Prime Editing Advances for In Vivo Therapies

Broad researchers enhanced several prime editing components: the motifs that protect the guide pegRNA (in red), the reverse transcriptase enzyme (in purple), and delivery via lipid nanoparticles (yellow).
Image Credit: Susanna Hamilton, Broad Communications 

Scientific Frontline: Extended "At a Glance" Summary: Prime Editing Advancements

The Core Concept: Prime editing is a precise genome-editing technology that replaces disease-causing DNA sequences with corrected segments without requiring double-strand DNA breaks.

Key Distinction/Mechanism: Unlike traditional CRISPR systems that rely on blunt DNA breaks, prime editing utilizes a prime editing guide RNA (pegRNA) to instruct a reverse transcriptase enzyme to write new genetic information directly into a targeted DNA site. Recent advancements enhance this mechanism by increasing component stability and delivery efficiency for in vivo applications.

Major Frameworks/Components:

• pegRNA Stabilization: The use of laboratory evolution to discover and implement novel structural motifs that shield pegRNA, extending its cellular lifespan and abundance.
• AI-Guided Enzyme Optimization: The application of artificial intelligence to redesign the reverse transcriptase enzyme, yielding highly mutated variants that maintain potent editing capabilities while demonstrating greater cellular stability.
• Lipid Nanoparticle (LNP) Delivery: The optimization of RNA packaging workflows to efficiently deliver prime editing components directly to target tissues, successfully demonstrated in mouse models.

Complete Fruit Fly Connectome Mapped

The connectome maps how neurons in the fruit fly brain connect to those in its body via its spinal cord equivalent.
Image Credit: Tyler Sloan

Scientific Frontline: Extended "At a Glance" Summary: Complete Fruit Fly Connectome

The Core Concept: A complete connectome is a highly detailed, three-dimensional wiring diagram mapping all neural connections between the brain and the nerve cord (the spinal cord equivalent) of an adult fruit fly. This comprehensive map allows scientists to observe all neurons and their synaptic connections as a single, holistic functional unit.

Key Distinction/Mechanism: Unlike previous mapping efforts that isolated the brain, bridging the brain and nerve cord revealed that motor control is highly decentralized. Rather than relying on a central brain hub to command movement, actions like walking are managed primarily by local neural circuits in the appendages communicating directly with one another.

Major Frameworks/Components:

• Serial Sectioning and Electron Microscopy: The creation of thousands of microscopic slices of a single fruit fly, which were imaged at high resolution to capture millions of neurons.
• AI-Assisted 3D Mapping: The utilization of artificial intelligence tools to align, stitch, and render electron microscopy images into a cohesive spatial map.
• Synapse-Level Connectomics: The precise mapping of connections on an individual neuron-to-neuron basis across both the brain and the nerve cord.
• Distributed Local Modules: A neurobiological framework highlighting a shift from centralized brain control to distributed local circuits for motor function and complex behavior.

MouseMapper: AI Analyzes Bodies at the Cell Level

Whole-Body Analysis
MouseMapper automatically segments 31 organs and tissue types in a mouse while simultaneously mapping neural and immune cells throughout the body. This enables comprehensive multi-organ analyses in intact mice.
Image Credit: © Ertürk Lab | Helmholtz Munich

Scientific Frontline: Extended "At a Glance" Summary: MouseMapper AI-Powered Whole-Body Analysis

The Core Concept: MouseMapper is an advanced, AI-powered imaging and analytical system that enables the whole-body analysis of mice down to the single-cell level. It automatically maps neural pathways, immune cells, and organs to visualize pathological changes throughout the entire organism.

Key Distinction/Mechanism: Unlike classical AI systems built for single tasks, MouseMapper utilizes "foundation models"—large AI models trained on vast datasets to recognize general patterns. Combined with tissue clearing and light-sheet microscopy, this deep learning framework flexibly adapts to various datasets to systematically compare changes across 31 different organs and tissues.

Major Frameworks/Components: 

• Tissue Clearing and Light-Sheet Microscopy: Imaging techniques utilized to process and visualize the complex anatomy of the organism at high resolutions.
• Foundation Models: Deep learning AI structures trained to recognize generalized patterns, allowing the flexible mapping of the finest nerve structures and immune cell accumulations.
• Molecular Analysis Integration: The system flags conspicuous regions for further molecular examination to connect cellular damage to specific signaling pathways.

