|
|
|
|||||||
|
|||||||||
(Received for publication, April 11, 1995; and in revised form, June
12, 1995) From the
Tumors obtained from v-Ha-ras-transformed PB-3c cells
are characterized by autocrine interleukin-3 (IL3) expression, which
occurs either without (class I tumors) or with enhanced transcription
(class II tumors). To address possible post-transcriptional mechanisms
of IL3 expression, IL3 mRNA stability was examined in both tumor
classes. Increased stability of IL3 mRNA was detected in class I tumor
lines (t > 3 h), whereas rapid decay of IL3 transcripts (t < 0.5 h) was found in class II tumor lines. In both
tumor classes, the c-myc and interleukin-6 transcripts were
short-lived. Transcripts of a constitutively expressed IL3 reporter
gene were stable in class I tumor cells but unstable in class II tumor
cells, suggesting that IL3 mRNA stabilization involved a trans-acting mechanism. Rapid decay of IL3 reporter
transcripts was observed in untransformed PB-3c as well as in
v-Ha-ras expressing precursor cells linking transcript
stabilization to the tumor stage. Reporter transcript stabilization in
class I tumor cells correlated with increased IL3 production. Deletion
of the AU-rich element from the IL3 reporter gene further augmented IL3
mRNA levels as well as IL3 production, suggesting that the stabilizing
mechanism in class I tumor cells is not equivalent to AU-rich element
deletion.
Escape from proliferation control is a central feature of
tumorigenesis. In autocrine tumors, growth autonomy is accomplished by
the unregulated production of self-stimulating
mitogens(1, 2) . Where identified in experimental or
clinical hemopoietic malignancies, aberrant growth factor expression
mostly involved rearrangements of the respective genes. Thus, a t
(5;14) chromosomal translocation in a human B-cell leukemia placed the
IL3 gene in the vicinity of the immunoglobulin heavy chain
enhancer(3, 4) . In murine malignancies, activation of
IL3, ( We are characterizing a murine tumor model
in which v-Ha-ras-transduced IL3-dependent PB-3c cells
progress in vivo to two different classes of tumors with
autocrine IL3 production(11, 24) . Class I tumor lines
lack detectable rearrangements of the IL3 gene. Nuclear transcription
run-on data showed that IL3 mRNA expression in class I tumor lines did
not involve a transcriptional mechanism. The mechanism appeared to be
recessive because down-regulation of autocrine IL3 expression,
reversion to IL3 dependence, and inhibition of tumor cell growth was
observed following somatic cell fusion to IL3-dependent PB-3c
cells(11, 25) . More recently, we found that autocrine
proliferation of class I tumor cells could also be inhibited by
treatment with the immunosuppressant cyclosporin-A, which appeared to
act by promoting the degradation of tumor-expressed IL3 transcripts (26) . In contrast, autocrine class II tumor lines are
characterized by IL3 gene rearrangements due to the insertion of
endogenous retroviral elements (intracisternal A-particles), enhanced
IL3 transcription rates, and maintained IL3 expression in somatic cell
hybrids(11) . We now provide evidence that class I and class II
tumor lines differ in the post-transcriptional regulation of IL3 mRNA.
Autocrine IL3 expression in class I tumor lines involves IL3 transcript
stabilization by a trans-acting mechanism that is not
operative in autocrine class II tumor lines, or in the IL3-dependent
precursor cells.
Figure 2:
Summary of the transfection experiments
with the IL3 reporter constructs and the respective mRNA half-life. For
constitutive expression, the Moloney murine leukemia virus long
terminal repeat (MoMuLVLTR) enhancer is set in front of the
TATA box of the genomic IL3 reporter gene, which is marked by two
silent point mutations (circle with cross-hairs) in exon 3.
Expression of the transfected Mx-IL3 reporter gene is monitored by
RNase A/T1 protection using the exon 1-5 probe, which yields two
fragments of 209 and 156 nt, whereas a fragment of 368 nt is protected
by the endogenous IL3 transcript. Deletion or replacement of the 220-bp NcoI/StyI inserts in the 3`-UTR with the indicated
fragments is described under ``Materials and
Methods.''
For electroporation using the gene pulser
(Bio-Rad), 1
Figure 5:
IL3 production of Mx-IL3-transfected tumor
lines with or without Act-D treatment. Cells were treated as outlined
in the flow chart, and the mitogenic activity in the diluted
supernatants was assayed by [
Figure 1:
IL3 mRNA levels in
IL3 autocrine tumor lines after Act-D treatment. Poly(A) RNA was
prepared from total cytoplasmic RNA extracted before or at the
indicated times after Act-D addition from the indicated tumor lines and
analyzed by Northern blotting with the indicated cDNA probes (top) as described under ``Materials and Methods.''
