Caspase-3 activation by beta-amyloid and prion protein peptides is independent from their neurotoxic effect.

Synthetic peptides corresponding to residues 25-35 of beta-amyloid (beta 25-35) and 106-126 of prion protein (PrP 106-126) are amyloidogenic and cause neuronal death by apoptosis in vitro. We evaluated, in rat cortical neurons, the role of caspases activation in the peptides neurotoxicity by measuring of caspase-3 (CPP32) activity and applying a non-selective caspase inhibitor (z-VAD-fmk) or CPP32-specific inhibitor (Asp-Glu-Val-Asp-CHO (DEVD-CHO)). CPP32 was dose-dependently activated by both peptides (2.5-50 microM). The caspase inhibitors completely abolished the CPP32 activation induced by the peptides. However, the neurotoxic effect was partially attenuated with z-VAD-fmk, while no antagonism was found with DEVD-CHO. Thus, although beta 25-35 and PrP 106-126 robustly activated CPP32, their neurotoxic effect was independent of this caspase activation.

Tetracycline affects abnormal properties of synthetic PrP peptides and PrP(Sc) in vitro.

Prion diseases are characterized by the accumulation of altered forms of the prion protein (termed PrP(Sc)) in the brain. Unlike the normal protein, PrP(Sc) isoforms have a high content of beta-sheet secondary structure, are protease-resistant, and form insoluble aggregates and amyloid fibrils. Evidence indicates that they are responsible for neuropathological changes (i.e. nerve cell degeneration and glial cell activation) and transmissibility of the disease process. Here, we show that the antibiotic tetracycline: (i) binds to amyloid fibrils generated by synthetic peptides corresponding to residues 106-126 and 82-146 of human PrP; (ii) hinders assembly of these peptides into amyloid fibrils; (iii) reverts the protease resistance of PrP peptide aggregates and PrP(Sc) extracted from brain tissue of patients with Creutzfeldt-Jakob disease; (iv) prevents neuronal death and astrocyte proliferation induced by PrP peptides in vitro. NMR spectroscopy revealed several through-space interactions between aromatic protons of tetracycline and side-chain protons of Ala(117-119), Val(121-122) and Leu(125) of PrP 106-126. These properties make tetracycline a prototype of compounds with the potential of inactivating the pathogenic forms of PrP.

Influence of mutations associated with familial prion-related encephalopathies on biological activity of prion protein peptides.

In transmissible spongiform encephalopathies (TSEs), an altered form of prion protein (PrP), PrPres, aggregates in amyloid fibrils and accumulates in the brain. Several point mutations of the PrP gene have been associated with the TSEs, so, to investigate how the mutations affect the biological activity of PrP, we analyzed the biological effects and chemicophysical characteristics of the peptide homologous to the wild-type and mutated sequence of PrP fragments. The mutation P102L altered the biological activity of PrP 89-106, which became neurotoxic without changing its fibrillogenic capacity. The mutation (D178N) in the PrP 169-185 strongly increased the neurotoxic activity of the native sequence. In this case, there was also a clear alteration of the structural conformation. None of the other mutations considered, including A117V, seemed to influence the biological activities of the respective peptides. These data identify new neurotoxic fragments of PrP in the mutated form and elucidate their genetic influence on the pathogenesis of TSEs.

Extracellular calcium deprivation in astrocytes: regulation of mRNA expression and apoptosis.

