BACE1 gene variants do not influence BACE1 activity, levels of APP or Aβ isoforms in CSF in Alzheimer's disease

The BACE1 gene encodes the beta-site APP-cleaving enzyme 1 and has been associated with Alzheimer's disease (AD). BACE1 is the most important β-secretase responsible for the generation of Alzheimer-associated amyloid β-proteins (Aβ) and may play a role in the amyloidogenic process in AD. We hypothesized that BACE1 gene variants might influence BACE1 activity or other markers for APP metabolism in the cerebrospinal fluid (CSF) and thereby contribute to the development of AD. We genotyped a Swedish sample of 269 AD patients for the rs638405 single nucleotide polymorphism (SNP) in the BACE1 gene and correlated genotype data to a broad range of amyloid-related biomarkers in CSF, including BACE1 activity, levels of Aβ40, Aβ42, α- and β-cleaved soluble APP (α-sAPP and β-sAPP), as well as markers for Alzheimer-type axonal degeneration, i.e., total-tau and phospho-tau181. Gene variants of BACE1 were neither associated with amyloid-related biomarkers, nor with markers for axonal degeneration in AD.


Findings
Cleavage of the amyloid precursor protein (APP) by βand γ-secretase gives rise to the amyloid β-protein (Aβ) found in senile plaques (SP) in Alzheimer's disease (AD). The BACE1 gene encodes the beta-site APP-cleaving enzyme 1 (OMIM 604252), which is involved in βsecretase activity [1][2][3][4]. The BACE1 gene is associated with AD [5,6] and the BACE1 activity is elevated both in brain tissue and in CSF in AD [7,8]. We hypothesized that BACE1 gene variants might influence the BACE1 activity or other amyloid-related biomarkers related to amyloid in the cerebrospinal fluid (CSF) and thereby contribute to developing AD. We tested a single nucleotide polymorphism (SNP) in the BACE1 gene to evaluate the genetic influence on BACE1 activity, levels of Aβ 40 , Aβ 42, αand β-cleaved soluble APP (α-sAPP and β-sAPP) in CSF from AD patients. Further we assessed the BACE1 genetic influence on markers for Alzheimertype axonal degeneration (total-tau and phospho-tau 181 ). The rs638405 SNP have a high allele frequency with a global frequency of the least common variant of 0.32.
To our knowledge this is the first study investigates BACE1 genotype data in relation to BACE1 activity and other amyloid-related biomarkers in CSF from AD patients.
We studied a Swedish Caucasian sample of 269 AD patients (90 men and 179 women, mean age 74.7 ± 6.3 years) where CSF levels of total-tau, phospho-tau 181 and Aβ 42 were known. We measured the BACE1 activity, levels of α-sAPP, β-sAPP and Aβ 40 in CSF samples from 84 of the patients. All participants were recruited at the Memory Clinic at the Malmö University Hospital. The patients gave informed consent to participate in the study, which was conducted according to the provisions of the Helsinki Declaration and approved by the local ethic committee. The diagnosis of "probable AD" was made according to the NINCDS-ADRDA criteria [9]. No AD patient had any of the known familial forms of autosomal dominant AD. CSF samples were taken by lumbar puncture. BACE1 activity was determined with a sensitive and specific solution-based assay previously described [10]. Levels of total-tau, phospho-tau 181 and Aβ 42 were measured using established ELISA methods [11,12] while α-sAPP, β-sAPP and Aβ 40 were quantified using MSD immunoassays (Cat#: K11120E and K111FTE), (Meso Scale Discovery, Gaithersburg, MD, USA).
APOE genotyping was performed using minisequencing as described before [14]. Gene designations follow the recommendations of HUGO Gene Nomenclature Committee [15].
The genotype and allele frequencies of the BACE1 rs638405 and APOE ε4 are shown in table 1. The analysis of variance (ANOVA) was used to analyze the effects of BACE1 genetic variants on MMSE, BACE1 activity and CSF protein levels. To test the effects of known risk factors, e.g., age, sex and APOE ε4 we identified significantly relevant covariates for each outcome variable (MMSE and levels of AD CSF biomarkers) using forward stepwise linear regression. Hardy-Weinberg equilibrium was assessed by χ 2 statistics. The criterion for significance was set at p < 0.05 for all statistical tests. Statistical analyses were performed with the SYSTAT11 (SYSTAT Software GmbH, Erkrath, Germany) software.
We studied BACE1 gene variants in relation to BACE1 activity and levels of amyloid-related biomarkers in CSF from AD patients. In the linear regression analysis we found APOE ε4 to significantly interact with phospho-tau18, Aβ 42 and MMSE. Subsequently, APOE ε4 was included as covariate in the statistic model. We found no effect of BACE1 gene variants on BACE1 activity or levels of Aβ 40 , Aβ 42, α-sAPP and β-sAPP (Table 2). We compared CSF levels of markers for axonal degeneration (total-tau and phospho-tau 181 ) between BACE1 gene variants and found no significant differences in protein levels ( Table 2). The BACE1 SNP showed no association with MMSE (data not shown).
The APOE allele distribution followed the expected frequencies seen in AD populations (Table 1). Genotype frequencies conformed to Hardy-Weinberg equilibrium in AD patients.
We hypothesized that BACE1 gene variants might influence BACE1 activity or levels of amyloid-related biomarkers in the cerebrospinal fluid (CSF) and thereby contribute to the development of AD. A genome-wide screen of AD families has shown linkage to marker close to BACE1 [16] and the rs638405 SNP has been tested in a number of studies [5,17]. Other studies have found association between the BACE1 gene and AD [5,18]. Even though this SNP does not change the protein sequence it still can have functional effects based on earlier findings where BACE1 influenced levels of Aβ in CSF [6]. We tested if the rs2069456 SNP was associated with BACE1 activity or with levels of Aβ 40 , Aβ 42, α-sAPP and β-sAPP in CSF from AD patients. Further we assessed the influence on total-tau and phospho-tau 181 . Gene variants of BACE1 were neither associated with amyloid-related biomarkers nor with markers for axonal degeneration. Our results do not support that variants of the BACE1 gene affects BACE1 activity or CSF levels of APP or Aβ isoforms in AD.

Acknowledgements
This work was supported by grants from Demensfonden.