Amyloid beta peptide load is correlated with increased beta-secretase activity in sporadic Alzheimer's disease patients

Abstract

Whether elevated beta-secretase (BACE) activity is related to plaque formation or amyloid beta peptide (Abeta) production in Alzheimer's disease (AD) brains remains inconclusive. Here, we report that we used sandwich enzyme-linked immunoabsorbent assay to quantitate various Abeta species in the frontal cortex of AD brains homogenized in 70% formic acid. We found that most of the Abeta species detected in rapidly autopsied brains (<3 h) with sporadic AD were Abeta(1-x) and Abeta(1-42), as well as Abeta(x-42). To establish a linkage between Abeta levels and BACE, we examined BACE protein, mRNA expression and enzymatic activity in the same brain region of AD brains. We found that both BACE mRNA and protein expression is elevated in vivo in the frontal cortex. The elevation of BACE enzymatic activity in AD is correlated with brain Abeta(1-x) and Abeta(1-42) production. To examine whether BACE elevation was due to mutations in the BACE-coding region, we sequenced the entire ORF region of the BACE gene in these same AD and nondemented patients and performed allelic association analysis. We found no mutations in the ORF of the BACE gene. Moreover, we found few changes of BACE protein and mRNA levels in Swedish mutated amyloid precursor protein-transfected cells. These findings demonstrate correlation between Abeta loads and BACE elevation and also suggest that as a consequence, BACE elevation may lead to increased Abeta production and enhanced deposition of amyloid plaques in sporadic AD patients.

Fig. 1.

Analysis of Aβ in formic acid extracts from AD and ND brains by ELISAs. Aβ_1-x, Aβ_x-42, Aβ_x-40, and total Aβ levels were obtained from temporal cortex in eight AD and eight ND cortex regions. (A) We used 4G8 as a capture antibody, and biotinylated 6E10 as a detection antibody for Aβ_1-x measurement. (B) To measure Aβ_x-42 or Aβ_x-40, we used 4G8 as a capture antibody and biotinylated AB42 and AB40 to assay Aβ_x-42 and Aβ_x-40, respectively. (C) Last, we used 6E10 as a capture antibody and biotinylated AB42 or AB40 for Aβ_1-42 or Aβ_1-40 assays.

Fig. 2.

Quantitative determination of BACE levels in AD brains by using BACE ELISA. The brain homogenate samples, including ones used for Western blot study from the temporal cortex (TC; n = 39) and hippocampus (HIPP; n = 8) of AD with short autopsy (< 3 h) were assayed by using BACE ELISA in a double-blinded manner. The capture BACE antibody was SECB1 (1:1000), and the detection antibody was the biotin-labeled antibody SECB2. The BACE-transfected cell lysates were used as a positive control, and BACE knockout cell lysates were used as negative control. The sensitivity is 8 pg/ml. All quantitations are the mean ± SD of at least three independent measurements. Significant correlation: ^*, P < 0.01; ^**, P < 0.001. Similar results were obtained by using SECB1 as a capture antibody, and the antibody against BACE C terminus as a detection antibody.

Fig. 3.

(A) BACE mRNA in AD brains examined by Northern blot. Total RNA was isolated from AD and ND brains. Twenty micrograms of total RNA was loaded in each lane. For all experiments, the 500-bp cDNA isolated from the initial two-hybrid screen was used as a BACE probe. The blot was sequentially hybridized with radiolabeled BACE and GAPDH under stringent conditions as described in Materials and Methods. (B) BACE1 mRNA expression in AD brains confirmed by RNase protection assay. The RNAs were hybridized with the BACE antisense probe. After treatment with RNase A and T1, protected bands were run through an 8% polyacrylamide gel, and the dried gel was exposed to x-ray film.

Fig. 4.

BACE enzymatic activities are increased in AD brains. The BACE activity was evaluated by using BACE cell lysates or AD-enzymatic crude extracts to incubate with fluorescent-labeled peptides bearing β-site APP, from either APP_WT or APP_sw. The specific fluorescent peptide was detected when labeled APPwt and APPsw peptides were cleaved by BACE. The measurement was performed in a double-blinded manner. Crude extract from AD brains showed significantly higher BACE activity in cleaving APPwt than that from ND brains. Moreover, a mixture of AD crude extracts and the fluorescent labeled APPsw peptide generated a significantly higher fluorescent fragment product than that of the mixture of APPsw and ND extract. All quantitations are the mean ± SD of at least three independent measurements. The value of the transfected BACE is considered as a positive control. Significant correlation: ^*, P < 0.01; ^**, P < 0.001.

Fig. 5.

BACE expression was analyzed in transfected cells. Poly (A)⁺ RNA was extracted from APPsw or mutated PS1-transfected cells. The multiple group blot containing 2 μg of poly (A)⁺ RNA per lane was hybridized with a ³²P-labeled fragment containing human BACE cDNA probe, according to the manufacturer's protocol. The blot was then stripped and rehybridized with ³²P-labeled human GAPDH probe as a control.

Fig. 6.

Linear regression analysis of BACE, plaque numbers, and disease history. The activity level of BACE in the temporal cortex of AD brains is positively correlated with plaque scores. (A) Temporal cortex samples from clinically diagnosed, neuropathologically confirmed AD patients were rapidly autopsied brains with sporadic AD. Postmortem intervals averaged <3 h. We randomly selected 18 cases of AD subjects. Average ages from AD (X_AD) cases are 76 ± 4 years old and from ND cases are (X_ND) 74 ± 6 years old. There is also a positive correlation between the level of BACE enzymatic activity and the disease history in the individual subject. (B) The disease duration of AD from these patients is 7.5 ± 1.3 years.