Paeoniflorin alleviates high glucose-induced endothelial cell apoptosis in diabetes mellitus by inhibiting HRAS-activated RAS pathway

Abstract

Paeoniflorin (Pae) can improve diabetes mellitus (DM), especially endothelial dysfunction induced by high glucose (HG). Molecularly, the mechanism pertinent to Pae and DM lacks further in-depth research. Hence, this study determined the molecular mechanism of Pae in treating DM through network pharmacology. The target of Pae was analyzed by TCMSP database, and DM-related genes were dissected by Genecards database and Omim database. PPI network was constructed for cross targets through Cytoscape 3.9.1 and STRING platform. GO and KEGG analyses were carried out on the cross targets. Protein molecular docking verification was completed by AutoDockTools and Pymol programs. Human umbilical vein endothelial cells (HUVECs) were separately treated with HG, Pae (5, 10, 20 μM) and/or HRAS overexpression plasmids (oe-HRAS). The cell viability, apoptosis and the protein expressions of HRAS and Ras-GTP were evaluated. There were 50 cross targets between Pae and DM, and VEGFA, EGFR, HRAS, SRC and HSP90AA1 were the key genes identified by PPI network analysis. GO and KEGG analyses revealed signal paths such as Rap1 and Ras. Molecular docking results confirmed that Pae had a good binding ability with key genes. In HG-treated HUVECs, Pae dose-dependently facilitated cell viability, attenuated cell apoptosis, and dwindled the expressions of HRAS and Ras-GTP, but these effects of Pae were reversed by oe-HRAS. In conclusion, Pae regulates the viability and apoptosis of HG-treated HUVECs by inhibiting the expression of HRAS.

 Introduction

Diabetes mellitus (DM) is a common chronic disease that seriously harms human health. China faces a great challenge from the high incidence of DM, which has become a public health problem that cannot be ignored [1]. Diabetic angiopathy is one of the common complications of DM [2, 3]. Vascular endothelial cells are the barrier between blood vessel wall and blood flow, which can be directly stimulated by various factors in blood, and also can synthesize and release assorted active substances that play an important role in regulating endothelial function and maintaining vascular homeostasis [4]. However, under the induction of high glucose (HG), the function of endothelial cells is affected, and problems such as angiogenesis disorder and vasodilation damage may occur [5]. Up to 75% of DM patients succumb to vascular diseases and endothelial dysfunction. Accordingly, endothelial dysfunction is a vital contributing factor of diabetic vascular disease, and protecting endothelial cells from damage provides a source of idea in developing novel treatment modality targeting DM.

Paeoniflorin (Pae) is a compound extracted from the root of Paeonia lactiflora, a perennial herb, which has the characteristics of easy absorption and high oral bioavailability [6]. Modern pharmacological research showed that Pae has antioxidant, anti-inflammatory, immunomodulatory, cardiovascular expansion and other pharmacological effects [6, 7], and functions in DM-related diseases, such as gestational diabetes [8], diabetic liver injury [9] and diabetic nephropathy [10]. Noteworthily, Pae exerts a protective effect on vascular endothelial cell injury induced by HG [11]. However, the mechanism of Pae improving HG-induced endothelial dysfunction has not been fully expounded, and needs further exploration.

Given the complex composition and mechanism of natural products, how to screen effective components and target genes related to diseases at low cost and high efficiency is a difficult point in drug development using natural products. Network pharmacology is a drug research method grounded in bioinformatics, systems biology and pharmacology. Studying the mechanism of drugs from a holistic and systematic perspective can better clarify the relationship among drugs, diseases and targets [12]. Therefore, this study explored the biological pathway of Pae in treating HG-induced endothelial dysfunction by network pharmacology, and performed verification at the cellular level according to the predicted results, so as to reveal the pharmacological effect and molecular mechanism of Pae in improving HG-induced endothelial dysfunction.

 Methods

 Screening of targets of Pae

The active components of Pae were selected from TCMSP database (https://tcmsp-e.com/tcmsp.php) with the settings of oral bioavailability (OB) ≥30% and drug-likeness (DL) ≥0.18. The obtained active components were searched in Pubchem database (https://pubchem.ncbi.nlm.nih.gov/), and the corresponding SMILE structure was retrieved. The SMILE structure was input into Swisstargetprediction Database (http://www.swisstargetprediction.ch/) for prediction, and the effective target of Pae was obtained by screening with probability >0.1.

