Lysosomotropic compounds and spermine enzymatic oxidation products in cancer therapy (review).

Apoptotic pathways represent the mechanisms of programmed cell death that counteract initiation and progression of cancer. New therapeutic targets are currently being explored on the basis of our detailed knowledge of the mechanisms and factors involved in apoptosis. In recent years, numerous proteins have been identified, which act as tumour suppressors or as oncoproteins in caspase-independent programmed cell death mechanisms, in which lysosomes are implicated for their lysosomal functions in cancer, mainly attributed to lysosomal proteinases, particularly the cathepsins. If cathepsins are released from the lysosomal lumen into the cytoplasm they initiate a number of processes that may cause either apoptotic or non-apoptotic (necrotic) cell death. The release of cathepsin D into the cytoplasm by vacuolar-type ATPase (V-ATPase) inhibitors produces the characteristic signs of apoptotic cell death, including caspase-3 activation and DNA laddering. For the destabilisation of the lysosomal membrane, two methods are available having therapeutic potential: the formation of reactive oxygen species (ROS) by irradiation or by enzymatic reactions and the lysosomal membrane permeabilisation by lysosomotropic compounds. Findings also suggest that the deregulation of polyamine metabolism or cytotoxic metabolites generated from the oxidative deamination of spermine by amine oxidases in association with lysosomotropic compounds may induce apoptosis. Cross-resistance of cells to cytotoxic actions of a wide variety of natural and synthetic anticancer drugs is the well-known phenomenon called multidrug resistance (MDR), due to glycoprotein P that functions as an ATP-dependent pump. The sensitisation of tumour cells to anticancer drugs by lysosomotropic compounds, and particularly the sensitisation of MDR-resistant cells recommend scrutinizing the potential of lysosomotropic drugs in cancer therapy.

Polyamines and the intestinal tract.

Owing to their high turnover, the intestinal mucosal cells have a particularly high requirement for polyamines. Therefore, they are an excellent charcol for the study of polyamine function in rapid physiological growth and differentiation. After a cursory introduction to the major aspects of polyamine metabolism, regulation, and mode of action, we discuss the contribution of the polyamines to the maintenance of normal gut function, the maturation of the intestinal mucosa, and its repair after injuries. Repletion of cellular polyamine pools with (D,L)-2-(difluoromethyl)ornithine has considerably improved our understanding of how the polyamines are involved in the regulation of normal and neoplastic growth. Unfortunately, the attempts to exploit polyamine metabolism as a cancer therapeutic target have not yet been successful. However, the selective inactivation of ornithine decarboxylase appears to be a promising chemopreventive method in familial adenomatous polyposis. Presumably, it relies on the fact that ornithine decarboxylase is a critical regulator of the proliferative response of the protooncogene c-myc, but not of its apoptotic response.

Synergism between apple procyanidins and lysosomotropic drugs: potential in chemoprevention.

BACKGROUND: Procyanidins are apple constituents with potential in colon cancer chemoprevention. MATERIALS AND METHODS: Human colon cancer derived metastatic cells (SW620), growing under standardized conditions, were exposed to procyanidins and lysosomotropic compounds. Growth, apoptosis and lysosomal integrity was determined using published methods. RESULTS: Lysosomotropic drugs (MDL 72527, phenylalanine methylester and chloroquine) amplified procyanidin-induced growth inhibition and apoptosis in SW620 cells at non-cytotoxic concentrations. The improved toxicity of the drug combinations relies primarily on the enhancement of lysosomal membrane permeability. CONCLUSION: Combinations with non-toxic concentrations of lysosomotropic compounds improve the anti-carcinogenic properties of apple procyanidins.

Sensitization of human colon adenocarcinoma cells (LoVo) to reactive oxygen species by a lysosomotropic compound.

