Scripps Research Institute

Generalist biological artificial intelligence (GBAI) represents a transformative approach to modeling the 'language of life'-the flow of information from DNA to cellular function. This Review synthesizes rapid advances in biological AI to interpret and generate DNA, RNA, proteins and cellular systems. We chart a course toward comprehensive systems that can concurrently process and predict across these domains, performing several critical biological tasks simultaneously. Substantial opportunities lie in synergizing language and structural AI, leveraging specialized models and improving AI agents for autonomous discovery. After addressing challenges in data, biological complexity, scalability and experimental validation, GBAI has the potential to deepen our understanding of disease pathways and biomarkers, advance automated therapeutic design and evaluation, and integrate within virtual cells to meaningfully simulate biological activity.

SUMMARYModification by ubiquitination governs the half-lives of thousands of proteins that are fated for elimination by either the proteasome or autophagy pathways, depending on the intricate architectures of ubiquitin modification. This system mediates quality control for individual proteins, protein complexes, and organelles, as well as myriad purely regulatory functions. Here we provide a comprehensive survey of the ubiquitin-proteasome system (UPS), the scope of which is at present poorly defined. The UPS, with the inclusion of pathways involving ubiquitin-like modifiers, comprises in our estimate over 1400 distinct proteins in humans, a vast set of activities whose collective impact on the biology of the cell is pervasive. The UPS is an integral component of the proteostasis network (PN), the remainder of which we have also surveyed in recent studies. With the addition of molecular chaperones, proteins from autophagy-lysosome pathway, and related activities, the PN includes in total over 3100 components by our estimates. Comprehensive and systematic definition of these pathways should support a range of ongoing investigations in the areas of genomics, proteomics, biochemistry, cell biology, and disease research.

Most enveloped viruses rely on furin for maturation of their surface glycoprotein. In contrast, mammarenaviruses process their glycoprotein precursor (GPC) using host site-1 protease (S1P), yet the biological implications of this unique reliance on S1P remain unclear. Here, we characterized a furin-dependent recombinant form (rCl13-RRRR) of the persistent clone 13 variant of LCMV (rCl13). Although rCl13-RRRR exhibited fitness comparable to rCl13 in cultured cells, it was highly attenuated in vivo and failed to establish persistence in immunocompetent mice. Clearance of rCl13-RRRR required interferon and CD8+ T cells, and immunization with rCl13-RRRR conferred protective immunity against a subsequent lethal LCMV challenge. Our results demonstrate that S1P-mediated processing of GPC is a key determinant of mammarenavirus fitness and immune evasion in vivo and highlight S1P as a promising and druggable target for host-directed antiviral strategies against human pathogenic mammarenaviruses.

Web-based molecular graphics have transformed the interactive visualization of molecular data, leveraging modern web technologies that provide GPU acceleration, optimized JavaScript engines, and seamless access across devices without additional software installation. We present the graphics engine at the core of the Mol* toolkit, a high-performance, open-source framework that is widely adopted in academia and industry, including by the Protein Data Bank, UniProt, EMDB, and AlphaFold DB. The engine combines a comprehensive styling system with a suite of optimized rendering primitives, including real-time surface generation, to deliver both flexibility and visual fidelity. Efficient handling of large-scale molecular scenes is achieved through level-of-detail management, GPU instancing, spatial acceleration structures, and frustum/occlusion culling. A screen-space global illumination model provides scalable, high-quality lighting, while integrated AR/VR support enables immersive molecular exploration. Together, these capabilities enable engaging, real-time, high-fidelity visualization of molecular systems, across a wide range of scales, from single atoms to billion-atom mesoscale assemblies, demonstrating the strengths of a bespoke web-native rendering engine for molecular graphics, available at https://molstar.org.

Immune imprinting1 or original antigenic sin2 is a phenomenon whereby the immune system preferentially recalls its initial response to a related, often evolving pathogen after subsequent exposure. Despite its important implications for vaccine development, the causes of imprinting remain unclear. Here, to understand the basis and impact of imprinting by influenza A viruses, we characterized the B cell responses of young children after consecutive first infections with divergent H1N1 and H3N2 strains of influenza. Children had a primary but otherwise similar B cell response to that of adults. Adult B cells commonly cross-reacted with past strains using more stereotyped and mutated immunoglobulin genes, indicating substantial homosubtypic imprinting. In children, after consecutive heterosubtypic primary infections, up to 6% of memory B cells are H1/H3 cross-reactive and bind to the highly conserved central stalk epitope-a lead target for broadly protective vaccine candidates. Over 90% of these B cells had a higher affinity for the imprinting H3N2 strain, resulting in reduced breadth and neutralization potency against H1N1 strains. Mechanistically, the imprinting H3 strains and affected H1 strains shared a residue change in the stalk epitope (D46N) that was central to the nearly universal shift in reactivity, despite differing by only a single atomic group. In conclusion, imprinting by influenza viruses can cause a deleterious shift of nearly the entire memory recall response against key, conserved epitopes.

