This article is originally compiled by brain science world team of brainnews, please contact permission to reprint.
Homeostasis of neural activity is thought to be a mechanism that may promote Alzheimer's disease (AD) lesions.
It is well known that 10-20 years before the clinical manifestations of AD (impaired cognitive function, etc.), the deposition of amyloid (Aβ) has occurred, and during this long asymptomatic period, whether the homeostasis mechanism of neural activity has been dysfunctional is not yet known.
The team led by Professor Inna Slutsky conducted a series of studies on this problem and found that in AD model mice in the awake state before the symptoms of cognitive impairment, the hippocampal neural circuit was not damaged, but during non-RAPID EYE movement sleep (NREM) and in the general anesthesia state (neuronal activity decreased), the neural circuit homeostasis of the hippocampus was dysfunctional.
The team of professors published an article titled "Disrupted neural correlates of anesthesia and sleep reveal early circuit dysfunctions in Alzheimer models" in Cell Reports magazine in January 2022.
1.APP/PS1 mice had normal activity of hippocampal CA1 neurons in the awake state
The mice used in the experiment were 4-5 months old APP/PS1 gene mutation mice, which showed significant amyloid deposition (Aβ40, Aβ42 and Aβ42 to Aβ40 ratio increased) at the same time, their cognitive function was not significantly impaired (spatial memory, condition-dependent fear memory was not impaired, but at 9 months of age).
The experiment used a wide-field head-mounted fluorescence miniature microscope to track the dynamic changes of Ca2+ at the level of a single neuron quickly and for a long time to show neuronal activity (Figure 1A). The results showed that the Ca2+ activity rate of HIPC CA1 neuron cells in the awake state was not significantly different from that of wild-type (WT) mice (Figures 1B and C), and there were no significant differences in the mean Ca2+ activity rate (mCaR) (Figure 1D), the number of active neurons (Figure 1E), and the total activity of neurons (Figure 1F).
At the same time, the experiment also employs long-term implanted four-stage rods to directly record the firing activity of individual neurons in mice. The results also showed that the average rate of discharge (MFR) of HIPPOC CA1 neuron cells in the awake state of APP/PS1 mice was not significantly different from that of WT mice (Figure 1G-I).
In addition, in recorded observations of field excitatory postsynaptic potentials outside cells, synaptic transmission and short-term synaptic plasticity of HIPPOC CA3-CA1 in awake states were found to be insignificantly different from those of WT mice (Figure J-K).
Figure 1.APP/PS1 mice were not impaired in CA1 neural network activity and CA3-CA1 synaptic transmission in the awake state
2.APP/PS1 mice developed hippocampal CA1 neuronal MFR abnormalities during non-REM sleep
Normally, during NREM, the activity of hippocampal CA1 neurons is down-regulated (this is known as a local homeostatic mechanism). It can be seen that the overall hippocampal CA1 neuronal activity of the WT mice in this experiment decreased by about 60% during NREM (Figures 2A-D, I), but during the NREM period of APP/PS1 mice, the activity of hippocampal CA1 neurons did not change significantly with during waking (Figures E-H, J).
Therefore, it can be concluded that a typical negative regulation of HIPC CA1 neurons in APP/PS1 mice during NREM has a significant dysfunction (Figure 2K), indicating that the steady-state regulatory function of CA1 neuron MFR is impaired.
Figure 2. APP/PS1 mice experienced abnormalities in negative regulatory function of hippocampal CA1 neurons during non-rapid eye movement sleep (NREM).
The abnormal local firing rate of HIPC CA1 neurons during NREM in 3.APP/PS1 mice predates the overall slow-wave oscillation abnormality
In WT mice, MFR decreases during NREM than in the awake state (Figures 3A, C), where the part of the cell below the median neuronal MFR value (1.4HZ) in the waking state does not show a significant change in activity during the NREM period, while the part of the cell above the median value shows a significant decrease in activity (Figure 3D).
However, in APP/PS1 mice, it was found that the MFR did not change significantly in either the awake state or the NREM state (Figures 3B, E), possibly because the activity of the part of the cell below the median value of the neuronal MFR in the awake state increased during the NREM period, while the activity of the part of the cell that was originally above the median value did not change (Figure 3F).
Considering that the overall slow-wave oscillation abnormalities have previously occurred in AD patients and AD mouse models, in order to explore whether the above changes in neuronal activity are due to slow-wave oscillation abnormalities, the experiment detected the EEG of mice and found that there was no significant difference between the slow-wave EEG (Figure 3G) and the slow-wave of the local field potential (LFP) of app/PS1 mice and WT mice (Figure 3H). This indicates that the abnormal local firing rate of hippocampal CA1 neurons during APP/PS1 mouse NREM predates the overall slow-wave oscillation abnormality.
