Zero-shot learning enables instant denoising and super-resolution in optical fluorescence microscopy.

Computational super-resolution methods, including conventional analytical algorithms and deep learning models, have substantially improved optical microscopy. Among them, supervised deep neural networks have demonstrated outstanding performance, however, demanding abundant high-quality training data, which are laborious and even impractical to acquire due to the high dynamics of living cells. Here, we develop zero-shot deconvolution networks (ZS-DeconvNet) that instantly enhance the resolution of microscope images by more than 1.5-fold over the diffraction limit with 10-fold lower fluorescence than ordinary super-resolution imaging conditions, in an unsupervised manner without the need for either ground truths or additional data acquisition. We demonstrate the versatile applicability of ZS-DeconvNet on multiple imaging modalities, including total internal reflection fluorescence microscopy, three-dimensional wide-field microscopy, confocal microscopy, two-photon microscopy, lattice light-sheet microscopy, and multimodal structured illumination microscopy, which enables multi-color, long-term, super-resolution 2D/3D imaging of subcellular bioprocesses from mitotic single cells to multicellular embryos of mouse and C. elegans.

Transfer of polarity information via diffusion of Wnt ligands in C. elegans embryos.

Different signaling mechanisms concur to ensure robust tissue patterning and cell fate instruction during animal development. Most of these mechanisms rely on signaling proteins that are produced, transported, and detected. The spatiotemporal dynamics of signaling molecules are largely unknown, yet they determine signal activity's spatial range and time frame. Here, we use the Caenorhabditis elegans embryo to study how Wnt ligands, an evolutionarily conserved family of signaling proteins, dynamically organize to establish cell polarity in a developing tissue. We identify how Wnt ligands, produced in the posterior half of the embryos, spread extracellularly to transmit information to distant target cells in the anterior half. With quantitative live imaging and fluorescence correlation spectroscopy, we show that Wnt ligands diffuse through the embryo over a timescale shorter than the cell cycle, in the intercellular space, and outside the tissue below the eggshell. We extracted diffusion coefficients of Wnt ligands and their receptor Frizzled and characterized their co-localization. Integrating our different measurements and observations in a simple computational framework, we show how fast diffusion in the embryo can polarize individual cells through a time integration of the arrival of the ligands at the target cells. The polarity established at the tissue level by a posterior Wnt source can be transferred to the cellular level. Our results support a diffusion-based long-range Wnt signaling, which is consistent with the dynamics of developing processes.

C. elegans Afadin is required for epidermal morphogenesis and functionally interfaces with the cadherin-catenin complex and RhoGAP PAC-1/ARHGAP21.

During epithelial morphogenesis, the apical junctions connecting cells must remodel as cells change shape and make new connections with their neighbors. In the C. elegans embryo, new apical junctions form when epidermal cells migrate and seal with one another to encase the embryo in skin ('ventral enclosure'), and junctions remodel when epidermal cells change shape to squeeze the embryo into a worm shape ('elongation'). The junctional cadherin-catenin complex (CCC), which links epithelial cells to each other and to cortical actomyosin, is essential for C. elegans epidermal morphogenesis. RNAi genetic enhancement screens have identified several genes encoding proteins that interact with the CCC to promote epidermal morphogenesis, including the scaffolding protein Afadin (AFD-1), whose depletion alone results in only minor morphogenesis defects. Here, by creating a null mutation in afd-1, we show that afd-1 provides a significant contribution to ventral enclosure and elongation on its own. Unexpectedly, we find that afd-1 mutant phenotypes are strongly modified by diet, revealing a previously unappreciated parental nutritional input to morphogenesis. We identify functional interactions between AFD-1 and the CCC by demonstrating that E-cadherin is required for the polarized distribution of AFD-1 to cell contact sites in early embryos. Finally, we show that afd-1 promotes the enrichment of polarity regulator, and CCC-interacting protein, PAC-1/ARHGAP21 to cell contact sites, and we identify genetic interactions suggesting that afd-1 and pac-1 regulate epidermal morphogenesis at least in part through parallel mechanisms. Our findings reveal that C. elegans AFD-1 makes a significant contribution to epidermal morphogenesis and functionally interfaces with core and associated CCC proteins.

