Age-dependent differential expression of BACE splice variants in brain regions of tg2576 mice

Neurobiology of Aging 26 (2005) 1167–1175

Age-dependent differential expression of BACE splice variants in brain regions of tg2576 mice O. Zohara,∗ , C.G. Pickb , S. Cavallaroc , J. Chapmand,e , A. Katzave , A. Milmanb , D.L. Alkona a

Blanchette Rockefeller Neurosciences Institute, Johns Hopkins University Academic and Research Building, 3rd Floor, Rockville, MD 20850, USA b Department of Anatomy and Anthropology, Tel-Aviv University Sackler School of Medicine, Tel-Aviv, Israel c Institute of Neurological Sciences, Italian National Research Council, 95123 Catania, Italy d Department of Neurology, Shiba Medical Center, Tel-Aviv University Sackler School of Medicine, Tel-Aviv, Israel e Department of Pharmacology and Physiology, Tel-Aviv University Sackler School of Medicine, Tel-Aviv, Israel Received 29 April 2004; received in revised form 20 September 2004; accepted 5 October 2004

Abstract Plaques found in the brains of patients suffering from Alzheimer’s disease (AD) mainly consist of ␤-amyloid (A␤), which is produced by sequential cleaving of amyloid precursor protein (APP) by two proteolytic enzymes, ␤- and ␥-secretases. Any change in the fine balance between these enzymes and their substrate may contribute to the etio-pathogenesis of AD. Indeed, the protein level and enzymatic activity of ␤-secretase (BACE), but not its mRNA level, were found elevated in brain areas of AD patients who suffer a high load of A␤ plaque formation. Similarly, increased BACE activity but no mRNA change was observed in a transgenic mouse model of AD, tg2576, in which over expression of the Swedish mutated human APP leads to A␤ plaque formation and learning deficits. Based on the recent demonstration of four BACE splice variants with different enzymatic activity, the discrepancy between BACE activity and mRNA expression may be explained by the altered BACE alternative splicing. To test this hypothesis, we studied the expression of all BACE splice variants in different brain areas of tg2576 mice at age of 4 months and 1 year old. We found developmental and regional differences between wild-type and tg2576 mice. Our results indicate that over expression of APP in tg2576 mice leads to the altered alternative splicing of BACE and the increase of its enzymatically more active splice variant (I-501). © 2004 Elsevier Inc. All rights reserved. Keywords: ␤-Secretase; Alternative splice variants; Brain regions; Age dependent; APP; ␤-Amyloid; tg2576 transgenic mice; Alzheimer disease

1. Introduction Aggregates of ␤-amyloid peptides (A␤) are the main constituent of senile plaques found in the brains of Alzheimer’s disease (AD) patients [5,9,18]. A␤ is a product of sequential cleaving of amyloid precursor protein (APP) by two proteolytic enzymes ␤- and ␥-secretases, which cleaves the APP at the N and C terminus of A␤, respectively. Any change in the fine balance between these enzymes and their substrate may contribute to the etio-pathogenesis of AD [11]. Increase in A␤ production, indeed, may result from upregulation of the ∗

Corresponding author. Tel.: +1 301 294 7174; fax: +1 301 294 7007. E-mail address: [email protected] (O. Zohar).

0197-4580/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.neurobiolaging.2004.10.005

substrate itself (i.e., APP) as in the case of Down’s syndrome (trisomy 21), or unbalanced production or activity of ␤- or ␥-secretases. ␤-Secretase is a type I transmembrane aspartic proteinase which belongs to the Asp2 family [13,15,20,24,27]. In mammals, BACE mRNA and protein are expressed mainly in the pancreas and the brain, but low expression levels can be found in many other body tissues [20,27,29]. BACE cleaves the APP at the N terminus of A␤, splitting it into a 100 kD soluble segment (APPs␤) and a 12 kD membrane anchored segment (C99) [10,27]. Next, the C99 is cleaved by ␥-secretase, creating the A␤ peptide. Three new alternative splice variants of the BACE gene were recently described [25]. These variants are the result of the alternative splicing of parts of exon

1168

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

3 and/or exon 4, which produces in frame deletion of 75 (I476), 132 (I-457) and 207 (I-432) nucleotides, and encodes protein isoforms with different enzymatic activity [3,6,25] and expression [31]. Increased expression of BACE protein and enzymatic activity have been recently demonstrated in neocortical and hippocampal brain regions of AD patients [7,12,30]. However, the increase in BACE protein has not been associated with elevation of its mRNA levels [7,12]. Similar to AD patients, BACE activity and not mRNA levels were found elevated in transgenic mice, termed tg2576, which express the Swedish mutation of human APP and develop age-dependent A␤ plaques and learning deficits [1,16,23]. Brains of tg2576 show about a 10-fold increase in A␤ levels by the age of 1 year [4,14,17,28]. In light of the existence of different BACE splice variants, the discrepancy between BACE activity and mRNA expression may be the result of altered expression of specific BACE isoforms. To test this hypothesis, we have determined the age-dependent differential expression of BACE and its four known splice variants in six brain regions of young and old tg2576 mice and compared them to their wild-type littermates. We compared the expression of all BACE splice variants at two different ages and in different brain areas of tg2576 and wild-type mice.

