|
|
|
|
||||||
|
|
|
|||||
|
|
||||||
|
Welcome to the homepage of the NYU Molecular Genetics Lab of Claude Desplan. Search PubMed for Lab Publications (Enter: "Desplan C")
RESEARCH PROGRAMS
Our laboratory focuses on two major questions of development, retinal specification for color vision and the evolution of early embryonic development.. These systems represent paradigms for understanding how pattern formation is genetically controlled. Recently, we have additionally embarked on a functional analysis of the neural network that supports color vision in the optic lobe of Drosophila.
Detection and processing of color information:
The majority of the lab is focused on understanding the establishment of the retinal mosaic that supports color vision in Drosophila. Color vision is achieved through the comparison in the brain of inputs coming from photoreceptor cells containing Rhodopsin photopigments with different wavelength specificity. These photopigments are expressed in mutually exclusive patterns. The fly compound eye is composed of 800 ommatidia, each containing 8 photoreceptor cells which form widely expanded membrane structures called rhabdomeres. The six outer photoreceptors (R1-R6) express the broad spectrum Rhodopsin1 and mediate motion detection. The rhabdomeres of the two inner photoreceptors, R7 and R8, are positioned above each other and thus share the same optic path. Each R7 and R8 expresses one of four rhodopsins, defining three subtypes of ommatidia: ‘yellow’ (y), ‘pale’ (p), and the ‘dorsal rim area’ (DRA). While DRA ommatidia function as polarized light detectors, color vision depends on the y and p ommatidial subtypes that are randomly distributed throughout the main part of the retina. In the p subtype, R7 expresses the UV-sensitive Rh3 and R8 the blue-Rh5. In the y subtype, R7 expresses a distinct UV-Rh4 while R8 expresses the green- Rh6. Therefore, there is coordination in the type of rhodopsins expressed in R7 and R8. While the p subtype is better suited to discriminate among shorter wavelengths, the y subtype discriminates among longer wavelengths.
Differentiation of R7 and R8 color Photoreceptors: Photoreceptor development starts in the eye imaginal disc when cells posterior to the morphogenetic furrow differentiate and eight photoreceptors are recruited to each ommatidium; immediately thereafter the photoreceptor axons project to their respective optic lobes. Later, during pupation, photoreceptors undergo terminal differentiation, which includes formation of rhabdomere and initiation of rhodopsin expression. Although each photoreceptor is specified differently during development, the adult eye contains only two types of photoreceptors: outer and inner photoreceptors. We have shown that all photoreceptors tend to differentiate toward an outer photoreceptor ground state. However, expression of the spalt gene complex in R7 and R8 redirects these cells toward an ‘inner photoreceptor’ fate (Mollereau et al. Nature 2001). Presumably spalt is turned on by two independent pathways in R7 and R8 where it directs the inner photoreceptor fate. In the absence of spalt, all photoreceptors develop as outer, while forced expression of spalt turns all photoreceptors into inner-like photoreceptors. In both cases, however, the projections to the lamina or medulla are not grossly affected, reinforcing the notion that this second phase of photoreceptors differentiation is distinct from the earlier neuronal phase. Once they are specified as ‘generic’ inner photoreceptors, R7 and R8 are then further specified. prospero is specifically expressed in R7 where it represses the R8 fate by directly repressing expression of R8 rhodopsins while inducing R7 rhodopsin expression (Cook et al. Dev. Cell 2003). Another transcription factor, Senseless, expressed in R8, plays the symmetrical role and specifies the R8 fate.
