Superhuman cell death detection with biomarker-optimized neural networks.

Cellular events underlying neurodegenerative disease may be captured by longitudinal live microscopy of neurons. While the advent of robot-assisted microscopy has helped scale such efforts to high-throughput regimes with the statistical power to detect transient events, time-intensive human annotation is required. We addressed this fundamental limitation with biomarker-optimized convolutional neural networks (BO-CNNs): interpretable computer vision models trained directly on biosensor activity. We demonstrate the ability of BO-CNNs to detect cell death, which is typically measured by trained annotators. BO-CNNs detected cell death with superhuman accuracy and speed by learning to identify subcellular morphology associated with cell vitality, despite receiving no explicit supervision to rely on these features. These models also revealed an intranuclear morphology signal that is difficult to spot by eye and had not previously been linked to cell death, but that reliably indicates death. BO-CNNs are broadly useful for analyzing live microscopy and essential for interpreting high-throughput experiments.

Genetically encoded cell-death indicators (GEDI) to detect an early irreversible commitment to neurodegeneration.

Cell death is a critical process that occurs normally in health and disease. However, its study is limited due to available technologies that only detect very late stages in the process or specific death mechanisms. Here, we report the development of a family of fluorescent biosensors called genetically encoded death indicators (GEDIs). GEDIs specifically detect an intracellular Ca2+ level that cells achieve early in the cell death process and that marks a stage at which cells are irreversibly committed to die. The time-resolved nature of a GEDI delineates a binary demarcation of cell life and death in real time, reformulating the definition of cell death. We demonstrate that GEDIs acutely and accurately report death of rodent and human neurons in vitro, and show that GEDIs enable an automated imaging platform for single cell detection of neuronal death in vivo in zebrafish larvae. With a quantitative pseudo-ratiometric signal, GEDIs facilitate high-throughput analysis of cell death in time-lapse imaging analysis, providing the necessary resolution and scale to identify early factors leading to cell death in studies of neurodegeneration.

In Silico Labeling: Predicting Fluorescent Labels in Unlabeled Images.

Microscopy is a central method in life sciences. Many popular methods, such as antibody labeling, are used to add physical fluorescent labels to specific cellular constituents. However, these approaches have significant drawbacks, including inconsistency; limitations in the number of simultaneous labels because of spectral overlap; and necessary perturbations of the experiment, such as fixing the cells, to generate the measurement. Here, we show that a computational machine-learning approach, which we call "in silico labeling" (ISL), reliably predicts some fluorescent labels from transmitted-light images of unlabeled fixed or live biological samples. ISL predicts a range of labels, such as those for nuclei, cell type (e.g., neural), and cell state (e.g., cell death). Because prediction happens in silico, the method is consistent, is not limited by spectral overlap, and does not disturb the experiment. ISL generates biological measurements that would otherwise be problematic or impossible to acquire.

A Newly Defined and Xeno-Free Culture Medium Supports Every-Other-Day Medium Replacement in the Generation and Long-Term Cultivation of Human Pluripotent Stem Cells.

Human pluripotent stem cells (hPSCs) present an unprecedented opportunity to advance human health by offering an alternative and renewable cell resource for cellular therapeutics and regenerative medicine. The present demand for high quality hPSCs for use in both research and clinical studies underscores the need to develop technologies that will simplify the cultivation process and control variability. Here we describe the development of a robust, defined and xeno-free hPSC medium that supports reliable propagation of hPSCs and generation of human induced pluripotent stem cells (hiPSCs) from multiple somatic cell types; long-term serial subculturing of hPSCs with every-other-day (EOD) medium replacement; and banking fully characterized hPSCs. The hPSCs cultured in this medium for over 40 passages are genetically stable, retain high expression levels of the pluripotency markers TRA-1-60, TRA-1-81, Oct-3/4 and SSEA-4, and readily differentiate into ectoderm, mesoderm and endoderm. Importantly, the medium plays an integral role in establishing a cGMP-compliant process for the manufacturing of hiPSCs that can be used for generation of clinically relevant cell types for cell replacement therapy applications.

Investigating the role of the extracellular environment in modulating hepatic stellate cell biology with arrayed combinatorial microenvironments.

Hepatic stellate cells (HSCs) are a major cell type of the liver that are involved in liver homeostasis. Upon liver damage, HSCs exit their normally quiescent state and become activated, leading to an increase of their proliferation, production of abnormal extracellular matrix proteins (ECMPs) and inflammatory mediators, and eventually liver fibrosis and cirrhosis. Current in vitro approaches to identify components that influence HSC biology typically investigate one factor at a time and generally ignore the complex crosstalk among the myriad of components that comprise the microenvironments of quiescent or activated HSCs. Here we describe a high throughput screening (HTS) approach to identify factors that affect HSC biology. Specifically, we integrated the use of ECMPs and signaling molecules into a combinatorial cellular microarray technology platform, thereby creating comprehensive "microenvironments". Using this technology, we performed real-time simultaneous screening of the effects of hundreds of unique microenvironments composed of ECMPs and signaling molecules on HSC proliferation and activation. From these screens, we identified combinations of microenvironment components that differentially modulate the HSC phenotype. Furthermore, analysis of HSC responses revealed that the influences of Wnt signaling molecules on HSC fate are dependent on the ECMP composition in which they are presented. Collectively, our results demonstrate the utility of high-content, array-based screens to provide a better understanding of HSC biology. Our results indicate that array-based screens may provide an efficient means for identifying candidate signaling pathways to be targeted for anti-fibrotic therapies.

Defining long-term maintenance conditions of human embryonic stem cells with arrayed cellular microenvironment technology.

The optimization of defined growth conditions is necessary for the development of clinical application of human embryonic stem cells (hESCs). Current research has focused on developing defined media formulations for long-term culture of hESCs with little attention on the establishment of defined substrates for hESC proliferation and self-renewal. Presently available technologies are insufficient to address the full complement of factors that may regulate hESC proliferation and maintenance of pluripotency. Here, we report the application of a multifactorial array technology to identify fully defined and optimized culture conditions for the proliferation of hESCs. Through the systematic screening of extracellular matrix proteins (ECMPs) and other signaling molecules, we developed and characterized a completely defined culture system for the long-term self-renewal of three independent hESC lines. In the future, the novel array platform and analysis procedure presented here will be applied toward the directed differentiation of hESCs and maintenance of other stem and progenitor cell populations.