Tiao-Bu-Fei-Shen Formula Improves Glucocorticoid Resistance of Chronic Obstructive Pulmonary Disease via Downregulating the PI3K-Akt Signaling Pathway and Promoting GRα Expression

Objective To predict and determine the mechanism through which Tiao-Bu-Fei-Shen (TBFS) formula improves glucocorticoid resistance in chronic obstructive pulmonary disease (COPD), using network pharmacology, molecular docking technology, and in vitro studies. Methods The main active components and associated targets of TBFS were screened using the systems pharmacology database of traditional Chinese medicine database (TCMSP). The main COPD targets were retrieved from the Human Gene (GeneCards) and DrugBank databases. A protein-protein interaction (PPI) network was constructed using the protein interaction platform STRING and Cytoscape 3.6.1. Gene ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genome Pathway (KEGG) analyses were performed using the biological information annotation database Metascape. Molecular docking was performed using the AutoDock Vina software. THP-1 monocytes were treated with TBFS-containing serum and cigarette smoke extract (CSE) for 48 h, and cell proliferation in each group was determined using cell counting kit-8 (CCK-8). A COPD cell model was constructed by stimulating THP-1 monocytes with CSE for 12 h. A lentivirus vector for RNA interference of histone deacetylase 2 (HDAC2) gene was constructed and transfected into the THP-1 monocytes, and the transfection efficiency was verified using quantitative polymerase chain reaction (qPCR) and western blotting (WB). The expression of HDAC2 in each group of cells was detected using qPCR, and the expression of HDAC2, phosphoinositide-3 kinase (PI3K) p85α, glucocorticoid receptor α (GRα), and P-AKT1 in each group of cells was detected through WB. Results A total of 344 TBFS active components, 249 related drug targets, 1,171 COPD target proteins, and 138 drug and disease intersection targets were obtained. Visual analysis of the PPI network map revealed that the core COPD targets of TBFS were AKT1, IL-6, TNF, TP53, and IL1-β. KEGG pathway enrichment analysis resulted in the identification of 20 signaling pathways as the main pathways involved in the action of TBFS against COPD, including the PI3K-Akt, TNF, and IL-17 signaling pathways. Molecular docking experiments revealed a strong binding capacity of kaempferol, luteolin, and quercetin to the ATK1 protein in TBFS, with quercetin performing the best. PCR results showed that treatment with TBFS significantly increased the expression levels of HDAC2 in the COPD model. WB results showed that TBFS treatment significantly increased the expression levels of GRα and HDAC2 in the COPD model, while reducing the expression levels of P-AKT1. Conclusion TBFS treatment improves glucocorticoid resistance observed in COPD through downregulation of the PI3K-Akt signaling pathway and promotion of GRα expression.


