Identification of critical genes in nucleus pulposus cells isolated from degenerated intervertebral discs using bioinformatics analysis
- Authors:
- Published online on: May 31, 2017 https://doi.org/10.3892/mmr.2017.6662
- Pages: 553-564
-
Copyright: © Zhu et al. This is an open access article distributed under the terms of Creative Commons Attribution License.
Abstract
Introduction
Intervertebral disc (IVD) degeneration, also termed degenerative disc disorder or degenerative disc disease, is a pathological process that may induce acute or chronic lower back pain (1,2). Lower back pain is one of the primary health problems in developed countries (3). The risk factors for disc degeneration include genetic inheritance and environmental risk factors, including smoking cigarettes and repetitive and high mechanical loading (4). IVD degeneration is a rapidly progressing disease without an effective therapeutic method (5). Therefore, it is necessary to explore the mechanisms of IVD degeneration in order to be able to develop a novel treatment scheme.
IVD degeneration and the underlying molecular mechanisms have been previously investigated. The aggrecanases ADAM metallopeptidase with thrombospondin type 1 motif (ADAMTS)-1, -4, -5, -9 and -15 may promote extracellular matrix (ECM) alterations during IVD degeneration, and may be used for preventing IVD degeneration and its morbidity (6). In disc cells, reduced expression of SRY-type high mobility group box 9 (SOX9) may be associated with disc degeneration and disc ageing via inhibition of type II collagen expression (7). The growth differentiation factor-5 (GDF-5) cDNA and the recombinant GDF-5 protein may promote the expression of ECM protein-coding genes in mouse IVD cells (8). Previous studies have detected overexpressed tumor necrosis factor α (TNF-α) and interleukin (IL)-1 in aged and degenerative IVDs obtained from human and animal models (9,10). IL-1 has been identified to be involved in IVD degeneration via directly inhibiting matrix synthesis and promoting matrix degradation (11,12). Cytokines of IL-1 and TNF-α may be associated with the pathogenesis of IVD degeneration; however, IL-1 may have a greater contribution to IVD degeneration and may be a more suitable therapeutic target for the disease (13).
In 2013, Markova et al (14) established a rat disc organ culture model that mimicked IVD degeneration via culturing rat IVDs in the presence of IL-1β, TNF-α and serum-limiting conditions. They obtained 1036 differentially expressed genes (DEGs) between experimental and control groups following gene expression analysis for microarray data. The present study used the data from Markova et al (14) and the DEGs between degenerated and normal nucleus pulposus cells were identified, and their possible functions were predicted using enrichment analysis. Additionally, protein-protein interaction (PPI) networks were visualized and module analysis was conducted to screen for key genes in degenerated nucleus pulposus cells.
Materials and methods
Microarray data
Microarray data obtained from GSE42611 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE42611), which was downloaded from the database of Gene Expression Omnibus (GEO), were sequenced on the platform of GPL6247 Affymetrix Rat Gene 1.0 ST Array [transcript (gene) version]. GSE42611 included 4 nucleus pulposus samples isolated from degenerated IVDs and 4 nucleus pulposus samples separated from normal IVDs. The procedure that had been used to obtain the rat lumbar disc specimens (n=4 specimens/group) was as follows, according to the method of Ponnappan et al (8): Whole lumbar IVDs with endplates had been dissected and preserved in organ culture. Lumbar discs in the experimental group had been cultivated in Dulbecco's modified Eagle's medium (DMEM) containing 100 ng/ml TNF-α, 10 ng/ml IL-1β, 50 µg/ml L-ascorbate, 40 mM NaCl, 1% fetal bovine serum (FBS), antibiotics and antimycotics. Lumbar discs in the control group had been cultured in DMEM containing 50 µg/ml L-ascorbate, 40 mM NaCl, 10% FBS and antibiotics without cytokines. The discs had been cultured for a total of 10 days (14). GSE42611 used in this study was downloaded from a public database; therefore, patient consent or ethics committee approval were not required.
