Tau pathology in the dorsal raphe may be a prodromal indicator of Alzheimer's disease

Abstract

Protein aggregation in brainstem nuclei is thought to occur in the early stages of Alzheimer's disease (AD), but its specific role in driving prodromal symptoms and disease progression is largely unknown. The dorsal raphe nucleus (DRN) contains a large population of serotonin (5-hydroxytryptamine; 5-HT) neurons that regulate mood, reward-related behavior, and sleep, which are all disrupted in AD. We report here that tau pathology is present in the DRN of individuals 25-80 years old without a known history of dementia, and its prevalence was comparable to the locus coeruleus (LC). By comparison, fewer cases were positive for other pathological proteins including α-synuclein, β-amyloid, and TDP-43. To evaluate how early tau pathology impacts behavior, we overexpressed human P301L-tau in the DRN of mice and observed depressive-like behaviors and hyperactivity without deficits in spatial memory. Tau pathology was predominantly found in neurons relative to glia and colocalized with a significant proportion of Tph2-expressing neurons in the DRN. 5-HT neurons were also hyperexcitable in P301L-tau^DRN mice, and there was an increase in the amplitude of excitatory post-synaptic currents (EPSCs). Moreover, astrocytic density was elevated in the DRN and accompanied by an increase in IL-1α and Frk expression, which suggests increased inflammatory signaling. Additionally, tau pathology was detected in axonal processes in the thalamus, hypothalamus, amygdala, and caudate putamen. A significant proportion of this tau pathology colocalized with the serotonin reuptake transporter (SERT), suggesting that tau may spread in an anterograde manner to regions outside the DRN. Together these results indicate that tau pathology accumulates in the DRN in a subset of individuals over 50 years and may lead to behavioral dysregulation, 5-HT neuronal dysfunction, and activation of local astrocytes which may be prodromal indicators of AD.

Figure 1:. Neurodegenerative markers in the DRN and LC of cognitively normal individuals and those with Alzheimer’s Disease.

Human dorsal raphe nucleus (DRN) sections were taken from post-mortem cases of individuals cognitively normal (Control) and diagnosed with AD at the time of death (AD). a-j DAB staining in control (top row) and AD cases (bottom row) for the following: AT8 (ptauSer202/Thr205; a-b); serotonergic cell marker Tryptophan hydroxylase 2 (Tph2, c-d), delineating the anatomical boundaries of the DRN; α-syn (pS129-α-synuclein, e-f); TDP-43 (pS409/410-TDP-43, g-h); and β-amyloid (i-j).Black arrowheads in f and h indicate staining magnified in inset. Counterstain in a-j is hematoxylin. aq=cerebral aqueduct. mlf=medial longitudinal fasciculus. scale bars = 200 μm for a-d and 100 μm for e-j. For the insets of f,h,j scale bar = 20um. k-l Immunofluorescent staining of AT8 (red) and Tph2 (cyan) in tau+ control (l) and AD (m) cases (40X magnification). White arrowheads point to colocalization of AT8 and Tph2 fluorescence. Asterisks denote larger regions of colocalization, such as cell bodies. Scale bar in k-l = 50 μm. m Bar graph of number of cases that were positive or negative for ptau or α-syn in DRN. n Mean AT8 % immunoreactive area values in control and AD cases, separated by presence or absence of DR tau staining. In the single tau- case among the AD group, OD=0 and does not appear on the axes. One control case was in poor condition and was not quantified. o Mean AT8 %IR values in control and AD cases, collapsed across DRN tau- and DRN tau+ cases. In a subset of available cases, the locus coeruleus (LC) was also evaluated for pathology. p-s DAB immunohistochemistry for ptau (AT8; p,q) and α-synuclein pathology (α-syn; r,s) in one control and AD patient in the LC. Stars indicate DAB+ neuromelanin staining, characteristic of noradrenergic cells of the LC. Black arrowheads denote DAB+ pathology for AT8 or α-syn. Counter stain is hematoxylin. Large images were taken at 10X and scale bars are 100μm. Inset scalebars are 20μm and images were taken at 40X. t-v Immunofluorescent stain of AT8, α-syn, and TH in human LC. t Control case positive for AT8 in the LC. u AD case positive for AT8 in the LC. v AD case positive for AT8 and α-syn in the LC. White arrowheads point to ptau and TH double-labeling. White arrows indicate α-syn and TH double-labeling. Asterisks indicate triple labeling of ptau, α-syn, and TH. Images taken at 40X, 1.7 zoom. Scalebar=50 μm. red=AT8, magenta=α-syn, cyan=TH, blue=DAPI. w Bar graph showing the number of control and AD cases positive for AT8 or α-syn in the LC. TDP-43 and β-amyloid images are not pictured, as all cases were negative for staining in the LC. Data are represented as mean ± SEM. *p<0.05.

