Science1 distinct publisher3 min readPublished
In an aligned fluid, a Caltech group reports in PNAS, colonies grow as chains one cell wide that run straight and then kink in a single tight spot, which hands part of colony architecture to the medium.
The Scientist · Science desk

Compiled by The ScientistSomething wrong?How this is made
The mechanism rests on an energy argument. Picture matches packed in a matchbox as the aligned molecules and the bacterium as a pencil in a very large box, which is roughly the size ratio, since the cell is much larger than the liquid crystal molecules [9]. The matches resist being bent out of their orientation and push back, forcing the pencil to lie along with them, which is bending elasticity [10]. A pencil that grows and divides end-on into more pencils, all in one aligned chain, is the cheapest arrangement on offer [15]. The cost turns up at the tip, where the matches must bend to make room for the chain's width, and it climbs steeply if the chain thickens [11]. In the model, then, it is the surrounding molecules that do the work: they push on the cells, and the chain stays one cell wide, without any need for the cells themselves to hold the line [12].
The odd part is the failure mode. A slender beam under compression is textbook mechanics, but this beam replicates itself, which is why Datta calls it a weird version of an old problem [7]. Instead of bowing in an arc along its length, the bacterial chain kinks locally, with a short section bending hard [5]. The group credits the same bending elasticity, along with the liquid crystal's high viscosity [13]. Applied mathematicians at Wisconsin-Madison and UNC Chapel Hill built the model behind that account [8].
What carries the "medium sets the architecture" reading is a comparison across two studies rather than one paired experiment. In polymeric fluids whose molecules point every which way, the same lab saw cables several cells wide, intertwined into a living gel [6]; in the aligned fluid, chains one cell wide [4]. The account of the new work names species only for the earlier polymeric-fluid experiments [17], so a reader cannot confirm from it that the same organisms were grown in both media. That gap is the join in the argument.
The account also carries no counts. Missing are the number of chains, the length reached before buckling, the fraction that buckled, and the curvature at the kink. The one quantitative output claimed is the degree to which a chain aligns with the fluid, predicted from established liquid crystal physics [12], and alignment is a measurable thing, so that is the claim with a handle on it.
The thing this does not tell you is whether any of it changes an infection. Colony shape and its effect on survival, growth and resistance to treatment is where the researchers put their new questions, not their results [14]. The pull toward a clinical reading is obvious: aligned fluids include biofilm matrices and the mucus lining of the airways and gut [3], and two of the three species named in the earlier work are described as human pathogens, one causing cholera and one common in hospitalised or immunocompromised patients [16]. Drug penetration and killing are not part of what this work measures.
My reading, with its condition attached: the physics looks sound and the biology is untested. For anyone modelling bacterial growth in mucus, the fluid's bending elasticity now looks like a parameter to carry rather than a constant to assume [12].
Ranked by verification strength, evidence, and original report placement.
In an aligned liquid crystal fluid, bacteria build single-cell-wide chains that grow and lengthen in relatively straight lines until they suddenly buckle.
A compressed steel beam bends gradually in an arc-like fashion along its entire length, whereas the buckling of the bacterial beam in liquid crystal is localized, with only a small section bending very tightly.
Sujit Datta is a professor of chemical engineering, bioengineering and biophysics at Caltech.
Datta and his former graduate student Sebastian Gonzalez La Corte studied bacteria growing in liquid crystal fluids, whose molecules are elongated and all point in the same direction; the findings are described in a paper in PNAS.
The aligned liquid crystal state is known to occur in some biological fluids, such as certain biofilm matrices and the mucus linings of the airways and gut, yet laboratory experiments typically study bacteria in fluids whose constituents have no preferred direction.
Datta's group previously found that in polymeric fluids with randomly arranged molecules, several commonly studied species including Escherichia coli, Vibrio cholerae and Pseudomonas aeruginosa grow into long cables several cells wide that intertwine and form a kind of living gel.
Distinct publishers with included, body-backed reporting in this cluster.
phys.org
1 article · September 1, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
science
Simulations show the fastest earthquake ruptures shake harder than hazard models assume1 distinct publisher
science
Caltech model says fast-dividing tissue may be proofreading, not just accumulating risk1 distinct publisher
product
Geologic hydrogen's firmest numbers come from mine vents, not reservoirs1 distinct publisher
science
A phage kinase with no target list: EMBL finds one enzyme that breaks several bacterial defences1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
Peer-reviewed upstream, single-channel downstream
The underlying work is a PNAS paper with confocal imaging and a model built alongside mathematicians at two universities, which is real weight. It reaches readers, though, through one Phys.org write-up that quotes nobody but the principal investigator and reproduces no data: the phrase 'quantitatively predict' arrives without a single quantity attached to it. Strong provenance, thin verification.
Nothing yet to take up
This is a bench result about the shape of growing colonies. Our sourcing shows no one outside the lab using it, no follow-on study, no clinical or industrial application, and the reporting itself says the biological consequences are unknown. Scoring uptake here would be inventing it.
Undersold, slightly
Rare for a university-sourced science story: the restraint runs ahead of the finding. The strongest sentence in this reporting is buried near the end, where Datta says arrangement governs communication, nutrients and antibiotic tolerance and therefore what sets colony shape matters. Around it, the piece keeps hedging that implications are unknown. Meanwhile the actual claim — an aligned medium overriding the organism's own architecture, and a beam that compresses itself by dividing — is stated more plainly than it is sold.
One lab telling its own story
The only voice is the senior author, the only metaphors are his, and the contrast case that makes the new result look surprising is his group's earlier polymeric-fluid work. That is the shape of institutional promotion rather than of a commercial pitch — nothing is being sold, no product or position rides on it. What is missing is the disclosure furniture: no funders, no competing-interest note, and no outside mechanician asked whether decades-old buckling theory really fails here.
Firm on mechanism, empty on consequence
We would stand behind the physics as reported: the observation is concrete, the energy argument is internally coherent, and it has been through PNAS review. We would not stand behind anything downstream of it. With one publisher, one interviewee, no species named for the key experiments and no numbers, the story is credible where it is careful and unresolved everywhere it gestures at mucus, biofilms or antibiotic tolerance.