Distribution of endogenous normal tau in the mouse brain

2.1. Animals and human tissues

In this study, we used wild‐type nontransgenic mice, P301L tau‐transgenic mice (P301L‐Tg; [Kimura et al., 2010]), and tau knockout mice (tau‐KO; [Dawson et al., 2001]). All animal experiments were approved by the institutional animal care and use committees. Both male and female animals were used. The autopsy brain tissues were obtained from the Brain Bank for Aging Research, Tokyo Metropolitan Institute of Gerontology (TMIG), Japan (URL: www.mci.gr.jp/BrainBank/index.cgi) with written informed consents for their use in medical research from the patients or their families. Their use in this particular study was approved by the ethics committee at Doshisha University and TMIG.

2.2. Recombinant proteins

Escherichia coli expression vectors, pRK172‐0N4R human tau was provided from Dr. M. Goedert. Mouse tau cDNA corresponding to the whole sequence of 0N4R isoform was amplified from an adult mouse brain cDNA library (Takara bio Inc., Shiga, Japan). Each cDNA fragment was subcloned into NdeI‐ and EcoRI‐ digested site of pRK172 vector (Goedert & Jakes, 1990). Recombinant tau was purified as described previously (Xie et al., 2014). Briefly, expression of human or mouse 0N4R tau were induced in E. coli, BL21(DE3) by 0.5 mM of isopropyl β‐D‐1‐thiogalactopyranoside for 2 hr. Tau expressing bacteria were harvested and lysed in Homogenize buffer (50 mM PIPES, 1 mM EGTA, and 1 mM DTT [pH 6.4]) with sonication. After centrifugation at 3000g at 4°C for 10 min, the soluble fractions were applied to phosphocellulose columns (P11, Whatmann) and eluted by a stepwise gradient of 0.1 M–0.3 M NaCl. The soluble fractions were then precipitated by dialyzing against saturated ammonium sulfate and collected by ultracentrifugation at 100,000g at 2°C for 15 min. Resultant pellets were resuspended in Homogenize buffer containing 0.5 M NaCl and 2% 2‐mercaptoethanol and heated at 100°C for 5 min. Insoluble materials were removed by centrifugation at 15,000g for 15 min. The soluble heat stable fractions containing tau proteins were further fractionated using reverse‐phase HPLC with Cosmosyl Protein‐R column (Nacalai tesque). These fractions were dried and stored in −80°C. Recombinant N‐terminal fragment of MAP2 used for anti‐MAP2N production was expressed in the E. coli BL21(DE3) strain and the soluble fraction was prepared similar to tau. MAP2N fragment was purified by ammonium sulfate precipitation followed by heating at 100°C for 5 min. After centrifugation, heat stable fraction was subjected to the SDS‐PAGE. Peptide of interest was electro eluted from gel‐piece corresponding to 25 kDa. Purity and concentrations of proteins were evaluated by SDS‐PAGE followed by Coomassie Brilliant Blue staining. We verified that purity of all recombinant proteins used here were better than 95% of total protein.

2.3. Antibody production

Anti‐tauN and anti‐MAP2N rabbit polyclonal antibodies were raised against the N‐term peptide (AEPRQEFEVMEDHAGGGC) of human tau and the N‐terminal 150 amino acid fragment of recombinant human MAP2, respectively. Anti‐human tau peptide rabbit polyclonal antibody was raised against the peptide (GTYGLGDRKDQGGYTMHQGGC). Antirodent tauN (anti‐RtauN) rabbit polyclonal antibody was raised against the peptide (DTMEDHAGDYTLLQDEG) corresponding to the N‐terminal potion of mouse tau. Anti‐total tau (RTM38), anti‐human tau specific (RTM49) and anti‐mouse tau specific (RTM47) rat monoclonal antibodies were raised against purified recombinant human and mouse tau, respectively. All antibodies were screened by one site enzyme‐linked immunosorbent assay using the antigenic peptides or recombinant tau proteins. Positive clones were further screened based on their specificity to tau in immunohistochemistry and Western blotting using wild‐type, tau‐KO, and tau‐Tg mice brains (Dawson et al., 2001; Kimura et al., 2010). In this study, tissue culture media of hybridoma were used here.

