Forsythoside A (purity>98%), phillyrin (purity>98%), and phillygenin (purity>97%) were purchased from Dalian Meilun Biotechnology (China), and methanol (purity>99%) and acetonitrile (purity>99%) were obtained from Thermo Fisher Scientific—CN. Antibodies specific to α-tubulin, β-Actin, TBP, HO-1, Nrf2, KEAP1, P-NF-κB p65, p-IKKα/β were purchased from Cell Signaling Technology and Lipolysaccharide (LPS) and Penicillin-Streptomycin from Sigma-Aldrich.

In total, 180 FF samples (30 g) were acquired from apiculture producers from the Shanxi, Shaanxi, and Henan Provinces in China in August and October. In total, 180 samples were from different times and different provinces [FF in August from Shanxi Province (SXFF-A), October from Shanxi Province (SXFF-O), August from Shaanxi Province (SAXFF-A), October from Shaanxi Province (SAXFF-O), August from Henan Province (HNFF-A), and October from Henan Province (HNFF-O)]. FF samples were broken to pieces and sieved through 60 pieces of mesh, dissolved in methanol, and extracted by ultrasound for 30 min. On the next day, samples were extracted again by ultrasound for 30 min and then centrifuged at 3,000 rpm. The upper methanol layer was filtered using a 0.22 μm PVDF membrane and stored at −80°C.

The FF were powdered and screened through 60 meshs, and evenly poured on to tape. After gold-plating, a scanning electron microscope (SEM) (SUPRA 55 VP, Zeiss) was used to collect samples at an accelerating voltage of 5.00 kV with a magnification of 4,500 times.

After 60 mesh screening, the 1 mg FF samples were mixed with 100 mg KBr powder, and the mixture was ground to less than 2 μm and put into an infrared tablet press with 20–24 MPa for approximately 1 min form potassium bromide tablets. The analysis was performed in the frequency range of 4,000 cm⁻¹ to 400 cm⁻¹ using a Vertex 70 FT-IR spectrometer (Brooke Biotechnology, Germany).

Analysis of constituents in FF was performed on an 1525-2707-2489 series HPLC system (Waters, United States) equipped with a binary solvent manager, sample manager, column compartment, UV detector with 280 nm, and LC-Solution software. The separation was performed on an Agilent TC-C₁₈ column (5 μm, 4.6 × 250 mm). The mobile phase consisted of water containing 0.1% acetic acid (A) and acetonitrile (B). The linear gradient was as follows: 0–30 min, 10–25% B; 30–45 min, 25–75% B; and 45–50 min, 75% B at a flow rate of 1.0 ml/min. The column temperature was maintained at 28°C and the injection volume was 10 μL (Lee et al., 2018). Chemicals used as a standard are Forsythoside A, phillyrin, and phillygenin. The constituents were identified by comparison of their retention times to those of standard compounds under identical analysis conditions and the UV spectra with our in-house DAD library.

The NIR spectrum of FF was measured by an Antaris II FT-NIR spectrometer (Thermo Fisher Scientific, Verona, United States), with a wavenumber range from 10,000 to 4,000 cm⁻¹, scanning times of 64, and a resolution of 8 cm⁻¹. The signals were generated in reflectance (%R) mode and showed using log 1/R. The NIR spectrum was recorded in triplicate for each sample, and the average spectrum was used in the data analysis. Analysis of NIR spectrum was performed with TQ Analyst 9.0 software.

J774A.1 cells purchased from the American Type Culture Collection were cultured according to a previously described method (Alizadeh et al., 2021). Cells were cultured in high glucose DMEM containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco; Thermo Fisher Scientific, Inc.) in a humidified incubator at 37°C with 5% CO₂.

