Extracting multiple surfaces from 3D microscopy images in complex biological tissues with the Zellige software tool

Background

Efficient tools allowing the extraction of 2D surfaces from 3D-microscopy data are essential for studies aiming to decipher the complex cellular choreography through which epithelium morphogenesis takes place during development. Most existing methods allow for the extraction of a single and smooth manifold of sufficiently high signal intensity and contrast, and usually fail when the surface of interest has a rough topography or when its localization is hampered by other surrounding structures of higher contrast. Multiple surface segmentation entails laborious manual annotations of the various surfaces separately.

Results

As automating this task is critical in studies involving tissue-tissue or tissue-matrix interaction, we developed the Zellige software, which allows the extraction of a non-prescribed number of surfaces of varying inclination, contrast, and texture from a 3D image. The tool requires the adjustment of a small set of control parameters, for which we provide an intuitive interface implemented as a Fiji plugin.

Conclusions

As a proof of principle of the versatility of Zellige, we demonstrate its performance and robustness on synthetic images and on four different types of biological samples, covering a wide range of biological contexts.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12915-022-01378-0.

SARS-CoV-2 infection induces the dedifferentiation of multiciliated cells and impairs mucociliary clearance

Understanding how SARS-CoV-2 spreads within the respiratory tract is important to define the parameters controlling the severity of COVID-19. Here we examine the functional and structural consequences of SARS-CoV-2 infection in a reconstructed human bronchial epithelium model. SARS-CoV-2 replication causes a transient decrease in epithelial barrier function and disruption of tight junctions, though viral particle crossing remains limited. Rather, SARS-CoV-2 replication leads to a rapid loss of the ciliary layer, characterized at the ultrastructural level by axoneme loss and misorientation of remaining basal bodies. Downregulation of the master regulator of ciliogenesis Foxj1 occurs prior to extensive cilia loss, implicating this transcription factor in the dedifferentiation of ciliated cells. Motile cilia function is compromised by SARS-CoV-2 infection, as measured in a mucociliary clearance assay. Epithelial defense mechanisms, including basal cell mobilization and interferon-lambda induction, ramp up only after the initiation of cilia damage. Analysis of SARS-CoV-2 infection in Syrian hamsters further demonstrates the loss of motile cilia in vivo. This study identifies cilia damage as a pathogenic mechanism that could facilitate SARS-CoV-2 spread to the deeper lung parenchyma.

DNABarcodeCompatibility: an R-package for optimizing DNA-barcode combinations in multiplex sequencing experiments

Summary

Using adequate DNA barcodes is essential to unambiguously identify each DNA library within a multiplexed set of libraries sequenced using next-generation sequencers. We introduce DNABarcodeCompatibility, an R-package that allows one to design single or dual-barcoding multiplex experiments by imposing desired constraints on the barcodes (including sequencer chemistry, barcode pairwise minimal distance and nucleotide content), while optimizing barcode frequency usage, thereby allowing one to both facilitate the demultiplexing step and spare expensive library-preparation kits. The package comes with a user-friendly interface and a web app developed in Java and Shiny (https://dnabarcodecompatibility.pasteur.fr), respectively, with the aim to help bridge the expertise of core facilities with the experimental needs of non-experienced users.

Availability and implementation

DNABarcodeCompatibility can be easily extended to fulfil specific project needs. The source codes of the R-package and its user interfaces are publicly available along with documentation at [https://github.com/comoto-pasteur-fr] under the GPL-2 licence.

Supplementary information

Supplementary data are available at Bioinformatics online.

PreMosa: extracting 2D surfaces from 3D microscopy mosaics

Motivation

A significant focus of biological research is to understand the development, organization and function of tissues. A particularly productive area of study is on single layer epithelial tissues in which the adherence junctions of cells form a 2D manifold that is fluorescently labeled. Given the size of the tissue, a microscope must collect a mosaic of overlapping 3D stacks encompassing the stained surface. Downstream interpretation is greatly simplified by preprocessing such a dataset as follows: (i) extracting and mapping the stained manifold in each stack into a single 2D projection plane, (ii) correcting uneven illumination artifacts, (iii) stitching the mosaic planes into a single, large 2D image and (iv) adjusting the contrast.

Results

We have developed PreMosa, an efficient, fully automatic pipeline to perform the four preprocessing tasks above resulting in a single 2D image of the stained manifold across which contrast is optimized and illumination is even. Notable features are as follows. First, the 2D projection step employs a specially developed algorithm that actually finds the manifold in the stack based on maximizing contrast, intensity and smoothness. Second, the projection step comes first, implying all subsequent tasks are more rapidly solved in 2D. And last, the mosaic melding employs an algorithm that globally adjusts contrasts amongst the 2D tiles so as to produce a seamless, high-contrast image. We conclude with an evaluation using ground-truth datasets and present results on datasets from Drosophila melanogaster wings and Schmidtae mediterranea ciliary components.

