Animal husbandry

Experiments were performed either on C57BL/6 (B6) animals acquired from the Institute for Basic Science Animal Facility (larf.ibs.re.kr) or on AD mouse model mouse APPswe/PSEN1dE9 (APP/PS1) mice of B6C3 hybrid background (RRID: MMRRC_034829-JAX), originated from Jackson Laboratory (USA, stock number 004462) and maintained as hemizygotes by crossing transgene-carrying mice with B6C3 F1 animals. Genotype of the APP/PS1 animals were determined by PCR using the following primers—APP/PS1_F—5’ AAT AGA GAA CGG CAG GAG CA 3’; APP/PS1_R—5’ GCC ATG AGG GCA CTA ATC AT 3’. All mice used in this study were maintained in a 12-h light/dark cycle (lights out at 2000 h) and were allowed ad libitum access to food and water. Animals were maintained in accordance with the guidelines stated by the Institutional Animal Care and Use Committee of IBS (Daejeon, South Korea). B6 littermates at 8-weeks of age were used for virus injection experiments, followed by slice patch clamp recordings at 14 weeks old. Virus injection and subsequent slice patch, behavioral tests and IHC were performed on 12–16 month old APP/PS1 animals and age-matched wild type (WT) littermates.

Primary astrocyte culture

Primary astrocytes were cultured from P1 pups of C57BL/6 J mice as previously described [26]. Briefly, the cerebral cortex and hippocampus were dissected and cleaned of meninges and midbrain before dissociation into a single cell suspension by trituration in astrocyte culture medium. The medium was prepared by using Dulbeccos’ modified Eagle’s Medium (DMEM, Corning) supplemented with 4.5 g/L glucose, L-glutamine, sodium pyruvate, 10% heat-inactivated horse serum, 10% heat-inactivated fetal bovine serum and 1000 units/mL of penicillin–streptomycin. Cells were plated onto culture dishes coated with 0.1 mg/mL poly-D-lysine (Sigma) and maintained in astrocyte culture medium at 37 °C in a humidified atmosphere containing 5% CO2. Three days later (at DIV4), cells were vigorously washed with Dulbecco’s phosphate buffered saline by repeated pipetting and the media was replaced.

Illumina Hiseq library preparation and RNA sequencing

RNA was isolated from cultured primary astrocytes using Qiagen RNEasy Kit (Qiagen, #74,104). Sample libraries were prepared using the Ultra RNA Library Prepkit (NEBNext, #E7530), Multiplex Oligos for Illumina (NEBNext, #E7335) and polyA mRNA magnetic isolation module (Invitrogen, #61,011) following manufacturers’ instructions. Full details of the library preparation and sequencing protocol are provided on the website and previously described [27]. The Agilent Bioanalyser and associated High Sensitivity DNA Kit (Agilent Technologies) were used to determine the quality, concentration, and average fragment length of the libraries. The sample libraries were prepared for sequencing according to the HiSeq Reagent Kit Preparation Guide (Illumina, San Diego, CA, USA). Briefly, the libraries were combined and diluted to 2 nM, denatured using 0.1N NaOH, diluted to 20 pM by addition of Illumina HT1 buffer and loaded into the machine along with read 1, read 2 and index sequencing primers. After the 2 × 100 bp (225 cycles) Illumina HiSeq paired-end sequencing run was complete, the data were base called and reads with the same index barcode were collected and assigned to the corresponding sample on the instrument, which generated FASTQ files for analysis.

NGS Data analysis

BCL files obtained from Illumina HiSeq2500 were converted to fastq and demultiplexed based on the index primer sequences. The data was imported to Partek Genomics Suite (Flow ver 10.0.21.0328; copyright 2009, Partek, St Louis, MO, USA), where the reads were further processed. Read quality was checked for the samples using FastQC. High quality reads were aligned to the Mus musculus (mouse) genome assembly GRCm38 (mm10, NCBI) using STAR (2.7.3a). Aligned reads were quantified to the mouse genome assembly (mm10, RefSeq transcripts 93) and normalized to obtain fragments per kilobase million (or FPKM) values of positively detected and quantified genes. Gene read counts were also normalized to Transcripts per million (TPM), which was used to identify alternate splice variants of the positively detected genes. Differential gene analysis was carried out by normalizing the quantified and annotated gene reads to the Median Ratio and performing DeSeq2 (available on Partek Genomics Suite).

cDNA synthesis and qRT-PCR

cDNA was synthesized from 500 μg RNA as previously described, using Superscript III Reverse Transcriptase (Enzynomics, #RT006M). 20 ng cDNA (per reaction tube) was used to perform qRT-PCR using Power SYBR Green PCR Master Mix (Applied Biosystems, #4,367,659) to check the expression level of SIRT2 and compare against GAPDH (primers used listed below).

