SEARCH

SEARCH BY CITATION

Keywords:

  • amyloid precursor protein;
  • β-amyloid precursor protein cleaving enzyme;
  • β-cleavage;
  • β′-cleavage;
  • species-specificity

Abstract

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

β-amyloid peptides (Aβ) are produced by a sequential cleavage of amyloid precursor protein (APP) by β- and γ-secretases. The lack of Aβ production in beta-APP cleaving enzyme (BACE1)–/– mice suggests that BACE1 is the principal β-secretase in mammalian neurons. Transfection of human APP and BACE1 into neurons derived from wild-type and BACE1–/– mice supports cleavage of APP at the canonical β-secretase site. However, these studies also revealed an alternative BACE1 cleavage site in APP, designated as β′, resulting in Aβ peptides starting at Glu11. The apparent inability of human BACE1 to make this β′-cleavage in murine APP, and vice versa, led to the hypothesis that this alternative cleavage was species-specific. In contrast, the results from human BACE1 transgenic mice demonstrated that the human BACE1 is able to cleave the endogenous murine APP at the β′-cleavage site. To address this discrepancy, we designed fluorescent resonance energy transfer peptide substrates containing the β- and β′-cleavage sites within human and murine APP to compare: (i) the enzymatic efficiency; (ii) binding kinetics of a BACE1 active site inhibitor LY2039911; and (iii) the pharmacological profiles for human and murine recombinant BACE1. Both BACE1 orthologs were able to cleave APP at the β- and β′-sites, although with different efficiencies. Moreover, the inhibitory potency of LY2039911 toward recombinant human and native BACE1 from mouse or guinea pig was indistinguishable. In summary, we have demonstrated, for the first time, that recombinant BACE1 can recognize and cleave APP peptide substrates at the postulated β′-cleavage site. It does not appear to be a significant species specificity to this cleavage.

Abbreviations used

β-amyloid peptide

Abz

aminobenzoic acid

AD

Alzheimer's disease

APP

amyloid precursor protein

BACE1

beta-APP cleaving enzyme

FRET

fluorescent resonance energy transfer

MBP

maltose binding protein

mca

methylcoumarin

MES

2-(N-morpholino)ethanesulfonic acid

Y(NO2)

nitrotyrosine

β-amyloid peptides (Aβ), the main component of the amyloid plaque, are generated in a sequential proteolytic process by β- and γ-secretases on the amyloid precursor protein (APP). Genetic studies of Alzheimer's disease (AD) families revealed that mutations at or around the β- and γ-secretase cleavage sites are associated with the early onset of AD (Selkoe 1996). The mutation at the β-secretase cleavage site [Met596–Asp597 bond of APP695, or Met(− 1)–Asp(+ 1) bond of Aβ peptide] is known as the Swedish mutation and is characterized by the change of Lys595–Met596 to Asn595–Leu596 in APP695. Patients carrying this Swedish mutation display an early onset of the disease (Rossor et al. 1993; Citron et al. 1994; Haass et al. 1995) and pathological features of AD characterized by the massive amount of β-amyloid plaques in the autopsied brains (Rossor et al. 1993).

BACE1 (EC 3.4.23.46), beta-APP cleaving enzyme, was independently discovered through conventional protein purification (Sinha et al. 1999), expression cloning (Vassar et al. 1999) and bioinformatics (Yan et al. 1999; Hussain et al. 2000; Lin et al. 2000) approaches. This enzyme is known by several names, e.g. BACE1, BACE, memapsin 2 and Asp 2. BACE1 has characteristics attributed to the long sought β-secretase (Selkoe 1996). It is a typical type I transmembrane protein, containing a single transmembrane domain and a short cytoplasmic tail at the carboxyl terminus. Both, the transmembrane and cytoplasmic domains regulate the cellular localization and trafficking of BACE1 (Yan et al. 2001; Pastorino et al. 2002). Ser498 of BACE1, located within the carboxyl terminus, is phosphorylated by casein kinase I (Walter et al. 2001). The phosphorylation of Ser498 plays important roles in intracellular trafficking of BACE1. The expression of BACE1 in astrocytes is stimulated by interferon in vitro (Hong et al. 2003). These findings suggest that the activity and gene expression of BACE1 may be highly regulated.

BACE1 co-localizes with APP in the endosomal compartment (Hussain et al. 1999; Vassar et al. 1999; Lin et al. 2000). Genetically engineered mice deficient in BACE1 expression fail to produce detectable Aβ in the brain, suggesting that BACE1 is the principal β-secretase in neurons (Cai et al. 2001; Luo et al. 2001; Roberds et al. 2001). Thus, pharmacological inhibition of BACE1 may be a therapeutic strategy for diminishing Aβ production.

The transfection of the human APP gene into mouse neurons, expressing only endogenous murine BACE1, yielded principally Aβ(1–40) and Aβ(1–42) arising from the cleavage of both murine and human APP. Notably, an amino terminus truncated murine Aβ peptide, Aβ(11–40), was also detected. This murine Aβ(11–40) appeared to be derived from an alternative cleavage of endogenous murine APP by murine BACE1, designated β′- cleavage. No human Aβ(11–40) could be detected unless both human BACE1 and human APP genes were co-transfected into the mouse neurons (Cai et al. 2001). These findings led Cai et al. to propose that β′-cleavage of APP by BACE1 is a species-specific event, i.e. that murine BACE1 does not cleave the human APP at the β′-cleavage site (Tyr10–Glu11 bond of the human Aβ peptide). In contrast, both murine C99 (carboxyl terminal fragment of the β-cleavage product of APP by BACE1) and C89 (carboxyl terminal fragment of the β′-cleavage product of APP by BACE1) fragments are generated in the human BACE1 transgenic mice (Bodendorf et al. 2002). The latter results suggest that human BACE1 is able to perform the proteolytic cleavage of endogenous murine APP at the β′-site. However, the amount of BACE1 protein expressed in the human BACE1 transgenic mice is much higher than that of endogenous BACE1 in the wild-type mice. The excess of enzyme in the transgenic mice may alter the kinetics and the apparent specificity of BACE1 in this mouse model.

Human and murine BACE1 share 95% identity in their primary amino acid sequence. However, the high similarity of the primary sequence between human and murine BACE1 does not rule out the possibility that differences may exist in enzymatic efficiency and distinct pharmacological profiles of these two orthologs. A recent finding demonstrated a striking difference in molecular recognition despite a very high structural identity (Wise et al. 2003). Therefore, it is particularly important to determine whether there is any species specificity in the enzymatic properties and pharmacological profiles of BACE1, which might preclude the use of rodent models for developing BACE1 inhibitor.

