Assessment of NMDA receptor genes (GRIN2A, GRIN2B and GRIN2C) as candidate genes in the development of degenerative lumbar scoliosis

Degenerative lumbar scoliosis (DLS) progresses with aging after 50–60 years. The genetic association of DLS remains largely unclear. In this study, the genetic association between glutamate receptor, ionotropic, N-methyl D-aspartate (NMDA, GRIN) receptor genes and DLS was investigated. A total of 9 coding single nucleotide polymorphisms (cSNPs) in NMDA receptor genes [GRIN2A (rs8049651, Leu425Leu; rs9806806, Tyr730Tyr); GRIN2B (rs7301328, Pro122Pro; rs35025065, Asp447Asp; rs1805522, Ile602Ile; rs1806201, Thr888Thr; rs1805247, His1399His); and GRIN2C (rs689730, Ala33Ala; rs3744215, Arg1209Ser)] were selected and genotyped using direct sequencing in 70 patients with DLS and 141 healthy controls. Multiple logistic models (codominant, dominant and recessive) were calculated for the odds ratio (OR), 95% confidence interval (CI) and corresponding P-values. The SNPStats, SNPAnalyzer and HelixTree programs were used for the evaluation of the genetic data. Among the SNPs examined, no significant associations were observed between the NMDA receptor genes and DLS. When the patients were divided into two groups according to clinical characteristics based on Cobb’s angle (<20° or ≥20°) and lateral listhesis (<6 mm or ≥6 mm), associations were observed between rs689730 of GRIN2C and Cobb’s angle (codominant, P=0.038; dominant, P=0.022) and between rs7301328 of GRIN2B and lateral listhesis (codominant, P=0.003; dominant, P=0.015; recessive, P=0.015). These results indicate that the GRIN2A, GRIN2B and GRIN2C genes do not affect the development of DLS. However, the GRIN2C gene may be associated with Cobb’s angle, while the GRIN2B gene may be associated with lateral listhesis.


Introduction
Degenerative scoliosis develops after skeletal maturity without a previous history of scoliosis and typically presents in the lumbar or thoracolumbar spine (1,2). The etiology of degenerative lumbar scoliosis (DLS) remains unclear, although several factors, including hormones, genetics and degenerative diseases of the spine, may be involved (3). It has been suggested that the cause of DLS is associated with disc degeneration and osteoporosis (4).
During an individual's lifetime, modeling, remodeling and repair of the bone occur as a result of the activity of osteoclasts and osteoblasts (5). The control of bone remodeling is performed by a few neurotransmitters, such as glutamate, via corresponding receptors in the bone cells (6). Glutamate is a major excitatory neurotransmitter in the central nervous system (CNS) (7) and acts through a variety of receptors including N-methyl-D-aspartate (NMDA) receptors. The NMDA receptors (NMDARs) are composed of NMDAR1, NMDAR2 and NMDAR3 and are expressed in the bone cells (8,9). In addition, glutamate is a known regulator of the maturation and differentiation of osteoclasts and osteoblasts (10,11).
We hypothesized that genetic polymorphisms in the NMDAR genes may be associated with DLS. However, there have been no published studies concerning the association between DLS and single nucleotide polymorphisms (SNPs) of NMDAR genes. Therefore, the 9 coding SNPs (cSNPs) of NMDAR2A (GRIN2A), NMDAR2B (GRIN2B) and NMDAR2C (GRIN2C) were analyzed in 70 patients with DLS and 141 control subjects.

Materials and methods
Subjects. The DLS group comprised 70 unrelated individuals (10 males and 60 females), aged 52-82 years (mean ± SD, 67.63±7.05 years). The DLS patients were selected from the Spine Center of Kyung Hee University East-West Neo Medical Center (Seoul, Korea) and the National Medical Center (Seoul, Korea), out of ~9,000 patients per year over three years. Each patient was diagnosed by two specialized spine surgeons and all patients fulfilled the physical examination and radiographic criteria (Cobb's angle, >10˚). Informed consent was obtained from all individuals according to the Declaration of Helsinki guidelines. The study was approved by the Ethics Review Committee of the Medical Research Institute, Kyung Hee University Medical Center (Seoul, Korea). The control group comprised 141 individuals (25 males and 116 females), aged 53-85 years (mean ± SD, 66.74±8.23 years). Control group individuals were recruited following a general health check-up program to confirm that they had no clinical evidence of DLS according to radiographic criteria. The ages and genders of the control group individuals were matched to those of the DLS group (Table I).

