A fluorescent splice-switching mouse model enables high-throughput, sensitive quantification of antisense oligonucleotide delivery and activity.

While antisense oligonucleotides (ASOs) are used in the clinic, therapeutic development is hindered by the inability to assay ASO delivery and activity in vivo. Accordingly, we developed a dual-fluorescence, knockin mouse model that constitutively expresses mKate2 and an engineered EGFP that is alternatively spliced in the presence of ASO to induce expression. We first examined free ASO activity in the brain following intracerebroventricular injection revealing EGFP splice-switching is both ASO concentration and time dependent in major central nervous system cell types. We then assayed the impact of lipid nanoparticle delivery on ASO activity after intravenous administration. Robust EGFP fluorescence was observed in the liver and EGFP+ cells were successfully isolated using fluorescence-activated cell sorting. Together, these results show the utility of this animal model in quantifying both cell-type- and organ-specific ASO delivery, which can be used to advance ASO therapeutics for many disease indications.

Correlating the Structure and Gene Silencing Activity of Oligonucleotide-Loaded Lipid Nanoparticles Using Small-Angle X-ray Scattering.

With three FDA-approved products, lipid nanoparticles (LNPs) are under intensive development for delivering wide-ranging nucleic acid therapeutics. A significant challenge for LNP development is insufficient understanding of structure-activity relationship (SAR). Small changes in chemical composition and process parameters can affect LNP structure, significantly impacting performance in vitro and in vivo. The choice of polyethylene glycol lipid (PEG-lipid), one of the essential lipids for LNP, has been proven to govern particle size. Here we find that PEG-lipids can further modify the core organization of antisense oligonucleotide (ASO)-loaded LNPs to govern its gene silencing activity. Furthermore, we also have found that the extent of compartmentalization, measured by the ratio of disordered vs ordered inverted hexagonal phases within an ASO-lipid core, is predictive of in vitro gene silencing. In this work, we propose that a lower ratio of disordered/ordered core phases correlates with stronger gene knockdown efficacy. To establish these findings, we developed a seamless high-throughput screening approach that integrated an automated LNP formulation system with structural analysis by small-angle X-ray scattering (SAXS) and in vitro TMEM106b mRNA knockdown assessment. We applied this approach to screen 54 ASO-LNP formulations while varying the type and concentration of PEG-lipids. Representative formulations with diverse SAXS profiles were further visualized using cryogenic electron microscopy (cryo-EM) to help structural elucidation. The proposed SAR was built by combining this structural analysis with in vitro data. Our integrated methods, analysis, and resulting findings on PEG-lipid can be applied to rapidly optimize other LNP formulations in a complex design space.

Lipid nanoparticle delivery limits antisense oligonucleotide activity and cellular distribution in the brain after intracerebroventricular injection.

Antisense oligonucleotide (ASO) therapeutics are being investigated for a broad range of neurological diseases. While ASOs have been effective in the clinic, improving productive ASO internalization into target cells remains a key area of focus in the field. Here, we investigated how the delivery of ASO-loaded lipid nanoparticles (LNPs) affects ASO activity, subcellular trafficking, and distribution in the brain. We show that ASO-LNPs increase ASO activity up to 100-fold in cultured primary brain cells as compared to non-encapsulated ASO. However, in contrast to the widespread ASO uptake and activity observed following free ASO delivery in vivo, LNP-delivered ASOs did not downregulate mRNA levels throughout the brain after intracerebroventricular injection. This lack of activity was likely due to ASO accumulation in cells lining the ventricles and blood vessels. Furthermore, we reveal a formulation-dependent activation of the immune system post dosing, suggesting that LNP encapsulation cannot mask cellular ASO backbone-mediated toxicities. Together, these data provide insights into how LNP encapsulation affects ASO distribution as well as activity in the brain, and a foundation that enables future optimization of brain-targeting ASO-LNPs.

Predictive high-throughput screening of PEGylated lipids in oligonucleotide-loaded lipid nanoparticles for neuronal gene silencing.