The signals were quantitated, normalized to
Figure 3:
Mx-IL3 and Mx-(
Figure 4:
Mx-IL3 mRNA decay in the
v-Ha-ras-expressing precursor 15V4 (left) and the
clonal tumor derivative 15V4T2 (right). Total RNA was
extracted from the indicated cell lines before or after the indicated
times of Act-D treatment and analyzed by RNase A/T1 protection assay
using the exon 1-5 and the hph probe as described under
``Materials and Methods.''
Figure 6:
Decay of IL3 reporter transcripts with
replaced ARE in the untransformed PB-3c cell line. A, RNase
A/T1 protection assay. Total RNA was extracted from the indicated cells
before or at the indicated times after actinomycin-D treatment and
analyzed by RNase A/T1 protection assay using the exon 1-5 probe
and the hph probe as described under ``Materials and
Methods.'' B, reporter transcript decay. IL3 mRNA levels
and hph mRNA levels were quantitated and plotted as described
in the legend of Fig. 3.
The present report shows that autocrine IL3 expression in
class I tumor cells involves stabilization of IL3 transcripts that is
not found in class II tumor cells. The mechanism appears to be active
in trans because IL3 transcripts are stabilized despite the
presence of intact ARE instability determinants. This conclusion is
based on (i) the prolonged metabolic stability of IL3 mRNA in class I
tumor cells compared to class II tumor cells (Fig. 1); (ii) the
absence of cis-acting alterations in RNase protection and ARE
sequence analysis (data not shown); and (iii) the stability of a
Moloney murine leukemia virus-driven exogenous IL3 transcript in class
I V2D1 tumor cells (t > 3 h) but not in the class II tumor
R56VT (t of Increased stability of growth factor transcripts as an
oncogenic event has been discussed in a detailed study by Schuler and
Cole on the c-myc-transduced monocytic tumor line
2.3(22, 40) . Akin to IL3 in class I tumor lines,
constitutive GM-CSF expression in the 2.3 tumor line involved increased
stability of the transcript (t of 2.4 h) in the absence of
detectable gene alterations by Southern blot. The hypothesis of a trans-acting alteration was supported by the observation that
transcripts of a transfected neomycin cDNA fused to the 3`-UTR of
GM-CSF were stable in this 2.3 tumor but not in two other cell lines.
In contrast, reporter transcript destabilization mediated by the entire
3`-UTR of the protooncogenes c-fos or c-myc was still
functional. These experiments clearly indicated that GM-CSF mRNA and
protooncogene decay were differently regulated in the 2.3 tumor line
and could be put into effect by the respective 3`-UTR in cis. Our present study on the v-Ha-ras-dependent tumor formation
of PB-3c mast cells independently confirms the hitherto unique
observation by Schuler and Cole (22) that growth factor mRNA
stabilization in trans may be an oncogenic target in autocrine
tumor formation. Furthermore, with the precursor cells at hand, we
observed that IL3 mRNA stabilization did not occur in the untransformed
PB-3c cells nor in the v-Ha-ras-expressing tumorigenic
precursor cells 15V4 and R56V but in the class I tumor lines 15V4T2 and
V2D1 ( Fig. 3and 4 and data not shown). Thus, we suggest that
IL3 mRNA stabilization is a property acquired as a late step in the
PB-3c tumor model. No significant differences between class I and
class II tumor lines could be detected for the decay rates of the
short-lived c-myc and IL6 transcripts showing half-lives of
less than 0.6 h (Fig. 1). Although c-myc mRNA
degradation is known to be mediated by more than one instability
determinant(17, 22) , these data support the view that
class I tumor cells are not defective in a more general mechanism of
cytoplasmic mRNA turnover. To evaluate the specificity of IL3 mRNA
stabilization in class I tumor lines, we have tried to target the
stable IL3( At present, the molecular
basis of the IL3 mRNA stabilization in class I tumor lines is not
known. Our recent observation of IL3 transcript destabilization and the
reversion to IL3 dependence of class I tumor by treatment with the
immunosuppressant cyclosporin-A suggests involvement of an
immunophilin-targeted step(26) . Given the role of the six
AUUUA repeats as necessary cis-elements for IL3 mRNA decay (Fig. 6) (30) and the loss of autocrine IL3 expression
in somatic cell hybrids(11, 25) , the simplest model
would be that trans-acting factor(s) involved in recognition
of the ARE instability determinant are inactive in class I tumor cells.
