Molecular Therapy Nucleic Acids

Exon-skipping therapy mediated by antisense oligonucleotides (ASOs) is expected to provide a therapeutic option for Duchenne muscular dystrophy (DMD). ASOs for exon skipping reported so far target a single continuous sequence in or around the target exon. In the present study, we investigated ASOs for exon 44 skipping (applicable to approximately 6% of all DMD patients) to improve activity by using a novel ASO design incorporating two connected sequences. Phosphorodiamidate morpholino oligomers targeting two separate sequences in exon 44 were created to simultaneously target two splicing regulators in exon 44, and their exon 44 skipping was measured. NS-089/NCNP-02 showed the highest skipping activity among the oligomers. NS-089/NCNP-02 also induced exon 44 skipping and dystrophin protein expression in cells from a DMD patient to whom exon 44 skipping is applicable. We also assessed the in vivo activity of NS-089/NCNP-02 by intravenous administration to cynomolgus monkeys. NS-089/NCNP-02 induced exon 44 skipping in skeletal and cardiac muscle of cynomolgus monkeys. In conclusion, NS-089/NCNP-02, an ASO with a novel connected-sequence design, showed both in vitro and in vivo exon-skipping activity.

Oligonucleotide therapeutics are an established class of drugs for the treatment of genetic disorders. Their clinical development is challenging, however, as they typically distribute poorly to extra-hepatic tissues after systemic injection. Here we tested the heteroduplex oligonucleotide (HDO) platform for systemic delivery of SMN2 splice-switching oligonucleotides of 2-O-methoxyethyl phosphorothioate or phosphorodiamidate morpholino oligomer chemistries. We first showed that splice-switching HDO cargoes correct SMN2 splicing in cells derived from spinal muscular atrophy (SMA) patients, and validated extra-hepatic activity in spinal cord and muscle in a mouse model of SMA following systemic delivery. Our study raises prospects for delivery of nusinersen, the 2-O-methoxylethyl phosphorothioate oligonucleotide therapy approved for SMA and currently delivered by intrathecal injection, by systemic injection exploiting the HDO chemistry platform. Our findings also suggest that oligonucleotide drugs lacking convincing in vivo efficacy in muscle tissue could be delivered effectively by the HDO technology.

Di-valent short interfering RNA (siRNA) is a promising therapeutic modality that enables sequence-specific modulation of a single target gene in the central nervous system (CNS). To treat complex neurodegenerative disorders, where pathogenesis is driven by multiple genes or pathways, di-valent siRNA must be able to silence multiple target genes simultaneously. Here we present a framework for designing unimolecular "dual-targeting" di-valent siRNAs capable of co-silencing two genes in the CNS. We reconfigured di-valent siRNA - in which two identical, linked siRNAs are made concurrently - to create linear di-valent siRNA - where two siRNAs are made sequentially attached by a covalent linker. This linear configuration, synthesized using commercially available reagents, enables incorporation of two different siRNAs to silence two different targets. We demonstrate that this dual-targeting di-valent siRNA is fully functional in the CNS of mice, supporting at least two months of maximal target silencing. Dual-targeting di-valent siRNA is highly programmable, enabling simultaneous modulation of two different disease-relevant gene pairs (e.g., Huntingtons disease: MSH3 and HTT; Alzheimers disease: APOE and JAK1) with similar potency to a mixture of single-targeting di-valent siRNAs against each gene. This work potentiates CNS modulation of virtually any pair of disease-related targets using a simple unimolecular siRNA.

Antisense oligonucleotides (ASOs) are a powerful class of drugs with the potential to treat a wide range of human diseases. However, the prediction of ASO efficacy remains challenging, as large-scale and costly experimental screens are typically required to identify optimal candidates for a specific therapeutic target. To address this challenge, we compiled ASO Atlas, a database comprising 188,521 RNase H-mediated ASO sequences targeting 334 unique genes with corresponding knockdown efficacy measurements extracted from published patents. Using ASO Atlas, we trained OligoAI, a deep learning model capable of jointly modelling RNA target context, ASO sequence, sugar and backbone chemistries, and dosage to predict in vitro efficacy. We experimentally validated OligoAI by targeting KCNT2, achieving a 5.72-fold reduction in screening effort compared to random selection. ASO Atlas provides the first systematic resource to rigorously evaluate hypotheses regarding key parameters in ASO design, including sequence composition, chemical modifications, and target region selection. Both ASO Atlas and OligoAI have been made freely accessible through an online web-tool with the aim of facilitating the accelerated optimisation of ASO design.

