|Home | About | Journals | Submit | Contact Us | Français|
Dopa-responsive dystonia is a familial childhood-onset disease characterized by fluctuating dystonia, associated with tremor and parkinsonism in some patients. In most families the disease displays autosomal dominant inheritance due to mutations in the GTP cyclohydrolase 1 gene (GCH1). Penetrance and symptom severity display strong female predominance for which gender-specific GCH1 expression has been hypothesized. In this study, GCH1 mRNA expression was measured in cerebellar tissue from 66 healthy human subjects (30 women), and in cerebellar and nigral tissue from 8 individuals. No significant difference was found between men and women with small effect sizes observed. Although the correlation between cerebellar and nigral GCH1 expression remains to be further examined, this exploratory study does not support gender-specific GCH1 expression being the basis for the skewed gender distribution observed in DRD patients.
Dopa-responsive dystonia (DRD) is a familial childhood-onset disease characterized by fluctuating foot dystonia which tends to generalize and may lead to severe motor impairment [2, 9–11]. Additional symptoms include tremor and parkinsonism with variable combinations of bradykinesia, tremor, rigidity and postural instability. Response to levodopa is dramatic and long-lasting. Most patients display autosomal dominant inheritance of disease with reduced penetrance, due to mutations in the GTP cyclohydrolase 1 gene (GCH1; DYT5; MIM: 600225) on chromosome 14q22.1-q22.2 . GCH1 encodes an enzyme critical for tetrahydrobiopterin (BH4) synthesis. BH4 is a co-factor of tyrosine hydroxylase, the rate-limiting enzyme for dopamine synthesis. Mutations in GCH1 cause DRD by reducing endogenous dopamine production.
Gender specificity appears to be a consistent, albeit enigmatic finding in a variety of disorders associated with the dopaminergic system, including DRD, Parkinson’s disease and restless legs syndrome. In DRD families, the female-to-male ratio is generally around four, with disease penetrance reported to be 2.3 times higher in women than in men [3, 15]. Affected women also tend to display more severe symptoms than men . It has been suggested that such differences may stem from lower basal GCH1 expression in women. Herein, we provide evidence against this hypothesis by showing there is no gender difference in GCH1 mRNA expression in a series of cerebellar brain samples from healthy men and women. We also examine the correlation between nigral and cerebellar GCH1 expression in a limited sample.
Sixty-six human cerebellar brain samples were collected under approved ethical committee protocols from each institution (available upon request) (Table 1). Individuals who donated their brain had no known neurological condition and did not display pathologic abnormalities suggestive of Alzheimer’s disease or Lewy body disease. The project was conducted in accordance with the principles of the WMA Declaration of Helsinki. Subjects were free of neurological disease and had no personal or familial history of parkinsonism or dystonia. Total RNA was extracted from cerebellar tissue using TRIzol PureLink total RNA purification System (Invitrogen). RNA integrity was checked using the Agilent 2100 Bioanalyzer (Agilent) and expressed as a RNA quality score (RIN) from 1 to 10. Samples with a RIN < 5 are considered degraded, between 5.0 and 6.0 adequate, and > 6.5 of high quality. All samples had a RIN ≥ 5.2 and 76% had a RIN > 6.0. cDNA was synthesized from 1.5ug RNA with Applied Biosystems High Capacity cDNA Archive Kit and used as a template for relative quantitative PCR employing ABI Taqman® chemistry (Applied Biosytems). GCH1 expression was quantified using fluorescein (6-FAM)-labeled probes for GCH1 exon 1-exon 2 (X1X2, sequence available on request) and GCH1 exon 4-exon 5 (X4X5) boundaries (Hs00609198_m1). Results were normalized to the geometric means of three housekeeping genes, YWHAZ (Hs00852925_sH), GAPDH (Hs99999905_m1) and HPRT (Hs99999909_m1). Each sample was run in replicates of four on an ABI 7900 and analysis was performed using SDS2.2.2 software (Applied Biosystems).
Numerical variables were summarized with the sample median, 25th percentile, and 75th percentile. Linear regression models adjusted for age of death and RIN were used to investigate the association between gender and GCH1 expression. Estimated multiplicative effects and corresponding 95% confidence intervals (CI’s) were calculated. Estimated multiplicative effects are interpreted as the multiplicative increase on the median GCH1 expression, resulting from the exponentiation of the least squares estimates from a model where GCH1 expression was considered on the natural logarithm scale due to its skewed distribution. Pearson’s correlation was used to examine the correlation between GCH1 expression in the cerebellum and substantia nigra. Statistical significance was determined at the 5% level.
