Biomarkers and coptis chinensis activity for rituximab-resistant diffuse large B-cell lymphoma: Combination of bioinformatics analysis,network pharmacology and molecular docking

Abstract

BACKGROUND:

Rituximab resistance is one of the great challenges in the treatment of diffuse large B-cell lymphoma (DLBCL), but relevant biomarkers and signalling pathways remain to be identified. Coptis chinensis and its active ingredients have antitumour effects; thus, the potential bioactive compounds and mechanisms through which Coptis chinensis acts against rituximab-resistant DLBCL are worth exploring.

OBJECTIVE:

To elucidate the core genes involved in rituximab-resistant DLBCL and the potential therapeutic targets of candidate monomers of Coptis chinensis.

METHODS:

Using the Traditional Chinese Medicine System Pharmacology Database and Analysis Platform (TCMSP), the Similarity Ensemble Approach and Swiss Target Prediction, the main ingredients and pharmacological targets of Coptis chinensis were identified through database searches. Through the overlap between the pharmacological targets of Coptis chinensis and the core targets of rituximab-resistant DLBCL, we identified the targets of Coptis chinensis against rituximab-resistant DLBCL and constructed an active compound-target interaction network. The targets and their corresponding active ingredients of Coptis chinensis against rituximab-resistant DLBCL were molecularly docked.

RESULTS:

Berberine, quercetin, epiberberine and palmatine, the active components of Coptis chinensis, have great potential for improving rituximab-resistant DLBCL via PIK3CG.

CONCLUSION:

This study revealed biomarkers and Coptis chinensis-associated molecular functions for rituximab-resistant DLBCL.

Keywords

Rituximab resistance differentially expressed genes WGCNA Coptis chinensis network pharmacology

1. Introduction

Diffuse large B-cell lymphoma (DLBCL) accounts for 85% to 90% of all lymphomas [1]; 20% of patients are primary refractory (progressing during or immediately after treatment) to first-line standard therapy (R-CHOP), and 30% of patients relapse after achieving complete remission (CR) with R-CHOP [2]. The objective response rate (ORR) for patients with recurrent/refractory disease receiving second-line therapy was 26%, with a median overall survival of only 6.3 months [3]. Therefore, R-CHOP, as the first-line regimen, remains the core treatment for DLBCL. Patients with DLBCL who fail first-line therapy, especially those with primary refractory disease and early recurrence, have a poor prognosis and short survival. Rituximab resistance is an important cause of R-CHOP failure [4]. Approximately 30%–60% of indolent non-Hodgkin’s lymphoma (iNHL) patients who are not receiving rituximab treatment (including those who relapse or receive only chemotherapy) have certain drug resistance to rituximab [5]. Nearly 60% of DLBCL patients who were previously sensitive to rituximab therapy developed rituximab resistance at the time of retreatment [6]. Rituximab resistance is one of the great challenges in the treatment of DLBCL.

Progression during rituximab treatment or relapse within 6 months after treatment was considered rituximab resistance [7]. Downregulation, adjustment, mutation or loss of CD20 is an important mechanism of rituximab resistance. Acquired CD20 mutations were reported by Terui et al. These mutations diminish the mean fluorescence intensity of CD20, which may affect antibody binding and induce drug resistance [8]. Another mechanism of CD20 loss in rituximab-resistant tumour cells is known as “shaving”, or knockout. Binding of rituximab to CD20 results in the translocation of CD20 to lipid valves within the cell membrane, after which Fc $\gamma$ RIIb-mediated downregulation of CD20 occurs, leading to drug resistance [9]. Recognition of antigen in combination with rituximab by cells carrying the Fc receptor triggers “shaving” [10]. In addition, the complement-dependent cytotoxicity (CDC) of rituximab is blocked by membrane complement regulatory proteins (mCRPs), such as CD46, CD55, CD35 and CD59, which are possible causes of rituximab resistance [11]. Impaired apoptotic signalling pathways in tumour cells may also result in resistance to rituximab. In rituximab-resistant cell lines, NF-kB pathway overactivation upregulates the expression of the Bcl-2 family of antiapoptotic proteins, causing apoptosis resistance [12]. Antibody-dependent cell-mediated cytotoxicity (ADCC) is the main effector mechanism of rituximab and is based on the activation of effector cells. In lymphoma patients, effector cells are depleted, which induces ADCC resistance and affects the effectiveness of rituximab [13]. Signal transduction in the tumour microenvironment may play a role in rituximab resistance, but the understanding of the underlying microenvironmental resistance mechanisms is still limited. However, the variations in cell lines and animal models, the different types of drug-resistant clones and the heterogeneity of tumours make these mechanisms uncertain and limited. The phenomenon and mechanism of rituximab resistance need to be further explored, which may also be helpful for solving the problems associated with resistance to other monoclonal antibodies in the future.

Figure 1.

Workflow of the study.

