Integrated hepatic transcriptome and metabolome reveal the mechanisms of Jiangtang Tiaozhi formula on improving glycolipid metabolic disorder

Abstract

BACKGROUND

Glycolipid metabolic disorder includes a series of chronic diseases that are closely associated with disturbances in both glucose and lipid metabolism. Jiangtang Tiaozhi formula (JTTZF) demonstrating significant hypoglycemic, lipid-modifying, and anti-inflammatory effects. However, the specific molecular mechanisms underlying JTTZF’s hepatoprotective effects and its ability to ameliorate glycolipid metabolic disorder remain largely unexplored.

AIM

To investigate how JTTZF improves glycolipid metabolic disorder using hepatic transcriptome and metabolome analyses.

METHODS

To induce glycolipid metabolic disorder, male C57BL/6J mice were fed a high-fat diet (HFD) for 12 weeks, after which they received an 8-week administration of JTTZF. Liver tissues were analyzed using transcriptomics and metabolomics. Real-time quantitative polymerase chain reaction validated key gene expression.

RESULTS

Metabolomics data revealed that JTTZF significantly regulated HFD-induced alterations in glycolipid metabolism, with notable changes in pathways such as the pentose phosphate pathway, steroid hormone biosynthesis, and purine metabolism. Transcriptomics profiles indicated that JTTZF exerted regulatory effects on lipid and glucose metabolism, primarily through pathways including peroxisome proliferators-activated receptor signaling, drug metabolism other enzymes, and regulation of lipolysis in adipocytes. Real-time quantitative polymerase chain reaction confirmed that JTTZF modulated pivotal genes associated with fatty acid synthesis, lipolysis, insulin resistance, energy metabolism, and inflammation. These findings suggest that JTTZF may act through multiple pathways to improve glycolipid metabolic disorder.

CONCLUSION

JTTZF can ameliorate the glycolipid metabolic disorder induced by HFD-diet by regulating lipid metabolism and improving insulin tolerance.

Key Words: Glycolipid metabolic disorder; Hepatic transcriptome; Hepatic metabolome; Jiangtang Tiaozhi formula; Multi-omics

Core Tip: This study explores the effects of Jiangtang Tiaozhi formula (JTTZF) on glycolipid metabolic disorder in high-fat diet-induced mice. By integrating hepatic transcriptome and metabolome analyses, we find that JTTZF regulates key genes and metabolites linked to glucose and lipid metabolism. Our results highlight the potential of JTTZF as a therapeutic agent for glycolipid metabolic disorder and provide valuable insights into its underlying mechanisms.

INTRODUCTION

Glycolipid metabolic disorder is caused by a variety of risk factors, including genetic, molecular, psychological, and environmental factors[1]. Glycolipid metabolic disorder, encompassing conditions like non-alcoholic fatty liver disease, type 2 diabetes mellitus (T2DM), and obesity, represent a major global public health challenge driven by lifestyle factors and unhealthy eating and living habits[2,3]. These disorders arise from dysregulated glucose and lipid metabolism, underpinned by core pathologies including insulin resistance, oxidative stress, inflammation, and gut microbiota dysbiosis, which depends on the high cooperation of various organs and tissues[1]. The liver plays a crucial role in regulating systemic glucose and lipid concentration. It is an intermediate organ between endogenous and exogenous energy supply to extrahepatic organs and serves as the major metabolic control hub for the synthesis and metabolism of glucose, lipids, carbohydrates, and proteins[4]. Consequently, understanding the precise molecular mechanisms governing hepatic glycolipid metabolism is critical, yet remains incompletely elucidated. The challenging task of researching and developing effective therapies and drugs to improve glycolipid metabolic disorder is still ongoing.

Traditional Chinese medicines are characterized by multi-components, multi-pathways, and multi-targets. Substantial preclinical and clinical evidence supports the efficacy of various traditional Chinese medicines modalities (monomers, compounds, decoctions) in managing glycolipid metabolic disorder with favorable safety profiles[5-7]. Jiangtang Tiaozhi formula (JTTZF), derived from Dahuang Huanglian Xiexin decoction, is composed of Coptis chinensis, Rhizoma anemarrhenae, Monascus, Aloe, Salvia miltiorrhiza, Schisandrae chinensis fructus, Momordica charantia, and Zingiberis rhizoma[8]. It is a clinically established formulation for treating T2DM and dyslipidemia, demonstrating significant hypoglycemic, lipid-lowering, and anti-inflammatory effects[8,9]. Modern pharmacological studies show that the prescription exerts its effects through multiple components, targets, and pathways, which makes JTTZF a promising candidate for treating the complex network dysregulation of glycolipid metabolic disorder, potentially outperforming single-target therapies. However, despite its clinical utility, the specific molecular mechanisms underlying JTTZF’s hepatoprotective effects and its ability to ameliorate glycolipid metabolic disorder in vivo, particularly at a systems level, remain largely unexplored. A critical gap exists in the comprehensive understanding of how JTTZF modulates hepatic gene expression and metabolic pathways.

