Evaluation of the association between a single-nucleotide polymorphism of bone morphogenetic proteins 5 gene and risk of knee osteoarthritis.

BACKGROUND: Osteoarthritis (OA) is a chronic degenerative disorder probably affected by both genetic and environmental causes. Bone morphogenetic proteins (BMPs) are bone-derived factors that can induce new bone formation. Single-nucleotide polymorphisms (SNPs) of BMP5 gene alters the transcriptional activity of the BMP5 promoter that has been involved in OA susceptibility. This case-control study investigated the association of rs1470527 and rs9382564 SNP of BMP5 gene with susceptibility to knee OA (KOA).
MATERIALS AND METHODS: A total of 499 cases with radiographic KOA and 458 age- and sex-matched healthy controls were enrolled. Venous blood samples were obtained from all the cases as well as controls for polymerase chain reaction-restriction fragment length polymorphism.
RESULTS: The genotype distribution for rs1470527 and rs9382564 SNP was significantly different in cases and controls (P < 0.0001). Within both the SNPs of BMP5 gene, genotype CT and TT were significantly (P < 0.0001) associated with KOA as compared to the CC genotype. T allele of both the studied SNP was significantly associated with KOA (P < 0.0001). The allele frequencies of rs1470527 were 0.56(T) and 0.44(C) in cases and 0.33(T) and 0.67(C) in controls and in rs9382564 were 0.57(C) and 0.43(T) in cases and 0.71(C) and 0.29(T) in controls. Further in relation with clinical severity of OA, we observed signification association of TT genotype with both visual analog scale (P < 0.0001) and Western Ontario and McMaster Universities score (P < 0.05).
CONCLUSION: Our results indicate significant association of rs1470527 and rs9382564 polymorphism of BMP5 gene with KOA.

Introduction

Osteoarthritis (OA) is the second most common rheumatological problem and is the most frequent joint disease with the prevalence of 80% in population having radiographic evidence. About 60% radiographically evidenced subjects were symptomatic.[1] In India, more than 56.6% of population suffers from OA beyond the age of 65 years.[2] OA is a multifactorial disease with both genetical and environmental determinants and all cases are probably affected by both, with a continuous distribution between the extremes of genetic or predominantly environmental causes.[3] Role of genetics is emerging as an important etiological factor in recent times. Several candidate genes such as CALM-1 and GDF-5 have been associated with knee OA (KOA) in Indian population.[45]

Bone morphogenetic protein (BMP5), a member of the transforming growth factor (TGF-β) superfamily, is located on human chromosome 6p12.1 and found to be involved in synovial joint development and joint tissue homeostasis.[6] BMPs can trigger the entire process of cartilage and bone formation when implanted at ectopic sites in adult animals[7] and are normally expressed in and around early cartilage and bone precursors during embryonic development.[8]

The normal development and repair of the synovial joint are influenced by the activity of BMPs, so it is reasonable to speculate that variation in the activity or action of these molecules could influence the development of arthritic phenotypes.[69] Wilkins et al.[10] demonstrated a significant association of 5 (microsatellite D6S1276 and single-nucleotide polymorphisms [SNPs] rs921126, rs1470527, rs3798821, and rs9382564) among the 18 polymorphisms on 146 female hip OA cases and 215 age-matched female controls within case–control cohort (CCC1). These 5 polymorphisms were then genotyped in CCC 2 on 619 female hip OA cases and 566 female controls. None of these polymorphisms were associated with CCC2 alone. Further, when CCC1 and CCC2 were combined, however, D6S1276 and rs921126 demonstrated association with the combined cohort, and rs1470527 showed a trend toward association. Results of the study showed that allelic variants of the microsatellite are responsible for altered transcriptional activity of the BMP5 promoter which implies that polymorphism in cis-regulation of BMP5 is involved in OA susceptibility. Based on these findings, we performed a genetic study of rs1470527 and rs9382564 polymorphism of BMP5 gene in KOA.

