Attenuation of microglial activation in a mouse model of Alzheimer’s disease via NFAT inhibition

Anti-NFATc1 (NFAT2), anti-NFATc4 (NFAT3), anti-NFATc3 (NFAT4), anti-phosphoNFATc2 (NFAT1), anti-ERK2, and α-tubulin antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-MAP2 antibody was from Cell Signaling Technology Inc (Danvers, MA, USA). HRP-conjugated secondary antibodies were also obtained from Santa Cruz Biotechnology (Dallas, TX, USA). The anti-NFATc2 (NFAT1) antibody and FK506 were obtained from Abcam (Cambridge, MA, USA). Anti-CD68 antibody was from AbD Serotec (Raleigh, NC, USA). Anti-APP antibody was from Zymed Laboratories (San Francisco, CA, USA). Elite Vectastain ABC Avidin and Biotin, Vector VIP, biotinylated anti-rat and anti-mouse antibodies were obtained from Vector Laboratories Inc (Burlingame, CA, USA). Synaptophysin antibody was purchased from Chemicon International, Inc (Temecula, CA, USA). Phospho-NFκB (p65) antibody, NFκB (p65) antibody, PSD95 antibody, and beta site APP cleaving enzyme (BACE) antibody were from Cell Signaling Technology Inc (Danvers, MA, USA). Anti-Aβ (4G8) antibody was from Covance (Emeryville, CA, USA). The c-Fos antibody was purchased from Novus Biologicals, LLC (Littleton, CO, USA). Lipopolysaccharide (LPS) and MTT were purchased from Sigma-Aldrich (St. Louis, MO, USA). Human Aβ1-42 was purchased from Bachem (Torrance, CA, USA) and fibrillized before use according to our prior protocol [80]. A VIVIT peptide sequence, MAGPHPVIVITGPHEE, that interacts with NFAT at its calcineurin binding site and inhibits its activation was used along with its negative control, VEET, peptide [73]. A cell-permeable variation of VIVIT peptide, tat-VIVIT, fused with an HIV-1 TAT protein cationic transduction domain [76,77], was generated to examine the potential of this novel peptide construct for NFAT inhibition as in our prior work [70]. The inhibitory peptide tat-VIVIT (H-YGRKKRRQRRR-AA-MAGPHPVIVITGPHEE-NH₂) and negative control peptide, tat-VEET (H-YGRKKRRQRRR-AA-MAGPPHIVEETGPHVI-NH₂), were synthesized by Dr. Satya Yadav at the Molecular Biotechnology Core Laboratory at the Cleveland Clinic Foundation (Cleveland, OH). Cytokine enzyme-linked immunosorbent assay (ELISA) kits for TNFα, interleukin-6 (IL-6), IL-1β were purchased from R&D Systems (Minneapolis, MN, USA). Fluorescein isothiocyanate (FITC) bioparticles were purchased from Invitrogen (Carlsbad, CA, USA).

All animal use was approved by the University of North Dakota Institutional Animal Care and Use Committee (UND IACUC). The Jackson Laboratory (Bar Harbor, ME) transgenic mouse line, strain 005864 B6.Cg-Tg (APPswe,PSEN1dE9)85Dbo/J, and their wild-type littermate controls were used for this study. Mice were provided food and water ad libitum and housed in a 12-h light:dark cycle. The investigation conforms to the National Research Council of the National Academics Guide for the Care and Use of Laboratory Animals (eighth edition). These APP/PS1 transgenic mice express the human APP, Aβ precursor protein (A4) with the Swedish mutations K595N/M596L, and the human PSEN1, presenilin 1 with the DeltaE9 mutation under control of the mouse prion promoter.

Primary microglia were derived, as described previously [8], from the brains of postnatal day 1 to 3 wild-type C57BL/6 wild type mice. Briefly, meninges-free cortices were removed, trypsinized, and triturated in microglia media (DMEM/F12 media containing L-glutamine (Invitrogen, Carlsbad, CA, USA) and 20% heat-inactivated FBS) and placed in T-75 flasks. Media in the flasks was replaced completely after 24 h and partially after 7 days with fresh media. Cells were harvested, counted, and used at day 14 in vitro.

Neurons were cultured from cortices of wild-type embryonic day 16 (E16) mice (C57BL/6). Meninges-free cortices were isolated, trypsinized, and plated onto poly-L-lysine-coated tissue culture wells containing Neurobasal media with L-glutamine and B27 supplements (Invitrogen, Rockville, MD, USA). Neurons were allowed to grow in the media for 7 days in vitro in order to provide >95% purity.