Cell-cycle control: Timing is everything for the Plk1-Bub1 partnership.

Bub1 and Polo kinases are well-known multitasking regulators of mitosis. New work shows that they team up at kinetochores to determine the mitotic duration of embryonic divisions in nematodes. As is often the case, the key effector is Cdc20 activity.

Inheritance of paternal DNA damage by histone-mediated repair restriction.

How paternal exposure to ionizing radiation affects genetic inheritance and disease risk in the offspring has been a long-standing question in radiation biology. In humans, nearly 80% of transmitted mutations arise in the paternal germline1, but the transgenerational effects of ionizing radiation exposure has remained controversial and the mechanisms are unknown. Here we show that in sex-separated Caenorhabditis elegans strains, paternal, but not maternal, exposure to ionizing radiation leads to transgenerational embryonic lethality. The offspring of irradiated males displayed various genome instability phenotypes, including DNA fragmentation, chromosomal rearrangement and aneuploidy. Paternal DNA double strand breaks were repaired by maternally provided error-prone polymerase theta-mediated end joining. Mechanistically, we show that depletion of an orthologue of human histone H1.0, HIS-24, or the heterochromatin protein HPL-1, could significantly reverse the transgenerational embryonic lethality. Removal of HIS-24 or HPL-1 reduced histone 3 lysine 9 dimethylation and enabled error-free homologous recombination repair in the germline of the F1 generation from ionizing radiation-treated P0 males, consequently improving the viability of the F2 generation. This work establishes the mechanistic underpinnings of the heritable consequences of paternal radiation exposure on the health of offspring, which may lead to congenital disorders and cancer in humans.

A condensate dynamic instability orchestrates actomyosin cortex activation.

A key event at the onset of development is the activation of a contractile actomyosin cortex during the oocyte-to-embryo transition1-3. Here we report on the discovery that, in Caenorhabditis elegans oocytes, actomyosin cortex activation is supported by the emergence of thousands of short-lived protein condensates rich in F-actin, N-WASP and the ARP2/3 complex4-8 that form an active micro-emulsion. A phase portrait analysis of the dynamics of individual cortical condensates reveals that condensates initially grow and then transition to disassembly before dissolving completely. We find that, in contrast to condensate growth through diffusion9, the growth dynamics of cortical condensates are chemically driven. Notably, the associated chemical reactions obey mass action kinetics that govern both composition and size. We suggest that the resultant condensate dynamic instability10 suppresses coarsening of the active micro-emulsion11, ensures reaction kinetics that are independent of condensate size and prevents runaway F-actin nucleation during the formation of the first cortical actin meshwork.

Modeling protein dynamics in Caenorhabditis elegans embryos reveals that the PLK-1 gradient relies on weakly coupled reaction-diffusion mechanisms.

SignificanceIntracellular gradients have essential roles in cell and developmental biology, but their formation is not fully understood. We have developed a computational approach facilitating interpretation of protein dynamics and gradient formation. We have combined this computational approach with experiments to understand how Polo-Like Kinase 1 (PLK-1) forms a cytoplasmic gradient in Caenorhabditis elegans embryos. Although the PLK-1 gradient depends on the Muscle EXcess-5/6 (MEX-5/6) proteins, we reveal differences in PLK-1 and MEX-5 gradient formation that can be explained by a model with two components, PLK-1 bound to MEX-5 and unbound PLK-1. Our combined approach suggests that a weak coupling between PLK-1 and MEX-5 reaction-diffusion mechanisms dictates the dynamic exchange of PLK-1 with the cytoplasm, explaining PLK-1 high diffusivity and smooth gradient.

Reprogramming the piRNA pathway for multiplexed and transgenerational gene silencing in C. elegans.