This model is challenged by the increase of IL3 mRNA and IL3 protein
expressed in class I tumor cells when the ARE is altered or deleted
from IL3 reporter transcripts (Fig. 3D and 5B and data not shown). One explanation for this might be that
ARE-mediated degradation of IL3 transcripts in class I tumor cells is
not completely inactive. On the other hand, IL3 mRNA stabilization in
class I tumors might be mediated by sequences other than the ARE, as
discussed for GM-CSF mRNA stabilization in a lung cancer
line(42) . In addition, the deletion of the ARE instability
determinant might contribute to IL3 production by removing other
restrictions of gene expression, for example at the level of
translation(19, 43, 44, 45) . These
intriguing possibilities will have to be resolved in further studies.
Volume 270,
Number 35,
Issue of September 01, pp. 20629-20635, 1995
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
)GM-CSF, CSF-1, interleukin-5, and interleukin-6 (IL6)
by insertion of retroviral elements has been
described(5, 6, 7, 8, 9, 10, 11, 12) .
In most of these cis-acting alterations, transcriptional
activation of the growth factor locus has either been shown or
implicated. Given the importance of post-transcriptional regulation for
cytokine expression (13-15, for reviews see (16, 17, 18, 19) ), it is
conceivable that perturbance of post-transcriptional control mechanisms
might also play a role in oncogenesis. Indeed, constitutive growth
factor expression associated with stable transcripts has been described
in leukemic cell lines(20, 21, 22) .
Consistent with the role of the AU-rich elements (ARE) as mRNA
instability determinants(13, 23) , the stabilizing
alterations involved truncations of ARE from the 3`-UTR of the
respective growth factor transcripts by insertion of endogenous
retroviral elements(20, 21) . On the other hand,
stabilization of the GM-CSF mRNA by a trans-acting mechanism
has been described in a c-myc-transduced monocytic tumor line.
In this tumor line, heterologous transcripts fused to the 3`-UTR of
GM-CSF were stable, whereas the 3`-UTRs of the protooncogenes c-fos and c-myc were still effective to direct rapid transcript
degradation(22) .
Cells and Tissue Culture
PB-3c is a cloned
IL3-dependent mast cell line derived from murine DBA/2 bone
marrow(27) . The IL3 autocrine tumor lines V2D1, 15V4T2, V4D6,
and R56VT have been described previously(11, 24) .
15V4 and R56V were obtained by v-Ha-ras transduction of the
IL3-dependent PB-3c clone 15 and R56,
respectively(11, 28) . The lines are cultured in
Iscove's modified Dulbecco's medium supplemented with 10%
fetal calf serum (Life Technologies, Inc.), 100 units/ml penicillin,
100 mg/liter steptomycin, and 50 µM of 2-mercaptoethanol
and is referred to as IMDM/10% FCS. For the culture of IL3-dependent
lines, conditioned medium from the X63-mIL3 line (29) was added
in saturating amounts, typically 1% (referred to as IMDM/10% FCS/IL3).IL3 Reporter Constructs and Electroporation
In the
Mx-IL3 construct(30) , the 0.6-kilobase pair HindIII/SacI fragment of the Moloney murine leukemia
virus long terminal repeat enhancer drives the 2.5-kilobase pair ApaI/SpeI fragment of the mouse genomic IL3 gene (31) starting 25 nucleotides upstream of the TATA box and
ending 256 bp downstream of the polyadenylation site. This plasmid
construct contains also the hygromycin-B phosphotransferase gene hph(32) under the control of the SV40 regulatory
regions to allow selection of stable transfectants and serves as
control for a stable mRNA. In the Mx-(
AU)IL3, the 216-bp NcoI/StyI fragment in the 3`-UTR of IL3 was deleted
(see Fig. 2)(30) . In the AUIL3, AUFOS, and AUMYC
derivatives of Mx-IL3, the NcoI/StyI fragment was
replaced with the respective sequences indicated in Fig. 2as
follows. First the annealed sense and antisense oligonucleotides were
phosphorylated, cloned into the SmaI site of KS-Bluescript
vector and sequenced, then cut out as NcoI/StyI
fragment, ligated into the corresponding sites of the 568-bp BglII fragment of IL3 present in the pSP72 vector, and finally
inserted into the Mx-IL3 gene as BglII fragment. The
Mx-(AU6)IL3 and the Mx-(AU3)IL3 were obtained by polymerase chain
reaction using the sense primers M820
5`-CCCATGGCTATTTATTTATGTATTTATGT-3` or M821
5`-CCCATGGTGTATTTATTTATTTATTGCC-3`, respectively, with the antisense
primer M822 5`-GATACATGTTGCATGCTGTGT-3` on Mx-AUIL3 plasmid DNA as
template. The amplicon was cut with NcoI and SphI and
cloned into the correspondingly cut BglII-fragment of IL3 (in
pSP72), which was inserted into the IL3 gene. Sequence and orientation
was verified in all final constructs by dideoxysequencing (U. S.
Biochemical Corp.).