Cell viability was measured by the MTT assay, which is described in the previous research (Hung et al., 2019). Briefly, the cells were seeded in a 96-well plate at a density of 2 × 10⁴ cells/well, treated with 1, 0.5, 0.2, 0.1 and 0.05 mg/ml SXFF-A and 1, 0.5, 0.2, 0.1 and 0.05 mg/ml SAXFF-A for 24 h. After incubation, the supernatant was removed, and then 100 μL of MTT solution (0.5 mg/ml) was added to each well and incubated for 4 h at 37°C. Next, the cell culture supernatant was removed and the resulting formazan crystals were dissolved in 100 μL DMSO. The absorbance was evaluated at 540 nm using a Multi-Mode Microplate Reader (Thermo Fisher Scientific, Inc.).

The levels of NO in the J774A.1 cells and the supernatant were performed with Griess Reagent I and II (Beyotime Co., Ltd.). The cells were seeded in a 96-well plate at a density of 2 × 10⁴ cells/well, treated with or without 0.5, 0.2, and 0.1 mg/ml SXFF-A and 0.5, 0.2, and 0.1 mg/ml SAXFF-A for 0.5 h followed by the addition of LPS (1 μg/ml) for 24 h. After incubation, the cell culture supernatant was transferred to 96-well, and the remaining cells were cleaved with Cell lysis buffer for Western and IP (Beyotime Co., Ltd.) to measure NO levels in supernatant and cells according to the manufacturer’s instructions. Briefly, 50 μL cell culture supernatant and 50 μL centrifuged cell lysis supernatant was mixed with 50 μL Griess Reagent I and II respectively for 10 min, and the absorbance was evaluated at 540 nm using a Multi-Mode Microplate Reader.

J744A.1 cells were plated onto a 12-well plate at a density of 4 × 10⁵ cells/well, treated with or without 0.5, 0.2, and 0.1 mg/ml SXFF-A and 0.5, 0.2 and 0.1 mg/ml SAXFF-A for 0.5 h, and then had LPS (1 μg/ml) added to them for 12 h. Total RNA was extracted from collected cells using FastPure^® Cell/Tissue Total RNA Isolation Kit (Vazyme Biotech Co., Ltd.) according to the manufacturer’s instructions. cDNA was synthesized with total RNA (1 μg) using HiScript III RT SuperMix for qPCR (+gDNA wiper) (Vazyme) and subjected to real-time quantitative PCR (RT-qPCR) amplification using the ChamQ universal SYBR qPCR Master Mix (Vazyme) in a QuantStudio^®5 (Thermo Fisher Scientific, Inc.). The primer sequences are shown in Table 1, and the data were analyzed and expressed as relative gene expression to β-Action using the 2^−△△CT method.

J744A.1 cells were plated onto a 6-well plate at a density of 6 × 10⁵ cells/well, treated with or without 0.5, 0.2, and 0.1 mg/ml SXFF-A and 0.5, 0.2, and 0.1 mg/ml SAXFF-A for 0.5 h, then had LPS (1 μg/ml) added to them for 12 h. The total protein and nuclear protein were extracted using Enhanced RIPA Lysis buffer with a 1 mM PMSF and phosphatase inhibitor cocktail (Beyotime) and a Nuclear Extract kit (EMD Millipore Corp), respectively, according to the manufacturer’s instructions. The concentrations of protein were determined by a BCA protein assay kit (Beyotime) according to the manufacturer’s instructions. Then, briefly, the protein (10 μg/lane) was separated with SDS-PAGE and transferred to a PVDF membrane, followed by blocking with 5% non-fat milk. After blocking, they were incubated with primary antibodies at 1:1,000 dilution in primary antibody dilution buffer overnight at 4°C and then incubated with the following HRP-conjugated secondary antibodies (1:10,000) in TBST at room temperature for 1 h, and immune-reacted bands were detected with enhanced chemiluminescence reagents. The relative density of the western blot bands was determined using the BLT GelView 6,000 Pro software.

SPSS software (version 26.0), Origin 2021, and GraphPad (version 8.0) were used for statistical analysis. Statistical comparisons were assessed by one-way analysis (ANOVA), followed by Least significant difference (LSD) test. All experiments were performed at least three times independently.

