Background: Alzheimer's disease (AD) is characterized by two complementary brain lesions: amyloid plaques and neurofibrillary tangles. Amyloid plaques occur up to 20years before the first clinical signs of the disease and their detection is thus critical for an early diagnostic and to follow-up potential treatments. We developed methods of ex-vivo or in-vivo amyloid plaques detection based on the use of a non-targeted gadolinium (Gd) contrast agent for magnetic resonance imaging (MRI). Even if numerous mouse models of amyloidosis have been developed, the age of appearance, the size and the composition of the plaques in these models are different. Here, we compared the MRI detection of amyloid plaques by in-vivo and ex-vivo Gd-staining in different strains of mouse model of amyloidosis.
      Methods: Five strains of mice developing amyloid plaques before 10months were used: APP sl/PS1 M146L, APP/PS1dE9, B6.PS2APP, APP-SDI and 3xTg (APP swe/PS1 M146VKI/Tau P301L) aged of 14-15months (n=2/strain). Mice received an intracerebroventricular (ICV) injection of Gd (500mM, 1μl/side) and were imaged by in-vivo MRI one hour later. They were then sacrificed and their brains were extracted and incubated in a Gd solution (2.5mM - 48h) before ex-vivo high-resolution MRI.
      Results: Following in-vivo ICV infusion of Gd, MR images show numerous hypointense spots (upper pictures) which were previously demonstrated to be amyloid plaques (see Petiet et al., 2012 for examples). The number, size and contrast of the hypointense spots were highly variable in the different strains. This inter-strain difference was confirmed by the high signal/noise ratio ex-vivo Gd-stained MR images (lower figures). Plaques were more visible in APP sl/PS1 M146L > B6.PS2APP > APP/PS1dE9 > APP-SDI > 3xTg mice and only few plaques appeared in the two latter strains.
      Conclusions: These results demonstrate that depending on the strain, amyloid plaques display highly different aspects in MRI. These differences appear to be mainly due to the size of amyloid plaques. The contrast/noise ratio of the plaques on MR images is also critical for the detection of the plaques. This parameter can be modulated by the composition of the plaques, for example by their iron concentration or the Aβ40/42 ratio that can modulate their hydrophobicity and interaction with the contrast agent.