2.4. Antibody dilutions

Dilutions of the antibodies for immunostaining are listed below and in Table 1; anti‐tauN, anti‐MAP2N, and anti‐RtauN (1:1000), anti‐tau12 (tau12, mouse monoclonal antibody, 1:2000, Abcam Cat# ab74137, RRID:AB_1281142), tau1 (tau1, mouse monoclonal antibody, 1:1000, Millipore Cat# MAB3420, RRID:AB_94855), anti‐MAP2 (HM2, mouse monoclonal antibody, 1:1000, Sigma‐Aldrich Cat# M4403, RRID:AB_477193), RTM49, RTM47 and RTM38 (1:300), tau5 (tau5, mouse monoclonal antibody, 1:1000, Millipore Cat# MAB361, RRID:AB_94944), BD‐anti‐tau (anti‐tau, mouse monoclonal antibody, 1:1000, BD Biosciences Cat# 610672, RRID:AB_397999), anti‐NeuN (mouse monoclonal antibody, 1:500, Abcam Cat# ab104224, RRID:AB_10711040), anti‐S100b (rabbit monoclonal antibody, 1:500, Novus Cat# NB110‐57478, RRID:AB_922392), Iba‐1 (rabbit polyclonal, 1:1000, Wako Cat# 019–19,741, RRID:AB_839504), anti‐olig2 (rabbit polyclonal, 1:500, Proteintech Group Cat# 13999‐1‐AP, RRID:AB_2157541), anti‐Drebrin (Drebrin, mouse monoclonal antibody, 1:1000, Novus Cat# NB 100–1951, RRID:AB_530937), anti‐VGlut1 (VGlut1, guinea pig polyclonal antibody, 1:1000, Synaptic Systems Cat# 135304, RRID:AB_887878), anti‐Pan‐Axonal Neurofilament maker (SMI‐312, mouse monoclonal antibody, 1:1000, Covance Research Products Inc Cat# SMI‐312R, RRID:AB_2314906), antiphosphorylated neurofilament (SMI31, 1:500, Covance Research Products Inc Cat# SMI‐31P‐100, RRID:AB_2028812), anti‐amyloid β (anti‐human amyloid β[N], rabbit polyclonal antibody, 1:500, IBL, Takasaki, Japan), AT8 (mouse monoclonal antibody, 1:200, Thermo Fisher Scientific Cat# MN1020B, RRID:AB_223648), anti‐MBP (rat monoclonal antibody, 1:200, Abcam Cat# ab7349, RRID:AB_305869) and Alexa 488‐conjugated DM1A (DM1A‐488, 1:100, Millipore Cat# 16–232, RRID:AB_441973).

2.5. Hypothermia induction and Western blotting

Hypothermia was induced by anesthesia as previously reported (Planel et al., 2008). Briefly, adult mice (3–6 months old) were anesthetized by intraperitoneal injections of pentobarbital (Somnopentyl, 50 mg/kg; Kyoritsu Seiyaku, Tokyo, Japan) and kept at room temperature. Mouse body temperature was monitored with a rectal probe (Thermalert TH‐5, Physitemp). After confirming that rectal temperatures drop below 30°C, the mice were subjected to the perfusion fixation as described below. Control mice were immediately perfused after anesthesia induction. For the biochemical analyses, brains were harvested and frozen in liquid nitrogen immediately after confirming that the body temperature, and stored in −80°C. The tissues were homogenized in O+ buffer and prepared the total lysates as described in previously (Planel et al., 2008). Solubilized proteins were fractioned by 10% SDS‐PAGE and electro‐transferred onto polyvinylidene difluoride membrane. After blocking by 5% skim milk in TBS‐T, the membranes were incubated with primary antibodies. Bound antibodies were detected by anti‐IgG antibodies (Jackson Immunoresearch, West Grove, PA), and developed by enhanced chemiluminescence (ECL; Amersham, Buckingham, UK).

2.6. Immunostaining of brain tissue sections

Animals were anesthetized with pentobarbital and fixed via transcardial perfusion with 4% paraformaldehyde (PFA) in phosphate buffered saline (PBS). Heads were post‐fixed in the same fixative for 48 hr at room temperature. Brains were harvested and sliced into 6 μm‐thick paraffin embedded sections or 50 μm‐thick sections using a vibrating microtome (LinearSlicer Pro7, Dosaka EM Co. Ltd, Kyoto, Japan). The formalin‐fixed human brain tissues were sliced at 50 μm‐thickness and stored in PBS. The paraffin embedded sections were deparaffinized and hydrated, and followed by blocked by 10% goat serum in Tris buffered saline (TBS; 150 mM NaCl, 50 mM Tris, pH 7.6). The sections were incubated with primary antibodies diluted in 1% BSA in TBS with 0.1% Tween 20 (TBS‐T) overnight at room temperature. After brief washing by TBS‐T, the sections were immersed in 1% BSA in TBS‐T containing biotinylated second antibodies (Vector Laboratories, Inc., Burlingame, CA). Bound antibodies were labeled by avidin and biotinylated HRP (Vectastain ABC kit; Vector Laboratories, Inc.) and developed with 3,3‐diaminobenzidine (DAB) in the presence of hydrogen peroxide. For fluorescent immunostaining of vibrating microtome sections, the tissue sections were permeabilized with methanol and blocked in 10% goat serum in PBS with 0.1% Tween 20 (PBS‐T) for 60 min. They were then incubated with primary antibodies diluted in 1% BSA‐PBS‐T for 48 to 72 hr at room temperature. After the sections were rinsed with PBS‐T, bound antibodies were visualized with secondary antibodies that were conjugated to Alexa Fluor Dyes (Thermo Fisher Scientific, Waltham, MA). The human autopsy tissue sections were stained by Sudan Black B to eliminate lipofuscin autofluorescence (Miyasaka et al., 2005; Schnell, Staines, & Wessendorf, 1999), and then immunostained as above except that the PBS was used in all procedure instead of PBS‐T. The specimens were observed under a LSM700 (Carl Zeiss Inc., Jena, Germany), a TCS SP8 LASER scanning confocal (Leica Microsystems, Wetzlar, Germany), or a Leica TCS SP8 STED (Leica Microsystems) microscope using ×10 and ×40 dry objectives and ×63 and ×100 oil objectives.

