One of the hallmark pathological characteristics of Alzheimer's disease (AD) is the accumulation of dense plaques composed mainly of amyloid-β (Aβ) peptides. Aβ peptides result from the sequential proteolytic cleavage of amyloid-precursor protein (APP) by β-secretase and γ-secretase (for review, see Chapman et al., 2001). Mutations in the APP gene at or near these cleavage sites increase Aβ levels in the brains and are associated with early-onset, familial AD (Citron et al., 1992; Suzuki et al., 1994). Similarly, mutations in genes encoding presenilins, proteins implicated in the γ-secretase cleavage of APP, also increase Aβ peptide levels and cause early-onset, familial AD (for review, see Tanzi et al., 1996). Together, these data have formed the amyloid hypothesis of AD that asserts that Aβ may be involved in the pathogenesis and progression of AD. This hypothesis is supported by studies of transgenic mice expressing human APP with mutations associated with early-onset AD. These mice show age-dependent accumulation in Aβ(1-40) and Aβ(1-42), amyloid plaque deposition in the brain as well as functional deficits in measures of learning and memory, although the latter deficits are not replicated by all investigators (for review, see Janus and Westaway, 2001).

The focus on Aβ as a key pathological component of AD has led to examination of β- and γ-secretases as possible targets for therapeutic intervention. The inhibition of either β- or γ-secretase in vitro decreases the production of Aβ (Rishton et al., 2000; Shearman et al., 2000; Dovey et al., 2001; Steinhilb et al., 2001). The availability of transgenic mice simulating clinical brain amyloidosis has enabled in vivo, preclinical proof-of-concept studies to investigate the utility of β- or γ-secretase inhibitors for potential disease-modifying treatment of AD. Using young PDAPP mice expressing human APP with the “Indiana” mutation (V717F), Dovey et al. (2001) demonstrated dose-dependent decreases in cortical Aβ levels after acute administration of the γ-secretase inhibitor N-[N-(3,5-difluorophenacetyl)-l-alanyl]-S-phenylglycine t-butyl ester (DAPT). In addition, chronic administration of DAPT to young PDAPP mice significantly prevents age-dependent amyloid plaque deposition in the brain (May et al., 2001). These data strengthen the rationale to test the efficacy of γ-secretase inhibitors in treatment of AD.

To facilitate clinical studies of γ-secretase inhibitors, it is critical to have a pharmacodynamic biomarker that can provide evidence of pharmacological activity of the agent being tested. Because Aβ peptides are secreted in the extracellular fluid, including the cerebrospinal fluid (CSF) and plasma of AD patients as part of the constitutive APP metabolism (Ida et al., 1996; Seubert et al., 1992), we set out to test the hypothesis that Aβ levels in these clinically accessible fluids may provide a biomarker of γ-secretase inhibition. To this end, we had two specific aims. First, to replicate the observations of Dovey et al. (2001) on DAPT effects on cortical Aβ levels in young Tg2576 mice expressing human APP with the “Swedish” mutation (Hsiao et al., 1996). Second, we examined whether the γ-secretase inhibitor would reduce CSF and plasma Aβ levels in the Tg2576 mouse, independent of the status of amyloid plaque load in the brain. The latter study was critical because several reports suggest profound alterations in CSF Aβ and moderate changes in plasma Aβ levels in AD patients that indicate altered Aβ kinetics among the three interrelated compartments: the brain, CSF and plasma. Specifically, Aβ(1-42) levels in the CSF of AD patients are reduced by about 50% compared with age-matched controls (Motter et al., 1995; Nitsch et al., 1995; Andreasen et al., 1999; Mehta et al., 2000), although a few studies have not replicated these findings (for review, see Andreasen and Blennow, 2002). There have been two reports of an increase in Aβ levels in the plasma of AD patients (Matsubara et al., 1999; Mayeux et al., 1999), but several others show no change (Ida et al., 1996; Scheuner et al., 1996; Tamaoka et al., 1996; Mehta et al., 2000). Although not fully understood, one factor putatively contributing to the alteration in Aβ kinetics is thought to be the progressive accumulation of amyloid plaques in the brain. Indeed, both CSF and plasma Aβ levels are reduced age dependently in Tg2576 mice such that an inverse relationship exists between brain amyloid plaque deposition and CSF or plasma Aβ levels (Kawarabayashi et al., 2001). Hence, we considered it critical to evaluate modulation of Aβ dynamics in the CSF, brain, and plasma by amyloid plaque load after γ-secretase treatment. Our data demonstrate that γ-secretase inhibition lowers Aβ levels in both the CSF and plasma in the APP transgenic mouse and that amyloid plaque burden does not affect this response. The significance of these data is discussed.

