Research Projects
Experimental and mathematical analysis of foxa cis-regulation in the sea urchin embryo
Smadar Ben-Tabou de-Leon, Eric H. Davidson
We combine mathematical modeling and experimental analysis to study the cis-regulation of the foxa gene. The foxa transcription factor is one of the key regulatory genes in the gene regulatory network governing endomesoderm development in the sea urchin embryo. Initially expressed in the endomesoderm progenitor field, foxa is necessary for endoderm specification and for exclusion of mesodermal fate in endoderm cells. As the segregation of the two domains occurs, foxa leaves the mesoderm territory and expresses only in the endoderm cells. FoxAprotein exerts negative feedback on its own expression and its levels oscillate with time. In order to understand foxa regulation we perform a cis-regulatory analysis of the foxa gene. Interspecies sequence comparison and reporter construct assays were used to identify a 2.5 kb sequence that drives the correct expression pattern up to 72 h after fertilization. The sequenceincludes three distinct modules located upstream of the transcription start site: a proximal module responds to ubiquitous activation, a middle module responds to endodermal activators and a distal module restricts the expression to the correct spatial pattern. We tested putative consensus sites of possible inputs by mutating the sites in GFP reporter constructs and monitoring the GFP expression. The distal module contains a TCF site that represses expression in the ectoderm and permits expression only in the endomesoderm territory. The middle module contains a Brachyury site that contributes positively to the endoderm expression of foxa. The three modules contain Otx sites that provide ubiquitous positive activation. The distal module contains five putative binding sites for FoxA itself; though these have not been tested yet, they might mediate the observed auto‑repression. We combined in a mathematical model the cis‑regulatory analysis results and the dynamic expression levels of the input genes measured by quantitative real-time PCR. The mathematical model consists of a set of differential equations for the mRNA and protein levels of the input genes and foxa. We included the positive inputs of Brachyury and Otx and also foxa auto-repression. We were able to simulate the experimentally observed time course of foxa expression up to 27 h after fertilization. Our study elucidates the mechanism by which a static genomic code controls a dynamic expression pattern
|
The Sea Urchin Genome Database - SpBase
R. Andrew Cameron, Qiu Autumn Yuan, Wick Gankanda, David Felt
The Sea Urchin Model Organism Database (http://spbase.org) was made publicly available on March 14, 2008. The web site and the database software that supports it have been in development during this past year. We have been tailoring our datasets to use the open source GMOD programs (http://gmod.org). These provide a consistent way to display, update and maintain the annotation data that emerged from the draft genome sequence done by BCM‑HGSC. A considerable amount of our effort was directed toward formatting the annotations from the draft genome project into a suitable structure for loading into a very complete data schema, dubbed Chado by the software developers. The basic system includes the database itself, a graphical sequence browser, a sequence search function and a gene search function to contain the information from manual annotation. We are adding features to this basic structure as rapidly as possible. A graphical sequence curation tool has been tied into the system as well. It is the familiar Apollo package developed at Berkeley. We have retained most of the additional information accumulated by BCM_HGSC during the curation process. Our last task before going public was to map all of the original features to the newest version of the genome assembly (version 2.1).
The Sea Urchin Genome Resource that provided a searchable sequence database (http://sugp.caltech.edu/) from which the user can obtain clone information and order clones is now obsolete. It will remain publicly available until we are assured that all of the information it contains has been successfully moved to SpBase.
|
The sea urchin repeat sequence complement
Susan Ernst, Emmanuelle Morin, Manoj Samanta, R. Andrew Cameron
Although repeat sequences make up 35% of the genome in sea urchins, they are difficult to analyze thoroughly in draft genome sequence assemblies. Almost invariably, gaps in sequence lie at repeat margins and unassembled repeat sequences cannot be placed or precisely counted. Nevertheless, we have collected a dataset of repeats from the version 2.1 assembly of the sea urchin. The basic strategy was to iterate through an all-by-all blast analysis followed by repeat clustering and genome sequence masking on ever increasing portions of the genome. Less simple sequence repeats, this collection yields 2500 consensus repeat sequences of which the most abundant was estimated to be present 17,000 times. To complement this approach, another method, k-mer analysis, is being used. We expect these will complement each other and give us a better estimate of the number and variety of the repeats. We have also begun to annotate the repeat families in the consensus set: Piggybac and Mariner DNA transposons are two distinctive families that are quite abundant in sea urchins. A measure of the divergence among the members of these two classes will allow an estimate of how recent these may have been active in the genome. Not unexpectedly, we are finding good representatives of both LTR and non-LTR retrotransposons and a rich complement of short interspersed nuclear elements or SINES is also detected.
