Tau pathology in the olfactory system and associated CNS regions in PS19 mice

The presence of hyperphosphorylated tau protein (pTau) was evidenced using the AT8 antibody (pTau (Ser202, Thr205)) in coronal sections of the OE from WT and PS19 mice. To that end, a specific protocol of decalcification has been set up to preserve the integrity of the OE. AT8 immunoreactivity was observed only in PS19 mice and was localized to the OE middle stratum and to the axon bundles (ABs), located in the lamina propria underlying the OE (Fig. 1a-d). The AT8 signal in the OE was assessed and compared at different ages: 1.5, 3, 6, and 9 months (Fig. 1a-d). AT8 immunoreactivity was detected as soon as 1.5 months (Fig. 1a, a’), increased progressively up to 6 months (Fig. 1c, c’), and decreased at 9 months, especially in the middle stratum (Fig. 1d, d’). This region did not exhibit any significant bilateral asymmetry in tau pathology. Based on morphology, no NFT-like tau accumulations were observed in this region for up to 9 months. A closer examination of the nasal cavity showed no pTau signal in the RE of PS19 mice (Fig. 1e). No pTau signal was detected in the OE or ABs of 6-month-old WT mice (Fig. 1f), indicating that the AT8 signal is specific to PS19 mice and related to the expression of pathological human tau. The results were confirmed independently by Western Blotting (Fig. S1) using the PHF-1 antibody. This antibody is generated against paired helical filaments isolated from the brains of AD patients, that recognizes tau phosphorylated at Ser396 and Ser404. It binds to tau filaments and is a recognized standard to evidence the pathological conformations found in NFTs [30, 50]. Immunostaining with the PHF-1 antibody and a human tau (hTau)-specific antibody showed similar results to those obtained with the AT8 antibody in PS19 mice (Fig. S2).

Fig. 1 

pTau expression in the OE of PS19 mice. The OE from 1.5-month (a, a’), 3-month (b, b’), 6-month (c, c’, e), 9-month-old PS19 mice (d, d’), and 6-month-old WT mice (f) were fixed and stained with the AT8 antibody (pTau (Ser202,Thr205)). pTau recognized by the AT8 antibody is found in the middle stratum of the OE (white arrows) and in the ABs (black arrows) as soon as 1.5 months in PS19 mice (a, a’). The signal increases up to 6 months (c, c’, e) and appears to decrease at 9 months (d, d’). No signal is detected in WT mice (f). All the mice used were males. The immunostainings shown are representative of n = 4 mice per time point. High magnification of the OE (white arrows) is shown in the inset (e’). ABs: axon bundles, OE: olfactory epithelium, RE: respiratory epithelium

Additional staining with a total tau antibody, detecting both endogenous murine tau protein and human tau protein, confirmed the presence of endogenous tau in the OE of WT mice (Fig. S2). RT-qPCR experiments indicated that the expression of endogenous tau transcripts was stable from 3 to 9 months in the OE of WT mice (Fig. S3). Since tau aggregation does not naturally occur in WT mice, overexpression of human tau with aggregation-prone mutations is required to induce tau pathology in animal models. Together, these observations indicate (i) that tau is endogenously expressed in the mouse OE, (ii) that the specific AT8-positive signal observed in the OE of PS19 mice is not an artifact simply related to ectopic expression of the human tau transgene in in OE of PS19 mice.

Colocalization analyses were carried out via immunofluorescence to confirm the cellular localization of the pTau signal in the OE of PS19 mice. The following antibodies were used: total tau antibody (TTau), olfactory marker protein (OMP) antibody (specific to mature OSNs and their neurites), and AT8 antibody (Fig. 2a–c). Colocalization was observed between the OMP signal and both TTau and AT8 signals in the cellular bodies and neurites of the OSNs, as well as in the ABs of the PS19 mice at 3 months (Fig. 2d, e). OMP and AT8 were also specifically colocalized in the superficial layer where dendritic knobs of OSNs are located (Fig. 2e). This result confirmed that pTau is found not only in mature OSNs of the OE but also in ABs within the lamina propria (Fig. 2e).

