HLA-DRB1 Alleles Are Associated with the Susceptibility to Sporadic Parkinson’s Disease in Chinese Han Population

Immune disorders may play an important role in the pathogenesis of Parkinson's disease (PD). Recently, polymorphisms in the HLA-DR region have been found to be associated with sporadic PD in European ancestry populations. However, polymorphisms in the HLA complex are highly variable with ethnic and geographic origin. To explore the relationships between polymorphisms of the HLA-DR region and sporadic PD in Chinese Han population, we genotyped 567 sporadic PD patients and 746 healthy controls in two independent series for the HLA-DRB1 locus with Polymerase chain reaction-sequence based typing(PCR-SBT). The χ2 test was used to evaluate the distribution of allele frequencies between the patients and healthy controls. The impact of HLA-DRB1 alleles on PD risk was estimated by unconditional logistic regression. We found a significant higher frequency of HLA-DRB1*0301 in sporadic PD patients than in healthy controls and a positive association, which was independent of onset age, between HLA-DRB1*0301 and PD risk. Conversely, a lower frequency of HLA-DRB1*0406 was found in sporadic PD patients than in healthy controls, with a negative association between HLA-DRB1*0406 and PD risk. Furthermore, a meta-analysis involving 195205 individuals was conducted to summarize the frequencies of these two alleles in populations from various ethnic regions, we found a higher frequency of HLA-DRB1*0301, but a lower frequency of HLA-DRB1*0406 in European ancestry populations than that in Asians, this was consistent with the higher prevalence of sporadic PD in European ancestry populations. Based on these results, we speculate that HLA-DRB1 alleles are associated with the susceptibility to sporadic PD in Chinese Han population, among them HLA-DRB1*0301 is a risk allele while the effect of HLA-DRB1*0406 deserves debate.


Introduction
Parkinson's disease (PD), the second most common human neurodegenerative disease, is characterized by progressive degeneration of dopaminergic neurons in the substantia nigra (SN) and other brainstem nuclei [1]. To date, the exact etiology of sporadic PD still remains unknown. Immune dysfunctions have been confirmed to be one of the causes of this disease [2][3][4]. In the brain of PD patients, large numbers of microglia expressing human leucocyte antigen (HLA)-DR have been detected in the SN, particularly in areas of maximal neurodegeneration [5]. The HLA-DR antigen, encoded by HLA-DRA and HLA-DRB alleles, belongs to HLA class II molecules and acts as an antigen-presenting molecule or regulatory molecule involved in the specific immune response and innate immune response [6]. The highly polymorphic HLA complex plays an important role in genetic susceptibility to human diseases, and different alleles of HLA loci are responsible to variable immune responses among individuals [7].
Associations between genotype of HLA loci and degenerative diseases in the central nerve system (CNS) have been found in many populations. For example, the association between HLA-A*01, A*2402, HLA-DRB1*03 or DRB1*1501 and high risk for Alzheimer disease has been reported [8][9][10][11]. Possible associations of HLA-A*2 and A*28, HLA-B*17 and B*18, and HLA-DQB1*06 with PD have also been reported [12][13][14][15]. However, in those studies, the enrolled subjects were relatively small by low-resolution genotyping analysis, and therefore, the conclusions could be ambiguous. Recently, Saiki et al. found that HLA-DRB1*03 was more common among British PD patients than healthy controls with low-resolution typing [16]. In addition, large-scale genome-wide association studies indicated that genetic susceptibility loci for late-onset sporadic PD might have existed in the HLA-DR region, such as rs3129882 in HLA-DRA and chr6:32588205 in HLA-DRB5 [17][18][19]. These studies suggest that the HLA-DR region is susceptive to the development of PD in European ancestry populations.
However, the HLA complex is highly polymorphic with different ethnic and geographic origins, PD genome-wide association studies (GWAS) study in Japanese shown no evidence of HLA association [20], while the relationship between HLA-DR polymorphism and PD in Chinese Han population has not yet been extensively investigated. Because HLA-DRB1 is the most polymorphic locus in the HLA-DR region [21], we systematically analyzed the polymorphism of HLA-DRB1 alleles among sporadic PD patients and healthy controls from a Han population living in the Guangdong province of the People's Republic of China (PRC) through polymerase chain reaction-sequencing based typing (PCR-SBT). The potential association of HLA-DRB1 alleles with PD in this population was evaluated, and gender and the onset age were taken into account.