Jurkat cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA) and maintained in RPMI-1640 medium (Gibco RBL, Rockville, MD, USA) containing 10% heat-inactivated fetal bovine serum (US Biotechnologies Inc., Parkerford, PA, USA), 5 mM HEPES, and 1.5 μg/mL antibiotics (penicillin/streptomycin/neomycin).

To determine the levels of proinflammatory cytokines, cells were pretreated with FK506 or tat-VIVIT for 30 min followed by 10 μM Aβ 1-42 stimulation for 24 h. Media was collected for ELISA.

In order to understand the effect of NFAT inhibition on neuronal cell viability, 7-day in vitro primary neurons were grown in microglial conditioned media for 72 h as previously described [25]. Briefly, 96-well cell culture plates were coated with Aβ 1-42 fibrils (48 pmole/mm²) in the presence or absence of 0.1 μM FK506 and 1 μM FK506 in serum-free Neurobasal media with B27 supplements. Microglia cells were plated on the surface-bound Aβ fibrils for 48 h. Following the stimulation, media from wells with only Aβ (negative control), wells with only microglia, and conditioned media from wells with microglia stimulated with Aβ were transferred to neurons for 72 h. Neurons were also left untreated or treated with 0.1 μM FK506 or 1 μM FK506 added to the neurons along with the conditioned media. Following all the treatments, neurons were fixed with 4% paraformaldehyde and immunostained with anti-MAP2 antibody. As in our prior work [25], surviving cells immunostained positive for MAP2 were counted, averaged, and plotted (±SD). Stimulations were performed in duplicate and repeated three times. A counting grid was placed over the wells to count the number of neurons from eight identical fields for each condition.

The levels of proinflammatory cytokines TNFα, IL-6, and IL-1β were measured using ELISA. Media from cell stimulation experiments was transferred onto ELISA plates and levels measured using the manufacturer’s protocol. For brain and spleen ELISA, flash-frozen hippocampi or spleen tissue was lysed in radioimmunoprecipitation assay (RIPA) buffer and lysates were centrifuged at 14,000 rpm, 4°C, for 10 min to remove any insoluble material. Supernatants were then transferred onto the ELISA plates and cytokine levels were measured and averaged (±SD) as per the manufacturer protocol.

Cell viability was measured using the MTT reduction assay. After 24-h stimulation with Aβ in the presence or absence of FK506 or tat-VIVIT, cells were treated with 0.1 mg/mL MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) for 4 h at 37°C. The media was aspirated and precipitated formazan dissolved with isopropanol. Optical density was measured at 560 nm, averaged (±SD), and plotted.

In order to determine the phagocytic ability of Aβ-stimulated microglia in the presence or absence of NFAT inhibitors, microglia were plated onto 96-well culture plates and were either untreated or treated with 1 μM FK506 for 24 h. Cells were then treated with either 500 nM FITC-conjugated Aβ1-42 fibrils or 0.25 mg/mL FITC-conjugated bioparticles for an additional 6 h. After incubation, the media was removed and fluorescence from extracellular peptide or membrane-bound bioparticles was quenched via a rinse with 0.25% trypan blue. Fluorescence intensity from phagocytosed peptide and bioparticles was measured using a fluorescent plate reader (480-nm excitation and 520-nm emission) and averaged (±SD).

NFAT inhibitors FK506, tat-VIVIT, and tat-VEET were infused subcutaneously into male APP/PS1 mice at 12 months of age. The compounds were delivered via mini-osmotic pumps (model 1004, 0.11 μL/h delivery rate for 28 days, Alzet, Cupertino, CA, USA). Pumps delivered either vehicle (DMSO/HEPES) (n = 8), FK506 (1 mg/kg/day) (n = 7), VIVIT (0.5 mg/kg/day) (n = 8), or negative control scrambled peptide, VEET (0.5 mg/kg/day) (n = 8), for 28 days. At the end of the infusion period, mice were behaviorally tested (day 28 of 28-day infusion) then euthanized and perfused with PBS-CaCl₂, and brains and spleens were rapidly collected. Control untreated APP/PS1 mice were collected at a comparable age of completion, 13 months (n = 6). The right hemispheres of the brains were fixed in 4% paraformaldehyde, and the left hemispheres were flash frozen in liquid nitrogen for biochemical analysis.