The in situ formation of cytotoxic metabolites by an enzyme-catalyzed reaction is a recent approach in cancer therapy. The present results show that multidrug-resistant human colon adenocarcinoma cells (LoVo) are significantly more sensitive than corresponding wild-type cells to hydrogen peroxide and aldehydes, the products of bovine serum amine oxidase (BSAO)-catalyzed oxidation of spermine. Pre-treatment of the cells with N1,N4-bis(2,3-butadienyl)-1,4-butanediamine (MDL 72527), a lysosomotropic compound, sensitized both cell lines to the subsequent exposure to spermine metabolites, as was evident from the decrease of cell survival by a log unit. The sensitizing effect was greater in the case of the multidrug-resistant cell line, an aspect of particular importance with respect to potential therapeutic applications of the method, since conventional cancer therapy suffers from the development of drug resistance. Cell viability was determined using a clonogenic assay. MDL 72527 (at 300 microM) produced numerous cytoplasmic vacuoles, presumably of lysosomal origin, after 6-h exposure, which decreased in size and number (in the presence of the drug) by 24 h and had almost disappeared completely at 48 h. Mitochondrial damage, as observed by transmission electron microscopy, seemed to correlate better with the cytotoxic effects of the treatment than the formation of vacuoles. We suggest that the release of lysosomal enzymes into the cytosol by MDL 72527 is the main reason for its sensitizing effect. It is known that lysosomotropic compounds, which release lysosomal enzymes, produce oxidative stress and apoptosis.

Toxicity of enzymatic oxidation products of spermine to human melanoma cells (M14): sensitization by heat and MDL 72527.

In situ formation of cytotoxic metabolites by an enzyme-catalyzed reaction is a recent approach in cancer chemotherapy. We demonstrate that multidrug resistant human melanoma cells (M14 ADR) are more sensitive than the corresponding wild type cells (M14 WT) to hydrogen peroxide and aldehydes, the products of bovine serum amine oxidase (BSAO)-catalyzed oxidation of spermine. Hydrogen peroxide was mainly responsible for the loss of cell viability. With about 20%, the aldehydes formed from spermine contribute also to cytotoxicity. Elevation of temperature from 37 degrees C to 42 degrees C decreased survival of both cell lines by about one log unit. Pre-treatment with N1,N4-bis(2,3-butadienyl)-1,4-butanediamine (MDL 72527), a lysosomotropic compound, sensitized cells to toxic spermine metabolites. MDL 72527 (at 300 microM) produced in M14 cells numerous cytoplasmic vacuoles which, however, disappeared by 24 h, even in the presence of the drug. Mitochondrial damage, as observed by transmission electron microscopy, correlated better with the cytotoxic effects of the treatment than vacuole formation. Since the release of lysosomal enzymes causes oxidative stress and apoptosis, we suggest that the lysosomotropic effect of MDL 72527 is the major reason for its sensitizing effect.

Potentiation of apple procyanidin-triggered apoptosis by the polyamine oxidase inactivator MDL 72527 in human colon cancer-derived metastatic cells.

Apple procyanidins have chemopreventive properties in a model of colon cancer, they affect intracellular signalling pathways, and trigger apoptosis in a human adenocarcinoma-derived metastatic cell line (SW620). In the present study we investigated relationships between procyanidin-induced alterations in polyamine metabolism and apoptotic effects. Apple procyanidins diminish the activities of ornithine decarboxylase and S-adenosyl-L-methionine decarboxylase, key enzymes of polyamine biosynthesis, and they induce spermidine/spermine N(1)-acetyltransferase, which initiates retroconversion of poly-amines. As a consequence of the enzymatic changes polyamine concentrations are diminished, and N(1)-acetyl-polyamines accumulate in SW620 cells. In contrast with expectations MDL 72527, an inactivator of polyamine oxidase (PAO), improved the anti-proliferative effect of procyanidins, and caused an increase of the proportion of apoptotic cells, although it prevented the formation of hydrogen peroxide and 3-acetamidopropanal, the cytotoxic products of PAO-catalysed degradation of N(1)-acetylspermidine and N1-acetylspermine. Addition of 500 microM N1-acetylspermidine to the culture medium in the presence of procyanidins mimicked the effect of MDL 72527. Therefore we presume that the enhanced procyanidin-triggered apoptosis by MDL 72527 is mediated by the accumulation of N(1)-acetyl-polyamines. The observation that apple procyanidins enhance polyamine catabolism and reduce polyamine biosynthesis activity similar to known inducers of SSAT, without sharing their toxicity, and the potentiation of these effects by low concentrations of MDL 72527 suggests apple procyanidins for chemopreventive and therapeutic interventions.

