Department of Genetics, School of Medicine,
Case Western Reserve University, Cleveland, Ohio 44106-4955
Received 7 October 1997/Returned for modification 24 November
1997/Accepted 30 January 1998
During the transition from the maternal to the zygotic
developmental program, the expression of genes important for pattern formation or cell cycle regulation changes dramatically. Rapid changes
in gene expression are achieved in part through the control of mRNA
stability. This report focuses on bicoid, a gene essential for formation of anterior embryonic structures in Drosophila
melanogaster. bicoid mRNA is synthesized exclusively
during oogenesis. Here, we show that bicoid mRNA stability
is regulated. While bicoid mRNA is stable in retained
oocytes, in unfertilized eggs, and during the first 2 h of
embryogenesis, specific degradation is activated at cellularization of
the blastoderm. To identify cis-acting sequences required
for bicoid mRNA's regulated stability, fusions between
bicoid and genes producing stable mRNAs were introduced into the Drosophila germ line by P-element-mediated
transformation. The analysis of the fusion mRNAs identified a
bicoid instability element (BIE) contained within a
43-nucleotide sequence immediately following the stop codon. The BIE is
sufficient to destabilize the otherwise-stable ribosomal protein A1
mRNA and is separable from the previously identified bicoid
mRNA localization signals and from the "nanos response
element." Similar mechanisms may regulate a class of developmentally
important maternal genes whose mRNA has a temporal profile similar to
that of bicoid.
 |
INTRODUCTION |
The degradation of maternally
encoded mRNAs at the transition from maternal to zygotic expression is
a common occurrence in animal development (4, 13, 14, 29,
36). In Drosophila melanogaster, the degradation of
two maternal mRNAs, string and twine, which
encode Cdc25 phosphatases, is essential for embryonic development
(13). Many other genes important for embryonic development are similarly regulated (33). One of them is
bicoid (16), a gene essential for the
morphogenesis of the anterior half of the embryo.
Transcription of bicoid is strictly maternal; the mRNA is
synthesized in the ovary by the nurse cells and then transported to the
oocyte (6). bicoid mRNA localizes to the anterior
ends of oocytes and early embryos by a mechanism involving multiple steps (31). Anterior localization is mediated by a
625-nucleotide sequence in the bicoid untranslated region
(UTR) (18). After egg deposition, bicoid mRNA is
translated (10) and the unstable Bicoid protein forms a
concentration gradient decreasing from the anterior tip of the embryo.
Different protein concentration thresholds activate different gap genes
along the anterior half of the embryo (11, 12). By the
beginning of gastrulation, bicoid mRNA has been completely
degraded. Thus, bicoid function depends on at least five
posttranscriptional processes: (i) mRNA localization at the anterior
pole of the oocyte, (ii) translational repression during oogenesis,
(iii) diffusion of the protein from a localized source, (iv)
instability of the protein, and (v) destabilization of
bicoid mRNA. This report focuses on the regulation of
bicoid mRNA stability.
Previously, Berleth et al. (6) had determined that
bicoid mRNA is absent from embryos over 4 h old.
However, it is unknown how degradation of bicoid mRNA is
controlled. It is possible that fertilization triggers the
destabilization and gradual degradation of bicoid mRNA.
Alternatively, the message may be initially stable and then rapidly
degraded before the onset of gastrulation. Moreover, the sequences that
regulate bicoid mRNA stability are unknown. In the present
report we consider the following questions. Is bicoid mRNA
constitutively unstable, or is its stability regulated? If regulated,
is bicoid mRNA degradation activated at fertilization or
later in embryogenesis? What cis-acting elements mediate the selective destabilization of bicoid mRNA? Surprisingly, we
found that bicoid mRNA is stable even after egg activation
and fertilization and that its degradation is activated during
embryogenesis, at cellularization of the blastoderm. Moreover, by
assaying different hybrid genes in embryos from transgenic flies, we
defined a cis-acting bicoid instability element
(BIE) that is both necessary and sufficient for regulated message
decay.
| |
MATERIALS AND METHODS |
Construction of plasmids.
The genomic bicoid
8.7-kb EcoRI fragment was obtained from the plasmid p902
(provided by P. Macdonald). The relevant parts of the bicoid
sequence used in this study are numbered as in reference 6. The ribosomal protein A1 gene (rpA1) 2.4-kb
BamHI fragment and the 
1-tubulin gene 4.5-kb
HindIII fragment were obtained from the plasmids p5D
(24) and pDmt-1 (17), respectively.
Construction of the bicoid-rpA1 and
bicoid-1-tubulin hybrid genes. (i) 5B
(5'bicoid/3'rpA1).
To construct the chimeric
gene 5B (5'bicoid/3'rpA1) the vector pGEM3 (Promega) was
modified as follows. A linker containing the NotI site was
inserted to replace the HindIII site (modified vector
kindly provided by A. Riedl). The SalI site of pGEM3 was removed by linearization of the plasmid with SalI and
filling in of the restriction site with T4 DNA polymerase, followed by ligation. The 2.4-kb BamHI fragment from the p5D plasmid was
subcloned into the modified pGEM3 vector with the NotI site
in the 5' direction from the rpA1 gene to yield pGEM3rpA1. A
bicoid 1.5-kb PstI-SalI fragment
obtained from p902 by partial digestion was ligated to pGEM3rpA1, which
had been digested with PstI-SalI and
dephosphorylated. The resulting 2.8-kb insert contained 1,519 bp of the
bicoid gene, from the cap site to the SalI site
(bicoid positions 1245 to 2764), and a 1.3-kb rpA1 sequence
from the unique SalI site within the protein-coding region
to the end of the rpA1 mRNA plus 3' flanking sequences. The HA1 DNA
cassette encoding the HA1 epitope YPYDVPDYA from the influenza virus
hemagglutinin HA1 protein (32) was inserted at the
SalI junction in the aforementioned plasmid by digesting the
plasmid with SalI, filling in the site with T4 DNA polymerase, and ligating the ends to a 55-bp
StuI-ScaI DNA fragment encoding the HA1 epitope.
This plasmid was designated pGEM-5B (Fig.
1). The insertion of the DNA cassette
fuses in frame the protein-coding sequences of bicoid and
the rpA1 gene. The NotI-BamHI fragment containing
the 5'bicoid/3'rpA1 hybrid was subcloned into the
transformation vector CaSpeR4/GERM4 to yield CaSpeR-5B. CaSpeR4/GERM4 was constructed as follows: the fragment
NotI-EcoRI, from the plasmid pGERM4 (kindly
provided by K. Cheung and R. Cohen), which contained a nurse
cell-specific enhancer from the heat shock gene hsp26 (388 bp) and the sgs3 promoter from bp 
127 to +33, was subcloned into CaSpeR4.
View larger version (17K):
[in this window]
[in a new window]
|
FIG. 1.
Constructs used to localize the BIE. The structure of
the bicoid gene is drawn from the work of Berleth et al.
(6). 5B contains a hybrid between bicoid 5'
sequences and rpA1 gene 3' sequences. The glue gene sgs3
promoter containing the nurse cell-specific hsp26 enhancer
was used to drive the expression of this construct. 3B contains a
hybrid between rpA1 gene 5' sequences and bicoid 3'
sequences. This hybrid gene is under the control of the rpA1 promoter.
5T contains a hybrid between 1-tubulin gene 5' sequences and
bicoid 3' sequences. The 1-tubulin promoter was used for
this construct. In the BBT construct, the bicoid 3' UTR
sequence downstream of the MluI site was replaced by that of
the tubulin gene 3' UTR. The bicoid promoter was used for
this construct (a kind gift from P. Macdonald [18]).
3BA, 3BB, 3BC, and 3BD are four overlapping deletion constructs of the
3B construct. The  NRE construct contains a bicoid gene
with a deletion of 45 nucleotides in the 3' UTR between the
HpaI and EcoRV restriction sites, downstream of
the stop codon (a kind gift from R. Wharton [34]).
Horizontal lines, nontranscribed flanking regions;  s, introns,  ,
bicoid protein-coding region;  , bicoid UTRs;
 , rpA1 protein-coding region;  , rpA1 gene
UTRs;  , tubulin protein-coding region;  , tubulin gene UTRs. A DNA
cassette encoding the influenza virus HA1 hemagglutinin epitope was
inserted (not shown [32]) to maintain the protein
reading frame in constructs 5B, 3B, 5T, 3BA, 3BB, 3BC, and 3BD.
|
|
(ii) 3B (5'rpA1/3'bicoid).
The pGEM3rpA1 plasmid
was digested with SalI and EcoRI to remove the 3'
end of the gene. The resulting DNA was dephosphorylated and ligated to
a bicoid 3.0-kb SalI-EcoRI fragment
from p902, yielding pGEM-3B (Fig. 1). The 4.1-kb insert from pGEM-3B
contains 1.1 kb of 5' rpA1 gene sequences, including the promoter and
coding region up to the SalI site. The remaining 3.0 kb
contain the 3'-most 2,044 bp of the bicoid gene, from the
SalI site (bicoid position 2765) to the
polyadenylation site plus 3' flanking sequences extending to the 3'
EcoRI site from the bicoid 8.7-kb
EcoRI genomic fragment. The HA1 DNA cassette was inserted at
the SalI junction of the aforementioned plasmid by digesting
the plasmid with SalI, filling in the site with T4 DNA
polymerase, and ligating the ends to a 69-bp
NaeI-SmaI fragment containing the sequence
encoding the HA1 epitope (32). The insertion of the HA1 DNA
cassette fuses in frame the protein-coding sequences of the rpA1 gene
and bicoid. The NotI-EcoRI fragment
containing the 5'rpA1/3'bicoid hybrid was subcloned into
CaSpeR4 to yield CaSpeR-3B.
(iii) 5T (5'1-tub/3'bicoid).
To
construct the chimeric gene 5T
(5'/1-tub/3'bicoid), the vector pBluescript II
KS+ (pBSIIKS+; Stratagene) was modified as follows. The
BamHI-PstI portion in the polylinker of pBSIIKS+ was removed by digestion with PstI, treatment with nuclease
S1, digestion with BamHI, filling with T4 DNA polymerase,
and ligation. The 4.5-kb HindIII fragment from pDm
t-1 was subcloned into modified pBSIIKS+ to yield
pBSIIKS+1-tub. The bicoid 3.0-kb
SalI-EcoRI fragment obtained from p902 and a
125-bp BamHI-SalI fragment containing the
sequence encoding the HA1 epitope (32) were subcloned by a
three-fragment ligation into pBSIIKS+1-tub digested with
BamHI and EcoRI and dephosphorylated, yielding
pBSII-5T (Fig. 1). The resulting 5.6-kb insert in pBSIIKS+ contains 5'
sequences of the tubulin gene, including the promoter and the first 750 nucleotides of the tubulin mRNA sequence, joined with the 3'-most 2,044 bp of the bicoid gene, from the SalI site to the
polyadenylation site plus the 3' flanking sequences (bicoid
position 2765 up to the EcoRI site). The insertion of the
HA1 DNA cassette fuses in frame the protein coding sequences of
1-tub and bicoid. The
XhoI-NotI fragment containing the
5'1-tub/3'bicoid hybrid was subcloned into
CaSpeR4 to yield CaSpeR4-5T.
The DNA sequences of the blunt-end junctions and the insertion sites of
the HA1 DNA cassettes were all checked by DNA sequencing with nearby
primers.
Construction of the 5' rpA1/3' bicoid deletion
derivatives.
Four DNA fragments were prepared by PCR with pGEM-3B
as a template and the primers listed in Table
1. These PCR products were digested with
NheI and MluI, and the resulting fragments were
inserted into the plasmid pGEM-3B, which had been digested with
NheI and MluI and dephosphorylated, to yield four
pGEM-5'rpA1/3'bicoid plasmids with various deletions in
the bicoid sequence: pGEM-3BA (deletion at positions 2765 to
4006), -3BB (deletion at positions 4007 to 4104), -3BC (deletion at
positions 3137 to 4104), and -3BD (deletion at positions 2765 to 3136)
(Fig. 1).
The parts inserted by PCR in pGEM-5'rpA1/3'bicoid
plasmids were verified by DNA sequencing. The
NotI-EcoRI fragments containing the
5'rpA1/3'bicoid hybrids were subcloned into CaSpeR4 to
yield CaSpeR-3BA, -3BB, -3BC, and -3BD.
rpA1-BIEsense and rpA1-BIEantisense.
To distinguish the
BIE-containing rpA1 gene from the endogenous rpA1 gene, a fragment of
500 nucleotides from the human transferrin receptor 3' UTR, located
upstream of the iron regulatory sequence (28), was made by
PCR and inserted at the NgoMI site of the 3' UTR of the rpA1
gene, which was cloned in pGEM3. Then, the 5' NgoMI site was
destroyed by partial digestion and filling in. Two complementary
oligonucleotides, the sense BIE (BIES) and the antisense BIE (BIEA)
(Table 1), were annealed, and the resulting fragment was inserted at
the remaining NgoMI site.
The rpA1 gene (a 2.4-kb BamHI fragment described in
reference 24), containing the 0.5-kb transferrin
fragment and the 43-nucleotide BIE in either orientation, was subcloned
into CaSpeR4 to yield CaSpeR-BIES and CaSpeR-BIEA.
P-element-mediated transformation and establishment of
transformed lines.
CaSpeR-5B, -3B, -5T, -3BA, -3BB, -3BC, -3BD,
-BIES, and -BIEA constructs (500 µg/ml) were coinjected with phs-
helper plasmid DNA (100 µg/ml) (30) into yellow
white mutant embryos (23, 26). Transformants were
selected by mating G0 adults with yellow white
flies and screening for flies having orange eyes. Homozygous stocks
were generated for all the transformant lines used in this work.
Transformant lines with BBT and with the nanos response element deleted (NRE transformants) were kind gifts from P. Macdonald and R. Wharton, respectively (18, 34).
RNA analysis.
Ovaries were taken from virgin females
deprived of yeast for 4 to 12 days (to prevent egg laying) and from
well-fed, mated females. All dissections were done in Ringer's
solution (3). Embryos were collected for 1 h from well-fed
females on agar-molasses plates and left to age at 25°C for different
lengths of time. The synchrony of the collections was checked by
staging an aliquot of dechorionated embryos. The percentage of
unfertilized eggs was determined, and only collections containing less
than 5% unfertilized eggs were analyzed. A low proportion of
unfertilized eggs is important for quantitative analysis of
bicoid mRNA decay because bicoid mRNA is stable
in laid, unfertilized eggs (see Results). Ovaries and embryos were kept
at 80°C after being frozen in liquid nitrogen.
Each sample of ovaries and embryos was homogenized in 1× binding
buffer (0.5 M NaCl, 10 mM Tris-HCl [pH 8.0], 1 mM EDTA, 1% sodium
dodecylsulfate) containing proteinase K (100 mg/ml) and incubated at
55°C for 2 h. After centrifugation for 10 min at 16,000 × g, the homogenates were passed three times on oligo(dT) columns to recover the poly(A)+ RNA. The columns were
washed with 1X binding buffer without sodium dodecyl sulfate and the
poly(A)+ RNA was recovered by elution with Tris-EDTA (10 mM
Tris [pH 8.0], 1 mM EDTA) and precipitated with ethanol. For the
experiments involving BIES and BIEA constructs, total RNA instead of
poly(A)+ RNA was analyzed. Oligonucleotide-mediated RNase
cleavage of bicoid mRNA was carried out according to the
method of Brown and Harland (7). The RNA samples were
fractionated by electrophoresis in denaturing 1 or 1.5%
agarose-formaldehyde gels (27) and transferred to GeneScreen
nylon membranes (NEN Research Products) by using the manufacturer's
recommendations. The membranes were UV irradiated, baked, prehybridized
in Church's buffer (8), and hybridized in the same buffer
with -32P-labeled probes synthesized by random primer
labeling (15). Northern blots were quantified with a
PhosphorImager (Molecular Dynamics). The following plasmids were used
to prepare the probes: bicoid cDNA p1122, pGEM3rpA1, and
pBSIIKS+1-tub.
| |
RESULTS |
bicoid mRNA stability is developmentally
regulated.
To determine whether bicoid mRNA is
constitutively unstable or whether its stability is regulated in
development, we measured mRNA decay in retained oocytes, in
unfertilized eggs, and in early embryos.
To measure mRNA stability in retained oocytes, we took advantage of the
fact that females deprived of yeast retain their oocytes for extended
periods of time (35). These females accumulate mature
oocytes in which transcription and translation are arrested. Since no
transcription takes place in these mature oocytes, determination of
bicoid mRNA abundance as a function of time provides a
direct measure of turnover rate. Polyadenylated RNAs from ovaries of rapidly laying females and from ovaries of females that retained their
oocytes for 4, 8, and 12 days were analyzed on Northern blots. Like
that of the stable rpA1 mRNA, the intensity of the bicoid
signal remained constant over 12 days (Fig.
2A). Thus, bicoid mRNA is
completely stable for up to 12 days in retained oocytes.
View larger version (31K):
[in this window]
[in a new window]
|
FIG. 2.
bicoid mRNA is stable for at least 12 days in
retained oocytes and for at least 8 h in activated, unfertilized
eggs. Polyadenylated mRNAs were prepared from ovaries of 3BA transgenic
females that retained their oocytes for 0, 4, 8, or 12 days (A) and
from laid (activated), unfertilized eggs from 3BA transgenic females, 0 to 2, 2 to 4, 4 to 6, or 6 to 8 h after oviposition (B), and
analyzed on Northern blots with bicoid and rpA1 mRNA probes.
Similar results were obtained in at least two independent experiments.
Identical results were obtained for the 3BA hybrid mRNA (data not
shown). rpA1 mRNA served as a loading control. mRNA sizes:
bicoid mRNA, 2.6 kb; rpA1 mRNA, 0.6 kb.
|
|
In many instances, ongoing translation is required for mRNA degradation
(5, 33). The absence of translation could explain why
bicoid mRNA is completely stable in retained oocytes.
However, bicoid mRNA translation is activated in laid eggs,
even if they are unfertilized (11). We measured the
stability of bicoid mRNA in laid, unfertilized eggs of
different ages. As shown in Fig. 2B, the intensity of the
bicoid signal relative to the rpA1 signal was constant for
up to 8 h. Thus, bicoid mRNA in eggs is stable even if
it is translated.
The disappearance of bicoid mRNA during early embryogenesis
could be due either to the activation of a degradation pathway soon
after fertilization or to the onset of degradation at a later time.
Since bicoid is strictly maternal, changes in
bicoid mRNA abundance in embryos provides a direct measure
of decay. Figure 3 shows a detailed
analysis of bicoid mRNA abundance during the first 4 h
of embryogenesis. Precisely staged embryos were analyzed on Northern
blots (Fig. 3A), and the signals were quantified with a PhosphorImager
(Fig. 3B). The results showed that bicoid mRNA is stable
during the first 2 h of embryogenesis and is then rapidly degraded
between 2 and 3 h after fertilization. After 3 h of
embryogenesis, bicoid mRNA was barely detectable. From these
results, we estimate the bicoid mRNA half-life to be less
than 30 min between 2 and 3 h of development. Thus,
bicoid mRNA decay is activated at the end of the syncytial
blastoderm stage of embryogenesis.
View larger version (19K):
[in this window]
[in a new window]
|
FIG. 3.
bicoid mRNA is stable during the first 2 h of embryogenesis and is rapidly degraded after this time. (A)
Polyadenylated RNAs prepared from 0- to 1-, 1- to 2-, 2- to 3-, and 3- to 4-h-old embryos were analyzed on a Northern blot probed with
bicoid and with rpA1 mRNA as a loading control. mRNA sizes
are indicated in the legend to Fig. 2. (B) This histogram represents a
quantitative analysis of data from seven independent experiments
similar to the experiment presented in panel A. The intensity of the
bicoid (bcd) mRNA signal (relative to the 0- to
1-h signal, given in arbitrary units) was plotted versus the time of
development, in hours. The bicoid mRNA signals were
normalized to those of the rpA1 loading control. Error bars represent
standard deviations.
|
|
A bicoid mRNA destabilizing sequence is contained in
the 3' half of the message.
The initial stability of
bicoid mRNA and its subsequent rapid degradation suggest two
possible regulatory mechanisms: (i) protection from degradation until
2 h of development and (ii) activation of degradation after 2 h of development. The first mechanism predicts that deletion of the
regulatory element would destabilize bicoid mRNA. The second
mechanism predicts that deletion of the regulatory element would
prevent the destabilization of bicoid mRNA. To distinguish
between these two possibilities and to localize the putative regulatory
determinant in bicoid mRNA, we constructed fusion genes
containing sequences from bicoid and from stable mRNAs.
These genes were transformed into flies and mRNA stability was measured
in transgenic embryos.
Three constructs were initially tested: 5B, 3B, and 5T (Fig. 1).
Construct 5B contains the entire 5' half of bicoid mRNA, while the reciprocal constructs 3B and 5T contain the 3' half of
bicoid mRNA (Fig. 1). All constructs are driven by promoters that are active during oogenesis and inactive during the first 3 h
of embryogenesis (1-tubulin, rpA1, and
hsp26/sgs3 promoters). Thus, the decrease in mRNA
abundance during early embryogenesis serves as a measure of their
degradation rates. Polyadenylated RNAs were extracted from staged
embryos of transgenic lines. The abundance of bicoid,
hybrid, and reference (rpA1 or tubulin) mRNAs was measured by Northern
blot analysis (Fig. 4). In each lane, the
signal from the reference mRNA provided a measure of the amount of mRNA
analyzed.
View larger version (20K):
[in this window]
[in a new window]
|
FIG. 4.
5B mRNA is significantly more stable than
bicoid mRNA, while 3B and BBT mRNAs are unstable.
Polyadenylated RNAs prepared from 0- to 1-, 1- to 2-, 2- to 3-, and 3- to 4-h-old embryos derived from 5B, 3B 5T, and BBT transgenic flies
were analyzed on Northern blots hybridized with bicoid and
rpA1 or tubulin mRNA probes. The rpA1 and tubulin signals served to
quantify the amount of RNA analyzed in each lane. mRNA sizes:
bicoid mRNA, 2.6 kb; 5B mRNA, 1.5 kb; 3B mRNA, 1.9 kb; BBT
mRNA, 2.2 kb; and rpA1 mRNA, 0.6 kb.
|
|
Figures 4 and 5 present the results.
While none of the hybrid mRNAs was completely stable, the mRNAs
containing 3' bicoid sequences (3B and 5T) were degraded
more efficiently than the mRNAs containing 5' bicoid
sequences (5B). Quantitative analysis showed that 5B mRNA is threefold
more abundant than bicoid mRNA at 2 to 3 h of
development and eightfold more abundant at 3 to 4 h (Fig. 5). In
contrast, the stability of the 3B and 5T mRNAs was regulated similarly
to that of bicoid mRNA (Fig. 4B and 5). These results
suggest that a major destabilizing sequence resides in the 3' half of
bicoid mRNA, while a weaker destabilizing sequence resides
in the 5' half.
View larger version (20K):
[in this window]
[in a new window]
|
FIG. 5.
The 3' half of bicoid mRNA contains the main
destabilizing element. Each histogram represents the quantitative
analysis of experiments such as those shown in Fig. 4 and combines data
from at least three independent experiments. The intensities of
bicoid (bcd) mRNA signals (relative to the 0- to
1-h signal, given in arbitrary units) were plotted versus the time of
development, in hours. The abundance of bicoid mRNA was
normalized to rpA1 or tubulin loading controls. Error bars represent
standard deviations. No standard deviation is shown for 3B, since the
histogram shows the quantification of one experiment. However, the same
result as for 3B was obtained for 5T, a fusion with the tubulin gene
containing exactly the same bicoid sequence.
|
|
bicoid mRNA is localized at the anterior tip of the embryo
through sequences contained within 625 nucleotides of its 3' UTR (18). The selective degradation of bicoid mRNA at
cellularization could be dependent on its localization. To address this
hypothesis, we measured the stability of a bicoid mRNA in
which most of the 3' UTR was replaced by the 3' UTR of the tubulin gene
(Fig. 1, BBT). Figures 4C and 5 show that BBT mRNA is initially stable and is later degraded to become barely detectable at 3 to 4 h of
development, as is the case for endogenous bicoid.
The instability of BBT mRNA suggests that localization of
bicoid mRNA is not necessary for its specific degradation in
the early embryo and that the main bicoid destabilization
sequence is distinct from the localization sequence. This assumption is supported by experiments presented later in this paper.
The main BIE is contained within the 92 nucleotides immediately
following the translation termination codon.
The results above
suggested that the main instability element is located within the 3'
half of bicoid mRNA, between the SalI and
MluI sites (Fig. 1). To determine the location of the
instability element more precisely, a series of deletion derivatives
(3BA, 3BB, 3BC, and 3BD) from the 3B construct were generated (Fig. 1).
All constructs retained the terminal 733-nucleotide segment which is
deleted in BBT. In 3BA the complete bicoid 3' UTR remained, in 3BB the 92 nucleotides after the stop codon were deleted, in 3BC the
last quarter of the bicoid translated region plus the 92 nucleotides after the stop codon were deleted, and in 3BD, the third
quarter of the bicoid translated region was deleted. All
these constructs were under the control of the rpA1 promoter. The
constructs were transformed into the germ line and the abundance of the
corresponding mRNAs was analyzed as before. As shown in Fig.
6 and 7,
the 3BA and 3BD mRNAs were unstable, while the 3BB and 3BC mRNAs were
stable. Thus, the 92-nucleotide sequence that was deleted in 3BB is
required for mRNA destabilization at the time of cellularization of the
blastoderm.
View larger version (39K):
[in this window]
[in a new window]
|
FIG. 6.
3BA and 3BD mRNAs are unstable, while 3BB and 3BC mRNAs
are stable. Polyadenylated RNAs prepared from 0- to 1-, 1- to 2-, 2- to
3-, and 3- to 4-h-old embryos from 3BA, 3BB, 3BC, and 3BD transgenic
flies were analyzed on Northern blots hybridized with bicoid
and rpA1 mRNA probes. rpA1 mRNA served as a loading control. mRNA
sizes: bicoid mRNA, 2.6 kb; 3BA mRNA, 1.2 kb; 3BB mRNA, 1.8 kb; 3BC mRNA, 1.4 kb; 3BD mRNA, 1.6 kb; and rpA1 mRNA, 0.6 kb.
|
|
View larger version (24K):
[in this window]
[in a new window]
|
FIG. 7.
The main bicoid mRNA destabilizing element is
located within the first 43 nucleotides after the stop codon. Each
histogram represents the quantitative analysis of experiments such as
those shown in Fig. 8 and combines data from at least three independent
experiments. The intensities of the bicoid (bcd)
mRNA signals (relative to the 0- to 1-h signal, given in arbitrary
units) were plotted versus the time of development, in hours. The
abundance of bicoid mRNA was normalized to the rpA1 loading
control. Error bars represent standard deviations.
|
|
The main instability element is distinct from the NRE.
The 92 nucleotides after the stop codon of the bicoid mRNA contain
an NRE sequence. This element is also present in hunchback mRNA, where it is necessary for translational repression in the posterior part of the embryo. It is not clear why an NRE is present in
bicoid mRNA because Nanos protein and bicoid mRNA
are not normally colocalized. However, translation of bicoid
mRNA is repressed via the NRE when Nanos is ectopically localized in
the anterior part of the embryo in BicaudalD mutants
(34). To assess the role of the NRE in bicoid
mRNA destabilization, we analyzed NRE transgenic flies carrying a
variant of the bicoid gene with a deletion of the 45 3'-most
nucleotides of the 92-nucleotide element (Fig. 1). The size difference
between the messages derived from wild-type and NRE
bicoid genes was too small to separate them on Northern
blots. Instead, an RNase H assay was used. Poly(A)+ RNA
from different stages was hybridized with an oligonucleotide complementary to a sequence within the NRE and digested with RNase H. This treatment was expected to cleave bicoid mRNA into two fragments (1.6 and 1.0 kb) and leave the NRE mRNA intact. The Northern blot in Fig. 8 and quantitative
analysis in Fig. 7 show that the NRE mRNA (2.6-kb fragment) is as
unstable as the wild-type bicoid mRNA (1.6- and 1.0-kb
fragments). Consequently, the 45 nucleotides containing the NRE are not
necessary for the degradation of bicoid mRNA in early
embryos. Moreover, these results show that the BIE is contained within
the first 43 nucleotides after the stop codon.
View larger version (52K):
[in this window]
[in a new window]
|
FIG. 8.
Deletion of the NRE does not affect bicoid
mRNA stability. The hollow arrow in the diagram at the top
indicates the position of the NRE sequence that was deleted from the
bicoid gene. Polyadenylated RNAs were prepared from control
(0- to 1-h-old untransformed embryos) and 0- to 1-, 1- to 2-, 2- to 3-, and 3- to 4-h-old embryos from NRE transgenic flies. The RNAs were
hybridized with an oligonucleotide (Table 1) complementary to a
sequence missing in the NRE construct and digested with RNase H. The
RNAs were analyzed on a Northern blot hybridized with bicoid
(bcd) and rpA1 mRNA probes. The bicoid mRNA is
cleaved into two fragments of 1.6 and 1.0 kb, while the NRE mRNA is
unaffected by the treatment. rpA1 mRNA served as a loading control. The
control lane shows that wild-type bicoid mRNA is cleaved to
completion by the RNase H treatment. mRNA sizes: NRE mRNA, 2.6 kb,
and rpA1 mRNA, 0.6 kb.
|
|
The BIE is sufficient to destabilize an otherwise-stable mRNA.
To determine whether the 43-nucleotide BIE is sufficient to confer
proper regulation on a heterologous mRNA, the BIE was inserted in the
3' UTR of the otherwise-stable rpA1 mRNA to yield rpA1-BIE. A 0.5-kb
fragment of an unrelated sequence (human transferrin receptor 3' UTR)
was inserted into the rpA1 gene 3' UTR to distinguish the rpA1-BIE mRNA
from the endogenous rpA1 mRNA. The stability of rpA1 mRNA was not
affected by the insertion of this sequence. Figure
9 shows that when the BIE is inserted in
the sense orientation, rpA1 mRNA is destabilized with the same kinetics
as bicoid mRNA (Fig. 9, left; compare with Fig. 3). In
contrast, when the BIE is inserted in the antisense orientation, rpA1
is completely stable (Fig. 9, right). Therefore, the BIE is sufficient
to destabilize a stable mRNA.
View larger version (39K):
[in this window]
[in a new window]
|
FIG. 9.
The BIE is sufficient to confer regulated instability on
a heterologous mRNA. (Top) Schematic diagram of the constructs used to
obtain transgenic flies. Stippled rectangles, rpA1 gene UTRs; black
rectangles, rpA1-coding regions; white rectangles, 500-nucleotide-long
fragments from the transferrin gene (3' UTR TfR) inserted to tag the
recombinant gene; arrows, the 43-nucleotide BIE (BIES, right-pointing
arrow; BIEA, left-pointing arrow). (Middle) Representative Northern
blots of total RNA extracted from synchronized embryos (ages are at the
bottom of each lane) from transgenic flies carrying either the BIES or
the BIEA construct. (Bottom) Graphs showing quantification of three
independent experiments similar to those shown in the middle part.
Error bars represent standard deviations.
|
|
| |
DISCUSSION |
Little was known about bicoid mRNA metabolism beyond
the fact that by 4 h of embryogenesis, the message is completely
degraded (6). Among possible models to explain the observed
pattern
constitutive instability, destabilization at fertilization,
and activation of degradation in developing embryoswe found that the
latter model applies. bicoid mRNA is destabilized at the
time of cellularization of the blastoderm, which is the time of major
activation of the zygotic genome (2, 21), raising the
possibility that bicoid degradation requires de novo
transcription.
We found that bicoid mRNA is stable for over 1 week in
retained oocytes. Perhaps this is not surprising, since these oocytes are metabolically inactive. However, oviposited, unfertilized eggs
become completely activated (19). For instance, meiosis is
completed and translation is activated, including that of
bicoid mRNA (11). Surprisingly, bicoid
mRNA was completely stable for at least 8 h in oviposited,
unfertilized eggs. This observation makes a "clock" mechanism less
likely and supports the concept that zygotic transcription is required
for bicoid destabilization. Moreover, the stability of
bicoid mRNA in unfertilized eggs may explain the
larger-than-normal Bicoid protein amounts in such eggs (11).
In contrast to bicoid, an rpA1/fushi tarazu
hybrid mRNA is rapidly degraded in activated, unfertilized eggs
(25), suggesting that the activities that degrade
bicoid and fushi tarazu mRNAs are distinct.
All mRNAs carrying the 43-nucleotide BIE (3B, 5T, BBT, 3BA, 3BD,
NRE, and BIES) are unstable after 2 h of development, while all
mRNAs lacking the BIE (3BB and 3BC) are stabilized. Therefore, the BIE
is necessary for regulation of bicoid mRNA stability. Moreover, since mRNAs lacking the BIE (5B, 3BB, and 3BC) are much more
stable in early embryos, it can be argued that the BIE acts to
destabilize bicoid mRNA after 2 h of development as
opposed to being necessary to protect bicoid mRNA from
degradation until 2 h of development. This is in complete
agreement with the finding that the BIE is sufficient to destabilize
the heterologous rpA1 mRNA. The position of the BIE does not seem to be
important because in the BIES construct, the BIE is 600 nucleotides
from the stop codon instead of next to it. The BIE appears to act as an
independent, developmentally regulated target of mRNA degradation
because the sequence context in which it is placed is irrelevant. The
BIE is likely to provide a binding site for a factor or complex
involved in mRNA degradation. Binding sites could be recognized as
secondary structures (5). However, no stable secondary
structure was predicted when the computer program of Zuker
(38) was used to fold the BIE sequence.
The 5B transcript is much more stable than endogenous bicoid
mRNA but is not as stable as rpA1 mRNA. The 5' half of
bicoid mRNA has much weaker destabilizing activity than the
BIE, since 5B mRNA abundance is reduced 2-fold from the first hour to
the fourth hour, while 3B mRNA abundance is reduced 20-fold. Possibly, another destabilization element much weaker than the BIE is present in
the 5' half of bicoid mRNA. Alternative explanations, such as the creation of a fortuitous destabilizing element at the junction of the chimeric construct, cannot be ruled out.
A connection between mRNA translation and stability was suggested for a
number of mRNA degradation pathways (for reviews, see references
5 and 33). However, in the case
of bicoid, the translation status seems not to be critical
for control of mRNA decay. First, while bicoid mRNA is
continuously translated from the time of fertilization, its degradation
is activated only at cellularization of the blastoderm. Thus,
bicoid mRNA is stable both when translated (in activated,
unfertilized eggs and in early embryos) and when not translated (during
oogenesis and in retained eggs). Second, when a premature stop codon
was introduced by site-directed mutagenesis into codon 218 of the
bicoid cDNA and the construct was analyzed in transgenic
flies, the transgenic mRNA was degraded with a temporal profile
indistinguishable from that of wild-type bicoid mRNA
(21a). Third, the rate of translation appears to have no
effect on the regulation of bicoid mRNA stability. The 3B
mRNA contains the rpA1 gene 5' UTR, which is poorly translated during
early embryogenesis (22). Despite the presence of the rpA1
gene 5' UTR, 3B mRNA is as unstable as the endogenous bicoid mRNA. Moreover, 5T contains the 5' half of the actively translated tubulin mRNA linked to the same 3' half of the bicoid mRNA
as in 3B. 5T is degraded with the same kinetics as 3B.
Transcript localization could be a mechanism for targeting specific
mRNAs for degradation (20, 37). However, BBT mRNA, which
lacks localization sequences but contains the BIE, is unstable. Moreover, bicoid mRNA and fushi tarazu mRNA are
apically localized in early embryos (9) and yet, at the same
stage, the former is stable while the latter is unstable. Most
importantly, BIES mRNA is presumed not to be localized and is degraded
with the same kinetics as bicoid mRNA. These observations
suggest that bicoid mRNA localization and mRNA degradation
are two independent processes.
While no similarities were found by comparing the entire BIE sequence
with sequences in databases (using the BLAST program [1]), the short sequence UUUCAUU present in
the BIE was found to appear in a subset of the genes listed in the EMBL
UTR database. Of about 1,800 UTRs searched, 112 (~6%) contained the
UUUCAUU motif. Additional genes not listed in the UTR
database were also found to have this motif (data not shown).
Interestingly, a significant proportion of the genes retrieved from the
UTR database (~35%) could be classified among four classes of
regulatory genes whose mRNA and protein concentrations are expected to
change rapidly as a function of time. In the majority of the genes
(~82%), the motif was present in the 3' UTR. The consensus for the
21 nucleotides containing the UUUCAUU motif (from a total of
137 motifs; the Genetics Computer Group package, University of
Wisconsin, was used) is DDDDDHDUUUACUUDDDHWHH (D = not C, H = not G, and W = A or T), indicating that the
UUUCAUU motif is embedded in a region from which the C
residue is usually excluded. The presence of the UUUCAUU
motif in several classes of genes suggests the existence of a
common pathway for mRNA destabilization. Further research is needed to
verify the validity of this hypothesis.
We thank Robert Cohen and Kam Cheung (Columbia University) for
providing a plasmid containing the sgs3 promoter and the
nurse cell-specific enhancer from hsp26. We thank Paul
Macdonald (Stanford University) for flies transformed with the BBT
construct and for the plasmid p1122, containing the bicoid
cDNA. We thank Robin Wharton (Duke University) for flies transformed
with the NRE construct. P.S. is very grateful to Lukas Kuehn (ISREC,
Epalinges, Switzerland) for giving him the opportunity to finish this
study in his laboratory and for his continuous support. We thank Claude Bonnard for his help in use of the Genetics Computer Group package. We
thank Marilyn Doman, Ann Riedl, and Maria Goreti Freitas-Sibajev for
their comments on the manuscript.
| 1.
|
Altschul, S. F.,
W. Gish,
W. Miller,
E. W. Myers, and D. J. Lipman.
1990.
Basic local alignment search tool.
J. Mol. Biol.
215:403-410[Medline].
|
| 2.
|
Anderson, K. W., and J. A. Lengyel.
1979.
Rates of synthesis of major classes of RNA in Drosophila embryos.
Dev. Biol.
70:217-231[Medline].
|
| 3.
|
Ashburner, M.
1989.
In
Drosophila: a laboratory manual.
Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
|
| 4.
|
Audic, Y.,
F. Omilli, and H. B. Osborne.
1997.
Postfertilization deadenylation of mRNAs in Xenopus laevis embryos is sufficient to cause their degradation at the blastula stage.
Mol. Cell. Biol.
17:209-218[Abstract].
|
| 5.
|
Belasco, J., and G. Brawerman.
1993.
In
Control of messenger RNA stability.
Academic Press, San Diego, Calif.
|
| 6.
|
Berleth, T.,
M. Burri,
G. Thoma,
D. Bopp,
S. Richstein,
G. Frigerio,
M. Noll, and C. Nüsslein-Volhard.
1988.
The role of localization of bicoid RNA in organizing the anterior pattern of the Drosophila embryo.
EMBO J.
7:1749-1756[Medline].
|
| 7.
|
Brown, B. D., and R. M. Harland.
1990.
Endonucleolytic cleavage of a maternal homeobox mRNA in Xenopus oocyte.
Genes Dev.
4:1925-1935[Abstract/Free Full Text].
|
| 8.
|
Church, G. M., and W. Gilbert.
1984.
Genomic sequencing.
Proc. Natl. Acad. Sci. USA
81:1991-1995[Abstract/Free Full Text].
|
| 9.
|
Davis, I., and D. Ish-Horowicz.
1991.
Apical localization of pair-rule transcripts requires 3' sequences and limits protein diffusion in the Drosophila blastoderm embryo.
Cell
67:927-940[Medline].
|
| 10.
|
Driever, W., and C. Nüsslein-Volhard.
1988.
A gradient of bicoid protein in Drosophila embryos.
Cell
54:83-93[Medline].
|
| 11.
|
Driever, W., and C. Nüsslein-Volhard.
1989.
The bicoid protein is a positive regulator of hunchback transcription in the early Drosophila embryo.
Nature
337:138-143[Medline].
|
| 12.
|
Driever, W.,
G. Thoma, and C. Nüsslein-Volhard.
1989.
