Soluble oligomers of amyloid Beta protein facilitate hippocampal long-term depression by disrupting neuronal glutamate uptake

Abstract

In Alzheimer's disease (AD), the impairment of declarative memory coincides with the accumulation of extracellular amyloid-beta protein (Abeta) and intraneuronal tau aggregates. Dementia severity correlates with decreased synapse density in hippocampus and cortex. Although numerous studies show that soluble Abeta oligomers inhibit hippocampal long-term potentiation, their role in long-term synaptic depression (LTD) remains unclear. Here, we report that soluble Abeta oligomers from several sources (synthetic, cell culture, human brain extracts) facilitated electrically evoked LTD in the CA1 region. Abeta-enhanced LTD was mediated by mGluR or NMDAR activity. Both forms of LTD were prevented by an extracellular glutamate scavenger system. Abeta-facilitated LTD was mimicked by the glutamate reuptake inhibitor TBOA, including a shared dependence on extracellular calcium levels and activation of PP2B and GSK-3 signaling. In accord, synaptic glutamate uptake was significantly decreased by soluble Abeta. We conclude that soluble Abeta oligomers perturb synaptic plasticity by altering glutamate recycling at the synapse and promoting synapse depression.

Figure 1. Soluble Aβ facilitates long-term depression in the CA1 region of hippocampal slices

(A) Western blot of the 3 sources of soluble Aβ used for LTD experiments in this study. All 3 contain a SDS-stable band at ∼8 kDa, and this has been confirmed previously as an Aβ dimer by mass spectrometry (Shankar et al, 2008). 7PA2 CM also contains a higher, apparent trimer species. CHO- CM is devoid of these human Aβ species, as expected, and serves as a negative control throughout these studies. All samples were immunoprecipitated with polyclonal Aβ antibody AW8 and blotted with combined Aβ monoclonal antibodies 2G3 (Aβ₄₀) and 21F12 (Aβ₄₂). SEC: fractions rich in oligomers (left) or monomers (right) separated by SEC of 7PA2 CM (B) A train of 300 single pulses at 1 Hz (5 min; small grey bar) did not induce LTD in acute mouse hippocampal slices in the presence of CHO- CM (blue squares, n=6) but induced a significant LTD in the presence of 7PA2 CM (red circles, n=7). C-F A standard LTD protocol of 900 single pulses at 1 Hz (15 min; long grey bar) was applied to slices treated with 7PA2 CM (C), synthetic Aβ_1-42 (D), AD brain TBS extract (E), or SEC fractions from 7PA2 CM (F). All are not significantly different from their respective controls. Insets in B-F represent typical field excitatory postsynaptic potentials (fEPSPs) recorded before (light) and 50 min after (dark) LFS. Horizontal calibration bar, 10 ms; vertical bar, 0.5 mV.

Figure 2. Soluble Aβ enhances hippocampal LTD through mGluR or NMDAR, depending on the stimulation protocol

(A) LTD induced by the 300-pulse protocol (grey bar) in the presence of 7PA2 CM was blocked upon coadministration of the non-selective group I/II mGluR antagonist, MCPG (500 μM, red circles, n=6), but not the NMDAR antagonist, AP5 (50 μM, black squares, n=5). Horizontal colored bars represent the corresponding means ± SEMs from data shown in Fig.1B. (B) LTD induced by the 900-pulse protocol (grey bar) is independent of mGluR activation. Horizontal colored bars represent the corresponding means ± SEMs from data shown in Fig.1C. (C) The 900-pulse LTD induced in slices in CHO- CM was blocked by co-administering the NMDAR antagonist, AP5 (50 μM, black squares, n=8), whereas LTD in slices in 7PA2 CM was unaltered (red circles, n=6). Horizontal colored bars from data in Fig.1C. (D) 900-pulse LTD in Aβ monomer treated slices was blocked upon co-perfusing AP5 (50 μM, black squares, n=5), whereas the LTD in oligomer-treated slices was not (red circles, n=6). Horizontal colored bars represent the corresponding means ± SEMs from Fig.1F. (E) 7PA2 CM enhanced LTD was blocked by treatment with D-AP5 at 100 μM (n=5). Horizontal colored bars represent the corresponding means ± SEMs from Fig.2C. (F) Does-response curves of LTD blockade by AP5 in either CHO- CM (black squares) or 7PA2 CM (red circles). Insets in A-E represent typical fEPSPs recorded before (light) and 50 min after (dark) the LFS. Horizontal calibration bar, 10 ms; vertical bar, 0.5 mV.

