Neuronal diversity and convergence in a visual system developmental atlas.

Deciphering how neuronal diversity is established and maintained requires a detailed knowledge of neuronal gene expression throughout development. In contrast to mammalian brains1,2, the large neuronal diversity of the Drosophila optic lobe3 and its connectome4-6 are almost completely characterized. However, a molecular characterization of this neuronal diversity, particularly during development, has been lacking. Here we present insights into brain development through a nearly complete description of the transcriptomic diversity of the optic lobes of Drosophila. We acquired the transcriptome of 275,000 single cells at adult and at five pupal stages, and built a machine-learning framework to assign them to almost 200 cell types at all time points during development. We discovered two large neuronal populations that wrap neuropils during development but die just before adulthood, as well as neuronal subtypes that partition dorsal and ventral visual circuits by differential Wnt signalling throughout development. Moreover, we show that the transcriptomes of neurons that are of the same type but are produced days apart become synchronized shortly after their production. During synaptogenesis we also resolved neuronal subtypes that, although differing greatly in morphology and connectivity, converge to indistinguishable transcriptomic profiles in adults. Our datasets almost completely account for the known neuronal diversity of the Drosophila optic lobes, and serve as a paradigm to understand brain development across species.

Integration of temporal and spatial patterning generates neural diversity.

In the Drosophila optic lobes, 800 retinotopically organized columns in the medulla act as functional units for processing visual information. The medulla contains over 80 types of neuron, which belong to two classes: uni-columnar neurons have a stoichiometry of one per column, while multi-columnar neurons contact multiple columns. Here we show that combinatorial inputs from temporal and spatial axes generate this neuronal diversity: all neuroblasts switch fates over time to produce different neurons; the neuroepithelium that generates neuroblasts is also subdivided into six compartments by the expression of specific factors. Uni-columnar neurons are produced in all spatial compartments independently of spatial input; they innervate the neuropil where they are generated. Multi-columnar neurons are generated in smaller numbers in restricted compartments and require spatial input; the majority of their cell bodies subsequently move to cover the entire medulla. The selective integration of spatial inputs by a fixed temporal neuroblast cascade thus acts as a powerful mechanism for generating neural diversity, regulating stoichiometry and the formation of retinotopy.

Knock-in mutations of scarecrow, a Drosophila homolog of mammalian Nkx2.1, reveal a novel function required for development of the optic lobe in Drosophila melanogaster.

scarecrow (scro) gene encodes a Drosophila homolog of mammalian Nkx2.1 that belongs to an evolutionally conserved NK2 family. Nkx2.1 has been well known for its role in the development of hypothalamus, lung, thyroid gland, and brain. However, little is known about biological roles of scro. To understand scro functions, we generated two types of knock-in mutant alleles, substituting part of either exon-2 or exon-3 for EGFP (or Gal4) by employing the CRISPR/Cas9 genome editing tool. Using these mutations, we characterized spatio-temporal expression patterns of the scro gene and its mutant phenotypes. Homozygous knock-in mutants are lethal during embryonic and early larval development. In developing embryos, scro is exclusively expressed in the pharyngeal primordia and numerous neural clusters in the central nervous system (CNS). In postembryonic stages, the most prominent scro expression is detected in the larval and adult optic lobes, suggesting that scro plays a role for the development and/or function of this tissue type. Notch signaling is the earliest factor known to act for the development of the optic lobe. scro mutants lacked mitotic cells and Delta expression in the optic anlagen, and showed altered expression of several proneural and neurogenic genes including Delta and Notch. Furthermore, scro mutants showed grossly deformed neuroepithelial (NE) cells in the developing optic lobe and severely malformed adult optic lobes, the phenotypes of which are shown in Notch or Delta mutants, suggesting scro acting epistatic to the Notch signaling. From these data together, we propose that scro plays an essential role for the development of the optic lobe, possibly acting as a regional specification factor.

Slit neuronal secretion coordinates optic lobe morphogenesis in Drosophila.

