|
|
|
|
|
|
|
||||
| Home Help Feedback Subscriptions Archive Search Table of Contents | ||||
First published online 16 August 2006
doi: 10.1242/dev.02512
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1 Department of Human Genetics, Howard Hughes Medical Institute, University of
Utah School of Medicine, 15 N 2030 E Room 5100, Salt Lake City, UT 84112-5331,
USA.
2 Banting and Best Department of Medical Research, Graduate Department of
Molecular and Medical Genetics, University of Toronto, Donnelly Centre for
Cellular and Biomolecular Research, 160 College Street, Toronto, Ontario, M5S
3E1, Canada.
Author for correspondence (e-mail:
h.krause{at}utoronto.ca)
Accepted 28 June 2006
 |
SUMMARY |
|---|
 TOP SUMMARY INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key words: Nuclear receptor, Hormone, GAL4, Ligand, Drosophila
|
INTRODUCTION |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Ligand binding alters the LBD structure, leading to changes
in subcellular localization, DNA binding, dimerization, co-factor binding
and/or transcriptional activity (reviewed by
Nagy and Schwabe, 2004).
Nuclear receptor ligands tend to be small lipophilic compounds such as
steroids, fatty acids and vitamins. Despite extensive studies of NR structure,
function and regulation, approximately half of the 48 human NRs remain orphan
receptors - receptors for which no ligand has been identified.
NRs feature in most fundamental biological processes, functioning as key
control points in diverse signaling and metabolic pathways, including
electrolyte homeostasis (reviewed by
DeLuca, 2004;
Pearce, 2001), lipid
metabolism and homeostasis (reviewed by
Chawla et al., 2001), sex
determination (reviewed by Iyer and
McCabe, 2004), circadian rhythm and aging (reviewed in
Pardee et al., 2004). NRs also
play a central role in sensing xenobiotic compounds and coordinating an
appropriate detoxification response
(Willson and Kliewer, 2002).
Accordingly, NR mutations are associated with many common and lethal human
disorders, including cancer, diabetes and heart disease
(Agoulnik et al., 2004;
Alcalay et al., 1991;
Barroso et al., 1999;
Culig et al., 2000;
Gurnell et al., 2003;
Sarraf et al., 1999). Thus,
understanding NR function, and the ligands that regulate their activity,
provides an important opportunity to understand central aspects of growth,
metabolism, development and disease.
The fruit fly, Drosophila melanogaster, has 18 genes that encode
NRs. In spite of this relatively small number, the fly NRs span all major
subclasses of vertebrate receptors
(King-Jones and Thummel,
2005). Close fly orthologs of key vertebrate NRs include DHR3 (ROR
family members in vertebrates), DHR38 (NGFIB/NURR1), DHR78 (TR2/TR4),
Dissatisfaction (Dsf) and Tailless (Tll) (both orthologous to vertebrate Tlx),
E75 (Rev-Erb family members), ERR, DHR51 (PNR), FTZ-F1 (SF-1, LRH-1), HNF4,
Seven-up (SVP) (COUP-TF in vertebrates) and Ultraspiracle (USP) (RXR in
vertebrates). These features establish Drosophila as an ideal model
system for defining NR regulation and function. Although developmental and
genetic studies have been conducted on the majority of these NRs, ligands have
only been identified for two: E75, which binds heme and can use this
prosthetic group to exchange small diatomic gases
(Reinking et al., 2005); and
the ecdysteroid receptor EcR, which binds 20-hydroxyecdysone (20E) as a
heterodimer with USP (Riddiford et al.,
2001). Although not capable of direct hormone binding, DHR38 can
also be activated by ecdysteroids in combination with an activated form of USP
(Baker et al., 2003). 20E
directs the major developmental transitions in Drosophila, including
molting and metamorphosis (reviewed by
Riddiford, 1993;
Thummel, 2001). Many NRs are
transcriptionally induced by the 20E/EcR/USP complex and play crucial roles
during the larval-to-adult transition
(King-Jones and Thummel,
2005). Most Drosophila NRs, however, are also expressed
in embryos, larvae and adults - stages at which their functions are relatively
poorly understood (Sullivan and Thummel,
2003).
As part of an effort to gain comprehensive insights into NR regulation and
function, we have used an in vivo ligand detection system to follow NR LBD
activation patterns in intact developing animals. This bipartite detection
system consists of the LBD of each Drosophila NR fused to the
DNA-binding domain of yeast GAL4, along with a GAL4 UAS-controlled reporter
gene. As originally reported in cultured cells, in mouse tissues
(Mata De Urquiza et al., 1999;
Solomin et al., 1998) and
later in Drosophila (Han et al.,
2000; Kozlova and Thummel,
2002; Osterwalder et al.,
2001; Roman et al.,
2001), this system can respond properly to activating hormones.
Here, a heat-inducible promoter is used to drive ubiquitous expression of the
transgenic fusion proteins at different stages in an effort to document the
normal patterns of LBD activation during development, with the goal of using
these patterns to guide future studies of NR regulation and function. In
addition, a number of hypotheses were tested, leading to both suspected and
unexpected findings.
