Inhibiting mtDNA‐STING‐NLRP3/IL‐1β axis‐mediated neutrophil infiltration protects neurons in Alzheimer's disease

Abstract Neutrophil is a pathophysiological character in Alzheimer's disease. The pathogen for neutrophil activation in cerebral tissue is the accumulated amyloid protein. In our present study, neutrophils infiltrate into the cerebra in two models (transgenic model APP/PS1 and stereotactic injection model) and promote neuron apoptosis, releasing their cellular constituents, including mitochondria and mitochondrial DNA (mtDNA). We found that both Aβ1–42 and mtDNA could provoke neutrophil infiltration into the cerebra, and they had synergistic effects when they presented together. This neutrophillic neuroinflammation upregulates expressions of STING, NLRP3 and IL‐1β. These inflammatory cytokines with mtDNA constitute the mtDNA‐STING‐NLRP3/IL‐1β axis, which is the prerequisite for neutrophil infiltration. When any factor in this pathway is depleted, the migration of neutrophils into cerebral tissue is ceased, with neurons and cognitive function being protected. Thus, we provide a novel perspective to alleviate the progression of Alzheimer's disease.

In addition, when neuron cells disintegrated, numerous mitochondrial DNA (mtDNA) was released.mtDNA was evolutionarily homologous to bacteria DNA.7][28][29][30] In our previous studies, we found that mitochondria and mtDNA were important factors for inflammatory responses, with the capability to attract neutrophils to the lesion and exacerbate damage. 26In this research, we found that Aβ could induce impaired neurons activating the mtDNA-STING-NLRP3/IL-1β axis, initiate neutrophil infiltration in cerebra and induce central nervous system damage.With the knockout of relevant genes on this axis, the cerebral tissue was protected.Thereby, we revealed that in the pathogenesis course of Alzheimer's disease, amyloid protein acted as the initiator, and the activation of mtDNA-STING-NLRP3/IL-1β axis could induce subsequent damage to the nervous system.

| Mice and cell lines
C57BL/6 mice aged 6-8 weeks were obtained from Charles River.
Sting À/À knockout mice were purchased from The Jackson Laboratory, IL-1β À/À and Nlrp3 À/À knockout mice were provided by the University of Tokyo.The 8-month APP/PS1 double transgenic mice were purchased from Sairuike Biotechnology Co., Ltd (Wuhan), which had been identified as co-overexpressing the APP and PS1 genes.All animals were fed in specific pathogen-free (SPF) animal house and used strictly according to the treaty of the Sichuan University Animal Protection and Ethics Association.
Mice hippocampal neuron cell line HT22 was obtained from Procell Life Technology Co., Ltd (Wuhan).N2a and SK-N-BE cell lines were from the cell bank of Sichuan University.

| Mice tissue preparation and immunofluorescent assay
Mice were sacrificed for tissue harvest.Cut the sternum and ribs of the mice to expose the heart and perfused with ice-cold normal saline from the left ventricle.The collected samples were stored at À80 C until required for section or other experiments.Embedded mice brains with OCT and freeze at À20 C. Then sliced brain tissue with a freezing slicer with a thickness of 40 μm (LEICA, Germany).The frozen tissue sections were dried in the air for the staining process or stored at À20 C. Washed sections with PBS (phosphate buffered solution) for 5 min.1% Triton-100 drilled for 15 min.PBS washed three times, 5 min each time.5% BSA (bovine serum albumin) blocked for 30 min.
PBS washed three times, 5 min each time.Added primary antibody to cover the tissues (anti-Aβ antibody, 1:1000, Abcam, USA; anti-Ly6G antibody, 1:2000, Abcam, USA; anti-DNA/RNA damage 1:1000, Abcam, USA) at 4 C overnight.PBS washing three times, 5 min each time.Dripped fluorescent secondary antibodies (anti-rabbit FITC antibody, Abcam, USA; anti-rat TRITC antibody, Abcam, USA; anti-mouse FITC/TRITC antibody, Abcam, USA) onto the sections until covered completely, then incubated them in room temperature for 2 h.Used PBS to wash secondary antibodies and dripped DAPI.If primary antibodies are the same host species for a single slice, the procedure is to drip one sort of primary antibody and washed for one colour of the secondary antibody and washed this secondary antibody for another primary antibody and a different colour of secondary antibody.The experiment in vitro used mice hippocampal neurons HT22 cells, cultured with DMEM medium (Gibco, USA) on the 24-well climbing sheet (WHB scientific, China).After 24 h of administration, cells were fixed with polyformaldehyde solution.After that, cells were washed with PBS for 5 min.PBS solution with 1% Triton-100 drilled for 15 min.PBS washed three times.5% BSA sealed for 30 min.PBS washed three times.Diluted the primary antibody (anti-Caspase-3 antibody 1:1000, CST, USA) and dripped it onto a climbing sheet covering it completely and stayed at 4 C overnight.PBS washed three times and dripped fluorescent secondary (anti-rabbit FITC antibody, Abcam, USA) antibody and incubated at room temperature for 2 h.Used PBS to wash secondary antibodies and dripped DAPI.Then took photographs with the microscope (DM6B, LEICA, Germany) and randomly select five high power fields to count the total number of positive cells for statistics.