Figure 3. Selective metabolism of extracellular glutamate prevents soluble Aβ facilitated-LTD

(A) A glutamate scavenger system (glutamic pyruvic transaminase [GPT, 5 unit/ml] + pyruvate [2 mM]) has no significant effect on the fEPSP baseline and 900-pulse LTD in CHO- or 7PA2 CM treated slices. Horizontal colored bars represent the corresponding means ± SEMs from Fig.1C. Insets in A-C represent typical fEPSPs recorded before (light) and 50 min after (dark) LFS; horizontal calibration bar, 10 ms; vertical bar, 0.5 mV. (B) 900-pulse LTD (grey bar) was blocked by AP5 (50 μM) in slices exposed to GPT + pyruvate for 15 min prior to 7PA2 CM (n=6). Red horizontal bar shows the LTD resistant to the same dose of AP5 in the absence of exposure to scavengers. (C) 300-pulse LTD (grey bar) was significantly prevented by co-administering glutamate scavengers with the 7PA2 CM. Horizontal colored bars represent the corresponding means ± SEMs from Fig.1B. (D) Paired-pulse facilitation in slices exposed to 7PA2 CM, CHO- CM or 7PA2 CM + AP5 (50 μM) measured before (white) or 30 min after (grey) adding these media or else measured 50 min after (black) a 900-pulse induction of LTD. At right are typical traces of these field recordings from the 7PA2 group. (E) Similar paired-pulse facilitation recorded by whole-cell voltage clamping at -70 mV in CA1 pyramid cells before and 30 min after 7PA2 CM exposure. At right are typical traces. Data are means ± SEMs.

Figure 4. Soluble Aβ oligomers alter NMDAR-mediated EPSCs in whole-cell voltage clamp recordings

(A1) Isolated AMPA-EPSC and NMDA-EPSC were recorded from -70 mV and +45 mV holding potentials, respectively, plus pharmacological blockers. Black traces, CHO- CM treated cells; grey traces, 7PA2 CM treated cells. (A2) Summary data for CHO- CM (black) and 7PA2 CM (grey) groups, and for pre-treatment with cyclothiazide (CTZ, 100 μM) prior to 7PA2 CM (white). (A3) NMDA/AMPA EPSC ratios are different in CHO- CM and 7PA2 CM. (B1) Typical traces from CA1 pyramidal cells held at −70 mV (black), isolated NMDA-EPSC in low Mg²⁺ with NBQX (10 μM) and bicuculline (20 μM) at -70 mV (grey), and full blockage of the NMDA current by AP5 (50 μM) (light grey). (B2) Summary data for the isolated NMDA-EPSC before (pre) and 20-30 min after 7PA2 CM exposure. Peak amplitudes (grey) and total charge transfers (black) are expressed as means ± SEMs. (C1) The selective NR2B inhibitor, ifenprodil (3 μM), modestly reduces NMDA charge transfers in control but markedly reduces it in 7PA2 CM. (C2) 7PA2 CM significantly increases the extrasynaptic response when synaptic NMDAR are first blocked by MK-801. (D) NMDA-mediated EPSC kinetic analysis: left panel, representative scaled traces before (black) vs. after (grey) 7PA2 CM exposure; right panel, summary data of rise time plotted vs. decay time before (pre) and after (post)7PA2 CM exposure.

