JQJTT (190210) was provided by Longshunrong Pharmaceutical Factory of Tianjin Zhongxin Pharmaceutical Group Co., Ltd (Tianjin, China). PAL (CAS: 3486-67-7, purity >98%) was provided by Sichuan Vikqi Biotechnology Co., Ltd (Sichuan, China). FGF21 protein was obtained from Northeast Agricultural University. STZ and insulin were purchased from Sigma-Aldrich Co., Ltd (St. Louis, MO, United States). PD173074 was purchased from Selleck Chemicals (Houston, TX, United States). The glucose detection kit was obtained from Shanghai Yuanye Bio-Technology Co., Ltd (Shanghai, China). Trizol and the cDNA reverse transcription kit were purchased from Thermo Fisher Scientific (Carlsbad, CA, United States). SYBR Green qPCR Mix was provided by Shandong Sikejie Biotechnology Co., Ltd (Shandong, China). Primary antibodies targeting phosphorylated (p)-FGFR1 (ab173305), FRS2 (ab137458), AMPK (ab207442), p-AMPK (ab133448), and β-actin (ab8227) were purchased from Abcam (Cambridge, United Kingdom); those targeting p-FRS2 (YP0805), p-AKT (YT0185), and AKT (YP0006) were obtained from ImmunoWay (California,United States); and that targeting FGFR1 (9740T) was purchased from CST (Boston, MA, United States). Secondary antibodies were obtained from Absin Biotechnology Co., Ltd (Shanghai, China).

KM mice (6 weeks old, 20 ± 5 g) were obtained from Changchun Yisi Experimental Animal Technology Co., Ltd (Changchun, China). The mice were kept in a controlled environment (temperature: 24 ± 2°C; relative humidity: 50%–60%; 12-h light/12-h dark cycle) and had free access to food and water for 1 week. Type 2 diabetes models were established by STZ injection (50 mg/kg) twice in 3 days and were fed with high-fat diet at the same time. Fasting blood glucose level greater than 16.7 mmol/L was regarded as successful modeling. Model mice were then randomly divided into four groups (n = 12 per group), three of which received intragastric administration of JQJTT of differing concentrations (JQ-H: 800 mg/kg JQJTT; JQ-M: 400 mg/kg JQJTT; and JQ-L: 200 mg/kg JQJTT) for four consecutive weeks, while one (the model group) was fed the same amount of normal saline. The normal control group was also fed the same amount of normal saline. Changes in blood glucose in the mice were measured weekly. To further test the effect of PAL, mice were randomly divided into six groups (n = 12), PAL-H (80 mg/kg), PAL-M (40 mg/kg), PAL-L (20 mg/kg), and FGF21 (0.5 mg/kg) groups. Animals were treated via intraperitoneal injection for four consecutive weeks, when the control groups received an intraperitoneal injection of an equal amount of normal saline. Subsequently, the fasting blood glucose changes in mice were tested weekly. On day 28, the intraperitoneal glucose tolerance test (IPGTT) was performed. Furthermore, the mice were euthanized, and liver tissues were harvested for western blotting and quantitative PCR (qPCR) analysis. All experiments were carried out in strict accordance with the recommendations of the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and were approved by the Ethics Review Committee of Harbin University of Commerce (No. HSDYXY-2020006).

During drug treatment, once weekly, fasting blood glucose was directly measured in the tail vein of mice after a 12-h fast using a glucometer (Glucotrend; Roche Diagnostics Ltd, Mannheim, Germany). On day 28, changes in blood glucose in the mice were detected 1, 2, 4, and 8 h after PAL treatment. For the IPGTT, mice were fasted overnight (12–16 h) and then intraperitoneally injected with 20% glucose (2 g/kg). At 0, 30, 60, and 120 min after injection, blood was collected from the tail vein for the determination of blood glucose concentrations. Assessment was based on the area under the glucose curve (AUC), which was calculated using the following equation: AUC = 0.5 × (BG0 + BG30)/2 + 0.5 × (BG30 + BG60)/2 + 1× (BG60 + BG120)/2. The glucose uptake rate was calculated as follows: glucose concentration (mmol/L) = OD of the sample/OD of the standard × 5 mmol/L; glucose consumption rate (%) = [(C blank group glucose-C administration group glucose)/C blank group glucose] × 100%.

