Finally, a Biomarker to Distinguish Primary Tauopathies?
Quick Links
While researchers enjoy a wealth of Alzheimer’s disease biomarkers, none have existed to distinguish tauopathies from one another—until now. In the November 24 Nature Medicine, scientists led by Randall Bateman and Chihiro Sato, Washington University School of Medicine, St. Louis, reported that two fragments containing tau’s microtubule-binding region drop in the cerebrospinal fluid the same as they accumulate in the brain. Lower CSF levels distinguished corticobasal degeneration (CBD) from other tauopathies with 80 to 89 percent accuracy. These tau markers will enable researchers to enroll people who have CBD into clinical trials and, eventually, improve CBD diagnosis, the authors concluded.
- Tau fragments accumulate in the brains of people with CBD and FTLD-MAPT.
- These microtubule-binding region snippets also fall in the CSF.
- This drop distinguishes CBD from controls and other tauopathies.
“This is a very important paper. The lack of biomarkers for CBD has been a major obstacle to understanding and curing this disease,” wrote James Rowe, University of Cambridge, U.K. “This will enable the development and targeting of anti-CBD treatments, and the ability to study CBD in lifetime with confidence and specificity.” (Full comment below.)
Tauopathies are classified by the type of tau that aggregates. In CBD, progressive supranuclear palsy (PSP), argyrophilic grain disease (AGD), and frontotemporal dementia caused by mutations in the MAPT gene (FTLD-MAPT), tau containing four repeats (4R) predominantly forms deposits, while in Pick’s disease (PiD) it is 3R-tau. Neurofibrillary tangles in AD comprise equal parts 3R- and 4R-tau. Could tau species beyond 3R and 4R isoforms distinguish specific tauopathies?
Previously, first author Kanta Horie and colleagues found that tangles and cognitive decline in AD correlated with CSF concentrations of a tau fragment, MTBR-243, corresponding to amino acids 243 to 254 of the microtubule-binding region (Dec 2020 news). Suspecting that other MTBR fragments may be amiss in primary tauopathies, he used mass spectrometry to analyze insoluble tau fragments extracted from brain tissue of 47 tauopathy cases. Seven had had AD, eight had FTLD-MAPT, 16 PSP, 12 CBD, three PiD, and one had AGD. As controls, he tested samples from 12 tau-negative FTLD cases who had evidence of TDP aggregates (FTLD-TDP), and from three cognitively healthy older adults. All samples were from the Memory and Aging Center at the University of California, San Francisco. Horie measured the ratio of tau fragment concentrations to that for total tau.
In the brain, high levels of two insoluble tau fragments distinguished CBD and FTLD-MAPT from other tauopathies. Concentrations of MTBR-275, comprising amino acids 275 to 280, and MTBR-282, spanning residues 282 to 290, were fivefold higher in CBD than in most other tauopathies and controls, and threefold higher in FTLD-MAPT. MTBR-282, but not MTBR-275, was also 2.5-fold higher in AD cases than in controls (see image below).
Brain Differences. MTBR-275 (left) and MTBR-282 (right) accumulate to higher levels in the brains of people with CBD and FTLD-MAPT than they do in other tauopathies or healthy control brains. MTBR-282 was also higher in AD. [Courtesy of Horie et al., Nature Medicine, 2022.]
What about in the CSF? Horie measured the two fragments in samples from 29 healthy older adults, 21 people pathologically confirmed to have had FTLD-TDP, and 62 who had tauopathies. Of the latter, 10 had AD, five FTLD-MAPT, 22 PSP, 18 CBD, five PiD, and two had AGD. All were from UCSF Memory and Aging Center. People who had CBD, FTLD-MAPT, or AD had about 20 percent less of both MTBR fragments in their CSF than did people with the other tauopathies or controls.
The data suggest that as these fragments accumulate in the brain, their concentration falls in the CSF. Indeed, for 54 participants for whom brain and CSF samples were available, MTBR-275 and MTBR-282 concentration in the latter inversely correlated with that in the brain (see image below). “These fragments seem to behave like the Aβ42/40 ratio; in the presence of pathology, you see fewer MTBR fragments in relation to total tau,” wrote Henrik Zetterberg of the University of Gothenburg in Sweden (comment below).