Materials and Methods

Animals and in Vivo Drug Administration. Six- and 17-month old female transgenic mice overexpressing the human APP gene with the Swedish double mutation (K670N/M671L) under the transcriptional control of the hamster prion protein promoter (Tg2576 line; Hsiao et al., 1996) were used for the studies. In the first study, young and aged mice were dosed s.c. with vehicle (5% ethanol in corn oil) or DAPT solution at the final volume of 10 ml/kg. The doses of DAPT (10, 30, 100 mg/kg s.c.) and vehicle were selected to replicate the study by Dovey et al. (2001). In the second experiment, young and aged mice were given 100 or 200 mg/kg DAPT p.o., in the same vehicle. The route of administration was changed due to difficulty in dissolving the higher dose (200 mg/kg) of DAPT in the vehicle and to minimize the potential for drug precipitation after s.c. administration. For the oral study, DAPT was administered as a suspension at 10 ml/kg after an overnight fast. Each treatment group was composed of 15 mice at each age.

Tissue Isolation and Processing. Three hours after vehicle or drug administration, mice were anesthetized with a mixture of ketamine/xylazine (200/5 mg/kg s.c.). Muscle tissue was dissected away from the cisterna magna, the skull around the cisterna magna was cleaned to remove traces of blood, and CSF was collected via a fine-tipped pipette. Samples were frozen immediately on dry ice. Blood samples were obtained by cardiac puncture, collected in EDTA-coated tubes, and centrifuged (2400 rpm; 20 min) within1hof collection to separate the plasma. In the first study of s.c. DAPT administration, whole brains were removed and frozen immediately over liquid nitrogen. The dorsal hippocampus and prefrontal cortex were later dissected from 1-mm-thick coronal sections using a surgical knife. In all subsequent studies, the entire hippocampus (bilateral) and overlying cerebral cortex from one hemisphere were removed immediately after excising the brains from the skulls and frozen on dry ice. The brain samples were homogenized in 5 M guanidine buffer (5M guanidine HCl in 50 mM Tris-Cl, pH 8.0) at 1:10 dilution. Homogenates were agitated at room temperature for 3 to 4 h and then stored at —20°C. The day before running the ELISA, the 1:10 samples in 5 M guanidine were diluted further 1:10 for a final guanidine concentration of 0.5 M. The homogenates were then spun down at 14,000 rpm for 20 min at 4°C, and supernatants were used for Aβ determination by ELISA.

Aβ ELISA. CSF and plasma samples were thawed and diluted in blocking buffer (phosphate-buffered saline, 0.05% Tween 20, 1% bovine serum albumin); CSF volumes of 2 to 12 μl were diluted up to 125 μl, and plasma was diluted 1:2. Brain extracts from the young mice were diluted 1:100. Due to very high Aβ levels in the brains of old animals, cortical samples were diluted 1:20,000 for Aβ(1-40) and 1:12,000 for Aβ(1-42); hippocampal samples were diluted 1:10,000 for Aβ(1-40) and 1:8,000 for Aβ(1-42).