Duchenne muscular dystrophy (DMD) is caused by mutations in the DMD gene that disrupt the reading frame and abolish expression of functional dystrophin protein. Antisense oligonucleotides (ASO) can restore production of partially functional dystrophins by inducing exon skipping to restore the reading frame of dystrophin transcripts. While exon skipping is an FDA approved therapeutic strategy, there are currently no approved therapies for patients amenable to exon 44 skipping (8% of DMD patients). Here, we carried out a discovery campaign to identify phosphorothioate (PS) ASOs that efficiently induce exon 44 skipping and to define key sequence and chemistry features associated with activity. A tiling and micro-tiling approach with 18mer fully PS and 2-O-methoxyethyl (2MOE) modified ASOs in patient-derived myotubes identified five exonic target regions that promote skipping. ASO activity was strongly correlated across skeletal muscle and iPSC-derived cardiomyocytes, indicating similar exon 44 splicing regulation across cell types. Optimization studies showed that for 2MOE PS ASOs, 16-20mers were generally most active, while longer ASOs often had reduced activity due in part to impaired productive uptake into cells. Swapping out 2MOE modifications at both terminal positions for locked nucleic acids (LNAs) rarely improved activity and could also reduce it. Finally, promising candidates were tested in a humanized mouse model with an exon 44 skippable deletion, where one ASO outperformed others, inducing dose-dependent exon 44 skipping and dystrophin restoration in mouse models. These findings define practical design rules for exon 44-targeted ASOs and provide a foundation for therapeutic development.

Osteoarthritis (OA) is a progressive degenerative joint disorder characterized by cartilage degradation, chronic pain, and impaired joint function. The avascular nature of cartilage isolates chondrocytes from systemic circulation, presenting significant challenges for therapeutic intervention. Despite extensive efforts, no clinically effective disease-modifying osteoarthritis drugs (DMOADs) are currently available. Targeting chondrocyte-specific receptors has therefore emerged as a promising strategy. The osteoclast-associated receptor (OSCAR), expressed on chondrocytes, has been implicated in the regulation of cartilage homeostasis and OA pathogenesis. Here, we applied sBEAR (Structurally similar Bioactive compound Enrichment by Assay Repositioning), a bioactivity-driven virtual screening framework independent of target structural information, to identify small-molecule inhibitors of the OSCAR-collagen interaction. By mining large-scale bioactivity profiles, we identified adefovir (ADV) and brivudine (BRV), as candidate OSCAR inhibitors. Molecular docking analyses indicated that both compounds occupy the collagen-recognition pocket within the OSCAR D2 domain. Intra-articular administration of these compounds in a post-traumatic OA mouse model significantly attenuated OA progression and enhanced chondrocyte regeneration. Both compounds increased Sox9 expression, and transcriptomic analyses revealed that BRV reverses inflammatory and extracellular matrix-degrading transcriptional programs. Together, these findings establish OSCAR as a therapeutically actionable target in OA and highlight ADV and BRV as potential DMOAD candidates.

