ZishenWan reduces obesity in db/db mice by inhibiting p21 expression in adipose tissues

Abstract

Background

Obesity is a major risk factor for metabolic disorders, including type 2 diabetes, metabolic syndrome, and cardiovascular diseases. Recent studies suggest that adipocyte cell cycle regulation is linked to metabolic dysfunction, making cell cycle regulatory factors potential therapeutic targets. ZishenWan (ZSW), a traditional Chinese medicinal formula composed of three herbs, has shown therapeutic potential against diabetes and obesity. This study aimed to investigate the mechanisms underlying the weight-regulatory effects of ZSW.

Methods

ZSW was orally administered to db/db mice for 5 weeks, with metformin used as a positive control. Body weight and fasting blood glucose levels were monitored weekly, and an oral glucose tolerance test was conducted at the end of the intervention. Insulin levels and serum lipid profiles were also assessed. Histological analysis was performed on adipose tissue, and epididymal adipose tissue (EAT) samples were collected for transcriptomic analysis. The findings from transcriptomic analysis were further validated using qPCR, immunohistochemistry, and western blotting.

Results

ZSW significantly reduced body weight, inhibited adipocyte hypertrophy, and improved insulin sensitivity and insulin secretion in db/db mice without affecting liver or kidney function. Transcriptomic analysis revealed that ZSW altered the expression of genes involved in lipid metabolism and inhibited p21 expression. Additionally, ZSW activated the PI3K/AKT/FoxO1 signaling pathway, resulting in reduced p21 expression, increased apoptosis, and enhanced UCP1 expression in EAT.

Conclusions

ZSW may improve obesity-related metabolic disorders by activating the PI3K/AKT pathway and downregulating p21 expression in EAT. These effects promote adipocyte apoptosis, enhance thermogenesis, inhibit adipocyte hypertrophy, and improve insulin sensitivity.

Graphical abstract

1. Introduction

Obesity, often accompanied by type 2 diabetes mellitus (T2DM), nonalcoholic fatty liver disease, cardiovascular diseases, and dyslipidemia, is among the most prevalent non-communicable diseases and poses a significant global health challenge.^1,2 The World Health Organization classifies overweight as body mass index (BMI) ≥ 25 kg/m²,³ and approximately 2 billion individuals worldwide are estimated to be overweight or obese.⁴ Weight gain typically results from an energy imbalance caused by excessive nutrient intake and reduced energy expenditure. Adipocyte hypertrophy, commonly seen in obese humans and mice, is linked to metabolic diseases.⁵ Under over-nutritional conditions, adipocytes expand to store excess energy in lipid, thus exerting a nutrient-buffering function. However, this capacity is limited, and once exceeded, ectopic lipid accumulation and impaired adipocyte function can occur.⁶

Excessive adipocyte enlargement is closely associated with insulin resistance, inflammatory cell infiltration, and oxidative stress, all of which are pathological changes commonly observed in metabolic diseases.^{7, 8, 9} These observations link metabolic disorders to adipocyte growth, which is closely tied to the cell cycle and adipocyte differentiation.^10,11 The cell cycle of adipocytes is governed by several cell cycle regulatory genes, including those coding for cyclin, cyclin-dependent kinases (CDKs), and cyclin-dependent kinase inhibitors (CKIs).^12,13 Indeed, the expression of these proteins changes during adipogenesis and adipocyte differentiation, and regulation of cell cycle progression can, in turn, influence metabolic status through mechanisms like adipocyte senescence and inflammatory cell infiltration.^14,15 Therefore, cell cycle regulatory factors could serve as potential therapeutic targets for metabolic disorders.

Traditional Chinese medicine (TCM) plays an important role in complementary therapies for various chronic conditions. ZishenWan (ZSW), a traditional Chinese formula composed of Rhizoma Anemarrhenae (知母, zhī mǔ), Cortex Phellodendri (黄柏, huáng bò), and Cinnamomum cassia (肉桂, ròu guì), is frequently used for managing diabetes. ZSW, also known as TongguanWan, was described by Li Dongyuan during the Yuan Dynasty. According to TCM theory, this formular nourishes yin, clears heat, and could relieve symptoms such as polydipsia, polyphagia, and polyuria in diabetes. In modern clinical practice, this herb combination is also used in decoctions for managing T2DM.

Our previous studies examined the effects of ZSW in a genetic mouse model of diabetes (db/db mice).^16,17 ZSW was found to reduce body weight, improve insulin sensitivity, and ameliorate metabolic disorders in mice.^16,18 ZSW enhanced glycolysis and glycogen synthesis in the liver while reducing gluconeogenesis.¹⁶ These findings provided experimental support for the clinical use of ZSW. However, the mechanism underlying its weight-reducing effect remains unclear.

This study aimed to investigate the regulatory effects of ZSW on adipocyte apoptosis in db/db mice, focusing on potential underlying molecular mechanisms. Understanding how ZSW affects adipocyte apoptosis could reveal new therapeutic strategies for managing obesity and its associated metabolic complications. Based on transcriptomic analysis of epididymal adipose tissue (EAT), we explored the influence of ZSW on adipocyte apoptosis and sought to identify key molecular pathways involved.