The obtained SEM micrographs of FF from different harvest seasons and regions showed obvious differences in the surface morphologies, as shown in Figure 1A. The SEM images indicated that the surface structure of FF from the same regions in October was looser and had a more net-like structure than that in August, whereas no significant differences were observed among the three regions in August and October. There were also more pore-like structures on the surface of SXFF-A compared with that in SAXFF-A and HNFF-A (Figure 1A a–c).

The MIR spectra revealed that FF from different harvest seasons and regions had the same functional groups through the absorption bands of the phytocompounds (Figure 1B). Bands for FF were obtained at peak 3,389 cm⁻¹, 2,927.37 cm⁻¹,1738.17 cm⁻¹, 1,637.19 cm⁻¹, 1,245.86 cm⁻¹, 1,032.27 cm⁻¹, and so on. The results showed that FF in different harvest times and regions have the same absorption peaks and different transmittance. This suggested that FF were provided with identical major functional groups, but different MIR spectra transmittance, indicating that FF from different regions and harvest times had similar compounds, yet may have varied in the content of the compound.

The NIR spectrum of 180 FF samples from different gathering times and localities is shown in Figure 2, which shows that there is a significant distinction among them. FF samples were recorded by NIR spectra from 10,000 to 4,000 cm⁻¹ with a resolution of 8 cm⁻¹ (Figure 2A). Cluster analysis is used to analyze the similarity of the NIR spectrum of 180 FF samples by clustering samples based on their intimacy. K-means clustering analysis is an iterative clustering analysis method. Firstly, all data are pre-divided into K groups, and K objects are randomly selected as the initial cluster center. Then, the distance between each object and each seed cluster center is calculated, and each object is assigned to the nearest cluster center. According to the results of K-means cluster analysis, 180 FF samples were divided into two clusters: FF from Hehan, Shanxi, and Shaanxi in August and in October. The different producing areas of FF were not distinguished, indicating that the effect of harvest time on FF was more obvious than that of regions (Figure 2B).

Principal component analysis (PCA), which reduces its dimensionality into several main components, is applied to analyze the NIR spectrum of the 180 FF samples. PCA is a dimensionality reduction method that converts a large quantity of data into a few comprehensive indicators, reducing the dimension of observation and obtaining the most important information. In this study, the NIR spectrum of PCA based on variance decomposition was projected into the new coordinate system, and the principal components (PC1, PC2, and PC3) take the eigenvalues that can reflect the maximum variance value. Figure 2C showed the PC1, PC2, and PC3 loading plots of different FF PCA models that corresponded to 70.9, 28.6, and 0.2% of the variance. The three-dimensional scores of PCA showed there was a significant difference in the distribution of the FF between August and October; Henan, Shanxi, and Shaanxi in August were also distinguished, whereas there was no significant difference in October of the FF from Henan, Shanxi, and Shaanxi (Figure 2C).

Taken together, the K-means clustering and PCA of NIR spectrum showed that FF from different producing areas and harvest seasons had obvious characteristics of the harvest season. Meanwhile, FF in August showed obvious characteristics of producing area, while FF from different producing areas in October is not distinguished. These results indicated that the impact of harvest season on FF was significantly greater than that of producing areas.

To investigate the differences in the surface morphology, MIR spectra transmittance, and the NIR spectra of FF from different harvest seasons and regions, the contents of forsythiaside A, phillyrin, and phillygenin were determined by HPLC. The chromatograms of the forsythiaside A, phillyrin, and phillygenin from standards and samples were shown in Figures 3A,B. And the contents of the three components showed that different producing areas and harvest time played obvious effects on the FF (Figure 3C). Compared with HNFF, the forsythiaside A, phillyrin, and phillygenin of SXFF and SAXFF in August was markedly improved, while that of HNFF and SAXFF was significantly higher than that of SXFF in October. The level of forsythiaside A and phillyrin in SXFF-A was the highest and the phillygenin in SAXFF-A was the highest, which indicated that the SXFF-A and SAXFF-A perhaps had the highest constituents basis.

