Meeting report – Imaging the Cell

Before the start of the scientific part of the meeting, the conference opened on Wednesday morning with a brief presentation of the French Society for Cell Biology (www.sbcf.fr) by Jacky Goetz (INSERM U1109, Strasbourg, France), who has recently been elected to the SBCF council. It was followed by an introduction to the France-BioImaging infrastructure provided by Jean Salamero (Institut Curie, Paris, France). France-BioImaging is a French consortium to gather resources in cell and tissue imaging as part of the Euro-BioImaging initiative (http://www.eurobioimaging.eu/). After these introductions, optogenetics was the focus of the remainder of the session. Optogenetics uses light to control genetically modified cells by means of expressing light-sensitive proteins. This innovative technique was first developed to trigger and monitor the activities of individual neurons. Improvement of the spatio-temporal control of light stimulation allows to photoactivate neural circuits in patterns that mimic physiological processes. Claire Wyart (Neuroscience School of Paris, ENP, Paris, France) explained how she makes use of this sophisticated technique to reconstruct the connectivity map of sensory neurons that are involved in locomotion in the intact zebrafish (Fig. 1A). By using this approach, she was able to describe the sensorimotor integration in the vertebrate spinal cord, a work that opened new avenues for innovative neuromodulation strategies, in particular after spinal cord injury (Knafo and Wyart, 2015). Recently, optogenetics has been extended to the subcellular control of biological functions and, as such, has become a pivotal tool in cell biology. In this particular approach, a light signal that is directed to a selected region of a cell is used to control the activity of a cellular signalling protein. Mathieu Coppey (Institut Curie, Paris, France) presented the use of optogenetic activation for the spatiotemporal manipulation of Rho GTPases in the context of cell polarity establishment in migrating cells. By inducing specific RhoA perturbations with optogenetics and quantifying actomyosin dynamics in space and time, he suggested a mechanism of cell polarization that is based on a spatially coupled biochemical balance of substrates. In this model, the myosin turnover determines the dynamics of cell polarity when migration is driven at the rear.

The ability to automatically image samples and extract extensive sets of parameters from the data has permitted the development of high-content screening (HCS). This technology is based on the measurement of biological responses, such as cell phenotype, to conditions or molecules, such as drugs, peptides or small interfering (si)RNA. HCS not only requires complex instrumentation and sample preparation but also appropriate software in order to rapidly acquire and analyse the huge amount of produced data. Two speakers described how they used HCS to answer their specific biological question. Chris Bakal (The Institute of Cancer Research, London, UK) is interested in how signalling regulates cell shape and, reciprocally, how cell shape regulates signalling. By using HCS, he was able to delineate cellular processes that couple cell shape and the cellular microenvironment with the transcriptional activity of NF-κB. These new findings led to a model of how NF-κB is regulated in epithelial versus mesenchymal cancer cells (Sero et al., 2015). Frédéric Bard (Institute of Molecular and Cell Biology, Singapore) used HCS to perform an RNA interference (RNAi) screen to decipher which genes control the organization of the Golgi complex (Chia et al., 2012). His innovative approach of image analysis allows for an extensive description of Golgi complex phenotypes, dubbed the ‘deep phenotyping approach’. This method uses unsupervised clustering and extensive control of image quality, thereby yielding a genetic network that is based on phenotypic similarities. In the long term, the goal is to better understand how the Golgi complex is regulated and how the control of its dynamic organization can translate signals into a modulation of glycosylation and, ultimately, the display of specific glycans at the cell surface.

For a long time, most cell imaging was done in 2D. However, the past few years have seen the development of 3D and in vivo models to visualise cellular processes under relevant physiological conditions. Pierre Nassoy (Institut d'Optique d'Aquitaine, Bordeaux, France) developed a microfluidic method that is based on the encapsulation and the growth of cells inside permeable, elastic and hollow micro-spheres or micro-tubes (Alessandri et al., 2013). This elegant approach permits access to a variety of information about 3D tumour spheroids, such as the biomechanical regulation of adherent and non-adherent cells or of cell migration. As eluded to in the presentation, the potential biological applications of this encapsulation method are vast, such as production of in vitro epithelia, micro-tissues and organoids; in the future, these micro-tissues might be used for drug screening (Fig. 1B). Another approach to analyse cell behaviour in 3D is to directly visualise cells in vivo. This strategy is particularly important for studying cell invasion during the process of metastasis. David Sherwood (Duke University, Durham, NC) demonstrated that live-cell imaging of genetically modified Caenorhabditis elegans can provide unexpected and astonishing findings with regard to the cell migration process (Kelley et al., 2014; Matus et al., 2014), highlighting, for example, the existence of a new passive form of invasion (Fig. 1C).

