SummaryCell populations can be strikingly heterogeneous, composed of multiple cellular states, each exhibiting stochastic noise in its gene expression. A major challenge is to disentangle these two types of variability and to understand the dynamic processes and mechanisms that control them. Embryonic stem cells (ESCs) provide an ideal model system to address this issue because they exhibit heterogeneous and dynamic expression of functionally important regulatory factors. We analyzed gene expression in individual ESCs using single-molecule RNA-FISH and quantitative time-lapse movies. These data discriminated stochastic switching between two coherent (correlated) gene expression states and burst-like transcriptional noise. We further showed that the “2i” signaling pathway inhibitors modulate both types of variation. Finally, we found that DNA methylation plays a key role in maintaining these metastable states. Together, these results show how ESC gene expression states and dynamics arise from a combination of intrinsic noise, coherent cellular states, and epigenetic regulation.

Quantitative single-cell time-lapse microscopy is a powerful method for analyzing gene circuit dynamics and heterogeneous cell behavior. We describe the application of this method to imaging bacteria by using an automated microscopy system. This protocol has been used to analyze sporulation and competence differentiation in Bacillus subtilis, and to quantify gene regulation and its fluctuations in individual Escherichia coli cells. The protocol involves seeding and growing bacteria on small agarose pads and imaging the resulting microcolonies. Images are then reviewed and analyzed using our laboratory's custom MATLAB analysis code, which segments and tracks cells in a frame-to-frame method. This process yields quantitative expression data on cell lineages, which can illustrate dynamic expression profiles and facilitate mathematical models of gene circuits. With fast-growing bacteria, such as E. coli or B. subtilis, image acquisition can be completed in 1 d, with an additional 1–2 d for progressing through the analysis procedure.

In plants mechanical signals pattern morphogenesis through the polar transport of the hormone auxin and through regulation of interphase microtubule (MT) orientation. To date, the mechanisms by which such signals induce changes in cell polarity remain unknown. Through a combination of time-lapse imaging, and chemical and mechanical perturbations, we show that mechanical stimulation of the SAM causes transient changes in cytoplasmic calcium ion concentration (Ca2+) and that transient Ca2+ response is required for downstream changes in PIN-FORMED 1 (PIN1) polarity. We also find that dynamic changes in Ca2+ occur during development of the SAM and this Ca2+ response is required for changes in PIN1 polarity, though not sufficient. In contrast, we find that Ca2+ is not necessary for the response of MTs to mechanical perturbations revealing that Ca2+ specifically acts downstream of mechanics to regulate PIN1 polarity response.

A barrier to utilizing machine learning in seasonal forecasting applications is the limited sample size of observational data for model training. To circumvent this issue, here we explore the feasibility of training various machine learning approaches on a large climate model ensemble, providing a long training set with physically consistent model realizations. After training on thousands of seasons of climate model simulations, the machine learning models are tested for producing seasonal forecasts across the historical observational period (1980-2020). For forecasting large-scale spatial patterns of precipitation across the western United States, here we show that these machine learning-based models are capable of competing with or outperforming existing dynamical models from the North American Multi Model Ensemble. We further show that this approach need not be considered a ‘black box’ by utilizing machine learning interpretability methods to identify the relevant physical processes that lead to prediction skill.

The method of microtubule tracking and dynamics analysis, presented here, improves upon the current means of manual and automated quantification of microtubule behavior. Key contributions are increasing accuracy and data volume, eliminating user bias and providing advanced analysis tools for the discovery of temporal patterns in cellular processes. By tracking the entire length of each resolvable microtubule, as opposed to only the tip, it is possible to boost dynamics studies with positional information that is virtually impossible to collect manually. We demonstrate the method on the analysis of a microtubule dataset, which was manually tracked and analyzed in the study of βIII-tubulin isoform. Our results show that automated recognition of temporal patterns in cellular processes offers a highly promising potential.