Polarized light vision: the compass of the fly (Mathias Wernet, now in the Clandinin lab at Stanford): Polarization-sensitive ommatia are located in the DRA. They serve as detectors for the oscillation plane of polarized light, forming the basis for a compass that has been demonstrated in many insect species. In the DRA, R7 and R8 have uniquely adapted their configuration and morphology for polarized light detection. Both R7 and R8 express UV-Rh3: the polarization compass is thus color-blind, eliminating any ambiguity between information about color and polarization. DRA R7 and R8 project to the medulla and are believed to interact antagonistically via polarization-sensitive interneurons to enhance polarization sensitivity. We have shown that homothorax is expressed specifically in both R7 and R8 of the DRA. homothorax is both necessary and sufficient to induce DRA ommatidia. Over-expression of homothorax in all developing photoreceptors results in the expansion of Rh3 to all inner photoreceptors with loss of all other Rhodopsins. Reciprocally, loss of homothorax function results in the transformation of polarized sensitive ommatidia into color vision ommatidia. The highly localized expression of hth results from regulation by the dorsal selector genes of the IroC and from Wingless signaling emanating from the dorsal cuticle. The student who performed most of this work is now continuing his investigations of mechanisms for polarized light detection with T. Clandinin at Stanford in collaboration with T. Labhart in Zurich (Wernet et al., Cell 2003).
Control of the stochastic choice between two ommatidial fates: spineless (Mathias Wernet, Esteban Mazzoni & Arzu Celik) Stochastic events play an important role in many biological processes. The p and y ommatidial subtypes are distributed stochastically in the retina, similarly to human cone photoreceptor distribution. R7 cells first make the p vs. y choice and then impose fate onto R8. We have recently found that a transcription factor, the Dioxin receptor encoded by spineless, plays a critical role in this process. spineless is expressed in a short burst during mid-pupation, preceding rhodopsin expression by more than one day, in a large subset of R7 cells. Loss of spineless function results in all R7 and most R8 adopting the p fate, whereas over-expression of spineless induces y-R7 fate. The stochastic expression of spineless in R7 cells is therefore necessary and sufficient to induce y-R7 cell fate and thus, control the entire mosaic of ommatidia in the fly retina (Wernet et al., Nature 2006). The mechanisms controlling cell autonomous stochastic spineless expression remain a mystery. A first inkling into the mechanism controlling spineless expression comes from the observation that introduction of extra copies of the spineless promoter yields a reduction in the percentage of y fate ommatidia, presumably resulting from a reduction in the level of spineless expression. This observation is consistent with the existence of a specific upstream limiting factor that binds the spineless promoter and controls its expression. We are currently analyzing how this mechanism can be generalized to other examples of stochastic gene expression (Preet Lidder & Bob Johnston).
A bistable loop in R8 reinforces R8 fates induced by R7 to R8 Activin signaling. (Tamara Dvali, Mathias Wernet, Daniela Pistilo, David Jukam): The choice between the p and y fate is first made in R7 by the stochastic expression of ss. Once an R7 commits to the p fate and expresses rh3, it sends an instructive signal to the underlying R8, which then also commits to the p fate and expresses rh5. In the absence of the R7 signal, R8 commits to the y fate and expresses rh6. The signal from R7 to R8 appears to belong to the Activin pathway: both loss- and gain-of-function of the Activin receptor Baboon lead to disruption of the communication between R7 and R8. We have recently identified four genes required in R8 cells for ensuring the correct choice of y vs. p cell fate. The warts gene, which encodes a Ser/Thr kinase tumor suppressor gene (Lats), is necessary and sufficient for R8 to adopt the y fate. Another gene, melted plays the opposite role and specifically induces the p fate in R8. warts and melted are expressed in a complementary manner in the yR8 and pR8 subsets, respectively. The two genes repress each other’s transcription to form a bistable loop: melted responds to the R7 signal while warts regulates the output of the loop. Furthermore, the tumor suppressor partners of Warts/Lats, Hippo and Salvador, have phenotypes identical to warts. Interestingly, melted has recently been reported to regulate growth and fat metabolism in Drosophila. Thus, somewhat surprisingly, genes known to regulate both cell growth (melted) and proliferation (warts) interact antagonistically during retinal patterning. They represent the re-utilization for a postmitotic event of a cassette of gene regulation that is normally used for tissue size control (Mikeladze-Dvali et al., Cell, 2005).