 Screening of targets of DM

With “Diabetes” as the key word, the DM-related targets were searched in Genecards database (https://www.genecards.org/) and Omim database (https://omim.org/). According to the requirement of relevance score ≥ twice the median, the related genes were filtered out from Genecards database, and compared with the genes in Omim database after sorting and de-duplication, so as to remove overlapping items and finally determine the DM-related targets.

 Construction of protein–protein interaction (PPI) network

The targets of Pae and DM were input into the Draw Venn Diagram website (http://bioinformatics.psb.ugent.be/webtools/Venn/) to obtain the common target genes of drugs and diseases. Then these target genes were input into the STRING database (https://cn.string-db.org/), with the selection of “multiple proteins” and “homo sapiens,” and default confidence to get the protein interaction relationship. The obtained results were exported in tsv format and imported into Cytoscape 3.9.1 for data visualization. Based on the degree value, the obtained targets were arranged in descending order to obtain the relevant information of the core targets.

 Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses

The common target genes of Pae and DM were imported into DAVID database (https://david.ncifcrf.gov/), with the identifier “OFFICIAL_GENE_SYMBOL,” species “homo sapiens,” and the list “gene list.” By calculating the p value <0.05 and arranging it in descending order of count value, the top 10 results of biological processes, cell components and molecular function were screened out and subjected to GO analysis, and the five signal paths were obtained as the data of correlation analysis of KEGG pathway. The above data were imported into the website (http://www.bioinformatics.com.cn/) to realize the visualization of GO and KEGG analyses.

 Molecular docking

Mol2 file of Pae molecular structure was searched in TCMSP database. Uniport database (https://www.uniprot.org/) was used to input the top five targets of PPI network, with the selection of status “Reviewed,” species “Human,” and the protein structure with low resolution and relatively more protein sequences. The PDB format file of protein was downloaded from PDB database (https://www.rcsb.org/), and the protein structure was pretreated by Pymol 2.3.0 software to remove irrelevant proteins and crystalline water molecules. The protein structure was hydrotreated by AutoDockTools-1.5.6 software, and the binding energy of Pae and protein was compared. The docking complex and protein structure with the highest binding energy fraction were input into Pymol 2.3.0 software for 3D visualization of molecular docking. The correlation between small molecules and core proteins was analyzed by LigPlus software. The results were shown in Table 1.

Table 1 Molecular docking scoring results

Target gene	Hydrogen bond number	Hydrogen bond length	Number of hydrophobic small molecules	Binding energy
VEGFA	1	3.18	10	–5.82
EGFR	3	3.07, 2.81, 2.81	6	–4.66
HRAS	0	0	8	–6.3
SRC	0	0	7	–5.33
HSP90AA1	1	2.82	8	–6.68

 Cell culture

Human umbilical vein endothelial cells (HUVECs) purchased from ATCC (PCS-100-010, USA) were cultured in HUVECs specific medium (CM-0122, Procell, China) at 37°C under 5% CO₂.

 Cell treatment

HUVECs were treated with 25 mM glucose [13] (SNSP-001, Sunncell, China) for 24 hours to induce cell dysfunction. To study the therapeutic effect of Pae (HY-N0293, MedChemExpress, China), Pae with a final concentration of 5, 10 and 20 μM [14] was added to the culture medium of HUVECs.

 Transfection

HRAS cDNA sequence was inserted into pcDNA 3.1+ vector (TSPLA10008, Testobio, China) to overexpress HRAS (oe-HRAS), while the empty vector served as negative control (NC). Oe-HRAS and transfection reagent (40806ES01, YEASEN, China) were diluted in serum-free medium, followed by mixing and incubation with HUVECs. After 48 hours of transfection, HUVECs were subjected to quantitative real-time polymerase chain reaction (qRT-PCR).

