Methanogenic archaea are the main producers of the potent greenhouse gas methane1,2. In the methanogenic pathway from CO2 and H2 studied under laboratory conditions, low-potential electrons for CO2 reduction are generated by a flavin-based electron-bifurcation reaction catalysed by heterodisulfide reductase (Hdr) complexed with the associated [NiFe]-hydrogenase (Mvh)3-5. F420-reducing [NiFe]-hydrogenase (Frh) provides electrons to the methanogenic pathway through the electron carrier F420 (ref. 6). Here we report that under strictly nickel-limited conditions, in which the nickel concentration is similar to those often observed in natural habitats7-11, the production of both [NiFe]-hydrogenases in Methanothermobacter marburgensis is strongly downregulated. The Frh reaction is substituted by a coupled reaction with [Fe]-hydrogenase (Hmd), and the role of Mvh is taken over by F420-dependent electron-donating proteins (Elp). Thus, Hmd provides all electrons for the reducing metabolism under these nickel-limited conditions. Biochemical and structural characterization of Elp-Hdr complexes confirms the electronic interaction between Elp and Hdr. The conservation of the genes encoding Elp and Hmd in CO2-reducing hydrogenotrophic methanogens suggests that the Hmd system is an alternative pathway for electron flow in CO2-reducing hydrogenotrophic methanogens under nickel-limited conditions.

Social interaction can be regarded as a dynamic feedback loop between interacting individuals as they act and react to each other1,2. Here, to understand the neural basis of these interactions, we investigated inter-brain neural dynamics across individuals in both mice and artificial intelligence systems. By measuring activities of molecularly defined neurons in the dorsomedial prefrontal cortex of socially interacting mice, we find that the multi-dimensional neural space within each individual can be partitioned into two distinct subspaces-a shared neural subspace that represents shared neural dynamics across animals and a unique neural subspace that represents activity unique to each animal. Notably, compared with glutamatergic neurons, GABAergic (γ-aminobutyric acid-producing) neurons in the dorsomedial prefrontal cortex contain a considerably larger shared neural subspace, which arises from behaviours of both self and others. We extended this framework to artificial intelligence agents and observed that, as social interactions emerged, so too did shared neural dynamics between interacting agents. Importantly, selectively disrupting the neural components that contribute to shared neural dynamics substantially reduces the agents' social actions. Our findings suggest that shared neural dynamics represent a fundamental and generalizable feature of interacting neural systems present in both biological and artificial agents and highlight the functional significance of shared neural dynamics in driving social interactions.

Establishing a unified theory of cognition has been an important goal in psychology1,2. A first step towards such a theory is to create a computational model that can predict human behaviour in a wide range of settings. Here we introduce Centaur, a computational model that can predict and simulate human behaviour in any experiment expressible in natural language. We derived Centaur by fine-tuning a state-of-the-art language model on a large-scale dataset called Psych-101. Psych-101 has an unprecedented scale, covering trial-by-trial data from more than 60,000 participants performing in excess of 10,000,000 choices in 160 experiments. Centaur not only captures the behaviour of held-out participants better than existing cognitive models, but it also generalizes to previously unseen cover stories, structural task modifications and entirely new domains. Furthermore, the model's internal representations become more aligned with human neural activity after fine-tuning. Taken together, our results demonstrate that it is possible to discover computational models that capture human behaviour across a wide range of domains. We believe that such models provide tremendous potential for guiding the development of cognitive theories, and we present a case study to demonstrate this.

