Babesia is a genus of apicomplexan parasites that infect red blood cells in vertebrate hosts. Pathology occurs during rapid replication cycles in the asexual blood stage of infection. Current knowledge of Babesia replication cycle progression and regulation is limited and relies mostly on comparative studies with related parasites. Due to limitations in synchronizing Babesia parasites, fine-scale time-course transcriptomic resources are not readily available. Single-cell transcriptomics provides a powerful unbiased alternative for profiling asynchronous cell populations. Here, we applied single-cell RNA sequencing to 3 Babesia species (B. divergens, B. bovis, and B. bigemina). We used analytical approaches and algorithms to map the replication cycle and construct pseudo-synchronized time-course gene expression profiles. We identify clusters of co-expressed genes showing “just-in-time” expression profiles, with gradually cascading peaks throughout asexual development. Moreover, clustering analysis of reconstructed gene curves reveals coordinated timing of peak expression in epigenetic markers and transcription factors. Using a regularized Gaussian graphical model, we reconstructed co-expression networks and identified conserved and species-specific nodes. Motif analysis of a co-expression interactome of AP2 transcription factors identified specific motifs previously reported to play a role in DNA replication in Plasmodium species. Finally, we present an interactive web application to visualize and interactively explore the datasets.

Babesia is a genus of Apicomplexan parasites that infect red blood cells in vertebrate hosts. Pathology occurs during rapid replication cycles in the asexual blood-stage of infection. Current knowledge of Babesia replication cycle progression and regulation is limited and relies mostly on comparative studies with related parasites. Due to limitations in synchronizing Babesia parasites, fine-scale time-course transcriptomic resources are not readily available. Single-cell transcriptomics provides a powerful unbiased alternative for profiling asynchronous cell populations. Here, we applied single-cell RNA sequencing to three Babesia species (B. divergens, B. bovis, and B. bigemina). We used analytical approaches and algorithms to map the replication cycle and construct pseudo-synchronized time-course gene expression profiles. We identify clusters of co-expressed genes showing just-in-time expression profiles, with gradually cascading peaks throughout asexual development. Moreover, clustering analysis of reconstructed gene curves reveals coordinated timing of peak expression in epigenetic markers and transcription factors. Using a regularized Gaussian Graphical Model, we reconstructed co-expression networks and identified conserved and species-specific nodes. Motif analysis of a co-expression interactome of AP2 transcription factors identified specific motifs previously reported to play a role in DNA replication in Plasmodium species. Finally, we present an interactive web-application to visualize and interactively explore the datasets.

