Identification of inflammation-related genes and exploration of regulatory mechanisms in patients with osteonecrosis of the femoral head

doi:10.21203/rs.3.rs-1449489/v1

Download PDF

Research Article

Identification of inflammation-related genes and exploration of regulatory mechanisms in patients with osteonecrosis of the femoral head

https://doi.org/10.21203/rs.3.rs-1449489/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background:

Osteonecrosis of the femoral head (ONFH) is a systemic and disabling orthopedic disease. If left untreated, it can lead to hip joint damage and dysfunction, ultimately affecting quality of life. Previous studies have suggested that infiltration of immune cells significantly impacts the occurrence and development of ONFH. In this study, GSE123568 dataset with control and steroid-induced osteonecrosis of the femoral head (SONFH) samples were downloaded from Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) were detected by limma R package and weighted gene co-expression network analysis (WGCNA) was used to explore the co-expression genes and co-expression modules. We obtained inflammation-related genes (IRGs) from the Molecular Signatures Database (MSigDB). Then, the IRGs associated with SONFH (IRGs-SONFH) were screened out and analyzed by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis. A protein-protein interaction (PPI) network was established using the Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database, and hub genes were identified by the MCODE algorithms. Based on the hub genes, we constructed a lncRNA-miRNA-mRNA network.

Results:

We identified 535 DEGs between control and SONFH samples. The WGCNA clearly indicated that the brown module was most significantly associated with SONFH. We identified 25 IRGs-SONFH through WGCNA module genes, DEGs and IRGs. GO analysis mainly included response to molecule of bacterial origin, neutrophil activation, membrane microdomain, membrane raft, amide binding, and peptide binding. KEGG pathways mostly included lipid and atherosclerosis, tuberculosis, and neutrophil extracellular trap formation. A total of 4 hub genes (CD14, CYBB, NOD2, and TLR1) were identified by cytoscape. Receiver operating characteristic (ROC) curve analysis determined that the expressions of the four genes could distinguish SONFH from controls as evidenced by the area under the curve (AUC) all greater than 0.7. Finally, we constructed a competitive endogenous RNA (ceRNA) network which included 67 lncRNAs, 1 miRNA (hsa-miR-320a), and 1 mRNA (NOD2).

Conclusions:

Our study identified 4 hub genes as potential biomarkers for early diagnosis of SONFH. Moreover, we proposed a ceRNA network of lncRNAs targeting hsa-miR-320a, hsa-miR-320a, and NOD2 as a potential RNA regulatory pathway that control disease progression in ONFH.

Osteonecrosis of the femoral head

hub gene

weighted gene co-expression network analysis

inflammation-related gene

ceRNA network

Osteonecrosis of the femoral head (ONFH) is characterized by the death of bone due to the circulatory disruption of femoral head with traumatic or nontraumatic factors[1]. As a large range of diseases requires steroid usage, steroid-induced osteonecrosis of the femoral head (SONFH) accounts for a large proportion of ONFH[2]. There are approximately 8.12 million nontraumatic ONFH cases in the population aged 15 years and over in China, among which SONFH counted for 47.4%[3]. Association Research Circulation Osseous (ARCO) classification system[4] was developed for accurate staging, which is essential of successful treatment. As early symptom of ONFH is not obvious, patients are often diagnosed at the advanced stage (ARCO stage III-IV). Therefore, novel diagnostic biomarkers and therapeutic targets are urgently needed to provide for early intervention and treatment of SONFH patients.

In previous studies, inflammation has been shown to play an indispensable role in the development and progression of SONFH. Li et al. identified 8 candidate serum biomarkers of SONFH and found they were significantly involved in the immune regulation and inflammation[5]. Real-time imaging probes showed the accumulation of elevated neutrophils and macrophages at the tissue of osteonecrosis in mouse model[6]. Necrotic bone stimulated the macrophage-inflammatory responses through activation of a pattern recognition receptor (PRR) Toll-like Receptor 4 (TLR4) and upregulation of the downstream transcription factor, including Nuclear Factor-Kappa B (NF-κB) and Monocyte Chemotactic Protein 1 (MCP-1), for inflammatory proteins[7, 8]. In other studies, the levels of pro-inflammatory cytokines interleukin (IL)-9, IL-17, IL-23, and IL-33 produced by T cells have been reported to be associated with ONFH[9–12]. In addition, the activation of specific B cells and elevated levels of serum Tumor Necrosis Factor α (TNF-α) were associated with the development of ONFH[13]. Moreover, neutrophil granulocyte levels and percentage of neutrophil granulocytes were related to ONFH[14].

Bioinformatics has been used to identify hub genes, interaction networks, and pathways of SONFH to improve diagnosis and treatment. Early studies identified biomarkers of SONFH based on weighted gene co-expression network analysis (WCGNA) and differentially expressed genes (DEGs) screening and further revealed their correlation with immune infiltration[15, 16]. Competitive endogenous RNA (ceRNA) networks can reveal potential mechanisms of disease in transcriptional regulatory networks, while no study has been conducted to analyze it in osteonecrosis.

