From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
Effects of Iron-Depletion on Cell Cycle Progression in Normal Human T
Lymphocytes: Selective Inhibition of the Appearance of the
Cyclin A-Associated Component of the p33cdk2 Kinase
By Joseph J. Lucas, Agota Szepesi, Joanne Domenico,
Kozo Takase, Attila Tordai, Naohiro Terada,
and Erwin W. Gelfand
Iron removal bythe chelating-agent deferoxamine (DFO) arin late
rests cell cycle progression activated
of
human cells
T
G1 phase, before the G l / S border. The effects of the drug
on molecules that regulate progression through the cell cycle were defined. DFO (10 pmol/L) inhibited induction of
transcription of the cdc2 gene, but had no effect on accumulation of cdk2, cdk4, or interleukin (IL)-2 transcripts. No detectable ~ 3 4 protein
~ "
accumulated, but synthesis of the
p33protein was begun. It accumulated to normal levels
during the first 20 to 30 hours of incubationin the presence
of DFO. Furthermore,~
3 was3 activated
~
as
~ anH I histone
kinase. As active p33primarily represents complexes of
the p33 protein with cyclin E or cyclin A, the effects of DFO
on these cyclins were examined. Althoughthe induction of
synthesis and early accumulation of cyclin E and cyclin Eassociated kinaseactivity appeared normal,
the appearance
of cyclin A and cyclin A-associated kinase
activity were inhibited by DFO.However, the productionofcyclin A mRNA
appeared to be normal in the presence of DFO.
A major effect
of DFO in blocking cell cycle progression may be mediated
through inhibitionof the appearance of cyclin
A proteinand,
of ~
3 activity.
3
The
~ results
~
therefore, a major component
also indicate that the p33Ddm/cyclinE activity produced in
of the
the presence of DFO was not sufficient for completion
G1 phase of the cell cycle.
0 1995 by The American Society of Hematology.
T
tion of resting T cells and reaches a maximal level about 5
hours later, well before maximal expression of L - 2 receptor
mRNA," a requirement for iron in the first G1 phase of
activated T cells was not unexpected. We, therefore, initiated
studies to determine the effects of DFO on cell cycle regulatory events known to occur in mid-to-late G1 phase of the
T-cell cycle.
We previously characterized the DFO arrest point with
respect to several molecular events, using normal human T
lymphocytes activated to enter the cell cycle in vitro.''"3 In
the presence of DFO, the cellular RNA content accumulated
to an amount less than that characteristic of late G1 phase.
The synthesis of the key cell cycle regulatory kinase, ~34'~"',
which normally begins just before S phase entry, was inhibited; and the p1 loRbprotein, which plays a role in G1 phase
progression, was only partially phosphorylated. In additional
studies, we showed that treatment of a neuroblastoma cell
line with DFO inhibited normal increases in the cellular
content of ~ 3 4 " ~protein
"
and kinase activity."
The chain of events that occurs throughout G1 phase in
animal cells has been defined in some detail (reviewed by
Reed," and Shed'). Of major importance has been the discovery that the ~ 3 4 " ~kinase
"
is but one member of a family
of kinases, each of which, complexed with an appropriate
cyclin molecule, appears toplay a particular rolein cell
cycle
Using this new information, we have
further defined the point of DFO-induced cell cycle arrest
in mitogen-stimulated human lymphocytes. It is shown that
iron chelation inhibits transcription of the cdc2 gene but has
little or no effect on expression of the cdk2 gene. As shown
elsewhere, activation of the cdk2 and cdc2 genes normally
occurs at about 8 hours (in early G1 phase) and 24 hours
(just before S phase entry), respectively, after activation of
human T cells. 1327~28~28aIn activated human T cells, the products of these genes, the p33'"' and ~34'~''kinases, are first
detectable as active enzymes in mid-G 1 phase and at mitosis,
respectively.2Ra
Appearance of ~ 3 3 ' ~kinase
~ ' activity occurs
in two stages, with a modest increase occurring throughout
late G1 phase and a more dramatic increase (about 10-fold
higher) beginning after entry into S phase. The appearance
of this second wave of activity is markedly reduced by iron
HE DRUG DEFEROXAMINE (DFO), whichis frequently used to reduce iron content in individuals with
an overabundance the
of
essential
can also potentially result in undesirable effects on cells of the immune
~ y s t e m . ~It. ~has also been observed thatiron depletion,
through the use of chelators, can quickly inhibit the growth
of some cell types, with lymphocytes and cells of neuronal
origin being especially ser~sitive.~"'To further elucidate the
role of iron in cell cycle progression, the point of action of
the chelator has been defined.