Cell viability and gene expression were studied in primary astroglial cells cultured in a nominally calcium-free medium. Ca2+ deprivation reduced progressively the astrocytes' viability, starting from 12 h; the restoration of a normal Ca2+ concentration (1.8 mM) in the medium after 12-h deprivation reversed the degenerative effect within 24 h. Biochemical and morphological examinations indicated that cell death induced by Ca2+ deprivation was mediated by apoptosis. This was associated with the expression of c-fos, c-jun, and c-myc, which, with different time courses, were induced in astrocytes after Ca2+ deprivation. Furthermore, shifting to a Ca2+-free medium modified the expression of Ich-1S transcript and rapidly increased intracellular cyclic AMP, which has been implicated in the transcriptional activation of immediate-early genes. The absence of Ca2+ in the medium reduced the expression of constitutive proteins such as alpha-actin, clusterin, glial fibrillary acidic protein, amyloid precursor protein, and glucose-6-phosphate dehydrogenase. The expression of these mRNAs was reduced >50% after 8 h of Ca2+ deprivation, when the effect on cell viability was negligible. When Ca2+ deprivation was prolonged for 24 h the expression of mRNA dropped completely, and restoration of the Ca2+ ions in the medium for 48 h did not reverse this effect. In contrast with general assumption, the apoptotic machinery in astrocytes is activated similarly not only by increased Ca2+ influx but also with the extracellular Ca2+ deprivation.

Densitometric quantification of neuronal viability by computerized image analysis.

A new method is presented for the quantification of cell viability based on densitometry with computerized image analysis. Neuronal cells were stained with crystal violet and densitometric analysis was performed with an IBAS 2.0 image analyzer (Kontron/ Zeiss), using specially implemented dedicated software which integrates the optical density of the culture in each well with the area covered by the stained cells. To test the reliability of the densitometric method cortical cells were plated at different concentrations (5 x 10(4)-10(6)/ml); the standard curve obtained by analysis of crystal violet staining showed a linear proportion between cell number and optical density signal. The validation and accuracy of the method were assessed and compared with other methods using rat cortical cells cultured in vitro for 10 days and exposed to kainic acid (250 microM) for 24 h. Neuronal viability was reduced by 40-50% and comparison with direct cell counting, MTT assay, and spectrophotometric analysis confirmed that the method is simple, quick, and reliable.

Beta-amyloid fragment potentiates IL-6 and TNF-alpha secretion by LPS in astrocytes but not in microglia.

The effect of a peptide homologous to the biologically active fragment of beta amyloid 25-35 (beta 25-35) was studied on interleukin 6 (IL-6) and tumour necrosis factor (TNF-alpha) secretion induced by lipopolysaccharide (LPS) in primary rat astrocytes and microglia. Twenty-four hour exposure to LPS (50 ng/ml) induced IL-6 and TNF-alpha both in astrocytes and in microglial cells, while the effect of beta 25-35 (50 microM) per se was negligible in both cell types. In microglial cells, the application of beta peptide did not alter the production of either cytokine induced by LPS. However, beta 25-35 strongly amplified the production of both IL-6 and TNF-alpha in astrocytes. These findings confirm the complex interaction between cytokines and amyloidogenesis in Alzheimer's disease and indicate that astrocytes rather than microglia respond to the beta amyloid fragment, suggesting that these cells may be actively involved in cytokine-mediated events in AD.

Influence of cell culture conditions on the protective effect of antioxidants against beta-amyloid toxicity: studies with lazaroids.

The mechanisms of cell death of rat cortical neurons chronically exposed to the beta-amyloid (betaA) biologically active fragment beta-(25-35) involve oxidative stress. We examined the influence of culture conditions on the neuroprotective activity of antioxidants against beta-(25-35) toxicity. Common radical scavengers such as N-acetylcysteine (250 microM) and N-t-butyl-phenylnitrone (500 microM) only protected cortical cells cultured in the presence of fetal calf serum (FCS) from betaA insult. The neuroprotective effect of lazaroids (U74389G and U83836E), 21-aminosteroids with antioxidant activity, was tested in cells grown with or without FCS. U74389G did not interfere with beta-(25-35) toxicity in either condition, while U83836E at a very low concentration (15 nM) protected cortical cells exposed to the beta peptide only when the neurons were cultured in the presence of FCS. These data show that a lazaroid can prevent beta-(25-35) toxicity and that the antioxidants exerted their protective effect in certain conditions.

