Microbes have wide-ranging effects on host phenotype. However, whether these effects extend to the relationships among host traits remains unknown. We tested whether microbes affect phenotypic correlations among early life traits in the weedy legume Medicago lupulina. In a field common garden, we inoculated plants with two types of microbes: nitrogen-fixing bacteria and rhizosphere microbial communities. We found that microbes modify phenotypic integration (phenotypic correlations) in leaves, primarily due to effects on independent effects on leaflet area and trichome density, a defense trait. While microbial effects on leaflet size were associated with overall plant growth, variation in trichome density was decoupled from growth and not predicted by investment in mutualism. Our results highlight the unique insights that can come from a multivariate approach to organismal plasticity, and raise the intriguing possibility that microbes may have an outsized impact on defense traits and their integration with organismal function.

Genome size varies widely across eukaryotes, largely because of differences in non-coding DNA, but the physiological consequences of this variation remain unclear. To directly test how non-coding DNA abundance influences cellular physiology, we engineered a scalable genome-expansion system in the budding yeast S. cerevisiae that increases genome size while leaving the endogenous genome unchanged. By sequentially fusing yeast artificial chromosomes (YACs) carrying predominantly non-coding human DNA, we generated strains with up to 12.8 Mb of additional DNA, approximately doubling the native genome. Genome expansion reduced growth rate and increased cell size in proportion to the amount of non-coding DNA. Spike-in-normalized RNA-seq and ChIP-seq revealed that the non-coding DNA is pervasively transcribed, with a proportional amount of RNA polymerase II being redistributed from the endogenous genome to the added non-coding sequences. This resulted in a global decrease in the endogenous mRNA concentration. However, ribosome profiling and proteomics experiments revealed that there is little translation of YAC-associated transcripts. Our mathematical model shows that cellular growth rate decreases because non-coding DNA acts as a sink for transcriptional resources to lower the concentration of endogenous mRNA. Thus, our work links genome expansion to proliferative capacity and offers a mechanistic explanation for why the fastest-growing cells, such as yeast and bacteria, carry so little non-coding DNA.

The immune system uses paracrine signaling to spatially confine potent responses such as inflammation. A bio-orthogonal synthetic paracrine system could enable engineering of analogous multicellular circuits in which different cell types coordinate their functions in a spatially organized fashion. Here, using the plant hormone auxin as a bio-orthogonal chemical signal, we introduce programmable paracrine circuits that distribute sensing and effector functions to different cell types to spatially restrict responses in mouse xenografts. Cells engineered to express auxin biosynthetic genes generated auxin-dense regions with tunable length scales in vivo. This localized signaling ability enabled design of a multicellular sentinel-effector system, in which THP-1 sentinel cells conditionally produce auxin in regions expressing the tumor-specific antigen EGFRvIII, and Jurkat effector cells respond by locally modulating the activity of a chimeric antigen receptor (CAR). This two-cell type system was able to achieve localized activation of engineered effector cells in vivo. These results establish a foundation for engineering multicellular therapeutic systems that focus responses in specific tissue contexts or disease sites.

Methane is a potent greenhouse gas, nearly half of which is consumed anaerobically by anaerobic methanotrophic archaea (ANME) through methyl-coenzyme M reductase (Mcr). However, ANME cannot be grown as pure cultures, and obtaining active ANME Mcr in vitro remains extremely challenging, preventing previous efforts to engineer this key enzyme. Here, we used directed evolution in the methanogen Methanosarcina acetivorans to enhance ANME-1 Mcr (McrANME-1) activity for methane and carbon dioxide capture by selecting McrANME-1 variants with improved growth during methane-dependent cultivation. As a result, we discovered two beneficial substitutions in the catalytic α-subunit of McrANME-1, S60P and I154V, that increased biofilm growth as well as acetate production and methane capture. AlphaFold structural predictions suggest possible mechanistic explanations for these beneficial substitutions. These findings demonstrate that Mcr can be engineered to enhance methane and carbon dioxide capture, establishing a foundation for biological greenhouse gas mitigation and carbon utilization technologies.

The computational design of soluble analogues of membrane proteins has unlocked exciting opportunities for the integration of unique membrane protein functions into soluble proteins. Here, we use AF2seq to generate accurate soluble analogues of both animal and microbial rhodopsins, based on the membrane GPCR topology, and the microbial rhodopsin transmembrane fold. We characterize the analogues and demonstrate that they are well-folded and highly thermostable. Furthermore, they exhibit the expected red shift characteristic of retinal binding. Top-down mass spectrometry confirms the placement of retinal covalent attachment, while X-ray crystallography validates the structural fidelity of the microbial rhodopsin analogue. Notably, the microbial rhodopsin analogue retains the primary reaction of the retinal photocycle, closely matching that of the native membrane protein. Overall, this work advances the possibility to transfer unique membrane protein functions, such as retinal photoswitching, into the soluble proteome.