What Is: Xenobots

Scientific Frontline: Extended "At a Glance" Summary: What Are Xenobots? Programmable Biological Organisms

The Core Concept: Xenobots are microscopic, programmable biological machines constructed entirely from living cells without any genetic modification. Measuring less than a millimeter, they lack traditional mechanical parts and are entirely organic, biodegradable, and derived primarily from embryonic stem cells of the African clawed frog (Xenopus laevis).

Key Distinction/Mechanism: Unlike inorganic robots engineered with deterministic algorithms, Xenobots are developed using evolutionary algorithms on a supercomputer to optimize biological architectures for specific behavioral goals. They rely on morphological computation and autonomous self-assembly to exhibit ciliary locomotion, molecular memory, swarm intelligence, and kinematic self-replication—a purely mechanical, non-genetic form of reproduction.

Major Frameworks/Components:

• In Silico Morphogenesis: Supercomputer-driven evolutionary algorithms simulate and optimize cellular configurations, applying specific constraints and noise injection to overcome the "sim-to-real gap".
• Kinematic Self-Replication: Utilizing an AI-optimized "Pac-Man" topology to mechanically corral free-floating stem cells into functional offspring, effectively decoupling biological reproduction from genetic division.
• Transcriptomic Plasticity: An inherent cellular adaptation resulting in a "phylostratigraphic shift" toward ancient evolutionary gene expressions when stem cells are isolated from standard embryonic developmental pathways.
• Human-Derived Anthrobots: Motile, multicellular spheroids spontaneously cultivated from adult human tracheal cells that have demonstrated the ability to autonomously bridge and regenerate severed neural tissue in vitro.
• Neurobots: Engineered biobots augmented with neural precursor cells that successfully self-organize into functioning, calcium-firing neural networks capable of autonomous visual gene expression despite lacking eyes.

A Gene-Encoded Blueprint Tells Growing Neurons Which Brain Regions to Connect With

A 3D visualization of the 13 major regions in the mouse brain. Black dots mark the centers of the 213 subdivisions used by SPERRFY to analyze relationships between brain connectivity and gene activity patterns.
Image Credit: Koike et al., PNAS, 2026.
(CC BY 4.0)

Scientific Frontline: Extended "At a Glance" Summary: Genetic Neural Wiring and SPERRFY

The Core Concept: A newly decoded, gene-encoded blueprint functions as a spatial "wiring map" that guides growing nerve fibers (axons) to connect with the precise target regions in the developing brain.

Key Distinction/Mechanism: Unlike previous models that relied heavily on physical distance or isolated sensory circuits, researchers utilized SPERRFY—a machine learning method—to analyze the overlapping activity patterns of 763 genes across 213 brain regions. This approach demonstrated that gene expression gradients act as a "GPS," pairing source and target regions to predict whole-brain connectivity with high accuracy.

Major Frameworks/Components:

• SPERRFY Algorithm: A machine learning tool designed to decode unique molecular identities by matching the gene activity profiles of neuronal source and target regions.
• Gene Expression Gradients: Chemical signals that vary in strength and genetic activity, providing spatial coordinates for growing neurons.
• Dual-Level Map Operation: Broad genetic activity patterns outline the general organization between brain regions, while highly detailed patterns manage specific, localized connections.

Personalized vaccine shows promise against aggressive brain cancer

A WashU Medicine-led clinical trial conducted at Siteman Cancer Center has found that a personalized vaccine to treat glioblastoma is safe and could potentially improve outcomes. Trial participant Kim Garland (left) reviews a scan with the study’s primary investigator, Tanner Johanns, MD, PhD, a WashU Medicine oncologist.
Photo Credit: Courtesy of Scott Garland

Scientific Frontline: Extended "At a Glance" Summary: Personalized DNA Vaccine for Glioblastoma (GNOS-PV01)

The Core Concept: A personalized therapeutic DNA vaccine engineered to target unique neoantigens on a patient's tumor, stimulating the immune system to recognize and eliminate aggressive glioblastoma cells.