| Stereotactic injection
Mice were transferred into the isoflurane (RWD life technology.Ltd, China) anaesthesia box until they were completely anaesthetized.

| Flow cytometry
This experiment mainly used the apoptosis detection kit (BD Biosciences, USA), the brief steps were as follows: HT22 cells were cultured with DMEM medium on 24-well plates, when cells stayed in the logarithmic growth phase and added Aβ 1-42 into cultures.After 24 h, collecting cells and supernatant were in each well and centrifuged at 1000 rpm for

| Immunoblot
Collected the cells and tissues, then added RIPA (Beyotime, China) to cleavage cells and tissues.After that, centrifuged the samples at

| Image Quantification
Cerebral hippocampal sections of mice were sectioned at 40 μL thickness as described above.The analysis was performed by the image threshold adjust function of Image J 1.53e software (National Institutes of Health, USA).The brain regions of interest (Bregma À1.6 mm to À3.5 mm) with six sections per mouse were analysed.Nissl-stained (Beyotime, China) sections were photographed by microscope system (Axiolab 5, Zeiss, Germany).

| Neutrophil preparation and transwell assay
Collected mice neutrophils from their bone marrows.Isolated bone marrows from femur and tibia with sterile normal saline solution and gently repeatedly blew until there were no visible particles in the solution.Took a sterile 15 mL centrifuge tube, added Sigma cell separation solution (Sigma, USA) to the bottom and then gently dripped bone marrow solution onto it.Centrifuge at 800 g for 20 min to isolate neutrophils.More than 90% of the isolated cells were neutrophils as determined by flow cytometry.Washed neutrophils with RPMI-1640 medium (Gibco, USA) once and resuspended them in RPMI-1640 medium.Added the neutrophils (1 Â 10 6 /ml) evenly into the inner chamber of the transwell plate (Corning, USA), and added an appropriate amount of RPMI-1640 medium with Aβ 1-42 , mtDNA, and mixture of Aβ 1-42 /mtDNA into the outer chamber of the plate.After 24 h of incubation, collected the medium of the outer chambers, stained with haematoxylin and counted a number of neutrophil that migrated into the outer chamber.

| Barnes maze test
The Barnes maze used a round disc that is 91 cm in diameter and 90 cm above the floor (SA208, Jiangsu Science Biological Technology Co. Ltd.).All mice from the fourth day before the test were placed in the starting box in the centre of the maze for 3 min, and then the mice were gently nudged into the target hole with the shelter box for 3 min.Then, from the third day before the test, the mice were placed in the central starting box of the maze every day and then released.If the mice could not find the shelter within 90 s, gently nudged mice into it and stayed for 3 min.From the first day of the test, the mice were placed in the central starting box of the maze, and the starting box was placed in the centre of the maze, then the strong light and noise stimulation were turned on.Then, mice were released, and the latency of mice successfully getting into the shelter box was recorded (latency, the longest stimulation time was 90 s).Each mouse was tested three times a day for three consecutive days.

| Statistics
We used GraphPad Prism 8 to conduct statistical analysis of the data and created statistical charts.The results were expressed in mean ± standard error (SEM).The significant difference between samples was analysed by a t test method.