The analysis adjusted for age and RIN did not show any significant influence of gender on GCH1 expression in human cerebellar samples, with small effect sizes observed (Table 2, Figure 1). Estimated median GCH1 X1X2 expression was 1.07 times higher (95% CI: 0.82 – 1.39, P=0.63) in men than in women, while median GCH1 X4X5 expression was 1.05 times higher (95% CI: 0.81 – 1.37, P=0.69) in men compared to women. Likewise, age of death did not modify GCH1 expression (Table 2). There was evidence of a negative influence of the RIN on GCH1 expression (Table 2). Thus it was of interest to consider very high quality samples separately and perform a secondary analysis only in the 41 samples with a RIN ≥ 6.5, using linear regression models adjusted only for age of death. Lack of association between gender and GCH1 expression was also observed in this subgroup of subjects for GCH1 X1X2 (P=0.77) and GCH1 X4X5 (P=0.89), with similar effect sizes (1.05 and 1.02, respectively). Also, although our analysis was adjusted for age, there was a significant difference in age at death between men and women (Table 1). Therefore, we performed a sensitivity analysis using a subset of 38 individuals (19 men, 19 women) matched for age (± 3 years), and found no gender-specific difference in GCH1 expression (data not shown).
Nigral samples (tissue prepared from the substantia nigra exclusively) were available in a subset of eight individuals (two women; median age: 82 years, 25th–75th percentiles: 80–85 years; median RIN for cerebellum: 6.0, 25th–75th percentiles: 5.6–6.8; median RIN for nigra: 6.3, 25th–75th percentiles: 5.5–6.6). Exploratory analysis in these subjects showed cerebellar expression of GCH1 is representative of GCH1 expression in the substantia nigra, with similar results for GCH1 X1X2 (correlation coefficient: 0.71, P=0.12) (not shown) and GCH1 X4X5 (correlation coefficient: 0.69, P=0.059) (Figure 2). Despite relatively high correlation coefficients, P-values did not reach statistical significance in this small sample with low power to detect such correlations.
Animal studies have shown GCH1 mRNA expression [8, 14] and GCH1 protein levels  are low in nigrostriatal neurons compared to other monoaminergic neurons in rats [6, 8] and in mice . Similar low expression of GCH1 mRNA in nigrostriatal neurons has been reported in humans . Furthermore, in monoaminergic neurons, basal GCH1 mRNA expression was lower in female mice compared to males, which has led to the hypothesis that gender-specific basal GCH1 expression in humans may explain higher female susceptibility to develop DRD symptoms . Possible mechanisms include estrogen-mediated regulation of GCH1 transcription, as estradiol was shown to modulate GCH1 mRNA levels in rats  and to attenuate the GCH1 promoter response to cyclic AMP in PC12 cells . However, the present study does not support this hypothesis, as cerebellar GCH1 mRNA expression was similar in healthy men and women, with no significant difference observed. These results are in agreement with data from GCH1 enzymatic activity in stimulated mononuclear blood cells from healthy controls showing no gender-related difference .
Although our analysis included adjustment for age, all our samples came from individuals over age 49. Age-related changes in GCH1 activity in stimulated mononuclear blood cells have been reported in healthy controls . Likewise, a study found a significant age-related reduction of the number of GCH1-immunoreactive nigral neurons in aged humans (81.5%) and non-human primates (67.4%) compared to young cohorts . Therefore, our study may have overlooked GCH1 expression differences in younger individuals, particularly in women given estrogen influence over GCH1 transcription. However, a putative age-dependent gender specificity of GCH1 expression would not account for susceptibility to DRD in which symptoms worsen or are stable over time. An additional caveat may be the use of cerebellar tissue, as cerebellar GCH1 expression may not adequately reflect that of the substantia nigra. However, exploratory data in a limited number of samples indicates cerebellar GCH1 expression may be representative of nigral GCH1 expression; therefore use of cerebellar samples would appear to be a reasonable surrogate. Finally, sample quality is unlikely to have influenced the results as data analysis included adjustment for RNA quality, and a secondary analysis was performed only in samples with the highest RNA quality. The present study did not identify gender-specific differences in GCH1 expression; however, other mechanisms may explain the skewed gender distribution of DRD among GCH1 mutation carriers. Variants in GCH1 un-transcribed regulatory regions may induce post-transcriptional GCH1 mRNA modifications, thereby reducing GCH1 protein levels. Alternatively, genetic susceptibility may stem from differential expression of other genes implicated in the dopaminergic system by means of gene-gene, gene-mRNA or gene-protein interactions. Gender-specific striatal sensitivity to dopamine depletion or differences in dopaminergic system compensatory mechanisms may play a role, although this would likely also alter the risk of other conditions such as Parkinson’s disease.
C.W. is supported by the Swiss National Science Foundation (PASMP3-123268/1), the Swiss Parkinson Foundation, and the Robert H. and Clarice Smith and the M.L. Simpson Foundation Trust. M.J.F., Z.K.W. and D.W.D. are supported by the Morris K. Udall Center of Excellence for Parkinson’s Disease Research (P50-NS40256), and by the Pacific Alzheimer Research Foundation (PARF) grant C06-01. G.K. is supported by grant NS26081 from the NINDS.