HepG2 cells were a kind donation from the Heilongjiang University of Traditional Chinese Medicine (Heilongjiang, China). The cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, Carlsbad, California, United States) supplemented with 10% fetal bovine serum (FBS) (Gibco, Carlsbad, California, United States) and 1% penicillin–streptomycin (Thermo Fisher Scientific) at 37°C and with 5% CO₂.

Fishing experiments were performed on a BIACORE T200 instrument at 25°C. FGFR1 protein was immobilized on all four channels of a CM5 chip following a standard EDC/NHS protocol. JQ-R tablets (JQ-R is a mixture of refined extracts from Coptis chinensis, Astragalus membranaceus, and Lonicera japonica, the three main constituents of JQJTT) were fully dissolved in water, and the insoluble residue was pelleted by centrifugation and discarded. Then, the supernatant was then injected at a flow rate of 10 μl/min. All four flow cells were used for analyte capture and recovery. The bound material was eluted with 0.5% trifluoroacetic acid.

For the kinetics experiment, six or seven analytes were serially diluted twofold, prepared in the running buffer, and injected at a flow rate of 30 μl/min onto the CM5 chips, followed by 1 min of dissociation data acquisition. To obtain more extensive off-rate decay data, an additional injection of the analyte with the third-highest concentration was included in each experiment. All covalent surfaces were regenerated with 10 mM glycine–HCl (pH 2.5, GE HealthcaMolecular Dockingre). All experiments were conducted at 25°C.

The 3D structures of the FGFR1 (PDB ID: 4V05) and FGF21 (PDB ID: 5VAQ) proteins were downloaded from the RCSB PDB database (https://www.rcsb.org/), and the 2D structure of PAL was obtained from the PubChem database (https://pubch em.ncbi.nlm.nih.gov). Open Babel 2.4.1 software was used to convert PAL 2D structure results to “mol2” format. PyMOL 2.4.0 software was used for receptor dehydrogenation and ligand separation. The data were then imported into AutoDockTools 1.5.6 for pretreatment, including hydrogenation, to identify the active pocket in the docking region of the receptor. Using AutoDock Vina 1.1.2 software, FGFR1 was molecularly docked with PAL and the C-terminal domain of FGF21 (https://www.rcsb.org/structure/5VAQ). The docking results were analyzed using PyMOL software.

Total RNA was extracted from cells and animal livers using Trizol reagent and then reverse transcribed into cDNA using Oligo-dT primers. qPCR was used to measure the relative mRNA expression levels of GLUT-1 in cells or livers. Relative mRNA expression levels were calculated using the 2^−ΔΔCT method with GAPDH serving as the reference gene. The sequences of the primers used for qPCR were as follows: GAPDH forward, 5′-CCT​TCC​GTG​TTC​CTA​CCC​C-3′ and GAPDH reverse, 5′-GCC​CAG​GAT​GCC​CTT​TAG​TG-3′; GLUT-1 forward, 5′-CAT​CAA​TGC​CCC​CCA​GAA-3′ and GLUT-1 reverse, 5′-AAG​CGG​CCC​AGG​ATC​AG-3′.

Total protein was extracted from cells and liver tissue using radioimmunoprecipitation assay lysis buffer. Protein concentrations were measured using a UV spectrophotometer. Equal amounts of protein were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis, transferred to a polyvinylidene fluoride membrane (Millipore, Massachusetts, Germany), blocked with 5% skimmed milk for 1 h, and then incubated first with antibodies targeting FGFR1, p-FGFR1, FRS2, p-FRS2, AKT, p-AKT, AMPK, p-AMPK, and β-actin overnight at 4°C and then with horseradish peroxidase-labeled secondary antibody for 1 h at room temperature. Protein bands were detected using an enhanced chemiluminescence reagent (Cytiva, Logan, Utah, United States) in the dark.

Statistical analysis was performed using GraphPad Prism 8.0 (GraphPad Software, Inc., San Diego, CA, United States). Data are presented as means ± SD. The t-test was used to analyze differences between two groups. Comparisons among three or more groups were performed using one-way ANOVA followed by Tukey’s post hoc test. A p-value < 0.05 was considered significant.

