Suppression of LPS-induced tau hyperphosphorylation by serum amyloid A

Primary antibodies used in this study are listed in Additional file 1: Table S1. The BCA protein assay kit and 4,6-diamidino-2-phenylindole (DAPI) were obtained from Beyotime Institute of Biotechnology (Nantong, Jiangsu, China). Mouse and rabbit control IgGs were purchased from Santa Cruz Biotechnology (Dallas, TX). IRDye®800CW secondary antibodies were from LI-COR (Lincoln, NE). Dulbecco’s modified Eagle’s medium (DMEM), neurobasal-A, and B-27® supplements were purchased from Life Technologies (Carlsbad, CA). Other chemicals were obtained from Sigma Chemical Company (St. Louis, MO).

The Saa3 ^−/− (Saa3 knockout) mice in C57BL/6 genetic background were obtained from the Knockout Mouse Project (KOMP) Repository (Davis, CA). Age- and sex-matched littermates were used in the experiments. The C57BL/6 mice were purchased from SLACCAS Laboratory Animal Co., Ltd (Shanghai, China). The generation of the Saa3 transgenic mice is described in Additional file 2. All mice were housed (four to five animals per cage) with a 12/12 h light/dark cycle, with ad libitum access to food and water. The housing, breeding, and animal experiments were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, with procedures approved by the Biological Research Ethics Committee of Shanghai Jiao Tong University.

Lipopolysaccharide (LPS, from Salmonella enterica serotype Abortus equi, Sigma-Aldrich, Cat. No. L5886, Lot 032M4067) at a low concentration (5 mg/kg body weight) or a high concentration (15 mg/kg body weight) was intraperitoneally injected into 3-month-old C57BL/6 mice consisting of Saa3 ^−/− and their respective wild-type (WT) littermates (n = 4 per group; half males and half females). The mice received either LPS or an equal volume of normal saline. After 24 h, all mice were sacrificed by decapitation and their brains removed immediately. The hippocampi and cerebral cortices of the left hemisphere of the mouse brain were dissected, flash frozen in dry ice, and stored at −80 °C for biochemical analyses later. The right hemispheres of the mouse brains were fixed with 4 % paraformaldehyde in 0.1 M phosphate-buffered saline (PBS), followed by cryoprotection in 30 % sucrose. Coronal sections of 30-μm thickness were cut using a freezing sliding microtome. The sections were stored in glycol anti-freeze solution (ethylene glycol, glycerol, and 0.1 M PBS in 3:3:4 ratio) at −20 °C until immunohistochemical staining.

Three-month-old C57BL/6 mice were randomly divided into control (saline) and SAA (recombinant human apo-SAA; Pepro Tech, Rocky Hill, NJ) (n = 4 in each group; half males and half females). Mice were anesthetized by intraperitoneal injection of 50 mg/kg pentobarbital sodium (Sigma) and then restrained onto a stereotaxic apparatus (RWD Life Science, Shenzhen, China). For stereotaxic intracranial injection of SAA, an earlier method was used [24]. After surgically exposed, the duramater and drilled holes in the skull, SAA (20 μg/8 μl) dissolved in saline was bilaterally injected into the left and right hippocampi (two sites, 4 μl per site), using a 10-μl Hamilton syringe with a 26-gauge needle. The injection rate was 0.5 μl/min, and the needle was kept in each site for an additional 3 min before a gentle withdrawal. The bregma coordinates were determined as follows: anterior, −2.0 mm; lateral, ±1.3 mm; vertical, −2.2 mm. As controls, the mice stereotaxically received equal volume of normal saline. Postoperatively, all mice were placed on heating pads (37 °C) until recovered from surgery. After 48 h, the injected mice were sacrificed by decapitation and the brain samples were prepared as with the LPS-treated mice described above.

