The clinical relevance of genetic susceptibility to osteoarthritis

Osteoarthritis is a major musculoskeletal cause of disability in the elderly, but current therapeutic approaches are insufficient to prevent initiation and progression of the disease. Genetic studies in humans have identified molecules involved in signalling cascades that are important for the pathology of the joint components. These include the bone morphogenetic protein (BMP) signalling, the wingless-type signalling and the thyroid pathway as well as apoptotic-related molecules. There is emerging evidence indicating that inflammatory molecules related to cytokine production, prostaglandin and arachidonic acid metabolism are also involved in susceptibility to osteoarthritis. All of these pathways are likely targets for pharmacological intervention. Genetic variation also affects pain due to osteoarthritis highlighting molecular mechanisms for pain relief. Moreover, combinations of genetic markers can be used to identify individuals at high risk of osteoarthritis and risk of total joint arthroplasty failure, which should facilitate the application of preventive and disease management strategies.

Osteoarthritis (OA) is the most common joint disorder and the leading cause of disability in the elderly in the US and Europe . Physician-diagnosed arthritis occurs in more than 50% of adults older than age 65 years and in more than 30% of adults aged 45–64 years .

The loss of articular cartilage is the hallmark of OA , but all the joint components, including the ligaments, tendons, capsule, synovial lining and periarticular bone, undergo structural and functional alterations during the course of OA progression . Although the pathogenesis of osteoarthritis is not fully understood, it is strongly age related, being rare before 40 years, and rising in frequency with age, such that a large proportion of people over the age of 70 have radiographic evidence of osteoarthritis in some joints . OA is a multifactorial disease with genetic and environmental determinants. All cases are probably affected by both genetics and environment, with a continuous distribution between the extremes of predominantly genetic or predominantly environmental causes . For example, the risk of post-traumatic OA after a meniscal injury of the knee is strongly affected by a family history of osteoarthritis, by the presence of nodal osteoarthritis of the hand (the classical marker of generalised osteoarthritis), by obesity and by sex .

How do we know that genetics is important in OA?

Evidence for a genetic predisposition to osteoarthritis was reported as early as the 1940s . The clustering of OA in families has been measured by using the risk ratio for a relative of an affected individual compared with the population prevalence . For affected sib pairs, this sib recurrence risk is termed the lambda sib (λs). Large values of λs indicate that a gene involved in traits should be easier to map than if λs is low . Thus, it is possible to identify subjects with clinical disease with severe enough symptoms to lead to total joint replacement (TJR) and to compare the prevalence of OA in their siblings (who have a genetic exposure) with that in controls who are matched as closely as possible to the siblings.

A study in Nottingham assessed the prevalence of hip OA in siblings of individuals undergoing total hip replacement (THR) to the prevalence of radiographic hip OA in subjects undergoing intravenous urograms for investigation of a renal problem (i.e., controls). A similar study was carried out using total knee replacement (TKR) as the selection criterion . Similar data, but using self-reported total joint arthroplasty in a smaller data set were found by a study in Oxford . The data shown in Table 1 indicate a strong familial aggregation, although lower than autoimmune arthropathies, not much lower than rheumatoid arthritis and much higher than metabolic disorders such as Type 2 diabetes.

Table 1

Familial aggregation of osteoarthritis and of other disorders.

Type of disorder	Condition	Ascertained via	Sibling recurrence risk λ _S
Autoimmune arthropathies	Rheumatoid arthritis	Sibs with condition	5.0–8.0
	Juvenile arthritis	“	15
	Systemic lupus erythematosus	“	30
Metabolic osteoarthritis	Type 2 diabetes	“	1.2–1.6
	Knee OA (TF and/or PF)	“	2.08
	TKR	“	4.81
	Anteromedial OA	Sibs with UKR	3.21
	Hip osteophytes grade 3	Sibs with THR	4.27
	THR	“	1.87–8.53
	Hip KL grade > = 3	“	4.99
	Hip JSW < = 1.5 mm	“	5.07

TKR = total knee replacement, THR = total hip replacement, UKR = unicompartmental knee replacement, KL = Kellgren – Lawrence grade, JSW = joint space width, PF = patellofemoral TF = tibiofemoral.

An alternative method to assess the actual genetic contribution to a condition is the use of classical twin studies, which enable investigators to quantify the environmental and genetic factors that contribute to a trait or disease. Comparing the resemblance of identical twins for a trait or disease with the resemblance of non-identical twins offers the first estimate of the extent to which genetic variation determines variation of that trait or heritability. The heritability of OA has been calculated in twin sets after adjustment of the data for other known risk factors such as age, sex and body mass index (BMI). Such findings show that the influence of genetic factors in radiographic OA of the hand, hip and knee in women is between 39% and 65%, independent of known environmental or demographic confounding factors. Classical twin studies and familial aggregation studies have also investigated the genetic contribution to cartilage volume and progression of disease (see MacGregor et al. ).