Availability and Implementation

PreMosa is available under https://cblasse.github.io/premosa.

Contact

blasse@mpi-cbg.de or myers@mpi-cbg.de.

Supplementary information

Supplementary data are available at Bioinformatics online.

Cell dynamics underlying oriented growth of the Drosophila wing imaginal disc

Quantitative analysis of the dynamic cellular mechanisms shaping the Drosophila wing during its larval growth phase has been limited, impeding our ability to understand how morphogen patterns regulate tissue shape. Such analysis requires explants to be imaged under conditions that maintain both growth and patterning, as well as methods to quantify how much cellular behaviors change tissue shape. Here, we demonstrate a key requirement for the steroid hormone 20-hydroxyecdysone (20E) in the maintenance of numerous patterning systems in vivo and in explant culture. We find that low concentrations of 20E support prolonged proliferation in explanted wing discs in the absence of insulin, incidentally providing novel insight into the hormonal regulation of imaginal growth. We use 20E-containing media to observe growth directly and to apply recently developed methods for quantitatively decomposing tissue shape changes into cellular contributions. We discover that whereas cell divisions drive tissue expansion along one axis, their contribution to expansion along the orthogonal axis is cancelled by cell rearrangements and cell shape changes. This finding raises the possibility that anisotropic mechanical constraints contribute to growth orientation in the wing disc.

Triangles bridge the scales: Quantifying cellular contributions to tissue deformation

In this article, we propose a general framework to study the dynamics and topology of cellular networks that capture the geometry of cell packings in two-dimensional tissues. Such epithelia undergo large-scale deformation during morphogenesis of a multicellular organism. Large-scale deformations emerge from many individual cellular events such as cell shape changes, cell rearrangements, cell divisions, and cell extrusions. Using a triangle-based representation of cellular network geometry, we obtain an exact decomposition of large-scale material deformation. Interestingly, our approach reveals contributions of correlations between cellular rotations and elongation as well as cellular growth and elongation to tissue deformation. Using this triangle method, we discuss tissue remodeling in the developing pupal wing of the fly Drosophila melanogaster.

TissueMiner: A multiscale analysis toolkit to quantify how cellular processes create tissue dynamics

Segmentation and tracking of cells in long-term time-lapse experiments has emerged as a powerful method to understand how tissue shape changes emerge from the complex choreography of constituent cells. However, methods to store and interrogate the large datasets produced by these experiments are not widely available. Furthermore, recently developed methods for relating tissue shape changes to cell dynamics have not yet been widely applied by biologists because of their technical complexity. We therefore developed a database format that stores cellular connectivity and geometry information of deforming epithelial tissues, and computational tools to interrogate it and perform multi-scale analysis of morphogenesis. We provide tutorials for this computational framework, called TissueMiner, and demonstrate its capabilities by comparing cell and tissue dynamics in vein and inter-vein subregions of the Drosophila pupal wing. These analyses reveal an unexpected role for convergent extension in shaping wing veins.

DOI:http://dx.doi.org/10.7554/eLife.14334.001

Cells interact, divide, rearrange and change shape to build an organ during development. Modern microscopy and computer technology can follow each individual cell of an entire organ in a living organism. However, to understand how the complex choreography of cells leads to well-shaped organs, researchers need tools to help the store and analyze the large amounts of data generated. Tools are also needed to visualize and quantify the complex cell behaviors in an easy and flexible manner.

During its development, a fruit fly’s wings become divided into distinct regions separated by tubular supports called veins. Early on in development, the vein cells are indistinguishable from their neighbors, but at late stages, vein cells become a different shape. Veins also become narrower, which is assumed to be due to the number of vein cells falling. However, the way in which cells behave to bring about these changes has not been studied in detail.

Etournay, Merkel, Popović, Brandl et al. have now developed a toolkit called TissueMiner that enables users to store large amounts of data about cells and analyze how cells collectively shape an organ. TissueMiner was then used to identify vein cells at late stages of wing development and follow them backward in time to reveal their position at early stages. This showed that veins become narrower and more elongated because the cells that make up the veins shrink more than cells in other regions.

TissueMiner was then used to show that vein cells specifically rearrange and elongate to produce thinner regions, while the number of cells increases slightly because the cells divide. These results suggest that the cell behaviors responsible for making veins elongate and narrow are likely to be different from what had previously been assumed.

TissueMiner can be used in future studies to help understand the molecule signals that influence how cells behave in veins during wing development. The toolkit could also now be used to explore the changes involved in the development of other organs in other organisms.