In the present study, we proposed to identified novel inflammation-related biomarkers and evaluate its diagnostic value in SONFH on the basis of GSE123568 dataset. The DEGs screening, WCGNA, and inflammation-related gene (IRG) expression were used to identified the gene-network signature and IRGs associated with SONFH (IRGs-SONFH). Then, the IRGs-SONFH were analyzed by Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis and the protein-protein interaction (PPI) was constructed. Subsequently, 4 hub genes were screened out and verified by quantitative Reverse Transcription Polymerase Chain Reaction (qRT-PCR). Finally, we constructed a ceRNA network to get deep understanding of the pathogenesis of ONFH on the basis of predicted results of microRNAs (miRNAs) and long non-coding RNAs (lncRNAs). The research process of this study is showed in Fig. 1.

Data source

Microarray RNA expression dataset of GSE123568 was downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/geo/). The GSE123568 dataset was generated from samples of peripheral serum in 30 SONFH patients and 10 controls. The dataset was based on the platform GPL15207([PrimeView] Affymetrix Human Gene Expression Array). The array data for GSE89587 included the miRNA expression profiles of 10 ONFH patients and 10 controls to construct a ceRNA network. We obtained 200 IRGs from the Molecular Signature Database (MSigDB) (https://www.broadinstitute.org/msigdb)[17].

Identification of DEGs

The limma, a package in the R language, was used to identify DEGs with the cut-off point of adjust p<0.05 and |Log₂fold change|>1. Heatmap and volcano plot of DEGs from the databases were constructed with Pheatmap and ggplot2 R packages.

Construction of weighted gene co-expression networks

The R package WGCNA[18] was used to analyze the gene co-expression network of the GSE123568 dataset. First, the samples were clustered and the outliers were removed. Second, to construct a scale-free network, the soft threshold of β=18 was chosen with the function pickSoftThreshold. Based on the selected soft threshold, the adjacency matrix was converted to topological overlap matrix for constructing the network, and the gene dendrogram and module color were established by utilizing the degree of dissimilarity. Then, the correlations between modules and SONFH were calculated using the WGCNA package. Therefore, modules with high correlation coefficient were considered candidates relevant to SONFH, and were selected for subsequent analysis. The intersection of DEGs, genes in key modules, and IRGs were carried out using the “VennDiagram” R package and were defined as IRGs-SONFH, which were used for subsequent analysis.

KEGG and GO Enrichment Analysis

Functional annotation of IRGs-SONFH was presented with the R package “clusterProfiler”[19], containing GO and KEGG pathway analysis. GO terms were comprised of the biological process (BP), cellular component (CC), and molecular function (MF)[20] and were used to identify the biological properties of genes and gene sets in all organisms. The KEGG enrichment analysis was performed to obtain the associated enrichment pathways. Adjust p<0.05 was considered statistically significant.

Construction of PPI network

The Search Tool for the Retrieval of Interacting Genes/Proteins (STRING) database (https://cn.string-db.org/cgi/input?sessionId=bbRkt711IuEL&input_page_active_form=multiple_identifiers)[21] was used to construct a PPI network of the above genes. Next, we downloaded the interaction information and optimized the PPI network with Cytoscape software (https://www.cytoscape.org)[22] for better visualization. The MCODE plug-in in Cytoscape was used to identify significant gene clusters and obtain hub genes. The hub genes were evaluated using the geometric mean of semantic similarities in CCs and MFs by applying the “GOSemSim” package in R software[23]. Corrplot package was used to analyze the correlation of hub genes.

qRT-PCR

The serum samples of 24 patients with femoral neck fracture and 24 patients with SONFH were acquired for qRT-PCR verification in order to verify the predictive analysis results. Total RNA was extracted from the serum samples using TRIzol, and then, its concentration and purity were assessed by nanodrop. RNA samples from total RNA were reverse-transcribed to cDNA, and qRT-PCR was performed using the SuperScript III RT (ABI-Invitrogen, 11752050). b-Actin was used as an internal normalization standard. The 2^−ΔΔCt method was utilized to determine the relative expression of each selected gene between SONFH and controls. Sequences of primers used in the study are showed in Table 1.

The receiver operating characteristic (ROC) curve analysis and expression analysis

In the GSE123568 dataset, 30 SONFH samples and 10 control samples were utilized to plot ROC curves, from which we obtained their area under the ROC curves (AUC) through the “pROC” package. ROC curve is a helpful tool to evaluate the efficiency of gene diagnosis. The hub genes with AUC > 0.7 were deemed useful for disease diagnosis.

Expression levels of hub genes between SONFH and normal samples were shown using boxplots. The boxplots of hub genes expression were drawn using the “ggplot2” in R package.

Correlation analysis between hub genes and infiltrated immune cells

Immune infiltration analysis was performed by using the ssGSEA algorithm. Correlation analysis between crucial genes and 28 immune cells was calculated via the spearman method, and the results were visualized. Correlation analysis was conducted to find out the relationship between hub genes and differential immune cells.

