Gade-Andavolu R, MacMurray JP, Blake H, Muhleman D, Tourtellotte W, Comings DE. Association Between the γ-Aminobutyric Acid A3 Receptor Gene and Multiple Sclerosis. Arch Neurol. 1998;55(4):513-516. doi:10.1001/archneur.55.4.513
Copyright 1998 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.1998
In a prior study we observed an association between the dopamine D2 receptor gene (DRD2) and the age of onset and/or diagnosis of multiple sclerosis (MS). We hypothesized that this effect was mediated through the dopaminergic control of the release of prolactin, a modulator of immune response. Since γ-aminobutyric acid also modulates the release of prolactin, we examined the possible association between alleles of the GABRA3 (γ-aminobutyric acid A3 receptor) gene and MS.
We examined the GABRA3 alleles of 189 subjects with MS who died of their disease. They were divided into test group 1 (n=64) and retest group 2 (n=56). Each group had a separate set of controls (group 1, n=109; group 2, n=430). All subjects were white. All were tested at a dinucleotide cytosine-adenosine repeat polymorphism with 6 alleles representing 11 to 16 repeats.
In the first group there was a significant difference in the frequency of the GABRA3 alleles (P<.002), with the most notable difference being an increase in the frequency of the 16-repeat allele in subjects with MS and a relative decrease in the other alleles. In the replication group there was again a significant difference in the distribution of the GABRA3 alleles (P<.001), and again the greatest difference was an increase in the frequency of the 16-repeat allele in subjects with MS. For both groups combined, a significant difference in the frequency of the 16-repeat allele was noted (χ2=46.30; P<.001).
These results suggest the GABRA3 gene may be a risk factor for MS. As with the DRD2 gene, the effect may be mediated through its regulation of prolactin release.
MULTIPLE sclerosis (MS) is a degenerative disease of the nervous system that affects an estimated 300000 persons in the United States. The disease most often appears in individuals between 20 and 40 years of age and has a strong genetic component.1 In MS, the immune system attacks the myelin sheath, leading to a range of symptoms that include blurred vision, fatigue, numbness, loss of movement, memory impairment, and difficulties with bladder, bowel, and sexual function.
In an earlier study, we observed a significant association between the 2 haplotypes of the dopamine D2 receptor gene (DRD2) and the age of onset and/or diagnosis of MS (J. P. MacMurray, PhD; N. Gonzalez; D. Muhleman, MS; W. Tourtellotte, MD; D. E. Comings, MD; unpublished data, 1998). Since dopamine, especially at the dopamine D2 receptor, regulates the release of prolactin, a modulator of immune function, we hypothesized that this was the pathway by which variants of DRD2 were acting as a risk factor for MS.
γ-Aminobutyric acid (GABA) is the principal inhibitory neurotransmitter in the central nervous system. Since it also plays a role in the regulation of prolactin release, we sought to determine if genetic variants of GABA receptors might also act as risk factors for MS. A number of other studies have suggested a potential role of abnormalities for GABA in MS. Levels of GABA in the cerebrospinal fluid of patients with MS are significantly lower than levels in control subjects.2 Treatment of MS with isoniazid has been shown to produce a 4-fold increase in both free and conjugated GABA.3 Baclofen, a GABA agonist, has been shown to be effective in the treatment of MS spasticity.4,5 Viruses have been implicated in the cause of MS,6 and studies in hamsters have shown that viral infection–induced granule cell loss in the cerebellum selectively depletes GABA receptor binding.7 γ-Aminobutyric acid has been shown to participate in the physiological regulation of the immune response8 and has been implicated in the tonic inhibitory control of myelinated fiber input.9
A multitude of different GABAA receptors exist in the brain due to the combinatorial assembly of various subunits encoded by at least 15 genes. Interneuronal signaling in the brain is characterized not only by a multitude of chemical signals but also by a diversity in the responsiveness of neurons to particular chemical signals. The differential responses are largely due to the heterogeneity of neurotransmitter receptors. A cell-specific expression pattern was visualized by immunofluorescence staining using antibodies to neurotransmitter markers and antibodies to receptor subunits (α1 and α3).10 Serotoninergic and cholinergic neurons selectively expressed the α3 subunit, whereas neighboring GABA neurons of similar size and morphologic features were characterized by the α1subunit.10 We have found a significant association between long and short alleles of different neuropsychiatric candidate genes and various quantitative behavioral traits.11- 13 These and other observations suggest the potential importance of these repeat polymorphisms in gene regulation and polygenic inheritance.14 We chose to examine the γ-aminobutyric acid A3 receptor gene (GABRA3) for 2 reasons. First, a tentative association has been observed between GABRA3 and age of onset of Alzheimer disease in women (D.E.C., unpublished data, 1997). These findings suggest that this gene might be relevant to other degenerative neurologic disorders as well. Second, GABRA3 is an X-linked gene (Xq28).15 If these effects were dominant, this could help explain why approximately 73% of persons with MS are female.
DNA was isolated from 189 MS brain samples obtained from the Human Neurological Research Specimen Bank at West Los Angeles Veterans Affairs Medical Center, Los Angeles, Calif. The samples were divided into 2 groups based on their case number (eg, MS group 1 consisted of all MS cases numbered 1278 or lower, and MS group 2 consisted of all subsequent cases). The 2 control samples consisted of 173 adult students recruited from a nearby university and 430 parents of twins from the Minnesota Twin and Family Study. All subjects were white. Brain samples were used for the MS subjects because a large number of specimens from persons with histologically proven MS severe enough to result in death were available from the Neurological Research Specimen Bank. Blood samples were used for the controls because they were healthy, unaffected individuals. Of course, DNA is identical regardless of the tissue of origin. As a further non-MS neuropsychiatric control, a group of 95 patients with schizophrenia were genotyped, many of whom were also located through the Neurological Research Specimen Bank.
