Analysis of (CAG)n expansion in ATXN1, ATXN2 and ATXN3 in Chinese patients with multiple system atrophy

Multiple system atrophy (MSA) is a complex and multifactorial neurodegenerative disease, and its pathogenesis remains uncertain. Patients with MSA or spinocerebellar ataxia (SCA) show overlapping clinical phenotypes. Previous studies have reported that intermediate or long CAG expansions in SCA genes have been associated with other neurodegenerative disease. In this study, we screened for the number of CAG repeats in ATXN1, 2 and 3 in 200 patients with MSA and 314 healthy controls to evaluate possible associations between (CAG)n in these three polyQ-related genes and MSA. Our findings indicated that longer repeat lengths in ATXN2 were associated with increased risk for MSA in Chinese individuals. No relationship was observed between CAG repeat length in the three examined genes and age at onset (AO) of MSA.

MSA-P. This study also included 314 healthy controls with no history of neurodegenerative disease or other diseases.
The control and patient groups were matched with respect to age (age range, 30-74 years; mean age, 53.5 ± 7.6 years), sex ratio (203 males, 111 females), and region of residence. All patients and controls were from the Han Chinese population. The study was approved by the Ethics Committee of Xiangya Hospital, and written informed consent was obtained from all participants. DNA analysis and genotype classification. Genomic DNA was extracted from peripheral blood using standard phenol-chloroform extraction procedures 14 . Genotyping of ATXN1, 2 and 3 was determined by polymerase chain reaction (PCR) amplification of CAG tracts in combination with capillary electrophoresis, using GeneMarker software (SoftGenetics). At the SCA1, SCA2 and SCA3/MJD loci, the allele containing the larger repeat was designated the 'long' allele, and the other allele was regarded as the 'short' allele. The short and long alleles were considered separately in statistical models. With respect to the multimodal or skewed distributions in Fig. 1, repeats in long alleles of ATXN2 and ATXN3 were classified as short, short-medium, medium or long in accordance with the approach described in previous studies 15,16 . For ATXN1, since the size of the long allele exhibited a nearly normal distribution, we divided long alleles into two groups, short (≤29 repeats) and long (å 29 repeats), based on mean repeat size. Statistical analysis. Differences in age and sex between the patients with MSA and the controls were assessed using a t-test and a chi-square test. Descriptive statistics are expressed as mean ± standard deviation ( Table 1). Associations between size of (CAG) n and risk for MSA were determined via logistic regression, adjusting for age and sex. We used one-way factorial analysis of variance (ANOVA) or the Kruskal-Wallis test to investigate the association between (CAG) n size and AO for the patients with MSA. The Mann-Whitney U test or t-tests were used to test for differences in repeat length between patients with MSA-C and patients with MSA-P. A two-tailed P-value ≤ 0.05 was regarded as significant. Table 1 summarizes demographic data for all 514 participants. No pathological (CAG) n expansion in the three SCA genes was detected (in either patients or controls). Distributions of (CAG) n size are shown in Fig. 1. In ATXN1, the mean size was 29.3 ± 1.6, ranging 25-35 repeats in patients and 29.4 ± 1.6 in controls, ranging 25-36 repeats. In ATXN2, the mean size was 22.4 ± 1.6, ranging 22-31 repeats in patients and in controls 22.1 ± 0.7, ranging 22-29 repeats; 8 patients and 3 controls carried long expansions (Supplementary Table S2). In ATXN3, the mean size was 25.6 ± 5.9, ranging 17-38 repeats in patients and in controls 26.2 ± 6.1, ranging 17-40 repeats.

Results
For the ATXN2 locus, there was a significant difference in the distribution of CAG repeats between patients and controls (P = 0.011, OR = 1.253, 95% CI = [1.052-1.492]) ( Table 2). The number of CAG repeats in ATXN1 and ATXN3 did not significantly differ between patients with MSA and controls. There were no significant correlations between AO for MSA and repeat length in ATXN1, 2 or 3 (Table 3). In addition, the distributions of   (CAG) n size in any of these three genes did not significantly differ between patients with MSA-C and patients with MSA-P (Supplementary Table S1).

Discussion
None of our patients exhibited pathogenic expansion in any of the three examined polyQ-related genes, indicating that such expansion may not be a causative factor for MSA. Nevertheless, we found a significant association between CAG repeat sizes in ATXN2 and risk for MSA. The most common size (over 95%) of the (CAG) n in SCA2 is either 22 or 23 (range, 14-32) [17][18][19][20] . Ataxin2, which is encoded by ATXN2, is localized to the rough endoplasmic reticulum and plays a critical role in mRNA processing 21 . In the pathogenesis of SCA2, polyQ expansion of ataxin2 confers a gain-of-function mutation that induces neuronal impairment and triggers the disease phenotype 21 . Ataxin2 is also closely related to other neurodegenerative diseases, such as ALS, PD and spinocerebellar ataxia type 3 (SCA3/MJD) [22][23][24][25] . Functional studies have proven that ataxin2 interacts with TDP-43 via joint mRNA binding, aggravating TDP-43 toxicity and thereby further increasing the risk of developing ALS 22 . In a yeast model, ataxin2 was shown to be a modifier of α-synuclein biotoxicity in specific molecular pathways and a predictive nodal point in the α-synuclein network 6 . As an mRNA-related translation factor, ataxin2 has been associated with α-synuclein toxicity in neurons of patients with PD 6 . The neuropathological hallmark of MSA is the presence of glial cytoplasmic inclusions (GCIs) containing α-synuclein; given this characteristic, MSA can be regarded as a synucleinopathy, together with PD and Lewy body dementia (DLB) [26][27][28] . We can speculate that ATXN2 increases the risk for MSA by perturbing mRNA metabolism and translation and thereby influencing α-synuclein biotoxicity.
An association was found between MSA and CAG repeat sizes in ATXN2 but not CAG repeat size in ATXN1 or ATXN3, implying that ATXN2 may play a role as a risk factor for MSA (at least) in the Chinese population; however, no modifying effects of repeat lengths at SCA1, SCA2 or SCA3/MJD loci on AO of MSA were observed, possibly due to sample size. The small number of ATXN2 long-expansion, over 26 CAGs, found in patients and controls (due to their rarity in the general population) is, of course, a limitation that cannot be easily overcome. Further studies in different ethnic populations and a larger sample size are needed to confirm the present findings.
In terms of mechanisms for neurodegeneration, it would be important to shed some light on a possible pathogenic interaction between ataxin2 and α-synuclein in MSA. The genetic association between ATXN2 and MSA may contribute to a more comprehensive understanding of neurodegenerative disorders and to foster new therapies for such diseases.

Ethics statement. This study was approved by the Ethics Committee of Xiangya Hospital of Central South
University in China (equivalent to an institutional review board), and all methods were performed in accordance with approved guidelines. Written informed consent was obtained from all participants.