Brain vs. CSF. Levels of MTBR-275 (left) and MTBR-282 (right) in the brain (x-axis) inversely correlated with their respective levels in the CSF (y-axis), especially in people who had CBD (green). [Courtesy of Horie et al., Nature Medicine, 2022.]
Could the MTBR fragments be useful for differential diagnoses of primary tauopathies? The scientists focused on CBD because it is hard to diagnose clinically, with accuracies of about 50 percent at best, according to the authors. In a separate cohort, CSF MTBR-275/total tau and MTBR-282/t-tau ratios distinguished 18 people with CBD from 29 cognitively normal older adults with areas under the curve of 0.81 and 0.89, respectively. AUC is a measure of specificity and sensitivity in which 1.0 is the highest score. Notably, the MTBR fragments distinguished CBD from all other tauopathies, including 21 people with FTLD-TDP, 22 diagnosed with PSP, five with PiD, and two with AGD, with AUCs of 0.80 to 0.89. The authors believe that CSF MTBR-275/t-tau and MTBR-282/t-tau ratios may be the first specific biomarkers for CBD.
Jeffrey Dage of Indiana University, Indianapolis, agreed. “The MTBR-tau peptides are not good biomarkers on their own, but, when ratioed with a mid-domain peptide that varies by diagnostic group, it highlights differences in expression, processing, and/or clearance among proteoforms,” he wrote (comment below).—Chelsea Weidman Burke
References
Mutation Interactive Images Citations
News Citations
Further Reading
Primary Papers
- Horie K, Barthélemy NR, Spina S, VandeVrede L, He Y, Paterson RW, Wright BA, Day GS, Davis AA, Karch CM, Seeley WW, Perrin RJ, Koppisetti RK, Shaikh F, Lago AL, Heuer HW, Ghoshal N, Gabelle A, Miller BL, Boxer AL, Bateman RJ, Sato C. CSF tau microtubule-binding region identifies pathological changes in primary tauopathies. Nat Med. 2022 Dec;28(12):2547-2554. Epub 2022 Nov 24 PubMed.
Annotate
To make an annotation you must Login or Register.
Comments
Indiana University School Of Medicine
In this manuscript by Horie et al., Bateman and Sato have advanced a new immunoprecipitation (IP) mass spectrometry method to measure specific proteoforms of tau in CSF that can potentially identify pathological changes in primary tauopathies. These findings are encouraging and suggest there may be additional proteoforms of tau, some yet to be identified, that shed light on differences between the primary tauopathies.
Interestingly, they used a two-stage IP approach. First, they depleted samples of anything binding to N-terminal and mid-domain antibodies (Tau1, HJ8.5 and HJ8.7) and then captured what was left with a C-terminal anti-tau antibody (77G7). Despite the complexity of the method, they were able to demonstrate stable and reproducible measures over four months from repeat lumbar punctures. The method measures two peptides from the microtubule binding region (MTBR) ratioed to a mid-domain tau peptide. The two MTBR peptides are both from the second repeat region that is unique to the 4-repeat form of tau. There is significant overlap between diagnostic groups, but the incorporation of this new measure, CSF MTBR-tau275/t-tau, with other CSF biomarkers could help differentiate primary tauopathies.
Unfortunately, in the clinically diagnosed cohort (Extended Data Fig. 2), the overall levels were a bit different than the pathology-confirmed cohort and the cut-point (Extended Data Table 2; <0.00563) was not transferable. This would seem to indicate the presence of cohort effects that will need some additional studies to understand. Despite the observed cohort bias, there was still some significant lowering observed, and indications are that these new measures have potential to help differentiate primary tauopathies.