Upstream open reading frames (uORFs) are cis-regulatory motifs that are predicted to occur in the 5' untranslated region (UTR) of the majority of human protein-coding transcripts. uORFs are typically associated with repression of the downstream primary open reading frame (pORF) at either the level of translation, or by promoting mRNA turnover via the nonsense-mediated decay pathway. Interference with uORF activity provides a potential mechanism for targeted upregulation of the expression of specific transcripts. It was recently reported that steric block antisense oligonucleotides (ASOs) can bind to and mask uORF start codons in order to inhibit translation initiation, and thereby disrupt uORF-mediated gene regulation. Given the relative maturity of the oligonucleotide field, such a uORF blocking mechanism might have widespread therapeutic utility. Here, we re-synthesised three of the most potent ASOs targeting the RNASEH1 uORF described in the study by Liang et al. and investigated their potential for RNASEH1 protein upregulation. No upregulation (of endogenous or reporter protein expression) was observed with any of the oligonucleotides tested at doses ranging from 25 nM to 300 nM. Conversely, we observed downregulation of expression in some instances, consistent with well-established mechanisms of blocking ribosome procession. Experiments were performed using multiple transfection protocol setups, with care taken to replicate the conditions of the original study. Transfection efficiency was confirmed using a MALAT1-targeting gapmer ASO as a positive control. We conclude that previously-described RNASEH1 uORF-targeting steric block ASOs are incapable of upregulating pORF protein expression in our hands.

Delivery of therapeutic moieties to specific cell types, such as neurons remains a challenge. Genes present in neurons are also expressed in non-neuronal cell types such as glia where they mediate non-targeted related functions. Thus, non-specific targeting of these proteins/channels has numerous unwanted side effects, as is the case with current small molecules or drug therapies. Current methodologies that use nanoparticles, lipid-mediated uptake, or mannitol in conjunction with lipids to deliver double-stranded RNA (dsRNA) have yielded mixed and unreliable results. We used a neuroanatomical tracer (B subunit of Cholera Toxin (CTB)) that binds to the ganglioside receptors (GM1) expressed on cells, including primary sensory neurons to deliver encapsulated dsRNA. This approach greatly improved delivery of dsRNA to the desired cells by enhancing uptake, reducing vehicle-mediated toxicity and protecting nucleotides from degradation by endonucleases. The delivery complex is internalized, and once inside the cell, the dsRNA naturally dissociates itself from the carrier complex and is very effective in knocking down cognate targets, both in vivo and in vitro. Past methods have used CTB-fusion proteins or chemically modified oligos or DNA moieties that have been covalently conjugated to CTB. Furthermore, CTB conjugated to an antigen, protein, or chemically modified nucleic acid is a potent activator of immune cell (T and B cells, macrophages) response, whereas CTB admixed with antigens or unmodified nucleic acids does not evoke this immune response. Importantly, in our method, the nucleic acids are not covalently linked to the carrier molecules. Thus, our method holds strong potential for targeted delivery of therapeutic moieties for cell types expressing GM1 receptors, including neuronal cell types.

The rapid progress of mRNA therapeutics has underscored a persistent challenge: achieving high protein expression at low dose. The 5 untranslated region (5-UTR) is a key regulator of translation initiation efficiency, prompting the question of whether a simple and portable modification could enhance expression across diverse designs. Here, we systematically engineered short repeats of the Kozak "core" motif (5-GCCACC-3) immediately upstream of the start codon and evaluated constructs incorporated into two widely used human UTRs (APO and HBB) in HeLa and HEK293T cells. Translation enhancement displayed a non-monotonic dependence on the number of Kozak repeats, with three repetitions consistently outperforming the native sequence and any other configuration. In mice, intramuscular lipid nanoparticle delivery of the three-copy design increased luciferase expression by [~]4-fold relative to the wild-type context and by up to [~]23-fold compared to a licensed vaccine UTR benchmark, providing clear in vivo relevance. These findings demonstrate that fine-tuning AUG-proximal Kozak elements constitutes a broadly applicable, UTR-independent strategy to enhance translation efficiency, offering a simple yet powerful principle for dose-sparing mRNA design in therapeutic applications.