Data preprocessing and DEGs screening
GSE42611 was downloaded and the microarray data was preprocessed using the Affy package (15) in R. The process of data preprocessing included background correction, quantile normalization, summarization and probe ID to gene symbol transformation. Linear models for microarray data in the limma package (16) in R were used to analyze the DEGs between degenerated and normal nucleus pulposus cells. P-values of the DEGs were calculated separately and adjusted using the t-test method and the Benjamini & Hochberg method (17). P<0.05 and |log2 fold-change (FC)|>1 were used as the thresholds.
Functional and pathway enrichment analysis
The Database for Annotation, Visualization and Integrated Discovery (DAVID; david.abcc.ncifcrf.gov) software was used to interpret functions of extensive genes obtained from previous genome studies (18). The Gene Ontology database (GO; www.geneontology.org) contained structured ontologies or vocabularies that depict basic characteristics of genes and gene products (19). The Kyoto Encyclopedia of Genes and Genomes database (KEGG; www.genome.jp/kegg/) synthesizes information of biological systems from genomic, chemical and systemic functional aspects (20). Using the DAVID software, functional and pathway enrichment analyses were conducted separately, for upregulated and downregulated genes. P≤0.05 and >2 enriched genes were set as the thresholds.
PPI network construction and module analysis
The Search Tool for the Retrieval of Interacting Genes (STRING; string-db.org) database provide comprehensive and easily accessible interaction information derived from experiments and predictions (21). Cytoscape software (www.cytoscape.org) was used to integrate high-throughput expression data and biomolecular interaction networks into a unified framework (22). The PPIs obtained for the DEGs were searched using the STRING database (21), with the required confidence (combined score) >0.4 as the threshold. Using Cytoscape software version 2.8 (22), the PPIs were used to established a PPI network. In the network, the proteins were termed nodes and the number of edges involved were their degrees. Finally, the MCODE plug-in (23) in Cytoscape was used to perform module analysis of the PPI networks. The parameters were set at the default thresholds.
Results
DEG analysis
P<0.05 and |log2FC|>1 were set as thresholds and the DEGs between degenerated and normal nucleus pulposus cells were analyzed. There were 558 DEGs identified in the degenerated nucleus pulposus cells compared with normal nucleus pulposus cells, including 253 upregulated and 305 downregulated genes. There were more downregulated genes compared with upregulated genes.
Functional and pathway enrichment analysis
The upregulated genes in the degenerated nucleus pulposus cells were significantly enriched in 255 GO terms and 9 KEGG pathways. The top 10 functions are presented in Table IA, including response to wounding (P=2.35×10−8), inflammatory response (P=5.99×10−8) and response to organic substance (P=1.56×10−7).
Table I.The top 10 enriched functions for the differentially expressed genes in the degenerated nucleus pulposus cells |
The downregulated genes in the degenerated nucleus pulposus cells were significantly enriched in 263 GO terms and 10 KEGG pathways. The top 10 functions included M phase (P=9.18×10−12), cell cycle phase (P=2.81×10−10) and response to steroid hormone stimulus (P=2.95×10−9; Table IB).
Additionally, the upregulated genes were significantly enriched in cytokine-cytokine receptor interaction (P=2.86×10−4), apoptosis (P=3.95×10−4) and chemokine (P=1.60×10−3; Table IIA) signaling pathways.
Table II.Enriched pathways for the differentially expressed genes in the degenerated nucleus pulposus cells. |
The pathways enriched for the downregulated genes included ECM-receptor interaction [P=1.17×10−11, involving thrombospondin 1 (THBS1)], focal adhesion (P=1.90×10−9) and hematopoietic cell lineage (P=3.12×10−3; Table IIB).
PPI network construction and module analysis
PPI networks were constructed by Cytoscape software following a PPI search of the DEGs. The PPI networks for the upregulated (Fig. 1) and the downregulated (Fig. 2) genes separately had 360 and 1,112 interactions. Notably, IL-6 (degree=39) in the PPI network for the upregulated genes and vascular endothelial growth factor A (VEGFA; degree=37) in the PPI network for the downregulated genes had higher degrees. Using the MCODE plug-in in Cytoscape, four modules (µM1, µM2, µM3 and µM4) were identified from the PPI network for the upregulated genes (Fig. 3). Meanwhile, four modules (dM1, dM2, dM3 and dM4) were identified from the PPI network for the downregulated genes (Fig. 4). It is of note that collagen, type I, α1 (COL1A1), COL1A2, COL3A1, COL5A2, COL6A1, COL6A2, COL6A3, COL11A1, COL12A1 and integrin α4 (ITGA4) may interact with each other in the dM2 module.
