Premature death in transgenic mice that overexpress a mutant amyloid precursor protein is preceded by severe neurodegeneration and apoptosis

Pergamon PII:

Neuroscience Vol. 91, No. 3, pp. 819–830, 1999 Copyright 䉷 1999 IBRO. Published by Elsevier Science Ltd Printed in Great Britain. All rights reserved 0306-4522/99 $20.00+0.00 S0306-4522(98)00599-5

PREMATURE DEATH IN TRANSGENIC MICE THAT OVEREXPRESS A MUTANT AMYLOID PRECURSOR PROTEIN IS PRECEDED BY SEVERE NEURODEGENERATION AND APOPTOSIS D. MOECHARS,* K. LORENT* and F. VAN LEUVEN† Experimental Genetics Group (EGG), Center for Human Genetics (CME), Flemish Institute for Biotechnology (VIB), K.U. Leuven, Campus Gasthuisberg, B-3000 Leuven, Belgium

Abstract—A mutant amyloid precursor protein (APP/RK) designed to interfere with processing by asecretase caused a severe phenotype in transgenic mice, including behavioural abnormalities, i.e. neophobia, aggression, hypersensitivity to kainic acid, hyposensitivity to N-methyl-d-aspartate, and premature death [Moechars D. et al. (1996) Eur. molec. Biol. Org. J. 15, 1265–1274]. We now demonstrated that the APP/RK transgene did not disturb the expression of several other genes, i.e. endogenous amyloid precursor protein and amyloid precursor protein-like proteins, members of the low density lipoprotein receptor lipoprotein receptor family and several of their ligands, including apolipoprotein E, but expression of alpha-2-macroglobulin was never detected. Neither amyloid deposits nor neurofibrillary tangles were detected in the brain of APP/RK transgenic mice, even when 15-months-old. The tendency for seizures and hyposensitivity for N-methyl-d-aspartate was not due to or reflected in the distribution of the three major types of glutamate receptors. The major and consistent finding in transgenic APP/RK mice that died prematurely was extensive neurodegeneration and apoptosis, mainly in hippocampus and cortex, and accompanied by astrocytosis throughout the brain. Reduced synaptic density and dendritic damage was only observed in three transgenic mice that were killed shortly after positive observation of seizures. In addition, the distribution of cathepsin D and ubiquitin was abnormal in these mice. 䉷 1999 IBRO. Published by Elsevier Science Ltd. Key words: transgenic mice, amyloid precursor protein, Alzheimer’s disease, apoptosis, neurodegeneration.

Alzheimer’s disease (AD) is the most frequent neurodegenerative disorder, characterized by the progressive deterioration of cognitive functions, learning and memory. The definitive, post mortem pathological diagnosis is based on the generalized occurrence of extracellular senile plaques and intracellular neurofibrillary tangles in the cortex, hippocampus, and amygdala in the brain of AD patients. Although senile plaques contain numerous immunoreactive proteins, their major constituent are the b-amyloid peptides derived from the amyloid precursor protein (APP) by proteolytic enzymatic activities, named secretases (reviewed in Ref. 11). De-regulation of expression and proteolytic processing of APP is central to the development of AD and †To whom correspondence should be addressed. *These authors contributed equally to this work. Abbreviations: AD, Alzheimer’s disease; A2M, alpha-2-macroglobulin; AMPA, amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; APLP, APP-like protein; ApoE, apolipoprotein E; APP, amyloid precursor protein; GFAP, glial fibrillary acidic protein; LDL-R, low density lipoprotein receptor; LRP, lipoprotein receptor related protein; MAP-2, microtubule-associated protein 2; NMDA, N-methyl-d-aspartate; RAP, receptor-associated protein; TUNEL, terminal deoxynucleotide transferase-mediated dUTP-nick end-labelling; VLDL-R, very low density lipoprotein receptor.

is based on different lines of evidence. Over-expression in trisomy 21 patients causes pre-amyloid lesions in the brain, very early in life. Rare point mutations in APP cause early onset familial AD by deviating APP into an amyloidogenic direction with increased amyloid peptides, especially the long, less soluble bA(42) form. 34,44 Finally, mutations in presenilins cause most early onset familial AD cases and similarly increase amyloid peptides. 2,33 Since neurons of presenilin 1-deficient mice do not appreciably produce bA4 peptides, the mutations in presenilins cause an unknown gain of function. 7 Among the genes and proteins implicated in the pathogenesis of AD, epidemiological studies identified apolipoprotein E (ApoE) as the major risk factor (reviewed in Ref. 35). The molecular mechanism is unknown and might involve production, clearance or deposition of the amyloid peptides. The lipoprotein receptor-related protein (LRP) is a major ApoE receptor in brain and is found in the amyloid plaques, together with many of its ligands, e.g., ApoE, APP, alpha-2-macroglobulin (A2M), lipoprotein lipase, among others. 21,27,30,37,38 This possible link was strengthened by a genetic association of sporadic AD to the LRP1 gene on chromosome 12q13. 17

819

820

D. Moechars et al.

Fig. 1. Quantification of mRNA in hippocampus and cerebrum. Northern blots were hybridized for APP mRNA and autoradiograms were densitometrically scanned to quantify the expression of endogenous APP in the hippocampus (left) and cerebrum without the hippocampus (centre) of non-transgenic and APP/RK transgenic mice. (right) APP/RK transgene mRNA levels relative to endogenous APP mRNA in the hippocampus and cerebrum of APP/RK transgenic mice.