In mice with STZ-induced diabetes, fasting blood glucose levels were significantly decreased after 4 weeks of treatment with JQ-R (Figure 1A), whereas the phosphorylation level of FGFR1 and that of its intracellular substrate FRS2, as well as the protein levels of GLUT-1, were markedly increased. Moreover, these effects were JQ-R dose-dependent (Figures 1B,C). These results suggested that the FGFR1 pathway may exert hypoglycemic effects and that there may be effective molecules in JQJTT that directly activate FGFR1.

To explore the mechanism via which JQJTT activates the FGFR1 pathway, an SPR assay was used to identify active ingredients in JQJTT that can directly target FGFR1. Figure 2A illustrates the variation in response values during the fishing experiment. The recovery solution was subjected to UPLC–MS/MS, and the resultant chromatogram is shown in Figure 2B. HR-ESI-MS data in positive ion mode showed a peak at m/z 353.1549 [M + Na]⁺ (calcd. 352.1549), indicating that the molecular formula of the detected ingredient was C21H22NO4, which identified it as PAL. To verify that the candidate compound indeed had a strong affinity for FGFR1, an affinity assay was performed using Biacore T200 Evaluation Software Version 3.0. The resulting sensogram and fitting curve indicated that the KD value for the interaction between PAL and FGFR1 was 2.181 × 10^–6 kcal/mol, thus representing a good affinity between the two molecules (Figure 2C).

The interaction energy and binding sites between PAL and FGFR1 were predicted using Molecular Docking analysis. The predicted score for interaction energy (−6.5 kcal/mol) indicated that PAL could act as an activator of FGF21 and therefore also the FRFR1 signaling pathway (Figure 3A). Furthermore, as shown in Figure 3B, the amino acid residue Arg577 of FGFR1 tended to form hydrogen bonds in its interaction with PAL, while the amino acid residues Ser699, Pro702, Tyr701, His717, Asp720, Trp691, and Thr695 in FGFR1 enabled hydrophobic interactions with PAL. Figure 3C shows the binding pattern between FGFR1 and its ligand FGF21. Amino acid residues Ser565, Thr632, Ser602, Arg609, His541, Thr509, and Lys510 in FGFR1 interacted with Ser209, Tyr207, Pro205, Ser200, Ser191, Leu194, and Asp187 in FGF21 through hydrogen bonds. As a whole, these data showed that the site for FGFR1/PAL binding differs from that for the FGFR1/FGF21 combination.

To detect the effects of PAL in vitro, we examined the effect of PAL on the FGFR1 pathway in IR HepG2 cells. As expected, we found that PAL could activate FGFR1 and its downstream substrate FRS2 in a dose-dependent manner via improving IR HepG2 cell glucose absorption and GLUT-1 mRNA expression (Figures 4A,B) without affecting the FGF21 mRNA level (Figure 4C). This effect was not seen in IR model cells. In addition, PAL also promoted the activation of the AKT and AMPK pathways, which are the main pathways that regulate glucose and lipid metabolism downstream of FGFR1 (Figure 4D). These findings demonstrated that PAL has hypoglycemic properties.

In IR HepG2 cells, treatment with the FGFR1 inhibitor PD1730749 significantly inhibited the PAL-mediated promotion of glucose absorption (Figure 5A) and GLUT-1 mRNA expression (Figure 5B). Likewise, the PAL-mediated activation of FGFR1 and its downstream AKT and AMPK signaling pathways were also suppressed after FGFR1 inhibitor treatment (Figure 5C).

As shown in Figure 6A, after 4 weeks of PAL administration, fasting blood glucose levels in mice with STZ-induced diabetes were significantly decreased in each PAL dosage group. On day 28 of administration, blood glucose was measured 1, 2, 4, and 8 h after PAL administration. The results revealed that, in STZ-treated mice, the hypoglycemic effect of each PAL dose was greatest after 2 h of treatment (Figure 6B). On the final day of administration, the IPGTT test indicated that after 2 h of PAL intervention, blood glucose levels and the AUC in each PAL dose group underwent a pronounced reduction (Figure 6C). At the same time, the GLUT-1 mRNA level in the liver showed a marked increase after PAL treatment; however, no significant change was observed in the level of FGF21 mRNA (Figures 6D,E). In addition, western blot results indicated that PAL improved blood glucose levels in STZ-treated mice while simultaneously activating FGFR1 and its downstream AMPK and AKT signaling pathways (Figure 6F), suggesting that PAL also plays a hypoglycemic role in vivo via binding to FGFR1 and activating its downstream signaling.