Mouse Gapdh qRT-F: ACC CAG AAG ACT GTG GAT GG

Mouse Gapdh qRT-R: CAC ATT GGG GGT AGG AAC AC

Mouse SIRT2 qRT-F: TCA TCA GCA AGG CAC CAC TA

Mouse SIRT2 qRT-R: GTC CCT GTA AGC CTT CTT GG

Immunocytochemistry

For pharmacological study, astrocytes (DIV 7–10) were seeded on coverslips and incubated with 180 μM putrescine (Sigma-Aldrich, P5780) in the presence or absence of 10 μM DEAB, 200 nM EX527 or 3 μM AGK2 overnight. For genetic ablation study, DIV 7 astrocytes were detached from culture dish surface, electroporated with mCherry-tagged pSicoR vector carrying Scr sequence or shRNA sequences against SIRT2 or ALDH1A1 (shSIRT2 targeting 5’-GGAGCATGCCAACATAGATGC-3’ or shALDH1A1 targeting 5’-TTTCCCACCATTGAGTGCC-3’ respectively) and seeded onto coverslips. Two days later, they were treated with putrescine (180 μM) for 24 h. Cells on the coverslips were fixed with 4% paraformaldehyde (PFA; Sigma-Aldrich) in 0.1 M PBS at room temperature for 15 min. After fixation, the coverslips were washed 3 times with 0.1 M PBS for 10 min each, then blocked with 0.1 M phosphate-buffered saline (PBS) containing 0.3% Triton X-100 (Sigma, USA) and 10% Donkey Serum (Genetex) for 1.5 h at room temperature. The cells were then incubated with primary antibodies in a blocking solution in the following composition: guinea-pig anti-GABA antibody (1:500, AB175, Millipore), chicken anti-GFAP antibody (1:1000, AB5541, Millipore), rabbit-anti-SIRT2 (1:200, ab211033, abcam) for overnight (16 h or more) at 4 °C with gentle rocking. After washing 3 times with 0.1 M PBS, 10 min each, the cells were incubated with corresponding secondary antibodies in blocking solution in the following composition: conjugated Alexa 647 donkey-anti-chicken anti IgG (1:500, 703–605-155, Jackson), Alexa 488 donkey anti-guinea-pig IgG (1:500, 706–545-148, Jackson), Alexa 488 donkey-anti-rabbit IgG (1:200, 711–547-003, Jackson) for 2 h at room temperature with gentle rocking. The cells were then incubated with 1:2000 DAPI solution (Pierce) in 0.1 M PBS for 10 min followed by 3 rinses with 0.1 M PBS. Coverslips were finally mounted onto slide glass with fluorescence mounting solution (S3023, DAKO, USA). Images were acquired using a Nikon A1R confocal microscope (pharmacological study) or Zeiss LSM900 microscope (genetic ablation study) and analyzed using the ImageJ program (NIH).

