The onset of early Alzheimer's disease and tau pathology in the central region of the brain

Guide:

Early onset of symptoms and disease progression in Alzheimer's disease (AD) are closely related to the spread of tau protein, but the underlying mechanism is unknown
.
Recently, a paper entitled "Earlier Alzheimer's disease onset is associated with tau pathology in brain hub regions and facilitated tau spreading" was published in Nature Communications.
The authors found by resting fMRI combined with longitudinal tau-PET in control and biomarker-confirmed AD patients: the overall frontopatoparietal center (an area critical for maintaining AD cognition) in younger symptomatic AD patients showed more significant tau-PET
.
Tau-PET in this region is associated with subsequent rapid accumulation of tau, suggesting that tau in the overall junction region promotes junction-mediated tau propagation
.
In addition, more @strong tau-PET in the central region mediated an association between younger symptomatic AD patients and faster tau accumulation, suggesting faster cognitive decline
.
These findings suggest that younger onset of AD symptoms is associated with stronger tau pathology in brain centers and accelerates the spread of tau in connected brain regions and cognitive decline
.

Sporadic Alzheimer's disease (AD) is highly heterogeneous, with different clinical manifestations, age of symptom onset, and cognitive and neuropathological trajectories
.
In sporadic AD, earlier onset of symptoms is associated with accelerated tau accumulation, neurodegeneration, faster cognitive decline, and higher mortality, which together suggest that younger onset of symptoms may be driven
by more severe forms of AD 。 However, why do patients with sporadic AD with earlier onset of symptoms show faster pathological and clinical progression than patients with later onset of symptoms? From a histopathological perspective, early-onset and late-onset sporadic AD have the same molecular pathology, including amyloid-β (Aβ) plaques and neurofibrillary tangles formed by overphosphorylated tau pathology, suggesting that earlier symptom onset is unlikely to be specifically driven
by the unique molecular features of AD pathology.
In contrast, previous autopsy and tau-PET imaging studies have reported that early onset of AD symptoms correlates with different spatial distribution patterns of tau pathology, a key driver of
AD neurodegeneration and cognitive decline.
Specifically, autopsy evaluation of AD patients found that more pronounced neocortical and hippocampus-preserving patterns of neurofiber tau pathology were associated with younger age at symptom onset and faster cognitive decline before death, while tau pathology in spatially restricted limbic-dominated neurofibril patterns was associated
with older age at symptom onset and slower cognitive decline before death.
Similarly, tau-PET studies in AD patients found that younger AD patients and early-onset AD (before age 65) had stronger tau-PET uptake in the frontoparietal joint cortical region and relatively less
in the medial temporal lobe region.
Together, these findings suggest that the onset of early symptoms of AD is associated
with a pattern of pathological deposition of tau from the medial temporal lobe region of the uneven cortex to the neocortical frontoparietal joint cortex.
As shown by functional MRI studies, the frontoparietal region has a key hub for overall connectivity to the rest of the brain, is the center of cognitive function, and they facilitate the integration
of information across different brain networks during cognitive needs.

A series of previous functional MRI studies have demonstrated the importance of
the functional integrity of the frontopatoparietal control network for maintaining cognitive function in patients with aging, AD and other neurodegenerative diseases.
Since tau pathology has been shown to impair neuronal function and drive neurodegeneration, a stronger tau pathological load in the frontopatrital central region, which is highly associated with cognition, may drive early symptom manifestations
of AD.
In addition, TAU pathological deposition in the pivotal region of the overall connection may further accelerate the progression and spread of TAU pathology itself, which in turn is a powerful driver of cognitive decline
.
Specifically, preclinical studies have consistently shown that tau propagates
across synapses in interconnected neurons.
Similarly, previous studies in combination with tau-PET and MRI have shown preferential transmission of tau in functional and anatomically connected brain regions, where the pattern of connections in which tau contains a central region determines the mode of transmission of
subsequent tau pathology.
Thus, the early occurrence of tau pathology in the hub of the overall connection of AD may lead to faster and more widespread spread of tau, thereby driving earlier disease manifestations and faster clinical progression
.