|
Network gene annotation project
Qiu Autumn Yuan, R. Andrew Cameron
One of the strengths of the sea urchin embryo research model is the ease with which one can describe gene regulatory networks. As this experimental approach matures it is becoming clear that a gene annotation knowledge base is indispensible for designing experiments. Gene sequences, expression patterns and responses to perturbation are the central classes of information to be used here. Our previously built and currently maintained database of sequence and expression data for sea urchin regulatory genes fills that role. The data housed in this suite of databases continues to grow as additional genes, expression patterns, and network linkages are added to the existing information. The new ectoderm regulatory network studies from the Davidson laboratory have added a unique chunk of data to this effort. These unpublished data are available to a registered user group through a private Caltech web site. Our client base continues to grow as new users add more biological data. The latest addition is a function that graphically displays gene expression time-courses during embryogenesis.
|
The evolution of cis-regulatory module sequence in lower deuterostomes
R. Andrew Cameron, Eve Helguero, Qiu Autumn Yuan, Ping Dong, Julie Hahn
The echinoderm taxa make an unusually good set of species in which to examine the evolution of cis-regulatory modules (CRM). We have used genomic sequence comparisons covering a range of divergences from 20-540 million years (MY) in order to address a variety of questions about CRMs. Conserved sequence patches across 50 MY between Strongylocentrous purpuratus (Sp, our reference species) and Lytechinus variegatus (Lv) reveal candidate CRM sequences for gene regulatory network analysis. Interestingly, even though many of these conserved patches contain CRMs as shown by reporter analysis in the reference species, the exact position and order of the transcription factor binding sites are seldom conserved between the two species. However the number of sites is very similar if not identical. This observation suggests that simple sequence similarity will not distinguish CRMs.
Comparisons of functionally characterized cis‑regulatory modules from the Sp genome, and the orthologous regulatory and flanking sequences from the genome of a congener, S. franciscanus (Sf), cover a 20 MY divergence and reveal that single nucleotide substitutions and small indels occur freely at many positions within the regulatory modules of these two species, as they do without. However, large indels (>20 bp) are statistically almost absent within, though they are common in flanking intergenic or intronic sequence. We continue to expand the panel of comparisons to be made in this analysis. We now have available sequence from Allocentrotus fragilis (Af, also 20 MY diverged) provided by the Baylor College of Medicine Human Genome Sequencing Center to add to the analysis discussed above. The amount of genome coverage is expected to reach 2X this year and we expect more thorough four-species analyses to quickly emerge from these data.
A number of across species tests are being conducted to test the sequence analyses. We use recombinant BACs in which GFP or RFP is substituted for the gene-coding region while the surrounding genomic sequence is preserved. First, we are examining the significance of the small differences between Sp and Lv discussed above by injecting Sp constructs into Lv embryos. We are modifying the Sp sequences to resemble the Lv ones and test their transcriptional activity in Sp as well. From these tests we will be able to assess the significance of these small sequence differences that evolved in the CRMs.
