Oncogene (1999) 18, 117 ± 125
ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00
http://www.stockton-press.co.uk/onc
The polyadenylation inhibitor cordycepin (3'dA) causes a decline in c-MYC
mRNA levels without a€ecting c-MYC protein levels
Panayotis Ioannidis1,2, Nelly Courtis1, Maria Havredaki2, Emmanuel Michailakis3,
Chris M Tsiapalis1 and Theoni Trangas1
1
Papanikolaou Research Center of Oncology, St Savvas Hospital, 171 Alexandras Avenue, 11522 Athens; 2NCSR-Democritos,
Athens; 3Oncology Unit, Athens General Hospital, Athens, Greece
Study of the distribution of the poly(A) tail length of
c-myc mRNA in several cell lines revealed a distinct,
prevailing population with short poly(A) tails, derived
through sequential deadenylation. To elucidate the
possible in vivo function of this distinct short tailed
c-myc mRNA population, the polyadenylation inhibitor
cordycepin was used. This resulted in a decline in steady
state c-myc mRNA levels with the remaining messenger
mostly oligoadenylated. However, c-MYC proteins did
not follow the reduction of the c-myc mRNA. On the
other hand, in cells exposed to physiological agents
known to downregulate c-myc expression, the reduction
of mRNA steady state levels, was re¯ected upon c-MYC
protein levels. The dissociation between c-myc mRNA
and protein levels caused by cordycepin was not due to
the stabilization of the c-MYC proteins and was not an
indiscriminate e€ect since in the presence of cordycepin,
c-fos mRNA and protein levels concomitantly declined.
Our data indicate that under these conditions, a long
poly(A) tail is not instrumental for c-myc mRNA
translation and furthermore, the discrepancy in the
steady state of c-myc mRNA level: c-MYC protein ratio
between control cells and cells treated with cordycepin
indicates that c-myc mRNA is subjected to translational
control.
Keywords: c-myc mRNA; translation; poly(A) tail;
cordycepin
Introduction
In recent years, it has become increasingly clear that
post-transcriptional control is an important mechanism
for regulating gene expression. This is more prominent
in the regulation of genes involved in critical processes
of cellular physiology such as proto-oncogenes, growth
factors, genes involved in in¯ammatory response etc.
Regulation of these genes at the level of mRNA
stability has been established (reviewed in: Ross, 1995)
and furthermore a number of examples indicate
regulation at the level of translation, as well (Kruys
et al., 1988; Bernstein et al., 1997; Black et al., 1997;
Yang et al., 1997). The elements governing either of
these mechanisms are usually present in the untranslated regions of the mRNA. In the 3'UTR besides
speci®c cis elements, another general regulatory
Correspondence: T Trangas
Received 23 September 1997; revised 2 July 1998; accepted 3 July
1998
element exists, the poly(A) tail. It is present at the 3'
end of virtually all mRNAs and a€ects cytoplasmic
mRNA stability and translatability (Sachs and Wahle,
1993).
Withstanding mRNAs degraded through regulated
endonucleolytic cleavage, it is believed that both stable
and unstable mRNAs do not begin to decay until the
poly(A) tail is reduced to a minimum length (Jacobson
and Peltz, 1996). These observations support the
notion that as long as a long poly(A) tail is present
the mRNA is protected from rapid indiscriminate
degradation. Di€erences in the stability of mRNAs
re¯ect di€erences in the rate of deadenylation and/or
the rate of degradation of the oligoadenylated form of
the message (Decker and Parker, 1994; Chen and Shyu,
1995).
On the other hand, there is extensive evidence that
the adenylation status might play a role in the
translatability of mRNAs (Jackson and Standart,
1990; Sachs, 1990; Gallie, 1991). Findings demonstrating that the poly(A) binding protein-poly(A) tail
complex facilitates the recruitment of the 40S
ribosomal subunit (Tarun and Sachs, 1995) combined
with other evidence (Munroe and Jacobson, 1990;
Christensen et al., 1987) suggest that the poly(A) tail
and the cap structure may physically interact to
facilitate initiation of translation. This hypothesis
may account for the fact that most of the evidence
assigns a positive role for the poly(A) tail in
translatability, the most overwhelming examples being
those from developmental systems (Bachvarova, 1992).
In these systems regulated cytoplasmic polyadenylation
of maternal mRNAs provides a control mechanism for
the translational activation of stored mRNAs at
various times during oocyte maturation and early
development, when little or no mRNA is synthesized
(Sheets et al., 1994, 1995; Salles et al., 1994).
On the other hand, there are cases where translational activation correlates with deadenylation. Even
though during Xenopus maturation and early embryogenesis where adenylation correlating with translational activation is the general rule, histone mRNAs
deadenylation coincides with their translational activation. At least, in the case of histone H4 mRNA the
deadenylation dependent maturation involves complete
removal of the poly(A) tail (Ballantine and Woodland,
1985). Similar results have been obtained in mammalian spermiogenesis in which shortening of the poly(A)
tail of the protamine mRNA occurs at the same time
as translation of this mRNA commences (Kleene,
1989).
Gra® et al. (1993) have shown that beta-1 interferon
mRNA with long poly(A) tails is not eciently
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
118
translated in vitro, in contrast to its oligoadenylated
counterpart. This e€ect is attributed to elements in the
3'UTR which cooperating with the poly(A) tail, inhibit
translation. Furthermore, IFN-b transcripts with
elongated poly(A) tails formed in vivo, are associated
with reduced translation of this message (Dehlin et al.,
1996).
