From Cell Differentiation to Cell CollectivesBacillus subtilis Uses Division of Labor to Migrate

Jordi van Gestel, Hera Vlamakis, Roberto Kolter•View original

OverviewBalancedadam voice

A single bacterial colony, barely a millimeter across, is doing something a lone cell could never manage — moving. Not swimming or tumbling through liquid on spinning flagella, but spreading across a dry solid surface as a collective and building a living machine out of two specialized cell types that have never encountered a central coordinator, yet somehow cooperate with remarkable precision. This is Bacillus subtilis solving the problem of migration by inventing division of labor from scratch on a petri dish, in real time. The puzzle van Gestel, Vlamakis, and Kolter set out to crack is this: how does a bacterium migrate across solid surfaces when it cannot use flagella, the molecular propellers most bacteria rely on? The team grew wild B. subtilis on a medium that explicitly blocks swimming and swarming, both flagellum-dependent, yet the colony still spreads. They call this flagellum-independent movement sliding motility. It matters ecologically because soil microbes face patchy resources and fierce local competition. Being able to cross a dry surface to reach new nutrients, without liquid to swim through, is a genuine survival advantage. Watch a colony of B. subtilis on this medium for a day and you see two distinct phases unfold. First, thin branching extensions called dendrites reach outward. Then a different morphology takes over: broader, petal-shaped outgrowths push the colony edge further. Van Gestel and colleagues link this two-stage pattern to a precise temporal sequence in gene expression, first a rise in the srfA operon for surfactin synthesis, then a later sharp increase in the tapA operon for matrix production. Those two expression peaks map almost exactly onto the two phases you can see with your eye. So who are the players? The study identifies two distinct, mutually exclusive cell types. Surfactin-producing cells express srfA and make surfactin, a biosurfactant. Matrix-producing cells express tapA and secrete two things: an extracellular polysaccharide called EPS, and the structural protein TasA. Together, EPS and TasA form the scaffold that holds chains of cells together. Apart, neither type accomplishes much. Mutants lacking surfactin cannot expand the colony. Mutants lacking EPS cannot expand at all. Mutants lacking TasA expand, but poorly. The chimera experiments are where the division of labor becomes unmistakable. When van Gestel and colleagues mixed pairs of expansion-deficient mutants, one lacking surfactin and one lacking the matrix, colony expansion recovered, with some chimeras exceeding wild type. But the recovery was asymmetric: when the starting mix had many matrix-deficient cells and few surfactin-deficient cells, expansion stayed low. Flip the ratio — mostly surfactin-deficient and few matrix-deficient — and expansion soared. The conclusion is crisp: a small number of surfactin producers is enough to enable expansion, but a small number of matrix producers is not. Matrix production is the rate-limiting ingredient. The spatial story is just as striking. Fluorescent reporter imaging shows that matrix-producing cells organize themselves into tightly aligned chains, structures the authors name van Gogh bundles, an allusion to the swirling alignment that recalls Van Gogh's brushwork. Surfactin-producing cells tend to surround these bundles rather than join them. EPS is absolutely required for bundle formation; cells that cannot make EPS never form the aligned chains. TasA is not strictly required, but it fine-tunes the bundles from the inside. Localized to pole-to-pole contact points between cells, it affects how rigid the bundle is, and that rigidity turns out to matter enormously for what happens next. What the bundles actually do at the colony edge is the most dramatic part of the story. Time-lapse microscopy shows the bundles forming large filamentous loops — sometimes a few millimeters long — at the colony margin. Those loops then push themselves away from the