Half-volume, 96-well plates were coated overnight at 4°C with capture antibody 6E10 (Signet Laboratories, Dedham, MA) diluted 1:250 in 0.1 M NaHCO₃, pH 8.2. Standard curves were prepared in blocking buffer (or guanidine buffer for brain homogenates) from stock solutions of Aβ(1-40) and Aβ(1-42). Standards and samples were incubated for 3 h at room temperature. For detection, biotinylated rabbit polyclonal antibodies directed against Aβ(1-40) (Rb 209) and Aβ(1-42) (Rb 321) were used. After signal amplification with horseradish peroxidase-conjugated neutravidin, protein levels were visualized using a 3,3′,5,5′-tetramethylbenzidine peroxidase substrate kit (Kirkegaard and Perry Laboratories, Gaithersburg, MD); color development was stopped after 1 h with 1 M H₃PO4. Plates were read at 450 nm on a microplate reader, and 4-par logistics were used to extrapolate unknown values from the standard curves.

Measurement of DAPT Levels in Brain Tissue and Plasma. Analysis of DAPT concentration in brain tissue and plasma was performed by liquid chromatography mass spectroscopy after organic solvent extraction from brain homogenates and plasma. One hemisphere of frozen brain tissue was weighed and homogenized in 0.9% saline (1:1 by weight). Similarly, plasma samples were mixed with an equal volume of mobile phase buffer (20 mM ammonium acetate, pH 4.5). An equal volume of 25 μM flurbiprofen (internal standard) was added to each sample, and proteins were precipitated in acetonitrile. The samples were centrifuged and the resulting supernatants were used for DAPT measurement. For quantification, standard curves were constructed using brain homogenates or plasma from vehicle-treated animals containing DAPT and internal standard. Samples were injected onto the liquid chromatography mass spectroscopy and quantified by multireaction monitoring. The analysis was performed with a PE-Sciex, API 4000 triple-quadrapole mass spectrometer (Applied Biosystems, Toronto, ON, Canada) with a Turbo Ionspray source, equipped with a PerkinElmer series 200 micropump and series 200 autosampler (PerkinElmer Instruments, Norwalk, CT). Separation was performed with a Zorbax SB-CN (2.1 × 150 mm, 5 μm) column using acetonitrile and 20 mM ammonium acetate, pH 4.5, as mobile phases. Data acquisition and analysis were carried out with PE-Sciex Analyst software (version 1.2).

Statistical Analysis. For each age group, one-way analysis of variance was used to detect a significant treatment effect for Aβ(1-40) and Aβ(1-42) in the separate tissue compartments. After a significant treatment effect, individual dose differences were analyzed with a Fisher's protected least significant difference test; p < 0.05 was set as a statistically significant level for both tests.

Results

Effects of DAPT (10, 30, and 100 mg/kg s.c.) on Brain and Plasma Aβ Levels in Young Tg2576 Mice. As reported by Dovey et al. (2001), acute DAPT administration dose dependently reduced Aβ(1-40) (F_3,12 = 3.068, p < 0.05), and to some degree Aβ(1-42) (F_3,12 = 2.215, p < 0.10) in the cortex. The highest dose, 100 mg/kg s.c., was the only effective dose (p < 0.05) with respect to both Aβ(1-40) and Aβ(1-42), eliciting a 38 to 40% reduction compared with the respective vehicle-treated group (Fig. 1, A and B). No treatment effect was detected in the hippocampus for either Aβ(1-40) (F_3,12 = 2.146, p = 0.11) or Aβ(1-42) (F_3,12 = 2.248, p = 0.10).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 1.

A and B, effects of acute s.c. DAPT administration on Aβ levels in the cortex and hippocampus in young Tg2576 mice. Filled columns represent mean Aβ concentration ± S.E.M. for cortex; open columns represent mean Aβ concentration ± S.E.M. for hippocampus. C and D, mean Aβ concentration ± S.E.M. in plasma (*, p < 0.05; **, p < 0.001).

Consistent with greater levels of DAPT in the plasma than in the brain (Table 1), DAPT produced greater reductions in Aβ levels in the plasma than in the brain. Thus, a significant treatment effect was evident for both Aβ(1-40) (F_3,12 = 13.502, p < 0.001) and Aβ(1-42) (F_3,12 = 8.329, p < 0.001). The 100 mg/kg s.c. dose significantly reduced Aβ(1-40) (53% reduction, p < 0.001) and Aβ(1-42) (51% reduction, p < 0.001) compared with vehicle-treated mice (Fig. 1, C and D). Additionally, the 30 mg/kg s.c. dose effectively reduced Aβ(1-40) (35% reduction, p < 0.001) and Aβ(1-42) (27% reduction, p < 0.05); the 10-mg/kg dose had no significant effect.