We have previously shown that the rapid
deadenylation of c-myc mRNA leads to the formation
of a relatively more stable oligoadenylated c-myc
mRNA (Ioannidis et al., 1996). The question arises
whether this population represents an inactive degradation intermediate in the process of c-myc mRNA decay
or whether it serves a functional role. One way to
elucidate in vivo the function of oligoadenylated c-myc
mRNA was to use the polyadenylation inhibitor
cordycepin which is known to prevent the formation
of mRNA with long poly(A) tails but allows for the
correct splicing and transport of transcribed mRNA
(Zeevi et al., 1982). Addition of cordycepin had as a
result the dissociation of c-MYC protein from c-myc
mRNA levels. The discrepancy observed between
c-myc mRNA steady state levels and protein accumulation in cordycepin treated cells in comparison to the
control cells indicates that this mRNA is subjected to
translational regulation. Furthermore, it appears that
the short poly(A) tailed c-myc mRNA is very eciently
translated.
Results
Figure 1 Detection of a distinct population of c-myc mRNA
with short poly(A) tails. RNAse H mapping was performed on cmyc mRNA isolated from HeLa 1C5 cells using oligo 2 (lane 1),
oligo 1 (lane 2) and no oligo (control reaction, lane 3) after
induction with dexamethasone (1075 M for 1 h) and from resting
FS-4 cells after induction with 10% FCS for 1 h, using oligo 2
(lane 4). The brackets indicate the distribution of the poly(A) tail
carrying fragments derived from the major polyadenylation site
pA2 which extend over 471 nt when using oligo 1 (lane 2) and
411 nt when using oligo 2 (lanes 1 and 4). A distinct accumulation
of fragments was observed towards the smaller sizes as pointed by
the arrows. The positions of RNA molecular size markers are
indicated on the left
Enhanced presence of oligoadenylated c-myc mRNA
species in vivo
We have performed RNase H mapping of the 3' end
of c-myc mRNA which revealed that all the various
poly(A) tail sizes were not evenly represented (Figure
1). The species of c-myc mRNA whose presence was
more pronounced were those which possessed short
poly(A) tails, as indicated by the arrow. This
population was enhanced when overexpression of
c-myc mRNA was achieved in HeLa 1C5 cells upon
dexamethasone induced transcription of the c-myc
gene under the control of the MMTV-LTR. However,
the existence of a prominent amount of short tailed
c-myc mRNA was also demonstrated in the parental
cell line HeLa (data not shown) and in the human
diploid FS-4 ®broblastic cell line (Figure 1, lane 4)
which expresses normally regulated c-myc (Lin and
Vilcek, 1987). It has been reported that the 3'UTR
derived from unstable mRNAs direct rapid
deadenylation of chimaeric mRNAs leading to the
formation of oligoadenylated species which decay
further with ®rst order kinetics. Among those studied
the c-myc 3'UTR confers the fastest rate of
deadenylation upon the chimaeric mRNAs (Chen
and Shyu, 1994).
The enhanced representation of oligoadenylated
c-myc mRNA species can be attributed to the fact
that the decay of normal c-myc mRNA has been
shown to be a two step process (Swartwout and
Kinniburgh, 1989) of which the later appears to be rate
limiting (Chen and Shyu, 1994; Ioannidis et al., 1996).
The question arising from the above observations is
whether the oligoanenylated c-myc mRNA can be
translated or it represents an inactive decay intermediate.
E€ects of cordycepin addition upon c-myc mRNA
One way to elucidate the in vivo function of
oligoadenylated c-myc mRNA was to use the
polyadenylation inhibitor cordycepin which is known
to prevent the formation of mRNA with long poly(A)
tails but allows for the correct splicing and transport of
transcribed mRNA (Zeevi et al., 1982). Addition of the
polyadenylation inhibitor to the HeLa 1C5 cells,
resulted in a decrease of the steady state c-myc
mRNA levels but it did not a€ect the levels of the
stable GAPDH mRNA, within 3 h. The reduction of
c-myc mRNA levels was correlated to the amount of
inhibitor used (Figure 2A). At a concentration of
20 mg/ml a 4 ± 5-fold reduction in c-myc mRNA levels
could be measured.
In order to ascertain the expected alterations in the
structure and to follow the fate of c-myc mRNA
formed in the presence of the polyadenylation
inhibitor, RNase H mapping of the 3' end of c-myc
mRNA, isolated from cordycepin treated and control
cells at various time points after actinomycin D
addition, was performed. As shown in Figure 2 (B,C)
the c-myc mRNA is converted within 15 min to its
deadenylated form. Addition of cordycepin to the cells
resulted, as expected, in the loss of the c-myc mRNA
species with long poly(A) tails. Nevertheless, the
oligoadenylated c-myc mRNA synthesized in the
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
presence of cordycepin decays at similar rates as its
naturally formed counterpart, despite the modi®cation
at its 3' end (Figure 2).