The North Eurasian forest and forest-steppe zones have sustained millennia of sociocultural connections among northern peoples, but much of their history is poorly understood. In particular, the genomic formation of populations that speak Uralic and Yeniseian languages today is unknown. Here, by generating genome-wide data for 180 ancient individuals spanning this region, we show that the Early-to-Mid-Holocene hunter-gatherers harboured a continuous gradient of ancestry from fully European-related in the Baltic, to fully East Asian-related in the Transbaikal. Contemporaneous groups in Northeast Siberia were off-gradient and descended from a population that was the primary source for Native Americans, which then mixed with populations of Inland East Asia and the Amur River Basin to produce two populations whose expansion coincided with the collapse of pre-Bronze Age population structure. Ancestry from the first population, Cis-Baikal Late Neolithic-Bronze Age (Cisbaikal_LNBA), is associated with Yeniseian-speaking groups and those that admixed with them, and ancestry from the second, Yakutia Late Neolithic-Bronze Age (Yakutia_LNBA), is associated with migrations of prehistoric Uralic speakers. We show that Yakutia_LNBA first dispersed westwards from the Lena River Basin around 4,000 years ago into the Altai-Sayan region and into West Siberian communities associated with Seima-Turbino metallurgy-a suite of advanced bronze casting techniques that expanded explosively from the Altai1. The 16 Seima-Turbino period individuals were diverse in their ancestry, also harbouring DNA from Indo-Iranian-associated pastoralists and from a range of hunter-gatherer groups. Thus, both cultural transmission and migration were key to the Seima-Turbino phenomenon, which was involved in the initial spread of early Uralic-speaking communities.

Ancient Egyptian society flourished for millennia, reaching its peak during the Dynastic Period (approximately 3150-30 BCE). However, owing to poor DNA preservation, questions about regional interconnectivity over time have not been addressed because whole-genome sequencing has not yet been possible. Here we sequenced a 2× coverage whole genome from an adult male Egyptian excavated at Nuwayrat (Nuerat, نويرات). Radiocarbon dated to 2855-2570 cal. BCE, he lived a few centuries after Egyptian unification, bridging the Early Dynastic and Old Kingdom periods. The body was interred in a ceramic pot within a rock-cut tomb1, potentially contributing to the DNA preservation. Most of his genome is best represented by North African Neolithic ancestry, among available sources at present. Yet approximately 20% of his genetic ancestry can be traced to genomes representing the eastern Fertile Crescent, including Mesopotamia and surrounding regions. This genetic affinity is similar to the ancestry appearing in Anatolia and the Levant during the Neolithic and Bronze Age2-5. Although more genomes are needed to fully understand the genomic diversity of early Egyptians, our results indicate that contacts between Egypt and the eastern Fertile Crescent were not limited to objects and imagery (such as domesticated animals and plants, as well as writing systems)6-9 but also encompassed human migration.

The efficacy of chimeric antigen receptor (CAR) T cell therapy in solid tumours is limited by immunosuppression and antigen heterogeneity1-3. To overcome these barriers, 'armoured' CAR T cells, which secrete proinflammatory cytokines, have been developed4. However, their clinical application has been limited because of toxicity related to peripheral expression of the armouring transgene5. Here, we have developed a CRISPR knock-in strategy that leverages the regulatory mechanisms of endogenous genes to drive transgene expression in a tumour-localized manner. By screening endogenous genes with tumour-restricted expression, we have identified the NR4A2 and RGS16 promoters as promising candidates to support the delivery of cytokines such as IL-12 and IL-2 directly to the tumour site, leading to enhanced antitumour efficacy and long-term survival of mice in both syngeneic and xenogeneic models. This effect was concomitant with improved CAR T cell polyfunctionality, activation of endogenous antitumour immunity and a favourable safety profile, and was applicable in CAR T cells from patients.

Ultimate limits for the sensing of fields and forces are set by the quantum noise of a sensor1-3. Entanglement allows for suppression of such noise and for achieving sensitivity beyond standard quantum limits4-7. Applicability of quantum optical sensing is often restricted by fixed wavelengths of available photonic quantum sources. Another ubiquitous limitation is associated with challenges of achieving quantum-noise-limited sensitivity in the acoustic noise frequency range relevant for several applications. Here we demonstrate a tool for broadband quantum sensing by performing quantum state processing that can be applied to a wide range of the optical spectrum and by suppressing quantum noise over an octave in the acoustic frequency range. An atomic spin ensemble is strongly coupled to one of the frequency-tunable beams of an Einstein-Podolsky-Rosen (EPR) source of light. The other EPR beam of light, entangled with the first one, is tuned to a disparate wavelength. Engineering the spin ensemble to act as a negative-mass or positive-mass oscillator, we demonstrate frequency-dependent quantum noise reduction for measurements at the disparate wavelength. The tunability of the spin ensemble enables targeting quantum noise in a variety of systems with dynamics ranging from kHz to MHz. As an example of broadband quantum noise reduction in the acoustic frequency range, we analyse the applicability of our approach to gravitational-wave detectors (GWDs). Other possible applications include continuous-variable quantum repeaters and distributed quantum sensing.