The physiological impact of de-regulation of APP metabolism in brain has been modelled in vivo, in many different strains of transgenic mice. Previously, we have generated transgenic mice that overexpress a mutant form of APP, i.e. mutated at the a-secretase cleavage site (APP/RK). 24 These transgenic mice developed typical behavioural disturbances, including neophobia, aggression, agitation, glutamatergic disturbances and finally premature death. The intensity and the age of onset of these phenomena were related to the expression levels of the transgene in different strains. 24 Independent transgenic mouse strains that overexpress human APP develop essentially this same phenotype (unpublished observations). All APP transgenic mice were less sensitive to N-methyl-daspartate (NMDA), while more sensitive to kainic acid 24,25 (unpublished observations). Premature death is evident in all transgenic mouse lines that we have generated by overexpressing either mouse or human APP, either normal APP, known clinical mutants of APP or the designed asecretase mutant APP/RK, in either the FVB/N or the C57Bl6 genetic background 24,25 (unpublished observations). Premature death was also reported by others using other promoters to drive expression of APP in different mouse strains. 15,16,20 The precise cause is unknown. Premature death was absent in different transgenic lines in which we used the same mouse thy-1 gene promoter to drive expression of a variety of transmembrane, secreted or cytoplasmic proteins, i.e. presenilins, ApoE4, protein tau among others. The continued analysis of the APP/RK transgenic mice involved those aspects that are relevant for AD,

as introduced above. We investigated systematically at different ages, the mRNA levels of endogenous APP and APP-like proteins, APLP1 and APLP2, and of different members of the low density lipoprotein (LDL) receptor family, of their chaperone and of several ligands, including ApoE. EXPERIMENTAL PROCEDURES

Complementary DNA probes used were isolated, radiolabelled and used as described before. 22,23 Briefly: APP: bases 101 to 917 of mouse APP 695 cDNA; 5 APLP1 (APLP1): bases 1771 to 2301 of mouse APLP1 cDNA; 42 APLP2: a 240 bp SstI-DraII restriction fragment (bases 1687 to 1927) of mouse APLP2 cDNA; 41 PS1: a 834 bp BglI restriction fragment of mouse cDNA; A2MR/LRP: a 1.4 kb fragment (bases 776 to 2195) of mouse LRP cDNA; 40 RAP: bases 8 to 1078 of mouse cDNA; 8 ApoE: a 1013 bp EcoRI restriction fragment of mouse ApoE cDNA; 14 LPL: a 1 kb PstI restriction fragment of rat LPL cDNA (gift from J. Auwerx, Lille, France); A2M: bases 1188 to 1758 of mouse cDNA; 39 VLDL-R (VLDL-R): bases 1086 to 2181 of mouse cDNA; 9 LDL-R (LDL-R): a 700 bp EcoRI fragment of mouse cDNA clone mLDLRc90; 13 tPA: a 0.8 kb EcoRI fragment of mouse tPA cDNA, 31 Dissected hippocampi and remainder of cerebrum were quick-frozen in liquid nitrogen and stored at ⫺70⬚C before RNA extraction and northern blotting. 22,23 Appropriately exposed autoradiographs were quantified densitometrically and normalized for actin mRNA, hybridized to the same blot. Mice were anaesthetized and perfused transcardially with cold saline followed by 4% paraformaldehyde. Brains were removed and fixed overnight at 4⬚C, washed in phosphatebuffered saline, dehydrated and embedded in paraffin. Sections (5 mm) were cut and dried on glass slides. Brain was processed for in situ hybridization as described. 22,23 For in situ hybridization, in situ apoptosis detection, and immunohistochemistry with microwave treatment, glass slides were first coated with triaminopropylmethyldiethoxysilane. Alternatively, after perfusion with saline and 4%

Apoptosis in APP transgenic mice

821

Fig. 2. In situ hybridization of endogenous and transgenic APP. Bright-field autoradiograph of the distribution of mRNA encoding total APP in the hippocampus of non-transgenic (A) and transgenic (B) mice. Scale bar ˆ 100 mm.

paraformaldehyde and overnight postfixation, 40 to 50 mm Vibratome sections were cut for A2MR/LRP and RAP immunostaining. Coronal sections were routinely stained on paraffin sections with haematoxylin–eosin and Cresyl Violet (Nissl stain). Apoptosis was analysed by terminal deoxynucleotidyl transferase-mediated addition of digoxygenin-11-dUTP or fluorescein-conjugated 11-dUTP (Apoptag, Oncor, Gaithersburg, MD; Apoptosis Detection System, Promega, Madison, WI). Two versions of silver impregnation routinely used for pathological analysis of the brain of AD patients 4,10 were applied.

Antibodies used: B10/4: rabbit antiserum against APP carboxy-terminal; 6 goat antiserum against APP ectodomain (gift from B. Greenberg); anti-bovine glial fibrillary acidic protein (GFAP), anti-synaptophysin and anti-ubiquitin (Dakopatts, Denmark); Mab 22C11 12 and anti-microtubule-associated protein (MAP-2) (Boehringer Mannheim, Germany), anti-cathepsin D, 18 anti-A2MR/LRP (R777); 19 anti-RAP, 26 Tissue sections were deparaffinized, rehydrated and endogenous peroxidase inactivated. Primary antibodies were applied overnight, and appropriate secondary, peroxidase-conjugated or biotinylated antisera were reacted for 1 h. Immunostaining for APP and synaptophysin was

Fig. 3. Analysis of mRNA expression. Northern blotting of total cerebrum RNA from heterozygous APP/RK transgenic (lanes 1, 2, 5, 6, 9 and 10) and non-transgenic mice (lanes 3, 4, 7, 8, 11 and 12) aged three (lanes 1–4), nine (lanes 5–8) and 15 months (lanes 9–12), sequentially hybridized with cDNA probes specific for, respectively, APP (endogenous and transgene), APLP1, APLP2, tPA and actin.

D. Moechars et al.

Fig. 4.

