DWYG (E Yao Zhi Zi: Z20113160) is mainly composed of Radix Rehmanniae Praeparata (Rehmannia glutinosa, SDH), Chinese Magnolcavine Fruit (Schisandrae Chinensis Fructus, WWZ), Virgate Wormwood Herb (Artemisiae Scopariae Herba, YC), turmeric (Curcumae Longae Rhizoma, JH), and licorice roots of north-western origin (Radix Glycyrrhizae, SGC). Simulated DWYG (1/20 of the original DWYG with coloring excipients and flavor excipients). DWYG and simulated DWYG were obtained from the manufacturing laboratory of Hubei Provincial Hospital of Traditional Chinese Medicine.

Patients with HIV/AIDS who visited Wuhan Jinyintan Hospital between December 2019 and March 2021 were selected for this clinical trial. The trial included individuals who met the following inclusion criteria. 1) Patients should meet the diagnostic criteria for AIDS-INR. 2) Patients must be between 18 and 65 years old. 3) Patients must the belong to the yin deficiency type in the liver and kidney through TCM syndrome differentiation. 4) Patients must sign the informed consent form and thus can be followed up on time.

The exclusion criteria were as follows. 1) Serious opportunistic infections in patients are not controlled before enrollment. 2) Patients that have multi-system diseases. 3) Patients with tumors or allergies 4) Patients who do not fully understand the trial or cooperate well. 5) The investigator considered the patients unfit to participate in the trial in certain cases.

This controlled clinical trial was approved by the Ethics Committee of Hubei Hospital of Traditional Chinese Medicine (No. HBZY2018-C23-01).

The present study employed a randomized, controlled, double-blind, single-simulation method. The study registration number of the Chinese Clinical Trial Registry (ChiCTR) is ChiCTR1900024673. SPSS 19.0 software was used to divide the patients who met the inclusion and exclusion criteria into two groups according to a random number table and using the HAART medication regimen. Patients in the control group were administered simulated DWYG, whereas those in the trial group were administered DWYG. The dosage is 3 times a day, 5 capsules each time.

The ratio of CD4⁺ to CD8⁺T cells in peripheral blood at baseline and after 24 weeks of treatment were observed and recorded. The samples were examined using flow cytometry (BD FACSCanto II). The proportion of T lymphocyte subsets was determined using Diva 8.0.2 Software. The reagents used included CD4 FITC/CD8 PE (Becton, Dickinson and Company).

The SAS 9.1 software was used for Statistical analysis. Normally distributed measurement data are expressed as mean ± SD, and skewed distribution data are expressed as M (P₂₅, P₇₅). Baseline data were compared between the groups using the t-test, and 24 week data were compared between the groups using the Mann-Whitney rank-sum test. Differences between 24 weeks and baseline were compared between groups using the Wilcoxon rank-sum test. All statistical tests were two-sided, and p < 0.05 was considered statistically significant.

Waters Xevo G2-XS QTOF, ACQUITY UPLC M-Class, Waters ACQUITY REH C18 (1.7 μm, 2.1 × 100 mm). The following chromatographic parameters were used: column temperature, 40°C; mobile phase A, 0.1% formic acid in water; and mobile phase B, methanol. Gradient elution was employed with the following specifications: 0 min, 10% B; 15 min, 55% B; 35 min, 90% B; 40 min, 98% B; 45 min, 98% B; 46 min, 10% B; and 50 min, 10% B. The flow velocity was set at 0.3 mL/min, and the sample size was 2 μL.

The mass spectrometry conditions were as follows: ESI ion source was used to collect data in positive ion mode; primary and secondary mass spectrometry ions were collected in MSE mode, and the collection range was m/z 50–1,500 Da, cone voltage was 60 V, electrospray voltage was 3000 V, and the ESI ion source temperature was 100°C. The desolvation temperature was 500°C, the curtain gas flow rate was 50 L/h, and the desolvation gas (N₂) flow rate was 600 L/h. The primary collision energy was 10 eV, the secondary collision energy was 35 eV, the collision energy fluctuation was ±10 eV, and an enkephalin (m/z 556.2771) tuning solution was employed as the real-time calibration solution. Based on the liquid-phase retention time, accurate molecular weight, mass spectrometry, and secondary fragmentation (Shi et al., 2023).