Among the results obtained, we find that half of the 18 GAL4-LBD fusion
proteins show no detectable activity patterns, suggesting that these function
only as repressors. The other half reveal a variety of developmentally
regulated patterns of activity, with dynamic changes in activation in specific
cell types. In several cases, fusion proteins are active in the same tissues,
revealing common or related functions. As expected, we show that the
activation pattern of GAL4-EcR in early Drosophila embryos is
dependent on the ecdysteroid biosynthetic pathway and that it responds to
exogenously added ecdysone. By contrast, GAL4-DHR38 activity, which also
responds to exogenous ecdysone, continues to function in the absence of
ecdysone, suggesting that EcR and DHR38 respond to distinct hormonal signals
at this stage in development. We test the hypothesis that xenobiotic agonists
will activate DHR96, which was recently shown to contribute to insect
xenobiotic responses (King-Jones et al.,
2006). In addition, we test the hypothesis that the ligand sensor
system can be used to reveal regulatory interactions between NR partner
proteins. We further demonstrate that this system can be used to screen for
new NR agonists and antagonists in live embryos and cultured larval tissues,
identifying two new agonists for USP.
|
MATERIALS AND METHODS |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) for a detailed description of the ligand
sensor system in Drosophila, including a discussion of interpreting
activation patterns and the spatial and temporal resolution of this system.
For dib mutant analyses, embryos were fixed in 4% paraformaldehyde
prior to staining. Reporter gene expression was detected using rabbit anti-GFP
(Abcam, 1:500) or mouse anti-ß-galactosidase antibodies (Promega, 1:750).
Secondary antibodies used were Cy5-conjugated goat anti-rat IgG (Abcam,
1:1000), Cy5-conjugated goat anti-mouse IgG (Jackson Immunolabs, 1:1000) or
Alexa 488-conjugated donkey anti-rabbit IgG (Molecular Probes, 1:1000). DAPI
(4',6-diamidino-2-phenylindole, Sigma, 0.1 µg/ml) or propidium iodide
(Sigma, 0.5 µg/ml) was used as a nuclear counterstain.
Embryo permeabilization, using heptane, was performed as described
previously (Schreuders et al.,
1996; Strecker et al.,
1994) with the following modifications. Ligand sensor embryos were
heat shocked to induce transgene expression, dechorionated rinsed with water
and then transferred to scintillation vials containing 2 ml of modified basic
incubation media (MBIM) (Strecker et al.,
1994) and 6 ml of heptane. Embryos were then swirled gently for 2
minutes and then transferred in 
100 µl heptane to deep well slides.
The excess heptane was removed and the embryos allowed to air dry just long
enough to allow the remaining heptane to evaporate. Embryos were then
immediately covered with MBIM containing 5.0x10-6 M
20-hydroxyecdysone, CITCO, PCN or TCPOBOP (all compounds from Sigma;
100x stocks were dissolved in ethanol). Embryos were incubated for 15
minutes at 25°C, the MBIM subsequently removed and the embryos covered
with Halocarbon oil and allowed to develop for a minimum of 2 hours prior to
observation.
Larval and prepupal staging, fixing and staining
Animals carrying both the GAL4-LBD and UAS-nlacZ
transgenes were maintained on food containing 0.5% bromophenol blue
(Andres and Thummel, 1994).
Vials were heat treated in a water bath at 37°C for 30 minutes and allowed
to recover for 6-7 hours at 25°C. Partial blue gut larvae were selected
from this population of heat-treated animals to assess activation during the
late third instar, prior to the high titer ecdysone pulse
(Andres and Thummel, 1994).
White prepupae were identified after 3-4 hours of recovery time and aged an
additional 2-3 hours to assay GAL4-LBD activation in early prepupae, for a
total of 6-7 hours after heat treatment. For earlier timepoints, animals were
staged at the L2-L3 molt (-48 hours) or as fully grown, blue gut animals upon
harvest (-24 hours). Animals were fixed in 1% glutaraldehyde (Sigma) in PBS
for 20 minutes and stained in 0.2% X-gal (Roche) for 15 minutes to overnight
at 37°C, depending on the strength of activation. Negative lines were
stained overnight in an attempt to reveal low levels of activation, and very
strongly activating lines were limited to short staining times to see
cell-autonomous stains and overall tissue structure. Mid-third instar larvae
carrying the hs-Gal4-DHR3, UAS-nlacZ, and hs-E75B transgenes were heat
treated, staged and assayed as described above. hs-Gal4-DHR3,
UAS-nlacZ animals lacking the hs-E75B construct were tested in parallel
as a control.
Organ culture
Mid-third instar (blue gut stage)
(Andres and Thummel, 1994)
hs-GAL4-USP; UAS-nlacZ larvae were heat treated in a water
bath at 37°C and allowed to recover for 3-6 hours at 25°C before
dissection. They were bisected and the anterior half was rinsed in PBS + 0.1%
Triton-X, everted, and placed in a glass nine-well glass dish in oxygenated
Grace's Insect Medium (Invitrogen). Compounds were administered at 1-100 µM
in freshly oxygenated Grace's medium with appropriate solvent controls. For
juvenile hormone treatment, glass dishes were treated with 20% PEG 20,000
(Fluka) and rinsed before treatment, to prevent the hormone from sticking to
the dish. Animals were cultured at room temperature overnight in an oxygenated
chamber, and fixed and stained in the morning as described above. Selected
data are depicted in Fig. 7 for
each tissue because, as found in our earlier studies, not all tissues of a
particular animal show a response (Baker et
al., 2003; Kozlova and
Thummel, 2002).
|
RESULTS |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
).
|
8-14 hours before
pupariation, just prior to the high titer 20E pulse that triggers
pupariation); or (3) as 2 hour prepupae. The tissue- and stage-specificity of
the activation patterns are summarized in
Table 1 and examples are shown
in Figs S1 (embryos) and S2 (larvae) in the supplementary material.
|
), SVP (Zelhof et
al., 1995a), TLL (Yu et al.,
1994), DSF (Pitman et al.,
2002), DHR4 (King-Jones et
al., 2005) and DHR78 (Zelhof
et al., 1995b). DHR51 and DHR83 are also likely to act as
repressors based on studies of their closest vertebrate homologues
(Chen et al., 2005). A
different method will be required to examine the regulatory activities of
these NRs.