2. Materials and methods

2.1. Chemicals and reagents

Extracts of the three herbs ZSW, Rhizoma Anemarrhenae, Cortex Phellodendri, and C. cassia, were purchased from Zhongjing Wanxi Pharmaceuticals (Nanyang, Henan, China). The purchased herbal extracts were aqueous extracts in powdered solid form. The three extracts were mixed at a ratio of Rhizoma Anemarrhenae: Cortex Phellodendri: Cinnamomum cassia=10:10:1, based on the weight of the raw materials. The metformin hydrochloride tablets were manufactured by Sino-American Shanghai Squibb Pharmaceuticals (Shanghai, China). All reagents were dissolved in double-distilled water before gavage.

2.2. High-performance liquid chromatography

To assess the quality of the extracted ZSW, six key constituents of ZSW, namely, mangiferin (1), timosaponin B-II (2), berberine (3), phellodendrine (4), cinnamaldehyde (5), and cinnamic acid (6), were quantitatively measured using high-performance liquid chromatography (HPLC) (Alliance HPLC, Waters Corporation, MA, USA). Chemicals 1, 3, 4, 5, and 6 were quantified using HPLC-UV with an absorbance at 280 nm (2998 PDA, Waters Corporation, US), whereas chemical 2 was quantified using an HPLC-evaporative light scattering detector (ELSD) (Allchrom Model 6100 ELSD, Alltech Technology Ltd., BC, Canada). The measurements were repeated four times.

2.3. Animal experiments

Twenty-one male C57BL/KsJ-db/db mice (6 weeks old, 33 ± 3 g) and seven sex- and age-matched C57BL/KsJ-wt/wt mice were purchased from Beijing GemPharmatech. The mice were housed at the Animal Experiment Center of Beijing University of Chinese Medicine. All animal experiments were approved by the Animal Care and Ethics Committee of Beijing University of Chinese Medicine (approval no. BUCM-2023090409-3111). During treatment, the mice were housed at 24 ± 2 °C under a 12-h light–dark cycle and had ad libitum access to food and water.

After 1 week of adaptive feeding, the db/db mice were randomly divided into three groups: ZSW group, metformin group, and model group. The wt/wt mice were used as normal controls. All the drugs were administered via oral gavage. Mice in the ZSW group received 3.3 g/kg ZSW, those in the metformin group received 200 mg/kg metformin hydrochloride, and those in the model and control groups received daily distilled water. The doses of ZSW and metformin were selected based on previous studies.¹⁶

The treatment lasted for five weeks, during which body weight and overnight fasting blood glucose levels were measured weekly. Blood glucose levels were measured in drops obtained from the tail vein using a portable glucometer (Glucocard 01-mini, Arkray, Kyoto, Japan).

2.4. Oral glucose tolerance test and serum insulin measurement

The oral glucose tolerance test (OGTT) was performed at the end of the treatment. Before the test, the mice were fasted for 12 h and administered 2 g/kg glucose by gavage. Blood was drawn from the tail vein before and at 30, 60, and 120 min after glucose intake. Furthermore, approximately 100 μL of blood was collected from the orbital vein before and 30 min after glucose intake and centrifuged at 3000 rpm for 10 min to obtain serum. Serum insulin levels were measured using a commercially available ELISA kit (Cat.80-INSMSU-E01, ALPCO Diagnostics, NH, USA) according to the manufacturer's instructions.

The area under the blood glucose-time curve (AUC) was calculated as follows:

$$\text{AUC}=0.25\times \mathrm{G}0+0.5\times \mathrm{G}30+0.75\times \mathrm{G}60+0.5\times \mathrm{G}120\text{.}$$

where G0, G30, G60, and G120 represent blood glucose levels before and 30, 60, and 120 min after glucose intake, respectively.

Homeostatic model assessment of insulin resistance (HOMA-IR) scores were calculated as follows:

HOMA-IR = fasting blood glucose × fasting insulin/22.5

2.5. Biochemical measurements and histological observation

After five weeks of treatment, the mice were anesthetized by isoflurane inhalation. The body length from the tip of the nose to the tip of the tail was measured, and the BMI of each mouse was calculated. Blood samples were collected, and serum was acquired as described above. Serum total cholesterol (TC), triglyceride (TG), high-density lipoprotein (HDL), low-density lipoprotein (LDL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), blood urea nitrogen (BUN), and creatinine (CREA) levels were measured using a chemistry analyzer (Cat. BS-420, Shenzhen Mindray Bio-Medical Electronics, Shenzhen, China).

The liver and adipose tissues (including epididymal, mesenteric, perirenal, subcutaneous, and brown fat) were removed, fixed in 4 % paraformaldehyde solution, and subsequently embedded in paraffin. Weights of the liver, brown adipose tissue, and left-sided EAT were determined. Hematoxylin and eosin (H&E) staining was performed for histological analysis.

