Unveiling pharmacological targets of Rihimaside C for radiation-induced lung injury: An in silico and experimental integrated approach

Abstract

Background and aim

Radiation-induced lung injury (RILI) is a common complication during caner radiotherapy, mainly characterized by oxidative stress and inflammation. Rihimaside C, a novel dihydroflavonol compound isolated from Ribes himalense, exhibits significant anti-inflammatory and antioxidant properties. The study aims to investigate the therapeutic efficacy of Rihimaside C in treating RILI, as well as to uncover the potential targets and mechanisms involved.

Experimental procedure

Animal experiments were performed to evaluate the pharmacological efficacy of Rihimaside C for RILI. A computer-based strategy was employed to retrieve and screen potential targets for the therapy of Rihimaside C against RILI. STRING, DAVID databases, and Cytoscape software were utilized to construct a protein-protein interaction network and identify hub targets. Gene Ontology and Kyoto Encyclopedia of Genes and Genomes enrichment analyses were conducted to illuminate the underlying mechanisms. Molecular docking and Cellular Thermal Shift Assay (CETSA) were performed to further validate the hub targets.

Results and conclusion

The results of animal experiments showed that Rihimaside C effectively alleviated RILI. Four hub targets (TNF, HSP90AA1, ESR1 and HIF1A) among the 72 possible targets of Rihimaside C involved in the treatment of RILI were finally identified through network pharmacology, which were enriched in MAPK, IL-17, and PI3K/Akt signaling pathways. Molecular docking and CETSA analyses indicated that HSP90AA1 displayed highest binding affinity with Rihimaside C. This study investigated the therapeutic effects of Rihimaside C on RILI and identified potential targets, providing a novel strategy in treating RILI.

Graphical abstract

1. Introduction

Radiation-induced lung injury (RILI) is a notable adverse event resulting from ionizing radiation exposure, especially in thoracic radiotherapy.^1,2 Its incidence has been reported to reach up to 30 % in specific cases, making it a critical dose-limiting factor during thoracic radiation treatment.³ RILI comprises early radiation pneumonia and late radiation-induced pulmonary fibrosis, with oxidative stress and inflammatory response playing pivotal roles in its development.⁴ Consequently, glucocorticoids, as potent anti-inflammatory agents, have become the most widely used drugs for clinical RILI treatment. However, long-term administration of glucocorticoids is known to be associated with various side effects, such as hypertension, diabetes, osteoporosis, among others.⁵ The search for safe and effective alternative medications with fewer side effects is of paramount importance.

Traditional Chinese medicine (TCM) is a valuable asset known for its diverse components, broad pharmacological effects, and minimal side effects, making it increasingly utilized in clinical practice for preventing and treating inflammatory diseases.^6,7 Ribes himalense (糖茶藨子, táng chá biāo zǐ), a deciduous shrub belonging to the Saxifragaceae family in the Ribes genus, possesses a long history of medicinal use and abundant genetic resources. In TCM, it is widely employed for its antipyretic and detoxifying properties, as well as for managing various inflammatory conditions and autoimmune diseases.⁸ Ribes himalense contains a variety of active components, such as flavonoids, phenolic acids, and proanthocyanidins, which have been extensively researched and proven to possess health-promoting qualities, including anti-inflammatory and antioxidant effects.⁹ In our preliminary research, using targeted isolation techniques for antioxidant components from natural products, we have isolated a novel dihydroflavonoid from Ribes himalense, Rihimaside C, which has demonstrated significant in vitro antioxidant activity.^10,11 Further investigations are warranted to explore the effects and mechanisms of Rihimaside C in treating RILI.

Drug discovery is a pivotal process in diseases treatment, and accurately identifying drug targets remains a primary challenge for scientists.¹² Traditional drug development trials rely on time-consuming and costly biochemical experiments.¹³ In recent decades, the combination of theoretical research with wet experiments has gained popularity among researchers, facilitated by advancements in scientific thinking and computer technology.¹⁴ Network pharmacology, supplemented by wet experiments, has emerged as a promising strategy to predict and verify targets of active ingredients, including novel compounds in pharmaceuticals. It enables the exploration of potential disease treatments from both theoretical and practical perspectives. This approach has not only achieved significant progress in the screening of effective components in TCM but also has gained momentum in new drug discovery. Moreover, it offers advantages in terms of time, cost, and labor saving, allowing for better elucidation of the pharmacological mechanisms underlying drug treatments for various diseases.¹⁵