Introduction
Chronic obstructive pulmonary disease (COPD) has a heavy disease burden globally, and its prevention has proven to be arduous. Te disease is characterized by persistent respiratory symptoms and limited airfow and is mainly caused by airway and/or alveolar abnormalities caused by toxic particles or gases. COPD is currently the third leading cause of death and ffth leading cause of disease worldwide [1]. Te burden of COPD in China is particularly severe. Studies have shown that the prevalence of COPD in people over 40 years of age is as high as 13.7%, with approximately 90 million patients being afected nationwide [2]. Te direct medical expenses for patients with COPD in China are approximately 72-3,565 US dollars per person per year, accounting for 33.33-118.09% of the average annual local income [3]. Te number of people with COPD in China is expected to reach 103.3 million in 2039, while the total loss of qualityadjusted life years and excess deaths due to COPD are estimated to be 253.6 million and 3.9 million, respectively; the direct and indirect costs of COPD are estimated to be $3.1 trillion and $360.5 billion, respectively [4]. COPD is a persistent, airway infammatory disease, and infammation plays a key role in the development of this disease; therefore, anti-infammatory treatment is very important for COPD. Glucocorticoids, the most potent anti-infammatory drugs, are recommended by the GOLD and GINA guidelines for infammatory airway diseases such as COPD and asthma. In COPD, glucocorticoid resistance is widely observed, which leads to a severe weakening of the anti-infammatory efects of glucocorticoids [5]. Studies have revealed that the degree of resistance to glucocorticoids at diferent stages of COPD is inconsistent and closely associated with lung function. Te lower the forced expiratory volume in 1 s/forced vital capacity (FEV1/FVC) ratio, the more severe is the steroid resistance [6]. Uncontrolled airway infammation in COPD not only further deteriorates the clinical symptoms and quality of life of patients but also increases the risk of disability and death, leading to increased economic burdens on the families of patients as well as society. Although glucocorticoid resistance in COPD has gained increasing attention over the past years, the underlying mechanism remains to be understood, and there is a lack of efective intervention in clinical practice.
Te phosphatidylinositol-3-kinase (PI3K)/serinethreonine protein kinase (Akt) signaling pathway plays an important role in diseases involving chronic airway infammation via regulating the infammatory mediator release, infammatory cell activation, and airway remodeling. Recently, the role of the PI3K/Akt signaling pathway in the infammatory mechanism, glucocorticoid resistance, and anti-infammatory treatment of COPD has received extensive attention. Moreover, several studies have investigated the regulation of diferent signaling molecules in the PI3K/ Akt/NF-κB signaling pathway as a treatment strategy for COPD [7,8].
Histone deacetylase (HDAC) is an infammatory gene regulatory enzyme, and together with histone acetyltransferase (HAT) in the nucleus, it maintains the dynamic balance of histone acetylation and deacetylation, which play a key role in the transcription and silencing of infammatory genes; moreover, increased HAT or reduced HDAC expression results in signifcant upregulation of infammatory gene expression [9]. Recently, HDAC2 expression was reported to be closely associated with COPD glucocorticoid resistance [10]. Glucocorticoids not only form complexes with receptors and move into the nucleus but also recruit HDACs in specifc regions of the cells leading to their antiinfammatory efects. Tus, if the expression of HDACs is reduced, the anti-infammatory efects of glucocorticoids will be signifcantly reduced.
Previous studies have reported that the expression levels of PI3K δ, NF-κB, IL-6/8, and TNF-α as well as Akt phosphorylation were signifcantly increased in lung macrophages and peripheral blood monocytes in patients with COPD, while HDAC2 expression and activity were significantly reduced. FEV1%pred was found to be positively correlated with HDAC2 expression and HDAC activity [11,12]. In vitro and in vivo studies were used to confrm that by inhibiting the infammatory response under oxidative stress, blocking or knocking out PI3Kδ can signifcantly improve the activity of HDAC2 and the efcacy of glucocorticoid treatment [13,14]. Oxidative stress is a key mechanism involved in COPD glucocorticoid resistance; it acts through activation of PI3K β/signaling and downregulation of HDAC2 expression [5,7].
According to traditional Chinese medicine (TCM), COPD belongs to the category of lung distention diseases (Fei-Zhang disease) [15]. Lung-kidney Qi defciency syndrome is one of the most common syndromes of Fei-Zhang diseases, and Tiao-Bu-Fei-Shen (TBFS) therapies are commonly used to treat these diseases [16,17]. A multicenter clinical study reported that TBFS improved symptoms, reduced the frequency of exacerbations, and improved exercise tolerance and quality of life in patients with COPD [16]. Previous studies have also revealed that TBFS can signifcantly reduce pulmonary infammation responses, alleviate airway remodeling, and regulate T lymphocyte subsets and CD4 + CD25 + cells [18][19][20][21] (Zhebeimu); it has been successfully used to treat COPD at our center. In clinical practice, we observed that TBFS treatment achieved better efects in patients with COPD with glucocorticoid resistance. Compared with standard Western medicine, TBFS can rapidly improve clinical symptoms and lung function, reduce the need for glucocorticoids, and reduce repeated aggravation of the disease. Although TBFS has shown defnite clinical efcacy in patients with COPD, the specifc mechanism by which it improves glucocorticoid resistance remains unclear. In addition, TBFS is a TCM with multiple active components and multitarget regulatory effects. Terefore, it is difcult to use a single method to explain the scientifc basis and potential pharmacological mechanisms of action of this drug. Terefore, in the present study, we systematically predicted the mechanism of action of TBFS treatment in COPD at cellular, molecular, and genetic levels using network pharmacology; we also performed in vitro studies to verify our predictions. Tis study provides evidence of the various mechanisms through which TBFS treatment improves glucocorticoid resistance in COPD.

Collection of Chemical Components and Targets of TBFS.
Te TCM system pharmacology database and analysis platform (TCMSP, https://tcmspw.com/tcmsp.php) was used to retrieve the chemical constituents of the 13 herbs in TBFS. Screening was performed according to two attribute values, oral bioavailability (OB) ≥ 30%, and drug-likeness (DL) ≥ 0.18, to obtain eligible active compounds and their targets.