Small-molecule drug predication and gene set enrich analysis (GSEA)

The latent drugs for hub genes were predicted through the Drug-Gene Interaction Database (DGIdb) (https://dgidb.org/search_interactions)[24]. We used the R package “clusterprofiler” to conduct GSEA on hub genes. The chosen reference gene set was downloaded from the MSigDB. A p<0.05 was considered statistically significant.

Prediction of networks mutually regulated by miRNAs and transcription factors (TFs)

The upstream TFs and the miRNAs were predicted using the miRNet database (https://www.mirnet.ca)[25]. Subsequently, the results were visualized using Cytoscape software.

CeRNA Network Construction

To predict the regulatory relationship among mRNAs, lncRNAs, and miRNAs, lncRNAs were predicted using miRNet. Briefly, differentially expressed miRNAs (DEmiRNAs) in GSE89587 with the threshold criterion of adjusted P-value < 0.05 were screened using the limma package of the R software program. Then, they were intersected with the miRNAs predicted by the hub genes above to get the final target miRNAs. Target lncRNAs matched by target miRNAs were retrieved using miRNet database. The ceRNA regulatory network of lncRNA-miRNA-mRNA was visualized using Cytoscape software.

Identification of DEGs

In our study, 535 DEGs were identified between SONFH samples and control samples. Among them, 299 were upregulated and 236 were downregulated (SONFH vs. control). The volcano plot and heatmap of gene expression are shown in Figures 2A and 2B.

Construction of co-expression networks

The sample clustering tree indicated that there was no abnormal sample (Figure 3A-B). After calculation, the best soft-thresholding power was set at 18 (Figure 3C). Finally, each module was assigned a color, and a total of 9 modules in GSE123568 (Figure 3D) were identified in this study. Furthermore, the result of the module-feature relationship revealed that the brown module had the highest correlations with SONFH (cor = 0.68, p = 1e−05, Figure 3E). Thus, 850 genes in the brown module were selected for further exploration.

Identification of IRGs-SONFH and functional enrichment analysis

Then, we took the intersection of DEGs, genes in key modules, and IRGs and identified 25 IRGs-SONFH (Figure 4A). To explore the function of 25 IRGs-SONFH in SONFH, the GO terms are shown in Figure 3. In BP analysis (Figure 4B), IRGs-SONFH mainly participated in response to molecule of bacterial origin, neutrophil activation, response to lipopolysaccharide, cellular response to biotic stimulus, and cellular response to molecule of bacterial origin. In CC analysis (Figure 4C), IRGs-SONFH significantly participated in the membrane microdomain, membrane raft, secretory granule membrane, endocytic vesicle, and phagocytic vesicle. MF analysis showed that IRGs-SONFH significantly participated in amide binding, peptide binding, immune receptor activity, pattern recognition receptor activity, and lipopolysaccharide binding (Figure 4D). KEGG analysis was performed to explore the pathways of these 25 IRGs-SONFH. The KEGG terms of IRGs-SONFH are shown in Figure 4E. As shown, these IRGs-SONFH were mainly enriched in lipid and atherosclerosis, tuberculosis, neutrophil extracellular trap formation, TLR signaling pathway, and legionellosis.

Identification of hub genes

The PPI network between IRGs-SONFH was established using the STRING database, interactions of 25 IRGs-SONFH were displayed in Figure 5A. 4 hub genes (CD14, CYBB, NOD2, and TLR1) were identified by MCODE plug-in Cytoscape (Figure 5B).

The correlation analysis between hub genes and the functional similarity analysis of hub genes

The correlation between these 4 hub genes was investigated using the corrplot package, CD14 and TLR1 had the strongest correlation (r = 0.85) (Figure 5C). We analyzed the functional similarity of these hub genes by the “GOSemSim” package in R. The results showed that 3 hub genes, including CD14, NOD2, and TLR1 (similarity score>0.5), had higher functional similarity (Figure 5D).

Validation and efficacy evaluation of hub genes

We explored the expressions of these genes between SONFH and control samples in GSE123568 and found that those genes exhibited higher expression levels in SONFH (Figure 6A). In addition, the relative expressions of the above four hub genes were investigated by qRT-PCR. As shown in Figure 6C-F, the relative expressions of CD14, CYBB, NOD2, and TLR1 were also significantly increased in peripheral blood of SONFH patients compared to controls. Furthermore, we executed a ROC curve analysis to calculate their sensitivity and specificity for the diagnosis of SONFH (Figure 6B). The AUC values of CD14, CYBB, NOD2, and TLR1 were 0.847, 0.753, 0.767, 0.847, respectively, demonstrating that these genes have high sensitivity and specificity for SONFH diagnosis.

Correlation analysis of hub genes and immune cells

To further understand the role of these genes in immune infiltration, we used spearman correlation analysis to determine whether these hub genes were related to immune cell infiltration. Correlation analysis showed that 4 hub genes including CD14, CYBB, NOD2, and TLR1 had significantly positive relationship with Type 1 T helper cell, T follicular helper cell, regulatory T cell, plasmacytoid dendritic cell, neutrophil, natural killer T cell, natural killer cell, monocyte, memory B cell, myeloid-derived suppressor cell, mast cell, macrophage, immature dendritic cell, immature B cell, gamma delta T cell, eosinophil, effector memory CD8 T cell, effector memory CD4 T cell, central memory CD4 T cell, central memory CD8 T cell, activated dendritic cell (Figure 7).