2. Methods 2.1. Reverse transcription real-time PCR Subjects were tg2576 mice and their wild-type littermates in two age groups, 4 months or 1 year (young transgenic (Ytg); old transgenic (Otg); young wild type (Ywt); old wild type (Owt)). Transgenic mice were generated by crossing tg2576 female with tg2576 males, both derived from C57Bl6/SJL F2 backcrossed to C57B6 breeders [14]. Mice were genotyped by PCR using DNA obtained from tail clippings. According to standard convention, N10 and beyond are considered fully congenic. Therefore, the mice examined had little or no variation in strain background. Mice were housed 2–6 per cage with ad libitum food and water, and maintained on a 12:12 light/dark cycle in a constant temperature (23 ◦ C). Animals were sacrificed by halothane overdose, decapitated, and the brain regions were rapidly removed and freezed in liquid nitrogen. Following the dissection, total RNA from the various brain regions was extracted using Trizol (Invitrogen, Carlsbad, CA). BACE primer pairs were designed to fit the Mus musculus BACE sequence (GenBank accession no. NM 011792.2). Single strand cDNA was synthesized by incubating total RNA (5 ␮g) with the reverse primer P4 (5 CCAATGATCATGCTCCCTCC-3 , corresponding to bases 1121–1140 of BACE; for positioning relative to the splice junction, see Fig. 1), 0.5 mM dNTP mix, 10 mM DTT, 1× first strand buffer and 50 U of SuperScriptTM II RT (Invitrogen, Carlsbad, CA), at 42 ◦ C for 50 min and then incubating

Fig. 1. Schematic drawing of BACE splice variants, their exons and the positioning of primers P1–P4 relative to the splice junctions. The vertical lines denote the splice junctions. The hedged and full areas denote the splice deletions (hedge exon 3, full exon 4).

at 70 ◦ C for 15 min. For real-time quantitative PCR, triplicate aliquots of the brain regions’ reverse transcribed RNA (0.4 ␮g), together with known amounts of external standards (purified PCR products, 104 –108 copies) were amplified with BACE primers. One reaction containing water instead of template was used as negative control. To quantify BACE expression, we amplified a region upstream to the alternative spliced site (P1, 5 -AGACGCTCAACATCCTGGTG3 , forward primer corresponding to bases 684–703 and P2, 5 -CCTGGGTGTAGGGCACATAC-3 , reverse primer corresponding to bases 811–830; Fig. 1). To control for RNA integrity and for differences attributable to errors in experimental manipulation from tube to tube, we normalized the data by dividing it by the expression level of S18 ribosomal RNA (rS18-15 -TCACCAAGAGGGCTGGAGAA-3 , rS1825 -CAGTGGTCTTGGTGTGCTGA-3 ). Each PCR reaction (final volume 20 ␮l) contained 0.5 ␮M of either BACE or rS18 primer pair, 10 ␮l of 2× QuantiTect SYBR Green PCR (Qiagen, Valencia, CA) and either 0.4 ␮g of reverse transcribed RNA, known amounts of external standards, or water. The following four step program was used for both BACE and S18 rRNA: (i) denaturation of cDNA (1 cycle: 95 ◦ C for 15 min); (ii) amplification (40 cycles: 94 ◦ C for 15 s, 58 ◦ C for 20 s, 72 ◦ C for 20 s); (iii) melting curve analysis (1 cycle: 95 ◦ C for 1 s, 65 ◦ C for 10 s, 95 ◦ C for 1 s); (iv) cooling (1 cycle: 40 ◦ C for 30 s). Temperature transition rate was 20 ◦ C/s except for the third segment of the melting curve analysis where it was 0.2 ◦ C/s. Fluorimeter gain value was 7. Real-time detection of SYBR Green I fluorescence intensity, indicating the amount of PCR products formed, was measured at the end of each elongation phase. Quantification of the amplification products and melting curve analysis were done using the LightCycler software (LightCycler, Roche, Indianapolis, IN). The log-linear phase of the curves of the unknown PCR products was fitted using the second derivative method, and compared to those of the known standards. Specificity of the PCR products was verified by melting curve analysis followed by gel electrophoresis (Fig. 2) and sequencing. Differences between the BACE expression levels in the