10
of the indicated cell lines were
incubated in IMDM/10% FCS for 2 h before washing twice with ice-cold
phosphate-buffered saline. After resuspension in 0.8 ml of
phosphate-buffered saline, the cells were mixed with 15 µg of the
plasmid (linearized with Asp) and incubated for 10 min on
ice in the electroporation cuvette (0.4-µm electrode, Bio-Rad).
After pulsing with 300 V and 960 microfarad, the cells were again put
on ice for 10 min and then resuspended in prewarmed IMDM/10% FCS/IL3.
After 48 h, selection of stable transfectants was started by adding 1
mg/ml of hygromycin-B (Calbiochem) to culture medium.
RNA Extraction and Northern Blotting
Cells were
resuspended in fresh prewarmed IMDM/10% FCS at 1 10
cells/ml for 2 h. Total cytoplasmic RNA was extracted according
to Gough (33) either directly (time point, 0 min) or at the
times indicated after addition of 5 mg/liter Act-D. Poly(A)-enriched
RNA was obtained as described previously(11) . RNA was
subjected to electrophoretic separation in 1.1% (w/v) agarose gel
containing 0.66 M formaldehyde in MOPS buffer, pH 5.9,
according to Thomas(34) , blotted onto nitrocellulose filters
(BA-S85, Schleicher & Schuell). Prehybrization, hybridization, and
washing were done as described previously(11) . The IL3 exon
1-5 probe was generated from a SP6 vector containing a 368-bp IL3
cDNA fragment kindly provided by N. Gough using an in vitro transcription kit (Boehringer Mannheim). The chicken 
-actin
probe (570-bp PstI fragment), a gift from Y. Nagamine, was
labeled using a random priming kit (Boehringer Mannheim). The c-myc probe was generated by random priming from a 2.8-kilobase pair
fragment (35) spanning from exon 2 to exon 3. The IL6 cDNA EcoRI fragment, a gift of J. Van Snick, was labeled by random
priming.RNase A/T1 Protection Assay
The RNase A/T1
protection assay was done as described previously (30) hybridizing 5 or 10 µg of total RNA for 16 h with an
excess of labeled antisense probe. Antisense RNA probes labeled with
[
-P]GTP were synthesized from SP6 promoters
using the SP6 transcription kit (Boehringer Mannheim), and full-length
transcripts were purified by polyacrylamide gel electrophoresis. The
exon 1-5 probe was prepared as indicated above. The exon 1
template contained the genomic 274-bp ApaI/Nae1
fragment cut with ApaI, the exon 5 template the 627-bp XbaI/Spe1 fragment cut with XbaI, both of
which are derived from the 8.5-kilobase pair EcoRI genomic IL3
fragment (31) and cloned into the pGEM3Z vector. The template
for hygromycin resistance gene hph mRNA (26) was a
96-bp EcoRI/PstI fragment in the pGEM vector cut with EcoRI.
Quantitation of mRNA Levels
For quantitation, the
activity of the specific mRNA hybridization signals for IL3,
c-myc, hph, and -actin were determined after
exposure to a storage phosphor screen and analysis in a PhosphorImager
by the Image Quant program (Molecular Dynamics, Sunnyvale, CA). For
decay measurements, the time point 0 min values of the -actin- or hph-normalized IL3, c-myc, and IL6 signals were taken
as 100%, and the averaged values of two independent experiments were
plotted versus the time of Act-D exposure.IL3 Production in Culture Supernatants
The cells
were seeded at 2 10
/ml in IMDM/10% FCS after three
rounds of washing and incubated for 24 h. The culture supernatants were
collected following centrifugation and passed through 0.45-µm
filters (Acrodisc). Supernatant dilutions in IMDM/10% FCS were done in
triplicate in microtiter plates. After 24 h of incubation, mitogenic
activity was tested by [
H]thymidine incorporation
of IL3-dependent PB-3c cells (5 10
cells/100
µl) during 6 h. The mitogenic activity was inhibitable by the
addition of the anti-IL3 antibody from the rat anti-mouse IL3 hybridoma
ATCC HB10652 ( (36) and data not shown). Comparison of the 24-h
IL3 production of cell lines after 1 h of actinomycin-D exposure was
done as described in the legend to Fig. 5.
H]thymidine
incorporation of IL3-dependent PB-3c cells as described under
``Materials and Methods.'' The effect of Act-D treatment on
IL3 production is indicated in percentages on top of the bars. A, class I V2D1 Mx-IL3, class II R56VT Mx-IL3,
and untransfected V2D1; B, class I V2D1 Mx-(AU)IL3 and
class II R56VT Mx-(AU)IL3.
