This annual meeting had significant international participation with 14 different countries being represented and, whilst comprehensive, the meeting focused on the cholinergic system, therapeutics and new findings.
The opening lecture on current and near-future therapeutics for Alzheimer's disease (AD) was presented by E Giacobini, Geneva Medical School. The only currently approved therapeutics for AD are acetyl cholinesterase inhibitors (AChEIs), a fact that is likely to remain for some time until therapeutics are developed based on other strategies such as β-protein precursor (βPP), estrogen, antioxidants or anti-inflammatory agents. The use of AChEIs to treat AD is based on the finding that loss of cholinergic function occurs early in the disease due to abnormalities in basal forebrain neurons that eventually lead to their death. Furthermore, strong evidence links the cholinergic system to learning and memory. This leaves open the possibility of therapeutics based on either inhibiting acetylcholinesterase (AChE) or mimicking acetylcholine (ACh) pharmacologically. Success has been made with inhibiting AChE, which is no small feat because it is the brain's most efficient enzyme, and therefore even 70 percent inhibition has little therapeutic effect.
Alzheimer's disease pathology: C Geula of Harvard University spoke of the pathogenesis of AD, a condition characterized by its main elements of extracellular senile plaques, intraneuronal neurofibrillary tangles (NFT) and neuronal loss. The major component of the senile plaque is amyloid-β (Aβ), a 39 to 43 amino acid fragment of the transmembrane precursor protein, βPP. In addition to extracellular Aβ, senile plaques have astrocytic microglia and neuritic reactive elements, the latter of which contains the same phosphorylated tau comprising the abnormal filaments of NFT.
The relationship of senile plaques and NFT to neuronal loss is not established because, whilst NFT are found in neurons at risk of death in AD, some neuronal populations spared to NFT are subject to death during the course of the disease. Further complicating the picture is the fact that variants of AD, characterized solely by NFT, senile plaques or Lewy bodies, have been documented but not clarified vis-à-vis primary etiology. This suggests that AD may be a syndrome generated by diverse mechanisms. Modeling the apparent relationship of the lesions to neuronal death has been difficult, maybe a major factor being that most experiments have been in young rodents or tissue culture. More recent experiments performed by Geula have demonstrated profound Aβ toxicity in aged primates, suggesting that models closer to the aged human may by necessary to evaluate the role of Aβ in AD.
Analyses of boron, phosphorous and sulfur in AD and controls, by using inductively coupled atomic emission spectrometry (ICP-AES), was presented by E Andrási, Eotvös Lorand University, Budapest. The findings showed sulfur accumulation and decreases in boron and phosphorous in AD patients.
I Degrell of University of Debrecen Medical School, Debrecen, Hungary, suggested that cerebral spinal fluid (CSF) reflects the metabolic differences between AD, controls and other conditions such as multi-infarct dementia (MID). His data showed that levels of glucose and lactate were lower in the younger cases of AD, whilst ascorbic acid was lower for all cases of AD, 8.1 mg/ml versus 32.2 mg/ml. Xanthine was elevated in AD and MID. These findings suggest a decrease in antioxidant defenses in AD.
The history of AChE is long, the first inhibitor, physostigmine being isolated in 1864 (Jobst and Hesse) and used therapeutically in 1877 (Laquer). However, it was not until the development of tacrine (Warner-Lambert Co) that the clinical benefit of AChE inhibition was noted. Evaluating drugs that benefit AD patients require sensitive psychometric instruments, e.g., Alzheimer's disease assessment scale (ADAS), and an understanding of the nature of cognitive decay with time (eight to nine points/year in the ADAS scale). With tacrine, one sees an immediate improvement but with a continued rate of decline essentially identical to that seen before treatment. With some of the newer AChE inhibitors, e.g., metrifonate (Bayer Corp), in addition to acute improvement, an actual stabilization of condition occurs at least for six months to a year. This, together with their lack of hepatotoxicity, e.g., with metrifonate and rivastigmine (Novartis AG), makes the new agents clearly the second generation of AChE inhibitors. Still in evaluating their effectiveness, it is important to define the patient group since most AChEIs are most efficacious for females and least for carriers of the apolipoprotein (ApoE) E4 allele.