In order to address any potential species-specific difference in the interaction between enzyme (BACE1) and substrate (APP), we designed several fluorescent resonance energy transfer (FRET) peptide substrates containing the known β- (+ 1) and the β′- (+ 11) cleavage sites within human and murine APP. We further characterized the inhibitory profile of both human and murine BACE1, using selected in-house BACE1 inhibitors. The binding rates of LY2039911 (sequence: Asp-Val-Asn-Leu*-Ala-Ala-Glu-Phe, where Leu* represents the transition state isostere derived from leucine), a close analog of OM99-2 (Hong et al. 2000), to recombinant human and murine BACE1 were also compared. The inhibitory potency of LY2039911 was measured for recombinant human BACE1 and endogenous BACE1 from brain tissues of PDAPP mice, (overexpression of human APP bearing Indiana mutation, transgene driven by PDGF promoter) CD-1 mice and guinea pigs. Collectively, results presented in this study provide new insights into the seemingly contradictory findings in the literature regarding species-specific difference between human and murine orthologs of BACE1.

Materials

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

FRET (Fluorescent Resonance Energy Transfer) peptide substrates containing β- (+ 1) and β′- (+ 11) cleavage sites of human and murine APP sequences were designed based both on physical properties of fluorophore/quencher pair and substrate requirements for BACE1. These peptides and StatVal BACE1 inhibitor were custom synthesized by Ana Spec Inc. (San Jose, CA, USA). All FRET peptides contain aminobenzoic acid (Abz) and nitrotyrosine [Y(NO2)]to serve as fluorophore and quencher, respectively. Ammonium acetate was purchased from EB bioscience (Gibbstown, NJ, USA). Protein A-agarose, amylose-sepharose, [2-(N-morpholino)ethanesulfonic acid (MES)] and bovine serum albumin (BSA) were obtained from Sigma Chemical Co. (St Louis, MO, USA). Phosphate-buffered saline (PBS) was procured from Invitrogen (Carlsbad, CA, USA). Flat-bottomed black 96-well plates were purchased from Corning Corp. (Corning, NY, USA). OM99-2 (BACE1 inhibitor) was acquired from Calbiochem (San Diego, CA, USA). Hydrogenated Triton X-100 and Complete Protease Inhibitor cocktail were obtained from Roche Molecular Biochemicals (Indeanapolis, IN, USA).

Both human (accession number: AF190725) and murine (accession number: NM_011792) BACE1 were cloned from total brain cDNA by RT-PCR. The nucleotide sequences corresponding to amino acid sequences #1–460 were inserted into the cDNA encoding human IgG1 (Fc) polypeptide (Vassar et al. 1999). This fusion protein of BACE1(1–460) and human Fc, named huBACE1:Fc, was constructed into the pJB02 vector. Human BACE1(1–460):Fc (huBACE1:Fc) and murine BACE1(1–460):Fc (muBACE1:Fc) were transiently expressed in HEK293 cells. 250 µg cDNA of each construct was mixed with Fugene 6 (Roche Bioscience, Indianapolis IN, USA) and added to 1 L of HEK293 cell suspension. Four days after the transfection, conditioned media were harvested for purification.

Conditioned media of HEK293 cells transiently transfected with huBACE1:Fc or muBACE1:Fc cDNA were collected. Cell debris was removed by filtering the conditioned media through a 0.22 µm sterile filter. A 5 mL volume of Protein A-agarose (bed volume) was added to 4 L conditioned media. This mixture was gently stirred overnight at 4°C. The Protein A-agarose resin was collected and packed into a low pressure chromatography column. The column was washed with 20× bed volumes of PBS at a flow rate of 20 mL/h. Bound huBACE1:Fc or muBACE1:Fc protein was eluted with 50 mm acetic acid, pH 3.6, at a flow rate of 20 mL/h. Fractions (1 mL) of eluate were neutralized immediately with 0.5 mL 200 mm ammonium acetate, pH 6.5. The purity of the final product was assessed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) in the 4–20% Tris–glycine polyacryamide gel. The enzyme was stored at −80°C in small aliquots.

Protein concentrations of purified recombinant human and murine BACE1:Fc were determined by the Bradford assay. An active site BACE1 inhibitor [OM99-2, Ki = 1.6 nm (Hong et al. 2000)] was used to titrate the fraction of the active enzyme in the purified preparation. OM99-2 was incubated, at concentrations ranging from 0.1 nm to 20 nm, with 10 nm human or murine BACE1:Fc and 30 µm methylcoumarin (mca)FRET peptide in the final reaction mixture in 50 mm ammonium acetate, pH 4.6, 1 mm Triton X-100 and 1 mg/mL BSA. The enzyme reaction was performed at room temperature for 8 h. The progress of reaction was recorded with excitation and emission wavelengths set at 330 and 400 nm, respectively, on a GEMINI (Molecular Devices, Palo Alto, CA, USA) fluorescence plate reader. The initial rate was computed from the slope of the reaction progress curve using a linear equation with the SoftMax Pro program (Molecular Devices). The dependence of initial rate on the inhibitor concentration was fitted with the Morrison equation.

All FRET peptides were designed to include either the β- or the β′-cleavage sites as well as Abz and Y(NO2) for the optimal quenching efficiency. The stock solution for each FRET peptide substrate was prepared at 30 mm in dimethylsulfoxide (DMSO), except for the murine β-site FRET peptide. Due to the poor solubility, the stock solution for the murine β-site FRET peptide was dissolved in DMSO to the final concentration of 1 mm. To maintain the concentration of DMSO in the final reaction mixture below 3%, the highest concentration of the murine β-site FRET peptide tested in the assay was 30 µm. The huBACE1:Fc and muBACE1:Fc preparations were concentrated through YM10 Centricon (Millipore Corp., Bedford, MA, USA) to a final concentration of at least 7 mg/mL. The optimal enzyme concentration for each FRET peptide substrate was determined individually at 30 µm FRET peptide substrate in 50 mm ammonium acetate, pH 4.6, 1 mg/mL BSA and 1 mm Triton X-100. The enzymatic efficiency (kcat/Km) of either of the BACE1 orthologs toward individual FRET peptide substrates at 15, 30 and 100 µm was determined under the optimal conditions for each substrate. The progress of the reaction was monitored by measuring an increase of the emission signal at 420 nm with excitation wavelength set at 320 nm, using a GEMINI fluorescence plate reader (Molecular Devices). Amino acid conjugated aminobenzoate was used to convert the emission signal in the relative fluorescence units into the molar concentration of product generated in the reaction mixture. The initial phase of the time-dependence curve was fitted with a linear function whose slope was used to calculate the initial rate for huBACE1:Fc or muBACE1:Fc toward each peptide substrate. The kcat/Km values were calculated from the linear dependence of the initial rate on the concentration of each peptide.