GAL4-EcR activity is ligand-dependent and responsive to hormone during embryogenesis
The GAL4-EcR fusion protein exhibits transcriptional activity during
mid-embryogenesis in the amnioserosa (Fig.
2A) (Kozlova and Thummel,
2003a). To determine whether this activity pattern reflects the
presence of ligand, we tested the dependence of this localized GAL4-EcR
transcriptional activity on the biosynthesis of 
-ecdysone (E), which is
the immediate precursor of 20-hydroxyecdysone (20E), the active ecdysteroid in
insects (Gilbert et al.,
2002). This was achieved by crossing EcR ligand sensor flies with
flies carrying a mutation in the disembodied (dib) gene,
which encodes a cytochrome P450 enzyme required in the penultimate step of E
biosynthesis (Chavez et al.,
2000; Warren et al.,
2002). In a dib mutant background, GAL4-EcR activity in
the amnioserosa is no longer detectable, confirming that this response is 20E
dependent (Fig. 2B). By
contrast, no effects were observed on any of the other positively acting
ligand sensor lines when tested in the dib mutant background
(Fig. 2E,F,J; data not shown).
Together, these results show that EcR LBD activation in the embryonic
amnioserosa is dependent on zygotic ecdysteroid biosynthesis and that ligand
sensor fusion proteins function in a ligand-dependent and ligand-specific
fashion.
GAL4-EcR can be activated by exogenously added hormone during embryogenesis
To test whether ligand sensor proteins can be used to detect exogenously
added ligands, GAL4-EcR embryos were permeabilized and allowed to develop in
media supplemented with 20E. Fig.
2D shows a typical 20E-treated embryo, displaying widespread GFP
expression that extends significantly beyond the response to endogenous 20E in
the amnioserosa (Fig. 2A). By
contrast, GAL4-FTZ-F1 shows no changes from the untreated control
(Fig. 2G,H). These results,
which are similar to the previously published effects of 20E on GAL4-EcR and
GAL4-USP activity in cultured larval organs
(Baker et al., 2003;
Kozlova and Thummel, 2002),
show that the co-factors required for EcR ligand sensor activity are not
temporally or spatially limiting, and that the presence of ligand is
sufficient for ectopic activation. By exploiting high-throughput screening
strategies, it should therefore be possible to expand this effort by testing
large compound libraries for their effects on ligand sensor activities.
GAL4-DHR38 can be activated by 20E but is not dependent on dib
DHR38 has previously been shown to be activated by a set of ecdysteroids
that are distinct from those that significantly activate EcR, although this
effect appears to be achieved through a novel mechanism that does not involve
direct ligand binding (Baker et al.,
2003). Like the EcR ligand sensor, GAL4-DHR38 is also active in
the amnioserosa (Fig. 2I).
Interestingly though, this activity begins at an earlier stage than that of
the EcR ligand sensor (see Fig.
4, Table 1). In
addition, no effects were observed on DHR38 ligand sensor activity in
dib mutant embryos (Fig.
2J), possibly owing to its activation by maternally provided
ecdysteroids other than 20E. The DHR38 ligand sensor embryos treated with 20E
do, however, exhibit a modest but reproducible increase in their activation
pattern. The embryo in Fig. 2L
shows typical patches of responding cells, which, unlike the epidermal cells
that respond to endogenous ligand (Fig.
2K), tend to be contiguous and display weaker GFP fluorescence.
Thus, although GAL4-DHR38 is not dependent on E biosynthesis, it can respond
to the addition of exogenous 20E, although not as robustly as EcR, consistent
with the weak 20E activation of DHR38 previously seen in transient
transfection assays (Baker et al.,
2003).
|
).
DHR96 is also required for the proper transcriptional response of a subset of
phenobarbital-regulated genes. Accordingly, we used the ligand sensor system
to determine if known mammalian xenobiotic agonists could activate the DHR96
LBD. Embryos expressing GAL4-DHR96 were treated with the PXR-selective agonist
PCN and the CAR-selective agonists TCPOBOP and CITCO
(Blumberg et al., 1998;
Tzameli et al., 2000;
Maglich et al., 2003). Of
these, only CITCO gave reproducible, strong activation of the DHR96 ligand
sensor, indicating that the activation status of the DHR96 LBD can be
regulated by xenobiotic compounds in a manner similar to that of its
vertebrate orthologs (Fig.
2M,N). Interestingly, CITCO had no effect on GAL4-DHR96
subcellular localization, as it does with CAR
(Maglich et al., 2003). Thus,
the CITCO effect on the DHR96 LBD most probably occurs at the level of
co-activator recruitment.
ERR displays widespread and dynamic switches in ligand sensor activity
Although several ligand sensor lines showed shifts in their spatial and
temporal patterns of activation, the most dramatic changes were observed with
the ERR ligand sensor. GAL4-ERR activity is initially detected during
mid-embryogenesis in a subset of myoblasts (Fig.
3A, 6-11 hours). Its activity
then shifts to a different cell type at 14-17 hours after egg laying (AEL) -
the central nervous system (CNS) and a few cells in the peripheral nervous
system (Fig. 3A).
Interestingly, the timing of this shift in ERR tissue activity coincides with
a switch in ERR transcript sizes that occurs at 14-18 hours AEL
(Sullivan and Thummel,
2003).