Bdellovibrio bacteriovorus is a predatory bacterium that non-selectively preys on Gram-negative bacteria by invading the prey-cell periplasm, leaching host nutrients and ultimately lysing the infected cell to exit and find a new host1,2. The predatory life cycle of B. bacteriovorus is, in many ways, comparable to a bacteriophage. However, unlike phage defence, defence against B. bacteriovorus has not been widely investigated. Here we screened a collection of diverse Escherichia coli strains for resistance to B. bacteriovorus and identified that roughly one-third of strains robustly defended against predation by producing curli fibres. Curli fibres are oligomers of the functional amyloid protein CsgA, which is exceptionally durable3. Using genetics and microscopy, we demonstrate that curli fibres provide a barrier that protects susceptible cells independent of genes required for biofilm formation. This barrier further protected E. coli against attack by the predatory bacterium Myxococcus xanthus and select phages. Bioinformatic analysis of bacterial amyloids showed these systems are diverse and widespread in diderm bacteria (those with both inner and outer membranes). One of these, an evolutionarily distinct amyloid encoded by Pseudomonas aeruginosa, also protected against B. bacteriovorus. This work establishes that functional amyloids defend bacteria against a wide range of threats.

Acute inflammation is an essential response that our bodies use to combat infections1. However, in the absence of infections, chronic inflammation can have a pivotal role in the onset and progression of chronic diseases, such as arthritis, cancer, autoimmune disorders, metabolic-dysfunction-associated steatohepatitis (MASH), and most ageing-associated pathologies2,3. The underlying mechanisms that distinguish chronic inflammation from its acute counterpart remain unclear, posing challenges to the development of targeted therapies for these major diseases. Here we identify a mechanism that separates the two responses: during chronic but not acute inflammation, chromatin remodelling is influenced by nuclear autophagy, in which the WSTF protein of the ISWI chromatin-remodelling complex interacts with the ATG8 autophagy protein family in the nucleus. This interaction leads to WSTF nuclear export and subsequent degradation by autophagosomes and lysosomes in the cytoplasm. Loss of WSTF leads to chromatin opening over inflammatory genes, amplifying inflammation. Cell-penetrating peptides that block the WSTF-ATG8 interaction do not affect acute inflammation but suppress chronic inflammation in senescence as well as in MASH and osteoarthritis in mouse models and patient samples. The ability to specifically target chronic inflammation without blunting acute inflammation offers an approach for treating common chronic inflammatory diseases.

Lung cancer in never smokers (LCINS) accounts for around 25% of all lung cancers1,2 and has been associated with exposure to second-hand tobacco smoke and air pollution in observational studies3-5. Here we use data from the Sherlock-Lung study to evaluate mutagenic exposures in LCINS by examining the cancer genomes of 871 treatment-naive individuals with lung cancer who had never smoked, from 28 geographical locations. KRAS mutations were 3.8 times more common in adenocarcinomas of never smokers from North America and Europe than in those from East Asia, whereas a higher prevalence of EGFR and TP53 mutations was observed in adenocarcinomas of never smokers from East Asia. Signature SBS40a, with unknown cause6, contributed the largest proportion of single base substitutions in adenocarcinomas, and was enriched in cases with EGFR mutations. Signature SBS22a, which is associated with exposure to aristolochic acid7,8, was observed almost exclusively in patients from Taiwan. Exposure to secondhand smoke was not associated with individual driver mutations or mutational signatures. By contrast, patients from regions with high levels of air pollution were more likely to have TP53 mutations and shorter telomeres. They also exhibited an increase in most types of mutations, including a 3.9-fold increase in signature SBS4, which has previously been linked with tobacco smoking9, and a 76% increase in the clock-like10 signature SBS5. A positive dose-response effect was observed with air-pollution levels, correlating with both a decrease in telomere length and an increase in somatic mutations, mainly attributed to signatures SBS4 and SBS5. Our results elucidate the diversity of mutational processes shaping the genomic landscape of lung cancer in never smokers.