Antisense oligonucleotides (ASOs) were the first modality to pioneer targeted gene knockdown in the treatment of ALS caused by mutant superoxide dismutase 1 (SOD1). RNA interference (RNAi) is another mechanism of gene silencing with historically superior potency in which short interfering RNAs (siRNAs) guide the RNA-induced silencing complex (RISC) to cleave complementary transcripts. However, delivery to extrahepatic tissues like the central nerve system (CNS) has been a bottleneck in the clinical development of RNAi. Herein, we identify potent siRNA duplexes for the knockdown of human SOD1 (hSOD1) in which medicinal chemistry and conjugation to an accessory oligonucleotide (ACO) enables durable and potent activity in CNS tissues. Local delivery via intracerebroventricular (ICV) or intrathecal (IT) injection into SOD1G93A mice delayed disease progression and extended animal survival with superior efficacy compared to an ASO compound resembling Tofersen in sequence and chemistry. Treatment also prevented disease-related declines in motor function including improvements in animal mobility, muscle strength, and coordination. The ACO itself does not target any specific complementary nucleic acid sequence; rather, it imparts benefits conducive to bioavailability and delivery through its chemistry. The complete conjugate (i.e., siRNA-ACO) represents a novel modality for delivery of RNAi to the CNS in which we aim to pursue ALS as an exemplary indication for clinical development.

RNA interference (RNAi)-based medicines offer precise targeting of virtually any transcript, making them an appealing new drug class for addressing unmet needs in immune-oncological applications. While RNAi therapies show exceptional duration of effect in non-dividing cells, their efficacy in rapidly dividing cells, crucial for immune-oncology, remains largely unexplored. Unlike in non-dividing cells, full chemical modification in rapidly dividing cells has not consistently extended silencing duration, according to limited data available. In this study, we investigated key factors affecting the duration of effect for three main types of RNAi-based therapeutics (siRNA, miRNA mimics, and miRNA inhibitors) in rapidly dividing cancer and immune cells. Saturation of intracellular depots by multiple loading doses, a common strategy to prolong silencing duration in non-dividing hepatocytes, had minimal impact on siRNA duration of effect in rapidly dividing cells. However, modifying the antisense strand with a 5-(E)-vinylphosphonate (5-VP) to protect siRNAs from exonucleases and enhance AGO2 binding significantly extended siRNA silencing duration to over 30 days both in vitro and in vivo. For miRNA mimics, extensive stabilization of the antisense strand with phosphorothioates was not effective and led to reduced potency and silencing duration. Interestingly, a shorter duplex region commonly seen in therapeutic siRNAs partially rescued duration of silencing in miRNA mimics with extended phosphorothioate modifications. On the other hand, miRNA inhibitors demonstrated robust reversal of miRNA activity for an impressive 25 days in cancer cell lines. Our findings enable the rational design of the chemical architecture and administration regimens of RNAi-based therapies in oncology and immunology.

Targeted delivery and controlled expression of mRNA-LNPs are critical for the development of safe and effective mRNA medicines. However, efficient post-delivery regulation of mRNA-LNP expression remains challenging. In this study, we engineered segmented poly(A) tail variants that function as gene-specific regulatory elements for synthetic mRNAs. Specifically, we inserted microRNA target sites (MTS) for miR-122 or miR-142 at various positions of the poly(A) tail of synthesized luciferase reporter mRNA. These modifications significantly reduced luciferase expression in non-target tissues in vitro and in vivo, demonstrating position-dependent selective expression control. Furthermore, by incorporating triple-MTS sequences for miR-142, miR-126, and miR-148a in all possible combinations at the 5 end of the poly(A) tail, we identified triple-MTS arrangements that simultaneously decrease luciferase activities in three non-target hepatic cell types, while preserving robust expression in hepatocytes. The results highlighted the importance of MTS insertion order for optimal mRNA silencing. These triple-MTS modules significantly expanded the utility of single miRNA-responsive elements, enabling cell-type selective regulation for mRNA therapeutics. By complementing tissue-tropic delivery LNPs, our mRNA cargo regulatory elements have the potential to improve tissue and cell type selectivity as a novel platform for post-delivery regulation of mRNA-LNP.

