Subscribe
Home > Blogs & Communities > Origins > Visualizing Life's Origins  

A Modern Artist's Take on Darwin's Classic | Main | Did Metabolism Set the Stage for Life?

January 28, 2009

Visualizing Life's Origins

Janet Iwasa creates the pictures worth 1000 words. After earning a Ph.D. in cell biology, she decided to pursue a career in science animation and landed what she called a "dream" postdoctoral position with Jack Szostak, an origin-of-life researcher at Harvard University. Her job was to show what the earliest life looked like. Her work is showcased in the Boston Museum of Science, in the 2008 International Science & Engineering Visualization Challenge, and in Science's first monthly "Origins" essay, "On the Origin of Life on Earth." She is now a lecturer in molecular visualization at Harvard Medical School. Here's her story in her own words.

What did Earth's earliest forms of life look like? This was one of the central questions that I faced as I started a postdoc in Jack Szostak's lab at Massachusetts General Hospital in the fall of 2006. My postdoc was a very unconventional one--through NSF's Discovery Corps, and I was being funded to carry out a 2-year project to work with the Szostak lab and Boston's Museum of Science to launch a multimedia exhibit on the chemical origins of life. Not only was I jumping into a scientific field that felt far removed from my former area of expertise, but the numerous molecular animations I proposed to create for the project would require mastery of 3D animation techniques that I had only just begun learning. It was my dream postdoc, one that I hoped would launch my career in scientific visualization, and I couldn't wait to get started.


Several years earlier, while studying the actin cytoskeleton as a Ph.D. student at UCSF, I watched an animation of the motor protein kinesin walking along a microtubule. Based on biochemical and structural data, the movie, made by Graham Johnson, was essentially an animated model of kinesin's ATPase cycle--a process which I thought I understood but didn't completely appreciate until I watched the animation.

It wasn't long after viewing the kinesin animation that I resolved to become a scientific animator myself, and with the blessing of my adviser, Dyche Mullins, I started taking a class in 3D animation and creating some simple animated models of my labmates' research. I also started looking for postdoctoral funding that would allow me to focus on molecular animation and was beginning to get disheartened when a friend pointed me towards NSF's Discovery Corps program (which is, sadly, no longer being offered). I spoke to the program officer, who encouraged me to apply for the fellowship and pointed me towards the Szostak lab. The chemical origins of life was an ideal area to focus on; there was a huge amount of public interest in understanding life's molecular origins but very few visualizations to enable learning. Since origins-of-life research spans multiple disciplines, including astronomy, geology, chemistry, and biology, visualizations could aid in communicating scientific concepts not only to the public but also to other scientists within the research community.

My proposal was accepted in the spring of 2006, and a few months later I wrapped up my Ph.D. thesis and drove down to Los Angeles for a summer of 3D animation boot camp to kick-start my postdoc.

Upon landing in Boston with my freshly acquired animation skills, I set up my computer, ordered animation and compositing software, and pondered over where to begin. Protocells, as described by the Szostak lab, are composed of only two types of molecules: fatty acids and nucleic acids. I thought I'd start with what seemed like the slightly easier animation target: fatty acids.

The key to animating fatty acids was figuring out a way to show individual lipid dynamics within the larger context of a membrane. Jack was particularly interested in showing, at a molecular scale, how the fatty acid constituents of a vesicle membrane are constantly moving--both entering and exiting the membrane and flipping between the two bilayers--while the vesicle itself appears to be unchanging in size and shape.

I ended up trying several methods to create fatty acid vesicles in my animation software, called Maya. First, I created the chemical structures of the fatty acid in ChemDraw and exported the structures in the Protein Data Bank (PDB) format. PDB files can be directly read by Maya using a plug-in that places CPK-colored spheres at each atomic coordinate. I made a first quick attempt at a protocell using this method.

Image converted using ifftoany

An early attempt to illustrate a protocell

Having spheres for each atom was fine for looking at a small number of fatty acids, but the file size quickly became unmanageable when I tried to create a vesicle, even an unrealistically small one, and even just in cross section. This method also didn't seem particularly well-suited for animation.

For a second attempt at creating vesicles, I decided to go through a different route. Many 3D animation packages include particle systems that allow the animator to create a large number of objects that can be controlled individually or as a group. By using a type of particle known as a sprite, I created 2D cross sections of micelles and vesicles in which each fatty acid is essentially a looping movie whose movements can be choreographed using a script. This method allowed me to create animations that clearly showed the dynamics of individual fatty acids within a membrane bilayer.

fatty_acid_dynamics

Fatty acid dynamics (above) and vesicle formation (below).  Click the image to view the animation (also viewable at ExploringOrigins.org)

vesicle_formation_from_micelles

The sprite method worked well for creating animations and illustrations of vesicles that focused on the individual movements of the lipids, but I needed to find yet another technique to create larger, 3D vesicles to illustrate and animate protocells. Two methods came to mind: The first was to model the lipids using a fur system, which many animation packages include, and the second was to use a system that is used primarily in Maya to model vegetation, called Paint Effects. The latter method had been used by Drew Berry with beautiful success to animate phospholipid bilayers and ended up being the most convincing option for fatty acids as well, both in terms of their appearance and dynamics.

final_protocell

Final protocell illustration

The animations and illustrations I created in collaboration with the Szostak lab can be viewed in the completed online exhibit called "Exploring Life's Origins" at the Boston Museum of Science. In addition to the Web site, I gave a series of presentations to the public at the Museum of Science during the summer of 2008 and incorporated the visualizations into a touch-screen interactive kiosk which is now located on the museum floor.

—Janet Iwasa

Leave a comment

Thanks for your feedback. Please keep it polite and to the point.