In this study, animal experiments were performed firstly to evaluate the in vivo pharmacological efficacy of Rihimaside C during the development of RILI. Subsequently, network pharmacology was employed to retrieve and screen potential targets for the therapy of Rihimaside C against RILI. At last, molecular docking along with Cellular Thermal Shift Assay (CETSA) were utilized to validate the targets. Our study integrated computational technologies with in vivo and in vitro experiments to comprehensively explore the therapeutic potential of Rihimaside C against RILI. This approach will enable a deeper understanding of the molecular mechanism underlying the beneficial effects of Rihimaside C in RILI treatment, and offer substantial support for the further development of Rihimaside C as a potential therapeutic agent.

2. Materials and methods

2.1. Animal experiment

Male C57BL/6 mice (6–8 weeks old) were purchased from GemPharmatech Co., Ltd., Changzhou China. The experimental animal management and animal welfare ethics committee of Jiangnan University (ethic review number: JN. No 20221130c0800615488) reviewed and approved the animal research program. All experimental procedures were conducted with the ethical guidelines and regulations set forth in the National Institutes of Health Guide for the Care and Use of Laboratory Animals. The mice were housed in a specific pathogen-free (SPF) environment with ad libitum access to standard diet and water, and maintained under standard conditions with alternating 12 h light and dark cycles, at a constant temperature of 22 ± 2 °C and a relative humidity of 50 ± 5 %. The mice were randomly divided into four groups, each consisting of six mice: the normal control group (NC), the irradiation (IR) group, the IR group treated with Rihimaside C (IR + Rihimaside C), and the IR group treated with dexamethasone (IR + DXM). After anesthesia, the mice were positioned supine on a specialized wooden board and exposed to X-rays from a 6 MV linear accelerator to achieve a single 20 Gy bilateral lung field irradiation, successfully establishing the model of RILI in C57BL/6 mice. Except for the NC group, all mice in the other three groups received irradiation. Following irradiation, the mice in the IR + Rihimaside C and IR + DXM groups were injected with Rihimaside C (0.4 mg/kg) and DXM (0.4 mg/kg)¹⁶ by tail vein injection respectively, once daily for 2 consecutive weeks (the dose of Rihimaside C was determined in the pre-experiment shown in Supplementary Material). Meanwhile, the mice in NC and IR groups were injected with equivalent normal saline. In the fourth week after the RILI model was established, all mice were sacrificed, the plasma, lung tissue, and bronchoalveolar lavage fluid (BALF) were collected for the subsequent experiments.

2.2. Histological examination of the lung tissues

The left lung tissue specimens were extracted and rinsed three times with cold phosphate-buffered saline (PBS, Biosharp, Cat No.: BL302A, Beijing, China), and fixed in 4 % paraformaldehyde for 24–48 h, followed by a process of dehydration and paraffin embedding. The embedded tissues were then sectioned into 4 μm thick slices and stained with hematoxylin and eosin (H&E, Solarbio, Cat No.: G1120, Beijing, China). The lung histopathological changes were assessed in a blinded manner, and the severity of injury was classified on a scale of 0 (no injury), 1 (mild injury), 2 (moderate injury), or 3 (severe injury).¹⁷ The assessment was based on the presence of exudates, congestion, neutrophilic infiltration, alveolar hemorrhage, debris and cellular proliferation observed in the H&E-stained lung sections.

For immunofluorescence analysis, tissue paraffin sections were placed in a 55 °C oven for 30 min to aid in in their adhesion to the slides. After roasting, dewaxing and antigen retrieval, endogenous peroxidase was blocked by incubating the sections with 10 % FBS (Yeasen, Cat No.: 40130ES76, Shanghai, China) and 0.3 % Triton X-100 (Beyotime, Cat No.: P0096, Shanghai, China). Subsequently, the sections were incubated with Ly6g primary antibody (Proteintech, Cat No.: FITC-65078, Wuhan, China) at 4 °C overnight and then treated with a secondary antibody (FITC, Proteintech, Cat No.: SA00003-2, Shanghai, China) for 1 h at room temperature, followed by DAPI (Beyotime, Cat No.: P0126, Proteintech, Wuhan, China) staining for 5 min. Finally, the sections were sealed with anti-fluorescence attenuation sealing agent containing DAPI. The images were captured using a fluorescence microscope (Zeiss, Axio Imager Z2, Germany), and neutrophil infiltrations were quantified and analyzed using Image J software.

