Kinetic Analysis of Label-Free Microscale Collagen Gel Contraction Using Machine Learning-Aided Image Analysis

SpheroScan: A User-Friendly Deep Learning Tool for Spheroid Image Analysis

Background

In recent years, three-dimensional (3D) spheroid models have become increasingly popular in scientific research as they provide a more physiologically relevant microenvironment that mimics in vivo conditions. The use of 3D spheroid assays has proven to be advantageous as it offers a better understanding of the cellular behavior, drug efficacy, and toxicity as compared to traditional two-dimensional cell culture methods. However, the use of 3D spheroid assays is impeded by the absence of automated and user-friendly tools for spheroid image analysis, which adversely affects the reproducibility and throughput of these assays.

Results

To address these issues, we have developed a fully automated, web-based tool called SpheroScan, which uses the deep learning framework called Mask Regions with Convolutional Neural Networks (R-CNN) for image detection and segmentation. To develop a deep learning model that could be applied to spheroid images from a range of experimental conditions, we trained the model using spheroid images captured using IncuCyte Live-Cell Analysis System and a conventional microscope. Performance evaluation of the trained model using validation and test datasets shows promising results.

Conclusion

SpheroScan allows for easy analysis of large numbers of images and provides interactive visualization features for a more in-depth understanding of the data. Our tool represents a significant advancement in the analysis of spheroid images and will facilitate the widespread adoption of 3D spheroid models in scientific research. The source code and a detailed tutorial for SpheroScan are available at https://github.com/FunctionalUrology/SpheroScan.

A roadmap for developing and engineering in vitro pulmonary fibrosis models

Idiopathic pulmonary fibrosis (IPF) is a severe form of pulmonary fibrosis. IPF is a fatal disease with no cure and is challenging to diagnose. Unfortunately, due to the elusive etiology of IPF and a late diagnosis, there are no cures for IPF. Two FDA-approved drugs for IPF, nintedanib and pirfenidone, slow the progression of the disease, yet fail to cure or reverse it. Furthermore, most animal models have been unable to completely recapitulate the physiology of human IPF, resulting in the failure of many drug candidates in preclinical studies. In the last few decades, the development of new IPF drugs focused on changes at the cellular level, as it was believed that the cells were the main players in IPF development and progression. However, recent studies have shed light on the critical role of the extracellular matrix (ECM) in IPF development, where the ECM communicates with cells and initiates a positive feedback loop to promote fibrotic processes. Stemming from this shift in the understanding of fibrosis, there is a need to develop in vitro model systems that mimic the human lung microenvironment to better understand how biochemical and biomechanical cues drive fibrotic processes in IPF. However, current in vitro cell culture platforms, which may include substrates with different stiffness or natural hydrogels, have shortcomings in recapitulating the complexity of fibrosis. This review aims to draw a roadmap for developing advanced in vitro pulmonary fibrosis models, which can be leveraged to understand better different mechanisms involved in IPF and develop drug candidates with improved efficacy. We begin with a brief overview defining pulmonary fibrosis and highlight the importance of ECM components in the disease progression. We focus on fibroblasts and myofibroblasts in the context of ECM biology and fibrotic processes, as most conventional advanced in vitro models of pulmonary fibrosis use these cell types. We transition to discussing the parameters of the 3D microenvironment that are relevant in pulmonary fibrosis progression. Finally, the review ends by summarizing the state of the art in the field and future directions.

SpheroScan: a user-friendly deep learning tool for spheroid image analysis

BACKGROUND

In recent years, 3-dimensional (3D) spheroid models have become increasingly popular in scientific research as they provide a more physiologically relevant microenvironment that mimics in vivo conditions. The use of 3D spheroid assays has proven to be advantageous as it offers a better understanding of the cellular behavior, drug efficacy, and toxicity as compared to traditional 2-dimensional cell culture methods. However, the use of 3D spheroid assays is impeded by the absence of automated and user-friendly tools for spheroid image analysis, which adversely affects the reproducibility and throughput of these assays.

RESULTS

To address these issues, we have developed a fully automated, web-based tool called SpheroScan, which uses the deep learning framework called Mask Regions with Convolutional Neural Networks (R-CNN) for image detection and segmentation. To develop a deep learning model that could be applied to spheroid images from a range of experimental conditions, we trained the model using spheroid images captured using IncuCyte Live-Cell Analysis System and a conventional microscope. Performance evaluation of the trained model using validation and test datasets shows promising results.

CONCLUSION

SpheroScan allows for easy analysis of large numbers of images and provides interactive visualization features for a more in-depth understanding of the data. Our tool represents a significant advancement in the analysis of spheroid images and will facilitate the widespread adoption of 3D spheroid models in scientific research. The source code and a detailed tutorial for SpheroScan are available at https://github.com/FunctionalUrology/SpheroScan.

RIPK1 and RIPK3 regulate TNFα-induced β-cell death in concert with caspase activity

OBJECTIVE

Type 1 diabetes (T1D) is characterized by autoimmune-associated β-cell loss, insulin insufficiency, and hyperglycemia. Although TNFα signaling is associated with β-cell loss and hyperglycemia in non-obese diabetic mice and human T1D, the molecular mechanisms of β-cell TNF receptor signaling have not been fully characterized. Based on work in other cell types, we hypothesized that receptor interacting protein kinase 1 (RIPK1) and receptor interacting protein kinase 3 (RIPK3) regulate TNFα-induced β-cell death in concert with caspase activity.