The complexity of the nervous system requires the coordination of multiple cellular processes during development. Among them, we find boundary formation, axon guidance, cell migration and cell segregation. Understanding how different cell populations such as glial cells, developing neurons and neural stem cells contribute to the formation of boundaries and morphogenesis in the nervous system is a critical question in neurobiology. Slit is an evolutionary conserved protein essential for the development of the nervous system. For signaling, Slit has to bind to its cognate receptor Robo, a single-pass transmembrane protein. Although the Slit/Robo signaling pathway is well known for its involvement in axon guidance, it has also been associated to boundary formation in the Drosophila visual system. In the optic lobe, Slit is expressed in glial cells, positioned at the boundaries between developing neuropils, and in neurons of the medulla ganglia. Although it has been assumed that glial cells provide Slit to the system, the contribution of the neuronal expression has not been tested. Here, we show that, contrary to what was previously thought, Slit protein provided by medulla neurons is also required for boundary formation and morphogenesis of the optic lobe. Furthermore, tissue specific rescue using modified versions of Slit demonstrates that this protein acts at long range and does not require processing by extracellular proteases. Our data shed new light on our understanding of the cellular mechanisms involved in Slit function in the fly visual system morphogenesis.

Patterning mechanisms diversify neuroepithelial domains in the Drosophila optic placode.

The central nervous system develops from monolayered neuroepithelial sheets. In a first step patterning mechanisms subdivide the seemingly uniform epithelia into domains allowing an increase of neuronal diversity in a tightly controlled spatial and temporal manner. In Drosophila, neuroepithelial patterning of the embryonic optic placode gives rise to the larval eye primordium, consisting of two photoreceptor (PR) precursor types (primary and secondary), as well as the optic lobe primordium, which during larval and pupal stages develops into the prominent optic ganglia. Here, we characterize a genetic network that regulates the balance between larval eye and optic lobe precursors, as well as between primary and secondary PR precursors. In a first step the proneural factor Atonal (Ato) specifies larval eye precursors, while the orphan nuclear receptor Tailless (Tll) is crucial for the specification of optic lobe precursors. The Hedgehog and Notch signaling pathways act upstream of Ato and Tll to coordinate neural precursor specification in a timely manner. The correct spatial placement of the boundary between Ato and Tll in turn is required to control the precise number of primary and secondary PR precursors. In a second step, Notch signaling also controls a binary cell fate decision, thus, acts at the top of a cascade of transcription factor interactions to define PR subtype identity. Our model serves as an example of how combinatorial action of cell extrinsic and cell intrinsic factors control neural tissue patterning.

A Serrate-Notch-Canoe complex mediates essential interactions between glia and neuroepithelial cells during Drosophila optic lobe development.

It is firmly established that interactions between neurons and glia are fundamental across species for the correct establishment of a functional brain. Here, we found that the glia of the Drosophila larval brain display an essential non-autonomous role during the development of the optic lobe. The optic lobe develops from neuroepithelial cells that proliferate by dividing symmetrically until they switch to asymmetric/differentiative divisions that generate neuroblasts. The proneural gene lethal of scute (l'sc) is transiently activated by the epidermal growth factor receptor (EGFR)-Ras signal transduction pathway at the leading edge of a proneural wave that sweeps from medial to lateral neuroepithelium, promoting this switch. This process is tightly regulated by the tissue-autonomous function within the neuroepithelium of multiple signaling pathways, including EGFR-Ras and Notch. This study shows that the Notch ligand Serrate (Ser) is expressed in the glia and it forms a complex in vivo with Notch and Canoe, which colocalize at the adherens junctions of neuroepithelial cells. This complex is crucial for interactions between glia and neuroepithelial cells during optic lobe development. Ser is tissue-autonomously required in the glia where it activates Notch to regulate its proliferation, and non-autonomously in the neuroepithelium where Ser induces Notch signaling to avoid the premature activation of the EGFR-Ras pathway and hence of L'sc. Interestingly, different Notch activity reporters showed very different expression patterns in the glia and in the neuroepithelium, suggesting the existence of tissue-specific factors that promote the expression of particular Notch target genes or/and a reporter response dependent on different thresholds of Notch signaling.