3.1. Generation and validation of novel anti‐tau antibodies

To investigate the tissue distribution of tau unambiguously, we raised anti‐tau antibodies, as most of commercially available antibodies are not sufficient to detect endogenous unaggregated tau. Therefore, we raised phosphorylation‐independent anti‐tau antibodies, namely, RTM38 and anti‐tauN (which react with both mouse and human tau), RTM49 (human tau specific), and RTM47 and anti‐rodent tauN (mouse tau specific) (Figure 1). These antibodies showed strong immunolabeling on the mossy fiber axons in the wild‐type nontransgenic (non‐Tg) or human tau transgenic mice (Figure 1a), which were absent in either tau knockout (tau‐KO) or non‐Tg mouse hippocampi. The sensitivity and specificity of these tau antibodies were also demonstrated by Western blot analysis against recombinant human and mouse tau proteins (Figure 1b).

To demonstrate that our new antibodies are phosphorylation‐independent in immunohistochemistry, we induced hyperphosphorylation of tau in the mouse brain by applying anesthesia‐induced hypothermia to non‐Tg mice (Planel et al., 2008). This procedure has been shown to induce robust phosphorylation in multiple sites (Planel et al., 2008). We confirmed on Western blotting that the immunoreactivity of an anti‐phosphorylated tau antibody, AT8, was greatly increased in the hypothermia samples, whereas that of an anti‐phosphorylated tau antibody, tau1, was decreased (Figure 2a). RTM38 detected both higher molecular weight phosphorylated‐ and low molecular weight unphosphorylated‐tau (Figure 2a), and the total intensity of the immunoreactive bands were not different between the samples from control and hypothermic animals (Figure 2b). In contrast, the immunoreactivity of AT8 and tau1 significantly changed in these samples (Figure 2b). We then tested our antibodies on tissue staining. As expected, the immunoreactivity of tau1 was also greatly diminished in immunofluorescence staining of the hippocampus in hypothermic animals (Figure 2c). Thus, it is highly possible that tau1, which had been used for immunostaining of tau, had missed tau phosphorylated in its epitope. In contrast, RTM38 and anti‐tauN (not shown) antibodies exhibited consistent signals in both conditions. These results therefore indicate that this method can reliably visualize the distribution of tau from physiological to pathological state in tissue sections.

3.2. Overall distribution of endogenous tau in the mouse brain

Using these new antibodies, we investigated the detailed localization of endogenous mouse tau in non‐Tg mouse brains. As shown here, our new antibodies are available for the immunostaining of paraffin embedded sections. DAB immunostaining with RTM38 showed global but nonuniform tau labeling in the mouse brain (Figure 3a,b). Most regions of the brain were diffusely labeled including the cerebral cortex, thalamus, basal ganglia, and hypothalamus. However, there were several regions showing robust and fibrous labeling of tau as described below. The labeling was absent in tau knockout mouse brain sections (Figure 3a inset), confirming the specificity of the staining. We also provide a negative control without the primary antibody (Figure 3b inset).

