Epigenetics in the pathogenesis of rheumatoid arthritis
BMC Medicinevolume 12, Article number: 35 (2014)
An increasing number of studies show that besides the inherited genetic architecture (that is, genomic DNA), various environmental factors significantly contribute to the etiology of rheumatoid arthritis. Epigenetic factors react to external stimuli and form bridges between the environment and the genetic information-harboring DNA. Epigenetic mechanisms are implicated in the final interpretation of the encoded genetic information by regulating gene expression, and alterations in their profile influence the activity of the immune system. Overall, epigenetic mechanisms further increase the well-known complexity of rheumatoid arthritis by providing additional subtle contributions to rheumatoid arthritis susceptibility. Although there are controversies regarding the involvement of epigenetic and genetic factors in rheumatoid arthritis etiology, it is becoming obvious that the two systems (genetic and epigenetic) interact with each other and are ultimately responsible for rheumatoid arthritis development. Here, epigenetic factors and mechanisms involved in rheumatoid arthritis are reviewed and new, potential therapeutic targets are discussed.
More than 10 years after the completion of the human genome sequencing project  and numerous genome-wide association studies (GWAS) , we still do not fully understand the genetic basis of rheumatoid arthritis (RA). GWAS on patients with RA revealed more than 30 genomic risk loci, but identification of disease-promoting genes and their functional characterization remain to be accomplished [3, 4]. The delayed progress in RA genetics can be explained by the polygenic nature of the disease, the enormous genetic heterogeneity of the human population, and the difficulties with the interpretation of GWAS data since most of the significant genetic alterations (that is, mutations) are located in non-protein coding regions of the genome. Another observation that raises some doubt about a major role of genetic factors in RA pathogenesis is that the concordance rate in monozygotic twins is only approximately 15% . However, twin studies drew attention to the importance of epigenetic factors that mediate interactions between the genes and the environment [6–8].
In this commentary, we will first introduce the basic epigenetic mechanisms and then discuss the results of RA-related epigenetic studies. Finally, we will provide a brief description of epigenetic factor-based future therapeutics in RA.
Although there is no ‘carved-in-stone’ definition for epigenetics, it is broadly defined as the study of heritable changes in gene activity that do not involve any alterations in the primary DNA sequence . Epigenetics originally focused on DNA methylation and various histone modifications, but recently expanded to the field of non-coding RNAs. Ab ovo, each cell of the body inherits the same genetic information. What makes each cell unique is that, during ontogenesis, different sets of genes are turned on and off. Epigenetic mechanisms establish the proper nuclear milieu for cell-specific gene expression and are responsible for the cellular memory, that is, keeping and transmitting cell-specific gene expression patterns to daughter cells. Epigenetic factors can deposit, interpret and eliminate epigenetic information and, in this sense, they can be divided into distinct functional groups: epigenetic ‘writers’ or enzymes that modify DNA and histones; epigenetic ‘readers’ with specific protein domains that recognize DNA or histone marks; and epigenetic ‘erasers’ that can delete the existing signals to make room for new modifications (Figure 1A).
In studies on cancer and inflammatory and metabolic disorders, frequent errors have been found in epigenetic mechanisms that can result in the miswriting, misreading or faulty removal of epigenetic signals .
DNA methylation is catalyzed by DNA methyltransferases (writers) and associated with gene silencing . DNA methylation readers are the methyl-CpG-binding domain proteins, which promote gene silencing by recruiting histone modifiers. Erasers of DNA methylation have been enigmatic for a long time, but recent studies have revealed that demethylation proceeds via selective oxidation of methylated cytosine residues, which is catalyzed by members of the ten-eleven translocation protein family [12, 13]. Genomic DNA and associated special nuclear proteins (histones) comprise the nucleosomes that are the building blocks of eukaryotic chromatin and the primary targets of epigenetic modifiers . We briefly describe the two best characterized post-transcriptional modifications because they have been already implicated in RA.
