The Exopolysaccharide Matrix Modulates the Interaction between 3D Architecture and Virulence of a Mixed-Species Oral Biofilm

Picture dental plaque not as slime, but as a tiny city with scaffolding, alleys, and neighborhoods. The scaffolding is the exopolysaccharide matrix, or EPS, which is sticky glucans and other polymers that glue everything together. This matrix changes what life is possible inside that city. Xiao and colleagues, along with Kramer and co-workers, asked a simple and consequential question: when this scaffold thickens and organizes in three dimensions, does it flip a quiet community into a destructive one? Their answer is yes, and they show how. They built a tooth-like world on saliva-coated hydroxyapatite discs, which are essentially enamel stand-ins dressed in real saliva, and let three common oral residents settle: Streptococcus mutans, Actinomyces naeslundii, and Streptococcus oralis. Then they did something that sounds trivial but turns out to be the plot twist: they fed the biofilms differently. Sometimes the medium carried a lot of sucrose, sometimes just a little, and sometimes it had glucose instead. While the cells grew, the team watched with confocal microscopes, measured gene transcripts with reverse transcription quantitative polymerase chain reaction, or RT-qPCR, profiled proteins using Multidimensional Protein Identification Technology, or MudPIT, proteomics, and cleverly wove a fluorescent pH sensor into the matrix so they could see the acidity inside the living biofilm. Here's the big architectural reveal. When they switched to 1 percent sucrose, the matrix didn't just get thicker—it organized. Within about 43 hours, a patchy EPS layer appeared on the saliva-coated surface. By 67 hours, the cells were packed into three-dimensional islets, each one a dense knot of bacteria wrapped in glucans. At 115 hours, those islets were bigger and more interconnected, forming what the teams call EPS-microcolony complexes. Two types of these structures showed up: some anchored right on the surface layer; others tethered to the matrix up near the fluid interface. With sucrose, the matrix itself exploded in abundance—up to ten times more EPS than with 1 percent glucose by 115 hours—and the entire biofilm got thicker. With only 0.1 percent sucrose, you saw a wispy EPS layer and few of those three-dimensional islets. With glucose, you got a pretty flat, homogeneous sheet of cells and no real EPS-rich architecture. That kind of build doesn't happen by accident. In these mixed-species biofilms, sucrose activates the Streptococcus mutans assembly crew, specifically two secreted glucosyltransferases, GtfB and GtfC. Between 43 and 67 hours, their transcripts jump, and proteomics shows both enzymes are more abundant at that early window when the matrix is laying down its foundation. In mixed communities, those genes are even more active than in Streptococcus mutans alone, indicating that neighbors are part of the signal. The team points to the LuxS and autoinducer-2 quorum-sensing system, which provides autoinducer-2, or AI-2. This boosts the gtfB and gtfC genes, and luxS gene mutants regain that boost if AI-2 comes from other bacteria. Notably, Streptococcus oralis didn't supply the needed rescue in their tests, suggesting that not every neighbor speaks the same chemical language. The division of labor is neat. GtfC loves the pellicle, which is the protein film on the tooth, and seeds the first EPS layer right on the surface. GtfB prefers cell surfaces and makes highly insoluble, rigid glucans rich in one, three linkages. Lay GtfC down first, and you create sticky real estate; bring GtfB in, and you cement the residents into tough, EPS-enmeshed clusters. By 115 hours, roughly 80 percent of the cells are co-localized with EPS. In other words, most of the city's inhabitants are living inside the scaffold, not just next to it. Break the build crew, and the city never rises. Remove Streptococcus mutans from the consortium, or knock out both gtfB and gtfC, and those EPS-microcolonies don't form. Single mutants misbuild in different ways: without GtfB, you can get an initial surface layer, but you can't finish the islets; without GtfC, you struggle to lay that first sticky layer, and only a few microcolonies emerge. Even after the structures exist, you can pull the keystone and watch them fall. Treat mature biofilms with mutanase, which cuts insoluble one, three-linked glucans, and the matrix collapses along with the cellular biomass. There are other players—glucan-binding proteins like GbpC and GbpB, and even extracellular DNA likely pitch in—but the main architectural beams are those GtfB and GtfC made glucans. Now, architecture isn't just about looks; it changes the chemistry of the neighborhood. If you only measure the surrounding liquid, 1 percent sucrose and 1 percent glucose both drive the culture medium acidic, ending up around pH 4.2 to 4.7, while low sugar keeps it closer to neutral at pH 6.1 to 6.2. But Kramer and colleagues demonstrate that this is the view from 30,000 feet. Inside the intact biofilm, the story is pockets and gradients. Using a ratiometric Lysosensor woven into the EPS, they mapped pH in three dimensions and found that acidity doesn't spread evenly. It collects. Specifically, it pools inside the