The glycine transporter 1 gene (GLYT1) is associated with methamphetamine-use disorder†
Please cite this article as follows: Morita Y, Ujike H, Tanaka Y, Kishimoto M, Okahisa Y, Kotaka T, Harano M, Inada T, Komiyama T, Hori T, Yamada M, Sekine Y, Iwata N, Iyo M, Sora I, Ozaki N, Kuroda S. 2007. The Glycine Transporter 1 Gene (GLYT1) is Associated With Methamphetamine-Use Disorder. Am J Med Genet Part B 147B:54–58.
Abstract
Glycine transporter (GlyT)-1 plays a pivotal role in maintaining the glycine level at the glutamatergic synapse. Glycine is an allosteric agonist of N-methyl-D-aspartate (NMDA) receptors. Because activation of NMDA receptors is an essential step for induction of methamphetamine dependence and psychosis, differences in the functioning of GlyT-1 due to genetic variants of the GlyT-1 gene (GLYT1) may influence susceptibility. A case-control genetic association study of the GLYT1 gene examined 204 patients with methamphetamine-use disorder and 210 healthy controls. We examined three single nucleotide polymorphisms (SNPs), SNP1, IVS3 + 411C > T, rs2486001; SNP2, 1056G > A, rs2248829; and SNP3, IVS11 + 22G > A, rs2248632, of the GLYT1 gene and found that SNP1 showed a significant association in both genotype (P = 0.0086) and allele (P = 0.0019) with methamphetamine-use disorder. The T-G haplotype at SNP1 and SNP2 was a significant risk factor for the disorder (P = 0.000039, odds ratio: 2.04). The present findings indicate that genetic variation of the GLYT1 gene may contribute to individual vulnerability to methamphetamine dependence and psychosis. © 2007 Wiley-Liss, Inc.
INTRODUCTION
Abuse of methamphetamine induces a strong psychological dependence, and further consumption produces highly psychotic states, such as auditory hallucinations and persecutory delusions [Tatetsu, 1963; Ujike and Sato, 2004]. Conditioned place preference (CPP) and behavioral sensitization induced by methamphetamine treatment in rodents have been recognized as animal models of methamphetamine dependence and psychosis. Many lines of experimental evidence have shown that A10 dopamine neurons in the ventral tegmentum area (VTA) projecting into the accumbens, amygdala, and prefrontal cortex play central roles in the induction of CPP and sensitization to methamphetamine [Kalivas and Stewart, 1991; McBride et al., 1999; Ujike, 2002]. It was also demonstrated that activation of glutamate neurons in the prefrontal cortex projecting into the accumbens, amygdala, and VTA and activation of N-methyl-D-aspartate (NMDA) receptors are essential to the development of CPP and sensitization [Wolf, 1998]. For example, bilateral lesions of the prefrontal cortex prevented the induction of sensitization to psychostimulants [Wolf et al., 1995], and systemic treatment or intra-VTA injection of NMDA antagonists also prevented it [Karler et al., 1989; Vezina and Queen, 2000]. NMDA receptors are multimeric protein complexes that are activated by glutamate binding to the NR2 subunit [Laube et al., 1997]. Glycine also activates the NMDA receptors by binding to allosteric sites of the NR1 subunit [Johnson and Ascher, 1987]. Recent studies indicated that glycine concentrations around NMDA receptors in the forebrain are efficiently regulated and maintained at a subsaturated level by glycine transporters (GlyTs), which belong to a superfamily of 12 transmembrane Na+/Cl−-dependent transporters [Sato et al., 1995]. GlyT antagonists, e.g., N-[3-(4′-fluorophenyl)-3-(4′-phenylphenoxy)] propylsarcosine (NFPS), inhibited glycine uptake [Harsing et al., 2003] and robustly enhanced the NMDA receptor functions [Bergeron et al., 1998; Kinney et al., 2003]. GlyTs are divided into two subtypes, GlyT-1 and GlyT-2 encoded by GLYT1 (SLC6A9, MIM 601019) and GLYT2 (SLC6A5, MIM 604159), respectively. GlyT-1 is known to be predominantly expressed in glial cells in the central nervous system, especially the frontal cortex, and hippocampus [Borowsky et al., 1993; Zafra et al., 1995]. However, GlyT-1 was recently reported in neurons as well [Cubelos et al., 2005], where it is closely associated with NMDA receptors and regulates the extracellular glycine concentration at synapses [Smith et al., 1992]. GlyT-2 is mainly expressed in the spinal cord and brainstem, where it belongs to the family of Na+/Cl−-dependent plasma membrane transporters [Liu et al., 1993], and is colocalized with inhibitory glycine receptors [Gomeza et al., 2003]. Therefore, it is possible that altered function of GlyT due to genetic variants of the GLYT1, but not the GLYT2 gene, may affect individual susceptibility to methamphetamine-use disorder. We investigated the association between the GLYT1 gene polymorphisms and the disorder by a case-control study.