The top 5 functions enriched for the upregulated genes in modules included taxis (µM1; P=1.03×10−8), response to organic substance (µM2; P=1.21×10−4), response to cytokine stimulus (µM3; P=5.04×10−4) and chemical homeostasis (µM4, P=1.22×10−3; Table IIIA). The pathways enriched for the upregulated genes in modules included the chemokine signaling pathway (µM1; P=1.10×10−4) and the Jak-STAT signaling pathway (µM2; P=7.69 ×10−4; Table IIIB). Additionally, the top 5 functions enriched for the downregulated genes in modules, included M phase (dM1; P=2.70×10−16), extracellular matrix organization (dM2; P=1.79×10−7, including COL3A1, COL1A2, COL1A1, COL11A1 and COL5A2), G-protein coupled receptor protein signaling pathway (dM3; P=7.27×10−4) and wound healing (dM4; P=1.56 ×10−4; Table IVA). The pathways enriched for the downregulated genes in modules included cell cycle (dM1; P=4.40×10−5), ECM-receptor interaction (dM2; P=1.37×10−15, including COL3A1, COL6A3, COL1A2, COL6A2, COL6A1, ITGA4, COL1A1, COL11A1 and COL5A2) and neuroactive ligand-receptor interaction (dM3; P=9.49×10−4; Table IVB).
Table III.Top 5 functions and pathways enriched for the upregulated genes in µM1, µM2, µM3 and µM4 modules. |
Table IV.Top 5 functions and pathways enriched for the downregulated genes in dM1, dM2, dM3 and dM4 modules. |
Discussion
The present study identified a total of 558 DEGs in degenerated nucleus pulposus cells compared with normal nucleus pulposus cells, including 253 upregulated and 305 downregulated genes. Using the MCODE plug-in in Cytoscape, four modules (µM1, µ0M2, µM3 and µM4) were identified from the PPI network for the upregulated genes. Additionally, four modules (dM1, dM2, dM3 and dM4) were identified from the PPI network for the downregulated genes.
A previous study demonstrated that genetic variations of IL-6 may be associated with IVD degeneration, accompanied by sciatica (24). VEGFA was overexpressed in the nucleus pulposus and affects the survival of nucleus pulposus cells in an autocrine/paracrine manner (25). Injuries of IVDs may lead to increased VEGF levels, indicating that VEGF may be associated with discogenic back pain (26). Under co-culture conditions, VEGF induction may contribute to neo-vascularization of IVD tissue and may function in the resorption of herniated discs (27). The findings of the present study indicated that IL-6 (degree=39) in the PPI network for the upregulated genes and VEGFA (degree=37) in the PPI network for the downregulated genes had higher degrees. Therefore, IL6 and VEGFA may be key genes involved in IVD degeneration. A previous study observed the immunolocalization of THBS in human IVD (28). THBS1 and THBS2 are promising susceptibility genes in lumbar-disc herniation (LDH) that mediate the expression levels of matrix metalloproteinases (MMPs) 2 and 9, which are critical effectors of ECM remodeling (29). Mice with THBS1 or THBS2 deficiency exhibit abnormal spine curvature (30). Pathway enrichment performed in the present study revealed that downregulated THBS1 was enriched in ECM-receptor interactions, suggesting that THBS1 may have an important role in IVD degeneration.