|
Transcriptional control of the sea urchin Brachyury gene
Elly Chow, R. Andrew Cameron
Previously, we have identified a minimal enhancer for brachyury, a gene intermediate in the endomesoderm specification gene regulatory network. Up to the peak of embryonic expression at 24 hrs, it recapitulates, in time and space, the expression of the endogenous gene. This enhancer contains binding sites for Tcf, a transcription factor that is activated by the wnt pathway and Gatae, the output of the endomesoderm kernel subcircuit. An Otx binding site in the larger intronic sequence that contains the minimal enhancer also plays a role in activating this gene. To determine more precisely the expression pattern of our recombinant BAC GFP reporter and smaller constructs derived from it, we have built a red fluorescent protein (RFP) containing recombinant BAC. Co-injection of this control red marker with various reporter constructs precisely demonstrates the spatial extent of the constructs. Using our GFP-BAC and mutated blimp constructs developed by Joel Smith in the Davidson laboratory, we have demonstrated that the expression of brachyury is most likely kept out of the mesoderm by TCF-groucho repression. Our current hypothesis for brachyury early transcriptional control is: a wnt activity in veg2 endoderm precursors leads to the nuclearization of beta-catenin and the activation of TCF in those cells. Thus, brachyury transcription elevates beginning at about 10 hours pf. Later as Gatae and Otx begin to be expressed in these cells the expression is pushed to the maximum at 24 hours.
The next phase of brachyury expression is especially interesting. The zone of expression is stable in the blastopore lip even though the endoderm cells are moving through the zone of expression in order to form the archenteron. Thus, these cells must first turn on brachyury as they enter the zone and then turn it off as they exit. We will use our previous results as a base to explore the molecular basis of this dynamic expression pattern.
|
Re-engineering embryonic skeletogenesis at the gene-regulatory network level in the sea urchin, Strongylocentrotus purpuratus
Sagar Damle, Eric Davidson
Early development of skeletogenic and non-skeletogenic mesenchyme cells in S. purpuratus embryos are becoming well understood at level of expression and regulation of transcription factors and signaling molecules. However an ultimate demonstration of our understanding of these systems will be to rewire those regulatory relationships in novel ways to produce predictable developmental outcomes. We have used BAC recombinant constructs to reprogram development of skeletogenic precursors, known as primary mesenchyme cells (PMCs), at the gene regulatory level by driving the expression of the non-skeletogenic mesenchyme cell (NSM)-specific transcription factor, SpGcm, using the regulatory system controlling the PMC-specific T-box factor SpTbr. This BAC recombinant causes exogenous gcm expression in PMCs in a manner that precisely mirrors the early expression pattern of SpTbrain. PMCs expressing Gcm remain in the vegetal plate when non-expressing PMCs ingress and fail to participate in PMC‑syncitium, leading to a significant delay in skeletogenesis. Whole-mount in situ of blastula-stage embryos show PMCs expressing exogenous Gcm also show a loss of several transcription factors necessary for PMC-specification or skeletogenesis including SpAlx, Spfoxb and SpJun as well as a differentiation gene SpMsp130. At the same time these PMCs express markers of pigment cells, a subtype of NSMs. We have shown that expression of SpGcm in the context of another mesodermal cell type is sufficient to repress expression of genes that control that cell€™'s expected fate and redirect it towards a pigment cell fate.
Current work involves the cis-regulatory analysis of SpAlx, a factor expressed early during PMC specification. SpAlx is expressed at early blastula stage as a direct consequence of a double-repression gate involving the homeobox factor SpPmar1 and the hairy-enhancer-of-split ortholog, SpHes-C. Later in development Alx expression is maintained in part by Ets1. Hes-C binding sites have been identified in sequences conserved with the sea urchin Lytechinus variegatus and their function confirmed by mutational analysis with GFP reporters. Additionally, putative conserved Ets1 sites have been identified 3kb upstream of the basal promoter. Finally, it has been shown that Gcm overexpression is capable of repressing expression of a BAC-GFP reporter containing Alx regulatory DNA. At present, the binding sites mediating this repression are being investigated.
|
Evolution of developmental gene regulatory networks GRN)
Feng Gao
The developmental gene regulatory networks (GRNs) of sea urchin and sea star are providing almost the only direct, head-to-head comparative study so far available on gene regulatory network evolution. Comparative GRN analysis has revealed the diverse subcircuits of the orthologous GRNs of these animals evolve in dramatically different ways: some network subcircuits are astoundingly conserved in their linkages, which are termed the kernels of the GRN; others are changed, which are particularly evidenced as the same genes doing very different things (1, 2).