Figure 3.APP/PS1 the abnormality of local firing rate of hippocampal CA1 neurons during NREM in mice predates the overall slow-wave oscillation abnormality
4.APP/PS1 mice lose neuronal inhibition in general anesthesia
This experiment uses isoflurane (ISO) to induce the production of anesthesia. The study found that under the action of isoflurane, WT mice will experience significant inhibition of the CA1 neuronal cell population (Figures 4A, C), mainly due to a decrease in the number of active neuronal cells (Figure 4C).
However, there was no significant inhibition of the CA1 neuronal cell population in APP/PS1 mice (Figures 4B, D), which eventually resulted in no significant difference in overall neuronal activity under anesthesia and when awake (Figure E).
In addition, COMPARED WITH WT MICE, APP/PS1 MICE EXHIBITED SIGNIFICANT HYPERSYALIZATION, WITH GREATLY INCREASED DISCHARGE CELLS (FIGURE 4F) AND DISCHARGE IN NEURAL NETWORKS (FIGURE 4G). After the experimenters used other anesthetic drugs, the results of the appeal still appeared, indicating that the APP/PS1 mice lost the neuronal inhibition function that should have appeared in the general anesthesia state.
Figure 4. APP/PS1 mice lose neuronal inhibition in a state of general anesthesia
5. Different fAD mouse models exhibit hyperexcitability under anesthesia
The experiment used different transgenic AD mice to further confirm the above findings, i.e., neuronal inhibition dysfunction in APP/PS1 mice under anesthesia. The study found that the above phenomenon persisted not only in APP/PS1 mice, but also in 5xFAD mice, as well as APP-KI mice (Figures 5A-C). It is shown that this phenomenon is prevalent in AD model mice, rather than just a unique phenomenon in AD mice in a single model.
Figure 5. Different fAD mouse models exhibited hyperexcitability under anesthesia
6. FAD mutations lead to downregulated MFR steady-state imbalances
The experiment uses multi-electrode arrays (MEAs) of in vitro hippocampal neurons to detect the MFR steady-state mechanism. The study found that isoflurane use causes the MFR level of WT to be lowered, fluctuating in a new descending set-point (Figures 6A, C). However, also under the trial of isoflurane, the APP/PS1 mutation led to an immediate increase in the MFR to a higher level, which is equivalent to an increase in the calibration point (Figure 6B, D).
This suggests that compared with WT, app/PS1 mutations lead to significant hyperexcitability in hippocampal neurons (Figure 6E), but it does not impair the basic homeostatic regulation response of the MFR (Figure F-H).
Normally, MFR levels increase significantly after the use of γ-aminobutyric acid (GABA) receptor antagonists, but their regulatory mechanisms cause the level of MFR to slowly decrease after that, as is the case with WT hippocampal neurons (Figure 6I).
However, app/PS1 mutations lead to abnormal homeostasis of the process, and in hippocampal neurons with APP/PS1 mutations, MFR cannot be reduced to its original level, but is consistently maintained at a high level (Figure 6J-K).
Figure 6. In vitro hippocampal neuron experiments have found that fAD mutations lead to downregulated MMFR homeostasis imbalances
Inhibition of mitochondrial mitochondrial dihydrotactate dehydrogenase can reduce overexcitation of CA1 neurons under anesthesia
The mitochondrial DHODH enzyme can modulate the modulation of MFR. Triflomet (TERI) is an inhibitor of the DHODH enzyme, and this experiment injects TERI transcarrier (VEH) into the brain chamber of mice for experimentation. The study found that epileptic abnormal high-voltage radio waves in CA1 neurons of APP/PS1 mice under anesthesia can be suppressed by TREI (Figure 7C) and the rate of discharge is reduced (Figure 7D-F).
Figure 7. Telflamide (TERI) can reduce overexcitation of CA1 neurons in APP/PS1 mice under anesthesia
Conclusion
In the early asymptomatic phase of AD, the CA1 neural network has been abnormal during anesthesia or non-REM sleep in a low-excitability state, but has not changed significantly in the awake state. This hyperexcitability can be relieved by DHODH enzyme inhibitors that modulate the MFR.
bibliography
Zarhin et al., 2022, Cell Reports 38, 110268 January 18, 2022. https://doi.org/10.1016/j.celrep.2021.110268.
Compiled by: KK (Brainnews Creative Team)
Reviewer: Simon (Brainnews Editorial Board)