The sequence variation of the regulatory region of COL1A1 is closely associated with lumbar disc disease (LDD) in young military recruits who are newly diagnosed (31). Ribosomal protein L8, ribosomal protein S16 and ribosomal protein S23 have been identified to contribute to protein synthesis, and COL3A1 was involved in skeletal system processes in disc degeneration (DD), indicating that they may be used for diagnosis and therapy of DD (32). Polymorphisms of the COL9 and COL11 genes contribute to the progression of degenerative lumbar spinal stenosis (33). COL11A1 expression level was reduced in the IVD of patients with LDH and it had a negative association with the severity of disc degeneration in patients with LDH (34). In the dM2 module identified by the present study, COL1A1, COL1A2, COL3A1, COL5A2, COL6A1, COL6A2, COL6A3, COL11A1, COL12A1 and ITGA4 may interact with each other. Functional enrichment indicated that collagen genes were enriched in ECM organization. Therefore, collagen genes may contribute to the progression of IVD degeneration. Additionally, ITGA4 may also be implicated in IVD degeneration via interaction with collagen genes.
In conclusion, the present study investigated the underlying mechanisms of IVD degeneration via bioinformatics analysis. A total 558 DEGs were screened in the degenerated nucleus pulposus cells. IL6, VEGFA, THBS1, ITGA4 and collagen genes may be involved in the progression of IVD degeneration. These results suggested that the manipulation of these genes and their products may have potential as a novel therapeutic strategy for the treatment of patients with IVD. However, these findings were obtained by bioinformatics prediction and require further confirmation via further experimental studies.
Acknowledgements
The present study was supported by the Shandong Province Pharmaceutical Technology Development Program (grant no. 2015-261), the Projects of Medical and Health Technology Development Program in Shandong Province, China (grant no. 2014WS0502), the Taishan Medical University Cultivate High-level Task Projects (grant no. 2014GCC02) and the Projects of Health Science and Technology Association in Shandong Province, China (grant no. 2016BJ0009).
References
Sakai D, Mochida J, Yamamoto Y, Nomura T, Okuma M, Nishimura K, Nakai T, Ando K and Hotta T: Transplantation of mesenchymal stem cells embedded in Atelocollagen gel to the intervertebral disc: A potential therapeutic model for disc degeneratio. Biomaterials. 24:3531–3541. 2003. View Article : Google Scholar : PubMed/NCBI | |
Le Maitre CL, Pockert A, Buttle DJ, Freemont AJ and Hoyland JA: Matrix synthesis and degradation in human intervertebral disc degeneration. Biochem Soc Trans. 35:652–655. 2007. View Article : Google Scholar : PubMed/NCBI | |
Richardson SM, Walker RV, Parker S, Rhodes NP, Hunt JA, Freemont AJ and Hoyland JA: Intervertebral disc cell-mediated mesenchymal stem cell differentiation. Stem Cells. 24:707–716. 2006. View Article : Google Scholar : PubMed/NCBI | |
Battié MC, Videman T and Parent E: Lumbar disc degeneration: Epidemiology and genetic influences. Spine. 29:2679–2690. 2004. View Article : Google Scholar : PubMed/NCBI | |
Sakai D, Mochida J, Iwashina T, Hiyama A, Omi H, Imai M, Nakai T, Ando K and Hotta T: Regenerative effects of transplanting mesenchymal stem cells embedded in atelocollagen to the degenerated intervertebral disc. Biomaterials. 27:335–345. 2006. View Article : Google Scholar : PubMed/NCBI | |
Pockert AJ, Richardson SM, Le Maitre CL, Lyon M, Deakin JA, Buttle DJ, Freemont AJ and Hoyland JA: Modified expression of the ADAMTS enzymes and tissue inhibitor of metalloproteinases 3 during human intervertebral disc degeneration. Arthritis Rheum. 60:482–491. 2009. View Article : Google Scholar : PubMed/NCBI | |