Viruses play indispensable roles in ecosystems and human health, yet deciphering their molecular functions remains challenging. Many viral protein annotations are incomplete or poorly characterized. Existing tools typically predict functional categories without linking to verifiable evidence, hindering the credibility of functional interpretation. Here, we present VirProtRAG, a viral protein function annotation framework that integrates information retrieval with evidence‑grounded knowledge generation. It introduces three task-adapted components: a hybrid retrieval module combining keyword‑based and semantic dense retrieval to maximize literature coverage, synonym‑expanded and rank‑aware retrieval with reciprocal rank fusion for improved search effectiveness, and literature quality and evidence‑oriented re‑ranking to enhance reliability and interpretability. Results show that hybrid retrieval strategy performed best, with quality and evidence features further enhancing re‑ranking. Compared with direct LLM prompting without retrieved literature, it consistently improves generation performance, underscoring the critical role of external knowledge. Finally, we built a searchable database comprising all 17,484 reviewed Swiss‑Prot viral proteins, supporting both sequence‑ and text‑based queries. VirProtRAG introduced 32.53% non-overlapping function annotations beyond existing expert curation, and independently supported 56.34% of sequence-inferred function points with retrieved literature. Case studies further demonstrate its capability to augment and refine the characterization of previously unannotated or poorly understood viral proteins.

Pangenomics quantifies the conserved and variable gene repertoire among genomes, but popular implementations ignore gene synteny. Graph-based approaches incorporate both gene homology and synteny, but become difficult to interpret due to pervasive rearrangements. Here we present network-pruning and graph-layout algorithms that enable interactive, synteny-aware quantification and visualization of gene conservation and variability. Applied to 29 genomes of the marine genus Undatipelagibacter (formerly SAR11 subclade Ia.3.VI), we find that genomic variability forms not a few hypervariable islands against a static backbone but a structured continuum, whose variable regions differ in scale, topology, function, and evolutionary character. Genome variation spans from ancient, specialized regions of hundreds of genes whose propensity to vary is conserved across genera, to single hypervariable genes shaped by epistatic co-selection with partners dispersed genome-wide, and shows that chromosomal context carries evolutionary information synteny-unaware pangenomics cannot capture, and some evolutionary processes act on entire functional subsystems throughout a pangenome.

Mutation is the ultimate mechanism that produces genetic novelty, and thus a central ingredient of evolution. Mutation rates are therefore thought to be tuned by natural selection, for example to optimize a delicate balance between the generation of adaptive diversity and the accumulation of deleterious mutations. As this selection occurs over very long time scales, models and simulations have been powerful tools to understand how mutation rate evolves and which factors influence it. Most simulation methods are nevertheless limited by the over-simplicity of the genotype-to-phenotype map they feature, especially regarding the encoding of mutation rate. We modified Aevol, an evolutionary simulator inspired by bacterial genomics with a realistic genome structure and a complex genotype-to-phenotype layer, to allow organisms to evolve genes coding for higher replication fidelity. This setup permits several degrees of realism absent in other models: mutation-rate modifier genes themselves experience a realistic distribution of effects of mutations and diminishing- returns epistasis, similarly to fitness modifiers. Moreover, a lower mutation rate comes with the trade-off of a larger genome to encode the genes improving replication fidelity. We use this setup to test hypotheses regarding the evolution of prokaryotic mutation rate, and its link with genome size and genetic drift. We found that evolution systematically increases replication fidelity, even when this results in lower fitness. We highlight two factors which limit the mutation rate decrease: genetic drift and the supply of gain-of-fidelity mutations.