The fluorescence intensity of 50 µm (Abz)-E-I-S-E-V-N-L (BACE1 cleavage product peptide) was determined in the presence of an increasing concentration of substrate FRET peptide [f(S + Abz)], as well as alone for the product [f(Abz) or the substrate (f(S)] in 50 mm ammonium acetate, pH 4.6, with 1 mg/mL BSA and 1 mm Triton X-100. The fluorescence intensity of the product peptide at each substrate concentration was computed by subtracting f(S) from f(S + Abz), and then dividing by f(Abz) to calculate a correction factor for the inner filter effect. For any substrate concentration less than 30 µm, this correction factor was equal to 1 within the limits of experimental error. Thus, the inner filter effect for any substrate and product concentration reported here was negligible.

Each FRET peptide substrate was incubated overnight at 100 µm (in 50 mm ammonium acetate, pH 4.6, 1 mg/mL BSA and 1 mm Triton X-100) with an optimal enzyme concentration at room temperature. Due to the poor solubility of the murine β-site FRET peptide, 30 µm of that FRET peptide substrate were used for this experiment. The enzyme concentration for each FRET peptide was adjusted to obtain detectable peaks on an HPLC chromatogram. Specific conditions for each incubation mixture were as follows: (i) 1249 µg human BACE1:Fc or 732 µg murine BACE1:Fc were incubated with 100 µm human β′-site FRET substrate; (ii) 500 µg human BACE1:Fc or 293 µg murine BACE1:Fc were incubated with 100 µm murine β′-site FRET substrate; and (iii) 187 µg human BACE1:Fc or 110 µg murine BACE1:Fc were incubated with 100 µm human β-site FRETwt, 30 µm murine β-site FRET or 100 µm human β-site FRETswe peptide substrates, all in a 100 µL total volume. The reaction mixtures were analyzed via liquid chromatography–mass spectrometry (LC–MS) on an Agilent HP1100 system (Agilent Technologies, Wilmington, DE, USA) with electrospray interface and single quadrupole mass spectrometer by direct injection of the reaction mixture with no prior work-up. The mass spectrometer was run in positive ion mode with scanning over the range of 300–2500 atomic mass units (amu). The HPLC separation was carried out on a Zorbax Eclipse XDB-C8 (Agilent) 4.6 mm i.d. × 15 cm column (5 µm particle size) with a linear AB gradient of 10–100% B over 15 min, in which A = 0.05% trifluoroacetic acid (TFA)/H2O and B = 0.05% TFA in 60 : 40 acetonitrile (CH3CN):H2O and with a flow rate of 0.9 mL/min. Observed mass for each sample component was determined from deconvolution of multiple (generally two to three) charge species and averaging.

BACE1 inhibitors were dissolved in DMSO at 10 mm. Ten-point threefold serial dilutions in DMSO were prepared for each compound. Aliquots of 2 µL from all serial dilutions were added to the reaction mixture consisting of 15 µm mcaFRET peptide substrate (containing Swedish mutation at the BACE1 cleavage site) in 50 mm ammonium acetate, pH 4.6, with 1 mg/mL BSA and 1 mm Triton X-100. Either huBACE1:Fc or muBACE1:Fc in the same buffer was added to the mixture, to the final concentration of 0.2 nm, to initiate the reaction. The 96-well plate was covered with aluminum foil and kept on a shaker for 20 h at room temperature. The enzymatic activity was determined by the changes in the relative fluorescence during the 20 h incubation time. The 10-point inhibition curve was plotted and fitted with the four-parameter logistic equation to obtain the IC50 value.

The modified MBP-C125swe assay was adapted from Sinha et al. (1999). MBP-C125swe is a fusion protein of maltose binding protein (MBP) and the 125 residues of the carboxyl terminus of human APP bearing the Swedish mutation at the β-cleavage site. The fusion protein was expressed in Escherichia coliand purified through anionic exchange and amylose-sepharose affinity chromatography. Briefly, brain tissues of mice and guinea pigs were homogenized in 50 mm MES, pH 5.5, containing 0.1% Triton X-100 and 1X Complete Protease Inhibitor cocktail (Roche Biochemical). The homogenate was centrifuged at 1000 g for 30 min to remove cell debris. The resulting supernatant fluid (brain homogenate) was used as the enzyme source. The reaction mixture, containing 4 µg/mL MBP-C125swe in 50 mm ammonium acetate, pH 4.6, and 1 mm Triton X-100, was incubated for 1 h at room temperature. The product, MBP-C26swe, was quantitated by ELISA. The MBP-C26swe was captured with anti-MBP monoclonal antibody and detected with a rabbit polyclonal antibody recognizing the TEESEVNL sequence. The IC50 value for LY2039911 was determined with the same protocol.

Purified recombinant huBACE1:Fc or muBACE1:Fc was rapidly mixed with LY2039911 containing 15 µm mcaFRET substrate in 50 mm ammonium acetate, pH 4.6, with 1 mg/mL BSA and 1 mm Triton X-100. The progress of reaction was monitored by the changes in relative fluorescence emission using a FluoroMax2 spectrofluorometer (ISA Jobin Yvon Spex Instruments SA, Inc., Edison, NJ, USA). The resulting time dependence was fitted with Equations 1 and 2 to obtain kapp and subsequently on-rate (kon) and off-rate (koff). In eqn 1, a represents the fraction of the non-ligated enzyme; Vmax is the maximal velocity of the reaction; kapp stands for the observed on-rate; t is time in seconds; [I] denotes the inhibitor concentration.