Patients
A total of 567 sporadic PD patients and 746 unrelated healthy controls of Han population root from Guangdong province of the PRC were enrolled for this study(Basic information shown in Table 1). Each subject signed an informed consent before participating. The Ethics Committee of the First Affiliated Hospital of Sun Yat-sen University approved the protocol for this study. All PD patients were diagnosed according to the modified UK Brain Bank criteria [22]

DNA Extraction and HLA Typing
Whole blood samples were collected from PD patients and healthy controls. Genomic DNA was extracted from 4 ml of whole blood using the QIA-amp DNA Mini Kit (QIAGEN, German) according to the manufacturer's protocols. Extracted genomic DNA samples were stored at 2806C until genotyping was performed. The primer pairs recommended by the International Histocompatibilty Working Group (IHWG) were used to subtype the alleles of HLA-DRB1. HLA-DRB1 allele groups were typed by polymerase chain reaction-sequencing based typing (PCR-SBT) described previously [23]. Briefly, 22 group-specific primer pairs were used to amplify exon 2 of HLA-DRB1 using a thermal cycler (model 9700, Applied Biosystems, Foster City, CA, USA). PCR products were purified by incubating them with exonuclease at 376C for 30 min and then with shrimp alkaline phosphatase (SAP) at 806C for 15 min. We sequenced each purified PCR product using the ABI BigDye Terminatorv3.1 Cycle Sequencing Kit (Applied Biosystems, USA) and the ABI Prism 3100 Genetic Analyzer (Applied Biosystems). The sequence data were analyzed using the Match Tools and Navigator software (Match Tools Allele Identification package, Applied Biosystems, USA).

Statistic Analyses
Alleles were tested for deviation from Hardy-Weinberg equilibrium (HWE) using the SHEsis software platform [24]. Phenotype Frequency was defined as the total number of carriers in the population(Carriers/n), Allele Frequency was defined as the total number of copies of the allele in the population sample (Alleles/2n). The x 2 test was used to evaluate the distribution of allele frequencies between the patients and healthy controls, The Fisher exact test was used when expected frequencies were less than five, In addition, Bonferroni adjusted p values (pc) were used to avoid an alpha inflation when chaining statistical tests, the number of chaining statistical tests was defined as the number of the detected alleles. The association between PD and alleles was represented by odds ratio (OR), which was estimated by unconditional logistic models adjusted for the age of onset and gender. Statistical power of every allele can be calculated by the = the phenotype frequency of the allele in group1, p2 = the phenotype frequency of the allele in group2, n1 = the sample size of group1, n2 = the sample size of group2, p~n 1 Ã p1zn2 Ã p2 ð Þ = n1zn2 ð Þ , a = 0.05, Z a/2 = 1.960 (two sides). Calculations were carried out using SPSS-13.0 software (SPSS Inc., Chicago, IL). A pc,0.05, statistic Power.0.75 was regarded as statistically significant.

Results
The Distribution of HLA-DRB1 Alleles between PD Patients and Healthy Controls HLA-DRB1 genotype frequencies in both PD patients and healthy controls in two series were remained constant according to

The Relationship between HLA-DRB1*0301 and PD Risk
To further analyze the relationship between HLA-DRB1*0301 and PD risk, we compared the allele frequencies of HLA-DRB1*0301 between subgroups classified by the gender and the onset age of all subjects through 20 independent x2 tests and statistic power calculation described previously. We found that the allele frequencies of HLA-DRB1*0301 were higher in both PD subgroups (onset age #50 and onset age .50) than in their healthy control subgroups (0.0678 vs. 0.0263, p = 0.001, 0.0342 vs. 0.0213, p = 0.008, respectively), these differences were only occurred in male subgroups (0.0892 vs. 0.0286, p = 0.001, 0.0412 vs. 0.0230, p = 0.014, respectively). More interestingly, we found that the allele frequency of HLA-DRB1*0301 was higher in patients with an onset age #50 than in patients with an onset age .50 (0.0678 vs. 0.0342, p = 0.001 for all patients, 0.0892 vs. 0.0412, p = 0.003 for male patients and 0.0538 vs. 0.0265, p = 0.046 for female patients). After Bonferroni adjustment, the differences of allele frequency between PD subgroups (onset age #50) and healthy control subgroups (onset age #50) and PD subgroups (onset age .50) was statistically significant(Pc = 0.020 and Pc = 0.020, statistic Power .0.75, respectively) (Fig.1). Using stratification analyses and unconditional logistic regression analysis adjusted for gender, we found that the positive association between HLA-DRB1*0301 and PD risk was independent of onset age(onset age #50:OR     Figure 2&3).

