Tracking down fusion events

CSIRO and Sydney’s Garvan Institute have signed a three-year collaboration agreement to investigate cellular processes such as vesicle-membrane fusion events using a variety of microscopy and image analysis techniques.

The collaboration has already borne fruit, with the partners developing a new computer vision system to quantify vesicle fusion in pancreatic beta cells, using total internal reflection fluorescence microscopy (TIRFM).

TIRFM is a commonly used technique to study molecular events occurring close to the plasma membrane. It allows researchers to observe specific fluorophores in a limited region of the specimen without background fluorescence overwhelming them.

In biological processes such as vesicle-membrane fusion events, where hundreds of events occur at a time, capturing the events is not the problem – quantifying it is. Last year, the CSIRO/Garvan team announced they had developed software to track the objects in question and designed an algorithm to filter out false positives and false negatives.

The team – Dr Pascal Valloton, head of biotech imaging at CSIRO Mathematical and Information Sciences; Professor David James, director of the Garvan’s diabetes and obesity research program; and Dr Will Hughes, a group leader in the program and director of the Garvan’s molecular imaging facility – has developed the system to automatically identify the transport of green fluorescent protein-tagged phogrin vesicles and their recruitment to the plasma membrane in pancreatic beta cells. Phogrins are membrane proteins that localise to insulin-containing secretory granules in beta cells and are an autoantigen in type 1 diabetes.

“GFP-phogrin has been widely used to monitor vesicular dynamics in pancreatic beta cells,” the researchers wrote in a conference paper last year, titled ‘Towards fully automated identification of vesicle-membrane fusion events in TIRF microscopy’. “Upon glucose stimulation of pancreatic beta cells expressing GFP-phogrin, puffs of fluorescence can be observed under TIRFM, indicating fusogen activity.”

The problem is, however, that typical TIRFM movies contain thousands of image frames, in which hundreds of vesicles randomly diffuse, dock, undock and fuse, the researchers say. “The attention of the visual system is attracted to all of this activity. As a result, it is extremely difficult to focus on one or the other aspect, let alone attempt to quantify it.”

The team recorded a number of potential fusion events using object tracking software and by applying a filter kernel were able to exclude false positives by identifying ‘death events’, or the termination of vesicle trajectory.

Another important task the software does is to distinguish fusion events from so called “kiss and run” events where the vesicle approaches the membrane but fails to fuse. This is quite common.

“The manual method of looking for vesicle fusion is a bit like watching Big Brother,” Vallotton says. “You have to do a lot of viewing to find the interesting parts but, by using automated image analysis software, our team (was) able to home in on the right locations in the right images much faster. This allows (us) to quickly gather critical information about how cells and their activities are affected in diseases like diabetes.”

The team is planning further investigations using the method to identify modulators of fusion and transport activity in compound screening programs, to measure accurately the kinetics of vesicle trafficking and to use these measurements to model the vesicular transport system in beta and fat cells.

The TIRFM data analysis is just one project, however. Under the collaboration, the researchers are planning to develop algorithms to enable real time 3D deconvolution in wide-field microscopy, a way of reversing optical distortion. This will be of use in studying vesicle transport in 3D, and in microtubule dynamics, where they are planning to work on a high throughput technology to measure the kinetic rates at both extremities of single microtubules, by using fluorescence speckle microscopy and new image analysis techniques.

CSIRO Biotech Imaging has also developed a technique called differential aberration correction (DAC) microscopy, which allows nanometre-scale distances in proteins and molecular complexes to be measured. This will be used to study ligand-receptor interactions, receptor dimerisation and conformational changes in membrane protein receptors upon activation.

They also plan to develop new technologies in live cell imaging to study the movement of cells and to improve fluorescence microscopy by using latest fluorophores and probes to avoid the problems of photo-bleaching.

David James says the collaboration will open up new possibilities in cellular imaging. “Most academic institutions have fancy microscopes but the real challenge is data analysis,” he says.