Effects of DAPT (10, 30, and 100 mg/kg s.c.) on Brain, CSF, and Plasma Aβ Levels in Aged Tg2576 Mice. Plasma and brain DAPT levels in aged mice were comparable with those in young animals described above (Table 1). Acute DAPT had no effect on either cortical or hippocampal Aβ concentrations in the aged Tg2576 mice (Fig. 2, A and B). This is not surprising, given the fact that the brain extracts of older mice contained about 500 times as much Aβ as that seen in young mice, most of which likely represents Aβ extracted from amyloid plaque deposits.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 2.

Effects of acute s.c. DAPT administration on Aβ levels in the brains (A and B), CSF (C and D), and plasma (E and F) of old Tg2576 mice. Filled columns (A and B) represent cortical Aβ; open columns represent hippocampal Aβ (mean concentration ± S.E.M.; *, p < 0.05; **, p < 0.01, ***, p < 0.001).

In the CSF (Fig. 2, C and D), however, where Aβ concentrations resemble those of younger mice, DAPT significantly reduced Aβ(1-40) (F_3,12 = 4.246, p < 0.01) and Aβ(1-42) (F_3,12 = 3.58, p < 0.01). Treatment with 100 mg/kg s.c. DAPT produced a significant decrease in both Aβ(1-40) (42% reduction, p < 0.05) and Aβ(1-42) (49% reduction, p < 0.01). The 30 mg/kg s.c. dose produced a significant decrease (30%) in only Aβ(1-42) (p < 0.05).

In the plasma (Fig. 2, E and F), there was a strong treatment effect on both Aβ(1-40) (F_3,12 = 21.104, p < 0.001) and Aβ(1-42) (F_3,12 = 6.832, p < 0.001) as seen in the young mice. DAPT-treated mice had reduced Aβ(1-40) levels at both the 30-mg/kg (40% reduction, p < 0.001) and 100-mg/kg doses (65% decrease, p < 0.001). Aβ(1-42) was reduced in old mice by the 100-mg/kg dose only (55% reduction, p < 0.001).

Effects of DAPT (100 and 200 mg/kg p.o.) on Aβ Levels in the Brain, CSF, and Plasma in Young and Aged Tg2576 Mice. Oral DAPT produced somewhat more variable exposures in both the brain and plasma than those seen after s.c. administration (Table 1). However, the levels at 3 h after oral DAPT administration were comparable between young and aged mice as seen in the previous studies after s.c. administration. Additionally, DAPT concentrations in the plasma and brain did not differ between animals treated with 100- and 200-mg/kg doses.

