The background issue was clear.

My point was: why is ThT not abitually used in cells or tissue? Considering that the background in that case is lower, ThT should be better than ThS. LeVine explains the differences in term of fluorescence emission and charge between the two fluorophores in the paper, but I don't understand why ThS is preferable in cells or tissues. Is that a matter of charge? Or capability of passing cellular membranes?

It seems to me that these uses of ThT and ThS are more due to historical reasons than practical ones.

In this paper, the author compared ThS with congo red and found that they correlate. Other researchers must have read that and used ThS.

Acta Neuropathol. 1967 Feb 3;7(4):336-48.

Thioflavin S fluorescent and Congo red anisotropic stainings in the histologic demonstration of amyloid.

Kelényi G.

Summary

Two methods employed for the histological detection of amyloid, the recently developed Thioflavin S fluorescent microscopic procedure and the Congo red anisotropic staining were compared. It was observed that in senile alterations of the brain, heart, and pancreas and in secondary amyloidosis both methods demonstrate the same structures. Dichroism of Congo red stained structures was also studied and it was found that not only renal amyloid and cerebral senile plaques, but Alzheimer cells are dichroic, too. The specificity of the two methods is discussed

My understanding also is that the reasons for the different uses of ThT vs ThS are historic. Neither molecule is particularly specific for amyloid: they bind to cross-beta-sheet structures, which are a defining feature of amyloid fibrils, but also exist elsewhere. However, both ThT and ThS also bind to lipid membranes, as mentioned above by Luis Oliveira. Both molecules to have high background in cells.

I have also seen (though not yet published) that ThT (so, presumably, also ThS) does not accurately report on the amount of aggregation in vitro when anionic phospholipids are present. In other words, ThT reports more aggregation than I measure by mass balance, for example, but only in the presence of lipids. So the background / artifact issue with these dyes is not specific to the cellular environment.

It is interesting that the different use of the two molecules is probably due to historical reasons. This explanation could be also confirmed by the fact that none of the papers that I read about this issue mentioned an actual motivation for the choice of ThS in cells and tissue. Thank you everybody for the fruitful discussion!

Thioflavin T, in addition to beta-sheets, seems to stain RNA and at least some DNA, therefore the histological staining with it would be less specific.

Those of you who do use ThioS, which manufacturer do you prefer? There is a great variation of the color of the powder from Acros, Sigma and Santa Cruz. They are not quite as good as the old supplier Roboz!

Late to the discussion but see "

Methods Enzymol. 1999;309:274-84.

Quantification of beta-sheet amyloid fibril structures with thioflavin T.

LeVine, H 3rd." who says"Although commonly used for the histological demonstration of amyloid fibrils where long wavelength UV excitation can be used and the free dye can be washed away, commercial preparations of ThS are complex mixtures of molecules. Binding to amyloid fibrils enhances the emission intensity of ThS severalfold with no change in the excitation or emission spectra, resulting in a high background fluorescence in solution, which makes this dye unsuitable for quantitative analysis in solution"

Dunno but I am about to find out! Someone above suggested that ThioT undergoes an excitation/emission wavelength shift on binding, whereas ThioS does not. That might make ThioT more discriminatory.