822

Apoptosis in APP transgenic mice

823

Table 1. Summary of neurodegeneration, apoptosis and astrogliosis in the brain of APP/RK transgenic mice from the three groups (see text for details) Premature death Feature

Neurodegeneration Lesions corpus callosum Apoptosis Astrogliosis

Healthy

Seizures

Method

Haematoxylin–Eosin and Cresyl Violet Haematoxylin–Eosin and Cresyl Violet TUNEL staining GFAP immunostaining

Analysed

Positive

Analysed

Positive

Analysed

Positive

35

29

26

2

3

3

35

24

26

10

3

3

26 18

7 16

11 25

0 14

3 3

3 3

carried out after microwaving slides in 10 mM citrate buffer, pH 6 for 8 min at 450 W. Immunostaining for A2MR/LRP and RAP was performed on floating Vibratome sections. Negative control sections were processed in parallel without the first antibody. The specificity of antibodies was authenticated by western blotting, immunoprecipitation or immunoelectrophoresis. Experiments were carried out according to the European Communities Council Directive of 24 November 1986 (86/ 609/EEC). All efforts were made to reduce the number of animals used and to minimize animal suffering.

RESULTS

Analysis of messenger RNA expression RNA was extracted from hippocampus and remaining cerebrum from APP/RK transgenic mice aged three, nine and 15 months and age-matched non-transgenic mice. Northern blots were probed for 12 mRNA species: transgenic APP/RK and endogenous wild-type APP, the direct relatives APLP1 and APLP2, the members of the LDL-R family, i.e. A2MR/LRP, LDL-R, and VLDL-R, their chaperone RAP and their ligands ApoE, A2M, LPL and tPA, and finally, for mouse presenilin 1. The 3.9 kb transgene APP/RK mRNA was differentially detected alongside the 3.4 kb endogenous APP mRNA, the difference in size being due to the 3 0 untranslated region of the thy-1 gene construct. Expression of endogenous APP increased with age in hippocampus but not in the rest of the cerebrum, and this was not different in transgenic mice (Fig. 1). Expression of APP/RK mRNA relative to endogenous APP was higher in the hippocampus. This remained constant with age, demonstrating that expression of endogenous APP was unaffected by the transgene. In situ hybridization

(Fig. 2) and immunohistochemistry (see below) further identified APP/RK transgene expression to be high in pyramidal neurons, especially in hippocampus, cortex, amygdala and striatum. In addition to APP, mRNA levels of the following genes were quantified: APLP1, APLP2, A2MR/ LRP, LDL-R, VLDL-R, RAP, MAM, ApoE, LPL, tPA, and presenilin 1. This did not reveal major differences in expression levels in hippocampus and cerebrum at different ages of the APP/RK transgenic mice (Fig. 3). Spatial patterns of expression analysed by in situ hybridization for specified mRNA species were not different (results not shown), excluding these actors to be directly involved in the observed phenotype. It must specifically be noted that expression of A2M was never detected in mouse brain.

Histology Neurodegeneration and apoptosis was essentially analysed in three groups of APP/RK mice. The first group of 35 mice of three independent transgenic lines: 24 five t/RK/B/2 mice, 26 t/RK/F/4 mice and four t/RK/F/6 mice. All these transgenic mice died spontaneously and prematurely, between one-month and one-year-old, and were analysed with a variable post mortem delay estimated at between 5 min to 14 h. Three mice of this group were positively identified undergoing seizures before death. The second group contained 26 mice, i.e. 22 t/RK/F/4, one t/RK/F/6 and one t/RK/B/2, aged between two and 15 months, which were killed when active and healthy without signs of seizures. The third and smallest group consisted of three APP/RK transgenic mice, aged three, four and six months that

Fig. 4. Histology of the brain of APP/RK transgenic mice. (A–F) Haematoxylin–Eosin staining in hippocampal CA1 pyramidal layer of (A) non-transgenic and (B) transgenic t/RK/F/6 mice (group 3); dentate gyrus of (C) non-transgenic and (D) transgenic t/RK/ F/4 mice (group 3); dentate gyrus of transgenic t/RK/F/4 (E) and t/RK/F/6 (F) mice that died prematurely. Scale bars ˆ 20 mm. (G–L) GFAP immunostaining in the cortex of non-transgenic mouse (G) and of transgenic t/RK/F/4 mouse (H); GFAP staining in the hippocampus of (I) non-transgenic and (J) transgenic t/RK/F/6 mouse from group 3. Scale bars ˆ 100 mm. (K, L) Magnification of panels I and J, respectively, to show enlarged cell bodies and processes of astrocytes in the transgenic mouse. Scale bars ˆ 20 mm. (M–P) Immunostaining for APP (22C11) in the dentate gyrus of (M) non-transgenic and (N) transgenic t/RK/F/4 mouse and in the amygdala of (O) non-transgenic and (P) transgenic t/RK/F/4 mice. Scale bar ˆ 20 mm. (Q–T) TUNEL staining of dentate gyrus of (Q) non-transgenic and (R) transgenic t/RK/F/4 mice and of cortex of (S) non-transgenic and (T) transgenic t/RK/F/6 mice (group 3). Scale bars ˆ 20 mm.