In earlier work, we demonstrated that iron depletion arrests the growth of activated T lymphocytes before S phase
entry.'"13 Growth arrest could be seen in vitro with a dose
of DFO (10 pmol/L) that is achieved in humans in clinical
application^,'^.^^ a fact which may explain its immunosuppressive effects. As iron is required for the activity of ribonucleotide reductase,16it at first seemed likely, as we proposed
earlier,7 that iron depletion affected cell cycle progression
by blocking activity of this key enzyme. However, we and
others noted that iron-chelating agents, such as DFO or mimosine, in fact, arrested cells at a point in G1 phase, demonstrably earlier than the GUS bo~ndary,""~.'~
that can be
defined using the agent aphidicolin, which directly binds to
and inhibits DNA polymerase.'*~19As transferrin receptor
mRNA becomes detectable as early as 1 hour after stimulaFrom the Division of Basic Sciences, Department of Pediatrics,
National Jewish Center for Immunology and Respiratory Medicine,
Denver, CO.
Submitted November 14, 1994; accepted May 12, 1995.
Supported by American Cancer Society Grant No. IM-746 and
National Institutes of Health Grants No. AI-26490, AI-29704, and
AI-29903.
Address reprint requests to Joseph J. Lucas, PhD, Department of
Pediatrics, National Jewish Center for Immunology and Respiratory
Medicine, 1400 Jackson St, Denver, CO 80206.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1995 by The American Society of Hematology.
0006-4971#5/8606-0015$3.00/0
2268
Blood, Vol 86, No 6 (September 15).
1995:
pp 2268-2280
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CELLCYCLE ARREST BY IRON DEPLETION
2269
depletion. The first component, due to a complex of ~ 3 3 ' ~ ~ ' 2 x IO7 cells in 300 mmol/L sodium acetate (pH 5.5), 10 mmol/L
EDTA (pH 8.0), 2% sodium dodecyl sulfate (SDS), and 1 mg/mL
and cyclin E, appears in nearly normal amounts and with
heparin sodium (Grade I; obtained from Sigma), which was added
normal kinetics. However, the complete inhibition of cyclin
as an RNase inhibitor. Five microgram amounts of RNA were elecA synthesis, combined with a block to the further S-phasetrophoresed in a 1% formaldehyde-agarose gel, blotted onto a nylon
associated accumulation of the ~ 3 3 and
' ~cyclin E proteins,
membrane (MSI, Westboro, MA), and hybridized with multiple
prevents normal induction ofthesecondwaveof
~
3
3
~
~
primed-radiolabeled
probes, as described previo~sly.'~
Equal loading
activity.
of gel lanes was confirmed by monitoring 28s and 18s ribosomal
The system of DFO-arrested T cells thus permits examinaRNAs after ethidium bromide-staining of the gels. The probes used
tion of the differential roles and effects of the cyclin E- and
here were the excised inserts from plasmids (pSE999 and pSE1000)
containing full-length cDNA copies of the human ~ 3 4 ' ~and
' p3YdkZ
cyclin A-associated p33cdk2
kinases. Furthermore, as overpromRNAs, as described by Elledge and Spottswood?' For detection
duction andor activation of cyclins and cyclin-dependent
of cdk4 transcripts, the excised insert from a full-length human PSKkinases has been increasingly observed in certain tumors and
J3/Cdk4 cDNA probe:8 provided by Dr S.K. Hanks (Duke Univertumor-derived cell line^,"*'^-^^ the possibility of using ironsity, Durham, NC), was used. Interleukin (L)-2 mRNA was detected
chelating agents to target cells with such aberrations is
as described previously." For detection of cyclin A and E transcripts,
raised.
excised inserts from plasmids containing the human cyclin A and
cyclin E sequences,j9,@which were provided by Drs Tony Hunter
(Salk Institute, San Diego, CA) and Steven Reed (Scripps Research
Institute, Lalolla, CA), respectively, were used.