Current research lacks a comprehensive analysis of the hepatic transcriptomic and metabolomic changes induced by JTTZF in high-fat diet (HFD)-induced models. This deficiency restricts our comprehension of the molecular pathways through which JTTZF ameliorates glycolipid metabolic disorder. Transcriptomics and metabolomics analyses are powerful high-throughput technologies that are increasingly used to identify key genes and metabolites from large-scale interaction networks of multisystem diseases, which provides a new method and insight for traditional Chinese medicines research[10,11]. In this research, we combined transcriptomic and metabolomic approaches to identify key genes, metabolites, and pathways altered after JTTZF treatment and investigated the detailed mechanism of JTTZF in ameliorating glycolipid metabolic disorder. Our research is designed to elucidate the detailed mechanisms underlying JTTZF’s effects on glycolipid metabolism and to offer convincing evidence of its therapeutic potential.

MATERIALS AND METHODS

Animals and drugs

Male C57BL/6J mice (7 weeks old, specific pathogen-free) were obtained from the Beijing Vital River Laboratory Animal Technology Company (Beijing, China). These mice were raised in a specific pathogen free laboratory animal facility at the Guang’anmen Hospital, China Academy of Chinese Medical Sciences (Beijing, China), where the room temperature was maintained at 22 ± 2 °C and humidity at 55% ± 5%, with a 12-hour light/12-hour dark cycle. The JTTZF formula granules utilized in the experiment were supplied by Tianjiang Pharmaceutical Co., Ltd. (Jiangsu Province, China).

HFD-induced model and drugs administration

Randomly divided the mice into three groups (n = 8): Control group, model group, and JTTZF group. The control group received a standard rodent diet, while the model and JTTZF groups were given a HFD containing 60 kcal% from fat (D12492; Research Diets, New Brunswick, New Jersey, United States) for 6 weeks to form an HFD-induced model (weight gain ≥ 30%, fasting blood glucose ≥ 7.0 mmoL/L). Mice in the JTTZF group were administered with JTTZF, and those in the control and model groups were administered with normal saline. Based on a human dose of 84 g/day (for an average body weight of 70 kg), and a reference equivalent dose for mice that is 9.1 times higher. Thus, the calculated equivalent dose for mice is 10.92 g/kg. All groups were administered orally once daily for 8 weeks at a dose of 10 mL/kg.

Index detection

Fasting blood glucose and body weight were measured at fixed time points each week. oral glucose tolerance test (OGTT) was conducted on the first day after 8 weeks of administration. Mice were fasted overnight (12 hours) with free access to water and then given 2 g/kg glucose intragastrically. Whole blood was collected from the tail vein at 0, 15, 30, 60, and 120 minutes, blood glucose levels were measured using a Roche Accu Chek blood glucose meter, and area under the curve was calculated. triglyceride (TG) and total cholesterol (TC) were detected by the selective inhibition method on an automated biochemistry analyzer (Olympus AU480; Olympus Corporation, Tokyo, Japan).

Sample collection

After 8 weeks of continuous administration, blood samples were collected from the retro-orbital venous plexus of mice following a 12-hour fast. The serum sample was centrifuged at 1000 g for 20 minutes at 4 °C and then stored at -80 °C. The mice were cervically dislocated under pentobarbital sodium anesthesia, and liver tissue samples were immediately stored at -80 °C for further analysis.

Transcriptomics analysis

Liver tissue RNA was extracted using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, United States) following the manufacturer’s protocol. RNA quantity and integrity were evaluated with the RNA Nano 6000 assay kit on the Bioanalyzer 2100 system (Agilent Technologies, CA, United States). A complementary DNA library was constructed and sequenced on the Illumina Novaseq platform, generating an end reading of 150 bp pairing. The raw data were converted to FASTQ format and processed with FASTP software. Adapters, N-bases, and low-quality reads were removed to obtain clean reads. Hisat2 (v2.0.5) was used to index the reference genome and align paired-end clean reads to it[12]. Gene read counts were obtained using feature counts (v1.5.0-p3), and fragments per kilobase per million (FPKM) values were calculated based on gene length[13,14]. Differential expression analysis was conducted using the DESeq2 R package (1.20.0), with P values adjusted using the Benjamini and Hochberg method to control the false discovery rate (FDR). Genes with P < 0.05, |fold change (FC)| > 1.5, and FPKM > 1 were identified as differentially expressed genes (DEGs)[15,16]. Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses of DEGs were performed using the clusterProfiler R package (3.8.1)[17,18].

Metabolomics analysis

Injecting liver extract into liquid chromatograph mass spectrometer/mass spectrometer system (ThermoFisher, Germany) for analysis[19]. The raw data were processed using compound discoverer 3.1 (ThermoFisher). Metabolites were annotated using the KEGG database (https://www.genome.jp/kegg/pathway.html). In the multivariate statistical analysis, partial least squares discriminant analysis (PLS-DA) and principal components analysis (PCA) were conducted to visualize the metabolic difference after log transformation (log 10) in metaX[20]. The statistical significance of differences in metabolite levels between groups was assessed using the t test. P values obtained were adjusted for multiple comparisons using the Benjamini and Hochberg method to control the FDR. FC values were calculated to quantify relative differences in metabolite levels. Metabolites with P value < 0.05, variable important in projection (VIP) > 1, and FC > 1.5 or FC < 0.667 were considered as differentially expressed metabolites (DEMs)[21]. Hierarchical clustering analysis of DEMs was performed using R (heatmap package) to analyze metabolite expression patterns across samples. Volcano plots were utilized to filter metabolites of interest through R (ggplot2 package). Metabolic pathway enrichment analysis of DEMs was conducted using MetaboAnalyst (v 5.0) with a P value < 0.05 from the hypergeometric test.