Key Distinction/Mechanism: Unlike conventional treatments, this DNA-based platform can target up to 40 unique tumor-specific proteins simultaneously. It successfully transforms immunologically "cold" tumors—which typically evade immune detection—into "hot" tumors vulnerable to targeted immune-mediated eradication.

Major Frameworks/Components:

• Neoantigen Identification: Utilizes computational algorithms to accurately identify and select cancer-specific proteins from various regions of an individual's tumor.
• Synthetic DNA Encoding: Custom-manufactures specialized DNA molecules that encode the unique information for each patient's tumor neoantigens.
• Adjuvant Immunotherapy Intervention: Administered post-operatively during a patient's recovery and radiation treatment to systematically boost immune-cell activity against remaining cancer cells.

What Is: Synthetic Biology

Scientific Frontline: Extended "At a Glance" Summary: Synthetic Biology

The Core Concept: Synthetic biology is a transformative discipline that merges the biological sciences with rigorous, quantitative engineering principles to fundamentally redesign genetic sequences and construct entirely new biological parts, devices, and systems from the ground up.

Key Distinction/Mechanism: Unlike traditional "top-down" genetic engineering, which relies on retrofitting existing, naturally occurring cells by splicing or modifying small collections of genes, synthetic biology utilizes a predictable, "bottom-up" approach. It treats biology as an engineering discipline, building complex biological circuits and dynamic cellular functions entirely from scratch using rational design and computer science.

Major Frameworks/Components:

• Core Engineering Principles: The strict enforcement of standardization, modularity, and abstraction to bypass biological chaos and render cellular processes as predictable as microchip manufacturing.
• The Abstraction Hierarchy: A multi-tiered framework designed to manage biological complexity by intentionally hiding information across four levels: DNA (informational substrate), Bioparts/BioBricks (standardized sequences encoding isolated functions), Devices (assembled parts for specific tasks like logic gates), and complex Biological Systems functioning within a host cell "chassis."
• The Design-Build-Test-Learn (DBTL) Cycle: An iterative manufacturing workflow reliant on computer-aided design (CAD) and thermodynamic simulations (Design), automated gene synthesis and robotics (Build), high-throughput screening and multi-omics (Test), and artificial intelligence/machine learning for data parsing (Learn).

Scalable cell imaging method could help make drug safety testing faster, cheaper

Human liver cells treated with a cancer drug are imaged under a microscope using a method called Cell Painting.
Image Credit: Axiom Bio 

Scientific Frontline: Extended "At a Glance" Summary: Cell Painting for Drug Safety Testing

The Core Concept: Cell Painting is a scalable, image-based cellular profiling method that utilizes fluorescent dyes and artificial intelligence to measure thousands of molecular and structural changes in human cells following chemical exposure.

Key Distinction/Mechanism: Unlike conventional cell-based toxicity tests that typically measure single endpoints, Cell Painting labels eight different cellular components and leverages a trained AI model to simultaneously analyze thousands of morphological changes. This high-content approach allows researchers to detect cellular harm—including specifically which proteins and biochemical pathways are affected—at much lower chemical concentrations and with significantly greater detail than traditional assays.

Major Frameworks/Components: 

• Fluorescent Multiplexing: The application of specific dyes to visualize eight distinct cellular compartments and structures simultaneously under a microscope.
• High-Throughput Image Analysis: Culturing human primary cells (such as liver cells), exposing them to over 1,000 different chemicals at varying concentrations, and capturing the resulting morphological data.
• Artificial Intelligence and Machine Learning: The deployment of predictive computational models trained on extensive cell-based toxicity datasets to recognize biological signatures indicative of adverse effects.

Survival training in a safe space

A group of meerkats. These African mammals use controlled learning to prepare their young for the dangers of everyday life.
Photo Credit: Dušan Veverkolog

Scientific Frontline: Extended "At a Glance" Summary: Protected Learning Environments in Animal Development

The Core Concept: Protected learning is a biological mechanism in which adult animals create staged, risk-mitigated developmental spaces, allowing offspring to safely acquire essential survival skills without facing immediate lethal consequences.

