Mouse and Human Tau Expression in Different Brain Areas

Animals and tissue processing

B6.129S4(Cg)-Mapttm1(eGFP)Klt/J mice (C57 BL6J background) from The Jackson Laboratory (stock 029219) with two GFP copies (here referred to as Tau^GFP/GFP) are knockout mice for the microtubule-associated protein tau (MAPT) gene due to insertion of the transgene GFP [11]. Tau^GFP/+ mice with a single copy of GFP are hemizygous mice for MAPT gene, while the littermate Tau^+/+ are the wild-type mice. Mice were housed in a specific pathogen-free colony facility, in accordance and handle following European and local animal care protocols. Male and female mice (2, 6, 12, and 18 months old) were used for all experiments. Animal experiments received the approval of the CBMSO’s (AECC-CBMSO-13/172) and national (PROEX 102.0/21) Ethics Committee. Mice were fully anesthetized by an intraperitoneal injection of pentobarbital (Dolethal, Vetoquinol, 60 mg/mL) and transcardially perfused with 0.9% saline.

Brain samples

All human brain samples from hippocampus and frontal lateral cortex were obtained from Banco de Tejidos (Fundación CIEN, Instituto de Salud Carlos III, Madrid, Spain). Cerebellum samples were kindly provided by Dr. Isidro Ferrer from Bellvitge University Hospital, IDIBELL (Bellvitge Biomedical Research Centre), Barcelona, Spain. Hippocampus (HPC) and frontal lateral cortex (FLC) samples were from control subjects (not AD associated pathology). Due to absence of control subjects from cerebellum, samples on very early stages of the disease (Braak I and II stages of tau-related AD pathology) were used. See Table 1 for a detailed description of each subject and the regions of brain available from each individual. Written informed consent was obtained premortem from all patients.

RNA extraction and RT-qPCR

Brain samples were extracted using Qiazol lysis reagent (79306, Qiagen). One stainless steel bead (69989, Qiagen) was introduced in an Eppendorf with a small portion of each brain sample and was homogenized through agitation for 4 min at a frequency of 30/s on a Qiagen Retsch MM300 TissueLyser. All RNA extracts were used by qPCR. Relative expression of TIR¹²-MAPT was determined by qPCR. cDNA was isolated from previously extracted RNA samples with the PrimeScript 1^st strand cDNA Synthesis Kit (6110A, Takara) using oligo dT primer, during 45 min at 42°C, according to manufacturer’s protocols. Gene expression was then determined using quantitative PCR with fast SYBR Green reagent (4385612, Applied Biosystems) on a CFX384 Touch Real-Time PCR Detection System (Bio-Rad). The reaction per well was 10μL and contained: 25 ng of cDNA template per sample, 5μL of the SYBR Green PCR mix, and 1μL per oligonucleotide pair at 5μM (see Table 2). Amplification was carried out for 40 cycles of 1 s at 95°C followed by 20 s at 60°C. Three technical replicates per gene were used. Each primer pair showed a single, sharp peak, thereby indicating that the primers amplified only one specific PCR product. No amplification from the no template control was observed for genes of interest.

Immunofluorescence

Mice whose samples were employed were fixed with 4% PFA in 0.1 N phosphate buffer (PB) for 24 h at 4°C. They were then washed three times in PB and placed in a 10% sucrose/4% agarose matrix. Sagittal brain sections were obtained on a Leica VT1000 S vibratome (50μm). The sections were subjected to a floating immunolabeling process. For immunofluorescence after performing three washes with PB 0.1 N, the sections were incubated with the following primary antibodies in the blocking solution (1% BSA and 1% Triton X-100 in 0.1 N PB) at 4°C for 24 h: anti-Olig2 (rabbit, 1 : 1000, Millipore); anti-Iba1 (guinea pig, 1 : 1000, Synaptic systems); anti-GFAP (chicken, 1 : 500, Invitrogen); and anti-NeuN (mouse, 1 : 1000, Millipore). Next, five washes were performed with this same blocking solution, and the sections were incubated with the corresponding secondary antibodies conjugated with Alexa fluorophores at 4°C with gentle agitation for 24 h (1 : 1000, Molecular Probes). Finally, three washes were performed with PB and samples were incubated with DAPI diluted 1 : 5000 for 10 min, followed by three additional washes with PB. The sections were placed on slides using FluorSave Images that were taken using a Confocal Spinning Disk SpinSR10 attached to an IX83 inverted microscope (Olympus).