Collection of Disease-Associated Targets.
Te disease targets of COPD were screened in the GeneCards (https:// www.genecards.org/) and DrugBank (https://go.drugbank. com/) databases with the keywords of "Chronic Obstructive Pulmonary Disease" and "COPD." Te score value in the GeneCards database represents the closeness of the relationship between the disease and target. Te higher the score, the stronger is the association between the disease and target. If there are too many targets, those with a score greater than the median were set as potential COPD targets. Te disease targets obtained from the two databases were merged and duplicates were removed to identify the disease targets of COPD.

Construction of the Component-Common Target-Disease
Network. To clarify the interaction between COPD and TBFS drug targets, an online drawing tool (https://www. bioinformatics.com.cn/) was used to draw Venn diagrams, and the potential targets of TBFS in the treatment of COPD were obtained. Second, the intersection targets of TBFS and COPD were imported into the STRING database (https://cn.string-db.org/) to construct a proteinprotein interaction (PPI) network. Tird, the "Network Analyzer" function in CytoScape 3.7.1 was used to perform topology analysis of the PPI network. Finally, the node size and color depth were adjusted according to the degree value, and the pharmacological efects of key targets were analyzed.

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Enrichment Analyses.
To further understand the functions of the above-screened target proteins and genes and their roles in signaling pathways, the Metascape platform was used to conduct KEGG and GO analyses, with P < 0.01 as the screening criterion. GO biological process and KEGG signaling pathway enrichment analyses were performed. Because of the large number of enrichment results, only the top 10 enrichment results with the smallest P value or the largest number of enriched targets were selected in the GO analysis, and only the top 20 signaling pathways with the smallest P value were selected for the KEGG analysis.

Molecular
Docking. Protein crystal structures were obtained using the UniProt database; 3D structures of the main compounds were obtained using the PUBCHEM database and energy was minimized by the AVOGADR under the MMFF94 force feld. Molecular docking was performed using AutoDock Vina 1.1.2, and the docking results were visually analyzed using the academic opensource version of PyMol.

Preparation of the Cigarette Smoke Extract (CSE) and
TBFS Drug-Containing Serum. First, two unfltered cigarettes (brand: Jinsheng, China Tobacco Industry Co., Ltd, Jiangxi, China; tar content: 11 mg; nicotine content: 1 mg; carbon monoxide content: 12 mg) were lit, followed by continuous suction through a syringe. Te smoke was dissolved in 10 mL serum-free culture medium to make a suspension with a concentration of approximately 1000 mL/L. Finally, the pH was adjusted to 7.4 and the solution was fltered through a 0.22-μm flter for experiments [22].
Te TBFS formula contained 13 herbs. Te botanical compositions are listed in Table 1. Te herbs were purchased from Hospital of Chengdu University of Traditional Chinese Medicine (Chengdu, China), and their voucher specimens were kept in the TCM Pharmacy of Hospital of Chengdu University of Traditional Chinese Medicine. Te equivalent daily dose for rats (g/kg) � 6.3 × the daily clinical dose for humans (g/kg). Sixteen specifed pathogen-free 8-week-old Wistar rats (males; 200 ± 10 g) were purchased from Liaoning Changsheng Experimental Animal Co., Ltd., (Animal ethical approval number: 2020090401). Rats were treated via oral gavage twice daily for 7 days with prepared TBFS. Te specifc preparation method of TBFS drugcontaining serum was consistent with that of our previously reported study [21].

IL-8 Levels Were
Determined by Enzyme-Linked Immunosorbent Assay (ELISA). All reagents and components were frst allowed to cool to 20°C, and the standards, quality controls, and samples were prepared in duplicate wells. Te working solutions of the various components of the kit were prepared and used according to the manufacturer's instructions [23]. Te supernatants were harvested, and 100 µL of antibody dilutions (IL-8, MM-1558H1, Jiangsu, China) were added to each well and incubated for 1 h. Secondary antibodies were added, and the reaction was terminated with the termination solution. A wavelength of 450 nm was used to measure the absorbance. blank serum (cell + blank serum + CSE + dexamethasone + TNF-α), and drug serum (cell + TBFS drug-containing serum + CSE treatment + dexamethasone + TNF-α) groups. Control patients were treated only with diferent concentrations of dexamethasone and TNF-α. Te CSE group was stimulated with CSE overnight, incubated with dexamethasone for 2 h, and then stimulated with TNF-α overnight. Te drug serum group was pretreated with a TBFS-containing serum for 2 h, stimulated with CSE overnight, incubated with dexamethasone for 2 h, and stimulated with TNF-α overnight. Te blank serum group was treated in the same manner as the drug-containing serum group, except for the blank control serum. Cell supernatants from each group were collected after 48 h of intervention, and IL-8 levels were measured using ELISA. Microsoft Excel was used to calculate the median inhibitory concentration of dexamethasone according to the inhibition rate of IL-8 at diferent concentrations of dexamethasone in each group.

