Donald Ingber

The claim

The cell is a tension structure, and tension is how it hears

Donald Ingber established that the living cell is a tensegrity structure, held in shape by continuous tension and propped apart by discontinuous compression, and that this architecture is what allows a cell to feel a force and answer it. He set out the argument as a Commentary in Journal of Cell Science in 1993, volume 104, pages 613 to 627 (Ingber 1993). In the same year, with Ning Wang and James Butler, he published the experiment in Science, volume 260, pages 1124 to 1127 (Wang and Butler 1993). Bind a magnetic bead to an integrin receptor on the cell surface, twist it, and the entire cytoskeleton stiffens in direct proportion to the force applied. A pressurized bag does not do that. A rigged mast does.

Ask what that commits you to. If a cell holds its shape by prestress, then the resting level of that prestress is not a housekeeping detail. It is the setting that decides how the cell will answer the next load, before any molecule has time to diffuse across it. Change the tension and you have already changed the response. That is Ingber's finding, not ours. What this site adds is a single word for the variable he measured.

The inherited picture

Biology had no structural theory of the cell

For most of the twentieth century, cell biology worked with a bag. A membrane enclosed viscous cytoplasm, organelles floated in it, and the cytoskeleton was scenery. Signaling meant chemistry: a ligand meets a receptor, a cascade fires, a gene switches. Mechanics belonged to engineers who studied bone loading and arterial wall stress, and it stopped at the tissue. Nobody expected a molecular biologist to talk about struts and cables. Ingber was trained on both sides of that line. Born in 1956, he took a B.A. and M.A. in molecular biophysics and biochemistry at Yale in 1977, an M.Phil. in cell biology in 1981, and an M.D. together with a Ph.D. from Yale in 1984.

Why did the bag model last so long? Because it predicted enough to be useful. Diffusion explains a great deal of what cells do. But it does not explain speed, and it does not explain the fact that a cell knows what surface it is standing on. Mechanical signals travel along load-bearing elements. Chemical signals wander. When the two disagree about how fast a cell responds to a tug, the mechanics wins, and the model that cannot account for it is the model that has to change.

Fuller and Snelson

Tensegrity was named by one man and built by another

Credit for tensegrity is routinely given wrongly, so state it accurately. Buckminster Fuller coined the word, a compression of tensional integrity. The first tensegrity object was built by the sculptor Kenneth Snelson, whose X-Piece dates from 1948, after a summer at Black Mountain College where Fuller had been lecturing. The two later fell out over the credit in public, at a 1959 exhibition at the Museum of Modern Art. Fuller holds United States patent 3,063,521, granted 13 November 1962. Snelson holds United States patent 3,169,611, granted 16 February 1965. Snelson's Needle Tower of 1968 rises 26.5 meters on cables and floating struts, with no continuous rigid frame anywhere in it.

Ingber did not invent tensegrity and has never claimed to. He did something harder to argue with. He took a principle from sculpture and architecture, made it a testable hypothesis about living matter, and then went and tested it for thirty years. The related term biotensegrity, used mostly of the musculoskeletal system, is associated with Stephen Levin and developed on a separate track. Getting these attributions right matters, because a true idea carried on false credits is easy for a critic to throw out whole.

1981 to 1985

The first proposal came from a tumor that could not organize itself

Ingber's first published paper, with Joseph Madri and James Jamieson, appeared in Proceedings of the National Academy of Sciences in June 1981, volume 78, pages 3901 to 3905 (Ingber 1981). The subject was a transplantable carcinoma of the rat pancreas. The tumor cells were cytologically differentiated acinar cells that had lost their epithelial orientation entirely and formed no acini. Yet wherever those same cells made contact with the vasculature, they palisaded, repolarized, and rebuilt a basal lamina. Immunofluorescence told the rest. In normal pancreas, laminin and type IV collagen run in continuous lines around acini and vessels. In the disorganized tumor parenchyma, type IV collagen was not detected at all and laminin appeared only as scattered points. Reorientation happened along linear basal lamina and nowhere else.

Ask what that means. The cells had not lost the instruction. They had lost the surface that carries it. Give them a continuous, mechanically competent matrix to pull against and the architecture reassembles itself. In 1985 Ingber put the mechanism into a title that says the whole thesis out loud: Cells as Tensegrity Structures: Architectural Regulation of Histodifferentiation by Physical Forces Transduced over Basement Membrane. The extracellular matrix was not scaffolding for cells to sit on. It was the other half of a force balance, and the cell reads its own state by pulling on it.

