Search for: All records

Abstract A high-throughput drug screen revealed that veratridine (VTD), a natural plant alkaloid, induces expression of the anti-cancer protein UBXN2A in colon cancer cells. UBXN2A suppresses mortalin, a heat shock protein, with dominant roles in cancer development including epithelial–mesenchymal transition (EMT), cancer cell stemness, drug resistance, and apoptosis. VTD-dependent expression of UBXN2A leads to the deactivation of mortalin in colon cancer cells, making VTD a potential targeted therapy in malignant tumors with high levels of mortalin. VTD was used clinically for the treatment of hypertension in decades past. However, the discovery of newer antihypertensive drugs and concerns over potential neuro- and cardiotoxicity ended the use of VTD for this purpose. The current study aims to determine the safety and efficacy of VTD at doses sufficient to induce UBXN2A expression in a mouse model. A set of flow-cytometry experiments confirmed that VTD induces both early and late apoptosis in a dose-dependent manner. In vivo intraperitoneal (IP) administration of VTD at 0.1 mg/kg every other day (QOD) for 4 weeks effectively induced expression of UBXN2A in the small and large intestines of mice. Liquid chromatography–tandem mass spectrometry (LC–MS/MS) assays on tissues collected from VTD-treated animals demonstrated VTD concentrations in the low pg/mg range. To address concerns regarding neuro- and cardiotoxicity, a comprehensive set of behavioral and cardiovascular assessments performed on C57BL/6NHsd mice revealed that VTD generates no detectable neurotoxicity or cardiotoxicity in animals receiving 0.1 mg/kg VTD QOD for 30 days. Finally, mouse xenograft experiments in athymic nude mice showed that VTD can suppress tumor growth. The main causes for the failure of experimental oncologic drug candidates are lack of sufficient safety and efficacy. The results achieved in this study support the potential utility of VTD as a safe and efficacious anti-cancer molecule. 

Specialized domain knowledge is often necessary to ac- curately annotate training sets for in-depth analysis, but can be burdensome and time-consuming to acquire from do- main experts. This issue arises prominently in automated behavior analysis, in which agent movements or actions of interest are detected from video tracking data. To reduce annotation effort, we present TREBA: a method to learn annotation-sample efficient trajectory embedding for be- havior analysis, based on multi-task self-supervised learn- ing. The tasks in our method can be efficiently engineered by domain experts through a process we call “task program- ming”, which uses programs to explicitly encode structured knowledge from domain experts. Total domain expert effort can be reduced by exchanging data annotation time for the construction of a small number of programmed tasks. We evaluate this trade-off using data from behavioral neuro- science, in which specialized domain knowledge is used to identify behaviors. We present experimental results in three datasets across two domains: mice and fruit flies. Using embeddings from TREBA, we reduce annotation burden by up to a factor of 10 without compromising accuracy com- pared to state-of-the-art features. Our results thus suggest that task programming and self-supervision can be an ef- fective way to reduce annotation effort for domain experts. 

We propose a method for learning the posture and struc- ture of agents from unlabelled behavioral videos. Start- ing from the observation that behaving agents are gener- ally the main sources of movement in behavioral videos, our method, Behavioral Keypoint Discovery (B-KinD), uses an encoder-decoder architecture with a geometric bottle- neck to reconstruct the spatiotemporal difference between video frames. By focusing only on regions of movement, our approach works directly on input videos without requir- ing manual annotations. Experiments on a variety of agent types (mouse, fly, human, jellyfish, and trees) demonstrate the generality of our approach and reveal that our dis- covered keypoints represent semantically meaningful body parts, which achieve state-of-the-art performance on key- point regression among self-supervised methods. Addition- ally, B-KinD achieve comparable performance to supervised keypoints on downstream tasks, such as behavior clas- sification, suggesting that our method can dramatically re- duce model training costs vis-a-vis supervised methods.