Amyloid β oligomers elicit mitochondrial transport defects and fragmentation in a time-dependent and pathway-specific manner
Small oligomeric forms of amyloid-β (Aβ) are believed to be the culprit for declined brain functions in AD in part through their impairment of neuronal trafficking and synaptic functions. However, the precise cellular actions of Aβ oligomers and underlying mechanisms in neurons remain to be fully defined. Previous studies have identified mitochondria as a major target of Aβ toxicity contributing to early cognitive decline and memory loss in neurodegenerative diseases including Alzheimer’s disease (AD). In this study, we report that Aβ oligomers acutely elicit distinct effects on the transport and integrity of mitochondria. We found that acute exposure of hippocampal neurons to Aβ oligomers from either synthetic peptides or AD brain homogenates selectively impaired fast transport of mitochondria without affecting the movement of late endosomes and lysosomes. Extended exposure of hipoocampal neurons to Aβ oligomers was found to result in mitochondrial fragmentation. While both mitochondrial effects induced by Aβ oligomers can be abolished by the inhibition of GSK3β, they appear to be independent from each other. Aβ oligomers impaired mitochondrial transport through HDAC6 activation whereas the fragmentation involved the GTPase Drp-1. These results show that Aβ oligomers can acutely disrupt mitochondrial transport and integrity in a time-dependent and pathway-specific manner. These findings thus provide new insights into Aβ-induced mitochondrial defects that may contribute to neuronal dysfunction and AD pathogenesis.
KeywordsAlzheimer’s disease Transport Fragmentation Hippocampus HDAC6 Drp-1 GSK3β
Charge coupled device
Glycogen synthase kinase 3β
Alzheimer’s disease (AD) is a progressive neurodegenerative brain disorder that is characterized by two hallmarks: intracellular neurofibrillary tangles and extracellular amyloid-β (Aβ) plaques in the regions of the brain that are responsible for learning and memory. Aβ, peptides of 39–43 amino acids produced by the sequential cleavage of β- and r-secretase at the C terminal of amyloid precursor protein, are believed to play a central role in the neurodegeneration and subsequent cognitive abnormalities of AD brains . Recent evidence indicates that small oligomeric species of Aβ are responsible for impaired brain functions in the early stages of AD, in part through their detrimental actions on neuronal trafficking and synaptic functions [2, 3, 4, 5]. Increasing evidence suggests that mitochondrial abnormalities play a critical role in AD pathogenesis. However, our understanding of various Aβ effects on mitochondria and their contribution to AD brain's dysfunction and disease progression remains incomplete.
Mitochondria are a cell’s vital organelle of energy production and they participate in diverse cellular functions ranging from Ca2+ homeostasis to cell apoptosis. Mitochondria are mainly produced in the soma and transported to the specific subcellular regions via microtubule (MT)-based fast transport [6, 7]. In neurons, mitochondria are seen accumulated at the pre- and post-synaptic sites where they function in synaptic transmission and plasticity [8, 9, 10, 11, 12]. Such subcellular localization of mitochondria involves spatiotemporal regulation of MT-based motors as well as interactions with other cytoskeletal structures such as the actin cytoskeleton and intermediate filaments . The half-life of mitochondria is about 1 month and mitochondrial fusion and fission provide a mechanism for mitochondrial renewal and degradation [10, 13, 14]. Therefore, timely and proper delivery of new mitochondria from the cell body is important to support the health and function of mitochondria in neurons.
Disruption of mitochondrial transport, structure, and function is believed to be a major contributor to many neurodegenerative diseases [10, 15]. Interestingly, disrupted axonal transport has been observed prior to the major pathological hallmarks such as senile plaques in AD [16, 17]. In many of these cases, axons exhibit swellings and varicosities with accumulated mitochondria and other organelles. On the other hand, damaged mitochondria dynamics in dendrite has been associated with the synaptic deficit in AD [18, 19]. Mitochondria that are stuck in the swellings/varicosities often look unhealthy and appear to undergo degradation and fragmentation. The cellular mechanisms that lead to defected mitochondrial transport and structure remain to be fully defined. In this study, we utilized cultured hippocampal neurons to investigate the acute effects of small Aβ oligomers on mitochondria. In particular, we examined the effects of Aβ oligomers prepared from synthetic peptides and the human AD brain homogenates on mitochondrial transport and morphology. Our data show that Aβ oligomers generated two distinct effects on mitochondria that are temporally segregated: transport impairment preceded the fragmentation. We further investigated the underlying mechanisms that mediate the specific effects of small Aβ oligomers on mitochondria.
