Learning and Memory Deficits in APP Transgenic Mouse Models of ...

Neurochemical Research, Vol. 28, No. 7, July 2003 (© 2003), pp. 1029–1034

Learning and Memory Deficits in APP Transgenic Mouse Models of Amyloid Deposition* Dave Morgan1,2 (Accepted September 3, 2002)

Several different transgenic APP mice develop learning and memory deficits. In some cases the mice have deficits very early in life, while in other instances the mice exhibit deficits only after they have aged and amyloid deposits have accumulated. In many cases, there is a correlation in individual mice of the same age and genotype between the extent of learning and memory deficits and the amounts of deposited amyloid found in the central nervous system. While superficially this might imply that the deposited material is somehow toxic to cognition, it is likely that deposited amyloid is also an index of the overall rate of amyloid production in each mouse. Rate of production would be expected to modify not only the amounts of deposited amyloid, but also other amyloid pools, including soluble, oligomeric, conjugated (e.g. ADDLs) and intracellular. Thus, the deposited material may be an integrated reflection of total Aß production, in addition to indicating the amounts in fibrillar forms. As such, it is conceivable that other Aß pools may be more directly linked to memory deficits. Thus far, the one manipulation found to mitigate the learning and memory deficits in APP transgenic mice is immunotherapy for Aß, either using active or passive immunization against the peptide. These data together with other findings are leading to a conclusion that the fibrillar Aß deposits are not directly linked to the memory deficits in mice, and that some other Aß pool, more readily diminished by immunotherapy, is more directly linked to the mechanisms leading to poor performance in learning and memory tasks.

KEY WORDS: Memory; Aß amyloid; water maze; transgenic mice.

INTRODUCTION

to develop such transgenic models in which to experimentally study the direct effects of amyloid (2). One question not addressed in the PDAPP mouse described by Games et al. regarded the presence of behavioral deficits. The first communications indicated that the PDAPP mice were unable to perform in standard rodent learning and memory tasks even at rather young ages (3,4). This made it difficult to determine if there was a correlation between A load and memory, as the mice were performing at basement levels even without visible amyloid deposits. Another problem in ascertaining behavioral deficits in these mice was the limited numbers of investigators permitted access to the animals. Although these restrictions have improved over the years, it is still

In 1995, Games et al. (1) characterized the first APP transgenic mouse model that succeeded in depositing A into structures that in many ways resembled those found in Alzheimer disease brain tissue. This ended several years of frustration in attempting * Special Issue: Application of genetically altered mice models and their significance in neurodegenerative disease. 1 Alzheimer Research Laboratory, Department of Pharmacology, University of South Florida, Tampa, Florida 33612. 2 Address reprint requests to: Dave Morgan, Alzheimer Research Laboratory, 12901 Bruce B Downs Boulevard, MDC Box 9, University of South Florida, Tampa, Florida 33612-4799. Tel: 813– 974–3949; Fax: 813–974–2565; E-mail: [email protected]

1029 0364-3190/03/0700–1029/0 © 2003 Plenum Publishing Corporation

1030

Morgan

difficult to obtain PDAPP mice for routine comparison to other mouse models, whether the endpoint is behavioral or otherwise. A year later, Hsiao et al. (5) described a second transgenic mouse which deposited amyloid. These mice, designated Tg2576, not only exhibited amyloid deposits similar to those found in the PDAPP mice and with the same regional distribution, but also demonstrated deficits in the Morris water maze that appeared in this cohort at roughly the same age as the deposits were manifest. It should be noted that it is likely there was still some learning in these younger Tg2576 mice (6), and at conferences it was reported that deficits did not emerge in some cohorts until they were much older. Still, these data demonstrated that amyloid deposition and memory deficits were at least linked temporally. Moreover, there was some indication that spatial memory deficits were linked to deficits in LTP (7), although this LTP deficit is not observed under all conditions (8). Fortunately, the Tg2576 mouse has been generously distributed to academic laboratories. Karen Duff and John Hardy at USF started breeding the Tg2576 mouse with a PS1 transgenic mouse (9) in November of 1996, and our research group at USF was the first to evaluate these doubly transgenic APPPS1 transgenic mice behaviorally and pathologically. One key observation was that the PS1 mice, while never depositing amyloid when singly transgenic, greatly accelerated the deposition of amyloid when crossed into the Tg2576 background (10). Importantly, while the early work identified some differences in Y maze spontaneous alternation performance, the most consistent finding was a mild hyperactivity in these doubly transgenic mice. Importantly, when tested on the Morris water maze, at both 6 months and 9 months these mice performed admirably, demonstrating both learning of platform location over days and retention during the probe trial (11). At 9 months these mice possess an amy-