  • image(1)
  • image(2)

Sequence and structural comparison

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

The primary sequences of human and murine BACE1 share 95% identity. The majority of differences between these two BACE1 orthologs reside within regions of signal peptide (72% identity) and propeptide (83% identity). The catalytic domains of these two ortholog proteins are highly conserved, with 97% identity. The structural locations of these substituted residues are highlighted in red to show their relative distance to each subsite binding pocket of the active site in human BACE1, based on the published co-ordinates of the crystal structure (Fig. 1a) (Hong et al. 2000). A cluster of substitutions in murine BACE1 equivalent to residues between Val81 and Thr87 in the human BACE1 primary sequence appears to be the most divergent. Notably, this cluster of residues is proximal to the first active site aspartate (Asp93). These non-conservative substitutions are located approximately 1 nm from the S1 binding pocket. Thus, they are not likely to influence the binding of substrates or inhibitors directly. Two conservative substitutions between human and murine BACE1 (Val227 and Ile385) are in close proximity to the S2, S3 and S4 subsite binding pockets of the active site, as shown in Fig. 1(b). The Val227 (alanine in human BACE1), within the loop A, is in the vicinity (approximately 0.5 nm) of the S3 binding pocket of the active site. Among all noted substitutions between these two proteins, only Ile385 in the human BACE1 (valine in murine BACE1) is located in the S2/S4 binding pocket of the active site. Although differences near the active site pocket exist, most of them are conservative substitutions, except for Val81, Ser83, Pro85, Gln86 and Thr87 in human BACE1. To determine the impact of these non-conservative substitutions on the conformation of the active site binding pocket and subsite preference of the enzyme, we designed the following biochemical and kinetic characterization of the human and murine BACE1 orthologs.

image

Figure 1. 3-D structural alignment of human and murine BACE1. (a) This illustrates that all substitutions between human and murine BACE1 are indicated in the human BACE1 structure, using the published co-ordinates of human BACE1 3-D structure complex with OM99-2. (b) This shows two conservative substitutions between the two ortholog proteins, Ile385 and Val227 of human BACE1, and their distance to major subsites of the active site binding pocket of human BACE1.

Download figure to PowerPoint

Cleavage site specificity

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

Figure 2 illustrates the amino acid sequence alignment of human and murine APP within the Aβ peptide region, including the two reported cleavage sites [β- (+ 1) and β′- (+ 11)] by BACE1. Two of three residues, which differ between human and murine Aβ peptides (residues #10 and 13), constitute the P1 and P3′ subsites of the proposed β′-cleavage site. To determine whether these changes lead to a species-specific difference in the BACE1 cleavage of APP, we designed FRET peptide substrates containing the β- or β′-cleavage sites as well as the same internal quenching pair. All peptide substrates used in this study contain Abz as a fluorophore and Y(NO2), which mimics phenylalanine, as the fluorescence quencher. The concentration of BACE1 was adjusted for each substrate to assure a linear progress of reaction for up to 6 h of incubation (data not shown). The enzymatic efficiency was calculated based on the initial rate. The absolute values for the rates were obtained using the standard curve for the conjugated aminobenzoate (the fluorophore), measured under the same conditions as the enzymatic reaction (data not shown). Sequences of FRET peptide substrates containing human and murine β- and β′-cleavage sites as well as enzymatic efficiencies of both enzymes toward this set of peptide substrates, at 15 and 30 µm, are shown in Tables 1 and 2, respectively.

image

Figure 2. Amino acid sequence alignment of human and murine APP around the Aβ peptide region. Residues differing between human and murine APP sequences are highlighted in red. Two known BACE1 cleavage sites, β- and β′-cleavage, are indicated with arrows.

Download figure to PowerPoint

Table 1.  Enzymatic efficiency of huBACE1:Fc or muBACE1:Fc toward 15 µm FRET peptide substrates with human or murine APP β- or β′-cleavage sites. All enzymatic reactions were conducted with 15 µm FRET peptide substrate in 50 mm ammonium acetate, pH 4.6, with 1 mg/mL BSA and 1 mm Triton X-100. Detailed conditions are described in the Methods. The peptide sequence of each FRET peptide substrate is shown in Table 3. Initial rates were measured to calculate the enzymatic efficiency. Data are expressed as mean ± SD from three to five independent experiments, each in duplicate
Enzymatic efficiency: kcat/Km (/min/m)Human β-site FRETsw peptideHuman β-site FRETwt peptideMurine β-site FRET peptideHuman β′-site FRET peptideMurine β′-site FRET peptide
huBACE1:Fc58 079 ± 4238608 ± 60802 ± 1049.0 ± 0.345.0 ± 3.4
muBACE1:Fc80 509 ± 3921865 ± 901155 ± 12810.9 ± 0.259.9 ± 4.1
Table 2.  Enzymatic efficiency of huBACE1:Fc or muBACE1:Fc toward 30 µm FRET peptide substrates with human or murine APP β- or β′-cleavage sites. All enzymatic reactions were conducted with 30 µm FRET peptide substrate in 50 mm ammonium acetate, pH 4.6, with 1 mg/mL BSA and 1 mm Triton X-100. Detailed conditions are described in the Methods. The peptide sequence of each FRET peptide substrate is shown in Table 3. Initial rates were measured to calculate the enzymatic efficiency. Data are expressed as mean ± SD from three to five independent experiments, each in duplicate
Enzymatic efficiency: kcat/Km (/min/m)Human β-site FRETsw peptideHuman β-site FRETwt peptideMurine β-site FRET peptideHuman β′-site FRET peptideMurine β′-site FRET peptide
huBACE1:Fc50 478 ± 1144522 ± 16503 ± 329.6 ± 0.553.8 ± 2
muBACE1:Fc75 859 ± 1539808 ± 451139 ± 1815.5 ± 0.386 ± 2

We also performed the active site titration of the purified recombinant human and murine BACE1:Fc used in this study. As determined by the active site titration with OM99-2, the concentration of active enzyme was equal to 0.9 and 1.2 of the concentration determined by the Bradford protein assay for the human and murine BACE1:Fc, respectively (data not shown). Since these values were within experimental error of each other, we made no attempt to correct the values of enzymatic efficiency, reported below, for the concentration of the active enzyme.