For transmission electron microscopic (TEM) observation of β cells, pancreatic tissue was removed and fixed in 2.5 % glutaraldehyde (Cat. P1126, Solarbio Life Sciences, Beijing, China) and preserved at 4 °C. Tissue samples were post-fixed in 1 % osmium tetroxide for 2 h and washed three times in 0.1 M PBS. After dehydration, samples were embedded in 812 epoxy resin (Cat. 90529-77-4, SPI Supplies, PA, USA), polymerized for 48 h at 60 °C, sliced, and stained with 2 % uranium acetate and 2.6 % lead citrate. Finally, the samples were observed under a 120 kV Hitachi HT7800 TEM (Cat. HT7800, Hitachi High-Tech, Tokyo, Japan).

2.6. Transcriptome and bioinformatics analysis

The EAT of the mice was removed and snap-frozen in liquid nitrogen. Four samples were obtained from the normal, model, and ZSW groups for transcriptome analysis. Total RNA was extracted using the mirVana miRNA Isolation Kit (Cat. AM1560, Life Technologies, CA, USA). Libraries were constructed using TruSeq Stranded mRNA Sample Prep Kit (Cat. RS-122-2101, Illumina, San Diego, CA, USA) according to the manufacturer's instructions. mRNA sequencing was performed using the Illumina HiSeq X Ten platform.

Raw data from mRNA sequencing were processed using Trimmomatic,¹⁹ and clean reads were mapped to the mouse reference genome (NCBI database: GRCm39) using HISAT2.²⁰ Gene expression levels were quantified based on fragments per kb per million reads (FPKM)²¹ and differentially expressed genes (DEGs) (p < 0.05, fold change >1.5) were identified using DESeq2. Functional enrichment analysis of DEGs was performed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases.

2.7. Terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling assay

Apoptosis in EAT was examined using a terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling (TUNEL) assay conducted with a commercially available labeling (Cat. G1502, Servicebio Technology, Wuhan, China). Briefly, the paraffin-embedded adipose tissue was sliced, deparaffinized, rehydrated, and subjected to antigen retrieval using protease K. After treatment with 0.1 % Triton (Cat. G1204, Servicebio Technology), the slices were reacted with the TdT enzyme and dUTP at 37 °C for 1 h, followed by DAPI nuclear counterstaining. Tetramethylrhodamine-5-dUTP red fluorescence was used as a readout for DNA fragmentation.

2.8. Immunofluorescence and immunohistochemistry

Immunofluorescence staining was conducted based on tyramide signal amplification (TSA). F4/80 and p21 protein expression levels were determined using immunohistochemistry. Briefly, the samples were dewaxed and antigen retrieval was performed in a microwave oven. The samples were blocked with 3 % bovine serum albumin for 30 min and subsequently incubated with a primary antibody targeting p21 (Cat. 188224, Abcam) at 4 °C overnight. HRP-conjugated secondary antibodies (Cat. GB23303, Servicebio Technology) and iF488-Tyramide (Cat. G1231, Servicebio Technology) were added sequentially. Antigen retrieval was performed again, followed by incubation with a primary antibody against F4/80 (Cat. A23788, ABclonal Technology, Wuhan, China), followed by incubation with an HRP-conjugated secondary antibody and CY3-Tyramide (Cat. G1223, Servicebio Technology). Nuclei were counterstained with DAPI. Slides were observed and imaged using a fluorescence microscope.

Uncoupling protein 1 (UCP1) was detected via immunohistochemistry using a commercially available kit (Cat. PV-6000, ZSGB-Bio, Beijing, China). Paraffin-embedded EAT and brown adipose tissue were sectioned and incubated with primary antibodies against UCP1 (Cat. ab234430, Abcam) at 4 °C overnight, and then with the HRP-conjugated secondary antibody provided in the kit. The tissue samples were then stained with DAB for 30 s, followed by nuclear counterstaining with hematoxylin for 10 s. Images were quantified as the mean fluorescence intensity, and the immunohistochemical staining intensity was measured as the mean optical density using ImageJ software.

2.9. Reverse transcription quantitative polymerase chain reaction

For quantitative polymerase chain reaction (qPCR) analysis, total RNA was extracted from EAT using the mirVana RNA Isolation Kit (Cat. AM1561, Life Technologies). The extracted RNA was reverse-transcribed into cDNA using TransScript All-in-One First-Strand cDNA Synthesis SuperMix for qPCR (Cat. AT341-02, TransGen Biotech, Beijing, China). qPCR was performed using PerfectStart Green qPCR SuperMix (Cat. AQ601, TransGen Biotech) using a LightCycler 480 system (Cat. LC480II, Roche, Swiss). The primers used for qPCR (Gapdh, Pik3ca, Pik3cb, Pik3cd, Pik3cg, Akt1, Akt2, Vegfa, Pdgfd, Irs1, Irs2, and Cdkn1a) are listed in Supplementary Table 1. The relative expression levels of target mRNAs were calculated using the 2^−ΔΔCT method with Gapdh serving as an internal control.