loid burden roughly equivalent to that reported in Alzheimer disease (8–12%). One issue in the use of the open pool water maze is that it is a reference memory task. The platform is located in the same position for all trials even on consecutive days. As such, it might be considered the equivalent of learning where your car is parked in a crowded lot when you have a reserved space that is the same day after day. One perception is that the task is a bit too easy, or that the mice are overtrained in this task before memory is tested on probe trials (typically at the end). One approach we took to this issue was to test mice in a recently developed radial arm water maze (12). In this task, the pool is divided into swim alleys with a hidden platform at the end of one of the alleys. In addition to permitting errors (entries into incorrect arms) to be used to measure performance in addition to latencies, this task was readily converted to a working memory task, with a different platform location each day. The mice must demonstrate withinday learning, finding the platform location on the first trial and then remembering it for subsequent trials. This would thus be equivalent to remembering the location of your car in a crowded lot when it is parked in a slightly different location each day. We found in 15-month-old APPPS1 transgenic mice that learning and memory of platform location on this task was impaired in the transgenic animals, and that within the transgenic group, performance was correlated with amyloid deposition (13). Moreover, this decline was progressive, with mice at 6 months of age and even 12 months of age able to learn the task, while by 15 months most mice are not able to demonstrate memory of platform location (14,15). We have evaluated the APPPS1 mice in a number of behavioral tasks, including spontaneous alternation in the Y maze, reference memory in the Morris water maze and the circular hole (Barnes) maze. While occasional cohorts exhibit performance deficits in these other tasks, the most consistent and robust learning and

Table I. Evidence Correlating A Deposition with Memory Dysfunction in Individual Mice Mouse Model

Age

Task

PDAPP PDAPP

15–20 mo 10–12 mo

Chen et al., 2000 (18) Dodart et al., 2000 (19)

APP  PS1 APP Tg2576

15 mo 5–6 mo 21–22 mo 12 mo

Open pool water maze; Working Memory Object Recognition (no corr. Radial arm maze working or reference memory) Radial Arm Water maze; Working memory Open pool water maze; multiple probe trials Open pool water maze reference memory

Puolivali et al, 2002 (22)

APP  PS1

Reference

Gordon et al., 2001 (13) Westerman et al., 2002 (20)

Memory in APP Transgenic Mice memory performance deficits are observed in 15 month APPPS1 mice in working memory variant of the radial arm water maze (14–17). Others have also found correlations between behavioral performance and amyloid deposition. Chen et al. (18) studied PDAPP mice in a water maze. In this study, the open pool maze was converted to a working memory task by changing the hidden platform location daily, and measuring the number of trials required for mice to reach the platform within a certain criterion latency. As demonstrated previously, even young PDAPP mice with few plaques performed poorly on this task compared to age-matched control mice. However, as the PDAPP mice aged and accumulated more A deposits, their performance deteriorated further, leading ultimately to a significant age by genotype interaction on the task. Importantly, there was a significant correlation between learning capacity and plaque burden in the PDAPP mice. Dodart et al. (19) also reported a correlation of A deposits and performance on an object recognition task in 10 to 12-month-old PDAPP mice. Interestingly, at this age, performance on the working and reference memory components of a dry radial arm maze task did not correlate with amyloid deposition, although the transgenic animals were impaired at this age. Instead, it appeared that there were better correlations of the radial arm maze results with synapse loss and hippocampal atrophy. These indications of neurodegeneration and the behavioral deficits were found even in 3 mo old mice, prior to amyloid deposition. These authors interpreted these alterations as resulting from transgene expression rather than amyloid deposition. Westerman et al. (20,21), using Tg2576 mice, also modified the water maze task by including several probe trial throughout the procedure to avoid missing memory deficits due to overtraining. They found significant reductions in memory performance in mice as young as 6–11 months (but not 4–5 months) using an integrated memory measure. This performance deteriorated as the mice grew older, with little retention evident in 20 to 25-month-old mice. As expected, mice in the oldest age group had considerably more “insoluble” A (measured by ELISA) than the mice in the younger age groups. This was correlated with the performance on the integrated memory measure in both groups. Intriguingly, however, there were some old mice with 10,000 pmol/g A40 that performed better than younger mice with 10 pmol/g A40. Thus, memory performance appeared regulated in the two age groups not by the absolute amount of A in the brain, but by the relative amount compared to others in their