Prion protein fragment 106-126 differentially induces heme oxygenase-1 mRNA in cultured neurons and astroglial cells.

Heme oxygenase (HO), which catalyzes the degradation of heme, has two isozymes (HO-1 and HO-2). In brain the noninducible HO-2 isoform is predominant, whereas the inducible HO-1 is a marker of oxidative stress. Because brain oxidative stress might be present in prion-related encephalopathies (PREs), as in other neurodegenerative diseases, we investigated whether HO-1 mRNA was induced in neuronal and astroglial cell cultures by a peptide corresponding to residue 106-126 of human prion protein (PrP). This peptide is amyloidogenic, and when added in vitro to cultured cells it reproduces the neuronal death and astroglial proliferation and hypertrophy occurring in PREs. HO-1 mRNA did not accumulate in rat cultured neurons from hippocampus or cortex exposed to PrP 106-126 (50 microM for 5 days). PrP 106-126 induced HO-1 mRNA accumulation in rat astroglial cultures depending on the exposure time and concentration, being maximal (33-fold) after 7 days of exposure at 50 microM. The nonamyloidogenic amidated or amidated-acetylated PrP 106-126 was ineffective, as was a scrambled peptide used as control. N-Acetylcysteine reduced (50%) the accumulation of HO-1 mRNA in astroglial cells after PrP 106-126 (25 microM) given for 5 days. Thus, oxidative stress is apparently a feature of the toxicity of PrP 106-126, and it might also occur in PREs; induction of HO-1 could contribute to the greater resistance of astrocytes compared with neurons to PrP 106-126 toxicity.

Neuroprotective effect of somatostatin on nonapoptotic NMDA-induced neuronal death: role of cyclic GMP.

Somatostatin (SRIF) exerts a modulatory function on neuronal transmission in the CNS. It has been proposed that a reduction of calcium currents is the major determinant of the inhibitory activity of this peptide on synaptic transmission. Because the neurotoxicity induced by activation of the NMDA subtype of glutamate receptor is mediated through excessive Ca2+ influx, we investigated whether SRIF counteracted NMDA-induced neuronal cell death. Neurons from embryonic rat cerebral cortex were cultured for 7-10 days and then exposed to 0.5 and 1 mM NMDA for 24 h. The neuronal viability, as assessed by the colorimetric method, decreased by 40 and 60%, respectively, compared with the control condition. Morphological and biochemical evidence indicated that cell death occurred by necrosis and not through an apoptotic mechanism. SRIF (0.5-10 microM), simultaneously applied with excitatory amino acid, significantly reduced in a dose-dependent manner the neurotoxic effect of NMDA but not that of KA (0.25-0.5 mM). GABA (10 microM) partially protected neurons to a similar extent from NMDA- or KA-induced toxicity. SRIF type 2 receptor agonists, octreotide (SMS 201-995; 10 microM) and vapreotide (RC 160; 10 microM), did not influence the NMDA-dependent neurotoxicity. The intracellular mechanism involved in SRIF neuroprotection was investigated. Pertussin toxin (300 ng/ml), a G protein blocker, antagonized the protective effect of SRIF on NMDA neurotoxicity. Furthermore, the neuroprotective effect of SRIF was mimicked by dibutyryl-cyclic GMP (10 microM), a cyclic GMP analogue, whereas 8-(4-chlorphenylthio)-cyclic AMP (10 microM), a cyclic AMP analogue, was ineffective. The cyclic GMP content was increased in a dose-dependent manner by SRIF (2.5-10 microM). Finally, both specific (Rp-8-bromoguanosine 3',5'-monophosphate, 10 microM) and nonspecific [1-(5 isoquinolinylsulfonyl)-2-methylpiperazine (H7), 10 microM] cyclic GMP-dependent protein kinase (cGMP-PK) inhibitors did not interfere with NMDA toxicity but substantially reduced SRIF neuroprotection. Our data suggest a selective neuroprotective role of SRIF versus NMDA-induced nonapoptotic neuronal death in cortical cells. This effect is likely mediated by cGMP-PK presumably by regulation of the intracellular Ca2+ level.