Fig. 2

pTau expression in the OSNs and ABs of PS19 mice. Paraffin sections immunostained for total tau (TTau, red) (a, d), olfactory marker protein (OMP, green) (b, d, e), and pTau (AT8, purple) (c, e) show colocalization between TTau-OMP (yellow) (d) and AT8-OMP (white) (e) indicating tau hyperphosphorylation in the OSNs (white arrows) and in the ABs (blue arrows) of PS19 mice at 3 months. High magnifications are shown in the insets (d’, e’). OSNs: olfactory sensory neurons, ABs: axon bundles

Since OSN axons extend through the cribriform plate of the ethmoid bone to reach the OB, where they form its outermost layer (ONL) [14], we next investigated the AT8 staining profile in the OB of PS19 mice. The AT8 signal was specifically present in the ONL of the OB as early as 1.5 months in PS19 mice (Fig. 3a–d). In addition, AT8 immunoreactivity was also detected in the MCL. In some 9-month-old mice, a stronger pTau signal appeared in the central region (e.g., GCL) of the OB (Fig. 3d). Similarly to the OE, no notable bilateral differences in tau pathology were observed in this region. Here also, no AT8 signal was observed in the OB of WT mice (Fig. 3f).

Fig. 3

pTau expression in the OB of PS19 mice. The OB from 1.5-month (a, a’), 3-month (b, b’), 6-month (c, c’, e), 9-month-old PS19 mice (d, d’), and 6-month-old WT mice (f) were immunohistochemically stained with AT8 (a–d, a’–d’). pTau recognized by the AT8 antibody is found in the external layer of the OB (black arrows) and in the MCL (white arrows) from 1.5 months (a–d, a’–d’). The pTau signal seems stable over time (a’–d’). No signal is detected in WT mice (f). All the mice used for immunostainings were males. The immunostainings shown are representative of n = 4 mice per time point. High magnification of the MCL (white arrows) is shown in the inset (a’’). ONL: olfactory nerve layer, MCL: mitral cells layer

Accumulation of hyperphosphorylated tau in the OB of PS19 mice was confirmed by Western blotting of soluble protein extracts using the AT8 antibody (Fig. S4). The experiments were repeated and quantified using the PHF-1 antibody (Fig. S5). A band around 60 kDa was observed exclusively in PS19 mice, both in males and females, as early as 3 months of age. No positive signal for PHF-1 or AT8 was observed in WT mice at any age (Fig. S5a–c). Western blot quantification revealed no significant differences in pTau expression between males and females in the OB at 3, 6, and 9 months (Fig. S5a–c). Finally, the PHF-1 signal was quite stable over time in PS19 males, suggesting that pTau, although already present at 3 months, did not further accumulate in the OB (Fig. S5d). The presence of human tau in the OBs of PS19 mice was confirmed by immunostaining with the Tau-13 antibody, as well as endogenous tau expression in the OBs of WT mice using a total tau antibody (Fig. S2). Endogenous tau expression in the OBs of WT mice was also confirmed by RT-qPCR (Fig. S6).

Immunofluorescence analyses of the OB in 3-month-old PS19 mice (Fig. 4a–c) revealed colocalization between TTau and OMP, indicating tau protein expression in both the ONL and the GL (Fig. 4d). Additionally, strong colocalization between OMP and AT8 was observed in the ONL, providing evidence of pTau accumulation in the outermost layer of the OB in PS19 mice (Fig. 4e). Some AT8 signal, which did not colocalize with OMP, was also observed in the inner layers of the OB (Fig. 4c).

