Transgenic mice overexpressing amyloid beta protein are an incomplete model of Alzheimer disease Claudia Schwab, Masato Hosokawa, and Patrick L. McGeer * Kinsmen Laboratory of Neurological Research, Division of Neurology, University of British Columbia, Vancouver, BC, Canada V6T 1Z3 Received 24 July 2003; revised 10 December 2003; accepted 12 March 2004 Available online 26 April 2004 Abstract We compared lesions in elderly transgenic (tg) mice carrying the Swedish double mutation KM670/671NL with lesions in Alzheimer disease (AD) by histochemical and immunohistochemical techniques. Highly similar staining for beta-amyloid protein (Ah) was observed in AD and the mouse models. The abundant amyloid deposits in tg mice were in a consolidated state as revealed by strong Congo red birefringence. In both tg mice and AD, amyloid deposits were ApoE-positive and were surrounded by activated astrocytes. However, Bielschowsky silver staining and immunostaining with tau antibodies revealed no neurofibrillary tangles (NFTs) in the mice as opposed to abundant NFTs in AD. The microglial pattern was also distinctly different. Tg mice had only weakly activated microglia, which expressed low levels of the complement receptor CD11b. They were gathered around the periphery of the deposits. In contrast, AD lesions had strongly activated microglia, which expressed high levels of CD11b. They were associated with the plaque core. Immunostaining for complement proteins was weak in tg mice but very strong in AD deposits. We conclude that the weak inflammatory response and absence of NFTs indicate that tg mice are only a limited model of AD. Therapeutic strategies for the treatment of AD based on tg mouse models that overexpress Ah may be limited in their application. D 2004 Elsevier Inc. All rights reserved. Keywords: Inflammation; Complement; Complement receptors; Vaccination; Microglia; Tau; Neurofibrillary tangles Introduction A new phase of Alzheimer disease (AD) research com- menced when the first transgenic (tg) mouse model demon- strating deposits of human beta-amyloid protein (Ah) in brain was reported. This was achieved by introducing into mice a human amyloid precursor protein (hAPP) minigene with the V717F mutation under control of the platelet- derived growth factor promoter (Games et al., 1995). It is described as the AD-APP model. Since then, highly compa- rable transgenic models have been developed (Hsiao et al., 1996, Sturchler-Pierrat et al., 1997). They involve introduc- ing a disease-producing mutation of human APP behind an aggressive promoter. A more severe model involves a double transgene in which a disease-producing mutation in prese- nilin-1 (PS-1) is added to the human APP transgene (Hol- comb et al., 1998). These models have in common the commencement of amyloid deposition starting at a young age with progressive accumulation until death. They also have in common a failure to develop neurofibrillary tangles (NFTs), which are an essential hallmark of AD. The amyloid cascade hypothesis holds that accumulation of Ah is the true cause of AD, with the neurofibrillary tangles and dystrophic neurites developing because of the Ah accumula- tion (Hardy and Selkoe, 2002). Arguments supporting this hypothesis are that APP mutations in or near the Ah segment produce autosomal dominant AD, as do PS-1 and PS-2 mutations which enhance the formation of Ah (Selkoe, 1999). The hypothesis is strengthened by experiments with the JNPL3 mouse line that is transgenic for the P301L tau mutation. This mutation is causative for one form of fronto- temporal dementia and parkinsonism linked to chromosome 17 (FTDP-17). These mice develop NFTs but not amyloid deposits (Lewis et al., 2000). When the mice are crossed with the tg2576 model expressing mutant APP, Ah deposits develop similarly to the tg2576 parents, but NFT formation is enhanced (Lewis et al., 2001). While these experiments indicate that amyloid deposits are associated with an in- 0014-4886/$ - see front matter D 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.expneurol.2004.03.016 * Corresponding author. Kinsmen Laboratory of Neurological Re- search, Division of Neurology, The University of British Columbia, 2255 Wesbrook Mall, Vancouver, BC, Canada V6T 1Z3. Fax: +1-604-822-7086. E-mail address: mcgeerpl@interchange.ubc.ca (P.L. McGeer). www.elsevier.com/locate/yexnr Experimental Neurology 188 (2004) 52 – 64