Summary of Science publications, focusing on contents relevant to AI and life sciences
Using a deep learning simulation framework at 100-meter resolution across Europe, researchers modeled how wildfires, insect outbreaks, and storms will reshape forests through the end of the 21st century. Under a scenario of continued high emissions, the total area affected by these disturbances is projected to more than double compared to recent decades, with wildfires emerging as the dominant driver of change. This shift toward more frequent and widespread disturbances would dramatically increase the proportion of young forests across the continent, fundamentally restructuring European forest age composition. The findings carry significant consequences for forest carbon storage capacity and biodiversity, since older forests disproportionately support both functions. The authors argue that anticipated disturbance increases should move to the center of European forest policy and management planning rather than being treated as background noise in climate adaptation strategies.
Alzheimer's disease involves progressive amyloid buildup that drives neurodegeneration, and existing anti-amyloid therapies have meaningful limitations. Researchers developed a novel approach called CAR-A therapy, in which astrocytes are engineered to express chimeric antigen receptors designed to target amyloid-β. After validating two CAR-A designs in cell culture, they tested them in mouse models and found that both constructs reduced existing amyloid plaques and their associated pathology, while also preventing early plaque formation. Single-nucleus RNA sequencing revealed that CAR-A treatment triggers a distinct glial response involving coordinated activity between astrocytes and microglia, with each receptor design producing unique cell-type-specific effects. The work establishes proof-of-concept for repurposing CAR cell therapy—a strategy already used in oncology—to target a neurodegenerative disease, positioning CAR-A as a potentially disease-modifying intervention for Alzheimer's.
Researchers set out to explain how mTORC2, a multi-protein complex central to PI3K and Ras signaling pathways frequently disrupted in cancer and diabetes, selectively phosphorylates certain kinases like Akt while ignoring closely related kinases targeted by the sister complex mTORC1. To capture and visualize the interaction, the team engineered semisynthetic molecular probes that stabilized the mTORC2–Akt complex long enough to determine its structure. Surprisingly, unlike most kinases that read the amino acid sequence immediately surrounding the phosphorylation site, mTORC2 pays little attention to local sequence context. Instead, recognition depends on the three-dimensional shape of Akt — specific secondary and tertiary structural features that dock onto the mTORC2 subunit mSin1 at a location away from the catalytic site. These structural elements appear conserved across at least 18 related substrates, clarifying a long-standing selectivity puzzle and pointing toward a rational path for designing inhibitors specific to mTORC2.
Researchers tested whether a single injection of self-amplifying RNA (saRNA) encoding the natriuretic peptide precursor Nppa could protect the heart after a heart attack. They packaged the saRNA in lipid nanoparticles and delivered it intramuscularly to mice and pigs. One injection sustained pro-ANP secretion for four weeks; the cardiac protease corin then converted it into active ANP, yielding significant cardioprotection superior to conventional mRNA at equivalent doses. Single-nucleus transcriptomics revealed that Npr1-positive endothelial and epicardial cells are the primary responders, with the treatment reprogramming their paracrine signaling to encourage cardiomyocyte regeneration and reduce fibrosis. No systemic toxicity was detected over extended follow-up. The findings establish a one-shot saRNA-LNP platform as a durable cardiac therapy and suggest this approach could be broadly applicable to other heart conditions requiring sustained protein delivery.
Genetic bottlenecks can trap small populations in a cycle of low diversity and high inbreeding, but escaping this trap is rarely documented in real populations. Researchers sequenced whole genomes from 418 koalas to compare genetic health across populations with different demographic histories. Counterintuitively, populations with greater overall genetic diversity carried heavier mutational loads and were shrinking, while populations that had previously gone through bottlenecks but were now rebounding showed reduced mutational burden, growing effective population sizes, and the emergence of new rare variants. The explanation lies in rapid demographic expansion: population growth reshuffles genetic variation through recombination and shifts effective population size faster than standard diversity measures can detect. The findings challenge the assumption that bottlenecked wildlife populations are necessarily on an irreversible genetic decline, suggesting that strong demographic recovery can restore evolutionary potential even in heavily impacted threatened species.
