. Association of plasma amyloid beta with risk of dementia: the prospective Three-City Study. Neurology. 2009 Sep 15;73(11):847-53. PubMed.


Please login to recommend the paper.


  1. The paper is well written and the study seems to be carried out very well, at least in regard to the data analysis. (Though an experimental scientist that I am, I was a bit surprised to see only half of the name of one of the used kits—the INNO-BIA kit—without further information on the experimental procedures.) In addition, the study is one of the biggest of its kind, and similar to the Rotterdam Study, although with different methodology.

    Previously, different studies have reported increased Aβ1-40 and decreased Aβ1-42, and also decreased Aβ1-42/Aβ1-40 ratios in CSF, but there are conflicting data on the level of these Aβ species and the ratio of these species in blood.

    Another striking thing is the decreased level of Aβ1-42 with progression of dementia (that has not been commented on in the paper), which is usually speculated to be due to absorption to the senile plaques in the brain. But some recent studies have demonstrated that this might depend on the methodology used. In an ELISA-based assay, the Aβ1-42 level is decreased, while in a denaturing Western blot experiment, the level of Aβ1-42 is increased in the same model material.

    The main part of the results in this paper conforms with the data from the Rotterdam Study, with some differences in the details, for instance, finding that the Aβ1-42/Aβ1-40 ratio in this study "may be associated with the risk of dementia only in individuals diagnosed at two years of follow-up," while in the case of the Rotterdam Study this time period is eight years. The current study has also included the measurement of other Aβ pieces (Aβn-40 and Aβn-42), which might shed light on some details.

    Overall, my belief is that this paper, by confirming other well-done studies such as the Rotterdam Study, takes us one step closer to be convinced that blood Aβ might reflect brain and CSF Aβ. This, in turn, might take us to another level to try harder to detect and identify other toxic isomers of Aβ, such as Aβ oligomers in blood, which is a more non-invasive approach compared to the CSF. Of course, this might require that more powerful tools be developed and applied in the field. Having said that, I do not believe that the outcome of this paper will change the current situation and won't clarify much whether blood Aβ is a promising AD biomarker.

    View all comments by Masood Kamali-Moghaddam
  2. First, we would like to thank Dr Kamali-Moghaddam for his comments. Regarding the methodology, we apologize about the name of the kit used which is not fully written (INNO-BIA plasma Aβ forms). The strength of such a study is the use of a commercially available kit. Thus, experimental procedures are the manufacturer’s instructions. We did not think that it was necessary to add them in the materials and methods. However, for Alzforum readers, here is the experimental procedure. All plasma samples were diluted 1/3 with a buffer containing detergent before analysis.

    Quantification of Aβ isoforms in plasma was performed using INNO-BIA plasma Aβ forms assays (Innogenetics, Ghent, Belgium), a multiplex microsphere-based Luminex xMAP technique that allows simultaneous analysis of Aβ1-40 and Aβ1-42 (module A) an Aβn-40 and Aβn-42 (module B).

    The monoclonal antibodies (MAbs) 21F12 and 2G3, which specifically bind Aβ peptides ending at 42 and 40, respectively, were used as capture antibodies. The capture MAbs were covalently coupled to carboxylated beads of different regions (region 104 for Aβ42;, region 105 for Aβ40;, and region 102 for a non-Aβ binding MAb). MAb 3D6, which specifically binds Aβ peptides starting at Asp1, and MAb 4G8 which react with all N-terminally truncated Aβ peptides up to Val18, were used as detector antibodies.

    In short, Aβ isoforms ending either at Aβ40 or Aβ42 were selectively captured by beads that have been coated with either MAb 21F12 for Aβ42 or MAb 2G3 for Aβ40. A third class of beads was coated with MAb AT120, used to measure matrix effects due to heterophilic antibodies in the plasma sample. In module A, biotinylated MAb 3D6, which selectively binds Aβ peptides starting at Aβ1, is used as detector antibody, providing specific quantification of Aβ1-42 and Aβ1-40 isoforms. In module B, biotinylated MAb 4G8, which binds all N-terminally truncated Aβ peptides up to those starting at Aβ18, is used as detector antibody, providing specific quantification of Aβn-42 and Aβn-40 isoforms.

    The procedure could be described as follows: after sonication and vortexing, a mixture (100 μL/well) of the beads (either Module A or B) were added to 96 well filter plates (Millipore Corporation, Bedford, Massachusetts). After draining the wells using a vacuum manifold (Millipore Corporation, Bedford, Massachusetts), standards, blanks, or diluted (1:3) plasma samples were added (75 μL/well) in duplicate together with the biotinylated detector MAb (25 μL/well) (MAb 3D6 for Module A and MAb 2G3 for Module B) and incubated overnight at 2-8ºC in the dark (plates covered with aluminum foil) on a plate shaker (600 rpm). After washing, phycoerythrine-labeled streptavidine was added (100 μL/well) and incubated for one hour on a plate shaker. After a second wash step, 100 μL of phosphate-buffered saline was added. The assays were analyzed on a Luminex 200 IS instrument (Luminex, Austin, Texas). For each set of microspheres, 100 beads were analyzed, and the median fluorescence intensity (MFI) was used for quantification.

    A ready-to-use calibrator series is included in duplicate in each assay run and, using the calibration curve constructed with the median fluorescence values for each of the standards, concentrations were determined by sigmoidal curve fitting. Moreover, in each series two run-validation control samples are also included.

    I hope that this is helpful.

    View all comments by Luc Buee

Make a Comment

To make a comment you must login or register.