Guanidine- bridged nucleic acid (GuNA) is a bridged nucleic acid analog with high binding affinity towards complementary strands along with high nuclease resistance. GuNA has been developed to improve pharmacokinetics and safety profiles of phosphorothioate modified gapmers. Here, we evaluated antisense oligonucleotides (ASOs) modified with combination of GuNA and 2-O-methoxyethyl (MOE) could significantly improve KD activity in vitro and in vivo. Long-term efficacy evaluation showed that intracerebroventricularly administered GuNA modified gapmers stayed active for over 24 weeks in mouse brain. Furthermore, we found that GuNA-modified gapmers could evade lymphocyte-derived immune responses and kept ASO-induced toxicity in check. Taken together, the results of this study demonstrated that GuNA modification can improve the potential of ASOs, especially in the central nervous system.

Antisense oligonucleotides (ASOs) are a versatile therapeutic modality for inborn genetic diseases. ASOs can induce skipping of ''dispensable'' exons containing disease-causing variants to rescue protein amount and function, but this approach has been studied for only a small number of genes. We developed a high-throughput in silico tool for assessing exon dispensability and designing exon-skipping ASO sequences. Parameters were optimized using known dispensable and in-frame indispensable exons. Across 72,644 exons of 5,057 disease genes, we identified thousands of new targets for exon-skipping ASOs (3.4% of exons with most stringent filters, 24.5% with less stringent filters) that collectively include 0.97%-15.6% of disease-causing variants in large-scale databases. To facilitate recognition of DNA variants potentially amenable to exon skipping as a therapeutic strategy, we established the HAWK-EYE database as a repository of exon dispensability predictions and corresponding in silico-optimized ASO sequences, available as an open-access web application (https://hawk-eye.research.sickkids.ca/). To illustrate translational utility, we experimentally validated a subset of the in silico-optimized ASO sequences that were generated for all Dispensable exons in the HAWK-EYE database, and showed that skipping a Dispensable exon in SOX5 preserves protein function using in vivo and in vitro assays. This scalable platform approach to exon-skipping ASOs will accelerate identification and testing of amenable genetic variants.

A six nucleotide repeat expansion in intron-1 of the C9ORF72 gene is the most common genetic mutation affecting individuals with Amyotrophic Lateral Sclerosis and Frontotemporal Dementia. Bi-directional transcription of the repeat expansion generates sense and antisense repeat RNAs that can then be translated in all reading frames to produce six distinct dipeptide repeat (DPR) proteins with unique termini. The precise site of translation initiation of these proteins within the C9ORF72 repeat expansion remains elusive. We used CRISPR-Cas9 genome editing and steric-blocking antisense oligonucleotides (ASOs) to investigate the contribution of different AUG codons in the antisense repeat RNA to the production of DPR proteins, poly(GP) and poly(PR) in C9ORF72 expansion carrier motor neurons and lymphoblast cells. We then utilized ASOs targeting C9ORF72 sense repeat RNA to examine whether sense or antisense RNA is the major source of the poly(GP) protein - a question for which conflicting evidence exists. We found that these ASOs reduced the intended sense RNA target, but also the antisense RNA, thus preventing the production of poly(PR). Our data highlights the importance of the sequences preceding the antisense CCCCGG repeat expansion for the synthesis of antisense DPR proteins and supports the use of sense C9ORF72 ASOs to prevent the accumulation of both sense- and antisense-dependent DPR proteins in C9ORF72 ALS/FTD.

Oligonucleotides are a promising genetic medicine for myotonic dystrophy type 1 (DM1), the most common adult-onset muscular dystrophy. However, poor muscle distribution of nucleic acid drugs after systemic administration has hindered drug development, and no curative treatment exists. Additionally, DM1 pathology requires drug localization to the nucleus, where pathogenic mutant RNA is trapped, posing challenges after endocytosis and endosomal escape. Here, we show that a locked nucleic acid oligonucleotide targeting mutant CUGexp RNA tracts, conjugated to a bottlebrush polymer, exhibited improved muscle distribution and potent correction of DM1-associated splicing at low nanomolar doses in a DM1 mouse model. Significant improvements in myotonia, body weight, and grip strength were observed. The conjugates were well tolerated after 12 weeks of weekly intravenous dosing. These results suggest that bottlebrush polymer bioconjugates may overcome key limitations of traditional antisense drugs for muscular dystrophies, with the potential as potent, durable, and cost-efficient DM1 therapies. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=78 SRC="FIGDIR/small/668589v1_ufig1.gif" ALT="Figure 1"> View larger version (21K): org.highwire.dtl.DTLVardef@e3961corg.highwire.dtl.DTLVardef@194225dorg.highwire.dtl.DTLVardef@5c470dorg.highwire.dtl.DTLVardef@3cc46a_HPS_FORMAT_FIGEXP M_FIG C_FIG