GSEA

The function of our hub genes was explored via GSEA. Genes in the high-expression cohorts of CD14, and TLR1were all highly enriched in leishmania infection, Toll-like receptor signaling pathways and Fc gamma R-mediated phagocytosis (Figure 8A, 8D). Genes in the high-expression cohorts of CYBB, and NOD2 were all highly enriched in spliceosome, lysosome and B-cell receptor signaling pathways (Figure 8B, 8C). Genes in the low-expression cohorts of CD14, CYBB, NOD2, and TLR1 were all highly enriched in olfactory transduction, linoleic acid metabolism, and basal cell carcinoma (Figure 8). After considering the results of GSEA, we concluded that these four genes might be highly correlated with immune and inflammation.

Drug-gene networks

A total of 17 potential drugs for treating SONFH patients were identified when the drug-gene interactions were explored using DGIdb (Table 2). Additionally, drug-gene networks were constructed by Cytoscape (Figure 9A). However, we did not find any small molecule drugs that could target TLR1 in this database.

Prediction of key miRNAs and TF

The miRNA and TFs regulatory network of the 4 hub genes was constructed using miRNet. As illustrated in Figure 9B, the interaction network consisted of 4 hub genes and 59 miRNAs. Specifically, 9 miRNAs (ie, hsa-mir-335-5p, hsa-mir-100-5p, hsa-mir-3687) targeting CD14, 28 miRNAs (ie, hsa-mir-6826-3p, hsa-mir-6845-3p, hsa-mir-6859-3p) targeting CYBB, 12 miRNAs (ie, hsa-mir-215-5p, hsa-mir-122-5p, hsa-mir-320a) targeting NOD2, 15 miRNAs (ie, hsa-mir-34a-5p, hsa-mir-3662, hsa-mir-4511) targeting TLR1. The interaction network consisted of 4 hub genes and 30 TFs. We found that 12 TFs (ie, CEBPB, FOS, JUN) could regulate CD14. 8 TFs (ie, NFIC, NFYA, YY1) could regulate CYBB. 11 TFs (ie, MAX, USF1, USF2) could regulate NOD2. 8 TFs (ie, MEF2A, HINFP, TP63) could regulate TLR1.

CeRNA regulatory network construction

To elucidate the potential molecular mechanism of lncRNAs in SONFH, we constructed a lncRNA-miRNA-mRNA interaction network. Briefly, 20 DEmiRNAs with the threshold criterion of adjusted p<0.05 were screened by GSE89587. 1 miRNA (hsa-miR-320a) was obtained by taking the intersection of 64 miRNAs predicted by the hub genes above and 20 DEmiRNAs. We used the database miRNet to predict the lncRNAs that interacted with the selected miRNAs (hsa-miR-320a). Finally, we obtained a ceRNA network which included 67 lncRNAs, 1 miRNA (hsa-miR-320a), and 1 mRNA (NOD2) (Figure 9C).

ONFH is a progressive disease with necrosis of osteocyte and bone marrow as a result of intramedullary microvascular lesions and interruption of blood supply of the femoral head. In patients with end-stage ONFH, collapse of femoral head can lead to dysfunction of hip joint, ultimately affecting the quality of life. Hence, novel biomarkers for early diagnosis and individualized treatment are urgently needed. Multiple biological processes, including circulation, steroid metabolism, immunity, and bone formation, were involved in the development of ONFH[26]. Although existing theories have pointed to the roles of inflammation in the pathogenesis of SONFH[15], the molecular mechanism contributing to disease onset remains unclear. In addition, few studies have systematically screened the biomarkers related to inflammation and their value for assessing the process of SONFH.

In this study, a total of 535 DEGs were identified in the SONFH and control samples. In addition, 9 co-expression modules were obtained by WGCNA analysis. Among them, the brown module with a total of 850 genes was the most relevant to SONFH. Moreover, 200 IRGs were identified. We obtained 25 candidate genes for hub genes by taking the intersection of the above three gene lists. Bioinformatics databases, including GO and KEGG, are widely used in gene classification and signaling pathway analysis. As demonstrated form GO enrichment results, the candidate genes showed a major relationship to response to molecule of neutrophil activation. In addition, KEGG pathways were enriched in neutrophil extracellular trap formation, and TLR signaling pathway. Although SONFH is not an inflammatory disease, the enrichment results suggest that inflammation may play a critical role in its pathophysiological mechanism.

Through PPI network, we identified 4 hub genes, namely CD14, CYBB, NOD2, and TLR1. As shown by the ROC curve, they had high sensitivity and specificity for SONFH and could be used as biomarkers. Furthermore, qRT-PCR showed that the relative expression of CD14, CYBB, NOD2, and TLR1 in peripheral blood sample of SONFH was increased compared with normal groups. Finally, we constructed a ceRNA network to clarify the pathogenesis of ONFH form the transcriptomic level.

