Aβ Metallobiology and the Development of Novel Metal-Protein Attenuating Compounds (MPACs) for Alzheimers Disease

Over a decade of studies have pointed to metal mediated neural oxidative damage as an attractive target for the treatment of Alzheimer’s disease. Because of the nature of the blood brain barrier, systemic depletion of the metals, copper, zinc and possibly iron, is not a viable approach. However preliminary studies with CQ, a blood brain barrier penetrating chelating agent, are showing promise. CQ probably works by combining with the metal centres, primarily copper and zinc complexes of Aβ, in the neuropil. This review discusses some of the background that resulted in CQ becoming a lead compound and how we might advance our understanding of its action METALLOPROTEINS AND OXIDATION DAMAGE IN ALZHEIMER’S DISEASE Increasing evidence emphasises the importance of metals in neurobiology. For example, copper-binding proteins in the central nervous system may possess oxidant or anti-oxidant properties, possibly affecting neuronal function or triggering neurodegeneration. Among the copper-binding proteins related to neurodegenerative disease is the amyloid precursor protein (APP) of Alzheimer’s disease (AD) that has two copper-binding sites APP135-156, and near its N-terminus, APP1. APP is a highly conserved and widely expressed integral membrane protein with a single membrane-spanning domain. The amyloid β peptides (Aβ) are 39–43 residue polypeptides derived from proteolytic cleavage of APP, by the combined action of two proteases, BACE and γsecretase. A characteristic central nervous system histological marker in AD patients is accumulation of morphologically heterogeneous neuritic plaques and cerebrovascular deposits of Aβ [1]. Both the APP135 – 156 and As have been shown to have copper reducing activity with concomitant production of reactive oxygen species (ROS) [2, 3]. It has been long-established that oxidative damage to many classes of biological molecule, including sugars, lipids, proteins and nucleic acids, is increased in AD [4-6]. Cu and Fe interact with Aβ to make it toxic in cell culture. In vitro Aβ catalyses H2O2 generation through the reduction of Cu and Fe, using O2 and biological reducing agents, such as cholesterol, vitamin C and catecholamines, as substrates [710]. Consistent with these biochemical properties being responsible for disease, the neurotoxicity of Aβ in culture is mediated by the Aβ:Cu (or Aβ:Fe) forming H2O2 [8, 11]. *Address correspondence to this author at the Laboratory for Oxidation Biology, Genetics and Aging Research Unit, Massachusetts General Hospital East, Bldg 114, 16 Street, Charlestown, MA 02129, USA; Tel: 617-726-8244; Fax: 617-724-1823; E-mail: bush@helix.mgh.harvard.edu Aβ generation alone was once believed to engender toxicity. However, we found that Aβ was not toxic in the absence of Cu or Fe [3]. Although there have been reports of toxic fibrillar and toxic soluble oligomeric species of As, those studies have not yet excluded the possibility that the toxicity of the modified As species is dependent upon recruiting Cu or Fe from the culture medium. H2O2, being freely permeable across all tissue boundaries, unless scavenged by defences such as catalase and glutathione peroxidase, will react with Fe and Cu to generate OH• radicals by the Fenton reaction [10] that, in various cellular compartments, generates the lipid peroxidation adducts, protein carbonyl modifications, and nucleic acid adducts such as 8-OH guanosine that are typical of AD neuropathology [4, 6, 12] and precede Aβ deposition [13, 14]. Metal-centered ROS generation reactions of this kind have also been reported to potentially mediate the neurotoxicity of PrP in transmissible spongioform encephalopathies and alpha-synuclein in Parkinson’s disease [10, 15]. METAL MEDIATED OXIDATIVE DAMAGE AS A THERAPEUTIC TARGET The oxidative damage hypothesis of AD gave rise to a great deal of discussion about the part that various types of antioxidants and free radical scavengers might play in preventing or delaying the onset of the disease. This became part of the wider interest in the role of such compounds in promoting general health. Unfortunately, there is little hard evidence of the beneficial effect of their consumption on Alzheimer’s or any other neurodegenerative disease. Neither has any credible therapy been based on an antioxidant or free radical scavenger. On the other hand, experimental in vitro studies had shown that it was possible by sequestering the metal ion with a suitable chelator to block the