Biomarkers could be invaluable diagnostic and prognostic indicators for Alzheimer’s and other neurodegenerative diseases. Hardly a meeting goes by these days without mention of a new potential marker. Neurodegenerative Diseases: The Molecular and Cellular Basis of Neurodegeneration, a Keystone Symposium held 21-26 February 2011 in Taos, New Mexico, was no exception. Plasma, cerebrospinal fluid, imaging, and antibody markers were all being discussed. But lest anyone get too optimistic, there also were sobering thoughts about markers in general, not least being a reminder of how few have actually been approved for diagnostic use.

John Trojanowski, University of Pennsylvania, reviewed the biomarker data that has emerged from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). There is consensus among AD researchers that ADNI has been a success, and most researchers were happy to see that funding was found to continue the project with ADNI2 (see ARF related news story). One thing to have emerged from ADNI is the potential for finding markers for various stages of AD, from pre-symptomatic to mild cognitive impairment to AD itself, said Trojanowski. He stressed that big science projects such as ADNI, and more recently, the Parkinson’s Progression Markers Initiative (see ARF related news story) are crucial to moving the field forward, not least because they can address technical issues such as assay standardization and validation. Simply finding a potential marker is not enough, he stressed. A recent Nature review on the subject by George Poste, Arizona State University, Scottsdale, surveyed the literature to find some 150,000 papers documenting thousands of biomarkers, of which fewer than 100 are used in routine clinical practice (Poste, 2011). Trojanowski said that many of these biomarkers never make it to the clinic because methods are not standardized, which makes validation difficult. Leroy Hood, president of the Institute for Systems Biology, Seattle, Washington, echoed this view. While Hood sees great promise for medicine in the twenty-first century, including the ability to cut medical care costs, he said that the fundamental challenge for medicine is to deal with incredible complexity. When researchers start to deal with large datasets, noise, both biological and technological, becomes a huge concern, he said. He predicted, for example, that many of the hits that have come from large genomewide association studies will turn out to be noise.

David Holtzman, Washington University, St. Louis, Missouri, also emphasized the methodological challenges in searching for meaningful biomarkers. His lab has been conducting high-throughput, unbiased proteomic screens of cerebrospinal fluid (CSF) and plasma. In Taos, he stressed that reproducibility of CSF analysis can be problematic, and that replication of the analysis is essential. He described new potential markers that came from a CSF proteomic screen. One of them, YKL-40, has no known function but might be a marker of inflammation. Produced by astrocytes, it strongly correlates with tau in CSF, said Holtzman, and his group has validated the marker in hundreds of samples by ELISA. When combined with Aβ42, it predicts who will progress from normal to cognitively impaired over five years as well as does the tau/Aβ42 ratio (see Craig-Schapiro et al., 2010). A second marker, visinin-like protein-1 (VILIP-1), does even better. It, too, is elevated in the CSF of people with AD compared to normal controls or people with non-AD dementia. It better predicts conversion to cognitively impaired when combined with Aβ42 than does tau in their dataset. Holtzman said this was striking over a four-year period. VILIP-1, a cytosolic calcium-binding protein, is found in neurons. VILIP-1 CSF levels also correlate with those of tau. This led Holtzman to propose that tau is elevated in AD CSF, not because of the presence of neurofibrillary tangles that comprise tau, but because synaptic or neuritic degeneration releases proteins like tau and VILIP-1 into the parenchyma.

In response to a question from the audience about whether YKL-40 is a marker of general astrogliosis, Holtzman said that his lab looked at CSF samples taken from about 30 people with other neurodegenerative disorders, including frontotemporal dementia and progressive supranuclear palsy. In most samples, neither YKL-40 nor VILIP-1 was elevated. This could be because those disorders involve fewer neuritic changes than does AD, he suggested. Others suggested that not all glioses are the same, and that markers of the process may not necessarily be common among different disorders.

Several poster presentations from Michael Sierks’s group at Arizona State University in Tempe described a different type of CSF analysis. Sierks uses a combination of phage display and atomic force microscopy to generate and identify antibody fragments (nanobodies) that react with different oligomeric proteins. The field is still grappling with ways of identifying such oligomers in biological samples, and some of them could turn out to be early diagnostic markers. Sierks isolated several nanobodies to oligomers of amyloid-β (including A4) and α-synuclein (including D5 and 10H). The D5 nanobody bound to antigens in CSF of PD patients and AD patients, but not of normal controls or of people with multiple system atrophy, a synucleinopathy that is difficult to diagnose. The A4 nanobody recognized antigens specifically from the CSF of AD patients. Though the sample sizes were small, the technology may prove useful for identifying specific oligomeric species in human tissue. Eliezer Masliah told ARF that the technology is promising; Masliah is collaborating with Sierks to study α-synuclein oligomers. On a separate poster, Sharareh Emadi in Sierks’s lab showed that D5 and another nanobody, 10H, react with distinct forms of α-synuclein and also with different cell types. The D5 nanobody reacts with smaller antigens and with dopaminergic cells, while 10H reacted with larger oligomers and with both dopaminergic and cholinergic cells. The differential reactivity of the nanobodies suggests that specific forms of oligomeric α-synuclein may form in different cell types.

Moving from molecules to whole-brain imaging, Erik Musiek from the University of Pennsylvania, Philadelphia, described a comparison between fluorodeoxyglucose (FDG) PET and arterial spin labeling MRI as measures of brain activity. FDG-PET is a good tracker, said Musiek, but it is expensive and the radiation limits repeat measurement. Arterial spin labeling MRI is not invasive, is cheaper, and can be carried out repeatedly with little risk. Whereas FDG-PET measures brain activity based on glucose consumption, arterial spin labeling (ASL) does it by measuring blood perfusion. While data indicate ASL-MRI is altered in AD patients, no one had ever before done a head-to-head comparison with FDG-PET, said Musiek.

He described a small pilot study that did just that. Eighteen control volunteers and 16 AD patients were injected with FDG while undergoing ASL scans, and were then immediately taken for PET scans of glucose consumption. Musiek found that both techniques lit up comparable areas in the brains of controls. In AD brains, regions where FDG-PET showed deficits matched with those showing perfusion deficits by ASL. Quantitatively, AD patients had a 31 percent decrease in ASL compared to controls; that difference was 14 percent with FDG-PET. Voxel-based analysis showed good matching between areas that showed reductions in both types of analysis. Musiek concluded that this small pilot study suggests that ASL and FDG-PET perform with similar diagnostic accuracy. He said ADNI2 volunteers will undergo ASL, and analysis from them should help confirm if the technique has clinical potential.—Tom Fagan.


No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Research Brief: Announcing ADNI2—Funding for Five More Years
  2. PPMI: Parkinson's Field’s Answer to ADNI

Paper Citations

  1. . Bring on the biomarkers. Nature. 2011 Jan 13;469(7329):156-7. PubMed.
  2. . YKL-40: a novel prognostic fluid biomarker for preclinical Alzheimer's disease. Biol Psychiatry. 2010 Nov 15;68(10):903-12. PubMed.

Further Reading

No Available Further Reading