| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online March 3, 2005 as doi:10.1096/fj.04-3210fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Pathology, Division of Neuropathology, University of Washington, Seattle, Washington, USA
1 Correspondence: Department of Pathology, Division of Neuropathology, University of Washington, Seattle, Washington 98105, USA. E-mail: tmontine{at}u.washington.edu
SPECIFIC AIMS
Neurofibrillary tangles (NFTs), hallmark pathologic inclusions of Alzheimer’s disease (AD) most closely associated with dementia, are composed primarily of an aberrant form of the microtubule binding protein tau. The aim of this study was to perform proteomic analysis using liquid chromatography (LC)/mass spectrometry (MS)-MS of tryptic digests of laser-captured NFTs to identify other NFT-associated proteins and to pursue immunohistochemical and biochemical validation and characterization of selected proteins identified by this unbiased survey.
PRINCIPAL FINDINGS
1. Discovery: proteomic analysis of tryptic digests from laser-captured NFTs from AD brain using LC/MS-MS
NFTs were visualized by anti-tau immunohistochemistry and obtained by laser capture microdissection (LCM) from pyramidal neurons in hippocampal sector CA1, solubilized in formic acid with sonication, digested with trypsin, the resulting peptides analyzed by LC/MS-MS, and the data analyzed by SEQUEST against the International Protein Index (IPI) database, followed by probability calculation with PeptideProphet and ProteinProphet. We identified a total of 155 proteins in laser-captured NFTs, 72 of which were identified by multiple unique peptides. Of the 72, 63 proteins had no previously known association with NFTs. Functional class distribution of these 72 proteins was 18% cytoskeletal or structural, 29% energy and metabolism, 10% extracellular matrix, 12% intracellular trafficking or synaptic, 4% nuclear, 10% signaling, 10% stress response or chaperon, and 7% unknown or unassigned. Expected proteins such as tau, apolipoprotein E (identified by single peptide), and 
-synuclein were identified in our proteomic screen, confirming the ability of this approach to identify NFT-associated proteins.
2. Validation: GAPDH colocalizes to most NFTs and binds to PHF-tau in vivo
Given the limitations of LCM and the need to validate proteins identified in proteomic database searches, we set out to confirm selected candidate proteins by immunohistochemistry in AD (n=3) and age-matched control (n=2) hippocampus. We selected three proteins identified by multiple unique peptides in our proteomic study that had no known prior association with NFTs and that had commercially available antibodies; these were clathrin, ubiquitin carboxyl-terminal hydrolase isozyme L1 (UCH-L1), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Both clathrin and UCH-L1 showed widespread immunolocalization in gray matter to cell soma and neuropil, demonstrating that these proteins were not specifically localized to NFTs. In contrast, GAPDH colocalized with the majority of NFTs as well as plaque-like structures in sections from AD patients (Fig. 1
); there were no immunoreactive structures in sections from controls. As biochemical validation, we confirmed that GAPDH was detectable in LCM NFTs by Western blot and proteolyzed by trypsin. Thus, our immunohistochemistry and immunoblot experiments validated our proteomics investigation that identified GAPDH as a protein component of NFTs, and suggested that GAPDH may also be a component of senile plaques (SPs). Others have shown that some antibodies to Aß peptides, the major protein constituent in senile plaques of AD, cross-react with GAPDH, raising the possibility that the converse may be true, viz., GAPDH antibody may have recognized Aß peptides. We ruled this out using Western blots to show that our GAPDH antibody did not react with Aß peptides in homogenates from AD temporal cortex. In addition, immunoprecipitation of homogenates from AD hippocampus with PHF-1, an antibody that recognizes an abnormal form of tau in AD, coprecipitated GAPDH while immunoprecipitation with tau antibodies did not coprecipitate GAPDH; stripping and reprobing blots showed that both antibodies immunoprecipitated tau. Because PHF-tau is present in pathologic structures other than NFTs (e.g., dystrophic neurites in SPs), these results do not prove localization of GAPDH to NFTs. Nevertheless, in combination with our immunohistochemistry and immunoblotting results, these findings strongly support a direct interaction between GAPDH and PHF-tau in NFTs in vivo.