IL3 Transcripts Are Stable in Class I Tumor Lines but
Not in Class II Tumor Lines
We have previously reported that
autocrine IL3 expression in tumors derived from
v-Ha-ras-expressing PB-3c cells involved two different
mechanisms: transcriptional activation by insertion of an
intracisternal A-particle in class II tumor lines and an unknown
post-transcriptional mechanism in class I tumor lines that lack IL3
gene rearrangements(11) . To further characterize class I and
class II tumors, we have now compared the rates of IL3 mRNA decay in
both tumor classes after inhibition of transcription with Act-D. The
analysis of mRNA decay was limited to 3 h throughout this study to
limit the side effects of Act-D. As shown by Northern blotting (Fig. 1, top), IL3 mRNA was stable in the class I tumor
line V2D1 over the 3 h of Act-D treatment, whereas IL3 mRNA levels
decreased in the class II tumor line V4D6 over the same time period.
Unlike the IL3 transcripts, IL6 and c-myc transcript levels
decreased in both tumor classes ( Fig. 1and data not shown).
Rehybridization for -actin provided the reference of a stable mRNA (Fig. 1, top). Analysis of two other tumor lines 15V4T2
(class I) and R56VT (class II) showed a corresponding pattern for each
class (data not shown). The specific mRNA signals were quantitated by
storage phosphor screens, normalized to -actin expression, and
plotted versus the time of Act-D treatment (Fig. 1, bottom). As reported for other
transcripts(17, 37) , a biphasic decay pattern of the
normalized IL3 levels was observed in class II tumor lines. After 1 h
of Act-D treatment, IL3 mRNA decay leveled off at around 30% of the
time point 0 min value. For the initial phase, a half-life of about 0.5
h was determined for IL3 mRNA in the class II tumor lines V4D6 and
R56VT. In contrast, IL3 mRNA was stable (t > 3 h) in the
class I tumor lines V2D1 and 15V4T2. In all tumor lines, c-myc transcripts showed rapid decay rates with a half-life of less than
0.6 h. IL6 mRNA decay was not impaired in the class I tumor V2D1
decaying as rapidly as in the class II tumor line V4D6 (t <
0.5 h) (Fig. 1, bottom, and data not shown). We
conclude from these data that autocrine IL3 expression in the class I
tumors V2D1 and 15V4T2 involves a post-transcriptional alteration at
the level of mRNA stability that is not found in the class II tumor
lines V4D6 and R56VT.
-actin levels, and
plotted versus the time of Act-D treatment (bottom)
taking the time 0 min value as 100%. , IL3 class I V2D1;
,
IL3 class II V4D6; 
, IL6 class II V4D6; , IL6 class I V2D1;
, c-myc class I V2D1;
, c-myc class II
V4D6.
No Evidence for Altered IL3 Transcripts in Class I Tumor
Cells
The rapid degradation of short-lived transcripts has been
shown to require specific cis-acting nucleotide sequences that
function as instability determinants(13, 17) . With
respect to IL3, the deletion of the entire ARE or specific point
mutations in AUUUA motifs have been shown to abolish rapid IL3 mRNA
degradation and result in stable transcripts(30) . We therefore
analyzed the IL3 transcripts for alterations by RNase A/T1 protection
assay using three probes that were chosen to cover the entire IL3
transcript from the 5`-UTR to the polyadenylation site in the 3`-UTR.
In the class I tumor lines V2D1 and 15V4T2, the protected fragments
corresponded to the sizes of 192, 368, and 359 nt as expected for
unaltered transcripts (data not shown). The absence of point mutations
in the ARE was confirmed directly by dideoxy sequencing of the IL3 cDNA
obtained by reverse transcriptase polymerase chain reaction from the
class I tumor line V2D1 (data not shown). Thus, cis-acting
alterations were not very likely to account for the stability of IL3
transcripts found in the class I tumor lines V2D1 and 15V4T2.Evidence for IL3 mRNA Stabilization by a trans-Acting
Mechanism in Class I Tumor Cells
To identify differences in
the trans-acting regulation of IL3 mRNA stability, an
exogenous IL3 reporter gene termed Mx-IL3 (30) was transfected
into both tumor classes. For high constitutive expression of the
Mx-IL3, the genomic IL3 transcription unit was driven by the Moloney
leukemia virus long terminal repeat enhancer (depicted in the upper
part of Fig. 2). Due to two silent point mutations in exon
3 (indicated as focus in Fig. 2), Mx-IL3 expression levels could
be distinguished as two bands of 206 and 156 nt from the endogenous
transcript of 368 nt using the exon 1-5 probe in an RNase A/T1
protection assay (Fig. 3A). The hph probe for
the hygromycin-B resistance gene was included as loading control. As
shown in the Act-D experiment (Fig. 3A), Mx-IL3
specific transcripts decayed rapidly in the transfected class II tumor
line R56VT. In contrast, Mx-IL3 mRNA levels were stable in the class I
tumor line V2D1 (Fig. 3A). By quantitation of the
Mx-IL3 mRNA expression levels and normalization to the hph loading control, the half-life of the Mx-IL3 mRNA was calculated
by linear regression to be around 0.5 h in the class II tumor line
R56VT but greater than 3 h in the class I tumor V2D1 (Fig. 3C). The data indicate that IL3 mRNA
stabilization in class I tumor cells involves a trans-acting
mechanism that counters the ARE instability function.