1031 age-matched cohort. One possible interpretation of this result is that there is some other A pool which does not accumulate with age, but whose level is dependent upon the rate of A formation. Mice generating more A would both have larger pools of memory modifying A and depositing A, with the difference being that once deposited the insoluble A would accumulate with age, while the memory modifying pool would achieve some steady state that was ultimately age-independent. The worsening performance as mice age could be an interaction of the debilitating effects of A, and the age-associated memory impairments found in rodents. Puolivali et al. (22) recently reported on a different APPPS1 mouse model. Using the standard Morris water maze, they obtained memory deficits in these mice at 11–12 months of age, but not at 3–4 months, identifying yet another mouse model in which the deficits in maze performance are progressive. They, too, identified a correlation between behavioral performance and A levels in this case restricted to the hippocampal A42 levels. While no correlation was obtained for amyloid burden in these mice, the number of deposits may have been too few to expect to see a correlation at this age. In July 1999, Schenk et al. (23) described the remarkable effects of A vaccines in lowering amyloid deposition in the PDAPP mouse model. Knowing of the difficulties in assessing behavioral deficits in these mice and having just obtained the results published in Gordon et al. (13), we decided to evaluate the functional consequences of the vaccine in the APPPS1 mouse model. We were concerned that while inoculation with vaccines against the A peptide might clear or prevent A deposition, that the ensuing activation of microglia might cause an inflammatory reaction which could impair learning and memory rather than benefit it. Thus, we inoculated mice in August 1999 with A and control (keyhole limpet hemocyanin) vaccines starting at 7 months of age when amyloid deposits are already present. We first tested mice at 12 months, an age when we expected most transgenic animals could learn the task rationalizing that any functional impairment caused by the vaccine should be manifest with that treatment duration and mouse age. To our surprise, we found that all mice learned the maze well, with not even a hint of a deficit in those inoculated with A vaccines (15). We then decided to continue the study to 15 months of age when mice were retested in the radial arm water maze task (one advantage of the working memory version of the water maze is lack of interference caused by prior experience