To address these issues, the authors combined tau-PET and resting fMRI studies with the primary objectives of the assessment: (1) whether younger age of symptom onset in the overall connected central region is associated with stronger tau pathology in symptomatic AD patients compared with non-central regions; (2) whether the pathology of tau in the central region is related to the subsequent accelerated spread of tau; (3) Whether the association between younger and faster tau accumulation rates in symptomatic AD is mediated
by stronger tau pathology in the hub region of the overall connection.
To do this, the authors used two independent samples covering the entire AD spectrum, including 240 participants from the AD Neuroimaging Initiative (ADNI) and 57 subjects from the BioFINDER study, all with available baseline amyloid-PET and longitudinal Flortaucipirtau-PET
.
To map the topology of the overall connectivity hub across the brain, the authors used high-resolution resting state fMRI data
from 1,000 healthy participants from the Human Connectome Program (HCP).
Based on these resting fMRI data, the authors estimated a whole-brain map of healthy individuals not affected by AD pathology based on the weighting degree of the map theory metric (also known as the global connection
).
By mapping individual tau-PET patterns from AD patients to the topology of the globally connected brain centers, the authors determined the extent to which
individual tau-PET patterns are expressed in the overall connective central region of the frontopareal cortex versus weakly connected non-central regions, such as the temporal margin and visual cortex regions 。 Based on these data, the authors found: (1) tau-PET deposition was more pronounced in younger patients with symptomatic AD in the central region of the overall connection and was associated with early symptom onset; (2) a more pivot-like pattern of tau pathological deposition at baseline was associated with subsequent acceleration of tau accumulation per year; (3) The association between younger and faster tau accumulation rates in symptomatic AD may be mediated by a central-like tau-PET pattern, leading to faster cognitive decline
.

outcome

The authors included two completely separate samples (ADNI and BioFINDER) with baseline amyloid-PET and longitudinal Flortaucipirtau-PET baseline data
.
The authors further included longitudinal cognitive data for memory synthesis as well as age estimates of symptom onset in patients with symptomatic AD, which could be used in a subset
of ADNI 。 Both samples included cognitively normal (CN) control subjects (ADNI/BioFINDER, n=93/16), CN amyloid-positive individuals (i.
e.
, preclinical AD, ADNI/BioFINDER, n=60/16), and symptomatic AD patients (including mild cognitive impairment (MCI) and ADADNI/Bio-FINDER, n=89/25).

Participant demographics for both samples are shown in
Table 1.

Table 1.
Participant demographics for both samples

Resting state fMRI data from 1,000 HCP participants were used to estimate healthy connectivity templates covering the Schaefer200 region of interest (ROI) atlas of the neocortex (Figure 1A).

Among 1,000 HCP participants, the authors create a group average functional connection matrix with a density threshold of 30% to eliminate potentially weak and noisy connections and convert to distance based on connections (i.
e.
, higher functional connections = shorter connection distances, Figure 1B).

Then, for each ROI, the authors determined the connection-based average distance from that ROI to the rest of the ROIs as a measure of centricity (i.
e.
, shorter distance to the rest of the brain = higher centrivity, Figure 1C), which corresponds to the weighting of
chart theory measurements.
Consistent with previous studies, the pivotal region is mainly located in the
frontopatogram symphysis cortex.
To determine the central graph of brain regions that scales between central and non-central, the group mean global connectivity plot (Figure 1C) is rescaled between 1 (i.
e.
, hub) and -1 (i.
e.
, non-hub) (Figure 1D) to assess whether a single tau-PET pattern matches
a central or non-central pattern in a later step 。 Segment the Tau-PETSUVR data (i.
e.
, normalized to the intensity of subcerebellar gray) using the same 200 ROI segmentation (Figure 1A) and apply a pre-established two-component Gaussian mixture model (Figure 1E) to transform tau-PETSUVR (normalized absorption value ratio) to tau-PET positive in order to distinguish
off-target from in-target tau-PET binding 。 In another exploration step, the authors also included tau-PETSUVR from a pre-established hippocampal ROI, which excluded the choroid plexus and thus minimized the effect
on off-target binding of Flortaucipir.
As for the remaining cortex, hippocampal tau-PETSUVR performed two-component Gaussian mixture modeling to further reduce the effects of
off-target binding.
The baseline tau-PET positive surface presentation of the diagnostic group of ADNI and BioFINDER is shown in Figure 1F, and in the amyloid-negative control group, there is no evidence of tau-PET positivity, temporal lobe tau levels are slightly elevated in preclinical AD, and tau-PET positivity gradually increases
in symptomatic AD patients with MCI and AD dementia.

