Abstract
Osteoarthritis (OA) is the most common age-related arthritic disorder and is characterized by aberrant extracellular matrix (ECM) content and surface disruptions that range from fibrillation, clefting and delamination, leading to articular surface erosion. Worldwide, over 20% of the population is affected with OA and 80% of these patients have limitations in movement, whereas 25% experience inhibition in major daily activities of life. OA is the most common disabling arthritic disease; nevertheless, no disease-modifying treatment is available except for the expensive total joint replacement surgery at end-stage disease. Lack of insight into the underlying pathophysiological mechanisms of OA has considerably contributed to the inability of the scientific community to develop disease-modifying drugs. To overcome this critical barrier, focus should be on translation of identified robust gene deviations towards the underlying biological mechanisms.
Osteoarthritis disease background
Osteoarthritis (OA) is a prevalent, complex, disabling disease affecting the articular joints. Age, genetic factors and mechanical stress are some of the risk factors of OA. Currently, there is no adequate therapy to reverse or slow down the disease. Analgesia, physiotherapy and, in severe forms of the disease, joint replacement are the main treatment options. As a result, OA has a large and detrimental impact on the quality of life of elderly individuals and this increased morbidity contributes indirectly to an increased mortality .
Comprehensive genome-wide searches for genetic variants conferring risk for OA have resulted in robust genome-wide significant signals . Functional follow-up studies, to show biological relevance, however, have only recently started to be implemented as a natural next step. Consequently, little progress has been made in clinical translation of these findings, let alone identification of novel evidence-based treatment options or disease-modifying OA drugs. This chapter provides a brief summary of the field of OA genomics and suggested common underlying pathways. Subsequently, an outline of proposed functional genomic research, including human joint tissues and three-dimensional (3D) in vitro cell and in vivo animal models, is provided and is exemplified by work on the susceptibility genedeiodinase iodothyronine type-2 (D2) gene ( DIO2 ).
OA pathophysiology: role of dynamic regulation of gene expression
Articular cartilage is a connective tissue that covers the ends of long bones. It has a smooth, wear-resistant lubricated surface that allows bones to glide over one another with minimal friction, yet eligible to absorb impact forces. Chondrocytes in articular cartilage reside in a maturational arrested state without detectable proliferation and at a low metabolic activity . Nonetheless, articular chondrocytes maintain cartilage tissue homeostasis throughout life and as such need to apply remodelling of extracellular matrix (ECM) upon stresses and micro-traumas . To secure cartilage integrity and cope with these challenges, the chondrocyte is required to continuously and dynamically adjust expression of catabolic and anabolic genes, while maintaining its capacity to restore its maturational arrested steady-state phenotype . This plasticity is likely maintained by virtue of epigenetic control mechanisms . Multiple epigenetic layers exist, such as histone modifications, microRNAs and long non-coding RNAs. The most studied epigenetic mechanism is DNA methylation, the phenomenon in which the cytosine nucleotide that is located in cytosine–guanine (CpG) residue pairs gets methylated. The amount of CpG methylation facilitates dynamical adjustment of gene expression in a very context-specific manner and is directed by cellular requirements upon environmental exposures. Hence, epigenetic regulation of gene expression has been found to be an important mechanism by which the maturational arrested articular chondrocytes are able to appropriately adapt to a changing environment encountered due to, for example, biomechanical stresses yet allowing securing a steady-state tissue homeostasis . Genetic variations and stochastic events during ageing have, however, been shown to cause changes in epigenetic marks resulting in subtle differential (allelic) gene expression. Moreover, such changes have frequently been found to be associated with pathological conditions , including OA . Given that chondrocytes in OA-affected cartilage have recapitulated a growth plate morphology and exhibit an articular cartilage-debilitating expression profile similarly to that occurring at end-stage endochondral ossification , it was hypothesized that deviations in epigenetically controlled gene expression could be an underlying mechanism of OA pathophysiology .
Growth plate endochondral ossification is a developmental process of growth and bone formation. Herein, differentiation of chondrocytes results in the formation of a complex, longitudinally organized cartilage template which following elongation through chondrocyte proliferation will be replaced by bone. To allow gradual replacement of cartilage to bone matrix, growth plate chondrocytes become hypertrophic and commence terminal differentiation, followed by mineralization of the cartilage, apoptosis of the chondrocytes and vascular invasion . In this process, local, intracellular active thyroid hormone (T3) is known to be essential in signalling terminal maturation of hypertrophic chondrocytes . As such, endochondral ossification is essential during skeletal development; however, it is considered deleterious for mature articular cartilage and respective joint function ( Fig. 1 ).
