Preliminary analysis of hippocampus synaptic apoptosis and microglial phagocytosis induced by severe restraint stress

Abstract

Aim

Several studies reported stress-induced microglial phagocytosis, but the biochemical mechanisms by which stress alters microglial synaptic phagocytosis are not fully uncovered. Local or limited apoptosis without cell death was observed at neuronal synapses in previous studies, and proposed as an upstream mechanism for microglial synapse elimination. In this micro-report, we aimed to preliminary examine local synaptic apoptosis in the mouse hippocampus following severe restraint stress, and its effect on microglial phagocytosis.

Methods

Mice were exposed to 10-day water immersion restraint stress (WIRS). Brain sections including stratum lucidum in the hippocampal CA3 subfield were stained with antibodies against cleaved caspase 3, ionized calcium-binding adapter molecule1 (Iba1), lysosomal-associated membrane protein1 (LAMP1), vesicular glutamate transporter1 (VGLUT1). Co-localization among these proteins were calculated.

Results

Our image analysis revealed that synaptic apoptosis was increased while there were no significant changes in microglial phagocytic activity and synaptic phagocytosis following 10-day WIRS.

Conclusion

Severe restraint stress enhanced pre-synaptic apoptosis in mouse CA3 stratum lucidum region, but did not promote microglial phagocytosis. The phenomenon microglia fail to phagocytose weakened and unnecessary synapses may be related to pathology of stress.

INTRODUCTION

The phenomenon of synaptically localized apoptosis was observed in previous studies.^1–3 Apoptotic biochemical changes such as caspase activation, loss of membrane phospholipid asymmetry, and mitochondrial alteration occur in synapses without neuronal cell death. Synaptic transmission is reduced at these synapses, and some of them could be eliminated.⁴ Microglia, the major immunocompetent cells in the brain, are a strong candidate for pruning apoptotic synapses.⁵ Cleaved caspase 3, an executioner of apoptosis, exposes phosphatidylserine (PS) to the cell surface by inactivating scramblase and activating flippase in the cell membrane.^6,7 Györffy et al revealed the preferential co-localization between C1q, a component of immune complement pathway, and cleaved caspase3 in adult mouse hippocampal pre-synapses.⁸ Both PS and C1q are well known “eat me” signal for initiating microglial phagocytosis.^9–11 Therefore, it is likely to consider the elimination of apoptotic synapses by microglia via externalized PS or C1q deposition at synapses.

The pathological significance of synaptically apoptotic mechanisms leading to synaptic dysfunction or loss has been documented in schizophrenia and Alzheimer's disease,^2,12 but it has never been linked to stress-related mental disorders. Several preclinical studies showed stress-induced microglial phagocytosis accompanied by depressive behaviors or cognitive deficits,^13–16 but biochemical mechanism by which stress drives alteration in microglial synaptic phagocytosis remain unclear.

In this micro-report, we preliminary examined local synaptic apoptosis in the CA3 stratum lucidum region of the mouse hippocampus following 10-day water immersion restraint stress (WIRS). After we found 10-day WIRS increase synaptic apoptosis significantly, assuming it affects microglia, we examined microglial phagocytosis.

METHODS Animals

Male 8 weeks C57BL/6 mice were used for experiments.5 mice were housed per cage with food and water freely accessible in a temperature-controlled, circadian rhythm-adjusted room. All animal treatments were in accordance with the Guidelines for Animal Experiments at Kyushu University, and this study was approved by the Committee on the Ethics of Animal Experiments, Kyushu University.

Stress

Mice were exposed to the WIRS for 10 days. In the WIRS, mice were restrained in a 50 mL conical tube with several holes and the surplus space was packed with Kimtowel (Nippon Paper Crecia Co., Ltd., Japan).The conical tube was immersed in a temperature-controlled (23°C) water at chest height for 2 hours in a day. Mice were randomly assigned to a control group or 10-day WIRS group.