The fibronectin ladder

One molecule, three fates, decided by how hard the matrix pulls back

In 1989 Ingber and Judah Folkman published in Journal of Cell Biology, volume 109, pages 317 to 330 (Ingber and Folkman 1989). They took capillary endothelial cells, held basic fibroblast growth factor constant, and varied only the coating density of fibronectin on the dish. Above roughly 500 nanograms per square centimetre the cells spread and divided. Below roughly 100 they rounded up and detached. In the band between 100 and 500 they shortened, elevated, and assembled themselves into tubular capillary networks within 24 to 48 hours. Time-lapse cinematography caught the retraction that made the tubes form.

Same growth factor. Same cells. Three different fates. The only variable was mechanical: how much resistance the matrix could offer to the cell's own contraction. Force the implication and it is uncomfortable for a purely chemical biology. If one soluble signal produces growth, differentiation or death depending on the tension a cell can generate against its surroundings, then the chemical message is not the instruction. It is one input to a decision that mechanics arbitrates. Ingber and Folkman called it mechanochemical switching. The name is exact.

1993

The year the model finally got its experiment

A structural theory that cannot be broken by an experiment is not worth much. In 1993 Ingber published both halves within months. The Commentary in Journal of Cell Science laid out the rules (Ingber 1993). The paper in Science on 21 May, with Ning Wang and James Butler, supplied the measurement (Wang and Butler 1993). The method was magnetic twisting cytometry. Ferromagnetic beads were coated with specific ligands, allowed to bind receptors on the cell surface, magnetized, then twisted by an applied field while the resulting rotation was recorded. Beads bound to integrin beta 1 induced focal adhesion formation and produced a force-dependent stiffening response. Beads bound to non-adhesion receptors produced nothing of the kind.

Two results carried the argument. Cytoskeletal stiffness, the ratio of stress to strain, rose in direct proportion to the applied stress. And the response required intact microtubules and intermediate filaments as well as microfilaments. That second point is the one that discriminates. A fluid-filled bag stiffens by internal pressure and does not care which filament class you disable. A tensioned network cares about all of them, because removing any element redistributes load across the rest. The authors concluded that integrins act as mechanoreceptors. Ask what a receptor is, then. Not only a chemical lock. A structural anchor through which the outside world reaches in.

The hard wiring

A pull at the surface moves the nucleus

On 4 February 1997, Andrew Maniotis, Christopher Chen and Ingber published in Proceedings of the National Academy of Sciences, volume 94, pages 849 to 854 (Maniotis 1997). They bound microbeads to integrins and pulled, with micromanipulators and micropipettes, and watched what happened inside. Cytoskeletal filaments reoriented. Nuclei distorted. Nucleoli redistributed along the axis of the applied tension field. The effect was specific to integrins, was independent of distortion of the cortical membrane, and was carried by direct linkages between cytoskeleton and nucleus.

The division of labor was clean. Actin microfilaments transferred force to the nucleus at low strain, but the actin gel tore under greater distortion. Intermediate filaments carried force under both conditions and acted, in the authors' phrase, as molecular guy wires that stiffen the nucleus and anchor it in place. Microtubules held the intermediate filament lattice open and stabilized the nucleus against lateral compression. Now force the implication. Nucleoli are where ribosomes are assembled. If a tug at the cell surface rearranges them in the time it takes to pull, then mechanical force has reached the machinery of gene expression without waiting on a diffusing messenger. Tension is not merely registered at the membrane. It arrives at the genome.

Shape decides

Give a cell room to spread and it lives, deny it and it dies

Later in 1997, Christopher Chen, Milan Mrksich, Sui Huang, George Whitesides and Ingber published Geometric control of cell life and death in Science, volume 276, pages 1425 to 1428 (Chen 1997). Using micropatterned substrates, they printed adhesive islands of different sizes and held the total area of cell to matrix contact constant while varying how far the cell could spread across it. Human and bovine capillary endothelial cells were used. Cells permitted to spread grew. Cells confined to small islands entered programmed cell death. The outcome held regardless of which matrix protein or which integrin antibody was used to make the attachment.

That is an unusually clean result. The chemistry of adhesion was matched between conditions. The geometry was not. Ask what decided the fate and the only remaining answer is the cell's ability to extend and to generate tension against its surroundings. The authors proposed geometric control as a fundamental mechanism for developmental regulation, and the proposal has held. The same principle appears at every scale you look at. Form is not stamped onto living parts from outside. It is negotiated between them, through tension, continuously.