Histone acetylation and methylation exert their effects on gene expression by regulating the accessibility of DNA for transcription factors. As a general rule, modifications decrease the compactness of chromatin structure and promote gene expression (Figure 1B) . Histone acetylation in any position favors transcription activation. Writers are histone acetyltranferases (HATs), erasers are histone deacetylases (HDACs), and bromodomain-containing proteins are the readers of this type of histone modification. Histone methylation represents a diverse set of epigenetic signals  for at least three reasons: first, it can occur on various residues (lysine or arginine); second, it exerts its effect on transcription by determining the degree of methylation (that is, mono-, di- or trimethylation); and third, depending on the location of the modified residue, histone methylation can either positively or negatively affect gene expression. Histone methyltransferases, histone demethylases and chromo-, Tudor- or plant homeodomain-containing proteins are the writers, erasers and readers of this type of post-transcriptional modification, respectively (Figure 1A). Different chromatin modifications act together and a highly specific combination of various post-transcriptional modifications creates the histone code that ultimately determines the transcriptional status of a gene .
Unlike genomic DNA (that is, genome), epigenetic signals (that is, epigenome) are highly dynamic and show cell type-specific patterns. Each type of cell has its own characteristic epigenome profile with unique gene expression patterns; therefore, studies must be highly specific regarding the investigated cell type.
Epigenetic alteration in rheumatoid arthritis synovial cells
Early studies found widespread DNA hypomethylation in RA synovial fibroblasts, including hypomethylation of the promoter of the CXCL12 gene  and the LINE1 retrotransposons  that are repetitive elements normally repressed by DNA methylation. In these cases, loss of the repressive DNA methylation signal results in increased gene expression. A recent genome-wide study on RA synovial fibroblasts revealed a number of differentially (hypo- and hyper-) methylated genomic regions . Most of the affected genes appear to be involved in inflammation, matrix remodeling, leukocyte recruitment and immune responses . Another study found that the HAT to HDAC activity ratio in arthritic joints was shifted towards HAT dominance, favoring histone acetylation , ultimately leading to an increase in gene transcription.
Epigenetic changes of the adaptive immune system
A genome-wide DNA methylation profiling study in peripheral blood mononuclear cells reported differentially methylated regions in the major histocompatibility complex loci that make a significant contribution to the genetic risk of developing RA . Our group performed the first study on arthritis-related epigenetic modifiers , in which chromatin-modifying enzymes were analyzed in B and T cells from arthritic mice and peripheral blood mononuclear cells from patients with RA. All chromatin-modifying enzyme families were represented in the repertoire of genes with arthritis-specific expression, including histone kinases, acetyltransferases, deacetylases, methyltransferases and demethylases, as well as ubiquitin ligases. The most strongly upregulated genes were those encoding Aurora kinase (A and B) enzymes in both arthritic animal and human lymphocytes, and this was accompanied by phosphorylation of serine 10 in the tail of histone H3. This type of histone phosphorylation is a pivotal epigenetic signal for the recruitment of the transcription factor nuclear factor-kappaB (NF-κB) to the promoter of cytokine genes , resulting in a cytokine-driven pro-inflammatory response. We found that VX-680, an Aurora kinase-specific inhibitor, significantly reduced the severity of arthritis and promoted B cell apoptosis in the proteoglycan-induced arthritis (PGIA) model of RA. The significance of VX-680-induced B cell apoptosis is that patients with RA who do not respond to anti-tumor necrosis factor therapy are frequently treated with a monoclonal anti-CD20 antibody to eliminate autoantibody-producing B cells . Our findings suggest that drug (VX-680)-induced B cell depletion may provide an alternative to the CD20 antibody-based therapy.
In addition to Aurora kinases, several members of the HAT family are also significantly upregulated in arthritic mice and patients with RA, with the gene encoding Esco2 showing the strongest increase in expression. Esco2 is thought to be required for the establishment of sister chromatid cohesion and it also couples cohesion and DNA replication to ensure that only sister chromatids are paired together [23, 24]. Because Esco2 belongs to the HAT family of epigenetic modifiers, it is reasonable to assume that it acts as a selective activator of certain target genes. Anacardic acid (ACA) inhibits HATs  and indirectly suppresses NF-κB activation . We tested the therapeutic potential of ACA in mice with established PGIA. Mice treated with ACA displayed significantly reduced arthritis progression as compared to untreated control animals (unpublished observations; Figure 2).