Single-guide RNAs can target exogenous CRISPR-Cas proteins to unique DNA locations, enabling genetic tools that are efficient, specific and scalable. Here we show that short synthetic guide Piwi-interacting RNAs (piRNAs) (21-nucleotide sg-piRNAs) expressed from extrachromosomal transgenes can, analogously, reprogram the endogenous piRNA pathway for gene-specific silencing in the hermaphrodite germline, sperm and embryos of Caenorhabditis elegans. piRNA-mediated interference ('piRNAi') is more efficient than RNAi and can be multiplexed, and auxin-mediated degradation of the piRNA-specific Argonaute PRG-1 allows conditional gene silencing. Target-specific silencing results in decreased messenger RNA levels, amplification of secondary small interfering RNAs and repressive chromatin modifications. Short (300 base pairs) piRNAi transgenes amplified from arrayed oligonucleotide pools also induce silencing, potentially making piRNAi highly scalable. We show that piRNAi can induce transgenerational epigenetic silencing of two endogenous genes (him-5 and him-8). Silencing is inherited for four to six generations after target-specific sg-piRNAs are lost, whereas depleting PRG-1 leads to essentially permanent epigenetic silencing.

Computable early Caenorhabditis elegans embryo with a phase field model.

Morphogenesis is a precise and robust dynamic process during metazoan embryogenesis, consisting of both cell proliferation and cell migration. Despite the fact that much is known about specific regulations at molecular level, how cell proliferation and migration together drive the morphogenesis at cellular and organismic levels is not well understood. Using Caenorhabditis elegans as the model animal, we present a phase field model to compute early embryonic morphogenesis within a confined eggshell. With physical information about cell division obtained from three-dimensional time-lapse cellular imaging experiments, the model can precisely reproduce the early morphogenesis process as seen in vivo, including time evolution of location and morphology of each cell. Furthermore, the model can be used to reveal key cell-cell attractions critical to the development of C. elegans embryo. Our work demonstrates how genetic programming and physical forces collaborate to drive morphogenesis and provides a predictive model to decipher the underlying mechanism.

Functional midbody assembly in the absence of a central spindle.

Contractile ring constriction during cytokinesis is thought to compact central spindle microtubules to form the midbody, an antiparallel microtubule bundle at the intercellular bridge. In Caenorhabditis elegans, central spindle microtubule assembly requires targeting of the CLASP family protein CLS-2 to the kinetochores in metaphase and spindle midzone in anaphase. CLS-2 targeting is mediated by the CENP-F-like HCP-1/2, but their roles in cytokinesis and midbody assembly are not known. We found that although HCP-1 and HCP-2 mostly function cooperatively, HCP-1 plays a more primary role in promoting CLS-2-dependent central spindle microtubule assembly. HCP-1/2 codisrupted embryos did not form central spindles but completed cytokinesis and formed functional midbodies capable of supporting abscission. These central spindle-independent midbodies appeared to form via contractile ring constriction-driven bundling of astral microtubules at the furrow tip. This work suggests that, in the absence of a central spindle, astral microtubules can support midbody assembly and that midbody assembly is more predictive of successful cytokinesis than central spindle assembly.

Asymmetry is defined during meiosis in the oocyte of the parthenogenetic nematode Diploscapter pachys.

Asymmetric cell division is an essential feature of normal development and certain pathologies. The process and its regulation have been studied extensively in the Caenorhabditis elegans embryo, particularly how symmetry of the actomyosin cortical cytoskeleton is broken by a sperm-derived signal at fertilization, upstream of polarity establishment. Diploscapter pachys is the closest parthenogenetic relative to C. elegans, and D. pachys one-cell embryos also divide asymmetrically. However how polarity is triggered in the absence of sperm remains unknown. In post-meiotic embryos, we find that the nucleus inhabits principally one embryo hemisphere, the future posterior pole. When forced to one pole by centrifugation, the nucleus returns to its preferred pole, although poles appear identical as concerns cortical ruffling and actin cytoskeleton. The location of the meiotic spindle also correlates with the future posterior pole and slight actin enrichment is observed at that pole in some early embryos along with microtubule structures emanating from the meiotic spindle. Polarized location of the nucleus is not observed in pre-meiotic D. pachys oocytes. All together our results are consistent with the idea that polarity of the D. pachys embryo is attained during meiosis, seemingly based on the location of the meiotic spindle, by a mechanism that may be present but suppressed in C. elegans.