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

1169

Fig. 2. Real time quantification of BACE mRNA levels by RT-PCR. Total RNA from different brain areas and a known amount of external standards (purified PCR products, 104 –108 copies) were amplified in a parallel reaction. (A) Fluorimetric intensity of the external standards was measured at the end of each elongation phase. (B) Fluorescence values measured at the log-linear phase of the amplification were fitted by the second derivative maximum method and used to produce the standard curve that in turn was used to estimate the concentration of the unknown sample. (C) Gel electrophoresis of the amplificants verified the existence of one product of the correct length. Abbreviations: F: frontal cortex, P: parietal cortex, T/O: temporal and occipital cortex, H: hippocampus, C: cerebellum, S: brain stem, Ytg: young transgenic, Otg: old transgenic, Ywt: young wild type, Owt: old wild type.

mice brain areas were revealed using a set of four two-way ANOVA tests followed by the Holm–Sidak test for pair-wise comparisons (SigmaStat, SPSS Inc., Chicago, IL). We analyzed the difference between the following groups: Ytg versus Ywt, Ytg versus Otg, Otg versus Owt and Ywt versus Owt. 2.2. Detection of BACE splice variants Distribution of BACE splice variants was characterized by amplifying 0.4 ␮g of reverse transcribed RNA of the above four experimental groups using a primer pair flanking the splice junction (P3 and P4; Fig. 1). Each PCR reaction (final volume 20 ␮l) contained, 0.5 ␮M of the following primers: P3, 5 -CCCTACACCCAGGGCAAGTG-3 (forward primer, corresponding to bases 818–837) and P4, 5 -CCAATGATCATGCTCCCTCC-3 (reverse primer, corresponding to bases 1121–1140), 2.5 mM MgCl2 , 0.2 mM dNTP, 2 ␮l PCR Gold 10× buffer and 0.5 U of AmpliTaq Gold (Applied Biosystems, Foster City, CA). PCR ampli-

fications were performed at the log-linear phase using the following cycle programs: denaturation, 95 ◦ C for 10 min; amplification, 25 cycles of 95 ◦ C for 30 s, 56 ◦ C for 25 s and 72 ◦ C for 30 s; final extension, 72 ◦ C for 5 min. Amplification products were separated on a bis-acrylamide gel (8% TBE, Invitrogen, USA), stained with ethidium bromide imaged using a cooled CCD camera (LAS-1000 Plus; Fujifilm Corp., Tokyo, Japan) and analyzed using ImageGauge software (Fujifilm Corp., Tokyo, Japan). The mean density of each band area was measured and assigned a gray level between 0 (black) and 255 (white) (N = 6 mice per group). In each specific band lane a background area was selected and its mean intensity was subtracted from the specific band mean intensity. The size of the background area was identical in all lanes analyzed. Differences between the expression levels of BACE splice variants of the mice’ brain areas were assessed using a set of eight two-way ANOVA tests (SigmaStat 3.0, Chicago, IL). In the first four ANOVA tests we examined the differential expression of the splice variants in the brain re-

1170

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

gions of each experimental group (Ytg, Otg, Ywt and Owt). In four additional ANOVA tests we examined the difference in each of the splice variants’ expression within the brain regions of the following groups: Ytg versus Ywt, Ytg versus Otg, Ywt versus Owt and Otg versus Owt. Each ANOVA test was followed by Holm–Sidak test for pair-wise comparisons between either the specific brain region, or the specific splice variant of the compared groups.

was fitted and calculated by the second derivative maximum method to produce a standard curve (Fig. 2B). The concentration of the unknown BACE products was calculated using this standard curve. The specificity of the PCR products was verified by gel electrophoresis (Fig. 2C). Results obtained (Fig. 3) showed no significant change in total BACE expression between the four experimental groups: Ytg, Otg, Ywt and Owt. 3.2. Distribution of BACE splice variants

3. Results 3.1. Differential expression of BACE mRNA To quantify the absolute levels of BACE mRNA, we used real-time quantitative PCR and a primer pair for a region upstream of the splice junction, which does not discriminate between the different splice variants since it targets a region upstream of the splice junction (primers P1 and P2; Fig. 1). We explored BACE expression in six brain regions (frontal, parietal, temporal and occipital cortices, cerebellum and brain stem) of animals in four experimental groups; wt and tg mice at two developmental stages (4 months and 1 year old). BACE mRNA levels were calculated using known amounts of external standards (purified PCR product, 104 –108 copies) that were amplified in parallel reactions with the brain regions’ cDNA (Fig. 2) and normalized to the expression of rS18 which was also amplified in parallel with the above reactions. At the end of each elongation phase, the fluorescence intensity of the brain regions’ mRNA and the external standards were measured, indicating the amount of PCR product formed (Fig. 2A). The log-linear phase of the amplification