DOI:http://dx.doi.org/10.7554/eLife.14334.002

Interplay of cell dynamics and epithelial tension during morphogenesis of the Drosophila pupal wing

How tissue shape emerges from the collective mechanical properties and behavior of individual cells is not understood. We combine experiment and theory to study this problem in the developing wing epithelium of Drosophila. At pupal stages, the wing-hinge contraction contributes to anisotropic tissue flows that reshape the wing blade. Here, we quantitatively account for this wing-blade shape change on the basis of cell divisions, cell rearrangements and cell shape changes. We show that cells both generate and respond to epithelial stresses during this process, and that the nature of this interplay specifies the pattern of junctional network remodeling that changes wing shape. We show that patterned constraints exerted on the tissue by the extracellular matrix are key to force the tissue into the right shape. We present a continuum mechanical model that quantitatively describes the relationship between epithelial stresses and cell dynamics, and how their interplay reshapes the wing.

Cochlear outer hair cells undergo an apical circumference remodeling constrained by the hair bundle shape

Epithelial cells acquire diverse shapes relating to their different functions. This is particularly relevant for the cochlear outer hair cells (OHCs), whose apical and basolateral shapes accommodate the functioning of these cells as mechano-electrical and electromechanical transducers, respectively. We uncovered a circumferential shape transition of the apical junctional complex (AJC) of OHCs, which occurs during the early postnatal period in the mouse, prior to hearing onset. Geometric analysis of the OHC apical circumference using immunostaining of the AJC protein ZO1 and Fourier-interpolated contour detection characterizes this transition as a switch from a rounded-hexagon to a non-convex circumference delineating two lateral lobes at the neural side of the cell, with a negative curvature in between. This shape tightly correlates with the 'V'-configuration of the OHC hair bundle, the apical mechanosensitive organelle that converts sound-evoked vibrations into variations in cell membrane potential. The OHC apical circumference remodeling failed or was incomplete in all the mouse mutants affected in hair bundle morphogenesis that we tested. During the normal shape transition, myosin VIIa and myosin II (A and B isoforms) displayed polarized redistributions into and out of the developing lobes, respectively, while Shroom2 and F-actin transiently accumulated in the lobes. Defects in these redistributions were observed in the mutants, paralleling their apical circumference abnormalities. Our results point to a pivotal role for actomyosin cytoskeleton tensions in the reshaping of the OHC apical circumference. We propose that this remodeling contributes to optimize the mechanical coupling between the basal and apical poles of mature OHCs.

Shroom2, a myosin-VIIa- and actin-binding protein, directly interacts with ZO-1 at tight junctions

Defects in myosin VIIa lead to developmental anomalies of the auditory and visual sensory cells. We sought proteins interacting with the myosin VIIa tail by using the yeast two-hybrid system. Here, we report on shroom2, a submembranous PDZ domain-containing protein that is associated with the tight junctions in multiple embryonic and adult epithelia. Shroom2 directly interacts with the C-terminal MyTH4-FERM domain of myosin VIIa and with F-actin. In addition, a shroom2 fragment containing the region of interaction with F-actin was able to protect actin filaments from cytochalasin-D-induced disruption in MDCK cells. Transfection experiments in MDCK and LE (L fibroblasts that express E-cadherin) cells led us to conclude that shroom2 is targeted to the cell-cell junctions in the presence of tight junctions only. In Ca(2+)-switch experiments on MDCK cells, ZO-1 (also known as TJP1) preceded GFP-tagged shroom2 at the differentiating tight junctions. ZO-1 directly interacts with the serine- and proline-rich region of shroom2 in vitro. Moreover, the two proteins colocalize in vivo at mature tight junctions, and could be coimmunoprecipitated from brain and cochlear extracts. We suggest that shroom2 and ZO-1 form a tight-junction-associated scaffolding complex, possibly linked to myosin VIIa, that bridges the junctional membrane to the underlying cytoskeleton, thereby contributing to the stabilization of these junctions.

PHR1, an integral membrane protein of the inner ear sensory cells, directly interacts with myosin 1c and myosin VIIa

By using the yeast two-hybrid technique, we identified a candidate protein ligand of the myosin 1c tail, PHR1, and found that this protein can also bind to the myosin VIIa tail. PHR1 is an integral membrane protein that contains a pleckstrin homology (PH) domain. Myosin 1c and myosin VIIa are two unconventional myosins present in the inner ear sensory cells. We showed that PHR1 immunoprecipitates with either myosin tail by using protein extracts from cotransfected HEK293 cells. In vitro binding assays confirmed that PHR1 directly interacts with these two myosins. In both cases the binding involves the PH domain. In vitro interactions between PHR1 and the myosin tails were not affected by the presence or absence of Ca2+ and calmodulin. Finally, we found that PHR1 is able to dimerise. As PHR1 is expressed in the vestibular and cochlear sensory cells, its direct interactions with the myosin 1c and VIIa tails are likely to play a role in anchoring the actin cytoskeleton to the plasma membrane of these cells. Moreover, as both myosins have been implicated in the mechanotransduction slow adaptation process that takes place in the hair bundles, we propose that PHR1 is also involved in this process.