Remarkably, the muscles and CNS also display GAL4-ERR activity in third
instar larvae, along with restricted activation in the midgut
(Fig. 3B). Moreover, the ERR
ligand sensor shows a dramatic switch in its activation pattern at this later
stage in development. GAL4-ERR activity is undetectable in early third instar
larvae, peaks at 24 hours after the L2-to-L3 molt, and then rapidly drops
to background levels again by late third instar
(Fig. 3B). This type of
widespread transient LBD activation has only been seen for the EcR and USP
ligand sensors at puparium formation, in response to the high titer late
larval pulse of 20E. Thus, ERR appears to be responding to a widespread,
temporally restricted activating signal that occurs in the mid-third
instar.
Temporally distinct patterns of ligand sensor activation in the amnioserosa
One of the advantages of studying all of the Drosophila nuclear
receptors in parallel is that common as well as unique features become
apparent. For example, although most tissues appear to support ligand sensor
activity at some stage, several tissues are particularly prevalent sites of
activity. These include the amnioserosa, yolk, regions of the midgut and fat
body. In some cases, the dynamics of these activity patterns suggest that
different ligand sensors may be responding to related sets of ligands or act
in functional hierarchies. The patterns of E78, DHR38, DHR3, HNF4 and EcR
ligand sensor activation in the amnioserosa provide one such example
(Fig. 4). The amnioserosa is a
dorsally located sheet of extra-embryonic polyploid cells that controls
essential morphogenetic movements such as retraction of the germ band and
dorsal closure (Kozlova and Thummel,
2003a; Narasimha and Brown,
2004; Reed et al.,
2004; Scuderi and Letsou,
2005). Interestingly, the E78 ligand sensor, which is active in
most embryonic and larval tissues, displays its first high level of activation
at about stage 9 in the amnioserosa (Fig.
4A, arrowheads). The DHR38 and DHR3 ligand sensors respond at
about the same time or soon after (Fig.
4D,G, arrowheads), with downregulation of E78, DHR38, and DHR3
ligand sensor activity in the amnioserosa at stages 13-14
(Fig. 4C,F,I). By contrast,
GAL4-HNF4 is active in yolk nuclei at early times
(Fig. 4J, arrows), only
switching to the amnioserosa at stage 12
(Fig. 4K,L, arrowheads). The
EcR and USP ligand sensors are the last to display activity in the
amnioserosa, beginning at about stage 13 (arrowheads in
Fig. 4O for EcR;
Table 1 and data not shown).
Thus, not only is the amnioserosa a hotspot for ligand sensor activation, but
the different timing of these responses may be due to distinct threshold
responses to the same or related set of ligands or to hierarchical
interactions between NRs and/or co-factors.
|
|
Following hatching and the onset of feeding, DHR3, DHR38, HNF4 and FTZ-F1 ligand sensor activities continue within regions of the midgut (Fig. 5K-T). Of these, GAL4-DHR3 has the highest and most uniform pattern of activity, spanning most of the midgut and gastric caeca (Fig. 5M-N). Although DHR3 ligand sensor activity is evident in the proventriculus of the midgut of third instar larvae, it is downregulated at puparium formation (Fig. 5M,N arrow). By contrast, DHR38 and HNF4 ligand sensor activities are restricted to the cells that span the junction of the midgut, proventriculus and gastric caeca (Fig. 5O-R), suggesting that these receptors may be responding to similar signal(s). Interestingly, GAL4-FTZ-F1 is activated in the midgut only after feeding has ceased, at puparium formation (Fig. 5S-T).
|
Identification of new receptor agonists by compound screening
Larval organ culture provides an accurate and simple means of testing
compounds for hormonal activity within a normal physiological context
(Ashburner, 1972). We thus
asked whether larval organ culture could be combined with our ligand sensors
to identify novel receptor agonists, screening for activation of GAL4-USP. Two
properties of USP make it a good prototype for this study. First, like its
vertebrate ortholog RXR, USP can dimerize with multiple Drosophila
nuclear receptors (Sutherland et al.,
1995), increasing the likelihood of obtaining a positive response
to a new compound. Second, the ability of GAL4-USP to interact with EcR, and
to activate reporter gene expression in the presence of the EcR ligand 20E,
permits the use of 20E as a positive control
(Kozlova and Thummel,
2002).
|
).
Over 40 other compounds were also tested for their ability to activate
GAL4-USP in this assay (Table
2). We focused on a range of known and potential insect hormones
and plant-derived compounds, including ecdysteroids, juvenoids, plant hormones
and psoralen-derived xenochemicals. We also tested two forms of vitamin D as
well as fatty acids and xenobiotics, which can regulate vertebrate nuclear
receptors (Chawla et al.,
2001). Each assay was repeated at least twice, and 20E was
included in each experiment as a positive control.
|
-ecdysone, 2-deoxy-20-hydroxyecdysone,
20,26-dihydroxyecdysone, 20-hydroxyecdysone 22-acetate and makisterone A
(Table 2), consistent with an
earlier study of compounds that activate the EcR/USP heterodimer in cultured
cells (Baker et al., 2000). The
ability of -ecdysone to activate GAL4-USP is most probably due to its
conversion to 20E (Petryk et al.,
2003). The plant ecdysteroids ajugasterone C, azadirachtin,
cyasterone, muristerone A, polypodine B and ponasterone A are also able to
activate GAL4-USP in this system, consistent with their ecdysteroid properties
and earlier cell culture studies (Baker et
al., 2000). Although linear psoralens and brassinosteroids did not
activate GAL4-USP in this assay, an angular psoralen, the furanocoumarin
angelicin, is a relatively strong activator of GAL4-USP
(Fig. 7E-H). Interestingly,
angelicin has no effect on GAL4-EcR (data not shown).