colony, driving expansion. No single cell is doing this. The directed outward movement emerges entirely from local cell-to-cell interactions and simple growth, elongation, division, and turning. No central coordinator. No external signal orchestrating the charge. Just cells sticking to their neighbors, growing, and folding. The alignment inside van Gogh bundles is quantifiably extreme. The average angle between neighboring cell segments inside a bundle is 4.5 degrees, compared to 21 degrees in ordinary single-cell populations. Cells inside bundles are also significantly longer than their single-cell siblings, a difference confirmed by a two-sample t-test with a p-value below one in ten thousand. These are not subtle differences. The bundles are genuinely distinct structures, and their geometry turns out to control the pace of the whole colony. That connection between bundle geometry and migration rate is what the team's mathematical model was built to reveal. The model is deliberately simple: a chain of cells in a two-dimensional space, where at each time step a randomly chosen cell either elongates, divides, or turns. A bending-rigidity parameter governs how strongly neighboring cells resist changes in their relative angle — high rigidity means the chain stays straighter. A maximal cell-length parameter determines when a cell divides. From these three local rules, the model reproduces what the microscopy shows: growth generates small undulations; those undulations, under continued compression, fold into loops; loops expand outward and drive colony advance. The key insight lies in how the parameters shift the outcome. Higher bending rigidity and longer maximal cell length both reduce folding, and reduced folding means faster migration. The bundle shape is not just aesthetics; it is the migration engine's gear setting. The experiment that validates this prediction involves the tasA mutant. TasA localizes to pole-to-pole contacts and is proposed to increase bending rigidity. Lose TasA, and the model predicts loops should fold more and migrate more slowly. That is exactly what happens. The team measured fold angles from five hundred and nine concatenated line segments in tasA mutant loops and six hundred and twenty-five from wild-type loops. The tasA loops showed more and stronger folds — confirmed by a Mann-Whitney U test with a p-value again below one in ten thousand, and a W statistic of 206,266. TasA mutant colonies showed reduced expansion in the petal phase. The model predicted it; the experiment confirmed it. Step back from the petri dish and the significance becomes clear. Van Gestel, Vlamakis, and Kolter have shown that two differentiated cell types, interacting through nothing more than physical contact and secreted molecules, can produce collective migration across a solid surface — a capability neither type possesses alone. Surfactin producers lower friction. Matrix producers build the aligned scaffolds. The scaffolds fold into loops. The loops push the colony forward. Each step follows from the previous one, and none requires anything more than local rules. The authors connect this explicitly to the broader question of why multicellularity evolved, and the connection is worth sitting with. The prevailing assumption is that multicellularity requires complexity: differentiated tissues, signaling cascades, developmental programs honed over millions of years. What this system shows is that the core logic of multicellularity, cells doing different jobs in service of a shared outcome, can emerge from startlingly simple ingredients. Two cell types, a biosurfactant, a structural protein at the poles of touching cells, and a surface. The simplicity is not a limitation of the B. subtilis system; it is the argument. If collective behavior with a clear ecological payoff can arise from this little, then the repeated, independent origins of multicellularity across the tree of life become less mysterious. The first steps toward being many rather than one may not have required much at all — just the right division of labor and somewhere to go. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.