The enzymatic efficiency of huBACE1:Fc (608 ± 60/min/m) toward the β-cleavage site containing the wild-type human APP sequence is similar to that of murine enzyme (865 ± 90/min/m). This rate was nearly 100-fold slower than that toward the APP sequence containing the Swedish mutation in the FRET assay for huBACE1:Fc (58079 ± 4238/min/m). Similar trends were also found for both BACE1 orthologs toward the murine APP β-cleavage site. The muBACE1:Fc was capable of processing both human (10.9 ± 0.2/min/m) and murine (59.9 ± 4.1/min/m) APP substrates at the postulated β′-cleavage site. Likewise, the huBACE1:Fc was also able to cleave both human (9.0 ± 0.3/min/m) and murine (45.0 ± 3.4/min/m) APP substrates at the postulated β′-cleavage site. The enzymatic efficiencies for huBACE1:Fc (45.0 ± 3.4/min/m) and muBACE1:Fc (59.9 ± 4.1/min/m) toward the murine APP sequence at the postulated β′-cleavage site were similar. Notably, the activities of both enzymes toward murine β′-cleavage site were five times greater than those toward the human β′-cleavage site.

Enzymatic efficiency of human and murine BACE1 toward these FRET peptide substrates at 30 µm (Table 2) was similar to those at 15 µm (Table 1) and 100 µm (data not shown), when corrected for the inner filter effect. These results suggest that both substrate concentrations are below Km values for these peptide substrates. Consequently, only kcat/Km values could be calculated. This conclusion is consistent with the linear dependence of enzymatic activity on substrate concentration in the range of 10–300 µm for both BACE1 orthologs (data not shown), except for murine β-site FRET peptide (due to poor solubility of this peptide. For details, see Methods). Consequently, the IC50 values of BACE1 inhibitors measured under the experimental conditions approximate the Ki values of these inhibitors well.

Table 3 shows the cleavage site by human and murine BACE1 within each FRET peptide as determined by LC/MS (example shown in Figs 3a–c). In order to achieve a detectable peak in the HPLC chromatogram, the enzyme concentration used in this experiment was five to 50 times higher than that used in the determination of enzymatic efficiency. Nevertheless, all human and murine APP FRET peptides were cleaved only at a single site by either of the BACE1 orthologs.

Table 3.  Determination of cleavage site within FRET peptides derived from human and murine APP sequences. Detailed conditions and procedures are described in the Methods. The bond between the two amino acids denoted by bold- and italic-font is the cleavage site identified by LC/MS
FRET substratesIntact peptide massAmino terminal fragment massCarboxyl terminal fragment mass
NameSequence Mcalc Mobs McalcMobs (huBACE1)Mobs (muBACE1) McalcMobs (huBACE1)Mobs (muBACE1)
  1. Mcalc = calculated mass; Mobs = observed mass.

Human β-site FRETsw(Abz)-EISEVN L D AEY(NO2)RHDKG2039.12038.4922921.5921.31135.11134.71134.4
Human β-site FRETwt(Abz)-EISEVK M D AEY(NO2)RHDKG2071.22070.7954.1953.5953.51135.11134.41134.4
Murine β-site FRET(Abz)-EISEVK M D AEY(N)2)GHDKG1972.11971.4954.1953.5953.510361035.51035.5
Human β-site FRETY(NO2)-RHDSG Y E VHHQK-(Abz)1819.91819.4941.9941.2941.3896895.4898.5
Murine β-site FRETY(NO2)-GHDSG F E VRHQK-(Abz)1723.81723.3826.8826.2826.2915914.5914.5
image

Figure 3. Cleavage of human β′-site FRET peptide by huBACE(– TM) and muBACE:Fc. Detailed experimental conditions are described in the Methods. LC/MS spectrum of (a) enzyme blank control, (b) human BACE1:Fc and (c) murine BACE1:Fc.

Download figure to PowerPoint

In summary, there was a small but notable difference in the enzymatic efficiency of the human and murine BACE1 toward the FRET peptide substrate containing human β′-cleavage site. We further investigated whether this small difference in the subsite preference of these two ortholog enzymes affects the pharmacological profiles and binding rates of LY2039911.

Pharmacological characterization of human and murine BACE1

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

We used eight BACE1 inhibitors, including two known BACE1 inhibitors [OM99-2 (Hong et al. 2000) and StatVal (Sinha et al. 1999)] to characterize these two orthologs. Figures 4(a) and (b) show the molecular structures of OM99-2 and StatVal. A new FRET substrate containing methylcoumarine (mca) was used for this experiment to increase the sensitivity (0.2 nm compared with 20 nm sensitivity for Abz containing FRET substrates) of the assay to measure IC50 values for potent compounds. The potency of these molecules ranges over more than three orders of magnitude (Table 4). Comparison of the inhibitory potency for these eight and additional 51 (data not shown) inhibitors against huBACE1:Fc and muBACE1:Fc shows a strong correlation (r2 = 0.98) with an intercept near zero (data not shown). This result suggests that there is little difference in the pharmacological profile of human and murine BACE1.

image

Figure 4. Molecular structures of (a) OM99-2, (b) StatVal and (c) LY2039911.

Download figure to PowerPoint

Table 4.  Comparison of potency of BACE1 inhibitors toward human and murine BACE1:Fc. Detailed experimental procedure of measuring IC50 values for each compound is described in the Methods. Data are expressed as mean ± SD from three independent experiments
Designation of compoundAverage IC50 human StructureAverage IC50 murine
BACE1 (µM)BACE1 (µM)
Compound AEVY(statine)VAEF0.118 ± 0.0110.126 ± 0.013
Compound BEVG(statine)VAEF5.319 ± 0.1324.673 ± 0.155
Compound CEVE(statine)VAEF0.450 ± 0.0220.452 ± 0.017
Compound DEVL(statine)VAEF0.265 ± 0.0210.251 ± 0.021
Compound EEVW(statine)VAEF7.544 ± 0.7026.666 ± 0.263
Compound FEVD(statine)VAEF0.651 ± 0.0220.603 ± 0.020
StatValKTEEISEVN(statine)VAEF0.055 ± 0.0030.049 ± 0.003
OM99-2EVNL*A AEF0.0016 ± 0.00010.0014 ± 0.0002

Potency for LY2039911 against endogenous and recombinant BACE1

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

We further compared the inhibitory potency of LY2039911 (structure shown in Fig. 4c) toward recombinant huBACE1:Fc and native BACE1, obtained from brain tissues of PDAPP mice, CD-1 mice and guinea pigs, in the MBP-C125swe assay (Sinha et al. 1999). In this assay, the polypeptide substrate contains 125 residues at the carboxyl terminus of human APP containing the Swedish mutation at the β-cleavage site. The products (MBP-C26swe) were quantitated with an ELISA specifically detecting the β-cleavage peptides. The IC50 values for LY2039911 were 2.3, 2.9 and 2.1 nm against native BACE1 enzyme from brain tissues of PDAPP mice, CD-1 mice and guinea pigs, respectively (Fig. 5a). The IC50 value for this compound was 2.5 nm when measured against recombinant huBACE1:Fc in the mcaFRET assay (Fig. 5b). These results suggest that there is little difference in the interaction of LY2039911 with native, full-length, membrane-associated BACE1 from rodents compared with a human recombinant BACE1:Fc.