Discussion
In this study, we found that HLA-DRB1 alleles may be contributing to susceptibility to sporadic PD in the Chinese Han population in the Guangdong Province of PRC. The allele frequency of HLA-DRB1*03 was significantly higher in PD patients than in healthy controls (0.0811 vs. 0.0449, pc = 1.789E-3, statistic power .0.75), which was identical with the report in European ancestry populations [16]. Furthermore, we found that HLA-DRB1*0301, the most common subtype of HLA-DRB1*03, had a strong association with PD development.
HLA-DRB1*0301 has been reported contribute to the genetic susceptibility of some autoimmune diseases such as type 1 diabetes and multiple sclerosis [53][54]. In our study, the allele frequency of HLA-DRB1*0301 was significantly higher in PD patients than in healthy controls (0.0811 vs. 0.0449, pc = 4.572E-3, statistic power .0.75). Furthermore, the allele frequency was significantly higher in patients with an onset age #50 than in patients with an onset age .50(0.0678 vs. 0.0342, pc = 0.012, statistic Power .0.75), suggesting that people carrying the HLA-DRB1*0301 allele may have a tendency for an early onset of PD. Molecules encoded by polymorphic HLA alleles elicit different T-cell killing effects by presenting different peptide-binding preferences [55]. In pulmonary sarcoidosis, patients carrying the HLA-DRB1*0301 allele have shown more AV2S3 positive T-cells, but less regulatory Tcells than those carrying other HLA-DRB1 alleles [56][57]. Interestingly, several studies reported that regulatory T-cells could mediate neuroprotection through modulation of microglia oxidative stress and inflammation, while profound T-cell responses could exacerbate neuroinflammation and induce dopaminergic neurodegeneration [58][59][60][61]. Thus, we speculated that the HLA-DRB1*0301 allele may be related to younger patients through eliciting strong inflammatory responses. The HLA-DRB1 *0406 allele, a subtype of HLA-DRB1*04, whether play a role to prevent the development of sporadic PD deserves debate. The HLA-DRB1*0406 allele is rare in European ancestry populations, but common in Asian ancestry populations . Previous studies have shown that it is associated with susceptibility to insulin autoimmune syndrome, silicosis, prostate cancer, and pemphigus in Asian ancestry populations [62][63][64][65], which implied that HLA-DRB1*0406 is a risk allele for immune dysfunction diseases. In our study, the allele frequency of HLA-DRB1*0406 in PD patients was significantly lower than in healthy controls (0.0013 vs.0.0308, pc = 3.295E-5, statistic Power .0.75) in series 1, but the difference was not achieve statistical significance in series 2 (0.0078 vs. 0.0267, p = 0.097, pc = 1), this inconsistency may be due to a sampling bias. We noticed that the allele frequencies of HLA-DRB1*0403, another subtype of HLA-DRB1*04, was presented higher in PD patients than in healthy controls in series 1 although the difference was not significant after Bonferroni adjustment (0.0307 vs. 0.0159, p = 0.035, pc = 1). There was no report that HLA-DRB1*04 have two opposing effects in the disease. Therefore, whether HLA-DRB1*0406 plays a unique neuroprotective role in the PD development of Chinese Han population still needs to be investigated.
After summarizing the allele frequencies of HLA-DRB1*0301 and HLA-DRB1*0406 in populations from various ethnic regions, we confirmed that the distribution of the two HLA-DRB1 alleles was highly polymorphic with different ethnic and geographic, this phenomenon was more obviously in vast and complicated races regions, such as mainland China, China Taiwan, and America. We also found a higher frequency of HLA-DRB1*0301, but a lower frequency of HLA-DRB1*0406 in European ancestry populations than in Asian ancestry populations. This was consistent with the higher prevalence of PD in European ancestry populations than in Asian ancestry populations [66][67][68][69]. However, we could not analyze the correlation between the allele frequencies of these loci and prevalence of sporadic PD because the allele frequency data was incomplete and the prevalence data was collected using different methodologies in the worldwide population.
We noticed that the allele frequencies of HLA-DRB1*1202, DRB1*1312, and DRB1*1454 were also higher in PD patients than in healthy controls, However, the difference was not significant after Bonferroni adjustment, the statistic power in those alleles are at moderate level (0.5, statistic Power ,0.75). Several studies reported these alleles unequivocally associated with immune-related diseases [70][71][72][73][74]. The possibility cannot be ruled out that these alleles may be either susceptibility genes or linkage disequilibrium with other susceptibility genes contributing to the etiology of PD. In addition, some rare alleles were detected because of random sampling, we lost most of the statistical power in these alleles due to their very small allele frequencies.
Recent studies have shown that some common HLA alleles can be marked with SNP-based tags [75][76][77][78][79][80]. We listed tag SNPs of the HLA-DRB1 alleles (Table 4), and compared it with the reported genetic susceptibility SNPs in the HLA-DR region to sporadic PD in European ancestry populations. We found that (i) these tag SNPs are all located at a considerable distance from the HLA-DRB1 allele. (ii) tag SNPs are likely to differ between populations, HLA-DRB1*0406 in JPT(Japanese in Tokyo) and HLA-DRB1*0403 in CHB(Han Chinese in Beijing) shared the same tag SNP:rs2395175, our study showed that HLA-DRB1*0403 was more common in PD patients than in controls, but the difference was no statistic significance, that partially explained why PD GWAS study in Japanese saw no evidence of HLA association even though the HLA-DRB1*0406 allele appears more frequent in Japan [20]. (iii) rs3129859 (tag HLA-DRB1*0701 in CHB), rs3129888(tag HLA-DRB1*0802,-DRB1*1406 in JPT), and rs3763313(tag HLA-DRB1*0803 in JPT) have been reported in association with the genetic susceptibility to PD in European ancestry populations [17][18][19], but there was no evidence showed that HLA-DRB1*0701, -DRB1*0802, -DRB1*0803,-DRB1*1406 were susceptibility loci to PD. The inconsistency may be arising from highly ethnic and region differences in HLA allele and SNP frequencies. In addition, reports on map of HLA alleles and its tag SNPs were few.
In conclusion, our study indicated that HLA-DRB1 alleles are associated with sporadic PD in a Chinese Han population, further research will be required to explore the role of HLA-DRB1 alleles in the pathogenesis of Parkinson' disease.