Rituximab resistance promoted the development of new CD20 monoclonal antibodies. Humanized anti-CD20 agents, including ofatumumab, IMMU106 (veltuzumab), and ocrelizumab, which are considered second-generation anti-CD20 agents and 3rd-generation anti-CD20 agents, include AME-133v (ocaratuzumab), PRO131921 and GA101 (obinutuzumab), respectively. These new monoclonal antibodies do not embed into lipid valves after binding to CD20 and induce high levels of ADCC and apoptosis without triggering CDC. However, these treatments have a limited effect on rituximab-resistant patients. In the GAUSS study, 175 iNHL patients who relapsed after treatment with rituximab regimens were randomized to receive either rituximab (375 mg/m2 $\times$ 4) or otuzumab (1000 mg $\times$ 4) once weekly for 4 consecutive weeks. Progression-free patients received maintenance therapy at the same dose every 2 months for 2 years. The ORR in the two groups was 44.6% and 33.3%, respectively, with otuzumab being superior, but the $p$ value of 0.08 was not significantly different. Moreover, there was no significant difference in PFS between the two groups [14]. On the other hand, rituximab resistance can be overcome by synergistic antitumour effects of other drugs in combination with rituximab, such as increasing the expression of the CD20 antigen, promoting apoptosis, and enhancing ADCC and CDC effects. Shimizu et al. reported that histone deacetylase inhibitors (HDACs) could augment the cytotoxic effects of rituximab by promoting CD20 expression and CDC [15]. The proteasome inhibitor bortezomib enhances rituximab sensitivity by inducing apoptosis in NF- $\kappa$ B-expressing cells [16]. The protein kinase C activator Bryostatin-1 (Bryostatin-1) increases the antitumour efficacy of rituximab by increasing CD20 expression on the surface of human B cells at the mRNA and protein levels through extracellular signal-regulated kinase signalling pathways [17]. However, these are mostly rituximab combination studies at the cellular level or even in non-rituximab-resistant cell lines. Some studies even showed opposite results [18, 19]. The improvement of rituximab resistance with these drugs is unknown, and additional drugs have to be developed for addressing rituximab resistance in DLBCL treatment.

Coptis chinensis and its active ingredients have been shown to have antitumour effects on many cancer species by inducing cell cycle arrest and apoptosis, inhibiting angiogenesis and inflammation, and suppressing invasion and metastasis [20]. Berberine, the most representative active ingredient of Coptis chinensis, eliminates DLBCL cells in vitro and in vivo by inhibiting c-myc expression, downregulating CD47 expression, and enhancing phagocytosis by macrophages [21]. In addition, berberine in combination with other chemotherapeutic agents has been reported to inhibit tumour progression. For example, the coadministration of berberine with evodiamine augmented the apoptosis of SMMC-7721 cells [22]. Berberine combined with cisplatin has a strong synergistic inhibitory effect on the growth of HeLa cells [23]. Coptis chinensis contains a variety of structural analogues of berberine, as well as a number of other components that have not been studied in depth; these compounds are potential antitumour agents with low price and few side effects. However, the impact of the monomeric components of Coptis chinensis on rituximab-resistant DLBCL has not been reported.

In this study, key genes associated with rituximab-resistant DLBCL were obtained by analysing differentially expressed genes (DEGs) and performing weighted gene coexpression network analysis (WGCNA). The biological functions and related signalling pathways of these genes were further determined by functional enrichment analysis, gene set enrichment analysis (GSEA), protein-protein interaction (PPI) network-based analysis, etc. Additionally, using network pharmacology and various online databases, we identified active targets of Coptis chinensis for the treatment of rituximab-resistant DLBCL and performed molecular docking to validate drug-target interactions, providing new references for the diagnosis and treatment of this disease. A schematic diagram of the integrated process is displayed in Fig. 1.

2. Materials and methods

2.1 Bioinformatics analysis

2.1.1 Identification of DEGs

Gene profiles of patients with rituximab-resistant DLBCL were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The DEGs in patients with rituximab-resistant DLBCL were screened and obtained using the ‘limma’ package of the R-language Bioconductor, for which the absolute values of log^{fold change (FC)} 1 and $p$ value $<$ 0.05 were used.

2.1.2 Coexpression network construction for WGCNA

First, we evaluated the variant genes to test their availability and used the R package WGCNA to construct a gene coexpression network. Then, we constructed an adjacency matrix to describe the correlation strength between the nodes. The matrix was constructed using the power function A_mn $=$ $|$ C_mn $|^{\wedge}\beta$ (C_mn $=$ Pearson’s correlation between Gene_m and Gene_n; A_mn $=$ adjacency between Gene m and Gene n). $\beta$ was a soft-thresholding parameter that could emphasize strong correlations between genes and penalize weak correlations. After choosing the power of 9, the adjacency was transformed into a topological overlap matrix (TOM), which measures the network connectivity of a gene defined as the sum of its adjacency with all other genes for the network gene ratio, and the corresponding dissimilarity (1-TOM) was calculated. Next, we performed hierarchical clustering to identify modules, each containing at least 30 genes (minModuleSize $=$ 30). To further analyse the modules, we calculated the dissimilarity of the module eigengenes, selected a cut-off for the module dendrogram and merged some of the modules. Finally, the module membership (MM) and gene significance (GS) of the target modules were calculated, and the genes screened with the criteria of $|$ MM $|$ $>$ 0.8 and $|$ GS $|$ $>$ 0.2 were considered hub genes.

2.1.3 Acquisition of core targets

Overlapping genes of DEGs and hub genes were obtained as core targets of rituximab-resistant DLBCL using Venny 2.1.0.

2.1.4 GSEA

GSEA software (version 3.0) was obtained from the following website: DOI:10.1073/pnas.0506580102, http://software.broadinstitute.org/gsea/index.jsp). The fold change in gene expression between the rituximab-resistant group and the rituximab-sensitive group was calculated, and a gene list was generated according to the change in $|$ log2FC $|$ . Then, we utilized GSEA-based enriched Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. A $P$ value $<$ 0.05 and an FDR $<$ 0.25 were considered to indicate statistical significance.