Cell preparation and culture. Peripheral blood T cells isolated
Immunoblot analysis. For immunoblot analysis, cells were
from healthy volunteers were used in all experiments. Mononuclear
washed with PBS and then lysed at 4°C in a buffer containing 25
cell suspensions were prepared by Ficoll-Hypaque gradient centrifugation, and T cells (EC) were obtained by E-rosette enri~hment.3~ mmoVL Tris-HC1 pH7.4,50 mmol/L NaCl, 0.5% sodium deoxycholate, 2% Nonidet P-40, 0.2% SDS, 1 mmoVL phenylmethylsulfonyl
This population contained greater than 95% T cells as determined
fluoride (PMSF), 50 pg/mL aprotinin, 50 pmol/L leupeptin, and 100
by immunofluorescence with anti-CD3 antibody. Cells were cultured
pmovL sodium orthovanadate." After removal of cellular debris
in RPM1 medium containing 10% (voVvol) fetal calf serum and 2
by centrifugation, lysates were prepared for electrophoresis, and
mmoVL glutamine. When using normal lymphocytes, experiments
polyacrylamide gel electrophoresis in the presence of SDS (SDSare routinely repeated at least three times using cells from different
PAGE) was performed as described by Laemmli!' After electrophodonors to establish the generality and reproducibility of the findings.
retic transfer of proteins to nitrocellulose membranes and blocking
Representative results are presented in the figures. T cells were
of the filters with a solution containing bovine serum albumin:'
activated by addition of phorbol dibutyrate (PDB; 10 nmoVL) and
filters were incubated with the appropriate primary antibody, as folionomycin (0.5 pmol/L). PDB, obtained from Sigma (St Louis, MO),
lows: for detection of ~34'~'', a mouse monoclonal antibody directed
was dissolved at a concentration of IO-' molL in dimethyl sulfoxide.
against a peptide comprising the C-terminal 15 amino acids of the
Ionomycin, obtained from Calbiochem (San Diego, CA) was preprotein (Zymed, San Francisco, CA); for p33"=, a rabbit polyclonal
pared as a 5 mmol/L stock solution in dimethyl sulfoxide. DFO was
antibody directed against a peptide comprising the C-terminal 12
obtained from Ciba-Geigy (Mississauga, Canada) and prepared as a
amino acids of the protein (UBI, Lake Placid, NY); for cyclin A, a
stock solution in distilled water at a concentration of 200 mmofi.
rabbit polyclonal antibody directed against a peptide comprising a
Metabolic labeling of cellular proteins was performed by preincubatC-terminal portion of the protein (residues 359-432; UBI); for cyclin
ing cells in methionine-free medium containing dialyzed fetal calf
E, a mouse monoclonal antibody (HE 12) prepared using recombiserum for 30 minutes, adding 150 pCi/mL of %methionine (spenant cyclin E protein (Pharmingen, San Diego, CA); and for pllORb,
cific activity, 1,200 Ci/mmol; Dupont NEN, Wilmington, DE), and
a mouse monoclonal antibody (1 F8) prepared using recombinant
incubating for the times indicated.
Simultaneous pow cytometric determination of DNA and ~ 3 4 ' ~ ' ' Rb protein (Santa Cruz Biotech, Santa Cruz, CA). Proteins were
detected by an enhanced chemiluminescence (ECL) method using
content The method described by Takase et a135was used. After
reagents and procedures provided by Amersham International ( A r a 60-minute fixation with 70% ethanol at 4"C, cells (about lo6)
lington Heights, IL). The secondary antibodies used were either a
were pelleted and washed. After a 5-minute incubation in 250 p1 of
sheep antimouse or donkey antirabbit antibody, linked to horseradish
phosphate-buffered saline (PBS) containing 0.5% Tween 20, 0.5%
peroxidase.