As described earlier, many of the epigenome modifiers can directly or indirectly affect the activity of NF-κB, a master regulator of the transcription of inflammation-related genes. With regard to autoimmune or inflammatory diseases such as RA, the emerging consensus is that epigenetic factors (enzymes) supporting repressive signals are downregulated, whereas those that promote transcription are upregulated. A combination of these activities in immune cells ultimately results in the strengthening of pro-inflammatory pathways and the weakening of anti-inflammatory mechanisms. For example, disease-linked expression of KDM6B, a histone methyltransferase responsible for eliminating a repressive epigenetic signal (that is, histone H3 K27 trimethylation), is involved in macrophage activation , and repression of the SETD6 gene, which encodes a known negative regulator of NF-κB, leads to runaway activity of this transcription factor .
The results of epigenetic studies in RA raise the question whether the reported epigenetic alterations play a causative role or are the consequences of other pathologic processes that take place in RA. To answer this question, there is a need for further epigenome-wide studies on all types of cells involved in RA, exploration of a larger repertoire of epigenetic signals, and investigation of the epigenetic landscape at different phases of arthritis. It is possible that significant advances will be achieved in the near future because the technologies and model systems, including genome and epigenome-wide analysis tools (such as whole-genome sequencing, chromatin immunoprecipitation sequencing and RNA sequencing) and animal models, are readily available.
Information from RA-associated epigenetic studies can be useful for diagnostic and therapeutic purposes because investigation of the epigenetic landscape can provide both potential biomarkers and therapeutic targets. There have been numerous clinical trials involving patients with cancer that have tested such inhibitors as therapeutics against malignancies . Although we have demonstrated the beneficial effect of specific Aurora kinase and HAT inhibitors , and HDAC inhibitors have been tested by other groups  in preclinical studies, unlike in the cancer field, there is still no epigenetics-based drug on the market of RA therapeutics.
A common outcome of genetic and epigenetic mutations is that both ultimately lead to aberrant gene expression. The mechanisms by which genetic mutations affect gene expression are well known, including shorter or longer deletions, insertions, inversions, translocations, or single nucleotide changes within transcription factor binding sites. Mutations hitting genes that encode epigenetic regulators may result in aberrant expression or functional impairment of the affected epigenetic factors [31–33]. The connection between epigenetically provoked and epigenetics-independent genetic mutations is not obvious and is currently under investigation. Both DNA hyper- and hypomethylation can trigger genetic mutations. DNA hypermethylation-mediated silencing of DNA repair genes (for example, MGMT and MLH1) can result in inactivation of cellular mechanisms responsible for keeping the genetic mutation rate low [34, 35] or in induction of microsatellite instability as described in certain types of cancer [36, 37]. DNA hypomethylation can reactivate retrotransposons (for example, long and short interspersed nuclear elements), which then promote genetic mutations by inserting extra nucleotides into the exons or regulatory regions of genes [38, 39].
Alteration in epigenetic mechanisms can trigger genetic mutations and genetic mutations in epigenetic regulators can lead to an altered epigenetic profile. Therefore, genetics and epigenetics can be considered two sides of the same coin, as has been established in the field of cancer research . It is very likely that in the near future the same conclusion will be reached regarding autoimmune diseases such as RA.
TTG and KM are Professors at Rush University Medical Center, and founding members of the Section of the Molecular Medicine. They have been studying immunological aspects of rheumatoid arthritis and ankylosing spondylitis in patients and corresponding animal models for more than three decades. They first described cartilage proteoglycan/aggrecan-induced arthritis (PGIA) and spondylitis (PGISpA) in genetically susceptible mice, and this pioneering work was honored by the Carol Nachman Price. TAR is an Assistant Professor at Rush University Medical Center. He is an expert in disease-associated epigenetic modifications of DNA and histones in cancer, and most recently, in rheumatoid arthritis.
genome-wide association studies
Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Baldwin J, Devon K, Dewar K, Doyle M, FitzHugh W, Funke R, Gage D, Harris K, Heaford A, Howland J, Kann L, Lehoczky J, LeVine R, McEwan P, McKernan K, Meldrim J, Mesirov JP, Miranda C, Morris W, Naylor J, Raymond C, Rosetti M, Santos R, Sheridan A, Sougnez C, et al: Initial sequencing and analysis of the human genome. Nature. 2001, 409: 860-921. 10.1038/35057062.
Manolio TA: Genomewide association studies and assessment of the risk of disease. N Engl J Med. 2010, 363: 166-176. 10.1056/NEJMra0905980.