2.3. Bronchoalveolar lavage fluid (BALF) analysis

To collect BALF, 1.5 mL of pre-chilled PBS was administered separately into the left lung through intratracheal administration. The fluid was then gently pumped back and forth three times. The BALF was centrifuged at 1000 × rpm for 5 min at 4 °C, leading to the separation of the supernatant and cell pellets. The protein concentration of supernatant was measured using a BCA protein assay kit (Vazyme, Cat. No.: E112-01, Nanjing, China) as per the manufacturer's instruction. For the cell pellets, 0.3 mL of red blood cell lysis buffer (Biosharp, Cat. No.: BL503A, Beijing, China) was added, followed by centrifugation at 12000×g for 5 min at 4 °C after standing at room temperature. The resultant precipitate was then suspended with 0.6 mL PBS. The total number of cells in BALF were counted using a hemacytometer.

2.4. Measurement of malondialdehyde (MDA) level

MDA is a common biomarker used to assess oxidative stress and lipid peroxidation. MDA concentration was detected using Lipid Peroxidation (MDA) Assay kit (Abbkine, Cat. No.: KTB1050, Wuhan, China). Lung tissue collected from the mice was processed following the manufacturer's instruction after obtaining the fresh materials.

2.5. RILI and Rihimaside C related targets identification

Potential targets of RILI were identified through a systematic search using various keywords, including “radiation-induced lung injury”, “radiation pneumonitis”, “radiation-induced pulmonary fibrosis”, “pneumonia”, “pulmonary fibrosis”, and “radiation injury”. Specialized online databases, including GeneCards (https://www.genecards.org/), Online Mendelian Inheritance in Man (OMIM, https://www.omim.org/statistics/geneMap), Therapeutic Target Database (TTD, https://db.idrblab.net/ttd/), Comparative Toxicogenomics Database (CTD, https://ctdbase.org/), DrugBank Online (https://go.drugbank.com/), and DisGeNET database (https://www.disgenet.org/), were utilized. In the meanwhile, targets associated with Rihimaside C were obtained from SwissTargetPrediction (http://www.swisstargetprediction.ch/), PharmMapper Server (http://lilab-ecust.cn/pharmmapper/index.html), TargetNet (http://targetnet.scbdd.com/home/index/), and SuperPred (https://prediction.charite.de/index.php). The molecular formula and canonical SMILES number of Rihimaside C were submitted to websites, and species selection was limited to Homo sapiens. Subsequently, all of the collected targets from above resources were integrated together after eliminating duplications. For further analysis, these targets were converted to standard gene names using the UniProt database (https://www.uniprot.org/).

2.6. Determination of hub targets using protein-protein interaction (PPI) network

The common targets shared between RILI and Rihimaside C were acquired via Venny 2.1.0 (https://bioinfogp.cnb.csic.es/tools/venny/). Then, these common targets were input to STRING database (https://cn.string-db.org/), with the organism screening limited to “Homo sapiens” in the operation interface, to construct a PPI network. Meanwhile, a confidence score of ≥0.400 was set, and disconnected nodes in the network were hidden. After that, the interaction information obtained from STRING database for the common targets was imported into Cytoscape 3.9.1 software. The CytoNCA and CytoHubba plugins within Cytoscape were employed to identify hub targets for Rihimaside C in the treatment of RILI. These plugins offered 11 topological analysis methods, containing Degree, Edge Percolated Component, Maximum Neighborhood Component, Density of Maximum Neighborhood Component, Maximum Cluster Centrality (MCC) and six centralities (Botteleneck, EcCentricity, Closeness, Radiality, Betweenness, Stress). Among these 11 methods, MCC showed superior performance in predicting essential proteins from PPI networks. The top 4 targets comprehensively evaluated based on a comprehensive consideration of MCC, degree (DC), betweenness centrality (BC), and closeness centrality (CC) were regarded as hub targets of Rihimaside C for treatment RILI.

