Dustin M. Graham

CRISPR/Cas9 corrects retinal dystrophy in rats.

Gene therapy is a therapeutic strategy where, as opposed to using traditional drugs, doctors manipulate specific genes in patients suffering from inherited genetic diseases, such as cystic fibrosis and severe combined immunodeficiency (SCID). Although conceptualized in the 1970s, gene therapy has had little clinical success and has suffered several setbacks as an approach to treating disease. Much of the difficulty in applying gene therapy lies in the technical difficulty of targeting and disrupting faulty genes. However, recent advances in gene editing technology, including the development of the clustered regularly interspaced short palindromic repeats (CRISPR)/Cas9 system, are providing new options and excitement for realizing the potential of gene therapy. As a proof-of-principle study, Benjamin Bakondi and colleagues at Cedars-Sinai Medical Center (Los Angeles, CA) have now successfully applied CRISPR/Cas9 to a rat model of retinitis pigmentosa, an inherited degenerative eye disease that causes severe loss of photoreceptors leading to visual impairment and blindness (Mol. Ther. doi: 10.1038/mt.2015.220; published online 15 December 2015). Transgenic S334ter rats suffer from an autosomal-dominant mutation in an allele for the rhodopsin gene (Rho), and display similar visual phenotypes to humans suffering from retinitis pigmentosa caused by Rho mutations. Using CRISPR/Cas9, Bakondi et al. were able to disrupt the allele-specific RhoS334 and rescue the effects of retinitis pigmentosa. Bakondi et al. injected plasmids containing targeting-guide RNA and the Cas9 enzyme directly into the eyes of S334ter rat pups at age P0, along with a fluorescent dye for histological verification of the spread of the injection. Using fluorescent confocal microscopy, they confirmed that the plasmids were successfully taken up by photoreceptors in the injected retinas. To test how well the CRISPR/ Cas9 strategy functionally rescued retinal degeneration, the researchers performed histology on injected retinal tissue at different periods throughout development of the S334ter rats and into adulthood. Bakondi et al. found that injected retinas had significantly higher levels of photoreceptors compared with retinas in non-injected eyes. Further, they showed immunohistochemical evidence that synaptic connectivity was maintained between photoreceptors and downstream neurons. Most importantly, they found that intact eyes with the injections had enhanced visual acuity (as tested by the optokinetic reflex response) compared with eyes that were not injected, demonstrating that this CRISPR/Cas9 strategy functionally rescued the deleterious effects of the RhoS334 mutation. Although only a proof-of-principle study, this work provides an important advancement for developing gene therapy as a treatment option for patients suffering from genetic disorders. Dustin M. Graham Using epifluorescence image capture, Tsutsumi et al. were able to reconstruct three-dimensional images that demonstrated the structure of this regenerated limb and the concave, socket-like morphology of the regenerated joint. They also observed in these images that the major flexor and extensor muscles (the biceps and triceps, respectively) inserted into the regenerated spike in a pattern than resembled that of the intact forelimb. Together these findings suggest that a functional elbow structure was regenerated at the elbow-end of the spike cartilage. To explain these findings, the researchers propose that regeneration depends upon interactions between remaining tissues and the regenerating tissues. Such interactions guide the growth and integration of regenerated tissues into remaining tissues in a process that Tsutsumi et al. call ‘reintegration’. By further studying these interactions and the process of reintegration, the scientists hope to unravel the mechanisms that allow functional regeneration in vertebrate species. Gregory D. Larsen described the regeneration of amputated joints in a species of newt, Tsutsumi’s team hypothesized that an adult frog might be able to regenerate a complete elbow joint under specific circumstances. The researchers amputated limbs at the elbow and studied the cartilaginous spike that grew in its place. Although they had removed the distal side of the elbow joint, frogs were able to bend and stretch the regenerated forelimb at the elbow (Regeneration doi:10.1002/reg2.49; published online 6 January 2016). Amphibians are an interesting clade for the study of regenerative medicine. Whereas most other tetrapods are unable to regenerate lost limbs, some amphibians can. Throughout their lives, urodeles (newts and salamanders) are able to regenerate the complete structure of a lost limb in a smaller but fully functioning replacement. Juvenile anurans (frogs and toads) have a similar ability during the larval stage, but adults only regenerate ‘spikes’ or cartilaginous rods that lack the joints and structure of the lost limb. This is the case with adult Xenopus laevis, the African clawed frog that has become a popular model for biomedical research. Because anurans show this incomplete regeneration in adulthood, they are considered an intermediate model between the complete regenerative ability of urodeles and the absence of regenerative ability in mammals. Researchers, including Rio Tsutsumi and colleagues at Kyoto University (Japan), hope that by achieving functional limb regeneration in frogs they can help bridge the gap between regenerative amphibians and non-regenerative tetrapods, such as humans. Building upon previous work that Amputated amphibians advance regenerative medicine