Genomic profiling of OA cartilage
In an attempt to acquire insight into the complex mechanisms that mark OA in joint tissues, methylome-wide studies of articular cartilage in OA have revealed considerable heterogeneity in disease processes and numerous differentially methylated loci between healthy and diseased tissue . Moreover, to assess underlying aberrant dynamic transcriptional processes that occur in OA , genome-wide interactions between the transcriptome and methylome of OA joint tissue samples were investigated and indicated, among differential genes, significant enrichment for those that particularly act within developmental pathways . Being a selection of genes that are epigenetically regulated by OA responsive methylation, these genes may possibly reflect the reactivation of endochondral ossification pathways . Such data represent an important prelude to strategies that aim to target aberrant processes of articular chondrocytes and develop necessary novel therapeutic options . By additionally analysing genetic variants proximal to these cartilage-specific OA relevant epigenetically regulated genes, single nucleotide polymorphisms (SNPs) were identified that likely affect such dynamic transcriptional processes occurring with OA pathophysiology . Ideally, these functionally relevant DNA variations should now be followed up in large genome-wide association consortia to assess and confirm their ability to confer risk to common OA.
Genetic susceptibility to OA necessitates functional follow-up studies
Common underlying pathway
OA has a considerable and complex genetic component ; many single nucleotide polymorphisms (SNPs) with small effects are expected to influence the onset and course of the disease . Genetic studies aiming to identify the underlying genes have, despite the complexity, consistently provided evidence for genes regulating the maturational process of growth plate chondrocytes, and as of such coordinate formation of cartilage and eventually bone during endochondral ossification as elegantly reviewed previously .
By addressing functionality of the identified OA risk alleles, it was shown, for example, for GDF5 , ALDH1A2 , NCOA3 , GNL3 and SPCS1 and DIO2 , that they modulate OA pathology due to altered transcription of a positional gene in cis also referred to as allelic imbalance. It is, therefore, acknowledged that allelic imbalanced expression of OA risk alleles may play a substantial role in OA susceptibility , as is also thought true for complex traits in general .
OA relevant functional follow-up studies
To allow for functional follow-up studies, such as addressing allelic imbalances of risk alleles, biobanks with multiple disease-relevant tissues tailored to allow histology and isolation of RNA, DNA and proteins are necessary. Hence, in OA, such tissues are readily collected from discarded surgical material of patients who undergo a joint replacement surgery as result of end-stage disease . In addition, isolated cells (e.g. bone marrow-derived stem cells, chondrocytes and bone cells) could readily be employed for in vitro culture experiments, allowing for functional validation and causal inference of the identified susceptibility genes. For example, introducing the identified genetic variant in chondrocyte cultures and/or mechanically perturbing the cultured cells might reveal the actual mechanism of action that is asserted by the respective genes . Moreover, animal models are particularly suitable to effectively study the initiation and progression of OA and improve our understanding of the molecular mechanisms underlying genetic association, driving the joint pathology. As most of these OA animal models imply acute injury or inflammation, mechanical loading by forced exercise was recently proposed as a physiological model of disease induction and progression . Along these lines, we have previously outlined a pipeline able to effectively unravel the underlying mechanisms of OA susceptibility alleles . In the following paragraphs, the effectivity of the pipeline is exemplified by the functional follow-up studies applied to the OA susceptibility gene DIO2 .
Genetic susceptibility to OA necessitates functional follow-up studies
Common underlying pathway
OA has a considerable and complex genetic component ; many single nucleotide polymorphisms (SNPs) with small effects are expected to influence the onset and course of the disease . Genetic studies aiming to identify the underlying genes have, despite the complexity, consistently provided evidence for genes regulating the maturational process of growth plate chondrocytes, and as of such coordinate formation of cartilage and eventually bone during endochondral ossification as elegantly reviewed previously .
By addressing functionality of the identified OA risk alleles, it was shown, for example, for GDF5 , ALDH1A2 , NCOA3 , GNL3 and SPCS1 and DIO2 , that they modulate OA pathology due to altered transcription of a positional gene in cis also referred to as allelic imbalance. It is, therefore, acknowledged that allelic imbalanced expression of OA risk alleles may play a substantial role in OA susceptibility , as is also thought true for complex traits in general .
OA relevant functional follow-up studies
To allow for functional follow-up studies, such as addressing allelic imbalances of risk alleles, biobanks with multiple disease-relevant tissues tailored to allow histology and isolation of RNA, DNA and proteins are necessary. Hence, in OA, such tissues are readily collected from discarded surgical material of patients who undergo a joint replacement surgery as result of end-stage disease . In addition, isolated cells (e.g. bone marrow-derived stem cells, chondrocytes and bone cells) could readily be employed for in vitro culture experiments, allowing for functional validation and causal inference of the identified susceptibility genes. For example, introducing the identified genetic variant in chondrocyte cultures and/or mechanically perturbing the cultured cells might reveal the actual mechanism of action that is asserted by the respective genes . Moreover, animal models are particularly suitable to effectively study the initiation and progression of OA and improve our understanding of the molecular mechanisms underlying genetic association, driving the joint pathology. As most of these OA animal models imply acute injury or inflammation, mechanical loading by forced exercise was recently proposed as a physiological model of disease induction and progression . Along these lines, we have previously outlined a pipeline able to effectively unravel the underlying mechanisms of OA susceptibility alleles . In the following paragraphs, the effectivity of the pipeline is exemplified by the functional follow-up studies applied to the OA susceptibility gene DIO2 .