Even though the benefits of anti-cholingerics are small, they are clinically profound, since they delay loss of non-cognitive function, e.g., activity of daily living, a major factor in determining nursing home entry. In the US, with over four million AD cases, delaying onset by just six months would result in 150,000 less afflicted individuals.
Abraham Fisher, Israel Institute for Biological Research, Ness-Ziona, Israel, presented his work on agonists to the M1 cholinergic receptor as a means of stimulating cholinergic function. He focused on M1, of the five muscarinic receptors, since it is most closely associated with learning and memory. These novel agonists, such as cevimeline (Snow Brand Milk Products Co Ltd), as well as more conventional agonists such as carbocol increase secreted bPP but not the related amyloid protein-like protein 1 or 2. In addition to affecting cholinergic function, M1 has "cross-talk" with growth factor pathways, e.g., bFGF, NFG and Grb, as well as signal transduction pathways, e.g., PKC and Ras. Therefore, it is not surprising that the M1 agonist AF-150 (Snow Brand) can act synergistically with NGF and bFGF to promote neuritic growth. Furthermore, M1 activation leads to tau dephosphorylation, suggesting that M1 is linked to cytoskeleton homeostasis. ApoE, the major genetic risk factor for late-onset AD, also appears to be linked to M1 since treatment with an M1 agonist can reverse the cognitive decline of ApoE-knockout mice. Moreover, these mice show reduced AChE.
However, the clinical use of M1 agonists has been relatively unsuccessful. Milameline (Parke-Davis & Co), a non-selective agonist, failed in phase III trails and, whilst xanomeline (Novo Nordisk A/S) was effective in maintaining ADAS cognitive assessment, it produced side-effects. More recently, nonetheless cevimeline has shown some promise of efficacy.
The issues related to inhibition of AChE as a therapeutic for AD were discussed by Bernard H Schmidt of Bayer CNS Research, Cologne, Germany. Not only is AChE one of the most active enzymes in the human body, it must also be inhibited by over 70 percent to see a clinical effect, requiring agents with minimal side effects. Of the currently available AChEIs, tacrine (80 to 160 mg), rivastigmine (6 to 12 mg), donepezil (Eisai Co Ltd; 5 to 10 mg) and metrifonate (50 to 80 mg), the major side-effects are gastrointestinal, and in the case of tacrine, additionally hepatotoxicity. Side effects are the major factor for patients leaving a study; thus, evaluation of the number leaving compared to placebo is a sensitive measure of side effects. Inhibition of AChE has no relationship to side effects. The most effective inhibitor would have a smooth onset of inhibition with long duration of action, which is the case for metrifonate.
Data on the issue of inhibition of butylcholinesterase (BuChE) versus AchE were presented by Z Rakonezay of Albert Szent-Györgi Medical University, Szeged, Hungary. Normally, AChE is in 3,000-fold excess over BuChE, although whilst AChE decreases in AD, BuChE often increases. Inhibition of both AChE and BuChE is therapeutically important and there is no need for specificity in inhibiting either enzyme.
Research into the of selective models of cholinergic deficits, surgical or chemical, were discussed by Dr Israel Hanin of Loyola University, Chicago. Either intervention drops ACh levels whilst having only a small effect on other systems such as norephinepherine. Memory and other deficits parallel those found in AD, including oxidative stress.
RG Wiley of Vanderbilt University, Nashville, Tennessee, is the developer of the immunotoxin, 192-saporin, that effectively can destroy the cholinergic system of the brain. In rats, 4 mg of this agent completely destroys cholinergic neurons in basal forebrain leading to memory deficits. Using this model, behavioral effects such as radial or water maze performance have been assessed, with the rats exhibiting decreased performance. Furthermore, changes in working versus reference memory have been investigated, with only the former being decreased. One interpretation of work on these animals is that loss of cholinergic function leads to lack of attention span, i.e., the ability to focus sufficiently to acquire new memories.
