Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Metal Ionophore Treatment Restores Dendritic Spine Density and Synaptic Protein Levels in a Mouse Model of Alzheimer's Disease

  • Paul A. Adlard ,

    p.adlard@mhri.edu.au

    Affiliations Oxidation Biology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia, Synaptic Neurobiology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia, Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia

  • Laura Bica,

    Affiliations Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia, Centre for Neuroscience, The University of Melbourne, Parkville, Victoria, Australia

  • Anthony R. White,

    Affiliations Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia, Centre for Neuroscience, The University of Melbourne, Parkville, Victoria, Australia

  • Milawaty Nurjono,

    Affiliation Oxidation Biology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia

  • Gulay Filiz,

    Affiliation Oxidation Biology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia

  • Peter J. Crouch,

    Affiliations Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia, Centre for Neuroscience, The University of Melbourne, Parkville, Victoria, Australia

  • Paul S. Donnelly,

    Affiliations Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, Victoria, Australia, School of Chemistry, The University of Melbourne, Parkville, Victoria, Australia

  • Roberto Cappai,

    Affiliations Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, Parkville, Victoria, Australia

  • David I. Finkelstein ,

    Contributed equally to this work with: David I. Finkelstein, Ashley I. Bush

    Affiliations Oxidation Biology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia, Synaptic Neurobiology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia, Centre for Neuroscience, The University of Melbourne, Parkville, Victoria, Australia

  • Ashley I. Bush

    Contributed equally to this work with: David I. Finkelstein, Ashley I. Bush

    Affiliations Oxidation Biology Laboratory, The Mental Health Research Institute, Parkville, Victoria, Australia, Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia

Metal Ionophore Treatment Restores Dendritic Spine Density and Synaptic Protein Levels in a Mouse Model of Alzheimer's Disease

  • Paul A. Adlard, 
  • Laura Bica, 
  • Anthony R. White, 
  • Milawaty Nurjono, 
  • Gulay Filiz, 
  • Peter J. Crouch, 
  • Paul S. Donnelly, 
  • Roberto Cappai, 
  • David I. Finkelstein, 
  • Ashley I. Bush
PLOS
x

Abstract

We have previously demonstrated that brief treatment of APP transgenic mice with metal ionophores (PBT2, Prana Biotechnology) rapidly and markedly improves learning and memory. To understand the potential mechanisms of action underlying this phenomenon we examined hippocampal dendritic spine density, and the levels of key proteins involved in learning and memory, in young (4 months) and old (14 months) female Tg2576 mice following brief (11 days) oral treatment with PBT2 (30 mg/kg/d). Transgenic mice exhibited deficits in spine density compared to littermate controls that were significantly rescued by PBT2 treatment in both the young (+17%, p<0.001) and old (+32%, p<0.001) animals. There was no effect of PBT2 on spine density in the control animals. In the transgenic animals, PBT2 treatment also resulted in significant increases in brain levels of CamKII (+57%, p = 0.005), spinophilin (+37%, p = 0.04), NMDAR1A (+126%, p = 0.02), NMDAR2A (+70%, p = 0.05), pro-BDNF (+19%, p = 0.02) and BDNF (+19%, p = 0.04). While PBT2-treatment did not significantly alter neurite-length in vivo, it did increase neurite outgrowth (+200%, p = 0.006) in cultured cells, and this was abolished by co-incubation with the transition metal chelator, diamsar. These data suggest that PBT2 may affect multiple aspects of snaptic health/efficacy. In Alzheimer's disease therefore, PBT2 may restore the uptake of physiological metal ions trapped within extracellular β-amyloid aggregates that then induce biochemical and anatomical changes to improve cognitive function.

Introduction

Alzheimer's disease (AD) is characterized by deficits in higher order cognitive processes. One of the principal substrates for this dysfunction is spine/synaptic loss, which studies in AD animal models have demonstrated is most pronounced in the immediate vicinity of Aβ plaques [1], [2], [3]. Furthermore, Aβ oligomers impair NMDAR-dependent signalling cascades, ultimately resulting in the progressive loss of dendritic spines and glutamatergic synapses [4], [5], [6].

One of the mechanisms likely to mediate such deleterious effects is perturbed trafficking of essential synaptic metal ions, such as zinc. Synaptic Zn2+ functions as a neuronal messenger and a modulator of synaptic transmission and plasticity through targeted interactions with proteins such as TrkB and NMDAR2b [7], [8]. The attraction of Aβ oligomers to Zn2+ emanating from the glutamatergic synapse selectively occludes the NMDAR2b subunit [9]. Thus, the sequestration of Zn2+ in oligomeric Aβ-Zn complexes may lead to a reduction in zinc turnover at the synapse, limiting the trans-synaptic movement of zinc to modulate post-synaptic targets, and resulting in impaired cognition.

Drug candidates clioquinol (CQ) and PBT2 (Prana Biotechnology Ltd) have a moderate affinity for metal ions, and rather than deplete biological metals in cell culture, promote the uptake of Cu and Zn [10]. We have previously demonstrated [10] that brief administration of CQ and PBT2 (11 days, 30 mg/kg) to both Tg2576 and APP/PS1 transgenic mice resulted in decreased brain Aβ burden with rapid (within 5 days) and marked improvements in learning and memory performance on the Morris water maze. PBT2 also induced cognitive benefits in a phase IIa clinical trial [11], [12]. To understand the mechanisms that underlie this restoration of cognition, we studied the effect of PBT2 on synaptic plasticity-related endpoints in both cell culture and Tg2576 mice.