2.10. Western blot

Western blot analysis was performed to examine the protein expression of p21, phosphorylated phosphoinositide 3-kinase (p-PI3K), phosphorylated protein kinase B (p-AKT), caspase-3, forkhead box protein O1 (FoxO1), and p-FoxO1 in the EAT. Total protein was extracted from samples using a commercial kit (Cat. KGP250, KeyGen Biotech, Nanjing, China). Protein was separated by 10 % SDS–PAGE, transferred to polyvinylidene fluoride membranes, blocked, and incubated with diluted primary antibodies (Supplementary Table 2) at 4 °C overnight. The following day, the membranes were incubated with HRP-conjugated secondary antibody (Cat. SA00001-2, Proteintech Group, Wuhan, China) for 1 h at room temperature, and subsequently detected using an ECL Super Kit (Cat. RM02867, ABclonal Technology) using a gel imaging system (Cat. ChemiDoc XRS+, Bio-Rad Laboratories, CA, USA). Relative protein expression was calculated using Image Lab software (Version 5.2.1, Bio-Rad Laboratories) with β-actin serving as an internal control.

2.11. Statistical analysis

Data were analyzed with IBM SPSS Statistics software (v.23.0) and are presented as means ± S.E. Data plotting was performed in GraphPad Prism (v.9.5.1). Differences among groups were analyzed using one-way analysis of variance (ANOVA) followed by Fisher's least significant difference (LSD) or Dunnett's test.

3. Results

3.1. Quantitative analysis of the main chemicals in ZSW

To assess the quality of ZSW, six chemicals were quantified: mangiferin and timosaponin B-II from Rhizoma Anemarrhenae, berberine and phellodendrine from Cortex Phellodendri, and cinnamaldehyde and cinnamic acid from C. cassia.^{22, 23, 24, 25, 26, 27} The chemicals were quantified by HPLC. As the 280 nm UV absorbance for timosaponin B-II was low, the timosaponin B-II content was assessed using an ELSD (Fig. 1). All six chemicals were detected in the ZSW extracts, and consistent results were obtained across different batches (Table 1).

Fig. 1.

Chromatographic fingerprint for the identification of six compounds in ZishenWan. (a) Chromatographs from the HPLC-UV analysis and (b) the HPLC-ELSD analysis. Mix: a mixture of five standard references. B-II: timosaponin B-II standard reference. The peaks represent (1) mangiferin, (2) timosaponin B-II, (3) berberine, (4) phellodendrine, (5) cinnamaldehyde, and (6) cinnamic acid.

Table 1.

Percent content (w/w) of six main compounds in ZishenWan.

Compound	Batch				Average content
	1	2	3	4
Mangiferin	0.44	0.41	0.43	0.42	0.425
Timosaponin B-II	2.40	2.38	2.39	2.46	2.408
Berberine	4.03	4.02	4.11	4.09	4.063
Phellodendrine	1.18	1.19	1.20	1.19	1.190
Cinnamaldehyde	0.0702	0.0707	0.0708	0.0680	0.070
Cinnamic acid	0.0254	0.0255	0.0254	0.0247	0.025

3.2. ZSW decreased body weight and suppressed adipocyte enlargement in db/db mice

Compared with normal controls, db/db mice showed significantly elevated body weights at the beginning of the experiment (6-week-old). ZSW treatment effectively reduced weight gain in db/db mice (Fig. 2a). At the end of the intervention, the body mass index (BMI) of the ZSW group was significantly lower compared to the model group (Fig. 2b). Additionally, we weighed the EAT, liver, and brown fat and performed hematoxylin and eosin (H&E) staining of adipose tissue. ZSW significantly reduced epididymal adipocyte hypertrophy and liver weight, without affecting brown fat mass (Fig. 2c–f). In comparison, metformin treatment reduced liver weight but did not significantly impact body weight, BMI, or adipose tissue weight (Fig. 2a–e). H&E staining showed that ZSW prevented the pathological enlargement of adipocytes seen in the model group, whereas metformin had no effect (Fig. 2f). ZSW treatment also increased serum HDL levels and decreased LDL levels, with a non-significant reduction in total cholesterol (TC) content (Fig. 2g–j). Importantly, ZSW did not impair liver or kidney function in db/db mice (Supplementary Fig. 1).

Fig. 2.

ZishenWan decreased body weight and suppressed adipocyte enlargement in db/db mice. (a) The body weight of mice during the intervention (n = 7). (b) The body mass index (BMI) of mice at week 5 (n = 7). (c–e) The weight of epididymal adipose tissue (EAT), liver, and brown adipose tissue of mice (n = 7). (f) Histological observation of hematoxylin and eosin (H&E)-stained adipose tissue at × 200 magnification. (g–j) The serum lipid profile of mice (n=7). ∗p < 0.05, ∗∗p < 0.01, compared with the model group. Nor: normal control group; Mod: model group; Met: metformin group; ZSW: ZishenWan group.