Juvenile hormone (JH) and several JH analogs were also tested for their
ability to activate GAL4-USP, following up on studies suggesting that USP is a
receptor for this insect hormone (Xu et
al., 2002). However, JHI, JHII, JHIII and two well-studied JH
analogs, pyriproxifen and methoprene, were unable to activate GAL4-USP
(Table 2). Curiously, however,
we observed weak activation by the insecticide fenoxycarb in tissues from some
animals but not others, suggesting that activation may be influenced by
developmental stage or the physiological state of the animal
(Fig. 7I-L). Fenoxycarb is a
carbamate insecticide that mimics the action of JH on several physiological
pathways, including molting and reproduction. Given the absence of an effect
with natural JH, pyriproxifen or methoprene, however, the observed activation
by fenoxycarb may represent a xenobiotic response, rather than an effect
caused by its JH-like activity. Neither angelicin nor fenoxycarb had a
significant effect on the DHR96 ligand sensor in larval organ culture,
suggesting that USP either responds directly to these compounds, or acts as a
heterodimer with another NR (L.P. and M. Horner, unpublished).
A repressive heterodimer partner can regulate ligand sensor activity
Earlier studies have shown that DHR3 induces ßFTZ-F1 at the onset of
metamorphosis, and that the activation function of DHR3 in cultured cells can
be blocked by heterodimerization with E75B, an isoform of E75 that is missing
its DNA binding domain but contains an intact LBD
(Segraves and Hogness, 1990;
White et al., 1997). We used
the ligand sensor system to test if this functional interaction also occurs in
vivo and to determine if the ligand sensor system can be used to monitor
repressive protein-protein interactions. As described above, GAL4-DHR3 is
widely active in late third instar larval tissues
(Fig. 5M,N,
Fig. 6,
Fig. 8A; see also Fig. S2 in
the supplementary material). This pattern changes dramatically, however,
following ectopic co-expression of full-length E75B protein, with a
significant reduction in DHR3 ligand sensor activity in the epidermis,
proventriculus of the midgut, CNS and fat body
(Fig. 8B). Importantly, this
pattern reflects that normally seen in early prepupae with GAL4-DHR3
(Fig. 8C), suggesting that the
change in DHR3 ligand sensor activation during the onset of metamorphosis can
be accounted for by ecdysone-induced expression of endogenous E75B at puparium
formation. This observation supports previous evidence that E75B is sufficient
to block the activation function of DHR3 at the onset of metamorphosis
(White et al., 1997). In
addition, the specific inability of E75B to block GAL4-DHR3 activity in the
larval midgut suggests that either the DHR3/E75B heterodimer cannot form in
this tissue or, more likely, that modifying ligand(s) or co-factors may block
the repressive function of E75B in this tissue. Given the recent discovery
that E75 binds heme and responds to diatomic messenger gases, it is possible
that E75B activity may be differentially regulated in a tissue-specific manner
(Reinking et al., 2005).
Moreover, this experiment demonstrates that ligand sensor fusion proteins can
be used to assess regulatory responses due to protein partners and co-factors,
as well as to detect ligand-regulated responses.
|
|
|
DISCUSSION |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Hormonal regulation of GAL4-LBD activation in the amnioserosa and yolk
Examination of the nine active ligand sensor lines provided a number of
insights into possible relationships between their corresponding NRs. For
example, although each of these ligand sensors displays unique temporal and
spatial patterns of activity, activation in specific tissues and stages is
common to many. These common sites of LBD activity may indicate shared
functions, hierarchical or physical interactions, or related ligands. Examples
of tissues that represent hotspots for GAL4-LBD activation include the
amnioserosa, yolk, midgut and fat body.
Each of these tissues, and the stages at which they score positively,
correlates well with the presence of putative ligands. The yolk, for example,
is believed to act as a storage site for maternally provided ecdysteroids
during embryogenesis. Work with other insects has shown that these
ecdysteroids are conjugated in an inactive form to vitellin proteins via
phosphate bridges (Hoffmann and Lagueux,
1985). Around mid-embryogenesis, these yolk proteins and phosphate
bonds are cleaved, thereby releasing what are presumed to be the earliest
biologically active ecdysteroids in the embryo
(Bownes et al., 1988).
Interestingly though, GAL4-EcR activation in the amnioserosa depends on
dib function (Fig.
2B), suggesting that the final steps in the linear E biosynthetic
pathway are required for EcR function in this tissue
(Chavez et al., 2000;
Warren et al., 2002) and
contradicting the prediction that this activity would be dependent on maternal
ecdysteroids and independent of the zygotic biosynthetic machinery
(Kozlova and Thummel, 2003a).
The mechanisms by which dib exerts this essential role in providing
an EcR ligand, however, remain to be determined.
The response of the EcR and USP ligand sensors in the adjacent amnioserosa
tissue shows that active ecdysteroids are not present until the hormone
reaches the amnioserosa. A recent study of yolk-amnioserosa interactions has
revealed dynamic transient projections that emanate from one tissue and
contact the other, suggesting that there may be functional interactions
between these two cell types (Reed et al.,
2004). It is possible that these projections mediate the transfer
of lipophilic ligand precursors from the yolk to the amnioserosa. This
transfer, in turn, could determine the proper timing of EcR activation in the
amnioserosa, thus triggering the major morphogenetic movements that establish
the body plan of the first instar larva
(Kozlova and Thummel,
2003a).
|
). The
activation of GAL4-DHR38 that we observe in the embryonic amnioserosa is
consistent with this model of DHR38 regulation. First, exogenous 20E can only
weakly activate GAL4-DHR38, relative to the strong ectopic activation seen
with 20E on the EcR ligand sensor (Fig.
