The FASEB Journal express article 10.1096/fj.02-0186fje. Published online October 18, 2002.
β-Amyloid inhibits NOS activity by subtracting NADPH availability Giorgio Venturini,* Marco Colasanti,* Tiziana Persichini,* Emanuela Fioravanti,* Paolo Ascenzi,* Letizia Palomba,† Orazio Cantoni,† and Giovanni Musci‡ *Department of Biology, University “Roma Tre”, Rome, Italy; †Institute of Pharmacology and Pharmacognosy, University of Urbino, Urbino, Italy; ‡Department of Microbiological, Genetic and Molecular Sciences, University of Messina, Messina, Italy. Corresponding author: Giovanni Musci, Department of Microbiological, Genetic and Molecular Sciences, University of Messina, Salita Sperone, 31, 98166 Messina, Italy. E-mail: musc[email protected]
ABSTRACT The amyloid peptides Aβ1-42 and Aβ25-35 strongly inhibited the activity of constitutive neuronal and endothelial nitric oxide synthases (i.e., NOS-I and NOS-III, respectively) in cellfree assays. The molecular mechanism of NOS inhibition by Aβ fragments was studied in detail with Aβ25-35. The inhibitory ability was mostly NADPH-dependent and specific for the soluble form of Aβ25-35. Optical, fluorescence, and NMR spectroscopy showed that the soluble, but not aggregated, Aβ25-35 interacted with NADPH, thus suggesting that a direct recruitment of NADPH may result in diminished availability of the redox cofactor for NOS functioning. To assess the physiological relevance of our findings, rat neuronal-like PC12 and glioma C6 cell lines were used as cellular models. After Aβ25-35 internalization into cells was verified, the activity of constitutive NOS was measured using the DAF-2DA detection system and found to be severely impaired upon Aβ25-35 uptake. Consistent with previous results on the molecular cross-talk between NOS isoforms, repression of constitutive NOS by Aβ25-35 resulted in enhanced expression of inducible NOS (NOS-II) mRNA in C6 cells. Our results represent the first evidence that amyloid fragments impair constitutive NOS activity in cell-free and cellular systems, providing a possible molecular mechanism for the onset and/or maintenance of Alzheimer’s disease. Key words: Alzheimer’s disease • nitric oxide • constitutive NOS inhibition
lzheimer’s disease (AD) is the most common form of senile dementia, primarily defined by cognitive deficits. AD is neuropathologically characterized by senile plaques and neurofibrillary tangle deposition, basal forebrain cholinergic deficit and extensive neuronal loss, and synaptic changes in the cortex and hippocampus (1). The presence of neuritic plaques containing β-amyloid (Aβ) constitutes the major pathohistological feature in AD. Aβ is a 39- to 43-amino acid β-sheet peptide derived from proteolytic processing at the N-terminus of the amyloid precursor protein (APP), a 770-residue membrane-associated protein with a single transmembrane domain consisting of 23 residues (2). From a physiological point of view, Aβ1-
40 and Aβ1-42 are the most relevant forms; however the derivative Aβ25-35 has been demonstrated to be the shortest fragment exhibiting large β-sheets fibrils. Also, Aβ25-35 has been postulated to be the biologically active region in that it retains the toxicity of the full-length peptide(s) (3–6). To date, attention has been focused on the mechanisms by which Aβ and/or its fragments exert the neurotoxic effect on cells at the level of the plasma membrane. The assumption of a membrane-mediated neurotoxicity is based on the knowledge that amyloid deposits are typically extracellular. However, intracellular Aβ accumulation has been observed in various cell types, including neurons and endothelial cells (7–13). Moreover, Aβ-induced cytotoxicity has been associated to the binding of the peptide with the intracellular endoplasmic reticulum amyloid βpeptide-binding protein (ERAB) (14), i.e., the brain short chain L-3-hydroxyacyl-CoA dehydrogenase (15). Therefore, a central question about the role of Aβ in AD is whether the disease initiates with the extracellular Aβ deposition or the