By Martin Ingelsson, co-organizer of the meeting and one of the speakers.
This is Part 2 of a two-part series. See Part 1.
4 August 2009. Biomarkers were top of the agenda on the second day of Neurodegenerative Disorders—Immunotherapy and Biomarkers, an international conference held 28-29 May at Uppsala, Sweden. Henrik Zetterberg, from Gothenburg University, Sweden, started out by providing an update on the use of cerebrospinal fluid (CSF) for diagnostic purposes. It has been known for a rather long time that tau levels in CSF increase and Aβ42 levels decrease in Alzheimer disease (AD). Zetterberg and his colleagues have now also analyzed differently truncated versions of Aβ and found that a majority of them decrease but that some actually increase in diseased subjects. Among all β markers, Aβ1-42 indeed turned out to have the best disease specificity, but a combined measure of Aβ1-16, Aβ1-33, Aβ1-39, and Aβ1-42 improved both sensitivity and specificity somewhat as compared to Aβ42 alone. Interestingly, Aβ1-39, together with certain other Aβ species, was particularly low in early-onset familial cases, pointing to their potential role in early disease stages. Moreover, certain Aβ forms such as Aβ1-14, Aβ1-15, and Aβ1-16 were found to increase with γ-secretase inhibition, indicating their usefulness for monitoring therapeutic intervention. Finally, Zetterberg and his team have also detected fragments of the amyloid precursor protein (APP) in CSF, some of which reach outside the Aβ domain and span the cleavage site of β-secretase. Ongoing studies aim at investigating their diagnostic relevance for AD.
The next speaker, Frida Ekholm Pettersson from Uppsala University, Sweden, has focused on Aβ protofibrils, because observations indicated that such soluble oligomeric Aβ forms correlate well with cognitive performance in transgenic mice. It was hypothesized that Aβ protofibrils may also be increased in Alzheimer patients. By analyzing CSF from 40 patients and controls, Ekholm Pettersson and Hillevi Englund, also at Uppsala University, found that soluble Aβ oligomers were significantly increased in both Alzheimer disease and MCI. The levels of Aβ oligomers were determined by calculating the ratio between Aβ42 as measured by denaturing Western blot and by non-denaturing ELISA (see Englund et al., 2009). The Uppsala group previously showed that Aβ42 levels are underestimated when measured by non-denaturing ELISA as compared to when measured under the denaturing conditions in Western blot (see Stenh et al., 2005). In fact, the typical observation of decreased Aβ42 in Alzheimer CSF may be due to masking of Aβ epitopes caused by oligomerization. Their ongoing work aims to investigate the presence of Aβ protofibrils in CSF with a more direct approach, i.e., by using Aβ oligomer-specific antibodies.
Agneta Nordberg from the Karolinska Institute, Stockholm, Sweden, has pioneered the clinical use of Aβ-specific ligands for imaging. The Pittsburgh compound B (PIB) was developed by William Klunk and colleagues and PET-PIB holds great promise as a biomarker to discriminate Alzheimer cases from healthy controls.
Nordberg, Anton Forsberg, and other colleagues at Karolinska Institute are investigating whether PET-PIB can be used to identify patients with mild cognitive impairment (MCI) who will eventually progress to Alzheimer disease. In a recent study, 21 MCI subjects were included. At a four-year follow-up examination, seven of the 11 cases that were PET-PIB-positive at baseline had converted to AD, whereas all of the 10 PIB-negative cases remained as MCI. Moreover, in a recent five-year follow-up of the same subjects, the mean PIB retention was unchanged overall, although certain patients displayed either an increased or a decreased amyloid load. Finally, occasional healthy subjects with high amyloid retention have now been followed for up to five years with no obvious decline in their cognitive abilities.
Nordberg has also evaluated the use of PET-PIB as a biomarker to monitor clinical trials. Together with Ahmadul Kadir and other colleagues at the Karolinska Institute, she found that patients who received phenserine, a weak acetylcholine esterase inhibitor, displayed decreased amyloid retention in the brain along with an increase in Aβ40 in CSF after six months of treatment, as well as an improvement in attention after 12 months.
Finally, Nordberg reported that more than 3,000 subjects worldwide now have undergone PET-PIB scanning. Of these, 228 have been examined within DiMI, the European Diagnostic Molecular Imaging initiative, which is conducting a multicenter study. Of the 52 controls included, only five were PIB-positive, which is in stark contrast to the experience from ADNI (the large American multicenter study), where approximately 50 percent of the controls turned out to be PIB-positive.
Similar to Alzheimer disease investigators, researchers are trying to develop novel diagnostics for Parkinson disease and other Lewy body disorders. Brit Mollenhauer from Paracelsus-Elena-Klinik in Kassel, Germany, has, together with Omar El-Agnaf, United Arab Emirates University, Al Ain, and Michael Schlossmacher, Ontario Health Research Institute, Ottawa, Canada, developed an ELISA that specifically recognizes oligomers consisting of two or more α-synuclein molecules. The research team found that patients with Parkinson disease, dementia with Lewy bodies, and multiple system atrophy have significantly lower levels of such multimeric α-synuclein in CSF (see Mollenhauer et al., 2008). However, they also found a considerable overlap between diseased and non-diseased subjects, and other studies have failed to find any differences in plasma α-synuclein between patients and controls (see Noguchi-Shinohara et al., 2009).
When the research team used their ELISA on plasma samples, they somewhat surprisingly found higher α-synuclein levels as compared to CSF. However, more studies are needed to elucidate whether plasma levels of α-synuclein are altered in patients as compared to healthy controls. Large ongoing studies in the U.S. (Harvard PD biomarker discovery study) and Europe (diagnostic study at Kassel) will investigate further the potential of finding novel biomarkers, based on α-synuclein and/or other disease-related molecules in plasma and CSF.
To date, most diagnostic biomarkers for neurodegenerative diseases have been discovered as a result of their neuropathological significance and are analyzed with traditional methods, such as ELISA. However, thanks to a rapid development in proteomics, novel analytical tools may soon reveal additional markers. Jonas Bergkvist from Uppsala University studies the use of multiplex quantitative mass spectrometry, and, in a collaborative effort with researchers in Switzerland and the U.S., his group is establishing a blueprint list of all molecular species that can be identified in human CSF. One molecule of interest is vascular endothelium derived growth factor (VEGF). Early observations indicate that some of the 19 detected isoforms of VEGF present may be decreased in CSF from AD and amyotrophic lateral sclerosis (ALS) patients. In other ongoing projects, Bergkvist and colleagues are performing matrix-assisted laser desorption/ionization (MALDI) imaging on brain sections as well as laser capture microscopy-based analyses of affected cells in ALS and other neurodegenerative disorders.
In the final talk of the conference, Masood Kamali-Moghaddam from Uppsala University presented other novel molecular tools for protein analyses. This research group, headed by Ulf Landegren, developed the proximity ligation assay. This technique, based on the simultaneous recognition of two or more neighboring epitopes with DNA-conjugated antibodies followed by enzymatic ligation and PCR, allows analyses of minute amounts of analyte with high specificity and sensitivity. With this technique, the research group recently demonstrated increased plasma levels of GAD65 as a biomarker for diabetes mellitus. For neurodegenerative disorders, they have developed an assay for the detection of Aβ oligomers and are currently applying the technique for analyses of both CSF and plasma from Alzheimer patients.
This is Part 2 of a two-part series. See Part 1.