Based on the high levels of forsythiaside A, phillyrin, and phillygenin by HPLC, SXFF, and SAXFF in August were selected to treat J774A.1 cells. The results of the MTT assay suggested that 0.5 mg/ml, 0.2 mg/ml, and 0.1 mg/ml of SXFF and SAXFF in August were not toxic to J774A.1 cells (Figure 4A), which indicated that the effects of SXFF and SAXFF on J774A.1 cells were not due to their cytotoxicity. The NO levels in supernatant and cells of LPS-induced J774A.1 cells were significantly reduced by SXFF and SAXFF (Figure 4B). In this study, the anti-inflammatory activity of FF on LPS-induced J774A.1 cells was investigated. As shown in Figure 5A, the relative expression levels of inflammatory cytokines and mediators, namely, TNF-α, IL-1β, NF-κB, and iNOS were markedly downregulated by SXFF and SAXFF (August) in LPS-induced J774A.1 cells. Moreover, SXFF obviously downregulated the relative expression levels of IL-1β and iNOS mRNA, and SAXFF significantly attenuated the expression of IL-1β and NF-κB in a dose-dependent manner.

NF-κB is a transcription factor in promoting inflammation, so we investigate whether the anti-inflammatory effect of SXFF and SAXFF is related to the suppression of NF-κB. Western blot analysis was performed to analyze the protein expression levels of phosphorylated total IKKα/β and nuclear NF-κB. As shown in Figure 5B, the levels of p-IKKα/β and nuclear p-NF-κB were markedly increased by LPS and remarkably decreased by SXFF and SAXFF in LPS-induced J774A.1 cells, and the levels of p-IKKα/β were decreased by SAXFF in a dose-dependent manner.

Taken together, these results showed that SXFF and SAXFF suppressed the relative expression levels of TNF-α, IL-1β, NF-κB, and iNOS mRNA and the protein expression levels of p-IKKα/β and nuclear p-NF-κB, suggesting that SXFF and SAXFF have evident anti-inflammatory function in vitro, and the anti-inflammatory effects may be related to suppressing the NF-κB signaling pathway.

Nrf2 is an important transcription factor regulating the cellular oxidative stress response, and also a central regulator maintaining intracellular redox homeostasis. Therefore, we investigated how SXFF and SAXFF regulate Nrf2 signaling to exert an antioxidant role, and whether there is a difference in antioxidant effect between SXFF and SAXFF. As shown in Figure 6A, the mRNA expression levels of HO-1 were significantly increased by SXFF and SAXFF with a dose of 0.5 mg/ml, and the NQO1 were markedly increased by SAXFF with a dose of 0.5 mg/ml in LPS-induced J774A.1 cells.

To further determine the antioxidant effect of SXFF and SAXFF, Western blot analysis was performed to analyze the protein expression levels of total HO-1, KEAP1, and nuclear Nrf2. As shown in Figure 6B, SXFF and SAXFF significantly improved the protein expression levels of total HO-1 and nuclear Nrf2. Meanwhile, the protein expression levels of total KEAP1were obviously reduced by SXFF and SAXFF in a dose-dependent manner, and the levels of KEAP1 in SAXFF were lower than that in SXFF.

Collectively, these results showed that SXFF and SAXFF promoted the mRNA expression levels of HO-1 and NQO1 and the protein expression levels of HO-1 and nuclear Nrf2 and suppressed the protein expression levels of KEAP1, suggesting that the antioxidant activity of SXFF and SAXFF on LPS-induced J774A.1 cells is correlated with the Nrf2 signaling, and the antioxidant activity in SAXFF was stronger than that in SXFF.

In view of the different contents of forsythoside A, phillyrin, and phillygenin in FF from different localities, we analyzed the correlation between them and the anti-inflammatory and antioxidant effects of FF based on Spearman correlation. As shown in Figure 6C, the results showed that the correlation between the levels of phillygenin and the protein expression levels of KEAP1, Nrf2, and p-IKKα/β was higher than that of forsythoside A and phillyrin, whereas the correlation between the phillyrin and the protein of HO-1 and p-NF-κB was higher than that of forsythoside A and phillygenin. Phillyrin and phillygenin belong to bicyclo lignans, which indicates that the anti-inflammatory and antioxidant effects of FF may be mainly exerted by the lignans.