A walk on the wild side

wild, not yet at least. “These guys were bred in a lab at New York University, actually, so they’re big city mice.” Spotting a trap with some movement inside, Graham scans the first catch of the day, making use of the radio-frequency chip implanted in each mouse for tracking. “Mouse 51-51-57,” Graham announces. “We caught it last time,” says Sriveena Chittamuri, a visiting undergraduate. “We have enough poop samples, but we still need urine.” the wild(ish) things are, and where Andrea Graham is about to start the safari. Graham, an evolutionary ecologist at Princeton University, hops over the enclosure’s fence and gathers her students for their weekly reconnaissance. Walking through the weeds, she inspects the Longworth traps her team laid out the night before, seeing if any have been tripped. Longworth traps are designed for behavioral ecologists to capture, without injury, small wild mammals, such as mice. But as Graham explains, the mice they are For most immunologists, the big red barn that sits on a farm near Princeton, New Jersey, would be an odd place to store data. It has no lab benches or freezers. The inside walls are covered in cobwebs, and overhead, dusty blue tarps stretch across the rafters, arched downwards from the weight of petrified bat guano. But standing on the barn’s floor is a server, humming away as it collects wireless data from a unique source outside. The source is a circular 1,500 square-meter open-air enclosure sitting in a field at the farm’s edge. This is where A walk on the wild side

Exploring social cognition with marmosets.

tions, combined with their innate prosocial behaviors, enables scientists to use marmosets as models of social communication and behavior4. Recent work with marmosets demonstrated that this species shares several similarities of social behavior and language development with humans, despite their relatively wide evolutionary distance. Marmosets exhibit prosocial behaviors, pair-bond with mates and engage in cooperative raising of young1,4. Cooperative breeding is a particularly rare trait among primates, making humans and marmosets potentially well suited for comparative studies of social behavior. Additionally, marmoset infants rely on parental interaction to develop vocalizations, which is uncommon to most primates but a shared feature of human language development4. The brain structure of marmosets is also similar to that of other primates, including humans4. Taking into consideration recent developments in genetics and miniaturized brainrecording tools, marmosets are poised to become an extremely valuable model for understanding the neural basis of social cognition and language development. Although novel technologies, such as brain recording techniques and transgenics, can also be applied to other nonhuman primates, it remains a challenge to create behavioral paradigms with rich social structures that mimic those that accompany human development. Marmosets, therefore, will increasingly be relied upon to help bridge this gap in the ability of scientists to study the development and neural mechanisms underlying complex social behaviors, which might lead to a better understanding of neuropsychiatric disorders involving social development. Overall, these small monkeys are ready to make a big impact on social neuroscience and the myriad of complex mental disorders that affect humans.

Dirty mice might make better models.

microbes that normal mice encounter in the wild.” In their experiments, the researchers focused on studying the composition of memory T cells, which are critical for adaptive immune responses to infections and cancer. The group compared the immune systems of human newborns and adults with those of ‘clean’ laboratory mice and wildcaught or ‘dirty’ pet-store mice. Unlike the human adult immune system—but similar to that of human newborns— lab mice lacked effector-differentiated and mucosally distributed T cells. To determine if this trend in lab mice was due to a lack of challenge by typical environmental microbes, the research group tested T cell types in wild-caught feral mice and petstore mice. They found that the immune systems of both wild-caught and pet-store mice had significantly higher levels of effector-differentiated T cells, a signature of the human adult immune system. They further demonstrated that, after adding pet-store mice into the cages of lab mice, the immune By housing laboratory mice in hygienic and specific pathogen free barrier facilities, scientists can tightly control conditions for studying immune system function and disease. However, owing to recent failures in translating findings from mouse models to the clinic, there are growing concerns that mice might not be an appropriate species for modeling human disease. In a recent report, a group led by Stephen Jameson and David Masopust at the University of Minnesota show that, rather than blaming the mice, scientists should start taking into account their facility’s housing conditions (Nature 532, 512–516; 2016). The study’s goal was to compare the immune systems of laboratory mice with those of humans, to determine what effects standard barrier facility housing conditions might have on translating mouse models to adult humans. According to Jameson, “Standard lab mice don’t reflect important features of the adult human immune system. We wanted to know whether this is because lab animals are shielded from systems of clean lab mice began to look less like those of human newborns, and more like those of human adults. Overall, the results demonstrate the importance of the environment in shaping the makeup and function of the immune system. They also highlight that tight control over lab and vivarium conditions comes with a translational price tag. As said by Masopust, “Utilizing this [‘dirty’] model to test vaccinations and therapeutics for cancer or transplantation may better predict how these will perform in humans.” Dustin M. Graham