A neurotoxic and gliotrophic fragment of the prion protein increases plasma membrane microviscosity.

Prion-related encephalopathies are characterized by astrogliosis and nerve cell degeneration and loss. These lesions might be the consequence of an interaction between the abnormal isoform of the cellular prion protein that accumulates in nervous tissue and the plasma membranes. Previously we found that a synthetic peptide, homologous to residues 106-126 of the human prion protein, is fibrillogenic and toxic to neurons and trophic to astrocytes in vitro. This study dealt with the ability of the peptide to interact with membranes. Accordingly, we compared PrP 106-126 with different synthetic PrP peptides (PrP 89-106, PrP 127-147, a peptide with a scrambled sequences of 106-126, and PrP 106-126 amidated at the C-terminus) as to the ability to increase the microviscosity of artificial and natural membranes. The first three had no effect on nerve and glial cells in vitro, whereas the amidated peptide caused neuronal death. Using a fluorescent probe that becomes incorporated into the hydrocarbon core of the lipid bilayer and records the lipid fluidity, we found PrP 106-126 able to increase significantly the membrane microviscosity of liposomes and of all cell lines investigated. This phenomenon was associated with the distribution of the peptide over the cell surface, but not with changes in the membrane lipid or protein content, or with membrane lipid phase transitions. Accordingly, we deduced that increased membrane microviscosity was unrelated to changes in the membrane native components and was the result of increased lipid density following PrP 106-126 embedding into the lipid bilayer. No control peptides had comparable effects on the membrane microviscosity, except PrP 106-126 amidated at the C-terminus. Since the latter was as neurotoxic, but not as fibrillogenic, as PrP 106-126, we argued that the ability of PrP 106-126 to increase membrane microviscosity was unrelated to the propensity of the peptide to raise fibrils. Rather, it could be connected with the primary structure of PrP 106-126, characterized by two opposing regions, one hydrophilic and the other hydrophobic, that enabled the peptide to interact with the lipid bilayer. Based on these findings, we speculated that the glial and nerve cell involvement occurring in prion-related encephalopathies might be caused by the interaction with the plasma membrane of a PrP 106-126-like fragment or of the sequence spanning residues 106-126 of the abnormal isoform of the prion protein.

Amidation of beta-amyloid peptide strongly reduced the amyloidogenic activity without alteration of the neurotoxicity.

Beta-amyloid accumulates in cerebral deposits in Alzheimer's disease, so to test the correlation between the neurotoxic and fibrillogenic capacity of beta-amyloid, we synthesized a peptide homologous to fragment 25-35 of beta-amyloid (beta25-35) and amidated at the C-terminus (beta25-35-NH2). As the amidation strongly reduced the amyloidogenic capacity of beta25-35, we compared its neurotoxic activity in the amidated (beta25-35-NH2) and nonamidated forms. The viability of primary cultures from fetal rat hippocampus was reduced in a dose-related manner (10-100 microM) similarly by beta25-35 and beta25-35-NH2, whereas a scrambled peptide, amidated or nonamidated, did not alter the neuronal viability. The neurotoxic activity of beta25-35-NH2 is mediated by apoptosis as demonstrated by morphological and biochemical investigations. Electron microscopy examination of culture media with beta25-35 or beta25-35-NH2 incubated with neuronal cells for 7 days confirmed the high level of fibrillogenic activity of beta25-35 and the almost total absence of fibrils in the solution with beta25-35-NH2. Furthermore, staining with thioflavine S was used to identify amyloid fibrils, and only the cultures exposed to beta25-35 exhibited intense staining associated with neuronal membranes. These data indicate that the neurotoxic activity of the beta-amyloid fragment is independent of the aggregated state of the peptide.

