Code on a Grecian urn

Haley Bridger, September 20th, 2012
  • The cover of the September 14 issue of Molecular Cell
    (2012 Volume 47, Issue 5).

A layer cake? No, not interactive enough. A spinning wheel? Too mechanistic. What about a film strip? No, that’s not quite right either.

Broad researcher Nir Yosef and scientific illustrator Sigrid Knemeyer were in search of the perfect metaphor to communicate the complex scientific concepts encapsulated in a paper by Yosef and his colleagues. The two spoke by phone and over email, exchanging ideas about how to capture the paper’s main messages. Through sketches and descriptions, one idea began to take shape: a Greek vase, depicting Gods, Olympians, and Warriors, as well as DNA and genes. The image was selected by the journal’s editors, and last week graced the cover of Molecular Cell.

To understand why this particular metaphor works so well, I spoke to both Yosef and Knemeyer about the paper and how the cover concept emerged.

The paper itself stems from a technological breakthrough, Yosef explained. Yosef, a postdoctoral fellow in Aviv Regev’s lab at the Broad and in Vijay Kuchroo’s lab at Brigham and Women’s Hospital, and his colleagues at the Broad Ido Amit (now a principal investigator at the Weizmann Institute) and Manuel Garber (now a principal investigator at the University of Massachusetts) are interested in studying a class of proteins known as transcription factors. These proteins control which genes are turned on or off in the genome. Mapping the locations of where these factors bind has, until now, been very labor intensive and very expensive, relying on a protocol called Chromatin ImmunoPrecipitation (ChIP). But Ido Amit helped to change that.

“He streamlined the protocol in many ways – it’s hard to put your finger on just one innovation because he innovated a lot of steps along the process,” says Yosef. The end of result of Amit’s work is a highly efficient method that allows researchers to map the binding of many transcription factors simultaneously. Researchers are not only interested in location but also in timing: they want to know both where and when factors bind. Using Amit’s streamlined protocol, Yosef and his colleagues were able to look at 25 transcription factors at once, and at four different time points.

“With this map in our hands, we were able to start asking deeper questions about how transcriptions factors cooperate with each other,” says Yosef. The team chose to study dendritic cells – immune cells that play a critical role in distinguishing different kinds of pathogens and directing the immune system’s response (more here).

By looking across time, the team found that there are three “layers” of transcription factors, with different – but interactive – roles. The researchers call the first layer “pioneer factors.” These proteins bind early in a cell’s development, and act like a foundation, changing very little over time. The second layer – priming factors – are more selective than the pioneers, binding only to a subset of genes. These are the genes that will be turned on in the future after a cell is exposed to pathogens, at which time the final layer of transcription factors will bind to them.

Yosef explained this hierarchy of layers to Knemeyer, and the two began thinking about the best way to illustrate them. “Our first thought was a layer cake,” Knemeyer recalls. “But those layers don’t interact with each other. Nir also mentioned that the layers are very dynamic – there’s a lot of action going on – so my next thought was a strip of movie roll, with different scenes depicted.”

But then Yosef brought up the idea of a Greek vase, which could show layers but also show different kinds of activity. The idea clicked, and both Yosef and Knemeyer began thinking about creative ways to elaborate on it.

“Once we narrowed in on the Greek vase, we really got into analogies,” says Knemeyer. In the top hierarchy on the vase, two Greek gods (Zeus and Hera) wrap yarn-like DNA around histones – they are meant to represent the pioneer factors, regulating the epigenomic state of the cell. They in turn influence the second tier of Olympians – primer transcription factors that bind to only a subset of genes. Finally, at the bottom of the vase are the warriors, which activate specific pathways such as inflammation in response to pathogens.

Knemeyer added each of the figures to the vase, trying to make the image look as realistic as possible. “I looked at old Greek vases in my art history books and tried to come up with a style that represents them best,” she says. “I set up my actual rendering in Photoshop and it reminded me of painting again with oils. There’s a very similar basic painting principle that you apply for getting it to look like a vase. It reminded me of how much I love painting and creating something that looks three dimensional.”

In addition to Yosef, Garber, and Amit, other researchers who also contributed to this work include Alon Goren, Raktima Raychowdhury, Anne Thielke, Mitchell Guttman, James Robinson, Brian Minie, Nicolas Chevrier, Zohar Itzhaki, Ronnie Blecher-Gonen, Chamutal Bornstein, Daniela Amann-Zalcenstein, Assaf Weiner, Dennis Friedrich, James Meldrim, Oren Ram, Christine Chang, Andreas Gnirke, Sheila Fisher, Nir Friedman, Bang Wong, Bradley Bernstein, Chad Nusbaum, Nir Hacohen, and Aviv Regev.