Inhibition of digestive carbohydrase activity

In this study, RRTPs collected from eight major production areas in Guizho were evaluated for their inhibitory effects on α-amylase and α-glucosidase, with acarbose serving as a pharmaceutical benchmark. All RRTP samples demonstrated concentrationdependent inhibition of α-amylase, with activities at 7.0 mg/mL ranging from 52.45% (RRTP-SN) to 79.64% (RRTP-CJ) (Figure 4A). Based on the IC₅₀ values, the analyzed samples were ranked as follows: RRTP-CJ (2.425 mg/mL) < RRTP-ZA (3.175 mg/mL) < RRTP-LL (4.189 mg/mL) <RRTP-PZ (4.497 mg/mL) <RRTP-SC (5.068 mg/mL) < RRTP-HZ (5.441 mg/mL) < RRTP-QXG (5.817 mg/mL) = RRTP-SN (5.971 mg/mL). All RRTP samples had higher IC₅₀ values than acarbose, indicating weaker α-amylase inhibition under the assay conditions. In contrast, the RRTPs showed stronger inhibition of α-glucosidase, yielding 67.73– –92.46% inhibition at 7.0 mg/mL compared with 66.49% determined for acarbose (Figure 4B). Based on the IC₅₀ values, they were ranked as follows: RRTP-CJ (0.747 mg/mL) < RRTP-ZA (0.826 mg/mL) < RRTP-LL (1.022 mg/mL) < RRTP-PZ (1.101 mg/mL) < RRTP-SC (1.434 mg/mL) < RRTP-SN (1.655 mg/mL) < RRTP-HZ (1.787 mg/mL) < RRTP-QXG (2.035 mg/mL) < acarbose (2.873 mg mL). The stronger inhibition of α-glucosidase than α-amylase observed for the RRTPs was consistent with previous findings for R. roxburghii fruit polysaccharides obtained using different extraction solvents, which showed IC₅₀ values of 0.494– –1.234 mg/mL against α-glucosidase and 0.758–2.003 mg/mL against α-amylase [Chen et al., 2025]. The excellent inhibitory activity of RRTP-CJ and RRTP-ZA may be due to their relatively small molecular weight, a higher galacturonic acid content, and longer chains, which enhance their interaction with enzyme active sites [Zhang et al., 2025b].

Figure 4

Inhibitory effects of polysaccharides from Rosa roxburghii Tratt. fruits (RRTP) collected in different producing areas of Guizhou Province in China on α-amylase activity (A) and α-glucosidase activity (B). Left graphs show RRTP concentration-dependent curves, and the right graphs show half-maximal inhibitory concentration (IC₅₀). IC₅₀ values not sharing a common letter differ significantly (p<0.05). Sample codes are listed in Table 1.

Kinetics of α-glucosidase inhibition

Lineweaver-Burk plots of α-glucosidase inhibition by RRTPs are shown in Figure S5 in Supplementary Materials. For each RRTP, three inhibitor concentrations generated straight lines intersecting in the third quadrant, indicating a mixed-type mode of inhibition. Increasing polysaccharide concentrations reduced both V_max and K_m (Table 3), consistent with simultaneous binding to both the free enzyme (E) and the enzyme-substrate complex (ES). Secondary replots yielded apparent dissociation constants, K_i (E-I) and K_is (ES-I) (Figure S6 in Supplementary Materials). For all RRTPs, K_is<K_i (ratios 0.19–0.60), indicating preferential binding to the ES complex and a predominant non-competitive inhibition component [Zhang et al., 2025a]. Among all preparations, RRTP-CJ displayed the strongest association (K_is = 0.336 mg/mL; K_is/K_i = 0.190) and the lowest IC₅₀, substantiating its position as the most potent inhibitor in this series. While all RRTPs moderated digestive carbohydrase activity, the preparation from Congjiang (RRTP-CJ) consistently outperformed others, exhibiting potent α-glucosidase inhibition through a predominantly non-competitive, mixed-type mechanism. Given their additional antioxidant activity, RRTP-CJ, and to a lesser extent RRTP-ZA and RRTP-LL, emerge as promising functional ingredients for the dietary management of type 2 diabetes.