Microorganisms have driven Earth's sulfur cycle since the emergence of life1-6, yet the sulfur-cycling capacities of microorganisms and their integration with other element cycles remain incompletely understood. One such uncharacterized metabolism is the coupling of sulfide oxidation with iron(III) oxide reduction, a ubiquitous environmental process hitherto considered to be strictly abiotic7,8. Here we present a comprehensive genomic analysis of sulfur metabolism across prokaryotes, and reveal bacteria that are capable of oxidizing sulfide using extracellular solid phase iron(III). Based on a phylogenetic framework of over hundred genes involved in dissimilatory transformation of sulfur compounds, we recorded sulfur-cycling capacity in most bacterial and archaeal phyla. Metabolic reconstructions predicted co-occurrence of sulfur compound oxidation and iron(III) oxide respiration in diverse members of 37 prokaryotic phyla. Physiological and transcriptomic evidence demonstrated that a cultivated representative, Desulfurivibrio alkaliphilus, grows autotrophically by oxidizing dissolved sulfide or iron monosulfide (FeS) to sulfate with ferrihydrite as an extracellular iron(III) electron acceptor. The biological process outpaced the abiotic process at environmentally relevant sulfide concentrations. These findings expand the known diversity of sulfur-cycling microorganisms and unveil a biological mechanism that links sulfur and iron cycling in anoxic environments, thus highlighting the fundamental role of microorganisms in global element cycles.

Prenatal stress (PS) is a repeated exposure to aversive situations during pregnancy, including high emotional strain, which is suspected to affect homeostatic systems in infants. Paediatric eczema develops quickly after birth at flexural sites subjected to continuous mechanical constraints1,2. Although epidemiological studies have suggested an association between PS and a higher risk of eczema in children3-6, no causative biological link has yet been identified. Here we show that eczema at birth originates from molecular dysregulations of neuroimmune circuits in utero, triggered by fluctuations in the maternal hypothalamic-pituitary-adrenal axis. We found that offspring of stressed pregnant dams have dysregulated mast cells and skin-projecting neurons and quickly develop eczema in response to harmless mechanical friction. We demonstrated that PS transiently modulates amniotic fluid corticosterone concentrations, which directly alters the activation program of skin mast cells expressing the glucocorticoid receptor Nr3c1 and the adjacent sensory neurons conveying mechanosensation. Therapeutic normalization of maternal corticosterone concentrations or genetic depletion of Mcpt5+ mast cells during stressed gestation prevents fetal immune dysregulation and protects against eczema development after birth. Our findings support a new model in which early-onset paediatric eczema originates from dysregulations in the fetal immune system, caused by fluctuations in maternal glucocorticoids induced by stress.

The forthcoming sixth-generation and beyond wireless networks are poised to operate across an expansive frequency range-from microwave, millimetre wave to terahertz bands-to support ubiquitous connectivity in diverse application scenarios1-3. This necessitates a one-size-fits-all hardware solution that can be adaptively reconfigured within this wide spectrum to support full-band coverage and dynamic spectrum management4. However, existing electrical or photonic-assisted solutions face a lot of challenges in meeting this demand because of the limited bandwidths of the devices and the intrinsically rigid nature of system architectures5. Here we demonstrate adaptive wireless communications over an unprecedented frequency range spanning over 100 GHz, driven by a thin-film lithium niobate (TFLN) photonic wireless system. Leveraging the Pockels effect and scalability of the TFLN platform, we achieve monolithic integration of essential functional elements, including baseband modulation, broadband wireless-photonic conversion and reconfigurable carrier and local signal generation. Powered by broadband tunable optoelectronic oscillators, our signal sources operate across a record-wide frequency range from 0.5 GHz to 115 GHz with high-frequency stability and consistent coherence. Based on the broadband and reconfigurable integrated photonic solution, we realize full-link wireless communication across nine consecutive bands, achieving record lane speeds of up to 100 Gbps. The real-time reconfigurability further enables adaptive frequency allocation, a crucial ability to ensure enhanced reliability in complex spectrum environments. Our proposed system represents a marked step towards future full-spectrum and omni-scenario wireless networks.