3.3. ZSW improved insulin sensitivity and insulin secretion in db/db mice

Compared to wt/wt mice, db/db mice showed elevated fasting blood glucose levels throughout the intervention period. Both ZSW and metformin decreased hyperglycemia and improved glucose tolerance in db/db mice (Fig. 3a–c). Notably, both treatments also ameliorated hyperinsulinemia in db/db mice, which reflects insulin resistance in insulin-targeted organs (Fig. 3d), and significantly decreased HOMA-IR scores (Fig. 3e). H&E staining of the pancreatic islets of mice showed that compared with the normal controls, islet morphology was disrupted (indicated by a skewed islet shape), and lipid droplets accumulated in the model group. The number of lipid droplets did not decrease in metformin-treated groups. However, ZSW treatment rescued the aberrant islet morphology and diminished the lipid droplet size in the islets observed in the model group (Fig. 3f). Although basal insulin levels were high in the model mice, glucose-stimulated insulin secretion was impaired. ZSW treatment significantly increased insulin secretion following glucose administration (Fig. 3g). TEM of β cells showed that mitochondria in the model group were swollen and distorted, with fewer secretory vesicles. ZSW treatment partially restored mitochondrial structure and increased the number of secretory vesicles (Fig. 3h).

Fig. 3.

ZishenWan improved insulin sensitivity in db/db mice. (a) Fasting blood glucose levels of mice during the intervention (n = 7). (b) Blood glucose levels of mice in the oral glucose tolerance test (OGTT) (n = 7). (c) Area under the blood glucose-time curve for the OGTT (n = 7). (d) Fasting serum insulin levels (n = 5). (e) HOMA-IR scores (n = 5). (f) Images of H&E staining of pancreatic islets of mice at × 200 magnification. (g) Insulin levels at 0 and 30 min in the OGTT (n = 6). (h) Transmission electron microscopic observation of pancreatic β cells at × 3000 and × 8000 magnification. ∗p < 0.05, ∗∗p < 0.01, compared with the model group. Nor: normal control group; Mod: model group; Met: metformin group; ZSW: ZishenWan group.

3.4. Epididymal adipose tissue transcriptomics analysis

To further explore the possible mechanisms underlying the therapeutic effects of ZSW in db/db mice, we analyzed the EAT transcriptome (n = 4). One sample from the model group was excluded from the final analysis because of its high deviation from the other samples. Principal component analysis showed that the transcriptomic profile of db/db mice markedly differed from that of normal controls, as evidenced by the high number of identified DEGs (Fig. 4a–c). KEGG pathway and GO enrichment analyses revealed that the DEGs were mainly associated with lipid metabolism, including cholesterol and lipoprotein synthesis and breakdown, potentially explaining the effects on body weight and serum lipid profile (Fig. 4d and e). PI3K/AKT signaling regulates critical steps in cell growth, proliferation, and insulin signaling. We identified several DEGs involved in this pathway. Factors upstream of PI3K/AKT were upregulated, whereas p21—a key cell cycle regulator—was significantly downregulated (Fig. 4f).

Fig. 4.

Results of the epididymal adipose tissue transcriptomics analysis. (a) Principal component analysis of mRNA expression. (b) Summary of the differentially expressed genes (DEGs). (c) Venn diagram of the differentially expressed genes. (d) KEGG metabolic pathway enrichment for the DEGs between the ZSW and model groups. (e) GO biological process enrichment for the DEGs between the ZSW and model groups. (f) Enrichment of the DEGs in the KEGG PI3K/AKT signaling pathway map (red: upregulated, dark green: downregulated).

3.5. ZSW reduced p21 expression in epididymal adipose tissue of db/db mice

The transcriptomic results showed that the expression of p21 in the ZSW group was significantly downregulated compared to that in the model group. p21 belongs to the CKI family and regulates cell cycling and apoptosis,^28,29 both of which are highly related to the size of adipocytes. It is also involved in regulating the metabolic status of the body, including insulin sensitivity and lipid metabolism.³⁰ High expression of p21 is associated with inflammation and lipid accumulation.^31,32 We subsequently validated the expression of the p21-encoding gene (Cdkn1a), which was shown to be reduced in the transcriptomic analysis, using qPCR and western blotting. The p21 levels in the EAT were significantly higher in the model group than those in the normal control group, whereas ZSW treatment significantly attenuated this tendency (Fig. 5a and b). The expression of p21 and F4/80, which are mature macrophage markers in mice, was assessed using immunofluorescence. The results showed that F4/80 expression was induced in areas with strong p21 expression in the model group. However, ZSW treatment reduced the level of F4/80, although not significantly (Fig. 5c).

Fig. 5.

ZishenWan reduced p21 expression in epididymal adipose tissue of db/db mice. (a) The expression of Cdkn1a in epididymal adipose tissue (EAT) as detected by qPCR (n = 4). (b) The protein expression of p21 in EAT (n=3). (c) Immunofluorescence images of EAT (p21: green, F4/80: red, nucleus: blue) captured at × 200 magnification. Immunofluorescence was quantified as mean fluorescence intensity (n = 5). ∗p < 0.05, ∗∗p < 0.01, compared with the model group. Nor: normal control group; Mod: model group; ZSW: Zishen Wan group.