Putting the brakes on CRISPR-Cas9 gene drive systems.

Throughout the brain in a broad range of animal species, synaptic connections between neurons are arranged in layers, or laminae. Laminae are formed by the axons of input neurons connecting with the dendrites of target neurons, and laminae are distinctly organized such that a single lamina contains synapses that share similar functional properties. While research over the last few decades has elucidated many of the molecular mechanisms that guide the formation of this synaptic layering, the purpose of synaptic lamination remains largely unclear. In a recent study, researchers Nikolas Nikolaou and Martin P. Meyer of King’s College London (UK) asked whether synaptic lamination is crucial to the development and function of neural circuits (Neuron 88, 999–1013; 2015). The authors addressed this question using a mutant zebrafish line that lacks the typically organized synaptic lamination of input axons from retinal ganglion cells (RGC) of the eye into the optic tectum, a brain structure that controls higher level visual functions such as prey capture. Neurons in the optic tectum are direction-selective, meaning that they respond specifically to visual stimuli moving in one direction. This directional specificity in tectal neurons is driven by precise connections with RGCs that are themselves direction-selective. By comparing direction-selectivity in tectal neurons at early and late stages of synapse formation, Nikolaou and Meyer found that although the absence of synaptic layering slowed circuit development in the mutant zebrafish, ultimately the tectal circuits in mutant and wildtype zebrafish became functionally indistinguishable. Nikolaou and Meyer also found that although the mutant zebrafish never regained synaptic lamination, the tectal circuits exhibited full functional recovery of direction-selectivity. When they examined how such a recovery could occur, the authors discovered that the direction-selective tectal neurons adjusted their dendrites to find and connect with incoming RGC axons that they needed to pair with in order to receive directional information. Nikolaou and Meyer attribute this connection in the absence of laminar organization to structural plasticity, or the ability of neurons to alter their morphology and synapses. These findings demonstrate that the developing brain can overcome the loss of even fundamental and conserved processes such as synaptic lamination. This work suggests that one important purpose of synaptic lamination might be to speed the development of neuronal circuits, and it shows that if given enough time, a disordered brain can set itself straight. Jocelyn Lippman-Bell genes escaping into wild populations (Nat. Biotechnol. 33, 1250–1255; 2015). Typically, gene drive systems based on CRISPR-Cas9 are self-sufficient, with all the components built-in that are necessary for targeting and driving a specific gene. DiCarlo et al. split the drive system into two separate components: the Cas9 endonuclease, which physically cuts targeted genes, and the guide RNA that is necessary to drive the inheritance of the edited gene. This splitdrive system ensures that even if genetically altered yeast were to escape from the lab, mating with wild-type yeast would quickly separate Cas9 from the RNA drive, greatly slowing the spread of altered genes through the wild population and minimizing their impact. “The gene drive research community has been actively discussing what should be done to safeguard shared ecosystems, and now we have demonstrated that the proposed safeguards work extremely well and should therefore be used by every gene drive researcher in every relevant lab organism”, commented Kevin Esvelt, a senior author of the study, in a press release. Dustin M. Graham While this can be beneficial, for instance when changing a malfunctioning gene that is associated with a specific disease, if an offtarget gene is accidentally altered and then spread through a population, it could have harmful and permanent consequences. To address these issues, James DiCarlo and colleagues at the Wyss Institute for Biologically Inspired Engineering at Harvard University (Cambridge, MA), developed an experimental paradigm with the CRISPR-Cas9 gene drive system that minimizes the risk of synthetic Genetic engineering of animals, plants and microorganisms has been a mainstay technique in biomedical science and comparative medicine for decades and has greatly increased our knowledge of disease-states and basic biological function. Recent discoveries and breakthroughs in genetic engineering, including zinc finger nucleases, transcription activator-like effector nucleases and clustered regularly interspaced palindromic repeats (CRISPR), have made it easier and faster than ever to produce powerful genetic models. The potential for this technology extends beyond biomedical research and into the biotechnology sector where researchers are already seeking applications for engineering pest-resistant crops and drug-producing yeast. However, as gene editing technologies have grown faster and easier to use, so too have concerns about potential unintended consequences. ‘Gene drive systems’ that use CRISPR and the associated Cas9 endonuclease enable edited genes to be inherited and passed along through a population at exponential rates that are much higher than those of typical Mendelian inheritance. Putting the brakes on CRISPR-Cas9 gene drive systems

