The aqueous extract of TTF was provided by Huisong Co., Ltd. (Hangzhou, China). TTF contains 15 g Trichosanthes kirilowii Maxim (Cucurbitaceae; Trichosanthis Fructus), 15 g Salvia miltiorrhiza Bunge (Lamiaceae; Salviae miltiorrhizae radix et rhizoma), 10 g Allium macrostemon Bunge (Amaryllidaceae; Bulbus Allii Macrostemonis), 5 g Leech (Hirudinidae; Hirudo), 10 g Acorus gramineus Aiton (Araceae; Acori Tatarinowii Rhizoma), 15 g Curcuma aromatica Salisb (Zingiberaceae; Curcumae Radix), 20 g Poria cocos [Polyporaceae; Wolfiporia cocos (F.A. Wolf) Ryvarden and Gilb], and 10 g Citrus × aurantium L (Rutaceae; Citri Reticulatae Pericarpium) (Table 1). All drugs in TTF were authenticated by Director Jinxia Wang and Minxia Zheng of the First Affiliated Hospital of Zhejiang Chinese Medical University (Hangzhou, China). Voucher specimens were deposited at the Herbarium of Zhejiang University. The procedure used was as follows: After soaking all drugs (100 g) in 1- L distilled water 2 times for 30 min, the mixture was boiled for 1 h. After collecting the filtrate, an additional 1- L distilled water was added and boiled for 1 h, and the filtrate obtained twice was mixed and concentrated to 0.6 mg/ml and 2.25 g/ml, respectively. Hence, crude extracts at low and high concentrations were obtained.

The aqueous extract of TTF (200 µL) was added to 1 ml of 80% methanol, swirled, and shaken for 1 min. High-speed centrifugation was performed at 14,000 rpm for 10 min, and the supernatant was collected. The supernatant was filtered through a 0.22-µm filter membrane. Chromatographic analysis collection and integration of uracil, hypoxanthine, adenosine, lithospermic acid B, lithospermic acid, tanshinone I, and quercetin in TTF were evaluated using Xcilabur4.0 software (Thermo Fisher Scientific, Waltham, MA, United States). A QExactive high-resolution mass spectrometer (Thermo Fisher Scientiic) coupled with an UltiMate 3000 RS (Thermo Fisher Scientific) was employed for chemical identification.

Mass spectrometry analysis was performed in both positive and negative modes under the following parameters: scan range, 50.0–500.0 m/z; ion spray pressure, 3.2 kV (positive); capillary temperature, 300°C. The chromatographic conditions were as follows: chromatographic column, RP-C18 150 × 2.1 mm, Welch; flow rate, 0.300 ml/min; aqueous phase, 0.1% formic acid in aqueous solution (mobile phase A); organic phase, acetonitrile (mobile phase B); needle liquid, methanol; column temperature: 35°C; automatic sampler temperature, 10.0°C.; chromatographic gradient: 0–5 min, 2 %B; 5–10 min, 20 %B; 10–15 min, 50 %B; 15–20 min, 80 %B; 20–25 min, 95 %B; 25–30 min, 2% B.

The study was approved by the Animal Ethics Committee of Zhejiang Chinese Medical University. Sixty male ApoE^-/- mice (6–8 weeks old, 18–22 g) in the C57/BL6 background were obtained from the Department of Laboratory Animal Science at Nanjing University. All mice were kept in a temperature-controlled room on a 12-h light/dark cycle with food and water available ad libitum. Animal experiments were carried out in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 8023, revised 1978).

All mice were adaptively fed for 1 week, after which high-fat diet feeding and TTF intragastric administration were started simultaneously. The mice were randomly divided into four experimental groups (15 mice per group): the control group was fed a common diet and HFD group was fed a high-fat diet (Research Diets, Inc, New Brunswick, NJ, United States; Lot: D12108C). The formulation of the high-fat diet is shown in Supplementary Table S1; the TTF-L and TTF-H groups were fed the HFD together with 0.2 ml low and high concentrations of TTF daily by oral gavage, respectively. After 16 weeks, the mice (n = 8–10 mice per group) were anesthetized with 1% sodium pentobarbital by intraperitoneal injection, and the hearts, aorta, and blood were harvested. The mice were perfused with normal saline through the heart before harvesting the tissues. Finally, the heart with the entire aorta and other tissues was isolated and stored at −80°C for subsequent histological evaluation or gene expression analysis.