 Gene expression

Total RNA was extracted from HUVECs using Trizol reagent (R21086, Shyuanye, China), and reversely transcribed into cDNA by RT kit (MF949, Mei5bio, China). QRT-PCR was performed using cDNA with qPCR Probe Master Mix (MF797, Mei5bio, China). The specific primer sequences were as follows: HRAS sense: 5'-CCAGCAAGCGGTGGGG-3', antisense 5'-TCCATGCGAAGGTCTTGGTC-3'; Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) sense: 5'-CCCTTAAGAGGGATGCTGCC-3', antisense 5'-ACTGTGCCGTTGAATTTGCC-3'. Gene expression results were normalized to GAPDH as per 2^–ΔΔCT method [15].

 Cell viability

The viability of HUVECs was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) kit (R20228, Shyuanye, China). Cells were cultured in 96-well plates (3,000 cells/well) for 24 hours, and successively reacted with 10 μL MTT solution (4 hours) and 110 μL formazan solvent. The OD value at 570 nm was read using a microplate reader (KD-810A, KDYXBIO, China).

 Flow cytometry

The apoptosis of HUVECs was assessed by ANNEXIN V-FITC/PI kit (G65873, JKBIO, China). HUVECs (1 × 10⁵) were re-suspended in ANNEXIN V-FITC binding buffer, and then stained with ANNEXIN V-FITC and PI, respectively. Following 20 minutes of incubation in the dark, the results were detected by a flow cytometer (Epics-XLⅱ Ⅱ, Beckman, USA).

 Western blot

HUVECs were lysed with cell lysis buffer (BL504A, Biosharp, China) to obtain the protein. After concentration measurement using BCA Protein Assay kit (BL521B, Biosharp, China), the proteins were loaded on SDS-PAGE (R21150, Shyuanye, China), and then transferred to nitrocellulose membrane. Post blockage in the tris-buffered saline tween-20 (TBST, JLC-SJ2538, JKBIO, China) containing BSA (S12012, Shyuanye, China) for 1 hour, the membrane was incubated with primary antibodies (Table 2) at 4°C and secondary horseradish peroxidase-conjugated antibodies (Table 2). Activation of Ras (Ras-GTP) was detected by Active Ras Detection Kit (#8821, cell signaling technology, USA). Thereafter, the membrane was photographed in an imaging system (SynGene, USA) using ECL substrate kit (JLC-SJ2635, JKBIO, China). GAPDH or total Ras served as the internal reference.

Table 2 Antibodies used in this study

Name	Catalog	Molecular weight	Dilution	Manufacturer
HRAS	ab191595	21 kDa	1/2,000	abcam, UK
GAPDH	ab8245	36 kDa	1/10,000	abcam, UK
goat anti rabbit	ab205718	—	1/2,000	abcam, UK
goat anti mouse	ab205719	—	1/2,000	abcam, UK

Abbreviation: GAPDH: Glyceraldehyde-3-phosphate dehydrogenase

 Statistical analysis

The experiments were repeated at least three times. One-way ANOVA was used for comparison among groups. All experimental results were analyzed by SPSS 21.0 system (SPSS Inc., USA) and expressed as the Mean ± Standard deviation (SD). P < 0.05 was considered statistically significant.

 Results

 Drug–disease intersection targets and PPI network analysis

63 Pae-related targets screened from Swisstargetprediction database were intersected with 4,758 DM-related targets, and 50 disease-drug common targets were obtained (Fig. 1A). PPI network was analyzed by STRING database, and 50 interacting nodes with 188 edges were formed (Fig. 1B). According to the analysis of Cytoscape 3.9.1, the top five genes were vascular endothelial growth factor A (VEGFA), epidermal growth factor receptor (EGFR), HRAS, SRC and heat shock protein 90 alpha family class A member 1 (HSP90AA1) (Fig. 1B), indicating that these five genes were more relevant to Pae in treating DM.

Fig. 1  Bioinformatics analysis of genes and pathways of paeoniflorin affecting diabetes

(A) Venn diagram of intersection genes of Paeoniflorin and diabetes. (B) Peoniflorin-diabetes related target visualization network. (C) GO function analysis of potential targets of Paeoniflorin in diabetes treatment. (D) KEGG pathway analysis of potential targets of Paeoniflorin in diabetes treatment.

