Peptide Conjugates of a 2'-O-Methoxyethyl Phosphorothioate Splice-Switching Oligonucleotide Show Increased Entrapment in Endosomes.

Antisense oligonucleotides (ASOs) are short, single-stranded nucleic acid molecules that alter gene expression. However, their transport into appropriate cellular compartments is a limiting factor in their potency. Here, we synthesized splice-switching oligonucleotides (SSOs) previously developed to treat the rare disease erythropoietic protoporphyria. Using chemical ligation-quantitative polymerase chain reaction (CL-qPCR), we quantified the SSOs in cells and subcellular compartments following free uptake. To drive nuclear localization, we covalently conjugated nuclear localization signal (NLS) peptides to a lead 2'-O-methoxyethyl phosphorothioate SSO using thiol-maleimide chemistry. The conjugates and parent SSO displayed similar RNA target-binding affinities. CL-qPCR quantification of the conjugates in cells and subcellular compartments following free uptake revealed one conjugate with better nuclear accumulation relative to the parent SSO. However, compared to the parent SSO, which altered the splicing of the target pre-mRNA, the conjugates were inactive at splice correction under free uptake conditions in vitro. Splice-switching activity could be conferred on the conjugates by delivering them into cells via cationic lipid-mediated transfection or by treating the cells into which the conjugates had been freely taken up with chloroquine, an endosome-disrupting agent. Our results identify the major barrier to the activity of the peptide-oligonucleotide conjugates as endosomal entrapment.

Deep learning-based behavioral analysis reaches human accuracy and is capable of outperforming commercial solutions.

To study brain function, preclinical research heavily relies on animal monitoring and the subsequent analyses of behavior. Commercial platforms have enabled semi high-throughput behavioral analyses by automating animal tracking, yet they poorly recognize ethologically relevant behaviors and lack the flexibility to be employed in variable testing environments. Critical advances based on deep-learning and machine vision over the last couple of years now enable markerless tracking of individual body parts of freely moving rodents with high precision. Here, we compare the performance of commercially available platforms (EthoVision XT14, Noldus; TSE Multi-Conditioning System, TSE Systems) to cross-verified human annotation. We provide a set of videos-carefully annotated by several human raters-of three widely used behavioral tests (open field test, elevated plus maze, forced swim test). Using these data, we then deployed the pose estimation software DeepLabCut to extract skeletal mouse representations. Using simple post-analyses, we were able to track animals based on their skeletal representation in a range of classic behavioral tests at similar or greater accuracy than commercial behavioral tracking systems. We then developed supervised machine learning classifiers that integrate the skeletal representation with the manual annotations. This new combined approach allows us to score ethologically relevant behaviors with similar accuracy to humans, the current gold standard, while outperforming commercial solutions. Finally, we show that the resulting machine learning approach eliminates variation both within and between human annotators. In summary, our approach helps to improve the quality and accuracy of behavioral data, while outperforming commercial systems at a fraction of the cost.