Pulmonology
Anatomy of an Asthma Attack

Complexity
Precursor Cells Follow Different Paths to Same Cell Fate

Genetics
Gene Network Predicts Stroke Risk in Sickle Cell Anemia

Education
Harvard Approves MD–PhD Program in Social Sciences

No Link Seen Between Dietary Patterns and Pancreatic Cancer Risk

New Heart Attack Therapy May Be Coming

Proceedings of the HMS Faculty Council

New Appointments to Full Professor

Stem Cell Research at BID Gains $6 Million Gift

Honors and Advances

News Brief: Soros New American Fellowship

In Memoriam

Survey Seeks to Improve Student Life on Longwood

Front Page

Precursor Cells Follow Different Paths to Same
Cell Fate

To peer inside a developing cell is to stare directly into the face of a conundrum. A human liver cell precursor has the same 30,000 genes as, say, a nascent kidney cell, and yet once launched on their developmental paths, the cells turn their genes on and off in a very different fashion. How do developing cells emerge from the thicket of genetic possibilities to become cells of a particular type? How do they know which genes to turn on and off? And what allows them to follow their genetic fates so unswervingly?

Photo by Steve Gilbert “Our view is that a cell’s differentiation state is there as a pre-existing program, an attractor, and all you need is to tip over into it,” said Sui Huang. “You do not need very specific instructional inputs.”

HMS researchers, working with a scientist from the New England Complex Systems Institute (NECSI), have provided the first experimental confirmation of a theory that has intrigued and inspired controversy among researchers for nearly 40 years. The idea is that a developing cell is guided toward its fate by the stability-seeking behavior of its entire network of genes and proteins. This contrasts with the view held by many molecular biologists that development is the result of the turning on and off of a specific sequence of genes that essentially teaches the cell what to become. The study by Sui Huang, HMS assistant professor of surgery at Children’s Hospital Boston, and his colleagues could have important implications for two hot areas of research—stem cell and systems biology.

“Systems biology is saying we need to understand how complex behavior emerges out of collective interactions,” said Donald Ingber, the Judah Folkman professor of vascular biology in the Department of Pathology at HMS and Children’s. The paper by Huang, Ingber, Yaneer Bar-Yam, president of NECSI, and Gabriel Eichler, a technician in Ingber’s lab and now a graduate student at Boston University, appears in the April 1 online version of Physical Review Letters.

Complex Behavior
With so many genes turning on and off inside the developing cell, the potential for instability is great. In the 1960s, the theoretical biologist Stuart Kaufmann proposed that if a cell were to hit upon an arrangement in which every element affected what the others were doing, much in the manner of a feedback system, that cell would better resist perturbation. In fact, evolution appears to have selected for just such a system—genes and proteins often inhibit or activate one another. The collective behavior of this network creates a system that is so stable and self-correcting that it exerts an almost gravitational pull—the developing cell is compelled toward its fate, for instance, to become a liver cell. Kaufmann called these stable networks “attractors,” a term he borrowed from physics. But evidence for the existence of attractors was lacking.

In one respect, it is easy to see why. In Kaufmann’s view, a cell hits upon a stable genetic configuration and develops into a particular cell type. But demonstrating that the entire network functions as an attractor, pulling the cell down a particular developmental path, is problematic. One could argue, as molecular biologists have, that the network is simply the consequence of individual signaling pathways instructing the cell to develop in a particular direction. There is another approach. According to Kaufmann, attractors are very robust—they cast a wide net. Two developing cells might initially exhibit wildly differing constellations of gene activity and yet they could eventually be caught up by the same stable configuration, or attractor, and become the same kind of cell. That is what Huang and his colleagues have shown.

“We start at the same state with the same set of genes and end with the same state, but the cells’ two paths are totally separate,” said Huang. “During day one they are really extremely different. Then they converge to a similar state.”

They compared how neutrophils develop from two different directions. For decades, lab workers have been culturing this type of white blood cell by exposing a precursor cell to one of two reagents, DMSO and all-trans-retinoic acid (atRA). Using DNA microarrays, Huang and his colleagues monitored the expression of 12,600 genes in cells undergoing these two types of treatment. They measured activity in the entire array of genes at regular intervals—0, 2, 4, 6, 8, 12, and 18 hours and then daily until the cells differentiated into neutrophils on day seven. They found that the two sets of cells exhibited very different—even diametrically opposed—expression patterns. Yet by the end of seven days their gene expression patterns were congruent. And both sets of cells had turned into neutrophils.

“We start at the same state with the same set of genes and end with the same state, but the cells’ two paths are totally separate,” said Huang. “During day one they are really extremely different. Then they converge to a similar state. What is the driving force that makes them converge? One solution totally obvious for mathematicians is that the final state is an attractor state. It is robust. You could put any state in its vicinity and it would converge.

Finding One’s Marbles
Even the most pathway-conscious molecular biologist may be familiar with the idea that developing cells are drawn to their fates. One of the iconic images of biology, portrayed by the early 20th century biologist C.H. Waddington, shows a marble poised between two peaks. The marble, intended to represent a cell, is about to roll into a series of ever narrowing valleys, each representing a possible differentiation state. Waddington’s picture of the “epigenetic landscape” was given new life in the 1960s, when Kaufmann ran a series of computer simulations. He found that under certain conditions, random collections of tens of thousands of genes would reliably produce a relatively small number of stable networks. He called these stable systems “high dimensional attractors” because they contained so many elements.

Huang’s goal was to update Waddington’s image even further. “I wanted to chart the molecular biology of the epigenetic landscape,” he said. The advent of DNA microarrays in the late 1990s made it possible to explore the collective behavior of thousands of genes at a time, and yet no one else seemed to be using the new technology for that purpose. “People were using microarrays just to find another differentially expressed gene,” said Huang. He and colleagues began working on an experimental design. “The way we decided to study it is by comparing how cell behaviors respond to two different stimuli,” said Bar-Yam. Ingber recalled from his undergraduate days that researchers down the hall from him were culturing neutrophils by using two all-purpose reagents, DMSO and atRA. People thought both reagents might be triggering a specific neutrophil differentiation pathway, but no one had ever found it.

Neutrophils develop along two different paths. Precursor cells treated with two different reagents diverge, and at times take opposite tacks, before exhibiting the same gene expression pattern—that of a neutrophil. Mosaic insets represent gene expression profiles at various times. The convergence on a common endpoint from such different directions suggests the existence of an attractor. (Image courtesy of Sui Huang)

The new study suggests that the DMSO- and atRA-induced precursors may simply be caught up by the same attractor. What is interesting is that although the researchers predicted the two sets of cells would initially diverge, they did not expect them to vary as much as they did. A gene that was overexpressed in the DMSO cells was often underexpressed in the atRA-induced ones. Fifty percent of the genes exhibited this diametrical pattern. “What is shocking is that they really go in almost opposite directions,” said Huang.

The researchers began their neutrophil study in 2000. It has been a long road to get their findings published. There were technical difficulties, but the real problem was winning over the reviewers of the biological journals to which they first submitted their study. “In general, biologists think in terms of individual components—genes and gene interactions,” said Bar-Yam. Their paper could find a warmer welcome in biology now, due in part to the rise of systems and stem cell biology. “One of the goals of systems biology is to open people’s minds to network level properties that cannot be defined by looking at any one of the subcomponents,” said Ingber.

“I think it is going to have real relevance for stem cell biology,” added Huang. “What is really interesting about stem cell differentiation is that the first stretch of the path is often the same. So it is the deeper structure of constraints that must be organized to create diversity.”

—Misia Landau

top