Fig. 4

pTau expression in the ONL of PS19 mice. Paraffin sections immunostained for total tau (a, d) (TTau, red), OMP (b, d, e) (green), and AT8 (c, e) (purple) show colocalization between TTau-OMP (yellow) (d) and AT8-OMP (white) (e) indicating tau protein expression both in ONL and GL as well as pTau expression in the ONL and inner layers of the OB from 3-month-old PS19 mice. ONL: olfactory nerve layer, GL: glomerular layer

Next, we explored tau pathology in CNS regions of PS19 mice that are connected to the OB and involved in the processing of olfactory information along with olfactory memory: the PC, the EC, and the hippocampus. All these CNS regions are connected by neuroanatomical pathways [43]. Efferent projections from the OB reach both the PC and EC, and the perforant path wires the EC to hippocampal subregions [3]. In the PC, faint pTau immunoreactivity (AT8), indicative of the first step of abnormal tau modification [4], appeared in the neuronal cell bodies at 3 months (Fig. 5b, b’). By 6 months, the first perikaryal NFT-like tau accumulations were observed, with an accumulation of pTau in both the cell bodies and neurites of neurons within the PC pyramidal layer (Fig. 5c, c’). These tau accumulations progressively increased up to 9 months and eventually spread throughout the entire PC (Fig. 5d, d’). 9-month-old WT mice showed no AT8-positive accumulations in the PC (Fig. 5e). In the EC, AT8 immunoreactivity appeared in the neuronal cell bodies as early as 1.5 months and increased over time (Fig. 5f–i). Perikaryal NFT-like tau accumulations were detectable at 6 months, with signal present in both the cell bodies and neurites in the lateral part of the EC (Fig. 5h, h’). Tau lesions gradually accumulated up to 9 months and spread to the whole EC (Fig. 5i, i’). The accumulation pattern of tau pathology was similar to that observed in the PC, although with an earlier onset. In WT mice, AT8 signal was not detected until 9 months of age (Fig. 5j).

Fig. 5

pTau expression in the PC and EC of PS19 mice. The PC (Bregma − 1.94 mm) from 1.5-month (a, a’), 3-month (b, b’), 6-month (c, c’), 9-month-old PS19 mice (d, d’), and 9-month-old WT mice (e, e’) were immunohistochemically stained with AT8 (a–e). pTau recognized by the AT8 antibody is found in the PC as soon as 3 months (b, b’) with an accumulation up to 9 months (b–d, b’–d’). The first NFT-like tau accumulations appear at 6 months (c, c’). There is no AT8 signal in the PC of WT mice (e, e’). The EC (Bregma − 3.08 mm) from 1.5-month (f, f’), 3-month (g, g’), 6-month (h, h’), 9-month-old PS19 mice (i, i’), and 9-month-old WT mice (j, j’) were immunohistochemically stained with AT8 (f–j). pTau recognized by the AT8 antibody is found in the EC as soon as 1.5 months (f, f’) with an accumulation up to 9 months (f–i, f’–i’). PS19 mice develop NFT-like tau accumulations in this region at 6 months (h, h’). No AT8 signal is detected in the EC of WT mice (j, j’). All the mice used were males. The immunostainings shown are representative of n = 4 mice per time point. Insets were taken from the PC area (a’–e’) and in the lateral EC (f’–j’)

In the hippocampus, AT8 immunoreactivity was faintly detected in the CA3 subregion at 3 months (Fig. 6b). By 6 months, pTau (AT8) was clearly present in the CA3 region and in the dentate gyrus (DG) (Fig. 6c), with an increased signal at 9 months (Fig. 6d). In the CA3 region, pTau was found in the CA3 mossy fibers and in the stratum pyramidale (Sp). In the DG, pTau was located primarily in the granular cell layer (Fig. 6e). No AT8 signal was detected in the hippocampus of WT mice (Fig. 6f).