Our study is to exemplify and test these discoveries, and try to produce a causal explanation of specific evolutionary differences in gene regulatory network architecture between sea urchins and sea stars, which had a common ancestor half a billion years ago.
GRN mapping indicated the skeletogenic regulatory apparatus may have existed as a modular GRN unit since early in echinoderm history, and it was this same unit that was inserted into the sea urchin embryological developmental address defining the embryonic skeleton (3). There is presumptive evidence for a highly conserved GRN kernel. But this proposition could only be substantiated only by determining whether the actual cis-regulatory linkages are as pleisiomorphic as is the skeletogenic gene expression complex. This could be tested by reintroduction into sea urchin eggs the cis-regulatory modules of the sea star skeletogenic regulatory genes, together with computational genomics. If there is indeed a pan-echinoderm skeletogenic kernel, then the sea star regulatory modules thereof should display most of the same functional linkages and require most of the same inputs when introduced in reporter constructs into sea urchin eggs as are portrayed in the sea urchin GRN.
The GRN analysis predicts that the changes in GRN architecture are due to differently designed cis-regulatory modules of orthologous genes, which receive different inputs, and function differently to produce their specific modes of expression. This could be tested by isolation and direct functional comparison of the respective cis-regulatory modules of orthologous genes. Three candidate genes to be studied are Tbr, Delta and GataC, which are known to play different roles in the GRNs of sea star and sea urchin.
|
The subnetwork underlying specification of mesoderm in sea urchins
Stefan C. Materna, Eric Davidson
We have identified a multitude of transcription factors that are specifically expressed in the early mesoderm of the sea urchin. Currently, we are probing the regulatory interactions of these with other known mesodermal genes to better understand the regulatory network that regulates the specification of mesodermal cells.
Initially, nuclearization of beta-catenin – as part of a re-enforcing, regulatory circuit that involves the blimp1/krox and the wnt8 gene – activates endomesodermal genes. This subcircuit is first active in the cells of the skeletogenic lineage and then moves in a torus‑like motion across the vegetal half of the embryo. Downstream of this 'wave' are some transcription factors that have an important function in pigment cell formation, as the z13 gene. Other genes include the previously known hox11/13/b, ets1/2, and eve genes as well as the newly identified z48. Following the subdivision of endomesoderm into endodermal and mesodermal cells, a Delta signal is expressed by the skeletogenic cells and received by the mesodermal cells via the Notch receptor. Downstream of this signal are exclusive mesodermal genes such as glial cells missing. It is essential for the specification of pigment cells and activates markers, as the polyketide synthase gene that, are specific to pigment cells. At early gastrulation gcm expression recedes from the oral quadrant of the mesodermal territory; simultaneously genes like gata-c are activated specifically in these cells. The oral quadrant will produce primarily blastocoelar cells that are thought to fulfill immune functions. Interestingly, a number of bHLH transcription factors whose vertebrate orthologs are involved in specifying immune cells are expressed in the same cells, among them scl and the ubiquitously expressed e2a and id genes. Other newly identified genes include the orally expressed prox gene, and the aboral six1/2 gene. While it is not clear which signaling system is involved in the subdivision of the mesodermal cells, many of the mesodermal genes on the oral and aboral side are activated by Notch signaling. Current perturbation experiments are aimed at linking all newly identified genes to the beta-catenin/blimp/wnt8, and Delta/Notch systems as well as establishing the regulatory linkages among them.