Figure 1.
fMRI and tau-PET treatment

In symptomatic AD, younger age appears to correlate with tau-PET patterns in the center

To assess the extent to which individual tau-PET deposition represents a pivotal or non-pivotal pattern, the authors mapped patient-specific tau-PET positives onto the zoom pivot plot shown in
Figure 1D.
Specifically, the authors multiplied the tau positivity of a specific ROI-specific individual by the pivot plot shown by the scaling scale (Figure 1D) to determine the center-weighted tau positivity value, which was subsequently averaged and divided by the overall tau-PET positive value for all 200 ROIs to adjust for the overall tau deposition
.
The authors adjusted the overall tau level so that the tau center ratio reflected the deposition pattern of tau rather than the overall tau severity
.
The mapping of the tau-PET pattern to the scaled pivot plot for a specific individual is shown in Figure 1 G-M, yielding a numerical index (tau-center ratio) that indicates whether the tau-PET pattern of a particular individual is more representative of the central region (i.
e.
, more positive) or non-central region (i.
e.
, more negative), while adjusting for overall tau-PET positivity
。 No difference in tau center ratio between groups was judged using the ANOVA method (Table 1, ADNI, p=0.
090, BioFINDER, p=0.
566), indicating that tau center ratio did cause spatial patterns of tau deposition, but did not add diagnostic errors
.
The authors then used linear regression, which controlled for sex, education, and diagnosis, to determine whether younger age at symptomatic onset correlated
with a higher tau-centered ratio in patients with symptomatic AD (i.
e.
, MCI and dementia).
The authors found that the younger age of symptomatic AD patients in two samples was associated with a stronger positive tau-centered ratio (ADNI: β=-0.
238, p=0.
024, Figure 2A; BioFINDER: β=-0.
482, p=0.
018, Figure 2 B).

These results were confirmed by accurate testing of 1000 beta values of the empty model tau center ratio generated using the shuffle connection group as a reference (ADNI: p=0.
016; BioFINDER: p=0.
019).

Furthermore, among 38 symptomatic ADNI participants with available resting fMRI data, this association was consistent when the tau-center ratio was determined using subject-level connection data and tau-PET (b=-0.
307, p=0.
037).

Results were also still significant
when controlling for overall Aβ levels or the ApoE subtype.
Conversely, for asymptomatic patients with preclinical AD, no association between age and tau center ratio was detected (ADNI: β=0.
087, p=0.
542, Figure 2A; BioFINDER: β=-0.
307, p=0.
901, Figure 2B), where the abnormal tau-PET signal is minimal
.
When the above analysis was repeated for the actual age associated with cognitive symptom onset (i.
e.
, 77/89 symptomatic ADNI participants) based on the informant assessment (i.
e.
, ADNI participants with symptoms at 77/89), the authors found consistent results indicating that the younger the age of symptom onset, the higher the central proportion of tau (β=-0.
256, p=0.
024, Figure 3A).

。 In addition, the authors tested whether the center of tau pathology (i.
e.
, the 10% of brain regions with the highest baseline tau-PET positive) was more likely to be located in the overall connection region of young symptomatic AD patients mapped to the center (the anatomical location of the center is labeled in the source data file).