Materials and Methods

This study was conducted in the Department of Orthopaedic Surgery, King George's Medical University, Lucknow. The study was approved by the Institutional Ethics Committee and was in accordance to the Declaration of Helsinki, 1975. A total of 499 cases and 458 age- and sex-matched controls were recruited from the outpatient clinic of the department. The sample size has been calculated using 80% power of the study. Informed consent was taken from all the participants. This case–control study consisted of men and women ≥40 years[11] who fulfilled the American College of Rheumatology clinical and radiographic criteria for KOA. These patients were profiled for demographic, clinical, radiological, and biochemical features. Radiographic features for severity of disease were assessed by Kellgren-Lawrence (KL) scores ranging from 0 to 4. KOA was classified into mild (KL grade 2), moderate (KL grade 3), and severe (KL grade 4). KOA symptoms were assessed through knee-specific Western Ontario and McMaster Universities (WOMAC) score[12] evaluating pain, stiffness and physical function, and interpretation response by a 5-point scale ranging from 0 to 4. Visual analog scale (VAS) was also assessed for knee pain.

DNA isolation and genotype analysis

Venous blood samples were obtained in ethylenediaminetetraacetic acid anticoagulant (20 g/L) for genomic DNA extraction. Samples were stored at −80°C until analysis. Isolated DNA was subsequently used for genotyping.

Detection of rs1470527 and rs9382564 polymorphism of bone morphogenetic protein 5 gene

Reaction was performed with 50-200 ng of genomic DNA, 0.2 μmol of each primer, 200 μmol of dNTP (Fermentas, USA), 1.2 mmol of MgCl2, 1 unit of Taq polymerase (Fermentas, USA), and sterile MilliQ water in a total volume of 20 μl. Amplification was performed on GeneAmp polymerase chain reaction (PCR) system 9700 (Applied Biosystem) using the following specific PCR conditions for each polymorphism.

rs1470527 polymorphism

A 384-bp PCR product of SFNaI (C/T; rs1470527) polymorphism of BMP5 gene was amplified by PCR using the following pair of primers: forward: 5′TACCACGAGCACTTCAGGTTG-3′ and reverse: 5′ CACACAGTCTTTGAATGGCCC 3′. Initial denaturation at 94°C for 5 min followed by 35 cycles of denaturation at 94°C for 45s, annealing at 58.3°C for 45s, and extension at 72°C for 1 min. It was further followed by final extension at 72°C for 10 min. PCR products (10 μL) were digested with 10U of SfaN1 restriction enzyme (MBI Fermentas, Germany) to identify the presence of polymorphic sites in BMP5 gene. Digestion of 384 bp PCR product of BMP5 gene with SfaN1 restriction enzyme into two fragments of 327 bp and 57 bp indicated the presence of TT genotype and the presence of fragments of three sizes (384 bp, 327 bp, and 57 bp) was indicative of CT genotype while the undigested 384 bp PCR fragment was indicative of CC genotype [Figure 2].

Figure 2

Restriction fragment length polymorphism image of single-nucleotide polymorphism rs1470527

rs9382564 polymorphism

A 276-bp PCR product of NspI (C/T; rs-9382564) polymorphism of BMP5 gene was amplified by PCR using the following pair of primers: forward: 5′ CAAGGAATCAAGGAAGGGATTCTG-3′ and reverse: 5′ CCACCATGAAGTACCAGCTACAAC 3′. Initial denaturation at 94°C for 5 min followed by 35 cycles of denaturation at 94°C for 45s, annealing at 56.7°C or 45s, and extension at 72°C for 1 min. It was further followed by final extension at 72°C for 10 min. PCR products (10 μL) were digested with 10U of Nsp1 restriction enzyme (MBI Fermentas, Germany) to identify the presence of polymorphic sites in BMP5 gene. Digestion of 276 bp PCR product of BMP5 gene with Nsp1 restriction enzyme into two fragments of 143 bp and 133 bp indicated the presence of CC genotype and the presence of fragments of three sizes (276 bp, 143 bp and 133 bp) was indicative of the CT genotype while the undigested 276 bp PCR fragment was indicative of TT genotype [Figure 3].