In his talk in this session, Olivier Hamant (ENS Lyon, France) explored two central questions in biology: how does cell shape in plants control intracellular dynamics, and how do mechanical forces in tissues control cell positioning and differentiation? The relatively slow growth of plants makes them excellent models to address these questions. By using live-cell imaging combined with modelling of cell mechanics, Oliver derived a model by which mechanical forces control plant morphogenesis (Fig. 1D). Furthermore, this approach allowed him to identify an amplification mechanism that acts at the level of microtubules, through which mechanical stress promotes the response of the microtubule cytoskeleton by increasing a microtubule-severing activity (Hamant, 2013; Uyttewaal et al., 2012). Last but not least, Louis Hodgson (Albert Einstein College of Medicine of Yeshiva University, New York, USA) presented a wide range of biosensors his group had developed to specifically analyse the localization, activation and spatiotemporal dynamics of Rho family GTPases (Donnelly et al., 2015; Hanna et al., 2014). These new probes can be applied in Förster resonance energy transfer (FRET)-based assays and allow the analysis of how a very small fraction of RhoGTPases turn on or off at different locations and times in order to regulate cancer cell invasion (Fig. 1E). Isoform-specific probes are of particular interest because different RhoGTPases tightly coordinate their activities in response to environmental cues during cell migration.

Super-resolution imaging technologies have tremendously improved in the last decade, and these approaches now allow the cell biologist to visualise and follow cellular processes with unprecedented accuracy and detail. In this session, which featured three local but well-known ‘Bordelaise’ speakers, some of the currently most exciting developments in the field, as well as their applications, were highlighted. First, Grégory Giannone (Interdisciplinary Institute for NeuroScience, Bordeaux, France) presented the use of single-particle tracking (spt) combined with photoactivated localisation microscopy (PALM) – so called sptPALM – and super-resolution microscopy to study the displacement of integrins and their relative distribution between the outside and inside of mature focal adhesions (Rossier et al., 2012). He also applied the same approach to study the dynamic organization of F-actin regulators in neuronal dendritic spines (Chazeau et al., 2014). On the basis of his data, he suggested spatial segregation of specific protein interactions into distinct nanodomains to be a general mechanism in order to locally regulate protein activity within sub-cellular structures. Second, Valentin Nägerl (Interdisciplinary Institute for NeuroScience, Bordeaux, France) spoke about the development of stimulated emission depletion (STED) microscopy to uncover the nano-physiological rules that govern the plasticity of neurons and glial cells within brain slices. To that end, he combined STED with either fluorescence recovery after photobleaching (FRAP) or specific neurophysiological approaches – such as patch-clamp electrophysiology or two-photon glutamate uncaging (Chéreau et al., 2015; Tønnesen et al., 2014). Third, Daniel Choquet (Interdisciplinary Institute for NeuroScience, Bordeaux, France) demonstrated the use of different super-resolution imaging methods, including sptPALM, STED and universal point accumulation imaging in nanoscale topography (uPAINT), to analyse the localisation of ionotropic AMPA glutamate receptors (AMPAR) and their dynamic in synapses (Nair et al., 2013; Opazo et al., 2012). By combining these approaches with electrophysiological, genetic and pharmacological methods, he was able to investigate the nanoscale organization and function of synapses (Fig. 1F).

It was evident from the stimulating discussions in this session and throughout the meeting that techniques in super-resolution microscopy are likely to become an important, fundamental tool in cell biology in the near future – they have already been acknowledged by the Nobel Prize in Chemistry 2014, awarded to Eric Betzig, Stefan W. Hell and William E. Moerner.

Despite the development of super-resolution techniques, today, electron microscopy (EM) still provides the highest resolution detail of cellular ultrastructure. Thanks to a number of recent developments, EM is able to provide high-resolution 3D images of cellular structures, emphasising that EM remains an invaluable technique for cell biologists. Wolfgang Baumeister (Max Planck Institute of Biochemistry, Martinsried, Germany) has been using electron cryotomography (ECT) to study macromolecular and supramolecular structures in situ for a long time. He presented the use of a new type of phase plate, the Volta potential phase plate, which shows higher contrast and no fringing artefacts, and was used to analyse the proteasome 26S complexes in hippocampal neurons that had been grown on EM grids (Danev et al., 2014). Its application allowed the identification of different assembly states and conformations that reflect the activity status of individual 26S complexes (Asano et al., 2015). Christel Genoud (Friedrich Miescher Institute for Biomedical Research, Basel, Switzerland) talked about her latest improvements in scanning electron microscopy (SEM) in order to investigate the ultrastructure of entire tissues. To obtain a 3D volume at the ultrastructural level, she used serial block-face scanning electron microscopy (SBEM) – an approach that is increasing in popularity as it can be utilised in various disciplines of biology (Wanner et al., 2015). Using this technique, she obtained an image of the entire zebrafish larvae olfactory bulb at ultrastructure level with the unprecedented resolution of 12 nm per pixel. The image quality is such that it allows the reconstruction of the morphology of individual neurons. This type of EM approach allows to investigate samples of widely different dimension – from the live organism to the reconstruction of a neuronal circuit, as well as to link a structure to its function.

As one single imaging technique often cannot answer all the questions a cell biologist might have, combining different methods is a feasible strategy. Accordingly, Marko Kaksonen (European Molecular Biology Laboratory, Heidelberg, Germany) demonstrated how he can track average positions of endocytic proteins with <1 s accuracy and a spatial resolution of ∼10 nm, which allowed him to visualise the dynamic architecture of the yeast endocytic machinery (Picco et al., 2015). He then combined the tracking data with correlative light and electron microscopy (CLEM) to define the precise events during clathrin-mediated endocytosis in the budding yeast (Avinoam et al., 2015). Jacky Goetz (INSERM U1109, Strasbourg, France) also applied CLEM to image live zebrafish embryos. By using a correlative combination of live-cell imaging at high temporal resolution as well as EM coupled to electron tomography, he dissected the mechanosensing mechanism that allows the developing endothelium to sense flow-mediated forces. This approach also enabled the characterisation of the architecture and the behaviour of endothelial primary cilia in intact zebrafish with unprecedented precision (Goetz et al., 2014; Karreman et al., 2014). These results demonstrate that the onset of blood flow is mechanically sensed by cilia, whose deflection fine-tunes the resulting Ca²⁺ influx and regulates endothelial homeostasis. More recently, Jacky applied correlated intravital fluorescence microscopy together with electron microscopy to investigate with ultrastructural accuracy the behaviour of tumour cell in vivo (Karreman et al., 2014). By using artificial and endogenous structural features of the sample as reference points, he was able to retrieve single tumour cells that had been imaged deep within the tissue, allowing to accurately link the metastatic potential of a tumour to its structure (Fig. 1G). Finally, Jim Swoger (Centre for Genomic Regulation, Barcelona, Spain) talked about his recently developed hybrid platform, the so-called OPTiSPIM, that combines optical projection tomography (OPT) and selective plane illumination microscopy (SPIM) (Mayer et al., 2014). These two techniques are complementary in that SPIM can generate high-resolution data sets in large samples, whereas OPT enables the capture of fluorescence and absorption contrast. Jim presented how the combination of SPIM and spinning disc confocal microscopy allowed to quantify the 4D geometry of the amnioserosa during dorsal closure in Drosophila. Analysis of the data sets, together with mathematical modelling, showed that a decrease in the volume of the amnioserosa cells generates contractile forces that help to drive closure (Saias et al., 2015) (Fig. 1H).

Following the highly stimulating morning talks, ‘Imaging the Cell’ afternoons were devoted to practical sessions. At three different locations – Bordeaux University, Institute for Optics and the INRA – almost 40 different workshops were held. During the meeting, each participant had the opportunity to take part in up to five workshops, either in a classroom-type setting for image processing and analysis workshops, or within a small group next to a microscope for practical sessions. All topics of the morning sessions were covered by using hands-on approaches, and discussions were aimed at exchanging tricks and tips to ensure that these sophisticated technologies become more accessible to the wider cell biology community.

This meeting presented the most recent advances in imaging techniques available to cell biologists, both from a theoretical and practical point of view. Cellular imaging is beginning to break through optics-imposed limits; it now goes further and deeper, making it possible to visualise entire organs with an increasingly high level of resolution. The combination of several imaging techniques is currently the most feasible approach to obtain convincing and relevant images to answer our most pertinent questions regarding cell behaviour.

We are grateful to the Bordeaux Imaging Center (www.bic.u-bordeaux.fr) for its contribution to the organization of the meeting and to all institutes that hosted the practical workshops within their laboratories. The meeting was generously sponsored by academic institutions, such as Bordeaux University, INSERM, CNRS, Région Aquitaine, Aviesan-ITMO BCDE, Labex Brain, France BioImaging, IdEx Bordeaux, SFR Bordeaux Neurocampus, as well as by a number of companies and other organisations, including Journal of Cell Science and The Company of Biologists. We thank all the people that were involved in the organisation of the workshops and the industrial sponsors for their participation.