Kurko J, Besenyei T, Laki J, Glant TT, Mikecz K, Szekanecz Z: Genetics of rheumatoid arthritis - a comprehensive review. Clin Rev Allergy Immunol. 2013, 45: 170-179. 10.1007/s12016-012-8346-7.
Lessard CJ, Ice JA, Adrianto I, Wiley GB, Kelly JA, Gaffney PM, Montgomery CG, Moser KL: The genomics of autoimmune disease in the era of genome-wide association studies and beyond. Autoimmun Rev. 2012, 11: 267-275. 10.1016/j.autrev.2011.10.003.
Silman AJ, MacGregor AJ, Thomson W, Holligan S, Carthy D, Farhan A, Ollier WE: Twin concordance rates for rheumatoid arthritis: results from a nationwide study. Br J Rheumatol. 1993, 32: 903-907. 10.1093/rheumatology/32.10.903.
Bell JT, Spector TD: A twin approach to unraveling epigenetics. Trends Genet. 2011, 27: 116-125. 10.1016/j.tig.2010.12.005.
Karlson EW, Deane K: Environmental and gene-environment interactions and risk of rheumatoid arthritis. Rheum Dis Clin North Am. 2012, 38: 405-426. 10.1016/j.rdc.2012.04.002.
Costenbader KH, Gay S, Alarcon-Riquelme ME, Iaccarino L, Doria A: Genes, epigenetic regulation and environmental factors: which is the most relevant in developing autoimmune diseases?. Autoimmun Rev. 2012, 11: 604-609. 10.1016/j.autrev.2011.10.022.
Virani S, Colacino JA, Kim JH, Rozek LS: Cancer epigenetics: a brief review. ILAR J. 2012, 53: 359-369. 10.1093/ilar.53.3-4.359.
Zhou VW, Goren A, Bernstein BE: Charting histone modifications and the functional organization of mammalian genomes. Nat Rev Genet. 2011, 12: 7-18.
Jones PA: Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nat Rev Genet. 2012, 13: 484-492. 10.1038/nrg3230.
Tan L, Shi YG: Tet family proteins and 5-hydroxymethylcytosine in development and disease. Development. 2012, 139: 1895-1902. 10.1242/dev.070771.
Bergman Y, Cedar H: DNA methylation dynamics in health and disease. Nat Struct Mol Biol. 2013, 20: 274-281. 10.1038/nsmb.2518.
Bannister AJ, Kouzarides T: Regulation of chromatin by histone modifications. Cell Res. 2011, 21: 381-395. 10.1038/cr.2011.22.
Karouzakis E, Rengel Y, Jungel A, Kolling C, Gay RE, Michel BA, Tak PP, Gay S, Neidhart M, Ospelt C: DNA methylation regulates the expression of CXCL12 in rheumatoid arthritis synovial fibroblasts. Genes Immun. 2011, 12: 643-652. 10.1038/gene.2011.45.
Neidhart M, Rethage J, Kuchen S, Kunzler P, Crowl RM, Billingham ME, Gay RE, Gay S: Retrotransposable L1 elements expressed in rheumatoid arthritis synovial tissue: association with genomic DNA hypomethylation and influence on gene expression. Arthritis Rheum. 2000, 43: 2634-2647. 10.1002/1529-0131(200012)43:12<2634::AID-ANR3>3.0.CO;2-1.
Nakano K, Whitaker JW, Boyle DL, Wang W, Firestein GS: DNA methylome signature in rheumatoid arthritis. Ann Rheum Dis. 2013, 72: 110-117. 10.1136/annrheumdis-2012-201526.
Huber LC, Brock M, Hemmatazad H, Giger OT, Moritz F, Trenkmann M, Distler JH, Gay RE, Kolling C, Moch H, Michel BA, Gay S, Distler O, Jungel A: Histone deacetylase/acetylase activity in total synovial tissue derived from rheumatoid arthritis and osteoarthritis patients. Arthritis Rheum. 2007, 56: 1087-1093. 10.1002/art.22512.
Liu Y, Aryee MJ, Padyukov L, Fallin MD, Hesselberg E, Runarsson A, Reinius L, Acevedo N, Taub M, Ronninger M, Shchetynsky K, Scheynius A, Kere J, Alfredsson L, Klareskog L, Ekstrom TJ, Feinberg AP: Epigenome-wide association data implicate DNA methylation as an intermediary of genetic risk in rheumatoid arthritis. Nat Biotechnol. 2013, 31: 142-147. 10.1038/nbt.2487.