Multiview confocal super-resolution microscopy.

Confocal microscopy1 remains a major workhorse in biomedical optical microscopy owing to its reliability and flexibility in imaging various samples, but suffers from substantial point spread function anisotropy, diffraction-limited resolution, depth-dependent degradation in scattering samples and volumetric bleaching2. Here we address these problems, enhancing confocal microscopy performance from the sub-micrometre to millimetre spatial scale and the millisecond to hour temporal scale, improving both lateral and axial resolution more than twofold while simultaneously reducing phototoxicity. We achieve these gains using an integrated, four-pronged approach: (1) developing compact line scanners that enable sensitive, rapid, diffraction-limited imaging over large areas; (2) combining line-scanning with multiview imaging, developing reconstruction algorithms that improve resolution isotropy and recover signal otherwise lost to scattering; (3) adapting techniques from structured illumination microscopy, achieving super-resolution imaging in densely labelled, thick samples; (4) synergizing deep learning with these advances, further improving imaging speed, resolution and duration. We demonstrate these capabilities on more than 20 distinct fixed and live samples, including protein distributions in single cells; nuclei and developing neurons in Caenorhabditis elegans embryos, larvae and adults; myoblasts in imaginal disks of Drosophila wings; and mouse renal, oesophageal, cardiac and brain tissues.

Insight into the interaction between the RNA helicase CGH-1 and EDC-3 and its implications.

Previous studies indicated that the P-body components, CGH-1 and EDC-3 may play a crucial role in the regulation of lifespan in Caenorhabditis elegans. Homo sapiens DDX6 or Saccharomyces cerevisiae Dhh1p (CGH-1 in C. elegans) could form complexes with EDC3 (Edc3p in yeast), respectively, which is significant for translation inhibition and mRNA decay. However, it is currently unclear how CGH-1 can be recognized by EDC-3 in C. elegans. Here, we provided structural and biochemical insights into the interaction between CGH-1 and EDC-3. Combined with homology modeling, mutation, and ITC assays, we uncovered an interface between CGH-1 RecA2 domain and EDC-3 FDF-FEK. Additionally, GST-pulldown and co-localization experiments confirmed the interaction between CGH-1 and EDC-3 in vitro and in vivo. We also analyzed PATR-1-binding interface on CGH-1 RecA2 by ITC assays. Moreover, we unveiled the similarity and differences of the binding mode between EDC-3 and CAR-1 or PATR-1. Taken together, these findings provide insights into the recognition of DEAD-box protein CGH-1 by EDC-3 FDF-FEK motif, suggesting important functional implications.

Novel LOTUS-domain proteins are organizational hubs that recruit C. elegans Vasa to germ granules.

We describe MIP-1 and MIP-2, novel paralogous C. elegans germ granule components that interact with the intrinsically disordered MEG-3 protein. These proteins promote P granule condensation, form granules independently of MEG-3 in the postembryonic germ line, and balance each other in regulating P granule growth and localization. MIP-1 and MIP-2 each contain two LOTUS domains and intrinsically disordered regions and form homo- and heterodimers. They bind and anchor the Vasa homolog GLH-1 within P granules and are jointly required for coalescence of MEG-3, GLH-1, and PGL proteins. Animals lacking MIP-1 and MIP-2 show temperature-sensitive embryonic lethality, sterility, and mortal germ lines. Germline phenotypes include defects in stem cell self-renewal, meiotic progression, and gamete differentiation. We propose that these proteins serve as scaffolds and organizing centers for ribonucleoprotein networks within P granules that help recruit and balance essential RNA processing machinery to regulate key developmental transitions in the germ line.