The argument

The model was contested in print, and it survived by being measured

This page will not pretend the idea walked in unopposed. In 2000 the Journal of Applied Physiology published Opposing views on tensegrity as a structural framework for understanding cell mechanics, volume 89, issue 4, opening at page 1663, with Steven Heidemann, Phillip Lamoureux and Robert Buxbaum putting the case against (Heidemann 2000). The objections were technical and serious. Do microtubules really bear compression inside a living cell, or do they only look like struts in cartoons? Is prestress genuinely the dominant determinant of stiffness, or can simpler continuum models predict the same curves? A theory that could not answer those questions deserved to be doubted.

The answer came from instruments rather than argument. In 2006, Clifford Brangwynne and colleagues, with Ingber and David Weitz among the nine authors, published in Journal of Cell Biology, volume 173, pages 733 to 741 (Brangwynne 2006). Microtubules do bear large-scale compressive loads in living cells. But they buckle at a characteristic wavelength of 2 to 3 micrometres, far shorter than an isolated filament predicts, because the surrounding elastic cytoskeleton reinforces them laterally along their length. An isolated microtubule has a persistence length near 2 millimetres. Inside a cell it behaves like something a thousandfold less stiff. Neither camp had been entirely right. The network changes what every element in it can do, which is the tensegrity point restated by its critics' own measurement.

Prestress at every scale

The word Ingber reached for was tone

In 2008 he published two reviews that pulled the scales together. Tensegrity-based mechanosensing from macro to micro appeared in Progress in Biophysics and Molecular Biology, volume 97, pages 163 to 179 (Ingber 2008). A short companion piece, Tensegrity and mechanotransduction, appeared in Journal of Bodywork and Movement Therapies, volume 12, pages 198 to 200 (Ingber 2008). The structural claim in both is hierarchical. Bodies are systems within systems, so that a single tension or compression element at one scale is itself a whole tensegrity built of smaller tension and compression elements at the scale below. Organ, tissue, cell, cytoskeleton, molecule. Force applied at the level of the organ is channeled over extracellular matrix and linked integrins and focused down onto focal adhesions, where mechanochemical transduction proceeds.

Then comes the sentence that earns this figure a place in a history of tone. Ingber, writing as a cell biologist about the physics of living structure, names the governing variable, and the word he selects is tone. Here is the unique contribution of this page to the story: Ingber supplies the physical mechanism by which tone stops being a metaphor, because he measured a resting tension that sets the response of a living structure at every scale from organ to genome. Note plainly what he did not do. He did not identify the regulator that sets that tension in a whole organism. That question belongs to the central autonomic network, and it is the next node in this timeline for exactly that reason.

From cell to chip

What the work built, and what this page does not claim

A structural theory earns its keep by making things. The Wyss Institute for Biologically Inspired Engineering at Harvard was founded in 2009 on a gift of 125 million dollars from Hansjörg Wyss, at the time the largest philanthropic gift in Harvard's history, with Ingber as founding director. He holds the Judah Folkman Professorship of Vascular Biology at Harvard Medical School and Boston Children's Hospital, and has authored more than 500 publications and more than 200 patents. In 2010, Dongeun Huh, Benjamin Matthews, Akiko Mammoto, Martín Montoya-Zavala, Hong Yuan Hsin and Ingber published the human lung on a chip in Science, volume 328, pages 1662 to 1668 (Huh 2010). The device reconstitutes the alveolar and capillary interface, and it breathes. Cyclic mechanical strain accentuated the toxic and inflammatory response to silica nanoparticles, increased uptake by both epithelium and endothelium, and drove their transport into the microvascular channel below. The same effects of physiological breathing on nanoparticle absorption were then confirmed in whole mouse lung. Hold the chemistry constant, add rhythmic motion, and the biology changes.

Ingber studies cells, tissues and engineered devices. He has not endorsed any clinical method, school or profession, and nothing written here should be read as his endorsement of one. His documented claims are that living cells are prestressed tensegrity structures, that integrins act as mechanoreceptors, that force applied at the cell surface reaches the nucleus through discrete molecular linkages, that cell shape and matrix mechanics govern growth and death, and that the level of prestress sets responsiveness at every scale of a body. What this site adds is a reading, and the reading should be labeled as one. We hold that the living state of the nervous system is expressed as tissue tension at every scale, and that this single variable is the one worth tracking across the whole history of the healing arts. Ingber did not say that. He supplied the floor it stands on, which is the demonstration that tension in living tissue is a real, measurable, regulating quantity rather than a figure of speech. A model that has to be defended by overstating its sources is a model in trouble, and this one does not need the help.