Immunohistochemistry

Mice were sacrificed 2 hours later after the last stress procedure. Mice were perfused with phosphate-buffered-saline (PBS) and then 4% paraformaldehyde (PFA). Brains were removed, post-fixed overnight in 4% PFA, and stored in 30% sucrose. 40 μm coronal sections containing dorsal hippocampus were cut on a cryostat (Leica CM1950, Wetzlar, Germany). Sections were blocked for 2 hours with 1% bovine serum albumin in PBS containing 0.3% Triton X-100 and 0.1% sodium azide at a room temperature. For evaluating local synaptic apoptosis, sections were double stained with rabbit anti-cleaved caspse-3 (1:300; Cell Signaling, #9664) and guinea pig anti- vesicular glutamate transporter1 (VGLUT1) antibody (1:2000; synaptic systems, #135304) for 6 days at 4°C. VGLUT1 is a marker for pre-synapses and expressed abundantly in the hippocampus. For phagocytosis assay, sections were triple stained with rabbit anti- ionized calcium-binding adapter molecule1 (Iba-1) antibody (1: 2000; Wako, #019–19 741) and rat anti- lysosomal-associated membrane protein1(LAMP1) antibody (1:100; BD Biosciences, #553792) and anti-VGLUT1 antibody above mentioned (1:2000; synaptic systems, #135304) overnight at 4°C. LAMP1 is a marker for lysosomes and the number of lysosomes in microglia is an indicator of their phagocytic capacity. After washing with PBS, sections were incubated with Alexa488-, Alexa568-, or Alexa647-conjugated secondary antibodies (1:300, Invitrogen) overnight at 4°C, then mounted with Vectashield (Vector Laboratories, Burlingame, CA).

Image analysis

Four brain sections per animal were used for apoptotic signal and phagocytosis assays respectively. Three to four confocal image stacks from CA3 stratum lucidum region were acquired per section with Zeiss LSM700 microscope. We analyzed parts where VGLUT1 is evenly stained along Z axis. In apoptotic signal assay, stacks were created using ×40 objective lens with 2.5× digital zoom, as 64.02 μm × 64.02 μm in XY, at 0.41 μm intervals along Z axis. Each stack was composed of 13 image slices. Images were converted, their backgrounds were subtracted, and binary probability images were generated for each stack using Keyence BZ-X800 image converter and analyzer. We set thresholds for each channel manually and kept them constant throughout. We calculated cleaved caspase3⁺ VGLUT1⁺ area / VGLUT1⁺ area for individual images and averaged them per stacks and individual animals to evaluate synaptic apoptosis. In phagocytosis assay, stacks were created using ×63 objective lens with 1.6× digital zoom, as 63.51 um × 63.51 um in XY, at 0.37 um intervals along Z axis. Each stack was composed of 14 image slices. After background subtraction for all channels, 3D reconstruction for Iba1 signals was performed using Imaris 7.0 software.¹⁷ LAMP1 signals within Iba1⁺ structures were picked up using the mask function in Imaris and transformed into 3D reconstruction. VGLUT1 signals in the 3D LAMP1 structures whose intensity value are greater than the specific threshold were picked up by Intensity mean filter in Imaris and transformed into 3D reconstruction. We calculated LAMP1⁺ Iba1⁺ volume / Iba1⁺ volume to evaluate microglial phagocytic activity, and LAMP1⁺ VGLUT1⁺ Iba1⁺ volume / Iba1⁺ volume to evaluate microglial synaptic phagocytosis for each stack and averaged them per individual animals. We exclude Iba1 structures located in blood vessels from our analysis. Thresholds for each channel were kept constant throughout.

Statistical analysis

For comparison of two groups, two-tailed unpaired t-tests were performed. P-values <0.05 were defined significant. We used R statistical packages.

RESULTS

Representative converted and binarized 2D images in control and stressed groups are shown in Figure 1A. 2D image analysis showed a significant increase of the ratio of the area where cleaved caspase3 co-localizes with VGLUT1 to the VGLUT1⁺ area following stress exposure (Figure 1B). Representative 3D reconstruction images in control and stressed groups are shown in Figure 2A. 3D image analysis showed no significant difference in the ratio of the volume where Iba1 colocalizes with LAMP1 to the volume of Iba1+ structures and the ratio of volume where VGLUT1 colocalizes with LAMP1 contained in Iba1+ structures to the volume of Iba1+ structures, between control and stressed groups (Figure 2B, C).

FIGURE 1. We acquired images from CA3 stratum lucidum region in control and stressed male C-57 BL6 mice using confocal laser scanning microscopy, subsequently for analysis converted and binarize them using image converter/analyzer. Scale bar, 5 μm. (A) representative converted (left side) and binarized (right side) images in control and stressed mice. (B) ratios of the area where VGLUT1 co-localizes with caspase 3 to the VGLUT1+ area.10-d water immersion restraint stress significantly increase the value (control: n = 4 animals, stress: n = 3 animals, t5 = 3.0265, P = .0292). Error bars, SD