2.1.5 Enrichment analyses and network visualization

The potential gene functional annotations and pathway enrichment associated with the common DEGs were elucidated. GO and KEGG analyses were performed using the clusterProfiler package of R software (version 3.10.1). The enrichplot and DOSE packages were used to visualize the enrichment results to aid interpretation.

2.1.6 Establishment of a protein-protein interaction (PPI) network

PPI data of the obtained core targets were extracted from the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING, https://stringdb.org/). The core targets were all inputted into the STRING interaction database, and the Homo sapiens category was selected for an interactive network. The confidence score for the construction of an interactive network was set up with a medium score of 0.4 to 0.7. Cytoscape software (version 3.9.1) was used for module construction of the clustered networks.

2.2 Network pharmacology

2.2.1 Selection of monomers in Coptis chinensis

The main compounds of Coptis chinensis were retrieved from the TCMSP database (http://lsp.nwu.edu. cn/tcmsp.php) using Coptis chinensis as the key word. The main ingredients were those with an oral bioavailability (OB) $\geqslant$ 30%, a drug likeness property (DL) $\geqslant$ 0.18 and a cell permeability (Caco-2) $\geqslant-$ 0.4.

2.2.2 Acquisition of pharmacological targets of Coptis chinensis

The selected main components of Coptis chinensis were input into the Similarity Ensemble Approach (https://sea.bkslab.org/) and Swiss Target Prediction (http://swisstargetprediction.ch/) databases, with Homo sapiens serving as the species. The results were merged, and duplicates were deleted to obtain the pharmacological targets.

2.2.3 Identifying targets of Coptis chinensis against rituximab-resistant DLBCL

The core targets of rituximab-resistant DLBCL were intersected with the pharmacological targets of Coptis chinensis to identify potential targets against rituximab-resistant DLBCL. The APN was constructed and visualized via Cytoscape software (version 3.9.1).

2.2.4 Molecular docking analysis

Molecular docking analysis was performed using AutoDock Vina software (version 1.5.7). The molecular structures of the monomers in Coptis chinensis were obtained from the TCMSP database. The protein structures were acquired from the PDB database (https://www.rcsb.org/). All protein structures were processed with receptor grids to obtain various binding poses for the ligand at its active sites. The binding conformation with the lowest free binding energy was selected. Furthermore, the amino acid interactions between the protein targets and the ligand were visualized using Schrodinger-Maestro (version 8.5).

Figure 2.

Analysis of core targets on rituximab-resistant DLBCL. (a) Volcano map of DEGs related to rituximab-resistant DLBCL. (b) Heat map of DEGs related to rituximab-resistant DLBCL. (c) Cluster dendrogram of the co-expression network modules (1-TOM). (d) Cluster plot analysis of the relationship between modules. (e) Heatmap of the correlation between the module eigengenes and clinical traits of DLBCL. (f) Scatter plot analysis of the antiquewhite2 module. (g) Venn diagram depicting the overlap of DEGs and hub genes from WGCNA.

Figure 3.

Gene Set Enrichment Analysis. (a) Biological process pathways. (b) Molecular function pathways. (c) KEGG pathways.

3. Results

3.1 Identification of the core targets of rituximab-resistant DLBCL

A total of 194 DEGs related to Rituximab-resistant DLBCL were identified (Fig. 2a), of which 59 genes were upregulated and 135 were downregulated (Fig. 2b). For WGCNA, a total of 25 modules were identified using the dynamic tree cut package (Fig. 2c and d). We determined the correlations between the modules and the occurrence of rituximab-resistant DLBCL by establishing a module-trait relationship. The antiquewhite2 module (Fig. 2e) was significantly correlated with the occurrence of rituximab-resistant or sensitive DLBCL (coefficient of $\pm$ 0.93 and p value of 2.5e-3, respectively). The gene distribution in the antiquewhite2 module showed that GS and MM were strongly correlated, indicating that genes in this module were highly significantly associated with rituximab-resistant or sensitive DLBCL (Fig. 2f). A total of 166 hub genes were screened from the antiquewhite2 module as hub genes by the thresholds of $|$ GS $|$ $>$ 0.2 and $|$ MM $|$ $>$ 0.8. A comparison of the DEGs and hub genes identified by WGCNA revealed 34 crossover genes as core targets for rituximab-resistant DLBCL (Fig. 2g).

3.2 GSEA

GSEA was also conducted between rituximab-resistant and rituximab-sensitive DLBCL patients to explore the possible pathways and gene sets involved. The results revealed 5 GO-biological process (BP) pathways, 1 GO-molecular function (MF) pathway and 7 KEGG pathways. The GO-BP pathways were associated with N-negative regulation of the lipoprotein metabolic process, histone H2A acetylation, regulation of mitotic recombination, response to ammonium ions and dephosphorylation of the RNA polymerase IIC terminal domain (Fig. 3a). Only the acid phosphatase activity pathway in the GO-MF was related to rituximab resistance (Fig. 3b). The representative pathways identified by KEGG analysis included autoimmune thyroid disease, allograft rejection, and asthma (Fig. 3c).