A careful comparison of the expression of mRNA and protein for a7 and a4 nicotinic receptors in neurons in cases of AD and controls was reported by A Wevers, University of Cologne, Germany. She found that whilst mRNA levels were unchanged in AD compared to controls, protein levels were depressed for both proteins. These findings suggest dysregulation in protein expression in AD, and it would be interesting to know if it is specific to expression of these receptors or instead was pleotrophic.
Dr B Csillik of Yale University, Connecticut, showed how dendrites of cholinergic neurons decreased from 512 neurons/mm3 in young to 160 neurons/mm3 in old monkeys, and how these cells are controlled by a number of features, the most notable being calcitonin gene-related protein (CGRP) axons that intimately surround each dendrite. Therefore, agents based on CGRP may be particularly useful in maintaining cholinergic neurons.
Alzheimer's disease proteins:
The major proteins involved in AD that play physiological roles in brain metabolism were discussed by Konrad Beyreuther of the University of Heidelberg, Germany. The transmembrane protein βPP is transported by fast anterograde axonal transport, requiring the Aβ domain for transport. ApoE plays a role in cholesterol transport into neurons where it directly affects βPP metabolism since cholesterol depletion favored the β-secretase pathway yielding increased Aβ. Estrogen (17 β-estradiol) directly effects cholesterol and reduces Aβ production by up to 45%. These findings suggest that by altering estrogen and cholesterol, new therapeutic interventions may be opened in AD.
The argument for the decisive role of Aβ in AD was discussed by B. Penke of Albert Szent-Gyorgi Medical University, Szeged, Hungary. He stated that Aβ is a neurotoxic peptide that, in a fibrous form, promotes astrogliosis and apoptosis. Dr Penke presented studies that monitored Aβ aggregation by infrared spectroscopy in solutions of water or dimethylsulfoxide and discussed the major limitation of these and other in vitro models of Aβ fibrillogenesis. Primary among the limitations is the need for non-physiological high concentrations of Aβ.
The issue of racemization, a time-dependent process that modifies about 0.15 percent of aspartate residues/year was presented by E Lang of Eötvös Loránd University, Budapest, Hungary. Aβ preparations derived from brain show about 4.57 ± 1.24 percent of D-aspartate, compared to 7.4 percent for myelin basic protein. However, in the presence of aluminum, and possibly other cations, racemization is increased by 182%. Data were presented showing racemization of aspartate 23 of the Aβ alters the peptide structure in such a way that allows the C-terminal portion of Aβ to more easily form a β-sheet. It was suggested that the link between aging and increased Aβ deposition may be Aβ aggregation related to racemization.
The lemur, Mircocebus murinus, as a model for primate aging was described by Nicole Bons, University of Montpellier, France. With a lifespan of nine to 13 years, "old" animals are easily obtained in captivity. The old animals show ventricular dilation and cortical atrophy as well as Aβ deposits, diffuse in younger aged animals and compact in the most aged. tau deposits outlined neurons in older animals. Significantly, the changes occur on a genetic backgroup that is homologous to humans by 92.7 percent for ApoE, 95.7 percent for βPP, 95.3 percent for presenilin-1 and 95.6 percent for presenilin-2.
The involvement of presenilins 1 and 2 in AD was discussed by Dora M Kovacs of Harvard University. Presenilins contain eight transmembrane domains and seem to play a role in neuronal apoptosis as well as Aβ cleavage. In presenilin mutants linked to AD, Aβ is preferentially cleaved to the more amyloidogenetic Aβ1-42. Furthermore, presenilin mutants affect caspase 3 activation. There was discussion on whether these findings implicate apoptosis in AD or instead avoidance of apoptosis.