2L,D). This correlates with the weak ability of 20E to activate
DHR38 in cell culture transfection assays relative to the strong 20E
activation of EcR (Baker et al.,
2003). Second, the DHR38 ligand sensor is activated in the
amnioserosa earlier than the EcR construct, suggesting that it is responding
to a different signal (Fig. 4).
It is possible that this signal is an ecdysteroid precursor that can act on
DHR38 but not EcR - paralleling the ability of DHR38 to be activated by E, the
precursor to 20E, which activates EcR. This putative ecdysteroid must be
produced in a manner independent of the conventional ecdysteroid biosynthetic
pathway, however, as a zygotic dib mutation has no effect on
GAL4-DHR38 activation in the amnioserosa. Rather, this early activation may be
due to maternal ecdysteroids that are conjugated and inactive in the yolk and
transferred to the amnioserosa. These studies highlight the value of combining
mutations in hormone biosynthesis with ligand sensor activation as a powerful
means of dissecting hormone signaling pathways. Further studies of DHR38
function and regulation in embryos could help clarify the potential
significance of this distinct activation response.
DHR3, DHR38 and HNF4 ligand sensors appear to respond to metabolic signals
Interestingly, the midgut continues to be a hotspot for ligand sensor
activity long after it has engulfed the yolk during embryogenesis. This seems
logical, as the midgut is responsible for most lipid absorption and release,
and many vertebrate NRs are involved in fatty acid, cholesterol and sterol
metabolism and homeostasis (Chawla et al.,
2001). The observed restriction of ligand sensor activity to a
narrow group of cells located at the base of the gastric caeca is of
particular interest (Fig.
5M-R). This is the site where nutrients in a feeding larva are
absorbed into the circulatory system
(Chapman, 1998). The
activation of DHR3, DHR38 and HNF4 ligand sensors in this region of the
gastric caeca suggests that these receptors are activated by one or more small
nutrient ligands (Fig. 5M-R).
Moreover, this suggests that the corresponding receptors may exert crucial
metabolic functions by acting as nutrient sensors.
Further evidence of metabolic functions for DHR3, DHR38 and HNF4 arises
from their ligand sensor activation patterns in the embryonic yolk and larval
fat body (Figs 4,
6). The yolk is the main
nutrient source for the developing embryo and represents an abundant source of
lipids, correlating with specific activation of DHR3, DHR38 and HNF4 ligand
sensors in this cell type during embryogenesis
(Fig. 4C,E,G). Upon hatching
into a larva, the fat body acts as the main metabolic organ of the animal,
functionally equivalent to the mammalian liver. Upon absorption by the gastric
caeca, nutrients travel through the circulatory system and are absorbed by the
fat body, where they are broken down and stored as triglycerides, glycogen and
trehalose. Once again, the efficient activation of the DHR3, DHR38 and HNF4
ligand sensors in the fat body of metabolically active third instar larvae,
and lack of sensor activity in non-feeding prepupae, supports the model that
the corresponding NRs operate as metabolic sensors
(Fig. 6). This proposed
function is consistent with the roles of their vertebrate orthologs. Mammalian
ROR, the ortholog of DHR3, binds cholesterol and plays a crucial role in lipid
homeostasis (Kallen et al.,
2004; Lau et al.,
2004). Similarly, mammalian HNF4 can bind C14-18 fatty acids, is
required for proper hepatic lipid metabolic gene regulation and lipid
homeostasis, and is associated with human Maturity-Onset Diabetes of the Young
(MODY1) (Dhe-Paganon et al.,
2002; Hayhurst et al.,
2001; Shih et al.,
2000; Stoffel and Duncan,
1997; Wisely et al.,
2002). The studies described here suggest that DHR3 and HNF4 may
perform similar metabolic functions in flies, defining a new genetic model
system for characterizing these key NRs.
New insights into the regulation of Drosophila xenobiotic responses
Several vertebrate NRs play a central role in xenobiotic responses by
directly binding toxic compounds and inducing the expression of key
detoxification enzymes such as cytochrome P450s and glutathione transferases
(Willson and Kliewer, 2002).
Ligand sensor activation observed in the gut, epidermis, tracheae or fat body
could represent xenobiotic responses insofar as toxic compounds could enter
the organism through any of these tissues. Directed screens that test
xenobiotic compounds for their ability to activate Drosophila NR
ligand sensors will provide a means of identifying potential xenobiotic
receptors. Understanding these response systems, in turn, could facilitate the
production of insect resistant crops and the development of more effective
pesticides. In this regard, we have shown that DHR96, which is required for
proper xenobiotic responses in Drosophila, can be activated by the
CAR-selective agonist CITCO, suggesting that it may be regulated in a manner
similar to that of the vertebrate xenobiotic receptors
(Fig. 2M,N). It is also
interesting to note that angelicin was found to activate the USP ligand sensor
fusion (Fig. 7E-H). Angelicin
is an angular furanocoumarin that has the furan ring attached at the 7,8
position of the benz-2-pyrone nucleus. Detailed studies have shown that
insects have adapted to the presence of furanocoumarins in their host plants
by expressing specific cytochrome P450 enzymes that detoxify these compounds
(Hung et al., 1996). In the
black swallowtail butterfly (Papilio polyxenes), furanocoumarins
induce the transcription of P450 genes through an unknown regulatory pathway,
thereby aiding in xenobiotic detoxification
(Berenbaum, 2002). Our
observation that angelicin, and not the linear furanocoumarins
8-methoxypsoralen (xanthotoxin) or 5-methoxypsoralen (bergapten), can activate
GAL4-USP suggests that NRs may mediate this detoxification response and may be
capable of distinguishing between the linear and angular chemical forms. It is
possible that USP may mediate this effect on its own or, more likely, as a
heterodimer partner with another NR. Similarly, the activation of GAL4-USP by
fenoxycarb may represent a xenobiotic response
(Fig. 7I-L). This activation,
however, is weaker and more variable than the activation we observed with
angelicin. Identifying other factors that mediate xenobiotic responses in
Drosophila would provide a new basis for dissecting the control of
detoxification pathways in higher organisms.