2.7. GO and KEGG enrichment analysis

The Database for Annotation, Visualization and Integrated Discovery (DAVID, https://david.ncifcrf.gov/) was utilized to perform comprehensive functional annotation tools for GO and KEGG enrichment analysis. GO includes biological process (BP), cellular component (CC), and molecular function (MF). Enrichment analysis of metabolic pathways was performed by KEGG. The results were visualized using histograms and bubble plots through bioinformatics tools available on an online website, which allowed for personalized data analysis and scientific research mapping. P-values were obtained from the DAVID database and set at a significance level below 0.05.

To visualize the interactions among potential targets and the pathways associated with RILI and Rihimaside C, the data on correlated targets and pathways obtained from KEGG enrichment analysis were integrated and uploaded into Cytoscape 3.9.1 software, constructing a “disease-drug-target-pathway” network.

2.8. Molecular docking

Molecular docking is a typical approach to predict the binding mode and affinity between biological macromolecules and small molecule drugs, offering a means to validate network pharmacology prediction. In this study, crystal structures of the four hub target proteins with the highest degree of the network were obtained from the RCSB PDB database (https://www.rcsb.org/) and the PDB format files were downloaded accordingly. PyMol software was used to eliminate water molecules and separate the original ligands from the hub target proteins. The processed proteins were then imported into AutoDock Tools software, where hydrogenation, charge calculation, and atomic type assignment were performed, saving the files in PDBQT format. Subsequently, AutoDock Tools software was used to conduct molecular docking, encompassing protein macromolecule detection, small drug molecule insertion, and configuration of operating methods and docking parameters. The binding energy less than −5.0 kcal/mol was considered as a basis to evaluate prediction reliability, favoring ligand-macromolecule complexes with lower binding energy as more favorable structures. In the end, PyMol software were used to visualize the docking results and establish the docking interaction pattern.

2.9. Cell culture and CETSA

Numerous cells and cytokines mediate the pathogenic alteration of RILI. An increasing body of research has shown that epithelial cells are involved in radiation-induced early lung inflammation. MLE-12 cells are derived from murine lung epithelial cells and have been widely used as an in vitro model to study lung-related diseases and injuries, including RILI. The target proteins are stably expressed in MLE-12 cells, so this study selected the cells for CETSA experiment. Mouse lung epithelial-12 (MLE-12) cells purchased from Cell Bank of the Chinese Academy of Sciences (Shanghai, China) were maintained in Dulbecco's Modified Eagle Medium (DMEM, Biosharp, Cat. No.: BL304A, Beijing, China) containing 10 % fetal bovine serum (FBS, Yeasen, Cat. No.: 40131ES76, Shanghai, China) and 1 % penicillin-streptomycin (Biosharp, BL505A, Beijing, China). The cells were cultured in a humidified atmosphere with 5 % CO₂ at 37 °C.

To evaluate the binding affinities between Rihimaside C and target proteins TNF, HSP90AA1, HIF1A and ESR1, we carried out CETSA assay. Firstly, MLE-12 cells were collected after several serial passages and lysed using M-per buffer (Thermo Fisher, Cat. No.: 78501, USA) with freshly added 1 % protease inhibitor cocktail (MCE, Cat. No.: HY-K0010, USA) for 15–20 min. The total protein was extracted and quantified using a BCA protein assay kit after centrifugation at 13,000×g at 4 °C for 15 min. Next, the protein was equally distributed into two tubes and subsequently incubated with Rihimaside C (100 μmol/L) in one tube and an equal volume of DMSO as control in the other tube for 1 h at room temperature. After that, each tube was divided into eight 0.2 mL PCR tubes, which were then heated for 4 min at different designated temperatures with a defined temperature gradient. After balancing for 3 min at room temperature, the tubes were centrifuged at 20,000×g at 4 °C for 10 min. Eventually, the soluble supernatant was mixed with SDS-PAGE loading buffer, boiled immediately at 95 °C for 8 min using a PCR instrument, and analyzed by Western blot.

2.10. Statistical analysis

Statistical analysis and graphical representation of the data were performed using GraphPad Prism 8.0 software. For the comparison between two groups, Student's two-tailed t-test was utilized, while one-way analysis of variance (ANOVA) was employed for comparisons involving more than two groups. The data were presented as the mean ± standard deviation (SD) of three independent experiments, unless stated otherwise.