3.3 Functional enrichment and PPI network of the core targets

The 34 core genes were subjected to GO and KEGG enrichment analyses. GO enrichment analysis revealed that genes involved in the positive regulation of alkaline phosphatase activity, antigen receptor-mediated signalling pathway, immune response-activating cell surface receptor signalling pathway and other pathways were involved (Fig. 4a). Additionally, in the KEGG pathway analysis, the top 10 pathways were found to be involved in cell adhesion molecule (CAM) metabolism, viral myocarditis and inflammatory bowel disease (IBD) (Fig. 4b). Next, STRING analysis was used to determine the PPIs mediated by the 34 core genes (Fig. 4c), and the 5 genes associated with the genes according to degree were RAC2, MS4A1, PIK3CG, EZR and EVI2B.

3.4 Screening of active ingredients and related pharmacological target genes

According to the TCMSP database, Coptis chinensis contains 48 ingredients in total. The patients were screened for OB $\geqslant$ 30%, DL $\geqslant$ 0.18 and Caco-2 $\geqslant-$ 0.4. Consequently, 13 ingredients were obtained (Table 1). Their targets were predicted by the similarity ensemble approach and Swiss Target Prediction, and 304 and 454 targets were identified, respectively. After the targets were merged and duplicate items were deleted, 623 pharmacological targets were obtained.

Table 1
Main compounds of Coptis chinensis

ID	Drug	OB (%)	DL (%)	Caco-2
MOL001454	Berberine	36.86	0.78	1.24
MOL013352	Obacunone	43.29	0.77	0.01
MOL002894	Berberrubine	35.74	0.73	1.07
MOL002897	Epiberberine	43.09	0.78	1.17
MOL002903	(R)-Canadine	55.37	0.77	1.04
MOL002904	Berlambine	36.68	0.82	0.97
MOL000622	Magnograndiolide	63.71	0.19	0.02
MOL000762	Palmidin A	35.36	0.65	$-$ 0.38
MOL000785	Palmatine	64.6	0.65	1.33
MOL000098	Quercetin	46.43	0.28	0.05
MOL001458	Coptisine	30.67	0.86	1.21
MOL002668	Worenine	45.83	0.87	1.22
MOL008647	Moupinamide	86.71	0.26	0.55

Figure 4.

Functional enrichment and PPI network of the core genes. (a) Top 10 results of GO enrichment analysis. (b) Top 10 results of KEGG enrichment analysis. (c) Analysis results of PPI network.

Figure 5.

Coptis chinensis-associated molecular functions for rituximab-resistant DLBCL. (a) Venn diagram depicting the overlap of the core genes and the pharmacological targets. (b) Active ingredient-target network. (c) Molecular docking of berberine, quercetin, epiberberine and palmatine to PI3K $\gamma$ . (d) Molecular docking of berberrubine to Ezrin.

3.5 Construction and validation of the active ingredient-target network against rituximab-resistant DLBCL

As shown in Fig. 5a, 2 intersection genes, PIK3CG and EZR, were potential targets of Coptis chinensis against rituximab-resistant DLBCL. The targeting relationship between Coptis chinensis and the intersection genes was determined by the active ingredient-target network (Fig. 5b). PIK3CG was the gene associated with the highest number of active components and included berberine, quercetin, epiberberine and palmatine. The corresponding ingredient of EZR was berberrubine. The interaction between the active ingredients and targets was validated using molecular docking analysis. The results illustrated that phosphoinositide 3-kinase subunit gamma (PI3K $\gamma$ ), which is encoded by PIK3CG, has very low binding energies, with all four active components ranging from $-$ 7.6 to $-$ 9.1 kcal/mol (Fig. 5c). However, the EZR-encoded protein Ezrin bound poorly to berberrubine (Fig. 5d). The binding energies and positions are given in Table 2.

Table 2
Molecular docking information on active ingredient-target of Coptis chinensis against rituximab-resistant DLBCL

Protein target	Drug	Binding energy	Residues involved in H-bond	Residues involved in Hydrophobic	Residues involved in charged (positive)
PI3K $\gamma$	Berberine	$-$ 7.9	None	Leu-791, Cys-817, Ala-822, Leu-823, Ile-828, Ile-881, Val-882	Lys-883
	Quercetin	$-$ 7.6	Asp-904	Met-804, Trp-812, Ile-831, Tyr-867, Ile-879, Ile-881, Val-882, Ala-885, Met953, Phe-961, Ile-963, Phe965	Lys-802, Lys-833, Lys-890
	Epiberberine	$-$ 9.1	None	Trp-201, Tyr-787, Leu-657, Phe-694, Phe-698	Arg-294, Arg-690, Arg-849
	Palmatine	$-$ 8.1	None	Trp-201, Leu-657, Leu-661, Phe-694, Phe-698, Tyr-787, Met-842, Cys-869, Ile-870	Arg-690, Arg-849
Ezrin	Berberrubine	$-$ 6.7	None	Trp-43, Tyr-44, Ile-94, Leu-290	Arg-293, Arg-294,

4. Discussion

Rituximab, a targeted monoclonal antibody against CD20, is effective at prolonging survival in DLBCL patients and has become a standard component of DLBCL treatment. However, the main obstacle to the efficacy of RTX is its resistance. Therefore, it is highly clinically important to potentiate the R-CHOP regimen by investigating the mechanisms of rituximab resistance and identifying drugs that may block the relevant mechanisms. This study is the first to identify core targets in rituximab-resistant DLBCL patients using differentially expressed gene (DEG) analysis, weighted gene coexpression network analysis (WGCNA) and other bioinformatics techniques and reveal that berberine, quercetin, epiberberine and palmatine may bind to Coptis chinensis through the core gene PIK3CG via network pharmacology and molecular docking analysis.