Gruber HE, Norton HJ, Ingram JA and Hanley EN Jr: The SOX9 transcription factor in the human disc: Decreased immunolocalization with age and disc degeneration. Spine (Phila Pa 1976). 30:625–630. 2005. View Article : Google Scholar : PubMed/NCBI | |
Ponnappan RK, Markova DZ, Antonio PJ, Murray HB, Vaccaro AR, Shapiro IM, Anderson DG, Albert TJ and Risbud MV: An organ culture system to model early degenerative changes of the intervertebral disc. Arthritis Res Ther. 13:R1712011. View Article : Google Scholar : PubMed/NCBI | |
Weiler C, Nerlich AG, Bachmeier BE and Boos N: Expression and distribution of tumor necrosis factor alpha in human lumbar intervertebral discs: A study in surgical specimen and autopsy controls. Spine (Phila Pa 1976). 30:44–54. 2005. View Article : Google Scholar : PubMed/NCBI | |
Bachmeier BE, Nerlich AG, Weiler C, Paesold G, Jochum M and Boos N: Analysis of tissue distribution of TNF-alpha, TNF-alpha-receptors and the activating TNF-alpha-converting enzyme suggests activation of the TNF-alpha system in the aging intervertebral disc. Ann N Y Acad Sci. 1096:44–54. 2007. View Article : Google Scholar : PubMed/NCBI | |
Le Maitre CL, Freemont AJ and Hoyland JA: The role of interleukin-1 in the pathogenesis of human intervertebral disc degeneration. Arthritis Res Ther. 7:R732–R745. 2005. View Article : Google Scholar : PubMed/NCBI | |
Hoyland J, Le Maitre C and Freemont A: Investigation of the role of IL-1 and TNF in matrix degradation in the intervertebral disc. Rheumatology (Oxford). 47:809–814. 2008. View Article : Google Scholar : PubMed/NCBI | |
Le Maitre CL, Hoyland JA and Freemont AJ: Catabolic cytokine expression in degenerate and herniated human intervertebral discs: IL-1beta and TNFalpha expression profile. Arthritis Res Ther. 9:R772007. View Article : Google Scholar : PubMed/NCBI | |
Markova DZ, Kepler CK, Addya S, Murray HB, Vaccaro AR, Shapiro IM, Anderson DG, Albert TJ and Risbud MV: An organ culture system to model early degenerative changes of the intervertebral disc II: Profiling global gene expression changes. Arthritis Res Ther. 15:R1212013. View Article : Google Scholar : PubMed/NCBI | |
Gautier L, Cope L, Bolstad BM and Irizarry RA: affy-analysis of Affymetrix GeneChip data at the probe level. Bioinformatics. 20:307–315. 2004. View Article : Google Scholar : PubMed/NCBI | |
Smyth GK: Limma: Linear Models for Microarray Data. Bioinformatics and Computational Biology Solutions Using R and Bioconductor Springer. 397–420. 2005. View Article : Google Scholar | |
Benjamini Y and Hochberg Y: Controlling the false discovery rate: A practical and powerful approach to multiple testing. J Royal Statistical Soci. 57:289–300. 1995. | |
Huang DW, Sherman BT, Tan Q, Kir J, Liu D, Bryant D, Guo Y, Stephens R, Baseler MW, Lane HC and Lempicki RA: DAVID bioinformatics resources: Expanded annotation database and novel algorithms to better extract biology from large gene lists. Nucleic Acids Res. 35:W169–W175. 2007. View Article : Google Scholar : PubMed/NCBI | |
Carbon S, Ireland A, Mungall CJ, Shu S, Marshall B and Lewis S: AmiGO Hub; Web Presence Working Group: AmiGO: Online access to ontology and annotation data. Bioinformatics. 25:288–289. 2009. View Article : Google Scholar : PubMed/NCBI | |
Kanehisa M, Araki M, Goto S, Hattori M, Hirakawa M, Itoh M, Katayama T, Kawashima S, Okuda S, Tokimatsu T and Yamanishi Y: KEGG for linking genomes to life and the environment. Nucleic Acids Res. 36:(Database issue). D480–D484. 2008. View Article : Google Scholar : PubMed/NCBI | |