2-cell sniffer patch clamp recording

Primary astrocyte cultures were prepared from P1 C57BL/6 mouse pups as described above. As required, the cells were seeded onto poly-D-lysine-coated cover glass and either electroporated with respective shRNA constructs (genetic ablation experiments) or treated with inhibitors in the presence of putrescine (pharmacological inhibition) on DIV7. On the day of sniffer patch, HEK 293-T cells expressing GFP-tagged GABAC receptors were seeded onto the astrocytes and allowed to settle for at least 1 h before patching. The cover glasses were then immersed in 5 μM Fura-2-AM (in 1 mL external HEPES solution containing 5μL of 20% pluronic acid) for 40 min to allow Fura incorporation into the cell, washed at room temperature (with external HEPES solution, described later) and subsequently transferred to the microscope stage. The external HEPES solution of the following composition (in mM): 150 NaCl, 10 HEPES, 3 KCl, 2 CaCl2, 2 MgCl2, 5.5 glucose (pH adjusted to 7.3, osmolality to 320 mOsmol kg−1) was allowed to continuously flow over the cells during the experiment and was replaced with one containing 100 μM GABA during full activation recording. Images at 510 nm wavelength were taken after excitation by 340 nm and 380 nm light using pE-340fura (CoolLED) to record calcium transients within the cells. The two resulting images were used for ratio calculations in Axon Imaging Workbench (version 11.3, Axon Instruments). To perform the sniffer patch, the astrocytic TRPA1 receptor was activated by pressure poking with a glass pipette and the resulting GABA release was recorded as inward current in the GABAC-GFP-expressing HEK 293 T cells under voltage clamp (Vh = −60 mV) using Axopatch 200A amplifier (Axon Instruments), acquired with pClamp 11.3. Recording electrodes (4–10 MΩ) were filled with the following internal solution (in mM): 110 Cs-gluconate, 30 CsCl, 0.5 CaCl2, 10 HEPES, 4 Mg-ATP, 0.3 Na3-GTP and 10 BAPTA (pH adjusted to 7.3 with CsOH, osmolality adjusted to 300 mOsm kg−1 with sucrose). For the simultaneous recording of calcium response with the patch and poking pipettes, Imaging Workbench was synchronized with pClamp 11.3. To normalize for the differences in GABAC receptor expression on the HEK cells, a saturating concentration of 100 μM GABA (in HEPES solution) was applied to record maximal GABA current from the cell, and the poking-induced current was normalized as a percentage of full activation current on application of GABA from the HEK cell.

Metabolite analysis

Metabolite analysis for putrescine, acetyl-GABA, and GABA was performed using electrospray ionization UPLC-MS/MS system. The system used for the analyses was an Exion LC™ AD UPLC coupled with a Triple Quad 4500 MS/MS System (AB Sciex LLC, Framingham, USA) using an Acquity® UPLC BEH HILIC column (1.7 μm particle size, 2.1 mm x 100 mm, Waters, USA) at 30 °C, controlled by Analyst 1.6.2 software (AB Sciex LP, Ontario, Canada). Into the astrocyte sample pellets, 70% methanol (100μL) was added and was vortexed for 30 s. Three freeze/thaw cycles with liquid nitrogen were used to lyse cells, followed by a 10 min centrifugation at 20,817 g (14,000 rpm). DNA normalization was performed using 5μL of each sample's supernatant. For sample normalization, DNA concentrations in each sample were analyzed using a Nano-MD UV–Vis spectrophotometer (Scinco, Seoul, Korea). The internal standard (d2-GABA at a final concentration of 4 μM) was added to 20μL of the supernatant from each sample, and the mixture was vortexed for 30 s. Then the mixture was evaporated with gentle stream of nitrogen to dryness at 37℃ using a TurboVap evaporator (Biotage, Uppsala, Sweden). The residue was reconstituted by vortexing for 30 s and sonicating for 15 min using 25μL of the mobile phase A (0.1% formic acid in acetonitrile):B (50 mM ammonium formate, pH 4) = 8:2 solvent. The initial chromatographic conditions were at 80% solvent A at a flow rate of 0.4 mL/min. After 7 min at 80% A, solvent A was set to 5% over the next 0.5 min, and these conditions were retained for an additional 1 min. The system was then returned to the initial conditions over the next 0.5 min. The system was then left for 1.5 min in the initial conditions for re-equilibration. The total sample running time was 10.5 min. All samples were placed in the auto-sampler, maintained at 7 °C during the analysis, and the injection volume was 5μL. The analysis was performed using positive ESI mode. The ion spray voltage and vaporizer temperature were 5.5 kV and 380℃, respectively. The curtain gas was kept at 35 psi, and the collision gas was maintained at 8 psi. The nebulizer gas was 60psi, while the turbo gas flow rate was 70psi. The metabolites were detected selectively using their unique multiple reaction monitoring (MRM) pairs. The MRM parameter (Q1/Q3) for putrescine, acetyl-GABA, and GABA are (89.062 / 72.1), (146.011 / 86.015), and (103.979 / 87.1), respectively, to monitor specific parent-to-product transitions. The standard calibration curve for each metabolite was used for absolute quantification.