intracellular Aβ accumulation. Although an intracellular Aβ accumulation may reflect uptake of secreted Aβ, the intraneuronal Aβ appears to precede its secretion, neurofibrillary tangle and Aβ plaque deposition, thus suggesting that intracellular buildup of Aβ generated within the cell may actually represent an early event in AD (9, 16, 17). Over the past few years, a relationship between AD and nitric oxide (NO) has been widely reported, remarking on the neurotoxic effect of NO as induced by extracellular Aβ species (for a recent review, see ref 18). In particular, Aβ has been found to stimulate inducible NO synthase (NOS-II) expression and NO production through a tumour necrosis factor α (TNFα) receptorassociated factor (TRAF6 and 2) and NF-κB-inducing kinase (NIK)-dependent signaling mechanism (19, 20). Although this mechanism might be involved in neurodegeneration, it is important to remember that NO produced by constitutive NOS (i.e., NOS-I and/or NOS-III, collectively termed cNOS) regulates critical neuronal/endothelial functions. Indeed, an interesting hypothesis has recently emerged, providing a new link between NO and AD. An underlying Aβ-driven process may very early contribute to promote AD through the lack of an adequate basal NO (i.e., constitutive NO) (21, 22). Consistently, in the early stages of AD, basal levels of NO maintained from cNOS would play a fundamental role in neuro- and vascular protection (21), thus confirming the dual personality of NO (23). Despite this crucial point of the question, a direct effect of intracellular Aβ or its fragments on cNOS activity and on NO pathway has never been demonstrated. For the first time, we show here that Aβ1-42 and Aβ25-35 are able to inhibit the activity of cNOS isoforms by lowering NADPH availability and that this inhibition can take place also in vivo under proper conditions, leading to impaired production of constitutive NO. MATERIALS AND METHODS Aβ1-42, Aβ25-35, Escherichia coli lypopolysaccharide (LPS), lactic dehydrogenase (LDH), nitro-L-arginine methyl ester (L-NAME), thapsigargin (Tg), the calcium ionophore A23187, sodium nitroprusside (SNP), and cytochrome c were purchased from Sigma-Aldrich (Milan, Italy). Arginine-L-[2,3-3H] (2,0 TBq/mmol) was purchased from NEN (Boston, CA). 4,5Diaminofluorescein diacetate (DAF-2DA) was from Vinci-Biochem/Alexis Italia (Vinci, Italy).
Aβ25-35, N-α-biotinyl hexanoate was from BioSource International (Camarillo, CA). Stock solution of the soluble amyloid peptides was prepared at 8 mM concentration in DMSO according to Boland et al. (24) and kept frozen at –20°C. Stock solution of the aggregated form of Aβ25-35 was prepared according to the manufacturer’s instructions by leaving a 4-mM Aβ2535 solution in water at room temperature for 2 h. Thawing and dilution to the final concentration in the proper medium was performed immediately before use. Tg and A23187 were dissolved in 95% (v/v) ethanol and in DMSO, respectively. NOS-I was prepared from rat brain homogenates, as already reported (25, 26). NOS-II was obtained from lung homogenate of rats treated with LPS (10 mg/kg i.p., 6 h). All samples for NOS detection were homogenized in 50 mM Hepes at pH 7.5, containing 0.5 mM EGTA, 1 mM DTT, and 0.1 mg/ml PMSF. Homogenates were then desalted by chromatography over disposable PD-10 columns packed with Sephadex G-25 medium (Amersham Pharmacia Biotech Italia, Milan, Italy). Recombinant NOS-III was from Vinci-Biochem/Alexis Italia (Vinci, Italy). Rat liver microsomes for cytochrome c reductase activity were prepared as described elsewhere (27). When not specified, chemicals were of analytical grade and used without further purification. Enzymatic assays NOS activity was assessed in vitro by evaluating the conversion of [3H]L-arginine to [3H]Lcitrulline, in 50 mM Hepes buffer at pH 7.5 and 37.0°C, according to the modification of the method described by Bredt and Snyder (28). For NOS-I and NOS-III activity, an aliquot of enzymes was added to a reaction mixture containing 0.01–1 mM