To characterize the specific expression of BACE splice variants in this mouse model of AD, we have performed extensive analyses of the variants’ distribution in the brain regions of the wild type and the transgenic mice. Later, we studied the effects of the transgene on the variants distribution and ontogeny. We found BACE variants to be expressed in all brain regions tested. However, the splices’ expression significantly varied between the transgenic and wild-type mice, was age dependent, and varied between brain regions. For the sake of clarity, the results will be organized in the following order: first we will briefly describe the methods used; next the results for each expression of the splice variant and the differences in their distribution with transgene and age will be described. The last two paragraphs will describe the ontogeny of all the splice variants together and the transgene effect on their distribution. To characterize the specific expression of BACE splice variants, we used RT-PCR with a primer pair flanking the splice junction (primers P3 and P4; Fig. 1). This primer pair produces four PCR products of 321, 246, 189 and 114 bp, which are indicative of each of the known BACE splice

Fig. 3. Expression levels of BACE-mRNA in various brain areas measured by real-time quantitative PCR. Copy number was calculated using the standard curve described in Fig. 2.

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

variants: I-501 (full length BACE), I-476, I-457 and I-432 BACE-mRNAs (for representative results, see Fig. 4). The expression of all BACE splice variants was assessed in six brain areas (frontal, parietal, temporal and occipital cortices, cerebellum and brain stem) of animals in four experimental groups, wt and tg mice at two developmental stages (4 months and 1 year old). Since the relative distribution of splice variants in a certain tissue is interdependent, we have used two sets of four two-way ANOVA tests to reveal the interactions. In the first set, we have compared the distribution of the splice variants in the various brain regions in each of our experimental groups (Tables 1 and 2). In the second set, we used a series of comparisons between the experimental groups (Ytg versus Ywt, Ytg versus Otg, Ywt versus Owt and Otg versus Owt) to reveal the differences in the distribution of each specific splice variant in the various brain regions (Table 3). Due to the complexity of the data analyses, only the most significant differences found will be described here. Of the four known splice variants of BACE, I-501 was expressed most and showed the biggest differences between our experimental groups. The most striking difference was found between the young transgenic and wild-type mice. In young transgenic mice, I-501 was expressed twice as much as the wild-type mice (Fig. 5A versus C; p < 0.05). This increase was evident when comparing either the I-501 expression levels from all the brain regions pooled together, or when comparing between each of the brain regions. Interestingly, the transgene had caused the ontogeny of I-501 expression levels

1171

Fig. 4. Distribution of BACE splice variants in the brain areas of young transgenic and wild-type mice. Please note the different expression patterns between transgenic and wild-type brain areas. In the transgenic mice only I501 was highly expressed, while brain areas of wild-type mice showed higher expression of both I-501 and I-476 when compared to the other splice variants. Numbers on the left denote the base number. I-501, I-476, I-457 and I-432 denote the four splice variants of BACE. Abbreviations: F: frontal cortex, P: parietal cortex, T/O: temporal and occipital cortex, H: hippocampus, C: cerebellum, S: brain stem.

Table 1 Differences between the expression levels of BACE splice variants within the brain regions of each experimental groups Frontal

Parietal

Tem/Occ

Hipp

Cereb

Stem

Young tg

501 > all

501 > all

501 > all

501 > all

501 > all 432 < all

501 > all 432 < all

Old tg

501 > all 476 > 457, 432

501 > all 457 > all

501 > all 476 > 457, 432 457 > all

501 > all 476 > 457, 432

501 > all 476 > 457, 432 457 > all

501 > all 476 > 457, 432

Young wt

501 > all 457 < all

501 > all 457 < all

501 > 476, 457 457 < all

501 > all

501, 476> 432 < 457

501 > all

Old wt

501 > all 476 > 432 > 457

501 > all 476 > 457

501 > all 476 > 432 > 457

501 > all 476 > 432, 457

501 > all 476 > 432, 457

501 > all 476 > 432 > 457

Variance were revealed using two-way ANOVA, followed by pair-wise multiple comparison procedure (Holm–Sidak method). Only significant differences are noted (p < 0.025 or smaller). Abbreviations: Frontal: frontal cortex; Parietal: parietal cortex; Tem/Occ: temporal and occipital cortex; Hipp: hippocampus; Cereb: cerebellum; Stem: brain stem; tg: transgenic mice; wt: wild-type mice; all: all splice variants; 501, 476, 457 and 432: the corresponding splice variants. Table 2 Differences within the expression levels of BACE splice variants between the various brain regions of each experimental group