Live-cell imaging of oROS-G in primary cultured mouse astrocytes

On DIV7, primary cultured astrocytes were transfected with AAV5-GFAP-oROS-GFP viral vector and seeded into a glass-bottom 96-well plate (ibidi; #89,627). 2 days later, cells were treated with putrescine (180 μM) in the presence or absence of enzyme inhibitors (AGK2, 3 μM or KDS2010, 1 μM) or vehicle and confocal live-cell imaging was performed using a Nikon A1R confocal microscope mounted onto a Nikon Eclipse Ti body with 20X objective lens. The plate was housed in a live-cell imaging chamber and maintained at 10% CO2 and 37 °C for 47 h of continuous recording. Images were acquired and analysed using the NIS-element AR (Nikon).

Virus injection

Mice were anesthetized using isoflurane and head-fixed onto stereotaxic frames (Kopf). The scalp was incised, and a hole was drilled into the skull above the hippocampus (A/P −1.8, D/V −1.9 from bregma, M/L ± 1.2 from the skull surface). Viruses were loaded into a stainless steel needle in a 1:1 ratio (lentivirus:AAV) and injected bilaterally into the dentate gyrus at the rate of 0.1μL/min for 10 min (1μL in each hemisphere). Viruses used: AAV-GFAP-Cre-mCherry, lenti-pSico-Scr/shSIRT2/shALDH1A1-eGFP, were generated at the Institute for Basic Science Virus Facility (https://www.ibs.re.kr/virusfacility/). Mice were used 2–3 weeks after injection for behavior experiments and 6 weeks after injection for electrophysiological recordings.

Acute brain slicing

Brain slices were prepped from 14–15 weeks old B6 animals or 12–16 months old transgenic (and WT littermate) APP/PS1 animals. Briefly, animals were deeply anaesthetized using isoflurane and the brain was swiftly excised from the skull in ice-cold high-sucrose artificial cerebrospinal fluid (aCSF) with the following composition (in mM): 212.5 sucrose, 26 NaHCO3, 10 d-( +)-glucose, 5 MgCl2, 3 KCl, 0.1 CaCl2, 1.25 NaH2PO4. Coronal slices of the hippocampus of 300 μm thickness were prepared and incubated at room temperature in extracellular aCSF for at least 1 h before recording. The extracellular aCSF solution used for incubation and recordings was composed as follows (in mM): 124 NaCl, 3 KCl, 24 NaHCO3, 2 CaCl, 1.25 NaH2PO4, 1 MgCl2, 10 d-( +)-glucose (osmolarity adjusted to 310 mOsm kg−1, pH 7.4). All solutions were maintained with constant bubbling with a mixture of 95%O2 and 5%CO2.

Tonic GABA recording

Electrophysiological recordings were performed by placing the slices in a recording chamber continuously perfused with extracellular aCSF (flow rate 0.2 mL/min), mounted on the stage of an upright Olympus microscope and viewed with a 4X and a 60X water-immersible objective lens (0.90 numerical aperture) with infrared DIC optics. The cells were visualized using a charged-couple device camera and the Imaging Workbench software (INDEC Biosystems). Whole-cell patch clamp recordings were made from the granule cells located in the dentate gyrus, held at a potential of −70 mV. Patch pipettes (5-9MΩ) were filled with an internal solution of the following composition (in mM): 135 CsCl, 4 NaCl, 0.5 mM CaCl2, 10 HEPES, 5 EGTA, 10 QX-314, 2 Mg-ATP and 0.5 Na2-GTP. Once whole-cell configuration was obtained, baseline current was stabilized using d-AP5 (50 mM, Tocris) and 6-cyano-7-nitroquinoxaline-2,3-dione (20 mM; Tocris) after which aCSF additionally containing bicuculline (100 μM; Tocris) was flowed through the recording chamber to record the tonic GABA current. Electrical signals were digitized and samples at 50 μs intervals with Digidata 1440A and a Multiclamp 700B amplifier (Molecular Devices) using pCLAMP10.2 software. The amplitude of the tonic GABA current was measured as the baseline shift after bicuculline application using the Clampfit program.