Figure 3

Restriction fragment length polymorphism image of single-nucleotide polymorphism rs9382564

Polymorphism in BMP5 gene was also compared with clinical symptoms of OA such as the knee-specific WOMAC index and VAS for knee pain. OA patients who are functionally or symptomatically poor (poor index) are defined as those with a WOMAC index score and VAS more than 33 and 6 while those with WOMAC index score and VAS ≤33 and 6 are classified as functionally or symptomatically good (good index).

Statistical analysis

Association of genotype of rs1470527 and rs9382564 polymorphism of BMP5 gene with OA knee was determined using standard Chi-square test. For comparison of continuous data of general characteristic of cases and controls, unpaired student t-test was performed. Haplotype analysis and linkage disequilibrium pattern were done by Haploview 4.1 software. Associations between the BMP5 variants and OA risk were estimated by calculating the odds ratios (ORs) and 95% confidence intervals (CIs) using both univariate and multivariate logistic regression analyses with adjustments for age, sex, and BMI. All statistical analyses were performed with the SPSS software package (version 16.0 for Windows; SPSS Chicago, IL, USA).

Results

A total of 499 radiographically confirmed KOA (KL-Grade 2–4) patients and 458 controls attending orthopedic OPD were recruited in the study. There were no significant differences in demographic characteristics between the cases and controls [Table 1]. Among the cases, 187 were male and 312 were female with age ranged from 40 to 72 years. Among controls, 168 were male and 290 were female with age ranged 40–70 years.

Table 1

General characteristics of cases and controls

	Case (n=499)	Control (n=458)	P	95% CI
Age (years)	54.56±9.30	53.93±8.55	0.265	−0.480-1.74
Gender, n (%)
Male (n=355)	187 (52.68)	168 (47.32)	0.79	−0.74-1.25
Female (n=602)	312 (51.83)	290 (48.17)
Height (m)	15,927±8.60	16,012±8.04	0.106	−1.884-0.184
Weight (kg)	63.64±9.15	63.91±8.31	0.62	−1.356-0.816
BMI (kg/m2)	25.44±3.24	24.99±4.29	0.06	−0.022-0.922

Data were as mean±SD. Difference between groups was calculated using unpaired Student's t-test (P value considered statistically significant). BMI: Body mass index, CI: Confidence interval, SD: Standard deviation

The genotype and allele distributions of the rs1470527and rs9382564 polymorphisms in cases with KOA and healthy controls are shown in Table 2. The genotype distribution for rs1470527 and rs9382564 was significantly different between the cases and controls (P < 0.0001). The major (T) and minor (C) allele frequencies of rs1470527 in cases were 0.56 (T) and 0.44 (C) and in controls were 0.33 (T) and 0.67 (C). Likewise, major (C) and minor (T) allele frequencies of rs9382564 in cases were 0.57 (C) and 0.43 (T) and in controls were 0.71 (C) and 0.29 (T).

Table 2

Comparison of genotype rs1470527 and rs9382564 polymorphism of bone morphogenetic proteins 5 gene between case and control