|
New experimental approaches for system-wide cis-regulatory analyses
Jongmin Nam, Eric H. Davidson
Deciphering detailed features of trans- and cis-regulatory gene network codes is a key step in understanding how life forms develop and evolve. Building a system-wide trans-regulatory gene network is becoming a common procedure with the sea urchin S. purpuratus and some pioneering studies successfully demonstrated its power. However, system-wide cis-regulatory network analysis is still practically impossible in multicellular systems. Our goal in this project is to develop a set of system-wide and general experimental approaches for building cis-regulatory gene networks and to better relate cis- and trans-regulatory gene networks using the sea urchin as a model system. The proposed experimental approaches are as follows: Aim 1) Active cis-regulatory modules will be identified by a heuristic version of phylogenetic footprinting followed by simultaneous Q-PCR detection of 13 unique DNA Tags driven by 13 different candidate modules in one experiment. Aim 2) Spatial activity of each of the active modules will be examined by fluorescence-activated cell sorting followed by quantitative expression analysis of a large set of endogenous genes. By comparing the known spatial expression pattern of each gene to its enrichment or depletion in GFP positive cells it is possible to infer the spatial activity of a module in a quantitative manner. This method also gives an averaged regulatory state in cells where the module is active. Aim 3) Potential transcriptional inputs will be tested by simultaneously examining the effect of gene perturbation on a large set of active cis-regulatory modules (~100 modules) discovered from Aim 1 and on endogenous genes. Either Q-PCR or Nanostring will be used for the detection of expression. Predicted inputs will further be tested by mutating putative binding sites.
The method of Aim 1 has been successfully applied to a set of about 25 genes, and more genes will be analyzed by using this method in the near future. A proof of principle experiment for the method stated in Aim 2 was conducted with a partial success. For Aim 3 a set of over 100 unique DNA Tags have been developed and are being tested for ≤10 embryos with promising results. Once established, the set of quantitative and large-scale methods proposed here are easily applicable to many other systems including mammalian and plant systems. Therefore, these methods will have a significant impact on the mechanistic understanding of how life forms develop and evolve, and of abnormalities in disease conditions.
Publication
- Nam, J., Su., Y.H., Lee, P.Y., Robertson, A.J., Coffman, J.A. and Davidson, E.H. (2007) cis-Regulatory control of the nodal gene, initiator of the sea urchin oral ectoderm gene network. Dev. Biol. 306(2):860-869.
|
Regulatory logic of endoderm development in sea urchin embryos
Isabelle Peter, Yi Fan, Jina Yun, Eric H. Davidson
Embryonic endoderm development relies on the transition of a common endomesoderm regulatory state to the distinct endoderm regulatory states that specify different domains of the tripartite gut. The logical sequence in this process is determined by the regulatory wiring among endoderm transcription factors and signals acting on presumptive endoderm cells. We aim at solving the architecture of the complete endoderm gene regulatory network (GRN) in pre-gastrula stage sea urchin embryos. An earlier model of the endoderm GRN included 12 regulatory genes and genome-wide characterizations of transcription factor gene expression patterns have revealed an additional set of 11 candidate genes. We have further analyzed their spatial expression pattern, confirming the expression of seven of the candidate genes in endoderm precursor cells. To identify the regulatory linkages between all endoderm regulatory genes, we have perturbed the expression of each transcription factor by morpholino injection and detected the resulting changes in expression levels of all endoderm network genes.
Our results indicate that common endomesoderm transcription factors, regulated by nuclear b-catenin, control the expression of early endoderm transcription factors, which are at first expressed in a broad endomesoderm area but become restricted to endoderm precursor cells soon thereafter. Among these early endoderm factors are FoxA and Brachyury (Bra). Our analyses has identified several transcription factor genes as regulatory targets of Bra: foxA (later expressed in foregut and hindgut), foxP (foregut), hnf1 and tgif (both mid- and hindgut). Bra therefore constitutes an important node in the early endoderm GRN, controlling the expression of genes later involved in the formation of all three-gut compartments. In addition to Bra, Hnf1 responds also to an activating input from GataE, which represents a conserved regulatory linkage.
