Alzheimer's disease beta-amyloid peptide is increased in mice deficient in endothelin-converting enzyme.

The abnormal accumulation of beta-amyloid (Abeta) in the brain is an early and invariant feature in Alzheimer's disease (AD) and is believed to play a pivotal role in the etiology and pathogenesis of the disease. As such, a major focus of AD research has been the elucidation of the mechanisms responsible for the generation of Abeta. As with any peptide, however, the degree of Abeta accumulation is dependent not only on its production but also on its removal. In cell-based and in vitro models we have previously characterized endothelin-converting enzyme-1 (ECE-1) as an Abeta-degrading enzyme that appears to act intracellularly, thus limiting the amount of Abeta available for secretion. To determine the physiological significance of this activity, we analyzed Abeta levels in the brains of mice deficient for ECE-1 and a closely related enzyme, ECE-2. Significant increases in the levels of both Abeta40 and Abeta42 were found in the brains of these animals when compared with age-matched littermate controls. The increase in Abeta levels in the ECE-deficient mice provides the first direct evidence for a physiological role for both ECE-1 and ECE-2 in limiting Abeta accumulation in the brain and also provides further insight into the factors involved in Abeta clearance in vivo.

Alzheimer's disease (AD), 1 the most common cause of dementia in the elderly, is characterized pathologically by the accumulation of ␤-amyloid peptides (A␤40 and A␤42) in the brain. Although considerable attention has been focused on understanding the enzymes and processes involved in the production of A␤, very little is known regarding the processes by which A␤ is normally removed. This removal can be in the form of trans-port of the peptide into the cerebrospinal fluid for peripheral clearance, by binding to proteins that sequester the peptide in a nonreactive form, or by direct catabolism of A␤. A significant role for A␤ catabolism has been highlighted in a recent report by Saido and colleagues (1) in which they showed that infusion of the metalloprotease inhibitor thiorphan into the hippocampus of rats resulted in localized A␤ deposition, reportedly through the inhibition of A␤ degradation by neprilysin (NEP).
Using pharmacological, molecular, and biochemical approaches, we have previously characterized endothelin-converting enzyme-1 (ECE-1) as a novel A␤-degrading enzyme that appears to act intracellularly to reduce the amount of A␤ available for secretion (16). Overexpression of ECE-1 in cells that lack endogenous ECE activity results in a pronounced decrease in A␤ accumulation in the conditioned media, and ECE-1 can directly hydrolyze A␤ at multiple sites with a favorable kinetic profile. ECEs are type II integral membrane zinc metalloproteases (active site luminal) named for their ability to cleave and activate big endothelins (big ETs) to the potent vasoconstrictor endothelins (17). Two different ECEs have been cloned. The first, ECE-1, is abundantly expressed in the vascular endothelial cells of all organs and is also widely expressed in many nonvascular cells (18 -20). In the brain, ECE-1 immunoreactivity has been detected in fibers within the glial limitans, in neuronal processes and cell bodies of the cerebral cortex, in pyramidal cells of the neocortex and hippocampus, in astrocytes, and in Purkinje cells in the cerebellum (19,(21)(22)(23). ECE-2 is an homologous enzyme that shares catalytic activity with ECE-1 and is localized to an intracellular compartment (24). ECE-2 is most abundantly expressed in neural tissues including the cerebral cortex, cerebellum, and adrenal medulla (24). In mouse brain, ECE-2 is expressed in heterogeneous populations of neurons in the thalamus, hypothalamus, amygdala, dentate gyrus, and CA3 (25).
A reasonable test for the physiological involvement of ECE or any enzyme in contributing to A␤ accumulation in the brain is to examine the effect of animals that are null for the enzyme. Both ECE-1 and ECE-2 null animals have been described previously (25,26). ECE-1 homozygous knock-out mice have craniofacial and cardiac abnormalities, resulting in lethality beginning at day E12.5 and extending to within 30 min of birth (26). ECE-1 heterozygotes and ECE-2 null animals, however, do survive and appear healthy. In this study we examined the physiological role of ECE activity on A␤ accumulation in the brain by comparing the amount of A␤ in animals deficient in ECE with the amount found in the brains isolated from their wild-type littermates.