Total RNA was extracted from the cortex and hippocampus of the mouse brain and primary cultures of neurons, astrocytes, and microglial cells using TRIzol reagent (Invitrogen). One microgram of the RNA was used for reverse transcription using the Reverse Transcription System A3500 kit (Promega, Madison, WI). The complementary DNA (cDNA) was subsequently subjected to Real-Time PCR to quantify the transcripts of TNF-α, IL-6, IL-10, Saa1, and Saa2, and Saa3 using SYBR® Green Real-time PCR Master Mix (TOYOBO, Osaka, Japan). The following primers were used: TNF-α (5′-TTC TCA TTC CTG CTT GTG G-3′; 5′-ACT TGG TGG TTT GCT ACG-3′), IL-6 (5′-CTT CTT GGG ACT GAT G-3′; 5′-CTG GCT TTG TCT TTC T-3′), IL-10 (5′-AGG GTT ACT TGG GTT GC-3′; 5′-TGA GGG TCT TCA GCT TC-3′), Saa1 (5′-TTG TTC ACG AGG CTT TC-3′; 5′-TTT GTC AGG CAG TCC AG-3′), Saa2 (5′-TGA TGC TGC CCA AAG G-3′; 5′-GCC AGG AGG TCT GTA GTA A-3′), and Saa3 (5′-CCT TCC ATT GCC ATC A-3′; 5′-GGG TCT TTG CCA CTC C-3′). The primers for the mouse housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were 5′-CCT TCC GTG TTC CTA CC-3′ and 5′-CAA CCT GGT CCT CAG TGT A-3′. PCR was performed according to the following conditions: 94 °C for 3 min, 40 cycles of denaturation at 94 °C for 30 s, annealing at 56 °C for 45 s, and extension at 72 °C for 30 s, followed by a final extension at 72 °C for 10 min. The quantitative fold changes in messenger RNA (mRNA) in each sample were normalized to GAPDH expression and calculated using the 2^{exp(−ΔΔCt)} method.

Mouse brain tissues and cultured neuronal cells were homogenized in lysis buffer containing 50 mM Tris-HCl (pH 7.4), 8.5 % sucrose, 1.0 mM EGTA, 1.0 mM EDTA, 1.0 % sodium dodecylsulfate, 10 mM β-mercaptoethanol, protease, and phosphatase inhibitor cocktails (Sigma-Aldrich). Protein concentrations were determined using a BCA Protein Assay Kit (Beyotime Biotechnology, Nantong, China) and then heated in 5×SDS-PAGE loading buffer (Beyotime Biotechnology) at 99 °C for 7 min. Tissue and neuronal cell homogenates were separated on 10 or 12 % SDS-PAGE and transferred onto nitrocellulose membranes (Whatman Protran, Dassel, Germany). Blots were blocked with 5 % non-fat milk for 1 h at room temperature and incubated overnight at 4 °C with primary antibodies including Tau5 (1:2000); the anti-phospho-tau antibodies pS199 (1:2000), pS205 (1:2000), pS396 (1:2000), pS404 (1:2000), anti-Saa3(1:500), and anti-β-actin (1:10000), followed by the respective IRDye®800CW secondary antibodies (LI-COR Biosciences, Lincoln, NE). The membranes were scanned using the 800-nm channel of an Odyssey® CLX Infrared Imaging System. The immunoreactive bands were quantified using the NIH Image J software.

Sections of mouse brain were washed with 0.05 M Tris-buffered saline (TBS) (containing 0.05 M Tris buffer and 9 g/L NaCl, pH 7.4) for 15 min and then permeabilized with 0.1 % Triton X-100 in TBS for 15 min and blocked with 5 % normal goat serum (NGS) for 30 min at room temperature. The sections were incubated overnight at 4 °C with anti-Iba1 (1:500) in TBS with 5 % NGS. After washing with TBS for 15 min, the sections were incubated with Alex Fluor®568 donkey anti-rabbit antibody (1:500, Invitrogen) at room temperature for 1 h. Again, after three washes in TBS, the sections were stained for nuclei with 5 μg/ml of DAPI for 10 min at room temperature and mounted on glass slides. To identify the expression of Saa3 in neuron, microglia, or astrocyte, the brain sections were first stained with anti-MAP-2 (1:200), anti-CD11b (1:200), or anti-glial fibrillary acidic protein (GFAP)-Cy3^TM (1:500) antibody overnight and incubated in Alex Fluor®633 donkey anti-mouse or Alex Fluor®555donkey anti-rat IgG (1:500; Invitrogen) for 1 h. The sections were rinsed in TBS, stained with anti-Saa3 antibody (1:200) overnight at 4 °C, and incubated with an Alex Fluor®488 donkey anti-rabbit antibody (1:500; Invitrogen) for 1 h. Sections were counterstained with DAPI for 10 min, and then fluorescent confocal images were captured using a laser-scanning confocal fluorescence microscope (TCS SP8, Leica Microsystems, Wetzlar, Germany).