bovine serum albumin, and 1.5% human y-globulin (Sigma) at room
Immunoprecipitation of metabolically (3SS-methionine)labeled
temperature, 2 pg of mouse monoclonal antihuman ~ 3 4 " antibody
~'
proteins. After labeling of cells, as described above, they were
(Zymed, San Francisco, CA) was added, and the mixture was incucollected by centrifugation and washed twice with cold PBS. Cells
bated at 4" for 120 minutes. This antibody, directed against a peptide
were disrupted by a 15-minute incubation in RIPA buffer (150 mmoV
comprising the C-terminal 15 amino acids does not appear to crossL NaCl, 50 mmoVL Tris pH 7.4, 0.1% Triton X-100, and 0.5%
react with any of the other known cdks present in T cells (Lucas et
sodium deoxycholate) containing 1 mmoVL PMSF, 10 pg/mL leual, unpublished data, June 1994). The cells were washed and further
peptin, 20 pg/mL aprotinin, 0.1 mmoUL sodium orthovanadate, and
treated with 3 p1 of a fluorescein isothiocyanate (F1TC)-conjugated
F(ab')* fragment of affinity-purified sheep antimouse immunoglobu50 mmolfL sodium fluoride?' Cell lysates were clarified by centrifulin (Sigma) at 4°C for 60 minutes. The cells were washed and then
gation, and supernatants were incubated with Protein G coupled to
incubated with 0.5 mL of 0.25 mg/mL ribonuclease (type 111-A,
Sepharose beads (Zymed) for 30minutes at 4°C.After removal of the
Sigma) in phosphate-buffered saline (PBS) at 37°C for 10 minutes.
Protein G-beads, supernatants were incubated with the appropriate
The cell suspension was mixed with 0.5 mL of a solution of propidspecific primary or control antibody for 2 hours at 4°C.For detection
ium iodide (50 p g / d in PBS) and, after 60 minutes, was analyzed
of the ~ 3 3protein,
' ~ ~the precipitating antibody used was the rabbit
using a flow cytometer (EPICS Profile; Coulter, Hialeah, K),colpolyclonal antibody described above; control antibody was IgG oblecting green fluorecence (520 to 540 nm) andred fluorescence
tained from an irrelevantly immunized rabbit. A 10-pg amount of
(greater than 600 nm) in linear scales with 488 nm excitation.
primary antibody was used for an extract prepared from IO' cells,
Northern blot analysis. Total cellular RNA was extracted from
an amount determined to result in near complete recovery of the
MATERIALS AND METHODS
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
2270
LUCAS ET AL
antigen. Immunocomplexes were recovered by binding toprotein
G-beads. Complexes were washed five times with RIPA buffer before being suspended in electrophoresis sample buffer.";" Samples
were heated for 5 minutes at 9 0 T , and beads were pelleted by
centrifugation and discarded. The denatured proteins were resolved
by SDS-PAGE and visualized by fluorography."."
Immunoprecipitation of activecyclin-dependentkinase/cyclin
complexes. For isolation of active complexes for in vitro kinase
assays, the method of Howe et
wasused,with some modifications. Cells (2 X IO6 to S X 10') were washed with PBS and then
lysed in a buffer containing S0 mmol/L Tris-HC1 pH 7.4, S mmol/
L EDTA, 250 mmol/L NaC1, 1 mmol/L PMSF, 50 pg/mL aprotinin,
50 pmoVL leupeptin, 100 pmol/L sodium orthovanadate, and 0.1 %
Nonidet-P40 (NP-40 lysis buffer). Complexes were precipitated with
the same antibodies used for immunoblotting, except in the case of
cyclin E, for which the mouse monoclonal antibody HE 67 (Pharmingen) was used. Optimal amounts of primary antibodies were
used, as determined by prior immunoprecipitation experiments using
radiolabeled (35S-methionine)protein extracts. Complexes were recovered by precipitation with protein G chemically bound to Sepharose beads (Zymed).
Enzymatic assay for cdk activity. The complexes, prepared by
immunoprecipitation, were washed first with the NP-40 lysis buffer
and then with a solution containing S0 mmolL Tris-HC1 pH 7.4,
10 mmoUL magnesium chloride, and 1 mmol/L dithiothreitol. They
were incubated in a kinase reaction buffer (25 p L per sample) containing S0 mmoVL Tris-HC1 pH 7.4, 10 mmolL magnesium chloride, 1 mmolL dithiothreitol, 80 pmoVL adenosine triphosphate
(ATP), 10 pg of histone H1 (Boehringer Mannheim, Indianapolis,
All reIN), and 40 pCi of [y-32P]-ATPat 30°C for 10minutes:'
actants were in excess, and the reactions were stopped by addition
of an SDS-containing electrophoresis sample buffer while in a linear
range of reaction. After heating at 90°C for S minutes andthen
cooling, the protein G-immunobeads were pelleted by centrifugation,
and proteins in the reaction mixtures were resolved by SDS-PAGE.