Lipid nanoparticles (LNPs) are gaining traction in the field of nucleic acid delivery following the success of two mRNA vaccines against COVID-19. As one of the constituent lipids on LNP surfaces, PEGylated lipids (PEG-lipids) play an important role in defining LNP physicochemical properties and biological interactions. Previous studies indicate that LNP performance is modulated by tuning PEG-lipid parameters including PEG size and architecture, carbon tail type and length, as well as the PEG-lipid molar ratio in LNPs. Owing to these numerous degrees of freedom, a high-throughput approach is necessary to fully understand LNP behavioral trends over a broad range of PEG-lipid variables. To this end, we report a low-volume, automated, high-throughput screening (HTS) workflow for the preparation, characterization, and in vitro assessment of LNPs loaded with a therapeutic antisense oligonucleotide (ASO). A library of 54 ASO-LNP formulations with distinct PEG-lipid compositions was prepared using a liquid handling robot and assessed for their physiochemical properties as well as gene silencing efficacy in murine cortical neurons. Our results show that the molar ratio of anionic PEG-lipid in LNPs regulates particle size and PEG-lipid carbon tail length controls ASO-LNP gene silencing activity. ASO-LNPs formulated using PEG-lipids with optimal carbon tail lengths achieved up to 5-fold lower mRNA expression in neurons as compared to naked ASO. Representative ASO-LNP formulations were further characterized using dose-response curves and small-angle X-ray scattering to understand structure-activity relationships. Identified hits were also tested for efficacy in primary murine microglia and were scaled-up using a microfluidic formulation technique, demonstrating a smooth translation of ASO-LNP properties and in vitro efficacy. The reported HTS workflow can be used to screen additional multivariate parameters of LNPs with significant time and material savings, therefore guiding the selection and scale-up of optimal formulations for nucleic acid delivery to a variety of cellular targets.

Multivalent interactions make adherens junction-cytoskeletal linkage robust during morphogenesis.

Embryogenesis requires cells to change shape and move without disrupting epithelial integrity. This requires robust, responsive linkage between adherens junctions and the actomyosin cytoskeleton. Using Drosophila morphogenesis, we define molecular mechanisms mediating junction-cytoskeletal linkage and explore the role of mechanosensing. We focus on the junction-cytoskeletal linker Canoe, a multidomain protein. We engineered the canoe locus to define how its domains mediate its mechanism of action. To our surprise, the PDZ and FAB domains, which we thought connected junctions and F-actin, are not required for viability or mechanosensitive recruitment to junctions under tension. The FAB domain stabilizes junctions experiencing elevated force, but in its absence, most cells recover, suggesting redundant interactions. In contrast, the Rap1-binding RA domains are critical for all Cno functions and enrichment at junctions under tension. This supports a model in which junctional robustness derives from a large protein network assembled via multivalent interactions, with proteins at network nodes and some node connections more critical than others.

Asterless is a Polo-like kinase 4 substrate that both activates and inhibits kinase activity depending on its phosphorylation state.

Centriole assembly initiates when Polo-like kinase 4 (Plk4) interacts with a centriole "targeting-factor." In Drosophila, Asterless/Asl (Cep152 in humans) fulfills the targeting role. Interestingly, Asl also regulates Plk4 levels. The N-terminus of Asl (Asl-A; amino acids 1-374) binds Plk4 and promotes Plk4 self-destruction, although it is unclear how this is achieved. Moreover, Plk4 phosphorylates the Cep152 N-terminus, but the functional consequence is unknown. Here, we show that Plk4 phosphorylates Asl and mapped 13 phospho-residues in Asl-A. Nonphosphorylatable alanine (13A) and phosphomimetic (13PM) mutants did not alter Asl function, presumably because of the dominant role of the Asl C-terminus in Plk4 stabilization and centriolar targeting. To address how Asl-A phosphorylation specifically affects Plk4 regulation, we generated Asl-A fragment phospho-mutants and expressed them in cultured Drosophila cells. Asl-A-13A stimulated kinase activity by relieving Plk4 autoinhibition. In contrast, Asl-A-13PM inhibited Plk4 activity by a novel mechanism involving autophosphorylation of Plk4's kinase domain. Thus, Asl-A's phosphorylation state determines which of Asl-A's two opposing effects are exerted on Plk4. Initially, nonphosphorylated Asl binds Plk4 and stimulates its kinase activity, but after Asl is phosphorylated, a negative-feedback mechanism suppresses Plk4 activity. This dual regulatory effect by Asl-A may limit Plk4 to bursts of activity that modulate centriole duplication.

