Elevated Plasma Tau-PT217 Linked to Decreased Hippocampal Functional Connectivity in Patients with Knee Osteoarthritis

Abstract

Background:

Osteoarthritis is associated with a higher risk of developing dementia, though the underlying biological mechanisms have remained unclear. Recent studies suggest that blood phosphorylated tau proteins, particularly Tau-PT217, are sensitive biomarkers capable of detecting cognitive decline in its early stages, making it useful for early diagnosis of Alzheimer’s disease and other forms of cognitive impairment.

Methods:

In this study, we investigated the plasma phosphorylated tau protein levels (Tau-PT217 and Tau-PT181), hippocampus functional connectivity, and cognitive function in people with knee osteoarthritis compared to age and gender matched pain-free controls.

Results:

We found that knee osteoarthritis was associated with increased plasma levels of Tau-PT217 (but not Tau-PT181), and that the Tau-PT217 is also correlated with reduced hippocampal functional connectivity with middle cingulate cortex. Our findings suggest a potential biological correlation between knee osteoarthritis and an elevated risk of dementia, contributing valuable insights that may guide the formulation of early intervention and preventative strategies to mitigate dementia in individuals diagnosed with knee osteoarthritis.

Introduction

As chronological age increases, there is a tendency for a higher risk of cognitive decline and dementia. This trend may result from cumulative damage to brain cells and tissues over time. While some generally healthy individuals reach the threshold for symptom onset at roughly the same age, others may experience different rates of biological aging due to genetic variations or exposure to various factors or disorders. These differences may subtly accelerate the accumulation of age-related damage.

One such contributing factor is osteoarthritis (OA), the most common form of arthritis and a degenerative joint disease often resulting in chronic joint pain. Many studies have found that people with OA are more likely to develop dementia than those without OA^1–5. The biological links between OA and dementia are complex and may involve multiple factors. For example, both conditions are associated with chronic inflammation. OA causes inflammation of the joint lining and releases pro-inflammatory cytokines into the bloodstream⁶, which could contribute to the neuroinflammation that is associated with cognitive decline and dementia. A study using the mouse model of OA found that OA accelerated and exacerbated neuroinflammation and subsequently led to Alzheimer’s disease (AD) pathology⁷. The altered expression of pro-inflammatory cytokines in OA may induce reactive oxygen species (ROS) production, which causes elevated oxidative stress throughout the body^8,9. The brain is particularly susceptible to oxidative damage, which plays a key role in neurodegenerative processes leading to dementia^10–12. Vascular changes, including abnormal blood flow and perfusion in the subchondral bone, are also associated with pathogenesis and progression of OA¹³. These vascular changes specifically reduced cerebral blood flow that contributes to cognitive impairment and dementia¹⁴.

All of these factors discussed above, inflammation, oxidative stress, as well as vascular changes are reported to be associated with tau deposition^15–17, the abnormal accumulation of tau proteins in the brain, which could lead to cognitive decline. A recent study investigating the AD pathology related to OA, APOE-ε4 allele and tau accumulation found OA was indeed associated with higher tau deposition¹⁸.

Tau proteins, a group of proteins primarily found in neurons, are responsible for stabilizing microtubules in axons, which are crucial for maintaining neuronal structures. Their functions are regulated by phosphorylation and dephosphorylation. Hyperphosphorylation of tau is a hallmark of AD^19,20. Plasma Tau-PT217 (tau phosphorylated at threonine 217) and Tau-PT181 (tau phosphorylated at threonine 181) have been increasingly recognized as valuable biomarkers over the years for AD diagnosis due to their accuracy, non-invasive nature and high sensitivity^21–23. Furthermore, existing research has shown significant correlations between plasma tau biomarkers and tau PET imaging. Additionally, plasma tau biomarkers have exhibited robust sensitivity in detecting brain changes associated with Alzheimer’s disease^21,24,25. Recently, the potential role of tau protein in the elevated prevalence of Alzheimer’s disease among chronic pain patients has attracted considerable attention from researchers. An animal study found chronic pain led to tau accumulations in the hippocampus of mice²⁶. Some investigators propose that tau proteins may represent a crucial molecular link between chronic pain and the pathogenesis of dementia²⁷. A 2023 study observed increased cerebrospinal fluid (CSF) levels of Tau-PT181 in patients with chronic pain²⁸. A more recent study showed that men with probable chronic pain had higher plasma total tau proteins, as well as smaller hippocampal volume compared to men without probable chronic pain²⁹. Given the increased dementia risk associated with OA, we hypothesize there might be detectable changes in plasma tau in patients with OA.