Results
Of the nine cSNPs of GRIN2A, GRIN2B and GRIN2C genes examined, eight were polymorphic and one (rs35025065) was monomorphic (Table III). The genotype distributions of the eight cSNPs were in HWE (P>0.05, data not shown). As shown in Table III, there was no significant difference in the genotypic frequency of polymorphism rs8049651 between the DLS (CC 87%, CG 13% and GG 0%) and control groups (CC 87%, CG 11% and GG 1%; P=0.76). Additionally, the other seven cSNPs were not associated with DLS. Via the Gabriel method (13), two cSNPs (rs1805247 and 1806201 of GRIN2B) were used to construct a one linkage disequilibrium (LD) block (data not shown). Three haplotypes in the block had frequencies >0.1. However, there was no significant difference in haplotype frequencies between the DLS and control groups (data not shown).
However, when DLS patients were divided into two groups according to clinical characters based on Cobb's angle (<20˚ or ≥20˚) and lateral listhesis (<6 mm or ≥6 mm), associations were observed between the GRIN2B and GRIN2C polymorphisms and clinical characteristics. As shown in Table IV (Table IV).

Discussion
A number of individuals present degenerative diseases of the spine, such as spondylolisthesis, lateral listhesis, spinal stenosis and degenerative scoliosis with increasing age (3,14). Scoliosis Table III. Genotype frequencies of cSNPs in GRIN2A, GRIN2B and GRIN2C genes. is a spinal deformity with lateral curvature or angulation and may be accompanied by rotation of the affected vertebrae in the vertical axis (15). Complicated DLS induces severe back or leg pain with neurological deficits by compressing neural elements and causing nerve root ischemia (16). Despite uncertainty with regard to the etiology of DLS, degeneration of the spinal column occurs in the development of the disease (3). Considering that the etiology of DLS is unclear, the identification of candidate genes associated with the development of DLS may potentially be used to identify at-risk individuals. Therefore, the potential of NMDAR2 genes as candidate genes in the development of DLS was investigated through a case-control analysis in the current study.
In the current study, 9 cSNPs in the genes GRIN2A, GRIN2B and GRIN2C were analyzed to assess the association between these polymorphisms and DLS. The GRIN2A, GRIN2B and GRIN2C genes are located on chromosomes 16p13.2, 12p12 and 17q25, respectively. In the coding regions of each gene, 15, 22 and 14 cSNPs have been discovered, respectively. Of these 51 cSNPs, 9 exhibited heterozygosity >0.1 (www.ncbi.nlm.nih.gov/SNP, dbSNP Build 130) and were selected for analysis in thist study. Overall, none of the genotype frequencies or haplotype distributions of the NMDRs (GRIN2A, GRIN2B and GRIN2C) were observed to be associated with DLS.
Several studies have shown associations between SNPs of GRIN2A and/or GRIN2B and neurological or psychiatric diseases, such as Huntington's disease (22), schizophrenia (23), bipolar disorder (24) and Alzheimer's disease (25). In addition, Kim et al (26) reported that rs1806201 of GRIN2B was associated with alcoholism in a Korean population, but rs18054247 was not. However, Tadic et al (27) reported that rs1806201 was not associated with alcohol dependence, alcohol withdrawal-induced seizures and delirium tremens in a Caucasian population.
The possible associations between nine cSNPs of GRIN2A, GRIN2B and GRIN2C genes and DLS were investigated, but the results were negative. To compare each genotype frequency in various ethnic populations, the human SNP database (www. ncbi.nlm.nih.gov-SNP, dbSNP Build 130) was searched. The genotype distributions of all cSNPs in the control group were similar to those of the Asian populations, particularly the Japanese population (data not shown).
When the patients were divided into two groups according to clinical characteristics (Cobb's angle and lateral listhesis), GRIN2B and GRIN2C were associated with the severity of DLS. As shown in Table IV, rs689730 of GRIN2C was associated with Cobb's angle in the codominant (P=0.038) and dominant models (P=0.022). Furthermore, rs7301328 of GRIN2B was associated with lateral listhesis in the codominant (P=0.003), dominant (P=0.015) and recessive models (P=0.015).
The present results indicate that GRIN2A, GRIN2B and GRIN2C do not affect the development of DLS in the Korean population. However, significant associations were observed between GRIN2C and Cobb's angle and between GRIN2B and lateral listhesis in DLS.