The case for tau involvement in neurodegeneration beginning with Blessed's studies from the 1960s linking NFT to dementia was presented by Michael Novak of the Slovak Academy of Sciences, Bratislava. He showed that pronase treatment can remove 50 percent of tau mass from NFT and leave the filament intact. The pronase treated filaments are defined as the core, which he feels is the essential structure of the NFT. Additionally, tau is modified by phosphorylation (25 sites and involving 10 different kinases and 3 phosphatases), glycation, glycosylation and ubiquitination. Yet it is tau truncation that is responsible not only for NFT formation but also for other neuronal abnormalities. Overexpression of truncated tau by transfection led to apoptosis in cultured cells, an effect inhibited by caspase or calpain inhibitors.
The possibility that impairment of βPP metabolism plays a crucial role in AD was the focus of the talk by A Matsumoto, Kobe University, Japan. A 40 kDa protease is thought to be a key element in βPP metabolism. Antibodies to the 40 kDs protease recognize microglia as well as degenerating neurites of senile plaques. Whether the protease was an α- or β-secretase is not established.
The effect of Aβ on membrane potential of M213-20 cells by using flow-cytometry was discussed by G Laskay, Józef Attila University, Szeged, Hungary. Aβ induces hyperpolarization suggesting that it has ionophore properties. Carmela Abraham of Boston University studied βPP processing and identified four brain proteases capable of acting to generate Aβ. One, bleomyocin hydrolase, that shows a polymorphism genetically linked as a risk factor to cases of AD in Ashkenazi Jews, appears pivotal in Aβ generation. Her work also supports a role for the coordinate regulation of a serine protease by a metalloproteinase in Aβ formation.
Dr A Mitro of the Slovak Academy of Sciences, Bratislava, generated a monoclonal antibody to Aβ, termed MN5, that does not recognize Aβ but instead recognizes phosphorylated tau. These findings either suggest that Aβ and tau share structure similarities or that Aβ leads to tau release by the immunized mice and subsequent antibody production.
Dr T Harkany, Haynal Imre University, Budapest, injected Aβ into the basal forebrain of rats and found vitamins C and E as well as the calcium channel blocker, MK-801 (Merck & Co Inc), protect neurons from toxicity. The mechanisms of Aβ toxicity may involve mitochondrial toxicity from Ca2+ infusion, since manganese superoxide dismutase is 40 percent depleted by Aβ. These findings suggest that calcium blockers together with antioxidants may be of particular therapeutic value in AD.
Experiments using fura-2 to show Aβ increased Ca2+ levels in astrocytes, an effect preventable with a peptide homologous to Aβ, Pr-Ile-Ile-Gly-Leu-NH2 were described by G Laskay, Jósef Attila University. These findings suggest peptide antagonists can be effective in blocking Aβ toxicity.
The effect of metals and antioxidants as agents that provide protection from Aβ toxicity was discussed by T Harkany. It was found that when trace metals are administered with antioxidants in ischemia, greater protection is noted suggesting metal balance is critical in antioxidant defenses.
Linking the cholinergic and amyloid hypotheses Research on Aβ-AChE interactions were presented by Nidoba C Inestrosa, Catholic University of Chile, Santiago. The findings are based on the discovery that this enzyme is tightly complexed with senile plaques. In in vitro studies, Inestrosa found that AChE promotes Aβ fibrillogenesis. AChE interacts by a non-active site domain encompassed in a 3 kDa fragment. AChE-Aβ interaction is physiologically important to Aβ homeostasis since injection of AChE into rat brain can lead to Aβ deposition.
Dr S Kar of McGill University, Montreal, linked Aβ and the cholinergic deficits by showing that Aβ, aggregated or non-aggregated, affects ACh release. Peter Kasa, Albert Szent-Györgyi Medical University, also presented work linking Aβ to cholinergic deficits through its toxicity. In culture, cholinergic neurons show a major redistribution of AChE from perinuclear to perikarya following administration of Aβ or fragments of it. As shown by Kar, Aβ aggregation is not essential for Aβ toxicity.
M Pakaski, Albert Szent Medical University, showed that treating cholinergic neurons in culture with the AChEI scopolamine increases βPP levels, suggesting a link between βPP and cholinergic homeostasis.