Clusterin (SGP-2) induction in rat astroglial cells exposed to prion protein fragment 106-126.

Prion-related encephalopathies are characterized by the accumulation of an abnormal prion protein isoform (PrPSc) associated with neuronal degeneration and astrogliosis. The synthetic peptide homologous to PrP fragment 106-126 (PrP 106-126) induced in vitro neuronal apoptosis and glial proliferation. We used Northern blot analysis and the RNA polymerase chain reaction to assess the expression of several genes associated with programmed cell death and proliferation. Blots of total RNA extracted from neuronal and astroglial cells exposed to PrP 106-126 for between 1 h and 7 days were hybridized with probes recognizing c-fos, c-jun, c-myc, p53, hsp-70 and bcl-2 mRNA. Except for a slight decrease in bcl-2 mRNA in neuronal cells, no change in other transcripts was evident. Since clusterin (apolipoprotein J) mRNA levels are increased in prion-related encephalopathies and clusterin immunoreactivity has been located in association with PrPSc in Gerstmann-Sträussler-Scheinker brain, the expression of clusterin was determined in neuronal and astroglial cells chronically exposed to PrP 106-126. Although the induction of clusterin has been involved in the apoptotic mechanism in other experimental conditions, its expression was unchanged in PrP 106-126-treated neurons, while a three-fold induction of clusterin mRNA was observed in astrocytes exposed to PrP 106-126. To investigate whether the clusterin up-regulation was simply associated with the astroglial proliferative stimulus of PrP 106-126 or was specifically induced by the peptide, we measured clusterin expression in astrocytes cultured in fetal calf serum-free medium and exposed to PrP 106-126 or fetal calf serum restoration. In this condition the PrP peptide, like fetal calf serum, increased the glial proliferation rate, but only PrP 106-126 doubled clusterin mRNA. The selectivity of this effect indicates that PrPSc is directly involved in the clusterin up-regulation seen in prion-related encephalopathies and is associated with astroglial cells.

Oxidative stress after acute and chronic application of beta-amyloid fragment 25-35 in cortical cultures.

The aim of this work was to investigate whether free radical reactions play a role in beta-amyloid neurotoxicity. Rat cortical neurons were exposed acutely (24 h) or chronically (3, 7 days) to beta-amyloid biologically active fragment beta 25-35 (50 microM). In these conditions, where only the longest exposure induced neuronal death, superoxide dismutase activity was increased after acute exposure but no change was detected after chronic treatments, whereas a different pattern was observed for glutathione peroxidase. In the basal condition, there was an eight-fold increase in dichlorofluoroscein, used as peroxide production marker, in neuronal cells after 7 days treatment with beta 25-35. Moreover, the intracellular peroxide production induced by Fe2+/ascorbate stimulation was amplified by beta 25-35, increasingly up to 7 days of exposure, by which time the dichlorofluoroscein-stimulated levels were 33 times higher than in controls. In conclusion, our results show that oxidative stress and free radical production are linked to beta 25-35 exposure and may contribute to neurodegenerative events associated with beta-amyloid deposits in Alzheimer's disease.

Reciprocal control of inflammatory cytokines, IL-1 and IL-6, and beta-amyloid production in cultures.

To investigate the role of IL-6 in the pathogenesis of Alzheimer's disease (AD) its effect on amyloid precursor protein (APP) mRNA expression was evaluated. The levels of APP mRNA were determined by Northern blot analysis in primary cultured rat cortical neurons and glial cells exposed to IL-6 (50-200 ng/ml). The cytokine increased neuronal APP mRNA expression about 100% at the highest dose after 6 h of exposure. APP mRNA expression was unaffected in astroglial cells exposed to IL-6. Since IL-1 beta also increased neuronal APP mRNA, the combination of IL-1 beta and IL-6 was tested. The effects were partially additive. The ability of beta-amyloid fragment 25-35 to induce IL-1 or IL-6 mRNA was also investigated in astroglial cells. IL-1 beta mRNA was strongly induced by beta 25-35 (25-100 microM) while the expression of IL-6 mRNA remaining unchanged. The results suggest roles for both IL-1 and IL-6 in the neuronal mechanisms related to beta-amyloid protein deposition in AD.