| Aβ induces neutrophil infiltration and neuron impairment
We performed a genetically modified animal experiment to test the role of neutrophils in Aβ presence condition.APP/PS1 mice were characterized by highly expressing Aβ.We adopted 8 month APP/PS1 mice as observation subject, which had heavy accumulation in cerebra and were controlled by the wild-type (WT) mice of the same age.
Ly6G is the characteristic marker of neutrophil.The results revealed that Ly6 G-positive cells were found in both cortex and hippocampus in APP/PS1 mice (Figure 1A).However, there was no significant infiltration of neutrophils in cerebra of WT mice, suggesting that neutrophils migration depended on the Aβ accumulation.
Next, we observed that a large amount of oxidative damaged DNA was presented in the cerebra of APP/PS1 mice.This damaged DNA was in the same place as the accumulated amyloid proteins (Figure 1B), suggesting that damaged DNA bound to Aβ to form the DNA-Aβ complex.
To confirm Aβ impaired neuron cell will release intracellular DNA, we conducted in vitro experiments.First, we used mouse hippocampal neurons HT22 cell as the experimental subject and added different concentrations of Aβ 1-42 .At lower concentrations, the toxicity of Aβ 1-42 was weak.With the increase in the concentration, the toxicity gradually enhanced.Besides, 2 μmol/L of Aβ 1-42 could induce significantly increased expression of caspase-3 in HT22 cells, which was a crucial marker for cellular apoptosis and necrosis (Figure 1C).
Flow cytometry experiments had similar results in immunofluorescence (Figure 1D).All these results indicated that 2 μmol/L Aβ 1-42 was the threshold concentration to induce toxicity for cells in vitro.
Second, we took this margin concentration as a condition to observe the release of mitochondria directly.To exclude accidental phenomena, we conducted the experiment by three cell lines.We found that all cell lines (HT22, N2a and SK-N-BE) released mitochondria when exposed to Aβ 1-42 (Figure 1E).Additionally, the concentrations of mitochondrial DNA (mtDNA) were increased with the higher concentration of Aβ 1-42 (Figure 1F).Thus, Aβ 1-42 could induce neutrophil infiltration, neuron apoptosis and necrosis.These impaired neurons released mitochondria and mtDNA.mtDNA could bind to amyloid proteins.

| Neutrophil neuroinflammation activates inflammatory cytokines expression
Next, we investigated the toxic effect of Aβ 1-42 on cerebra in vivo.
We injected fMLF, Aβ 1-42 , mtDNA and Aβ/mtDNA mixture into the cerebral tissue of wild-type C57BL/6 mice respectively.fMLF is a short peptide chain that is the chemotactic signal for neutrophil migration.In this experiment, fMLF was used as the positive control.We injected these subjects into the brain.We found that all these subjects could induce neutrophil infiltration (Figure 2A).Wherein, the number of neutrophils migrated by Aβ and mtDNA was basically even, while the number of neutrophils in Aβ/mtDNA mixture group was significantly higher than that in Aβ or mtDNA groups.Namely, Aβ and mtDNA could activate the migration of neutrophils into cerebral tissue.Besides they had synergistic effects in chemotactic function when they presented together.
To explore the toxicity of Aβ in vivo, we conducted TdT-mediated dUTP Nick-End Labeling (TUNEL) assay.When neurons apoptosis began, DNA in cells would be fragmented.The fragmented DNA could be detected by TdT-mediated dUTP.We observed that the injections of fMLF, Aβ 1-42 , mtDNA and Aβ/mtDNA mixture could induce apoptosis (Figure 2B).Similarly to the results of Ly6G cells infiltration, the number of TUNEL-positive cells in the Aβ/mtDNA mixture group was much higher than that in Aβ or mtDNA group.Interestingly, there was no evidence showing that fMLF had direct cellular toxicity to neurons, but it could induce a large number of neurons for apoptosis, suggesting that the toxicity of fMLF might come from the infiltrated neutrophils.
Furthermore, we noticed the homology between mtDNA and bacterial DNA in terms of structure and evolution.And our previous study showed that mtDNA could induce the expression of STING to activate inflammation, 26 we used immunoblot to detect the expression levels of STING after injection (Figure 2C).The results showed that the expression levels of STING were also parallel with the increases in neutrophil infiltration.
As neuroinflammation had been activated, we observed expressions of various inflammatory factors.Among them, the expression of IL-1β, NLRP3 and IFN-γ was upregulated (Figure 2D,E