Construction of HDAC2-Small Interfering RNA (siRNA) Interference Vector and HDAC2 Detection. Transfection
Reagent Lipofectamine ™ 3000 (L3000015, Invitrogen ™ , USA) and diluted HDAC2 siRNA (Abbexa, Cambridge, UK) were added to the Opti-MEM transfection medium at a specifc ratio [23]. Te HDAC2-siRNA interference vector was constructed and transfected into the target cells (THP-1 cells, CL-0109, China), and the transfection efciency was verifed through quantitative PCR (qPCR) and western blotting (WB). After successful transfection, the experimental groups were set as follows: control group, CSE group, blank serum group, drug serum group, empty load group (HDAC2 siRNA NC), and HDAC2-siRNA group. Te expression of HDAC2 in each group was detected using qPCR, and the expression of HDAC2, PI3K p85α, GRα, and P-AKT1 in each group was detected using WB.

Cell Counting Kit-8 (CCK-8) Assay.
First, the cells were digested, resuspended, counted, and plated at a density of 5 × 10 3 cells/well. Te cells were cultured for 48 h for detection. Te cells in the 96-well plate were then replaced with the same medium, such that each well contained 100 μL. Subsequently, 10 μL of CCK-8 reagent was added to each well, and the cells were incubated in an incubator for 2 h. Finally, the microplate reader was used to measure the absorbance of each well at a wavelength of 450 nm and the survival rate was calculated [21].
2.11. WB Analysis. Te WB assay was performed according to the manufacturer's instructions. After the total protein concentration was determined, we separated equal amounts of proteins from each well by vertical electrophoresis using sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE). We transferred the proteins to polyvinylidene fuoride membranes, followed by immunoblotting [23]. Te gray value of each band was analyzed using the Image J software.

Quantitative Real-Time Quantitative PCR (qPCR).
Total RNA was extracted from THP-1 cells using TRIzol reagent (CW0580S, China) and reverse transcribed into cDNA using a reverse transcription kit (R223-01, China). PCR amplifcation reaction was then performed. Te amplifcation conditions were 95°C for 10 min, 40 cycles of denaturation at 95°C for 10 s, annealing at 58°C for 30 s, and extension at 72°C for 30 s. β-Actin was used as an internal reference, and the relative expression of HDAC2 was calculated according to the 2 −△△Ct method [21]. Te primers and their sequences are listed in Table 2. 2.13. Statistical Processing. SPSS19.0 software was used for statistical analysis. All experiments were repeated three times, and the quantitative results are expressed as mean-± standard deviation (X ± S). Quantitative numerical comparisons among multiple groups were performed using oneway analysis of variance (ANOVA), and pairwise comparisons were performed using the Least Signifcant Diference (LSD) method. Te inspection level was α � 0.05, the graphs

Screening of Active Compounds in TBFS.
A total of 1818 compounds present in TBFS were preliminarily extracted from the TCMSP database. A total of 344 active compounds were obtained after screening based on OB ≥ 30% and druglike properties (DL) ≥ 0.18. A total of 249 TBFS targets were obtained after the corresponding gene names were merged, and duplicates were deleted (Supplementary Materials).

Known Terapeutic Targets Acting on COPD.
A total of 1,983 COPD targets were obtained from the GeneCards database, setting targets with a score greater than the median as potential targets for COPD. Te maximum COPD target score obtained by GeneCards was 36.72, and the minimum value was 0.10. A target with a score greater than the median was set as a potential target, and a total of 1128 targets were obtained. Combined with the DrugBank database to supplement COPD-related targets, duplicate values were deleted after merging to yield 1,171 COPD disease targets. Te Venn diagram shows 138 intersection targets for TBFS and COPD ( Figure 1).

Component-Common Target-Disease Network Construction.
Compounds were linked to the target site using CytoScape 3.7.1 to obtain the "TCM-active compound-target" network. Te network consisted of 165 nodes and 138 edges (Figure 2), where the nodes represented the TCM active compound and corresponding target, and the edges represented the interaction between the active compound and target protein. Te top three compounds with the most targets were quercetin, luteolin, and kaempferol, which may be the key compounds in TBFS that play a role in the treatment of COPD.

