Q-Exactive Plus mass spectrometer (Thermo Scientific); UHPLC Nexera LC-30, an ultra-performance liquid chromatography system (Shimadzu); Low temperature high-speed centrifuge (Eppendorf 5430R); Chromatographic column: Waters, Acquity UHPLC CSH C18, 1.7 μm, 2.1 × 100 mm column; acetonitrile (Thermo Fisher), isopropyl alcohol (Thermo Fisher), methyl alcohol (Thermo Fisher), and ammonium formate (Sigma-Aldrich, 70221) were employed in this study.

IL-1β, IL-6, IL-2, and TNF-α ELISA kits were obtained from Neobioscience Biotechnology Co., Ltd. (Shenzhen, China). We purchased Enterotoxigenic Escherichia coli O101 (bacteria number: CVCC231) from the Chinese Institute of Veterinary Drug Control.

Table 1 shows the relevant information for PD. The mixture of Pulsatillae radix (Bge.) Regel, Phellodendron chinense Schneid, Coptidis chinense Franch and Fraxinus rhynchophylla Hance was soaked for approximately 40 min in 10 x (v/w) distilled water and boiled for 30 min, twice. The filtering medium was collected and concentrated to 1.00 g/mL (crude drugs).

The main components of PD were further analyzed by HPLC (Agilent Technologies, Santa Clara, CA, United States), with an Agilent 1260 HPLC system and UV detection system. The analytes were isolated using an Agilent ZORBAX SB C18 column (4.6 × 250 mm, 5 μm) at a column temperature of 30°C. The mobile phase comprised of A (acetonitrile) and B (0.1% H₃PO₄ acidified water). The detection wavelengths were set at 205 and 335 nm, and the absorption spectra of compounds were recorded between 205 and 340 nm.

Thirty-two Wistar rats (weight, 180–220 g; age, 6–8 weeks; SCXK (Gan) 2015-0006) were obtained from the experimental Animal Center of Lanzhou University. Rats were housed at room temperature (22 ± 3°C) and constant humidity (55 ± 15%) under a natural light cycle in the laboratory. Food and tap water were provided ad libitum. All procedures in this study were strictly performed according to the Guidelines of the Animal Care and Use Committee of Gansu Agricultural University, and the local Animal Research Welfare Committee approved the study. Rats were randomly divided into four groups: model group, control group, PD group, and self-healing group.

Based on our previous study, the experiment comprised of four stages: stage 1, the high fat and high sugar stage (10 days) where rats in the model, self-healing, and PD groups were fed sufficient chow combined with an oral gavage of lard (4 mL/200 g weight) on alternate days; stage 2, the high temperature and high humidity environment stage (5 days) where white wine (56 degrees Red Star Erguotou, 2 mL/200 g weight) were given to rats in the model, self-healing, and PD groups by gavage every morning, and the rats placed in a high temperature (33°C ± 2°C) and humid (93% ± 2%) chamber for 8 h once daily; stage 3, the Escherichia coli injection stage (3 days) where rats in the model, self-healing, and PD groups were intraperitoneally injected with 1.06 × 109 CFU/mL Escherichia coli suspension (0.2 mL/200 g) twice at intervals for 24 h, followed by feeding in a natural environment for 1 day. Rats in the control group were fed standard diet in the natural conditions and intraperitoneally injected with an equal volume of normal saline for 18 days. Rats in the model, self-healing, and PD groups were also given honey water (30%) in all three stages (18 days). For the final (fourth) stage (the treatment stage), the PD crude drug (2 g/kg × 1.88 = 3.76 g/kg/day, the formula for the dose translation was as follows: human dose of crude herbs in clinic × 0.0188/200 × 1,000 × the multiple of clinical equivalency dose. The human dose of crude herbs in clinic is 40 g in the Table 1. The dose of the PD extracts was equivalent to that of crude herbs based on the TCM prescription and the multiple of clinical equivalency dose is 1.) was administered by gavage once per day for 5 consecutive days; rats in the control and self-healing groups were gavaged with a volume of normal saline. Rats in the model group were sacrificed at this stage.

On the 24th day, serum samples were collected by administering 1% pentobarbital anesthesia to rats for pain relief. Serum samples were used to analyze metabolite variations and the results were shown in reference (Hua et al., 2019). The ileum and colon were removed, quickly subpackaged, and immediately fixed in 10% neutral formalin. A section of the colon was also obtained and stored at −80°C for lipidomics analysis.

Using the LipidSearch^TM software version 4.1, peak recognition, lipid extraction (secondary appraisal), peak, peak alignment, quantitative processing, etc. were performed using the following parameters: precursor tolerance, 5 ppm and product ion threshold, 5%. Lipid molecules with a RSD >30% were deleted. LipidSearch^TM processing software can perform original data extraction, lipid and peak identification, peak alignment, and quantitative analysis of integration. LipidSearch^TM includes eight categories, 300 class types, and approximately 1.7 million types of lipid molecules in the MS2 and MS3 databases. Data from the Orbitrap^TM mass spectrometer can be used to generate high resolution, high quality, fine degree, parent ion, ion and neutral loss scanning identification algorithms, and perform reliable lipid qualitative analysis systematically (Supplementary Table S1).