Translating genomics into mechanisms of disease: DIO2 as example
DIO2 as OA susceptibility gene
The OA susceptibility gene DIO2 was identified by applying genome-wide linkage analyses and replicated by association of the risk allele C of the nonsynonymous single nucleotide polymorphism (SNP) rs225014 located in the coding region of DIO2 . The protein product of the gene, D2, is an enzyme that enhances intracellular thyroid (T3) bio-availability in specific target tissues such as the growth plate where it signals breakdown and mineralization of the cartilage to allow transition to bone . Based on this function, it was hypothesized that DIO2 might confer risk to OA either by affecting early the process of endochondral ossification or late, the propensity of maturational arrested articular chondrocytes to recuperate growth plate morphology against environmental challenges . Moreover, the identification of DIO2 , for the first time, underscored the significant role of local thyroid hormone in the aetiology of symptomatic OA. More recently, large-scale genome-wide association efforts indicated that additional genes involved in T3-signalling confer consistent risk to OA, such as NCOA3 and ALDH1A2 ( Fig. 2 ).
Effect of DIO2 on hip-shaped morphology
Being essential for the skeletal developmental process of endochondral ossification, it was investigated whether the DIO2 risk alleles confer risk to OA via suboptimal shape of the bones in the joint which could, consequently, lead to recurrent damage of the cartilage and eventually triggering the OA onset . To address this possible early effect of DIO2 on OA susceptibility, hip shapes among healthy and OA-affected individuals were quantified by statistical shape models (SSM) . The study showed that the DIO2 OA risk allele did not directly affect hip geometry; however, the DIO2 OA risk allele affected cartilage structure or metabolism in such a way that the cartilage may become more vulnerable to OA as result of a sub-optimal hip morphology and, respectively, induced biomechanical stresses .
Transcription of DIO2 in articular cartilage
To assess the effect of the identified DIO2 risk allele C of SNP rs225014, allelic imbalanced (AI) mRNA expression in joint tissues was explored among heterozygotes of this coding SNP which showed that the OA risk allele C was highly significant and consistently more abundantly present than the wild-type allele T. The consistency of AI marked by rs225014 indicates that either the polymorphism itself or a polymorphism in strong linkage disequilibrium is causing the effect.
Multiple microarray expression studies of OA affected as compared to healthy articular cartilage in both mice and humans demonstrated significant upregulation of DIO2 expression in OA-affected cartilage as well as a marked upregulation of D2 protein . These DIO2 expression patterns, together with other transcriptomic gene networks, mark a hypertrophic state of OA-affected articular cartilage . Nevertheless, OA-specific upregulation of DIO2 was reported in lesioned as well as in macroscopically preserved articular cartilage from OA-affected joints, thereby indicating that DIO2 -expressing chondrocytes do not necessarily mark cartilage destruction . Upon DIO2 induction, additional perturbation appears necessary to prime articular chondrocytes towards hypertrophy and terminal maturation with respective breakdown of cartilage. Given that DIO2 -associated articular cartilage destruction in humans occurs particularly at loading hotspots of joints , mechanical overloading may be the additional trigger necessary to engage an OA-like state.
Epigenetics of DIO2
In order to elucidate the possible influence of epigenetic modulation of DIO2 gene expression in OA-affected cartilage, mRNA expression as well as methylation status of several CpG dinucleotides across the DIO2 locus was compared between preserved and OA-affected cartilage of patients undergoing total arthroplasty of the knee or hip . Methylation at a single CpG dinucleotide, 2031 base pairs upstream of the DIO2 transcription start site, appeared highly sensitive to the ongoing OA process and showed positive association with DIO2 expression. Moreover, differential methylation at this particular CpG dinucleotide correlated with an enhanced up-regulation of DIO2 expression, primarily among rs225014 risk allele carriers. The regulatory properties of DNA methylation on DIO2 expression was subsequently confirmed by applying 5-aza-2′-deoxycytidine (AZA) treatment eliciting general demethylation of the DNA, concomitant with a decrease in methylation at CpG-2031 and DIO2 expression . Given that the CpG-2031 maps within an active CCCTC-transcription factor (CTCF)-binding site, it was hypothesized that methylation-dependent binding of CTCF could act as a positional isolator of DIO2 expression ( Fig. 3 ).
Related posts:
HLA and anti-citrullinated protein antibodies: Building blocks in RA
Refine, reduce, replace: Imaging of fibrosis and arthritis in animal models
Lipid mediators of inflammation in rheumatoid arthritis and osteoarthritis
New biological therapies in Sjögren’s syndrome
Metabolomics to identify biomarkers and as a predictive tool in inflammatory diseases
Corrigendum to “Risk of infection with biologic antirheumatic therapies in patients with rheumatoid arthritis” [Best Pract Res Clin Rheumatol 29 (2) (2015) 290–305]
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