Szklarczyk D, Franceschini A, Kuhn M, Simonovic M, Roth A, Minguez P, Doerks T, Stark M, Muller J, Bork P, et al: The STRING database in 2011: Functional interaction networks of proteins, globally integrated and scored. Nucleic Acids Res. 39:(Database issue). D561–D568. 2011. View Article : Google Scholar : PubMed/NCBI | |
Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, Amin N, Schwikowski B and Ideker T: Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res. 13:2498–2504. 2003. View Article : Google Scholar : PubMed/NCBI | |
Bader GD and Hogue CW: An automated method for finding molecular complexes in large protein interaction networks. BMC bioinformatics. 4:22003. View Article : Google Scholar : PubMed/NCBI | |
Noponen-Hietala N, Virtanen I, Karttunen R, Schwenke S, Jakkula E, Li H, Merikivi R, Barral S, Ott J, Karppinen J and Ala-Kokko L: Genetic variations in IL6 associate with intervertebral disc disease characterized by sciatica. Pain. 114:186–194. 2005. View Article : Google Scholar : PubMed/NCBI | |
Fujita N, Imai J, Suzuki T, Yamada M, Ninomiya K, Miyamoto K, Iwasaki R, Morioka H, Matsumoto M, Chiba K, et al: Vascular endothelial growth factor-A is a survival factor for nucleus pulposus cells in the intervertebral disc. Biochem Biophys Res Commun. 372:367–372. 2008. View Article : Google Scholar : PubMed/NCBI | |
Sato J, Sakuma Y, Yamauchi K, Orita S, Kubota G, Oikawa Y, Inage K, Sainoh T, Fujimoto K, Takahashi K, et al: Elevated VEGF in degenerative intervertebral discs in rats with injured intervertebral discs of the caudal vertebrae. Global Spine J. 4:po. 165. 2014. View Article : Google Scholar | |
Haro H, Kato T, Komori H, Osada M and Shinomiya K: Vascular endothelial growth factor (VEGF)-induced angiogenesis in herniated disc resorption. J Orthop Res. 20:409–415. 2002. View Article : Google Scholar : PubMed/NCBI | |
Gruber HE, Ingram JA and Hanley EN Jr: Immunolocalization of thrombospondin in the human and sand rat intervertebral disc. Spine (Phila Pa 1976). 31:2556–2561. 2006. View Article : Google Scholar : PubMed/NCBI | |
Hirose Y, Chiba K, Karasugi T, Nakajima M, Kawaguchi Y, Mikami Y, Furuichi T, Mio F, Miyake A, Miyamoto T, et al: A functional polymorphism in THBS2 that affects alternative splicing and MMP binding is associated with lumbar-disc herniation. Am J Med Genet. 82:1122–1129. 2008. | |
Lawler J, Sunday M, Thibert V, Duquette M, George EL, Rayburn H and Hynes RO: Thrombospondin-1 is required for normal murine pulmonary homeostasis and its absence causes pneumonia. J Clin Invest. 101:982–992. 1998. View Article : Google Scholar : PubMed/NCBI | |
Tilkeridis C, Bei T, Garantziotis S and Stratakis CA: Association of a COL1A1 polymorphism with lumbar disc disease in young military recruits. J Med Genet. 42:e442005. View Article : Google Scholar : PubMed/NCBI | |
Yang Z, Chen X, Zhang Q, Cai B, Chen K, Chen Z, Bai Y, Shi Z and Li M: Dysregulated COL3A1 and RPL8, RPS16, and RPS23 in disc degeneration revealed by bioinformatics methods. Spine (Phila Pa 1976). 40:E745–E751. 2015. View Article : Google Scholar : PubMed/NCBI | |
Noponen-Hietala N, Kyllönen E, Männikkö M, Ilkko E, Karppinen J, Ott J and Ala-Kokko L: Sequence variations in the collagen IX and XI genes are associated with degenerative lumbar spinal stenosis. Ann Rheum Dis. 62:1208–1214. 2003. View Article : Google Scholar : PubMed/NCBI | |
Mio F, Chiba K, Hirose Y, Kawaguchi Y, Mikami Y, Oya T, Mori M, Kamata M, Matsumoto M, Ozaki K, et al: A functional polymorphism in COL11A1, which encodes the alpha 1 chain of type XI collagen, is associated with susceptibility to lumbar disc herniation. Am J Med Genet. 81:1271–1277. 2007. |