Translational eciency of oligoadenylated c-myc mRNA
Addition of cordycepin did not a€ect c-myc mRNA
half life and thus, in the presence of the inhibitor, all
the available c-myc mRNA is oligo or de-adenylated
within a short time. Due to the fact that in mammals
the average half-life of stable mRNAs ± which
comprise more than 90% of total mRNAs ± is over
12 h, 3 h after cordycepin addition, the levels of the
majority of stable mRNAs synthesized before the
addition of the inhibitor were not grossly a€ected, as
exempli®ed by the levels of GAPDH mRNA. Thus,
cordycepin addition resulted within 3 h in the
reduction of c-myc mRNA steady state levels with
the remaining population consisting of the oligoadenylated form of the message. Given the fact that
c-MYC proteins are very unstable ± 30 min half life
(Hann and Eiseman, 1984) ± the protein detected 3 h
after cordycepin addition would be the translation
product of the mRNA synthesized in the presence of
Figure 2 E€ects of cordycepin addition upon c-myc mRNA. (A)
Northern blot of total RNA isolated from HeLa 1C5 cells 3 h
after cordycepin addition at the concentrations indicated. (Ctl:
control). (B.1). HeLa 1C5 cells were induced with dexamethasone
or dexamethasone in the presence of cordycepin (20 mg/ml). Three
hours later actinomycin D (5 mg/ml ®nal concentration) was
added and total RNA was isolated at the times indicated. RNase
H mapping was performed using oligo 2. (B.2). Graphic
representation obtained by scanning the autoradiogram at the
pA2 site of polyadenylation
the inhibitor. If the length of the poly(A) tail was
inconsequential to the rate of translation of c-myc
mRNA one would expect that the levels of c-MYC
protein would re¯ect the reduced steady state levels of
the corresponding mRNA observed upon addition of
cordycepin. If the shortening of the poly(A) tail had
an adverse e€ect upon the translatability of this
message then the levels of c-MYC protein should be
further diminished. To address this question Western
blot analysis was performed after addition of
cordycepin. The data in Figure 3 (A and B)
demonstrated that addition of cordycepin up to
60 mg/ml, despite the decrease in the steady state
c-myc mRNA levels, had no measurable e€ect upon
c-MYC protein levels. These ®ndings were con®rmed
with four di€erent monoclonal antibodies and one
polyclonal antibody recognizing di€erent epitopes of
the protein. The sole or the most prominent band
recognized by all of the above antibodies corresponded to a protein of approximately 64 kD
molecular weight. The recognition of this protein
was abolished when the incubation was performed in
the presence of the immunizing peptide (C-19).
Incubation in the presence of GST-c-MYC fusion
protein abolished (C-33) or diminished (C-8) binding
(Figure 3B). Thus, using all the speci®c antibodies no
changes in the c-MYC protein levels could be
measured with increasing concentrations of cordycepin, despite the gradual reduction in c-myc mRNA
levels. At a concentration of 20 mg/ml a 4 ± 5-fold
decrease in c-myc mRNA levels could be measured
(Figure 2A). At 40 mg/ml the reduction was estimated
to be larger than sevenfold. However, this is a rough
approximation since at this and higher cordycepin
concentrations, no accurate measurements could be
made ± overexposure of the autoradiograms was
inevitable in order to detect the remaining low levels
of c-myc mRNA, resulting in loss of linearity.
Similar results were obtained when cordycepin was
added to other established cell lines. In addition, the
expression of c-MYC proteins was determined after
the exposure of cells to physiological agents known to
decrease the c-myc mRNA steady state levels. It is
documented that IFN-a decreases c-myc mRNA
steady state levels up to 50% in Daudi cells (Dani
et al., 1985). IFN-a (1000 units/ml) was added for
24 h to Daudi cells and c-MYC protein levels in these
cells were compared to those of control cells and cells
exposed to cordycepin for 3 h. In ®gure 4 is shown
that cordycepin caused a dose dependent decrease in
c-myc mRNA levels which was not re¯ected in the
protein levels. On the other hand, IFN-a which
downregulates c-myc expression, without a€ecting the
distribution of poly(A) tail size of c-myc mRNA,
produced a concomitant decrease in protein levels.
Similar results (Figure 5) were also obtained with the
cell line HL-60 using the di€erentiation agent retinoic
acid which results in the reduction the c-myc mRNA
levels due to transcriptional inhibition (Krumm et al.,
1992).
These ®ndings indicate that the reduction of c-myc
mRNA steady state levels is re¯ected upon protein
levels when this is caused by physiological agents. On
the other hand, cordycepin resulted in the reduction of
c-myc mRNA levels and the enrichment of the
remaining message with short tailed species but did
119
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
120
not a€ect the levels of c-MYC protein. This observation may lead to the hypothesis that cordycepin
stabilizes the c-MYC protein. To test this possibility,
we followed the protein decay process after addition of
cycloheximide in control and cordycepin treated cells.
The data in Figure 6 did not reveal any stabilizing
e€ect of cordycepin upon c-MYC proteins since they
decayed at comparable rates in both control and
cordycepin treated cells.
The above indicate that cordycepin addition resulted
in the dissociation between c-myc mRNA and protein
levels.
E€ect of cordycepin addition upon c-fos expression
In order to ascertain whether the e€ect of the
adenylation inhibitor upon c-myc expression was
generalized and a€ected other genes in a similar
manner, we studied the expression of c-fos in the
presence of cordycepin. c-fos is a tightly regulated (at
the level of transcription) early response gene, coding
for short lived mRNA (Greenberg and Zi€, 1984) and
protein products (Kovary and Bravo, 1991) and its
expression is induced by many growth factors present
in serum. HeLa 1C5 cells proliferate even at low serum
Figure 3 Cordycepin addition to HeLa 1C5 cells did not a€ect the levels of c-MYC proteins. (A) Northern blot analysis of RNA
isolated form HeLa 1C5 cells 3 h after the addition of 40 or 60 mg/ml cordycepin. (B) Western blot analysis from HeLa 1C5 control
cells (ctl) and cells treated for 4 h with 20, 40 and 60 mg/ml cordycepin using ®ve di€erent MYC speci®c antibodies. The stable bcl-2
protein was used as an internal reference. The antibodies used are indicated on the left of the blot. ctl*: lysate from control cells
treated with antibody which had been pre-incubated with the corresponding immunizing peptide (C-19) or the GST-c-myc protein
(C-33, C-8) as described in Materials and methods
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
121
concentrations. However, it has been shown that unlike
c-myc whose levels are independent of cell cycle stage
(Thompson et al., 1985; Hann et al., 1985), c-fos
mRNA in cycling cells is barely detectable but its
induction is as sensitive to growth factors as in
quiescent cells (Bravo et al., 1986). Thus, high levels
of c-fos expression were achieved in HeLa 1C5 cells by
FCS addition to cells maintained at low serum
concentration for 48 h. As shown in Figure 7 lower
levels of c-fos mRNA were induced in the presence of
cordycepin resulting in lower protein levels as well.