Young tg Old tg Young wt Old wt

I-501

I-476

N > all H > F, C, S; P > F, C F > T/O, C P < F, H

C>F C > T/O, H, S P < all

I-457

I-432

T/O, C > all

N > all S < all P > C; S < all F > C, S; T/O > all; S < P, H

Variance were revealed using two-way ANOVA, followed by pair-wise multiple comparison procedure (Holm–Sidak method). Only significant differences are noted (p < 0.025 or smaller). Abbreviations: F: frontal cortex; P: parietal cortex; T/O: temporal and occipital cortex; H: Hippocampus; C: cerebellum; S: brain stem; N: neocortex areas (i.e., frontal, parietal and temporal/occipital cortices). tg: transgenic mice; wt: wild-type mice; all: all other brain regions.

1172

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

Table 3 Differences between the experimental groups of BACE splice variants expression levels within the brain regions

I-501 I-476 I-457

I-432

All regions

Frontal

Parietal

Tem/Occ

Hipp

Cereb

Stem

Yt > Yw, Ot Ow > Yw Ow > all Yw > Yt Yt > Ot, Yw Ow > Ot

Yt > Yw, Ot

Yt > Yw, Ot Ow > Yw

Yt > Yw, Ot Ow > Yw Ow > all

Yt > Yw Ow > Yw Ow > all

Yt > Yw, Ot Ow > Yw Ow > all

Yt > Ot, Yw Ow > Ot

Ow > Ot

Yt > Ot, Yw Ow > Yw

Ow > Ot

Yt > Yw Ow > Yw Ow > all Yw > Yt Yt > Yw

Yw > Yt

Yw > Yt

Ow > Yw

Yw > Yt

Ow > all

Ow > Ot

Ow > Yw Yw > Yt

Yw > Yt Ot > all

Yt > Yw Ow > Ot

Ot > Yt

Ow > Ot Ow > Yw

Variance were revealed using two-way ANOVA, followed by pair-wise multiple comparison procedure (Holm–Sidak method). Only groups who were significantly different are noted (p < 0.05). Abbreviations: Frontal: frontal cortex; Parietal: parietal cortex; Tem/Occ: temporal and occipital cortex; Hipp: hippocampus; Cereb: cerebellum; Stem: brain stem; Yt: young transgenic mice; Ot: old transgenic mice; Yw: young wild-type mice; Ow: old wild-type mice; all: all other experimental groups.

to change in opposite directions. In old transgenic mice, aging had reduced I-501 expression by one-third in the pooled brain areas (Fig. 5A versus B; p < 0.05). However, comparison of the specific brain areas revealed that the expression was reduced only in the frontal, parietal, temporal and occipital cortices and in brain stem while the hippocampal and the cerebellar levels did not change (p < 0.05). In the old wild-type mice on the other hand, aging had caused I-501 expression

to increase by 1.5-fold in the pooled brain areas and also in all pairs of brain regions, except the frontal cortex where it did not change (Fig. 5C versus D; p < 0.05). It is important to note, that the opposite effects of aging had caused the I-501 levels to be similar in the old transgenic and the wild-type mice. In contrast to I-501, I-476 was expressed more in the wild type than the transgenic mice. The frontal cortex and the

Fig. 5. Distribution of BACE splice variants in the brain regions of the experimental groups. Please note that in the young transgenic mice I-501 was expressed more than all the other groups. Also, the balance of the variants expression was different between the transgenic and wild-type mice. In the transgenic mice, I-501 was the main variant expressed while in the wild type the expression was more balanced between I-501 and I-476. Frontal, parietal, and temporal/occipital denote the corresponding cortices.