METHODS

We evaluated TNFα-induced cell death, caspase activity, and TNF receptor pathway molecule expression in immortalized NIT-1 and INS-1 β-cell lines and primary mouse islet cells in vitro. Our studies utilized genetic and small molecule approaches to alter RIPK1 and RIPK3 expression and caspase activity to interrogate mechanisms of TNFα-induced β-cell death. We used the β-cell toxin streptozotocin (STZ) to determine the susceptibility of Ripk3+/+ and Ripk3-/- mice to hyperglycemia in vivo.

RESULTS

Expression of TNF receptor signaling molecules including RIPK1 and RIPK3 was identified in NIT-1 and INS-1 β cells and isolated mouse islets at the mRNA and protein levels. TNFα treatment increased NIT-1 and INS-1 cell death and caspase activity after 24-48 h, and BV6, a small molecule inhibitor of inhibitor of apoptosis proteins (IAPs) amplified this TNFα-induced cell death. RIPK1 deficient NIT-1 cells were protected from TNFα- and BV6-induced cell death and caspase activation. Interestingly, small molecule inhibition of caspases with zVAD-fmk (zVAD) did not prevent TNFα-induced cell death in either NIT-1 or INS-1 cells. This caspase-independent cell death was increased by BV6 treatment and decreased in RIPK1 deficient NIT-1 cells. RIPK3 deficient NIT-1 cells and RIPK3 kinase inhibitor treated INS-1 cells were protected from TNFα+zVAD-induced cell death, whereas RIPK3 overexpression increased INS-1 cell death and promoted RIPK3 and MLKL interaction under TNFα+zVAD treatment. In mouse islet cells, BV6 or zVAD treatment promoted TNFα-induced cell death, and TNFα+zVAD-induced cell death was blocked by RIPK3 inhibition and in Ripk3-/- islet cells in vitro. Ripk3-/- mice were also protected from STZ-induced hyperglycemia and glucose intolerance in vivo.

CONCLUSIONS

RIPK1 and RIPK3 regulate TNFα-induced β-cell death in concert with caspase activity in immortalized and primary islet β cells. TNF receptor signaling molecules such as RIPK1 and RIPK3 may represent novel therapeutic targets to promote β-cell survival and glucose homeostasis in T1D.

Machine learning approaches for biomolecular, biophysical, and biomaterials research

A fluent conversation with a virtual assistant, person-tailored news feeds, and deep-fake images created within seconds-all those things that have been unthinkable for a long time are now a part of our everyday lives. What these examples have in common is that they are realized by different means of machine learning (ML), a technology that has fundamentally changed many aspects of the modern world. The possibility to process enormous amount of data in multi-hierarchical, digital constructs has paved the way not only for creating intelligent systems but also for obtaining surprising new insight into many scientific problems. However, in the different areas of biosciences, which typically rely heavily on the collection of time-consuming experimental data, applying ML methods is a bit more challenging: Here, difficulties can arise from small datasets and the inherent, broad variability, and complexity associated with studying biological objects and phenomena. In this Review, we give an overview of commonly used ML algorithms (which are often referred to as "machines") and learning strategies as well as their applications in different bio-disciplines such as molecular biology, drug development, biophysics, and biomaterials science. We highlight how selected research questions from those fields were successfully translated into machine readable formats, discuss typical problems that can arise in this context, and provide an overview of how to resolve those encountered difficulties.

Contracting scars from fibrin drops

This paper describes a microscale fibroplasia and contraction model that is based on fibrin-embedded lung fibroblasts and provides a convenient visual readout of fibrosis. Cell-laden fibrin microgel drops are formed by aqueous two-phase microprinting. The cells deposit extracellular matrix (ECM) molecules such as collagen while fibrin is gradually degraded. Ultimately, the cells contract the collagen-rich matrix to form a compact cell-ECM spheroid. The size of the spheroid provides the visual readout of the extent of fibroplasia. Stimulation of this wound-healing model with the profibrotic cytokine TGF-β1 leads to an excessive scar formation response that manifests as increased collagen production and larger cell-ECM spheroids. Addition of drugs also shifted the scarring profile: the FDA-approved fibrosis drugs (nintedanib and pirfenidone) and a PAI-1 inhibitor (TM5275) significantly reduced cell-ECM spheroid size. Not only is the assay useful for evaluation of antifibrotic drug effects, it is relatively sensitive; one of the few in vitro fibroplasia assays that can detect pirfenidone effects at submillimolar concentrations. Although this paper focuses on lung fibrosis, the approach opens opportunities for studying a broad range of fibrotic diseases and for evaluating antifibrotic therapeutics.

Aqueous Two-Phase Systems and Microfluidics for Microscale Assays and Analytical Measurements

Phase separation is a common occurrence in nature. Synthetic and natural polymers, salts, ionic liquids, surfactants, and biomacromolecules phase separate in water, resulting in an aqueous two-phase system (ATPS). This review discusses the properties, handling, and uses of ATPSs. These systems have been used for protein, nucleic acid, virus, and cell purification and have in recent years found new uses for small organics, polysaccharides, extracellular vesicles, and biopharmaceuticals. Analytical biochemistry applications such as quantifying protein-protein binding, probing for conformational changes, or monitoring enzyme activity have been performed with ATPSs. Not only are ATPSs biocompatible, they also retain their properties at the microscale, enabling miniaturization experiments such as droplet microfluidics, bacterial quorum sensing, multiplexed and point-of-care immunoassays, and cell patterning. ATPSs include coacervates and may find wider interest in the context of intracellular phase separation and origin of life. Recent advances in fundamental understanding and in commercial application are also considered.