3.6. ZSW regulated p21 expression by activating the PI3K/AKT pathway

The transcription of p21 is regulated by PI3K/AKT signaling. Thus, we evaluated the expression levels of PI3K/AKT-encoding genes, as well as the expression of the activated forms of these two proteins. The mRNA levels of Pik3cb, Pik3cd, Pik3cg, and Akt2 were significantly upregulated in the ZSW group (Fig. 6a–c). FoxO1 expression did not differ significantly between the model and ZSW groups, although the rate of FoxO1 phosphorylation was much higher in the latter group (Fig. 6c). Based on the results of the transcriptomic analysis, the mRNA expression levels of genes acting upstream of PI3K/AKT, including two growth factor-encoding genes, Vegfa and Pdgfd, and two insulin receptor substrate-encoding genes, Irs1 and Irs2, were subsequently validated by qPCR, and the results are shown in Fig. 6d and e. Taken together, these findings suggest that the diminished expression of p21 in ZSW-treated db/db mice might be due to activation of the PI3K/AKT pathway.

Fig. 6.

ZSW regulated p21 expression by activating the PI3K/AKT pathway. (a, b) The expression levels of PI3K-encoding mRNA (a) and AKT-encoding mRNA (b) in mouse epididymal adipose tissue (EAT) (n = 4). (c) Western blot analysis of p-PI3K p55, p-AKT, p-FoxO1, and FoxO1 protein expression in mouse EAT (n = 3). (d, e) The mRNA expression levels of Vegfa and Pdgfd(d) and Irs1 and Irs2(e) in mouse EAT (n = 4). ∗p < 0.05, ∗∗p < 0.01, compared with the model group. Nor: normal control group; Mod: model group; ZSW: ZishenWan group.

3.7. ZSW enhanced white fat browning and adipocyte apoptosis in db/db mice

The CKI p21 regulates apoptosis and adipogenesis.³³ Low p21 expression levels have also been shown to be associated with the activation of white fat browning (beige expansion) and the enhancement of thermogenesis, which contribute to energy consumption in white adipose tissue.³⁴ In this study, the levels of UCP1 in brown fat and EAT were evaluated by immunohistochemistry, and the TUNEL assay was used to detect apoptosis in the EAT. The results showed that, compared with the normal control group, UCP1 expression was decreased in the model group but increased after ZSW treatment (Fig. 7a). When highly expressed, p21 suppresses apoptosis by interacting with apoptosis-related proteins including procaspase-3, caspase-8, and caspase-10.^29,35 TUNEL assay and Western blot analysis demonstrated that reduced p21 levels in the EAT of mice in the ZSW group were associated with increased rates of apoptosis (Fig. 7b and c).

Fig. 7.

ZishenWan enhanced white fat browning and adipocyte apoptosis in db/db mice. (a) Immunohistochemical examination of UCP1 expression in brown adipose tissue and epididymal adipose tissue (EAT). Staining intensity was measured as mean optical density (n = 5). (b) TUNEL staining of EAT (red, TUNEL-positive; blue, nuclei). Immunofluorescence was quantified as mean fluorescence intensity (n = 4). (c) The expression of caspase-3 in EAT (n = 3). (d) Diagram of the observed pharmaceutical effects of ZSW in db/db mice. ∗p < 0.05, ∗∗p < 0.01, compared with the model group. Nor: normal control group; Mod: model group; ZSW: ZishenWan group.

4. Discussion

In this study, we investigated the effects of ZSW, a traditional Chinese formula, on genetically obese db/db mice and explored the underlying mechanisms of its weight-reducing action. We found that ZSW treatment reduced body weight in db/db mice without affecting liver or kidney function, suggesting that it may be a safer alternative compared to some conventional weight loss treatments, which often have adverse side effects. Consistent with previous studies,^16,18 ZSW also enhanced whole-body insulin sensitivity and improved glucolipid metabolism in db/db mice.

ZSW comprises three traditional Chinese herbs: Zhimu, Huangbo, and Rougui, combined in a ratio of 10:10:1, respectively. Based on early studies of the chemical components of these herbs,^{22, 23, 24, 25, 26, 27} a total of six chemicals were selected and quantified to confirm the quality of the extractions used in this study. All six chemicals were identified in the extracts, with higher levels of mangiferin, timosaponin B-II, berberine, and phellodendrine than cinnamaldehyde and cinnamic acid. This is likely due to the herb ratio in ZSW.

TCM formulas like ZSW are complex and exhibit multi-target, multifunctional, and multi-mechanism effects, which make them potentially effective for managing multifaceted conditions like obesity and diabetes. In our previous study, ZSW restored glucose homeostasis by increasing glycogen synthesis and glycolysis and reducing gluconeogenesis in the liver of db/db mice.¹⁶ Since obesity and T2DM involve systemic inflammation and multi-organ dysfunction,^{36, 37, 38} we extended our investigation to understand how ZSW affects adipose tissue using transcriptomic analysis, in order to provide valuable insights into its weight-reducing effects.