Hippocampal culture, transfection and live imaging
All the experiments involving the use of vertebrate animals were carried out in accordance to the NIH guideline for animal use and have been approved by the Institutional Animal Care and Use Committee of Emory University. Hippocampal neurons from embryonic day 18 rats were obtained as described previously . About ~200,000 cells were plated in a 35 mm glass bottom culture dish (Warner Instruments, Hamden, CT) that were coated with 100 μg/ml poly-D-lysine (Sigma, St. Louis, MO). Hippocampal neurons were cultured in the Neurobasal medium containing B27 and Glutamax (Invitrogen) and typically imaged after 3–7 days in culture. Transfection of hippocampal neurons was performed using the CalPhos Mammalian Transfection Kit (Clontech, Mountain View, CA), which was typically done one day before the imaging . The DNA constructs for transfection were prepared by plasmid maxi kit (Qiagen, Valencia, CA). Mito-DsRed and Mito-GFP were provided by Dr. Zheng Li at NIH/NIMH. All the labeling and imaging were carried out in Krebs’–Ringer’s buffer (KRB, in mM: 150 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 10 glucose, and 10 HEPES, pH 7.4) [19, 21, 22]. For simultaneous imaging of mitochondria and lysosome trafficking, neurons were sequentially labeled by MitoTracker Green (20 nM for 20 min; Invitrogen, Eugene, OR) and LysoTracker Red (100 nM for 30 min; Invitrogen), followed by three rinses and additional 10 min incubation at 37 °C before imaging.
Dual-channel fluorescent time-lapse recordings were performed on an inverted microscope (TE2000, Nikon) equipped with a multispectral imager (Dual-View; Optical Insights) using either a 40× with 1.3 numeric aperture (NA) S Fluor oil immersion objective or a 60× 1.4 NA Plan Apo oil immersion objective with identical settings between the control and experimental groups. Time-lapse images were captured with a charge-coupled device (CCD) camera (SensiCam QE, Cooke Scientific) using the IPLab imaging software (BD Biosciences). We typically recorded neurons at a sampling rate of one frame every 5 s for 5 min, with the CCD exposure at 50 ms and 2x2 binning. For each experiment, a population of neurons was imaged for a 5 min control period before the application of Aβ molecules, followed by another 5 min time-lapse recording after Aβ exposure for 30 min or 2 h. All of the experiments were performed on the microscope stage with the 35 mm dish housed in a temperature controlled chamber (Warner Instruments, New Haven, CT) with the temperature set at ~35 °C. For different inhibitors, we applied the inhibitor 20 min before the onset of Aβ exposure. Tubacin was provided by Dr. Stuart Schreiber at Broad Institute of Harvard and MIT through the support by the Initiative for Chemical Genetics, National Cancer Institute. SB415286 and SB216763 were purchased from Tocris Bioscience (Ellisville, MO), Trichostatin A (TSA) was from Sigma (St. Louis, MO), and MS-275 was from Selleckchem (Houston, TX).
