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

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).

Table I. Characteristics of DLS and control subjects.

Table I.

Characteristics of DLS and control subjects.

Factor	DLS (n=70)	Controls (n=141)
Age (years, mean ± SD)	67.63±7.05	66.4±8.23
Gender (male/female)	10/60	25/116
Cobb’s angle (°, mean ± SD)	17.07±2.95	NA
Lateral listhesis (mm, mean ± SD)	4.78±2.30	NA

[i] DLS, degenerative lumbar scoliosis; NA, not applicable.

SNP genotyping

The cSNPs of GRIN2A, GRIN2B and GRIN2C were searched. Information relating to the cSNPs was obtained from the SNP database (www.ncbi.nlm.nih.gov/SNP; dbSNP Build 130) of the National Center for Biotechnology Information (NCBI). cSNPs with unknown heterozygosity and minor allele frequencies (<5%) were excluded. Finally, 9 cSNPs were selected [GRIN2A (rs8049651, Leu425Leu; rs9806806, Tyr730Tyr); GRIN2B (rs7301328, Pro122Pro; rs35025065, Asp447Asp; rs1805522, Ile602Ile; rs1806201, Thr888Thr; rs1805247, His1399His); and GRIN2C (rs689730, Ala33Ala; rs3744215, Arg1209Ser)] for analysis. Genomic DNA was extracted from blood samples collected in EDTA using a commercially available Qiagen DNA extraction kit (Qiagen, Tokyo, Japan). Genomic DNA was amplified using primers for each cSNP (Table II). The PCR products were sequenced using an ABI PRISM 3730XL analyzer (PE Applied Biosystems, Foster City, CA, USA). Sequence data were analyzed using SeqManII software (DNASTAR Inc., Madison, WI, USA).

Table II. Sequences of primers for cSNPs of GRIN2A, GRIN2B and GRIN2C genes.

Table II.

Sequences of primers for cSNPs of GRIN2A, GRIN2B and GRIN2C genes.

Gene	SNP	Direction	Sequence (5′→3′)	Size (bp)
GRIN2A	rs8049651, Leu425Leu	Forward	GATTTGCCTCTCCAGAAATCAG	321
		Reverse	CTATTTCAAAGGGTTGGGCACG
	rs9806806	Forward	TGCATGCATTTACCTCCTAACA	380
	Tyr730Tyr	Reverse	AATGGGAGCAGATAGGAACTGA
GRIN2B	rs7301328, Pro122Pro	Forward	CGAGAAAGATGATTTCCACCAT	391
		Reverse	GATTAATTGCTGGCCTATCCAC
	rs35025065, Asp447Asp	Forward	GGTCTGTTGCCTGCTTTATTTC	322
		Reverse	CAGGTTCCATTGATTTTCTTCC
	rs1805522, Ile602Ile	Forward	TGAGAGTCCCTGGAAAAGAGAG	392
		Reverse	CCAGAAACCTGGTCCACATATT
	rs1806201, Thr888Thr	Forward	TTGTGGTCATTTCTAGCCTCTC	371
		Reverse	CCGAACGTTCTCTCTACCTCAC
	rs1805247, His1399His	Forward	CCTACGCCCACATGTTTGAGAT	381
		Reverse	GGTTTTTGTTGTTAGGCACACA
GRIN2C	rs689730, Ala33Ala	Forward	AACCTCTGTCTCTCTCCCTGTG	394
		Reverse	GGAGATGAAGTCAAGGATCTGG
	rs3744215, Arg1209Ser	Forward	CTGTCTGTCCTCACCTTCCACC	392
		Reverse	CACCATGATTTGAGGCTACTGA

[i] SNP, single nucleotide polymorphisms.

Statistical analysis

The Hardy-Weinberg equilibrium (HWE) was assessed using SNPStats (http://bioinfo.iconcologia.net/snpstats) (12). Multiple logistic regression models (codominant, dominant and recessive) were used to calculate the odds ratio (OR), 95% confidence interval (CI) and corresponding P-values, with age and gender as covariables. To analyze the associations of cSNPs and haplotypes, SNPStats, HapAnalyzer and HelixTree (Golden Helix, Inc., Bozeman, MT, USA) were used.

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).

Table III. Genotype frequencies of cSNPs in GRIN2A, GRIN2B and GRIN2C genes.

Table III.

Genotype frequencies of cSNPs in GRIN2A, GRIN2B and GRIN2C genes.

