Summary of Nature publications, focusing on contents relevant to AI and life sciences
Colorectal Neoplasms, Tumor Microenvironment, Humans, Animals, Fibroblasts, Cell Plasticity, Organoids, Female, Cell Transformation, Neoplastic, Disease Progression, Neoplasm Invasiveness, Neoplasm Metastasis, Mice, Male, Coculture Techniques
Metastatic ability in colorectal cancer has long been assumed to arise late in the disease. Using multiregional organoid models spanning early tumor stages, whole-genome sequencing, and single-cell spatial mapping of the tumor microenvironment, the researchers found that metastasis-associated oncofetal cell states appear already at the earliest stages, tied to formation of the invasive front. These states were present in all early non-metastatic tumors, so cancer-intrinsic genetic changes alone couldn't explain their emergence. Instead, the first cancer-associated fibroblasts recruited during submucosal invasion resembled normal submucosal trophocytes and co-localized with oncofetal cells at invasive fronts. Coculture experiments confirmed these trophocyte-like fibroblasts actively induce the plastic switch to oncofetal states. This reveals that the immediate cross-talk between tumor cells and surrounding submucosal fibroblasts right after malignant transformation determines when and where metastatic plasticity first develops.
Humans, China, Genetic Variation, Genetics, Medical, Genetics, Population, Genome, Human, Genomic Structural Variation, Haplotypes, HLA Antigens, Multigene Family, Quantitative Trait Loci, Tandem Repeat Sequences, East Asian People
Researchers built a large pangenome dataset from 1,116 Chinese diploid genome assemblies, combining 55 novel assemblies with 1,061 reconstructed ones. This resource adds roughly 405 million base pairs of sequence not found in standard reference genomes, including many potentially functional gene and regulatory regions. The team mapped a broad range of genetic variants—millions of small mutations, structural variants, tandem repeats, and previously hidden variants in novel sequences—and connected some to gene expression changes. Specifically, they discovered over 3,200 expression-linked variants involving complex structures, revealing intricate regulatory effects. They also created a reference panel with diverse variant types to improve future genetic association studies. Overall, this effort expands our understanding of complex human genetic variation and provides a powerful tool for medical and population genetics research, particularly for variants outside typical references.
Humans, Mutation, Autoimmunity, Receptors, Tumor Necrosis Factor, Member 14, Thyroid Gland, B-Lymphocytes, B7-H1 Antigen, Animals, Spatial Transcriptomics, Female
Researchers investigated whether somatic mutations in immune regulatory genes allow self-reactive lymphocytes to escape tolerance checkpoints and trigger autoimmune disease, testing this in autoimmune thyroid disease. Using whole-exome sequencing and targeted NanoSeq, a highly accurate single-molecule DNA sequencing method, they scanned for driver mutations. They discovered many B cell clones independently acquiring loss-of-function mutations in immune checkpoint genes, notably TNFRSF14 (HVEM) and CD274 (PD-L1), alongside less frequent alterations in other immune genes. Highly inflamed tissues harbored tens to hundreds of distinct mutant clones. Through techniques like laser microdissection, spatial transcriptomics, and single-nucleus sequencing, they confirmed these mutations occurred in B cells, some self-reactive, with widespread biallelic TNFRSF14 loss and clones carrying four to six driver mutations. Although individual clones represented small fractions, collectively mutant clones comprised a substantial share of B cells. These findings support the idea that somatic mutations drive a polyclonal evolutionary cascade enabling autoreactive lymphocytes to bypass tolerance, illuminating autoimmune disease mechanisms.
Animals, Female, Immunity, Innate, Eosinophils, Lactation, Mice, Reproduction, Pregnancy, Goblet Cells, Cell Differentiation, Mucus, Intestine, Small, Intestines, Intestinal Barrier Function, Mice, Inbred C57BL
Pregnancy and lactation trigger immune shifts in the gut that were previously unclear. Researchers found that eosinophils—immune cells usually linked to allergies and parasites—accumulate in the small intestine during pregnancy and peak during lactation, without any infection. Using mouse models, organoid cultures, and single-cell and spatial transcriptomics, they showed eosinophils act on intestinal stem cells to promote goblet cell differentiation, boosting mucus production. This strengthened barrier limited bacterial entry and spread, providing broad innate protection against gut infections. Notably, these protective changes persisted for weeks after lactation ended. The findings reveal that, despite overall immune tolerance during reproduction, the maternal intestine actively strengthens its defenses—likely an evolutionary safeguard for mothers and offspring in pathogen-rich settings. This work opens avenues for studying tissue-specific immune adaptation across reproduction and suggests lasting effects on host defense and women's health.
Animals, Female, Male, Mice, Competitive Behavior, Cooperative Behavior, Dopamine, Dopaminergic Neurons, Reinforcement Machine Learning, Sex Characteristics, Social Behavior, Ventral Tegmental Area, Behavior, Animal
Using automated behavioral tracking in semi-natural enclosures, neural recordings, and reinforcement-learning models, researchers asked how social roles emerge spontaneously from individual interactions. They found that genetically identical male mice, when solving a foraging task under social constraints, developed stable worker–scrounger roles shaped by competition, whereas females adopted uniform, cooperative strategies—despite similar solo behavior. Dopamine signaling in the ventral tegmental area drove these sex-divergent role patterns. Computational modeling showed that small individual differences in resource exploitation, amplified by contingent social interactions and competition, could cascade into distinct social profiles. The roles proved plastic: mixing sexes or introducing experienced animals into naïve groups reshaped role distributions, and dopaminergic manipulations likewise altered group structure. The results suggest a feedback loop in which social context changes neural states, reinforcing behavioral specialization and stabilizing social organization.