Glant TT, Besenyei T, Kadar A, Kurko J, Tryniszewska B, Gal J, Soos G, Szekanecz Z, Hoffmann G, Block JA, Katz RS, Mikecz K, Rauch TA: Differentially expressed epigenome modifiers, including aurora kinases a and B, in immune cells in rheumatoid arthritis in humans and mouse models. Arthritis Rheum. 2013, 65: 1725-1735. 10.1002/art.37986.
Wolter S, Doerrie A, Weber A, Schneider H, Hoffmann E, von der OJ, Bakiri L, Wagner EF, Resch K, Kracht M: c-Jun controls histone modifications, NF-kappaB recruitment, and RNA polymerase II function to activate the ccl2 gene. Mol Cell Biol. 2008, 28: 4407-4423. 10.1128/MCB.00535-07.
Leandro MJ, Becerra-Fernandez E: B-cell therapies in established rheumatoid arthritis. Best Pract Res Clin Rheumatol. 2011, 25: 535-548.
Kim BJ, Kang KM, Jung SY, Choi HK, Seo JH, Chae JH, Cho EJ, Youn HD, Qin J, Kim ST: Esco2 is a novel corepressor that associates with various chromatin modifying enzymes. Biochem Biophys Res Commun. 2008, 372: 298-304. 10.1016/j.bbrc.2008.05.056.
Vega H, Waisfisz Q, Gordillo M, Sakai N, Yanagihara I, Yamada M, van GD, Kayserili H, Xu C, Ozono K, Jabs EW, Inui K, Joenje H: Roberts syndrome is caused by mutations in ESCO2, a human homolog of yeast ECO1 that is essential for the establishment of sister chromatid cohesion. Nat Genet. 2005, 37: 468-470. 10.1038/ng1548.
Balasubramanyam K, Swaminathan V, Ranganathan A, Kundu TK: Small molecule modulators of histone acetyltransferase p300. J Biol Chem. 2003, 278: 19134-19140. 10.1074/jbc.M301580200.
Sung B, Pandey MK, Ahn KS, Yi T, Chaturvedi MM, Liu M, Aggarwal BB: Anacardic acid (6-nonadecyl salicylic acid), an inhibitor of histone acetyltransferase, suppresses expression of nuclear factor-kappaB-regulated gene products involved in cell survival, proliferation, invasion, and inflammation through inhibition of the inhibitory subunit of nuclear factor-kappaBalpha kinase, leading to potentiation of apoptosis. Blood. 2008, 111: 4880-4891. 10.1182/blood-2007-10-117994.
Satoh T, Takeuchi O, Vandenbon A, Yasuda K, Tanaka Y, Kumagai Y, Miyake T, Matsushita K, Okazaki T, Saitoh T, Honma K, Matsuyama T, Yui K, Tsujimura T, Standley DM, Nakanishi K, Nakai K, Akira S: The Jmjd3-Irf4 axis regulates M2 macrophage polarization and host responses against helminth infection. Nat Immunol. 2010, 11: 936-944. 10.1038/ni.1920.
Levy D, Kuo AJ, Chang Y, Schaefer U, Kitson C, Cheung P, Espejo A, Zee BM, Liu CL, Tangsombatvisit S, Tennen RI, Kuo AY, Tanjing S, Cheung R, Chua KF, Utz PJ, Shi X, Prinjha RK, Lee K, Garcia BA, Bedford MT, Tarakhovsky A, Cheng X, Gozani O: Lysine methylation of the NF-kappaB subunit RelA by SETD6 couples activity of the histone methyltransferase GLP at chromatin to tonic repression of NF-kappaB signaling. Nat Immunol. 2011, 12: 29-36. 10.1038/ni.1968.
Stolfa DA, Einsle O, Sippl W, Jung M: Current trends in epigenetic drug discovery. Future Med Chem. 2012, 4: 2029-2037. 10.4155/fmc.12.144.
Grabiec AM, Reedquist KA: The ascent of acetylation in the epigenetics of rheumatoid arthritis. Nat Rev Rheumatol. 2013, 9: 311-318. 10.1038/nrrheum.2013.17.