Isoflurane impairs oogenesis through germ cell apoptosis in C. elegans.

Anesthetic isoflurane has been reported to induce toxicity. However, the effects of isoflurane on fecundity remain largely unknown. We established a system in C. elegans to investigate the effects of isoflurane on oogenesis. Synchronized L4 stage C. elegans were treated with 7% isoflurane for 4 h. Dead cells, ROS, embryos, and unfertilized eggs laid by hermaphrodites were measured by fluorescence imaging and counting. The C. elegans with losses of ced-3, cep-1, abl-1, male C. elegans, and oxidative stress inhibitor N-acetyl-cysteine were used in the interaction studies. We found that isoflurane decreased the numbers of embryos and unfertilized eggs and increased the levels of dead cells and ROS in C. elegans. The isoflurane-induced impairment of oogenesis was associated with abl-1, ced-3, but not cep-1. N-acetyl-cysteine attenuated the isoflurane-induced impairment of oogenesis in C. elegans. Mating with male C. elegans did not attenuate the isoflurane-induced changes in oogenesis. These findings suggest that isoflurane may impair oogenesis through abl-1- and ced-3-associated, but not cep-1-associated, germ cell apoptosis and oxidative stress, pending further investigation. These studies will promote more research to determine the potential effects of anesthesia on fecundity.

A balance between antagonizing PAR proteins specifies the pattern of asymmetric and symmetric divisions in C. elegans embryogenesis.

Coordination between cell differentiation and proliferation during development requires the balance between asymmetric and symmetric modes of cell division. However, the cellular intrinsic cue underlying the choice between these two division modes remains elusive. Here, we show evidence in Caenorhabditis elegans that the invariable lineage of the division modes is specified by the balance between antagonizing complexes of partitioning-defective (PAR) proteins. By uncoupling unequal inheritance of PAR proteins from that of fate determinants during cell division, we demonstrate that changes in the balance between PAR-2 and PAR-6 can be sufficient to re-program the division modes from symmetric to asymmetric and vice versa in two daughter cells. The division mode adopted occurs independently of asymmetry in cytoplasmic fate determinants, cell-size asymmetry, and cell-cycle asynchrony between sister cells. We propose that the balance between PAR proteins represents an intrinsic self-organizing cue for the specification of the two division modes during development.

A 4D single-cell protein atlas of transcription factors delineates spatiotemporal patterning during embryogenesis.

Complex biological processes such as embryogenesis require precise coordination of cell differentiation programs across both space and time. Using protein-fusion fluorescent reporters and four-dimensional live imaging, we present a protein expression atlas of transcription factors (TFs) mapped onto developmental cell lineages during Caenorhabditis elegans embryogenesis, at single-cell resolution. This atlas reveals a spatiotemporal combinatorial code of TF expression, and a cascade of lineage-specific, tissue-specific and time-specific TFs that specify developmental states. The atlas uncovers regulators of embryogenesis, including an unexpected role of a skin specifier in neurogenesis and the critical function of an uncharacterized TF in convergent muscle differentiation. At the systems level, the atlas provides an opportunity to model cell state-fate relationships, revealing a lineage-dependent state diversity within functionally related cells and a winding trajectory of developmental state progression. Collectively, this single-cell protein atlas represents a valuable resource for elucidating metazoan embryogenesis at the molecular and systems levels.

Two isoforms of the essential C. elegans Argonaute CSR-1 differentially regulate sperm and oocyte fertility.

