Impacts of CR1 genetic variants on cerebrospinal fluid and neuroimaging biomarkers in alzheimer’s disease

Background The complement component (3b/4b) receptor 1 gene (CR1) gene has been proved to affect the susceptibility of Alzheimer’s disease (AD) in different ethnic and districts groups. However, the effect of CR1 genetic variants on amyloid β (Aβ) metabolism of AD human is still unclear. Hence, the aim of this study was to investigate genetic influences of CR1 gene on Aβ metabolism. Methods All data of AD patients and normal controls (NC) were obtained from alzheimer’s disease neuroimaging initiative database (ADNI) database. In order to assess the effect of each single nucleotide polymorphism (SNP) of CR1 on Aβ metabolism, the PLINK software was used to conduct the quality control procedures to enroll appropriate SNPs. Moreover, the correlation between CR1 genotypes and Aβ metabolism in all participants were estimated with multiple linear regression models. Results After quality control procedures, a total of 329 samples and 83 SNPs were enrolled in our study. Moreover, our results identified five SNPs (rs10494884, rs11118322, rs1323721, rs17259045 and rs41308433), which were linked to Aβ accumulation in brain. In further analyses, rs17259045 was found to decrease Aβ accumulation among AD patients. Additionally, our study revealed the genetic variants in rs12567945 could increase CSF Aβ42 in NC population. Conclusions Our study had revealed several novel SNPs in CR1 genes which might be involved in the progression of AD via regulating Aβ accumulation. These findings will provide a new basis for the diagnosis and treatment AD.

1. We found that five SNPs were linked to Aβ accumulation in brain. 2. The rs17259045 decreased Aβ accumulation among AD patients. 3. The rs12567945 could increase CSF Aβ42 in NC population.

Background
Alzheimer's disease (AD) has been regarded as a neurodegenerative disease of the elderly, which has accounted for 47 million people worldwide with numbers predicted to rise double by 2030 and triple by 2050 [1]. As one of the most common dementia, AD has the characteristics of poor language, memory, perception, behavior and activities of daily living. Moreover, the extracellular neurotoxic amyloid-β (Aβ) plaques and intracellular neurofibrillary tangles have been regarded as the neuropathological hallmarks of AD [2]. It has been widely confirmed that AD is a multifactorial disease, and genetic factors is proved to play a vital role in AD [3,4]. However, in spite of the progress in understanding risk factors related to AD development, the underlying mechanisms involved in this disease have not been completely understood till now, and to date there is no curative treatment for AD [5,6]. Now many genes are proved to significantly influence AD risk, among which the complement component (3b/4b) receptor 1 gene (CR1) has been proved to affect AD susceptibility across different ethnic and districts groups [7][8][9][10][11][12]. Currently, CR1 has been postulated to be a key factor for AD pathogenesis because of its role in regulating complement activity by acting as a receptor of complement C3b protein [13]. More importantly, in AD patients, CR1 is found to be associated with neuronal death [14] and hence has received increasing attention. Although a significant association between AD and single nucleotide polymorphisms (SNPs) in several novel AD loci of large case-control datasets is identified, CR1 is considered as one of the most important genetic susceptibility loci in AD according to the Alzgene database [15][16][17]. As well known, accumulation of Aβ in brain is one important pathological hallmark of AD, moreover, it is considered to induce a deleterious neurodegenerative cascade and finally cause cognitive impairments [18]. Furthermore, it has been shown that CR1 takes part in AD pathology by regulating the amyloid protein (Aβ) metabolism [19], and Johansson et al. [20] reveals that the single nucleotide polymorphisms (SNPs) in CR1 gene were associated with increased erythrocyte CR1 which will finally decreased AD risk. Hence, it would be meaningful to discover the genetic variants of CR1 in AD development.
In this study, we enrolled the participants from the alzheimer's disease neuroimaging initiative (ADNI) database (http://www.loni.ucla.edu/ADNI), which is a multicenter project to assess the role of genetic factors in neuroimage biomarkers and cerebrospinal fluid (CSF) proteins. Next, we used PLINK software to conduct the quality control procedures to enroll appropriate SNPs in CR1, and then investigated genetic influences of CR1 gene on Aβ metabolism, in order to explore the role of CR1 genetic variants in the progression of AD.