Replication protein a links cell cycle progression and the onset of neurogenesis in Drosophila optic lobe development.

Stem cell self-renewal and differentiation must be carefully controlled during development and tissue homeostasis. In the Drosophila optic lobe, neuroepithelial cells first divide symmetrically to expand the stem cell population and then transform into asymmetrically dividing neuroblasts, which generate medulla neurons. The mechanisms underlying this cell fate transition are not well understood. Here, we show a crucial role of some cell cycle regulators in this transition. We find that loss of function in replication protein A (RPA), which consists of three highly conserved protein subunits and functions in DNA replication, leads to disintegration of the optic lobe neuroepithelium and premature differentiation of neuroepithelial cells into medulla neuroblasts. Clonal analyses of RPA loss-of-function alleles indicate that RPA is required to prevent neuroepithelial cells from differentiating into medulla neuroblasts. Inactivation of the core cell cycle regulators, including the G1/S regulators E2F1, Cyclin E, Cdk2, and PCNA, and the G2/M regulators Cyclin A, Cyclin B, and Cdk1, mimic RPA loss-of-function phenotypes, suggesting that cell cycle progression is required for both maintaining neuroepithelial cell identity and suppressing neuroblast formation. We further find that RPA or E2F1 inactivation in the neuroepithelial cells correlates with downregulation of Notch signaling activity, which appears to result from Numb mislocalization. Thus, we have shown that the transition from neuroepithelial cells to neuroblasts is directly regulated by cell cycle regulators and propose a model in which the inhibition of neuroepithelial cell cycle progression downregulates Notch signaling activity through Numb, which leads to the onset of neurogenesis.

A multipotent transit-amplifying neuroblast lineage in the central brain gives rise to optic lobe glial cells in Drosophila.

The neurons and glial cells of the Drosophila brain are generated by neural stem cell-like progenitors during two developmental phases, one short embryonic phase and one more prolonged postembryonic phase. Like the bulk of the adult-specific neurons, most of glial cells found in the adult central brain are generated postembryonically. Five of the neural stem cell-like progenitors that give rise to glial cells during postembryonic brain development have been identified as type II neuroglioblasts that generate neural and glial progeny through transient amplifying INPs. Here we identify DL1 as a novel multipotent neuroglial progenitor in the central brain and show that this type II neuroblast not only gives rise to neurons that innervate the central complex but also to glial cells that contribute exclusively to the optic lobe. Immediately following their generation in the central brain during the second half of larval development, these DL1 lineage-derived glia migrate into the developing optic lobe, where they differentiate into three identified types of optic lobe glial cells, inner chiasm glia, outer chiasm glia and cortex glia. Taken together, these findings reveal an unexpected central brain origin of optic lobe glial cells and central complex interneurons from one and the same type II neuroglioblast.

Waves of differentiation in the fly visual system.

Sequential progression of differentiation in a tissue or in multiple tissues in a synchronized manner plays important roles in development. Such waves of differentiation are especially important in the development of the Drosophila visual system, which is composed of the retina and the optic lobe of the brain. All of the components of the fly visual system are topographically connected, and each ommatidial unit in the retina corresponds to a columnar unit in the optic lobe, which is composed of lamina, medulla, lobula and lobula plate. In the developing retina, the wave of differentiation follows the morphogenetic furrow, which progresses in a posterior-to-anterior direction. At the same time, differentiation of the lamina progresses in the same direction, behind the lamina furrow. This is not just a coincidence: differentiated photoreceptor neurons in the retina sequentially send axons to the developing lamina and trigger differentiation of lamina neurons to ensure the progression of the lamina furrow just like the furrow in the retina. Similarly, development of the medulla accompanies a wave of differentiation called the proneural wave. Thus, the waves of differentiation play important roles in establishing topographic connections throughout the fly visual system. In this article, we review how neuronal differentiation and connectivity are orchestrated in the fly visual system by multiple waves of differentiation.

Cell migration in Drosophila optic lobe neurons is controlled by eyeless/Pax6.