Quantitative polymerase chain reaction assay

The messenger RNA (mRNA) expression of genes related to glycolipid metabolic disorder was assessed via real-time polymerase chain reaction (PCR). Total hepatic RNA was isolated from samples using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, United States). Reverse transcription of total RNA was performed using a SweScript All-in-One RT SuperMix for quantitative PCR (Servicebio, Wuhan, Hubei Province, China). The target gene was measured using SYBR Green quantitative PCR master mix on BioRad Laboratories CFX Connect^TM real-time PCR detection system. Glyceraldehyde-3-phosphate dehydrogenase was used as an internal reference gene, the expression was calculated using 2^-△△ct. The PCR protocol included initial activation at 95 °C for 30 seconds, followed by 40 cycles of denaturation at 95 °C for 15 seconds, and annealing and extension at 60 °C for 30 seconds. Fluorescence signals were collected once every 0.5 °C increase in temperature from 65 °C to 95 °C[22]. Primer sequences are detailed in Table 1.

Table 1 Forward and reverse primer sequences for real-time polymerase chain reaction.

Gene name	Forward	Reverse
Cry1	TACTGAGGCTGGCGTGGAA	AGTGGCTCCATCTTGCTGACG
Ces2a	CTAAATCTGGTGTCCATGCCTTC	ATTCATTTCCTTCGCATCCTCT
Ces1d	CATTGCTGGTCTGGTTGCTACT	TGCTTGTTGATGCCCACTATGT
Gpnmb	GCCAAGCGATTTCGTGATGT	AGTCCTTCCACCTGCCGTCT
Cidea	GTCCTACTGACCCCCCTCATA	TCCTCCAGCACCAGCGTAA
Gprc5b	TGGCAAGTTCAAACGGTGG	TGGGAGAAGGGTGTAGTGGAT
Fabp4	GTGATGCCTTTGTGGGAACC	CCAGTTTGAAGGAAATCTCGGT
Fabp3	GACCAAGCCTACTACCATCATCG	TGAGTTTGCCTCCGTCCAGC
Cry1	TACTGAGGCTGGCGTGGAA	AGTGGCTCCATCTTGCTGACG

Statistical analysis

Data analysis was conducted using GraphPad Prism 9.5.0 (GraphPad, La Jolla, CA, United States), and results are presented as mean ± SE. Multiple comparisons were analyzed by one-way analysis of variance and the Kruskal-Wallis test. Differences were considered statistically significant at the level of ^aP < 0.05, ^bP < 0.01, ^cP < 0.001.

RESULTS

Effects of JTTZF on metabolic parameters in HFD-induced glycolipid metabolic disorder mice

As shown in Figure 1, HFD-induced glycolipid metabolic disorder was observed in the model group. After 8 weeks treatment with JTTZF, a notable reduction in blood glucose was observed (Figure 1A, P < 0.001). The OGTT results indicated that, after 60 minutes, the blood glucose level in the JTTZF group was significantly lower compared with the control group (Figure 1B). Compared with the control group, the model group had a higher body weight gain (Figure 1C, P < 0.0001). Although JTTZF treatment appeared to reduce weight gain, this effect did not reach statistical significance compared with the model group. However, the trend may still be biologically meaningful, it could suggest a longer term effect not captured within the study period. The serum TG and TC levels were significantly elevated in HFD-induced mice (Figure 1D and E, P < 0.01). JTTZF treatment led to a marked reduction in serum TG levels. While JTTZF also appeared to decrease TC levels, this change was not statistically significant compared with the model group (Figure 1F, P < 0.001). Moreover, JTTZF significantly reduced fasting glucose levels and the area under the curve of the OGTT compared with the model group (Figure 1G, P < 0.01). In summary, JTTZF can reduce body weight, blood lipids, and blood glucose in HFD-induced mice, and ameliorate the glycolipid metabolic disorder.

Figure 1 Jiangtang Tiaozhi formula attenuated glycolipid metabolic disorder in vivo. A: Fasting glucose of mice during treatment (n = 8); B: Oral glucose tolerance test (OGTT) (n = 8); C: Body weight (n = 8); D: Triglyceride (n = 8); E: Total cholesterol (n = 8); F: Fasting glucose of mice after 8 weeks of treatment (n = 8); G: Area under the curve of OGTT (n = 8). ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. NS: Not significant; JTTZF: Jiangtang Tiaozhi formula; OGTT: Oral glucose tolerance test; TG: Triglyceride; TC: Total cholesterol; AUC: Area under the curve.

