250Following Abeta fibrillization/aggregation in real-time using a FLUOstar Omega microplate reader

Frank Baumann, Hertie Institute for Clinical Brain Research, 12/2014
  • Long-term shaking capability of the microplate reader used to monitor Aß aggregation
  • Lag times derived from signal curves proved to be a useful measure for the fibrillization process


Aggregation of the amyloid-ß (Aß) peptide is a fundamental hall-mark for Alzheimer’s disease. The formation of extracellular senile plaques will lead to synaptic and neuronal damages in clinical demented patients. The aggregation process of Aß peptide is seen as seed driven. These seeds consist of small stable aggegrates of Aß. It is thought that these aggregates are already present in early stages of Alzheimer’s even before a patient experiences any symptoms. If this is true, determination of these early aggregates (aggregation seeds) would be an excellent diagnostic tool.

Here we present a cell-free assay (FRANK-Assay = Fibrilization of recombinant Aß nucleation kinetics) that allows determination of the amount of aggregation seeds from brain tissue homogenates. The assay is run over 2-3 days using the FLUOstar® Omega microplate reader from BMG LABTECH.

Assay Principle

The assay uses Thioflavin T to follow the amyloid formation (Fig. 1). Thioflavin T is a benzothiazole salt that is known to show increased fluorescence when bound to beta sheet-rich structures, such as in amyloid fibrils of Aß.

Fig. 1: Fibrillization process followed over time.Before aggregation or fibrillization can start a critical amount of initial aggregation seeds need to be present or spontaneously formed. This is a thermodynamically unfavoured process even in the presence of excess monomeric Aß, kinetically slow and results in a delay in time before measurable aggregation starts. Once enough seeds have formed a massive and steep increase of fibrillization can be monitored by following the increase of ThT fluorescence due to incorporation into newly  formed fibrils. After some time a plateau is reached indicating the end of the reaction. The delay time is significant and can be shortened by exogenous addition of aggregation seeds. These seeds accelerate increase in fluorescence in relation to their amount.

Materials & Methods

  • FLUOstar Omega, BMG LABTECH, Germany

  • black, clear bottom 96-well plates, Greiner

  • Sealing films, non-sterile, Excel Scientific

All other chemicals were purchased through normal distribution channels.

Donor brain tissue and extract preparation
APP23 transgenic (tg) mice were used as seed donors for all studies. This transgenic mouse model is characterized by an overexpression of mutant human  APP. Transgenic APP (amyloid precurser protein) is subject to proteolytic cleavage and gives rise to ß-amyloid peptide which in turn is aggregation prone and results in amyloid plaques consisting of ß-amyloid deposits.

Donor brain tissues were obtained from APP23 tg mice as well as from non tg wild type (WT) mice. After removal, the brain was divided into hemispheres. One hemisphere was immediately fresh-frozen (but not fixed) in dry ice while the other hemisphere was immersion-fixed in formaldehyde solution. After finishing the fixation process this hemisphere was frozen on dry ice as well and stored at -80°C until use (Fig. 2).Fig. 2: Method to obtain comparable fixed and fresh-frozen brain samples from either APP23 mice or wild type mice.Prior to use fixed and fresh-frozen tissues were homogenized, centrifuged and the supernatant aliquoted. For all experiments, a 10 % (w/v) extract was used.

Thioflavin T aggregation assay
1 μl of brain extract, protease inhibitor cocktail (Complete, Roche), 20 μM Thioflavin T, 25 μM Aß1-40 were incubated in aggregation buffer (50 mM phosphate and 150 mM NaCl) at 37°C. Each brain homogenate was present in 8 replicate wells. The fluorescence increase was measured every 30 min from the bottom of the well. Before each measurement, the microplate was shaken (double-orbital mode) for 30 sec at 500 rpm. The progress of the measurement was followed in the current state window and the reaction was stopped after all samples showed a plateau. Remote control (TeamViewer) was used to monitor and eventually stop the measurement during the weekends. Raw data were fitted and lag times were determined with either using BMG LABTECH’s MARS Data Analysis Software or GraphPad Prism.

Instrument Settings


Detection Mode: Fluorescence Intensity, Plate Mode
Optic: Bottom optic
Orbital Averaging: Activated, 4 mm diameter
Filters: 440-10/480-10
Number of cycles: 90
Cycle time: 1800 sec
No. of flashes: 10
Temperature: 37°C

Results & Discussion

A typical result for signal curves over time is shown in figure 3:

Fig. 3: Signal curves for samples containing either fixed or fresh-frozen wild type or APP23 brain homogenates. Error bars represent deviation of replicate wells within one plate from mean.

All signal curves show a clear increase in fluorescence after a certain time. This increase illustrates the incorporation of Thioflavin T into the newly formed Ab fibrils. After some time a plateau is reached that is considered as endpoint of amyloid formation and Thioflavin T incorporation process.

The time until the signal starts to increase is the lag time. The MARS data analysis software offers the possibility to create 4-parameter fits of the signal curves from which the lag times are calculated (lag times correspond to the EC20 value of the fit). Initial fibril seeds are formed until the lag time is reached. Considering this, the lag time can be used as a measure to compare different brain homogenates (Fig. 4).

Fig. 4: Lag times created from signal curves shown in figure 3. Fixed and fresh frozen tg and WT mice are compared. As a control thioflavin T only (ThT) was measured on the same microplate (n=1). Error bars refer to 3 biological replicates.From figure 4 it can be followed that the lag times of the wild type are bigger compared to the lag times obtained for the tg mice. Further, a difference can be seen between fixed and fresh-frozen APP23 samples. As expected the fresh samples induce Ab deposition much faster. Nonetheless, the fixed APP23 samples show compared to the WT a significantly lower lag time indicating that fixation in formaldehyde is not sufficient to prevent Aß aggregation.


With the help of the FLUOstar Omega microplate reader, it is possible to prove that the in vitro assay is reliable to detect seeding activity in brain samples. In addition, it allows quantitative comparison of seeding activity which with only little effort can be statistically validated. The method is also open for bigger sample cohorts which would easily burst the limited capacity of the in vivo approach. Since data can be obtained much easier (no animal surgery and animal permission needed) and earlier (2-3 days vs. 4-6 months) compared to the in vivo assay it is meant to replace the latter for qualitative and quantitative seed determination measurements in the future.

Newsletter Sign-up