image

Figure 5. IC50 values for LY2039911 against endogenous BACE1 from PDAPP mice, CD-1 mice and guinea pig brain as well as huBACE:Fc. Detailed experimental conditions are described in the Methods. (a) Concentration dependence of inhibition by LY2039911 for endogenous BACE1 from PDAPP mice (filled circles), CD-1 mice (filled triangles) and guinea pig (filled squares) brain homogenate was determined in the MBP-C125swe assay. The IC50 values for LY2039911 against each endogenous enzyme were determined in two independent experiments. The average IC50 values for LY2039911 were 2.3, 2.9 and 2.1 nm for brain homogenate of PDAPP mice, CD-1 mice and guinea pig, respectively. (b) The IC50 value, 2.5 nm, for LY2039911 against recombinant huBACE:Fc was also determined in the mcaFRET assay.

Download figure to PowerPoint

Binding rate of LY2039911

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

We further used binding kinetics of a BACE1 inhibitor, LY2039911, to probe for possible differences between active sites of human and murine BACE1. The time-dependent progression of reaction was recorded with a spectrofluorometer set at excitation and emission wavelengths of 330 nm and 406 nm, respectively. The resulting time-course curves (data not shown) were fitted with eqn 1 to determine the apparent on-rate (kapp). Subsequently, the apparent on-rates (kapp) determined at different inhibitor concentration were re-plotted in the secondary plot (data not shown) to determine the actual on-rate (kon) and off-rate (koff) (eqn 2). Results, summarized in Table 5, demonstrate that the binding kinetics for LY2039911 and either of these two BACE1 orthologs are very similar. To minimize the impact of tight binding on the accuracy of the rate determination, under all conditions, the concentration of the inhibitor was five times that of the enzyme. As judged from the time course of the progress of reaction, the binding of the substrate to BACE was at least 10 times faster than that of the inhibitor (data not shown). Consequently, the interaction of substrate with enzyme was considered to be in rapid equilibrium.

Table 5.  Determination of on-rate and off-rate for LY2039911 binding to huBACE1:Fc and muBACE1:Fc. Both huBACE1:Fc and muBACE1:Fc were expressed in HEK293 cells. All kinetic rates for LY2039911 were measured with 15 µm BACE1mcaFRET peptide in 50 mm ammonium acetate, pH 4.6, 1 mg/mL BSA and 1 mm Triton X-100. The progress of reaction was monitored by the changes in relative fluorescence emission at 406 nm with excitation wavelength at 330 nm, using a FluoroMax2 spectrofluorometer. The resulting time dependence was fitted with eqn 1 and eqn 2 to obtain kapp and subsequently on-rate (kon) and off-rate (koff)
Enzymekon (/m/s)koff (/s)
  1. Data are expressed as mean ± SE from two independent measurements.

huBACE:Fc(5.7 ± 0.27) × 104(5.2 ± 3.3) × 10−5
muBACE:Fc(3.8 ± 0.45) × 104(1.06 ± 0.55) × 10−4

Discussion

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References

High sequence homology and a small number of substitutions around the active site of the enzyme may lead to the conclusion that there should be little difference in the molecular recognition of substrates by either of these two BACE1 orthologs. However, a recent publication provided a striking example of minute differences in the primary sequence associated with very large changes in the affinity of molecular interaction. The human nicotinic acid receptor (HM74A) and its close paralog HM74 share 96% identity in the primary sequence and differ by only 15 amino acids within the overlapping sequence. Nevertheless, their affinity for nicotinic acid differs by a factor of 1000 (Wise et al. 2003). Therefore, considering apparently conflicting reports, we compared: (i) the enzymatic efficiency of human and murine BACE1 toward FRET substrates containing β- and β′-cleavage sites of APP; (ii) the pharmacological profiles of these two enzymes toward a set of BACE1 inhibitors; and (iii) the binding kinetics of LY2039911 toward these two orthologs of BACE1 to delineate any putative differences between them.

The major advantage of this FRET system, using the aminobenzoate and nitrotyrosine internal quenching pair, is that these two molecules are relatively small compared with other fluorophores and quenchers. In addition, within each APP FRET peptide, there is only one cleavage site for either of these two BACE1 orthologs, as identified by LC/MS, even at high enzyme concentration and with prolonged incubation. Using peptides containing this quenching pair, we have conducted a comprehensive species-specific study for human and murine BACE1 toward their respective APP-derived FRET peptide substrates containing β- and β′-cleavage sites. Results of this study demonstrate that there is little species specificity for the enzyme and substrate at the β′-cleavage site. Results from human BACE1 transgenic mice studies showed that C99 (the β-cleavage product) is the major proteolytic product of APP generated in the brain, as compared with C89 (the β′-cleavage site) (Bodendorf et al. 2002). The same study also showed that human BACE1 is able to process the murine APP at the β′-cleavage site to generate murine C89 fragments. These findings from the human BACE1 transgenic mice (Bodendorf et al. 2002) are consistent with results from this study, using in vitro biochemical tools. The in vitro biochemical evidence presented here provides a molecular hypothesis to explain the apparent species specificity noted in the BACE1 knockout mice study (see below for details) (Cai et al. 2001). Collectively, the in vitro biochemical evidence in this study and the results from human BACE1 transgenic mice suggest that there is no significant species specificity difference of BACE1 toward APP substrate at the β′-cleavage site.