We quantified the Aβ effects on organelle movement by determining the percentage of moving mitochondria before and after Aβ exposure [19, 22]. Here, quantification was done by repeatedly playing back the time-lapse sequences and counting the number of moving mitochondria or other organelles (moved >3 μm) in each 5 min time-lapse sequence. We normalized the number of moving mitochondria or other organelles in the 5-min sequence against that before the Aβ application. A value of 100 % indicates that same numbers of moving mitochondria or vesicles were observed in both recording periods [19, 22]. To generate the movement traces of mitochondria and other organelles, ImageJ (National Institutes of Health, Bethesda, MD) was used to first process the image sequence using the Zprojection function (maximum intensity), followed by division against the first frame to produce the final image of the moving traces . The movement traces were also used to confirm the number of moving mitochondria over the 5 min time-lapse. Kymograph analysis was done using ImageJ with the multiple kymograph plugin written by J. Rietdorf (FMI Basel) and A. Seitz (EMBL Heidelberg). Here, Zprojection of the time-lapse sequence was used to identify the moving organelles and segmented line regions of interest (ROIs) were created from the movement traces. These segmented line ROIs were then applied to the time-lapse sequence to produce kymographs that allow the measurement of the average speed of each moving organelle.
Aβ oligomeric preparation and treatment
Aβ oligomers were prepared from either synthetic Aβ1-42 (American Peptide Company Inc, Sunnydale,CA) or recombinant Aβ1-42 (r-Peptide Company, Bogart, GA) according to the procedure described previously . In brief, Aβ1-42 was dissolved in hexafluoro-2-propanol (HFIP) and aliquoted to microcentrifuge tubes. HFIP was subsequently removed by evaporation in a speed-vacuum and desiccated Aβ aliquots were stored at -20 oC. To make Aβ oligomers, Aβ1-42 was dissolved in DMSO to make a 5 mM stock solution. The stock solution was then diluted to 100 μM with KRB and kept at 4 oC for 24 h before use. Bath application of Aβ was achieved through a two-step dilution procedure. First, the Aβ stock solution was diluted in KRB to twice the designated concentration (2× working stock). The 2× working stock solution was warmed to ~35 oC and then gently added to and mixed with the bath saline of the cells in an equal volume to reach the desired final concentration. In a typical experiment, 1 ml of the 2× stock solution was added to 1 ml of the bath solution in the culture/imaging dish on the microscope stage.
Postmortem human brain tissues were obtained from the tissue bank of the Emory Alzheimer’s Disease Research Center and the Emory Center for Neurodegenerative Disease. Frozen tissues from the frontal cortex of age- and gender-matched control (n = 3) and AD (n = 3) subjects were provided. Control subjects had no clinical history of neurological disease and were free of neurodegenerative disease pathology at autopsy. AD subjects met both CERAD  and NIA-Reagan  criteria for the neuropathologic diagnosis of AD. The frozen tissues were weighted and homogenized in ice-cold phosphate-buffered saline at 20 g/100 ml using a Dounce homogenizer (A size pestle for 15 strokes, B size pestle for 15 strokes). Homogenate was then sonicated 3 times for 5 s (power around 4), vortexed, and centrifuged for 5 min at 3000 g. The supernatant was collected, aliquoted in eppendorf tubes, and stored at -80 oC. Before application to the cells, each aliquot was thawed and centrifuged at 10,000 rpm for 5 min to remove any aggregates and precipitates. Similar to the application of Aβ oligomers from synthetic peptides, we applied the soluble human brain homogenate through a two-step dilution procedure: the 2× working stock and then 1:1 (volume:volume) to the cells.
Silver staining, western blotting and immunoprecipitation
Aβ samples (200 ng each) from American Peptide Company or r-Peptide Company were loaded to sample buffer with 50 mM DTT and heated at 85 °C for 5 min. Samples were loaded and fractioned by SDS-PAGE on 10–20 % Tris-Tricine gel (Invitrogen) for 90 min. After electrophoresis, the gel was briefly rinsed with ultrapure water and subjected to silver staining using SilverXpress® Silver Staining Kit (Invitrogen). For western blotting, different Aβ samples from synthetic peptides or human brain tissues were loaded in equal volume and fractioned by SDS-PAGE using 10–20 % Tris-Tricine gel and subsequently transferred to nitrocellulose membrane. The membrane was boiled for 10 min in PBS and blocked with 5 % non-fat dry milk in TBS with 0.05 % Tween-20 for 1 h at room temperature. The membrane was then incubated with a mixture of two anti-Aβ antibodies: 6E10 (1:1000) and 4G8 (1:1000) antibodies (Signet, Dedham, MA) in blocking buffer overnight at 4 °C. Bound antibodies were detected by HRP-conjugated secondary antibody, visualized by chemiluminescence using ECL (Thermo Scientific, Rockford, IL), and quantified using the gel analysis routine of ImageJ software (NIH).