Temporal cortices were dissected out of the left brain hemispheres from the control, the vehicle-treated, or the inhibitor-infused mice and lysed, sonicated in RIPA buffer, and quantitated using the Bradford method [81]. The lysates were resolved on 10% SDS-PAGE and transferred to polyvinylidene difluoride membranes for Western blotting using anti-APP, anti-BACE, anti-synaptophysin, anti-PSD95, anti-pNFκB (p65), and anti-pNFATc2 antibodies with anti-α-tubulin, NFκB (p65) (for pNFκB), and NFATc2 (for pNFATc2) antibodies used for loading controls. Western blots were quantified using Adobe Photoshop software. Optical densities (O.D.) of bands were normalized against their respective loading controls and averaged (±SD).

Right hemispheres from brains collected from untreated or NFAT inhibitor-infused mice were fixed in 4% paraformaldehyde and cryoprotected in 30% sucrose (in 0.1 M phosphate buffer) by replacing the sucrose solution every 3 to 4 days at least three times. The fixed tissue was then embedded in a 15% gelatin matrix, and the gelatin block again fixed in 4% paraformaldehyde for 3 to 4 days followed by cryoprotection in 30% sucrose solution. The block was then flash frozen using dry ice/isopentane and serially cut (40 μm) on a freezing microtome. Serial sections were used for immunostaining using anti-Aβ (4G8) and anti-CD68 (macrophage/microglia marker) antibodies using Vector VIP as the chromogen. In order to quantitate immunostaining, ×1.25 pictures of three consecutive sections (960 μm apart) throughout the temporal cortex region were taken. Optical densities from hippocampi from each brain section were measured using Adobe Photoshop software. The optical density values per condition (five conditions), per brain (six to eight brains per condition) and per section (three sections per brain) were obtained, averaged (±SD), and plotted.

T-maze analysis was performed as described by Wenk [82]. Briefly, on the 28th day after initiation of infusion, the mice were placed into the starting arm of the T-maze and the door was raised to allow the mice to walk down the stem and enter an arm. Once a mouse entered an arm with all four paws, it was returned to the starting arm. The door was closed for 30 s, then raised, and the mouse was allowed to move freely until it entered an arm again. The process was repeated for nine trials and the choice of arms that mice entered was noted. The number of alternations per mouse in each condition was averaged (±SD) and plotted.

Data are presented as mean values ± SD. Statistical significances were calculated using one-way ANOVA with Tukey-Kramer’s post hoc comparison and considered significant when p values were <0.05.

In order to compare microglia expression of NFAT to peripheral immune cells, we compared primary murine microglia to the human T cell line, Jurkat. Based upon the fact that NFAT isoforms in both humans and mice each demonstrate a multitude of alternative splice patterns, we expected a diverse range of protein expression patterns per isoform [83]. As predicted, Western blot analysis demonstrated that Jurkat cells expressed multiple molecular weight NFAT isoforms (Figure 1). Although primary murine microglia also displayed detectable amounts of numerous molecular weight species, only NFATc1, NFATc2, and NFATc4, and not NFATc3, were detectable (Figure 1).

In order to determine whether increased NFAT activity was required for altered protein expression in microglia, we next collected media from microglia stimulated 24 h with Aβ or LPS in the absence or presence of tat-VIVIT or FK506 treatments to inhibit NFAT [70,73]. ELISAs for TNFα and interleukin-6 (IL-6) were performed from the media to quantitate any changes in cytokine secretion. Although Aβ treatment alone had an expected slight decrease in cell viability, neither tat-VIVIT nor FK506 further decreased survival (Figure 2). On the other hand, both tat-VIVIT and FK506 significantly attenuated both Aβ- and LPS-stimulated secretion of either cytokine (Figure 2).

Decreasing proinflammatory secretion from microglia is arguably an advantageous condition in the AD brain, but decreasing phagocytic ability in parallel may lead to an undesirable increase in fibrillar Aβ plaque deposition. To determine whether NFAT inhibition attenuated microglial ability to clear fibrillar Aβ via phagocytic uptake, microglia were stimulated with FK506 and uptake of FITC-conjugated Aβ1-42 fibrils was quantified. In contrast to the inhibitory consequence of FK506 treatment with regard to cytokine secretion, drug treatment actually increased phagocytic uptake of Aβ fibrils (Figure 3). Moreover, this was not limited to Aβ. Phagocytosis of FITC-labeled bioparticles was also increased following FK506 treatment suggesting that the benefits were not specific to a particular particle type (Figure 3).