Maintenance of mutual repression of rhodopsin genes (Daniel Vasiliauskas): A decision made in R7 to express Rh4 commits the entire ommatidia to the y fate. The remaining R7 take on the default p state and send an instructive Activin-like signal from R7 to R8 to indicate its choice. The Wats/melted bistable loop ensures that the decision in R8 is made without ambiguity. However, expression of these genes is only maintained until shortly after eclosion. This leads to a question of how Rhodopsin expression is maintained throughout adulthood thereby preserving the fidelity of the neural network for color discrimination. We have shown that, at least in some cases, expression of Rhodopsin proteins themselves is required for this maintenance. This is reminiscent of the mechanisms employed by the olfactory system for the receptor choice. For instance, in two-week old flies carrying a rh6 mutation, Rh5 expands to all R8 suggesting that Rh6 normally acts to exclude rh5 from yR8. We have also evidence that Rh5 and Rh3 act in mutually antagonistic ways to repress each other’s transcription when mutant situations would lead to their co-expression in R7 photoreceptors. This appears to be a general feature of sensory receptors and we are pursuing the molecular mechanisms underlying this function.
The neural network in the medulla for color vision (Javier Morante & Daniel Vasiliauskas): The Drosophila optic lobe is formed by approximately 60,000 cells and can be divided in four neuropils: lamina, medulla, lobula and lobula plate. The lamina is the first optic lobe neuropil and receives innervation from R1-R6. R7-R8 projections target two distinct medulla layers, with R7 projecting deeper than R8. The medulla is formed by ~40,000 cells whose cell bodies are located in the medulla cortex. Associated with each set of R7/R8 projections, there are ~800 ‘columns’, defined as fixed cassettes of cells that repetitively contact every R7-8 fascicule. Columns likely represent the functional units in the medulla. The main objective of this part of our research is to understand how medulla cells process color information coming from R7 and R8. In order to manipulate the network required for color vision, it is necessary to understand its organization. This requires a precise description of the anatomy of the system and of the logic behind its development. We hypothesize that, in each column, several medulla neuron types contact one R7 and one R8 of the same ommatidium and compare their output (vertical integration, e.g. UV vs. blue in p; UV vs. green in y ommatidia). Other medulla cells may compare the information between p and y ommatidia (horizontal integration, e.g. blue vs. green). Additional comparisons might also occur: Interneurons might modulate the activity of photoreceptors and their target neurons while other neurons might integrate retinotopic and motion detection information. We have identified a series of marker Gal4 lines (most of them inserted in genes encoding transcription factors) that are expressed in subsets of medulla neurons. Using these lines and MARCM analysis, we have reconstructed the morphology of most medulla neurons. Based on this morphology, we have classified them in 15 different cell types, which roughly correspond to classes identified by Cajal and Fischbach using Golgi impregnation technique. Among these cells are short local neurons and long-projecting neurons that project deep in the optic lobes after contacting R7 and R8. Six types are ‘columnar neurons’ that contact only one R7/8 termination pair while the remaining seven neuronal types are ‘non-columnar’ as they contact more than one R7/8 and are likely not present in every column. Each R7/R8 termination pair is therefore surrounded by at least 13 different cell types. In addition, there are also at least two other cell types that do not contact photoreceptors: one short type of projecting neuron connecting pre-synaptically to the serpentine layer and one long type of projecting neuron connecting post-synaptically to the same layer. Although we do not yet have lines for each neuronal class, we are screening a large set of new Gal4 lines to be able to modulate activity in each class of neurons.
Connectivity and directionality of the neural network for color vision (Javier Morante & Alberto delValle): The complexity in number of cells and synapses in the medulla has made it extremely difficult to study its ultramicrostructure using EM. Axonal and dendritic compartments are also poorly separated in Drosophila, whose neurons are for the most part monopolar. To establish connectivity in the medulla, i.e. to establish which pre-synaptic termination contacts which post-synaptic arborization, we have co-expressed in single-cell clones transgenes that mark either axonal or dendritic compartments. This allows us to understand the directionality of information between photoreceptors and medulla cells, and among medulla cells by determining which terminations are pre- or postsynaptic as well as the types of connectivity established between cells. We have also investigated which neurotransmitters, excitatory or inhibitory, are expressed in the various medulla cells. We have also started to investigate the lineage of medulla cells in order to define the logic of optic lobe development: Is it based on a columnar organization whereby one neuroblast gives rise to distinct neurons organized around the terminations of one ommatidium; or are neurons generated as a pool with regulatory interactions leading to the specification of the correct number of neuronal subtypes?
