Neurofibrillary pathology made up of hyperphosphorylated tau is the common feature of a family of neurodegenerative disorders collectively termed tauopathies, which include Alzheimer’s disease (AD), argyrophilic grain disease (AGD), Pick’s disease (PiD), progressive supranuclear palsy (PSP), corticobasal degeneration (CBD) and several others (Tolnay and Probst, 1999; Lee et al., 2001). In the brain with AD or AGD, tau pathology is initiated in the entorhinal cortex (EC) and follows a stereotypical pattern to progressively propagate to the limbic system and eventually to widespread isocortex regions (Braak and Braak, 1991; Saito et al., 2004; Braak and Del Tredici, 2011). Of note, in AD cases some brain areas such as the cerebellum are relatively spared from tau pathology (Larner, 1997). Thus, it seems that some regions of the brain are more vulnerable whereas the others are more resistant to the development of tau lesions (Kaufman et al., 2016). Despite that microarray analysis has revealed different mRNA expression profiles in different brain regions (Liang W. S. et al., 2007; Liang et al., 2008), and that phylogenetic diversity of tau isoforms expressed in different mammalian species has been proposed to link to particular vulnerability of the human brain to tau pathology (Janke et al., 1999), the exact nature of molecular basis underlying the differential vulnerability to tau pathology is not understood.

Although several transgenic animal models of tauopathy, which overexpress wild-type or mutated human tau up to many folds of the endogenous murine tau, have been developed in the last two decades (Andorfer et al., 2003; Oddo et al., 2003; Santacruz et al., 2005; Yoshiyama et al., 2007; de Calignon et al., 2012; Rockenstein et al., 2015), filamentous tau pathology has not been observed in wild-type rodents expressing murine tau only (Ojo et al., 2016), suggesting that the level of tau expression may be critical for the development of tau pathology. Tau is known to be phosphorylated by several protein kinases and dephosphorylated by several protein phosphatases (Wang et al., 2007) and therefore, hyperphosphorylation of tau can be considered as the consequence of an imbalance between activities of tau kinases and tau phosphatases (Iqbal et al., 2016). However, whether tau and its kinases and phosphatases are differentially expressed across brain regions and how they are associated with the susceptibility of tau lesions in the corresponding region remain elusive.

To understand the possible role of differential topographic expression of proteins involved in regional-specific vulnerability to tau pathology, in the present study we first compared the phosphorylation and total levels of tau protein between frontal cortex (FC) and cerebellum in AD and in control brains. We then dissected nine different brain regions of the rat (Rattus norvegicus), a species known to have phosphorylated tau and predominantly tau with four microtubule-binding repeats yet all six isoforms of tau in the brain (Gärtner et al., 1998; Janke et al., 1999; Hanes et al., 2009), and measured biochemically the protein expression levels of tau, tau kinases, tau phosphatases and other tau pathology-related proteins in different brain regions.

Tau pathology in AD brain occurs sequentially in the EC, the hippocampus and widespread isocortex areas, but is never seen in the cerebellum (Braak and Braak, 1995; Larner, 1997; Braak and Del Tredici, 2011). It is of significance to understand the molecular basis of the differential vulnerability/resistance of different brain regions to tau pathology. In the present study, we found that unlike hyperphosphorylation of tau in the FC, tau was not hyperphosphorylated in the cerebellum of the same AD brains. Cerebellum expressed one fourth of the level of tau in the human FC. Similar to human brain, the rat cerebellum expressed much less tau than EC, the hippocampus and other regions of the forebrain. We observed comparable levels of 3R-tau, but different levels of 4R-tau, expression in different rat brain regions. Since 4R-tau is the dominant isoform of tau in the rat brain (Janke et al., 1999; Hanes et al., 2009), it is likely that regional differential expression of 4R-tau may account for varying expression levels of total tau in different brain regions. EC expressed the highest level of tau, from where tau lesions are initiated, and lowest expression of tau in cerebellum, which is spared from tau pathology. Thus, the level of tau may determine or play an important role in the onset and severity of tau pathology. Consistent with this speculation, inhibition of tau expression in tau transgenic mice by anti-sense oligonucleotides not only prevents but also rescues tau pathology (DeVos et al., 2017) and spatial reference memory (Fu et al., 2017), and injection of hyperphosphorylated tau seeds into the brain induces pronounced tau pathology in transgenic mice overexpressing human tau but only produces minimal tau pathology in wild-type mice (Clavaguera et al., 2009; Hu et al., 2016). In the present study we determined the level of rat tau in brain homogenate. We would not expect marked presence of insoluble tau species in the brain of 5–6-month old wild-type Wistar rats, however, the presence and possible difference in the level of insoluble tau between brain regions is not completely excluded, which merits future investigation.