Neuronal cultures were prepared in cortices from newborn (postnatal day 0) WT mouse pups as described before [25]. In brief, cerebral cortices were removed from the brains of mice, the meanings and microvessels were removed, and tissues were minced with a sterile razor blade. Tissues were digested with 0.025 % trypsin (Sigma) and 0.01 % DNase I (Sigma) at 37 °C for 10 min. The cell suspension was filtered through a 200-mesh sieve, and cells were plated on poly-d-lysine (Sigma)-coated 24-well plates at a density of 5 × 10⁵ cells per well. Two hours later, the DMEM medium was replaced with neurobasal medium containing 2 % B-27® supplements for 2 days. Culture medium were changed to neurobasal with 10 % fetal bovine serum (FBS) and 3 μg/ml cytosine-β-D-arabinofuranoside (Ara-C, Sigma) in the following 3 days and then again switched back to neurobasal medium containing 2 % B-27® supplements. Experiments were performed on days 7–8 after initiation of the culture.

Microglial and astrocyte cultures were prepared from newborn (postnatal day 0) WT mouse pups as described before [6, 25, 26]. Specifically, separated cells were cultured in poly-D-lysine coated 75 cm² flasks with DMEM medium (containing 10 % FBS, 100 U/ml penicillin, and 100 g/ml streptomycin sulfate). The medium was replenished on day 1 and day 3. On day 7, microglia cells in the culture flasks were shaken off at 260 rpm for 2.5 h, and the remaining astrocytes were maintained in DMEM with 10 % FBS. Experiments were performed after two passages of the cells.

Primary neuronal cells, astrocyte cells, and microglial cells were stimulated with LPS at indicated concentrations or with PBS for 24 h. Total RNA was extracted for real-time PCR to detect cell-specific expression of Saa3. For the analysis of the effect of SAA on neuronal tau phosphorylation, primary neuronal cells were treated with the recombinant human SAA (PeproTech, Rocky Hill, NJ; 0.5 μΜ for 3, 6 and 12 h), and the neuronal lysates were prepared for Western blot analysis as mentioned above.

The microglia cells and astrocyte were plated at a density of 1 × 10⁶ cells per well onto a 12-well plate for experiments involving the collection of conditioned media. The cells were treated with 0.5 μΜ SAA or PBS (control) for 6 h. The medium was removed and replaced with fresh DMEM. The cells were incubated for another 12 h, and the conditioned medium (CM) was collected. Half of the medium in the primary neurons was replaced with equal volume of CM from SAA-stimulated or PBS (control) cell culture. After 3, 6, and 12 h of incubation, neuronal lysate was prepared and blotted for Western analysis. To determine the effect of IL-10, an IL-10-neutralizing antibody (25 μg/ml; BD Pharmingen, San Diego, CA) was added to CM 30 min before application to primary cultures of neurons. After 12 h of incubation, neuronal lysate was prepared and Western blot analysis was conducted.

The IL-10 in the microglial conditioned medium was measured using enzyme-linked immunosorbent assay (ELISA) kit (eBioscience, San Diego, CA), according to the instructions of the vender. Briefly, the conditioned medium and mouse IL-10 standard samples were incubated in the IL-10 ELISA plates at 4 °C overnight. After washed plates three times with PBST (PBS + 0.5 % Tween-20) and blocked with 3 % BSA in PBS for 1 h at 37 °C, biotin-conjugate anti-mouse IL-10 antibody was incubated in the wells at 37 °C for 90 min. The plate was emptied and washed three times with PBST. Substrate solution was pipetted into each well after drying and incubated in 37 °C for 10 min. The enzymatic reaction was stopped by adding stop solution per well. The concentration of IL-10 was determined by measuring absorbance at 450 and 630 nm. All the experiments were done in three separate experiments.

Data are presented as mean ± SEM from at least three experiments. One-way ANOVA followed by Newman-Keuls test or student’s t test was performed using the statistic software GraphPad Prism 5 (San Diego, CA). p values less than 0.05 was considered statistically significant.

To evaluate the function of SAA in tau phosphorylation, SAA expression and distribution in the mouse brain was examined using a systemic inflammation model [27, 28]. Three-month-old C57BL/6 mice were given a single dose of LPS at 5 and 15 mg/kg through intraperitoneal injection. The control mice received an equal volume of normal saline. After 24 h, brain extracts from hippocampus and cortex were collected (Fig. 1a). Systemic administration of LPS induced neuroinflammation in the brain, as evidenced by a modest but significant increase in the transcripts of the pro-inflammatory cytokines IL-6 and TNF-α in the hippocampus and cortex (Fig. 1b). The effect of systemic LPS administration in SAA expression in the brain was next examined. All three inducible mouse Saa transcripts were elevated in the hippocampus and cortex (Fig. 1c). Of note, there was a 3500-fold increase in the hippocampus and a 600-fold increase in the cortex of the Saa3 transcript in mice receiving 15 mg/kg of LPS compared to mice receiving normal saline (Fig. 1c). These results suggest that, in the mouse brain, Saa3 is the predominant form of SAA induced by LPS. The inducible expression of Saa3 was confirmed at the protein level using a specific antibody against Saa3, as shown in Western blots with two randomly chosen brain samples from mice receiving 15 mg/kg of LPS compared to those receiving normal saline (Fig. 1d).

