Analysis of critical molecules and signaling pathways in osteoarthritis and rheumatoid arthritis
- Authors:
- Published online on: December 4, 2012 https://doi.org/10.3892/mmr.2012.1224
- Pages: 603-607
Abstract
Introduction
Osteoarthritis (OA) is a group of states associated with defective articular cartilage and changes in the underlying bone.OA is divided into erosive or non-erosive. Erosive OA is more abrupt and commonly exhibits subchondral bone erosions (1). Pathological changes in articular cartilage and subchondral bone result from chondrocyte imbalance in the extracellular matrix. Although numerous studies have reported probable chemical or mechanical causes of cartilage destruction (2,3), this area of research requires more detailed investigation.
Rheumatoid arthritis (RA) is a complex, chronic multisystemic autoimmune disease, which affects the synovial membranes of multiple joints, cartilage and bone as well as bursa and tendon sheaths (4). RA is a prevalent chronic inflammatory joint disease affecting 0.5–1% of the world’s population (5). RA leads to severe morbidity and disability if incorrectly treated, imposing a substantial economic burden on the affected individuals and society. The inflammatory process associated with RA is primarily observed in the synovial tissue. Synovial hyperplasia results from synovial outgrowths or synovial villi, comprised of macrophages, synovial lining cells, lymphocytes and blood vessels (6). Joint destruction occurs when the synovial pannus produces enzymes resulting in cartilage penetration, cartilage damage and joint erosion (7).
Although RA and OA share similar symptoms, it has been demonstrated that RA follows an alternative inflammatory pathway of pathogenesis to OA. Diagnosis and assessment of RA and OA is largely based on semi-quantitative methods of diagnosis, including symptoms, joint damage and physical function (8). At present, no cure exists for RA and OA and the management of these diseases depends upon early detection and aggressive treatment. Therefore, it is increasingly important to explore the molecular mechanisms of these diseases and analyze the associated signaling pathways, in order to uncover an effective therapeutic approach. In this study, we analyzed gene expression profiles of OA and RA cells to determine differentially expressed genes (DEGs) in the two forms of arthritis. Furthermore, through comparison we determined changed metabolic and non-metabolic pathways, small bioactive molecules and SNP corresponding genes associated with RA and OA.
Materials and methods
Gene expression profiles of synovial tissue samples from RA and OA patients and normal donors
The transcription profile of GSE1919 was obtained from the National Center for Biotechnology Information Gene Expression Omnibus database (http://www.ncbi.nlm.nih.gov/geo/) and was based on the Affymetrix Human Genome U95A Array (Santa Clara, CA, USA). A total of 15 chips were used in this study, including 5 OA tissue chips, 5 RA tissue chips and 5 normal donor (ND) tissue chips (9). The study was approved by the Ethics Committee of the Charité Universitätsmedizin Berlin and has been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. All patients gave their informed consent prior to their inclusion in the study.
Analysis of DEGs
Raw data were normalized using the robust multichip average method (10) with the default settings implemented in the R affy package (version 2.13.0). The Limma (linear models for microarray data) method was used to identify DEGs (11). The original expression datasets from all conditions were extracted into expression estimates and used to construct the linear model. Significance of gene expression differences between OA, RA and ND cells were tested by classical t-test and P-values were adjusted for multiple comparisons using the false discovery rate (FDR) of Benjamini and Hochberg (12). FDR-corrected P<0.05 was considered to indicate a statistically significant difference.
Pathway analysis
Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database is a collection of manually drawn pathway maps of the molecular interaction and reaction networks. A total of 130 pathways, involving 2,287 genes, were collected from KEGG (updated 2011/06). The Database for Annotation, Visualization and Integrated Discovery (DAVID), a high-throughput and integrated data-mining environment, analyzed gene lists derived from high-throughput genomic experiments. The DAVID (13,14) was used to identify over-represented pathways based on the hypergeometric distribution. Pathways with P<0.05 and count >2 were considered to be significant.