Acute oral administration of DAPT produced a significant reduction of cortical Aβ(1-40) (F_2,13 = 31.575, p < 0.001) and Aβ(1-42) (F_2,13 = 5.293, p < 0.01), as well as hippocampal Aβ(1-40) (F_2,13 = 37.475, p < 0.001) and Aβ(1-42) (F_2,13 = 10.003, p < 0.001) in young mice (Fig. 3, A and B). Both 100 and 200 mg/kg DAPT significantly reduced Aβ(1-40) and Aβ(1-42) in cortex (p < 0.01) and hippocampus (p < 0.001) compared with the vehicle. DAPT (100 mg/kg p.o.) reduced Aβ(1-40) by 46 and 50% in the cortex and hippocampus, respectively. These reductions rose to 57% in the cortex and 56% in the hippocampus with the 200-mg/kg p.o. dose, but were not significantly different from the reductions seen after 100 mg/kg. In older animals, such a robust effect of DAPT on brain Aβ levels was not seen (Fig. 3, C and D), although a significant treatment effect was detected in only the cortex for Aβ(1-40) (F_2,13 = 4.012, p < 0.05), but not Aβ(1-42) (F_2,13 = 1.513, p = 0.23). The former was significantly reduced by the 100-mg/kg (23% reduction, p < 0.05) and 200-mg/kg (28% reduction, p < 0.05) dose. This treatment effect in the brains of old Tg2576 mice was not replicated, however, in a separate group of 16-month-old Tg2576 mice given either the vehicle or DAPT (100 mg/kg) orally. After vehicle, Aβ(1-40) was 45,754 ± 4,750 nM in the cortex and 28,090 ± 3,400 in the hippocampus. DAPT treatment resulted in Aβ(1-40) level of 40,959 ± 2,339 nM in the cortex (F_1,14 = 1.064, p = 0.31) and 24,889 ± 1,264 nM Aβ(1-40) in the hippocampus (F_1,14 = 0.693, p = 0.41). Levels of Aβ(1-42) were 3,512 ± 232 nM in cortex and 2,467 ± 189 in hippocampus after vehicle treatment, and 3,359 ± 169 nM in cortex (F_1,14 = 0.282, p = 0.6) and 2,214 ± 140 in hippocampus after DAPT (F_1,14 = 1.153, p = 0.29).

• Download figure
• Open in new tab
• Download powerpoint

Fig. 3.

Effects of acute p.o. DAPT administration on Aβ levels in the cortex and hippocampus in young (A and B) and old (C and D) Tg2576 mice. Filled columns represent mean Aβ concentration ± S.E.M. for cortex; open columns represent mean Aβ concentration ± S.E.M. for hippocampus.

Unlike the brain tissue, CSF did not show a significant difference in either Aβ(1-40) or Aβ(1-42) between young and old vehicle-treated mice. Acute, oral DAPT exerted a strong treatment effect on CSF Aβ(1-40) (F_2,13 = 13.536, p < 0.001) and Aβ(1-42) (F_2,13 = 6.018, p < 0.001) in both young and old mice. Consistent with similar exposures in the brain, both doses produced equivalent decreases in CSF Aβ peptide levels. CSF Aβ(1-40) was reduced by 72 to 74% in young mice (p < 0.001) and 53 to 58% in old mice (p < 0.001; Fig. 4A). Aβ(1-42) decreased by 46 to 58% in the young mice (p < 0.01) and 51 to 62% in the old mice (p < 0.001; Fig. 4B). There were no significant differences between the magnitude of DAPT-induced Aβ reduction in the CSF between the young and aged mice.

• Download figure
• Open in new tab
• Download powerpoint

Fig. 4.

Effects of acute p.o. DAPT administration on CSF (A and B) and plasma (C and D) Aβ. Filled columns represent young mice; open columns represent old mice (*, p < 0.05; **, p < 0.01; ***, p < 0.001).

As seen with s.c. doses of DAPT, the treatment effect in plasma was very robust for Aβ(1-40) (F_2,13 = 19.710, p < 0.001) and Aβ(1-42) (F_2,13 = 4.014, p < 0.001) in both young and old Tg2576 mice. The two doses tested did not differ from each other with respect to the magnitude of Aβ reductions, as expected on the basis of evidence of similar plasma concentrations (Table 1). Plasma Aβ(1-40) was reduced by 60 to 74% in young mice (p < 0.001), and by 51 to 53% in aged mice (p < 0.001; Fig. 4C). Plasma Aβ(1-42) decreased 35 to 39% in young mice (p < 0.001) and up to 30% in the old mice (p < 0.05; Fig. 4D). Levels of Aβ(1-40) and Aβ(1-42) did not significantly differ between young and old mice in vehicle-treated animals.

