Noninvasive Quantitation of Early Biochemical Changes in Alzheimer's Disease

Ravinder Reddy, Ph.D.

University of Pennsylvania School of Medicine

Funded in June, 2004: $300000 for 5 years
LAY SUMMARY . ABSTRACT . HYPOTHESIS . FINDINGS . SELECTED PUBLICATIONS .

LAY SUMMARY

back to top

Developing an MRI Method for Indirectly Detecting Amyloid Early in Alzheimer’s Disease

A form of MRI imaging may provide a non-invasive indirect method to identify and quantify brain amyloid deposits in Alzheimer’s disease.  Investigators at the University of Pennsylvania will try to validate this approach using a “humanized” mouse model of the disease.  If successful, this non-invasive method could be easily applied to human diagnosis and assessment of experimental therapies designed to reduce amyloid plaques or prevent their further accumulation.  

First, investigators will see whether use of high resolution “T1r-weighted” MRI in a mouse model of Alzheimer’s disease provides direct evidence of amyloid accumulation.  To do this, they first will take brain images of the animals’ brains over time as amyloid deposits increase.  The investigators also will generate computerized maps of water in the brain that is displaced around the amyloid plaques.  The researchers hypothesize that the extent of water displacement in the brain will provide an accurate indirect measure of amyloid accumulation and changes in this accumulation over time.

The investigators then will see whether the high resolution images of amyloid plaques and the maps of water displacement indicating amyloid accumulation reflect the actual sites of amyloid that are found when the animals’ brains are autopsied. If the high resolution images and maps accurately reflect amyloid deposits, the researchers then will see whether a low resolution form of this T1r-weighted MRI, which can be used in humans, provides a valid indirect measure of amyloid.

To validate this low-resolution form of MRI, the investigators will use a “humanized” mouse model, in which a human gene has been inserted in mice of various ages. This age span of the animals corresponds to the increasing levels of amyloid deposits that occur in humans as they age. If the low resolution and high resolution imaging techniques generate similar maps of water displacement in the brain that are presumably produced by amyloid build-up, the low resolution technique will prove to be a valid, indirect method for imaging human amyloid and changes in its accumulation over time.

Investigators propose to further validate this technique to ensure that changes in maps of water displacement seen over time solely reflect the build-up of amyloid.  They will determine whether this is the case by comparing water displacement maps generated from the animal disease model to those generated by healthy mice.  Only diseased animals should show water displacement associated with build-up of amyloid.

Finally, the researchers will test the hypothesis that amyloid accumulation similarly alters cerebral blood flow patterns, by computing cerebral blood flow maps.  They will confirm at autopsy that altered cerebral blood flow is correlated with the presence of amyloid.  If investigators prove that both water displacement and alterations in cerebral blood flow are indirect measures of the extent of amyloid accumulation, this T1r-imaging technique would provide a way to non-invasively diagnose brain amyloid accumulation in humans.

Significance:  Currently, definitive confirmation of amyloid accumulation associated with Alzheimer’s disease is possible only at autopsy.  If this non-invasive MRI imaging technology can be validated as an accurate indirect measure of brain amyloid accumulation, including changes in amyloid build-up over time, the technology could be easily used in humans.  The technique also could be used to assess the effectiveness of experimental therapies in reducing amyloid early in the disease process, when they are most likely to be effective.

ABSTRACT

back to top

Noninvasive Quantitation of Early Biochemical Changes in Alzheimer's Disease

Alzheimer's disease (AD) is the most common form of dementia in the elderly. Classic symptoms of the disease include memory loss, confusion, and biological features such as the formation of neurofibrilary tangles, senile plaques (SP) of Amyloid-b (Ab) depositions and neuronal atrophy in the brain. Early detection of the disease provides greater understanding of the underlying pathobiology, contribute to presymptomatic diagnosis and, eventually, to the development of disease-modifying therapies for use in the presymptomatic stage.