PPI Network Construction.
Te 138 intersecting targets were imported into the String11.0 platform to obtain the interaction between the targets. As shown in Figure 3, under the condition of moderate confdence of 0.4, the network graph was revealed to contain 138 nodes. Te obtained results were imported into CytoScape 3.7.1 software to construct the PPI network. Te nodes were the targets, the edges were the interactions between the targets, and the size and color of the nodes refected the magnitude of the degree value. Te degree values of 15 targets, including AKT1, CASP3, CXCL8, EGFR, ESR1, FOS, IL-1β, IL-6, JUN, MMP9, PPARG, PTGS2, TNF, TP53, and VEGFA were high, and it was speculated that these targets may be key targets of TBFS in the treatment of COPD. Among these, the top fve targets were considered core targets, namely, AKT1, IL-6, TNF, TP53, and IL1β.

GO and KEGG Enrichment
Analyses. GO and KEGG enrichment analyses were performed using Metascape to further analyze the relationship between TBFS and COPD. Te results of GO enrichment analysis with P < 0.01 as the screening criterion showed that a total of 1740 items were involved in biological processes, 99 items were involved in the cellular components, and 169 items were involved in molecular function. Te top ten enrichment results were selected for analysis, and the entries are presented as bar plots (Figures 4(a) and 4(b)). GO analysis results showed that the biological process of TBFS in the treatment of COPD mainly involved positive regulation of calcidiol 1monooxygenase activity, heat generation, adenylate cyclaseinhibiting G protein-coupled acetylcholine receptor signaling pathway, intracellular steroid hormone receptor signaling pathway, and hormone response. Cellular components mainly included membrane rafts, membrane microdomains, vehicle lumens, and Bcl-2 family protein complexes. Molecular functions included G protein-coupled neurotransmitter receptor activity, G protein-coupled acetylcholine receptor activity, steroid hormone receptor activity, histone acetyltransferase binding, and histone deacetylase binding. Te details are shown in Figures 4(a) and 4(b). KEGG pathway enrichment analysis yielded 184 KEGG signaling pathways, and the top 20 were selected according to the logP value standard from small to large. Te enriched pathways were visually displayed using a bubble chart. Te main pathways of TBFS in the treatment of COPD included    Figure 4(c).

Molecular Docking.
Docking simulation technology is a convenient and efective means of exploring the interactions between small molecules and targets. We used AutoDock Vina 1.1.2 software to conduct docking studies on small molecules such as kaempferol, luteolin, quercetin, and ATK1, and the binding energy scores are shown in Table 3. A negative binding energy indicates the possibility of binding, and a value less than −5 kcal/mol is generally considered to indicate a high likeliness to bind. As shown in Table 3, all combinations have binding afnities below −5 kcal/mol, implying that these molecules have a potential active efect on all three proteins. Moreover, by comparing the binding afnity size, the small molecules kaempferol, luteolin, and quercetin showed similar binding abilities to the ATK1 protein, with quercetin performing the best.
Overall, the main interactions between ATK1 protein and kaempferol, luteolin, and quercetin were hydrogen bonding and hydrophobic interactions, which may be the main reason for the efect of these three small molecules on the ATK1 protein.

Screening of CSE and TBFS Drug-Containing Serum.
Compared with the control group, the TBFS-containing serum with diferent concentrations of low, medium, and high doses had a certain degree of promoting efect on cell proliferation. Based on these results, 10% drugcontaining serum was selected for subsequent experiments. Compared with the control group, the cell proliferation ability after 25, 50, and 100 mL/L CSE treatment was signifcantly decreased, and the diference was statistically signifcant (P < 0.05). To ensure that a certain degree of cell damage would not prevent subsequent detection due to low cell viability, 25 mL/L CSE was selected for subsequent experiments ( Figure 6).  Evidence-Based Complementary and Alternative Medicine

Efect of TBFS-Containing Serum on the Half-Inhibitory
Concentration of Dexamethasone. Te inhibition rate of IL-8 in the CSE and blank serum groups decreased with an increase in dexamethasone concentration, indicating that the anti-infammatory efect of glucocorticoids on the COPD model decreased, suggesting the existence of glucocorticoid resistance. Te inhibition rate of IL-8 in the control and TBFS-containing serum groups increased with an increase in dexamethasone concentration, indicating that the drug-containing serum could enhance the anti-infammatory efect of glucocorticoids and improve the resistance to glucocorticoids in the COPD model. Details are presented in Table 4.