Mice were fasted overnight, and then the retro-orbital sinus blood was collected, allowed to coagulate, and centrifuged at 1,000 ×g (RCF) at room temperature to isolate the serum. The serum was immediately frozen and stored at −80°C. The serum lipid profile, including cholesterol (CHOL), triglycerides (TG), low-density lipoprotein-cholesterol (LDL-C), and high-density lipoprotein-cholesterol (HDL-C), was measured at the Zhejiang Chinese Medical University with an AU 400 fully automated chemistry analyzer (Olympus, Tokyo, Japan) using the enzymatic-colorimetric method. Serum tumor necrosis factor (TNF)-α, interleukin (IL)-6, and IL-1β levels were evaluated by enzyme-linked immunosorbent assay according to the manufacturer’s instructions.

The en face pinned aortas were stained with Oil-Red-O. Images of the aortas were captured using a stereoscope (Nikon, Tokyo, Japan). The proximal aortas attached to the heart were embedded in OCT compound and frozen at −80°C. The aortic root was serially sectioned into 7-μm sections from the site where the aortic valve appeared. A set of three consecutive sections was stained with Oil-Red-O and Masson’s trichrome for morphological analysis of atherosclerotic plaques. Macrophages and the expression of Arg1 in the aortic root were detected by immunofluorescence staining of sequential sections with rat anti–monocyte macrophage (CD68) monoclonal antibody (1 μg/ml, abcam, ab237968) and rabbit polyclonal anti-Arg1 (1:1,000, CST, 93668S) at 4°C overnight, followed by the corresponding secondary antibodies conjugated with Alexa Fluor® 488 and 647 (1:5,000, abcam, ab201844 and 1:5,000, CST, 43279S) for fluorescence detection. Images were captured using a Nikon A1R digital camera (Nikon). The sections were quantitatively analyzed using Image-Pro Plus 6.0 software (Media Cybernetics, Rockville, MD, United States).

Primary peritoneal macrophages (PMs) were isolated from successful modeling ApoE^-/- mice in the control, HFD, TTF-L, and TTF-H groups. Briefly, the mice were

Sacrificed by cervical dislocation, 4 ml of ice-cold RPMI medium was injected into the peritoneal cavity, the peritoneum was gently massaged for 1 min, the RPMI medium was collected and centrifuged at 800 ×g (RCF) for 5 min, and the PMs settled at the bottom of the test tube. The PMs were resuspended and cultured at 5% CO₂ and 37°C for 2 h in RPMI containing 10% fetal bovine serum, nonadherent cells were removed, and the cells were cultured for subsequent assays.

Ten microliters RPMI were added to each well of a 96-well plate, and then 90 μL PMs were seeded into the plates at a density of 5 × 10⁴ cells/well. After culturing for 2 h, nonadherent cells were removed and 90 μL RPMI containing TTF was added at gradient of concentrations for 24 h. Next, 10 μL CCK-8 reagent was added. The cells were placed in an incubator in the dark for 2 h. The optical density of the medium was measured using a microplate reader at a wavelength of 450 nm.

PMs isolated from ApoE^-/- mice were incubated with 0.5 mg/ml TTF for 6 h at 37°C and then incubated with 80 μg/ml oxidized LDL (ox-LDL; Yiyuan Biotechnologies, Guangzhou, China) for 24 h at 37°C. Total RNA was isolated from PM cells using TRIzol reagent (Invitrogen, Carlsbad, CA, United States). Total RNA (2 μg) was reverse-transcribed to cDNA using M-MLV transcriptase (Thermo Fisher Scientific) according to the manufacturer’s instructions. Real-time polymerase chain reaction was performed on an ABI 7500 System cycler (Applied Biosystems, Foster City, CA, United States) using SYBR Green PCR Master Mix (Transgen Biotic). The expression of six genes (Fizz1, Arg1, Mrc, TNF-α, IL-1β, and IL-6) was normalized against the expression of 18S rRNA. Data were analyzed using the 2^-ΔCt method. The primers used are shown in Table 2.

PMs isolated from ApoE^-/- mice were incubated with 0.5 mg/ml TTF for 6 h at 37°C and then incubated with 80 μg/ml oxidized LDL (ox-LDL; Yiyuan Biotechnologies, Guangzhou, China) for 24 h at 37°C. The RNA-seq library was prepared for sequencing using standard Illumina protocols. Total RNA samples from ApoE^-/- PMs were isolated using TRIzol reagent. Library construction and sequencing were performed using BGI (Shenzhen, China). For data analysis, base calling was performed using SOAPnuke (v1.5.2). Clean reads were aligned to the genome using HISAT2 v2.0.5. Differential expression was determined using RSEM (v1.2.8), and the significance of differential gene expression was defined by DEGseq according to the combination of the absolute value of log2-fold-change ≥ 2 and p value ≤0.001. Gene Ontology and pathway annotation and enrichment analyses were based on the GO database (http://www.geneontology.org/) and Kyoto Encyclopedia of Genes and Genomes pathway database (http://www.genome.jp/kegg/), respectively. The R package heatmap was used for hierarchical cluster analysis of gene expression patterns.