The reason for exploring the ratio when these specific MTBR-tau peptides on their own show no differences between groups (Supplementary Fig. 1) is also intriguing, especially considering that not all MTBR peptides, or even 4R MTBR peptides, show the same lack of difference. The MTBR-tau peptides are not good biomarkers on their own, but ratioed against a mid-domain peptide that does show variation by diagnostic group, it acts to highlight differences in expression, processing, and/or clearance among proteoforms. Perhaps the authors will ratio the MTBR peptides with p-tau217 next to maximize the effects of both hyperphosphorylation and aggregation?
University of Gothenburg
This is such a clever way of visualizing C-terminal tau fragments in CSF. The authors immunodepleted the more abundant N- and mid-domain tau fragments and then specifically pulled out C-terminal fragments, which are much less abundant, and then analyzed them by mass spectrometry. These C-terminal fragments seem to reflect brain tissue tau, and the resulting biomarker behaves a little bit like the Aβ42/40 ratio—in the presence of pathology, you see a reduction in relation to total-tau.
Similar results have been seen for another C-terminal tau fragment, tau368, in Alzheimer’s disease (Blennow et al., 2020), but this is the first time that non-AD tauopathies have been examined in this way. These results show that non-AD tauopathies result in a discernible tau fluid biomarker change, the absence of which has been very puzzling before.
References:
Blennow K, Chen C, Cicognola C, Wildsmith KR, Manser PT, Bohorquez SM, Zhang Z, Xie B, Peng J, Hansson O, Kvartsberg H, Portelius E, Zetterberg H, Lashley T, Brinkmalm G, Kerchner GA, Weimer RM, Ye K, Höglund K. Cerebrospinal fluid tau fragment correlates with tau PET: a candidate biomarker for tangle pathology. Brain. 2020 Feb 1;143(2):650-660. PubMed.
Cambridge University
This is a very important paper. The lack of positive biomarkers for corticobasal degeneration has been a major obstacle to understanding and curing this disease. CBD is often mimicked by other diseases (i.e., there are multiple causes of corticobasal syndrome), while CBD in turn can mimic other diseases. In recent years, Alzheimer’s disease-specific biomarkers have been helpful in ruling out AD as the cause of a corticobasal syndrome, but this diagnosis by exclusion has been problematic—if, for example, there is co-pathology, or if the CBS is caused by neither CBD nor AD. Recent cryo-EM studies show how tau deposited in CBD is distinctive. Now there is a fluid biomarker that can also distinguish CBD from other dementias, even other “4R-tau” diseases.
At last, we now have Horie et al.’s result, describing a fluid biomarker that is strongly indicative of CBD more than of PSP, and distinguishes CBD from Alzheimer’s and Parkinson’s disease. The MTBRs may have even better specificity and sensitivity for FTLD-tau’s when used in conjunction with other fluid biomarkers (e.g., p-tau’s and neurofilament light chain). A future blood-based analogue would increase scalability and accessibility.
This will enable the development and targeting of anti-CBD treatments, and the ability to study CBD in a person’s lifetime with confidence and specificity.
The study is strengthened by the mapping of co-pathologies, the comparison of brain versus CSF, and suggested CSF thresholds for diagnosis. They also demonstrate reliability on repeat assessment, and good performance in clinically diagnosed cohorts, which are vital for the assay to be useful. Larger, longitudinal, and early stage or presymptomatic cohorts will now be able to validate the assay independently, and explore the relationship to severity and progression, as a prelude to clinical trials. The strength of correlation between brain and CSF MTBRs is, in part, driven by group differences, and correlations are weak within groups (an example of Simpson’s paradox), but this may reflect the relatively small group sizes.
We can look forward to accelerated progress in diagnosis and clinical trials for FTLD-tau disorders, with special optimism for CBD, which cannot readily be confirmed by other means in vivo.
Washington University in St. Louis
I'm continually impressed by the wealth of information that the different MTBR species can provide. Although the increasing arsenal of ptau sites have provided additional insights, they primarily seem to reflect a response to amyloidosis. This work, along with prior work from Dr. Horie, clearly shows that MTBR can track tauopathies.
Make a Comment
To make a comment you must login or register.