BACE1 (β-secretase) transgenic and knockout mice: identification of neurochemical deficits and behavioral changes

Abstract

BACE1 is a key enzyme in the generation of Aβ, the major component of senile plaques in the brains of Alzheimer's disease patients. We have generated transgenic mice expressing human BACE1 with the Cam Kinase II promoter driving neuronal-specific expression. The transgene contains the full-length coding sequence of human BACE1 preceding an internal ribosome entry site element followed by a LacZ reporter gene. These animals exhibit a bold, exploratory behavior and show elevated 5-hydroxytryptamine turnover. We have also generated a knockout mouse in which LacZ replaces the first exon of murine BACE1. Interestingly these animals show a contrasting behavior, being timid and less exploratory. Despite these clear differences both mouse lines are viable and fertile with no changes in morbidity. These results suggest an unexpected role for BACE1 in neurotransmission, perhaps through changes in amyloid precursor protein processing and Aβ levels.

Introduction

Alzheimer's disease (AD) is the most prevalent of all adult-onset neurodegenerative disorders and is characterized by a progressive memory loss that may also be associated with other neuropsychiatric behavioral impairments. The principle neuropathological features of this disease are overt neuronal and synaptic loss, extracellular accumulation of amyloid plaques, composed primarily of amyloid β-peptide (Aβ) deposits, and intracellular neurofibrillary tangles composed of hyperphosphorylated tau (see Hardy and Selkoe, 2002). These defining hallmarks of AD neuropathology have been the focus of intense research over the past two decades and have led to the identification of the secretase enzymes involved in Aβ production, as well as to the identification of a variety of kinases that can potentially hyperphosphorylate tau in vivo.

Aβ peptide is an internal peptide cleaved from a type 1 membrane-spanning precursor protein, the amyloid precursor protein (APP) through two sequential proteolytic cleavage events catalyzed by β- and γ-secretases. Several amyloid-depositing mouse models have been generated using different promoters, constructs, and APP mutations linked to familial forms of the disease (see Janus et al., 2000, Emilien et al., 2000b, van Leuven, 2000. In all of these models, there is an age-dependent increase in brain Aβ levels and concomitant amyloid plaque deposition associated with astrogliosis, microgliosis, and the appearance of dystrophic neurites Janus et al., 2000, Emilien et al., 2000b, Irizarry et al., 1997. There are also reports of reduced synaptophysin staining in PDAPP mice Games et al., 1995, Mucke et al., 2000 and some neuronal loss in APP23 mice (Bondolfi et al., 2002), although none of these mouse lines demonstrate a convincing level of neuronal degeneration. Behaviorally, several of the APP transgenic lines exhibit deficits in learning and memory Hsiao et al., 1996, Chapman et al., 1999, King et al., 1999, Pompl et al., 1999, Moechars et al., 1999, lending support to the notion that elevated Aβ levels and plaque deposition can cause behavioral changes despite a lack of neuronal loss.

The identity of the elusive β-secretase was uncovered by four independent groups using a variety of techniques Hussain et al., 1999, Vassar et al., 1999, Yan et al., 1999, Sinha et al., 1999. The molecular, enzymatic, and cell biological characterization of β-secretase APP cleaving enzyme 1 (BACE1) has significantly increased our understanding of the role of this enzyme in Aβ generation and presented the scientific community with a new therapeutic target for the treatment of AD. Further validation of BACE1 as a key enzyme in the generation of Aβ has been made possible using mice lacking BACE1. In such animals there is an almost complete inhibition of Aβ production, even when these animals are crossed with mice overexpressing APP transgenes (Luo et al., 2001). In addition, BACE1 knockout (KO) mice are viable and fertile, with an absence of any deleterious or adverse findings Luo et al., 2001, Cai et al., 2001, Roberds et al., 2001. This is in sharp contrast to mice lacking presenilin-1, one of the components of the γ-secretase enzyme complex Edbauer et al., 2003, Kimberly et al., 2003, which have an embryonic lethal phenotype (Shen et al., 1997). These observations, coupled to the clear correlation between Aβ burden and cognitive decline (Naslund et al., 2000), provide good evidence that inhibition of BACE1 could be of benefit to AD patients and potentially be less deleterious than approaches targeting inhibition of γ-secretase.

In addition to the neuropathological features described above, the brains of AD patients are deficient in a multitude of neurotransmitter systems. These include acetycholine and glutamate, upon which current symptomatic therapies such as cholinesterase inhibitors and NMDA antagonists are based, as well as noradrenaline, serotonin, and neuropeptides such as somatostatin and neuropeptide Y Arai et al., 1991, Chan-Palay et al., 1985, Richter et al., 1980. The evaluation of neurotransmitter deficits in transgenic mouse models of AD has only recently been undertaken. Several reports have suggested the existence of cholinergic deficits in APP transgenic mice Hernandez et al., 2001, Bednar et al., 2002. In addition, elevated levels of neuropeptides such as galanin, neuropeptide Y, cholecystokinin, and enkephalin have been reported in aged brains of PDAPP mice (Diez et al., 2000). However, the extent to which these changes are typical of mouse models of AD and their direct relevance to changes observed in AD patients remain to be investigated.

In this study we have generated a novel BACE1 knockout mouse, and also report the characterization of a BACE1 transgenic mouse. Both mouse lines were viable, fertile, and healthy. However, a detailed neurochemical and behavioral analysis revealed some interesting and novel phenotypes. Mice overexpressing BACE1 exhibited an increase in 5-hydroxytryptamine (5-HT) turnover and were less anxious and more exploratory in their behavior, consistent with the known correlation between anxiety and decreased 5-HT turnover (Deakin, 1998). In contrast, mice lacking BACE1 displayed an anxious and less exploratory phenotype. These results extend our current knowledge of BACE1 mouse models and reveal an unexpected link between BACE1, serotonergic neurotransmission, and behavior.