3. Results

3.1. Rihimaside C exerts a protective effect against RILI in vivo

To investigate the therapeutic effect of Rihimaside C on RILI, histological examinations were employed to assess the pathological changes in lung tissues of the RILI mouse model after intravenous injection of Rihimaside C. The results of H&E staining and immunofluorescence assay demonstrated that radiation exposure induced significant injuries in the lung tissues, characterized by alveolar and interstitial exudation, congestion, neutrophil infiltration, and interalveolar hemorrhage, when compared to the control group. However, these alterations were alleviated by Rihimaside C and DXM (Fig. 1A and B). As shown in Fig. 1C and D, the total cell counts and protein levels in BALF were significantly elevated in IR group, indicating increased permeability of the alveolar vascular barrier and the presence of pulmonary edema. While, treatment with Rihimaside C and DXM partially ameliorated these effects, suggesting that, similar to DXM, Rihimaside C has a beneficial impact on alveolar vascular barrier integrity and reduction in pulmonary edema. Additionally, the levels of MDA in the Rihimaside C and DXM treatment groups were significantly lower than those in the IR control group, indicating that both Rihimaside C and DXM can reduce the degree of membrane lipid peroxidation (Fig. 1E). These findings suggested that Rihimaside C possessed a protective effect against RILI.

Fig. 1.

Rihimaside C exerts protective effect against RILI in C57BL/6 mice.

3.2. Identification of hub targets for Rihimaside C in the treatment of RILI

Rihimaside C shares certain common features with dihydroflavonoid and has the potential to exert pharmacological effects through multiple pathways and targets. Fig. 2A presented the chemical structural formula of Rihimaside C. We employed network pharmacology to predict the targets associated with Rihimaside C and RILI. The results from GeneCards were screened using a relevance score of ≥10, and targets with an inference score of ≥50 were selected from CTD. To be included in the study, the DisGeNET Score_gda had to be > 0.01. Ultimately, we obtained 814, 862, 33, 3550, 250, and 748 targets from GeneCards, OMIM, TTD, CTD, DrugBank, and DisGeNET databases, respectively. A total of 4880 disease targets corresponding to RILI were screened after removing duplicate targets (Table S1). Concurrently, we integrated 201 potential targets of Rihimaside C based on its molecular formula, SMILES number and mol2 format from SwissTargetPrediction, PharmMapper Server, TargetNet and SuperPred databases (Table S2). As indicated in Venn diagram, 72 drug-disease common targets were captured, representing the possible targets of Rihimaside C involved in the treatment of RILI (Fig. 2B).

Fig. 2.

PPI network and identification of hub targets for Rihimaside C reducing RILI.

Afterwards, the 72 common targets were imported into STRING database to construct a PPI network, with a medium confidence level set to 0.4 (Fig. 2C). The PPI network comprised 72 nodes and 236 edges, which were input into the Cytoscape 3.9.1 software for subsequent analysis and identification of hub targets. In Fig. 2D, the nodes with a redder color and larger size represent the targets ordered from small to large based on their degree values. As shown in Table 1, several parameters, including MCC, DC, BC, and CC, were calculated using CytoHubba and CytoNCA plugin within Cytoscape 3.9.1 software. Based on these computed parameters, TNF, HSP90AA1, HIF1A and ESR1 were identified as the hub targets associated with the benefical effects of Rihimaside C in mitigating RILI (Fig. 2E–H).

Table 1.

Top ten target proteins ranked by degree.

NUM	Protein	Degree	Betweenness	Closeness	MCC
1	TNF	34	1168.7207	0.63809526	2670
2	HSP90AA1	31	994.0768	0.62616825	2581
3	ESR1	30	782.043	0.6203704	2600
4	HIF1A	28	556.75183	0.5826087	2666
5	RELA	15	112.95344	0.51937985	1895
6	CYP3A4	14	290.24585	0.5153846	566
7	ABCB1	13	43.043503	0.5153846	1452
8	CYP1A2	12	99.047676	0.45578232	510
9	HNF4A	12	80.92369	0.4962963	332
10	ESR2	10	45.766033	0.4890511	152

3.3. Enrichment analysis and construction of disease-drug-target-pathway network

To extract functional information from the 72 common targets, a GO enrichment analysis was conducted, revealing a total of 293 enriched items across three categories: 195 BP terms, 73 MF terms, and 25 CC terms. The top 10 enriched BP terms, MF terms, and CC terms were presented using bubble charts (Fig. 3A). The results indicated that the anti-RILI mechanism of Rihimaside C predominantly involves the regulation of gene transcription and expression, inflammatory response, and response to hypoxia. Among the enriched cellular components, significant associations were observed with cytosol, nucleus, nucleoplasm and extracellular region. In addition, the enriched molecular functions primarily included protein binding, transcription factor activity, and enzyme binding.