Small molecule expression analysis
CMap (the connectivity map) is a collection of gene expression profiles from cultured human cells treated with bioactive small molecules. The database contains 6,100 bioactive small molecule-interfering tests and 7,056 corresponding gene expression profiles (15). DEGs were analyzed through the CMap database to identify small bio-active molecules which resulted in similar or adverse gene expression. Probes of DEGs were converted to the accession number of GeneBank and then probe numbers for use in the CMap.
RA- and OA-related SNP analysis
RA- and OA- related SNPs were obtained following a dbSNP database search (http://www.ncbi.nlm.nih.gov/projects/SNP/) using the keywords ‘osteoarthritis’ and ‘osteoporosis’. SNPs of RA and OA, which were suitable for probing, were acquired through the comparison of their corresponding genes with DEGs under pathological conditions.
Results
Recognition of DEGs in different samples
Analysis of GSE1919 using the Limma method identified a total of 460 DEGs in OA tissues compared with normal tissues. In RAt issues 1,148 DEGs were identified in comparison with ND (P<0.05; Fig. 1). Only 196 DEGs were identified when we compared the gene expression in OA tissues with that of RA. These results indicate similarities in the molecular mechanisms of OA and RA.
Analysis of pathways induced by arthritis
In arthritic tissues, gene expression profiles were significantly changed compared with the normal tissues. Therefore, DEGs were adopted to perform KEGG sub-pathway enrichment analysis. Significantly changed pathways with ≥2 genes included and P<0.05 in arthritis were obtained. It was demonstrated that the majority of pathways were involved in metabolic and non-metabolic processes, indicative of a large number of changes in arthritic tissues compared with normal tissues (Table I). These results are likely to be important for drug discovery for the treatment of arthritis.
Analysis of small molecules resulting in RA and OA
DEGs were first divided into upregulated and downregulated genes and then enriched with significantly changed genes obtained from treatment of small molecules from the CMap database. Targeted molecules observed to induce similar effects to arthritis were selected and the 20 targeted molecules with the lowest P-values were enumerated (Table II). Parthenolide and alsterpaullone were identified in OA and RA tissue analysis and are highly relevent molecules.
 | Table IIIntersection of gene expressions between small bio-active molecules and the differentially expressed genes of arthritis. |
Analysis of RA- and OA-related SNPs
A total of 10 SNPs were obtained from the dbSNP database with the keyword ‘osteoarthritis’ and 15 were obtained with the keyword ‘osteoporosis’. Following comparison of these acquired SNP corresponding genes with DEGs, we revealed that no arthritis-related SNP corresponding genes were the same as the previously identified DEGs and three osteoporosis-related SNP corresponding genes were identified in the DEGs (Table III).
Discussion
During the development of OA and RA, significant changes in gene expression occur. The present study demonstrated that more than 320 genes changed in OA and RA. Study of these common DEGs may help identify potential broad-spectrum anti-arthritis drugs. Furthermore, only 196 DEGs were identified between OA and RA, including interleukin 3 receptor α (IL3RA), transforming growth factor β receptor III and CRYAB, indicating that the two diseases are correlated and drugs that simultaneously treat these diseases may exist.
Cluster analysis of DEGs demonstrated several common pathways associated with these diseases, including the classic mitogen-activated protein kinase (MAPK) signaling pathway and the insulin signaling pathway. The signaling pathways leading to MAPK activation have been linked to various catabolic responses in diseases, including arthritis (16,17). Two immune pathways, antigen processing and presentation and intestinal immune network for IgA production, were also activated in the arthritic tissues, indicating that the immune response is involved in these diseases. Inflammation and cytokines play significant roles in RA and in certain cases of OA (17). Furthermore, changes of several cell adhesion molecules, including integrin β2 (ITGB2) (18) and protein tyrosine phosphatase receptor type c (PTPRC) (19) and lysosome-related molecules, including phospholipase A2 group XV (PLA2G15) (20) and adaptor-related protein complex 1β (AP1B1) (21) in OA and RA cells demonstrated that the two diseases altered their microenvironment and removed the exogenous substances by the lysosome. In addition, graft-versus-host (22) and autoimmune thyroid disease-related pathways were also activated in OA and RA (23). Further investigation of these pathways is likely to be shed light the network of signal pathways under OA and RA. More studies on the DEGs between OA and RA may provide useful information to differentiate the molecular mechanisms associated with OA and RA.