Brain imaging studies found that patients with knee osteoarthritis (KOA) had brain structural and functional changes³⁰. Altered brain functional connectivity (FC) between the hippocampus and thalamus/superior frontal gyrus was observed in people with KOA, which was correlated with serum brain-derived neurotrophic factor (BDNF) levels and memory scores³¹. There are strong links between the hippocampus and dementia, and hippocampal FC changes precedes brain structure changes in dementia^32–34. Studies found that the hippocampal changes were associated with MoCA (Montreal Cognitive Assessment) scores³⁵. Memory function improvements from mind-body exercises were observed to be correlated with hippocampal FC changes³⁶. Studies also found that hippocampal FC changes were associated with tau deposition³⁷.

In this study, we investigated the associations between cognitive decline and KOA by examining the plasma levels of Tau-PT217 and Tau-PT181 in patients with KOA, together with hippocampus resting state FC analysis and cognitive assessment using MoCA.

Methods

Study population

Inclusion/exclusion criteria for patients with KOA

Inclusion Criteria:

a) Volunteers 40–75 years of age; b) Meet the Classification Criteria of the American College of Rheumatology for osteoarthritis of the right and/or left knee for at least the past 3 months; c) Ability to read and understand English.

Exclusion Criteria:

a) Presence of any contraindications to MRI scanning (e.g. cardiac pacemaker, metal implants, fear of closed spaces, pregnancy). b) History of a chronic disease that, in the investigator’s judgment, precludes participation in the study because of a heightened potential for adverse outcome (e.g. asthma or claustrophobia).

Inclusion/exclusion criteria for HC (heathy controls)

Inclusion Criteria:

a) Volunteers 40–75 years of age; b) Have not experienced knee pain or other chronic pain in the past three months; c) Have not experienced any acute pain in the past week; d) Ability to read and understand English.

Exclusion Criteria:

a) Presence of any contraindications to MRI scanning (e.g. cardiac pacemaker, metal implants, fear of closed spaces, pregnancy). b) History of a chronic disease that, in the investigator’s judgment, precludes participation in the study because of a heightened potential for adverse outcome (e.g. asthma or claustrophobia).

The study was approved by the MGH Institutional Review Board in accordance with the Declaration of Helsinki. Informed consent was obtained from all participants after explaining the study’s purpose, procedures, risks, and benefits, and was documented in writing before the experiment began.

MRI data Acquisition

Each participant underwent brain MRI scan. The MRI scans were acquired at the Martinos Center for Biomedical Imaging using a 32-channel radiofrequency head coil in a 3T Siemens scanner. Structural brain images were obtained with a T1-weighted 3D multiecho magnetization-prepared rapid gradient-echo (MPRAGE) sequence with the following parameters: voxel size 1mm³, TR=2500ms, TE=1.69ms, flip angle=7°, slices=176, and field of view=256. Resting state functional MRI (fMRI) scans included 44 slices with slice thickness of 2.59mm, TR=3000ms, and TE=30ms. Participants were asked to keep their eyes open during the resting state fMRI scan.

Cognitive assessment - MoCA

All the participants underwent the MoCA which has 7 different sub-categories: visuospatial executive, naming, attention, language, abstraction, delayed recall and orientation scores. A total MoCA score as well as sub-scores for each category were collected for each participant. All assessments were performed by well-trained research staff.

Measurements of plasma Tau-PT217 and Tau-PT181

Participants underwent blood drawing at the time of the MRI visit. Peripheral blood samples were collected by venipuncture in EDTA tubes. The samples (about 10 mL) were immediately placed on ice and centrifuged for acquisition of plasma within one hour. The blood samples were processed at Massachusetts General Hospital. Blood samples were centrifuged at 2000 g for 15 minutes, and the plasma supernatant was collected into an EDTA tube. Tau-PT217 and Tau-PT181 were measured by nanoneedle (NanoMosaic, Waltham, MA) as described in previous studies^38,39. Blank nanoneedle chips were supplied by NanoMosaic. 5 μg/ml capture antibody was incubated with the nanoneedle chip overnight. It was then washed in PBS and incubated in the blocking solution for 1 hour. 5 μL of plasma samples were diluted 2 X into the dilution buffer provided by NanoMosaic. The diluted sample was incubated on the chip for 2 hours. After washing, 0.5 ug/ml detection antibody was incubated on the chip. We used phospho-specific antibodies to Tau-PT217 and Tau-PT181 (Cat#44–744 and Cat# MN1050, ThermoFisher) as the detection antibody. We used Tessie TM instrument (NanoMosaic) to image the chip, and the software provided by NanoMosaic to analyze the colors of all nanoneedles and reported the percentage of the color-shifted number of nanoneedles. The protein concentrations were reported in relative units specific to the nanoneedle technology (i.e., Nano Unit). Three different values for each pTau protein were measured and averaged in a single value for further analyses.