Mouse brain tissue sectioning and immunohistochemistry

Animals were anaesthetized using isoflurane and perfused with 0.9% saline, followed by ice-cold 4% PFA in 0.1 M PBS. The brain was excised and stored in 4% PFA at 4 °C overnight for post-fixation, followed by dehydrolysation in 30% sucrose for 48 h. Coronal sections of 30 μm thickness were prepared in a cryostat and stored in a glycerol-based storage solution at 4 °C till use. Before staining, the slices were washed in 0.1 M PBS thrice and incubated for 1 h in blocking solution (4% Donkey Serum, 0.3% Triton X-100 in 0.1 M PBS). Primary antibodies were added to the blocking solution at desired dilution and incubated overnight at 4 °C with gentle rocking. Unbound antibodies were washed off by rocking and rinsing the slices thrice with 0.1 M PBS, followed by 2 h of incubation at room temperature with corresponding fluorescence-tagged secondary antibodies (diluted in blocking solution). Unbound secondary antibodies were washed by rocking and rinsing thrice with 0.1 M PBS, the first wash of which contained 1:1000 DAPI for nuclear visualization. The slices were then mounted using a fluorescence mounting medium (Dako) and dried.

Antibodies used in the experiments were as follows (dilutions in blocking solution)—rabbit-anti-ALDH1A1 (1:200, ab52492, abcam), chicken-anti-GFAP (1:500; AB5541, Millipore), rabbit-anti-SIRT2 (1:200, ab211033, abcam), Alexa 647 donkey-anti-chicken anti IgG (1:500, 703–605-155, Jackson), Alexa 488 donkey-anti-rabbit IgG (1:200, 711–547-003, Jackson), PyrPeg (1 μM, incubated along with secondary antibodies). 22–24 μm Z-stacked images in 2 μm steps were processed using the ZEN Digital Imaging for Light Microscopy blue system (Zeiss, ver. 3.2) and ImageJ (NIH, ver. 1.54b) software.

Mouse behavior tests

Behavior experiments were performed 2–3 weeks after virus injection. All test subjects were handled for 10 min by the experimenter every day for 5 days before behavior recording at the same time during the day/night cycle in the same room as the behavior setup. Behavior experiments were carried out in the light phase circadian cycle.

For the Open Field test, animals were placed in a 40 cm × 40 cm chamber with no markings and allowed to explore uninterrupted for 10 min. They were tracked for the last 8 min of the recording (Ethovision XT, Noldus) to measure the total distance travelled and the velocity.

On the day of the Y-maze experiment, the test animals were put in the middle of a symmetrical Y maze with 3 identical arms (30 cm long x 15 cm high) and allowed to explore freely for 10 min. The animal was recorded and the last 8 min of the videos were studied by an experimenter, blinded to the animal test condition, to analyze the total number of arm entries and the alternation behavior. Data was analyzed to exclude any abnormal behavior (one test subject was removed from the study due to reduced mobility). An entry was considered only when all 4 limbs of the animal were within the arm. The percentage of alternation was calculated as follows:

$$Percent\;alternation=\lbrack(total\;number\;of\;alternations)/(total\;number\;of\;arm\ entries-2)\rbrack\ast100.$$

For the Novel Place Recognition test, animals were placed into a 40 cm x 40 cm chamber with one marked wall and 2 identical objects placed close to the marked wall (see schematic in Fig. S2 for reference). Animals were allowed to roam freely and explore the objects for 8 min before returning to their home cage. 60 min later, one object was moved further away from the marked wall and the animal was returned to the chamber. Animal exploration of each object was tracked for 8 min. Recorded videos were analyzed by an experimenter blinded to animal condition. Exploring was considered when the animal sniffed at the object without attempting to climb it in the recorded video. Ratio of preference (Ret./Acq.) was calculated as follows:

$$\beginPreference\;in\;Acq\;=\;(Time\;spent\;exploring\;object-to-be-moved)/Total\;exploration\;time\\Preference\;in\;Ret\;=\;(Time\;spent\;exploring\;moved\;object)/Total\;exploration\;time\\Ratio=(Preference\;in\;Retention)/(Preference\;in\;Acquisition)\end$$