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

cerebellum of the young wild type showed 1.5-fold more I476 than the transgenic mice (Fig. 5A versus C; p < 0.05). Old wild-type mice showed the highest expression levels of I-476. In these mice, I-476 was expressed twice that of all other groups and in all pair-wise comparisons of the brain areas except the parietal cortex (Fig. 5; p < 0.05). In young transgenic mice, I-457 levels were the same as I476. However, in each of the other experimental groups it was the least expressed variant. Interestingly, I-457 expression showed age-dependent changes similar to I-501. Here also, young transgenic mice expressed more than the old transgenic mice and the old wild-type mice expressed more than the young wild-type mice (Fig. 5; p < 0.05). The expression of I-432 was the lowest in the young transgenic mice when compared to all other groups (p < 0.05). The ontogeny of the splice variants’ expression pattern differed between the wild-type and the transgenic mice. In the wild-type animals, the expression level of I-501, I-476 and I-457 was increased with age by about 150% in all brain regions. However, their expression pattern between brain regions remained similar (Fig. 5C and D). The expression of I-432 on the other hand, did not change with age. In both ages I-457 had the lowest expression level in the parietal cortex and the cerebellum and I-432 had higher expression levels in all the cortices and the hippocampus (p < 0.05). In the transgenic mice, the level of I-501 expression decreased with age in all the cortices and the brain stem by about one-third, while I-476 expression did not change. I-457 expression also decreased with age in the frontal, temporal and occipital cortices (p < 0.05). Expression of I-432 on the other hand, increased with age in the hippocampus, cerebellum and the brain stem (p < 0.05). The most striking difference between the transgenic and wild-type mice was the relative expression levels of I-501 (Fig. 5). In both young and old transgenic mice, I-501 was expressed 2–4-fold more than the other variants, while the other splice variants (I-476, I-457 and I-432) had comparable expression levels in both ages. In wild-type mice, I-501 was also the dominant variant expressed, but to a much lesser extent (1.5–2-fold). In contrast to the transgenic mice, the wild-type mice showed comparable expression levels of I501 and I-476 in both ages. In the young wild type I-432 also had a comparable level to I-501 and I-476, while in the old ones both I-457 and I-432 had lower expression levels than I501 and I-476. It is interesting to note that when all the splice variants’ densities were combined, their levels between the experimental groups were comparable and thus corroborate our light cycler results.

4. Discussion The recent identification of BACE has facilitated research on its role and regulation in AD. Its possible importance to AD pathology was furthered by the demonstration that BACE protein levels and enzymatic activity are elevated in areas of

1173

diseased brains that suffer from a high level of A␤ plaque formation [7,12,30]. Interestingly, this increase in protein production was not accompanied by a similar increase in BACE mRNA expression [8,12]. Similarly, brains of tg2576 mice which over express human APPsw and form A␤ plaques in several brain areas by the age of 9 months, do not express more total BACE mRNA compared to their wild-type littermates [1,16,23]. By itself, the over expression of APP cannot be the only factor explaining the increase in A␤ production in these mice. A necessary step to increase APP cleavage by ␤- and ␥-secretases leading to A␤ production must follow. Since BACE mRNA was not found to increase in the tg2576 mice, but its splice variants are known to have differential enzymatic activity, we have tested the hypothesis that the APP over expression in those mice affected the differential expression of BACE splice variants in the brain. Indeed, we observed that BACE expression is regulated by age-dependent post-transcriptional regulatory mechanisms in the brain regions of tg2567 mice. This is without changing the amounts of total BACE mRNA. We suggest that the differential expression of BACE may explain the observed increase in A␤ production and plaque formation in the mice brain. Similar to previous reports [1,8,12,16,23], we found that total BACE mRNA expression did not change in tg2576 mice and did not parallel the age-dependent A␤ production observed in their brains [14,19,26]. However, since the different splice variants of BACE show different enzymatic activity [3,25], we have further investigated the age-dependent distribution of BACE splice variants in the brains of tg2576 mice. We observed that the over expression of the human APPsw in the mice brain affected the differential expression of BACE splice variants in their brains. This suggested a possible regulatory role of APP on BACE transcripts. Although all the mice used in this study expressed the four previously described BACE splice variants [25], their expression profiles varied between the transgenic and wildtype mice in an age-dependent manner. In the transgenic mice of both ages, I-501 (the full length variant and the most enzymatically active) was expressed significantly more than the other splice variants. The wild-type mice, on the other hand, showed more balanced variants expression. In both transgenic and wild type, the expression of I-501 was age dependent, but behaved differently from the transgene. The transgenic young mice expressed two-fold more I-501 than their wild-type littermates while old transgenic and wild-type mice showed similar levels of I-501. Surprisingly, the ontogeny of I-501 differed between the wild-type and transgenic mice. In the transgenic mice, I-501 expression decreased with age while in the wild type its expression increased. Young transgenic mice also expressed slightly more I-501 in cortical areas, while in all other experimental groups I-501 expression was more balanced between the brain areas. In contrast to I-501, I-476 was expressed more in the wild type than the transgenic mice and showed age-dependent differential expression only