A neurotoxic prion protein fragment induces rat astroglial proliferation and hypertrophy.

Prion-related encephalopathies are characterized by the accumulation of an abnormal prion protein isoform (PrPSc) and the deposition of PrP amyloid in the brain. This process is accompanied by neuronal loss and astrogliosis. We recently showed that a synthetic peptide corresponding to residues 106-126 of human PrP is amyloidogenic and causes neuronal death by apoptosis in vitro. In the present study we investigated the effects of 1- and 14-day exposures of rat astroglial cultures to micromolar concentrations of this peptide as well as peptides homologous to other portions of PrP, a peptide corresponding to residues 25-35 of amyloid-beta protein, and a scrambled sequence of PrP 106-126. No significant changes were observed after 1-day exposure of cultures to any peptide. Conversely, 14-day treatment with PrP 106-126 (50 microM) resulted in a 5-fold increase in glial fibrillary acidic protein (GFAP) expression, as evaluated by Northern and Western blot analyses, and a 1.5-fold increment in cell number. Light and electron microscopy immunohistochemistry showed an enlargement in size and density of astroglial processes, and an increase in GFAP-immunoreactive intermediate filaments. These changes were not observed after 14-day treatment of cultures with the other peptides, including PrP 106-126 scrambled. The increase in GFAP expression of astroglial cultures exposed to PrP 106-126 was quantitatively similar to that found in scrapie-infected hamster brains.(ABSTRACT TRUNCATED AT 250 WORDS)

Neuroprotective activity of acetyl-L-carnitine: studies in vitro.

The neuroprotective properties of acetyl-L-carnitine (ALCAR) were investigated in primary cell cultures from rat hippocampal formation and cerebral cortex of 17-day-old rat embryos. Chronic exposure to ALCAR (10-50 microM for 10 days) reduced the cell mortality induced by 24 hr fetal calf serum deprivation. Protection was partial when the neuronal cells, chronically treated with ALCAR (50 microM), were exposed to glutamate (0.25-1 mM) and kainic acid (250-500 microM) for 24 hr. The neurotoxicity induced by N-methyl-D-aspartate (NMDA, 250 microM) was attenuated by the acute co-exposure with ALCAR (1 mM), the chronic treatment with ALCAR (50 microM) significantly reduced the neuronal death induced by NMDA (0.25-1 mM). Cell mortality was also investigated in ALCAR-treated hippocampal cultures chronically treated with beta-amyloid fragment 25-35. ALCAR appeared to have neuroprotective activity. This suggests an explanation of the positive results obtained with ALCAR in the treatment of Alzheimer's disease.

Nerve growth factor does not influence the expression of beta amyloid precursor protein mRNA in rat brain: in vivo and in vitro studies.

We investigated the effect of NGF on amyloid precursor protein (APP) mRNA levels in the rat septal/nucleus basalis system. Total APP mRNA and APP 695 mRNA were determined in basal forebrain primary cell cultures exposed acutely and chronically to NGF (150-300 ng/ml) and, in vivo, in the septal area and striatum of rat pups after multiple intracerebroventricular injections of NGF. The trophic factor was able to affect cholinergic neurons in both paradigms, as evidenced by the significant increase of choline acetyltransferase (ChAT) activity induced by NGF in cell cultures (+80%) and in the striatum (+240%) of rat pups. In spite of this effect, no significant change of APP mRNA expression was observed in neuronal cultures and brain tissues. These data indicate that the neurotrophic effect of NGF on forebrain cholinergic neurons is not always associated with an alteration of APP expression.