The protein encoded by CD14 is a surface antigen which is preferentially expressed on monocytes and macrophages. Generally, macrophages are known as a critical role in innate immune response and they can polarize into pro-inflammatory (M1) or anti-inflammatory (M2) phenotypes depending on microenvironment[27]. The repolarization of macrophages from M1 phenotype to M2 phenotype could promote survival of osteocytes and decrease inflammatory cytokines, which was effective in alleviation of SONFH[28]. Early study has shown that CD14+ macrophages increased in the fibrovascular repair tissue during the induction of ONFH[7]. Consistent with the present study, we found that CD14 expression was up-regulated in SONFH patients and the ROC curve showed that CD14 had high diagnostic value (AUC = 0.847).

CYBB is a transmembrane protein of microbicidal oxidase system of phagocytes[29]. CYBB deficiency can lead to the disorder of reactive oxygen species (ROS) production, resulting in the disability of phagocytes to kill most pathogens, which is associated with the rare immune deficiency disorder, chronic granulomatous disease[30]. However, uncontrolled neutrophil ROS production can lead to persistent vascular inflammation reactions contributing to some inflammatory disease[31]. Moreover, the impaired blood vessels caused by dysregulation of bone endothelial cells is one of the most convincing mechanisms of SONFH[32]. Therefore, the vascular inflammation caused by high expression of CYBB may be related to SONFH. In current study, we found that CYBB expression was up-regulated in SONFH patients and it may be a diagnostic biomarker for SONFH (AUC = 0.753).

NOD2 is one of PRRs of the NOD-like receptor (NLR) family that sense conserved motifs in bacterial peptidoglycan and activate intracellular signaling pathways that drive pro-inflammatory and anti-microbial responses[33]. NOD2 regulates multiple pathways involved in a variety of inflammatory responses via the activation of NF-κB, MAPK, and type I interferons (IFN)[34, 35]. NOD2 also interplays with TLRs during systemic bacterial infection to enhance immune response and promote immune responses after toleration by TLR ligands[36]. Furthermore, the activation of TLR4/NF-κB pathway results in the gene expression of molecules responded for inflammatory cytokine responses in macrophages, which may contribute to SONFH[37, 38]. Therefore, combined with the results of the ROC curve (AUC = 0.767), we hypothesize that NOD2 may be an effective biomarker for diagnosis of SONFH.

TLR1 is a member of the TLR family which is responsible for the recognition of pathogen-associated molecular patterns (PAMPs) and induction of inflammatory immune response[39]. The formation of TLR1-TLR2 heterodimer brings the intracellular Toll/interleukin 1 receptor (TIR) domains into close proximity and initiates signaling[40]. MyD88, an intracellular TIR-containing adaptor used by TLR1, interacts with interleukin-1 receptor-associated kinases (IRAKs) and eventually leads to the activation of NF-κB and IFN-regulatory factors (IRFs) to elicit anti-pathogen responses and inflammation[41]. Our study showed that the expression of TLR1 increased in SONFH patients and TLR1 had high diagnostic value according to the ROC curve (AUC = 0.847). Therefore, we believe that TLR1 is a new and effective biomarker for the diagnosis of SONFH.

Furthermore, miRNAs and TFs targeting CD14, CYBB, NOD2, and TLR1 were predicted and a total of 20 differentially expressed miRNAs were identified in patients with osteonecrosis and controls. Among the 20 differentially expressed miRNAs, hsa-miR-320a was found as a regulatory miRNA of NOD2. Consequently, lncRNAs targeting hsa-miR-320a were searched from database and a ceRNA network of 67 lncRNAs targeting hsa-miR-320a, hsa-miR-320a, and NOD2 was constructed. In early bioinformatics researches, hsa-miR-320a has been identified as diagnostic biomarkers of atherogenesis[42], multiple sclerosis[43], gastric cancer[44], and metabolic syndrome[45]. Our study indicated that hsa-miR-320a was a key regulator of NOD2 associated with inflammation contributing to the progression of osteonecrosis.

This study had several limitations. The sample size for analysis and validation was relatively small. Moreover, most SONFH cases have other comorbidities which have been treated with glucocorticoids, and the different primary diseases may influence our results. Therefore, future studies need to increase the sample size and control the effects of primary disease to further confirm our results.

This study identified and validated 4 hub genes, CD14, CYBB, NOD2, and TLR1, as potential biomarkers for early diagnosis of SONFH, and provided clues to the mechanism of disease development of SONFH at the transcriptome level. Moreover, we proposed a ceRNA network of lncRNAs targeting hsa-miR-320a, hsa-miR-320a, and NOD2 as a potential RNA regulatory pathway that control disease progression in ONFH.