 GO and KEGG analyses

The 50 genes listed in Fig. 1A were imported into DAVID database for GO enrichment analysis, and finally 243 GO enrichment entries were obtained, including 171 biological processes (BP), 27 cellular components (CC) and 45 molecular functions (MF). As delineated in Fig. 1C, Pae played an important role in biological processes through promoting protein phosphorylation, inhibiting cell proliferation, responding to xenobiotic stimulus, boosting proteolysis, and thus affecting biological growth and development. Cell components were enriched in plasma membrane, cytosol, cytoplasm, extracellular region and extracellular space. The molecular functions were expressed by protein binding, ATP binding, identical protein binding, zinc ion binding, metalloendopeptidase activity and so on. The enrichment of KEGG pathway showed that 50 intersecting genes were significantly enriched in 38 pathways (p < 0.05), and the top 20 related signal pathways were enriched (Fig. 1D). The mechanism of Pae in the treatment of DM was closely related to the signal pathways, such as Rap1, Ras, calcium and PI3K-Akt, and the interaction of neuroactive ligand receptors.

 Molecular docking study

We analyzed the binding energy (Table 1) of the first five genes with Pae in Fig. 1B, and the scores within 0 to 5 indicated average results and the scores within –5 to –10 implied good results. The docking results of Pae and VEGFA were good, and hydrogen bonds were generated at C1 and N sites. The docking results of Pae and EGFR were average, and hydrogen bonds were formed at O8 and NE, O9 and N, and O10 and NE2, respectively, with hydrogen bond lengths of 3.07, 2.81 and 2.81. Pae had a good docking result with HRAS and SRC, but without hydrogen bond. The docking result of Pae and HSP90AA1 was good, and hydrogen bonds were formed at O2 and N sites, with 2.82 in length, which were polar and hydrophobic.

 Effects of Pae on HG-treated HUVECs

Pae had no effect on the viability of normally cultured HUVECs (Fig. 2A), yet concentration-dependently promoted the viability of HG-treated HUVECs (Fig. 2B, p < 0.05). The increase of apoptosis of HG-induced HUVECs was also reversed by Pae in a concentration-dependent manner (Fig. 2C, D, p < 0.001). Moreover, HG elevated the expressions of HRAS and Ras-GTP, which was offset by Pae in a dose-dependent manner in HUVECs (Fig. 2E–H, p < 0.05). PAE at 20 μM was selected for subsequent experiments owing to its best effect.

Fig. 2  Effects of paeoniflorin on HG-treated endothelial cells

(A) Effect of paeoniflorin on cell viability was determined by MTT. For (B–H), endothelial cells were treated with 25  mM glucose for 24 hours in the presence or absence of paeoniflorin (5, 10, 20 μM). (B) Cells viability was detected by MTT. (C, D) Cell apoptosis was assessed by flow cytometry. (E, F) The protein level of HRAS was measured by Western blot. GAPDH was used as internal control. (G, H) The protein level of Ras-GTP was determined by Western blot. Total Ras was used as internal control.

**p < 0.01, ***p < 0.001 vs. control. ⁺p < 0.05, ⁺⁺p < 0.01, ⁺⁺⁺p < 0.001 vs. HG.

HG: high glucose. MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide. GAPDH: Glyceraldehyde-3-phosphate dehydrogenase.

 Oe-HRAS reversed the effects of Pae on HG-treated HUVECs

To explore whether the regulatory effect of Pae on HG-treated HUVECs was related to HRAS, we constructed an overexpression plasmid of HRAS, and confirmed its promoting effect on HRAS expression in HUVECs (Fig. 3A, p < 0.001). In HG-treated HUVECs, oe-HRAS reversed the promotive effect of Pae on the viability, and the inhibitory effect of Pae on the apoptosis and the expression of Ras-GTP (Fig. 3B–F, p < 0.001).

Fig. 3  Oe-HRAS reversed the effects of paeoniflorin on HG-treated endothelial cells

(A) The transfection efficiency of oe-HRAS was determined by qRT-PCR. GAPDH was used as internal control. For (B–F), endothelial cells transfected with NC or oe-HRAS were treated with 25  mM glucose for 24 hours in the presence or absence of Paeoniflorin (20 μM). (B) Cell viability was detected by MTT. (C, D) Cell apoptosis was tested by flow cytometry. (E, F) The protein level of Ras-GTP was examined by Western blot. Total Ras was used as internal control.