In addition to immunoblotting, immunofluorescence staining was performed to confirm the increase in Saa3 expression and its distribution in the mouse brain. Staining of serial slices from WT mouse brain with an anti-Saa3 antibody and Alexa Fluor 488-conjugated secondary antibody identified a significant up-regulation of Saa3 (green fluorescence) in the CA1 (Fig. 2a–c) and DG (Additional file 3: Figure S1) regions of the hippocampus as well as in the cortex (Additional file 3: Figure S2) of mice receiving LPS, compared with mice receiving saline. To identify the cellular origin of Saa3, double immunostaining was performed for Saa3 and the cell-specific markers MAP2 (neurons), CD11b (microglia), and GFAP (astrocytes). The Saa3 protein was colocalized with MAP2 (Fig. 2a, Additional file 3: Figures S1A and S2A) and, to a lesser extent, with GFAP in the DG area of the hippocampus (Additional file 3: Figure S1C) and in the cortex (Additional file 3: Figure S2C). No Saa3 protein colocalization was visible in CD11b-positive cells, although induced expression of CD11b was evident in microglia from the LPS-injected mice (Fig. 2b, Additional file 3: Figures S1 and S2). These results show a much higher level of inducible Saa3 expression in neurons than in astrocytes, but not in microglia. Since Saa3 is reported to stimulate glial cells [6], its role in the activation of microglia and astrocytes after LPS injection was next examined. The expression of GFAP, a protein abundant in activated astrocytes, did not change in mice receiving LPS (Fig. 2c, d). In contrast, the CD11b-immunoreactive cells displayed a hypertrophic morphology, a sign of microglial activation. The expression of CD11b in these cells was significantly higher than that in the control (saline) samples (Fig. 2c, d). Altogether, these results demonstrate that Saa3 is the major form of SAA proteins induced by LPS in mouse neurons and astrocytes.

To determine whether the elevated Saa3 expression plays a role in neuroinflammation and AD development, Saa3 gene knockout (Saa3 ^−/−) mice were generated (Fig. 3a). The success of Saa3 deletion was functionally confirmed in mice receiving systemic LPS administration, which led to elevated Saa3 expression in WT but not the Saa3 ^−/− mice (Fig. 3b, c). These mice were next compared for the extent of tau hyperphosphorylation, a major pathological feature of AD [1]. The contents of total tau (detected by the Tau5 antibody), non-phosphorylated tau (detected by the Tau1 antibody), and phosphorylated tau at two AD-related amino acid positions (detected by antibodies against pT205 and pS396, respectively) were determined in the hippocampus of Saa3 ^−/− mice and their WT littermates that received LPS or saline. As shown in Fig. 3d, e, there was an increase in tau phosphorylation at T205 and S396 in WT and Saa3 ^−/− mice receiving LPS injection, compared to mice receiving saline injection. A decrease in the level of non-phosphorylated tau in LPS-injected mice was also evident when normalized against total tau (Tau5). When the Saa3 ^−/− mice were compared with their WT littermates, the Saa3 ^−/− mice showed an additional increase in tau phosphorylation at T205 and S396, along with a more evident decrease in non-phosphorylated tau between the saline-injected mice and LPS-injected mice (Fig. 3d, e). The results were further confirmed by immunofluorescence staining analysis. As expected, after systemic LPS treatment, the expression of phosphorylated tau at T205 had an additional increase in neuronal cells from the Saa3 ^−/− mouse (Additional file 3: Figure S3), but not in astrocytes (Additional file 3: Figure S4). These findings suggest a role for Saa3 in restraining tau hyperphosphorylation induced by systemic LPS administration.

To further evaluate the potential involvement of Saa3 in regulating tau hyperphosphorylation, transgenic mice expressing Saa3 were generated. Based on our finding that induced Saa3 expression was seen mainly in neurons, the Saa3 transgenic (Saa3-Tg) mice were prepared using the neuron-specific synapsinI (SYNI) promoter (Additional file 3: Figure S5A–B; see Additional file 2 for detail). Transgenic expression of Saa3 increased the basal and inducible Saa3 mRNA level (Additional file 3: Figure S5C). Of note, the basal and inducible expression of Saa3 was also enhanced at the protein level in the Saa3-Tg mice (Additional file 3: Figure S5D–E). To determine whether Saa3 was secreted by neurons, primary neuronal cultures prepared from newborn Saa3-Tg mouse pups and their WT littermates were stimulated with 0.1 or 1 μg/ml of LPS for 24 h. The ELISA result showed a higher level of Saa3 protein in the culture medium from Saa3-Tg mice than non-Tg mice, either with or without LPS stimulation (Additional file 3: Figure S5F). These results validate the transgenic expression of Saa3 in the mouse brain.

