Minhua Qiu — 2010-11 Fellow
Alzheimer’s disease (AD) is a progressive and fatal brain disease that is characterized by slow irreversible loss of memory and cognitive skills and massive death of neurons. According to the 2010 statistics from the Alzheimer’s Association, currently as many as 5.3 million Americans are affected by AD . Because the risk of developing AD doubles every five years beyond age 65  and the U.S population is aging, the number of AD patients will increase significantly unless effective prevention and treatment are provided.
AD is featured by two pathological hallmarks related to the deposition of filamentous tau and β amyloid (Aβ): accumulation of tau forms neurofibrillary tangles while the deposition of Aβ forms neural plaques as observed in diseased brain tissue. Pathogenic tau and Aβ both induce defects in axonal transport , a basic process used by neuronal cells for the translocation and delivery of various essential cargoes (vesicles, organelles, and macromolecular protein complexes, etc.) along axons. Movement of cargoes is driven by molecular motor protein kinesin and dynein. As axonal transport is critical to the survival and function of neurons, defects of this process may be the main cause of neuronal cell death in AD. Quantitative characterization and analysis of axonal transport defects is critical to understanding the underlying molecular mechanisms of AD and development of treatments for AD patients.
In this study I propose to develop a novel nanometer resolution in vivo imaging and computational analysis assay to quantitatively and automatically evaluate axonal cargo transport in segmental nerves of Drosophila, and to carry out experimental and computational analysis to characterize and understand transport defects in Drosophila models of AD .
Previous work in our lab has established that the segmental nerves in Drosophila larva is a powerful model system of axonal transport  because of its simple geometry and the wide variety of Drosophila genetics tools. Time-lapse videos are collected for axonal transport of yellow fluorescent protein (YFP) tagged amyloid precursor protein (APP), whose degradation produces Aβ . These videos visualize axonal cargo transport. However, quantitative analysis of these videos poses substantial challenges on related image processing and investigation of underlying molecular mechanisms. Here I propose to overcome these challenges by achieving the following specific aims.
Aim 1: To automatically detect individual axonal cargoes with nanometer resolution and to fully recover their trajectories.
We will design software to achieve: (1) Nanometer resolution detection. Because of diffraction effect, a fluorescence tagged particle is observed as an Airy disk spanning over several pixels. When such particles approach to each other under imaging, their intensity profile is deformed and therefore difficult to be resolved using traditional methods. The detection part of this software is capable of partially overcoming the diffraction barrier and locate individual fluorescence particles with nanometer resolution using either a Gaussian correlation algorithm or a multi-kernel fitting algorithm; (2) Full trajectories generation. For visualization, the complex motion of axonal transport is projected onto a two-dimension kymograph, which visualizes crossing and overlapping vesicle trajectories for tracking. We propose to integrate statistic tools and computer vision techniques into this software package to achieve unambiguous and complete trajectory recovery. (3) Minimal human input. Manually tracking and connecting trajectories is infeasible given the complex traffic along axons. Furthermore, to fully understand the behavior of axonal transport, we image totally four regions along one segmental nerve of Drosophila larva by 150 frames per region and repeat the procedure on hundreds of larvae for both control and mutant case. Automated analysis is essential for analyzing such a large volume of image data.
Aim 2: To characterize axonal transport and to extract internal information of the molecular motor machinery by using a hidden Markov model.
The recovered trajectories of fluorescence particles will be used to characterize axonal transport by: direction (anterograde, retrograde or stationary), velocity, pause duration, run length, etc. Furthermore, we will also extract information regarding the lateral motion of cargoes and the underlying states of axonal transport within experimental trajectories. To characterize the lateral displacement, I plan to retrieve the axon’s location by linear least-square fitting of the path of fluorescence particles and compare each particle with its projection on axon. To recover the potential states of axonal transport, we will regard the particle trajectories as the outcome of a multi-state hidden Markov model (HMM) and derive model parameters by learning algorithm. The analysis tool and modeling method proposed here will pave the way for characterizing axonal transport defects in Drosophila mutant models of AD.
Aim 3: To identify causes of transport defects in Drosophila models of AD by pattern matching with existing library of transport defects of known causes.
Previous work in our lab has generated a comprehensive library of transport defects under known genetic mutations of different parts of the molecular motor machinery of axonal transport. Using statistical pattern recognition and machine learning techniques, I propose to match measurements of transport defects in Drosophila models of AD with those in the library to identify at the molecular level which part of the molecular machinery is affected. We will start the comparison and match by analyzing axonal transport in tau mutant. Drosophila stocks for control are maintained in the lab and will be crossed with the mutants. Crossed Drosophila’s larvae will be cropped at their 3rd stage and dissected to expose their segmental nerves whose images will be collected and analyzed by our developed software and modeling tools.
Successful completeness of this project will elucidate the molecular mechanisms of axonal transport defects in terms of motor function failure and may ultimately suggest a hopeful direction for diagnosis of AD at its early stage and for treatment. However, the application of this research will not be restricted to AD, since many other neurodegenerative diseases are also closely associated with axonal transport defects. Examples of these neurodegenerative diseases include Huntington’s disease, the early stage of amyotrophic lateral sclerosis, hereditary spastic paraplegias and Parkinson’s disease . Molecular mechanisms associated with the impairment of axonal transport in these cases are still elusive. The automated analysis tools and experimental procedures developed in this project will benefit the study of these diseases.
 Alzheimer’s Association, “2010 Alzheimer’s Disease Facts and Figures,” Alzheimer’s & Dementia.
 G.A. Morfini, M. Burns, L.I. Binder, N.M. Kanaan, et al., “Axonal Transport Defects in Neurodegenerative Diseases,” Journal of Neuroscience, 29(41): 12776-12786, 2009.
 A.M. Celotto and M.J. Palladino, “Drosophila: A “Model” Model System to Study Neurodegeneration,” Molecular Interventions, 5(5): 292-303, 2005.
 G. Yang, G. Reis, L. Szpankowski, G. Danuser, L.S.B. Goldstein, “Genetic Evidence for a Coordination-Competition Mechanism between Kinesin and Dynein in Axonal Transport of APP vesicles,” Submitted.
 P.R. Turner, K. O’Connor, W.P. Tate, W.C. Abraham, “Roles of Amyloid Precursor Protein and Its Fragments in Regulating Neural Activity, Plasticity and Memory,” Progress in Neurobiology, 70(1): 1-32, 2003.