	Case (n=499), n (%)	Control (n=458), n (%)	OR (95% CI)	P
rs1470527 polymorphism (SFNaI C/T)
Comparison of genotype with OA
CC (Wild type)	78 (15.63)	265 (57.86)	Reference group
CT (Hetero)	287 (57.51)	90 (19.65)	10.83 (7.7-15.31)	<0.0001*
TT (Mutant)	134 (26.85)	103 (22.48)	4.41 (3.1-6.3)	<0.0001*
Allele frequency
C (n=1063)	443 (44.39)	620 (67.69)
T (n=851)	555 (55.61)	296 (32.31)	2.62 (2.2-3.2)	<0.0001*
rs9382564 polymorphism (NspI C/T)
Comparison of genotype with OA
CC (n=428)	169 (33.86)	259 (56.55)
CT (n=366)	228 (45.69)	138 (30.13)	2.53 (1.9-3.4)	<0.0001*
TT (n=163)	102 (20.44)	61 (13.31)	2.6 (1.8-3.7)	<0.0001*
Allele frequency
C (n=1222)	566 (56.71)	656 (71.62)
T (n=692)	432 (43.29)	260 (28.38)	1.92 (1.6-2.33)	<0.0001*

Values are given as frequency. Differences were tested by Chi-square test. (*P<0.05 considered statistically significant). OR: Odds ratio, CI: Confidence interval, OA: Osteoarthritis

Within both the SNPs of BMP5, a significantly increased risk for KOA was found to be associated with the genotype CT and TT as compared with the CC genotype. In addition, those bearing at least one T allele (CT + TT) in both the SNPs has also been observed to be significantly associated with KOA (P < 0.0001).

Cases with TT genotype of both the SNP (rs1470527 and rs9382564) of BMP5 gene were found to be associated with both VAS (P < 0.001) and WOMAC scale (P < 0.05) [Table 3]. The mean WOMAC and VAS scores were 35.01 and 6.2, respectively.

Table 3

Distribution of cases of different genotype according to visual analog scale subgroups (<6 and ≥6) and Western Ontario and McMaster Universities subgroups (<33 and ≥33)

	CC (%)	TT (%)	CT (%)	P	χ2
VAS
rs1470527
<6 (n=242)	58 (23.96)	99 (40.90)	85 (35.12)	<0.001*	26.97
≥6 (n=257)	20 (7.78)	188 (73.15)	49 (19.06)
rs9382564
<6 (n=239)	110 (46.02)	68 (28.45)	61 (25.52)	<0.001*	36.49
≥6 (n=260)	60 (23.07)	160 (61.53)	40 (15.38)
WOMAC
rs1470527
<33 (n=234)	39 (16.66)	129 (55.12)	66 (28.20)	0.035*	4.44
≥33 (n=265)	39 (14.71)	158 (59.62)	68 (25.66)
rs9382564
<33 (n=234)	73 (31.19)	113 (48.29)	48 (20.51)	0.023*	5.14
≥33 (n=265)	97 (36.60)	115 (43.39)	53 (20.00)

VAS: Visual analog scale, WOMAC: Western Ontario and McMaster Universities

Haplotype analysis was done by Haploview 4.1 software. Haplotype block structure was determined using the genotype data from case and controls groups. In the North Indian population, the pairwise “D” values were ≥0.09 and both SNPs were part of on haplotype block [Figure 1].

Figure 1

Linkage disequilibrium block image

The haplotype approach was used to explore the effect of the interaction between two SNPs of the BMP5 gene and the resulting risk of developing KOA. Total numbers of four possible haplotype were observed for these two SNPs rs1470527 and rs9382564 in all cases and control. Haplotype carrying all wild-type alleles of C-C was the reference haplotype. All remaining three haplotype (T-T, T-C, C-T) were significantly associated with KOA in cases as compared to the controls [Table 4].

Table 4

Distribution of bone morphogenetic protein 5 (rs1470527 and rs9382564) haplotype in cases and controls and knee osteoarthritis risks

Haplotype (allele combination)	Case (%)	Control (%)	OR (95% CI)	P
rs1470527 and rs9382564
C-C	141 (14.13)	393 (42.90)	1 reference
T-T	408 (40.88)	97 (10.59)	11.72 (8.74-15.72)	<0.0001
T-C	300 (30.06)	186 (20.31)	4.49 (3.4-5.9)	<0.0001
C-T	149 (14.93)	240 (26.20)	1.79 (1.3-2.29)	<0.0001

OR: Odds ratio, CI: Confidence interval

Stratification analysis was performed to evaluate the potential association of genetic variants of polymorphism rs1470527 with KOA risk in subgroups based on demographic characteristics. When stratified by age, both young (≤55 years) and old (>55 years) patients showed significant differences in genotype frequencies compared with controls (P < 0.0001) and similarly in both genders (male and female) [Table 5]. However, in polymorphism rs9382564 only old (>55 years) and female groups showed significant differences in genotype frequencies compared with controls (P < 0.0001) [Table 6].