Protein extracts and western blots

Mouse brain tissue preserved at –80°C was homogenized to obtain the extracts in RIPA buffer as described before [9]. Next, the protein concentration of each homogenate was determined by the Bradford method [12], using the BCA test (Thermo Fisher). Finally, the SDS-PAGE buffer used in [9] was added to the protein extracts obtained. The extracts were boiled in a thermoblock at 100°C for 5 min. Thirty micrograms of protein from each sample were loaded in as many wells. The proteins were separated on 10% acrylamide/bisacrylamide gels in the presence of SDS at 120 mV for approximately 1 h. Those present in the gel were transferred to nitrocellulose membranes (Schleicher and Schuell, Keene, NH, USA) at 150 mA for 45 min, using the Bio-Rad Mini-Protean system. Subsequently, the membranes were blocked using 5% milk powder in 0.1% Tween PBS for 1 h. The membranes were then washed twice with 0.1% PBS Tween-20 (v/v) under stirring for 10 min. The membrane was stained with Ponceau dye (Ponceau 0.3% in TCA 3%) to check the transfer efficiency. Finally, they were incubated with the appropriate primary antibody overnight at 4°C: anti-total Tau5 (mouse, 1 : 1000, Calbiochem); anti-GFP (rabbit, 1 : 1000, Invitrogen); and anti-Vinculin (rabbit, 1 : 5000, Abcam) as load control. Protein expression was detected using the secondary antibody (1 : 1000), which were incubated for 1 h at room temperature. After performing three 10-min washes in the wash solution, the immunoreactive proteins were detected using an ECL kit (Enhanced Chemiluminescence Detection System, Amersham).

Statistical analysis

Mean data were analyzed with one-way ANOVA followed by Tukey’s test post hoc analysis or by two tails T-test. Differences between groups were considered significant at *p<0.05, **p<0.01, and ***p<0.001.

Characterization of TauGFP knock-in/knock-out mice

TauGFP knock-in/knock-out mice [11] have an eGFP-coding sequence inserted into the first exon of the microtubule-associated protein tau gene (Mapt), disrupting expression of the gene and generating a cytoplasmic eGFP expression under the endogenous tau promoter (Fig. 1A–C). This model allows us to analyze where tau promoter is expressed at each moment. Thus, we firstly analyzed protein expression in cortex, hippocampus, and cerebellum in Tau^GFP/GFP, hemizygous (Tau^GFP/+),and wild type (Tau^+/+) mice (Supplementary Figure 1). These experiments demonstrated that the amount of tau protein and eGFP is halved in Tau^GFP/+ mice compared with Tau^+/+ and Tau^GFP/GFP respectively. This proves that protein levels correlate with the number of tau and GFP gene copies.

OPEN IN VIEWER

Fig. 1

B6.129S4(Cg)-Mapttm1(eGFP)Klt/J Mice (Tau ^GFP/GFP mice). A) Fluorescence loupe image of a brain coronal section in a Tau^GFP/GFP and Tau^+/+ (wild type) mice. B) Gene Mapt schematic with GFP insertion in the first exon and (C) scheme of the resulting protein with the first 31 amino acids of tau protein followed by cytoplasmic eGFP protein that is expressed under the endogenous tau promoter. D) Representative fluorescence confocal microscopy images with different markers of glial cells as Olig2 (oligodendrocytes), Iba1 (microglia), GFAP (astrocytes), and NeuN (neurons) and their colocalization with the endogenous signal in Tau^GFP/GFP mice. Scale bar 10μm.