Based on the DEGs and data from the CMap database, we acquired a series of small molecules. Two of these small molecules, parthenolide and alsterpaullone, demonstrated significant similarity in OA and RA tissues (P<0.05) and required additional analysis to determine their suitability as broad spectrum anti-arthritis drugs.
Parthenolide is a sesquiterpene lactone of the germacranolide class which occurs naturally in the plant feverfew (Tanacetum parthenium). Parthenolide modulates the NF-κB-mediated inflammatory responses in experimental atherosclerosis (24) and blocks lipopolysaccharide-induced osteolysis through suppression of NF-κB activity (25). Parthenolide induces apoptosis in acute myelogenous leukemia cells, leaving normal bone marrow cells relatively unscathed (26). Pharthenolide also exhibits microtubule-interfering activity (27), anti-inflammatory and anti-hyperalgesic effects (28) and activity against the parasite Leishmania amazonensis(29). Since numerous cases of OA and RA result from the interruption of immune responses, including inflammation, it is probable that parthenolide is a suitable therapeutic for these ailments.
Alsterpaullone is a potent, ATP-competitive inhibitor of the cell cycle regulating cyclin-dependent kinases CDK1/cyclin B (IC50 = 0.035 μM) and an inhibitor of GSK-3β and the neuronal CDK5/p25 (30). In addition, alsterpaullone induces apoptosis by activation of caspase-8 and -9 followed by disruption of mitochondrial potential (31).
Through analysis of RA- and OA-related SNPs, we identified that three osteoporosis-related SNP corresponding genes (IGF1, COL1A2 and SATB2) were differentially expressed. Insulin-like growth factor 1 (IGF-1), also called somatomedin C, is a protein encoded by the IGF1 gene in humans (32,33). IGF-1 is expressed and produced by chondrocytes and is one of the anabolic growth factors associated with cartilage (34) and thus is involved in arthritis. Collagen α2(I) chain is a protein encoded by the COL1A2 gene in humans (35,36). Mutations in this gene are associated with osteogenesis imperfecta, Ehlers-Danlos syndrome, idiopathic osteoporosis and atypical Marfan syndrome. Special AT-rich sequence-binding protein 2 (SATB2), also known as DNA-binding protein SATB2, is a human protein encoded by the SATB2 gene (37). SATB2 has been identified to be disrupted in two unrelated cases with de novo apparently balanced chromosome translocations associated with cleft palate and Pierre Robin Sequence (38). The present study also demonstrated that these DEGs were often mutated under arthritis and thereby more studies should focus on their roles in OA and RA.
The present findings shed new light on the molecular mechanisms of OA and RA. Results revealed more than 320 DEGs in both diseases which may be involved in OA and RA development via MAPK and insulin signaling pathways, antigen processing and presentation, intestinal immune network, graft-versus-host disease and autoimmune thyroid disease-related pathways. Notably, parthenolide and alsterpaullone were identified as important small molecules involved in the induction of anti-inflammatory and apoptosis-related gene expression and thus we suggest that these molecules may be suitable anti-arthritis drugs for OA and RA. Furthermore, mutations of IGF1, COL1A2 and SATB2 genes were critical for the pathogenesis of OA and RA. However, there are specific limitations in our study. The pathway enrichment was only based on the connection between genes and therefore genes without strong neighbors are likely to be excluded from the analysis (13). In addition, further experimental analysis is required to confirm the conclusions of the present study.