Materials and Methods

All animal procedures were approved by the Howard Florey Institute animal ethics committee (Melbourne, Australia), and were carried out in accordance with the Australian code of practice for the care and use of animals for scientific purposes as described by the National Health and Medcial Research Council of Australia. Female Tg2576 animals were group-housed and aged to 4 or 14 months. For 11 days prior to culling, animals were given a daily oral gavage of sham (standard suspension vehicle [10]) or PBT2 (30 mg/kg, provided by Prana Biotechnology Ltd, Melbourne, Australia) in the same vehicle (n = 8–12 mice/group total; split between golgi and biochemical analyses). One hour following the last drug dose, animals were culled and the brain either immediately processed for Golgi analysis, or frozen at −80°C for subsequent biochemical analyses. The dose of PBT2 was chosen based upon our own historical data, where we have shown that this level of exposure to the compound effectively modulates a number of biochemical and behavioural parameters in various transgenic mouse models of AD, but does not alter bulk brain metal levels [10]. With regards to how this relates to a human dose, based upon FDA guidelines a dose of 30 mg/kg in a mouse equates to a dose of 183 mg/day in a 75 kg man. Our published human clinical trial data demonstrated that PBT2 was found to be therapeutic at 250 mg/day [11], [12].

Golgi staining

A subset of animals (n = 3–5 mice/group) were deeply anaesthetized with sodium pentobarbitone (100 mg/kg) and then transcardially perfused with 0.01 M cold PBS. Brains were then rapidly removed and cut into 4 mm blocks. The blocks containing the hippocampus were incubated in rapid golgi solutions (Rapid Golgi Stain Kit, MTR scientific), based on the methods of Glasser and Van der Loos [13]. The blocks were infused with a solution containing potassium dichromate, potassium chromate and mercuric chloride for two weeks. The blocks were then snap frozen and cut at 90 µm using a microtome (Cryostat, Leica CM18-50, Leica Microsystems, Nussloch GmbH, Germany). Sections were collected at the level of the hippocampus (bregma −1.40 to −2.70) and thaw-mounted onto gelatinized microscope slides. The slides were dried, dehydrated, cleared with xylene and mounted with distyrene plasticized xylene (DPX). Golgi-treated brain tissue was analysed with a light microscope with an oil immersion lens (63x, NA 1.3; Leica DM4000B, Leica Microsystems; with a further magnification from the projection lens to the camera (10x) used for all measurements, providing a total magnification of 630 times). The whole neuron was traced manually.

Five neurons from each animal (n = 3–5 animals/group) were randomly selected for hippocampal spine analysis. Neurons were selected only if; they were clearly identified as being from the CA1 subfield, the neurons appeared completely filled and they were far enough from the neighbouring Golgi stained cells to be individually identified. Tertiary or greater order apical and basal dendrites (n = 5 apical and n = 5 basal) were selected and analysed to determine spine density, dendritic length and dendritic number (a cartoon outlining this method is shown in Figure S1). Thus, 50 dendrites were sampled from each animal. The standard deviations indicate the uniformity of the analyses and demonstrate that this is a representative sampling.

Biochemical analyses

Western blot, as previously described [14], was used for the quantitation of protein levels in individual hippocampi taken from a subset of animals (n = 5–7 mice/group). The antibodies utilized are outlined in Table S1.

Cell culture

Diaminosarcophagine (1,8-diamino-3,6,10,13,16,19 hexaazabicyclo[6.6.6]icosane), abbreviated to (NH2)2sar and given the trivial name “diamsar”, was synthesised as previously reported [15].

For biochemical studies, SH-SY5Y cells were utilised (in DMEM+10% serum), cells harvested after one hour exposure to the various treatments and then western blots performed (n = 6 replicates for each treatment).

Neurite outgrowth studies, which were conducted separately from the biochemical studies, utilised PC12 cells (grown in DMEM+10% fetal bovine serum) treated with NGF at 50 ng/ml in serum-free DMEM for 48 h. DMSO vehicle, 150 nM PBT2, 150 nM CuCl2, 150 nM ZnCl2, 150 nM PBT2-Zn (equimolar PBT2 and ZnCl2), 5 µM diamsar, 150 nM PBT2 plus 5 µM diamsar or 150 nM PBT2-Zn plus 5 µM diamsar were then added for 24 h. Treated cells were photographed and Image J software (NIH) used to determine the proportion of cells that had neurites 2x or longer than the width of the cell body, and to obtain a mean neurite length (n≥6 images from 3 cultures for each treatment).

Statistics

All analyses were performed by an operator that was blinded to the experimental conditions. An ANOVA followed by a post-hoc Tukey's multiple comparison test was performed for all comparisons (GraphPad Prism Version 5.0b).

Results

Initial assessment of 4 month old wildtype animals did not reveal any significant effect of PBT2 on dendritic spine density (apical  = −7.5% and basal  = −1.8%; both compared to sham-treated wildtypes), suggesting that there was no generic effect of PBT2 on this endpoint. This is consistent with our previous report, where there was no effect of PBT2 on the performance of wildtype animals in the Morris water maze [10].

PBT2 increases hippocampal apical spine density in Tg2576 mice