This proposal deals with the development and optimization of novel magnetic resonance imaging (MRI) methods to detect and quantify early biochemical changes due to Alzheimer's disease in a transgenic (tg) mouse model. Specifically, T1r-weighted MRI will be optimized to obtain high-resolution images of normal brains in vivo and relaxation maps will be calculated from these images. These methods will then be used to match SP from T1r-weighted images of tg mice in vivo with immuno-stained histology sections of the brain. Then the MRI methodology will be used to measure changes in T1r relaxation time (from calculated T1r maps) in plaque-rich regions as determined from both high-resolution T1r-weighted images and histology. T1r maps will also be calculated in several age groups of animals to determine if there exist any plaque burden-dependent changes in T1r values. Finally, we will optimize a previously developed MRI method to map cerebral blood flow (CBF) distribution and determine any changes in flow resulting from the presence of SP.

The accomplishment of the proposed aims will result in a non-invasive MRI method for visualizing (via T1r-weighted imaging) and quantifying (via T1r and CBF mapping) the early degenerative changes in AD in the form of senile plaques. Since the methodology is completely non-invasive, it can be translated into human studies and implemented in a clinical setting.

HYPOTHESIS

back to top

Hypothesis:
Early biochemical changes in Alzheimer's disease (AD), which lead to the formation of senile plaques (SP) consisting of amyloid deposits, induce changes in the T1r relaxation times of water in the brain.  These changes can be visualized by high-resolution T1r-weighted magnetic resonance imaging (MRI) and can be quantified by in vivo T1r-relaxation maps. Further, changes in cerebral blood flow resulting from AD-related atrophy can be quantified using an optimized MRI-based method.

Goals:
AD is the most common form of dementia in the elderly. Early detection of the disease provides greater understanding of the underlying pathobiology, contribute to presymptomatic diagnosis and, eventually, to the development of disease-modifying therapies for use in the pre-symptomatic stage. Goals of this proposal are to develop sensitive MR imaging methods for visualizing and quantifying early changes due to AD. Since the MRI techniques being developed are noninvasive and MRI scanners are in widespread use, the methodology can be immediately be implemented on human studies.

Methods:
Mutations in the amyloid precursor protein (APP) and presenilin 1 (Ps1) genes are genetic markers of AD.  Elevated Ab protein levels have been observed in mutant APP/Ps1 transgenic (tg) mice line (tg2576) by 9-12 months in age.  The presence of macromolecules is expected to alter the T1r relaxation time of water in the brain.  We first optimize the T1r-imaging methodology in terms of resolution and imaging time, as applied to mouse model of AD.The optimized methodology will then be used to obtain high-resolution T1r-weighted images on a group of APP/Ps1 tg mice. Following the imaging experiments, the brains will be removed and immuno-stained to highlight the SP on histologic sections.  The high-resolution T1r-weighted images will be matched with corresponding histologic sections using specialized computer software to determine if SP are visible on MR images. 

Specifically, we will obtain both high-resolution T1rr-weighted images and low-resolution T1r-relaxation maps on different age groups of tg. Whereas MR images simply visualize plaques, maps provide quantitative T1r values at each pixel location.  Histological sections will be used to generate a 3D model of the brain from which "virtual sectioning" of 130µm thick-sections will be performed and then co-registered with the MR images by inspecting anatomical landmarks from a standard mouse brain atlas. Senile plaque density, from histology, will be correlated with T1r values from identical regions in the maps.  Analysis will also be performed on similar locations in the brain of healthy animals to account for normal age-related effects on T1r relaxation times.  This validation will enable us to use T1r-relaxation maps as a surrogate for directly quantifying plaque burden.  Further, we will compute cerebral blood flow (CBF) maps of both tg and healthy mice in vivo with a previously validated MRI method.  Changes in CBF will be also be correlated to T1r relaxation times due to the presence of SP.