White adipose tissue primarily stores excess energy through the synthesis of triglycerides in the cytoplasm and acts as an endocrine organ by secreting adipokines, such as leptin and adiponectin, which regulate whole-body energy homeostasis by controlling food intake, glucose output from the liver, lipolysis, and gluconeogenesis.¹⁰ Increased fat mass results from adipocyte hyperplasia (increase in cell number) or hypertrophy (increase in cell size). Adipocyte hypertrophy is associated with adipocyte dysfunction and systematic insulin resistance,^5,7,39 while adipocyte hyperplasia, also known as adipogenesis, is associated with low levels of hypoxic stress and inflammation, potentially leading to “metabolically healthy obesity”.^10,40 However, because the number of adipocytes typically stabilizes after childhood, hypertrophy, rather than hyperplasia, contributes to weight gain in adulthood.⁴¹

Our results indicate that ZSW suppresses adipocyte hypertrophy in white adipose tissue. Transcriptomic analysis revealed that p21 was a core regulatory target of ZSW in EAT. p21, also known as p21WAF1/^cip1, belongs to the CKI family, which regulates the cell cycle by inhibiting CDK activity and interferes with cyclin kinase complexes to arrest the cell cycle at both the G1/S and G2/M transitions.^28,42 Additionally, p21 inhibits apoptosis, mainly by suppressing the activity of several caspases, especially caspase-3 and caspase-9.⁴²

In adipogenesis, cytoplasmic triglyceride accumulation occurs at the final stage when adipocytes exit the cell cycle and exhibit mature adipocyte morphology with corresponding metabolic features.^11,43 Large adipocytes are more vulnerable to apoptosis due to increased local hypoxic stress, pro-inflammatory adipokine secretion, and inflammatory cell infiltration.⁴⁴ In our study, the rate of EAT apoptosis was significantly higher in db/db mice than that in wt/wt mice, likely due to overnutrition-induced adipocyte hypertrophy. Elevated p21 expression in adipose tissue inhibits the apoptosis of oversized adipocytes and halts the cell cycle, reducing adipocyte turnover.⁴⁵ Which means the accumulation of p21 in adipocytes is associated with senescence-related features, including larger cell size and reduced cell proliferation.⁴⁶

In addition to white adipocytes, brown and beige adipocytes play a role in energy metabolism by utilizing energy for thermogenesis.⁴⁷ These adipocytes are characterized by high expression levels of UCP1, a mitochondrial protein that decreases ATP generation and promotes heat genesis through driving high levels of fatty acid oxidation.⁴⁸ Beige adipocytes, found in white adipose tissue, can originate from reprogrammed white adipocytes or bipotent preadipocytes.⁴⁷ The overexpression of p21 has been shown to inhibit beige adipocyte formation, whereas p21 deletion enhances it, likely through a PPARγ co-activator-1α (PGC1α)-dependent mechanism.^34,45,49 In our study, ZSW treatment increased UCP1 protein expression in both brown and white adipose tissues, indicating that ZSW enhances thermogenesis and energy expenditure in adipocytes.

Cells expressing high levels of p21 were observed to accumulate in adipose tissues in high-fat-diet induced obese mice.⁵⁰ The clearance of approximately 25 % of these cells in gonadal visceral adipose tissue significantly improved glucose tolerance and whole-body insulin sensitivity, accompanied by a reduction in pro-inflammatory activities. Additionally, the ablation of p21 in white adipose tissue improved insulin sensitivity and reduced obesity in mice fed a high-fat/high-sugar diet, while apoptosis was induced in EAT.^45,50 In our study, ZSW treatment significantly reduced p21 mRNA and protein expression in EAT, accompanied by increased apoptosis, as shown by elevated caspase-3 levels and TUNEL assay. The expression of the macrophage marker F4/80 is upregulated during obesity and thus serves as an indicator of increased macrophage and related inflammatory cell infiltration in adipose tissue.⁵¹ The expression of F4/80 tended to increase in the model group and decline in ZSW-treated mice, which suggested that ZSW reduced the inflammatory response in EAT.

The levels of molecules acting upstream of p21 were also examined in this study. PI3K/AKT activation leads to the phosphorylation of the transcription factor FoxO1, which promotes its translocation from the nucleus and diminishes its function. p21 promoter activation is dependent on FoxO1 activity.^30,52 In the present work, ZSW significantly induced the PI3K/AKT signaling pathway in EAT, which may have led to a decrease in p21 expression through the phosphorylation and subsequent inactivation of FoxO1.