NADPH, 0.4 mM CaCl2, 1 µg/ml calmodulin, 10 µM FAD, 10 µM FMN, 100 µM tetrahydrobiopterin (BH4), [3H]L-arginine (from 12 kBq to 185 kBq) and 50 µM L-arginine. For NOS-II, CaCl2 and calmodulin were omitted. After 15 min of incubation, the reaction was stopped with ice-cold 20 mM Hepes solution, pH 5.5, containing 2 mM EDTA. [3H]L-citrulline was separated from [3H]L-arginine by ion exchange chromatography on Dowex 50WX8 (Fluka Chemie AG). The enzyme activity was linear up to 30 min of incubation, and results were expressed as nmol citrulline×min–1×mg–1. Cytochrome c reductase activity was measured by following the formation of the 550-nm peak in the optical spectrum due to reduced cytochrome c (29). The assays were run in 0.2 phosphate buffer, pH 7.2, containing 10 µM NADPH and 100 µM cytochrome c. LDH activity was assessed in 50 mM Hepes buffer at pH 7.5, monitoring the optical absorbance change at 340 nm due to reduced NAD(P)H. Spectroscopic measurements Absorption and fluorescence spectra were recorded on a Perkin-Elmer Lambda 9 spectrophotometer and on a Perkin-Elmer LS50 spectrofluorimeter, respectively. NMR spectra were acquired with a Varian Gemini 300 spectrometer. Cell cultures and treatment conditions PC12 or C6 glioma cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; HyClone Laboratories, Logan, UT) supplemented with 5% (PC12) or 10% (C6) fetal bovine serum (HyClone), penicillin (50 units/ml), and streptomycin (50 µg/ml) (Sera-Lab, Crawley Down, UK), at 37°C in T-75 tissue culture flasks (Sarstedt S.r.l, Verona, Italy), gassed with an
atmosphere of 95% air/5% CO2. The culture medium used for PC12 cells was also supplemented with 10% horse serum (HyClone). At the treatment stage, the final concentration of ethanol or DMSO never exceeded 0.05% (v/v). β-Amyloid uptake measurements Cells (1.0 × 105) were plated onto glass coverslips or 35-mm tissue culture dishes and grown for 24 h. Cells were treated with increasing concentrations of the biotinylated soluble form of Aβ2535 for 2 h. The media were aspirated, and excess Aβ25-35 was removed by five washings with fresh DMEM. Cells on glass coverslips were fixed in 4% paraformaldehyde for 20 min and rinsed twice in PBS. To visualize the intracellular peptide, cells were permeabilized with 0.1% saponin in PBS for 5 min and incubated at room temperature for 30 min with R.T.U. Vectastain ABC reagent (Vector Laboratories, Burlingame, CA) supplemented with 0.1% saponin. After three washings in PBS without saponin, cells were then incubated for 10 min in 0.08% diaminobenzidine (DAB) and 0.025% H2O2. Coverslips were finally rinsed in PBS and placed onto uncoated slides with a small drop of the mounting medium glycerol/PBS (1:10). Finally, the slides were viewed with an Axioplan II microscope (Carl Zeiss, Jena, Germany) equipped with a cooled CCD camera SPOT (Diagnostic Instruments, Sterling Heights, MI). The shutter aperture and the exposure/integration settings were kept constant to allow quantitative comparisons of relative staining intensity of cells between treatment groups. Images were digitally acquired and processed for densitometric determination at the single-cell level, using the public domain NIH Image 1.61 program (developed at the U.S. National Institutes of Health and available on the Internet at http://rsb.info.nih.gov/nih-image/). To make sure that the observed staining was due to intracellularly internalized Aβ25-35 and not to any surface-bound peptide, the cells grown in culture dishes were treated for 10 min with PBS containing trypsin. Detached cells were collected and spinned onto glass coverslips, using a CytoSpin3 centrifuge (Shandon, Pittsburgh, PA). Cell fixing and peptide detection was then performed as previously described. The staining did not appreciably differ from the untrypsinized cells grown directly on glass coverslips. Nitric oxide detection system The production of NO in PC12 or C6 glioma cells was measured using the DAF-2DA detection system, as described previously (30). In brief, 1.5×105 cells