Validation of HDAC2 siRNA Transfection Efciency.
Te PCR and WB results showed that the relative expression levels of HDAC2 mRNA and protein were both signifcantly reduced in interference groups 1, 2, and 3 when compared with the control group, suggesting a successful interference validation (Figure 7).

Expression of HDAC2 after HDAC2 siRNA Interference.
Compared with the control group, the expression levels of HDAC2 in the CSE and the TBFS drug-containing serum groups were signifcantly increased, and the expression levels of HDAC2 in the blank serum and the HDAC2 siRNA groups were signifcantly decreased; the diferences were statistically signifcant (P < 0.05). Compared with the blank serum group, the expression levels of HDAC2 in the TBFScontaining serum and the HDAC2 siRNA NC groups were signifcantly increased, and the diference was statistically signifcant (P < 0.05). It is suggested that TBFS-containing serum can increase the expression of HDAC2 in human monocyte-macrophage THP-1 cells (Figure 8).

Expression of Key Proteins in the PI3K-AKT Signaling
Pathway. Compared with the CSE group, the expression level of GRα in the TBFS-containing serum group and the HDAC2 siRNA NC group was signifcantly increased, while the expression level of GRα in the HDAC2 siRNA group was signifcantly decreased. Compared to the blank serum group, Circular node represents the pathway, size represents the number of targets enriched by the pathway, and color represents P value. the expression level of GRα in the TBFS-containing serum group and the HDAC2 siRNA NC group was signifcantly increased, while the expression level of GRα in the HDAC2 siRNA group was signifcantly downregulated. Experiments showed that drug-containing serum signifcantly increased the expression of GRα in the COPD model (Figure 9(a)). Compared with the control group, the level of P-AKT1 in the CSE group was signifcantly increased, and the levels of P-AKT1 in the TBFS-containing serum, HDAC2 siRNA NC, and HDAC2 siRNA groups were signifcantly downregulated. Experimental results revealed that the drug-containing serum reduces the expression of P-AKT1 (Figure 9(b)).
Compared with the CSE group, the expression level of HDAC2 in the blank serum group was signifcantly decreased, whereas the expression level of HDAC2 in the TBFS-containing serum, HDAC2 siRNA NC, and HDAC2 siRNA groups was signifcantly increased. Tis study showed that treatment with drug-containing serum increased HDAC2 levels (Figure 9(c)).
Compared with the control group, there was no signifcant change in PI3k expression in the CSE and blank serum groups, while the expression of PI3k in the TBFScontaining serum, HDAC2 siRNA NC, and HDAC2 siRNA groups was increased (Figure 9(d)).