Fig. 3.

Enrichment analysis and construction of disease-drug-target-pathway network.

To further elucidate the molecular mechanism underlying the anti-RILI effects of Rihimaside C, a KEGG pathway analysis was performed on the relevant targets. The result revealed high enrichment across 58 pathways, and subsequently, the top 30 highly enriched pathways were filtered according to gene counts (Fig. 3B). Taking these results into account, it became evident that the potential targets were principally associated with “Apoptosis”, “MAPK signaling pathway”, “IL-17 signaling pathway” and “PI3K/Akt signaling pathway”. Disease-drug-target-pathway network, constructed using Cytoscape 3.9.1 software, was depicted in Fig. 3C.

3.4. High binding activity of Rihimaside C with HSP90AA1 and TNF

The binding ability of four potential targets (TNF, HSP90AA1, ESR1 and HIF1A) with Rihimaside C were calculated by molecular docking. The PDB structures of four key target proteins were downloaded from PDB database. Both the PDB protein structures and Rihimaside C were preprocessed. The gridbox settings, essential for molecular docking, along with the spatial coordinates, targets and compound details, were provided in Table 2. Additionally, Table 2 displayed the binding energies (kcal/mol) of Rihimaside C with target proteins. The bind energy between Rihimaside C and TNF was calculated to be −6.43 kcal/mol, and this binding affinity resulted from the formation of hydrogen bonds with Gly121 and Tyr151 residues (Fig. 4A). For the interaction between Rihimaside C and HSP90AA1, the binding energy was −9.39 kcal/mol, facilitated by 5 hydrogen bonds involving Leu103, Asp102, Gly135, Phe138, and Gly108 residues (Fig. 4B). Rihimaside C showed binding energies of −5.48 kcal/mol and −4.78 kcal/mol with ESR1 and HIF1A, respectively (Fig. 4C and D). The results suggested that Rihimaside C presented high binding affinity with HSP90AA1 and TNF. Furthermore, the interactions between Rihimaside C and the targets were visualized using PyMol software.

Table 2.

The binding energies and offset of center grid box.

Protein	PDB ID	Binding energy (kcal/mol)	X center	Y center	Z center
TNF	2az5	−6.43	−0.389	−0.889	8.194
HSP90AA1	3o0i	−9.39	−4.417	1.944	2.306
ESR1	2r6w	−5.48	12.000	8.889	0.972
HIF1A	4h6j	−4.78	5.333	1.778	−0.556

Fig. 4.

Verification of the binding affinity between Rihimaside C and target proteins by molecular docking.

3.5. HSP90AA1 as a direct target of Rihimaside C

The binding affinity of a drug to hub targets is a crucial factor in assessing the reaction intensity. The determination of the binding affinity between drugs and targets is the basis for the development of small molecule drugs.¹⁸ Given the therapeutic effects of Rihimaside C on inflammation and oxidative stress in RILI observed in animal experiments, we employed CETSA to verify and quantify the extent of the interaction between Rihimaside C and TNF, HSP90AA1, ESR1, and HIF1A in MLE-12 cells. At least three experiments were performed after exploring an appropriate temperature range in the early stage. Compared to DMSO, incubation with Rihimaside C resulted in improved thermal stability of TNF (Fig. 5A) and HSP90AA1 (Fig. 5B) in cells lysates, whereas no significant differences were observed in the thermal stability of ESR1 (Fig. 5C) and HIF1A (Fig. 5D). These findings provided evidence that Rihimaside C enhanced the thermal stability of TNF and HSP90AA1 within a certain temperature range, consistent with the results obtained from molecular docking. Moreover, HSP90AA1 was regarded as the core target, showing direct and intensive binding with Rihimaside C.

Fig. 5.

Identification of the direct binding targets for Rihimaside C using CETSA.