Fig. 6

pTau expression in the hippocampus of PS19 mice. The hippocampus (Bregma − 1.82 to − 2.18 mm) from 1.5-month (a), 3-month (b), 6-month (c), 9-month-old PS19 mice (d, e), and 9-month-old WT mice (f) were immunohistochemically stained with AT8 (a–f). pTau recognized by the AT8 antibody is found in specific hippocampal subregions like the CA3 (e’’) and DG (e’) as soon as 6 months with an increased signal at 9 months (c–e). No AT8 signal is found in the hippocampus of WT mice (f). All the mice used for immunostainings were males. The immunostainings shown are representative of n = 4 mice per time point. DG: Dentate gyrus, CA: Cornu Ammonis, Sp: Stratum pyramidale

pTau expression in soluble hippocampal extracts was confirmed by Western blotting with the AT8 antibody (Fig. S7). Equivalent results were obtained and quantified using the PHF-1 antibody. pTau was detected as a band around 60 kDa and was present only in PS19 mice, as soon as 3 months of age in both males and females. No positive signal for PHF-1 or AT8 was observed in the WT mice regardless of the aging stage (Fig. S8a–c). At the same age, the average pTau levels were significantly greater in males than in females. Although the difference was already significant at 3 months, it was more pronounced at 9 months (Fig. S8a-c). Western blot analysis finally revealed a significant increase in the pTau signal between 3 and 9 months in PS19 males, consistent with the staining results, suggesting that pTau accumulation in the hippocampus occurs in a time-dependent manner (Fig. S8d).

To distinguish AT8-positive NFT-like tau accumulations from mature filamentous tau aggregates referred to as mature NFTs, Gallyas silver staining was performed in all previously studied regions and time points. No NFTs were observed in the OE (Fig. 7a–d) and OB (Fig. 7f–i) of PS19 mice. In the PC and EC, Gallyas-positive NFTs were detected only at 9 months of age in PS19 mice (Fig. 7n, s), and were very sparse compared to the AT8 signal observed previously (Fig. 5). In the hippocampus of PS19 mice, no NFTs were observed at any age (Fig. 7u–x). As a control, no NFTs were detected by silver staining in any region of interest in WT mice (Fig. 7e, j, o, t, y).

Fig. 7

Gallyas staining in the OE, OB, PC, EC, and hippocampus of PS19 mice. The OE (a–e), OB (f–j), PC (k–o), EC (p–t), and hippocampus (u–y) of PS19 mice at 1.5, 3, 6, and 9 months, as well as 9-month-old WT mice were stained with Gallyas silver staining to highlight NFT inclusions. A few Gallyas-positive inclusions (black arrows) are only observed in the PC (n) and EC (s) from 9 months PS19 mice. No NFTs are detected in the OE, OB, or hippocampus at any time point, nor in the PC or EC before 9 months. No signal is detected in WT mice (e, j, o, t, y). High magnifications of the Gallyas-positive aggregates (black arrows) are shown in the inset (n’) and (s’). All the mice used for Gallyas silver staining were males. The Gallyas stainings shown are representative of n = 3 mice per time point

These results clarify some aspects of the temporal and spatial staging of tau pathology in PS19 mice by distinguishing pTau (i.e., recognized by AT8 or PHF-1 antibodies), perikaryal NFT-like tau accumulations or pretangles (AT8-positive) and mature NFTs evidenced by Gallyas staining. Strikingly, pathological pTau accumulated very early (from 1.5 months of age) in the OE without the formation of typical NFTs, as shown by the Gallyas staining. A similar profile (pTau staining without NFTs) was observed in the OBs. pTau immunoreactivity was found in the EC and OE from 1.5 months on and in the hippocampus from 6 months. NFT-like tau accumulations appeared later, from 6 months in the PC and EC, and at 9 months in the hippocampus (Sp). While only a few NFTs (Gallyas-positive) were present in the PC and EC at 9 months. Taken together, our extensive analysis of the accumulation of pTau in connected olfactory regions gives a solid support to the hypothesis that tau pathology progresses along the olfactory system to reach regions of the temporal lobe that are particularly vulnerable in AD (e.g., EC and hippocampus).