Discussion

The present study replicates and extends the findings of Dovey et al. (2001) by demonstrating that acute treatment with the γ-secretase inhibitor, DAPT, dose dependently reduces cortical Aβ levels in young Tg2576 mice expressing human APP with Swedish familial mutation (hAPP_Sw). The magnitude (38–40%) of reduction in brain Aβ(1-40) and Aβ(1-42) levels in young mice treated with DAPT (100 mg/kg s.c.) was consistent with the 40 to 50% reduction in total Aβ(1-x) levels reported by Dovey et al. (2001). In addition, the data provide evidence that DAPT dose dependently reduces Aβ peptide levels in both the CSF and the plasma of aged, plaque-bearing Tg2576 mice. The magnitude of Aβ reduction in the CSF and plasma of aged mice was comparable with that seen in the respective fluids from young Tg2576 mice. These results suggest that Aβ reduction provides a mechanism-based measure of γ-secretase inhibition, regardless of the presence of amyloid plaques. Hence, the data reported here provide support to a recently published mathematical model (Craft et al., 2002) of Aβ accumulation in AD that predicts that agents (such as β-or γ-secretase inhibitors) that inhibit the production of Aβ are likely to cause reductions in Aβ concentrations in the three major interrelated compartments: the brain, CSF, and plasma.

The absolute levels of hippocampal and cortical Aβ(1-40) and Aβ(1-42) differed between the two studies (s.c. versus p.o. administration) described here. We believe this may be due to the differential tissue dissecting techniques applied in the two studies. In the first study when DAPT was administered via the s.c. route, Aβ peptide levels were assessed in dorsal hippocampus and prefrontal cortex excised out of frozen brains from both young and aged mice. In contrast, the subsequent studies of oral DAPT administration used the entire cerebral cortex removed from one hemisphere and bilaterally dissected hippocampal formation from freshly removed brains. The assessment of only dorsal hippocampal Aβ concentration could be the reason also why hippocampal Aβ decreases were not evident in the young mice in the study assessing s.c. DAPT effects. It should be noted that the absolute levels of Aβ(1-40) and Aβ(1-42) when assessed from the whole cortex and hippocampus are in pretty good agreement with those reported by Kawarabayashi et al. (2001) for the Tg2576 mice.

In contrast to the CSF and plasma, DAPT reduced Aβ levels in the brain primarily in the young Tg2576 mice. Although a small decrease in brain Aβ(1-40) level was detected in aged mice after one of the oral DAPT studies (Fig. 3), this effect was not seen after either s.c. DAPT treatment or a subsequent oral DAPT efficacy study. The differential reduction in Aβ levels in young versus aged mice is likely attributed to different sources of Aβ in brain extracts from the two age groups. Thus, Aβ measured in brain extracts from young mice largely reflects undeposited, newly processed APP metabolism products, whereas in the aged mice, amyloid deposits are the major source of the Aβ peptides. This is supported by the data that brain Aβ concentrations in vehicle-treated, 17-month-old mice were about 500 times greater than those in young control mice. On the other hand, vehicle-treated mice showed no age-related differences in CSF or plasma Aβ concentrations. A recent report by Kawarabayashi et al. (2001) demonstrated age-dependent reduction in CSF and plasma Aβ levels in Tg2576 mice; however, a significant difference was not seen before 18 to 23 months of age in these mice. Hence, the inability to detect a significant age-related reduction in plasma or CSF Aβ levels in the present study may be due the somewhat younger age (16–17 months) of the mice used in our study.

The relationship between brain, CSF, and plasma Aβ pools with progression of the disease in AD patients remains to be fully elucidated. As discussed previously, several, although not all, studies demonstrate that AD patients have higher plasma and lower CSF Aβ peptide levels than age-matched controls (Motter et al., 1995; Nitsch et al., 1995; Matsubara et al., 1999; Mehta et al., 2001). These studies have led to the suggestion that sequestration of Aβ peptides by amyloid plaques may alter the kinetics of Aβ and thereby modulate also Aβ pharmacodynamics in response to therapeutic agents aimed at reducing brain Aβ levels. The data shown here clearly demonstrate that the ability of the γ-secretase inhibitor, DAPT, to reduce CSF and plasma Aβ levels is independent of brain plaque burden. Thus, in clinical trials of AD where all patients are likely to have a significant amyloid plaque burden, CSF Aβ may offer a viable biomarker of γ- and β-secretase inhibitors that block processing of APP to Aβ. Additionally, our data indicate that the decreases in CSF and plasma Aβ levels in older transgenic mice (Kawarabayashi et al., 2001), and some clinical samples from Alzheimer's patients may not result simply due to Aβ deposition into plaques; other factors such as neuronal degeneration or dysfunction may be involved.