EXPERIMENTAL PROCEDURES
Animals-ECE-1 and ECE-2 knock-out mice were a gift from Dr. Masashi Yanagisawa. Offspring from heterozygous crosses were compared in all studies. ECE enzymatic levels were determined using a big * This work was supported by a grant from the Alzheimer's Association (to E. A. E.), by the generosity of Mr. and Mrs. Robert H. Smith, and by the Mayo Foundation for Medical Education and Research. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
DEA Extraction and A␤ Measurements-DEA extractions were performed using a protocol similar to those described previously (27,28). One hemisphere of the brain was homogenized in 50 mM NaCl, 0.2% DEA at a concentration of 100 mg/ml. The extract was then centrifuged at 4°C for 1 h at 100,000 ϫ g. The supernatant was removed and neutralized by the addition of a 1/10 volume of 0.5 M Tris-HCl, pH 6.8. Neutralized extracts (100 l) were added to assay plates containing 50 l of EC buffer (0.02 M sodium phosphate, pH 7.0 containing 0.002 M EDTA, 0.4 M NaCl, 0.2% bovine serum albumin, 0.05% CHAPS, 0.4% Block Ace, 0.05% NaN 3 ) and were analyzed using the well characterized BNT77/BA27 and BNT77/BC05 sandwich ELISA systems to detect A␤40 and A␤42, respectively (29). A␤ concentration was determined by comparing absorbance values obtained for samples to those obtained for synthetic A␤ standards prepared in EC buffer. Results from 2-3 experiments were pooled to generate the data shown by normalizing values to the means of the wild-type animals.
Western Blot Analysis of ␤APP and C-terminal Fragments-Neutralized DEA extracts were analyzed by Western blot as described previously (30). CT20, a rabbit polyclonal antibody raised against the 20 C-terminal amino acids of ␤APP, was used for quantitation of C-terminal fragments (CTFs), and 22C11 (Roche Molecular Biochemicals) was used for quantitation of ␤APP. Values reported are the mean from two scans each of two independent experiments, each performed in triplicate, and are presented in arbitrary units. In all cases multiple exposures were examined to make certain that the film was not "saturated." Values obtained were normalized to the mean of the wild-type animals to enable comparisons between experiments.
Statistical Analysis-Data were analyzed using the Mann-Whitney nonparametric test.

RESULTS
To determine whether reduced ECE-1 activity would lead to increased A␤ levels in the brain, we compared the amount of A␤ in brains from ECE-1 (ϩ/Ϫ) mice to that in wild-type littermate controls. The levels of both A␤40 and A␤42 were significantly increased in the mice with reduced ECE-1 activity ( Fig. 1 and Table I). Although modest, these increases are in the range observed for some of the familial Alzheimer's diseaselinked mutations (29,(31)(32)(33)(34)(35)(36) and are particularly noteworthy, as ECE-1 (ϩ/Ϫ) mice have only an ϳ27% reduction in ECE-1 activity (26 and data not shown) . To determine whether the increase in A␤ concentration we observed in the brains of these animals might be due to an increase in expression and/or processing of the A␤ precursor protein (APP), we examined the levels of APP and the C-terminal fragment of APP, which serves as the immediate precursor to A␤ (CTF␤). Consistent with the increase in A␤ concentration being a result of decreased degradation of A␤, the levels of APP and CTF␤ were unchanged in the ECE-1 (ϩ/Ϫ) mice compared with their wildtype littermates (Fig. 2).
Given the catalytic similarity between ECE-1 and ECE-2 and that ECE-2 expression is greater in the nervous system than in other tissues, we also examined whether A␤ concentration is affected by the level of ECE-2 activity in the brain. ECE-2 (Ϫ/Ϫ) animals are viable and appear healthy (25). Thus we were able to analyze both ECE-2 (ϩ/Ϫ) and ECE-2 (Ϫ/Ϫ) mice and compare the amount of A␤ in the brains of these animals to that found in littermate controls with normal ECE-2 levels. In these studies we found a significant elevation in the concentrations of both A␤40 and A␤42 in the brains of animals with reduced ECE activity compared with their control littermates ( Fig. 1 and Table I). The concentration of A␤ in the brains of ECE-2 heterozygous mice was intermediate between that observed in the wild-type controls and the complete nulls, suggesting a gene-dosage effect. As in the ECE-1 (ϩ/Ϫ) mice, the levels of APP and CTF␤ were unchanged in the ECE-2 (Ϫ/Ϫ) mice compared with controls (Fig. 2). DISCUSSION The increase in A␤ concentration in the brains of ECEdeficient mice provides direct evidence of a physiological role for ECEs in limiting A␤ accumulation in the brain. It will be important in future studies to determine whether these elevations will enhance deposition and plaque formation either by themselves or when crossed with mouse models that normally deposit A␤, such as the Tg2576 model (37). The same holds true for the neprilysin-efficient animals and for animals deficient in other enzymes that result in increased A␤ levels.
Given the evidence that A␤ plays a significant role in AD pathogenesis, factors that influence ECE activity, whether pharmacological, genetic, or other, may influence the risk for developing AD. The increases observed in the ECE-1 (ϩ/Ϫ) animals are particularly noteworthy, as these animals have only a 27% reduction in ECE-1 activity (26), suggesting that even modest changes in ECE activity can have significant effects on A␤ accumulation. It will be interesting in future studies to determine whether variations in ECE activity correlate with either soluble and/or insoluble A␤ levels in human brain. Reports have suggested the linkage of a possible AD locus to chromosome 3, where two genes involved in A␤ catabolism, neprilysin and ECE-2, are located (38 -40). Our results argue for the inclusion of ECE-2 in candidate gene approaches in this region.  1. Scatter plot of A␤40 detected in the brains of ECE-1 ؉/؊ and ECE-2 ؊/؊ mice. A␤ concentration was determined using the well characterized BNT77/BA27 sandwich ELISA. Bars indicate the mean value of each group. A, A␤40 concentration is significantly elevated in ECE-1 (ϩ/Ϫ) mice (p ϭ 0.0002) compared with wild-type littermate controls. B, A␤40 concentration is significantly elevated in ECE-2 (Ϫ/Ϫ) mice (p Ͻ 0.0001) compared with wild-type littermate controls.