were inoculated into 35-mm tissue culture dishes and grown for 18–24 h. Cells were loaded with 10 µM DAF-2DA for 15 min at 37°C in 8.182 g/l NaCl, 0.372 g/l KCl, 0.336 g/l NaHCO3, and 0.9 g/l glucose. After accurate washings, cells were treated with soluble Aβ25-35 as detailed in Results, and cellular fluorescence was imaged using a confocal laser microscope (Bio Rad DVC 250, Bio-Rad, Richmond, CA) equipped with a Hamamatsu 5985 (Hamamatsu Italy, Milan, Italy) chilled CCD camera. Cells were illuminated with the 488-nm line of the argon laser, and the fluorescence emitted was monitored at λ>515 nm. The laser intensity, the shutter aperture, and the exposure/integration settings were kept constant to allow quantitative comparisons of relative fluorescence intensity of cells between treatment groups. Images were digitally acquired and processed for fluorescence determination at the single-cell level on a Macintosh 6100/66 computer, using the NIH Image 1.61 program.
Analysis of NOS-II mRNA expression Reverse transcriptase-polymerase chain reaction (RT-PCR) was carried out to analyze the effect of soluble Aβ25-35 on NOS-II mRNA expression in rat C6 glioma cells, as suboptimally induced by LPS (1 µg/ml) plus IFN-γ (100 U/ml), for 4 h. In brief, total cellular RNA was purified from 1×106 rat C6 glioma cells and reverse-transcribed into cDNA, as previously reported (31). cDNA was amplified for the NOS-II gene, using rat NOS-II-specific primers, as described previously (31). The mRNA for the constitutive GAPDH enzyme was examined as the reference cellular transcript and was present at equivalent levels in all cell lysates. Estimates of the relative NOS-II mRNA amounts were obtained by dividing the area of the NOS-II band by the area of the GAPDH band (Bio-Rad Multi-Analyst/PC Version 1.1). Molecular weight (MW) was the 100-bp DNA ladder (Life Technologies). RESULTS Aβ1-42 and Aβ25-35 inhibit the activity of NOS To verify the hypothesis that amyloid Aβ may affect NOS, we first analyzed the effect of either Aβ1-42 or Aβ25-35 on the catalytic activity of neuronal NOS-I and endothelial NOS-III by measuring the conversion of [3H]arginine to [3H]citrulline in vitro. As indicated in Table 1, the activity of both NOS-I and NOS-III was significantly inhibited by Aβ1-42 or Aβ25-35. The inhibitory effect of the peptides could be relieved by increasing the concentration of NADPH in the assay mixture, suggesting an action for Aβ in either directly recruiting NADPH or interacting with the NADPH-binding site on NOS. However, addition of excess arginine, FAD, FMN, or Ca2+-calmodulin could not reverse the inhibition by Aβ. The inhibition of NOS by Aβ was studied in greater detail with the shorter Aβ25-35 peptide. The inhibitory action was restricted to the soluble form of the peptide, whereas the aggregated form of Aβ25-35, at the same concentration, was totally uneffective (Fig. 1A). The inverted Aβ35-25 peptide was also uneffective (Fig. 1A), proving that the inhibition of NOS activity was specifically elicited by the Aβ25-35 aminoacid sequence derived from the amyloid precursor protein. The compensatory effect of NADPH was also better specified for NOS-I (Fig. 1B) and NOS-III (Fig. 1C). The inhibition of NOS by Aβ25-35 was progressively relieved by increasing concentrations of NADPH. As expected, the effect of Aβ25-35 was not limited to cNOSs, inducible NOS-II also being inhibited in a similar pattern of NADPH dependence (data not shown). A further enzymatic assay helped to define the simple mechanism of NADPH recruitment by Aβ. This was achieved by testing another NADPH-dependent enzyme, namely microsomal cytochrome c reductase. In line with the results obtained on NOS, Aβ25-35 significantly inhibited the enzymatic reduction of cytochrome c by microsomes, the inhibition being dependent on NADPH concentration. In particular the peptide, at 100 µM concentration, caused ∼60% and 20% inhibition of the cytochrome c reduction rate in the presence of 50 µM and 200 µM NADPH, respectively (data not shown).