In China, low pathogenic avian influenza (LPAI) H3N8 virus is widespread among chickens and has recently caused three zoonotic infections, with the last one in 2023 being fatal. Here we evaluated the relative pandemic risk of this 2023 zoonotic H3N8 influenza virus, utilizing our previously published decision tree. Serological analysis indicated that a large proportion of the human population does not have any cross-neutralizing antibodies against this H3N8 strain. LPAI H3N8 displayed a dual affinity for α2-3 and α2-6 sialic acids and replicated efficiently in human bronchial epithelial cells. Furthermore, we observed H3N8 transmission via direct contact but not aerosols to ferrets with pre-existing H3N2 immunity. Although pre-existing H3N2 immunity resulted in a shortened disease course in ferrets, it did not reduce disease severity or replication in the respiratory tract. This study suggests that this zoonotic H3N8 strain has moderate pandemic potential and emphasizes the continued need for avian influenza surveillance.

Detecting antigen-specific lymphocytes is crucial for immune monitoring in vaccination, infection, cancer and autoimmunity. However, their low frequency and dispersed distribution make reliable detection challenging. We developed a strategy exploiting the functions of tissue-resident memory T cells (TRMs) to locally concentrate target circulating immune cells in the skin, enabling their non-invasive sampling using a microneedle (MN) skin patch. TRMs were first induced by antigen sensitization and subsequently restimulated by intradermal inoculation of the same antigen to trigger the 'alarm' and immune recruitment functions of these cells, leading to accumulation of antigen-specific T cells from the circulation over several days. In mouse vaccination models, MN patches with optimized hydrogel coatings effectively isolated thousands of live antigen-specific lymphocytes and innate immune cells. In a human participant with allergic contact dermatitis, allergen-induced TRM stimulation followed by MN application recovered diverse lymphocyte populations absent from untreated skin sites. These results establish TRM restimulation coupled with MN sampling as a non-invasive platform to monitor both local and systemic antigen-specific immune responses across disease or vaccination settings.

Described here is the synthesis and assembly of a new-generation mRNA-bifunctional lipid nanoparticle (mRNA-BLNP3) for selective delivery of mRNA to antigen-presenting cells (APCs). Compared to mRNA-BLNP1 and BLNP2 (BioRxiv,doi.org/10.1101/2023.12.26.572282), the mRNA-BLNP3, incorporating a glycolipid with the lipid moiety designed to target the CD1d receptor on dendritic cells (DCs) and the sugar head group targeting the mannose-binding receptors on APCs is more selective with stronger target-specific immune responses. It was shown that vaccination in mice with mRNA-BLNP1 elicited enhanced cytokine induction and antibody responses compared to traditional mRNA-LNPs, and the mRNA-BLNP2 vaccine, incorporating a mannose-glycolipid, further improved DC targeting. However, BLNP3, incorporating the glycolipid with an aryl-mannose head group targeting the mannose receptor on APCs and the same lipid moiety targeting the Cd1d receptor on DCs, showed superior lymph node targeting in vivo with reduced liver accumulation, enhanced mRNA expression in DCs and macrophages, and increased DC maturation. Immunization in mice with mRNA-BLNP3 elicited enhanced humoral and cellular immune responses compared to mRNA-BLNP1 and mRNA-BLNP2, with higher antigen-specific IgG titers and granzyme B–producing CD8+ T cells, demonstrating that BLNP3 is a promising bifunctional lipid nanoparticle for delivery of mRNA vaccines with improved efficacy and safety.

Recognition of dsRNA triggers antiviral defense mediated by PKR and OAS3/RNase L pathways through translational arrest and RNA decay. This is accompanied by assembly of distinct cytoplasmic ribonucleoprotein (RNP) condensates termed stress granules (SGs) and RNase L-dependent bodies (RLBs). Here we show that adenovirus infection differentially modulates dsRNA sensors and RNP granule assembly. Infection with splicing-defective ΔE4 mutant leads to dsRNA accumulation and activation of both PKR and OAS3/RNase L, promoting formation of RLB-like granules. In contrast, mutants lacking virus-associated (VA) RNAs trigger PKR activation and assembly of SGs despite absence of detectable dsRNA. Proteomic analysis revealed distinct protein compositions of canonical SGs and RLBs, which were reflected in virus-induced granules. While ΔVA-induced granules were PKR-dependent, ΔE4 mutants induced RLB-like granules independently of PKR and RNase L. In these cells, granule assembly coincided with translational arrest independent of eIF2α phosphorylation, indicating additional pathways linking nuclear dsRNA sensing to translational control and RNP granule assembly during viral infection. These findings provide novel insights into how distinct dsRNA sensors modulate translation and RNP condensates in response to stress.