To deplete Aβ molecules from the soluble human brain homogenate, we performed immunoprecipitation (IP) using Protein A-agarose beads (Santa Cruz). Here, the homogenate samples were incubated with 10 μl (2 μg) polyclonal anti-Aβ antibody A8326 (Sigma) for 1 h at 4 °C, followed by incubation with Protein A-agarose beads at 4 °C overnight on a rocker platform. The samples were subjected to centrifugation at 2500 rpm for 5 min at 4 °C to allow the collection of the supernatant (referred to as Thru 1st) and pellet (referred to as Output 1st) respectively. Parts of the supernatant from the first round IP (Thru 1st) was used to test the effects on organelle trafficking. Another round of IP was done with the remaining supernatant to obtain the second round IP supernatant (Thru 2st) and pellet (Output 2st). All these samples after IP were examined by western blot as described above.
Aβ oligomers rapidly impair mitochondrial transport
To better depict the effects of Aβ-O on organelle transport, we normalized the percentage of moving organelles after different treatments to that before the treatments. As shown in Fig. 1c, the inhibitory effect of Aβ-O on mitochondrial transport was rapid (e.g., 30 min) and dose-dependent. Significant impairment of mitochondrial transport was observed at 0.04 and 1 μM total Aβ concentrations (corresponding to 4 nM and 100 nM of Aβ oligomers) for 2 h and 30 min exposure, respectively (p < 0.005 comparing to that before Aβ-O application, Student’s t-test). Higher doses of Aβ-O were found to cause severe impairment of mitochondrial transport, whereas the transport of endo/lysosomes labeled by LysoTracker remained unaffected. Importantly, no effect was observed for the reverse peptide Aβ42-1 prepared the same way. Consistent with previous findings [22, 26], the impairment of mitochondrial transport by Aβ-O was abolished by inhibition of GSK3β using 1 mM LiCl (Additional file 1: Figure S1).
In addition to reducing the number of moving mitochondria, Aβ-O also slowed down the moving mitochondria. Kymograph analysis showed that the average speed of moving mitochondria was reduced to about half of that of the control period (Fig. 1d). It should be noted that all of our experiments were performed on relatively high density hippocampal cultures that contain mixed populations of highly elaborated and branched axons and dendrites. As a result, it is nearly impossible to distinguish between the MT plus end-directed anterograde and minus end-directed retrograde movements in these processes. The speed reported here represents the average of mitochondrial transport in both directions and from both axonal and dendritic processes. Consistently, Aβ-O did not significantly affect the speed of endo/lysosomes movement. We found that the average speed of endo/lysosomes was 0.20 ± 0.16 μm/s (Mean ± SD) before and 0.21 ± 0.20 (Mean ± SD) μm/s after Aβ–O treatment. Together, these results indicate that transport impairment of Aβ-O on mitochondria was not a result of non-specific and gross disruption of the transport machinery, and may involve specific signaling pathways that target mitochondria.
Extended exposure to Aβ-O leads to mitochondrial fragmentation
Distinct Aβ effects on mitochondria are mediated by specific pathways
We next investigated how Aβ oligomers elicit transport impairment and fragmentation of mitochondria, two effects with distinct time courses and dose-dependence. We first examined if prolonged Aβ-O exposure may disrupt the mitochondrial membrane potential leading to transport impairment and fragmentation. Live cell imaging using the TMRE dye [27, 28] showed that 2 h exposure to 5 μM (total concentration) Aβ-O did not affect the TMRE signal (Additional file 1: Figure S4). We next examined if Aβ oligomers caused cell death, which in turn resulted in mitochondrial transport defects and fragmentation. The cell viability assay  showed that cell death only started to be observed after 6 h exposure to Aβ-O (Additional file 1: Figure S5a). Similar results were also obtained using a different method in which a Hoechst dye was used to identify apoptotic nuclei (Additional file 1: Figure S5b). The very late onset of cell death upon Aβ-O exposure argues against cell death as the cause of the acute mitochondria defects induced by Aβ oligomers.