824

D. Moechars et al.

Fig. 5. Garvey and Bielschowsky silver impregnation of mouse brain. Photomicrographs of brain stained by Garvey (A, C, E) and Bielschovsky (B, D, F) silver impregnation methods of APP/London(V I) transgenic mouse (A, B), non-transgenic mouse (C, D) and APP/RK transgenic mouse (group 3) with extensive neurodegeneration (E, F). Scale bars: (A, B) ˆ 30 mm; (C–F) ˆ 100 mm.

were killed immediately after undergoing seizures when still apathic and recovering. Haematoxylin–Eosin staining of the brain of APP/RK mice of group 1 revealed, in 29 animals (83%), neurons with abnormal morphology, i.e. a compact and densely-stained nucleus surrounded by clear cytoplasm (Fig. 4; Table 1). These neurons were prominently concentrated on the boundary of the granular layer and the hilus of the dentate gyrus and otherwise dispersed throughout the granular layer (Fig. 4E, F) and thalamus in all these animals. The morphology of these neurons in the dentate gyrus was reminiscent of apoptotic cells which was authenticated by TUNEL staining (see below). Occasionally, similarly degenerating neurons were present in the posterior cingulate cortex, the striatum, the hypothalamus, and the CA1 and CA2 region of the hippocampal pyramidal layer. Another typical finding was distortion of the morphology of

nuclei of glial cells in the corpus callosum and fimbria, which in normal mice were ellipsoid in shape, but appeared rounded and more densely stained in 63% of APP/RK transgenic mice. In the corpus callosum of 24 mice in this group (68.5%) morphological lesions were present (Table 1). In only four mice, aged between two and nine months, the blood vessels in the brain were dilated. In the second group of APP/RK mice, representing healthy animals aged between two and 15 months, this abnormal neuronal morphology was rare or absent. Only in two mice (five- and sixmonths-old) some collapsed neurons in the amygdala were evident, while in 10 mice out of the 26 examined (41%) the morphological lesions in the corpus callosum were evident. In only one mouse signs of dilated cerebral blood vessels were similar to the four mice in group 1. The brain of three transgenic APP/RK mice (aged

Apoptosis in APP transgenic mice

three, four and six months) with a positive history of seizures and killed immediately after seizures, were most severely damaged: many collapsed neurons were evident in the hippocampal pyramidal layer (Fig. 4B), the subiculum, the cortex, and the amygdala, and in two mice also in the dentate gyrus granular layer (Fig. 4D) and in striatum. Lesions in the corpus callosum were seen in all three transgenic mice (Table 1). The two silver impregnation methods used to stain typical lesions in human AD brain and in human APP transgenic mouse brain (Fig. 5A, B) did not reveal either plaques or tangles in the brain of APP/RK transgenic mice of up to 15-months-old. The Bielschowsky method stained axons and white matter dark brown on a lightly coloured background (Fig. 5D, F). The Garvey method resulted in purpleblue staining of white matter and axons with chromatin and nuclear membrane appearing purple with black nucleoli. Degenerated neurons and apoptotic bodies stained darkly (Fig. 5C, E).

825

for GFAP, cathepsin D, ubiquitin, synaptophysin, MAP-2, A2MR/LRP and its chaperone RAP. Amyloid precursor protein. The APP/RK mice over-express APP in brain at levels that were estimated by western blotting with monoclonal antibody 22C11 to be about 75% (line t/RK/F/6) and 200% (line t/RK/F/4) higher than in non-transgenic control mice (results not shown). In wild-type mice, APP immunoreactivity was distributed in a granular staining pattern in the cell body and proximal neurites of neurons, representing different vesicles (reviewed in Ref. 28). In the brain of APP/RK transgenic mice strong immunoreactivity was evident in neurons of the cortex, amygdala (Fig. 4P), and striatum, and in the hippocampal pyramidal layer and in granular cells of the dentate gyrus (Fig. 4N, M). Overall, the distinct neuronal staining in APP/RK transgenic animals appeared more coarse than in wild-type mice, most evident in cortical neurons. Collapsed neurons with condensed nuclei, when present in the sections examined, displayed intense immunoreactivity for APP.

Apoptosis by TUNEL staining The degenerating granular cells in the dentate gyrus were typified by condensed nuclei with crescent-shaped chromatin aggregated at the nuclear membrane. This represented a process of apoptosis as demonstrated by terminal transferase-mediated end-labelling of fragmented DNA. In the brain of seven mice in group 1, variable numbers of apoptotic neurons were present in the dentate gyrus granular layer (Fig. 4R). In one transgenic animal, additional apoptotic neurons were spread over the cortex only and in another animal were also in hippocampus, amygdala and striatum. All transgenic APP/RK mice in group 3 displayed shrunken neurons that stained positively for DNA fragmentation while only in these animals typical apoptotic bodies were observed in the cortex (Fig. 4T). In the brain of all mice of group 2 we failed to detect apoptotic cells by this technique.

Immunohistochemistry Since mouse APP cDNA was the basis of the APP/RK construct used to generate these transgenic mice, it was important to determine whether the transgenic mice developed b-amyloid deposits of any type that escaped detection by both silver staining methods. A battery of antibodies against the bA4 peptide corroborated the conclusion that neither plaques nor deposits of any type were present in the brain of transgenic mice even as old as 15 months. Similarly, immunostaining with several anti-tau antibodies did not reveal neuritic pathology nor abnormal cellular staining (results not shown). Next to APP, bA4 peptides and protein tau, additional immunohistochemical analysis was performed

Glial fibrillary acidic protein. Immunostaining for GFAP revealed astrogliosis in the brain of most APP/RK mice of group 1, in more than half of the mice of group 2 and very intense and widespread in all animals of group 3 (Table 1). The astrocytic reaction was always most prominent in the hippocampus (Fig. 4I–L), but was also evident in amygdala and cortex, especially the primary olfactory cortex and the entorhinal cortex, of all APP/RK mice. Furthermore, anti-GFAP-stained astrocytes in the striatum, the thalamus and the corpus callosum in half of the APP/RK transgenic mice. Interestingly, in the brain of most transgenic mice from group 2, i.e. mice that were killed when apparently healthy, astrogliosis was already evident in the hippocampus, the amygdala and the cortex (Fig. 4H). Hematoxylin–Eosin staining showed little or no neurodegeneration in these regions. Synaptophysin. Synaptophysin is located in presynaptic vesicles in axonal endings and is used as a marker for synaptic integrity. 1,36,43 In both wildtype and APP/RK transgenic mice a similar pattern of immunostaining was observed in the neuropil in a characteristic granular pattern, distributed around the neuronal somata and along the dendritic arbors (Fig. 6A, B). Only the transgenic mice in group 3 displayed a disorganized synaptophysin staining pattern in the areas with degenerated neurons (Fig. 6E). Microtubule-associated protein 2. MAP-2 is a dendritic marker, most evident in cortical neurons (Fig. 6I), the hilus of the dentate gyrus, stratum radiatum (Fig. 6F, G), and the amygdala. Deviating patterns were again only observed in the transgenic