production of ROS by copper complexed Aβ peptides. [16-18]. These findings suggested that it might be possible to develop a therapy based on metal complexation. Targeting a metal310 Curr. Med. Chem. – Immun., Endoc. & Metab. Agents, 2003, Vol. 3, No. 4 Curtain et al. binding site on a protein is not novel, and is quite different to chelation therapy, which aims to lower toxic metal burdens by sequestration. Examples of agents directed at metal sites are disulfiram, which chelates the zinc-catalytic site of alcohol dehydrogenase, blocking its activity [19], and the non-steroidal anti-inflammatory drugs, aspirin, diflunisal, ibuprofen, naproxen sodium, Indomethacin and D-penicillamine, which block the heme-iron catalytic site in the cyclooxygenase/arachidonic acid pathway [20]. An initial attempt to treat AD with the chelator desferrioxamine (DFO) showed a reduction in the rate of progression of the dementia [21], but this approach appears not to have been followed up. The reasons were probably that DFO requires twice-daily intramuscular injections and is a broad-spectrum chelator with a high affinity for iron, copper, zinc and aluminium that can cause systemic metal depletion. The affinity for aluminium was the rationale for the trial, since it was believed at the time that this metal played a role in producing the amyloid deposits and tau protein tangles in the brain, characteristic of the disease. Studies of the action of a range of metal chelators on post-mortem AD-affected brain tissue found that the dissolution of AβP in amyloid plaques was correlated with the release of Cu and Zn, but not Fe [22]. Recently, Dong et al. [23] have shown with the aid of Raman microscopy that treatment of amyloid plaques with the chelator tetraethyldiamine tetraacetate leads to a loosening of their characteristic β−structure owing to a reversal of Cu binding by the histidine residues of the plaques’ constituent Aβ peptides. Consideration of the bioavailability of the chelators led to in vivo trials of clioquinol (5-chloro-7-iodo-8-hydroxyquinoline, CQ), (Fig. 1). Oral treatment of human APP-expressing Tg2576 transgenic mice for 9 weeks with CQ caused a 49% decrease in brain Aβ deposition (-375 μg/g wet weight, p = 0.0001) in a blinded study [24]. Neurotoxicity was absent, Fig. (1). Clioquinol (5-chloro-7-iodo-8-hydroxyquinoline). and general health and body weights were significantly more stable in the treated animals, which were conspicuously improved after only 16 days of treatment. Compared with the results of a study of an anti-Aβ vaccine in the PDAPP transgenic mouse [25], the inhibition of the absolute amount of cerebral Aβ deposition by CQ was more extensive, and occurred more rapidly. CQ treatment, therefore, appears like the vaccine therapy to be a potent inhibitor of Aβ accumulation. Treatment of the Tg2576 mice with the a hydrophilic copper chelator, tetraethyl triamine acetic acid (TETA) that could not cross the blood brain barrier did not inhibit amyloid deposition [24], indicating that systemic metal depletion (eg “chelation therapy”) is not likely to be a useful therapeutic approach for AD. CQ has now proceeded into phase I [25] and phase II [27] clinical trials in AD patients. These early findings with CQ have been treated with cautious optimism in some quarters because of the reputation that the compound has of being a neurotoxin itself. Thirty years ago, CQ was removed from the market as an antibiotic after its use was linked to some 10, 000 cases of subacute myelo-optic neuropathy (SMON), mostly in Japan [28], although it is still used in topical antibiotic preparations. SMON symptoms and distal axonopathy could be reproduced by giving high doses of CQ to dogs and cats [29], and more recently [30] it has been suggested that CQ zinc chelates were the neurotoxin involved in SMON. It has also been suggested that the high incidence of the disease in Japan might have been a consequence of that population’s impaired vitamin B12 status after WWII [31]. In the Phase II trial the vitamin was coadminstered with CQ and no serious adverse effects were observed [27]. Another concern had been that dissolution of the plaques could cause a rise in soluble Aβ, which has been shown to be more neurotoxic than the insoluble form [32]. Again this concern has not been supported either by the transgenic mouse studies or in the clinical trial. Nevertheless, until its mechanism of action and its targets are thoroughly understood, there will be some concerns about long-term side effects with sustained maintenance or prevention therapy with CQ.