Whether CSF and plasma Aβ reduction in the Tg2576 mice reflects inhibition of APP processing in the brain or the peripheral tissue is not entirely clear from the data shown here. In the Tg2576 mice, it is likely that Aβ is being generated in the periphery because the transgene (hAPP_Sw) expression can be detected in several peripheral organs (Kawarabayashi et al., 2001). Hence, peripheral γ-secretase inhibition may contribute to the observed plasma Aβ decreases. On the other hand, central injections of Aβ in mice and rats have been shown to diffuse fairly quickly into the blood and parenchymal vasculature (Ghersi-Egea et al., 1996; Shibata et al., 2000). These data suggest that central γ-secretase inhibition also may be reflected in plasma Aβ decreases after DAPT treatment. Thus, the utility of plasma Aβ decrease as a marker of central inhibition of APP processing needs to be further established. In contrast, it is likely that CSF Aβ reduction may primarily reflect inhibition of γ cleavage of APP in the brain. Although an active transport of Aβ from the plasma to the brain has been demonstrated (Zlokovic et al., 1993; Martel et al., 1996), the net contribution of the plasma Aβ pool to the total CSF Aβ pool may be relatively small because plasma Aβ levels are about 1/50th of those found in the brain. However, in view of the active transport mechanism for Aβ across the blood-brain barrier, definitive studies will be needed to clarify the role of peripheral γ-secretase inhibition on kinetics of central Aβ compartments. Recent studies of Aβ sequestering agents such as a high-affinity anti-Aβ antibody (DeMattos et al., 2001, 2002), gelsolin, and ganglioside GM1 (Matsuoka et al., 2003) suggest that sustained sequestration of peripheral Aβ reduces central Aβ levels. These studies suggest that there may be dynamic interactions between the central and peripheral pools of Aβ. It remains to be seen whether compounds that reduce plasma Aβ levels by inhibiting its production can also affect central Aβ levels. It would be interesting to test the effects of a nonbrain-penetrant γ- or β-secretase inhibitor to understand further the dynamics of Aβ pools in the plasma and central compartments.

Because CSF is in direct contact with the extracellular fluid in the brain, it is likely that CSF Aβ levels reflect Aβ in extracellular brain levels. The fact that DAPT reduces brain and CSF Aβ levels in the young, nonplaque-bearing mice but only CSF Aβ levels in the aged mice has led us to hypothesize that CSF Aβ levels are indicative of a nondeposited form of Aβ that can be modulated by γ-secretase. Future studies using differential extraction procedures to assess soluble versus nonsoluble pools of brain Aβ and their correlation with CSF Aβ will help characterize further the Aβ pools affected by after β- or γ-secretase inhibition.

In summary, the current study demonstrates the efficacy of an acute dose of a γ-secretase inhibitor in reducing Aβ(1-40) and Aβ(1-42) in vivo in both central and peripheral compartments of Tg2576 mice. To our knowledge, this is the first report examining the relationship between Aβ in brain, CSF, and plasma compartments in the presence or absence of brain amyloid plaques after inhibition of APP processing by a γ-secretase inhibitor. The data presented here suggest that CSF Aβ levels may be used as a biochemical biomarker to establish proof of mechanism in clinical studies of β- or γ-secretase inhibitors in Alzheimer's disease patients. However, studies of secretase inhibition on endogenous APP processing using nontransgenic animals are required to confirm the utility of plasma and CSF Aβ levels as a biomarker for clinical studies of sporadic Alzheimer's Disease patients. Of course, the clinical utility of β- or γ-secretase inhibitors as disease-modifying therapies for Alzheimer's disease will have to take into account the potential liability resulting from inhibition of processing of other substrates of these enzymes.