The histone H1 band was visualized by a brief staining of the gels
with Coomasie Blue (R-250) in fixative (10% acetic acid/40% methanoUSO% water). The gels were then dried, and the labeled histone
bands were detected by autoradiography. The Coomassie-stained
histone bands were excised from the dried gels and quantitated by
scintillation counting. Unless noted in the text, values were corrected
for nonspecific activity, which was determined from duplicate samples in which the primary antibody step was omitted. As comparisons
are made between levels of activities obtained using complexes immunoprecipitated using different antibodies (eg, cyclin A- v cyclin
E-associated activities), data are expressed as the percentage of maximal activity obtained with a particular antibody. Because many
factors associated with the reactions, for example, the differential
effects of the antibodies on reactions performed by particular cdk/
cyclin complexes, are unknown, direct comparisons of the absolute
values (counts per minute incorporated into the H1 histone substrate)
are not made.
RESULTS
Effects of DFO on cell cycle progressionin normal human
T lymphocytes. As shown previously," dual flow cytometric analysis of DNA and RNA contents have permitted a
detailed delineation of the kinetics of cell cycle progression
in normal human T lymphocytes. In brief, it was shown that
T cells activated with the potent combination of PDB and
ionomycin (PDB/I) rapidly enter the G1 phase of the cell
cycle, as indicated by increasing content of RNA. S-phase
entry, indicated by increased DNA content, begins at about
30 hours after activation. By performing flow cytometric
analysis of cells that have been activated with PDB/I and
then accumulated about the G 2 M border with demecolcine,
it has been estimated that 80% to 90% of activated T cells
enter and progress throughthecell cycle4j (Takase et al,
unpublished results, April 1994). DFO has been shown to
arrest T-cell proliferation before S-phase entry. Determinations of RNA content indicated that this point occurred in
mid- to late G1 phase, as cells possessed an overallRNA
content significantly lower than T cells poisedto enter S
phase."
T-cell progression andthe effects of DFO on thecell
cycle were further delineated using a new method for the
simultaneous determination of DNA and~ 3 4 ' ~ cellular
"'
content~.~'
As shown in Fig 1, resting T cells represent a homogeneous population of cells containing a G1 content (2N) of
DNA and low ~ 3 4 " ~content.
"
As shown previously. resting
T cells isolated from peripheral blood, in fact, contain no
P3PdC2
, as measured by sensitive immunoblotting or immunoprecipitation method^,'^.'^ and so the levels of fluorescence
indicated in Fig 1A permit definition of background levels
for the technique. As shown in Fig IB, increased DNA and
P3QdC2contents are clearly detectable at 30 hours and are
dramatically enhanced by 40 hours (Fig ID). Little change
is seen in either DNA or ~ 3 4 " ~content
"
in DFO-treated cells
(Fig 1C and E), in agreement with biochemical analysis
presented elsewhere"," and below. Because ~ 3 4 ' ~content
'~
does not return to the low levels characteristic of T cells in
their first G1 phase, cells that have progressed through the
cell cycle and re-entered G1 phase appear, as in Fig ID, as
cells with a GI (2N) content of DNA but high~ 3 4 ' content.
~'~
Therefore, these results provide confirmationthat a large
portion of PDBD-activated T cells successfully and synchronously progress through the cell cycle during thefirst 40
hours after activation and that DFO effectively blocks cells
before both ~ 3 4 " ~accumulation
"
and DNA replication.
Effects of DFO on expression of T-cell cycle regulatory
genes. In previous reports,'o"3 we showed that DFO inhibited cell cycle progression at a point in late G1 phase, before
the GUS-phase transition, which was earlier than and distinguishable from the arrest point induced by either aphidicolin
or hydroxyurea. T lymphocytes arrested by DFO contained
no detectable p34'd'2 protein, a protein whose synthesis begins just before and independently of S-phase e n t ~ y . ~In
'.~~
the first cell cycle of activated human T lymphocytes, the
cdk2 and cdc2 genes are first expressed beginning at about
8 and 24 hours after activation, r e s p e c t i ~ e l y . AS
~ ~shown
~~~~~~"
inFig2,
accumulation of transcripts of these two genes
shows a differential sensitivity to the chelating agent. The
steady-state levels of cdk2 mRNAs accumulated are the same
in the presence or absence of the drug, even at a concentration of 100 ymol/L. Accumulation of cdc2 mRNA, on the
other hand, is almost completely suppressed by DFO, even
at the low dose of 10 PmoVL. As shown in Fig 3, expression
of the cdk4 and ZL-2 genes, which normally is first detected
within 3 to 4 hours after a c t i v a t i ~ n , ~ was
" ~ ~also
~ " ~not
~~~~~
affected by DFO.