Interaction of bovine serum amine oxidase with the polyamine oxidase inactivator MDL 72527.

MDL 72527 was considered a selective inhibitor of FAD-dependent polyamine oxidases. In the present communication, we demonstrate that MDL 72527 inactivates bovine serum amine oxidase, a copper-containing, TPQ-enzyme, time-dependently at 25 degrees C. In striking contrast, the enzyme remained active after incubation with excessive MDL 72527 at 37 degrees C, even after 70 h of incubation. Inactivation of BSAO with MDL 72527 at 25 degrees C did not involve the cofactor, as was shown by spectroscopy and by reaction with phenylhydrazine. Docking of MDL 72527 is difficult, owing to its size and two lipophilic moieties, and it has been shown that minor changes in reaction rate of substrates cause major changes in K(m) and k(cat)/K(m). We hypothesise that subtle conformational changes between 25 and 37 degrees C impair MDL 72527 from productive binding and prevent the nucleophilic group from reacting with the double bond system.

Cytotoxicity of the polyamine oxidase inactivator MDL 72527 to cancer cells: comparison with a saturated structural analogue.

MDL 72527 (N1,N4-di-2,3-butadienyl-1,4-butanediamine) is a selective inactivator of polyamine oxidase with therapeutic potential. However, the development of lethal toxic effects due to prevention of spermine degradation is a considerable disadvantage of the compound. Since the cytotoxicity of MDL 72527 was postulated to be independent of its anti-polyamine oxidase activity, its cytotoxicity to cancer cells was compared with that of a close analogue that is devoid of structural features enabling mechanism-based inactivation of polyamine oxidase. N1,N4-di-n-butyl-1,4-butanediamine proved to be a cytotoxic agent of considerable potency, which induces mainly non-apoptotic cell death, whereas MDL 72527 causes under identical conditions both, apoptotic and non-apoptotic cell death. The sensitivity of cells to both compounds is presumably dependent of their glutathione content.

Polyamines and apoptosis.

The natural polyamines putrescine, spermidine and spermine are in multiple ways involved in cell growth and the maintenance of cell viability. In the course of the last 15 years more and more evidence hinted also at roles in gene regulation. It is therefore not surprising that the polyamines are involved in events inherent to genetically programmed cell death. Following inhibition of ornithine decarboxylase, a key step in polyamine biosynthesis, numerous links have been identified between the polyamines and apoptotic pathways. Examples of activation and prevention of apoptosis due to polyamine depletion are known for several cell lines. Elevation of polyamine concentrations may lead to apoptosis or to malignant transformation. These observations are discussed in the present review, together with possible mechanisms of action of the polyamines. Contradictory results and incomplete information blur the picture and complicate interpretation. Since, however, much interest is focussed at present on all aspects of programmed cell death, a considerable progress in the elucidation of polyamine functions in apoptotic signalling pathways is expected, even though enormous difficulties oppose pinpointing specific interactions of the polyamines with pro- and anti-apoptotic factors. Such situation is quite common in polyamine research.

Pharmacological aspects of cytotoxic polyamine analogs and derivatives for cancer therapy.