We have almost completed the perturbation analysis for all known early endoderm transcription factors and we are currently attempting to fill the remaining gaps in the endoderm GRN model by analyzing perturbation-induced changes in the spatial organization of gene expression.
|
cis-Regulatory linkages coordinating the sequential expression of Wnt and Delta ligands during early sea urchin embryogenesis
Joel Smith, Eric H. Davidson
We dissect the transcriptional regulatory relationships that align and coordinate the dynamic expression patterns of two signaling genes, wnt8 and delta, central to sea urchin endomesoderm specification. By cis-regulatory analysis we find delta transcription, while being broadly activated by the widely-expressed Runx transcription factor, is restricted by HesC-mediated repression through a site in the delta 5´UTR. Further cis-regulatory analysis reveals transcription of the hesC gene is itself shut down by Blimp1 repression. Blimp1 therefore represses the repressor of delta, thereby permitting its transcription. As Blimp1 autorepression is the mechanism responsible, indirectly, for extinguishing wnt8 expression in the center of an expanding torus pattern of expression, Blimp1 repression of hesC couples delta expression with the extinction of wnt8 transcription, resulting in side-by-side alignment of wnt8 and delta expression domains and their coordinated movement. The specific cis-regulatory linkages of the gene regulatory network thus, guarantee a reproducible developmental sequence.
|
A spatially dynamic cohort of regulatory genes in the endomesodermal gene network of the sea urchin embryo
Joel Smith, Ebba Kraemer*, Hongdau Liu, Christina Theodoris, Eric H. Davidson
A gene regulatory network subcircuit comprising the otx, wnt8 and blimp1 genes accounts for a moving torus of gene expression that sweeps concentrically across the vegetal domain of the sea urchin embryo. Here we confirm by mutation the inputs into the blimp1 cis‑regulatory module predicted by network analysis. Its essential design feature is that it includes both activation and autorepression sites. The wnt8 gene is functionally linked into the subcircuit in that cells receiving this ligand generate a b-catenin/Tcf input required for blimp1 expression, while the wnt8 gene in turn requires a Blimp1 input. Their torus-like spatial expression patterns and gene regulatory analysis indicate that the genes even-skipped and hox11/13b are also entrained by this subcircuit. We verify the cis-regulatory inputs of even-skipped predicted by network analysis. These include activation by b‑catenin/Tcf and Blimp1, repression within the torus by Hox11/13b, and repression outside the torus by Tcf in the absence of Wnt8 signal input. Thus, even-skipped and hox11/13b, along with blimp1 and wnt8, are members of a cohort of torus genes with similar regulatory inputs and similar, though slightly out-of-phase, expression patterns, which reflect differences in cis-regulatory design.
*IBG, Uppsala Biomedicinska Center, Uppsala University, S-75124, Uppsala, Sweden
|
Dissecting a three-gene positive feedback subcircuit
Qiang Tu
A regulatory state stabilization subcircuit is commonly used in gene regulatory networks (GRNs) for developmental specification. In sea urchin skeletogenic micromere specification, a 3-gene positive feedback loop, which is composed of erg, hex and tgif, is identified to be the stabilization device by large-scale perturbation analysis. Perturbation of any of the three genes will affect the other two, thereby also affecting the downstream differentiation gene battery. The cis-regulatory analysis of the genes in this loop is particularly interesting due to the complicated interactions among them. New tools are used to elucidate the subcircuit and investigate the cis‑regulatory modules.
One tool, 'qPCRplot,' is a computational program we developed for qPCR data processing and visualization. In the course of perturbation experiments, large amounts of quantitative measurements of gene expression are generated. The program is able to process the raw data, run all calculations, and visualize the result. The inputs are simply the files generated from qPCR experiments, and the output graphs all measurements of gene expressions in experimental/control perturbations at different time points, with indicators of significant changes, experimental variations, and rough expression levels. This program is currently being used to elucidate the sea urchin embryonic skeletogenesis GRN and other GRNs.
Another tool is the Solexa high-throughput sequencing platform, which facilitates discovery of cis-regulatory modules. Comparative sequence analyses have been shown to be reliable techniques to find cis-regulatory modules. Solexa technology is able to provide large amount of sequence data sufficient for this purpose. Here we apply the technology in the dissection of the cis‑regulatory modules of the three genes in this subcircuit. Several modules have been identified which drive the expression of reporter genes in the sea urchin embryo, including two tgif modules driving the expression in the primary mesenchyme cells and the endoderm at mesenchyme blastula stage, respectively. These two modules fully recapitulate the endogenous expression pattern of tgif at this stage.
|
|
|