Large urban gullies cause damage in many tropical cities across the Global South1,2. They can result from inappropriate urban planning and insufficient infrastructure to safely store and evacuate rainfall in environments that are already highly sensitive to soil erosion1,3,4. Although they can cause large destruction and societal impacts such as population displacement1,2,5, the magnitude of this geo-hydrological hazard remains poorly documented and understood6,7. Here we provide an assessment of the extent and impact of urban gullies at the scale of the Democratic Republic of the Congo (DRC). Through mapping, we identify 2,922 urban gullies across 26 cities. By combining their formation and growth rates with population density data8, we estimate that around 118,600 people (uncertainty range: ± 44,400 people) have been displaced by urban gullies over the period 2004-2023. We find that average displacement rates increased from about 4,650 persons yr-1 (pre-2020) to about 12,200 persons yr-1 (post-2020). Between 2010 and 2023, the number of people living in the potential expansion zone of urban gullies doubled from 1.6 (±0.6) to 3.2 (±1.3) million, with more likely to be exposed due to urban sprawl9,10 and climate change11. We suggest that there is a need for tools and strategies to prevent and mitigate this hazard.

To orchestrate ribosomal peptide synthesis, transfer RNAs (tRNAs) must be aminoacylated, with activated amino acids, at their 2',3'-diol moiety1,2, and so the selective aminoacylation of RNA in water is a key challenge that must be resolved to explain the origin of protein biosynthesis. So far, there have been no chemical methods to effectively and selectively aminoacylate RNA-2',3'-diols with the breadth of proteinogenic amino acids in water3-5. Here we demonstrate that (biological) aminoacyl-thiols (1) react selectively with RNA diols over amine nucleophiles, promoting aminoacylation over adventitious (non-coded) peptide bond formation. Broad side-chain scope is demonstrated, including Ala, Arg, Asp, Glu, Gln, Gly, His, Leu, Lys, Met, Phe, Pro, Ser and Val, and Arg aminoacylation is enhanced by unprecedented side-chain nucleophilic catalysis. Duplex formation directs chemoselective 2',3'-aminoacylation of RNA. We demonstrate that prebiotic nitriles, N-carboxyanhydrides and amino acid anhydrides, as well as biological aminoacyl-adenylates, all react with thiols (including coenzymes A and M) to selectively yield aminoacyl-thiols (1) in water. Finally, we demonstrate that the switch from thioester to thioacid activation inverts diol/amine selectivity, promoting peptide synthesis in excellent yield. Two-step, one-pot, chemically controlled formation of peptidyl-RNA is observed in water at neutral pH. Our results indicate an important role for thiol cofactors in RNA aminoacylation before the evolution of proteinaceous synthetase enzymes.

Generative models cover various application areas, including image and video synthesis, natural language processing and molecular design, among many others1-11. As digital generative models become larger, scalable inference in a fast and energy-efficient manner becomes a challenge12-14. Here we present optical generative models inspired by diffusion models4, where a shallow and fast digital encoder first maps random noise into phase patterns that serve as optical generative seeds for a desired data distribution; a jointly trained free-space-based reconfigurable decoder all-optically processes these generative seeds to create images never seen before following the target data distribution. Except for the illumination power and the random seed generation through a shallow encoder, these optical generative models do not consume computing power during the synthesis of the images. We report the optical generation of monochrome and multicolour images of handwritten digits, fashion products, butterflies, human faces and artworks, following the data distributions of MNIST15, Fashion-MNIST16, Butterflies-10017, Celeb-A datasets18, and Van Gogh's paintings and drawings19, respectively, achieving an overall performance comparable to digital neural-network-based generative models. To experimentally demonstrate optical generative models, we used visible light to generate images of handwritten digits and fashion products. In addition, we generated Van Gogh-style artworks using both monochrome and multiwavelength illumination. These optical generative models might pave the way for energy-efficient and scalable inference tasks, further exploiting the potentials of optics and photonics for artificial-intelligence-generated content.