			DLS		Control
Gene	SNP	Genotype	n	Freq.	n	Freq.	Models	OR (95% CI)	P-value
GRIN2A	rs8049651, Leu425Leu	C/C	61	0.87	123	0.87	Codominant	0.88 (0.40–1.94)	0.76
		C/T	9	0.13	16	0.11	Dominant	0.99 (0.42–2.35)	0.99
		T/T	0	0.00	2	0.01	Recessive	0.00 (0.00–NA)	0.17
	rs9806806, Tyr730Tyr	C/C	63	0.90	120	0.85	Codominant	0.60 (0.26–1.37)	0.20
		C/G	7	0.10	18	0.13	Dominant	0.64 (0.25–1.59)	0.32
		G/G	0	0.00	3	0.02	Recessive	0.00 (0.00–NA)	0.11
GRIN2B	rs7301328, Pro122Pro	C/C	12	0.18	31	0.22	Codominant	0.86 (0.57–1.29)	0.46
		G/C	32	0.47	61	0.43	Dominant	0.91 (0.49–1.69)	0.77
		G/G	24	0.35	49	0.35	Recessive	0.69 (0.32–1.46)	0.32
	rs1805522, Ile602Ile	A/A	5	0.07	7	0.05	Codominant	0.93 (0.57–1.51)	0.76
		G/A	16	0.23	41	0.29	Dominant	0.84 (0.45–1.56)	0.58
		G/G	49	0.70	93	0.66	Recessive	1.22 (0.36–4.16)	0.75
	rs1806201, Thr888Thr	A/A	14	0.21	34	0.24	Codominant	0.89 (0.58–1.36)	0.60
		G/A	37	0.54	72	0.51	Dominant	0.95 (0.48–1.87)	0.88
		G/G	17	0.25	34	0.24	Recessive	0.77 (0.38–1.57)	0.47
	rs1805247, His1399His	A/A	46	0.66	91	0.65	Codominant	0.84 (0.50–1.40)	0.50
		A/G	23	0.33	42	0.30	Dominant	0.96 (0.53–1.77)	0.91
		G/G	1	0.01	8	0.06	Recessive	0.20 (0.02–1.68)	0.08
GRIN2C	rs689730, Ala33Ala	C/C	36	0.54	81	0.58	Codominant	1.28 (0.77–2.13)	0.34
		C/T	26	0.39	56	0.40	Dominant	1.14 (0.63–2.06)	0.66
		T/T	5	0.07	3	0.02	Recessive	3.31 (0.76–14.48)	0.10
	rs3744215, Arg1209Ser	G/G	26	0.39	44	0.31	Codominant	0.97 (0.64–1.47)	0.88
		G/T	27	0.40	74	0.52	Dominant	0.75 (0.41–1.39)	0.36
		T/T	14	0.21	23	0.16	Recessive	1.38 (0.66–2.91)	0.40

[i] SNP, single nucleotide polymorphism; cSNPs, coding SNPs; DLS, degenerative lumbar scoliosis; freq, frequency; OR, odds ratio; 95% CI, 95% confidence interval; NA, not applicable.

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, rs689730 of GRIN2C was associated with Cobb’s angle in the codominant (P=0.038; OR, 2.72; 95% CI, 0.98–7.55) and dominant models (P=0.022; OR, 3.79; 95% CI, 1.15–12.54). The present study showed that the CT or TT genotype frequency with Cobb’s angle ≥20° was ∼2-fold higher than with Cobb’s angle <20°. In the analysis of lateral listhesis, rs7301328 of GRIN2B showed significant differences in the codominant (P=0.0027; OR, 3.21; 95% CI, 1.41–7.29), dominant (P=0.015; OR, 3.91; 95% CI, 1.22–12.51) and recessive models (P=0.015; OR, 5.39; 95%CI, 1.26–23.08). The results indicated that the CC genotype frequency with lateral listhesis ≥6 mm was ∼4-fold higher than with lateral listhesis <6 mm (Table IV).

Table IV. Genotype frequencies of rs689730 and rs7301328 in DLS patients with Cobb’s angle or lateral listhesis.

Table IV.

Genotype frequencies of rs689730 and rs7301328 in DLS patients with Cobb’s angle or lateral listhesis.

A, Cobb’s angle
		<20°		≥20°
SNP	Genotype	n	Freq.	n	Freq.	Model	OR (95% CI)	P-value
GRIN2C,	C/C	15	0.75	21	0.45	Codominant	2.72 (0.98–7.55)	0.038
rs689730,	C/T	4	0.20	22	0.47	Dominant	3.79 (1.15–12.54)	0.022
Ala33Ala	T/T	1	0.05	4	0.09	Recessive	1.82 (0.18–17.97)	0.590

B, Lateral listhesis
		<6 mm		≥6 mm
SNP	Genotype	n	Freq.	n	Freq.	Model	OR (95% CI)	P-value
GRIN2B,	C/C	3	0.08	9	0.32	Codominant	3.21 (1.41–7.29)	0.003
rs7301328,	G/C	18	0.45	14	0.50	Dominant	3.91 (1.22–12.51)	0.015
Pro122Pro	G/G	19	0.48	5	0.18	Recessive	5.39 (1.26–23.08)	0.015

[i] DLS, degenerative lumbar scoliosis; SNP, single nucleotide polymorphism; freq, frequency. OR, odds ratio; 95% CI, 95% confidence interval.

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 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.

Glutamate is the most prominent excitatory neurotransmitter in the CNS. It has been studied intensively in a number of neuropathological conditions, including stroke, epilepsy and neurodegenerative diseases (17,18). Glutamate exerts its effects through ionotropic ligand-gated ion channels and metabotropic G-protein-coupled glutamate receptors. The ionotropic receptors are further divided into NMDA, α-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) and kainate receptors (19–21). Spencer et al(6) conducted a review of NMDA-type glutamate signalling.

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.

Acknowledgements

The authors would like to thank Eri Kwon for proofreading the manuscript in English.

References