826

D. Moechars et al.

mice of group 3, with no immunopositive reaction detectable in degenerated neurons (Fig. 6H) in the hilus of the hippocampus and in cortical areas. Reduced immunoreactivity was evident in CA3 and amygdala while dendrites in cortex appeared dilated (Fig. 6J). Cathepsin D. Cathepsin D was implied in AD pathology 3 and present in senile plaques, but the enzyme is not involved in processing of APP. 32 We observed cathepsin D in a granular pattern within neurons and glial cells, most intensely in cortex, amygdala, globus pallidus, and hippocampal CA3 and CA4 pyramidal layers. The distribution and intensity of immunoreactivity for cathepsin D was not different in any of the APP/RK transgenic mice analysed. Ubiquitin. Ubiquitin has been associated with senile plaques and tangles in AD brain. 29 In mouse brain, ubiquitin immunoreactivity was located in nuclei and cytoplasm of all cell types and in the dendrites. Variable numbers of neurons presented different ratios of immunostaining in nucleus and cytoplasm, but this was not different in APP/RK transgenic mice relative to wild-type animals. Lipoprotein receptor related protein. LRP was immunohistochemically detected in punctate patterns, representing endocytotic vesicles, in many cell types in both transgenic and wild-type mouse brain. Neurons carried LRP predominantly on cell bodies and proximal processes and no consistent differences in distribution were evident between transgenic APP/RK and wild-type mice (results not shown). Receptor-associated protein. RAP, the ERbased chaperone of all members of the LDL receptor family, was immunochemically revealed in most areas of mouse brain. Differences between APP/ RK transgenic and control mice were restricted to apoptotic neurons in the brain of APP/RK animals in group 3, which were almost devoid of immunoreactive RAP (Fig. 6K, L). Glutamate receptors distribution Alterations in subtypes of glutamate receptor activity resulting in excitotoxicity and neuronal damage might, directly or indirectly, contribute to AD pathology. Many AD patients suffer from

seizures in which glutamate, as the most important excitatory neurotransmitter is thought to be involved. APP/RK transgenic mice have an increased tendency to spontaneous seizures and a markedly reduced sensitivity towards NMDA and increased reactivity to kainic acid. 24 To approach this aspect, we examined the spatial distribution of three subtypes of glutamate receptors by autoradiographic ligand binding. A marked regional heterogeneity was observed in binding levels of all three investigated glutamate receptor ligands with the highest concentration of NMDA and amino-3-hydroxy-5methyl-4-isoxazolepropionic acid (AMPA) binding sites in the hippocampus, whereas kainate binding sites were distributed complementary to NMDA sites. No marked differences were observed in regional distribution for the receptor ligands between mice from strain t/RK/F/4 and t/RK/F/6 and age-matched wild-type mice (Fig. 7). DISCUSSION

Mutant APP/RK was designed and proven to decrease secretion by a-secretase processing in neurons and transgenic mice expressing this mutant exhibited behavioural abnormalities, i.e. neophobia, aggression, increased tendency for spontaneous seizures, hyposensitivity to NMDA and hypersensitivity to kainic acid, and premature death. 24 We have generated additional transgenic mice that overexpress different isoforms of human APP resulting in essentially the same phenotype as the APP/RK mice. A more detailed analysis of these mice, including a thorough biochemical analysis, which is difficult to perform on the APP/RK mice because well-characterized antibodies are not available, and correlation to the phenotypic characteristics is ongoing. The present study established that the APP/RK transgene did not affect the expression of other genes, including endogenous APP. In addition, the phenomena of neurodegeneration, premature death and differential reactivity to glutamate analogues were approached in three different groups of APP/RK transgenic mice. First, expression of the transgene was defined by in situ hybridization and by immunohistochemistry. Transgenic APP/RK mRNA and protein, as evidenced by elevations in total APP, were detected prominently in neurons in the hippocampus, cortex and amygdala, which are regions associated with learning, memory and affective behaviour and are the most affected areas in AD patients. Strong

Fig. 6. Immunohistochemistry of mouse brain. (A–E) Synaptophysin immunoreactvity in hippocampus of (A) non-transgenic and (B) transgenic t/RK/F/4 mouse; same staining in hippocampal CA1 pyramidal layer and the strata radiatum and lacunosum moleculare of (C) non-transgenic, (D) transgenic t/RK/F/4 (group 2) and (E) transgenic t/RK/F/6 (group 3). Scale bars ˆ 100 mm. (F–H) MAP-2 immunostaining in hippocampus of (F) nontransgenic, (G) transgenic t/RK/F/4 (group 2) and (H) transgenic t/RK/F/6 mouse (group 3). Scale bars ˆ 100 mm. (I, J) MAP-2 immunoreactivity in cortex of (I) non-transgenic and (J) transgenic t/RK/F/6 mouse (same as H). Scale bars ˆ 20 mm. (K, L) RAP immunoreactivity in hippocampal CA1 of (K) non-transgenic and (L) transgenic t/RK/F/4 (group 3). Scale bars ˆ 100 mm.

Apoptosis in APP transgenic mice

Fig. 6.