In the developing Drosophila optic lobe, eyeless, apterous and distal-less, three genes that encode transcription factors with important functions during development, are expressed in broad subsets of medulla neurons. Medulla cortex cells follow two patterns of cell movements to acquire their final position: first, neurons are arranged in columns below each neuroblast. Then, during pupation, they migrate laterally, intermingling with each other to reach their retinotopic position in the adult optic lobe. eyeless, which encodes a Pax6 transcription factor, is expressed early in progenitors and controls aspects of this cell migration. Its loss in medulla neurons leads to overgrowth and a failure of lateral migration during pupation. These defects in cell migration among medulla cortex cells can be rescued by removing DE-Cadherin. Thus, eyeless links neurogenesis and neuronal migration.

A role for juvenile hormone in the prepupal development of Drosophila melanogaster.

To elucidate the role of juvenile hormone (JH) in metamorphosis of Drosophila melanogaster, the corpora allata cells, which produce JH, were killed using the cell death gene grim. These allatectomized (CAX) larvae were smaller at pupariation and died at head eversion. They showed premature ecdysone receptor B1 (EcR-B1) in the photoreceptors and in the optic lobe, downregulation of proliferation in the optic lobe, and separation of R7 from R8 in the medulla during the prepupal period. All of these effects of allatectomy were reversed by feeding third instar larvae on a diet containing the JH mimic (JHM) pyriproxifen or by application of JH III or JHM at the onset of wandering. Eye and optic lobe development in the Methoprene-tolerant (Met)-null mutant mimicked that of CAX prepupae, but the mutant formed viable adults, which had marked abnormalities in the organization of their optic lobe neuropils. Feeding Met(27) larvae on the JHM diet did not rescue the premature EcR-B1 expression or the downregulation of proliferation but did partially rescue the premature separation of R7, suggesting that other pathways besides Met might be involved in mediating the response to JH. Selective expression of Met RNAi in the photoreceptors caused their premature expression of EcR-B1 and the separation of R7 and R8, but driving Met RNAi in lamina neurons led only to the precocious appearance of EcR-B1 in the lamina. Thus, the lack of JH and its receptor Met causes a heterochronic shift in the development of the visual system that is likely to result from some cells 'misinterpreting' the ecdysteroid peaks that drive metamorphosis.

Looming detection by identified visual interneurons during larval development of the locust Locusta migratoria.

Insect larvae clearly react to visual stimuli, but the ability of any visual neuron in a newly hatched insect to respond selectively to particular stimuli has not been directly tested. We characterised a pair of neurons in locust larvae that have been extensively studied in adults, where they are known to respond selectively to objects approaching on a collision course: the lobula giant motion detector (LGMD) and its postsynaptic partner, the descending contralateral motion detector (DCMD). Our physiological recordings of DCMD axon spikes reveal that at the time of hatching, the neurons already respond selectively to objects approaching the locust and they discriminate between stimulus approach speeds with differences in spike frequency. For a particular approaching stimulus, both the number and peak frequency of spikes increase with instar. In contrast, the number of spikes in responses to receding stimuli decreases with instar, so performance in discriminating approaching from receding stimuli improves as the locust goes through successive moults. In all instars, visual movement over one part of the visual field suppresses a response to movement over another part. Electron microscopy demonstrates that the anatomical substrate for the selective response to approaching stimuli is present in all larval instars: small neuronal processes carrying information from the eye make synapses both onto LGMD dendrites and with each other, providing pathways for lateral inhibition that shape selectivity for approaching objects.

SAGA-mediated H2B deubiquitination controls the development of neuronal connectivity in the Drosophila visual system.

Nonstop, which has previously been shown to have homology to ubiquitin proteases, is required for proper termination of axons R1-R6 in the optic lobe of the developing Drosophila eye. Herein, we establish that Nonstop actually functions as an ubiquitin protease to control the levels of ubiquitinated histone H2B in flies. We further establish that Nonstop is the functional homolog of yeast Ubp8, and can substitute for Ubp8 function in yeast cells. In yeast, Ubp8 activity requires Sgf11. We show that in Drosophila, loss of Sgf11 function causes similar photoreceptor axon-targeting defects as loss of Nonstop. Ubp8 and Sgf11 are components of the yeast SAGA complex, suggesting that Nonstop function might be mediated through the Drosophila SAGA complex. Indeed, we find that Nonstop does associate with SAGA components in flies, and mutants in other SAGA subunits display nonstop phenotypes, indicating that SAGA complex is required for accurate axon guidance in the optic lobe. Candidate genes regulated by SAGA that may be required for correct axon targeting were identified by microarray analysis of gene expression in SAGA mutants.