A new window sheds light on lung tumor metastasis

implanted over the lungs for high-resolution cellular imaging. Because the procedure completely reseals the thoracic cavity after implantation, the mice can recover and breath comfortably on their own, ready for reimaging over the course of weeks. Using mouse models of lung tumor metastasis, the team was able to image individual tumor cells invading and expanding in the murine lungs, an accomplishment that opens up new avenues for testing long-held assumptions about tumor metastasis in the lungs. “Our interests going forward are to focus in on the lung and focus in on metastasis,” says Entenberg. “Before it was a black box and now we have the ability to directly visualize it over days and days and see what actually happens. We have the ability to look at that physiological process over time and test conclusions that have been based upon old experimental metastasis assays, and find out which ones hold up under more realistic biological conditions.” Dustin M. Graham experience. Sonia Voiculescu, a surgical resident at Einstein, joined the Condeelis lab, and teamed up with Entenberg to provide the surgical expertise and perspective necessary to turn Entenberg’s crazy idea into a reality. The two set out to develop a novel imaging window in the mouse lungs that would enable the animals to wake up and fully recover after surgery, allowing for serial microscopy in the same lungs at cellular resolution. “I had developed the engineering design for the window,” says Entenberg, “and Sonia came in with her surgical skills and together we started pushing forward what we call surgical engineering.” Rather than developing a robot or new material for the operating room, which, according to Entenberg, is the typical province of surgical engineering, “we flipped the concept on its head and brought the knowledge and skills that surgeons have in the OR and combined it with our engineering to develop a novel protocol.” The protocol, described in detail in their paper, uses a small glass window The cellular mechanisms of tumor invasion and expansion during metastasis in the lungs have been a black box for decades, understood only by inference using relatively low-resolution imaging techniques. But thanks to some moxie and a T32 training grant at Montefiore’s Department of Surgery, the precise details in mouse models are beginning to reveal themselves, as demonstrated in a recent Nature Method’s paper (Nat. Methods 15, 73–80; 2018). David Entenberg is a physicist and engineer who develops novel instrumentation for microscopy at the biophotonics center at the Einstein College of Medicine, where he has been collaborating with cancer researcher John Condeelis since 2006. A couple of years ago the two teamed up to apply a vacuum lung imaging window to study tumor biology in mouse lungs using intravital imaging. Although successful, the group had pushed the surgical technique to its limits and were frustrated by the small experimental timeframe allowed by the method, which requires a highly invasive surgical procedure in terminally anesthetized animals. To capture cellular mechanisms of events like tumor metastasis, a common occurrence in the lungs and a primary scientific aim of the Condeelis lab, the team needed to be able to image lungs in mice over a timespan of weeks, not the 12 hours typical of terminally anesthetized preps. “So I tossed out the idea, why don’t we try to make a window where the mouse could survive and wake-up,” says Entenberg, “and everybody basically said, there’s no way that’s going to be possible.” This happened to coincide with the Montefiore Medical Center (the university hospital of Einstein) receiving a T32 training grant from the NIH to specifically bring surgeons into the lab to get some science A new window sheds light on lung tumor metastasis