An ordered pattern of Ana2 phosphorylation by Plk4 is required for centriole assembly.

Polo-like kinase 4 (Plk4) initiates an early step in centriole assembly by phosphorylating Ana2/STIL, a structural component of the procentriole. Here, we show that Plk4 binding to the central coiled-coil (CC) of Ana2 is a conserved event involving Polo-box 3 and a previously unidentified putative CC located adjacent to the kinase domain. Ana2 is then phosphorylated along its length. Previous studies showed that Plk4 phosphorylates the C-terminal STil/ANa2 (STAN) domain of Ana2/STIL, triggering binding and recruitment of the cartwheel protein Sas6 to the procentriole assembly site. However, the physiological relevance of N-terminal phosphorylation was unknown. We found that Plk4 first phosphorylates the extreme N terminus of Ana2, which is critical for subsequent STAN domain modification. Phosphorylation of the central region then breaks the Plk4-Ana2 interaction. This phosphorylation pattern is important for centriole assembly and integrity because replacement of endogenous Ana2 with phospho-Ana2 mutants disrupts distinct steps in Ana2 function and inhibits centriole duplication.

Stu2 uses a 15-nm parallel coiled coil for kinetochore localization and concomitant regulation of the mitotic spindle.

XMAP215/Dis1 family proteins are potent microtubule polymerases, critical for mitotic spindle structure and dynamics. While microtubule polymerase activity is driven by an N-terminal tumor overexpressed gene (TOG) domain array, proper cellular localization is a requisite for full activity and is mediated by a C-terminal domain. Structural insight into the C-terminal domain's architecture and localization mechanism remain outstanding. We present the crystal structure of the Saccharomyces cerevisiae Stu2 C-terminal domain, revealing a 15-nm parallel homodimeric coiled coil. The parallel architecture of the coiled coil has mechanistic implications for the arrangement of the homodimer's N-terminal TOG domains during microtubule polymerization. The coiled coil has two spatially distinct conserved regions: CRI and CRII. Mutations in CRI and CRII perturb the distribution and localization of Stu2 along the mitotic spindle and yield defects in spindle morphology including increased frequencies of mispositioned and fragmented spindles. Collectively, these data highlight roles for the Stu2 dimerization domain as a scaffold for factor binding that optimally positions Stu2 on the mitotic spindle to promote proper spindle structure and dynamics.

TOG-tubulin binding specificity promotes microtubule dynamics and mitotic spindle formation.

XMAP215, CLASP, and Crescerin use arrayed tubulin-binding tumor overexpressed gene (TOG) domains to modulate microtubule dynamics. We hypothesized that TOGs have distinct architectures and tubulin-binding properties that underlie each family's ability to promote microtubule polymerization or pause. As a model, we investigated the pentameric TOG array of a Drosophila melanogaster XMAP215 member, Msps. We found that Msps TOGs have distinct architectures that bind either free or polymerized tubulin, and that a polarized array drives microtubule polymerization. An engineered TOG1-2-5 array fully supported Msps-dependent microtubule polymerase activity. Requisite for this activity was a TOG5-specific N-terminal HEAT repeat that engaged microtubule lattice-incorporated tubulin. TOG5-microtubule binding maintained mitotic spindle formation as deleting or mutating TOG5 compromised spindle architecture and increased the mitotic index. Mad2 knockdown released the spindle assembly checkpoint triggered when TOG5-microtubule binding was compromised, indicating that TOG5 is essential for spindle function. Our results reveal a TOG5-specific role in mitotic fidelity and support our hypothesis that architecturally distinct TOGs arranged in a sequence-specific order underlie TOG array microtubule regulator activity.