^{^^^}p < 0.001 vs. NC. ⁺⁺⁺p < 0.001 vs. HG. ^&&&p < 0.001 vs. HG + Pae + NC.

Oe-HRAS: HRAS overexpression plasmid. NC: negative control. qRT-PCR: quantitative real-time polymerase chain reaction.

 Discussion

In recent years, network pharmacology combined with molecular docking contributes to the study on the molecular mechanism of traditional Chinese medicine and natural compounds [16]. This is because network pharmacology emphasizes the multi-channel and multi-target relationship between drugs and diseases, and reveals the complex interaction between drugs and diseases, while molecular docking can predict the ligand-target interaction at the molecular level and thus help screen suitable active compounds. In this study, we employed network pharmacology and molecular docking technology to explore the relationship between Pae and DM.

Pae and DM had five important targets, namely VEGFA, EGFR, HRAS, SRC and HSP90AA1, which had good binding force with Pae. VEGFA is an active growth factor in angiogenesis and endothelial cell growth, and can promote the proliferation and migration of endothelial cells and induce vascular permeability [17]. It has been reported that Pae can improve endothelial dysfunction by up-regulating VEGFA [18]. EGFR, a member of epidermal growth factor receptor family, abnormal activation of EGFR contributes to the vascular dysfunction associated with type 2 diabetes [19, 20]. Wang et al. reported that Pae protects mouse brain microvascular endothelial cells by activating EGFR [21]. HRAS, a small GTPase, is a biological switch in cell process [22]. The activation of HRAS protein promotes the apoptosis of diabetic retinal capillary cells [23]. However, there is no report regarding the regulatory effect of Pae on HRAS. SRC is a member of SRC kinase family, which plays a key role in cell morphology, migration, proliferation and angiogenesis [24]. Down-regulation of VEGF/ERK/FAK/Src pathway signals can prevent angiogenesis in diabetic retinopathy [25]. Of note, Pae inactivates Src to improve the permeability of cardiac microvascular endothelial cells [26]. HSP90AA1 encodes Hsp90α protein to promote the proper folding of target protein during cell stress [27]. In addition, HSP90AA1 has been identified as a key node in DM and its complications by bioinformatics analysis [28]. Pae can target HSP90AA1 to prevent cisplatin-induced acute kidney injury [29], but whether Pae targets HSP90AA1 to participate in HG-induced endothelial dysfunction remains obscure.

Then GO and KEGG analyses showed that Pae regulated the pathogenesis of DM through multiple pathways (such as protein binding and plasma membrane) and multiple signal pathways (such as Rap1 signal pathway and RAS signal pathway). Rap1 is related to the increase of cell adhesion and the enhancement of barrier function of endothelial cells [30]. Rap1 can mitigate renal tubular injury in diabetic nephropathy [31]. RAS protein is a small GTPase with two forms: RAS-GDP and RAS-GTP, and the latter can interact with more than 20 effector molecules to initiate downstream signals [32]. There are three main RAS genes: HRAS, KRAS and NRAS, which have always been considered as the driving factors of cancer [22]. However, it was found that DM can increase the expression of HRAS [33], and the activation of HRAS can contribute to the microvascular pathological characteristics of diabetic retinopathy [34]. Considering that HRAS was also predicted as the target of Pae in treating DM in this study, and no report revealed the regulatory effect of Pae on HRAS, we chose HRAS for verification via cell experiments.

In the cell experiment, we first observed the effect of Pae on HUVECs, and found Pae did not affect the viability of normal HUVECs, manifesting that Pae was nontoxic to HUVECs. In addition, Pae increased the viability yet decreased the apoptosis of HG-treated HUVECs, coinciding with a previous study [11]. We also provided a new finding that Pae inhibited the expressions of HRAS and RAS-GTP. Moreover, oe-HRAS reversed the effect of Pae on HG-treated HUVECs, hinting that Pae may exert its therapeutic effect on HG-treated HUVECs by diminishing the expression of HRAS. However, there are limitations in this study. For instance, the results of this study need to be further validated in animal experiments; and an in-depth research on the downstream mechanism of HRAS implicated in Pae affecting DM is needed.

 Conclusion

In conclusion, this study suggested that Pae may regulate the viability and apoptosis of HG-treated HUVECs by inhibiting the expression of HRAS, which will provide a new direction for developing treatment methods towards DM-related endothelial dysfunction.