During the past 20 years, numerous derivatives and analogues of spermidine (Spd) and spermine (Spm) were synthesized with the aim to generate a new type of anticancer drug. The common denominator of most cytotoxic polyamine analogues is their lipophilicity, which is superior to that of the parent amines. The natural polyamines bind to polyanions and to proteins with anionic binding sites. Their hydrophilicity/hydrophobicity is balanced, allowing them to perform physiological functions by interacting with some of these anionic structures, without impairing the functionality of others. Because the attachment of lipophilic substituents to the polyamine backbone increases the binding energy, lipophilic polyamine derivatives affect secondary and tertiary structures of a larger number of macromolecules than do their natural counterparts. In addition, lipophilicity improves the blood-brain barrier transport and thus enhances CNS toxicity. Close structural analogues of spermidine and spermine mimic the natural polyamines in regulatory functions. The cytotoxic mechanisms of analogues with a less close structural resemblance to spermidine or spermine have not been completely clarified. The displacement of spermidine from functional binding sites and the consequent prevention of its physiological roles is a likely mechanism, but many others may play a role as well. Up to now, polyamine analogues were conceived without specific growth-related targets in mind. To develop therapeutically useful drugs, it will be imperative to identify specific targets and to design compounds that interact selectively with the target molecules. It will also be necessary to include, at an early state of the work, pharmacological and toxicological considerations, to avoid unproductive directions.

Chemopreventive properties of apple procyanidins on human colon cancer-derived metastatic SW620 cells and in a rat model of colon carcinogenesis.

Apples contain several classes of polyphenols: monomers (catechins, epicatechins) and oligomers/polymers, such as the procyanidins. Our aim was (i) to study anti-proliferative mechanisms on human metastatic colon carcinoma (SW620 cells) of apple polyphenol fractions (monomers or procyanidins) and (ii) to evaluate their anti-carcinogenic properties in vivo. Two polyphenol-enriched fractions were isolated from apples. Fraction non-procyanidins contained 73% phenolic monomers and no procyanidins, while fraction procyanidins contained 78% procyanidins and no monomers. Inhibition of SW620 cell growth was only observed with fraction P (IC50 = 45 microg/ml). After a 24-h exposure of cells to fraction P, protein kinase C activity was inhibited by 70% and a significant increase in extracellular signal-regulated kinases 1 and 2 and c-jun N-terminal kinases expression was observed together with the down-regulation of polyamine biosynthesis and the activation of caspase-3. Colon carcinogenesis was induced in rats by intraperitoneal injections of azoxymethane, once a week for 2 weeks. Seven days after the last injection, Wistar rats received fraction P (0.01%) dissolved in drinking water. After 6 weeks of treatment, the colon of rats receiving procyanidins showed a significant (P < 0.01) reduction of the number of preneoplastic lesions when compared with controls receiving water. The total number of hyperproliferative crypts and of aberrant crypt foci was reduced by 50% in rats receiving 0.01% apple procyanidins in their drinking water. Our results show that apple procyanidins alter intracellular signaling pathways, polyamine biosynthesis and trigger apoptosis in tumor cells. These compounds antagonize cancer promotion in vivo. In contrast with absorbable drugs, these natural, non toxic, dietary constituents reach the colon where they are able to exert their antitumor effects.

Polyploidisation of metastatic colon carcinoma cells by microtubule and tubulin interacting drugs: effect on proteolytic activity and invasiveness.

When SW620 colon cancer-derived metastatic cells were exposed to nanomolar concentrations of Taxol, colchicine or (Z)-3,5,4'-trimethoxystilbene (R3), huge aneuploid, polynuclear cells survived the treatment. These cells released considerable amounts of the matrix metalloproteinase matrilysin (MMP-7), and tissue-type plasminogen activator (tPA) into the surrounding culture medium. MMP-7, and other proteolytic enzymes were highly expressed by these cells. In spite of their enormous size, the polyploid cells exhibited a considerable migratory capacity, as was demonstrated by their migration through an artificial basement membrane. While colchicine and R3-treated cells showed an inverse relationship between drug concentration and invasiveness, treatment with Taxol increased the capacity of the SW620 cells to penetrate through the membrane. The invasive capacity was not correlated with the induction and release of proteolytic enzymes. The idea that expression and release of proteolytic enzymes is a fundamental prerequisite of tumour cell invasiveness is generally accepted. The ability of the cells to respond to chemotactic signalling, and the filamentous structures of the cells, together with several cell adhesion factors, which are the basis of cell migration, are prerequisites of invasiveness. These factors are presumably different in the aneuploid cells produced by Taxol, colchicine and R3, and await scrutiny.