AU)IL3 decay in tumor
cells after Act-D treatment. Total RNA was extracted from the indicated
cell lines before or at the indicated times after Act-D treatment and
analyzed by RNase A/T1 protection assay using the exon 1-5 and
the hph probe as described under ``Materials and
Methods.'' The expression levels of the IL3 reporter transcripts
were quantitated by storage phosphor technique, and the values plotted
after normalization to the respective hph signal. To plot the
mRNA levels remaining, the normalized time point 0 min values were
taken as 100%. A, class I V2D1 Mx-IL3 (left) and
class II R56VT Mx-IL3 (right). B, class I V2D1
Mx-(AU)IL3 (left) and class II R56VT Mx-(AU)IL3 (right). C, decay of IL3 reporter transcripts.
, Mx-(
AU)IL3 class I V2D1;
, Mx-(AU)IL3 class
II R56VT;
, Mx-IL3 class I V2D1;
, Mx-IL3 class II R56VT. D, relative expression levels of IL3 reporter transcripts at 0
min. Thick hatching lines, Mx-IL3; thin hatching
lines, Mx-(AU)IL3.
ARE Deletion Increases the Steady-state Levels of the IL3
Transcripts in Both Tumor Classes
Because ARE deletion rendered
IL3 reporter transcripts stable in untransformed PB-3c
cells(30) , it seemed possible that the trans-acting
stabilization in class I tumor cells represented a defect in factor(s)
recognizing the ARE instability determinant. If that was the case, ARE
deletion from the reporter transcripts would be predicted to remain
without effect in class I tumor cells but should abrogate the
differences to the class II tumor cells. We therefore transfected
Mx-(AU)IL3 reporter constructs into both tumor classes (see Fig. 2). As expected, the (AU)IL3 transcripts were now
stable in the class II tumor cells (Fig. 3C). Little
difference in the stability of the Mx-IL3 and the Mx-(AU)IL3
transcripts were detected in the class I tumor line V2D1 over the 3 h
of the Act-D experiment. In both tumor classes, however, the
(AU)IL3 reporter transcripts were present at higher levels
relative to the hph reference transcripts or the endogenous
IL3 transcripts (Fig. 3B). Previous studies in PB-3c
cells have shown that the steady-state levels of the Mx-IL3 mRNA
correlate with increasing transcript stability(30) . To
estimate the steady-state levels, we determined the ratio of the
exogenous IL3 signal (lower band of 156 nt) to the hph signal
at time point 0 min (Fig. 3D). The results show that
the Mx-IL3 levels are about 5-fold higher in the class I tumor cells
than in the class II tumor cells. The levels of the reporter
transcripts lacking the ARE were found to be 4-fold higher in class I
tumor cells but 20-fold higher in class II tumor cells, thereby
reaching nearly equal levels (Fig. 3D). Thus, IL3 mRNA
accumulation could be increased by ARE deletion not only in the class
II but also in the class I tumor cells. The data suggest that the trans-acting mechanism operating in class I tumor cells is not
equivalent to ARE deletion.IL3 mRNA Stabilization Is Linked to Tumor
Progression
We also compared Mx-IL3 transfectants of the
v-Ha-ras expressing PB-3c clone 15V4 with the clonal class I
tumor derivative 15V4T2 (Fig. 4). The results showed that the
exogenous Mx-IL3 transcripts are stable in the class I tumor line
15V4T2 but decay rapidly in the v-Ha-ras-expressing precursor
line 15V4 Mx-IL3. Due to the weak expression levels, the endogenous
transcript in 15V4T2 is not apparent on this exposure. Rapid decay was
also observed in the v-Ha-ras-transduced precursor clone R56V
as well as in the untransformed PB-3c cells (data not shown). The
comparison between precursor and tumor lines suggests that the
stabilization of IL3 transcripts is linked to the autocrine class I
tumor stage.
IL3 Production Correlates with Transcript Stability in
Class I Tumor Cells
As evident from Fig. 3A, a
substantial difference in Mx-IL3 mRNA levels of the transfected class I
and class II tumor lines could be observed after 1 h of Act-D
treatment. To investigate whether differences of IL3 mRNA stability in
tumor cells would be translated into corresponding amounts of protein,
we compared the 24 h IL3 production of Mx-IL3 transfected tumor cell
lines with or without 1 h of preincubation with Act-D (Fig. 5A). Because Act-D inhibits transcription by
intercalation of double-stranded DNA, its effect is essentially
irreversible. Before seeding the cells in IL3-free medium, unbound
Act-D was removed from the cell cultures by washing. The cell number
was determined before and after each protocol, confirming that both
lines had doubled only when not exposed to Act-D. As shown for two
dilutions of the culture supernatants, the class I V2D1 Mx-IL3 cells
secreted about 4-fold more IL3 than the class II R56VT Mx-IL3 cells.