Small structures on a big scale

Using 4 different designed structures (2 novel, and 2 previously generated in other reports), the team demonstrated their protocol’s generalizability. Likewise, the group showed their ability to produce large quantities of DNA origami structures (~167 mg) using a laboratory-scale stirred-tank bioreactor. The team estimated the cost of their production protocol to be €23 per milligram, but note that with even more efficient and standard industry methods for bacteriophage production, the cost could drop to as low as €0.18 per milligram. Dustin M. Graham are held in place using shorter ‘staple strands.’ While long scaffolds can be easily generated using standard bacteriophage methods in the lab, producing short staple strands requires much more expensive synthesis processes. To solve this problem, Dietz and colleagues incorporated small self-cleaving DNA cassettes (called DNAzymes) into longer stretches of DNA that can also be easily produced en mass by bacteriophages. These cassettes, when activated by addition of zinc, automatically cleave themselves out, producing shorter staple strands that combine with scaffolds produced in parallel to form origami DNA structures. DNA origami, the bottom-up self-assembly of designed 3-D nanostructures, is a promising new method for producing complex materials and delivery vehicles for therapeutic agents. But the process for developing DNA origami structures is expensive and time-consuming, limiting the batch sizes of final products to just micrograms. In a recent paper in Nature, a group in Germany led by Hendrik Dietz has developed an efficient workflow for mass production of DNA origami using standard laboratory methods that won’t break the bank (Nature 552, 84–87; 2017). DNA origami structures are assembled from a long single-stranded scaffolds, which

Putting down the pineal gland in zebrafish

zebrafish, the master circadian clock has traditionally been thought to reside in the pineal gland, but recent results demonstrating clock function in other tissue has brought this view into question. To further assess the role pineal cells play in driving circadian rhythms, the group developed a novel model where melatonin expressing pineal cells had their molecular clock selectively blocked with a dominantnegative strategy. The group found evidence that circadian function in locomotor and place preference had been disrupted, but other tissues outside the pineal gland maintained their core clock function. Overall the results suggest that the pineal gland plays an important role in modulating rhythmic behaviors, but may be part of a multi-pacemaker distributed system governing organismal-level circadian function. Importantly, the novel transgenic model developed in this work may help future studies uncover important mechanisms for multi-system clock function. Dustin M. Graham rhythms, a systems level understanding is only beginning to come into focus. For instance, although the SCN is necessary to maintain circadian rhythms, it is also known that multiple tissues can maintain their own independent clocks, suggesting the SCN may act more akin to a conductor guiding a larger orchestra of rhythms. Understanding how circadian rhythms arise and are maintained at this systems level can be difficult in mammalian models, and some researchers turn to other model organisms. One group of researchers, led by Yoav Gothilf at Tel-Aviv University, has recently applied a genetic approach in zebrafish to generate a novel knock-down model to better understand circadian function in this species (PLoS Genetics 12, e1006445; 2016). In Circadian rhythms are a critical part of normal biology, as evidenced by virtually all animals displaying a roughly 24-hr cycle of key physiological functions. Although typically studied at higher behavioral levels in vertebrates, like sleep-wake cycles, recent research indicates that circadian rhythms can have important effects at the cellular and network level, and can influence a wide-range of systems processes, including immune function and cancer development. Owing to these facts, scientists have consistently tried to crack the mechanisms driving circadian rhythms in whole organisms. In mammals, the master circadian clock is located in the suprachiasmatic nucleus (SCN), with individual SCN neurons independently capable of maintaining rhythmic changes in activity owing to a precise molecular network of genes switched on and off over a roughly 24-hr cyclic period. However, despite a relatively clear picture of the molecular framework underlying cellular circadian M irk o_ R os en au /G et ty