Figure 2:. DRN tau pathology promotes depressive-like behaviors in male mice.

The P301L-tau mutation or green fluorescent protein (GFP) was delivered via adeno-associated virus into the DRN of C57BL/6J mice (n=10 per group, 20 total). a Schematic of viral infusion site in the DRN. b-c Representative DAB images of viral P301L-tau placement staining in GFP (b) and P301L-tau animals (c). d Graph of mean time GFP (gray) and P301L-tau mice (magenta) spent interacting with a stranger mouse and object in the social interaction assay. e X-Y plot of percent sucrose preference across all days of the sucrose preference assay. f Bar graph of percent sucrose preference on test day for both groups (day 5). g Bar graph depicting the percent of time mice spent immobile in the forced swim assay. h-i Time spent in the open arms (h) and distance moved (i) in the elevated plus maze (EPM). j-k Time mice spent in the center of the arena (j) and distance moved (k) in the open field test (OFT). l-m Latency of first visit (l) and number of revisits (m) to target and non-target holes in the Barnes Maze task. n Percent time mice spent immobile in contextual fear conditioning task. Data are represented as mean ± SEM. Scale bar= 500 μm in (b) and 200 μm in (c). *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Figure 3:. Tau pathology colocalizes with serotonergic DRN neurons and results in astrocytic activation and neuroinflammation in the DRN of P301L-tau^DRN mice.

a,b Representative images of Tph2 (cyan) and AT8 (red) immunofluorescent staining in the DRN of a mouse that received AAV-GFP at 15X and 40X (inset), respectively. c,d Representative images of a mouse that received AAV-P301L-tau, also at 15X and 40X (inset). d White arrowheads point to instances of Tph2 and AT8 colocalization. e Bar graph of AT8 optical density (OD) of GFP (gray) and P301L-tau (magenta) animals across subregions of the DRN and pooled across subregions (overall). f Bar graph of Tph2+ cell density in ROIs that contained AT8+ staining. g Graph of Tph2 OD in ROIs that were AT8+. We observed no differences between GFP and P301L-tau mice. h Simple linear regression between Tph2+ cell density and AT8 OD in P301L-tau animals. i Bar graph of the percentage of Tph2+ cells that were also AT8+ in ROIs for P301L-tau animals. j-m Immunofluorescent staining for astrocytic marker GFAP (yellow) and microglial marker Iba1 (red) in GFP (j) and P301L-tau animals (l) at 20X magnification. k,m GFAP and Iba1 IF in GFP (j) and P301L-tau mice (l) at 40X magnification (inset). n Bar graphs displaying mean OD of GFAP+ and Iba1+ staining in ROIs expressing AT8. o Bar graph displaying GFAP+ and Iba1+ cell density in ROIs. Data are represented as mean ± SEM. g Heatmap representation of gene expression analysis for cytokines, Tph2, and Frk in the DRN of GFP and P301L-tau injected mice (SE of difference: Il1a=0.16; Il12b=0.39; Frk=0.21). Scale represents fold change in mRNA levels from <0.5 (white) to >1.5 (blue). Scalebar in a and c= 200 μm; b, d, k, and m = 20 μm. j and l =100 μm. aq=cerebral aqueduct. *p<0.05, **p<0.01

Figure 4:. Tau pathology preferentially colocalizes with neurons instead of glia in the DRN.

