Abstract Objective Food accumulation fever (FAF), a common clinical disease in children, is generally induced by the excessive intake of high-calorie or high-fat foods. Zhiqiao Chuanlian decoction (ZQCLD) is a classical traditional Chinese medicine (TCM) that may have therapeutic effects on FAF. Methods Network pharmacological analyses of ZQCLD and FAF were conducted. Animal experiments lasted for 14 days. Rats in the model, positive control, and low-, medium-, and high-dose groups were fed a high-calorie diet. On days 11–14, the positive group was given a domperidone solution. The low-, medium-, and high-dose groups were administered different concentrations of ZQCLD. The body temperature, gastric emptying rate, and intestinal propulsion rate were measured. Relevant indicators were determined by ELISA. Results The main target proteins included IL-1β, C–C motif chemokine 2 (CCL2), prostaglandin G/H synthase 2 (PTGS2), transcription factor AP-1 (JUN), haem oxygenase 1 (HMOX1), interferon-gamma (IFN-γ), peroxisome proliferator-activated receptor-gamma (PPAR-γ), and inducible nitric oxide synthase (NOS2/iNOS). Compared with those in the control group, body weight, gastric emptying rate, intestinal propulsion rate, and neuronal nitric oxide synthase (NOS1/nNOS) levels were significantly lower in the model group, whereas body temperature and endotoxin, interleukin-1β (IL-1β), PGE2, and iNOS levels were increased. In each treatment group, body temperature and PGE2 levels returned to normal levels. Compared with those in the model group, the gastric emptying rates in the positive group and the low- and medium-dose groups increased; the intestinal propulsion rates were higher in the medium- and high-dose groups, whereas the endotoxin and IL-1β levels were lower; and the nNOS level was higher in the high-dose group, whereas the iNOS level was lower. Conclusions ZQCLD may treat FAF by regulating jejunal IL-1β and nNOS, serum endotoxin, and hypothalamic PGE2 and iNOS levels. Keywords: Traditional Chinese medicine, Aurantii Fructus, Coptidis Rhizoma, Nitric oxide synthase, Network pharmacology 1. Introduction With improvements in living standards in recent years, children's diets have exhibited a trend towards eutrophication. An increasing number of delicious, high-calorie, and high-nutrient foods are available for this purpose. The parents' desire for adequate nutrition for their children may result in a high food consumption by the children [[35]1]. Children may be attracted to the enticing flavours and textures of these foods, leading to excessive intake [[36][2], [37][3], [38][4]]. According to traditional Chinese medicine (TCM) theory, continuous excessive intake of high-calorie or high-fat foods may result in bloating or abdominal pain, hiccups, nausea and vomiting, loss of appetite, constipation, or loose stools, which is called food accumulation (FA). This is similar to the functional dyspepsia (FD) in modern medicine [[39]5]. Owing to an underdeveloped digestive system, children are more susceptible to damage from overeating [[40]6]. This results in FA being more common in children than in adults. Furthermore, a particular phenomenon has been observed in clinical practice in which FA can lead to fever without coinfection. This condition is called FA fever (FAF) and has been documented in ancient TCM books. Because there is no established standard definition or accepted diagnostic criteria, FAF is typically treated as one of two independent disorders: dyspepsia or fever. Consequently, antibiotics, antipyretics, and digestive system stimulants are frequently used to treat the symptoms. However, these separate therapeutic approaches can quickly result in disease recurrence and other gastrointestinal symptoms [[41]7,[42]8]. When FA and FAF are not treated in a timely manner, they may further affect the subsequent growth and development of children [[43]9,[44]10]. Therefore, it is necessary to study FAF, which has been well recorded in ancient texts, but has received little attention in recent years. Previous animal studies have shown that long-term (>21 days) free intake of high-calorie, high-sugar, or high-fat diets can lead to decreased gastrointestinal motility in animal models [[45]11,[46]12], accompanied by gastrointestinal mucosal barrier damage [[47]13] and local intestinal inflammation [[48]14,[49]15]. Moreover, systemic inflammation may increase [[50]16]. Some short-term experiments (6–21 d) [[51][17], [52][18], [53][19], [54][20]] revealed that these dietary patterns have an impact on gastrointestinal motility and the gastrointestinal mucosal barrier, accompanied by local intestinal inflammation. However, the impact of short-term but excessive high-calorie dietary intake on young individuals remains unclear. Moreover, the reason for the FA-induced increase in body temperature due to FA in children remains unclear. The present study was conducted to address this issue. In China, TCM has been used for more than 2000 years. With the advantages of abundant resources, few side effects, stable efficacy, and multiple pathways of action, TCM has shown significant therapeutic potential in many basic and clinical studies [[55]21]. Zhiqiao Chuanlian decoction (ZQCLD) is a classical TCM prescription composed of Aurantii Fructus and Coptidis Rhizoma, which was recorded in ancient texts to relieve FAF caused by consuming excessive amounts high-fat or high-sugar foods. ZQCLD is composed of only two