The varied morphological and biochemical forms in which Aβ deposits in brain of Alzheimer's disease (AD) patients are complex and the mechanisms that drive their formation are not yet properly understood. Aβ is predominantly deposited in brain as dense-core and diffuse plaques and in cerebral vascular walls. The dense-core plaques and vascular amyloid, but not diffuse plaques, are associated with tau pathological changes in the surrounding brain tissue, suggesting that such amyloid deposits are toxic, although structural changes associated with diffuse plaques have not yet been studied in detail. Mutations in amyloid precursor protein or presenilin genes cause an increased production of Aβ42 compared to the more physiological and soluble Aβ40 isoform. However, in the majority of AD patients, no known cause is identified and is postulated to be due to failure of Aβ clearance or other mechanisms. Nevertheless, mutations identified in familial AD have allowed development of a number of mouse models that mimic most of the pathological features of AD and have been highly instrumental in furthering our understanding of different aspects of AD neuropathology. These mouse models are also being extensively utilized for testing drugs, in what we call as preclinical mouse trials. These mice are eventually evaluated for drug efficacy based on reversion of behavioural and cognitive abilities, however, these methods are cumbersome and time-consuming and not useful in the high-throughput first screens. For drugs that target Aβ, reduction of plaque load is a good indicator of drug efficacy, but an increasing number of drug candidates do not target Aβ directly. Thus, biomarkers including those based on histopathology, as AD is definitively diagnosed on neuropathology, are urgently needed in the field.
As a part of an international consortium, we had earlier identified an important association of a subset of amyloid plaques (namely, dense core plaques) with vessel walls (Kumar-Singh et al., 2005). Importantly, structural microvascular defects were also identified not only in amyloidogenic vessels but also in non-amyloidogenic vessels, suggesting a key role of vessels in the etiopathogenesis of AD (Kumar-Singh et al., 2005). However, structural microvascular defects were most robustly identified only on electron microscopy, a technique that is cumbersome and time-consuming, and therefore again not suitable for routine analysis in mouse preclinical trials. To test the feasibility of a faster approach, we employed tissue fractal dimension image analysis on two AD mouse models and compared them with AD patients and aged controls. We showed that the various types of plaques present in humans and transgenic mice have comparable fractal dimensions that also accurately differentiated various types of plaques in both species. These data not only suggest that mouse models reproduce plaque pathology present in AD, but also that image analysis could be a valuable tool for objective, computer-oriented diagnosis of AD-associated changes. The project aims to extend these studies in larger series and also intends to quantify parenchymal and especially vascular parameters. The identified image analysis parameters would also be evaluated for reversion to normal values in mice treated with Aβ-clearing and vasoactive compounds. A successful outcome of this project promises to deliver early amyloidotic and non-amyloid markers that will not only provide diagnostic tissue image parameters for high-throughput drug screening in mouse models, but will also give important insights into AD etiopathogenesis.