MATERIALS AND METHODS
Subjects
The subjects were 204 patients with methamphetamine-use disorder (167 males and 37 females; mean age, 37.4 years; SD 11.9 years) who met the ICD-10-DCR criteria (F15.2) and who were outpatients or inpatients in psychiatric hospitals of the Japanese Genetics Initiative for Drug Abuse (JGIDA), and 210 age-, gender-, and geographical origin-matched normal controls (163 males and 47 females; mean age, 36.5 years; SD 10.6 years), who were mostly medical staff members without a past individual or family history of drug dependence or major psychotic disorders. Patients who had a comorbidity of any other major psychiatric disorders, e.g., schizophrenia and bipolar disorder, were excluded. Assessment for the diagnosis of methamphetamine-use disorder and controls were performed by trained psychiatrists on the basis of all available information, including hospital notes. These assessments were done by unstructured interview.
The patients started methamphetamine abuse at 21.0 ± 5.5 years. As to multi-substance abuse status, 37.2% patients concurrently abused other illicit drugs besides methamphetamine. Cannabinoids were most frequently abused (34.0%), and followed by LSD (14.1%), cocaine (13.1%), opioids (12%), and hypnotics (9.9%). More than 60% patients abused methamphetamine solely, but about half of them had a past history of organic solvent abuse in their teenage. One hundred eighty-eight patients have or had the comorbidity of methamphetamine psychosis (F15.50). More details about the backgrounds of the patients have been published elsewhere [Ujike et al., 2003]. After the study was described, written informed consent was obtained from all participants. This study was approved by the Ethics Committee at each institute of the JGIDA.
Genotyping
The genomic DNA was extracted from peripheral leukocytes using the standard phenol/CHCl3 method. Over 100 single nucleotide polymorphisms (SNPs) spanning the GLYT1 gene were listed in the NCBI SNP database and the International HapMap project data. Since there was no non-synonymous SNP in the gene which should have potential physiological effects, we selected SNPs which fulfilled the following conditions: (1) those allele frequencies in a Japanese population were already known by the database and their minor allele frequencies were over 10%, (2) they were recognized by inexpensive restriction enzymes, (3) they were not located in deep introns, (4) they were located apart from each other to cover the entirety of the gene. Finally, we selected three SNPs, rs2486001 (IVS3 + 411C > T, SNP1), rs2248829 (1056G > A, SNP2), rs2248632 (IVS11 + 22G > A, SNP3). The three SNPs of the GLYT1 gene were individually amplified by polymerase chain reaction (PCR) using the primers listed in Table I. PCR was carried out in a total volume of 15 µl with 3% dimethyl sulfoxide and 0.75 units of Taq DNA polymerase in the reaction mixture. Initial denaturation was performed for 5 min at 95°C; 35 cycles were then performed (30 sec of denaturing at 95°C, 30 sec of annealing at the appropriate temperature, and 30 sec of extension at 72°C), followed by a final extension at 72°C for 5 min. The PCR products were then analyzed on 3.0% agarose gels after digestion with Eco47I (SNP1), HincII (SNP2), and MnlI (SNP3), respectively (Table I). Genotyping of SNP1, SNP2, and SNP3 were confirmed by direct sequencing of a part of the samples. To ensure the positive association for SNP1, we confirmed the genotypes of all samples by a different PCR mismatch primer set to produce a recognition site for a different restriction enzyme, Eco0109I. (Table I).