。 The authors found that in symptomatic AD patients, younger age did correlate with a stronger centrality (i.
e.
, shorter connection distance to the rest of the brain) in the tau center (ADNI: β=0.
300, p=0<0.
001, Figure 2C; BioFINDER: β=0.
646, p<0.
001, Figure 2D) further supports this view
.
Under the premise of controlling for gender and educational attainment, this pattern of outcomes was not observed in preclinical AD (Figure 2D).

At ADNI, the authors repeated these analyses with age-based onset of cognitive symptoms based on the insider, finding consistent results showing that younger ages of onset were associated with stronger centrality of tau (i.
e.
, shorter connection-based distances to the rest of the brain) (β=0.
352, p=0.
002, Figure 3B).

Together, these results support the hypothesis that younger episodes of AD symptoms are associated with stronger tau pathology in functional center regions that are tightly
connected to other parts of the brain.

The authors further explored whether amyloid-positive ApoE4 noncarriers exhibited a stronger tau-center ratio, as previous studies have shown that in AD patients, ApoE4 carriers are associated with margin-dominated tau pathological patterns, while amyloid-positive ApoE4 noncarriers exhibit more neocortical tau pathological patterns
.
All of the above results were consistent
when the hippocampal tau-PET was included in the analysis.

Figure 2.
Lower age is associated with a higher central rate of symptomatic AD

Figure 3.
Earlier episodes of symptoms are associated with a higher tau-centered ratio

Younger time of symptom onset and higher tau center rates are associated with faster tau accumulation in symptomatic AD

Further, the authors tested whether younger age was associated
with a faster rate of tau accumulation (i.
e.
, the annual rate of change in overall tau positivity) in symptomatic AD patients.
Consistent with previous work, the authors used linear regression to control for sex, education, and diagnosis that younger age correlated with a faster annual rate of change in overall tau-PET positivity rates in symptomatic AD patients (ADNI: β=-0.
284, p=0.
009, Figure 4B; BioFINDER: β=-0.
836, p<0.
001, Figure 4C).

As expected, an association between younger age and faster tau accumulation was not detected in preclinical AD patients (ADNI: β=-0.
119, p=0.
389, Figure 4B; BioFINDER: β=0.
208, p=0.
448, Figure 4C), indicating that faster tau accumulation or generally tau accumulation at younger ages is specific for symptomatic AD subjects
.
Next, the authors tested whether a higher tau-centered ratio (indicating a more hub-like tau-PET deposition pattern) was associated
with faster overall tau accumulation.
This analysis, inspired by the authors' previous findings, showed that tau diffuses from specific central regions to connected brain regions
.
Therefore, tau pathology in the hub region of the overall connection should lead to wider tau spread and faster tau accumulation
.
Conversely, tau in weakly connected non-hub regions should result in slower propagation and tau accumulation (Figure 4A).

In support of this view, the authors found that a higher tau-centered ratio at baseline was associated with faster subsequent tau accumulation in patients with symptomatic AD (ADNI: β=0.
347, p=0.
006, Figure 4D; BioFINDER: β=0.
668, p=0.
002, Figure 4E).

Accurate testing using 1000 beta values from the empty model tau center ratio, using shuffled connection groups as a reference, confirmed these results (ADNI: p=0.
010; BioFINDER: p=0.
018).

The results were consistent when additional age was controlled (ADNI: β=0.
259, p=0.
009; BioFINDER: β=0.
311, p=0.
026), suggesting that the tau center ratio has a unique contribution
to faster tau accumulation 。 In preclinical AD patients, the authors also found a significant association between a higher tau-centered ratio at baseline and a faster rate of subsequent tau accumulation in ADNI (β=0.
347, p=0.
006, Figure 4D), but this finding could not be replicated in BioFINDER (β=0.
199, p=0.
469, Figure 4E).

Together, these results suggest that in symptomatic AD patients, a smaller age at symptom onset and a more central-like tau-PET pattern are associated
with a faster accumulation of tau pathology.
Again, these results were consistent
when the hippocampal tau-PET was included in the analysis.

