Inhibition of Alzheimer's disease β-amyloid aggregation, neurotoxicity, and in vivo deposition by nitrophenols: implications for Alzheimer's therapy.

SPECIFIC AIMS We have examined the hypothesis that small molecular weight compounds capable of inhibiting the aggregation of the beta-amyloid peptide (Aβ) could protect neurons from the toxic effect of Aβ and prevent cerebral amyloid deposition. The anti-aggregating and neuroprotective effects of nitrophenols were investigated in vitro using primary cultures of rat hippocampal neurons, and the capacity to inhibit amyloid formation was evaluated in vivo in rats using a model system of cerebral amyloid deposition. PRINCIPAL FINDINGS 1. Aβ fibrils are stabilized by hydrophobic interactions mediated by the carboxyl-terminal domain of the peptide Amyloid fibrils were formed in vitro by dissolving either full-length Aβ (Aβ1-42 and Aβ1-43) or the carboxyl-terminally truncated Aβ1-28 peptide in aqueous buffer, and their stabilities in denaturant solutions were investigated. We found that the stability of fibrillar Aβ in guanidine hydrochloride (GdnHCl) solutions was markedly dependent on peptide chain length. For Aβ1-28, complete disaggregation (as indicated by light scattering measurements) was observed at 3 M GdnHCl, whereas full disaggregation of Aβ1-42 required 5-6 M GdnHCl. All polar and charged amino acid residues of Aβ are located in the 28 amino acid-long amino-terminal portion of the peptide. Residues 29-42 comprise a cluster of nonpolar amino acids contained in a transmembrane sequence of the amyloid precursor protein. Thus, the higher stabilities of Aβ1-42 and Aβ1-43 relative to Aβ1-28 suggest that the carboxyl-terminal nonpolar amino acid sequence in the former two peptides mediates hydrophobic interactions that are important for the stability of the fibrils. Hydrophobic interactions are known to be destabilized by low temperatures, due to the decrease in the entropic contribution to the hydrophobic effect. In line with this, we found that decreasing temperature from 25°C to 1°C caused reversible and nearly complete disaggregation of Aβ. These results indicate that a significant contribution to the stability of Aβ aggregates comes from entropy-driven hydrophobic interactions, leading to the hypothesis that hydrophobic compounds-such as nitrophenols-could be effective in destabilizing and disaggregating amyloid fibrils. 2. Nitrophenols inhibit the aggregation and disaggregate fibrillar Aβ Addition of micromolar concentrations of 2,4-dinitrophenol (DNP) or 3-nitrophenol (NP) caused marked disaggregation of previously formed Aβ fibrils, as revealed by light scattering measurements. IC50 values of approximately 7 μM and 80 μM were found for DNP and NP, respectively. DNP (20 μM) caused complete disaggregation of amyloid fibrils. Furthermore, direct demonstration that DNP and NP inhibited the aggregation of fibrillar amyloid was obtained by transmission electron microscopy. Abundant fibrils were observed in control samples of Aβ1-42, whereas samples in which Aβ was added to the medium in the presence of NP or DNP were completely devoid of fibrils and contained only occasional scattered amorphous aggregates. 3. Nitrophenols block the neurotoxicity of Aβ To investigate the possible neuroprotective effects of nitrophenols against Aβ-induced neurotoxicity, 48 h primary cultures of E18 rat hippocampal neurons were used. Aβ1-42 (44 μM) was added to the medium and incubation was continued for 72 h. Whereas control hippocampal neurons exhibited large cell bodies and long, branched neurites (Fig. 1a ⤻ ), significant neuronal degeneration and death was observed after 72 h of culture in the presence of Aβ1-42 (Fig. 1b ⤻ ). Large numbers of Aβ-treated neurons became detached from the plate during the immunostaining washes (Fig. 1b ⤻ ), suggesting that neuronal adhesion was impaired. Furthermore, the remaining cell bodies of Aβ-treated neurons were attached to the plate but their neurites were retracted and thin, and sometimes detached from the plate. When incubation with Aβ was carried out in the presence of NP or DNP, a marked protection against neurotoxicity was observed (Fig. 1c ⤻ , d ⤻ , respectively). In the presence of nitrophenols, neurons treated with Aβ showed large cell bodies and long neurites with good adhesion properties, and the morphological aspect of the cultures was similar to a control 5 day hippocampal culture. [Figure: see text] The results obtained in cell culture experiments were quantitated by measuring neuronal survival by trypan blue exclusion. Incubation with Aβ1-42 caused marked cell death, with only 25% neuronal survival vs. 86% survival in control cultures. Remarkably, addition of nitrophenols to the incubation medium almost completely blocked Aβ-induced cell death (74% and 65% neuronal survival in the presence of Aβ plus DNP or NP, respectively). 