ERR activity appears to be regulated by a temporally restricted and widespread signal
GAL4-ERR displays a remarkable switch in activity during mid-embryogenesis,
from strong activation in the myoblasts to specific and strong activation in
the CNS (Fig. 3A). The ERR
ligand sensor also shows widespread transient activation in the mid-third
instar (Fig. 3B), a time when
larval ERR gene expression begins
(Sullivan and Thummel, 2003),
together with a global switch in gene expression that prepares the animal for
entry into metamorphosis 1 day later
(Andres et al., 1993;
Cherbas et al., 2003). This
so-called mid-third instar transition includes upregulation of EcR, providing
sufficient receptor to transduce the high titer late larval 20E hormone pulse
(Talbot et al., 1993),
upregulation of the Broad-Complex, which is required for entry into
metamorphosis (Kiss et al.,
1988), and induction of the genes that encode a polypeptide glue
used to immobilize the puparium for metamorphosis
(Lehmann, 1996). The signal
and receptor that mediate this global reprogramming of gene expression remain
undefined. The widespread activation of GAL4-ERR at this stage raises the
interesting possibility that it may play a role in this transition. Moreover,
given that the only ligand sensors to display widespread transient activation
are EcR and USP, in response to 20E, it is possible that this response
reflects a systemic mid-third instar pulse of a ERR hormone. Vertebrate
members of the ERR family can bind the synthetic estrogen diethylstilbestrol
and the selective ER modulator tamoxifen, as well as its metabolite,
4-hydroxytamoxifen, suppressing their otherwise constitutive activity in cell
culture (Coward et al., 2001).
This is notably different from the highly restricted patterns of ERR ligand
sensor activity that we detect in Drosophila, which suggests that it
does not function as a constitutive activator in vivo. Rather, we envision
that the patterns of ERR activation are precisely modulated by protein
co-factors and/or one or more ligands to direct the dynamic shifts in
activation that we detect during embryogenesis and third instar larval
development. Functional studies of the Drosophila homolog of the ERR
receptor family may provide a basis for understanding these dynamic shifts in
LBD activation, as well as revealing a natural ligand for this NR.
Future directions
This study provides, for the first time, a comprehensive analysis of the
activation patterns of NR LBDs in a developing organism, uncovering a wide
range of dynamic and localized changes in activity that occur as the animal
undergoes massive developmental and physiological changes during embryogenesis
and early metamorphosis. Our data provide a foundation for biochemical and
genetic studies aimed at defining the molecular and functional basis for these
LBD activation responses. We anticipate that this work will lead to new
insights into NR regulation and function, including the discovery of new NR
partner proteins and endogenous ligands. In addition, extensions of this work
could have practical consequences by identifying novel agonists and
antagonists that could be used for insect population control, potentially
impacting deadly insect-borne human diseases such as malaria and providing
more effective means of crop protection. Finally, characterization of the
activity patterns described here should lead to novel insights into embryonic
patterning, metabolic control, xenobiotic metabolism, immunity, circadian
rhythms and aging, with direct implications for how these pathways might be
controlled by orthologous NRs in humans.
Supplementary material
Supplementary material for this article is available at
http://dev.biologists.org/cgi/content/full/133/18/3549/DC1
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
|
REFERENCES |
|---|
|
TOPSUMMARYINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Agoulnik, I. U., Tong, X. W., Fischer, D. C., Korner, K.,
Atkinson, N. E., Edwards, D. P., Headon, D. R., Weigel, N. L. and Kieback, D.
G. (2004). A germline variation in the progesterone receptor
gene increases transcriptional activity and may modify ovarian cancer risk.
J. Clin. Endocrinol. Metab.
89,6340
-6347.
Alcalay, M., Zangrilli, D., Pandolfi, P. P., Longo, L.,
Mencarelli, A., Giacomucci, A., Rocchi, M., Biondi, A., Rambaldi, A., Lo Coco,
F. et al. (1991). Translocation breakpoint of acute
promyelocytic leukemia lies within the retinoic acid receptor alpha locus.
Proc. Natl. Acad. Sci. USA
88,1977
-1981.
