By Aaron Krol
December 20, 2013 | Don Ingber, of Harvard University’s Wyss Institute for Biologically Inspired Engineering, is a mechanobiologist – a member of the small but growing community of scientists who look at biological systems through the lens of physics. Mechanobiologists hope to add to our understanding of physiology, embryology, and disease ontogeny by observing the mechanical forces at work in living cells and tissues. Ingber sees this work as part of a rich tradition dating to the birth of modern biology. “A hundred years ago,” he told Bio-IT World in an interview, “all biologists in developmental biology were focused on mechanics in the embryo. They watched development on the scope with their eyes, and they saw all these incredible mechanical transformations and constrictions and movement… But then, when chemicals and genes came in, it was thrown out.”
At a time when big data systems are drilling through the human genome to link the DNA code to health and development, and computer models are inventing novel biomolecules by the thousands to test their hypothetical properties, the methods of mechanobiologists, who still spend much of their time poring over microscopes and manipulating tissues by hand, can look primitive. But just because technical innovations in this field have not kept pace with its more fashionable cousins does not mean it’s dated or irrelevant. As founding director of the Wyss Institute, Ingber is working to bring cutting-edge technology to bear on the problems of cellular and tissue mechanics.
One particularly thorny puzzle has been how to measure force in vivo on the cellular level.
Force in Nature
Quantifying force is fundamental to describing any physical system, especially when you need to understand how small changes in force affect the system as a whole. Yet measuring force in living tissue – with its three-dimensional structure, complex network of cells exerting different forces, and sensitive chemical interactions with foreign objects – has been “a really complex problem that has held up the field for fifty, a hundred years,” says Ingber.
Don Ingber, Founding Director of the Wyss Institute at Harvard Medical School. Image credit: Wyss Institute
Until recently, mechanobiologists have had to content themselves with measuring force in vitro, in two-dimensional slices of cells in culture: for instance, by measuring the deflection of an atomic force microscope probe from the culture as it scans across its surface. When examining living tissue, researchers could do little better than slicing a sample open with a laser and eyeballing the tension as cells pulled apart. So “you could estimate force… [but] you can’t tell how much,” says Ingber. “And therefore it’s very hard to relate cause and effect,” if you want to demonstrate that force is related to other biological properties like genetic variation or disease progression.
That was the state of affairs when Otger Campàs came to Harvard as a postdoc in 2006. Campàs had studied biophysics at the Institut Curie in Paris, where he became interested in the forces exerted by actin filaments: nanoscale cellular structures that form dense networks to move substances through cells. Senior researchers at the Institut Curie came up with a novel method to measure those forces, using droplets of cooking oil.
“The way it worked was very simple,” Campàs told Bio-IT World. “They would actually make these filaments grow from the surface of these cooking oil droplets, and then the forces that these filaments would generate would deform the droplets. And from the deformation, we could understand what was the force that these filaments were generating.” The method relied on computational imaging analysis that could precisely quantify the shape of the droplets – very similar to techniques the Wyss Institute was already using to track the shape of cells in three dimensions.
Ingber was intrigued by the oil droplet method, and with Campàs’ help, the Wyss Institute decided to try the technique in 3D tissues, marking an oil droplet with fluorescent molecules, injecting it in vivo and photographing it with a confocal microscope. Campàs remembers that early trials had underwhelming results. “Droplets that are made of cooking oil mix with the membranes of the cells… you could see it in all cell membranes afterwards.” But when the team started using highly inert fluorocarbon oils, the droplets stabilized.
This mechanobiological breakthrough was aided by advances in more mainstream disciplines, which the Wyss Institute specializes in synthesizing into bioengineering solutions. A thorough knowledge of cellular chemistry allowed the researchers to coat the surface of their oil droplets with molecules that cells use to adhere to each other, and to the extracellular matrix. “A cell on the outside of it just thinks it’s another cell,” says Ingber. “The matrix interacts with it like it’s a cell, and it interacts with the matrix like it’s a cell. So it’s basically a decoy.” Computational biology also played a key role. In addition to being a biophysicist, Campàs is a skilled programmer, and it was his coding that allowed the raw fluorescent images to be converted into useful information about a droplet’s shape.