Van Gestel and colleagues link this two-stage pattern to a precise temporal sequence in gene expression, first a rise in the srfA operon for surfactin synthesis, then a later sharp increase in the tapA operon for matrix production. Those two expression peaks map almost exactly onto the two phases you can see with your eye. So who are the players? The study identifies two distinct, mutually exclusive cell types. Surfactin-producing cells express srfA and make surfactin, a biosurfactant. Matrix-producing cells express tapA and secrete two things: an extracellular polysaccharide called EPS, and the structural protein TasA. Together, EPS and TasA form the scaffold that holds chains of cells together. Apart, neither type accomplishes much. Mutants lacking surfactin cannot expand the colony. Mutants lacking EPS cannot expand at all. Mutants lacking TasA expand, but poorly. The chimera experiments are where the division of labor becomes unmistakable. When van Gestel and colleagues mixed pairs of expansion-deficient mutants, one lacking surfactin and one lacking the matrix, colony expansion recovered, with some chimeras exceeding wild type. But the recovery was asymmetric: when the starting mix had many matrix-deficient cells and few surfactin-deficient cells, expansion stayed low.

Flip the ratio — mostly surfactin-deficient and few matrix-deficient — and expansion soared. The conclusion is crisp: a small number of surfactin producers is enough to enable expansion, but a small number of matrix producers is not. Matrix production is the rate-limiting ingredient. The spatial story is just as striking. Fluorescent reporter imaging shows that matrix-producing cells organize themselves into tightly aligned chains, structures the authors name van Gogh bundles, an allusion to the swirling alignment that recalls Van Gogh's brushwork. Surfactin-producing cells tend to surround these bundles rather than join them. EPS is absolutely required for bundle formation; cells that cannot make EPS never form the aligned chains. TasA is not strictly required, but it fine-tunes the bundles from the inside. Localized to pole-to-pole contact points between cells, it affects how rigid the bundle is, and that rigidity turns out to matter enormously for what happens next. What the bundles actually do at the colony edge is the most dramatic part of the story. Time-lapse microscopy shows the bundles forming large filamentous loops — sometimes a few millimeters long — at the colony margin. Those loops then push themselves away from the colony, driving expansion. No single cell is doing this. The directed outward movement emerges entirely from local cell-to-cell interactions and simple growth, elongation, division, and turning. No central coordinator.

No external signal orchestrating the charge. Just cells sticking to their neighbors, growing, and folding. The alignment inside van Gogh bundles is quantifiably extreme. The average angle between neighboring cell segments inside a bundle is 4.5 degrees, compared to 21 degrees in ordinary single-cell populations. Cells inside bundles are also significantly longer than their single-cell siblings, a difference confirmed by a two-sample t-test with a p-value below one in ten thousand. These are not subtle differences. The bundles are genuinely distinct structures, and their geometry turns out to control the pace of the whole colony. That connection between bundle geometry and migration rate is what the team's mathematical model was built to reveal. The model is deliberately simple: a chain of cells in a two-dimensional space, where at each time step a randomly chosen cell either elongates, divides, or turns. A bending-rigidity parameter governs how strongly neighboring cells resist changes in their relative angle — high rigidity means the chain stays straighter. A maximal cell-length parameter determines when a cell divides.

From these three local rules, the model reproduces what the microscopy shows: growth generates small undulations; those undulations, under continued compression, fold into loops; loops expand outward and drive colony advance. The key insight lies in how the parameters shift the outcome. Higher bending rigidity and longer maximal cell length both reduce folding, and reduced folding means faster migration. The bundle shape is not just aesthetics; it is the migration engine's gear setting. The experiment that validates this prediction involves the tasA mutant. TasA localizes to pole-to-pole contacts and is proposed to increase bending rigidity. Lose TasA, and the model predicts loops should fold more and migrate more slowly. That is exactly what happens. The team measured fold angles from five hundred and nine concatenated line segments in tasA mutant loops and six hundred and twenty-five from wild-type loops. The tasA loops showed more and stronger folds — confirmed by a Mann-Whitney U test with a p-value again below one in ten thousand, and a W statistic of 206,266. TasA mutant colonies showed reduced expansion in the petal phase. The model predicted it; the experiment confirmed it.

Step back from the petri dish and the significance becomes clear. Van Gestel, Vlamakis, and Kolter have shown that two differentiated cell types, interacting through nothing more than physical contact and secreted molecules, can produce collective migration across a solid surface — a capability neither type possesses alone. Surfactin producers lower friction. Matrix producers build the aligned scaffolds. The scaffolds fold into loops. The loops push the colony forward. Each step follows from the previous one, and none requires anything more than local rules. The authors connect this explicitly to the broader question of why multicellularity evolved, and the connection is worth sitting with. The prevailing assumption is that multicellularity requires complexity: differentiated tissues, signaling cascades, developmental programs honed over millions of years. What this system shows is that the core logic of multicellularity, cells doing different jobs in service of a shared outcome, can emerge from startlingly simple ingredients. Two cell types, a biosurfactant, a structural protein at the poles of touching cells, and a surface. The simplicity is not a limitation of the B. subtilis system; it is the argument. If collective behavior with a clear ecological payoff can arise from this little, then the repeated, independent origins of multicellularity across the tree of life become less mysterious.

The first steps toward being many rather than one may not have required much at all — just the right division of labor and somewhere to go. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.

Related lectures