Three of the 34 core targets of rituximab-resistant DLBCL, CD99, RAC2, and MS4A1, were reported to be associated with the efficacy or adverse effects of the R-CHOP regimen. Junshik Hong et al. reported that, in DLBCL patients receiving the R-CHOP regimen, 2-year event-free survival (EFS) and overall survival (OS) were better in CD99 $+$ patients with the germinal centre B-cell (GCB) subtype than in those with the GCB/CD99 $-$ subtype. Conversely, among patients in the non-GCB subgroup, inferior EFS and OS were reported in non-GCB/CD99 $+$ patients. Superior 2-year EFS ( $p=$ 0.004) and 2-year OS ( $p=$ 0.003) were observed in patients with GCB/CD99 $+$ and non-GCB/CD99 $-$ tumours compared to the other patients, and the combination classification was found to be an independent prognostic factor [24]. For RAC2, the rs13058338 AT genotype is an independent predictor of the toxicity of the R-CHOP regimen [25]. MS4A1 is the coding gene of CD20, and its missense mutation can directly lead to rituximab resistance by reducing the expression and/or stability of CD20 [26].

GSEA of all genes, as well as GO and KEGG enrichment analyses of 34 crossover genes, suggested that phosphatase activity, cell adhesion molecules (CAMs), viral myocarditis, asthma, and allograft rejection play important roles in rituximab-resistant DLBCL. The level of alkaline phosphatase (ALP) was reported by Ching-Qing Cai et al. to be a new independent predictor of central nervous system relapse in DLBCL patients receiving the R-CHOP regimen [27]. In CAMs, cellular experiments have shown that knockdown of ICAM-1 can cause rituximab resistance [28]. There have also been reports about severe and fatal enterovirus-associated myocarditis after the use of rituximab [29, 30] and, conversely, the efficacy of rituximab for different types of asthma [31, 32] and renal transplant patients [33].

This network pharmacology study revealed that EZR and PIK3CG are potential key genes involved in the inhibition of rituximab resistance in Coptis chinensis. Ezrin, the coding product of EZR, is expressed in the stomach, kidney and intestine [34]. It is usually located in cell folds, cell pseudopodia, microvilli, cleavage grooves and contraction rings and participates in the regulation of cell adhesion [35]. EZR is an important factor related to tumour invasion [36]. PIK3CG encodes PI3K $\gamma$ and subsequently interacts with the G $\beta\gamma$ subunit of G protein-coupled receptors (GPCRs) through its regulatory subunit in response to related signals, thus contributing to many cellular physiological activities, such as metabolism, proliferation, survival, and migration [37, 38]. As reported by Ahmed Y. Ali et al., knockdown of PI3K $\gamma$ reduced the migration of CLL cells and cell lines. The expression of the PI3K $\gamma$ subunits increased in CLL cells in response to CD40L/IL-4, whereas BCR cross-linking had no effect. Overexpression of PI3K $\gamma$ subunits enhanced cell migration in response to SDF1 $\alpha$ /CXCL12, with the strongest effect observed in the ZAP70 $+$ CLL samples. Microscopic tracking of cell migration within chemokine gradients revealed that PI3K $\gamma$ functions in gradient sensing and impacts cell morphology and F-actin polarization. Similarly to idelalisib, PI3K $\gamma$ inhibition also reduced CLL adhesion to stromal cells. These findings provide evidence that PI3K $\gamma$ has unique functions in malignant B cells [38]. This paper supplements the above literature by highlighting the potential value of Ezrin and PI3K $\gamma$ as targets for rituximab-resistant DLBCL.

The candidate active ingredients in Coptis chinensis that block rituximab resistance are berberine, quercetin, epiberberine, and palmatine. There are no reports of epiberberine or palmatine combined with rituximab or DLBCL. Berberine exerts antitumour effects on DLBCL by modulating the c-myc/CD47 axis, but its effect on rituximab-resistant DLBCL is unclear [21]. CD47 enhances tumour cell viability and migration via activation of the PI3K/Akt pathway [39, 40]. Xin Li et al. reported that quercetin enhances the apoptosis of DLBCL cells through the inhibition of the STAT3 pathway [41]. In DLBCL, the activation of STAT3 and its kinase JAK1 is caused by the autocrine production of IL-6 and IL-10 in the activated B-cell-like subtype (ABC). STAT3 regulates multiple oncogenic signalling pathways, including the NF- $\kappa$ B pathway, a cell cycle checkpoint, the PI3K/AKT/mTORC1 pathway, and the STAT3 pathway itself [42]. JAK/STAT signalling regulates the cytoskeleton and cell-cell communication (EZR, GJA1) [43]. Network pharmacology is a field of research based on systems biology, genomics, proteomics and other disciplines and is a method for identifying new drug targets and molecular mechanisms. An increasing number of studies have used network pharmacology to construct and analyse drug-gene-disease networks. Jiaxin Lin et al. revealed the main candidate components, targets, and pathways involved in the action of Andrographis paniculata against lung cancer through network pharmacology and molecular docking [44]. Another study investigated the mechanism of action of Ginkgo Folium in the treatment of nonalcoholic fatty liver disease via network pharmacology and molecular docking [45]. These findings indirectly suggest the feasibility of the targets and drugs screened in this study.