Morin RD, Mendez-Lago M, Mungall AJ, Goya R, Mungall KL, Corbett RD, Johnson NA, Severson TM, Chiu R, Field M, Jackman S, Krzywinski M, Scott DW, Trinh DL, Tamura-Wells J, Li S, Firme MR, Rogic S, Griffith M, Chan S, Yakovenko O, Meyer IM, Zhao EY, Smailus D, Moksa M, Chittaranjan S, Rimsza L, Brooks-Wilson A, Spinelli JJ, Ben-Neriah S, et al: Frequent mutation of histone-modifying genes in non-Hodgkin lymphoma. Nature. 2011, 476: 298-303. 10.1038/nature10351.
Chase A, Cross NC: Aberrations of EZH2 in cancer. Clin Cancer Res. 2011, 17: 2613-2618. 10.1158/1078-0432.CCR-10-2156.
Yan XJ, Xu J, Gu ZH, Pan CM, Lu G, Shen Y, Shi JY, Zhu YM, Tang L, Zhang XW, Liang WX, Mi JQ, Song HD, Li KQ, Chen Z, Chen SJ: Exome sequencing identifies somatic mutations of DNA methyltransferase gene DNMT3A in acute monocytic leukemia. Nat Genet. 2011, 43: 309-315. 10.1038/ng.788.
Shen L, Kondo Y, Rosner GL, Xiao L, Hernandez NS, Vilaythong J, Houlihan PS, Krouse RS, Prasad AR, Einspahr JG, Buckmeier J, Alberts DS, Hamilton SR, Issa JP: MGMT promoter methylation and field defect in sporadic colorectal cancer. J Natl Cancer Inst. 2005, 97: 1330-1338. 10.1093/jnci/dji275.
Watanabe T, Katayama Y, Komine C, Yoshino A, Ogino A, Ohta T, Fukushima T: O6-methylguanine-DNA methyltransferase methylation and TP53 mutation in malignant astrocytomas and their relationships with clinical course. Int J Cancer. 2005, 113: 581-587. 10.1002/ijc.20625.
Fleisher AS, Esteller M, Wang S, Tamura G, Suzuki H, Yin J, Zou TT, Abraham JM, Kong D, Smolinski KN, Shi YQ, Rhyu MG, Powell SM, James SP, Wilson KT, Herman JG, Meltzer SJ: Hypermethylation of the hMLH1 gene promoter in human gastric cancers with microsatellite instability. Cancer Res. 1999, 59: 1090-1095.
Leung SY, Yuen ST, Chung LP, Chu KM, Chan AS, Ho JC: hMLH1 promoter methylation and lack of hMLH1 expression in sporadic gastric carcinomas with high-frequency microsatellite instability. Cancer Res. 1999, 59: 159-164.
Martinez-Garay I, Ballesta MJ, Oltra S, Orellana C, Palomeque A, Molto MD, Prieto F, Martinez F: Intronic L1 insertion and F268S, novel mutations in RPS6KA3 (RSK2) causing Coffin-Lowry syndrome. Clin Genet. 2003, 64: 491-496. 10.1046/j.1399-0004.2003.00166.x.
van den Hurk JA, van de Pol DJ, Wissinger B, van Driel MA, Hoefsloot LH, de Wijs IJ, van den Born LI, Heckenlively JR, Brunner HG, Zrenner E, Ropers HH, Cremers FP: Novel types of mutation in the choroideremia (CHM) gene: a full-length L1 insertion and an intronic mutation activating a cryptic exon. Hum Genet. 2003, 113: 268-275. 10.1007/s00439-003-0970-0.
You JS, Jones PA: Cancer genetics and epigenetics: two sides of the same coin?. Cancer Cell. 2012, 22: 9-20. 10.1016/j.ccr.2012.06.008.
We thank to our laboratory members for helpful discussion of the manuscript. We apologize to those whose work was not directly cited because of space constraints. This study was supported in part by NIH grants R01 AR059356, R01 AR062991 (TTG), R01 AR064958 (KM) and R21 AR064948 (TAR).
The authors declare that they have no competing interests.
All authors contributed to the manuscript. TAR wrote the first draft and all other authors amended the manuscript. All authors read and approved the final manuscript.