Seed-Based Functional Connectivity Analysis

The fMRI data was processed using CONN toolbox (v22a⁴⁰). In brief, pre-processing steps included: slice-timing correction, realignment, outlier detection, indirect segmentation and normalization on the MNI 152 template, smoothing with a Gaussian kernel of FWHM 8 mm, regression of nuisance covariates and head motion scrubbing and filter applied with a frequency range of 0.008–0.09 Hz. In addition, we employed segmentation of grey matter, white matter, as well as CSF for the removal of temporal confounding factors (white matter and CSF)⁴⁰. Finally, data was checked for motion correction using the artifact detection toolbox (ART - http://www.nitrc.org/projects/artifact_detect)^41–43.

A seed-based FC analysis was then conducted. The bilateral hippocampus was used as the region of interest (ROI) to assess the connectivity between the seed and rest of the brain. The ROI was extracted using automated anatomical labeling (AAL). FC measures were then computed between the seed and every other voxel in the brain. First-level correlation maps were produced by extracting the residual BOLD time course from the seed and by computing Pearson’s correlation coefficients between that time course and the time courses of all other voxels in the brain. Correlation coefficients were transformed into Fisher z-scores to increase normality and allow for improved second-level analyses.

Group analysis was applied in CONN using ANCOVA, including two groups (KOA and HC), with age and gender included as covariates. In addition, multiple regression analyses were applied between resting state FC, Tau-PT217 and the MoCA score across all subjects and the two groups separately. A threshold of voxel wise p < 0.005 uncorrected and p < 0.05 familywise error (FWE) corrected were used for whole brain analysis. Based on previous studies, we predefined cingulate cortex, medial prefrontal gyrus, and thalamus as regions of interest (ROIs) due to their important role in pain processing^39,40 and memory functions²³. A threshold of voxel wise p < 0.005 and p < 0.05 using the 3dFWHMx and 3dClustSim (as part of the Analysis of Functional NeuroImages program [http://afni.nimh.nih.gov] released in July 2017) were applied for the p value correction for pre-defined ROIs.

Statistical Analysis

Groups differences for age and gender were assessed by performing a t-test and chi-square test, respectively.

Linear regression was used to assess differences between groups in the protein’s levels concentration and MoCA scores, including age and gender as covariates. Clusters identified from the seed-based FC analyses of the hippocampus in groups comparison (HC vs KOA) were extracted. A linear regression model was then used to test whether hippocampus FC changes between groups were associated with the protein’s Tau protein concentration or MoCA scores. To further explore the relationship between KOA, Tau-PT217, and hippocampus FC, we conducted a mediation analysis to explore whether the Tau-PT217 was a mediator between hippocampus FC and group status (HC vs KOA). R programming language (v3.6.0) (R Core Team (2023). _R: A Language and Environment for Statistical Computing_. R Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org/) used to perform statistical analyses.

Results

Participants

22 KOA and 37 HC were included in data analysis. No significant differences were detected for age (p=0.45) and gender (p=0.97) between groups. Descriptive statistics of the dataset can be found in Table 1.

Table 1.

Demographic characteristic of the dataset used for the analysis.

Group	N	Age	Sex (n. female)	Tau 181	Tau 217	MoCA (Total Score)
All Subjects	59	58.2 ± 7.4	28 (47.4%)	0.59 ± 0.12	0.41 ± 0.07	27.9 ± 2.1
HC	37	58.8 ± 7.8	17 (45.9%)	0.58 ± 0.14	0.38 ± 0.06	28.1 ± 2.4
KOA	22	57.4 ± 6.7	11 (50%)	0.59 ± 0.09	0.45 ±0.08	27.6 ± 1.6

HC= Healthy Controls; KOA= Knee osteoarthritis

Proteins concentration – elevated plasma Tau-PT217 in KOA group

Linear regression model for groups comparison for the Tau-PT217 concentration showed significant differences between HC and KOA (p=0.0014), with higher Tau-PT217 concentration detected in the KOA group (mean ± SD, Tau-PT217= 0.45 ± 0.08) in comparison to the HC group (mean ± SD, Tau-PT217= 0.38 ± 0.06). The same analysis performed on Tau-PT181 did not show significant differences between the two groups (p=0.82, mean group ± SD, HC= 0.58 ± 0.14, mean group ± SD, KOA= 0.59 ± 0.09).