The possibility that oxidative stress plays an important role in AD, Parkinson's disease, amyotrophic lateral sclerosis and Huntington's disease was discussed by P Klivényi of Albert Szent-Györgi Medical University. One of the best models to approach this issue is MPTP intoxication which blocks site I of mitochondrial respiration. Mice overexpressing Cu/Zn superoxide dismutase are protected from MPTP whilst mice deficient in glutathione peroxidase are more vulnerable.
Professor George Perry of Case Western Reserve University, Cleveland, Ohio, indicated that the role of mitochondrial deficiencies in AD may link oxidative stress to metabolic failure. Mitochondrial proliferation is a hallmark response to metabolic failure and has been demonstrated with in situ hybridization with cDNA probes directed to wild-type as well as mtDNA with the common 5 kb deletion. Additionally, cytochrome oxidase immunocytochemistry has been performed. Comparison of hippocampal neurons in AD with age-matched controls has shown a two- to threefold mitochondrial proliferation in AD patients, specifically for large neurons. The neurons displaying increased mtDNA show extensive RNA oxidative damage marked by 8OH-guanosine. Surprisingly, quantitative analysis has demonstrated that the increase in mtDNA is inversely correlated with the extent of 8OH-guanosine in the same neurons, suggesting that, whilst mitochondria may be responsible for much of the neuronal oxidative damage in AD, oxidative damage actually decreases during the course of the disease due to failure of mitochondrial respiration. Ultrastructural analysis of biopsy samples from AD and control cases have substantiated the profound mitochondrial abnormalities marking neurons in AD. Intriguingly, immunoelectron microscopy of βPP in well-preserved animal tissue has shown that much of the cellular βPP is associated with mitochondria.
The use of rat astrocytes and luminol fluorescence to study free radical production was discussed by C Torday of Albert Szent-Györgyi Medical University. Results of the studies showed dynamic interaction between NO and O2- species.
I Kovacs of Albert Szent-Györgyi Medical University showed that, whilst in rats, cholinergic axons of the olfactory bulb possibly innervate vessels, in humans, there is no such association. The major question raised by this study is the significance of cholinergic innervation since microvessels lacking smooth muscle are not thought to respond to innervation.
The fact that AD is associated with reduced vascular perfusion, the greater the deficit the greater the dementia, was discussed by Paul GM Luiten of the University of Groningen, The Netherlands. Structurally, vessels in AD show basement membrane thickening and periocyte degeneration. To model the reduced perfusion of AD, Luiten has performed partial ligation of the carotid artery in rats. These animals show reduced performance in memory tests correlated with perivascular basement membrane thickening.
Aβ entry into the brain from the vasculature in a model of blood barrier breakdown was examined by Maria Barcikowska, Polish Academy of Sciences, Warsaw. Global ischemia single (10 min) or repeated (3 x 10 min) was performed by inducing cardiac arrest followed by Aβ (5 mg) infusion. Barcikowska found that Aβ can enter the brain following a single ischemic event. Initially, one sees periocytes staining, but diffuse Aβ deposits can be noted even three months after Aβ infusion. Without the ischemic event, no Aβ enters the brain, and without Aβ infusion there is no depositon Aβ following ischemia. These findings not only show that Aβ can be cleared from the brain but also that under normal circumstances, Aβ cannot enter the brain from the circulation.
Dr G Jancsó of Albert Szent-Györgyi Medical University, as Barcikowska, infused Aβ into the brain vasculature but in this case to ask whether Aβ disrupts blood-brain barrier function. Aβ was infused into the left carotid artery of the rat followed by Evans blue. Jancsó found at 10-5 to 10-4M Aβ, aggregated or soluble, led to a substantial breakdown of barrier function in the ipsalateral but not the contralateral side. These findings suggest Aβ can alter brain permeability.
The important role of glia to maintain the blood-brain barrier and ionic balance was presented by L Latzkovits, Albert Szent-Györgyi Medical University.
This meeting highlighted the worldwide interest in development of novel therapeutics in the treatment of AD [Ref IDdb].—Reported by George Perry, Institute of Pathology, Case Western Reserve University, Cleveland, Ohio.
No Available References
No Available Further Reading