References
|
Roach HI: The complex pathology of osteoarthritis: even mitochondria are involved. Arthritis Rheum. 58:2217–2218. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Berenbaum F: New horizons and perspectives in the treatment of osteoarthritis. Arthritis Res Ther. 10(Suppl 2): S12008. View Article : Google Scholar : PubMed/NCBI | |
|
Dieppe P: Osteoarthritis of the knee in primary care. BMJ. 336:105–106. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Gay S, Gay RE and Koopman WJ: Molecular and cellular mechanisms of joint destruction in rheumatoid arthritis: two cellular mechanisms explain joint destruction? Ann Rheum Dis. 52(Suppl 1): S39–S47. 1993. View Article : Google Scholar : PubMed/NCBI | |
|
Kvien TK: Epidemiology and burden of illness of rheumatoid arthritis. Pharmacoeconomics. 22:1–12. 2004. View Article : Google Scholar : PubMed/NCBI | |
|
Beasley J: Osteoarthritis and rheumatoid arthritis: conservative therapeutic management. J Hand Ther. 25:163–171. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Scott DL and Symmons DP: The role of specialists in managing established rheumatoid arthritis. Rheumatology (Oxford). 47:237–238. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Biswas S, Manikandan J and Pushparaj PN: Decoding the differential biomarkers of Rheumatoid arthritis and Osteoarthritis: A functional genomics paradigm to design disease specific therapeutics. Bioinformation. 6:153–157. 2011. View Article : Google Scholar : PubMed/NCBI | |
|
Ungethuem U, Haeupl T, Witt H, et al: Molecular signatures and new candidates to target the pathogenesis of rheumatoid arthritis. Physiol Genomics. 42A:267–282. 2010. View Article : Google Scholar : PubMed/NCBI | |
|
Irizarry RA, Hobbs B, Collin F, et al: Exploration, normalization and summaries of high density oligonucleotide array probe level data. Biostatistics. 4:249–264. 2003. View Article : Google Scholar | |
|
Smyth GK: Linear models and empirical bayes methods for assessing differential expression in microarray experiments. Stat Appl Genet Mol Biol. 3:32004.PubMed/NCBI | |
|
Benjamini Y and Hochberg Y: Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Series B Stat Methodol. 57:289–300. 1995. | |
|
Huang DW, Sherman BT and Lempicki RA: Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nat Protoc. 4:44–57. 2009.PubMed/NCBI | |
|
Huang da W, Sherman BT and Lempicki RA: Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res. 37:1–13. 2009.PubMed/NCBI | |
|
Lamb J, Crawford ED, Peck D, et al: The connectivity map: using gene-expression signatures to connect small molecules, genes and disease. Science. 313:1929–1935. 2006. View Article : Google Scholar | |
|
Sondergaard BC, Schultz N, Madsen SH, Bay-Jensen AC, Kassem M and Karsdal MA: MAPKs are essential upstream signaling pathways in proteolytic cartilage degradation-divergence in pathways leading to aggrecanase and MMP-mediated articular cartilage degradation. Osteoarthritis Cartilage. 18:279–288. 2010. View Article : Google Scholar | |
|
Lawrence MC, Jivan A, Shao C, et al: The roles of MAPKs in disease. Cell Res. 18:436–442. 2008. View Article : Google Scholar : PubMed/NCBI | |
|
Kotovuori A, Pessa-Morikawa T, Kotovuori P, Nortamo P and Gahmberg CG: ICAM-2 and a peptide from its binding domain are efficient activators of leukocyte adhesion and integrin affinity. J Immunol. 162:6613–6620. 1999.PubMed/NCBI | |
|
Kaplan R, Morse B, Huebner K, et al: Cloning of three human tyrosine phosphatases reveals a multigene family of receptor-linked protein-tyrosine-phosphatases expressed in brain. Proc Natl Acad Sci U S A. 87:7000–7004. 1990. View Article : Google Scholar | |
|
Wang A and Dennis EA: Mammalian lysophospholipases. Biochim Biophys Acta. 1439:1–16. 1999. View Article : Google Scholar | |
|