1032 permitting longitudinal testing). In this instance we observed the expected deficits in transgenic mice treated with the control vaccine, and strong learning performance in the nontransgenic littermates. The transgenic mice vaccinated with A, while slower to learn platform location, ultimately performed the same as nontransgenic mice, averaging less than one error on the fifth trial. When we analyzed the amyloid loads in these mice, we found a modest 20% reduction overall in the A-vaccinated APPPS1 transgenic mice (the APPonly transgenics also used in the study had a larger percentage reduction). Since these mice still had intact memory function, it appeared that A loads were not all that critical in regulating memory performance. In parallel to our work, Janus et al. (24) tested A vaccines in a new transgenic mouse model with two human APP mutations (TgCRND8). These mice were deficient in the Morris water maze as early as 11 weeks of age, and this deficit was also observed at 24 weeks compared to nontransgenic animals. The A vaccine improved the performance of these mice such that they were no longer worse than nontransgenic mice, yet there was not a significant difference between the A vaccinated and control vaccinated transgenic mice. At necropsy, there was a substantial number of deposits for mice this young (6 months) and these were reduced by 50% in the A vaccinated mice. However, unexplainedly, ELISA measurements of acid extractable A from the brains of these mice were unaltered by the vaccination. Again taking the lead, the Elan group (25) demonstrated that passive immunization with anti-A antibodies could also reduce amyloid load in the PDAPP mice. This led recently to two publications evaluating passive immunization which suggest that the removal of amyloid deposits is not necessary for vaccines to have a beneficial behavioral effect in APP transgenic mice. Dodart et al. (26) demonstrated that a single injection of anti-A antibody could dramatically improve performance of PDAPP mice on both a hole board task integrating both working and reference spatial memory and an object recognition task. As even 6 weekly administrations of the antibody could not reduce deposited A, it would appear most likely that reduction of some other A pool that can be rapidly modified is responsible for the intact memory function in the PDAPP mice. More recently, a second study in 9 to 11-monthold Tg 2576 mice demonstrated a similar benefit of passive immunization (27). These mice were first demonstrated to be memory impaired in the reference

Morgan memory version of the Morris maze. They were then administered three anti-A antibody injections over a period of a week and retested. The mice that were impaired on the first test and given anti-A antibody injections improved in their memory test performance, while impaired mice given a control antibody preparation showed no improvement. There was no effect of the antibody treatment on saline-soluble, SDS-soluble, or formic acid–soluble A pools, suggesting none of these pools were responsible for the observed improvement in performance. These general arguments are supported by a considerable literature demonstrating that injections of A peptide into the brains of mice and rats can result in rapid disruption of learning and memory performance (reviewed in [28]). Yet another human APP751 transgenic mouse develops diffuse A deposits but fails to exhibit neuritic fibrillar A plaques. This mouse too develops age-dependent deficits in reference memory water maze performance (29). Again, this would suggest that deposited A is not responsible for the memory deficits in APP transgenic mice, and supports the notion that there is some other A pool that is responsible. Thus, over the last several years, there is a consensus emerging that A is in some way linked to the memory deficits that develop in transgenic mice over expressing APP. It is important to recognize that strain background may have a significant influence not only on learning and memory performance, with heterogeneous backgrounds exhibiting superior learning and memory performance (30,31), but also on the apportionment of A into different pools. This makes comparisons across different transgenic lines virtually impossible, and likely makes comparisons of the same lines from different laboratories difficult. Still, just as most transgenic APP mice deposit A primarily in cerebral cortex and hippocampus irrespective of the promoter, so too do these mice express impaired learning and memory in standard rodent behavioral tests. However, the default assumption from the amyloid hypothesis (32), that fibrillar amyloid plaques are the main culprit, does not seem to apply in these mouse models. In particular, the results from vaccination trials indicate that memory deficits can be reversed rapidly without significant depletion of the fibrillar A pool. The question then becomes, what does this mean for Alzheimer’s disease? Is the memory deficit found in AD due to some rapidly modifiable A pool? Perhaps early in the disease it is. However, a critical difference between the APP transgenic

Memory in APP Transgenic Mice mouse models and AD is the absence of neuron loss in mice. Even in the few instances where neuron loss does occur (33) it is not as extensive as in AD, and largely restricted to he hippocampus. Reliable neuron loss has not yet been reported in the PDAPP, Tg2576, or the APPPS1 model (34–36). It is probable that the early memory deficits in mild cognitive impairment or early stage AD may result from a mechanism similar to that found in the transgenic mouse models. Modifying this pool with passive immunization approaches would appear to be a reasonable test of the hypothesis, with careful monitoring of patients to look for signs of CNS inflammation. ACKNOWLEDGMENTS This work was supported by grants NIA AG 15490, AG18478, and AG20227.