Although most mammalian transcriptional enhancers regulate their cognate promoters over distances of tens of kilobases, some enhancers act over distances in the megabase range1. The sequence features that enable such long-distance enhancer-promoter interactions remain unclear. Here we used in vivo enhancer-replacement experiments at the mouse Shh locus to show that short- and medium-range limb enhancers cannot initiate gene expression at long-distance range. We identify a cis-acting element, range extender (REX), that confers long-distance regulatory activity and is located next to a long-range limb enhancer of Sall1. The REX element has no endogenous enhancer activity. However, addition of the REX to other short- and mid-range limb enhancers substantially increases their genomic interaction range. In the most extreme example observed, addition of REX increased the range of an enhancer by an order of magnitude from its native 73 kb to 848 kb. The REX element contains highly conserved [C/T]AATTA homeodomain motifs that are critical for its activity. These motifs are enriched in long-range limb enhancers genome-wide, including the ZRS (zone of polarizing activity (ZPA) regulatory sequence), a benchmark long-range limb enhancer of Shh2. The ZRS enhancer with mutated [C/T]AATTA motifs maintains limb activity at short range, but loses its long-range activity, resulting in severe limb reduction in knock-in mice. In summary, we identify a sequence signature associated with long-range enhancer-promoter interactions and describe a prototypical REX element that is necessary and sufficient to confer long-distance activation by remote enhancers.

Barrier tissues isolate organisms from their surrounding environment. Maintaining the integrity of the tissues is essential for this function. In many seed plants, periderm forms as the outer barrier during secondary growth to prevent water loss and pathogen infection1. The periderm is regenerated when its integrity is lost following injury; however, the underlying mechanism remains largely unknown, despite its importance for plant survival. Here we report that periderm integrity in Arabidopsis roots is sensed by diffusion of the gases ethylene and oxygen. Following injury of the periderm, ethylene leaks out through the wound and oxygen enters, resulting in attenuation of ethylene signalling and hypoxia signalling. This condition promotes periderm regeneration in the root. When regeneration is complete and barrier integrity is re-established, pre-injury levels of ethylene and hypoxia signalling are regained. Gas diffusion monitoring is also used to re-establish the barrier in inflorescence stems after the epidermis is injured. We thus propose that gas diffusion is used by plants as a general principle to monitor and re-establish barrier integrity.

Understanding how animals and humans learn from experience to make adaptive decisions is a fundamental goal of neuroscience and psychology. Normative modelling frameworks such as Bayesian inference1 and reinforcement learning2 provide valuable insights into the principles governing adaptive behaviour. However, the simplicity of these frameworks often limits their ability to capture realistic biological behaviour, leading to cycles of handcrafted adjustments that are prone to researcher subjectivity. Here we present a novel modelling approach that leverages recurrent neural networks to discover the cognitive algorithms governing biological decision-making. We show that neural networks with just one to four units often outperform classical cognitive models and match larger neural networks in predicting the choices of individual animals and humans, across six well-studied reward-learning tasks. Critically, we can interpret the trained networks using dynamical systems concepts, enabling a unified comparison of cognitive models and revealing detailed mechanisms underlying choice behaviour. Our approach also estimates the dimensionality of behaviour3 and offers insights into algorithms learned by meta-reinforcement learning artificial intelligence agents. Overall, we present a systematic approach for discovering interpretable cognitive strategies in decision-making, offering insights into neural mechanisms and a foundation for studying healthy and dysfunctional cognition.

Biomolecular condensates are key features of intracellular compartmentalization1,2. As the most prominent nuclear condensate in eukaryotes, the nucleolus is a multiphase liquid-like structure in which ribosomal RNAs (rRNAs) are transcribed and processed, undergoing multiple maturation steps to form the small (SSU) and large (LSU) ribosomal subunits3-5. However, how rRNA processing is coupled to the layered organization of the nucleolus is poorly understood owing to a lack of tools to precisely monitor and perturb nucleolar rRNA processing dynamics. Here we developed two complementary approaches to spatiotemporally map rRNA processing and engineer de novo nucleoli. Using sequencing in parallel with imaging, we found that rRNA processing steps are spatially segregated, with sequential maturation of rRNA required for its outward movement through nucleolar phases. By generating synthetic nucleoli in cells using an engineered rDNA plasmid system, we show that defects in SSU processing can alter the ordering of nucleolar phases, resulting in inside-out nucleoli and preventing rRNA outflux, while LSU precursors are necessary to build the outermost layer of the nucleolus. These findings demonstrate how rRNA is both a scaffold and substrate for the nucleolus, with rRNA acting as a programmable blueprint for the multiphase architecture that facilitates assembly of an essential molecular machine.