Apoptosis mediated neurotoxicity induced by chronic application of beta amyloid fragment 25-35.

To investigate whether and how amyloid-beta protein (A beta) is involved in the neurodegenerative changes characteristic of Alzheimer's disease (AD), primary hippocampal neurones from foetal rat brain were exposed acutely and chronically to micromolar concentrations of a synthetic peptide homologous to residues 25-35 of A beta (beta 25-35). A single application of this peptide (25-100 microM) was ineffective but when the neuronal cultures were exposed to beta 25-35 (25-100 microM) repeatedly every two days for ten days, cell survival was dramatically reduced. The structural changes and the DNA fragmentation of cells chronically exposed to the peptide suggested that neuronal death occurred by apoptosis. Furthermore, beta 25-35 showed the intrinsic ability to polymerize into amyloid-like fibrils in vitro. These results confirm the potential pathogenic role of A beta in AD, and indicate that amyloid fibrils may induce neuronal death through a specific programmed process.

Neurotoxicity of a prion protein fragment.

The cellular prion protein (PrPC) is a sialoglycoprotein of M(r) 33-35K that is expressed predominantly in neurons. In transmissible and genetic neurodegenerative disorders such as scrapie of sheep, spongiform encephalopathy of cattle and Creutzfeldt-Jakob or Gerstmann-Sträussler-Scheinker diseases of humans, PrPC is converted into an altered form (termed PrPSc) which is distinguishable from its normal homologue by its relative resistance to protease digestion. PrPSc accumulates in the central nervous system of affected individuals, and its protease-resistant core aggregates extracellularly into amyloid fibrils. The process is accompanied by nerve cell loss, whose pathogenesis and molecular basis are not understood. We report here that neuronal death results from chronic exposure of primary rat hippocampal cultures to micromolar concentrations of a peptide corresponding to residues 106-126 of the amino-acid sequence deduced from human PrP complementary DNA. DNA fragmentation of degenerating neurons indicates that cell death occurred by apoptosis. The PrP peptide 106-126 has a high intrinsic ability to polymerize into amyloid-like fibrils in vitro. These findings indicate that cerebral accumulation of PrPSc and its degradation products may play a role in the nerve cell degeneration that occurs in prion-related encephalopathies.

Expression of amyloid precursor protein mRNAs in endothelial, neuronal and glial cells: modulation by interleukin-1.

The origin of beta-amyloid deposited in senile plaques in Alzheimer's disease (AD) is not known. We compared the expression of protein precursor of beta-amyloid (APP) in the cell types involved in plaque formation. The levels of APP mRNA were determined in primary rat neurons and glial cells in culture, human endothelial cells and in a murine brain-derived endothelial cell line. Northern blot analysis was performed using an APP cDNA probe to detect the general APP sequence and an oligonucleotide (40 mer) complementary to the sequence of the Kunitz protease inhibitor (APP-KPI). The APP mRNA transcripts were abundant in all three cell types. The highest level of APP, normalized to beta-actin mRNA content, was expressed in neurons, followed by glial cells, where the APP expression was similar (94%) while in endothelial cells was lower (53%). The proportion between APP-KPI mRNA and total APP mRNA was high in endothelial, intermediate in glial and low in neuronal cells. We compared the effects of exposure to interleukin-1 (IL-1), a cytokine involved in several biological processes and elevated in AD, on APP mRNA expression in neuronal, glial and endothelial cells. In human endothelial and in brain-derived murine endothelial cells we observed a similar increase (50%) of total APP mRNA or APP-KPI mRNA after treatment with human recombinant IL-1 beta. In neuronal cells, IL-1 (200 ng/ml) substantially increased APP mRNA (175%), detected with both probes. In glial cells, the expression of APP mRNA did not appear to be altered by IL-1 (50-400 ng/ml). The results suggest a role of IL-1 in the neuronal mechanisms related to beta-amyloid protein deposition in AD.