Andres, A. J. and Thummel, C. S. (1994). Methods for quantitative analysis of transcription in larvae and prepupae. Methods Cell Biol. 44,565 -573.[Medline]
Andres, A. J., Fletcher, J. C., Karim, F. D. and Thummel, C. S. (1993). Molecular analysis of the initiation of insect metamorphosis: a comparative study of Drosophila ecdysteroid-regulated transcription. Dev. Biol. 160,388 -404.[CrossRef][Medline]
Ashburner, M. (1972). Patterns of puffing activity in the salivary gland chromosomes of Drosophila. VI. Induction by ecdysone in salivary glands of D. melanogaster cultured in vitro. Chromosoma 38,255 -281.[CrossRef][Medline]
Baker, K. D., Warren, J. T., Thummel, C. S., Gilbert, L. I. and Mangelsdorf, D. J. (2000). Transcriptional activation of the Drosophila ecdysone receptor by insect and plant ecdysteroids. Insect Biochem. Mol. Biol. 30,1037 -1043.[CrossRef][Medline]
Baker, K. D., Shewchuk, L. M., Kozlova, T., Makishima, M., Hassell, A., Wisely, B., Caravella, J. A., Lambert, M. H., Reinking, J. L., Krause, H. et al. (2003). The Drosophila orphan nuclear receptor DHR38 mediates an atypical ecdysteroid signaling pathway. Cell 113,731 -742.[CrossRef][Medline]
Barroso, I., Gurnell, M., Crowley, V. E., Agostini, M., Schwabe, J. W., Soos, M. A., Maslen, G. L., Williams, T. D., Lewis, H., Schafer, A. J. et al. (1999). Dominant negative mutations in human PPARgamma associated with severe insulin resistance, diabetes mellitus and hypertension. Nature 402,880 -883.[Medline]
Berenbaum, M. R. (2002). Postgenomic chemical ecology: from genetic code to ecological interactions. J. Chem. Ecol. 28,873 -896.[CrossRef][Medline]
Blumberg, B., Sabbagh, W., Jr, Juguilon, H., Bolado, J., Jr, van
Meter, C. M., Ong, E. S. and Evans, R. M. (1998). SXR, a
novel steroid and xenobiotic-sensing nuclear receptor. Genes
Dev. 12,3195
-3205.
Bownes, M., Shirras, A., Blair, M., Collins, J. and Coulson,
A. (1988). Evidence that insect embryogenesis is regulated by
ecdysteroids released from yolk proteins. Proc. Natl. Acad. Sci.
USA 85,1554
-1557.
Chapman, R. (1998). The Insects: Structure and Function. Cambridge: Cambridge University Press.
Chavez, V. M., Marques, G., Delbecque, J. P., Kobayashi, K., Hollingsworth, M., Burr, J., Natzle, J. E. and O'Connor, M. B. (2000). The Drosophila disembodied gene controls late embryonic morphogenesis and codes for a cytochrome P450 enzyme that regulates embryonic ecdysone levels. Development 127,4115 -4126.[Abstract]
Chawla, A., Repa, J. J., Evans, R. M. and Mangelsdorf, D. J.
(2001). Nuclear receptors and lipid physiology: opening the
X-files. Science 294,1866
-1870.
Chen, J., Rattner, A. and Nathans, J. (2005).
The rod photoreceptor-specific nuclear receptor Nr2e3 represses transcription
of multiple cone-specific genes. J. Neurosci.
25,118
-129.
Cherbas, L., Hu, X., Zhimulev, I., Belyaeva, E. and Cherbas,
P. (2003). EcR isoforms in Drosophila: testing
tissue-specific requirements by targeted blockade and rescue.
Development 130,271
-284.
Coward, P., Lee, D., Hull, M. V. and Lehmann, J. M.
(2001). 4-Hydroxytamoxifen binds to and deactivates the
estrogen-related receptor gamma. Proc. Natl. Acad. Sci.
USA 98,8880
-8884.
Culig, Z., Hobisch, A., Bartsch, G. and Klocker, H. (2000). Androgen receptor - an update of mechanisms of action in prostate cancer. Urol. Res. 28,211 -219.[CrossRef][Medline]
DeLuca, H. F. (2004). Overview of general
physiologic features and functions of vitamin D. Am. J. Clin.
Nutr. 80,1689S
-1696S.
Dhe-Paganon, S., Duda, K., Iwamoto, M., Chi, Y. I. and Shoelson,
S. E. (2002). Crystal structure of the HNF4 ligand
binding domain in complex with endogenous fatty acid ligand. J.
Biol. Chem. 277,37973
-37976.
Gilbert, L. I., Rybczynski, R. and Warren, J. T. (2002). Control and biochemical nature of the ecdysteroidogenic pathway. Annu. Rev. Entomol. 47,883 -916.[CrossRef][Medline]
Gurnell, M., Savage, D. B., Chatterjee, V. K. and O'Rahilly,
S. (2003). The metabolic syndrome: peroxisome
proliferator-activated receptor gamma and its therapeutic modulation.
J. Clin. Endocrinol. Metab.
88,2412
-2421.
Han, D. D., Stein, D. and Stevens, L. M. (2000). Investigating the function of follicular subpopulations during Drosophila oogenesis through hormone-dependent enhancer-targeted cell ablation. Development 127,573 -583.[Abstract]
Hayhurst, G. P., Lee, Y. H., Lambert, G., Ward, J. M. and
Gonzalez, F. J. (2001). Hepatocyte nuclear factor 4
(nuclear receptor 2A1) is essential for maintenance of hepatic gene expression
and lipid homeostasis. Mol. Cell. Biol.
21,1393
-1403.
Hoffmann, J. A. and Lagueux, M. (1985). Endocrine aspects of embryonic development in insects. In Embryogenesis and Reproduction. Vol.1 (ed. G. A. Kerkut and L. I. Gilbert), pp.435 -460. Oxford: Pergamon Press.
Hung, C. F., Holzmacher, R., Connolly, E., Berenbaum, M. R. and
Schuler, M. A. (1996). Conserved promoter elements in the
CYP6B gene family suggest common ancestry for cytochrome P450 monooxygenases
mediating furanocoumarin detoxification. Proc. Natl. Acad. Sci.
USA 93,12200
-12205.
Iyer, A. K. and McCabe, E. R. (2004). Molecular mechanisms of DAX1 action. Mol. Genet. Metab. 83, 60-73.[CrossRef][Medline]
Kallen, J., Schlaeppi, J. M., Bitsch, F., Delhon, I. and Fournier, B. (2004). Crystal structure of the human RORalpha Liga