Effects of JTTZF supplementation on hepatic transcriptome

Figure 2 illustrates the gene transcript expression levels of the control, model and JTTZF groups as determined by RNA sequencing. The PCA score plot (Figure 2A) indicates a distinct separation among the three groups. The JTTZF group exhibits a tendency to return to the control group, indicating that JTTZF supplementation significantly alters the liver transcriptome of mice with glycolipid metabolic disorder. Compared with the control group, there are 1220 DEGs in the model group, including 607 up-regulated and 613 down-regulated genes. In contrast to the model group, the JTTZF have 1206 DEGs, including 990 up-regulated and 216 down-regulated genes (Figure 2B and C). The heat map shows expression of representative DEGs (Figure 2D). Relative to the control group, 215 genes were down-regulated in the model group but up-regulated in the JTTZF group. These genes are primarily clustered in pathways such as mainly clustered in linoleic acid metabolism, biosynthesis of unsaturated fatty acids, inflammatory mediator regulation of transient receptor potential channels (Figure 2E). 90 genes were up-regulated in the model group but down-regulated in the JTTZF group. These genes were enriched in the in pathways such as ErbB signaling pathway, peroxisome proliferators-activated receptor signaling pathway, Regulation of lipolysis in adipocytes, interleukin-17 signaling pathway, Fat digestion and absorption, fructose and mannose metabolism, ferroptosis (Figure 2F). The specific information of these potential genes are summarized in Table 2. To identify the transcriptional profile results, we measured the mRNA levels of 8 potential genes by real-time PCR, the results showed a high consistency with the transcriptome data (Figure 3). The results of our transcriptome analysis reveal significant changes in gene expression patterns in the liver of glycolipid metabolic disorder mice treated with JTTZF. The PCA analysis confirmed distinct separation among the control, model, and JTTZF groups, indicating that JTTZF treatment has a profound effect on the liver transcriptome. The volcano plots and heat map further illustrate the differential expression of genes, with JTTZF reversing many of the changes observed in the model group. This suggests that JTTZF may act by modulating specific pathways involved in glycolipid metabolism.

Figure 2 Jiangtang Tiaozhi formula modulates liver transcriptome profiles. A: Principal components analysis [control vs model vs Jiangtang Tiaozhi formula (JTTZF) mice, n = 8]; B: Volcano plots (model vs control mice, n = 8); C: Volcano plots (JTTZF vs model mice, n = 8); D: Heatmap of representative differentially expressed genes (control vs model vs JTTZF mice, n = 8), with red indicating up-regulation and blue indicating down-regulation; E: Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis to annotate the biological function of differential genes(model vs control mice, n = 8); F: KEGG pathway analysis to annotate the biological function of differential genes (JTTZF vs model mice, n = 8). JTTZF: Jiangtang Tiaozhi formula; PC: Principal component.

Figure 3 The messenger RNA levels of potential genes by real-time polymerase chain reaction. A: Ces1d; B: Ces2a; C: Cry1; D: Cidea; E: Fabp3; F: Fabp4; G: Gprc5b; H: Gpnmb. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001.

Table 2 The specific information of these potential genes.

No.	Gene ID	Gene name	Model vs control		JTTZF vs model		Pathway
			log2FC	Trend	log2FC	Trend
1	ENSMUSG00000020038	Cry1	-1.65	Db	1.92	Ub
2	ENSMUSG00000055730	Ces2a	-3.09	Db	1.64	Ub	Pathway1
3	ENSMUSG00000056973	Ces1d	-1.91	Db	1.91	Ub
4	ENSMUSG00000029816	Gpnmb	3.55	Ub	-2.79	Db
5	ENSMUSG00000024526	Cidea	4.14	Ub	-2.99	Db
6	ENSMUSG00000008734	Gprc5b	1.73	Ub	-2.15	Db
7	ENSMUSG00000062515	Fabp4	1.52	Ub	-1.58	Db	Pathway2,3
8	ENSMUSG00000028773	Fabp3	4.51	Ub	-2.90	Db	Pathway2

^bP < 0.01.

¹Drug metabolism other enzymes.

²Peroxisome proliferators-activated receptor signaling pathway.

³Regulation of lipolysis in adipocytes.

JTTZF: Jiangtang Tiaozhi formula; D: Down-regulated; U: Up-regulated; FC: Fold change.

Effects of JTTZF on hepatic metabolic profiles

To explore the impact of JTTZF on hepatic metabolites in mice with HFD-induced metabolic disorders, we conducted untargeted metabolomics analysis (Figure 4). The PCA model showed good differentiation between the control and model groups, indicating significant differences in metabolic liver function (Figure 4A). The model and JTTZF groups also exhibited clear differentiation, indicating changes in the amount and concentration of metabolites after drug intervention, although PCA analysis revealed slight overlap (Figure 4B). Furthermore, the control and model groups were significantly separated, with explanatory power (R2) Y = 0.99 and predictive power (Q2) Y = 0.96 (Figure 4C) in the positive ion mode and (R2) Y = 0.99 and (Q2) Y = 0.96 (Figure 4D) in the negative ion mode, respectively. Similarly, the model and JTTZF groups showed significant separation, with R2Y = 0.97 and Q2Y = 0.68 in the positive ion mode (Figure 4E), and R2Y = 0.97 and Q2Y = 0.62 in the negative ion mode (Figure 4F). These results indicated the robust predictive reliability of the orthogonal PLS-DA model. The PLS-DA model’s permutation test demonstrated stability and strong predictive power, suitable for identifying DEMs (Figure 4G-J).