4. Discussion

Radiation therapy in patients with thoracic tumors and exposure to a certain amount of ionizing radiation are prominent risk factors contributing to the development of RILI. The pathological manifestations of RILI encompass alveolar and interstitial exudative changes, edema, inflammatory infiltration, and pulmonary interstitial fibrosis.¹⁹ Available researches have illuminated the multifaceted and dynamic nature of RILI, influenced by a myriad of cytokines, including TNF-α, IL-1β, IL-6, IL-17, TGF-β.^20,21 The incipient mechanisms instigating RILI entail direct DNA damage and the generation of ROS, initiating intricate signaling pathways like TGF-β/Smad, HMGB1/TLR4 and Nrf2/ARE, culminating in the release of cytokines and molecules that provoke inflammatory and immune responses.^22,23 Therefore, oxidative stress-induced injury, coupled with subsequent inflammatory cascades, assumes a cardinal role in the pathogenesis of RILI.^24,25 Despite the swift and unceasing advancements in radiotherapy techniques, such as stereotactic body radiotherapy and flash radiotherapy, dose-limiting toxicity perpetually impedes the realization of optimal local tumor control through radiotherapeutic interventions.^26,27 Presently, therapeutic options for RILI remain relatively circumscribed.

TCM and natural products have emerged as pivotal sources in the exploration of novel pharmaceuticals. A large number of TCM and natural products contain antioxidant and anti-inflammatory components, which have been garnering increasing attention in the field of radiation protection due to their wide availability and low toxicity profile in patients.^28,29 Among these, flavonoids, extensively found in various natural plants, are renowned for their beneficial health effects.²⁹ Several studies have provided evidence supporting the ability of flavonoids to enhance the antioxidant and free radical scavenging capacity in animals, along with their potential anti-inflammatory effects.^29,30 Notably, many flavonoids have demonstrated potential in mitigating radio-induced lung damage. For instance, baicalin, an active compound extracted from the dry roots of the Chinese herb Scutellaria baicalensis Georgi, exhibits the ability to alleviate radiation-induced cell damage in primary AEC II.³¹ Hesperidin, a member of the dihydroflavonoids and a natural polyphenolic structure presenting in fruits and vegetables, has been proven to significantly diminish γ-radiation-induced damage in rat lung tissues.³² Interestingly, the potential of flavonoids as effective COX inhibitors in treating inflammation has also been reported.²⁹

Ribes himalense, a plant used in traditional Tibetan medicine, has been demonstrated to have significant antioxidant and anti-inflammatory properties.⁸ In our previous experimental study, we established a targeted isolation techniques for antioxidant components in natural products. In this process, an online HPLC-DPPH activity screening system is utilized in conjunction with medium-pressure/high-pressure liquid chromatography for the qualitative identification, targeted enrichment, and directed separation of free radical scavengers from stems and leaves of Ribes himalense. Rihimaside C was obtained and in vitro antioxidant activity was validated.¹¹ Considering its antioxidative properties and the primary role of oxidative damage in RILI, we speculate that Rihimaside C may have a mitigating effect on RILI.

In the current investigation, the therapeutic action of Rihimaside C against RILI was meticulously explored using a combination of in silico methods and wet experiments. Animal experiments corroborated the protective effects of Rihimaside C in RILI, evidenced by the restraint of inflammatory cell infiltration, improvement in alveolar capillary barrier permeability, reduction of pulmonary edema, and attenuation of oxidative stress levels. The result of network pharmacology suggested that TNF, HSP90AA1, ESR1, and HIF1A were pivotal potential targets of Rihimaside C in RILI treatment. GO enrichment analysis further revealed their crucial involvement in hypoxia and inflammation responses, which bear direct relevance to RILI initiation and progression.³³ Subsequent KEGG enrichment analysis substantiated that Rihimaside C might exert a defensive effect in RILI by targeting hub targets and regulating inflammation-related signaling pathways such as MAPK, IL-17, and PI3K/Akt signaling pathways. Interestingly, HSP90AA1 was found to be involved in all above pathways. These findings indicated that HSP90AA1 might be a more important target in the process of treating RILI with Rihimaside C. Subsequently, molecular docking results exhibited favorable docking affinity between Rihimaside C and TNF, HSP90AA1, ESR1, and HIF1A, with valueed less than −4.0 kJ/mol, indicative of stable docking interactions and the potential for binding to these hub targets against RILI. Numerous cells and cytokines mediate the pathogenic alteration of RILI. An increasing body of research has shown that epithelial cells are involved in radiation-induced early lung inflammation. In the pursuit of further elucidating the mechanism underlying the therapeutic potential of Rihimaside C in RILI, the CETSA assay ultimately identified HSP90AA1 as the most pivotal target, showcasing direct and tight binding interactions with Rihimaside C.