Linkage analyses

Genetic linkage occurs when a locus involved in the trait of interest (in this case OA) and alleles at nearby markers are inherited jointly. At least five genome-wide linkage scans exist in the literature based on small families or twins of affected relatives collected in the U.K., Finland, Iceland and the US (see Lee et al. . for references). These genome-wide linkage scans were performed on patients ascertained for hip, knee or hand OA and have identified a large number of relatively broad genomic intervals that may harbour OA susceptibility in chromosomes 2, 4, 6, 7, 11, 16, 19 and the X chromosome. Recently, Lee and co-workers conducted a meta-analysis of OA whole-genome scans from 893 families with 3000 affected individuals taking part in three studies (Iceland, U.K. and US). Their analysis provided summarised linkage loci of OA across whole-genome scan studies and, based on their data, they concluded that genetic regions in 7q34–7q36.3, 11p12–11q13.4, 6p21.1–6q15, 2q31.1–2q34 and 15q21.3–15q26.1 were the most likely to harbour OA susceptibility genes. However, to date, no susceptibility genes for large joint OA have been identified directly by linkage alone.

Genetic associations

Genetic association studies provide a means of quantifying the effects of specific gene variants on disease occurrence. If a genetic association is present, a particular allele, genotype at a given polymorphic locus will be seen more often than expected by chance in an individual affected by the disease. With this type of analysis, the frequencies of genotypes or alleles are compared usually in a classical cases-control design, although sometimes association studies are carried also in family settings.

How does genetic variation influence risk of OA?

Genetic variation can influence susceptibility to OA through diverse pathways and at different stages (see Fig. 1 ), and a select list of genes found to be associated with OA in various studies is presented in Table 2 .

Table 2

Selected published genetic associations with OA and biochemical pathways involved.

Pathway	Symbol	Gene name	Trait associated with	Known or putative function
BMP	ASPN	Asporin	Hip OA/Knee OA	Cartilage extracellular protein that regulates the activity of TGFβ
BMP	BMP2	Bone morphogenetic protein 2	Knee OA	Growth factor involved in chondrogenesis and osteogenesis
BMP	CILP	Cartilage intermediate layer protein	Knee OA,	Inhibits TGFß1–mediated induction of cartilage matrix genes
BMP	GDF5	Growth differentiation factor 5	Hip OA	Member of the BMP family, regulator of growth and differentiation
BMP/Wnt	OPG	Osteoprotegerin	Knee OA	Regulation of osteoclastogenesis
Inflammation	PLA2G4A	Cytosolic phospholipase A2	Knee OA	Catalyzes the rate-limiting step in the production of pro-inflammatory eicosanoids and free radicals
Inflammation	PTGS2	Cyclooxygenase 2	Knee OA, spine OA	COX-2-produced prostaglandin-E2 modulates cartilage proteoglycan degradation in OA
Inflammation	IL1 gene cluter	Interleukin (IL-)1 alpha, IL- beta and IL-1 receptor antagonist	Hip/knee/Hand OA	Regulation of metalloproteinase gene expression in synovial cells and chondrocytes
Inflammation	IL6	Interleukin 6	Hip/knee/Hand OA	Pro-inflammatory cytokine, involved in the cartilage degradation but also induces ILRa
Inflammation	IL10	Interleukin 10	Knee/Hand OA	Anti-inflammatory cytokine inhibits the synthesis of IL-1
Other	VDR1	Vitamin D receptor	Knee OA	Nuclear receptor, mediates effects of vitamin D whose serum levels affect incidence severity and progression of OA
Wnt, other	ANP32A	Acidic leucine-rich nuclear phosphoprotein 32 (pp32 or PHAPI)	Hip OA	Regulator of apoptosis and of Wnt signaling
Other	ADAM12	A disintegrin and metalloproteinase domain 12	Knee OA	Metalloprotease involved in osteoclast formation and cell-cell fusion
thyroxin	DIO2	Iodothyronine-deiodinase enzyme type 2	Hip OA, Generalized OA	Thyroxin signalling: Regulates intracellular levels of active thyroid hormones in target tissues
Wnt	FRZB	Secreted frizzled-related protein 3	Hip/knee OA, GOA	Wnt antagonist and modulator of chondrocyte maturation
Wnt	LRP5	Low- Density Lipoprotein Receptor-Related Protein 5	Knee OA	Receptor involved in Wnt signalling via the canonical beta-catenin pathway
Other	COMT	Catechol-O-methyltransferase	Pain in hip OA	Key regulator of pain perception

A comprehensive list of all the genetic associations with OA reported to date is beyond the scope of this review but in addition to the genes listed in Table 2 , several others have been reported, such as extracellular matrix molecules such as COL2A1 , MATN3 or COMP . A more extensive list can be found at Valdes and Spector . Here, we review some of those genes that are particularly clinically relevant, along with the molecular mechanisms involved in the pathology of OA.