In order to further confirm that it is indeed the spatial distribution of tau pathology in the center, rather than the overall severity of tau pathology that leads to the rapid accumulation of tau and early disease manifestations, the authors conducted further analysis
.
The authors performed a global tau sliding window analysis from low to high in Aβ+ subjects from ADNI and BioFINDER and assessed the mean annual tau-PET change and age for each window, stratified by high-to-low tau hierarchical set center ratio (i.
e.
, median segmentation) or stratified
based on distance based on high-to-low center connection.
The authors found that the overall tau-PET and tau-center ratios had a significant interaction with the annual tau accumulation rate, and the higher the tau-PET baseline level, the higher the tau-center ratio, the faster
the tau accumulation.
Similarly, the authors found an interaction between the overall tau-PET and tau-center ratio in symptomatic AD patients, with higher tau-PET levels and higher tau-center ratios leading to younger age at symptomatic symptoms
.
These analyses support the idea that a higher tau-centered ratio is associated with faster tau protein accumulation at pathological levels of a given tau protein in AD and younger tau protein accumulation in symptomatic AD
.
Similarly, more overall connectivity centers are associated
with subsequent faster tau accumulation.

Figure 4.
A higher tau center ratio contributes to tau diffusion

A higher tau-centered ratio mediates an association between younger age and faster tau accumulation in symptomatic AD

Finally, the authors assess in symptomatic AD patients whether the effect of a younger age of symptom onset on the faster rate of tau accumulation is mediated
by a higher tau-centered ratio.
Through 10,000 iterations of self-mediation, the authors found significant mediating effects
between ADNI (β=-0.
065, 95%CI[-0.
153;0.
00], p=0.
039) and BioFINDER (β=-0.
149, 95%CI[-0.
371;0.
00], p=0.
046) under the premise of controlling for gender, educational attainment, and diagnosis.
As shown in Figure 5A: a higher tau-centered ratio mediates an association
between younger and faster tau accumulation rates in patients with symptomatic AD.
These data were consistent
when included in hippocampal tau-PET.
Using the longitudinal cognitive data available in symptomatic AD patients in ADNI samples, it is further suggested that more accumulation of tau protein per year is associated with faster memory decline (i.
e.
, ADNI-mem, β=0.
322, p=0.
004, Figure 5B, controlling for age, sex, and educational attainment), suggesting that rapid accumulation of tau mediated by a higher tau-centered ratio may lead to faster clinical deterioration
in younger patients with clinical AD.
Again, this association was consistent when hippocampal tau-PET was included (β-0.
315, p=0.
004).

Using the sliding window method described above to confirm these analyses, the authors found that the interaction between baseline global tau-PET and tau-centered ratios affected annual changes in ADNI-MEM, with higher tau-centered ratios associated
with faster cognitive decline at higher overall tau levels.

Figure 5.
The association between younger age and faster tau accumulation is mediated by a higher tau-centered ratio