The HSP90 family constitutes a cluster of chaperone proteins that play a crucial role in the folding and maturation of various proteins, thereby precisely regulating gene expression, cell cycle, proliferation, signaling, and other biological processes.³⁴ HSP90AA1, belonging to the HSP90 family, displays significantly elevated in tumor cells and under conditions of heat stress. While research on HSP90 has predominantly focused on cancer and diseases linked to protein misfolding, its involvement in RILI has been relatively limited. Nevertheless, its role in apoptosis and inflammation-related signaling pathways has been extensively studied. It has been implicated in controlling various forms of programmed cell death (PCD), including apoptosis, necroptosis, autophagy, ferroptosis, and others.³⁵ Inhibition of HSP90 activity has been demonstrated to restrict the PI3K/Akt pathway, leading to downregulation of survival genes and promoting apoptosis of target cells.³⁶ Li et al. revealed that radiation-induced oxidative stress triggered ferroptosis by activating the P62-Keap1-NRF2 pathway in RILI cells.³⁷ The PI3K/Akt signaling pathway has been identified as a co-upstream regulator of inflammation-related pathways, including NF-κB, TNF, and MAPK signaling pathways, all of which were enriched in KEGG analysis of this study.

Furthermore, HSP90AA1 was also found to be enriched in the IL-17 signal pathway. IL-17R-Act1 complex actives TRAF6, thereby triggering NF-κB, C/EBPβ, C/EBPδ, and MAPK pathways, leading to the secretion of a series of cytokines and chemokines involved in the regulation of autoimmune pathology, neutrophil recruitment, and immunity against extracellular pathogens.³⁸ Notably, our laboratory has uncovered evidence suggesting that IL-17A may play a role in the occurrence and development of RILI by mediating the secretion of S100A9, with subsequent experiments exploring the potential underlying mechanism.³⁹ Based on these discoveries, we posit that Rihimaside C may exhibit its therapeutic effects by targeting HSP90 and inhibiting the activation of the IL-17 signaling cascade, ultimately mitigating the advancement of RILI.

In contemporary drug development, the amalgamation of computational strategies with experimental validation has emerged as a preferred avenue for researchers. Network pharmacology involves the systematic analysis of complex biological networks, including protein-protein interaction networks and signaling pathways, to understand the interactions between drugs and their targets within biological systems. In the study of novel compounds, such as Rihimaside C in our research, network pharmacology serves as a valuable tool for elucidating potential therapeutic mechanisms and identifying molecular targets. However, it is essential to recognize the limitations of network pharmacology when applied to novel compounds. Firstly, the predictive accuracy of network pharmacology relies heavily on the availability and quality of data used to construct biological networks and predict drug-target interactions. For novel compounds, limited experimental data may hinder the accuracy of predictions and necessitate cautious interpretation of results. Additionally, network pharmacology alone may not provide definitive insights into the pharmacological properties of novel compounds. Experimental validation of predicted drug-target interactions and biological activities is crucial to confirm the relevance of network pharmacology findings and establish the therapeutic potential of the compound. In this study, we not only validated the therapeutic effect of Rihimaside C on RILI through animal experiments but also corroborated the predicted target proteins from network pharmacological analysis through molecular experiments. Regarding the limitations of our current research, a comprehensive understanding of the treatment mechanism of Rihimaside C in addressing RILI necessitates further investigation through meticulous in vivo and in vitro experiments.

5. Conclusion

The integration of in silico strategy with wet experiments has yielded novel insights into the pharmacological activity and mechanism of new compounds. In this study, we explored the pharmacodynamics of Rihimaside C in addressing RILI, employing a multi-faceted approach involving network pharmacology, animal experiments, molecular docking, and CETSA. We postulated that Rihimaside C could effectively treat RILI via binding to HSP90AA1and regulating IL-17, PI3K/Akt and MAPK signal pathways. This significant finding holds great potential for the rational development and application of Rihimaside C in clinical practice.

Funding

This work was supported by the Translational Medicine Project of Wuxi Commission of Health (grant number ZH202101) and the “Taihu lake” Science and Technology Project of Wuxi (grant number K20221020).

Declaration of competing interest

The authors declared that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Appendix A. Supplementary data

The following are the Supplementary data to this article: