Retinoid X receptor and retinoic acid response in the marine sponge Suberites domuncula
1 Institut für Physiologische Chemie, Abteilung Angewandte
Molekularbiologie, Universität, Duesbergweg 6, D-55099 Mainz,
Germany
2 Center for Marine Research, `Ruder Boskovic' Institute, HR-52210 Rovinj,
Croatia
* Author for correspondence (e-mail: WMUELLER{at}mail.UNI-Mainz.DE)
Accepted 13 June 2003
![]() |
Summary |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: sponge, Porifera, Suberites domuncula, retinoid X receptor, retinoic acid, canal formation, morphogenesis, primmorphs, functional molecular evolution
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Sponges require morphogenetic factors acting extracellularly, e.g.
myotrophin (Schröder et al.,
2000), and nuclear receptors, acting intracellularly, e.g.
allograft inflammatory factor in conjunction with the Tcf-like transcription
factor, for their differentiation
(Müller et al., 2002
).
Sponges even possess specific homeodomain proteins, e.g. a LIM/homeobox
protein (Wiens et al.
, in
press). The introduction of primmorphs, special three-dimensional aggregates
composed of proliferating cells that have retained their differentiation
potency and thus represent an in vitro cell culture system
(Müller et al., 1999
),
was another very important step towards the elucidation of potential
morphogenetic factors/processes. This system provided proof that homologous
cell substrate molecules, e.g. galectin, cause signaling events in
non-structured primmorphs that ultimately result in the formation of canals, a
process very likely correlated with the expression of LIM/homeobox
transcription factors (Wiens et
al.
, in press). Canal formation not only occurs in primmorphs but
is also one of the major morphogenetic processes observed during hatching of
sponges from gemmules, the asexual reproduction bodies
(Imsiecke et al., 1994
), and
during embryogenesis (see Simpson,
1984
; Weissenfels,
1989
).
Gemmules are formed by many sponges in response to adverse physiological or
ecological factors (see Simpson,
1984). For example, gemmulation occurs in response to an increased
bacterial load (Rasmont, 1963
)
or change in light intensity (Rasmont,
1970
); in addition to exogenous factors, endogenous factors are
also known to initiate gemmule formation
(Simpson and Gilbert, 1973
).
Focusing on S. domuncula, under experimental conditions
(Böhm et al., 2001
) and in
the field, the formation of gemmules is very frequently seen after infection
of the specimens with bacteria.
Retinoids and carotenoids have been detected in sponges
(Biesalski et al., 1992). In
Geodia cydonium retinoic acid reduces cell metabolism
(Biesalski et al., 1992
) but
does not induce gemmule formation. In the freshwater sponge Ephydatia
muelleri, however, retinoic acid initiates morphogenetic events leading
to formation of buds/gemmules (Imsiecke et
al., 1994
). At micromolar concentrations, retinoic acid causes a
downregulation in the expression of the EmH-3 homeobox-containing
gene in E. muelleri (Nikko et
al., 2001
), which led the authors to conclude that retinoic acid
and the responsive EmH-3 gene are involved in differentiation and
re-differentiation of archaeocytes to choanocytes, and hence in formation of a
functional aquiferous system.
In the coral Acropora millepora, receptors were detected that
share some similarity to retinoid X receptors (RXRs)
(Grasso et al., 2001), but to
date molecular data on the presence of nuclear hormone receptors in the
Porifera, the phylogenetically oldest metazoan phylum, are missing. Several
RXRs are activated by known regulatory ligands; receptors for whom the ligands
have not yet been identified were termed `orphan receptors'
(Aranda and Pascual, 2001
).
Experimental identification of nuclear receptors in sponges would reinforce
the theory of monophyletic evolution of animal genomic regulatory systems, as
with e.g. cell surface receptors and the receptor tyrosine kinases
(Schäcke et al., 1994
),
both establishing the basis for common cellcell/cellmatrix
regulatory networks present in all Metazoa
(Müller et al.,
1994
).
For an initial search of a nuclear hormone receptor in the phylogenetically
oldest animals we used the primmorph system of S. domuncula and
concentrated on the retinoid X receptor (RXR). The decision to use RXR was
based on the facts that (i) in sponges, retinoic acid causes morphogenetic
responses, (ii) related receptors had already been isolated from the coral
Acropora millepora (Grasso et
al., 2001) and the Cnidarian Tripedalia cystophora
(Kostrouch et al., 1998
) and
(iii) RXR heterodimerizes with other nuclear receptors. RXRs form heterodimers
with orphan receptors involved in lipid metabolism; these receptors bind their
ligands with lower affinity than the endocrine steroid receptors and function
as lipid sensors/regulators (Chawla et al.,
2001
). It is well established that in several biological systems
retinoic acid is involved in the induction of apoptosis (see
Sato et al., 1999
).
In the present work we have studied the morphogenetic effect of retinoic
acid in vivo and in vitro, particularly in the primmorph
system of the sponge S. domuncula. The amino acid sequence of RXR
from S. domuncula is described and we demonstrate its modulated
expression in response to retinoic acid. We investigated whether the
morphogenetic effect of retinoic acid in the primmorph system occurs
via caspase-mediated apoptosis or via a controlled
expression of homeotic genes, by studying expression of the S.
domuncula caspase gene (SDCASPR; accession number
AJ426651; Wiens et al., 2003a) and the LIM/homeobox protein
(SDLIM4; AJ493059) in parallel. The S. domuncula caspase is
also involved in apoptosis (Wiens et al.,
2003).
Finally, the potential effect of retinoic acid on the expression of
SDCYP4 (accession number Y17616;
Müller et al., 1999), a
gene belonging to the cytochrome P450 superfamily
(Goksøyr and Förlin,
1992
), was investigated.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Sponges
Live specimens of Suberites domuncula Olivi (Porifera,
Demospongiae, Hadromerida) were collected near Rovinj (Croatia). The sponges
(2040 specimens) were kept in 200 litre large aquaria in Mainz
(Le Pennec et al., 2003).
S. domuncula lives in symbiosis with a hermit crab Pagurites
oculatus Herbst (Decapoda: Paguridea), which resides in the snail shell
Trunculariopsis trunculus L. (Gastropoda: Muricidae;
Herland-Meewis, 1948
) on which
the sponge settles.
In the incubation experiments 35 specimens were kept under conditions of optimal aeration in 20 litre aquaria containing 150 µmol l1 of retinoic acid for up to 10 days.
Dissociation of cells and formation of primmorphs
Primmorphs from dissociated cells were obtained as described
(Müller et al., 1999).
After dissociation and washing, the dissociated cells at a density of
3x106 cells ml1 were added to each well of
12-well tissue culture test plates (2 ml per well) in natural seawater
(Sigma), supplemented with 0.1% of RPMI1640-medium, silicate to the optimal
concentration of 60 µmol l1
(Krasko et al., 2000
) and 30
µmol l1 Fe3+ (added as ferric citrate)
(Krasko et al., 2002
).
To stimulate primmorph formation the cells were cultivated either on
non-coated dishes, or on tissue culture test plates coated with poly-L-lysine
(Mr 30 00070 000) or galectin
(Wiens et al., in press).
Homologous galectin (rGALEC1_SUBDO) was obtained by recombinant techniques
from a complete cDNA, termed SDGALEC1 (EMBL/GenBank accession number
AJ493055; Wiens et al.
, in
press). To obtain coated plates, 500µl of poly-L-lysine solution (10 µg
ml1) or a recombinant galectin solution (10 µg
ml1) were added per well. After 3 h at room temperature the
wells were washed and the dissociated cells added.
Primmorphs formed after 58 days; those that developed on non-coated plates had diameters of 36 mm as observed by inspection with an Olympus OSP-MBI binocular light microscope. When cultivation was continued on non-coated dishes the size and the shape of the primmorphs remained unchanged during a prolonged period of more than 3 weeks. If however, primmorphs were transferred to galectin- or poly-L-lysine-coated tissue culture test plates after an initial period of 5 days, and cultured for up to 10 days, they developed canal-like structures; these canal-containing primmorphs were approximately 10 mm in diameter.
Treatment of primmorphs with retinoic acid
The potential effect of retinoic acid on canal-formation in primmorphs was
studied by first cultivating primmorphs for 5 days on non-coated plates,
followed by transfer to either galectin- or poly-L-lysine-coated tissue
culture test plates; 2-3 primmorphs were retained per well. At day 5 after
transfer, retinoic acid was added at concentrations of 150 µmol
l1. At the end of the incubation the three-dimensional-cell
aggregates were collected for further analysis.
Determination of apoptosis
The amount of apoptotic cells was assessed using the photometric
immunoassay `Cell Death Detection ELISA plus' kit. Primmorphs were treated
with Ca2+- and Mg2+-free artificial seawater containing
EDTA (Rottmann et al., 1987)
to facilitate dissociation into single cells. Five parallel determinations
from two primmorphs were performed for each experiment. As described by the
manufacturer (Roche), the cells were centrifuged (500 g; 5
min) and the pellets treated with lysis buffer (30 min; room temperature).
After collecting the nuclei by centrifugation (2000 g; 5 min),
the released nucleosomes were assayed using the one-step sandwich immunoassay.
The nucleosomes in the lysate were immobilized on streptavidin-coated wells
and the amount of nucleosomes determined using the peroxidase substrate
2,2'-azino-di-[3-ethylbenzthiazoline sulfonate]. The absorbance of the
colored product, i.e. immobilized nucleosomes, at 405 nm was determined and
corrected for the background absorbance at 480 nm. The values obtained were
correlated with the amount of DNA present in the samples before lysis. The
mean values and standard deviations were analyzed by paired Student's
t-test (Sachs, 1984
).
DNA concentration was determined by standard assay
(Kissane and Robins,
1958
).
Isolation of the S. domuncula RXR-related cDNA
clone
One complete cDNA, encoding the putative retinoid X receptor, was isolated
from a cDNA library obtained from S. domuncula
(Kruse et al., 1997) by
polymerase chain reaction (PCR). A forward primer was designed against the
conserved amino acids (aa) within the first zinc finger module, P-box, of RXR
and RXR-related receptors. The degenerate primer is located with the sponge
receptor between aa 7480 and reads: YKRAARAAIVHIGYRCAISC-3'
(Y=C,T; K=G,T; R=A,G; V=A,G,C; H=A,C,T; I=inosine). It was used in conjunction
with a vector-specific reverse primer. PCR was carried out using a GeneAmp
9600 thermal cycler (Perkin Elmer, Boston, MA, USA), with an initial
denaturation at 95°C for 3 min, then 35 amplification cycles each at
95°C for 30 s, 51°C for 45 s, 74°C for 1.5 min, and a final
extension step at 74°C for 10 min as described
(Ausubel et al., 1995
;
Pancer et al., 1997
). The
amplified product (approx. 0.8 kb) was purified and used for screening of the
library (Ausubel et al., 1995
).
The plasmid DNAs were sequenced using an automatic DNA sequenator (Li-Cor
4200; Lincoln, NE, USA). The sequence obtained was termed SDRXR and
was 1949 nt long, excluding the poly(A) tail.
Sequence analysis
The sequences were analyzed using computer programs BLAST (1997;
ncbi.nlm.nih.gov:
80/cgi-bin/BLAST/nph-blast) and FASTA (1997;
http://www.ebi.ac.uk/fasta33/).
Multiple alignments were performed with CLUSTAL W Version 1.6
(Thompson et al., 1994).
Phylogenetic trees were constructed on the basis of aa sequence alignments by
neighbour-joining, as implemented in the `Neighbor' program from the PHYLIP
package (Felsenstein, 1993
).
Distance matrices were calculated using the Dayhoff PAM matrix model as
described (Dayhoff et al.,
1978
). The degree of support for internal branches was further
assessed by bootstrapping (Felsenstein,
1993
). Graphic presentations were prepared using GeneDoc
(Nicholas and Nicholas,
2001
).
Northern blotting
Primmorphs were grown first in the non-attached state (for 5 days) and then
for 15 days on a galectin matrix with or without retinoic acid (150
µmol l1). RNA was then extracted and subjected to
northern blot analysis to determine the steady-state level of expression of
the following sponge genes: SDRXR, SDCASPR, SDLIM4 or the sponge
SDCYP4 gene (see below). The level of expression of the genes in
canal-forming primmorphs from S. domuncula was determined
semiquantitatively by northern blotting.
RNA was extracted from primmorphs pulverized in liquid nitrogen with TRIzol
Reagent (GibcoBRL, Grand Island, NY, USA). 5 µg of total RNA was then
electrophoresed through a 1% formaldehyde/agarose gel and blotted onto
Hybond-N+ nylon membrane following the manufacturer's instructions
(Amersham, Little Chalfont, Bucks, UK)
(Wiens et al., 1998).
Hybridization was performed with 400600 bp probes, derived from the
following S. domuncula cDNAs: SDRXR, encoding the putative
sponge retinoid X receptor; SDCASPR, encoding the caspase CASPR_SD
(accession number AJ426651); SDLIM4, encoding the LIM/homeobox
protein (accession number AJ493059) or SDCYP4 (Y17616), encoding the
putative CYP4_SD protein, a polypeptide belonging to the CYP4A (clofibrate)
subfamily cytochrome P-450 proteins. These probes were labeled with
the `PCR-DIG-Probe-Synthesis Kit' according to the manufacturer's
instructions. After washing, digoxygenin (DIG)-labeled nucleic acid was
detected with anti-DIG Fab fragments (conjugated to alkaline phosphatase;
diluted 1:10,000) and visualized by a chemiluminescence technique using
CDP-Star: disodium
2-chloro-5-(4-methoxyspiro{1,2-dioxetane-3,2'-(5'-chloro)-tricyclo[3.3.-1.13,7]decan}-4-yl)-1-phenyl
phosphate, the chemiluminescence substrate alkaline phosphatase, according to
the manufacturer's instructions (Roche).
Northern blot signals were quantitated by a chemiluminescence procedure
(Stanley and Kricka, 1990).
The screen was scanned with the GS-525 Molecular Imager (Bio-Rad,
München, Germany). The relative values for expression of the four genes
selected were correlated with the intensities of the bands measured in
controls (primmorphs not treated with retinoic acid).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
From these observations we conclude firstly that retinoic acid caused an involution in vivo and secondly, in consequence, that the respective receptor, the RXR, must be present. To verify the effect seen in vivo under controlled cell culture conditions, the potential morphogenetic effect of retinoic acid was studied in the in vitro primmorph system.
Induction of canals in primmorphs through attachment to the
substratum
Primmorphs are characterized by a compact smooth and `waxy' surface
(Fig. 1D). Recently, we
succeeded in inducing canal formation in primmorphs following adhesion to
substrata. We first used homologous recombinant galectin as the matrix
(Fig. 1E). Under these
conditions the primmorphs attached to the matrix and formed canal-like
structures, which could also be visualized in cross sections of the
primmorphs. Poly-L-lysine was also a suitable substrate for canal formation in
primmorphs (Fig. 1F). Fig. 1G shows the canal system
in the primmorphs at higher magnification. As outlined in Materials and
methods, canal formation generally starts approximately 10 days after transfer
of round-shaped, non-attached primmorphs (incubated for 5 days) to
galectin/poly-L-lysine-coated plates.
Effect of retinoic acid on canal formation in primmorphs
For these studies we used primmorphs that had already formed canals after
incubation on non-coated plates followed by 10 days of incubation on
galectin-coated plates. Retinoic acid was applied at 1 µmol
l1 (Fig.
1H,I) or 50 µmol l1
(Fig. 1JL) to determine
its effect on the integrity of canals.
The canals disappeared after 5 days of incubation with 1 µmol l1 retinoic acid in every one of the ten primmorphs studied (Fig. 1H). Additionally round shaped bodies of a size (<1 mm) smaller than the gemmules (diameter 13 mm) were released (Fig. 1I). After incubation with 50 µmol l1 retinoic acid the canals in the primmorphs disintegrated (Fig. 1J) and most of the primmorph tissue involuted (Fig. 1K,L). To determine whether the change in primmorph morphology caused by retinoic acid was reversible, the compound was washed out after a total incubation period of 20 days. Approximately 5 days later all the primmorphs studied started to form canals again (N=6; data not shown).
Effect of retinoic acid on cell viability
We determined whether the sponge primmorphs, cultivated on a
galectin-coated matrix for up to 15 days (after the initial incubation period
of 5 days in the non-attached state), and in the presence of retinoic acid,
undergo apoptotic cell death. We used a photometric immunoassay to quantitate
the amount of nucleosomes as outlined in Materials and methods. Untreated
primmorphs showed low apoptotic degradation of DNA, with absorbance values
(correlated to 50 ng of DNA) between 0.7 (day 3 after transfer to the galectin
matrix) and 1.3 (day 15; Fig.
2). Similar values were found for primmorphs that had been
transferred after 5 days to a galectin-coated matrix and were further
cultivated for 10 or 15 days, with retinoic acid (1 µmol
l1 or 50 µmol l1) during the last 2
days of incubation. In these primmorphs the amount of released nucleosomes
(measured as absorbance at 405 nm) varied non-significantly
(P>0.1) around values of 0.81.2, irrespective of the
retinoic acid concentration.
|
These data suggest that retinoic acid does not cause apoptosis, which features fragmentation of chromatin to nucleosomes.
Reversibility of the retinoic acid effect
The reversibility of the effects caused by retinoic acid on primmorphs was
tested by washout experiments. After the total incubation of 15 days the
primmorphs were transferred into the normal culture medium
(seawater/RPMI1640/silicate/Fe3+). Primmorphs restored their
original morphology regardless of whether they had been treated at low (1
µmol l1) or higher (50 µmol l1)
retinoic acid concentrations. In more than 80% of the 3D-cell aggregates
tested (N=12), the canal organization reappeared after 5 days (not
shown). We therefore conclude that the retinoic acid-mediated effect on canal
formation is reversible.
Cloning of the S. domuncula retinoid X
receptor
The complete clone for the sponge RXR cDNA, termed SDRXR, was
isolated using PCR and degenerate primers. The 1949 nt cDNA has one open
reading frame (ORF) with the start-methionine at nt 180182 and the stop
codon at nt 18241826. The 548-aa deduced protein has a calculated size
of 61346 (PC/GENE 1995; Data Banks CD-ROM; Release 14.0. Mountain View, CA,
USA: IntelliGenetics, Inc.). The instability index was computed as 65.16,
indicating an unstable protein. The putative protein, the S.
domuncula retinoid X receptor RXR_SUBDO, comprises two motifs with a high
significance score, the DNA-binding (hormone-receptor) domain including the
zinc finger (C4 type; two domains) modules (PFAM00105) between aa 153 and 228,
and the ligand-binding domain of nuclear hormone receptors [PFAM00104
(www.ncbi.nlm.nig.gov)]
spanning the protein sequence between aa 361 and 545
(Fig. 3). The nuclear receptor
signature, W-a-b[basic residue]-x-h[hydrophobic]-P-x-F-x-x-L-x-x-x-D-Q-x-x-L-L
(Wurtz et al., 1996), is
completely present in the sponge sequence between helices 3 and 4, aa
370389 (Fig. 3).
Northern blot studies revealed that the S. domuncula SDRXR transcript
is 2.0 kb in size (see below), indicating that the complete cDNA was isolated.
Furthermore, genomic DNA prepared from S. domuncula was used
successfully to perform Southern blots with SDRXR (not shown).
|
Sequence analysis of the sponge RXR
It is important to note that the two domains, the ligand-binding domain of
nuclear hormone receptors and the DNA-binding (zinc finger) domain, exist
exclusively in proteins from metazoans. For comparative analysis we list the
highest similarity score values `Expect value' (E; Blast-NCBI;
Coligan et al., 2000) of the
S. domuncula domains to sequences from Protostomia, Deuterostomia,
fungi and plants.
The ligand-binding domain for the nuclear hormone receptor: from two protostomians, the hormone receptor/zinc finger protein of Caenorhabditis elegans (NM_058702; E=1x1016) and the Drosophila melanogaster hepatocyte nuclear factor 4 (S36218; E=2x1024), the deuterostomian (human) retinoid X receptor, gamma (BC012063; E=9x1035) and, in comparison, the fungus glutathione reductase from Saccharomyces cerevisiae (D37871; E=3.6) as well as from the plant Arabidopsis thaliana an unknown protein (AY086748; E=5.0).
A likewise exclusively high relationship of the S. domuncula zinc finger (C4 type) domain is found to that in metazoan proteins; the scores are: for the C. elegans nuclear hormone receptor (NM_077038; E=8x1015), the D. melanogaster steroid receptor protein svp 2 (NM_079601; E=5x1027), the human retinoic acid receptor RXR, alpha (X52773; E=9x1027), as well as for the conserved hypothetical protein of the fungus N. crassa (AL390189; E=5.4) and the putative cellulose synthase catalytic subunit from A. thaliana (NM_128111; E=1.9).
Based on these comparisons it can be concluded that the sponge nuclear hormone receptor must be grouped with the other metazoan receptors of this class and that it shares highest similarity to the corresponding human sequences.
A more detailed analysis was elaborated for the ligand-binding domain of
nuclear hormone receptors (reviewed in
Giguère, 1999). This
part of the sequence is involved in ligand-binding, dimerization, interaction
with heat shock proteins, nuclear localization and transactivation. Based on
X-ray crystallographic experiments, the ligand-binding domain can be
subdivided into up to 13 helices that bury the ligand-binding site; these
helices H1H5 and H7H10 are shown in
Fig. 3. At the carboxy-terminal
end of the ligand-binding domain the sponge RXR also shows the activator
function-2 stretch, which is involved in the recruitment of coactivators
(Onate et al., 1998
).
The zinc finger (C4 type) domain represents the DNA-binding domain (see
Glass, 1994). It comprises two
zinc finger modules, which are encoded in the sponge RXR by 76 aa (or 92 aa,
according to Giguère,
1999
) (Fig. 3). The
four determinant cysteine residues exist at the same positions/distances as
those found in other nuclear hormone receptor proteins
(Giguère, 1999
)
(Fig. 3). The zinc finger
modules are further subdivided into subdomains that are involved in
recognition of the core half-site sequences (P-box;
Umesono and Evans, 1989
) and
the dimerization determinants (D-box;
Zechel et al., 1994
); also
their conserved residues are present in the sponge protein
(Fig. 3). The zinc finger
modules are followed by the carboxy-terminal extension stretch, which in
RXR_SUBDO comprises the consensus length. Between the DNA-binding (zinc
finger) domain and the ligand-binding domain is the hinge region which, in
comparison to other receptors of this group, is very variable. The function of
this region is to support dimerization
(Glass, 1994
). At the
aminoterminal end of the sponge RXR is the modulator domain, containing highly
variable amino acid residues
(Giguère, 1999
). Here
the transcriptional activation function (AF-1) resides.
Phylogenetic analysis of sponge RXR: total sequence
The sequence alignment of the sponge RXR, RXR_SUBDO, with the related
metazoan sequences in Fig. 3
also shows the nuclear receptor AmNR8 (from the coral Acropora
millepora). Based on the Expect value (E), sequences were
selected with the highest similarity between S. domuncula RXR and the
corresponding sequences from human, D. melanogaster and C.
elegans as well as from A. millepora. The overall protein
sequence similarity/identity of the sponge RXR to these metazoan sequences is
around 2035%. No nuclear receptor/nuclear hormone receptor molecule
other than in Metazoans has been found in, for example, plants or yeast (see
Grasso et al., 2001),
indicating that the sponge RXR is phylogenetically the oldest one known.
In 1999 a unified nomenclature system for the nuclear receptor superfamily
(Nuclear Receptors Nomenclature Committee;
NRNC, 1999) was proposed. In
accordance with this subfamily grouping, representative members from the
different groups were selected from the databases and analyzed. After
alignment an obvious classification of the S. domuncula protein
became possible based on an unrooted tree
(Fig. 4). The sponge
polypeptide does not belong to the nuclear receptors of the subfamilies 1, 5
or 6 but to subfamily 2 (retinoic acid receptors;
NRNC, 1999
;
Escriva et al., 2000
;
Chawla et al., 2001
). This
subfamily is subdivided again into six fractions: group 2A (e.g. human
hepatocyte nuclear factor 4), group 2B (e.g. human retinoic acid receptor-like
protein), group 2C (e.g. the human steroid receptor TR2-11), group 2D (e.g.
the insect THR6 nuclear receptor), group 2E (e.g. the photoreceptor-specific
nuclear receptor PNR) and group 2F (e.g. the human v-erbA related ear-2
receptor); Fig. 4. Other
receptors identified apart from those of Protostomia or Deuterostomia are from
the coral A. millepora (Phylum Cnidaria; Anthozoa), from which a
series of nuclear receptors has been cloned
(Grasso et al., 2001
); among
those the NR7 receptor falls into group 2F, the TLL receptor to group 2E,
while NR8 does not significantly belong to a known group. A further diploblast
receptor identified is the retinoic acid X receptor from Tripedalia
cystophora (Cnidaria; Cubozoa;
Kostrouch et al., 1998
); this
protein belongs to group 2B. The new, herein described sponge polypeptide,
which represents the phylogenetically oldest nuclear hormone receptor known to
date, falls in the branch of group 2A (with the human hepatocyte nuclear
factor) / group 2B (including the retinoid X receptors) with a high
significance (>98%). In general, the receptors of subfamily 2, comprising
the mentioned groups 2A2F, are adopted orphan receptors that bind with
low-affinity dietary lipids (Chawla et al.,
2001
).
|
Expression of selected genes in response to retinoic acid
incubation
Differential expression of the genes, in response to retinoic acid was seen
(Fig. 5). Expression of the
putative sponge retinoid X receptor at low retinoic acid concentration (1
µmol l1; Fig.
5, lane b) was 1.5-fold higher than that of the controls
(primmorphs cultivated in solution; Fig.
5, lane a); the size of the transcript was 2 kb. At higher
retinoic acid concentrations, of 3 µmol l1
(Fig. 5, lane c) and 50 µmol
l1 (Fig. 5,
lane d), expression levels increased even to 5- and 22- fold,
respectively.
|
A different expression pattern was seen for the caspase gene SDCASPR, where the steady-state expression remained almost unchanged even after the different retinoic acid treatments; the transcript size was 1.6 kb. In contrast, the expression of SDLIM4 (transcript size: 1.7 kb) the sponge LIM/homeobox protein, increased at 1 µmol l1 retinoic acid; higher concentrations of retinoic acid caused a further upregulation of the expression of this gene. A similar pattern as for the caspase gene was seen for the gene encoding the CYP4A related cytochrome P-450 protein, SDCYP4 (transcript size: 1.8 kb); no significant change in the expression pattern was seen in the primmorphs after treatment with different concentrations of retinoic acid (Fig. 5).
From these data we conclude that retinoic acid (at 150 µmol l1) causes a dose-dependent increase in the expression of the sponge retinoid X receptor gene and SDLIM4, the LIM/homeobox protein. The expression levels of the P-450 protein gene and the caspase gene remained unaffected.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Retinoic acid was previously found to trigger sponges to undergo involution
(gemmule or bud formation; see Simpson
1984). We show using the primmorph system that retinoic acid
causes both an involution of the intact S. domuncula organisms under
formation of the asexual reproduction bodies, the gemmules, and a similar
regression in vitro. Using retinoic acid concentrations that cause
similar effects in other sponge systems, especially in freshwater sponges
(Imsiecke et al., 1994
;
Nikko et al., 2001
). Sponges,
individuals and primmorphs, reacted to concentrations of 150 µmol
l1 retinoic acid by tissue regression.
In order to substantiate the observed effect of retinoic acid, a cDNA
library from the sponge S. domuncula was screened for the presence of
RARs and RXRs. So far we have been unable to detect an RAR in the library, but
the search for the RXR was successful. The full-length clone of a RXR was
identified, the deduced protein comprising the characteristic features of
other metazoan RXRs, e.g. the zinc finger (C4 type) domain and the activator
function-2 stretch, a segment that is important for the recruitment of
coactivators (see Results). The S. domuncula RXR showed high
similarity to the subfamily 2 of the nuclear receptor superfamily and, more
specificly, to group 2A/2B of these receptors. Hence this S.
domuncula RXR is phylogenetically the oldest nuclear hormone receptor
identified to date. The RXR from the Cnidarian jellyfish Tripedalia
cystophora (Kostrouch et al.,
1998), which until now has been the phylogenetically oldest one
identified, must be evolutionarily younger. With these data in hand it is now
possible to answer the question of the origin of the nuclear hormone receptors
and their possible existence in sponges
(Mendoza et al., 1999
). Our
data demonstrate that with the evolutionary transition to the Metazoa and with
the Porifera as the oldest metazoan phylum, the nuclear hormone receptors had
already appeared. This finding is amazing since sponges do not contain a
circulation vessel system that would allow an efficient transport of hormones,
e.g. steroids, thyroid hormones or ecdysones.
The data, however, clearly demonstrate the existence of the RXR gene in
S. domuncula. In addition, the data show that retinoic acid exerts a
morphogenetic effect on this sponge both in vivo and in
vitro. In order to investigate the most likely mode of action of retinoic
acid/RXR at the subcellular level, gene expression studies were performed
using the northern blot technique. It was found that retinoic acid causes a
strong upregulation of the expression of the RXR gene, a finding that
is in accordance with the related receptors present in triploblasts; in these
animals a tissue- and stage-dependent expression of RXRs is known (see
Rana et al., 2002). Since this
effect is observed in S. domuncula with all-trans retinoic
acid we assume that the morphogen is isomerized in the animal to
9-cis-retinoic acid, the characteristic ligand for RXRs.
Further genes were analyzed in parallel with the expression studies of RXR
in the in vitro system of S. domuncula. A caspase gene was
included, since retinoic acid is known to cause apoptosis in some systems,
e.g. leukemia cells (Nervi et al.,
1998). In a recent study it was established that in the sponge
system caspases are also involved in the induction of apoptosis
(Wiens et al., 2003
).
Surprisingly it was found that in the presence of retinoic acid the expression
level of the caspase gene does not alter, and this finding was also supported
by a series of experiments which showed that there is no parallel increase in
the formation of free nucleosomes during the retinoic acid-mediated regression
of tissue.
Another gene involved in detoxification of drugs
(Nebert et al., 1991;
Chawla et al., 2001
), the
sponge CYP4_SD gene, was selected together with SDLIM4
(encoding LIM/homeobox protein), which mediates differentiation events. Using
the northern blot approach, it could be demonstrated that the steady-state
expression of the caspase gene remained unchanged, irrespective of the
retinoic acid concentration used. However, a strong increase in expression of
the SDLIM4 gene was observed. In triploblasts, both in Protostomians
and in Deuterostomians, the Lim-class homeodomain proteins are involved in
organogenesis (see Bach, 2000
).
The Lim-class HD proteins still await identification in Cnidaria, but the
finding presented in this report is in agreement with our recent observation
that the LIM/homeobox protein gene is upregulated after induction of
canal-like structures in primmorphs (Wiens
et al.
, in press) and emphasises that the LIM/homeobox protein,
like the Paired-class homeobox genes, is involved in the control of the
patterning (Galliot and Miller
2000
). Likewise, elucidation of the role of the HOX-like molecules
(e.g. Degnan et al., 1995
) in
sponges will provide further clues on the regulation of the bodyplan formation
in these animals.
The results presented in this paper suggest that RXR(s), retinoic acid, and its derivatives may play important roles during morphogenesis in sponges. Identifying the proteins/receptors which `transduce/facilitate' the retinoic acid signal to the transcription factor(s), and in turn are involved in the cytological/morphological changes, will be the next major steps forward in the understanding of the molecular biology of retinoid responses in the earliest metazoan phylum, which branched off from the common animal ancestor, the Urmetazoa.
![]() |
Acknowledgments |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Aranda, A. and Pascual, A. (2001). Nuclear
hormone receptors and gene expression. Physiol. Rev.
81,1269
-1304.
Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Smith, J. A., Seidmann, J. G. and Struhl, K. (1995). Current Protocols in Molecular Biology. New York: John Wiley and Sons.
Bach, I. (2000). The Lim domain: regulation by association. Mech. Dev. 91, 5-17.[CrossRef][Medline]
Biesalski, H. K., Doepner, G., Tzimas, G., Gamulin, V., Schröder, H. C., Batel, R., Nau, H. and Müller, W. E. G. (1992). Modulation of myb gene expression in sponges by retinoic acid. Oncogene 7,1765 -1774.[Medline]
Böhm, M., Hentschel, U., Friedrich, A., Fieseler, L., Steffen, R., Gamulin, V., Müller, I. M. and Müller, W. E. G. (2001). Molecular response of the sponge Suberites domuncula to bacterial infection. Mar. Biol. 139,1037 -1045.[CrossRef]
Cha, J. N., Shimizu, K., Zhou, Y., Christianssen, S. C.,
Chmelka, B. F., Stucky, G. D. and Morse, D. E. (1999).
Silicatein filaments and subunits from a marine sponge direct the
polymerization of silica and silicones in vitro. Proc. Natl. Acad.
Sci. USA 96,361
-365.
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.
Coligan, J. E., Dunn, B. M., Ploegh, H. L., Speicher, D. W. and Wingfield, P. T. (2000). Current Protocols in Protein Science. Chichester: John Wiley & Sons.
Dayhoff, M. O., Schwartz, R. M. and Orcutt, B. C. (1978). A model of evolutionary change in protein. In Atlas of Protein Sequence and Structure (ed. M. O. Dayhoff), pp. 345-352. Washington DC: Nat. Biomed. Res. Foundation.
Degnan, B. M., Degnan, S. M., Giusti, A. and Morse, D. E. (1995). A hox/hom homeobox gene in sponges. Gene 155,175 -177.[CrossRef][Medline]
Egea, P. F., Mitschler, A., Rochel, N., Ruff, M., Chambon, P.
and Moras, D. (2000). Crystal structure of the human
RXR ligand-binding domain bound to its natural ligand: 9-cis retinoic
acid. EMBO J. 19,2592
-2601.
Escriva, H., Delaunay, F. and Laudet, V. (2000). Ligand binding and nuclear receptor evolution. BioEssays 22,717 -727.[CrossRef][Medline]
Felsenstein, J. (1993). PHYLIP, version 3.5. University of Washington, Seattle.
Galliot, B. and Miller, D. (2000). Origin of anterior patterning how old is our head? Trends Genet. 16,1 -5.[CrossRef][Medline]
Giguère, V. (1999). Orphan nuclear
receptors: from gene to function. Endocrinol. Rev.
20,689
-725.
Glass, C. K. (1994). Differential recognition of target genes by nuclear receptors monomers, dimers, and heterodimers. Endocrinol. Rev. 15,391 -407.[Medline]
Goksøyr, A. and Förlin, L. (1992). The cytochrome P-450 system in fish, aquatic toxicology and environmental monitoring. Aquat. Toxicol. 22,287 -312.[CrossRef]
Grasso, L. C., Hayward, D. C., Trueman, J. W. H., Hardie, K. M., Janssens, P. A. and Ball, E. E. (2001). The evolution of nuclear receptors: evidence from the coral Acropora. Mol. Phyl. Evol. 21,93 -102.[CrossRef][Medline]
Heery, D. M., Zacharewski, T., Pierrat, B., Gronemeyer, H. and Chambon, P. (1993). Efficient transactivation by retinoic acid receptors in yeast requires retinoid X receptors. Proc. Natl. Acad. Sci. USA 90,4281 -4285.[Abstract]
Herland-Meewis, H. (1948). La gemmulation chez Suberites domuncula (Olivi) Nardo. Arch. Anat. Microsc. 37,289 -322.
Imsiecke, G., Borojevic, R. and Müller, W. E. G. (1994). Retinoic acid acts as a morphogen in freshwater sponges. Invert. Reprod. Dev. 26,89 -98.
Kissane, J. M. and Robins, E. (1958). The
fluorometric measurement of deoxyribonucleic acid in animal tissue with
special reference to the central nervous system. J. Biol.
Chem. 233,184
-188.
Kostrouch, Z., Kostrouchova, M., Love, W., Jannini, E.,
Piatigorsky, J. and Rall, J. E. (1998). Retinoic acid X
receptor in the diploblast Tripedalia cystophora. Proc. Natl. Acad.
Sci. USA 95,13442
-13447.
Koziol, C., Borojevic, R., Steffen, R. and Müller, W. E. G. (1998). Sponges (Porifera) model systems to study the shift from immortal- to senescent somatic cells: the telomerase activity in somatic cells. Mech. Ageing Dev. 100,107 -120.[CrossRef][Medline]
Krasko, A., Batel, R., Schröder, H. C., Müller, I. M.
and Müller, W. E. G. (2000). Expression of silicatein
and collagen genes in the marine sponge Suberites domuncula is
controlled by silicate and myotrophin. Eur. J.
Biochem. 267,4878
-4887.
Krasko, A., Schröder, H. C., Batel, R., Grebenjuk, V. A., Steffen, R., Müller, I. M. and Müller, W. E. G. (2002). Iron induces proliferation and morphogenesis in primmorphs from the marine sponge Suberites domuncula. DNA Cell Biol. 21,67 -80.[CrossRef][Medline]
Kruse, M., Müller, I. M. and Müller, W. E. G. (1997). Early evolution of metazoan serine/threonine- and tyrosine kinases: identification of selected kinases in marine sponges. Mol. Biol. Evol. 14,1326 -1334.[Abstract]
Le Pennec, G., Perovic, S., Ammar, M. S. A., Grebenjuk, V. A., Steffen, R., Brümmer F. and Müller, W. E. G. (2003). Cultivation of primmorphs from the marine sponge Suberites domuncula: morphogenetic potential of silicon and iron. A review. Mar. Biotechnol. 100,93 -108.
Mangelsdorf, D. J., Thummel, C., Beato, M., Herrlich, P., Schütz, G., Umesono, K., Kastner, P., Mark, M., Chambon, P. and Evans, R. M. (1995). The nuclear receptor superfamily: the second decade. Cell 83,835 -839.[Medline]
Mendoza, R. L., Escriva, H., Vanacker, J. M., Bouton, D., DeLannoy, S., Pierce, R. and Laudet, V. (1999). Nuclear hormone receptors and evolution. Amer. Zool. 39,704 -713.
Morris-Kay, G. M. (1997). Retinoids in mammalian embryonic development. In Retinoids: Their Physiological Function and Therapeutic Potential (ed. G. V. Sherbet), pp.79 -93. Greenwich CT: JAI Press.
Müller, W. E. G. (1982). Cell membranes in sponges. Int. Rev. Cytol. 77,129 -181.
Müller, W. E. G. (1997). Origin of metazoan adhesion molecules and adhesion receptors as deduced from their cDNA analyses from the marine sponge Geodia cydonium. Cell Tissue Res. 289,383 -395.[CrossRef][Medline]
Müller, W. E. G. (2001). How was metazoan threshold crossed: the hypothetical Urmetazoa. Comp. Biochem. Physiol. 129A,433 -460.
Müller, W. E. G. (2002). Telomerase activity in sponges (Porifera) the closest related taxa of the hypothetical ancestral animal the Urmetazoa. In Telomeres and Telomerases: Cancer and Biology (ed. G. Krupp and R. Parwaresch), pp.300 -313. New York: Eurekah Pub.
Müller, W. E. G., Müller, I. M. and Gamulin, V. (1994). On the monophyletic evolution of the Metazoa. Brazil. J. Med. Biol. Res. 27,2083 -2096.
Müller, W. E. G., Wiens, M., Batel, R., Steffen, R., Borojevic, R. and Custodio, M. R. (1999). Establishment of a primary cell culture from a sponge: primmorphs from Suberites domuncula.Mar. Ecol. Progr. Ser. 178,205 -219.
Müller, W. E. G., Schröder, H. C., Skorokhod, A., Bünz, C., Müller, I. M. and Grebenjuk, V. A. (2001). Contribution of sponge genes to unravel the genome of the hypothetical ancestor of Metazoa (Urmetazoa). Gene 276,161 -173.[CrossRef][Medline]
Müller, W. E. G., Krasko, A., Skorokhod, A., Bünz, C., Grebenjuk, V. A., Steffen, R., Batel, R., Müller, I. M. and Schröder, H. C. (2002). Histocompatibility reaction in the sponge Suberites domuncula on tissue and cellular level: central role of the allograft inflammatory factor 1. Immunogenetics 54,48 -58.[CrossRef][Medline]
Nebert, D. W., Nelson, D. R., Coon, M. J., Estabrook, R. W., Feyereisen, R., Fujii-Kuriyama, Y., Gonzalez, F., Guengerich, F. P., Gunsalus, I. C., Johnson, E. F., Loper, J. C., Sato, R., Waterman, M. R. and Waxman, D. J. (1991). The P-450 superfamily: update on new sequence, gene mapping, and recommended nomenclature. DNA Cell Biol. 10,1 -4.[Medline]
Nervi, C., Ferrara, F. F., Fanelli, M., Rippo, M. R., Tomassini,
B., Ferrucci, P. F., Ruthardt, M., Gelmetti, V., Gambacorti-Passerini, C.,
Diverio, D., Grignani, F., Pelicci, P. G. and Testi, R.
(1998). Caspases mediate retinoic acid-induced degradation and
the acute promyelocytic leukemia PML/RAR fusion protein.
Blood 92,2244
-2251.
Nicholas, K. B. and Nicholas, H. B., Jr (2001). GeneDoc: a tool for editing and annotating multiple sequence alignments. Version 1.1.004. http://www.psc.edu/biomed/genedoc/gddl.htm; INTERNET.
Nikko, E., Van de Vyver, G. and Richelle-Maurer, E. (2001). Retinoic acid down-regulates the expression of the EmH-3 homeobox-containing gene in the freshwater sponge Ephydatia muelleri. Mech. Ageing Dev. 122,779 -794.[CrossRef][Medline]
NRNC (1999). A unified nomenclature system for the nuclear receptor superfamily. Cell 97,161 -163.[Medline]
Onate, S. A., Boonyaratanakornkit, V., Spencer, T. E., Tsai, S.
Y., Tsai, M. J., Edwards, D. P. and O'Malley, B. W. (1998).
The steroid receptor coactivator-1 contains multiple receptors interacting and
activation domains that cooperatively enhance the activation function 1 (AF1)
and AF2 domains of steroid receptors. Proc. Natl. Acad. Sci.
USA 93,10069
-10073.
Osburn, D. L., Shao, G., Seidel, M. and Schulman, I. G.
(2001). Ligand-dependent degradation of the retinoid X receptors
does not require transcriptional activity or coactivator interactions.
Mol. Cell. Biol. 21,4909
-4918.
Pancer, Z., Kruse, M., Müller, I. M. and Müller, W. E. G. (1997). On the origin of adhesion receptors of Metazoa: cloning of the integrin alpha subunit cDNA from the sponge Geodia cydonium. Mol. Biol. Evol. 14,391 -398.[Abstract]
Rana, B., Veal, G. J., Pearson, A. D. J. and Redfern, P. F. (2002). Retinoid X receptors and retinoid response in neuroblastoma cells. J. Cell Biochem. 86, 67-78.[CrossRef][Medline]
Rasmont, R. (1963). Le rôle de la taille et de la nutrition dans le déterminisme de la gemmulation chez les spongillides. Dev. Biol. 8, 243-271.
Rasmont, R. (1970). Some new aspects of the physiology of fresh-water sponges. In The Biology of the Porifera. Symp. Zool. Soc. Lond. 25 (ed. W. G. Fry), 415-422. London: Zoological Society.
Redfern, C. P. F. (1997). Molecular mechanisms of the retinoid function. In Retinoids: Their Physiological Function and Therapeutic Potential (ed. G. Sherbet), pp.35 -77. Greenwich CT: JAI Press.
Rottmann, M., Schröder, H. C., Gramzow, M., Renneisen, K., Kurelec, B., Dorn, A., Friese, U. and Müller, W. E. G. (1987). Specific phosphorylation of proteins in pore complex-laminae from the sponge Geodia cydonium by the homologous aggregation factor and phorbol ester. Role of protein kinase C in the phosphorylation of DNA topoisomerase II. EMBO J. 6,3939 -3944.[Abstract]
Sachs, L. (1984). Angewandte Statistik. Springer-Verlag, Berlin.
Sato, A., Imaizumi, M., Saito, T., Yoshinari, M., Suzuki, H., Hayashi, Y. and Iinuma, K. (1999). Detection of apoptosis in acute promyelocytic leukemia cells in vivo during differentiation-induction with all-trans retinoic acid in combination with chemotherapy. Leukemia Res. 23,827 -832.[CrossRef][Medline]
Schäcke, H., Schröder, H. C., Gamulin, V., Rinkevich, B., Müller, I. M. and Müller, W. E. G. (1994). Molecular cloning of a tyrosine kinase gene from the marine sponge Geodia cydonium: a new member belonging to the receptor tyrosine kinase class II family. Mol. Membr. Biol. 11,101 -107.[Medline]
Schröder, H. C., Krasko, A., Batel, R., Skorokhod, A.,
Pahler, S., Kruse, M., Müller, I. M. and Müller, W. E. G.
(2000). Stimulation of protein (collagen) synthesis in sponge
cells by a cardiac myotrophin-related molecule from Suberites domuncula.FASEB J. 14,2022
-2031.
Simpson, T. L. (1984). The Cell Biology of Sponges. New York, NY: Springer-Verlag.
Simpson, T. L. and Gilbert, J. J. (1973). Gemmulation, gemmule hatching and sexual reproduction in fresh-water sponges. I. The life cycle of Spongilla lacustris and Tubella pennsylvanica. Trans. Am. Microsc. Soc. 92,422 -433.
Stanley, P. E. and Kricka, L. J. (1990). Bioluminescence and Chemiluminescence: Current Status. New York, NY: John Wiley & Sons.
Thompson, J. D., Higgins, D. G. and Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positions-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22,4673 -4680.[Abstract]
Umesono, K. and Evans, R. M. (1989). Determinants of target gene specificity for steroid/thyroid hormone receptors. Cell 57,1139 -1146.[Medline]
Uriz, M. J., Turon, X. and Becerrro, M. A. (2000). Silica deposition in demosponges: spiculogenesis in Crambe crambe. Cell Tissue Res. 301,299 -309.[CrossRef][Medline]
Weissenfels, N. (1989). Biologie und Mikroskopische Anatomie der Süßwasserschwämme (Spongillidae). Stuttgart: Gustav Fischer-Verlag.
Wiens, M., Koziol, C., Hassanein, H. M. A., Batel, R. and Müller, W. E. G. (1998). Expression of the chaperones 14-3-3 and HSP70 induced by PCB 118 2,3',4,4',5-pentachlorobiphenyl in the marine sponge Geodia cydonium. Mar. Ecol. Prog. Ser. 165,247 -257.
Wiens, M., Luckas, B., Ammar, M. S. A., Steffen, R., Batel, R., Diehl-Seifert, B., Schröder, H. C. and Müller, W. E. G. (2003). Okadaic acid: a potential defense toxin for the sponges Suberites domuncula. Mar. Biol. 142,213 -223.
Wiens, M., Mangoni, A., D'Esposito, M., Fattorusso, E., Korchagina, N., Schröder, H. C., Grebenjuk, V. A, Krasko, A., Batel, R., Müller, I. M. and Müller, W. E. G. (in press). The molecular basis for the evolution of the metazoan bodyplan: extracellular matrix-mediated morphogenesis in marine demosponges. J. Mol. Evol., in press.
Wimmer, W., Blumbach, B., Diehl-Seifert, B., Koziol, C., Batel, R., Steffen, R., Müller, I. M. and Müller, W. E. G. (1999a). Increased expression of integrin and receptor tyrosine kinase genes during autograft fusion in the sponge Geodia cydonium.Cell Adhes. Commun. 7,111 -124.[Medline]
Wimmer, W., Perovic, S., Kruse, M., Krasko, A., Batel, R. and
Müller, W. E. G. (1999b). Origin of the
integrin-mediated signal transduction: functional studies with cell cultures
from the sponge Suberites domuncula. Eur. J. Biochem.
260,156
-165.
Wurtz, J. M., Bourget, W., Renaud, J. P., Vivat, V., Chambon, P., Moras, D. and Gronemeyer, H. (1996). A canonical structure for the ligand-binding domain of nuclear receptors. Nature Struct. Biol. 3,87 -94.[Medline]
Zechel, C., Shen, X. Q., Chambon, P. and Gronemeyer, H. (1994). Dimerization interfaces formed between the DNA binding domains determine the cooperative binding of RXR/RAR and RXR/TR heterodimers to DR5 and DR4 elements. EMBO J. 13,1414 -1424.[Abstract]