Otger Campàs, Associate Professor of Mechanical Engineering at UC Santa Barbara. Image courtesy of Otger Campàs
Campàs wrote a code for the MATLAB platform to plot individual points along a droplet’s surface, at 40 nm resolution, and create a 3D mesh representing the entire object. He then exported that information into Mathematica, where his algorithms derived, from the shape’s degree of divergence from a perfect sphere, the stresses applied to the droplet. “It’s incredibly elegant,” says Ingber, “in that all you do is measure deformation, and his algorithms get out stress.” With the new biochemistry, and sophisticated analytical tools, the droplet method began providing highly sensitive measures of cellular force in living tissue.
There are limitations to the technique – mainly that it is hard to verify its accuracy. “There is no other technique to measure in vivo,” points out Campàs, so researchers can’t take a separate set of measurements to double-check. But they have compared the technique to in vitro methods, relating 3D forces to the 2D forces exerted by the same cells. They also caught a lucky break when Christopher Chen’s lab at the University of Pennsylvania, at around the same time, created a second method of measuring 3D cellular force, by immersing cells in a 3D gel and measuring the distortions to this matrix. The methods aren’t directly comparable – the U Penn technique can’t be used in live tissue – but they do yield similar force measurements, an encouraging sign that both methods are at least approximately accurate.
The Wyss Institute’s proof-of-concept was a measurement of forces exerted by tooth mesenchymal cells in living mouse mandibles – not the most thrilling study, but a useful way to validate the droplet method in a real in vivo environment. It took years to find a home for the resulting paper, but on December 8, 2013, it finally appeared in Nature Methods, a high-profile venue that should help inspire new researchers to take up the technique.
A droplet embedded in mesenchymal cells from a mouse mandible. From Nature Methods, reprinted with permission of Otger Campàs
In the meantime, the authors have had plenty of time to apply their novel methods to more pressing research. Ingber is interested in diseases that cause stress to tissues, like asthma, hypertension, irritable bowel syndrome, and diverse cancer types. In this area, force measurement could help understand disease progression, or gauge the severity of different cases. It could also be a useful tool for quantifying the success of interventions, for instance by comparing two asthma drugs to see which relieves more pressure on the smooth tissue of the lungs.
Campàs, meanwhile, has moved on to the University of California, Santa Barbara, where he holds the Mellichamp Chair in Systems Biology and runs his own lab. The Campàs Group is focused on embryology, trying to discover the forces at work in developing organisms. For this, Campàs needs to adapt the droplet method to the multicellular level. “If you make bigger droplets,” he says, “then you start measuring tissue-scale forces. The problem with that is, in embryonic tissues, if you make droplets that are very big, then you are affecting the development of the tissue.” Instead, the team uses many droplets dispersed throughout embryonic tissue to measure cellular forces at multiple points, and scales that information up to determine whole-tissue forces.
For this work, Campàs has teamed up with the Jerome Gros lab at the Institut Pasteur in Paris, measuring force in developing limb buds in chicken and zebrafish embryos. With this research, the partners are cementing the relationship between genetics and mechanobiology. By performing gene knockdown studies on their embryos, they can show which genes contribute to cellular forces during embryological development, filling in the picture of how genetic instructions ultimately shape whole organisms.
A computer-generated 3D image of an oil droplet. From Nature Methods, reprinted with permission of Otger Campàs
Campàs also seeks to make the technology behind the droplet method available to scientists around the world. In the years since the first steps toward measuring cellular force in vivo were taken at the Wyss Institute, Campàs’ techniques have grown steadily more sophisticated. His lab is nearing completion of a second, more advanced version of the image analysis software, and has been rapidly modifying the molecular coats of the droplets to make them compatible with new tissues and environments. “The coating of the droplet is extraordinarily modular,” Campàs says, “in the sense that you can put on the surface of the droplet whatever molecule you want…. We are able to do very specific and versatile coatings.” He’s even trying to reverse the engineering process, building a library of coats that can later be filled with oil droplets, rather than synthesizing droplets and then attaching the desired molecules to their surface.
“Many people have asked me to send them drops,” continues Campàs, “and unfortunately I cannot handle that…. We don’t have the resources at the moment to scale up the system for many groups.” The process is patented, and both Ingber and Campàs are eager for a commercial company to pick it up and begin producing the droplets on an industrial scale. “What we’re trying to do now is to license these droplets,” says Campàs, “to make a package, so people can just buy these droplets with the coatings that they want, and with the software.”
They say that almost any biological lab should be able to use their method with standard equipment and little specialized training. The biggest hurdle is the computational skill needed to analyze the 3D images for deformation. But in the interest of spurring innovation in this neglected field, Campàs is prepared to make his algorithms easily available. “I’m more than willing to share [the software] for scientific purposes,” he says.