Olfactory function in PS19 mice

The impact of tau lesions in the olfactory system on olfactory function was assessed in PS19 mice using two specific olfactory tests: the olfactory discrimination test and the food-seeking test. The first test evaluated odor discrimination, whereas the food-seeking test measured odor threshold in mice. As a proof of concept, control and ZnSO4-treated mice were first evaluated, since intranasal irrigation with ZnSO4 induces OE stripping and transient anosmia [46].

In the olfactory discrimination test, the curve for the number of sniffing events in non-treated (control mice) showed two peaks at 6 and 12 min (Fig. S9a). The first peak corresponded to the presentation of the geraniol odor, which induced a greater number of head movements toward the cotton swab in the control mice. Following this peak, the mice exhibited fewer head movements as they became accustomed to the odor. A second peak was observed after 12 min, when the second odor, lime, was presented. Both peaks were absent in the mice that received intranasal ZnSO4 administration (Fig.S9a). Similarly, during the food-seeking test, none of the ZnSO4-treated mice were able to locate buried food before the end of the test (300 s) (Fig. S9a’).

The same tests were conducted in WT and PS19 mice at 3, 6, and 9 months (Fig. 8a–f) to evaluate the impact of tau pathology in the OE and associated CNS regions on olfactory function. In the olfactory discrimination test, both WT and PS19 mice showed two peaks corresponding to the introduction of each new odor. Each peak was followed by reduced reactions due to habituation to the odor (Fig. 8a–f). All mice appeared to react more strongly, with a greater number of head movements, when presented with the second odor of lime. The superposition of curves indicates that all the mice, independently of their genotype and age, have similar olfactory capacities and that PS19 mice do not perform worse as they age. Statistical analyses revealed a significant impact of time as the mice responded to odor changes. However, no significant effect of genotype on olfactory capacities was observed at 3, 6, and 9 months in both males and females (Fig. 8a–f). The results from the olfactory discrimination test suggested that the odor discrimination abilities of PS19 mice are not impaired by the presence of tau pathology in the OE. In the food-seeking test, there was no significant difference between WT and PS19 mice in the latency to find the pellet. This observation was consistent at 3, 6, and 9 months in both males and females. Compared to WT mice, PS19 mice did not exhibit an impaired ability to find the food (Fig. 8a’–f’). Together, these data indicate that despite progression of tau pathology along the olfactory regions with age, PS19 mice did not show a significant impairment in odor discrimination or detection threshold (sensitivity). Of note, another parameter (odor identification) difficult to address in mice was not evaluated here.

Fig. 8

Olfactory function assessment in PS19 mice. The olfactory discrimination (a–f) and food-seeking tests (a’–f’) show no significant effect of genotype (PS19) on olfaction at either 3, 6, or 9 months, both in males and females (a–f, a’–f’). For PS19 males; n = 9 (3 months), n = 9 (6 months), and n = 7 (9 months). For WT males; n = 9 (3 months), n = 9 (6 months), and n = 9 (9 months). For PS19 females; n = 8 (3 months), n = 8 (6 months), and n = 5 (9 months). For WT females; n = 8 (3 months), n = 8 (6 months), and n = 8 (9 months). Data were analyzed using two-way ANOVA with Šídák’s multiple comparisons test (olfactory discrimination test) or using independent t test (food-seeking test)

Modulation of tau pathology in the CNS via intranasal interventions

Given the staged progression of tauopathy in olfactory regions and its early manifestation in the OE, we administered localized treatments to the OE to either mitigate tau pathology or exacerbate its spread to the CNS.

First, we examined the impact of targeted removal of the source of pathological tau present in the OE on tau pathology progression in the CNS. Our approach was designed to specifically remove the pTau accumulation in the OE using ZnSO₄ nasal irrigation, known to strip the OE and ONL (OB), and cause transient anosmia [32, 45], which can later regenerate (OE) and recover (anosmia). The impact on tauopathy progression was assessed in regions with varying degrees of connectivity to the OB, such as the PC, EC, and amygdala. PS19 mice were intranasally administered either ZnSO₄ or phosphate-buffered saline (PBS, control). To ensure that the accumulation of tau in the OE was prevented for a sufficiently long period, nasal irrigation with ZnSO₄ was performed repeatedly at 1.5, 2.5, and 4 months prior to analysis of tauopathy at 6 months.