 Conflict of Interest

Not applicable.

 Acknowledgments

Not applicable.

 Funding

Not applicable.

References

• 1   Ma  RCW (2018) Epidemiology of diabetes and diabetic complications in China. Diabetologia 61: 1249–1260.
• 2   Viigimaa  M,  Sachinidis  A,  Toumpourleka  M,  Koutsampasopoulos  K,  Alliksoo  S, et al. (2020) Macrovascular complications of type 2 diabetes mellitus. Curr Vasc Pharmacol 18: 110–116.
• 3   Wei  J,  Tian  J,  Tang  C,  Fang  X,  Miao  R, et al. (2022) The influence of different types of diabetes on vascular complications. J Diabetes Res 2022: 3448618.
• 4   Clyne  AM (2021) Endothelial response to glucose: dysfunction, metabolism, and transport. Biochem Soc Trans 49: 313–325.
• 5   Maruhashi  T,  Higashi  Y (2021) Pathophysiological Association between Diabetes Mellitus and Endothelial Dysfunction. Antioxidants (Basel) 10: 1306.
• 6   Zhang  L,  Wei  W (2020) Anti-inflammatory and immunoregulatory effects of paeoniflorin and total glucosides of paeony. Pharmacol Ther 207: 107452.
• 7   Li  J,  Zhang  SH,  He  D,  Wang  JF,  Li  JQ (2019) Paeoniflorin reduced the cardiotoxicity of aconitine in h9c2 cells. J Biol Regul Homeost Agents 33: 1425–1436.
• 8   Zhang  Y,  Liang  Y,  Liu  H,  Huang  Y,  Li  H, et al. (2020) Paeoniflorin attenuates gestational diabetes via Akt/mTOR pathway in a rat model. Food Nutr Res 64: 4362.
• 9   Wang  A,  Gong  Y,  Pei  Z,  Jiang  L,  Xia  L, et al. (2022) Paeoniflorin ameliorates diabetic liver injury by targeting the TXNIP-mediated NLRP3 inflammasome in db/db mice. Int Immunopharmacol 109: 108792.
• 10   Wang  X,  Liu  XQ,  Jiang  L,  Huang  YB,  Zeng  HX, et al. (2022) Paeoniflorin directly binds to TNFR1 to regulate podocyte necroptosis in diabetic kidney disease. Front Pharmacol 13: 966645.
• 11   Wang  JS,  Huang  Y,  Zhang  S,  Yin  HJ,  Zhang  L, et al. (2019) A protective role of paeoniflorin in fluctuant hyperglycemia-induced vascular endothelial injuries through antioxidative and anti-inflammatory effects and reduction of PKCβ1. Oxid Med Cell Longev 2019: 5647219.
• 12   Nogales  C,  Mamdouh  ZM,  List  M,  Kiel  C,  Casas  AI, et al. (2022) Network pharmacology: curing causal mechanisms instead of treating symptoms. Trends Pharmacol Sci 43: 136–150.
• 13   Sun  F,  Liu  J,  Wang  Y,  Yang  H,  Song  D, et al. (2023) BASP1 promotes high glucose-induced endothelial apoptosis in diabetes via activation of EGFR signaling. J Diabetes Investig 14: 535–547.
• 14   Jiang  J,  Dong  C,  Zhai  L,  Lou  J,  Jin  J, et al. (2021) Paeoniflorin suppresses TBHP-induced oxidative stress and apoptosis in human umbilical vein endothelial cells via the Nrf2/HO-1 signaling pathway and improves skin flap survival. Front Pharmacol 12: 735530.
• 15   Livak  KJ,  Schmittgen  TD (2001) Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25: 402–408.
• 16   Jiashuo  WU,  Fangqing  Z,  Zhuangzhuang  LI,  Weiyi  J,  Yue  S (2022) Integration strategy of network pharmacology in Traditional Chinese Medicine: a narrative review. J Tradit Chin Med 42: 479–486.
• 17   Naito  H,  Iba  T,  Takakura  N (2020) Mechanisms of new blood-vessel formation and proliferative heterogeneity of endothelial cells. Int Immunol 32: 295–305.