The articular cartilage forms a biomechanical unit with the subchondral and cortical bone to attenuate forces through joints, particularly following impact loading. Hence, genes underlying the biology of articular cartilage are of the most likely ones to affect genetic risk of joint damage in OA. Besides the known features of the articular cartilage, thickening, sclerosis and osteophyte formation of the subchondral bone are very common in OA. Changes in the articular cartilage cause impairment in the absorption of shocks that it should normally attenuate, impacting the subchondral bone and subsequently destroying it and leading to secondary changes such as sclerosis and osteophyte formation . Therefore, genetic variations at molecules involved in bone remodelling are also involved in susceptibility to OA. Genetic susceptibility of risk of OA is thus due, at least in part, through genetic control of skeletal shape and development. Studies in animal models have shown that both skeletal development and skeletal shape are under tight genetic control and some studies have indicated a role for skeletal shape in the risk of OA. For example, Lane and colleagues , examining baseline and 8-year follow-up radiographs, found that an abnormal centre-edge angle and acetabular dysplasia were each associated with increased risk of incident hip osteoarthritis. More recently, Doherty and colleagues reported that both the femoral head shape and the femoral neck shaft angle of affected hips as well as of contralateral hips are very strongly associated with risk of hip OA, indicating that a non-spherical femoral head shape not only occurs as a consequence of OA, but itself may be a morphologic risk factor for development of hip OA.

Several of the genes known to control skeletal development in animal model systems, such as bone morphogenetic proteins (BMPs) and Wnt signalling genes, have indeed been associated with risk of OA in animal models and in humans, although whether this is due to an effect on skeletal shape has not been investigated. Wnt and BMPs are expressed in many overlapping tissues and dual regulation by Wnt and BMPs appear to be frequent in mammalian development . In addition, there is substantial cross-talk between these two pathways. These genes are also involved in bone remodelling and in chondrocyte biology and are therefore primary candidates for influencing risk of OA.

How does genetic variation influence risk of OA?

Table 2

Selected published genetic associations with OA and biochemical pathways involved.

Pathway	Symbol	Gene name	Trait associated with	Known or putative function
BMP	ASPN	Asporin	Hip OA/Knee OA	Cartilage extracellular protein that regulates the activity of TGFβ
BMP	BMP2	Bone morphogenetic protein 2	Knee OA	Growth factor involved in chondrogenesis and osteogenesis
BMP	CILP	Cartilage intermediate layer protein	Knee OA,	Inhibits TGFß1–mediated induction of cartilage matrix genes
BMP	GDF5	Growth differentiation factor 5	Hip OA	Member of the BMP family, regulator of growth and differentiation
BMP/Wnt	OPG	Osteoprotegerin	Knee OA	Regulation of osteoclastogenesis
Inflammation	PLA2G4A	Cytosolic phospholipase A2	Knee OA	Catalyzes the rate-limiting step in the production of pro-inflammatory eicosanoids and free radicals
Inflammation	PTGS2	Cyclooxygenase 2	Knee OA, spine OA	COX-2-produced prostaglandin-E2 modulates cartilage proteoglycan degradation in OA
Inflammation	IL1 gene cluter	Interleukin (IL-)1 alpha, IL- beta and IL-1 receptor antagonist	Hip/knee/Hand OA	Regulation of metalloproteinase gene expression in synovial cells and chondrocytes
Inflammation	IL6	Interleukin 6	Hip/knee/Hand OA	Pro-inflammatory cytokine, involved in the cartilage degradation but also induces ILRa
Inflammation	IL10	Interleukin 10	Knee/Hand OA	Anti-inflammatory cytokine inhibits the synthesis of IL-1
Other	VDR1	Vitamin D receptor	Knee OA	Nuclear receptor, mediates effects of vitamin D whose serum levels affect incidence severity and progression of OA
Wnt, other	ANP32A	Acidic leucine-rich nuclear phosphoprotein 32 (pp32 or PHAPI)	Hip OA	Regulator of apoptosis and of Wnt signaling
Other	ADAM12	A disintegrin and metalloproteinase domain 12	Knee OA	Metalloprotease involved in osteoclast formation and cell-cell fusion
thyroxin	DIO2	Iodothyronine-deiodinase enzyme type 2	Hip OA, Generalized OA	Thyroxin signalling: Regulates intracellular levels of active thyroid hormones in target tissues
Wnt	FRZB	Secreted frizzled-related protein 3	Hip/knee OA, GOA	Wnt antagonist and modulator of chondrocyte maturation
Wnt	LRP5	Low- Density Lipoprotein Receptor-Related Protein 5	Knee OA	Receptor involved in Wnt signalling via the canonical beta-catenin pathway
Other	COMT	Catechol-O-methyltransferase	Pain in hip OA	Key regulator of pain perception