Equal amounts of cells were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and western blotting analysis was performed using the mouse anti–CD68 antibody (1:1,000, abcam, ab955), rabbit anti-Arg1 antibody (1:1,000, CST, 93668S), rabbit anti-TNF-α antibody (1:1,000, CST, 3707S), mouse anti-PPARγ antibody (1:1,000, CST, 95128S), rabbit anti-extracellular signal-regulated kinase (ERK) antibody (1:2000, CST, 4370S), rabbit anti-phosphorylated (p)-ERK antibody (1:1,000, CST, 4695S), rabbit anti-AKT antibody (1:1,000,CST, 4685S), rabbit anti-p-AKT antibody (1:1,000, CST,4060S), and rabbit anti-β-actin antibody (1:2000, diagbio, db10001). Quantification of the band intensity was performed using ImageJ software 1.49v (NIH, Bethesda, MD, United States), and the values were normalized to the levels of β-actin.

Statistical analyses were performed using SPSS version 27.0 software (SPSS, Inc, Chicago, IL, United States). All results are presented as the mean ± standard error of the mean. Statistical significance was determined by t-test or one-way analysis of variance for data with a normal distribution and by Kruskal–Wallis test for non-normally distributed data or small samples. Statistical significance was set at p < 0.05. All cell experiments were performed at least 3 times.

Seven chemical compositions were identified in TTF using high-performance liquid chromatography: hypoxanthine (observed m/z: 110.035, retention time (RT): 1.51), uracil (observed m/z: 113.034, RT: 1.57); adenosine (observed m/z: 268.104, RT: 7.33); quercetin (observed m/z: 153.018, RT: 10.13), lithospermic acid B (observed m/z: 719.160, RT: 13.59), lithospermic acid (observed m/z: 539.119, RT: 16.68); and tanshinone I (observed m/z: 279.099, RT: 17.73) (Figure 1). Our results showed that the quality of TTF met the pharmacopeia requirements, which provided a reliable and controllable sample for subsequent experiments.

To investigate the role of TTF in atherosclerotic development, we established an AS model with ApoE^-/- mice fed a HFD for 16 weeks. Our experimental animal scheme is shown in Figure 2A. The body weight of mice in the TTF-H group was decreased compared with the control group and HFD group from the 8th week of modeling (Figure 2B). The serum lipid levels of mice in the four groups were measured; CHOL and TG levels in the HFD group were significantly higher than those in the control group (p < 0.001, p < 0.01). In the TTF-L group, the CHOL and TG levels piwere decreased compared to that in the HFD group, but the difference was not significant. However, the CHOL and TG levels in the TTF-H group were significantly lower than those in the HFD group (p < 0.01) (Figure 2C). The four groups showed no significant difference in LDL-C and HDL-C levels (all p > 0.1). Serum IL-6 and TNF-α levels were significantly higher in the HFD group than in the control group (p < 0.01). The levels of IL-1β, IL-6, and TNF-α in the TTF-H group were much lower than those in the HFD group (p < 0.01) (Figures 2D–F).

We assessed the plaque area and stability using Oil Red O and Masson’s trichrome staining. The vivisection of the aortic arch, brachiocephalic trunk, left common carotid artery, and left subclavian artery showed that the HFD significantly increased plaque area in ApoE^-/- mice, whereas TTF significantly decreased the plaque area in these mice (Figure 3A). Oil Red O staining of the en face aorta showed that the plaque area in the HFD group was significantly higher than that in the control group (p < 0.01). The plaque areas in the TTF-L and TTF-H groups were also higher than those in the control group (p < 0.01, p < 0.05); however, compared with the HFD group, they were significantly decreased (p < 0.05, p < 0.01) (Figures 3B,C). The results of Oil Red O staining of the aortic valve were consistent with those of the en face (Figures 4A,B). Masson’s trichrome staining was used to detect collagen deposition in the plaque region of the aortic valve. The collagen positive areas in the HFD, TTF-L, and TTF-H groups were significantly higher than that in the control group (p < 0.01, p < 0.05, p < 0.05), but they were slightly decreased in the TTF-L and TTF-H groups compared with the HFD group with no significant difference (Figures 4C,D). The staining results suggest that TTF can significantly reduce the lipids and slightly reduce the collagen content in atherosclerotic plaque, which is helpful to delay the progression of plaque (Rekhter, 1999; Baardman, et al., 2020).