| Impaired neurons and their mitochondria, mtDNA induce neutrophil neuroinflammation
Previously, we observed that Aβ 1-42 could induce the apoptosis of neurons, a large number of mitochondria leakage and release mtDNA, which activated neutrophils infiltration.To verify that the initial chemotactic factors for neutrophils infiltration were Aβ damaged neurons with their intracellular constituents, we injected Aβ 1-42 -treated hippocampus neurons HT22, as well as mitochondria and mtDNA leaked from these cells.We found that in the cerebral tissues of mice injected with Aβ 1-42 -treated cells, mitochondria and mtDNA, neutrophil infiltrations were significantly increased (Figure 3A).
Besides, results showed that 24 h after injection, the expression of STING in mice cerebral tissues increased outstandingly (Figure 3B).In addition, for inflammatory factors, Aβ 1-42 -treated cells and their mitochondria, mtDNA could also upregulate the expressions of NLRP3, IL-1β and IFN-γ (Figure 3C-E).Expression of TNF-α had no significant change compared to the control group (Figure 3F).By injecting Aβ 1-42 -treated hippocampus neurons, and their mitochondria and mtDNA, we found that these results were essentially the same as those of direct injections of   First, we injected Aβ 1-42 or mtDNA into the cerebra of STING (STING ÀnÀ ), NLRP3 (NLRP3 À/À ) and IL-1β (IL-1β ÀnÀ ) knockout mice.
We observed that there were barely neutrophils in the brain (Figure 4A,B), and their neutrophil level was basically consistent with that of the control group.Additionally, the inflammatory factors NLRP3 (Figure 4C,D) and IL-1β were also in trace levels (Figure 4E,F).
Additionally, from the results of the TUNEL apoptosis assay, we found that when knocked out one of STING, NLRP3 or IL-1β gene, not only neutrophil migration was inhibited, but also apoptotic cells were hardly to be found (Figure 4I,J).These results implied that silencing expressions of STING, NLRP3 or IL-1β gene could inhibit neutrophilic neuroinflammation and protect the central nervous system.We observed that neutrophils of wild-type mice could migrate towards higher concentrations of mtDNA.The Aβ/mtDNA mixture (A +M) had an enhanced chemotactic effect on neutrophils.But the number of migrated cells induced by Aβ 1-42 did not have a significant difference to that of NS (Figure 5A).Besides, neutrophils extracted from STING ÀnÀ , NLRP3 ÀnÀ and IL-1β ÀnÀ mice had little response with the stimulations of Aβ 1-42 and mtDNA, even with the mixture (Figure 5B-D).In summary, STING, NLRP3 and IL-1β genes were crucial factors for neutrophil migration, when one of these genes is depleted, neutrophil migration would be inhibited.
[33] So, we investigated their positional relationship in the signal transduction pathway.When neutrophils extracted from STING ÀnÀ stimulated by Aβ 1-42 or mtDNA, no significant expression alterations were observed in IL-1β and NLRP3 compared to the NS control group (Figure 5E,F).Nevertheless, when neutrophils from NLRP3 ÀnÀ and IL-1β ÀnÀ were exposed to mtDNA, the STING expression was upregulated (Figure 5G,H).But, it could not be increased by the stimulation of Aβ 1-42 .These results suggested that STING was the upstream factor of NLRP3 and IL-1β.Meanwhile, mtDNA was the major initiator to STING instead of Aβ to induce the infiltration of neutrophils.
Additionally, recent reports showed that mitochondria had neuron-impairment effect [40][41][42] and mtDNA had an inflammatorypromoting effect, [28][29][30] but the pathological mechanism was still oblivious.Our results showed when mtDNA interacted with cerebral tissue, neutrophil infiltration was activated and the neuron apoptosis process was initiated.Results of in vitro experiments also revealed that mtDNA could induce neutrophil chemotactic migration.Neutrophils are harmful to neurons and our data are supplements and improvements for the mechanism of the mtDNA pathophysiology.Namely, when mtDNA is released from cells, STING protein which recognizes double-stranded DNA will be activated.4][45][46][47] This also explains the reason that mtDNA, with evolutionary homology to bacterial DNA, can induce neuroinflammation.The classic downstream product of STING is the molecules in the interferon family. 48,49Our data showed that the expression of IFN-γ was upregulated when STING was activated, which suggested that mtDNA could activate the STING pathway.50,51 Our results revealed that levels of NLRP, and IL-1β increased parallelly, indicating that STING, NLRP3 and IL-1β were in the same signal transduction pathway.Furthermore, STING and NLRP3 expressions in neuro-degeneration disease were reported by researchers. 16,52,53We found that STING, NLRP3 and IL-1β participated in neutrophil neuroinflammation, but knocking out one of them could inhibit the injury progression.These results demonstrated that the activation of neutrophilic neuroinflammation requires mtDNA-STING-NLRP3/IL-1β axis in integrity.Surprisingly, we did not observe the upregulated expression of TNF-α in both in vivo and in vitro experiments.This differs from previous research results. 22,54,55It implies that the level of TNF-α is not dominated by Aβ 1-42 or mtDNA.Therefore the role of TNF-α in Alzheimer's disease models requires further exploration.
In conclusion, Alzheimer's disease is caused by neurodegeneration.In our research, we found that Aβ 1-42 could induce neuron damage, following mitochondria and mtDNA leakage, which initiated cellular apoptosis and necrosis.The pathophysiological mechanism is that amyloid protein induces mtDNA presenting in the brain, which activates neutrophil migration into cerebra and aggravates neuron injury.The prerequisite of cerebral neutrophil infiltration is the integrity of the mtDNA-STING-NLRP3/IL-1β axis.When any factor in this pathway is depleted, the migration of neutrophils into cerebral tissue will be ceased.Thus, the development of drugs or agents targeted on the knots of the mtDNA-STING-NLRP3/IL-1β axis can inhibit neuroinflammation and alleviate the progression of Alzheimer's disease.