Stalled and deformed mitochondria in clusters are one of the outstanding signs of many neurodegenerative diseases [35, 36]. In AD, disruption in mitochondrial transport and function contributes to the disease pathogenesis and brain dysfunction [37, 38, 39]. While Aβ plaques are a hallmark of degenerating AD brains, soluble Aβ oligomers are believed to be the culprit underlying a variety of toxic effects on neuronal functions leading to the cognitive decline of the AD brain. While the adverse effects of Aβ oligomers on neurons and synapses are diverse and involve different targets, mitochondria are one of the major targets that Aβ oligomers negatively impact. In this study, we present evidence that Aβ oligomers elicit two acute effects on mitochondria: impairment of fast transport and fragmentation. Importantly we show that these two acute effects of Aβ oligomers involve different exposure times and doses of Aβ oligomer and are mediated by distinct signaling pathways. The selective impairment of mitochondrial transport is rapidly induced by low concentrations of by Aβ oligomers, whereas the fragmentation requires longer exposure and higher doses. While both Aβ effects on mitochondria can be blocked by inhibition of GSK3β, the transport impairment requires the activation of HDAC6 and fragmentation involves the GTPase drp1. Given that fast transport and fusion/fission of mitochondria are crucial for their function at subcellular locations, impairment of mitochondria transport and disruption of fusion/fission balance could argument the local function of mitochondria, such as at synapses.
Acute impairment of mitochondrial transport by Aβ molecules has been reported previously [22, 40]. This study has not only substantiated previous findings, but also provided additional insights. First, our findings show that Aβ oligomers selectively impair mitochondrial transport without affecting that of lysosomes and endosomes. Second, we present the evidence that Aβ oligomers from human AD brains exert the similar impairment of mitochondria transport to that of synthetic Aβ oligomers, indicating that this effect could potentially occur in human brains. Finally, our data suggest that Aβ oligomers negatively impact mitochondrial transport more effectively in dendrites than in axons. Mitochondria in axonal and dendritic compartments display different morphology and motility that may reflect their difference in functions in these two compartments . The long and slender axons of neurons contain a limited space that is prone to road blocks and traffic jams for long range transport, especially under neurodegenerative conditions . Dendrites, on the other hand, are shorter and thicker but form more elaborated arbors that function in receiving and integrating hundreds of synaptic inputs. Trafficking of synaptic receptors, mitochondria and other organelles, and mRNA-ribosomal complex occur actively in dendritic compartments and play a crucial role in postsynaptic functions [6, 10, 42, 43]. Our finding that mitochondrial transport in dendrites appears to be more severely inhibited by Aβ oligomers is consistent with the emerging view of Aβ adverse effects on postsynaptic structure and function [19, 44, 45]. Disrupted mitochondrial transport in dendrites could negatively impact the distribution and health of mitochondria to affect a number of mitochondria-dependent postsynaptic events, such as Ca2+ signaling, receptor trafficking, spine structure and plasticity, leading to synaptic dysfunction and degeneration .
The signaling mechanisms underlying the acute effects of Aβ oligomers on mitochondria remain to be elucidated. While inhibition of GSK3β appears to alleviate both Aβ effects, the transport impairment and fragmentation of mitochondria induced by Aβ oligomers are selectively abolished by the inhibition of HDAC6 and Drp1, respectively. HDAC6 is a member of class II HDACs that is localized mostly in the cytoplasm with a preference for non-histone proteins . Increased HDAC6 level in cortex and hippocampus of AD brain has been reported . HDAC6 has been found to deacetylate multiple non-histone proteins  such as α-tubulin [46, 49], tau , HSP90  and cortactin . The lysine 40 (K40) of α-tubulin is posttranslationally modified through acetylation in cells and HDAC6 is known to be the main deacetylase for this site, but this posttranslational modification does not appear to affect microtubule motor-based transport [52, 53, 54]. Therefore, other potential targets of HDAC6, such as tau , need to be investigated for their role in Aβ impairment of mitochondrial transport.
