Science1 distinct publisher2 min readPublished
University of Oregon biologists rebuilt lactoferrin's bacteria-killing peptide across 160 million years. The strongest candidates came from recent ancestors; the oldest one let bacteria heal.
The Scientist · Science desk
Compiled by The ScientistSomething wrong?How this is made
The bacteria that repaired themselves and kept growing did the study a favour. That outcome separates permeabilisation from killing [5], and it says something specific to anyone designing a peptide drug: a molecule can damage the membrane and still lose. How much a peptide perturbs a bilayer is not the endpoint worth optimising.
The arithmetic on the potency gradient is less flattering than the 160-million-year figure implies. Take "a few million years" as about five, and the strongest variants sit in roughly the last 3 percent of the reconstructed span [12]. The deep nodes are also the least constrained. A reconstructed ancestor is a maximum-likelihood inference from living sequences [4], not a fossil, and the further back the node, the less of the original signal survives in the alignment to pin it down. The most expensive part of the reconstruction produced the molecule the bacteria tolerated.
That does not make the approach less interesting, it relocates the value. Ancestral sequence reconstruction was pioneered by Joseph Thornton, then at Oregon [c4b], as a way of testing evolutionary hypotheses, and each node it yields is a sequence some living animal once carried and survived with. As a source of starting points that is a stronger prior than a randomised library, and the slope itself is data: it marks where along the lineage activity was rising, and therefore which substitutions to interrogate.
What is missing is the pharmacology. The tests covered Pseudomonas aeruginosa, Staphylococcus aureus, Escherichia coli and Streptococcus [7], and the university's account carries no concentrations at all [13]. A twofold edge over the modern human peptide and a hundredfold edge read identically in a press release and are different projects in a lab. Nothing released addresses what these peptides do to host cells at the doses that kill bacteria.
Generalisation has a limit too. Lactoferrin's main job is withholding iron from pathogens [10]; the hole-punching peptide is an addition, and none of lactoferrin's close protein relatives kill bacteria [8], which places the origin of the trait inside the mammalian lineage. Families like that, invented recently and still being tuned, are where ancestral nodes should differ enough to justify synthesising them. A deeply conserved family would return a flat series and a wasted year.
Matt Barber, the senior author, says he is "definitely interested" in whether resurrected or engineered peptides could be used "as therapeutics down the road" [9]. Down the road is doing the work in that sentence. What exists now is a sequence series with a measured activity gradient and one negative result from the end of the Jurassic [2].
Ranked by verification strength, evidence, and original report placement.
In laboratory tests, the researchers found that some of the extinct peptides were more potent against drug-resistant bacteria than some of their present-day counterparts.
Peptides of later mammalian ancestors, a few million years old, displayed progressively stronger antimicrobial activity, sometimes outperforming the modern human versions.
University of Oregon biologists resurrected prehistoric proteins up to 160 million years old carrying natural antimicrobial properties; the work was described in a paper published in PLOS Biology on Aug. 25.
About 160 million years ago, near the end of the Jurassic Period, the ancestor of all placental mammals emerged, and so did lactoferrin, the protein at the centre of the study.
Lead author Titas Sil compared lactoferrin gene sequences in present-day organisms such as humans and cows, mapped their evolutionary relationships, and statistically inferred the most likely sequences of their common ancestors, reaching back about 160 million years; the predicted genes were then synthesised and the proteins regenerated in cells.
Ancestral sequence reconstruction was pioneered by Joseph Thornton, a former University of Oregon scientist whose previous lab space now houses Barber's group.
Follow any of these and your For You feed starts watching them — no settings page required.
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 but qualitatively reported
There is a real peer-reviewed artifact (PLOS Biology) behind a mechanistically coherent story: an established technique (ancestral sequence reconstruction), synthesis and expression of the inferred proteins, a named four-pathogen test panel, and a specific structural explanation (a single amino-acid change) for the potency shift. Evidence is held down because the only supplied account reports potency purely qualitatively - no concentrations, minimum inhibitory concentrations or host-cell toxicity values - results are in vitro with no animal data, and no independent replication or outside expert assessment appears in the cluster.
Publication-stage only
The only adoption-adjacent event in the supplied material is the research publication itself. There is no product, no clinical trial of the resurrected peptides, no company, license, partnership or deployment, and the authors state drug development from these peptides is unlikely to be immediate because the molecules are structurally unstable and rapidly degraded in the body.
Mildly overstated toward the therapeutic frame
Headline and lede framing ('resurrect ancient proteins, offering new antibiotic leads', extinct peptides beating present-day counterparts) runs ahead of what the account documents: qualitative in vitro comparisons with no numeric endpoints, no in vivo work, and the oldest reconstruction actually failing as bacteria repaired the damage. The gap is moderate rather than severe because the same article carries the authors' own caveats about peptide instability and non-immediate drug development, and because the strongest performers come from a very recent sliver of the advertised 160-million-year span.
Institutional promotion of own faculty and technique
The single account is a university-communications style piece about University of Oregon researchers, quoting only the paper's two authors, with no external voices. It also foregrounds institutional lineage - crediting the pioneering of ancestral sequence reconstruction to a former UO scientist whose lab space now houses the group - and closes with a therapeutics ambition of the kind that supports future grant and translational funding. No conflicts, funding sources or commercial interests are disclosed in the supplied material, so the incentive read rests on authorship and framing rather than declared stakes.
Single-source, single-publisher
Confidence is limited structurally: one publisher, one institutionally sourced item, no independent corroboration, and an incomplete DOI in the publication details. The underlying claims are internally consistent and peer-reviewed, which keeps confidence from being lower, but nothing in the cluster allows verification of the potency comparisons or triangulation against other reporting.
science
A phage kinase with no target list: EMBL finds one enzyme that breaks several bacterial defences1 distinct publisher
science
Half the resistance genes in livestock manure also show up in 875 wild farm mice1 distinct publisher
science
NASA puts survivable lunar microbe refuges at bootprint scale near the south pole1 distinct publisher
science
Stewardship is aimed at the wrong half: rich countries use the resistance drugs, LMICs need them1 distinct publisher
Distinct publishers with included, body-backed reporting in this cluster.
phys.org
1 article · August 25, 2026