A␤ Levels Are Increased in ECE-deficient Mice 2082
Perhaps most immediately important is that ECE inhibitors have received a significant amount of pharmaceutical interest for their potential as drugs for the treatment of hypertension and other ailments (41)(42)(43). Our results indicate that if these drugs enter the brain they may increase A␤ levels, potentially leading to the development and/or acceleration of AD in susceptible individuals. Based on the genetic mutations that cause AD, in which A␤ levels are elevated for decades prior to the development of the disease, this potential side effect of ECE inhibitors may not be observed for years and should be considered carefully before clinical use of this class of compounds. Ideally, these drugs should be examined directly for any effects on A␤ in preclinical studies.
Finally, our data suggest a possible novel approach for reducing A␤ levels in vivo by up-regulating ECE activity. This could be accomplished, for example, through gene therapy, transcriptional activation, or perhaps even by reducing ECE turnover itself. As with any approach, this strategy is not without potential problems, which will need to be examined carefully. One obvious concern with up-regulation of ECE is that patients may become hypertensive. However, up-regulation of ECE activity by intravenous injection of an adenoviral construct containing a secreted form of ECE does not result in increased circulating endothelin levels or in hypertension, indicating that ECE is not likely to be rate-limiting in the biosynthesis of ET, at least under these conditions (44). Because ECE has been shown capable of hydrolyzing several biologically active peptides in vitro, including bradykinin, neurotensin, and substance P (45,46), it will be important to monitor the levels of these potential targets as well.
In summary, our data provide direct evidence that ECE activity contributes to the steady state levels of A␤ in the brain and provide further insight into the factors involved in A␤ clearance. Thus, it appears that multiple enzymes are likely to be involved in A␤ catabolism in vivo. Catabolism of A␤ peptides by each of these enzymes would limit the accumulation of A␤, and disruption of this catabolism may be a risk factor for AD. Because removal of A␤ by multiple enzymatic activities appears to contribute to overall A␤ catabolism, the identification of these activities should be viewed as complementary rather than mutually exclusive. For example, certain enzymes such as the ECEs may degrade A␤ predominately intracellularly. Following secretion, other enzymes such as NEP may play a larger role. In addition, the principle mechanism of A␤ removal may vary by cell type and brain region and could potentially be influenced by other factors such as inflammation. Understanding these processes in more depth may provide new clues to the abnormal accumulation of A␤ found in patients with Alzheimer's disease.