827

828

D. Moechars et al.

Fig. 7. Distribution of glutamate receptors in brain of wild-type and APP/RK transgenic mice. Glutamate receptor radioassay with NMDA antagonist [ 3H]MK-801 (dizocilpine maleate; A, B), with [ 3H]kainic acid (C, D) and with [ 3H]AMPA (E, F). A, C and E show autoradiographs of sagittal cryosections of the brain of a nontransgenic mouse; B, D and F represent cryosections of the brain of transgenic APP/RK mice. Scale bars ˆ 1 mm.

immunoreaction for APP was striking in the dentate gyrus granular cells. The expression of the endogenous APP gene was clearly not affected by the transgene, proven by mRNA levels in mice aged from three to 15 months. Amyloid deposits and neurofibrillary tangles were not detected in the brain of APP/RK mice, neither by silver staining nor by immunohistochemistry with a battery of polyclonal and monoclonal antibodies against APP, the bA4 peptide and protein tau. In addition, patterns of immunostaining for many associated proteins, i.e. cathepsin D, ubiquitin, LRP, ApoE, among others, did not reveal consistent differences with wild-type mice. Post mortem histology of the brain of APP/RK mice that died prematurely revealed variable degrees of neurodegeneration. Similarly, in the brain of transgenic mice that were killed subsequent to an episode of seizures, neurodegeneration was generally evident as opposed to apoptosis, which was detected in a subset of APP/RK mice. The results obtained do not allow us to understand the mechanism of neurodegeneration in the presented experimental system. We can, however, exclude amyloid deposits, since these were absent in the brain of APP/RK mice. Apoptosis was not observed

in young mice or in mice that were healthy when killed, despite changed behaviour and occasional and infrequent mild seizures. APP/RK transgenic mice in the C57Bl genetic background, i.e. from line t/RK/B/2, did not show appreciable seizure activity, which is in line with known features of this mouse strain. The extent of neurodegeneration and apoptosis in these transgenic mice was, however, not markedly different from APP/RK mice in the FVB/N background, demonstrating that seizure activity as such was not a major determining factor. This also leads to the conclusion that apoptosis is not an early component of the clinical phenotype of the APP/RK mice. The general and extensive astrogliosis in the APP/ RK transgenic mice, with the exception of young animals, indicated that overexpression of APP was an insult for their brain. This type of activation of astrocytes is, however, rather unspecific and induced by diverse causes, demonstrating or indicating anything from subtle to severe injury. This was underlined by immunostaining for synaptophysin and MAP-2, revealing that in the vast majority of APP/RK transgenic mice, loss of synapses or damaged dendrites was not an early or major problem. As the number of factors and implied

Apoptosis in APP transgenic mice

components reported in the literature is extensive and growing, it remains a matter of strategy how this will be further analysed. CONCLUSION

The expression of other genes that code for proteins that are directly or indirectly associated with neurodegeneration in AD appeared largely unaffected by overexpression of the APP/RK transgene in brain: endogenous APP and its relatives APLP1 and APLP2, the receptors LRP, LDL-R and VLDL-R, their chaperone RAP, the LRP ligands, ApoE, lipoprotein lipase, and tissue plasminogen activator, and finally presenilin 1. This is both surprising and reassuring, because the phenotype of the APP/RK transgenic mice is severe and profound and deregulation of many unrelated brain systems might have been anticipated. We have now demonstrated that the observed phenotypic, behavioural

829

and other effects do not result from gross changes in neuronal metabolism. This is reassuring for the ongoing analysis of the functionality of the CNS in these and related transgenic mice, which continues to be promising for our understanding of the pathological processes in AD. Acknowledgements—The intellectual and material contributions of the following scientists are gratefully acknowledged: M. Gilis, C. Kuipe´ri, I. Laenen, B. De Strooper, P. Saftig, K. von Figura, D. Strickland, B. Greenberg and H. Van der Putten. This investigation was supported by the ’Fonds voor Wetenschappelijk Onderzoek-Vlaanderen (FWO-Vlaanderen), by FWOLotto, by the Action Program for Biotechnology of the Flemish government (VLAB-COT-008/IWT), by the 4th Framework EEC-Biotechnology program, by the Rooms-fund, by the Janssen Research Foundation, by Leuven Research and Development. DM thanks the IWT for a doctoral stipend and the K. U. Leuven Special Research Fund for a post-doctoral fellowship. We thank the K. U. Leuven for continuous support.