EPS-enmeshed microcolonies and at the spots where those islets touch the tooth-like surface. Many of those interfaces sit below pH 5.5, which is the clinical line where enamel dissolves quickly. Regions without three-dimensional microcolonies don't show those low values, and you can see nearly neutral zones right at the fluid-facing edges of the islets. Here's what's striking. Even if you flush the environment with neutral buffer, the pockets don't bounce back quickly. It took more than 120 minutes before those compartments began to noticeably neutralize. That tells you the EPS-rich architecture is restricting diffusion enough to trap acids, at least for a while. And size matters. When they measured a lot of individual islets, larger microcolonies tended to have lower pH at the surface interface. The relationships were modest but clear: pH tracked with both diameter and height, with coefficients of determination around 0.26 and 0.37, respectively. Bigger, taller islets correlate with worse chemistry underneath. Why does that matter? Because the real damage in caries starts at the surface where enamel meets the biofilm. If the lowest pH sits right at that interface, you've concentrated dissolution at exactly the vulnerable spot. That's a physical mechanism for targeted demineralization: build compartments that hold acid, press them against the tooth, and erode locally even if saliva outside is neutral. It also connects to a broader ecological story—these compartments are not just corrosive; they're selective. Feed the system sucrose, and you don't just get more matrix; you reshuffle who thrives. At baseline on those saliva-coated discs, Streptococcus mutans starts as the underdog. After sucrose, it rises fast and becomes dominant by 115 hours. The numbers make the point: under 1 percent sucrose, Streptococcus mutans reached roughly one times ten to the eleventh colony-forming units, while under sugar limitation it was closer to six point seven times ten to the ninth. Glucose, in contrast, didn't build the same architecture, and the population skewed toward Streptococcus oralis. In other words, the scaffold you build changes the roster of who lives there, favoring acidogenic, acid-tolerant, glucan-binding organisms. That scaffold also changes how the community responds to attack. When they pulsed the biofilms with 0.12 percent chlorhexidine, a standard antiseptic, location determined fate. Fewer than 10 percent of the cells inside EPS-microcolony complexes died after 15 minutes, while more than 70 percent of the cells outside those structures were killed. If you remove the architecture—either by growing in glucose so those islets never form, by using gtfB and gtfC mutants, or by digesting the matrix with mutanase—the tolerance vanishes, and the biofilm becomes highly susceptible. That pattern suggests diffusion-limited protection inside the EPS compartments rather than a uniform "EPS makes you tough" effect. If you want one line that ties architecture to disease, take it from the animal side. In a rodent caries model, gtfB and gtfC mutants, which are strains that can't build this matrix, are essentially non-virulent, and anti-GtfB antibodies reduce carious lesions. The build really is the weapon. A quick nod to the toolbox that made these pictures and conclusions possible. The teams grew intact, mixed-species biofilms in a saliva-conditioned environment, then paired three-dimensional confocal imaging with an in-matrix, ratiometric pH dye to see architecture and chemistry together in real time. They quantified microcolony size with image analysis tools like COMSTAT, used RT-qPCR and MudPIT proteomics to catch the enzymes in the act, and ran statistics with multiple-comparison corrections, such as Tukey-Kramer HSD for group contrasts and Pearson tests for the pH-size correlations. None of that is flashy on its own. But together, it lets you see how a sugar molecule is not just fuel; it's a contractor hiring enzymes, pouring concrete, and changing the neighborhood. So what's the through-line? Sucrose acts as a molecular lever. It flips on Streptococcus mutans' GtfB and GtfC, which lay down an EPS scaffold in two layers—GtfC on the pellicle and GtfB on the cells—turning a flat lawn of bacteria into a city of microcolonies. That three-dimensional architecture traps acids in localized pockets, with the lowest pH right where enamel dissolves fastest, and the pockets persist even when the outside world looks benign. The same architecture favors aciduric species and shelters residents from antimicrobials. Break the scaffold—genetically or enzymatically—and the city can't fortify itself or corrode its host. If you're thinking ahead to the clinic, here's a careful, grounded take. These studies suggest two practical angles. One is measurement. Three-dimensional, in situ pH mapping shows you the risky neighborhoods, not just the average acidity. That could become a way to assess caries risk by structure and chemistry, not just by who's there. The second is intervention at the build stage. Targeting GtfB and GtfC, or disrupting glucan-cell interactions, dismantles the compartments that make plaque both corrosive and drug-tolerant. None of that replaces brushing or diet changes. It just adds a finer lever: don't just starve the city—reshape it.