Cognitive assessment – MoCA

No significant differences were found in the total MoCA score between groups (p=0.46, mean ± SD, HC= 28.1 ± 2.4, mean ± SD, KOA = 27.6 ± 1.6). An exploratory analysis showed significant difference between groups in the abstraction sub-score of the MoCA, where the KOA participants showed lower scores, compared to the HC participants (p=0.04, mean group ± SD, HC= 1.94 ± 0.22, mean group ± SD, KOA= 1.76 ± 0.42). A marginally significant lower score was also observed for the naming score in the KOA group (p=0.06, mean group ± SD, HC=3.00 ± 0.00, mean group ± SD, KOA= 2.90 ± 0.29).

Seed-Based Functional Connectivity – decreased hippocampal FC in KOA group

The KOA group showed decreased hippocampus FC with the right temporal pole (RTP), middle and inferior temporal gyri (MTG, ITG), right fusiform gyrus (RFG), bilateral middle cingulate cortex (MCC), and bilateral anterior cingulate cortex (ACC)/medial prefrontal cortex (mPFC) in comparison with the HC participants (contrast: HC>KOA) (Table 2 and Figure 1). The opposite contrast (KOA > HC) did not show significant results.

Table 2.

Results the seed-based FC analysis. Increased hippocampus FC in HC, in comparison with KOA subjects (contrast HC > KOA).

Brain Region	Cluster Size	T	Peak Coordinate (x,y,z)
Right temporal pole, middle and inferior temporal gyri, fusiform gyrus	770	4.64	20, 6, 48
Bilateral middle cingulate gyrus	130*	3.93	6, 10, 32
Right ACC/medial prefrontal gyrus	73*	4.09	14, 48, 2

^*p<0.05 after 3dFWHMx correction

Figure 1.

Increased hippocampus FC in HC group, in comparison with KOA subjects. On the left: Increased FC between the hippocampus and bilateral middle cingulate cortex (A), and right ACC/medial prefrontal cortex (B). On the right: Scatter plot of the linear regression models between the hippocampus-bilateral MCC FC with Tau 217 protein (C) and hippocampus – right ACC/medial prefrontal cortex FC with MoCA scores (D).

Relationships between Tau-PT217, MoCA scores and Hippocampal FC

Association analysis showed that hippocampus - MCC FC was negatively correlated with Tau-PT217 concentration (p=0.026, β= −0.45). In addition, hippocampus-ACC/mPFC FC was positively correlated with total MoCA scores (p=0.008, β= 0.03) (Figure 1). No other significant association was detected.

Mediation analysis

We found that Tau-PT217 significantly mediated (p=0.044, β= −0.31 [−0.86 – −0.01]) the association between hippocampus-MCC FC and group category (HC vs KOA). A significant direct effect between hippocampus-MCC FC and group status (HC vs KOA) was also observed (p=0.02, β= −0.98 [−1.66 – −0.20]). The total effect was also significant (p<0.001, β= −1.29 [−1.91 – −0.62]).

Discussion

In this study, we investigated the plasma levels of Tau-PT217 and Tau-PT181, together with hippocampus resting state FC changes associated with KOA. We found that individuals with KOA showed an elevated level of plasma Tau-PT217 that was correlated with reduced FC between the hippocampus and MCC and that lower MoCA scores were correlated with reduced FC between the hippocampus and ACC/mPFC.

Increased plasma Tau-PT217

Both plasma Tau-PT217 and Tau-PT181 are biomarkers associated with the presence and progression of AD pathology and development⁴⁴. Studies found that Tau-PT217 had higher diagnostic accuracy and a stronger correlation with amyloid and tau pathology in the brain^45–47. In comparison to Tau-PT181, Tau-PT217 is a more sensitive biomarker as it shows changes earlier in the disease process⁴⁸ and detects longitudinal changes^21,49. Tau-PT217 may even be detectable years before clinical symptoms appear. In this study, elevated plasma Tau-PT217 in KOA patients indicates increased neuronal damage or degeneration. On the other hand, Tau-PT181 showed no significant changes between KOA and HC, suggesting that the KOA subjects in our study may still be in the early stages of tau pathology – given that Tau-PT217 is more sensitive than Tau-PT181. Lower MoCA sub-scores in KOA patients indicate potential cognitive decline to some degree.