Under the same conditions c-MYC protein levels were
not a€ected by cordycepin. This ®nding excludes the
possibility that the e€ect of cordycepin upon c-myc is
due to an indiscriminate e€ect upon the translational
machinery.
The data presented show that cordycepin addition to
cell cultures resulted in the dissociation of c-myc
Figure 4 E€ect of IFN-a or cordycepin addition to Daudi cells.
To exponentially growing Daudi cells 20 and 40 mg/ml cordycepin
were added for 3 h and IFN-a for 24 h (1000 U/ml). Total cell
lysates were prepared at a density of 56103 cells/ml. (A) Western
blot analysis of c-MYC proteins from control cells (ctl) and cells
treated with cordycepin or IFN-a using the monoclonal
antibodies indicated. (B) c-MYC levels were determined in equal
volumes of extracts from cells treated with cordycepin or IFN-a
and increasing amounts (titration) of control cell extracts using
the monoclonal antibody indicated. Schematic representation of
the results obtained by densitometry (*; control, ~; 20 mg/ml
cordycepin, &; 40 mg/ml cordycepin, ^; IFN-a). (C) Northern
blots of total RNA from Daudi cells treated with cordycepin for
3 h. (D) Total RNA (30 mg) from control cells and total RNA
(60 mg) from cells treated with IFN-a (1000 U/ml) was subjected
to RNase H analysis of the 3' end of c-myc mRNA. No di€erence
in the size distribution of the poly(A) tails between control and
IFN-a treated cells was detectable
Figure 5 E€ects of retinoic acid and cordycepin (cd) upon c-myc
mRNA and c-MYC protein levels in HL-60 cells. Exponentially
growing HL-60 cells were exposed to 10 mM ®nal concentration
retinoic acid (RA) for 24 h before RNA and protein isolation.
Alternatively exponentially growing HL-60 cells were exposed to
cordycepin (20 or 40 mg/ml ®nal concentration) and RNA was
isolated 3 h later while protein was collected 4 h after cordycepin
addition. Total cell lysates were prepared at a density of 66103
cells/ml. (A.1) Northern blot analysis of c-myc mRNA from cells
treated either with cordycepin or retinoic acid. (A.2) RNase H
analysis of the 3' end of c-myc mRNA from control and from cells
treated with cordycepin (20 mg/ml) and (A.3) densitometric scan of
the poly(A) tail size distribution. (B.1) Western blot analysis of cMYC protein levels. c-MYC levels were determined ± using the
extract volume indicated ± from cells treated with cordycepin or
retinoic acid and increasing extract volumes (titration) of control
cells. The antibody used is indicated on the left of the blot. (B.2)
Schematic representation of the results obtained by densitometry
(&; control, *; 40 mg/ml cordycepin, ~; RA)
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
122
mRNA/protein levels. Since this could not be
attributed to changes in the stability of the c-MYC
protein, it is indicative of translational control. Based
on the assumption that the ratio protein synthesized:
template mRNA re¯ects the translational eciency of a
given message, it appears that the oligoadenylated cmyc mRNA synthesized in the presence of cordycepin
was more eciently translated than its counterpart
with a heterogeneous size distribution of the poly(A)
tail. A four- to ®vefold increase in the protein/mRNA
ratio can be estimated when cells are treated with
20 mg/ml cordycepin. The ratio greatly increases at
higher concentrations of the inhibitor even though the
magnitude cannot be quantitated, due to very low
levels of remaining c-myc mRNA. The possibility exists
also that the determining factor for c-MYC levels is
not mRNA availability but either inhibitory mechanisms controlling c-myc mRNA translation and/or other
factors allowing a quota of c-myc mRNA molecules to
be translated. However, the fact remains that
elimination of a long poly(A) tail does not thwart the
translation of this mRNA.
resulted in a decline in c-myc mRNA levels while the
remaining message was enriched in the oligoadenylated
form, in all cell lines tested. The cell line HeLa 1C5 in
which high levels of c-myc mRNA could be induced,
was used in order to be able to follow the metabolic
fate of the message. The c-myc mRNA synthesized in
the presence of cordycepin despite its 3' end
modi®cation was neither stabilized nor destabilized
but had an apparent decay rate similar to that of its
naturally formed counterpart derived through
deadenylation (Figure 2). The proportional to the
amount of the inhibitor used decrease in c-myc mRNA
steady state levels could be attributed to transcriptional
inhibition (Beach and Ross, 1978). Within 3 h, stable
mRNAs representing the vast majority of the cellular
messages are not grossly a€ected by the addition of
cordycepin and the cell should not be depleted of
translatable mRNA, as exempli®ed by the levels of
Discussion
Study of the distribution of the poly(A) tail length of
c-myc mRNA has revealed a distinct prevailing
population with short poly(A) tails, derived through
sequential deadenylation. To elucidate whether this
population serves a function or whether it represents
an inactive degradation intermediate, the polyadenylation inhibitor cordycepin was used to further enrich the
oligoadenylated fraction of c-myc mRNA and observe
the e€ects upon c-MYC protein levels. It has been
demonstrated that the poly(A) de®cient mRNA
synthesized in the presence of cordycepin is correctly
spliced, transferred to the cytoplasm and enters the
polyribosomes (Zeevi et al., 1982). Cordycepin addition
Figure 6 Cordycepin addition did not a€ect the stability of cMYC proteins. The stability of c-MYC proteins was compared in
control and cells treated with cordycepin (60 mg/ml) for 3 h.
Cycloheximide (10 mg/ml) was added and the levels of c-MYC
protein were determined at the times indicated. The antibodies
used are indicated at the left side of the blots
Figure 7 E€ect of cordycepin addition on c-fos expression. HeLa
1C5 cells were kept for 48 h in medium containing 0.25% FCS. cfos induction was achieved by the addition of FCS (®nal
concentration 20%). Cordycepin was added (20 mg/ml) to the
cells 1 h before induction. Total RNA was isolated 30 min post
induction and protein at the times indicated. (A) Northern blot
analysis of c-fos mRNA levels. (B) Western blot analysis of c-FOS
protein levels. Lane 1, Control; Lanes 2 ± 7, FCS induced cells
collected at the times indicated. Lanes 3, 5 and 7: cells treated
with cordycepin. (C) Western blot analysis of c-MYC protein
levels after induction of serum starved cells by the addition of
FCS Lane 1: Control, Lane 2: FCS induced cells Lane 3: FCS
induced cells in the presence of 20 mg/ml cordycepin
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
GAPDH mRNA. Under these conditions the translatability of oligoadenylated c-myc mRNA is determined
in a competitive milieu. c-MYC proteins remained
highly unstable in the presence of cordycepin, thus the
protein detected 3 h after cordycepin addition and
thereafter re¯ected the translation product of the
oligodenylated message.
We observed that in the presence of cordycepin
c-MYC proteins do not follow the marked decrease in
c-myc mRNA levels. This was not proven to be an
indiscriminate e€ect of cordycepin since the reduction
in c-fos mRNA was re¯ected upon c-FOS protein.
Discrepancies between mRNA levels and protein
accumulation such as those observed with c-myc
mRNA in the presence of cordycepin are considered
as indicative of translational control.
The results presented above may be due to the fact
that only a small prescribed number of c-myc mRNA
molecules is translated at a given time. In that case the
reduction of c-myc mRNA which became dramatic at
high cordycepin concentrations, must not have fallen
under the threshold of the mRNA molecules meant to
be translated and thus, no changes in the protein levels
could be detected. The above hypothesis seems to be
contradicted by the data indicating that the reduction
in c-myc mRNA steady state levels caused by
physiological agents is re¯ected upon protein levels.
However, the changes in the physiology of the cell
induced by those agents should be taken into
consideration. From that view point the e€ect of
IFN-a upon Daudi cells and retinoic acid upon HL-60
cells, i.e. growth arrest and di€erentiation respectively,
may dictate the concomitant decrease in c-myc mRNA
and c-MYC protein levels. If c-myc mRNA is subjected
to translational control, changes in the physiological
conditions may modulate this mechanism either
augmenting it or relaxing it. Translational enhancement during the recovery phase of c-myc mRNA and
protein levels after induction of di€erentiation in MEL
cells, has been suggested before (Spotts and Hann,
1990). Additional examples have been demonstrated
for other mRNAs whose expression is translationally
regulated (Bernstein et al., 1995; Han et al., 1990).
Proto-oncogenes may be genuine candidates for
translational repression. For many proto-oncogenes'
mRNAs, the presence of a long 5'UTR burdened with
AUG initiation codons or high GC content led to the
prediction that they must be ineciently translated
(Kozak, 1991). There are observations as well as
experimental evidence indicating that removal of
portions of these sequences dramatically improves the
expression of these mRNAs (Rao et al., 1988; Marth et
al., 1988). Evidence of increased mobilization of c-myc
mRNA to polysomes due to a mutation in the 5'UTR
in multiple myeloma cell lines suggests that this mRNA
is not eciently translated (Paulin et al., 1996). c-myc
5'UTR, which is highly conserved among di€erent
species, has been proposed to exert negative translational control due to the formation of highly stable
secondary structure which inhibits the scanning
mechanism of translation as suggested by in vitro
experiments (Darveau et al., 1985). This inhibitory
e€ect can be circumvented by the presence of an
internal ribosome entry segment (IRES) identi®ed in
the 5' c-myc UTR which allows for cap independent
translation (Nanbru et al., 1997; Stoneley et al., 1998).
Translational repression or regulated translatability
may be exerted through the 3'UTR, as well (Yang et
al., 1997). This has been demonstrated using chimaeric
mRNAs carrying the 3'UTR of c-fos, GM-CSF, bInterferon (Kruys et al., 1988, 1989) or hu-TNF (Han
et al., 1990). Translational repression can be relaxed
due to alterations in cellular physiology, as has been
demonstrated for MEF2A (Black et al., 1997) and
lipoprotein lipase mRNA (Ranganathan et al., 1997).
In the latter case a regulatory trans factor recognizing
the 3'UTR has been identi®ed.