4. In vivo inhibition of amyloid deposition To evaluate the effects of nitrophenols as inhibitors of amyloid deposition in vivo, we have used a rat model of intracerebral Aβ deposition. Aβ was injected alone or in the presence of DNP into the left or right hippocampi of rats, respectively, and the areas occupied by amyloid deposits on either side were measured on consecutive sections stained with thioflavin S. The experimental protocol used was designed to minimize the influence of individual variability in animal response by injecting Aβ into one hemisphere and an identical amount of Aβ plus DNP into the other hemisphere for each individual rat. Figure 2a ⤻ shows a schematic diagram of the microinjection protocol, along with representative hippocampal sections stained with thioflavin S or directly visualized using the autofluorescence of amyloid, as indicated. The total volume occupied by amyloid deposits was integrated by image analysis over all consecutive sections spanning the site of deposition. Figure 2b ⤻ , c ⤻ , d ⤻ , e ⤻ shows results of the integration for four different rats. Coinjection of Aβ and DNP strongly prevented amyloid deposition, with a reduction of 86 ± 17% in the volume of amyloid deposits in rat brains relative to the volume occupied when Aβ was injected alone (P=0.05, n=4; one-tailed paired t test). [Figure: see text] CONCLUSIONS Despite considerable efforts aiming at understanding the molecular basis and physiopathology of Alzheimer's disease, there are currently no clinically accepted treatments to cure or stop the progression of this devastating disease. The β-amyloid peptide plays a central role in the neuropathology of AD, and considerable evidence indicates that fibrillar aggregation of Aβ and amyloid deposition are related to neurotoxicity. Our study shows that micromolar concentrations of nitrophenols inhibit the aggregation of Aβ in vitro and cause disaggregation of previously formed amyloid fibrils (Fig. 3 ⤻ ). Of greater interest, nitrophenols protect rat hippocampal neurons in culture from the neurotoxic effect of Aβ (Fig. 3) ⤻ . Furthermore, nitrophenols inhibit the formation of amyloid deposits in rat hippocampi in an in vivo model system of cerebral amyloid deposition. [Figure: see text] Nitrophenols are small water-soluble compounds, yet presumably hydrophobic enough to cross the blood-brain barrier and gain access to the central nervous system. DNP and other nitrophenols are known to be toxic at high concentrations. It is important to note, however, that the nitrophenols by themselves had no detectable toxic effects to neuronal primary cultures at the concentrations used in our studies and that the cytoarchitecture of rat brains (as revealed by staining with cresyl violet) also appeared normal in hemispheres injected with DNP. Finally, because there is no effective treatment currently available for amyloidoses, including Alzheimer's disease, type II diabetes, and prion-related spongiform encephalopathies, a possible clinical utilization of nitrophenols at low, subtoxic concentrations should be considered. Another possibility lies in the use of nitrophenols as lead compounds for the development of small molecule inhibitors of amyloidogenesis active at lower concentrations or with fewer undesirable side effects. In conclusion, we propose that nitrophenols (and/or their derivatives) should be explored as potential therapeutic agents (or lead compounds for further drug development) to prevent the aggregation and neurotoxicity of Aβ in Alzheimer's disease. 1 To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.00-0676fje ; to cite this article, use FASEB J. (March 20, 2001) 10.1096/fj.00-0676fje.