REFERENCES

1. Blennow K., Bogdanovic N., Alafuzoff I., Ekman R. and Davidsson P. (1996) Synaptic pathology in Alzheimer’s disease: relation to severity of dementia, but not to senile plaques, neurofibrillary tangles, or the ApoE4 allele. J. neural Transm. 103, 603–618. 2. Borchelt D. R., Thinakaran G., Eckman C. B., Lee M. K., Davenport F., Ratovitsky T., Prada C. M., Kim G., Seekins S., Yager D., Slunt H. H., Wang R., Seeger M., Levey A. I., Gandy S. E., Copeland N. G., Jenkins N. A., Price D. L., Younkin S. G. and Sisodia S. S. (1996) Familial Alzheimer’s disease-linked presenilin 1 variants elevate Abeta1-42/1-40 ratio in vitro and in vivo. Neuron 17, 1005–1013. 3. Cataldo A. M., Hamilton D., Barnett J. L., Paskevich P. A. and Nixon R. A. (1996) Properties of the endosomal-lysosomal system in the human central nervous system: disturbances mark most neurons in populations at risk to degenerate in Alzheimer’s disease. J. Neurosci. 16, 186–199. 4. Churukian C. J., Kazee A. M., Lapham L. W. and Eskin T. A. (1992) Microwave modification of Bielschowsky silver impregnation method for diagnosis of Alzheimer’s disease. J. Histotechnol. 15, 299–302. 5. De Strooper B., Van Leuven F. and Van den Berghe H. (1991) The amyloid b protein precursor or proteinase nexin II from mouse is closer related to its human homolog than previously reported. Biochim. biophys. Acta 1129, 141–143. 6. De Strooper B., Umans L., Van Leuven F. and Van den Berghe H. (1993) Study of the synthesis and secretion of normal and artificial mutants of murine amyloid precursor protein (APP): cleavage of APP occurs in a late compartment of the default secretion pathway. J. Cell Biol. 121, 295–304. 7. De Strooper B., Saftig P., Crassaerts K., Vanderstichele H., Guhde G., Annaert W., Von Figura K. and Van Leuven F. (1998) Deficiency of presenilin-1 inhibits the normal cleavage of amyloid precursor protein. Nature 391, 387–390. 8. Furukawa T., Masayuki O., Huang R. P. and Muramatsu T. (1990) A heparin binding protein whose expression increases during differentation of embryonal carcinoma cells to parietal endoderm cells: cDNA cloning and sequence analysis. J. Biochem. 108, 297–302. 9. Ga˚fvels M. E., Paavola L. G., Boyd C. O., Nolan P. M., Witmaack F., Chawla A., Lazar M. A., Bucan M., Angelin B. and Strauss J. F. (1994) Cloning of a complementary deoxyribonucleic acid encoding the murine homolog of the very low density lipoprotein/apoliprotein E receptor: expression pattern and assignment of the gene to chromosome 19. Endocrinology 135, 387–394. 10. Garvey W., Fathi A., Bigelow F., Jimenez C. L. and Carpenter B. F. (1991) Rapid, reliable and economical silver stain for neurofibrillary tangles and senile plaques. J. Histotechnol. 14, 39–42. 11. Haass C. and Selkoe D. J. (1993) Cellular processing of beta-amyloid precursor protein and the genesis of amyloid betapeptide. Cell 75, 1039–1042. 12. Hilbich C., Mo¨nning U., Grund C., Masters C. and Beyreuther K. (1993) Amyloid-like properties of peptides flanking the epitope of amyloid precursor protein specific monoclonal antibody 22C11. J. biol. Chem. 268, 26571–26577. 13. Hoffer M. J. V., van Eck M. M., Petrij F., van der Zee A., de Wit E., Meijer D., Grosveld G., Havekes L. M., Hofker M. H. and Frants R. R. (1993) The mouse low density lipoprotein receptor gene: cDNA sequence and exon-intron structure. Biochem. biophys. Res. Commun. 191, 880–886. 14. Horiuchi K., Tajima S., Menju M. and Yamamoto A. (1989) Structure and expression of mouse apolipoprotein E gene. J. Biochem. 106, 98–103. 15. Hsiao K. K., Borchelt D. R., Olson K., Johannsdottir R., Kitt C., Yunis W., Xu S., Eckman C., Younkin S., Price D., Iadecola C., Clark H. B. and Carlson G. (1995) Age-related CNS disorder and early death in transgenic FVB/N mice overexpressing Alzheimer amyloid precursor proteins. Neuron 15, 1203–1218. 16. Hsiao K., Chapman P., Nilsen S., Eckman C., Harigaya Y., Younkin S., Yang F. and Cole G. (1996) Correlative memory deficits, Ab elevation, and amyloid plaques in transgenic mice. Science 274, 99–102. 17. Kamboh M. I., Ferell R. E. and DeKosky S. T. (1998) Genetic association studies between Alzheimer’s disease and two polymorphisms in the low density receptor-related protein gene. Neurosci. Lett. 244, 65–68. 18. Kasper M., Lackie P., Haase M., Schuh D. and Muller M. (1996) Immunolocalization of cathepsin D in pneumocytes of normal human lung and in pulmonary fibrosis. Virchows Arch. 428, 207–215.

