Biomarker Discovery and Validation in Progressive Supranuclear Palsy (PSP)

Ted Dawson
Ted Dawson, MD, PhD
Johns Hopkins University (Baltimore, Maryland)
Alexander Pantelyat
Alexander Pantelyat, MD
Johns Hopkins University (Baltimore, Maryland)


The goals of this application are to identify cerebrospinal fluid (CSF) biomarkers that reliably distinguish between PSP, PD and healthy individuals using TMT-based multiplexing mass spectrometry technology for discovery and targeted parallel reaction monitoring (PRM) mass spectrometry for validation. In the discovery phase global pallidus tissue will also be analyzed from 15 PSP patients and 15 PD as well as 15 HC individuals using TMT-based multiplexing mass spectrometry.  A cohort of 40 PSP, 40 PD and 40 healthy control individuals will be recruited for this project and followed over a three year time period.  Biofluid biosamples will be collected and deposited with the NINDS BioSEND Repository and clinical data will be shared through the PDBP data management resource.


Progressive supranuclear palsy (PSP) is a devastating atypical parkinsonian disorder that currently lacks meaningful symptomatic therapies, reduces lifespan and greatly impairs daily function and quality of life. It is often difficult to distinguish from Parkinson disease (PD) clinically, which is crucial for appropriate and timely management, prognosis and clinical trial enrollment. Despite a critical need for a reliable diagnostic marker for parkinsonian disorders, there is currently no biomarker that can be used in routine clinical practice to distinguish between PSP and PD. The purpose of this project is to discover cerebrospinal fluid (CSF) biomarkers that reliably distinguish between PSP, PD and healthy individuals. The difficulty of identifying reliable biomarkers can be attributed to the variability of clinical samples, low abundance of proteins that are involved in the pathogenesis of PSP and PD, and the lack of reproducibility in validating biomarker candidates. To overcome these limitations, we propose use of a large CSF cohort with greater statistical power for true discovery, and deep proteome analysis to reveal PSP biomarkers that are involved in PSP pathogenesis, but are present at low abundance. In addition, multiplexed sample analysis by isobaric tandem mass tagging (TMT) with a common reference for data normalization will ensure robust analytical precision of quantitative proteomic data for discovery from a larger set of samples. Moreover, additional proteomic analysis of brain tissue will be used to select those biomarkers that show differential expression in CSF as well as the globus pallidus, a representative brain region used to pathologically define PSP. These discovery platforms will utilize a bioinformatics approach to select the most plausible candidates for targeted validation studies followed by an intensive validation of the discovered biomarker candidates. To achieve these goals, we propose four aims: Specific Aim 1: To prospectively collect CSF on patients with clinically well-characterized PSP. Specific Aim 2: To discover proteins that are differentially expressed in patients with PSP compared to controls and PD. We plan to carry out a quantitative proteomic analysis of CSF and globus pallidus samples from patients with PSP, PD and from controls by employing TMT-based multiplexing technology. With this approach, we expect to obtain a more comprehensive coverage of a larger number of proteins quantified across the analyzed samples. Specific Aim 3: To prioritize PSP biomarker candidates based on an integrative analysis of alterations in CSF and globus pallidus. By integrating the expression changes in CSF and brain tissue with a network approach that takes advantage of the known biological pathways that have been described in PSP, our proposal will be able to select reliable PSP biomarker candidates for validation by targeted PRM experiments. Specific Aim 4: To validate candidate protein biomarkers in a larger cohort using targeted parallel reaction monitoring (PRM) mass spectrometry using CSF samples from a PSP cohort at Johns Hopkins University, the University of Pennsylvania, UCSF and PDBP. Biomarkers that are selected by algorithms based on these PRM experiments will finally be confirmed using blinded PDBP CSF samples from PSP and will be compared to CSF samples from PD. Through the approaches outlined above, we expect to discover and validate reliable PSP biomarkers that are distinguishable from PD in a reproducible manner.


1. Establish biomarker cohort of 40 PD, 40 PSP and 40 healthy control individuals.  Biofluid biosamples (DNA, RNA, plasma and CSF) will be collected and deposited with the NINDS BioSEND repository.  Individuals will be followed for 3 years at 12 month intervals.

  1. *Limited number of 18 month visits were conducted due to the impact of COVID-19 related closures

2. Mass spectrometry and CSF samples and brain tissue will be used to identify protein signatures that differentiate between PD, PSP and healthy controls.




Please contact:

Anna J. Hall
Clinical Coordinator
Johns Hopkins University

Available Data Types