Participants
The data in our study were obtained from the ADNI database, which contains genetic information, neuroimaging information, and CSF proteins of AD, and normal controls (NC) (http:// www.adni-info.org). All participants of this study were included with the specific criteria according the protocol of ADNI, and then divided into two groups, including the AD group and NC group. Briefly, when participants met the National Institute of Neurological and Communicative Disorders (NINCDS) and Stroke/Alzheimer's Disease and Related Disorders Association (ADRDA) criteria for probable AD [21], they were diagnosed as AD.

Genotyping data
All genetic information of SNPs of CR1 were detected using the Illumina Infinium Human610-Quad Bead Chip (Illumina, Inc., San Diego, CA) or Illumina Human Omni Express Bead Chip. And the quality control procedures were performed by using PLINK software. The SNPs would be excluded when minimum minor allele frequency (MAF) was less than 0.01 or Hardy-Weinberg (H-W) equilibrium test's value was less than 0.05.

AV45-pet
The imaging data of PET with amyloid tracer, florbetapir (AV-45), was obtained from UC Berkeley-AV45 analysis database [22]. In order to define cortical grey matter regions of interest, these images were segmented and parcellated with Freesurfer (Version 5.3.0). After that, four regions, including the frontal, cingulate, parietal, temporal and florbetapir were involved in this study [23]. In addition, through averaging across the four cortical regions and dividing it by whole cerebellum florbetapir, the cortical standardized uptake values ratios (SUVR) were calculated [24]..

CSF Aβ 42 proteins
Similarly, the data about the level of CSF Aβ 42 was also got from ADNI database. Briefly, all samples of CSF were collected and transported to ADNI Biomarker Core laboratory at the University of Pennsylvania Medical Center. Following thawed at room temperature and gentle mixed, these samples were used for preparation of aliquots (0.5 ml). Finally, the level of CSF Aβ 42 was determined with multiplex xMAP luminex platform (Luminex Corp, Austin, TX) with immunoassay kit according to reagents [25] .

Statistical analyses
All statistical analyses were determined by using the SPSS 18.0 software (SPSS Inc., Chicago, IL, USA) and PLINK (http:// pngu.mgh.harvard.edu/wpurcell/plink/). The demographic characteristics were performed with means ± standard deviations (SD). The t-test or chisquare test were used for the analysis of demographics and genotypic frequencies. The correlation between CR1 genotypes and Aβ metabolism in all cohorts were estimated with multiple linear regression models. The false discovery rate (FDR) test was applied to control for multiple hypothesis testing [26], and a P ≤ 0.05 was considered to be statistically significant.

Characteristics of included participants
As shown in Table 1, a total of 329 individuals (48 AD and 281 NC) were enrolled in our study according to the quality control for genotype. Moreover, the AD group with 70.8% has higher frequency of the ε4 allele within apolipoprotein E (ApoE) gene than the NC group with 26.3%. According to the scores of different neuropsychological scales, the patients with AD have worse cognitive function in comparison to those NC group, respectively.

Characteristics of included SNPs of CR1
After quality control with PLINK software, a total of 83 SNPs of CR1 were enrolled in our study. Next, we used Haploview version 4.2 to explore the linkage disequilibrium (LD) patterns of these included SNPs of CR1 (Supplementary Fig. 1). The results showed these SNPs distributed from block 1 to 5 which indicated the SNPs capture most common variants in CR1. Furthermore, the characteristics (major allele, minor allele, MAF, functional consequence, position and H-W value) of included CR1 SNPs were illustrated in supplementary Table 1.
The MAF values of all included SNPs were more than 0.01, and the H-W values of included SNPs were more than 0.05.