Improved multi-label imaging using hyperspectral phaser analysis

version of the algorithm that would enable scientists to more easily take advantage of tools like brainbow mice and zebrafish for lineage analysis. With some small tweaks to the software, they hope to add some user-friendly features that can identify and isolate cells with the same combination of colors, allowing researchers to simply switch channels and follow daughter cells with distinct origins. Ultimately, Cutrale’s game plan is to help push HySP into the surgery room, and he and colleagues in the Fraser Lab are already in discussion with industry leaders like Intuitive Surgical Inc., which developed the Da Vinci Surgical System, a robotic surgical platform now commonly used for prostate surgeries. As Cutrale notes, “They already have the interface for this; it’s all digital, so the surgeon doesn’t touch the patient anymore. He is sitting somewhere else looking at an image. What if we could augment that image to better separate, for example, muscle from nerve?” Dustin M. Graham images. Second is photobleaching and subsequent damage to living tissue, which increases with additional probes and excitation lines. Additionally, autofluorescence can be a significant factor affecting the quality of images, especially in vivo where a variety of structures with different cell types and densities can create difficult backgrounds. Lastly, there is the issue of speed. Previous algorithms have been developed to de-noise data and solve some of the above problems, but are computationally intense and require far too much processing time, especially for live imaging experiments where close to real-time data is important. Working off of previously developed phasor analyses for multispectral imaging, which use computationally efficient fourier transformations of data, Cutrale et al. developed HySP and demonstrate its ability to quickly and accurately separate up to eight unmixed signals in developing zebrafish embryos over long imaging periods. HySP takes the high-dimensional multispectral information for each pixel in an image, transforms it into a single point on a 2D phasor plot, then applies an algorithm that rapidly reduces spectral noise, removes autofluorescence, and cleanly separates multiple signals from fluorescent markers with overlapping spectra. As demonstrated in their paper, images with multiple labels are significantly improved. Importantly, the denoising process enabled images to be collected under low signal-tonoise conditions, allowing for low laser levels to be used during time-lapse live imaging, with no noticeable damage to growing zebrafish embryos. For Cutrale and colleagues, this proof-ofprinciple is just the beginning. They see several future opportunities to apply their algorithm in a wide range of fields. The team is working on developing a semi-automated Software for separating spectra helps scientists, and potentially clinicians, sort through the signals. The need to identify friend and foe is nowhere more apparent than in the operating room, where surgeons work diligently to distinguish and remove tumors with minimal damage to surrounding muscle and nerve tissue. New software developed by Francesco Cutrale and colleagues in the lab of Scott Fraser at University of Southern California could potentially aid this process, and along the way, help scientists make the most of a variety of imaging tools in zebrafish and mice. For Cutrale, translating his work in microscopy methods across scientific disciplines and into the clinic is what it’s all about. “Biologists use genetics in animal models, because the genetics can be somewhat translated back to humans. In my case, I’m ignorant of genetics, but I work on a core, which is an algorithm, and that algorithm can be translated between fields, and that is the real strength of this method.” The algorithm he developed and packaged into a new software platform, called HyperSpectral Phasors—or HySP for short— allows unambiguous unmixing of multiple fluorophores with overlapping spectra, and significantly improves time-lapse in vivo imaging of multiple cell and tissue types (Nat. Methods 14, 149–152 (2017)). Simultaneous imaging of multiple structures and cells in living whole organisms has advanced greatly owing to continuous improvements in fluorescent probes. But several challenges still await the scientist that dares make use of more than two or three separate labels in the same tissue. First, although researchers have a wide variety of probes to choose from, many have overlapping excitation and emission spectra, making it difficult to accurately distinguish one from the other in resulting Improved multi-label imaging using hyperspectral phaser analysis

Flower power controls the flow of synaptic transmission

two stimulation paradigms, promoting CME independently of calcium channeling during moderate stimulus conditions, but driving calcium influxes leading to ADBE during in the presence of intense stimulation. Flower, therefore, may play a key role in linking stimulus-dependent forms of synaptic vesicle endocytosis. Dustin M. Graham participate in activity-dependent bulk endocytosis (ADBE). CME and ADBE are the two predominant forms of vesicle endocytosis, but occur under very different stimulation conditions: CME during moderate nerve stimulation, and ADBE under intense stimulation. Importantly, this new work shows that Flower proteins behave quite differently under the The calcium channel Flower (Fwe) helps regulate normal vs. bulk endocytosis during different stimulation paradigms. When an action potential reaches the end of the line at a nerve terminal, a complex series of molecular events culminates in the release of neurotransmitters: a process referred to as ‘synaptic transmission.’ The regulation of synaptic transmission is critical to understanding information processing in the nervous, and, importantly, how that processing can go awry. In new work published in PLoS Biology, Chi-Kuang Yao et al. take a deeper look at Flower (Fwe), a synaptic vesicle associated calcium channel they previously identified in a genetic screen of Drosophila (PLoS Biol. 15, e2000931; 2017). Previously shown to be associated with clathrin-mediated endocytosis (CME), a form of retrieving synaptic vesicles for future release of neurotransmitter, Flower is now demonstrated to also A model for different roles of Fwe in CME and ADBE. Adapted from PLoS Biol. 15, e2000931 (2017). ENDOCYTIC ZONE ADBE CME Moderate activity