FIGURE 2. (A) Representative 3D reconstruction images obtained from CA3 stratum lucidum region in control and stressed male C-57 BL6 mice. Iba1+ microglia and LAMP1+ structures are visualized at left side. Iba1+ microglia, LAMP1+ structures and VGLUT1+ structures are visualized at right side. Scale bar, 5 μm. (B) Ratios of the volume of the LAMP1+ structures contained in the Iba1+ microglia to the volume of the Iba1+ microglia. There was no significant difference between control and stressed groups in the ratio (control: n = 4 animals, stress: n = 3 animals, t5 = 0.8946, P = .412). Error bars, SD. (C) Ratios of the volume of engulfed VGLUT1 within LAMP1+ structures contained in the Iba1+ microglia to the volume of the Iba1+ microglia. There was no significant difference between control and stressed groups in the ratio (control: n = 4 animals, stress: n = 3 animals, t5 = 0.4414, P = .6773). Error bars, SD

DISCUSSION

In the present study, 10-day WIRS significantly increased synaptically localized apoptotic signals. It has been demonstrated that local synaptic apoptosis process drives synaptic failure.¹⁸ A causal relationship was found between reduced synaptic transmission and caspase 3 activation in synapses in an animal model of Alzheimer's disease.¹⁹ In Clq deposited apoptotic synapses, expressions of proteins involved in synaptic transmission were shown to be decreased.⁸ Microglia phagocytose inactive synapses with reduced transmission capacity.^20,21 Despite enhanced synaptic apoptosis which probably leads to PS exposure to the cell surface, C1q deposition in the synapses, and reduced synaptic transmission, microglial phagocytosis was not promoted in this study. Our results propose the possibility that the inability of microglia to phagocytize weakened and unnecessary synapses may participate in the pathology of stress related mental disorder. We speculate not only excessive synaptic phagocytosis, but also failure to remove dysfunctional synapses to a certain extent is also detrimental for homeostasis. In fact, most studies regard stress-induced microglial phagocytosis as detrimental, but a few studies noted its beneficial effect. Piirainen et al found increased CD206⁺ microglial synaptic phagocytosis by 10-day restraint stress is positively correlated strengthened spatial learning.²² In another study, restraint stress-induced depressive behavior was rescued by pharmacological treatment to enhance microglial phagocytosis.²³ The effects of stress on microglial phagocytosis are inconsistent. It may depend on types of stressors or brain regions. In this micro-report, stress induced synaptic apoptosis and microglial dysfunction were indicated. The distribution of cleaved caspase 3 expression in neuronal cells and its alterations with stress also should be addressed in the future, and further research is needed to understand the pathological significance of stress-induced alteration in microglial phagocytosis and its upstream biochemical mechanisms.

Limitation

This is a preliminary study because of low sample size. We used a VGLUT1 antibody in this study because, of the three isozymes (VGLUT1, VGLUT2, VGLUT3), VGLUT1 is mainly expressed in the hippocampus.²⁴ However, we need to note that the detected VGLUT1 contained in microglia do not entirely result from microglial engulfment because microglia express a low level of VGLUT1.²⁵ We hypothesize apoptotic synapses could be eliminated by microglia via externalized PS or C1q, but we did not evaluate co-localization of externalized PS or C1q and synaptic marker proteins in this study. Alteration in the exposure of PS to the cell surface and the level of Clq expression in the synaptic compartment following stress need to be examined. An ex vivo slices imaging study revealed microglia engulf and digest synaptic elements very quickly.²¹ Therefore, it is important to point out that the result in this study as to microglial synaptic phagocytosis may be highly depend on the timing of tissue sampling. Temporal dynamics of microglial synaptic phagocytosis following stress also need to be evaluated.

AUTHOR CONTRIBUTIONS

Takahiro A Kato was the principal investigator of the present research. Shingo Enomoto and Masahiro Ohgidani are the first authors. These two authors created the conception and design of the project and wrote the manuscript. Noriaki Sagata and Shogo Inamine are involved in the performance of experiments and data analyses/interpretation. All authors contributed substantially to the scientific process leading up to the writing of the present manuscript and approved this submission in its current form.

ACKNOWLEDGMENTS

We would like to thank members of the Research Support Center, Research Center for Human Disease Modeling, Kyushu university Graduate School of Medical Sciences for their technical assistance. We also thank Aya Yamada, Yuka Matsushita, Mayumi Tanaka and Miwa Irie for their technical assistance.

FUNDING INFORMATION

This study was supported in part by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (JSPS KAKENHI Grant Numbers: JP16H06403, JP18H04042, JP19K21591, JP20H01773, JP22H00494 and JP22H03000), and the Japan Agency for Medical Research and Development (AMED: JP18dk0307075, JP21wm0425010 & JP22dk0207065). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available in Appendices S1 and S2 of this article.

ANIMAL STUDIES

All animal experiments were approved by the Animal Care and Use Committee of Kyushu university.