In vehicle-treated PS19 mice, the structure of the OE remained intact, with strong AT8-positive tau immunoreactivity in OSNs and ABs. However, following intranasal ZnSO₄ administration, the OE was completely stripped, and the AT8 signal was almost absent at 6 months (Fig. 9a). In the OB of PBS-treated mice, all six layers were present and structurally preserved, with AT8-positive signal characteristic of PS19 mice. After ZnSO₄ treatment, the AT8 signal in the ONL was no longer detectable (Fig. 9a). The efficient stripping of the OE and the ONL after ZnSO₄ treatment was further confirmed by Masson’s trichrome blue staining (Fig. S10a–d). Tau pathology was next examined in regions receiving direct inputs from the OB, such as the PC, EC, and amygdala. At 6 months, vehicle-treated PS19 mice exhibited NFT-like tau accumulations in these regions. However, following ZnSO₄ treatment of PS19 mice, no NFT-like tau accumulations were detectable in the PC, and only a few remained in the EC and amygdala at 6 months compared to control PS19 mice (Fig. 9a). Quantification of the number of NFT-like tau accumulations per mm2 in the PC, EC, and amygdala (Fig. S11) further confirmed a massive reduction of pTau signal associated with tau pathology in the PC, EC, and amygdala following intranasal ZnSO4 administration (Fig. 9b). No overt morphological alterations were found in these regions after staining with Masson’s trichrome blue (Fig. S10e–j), indicating that ZnSO₄ treatment of the OE did not result in loss of density of the connected olfactory regions. In other terms, removing the source of pTau present in the OE sharply reduces tau pathology in the connected regions of the temporal lobe and limbic system.

Fig. 9

Modulation of tau pathology in the CNS via intranasal interventions. Sections of the OE, OB, PC, EC, and amygdala from 6-month-old PS19 males were fixed and stained with the AT8 antibody. After intranasal ZnSO4 irrigation, the OE and ONL from the OB are stripped, no more pTau signal is found in these regions (a). Quantification of the number of NFT-like tau accumulations/mm2 was performed in the PC, EC, and amygdala. In the PC, there are no NFT-like tau accumulations compared to the PBS-treated mice. In the EC and amygdala, there is a significant decrease in NFT-like accumulations compared to the control condition. **P < 0.01 (Mann–Whitney test, n = 6 mice/group) (b). OE sections immunostained for GFP (green), AT8 (purple), and DAPI showed efficient transduction following intranasal AAVs administration (c). In the PC, AT8 immunoreactivity appeared increased in AAV-treated mice compared to PBS controls (d). Quantification of the AT8-positive area in the PC showed a trend toward increased pTau signal following nasal delivery of hTau-P301S AAVs (Mann–Whitney, n = 5 mice per group) (e). ONL: olfactory nerve layer, AG: amygdala

To further validate the hypothesis that tau pathology spreads from olfactory regions, we aimed to determine whether increasing pathology at the OE level could accelerate the progression of tau pathology in the CNS. To this end, we performed intranasal administrations of AAVs encoding the P301S-mutated human tau protein (referred to as hTau-P301S AAV). The efficacy of AAV infection was first validated by the presence of GFP, which is co-expressed from the same vector. One week after the infection, a GFP-positive (green) signal was detected exclusively in the OE of AAV-treated mice, but not in PBS-treated controls (Fig. 9c), confirming successful transduction.