Our study has several limitations. First, the small sample size (3 sensitive and 3 resistant samples) may have affected the accuracy of the results. Second, bioinformatics and network pharmacology have challenges such as inherent variability and poor reproducibility of biological data; thus, the results need further validation at the animal or cellular level. Finally, further preclinical studies and clinical trials are needed to support the use of Coptis chinensis for the treatment of rituximab-resistant DLBCL.

5. Conclusion

In this study, we revealed potential efficacy predictive biomarkers in DLBCL patients treated with rituximab. Berberine, quercetin, epiberberine, and palmatine, which target PIK3CG, may be viable options for patients with rituximab-resistant DLBCL. Although further pilot studies are needed to determine the true clinical potential of these findings, our findings could lead to a new solution for treating rituximab resistance in DLBCL or other diseases.

Ethics statement

The data used in this study were obtained from public databases. No ethics approval was required for this study.

Funding

This study was supported by the Natural Science Foundation of Fujian Province (Grant No. 2023J011276), the Fujian Provincial Health Technology Project (Grant No. 2022CXA027) and Fujian Cancer Hospital Project (Grant No. 2023YN11).

Author contributions

Qiuling Zhao and Shengqiang Huang performed this study and contributed to the R language programming, acquisition, and interpretation of the data visualizing tasks. Lin Yang created the tables and figures. Ting Chen performed the molecular docking. Xiuliang Qiu modified the formatting procedure. Ruyi Huang and Liangliang Dong drafted and edited the manuscript. Wenbin Liu reviewed the manuscript. All authors read and approved the final manuscript.

Availability of data and materials

All data can be obtained from the corresponding author upon request.

Footnotes

Acknowledgments

Not applicable.

Conflict of interest

The authors declare that they have no competing interests.

References

Poonacha

. Level of scientific evidence underlying recommendations arising from the National Comprehensive Cancer Network clinical practice guidelines. J Clin Oncol. 2011; 29(2): 186-91. doi: 10.1200/jco.2010.31.6414.

Coiffier

Sarkozy

. Diffuse large B-cell lymphoma: R-CHOP failure-what to do? Hematology Am Soc Hematol Educ Program. 2016; 2016(1): 366-78. doi: 10.1182/asheducation-2016.1.366.

Crump

Neelapu

Farooq

Van Den Neste

Kuruvilla

Westin

, et al. Outcomes in refractory diffuse large B-cell lymphoma: results from the international SCHOLAR-1 study. Blood. 2017; 130(16): 1800-8. doi: 10.1182/blood-2017-03-769620.

Nastoupil

Rose

Flowers

. Diffuse large B-cell lymphoma: current treatment approaches. Oncology (Williston Park). 2012; 26(5): 488-95.

Rezvani

Maloney

. Rituximab resistance. Best Pract Res Clin Haematol. 2011; 24(2): 203-16. doi: 10.1016/j.beha.2011.02.009.

Bello

Sotomayor

. Monoclonal antibodies for B-cell lymphomas: rituximab and beyond. Hematology Am Soc Hematol Educ Program. 2007; 233-42. doi: 10.1182/asheducation-2007.1.233.

Sehn

Chua

Mayer

Dueck

Trněný

Bouabdallah

, et al. Obinutuzumab plus bendamustine versus bendamustine monotherapy in patients with rituximab-refractory indolent non-Hodgkin lymphoma (GADOLIN): A randomised, controlled, open-label, multicentre, phase 3 trial. Lancet Oncol. 2016; 17(8): 1081-93. doi: 10.1016/s1470-2045(16)30097-3.

Terui

Mishima

Sugimura

Kojima

Sakurai

Mishima

, et al. Identification of CD20 C-terminal deletion mutations associated with loss of CD20 expression in non-Hodgkin’s lymphoma. Clin Cancer Res. 2009; 15(7): 2523-30. doi: 10.1158/1078-0432.Ccr-08-1403.

Lim

Vaughan

Ashton-Key

Williams

Dixon

Chan

, et al. Fc gamma receptor IIb on target B cells promotes rituximab internalization and reduces clinical efficacy. Blood. 2011; 118(9): 2530-40. doi: 10.1182/blood-2011-01-330357.

10.

Beum

Kennedy

Williams

Lindorfer

Taylor

. The shaving reaction: Rituximab/CD20 complexes are removed from mantle cell lymphoma and chronic lymphocytic leukemia cells by THP-1 monocytes. J Immunol. 2006; 176(4): 2600-9. doi: 10.4049/jimmunol.176.4.2600.

11.

Fishelson

Donin

Zell

Schultz

Kirschfink

. Obstacles to cancer immunotherapy: Expression of membrane complement regulatory proteins (mCRPs) in tumors. Mol Immunol. 2003; 40(2-4): 109-23. doi: 10.1016/s0161-5890(03)00112-3.

12.

Seyfizadeh

Hasenkamp

Huerta-Yepez

. A molecular perspective on rituximab: A monoclonal antibody for B cell non Hodgkin lymphoma and other affections. Crit Rev Oncol Hematol. 2016; 97: 275-90. doi: 10.1016/j.critrevonc.2015.09.001.

13.

Kusowska

Kubacz

Krawczyk

Slusarczyk

Winiarska

Bobrowicz

. Molecular Aspects of Resistance to Immunotherapies-Advances in Understanding and Management of Diffuse Large B-Cell Lymphoma. Int J Mol Sci. 2022; 23(3). doi: 10.3390/ijms23031501.

14.