ONFH: Osteonecrosis of the Femoral Head;

SONFH: steroid-induced osteonecrosis of the femoral head;

ARCO: Association Research Circulation Osseous;

PRR: Pattern Recognition Receptor;

TLR4: Toll-like Receptor 4;

NF-kB: Nuclear Factor-Kappa B;

MCP-1: Monocyte Chemotactic Protein 1;

IL: Interleukin;

TNF-a: Tumor Necrosis Factor a;

WCGNA: Weighted Gene Co-expression Network Analysis;

DEG: Differentially Expressed Gene;

ceRNA: Competitive endogenous RNA;

IRG: Inflammation-Related Gene;

GO: Gene Ontology;

KEGG: Kyoto Encyclopedia of Genes and Genomes;

qRT-PCR: quantitative Reverse Transcription Polymerase Chain Reaction;

miRNA: microRNA;

lncRNA: long non-coding RNA;

GEO: Gene Expression Omnibus;

MSigDB: Molecular Signature Database;

BP: Biological Process;

CC: Cellular Component;

MF: Molecular Function;

PPI: Protein-Protein Interaction;

STRING: Search Tool for the Retrieval of Interacting Genes/Proteins;

ROC: Receiver Operating Characteristic;

AUC: Area Under Curve;

DGIdb: Drug-Gene Interaction database;

GSEA: Gene Set Enrich Analysis;

TF: Transcription Factor;

DEmiRNA: Differentially Expressed miRNA;

NLR: NOD-Like Receptor;

IFN: Interferons;

ROS: Reactive Oxygen Species;

PAMP: Pathogen-Associated Molecular Pattern;

TIR: Toll/Interleukin 1 Receptor;

IRAK: Interleukin-1 Receptor-Associated Kinase;

IRF: IFN-Regulatory Factor.

Ethics approval and consent to participate

This study was performed in accordance with the ethical standards of the Declaration of Helsinki (1964) and its subsequent amendments. This study was approved by ethical committees of China-Japan Friendship Hospital. All participants provided written informed consent before participating.

Consent for publication

Not applicable.

Availability of data and materials

The datasets generated and analyzed during the current study are available in the GEO, MSigDB, STRING, DGIdb, and miRNet databases. GEO (https://www.ncbi.nlm.nih.gov/geo/); MSigDB (https://www.broadinstitute.org/msigdb); STRING (https://cn.string-db.org/cgi/input?sessionId=bbRkt711IuEL&input_page_active_form=multiple_identifiers); DGIdb (https://dgidb.org/search_interactions); miRNet (https://www.mirnet.ca).

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the Elite Medical Professionals project of China-Japan Friendship Hospital (NO. ZRJY2021-QM08), Natural Science Foundation of Beijing, China (grant number 7202183), and National Natural Science Foundation of China (grant number 81972107).

Authors' contributions

TL, CH and WGW were involved in the concept and design of the study. TL, JHM, RD, and QDZ participated in the collection and analysis of data. TL and CH carried out interpretation of data and drafted the manuscript. All authors read and approved the final manuscript.

Acknowledgements

Not applicable.

Authors' information

Tong Li

Cheng Huang

Jinhui Ma

Ran Ding

Qidong Zhang

Weiguo Wang

Tong Li and Cheng Huang contributed equally for this work.

Institutional address for all authors: Department of Orthopedics, China-Japan friendship Hospital, No. 2, Yinghua East Street, Chaoyang District, Beijing, China.