Determination of spatial domains of zygotic gene expression in the Drosophila embryo by affinity of binding sites for the bicoid morphogen.
Nature
340:363-367[Medline].
|
| 13.
|
Edgar, B. A., and S. A. Datar.
1996.
Zygotic degradation of two maternal Cdc25 mRNAs terminates Drosophila's early cell cycle program.
Genes Dev.
10:1966-1977[Abstract/Free Full Text].
|
| 14.
|
Edgar, L. G.,
N. Wolf, and W. B. Wood.
1994.
Early transcription in Caenorhabditis elegans embryos.
Development
120:443-451[Abstract].
|
| 15.
|
Feinberg, A. P., and B. Vogelstein.
1984.
A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity.
Anal. Biochem.
137:266-267[Medline].
|
| 16.
|
Frohnhöfer, H. G., and C. Nüsslein-Volhard.
1986.
Organization of anterior pattern in the Drosophila embryo by the maternal gene bicoid.
Nature
324:120-125.
|
| 17.
|
Kalfayan, L., and P. C. Wensink.
1981.
-Tubulin genes of Drosophila.
Cell
24:97-106[Medline].
|
| 18.
|
Macdonald, P. M., and G. Struhl.
1988.
Cis-acting sequences responsible for anterior localization of bicoid mRNA in Drosophila embryos.
Nature
336:595-598[Medline].
|
| 19.
|
Mahowald, A. P.,
T. J. Goralski, and J. H. Caulton.
1983.
In vitro activation of Drosophila eggs.
Dev. Biol.
98:437-445[Medline].
|
| 20.
|
Mason, J. O.,
G. T. Williams, and M. S. Neuberger.
1988.
The half-life of immunoglobulin mRNA increases during B-cell differentiation: a possible role for targeting to membrane-bound polysomes.
Genes Dev.
2:1003-1011[Abstract/Free Full Text].
|
| 21.
|
McKnight, S. L., and O. L. Miller, Jr.
1976.
Ultrastructural patterns of RNA synthesis during early embryogenesis of Drosophila melanogaster.
Cell
8:305-319[Medline].
|
| 21a.
| Mouland, A., and M. Jacobs-Lorena. Unpublished
observations.
|
| 22.
|
Patel, R., and M. Jacobs-Lorena.
1992.
Cis-acting sequences in the 5'-untranslated region of the ribosomal protein A1 mRNA mediates its translational regulation during early embryogenesis of Drosophila.
J. Biol. Chem.
267:1159-1164[Abstract/Free Full Text].
|
| 23.
|
Pirrotta, V.
1988.
Vectors for P-mediated transformation in Drosophila.
Biotechnol. Ser.
10:437-456.
|
| 24.
|
Qian, S.,
J.-Y. Zhang,
M. A. Kay, and M. Jacobs-Lorena.
1987.
Structural analysis of the Drosophila rpA1 gene, a member of the eukaryotic 'A' type ribosomal protein family.
Nucleic Acids Res.
15:987-1003[Abstract/Free Full Text].
|
| 25.
| Riedl, A., and M. Jacobs-Lorena. Unpublished data.
|
| 26.
|
Rubin, G. M., and A. C. Spradling.
1982.
Genetic transformation of Drosophila with transposable element vectors.
Science
218:348-353[Abstract/Free Full Text].
|
| 27.
|
Sambrook, J.,
E. F. Fritsch, and T. Maniatis.
1989.
In
Molecular cloning: a laboratory manual, 2nd ed.
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
|
| 28.
|
Schneider, C.,
M. J. Owen,
D. Banville, and J. G. Williams.
1984.
Primary structure of human transferrin receptor deduced from the mRNA sequences.
Nature
311:675-678[Medline].
|
| 29.
|
Schultz, R. M.
1993.
Regulation of zygotic gene activation in the mouse.
Bioessays
15:531-538[Medline].
|
| 30.
|
Steller, H., and V. Pirrotta.
1986.
P transposons controlled by the heat shock promoter.
Mol. Cell. Biol.
6:1640-1649[Abstract/Free Full Text].
|
| 31.
| St. Johnston, D., W. Driever, T. Berleth, S. Richstein,
and C. Nüsslein-Volhard. 1989. Multiple steps in the
localization of bicoid RNA to the anterior pole of the
Drosophila oocyte. Development
107(Suppl.):13-19.
|
| 32.
|
Surdej, P., and M. Jacobs-Lorena.
1994.
Strategy for epitope tagging the protein coding-region of any gene.
BioTechniques
17:560-565.
[Medline] |
| 33.
|
Surdej, P.,
A. Riedl, and M. Jacobs-Lorena.
1994.
Regulation of mRNA stability in development.
Annu. Rev. Genet.
28:263-282[Medline].
|
| 34.
|
Wharton, R. P., and G. Struhl.
1991.
RNA regulatory elements mediate control of the Drosophila body pattern by the posterior morphogen nanos.
Cell
67:955-967[Medline].
|
| 35.
|
Wyman, R.
1979.
The temporal stability of the Drosophila oocyte.
J. Embryol. Exp. Morphol.
50:137-144[Medline].
|
| 36.
|
Yasuda, G. K., and G. Schubiger.
1992.
Temporal regulation in the early embryo: is MBT too good to be true?
Trends Genet.
8:124-127[Medline].
|
| 37.
|
Zambetti, G.,
J. Stein, and G. Stein.
1987.
Targeting of a chimeric human histone fusion mRNA to membrane-bound polysomes in HeLa cells.
Proc. Natl. Acad. Sci. USA
84:2683-2687[Abstract/Free Full Text].
|
| 38.
|
Zuker, M.
1989.
On finding all suboptimal foldings of an RNA molecule.
Science
244:48-52[Abstract/Free Full Text].
|
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
and
Top
Abstract
Introduction
Present address: Genetics Unit, Swiss Institute for Experimental
Cancer Research, Chemin des Boveresses, CH-1066 Epalinges, Switzerland.