REFERENCES 1. Games, D., Adams, D., Alessandrini, R., Barbour, R., Berthelette, P., Blackwell, C., Carr, T., Clemens, J., Donaldson, T., and Gillespie, F. 1995. Alzheimer-type neuropathology in transgenic mice overexpressing V717F beta-amyloid precursor protein. Nature 373:523–527. 2. Greenberg, B. D., Savage, M. J., Howland, D. S., Ali, S. M., Siedlak, S. L., Perry, G., Siman, R., and Scott, R. W. 1996. APP transgenesis: approaches toward the development of animal models for Alzheimer disease neuropathology. Neurobiol. Aging 17:153–171. 3. Justice, A. and Motter, R. 1997. Behavioral characterization of PDAPP transgenic Alzheimer mice. Neurosci. Abstr. 23(2):1637. 4. Dodart, J. C., Meziane, H., Mathis, C., Bales, K. R., Paul, S. M., and Ungerer, A. 1997. Memory and learning impairment precede amyloid deposition in the V717F PDAPP transgenic mouse. Neurosci. Abstr. 23(2):1637. 5. Hsiao, K., Chapman, P., Nilsen, S., Eckman, C., Harigaya, Y., Younkin, S., Yang, F., and Cole, G. 1996. Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 274:99–102. 6. Routtenberg, A. 1997. Measuring memory in a mouse model of Alzheimer’s. Science 277:839–840. 7. Chapman, P. F., White, G. L., Jones, M. W., Cooper-Blacketer, D., Marshall, V. J., Irizarry, M., Younkin, L., Good, M. A., Bliss, T. V., Hyman, B. T., Younkin, S. G., and Hsiao, K. K. 1999. Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nat. Neurosci. 2:271–276. 8. Fitzjohn, S. M., Morton, R. A., Kuenzi, F., Rosahl, T. W., Shearman, M., Lewis, H., Smith, D., Reynolds, D. S., Davies, C. H., Collingridge, G. L., and Seabrook, G. R. 2001. Agerelated impairment of synaptic transmission but normal longterm potentiation in transgenic mice that overexpress the human APP695SWE mutant form of amyloid precursor protein. J. Neurosci. 21:4691–4698. 9. Duff, K., Eckman, C., Zehr, C., Yu, X., Prada, C. M., Perez-tur, J., Hutton, M., Buee, L., Harigaya, Y., Yager, D., Morgan, D., Gordon, M. N., Holcomb, L., Refolo, L., Zenk, B., Hardy, J., and Younkin, S. 1996. Increased amyloid-beta42(43) in brains of mice expressing mutant presenilin 1. Nature 383:710–713.