Discussion
COPD is an incomplete, reversible, chronic infammatory airway disease characterized by progressive airfow restriction; the infammatory response is a core mechanism underlying the progression of COPD. Inhibiting the infammatory response is a key treatment for COPD; according to the GOLD guidelines, glucocorticoids can be used to manage acute aggravation treatment for severe COPD. However, patients with COPD may exhibit diferent degrees of glucocorticoid resistance, which often signifcantly attenuates the anti-infammatory efects of glucocorticoids [24]. Recently, signifcant progress has been made in elucidating the mechanism underlying the development of glucocorticoid resistance in COPD, wherein glucocorticoid receptor and isoform expression levels, PI3K/AKT signaling pathway, and histone deacetylase expression levels were reported to play a major role [25]. Studies have shown that the expression and activity of diferent GR isoforms are critical for glucocorticoid-mediated anti-infammatory activity. GR includes two isoforms, namely, GR-α and GR-β. GR-α, which is mainly found in the cytoplasm. When GR-α is activated, it can form a complex by recruiting HDAC2 and inhibiting the formation of the NF-κB/HAT complex. In contrast, if the expression of GR-β is low, it can directly bind to GRE, as GR-β is a GR-α antagonist and can attenuate the anti-infammatory activity of GR-α. When the expression of GR-α is downregulated or that of GR-β is upregulated, the anti-infammatory efect of glucocorticoids is signifcantly weakened [6]. Hyperactivation of the PI3K-Akt signaling pathway also causes glucocorticoid resistance in COPD. PI3K is a family of proteins, which mainly catalyze the phosphorylation of phosphoinositide-3-OH ends and can be divided into three categories; the most widely studied category of PI3K proteins is class I. Serine-threonine protein kinase (Akt) is a key signal transduction molecule involved in the PI3K signaling pathway. Abundant reactive oxygen species (ROS) levels are observed in patients with COPD. When PI3K is activated by ROS such as superoxide anions and hydroxyl radicals, the second messenger phosphatidylinositol-3,4,5-triphosphate (PIP3) is produced at the plasma membrane. PIP3 binds to the PH domaincontaining signaling protein Akt and phosphatidyl kinasedependent kinase 1 (PDK1) in cells, prompting PDK1 to phosphorylate Tr308 of the Akt protein, ultimately leading to the activation of Akt. Akt, through phosphorylation, activates downstream target proteins, such as NF-κB and caspase, to regulate the proliferation, diferentiation, apoptosis, and migration of proinfammatory cells. Conversely, downregulation of HDAC2 expression leads to a decrease in the anti-infammatory efects of glucocorticoids [26,27].
COPD is characterized by progressive infammation in the small airway and lung parenchyma, mediated by increased expression of multiple infammatory genes, and increased HDAC2expression suppresses this infammation. In COPD, HDAC2 activity and expression are reduced in the peripheral lungs and alveolar macrophages, leading to an amplifed infammatory response. Corticosteroid resistance is observed in COPD because corticosteroids require HDAC2 to suppress infammatory gene expression, and reduction in HDAC2 expression is often secondary to increased oxidative and nitrifcation stress in the lungs of patients with COPD [5]. Although the mechanisms underlying glucocorticoid resistance in COPD have not been fully elucidated, it is widely accepted that a key mechanism underlying this phenomenon depends on oxidative stress downregulating HDAC2 expression through the activation of PI3K β/AKT signaling [28,29]. Although some studies have shown that erythromycin, roxithromycin, theophylline, rofumilast, tiotropium bromide, carbocysteine, progesterone, and ubiquitin-specifc protease USP17 can improve glucocorticoid resistance in COPD cell models [30][31][32][33][34][35]; these drugs have diferent degrees of side efects or lack clinical evidence, limiting their clinical use.
IL-8, the strongest chemokine produced by neutrophils, plays a signifcant role in airway infammation of COPD, which is widely involved in the pathological process of COPD [23]. IL-8 can induce neutrophils to migrate towards the site of infammation, thus increasing the burden of the infammatory site. IL-8 levels are also a marker of the degree of infammation in COPD. Dexamethasone, as a common glucocorticoid, has a powerful anti-infammatory efect. In COPD, although the concentration of dexamethasone continues to increase, its inhibitory efect on IL-8 does not increase, indicating the presence of glucocorticoid resistance [36]. Tis study showed that the inhibitory efect of dexamethasone on IL-8 in the COPD cell model was signifcantly weakened, while TBFS could increase the inhibitory rate of IL-8, which revealed that TBFS could reduce glucocorticoid resistance and partially restore the sensitivity of THP-1 monocytes to dexamethasone.
TCM is efective for the treatment of COPD. Studies have shown that TCM can increase the expression of HDAC2, thereby improving the anti-infammatory efect of glucocorticoids in the treatment of COPD. Wu et al. showed that a Jingwei decoction combined with budesonide inhalation increased the expression of HDAC2 and reduced the expression of TNF-α, thereby improving the symptoms of COPD [37]. Li et al. reported that a Quanzhen Yiqi decoction could induce the apoptosis of COPD alveolar macrophages, regulate the expression of HDAC2, and produce an overall anti-infammatory efect [38]. Wu et al. [39] found that a Shenqi Bufei decoction can inhibit ASM proliferation in a COPD rat model with Lung Qi defciency syndrome and improve glucocorticoid resistance. Tis mechanism involves increased expression of HDAC2 and inhibition of NF-KB p65 activation. Siqing et al. showed that Erchen decoction may upregulate the expression of the HDAC2 gene in peripheral blood mononuclear cells (PBMCs) and inhibit the transcription and translation of the TGF-β1 gene, thereby antagonizing airway infammation in COPD rats and protecting the lung