N1-dansyl-spermine: a potent polyamine antagonist.

The potential polyamine antagonist action of N1-dansyl-spermine (a potent NMDA antagonist) was assessed in two in vivo mouse models of polyamine action. Co-administration of N1-dansyl-spermine (2-10 microg, i.c.v.) with spermine (100 microg, i.c.v.) resulted in a dose-dependent antagonism of the spermine-induced CNS excitation (body tremor and fatal tonic convulsions). In addition, the same dose of N1-dansyl-spermine antagonised spermine's enhancement of NMDA-induced convulsions. These results suggest that N1-dansyl-spermine is in vivo a potent antagonist of the CNS effects of spermine and of its action at the positive polyamine modulatory site on the NMDA receptor.

Thirty years of polyamine-related approaches to cancer therapy. Retrospect and prospect. Part 1. Selective enzyme inhibitors.

As soon as the natural polyamines (PAs), putrescine (Put), spermidine (Spd) and spermine (Spm), were recognized as ubiquitous constituents of eukaryotic cells, their involvement in growth-related processes attracted particular interest. The high activities of ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylase (AdoMetDC) in rapidly growing tissues and cells, particularly in tumour cells, suggested PA biosynthesis as a target for antineoplastic therapy. In the course of the years selective inhibitors have been developed for literally all enzymes of PA metabolism. Some became important as tools in the elucidation of the PA metabolic system, but only few of them were efficient as inhibitors of tumour growth. A major reason for the inefficacy of selective enzyme inhibitors as anticancer drugs is the sophistication of the system, which regulates intracellular PA pools. Selective blockade of a single enzyme induces changes of metabolism and transport, which compensate for the deficit. The selective impairment of tumour growth is in addition hampered by the ubiquitous occurrence of the PAs, their importance in normal functions of nearly all mammalian cells, and the ability or the mammalian organism to utilize exogenous (gastrointestinal) PAs. Among the inhibitors of PA-related enzymes, the ODC inactivator (R, S)-2-(difluoromethyl)ornithine (DFMO) became most famous. Although it was disappointing in most therapeutic attempts to use it as single drug, it has--based on its low toxicity--considerable potential in cancer chemoprevention, and it turned out to be a highly efficient anti-trypanosome agent. Very likely DFMO is suitable to improve the efficacy of some of the current cytotoxic drugs, and it may allow one to create new therapies in combination with other PA-directed drugs. Some of the less selective enzyme inhibitors, particularly those, which inhibit two or more enzymes of PA metabolism, appear to have had a chance to become practically useful, but they have not been developed energetically. Disregarding DFMO, the AdoMetDC inhibitor SAM486A is the only compound for which clinical trials were published. The future of this drug is unclear at present; presumably phase III clinical trials have been discontinued. One of the lessons that had to be learned from the work on selective enzyme inhibitors was that PA metabolism is a much more difficult target, than has been expected on the basis of the simplicity of the PA structures, and the simple reactions involved in their biosynthesis. In order to inhibit tumour growth several reactions or regulatory functions of PA metabolism have to be impaired at the same time. Recent efforts devoted to the development new types of anticancer drugs, which are based on the perturbation of PA metabolism by structural analogues of the natural PAs, take this message into account. These approaches are the topic of the 2nd part of this overview.

Thirty years of polyamine-related approaches to cancer therapy. Retrospect and prospect. Part 2. Structural analogues and derivatives.

