Nicholas T. Southern

The expression of genetic material governs brain development, differentiation, and function, and targeted manipulation of gene expression is required to understand contributions of gene function to health and disease states. Although recent improvements in CRISPR/dCas9 interference (CRISPRi) technology have enabled targeted transcriptional repression at selected genomic sites, integrating these techniques for use in non-dividing neuronal systems remains challenging. Previously, we optimized a dual lentivirus expression system to express CRISPR-based activation machinery in post-mitotic neurons. Here we used a similar strategy to adapt an improved dCas9-KRAB-MeCP2 repression system for robust transcriptional inhibition in neurons. We find that lentiviral delivery of a dCas9-KRAB-MeCP2 construct driven by the neuron-selective human synapsin promoter enabled transgene expression in primary rat neurons. Next, we demonstrate transcriptional repression using CRISPR sgRNAs targeting diverse gene promoters, and show superiority of this system in neurons compared to existing RNA interference methods for robust transcript specific manipulation at the complex Brain-derived neurotrophic factor (Bdnf) gene. Our findings advance this improved CRISPRi technology for use in neuronal systems for the first time, potentially enabling improved ability to manipulate gene expression states in the nervous system.

Domain insertion engineering is a powerful approach to juxtapose otherwise separate biological functions, resulting in proteins with new-to-nature activities. A prominent example are switchable protein variants, created by receptor domain insertion into effector proteins. Identifying suitable, allosteric sites for domain insertion, however, typically requires extensive screening and optimization. We present ProDomino, a machine learning pipeline to rationalize domain recombination, trained on a semisynthetic protein sequence dataset derived from naturally occurring intradomain insertion events. ProDomino robustly identifies domain insertion sites in proteins of biotechnological relevance, which we experimentally validated in Escherichia coli and human cells. Finally, we used light- and chemically regulated receptor domains as inserts and demonstrate the rapid, model-guided creation of potent, single-component opto- and chemogenetic protein switches. These include novel CRISPR-Cas9 and -Cas12a variants for inducible genome engineering in human cells. Our work enables one-shot domain insertion engineering and substantially accelerates the design of customized allosteric proteins.

Abstract Domain insertion engineering is a powerful approach to juxtapose otherwise separate biological functions, resulting in proteins with new-to-nature activities. A prominent example are switchable protein variants, created by receptor domain insertion into effector proteins. Identifying suitable, allosteric sites for domain insertion, however, typically requires extensive screening and optimization. We present ProDomino, a novel machine learning pipeline to rationalize domain recombination, trained on a semi-synthetic protein sequence dataset derived from naturally occurring intradomain insertion events. ProDomino robustly identifies domain insertion sites in proteins of biotechnological relevance, which we experimentally validated in E. coli and human cells. Finally, we employed light- and chemically regulated receptor domains as inserts and demonstrate the rapid, model-guided creation of potent, single-component opto- and chemogenetic protein switches. These include novel CRISPR-Cas9 and -Cas12a variants for inducible genome engineering in human cells. Our work enables one-shot domain insertion engineering and substantially accelerates the design of customized allosteric proteins.

The expression of genetic material governs brain development, differentiation, and function, and targeted manipulation of gene expression is required to understand contributions of gene function to health and disease states. Although recent improvements in CRISPR/dCas9 interference (CRISPRi) technology have enabled targeted transcriptional repression at selected genomic sites, integrating these techniques for use in non-dividing neuronal systems remains challenging. Previously, we optimized a dual lentivirus expression system to express CRISPR-based activation machinery in post-mitotic neurons. Here we used a similar strategy to adapt an improved dCas9-KRAB-MeCP2 repression system for robust transcriptional inhibition in neurons. We find that lentiviral delivery of a dCas9-KRAB-MeCP2 construct driven by the neuron-selective promoter human synapsin 1 enabled transgene expression in primary rat neurons. Next, we demonstrate transcriptional repression using CRISPR sgRNAs targeting diverse gene promoters, and show superiority of this system in neurons compared to existing RNA interference methods for robust transcript specific manipulation at the complex Brain-derived neurotrophic factor ( Bdnf ) gene. Our findings advance this improved CRISPRi technology for use in neuronal systems for the first time, potentially enabling improved ability to manipulate gene expression states in the nervous system.

Allostery, the transmission of locally induced conformational changes to distant functional sites, is a key mechanism for protein regulation. Artificial allosteric effectors enable remote manipulation of cell function; their engineering, however, is hampered by our limited understanding of allosteric residue networks. Here, we introduce a phage-assisted evolution platform for in vivo optimization of allosteric proteins. It applies opposing selection pressures to enhance activity and switchability of phage-encoded effectors and leverages retron-based recombineering to broadly explore fitness landscapes, introducing point mutations, insertions, and deletions. Applying this framework to the transcription factor AraC yielded near-binary optogenetic switches, with light-controlled activity spanning ~1000-fold dynamic range. Long-read sequencing across selection cycles enabled high-resolution tracking of evolving variant pools, revealing adaptive trajectories and context-dependent residue interactions. Mechanistically, we find that linker mutations promoting α-helix extension at the sensor-effector junction enhance conformational coupling between LOV2 and AraC. These variants emerge consistently across independently evolved pools, underscoring their functional relevance. Together, we develop a framework for the directed evolution of programmable allosteric switches in vivo. By coupling dynamic selection with deep mutational scanning and temporal sequencing, it enables both functional optimization and mechanistic insight into allosteric networks.