Effects of DFO on accumulation of cell cycle regulatory
proteins. Stimulation of Tcells with PDB/Iinducesa
large portion of the cells to synchronously enter the cell
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CELL CYCLE ARRESTBY IRON DEPLETION
2271
A
1
Resting T Cells
2N
DNA content
C PA - 30 hrs + DFO
I
7
B
P/I - 30 hrs
E P/I - 40 hrs + DFO
1
7
D
P/I - 40 hrs
Fig 1. Simulteneousflow cytometric detection of
DNA and ~ 3 4 content
' ~
in human T lymphocytes
activated in the presence or absence ofDFO.Normal
human T lymphocyteswere activated with PDB/l (P/
I) in the absence (A,B, and D)or presence (C and E)
of DFO and harvested at 0 (A),30 (B and C), and 40
(Dand E) hours after treatment. Cells were fixed and
stained for flow cytometric analysis as described in
Materials and Methods. In these two-dimensionalcytograms, cellular ~ 3 4 content
' ~
(horizontal axis) is
presented on a logarithmic scale, whereas DNA content (vertical axis) is on a linear scale.
cycle, with DNA synthesis first beginning at about 30
whose patterns of synthesis also appear concordant. Both
hours
after cell activation The
is
appearance of
are present in resting T cells at very low levels. The basal
four cell cycle regulatory proteins
cyclin
A,
level of p33""' appears to be inactive as a functional kinase.28a.45
These proteins begin to increase in amount in
p33""*, and cyclin E) and theeffects of DFO on their
accumulation are shown in the immunoblottingresults
early to mid-GI phase (at IO to 15 hours after stimulation),
presented in Fig 4. The p34'dc2protein first appears in
about the time atwhich enzymatically active~ 3 3 ' ~kinase
~'
activated T cells at 30 to 40 hours after activation, almost
is first detected. As shown in Fig 4, increased amounts of
concommitant with but independent of the onset of DNA
the two proteins appearin the presence of DFO. However,
replication.".*' Its appearance is completely inhibited by
only levels of the proteins characteristic of those seen at
I O pmol/L DFO, aslateas 40 hoursafterstimulation.
about 30 hours after normal activation (ie, in the absence
The normal pattern of accumulation of cyclin A, which
of DFO) are achieved in the presence of the drug. Thus,
forms complexes with both the ~ 3 4 and
~ ~~ ~3 ' 3 pro' ~ ~ although
~
induction of synthesis of the proteins occurs and
e i n . ~ 'appears to proceed normally for the first 30 hours after
teinS,28.3Y.46-4XIS similar to that of the ~ 3 4 ' ~ ' ' ~ p r ~ t Its
'
appearance is alsocompletely prevented by DFO. The
~ 3 3 " ' ~and
' cyclin E proteins comprise a pair of proteins
activation, continued production of the proteins, characteristic of S-phase progression, is partially inhibited. It is
k2
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
2272
LUCASETAL
cdc2
F
-285
i
'- 185
0 PA DFO DFO O'PA DFO DFO
10010uM
lOOuM
uM
10uM
Fig 2. Effect of DFO on accumulation of mRNAs encoding p34and p33in activated human T cells. T cells were stimulated with
PDBII and cultured for 48 hours in the absence or presence of DFO
(10 pmollL or 100 pmolIL1. The mRNA levels for the cdk2 and c d d
transcripts were determined by Northern blotting. RNA levels were
examined in resting cells (01, PDBII-treated cells (PII1, and cells
treated with either 10 pmolIL or 100 pmol/L DFO, as indicated. Equal
amounts of RNA were added to the lanes as assessed by ethidium
bromide-stainingof the gels.
through the so-called "restriction point," entry into S phase,
and/or progression through S p h a ~ e . ' ~ . . ' ~ . ~ ' . ~A
" -possible
~'.~~.~'
role for p3Ydk2at the G2/M phase transition has also been
disc~ssed.~'
As DFO arrested cell cycle progression in late
G1 phase, apparently without suppressing the initial production of the ~33'"'' protein, the effects of the drug on this
protein were examined in more detail.
