DeepD3, an open framework for automated quantification of dendritic spines.

Dendritic spines are the seat of most excitatory synapses in the brain, and a cellular structure considered central to learning, memory, and activity-dependent plasticity. The quantification of dendritic spines from light microscopy data is usually performed by humans in a painstaking and error-prone process. We found that human-to-human variability is substantial (inter-rater reliability 82.2±6.4%), raising concerns about the reproducibility of experiments and the validity of using human-annotated 'ground truth' as an evaluation method for computational approaches of spine identification. To address this, we present DeepD3, an open deep learning-based framework to robustly quantify dendritic spines in microscopy data in a fully automated fashion. DeepD3's neural networks have been trained on data from different sources and experimental conditions, annotated and segmented by multiple experts and they offer precise quantification of dendrites and dendritic spines. Importantly, these networks were validated in a number of datasets on varying acquisition modalities, species, anatomical locations and fluorescent indicators. The entire DeepD3 open framework, including the fully segmented training data, a benchmark that multiple experts have annotated, and the DeepD3 model zoo is fully available, addressing the lack of openly available datasets of dendritic spines while offering a ready-to-use, flexible, transparent, and reproducible spine quantification method.

Negative selection and stringency modulation in phage-assisted continuous evolution.

Phage-assisted continuous evolution (PACE) uses a modified filamentous bacteriophage life cycle to substantially accelerate laboratory evolution experiments. In this work, we expand the scope and capabilities of the PACE method with two key advances that enable the evolution of biomolecules with radically altered or highly specific new activities. First, we implemented small molecule-controlled modulation of selection stringency that enables otherwise inaccessible activities to be evolved directly from inactive starting libraries through a period of evolutionary drift. Second, we developed a general negative selection that enables continuous counterselection against undesired activities. We integrated these developments to continuously evolve mutant T7 RNA polymerase enzymes with ∼10,000-fold altered, rather than merely broadened, substrate specificities during a single three-day PACE experiment. The evolved enzymes exhibit specificity for their target substrate that exceeds that of wild-type RNA polymerases for their cognate substrates while maintaining wild type-like levels of activity.

Distributed chromatic processing at the interface between retina and brain in the larval zebrafish.

Larval zebrafish (Danio rerio) are an ideal organism for studying color vision, as their retina possesses four types of cone photoreceptors, covering most of the visible range and into the UV.1,2 Additionally, their eye and nervous systems are accessible to imaging, given that they are naturally transparent.3-5 Recent studies have found that, through a set of wavelength-range-specific horizontal, bipolar, and retinal ganglion cells (RGCs),6-9 the eye relays tetrachromatic information to several retinorecipient areas (RAs).10-13 The main RA is the optic tectum, receiving 97% of the RGC axons via the neuropil mass termed arborization field 10 (AF10).14,15 Here, we aim to understand the processing of chromatic signals at the interface between RGCs and their major brain targets. We used 2-photon calcium imaging to separately measure the responses of RGCs and neurons in the brain to four different chromatic stimuli in awake animals. We find that chromatic information is widespread throughout the brain, with a large variety of responses among RGCs, and an even greater diversity in their targets. Specific combinations of response types are enriched in specific nuclei, but there is no single color processing structure. In the main interface in this pathway, the connection between AF10 and tectum, we observe key elements of neural processing, such as enhanced signal decorrelation and improved chromatic decoding.16,17 A richer stimulus set revealed that these enhancements occur in the context of a more distributed code in tectum, facilitating chromatic signal association in this small vertebrate brain.

Properties of the Visible Light Phototaxis and UV Avoidance Behaviors in the Larval Zebrafish.

For many organisms, color is an essential source of information from visual scenes. The larval zebrafish has the potential to be a model for the study of this topic, given its tetrachromatic retina and high dependence on vision. In this study we took a step toward understanding how the larval zebrafish might use color sensing. To this end, we used a projector-based paradigm to force a choice of a color stimulus at every turn of the larva. The stimuli used spanned most of the larval spectral range, including activation of its Ultraviolet (UV) cone, which has not been described behaviorally before. We found that zebrafish larvae swim toward visible wavelengths (>400 nm) when choosing between them and darkness, as has been reported with white light. However, when presented with UV light and darkness zebrafish show an intensity dependent avoidance behavior. This UV avoidance does not interact cooperatively with phototaxis toward longer wavelengths, but can compete against it in an intensity dependent manner. Finally, we show that the avoidance behavior depends on the presence of eyes with functional UV cones. These findings open future avenues for studying the neural circuits that underlie color sensing in the larval zebrafish.