drugs, which avoids potential effects of excessive medication in children and is therefore suitable for the treatment of FAF in children. Previous studies on the effects of Aurantii Fructus and Coptidis Rhizoma and their underlying mechanisms have shown that Aurantii Fructus regulates gastrointestinal motility and alleviates intestinal inflammation [[56]22,[57]23], whereas Coptidis Rhizoma has anti-inflammatory and antipyretic properties [[58]24,[59]25]. These studies suggest that ZQCLD may play a therapeutic role by restoring gastrointestinal motility, preventing inflammation, and reducing body temperature. However, this hypothesis remains to be verified. Network pharmacology is a new branch of pharmacology that uses computer models [[60]26], network analysis methods [[61]27,[62]28] and principles of systems biology to analyse the multi-component, multi-target, and multi-pathway synergistic relationships between drugs, diseases, and targets [[63]29]. The mechanism of TCM prescription involves multiple targets and levels, similar to the integrity, systematisation, and comprehensiveness of network pharmacology. Network pharmacology is a promising method for revealing the potential mechanisms and targets of TCM in disease intervention [[64]30]. Therefore, this study aimed to explore (i) the mechanisms by which FA causes fever and (ii) the therapeutic effects of ZQCLD through network pharmacology analysis and animal experiments to enhance the scientific evidence in support of the TCM theory and provide options for the diagnosis and treatment of FAF in children. 2. Materials and methods 2.1. Network pharmacology analysis 2.1.1. Screening of the active ingredients and targets of ZQCLD The Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP) ([65]https://tcmsp-e.com/tcmsp.php) [[66]31] was used to screen the chemical components of ZQCLD (Aurantii Fructus and Coptidis Rhizoma) with the following criteria: oral bioavailability ≥30 % and drug-likeness ≥0.18 [[67]32]. Targets for these active components were identified using the TCMSP database and standardised using the UniProt database ([68]https://www.uniprot.org) [[69]33]. 2.1.2. Identification of disease-related targets and disease–drug action targets The GeneCards ([70]http://www.genecards.org), OMIM ([71]http://www.omim.org), and DisGeNET ([72]http://www.disgenet.org) databases were used to search for disease-related targets, using the queries ‘food accumulation’, ‘food retention’, ‘dyspepsia’, and ‘fever’. Because the number of targets was large, targets in the GeneCards database with relevance scores higher than the median were chosen, and the operation was repeated once [[73]34]. The targets obtained using the queries ‘food accumulation’, ‘food retention’, and ‘dyspepsia’ were summarised and deduplicated. The FAF-related targets were obtained by identifying the intersection of the above targets, and the targets were obtained using the query ‘fever’ using Venny 2.1.0 ([74]https://bioinfogp.cnb.csic.es/tools/venny). Next, the intersection with the targets of ZQCLD was identified to obtain the disease-drug targets. Inflammation may be a key factor in FAF; therefore, the Gene Ontology (GO) functions of disease drug targets were analysed against the UniProt database. Disease-drug targets related to ‘inflammatory response’ were selected for further studies. 2.1.3. Network construction and analysis A protein-protein interaction (PPI) network was constructed using the STRING database ([75]https://cn.string-db.org) with species limited to Homo sapiens. Core targets in the PPI network were selected using Cytoscape 3.8.0 [[76]35] with the CytoNCA plugin [[77]36]. Targets with scores above the median were selected for this study. The component–target network was constructed using Cytoscape 3.8.0, and the NetworkAnalyzer plugin was used to analyse the topology parameters. 2.1.4. Function enrichment and pathway analysis GO function enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of the core targets were performed using R 4.4.2 with the R packages ‘clusterProfiler’ [[78]37], ‘pathview’ [[79]38], ‘org.Hs.eg.db’, and ‘ggplot2’. A threshold of P < 0.05 was set. GO enrichment analysis was conducted for the biological process (BP), molecular function (MF), and cellular component (CC) categories. 2.2. Animals and feed Sixty 4-week-old SPF-grade healthy Sprague-Dawley rats (120 ± 10 g, 30 male and 30 female) were purchased from SiPeiFu Biotechnology Co., Ltd. (Beijing, China, licence number: SCXK (Beijing) 2019-0010). Rats were allowed to acclimate for 1 week before the experiments. All animals were housed in the animal laboratory of the Beijing University of Chinese Medicine. Ordinary irradiated feed was provided by SiPeiFu Biotechnology Co., Ltd. (Beijing, China). Special high-calorie feed was prepared from rice crust, chocolate wafers, beef grains, and commercial wheat flour in a 1:2:2:1 ratio. The high-calorie feed had the same appearance and texture as ordinary feed. SiPeiFu Biotechnology Co., Ltd. (Beijing, China) was performed the procurement of raw materials, feed production, radiation sterilisation, and quality control. 