Sehn

Assouline

Stewart

Mangel

Gascoyne

Fine

, et al. A phase 1 study of obinutuzumab induction followed by 2 years of maintenance in patients with relapsed CD20-positive B-cell malignancies. Blood. 2012; 119(22): 5118-25. doi: 10.1182/blood-2012-02-408773.

15.

Shimizu

Kikuchi

Wada

Ozawa

Kano

Furukawa

. HDAC inhibitors augment cytotoxic activity of rituximab by upregulating CD20 expression on lymphoma cells. Leukemia. 2010; 24(10): 1760-8. doi: 10.1038/leu.2010.157.

16.

Lin

Chen

Zhou

Fang

Luo

, et al. The role of bortezomib in newly diagnosed diffuse large B cell lymphoma: A meta-analysis. Ann Hematol. 2018; 97(11): 2137-44. doi: 10.1007/s00277-018-3435-1.

17.

Wojciechowski

Marshall

Dell’Agnola

Espinoza-Delgado

. Enhanced expression of CD20 in human tumor B cells is controlled through ERK-dependent mechanisms. J Immunol. 2005; 174(12): 7859-68. doi: 10.4049/jimmunol.174.12.7859.

18.

Bil

Winiarska

Nowis

Bojarczuk

Dabrowska-Iwanicka

Basak

, et al. Bortezomib modulates surface CD20 in B-cell malignancies and affects rituximab-mediated complement-dependent cytotoxicity. Blood. 2010; 115(18): 3745-55. doi: 10.1182/blood-2009-09-244129.

19.

Leonard

Kolibaba

Reeves

Tulpule

Flinn

Kolevska

, et al. Randomized Phase II Study of R-CHOP With or Without Bortezomib in Previously Untreated Patients With Non-Germinal Center B-Cell-Like Diffuse Large B-Cell Lymphoma. J Clin Oncol. 2017; 35(31): 3538-46. doi: 10.1200/jco.2017.73.2784.

20.

Wang

Lou

Zeng

Huang

, et al. Coptidis Rhizoma: A comprehensive review of its traditional uses, botany, phytochemistry, pharmacology and toxicology. Pharm Biol. 2019; 57(1): 193-225. doi: 10.1080/13880209.2019.1577466.

21.

Ren

Cai

Liu

Zhao

Ding

, et al. Berberine exerts anti-tumor activity in diffuse large B-cell lymphoma by modulating c-myc/CD47 axis. Biochem Pharmacol. 2021; 188: 114576. doi: 10.1016/j.bcp.2021.114576.

22.

Wang

Han

Yin

, et al. Enhancement of apoptosis of human hepatocellular carcinoma SMMC-7721 cells through synergy of berberine and evodiamine. Phytomedicine. 2008; 15(12): 1062-8. doi: 10.1016/j.phymed.2008.05.002.

23.

Youn

Cho

Kim

Lee

, et al. Berberine, a natural product, combined with cisplatin enhanced apoptosis through a mitochondria/caspase-mediated pathway in HeLa cells. Biol Pharm Bull. 2008; 31(5): 789-95. doi: 10.1248/bpb.31.789.

24.

Hong

Park

Jang

Ahn

Sym

, et al. CD99 expression and newly diagnosed diffuse large B-cell lymphoma treated with rituximab-CHOP immunochemotherapy. Ann Hematol. 2012; 91(12): 1897-906. doi: 10.1007/s00277-012-1533-z.

25.

Liu

Sun

Dong

Guo

Chi

, et al. ABCG2 and NCF4 polymorphisms are associated with clinical outcomes in diffuse large B-cell lymphoma patients treated with R-CHOP. Oncotarget. 2017; 8(35): 58292-303. doi: 10.18632/oncotarget.16869.

26.

Rushton

Arthur

Alcaide

Cheung

Jiang

Coyle

, et al. Genetic and evolutionary patterns of treatment resistance in relapsed B-cell lymphoma. Blood Adv. 2020; 4(13): 2886-98. doi: 10.1182/bloodadvances.2020001696.

27.

Cai

Geng

Chen

Rao

, et al. New risk factors and new tendency for central nervous system relapse in patients with diffuse large B-cell lymphoma: A retrospective study. Chin J Cancer. 2016; 35(1): 87. doi: 10.1186/s40880-016-0150-y.

28.

Liu

Yang

Tsai

Guo

Xue

, et al. The adhesion molecule ICAM-1 in diffuse large B-cell lymphoma post-rituximab era: Relationship with prognostic importance and rituximab resistance. Aging (Albany NY). 2020; 13(1): 181-93. doi: 10.18632/aging.202180.

29.

Diarra

Gantois

Lazrek

Verdier

Elsermans

Zephir

, et al. Fatal Enterovirus-related Myocarditis in a Patient with Devic’s Syndrome Treated with Rituximab. Card Fail Rev. 2021; 7: e09. doi: 10.15420/cfr.2020.33.

30.

Alonso

Cánovas

Rubio

. Lethal enterovirus myocarditis associated with rituximab and chemotherapy for follicular lymphoma. Med Clin (Barc). 2013; 141(10): 459-60. doi: 10.1016/j.medcli.2013.03.001.

31.

Asproudis

Kanari

Ntountas

Ragos

Goussia

Batistatou

, et al. Successful treatment with rituximab of IgG4-related disease coexisting with adult-onset asthma and periocular xanthogranuloma. Rheumatol Int. 2020; 40(4): 671-7. doi: 10.1007/s00296-019-04409-2.

32.