AUTHOR CONTRIBUTIONS
Yuquan Wei and Xiawei Wei designed the study.Xiangyu Xia, Xuemei He and Tingmei Zhao managed the project, collected the data and prepared the manuscript.Xiangyu Xia, Jingyun Yang, Zhenfei Bi, Qianmei Fu, Jian Liu and Danyi Ao analysed the data.

Aβ 1 - 2 . 5 |
42 (QYAOBIO, China) was dissolved in hexafluoroisopropanol (Sigma, USA), and gently shaken until no visible solid, then stored at 4 C overnight.Dried hexafluoroisopropanol in the fume hood and added dimethyl sulfoxide (DMSO) into Aβ 1-42 containers to dissolve it.Vortex Aβ 1-42 solution until there was no visible precipitation.Incubated the solution at 4 C overnight.Centrifuged the solution at 12000 rpm, 4 C for 10 min.The supernatant was soluble Aβ 1-42 and transferred the supernatant into a clean container.Used the BCA assay kit (ThermoFisher Scientific, USA) for concentration detection, then stored the sample at 4 C. Preparation of mtDNA Collected 2-5 g of fresh mice brain tissue and washed in TE (Tris-EDTA) buffer twice.Ground mice brain into homogenate with a grinder (Servicebio, China).Centrifuged the homogenate at 1000 rpm for 10 min, collected supernatant and repeated this step three times.The supernatant was centrifuged at 12000 rpm for 10 min, and the precipitate was crude mitochondria, washed the sediment twice with TE buffer.Prepare mitochondrial DNA lysate solution (NaOH 0.2 mol/L; 1% SDS v/w).Divided the crude mitochondria into 10 parts, added about 1 mL of lysate solution into each part and resuspended.Heated the solution at 65 C for 24 h.Then added the same volume of saturated phenol into the lysate solution, mixed them well, placed on the oscillator for 15 min, added chloroform with the same volume of phenol into the mixture and shaken for 15 min.Centrifuged solution at 12000 rpm for 5 min, collected the upper water phase solution and transferred them into clean containers.Added twofold the volume of anhydrous ethanol to the solution at 4 C, 30 min.Centrifuged the solution at 12000 rpm 4 C for 30 min, and the white flocculent precipitate was mitochondrial DNA.Washed mitochondrial DNA twice with anhydrous ethanol and dissolved with TE buffer.Confirmed the concentration of mtDNA (Nanodrop 2000, Thermo, USA) and store at 4 C.