Next, we checked the endogenous expression of tau in different cell types analyzing colocalization of Olig2, NeuN, Iba-1, and GFAP proteins with eGFP. As shown in Fig. 1D, there are endogenous expression in oligodendrocytes and neurons. Nevertheless, there are not endogenous expression of tau in microglia or astrocytes. Interestingly, we can observe that GFP signal in oligodendrocytes is more localized and intense, meanwhile the neuronal signal is diffuse. In the Fig. 1D, it can be observed how the eGFP protein spreads throughout the neuronal cytoplasm, including the axon (arrow in Fig. 1D).

Expression of GFP under endogenous tau promoter during aging in different brain areas

Previous studies have reported differences in tau protein levels in different brain regions [7, 8]. To analyze endogenous tau promotor expression in cortex, hippocampus, and cerebellum, we performed western blot in Tau^GFP/GFP mice at different ages (2, 6, 12, and 18 months). Figure 2A shows a representative western blot of cortical extracts. Figure 2B shows quantification of eGFP protein present in these brain areas in Tau^GFP/GFP mice during aging. Expression levels remain constant in all cases suggesting that the endogenous tau promoter in Tau^GFP/GFP mice do not change with aging in any structure. However, brain samples are complex and western blot does not allow us to analyze gene expression at cellular level. Thus, to analyze in more detail tau promoter expression, we quantified the intensity of eGFP autofluorescence in saggital sections of Tau^GFP/GFP mice (Fig. 2 C). Thanks to this approach, it has been possible to analyze the dentate gyrus and CA1 separately and not as a whole. Thus, it can be seen that we did not observe significant age-related differences in cortical regions. However, when the hippocampus was analyzed, some differences could be noted. Indeed, a tendency towards increased tau levels with age was observed in the CA1 region while a tendency towards decreased levels was observed in the dentate gyrus. These results suggest that the tau promoter is not expressed at a similar rate in pyramidal and granule neurons.

OPEN IN VIEWER

Fig. 2

Evolution of eGFP protein level in function of aging at 2, 6, 12, and 18 months. A) Representative western blot of cortex at different ages. B) eGFP protein quantification in Tau^GFP/GFP animals in function of aging in cortex, hippocampus, and cerebellum. C) Representative fluorescence confocal microscopy images in cortex, CA1, and dentate gyrus. The localization of eGFP was determined by direct fluorescence (green) in 12-month-old Tau^GFP/GFP mice. DAPI staining is shown in blue. Scale bar 20μm. D) eGFP intensity quantification in immunofluorescence sections in cortex and hippocampal region as dentate gyrus and CA1 at 2, 6, 12, and 18 months. To analyze eGFP expression we used 2-month-old mice to normalize the data. Thus, the intensity measured in 2-month-old mice was arbitrarily given a value of 100. The graph shows the mean value±SEM (n = 5–6 per group). No differences were found using two-way ANOVA.

Tau expression in human brain areas

Human MAPT expression by RT-qPCR was analyzed (Fig. 3A) in three different brain regions (hippocampus, cerebellum, and frontal lateral cortex) in control human brains. When analyzing differences between brain regions in non-demented individuals, we found that MAPT expression is consistently maintained on similar levels both across regions, like in mice brain (Fig. 2).

OPEN IN VIEWER

Fig. 3

Differences in expression mRNA of total MAPT and TIR ¹² -MAPT in hippocampus, cerebellum, and frontal lateral cortex. A) Relative expression of total MAPT in human control samples in hippocampus (HPC), cerebellum (CBL), and frontal lateral cortex (FLC) (B) and relative expression of TIR¹²-MAPT at the same samples. The graphs show the mean value±SEM (n = 3–10). **p = 0.004; ***p = 0.0008) using one-way ANOVA.

To go one step further, we have measured TIR¹²-MAPT expression in same samples. TIR¹²-MAPT has been recently described [10] as a new, human-specific tau isoform, which is generated by intron 12 retention in MAPT. This new tau isoform is able to bind to microtubules and is less prone to aggregate, suggesting a beneficial role in the pathology. As we can see in Fig. 3B, there are significant differences in expression between hippocampal regions, frontal lateral cortex, and cerebellum, being the amount of mRNA ^{TIR12 - MAPT} lower in cerebellum.