Translational regulation of c-myc mRNA expression, could be operative if a predetermined number of
molecules is `tagged' to be translated or through
feedback inhibition. Tagging for translation by a
cellular factor present in limited amounts has been
implied for ornithine decarboxylase mRNA (Lorenzini
and Sche‚er, 1997). Our original goal when using
cordycepin was to prevent the formation of long
poly(A) tails, to enrich the short tailed c-myc mRNA
population and investigate its function. This approach
would allow us to determine the translatability of this
predominant c-myc mRNA species in vivo and in a
competitive milieu. A variety of di€erent experimental
approaches have indicated that in addition to being a
modulator of mRNA turnover, the poly(A) status of
mRNA can be a determinant of mRNA translational
eciency. It has been postulated that this is achieved
by the interaction of the poly(A) binding protein PAB1
with the cap binding protein via the eIF-4G which
stimulates the recruitment of the 40 S ribosomal
subunit (Sachs et al., 1997). The eIF-4 holoenzyme is
instrumental in cap dependent initiation of translation
and its concentration is limiting. Furthermore, the
concentration requirement is considered to be positively correlated to the putative degree of secondary
structure at the 5' end of the mRNA (Koromilas et al.,
1992). On the other hand it is believed that mRNAs
translated in a cap independent manner require very
little if any eIF-4 (reviewed in: Jackson et al., 1995).
Even though the overwhelming majority of examples
attribute a positive role for the poly(A) tail in
translatability there has been evidence that for certain
mRNAs the oligoadenylated species represents the
translationally active form. Poly(A)+ vasopressin
mRNA upon injection to diabetic rats was 10-fold
less e€ective in eliciting physiological response and
measurable protein levels in plasma, compared with
poly(A)-mRNA. Poly(A) tail removal restored functional activity (Maciejewski-Lenoir et al., 1993). In
vitro studies suggest that a long poly(A) tail prevents
translation of human Beta-1 Interferon mRNA.
However, translational eciency was considerably
improved when the poly(A) tract was shortened to 11
A residues. This ®nding was attributed to the presence
of the element UUAUUUAU in the 3'UTR of this
message which co-operates with a long poly(A) tail to
inhibit translation (Gra® et al., 1993). This element is
responsible for the inhibitory role upon translation
which has been also demonstrated for c-fos, GM-CSF
and TNF 3'UTRs. This octanucleotide, or the core
sequence AUUUA, is a common feature in the 3'UTR
of unstable mRNAs and was initially identi®ed as an
instability element directing rapid deadenylation (Shaw
and Kamen, 1986; Zubiaga et al., 1995). The c-myc
mRNA contains two such elements which may link
123
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
124
deadenylation to translation, inhibiting the latter until
the message is deadenylated. It has been demonstrated
that IFN-b mRNA with elongated poly(A) tails,
formed in vivo, in virus infected cells, is not eciently
translated (Dehlin et al., 1996). The strategy utilized by
the virus to downregulate the defence speci®c protein
supports the in vitro data that this message is translated
in its oligoadenylated form.
The e€ect of the polyadenylation inhibitor on the
length of the poly(A) tail may account for the
increased translatability of c-myc mRNA. One
posibility is that the lack of poly(A) tails in the
newly synthesized c-myc mRNA thwarts the poly(A)
assisted cap dependent translation ± which would be
inecient for c-myc mRNA due to its highly
structured 5'UTR ± and promotes the utilization of
the highly ecient IRES. Another posibility is that a
long poly(A) tail inhibits translation. It has been
reported that a poly(A) tail longer than 80 nt is
bound simultaneously with the 3'UTR of c-myc
mRNA by Elav like proteins in vitro (Ma et al.,
1997) Such an interaction may impair the stimulatory
e€ect of the poly(A) tail upon cap depended
translation. Shortening of the poly(A) tail would
release the binding and allow it to resume its
function. This may imply that the more stable than
its precursor, naturally formed oligoadenylated c-myc
mRNA represents the translationally active form of
this message as is indicated for certain mRNAs. In
that case the rapid c-myc mRNA deadenylation
leading to relatively more stable oligoadenylated
intermediates may represent a prerequisite for the
translational maturation of this message. At the same
time the message is irrevocably on its way to its ®nal
degradation. It has been documented that regulation
of expression of the c-myc gene is e€ected by
modulating the half-life of its mRNA. Considering
the above, any modulation of the rate of
deadenylation, even though it may determine the
onset of the decay of the molecule, is not expected
to a€ect the functional half-life of this message.
Therefore, modi®cation of the functional half-life of
c-myc mRNA would either involve alterations in the
decay rate of the oligoadenylated form or an
alternative mechanism would be required for the
deactivation of this mRNA. There is evidence for an
endonucleolytic decay of this message without prior
deadenylation (Swartwout and Kinninburgh, 1989;
Ioannidis et al., 1996).
Materials and methods
Cell cultures
HeLa 1C5 cells were derived from a stable clone isolated
from HeLa cells transfected with the eukaryotic expression
vector pMSG-myc. Cell culture conditions and induction
with dexamethasone were performed as previously described (Ioannidis et al., 1996).
The human diploid ®broblastic cell line FS-4 was cultured
for 48 h in DMEM plus 10% FCS and subsequently in
DMEM with 0.25% FCS for 48 h before the induction with
fresh medium containing 10% FCS. Cordycepin (Sigma Co),
IFN-a2b (INTRON-A, Schering-Plough) or retinoic acid
(Sigma Co) were added to actively dividing Daudi and HL-60
cells, respectively.