Figure 4 Multivariate statistical analysis of liver samples among the control, model, and Jiangtang Tiaozhi formula groups. A: Principal components analysis (PCA) score plot of liver samples among control, model, and Jiangtang Tiaozhi formula (JTTZF) groups in the positive ion mode; B: PCA score plot of liver samples among control, model, and JTTZF groups in the negative ion mode; C: Scores plots of partial least squares discriminant analysis (PLS-DA) between control and model groups in the positive ion mode; D: Scores plots of PLS-DA between control and model groups in the negative ion mode; E: Scores plots of PLS-DA between and between model and JTTZF groups in the positive ion mode; F: Scores plots of PLS-DA between model and JTTZF groups in the negative ion mode; G: Permutation tests of PLD-DA models between control and model groups in the positive ion mode; H: Permutation tests of PLD-DA models between control and model groups in the negative ion mode; I: Permutation tests of PLD-DA models between model and JTTZF groups in the positive ion mode; J: Permutation tests of PLD-DA models between model and JTTZF groups in the negative ion mode. PC: Principal component; J: Jiangtang Tiaozhi formula group; Mo: Model group; C: Control group; Cor: Correlation.

DEMs were filtered using the following criteria: (1) VIP > 1.0; (2) P < 0.05; and (3) FC > 1.5 or FC < 0.667[23-25]. According to the volcano plots, in the positive ion mode, 108 metabolites were significantly up-regulated and 81 metabolites were down-regulated between control and model groups (Figure 5A and B). In the negative ion mode, 87 metabolites were increased and 88 metabolites were decreased (Figure 5C and D). These metabolites were enriched in pathways such as steroid hormone biosynthesis, tricarboxylic acid cycle (TCA cycle), pentose phosphate pathway, glycerophospholipid metabolism, glycolysis/gluconeogenesis (Figure 5E). Between the model and JTTZF groups, 53 metabolites were increased and 78 metabolites were decreased in the positive ion mode, while 62 metabolites were increased and 39 metabolites were decreased in the negative ion mode. These metabolites were enriched in TCA cycle, pentose and glucuronate interconversions, pentose phosphate pathway, arachidonic acid metabolism, fatty acid degradation (Figure 5F). The results gained under these two modes are shown in Figure 5. Compared with the control group, 41 metabolites were increased in the model group and decreased in the JTTZF group, while 34 metabolites were decreased in the model group and increased in the JTTZF group. The TCA cycle, pentose phosphate pathway, pyrimidine metabolism, steroid hormone biosynthesis, purine metabolism are common to all three groups, suggesting their involvement in glycolipid metabolic disorder. We selected 25 significantly altered endogenous metabolites associated with glycolipid metabolic disorder as potential biomarkers for liver tissue (Table 3).

Figure 5 Jiangtang Tiaozhi formula modulates liver metabolite profiles. A: Volcano plots under the positive ion mode (model vs control mice, n = 8); B: Volcano plots under the ion mode (model vs control mice, n = 8); C: Volcano plots under the positive ion mode [Jiangtang Tiaozhi formula (JTTZF) vs model mice, n = 8]; D: Volcano plots under the negative ion mode (JTTZF vs model mice, n = 8); E: Overview of metabolic pathway analysis of liver metabolism (model vs control mice, n = 8); F: Overview of metabolic pathway analysis of liver metabolism (JTTZF vs model mice, n = 8). J: Jiangtang Tiaozhi formula group; Mo: Model group; C: Control group; VIP: Variable important in projection; TCA: Tricarboxylic acid cycle.

Table 3 The endogenous metabolites associated with glycolipid metabolic disorder.