 

FINDINGS

back to top

We have been able carry out the experiments as planned.  Accomplishments of the project are as follows:

  • Optimized T1r MRI pulses sequence, built an optimal radiofrequency coil, and obtained high-resolution T1r weighted images of mouse brain depicting Ab plaques.
  • Obtained series of T1r weighted images of mouse brain and demonstrated the feasibility of computing corresponding T1r relaxation maps in vivo.
  •  Investigated the effect of dipolar-coupling on T1r in the mouse brain and found that T1r imaging could be used to measure dipolar oscillations in the brain.
  • Found that T1r relaxation rates of transgenic APP/PS1 mice brains were higher compared to that of healthy mouse.
  • In a limited number of animals, we demonstrated that the T1r relaxation rate progressively increases with age and hence the plaque burden of transgenic AD mice.
  • In a limited number of animals, we found that perfusion rates decrease appreciably in the APP/PS1 mouse at an early stage of neurodegeneration.
  • Further studies on a larger group of animals, including histological correlation, are in progress.

2009 Scientific Findings: The project was completed successfully and resulted in three published journal articles and four abstracts to international conferences describing the techniques as well as results showing variations in T1r across age groups and differences in perfusion between tg and control mice.  A significant decrease in T1r was observed in the cortex and hippocampus of 12- and 18-month-old animals compared to their age-matched controls. There was also a correlation between changes in T1r and the age of the animal, with the older animals demonstrating lower T1r values in the cortex and hippocampus regions that exhibit significant AD pathology.  Perfusion MRI was performed on mice and an 18% decrease in blood perfusion was measured in the tg animals’ brains compared to controls. We extended the T1r MRI studies to human subjects with AD and obtained promising preliminary results. More studies are planned to elucidate the mechanisms of T1r MRI contrast and the physiologic basis of the blood flow changes observed.

2009 Summary: The successful completion of the project led us to discover that mice that possessed a genetic mutation to produce AD-like plaques have lower T1r values in regions of the brain associated with plaque deposits.  Further, they also exhibited lower blood perfusion in the same regions. It is also possible to visualize AD-like plaques using T1rweighted MRI. The results from this preliminary study allowed us to develop and optimize our image acquisition and data processing methods for T1r MRI and perfusion MRI for future studies on mice as well as in humans.

SELECTED PUBLICATIONS

back to top

Borthakur A., Gur T., Wheaton A.J., Corbo M., Trojanowski J.Q., Lee V.M., and Reddy R. In vivo measurement of plaque burden in a mouse model of Alzheimer's disease.  J Magn Reson Imaging. 2006 Nov;24(5):1011-7.

Borthakur A., Hulvershorn J., Gualtieri E., Wheaton A.J., Charagundla S., Elliott M.A., and Reddy R.  A pulse sequence for rapid in vivo spin-locked MRI.  J Magn Reson Imaging. 2006 Apr;23(4):591-6.

Wang C., Witschey W.R.T, Niyogi S., Johannessen W., Borthakur A., Elliott D.M., and Reddy R. Simultaneous acquisition of Human Brain T1r and T2  Maps, Human Brain Mapping, Florence (2006).

Borthakur A., Moonis G., Wheaton A.J., Melhem E.R., Clark C.M., and Reddy R., Quantifying T1r in Alzheimer’s Disease, Proc. Thirteenth Intl. Soc. Mag. Reson. Med., Miami, 2005, p. 1178.

Borthakur A., Wheaton A.J., Regatte R.R., Akella S.V., and Reddy R., Effect of dipolar coupling on the T1r MRI signal in the mouse brain. Proc. Int. Soc. Magn. Reson. Medicine, Kyoto, 1245 (2004).

Borthakur A., Uryu K., Corbo M., Wheaton A.J., Melhem E.R., Trojanowski J.Q., Lee V.M., and Reddy R.  Visualizing amyloidal deposition in a transgenic mouse by MRI.  Proc. Int. Soc. Magn. Reson. Medicine, Kyoto, 1433 (2004).