Owing to their role in growth-related processes the natural polyamines (PAs), putrescine (Put), spermidine (Spd) and spermine (Spm) were identified about 30 years ago as potential targets for the development of anticancer drugs. It was presumed that inhibition of a key enzyme of PA biosynthesis, followed by the depletion of the intracellular PA pools results in the prevention of cell growth. Initial efforts were nearly exclusively focused on the design and synthesis of selective inhibitors of the PA biosynthetic enzymes. This period is reviewed in the 1st part. Selective inhibition of ODC caused in various cell lines growth inhibition, but was usually not sufficient to inhibit tumour growth, because the PA regulatory system outbalances selective enzyme blockade by enhancing compensatory reactions, and because exogenous PAs are used if de nova synthesis is impaired. When these facts were recognized, new targets were envisaged. Among these the PA uptake system and the deregulation of PA homeostasis became most attractive. They are the major topic of the present 2nd part. Inhibition of PA uptake from the cellular environment is expected to improve the efficacy of drugs, which rely on the depletion of intracellular PA pools. During the past few years several potent inhibitors of the PA uptake system became known. However, more work will be needed to allow their assessment as anticancer drugs in combination with DFMO and other compounds capable of depleting PA pools. The PA transport system also offers the possibility to improve the accumulation by tumors of compounds, which are tethered to PA structures. This can be achieved for the following reasons: (a) Structural requirements of the PA uptake systems are not stringent. (b) Tumour cells accumulate PAs more avidly than most non-transformed cells. (c) The transport rate for PAs is higher in cells with depleted PA pools, than in cells with a normal PA content. (d) In cells, which proliferate rapidly, PA depletion by biosynthesis inhibitors is more effective, than in slowly growing cells. The most actively pursued approach is currently based on the inhibition of tumour growth by cytotoxic structural analogues of the natural PAs. Some of these compounds mimic regulatory properties of the natural PAs. If a PA mimetic accumulates in cells, it induces catabolic processes, suppresses biosynthetic reactions, and depletes the pools of Put. Spd and Spm. N1,N11 -bis(ethyl)norspermine is a prototype of the PA mimetics. During the last decade a very large number of PA derivatives and structural analogues have been prepared, which are capable of inhibiting cell growth at low microM concentrations. Among the new PA-like structures several compounds were identified, which prevent cells from growing, without depleting PA pools to an extent that would be necessary to prevent cell growth. They may be considered as PA antagonists, although their mode of action is not well understood. A therapeutically useful drug has not yet been identified among the PA analogues. In many instances investigations were stopped at a preliminary stage. Recently synthesized compounds have not yet been pursued far enough to justify the initiation of clinical trials. Only very few toxicological results of the new structures have been reported, although the knowledge of the toxicology of Spm analogues is of eminent importance. PAs are ubiquitous cell constituents and are indispensable for normal cell function. However. extracellular PAs, and particularly extracellular Spm is cytotoxic and neurotoxic. These properties are shared by close structural analogues. A major difficulty in the development of PA analogues to therapeutically useful drugs is, therefore, the identification of structures, which do not share neurotoxic properties with Spm. Several tetramines are presently in early phases of clinical trials. It will be a matter of a few more years to allow one to decide, whether PA-related approaches of cancer therapy are a success or a failure.

The polyamine oxidase inactivator MDL 72527.