We would like to thank Drs. Jason Fritz and Marla Gearing at Emory CND for their help with human tissue samples. We also thank Dr. Stuart L. Schreiber from Howard Hughes Medical Institute for providing us the specific HDAC6 inhibitor tubacin.
This research was supported in part by grants from the National Institutes of Health (GM083889 and MH104632), a pilot grant from Emory Alzheimer’s Disease Resource Center (ADRC P50 AG025688), a NINDS core facilities grant (P30NS055077) to the Neuronal Imaging Core of Emory Neuroscience, and a postdoctoral fellowship from the Ellison Medical Foundation/American Federation for Aging Research to YR.
Availability of data and materials
YR designed and performed all the experiments and drafted the manuscript. JQZ oversaw and guided the study and wrote the manuscript together with YR. Both authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
Ethics approval and consent to participate
All the experiments involving the use of vertebrate animals were carried out in accordance to the NIH guideline for animal use and have been approved by the Institutional Animal Care and Use Committee of Emory University. No human subjects were included in this study.
- 18.Du H, Guo L, Yan S, Sosunov AA, McKhann GM, et al. Early deficits in synaptic mitochondria in an Alzheimer’s disease mouse model. Proc Natl Acad Sci U S A. 2010;107(43):18670-18675.Google Scholar
- 25.Hyman BT, Trojanowski JQ. Consensus recommendations for the postmortem diagnosis of Alzheimer disease from the National Institute on Aging and the Reagan Institute Working Group on diagnostic criteria for the neuropathological assessment of Alzheimer disease. J Neuropathol Exp Neurol. 1997;56:1095–7.CrossRefPubMedGoogle Scholar
- 26.Decker H, Lo KY, Unger SM, Ferreira ST, Silverman MA. Amyloid-beta peptide oligomers disrupt axonal transport through an NMDA receptor-dependent mechanism that is mediated by glycogen synthase kinase 3 beta in primary cultured hippocampal neurons. J Neurosci. 2010;30:9166–71.CrossRefPubMedGoogle Scholar
- 30.Simoes-Pires C, Zwick V, Nurisso A, Schenker E, Carrupt PA, et al. HDAC6 as a target for neurodegenerative diseases: what makes it different from the other HDACs? Mol Neurodegener. 2013;8:7. doi: 10.1186/1750-1326-8-7.
- 32.Chen SG, Owens GC, Makarenkova H, Edelman DB. HDAC6 regulates mitochondrial transport in hippocampal neurons. PLoS One. 2010;5(5):e10848. doi: 10.1371/journal.pone.0010848.
- 40.Decker H, Lo KY, Unger SM, Ferreira ST, Silverman MA. Amyloid-beta peptide oligomers disrupt axonal transport through an NMDA receptor-dependent mechanism that is mediated by glycogen synthase kinase 3beta in primary cultured hippocampal neurons. J Neurosci. 2010;30:9166–71.CrossRefPubMedGoogle Scholar
- 49.Ran J, Yang YF, Li DW, Liu M, Zhou J. Deacetylation of alpha-tubulin and cortactin is required for HDAC6 to trigger ciliary disassembly. Sci Rep. 2015;5:12917. doi: 10.1038/srep12917.
- 53.Walter WJ, Beranek V, Fischermeier E, Diez S. Tubulin acetylation alone does not affect kinesin-1 velocity and run length in vitro. PLoS One. 2012;7(8):e42218. doi: 10.1371/journal.pone.0042218.
- 54.Kalebic N, Sorrentino S, Perlas E, Bolasco G, Martinez C, et al. alpha TAT1 is the major alpha-tubulin acetyltransferase in mice. Nat Commun. 2013;4:1962. doi: 10.1038/ncomms2962.
- 55.Cook C, Stankowski JN, Carlomagno Y, Stetler C, Petrucelli L. Acetylation: a new key to unlock tau’s role in neurodegeneration. Alzheimers Res Ther. 2014;6(3):29. doi: 10.1186/alzrt259.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.