tissue [40]. Zhang et al. showed that baicalin ameliorated CS-induced airway infammation in rats by enhancing HDAC2 protein expression and inhibiting the Evidence-Based Complementary and Alternative Medicine expression of NF-κB and its downstream target PAI-1 [41]. Hu et al. not only showed that icariin reduced CSE-induced infammation, airway remodeling, and ROS production but also reported that treatment with the combination of icariin and glucocorticoids could reduce glucocorticoid resistance [42]. Although TCM has proved to be successful in improving glucocorticoid resistance in COPD, whether the underlying mechanism involves downregulation of PI3Kβ/AKT signaling and upregulation of HDAC2 expression remains unclear. Network pharmacology utilizes a combination of artifcial intelligence and medicine to facilitate biomedical research, analyze massive biomedical data, and establish transformation from data to knowledge. Network pharmacology has become a popular tool that is widely used in the elucidation of mechanisms in TCM pharmacology and screening of TCM active ingredients, drug repositioning, exploration of TCM compatibility mechanisms, and interpretation of the multicomponent, multitarget, and multipathway action mechanisms in TCM [43].
TBFS is a common TCM formula used by our team to treat COPD, and it has defnite clinical efcacy. TBFS comprises 13 Chinese medicines. Network pharmacology analysis showed that TBFS contained 1818 compounds, and 344 active compounds were obtained after screening, with   Te PI3K signaling pathway is extremely important for mediating various forms of cellular responses, ranging from cell survival, growth, proliferation, and diferentiation to DNA repair and apoptosis in diferent developmental and tissue contexts. Te expression of PI3K and its downstream mediators are upregulated during lung and airway remodeling in COPD [44]. Te diferential expression of PI3K during COPD progression implies dynamic regulation under pathological conditions. Dysregulation of PI3K signaling adversely afects not only the normal function of airway epithelial cells but also that of alveolar immune cells, leading to an excessive immune response [11]. Previous studies have confrmed that PI3Kβ/AKT signaling and HDAC2 expression play key roles in COPD glucocorticoid resistance. In the present study, we used network pharmacological analysis to show that TBFS can target AKT1 and  regulate PI3K-Akt signaling to ameliorate hormone resistance. To further validate the mechanism of action of TBFS in COPD glucocorticoid resistance, we constructed a COPD cell model and used a HDAC2 interference vector. PCR results showed that treatment with the TBFScontaining serum signifcantly increased the HDAC2 expression level in the COPD cell model; WB results showed that serum TBFS signifcantly increased the expression of GR-α and HDAC2 and decreased the expression of P-AKT1. Tus, we verifed the exact mechanism of action of TBFS in the treatment of glucocorticoid resistance in COPD using network pharmacology and in vitro experiments and provided evidence for the clinical application of TBFS. Although this study revealed the exact role of TBFS in ameliorating glucocorticoids in COPD, it had some limitations. First, limited by experimental funding, this study did not verify the role of TBFS at the global level in animals. Second, TBFS is a TCM compound, and network pharmacology analysis revealed that this prescription may act on COPD glucocorticoid resistance through a variety of signaling pathways and targets; however, this study only focused on the PI3K-Akt signaling pathway, and there may be a selection bias. In addition, network pharmacology research usually starts with the target proteins shared by TCM and diseases and rarely considers the combination of drug components with other biological functional molecules, such as metabolites, long noncoding RNA (lncRNA), and circular RNA (circRNA). Finally, although TBFS has been used for the treatment of COPD in our center for many years and clinical observations have revealed that it has the efect of improving glucocorticoid resistance, the observation sample number is limited, and there is a lack of rigorous randomized controlled trials to confrm these observations. Terefore, it is important to verify the role of TBFS in COPD glucocorticoid resistance at the global level in a multicenter, randomized, controlled study. Cytokine receptors Figure 10: Schematic illustration of TBFS formula treatment of COPD glucocorticoid resistance. Tis fgure summarizes the results presented in this study, in part. Green arrows indicated that TBFS treated COPD glucocorticoid resistance by downregulating the AKTsignaling pathway. Black arrows indicated that TBFS treated COPD glucocorticoid resistance by upregulating the expression of GRα and HDAC2.

Conclusion
Herein, we used network pharmacology to reveal that TBFS treatment may improve glucocorticoid resistance in COPD through multiple signaling pathways, such as the PI3K-Akt signaling pathway. We used an in vitro study to confrm that treatment with TBFS drugcontaining serum improves glucocorticoid resistance in COPD via the downregulation of the PI3K-Akt signaling pathway and promotion of GRα expression ( Figure 10).

Data Availability
Te data presented in this study can be obtained from the corresponding author upon request.

Conflicts of Interest
Te authors declare that they have no conficts of interest.

Authors' Contributions
Z-J and Z-WS designed the outline of the study. Z-PC, Y-W, and M-JL performed experiments, conceived the study, and drafted and revised the manuscript. M-JL and Y-W were involved in performing experiments, acquisition of data, and statistical analysis. Y-W and C-KL contributed to the data analysis and interpretation. All authors contributed toward data analysis, drafting, critically revising the paper, giving fnal approval of the version to be published, and agreeing to be accountable for all aspects of the work. Te authors Pengcheng Zhou, Jianli Ma, and Wei Yu have contributed equally to this work and share the frst authorship.