Spatio-temporal pattern of programmed cell death in the developing Drosophila optic lobe.

A large number of cells die via programmed cell death during the normal development of the Drosophila optic lobe. In this study, we report the precise spatial and temporal pattern of cell death in this organ. Cell death in the developing optic lobe occurs in two distinct phases. The first phase extends from the start of metamorphosis to the mid-pupal stage. During this phase, a large number of cells die in the optic lobe as a whole, with a peak of cell death at an early pupal stage in the lamina and medulla cortices and the region of the T2/T3/C neurons, and a smaller number of dead cells observed in the lobula plate cortex. The second phase extends from the mid-pupal stage to eclosion. Throughout this period, a small number of dying cells can be observed, with a small peak at a late pupal stage. Most of the dying cells are neurons. During the first phase, dying cells are distributed in specific patterns in cortices. The lamina cortex contains two distinct clusters of dying cells; the medulla cortex, four clusters; the lobula plate cortex, one cluster; and the region of the T2/T3/C neurons, one cluster. Many of the clusters maintain their distinct positions in the optic lobe but others extend the region they cover during development. The presence of distinct clusters of dying cells at different phases suggests that distinct mechanisms control cell death during different stages of optic lobe development in Drosophila.

Coordinated control of neuronal differentiation and wiring by sustained transcription factors.

The large diversity of cell types in nervous systems presents a challenge in identifying the genetic mechanisms that encode it. Here, we report that nearly 200 distinct neurons in the Drosophila visual system can each be defined by unique combinations of on average 10 continuously expressed transcription factors. We show that targeted modifications of this terminal selector code induce predictable conversions of neuronal fates that appear morphologically and transcriptionally complete. Cis-regulatory analysis of open chromatin links one of these genes to an upstream patterning factor that specifies neuronal fates in stem cells. Experimentally validated network models describe the synergistic regulation of downstream effectors by terminal selectors and ecdysone signaling during brain wiring. Our results provide a generalizable framework of how specific fates are implemented in postmitotic neurons.

Brain development in the yellow fever mosquito Aedes aegypti: a comparative immunocytochemical analysis using cross-reacting antibodies from Drosophila melanogaster.

Considerable effort has been directed towards understanding the organization and function of peripheral and central nervous system of disease vector mosquitoes such as Aedes aegypti. To date, all of these investigations have been carried out on adults but none of the studies addressed the development of the nervous system during the larval and pupal stages in mosquitoes. Here, we first screen a set of 30 antibodies, which have been used to study brain development in Drosophila, and identify 13 of them cross-reacting and labeling epitopes in the developing brain of Aedes. We then use the identified antibodies in immunolabeling studies to characterize general neuroanatomical features of the developing brain and compare them with the well-studied model system, Drosophila melanogaster, in larval, pupal, and adult stages. Furthermore, we use immunolabeling to document the development of specific components of the Aedes brain, namely the optic lobes, the subesophageal neuropil, and serotonergic system of the subesophageal neuropil in more detail. Our study reveals prominent differences in the developing brain in the larval stage as compared to the pupal (and adult) stage of Aedes. The results also uncover interesting similarities and marked differences in brain development of Aedes as compared to Drosophila. Taken together, this investigation forms the basis for future cellular and molecular investigations of brain development in this important disease vector.

Controlling nuclear JAKs and STATs for specific gene activation by IFNγ.