In a classic example of adaptation, harmless Batesian mimics gain protection from predators through resemblance to one or more unpalatable models1,2. Mimics vary greatly in accuracy, and explaining the persistence of inaccurate mimics is an ongoing challenge for evolutionary biologists3,4. Empirical testing of existing hypotheses is constrained by the difficulty of assessing the fitness of phenotypes absent among extant species, leaving large parts of the adaptive landscape unexplored5-a problem affecting the study of the evolution of most complex traits. Here, to address this, we created mimetic phenotypes that occupy hypothetical areas of trait space by morphing between 3D images of real insects (flies and wasps), and tested the responses of real predators to high-resolution, full-colour 3D-printed reproductions of these phenotypes. We found that birds have an excellent ability to learn to discriminate among insects on the basis of subtle differences in appearance, but this ability is weaker for pattern and shape than for colour and size traits. We found that mimics gained no special protection from intermediate resemblance to multiple model phenotypes. However, discrimination ability was lower in some invertebrate predators (especially crab spiders and mantises), highlighting that the predator community is key to explaining the apparent inaccuracy of many mimics.

Immune checkpoint blockade (ICB) therapy is effective against many cancers, although resistance remains a major issue and new strategies are needed to improve clinical outcomes1-5. Here we studied ICB response in a cohort of patients with ovarian clear cell carcinoma-a cancer type that poses considerable clinical challenges and lacks effective therapies6-8. We observed significantly prolonged overall survival and progression-free survival in patients with tumours with PPP2R1A mutations. Importantly, our findings were validated in additional ICB-treated patient cohorts across multiple cancer types. Translational analyses from tumour biopsies demonstrated enhanced IFNγ signalling, and the presence of tertiary lymphoid structures at the baseline, as well as enhanced immune infiltration and expansion of CD45RO+CD8+ T cells in the tumour neighbourhood after ICB treatment in PPP2R1A-mutated tumours. Parallel preclinical investigations showed that targeting PPP2R1A (by pharmacological inhibition or genetic modifications) in in vitro and in vivo models was associated with improved survival in the setting of treatment with several forms of immunotherapy, including chimeric antigen receptor (CAR)-T cell therapy and ICB. The results from these studies suggest that therapeutic targeting of PPP2R1A may represent an effective strategy to improve patient outcomes after ICB or other forms of immunotherapy, although additional mechanistic and therapeutic insights are needed.

Multi-omics studies, represented by connectomes and spatial transcriptomes, have entered the era of single-cell resolution, necessitating a reference brain atlas with spatial localization capability at the single-cell level1-4. However, such atlases are unavailable5. Here we present a whole mouse brain dataset of Nissl-based cytoarchitecture with isotropic 1-μm resolution, achieved through continuous micro-optical sectioning tomography. By integrating multi-modal images, we constructed a three-dimensional reference atlas of the mouse brain, providing the three-dimensional topographies of 916 structures and enabling arbitrary-angle slice image generation at 1-μm resolution. We developed an informatics-based platform for visualizing and sharing of the atlas images, offering services such as brain slice registration, neuronal circuit mapping and intelligent stereotaxic surgery planning. This atlas is interoperable with widely used stereotaxic atlases, supporting cross-atlas navigation of corresponding coronal planes in two dimensions and spatial mapping across atlas spaces in three dimensions. By facilitating the data analysis and visualization for large brain mapping projects, our atlas promises to be a versatile brainsmatics tool for studying the whole brain at single-cell level.