830

D. Moechars et al.

19. Kounnas M. Z., Morris R. E., Thompson M. R., FitzGerald D. J., Strickland D. K. and Saelinger C. B. (1992) The a2macroglobulin receptor/low density lipoprotein receptor-related protein binds and internalizes Pseudomonas exotoxin A. J. biol. Chem. 267, 12420–12423. 20. LaFerla F., Tinkle B., Bieberich C., Haudenschild C. and Jay G. (1995) The Alzheimer’s Ab peptide induces neurodegeneration and apoptotic cell death in transgenic mice. Nature Genet. 9, 21–29. 21. LaFerla F., Troncoso J. C., Strickland D. K., Kawas C. H. and Jay G. (1997) Neuronal cell death in Alzheimer’s disease correlates with ApoE uptake and intracellular Abeta stabilization. J. clin. Invest. 100, 310–320. 22. Lorent K., Overbergh L., Delabie J., Van Leuven F. and Van den Berghe H. (1994) Distribution of mRNA coding for alpha-2macroglobulin, the murinoglobulins, the alpha-2-macroglobulin receptor and the alpha-2-macroglobulin receptor associated protein during mouse embryogenesis and in adult tissues. Differentiation 55, 213–223. 23. Lorent K., Overbergh L., Moechars D., De Strooper B., Van Leuven F. and Van den Berghe H. (1995) Expression in mouse embryos and in adult mouse brain of three members of the amyloid precursor protein family, of the alpha-2-macroglobulin receptor/LDL receptor related protein and of its ligands apolipoprotein E, lipoprotein lipase, alpha-2-macroglobulin and the 40 kDa receptor associated protein. Neuroscience 65, 1009–1025. 24. Moechars D., Lorent K., De Strooper B., Dewachter I. and Van Leuven F. (1996) Expression in brain of amyloid precursor protein mutated in the a-secretase site causes disturbed behaviour, neuronal degeneration and premature death in transgenic mice. Eur. molec. Biol. Org. J. 15, 1265–1274. 25. Moechars D., Lorent K., Dewachter I., Baekelandt V., De Strooper B. and Van Leuven F. (1998) Transgenic mice expressing an alpha-secretase mutant of the amyloid precursor protein in the brain develop a progressive CNS disorder. Behav. Brain Res. 95/1, 55–64. 26. Nakamoto M., Ozawa M., Jacinto S. D., Furakawa T., Natori Y., Shirahama H., Yonezawa S., Nakayama T. and Muramatsu T. (1993) Mouse heparine binding protein-44 (HBP-44) associates with brushin, a high-molecular-weight glycoprotein antigen common to the kidney and teratocarcinomas. J. Biochem. 114, 344–349. 27. Namba Y., Tomonaga M., Kawasaki H., Otomo E. and Ikeda K. (1991) Apolipoprotein E immunoreactivity in cerebral amyloid deposits and neurofibrillary tangles in Alzheimer’s disease and Kuru plaque amyloid in Creutzfeldt-Jakob disease. Brain Res. 541, 163–166. 28. Ouimet C. C., Baerwald K. D., Gandy S. E. and Greengard P. (1994) Immunocytochemical localization of amyloid precursor protein in rat brain. J. comp. Neurol. 348, 244–260. 29. Perry G., Friedman R., Shaw G. and Chau V. (1987) Ubiquitin is detected in neurofibrillary tangles and senile plaque neurites of Alzheimer disease brains. Proc. natn. Acad. Sci. U.S.A. 84, 3033–3036. 30. Rebeck G. W., Harr S. D., Strickland D. K. and Hyman B. T. (1995) Multiple, diverse senile plaque-associated proteins are ligands of an apolipoprotein E receptor, the a2-macroglobulin receptor/low-density-lipoprotein receptor-related protein. Ann. Neurol. 37, 211–217. 31. Rickles R. J., Darrow A. L. and Strickland S. (1988) Molecular cloning of complementary DNA to mouse tissue plasminogen activator mRNA and its expression during F9 teratocarcinoma cell differentiation. J. biol. Chem. 263, 1563–1569. 32. Saftig P., Peters C., von Figura K., Craessaerts K., Van Leuven F. and De Strooper B. (1996) Amyloidogenic processing of human amyloid precursor protein in hippocampal neurons devoid of cathepsin D. J. biol. Chem. 271, 27,241–27,244. 33. Scheuner D., Eckman C., Jensen M., Song X., Citron M., Suzuki N., Bird T. D., Hardy J., Hutton M., Kukull W., Larson E., Levy-Lahad E., Viitanen M., Peskind E., Poorkaj P., Schellenberg G., Tanzi R., Wasco W., Lannfelt L., Selkoe D. and Younkin S. (1996) Secreted amyloid beta-protein similar to that in the senile plaques of Alzheimer’s disease is increased in vivo by the presenilin 1 and 2 and APP mutations linked to familial Alzheimer’s disease. Nature Med. 2, 864–870. 34. Selkoe D. J. (1996) Amyloid beta-protein and the genetics of Alzheimer’s disease. J. biol. Chem. 271, 18,295–18,298. 35. Strittmatter W. J. and Roses A. D. (1995) Apolipoprotein E and Alzheimer disease. Proc. natn. Acad. Sci. U.S.A. 92, 4725–4727. 36. Thiel G. (1993) Synapsin I, synapsin II, and synaptophysin: marker proteins of synaptic vesicles. Brain Pathol. 3, 87–95. 37. Tooyama I., Kawamata T., Akiyama H., Moestrup S. K., Gliemann J. and McGeer P. L. (1993) Immunohistochemical study of a2 macroglobulin receptor in Alzheimer and control postmortem human brain. Molec. chem. Neuropath. 18, 153–160. 38. Van Gool D., De Strooper B., Van Leuven F., Triau E. and Dom R. (1993) Alpha-2-macroglobulin expression in neuritic-type plaques in patients with Alzheimer’s disease. Neurobiol. Aging 14, 233–237. 39. Van Leuven F., Torrekens S., Overbergh L., Lorent K., De Strooper B. and Van den Berghe H. (1993) The primary sequence and the subunit structure of mouse a-2-macroglobulin, deduced from protein sequencing of the isolated subunits and from molecular cloning of the cDNA. Eur. J. Biochem. 210, 319–327. 40. Van Leuven F., Stas L., Raymakers L., Overbergh L., De Strooper B., Hilliker C., Lorent K., Fias E., Umans L., Torrekens S., Serneels L., Moechars D. and Van den Berghe H. (1993) Molecular cloning and sequencing of the murine alpha-2-macroglobulin receptor cDNA. Biochim. biophys. Acta 1173, 71–74. 41. Vidal F., Blangy A., Rassoulzadegan M. and Cuzin F. (1992) A murine sequence-specific DNA binding protein shows extensive local similarities to the amyloid precursor protein. Biochem. biophys. Res. Commun. 189, 1336–1341. 42. Wasco W., Gurubhagavatula S., Paradis M. D., Romano D. M., Sisodia S. S., Hyman B. T., Neve R. L. and Tanzi R. E. (1993) Isolation and characterization of APLP2 encoding a homologue of the Alzheimer’s associated amyloid b protein precursor. Nature Genet. 5, 95–100. 43. Wiedenmann B. and Franke W. W. (1985) Identification and location of synaptophysin, an integral membrane glycoprotein of Mr 38,000 characteristic of presynaptic vesicles. Cell 41, 1017–1028. 44. Younkin S. (1995) Evidence that A beta 42 is the real culprit in Alzheimer’s disease. Ann. Neurol. 37, 287–288. (Accepted 5 October 1998)