No.	Formula	RT (minute)	m/z	Metabolites	Ion mode	Model vs control			JTTZF vs model
						VIP	FC	Trend	VIP	FC	Trend
1	C23H45NO4	7.12	400.34	Palmitoylcarnitine	ESI +	2.47	6.69	Ub	2.45	0.26	Da
2	C23H44NO4	6.80	398.33	ACar 16:1	ESI +	1.88	4.10	Ub	2.42	0.27	Da
3	C25H48NO4	7.25	426.36	ACar 18:1	ESI +	2.26	6.38	Ub	2.11	0.28	Da
4	C21H42NO4	6.67	372.31	ACar 14:0	ESI +	1.21	2.22	Ub	2.54	0.34	Da
5	C28H54NO4	7.89	468.40	ACar 21:1	ESI +	2.34	11.32	Ub	1.55	0.27	Da
6	C10H14N5O7P	1.60	348.07	Adenosine 5’-monophosphate	ESI +	1.02	0.41	Db	1.22	0.45	Da
7	C26H48NO7P	8.34	562.32	LPC 18:3	ESI -	1.68	0.44	Db	1.68	1.82	Ua
8	C26H50NO8P	7.06	536.33	PC (5:0/13:1)	ESI +	2.54	0.38	Db	1.89	1.74	Ua
9	C23H40NO7P	7.90	472.25	LPE 18:4	ESI -	2.25	0.21	Db	1.53	2.25	Ua
10	C25H44NO4	6.72	422.33	ACar 18:3	ESI +	1.23	2.69	Ub	2.46	0.32	Da
11	C23H42NO7P	8.32	474.26	LPE 18:3	ESI -	1.55	0.37	Db	1.31	1.89	Ua
12	C26H52NO4	7.78	442.39	ACar 19:0	ESI +	2.06	3.17	Ub	2.18	0.38	Da
13	C26H52NO8P	8.43	538.35	PC (9:0/9:0)	ESI +	1.04	1.50	Ua	2.19	0.55	Da
14	C25H46NO4	6.95	424.34	ACar 18:2	ESI +	1.34	2.19	Ub	2.38	0.41	Da
15	C22H44NO7P	7.72	466.29	LPC 14:1	ESI +	1.84	0.46	Db	1.64	1.66	Ua
16	C10H15N5O10P2	1.47	428.04	Adenosine diphosphate	ESI +	1.28	0.33	Ub	1.20	0.36	Da
17	C9H9NO3	5.35	132.04	Hippuric acid	ESI -	1.42	0.42	Db	1.63	2.09	Ua
18	C28H50NO8P	7.62	560.34	PC (4:0/16:3)	ESI +	1.13	0.67	Da	2.27	2.03	Ua
19	C29H46NO4	6.93	472.34	ACar 22:6	ESI +	1.96	3.10	Ub	2.00	0.52	Db
20	C5H10O5	1.41	149.05	D-Ribose	ESI -	1.60	1.87	Ub	1.86	0.61	Db
21	C4H8O3	2.79	105.06	2-hydroxybutyric acid	ESI +	1.15	1.69	Ub	2.07	0.62	Db
22	C21H30O2	9.29	315.23	Progesterone	ESI +	1.02	2.15	Ub	2.02	0.60	Db
23	C23H44NO7P	7.40	478.29	LysoPE 18:2	ESI +	1.09	0.65	Da	1.45	1.52	Ub
24	C24H45O9P	8.26	507.27	LPG 18:2	ESI +	1.01	1.82	Ua	1.16	0.60	Db
25	C26H46NO7P	7.72	516.31	LPC 18:4	ESI +	1.99	0.10	Db	1.07	2.44	Ub

^aP < 0.05.

^bP < 0.01.

JTTZF: Jiangtang Tiaozhi formula; D: Down-regulated; U: Up-regulated; FC: Fold change; VIP: Variable important in projection; RT: Retention time; ESI: Electrospray ionization; ACar: Acetyl carnitine; LPC: Lysophosphatidyl choline; PC: Phosphatidylcholine; LPE: Lysophhosphatidyl ethanolamine; LysoPE: Lysophosphatidyl ethanolamine; LPG: Lysophosphatidyl glycerol.

Integrated enrichment analysis of transcriptome and metabolite profiles

To understand the changes occurring in the liver, the potential relationships between genes and metabolites were further analyzed. To systematically evaluate the metabolic disturbances and potential regulatory factors of JTTZF in ameliorating the HFD-induced glycolipid metabolic disorder, we jointly analyzed the DEGs in transcriptomics and the characteristic metabolic pathways in metabolomics. The DEGs and DEMs were simultaneously mapped to the KEGG pathway database to identify the common pathways. Through mapping the metabolic enzyme-related genes of metabolic pathways enriched in the metabolomics, we speculated that JTTZF may alleviate glycolipid metabolic disorder by regulating the pathways related to energy production, lipid metabolism, and glucose metabolism (Figure 6).

Figure 6 Integrated transcriptomics and metabolomics analyses of Jiangtang Tiaozhi formula on improving glycolipid metabolic disorder.

DISCUSSION

Glycolipid metabolic disorder is a complex chronic metabolic disease, and our previous study reported that JTTZF showed therapeutic effects for patients. Moreover, JTTZF can effectively and safely reduce blood glucose and lipids in glycolipid metabolic disorder patients. It also enhances insulin sensitivity, reduces chronic inflammation, and modifies the gut microbiota structure. Currently, identifying potential intervention targets for JTTZF to improve glycolipid metabolic disorder is an important goal. However, the characteristics of metabolites in glycolipid metabolic disorder have not been fully demonstrated and there is a lack of integrated transcriptome metabolome analysis of liver in HFD-induced glycolipid metabolic disorder mice. In this study, transcriptomics and the metabolomics of HFD-induced glycolipid metabolic disorder mice were performed. Integrated transcriptome metabolome analysis revealed the mechanism of JTTZF amelioration of the glycolipid metabolic disorder from the perspective of regulation of metabolites accumulation and expression of genes.