We previously showed that gamma interferon (IFNγ) and its receptor subunit, IFNGR1, interacted with the promoter region of IFNγ-activated genes along with transcription factor STAT1α. Recent studies have suggested that activated Janus kinases pJAK2 and pJAK1 also played a role in gene activation by phosphorylation of histone H3 on tyrosine 41. This study addresses the question of the role of activated JAKs in specific gene activation by IFNγ. We carried out chromatin immunoprecipitation (ChIP) followed by PCR in IFNγ treated WISH cells and showed association of pJAK1, pJAK2, IFNGR1, and STAT1 on the same DNA sequence of the IRF-1 gene promoter. The ß-actin gene, which is not activated by IFNγ, did not show this association. The movement of activated JAK to the nucleus and the IRF-1 promoter was confirmed by the combination of nuclear fractionation, confocal microscopy and DNA precipitation analysis using the biotinylated GAS promoter. Activated JAKs in the nucleus was associated with phosphorylated tyrosine 41 on histone H3 in the region of the GAS promoter. Unphosphorylated JAK2 was found to be constitutively present in the nucleus and was capable of undergoing activation in IFNγ treated cells, most likely via nuclear IFNGR1. Association of pJAK2 and IFNGR1 with histone H3 in IFNγ treated cells was demonstrated by histone H3 immunoprecipitation. Unphosphorylated STAT1 protein was associated with histone H3 of untreated cells. IFNγ treatment resulted in its disassociation and then re-association as pSTAT1. The results suggest a novel role for activated JAKs in epigenetic events for specific gene activation.

Role of JAK/STAT signaling in neuroepithelial stem cell maintenance and proliferation in the Drosophila optic lobe.

During Drosophila optic lobe development, proliferation and differentiation must be tightly modulated to reach its normal size for proper functioning. The JAK/STAT pathway plays pleiotropic roles in Drosophila development and in the larval brain, has been shown to inhibit medulla neuroblast formation. In this study, we find that JAK/STAT activity is required for the maintenance and proliferation of the neuroepithelial stem cells in the optic lobe. In loss-of-function JAK/STAT mutant brains, the neuroepithelial cells lose epithelial cell characters and differentiate prematurely while ectopic activation of this pathway is sufficient to induce neuroepithelial overgrowth in the optic lobe. We further show that Notch signaling acts downstream of JAK/STAT to control the maintenance and growth of the optic lobe neuroepithelium. Thus, in addition to its role in suppression of neuroblast formation, the JAK/STAT pathway is necessary and sufficient for optic lobe neuroepithelial growth.

Brain plasticity over the metamorphic boundary: carry-over effect of larval environment on froglet brain development.

Brain development shows high plasticity in response to environmental heterogeneity. However, it is unknown how environmental variation during development may affect brain architecture across life history switch points in species with complex life cycles. Previously, we showed that predation and competition affect brain development in common frog (Rana temporaria) tadpoles. Here, we studied whether larval environment had carry-over effects in brains of metamorphs. Tadpoles grown at high density had large optic tecta at metamorphosis, whereas tadpoles grown under predation risk had small diencephala. We found that larval density had a carry-over effect on froglet optic tectum size, whereas the effect of larval predation risk had vanished by metamorphosis. We discuss the possibility that the observed changes may be adaptive, reflecting the needs of an organism in given environmental and developmental contexts.

The conserved Ig superfamily member Turtle mediates axonal tiling in Drosophila.

Restriction of adjacent same-type axons/dendrites to separate single columns for specific neuronal connections is commonly observed in vertebrates and invertebrates, and is necessary for proper processing of sensory information. Columnar restriction is conceptually similar to tiling, a phenomenon referring to the avoidance of neurites from adjacent same-type neurons. The molecular mechanism underlying the establishment of columnar restriction or axonal/dendritic tiling remains largely undefined. Here, we identify Turtle (Tutl), a member of the conserved Tutl/Dasm1/IgSF9 subfamily of the Ig superfamily, as a key player in regulating the tiling pattern of R7 photoreceptor terminals in Drosophila. Tutl functions to prevent fusion between two adjacent R7 terminals, and acts in parallel to the Activin pathway. Tutl mediates homophilic cell-cell interactions. We propose that extrinsic terminal-terminal recognition mediated by Tutl, acts in concert with intrinsic Activin-dependent control of terminal growth, to restrict the connection made by each R7 axon to a single column.