FRC-QE: a robust and comparable 3D microscopy image quality metric for cleared organoids

mtFRC: depth-dependent resolution quantification of image features in 3D fluorescence microscopy

Motivation

Quantifying lateral resolution as a function of depth is important in the design of 3D microscopy experiments. However, for many specimens, resolution is non-uniform within the same optical plane because of factors such as tissue variability and differential light scattering. This precludes application of a simple resolution metric to the image as a whole. In such cases, it can be desirable to analyse resolution only within specific, well-defined features.

Results

An algorithm and software are presented to characterize resolution as a function of depth in features of arbitrary shape in 3D samples. The tool can be used to achieve an objective comparison between different preparation methods, imaging parameters, and optical systems. It can also inform the design of experiments requiring resolution of structures at a specific scale. The method is demonstrated by quantifying the improvement in resolution of two-photon microscopy over confocal in the central brain of Drosophila melanogaster. Measurement of image quality increases by tuning a single parameter, laser power, is also shown. An ImageJ plugin implementation is provided for ease of use via a simple Graphical User Interface, with outputs in table, graph, and colourmap formats.

Availability and implementation

Software and source code are available at https://www.imperial.ac.uk/rowlands-lab/resources/.

Princeton RAtlas: A Common Coordinate Framework for Fully cleared, Whole Rattus norvegicus Brains

Whole-brain clearing and imaging methods are becoming more common in mice but have yet to become standard in rats, at least partially due to inadequate clearing from most available protocols. Here, we build on recent mouse-tissue clearing and light-sheet imaging methods and develop and adapt them to rats. We first used cleared rat brains to create an open-source, 3D rat atlas at 25 μm resolution. We then registered and imported other existing labeled volumes and made all of the code and data available for the community (https://github.com/emilyjanedennis/PRA) to further enable modern, whole-brain neuroscience in the rat. Key features • This protocol adapts iDISCO (Renier et al., 2014) and uDISCO (Pan et al., 2016) tissue-clearing techniques to consistently clear rat brains. • This protocol also decreases the number of working hours per day to fit in an 8 h workday. Graphical overview.

Induced pluripotent stem cell-derived brain organoids as potential human model system for chemotherapy induced CNS toxicity

Neurotoxic phenomena are among the most common side effects of cytotoxic agents. The development of chemotherapy-induced polyneuropathy (CIPN) is a well-recognized adverse reaction in the peripheral nervous system, while changes of cognitive functions (post-chemotherapy cognitive impairment (PCCI)) are more diffuse and have only recently drawn scientific interest. PCCI in patients most often displays as short-term memory loss, reduced multitasking ability or deficits in language. Not least, due to a lack of preclinical human model systems, the underlying molecular mechanisms are poorly understood, and treatments are missing. We thus investigated whether induced pluripotent stem cell (iPSC)-derived brain organoids can serve as a human model system for the study of chemotherapy induced central nervous system toxicity. We robustly generated mature brain organoids from iPSC-derived neuronal precursor cells (NPC), which showed a typical composition with 1) dividing NPCs forming ventricle like structures 2) matured neurons and 3) supporting glial cells closer to the surface. Furthermore, upon stimulation the brain organoids showed functional signaling. When exposed to increasing concentrations of paclitaxel, a frequently used chemotherapy drug, we observed time dependent neurotoxicity with an EC50 of 153 nM, comparable to a published murine model system. Histological analysis after paclitaxel exposure demonstrated dose dependent apoptosis induction and reduced proliferation in the organoids with further Western blot analyses indicating the degradation of neuronal calcium sensor one protein (NCS-1) and activation of Caspase-3. We could also provide evidence that paclitaxel treatment negatively affects the pool of neuronal and astrocyte precursor cells as well as mature neurons. In summary our data suggests that human iPSC derived brain organoids are a promising preclinical model system to investigate molecular mechanisms underlying PCCI and to develop novel prevention and treatment strategies.

Fourier Ring Correlation and Anisotropic Kernel Density Estimation Improve Deep Learning Based SMLM Reconstruction of Microtubules

Single-molecule super-resolution microscopy (SMLM) techniques like dSTORM can reveal biological structures down to the nanometer scale. The achievable resolution is not only defined by the localization precision of individual fluorescent molecules, but also by their density, which becomes a limiting factor e.g., in expansion microscopy. Artificial deep neural networks can learn to reconstruct dense super-resolved structures such as microtubules from a sparse, noisy set of data points. This approach requires a robust method to assess the quality of a predicted density image and to quantitatively compare it to a ground truth image. Such a quality measure needs to be differentiable to be applied as loss function in deep learning. We developed a new trainable quality measure based on Fourier Ring Correlation (FRC) and used it to train deep neural networks to map a small number of sampling points to an underlying density. Smooth ground truth images of microtubules were generated from localization coordinates using an anisotropic Gaussian kernel density estimator. We show that the FRC criterion ideally complements the existing state-of-the-art multiscale structural similarity index, since both are interpretable and there is no trade-off between them during optimization. The TensorFlow implementation of our FRC metric can easily be integrated into existing deep learning workflows.

Perspective: Extending the Utility of Three-Dimensional Organoids by Tissue Clearing Technologies

An organoid, a self-organizing organ-like tissue developed from stem cells, can exhibit a miniaturized three-dimensional (3D) structure and part of the physiological functions of the original organ. Due to the reproducibility of tissue complexity and ease of handling, organoids have replaced real organs and animals for a variety of uses, such as investigations of the mechanisms of organogenesis and disease onset, and screening of drug effects and/or toxicity. The recent advent of tissue clearing and 3D imaging techniques have great potential contributions to organoid studies by allowing the collection and analysis of 3D images of whole organoids with a reasonable throughput and thus can expand the means of examining the 3D architecture, cellular components, and variability among organoids. Genetic and histological cell-labeling methods, together with organoid clearing, also allow visualization of critical structures and cellular components within organoids. The collected 3D data may enable image analysis to quantitatively assess structures within organoids and sensitively/effectively detect abnormalities caused by perturbations. These capabilities of tissue/organoid clearing and 3D imaging techniques not only extend the utility of organoids in basic biology but can also be applied for quality control of clinical organoid production and large-scale drug screening.