3.3. Tau labeling in myelinated and nonmyelinated axons

We confirmed that tau did not co‐localize with MAP2, which is a dendritic MT‐associated protein, evident in the hippocampal CA1, and that they were separately distributed in the axonal and somatodendritic compartments of adult neurons, respectively (Figure 7) (Dehmelt & Halpain, 2005; Deshpande, Win, & Busciglio, 2008; Xie & Miyasaka, 2016). As described earlier, tau immunoreactivity was evident in regions rich in nonmyelinated axons such as the hippocampal mossy fiber and the cerebellar molecular layer. In contrast, the staining was relatively poor and sparse in white matters, including the corpus callosum (cc), fornix (f), anterior commissure (ac), and the white matter of the cerebellum (Cb) (Figure 3a). Furthermore, triple labeling of the corpus callosum with anti‐tauN and antibodies for MBP and neurofilament revealed that tau is most abundant in areas where MBP and neurofilament immunoreactivity are relatively weak (Figure 8a,b). High magnification views and line scan analysis of fluorescence intensity further revealed that tau immunoreactivity was poor on MBP‐positive axons containing neurofilaments (Figure 8c). This was also evident in the hippocampal CA3. MBP‐positive axons in the area were virtually void of tau immunoreactivity, while the neighboring mossy fiber axons, which are nonmyelinated, are strongly labeled (Figure 8d).

3.4. Expression of mouse tau in glial cells

We also addressed tau expression in glial cells, which has been implicated previously (LoPresti, Szuchet, Papasozomenos, Zinkowski, & Binder, 1995; Seiberlich et al., 2015). In consistent with these reports, our anti‐mouse tau antibodies labeled small cells in the hippocampus (Figure 5), in which most of the signal was observed in the cytoplasm. These cells are not neurons, as they were negative for NeuN (data not shown). We therefore labeled each class of glial cells using antibodies for well‐known glial markers, S100β for astrocytes, Iba‐1 for microglia, and Olig‐2 for oligodendrocytes. Anti‐S100β and Iba‐1 antibodies labeled respective glial cells in the hippocampal CA3 area but not those positive for tau (Figure 9a,b). Only the anti‐Olig‐2 antibody labeled the nuclei of cells, of which cytoplasm was positive for tau (Figure 9c,d). Virtually all tau positive nonneuronal cells were positive for Olig‐2 throughout the brain regions we investigated including cerebral cortex, corpus callosum, internal capsule, anterior commissure, cerebellum, and medulla oblongata (data not shown). These results indicate that the endogenous tau is abundantly expressed in the cell bodies of oligodendrocytes, but not in astrocyte and microglia in the mouse brain.

3.5. Nondendritic and nonsynaptic localization of tau in the hippocampus

We also found that the tau immunoreactivity appeared punctate, particularly in the stratum radiatum of the hippocampal CA1 area (Figure 7a), suggesting that tau may colocalize with synapses, as previously postulated (Ittner et al., 2010; Zempel et al., 2010). To test potential peri‐synaptic localization of endogenous tau, we therefore examined the localization of tau in the brains of Thy1‐EGFP mice (Feng et al., 2000). In these mice, dendritic spines in individual neurons can be easily identified because of the sparse EGFP expression. Triple labeling of EGFP, tau, and MAP2 combined with line scan analyses confirmed that punctate tau immunoreactivity was scattered around dendritic spines but was never found within them (Figure 10a). Tau did not colocalize with drebrin or VGlut1, which are markers for spines and presynaptic terminals, respectively. This punctate axonal and peri‐synaptic localization of tau was also confirmed in the human hippocampus using our anti‐human tau antibodies (Figure 10b). Thus, in physiological condition, tau is localized within the axonal compartment in small clusters.

3.6. Discontinuous distribution of endogenous tau on the microtubules of axons

The punctate labeling of axonal tau was surprising for a MAP, which is expected to decorate the entire MT and appear fibrous under light microscopy. We therefore investigated the localization of tau at a super‐resolution (~50 nm) using the stimulated emission depletion (STED) microscopy (Tonnesen & Nagerl, 2013) to test whether tau is distributed sparsely on axonal MTs. Immunofluorescence staining of the hippocampal CA3 with anti‐tauN and anti‐tubulin (DM1A) antibodies showed overlapping but not colocalizing distribution of these proteins in mossy fiber axons, but not in apical dendrites of CA3 pyramidal neurons, under confocal microscopy (Figure 11a). With STED, tau labeling appeared more punctate and sparsely distributed on tubulin labeling (Figure 11a). The mossy fiber axons in CA3 have a diameter of 200–500 nm and are typically packed with several MTs observable with electron microscopy (Amaral, Scharfman, & Lavenex, 2007). Despite this density of MTs, tau labeling did not appear fibrous but punctate even in these axons (Figure 11b). This unique labeling was reproduced using the RTM38 antibody, of which epitope differs from that of anti‐tauN and is in the C‐terminal region of tau (data not shown). Therefore, the sparse labeling of endogenous tau may not be due to epitope masking. Considering that tubulins constituting MTs are far more abundant than tau in neurons and even in axons (Miyasaka et al., 2018), these data indicate that endogenous tau localized in the axon does not occupy all the binding sites of axonal MTs in the mouse brain.