2.3. Animal feeding and material collection The 60 rats were divided into control, model, positive, low-, medium-, and high-dose groups according to a random number table, with five males and five females in each group. To avoid experimental bias owing to subjective human factors, the principle of blinding was followed. During the whole process of the animal experiment, the operator and recorder were unaware of the group and specifics of the experimental animals. The experiment lasted for 14 days. During the experiment, all rats except those in the control group were given high-calorie feed and a special high-calorie suspension (Supplementary information 1) by gavage (1 mL per 100 g body weight) once a day. The rats in the control group were provided with basic feed and an equal volume of pure water. From days 11–14, rats in the positive control group were administered domperidone solution (0.315 mg/mL) (Supplementary information 1 and 2) by gavage (1 mL per 100 g body weight) once a day. Rats in the low-, medium-, and high-dose groups were administered ZQCLD solution (0.0475, 0.095, and 0.19 g/mL, respectively) (Supplementary information 1 and 2) by gavage (1 mL per 100 g body weight) once a day. The rats in the control and model groups were administered an equal volume of pure water. On the morning of day 15, after overnight fasting with free access to water, each rat was administered 2 mL (2 g) of alimentary semisolid paste (Supplementary information 1) by gavage. After 30 min, the rats were anaesthetised by intraperitoneal injection of 20 % urethane (0.7 mL per 100 g body weight). Blood was collected from the abdominal aorta, held at room temperature for 4 h, and centrifuged at 1000×g for 20 min. The serum was stored at −80 °C. The upper and lower ends of the rat gastric body were knotted, and the gastric body and small intestine were removed at 4 °C. Brain tissue was removed and placed on ice. The hypothalamus was immediately harvested, weighed, fully homogenised, and centrifuged at 5000×g for 10 min at 4 °C. The supernatant was stored at −80 °C. After the measurement of the intestinal propulsion rate, 200 mg of jejunal tissue was collected, fully homogenised, and centrifuged at 5000×g for 10 min at 4 °C. The supernatant was stored at −80 °C. 2.4. Detection of signs Prior to the experiment, the rectal temperatures of the rats in each group were recorded. Every morning, the rats were weighed and their daily food intake was recorded. The rectal temperature of the rats was measured. After material collection, the gastric emptying and intestinal propulsion rates were calculated (Supplementary Information 2). The levels of endotoxin in the serum, IL-1β and nNOS in the jejunum, and PGE2 and iNOS in the hypothalamus were measured using ELISA kits (Supplementary information 1). 2.5. Statistical analysis Statistical analysis was performed using SPSS 22.0, and GraphPad Prism 9.0.0. The results are expressed as the mean ± standard deviation (SD) (X ± s). One-way ANOVA was conducted to compare data between groups, followed by LSD test. Differences were considered statistically significant at P < 0.05. Chi-squared test, correlation analysis, and principal component analysis (PCA) were performed using SPSS 22.0, and R 4.2.2. The R packages ‘ggstatsplot’, ‘factoextra’, ‘corrplot’, ‘ggplot2’, and ‘FactoMineR’ [[80]39] were used. 3. Results 3.1. Active ingredients and targets of ZQCLD Nineteen components were obtained from ZQCLD: five belonged to Aurantii Fructus and 14 belonged to Coptidis Rhizoma. A total of 16 compounds in ZQCLD had corresponding targets: five belonged to Aurantii Fructus and 11 belonged to Coptidis Rhizoma (Supplementary information 3, [81]Table S1). After standardising the target names using the UniProt database and removing duplicates, 112 drug targets were obtained (Supplementary information 3, [82]Table S2). 3.2. Disease-related targets and disease–drug action targets A total of 1176 disease-related targets were identified after screening for deduplication and intersection. By intersecting the disease and drug targets, 67 disease–drug intersection targets were obtained, 20 of which were related to the ‘inflammatory response’ (Supplementary information 3, [83]Table S3). 3.3. PPI network and component–target network analyses A PPI network of disease–drug action targets was constructed using the STRING database and Cytoscape 3.8.0. It contained 20 nodes and 89 edges, and the average node degree was 8.9 ([84]Fig. 1A). The following core targets were obtained through the CytoNCA plugin: IL-1β, C–C motif chemokine 2 (CCL2), prostaglandin G/H synthase 2 (PTGS2), transcription factor AP-1 (JUN), haem oxygenase 1 (HMOX1), interferon-gamma (IFN-γ), peroxisome proliferator-activated receptor-gamma (PPAR-γ), and NOS2 ([85]Fig. 1B). The component–target network was constructed using Cytoscape 3.8.0 ([86]Fig. 1C). It contained 34 nodes (14 nodes belonging to the components and 20 nodes belonging to the targets) connected by 44 edges. The core components of ZQCLD, including quercetin (degree = 17), nobiletin (degree = 4), and naringenin (degree = 3), play important roles in this network, mainly by targeting PTGS2 and NOS2 (degree >2). Fig. 1. [87]Fig. 1 [88]Open in a new tab Network construction. A PPI network of disease-drug targets. B Selection of core targets. Yellow squares represent core targets. The size represents the degree of the relevant target. C Compound-target network. Light blue ellipses represent drug targets. The size represents the degree of the relevant target. Light yellow squares with green circles represent compounds from Coptidis Rhizoma. Light yellow squares with purple circles represent compounds from Aurantii Fructus. (For interpretation of the references to colour in this figure legend,