2. 6 |
Preparation for damaged cells and their mitochondria, mtDNA。 HT22 cells were cultured to the logarithmic growth phase.Added 8 μmol/L Aβ 1-42 and incubated for 24 h.Collected cells and supernatant.The supernatant was centrifuged at 12000 rpm for 30 min to collect mitochondria leaked from damaged cells.Part of mitochondria was used to extract mtDNA with the methods described above.The Aβ 1-42 -treated HT22 cells and their mitochondria, mtDNA were stored at À20 C for use.
5 min.Discarded the supernatant and used PBS resuspended cells and centrifuged at 1000 rpm for 5 min.Discarded the supernatant and added 300 μL binding buffer to resuspended cells.Added 5 μL PI and AnnexinV-FITC into each tube and incubated in the dark at room temperature for 15 min.Centrifuged at 1000 rpm for 5 min, discarded the supernatant and used PBS resuspended cells.After the last centrifugation, add 100 μL 1Â Binding Buffer for detection by flow cytometer (ACEA Biosciences, Inc., USA).

2. 9 |
ELISA for cytokinesThis experiment used the ELISA (enzyme-linked immunosorbent assay) kit (ThermoFisher Scientific, USA), and brief protocols were as follows: added 100 μL diluted capture antibody solution into each clean 96-well plate well, overnight at 4 C. Discarded the solution in each well, washed by PBST (phosphate buffered solution with 0.1% Tween20) for three times.Added 100 μL of ELISA/ELISPOT diluent to each well to reduce non-specific reaction.Incubated at room temperature for 1 h.Discarded the solution and washed them by PBST three times.Added the standard and samples at 4 C overnight.Discarded the liquid and washed it by PBST three times.Added 100 μL primary antibody diluent into each well.Then sealed the plate and incubated at room temperature for 1 h.Discarded the liquid and washed it three times by PBST.Then, added 100 μL HRP-coupled secondary antibody into each well and incubated at room temperature for 30 min.Discarded the liquid and washed it with PBST six times, then added 100 μL TMB solution and incubated at room temperature for 15 min.Added 100 μL reaction termination solution into each hole.Read the absorbance value of each well and draw the standard concentration curve.Then analyse the data.
12000 rpm, 4 C, 5 min.Collected the supernatant and used a BCA kit (ThermoFisher Scientific, USA) to detect the total concentration of protein.Then mixed the sample with a loading buffer at 95 C 10 min.Then stored them at À20 C. Prepared 12.5% SDS-PAGE gel.Loaded sample protein 20-50 μg each well and then conducted electrophoresis at 100 V voltage, and stopped electrophoresis when the indicator reached the bottom.Cut the PVDF film (Merck Millipore, USA) to the proper size and soaked it in methanol for 10 s to activate it.Taken out the SDS-PAGE gel for film-transfer process at 200 mA for 1 h.After transferring completed, washed the PVDF film with TBST (trisbuffered saline with 0.1% Tween20) and then put it into a 5% skimmed milk solution for sealing.Then, transferred PVDF film into the primary antibody solution (anti-NLRP3 antibody, 1:2000, LSbio, USA; anti-STING antibody, 1:2000, CST, USA; anti-GAPDH and antiβ-actin antibody, 1:2000, Santa Cruz biotechnology, USA) at 4 C overnight.After that washed the PVDF film with TBST three times.Then immersed film with HRP coupled secondary antibody (anti-rabbit IgG and anti-mouse IgG, 1:10000, Invitrogen, USA) and incubated at room temperature for 2 h.After the secondary antibody incubation, dripped the luminescent liquid (ECL substrate, BIO-RAD, USA) onto PVDF film, and conducted image collection process by bio-rad imager (ChemiDoc, BIO-RAD, USA).

