American Scientist

Our choice of careers as adults isn't determined by whether we slept on a cot or a feather mattress as a child. But then, we're not stem cells (at least, not any more). A pair of recent reports shows that when scientists grow stem cells in the laboratory, the physical properties of the cells' culture-dish homes influence when they will adopt a distinct path in life and what that path will be.

At the 2006 annual meeting of the American Society for Cell Biology in San Diego last December, Christopher J. Murphy of the University of Wisconsin-Madison presented data showing that embryonic stem cells are more likely to keep their pluripotency—their ability to become any type of cell—when they are grown on a surface stamped with a pattern of tiny ridges. The effect was independent of the scale of the ridges, which ranged from nanometers to micrometers in width. Traditionally, cultured cells are grown on smooth glass or plastic surfaces.

This finding challenges the ingrained culture of cell culture, which has long assumed that the molecules dissolved in the liquid medium that bathes cells, including growth factors and other chemical signals, had the last word in determining cell physiology. Murphy disagrees, citing work in his laboratory that shows physical properties of the surface to be "as fundamental an element [in determining cell behavior] as having growth factors in the media."

So what exactly does the physical topography do to the cell? "What doesn't it do?" asks Murphy. "It changes everything—adhesion, migration, proliferation, differentiation." From his systems-engineering perspective, the research offers potential benefits for the large-scale production of embryonic stem cells, which have the theoretical ability to divide indefinitely but often lose pluripotency for reasons that are poorly understood.

The Wisconsin team's findings echo previous work by scientists at the University of Pennsylvania. In the August 25, 2006, issue of the journal Cell, the Penn group showed that the fate of mesenchymal stem cells could be directed by the stiffness of the substrate on which they were grown. Mesenchymal stem cells come from adult bone marrow.

A research team led by Dennis E. Discher found that stem cells grown on the stiffest matrix became bone precursors. Those grown on the softest surface became nerve cells, and those grown on a medium-stiff substrate assumed the characteristics of muscle cells. The shapes of the cells and the suite of active genes contained within confirmed their new identities. Previous studies had shown that chemical cues could effect this kind of differentiation, but Discher's paper was the first to demonstrate that cell lineage could be controlled in the absence of soluble stimuli.

This property makes sense, Discher says, given the role of such cells in tissue regeneration and repair. "These [mesenchymal stem] cells leave the bone marrow and end up in different places—muscle, bone, fat." But to repair muscle, or become a neuron in the brain or cartilage at the end of a bone, the stem cell must become anchored and "physically engage the microenvironment" before becoming one of the gang.

Such differences in the relative stiffness of animal tissues are easy to appreciate, according to Discher. "You can feel it. You know brains, calf's brains, how soft they are?  I take you to the supermarket, you can feel the difference between steak and bone, or between fat and calf's brains."

But groceries aside, how exactly does a cell "feel" the surface it grows on? "It starts with attachment and contraction," explains Discher. Cells create what are called focal adhesions—points of attachment on the substrate—that provide a foundation for the cytoskeleton. A type of motor protein known as nonmuscle myosin II applies tension to the actin filaments of the cytoskeleton. At the end of the line, a complex of proteins near the focal adhesion appears to act as a "mechano-transducer," translating physical forces into intracellular signals. In the Cell paper, Discher and his colleagues showed that inhibiting myosin prevented the substrate-based differentiation.

The strong effect of physical properties on cell behavior in these experiments doesn't mean that scientists were wrong before about soluble signals. On the contrary, the two types of stimuli seem to work together. But Discher states—using language similar to Murphy's—that "physical cues are just as influential [on cell fate] as chemical cues. It's a balance." In his mesenchymal stem cells, differentiation was more complete when chemical and physical cues pointed in the same direction.

The potential implications of this work for cell biologists are profound, given that everything scientists know about the way cells behave in culture comes from experiments on hard, smooth surfaces. For this reason, Discher cautions that biologists need to "take those data [from rigid substrates] with a grain of salt." Given the tens of thousands of research papers that include such methods, that's a lot of salt.