Subsequently, we assessed tau pathology using the AT8 antibody. We focused our analysis on the PC, a region directly connected to the OB and one of the first affected areas in tauopathies. At the end of the treatment, when the mice were 4 months old, we observed that the intranasal delivery of hTau-P301S AAV increased AT8 immunoreactivity in the PC of AAV-treated mice compared to controls (Fig. 9d, e). Although the difference did not reach statistical significance—likely due to variability in transduction efficiency—the pattern was consistent: all control mice displayed similar baseline levels of pathology, whereas all AAV-treated mice exhibited more advanced tau pathology, with severity likely varying in relation to the efficacy of AAV delivery (Fig. 9d, e).

Overall, our data indicate that early accumulation of pTau in the OE can be instrumental to the progression of tau pathology that eventually reaches the EC and the hippocampal formation.

Pathological pTau protein in the human olfactory system

Finally, we investigated whether a similar pattern of tau pathology was observed in OE and OB collected from patients at early Braak stages I and II (patients 1–4, Table S2). In the OE collected from patient 4 (Braak stage II), pTau immunoreactivity (PHF-1 Ser396, Ser404) was detected and colocalizing with the OMP signal, indicating the presence of pTau in OSNs (Fig. 10b). PHF-1 signal was also detected in the ABs present within the OE structure and revealed by the OMP antibody (Fig. 10c–e). Still, PHF-1 and OMP signals are not fully overlapping. PHF-1 signal is mostly around the nuclei, while the OMP stains all the axonal fibers present in the ABs (Fig. 10f’). Interestingly, while colocalization was observed between OMP and PHF-1 within the ABs (Fig. 10f), AT8 signal was detected neither in the ABs (Fig. 10g) nor in the OSNs.

Fig. 10

pTau expression in human OE. The OE was obtained from a Braak stage II patient (patient 4). The sections were immunohistochemically stained with PHF-1 or with OMP (a–e). PHF-1 signal is observed in the OMP-positive OSNs (b) and ABs (c–e) of the OE. Paraffin sections immunostained for OMP (green) and PHF-1 (purple) show colocalization between OMP-PHF-1 (white) in the AB (f, f’). PHF-1 signal is also observed inside the tissue (f). Paraffin sections immunostained for OMP (green) and AT8 (purple) show no colocalization between OMP-AT8 (g, g’). No signal is detected in the negative controls without primary antibodies. Scale bar: 100 μm. OSNs: olfactory sensory neurons, ABs: axon bundles

In the OB of Braak stage I/II patients, PHF-1 and OMP colocalization was consistently detected in the ONL, and to a lesser extent, in the GL (Fig. 11a–d, d’). Some PHF-1 pTau signal was additionally found in the inner layers of the OB (Fig. 11a–d, d”). No colocalization was observed between OMP and AT8 in the ONL (Fig. 11e–h, h’), but similarly to PHF-1, neuropil threads highlighted by the AT8 antibody were observed in the inner layers of the OB from Braak stage I and II patients (Fig. 11e–h, h”). All OB included in the study were Gallyas-negative, highlighting the absence of mature NFTs at this stage (Braak stage I/II) of the pathology (data not shown). These results underscore the early appearance of pathological tau markers in the sensory regions of the human olfactory system in agreement with observations in the PS19 mouse model.

Fig. 11

pTau expression in human OB. The OB were obtained either from patients 1 and 2 (Braak stage I) (a, b, e, f) or from patients 3 and 4 (Braak stage II) (c, d, g, h). Paraffin sections immunostained for OMP (green) and PHF-1 (purple) show colocalization between OMP-PHF-1 (white) in the ONL of all patients (a–d, d’). PHF-1 signal is also found in the inner layers of the OB (a–d, d’’). Paraffin sections immunostained for OMP (green) and AT8 (purple) show no colocalization between OMP-AT8 (e–h, h’). AT8 signal is only found in the inner layers of the OB, similarly to PHF-1 (e–h, h’’). No signal is detected in the negative controls without primary antibodies. Scale bar: 100 μm. ONL: Olfactory nerve layer, GL: Glomerular layer