Table 5

Logistic regression analysis of rs1470527 in bone morphogenetic protein 5 genotype frequencies and risk of osteoarthritis with osteoarthritis of the knee and controls

	Cases				Controls				TT versus CC		CT versus CC
	n=499	CC	CT	TT	n=458	CC	CT	TT	Adjusted ratio	95% CI	Adjusted ratio	95% CI
Age group
≤55	n=206	39 (18.93)	100 (48.54)	67 (32.52)	n=206	107 (51.94)	46 (22.33)	50 (24.27)	3.7	2.19-6.17	7.6	4.5-12.93
>55	n=293	39 (13.31)	187 (63.82)	67 (22.87)	n=255	158 (61.96)	44 (17.25)	53 (20.78)	5.12	0.9-2.3	17.21	10.68-27.83
Gender
Male	n=187	35 (18.71)	88 (47.06)	64 (34.22)	n=168	84 (50.0)	39 (23.21)	45 (26.79)	3.41	1.9-5.9	5.41	3.14-9.3
Female	n=312	43 (13.78)	199 (63.78)	70 (22.43)	n=290	181 (62.41)	51 (17.59)	58 (20.0)	5.1	3.13-8.22	16.42	10.44-25.83

Data presented as n (%) of patients. Adjusted for the other covariate presented in this table and for BMI using a logistic regression model for each stratum. CI: Confidence interval, BMI: Body mass index

Table 6

Logistic regression analysis of rs9382564 in bone morphogenetic protein 5 genotype frequencies and risk of osteoarthritis with osteoarthritis of the knee and controls

	Cases				Controls				TT versus CC		CT versus CC
	n=499	CC	CT	TT	n=458	CC	CT	TT	Adjusted ratio	95% CI	Adjusted ratio	95% CI
Age group
≤55	n=206	66 (32.03)	94 (45.63)	46 (22.33)	n=203	83 (40.87)	77 (37.93)	43 (21.18)	1.3	0.79-2.23	1.5	1.0-2.6
>55	n=293	103 (35.15)	134 (45.73)	56 (19.11)	n=255	176 (69.01)	61 (23.92)	18 (7.06)	5.3	2.96-9.5	3.8	2.54-5.5
Gender
Male	n=187	65 (34.76)	79 (42.24)	43 (22.99)	n=168	75 (44.64)	59 (35.11)	34 (20.24)	1.4	0.83-2.6	1.5	0.96-2.5
Female	n=312	104 (33.33)	149 (47.76)	59 (18.91)	n=290	184 (63.45)	79 (27.24)	27 (9.31)	3.9	2.3-6.5	3.4	2.3-4.8

Data presented as n (%) of patients. Adjusted for the other covariate presented in this table and for body mass index using a logistic regression model for each stratum. CI: Confidence interval

When the association between T allele carriers and the risk of OA was evaluated using logistic regression analysis, CT genotype carried a 7.62-fold increased risk of OA in young patients group and 17.21-fold in old patient group, compared with CC genotype [Table 5] whereas TT carried 3.7 and 5.12. Similarly, CT genotype carried 5.41 and 16.42-fold increased risk of KOA in males and females and TT genotype carried 3.41 and 5.08, respectively, compared with CC genotype [Table 5].

In contrast, stratification analysis of age and gender in rs9382564 genetic variants showed strong association of CT heterozygote with 3.75-fold increased risk of OA only in old patients group and 3.36-fold in female patient group, compared with TT homozygote and TT mutant homozygous carried 5.31 and 3.86 [Table 6].