Northern hybridization
Total RNA was isolated as described by Chomczynski and
Sacchi (1987). Thirty mg of denatured total RNA were
electrophoresed in 1.2% formaldehyde-agarose gels,
transferred to nylon membranes (Zetaprobe, BioRad),
immobilized and hybridized according to the manufacturer's instructions. Filters were hybridized with the
speci®c c-myc (Pr3), c-fos (ON 118) and GAPDH end
labelled deoxyoligonucleotides from Oncogene Science.
Mapping of the c-myc mRNA 3'UTR
Oligonucleotide/RNase H treatment: c-myc mRNA was
annealed to a speci®c DNA oligonucleotide and treated
with RNase H(Gibco ± BRL). The speci®c antisense
oligonucleotides used were CCTTACGCACAAAGAGTTCCG (oligo 1) and CAAGTTCATAGGTGATTGCTC
(oligo 2) as described in Ioannidis et al. (1996). RNase H
mapping was performed according to Brewer and Ross
(1990) i.e. 40 mg of total RNA were ethanol precipitated,
resuspended in 20 ml of 1 mM EDTA pH 7.4 and heated at
708C for 10 min. Oligonucleotide (0.5 mg) was added, and
the mixture was incubated at 208C for 15 min. One ml of
4 M KCl was added and the mixture was incubated at 208C
for 15 min, followed by the addition of 20 ml of TM bu€er
(40 mM Tris HCl, pH 7.5, 60 mM MgCl2). RNase H was
added (®nal concentration 20 units/ml) and the digestion
was performed at 378C. Following digestion RNA was
ethanol precipitated. RNA samples were separated by
length using a denaturing 4% polyacrylamide-urea gel as
described by Stoeckle and Guan (1993) and transferred
electophoretically to a charged membrane. The probe used
was the end labelled deoxynucleotide GGCTAAATCTTTCAGTCTCAAGACTCAGCCAAGGTTGTGAGGTTG.
Western blot analysis
Cells were lysed in RIPA bu€er (1% NP-40, 0.5% Na
deoxycholate, 0.1% SDS in PBS) containing the following
protease inhibitors: PMSF, 0.1 mg/ml, aprotinin, 50 mg/ml,
sodium orthovanadate, 1 mM. Protein concentrations were
determined according to Lowry et al (1950).
Cell lysates were electrophoresed on SDS-PAGE gels and
transferred electrophoretically to Immobilon-P (Millipore)
membranes. The antibodies used in this study were for the cMYC proteins: pan-myc, 0.5 mg/ml (OM-11-904, Cambridge
Res. Biochem.), 9E10, 1 mg/ml (OM-11-908, Cambridge Res.
Biochem.), C-33, 0.1 mg/ml (SC-42, Santa Cruz Biotech. Inc.),
C-8 0.05 mg/ml (SC-41, Santa Cruz Biotech. Inc.) and C-19
0.1 mg/ml (SC-788, Santa Cruz Biotech. Inc.), for c-FOS: cfos (4), 0.1 mg/ml (SC-52, Santa Cruz Biotech. Inc.) and for
the bcl-2 protein: Ab-1, 0.2 mg/ml (OP60, Oncogene Sc.).
Blocking of the primary antibody C-19 was performed
overnight at 48C using 10-fold by weight excess of the
immunizing peptide (SC-788P Santa Cruz Biotech. Inc.) for
C-19; for C-8 and C-33 a 10-fold by weight excess of the
E. coli produced GST-MYC fusion protein containing aa 1 ±
262 of c-MYC (SC-4084, Santa Cruz Biotech. Inc.). Western
blots were visualized with the ECL system by Amersham
according to the instructions of the manufacturer. The bcl-2
protein (MW 26 kD) proven stable, was used as an internal
standard in HeLa 1C5 cells. The ®lms were scanned using the
UMAX Scanner (Vista-S6) and were quantitated using the
Image 1.44 program.
Acknowledgements
We would like to thank Professor J Vilcek for providing
the FS-4 human cell line. This work was partially
supported by a grant from the Greek Ministry of Health
to TT.
Cordycepin and translational control of c-myc mRNA
P Ioannidis et al
125
References
Bachvarova RF. (1992). Cell, 69, 895 ± 897.
Ballantine JEM and Woodland HR. (1985). FEBS Lett., 180,
224 ± 228.
Beach R and Ross J. (1978). J. Biol. Chem., 253, 2628 ± 2632.
Bernstein J, She¯er I and Elroy-Stein O. (1997). J. Biol.
Chem., 270, 10559 ± 10565.
Black BL, Lu J and Olson EN. (1997). Mol. Cell. Biol., 17,
2756 ± 2763.
Brewer G and Ross J. (1990). Meth. Enzymol., 181, 202 ± 209.
Bravo R, Burckhardt J, Curran T and Muller R. (1986).
EMBO J., 5, 695 ± 700.
Chen CY and Shyu AB. (1994). Mol. Cell. Biol., 14, 8471 ±
8482.
Chen CY and Shyu AB. (1995). TIBS, 20, 465 ± 470.
Chomczynski P and Sacchi N. (1987). Analyt. Biochem., 162,
156 ± 159.
Christensen AK, Kahn LE and Bourne CM. (1987). Am. J.
Anat., 178, 1 ± 10.
Dani Ch, Mechti N, Piechaczyk M, Lebleu B, Jeanteur P and
Blanchard JM. (1985). Proc. Natl. Acad. Sci. USA, 82,
4896 ± 4899.
Darveau A, Pelletier J and Sonenberg N. (1985). Proc. Natl.
Acad. Sci. USA, 82, 2315 ± 2319.