In vitro-transcribed messenger RNA (IVT mRNA) has emerged as a versatile protein expression platform with broad clinical potential. Current optimization strategies for IVT mRNA focus on untranslated regions (UTRs), mRNA stability, and codon usage, often guided by massively parallel screening and machine learning approaches. In contrast, the Kozak sequence, a key determinant of translation initiation, is often inconsistently incorporated into synthetic 5' UTR design, and its contribution to translation efficiency remains poorly defined. Here, we systematically varied the Kozak sequence across diverse UTR contexts and performed combinatorial optimization using synthetic, established, and viral UTRs to identify design principles for enhanced translation. We show that a single-nucleotide deviation from the consensus Kozak sequence consistently enhances protein expression across UTR contexts and coding sequences. This effect is conserved across in vitro and in vivo models, highlighting the generalizability of the optimized Kozak sequence. These findings redefine the role of the Kozak sequence in synthetic mRNA design and demonstrate its substantial contribution to translation efficiency when optimized, enabling improved mRNA-based therapeutics.

Dysregulation of the alternative splicing process results in aberrant mRNA transcripts, leading to dysfunctional proteins or nonsense-mediated decay that cause a wide range of mis-splicing diseases. Development of therapeutic strategies to target the alternative splicing process could potentially shift the mRNA splicing from disease isoforms to a normal isoform and restore functional protein. As a proof of concept, we focus on Stargardt disease (STGD1), an autosomal recessive inherited retinal disease caused by biallelic genetic variants in the ABCA4 gene. The splicing variants c.5461-10T>C and c.4773+3A>G in ABCA4 cause the skipping of exon 39-40 and exon 33-34 respectively. In this study, we compared the efficacy of different RNA-targeting systems to modulate these ABCA4 splicing defects, including four CRISPR-Cas13 systems (CASFx-1, CASFx-3, RBFOX1N-dCas13e-C and RBFOX1N-dPspCas13b-C) as well as an engineered U1 system (ExSpeU1). Using a minigene system containing ABCA4 variants in the human retinal pigment epithelium ARPE19, our results show that RBFOX1N-dPspCas13b-C is the best performing CRISPR-Cas system, which enabled up to 80% reduction of the mis-spliced ABCA4 c.5461-10T>C variants and up to 78% reduction of the ABCA4 c.4773+3A>G variants. In comparison, delivery of a single ExSpeU1 was able to effectively reduce the mis-spliced ABCA4 c.4773+3A>G variants by up to 84%. We observed that the effectiveness of CRISPR-based and U1 splicing regulation is strongly dependent on the sgRNA/snRNA targeting sequences, highlighting that optimal sgRNA/snRNA designing is crucial for efficient targeting of mis-spliced transcripts. Overall, our study demonstrated the potential of using RNA-targeting CRISPR-Cas technology and engineered U1 to reduce mis-spliced transcripts for ABCA4, providing an important step to advance the development of gene therapy to treat STGD1.