1174

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

in the wild-type mice. Old wild-type mice expressed twice as much I-476 than the young ones. The balance between the differential expression of the BACE splice variants was the biggest factor setting the experimental groups apart. In transgenic mice, I-501 was the major variant expressed while in the wild-type animals more balanced expression of the splice variants was observed. Several studies suggested that only a few fold change in BACE activity and production may cause AD pathology. Brains of AD patients and of transgenic mice expressing human BACE, for instance, show only a few fold increase in BACE production and activity, which does not correlate with the observed unchanged BACE mRNA levels in the same patients [2,7,12]. Here we also report unchanged mRNA levels in tg2576 mice; however, the post regulation mechanism in these mice brain skew BACE expression toward the increased expression of the most enzymatically active splice variant, I-501. Interestingly, the few fold increase of I-501 observed here, correlates well with the several fold increase of A␤ production in the plaque forming brains of tg2576 mice, over their wild-type littermates [14]. Taken together, these results suggest that the differential expression of BACE splice variants might contribute to increased A␤ production and plaque formation in the transgene brain. Alternative splicing is a powerful and commonly used mechanism of genetic control, processing pre-mRNA into multiple mRNA isoforms that affect animal physiology, development and disease. Changes in this versatile form of genetic regulation may affect the fine balance between substrate and the enzymes catalyzing its proteolytic processing. Thus, the differential expression of BACE variants with different brain distributions may have important implications for understanding the etiology of Alzheimer’s disease. The expression of these isoforms for example, could be differentially affected by feedback form its substrate, mutation, or polymorphism of the BACE gene and thus may provide a possible explanation for the broad spectrum of phenotypic abnormalities observed in patients with Alzheimer’s disease. The regulation, possible functions and brain distribution of BACE splice variants in the AD brain are still unclear. Nevertheless, the differential expression of its splice variants may sway APP turnover and A␤ production. The enzymatic activity of BACE variants was studied in several cellular systems. In HEK293 cells for instance, I-501 expression increased A␤ production while I-457 or I-476 expression induced less ␤-secretase activity [25]. Expressed in PtK2 or Cos7 cells, I-501 was found to be localized in the Golgi and showed ␤-secretase enzymatic activity, while I-457 and I-476 were found to be localized in the ER and had no affect on APP processing [3,6]. Interestingly, this correlates well with the findings described here. Most cases of AD are late onset and sporadic [22]. This is probably a result of multiple genetic and environmental risk factors such as head trauma [21,22]. These risk factors may influence the spatial and temporal distribution of BACE isoforms that in turn may affect A␤ plaque formation and age

of onset of AD in patients. Deciphering the physiological and pathological functions of the alternative-spliced transcripts of BACE lies ahead and may provide novel insights into the pathogenesis of Alzheimer’s disease and its cure.

References [1] Bigl M, Apelt J, Luschekina EA, Lange-Dohna C, Rossner S, Schliebs R. Expression of beta-secretase mRNA in transgenic tg2576 mouse brain with Alzheimer plaque pathology. Neurosci Lett 2000;292(2):107–10. [2] Bodendorf U, Danner S, Fischer F, Stefani M, Sturchler-Pierrat C, Wiederhold KH, et al. Expression of human beta-secretase in the mouse brain increases the steady-state level of beta-amyloid. J Neurochem 2002;80(5):799–806. [3] Bodendorf U, Fischer F, Bodian D, Multhaup G, Paganetti P. A splice variant of beta-secretase deficient in the amyloidogenic processing of the amyloid precursor protein. J Biol Chem 2001;276(15):12019–23. [4] Chapman PF, White GL, Jones MW, Cooper-Blacketer D, Marshall VJ, Irizarry M, et al. Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nat Neurosci 1999;2(3):271–6. [5] Citron M, Oltersdorf T, Haass C, McConologue L, Hung AY, Seubert P, et al. Mutation of the beta-amyloid precursor protein in familial Alzheimer’s disease increases beta-protein production. Nature 1992;360(6405):672–4. [6] Ehehalt R, Michel B, De Pietri TD, Zacchetti D, Simons K, Keller P. Splice variants of the beta-site APP-cleaving enzyme BACE1 in human brain and pancreas. Biochem Biophys Res Commun 2002;293(1):30–7. [7] Fukumoto H, Cheung BS, Hyman BT, Irizarry MC. Beta-secretase protein and activity are increased in the neocortex in Alzheimer disease. Arch Neurol 2002;59(9):1381–9. [8] Gatta LB, Albertini A, Ravid R, Finazzi D. Levels of beta-secretase BACE and alpha-secretase ADAM10 mRNAs in Alzheimer hippocampus. NeuroReport 2002;13(16):2031–3. [9] Golde TE, Eckman CB, Younkin SG. Biochemical detection of Abeta isoforms: implications for pathogenesis, diagnosis, and treatment of Alzheimer’s disease. Biochim Biophys Acta 2000;1502(1):172–87. [10] Gowing E, Roher AE, Woods AS, Cotter RJ, Chaney M, Little SP, et al. Chemical characterization of Abeta 17–42 peptide, a component of diffuse amyloid deposits of Alzheimer disease. J Biol Chem 1994;269(15):10987–90. [11] Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science 2002;297(5580):353–6. [12] Holsinger RM, McLean CA, Beyreuther K, Masters CL, Evin G. Increased expression of the amyloid precursor beta-secretase in Alzheimer’s disease. Ann Neurol 2002;51(6):783–6. [13] Hong L, Koelsch G, Lin X, Wu S, Terzyan S, Ghosh AK, et al. Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor. Science 2000;290(5489):150–3. [14] Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, Younkin S, et al. Correlative memory deficits, Abeta elevation, and amyloid plaques in transgenic mice. Science 1996;274(5284):99–102. [15] Hussain I, Powell D, Howlett DR, Tew DG, Meek TD, Chapman C, et al. Identification of a novel aspartic protease (Asp 2) as betasecretase. Mol Cell Neurosci 1999;14(6):419–27. [16] Irizarry MC, Locascio JJ, Hyman BT. Beta-site APP cleaving enzyme mRNA expression in APP transgenic mice: anatomical overlap with transgene expression and static levels with aging. Am J Pathol 2001;158(1):173–7. [17] Irizarry MC, McNamara M, Fedorchak K, Hsiao K, Hyman BT. APPSw transgenic mice develop age-related Abeta deposits and neu-