Decker CJ and Parker R. (1994). TIBS, 19, 336 ± 340.
Dehlin E, von Gabain A, Alm G, Dingelmaier R and
Resnekov O. (1996). Mol. Cell. Biol., 16, 468 ± 474.
Gallie DR. (1991). Genes Dev., 5, 2108 ± 2116.
Gra® G, Sela I and Galili G. (1993). Mol. Cell. Biol., 13,
3487 ± 3493.
Greenberg ME and Zi€ EB. (1984). Nature, 311, 433 ± 438.
Han J, Brown T and Beutler B. (1990). J. Exp. Med., 171,
465 ± 475.
Hann SR, Thompson CB and Eiseman RN. (1985). Nature,
314, 366 ± 369.
Hann SR and Eiseman RN. (1984). Mol. Cell. Biol., 4, 2486 ±
2497.
Ioannidis P, Havredaki M, Courtis N and Trangas T. (1996).
Nucl. Acids Res., 24, 4969 ± 4977.
Jackson RJ and Standart N. (1990). Cell, 62, 15 ± 24.
Jackson RJ, Hunt SL, Reynolds JE and Kaminski A. (1995).
Curr. Top. Microbiol. Immunol., 203, 1 ± 29.
Jacobson A and Peltz SW. (1996). Annu. Rev. Biochem., 65,
693 ± 739.
Kleene KC. (1989). Development, 106, 367 ± 373.
Koromilas A, Lazaris-Karatzas A and Sonenberg N. (1992).
EMBO J., 11, 4153 ± 4158.
Kovary K and Bravo R. (1991). Mol. Cell. Biol., 11, 2451 ±
2459.
Kozak M. (1991). J. Cell. Biol., 115, 887 ± 903.
Krumm A, Meulia T, Brunvand M and Groudine M. (1992).
Genes Dev., 6, 2201 ± 2213.
Kruys V, Marinx O, Shaw G, Deschamps J and Huez G.
(1989). Science, 245, 852 ± 854.
Kruys V, Wathelet MG and Huez GA. (1988). Gene, 72,
191 ± 200.
Lin J-X J and Vilcek J. (1987). J. Biol. Chem., 262, 11908 ±
11911.
Lorenzini EC and Sche‚er IE. (1997). Biochem. J., 326,
361 ± 367.
Lowry OH, Rosebrough NJ, Farr AL and Randall RJ.
(1950). J. Biol. Chem., 193, 265 ± 275.
Ma WJ, Chung S and Furneaux H. (1997). Nucl. Acids Res.,
25, 3564 ± 3569.
Maciejewski-Lenoir D, Jirikowski GF, Sanna PP and Bloom
FE. (1993). Proc. Natl. Acad. Sci. USA, 90, 1435 ± 1439.
Marth JD, Overell RW, Meier KE, Krebs EG and Permutter
RM. (1988). Nature, 332, 171 ± 173.
Munroe D and Jacobson A. (1990). Mol. Cell. Biol., 10,
3441 ± 3455.
Nabru C, Lafon I, Audigier S, Gensac M-C, Vagner S, Huez
G and Prats A-C. (1997). J. Biol. Chem., 272, 32061 ±
32066.
Paulin F, West MJ, Sullivan NF, Whitney RL, Lyne L and
Willis AE. (1996). Oncogene, 13, 505 ± 513.
Ranganathan G, Vu D and Kern PA. (1997). J. Biol. Chem.,
272, 2515 ± 2519.
Rao CD, Pech K, Robbins KC and Aaronson SA. (1988).
Mol. Cell. Biol., 8, 284 ± 292.
Ross J. (1995). Microbiol. Rev., 59, 423 ± 450.
Sachs A. (1990). Curr. Opin. Cell. Biol., 2, 1092 ± 1098.
Sachs A and Wahle E. (1993). J. Biol. Chem., 268, 22955 ±
22958.
Sachs A, Sarnow P and Hentze M. (1997). Cell, 89, 831 ± 838.
Salles FJ, Lieberfarb ME, Wreden C, Gergen JP and
Strickland S. (1994). Science, 266, 1996 ± 1999.
Shaw G and Kamen R. (1986). Cell, 46, 659 ± 667.
Sheets MD, Fox CA, Hunt T, Van De Wounde G and
Wickens M. (1994). Genes Dev., 8, 926 ± 938.
Sheets MD, Wu M and Wickens M. (1995). Nature, 374,
511 ± 516.
Spotts GD and Hann SR. (1990). Mol. Cell. Biol., 10, 3952 ±
3964.
Stoeckle M and Guan L. (1993). BioTeckniques, 15, 230 ±
231.
Stoneley M, Paulin F, Le Quesne J, Chappell S and Willis
AE. (1998). Oncogene, 16, 423 ± 428.
Swartwout SG and Kinninburgh AJ. (1989). Mol. Cell. Biol.,
9, 288 ± 295.
Tarun SZ and Sachs AB. (1995). Genes Dev., 9, 2997 ± 3007.
Thompson CB, Challoner PB, Neiman PE and Groudine M.
(1985). Nature, 314, 363 ± 366.
Yang Q, McDermott PJ, Duzic E, Pleij WA, Sherlock JD and
Lanier SM. (1997). J. Biol. Chem., 272, 15466 ± 15473.
Zeevi M, Nevins JR and Darnell Jr JE. (1982). Mol. Cell.
Biol., 2, 517 ± 525.
Zubiaga AM, Belasco JG and Greenberg ME. (1995). Mol.
Cell. Biol., 15, 2219 ± 2230.