Background & Aims: Adeno-associated virus (AAV) vectors are attractive delivery vehicles for therapeutic gene delivery, and a notable feature of most AAVs is their natural tropism for the liver, which leads to significant hepatic uptake following systemic administration. In previous work, we identified 266G as a conserved motif on a variable region on the capsid of many commonly used AAV variants that controls liver uptake in both mice and non-human primates. This single amino acid could be functionally leveraged to engineer AAVs to either de-target from or enhance tropism to the liver. Here, we explored whether these observations extended to the human context. Methods: Two human hepatocyte models were tested: Fah-/-/Rag2-/-/Il2rg-/- (FRG) mice with humanized livers and a bioengineered human microliver platform in vitro. A barcoded AAV capsid library including standard control serotypes were used to assess the role of the 266G motif on gene transfer and transgene expression in both liver systems. Results: In vivo, 266G containing AAVs indeed targeted human hepatocytes superiorly, with some noted dependency on the degree of human-hepatocyte replacement in the chimeric mouse model. Initial studies in the micropatterned primary human hepatocyte co-culture model however demonstrated enrichment of heparin-binding AAVs, and not 266G variants. Notably, incorporation of polyethylene glycol (PEG) into the system modified the AAV transduction potential of those capsids including the liver-targeting motif, recapitulating the hepatocyte transduction pattern observed in vivo. Importantly, when PEG was used, the two human models, both at the DNA and RNA level, did correlate significantly. Conclusions: Our results showed the potential of a combinatorial AAV library for model validation and revealed the human microliver platform-PEG as a reliable system for the development of AAV therapeutics.

BackgroundThe gene for apolipoprotein E4 (ApoE4 E4) confers an increased risk for development and lowers the age of onset of Alzheimers disease (AD), and is a highly suitable target for CRISPR-based editing because ApoE4 differs from ApoE3 by a single nucleotide polymorphism in the codon for residue 112 that codes for arginine (CGC) in E4 and cysteine (TGC) in E3. Editing of E4 to E3 could lower the risk of AD or ameliorate E4-related AD phenotypes. For AD, in order to deliver CRISPR components across the blood-brain barrier to the brain, we have developed a delivery platform termed Synthetic Exosomes (SEs) - microfluidically-synthesized deformable nanovesicles approximately the size of natural exosomes that have the ability to cross the BBB and deliver cargo to the brain. Here, we describe our use of SEs carrying CRISPR to successfully edit E4 to E3 in brain tissue of an E4-expressing mouse model. MethodsSeveral CRISPR guide RNAs (gRNA) and Cytosine Base Editor (CBE) mRNAs were synthesized by chemical and in vitro transcription syntheses, respectively. Four combinations of gRNA and CBE mRNA were tested in vitro for their relative activity to edit the E4 (cytosine) to E3 (thymine) in E4-expressing neuroblastoma (E4-N2A) and human Kelly neuroblastoma cells, to assess which combination produced the highest E4 to E3 base editing efficiency. The CRISPR RNA combination with the highest efficiency was encapsulated in SEs and injected intravenously (IV) via the tail vein into an AD model E4-expressing (E4-5XFAD) transgenic mouse; as a negative control, an E4-5XFAD mouse was injected with empty SEs. Five days after injection, mice were euthanized and brain, liver, and buffy coat (white blood cells (WBC)) collected to determine the editing of E4 to E3 measured by Next Generation Sequencing. In addition, E3 mRNA was measured in the brain and liver and compared to the %E3 gene editing. ResultsThe highest gRNA+CBE mRNA editing efficiency was [~]50% in E4-N2A cells and the same gRNA+CBE combination delivered in SEs to Kelly neuroblastoma cells showed 6.5% editing efficiency. In the E4-5XFAD mouse in vivo, five days after IV delivery of a single dose of the highest-activity SE-CRISPR gRNA+CBE mRNA, the percent of E4 edited to E3 was 0.14% in brain, 0.8% in liver, and 0.36% in WBCs. As evidence of functional editing, SE-CRISPR-treated mice had 0.03% E3 mRNA in brain and 0.09% E3 mRNA in liver. ConclusionsWhile this level of ApoE4 to E3 editing achieved five days after a single IV injection of SE-CRISPR is small, it provides initial in vivo proof-of-concept that the ApoE4 gene can be successfully edited, and editing results in functional expression of ApoE3 mRNA. The findings presented herein supports further optimization of the SE-CRISPR approach to increase the level of editing in brain as part of clinical development of SE-CRISPR as a powerful novel therapeutic approach for AD.