First, as shown in Fig SA, the levels of p3Ydk2protein
accumulated in the presence or absence of DFO were compared. In this experiment, the ~33'~''protein was prepared
by immunoprecipitation. After resolutionof proteins by electrophoresis, the protein was detected by immunoblotting. As
shown by comparison of lanes 1 and 2 in Fig 5A, the amounts
of p3Ydk2present at 30 hours after activation (characteristic
of late GI phase) were very similar in cells incubated in the
absence (lane 1) or presence (lane 2) ofDFO.However,
continued accumulation was suppressed, as shown by comparison of lanes 3 and 4. By 40 hours after activation, the
amount of ~33'~''present in DFO-treated cells (lane 4) was
less than that detected in untreated cells (lane 3). As shown
in lane 5, the inhibitory effect of DFO on accumulation of
the ~33'~''protein by 40 hours was effectively overcome by
the addition of excess iron.
Asnoted above, levels of cdk2 mRNAappeared to be
normal in cells activated in the presence of DFO, even as
late as 48 hours of incubation. To determine whether the
decreased accumulation of the protein was due to decreased
synthesis or enhanced degradation of the protein, the follow-
also demonstrated in Fig 4 that normal levels of the ~ 3 3 ' ~ ' ~
and cyclin E proteins accumulate in T cells that are activated in the presence of DFO and an excess of iron.
Also, it was determined that phosphorylation of the pl loRb
protein (shown by the appearance of the slower mobility
forms of the protein) was initiated in the presence of DFO
-1 8s
(Fig 4). However, levels of phosphorylation of pl IORh
achieved in the presence of DFO were characteristic of those
normally seen after about 30 hours after activation; attainment of the normal pattern of hyperphosphorylation, characteristic of S-phase entry and progression,".'~.5".4'was partially suppressed. Also shown in Fig 4, addition of excess
iron to the cultures almost completely restored the normal
pattern of accumulation and phosphorylation of p1 IORh, indicating that the effects of DFO were, indeed, due to chelation
of the metal.
Effects of DFO on metabolism of the p3.Yd2protein. The
results described above indicate that the induction of cdk2
1OuM
1 OuM
gene expression and p3Ydk2protein synthesis occurred normally in the presence of DFO and place the point of arrest
Fig 3. Effect of DFO on accumulation of mRNAs encoding
in late GI phase, before the first appearance of the ~34'~'' and 11-2 in activated human T cells. T cells were stimulated with
PDBII and cultured for 6 hours in the absence or presence of DFO
and cyclin A proteins, which normally begin just before the
110 pmolIL1. The mRNA levels for the cdk4 and/L-2 transcripts were
Gl/S-phase transition. The exact role(s) of the p3Y"' kinase
determined by Northern blotting. RNA levelswere examined in restin cell cycle progression is still unclear. It has been proposed
ing cells (01, PDBII-treated cells lP/l1, and cellstreated with 10 *mol/
that thep3Ydk2kinase, which has been found to be associated
L DFO, as indicated.Equal amounts of RNA were added to thelanes
withboth cyclin A and cyclin E, may regulate passage
as assessed by ethidium bromide-stainingof the gels.
0 PA DFO
0 PA DFO
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CELL CYCLE ARREST BYIRONDEPLETION
+PDB/I
+PDB/I
2273
+DFO
0 20 30 40 I20 30 40I40 hours
1
4
.
m-
.
2
1
2
3
4
5
6 7
'
p1 1ORk
8 Lane Number
Fig4.
Effect of DFO on accumulation of the p34-,
cyclin A,
~ 3 3 ' cyclin
~ ,
E, and p110" proteins in activated human T cells. T
cells were stimulated with PDB/I and incubated in the absence (lanes
l through 4) orpresence (lanes 5 through 8) of 10 pmol/L DFO.