| SNP | Primer sequence | Product size (bp) | Annealing temp. (°C) | Restriction enzyme | |
|---|---|---|---|---|---|
| SNP1 (IVS3 + 411C > T, rs2486001) | Forward-1 | 5′-CACCCCCAGCCTGTCTTTCAAC-3′ | 122 | 61 | Eco471 |
| Reverse-1 | 5′-CCCCTAAATGGTTCAGGGATGT-3′ | ||||
| Forward-2 | 5′-AGCCTGGGCTGAGGCACACCAC-3′ | 188 | 65 | Eco01091 | |
| Reverse-2 | 5′-TTCTATTCCCTGGGGTTCAGCA-3′ | ||||
| SNP2 (1056G > A, rs2248829) | Forward | 5′-TGGGCAGAGGCAGGCACCTC-3′ | 141 | 61 | Hincll |
| Reverse | 5′-GAGTGACCCTGGAGGGAGCCGTTGA-3′ | ||||
| SNP3 (IVS11 + 22G > A, rs2248632) | Forward | 5′-AGAAGAGGGGTGGTGGGAATCC-3′ | 131 | 63 | Mnll |
| Reverse | 5′-TACGGTGAGCACTCGAGCGTCC-3′ |
Statistical Analysis
Deviation of the genotype counts from Hardy–Weinberg equilibrium was tested using a chi-square goodness-of-fit test. The statistical significance of difference was assessed by a chi-square test (genotype comparison) or log likelihood ratio test (allele comparison) at a significance level of 0.05. The pairwise linkage disequilibrium (LD) and haplotype frequencies were estimated by the EH algorithm using the SNPalyze program (Dynacom Co., Mobara-shi, Chiba, Japan).
RESULTS
Both genotype and allele frequency distributions of patients with methamphetamine-use disorder and control subjects are shown in Table II. The genotype distributions of SNP1, 2, and 3 of patients and controls did not deviate significantly from Hardy–Weinberg equilibrium. We found significant differences in the frequency of genotypes and alleles of SNP1 (genotype, χ2 = 9.52, P = 0.0086; allele, G = 9.66, P = 0.0019) and SNP2 (genotype, χ2 = 6.04, P = 0.048; allele, G = 4.14, P = 0.042) between patients with methamphetamine-use disorder and control subjects. SNP3 showed no significant differences in allele or genotype between groups. After Bonferroni correction, the genotype and allele frequencies of SNP1, but not SNP2, remained significant. The odds ratio of the T allele of SNP1 for methamphetamine-use disorder was 1.69 (95% confidence interval: 1.22–2.36).
| Group | N | Genotype | P | Corrected P | Allele | P | Corrected P | |||
|---|---|---|---|---|---|---|---|---|---|---|
| SNP1 (rs2486001) | C/C (%) | C/T (%) | T/T (%) | C (%) | T (%) | |||||
| Control | 210 | 139 (66.2) | 63 (30.0) | 8 (3.8) | 341 (81.0) | 79 (19.0) | ||||
| METH-use disorder | 204 | 106 (51.9) | 82 (40.2) | 16 (7.8) | 0.0086 | 0.026 | 294 (72.1) | 114 (27.9) | 0.0019 | 0.0057 |
| SNP2 (rs2248829) | G/G (%) | G/A (%) | A/A (%) | G (%) | A (%) | |||||
| Control | 210 | 110 (52.4) | 76 (36.2) | 24 (11.4) | 296 (70.5) | 124 (29.5) | ||||
| METH-use disorder | 204 | 119 (56.7) | 75 (35.7) | 10 (4.8) | 0.048 | — | 313 (76.7) | 95 (23.3) | 0.042 | — |
| SNP3 (rs2248632) | G/G (%) | G/A (%) | A/A (%) | G (%) | A (%) | |||||
| Control | 210 | 113 (53.8) | 81 (38.6) | 16 (7.6) | 307 (73.1) | 113 (26.9) | ||||
| METH-use disorder | 204 | 122 (58.1) | 70 (33.3) | 12 (5.7) | 0.45 | — | 287 (76.3) | 89 (23.7) | 0.20 | — |
- METH, methamphetamine.
