Making models of cells

LA JOLLA, Calif.—The Scripps Research Institute (TSRI) and the University of California, San Diego (UCSD) have formed a new consortium with a big mission (in a tiny space): to map cells in space and time.

 

As TSRI Acting President and CEO Jim Paulson puts it: “The Visible Molecular Cell Consortium [VMCC] aims to bring together the best minds from different disciplines to understand and articulate how the body’s cells work, which will lay important groundwork for understanding health and disease.”

 

Pradeep K. Khosla, the chancellor of UCSD, originally approached Paulson about the partnership. “The collaboration [between UCSD and TSRI] will advance scientific excellence and research infrastructure at both institutions. The goal of building virtual cells poses an important challenge to researchers in fields from experimental biology to computation and information analysis,” Khosla said.

 

The VMCC will be directed jointly by TSRI professor Dr. Arthur Olson and Dr. Rommie Amaro, a UCSD associate professor of chemistry and biochemistry. The consortium will offer fellowship funding for 10 to 12 graduate students and postdoctoral fellows to work on collaborative projects that build bridges between the campuses and different disciplines, in order to assemble and simulate a virtual model of a cell, down to an atomic level of detail. They aim to visualize protein interactions in real time to better understand cellular function.

 

“Even the simplest living cells contain 1 to 2 million proteins of 3,000 to 4,000 different types,” notes Olson. “Figuring out how they work together over time will shed light on the cell as a living, working individual entity. Just like you couldn’t build a car from just its wiring diagram, we can’t have a complete understanding of a cell unless we know how all of its physical parts work together in 3D.”

 

This is a large data challenge, Olson points out, applied to the uncharted territory of cellular architecture and ecology. In recent years, better and more powerful imaging devices and automated programs in high-resolution image analysis and visualization have provided more detailed pictures of cells and their proteins than ever before. Scientists have not yet translated the enormous amounts of data into a single, atomic-level cellular model.

 

“This is a particularly exciting time for such efforts, due to a number of technological and scientific factors,” enthuses Amaro, and she adds: “Advances in various imaging technologies, modeling frameworks and cyber-infrastructure are enabling us to make new strides in the creation of 3D virtual cells. This timely new inter-institutional alliance will provide new insights into the inner workings of cell machinery, some of which may present opportunities for novel therapeutics.”

 

TSRI is well known for its structural biology using both cryo-electron microscopy and X-ray crystallography, and Olson’s lab develops and uses graphics programs (similar to those in the video game industry) to visualize complex cellular machinery. UCSD is home to the only publicly available supercomputer in California and to the National Biomedical Computation Resource, a National Institutes of Health-sponsored national resource that develops multiscale modeling tools.

 

According to Olson, he and Amaro plan on developing a dual mentorship fellowship program to cross-train graduate students and postdoctoral fellows with two mentors on different sides of exploration—for example, experimental versus computational, or observation light microscopy versus electron microscopy. The long-term goal is to train scientists capable of bridging modalities and synthesizing more complete ideas.

 

Visualization tools will help present all of this information to the human brain in a coherent fashion, but in order to achieve this visualization, the researchers will need both computational and algorithmic modeling of how the elements of the cell actually behave.

 

“We’re trying to represent as closely as possible the nature of a living cell as a physical and dynamic object,” Olson tells DDNews. “The technology we’re using can enable us to see and interact with extremely complex models that contain millions of protein molecules, each of which has tens of thousands of atoms, in real time.”

 

Olson predicts the researchers will be able to map a simple cell such as a red blood cell, down to the atomic level, within five years. More complex cells, such as nerve cells, will take more time to map completely.

 

“If you look at the island of Manhattan, there’s about 2 million people living there—about the same as the number of protein molecules in E. coli. And we currently have the technology to figure out what every person on Manhattan is doing, where they’re going and who they’re talking to,” says Olson. “Such is the magnitude of this cell mapping project. I’m a big fan of collaborating and using as much scientific and intellectual horsepower as we can to attack important projects. Science is a huge enterprise, and we need to be able to share data as best we can.”

 

The organizers anticipate the consortium will be particularly strong in neurological diseases and infectious diseases, such as influenza, HIV and Ebola virus, although the insights into cellular behavior will be applicable across many fields. Olson and Amaro plan to organize a biannual conference to encourage new collaborations and share results. Researchers who are interested in learning more about the consortium are encouraged to contact visiblemolecularcell@gmail.com.