Transcriptome analysis showed that JTTZF ameliorated the HFD-induced glycolipid metabolic disorder induced in several ways. Specifically, JTTZF regulated key genes associated with lipid metabolism. Compared with the control group, Cry1, Ces2a, and Ces1d were down-regulated, while Gpnmb, Cidea, Gprc5b, Fabp4, and Fabp3 were up-regulated in the model group. After JTTZF intervention, the expression levels of these genes were restored to levels comparable to those in the control group. Cry1 is related to the lipids utilization and the foemation of lipid droplets. Knocking down Cry1 inhibits the expression of adipogenic markers and reduces lipid droplet formation in cells under adipogenic induction[26,27]. High-fat feeding accelerates the degradation of autophagic Cry1, leading to obesity-related hyperglycemia[28]. Additionally, down-regulation of hepatic Cry1 is linked to heightened hepatic glucose production through the mediation of FOXO1 degradation[29-31]. Carboxylesterases, including mouse Ces2a and Ces1d, are important hydrolytic enzymes involved in lipid metabolism. Ces2a, a highly efficient diglyceride and monoglycerides hydrolase, is vital for energy and lipid metabolism[32-34]. Ces2a knockout mice exhibit increased body and liver weight, insulin resistance, and liver steatosis due to impaired diacylglycerol and lysophosphatidylcholine (LPC) catabolism[35,36]. Ces1d, a major hepatic cholesteryl ester hydrolase, is notably elevated in obese patients with T2DM. Ces1d knockout mice show increased fat mass and abnormal lipid droplet deposition in adipocytes, leading to ectopic TG accumulation in other tissues[37,38]. Cell death-inducing DNA fragmentation factor-like Cidea, highly expressed in thermogenic adipose tissues like brown and beige, regulates the britening/beiging of human adipocytes, thereby regulating energy expenditure[39,40]. HFD significantly increase the levels of Cidea mRNA in the liver and adipose tissue[41]. Knocking down Cidea in the liver of ob/ob mice markedly reduced small lipid droplets and hepatic lipid accumulation[42].

JTTZF also appears to modulate inflammation, a crucial factor in metabolic disorders. Gpnmb is associated with inflammation and serves as a crucial factor in combating obesity-related metabolic disorders by diminishing the inflammatory activity of macrophages[43]. Gpnmb secreted by the liver promoted lipogenesis in white adipose tissue and exacerbated insulin resistance and obesity. Inhibiting or specifically knocking down Gpnmb in the liver increases insulin sensitivity and reduces weight gain[44]. Gprc5b is highly expressed in both mice and human pancreatic islets. It is up-regulated 2.5-fold in islets with T2DM[45]. Gprc5b-deficient mice are protected against diet-induced insulin resistance and obesity by reducing inflammation in white adipose tissue and cytokine-induced apoptosis[45,46]. Fabp family genes encode proteins involved in the transport of long-chain fatty acids. These proteins also play a role in regulating phospholipid synthesis, lipid metabolism, and mitochondrial β-oxidation. The expression levels of Fabp3/Fabp4/Fabp5 increased during the differentiation of preadipocyte[47-49]. Elevated expression of these genes has been observed in obese mice and patients with T2DM[50-55]. Targeted silencing of Fabp4 and Fabp5 in white adipocytes can combat obesity and inflammation and reverse insulin resistance[56].

Metabolomic results showed that JTTZF regulates several metabolites involved in energy metabolism, such as acylcarnitines, LPC, lysophosphatidyl ethanolamine (LPE) and others. These metabolites are critical for energy storage, cell membrane integrity, and signal transduction. Lipids are among the most crucial biomolecules involved in energy storage, cell membranes, signal transduction, and so on[57]. Lipids mainly include lysophosphatidylserine, phosphatidyl inositol, phosphatidyl ethanolamine (PE), phosphatidylcholine (PC), phosphatidic acid, sphingomyelin. PC and PE can be hydrolyzed into LPC and LPE, respectively. These lysophospholipids are closely associated with diabetes-related oxidative stress and inflammation[58]. LPE 18:2 inhibits fatty acid biosynthesis and lipolysis, contributing to lipid droplet formation[59]. HFD upregulate exosomal PC, which exacerbates insulin resistance[60]. LPC is closely related to diabetes-related indicators such as glycosylated hemoglobin, high-density lipoprotein cholesterol, TG, and alanine aminotransferase[61]. In our study, the model group showed significant decreases in several lipids, including LPC 18:3, LPC 18:4, LPE 18:4, LPE 18:3, PC (5:0/13:1), PC (9:0/9:0), LPC 14:1, PC (4:0/16:3), and lysophosphatidylethanolamine 18:2. JTTZF treatment restored these lipid levels, which is associated with the improvement glycolipid metabolic disorder. Acylcarnitines, intermediates of fatty acid metabolism, are early biomarkers of insulin resistance in diabetes in animal studies and a population-based study[62-65]. They are important energy source for β cells, but their accumulation impairs insulin synthesis, energy metabolism, and the TCA cycle, leading to β-cell dysfunction[66]. Acylcarnitines were significantly increased in the model group, JTTZF can improve glycolipid metabolic disorder by decreasing them. Researches have indicated that palmitoyl carnitine was positively correlated with the risk of T2DM. It has been identified as a novel metabolic marker for incident T2DM in two prospective case-control studies involving Chinese adults[67,68]. It can induce significant insulin insensitivity, decrease muscles glucose uptake, and increase blood glucose levels Carnitine palmitoyl transferase 2 knockout in C2C12 cells exacerbated insulin resistance. Adenosine diphosphate (ADP) and Adenosine 5’-monophosphate (AMP) are important participants in energy metabolism processes, the mutual conversion between ATP and ADP is accompanied by the release and storage of energy, related to glycolysis and/or oxidative metabolism[69]. AMP can effectively improve high-density lipoprotein cholesterol, plasma TGs, hepatic lipids, and glucose levels[70,71]. An elevated ADP/ATP ratio activates adenosine 5’-monophosphate-activated protein kinase, which in turn inhibits fat synthesis and promotes fat degradation[69]. This mechanism is implicated in metabolic disorders such as diabetes and obesity. JTTZF improved glucose metabolism in HFD-induced glycolipid metabolic disorder by regulating ADP and AMP via the TCA cycle. The TCA cycle is the hub of amino acid, lipid, and carbohydrate metabolism, serving as a major energy source and pathway for glucose degradation. Hippuric acid levels are decreased in HFD-induced mice, impaired glucose tolerance, and obese patients[72,73]. Increased hippuric acid is linked to improved insulin secretion and fasting glucose levels in high-risk T2DM populations[74]. In our study, hippuric acid was significantly decreased in the model group. JTTZF increased its levels via phenylalanine metabolism, improving glycolipid metabolic disorder. D-ribose plays an important role in T2DM, it reacts with hemoglobin to produce hemoglobin A1c[75]. It is significantly increased in Zucker diabetic fatty rats and diabetic patients, which is consistent with our research findings[76]. 2-hydroxybutyric acid has been demonstrated as a clinical monitoring molecule and functioned as an insulin resistance biomarker, utilized for tracking disease progression during insulin resistance treatment[77]. It is significantly higher in T2DM mice and women with gestational diabetes mellitus, and a significant decrease of 2-hydroxybutyric acid was observed after six months in patients who underwent gastric bypass surgery, which is associated with improved insulin resistance[78,79]. In our study, JTTZF ameliorated glycolipid metabolic disorder by propanoate metabolism. Progesterone, a neurosteroid, plays a role in energy metabolism. Studies found a positive correlation between progesterone and T2DM, it was increased in db/db mice, and low levels of progesterone can reduce the incidence of T2DM[80-82]. Progesterone can lead to apoptosis of rat pancreatic islet primary β-cells and enhance the activity of the rate-limiting enzyme glucokinase to increase glucose metabolism in MIN6 beta-cells[83,84]. It also increased blood glucose via gluconeogenesis when insulin action is limited or impaired[85]. Progesterone receptor-knockout mice showed improved glucose tolerance[86]. In our study, progesterone levels were increased in the model group, JTTZF increased its level by ameliorating glycolipid metabolic disorder via steroid hormone biosynthesis.