The missing value in the group with >50% of lipid molecules was deleted and the abundance values were obtained using total peak area normalization based on LipidSearch data extraction. Multivariate data analysis was performed using the software SIMCA-P 14.1 (Umetrics, Umea, Sweden) after the process of Pareto-scaling. Multi-dimensional statistical analysis including unsupervised main points analysis, PCA, least squares discriminant analysis supervision (PLS-DA), and orthogonal partial least squares discriminant analysis (OPLS-DA) were used to screen different lipids. Unidimensional statistical analysis was performed using the Student’s t-test, and multiple variation analysis and R software map were employed for the volcano, hierarchical cluster analysis, and correlation analysis.

The experimental data were expressed as mean ± S.D. ANOVA followed by the Duncan’s multiple range test was performed using SPSS16.0 (Chicago, United States). A P-value <0.05 was considered statistically significant.

The HPLC chromatogram of the standard mixture and samples of PD is shown in Figure 1. All samples were significantly separated at the retention time of 20 min. Aesculin, esculetin, jateorhizine, palmatine, berberine, and anemoside B4 were quantitatively determined and the respective contents were 43.98, 31.13, 10.54, 15.31, 108.29, and 232.95 mg in 1 kg of the PD Crude herb.

Throughout the entire process of the experiment, rats in the control group did not display any clinical symptoms (23 days, Figure 2). At the first stage (10 days), rats in the three groups (model, self-healing, and PD groups) were fed a high-fat feed and honey sugar water. The excretory waste was characterized as soft and wet, and eventually became loose and appeared dry and yellow.

The waste collected from rats in the model, self-healing, and PD groups was soft and wet. In the days of fasting from water intake, urine output was increased and a pale-yellow color was apparent; urine during the days of feeding displayed high fat storage. The stool of rats went from soft and wet to loose, then finally appearing dry and yellow. Meanwhile, increased water intake, frequent urination and yellowish urine were observed, with a decrease observed on feeding days. The weight of rats increased on days of feeding and decreased on days of fasting. At the second stage (5 days), rats exhibited signs of depression and appetency, their hairs became lusterless, and stools appeared sticky and yellow, before becoming loose. In this stage, the weight of rats gradually decreased. At the third stage (3 days), rats in the model, self-healing, and PD groups displayed serious depression, sluggishness, chills, decreased food consumption, eyelid swelling and increased secretion, followed by diarrhea and increasing temperature. Meanwhile, stools were loose, sticky, yellow, pungent and overpowering; some stools also contained bloody mucus (Figure 2). Changes in weight were not significant over the 3 days; however, a gradual increase occurred in the subsequent experiments (Figure 3).

After PD administration (1 day), the spirit of rats began to recover, and a gradual increase in diet consumption and activity was observed. At 5 days following PD administration, symptoms were restored to normal and weights gradually increased.

These results demonstrated that the rat DHD model was successfully established, and the clinical symptoms and signs exhibited by the model rats could be improved by treatment with PD.

Table 2 shows the increased levels of WBC, NE, MO, RBC, LY, HGB, and PCV in the self-healing and model groups compared to the control group (P < 0.05). These indexes for the PD treatment group were also lower than in the model group. Based on these results, strong inflammatory response and dehydration in the process of DHD were demonstrated, with PD possessing the ability to alleviate these symptoms.

The levels of TNF-α, IL-1β, IL-6, and IL-2 in the model group were increased compared to the control group (P < 0.05) (Figure 4), while the levels of these serum inflammatory cytokines decreased in the PD group when compared to the model group. These results are consistent with those obtained in blood routine detection.

As shown in Figure 5, mucosal epithelium was complete and possessed a normal fuzzy morphology which cannot be observed in the LP venous congestion in the control group (Figures 5, 6A–A1). The integrity of the ileum and colon epithelium was disrupted, and inflammatory cellular infiltration and congestion were observed (Figures 5, 6B–B1) in the DHD rats and (Figures 5, 6C–C1) in the self-healing. In the PD group (Figures 5, 6D–D1), histopathological changes including the integrity of the ileum and colon epithelium, inflammatory cellular infiltration and congestion were effectively ameliorated or eliminated following treatment with PD.