discuss

Key findings: First, younger patients with symptomatic AD exhibited more central-like PET assessment tau deposition patterns, with strong tau pathology in highly connected brain regions, which is critical to overall cognition and has previously demonstrated that it is central
to maintaining cognitive performance in AD patients.
Importantly, these associations
were not detected in amyloid-positive but asymptomatic individuals (i.
e.
, preclinical AD).
This supports previous ideas that tau pathology is a key driver of AD symptoms; Important new evidence has also been added that stronger tau deposition in the central area of the overall connection is associated
with early presentation of AD symptoms.
Second, the authors found that in the global connectivity central area, the younger the age at which symptoms appear and the greater the pathological load of tau, predicting faster subsequent tau accumulation in symptomatic AD patients, which is associated with
faster memory loss.
Hypothesis that connectivity is a key vector for tau transmission, so these results support that early in the disease process, pathology of the overall connectivity center may lead to wider spread of tau pathology in connected brain regions, leading to a faster cognitive deterioration
in patients with early symptomatic AD.
In symptomatic AD patients, the association between younger and faster tau accumulation is mediated by a stronger TA pathology in the central region
.
Taken together, this article suggests that tau pathological deposition predispositions are critical to cognition-critical overall connectivity brain centers, may determine early manifestations of symptomatic disease and accelerate connection-mediated tau transmission
.
Patients with younger manifestations of AD symptoms exhibit stronger tau pathology
in the overall junction of the frontoparietal symphysis cortex.
This finding is very consistent with previous post-mortem and in vivo tau-PET evidence, suggesting that younger onset of symptomatic AD is associated with more neocortical marginal retention-tau pathologic patterns, while older age of symptom onset is associated
with borderline-dominated tau pathologic patterns.
Similarly, previous studies have shown that younger onset of symptomatic AD is associated with stronger frontopatoparietal gray matter atrophy, hypoperfusion, and reduced glucose metabolism, which have been shown to be closely related
to tau pathology.
Taken together, this suggests that it is the spatial pattern and not just the degree of tau pathology that determines the likelihood, type, and aggressiveness of the symptomatic manifestations of
AD.
The authors previously found that higher global connectivity to frontoparietal control network hubs, such as the left frontal cortex, was associated
with more efficient overall brain network topology and reduced effects of posticulate gyrus glucose metabolism, hippocampal atrophy, or endoolfactory tau pathology on cognitive performance.
This suggests that maintaining the integrity of the central region may promote resilience to the cognitive effects of AD pathology
.
In turn, damage to central regions has been shown to reduce effective communication between brain networks, while clinical studies in patients with cerebrovascular disease have shown that focal lesions in the central region lead to disruption of the whole brain network and are more severe
than lesions in non-central regions.
In preclinical and clinical studies, Tau pathology has been shown to disrupt neuronal activity and connections, so early occurrence of tau pathology in the hub region can lead to early impairment of overall brain network function, leading to earlier appearance of cognitive impairment
.
Here, the next step to focus on is whether higher tau pathology in the hub can cause the hub connection to be interrupted, leading to the appearance of
early symptoms.
For the authors' second finding, the authors demonstrate that earlier onset of symptoms in symptomatic AD is associated with a stronger tau-centered ratio associated with faster tau accumulation, an important extension of previous preclinical and clinical work suggesting that brain connections mediate tau diffusion
.
Studies in mouse and neuronal cell cultures have found that tau propagates across synapses, with higher synaptic activity promoting tau propagation
.
Similarly, the authors report in a combined MRI and tau-PET study of AD patients that tau-based seed ligations predict tau diffusion patterns, where tau preferentially diffuses from the center to the junction region
.
Thus, a tau center region with wide connections can enhance the diffusion
of tau to the connected region compared to non-central regions with fewer connections.
In support of this idea, the authors found that young, symptomatic AD patients had the highest tau pathological region (i.
e.
, center) in the overall connected brain region, and that a stronger tau-centered ratio was associated
with faster tau accumulation.
Importantly, the authors' findings also provide an explanatory mechanism
for previous results on faster tau accumulation in younger AD patients.
Importantly, the rapid accumulation of tau is associated with a rapid decline in memory, which supports the view that the
rapid accumulation of tau drives rapid clinical progression in AD patients with early symptomatic manifestations.

Together, the authors' findings suggest that symptomatic younger AD patients exhibit stronger tau pathology in the hubs of overall connectivity, which may drive faster tau transmission and accelerate cognitive decline
.
This suggests that early symptom presentation is not driven by specific pathophysiological features, but by a pattern of tau distribution that preferentially targets brain centers
that are important for cognitive function.
These results are in good agreement with previous studies, suggesting that different patterns of tau distribution are associated
with different clinical trajectories, different episodes of symptoms, and disease progression.
Therefore, identifying potential determinants of spatially variable tau pathogenesis will be a key goal for the future, such as differential gene expression patterns, pre-existing tau pathology, premorbid differences and/or heterogeneity
in brain network architecture.

： ，
。   ， ，