Representative images from C57BL/6J mice injected with AAV-P301L-tau into the DRN after 4 weeks (a-d) and 8 weeks (g-j) of viral transduction. a,g Images showing AT8 (magenta) and GFAP (cyan) immunflourescence staining. b,h Images showing AT8 and Iba1 (yellow) staining. c,i Images of AT8 and NeuN (green). d,j Combined views of the AT8-GFAP-Iba1-NeuN IF depicted in a-c and g-i. White arrows point to areas of colocalization. Images taken at 40X. Scale bar = 50 μm. e,k Pie chart depicting the mean percent of AT8+ staining that was colocalized with GFAP (cyan), Iba1 (yellow), or NeuN (green). Gray indicates AT8-only, presumably in processes not positive for one of the 3 cell markers. f,l Violin plots showing median % positive staining (black line), quartiles (dashed lines), and distribution of data points used in the % calculations.

Figure 5:. Neuronal excitability and glutamatergic transmission is elevated in P301L-tau^DRN mice.

a Stereotaxic injection of AAV encoding P301L-tau and GFP into the DRN. b Representative confocal image of recorded neuron containing biocytin. Left: Tph2 immunoreactivity (magenta) with recorded neuron indicated with a white arrow. Center: AT8 immunoreactivity (cyan) with recorded neuron indicated with a white arrow. Right: Merged image containing Tph2, AT8, and biocytin (yellow). c Top: Representative traces of action potential firing during a current ramp protocol used to compute rheobase (x-axis = 200 ms, y-axis = 50 mV). Bottom: Representative traces showing action potential firing during a 200 pA current step (x-axis = 50 ms, y-axis = 50 mV). d-f Histogram of mean action potential thresholds (rheobase), resting membrane potential (RMP), and input resistance for control and P301L-tau^DRN mice. g-h Histogram of action potential frequency as a function of current (0–200 pA) starting at RMP and at −70 mV. i Top: Representative traces of eEPSCs at a range of stimulation intensities (10 – 320 μA) (x-axis = 10 ms, y-axis = 500 pA). Bottom: Representative traces of eIPSCs (x-axis = 10 ms, y-axis = 200 pA). j-k Histogram of mean eEPSC amplitude and amplitude as a function of stimulation intensity. l Representative traces of PPR for eEPSCs (x-axis = 10 ms, y-axis = 500 pA) and eIPSCs (x-axis = 10 ms, y-axis = 200 pA). m Histogram of mean PPR for eEPSCs and eIPSCs. n Representative traces of sEPSCs (x-axis = 100 ms, y-axis = 50 pA) and sIPSCs (x-axis = 200 ms, y-axis = 50 pA) in control and P301L-tau^DRN mice. o-p Histograms of mean frequency and amplitude of sEPSCs and sIPSCs. q-r Cumulative frequency distributions of sEPSC and inter-event intervals (IEIs) in GFP and P301L-tau^DRN mice. Data are represented as mean ± SEM unless otherwise indicated. Scalebars = 100 μm. *p<0.05, ***p<0.001, ****p<0.0001.

Figure 6:. Tau pathology propagates from the DRN to other brain regions.

In a subset of male mice (n=4), we evaluated brain regions downstream of the DRN known to be involved in affective changes and memory for AT8 staining. a-f Representative images of staining in respective regions at 60X objective magnification. Insets in b, d-f are enlarged and cropped from larger images to highlight positive staining. Black arrowheads point to AT8+ staining. g Table summarizing semi-quantitation of AT8 staining in downstream regions. NA=no sections available of the region from that brain, (−) indicates no AT8 staining. (+) indicates very little pathology localized to single fields of view within a region. (++) indicates multiple areas in multiple fields of view with AT8 staining in a region. Scale bars=20 μm. EC=entorhinal cortex, Thal = thalamus, HP= hippocampus, HT=hypothalamus, AMY=amygdala, CPu=caudate/putamen