Tau is the major microtubule-associated protein in the central nerve system. Microtubules, the essential structures for intracellular transport and the maintenance of neuronal morphology, are non-covalently linked cytoskeletal polymers composed of heterodimer subunits of α- and β-tubulins. α-tubulin was expressed comparably throughout all rat brain regions with a slightly less expression in the FC, OTC and OB. βIII-tubulin, a neuronal specific β-tubulin, was highly expressed in the cerebellum, but expressed at an extremely low level in the OB. MAP1B is prominently expressed during early neuronal development but in the adult brain it remains in areas with high synaptic plasticity (Tortosa et al., 2011; Villarroel-Campos and Gonzalez-Billault, 2014). We observed that in adult rat brain the hippocampus, EC, FC and OB expressed higher level of MAP1B, suggesting that these areas have high synaptic plasticity. MAP2 is known to be expressed in dendrites of mature neurons, and in the cerebellum it is preferentially localized in finer and distal branches of Purkinje cell dendrites (Chauhan and Siegel, 1997); low level of MAP2 found in the present study in cerebellum, striatum and thalamus is consistent with low dendritic mass in these brain regions.

Tau lesion is made up of hyperphosphorylated tau, which could result from up-regulation of tau kinases or/and down-regulation of tau phosphatases. Multiple kinases, including cdk5, GSK-3β, PKA, CK1ε, Dyrk1A and CaMKII, are known to phosphorylate tau in vitro (Wang and Liu, 2008). There are two types of tau kinases, proline directed kinases (PDPK) and non-proline directed kinases (NPDPK) that phosphorylate Ser/Thr residues followed by proline or not proline, respectively (Avila, 2008). Tau contains 17 Ser/Thr proline directed sites and 63 Ser/Thr non-proline directed sites. Cdk5, GSK-3β and Dyrk1A belong to PDPK, and PKA, CaMKII and CK1 are NPDPK (Iqbal et al., 2005). Up-regulation of GSK-3β (Jin et al., 2015b), cdk5 (Lee et al., 2000), CK1ε and δ (Kannanayakal et al., 2006), and Dyrk1A (Jin et al., 2015a) and down-regulation of PKAcα and PKAcβ (Liang Z. et al., 2007) have been reported in AD brains. In the present study, we found that the cerebellum, in which there are no overt tau hyperphosphorylation or tau pathology, expressed lower level of PKAcα, PKAcβ, cdk5, Dyrk1A, GSK-3α and CaMKII than the EC, while EC expressed relative higher levels of CK1ε, PKAcα, PKAcβ and Dyrk1A.

It has been reported that major four phosphatases, PP1, PP2A, PP2B and PP5, catalyze tau dephosphorylation in vitro and in cultured cells (Liu et al., 2006). PP2A is the predominant tau phosphatase and accounts for ~70% of tau phosphatase activity in the human brain (Liu et al., 2005a). Its activity toward tau is regulated by the methylation of its catalytic subunit at leu309 (Longin et al., 2007; Janssens et al., 2008). PP2A is down-regulated in AD brain and this downregulation may contribute to tau pathology (Gong et al., 1994; Liu et al., 2005a). Activities of PP1 and PP5 are also reduced (Gong et al., 1994; Liu et al., 2005c), but PP2B activity is increased because of the proteolysis by activated calpain in AD brain (Liu et al., 2005b; Qian et al., 2011). Here, we found that rat cerebellum expresses low levels of PP1, PP2A and PP2B and higher level of PP5.