Peyrard M, Fransson I, Xie YG, et al: Characterization of a new member of the human beta-adaptin gene family from chromosome 22q12, a candidate meningioma gene. Hum Mol Genet. 3:1393–1399. 1994. View Article : Google Scholar : PubMed/NCBI | |
|
Ferrara JL, Levine JE, Reddy P and Holler E: Graft-versus-host disease. Lancet. 373:1550–1561. 2009. View Article : Google Scholar : PubMed/NCBI | |
|
Weetman AP and McGregor AM: Autoimmune thyroid disease: further developments in our understanding. Endocr Rev. 15:788–830. 1994.PubMed/NCBI | |
|
López-Franco O, Hernández-Vargas P, Ortiz-Muñoz G, et al: Parthenolide modulates the NF-kappaB-mediated inflammatory responses in experimental atherosclerosis. Arterioscler Thromb Vasc Biol. 26:1864–1870. 2006.PubMed/NCBI | |
|
Yip KH, Zheng MH, Feng HT, Steer JH, Joyce DA and Xu J: Sesquiterpene lactone parthenolide blocks lipopolysaccharide-induced osteolysis through the suppression of NF-kappaB activity. J Bone Miner Res. 19:1905–1916. 2004. View Article : Google Scholar : PubMed/NCBI | |
|
Guzman ML, Rossi RM, Karnischky L, et al: The sesquiterpene lactone parthenolide induces apoptosis of human acute myelogenous leukemia stem and progenitor cells. Blood. 105:4163–4169. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Miglietta A, Bozzo F, Gabriel L and Bocca C: Microtubule-interfering activity of parthenolide. Chem Biol Interact. 149:165–173. 2004. View Article : Google Scholar : PubMed/NCBI | |
|
Feltenstein MW, Schuhly W, Warnick JE, Fischer NH and Sufka KJ: Anti-inflammatory and anti-hyperalgesic effects of sesquiterpene lactones from Magnolia and Bear’s foot. Pharmacol Biochem Behav. 79:299–302. 2004.PubMed/NCBI | |
|
Tiuman TS, Ueda-Nakamura T, Garcia Cortez DA, et al: Antileishmanial activity of parthenolide, a sesquiterpene lactone isolated from Tanacetum parthenium. Antimicrob Agents Chemother. 49:176–182. 2005. View Article : Google Scholar : PubMed/NCBI | |
|
Leost M, Schultz C, Link A, et al: Paullones are potent inhibitors of glycogen synthase kinase-3beta and cyclin-dependent kinase 5/p25. Eur J Biochem. 267:5983–5994. 2000. View Article : Google Scholar : PubMed/NCBI | |
|
Lahusen T, De Siervi A, Kunick C and Senderowicz AM: Alsterpaullone, a novel cyclin-dependent kinase inhibitor, induces apoptosis by activation of caspase-9 due to perturbation in mitochondrial membrane potential. Mol Carcinog. 36:183–194. 2003. View Article : Google Scholar | |
|
Hoppener JW, de Pagter-Holthuizen P, Geurts van Kessel AH, et al: The human gene encoding insulin-like growth factor I is located on chromosome 12. Hum Genet. 69:157–160. 1985. View Article : Google Scholar : PubMed/NCBI | |
|
Jansen M, van Schaik FM, Ricker AT, et al: Sequence of cDNA encoding human insulin-like growth factor I precursor. Nature. 306:609–611. 1983. View Article : Google Scholar : PubMed/NCBI | |
|
Studer RK: Nitric oxide decreases IGF-1 receptor function in vitro; glutathione depletion enhances this effect in vivo. Osteoarthritis Cartilage. 12:863–869. 2004. View Article : Google Scholar : PubMed/NCBI | |
|
Retief E, Parker MI and Retief AE: Regional chromosome mapping of human collagen genes alpha 2(I) and alpha 1(I) (COLIA2 and COLIA1). Hum Genet. 69:304–308. 1985. View Article : Google Scholar : PubMed/NCBI | |
|
Wenstrup RJ, Cohn DH, Cohen T and Byers PH: Arginine for glycine substitution in the triple-helical domain of the products of one alpha 2(I) collagen allele (COL1A2) produces the osteogenesis imperfecta type IV phenotype. J Biol Chem. 263:7734–7740. 1988. | |
|
Kikuno R, Nagase T, Ishikawa K, et al: Prediction of the coding sequences of unidentified human genes. XIV The complete sequences of 100 new cDNA clones from brain which code for large proteins in vitro. DNA Res. 6:197–205. 1999. View Article : Google Scholar | |
|
FitzPatrick DR, Carr IM, McLaren L, et al: Identification of SATB2 as the cleft palate gene on 2q32-q33. Hum Mol Genet. 12:2491–2501. 2003. View Article : Google Scholar : PubMed/NCBI |