Western blot

The hippocampus was dissected from the acute brain slices after patch clamp recordings were done and immediately frozen. Protein was extracted from the samples by homogenization in RIPA buffer (Rockland; MB-030–0050) containing 100 × Phosphatase Inhibitor (GenDEPOT; P320-001) and estimated using the BCA assay (Thermo Scientific; #23255). 20 μg of protein was prepared (boiling with marker), loaded into a gradient 4–15% SDS protein gel (BioRAD; #4561084) and then separated via electrophoresis. The separated proteins were transferred onto a PVDF membrane (Invitrogen; B24002) and blocked for 1 h at room temperature with 5% skim milk (Difco) in TBST solution. The membrane was then rinsed thrice with TBST and incubated with primary antibodies in 3% BSA (GenDEPOT; A0100-010) in TBST and sodium azide overnight at 4 °C. Membranes were subsequently washed thrice with TBST and incubated with appropriate HRP-conjugated antibodies (in 5% skim milk in TBST) for 1 h at room temperature. Membranes were thoroughly washed with TBST (3–4 times) and the HRP was visualized using ECL Western Blotting Substrate (Thermo Scientific; #32106). Expression (or knockdown) of SIRT2 was calculated relative to the protein loading control β-actin. Antibodies used – 1:1000 Rabbit anti-SIRT2 (abcam; ab211033); 1:2000 Rabbit anti-beta actin (Cell Signaling; 490L); 1:3000 Goat-anti-Rabbit HRP (Sera Care; 5450–0010).

Human brain samples

Neuropathological examination of postmortem brain samples from normal subjects and AD patients was determined using procedures previously established by the Boston University Alzheimer’s Disease Center (BUADC). Demographic information of humans used in this study is provided in Table S1. Next of kin provided informed consent for participation and brain donation. Institutional review board approval for ethical permission was obtained through the BUADC center. This study was reviewed by the Institutional Review Board of the Boston University School of Medicine and was approved for exemption because it only included tissues collected from post-mortem subjects not classified as human subjects. The study was performed in accordance with institutional regulatory guidelines and principles of human subject protection in the Declaration of Helsinki.

Double chromogenic staining for human brain tissues

Sequential double staining was performed using horse radish peroxidase (HRP) and alkaline phosphatase (AP) substrates as chromogens for two different antigens, respectively.

First staining

Paraffin-embedded tissues were sectioned in a coronal plane at 10 μm. BLOXALL® Blocking solution (Vector Laboratories, Burlingame, CA, USA) was used to block endogenous alkaline phosphatase. Tissue sections were blocked with 2.5% normal horse serum (Vector Laboratories) for 1 h and then incubated with GFAP antibody (1:400; AB5541, Millipore) for 24 h. After reaction with secondary antibodies, tissue slides were processed with Vector ABC Kit (Vector Laboratories). The GFAP immunoreactive signals were developed with DAB chromogen (Thermo Fisher Scientific).

Second staining

Tissue slides stained with GFAP were incubated with SIRT2 antibody (1:200; ab211023, abcam) for 24 h to confirm localization of SIRT2 in reactive astrocytes. After washing three times with PBS, sections were incubated with ImmPRESS-AP anti-rabbit IgG (alkaline phosphatase) polymer detection reagent (Vector Laboratories: MP-5401) for 2 h at room temperature. A Vector Red alkaline phosphatase substrate kit (Vector Laboratories: SK-5105) was used to develop SIRT2 signals. Hematoxylin (Vector Laboratories: H-3401–500) was added to tissue slides to visualize the nuclei of the cells. Double-stained tissue slides were gradually processed back to xylene through an increasing ethanol gradient [70%, 80%, 90%, 95%, and 100% (1 time)], and subsequently mounted. The chromogenic signals of GFAP (brown) and SIRT2 (red) were examined under a light microscopy (BX63; Olympus, Japan) equipped with high definition (1920 × 1200 pixel) digital camera (DP74; Olympus).

Quantification and statistical analysis

All analyses were done blindly. The statistical tests used in each figure are listed in the legends. Numbers and individual dots refer to individual samples (or individual cells) unless mentioned otherwise in figure legends. N represents the number of animals used for the experiment. Data representation and statistical analysis was performed using GraphPad Prism (Graphpad Software). For image analysis, ImageJ (NIH) and Nikon A1R (Nikon) were used. Statistical significance was set at ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 and ∗∗∗∗p < 0.0001 (unless mentioned otherwise in figure legends).