Engineered transferases and a T-cell microRNA atlas each show selective molecular targeting
Manchester artificial allylic transferases exceed 99 per cent conversion on key bonds, while a mouse atlas maps 110 microRNAs that separate inflammatory from regulatory T cells.
Science··Midday
Artificial enzymes that build carbon bonds selectively
A University of Manchester team built a family of artificial enzymes by placing non-natural catalytic amino acids into proteins, where they form reactive imidazolium intermediates. Reported in Nature Catalysis, the variant ASB1.3 passed 99 per cent conversion in several reactions with high stereochemical purity, and reached 98 per cent at preparative scale. A second variant, ASA1.5, built all-carbon quaternary stereocentres at 98 per cent conversion, a bond arrangement that is awkward to make selectively by ordinary chemistry. The enzymes worked across substituted furans, indoles, pyrroles, cyanoesters, diketones, ketoesters, anilines and isatins, each giving a single major product with few byproducts. The authors say more development is needed before the enzymes can be used widely, so what stands here is a demonstrated catalytic range and not yet a process ready for a plant. The public finding is selective carbon-carbon and carbon-nitrogen bond formation under enzyme control at high conversion, not a claim that industrial production has already switched to these catalysts.[1]
MicroRNAs that mark inflammatory T-cell states
Carolina Cunha and colleagues built a triple reporter mouse for Ifng, Il17 and Foxp3, put it through experimental autoimmune encephalomyelitis, and compared the microRNA profiles of the resulting T cell subsets. They report 110 microRNAs that differ between the pro-inflammatory Th1 and Th17 cells and the Foxp3 regulatory cells, in a peer-reviewed resource paper in PLOS Biology. Two of them carry new functions: miR-122-5p limits how fast Th17 cells multiply, and miR-1247 shapes Th1 differentiation, so one affects the course of the disease model and the other its severity. Both fall away in the inflamed central nervous system, which is what leads the authors to describe them as peripheral brakes that the disease releases. This is mouse work in an induced model, with nine to eleven animals per group across three independent experiments, and multiple sclerosis in people is a different problem. The atlas maps molecular distinctions among T-cell states in that model; it does not transfer those microRNAs into a human therapy.[2]
Selection at the bond and at the cell state
Both reports are about selective molecular targeting, yet they operate at different layers of biology. The ASB1.3 and ASA1.5 enzymes force carbon-carbon and carbon-nitrogen bonds toward single major products with conversions above 99 per cent in several reactions and at 98 per cent on preparative and quaternary-centre cases, while the authors still mark industrial readiness as unfinished. The PLOS Biology mouse atlas sorts 110 microRNAs that separate inflammatory from regulatory T cells and assigns course and severity roles to two of them inside an induced autoimmune model. One story is synthetic chemistry expanded by non-natural catalytic residues; the other is immune-cell state control mapped in mice. Neither study claims the other field's outcome: bond-forming catalysts do not become immune-cell drugs here, and microRNA differences do not become factory chemistry. Held together they show how modern molecular work isolates a preferred path among many possible ones, whether the path is a stereochemical product or an immune-cell programme, and they keep those paths inside their stated scopes.[1], [2]