The identified compounds were imported into the traditional Chinese medicine systems pharmacology database and analysis platform (TCMSP) (Ru et al., 2014) and Swiss Target Prediction (Gfeller et al., 2014) databases to predict the targets of the active ingredients based on the probability criterion (probability >0). The keyword “INR” was searched in the Gene Cards databases (Safran et al., 2010), and the therapeutic targets associated with INR were summarized. A Venn diagram was constructed to visualize the intersection of the compounds and disease targets. Genes intersecting the Venn diagram were imported into the STRING database (Szklarczyk et al., 2023) for protein-protein interaction (PPI) analysis. The STRING database was then imported into Cytoscape for topological analysis. CytoNCA (Tang et al., 2015) was used to perform the calculations, and R was employed to filter the disease targets according to the criteria of the two conditions for the analysis of core targets in the PPI. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) (Kanehisa and Goto, 2000) analyses were performed using the ClusterProfile R package (Xu et al., 2024). The GO item and KEGG pathway (p < 0.05) were considered effective.

The two-dimensional (2D) structure of the core ingredients of DWYG was obtained from the PubChem database (Kim et al., 2023). A three-dimensional (3D) structure was constructed using ChemBio3D and the structure of the small-molecule ligand was obtained after structural optimization. The 3D structure of the protein receptor was obtained from the PDB database (Berman et al., 2020). The water molecules and small-molecule ligands of the protein receptor were removed using PyMOL, and the protein receptor was hydrogenated to determine the active pocket using AutoDockTools (Morris et al., 2009). Molecular docking was performed to calculate the binding energy using AutoDock Vina. The molecular docking effect was relatively stable, and the binding activity is strong when the binding energy is < −5 kcal/mol. The specific molecular docking position was identified and visualized in 3D using PyMOL software.

Twenty Sprague-Dawley (SD) rats, comprising an equal number of males and females, were used in this study. The rats were randomly and evenly divided into two groups: DWYG-G and blank control (BC-G). Rats in the BC-G group were fed 20 mL of 0.9% saline, and rats in the DWYG-G group were fed 20 mL of DWYG twice daily for 3 days. Prior to obtaining blood samples from the rats, a 12-h fast was implemented, while water was still permitted. Blood was collected via carotid artery catheterization 1 h after the rats were fed and kept awake. After allowing the blood to rest for 30 min, centrifugation was performed at 2,500 rpm for 15 min. Subsequently, the serum was transferred to a clean test tube and inactivated in a water bath at 56°C for 30 min. The serum was then divided into Eppendorf tubes with 10 mL per tube and stored at −20°C.

Approximately 10 mL of PMBCs was extracted from 10 HIV INRs and 10 HIV patients with normal immunologic reconstitution (NIR), and heparin anticoagulants were prepared. The PMBCs in the above two types of patients were isolated using the Ficoll method and added to medicine-containing serum and non-medicine-containing serum, respectively, and then cultured in a cell incubator. The groups were as follows. Group (1) comprised PBMC from NIR- and medicine-containing sera (NIR + M, n = 10). Group (2) comprised PBMC from NIR and non-medicinal serum (NIR + N, n = 10). Group (3) comprised PBMC from INR and medicine-containing serum (INR + M, n = 10). Group (4) comprised PBMC from INR and non-medicine sera (INR + N, n = 10).

RNA was extracted from the resuspended cells using TRIzol reagent. qRT-PCR was performed using the same amount of RNA and the One-Step PrimeScript RT-PCR Kit. The relative levels of TLR4 in different samples were calculated, and their trends were analyzed using the 2^−△△CT method using GAPDH as the reference gene. Flow cytometry was used to detect the content of major cytokines in the supernatant.

As shown in Table 1, no differences in baseline characteristics were found among the 62 patients in this study.

As shown in Table 2, the ratio of CD4⁺ to CD8⁺T cells in patients in both groups increased after 24 weeks of treatment. Compared with the baseline, the increase in the CD4:CD8 ratio in the trial group was significantly greater than that in the control group (p < 0.05).

