Illustrating the Invisible

Delicately shaded chemical bonds painted with the finest of brushes, explosions of brightly colored ribbons tracing a molecular backbone, densely packed watercolor landscapes teeming with organelles and proteins. Irving Geis could hardly have imagined what he was launching when he began painting the myoglobin molecule in 1961. But in the 50 years since, the field of molecular illustration has burst open in a fury of color and imagination.

Myoglobin was the first protein structure solved—and the first molecule most people had ever seen. Trained as an architect, Geis had been illustrating for Scientific Americansince 1948. He had a reputation for taking on difficult projects. So when it came time to illustrate John Kendrew’s article on the myoglobin protein’s structure, Geis was a perfect choice. It took him nearly six months to complete his labyrinthine watercolor painting of myoglobin—a far cry from the few weeks he normally needed to illustrate a scientific concept. But the end result was worth the wait. The painting, a lattice of Tinker Toy-like pieces strung together to show the location of each molecular bond, is an icon of scientific illustration.

Demanding as it was, Geis was hooked. He went on to create hundreds of other reproductions of molecules, revealing not only the way each one looked but how it worked. His initial paintings were similar to the myoglobin illustration—crowded frameworks of short color-coded lines and circles resembling jumbles of crumpled chicken wire. In some of his images, unseen spotlights cast an eerie glow, highlighting the binding sites and cofactors that powered the molecular machines. Occasionally, Geis used what he called “selective lying”— tiny distortions and exaggerations to resolve overlaps and create understandable images. Later, his work became more stylized, with thick tubes snaking their way across the canvas evoking the gross molecular structure and finer embellishments like brightly colored disks and shiny spheres standing in for various functional elements. HHMI houses a collection of Geis’s work at its headquarters in Maryland.

By 1980, researchers had solved about 70 protein structures and they needed a simple universal way to share these discoveries. After all, not everyone could hire Geis to paint their molecules. “It was pretty clear that something further was needed,” says Jane Richardson, a biochemist at Duke University. Several people had tried their hand at drawing proteins, with varied results. “Even when the proteins were related, people would show them from different points of view and with different representations so it was hard to see their similarities,” says Richardson.

Richardson decided to give it a shot. The outcome: proteins as flowing intertwined ribbons. She used these deceptively simple loops and lines to illustrate a 1981 review article about protein structure that she wrote with the help of her husband, David Richardson, for Advances in Protein Chemistry“I spent two years writing that 169-page article, and a year of that was learning how to make these drawings,” recalls Richardson. “There were only about 75 really different protein structures at that point and I made drawings of all of them.”

Like Geis’s tubes, Richardson’s ribbons accurately traced the path of a protein’s backbone in three dimensions. The flattened ribbons twisted and turned like delicately placed party streamers, suggesting the relationships between neighboring backbones. Her work was clean and unadorned. She drew alpha-helices as corkscrewed ribbons, beta-strands as arrows, and loops as thin tubes. Blended colors and shading added depth.

Soon, Richardson’s straightforward system became the standard for visualizing proteins. Computer programs that reproduced her fluid ribbons found their way into structural biology labs. They became the go-to way of representing a protein’s structure. Programmers even added interactivity so that viewers could move molecules around on the screen, zooming in and out, examining the images from all angles and then adding local details.

The Big Picture

The innovation didn’t stop there. As scientific techniques advanced, researchers focused on larger molecular complexes and whole cells. Molecular illustration needed to expand to cover the big picture—and computers couldn’t do it. In the late 1980s, David Goodsell, a molecular biologist at the Scripps Research Institute, used ink and watercolor to create cellular landscapes, like cross sections of the single-cell bacterium Escherichia coli.

Crammed with life, Goodsell’s paintings are a visual symphony of shape and color. Hundreds of molecules and organelles come alive on the canvas. “I try to take a chunk of a living cell and draw a picture that shows everything that you would see if you were there,” he explains. “I try to get all the molecules in the right place and the right size and the right concentration.”

Goodsell consults electron micrographs and published molecular structures, making sure his subjects’ shapes, sizes, and locations are spot on. He uses color to differentiate ribosomes from RNA and show some of the structure–function relationships within a cross section. The result is chaotic and dense, true to the invisible world inside a cell.

“He’s really changed everybody’s feel of what the cell is like,” says Richardson. “His illustrations of the crowded cytoplasm are great. People may have known intellectually, but I think they really didn’t appreciate how crowded it is.”

Although computer graphics are improving and may soon be able to replicate three-dimensional models of large cellular swaths, Goodsell doesn’t believe human illustrators will ever become obsolete. “A person that has some artistic knowledge will always be able to come up with an illustration that is clearer, that shows the concept, and gets people more excited.”

-- Nicole Kresge
HHMI Bulletin, Winter 2013

Return to main article >