• 18   Zhang  J,  Hua  W,  Zhao  X,  Yang  F,  Guo  T, et al. (2021) Paeoniflorin alleviates endothelial dysfunction caused by overexpression of soluble fms-like tyrosine kinase 1 and soluble endoglin in preeclampsia via VEGFA upregulation. Biosci Biotechnol Biochem 85: 814–823.
• 19   Sabbah  DA,  Hajjo  R,  Sweidan  K (2020) Review on epidermal growth factor receptor (EGFR) structure, signaling pathways, interactions, and recent updates of EGFR inhibitors. Curr Top Med Chem 20: 815–834.
• 20   Belmadani  S,  Palen  DI,  Gonzalez-Villalobos  RA,  Boulares  HA,  Matrougui  K (2008) Elevated epidermal growth factor receptor phosphorylation induces resistance artery dysfunction in diabetic db/db mice. Diabetes 57: 1629–1637.
• 21   Wang  H,  Ma  S,  Li  J,  Zhao  M,  Huo  X, et al. (2018) ADAM17 participates in the protective effect of paeoniflorin on mouse brain microvascular endothelial cells. J Cell Physiol 233: 9320–9329.
• 22   Chen  K,  Zhang  Y,  Qian  L,  Wang  P (2021) Emerging strategies to target RAS signaling in human cancer therapy. J Hematol Oncol 14: 116.
• 23   Kowluru  RA,  Kanwar  M (2009) Translocation of H-Ras and its implications in the development of diabetic retinopathy. Biochem Biophys Res Commun 387: 461–466.
• 24   Roskoski R  Jr (2004) Src protein-tyrosine kinase structure and regulation. Biochem Biophys Res Commun 324: 1155–1164.
• 25   Long  L,  Li  Y,  Yu  S,  Li  X,  Hu  Y, et al. (2019) Scutellarin prevents angiogenesis in diabetic retinopathy by downregulating VEGF/ERK/FAK/Src pathway signaling. J Diabetes Res 2019: 4875421.
• 26   Hong  X,  Li  L,  Guo  D,  Yang  Z,  Yan  J (2020) Paeoniflorin improves the permeability of cardiac microvascular endothelial cells by regulating Src/vascular endothelial-cadherin pathway. Zhonghua Wei Zhong Bing Ji Jiu Yi Xue 32: 83–87 (In Chinese).
• 27   Zuehlke  AD,  Beebe  K,  Neckers  L,  Prince  T (2015) Regulation and function of the human HSP90AA1 gene. Gene 570: 8–16.
• 28   Saik  OV,  Klimontov  VV (2020) Bioinformatic reconstruction and analysis of gene networks related to glucose variability in diabetes and its complications. Int J Mol Sci 21: 8691.
• 29   Zhang  MY,  Ma  LJ,  Jiang  L,  Gao  L,  Wang  X, et al. (2023) Paeoniflorin protects against cisplatin-induced acute kidney injury through targeting Hsp90AA1-Akt protein-protein interaction. J Ethnopharmacol 310: 116422.
• 30   Lakshmikanthan  S,  Sobczak  M,  Li Calzi  S,  Shaw  L,  Grant  MB, et al. (2018) Rap1B promotes VEGF-induced endothelial permeability and is required for dynamic regulation of the endothelial barrier. J Cell Sci 131: jcs207605.
• 31   Xiao  L,  Zhu  X,  Yang  S,  Liu  F,  Zhou  Z, et al. (2014) Rap1 ameliorates renal tubular injury in diabetic nephropathy. Diabetes 63: 1366–1380.
• 32   Roskoski R  Jr (2021) Blockade of mutant RAS oncogenic signaling with a special emphasis on KRAS. Pharmacol Res 172: 105806.
• 33   Vairaktaris  E,  Goutzanis  L,  Yapijakis  C,  Vassiliou  S,  Spyridonidou  S, et al. (2009) Diabetes enhances the expression of H-ras and suppresses the expression of EGFR leading to increased cell proliferation. Histol Histopathol 24: 531–539.
• 34   Kanwar  M,  Kowluru  RA (2008) Diabetes regulates small molecular weight G-protein, H-Ras, in the microvasculature of the retina: implication in the development of retinopathy. Microvasc Res 76: 189–193.