Polyamine oxidase is a FAD-dependent amine oxidase, which is constitutively expressed in nearly all tissues of the vertebrate organism. In 1985, N1,N4-bis(2,3-butadienyl)-1,4-butanediamine (MDL 72527) was designed as a selective enzyme-activated irreversible inhibitor of polyamine oxidase (EC 1.5.3.11). It inactivates, at micromolar concentration and time-dependently, the enzyme in cells, as well as in all organs of experimental animals, without inhibiting other enzymes of polyamine metabolism. MDL 72527 served during nearly two decades as a unique tool in the elucidation of the physiological roles of polyamine oxidase. The compound has anticancer and contragestational effects, and it improves the anticancer effect of the ornithine decarboxylase inactivator (D,L)-2-(difluoromethyl)ornithine (DFMO). Profound depletion of the polyamine pools of tumour cells and effects on different components of the immune defence system are responsible for the anticancer effects of MDL 72527/DFMO combinations. Recently a direct cytotoxic effect of MDL 72527 at concentrations above those required for polyamine oxidase inactivation was observed. The induction of apoptosis by MDL 72527 was ascribed to its lysosomotropic properties. Therapeutic potentials of the apoptotic effect of MDL 72527 need to be explored. Polyamine oxidase is the last enzyme of the polyamine interconversion pathway that awaits the detailed elucidation of its structure and regulation. MDL 72527 should be useful as a lead in the development of inactivators which are selective for the isoforms of polyamine oxidase. Isozyme-selective inhibitors will give more profound insights into and reveal a diversity of specific functions of polyamine oxidase.

(Z)-1,4-diamino-2-butene as a vector of boron, fluorine, or iodine for cancer therapy and imaging: synthesis and biological evaluation.

Polyamine vectors are attractive for tumor targeting. We envisaged (Z)-1,4-diamino-2-butene (Z-DAB), an unsaturated analogue of putrescine as vector of (10)B, (18)F and (131)I for boron neutron capture therapy (BNCT), and tumor imaging by positron emission tomography or scintigraphy respectively. In the present work, the synthesis and characterization of new derivatives of Z-DAB were reported. Z-DAB was actively transported in cells via the polyamine transport system and converted into the spermidine analogue.(E)-2-iodo-1,4-diamino-2-butene (E-I-DAB) was not taken up by the polyamine transport system and may not be suitable for tumor imaging. In contrast, (Z)-2-[4-(5,5-dimethyl-dioxaborinan-2-yl)phenyl]methyl-1,4-diamino-2-butene (Z-4-Bbz-DAB) was a substrate of the transport system and allowed significant boron accumulation in 3LL cells. Its potential in BNCT will be evaluated.

Ammonia and Alzheimer's disease.

Alzheimer's disease (AD) is the most common age-related neurodegenerative disorder. Behavioural, cognitive and memory dysfunctions are characteristic symptoms of AD. The formation of amyloid plaques is currently considered as the key event of AD. Other histological hallmarks of the disease are the formation of fibrillary tangles, astrocytosis, and loss of certain neuronal systems in cortical areas of the brain. A great number of possible aetiologic and pathogenetic factors of AD have been published in the course of the last two decades. Among the toxic factors, which have been considered to contribute to the symptoms and progression of AD, ammonia deserves special interest for the following reasons: (a) Ammonia is formed in nearly all tissues and organs of the vertebrate organism; it is the most common endogenous neurotoxic compounds. Its effects on glutamatergic and GABAergic neuronal systems, the two prevailing neuronal systems of the cortical structures, are known for many years. (b) The impairment of ammonia detoxification invariably leads to severe pathology. Several symptoms and histologic aberrations of hepatic encephalopathy (HE), of which ammonia has been recognised as a pathogenetic factor, resemble those of AD. (c) The excessive formation of ammonia in the brains of AD patients has been demonstrated, and it has been shown that some AD patients exhibit elevated blood ammonia concentrations. (d) There is evidence for the involvement of aberrant lysosomal processing of beta-amyloid precursor protein (beta-APP) in the formation of amyloid deposits. Ammonia is the most important natural modulator of lysosomal protein processing. (e) Inflammatory processes and activation of microglia are widely believed to be implicated in the pathology of AD. Ammonia is able to affect the characteristic functions of microglia, such as endocytosis, and cytokine production. Based on these facts, an ammonia hypothesis of AD has first been suggested in 1993. In the present review old and new observations are discussed, which are in support of the notion that ammonia is a factor able to produce symptoms of AD and to affect the progression of the disease.