We calculated the pairwise LD between SNP1, SNP2, and SNP3. D′ (absolute value) and r2 for pairwise LD are shown in Table III. High LD was detected between SNP1 and SNP2 (D′ = 0.76, r2 = 0.064), SNP1 and SNP3 (D′ = 0.92, r2 = 0.087), and SNP2 and SNP3 (D′ = 0.87, r2 = 0.70).
| SNP1 | SNP2 | |
|---|---|---|
| SNP2 | D′ = 0.76 | |
| r2 = 0.064 | ||
| SNP3 | D′ = 0.92 | D′ = 0.87 |
| r2 = 0.087 | r2 = 0.70 |
Because SNP1, SNP2, and SNP3 were shown to be located on the same LD block, the global haplotypic association was analyzed for every combination of the three SNPs (Table IV). Haplotypes consisting of SNP1–SNP2, SNP2–SNP3, and SNP1–SNP2–SNP3 showed significant association with patients with methamphetamine-use disorder. Among them, the haplotype comprising SNP1–SNP2 showed the smallest P value (P = 0.000011). The frequency of each haplotype consisting of SNP1–SNP2 is shown in Table V. The T-G haplotype of SNP1–SNP2 showed a significant excess in patients with methamphetamine-use disorder over control subjects (P = 0.000039), indicating that this haplotype was a risk factor for methamphetamine-use disorder, with an odds ratio of 2.04 (95% confidence interval, 1.45–2.86). These associations were still significant even after Bonferroni correction.
| SNP ID | 1SNP | 2SNP | 3SNP |
|---|---|---|---|
| SNP1 (C > T) | 0.0019 | ||
| 0.000011 | |||
| SNP2 (G > A) | 0.043 | 0.000050 | |
| 0.0040 | |||
| SNP3 (G > A) | 0.20 |
| SNP | Haplotype | METH | CON | P value |
|---|---|---|---|---|
| SNP1–2 | C-G | 0.48 | 0.54 | 0.080 |
| C-A | 0.24 | 0.27 | 0.30 | |
| T-G | 0.28 | 0.16 | 0.000039 | |
| T-A | 0.00 | 0.027 | — |
DISCUSSION
We found that the T allele of SNP1 (IVS3 + 411C > T, rs2486001) and the T-G haplotype consisting of SNP1 and SNP2 (1056G > A, rs2248829) of the GLYT1 gene showed a substantially significant association with methamphetamine-use disorder (allele P = 0.0018, haplotype P = 0.000039). The T-G haplotype of the gene approximately doubles the risk of predisposition to methamphetamine-use disorder.
GlyTs strictly maintain glycine concentrations in the vicinity of NMDA receptors. Glycine binds to glycine sites on the NR1 subunit of NMDA receptors and activates NMDA receptor signaling. Because the glycine concentration is set low by GlyTs, glycine sites are subsaturated in the physiological condition [Sato et al., 1995]. An increase of glycine concentration due to a glycine diet or administration of a glycine transport inhibitor, NFPS, enhanced the NMDA receptor function in vitro [Bergeron et al., 1998; Martina et al., 2004] and in vivo [Chen et al., 2003]. Heterozygous GLYT1 gene knockout mice showed enhancement of NMDA receptor function [Gabernet et al., 2005]. Therefore, increases and decreases in glycine induce stronger and weaker NMDA receptor neurotransmission. Because many lines of experimental evidence have shown that the glutamatergic system and NMDA receptor signaling in the brain play pivotal roles in the development of substance dependence on psychostimulants including amphetamine, methamphetamine, and cocaine [Cervo and Samanin, 1995; Bespalov, 1996; Kim and Jang, 1997; Wolf, 1998], it is possible that modulation of glycine sites or GlyTs also affects substance dependence. Induction of amphetamine-induced CPP in rodents was prevented by a glycine site antagonist L-701,324 (7-chloro-4-hydroxy-3-(2-phenoxy)phenyl-2(1H)-quinolone) [Mead and Stephens, 1999], and a glycine site partial agonist ACPC (1-aminocyclopropanecarboxylic acid), which disturbs the effects of endogenous glycine [Papp et al., 2002]. Combined with our findings, variants of the GLYT1 gene may affect susceptibility to methamphetamine dependence by modulating NMDA receptor function.