Flavonoids constitute a large class of natural products widely investigated for their bioactive properties, with microbial production offering a potentially scalable alternative to plant extraction. However, achieving structural diversification of these compounds in microbial systems remains challenging, as modification of the flavonoid B-ring typically relies on downstream tailoring enzymes. An alternative strategy is to exploit the intrinsic promiscuity of the canonical flavanone biosynthesis pathway to introduce structural variation at an early stage. Here, we sought to improve microbial production of diverse flavanones by systematically leveraging pathway promiscuity. By constructing a combinatorial library of pathways comprising 4-coumarate-CoA ligase, chalcone synthase, and chalcone isomerase, we enabled the conversion of a panel of ring-substituted cinnamic acid precursors into ten natural and non-natural flavanones. In parallel, we established a genetically encoded biosensor based on the transcriptional regulator FdeR and demonstrated its responsiveness across all ten flavanones. Leveraging this biosensor for high-throughput screening, we performed directed evolution of chalcone synthases from Hordeum vulgare and Arabidopsis thaliana, identifying enzyme variants that led to improved production of O-methylated flavanones, including isosakuranetin, hesperetin, and homoeriodictyol, as well as fluoro-substituted flavanones. In addition, we demonstrated that specific variants of H. vulgare chalcone synthase promoted the formation of isoferuloyl-derived derailment products. Collectively, this work establishes the FdeR-based biosensor as a versatile platform for pathway and enzyme engineering, enabling efficient early-stage diversification of flavanones in microbial systems and providing insight into the mutational landscape of chalcone synthases.

Natural CRISPR-Cas9 systems rely on crRNA-tracrRNA duplexes to guide DNA targeting. Prior work showed that tracrRNAs could be reprogrammed to hybridize to cellular RNAs, resulting in their conversion into non-canonical crRNAs that enabled RNA detection and recording. However, the fate of the cellular RNA and the engineering opportunities it affords remain unexplored. Here, we show that the hybridized RNA is not inactivated, allowing the recruitment of Cas9 to the RNA duplex to drive RNA base editing and trans-splicing. Fusing ADAR2dd to dSpyCas9 and systematically engineering the reprogrammed tracrRNA (Rptr) enabled efficient and tunable A-to-I RNA editing, with on- and off-target profiles comparable to dCas13. The methodology extended to the compact CjeCas9 that could be further tailored for RNA targeting by deleting the HNH domain and mutating the PAM-interacting domain. Finally, utilizing Rptrs to block splicing enabled 3′ and 5′ RNA trans-splicing. Thus, Rptrs offer a versatile alternative to conventional Cas9 guide RNA architectures for programmable RNA manipulation.

Cas9 precision editing is increasingly predictable because guide, donor and target-context effects have been systematically characterized. Extending this framework to other nucleases is essential for installing variants outside convenient Cas9 target space. Cas12a provides a T-rich protospacer-adjacent motif (PAM) alternative, but determinants of efficient donor-templated Cas12a editing remain poorly defined. Here, we systematically dissected Cas12a precision editing in Saccharomyces cerevisiae across nuclease, direct repeat, expression, crRNA, donor, genomic context and time-course variables. Reporter and amplicon-sequencing assays showed that cleavage activity alone did not predict precise editing. Highly active configurations often reduced viability or lost edited alleles over time, whereas attenuated configurations better preserved programmed edits. Enhanced AsCas12a edited rapidly and tolerated shorter crRNAs, resulting in a narrower editing window, while an attenuated FnCas12a configuration edited more slowly but maintained higher viability and better distal-edit recovery. Alternative repair outcomes were rare, target-dependent, and further suppressed by LexA-FHA donor recruitment. To define design parameters at scale, we established a pooled Cas12a platform with 530 barcoded edit cassettes and recovered programmed edits for 70.2% of designs. Successful editing was reduced with TTTG PAMs, a C upstream of the PAM and at distal edit positions. Excluding these features increased the edited fraction to 85.4% and adding high predicted cleavage scores further elevated it to 91.4%. Applied retrospectively, these criteria also identified poorly edited loci in the targeted panels. Together, these data define design principles for Cas12a-mediated precision editing and establish a scalable platform for genome-scale pooled variant engineering and phenotyping in yeast.

We developed an AI-guided design pipeline that generated and validated non-natural microbial rhodopsins with spectral properties not yet known in nature. The pipeline comprised a three-stage in silico design, a genetic algorithm (GA) for sequence generation, a stacked LASSO and XGBoost machine-learning (ML) regressor for spectral prediction and fitness ranking, and a Markov-based sequence plausibility filter to enforce proton pumping like characteristics. Four candidate rhodopsins (APR1, APR2, APR6, and APR7) targeting blue light absorption were designed and AlphaFold3 structural modelling predicted retinal binding pocket architecture consistent with outward proton-pumping function. Experimental characterisation confirmed that all four variants absorbed light at ∼410 nm and significantly promoted the growth of Cupriavidus necator under blue light illumination. This study demonstrates that AI-enabled design can engineer proteins with no natural precedent, generating light-harvesting rhodopsins with novel spectral properties while preserving biological function, marking a significant advance in programmable synthetic biology.