Discussion
In our study, we explored the relation between whole CR1 genetic variants and Aβ metabolism biomarkers, and the results showed that five SNPs, including rs10494884, rs11118322, rs1323721, rs17259045 and rs41308433 could significantly alter Aβ accumulation in brain. In further analyses, the results suggested rs17259045 might decrease Aβ accumulation among AD patients. In addition, the genetic variants in rs12567945 would increase CSF Aβ 42 in NC population.
As we all known, Aβ is one important pathological characteristic of AD [27], which may induce the activation of the classical complement pathway in AD brains [28,29]. Moreover, CR1 is a necessary component of complement system, and it has been reported to have a close connection with amyloid plaque burden during aging [30,31]. More importantly, CR1 genetic variants are found to link to intelligence decline, and may influence the eliminations of Aß plaques [30]. Hence, it is urgent to investigate whether CR1 polymorphisms take part in the pathogenesis and development of LOAD. Actually, previous studies have revealed the association between CR1 SNPs and amyloid plaque [30,[32][33][34], including the CSF Aβ levels [35][36][37]. However, the current studies only discuss the role of specific SNPs (rs6656401, rs3818361, rs670173, and rs1408077) in Aβ metabolism. Briefly, rs6656401 and rs3818361, within the CR1 gene, have association with LOAD susceptibility in Caucasians [17], which are found to be in moderate LD (D′ = Fig. 1 The analysis about the associations of the variants and Aβ accumulation. The rs17259045 decreased Aβ accumulation in frontal a, temporal b, SUVR c, and cingulate d of AD patients 0.824, r 2 = 0.328) [38]. Specially, rs3818361 is found to be in block 1 [37]. In our study, the results firstly revealed that rs17259045 could reduce the level of Aβ accumulation among AD patients, respectively; moreover, rs12567945 could increase CSF Aβ 42 in NC population. In fact, rs17259045 was in the missense of CR1 gene, and rs12567945 located in the intron variant of CR1 gene. We speculated the genetic variants in the two SNPs might modulate the level of CR1, influence the activation of complement system, and finally alter the Aβ metabolism in the clearance of Aβin the brain. Taken together, these results indicated that the detection of variants in CR1 gene may be useful to diagnose AD timely, and it may be a useful method to treat AD via altering CR1 level.
Our previous study had reported that several volume (entorhinal, middle temporal, posterior cingulate, precuneus, parahippocampal), volume of subcortical (amygdale and hippocampus) and CA1 (the most associated area with the AD-specific amnenstic syndrome in hippocampus) were significantly related to AD [39]. However, our study failed to find the association between the genetic variants of CR1 (rs17259045 and rs12567945) and the above regions of interest via using ADNI data. As well know, one characteristic feature of synaptic function and density is cerebral glucose metabolic activity. Moreover, the change of glucose metabolic activity in specific brain regions could be valued via FDG PET [40]. Our study indicated that AD patents with genetic variants in rs17259045 might have more level of glucose metabolic activity in right angular (P = 0.03278). Hence, we hypothesized that genetic variants in CR1 might influence cognitive function (Fig. 3), through regulating CSF Aβ level, changing Aβ accumulations in brains, influencing the glucose metabolic activity, as well as altering the synaptic function and density.

Conclusion
In summary, our study found five SNPs (rs10494884, rs11118322, rs1323721, rs17259045 and rs41308433) were significantly linked to Aβ accumulation in brain. In further analyses of positive results, rs17259045 was found to decrease Aβ accumulation among AD patients. In addition, our study indicated genetic variants in rs12567945 would increase CSF Aβ 42 in NC population. Taken together, our study revealed some novel SNPs in CR1 which might be involved in AD development through regulating the Aβ pathology. However, several limitations still exist in this study. Firstly, the numbers of included samples were relative small. Secondly, our study was explored only in Caucasians. Hence, further study with larger samples and different ethnicities is still necessary.