NMDA receptor signaling is also considered to be involved in psychotic disorders. Phencyclidine and ketamine, non-competitive antagonists of NMDA receptors, produce a psychotic state in healthy subjects and exacerbate symptoms in schizophrenics [Javitt and Zukin, 1991; Breier et al., 1997], and hypofunction of NMDA receptors is assumed to be a possible pathophysiology of schizophrenia. Mice with reduced expression of NR1 and the ε1 subunit of NMDA receptors due to genetic manipulation showed abnormal phenotypes similar to those observed in animal models of schizophrenia [Mohn et al., 1999; Miyamoto et al., 2001]. Recent human genetic studies also revealed a significant association of the GRIN1 and GRIN2A genes encoding the NR1 and NR2A subunits of NMDA receptors, respectively, with susceptibility to schizophrenia [Itokawa et al., 2003; Zhao et al., 2006]. In addition, the GlyT inhibitors SSR504734 (2-chloro-N-[(S)-phenyl[(2S)-piperidin-2-yl]methyl]-3-trifluoromethyl benzamide) and NFPS produced antipsychotic profiles in experimental paradigms [Javitt et al., 2004; Depoortere et al., 2005], and heterozygous GLYT1 knockout mice had reversed amphetamine-induced disruption of prepulse inhibition, one of the physiological phenotypes of psychosis [Tsai et al., 2004]. Because the majority of patients with methamphetamine-use disorder examined in the present study had a comorbid diagnosis of methamphetamine psychosis, it is possible that the GLYT1 gene may be involved in liability to psychotic symptoms. We re-analyzed the data and found that methamphetamine psychosis is also significantly associated with SNP1 of the GLYT1 gene (data not shown). Hence, the present findings may indicate that genetic variants of the GLYT1 gene could predict a risk of comorbidity of psychosis after repeated methamphetamine abuse.
The coding sequence of the human GLYT1 gene is divided into 14 exons (based on NM 1024845) distributed over a region of 133 kb, most of which (exons 2–14) cluster within 7 kb [Adams et al., 1995]. SNP1 (IVS3 + 411C > T) is located in intron 3, and SNP2 (1056G > A) is located in exon 7, but is synonymous. Usually, these SNPs are considered non-functional. However, recent advances in molecular genetics show intronic RNAs can actually be processed to smaller RNAs, snoRNAs, or miRNAs, which control various levels of gene expression in physiology and development, including transcription, RNA splicing, editing, translation, and turnover [Mattick and Makunin, 2006]. It is possible that SNP1 affects the GLYT1 gene expression and results in susceptibility to methamphetamine-use disorders. Alternatively, other functional mutations in the GLYT1 gene with LD with SNP1 or locating on the T-G haplotype of SNP1–SNP2 may be involved in an increased genetic risk for methamphetamine-use disorder because SNP1 was shown to be in an LD block extending at least from intron 3 to intron 11 in our Japanese samples. HapMap project data of a Japanese population (Rel 21a) showed a larger LD block from intron 1 to intorn 12. GLYT1 is expressed as three splice forms by alternative splicing in 5′-regions, GLYT1a, GLYT1b, and GLYT1c [Borowsky et al., 1993; Zafra et al., 1995]. It is possible that unexamined polymorphisms in intron 1 in LD with SNP1 may affect alternative splicing of the GLYT1 mRNA and result in susceptibility to methamphetamine-use disorders.