After Act-D exposure, the average 24-h production of IL3 is reduced to
about 17% in the class I tumor supernatants but to 4% in the class II
tumor supernatants (Fig. 5A). Comparison with the
untransfected tumor line V2D1 indicates that the contribution of the
endogenous IL3 is negligible at these dilutions. Similar results were
obtained examining the 5 h production, albeit at a lower level (data
not shown). Thus, the abundance of transcripts without and with Act-D
treatment (Fig. 3) correlates with the amount of IL3 produced.
These results support the notion that the increased stability of IL3
transcripts in class I tumor cells is a relevant mechanism for the
autocrine IL3 loop.Deletion of the ARE Increases IL3 Production of Class I
Tumor Cells
When class I and class II tumor lines transfected
with the ARE-deleted IL3 construct Mx-(AU)IL3 were analyzed, both
tumor classes produced nearly equal amounts but at a 30-fold higher
level (Fig. 5B). Pretreatment with Act-D lowered the
IL3 production of the transfected class I and class II tumor cells to
about 50 and 30%, respectively. Thus, the deletion of the ARE
instability determinant from the IL3 transcripts had not only abolished
the difference between the two tumor classes with respect to the
overall IL3 production but also reduced the effect of transcriptional
inhibition by Act-D on the IL3 yield as expected for stable
transcripts.The AU-rich Elements of c-fos and c-myc Are Not
Sufficient to Mediate Rapid Decay of IL3 Transcripts
The
insertion of ARE instability determinants derived from the 3`UTR of
GM-CSF, c-fos, or c-myc has been shown to mediate
destabilization of a heterologous stable
transcripts(13, 17, 22, 37, 38, 39) .
We thought of applying a similar approach to the characterization of
class I alteration by targeting the stable Mx-(AU)IL3 reporter
transcripts to the c-myc or the c-fos degradation
pathway. Therefore, we inserted the heterologous ARE of c-myc or c-fos as 78-nt fragments in place of the autologous
IL3 ARE (Fig. 2). The validity of this approach was tested in
PB-3c cells by analyzing the decay of IL3(AU) transcripts with
reinserted autologous ARE fragments of 57, 39, or 19 nt (Fig. 2). Rapid decay of transcripts with a half-life of less
than 30 min was observed for Mx-AUIL3 (data not shown) or Mx-(AU6)IL3
containing six AUUUA motifs (Fig. 6). For Mx-(AU3)IL3 containing
only the three downstream AUUUA-motifs of the ARE, transcript decay was
slowed down showing a half-life of approximately 85 min as revealed by
normalization to hph mRNA levels (Fig. 6B). In
contrast, reinsertion of ARE derived from c-fos or c-myc resulted in stable transcripts with a half-life of more than 3 h (Fig. 6, A and B). The lack of rapid
degradation upon reinsertion of the c-fos and c-myc ARE into the IL3 transcript was confirmed in the class II tumor
line R56VT (data not shown). Thus, this approach did not allow
characterization of the specificity of the class I alteration. However,
the data emphasize the role of the clustered AUUUA motifs for IL3
transcript degradation in PB-3c cells and indicate that the degradation
pathways for c-myc or c-fos are not efficiently
targeted in PB-3c cells by the respective AU elements alone in the
context of a constitutively expressed IL3 transcript.
, PB-3c Mx-(AUMYC)IL3; ,
PB-3c Mx-(AUFOS)IL3;
, PB-3c Mx-(AU3)IL3; , PB-3c
Mx-(AU6)IL3.
0.6 h) (Fig. 3, summarized in Fig. 2). After 1 h of Act-D exposure, the abundance of IL3 mRNA
signals differed considerably. To assess whether or not the IL3
transcripts stabilized in class I tumor cells represent functional
mRNA, we have determined the mitogenic IL3 activity in 24 h culture
supernatants of class I and class II Mx-IL3 transfectants with or
without 1 h of Act-D pretreatment (Fig. 5). The results show
that the Mx-IL3 transfected class I V2D1 cells produce 4-fold higher
and following Act-D exposure about 16-fold higher mitogenic IL3
activity than the Mx-IL3 transfected class II tumor line. The
differences between the tumor classes with respect to IL3 mRNA
stability and IL3 production disappeared upon deletion of the ARE from
the transfected IL3 reporter gene ( Fig. 3and Fig. 5).
