Live imaging in Drosophila: The optical and genetic toolkits

Live imaging in Drosophila: The optical and genetic toolkits Elena Rebollo, Katerina Karkali, Federica Mangione, Enrique Martín-Blanco Instituto de Biología Molecular de Barcelona, Consejo Superior de Investigaciones Científicas, Parc Científic de Barcelona, Baldiri Reixac 10, 08028 Barcelona, Spain Abstract Biological imaging based on light microscopy comes at the core of the methods that let us understanding morphology and its dynamics in synergy to the spatiotemporal distribution of cellular and molecular activities as the organism develops and becomes functional. Non-linear optical tools and superesolution methodologies are under constant development and their applications to live imaging of whole organisms keep improving as we speak. Genetically coded biosensors, multicolor clonal methods and optogenetics in different organisms and, in particular, in Drosophila follow equivalent paths. We anticipate a brilliant future for live imaging providing the roots for the holistic understanding, rather than for individual parts, of development and function at the whole-organism level. Introduction Organisms are made of cells and cells undergo dramatic dynamic rearrangements and changes in shape and motility during our lifetimes. Development is the result of the action of functionally interlaced hierarchical levels, genes, proteins, cells, tissues and organs. Each of them are just elements of a program used to build organisms. Knowledge on genes’ expression and function without understanding cellular behaviors development, morphogenesis or physiological functionality offers no convincing concepts or methods to grasp how system properties are built. Imaging thus arises as an essential tool to fully appreciate the cellular architecture and the dynamic functions and cooperative behavior of cells in this integrative context. Approaching a developing organism by imaging constitutes a great challenge. During development, proteins become expressed, traffic, change their subcellular localization or disappear, cells divide, specialize, differentiate and move or die, tissues aggregate, expand or fold and organs grow, shape and become physiologically active. Further, developing organisms display unique inhomogeneous optic properties. They can be opaque, refringent or autofluorescent, with these properties changing with time or locally. Thus, for each specific application the imaging protocols must be appropriately tailored. In synthesis, the correct optical tools with precise spatial and temporal resolution capabilities and the suitable structural or functional probes to employ must be carefully chosen and tuned. In this review, we will not comment on basic approaches or methodologies aimed to extract imaging information on fixed tissues or cultured cells, which are well covered throughout the literature. Our aim has been to present the latest methodological advances been employed in understanding Drosophila development. Our organism of choice, Drosophila, has constituted one of the major objects for basic research over the last 100 years. Many, if not most, of the major qualitative breakthroughs in the advances of biological sciences have taken place by using Drosophila as a model system. Its genetics and accessibility for imaging and physiological analyses make it ideal for a systems integrative approach. First, we present a thorough survey on the major optical techniques at place to undergo live imaging acquisition. Secondly, we review the major advances from the genetic point of view that have provided us with probes and methods to interrogate the organism by imaging means. The optical toolkit While bright field linear microscopy has brought a wealth of information for the understanding of morphology, cell and tissue behavior, it falls short accomplishing detailed information at the molecular level. In particular it is not suitable to detect those fast molecular events occurring in developmental or physiological processes. Fluorescence microscopy has become the method of choice to trace the molecular actions at the core of biological outcomes. Small intrinsically-fluorescent molecules, compounds derivatized with fluorescent reporters and modified genetically engineered protein chimeras all can be directly tracked in living cells and serve as sensitive sensors for expression, traffic or function of specific cellular components in a wide range of biological applications. A key advantage of using fluorescence in biological samples is that multiple protein species can be labeled and monitored concurrently in live specimens. The classic approach Fluorescence microscopy requires intense, near-monochromat9ic, illumination that in epifluorescence microscopes is focused through an objective lens that collects back the emission of the fluorophores prese10nt in the specimen. Although fluorescence microscopy can be extremely sensitive, the intrinsic design of classic fluorescence micro11scopes with full sample illumination leads to the capture of out of focus light from points outside the focal plane, w12hich reduces image clarity. In laser scanning confocal microscopy (LSCM) the objective focused laser spot is scanned across the sample point by point, to build up a pixel based image. Optical sectioning is achieved by introducing an aperture – pinhole – in the detection path. The pinhole discriminates out-of-focus light in an adjustable manner, delivering thin optical sections that build up an impressive three-dimensional view at an improved resolution compared to wide-field microscopes. However, this set up presents some inherent limitations that reduce its performance in live applications, where signal intensity is low by definition and temporal resolution in an important need. Traditionally, video-recordings have been performed using LSCM microscopes at the cost of reduced spatial resolution and optical sectioning. The progressive evolution of the LSCM though faster scanners and more sensitive detectors have helped improve the ratio between spatial and temporal resolution, to the point that this technology is one of the most widely used in standard live imaging applications in Drosophila (Fig. 1). A PubMed survey renders more than 1.000 references when combining the keywords ‘‘Drosophila’’ and ‘‘confocal’’ and is well beyond this review to comment the different applications in detail. Despite this overwhelming success, the reaction of the specimen to a continuous flux of intense illumination in LSCM imposes several limits that mainly show up in very demanding applications: (1) a significant fraction of the fluorescent molecules are already in the excited state (saturation), (2) a second photon may be absorbed leading to bleaching and (3) the specimen’s long term viability is compromised due to phototoxicity. An alternative to LSCM is the Spinning Disk Laser Confocal Microscope (SDLCM), which uses a multipoint scanning strategy based on a rotatory disk of pinholes, arranged in spirals arrays according to the original design by Paul Nipkow. The excitation laser source, rather than being concentrated in a single illumination spot, is shed through a portion of the rotatory disk and therefore divided in around 1000 illumination rays that get focused onto the sample by a high numerical aperture objective. Upon rotation of the disk at 5000–10,000 fpm, the focused laser spots scan the sample to generate an image acquired by parallel detection with a CCD camera. In this way, the system has the potential to collect images at an extremely high rate, limited only by the pixel clock rate of the CCD camera and the amount of available signal. This advantage has led to countless applications to the study of Drosophila development and physiology that we cannot analyze in this review. The most advanced confocal scanner unit (CSU), developed by Yokogawa Electric Corporation, incorporates a second disk, formed by microlenses arranged with identical pitch. This second disk compensates for the loss in illumination collection introduced by the interpinhole space and improves optical efficiency more than ten times compared to that of the conventional confocal Nipkow microscopes. The great achievement of the SDLCM for live imaging is that, by exposing the fluorophores to longer illumination times at lower intensity, fluorescence saturation is minimized without compromising the overall frame rate due to the multipoint design. Moreover, photobleaching is reduced and viability is improved during long recording sessions. However, in contrast to the highly spectral detection systems of most LSCM, the detection optical path of SDLCM is based on classical filter separation, which depending on the specimen and the fluorophores of choice may result in the need of longer exposure times to yield images of acceptable signal to noise ratio quality. This, together with the inherently low-light nature of confocal microscopes, limits the originally thought revolutionary speed of multichannel acquisitions on spinning disk microscopes. In conclusion, the advantages of using a LSCM or a SDLCM for live imaging applications will greatly depend on the biological question and the specimen characteristics. Non-linear optics The development of solid-state ultrafast lasers twenty-five years ago has lead to dramatic improvements of non-linear optics and the implementation of Multiphoton excited fluorescence microscopy (MPM) as the most effective tool for deep tissue analysis. Multiphoton molecular excitation consists of the simultaneous absorption of multiple photons that combine their energies to drive the chromophore to the excited state. The most widely used is the two-photon molecular excitation, whereby the simultaneous absorption of two photons of near infrared light results in the excitation of a fluorophore that normally absorbs UV or blue–green light. Such process will therefore drive fluorophore emission at a wavelength shorter than that of the excitation infrared source. Because at least two photons, not necessarily of identical wavelength, are required for each excitation, two-photon absorption is a rear non-linear event, restricted to those illumination points where photon concentration is high enough. Pulsed lasers provide optimal two-photon excitation by producing periodic sequences of extremely high photon density at the focal volume. Outside the latter, photon density is not high enough to generate an excitation event, which provides inherent three-dimensional resolution without the need of pinholes. Additionally, since all collected photons are ‘‘in focus’’, signal to noise ratio is dramatically improved, giving this set up extraordinary advantages when imaging thick, highly scattering tissues. Last but not least, the longer wavelengths used for excitation dramatically increase the penetration depth and ameliorate the overall specimen damage during long-term imaging experiments. Multiphoton excitation principles and capabilities are reviewed in many outstanding articles all of which cannot be included here, as a few examples see [11–14]. First generation multiphoton microscopes have enabled important advances in the understanding of relevant biological processes and have been applied to an increasing number of developmental models, in particular to Drosophila. The simple fact of being able to observed cell behaviors that remained unreachable by the one-photon approach, has made possible to place many published data within a new context. For instance, the widely accepted notion that Drosophila embryonic neural stem cells (neuroblasts), use spindle rotation to achieve asymmetric division in the embryo, turned out to be a unique characteristic of the very cell cycle at which the neuroblast detaches from the epithelium. Long term timelapse MPM of flies expressing fluorescent reporters has shown how a preset mode of spindle orientation, based on asymmetric centrosome behavior, is set right after this very first neuroblast division, thus creating a new scenario for the interpretation of relevant polarity events that occur during embryogenesis [15]. More sophisticated MPM approaches have led to: establish the correlation of cell movements with expression patterns, a key to understand the modulation of morphogenetic movements [16]; provide detailed insights into collective cell migration putting forward the trajectory of hundreds of cells during embryogenesis [17]; analyze the remodeling of the dendritic arborization of sensory neurons during metamorphosis and the role of juvenile hormone (JH) influencing the shape and complexity of the adult dendritic tree in a time-dependent manner [18] and resolve the forces that drive movements during ventral furrow formation and wound healing [16]. When combined with genetic and pharmacological manipulations, MPM has assisted to uncover the mechanisms involved in primordial germ cells (PGC) formation requiring an anaphase spindle-independent cleavage pathway [19]. MPM has also allowed other applications at the molecular level such as studies onto the spatiotemporal dynamics of DNA repair, RNA transcription and RNA transport. Multiphoton excitation with visible femtosecond pulses has been employed to produce thymine cyclopyrimidine dimers (CPDs) in polytene chromosomes from salivary glands. Live imaging then has let to observe the spatiotemporal recruitment of GFP-tagged topoisomerase I (TopI) to sites of localized DNA damage [20]. MPM has also been applied to visualize the assembly and dynamics of individual transcription factors and regulators and to dissect their functions at their endogenous gene targets in Drosophila polytene chromosomes in living salivary glands [21,22]. Last, imaging the dynamics of fluorescently labeled nos mRNA and Vas protein has been achieved at high resolution during early embryogenesis to image depths in excess of 80 µm. Visualization of detached germ plasm particles as they undergo rapid microtubule and dynein-dependent active transport toward proximal nuclei has helped elucidate the mechanism by which germ plasm components are sequestered away from the soma and into germ cells [23]. To an end, MPM has been employed to record neural activity. Using the highly sensitive calcium indicator GCaMP3 it has been possible to simultaneously monitor the activity of >100 MB neurons in vivo in mushroom bodies, an area involved in olfactory learning and memory [24]. It has also allowed for tracking the dynamics of PKA activity using a PKA FRET probe in the same structure [25]. The use of different calcium biosensors [Troponin C chimeras (TN-XXL) [26] and synthetic Ca2+ dyes [27]] has also been used to monitor fluorescence changes in presynaptic motoneuron boutons of transgenic Drosophila larvae with single-cell resolution. Besides all benefits of multiphoton microscopes, some limitations still hamper the use of twophoton scanning microscopy as the main tool for live imaging of thick specimens and the analysis of fast cellular and morphogenetic processes in vivo: (1) significant photodamage and rapid photobleaching at the focal plane, (2) low acquisition rates due to the point scanning system, (3) limited depth resolution and (4) difficulties to simultaneously acquire several discernible fluorophores. Different approaches have aimed to overpass these limitations. Regarding depth, three-photon excitation with an illumination wavelength of 1.7 µm with reduced light scattering has been demonstrated in proof-of-concept studies to reach into even the hippocampus in a living mouse [28]. On the other hand, adaptive optics (AO) have also shown utility for improving the resolution of 2PMs imaging deep in tissues [29]. AO measure optical aberrations and iteratively change the shape, phase, or intensity of the excitation sources. This leads to improve imaging at depth [30], providing near diffractionlimited imaging up to 450 µm inside tissues and enabling dramatic signal enhancement for small structures. When combined with long-wavelength laser illumination, such as for threephoton excitation, AO may become even more important. However, much work remains to make this a practical technique for day-to-day experimental studies. On the subject of time, a high pixel rate is necessary to capture fast processes such as fluid flows or Ca2+ spikes and even slow large-scale processes such as collective migration or cell division patterns if they involve a large number of pixels per image. In this context, pointscanning confocal or multiphoton approaches are too slow, as the image is recorded one pixel at a time. To solve this issue, ‘random-access’ two-photon microscopy rapidly imaging a series of discrete points can be performed using specialized acousto-optic deflectors, which enable positioning of a laser’s focus within a large volume essentially instantaneously. Alternatively, multiple excitation beams can be used in parallel to increase scanning speed but this occurs at the expense of limiting the penetration depth due to increased sensitivity to scattering [31]. Last, light-sheet fluorescence microscopy (LSFM) rises as the most promising way to improve the imaging speed of 2PMs (see below) and, indeed, has been successfully employed to track individual cell behaviors, mitotic lineages and cell movements in Drosophila embryos. Non-fluorescent non-linear contrasts (Second and Third Harmonic Generation) Besides fluorescence, nonlinear microscopy allows to employ alternative contrasts. Harmonic-generation contrast usually provides additional structural information on molecular or supramolecular order that would not be easily detectable with fluorescence strategies. In Second Harmonic Generation (SHG) two simultaneously incoming photons are transformed into one emitted photon with exactly half the wavelength. SHG is exclusively observed from dense organized non-centrosymmetric structures including fibrillar collagen, myofilaments, astroglial fibers, and polarized tubulin assemblies such as mitotic spindles. In Third Harmonic Generation (THG), the energy of three simultaneously incoming photons is combined to emit one photon with one-third the wavelength. THG does not require molecular asymmetry but is observed only near optical heterogeneities such as refraction index changes as occur, for example, between cell nuclei and cytoplasm or cytoplasm and interstitial fluid [32]. SHG and THG rely on coherent optical processes, and therefore have more complex contrast mechanisms than fluorescence. However, they are efficiently produced using femtosecond excitation pulses requiring a single laser, using standard two-photon microscopes (Fig. 2). Thus, SHG and THG offer optical sectioning capability, with exclusion of out-of-focus signal and 3D visualization for unlabeled preparations. Further, their combination with fluorescence microscopy is straightforward and, indeed, several studies have reported harmonic and multimodal harmonic/fluorescence imaging of embryos in various models, including Drosophila. SHG has been employed to investigate the differences between Drosophila larval and adult muscles. Indeed, a substantial decrease in SHG correlates with the reduction of the semi crystalline order of the anisotropic bands of the sarcomeres reached in mutant conditions by a reduction of myosin load [33]. SHG signal also reveals differences upon sarcomeres relaxation and contraction. Double peaked SHG profiles at rest, evolve into single peaks with a higher SHG intensity upon stretching [34]. Last, in combination with two-photon excited fluorescence, SHG has been employed in the analysis of the muscular architecture of living stage 2 larvae [35]. THG imaging has been used in combination with particle image velocimetry (PIV) analysis (velocimetric THG) to reach a detailed characterization at micrometer-scale of the morphogenetic movements taking place as the embryo develops up to dorsal closure stage in unstained wild-type and mutant embryos [36]. Sustained THG does not perturb embryos developmental dynamics. Further, THG together with two-photon excited fluorescence has been employed during gastrulation to explore both, long-distance modulations of morphogenetic movements and cytoskeleton perturbations, in response to femtosecond pulse-induced controlled microdissections. Indeed, THG microscopy performs well deep within the embryo and provides simultaneous information about the dynamics of tissue and yolk, revealing their long-range mechanical coupling [16]. It is important to note that non-fluorescent nonlinear signals provide additional structural information fully complementary to that gathered from fluorescent techniques. Multicolor imaging by wavelength mixing Multicolor labeling is opening new avenues in developmental biology and physiology. However, multicolor imaging with 2PMs has remained challenging, preventing the use of these strategies in thick tissues or live samples. New applications as the Brainbow clonal system employed in vertebrates and Drosophila (see Section 3) can be of tremendous help for automated cell tracking, but they come at the cost of the requirement to record multiple channels, which may significantly impact temporal sampling. In this scenario, multicolor twophoton tissue imaging by wavelength mixing can be employed for optimal and simultaneous two-photon imaging of three chromophores with distinct absorption spectra [37]. Wavelength mixing is obtained by synchronizing the pulse trains of a femtosecond optical parametric oscillator (OPO), which results in nonlinear processes produced by the different optical sources and their combinations. The combination of outputs can then be employed as an additional ‘virtual’ excitation wavelength. This scheme allows simultaneous two-photon excitation of different fluorophores (blue, green–yellow and red) with automatic submicrometer channels overlay. These features can become critical for quantitative approaches such as multicolor fluorescence correlation spectroscopy, FRET and ratiometric imaging. Multicolor two-photon imaging by wavelength mixing have been successfully applied to simultaneously detect multiple fluorophores, as those generated by Brainbow stochastic recombination both in the mouse cortex and in the chicken spinal cord. In Drosophila, during gastrulation, it has been employed to obtain in simultaneous recordings THG signal in parallel to three-channel fluorescence 3D+T data sets [37]. Such a technique, in combination with LSFM, promises to be an elegant way to facilitate multicolor-aided tracking of many cells. Light-sheet fluorescence microscopy (LSFM) LSFM takes a different approach to classic confocal microscopy by selectively illuminating and capturing only the current focal plane. Furthermore, in contrast to point scanning, the entire plane is illuminated at once and many planes can be acquired extremely fast using sensitive high-speed CCD or SCMOS cameras. LSFM successful application to developmental biology stems from the pioneer work of Jan Huisken, Ernst Stelzer and collaborators on the long-term live imaging of medaka fish and Drosophila embryos [38]. The most common LSFM designs feature a stage that moves the sample orthogonally through a thin laser light sheet. This light sheet is arranged perpendicular to the detection objective and illuminates the entire focal plane. Since fluorescence is only produced within the illuminated volume, optical sections are then confined to single slices with a well-defined thickness. Importantly, as the sample is continuously displaced between consecutive images through the laser sheet, only a small part is exposed at any time resulting in very low phototoxicity and bleaching [39]. This is essential for the long-term recording of live specimens. By mounting the sample in a rigid medium, e.g. agarose, and hanging it into the sample chamber in front of the detection lens, it is possible to rotate the sample and collect 3D stacks from multiple angles (SPIM). SPIM allows imaging of large samples with high resolution over extended periods of time delivering high signal-to-noise images. Multiple 3D data sets of the same object are collected from different directions and combined in a post-processing step. OpenSPIM is an Open Access platform (http://openspim.org) developed to facilitate the access of SPIM technology to biologists. This platform developed by the Tomancak and Huisken labs in collaboration with the Laboratory for Optical and Computational Instrumentation (LOCI) in Madison documents in great detail how to assemble an affordable light sheet set-up. The OpenSPIM hardware employs an open source microscopy control plugin (microManager), which includes SPIM specific functionalities such as sample rotation. All software is incorporated into Fiji (http://fiji.sc), an open source platform for biological image analysis, which offers advanced algorithms for processing SPIM image data [40]. Despite all its advantages, LSFM also present some drawbacks. The optical penetration depth in light sheet microscopes is limited by light scattering, which precludes systems-level imaging in thick biological model organisms. Employing non-linear excitation in multiphoton configurations does not ameliorate this problem. Neither does sample rotation and sequential acquisition of image stacks from multiple view angles that it is inherently slow and fails to capture fast-occurring processes. Lately, two simultaneous multiview imaging set ups presenting minor technical differences at the imaging and post-processing levels have been developed to overcome these spatiotemporal artifacts. On one hand, the SiMView light sheet microscopy platform allows simultaneous acquisition of four complementary views of the specimen for optimal physical coverage (Fig. 3). SiMView uses an orthogonal arrangement of four independently operated optical arms. One pair is used for bidirectional light sheet two-photon illumination with two long working distance air objectives. The other pair, arranged at a right angle to the first, is used for bidirectional fluorescence detection with high numerical aperture water-dipping objectives and fast SCMOS cameras. It is designed for high-speed long-term imaging under physiological conditions and can sustained data acquisition at 350 megabytes (175 million voxels) per second for up to several days [41]. MuVi-SPIM, as SiMView, uses two illumination and two detection objective lenses that are focused onto a specimen from four different directions along two perpendicular axes. Each one-photon light sheet is generated by scanning a Gaussian beam at constant velocity. Three-dimensional images are accumulated by recording a series of sections, while the specimen is moved through the light sheet. To construct a single image, all four images are transformed to a common coordinate system and then combined [42]. The first demonstration of the capacity of LSFM (SPIM) to image internal structures of opaque embryos was performed in Drosophila. With no need of multiview reconstruction, structures inside the embryo were clearly identifiable and traceable over a period of 17 h, without refocusing or realignment [27]. LSFM rises as the method of choice for lineaging direct observation. In higher organisms, including Drosophila, lineages are regulative, and cell migration and mixing are very prevalent during development. Thus, establishment of a full mitotic lineage requires the simultaneous tracking of cell movements over time. Combining LSFM and automated image processing circumvents this caveat and has let to generate the first comprehensive data set of cell positions both for early zebrafish [43] and for Drosophila embryogenesis [33]. Further, whole organisms analyses are now readily accessible. Longterm and fast multiple-view imaging of early whole Drosophila embryos has let to reconstruct cell positions over time to create a fly digital embryo [44]. An independent equivalent analysis has been performed employing two-photon scanned LSFM to cover the entire embryo development with excellent penetration depth [45]. Indeed, both MuVi-SPIM with scanned single-photon [42] and SiMView with scanned two-photon LSFM [33] have been applied to the long-term in toto imaging of the Drosophila embryo at 30 s time intervals covering essentially the entire embryonic development. Both studies additionally perform automated cell tracking during syncytial blastoderm stages, demonstrating the potential of this technique for the retrospective analysis of cell behavior and lineage in whole organisms. The feasibility of LSFM technology to address the analysis of more complex morphogenetic movements at different stages of Drosophila development has just been tested to monitor dendrite pruning during pupal stages [46]. The full potential of LSFM/SPIM itself or in combination with biosensors or clonal analysis still remains to be determined. Superesolution For centuries, advances in cell biology that rely on light microscopy techniques have been limited by the optical resolution of the conventional microscopes. This limit, imposed by the wavelength of light, is defined by the point-spread function (PSF), that is, the fixed minimum size at which a point light source appears to be convolved by the microscope optics. Being this size around 250 nm in the xy dimension and 450–700 nm in the z dimension, much of the macromolecular complexes behavior remains beyond the imaging reach, which hinders a molecular-level understanding of cell structure and dynamics. Total internal reflection fluorescence (TIRF) microscopy appeared in the early 1980’s as a revolutionary technology capable of fast imaging within a thin section – between 100 and 200 nm – of the specimen [47]. TIRF uses an angle of illumination such that all the incident light is internally reflected, generating an evanescence electromagnetic field at the glass–water interphase that decays exponentially, thus penetrating less than 200 nm into the sample. Due to its inherent capability for tracking vesicles, tiny organelles and even molecular events near the plasma membrane, this new achievement in z axial resolution allowed uncovering new questions related to membrane trafficking and cortical cytoskeleton dynamics. Nevertheless, resolution in the xy dimension still remained within the limits established by the optics. Over the past decade, several new technologies have been developed which overcome the optical resolution limit by temporally or spatially modulating the excitation light. These techniques, collectively called superesolution (SR) techniques have been able to surpass the diffraction limit at very different ranges between 100 and 10 nm in the xy dimension. In general, SR approaches can be classified into two broad categories: ensemblebased techniques and single molecule techniques. Ensemble-level methods increase resolution by shaping the PSF; among these methods, stimulated emission depletion (STED) microscopy reduces the effective diameter of the PSF by induced de-excitation of the offcentered fluorophores [48], and Structured illumination microscopy (SIM) uses an interference pattern that is rotated and translated over a series of images in order to gain access to higher successfully adapted to a variety of other systems and applications including Drosophila [96– 98]. dBrainbow constitutes an adapted version of the vertebrate Brainbow-1 system consisting of a reporter construct that can be restrictively expressed using the Gal4/UAS system and that can randomly generate different color patterns from three different cytoplasmic fluorescent proteins. It comprises four cassettes flanked by pairs of mutually exclusive lox sites. In the presence of Cre recombinase, recombination occurs and results in the irreversible selection of one of the fluorescent proteins. Fly stocks carrying two recombinant chromosomes allow the production of up to six color combinations. dBrainbow has been employed in morphological analysis of the adult nervous system to track motor neurons in the subesophageal ganglion to identify their specific proboscis muscle targets, and to study GABAergic lineages. In both cases, antibody labeling of epitopes rather than endogenous fluorescence was employed to trace fine processes over long distances [97,99]. dBrainbow has also been applied to analyze in detail the epithelial organization and to trace the surface and shape of individual tracheal cells in embryos [100]. Flybow is a Drosophila variant of the mouse Brainbow-2 system combining the Gal4-UAS and the inducible Flp-FRT systems to drive inversions and excisions of FRT cassettes. Sequences encoding different membrane-tethered fluorescent proteins [enhanced GFP (EGFP), mCitrine, mCherry and Cerulean] were arranged in pairs flanked by recombination sites. Flybow has been employed for cell tracing and identification of cell subtypes in the embryonic nervous system, the visual system (eye disc and optic lobe) and the third instar larval wing imaginal disc [96,101]. It has also been used to label and identify each individual embryonic peripheral glia (ePG) in the living embryo and across till third instar larva. This permitted to reach a description of the embryonic origin of all three glial layers and the tracing of larval differentiation for each ePG cell. Cell tracking via Flybow indicates that at least some ePG survive metamorphosis and persist until adulthood [102]. Flp in Flybow causes no apparent toxicity but recombines Brainbow cassettes at very low efficiencies. On the other hand, Cre in dBrainbow exhibits much higher per-cell recombination efficiency but fails to control when and where recombination takes place and can be toxic to the animals. To overcome these problems, photo-inducible split-Cre Drosophila transgenes (LOLLIbow) have been developed (Fig. 5). Two complementary fragments of Cre recombinase were each chimerized to two plant proteins capable of dimerization in a blue light-dependent manner. The N-terminal fragment of the recombinase was fused to nuclear tagged cryptochrome 2 (CRY2) and the Cterminal fragment to a nuclear tagged truncated version of CIB1 (CIBN). Both construct were placed under the control of UAS sequences and recombined into a single fly line. LOLLI-bow allows visualizing morphological details in intact animals with single cell resolution in vivo through recombination of cassettes that render each cell to inherit a permanent assignment among three membrane-targeted fluorescent proteins [98]. Optogenetics The ability to instruct cell activities non-intrusively, fast and in a precise, reproducible way with cellular resolution is the dream of any developmental biologists or physiologists. Optogenetics have developed through the combination of genetic and optical means to control and monitor the activities of individual cells and is primarily applied to neurons. The earliest genetically targeted method, which used light to control genetically sensitized neurons, employed Drosophila rhodopsin photoreceptors for controlling neural activity in cultured mammalian neurons [103]. This led subsequently to the development of single-component optogenetic systems using channelrhodopsin for the photostimulation of genetically circumscribed groups of neurons [104]. Optimized channelrhodopsins have been engineered to increased blue-light sensitivity and they have been used to increase neural activity in deep-brain genetically targeted cells [105]. These engineered tools are now been used in modulating complex behaviors in different model systems. In Drosophila, using spatially defined GAL4 drivers and UAS-channelrhodopsin-2 variants is possible to manipulate neuronal populations in response to illumination by blue light and to test whether the activation of defined neural circuits is sufficient to shape behaviors of interest. This approach has been used to activate larval nociceptors and nociception behaviors in the third-larval instar [106]. Technically, in Drosophila, optogenetics application has been limited by the inherent opacity of the adult fly cuticle to blue light. However, red activatable channelrhodopsin (ReaChR) variants have been recently developed allowing behavioral control in freely moving adult flies. This tool affords the opportunity to control neural activity over a broad dynamic range of stimulation intensities [107]. Optogenetic stimulation has been successfully applied amongst others to the large tangential horizontal system cells of the lobula plate to evaluate their behavioral role. This interference elicits robust yaw head movements and yaw turning responses in fixed and tethered flying flies, respectively [108]. Optogenetics in Drosophila is not just restricted to neurosciences Modulating activity with light has also been applied to the analysis of collective cell movements during oogenesis. The BCs of the Drosophila ovary undergo stereotypic movements from the posterior pole of the egg chamber to the oocyte posterior edge around stage 10 of oogenesis. This movement depends on active dynamical changes of these cells cytoskeleton. The small GTPase Rac induces actin polymerization, membrane ruffling and focal contact formation in cultured single cells and is a potential candidate to control collective epithelial cell movements. The group of Denise Montell took use of recently developed photoactivatable analogues of Rac (PA-Rac) allowing its rapid and reversible activation or inactivation using light to analyze its role in BCs movements. PA-Rac is a member of a series of probes encompassing different members of the Rho family of GTPases that have been grafted to the phototropin LOV (light, oxygen, or voltage) domain from the Arabidopsis NPH1 protein. Exposure to 400–500 nm light caused unwinding of the helix linking the LOV domain to the GTPase, relieving steric inhibition. The change is reversible and repeatable [77]. Using this tool it was found that that focal activation of Rac is sufficient to polarize the whole group of cells in vivo. Photoactivation, time-lapseimaging, and 3D morphological reconstruction were carried out using confocal microscopy. Moreover, activation or inactivation of Rac in one cell of the cluster caused an immediate response in its siblings, suggesting that cells responding to relative levels of Rac activity sense direction as a group [78]. These studies indicate that photoactivatable proteins are effective tools in vivo and can presumably be applied to other developmental processes and to other signaling molecules. Conclusions Imaging has become essential in addressing local and global, structural and functional events. New optical tools and genetic methods are under constant development and improvement. Computational approaches and physical models to transform all the information provided by these imaging tools into biological insights are still in their early stages, but we expect in the near future major qualitative and quantitative leaps in this direction. To understand the principles that govern the development of an organism, producing enough cells, making them different and organizing them in space and time, is all a problem of information processing. The original information behind this processing comes from the actions of cells and molecules in concert and from their synchronization. Live imaging helps us to acquire this information and the major technological revolution that has taken place in these last past years in the field is positioning us in a place where we should be able to undergo a qualitative leap in the way we grasp this information. Approaching cellular and tissue dynamics in entire vertebrate and higher invertebrate embryos, following cell movements, cell shape dynamics, subcellular protein localization, functional activities and changes in gene expression, all simultaneously, is now at hand. Acknowledgements We thank the Confocal Microscopy Unit, IBMB-PCB, the Advanced Digital Microscopy Unit, IRB Barcelona and members of the EMB laboratory for continuous support. We are grateful to Philipp Keller, Pavel Tomancak and Emmanuel Beaurepaire for encouragement and help with images. The Consolidated Groups Program of the Generalitat de Catalunya and DGI and Consolider Grants from the Ministry of Economy and Competitivity of Spain to E.M.B. and an Explora Grant to E.R. supported this work. References 1. K.D. Micheva, S.J. Smith, Neuron 55 (2007) 25–36. 2. P.J. Shaw, in: James B. Pawley (Ed.), Handbook of Biological Confocal Microscopy, Springer, 2006, pp. 1–19. 3. P.N. Dean, Curr. Cytometry 5 (2001) 2.8.1–2.8.12. 4. S.C. Tovey, P.J. Brighton, E.T. Bampton, Y. Huang, G.B. Willars, Calcium Signaling Protocols, Springer, 2013, pp. 51–93. 5. S. Inoué, in: James B. Pawley (Ed.), Handbook of Biological Confocal Microscopy, Springer, 2006, pp. 1–19. 6. James B. Pawley (Ed.), Handbook of Biological Confocal Microscopy, Springer, 2006, pp. 20–42. 7. R. Gräf, J. Rietdorf, T. Zimmermann (Eds.), Microscopy Techniques, Springer, 2005, pp. 57–75. 8. D. Toomre, in: J.B. Pawley (Ed.), Handbook of Biological Confocal Microscopy, Springer, 2006, pp. 221–238. 9. T. Tanaami, S. Otsuki, N. Tomosada, Y. Kosugi, M. Shimizu, H. Ishida, Appl. Opt. 41 (2006) 4704–4708. 10. E. Gratton, M.J. vandeVen, in: James B. Pawley (Ed.), Handbook of Biological Confocal Microscopy, Springer, 2006, pp. 80–125. 11.A. Diaspro, G. Chirico, M. Collini, Q. Rev. Biophys. 38 (2005) 97–166. 12. K. Svoboda, R. Yasuda, Neuron 50 (2006) 823–839. 13. R. Yasuda, Curr. Opin. Neurobiol. 16 (2006) 551–561. 14. W. Denk, Cold Spring Harb Protoc (2007) pdb. Top23. 15. E. Rebollo, M. Roldan, C. Gonzalez, Development 136 (2009) 3393–3397. 16. G.W. Brodland, V. Conte, P.G. Cranston, J. Veldhuis, S. Narasimhan, M.S. Hutson, A. Jacinto, F. Ulrich, B. Baum, M. Miodownik, Proc. Natl. Acad. Sci. U.S.A. 107 (2010) 22111– 22116. 17. A. McMahon, W. Supatto, S.E. Fraser, A. Stathopoulos, Science 322 (2008) 1546– 1550. 18. D.W. Williams, J.W. Truman, J. Neurosci. 24 (2004) 1541–1550. 19. R.M. Cinalli, R. Lehmann, Nat. Cell Biol. 15 (2013) 839–845. 20. M.K. Daddysman, C.J. Fecko, Biophys. J. 101 (2011) 2294–2303. 21. J. Yao, K.M. Munson, W.W. Webb, J.T. Lis, Nature 442 (2006) 1050–1053. 22. J. Yao, K.L. Zobeck, J.T. Lis, W.W. Webb, Methods 45 (2008) 233–241. 23. D.A. Lerit, E.R. Gavis, Curr. Biol. 21 (2011) 439–448. 24. Y. Wang, H.F. Guo, T.A. Pologruto, F. Hannan, I. Hakker, K. Svoboda, Y. Zhong, J. Neurosci. 24 (2004) 6507–6514. 25. N. Gervasi, P. Tchenio, T. Preat, Neuron 65 (2010) 516–529. 26. M. Mank, A.F. Santos, S. Direnberger, T.D. Mrsic-Flogel, S.B. Hofer, V. Stein, T. Hendel, D.F. Reiff, C. Levelt, A. Borst, T. Bonhoeffer, M. Hubener, O. Griesbeck, Nat. Methods 5 (2008) 805–811. 27. T. Hendel, M. Mank, B. Schnell, O. Griesbeck, A. Borst, D.F. Reiff, J. Neurosci. 28 (2008) 7399–7411. 28. N.G. Horton, K. Wang, D. Kobat, C.G. Clark, F.W. Wise, C.B. Schaffer, C. Xu, Nat. Photonics 7 (3) (2013) 205–209. 29. N. Ji, T.R. Sato, E. Betzig, Proc. Natl. Acad. Sci. U.S.A. 109 (2012) 22–27. 30. C. Wang, N. Ji, Opt. Lett. 37 (2012) 2001–2003. 31. V. Andresen, K. Pollok, J.L. Rinnenthal, L. Oehme, R. Gunther, H. Spiecker, H. Radbruch, J. Gerhard, A. Sporbert, Z. Cseresnyes, A.E. Hauser, R. Niesner, PLoS ONE 7 (2012) e50915. 32. W. Supatto, T.V. Truong, D. Debarre, E. Beaurepaire, Curr. Opin. Genet. Dev. 21 (2011) 538–548. 33. C. Greenhalgh, N. Prent, C. Green, R. Cisek, A. Major, B. Stewart, V. Barzda, Appl. Opt. 46 (2007) 1852–1859. 34. N. Prent, C. Green, C. Greenhalgh, R. Cisek, A. Major, B. Stewart, V. Barzda, J. Biomed. Opt. 13 (2008) 041318. 35. C.Y. Lin, V. Hovhannisyan, J.T. Wu, C.W. Lin, J.H. Chen, S.J. Lin, C.Y. Dong, J. Biomed. Opt. 13 (2008) 050502. 36. D. Debarre, W. Supatto, E. Farge, B. Moulia, M.C. Schanne-Klein, E. Beaurepaire, Opt. Lett. 29 (2004) 2881–2883. 37. P. Mahou, M. Zimmerley, K. Loulier, K.S. Matho, G. Labroille, X. Morin, W. Supatto, J. Livet, D. Debarre, E. Beaurepaire, Nat. Methods 9 (2012) 815–818. 38. J. Huisken, J. Swoger, F. Del Bene, J. Wittbrodt, E.H. Stelzer, Science 305 (2004) 1007–1009. 39. P.J. Keller, E.H. Stelzer, Curr. Opin. Neurobiol. 18 (2008) 624–632. 40. P.G. Pitrone, J. Schindelin, L. Stuyvenberg, S. Preibisch, M. Weber, K.W. Eliceiri, J. Huisken, P. Tomancak, Nat. Methods 10 (2013) 598–599. 41. R. Tomer, K. Khairy, F. Amat, P.J. Keller, Nat. Methods 9 (2012) 755–763. 42. U. Krzic, S. Gunther, T.E. Saunders, S.J. Streichan, L. Hufnagel, Nat. Methods 9 (2012) 730–733. 43. P.J. Keller, A.D. Schmidt, J. Wittbrodt, E.H. Stelzer, Science 322 (2008) 1065–1069. 44. P.J. Keller, A.D. Schmidt, A. Santella, K. Khairy, Z. Bao, J. Wittbrodt, E.H. Stelzer, Nat. Methods 7 (2010) 637–642. 45. T.V. Truong, W. Supatto, D.S. Koos, J.M. Choi, S.E. Fraser, Nat. Methods 8 (2011) 757–760. 46. J. Huisken, D.Y. Stainier, Development 136 (2009) 1963–1975. 47. A.L. Mattheyses, S.M. Simon, J.Z. Rappoport, J. Cell Sci. 123 (2010) 3621–3628. 48. S.W. Hell, J. Wichmann, Opt. Lett. 19 (1994) 780–782. 49. M.G. Gustafsson, J. Microsc. 198 (2000) 82–87. 50. E. Betzig, G.H. Patterson, R. Sougrat, O.W. Lindwasser, S. Olenych, J.S. Bonifacino, M.W. Davidson, J. Lippincott-Schwartz, H.F. Hess, Science 313 (2006) 1642–1645. 51. M.J. Rust, M. Bates, X. Zhuang, Nat. Methods 3 (2006) 793–795. 52. U. Agrawal, D.T. Reilly, C.M. Schroeder, Curr. Opin. Biotechnol. 24 (2013) 646–653. 53. S.A. Jones, S.H. Shim, J. He, X. Zhuang, Nat. Methods 8 (2011) 499–508. 54. R.L. Webb, O. Rozov, S.C. Watkins, B.M. McCartney, Dev. Dyn. 238 (2009) 2622– 2632. 55. R.D. Vale, J.A. Spudich, E.R. Griffis, J. Cell Biol. 186 (2009) 727–738. 56. P. Fabrowski, A.S. Necakov, S. Mumbauer, E. Loeser, A. Reversi, S. Streichan, J.A. Briggs, S. De Renzis, Nat. Commun. 4 (2013) 2244. 57. P. Kner, B.B. Chhun, E.R. Griffis, L. Winoto, M.G. Gustafsson, Nat. Methods 6 (2009) 339–342. 58. M. Sander, A.J. Squarr, B. Risse, X. Jiang, S. Bogdan, Eur. J. Cell Biol. 92 (2013) 349– 354. 59. T. Zobel, S. Bogdan, J. Microsc. 251 (2013) 224–231. 60. L. Gao, L. Shao, C.D. Higgins, J.S. Poulton, M. Peifer, M.W. Davidson, X. Wu, B. Goldstein, E. Betzig, Cell 151 (2012) 1370–1385. 61. K.J. Venken, H.J. Bellen, Nat. Rev. Genet. 6 (2005) 167–178. 62. A.H. Brand, N. Perrimon, Development 118 (1993) 401–415. 63. A. del Valle Rodríguez, D. Didiano, C. Desplan, Nat. Methods 9 (2012) 47–55. 64. V.K. Lund, Y. DeLotto, R. DeLotto, Proc. Natl. Acad. Sci. U.S.A. 107 (2010) 18028– 18033. 65. O. Cormier, N. Mohseni, I. Voytyuk, B.H. Reed, Autophagy 8 (2012) 252–264. 66. P.L. Bardet, G. Kolahgar, A. Mynett, I. Miguel-Aliaga, J. Briscoe, P. Meier, J.P. Vincent, Proc. Natl. Acad. Sci. U.S.A. 105 (2008) 13901–13905. 67. S.C. Albrecht, A.G. Barata, J. Grosshans, A.A. Teleman, T.P. Dick, Cell Metab. 14 (2011) 819–829. 68. L. Tian, S.A. Hires, T. Mao, D. Huber, M.E. Chiappe, S.H. Chalasani, L. Petreanu, J. Akerboom, S.A. McKinney, E.R. Schreiter, C.I. Bargmann, V. Jayaraman, K. Svoboda, L.L. Looger, Nat. Methods 6 (2009) 875–881. 69. J. Akerboom, N. Carreras Calderon, L. Tian, S. Wabnig, M. Prigge, J. Tolo, A. Gordus, M.B. Orger, K.E. Severi, J.J. Macklin, R. Patel, S.R. Pulver, T.J. Wardill, E. Fischer, C. Schuler, T.W. Chen, K.S. Sarkisyan, J.S. Marvin, C.I. Bargmann, D.S. Kim, S. Kugler, L. Lagnado, P. Hegemann, A. Gottschalk, E.R. Schreiter, L.L. Looger, Front. Mol. Neurosci. 6 (2013) 2, 1-29. 70. N. Mochizuki, S. Yamashita, K. Kurokawa, Y. Ohba, T. Nagai, A. Miyawaki, M. Matsuda, Nature 411 (2001) 1065–1068. 71. R.E. Itoh, K. Kurokawa, Y. Ohba, H. Yoshizaki, N. Mochizuki, M. Matsuda, Mol. Cell Biol. 22 (2002) 6582–6591. 72. D. Kamiyama, A. Chiba, Science 324 (2009) 1338–1340. 73. A. Leibfried, S. Muller, A. Ephrussi, Development 140 (2013) 362–371. 74. W. Lu, M.O. Casanueva, A.P. Mahowald, M. Kato, D. Lauterbach, E.L. Ferguson, PLoS Biol. 10 (2012) e1001357. 75. C.H. Fernandez-Espartero, D. Ramel, M. Farago, M. Malartre, C.M. Luque, S. Limanovich, S. Katzav, G. Emery, M.D. Martin-Bermudo, J. Cell Sci. 126 (2013) 2285–2293. 76. D. Ramel, X. Wang, C. Laflamme, D.J. Montell, G. Emery, Nat. Cell Biol. 15 (2013) 317–324. 77. Y.I. Wu, X. Wang, L. He, D. Montell, K.M. Hahn, Methods Enzymol. 497 (2011) 393– 407. 78. X. Wang, L. He, Y.I. Wu, K.M. Hahn, D.J. Montell, Nat. Cell Biol. 12 (2010) 591–597. 79. T. Tsuyama, J. Kishikawa, Y.W. Han, Y. Harada, A. Tsubouchi, H. Noji, A. Kakizuka, K. Yokoyama, T. Uemura, H. Imamura, Anal. Chem. 85 (2013) 7889–7896. 80. K. Takemoto, E. Kuranaga, A. Tonoki, T. Nagai, A. Miyawaki, M. Miura, Proc. Natl. Acad. Sci. U.S.A. 104 (2007) 13367–13372. 81. E. Kuranaga, T. Matsunuma, H. Kanuka, K. Takemoto, A. Koto, K. Kimura, M. Miura, Development 138 (2011) 1493–1499. 82. I. Almudi, I. Poernbacher, E. Hafen, H. Stocker, Cell Commun. Signal. 11 (2013) 26. 83. H.D. Vishwasrao, P. Trifilieff, E.R. Kandel, Biophys. J. 102 (2012) 1204–1214. 84. W. Yu, F. Kawasaki, R.W. Ordway, Mol. Cell Neurosci. 47 (2011) 19–27. 85. M. Mank, D.F. Reiff, N. Heim, M.W. Friedrich, A. Borst, O. Griesbeck, Biophys. J. 90 (2006) 1790–1796. 86. V. Lissandron, M.G. Rossetto, K. Erbguth, A. Fiala, A. Daga, M. Zaccolo, Cell Signal. 19 (2007) 2296–2303. 87. J. Zhang, C.J. Hupfeld, S.S. Taylor, J.M. Olefsky, R.Y. Tsien, Nature 437 (2005) 569– 573. 88. C.G. Vecsey, N. Pirez, L.C. Griffith, J. Neurophysiol. 111 (2014) 1033–1045. 89. N. Pirez, B.L. Christmann, L.C. Griffith, J. Neurophysiol. 110 (2013) 700–708. 90. O.T. Shafer, D.J. Kim, R. Dunbar-Yaffe, V.O. Nikolaev, M.J. Lohse, P.H. Taghert, Neuron 58 (2008) 223–237. 91. T. Lee, L. Luo, Neuron 22 (1999) 451–461. 92. R. Griffin, A. Sustar, M. Bonvin, R. Binari, A. del Valle Rodriguez, C. Bakal, A.M. Hohl, J.R. Bateman, C. Villalta, E. Heffern, Nat. Methods 6 (2009) 600–602. 93. C.J. Evans, J.M. Olson, K.T. Ngo, E. Kim, N.E. Lee, E. Kuoy, A.N. Patananan, D. Sitz, P. Tran, M.-T. Do, Nat. Methods 6 (2009) 603–605. 94. L. Centanin, B. Hoeckendorf, J. Wittbrodt, Cell Stem Cell 9 (2011) 553–562. 95. J. Livet, T.A. Weissman, H. Kang, R.W. Draft, J. Lu, R.A. Bennis, J.R. Sanes, J.W. Lichtman, Nature 450 (2007) 56–62. 96. D. Hadjieconomou, S. Rotkopf, C. Alexandre, D.M. Bell, B.J. Dickson, I. Salecker, Nat. Methods 8 (2011) 260–266. 97. S. Hampel, P. Chung, C.E. McKellar, D. Hall, L.L. Looger, J.H. Simpson, Nat. Methods 8 (2011) 253–259. 98. M. Boulina, H. Samarajeewa, J.D. Baker, M.D. Kim, A. Chiba, Development 140 (2013) 1605–1613. 99. C.L. Gatto, D. Pereira, K. Broadie, Neurobiol. Dis. 65 (2014) 142–159. 100. D. Forster, S. Luschnig, Nat. Cell Biol. 14 (2012) 526–534. 101. N. Shimosako, D. Hadjieconomou, I. Salecker, Methods Mol. Biol. 1082 (2014) 57–69. 102. C.M. von Hilchen, A.E. Bustos, A. Giangrande, G.M. Technau, B. Altenhein, Development 140 (2013) 3657–3668. 103. B.V. Zemelman, G.A. Lee, M. Ng, G. Miesenbock, Neuron 33 (2002) 15–22. 104. E.S. Boyden, F. Zhang, E. Bamberg, G. Nagel, K. Deisseroth, Nat. Neurosci. 8 (2005) 1263–1268. 105. K. Deisseroth, M.J. Schnitzer, Neuron 80 (2013) 568–577. 106. K. Honjo, R.Y. Hwang, W.D. Tracey Jr., Nat. Protoc. 7 (2012) 1470–1478. 107. H.K. Inagaki, Y. Jung, E.D. Hoopfer, A.M. Wong, N. Mishra, J.Y. Lin, R.Y. Tsien, D.J. Anderson, Nat. Methods 11 (2014) 325–332. 108. V. Haikala, M. Joesch, A. Borst, A.S. Mauss, J. Neurosci. 33 (2013) 13927–13934. Fig. 1. Time course of expression in the Drosophila embryonic CNS and PNS of puc, a JNK activity reporter. (A–F) 6 snapshots (1 h apart each) generated from a time-lapse recording of a ventrally positioned pucE69I-GAL4 > UAS-GFP embryo. Monitoring was initiated at embryonic stage 13 and lasted for 6 h (anterior is to the left). puckered positive cells, shown in green, are detected initially at the ventral midline and then both at the Central and Peripheral Nervous System, while puckered expression pattern increases its complexity as embryogenesis proceeds. Images were acquired live in a Zeiss LSM 700 confocal microscope. Fig. 2. Conformal scanning harmonic microscopy. Sample mounting for upright microscopy; excitation (Exc.) and detection geometry (2PEF epidetected, THG-SHG transmitted) (from Olivier et al., Science 239 (2010) 967–971, with permission).