Amyloid plaques are another hallmark brain lesion of AD. In addition, TDP-43 pathology is a hallmarks of ALS, but it also appears in some AD cases (Higashi et al., 2007; Josephs et al., 2014). In the present study we found that the cerebellum expressed low level of APP, but the highest level of TDP-43. We recently found that TDP-43 suppressed tau expression and tau exon 10 exclusion (Gu et al., 2017a,b). Thus, the high level of TDP-43 in the OB and the cerebellum may contribute to the low expression of 4R-tau in these regions.

Glial cells, which function primarily as the physical support for and nutrifying neurons and regulating the internal environment of the brain, account for nearly half the number of cells in the human brain (von Bartheld et al., 2016). During early embryogenesis glial cells direct the migration of neurons and produce neurotrophic factors that modify the growth of axons and dendrites (Campbell and Götz, 2002). Glial cells also play an important role in synaptic plasticity (De Pittà et al., 2016). Gliosis is another feature of AD and related neurodegenerative diseases (Leyns and Holtzman, 2017). Over-activation of microglia is found around Aβ plaques, which suggest the involvement of microglia in Aβ clearance (Mattiace et al., 1990). Recent studies have shown that microglia participate in the spread of tau pathology (Asai et al., 2015). In the present study, we found that EC expressed the highest and cerebellum expressed the lowest level of Iba1, and that EC expressed high level, but other cortical regions expressed low level of GFAP. CNPase is the marker of oligodendroglia (Kasama-Yoshida et al., 1997). High level of CNPase observed in striatum and thalamus is consistent with higher content of white matter in these regions. Strikingly, the level of Iba1 was very low in the cerebellum, indicating a possible low population of microglia in this brain region.

NeuN is widely used as a neuronal marker and expressed in the neuronal nucleus (Dredge and Jensen, 2011). Cerebellum is known to occupy 80% of brain’s neurons with 10% of brain’s mass (Azevedo et al., 2009). The ratio of glia/neuron is ~0.2:1 in cerebellum, ~4:1 in telencephalon and ~10:1 in the rest of brain regions (Azevedo et al., 2009; von Bartheld et al., 2016). In the present study, we observed that cerebellum expresses the highest while the striatum and thalamus the lowest level of NeuN, which is consistent with the highest density of neurons in the cerebellum (von Bartheld et al., 2016).

In the present study, we found that tau is hyperphosphorylated in the cerebral cortex but not in the cerebellum in AD brain. Cerebellum expresses only one fourth the level of tau seen in the frontal and temporal cortices in the human brain. The rat brain expresses comparable levels of 3R-tau throughout different regions, but the cerebellum expresses the lowest and EC the highest level of 4R-tau. The rat EC expresses higher levels of major tau kinases such as CK1ε, PKAcα and PKAcβ but similar level of PP2A, the major tau phosphatase, as compared to the hippocampus. Because the EC and the hippocampus are both critical brain regions involved in memory encoding and retrieval (Suh et al., 2011; Basu et al., 2016), the high neuronal activity in the EC may leave tau more prone to phosphorylation and therefore higher risk of developing tau lesions. In contrast, the rat cerebellum only expresses very low level of PKAcβ, cdk5 and CaMKII when taking into account the relative level of PP2A. Importantly, the average neuronal load of tau and tau kinases is even much lower in the cerebellum, considering the high density of neurons in it (Azevedo et al., 2009). In addition, other tau pathology-related proteins are also expressed differentially across various brain regions.

Taking together, higher levels of tau and/or tau kinases in the EC and low levels of these proteins in the cerebellum may account for the vulnerability and resistance of these representative brain regions to the development of tau pathology, respectively. The present study, for the first time, provides a critical reference point for further studies to unravel the molecular mechanisms involved in the selective vulnerability/resistance of particular brain regions to tau pathology in AD and related neurodegenerative disorders.

FL, WH, C-XG, and KI: conception of the research. WH, FW, YZ, and FL: performing experiments. WH, FW and FL: analyses and interpretation of results. WH and FL: drafting of the manuscript. C-XG, KI, and FL: critical revision of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.