Mont MA, Salem HS, Piuzzi NS, Goodman SB, Jones LC. Nontraumatic Osteonecrosis of the Femoral Head: Where Do We Stand Today?: A 5-Year Update. J Bone Joint Surg Am. 2020;102(12):1084–99.
Cui L, Zhuang Q, Lin J, Jin J, Zhang K, Cao L. et al. Multicentric epidemiologic study on six thousand three hundred and ninety five cases of femoral head osteonecrosis in China. Int Orthop. 2016;40(2):267–76.
Zhao DW, Yu M, Hu K, Wang W, Yang L, Wang BJ. et al. Prevalence of Nontraumatic Osteonecrosis of the Femoral Head and its Associated Risk Factors in the Chinese Population: Results from a Nationally Representative Survey. Chin Med J (Engl). 2015;128(21):2843–50.
Yoon BH, Mont MA, Koo KH, Chen CH, Cheng EY, Cui Q. et al. The 2019 Revised Version of Association Research Circulation Osseous Staging System of Osteonecrosis of the Femoral Head. J Arthroplasty. 2020;35(4):933–40.
Li T, Zhang Y, Wang R, Xue Z, Li S, Cao Y. et al. Discovery and validation an eight-biomarker serum gene signature for the diagnosis of steroid-induced osteonecrosis of the femoral head. Bone. 2019;122:199–208.
Phipps MC, Huang Y, Yamaguchi R, Kamiya N, Adapala NS, Tang L. et al. In vivo monitoring of activated macrophages and neutrophils in response to ischemic osteonecrosis in a mouse model. J Orthop Res. 2016;34(2):307–13.
Adapala NS, Yamaguchi R, Phipps M, Aruwajoye O, Kim HKW. Necrotic Bone Stimulates Proinflammatory Responses in Macrophages through the Activation of Toll-Like Receptor 4. Am J Pathol. 2016;186(11):2987–99.
Tian L, Wen Q, Dang X, You W, Fan L, Wang K. Immune response associated with Toll-like receptor 4 signaling pathway leads to steroid-induced femoral head osteonecrosis. BMC Musculoskelet Disord. 2014;15:18.
Geng W, Zhang W, Ma J. IL-9 exhibits elevated expression in osteonecrosis of femoral head patients and promotes cartilage degradation through activation of JAK-STAT signaling in vitro. Int Immunopharmacol. 2018;60:228–34.
Zou D, Zhang K, Yang Y, Ren Y, Zhang L, Xiao X. et al. Th17 and IL-17 exhibit higher levels in osteonecrosis of the femoral head and have a positive correlation with severity of pain. Endokrynol Pol. 2018;69(3):283–90.
Kim TH, Hong JM, Oh B, Cho YS, Lee JY, Kim HL. et al. Association of polymorphisms in the Interleukin 23 receptor gene with osteonecrosis of femoral head in Korean population. Exp Mol Med. 2008;40(4):418–26.
Zheng L, Wang W, Ni J, Li Z, Xiao T, Zhang Q. et al. Plasma interleukin 33 level in patients with osteonecrosis of femoral head: an alarmin for osteonecrosis of the femoral head? J Investig Med. 2014;62(3):635–7.
Zhang H, Xiao F, Liu Y, Zhao D, Shan Y, Jiang Y. A higher frequency of peripheral blood activated B cells in patients with non-traumatic osteonecrosis of the femoral head. Int Immunopharmacol. 2014;20(1):95–100.
Jiang J, Liu X, Lai B, Hu D, Lai L, Xu J. et al. Correlational analysis between neutrophil granulocyte levels and osteonecrosis of the femoral head. BMC Musculoskelet Disord. 2019;20(1):393.
Wu Z, Wen Y, Fan G, He H, Zhou S, Chen L. HEMGN and SLC2A1 might be potential diagnostic biomarkers of steroid-induced osteonecrosis of femoral head: study based on WGCNA and DEGs screening. BMC Musculoskelet Disord. 2021;22(1):85.
Yu R, Zhang J, Zhuo Y, Hong X, Ye J, Tang S. et al. ARG2, MAP4K5 and TSTA3 as Diagnostic Markers of Steroid-Induced Osteonecrosis of the Femoral Head and Their Correlation With Immune Infiltration. Front Genet. 2021;12:691465.
Liberzon A, Birger C, Thorvaldsdottir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;1(6):417–25.
Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.
Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS. 2012;16(5):284–7.
The Gene Ontology C. Expansion of the Gene Ontology knowledgebase and resources. Nucleic Acids Res. 2017;45(D1):D331-D8.
Szklarczyk D, Franceschini A, Wyder S, Forslund K, Heller D, Huerta-Cepas J. et al. STRING v10: protein-protein interaction networks, integrated over the tree of life. Nucleic Acids Res. 2015;43(Database issue):D447-52.
Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D. et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504.
Yu G, Li F, Qin Y, Bo X, Wu Y, Wang S. GOSemSim: an R package for measuring semantic similarity among GO terms and gene products. Bioinformatics. 2010;26(7):976–8.
Cotto KC, Wagner AH, Feng YY, Kiwala S, Coffman AC, Spies G. et al. DGIdb 3.0: a redesign and expansion of the drug-gene interaction database. Nucleic Acids Res. 2018;46(D1):D1068-D73.
Fan Y, Xia J. miRNet-Functional Analysis and Visual Exploration of miRNA-Target Interactions in a Network Context. Methods Mol Biol. 2018;1819:215–33.
Wang T, Azeddine B, Mah W, Harvey EJ, Rosenblatt D, Seguin C. Osteonecrosis of the femoral head: genetic basis. Int Orthop. 2019;43(3):519–30.
Yunna C, Mengru H, Lei W, Weidong C. Macrophage M1/M2 polarization. Eur J Pharmacol. 2020;877:173090.
Jiang C, Zhou Z, Lin Y, Shan H, Xia W, Yin F. et al. Astragaloside IV ameliorates steroid-induced osteonecrosis of the femoral head by repolarizing the phenotype of pro-inflammatory macrophages. Int Immunopharmacol. 2021;93:107345.
Belambri SA, Rolas L, Raad H, Hurtado-Nedelec M, Dang PM, El-Benna J. NADPH oxidase activation in neutrophils: Role of the phosphorylation of its subunits. Eur J Clin Invest. 2018;48 Suppl 2:e12951.
Marciano BE, Spalding C, Fitzgerald A, Mann D, Brown T, Osgood S. et al. Common severe infections in chronic granulomatous disease. Clin Infect Dis. 2015;60(8):1176–83.
Gomez-Moreno D, Adrover JM, Hidalgo A. Neutrophils as effectors of vascular inflammation. Eur J Clin Invest. 2018;48 Suppl 2:e12940.
Huang C, Wen Z, Niu J, Lin S, Wang W. Steroid-Induced Osteonecrosis of the Femoral Head: Novel Insight Into the Roles of Bone Endothelial Cells in Pathogenesis and Treatment. Front Cell Dev Biol. 2021;9:777697.
Inohara N, Ogura Y, Chen FF, Muto A, Nunez G. Human Nod1 confers responsiveness to bacterial lipopolysaccharides. J Biol Chem. 2001;276(4):2551–4.
Thay B, Damm A, Kufer TA, Wai SN, Oscarsson J. Aggregatibacter actinomycetemcomitans outer membrane vesicles are internalized in human host cells and trigger NOD1- and NOD2-dependent NF-kappaB activation. Infect Immun. 2014;82(10):4034–46.
Pandey AK, Yang Y, Jiang Z, Fortune SM, Coulombe F, Behr MA. et al. NOD2, RIP2 and IRF5 play a critical role in the type I interferon response to Mycobacterium tuberculosis. PLoS Pathog. 2009;5(7):e1000500.
Kim YG, Park JH, Shaw MH, Franchi L, Inohara N, Nunez G. The cytosolic sensors Nod1 and Nod2 are critical for bacterial recognition and host defense after exposure to Toll-like receptor ligands. Immunity. 2008;28(2):246–57.
Newton K, Dixit VM. Signaling in innate immunity and inflammation. Cold Spring Harb Perspect Biol. 2012;4(3).
Zhu D, Yu H, Liu P, Yang Q, Chen Y, Luo P. et al. Calycosin modulates inflammation via suppressing TLR4/NF-κB pathway and promotes bone formation to ameliorate glucocorticoid-induced osteonecrosis of the femoral head in rat. Phytother Res. 2021;35(5):2824–35.
Takeuchi O, Akira S. Pattern recognition receptors and inflammation. Cell. 2010;140(6):805–20.
Jin MS, Kim SE, Heo JY, Lee ME, Kim HM, Paik SG. et al. Crystal structure of the TLR1-TLR2 heterodimer induced by binding of a tri-acylated lipopeptide. Cell. 2007;130(6):1071–82.
O'Neill LA, Bowie AG. The family of five: TIR-domain-containing adaptors in Toll-like receptor signalling. Nat Rev Immunol. 2007;7(5):353–64.
Chen C, Wang Y, Yang S, Li H, Zhao G, Wang F. et al. MiR-320a contributes to atherogenesis by augmenting multiple risk factors and down-regulating SRF. J Cell Mol Med. 2015;19(5):970–85.
Regev K, Healy BC, Paul A, Diaz-Cruz C, Mazzola MA, Raheja R. et al. Identification of MS-specific serum miRNAs in an international multicenter study. Neurol Neuroimmunol Neuroinflamm. 2018;5(5):e491.
Yao Y, Ding Y, Bai Y, Zhou Q, Lee H, Li X. et al. Identification of Serum Circulating MicroRNAs as Novel Diagnostic Biomarkers of Gastric Cancer. Front Genet. 2020;11:591515.
Dandare A, Rabia G, Khan MJ. In silico analysis of non-coding RNAs and putative target genes implicated in metabolic syndrome. Comput Biol Med. 2021;130:104229.