Figure 5. Soluble Aβ-enhanced LTD involves impaired glutamate reuptake

(A) LTD was induced by 300 pulses (grey bar) when just the glutamate uptake inhibitor, TBOA (15 μM), was perfused (black squares, n=5). This TBOA-mediated LTD was prevented by mGluR antagonist MCPG (500 μM, blue triangles, n=6), but not by NMDAR antagonist AP5 (50 μM, red circles, n=6). Insets in A-F represent fEPSPs recorded before (light) and 50 min after (dark) either LFS (in A-C) or HFS (in E and F). Horizontal calibration bars, 10 ms; vertical bars, 0.5 mV. (B) A 900-pulse LTD facilitated by TBOA was resistant to a standard dose of AP5 (50 μM, red circles, n=6), but it could be blocked by D-AP5 (100 μM, blue triangles, n=6). (C) Saturation of 900-pulse LTD occurring in the presence of TBOA (black horizontal bar) occludes further LTD in the presence of 7PA2 CM (red horizontal bar). (D) Summary data from occlusion experiments as in C: left bar, evoked fEPSP was re-normalized to baseline values (i.e., point 3 in C) before addition of 7PA2 CM to the perfusate and an additional LFS (right bar). (E) LTP induced by high-frequency stimulation (HFS) was prevented in slices in 7PA2 CM (red circles) but unaffected in slices in CHO- CM (black squares). (F) HFS-induced LTP was similarly prevented by TBOA. (G) Dose-dependent inhibition of glutamate uptake by TBOA in hippocampal synaptosomes. (H) Pretreatment with soluble synthetic Aβ_1-42 impaired glutamate uptake in hippocampal synaptosomes in a fashion similar to TBOA treatment at 50 μM. Data are means ± SEMs as percentage of vehicle alone; **p<0.01.

Figure 6. LTD facilitated by soluble Aβ or TBOA share similar signaling pathways

(A) Conventional LTD (in CHO- CM, black squares) and Aβ-enhanced LTD (in 7PA2 CM, red circles) induced by the 900-pulse LFS in the presence of AP5 (50 μM) were plotted as a function of increasing extracellular calcium concentrations in the perfusate. Light blue area represents the LTD obtained in the CHO- CM alone (without AP5). (B) Ryanodine (20 μM) given 30 min prior to 900 pulses partially blocked the LTD in CHO CM (black squares, n=5) but produced no block of the 7PA2-facilitated LTD (red circles, n=6). (C) U73122 (10 μM) (applied 45 min prior to 900 pulses) fully blocked LTD in the presence of CHO- CM (black squares, n=6) but had no effect in the presence of 7PA2 CM (red circles, n=6). (D) 900-pulse LTD was blocked by p38 MAPK inhibitor, SB203580 (5 μM), in CHO- CM (black squares, n=7) but not in 7PA2 CM (red circles, n=7). (E) Calcineurin inhibitor, FK-506 (20 μM) perfused 40 min prior to 900 pulses prevented LTD in both CHO- CM (black squares, n=5) and 7PA2 CM (red circles, n=6). (F) GSK-3β inhibitor, SB415286 (10 μM) perfused 60 min prior to 900 pulses prevented LTD in both CHO- CM (black squares, n=6) and 7PA2 CM (red circles, n=7). (G) LTD enhanced by TBOA (15 μM) was likewise resistant to ryanodine (n=5). (H) TBOA-enhanced LTD was prevented by the calcineurin inhibitor (FK 506, black circles, n=5) but not by the p38 MAPK inhibitor (SB203580, red diamonds, n=5). (I) Summary data for actions of the signaling pathway modulators on Aβ- and TBOA-enhanced LTD. The horizontal gray bar represents mean LTD in TBOA alone (Fig. 5B), and the horizontal light orange bar represents mean LTD in 7PA2 CM alone (Fig. 1C).

Figure 7. Schematic of the principal pathways implicated by this study in conventional LTD (left) and in LTD facilitated by soluble Aβ oligomers (right)

Conventional LTD requires NMDAR-mediated influx of extracellular calcium and liberation of intracellular calcium stores. This ultimately activates PP2B, GSK-3β or p38 MAPK signaling pathways that induce LTD. Soluble Aβ oligomers lead to activation of more NMDAR, leading to extracellular calcium influx and activation of PP2B and GSK-3β pathways to facilitate LTD. Our data suggest that Aβ oligomers decrease glutamate uptake by neuronal transporters (red x's), resulting in the enhanced activation of NMDARs and thus facilitation of LTD-inducing pathways.