No abstract

Background: Neoadjuvant chemotherapy (NACT) used for triple-negative breast cancer (TNBC) eradicates tumors in only 45% of patients. TNBC patients with substantial residual cancer burden have poor metastasis-free and overall survival rates. Our previous studies demonstrated mitochondrial oxidative phosphorylation (OXPHOS) was elevated, suggesting a unique therapeutic dependency of residual tumor cells that survived after NACT. However, mechanisms underlying this enhanced reliance on OXPHOS are yet unknown. Mitochondria are morphologically plastic organelles that cycle between fission and fusion to maintain mitochondrial integrity and metabolic homeostasis. Methods: We modeled residual disease in human TNBC cells by treating with chemotherapeutic agents at the IC50 of cell killing, then evaluating surviving cells after 48 hours of treatment. We modeled residual TNBC in orthotopic patient-derived xenograft (PDX) model (PIM001p) by treating with standard front-line NACT (Adriamycin + cyclophosphamide; AC), then longitudinally harvesting tumors prior to treatment, residual, and upon regrowth. We analyzed mitochondrial morphology, mtDNA content and integrity, mitochondrial oxygen consumption rate, and metabolomic flux. We developed a U-Net based deep learning model that automatically detects and quantifies mitochondrial features in transmission electron micrographs. To test the functional dependency of mitochondrial structure in TNBC, we perturbed mitochondrial fusion genetically (by knocking down the fusion-driving protein Optic Atrophy 1, OPA1) and pharmacologically (using the first-in-class small molecule OPA1 inhibitor, MYLS22). Results: Pharmacologic or genetic disruption of mitochondrial fusion and fission resulted in decreased or increased OXPHOS rate, respectively, in TNBC cells, revealing for the first time that mitochondria morphology regulates OXPHOS in TNBC. Upon comparing mitochondrial effects of conventional chemotherapies, we found that DNA-damaging agents (adriamycin, carboplatin) increased mitochondrial elongation, mitochondrial content, flux of glucose through the TCA cycle, and OXPHOS, whereas taxanes (paclitaxel, docetaxel) instead decreased mitochondrial elongation and OXPHOS rate. Increased levels of the short protein isoform of OPA1 were observed in residual cells that not killed by DNA-damaging chemotherapy treatment. Treatment of cells with adriamycin followed by MYLS22 or given concurrently with MYLS22 drastically decreased cell growth. Conversely, cells treated with adriamycin, inducing fusion, followed by the DRP1 inhibitor Mdivi-1, further inducing fusion, were less sensitive to adriamycin than were vehicle-treated cells. Further, we observed heightened OXPHOS, OPA1 protein levels, and mitochondrial elongation in residual tumors of the PDX model following AC treatment. We found that sequential treatment first with AC, thus inducing mitochondrial fusion and OXPHOS, followed by MYLS22 to inhibit OPA1 in residual tumors, was able to suppress mitochondrial fusion and OXPHOS and significantly inhibited residual tumor regrowth. Our deep-learning algorithm identified distinct changes in mitochondrial phenotypes in residual tumors of multiple PDX models. Treatment of non-chemotherapy-treated mice with the OPA1 inhibitor MYLS22 as a single agent had no effect on tumor growth, revealing that post-AC residual tumors have an enhanced dependency on mitochondrial fusion compared to treatment-naïve tumors. Taken together, our findings establish a functional role for mitochondrial structure in chemotherapeutic response and metabolic reprogramming, which may confer survival advantage to TNBC cells. These results suggest that pharmacologic perturbation of mitochondrial structure can overcome chemoresistance in TNBC cells when administered rationally based on our understanding of chemotherapy-induced mitochondrial adaptations. Citation Format: Lily Baek, Junegoo Lee, Katherine E. Pendleton, Mariah J. Berner, Emily Goff, Lin Tan, Sara Martinez, Iqbal Mahmud, Argenis Arriojas, Alexander Zhurkevich, Tao Wang, Matthew Meyer, Bora Lim, James P. Barrish, Weston Porter, Kourosh Zarringhalam, Philip L. Lorenzi, Gloria V. Echeverria. Mitochondrial structure and function adaptation in residual triple negative breast cancer cells surviving chemotherapy treatment [abstract]. In: Proceedings of the 2022 San Antonio Breast Cancer Symposium; 2022 Dec 6-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2023;83(5 Suppl):Abstract nr P6-11-14.

In reinforcement learning (RL), the ability to utilize prior knowledge from previously solved tasks can allow agents to quickly solve new problems. In some cases, these new problems may be approximately solved by composing the solutions of previously solved primitive tasks (task composition). Otherwise, prior knowledge can be used to adjust the reward function for a new problem, in a way that leaves the optimal policy unchanged but enables quicker learning (reward shaping). In this work, we develop a general framework for reward shaping and task composition in entropy-regularized RL. To do so, we derive an exact relation connecting the optimal soft value functions for two entropy-regularized RL problems with different reward functions and dynamics. We show how the derived relation leads to a general result for reward shaping in entropy-regularized RL. We then generalize this approach to derive an exact relation connecting optimal value functions for the composition of multiple tasks in entropy-regularized RL. We validate these theoretical contributions with experiments showing that reward shaping and task composition lead to faster learning in various settings.