Plants must balance responsiveness to environmental flowering cues with the need to maintain an indeterminate stem cell pool that keeps producing flowers over time. Researchers used computational modeling combined with genetic and molecular experiments in Arabidopsis to investigate how this balance is achieved. They discovered a negative feedback loop between two key regulators: LEAFY (LFY), a transcription factor that promotes flower formation, and TERMINAL FLOWER1 (TFL1), a co-repressor that sustains inflorescence indeterminacy. When plants transition to reproductive growth, LFY directly activates TFL1 expression in a dose-dependent manner; TFL1 then suppresses LFY activity, preventing its overaccumulation. This mutual regulation acts as a buffer, shielding the inflorescence meristem from premature termination even when floral inductive signals are strong. The findings explain how two neighboring cell populations can respond differently to the same environmental input, maintaining architectural robustness throughout reproductive development.
Osteoarthritis, driven by aging or joint injury, causes progressive cartilage loss with few effective treatments available. Researchers discovered that the enzyme 15-hydroxy prostaglandin dehydrogenase (15-PGDH) is upregulated in articular cartilage of aged or injured mice. Using both systemic and local delivery of a small-molecule inhibitor of 15-PGDH, they observed cartilage regeneration and reduced OA-associated pain. Single-cell RNA sequencing and multiplexed immunofluorescence imaging revealed distinct chondrocyte subpopulations; inhibiting 15-PGDH shifted the balance away from hypertrophic-like chondrocytes toward matrix-producing articular chondrocytes. Importantly, this regeneration appeared to result from gene expression reprogramming within existing chondrocytes rather than stem or progenitor cell expansion. These findings position 15-PGDH inhibition as a promising disease-modifying and regenerative strategy for osteoarthritis, offering a potential therapeutic avenue beyond the symptomatic treatments currently available.
Bacteria, like animals and plants, use nucleotide signaling molecules as part of their immune responses against viruses, and phages have evolved proteins to neutralize these signals. Researchers developed a computational pipeline based on shared structural and biophysical features of known viral antidefense proteins to predict new phage proteins that interfere with bacterial nucleotide immunity. Experimental testing uncovered three previously unknown protein families—Sequestin, Lockin, and Acb5—each targeting bacterial immune systems called Thoeris and CBASS. Sequestin and Lockin work by binding and sequestering specific signaling nucleotides (3′cADPR and His-ADPR respectively), acting as molecular sponges, while Acb5 enzymatically cleaves cGAMP and related molecules. Structural and mutational studies clarified the precise mechanisms behind each activity. Thousands of related proteins appear across phage genomes, underscoring how widespread and varied phage strategies for disabling nucleotide-based bacterial immunity truly are.
Researchers investigated how the brain adjusts neural representations in visual cortex as animals learn new tasks, focusing on two competing theories: one predicting that learning reduces redundancy for efficiency, and another grounded in Bayesian inference predicting that learning instead spreads information redundantly across neurons. Using population recordings from macaque area V4 while monkeys trained on visual discrimination tasks over several weeks, the team found clear support for the Bayesian view. Training progressively increased redundancy in neural responses both across weeks and within individual trials. Crucially, this redundancy did not dilute the overall information available — rather, it boosted how much information individual neurons carried. The findings suggest the visual cortex operates more like a generative inference system, building internal models of sensory input, than a discriminative one that simply filters signals for categorization.
Chromosomal inversions are known to contribute to divergence between distinct ecotypes, but how they sustain continuous adaptive variation across environmental gradients is less clear. Researchers combined quantitative and population genetics, transcriptomics, and artificial selection experiments to study a widespread marine fish adapting along a steep thermal gradient despite substantial gene flow. They identified three chromosomal inversions linked to multiple adaptive traits, and these inversions showed the strongest signals of divergent selection across the genome. Notably, the inversions displayed different selection signatures at different latitudes, suggesting each inversion governs distinct components of complex traits and that together they enable modular adaptation to varying environmental pressures. The findings clarify how structural genomic variants can maintain continuous clinal adaptation in the face of gene flow, with implications for understanding how marine species may respond to shifting thermal environments.
Comments (0)
No comments yet. Log in to leave a comment.
You must be logged in to leave a comment.
Login to Comment