|
3. Characterization: GAPDH accumulates in the detergent-insoluble fraction in AD brain but not in controls
A key pathogenic event in AD and several other neurodegenerative diseases is the accumulation of abnormal forms of proteins that acquire new biochemical characteristics; in the case of AD this is widely determined as the formation of detergent (sarkosyl)–insoluble, but at least partially formic acid-soluble PHF-tau and Aß. Serial extractions of temporal cortex were prepared from five patients who died of AD (age=80±1 year, 3 women, PMI=4.3, median Braak stage of VI) and four age-matched controls (age=84±2 year, 1 woman, PMI=4.7 h). Western blots showed the expected serial extraction of Aß and PHF-tau with abundant Aß and PHF-tau immunoreactivity in the sarkosyl-insoluble/formic acid soluble fraction in all AD patients, but none of the controls. Remarkably, GAPDH was present in the sarkosyl-insoluble/formic acid-soluble fraction only from AD patients (Fig. 2
).
|
CONCLUSIONS AND SIGNIFICANCE
NFTs are hallmark pathologic inclusions formed in AD that are closely associated with dementia and are composed primarily of PHF-tau. Several other proteins have been colocalized to NFTs; however, we are unaware of any unbiased assessment of NFT constituents. We hypothesized that unbiased discovery of NFT constituent proteins would provide new insight into protein-protein interactions that might be important in AD pathogenesis. Here we tested this hypothesis by LC/MS/MS proteomic analysis of laser-captured NFTs, followed by characterization of selected proteins so identified.
Our novel findings were 1) discovery of 63 proteins in NFTs identified by multiple unique peptides that had no prior association with NFTs, 2) validation by immunohistochemistry that one of three proteins selected from among these 63, GAPDH, colocalized to most NFTs, was present in LCM NFTs by Western blot, and immunoprecipitated with PHF-tau from AD temporal cortex, and 3) characterization of GAPDH as one of the few proteins whose conversion to a detergent-insoluble state is associated with AD (see Fig. 3
).
|
GAPDH was initially described as a glycolytic enzyme but is now appreciated to have pleiotrophic actions independent of energy metabolism that may influence neurodegeneration; these include roles in membrane fusion and endocytosis, microtubule binding, regulation of mRNA, and nuclear translocation in apoptosis. Others have colocalized GAPDH to Lewy bodies in Parkinson disease and to proteins that translate trinucleotide repeat expansions, such as huntingtin. With respect to AD, GAPDH has been shown to bind amyloid precursor protein in vitro, and cultured human fibroblasts from AD patients have abnormal subcellular distribution of GAPDH.
Although many NFT-associated proteins we discovered here are yet to be validated and characterized, these data will serve as a starting point for other investigators pursuing protein-protein interactions that underlie NFT formation. Moreover, our results with GAPDH highlight the power of unbiased proteomic techniques in pursuit of previously overlooked interacting proteins that may influence the pathogenesis of neurodegenerative diseases.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.04-3210fje;
This article has been cited by other articles:
|
|
 |
|
  M. VanRollins, R. L. Woltjer, H. Yin, J. D. Morrow, and T. J. Montine F2-Dihomo-isoprostanes arise from free radical attack on adrenic acid J. Lipid Res., May 1, 2008; 49(5): 995 - 1005. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Nakajima, W. Amano, A. Fujita, A. Fukuhara, Y.-T. Azuma, F. Hata, T. Inui, and T. Takeuchi The Active Site Cysteine of the Proapoptotic Protein Glyceraldehyde-3-phosphate Dehydrogenase Is Essential in Oxidative Stress-induced Aggregation and Cell Death J. Biol. Chem., September 7, 2007; 282(36): 26562 - 26574. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. N. G. Anderson and S. G. N. Grant High throughput protein expression screening in the nervous system - needs and limitations J. Physiol., September 1, 2006; 575(2): 367 - 372. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Boutte, R. L. Woltjer, L. J. Zimmerman, S. L. Stamer, K. S. Montine, M. V. Manno, P. J. Cimino, D. C. Liebler, and T. J. Montine Selectively increased oxidative modifications mapped to detergent-insoluble forms of A{beta} and {beta}-III tubulin in Alzheimer's disease FASEB J, July 1, 2006; 20(9): 1473 - 1483. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