In conclusion, the conjoint analysis of transcriptome and metabolome showed that JTTZF could regulate glycolipid metabolic disorder through multiple metabolic pathways. These findings lay the groundwork for future research into JTTZF’s role in precision medicine approaches for metabolic disorders, which suggest that JTTZF has the potential to be developed as a therapeutic strategy for metabolic disorders. Given the increasing prevalence of metabolic disorders, JTTZF could offer a promising alternative or complementary therapy to existing treatments with its ability to modulate key metabolic pathways, could offer a promising alternative or complementary therapy to existing treatments. Future clinical trials should explore the safety and efficacy of JTTZF in human with metabolic disorders, particularly those who are unresponsive to conventional therapies. Long-term studies in human clinical trials are needed to evaluate the sustained effects of JTTZF on glycolipid metabolism and to determine the optimal dosing regimen. Furthermore, integrating JTTZF into multimodal treatment strategies, such as combining it with lifestyle modifications or other pharmacological agents, could be a promising approach to enhance therapeutic outcomes. While our study offers valuable insights into the mechanisms underlying JTTZF’s effects, several limitations of the current study should be acknowledged. Firstly, our study relied on an animal model of glycolipid metabolic disorder induced by a HFD. While animal models are valuable for initial mechanistic studies, they may not fully replicate the complexity and variability of human metabolic disorders. The short treatment duration is another limitation, as it does not allow us to assess the long-term effects of JTTZF on glycolipid metabolism. Additionally, it is important to note that multi-omics approaches are primarily correlational, we did not determine whether the beneficial effects of JTTZF on HFD mice depended on changes in liver metabolites. Establishing causality will require further validation through targeted studies, such as gene knockout or metabolite-targeting experiments. Therefore, the potential mechanism and causality between the anti-glycolipid metabolic disorder effects of JTTZF and changes in genes and metabolites of the liver deserve further investigation. Additionally, future research should explore the underlying mechanisms of JTTZF’s effects in more detail, including its impact on other organs and systems involved in metabolic regulation.

CONCLUSION

In conclusion, JTTZF can ameliorate the glycolipid metabolic disorder induced by HFD-diet by regulating lipid metabolism and improving insulin tolerance. We found that JTTZF can restore 8 genes abnormally expressed in the pathway associated with the glucose and lipid metabolism disorder and 25 DEMs involved in glucose, fatty acid and amino acid metabolism. These genes and metabolites may provide valuable insights into the molecular mechanisms underlying JTTZF’s effects, but its complex mechanisms require further research and validation. Importantly, these findings lay a solid foundation for future investigations into JTTZF’s potential role in precision medicine approaches for metabolic disorders. By elucidating the intricate interplay between gene expression and metabolite regulation, our study offers insights that could inform the design of future omics-based personalized therapies, ultimately contributing to more effective and tailored treatment strategies for patients with metabolic disorders.

ACKNOWLEDGEMENTS

We would like to thank all the authors for their contribution to the realization of this manuscript.