2. 11 |
TUNEL This experiment used the commercial TUNEL (TdT-mediated dUTP Nick-End Labeling) kit (Promega, USA) and operated according to its instructions.The brief steps were as follows: fixed sections in 4% polyformaldehyde PBS for 15 min.Immersed sections in PBS twice for 5 min each time.Added 100 μL balance buffer at room temperature for 5-10 min.Dripped 50 μL TdT reaction mixture covering tissues.Immersed the sample in SSC solution to stop reaction for 15 min.Washed the sample in PBS three times.Sections were restained with DAPI and observed by fluorescent microscope.
), and they were also consistent with the degrees of neutrophil infiltration.But, TNF-α had no significant alteration after injection of fMLF, Aβ 1-42 , mtDNA F I G U R E 2 Inflammatory cytokines expression activated by Aβ 1-42 and mtDNA-induced neuroinflammation.(A) Neutrophil infiltrations in cerebral tissues in wild-type mice after 24 h injection by NS, fMLF, Aβ 1-42 , mitochondrial DNA (mtDNA), the mixture of Aβ 1-42 and mtDNA (A +M).n = 6 per group.****p < 0.0001 versus NS; t test; data are means ± SEM.Scale bar 25 μm.(B) TUNEL+ cells (TdT-mediated dUTP Nick-End Labeling) in cortex and hippocampus with statistics after injection by 24 h.n = 6 per group.****p < 0.0001 versus NS; t test; data are means ± SEM.Scale bar 25 μm.(C,D) STING and NLRP3 expressions in cerebral tissue detected by immunoblot assay and the grayscale relative concentration analysis.n = 6 per group.**0.001 < p < 0.01 versus NS; ****p < 0.0001 versus NS; t test; data are means ± SEM. (E) The concentration of IL-1β, IFN-γ and TNF-α detected by ELISA (enzyme-linked immunosorbent assay).n = 6 per group.****p < 0.0001 versus NS; t test; data are means ± SEM. and Aβ/mtDNA mixture compared to the control group, which indicated that in our experimental model, TNF-α was not an effector in neutrophil infiltration.The expressions of inflammatory factors might explain the progressive aggravation of neuroinflammation in Alzheimer's disease.
Aβ 1-42 , untreated mitochondria and mtDNA.All of them could induce neutrophil infiltration and upregulate expressions of STING and inflammatory factors IL-1β, NLRP3 and IFN-γ.Namely, when neurons are exposed to Aβ 1-42 , their structure and metabolism could be damaged.Subsequently, a large amount of mitochondria and mtDNA were released.These cell constituents could chemotactically attract neutrophils infiltrated into cerebral tissues and upregulate the expressions of inflammatory factors, inducing neuroinflammation and disturbing physiological functions of the central nervous system.

3. 4 |
Knocking out STING, NLRP3 or IL-1β gene inhibits neutrophil infiltration Subsequently, we adopted specific gene knockout mice to detect the role of STING and inflammatory factor IL-1, NLRP3 in neuroinflammation.

3. 6 |
Neuron degeneration and cognitive impairment induced by Aβ 1-42 and mtDNA are inhibited by knocking out STING, NLRP3 or IL-1β gene In addition, we recorded the long-term results after the injection of Aβ 1-42 and mtDNA.For wild-type mice, after 5 weeks of stereotactic injection, we took brain sections for experiments.We found that the hippocampus region atrophied, especially in CA1 and DG regions, which were sensitive to injury stimulation (Figure 6A).Compared with the control group, mice injected Aβ 1-42 or mtDNA had significant atrophies in CA1 and DG regions.For the mice injected with Aβ/ mtDNA mixture, their hippocampus had a severer atrophy to mice injected with Aβ 1-42 or mtDNA.This confirmed that Aβ 1-42 or mtDNA had toxicity to cerebral tissues, and their toxicity had a synergistic effect when they mixed together.Next, we observed the cerebral changes in STING ÀnÀ , NLRP3 ÀnÀ or IL-1β ÀnÀ knockout mice within 5 weeks period after injection.Results showed that hippocampus atrophies in CA1 and DG regions of genetic knockout animals were less severe than that of the control group when exposed to Aβ 1-42 or mtDNA (Figure 6B,C).This indicated that the knockout of STING, NLRP3 or IL-1β gene could enhance the tolerance of cerebral tissue to impaired stimulations caused by Aβ 1-42 or mtDNA.Finally, we conducted Barnes maze cognitive test on mice.With 5 weeks exposed to Aβ 1-42 or mtDNA, wild-type mice had cognitive disorders manifested as the longer latency to find shelter (Figure 6D).When mice were injected with Aβ/mtDNA mixture, they had a severe cognitive disorders than that of Aβ 1-42 or mtDNA-injected mice.But when one of STING, NLRP3 or IL-1β gene knocked out, the cognition of mice had been improved (Figure 6E,F).These results were consistent with neutrophil migration degree and inflammatory factors expression levels.It proved that STING, NLRP3 or IL-1β gene had important roles in the progression of Alzheimer's disease.But as long as any knot on the mtDNA-STING-NLRP3/IL-1β axis was knocked out, cerebra could enhance tolerance to injury stimulation caused by Aβ 1-42 or mtDNA.