To investigate whether TTF regulated alternative macrophage activation in the atherosclerotic plaques of ApoE^-/- mice fed a HFD, we conducted immunofluorescence staining and analysis of Arg1 on aortic valve sections. Expression of the total macrophage marker CD68 in the HFD group was significantly higher, whereas expression of the alternative macrophage marker Arg1 was lower than that in the control group. In addition, compared with the HFD group, expression of CD68 in the TTF-L and TTF-H groups significantly decreased, and expression of Arg1 was increased (Figure 5A). The quantitative analysis of relative Arg1/CD68 positive area showed that the proportion of Arg1 positive macrophages in HFD group was significantly less than that in control group, while the proportion of arg1 positive macrophages in TTF-L and TTF-H groups were significantly higher than that in HFD group (Figure 5B). Moreover, to detect the expression levels of Macrophage markers in the peritoneal macrophages of ApoE^-/- mice by quantitative PCR assay, we determined the optimal drug concentration using the CCK-8 assay. The results showed that TTF (0.5 mg/ml) was an appropriate dosage for PMs (Figures 6A,B). Furthermore, quantitative PCR showed that TTF promoted mRNA expression of the alternative macrophage markers Fizz1, Arg1, and Mrc and decreased expression of the M1 macrophage markers TNF-α, IL-1β, and IL-6 following stimulation with ox-LDL (Figures 6C–H). Moreover, western blotting revealed significant up-regulation of the alternative macrophage-associated gene (Arg1) and down-regulation of the M1 macrophage-associated gene (TNF-α) in PMs treated with 0.25 and 0.5 mg/ml TTF (Figure 6I). Analysis of the grayscale image showed that Arg1/CD68 levels were significantly higher than TNF-α/CD68 levels in the TTF-L and TTF-H groups (Figure 6J).

Based on the results described above, TTF regulates M2 polarization in atherosclerotic plaques, but the underlying mechanism was unclear. To determine the specific effect of TTF on the RNA expression of macrophages and related signaling pathways in ApoE^-/- mice, we collected PMs from ApoE^-/- mice pretreated with TTF and stimulated with ox-LDL. RNA was isolated and subjected to RNA sequencing. Twelve samples were tested, with three samples from each group. Principal component analysis revealed overall transcriptomic similarity among the four groups (Figure 7A). Based on the gene expression levels of each sample, we detected differentially expressed genes between PBS- and TTF-pretreated macrophages induced by ox-LDL using a volcano plot (Figure 7B). The hierarchical clustering of differentially expressed genes are shown in Figure 7C and the results of pathway classification are shown in Figure 7D. Vital genes involved in alternative macrophage activation between TTF + ox-LDL group and PBS + ox-LDL group are shown as a heatmap (Figure 7E). Our results indicate that PPARγ, AMPK, and inflammatory mediator-related signaling pathways may play critical roles in the effects of TTF on the RNA expression of macrophages in ApoE^-/- mice.

According to the results of RNA sequencing, to detect the protein expression of key molecules in related signal pathways of TTF affecting macrophage polarization, we performed immunoblotting to determine the protein expression and phosphorylation levels of PPARγ, ERK1/2, and AKT in PMs. We found that 0.5 mg/ml TTF did not affect the expression of the tested proteins in PMs without ox-LDL stimulation. However, upon stimulation with 80 μg/ml ox-LDL, TTF decreased the phosphorylation levels of AKT and ERK and increased the expression of PPARγ significantly (Figures 8A–D). The regulation effect on PPARγ and AKT/ERK was related to the concentration of TTF, as 0.5 mg/ml TTF (TTF-H) had a stronger regulatory effect than 0.25 mg/ml TTF (TTF-L) (Figures 8E–H). This effect was remarkably reversed by an AKT activator (recilisib), ERK activator (pamoic acid disodium, PAD), or PPARγ inhibitor (mifobate). The quantitative PCR results revealed decreased relative mRNA levels of Fizz1, Arg1, and Mrc and increased relative mRNA levels of TNF-α, IL-1β, and IL-6 in PMs pretreated with TTF and 30 μM AKT activator (recilisib), 25 nM ERK activator (PAD), or 50 μM PPARγ inhibitor (mifobate) following stimulation with ox-LDL (Figure 9).