Discussion

BMP's are known to play a key role in both formation and repair of skeletal structures.[13] Initially discovered as promoters of bone growth and formation, BMPs are now known to be key participants in the morphogenesis of a variety of vertebrate tissues and organs with a well-defined role in the differentiation of chondrocytes through promotion of cell proliferation and matrix synthesis.[1415]

Bramlage et al.[16] stated that BMP5 has been expressed in normal synovial tissue and were decreased in OA and RA. In a study by Feldman et al.,[17] an association was reported between BMP5 protein expression and an array of human axial skeletal abnormalities. Csomor et al.[18] showed that BMP5 plays a role in the pathogenesis of otosclerosis by accelerating hypertrophic markers which is of relevance in both development and progression of diseases such as OA. BMP5 clearly exhibits stage-specific regulation of multiple signaling and promotes chondrogenic differentiation.[19] These results suggest that BMP5 may be important in joint homeostasis and are therefore potential candidates for joint regeneration. Thus, it is reasonable to speculate that variation in the activity or action of these molecules could influence the development of arthritic phenotype.[69]

Several candidate genes have previously been associated with OA to display significantly higher expression in OA cartilage than in normal donor cartilage.[20] BMP5 is a member of the TGF-β superfamily of secreted signaling molecules, initially discovered as promoters of bone growth and formation.[14] In a study by Liang et al. on polymorphism rs373444 of BMP5 gene, no significant association was found with KOA.[21] Based on a study by Wilkins et al.[10] wherein they demonstrated significant association of 5 among the 18 polymorphisms studied on one of their CCC on hip OA in females. We have selected two of these five SNPs as these two did not show any association when “Wilkins” combined both the CCC1 and CCC2 populations of hip OA whereas we are studying KOA.

We found that both these SNPs of BMP5 were associated with KOA and this was validated by haplotypic association also. Further the association between the risk of KOA and rs1470527 CT and TT genotype compared with TT homozygotes was stronger in both gender and age groups. However, rs9382564 CT and TT genotype were stronger only in female patients and those >55-year-old. Furthermore, our data showed a significant association of TT genotype of both SNPs in relation to KOA and also with the clinical VAS and WOMAC scores.

BMP5 is a candidate gene for OA because it has been associated with hip OA in UK women.[10] Its allelic variants showed a regulatory effect on the BMP5 promoter in vitro. BMPs are members of the TGF-β superfamily that were identified by their involvement in cartilage and bone development. BMP5 gene associated with OA with multiallelic polymorphisms is an intronic microsatellite. In our study, both these polymorphisms are intronic SNPs and both SNPs are significantly associated.

BMP5 is expressed in proliferating zone chondrocytes of the growth plate and is markedly increased by hypertrophic differentiation. In chondrocytes, phosphorylation and accumulation of phosphor-SMADs lead to overexpression of genes involved in cartilage homeostasis and hypertrophic differentiation markers.[19] As a result, chondrocyte proliferation and cartilage matrix synthesis increase at cellular level.[15] BMP5 e xpression is also decreased in synovial and bone.[1622] These changes in expression along with chondrogenic effect and involvement in chondrocyte hypertrophy indicate the role of BMP5 in pathogenesis of OA.

The data suggested that identification of genetic factors may enable better understanding of the pathogenesis of this complex disease. This genetic association could provide prognostic information in the context of other known risk factors such as age, gender, and BMI and may still be useful clinically in terms of changing risk behavior. The genetic marker (SNPs) may provide independent predictive power and surrogate measures of familial aggregation, such as family history. Although, currently, we cannot modify genes, the genetic information is likely to be helpful in planning other health interventions or initiating behavior changes.

Financial support and sponsorship

This study was supported by a research fund from the Department of Biotechnology, New Delhi.

Conflicts of interest

There are no conflicts of interest.