Climate assessments of civil aviation1,2 have consistently quantified the dominant climate-forcing components: (1) CO2 emissions, (2) NOx (NO + NO2) emissions and (3) persistent contrails. All three components exert a positive radiative forcing (RF) and lead to climate warming of similar magnitudes. The aviation community is actively seeking to reduce its climate footprint through advanced engine technologies, more sustainable aviation fuel and optimal routing plans3-12. These approaches usually involve a trade-off of CO2 against NOx or contrails (non-CO2), such as burning 1% more fuel to decrease contrail RF by 4%. Here, we show that a climate-trade-off risk curve derived from uncertainties in the RF components2,13-16 can give the probability that a specified trade-off ratio will produce a climate benefit. For each component, we calculate the integrated effective RF resulting from 1 year of flights: global warming per activity (GWA). The complementary cumulative probability distribution of the GWA(non-CO2) to GWA(CO2) ratio results in a climate-trade-off risk curve giving the likelihood of a positive climate outcome as a function of the trade-off-CO2 to trade-off-non-CO2 ratio, because the product, GWA × trade-off, should be the same for both. We find a likely (67%) chance of climate mitigation on a 100-year time horizon for the above suggested ratio of 1:4, favouring proposed non-CO2 mitigation efforts3-12 with ratios smaller than this.

In the past decade, hundreds of exoplanets have been discovered in extremely short orbits below 10 days. Unlike in the Solar System, planets in these systems orbit their host stars close enough to disturb the stellar magnetic field lines1. The interaction can enhance the magnetic activity of the star, such as its chromospheric2 and radio3 emission or flaring4. So far, the search for magnetic star-planet interactions has remained inconclusive. Here we report the detection of planet-induced flares on HIP 67522, a 17 million-year-old G dwarf star with two known close-in planets5,6. Combining space-borne photometry from the Transiting Exoplanet Survey Satellite and dedicated Characterising Exoplanets Telescope observations over 5 years, we find that the 15 flares in HIP 67522 cluster near the transit phase of the innermost planet, indicating persistent magnetic star-planet interaction in the system. The stability of interaction implies that the innermost planet is continuously self-inflicting a six times higher flare rate than it would experience without interaction. The subsequent flux of energetic radiation and particles bombarding HIP 67522 b may explain the remarkably extended atmosphere of the planet, recently detected with the James Webb Space Telescope7. HIP 67522 is, therefore, an archetype to understand the impact of magnetic star-planet interaction on the atmospheres of nascent exoplanets.

Natural chiral polymers, such as DNA, proteins, cellulose and poly[(R)-3-hydroxybutyrate] ((R)-P3HB), are prevalent in their enantiopure forms1,2. Existing methods to synthesize enantiopure polymers focus on enantiospecific polymerization, in which only one specific enantiomer is obtained from the corresponding chiral monomer3-6. Here we introduce a catalytic stereodivergent synthetic strategy to access all enantiopure di-isotactic poly(3-hydroxyalkanoate) (PHA) diastereomers from bacterial (R)-P3HB as the single chiral source. A series of enantiopure (R,R)-α-alkylated-β-butyrolactones are obtained from (R)-P3HB and then subjected to the catalyst-controlled diastereodivergent ring-opening polymerization (ROP) to enantiopure di-isotactic α-alkylated PHAs. Metal-catalysed coordination-insertion ROP results in threo-(R,R)-di-isotactic PHAs with chiral retention, whereas anionic ROP catalysed by an organic superbase produces erythro-(R,S)-di-isotactic PHAs with chiral inversion, achieving precision di-isotactic PHAs with exclusive regio- and stereoregularity. This strategy has also enabled the stereodivergent synthesis of all four [(R,R), (S,S), (R,S) and (S,R)] stereoisomers of α,α-dialkylated PHAs from (R)-P3HB, which can be depolymerized to chiral α,α-dialkylated-β-butyrolactones with high stereoselectivity. Overall, this catalyst-controlled regio- and stereoselective, stereodivergent synthetic methodology provides access to 16 enantiopure stereoisomers of α(α)-(di)substituted PHAs and enables the stereochemistry-defined structure-property relationship study of the di-isotactic PHAs, providing insights into the effects of main-chain stereoconfigurations and alkyl side chains on their thermal properties, melt processability, mechanical performance and supramolecular stereocomplexation.