Finally, we tested the effect of Aβ25-35 on lactic dehydrogenase (LDH), an enzyme that normally utilizes NADH as a cofactor, but is also able to use NADPH as an alternative electron source. Interestingly, LDH activity was inhibited by Aβ25-35 when either NADPH or NADH was used as a cofactor (data not shown). This suggests that Aβ25-35 is able to interact with both dinucletotides, although the result with NADH should be taken with caution, because some precipitation of NADH occurred in the presence of the peptide. Spectroscopic evidence of the Aβ-NADPH interaction Absorption and fluorescence spectroscopies were used to find evidence of a molecular interaction of NADPH with either Aβ1-42 or Aβ25-35. The optical spectrum of NADPH was greatly affected by the soluble amyloids, as representatively shown for Aβ25-35 at a peptide:NADPH ratio 1:2.5 (Fig. 2A, curves 1 and 2). The kinetics of the optical change was fast, and the intensity enhancement was complete within the measurement time. The optical spectrum of NADPH was not affected by either the inverted Aβ35-25 (curve 3) or the aggregated Aβ25-35 peptides (curve 4). The fluorescence spectrum of NADPH in the presence and absence of Aβ1-42 or Aβ25-35 (peptide:NADPH ratio 1:2.5) was recorded. As already observed in the case of the optical spectrum, Aβ significantly affected the spectral lineshape. Again, we representatively report the case of Aβ25-35, which induced an enhancement of the emission intensity (Fig. 2B, curves 1 and 2). The inverted (curve 3) and the aggregated (curve 4) peptides were again unable to induce any change. Finally, to further prove that Aβ25-35 interacts with NADPH, 1H-NMR spectroscopy was used. Significant changes of the proton NMR spectrum lineshape of NADPH were observed in the presence of Aβ25-35 (Fig. 3). The signal at –0.199 ppm (relative to dioxane) shifted to –0.123 ppm and to –0.108 ppm after addition of 0.5 and 0.75 equivalents of Aβ25-35, respectively. Beside the shift of the –0.199 ppm peak, increasing concentrations of Aβ25-35 induced the progressive broadening of other multiplets, in particular of that centered at ∼ 1.36 ppm. These changes are in line with a concentration-dependent modification of the nucleotide electronic structure by the peptide. Internalization of Aβ25-35 into PC12 and C6 cells The physiopathological relevance of these findings was assessed by testing whether Aβ25-35 impairs the enzymatic activity of cNOS directly in the cells. First, we sought to verify whether Aβ25-35, which is known to enter different cell types (32), did so under our experimental conditions. For this purpose, neuronal-like PC12 and glial-derived C6 cell lines were chosen as models of possible cellular targets of Aβ peptides in the central nervous system. The accumulation of Aβ25-35 in the intracellular compartment was assessed as detailed in Materials and Methods, and care was taken to rule out a significant contribution of surface-bound peptide in our measurements. Figure 4 shows the densitometric analysis of C6 cells incubated for 2 h with increasing concentrations of biotinylated Aβ25-35 and enzymatically stained with an avidin-peroxidase complex. In these experiments, the cells were treated with trypsin to remove
the membrane-bound peptide, but essentially the same results were obtained in untrypsinized cells. Clearly, Aβ25-35 dose-dependently accumulated in the cells, the staining intensity being significantly increased (P