The work was supported by the EU via three projects: MITOCHECK (Regulation of mitosis by phosphorylation - A combined functional genomics, proteomics and chemical biology approach), MITOSYS (Systems biology of mitosis) and SYSTEMS MICROSCOPY (Systems microscopy - a key enabling methodology for next-generation systems biology).
Advanced imaging techniques allow researchers to visualise highly complex processes within cells. However, a large sample of cells may contain only a few cells undergoing the process under investigation; this is particularly a problem for relatively rare processes. Finding these cells requires researchers to spend hours at the microscope, scanning the cells manually in the hunt for cells that can be used.
The current study describes a neat software solution to this problem. Called Micropilot, the system effectively scans samples for cells of interest and performs appropriate experiments on them.
The system includes a machine learning-based module so it can be rapidly trained by a user to automatically identify the cells the researcher is interested in. Once trained, Micropilot can be left alone to scan the sample in a fast, low-resolution mode.
When it identifies a cell matching the researchers' needs, the system switches to a complex imaging mode that automatically carries out more complicated experiments. These could range from fairly simple tasks like recording high-resolution time-lapse videos to more complicated experiments that use lasers to interfere with fluorescently tagged proteins.
The system requirements on the hardware front include a motorised microscope stage, as well as the abilities to automatically switch between objectives or laser scanner zoom, and to switch fluorescence filter and/or laser lines.
The team tested their software on phases of the cell division cycle that are relatively quick, and so hard to catch 'in action'. With the Micropilot software, the team succeeded in determining when structures called endoplasmic reticulum exit sites (ERESs) form and shed new light on the roles of two proteins, CBX1 and CENP-E, in condensing the genetic material into compact chromosomes and in forming the spindle which helps to align the chromosomes during cell division.
What makes Micropilot particularly exciting is its speed; in just 4 nights of (unattended) operation, it detected 232 cells in 2 particular stages of cell division and carried out complex imaging experiments on them. In contrast, it would have taken an experienced microscopist at least a month working full time to detect these cells in a sample of thousands.
"Micropilot [...] liberates cell biologists from the tedious work of repetitive manual data generation," the team notes. "It can be adapted to virtually any imaging system that allows automation and online control based on the results of image classification by machine vision."
The team concludes: "In three independent experimental setups, [Micropilot] allowed us to statistically analyse biological processes in detail and is thus a powerful tool for systems biology."
The software is set to be a key tool in the EU-funded projects MITOSYS and SYSTEMS MICROSCOPY, which received EUR 10.2 million and EUR 12 million under the Seventh Framework Programme (FP7).
The other EU-funded project that contributed to this study, MITOCHECK, was funded to the tune of EUR 8.6 million under the Sixth Framework Programme (FP6).
For further information, please visit:
- European Molecular Biology Laboratory (EMBL), http://www.embl.org
- Nature Methods, http://www.nature.com/nmeth
- The Micropilot software is available at http://www.embl.de/almf/almf_services/hc_screeing/micropilot/
- MITOCHECK, http://www.mitocheck.org
- MITOSYS, http://www.mitosys.org
Copyright ©European Communities, 2011
Neither the Office for Official Publications of the European Communities, nor any person acting on its behalf, is responsible for the use, which might be made of the attached information. The attached information is drawn from the Community R&D Information Service (CORDIS). The CORDIS services are carried on the CORDIS Host in Luxembourg - http://cordis.europa.eu. Access to CORDIS is currently available free-of-charge.