1033 10. Holcomb, L., Gordon, M. N., McGowan, E., Yu, X., Benkovic, S., Jantzen, P., Wright, K., Saad, I., Mueller, R., Morgan, D., Sanders, S., Zehr, C., O’Campo, K., Hardy, J., Prada, C. M., Eckman, C., Younkin S., Hsiao, K., and Duff, K. 1998. Accelerated Alzheimer-type phenotype in transgenic mice carrying both mutant amyloid precursor protein and presenilin 1 transgenes. Nat. Med. 4:97–100. 11. Holcomb, L. A., Gordon, M. N., Jantzen, P., Hsiao, K., Duff, K., and Morgan, D. 1999. Behavioral changes in transgenic mice expressing both amyloid precursor protein and presenilin-1 mutations: Lack of association with amyloid deposits. Behav. Gen. 29:177–185. 12. Diamond, D. M., Park, C. R., Heman, K. L., and Rose, G. M. 1999. Exposing rats to a predator impairs spatial working memory in the radial arm water maze. Hippocampus 9:542–551. 13. Gordon, M. N., King, D. L., Diamond, D. M., Jantzen, P. T., Boyett, K. L., Hope, C. E., Hatcher, J. M., DiCarlo, G., Gottschal, P., Morgan, D., and Arendash, G. W. 2001. Correlation between cognitive deficits and A deposits in transgenic APPPS1 mice. Neurobiol. Aging 22:377–385. 14. Arendash, G. W., King, D. L., Gordon, M. N., Morgan, D., Hatcher, J. M., Hope, C. E., and Diamond, D. M. 2001. Progressive behavioral impariments in transgenic mice carrying both mutant APP and PS1 transgenes. Brain Res. 891:45–53. 15. Morgan, D., Diamond, D. M., Gottschall, P. E., Ugen, K. E., Dickey, C., Hardy, J., Duff, K., Jantzen, P., DiCarlo, G., Wilcock, D., Connor, K., Hatcher, J., Hope, C., Gordon, M., and Arendash, G. W. 2000. A beta peptide vaccination prevents memory loss in an animal model of Alzheimer’s disease. Nature 408:982–985. 16. Austin, L. A., Arendash, G. W., Gordon, M. N., Diamond, D. M., DiCarlo, G., Dickey, C. A., Ugen, K. E., and Morgan, D. 2003. Short-term A vaccinations do not Improve cognitive performance in cognitively-impaired APPPS1 mice. Behav. Neurosci. in press. 17. Gordon, M. N., Holcomb, L. A., Jantzen, P. T., DiCarlo, G., Wilcock, D., Connor, K., Melachrino, J. O., O’Callaghan, J. P., and Morgan, D. 2001. Time course of the development of Alzheimer-like pathology in the doubly transgenic mPS1mAPP mouse. Exp. Neurol. 173:183–195. 18. Chen, G., Chen, K. S., Knox, J., Inglis, J., Bernard, A., Martin, S. J., Justice, A., McConlogue, L., Games, D., Freedman, S. B., and Morris, R. G. 2000. A learning deficit related to age and beta-amyloid plaques in a mouse model of Alzheimer’s disease. Nature 408:975–979. 19. Dodart, J. C., Mathis, C., Saura, J., Bales, K. R., Paul, S. M., and Ungerer, A. 2000. Neuroanatomical abnormalities in behaviorally characterized APP(V717F) transgenic mice. Neurobiol. Dis. 7:71–85. 20. Westerman, M. A., Cooper-Blacketer, D., Mariash, A., Kotilinek, L., Kawarabayashi, T., Younkin, L. H., Carlson, G. A., Younkin, S. G., and Ashe, K. H. 2002. The relationship between Abeta and memory in the Tg2576 mouse model of Alzheimer’s disease. J. Neurosci. 22:1858–1867. 21. Dineley, K. T., Westerman, M., Bui, D., Bell, K., Ashe, K. H., and Sweatt, J. D. 2001. Beta-amyloid activates the mitogenactivated protein kinase cascade via hippocampal alpha7 nicotinic acetylcholine receptors: In vitro and in vivo mechanisms related to Alzheimer’s disease. J. Neurosci. 21:4125–4133. 22. Puolivali, J., Wang, J., Heikkinen, T., Heikkila, M., Tapiola, T., van Groen, T., and Tanila, H. 2002. Hippocampal A beta 42 levels correlate with spatial memory deficit in APP and PS1 double transgenic mice. Neurobiol. Dis. 9:339–347. 23. Schenk, D., Barbour, R., Dunn, W., Gordon, G., Grajeda, H., Guido, T., Hu, K., Huang, J., Johnson-Wood, K., Khan, K., Kholodenko, D., Lee, M., Liao, Z., Lieberburg, I., Motter, R., Mutter, L., Soriano, F., Shopp, G., Vasquez, N., Vandevert, C., Walker, S., Wogulis, M., Yednock, T., Games, D., and Seubert, P. 1999. Immunization with amyloid-beta attenuates Alzheimer-disease-like pathology in the PDAPP mouse. Nature 400:173–177.