Furthermore, results obtained with these FRET peptide substrates showed that the enzymatic efficiency of BACE1 towards the canonical β-cleavage site is 10 to 50 times higher than that towards the β′-cleavage site. Importantly, the enzymatic efficiency of the murine BACE1 toward the FRET peptide substrate containing the β-cleavage site of human APP is approximately 50% higher than that of human BACE1 toward the same FRET peptide substrate. These results are well supported by the abundant Aβ(1–40) or Aβ(1–42) peptides generated in transgenic mice overexpressing human APP. Consistent with our findings, the amount of Aβ11–40 and Aβ11–42 peptides was significantly lower than that of Aβ1–40 and Aβ1–42 peptides in naïve HEK293 cells (Liu et al. 2002). In contrast, in the HEK293 cells overexpressing BACE1, Aβ11–40 and Aβ11–42 peptides were found to be the major Aβ variants (Liu et al. 2002). Nevertheless, Aβ peptides derived from the β-cleavage site, Aβ1–40 and Aβ1–42, are the major Aβ variants found in brain tissues of AD patients (sporadic and familial AD), human APP transgenic mice and human BACE1 transgenic mice (Naslund et al. 1994; Tamaoka et al. 1994; Bodendorf et al. 2002; Pype et al. 2003). It is unclear whether the β′-cleavage occurs in a different subcellular compartment than that of the β-cleavage. Overexpression of BACE1 in such a compartment may bias the APP cleavage by BACE1 toward the β′-site. Further work is required to elucidate the regulation of the β′-cleavage in a whole cell system.

The recombinant human BACE1 was able to hydrolyze the murine APP at the β′-cleavage site, as demonstrated in this study. This is consistent with the findings from human BACE1 transgenic mice (Bodendorf et al. 2002). Notably, the enzymatic efficiency of recombinant human or murine BACE1 toward the human APP at the β′-cleavage site is five times lower than that toward the murine APP at the β′-cleavage site. This lower efficiency in processing of human APP at the β′-cleavage site may provide an explanation for the lack of β′-cleavage product of human APP by murine BACE1, as reported by Cai et al. in their in vitro studies using neurons isolated from mice deficient in the BACE1 gene (Cai et al. 2001). The activity profiles of recombinant human or murine BACE1 toward the β′-cleavage site of human and murine APP shown in this study are consistent with the subsite preference determined by Turner et al. (2002). P1-tyrosine (found in human β′-cleavage site) is five times less favorable for human BACE1 than P1-phenylalanine (found in murine β′-cleavage site). P3′-histidine (found in human β′-cleavage site) is four times less preferred for human BACE1 than P3′-arginine (found in murine β′-cleavage site).

The enzymatic efficiency of human BACE1 toward human β-site FRET peptides containing the Swedish mutation is near 100-fold higher than that of the wild-type APP sequence. This relationship is consistent with the subsite preference of the human BACE1 reported earlier (Turner et al. 2002). These in vitro results are also consistent with the observations on the earlier onset of disease found with patients carrying Swedish mutations (Citron et al. 1994; Haass et al. 1995).

Lack of species-selective recognition of BACE1 substrates by either murine or human enzyme is extended to the interaction with inhibitors. For the eight peptidic BACE1 inhibitors with in vitro potency spanning three orders of magnitude, there was no significant difference in the inhibitory potency of those molecules against either human or murine BACE1. Brain homogenates from PDAPP mice, CD-1 mice and guinea pigs were used as the source of native BACE1 to determine the inhibitor potency of LY2039911, a close analog of OM99-2. LY2039911 differs from OM99-2 (Hong et al. 2000) by one residue at the amino terminus of the molecule, aspartate instead of glutamate in OM99-2. The IC50 values for LY2039911 in the MBP-C125swe assay displayed no significant difference against native BACE1 among these three animal models. Importantly, the IC50 value for LY2039911 against huBACE1:Fc in the FRET assay is similar to those against native BACE1 in the MBP-C125swe assay, containing polypeptide substrate derived from human APP. Consequently, biochemical properties of BACE1 suggest that both mice and guinea pigs could serve as suitable pre-clinical models for the testing of BACE1 inhibitor in vivo.

Marcinkeviciene et al. suggested that inhibitor binding to BACE1 is a two-step process, using StatVal (Marcinkeviciene et al. 2001). The first step is a very fast binding event in comparison with the second binding step, which involves a conformational change. Consequently, due to the large difference in these two binding rates, only the rate for the slower binding step is observed in the kinetic experiments. Similarly, we could only measure the on-rate of the second slow binding step of LY2039911 to BACE1. There was no significant difference in the inhibitor kinetics of either human or murine BACE1. Notably, the kon (5.7 × 104/m/s) value for LY2039911 (Ki = 2.5 nm) binding to huBACE1:Fc is similar to that [3.5 × 104/m/s (experimental value) and 5.0 × 104/m/s (experimental value)] for StatVal binding to recombinant human BACE1 devoid of transmembrane domain (Marcinkeviciene et al. 2001). The koff value for LY2039911 binding to huBACE1:Fc huBACE1:Fc is nearly 10-fold slower than that for StatVal (7.8 × 10−4/s) to human BACE1. This 10-fold difference in the koff value accounts fully for the 10-fold difference in Ki value measured for StatVal (Ki = 22 nm) binding to human BACE1.

In summary, we observed little difference in the enzymatic efficiencies toward human and murine β′-cleavage sites by human and murine BACE1 orthologs. The pharmacological profiles and inhibitor kinetics between human and murine BACE1 orthologs are indistinguishable. Thus, our in vitro biochemical evidence favors the conclusion of Bodendorf et al. (2002). These results suggest that murine BACE1 in APP transgenic mouse models should exhibit similar pharmacological and enzymatic profiles to those of human BACE1 and should thus be useful in the development of BACE inhibitors for the treatment of Alzheimer's disease.