Conservation of Natural Resources, Global Warming, Forests, Droughts, Atmosphere, Brazil
This study modeled how global warming and deforestation interact to threaten the Amazon biome, using a dynamical systems framework combined with atmospheric moisture tracking. Without deforestation, the model identified a critical warming level of 3.7–4.0 °C, beyond which up to one-third of the forest could lose stability. However, when deforestation was included, far larger transitions—affecting 62–77% of the area—occurred at much lower warming (1.5–1.9 °C) combined with 22–28% forest loss. Most of these shifts were driven by cascading drought effects that spread over hundreds to thousands of kilometers, as moisture transport links distant regions. The findings underscore that keeping warming below 1.5 °C, halting deforestation, and restoring degraded areas are essential to prevent widespread, self-reinforcing ecosystem collapse in the Amazon.
Animals, Humans, Agriculture, Biological Evolution, Dental Enamel, Diet, Durapatite, Fossils, History, Ancient, Meat, Nanoparticles, Primates
Researchers used a nanoscale imaging technique called PELICAN to measure the angular misalignment between neighboring enamel nanocrystals in teeth from 12 primate specimens, including humans and fossil relatives spanning nearly 18 million years. Across apes and monkeys, misorientation increased with dietary hardness—tougher foods coincided with greater crystal misalignment. In the human lineage, misorientation rose steadily over the past 1.6 million years, with notable jumps tied to the shift toward meat-eating and the later adoption of agriculture with stone-ground grains, but not with the Industrial Revolution. This suggests that enamel adapted to changing diets not just in overall shape but at the crystallographic nanoscale, potentially enhancing fracture resistance. The finding hints that engineered materials could mimic small misorientation angles to improve toughness and resilience.
Animals, Female, Humans, Male, Mice, Rats, Aging, Caloric Restriction, Cellular Senescence, Chronic Disease, Glucuronidase, Inflammation, Klotho Proteins, Longevity, Mammals, Transcriptome, Macaca
Researchers combined over 11,000 gene-expression datasets from four mammals (mouse, rat, macaque, human) spanning 25+ tissues to build machine-learning biomarkers that predict both chronological age and remaining lifespan. The models successfully forecast mortality, chronic disease, and the effects of lifespan-modulating interventions like caloric restriction. Age-related gene changes were conserved across species and cell types, with key mortality-linked genes like CDKN1A and LGALS3 also tied to death and multimorbidity in human biobank data. The team identified a modular network of ageing hallmarks—including inflammation, interferon signaling, mitochondrial function, and chromatin remodeling—and built module-specific clocks to track ageing in individual cellular subsystems. These clocks showed that different interventions act on different modules, and that transcriptomic and DNA-methylation ageing measures correlate in human blood, especially for the chromatin module. The work offers a framework for quantifying and potentially targeting ageing in specific tissues and pathways.
Forests, Climate Change, Carbon, Carbon Sequestration, Machine Learning, United States
This study examined whether carbon crediting projects that rely on forest carbon storage are adequately prepared for the rising risk of natural disturbances under climate change. Using forest inventory data, satellite observations, disturbance models, and machine learning, the team mapped the likelihood of carbon loss across forests in the contiguous United States over 100 years. They found that climate change notably increases this risk, especially in California and the Intermountain West. Comparing their risk estimates to the buffer pool—the reserve meant to cover unexpected carbon losses—of the region's largest forest mitigation program, they concluded that pool is on average 6.3 times too small, with that shortfall possibly ranging from 2.2 to 8.0 times depending on future climate and disturbance assumptions. The authors argue these findings show current methodologies for sizing buffer pools need revision, and they offer detailed maps of long-term carbon loss risk to support better planning.
Humans, Cell Dedifferentiation, Cell Line, Tumor, Chromatin Assembly and Disassembly, CRISPR-Cas Systems, Drug Resistance, Neoplasm, Gene Expression Regulation, Neoplastic, Mediator Complex, Melanoma, Phenotype, Single-Cell Gene Expression Analysis, Transcription Factors, Transcriptional Activation, Vemurafenib
The researchers developed PerturbFate, a high-throughput single-cell platform using combinatorial indexing and CRISPR interference to probe how many different genetic perturbations converge on similar cellular outcomes. They applied it to over 300,000 melanoma cells, testing more than 140 genes linked to vemurafenib resistance. Across these diverse perturbations, they observed a shared, dedifferentiated cell state driven by cooperative transcription factor activity. Examining Mediator complex components, they linked specific biochemical modules to convergent gene activation patterns. They also identified common regulatory nodes that produce similar phenotypes from unrelated genetic changes and showed how unrelated gene perturbations reshape cell states. PerturbFate offers a scalable way to connect multimodal gene regulation to disease-relevant phenotypes, aiding discovery of key regulators in resistance and other conditions.
Animals, Antigenic Variation, DNA Breaks, Double-Stranded, Immune Evasion, Rad51 Recombinase, Recombination, Genetic, Trypanosoma brucei brucei, Variant Surface Glycoproteins, Trypanosoma
Antigenic variation lets pathogens like Trypanosoma brucei dodge host antibodies by switching surface proteins, and this parasite further boosts its arsenal by generating new genes for variant surface glycoproteins (VSGs). Yet the mechanics were unclear due to a lack of tools. Researchers built a highly sensitive targeted sequencing method to track VSG diversification. They found that inducing a double-strand break in a VSG gene with Cas9 triggered recombination processes reliant on RAD51 and BRCA2, matching patterns seen during natural infections. The resulting new VSGs were antigenically distinct from the parent strain, enabling immune evasion. This work clarifies how T. brucei diversifies its antigen repertoire and offers an experimental approach to study these processes in other pathogens.
Comments (0)
No comments yet. Log in to leave a comment.
You must be logged in to leave a comment.
Login to Comment