Table 1. Primer information.
Target name		Primer
b-Actin	F	GACAGGATGCAGAAGGAGATTACT
	R	TGATCCACATCTGCTGGAAGGT
NOD2	F	TTGCCTAGTTCTGGAAGGCTG
	R	CCTCTTCCCCCATCAAAGCC
CYBB	F	GCAGGAAAGGAACAATGCCG
	R	CATTATCCCAGTTGGGCCGT
TLR1	F	GCCACCCTACTGTGAACCTC
	R	ATGAGCAATTGGCAGCACAC
CD14	F	ACAGGTGCCTAAAGGACTGC
	R	GATTCCCGTCCAGTGTCAGG

Table 2. Potential drugs for treating SONFH
Gene	Drug	Sources	Pmids
NOD2	CHEMBL1456848	DTC
NOD2	CHEMBL585502	DTC
NOD2	MIFAMURTIDE	ChemblInteractions	20596505\|21226638
NOD2	INARIGIVIR SOPROXIL	TTD
NOD2	CHEMBL599890	DTC
NOD2	CHEMBL578944	DTC
NOD2	CHEMBL577660	DTC
NOD2	MURABUTIDE	DTC	22716113
NOD2	TACROLIMUS	PharmGKB	23175667
NOD2	CHEMBL590706	DTC
NOD2	CHEMBL1351908	DTC
NOD2	CYCLOVALONE	DTC
CYBB	APIGENIN	DTC	23786520
CYBB	CHRYSIN	DTC	23786520
CYBB	LUTEOLIN	DTC	23786520
CD14	IC14	TdgClinicalTrial\|ChemblInteractions\|TTD
CD14	LOVASTATIN	NCI	7506029

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of inflammation-related genes and exploration of regulatory mechanisms in patients with osteonecrosis of the femoral head

Status:

Version 1

Abstract

Background:

Results:

Conclusions:

Figures

Background

Material And Methods

Results

Discussion

Conclusions

Abbreviations

Declarations

References

Tables

Additional Declarations

Status:

Version 1