References

  1. Top of page
  2. Abstract
  3. Experimental procedures
  4. Materials
  5. Methods
  6. Expression of human and murine BACE1
  7. Purification of huBACE1:Fc and muBACE1:Fc
  8. Active site titration of human and murine BACE1:Fc
  9. Determination of enzymatic efficiency of huBACE1:Fc and muBACE1:Fc toward APP FRET peptides
  10. Assessment of the inner filter effect of FRET peptides
  11. Determination of cleavage site of β- and β′-site FRET peptides derived from human and murine APP by human and murine BACE1
  12. Determination of EC50 values for BACE1 inhibitors in the mcaFRET assay
  13. Determination of the inhibitory potency of LY2039911 toward endogenous native BACE1 from mouse and guinea pig brains
  14. Determination of kon and koff of LY2039911 toward huBACE1:Fc and muBACE1:Fc
  15. Results
  16. Sequence and structural comparison
  17. Cleavage site specificity
  18. Pharmacological characterization of human and murine BACE1
  19. Potency for LY2039911 against endogenous and recombinant BACE1
  20. Binding rate of LY2039911
  21. Discussion
  22. Acknowledgements
  23. References
  • Bodendorf U., Danner S., Fischer F., Stefani M., Sturchler-Pierrat C., Wiederhold K., Staufenbiel M. and Paganetti P. (2002) Expression of human beta-secretase in the mouse brain increases the steady-state level of beta-amyloid. J. Neurochem. 80, 799806.
  • Cai H., Wang Y., Mccarthy D., Wen H., Borchelt D., Price D. and Wong P. (2001) BACE1 is the major beta-secretase for generation of Abeta peptides by neurons. Nat. Neurosci. 4, 233234.
  • Citron M., Vigo-Pelfrey C., Teplow D., Miller C., Schenk D., Johnston J., Winblad B., Venizelos N., Lannfelt L. and Selkoe D. (1994) Excessive production of amyloid beta-protein by peripheral cells of symptomatic and presymptomatic patients carrying the Swedish familial Alzheimer disease mutation. Proc. Natl Acad. Sci. USA 91, 11 99311 997.
  • Haass C., Lemere C., Capell A., Citron M., Seubert P., Schenk D., Lannfelt L. and Selkoe D. (1995) The Swedish mutation causes early-onset Alzheimer's disease by beta-secretase cleavage within the secretory pathway. Nat. Med. 1, 12911296.
  • Hong H., Hwang E., Sim H., Cho H., Boo J., Oh S., Kim S. and Mook-Jung I. (2003) Interferon gamma stimulates beta-secretase expression and sAPPbeta production in astrocytes. Biochem. Biophys. Res. Commun. 307, 922927.
  • Hong L., Koelsch G., Lin X., Wu S., Terzyan S., Ghosh A., Zhang X. and Tang J. (2000) Structure of the protease domain of memapsin 2 (beta-secretase) complexed with inhibitor. Science 290, 150153.
  • Hussain I., Powell D., Howlett D. et al. (1999) Identification of a novel aspartic protease (Asp 2) as beta-secretase. Mol. Cell. Neurosci. 14, 419427.
  • Hussain I., Powell D., Howlett D. et al. (2000) ASP1 (BACE2) cleaves the amyloid precursor protein at the beta-secretase site. Mol. Cell. Neurosci. 16, 609619.
  • Lin X., Koelsch G., Wu S., Downs D., Dashti A. and Tang J. (2000) Human aspartic protease memapsin 2 cleaves the beta-secretase site of beta-amyloid precursor protein. Proc. Natl Acad. Sci. USA 97, 14561460.
  • Liu K., Doms R. and Lee V. (2002) Glu11 site cleavage and N-terminally truncated A beta production upon BACE overexpression. Biochemistry 41, 31283136.
  • Luo Y., Bolon B., Kahn S. et al. (2001) Mice deficient in BACE1, the Alzheimer's beta-secretase, have normal phenotype and abolished beta-amyloid generation. Nat. Neurosci. 4, 231232.
  • Marcinkeviciene J., Luo Y., Graciani N., Combs A. and Copeland R. (2001) Mechanism of inhibition of beta-site amyloid precursor protein-cleaving enzyme (BACE) by a statine-based peptide. J. Biol. Chem. 276, 23 79023 794.
  • Naslund J., Schierhorn A., Hellman U. et al. (1994) Relative abundance of Alzheimer A beta amyloid peptide variants in Alzheimer disease and normal aging. Proc. Natl Acad. Sci. USA 91, 83788382.
  • Pastorino L., Ikin A., Nairn A., Pursnani A. and Buxbaum J. (2002) The carboxyl-terminus of BACE contains a sorting signal that regulates BACE trafficking but not the formation of total A (beta). Mol. Cell. Neurosci. 19, 175185.
  • Pype S., Moechars D., Dillen L. and Mercken M. (2003) Characterization of amyloid beta peptides from brain extracts of transgenic mice overexpressing the London mutant of human amyloid precursor protein. J. Neurochem. 84, 602609.
  • Roberds S., Anderson J., Basi G. et al. (2001) BACE knockout mice are healthy despite lacking the primary beta-secretase activity in brain: implications for Alzheimer's disease therapeutics. Hum. Mol. Genet. 10, 13171324.
  • Rossor M., Newman S., Frackowiak R., Lantos P. and Kennedy A. (1993) Alzheimer's disease families with amyloid precursor protein mutations. Ann. NY Acad. Sci. 695, 198202.
  • Selkoe D. (1996) Amyloid beta-protein and the genetics of Alzheimer's disease. J. Biol. Chem. 271, 18 29518 298.
  • Sinha S., Anderson J., Barbour R. et al. (1999) Purification and cloning of amyloid precursor protein beta-secretase from human brain. Nature 402, 537540.
  • Tamaoka A., Odaka A., Ishibashi Y. et al. (1994) APP717 missense mutation affects the ratio of amyloid beta protein species (A beta 1–42/43 and a beta 1–40) in familial Alzheimer's disease brain. J. Biol. Chem. 269, 32 72132 724.
  • Turner R. R., Loy J., Nguyen C., Devasamudram T., Ghosh A., Koelsch G. and Tang J. (2002) Specificity of memapsin 1 and its implications on the design of memapsin 2 (beta-secretase) inhibitor selectivity. Biochemistry 41, 87428746.
  • Vassar R., Bennett B., Babu-Khan S. et al. (1999) Beta-secretase cleavage of Alzheimer's amyloid precursor protein by the transmembrane aspartic protease BACE. Science 286, 735741.
  • Walter J., Fluhrer R., Hartung B., Willem M., Kaether C., Capell A., Lammich S., Multhaup G. and Haass C. (2001) Phosphorylation regulates intracellular trafficking of beta-secretase. J. Biol. Chem. 2761, 14 63414 634.
  • Wise A., Foord S., Fraser N. et al. (2003) Molecular identification of high and low affinity receptors for nicotinic acid. J. Biol. Chem. 278, 98699874.
  • Yan R., Bienkowski M., Shuck M. et al. (1999) Membrane-anchored aspartyl protease with Alzheimer's disease beta-secretase activity. Nature 402, 533537.
  • Yan R., Han P., Miao H., Greengard P. and Xu H. (2001) The transmembrane domain of the Alzheimer's beta-secretase (BACE1) determines its late Golgi localization and access to beta-amyloid precursor protein (APP) substrate. J. Biol. Chem. 276, 36 78836 796.