The Caenorhabditis elegans genome encodes nineteen functional Argonaute proteins that use 22G-RNAs, 26G-RNAs, miRNAs or piRNAs to regulate target transcripts. Only one Argonaute is essential under normal laboratory conditions: CSR-1. While CSR-1 has been studied widely, nearly all studies have overlooked the fact that the csr-1 locus encodes two isoforms. These isoforms differ by an additional 163 amino acids present in the N-terminus of CSR-1a. Using CRISPR-Cas9 genome editing to introduce GFP::3xFLAG into the long (CSR-1a) and short (CSR-1b) isoforms, we found that CSR-1a is expressed during spermatogenesis and in several somatic tissues, including the intestine. CSR-1b is expressed constitutively in the germline. small RNA sequencing of CSR-1 complexes shows that they interact with partly overlapping sets of 22G-RNAs. Phenotypic analyses reveal that the essential functions of csr-1 described in the literature coincide with CSR-1b, while CSR-1a plays tissue specific functions. During spermatogenesis, CSR-1a integrates into an sRNA regulatory network including ALG-3, ALG-4 and WAGO-10 that is necessary for fertility at 25°C. In the intestine, CSR-1a silences immunity and pathogen-responsive genes, and its loss results in improved survival from the pathogen Pseudomonas aeruginosa. Our findings functionally distinguish the CSR-1 isoforms and highlight the importance of studying each AGO isoform independently.

The Prop1-like homeobox gene unc-42 specifies the identity of synaptically connected neurons.

Many neuronal identity regulators are expressed in distinct populations of cells in the nervous system, but their function is often analyzed only in specific isolated cellular contexts, thereby potentially leaving overarching themes in gene function undiscovered. We show here that the Caenorhabditis elegans Prop1-like homeobox gene unc-42 is expressed in 15 distinct sensory, inter- and motor neuron classes throughout the entire C. elegans nervous system. Strikingly, all 15 neuron classes expressing unc-42 are synaptically interconnected, prompting us to investigate whether unc-42 controls the functional properties of this circuit and perhaps also the assembly of these neurons into functional circuitry. We found that unc-42 defines the routes of communication between these interconnected neurons by controlling the expression of neurotransmitter pathway genes, neurotransmitter receptors, neuropeptides, and neuropeptide receptors. Anatomical analysis of unc-42 mutant animals reveals defects in axon pathfinding and synaptic connectivity, paralleled by expression defects of molecules involved in axon pathfinding, cell-cell recognition, and synaptic connectivity. We conclude that unc-42 establishes functional circuitry by acting as a terminal selector of functionally connected neuron types. We identify a number of additional transcription factors that are also expressed in synaptically connected neurons and propose that terminal selectors may also function as 'circuit organizer transcription factors' to control the assembly of functional circuitry throughout the nervous system. We hypothesize that such organizational properties of transcription factors may be reflective of not only ontogenetic, but perhaps also phylogenetic trajectories of neuronal circuit establishment.

Parallel Rap1>RalGEF>Ral and Ras signals sculpt the C. elegans nervous system.

Ras is the most commonly mutated oncogene in humans and uses three oncogenic effectors: Raf, PI3K, and RalGEF activation of Ral. Understanding the importance of RalGEF>Ral signaling in cancer is hampered by the paucity of knowledge about their function in animal development, particularly in cell movements. We found that mutations that disrupt function of RalGEF or Ral enhance migration phenotypes of mutants for genes with established roles in cell migration. We used as a model the migration of the canal associated neurons (CANs), and validated our results in HSN cell migration, neurite guidance, and general animal locomotion. These functions of RalGEF and Ral are specific to their control of Ral signaling output rather than other published functions of these proteins. In this capacity Ral functions cell autonomously as a permissive developmental signal. In contrast, we observed Ras, the canonical activator of RalGEF>Ral signaling in cancer, to function as an instructive signal. Furthermore, we unexpectedly identified a function for the close Ras relative, Rap1, consistent with activation of RalGEF>Ral. These studies define functions of RalGEF>Ral, Rap1 and Ras signaling in morphogenetic processes that fashion the nervous system. We have also defined a model for studying how small GTPases partner with downstream effectors. Taken together, this analysis defines novel molecules and relationships in signaling networks that control cell movements during development of the nervous system.