Key Distinction/Mechanism: Unlike unassisted trial-and-error learning in the wild—which poses a significant threat to inexperienced juveniles hunting dangerous prey—this process relies on graduated risk exposure (e.g., adult meerkats offering dead, then disarmed, then fully intact venomous scorpions to their young). A critical finding is that if the developmental environment is too safe and diverges significantly from reality (analogous to "helicopter parenting"), maladaptation occurs, leaving the animal unprepared to cope with genuine risks in adulthood.

Major Frameworks/Components:

• Two-Phase Learning Framework: A developmental model simulating the transition from a protected juvenile stage to an unprotected, hazardous adult environment.
• Dynamic Programming: A mathematical optimization method used to calculate the theoretically ideal behavioral strategy under varying environmental conditions.
• Reinforcement Learning: A computational approach employed to simulate the trial-and-error processes through which individuals acquire survival strategies over time.

Computational Biology: In-Depth Description

Computational Biology is the interdisciplinary science that uses computational approaches, mathematical modeling, and algorithmic analysis to understand biological systems and relationships. Its primary goal is to extract meaningful insights from vast biological datasets—such as genetic sequences, protein structures, and cell signaling pathways—to simulate biological processes and predict outcomes in living systems.

Blueprints for Designing T Cells that Kill

This image shows killer T cells surrounding and attacking a cancer cell. A new atlas developed by researchers at UC San Diego could make it possible to design custom T cells for immunotherapy to maximize patient benefit while minimizing potential negative effects.
Image Credit: National Institutes of Health/NIAID

Scientific Frontline: "At a Glance" Summary

• Main Discovery: A comprehensive genetic atlas of CD8+ T cell states was developed, identifying specific transcription factors that determine whether these immune cells persist as effective defenders or succumb to dysfunctional exhaustion.
• Methodology: Researchers utilized advanced computational modeling, gene editing, and in vivo mouse studies to map nine distinct T cell states and experimentally manipulated genetic switches to decouple the pathways regulating immune memory from those driving exhaustion.
• Key Data: The study identified nine distinct CD8+ T cell states and discovered two previously unknown transcription factors, ZSCAN20 and JDP2, which, when inhibited, restored tumor-killing capacity without sacrificing long-term immune memory.
• Significance: This research fundamentally challenges the long-standing scientific belief that T cell exhaustion is an inevitable byproduct of chronic immune activation, proving instead that exhaustion and protective memory are distinct, separable genetic programs.
• Future Application: These findings provide a blueprint for engineering "custom" T cells in adoptive cell transfer and CAR T-cell therapies that are programmed to resist burnout while maintaining long-term potency against cancer and chronic infections.
• Branch of Science: Immunology, Oncology, and Computational Biology.

Scientists use quantum biology, AI to sharpen genome editing tool

ORNL scientists developed a method that improves the accuracy of the CRISPR Cas9 gene editing tool used to modify microbes for renewable fuels and chemicals production. This research draws on the lab’s expertise in quantum biology, artificial intelligence and synthetic biology.
Illustration Credit: Philip Gray/ORNL, U.S. Dept. of Energy

Scientific Frontline: Extended "At a Glance" Summary: Enhancing CRISPR Cas9 with Quantum Biology and AI

The Core Concept: Scientists have combined principles of quantum biology with explainable artificial intelligence (AI) to significantly improve the accuracy and efficiency of the CRISPR Cas9 genome editing tool for modifying microbes.

Key Distinction/Mechanism: While traditional CRISPR models were developed for mammalian or model species and often failed in microbes due to differing chromosomal structures, this new approach analyzes the quantum chemical properties of nucleotides (like electron distribution). An explainable AI model is then used to predict the most effective guide RNAs, ensuring the Cas9 enzyme binds successfully to the targeted microbial DNA.

Major Frameworks/Components: 

• CRISPR Cas9 Machinery: A powerful bioengineering tool that uses a unique guide RNA to direct an enzyme to cleave and modify specific genetic code.
• Quantum Biology: A field bridging molecular biology and quantum chemistry, used here to investigate how the electronic structures of nucleotides affect chemical reactivity and conformational stability.
• Explainable Artificial Intelligence (Iterative Random Forest): An interpretable AI model trained on roughly 50,000 guide RNAs targeting E. coli. Unlike "black box" algorithms, it reveals the specific molecular mechanisms and rules that dictate optimal cutting efficiency.
• Functional Genomics: The biological discipline of linking specific genetic codes (genotype) to physical traits (phenotype).