Adipose tissue is also an important insulin-targeted organ. Studies have shown that IRS, PI3K, and AKT are three key nodes in insulin signaling,⁵³ and all three proteins were upregulated in EAT after ZSW treatment. The PI3K/AKT pathway functions at the intersection of metabolic and cell proliferation pathways. PI3K/AKT activation can be induced by receptor tyrosine kinases (RTKs), the Ras superfamily of small GTPases, and G protein-coupled receptors.⁵⁴ Both insulin and growth factors exert their biofunction through binding to RTKs. Well-characterized RTKs include insulin receptors, EGFR, PDGFR, VEGFR, and FGFRs.⁵⁴ Here, ZSW upregulated the mRNA levels of Irs1, Irs2, Vegfa, and Pdgfd, important upstream activators of class I PI3K protein activity, leading to the increased phosphorylation of PI3K/AKT, and, consequently, the suppression of p21 expression, which improved insulin sensitivity. Notably, we have previously shown that ZSW significantly activated the PI3K/AKT pathway in the liver,¹⁶ showing that ZSW has a core set of targets irrespective of the tissue type.

This study used a combination of transcriptomics and molecular techniques, including qPCR, immunohistochemistry, and western blotting, to explore the effects of ZSW on obesity. It bridges TCM and modern molecular biology, leveraging the multi-target properties of traditional formulations. To the best of our knowledge, this is the first study to investigate the molecular mechanisms of ZSW in adipose tissue and its role in improving insulin resistance. These findings provide a comprehensive understanding of how ZSW modulates metabolic dysfunction.

This study has few limitations. Although we used the optimal dose identified in previous research, we did not explore whether ZSW has a dose-dependent effect on p21 expression. Future studies should investigate this to better understand the therapeutic potential of ZSW. Additionally, in vitro experiments could help clarify the direct effects of ZSW on adipocytes and its molecular mechanisms, offering more detailed insights into its mode of action.

Our findings indicate that ZSW targets p21 to modulate adipocyte function and improve metabolic health. Given that targeting adipocyte apoptosis has been proposed as a potential treatment for obesity,⁵⁵ ZSW's ability to reduce p21 expression and promote apoptosis in visceral adipose tissue highlights its potential as a therapeutic agent. Previous studies have evaluated drugs that target p21 in human fat tissue ex vivo, showing promising effects.⁵⁰ Collectively, our results provide experimental evidence for the potential benefits of ZSW in treating metabolic disorders.

5. Conclusions

This study demonstrated that ZSW reduces obesity and enhances insulin sensitivity in db/db mice. These effects are likely mediated by the inhibition of p21 expression through the activation of the PI3K/AKT signaling pathway in EAT. By downregulating p21, ZSW promotes adipocyte apoptosis, increases thermogenesis, and suppresses adipocyte hypertrophy, leading to weight loss and improved insulin resistance. These findings suggest that ZSW has promising potential as a therapeutic agent for obesity-related metabolic disorders, offering a multi-target approach to modulate adipose tissue function and metabolic health.

Author contributions

Y.W. designed the experiment, Y.W. and H.Q. completed the experiment, analyzed data and wrote the manuscript. Z.Q., Y.H., and X.L. helped with the animal experiments. Y.C. helped with the manuscript writing. L.Q. and M.J. reviewed and revised the manuscript. T.L. and L.W. supervised this study. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

The animal experiments conducted in this study were approved by the Animal Care and Ethics Committee of Beijing University of Chinese Medicine (approval No. BUCM-2023090409-3111) and conformed to internationally recognized principles of laboratory animal use and care.

Consent for publication

All authors agree to publish this article.

Availability of data and materials

The datasets presented in this article could be found in online repository. The raw data from transcriptomics are deposited in Sequence Read Archive database (BioProject No. PRJNA1091037).

Funding

This research was funded by the Creation and Talent Introduction Base of Prevention and Treatment of Diabetes and Its Complications with Traditional Chinese Medicine (Grant No. B20055).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Abbreviations

AKT

protein kinase B

ALT

alanine aminotransferase

AST

aspartate aminotransferase

AUC

area under the blood glucose-time curve

BMI

body mass index

BUN

blood urea nitrogen

CDKs

cyclin-dependent kinases

CKIs

cyclin-dependent kinase inhibitors

CREA

creatinine;

DEGs

differentially expressed genes

EAT

epididymal adipose tissue

ELSD

evaporative light-scattering detector

FoxO1

forkhead box protein O1

FPKM

ragments per kb per million reads

GO

Gene Ontology

H&E

hematoxylin and eosin

HDL:

high-density lipoprotein

HOMA-IR

homeostatic model assessment for insulin resistance

HPLC

high-performance liquid chromatography

KEGG

Kyoto Encyclopedia of Genes and Genomes

LDL:

low-density lipoprotein

OGTT

oral glucose tolerance test

PGC1α

PPAR-γ co-activator-1α

PI3K

phosphoinositide 3-kinase

RTK

receptor tyrosine kinase

T2DM

type 2 diabetes

TC

total cholesterol

TEM

transmission electron microscopy

TG

triglyceride;

TSA

tyramide signal amplification

TUNEL:

terminal deoxynucleotidyl transferase-mediated dUTP nick-end labeling

UCP1

uncoupling protein 1; ZSW: ZishenWan

Appendix A. Supplementary data

The following is the Supplementary data to this article: