Functional gamma-secretase inhibitors reduce beta-amyloid peptide levels in brain.

Converging lines of evidence implicate the beta-amyloid peptide (Ass) as causative in Alzheimer's disease. We describe a novel class of compounds that reduce A beta production by functionally inhibiting gamma-secretase, the activity responsible for the carboxy-terminal cleavage required for A beta production. These molecules are active in both 293 HEK cells and neuronal cultures, and exert their effect upon A beta production without affecting protein secretion, most notably in the secreted forms of the amyloid precursor protein (APP). Oral administration of one of these compounds, N-[N-(3,5-difluorophenacetyl)-L-alanyl]-S-phenylglycine t-butyl ester, to mice transgenic for human APP(V717F) reduces brain levels of Ass in a dose-dependent manner within 3 h. These studies represent the first demonstration of a reduction of brain A beta in vivo. Development of such novel functional gamma-secretase inhibitors will enable a clinical examination of the A beta hypothesis that Ass peptide drives the neuropathology observed in Alzheimer's disease.

Fibroblast growth factor-8 protects cultured rat hippocampal neurons from oxidative insult.

Basic fibroblast growth factor (bFGF) has been reported to have neuroprotective properties following excitotoxic, metabolic, and oxidative insults. We report here that another FGF family member, FGF-8 is able to protect rat hippocampal cultures from oxidative stress. The b isoform of FGF-8 protected hippocampal cultures from hydrogen peroxide with an EC50 of approximately 25 ng/ml. In a time course study, using pre-, co-, post-treatment paradigms, we report that bFGF and FGF-8b were neuroprotective when added as a pre-treatment, co-treatment, and even at 2 h post-insult. Using neuronal enriched cultures, we demonstrate that bFGF and FGF-8b neuroprotection partially results from a direct action of the growth factors on neurons. The direct action on neurons may work in concert with normal and FGF-stimulated glial secretion products to give the full FGF protective effect. FGF-8b showed maximal protection at 50 ng/ml, whereas bFGF showed maximal protection at 10 ng/ml. Despite requiring higher concentrations to elicit protection, FGF-8b is able to attain levels of protection equivalent to that of bFGF (attenuation of 75-80% of hydrogen peroxide induced death). We also report that bFGF and FGF-8b are able to protect the human neuroblastoma cell line, SK-N-MC, from peroxide-induced LDH release by 50%. From these studies, we conclude that FGF-8b is another member of the FGF family which may show in vivo efficacy for the treatment of oxidative insults, such as stroke.

Stereoselective neuroprotection by novel 2,3-benzodiazepine non-competitive AMPA antagonist against non-NMDA receptor-mediated excitotoxicity in primary rat hippocampal cultures.

Glutamate excitotoxicity has been implicated in a variety of acute and chronic neurodegenerative diseases but early phase clinical trials with competitive antagonists at both N-methyl-D-aspartate (NMDA)-receptors and alpha-amino-3-hydroxy-5-methyl-isoxazolepropionate (AMPA) receptors have been disappointing. A family of atypical 2,3 benzodiazepines, exemplified by GYKI 52466, have been described recently which function as non-competitive AMPA-receptor antagonists. We have investigated the neuroprotective efficacy of LY303070 and LY300164, two analogs of GYKI-52466, in an embryonic rat hippocampal culture model of non-NMDA receptor-mediated excitotoxicity using kainic acid (KA) as an agonist at the AMPA/KA receptor. Overnight treatment with 500 microM KA resulted in prominent neuronal excitotoxicity as assessed by lactate dehydrogenase efflux. LY300164 and LY303070 attenuated KA-excitotoxicity in a dose-dependent manner with IC50s of 4 and 2 microM, respectively. In contrast, their stereoisomers, LY300165 and LY303071 showed no neuroprotection at concentrations up to 25 microM. In addition, AMPA-mediated excitotoxicity in cyclothiazide pre-treated cultures was also completely blocked by LY303070. Finally, neuroprotection by this class of 2,3 benzodiazepines was not influenced by antagonism of the classical benzodiazepine receptor. LY303070 and LY300164 represent novel non-competitive AMPA-receptor antagonists which may offer unique advantages in the clinic over competitive AMPA-receptor antagonists.

Characterization of 8-epiprostaglandin F2alpha as a marker of amyloid beta-peptide-induced oxidative damage.

The amyloid beta-peptide (A beta) is a major component of the neuritic plaques that are a defining histological characteristic of Alzheimer's disease. A beta can be directly toxic and pro-inflammatory to cells in vitro. Numerous reports have shown that oxidative damage and reactive oxygen species play a role in A beta-mediated neurotoxicity. 8-Epiprostaglandin F2alpha (8-isoprostane) is a well characterized product of lipid peroxidation that is formed nonenzymatically in cell membranes following an oxidative insult. We report a time- and concentration-dependent increase in 8-isoprostane levels in rat hippocampal cultures treated with A beta(1-40) or hydrogen peroxide. As evidence that 8-isoprostane production is part of an A beta toxic pathway, alkaline-treated peptide, which shows minimal toxic activity, resulted in greatly attenuated 8-isoprostane production. Although the increase in 8-isoprostane levels preceded cell death, exogenously added 8-isoprostane had no cytotoxic effects. The antioxidants vitamin E and propyl gallate attenuated A beta-induced 8-isoprostane formation yet had no effect on A beta-induced lactate dehydrogenase release. Neither vitamin E nor propyl gallate had any effect on A beta's ability to adopt a beta-pleated sheet structure and deposit on cells as determined by thioflavine S fluorescence. We conclude that 8-isoprostane is an indicator of A beta-induced damage but not necessarily a mediator of A beta-induced neurotoxicity. Also, 8-isoprostane could be a useful marker for assessing oxidative damage in the CNS.

Characterization of LY231617 protection against hydrogen peroxide toxicity.

The compound LY231617 [2,6-bis(1,1-dimethylethyl)-4-[[(1-ethyl)amino]methyl]phenol hydrochloride] has been reported to afford significant neuroprotection against hydrogen peroxide (H2O2)-induced toxicity in vitro and global ischemia in vivo. We now report on further mechanistic studies of H2O2 toxicity and protection by LY231617. Brief exposure to H2O2 (15 min) elicited an oxidative insult comparable with that generated by overnight treatment. H2O2-mediated cellular degeneration was characterized using lactate dehydrogenase (LDH) release, changes in total glutathione, and a new marker of oxidative stress, 8-epiprostaglandin F2alpha (8-isoprostane). LY231617 attenuated H2O2-mediated degeneration under a variety of exposure conditions, including a more clinically relevant posttreatment paradigm. Levels of 8-isoprostane paralleled LDH release under various treatment paradigms of 100 microM H2O2 +/- 5 microM drug. In contrast, despite affording significant protection, LY231617 had modest to no effects on cellular levels of glutathione. Taken together, these results are consistent with a membrane site of action for LY231617 and suggest that the compound affords cytoprotection via its antioxidant properties.

Induction of sulfated glycoprotein-2 (clusterin) and glial fibrillary acidic protein (GFAP) RNA expression following transient global ischemia is differentially attenuated by LY231617.

Sulfated glycoprotein-2 (SGP-2) is a secreted glycoprotein that along with GFAP has emerged as a prominent molecular marker of neurodegeneration. In the present study, we have evaluated further the relationship between SGP-2, GFAP and neurodegeneration, by examining the effects of LY231617, a potent antioxidant, on expression of SGP-2 and GFAP following four vessel occlusion (4VO). GFAP and SGP-2 RNA levels increased several fold in hippocampus and caudate nucleus in response to 30 min of 4VO. LY231617 treatment markedly attenuated the induction of GFAP RNA in both hippocampus and caudate nucleus, consistent with the significant neuroprotection observed histologically. In contrast, LY231617 treatment blunted SGP-2 RNA expression only in the hippocampus; SGP-2 RNA expression in caudate nucleus was similar to vehicle-treated 4VO, despite the marked attenuation of neuronal damage in both areas by LY231617. These data suggest region-specific differential regulation of SGP-2 and GFAP RNA induction.

Clusterin (Apo J) protects against in vitro amyloid-beta (1-40) neurotoxicity.

Clusterin is a secreted glycoprotein that is markedly induced in many disease states and after tissue injury. In the CNS, clusterin expression is elevated in neuropathological conditions such as Alzheimer's disease (AD), where it is found associated with amyloid-beta (A beta) plaques. Clusterin also coprecipitates with A beta from CSF, suggesting a physiological interaction with A beta. Given this interaction with A beta, the goal of this study was to determine whether clusterin could modulate A beta neurotoxicity. A mammalian recombinant source of human clusterin was obtained by stable transfection of hamster kidney fibroblasts with pADHC-9, a full-length human cDNA clone for clusterin. Recombinant clusterin obtained from this cell line, as well as a commercial source of native clusterin purified from serum, afforded dose-dependent neuroprotection against A beta (1-40) when tested in primary rat mixed hippocampal cultures. Clusterin afforded substoichiometric neuroprotection against several lots of A beta (1-40) but not against H2O2 or kainic acid excitotoxicity. These results suggest that the elevated expression of clusterin found in AD brain may have effects on subsequent amyloid-beta plaque pathology.

Selective protection against AMPA- and kainate-evoked neurotoxicity by (3S,4aR,6R,8aR)-6-[2-(1(2)H-tetrazole-5-yl)ethyl]decahyd roisoquinoline- 3-carboxylic acid (LY293558) and its racemate (LY215490).

Glutamate receptor-mediated excitotoxicity is linked to the activation of multiple receptors including those activated by alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA), N-methyl-D-aspartate (NMDA), and kainate. In this study, the novel glutamate receptor antagonist, as its active isomer (3S,4aR,6R,8aR)-6-[2-(1(2)H-tetrazole-5-yl)ethyl]-decahyd roisoquinoline-3- carboxylic acid ((-)LY293558) and it's +/- racemate (LY215490), was examined for neuroprotectant effects against excitotoxic injury in vitro and in vivo. This agent selectively protected against AMPA and kainate injury in cultured primary rat hippocampal neurons, an in vivo rat striatal neurotoxicity model, and against agonist-evoked seizures in mice. Thus, (3S,4aR,6R,8aR)-6-[2-(1(2)H-tetrazole-5-yl)ethyl]decahydr -oisguino-line-3-carboxylic acid represents a novel receptor selective and potent systemically active AMPA/kainate receptor antagonist for exploring neuroprotection via non-NMDA receptor mechanisms.

Secondary structure of amyloid beta peptide correlates with neurotoxic activity in vitro.

Amyloid beta peptide (A beta), the major protein constituent of senile plaques in patients with Alzheimer's disease, is believed to facilitate the progressive neurodegeneration that occurs in the latter stages of this disease. Early attempts to characterize the structure-activity relationship of A beta toxicity in vitro were compromised by the inability to reproducibly elicit A beta-dependent toxicity across different lots of chemically equivalent peptides. In this study we used CD spectroscopy to demonstrate that A beta secondary structure is an important determinant of A beta toxicity. Solubilized A beta was maximally toxic when the peptide adopted a beta-sheet conformation. Three of the four A beta lots tested had a random coil conformation and were weakly toxic or inactive, whereas the single A beta lot exhibiting toxic activity at low peptide concentrations had significant beta-sheet structure. Incubation of the weakly toxic A beta lots in aqueous stock solutions for several days before use induced a time-dependent conformational transition from random coil to beta-sheet and increased A beta toxicity in three different toxicity assays. Furthermore, the secondary structure of preincubated A beta was dependent upon peptide concentration and pH, so that beta-sheet structures were attenuated when peptide solutions were diluted or buffered at neutral and basic pH. Our data could explain some of the variable toxic activity that has been associated with A beta in the past and provide additional support for the hypothesis that A beta can have a causal role in the molecular neuropathology of Alzheimer's disease.

Neurotoxicity of human amylin in rat primary hippocampal cultures: similarity to Alzheimer's disease amyloid-beta neurotoxicity.

Amylin, a 37-amino-acid amyloidogenic peptide, bears biophysical similarities to the amyloid-beta peptide (A beta) deposited in Alzheimer's disease. Using embryonic rat hippocampal cultures we tested whether amylin induces neurotoxicity similar to that previously observed with A beta(1-40). Treatment with human amylin(1-37) resulted in prominent toxicity as assessed by phase-contrast microscopy and quantification of lactate dehydrogenase in the medium. Amylin-induced neurotoxicity was morphologically similar to that induced by A beta(1-40). In contrast, the nonamyloidogenic rat amylin showed negligible neurotoxicity despite having 95% sequence similarity to human amylin. Only full-length human amylin was toxic; various amylin peptide fragments including amino acid residues 20-29 were nontoxic at similar concentrations. These studies suggest that unrelated amyloidogenic peptides like human amylin and A beta can adopt a similar neurotoxic conformation in vitro. Similar conformation-dependent neurotoxicity may drive the prominent neurite degeneration around compacted but not diffuse deposits of A beta in Alzheimer's disease.

Thyrotropin-releasing hormone gene expression and receptors are differentially modified in limbic foci by seizures.

Previous studies using two seizure paradigms, electroconvulsive shock and kindling, suggested potential sites of endogenous thyrotropin-releasing hormone (TRH) action in specific epileptogenic areas. We studied TRH gene expression and TRH receptors in rat limbic areas using the kindling model of epilepsy. Immunoassayable TRH increased 4- to 20-fold over control levels in specific subregions of the hippocampus 24 hours after a single stage 5 seizure. Concurrently, TRH receptor binding was significantly reduced in hippocampal (23-39%) and amygdaloid (21-22%) membranes. Dramatic temporal and spatial changes in prepro-TRH messenger RNA were visualized by in situ hybridization histochemistry in the hippocampal dentate gyrus, the piriform cortex, and the amygdala. Peak hybridization occurred 6 and 12 hours postictally in these loci and returned toward basal levels by 24 hours. These results are consistent with the hypothesis that TRH may have an important role in the pathophysiology epilepsy by modulating excitatory processes.