As shown in Figure 1, the statistical analysis showed that after 24 weeks of treatment, the proportion of CD4⁺ and CD4⁺CD25⁺CD127^low in both groups showed an upward trend. Moreover, compared to the control group, the increase in the experimental group was more significant with respect to the proportion of CD4 (mean 4.057% vs 2.670%, p < 0.01) and CD4⁺CD25⁺CD127^low (mean 0.99% vs 0.867%, p < 0.01). The proportion of CD8 in the two groups showed a downward trend, and compared to the control group, the decrease in the experimental group was more significant (mean −5.54% vs −4.03%, p < 0.01).

The compounds in the 80% methanol extract of DWYG were analyzed by UPLC-QTOF-MS/MS. A total of 210 chemical components were identified. The number of compounds from JH, SGC, YC, WWZ and SDH were 82, 56, 15, 43 and 4, respectively (Shi et al., 2023). Extraction-ion chromatograms for each compound are shown in Supplementary Figure S1, and chemical information for the 210 components in DWYG are shown in Supplementary Table S1. A total of 780 alternative compound targets were obtained by searching 210 chemical components using the Swiss Target Prediction and TCMSP databases. Furthermore, 958 INR-associated targets were retrieved from the GeneCard database. Additionally, 182 potential therapeutic targets of DWYG were identified using the Venn diagram, as depicted in Figure 2.

GO enrichment analysis of 182 potential therapeutic targets was performed (Figure 3). The first ten items were plotted in terms of the horizontal gene ratio and vertical count. The biological process mainly involved response to xenobiotic stimulus, response to molecules of bacterial origin, and positive regulation of the MAPK cascade, while the cellular component mainly involved the external side of the plasma membrane, membrane raft, membrane microdomain, and molecular function (MF), which mainly involved protein serine/threonine kinase activity, protein tyrosine kinase activity, and amide binding.

As shown in Figure 4, KEGG pathway analysis showed that 182 potential therapeutic targets were involved in 173 signaling pathways. The first 30 pathways sorted by p-values were further analyzed. Because the main target disease of this study was INR, the signaling pathways related to other diseases, such as cardiovascular disease and cancer (specific types), were removed, and the signaling pathways of infectious disease (viral and immune system subcategory) were retained. Including the Toll-like receptor signaling pathway, Th17 cell differentiation, Kaposi sarcoma-associated herpesvirus infection, human cytomegalovirus infection, hepatitis B, coronavirus disease-COVID-19, measles, Epstein-Barr virus infection, and influenza A, i.e., total of nine pathways were retained.

The 182 potential therapeutic targets were analyzed using STRING to further explore the interactions between the potential target proteins, which were screened to further explore the mechanism of action of the drug. The minimum required interaction score was set to high confidence (0.7) and network topology analysis was performed on all identified therapeutic targets to build an interaction network. The network consisted of 181 nodes and 1,091 edges with an average node degree of 12.1. Moreover, network centrality analysis was performed using CytoNCA and R programming languages, and 19 nodes and 264 edges were obtained, as shown in Figure 5. The central proteins of the PPI network included TLR4, TNF, STAT3, and JAK2, which may play important roles in DWYG’s promotion of immune reconstitution in DWYG.

As shown in Figure 6, combined with the GO and KEGG enrichment results, further molecular docking verification of TNF, MAPK14, CASP3, STAT3, MAPK3, TLR4, RELA, CCND1, JAK2, MAPK1, BCL2, and AKT1 was performed, and the results showed that the binding energy was < −5 kcal/mol.

As shown in Table 3, the expression level of TLR4 was higher in Group (4) (INR + N) than in Group (2) (NIR + N); however, the difference was not statistically significant (p > 0.05). The expression level of TLR4 was lower in Group (3) (INR + M) than in Group (1) (NIR + M) (p > 0.05).

Table 4 shows significant differences in the levels of IL-2, IL-10, and TNF-α among the four groups. In particular, the expression level of IL-2 in Group (3) (INR + M) was lower than that in Group (4) (INR + N) (p = 0.012). Moreover, the expression level of IL-10 in Group (2) (NIR + N) was lower than that in Group (4) (INR + N) (p = 0.009). Additionally, the expression level of TNF-α in Group (1) (NIR + M) was lower than that in Group (3) (INR + M) (p = 0.029), and the expression level of TNF-α in Group (2) (NIR + N) was lower than that in Group (4) (INR + N) (p = 0.002).