Genomic DNA was extracted from brain samples and whole blood by standard procedures. A dinucleotide repeat polymorphism at GABRA3 was used for the genetic studies.15 The amplified product size ranged from 161 base pair (bp) to 171 bp and the primers annealed to the CA and GT strands. Polymerase chain reaction was used to amplify the target DNA, using 0.1 µm of each fluorescence-labeled primer
in separate reactions with 0.5 µL of the polymerase chain reaction product added to 2.8 µL of a mixture of 50 µL of deionized formamide plus 5 µL of internal standard (ROX500, Applied Biosystems Inc, Foster City, Calif) plus 5 µL of blue dextran dye, then denatured for 2 seconds at 92°C. The sample was loaded on a 6% polyacrylamide gel electrophoresis DNA sequencer (373 DNA sequencer, Applied Biosystems Inc) and gel was run for 5 hours at 1500 V with a constant of 30 W. The gel was preprocessed and analyzed using the internal standard (ROX500). Two peaks were recognized by software (Genotyper, version 1.1, Applied Biosystems Inc) based on the color and the size of the fragments.
The allelic distributions in the subjects with MS and control subjects from group 1, group 2, and the total of both are shown in Table 1. Allele frequencies are shown on the left while actual numbers are on the right. Using the χ2 test, we calculated the statistical significance of the difference in allele frequencies between subjects with MS vs controls in group 1, between the subjects with MS vs controls in group 2, and between the total subjects with MS and total controls groups. The distribution of the 6 alleles consisting of 11 to 16 dinucleotide repeats was significantly different for the subjects with MS vs the controls in group 1 (χ2=19.15; P<.002), the subjects with MS vs the controls in group 2 (χ2= 34.11; P <.001), and the total for both (χ2= 46.30; P<.001). The most marked differences were for the 16-repeat allele (Table 1) that was present in 17 (9%) of the 189 subjects with MS vs 2 (3%) of the 603 controls (χ2= 6.12; P<.001). To compensate for these differences, there was a modest increase in the frequency of the 11-, 12-, 13-, and 15-repeat alleles. The research was set up as a test and replication protocol. The first (test) set of subjects and controls (group 1) showed an increase in the frequency of the 16-repeat allele and a decrease in most of the remaining alleles (P<.002). The replication study (group 2) confirmed these results (P<.001). While such a protocol does not require a Bonferroni adjustment, even with such an adjustment using α=.05/3 or .016, the results are still significant.
The frequency of the 16-repeat allele in the 95 subjects with schizophrenia was identical to that in the total controls: 3%.
A genetic component in the susceptibility to MS has been demonstrated by a population-based study of MS in twins.16 The concordance rate for MS in monozygotic twins (25.9%) was significantly higher than that in dizygotic twins (2.3%) and nontwin siblings (1.9%). Recent genome-wide searches for MS-associated loci have implicated genes on chromosomes 2, 3, 5, 6, 11, 17, and X.17- 20 On the basis of the suggested links between GABA and MS, and for the reasons given earlier, we sought to examine the possible association between GABRA3 dinucleotide repeat polymorphism and MS. A strength of this study, decreasing the possibility that the results were due to chance, is the fact that the significant results of the initial study showing an increase in the frequency of the 16-repeat allele and relative decrease of other alleles (P<.002) was verified in a replication study (P<.001). In addition, a group of subjects with schizophrenia used as a neuropsychiatric control gave results identical to the controls. Finally, a relatively large number of subjects (189 in the MS group and 603 in the control group) were examined. Combined, these findings are consistent with the theory of GABRA3 as an X-linked susceptibility factor in MS.
In addition to those reviewed earlier, we examined several other explanations why GABRA3 might play a role as an MS susceptibility factor, including via an effect on glutamic acid decarboxylase (GAD), cytokines, or prolactin.
γ-Aminobutyric acid is synthesized from glutamic acid by GAD, the rate-limiting enzyme of this pathway.21 A negative feedback exists between GABA and GAD in which GABA can down-regulate GAD. Since GAD antibodies are associated with some autoimmune disorders, 22- 24 this could represent a connection between GABA and MS. To test this hypothesis, we examined a dinucleotide repeat marker in the GAD2 gene25 in the same MS and control subjects. The results showed a modest association among the different alleles that were of borderline significance (P<.05).
Various cytokines have been implicated in the cause of MS.26 An alternative explanation for the role of GABRA3 in MS could be related to the numerous studies showing an interaction between cytokines, nerve growth factors, and GABA metabolism.27- 30
An additional explanation of these results could be given if the GABA A3 receptor subunit were expressed on lymphocytes and thus played a direct role in modifying the immune response. While there is abundant evidence that benzodiazepine receptors are present on lymphocytes,31- 33 we are not aware of direct evidence that the GABA A3 receptors are expressed on lymphocytes. Benzodiazepine receptors are also present in platelets, but the absence of evidence of GABAA receptors suggests the 2 are dissociated in platelets.34 This could also be the case in lymphocytes.
In an earlier study it was observed that a significant association existed between the DRD2 haplotype and the age of onset of MS35 and hypothesized that this association was related to the important role of prolactin as an immunomodulator and the role of the D2 receptor in the control of prolactin release. Since GABA also plays a role in the regulation of prolactin release,36 this could also explain why variants at GABAA receptor genes, such as GABRA3, could also act as risk factors for MS.
Accepted for publication September 2, 1997.
Reprints: David E. Comings, MD, Department of Medical Genetics, City of Hope National Medical Center, 1500 E Duarte Rd, Duarte, CA 91910.