To investigate whether transgenic expression of Saa3 affects tau phosphorylation, the level of total tau and phosphorylated tau at T205 and S396 was determined in Saa3-Tg mice and non-Tg controls following systemic LPS administration (Additional file 3: Figure S6). When normalized against total tau, an increase in tau phosphorylation at T205 and S396 was observed in non-Tg mice injected with LPS compared with that of saline controls. In the Saa3-Tg mice, however, the level of LPS-induced tau phosphorylation at T205 and S396 was lower than that of the non-Tg mice.

Intracerebral injection has been used widely to assess the effects of drugs and cell-derived factors in brain functions. To further evaluate the potential involvement of Saa3 in regulating tau hyperphosphorylation and to determine whether this effect is unique to Saa3 or generally applicable to other acute-phase SAA proteins, mice were given hippocampal injection of recombinant SAA or saline. The level of total tau and phosphorylated tau at T205 and S396 was determined in the hippocampus of mice receiving SAA or saline injection. As shown in Fig. 4, when normalized against total tau, a decrease in tau phosphorylation at T205 and S396 was observed in mice receiving SAA injection compared with mice receiving saline.

It was shown in Fig. 2 that Saa3 was found primarily in neurons, whereas its expression was absent from microglia. This finding was confirmed in primary cultures of brain cells. At the mRNA level, marked induction of the Saa3 transcript was observed in mouse neurons and astrocytes, but not in microglia (Fig. 5a). Experiments were conducted to determine whether microglia could respond to SAA and play a role in neuroinflammation, as suggested previously [6]. As shown in Fig. 5b and quantified in Fig. 5c, microglia was markedly activated in the hippocampus of mice receiving intracerebral injection of SAA, compared with mice receiving saline only. This result is consistent with our previous report that SAA could induce the activation of microglia in primary culture [6], suggesting a potential role for SAA in microglial activation.

Experiments were conducted to determine whether SAA acts directly on neurons to affect tau phosphorylation. As shown in Fig. 6a, SAA treatment of primary neurons in culture did not have a significant effect on tau phosphorylation, suggesting that SAA might influence tau phosphorylation through an indirect mechanism. Since SAA is known to regulate microglial activation, it was postulated that SAA-treated microglia might release mediator(s) that affect tau hyperphosphorylation. To test this possibility, primary neurons from the mice were cultured with conditioned media (CM) of microglia or astrocyte collected after exposure to SAA (Fig. 6b). As expected, these cells showed reduced tau phosphorylation in the presence of the CM from SAA-treated microglia (Fig. 6c). In comparison, the level of tau phosphorylation was not changed when neurons were cultured with CM from SAA-treated astrocyte (Fig. 6d). This finding confirms our hypothesis that SAA-stimulated microglia release mediator(s) that affects tau phosphorylation.

We found that the level of IL-10 mRNA was lower in Saa3 ^−/− mice than in their WT littermate controls following LPS stimulation (Fig. 7a). Consistent with this finding, the mRNA level of IL-10 was increased after intracerebral injection of SAA (Fig. 7b). There was also a remarkable increase in IL-10 protein production in microglial CM after SAA stimulation (Fig. 7c). To confirm that IL-10 was the mediator of interest, an anti–IL-10 monoclonal antibody was included in the culture. Notably, the effect of microglial CM on tau phosphorylation was partially reversed with the anti–IL-10 antibody (Fig. 7d), suggesting involvement of IL-10 in the regulation of tau phosphorylation.

To further explore the potential involvement of IL-10 in the SAA-induced reduction of tau hyperphosphorylation, mice were given hippocampal injection of recombinant SAA with or without the IL-10 neutralizing antibody. The control mice received hippocampal injection of saline. The level of total tau and phosphorylated tau at T205 and S396 was determined. As shown in Fig. 8, when normalized against total tau, a decrease in tau phosphorylation at T205 and S396 was observed in mice receiving intracerebral SAA injection compared with mice receiving saline. The inclusion of the IL-10 neutralizaing antibody with SAA injection reversed the decrease in tau phosphorylation at T205 and S396.