How ‘viral dark matter’ may help mitigate climate change

A network-based ecological interaction analysis showed the diversity of RNA viral species was higher than expected in the Arctic and Antarctic.
Photo Credit: Tara Ocean Foundation

Scientific Frontline: Extended "At a Glance" Summary: Marine RNA Viruses and Carbon Export

The Core Concept: Researchers have identified 5,500 new marine RNA virus species, uncovering their vital ecological role in driving atmospheric carbon into permanent storage on the ocean floor.

Key Distinction/Mechanism: Unlike marine DNA viruses that predominantly infect bacteria, these marine RNA viruses primarily target microbial eukaryotes and fungi. They utilize "stolen" auxiliary metabolic genes (AMGs) to reprogram host metabolism, forcing hosts—such as algae—to grow larger, die, and sink, thereby exporting digestible carbon to the deep ocean.

Major Frameworks/Components:

• Utilization of computational genomics to reconstruct host-virus relationships from small RNA sequence fragments.
• Classification of RNA virus communities into four marine ecological zones: Arctic, Antarctic, Temperate/Tropical Epipelagic, and Temperate/Tropical Mesopelagic.
• Application of network-based ecological interaction analysis, revealing unexpectedly high RNA viral diversity in polar regions driven by intense competition for limited host species.
• Discovery of 72 functionally distinct auxiliary metabolic genes (AMGs) across 95 RNA viruses, functioning as tools to hijack cellular carbon processing.
• Mapping of 1,243 RNA virus species to carbon export pathways, isolating 11 highly conserved targets for future ecological modeling.

Breaking down fungal biofilm defenses provides potential path to treating sticky infections

David Andes is a professor of medicine at the
University of Wisconsin School of Medicine and Public Health.

Scientific Frontline: Extended "At a Glance" Summary: Fungal Biofilm Defenses

The Core Concept: A fungal biofilm is a sticky, armor-like extracellular matrix that encases clusters of disease-causing microorganisms, protecting them from host immune responses and antimicrobial treatments.

Key Distinction/Mechanism: Unlike free-floating pathogens, organisms within a biofilm secrete a complex array of proteins that directly drive resistance to antifungal drugs, such as fluconazole, and regulate the dispersion of cells throughout the body.

Major Frameworks/Components:

• Extracellular Matrix: The physical barrier consisting of secreted materials, including specialized proteins, that shields the fungi.
• Protein Regulation: Specific proteins dictate both the biofilm's structural integrity against antifungal agents and its capacity to release cells for systemic spread.
• Targeted Gene Mutation: Experimental genetic alterations creating Candida albicans strains that lack specific biofilm-producing proteins, rendering the fungus significantly more susceptible to existing drugs.
• Turbinmicin: A recently discovered antifungal agent capable of blocking the pathogen's ability to secrete protective biofilm components.

Pathways to production

A graphic illustration of the kind of retrosynthetic analysis conducted by RetSynth software developed at Sandia National Laboratories. Using a novel algorithm, the software identifies the biological or chemical reactions needed to create a desired biological product or compound.
(Graphic by Laura Hatfield)

Scientific Frontline: "At a Glance" Summary: Pathways to Production

• Main Discovery: Biologists at Sandia National Laboratories developed a stand-alone software program called RetSynth that uses a novel algorithm to sort through large, curated databases of biological and chemical reactions.
• Methodology: The platform employs retrosynthetic analysis to map out the biological and chemical steps required to engineer and modify cellular molecules, rapidly evaluating all possible production pathways to determine the most efficient sequences.
• Key Data: The software algorithm optimizes manufacturing routes based on specific metrics: the fewest required steps, the highest economic viability utilizing available resources, and the maximum achievable theoretical yield of the desired bioproduct.
• Significance: This technology substantially accelerates the traditionally slow research and development process for bioproduction by rendering clear visual pathways and offering customizable biological, chemical, or hybrid production options.
• Future Application: The software is being commercially licensed to dramatically reduce manufacturing waste and emissions while producing next-generation therapeutics, biofuels, industrial chemicals, cosmetics, and agricultural compounds.
• Branch of Science: Synthetic Biology, Computational Biology, Bioengineering.