Casal Moura

Berti

Keogh

Volcheck

Specks

Baqir

. Asthma control in eosinophilic granulomatosis with polyangiitis treated with rituximab. Clin Rheumatol. 2020; 39(5): 1581-90. doi: 10.1007/s10067-019-04891-w.

33.

Parajuli

Mandelbrot

Muth

Mohamed

Garg

Aziz

, et al. Rituximab and monitoring strategies for late antibody-mediated rejection after kidney transplantation. Transplant Direct. 2017; 3(12): e227. doi: 10.1097/txd.0000000000000746.

34.

Kawaguchi

Yoshida

Hatano

Asano

. Pathophysiological roles of ezrin/radixin/moesin proteins. Biol Pharm Bull. 2017; 40(4): 381-90. doi: 10.1248/bpb.b16-01011.

35.

Wakayama

Nakata

Kurobo

Sai

Iseki

. Expression, localization, and binding activity of the ezrin/radixin/moesin proteins in the mouse testis. J Histochem Cytochem. 2009; 57(4): 351-62. doi: 10.1369/jhc.2008.952440.

36.

Bruce

Khanna

Ren

Landberg

Jirström

Powell

, et al. Expression of the cytoskeleton linker protein ezrin in human cancers. Clin Exp Metastasis. 2007; 24(2): 69-78. doi: 10.1007/s10585-006-9050-x.

37.

Vanhaesebroeck

Guillermet-Guibert

Graupera

Bilanges

. The emerging mechanisms of isoform-specific PI3K signalling. Nat Rev Mol Cell Biol. 2010; 11(5): 329-41. doi: 10.1038/nrm2882.

38.

Ali

Eissa

Hou

Ghia

Murooka

, et al. Distinct roles for phosphoinositide 3-kinases γ and δ in malignant B cell migration. Leukemia. 2018; 32(9): 1958-69. doi: 10.1038/s41375-018-0012-5.

39.

Liu

Wang

Chen

Yang

, et al. CD47 Promotes Human Glioblastoma Invasion Through Activation of the PI3K/Akt Pathway. Oncol Res. 2019; 27(4): 415-22. doi: 10.3727/096504018x15155538502359.

40.

Liu

Chang

Cai

Jiang

Jia

, et al. CD47 Enhances Cell Viability and Migration Ability but Inhibits Apoptosis in Endometrial Carcinoma Cells via the PI3K/Akt/mTOR Signaling Pathway. Front Oncol. 2020; 10: 1525. doi: 10.3389/fonc.2020.01525.

41.

Wang

Zhang

Sun

. Quercetin potentiates the antitumor activity of rituximab in diffuse large B-cell lymphoma by inhibiting STAT3 pathway. Cell Biochem Biophys. 2014; 70(2): 1357-62. doi: 10.1007/s12013-014-0064-8.

42.

Zhu

Zhang

Drennan

Kimpara

, et al. Gene regulation and suppression of type I interferon signaling by STAT3 in diffuse large B cell lymphoma. Proc Natl Acad Sci U S A. 2018; 115(3): E498-e505. doi: 10.1073/pnas.1715118115.

43.

Leventoux

Augustus

Azar

Riquier

Villemin

Guelfi

, et al. Transformation Foci in IDH1-mutated Gliomas Show STAT3 Phosphorylation and Downregulate the Metabolic Enzyme ETNPPL, a Negative Regulator of Glioma Growth. Sci Rep. 2020; 10(1): 5504. doi: 10.1038/s41598-020-62145-1.

44.

Liu

Han

. Mechanism of andrographis paniculata on lung cancer by network pharmacology and molecular docking. Technol Health Care. 2023; 31(4): 1407-27. doi: 10.3233/thc-220698.

45.

Sun

Wang

Tian

Jin

, et al. Investigation on the mechanism of Ginkgo Folium in the treatment of Non-alcoholic Fatty Liver Disease by strategy of network pharmacology and molecular docking. Technol Health Care. 2023; 31(S1): 209-21. doi: 10.3233/thc-236018.

Biomarkers and coptis chinensis activity for rituximab-resistant diffuse large B-cell lymphoma: Combination of bioinformatics analysis,network pharmacology and molecular docking

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2.1 Bioinformatics analysis

2.1.1 Identification of DEGs

2.1.2 Coexpression network construction for WGCNA

2.1.3 Acquisition of core targets

2.1.4 GSEA

2.1.5 Enrichment analyses and network visualization

2.1.6 Establishment of a protein-protein interaction (PPI) network

2.2 Network pharmacology

2.2.1 Selection of monomers in Coptis chinensis

2.2.2 Acquisition of pharmacological targets of Coptis chinensis

2.2.3 Identifying targets of Coptis chinensis against rituximab-resistant DLBCL

2.2.4 Molecular docking analysis

3.1 Identification of the core targets of rituximab-resistant DLBCL

3.2 GSEA

3.3 Functional enrichment and PPI network of the core targets

3.4 Screening of active ingredients and related pharmacological target genes

Table 1 Main compounds of Coptis chinensis

Table 2 Molecular docking information on active ingredient-target of Coptis chinensis against rituximab-resistant DLBCL

5. Conclusion

Ethics statement

Funding

Author contributions

Availability of data and materials

Footnotes

Acknowledgments

Conflict of interest

References

Table 1
Main compounds of Coptis chinensis

Table 2
Molecular docking information on active ingredient-target of Coptis chinensis against rituximab-resistant DLBCL