1034 24. Janus, C., Pearson, J., McLaurin, J., Mathews, P. M., Jiang, Y., Schmidt, S. D., Chishti, M. A., Horne, P., Heslin, D., French, J., Mount, H. T., Nixon, R. A., Mercken, M., Bergeron, C., Fraser, P. E., George-Hyslop, P., and Westaway, D. 2000. A beta peptide immunization reduces behavioural impairment and plaques in a model of Alzheimer’s disease. Nature 408:979–982. 25. Bard, F., Cannon, C., Barbour, R., Burke, R. L., Games, D., Grajeda, H., Guido, T., Hu, K., Huang, J., Johnson-Wood, K., Khan, K., Kholodenko, D., Lee, M., Lieberburg, I., Motter, R., Nguyen, M., Soriano F., Vasquez, N., Weiss, K., Welch, B., Seubert, P., Schenk, D., and Yednock, T. 2000. Peripherally administered antibodies against amyloid beta-peptide enter the central nervous system and reduce pathology in a mouse model of Alzheimer disease. Nat. Med. 6:916–919. 26. Dodart, J. C., Bales, K. R., Gannon, K. S., Greene, S. J., DeMattos, R. B., Mathis, C., DeLong, C. A., Wu, S., Wu, X., Holtzman, D. M., and Paul, S. M. 2002. Immunization reverses memory deficits without reducing brain Abeta burden in Alzheimer’s disease model. Nat. Neurosci 5:452–457. 27. Kotilinek, L. A., Bacskai, B., Westerman, M., Kawarabayashi, T., Younkin, L., Hyman, B. T., Younkin, S., and Ashe, K. H. 2002. Reversible memory loss in a mouse transgenic model of Alzheimer’s disease. J. Neurosci. 22:6331–6335. 28. Morgan, D. and Gordon, M. N. 2002. The amyloid hypothesis of cognitive dysfunction. In: Neuroinflammation Mechanisms and Management, Wood, P. M. (ed.) New York: Humana. 29. Koistinaho, M., Ort, M., Cimadevilla, J. M., Vondrous, R., Cordell, B., Koistinaho, J., Bures, J., and Higgins, L. S. 2001. Specific spatial learning deficits become severe with age in beta-amyloid precursor protein transgenic mice that harbor diffuse beta-amyloid deposits but do not form plaques. Proc. Natl. Acad. Sci. USA 98:14675–14680.

Morgan 30. Owen, E. H., Logue, S. F., Rasmussen, D. L., and Wehner, J. M. 1997. Assessment of learning by the Morris water task and fear conditioning in inbred mouse strains and F1 hybrids: Implications of genetic background for single gene mutations and quantitative trait loci analyses. Neurosci 80(4):1087–1099. 31. Gonzalez, R. G., Guimaraes, A. R., Moore, G. J., Crawley, A., Cupples, L. A., and Growdon, J. H. 1996. Quantitative in vivo 31P magnetic resonance spectroscopy of Alzheimer disease. Alzheimer. Dis. Assoc. Disord. 10:46–52. 32. Hardy, J. and Selkoe, D. J. 2002. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 297:353–356. 33. Calhoun, M. E., Burgermeister, P., Phinney, A. L., Stalder, M., Tolnay, M., Wiederhold, K. H., Abramowski, D., SturchlerPierrat, C., Sommer, B., Staufenbiel, M., and Jucker, M. 1999. Neuronal overexpression of mutant amyloid precursor protein results in prominent deposition of cerebrovascular amyloid. Proc. Natl. Acad. Sci. USA 96:14088–14093. 34. Takeuchi, A., Irizarry, M. C., Duff, K., Saido, T. C., Hsiao, A. K., Hasegawa, M., Mann, D. M., Hyman, B. T., and Iwatsubo, T. 2000. Age-related amyloid beta deposition in transgenic mice overexpressing both Alzheimer mutant presenilin 1 and amyloid beta precursor protein Swedish mutant is not associated with global neuronal loss. Am. J. Pathol. 157:331–339. 35. Irizarry, M. C., McNamara, M., Fedorchak, K., Hsiao, K., and Hyman, B. T. 1997. APPSw transgenic mice develop age-related A beta deposits and neuropil abnormalities, but no neuronal loss in CA1. J. Neuropathol. Exp. Neurol. 56:965–973. 36. Irizarry, M. C., Soriano, F., McNamara, M., Page, K. J., Schenk, D., Games, D., and Hyman, B. T. 1997. Abeta deposition is associated with neuropil changes, but not with overt neuronal loss in the human amyloid precursor protein V717F (PDAPP) transgenic mouse. J. Neurosci. 17:7053–7059.