The data support the view that ARE-mediated transcript decay is at
least partially inactivated in class I tumor cells and suggest that IL3
mRNA stabilization is of relevance to the autocrine IL3 loop in class I
tumor cells. In contrast, an increased rate in IL3 gene transcription
as found in class II tumor lines (11) or as constructed
experimentally in Mx-IL3 transfected PB-3c (30) appears to be
sufficient to counterbalance rapid transcript degradation in order to
maintain steady-state expression levels required for a tumorigenic
autocrine loop. Thus, post-transcriptional (class I) or transcriptional
up-regulation (class II) appear to occur as alternative mechanisms in
the autocrine tumor formation of v-Ha-ras expressing PB-3c
mast cells.AU) transcripts to the c-fos or the c-myc degradation pathway in PB-3c cells by inserting the respective ARE (Fig. 6). In contrast to the rapid degradation conferrable onto
stable reporter transcripts like -globin in
fibroblasts(17) , we found that IL3 reporter transcripts
containing the c-fos or the c-myc ARE were already
rather stable in the untransformed PB-3c (summarized in Fig. 2).
However, rapid degradation of IL3 reporter transcripts in PB-3c could
be restored by the IL3 ARE of 57 or 39 nt containing six AUUUA motifs (Fig. 6). A short sequence of 19 nt containing three AUUUA
motifs showed an intermediate IL3 decay rate with a half-life of around
85 min. This indicates that despite some similarity, the AREs of IL3,
c-fos, and c-myc are not functionally equivalent in
the context of the IL3 transcripts and the cell system used. Inspection
of the inserted ARE suggests that the structure and/or number of three
AUUUA motifs present in a cluster might be critical (Fig. 2), as
suggested by a detailed mutational analysis of AUUUA-motifs in the ARE
in IL3(30) . A recent study on the ARE of GM-CSF indicated that
rapid degradation (t < 1 h) may in fact require repeated
copies of a slightly larger UUAUUUA(U/A)(U/A)
determinant(23, 41) .
)
We thank our colleagues Drs. J. Garcia-Sanz, S. Hahn,
and Y. Nagamine for helpful discussions.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Groger, R. Loewe, W. Holnthoner, R. Embacher, M. Pillinger, G. S. Herron, K. Wolff, and P. Petzelbauer IL-3 Induces Expression of Lymphatic Markers Prox-1 and Podoplanin in Human Endothelial Cells J. Immunol., December 15, 2004; 173(12): 7161 - 7169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Tran, F. Maurer, and Y. Nagamine Stabilization of Urokinase and Urokinase Receptor mRNAs by HuR Is Linked to Its Cytoplasmic Accumulation Induced by Activated Mitogen-Activated Protein Kinase-Activated Protein Kinase 2 Mol. Cell. Biol., October 15, 2003; 23(20): 7177 - 7188. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X.-F. Ming, G. Stoecklin, M. Lu, R. Looser, and C. Moroni Parallel and Independent Regulation of Interleukin-3 mRNA Turnover by Phosphatidylinositol 3-Kinase and p38 Mitogen-Activated Protein Kinase Mol. Cell. Biol., September 1, 2001; 21(17): 5778 - 5789. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. L. Bernasconi, T. A. M. Wormhoudt, and I. A. Laird-Offringa Post-Transcriptional Deregulation of myc Genes in Lung Cancer Cell Lines Am. J. Respir. Cell Mol. Biol., October 1, 2000; 23(4): 560 - 565. [Abstract] [Full Text]
|
|
|
|
 |
|
 C. M. Brennan, I.-E. Gallouzi, and J. A. Steitz Protein Ligands to HuR Modulate Its Interaction with Target mRNAs In Vivo J. Cell Biol., September 25, 2000; 151(1): 1 - 14. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Stoecklin, X.-F. Ming, R. Looser, and C. Moroni Somatic mRNA Turnover Mutants Implicate Tristetraprolin in the Interleukin-3 mRNA Degradation Pathway Mol. Cell. Biol., June 1, 2000; 20(11): 3753 - 3763. [Abstract] [Full Text]
|
|
|
|
 |
|
 I.-E. Gallouzi, C. M. Brennan, M. G. Stenberg, M. S. Swanson, A. Eversole, N. Maizels, and J. A. Steitz HuR binding to cytoplasmic mRNA is perturbed by heat shock PNAS, March 28, 2000; 97(7): 3073 - 3078. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. P. K. Nair, H. H. Hirsch, M. Colombi, and C. Moroni Cyclosporin A Promotes Translational Silencing of Autocrine Interleukin-3 via Ribosome-Associated Deadenylation Mol. Cell. Biol., January 1, 1999; 19(1): 889 - 898. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |