Real-time prediction of cell division timing in developing zebrafish embryo

Predicting the future direction of cell movement with convolutional neural networks

Image-based deep learning systems, such as convolutional neural networks (CNNs), have recently been applied to cell classification, producing impressive results; however, application of CNNs has been confined to classification of the current cell state from the image. Here, we focused on cell movement where current and/or past cell shape can influence the future cell movement. We demonstrate that CNNs prospectively predicted the future direction of cell movement with high accuracy from a single image patch of a cell at a certain time. Furthermore, by visualizing the image features that were learned by the CNNs, we could identify morphological features, e.g., the protrusions and trailing edge that have been experimentally reported to determine the direction of cell movement. Our results indicate that CNNs have the potential to predict the future direction of cell movement from current cell shape, and can be used to automatically identify those morphological features that influence future cell movement.

Modulation of first-passage time for gene expression via asymmetric cell division

How to balance the size of exponentially growing cells has always been a focus of biologists. Recent experiments have uncovered that the cell is divided into two daughter cells only when the level of time-keeper protein reaches a fixed threshold and cell division in prokaryote is not completely symmetric. The timing of cell division is essentially random because gene expression is stochastic, but cells seen to manage to have precise timing of cell division events. Although the inter-cellular variability of gene expression has attracted much attention, the randomness of event timing has been rarely studied. In our analysis, the timing of cell division is formulated as the first-passage time (denoted by FPT) for time-keeper protein’s level to cross a critical threshold firstly, we derive exact analytical formulae for the mean and noise of FPT based on stochastic gene expression model with asymmetric cell division. The results of numerical simulation show that the regulatory factors (division rate, newborn cell size, exponential growth rate and threshold) have significant influence on the mean and noise of FPT. We also show that both the increase of division rate and newborn cell size could reduce the mean of FPT and increase the noise of FPT, the larger the exponential growth rate is, the smaller the mean and noise of FPT will be; and the larger the threshold value is, the higher the mean of FPT is and the lower the noise is. In addition, compared with symmetric division, asymmetric division can reduce the mean of FPT and improve the noise of FPT. In summary, our results provide insight into the relationship between regulatory factors and FPT and reveal that asymmetric division is an effective mechanism to shorten the mean of FPT.

Zebrafish: an emerging real-time model system to study Alzheimer's disease and neurospecific drug discovery

Zebrafish (Danio rerio) is emerging as an increasingly successful model for translational research on human neurological disorders. In this review, we appraise the high degree of neurological and behavioural resemblance of zebrafish with humans. It is highly validated as a powerful vertebrate model for investigating human neurodegenerative diseases. The neuroanatomic and neurochemical pathways of zebrafish brain exhibit a profound resemblance with the human brain. Physiological, emotional and social behavioural pattern similarities between them have also been well established. Interestingly, zebrafish models have been used successfully to simulate the pathology of Alzheimer’s disease (AD) as well as Tauopathy. Their relatively simple nervous system and the optical transparency of the embryos permit real-time neurological imaging. Here, we further elaborate on the use of recent real-time imaging techniques to obtain vital insights into the neurodegeneration that occurs in AD. Zebrafish is adeptly suitable for Ca²⁺ imaging, which provides a better understanding of neuronal activity and axonal dystrophy in a non-invasive manner. Three-dimensional imaging in zebrafish is a rapidly evolving technique, which allows the visualisation of the whole organism for an elaborate in vivo functional and neurophysiological analysis in disease condition. Suitability to high-throughput screening and similarity with humans makes zebrafish an excellent model for screening neurospecific compounds. Thus, the zebrafish model can be pivotal in bridging the gap from the bench to the bedside. This fish is becoming an increasingly successful model to understand AD with further scope for investigation in neurodevelopment and neurodegeneration, which promises exciting research opportunities in the future.

Adaptable, illumination patterning light sheet microscopy

Minimally-invasive optical imaging requires that light is delivered efficiently to limit the detrimental impact of photodamage on delicate biological systems. Light sheet microscopy represents the exemplar in tissue specific optical imaging of small and mesoscopic samples alike. However, further gains towards gentler imaging require a more selective imaging strategy to limit exposure to multiple yet discrete tissues without overexposing the sample, particularly where the information content is sparse or particularly optically sensitive tissues are present. The development of sample-adaptive imaging techniques is crucial in pursuit of the next generation of smart, autonomous microscopes. Herein, we report a microscope capable of performing 4D (x, y, z, t) light patterning to selectively illuminate multiple, rapidly reconfigurable regions of interest while maintaining the rapid imaging speed and high contrast associated with light sheet microscopy. We illustrate this utility in living zebrafish larvae and phantom samples.