Behavior assay for color vision: an operant paradigm (Satoko Yamaguchi, in collaboration with Reinhard Wolf in Würzburg): Our goal is to correlate anatomy and functionality with color vision in Drosophila. For this purpose, we developed a test for color discrimination. We first used a test based on innate color preference: Given a choice, flies move toward light of shorter wavelength if the same light intensity is used. We could show that the various classes of retinal photoreceptors mediate the expected attraction to their respective colors. However, in order to assay for true color vision, i.e. color discrimination, we must use an assay based on associative learning. Unfortunately, Pavlovian mass assay which work for studying olfactory behavior, failed so far to teach flies to associate color with punishments. We therefore setup a collaboration with Reinhart Wolf in Martin Heisenberg's lab in Würzburg to develop an operant paradigm. The ability of Drosophila to discriminate among widely separate wavelengths has been demonstrated using the flight simulator developed for Drosophila in the Heisenberg lab. We have been working for the last year in Würzburg and have been developing an operant paradigm to test for color discrimination. Although not appropriate for mutagenesis screens, this assay might allow us to test flies in which specific subsets of medulla neurons have been silenced and will be a powerful tool to understand the function of the medulla neural network.
Evolution of axis formation Also: Ava Brent & Miriam Rosenberg The Evolution and Development (Evo-Devo) project represents a long standing interest that originally stemmed from our analysis of the earliest steps of Drosophila development, the formation of the antero-posterior axis. It has since expanded into an attempt to analyze how the exquisitely well-understood regulatory gene network that controls segmentation in Drosophila is adapted to different developmental conditions in other insects. We are now approaching our goal of describing in similar detail the genetic network that controls segmentation in the hymenopteran wasp Nasonia vitripennis, which diverged from flies over 200 millions years ago. In Drosophila, the morphogenetic gradient of the Bicoid protein is essential for anterior development: This maternal transcription factor differentially controls the expression of target genes that specify head, thorax and much of the abdomen. However, bicoid is unique to higher dipterans and our goal has been to understand which factors played the role of bicoid in other species. There are already many early developmental mutants in Nasonia and we have adapted parental RNAi to knock down genes involved in early development in Nasonia [Lynch & Desplan, 2006] .
Orthodenticle: the ancestral anterior gene is required at both poles: (Jeremy Lynch now with Siegfried Roth in Cologne) We have shown that, while flies have bicoid mRNA localized at the anterior of the egg, Nasonia utilizes localized orthodenticle as an anterior morphogenetic center. Orthodenticle (otd1) plays an anterior patterning role in the beetle Tribolium. However, unlike the Bicoid gradient, the Tribolium Otd gradient forms through translational repression of otd mRNA by a posteriorly localized factor, i.e. from a posterior morphogenetic center. This reflects differences in modes of embryonic patterning. The long germ Drosophila embryo derives from the entire egg, and all segments are patterned at blastoderm stage, prior to gastrulation. In contrast, Tribolium undergoes short germ embryogenesis: The embryo arises from the posterior of the egg, and only anterior segments are patterned at blastoderm stage. The remaining segments arise post-gastrulation from a growth zone. In Nasonia, otd1 maternal mRNA is localized at both ends of the embryo, resulting in protein gradients that pattern both poles. Thus, localized Nasonia otd1 plays two major roles that allow long germ development. The anterior morphogenenetic gradient activates anterior targets in a manner reminiscent of the Bicoid gradient, thus leaving room at the posterior of the embryo for patterning from the posterior morphogenetic center. Otd1 is also required at the posterior to control posterior patterning by inducing gap genes prior to gastrulation. Short germ insects only induce these genes much later in the growth zone [Lynch et al, Nature, 2006].