Another demonstrative therapeutic advantage of NFAT inhibition in AD would be evidence of neuroprotection. To generate evidence of whether or not microglial NFAT inhibition could improve neuron survival in AD, we next stimulated microglia with and without Aβ fibrils in the absence or presence of FK506 and transferred this conditioned media to primary cortical neuron cultures. Our prior work has demonstrated that this Aβ-stimulated conditioned media is potently toxic to neuron cultures compared to media from unstimulated microglia [84]. Treatment of the microglia with FK506 attenuated the toxicity of Aβ-stimulated conditioned media (Figure 4). Importantly, this neuroprotection was via microglial inhibition since adding FK506 directly to neurons in the presence of the Aβ-stimulated conditioned media had no ability to offer protection (Figure 4).

In order to test the efficacy of NFAT inhibition for particularly microglial activation in vivo, we next relied on a transgenic mouse model of AD with robust microgliosis [80]. Using an APP/PS1 transgenic mouse model of AD, we tested the efficacy of NFAT inhibition as a therapeutic for AD. APP/PS1 mice were treated, for 28 days via mini-osmotic pumps, with either FK506 or the brain-permeable NFAT inhibitory peptide, tat-VIVIT. Immunostaining for CD68, the microglial marker, demonstrated that both FK506 and tat-VIVIT but not the negative control-scrambled peptide, tat-VEET [74], significantly attenuated CD68 immunoreactivity in the hippocampus of the mice as predicted (Figure 5).

In order to determine whether microglial inhibition resulted in any change in plaque load in these mice, Aβ immunoreactive plaques were stained from serial hippocampal sections from the mice using the anti-Aβ antibody, 4G8. Plaque density was quantified to reveal that both FK506 and tat-VIVIT attenuated Aβ peptide-containing plaque load (Figure 6). This suggested that decreasing reactive microglial phenotype via NFAT inhibition had an additional beneficial effect of also decreasing plaque load.

The brains and spleens of these mice were next used to measure changes in NFAT activity and cytokine levels to determine whether the decreases in microglial and Aβ plaque immunoreactivity correlated with an actual decrease in NFAT activity and anti-inflammatory effects after FK506 or tat-VIVIT treatment. Although FK506 significantly decreased IL-6, TNFα, and IL-1β levels in the spleens of mice, tat-VIVIT treatment significantly attenuated only IL-6 and TNFα levels (Figure 7). Unexpectedly, tat-VIVIT or FK506 treatment had no effect on levels of brain cytokines (Figure 8) in spite of the changes observed in the spleen (Figure 7) and the significant decrease in microglial immunoreactivity observed (Figure 5). This suggests that a higher concentration of either drug may be required for quantifiable brain anti-inflammatory, NFAT inhibitory effects.

In order to continue characterizing the effects of tat-VIVIT or FK506 treatment on the brain, we performed Western blot analyses to quantify any additional differences due to NFAT inhibitor treatment. Although NFATc1 activity was not different between treatment groups, it was possible that other NFAT isoforms were affected by treatments. Active levels of NFATc2 were quantified by measuring any change in phosphorylation state. However, similar to NFATc1 no increase in active, non-phosphorylated NFATc2 resulted from treatments (Additional file: Figure S1). To determine whether the decrease in Aβ plaque levels could be due to altered levels of APP or its processing, APP and BACE were quantified. Again, no differences were observed in any treatment (Additional file: Figure S1). To examine whether changes in other transcription factors might correlate with the effects we observed, protein levels of c-Fos and phosphorylated, active p65 NFκB were also examined. However, similar to NFATc2, no differences were observed between groups (Additional file: Figure S1). Finally, synaptic changes were assessed by quantifying protein levels of the presynaptic marker, synaptophysin, and the postsynaptic marker, PSD95. No differences were observed between groups (Additional file: Figure S1). 1 1 1 1

In order to determine whether an NFAT inhibition-mediated decrease in microgliosis had an effect on behavior of APP/PS1 mice, we performed T-maze analyses to quantify working memory. Mice infused with tat-VIVIT or FK506 did not show any changes in the performance on T-maze as measured by the number of alternations between the T-maze arms (Figure 9).