Uncovering developmental time and tempo using deep learning

Multicell-Fold: geometric learning in folding multicellular life

During developmental processes such as embryogenesis, how a group of cells fold into specific structures, is a central question in biology. However, it remains a major challenge to understand and predict the behavior of every cell within the living tissue over time during such intricate processes. Here we present a geometric deep-learning model that can accurately capture the highly convoluted interactions among cells. We demonstrate that multicellular data can be represented with both granular and foam-like physical pictures through a unified graph data structure, considering both cellular interactions and cell junction networks. Using this model, we achieve interpretable 4-D morphological sequence alignment, and predicting cell rearrangements before they occur at single-cell resolution. Furthermore, using neural activation map and ablation studies, we demonstrate cell geometries and cell junction networks together regulate morphogenesis at single-cell precision. This approach offers a pathway toward a unified dynamic atlas for a variety of developmental processes.

Creatine and L-carnitine attenuate muscular laminopathy in the LMNA mutation transgenic zebrafish

Lamin A/C gene (LMNA) mutations contribute to severe striated muscle laminopathies, affecting cardiac and skeletal muscles, with limited treatment options. In this study, we delve into the investigations of five distinct LMNA mutations, including three novel variants and two pathogenic variants identified in patients with muscular laminopathy. Our approach employs zebrafish models to comprehensively study these variants. Transgenic zebrafish expressing wild-type LMNA and each mutation undergo extensive morphological profiling, swimming behavior assessments, muscle endurance evaluations, heartbeat measurement, and histopathological analysis of skeletal muscles. Additionally, these models serve as platform for focused drug screening. We explore the transcriptomic landscape through qPCR and RNAseq to unveil altered gene expression profiles in muscle tissues. Larvae of LMNA(L35P), LMNA(E358K), and LMNA(R453W) transgenic fish exhibit reduced swim speed compared to LMNA(WT) measured by DanioVision. All LMNA transgenic adult fish exhibit reduced swim speed compared to LMNA(WT) in T-maze. Moreover, all LMNA transgenic adult fish, except LMNA(E358K), display weaker muscle endurance than LMNA(WT) measured by swimming tunnel. Histochemical staining reveals decreased fiber size in all LMNA mutations transgenic fish, excluding LMNA(WT) fish. Interestingly, LMNA(A539V) and LMNA(E358K) exhibited elevated heartbeats. We recognize potential limitations with transgene overexpression and conducted association calculations to explore its effects on zebrafish phenotypes. Our results suggest lamin A/C overexpression may not directly impact mutant phenotypes, such as impaired swim speed, increased heart rates, or decreased muscle fiber diameter. Utilizing LMNA zebrafish models for drug screening, we identify L-carnitine treatment rescuing muscle endurance in LMNA(L35P) and creatine treatment reversing muscle endurance in LMNA(R453W) zebrafish models. Creatine activates AMPK and mTOR pathways, improving muscle endurance and swim speed in LMNA(R453W) fish. Transcriptomic profiling reveals upstream regulators and affected genes contributing to motor dysfunction, cardiac anomalies, and ion flux dysregulation in LMNA mutant transgenic fish. These findings faithfully mimic clinical manifestations of muscular laminopathies, including dysmorphism, early mortality, decreased fiber size, and muscle dysfunction in zebrafish. Furthermore, our drug screening results suggest L-carnitine and creatine treatments as potential rescuers of muscle endurance in LMNA(L35P) and LMNA(R453W) zebrafish models. Our study offers valuable insights into the future development of potential treatments for LMNA-related muscular laminopathy.

Dev-ResNet: automated developmental event detection using deep learning

Delineating developmental events is central to experimental research using early life stages, permitting widespread identification of changes in event timing between species and environments. Yet, identifying developmental events is incredibly challenging, limiting the scale, reproducibility and throughput of using early life stages in experimental biology. We introduce Dev-ResNet, a small and efficient 3D convolutional neural network capable of detecting developmental events characterised by both spatial and temporal features, such as the onset of cardiac function and radula activity. We demonstrate the efficacy of Dev-ResNet using 10 diverse functional events throughout the embryonic development of the great pond snail, Lymnaea stagnalis. Dev-ResNet was highly effective in detecting the onset of all events, including the identification of thermally induced decoupling of event timings. Dev-ResNet has broad applicability given the ubiquity of bioimaging in developmental biology, and the transferability of deep learning, and so we provide comprehensive scripts and documentation for applying Dev-ResNet to different biological systems.