O. Zohar et al. / Neurobiology of Aging 26 (2005) 1167–1175

[18]

[19]

[20]

[21] [22]

[23]

[24]

ropil abnormalities, but no neuronal loss in CA1. J Neuropathol Exp Neurol 1997;56(9):965–73. Iwatsubo T, Odaka A, Suzuki N, Mizusawa H, Nukina N, Ihara Y. Visualization of Abeta 42(43) and Abeta 40 in senile plaques with end-specific Abeta monoclonals: evidence that an initially deposited species is Abeta 42(43). Neuron 1994;13(1):45–53. Kawarabayashi T, Younkin LH, Saido TC, Shoji M, Ashe KH, Younkin SG. Age-dependent changes in brain, CSF, and plasma amyloid (beta) protein in the tg2576 transgenic mouse model of Alzheimer’s disease. J Neurosci 2001;21(2):372–81. Lin X, Koelsch G, Wu S, Downs D, Dashti A, Tang J. Human aspartic protease memapsin 2 cleaves the beta-secretase site of betaamyloid precursor protein. Proc Natl Acad Sci USA 2000;97(4): 1456–60. Masters CL, Beyreuther K. Alzheimer’s disease. BMJ 1998; 316(7129):446–8. Rosenberg RN. The molecular and genetic basis of AD: the end of the beginning: the 2000 Wartenberg lecture. Neurology 2000;54(11): 2045–54. Rossner S, Apelt J, Schliebs R, Perez-Polo JR, Bigl V. Neuronal and glial beta-secretase (BACE) protein expression in transgenic tg2576 mice with amyloid plaque pathology. J Neurosci Res 2001;64(5):437–46. Sinha S, Anderson JP, Barbour R, Basi GS, Caccavello R, Davis D, et al. Purification and cloning of amyloid precursor protein betasecretase from human brain. Nature 1999;402(6761):537–40.

1175

[25] Tanahashi H, Tabira T. Three novel alternatively spliced isoforms of the human beta-site amyloid precursor protein cleaving enzyme (BACE) and their effect on amyloid beta-peptide production. Neurosci Lett 2001;307(1):9–12. [26] Terai K, Iwai A, Kawabata S, Tasaki Y, Watanabe T, Miyata K, et al. Beta-amyloid deposits in transgenic mice expressing human beta-amyloid precursor protein have the same characteristics as those in Alzheimer’s disease. Neuroscience 2001;104(2):299– 310. [27] Vassar R, Bennett BD, Babu-Khan S, Kahn S, Mendiaz EA, Denis P, et al. Beta-secretase cleavage of Alzheimer’s amyloid precursor protein by the transmembrane aspartic protease BACE. Science 1999;286(5440):735–41. [28] Westerman MA, Cooper-Blacketer D, Mariash A, Kotilinek L, Kawarabayashi T, Younkin LH, et al. The relationship between Abeta and memory in the tg2576 mouse model of Alzheimer’s disease. J Neurosci 2002;22(5):1858–67. [29] Yan R, Bienkowski MJ, Shuck ME, Miao H, Tory MC, Pauley AM, et al. Membrane-anchored aspartyl protease with Alzheimer’s disease beta-secretase activity. Nature 1999;402(6761):533–7. [30] Yang LB, Lindholm K, Yan R, Citron M, Xia W, Yang XL, et al. Elevated beta-secretase expression and enzymatic activity detected in sporadic Alzheimer disease. Nat Med 2003;9(1):3–4. [31] Zohar O, Cavallaro S, D’Agata V, Alkon DL. Quantification and distribution of beta-secretase alternative splice variants in the rat and human brain. Brain Res Mol Brain Res 2003;115(1):63–8.