Ingber and the model

What cellular tensegrity gave the Unified Model of Tone

The Unified Model of Tone singles Ingber out for evidentiary strength. The model holds that a living body is held in shape by distributed tension rather than by stacked compression, and that tone is the state of that tension. Stephen Levin argued that case for the musculoskeletal system over decades. Ingber established the same principle at the scale of the single cell, where the evidence is strongest and least disputed. The cell is a tensegrity structure, with microtubules as compression elements and the actin cytoskeleton as tension elements. That measurement is what the model builds on, and it is the reason the model can claim one architecture across every scale of the body.

Mechanical force applied at the tissue scale is transmitted through the soft tissue continuum to cell membranes. There it activates mechanotransduction pathways that change ion channel behavior, second messenger cascades and gene expression. This is how manual therapy reaches the cell. Touch propagates through the mechanical continuum all the way to the nucleus, whatever tissue first received the contact. The cytoskeleton's mechanical state reaches the nucleus and influences which genes are expressed, and the 1997 bead experiment is where that last step was measured rather than inferred.

Here the model takes a step Ingber did not take. Cellular tensegrity, as it has been argued in the literature, is a claim about load. The model adds a claim of its own. The tension network is the body's geometric self-registration, which is to say that its integrated tensional state is how the body knows its own shape. So a touch does not only move tissue. It changes the account the body is keeping of where it is.

Tone is fractal, and the single cell is where that claim is easiest to check. Neurophysiology already describes the cellular case and calls it the central integrative state. It is the running sum of every excitatory and inhibitory influence converging on a neuron at a given moment, the baseline that decides how that neuron answers the next signal. That much is established. What the model adds is the identification. The central integrative state is tone read at the scale of a single cell, and tone is that same state read at the scale of a whole organism. The construct does not change as the resolution changes. Only the instrument does.

The measurement is his and the extension is the model's. Ingber established that prestress governs how a living structure answers force at every scale, and he stopped there. The claim that this same tension is the body's running record of its own geometry belongs to the model, and it is stated here as a claim rather than as his finding.

What the record shows

Cellular tensegrity in seven dated findings

• 1993. Ingber set out cellular tensegrity in the Journal of Cell Science, volume 104, pages 613 to 627 (Ingber 1993). Microtubules bear compression, the actin cytoskeleton carries tension, and the cell holds its shape by the balance between them.
• 1993. Wang, Butler and Ingber, Science volume 260, pages 1124 to 1127 (Wang and Butler 1993), bound magnetic beads to integrin beta 1 receptors and twisted them. Cytoskeletal stiffness rose in direct proportion to the applied stress, which is what a tensioned network does and a pressurized bag does not.
• 1997. Maniotis, Chen and Ingber, Proceedings of the National Academy of Sciences volume 94, pages 849 to 854 (Maniotis 1997): pulling integrin-bound microbeads distorted nuclei and moved nucleoli along the axis of the tension field. Force reaches the machinery of gene expression by a continuous mechanical path.
• 1997. Chen, Mrksich, Huang, Whitesides and Ingber, Science volume 276, pages 1425 to 1428 (Chen 1997), showed geometric control of cell life and death. Give a cell room to spread and it lives, deny it the room and it dies, with the chemistry held constant.
• 2006. Brangwynne and colleagues, Journal of Cell Biology volume 173, pages 733 to 741 (Brangwynne 2006), measured microtubules carrying compressive loads in living cells and buckling at 2 to 3 micrometres, because the surrounding network reinforces them laterally. The dispute of 2000 was settled by refinement.
• 2008. In Progress in Biophysics and Molecular Biology, volume 97, pages 163 to 179 (Ingber 2008), Ingber named the governing variable tone. At every size scale, the level of prestress in these networks governs how a living network answers stress, mechanically and biochemically.
• 2010. Huh, Matthews, Mammoto, Montoya-Zavala, Hsin and Ingber published the human lung on a chip in Science, volume 328, pages 1662 to 1668 (Huh 2010). Cyclic mechanical strain increased the toxic and inflammatory response to silica nanoparticles, with the chemistry unchanged.