Prevalent use of maternal mRNA localization to replace bicoid function in Nasonia: (Eugenia Olesnicky) Although in Nasonia otd1 can fulfill much of the function of bicoid in flies, some of these roles cannot be played by otd1. For instance, Bicoid prevents translation of the maternal posterior gene caudal in the anterior of the embryo by directly binding to the cad mRNA 3'UTR. Since Otd cannot bind to cad mRNA, another mechanism must be involved to limit cad function to the posterior of the embryo in Nasonia. Moreover, since otd1 acts at both poles of the Nasonia embryo, its function must be modified by other factors to direct development of anterior (with hunchback) vs. posterior (with caudal) structures. We first tested whether cad functions as a major ancestral player in early posterior embryonic patterning in Nasonia. We have shown that (1) Nvit cad plays a greater role in posterior patterning than In flies, and this role expands far more anteriorly in the wasp embryo than in fly (2) a Nvit cad gradient is achieved through maternal mRNA localization rather than through Bcd repression as in fly, and (3) Nvit cad is an activator of gap gene expression, in contrast to its role as a pair-rule gene activator in Drosophila [Olesnicky et al., 2006]. During Nasonia oogenesis, mRNA localization thus appears to be used extensively to replace the function of bicoid for the initiation of patterning along the antero-posterior axis. Nasonia localizes both caudal and nanos to the posterior pole, whereas giant mRNA is localized to the anterior pole of the oocyte; otd-1 is localized to both the anterior and posterior poles. The abundance of differentially localized mRNAs during Nasonia oogenesis gave us a unique opportunity to study the different mechanisms involved in mRNA localization. Through genetics and pharmacological disruption of the microtubule network, we have found that, while a microtubule-dependent mechanism is involved in the localization of anterior otd1 and gt mRNA, and of posterior cad, a microtubule-independent mechanism is required for the posterior localization of nanos and otd1 mRNAs. Finally, we find that actin is important in anchoring posteriorly localized otd1 and nos mRNAs and the oosome, a structure containing the pole plasm, to the posterior cortex of the oocyte.
Our investigations focus on dissecting the entire gene network that controls segmentation in Nasonia. We have reconstructed most of the interactions among gap genes and are moving to understand control of pair-rule genes, which represent the basic segmentation machinery in flies (even-skipped and hairy in particular). Of particular interest will be a complete comparison between the modes of stripe formation for eve and hairy. Through a detailed analysis of gene regulation, we will compare the molecular events that control posterior stripe formation in Drosophila and those of Nasonia for which an assembled genome sequence was released in July 2006.
Selected papers:
Sprecher, S.G., & Desplan C. Brent A.E., Yucel G., Small
S. & Desplan C. Sprecher S. G, Pichaud, F. & Desplan C. Adult and larval photoreceptors use different mechanisms to specify the same Rhodopsin fates Genes & Devel. 21, 2182-2195 (2007). [Medline Abstract] [Paper]
Lynch J.A., Brent A.E., Leaf D.S., Pultz M.A. & Desplan C.
Localized maternal
orthodenticle patterns anterior and posterior in the long germ wasp
Nasonia
Mikeladze-Dvali T., Wernet M., Pistillo D. , Mazzoni E. O., Teleman A., Chen Y., Cohen S. & Desplan C.
The
growth regulators Warts/lats and Melted interact in a bistable loop to
specify opposite fates in R8 photoreceptors.
Mazzoni
E, Desplan C. & Blau J.
Wernet M., Labhart T., Baumann F., Mazzoni E., Pichaud F. & Desplan C.
Tahayato A., Sonneville R., Pichaud F., Papatsenko D., Beaufils P., Wernet
M., Cook T. & Desplan C.
Cook
T.,
Pichaud F.,
Sonneville
R, Papatsenko
D. &
Desplan C.
|
||||||
|
|
| New York University | Dept. Biology | GSAS | GSAS Application | CAS | CAS Application | Contact Us | |
|||||