Glycerophospholipid metabolism plays an important role in the DHD process. This is because diarrhea results from abnormalities in intestinal electrolyte transport (Singh et al., 2014; Das et al., 2018) which contributes to fluid and electrolyte homeostasis (Das et al., 2018). Intestinal ion transport mechanisms play a significant role in diarrhea (Sheikh et al., 2018). Therefore, the possible novel target for diarrhea therapy is the intestinal ion transporters. Some studies have reported that LPA, a form of phospholipids similar to growth factors, have been shown to prevent or restore gastrointestinal diseases, such as stomach ulcer and diarrhea in rats (Lin et al., 2010; Afroz et al., 2018). Diarrhea is caused by a decrease in the absorption of fluids and salts resulting from intestinal absorption and secretion disorders. Substantial absorption of sodium and water in the intestine is mediated by sodium/H + exchanger 3 (NHE3) found in the lumen of the intestinal cells. LPA is an effective stimulant of NHE3 and intestinal fluid absorption, which transmits signals through LPA5. As the regulation of LPA5 depends on its interaction with NHERF2, LPA may be useful in treating some diarrheal diseases (Lin et al., 2010). As LPA is a phospholipid acid (PA) formed by the activity of digesting PLA2, PA is a potential ingredient in the treatment of this gastrointestinal disease (Tanaka et al., 2012). The conversion of PA into diacylglycerol (DAG) by LPPs is the commitment step for the production of PC, PE, and PS. DAG is also converted into CDP-DAG, which is a precursor for PG, phosphatidylinositol (PI), and phosphoinositides (PIP, PIP2, PIP3) (Daum, 1999). In our results, the levels of PC, PE, PG, and PS were decreased in DHD rats compared to the levels in control rats. This is because of the decrease in production of LPA in the DHD process. LPA has been reported to ameliorate epithelial damage in chemical-induced colitis in rats (Lee et al., 2011). Through in vivo study, LPA has also been identified as a key regulator of epithelial cell proliferation and migration in the intestine. As loss of LPA impedes mucosal restitution of wounded areas in the colon, demon-FIG 9 LPA1 requires epithelial mucosal wound repair, reinforcing the clinical importance of LPA1 (Lee et al., 2013). PD also demonstrated the potential to increase the levels of these metabolites. Thus, treatment of DHD with PD has a relationship with intestinal ion transporter mechanisms. Owing to this, we believe PD can promote the production of LPA, which can inhibit intestinal chloride secretion in preclinical in vitro studies and in NHE3 upregulation of DRA expression (Singla et al., 2012), and contribute significantly to PD treatment of DHD.

Saturated PC is an essential lipid component of pulmonary surfactant (Li and Vance, 2008) and PC has antibacterial effects. A previous study showed that PD also has antibacterial effects (Yang Y. et al., 2018). Therefore, the interference of PD with PC may change the ability of the body to resist bacteria.

AA is an important component of the cell membrane, and possesses fluidity and flexibility. The four double bonds of AA contributes to its ability to easily oxidize, thereby producing excessive metabolites, which are of great significance to the normal function of the immune system, promotion of allergy and inflammation, response to inflammation, and appetite stimulation (Hanna and Hafez, 2018). In the present study, the decreased levels of AA, TNF-α, IL-1β, IL-6, and IL-2, and the pathological observations demonstrated that dampness-heat induced diarrhea is associated with inflammation. Indeed, inflammation plays an important role in DHD and this has been demonstrated (Yao et al., 2017; Schiller, 2018). Interestingly, some studies have proven that lipid metabolism disorder could increase the incidence of inflammatory bowel disease (Fujimoto et al., 2013; Paik et al., 2013). In addition, high-sugar and high-fat diet can induce AA metabolism disorder, and oxidative stress is the pathogenic link that bridges over-nutrition with inflammation via NF-kB activation. Activated NF-kB can trigger the generation of inflammatory molecules such as TNF- α and IL-6 (Kalivarathan et al., 2017). The levels of these molecules are also increased in high-sugar and high-fat diet-fed rats because of NF-kB activation. Heat stress can cause hemorheology disorder, and extensive microcirculatory obstruction. Microcirculatory obstruction may further result in local hypoxia and energy metabolism disturbance, as demonstrated by the severe inhibition of oxidative phosphorylation. Together with the stimulation of the biological factor of Escherichia coli, colonic epithelium was destroyed. Therefore, fat metabolism disorder and microcirculatory obstruction are the pathological bases of colon injury in DHD, and is caused by AA metabolism disturbance induced by high-sugar, high-fat diet, high temperature, and a highly damp environment.

Degradation of glycerophospholipids by PLA2 can cause AA release (Frisardi et al., 2011). Non-enzymatic and enzymatic oxidation of AA produce several lipid mediators (Yang L. et al., 2018), all of which are closely associated with the glycerophospholipid pathways involved in DHD; this suggests that an interplay among lipids occurs in the colon tissue (Kasuga et al., 2015). The results demonstrated that the level of AA was significantly increased in the model group, whereas a significant decrease in AA was observed when compared to the model group, with restoration to normal level observed in the PD group. These results indicate that PD may exhibit treatment effects in DHD by suppressing the production of AA. In addition, a relationship is demonstrated to exist with the increase in glycerophospholipids.