Decreased hippocampal FC

Hippocampus is a key brain structure involved in memory formation, spatial navigation, and learning. Reduced hippocampal FC suggests weaker communication between the hippocampus and other brain regions, implicating cognitive decline specifically in memory function. It may be used as a biomarker for early detection of cognitive decline^50–52. Thus, our finding that KOA patients had decreased hippocampal FC is consistent with previous findings³¹ and aligns with the epidemiological evidence that people with KOA have a higher risk developing dementia. The hippocampus works in concert with other brain regions in converting short-term memory to long-term memory. Decreased FC between the hippocampus and specific brain regions identified in this study- namely the RTP, ITG, MTG, RFG, bilateral MCC and ACC/ mPFC - may indicate impairments across various cognitive domains. These potentially affected functions include memory formation and retrieval, visual information processing, object and face recognition, emotional memory integration, as well as pain-related cognitive processes^53–55. This pattern of reduced connectivity suggests a broad impact on cognitive functioning.

Association between plasma Tau-PT217, hippocampal FC and MoCA score

We observed two correlations in KOA group: 1) elevated plasma Tau-PT217 was correlated with decreased FC between hippocampus and MCC; 2) decreased hippocampus – ACC/mPFC FC was correlated with lower MoCA scores. Interestingly, both MCC and ACC/mPFC play important roles in pain processing^56,57. This suggests that chronic pain from KOA may increase cognitive load which may affect hippocampal function and cognitive performance. These findings highlight the association between plasma Tau-PT217 levels and hippocampal FC as well as their potential role as biomarkers of early cognitive decline.

Interestingly, we found that Tau-PT217 significantly mediated the association between hippocampus-MCC FC and group category (HC vs KOA), which further endorsed the role of Tau-PT217 in cognitive decline.

Cognitive changes in KOA

In this study, we did not detect significant difference on MoCA score between KOA patients and HC, this is consistent with findings from Liu et al³¹ investigated the cognitive decline in KOA. In the study, Liu et al. found significant memory impairments as measured by the Wechsler Memory Scale (WMS), a scale that specifically focuses on measuring different facets of memory, including working memory, immediate and delayed recall, and visual and verbal memory.

We also evaluated memory performance using the delayed recall MoCA subscore, but did not detect a significant difference. Instead, we found significant differences in one of the MoCA subscores (the abstraction score), where lower scores were observed in KOA patients. Further study with a larger sample size is needed to validate our findings further.

The inflammation connection

Though there is emerging evidence from epidemiological, biological and clinical studies suggesting a significant association between KOA and cognitive decline, the underlying mechanism remains unclear. We propose inflammation may be a key factor contributing to brain functional changes in people with KOA. The chronic low-grade inflammation characteristic of KOA could become systemic and affect the brain. Pro-inflammatory cytokines (e.g., IL-1, TNF-α) elevated in KOA⁵⁸ can cross the blood-brain barrier and induce neuroinflammation, affecting brain structures like the hippocampus which is susceptible to inflammatory damage. These cytokines could also promote tau hyperphosphorylation, possibly explaining the elevated plasma Tau-PT217 levels. Moreover, inflammation in KOA may increase oxidative stress which is also linked to tau pathology and hippocampal dysfunction.

Collectively, inflammation may contribute to both KOA progression and cognitive decline. These findings highlight the complex interplay between peripheral inflammation and central nervous system function.

One potential limitation of the study is the small sample size, which may limit the generalizability of the findings. Future studies with larger sample sizes are needed to further validate our findings.

Conclusion

In this study, we observed elevated plasma Tau-PT217 together with reduced hippocampus – MCC FC, lower MoCA score together with hippocampus-ACC/ mPFC FC, as well as lower MoCA abstraction sub-scores in people with KOA. These findings suggest that Tau-PT217 may play a critical role in explaining the heightened susceptibility to dementia observed in individuals with osteoarthritis. This insight could be pivotal in guiding the development of early intervention and preventative strategies to mitigate the higher risk of dementia in those diagnosed with knee osteoarthritis.

Highlights.

1. Knee osteoarthritis is associated with increased plasma levels of Tau-PT217

2. Tau-PT217 is correlated with reduced hippocampal functional connectivity with middle cingulate cortex.

3. MoCA score is correlated with hippocampus- anterior cingulate cortex / medial prefrontal cortex.

4. MoCA abstraction sub-scores is lower in people with KOA comparing to controls.