Samples were harvestedat initiation of the experiment ( 0 hours, lane
1) and at 20 (lanes 2 and 5). 30 (lanes 3 and 61, and 40 (lanes 4 and
7) hours of incubation. Sampleswere also preparedat 40 hours from
a culture activated with PDB/I in the presence of both DFO and an
excess of iron,provided by 500 pmol/Lferric citrate (lane 8). Samples
were analyzed by immunoblotting for the ~ 3 4 cyclin
~ . A, ~33"''~.
cyclin E, and pllORbproteins. Immunoreactive protein bands of appropriatesize, ascomparedwith marker proteinsof known molecular
weight included in the same gels, are indicated. Protein extracts prepared from equal numbers of cells were loaded in the lanes.
ing experiments were performed. First, activated T cells incubated in the presence or absence of DFO were metabolically labeled with "S-methionine at 18 to 25 and 26 to 33
hours after stimulation. The ~ 3 3 ' ~protein
~ ' was isolated by
immunoprecipitation and then detected by autoradiography
after gel electrophoresis. As shown in Fig 5B, the extent of
incorporation into the protein was nearly identical at the
earlier time point in the presence (lane 2) or absence (lane
3) of DFO. By the later time point, the extent of incorporation was clearly reduced in the DFO-treated cells (lane 5)
as compared with control cells (lane 6). suggesting that the
rate of synthesis of the protein was decreased by this time
point.
Finally, to ascertain whether or not the decreased accumulation of the protein could be due at least partly to enhanced
protein degradation induced by the drug, cells were labeled
with "S-methionine from 42 to 45 hours after activation in
the absence or presence of DFO. A portion of each culture
was harvested at 45 hours. The remaining cells were washed
free of the radioisotope, and portions were incubated for an
additional 2 or 4 hours in normal growth medium. The
p3Ydk2protein was isolated from the six samples and visualized by autoradiography after electrophoresis. As shown in
Fig 5C, the amount of label incorporated during the labeling
period was, as expected, somewhat reduced in the presence
of DFO (lane 4) as compared with the control cells activated
in the absence of the drug (lane 1). However, the kinetics
of decay appeared to be essentially the same in the presence
or absence of DFO. A slight loss of radioactivity associated
with ~ 3 3 was
' ~ detected
~ ~ by 2 hours (lanes 2 and 5 ) after
the labeling period and became more pronounced by 4 hours
(lanes 3 and 6). It appeared that the protein was a relatively
long-lived one in the absence or presence of DFO. No rapid
or dramatic degradation of p3Ydk2occurred in the presence
of the drug, supporting the notion thatthe increased synthesis
normally occumng as cells entered and progressed through
S phase was, indeed, curtailed.
The effects of DFO on HI histone kinase activities in T
cells. The levels of H1 histone kinase activity due to ~34'~''
and ~ 3 3 after
' ~ T-cell
~ ~ activation in the presence or absence
of DFO were compared. In the experiment shown in Fig 6,
kinase activities associated with specific immunoprecipitates
of the two kinases were measured at 55 hours after stimulation. At this late time point, normal levels of activity due to
both kinases were very high. As expected considering its
inhibitory effect on the appearance of the protein, DFO almost completely prevented the appearance of ~34'~''kinase
activity. Activity due to ~ 3 3 ' ~was
~ ' also substantially reduced, to about 8% of its control level, as determined using
the extract prepared from cells activated in the absence of
the drug. However, that this represents a substantial level of
activity above background control levels or the low residual
level of ~34'~''activity is apparent from a comparison of the
two autoradiograms presented as insets in Fig 6.
As DFO appeared to inhibit the late accumulation of
p3Ydk2,but not its initial synthesis, ~ 3 3 activity
' ~ ~ was
~ next
measured at earlier times after activation in the presence or
absence of the drug. In these experiments, active cyclid
~ 3 3 complexes
' ~ ~ ~ were isolated by immunoprecipitation
from cells activated in the presence or absence of DFO.
Histone H1 kinase activities were determined using the in
vitro kinase reaction method described above. The normal
kinetics of activation of ~ 3 3 ' ~after
~ ' T-cell stimulation occurs in two distinguishable phases (Fig 7, solid bars).2s"
Activity is first detectable at about 15 to 20 hours after
stimulation. It increases slowly for up to 30 hours, until the
Gl/S-phase transition, reaching about 10% of subsequent
maximal levels. After entry into S phase, the second phase
of activation, marked by a rapid 10-fold increase in activity,
begins. In the presence of DFO (Fig 7, open bars), the initial
phase of activation of the kinase was observed; the levels
of ~ 3 3 ~ activity
~ ' ' ~ present at 20 hours were nearly the same
in cells activated in the presence or absence of DFO. By 30
hours, however, an inhibitory effect of the drug became
evident, and the dramatic 10-fold increase associated with
S-phase entry and progression was nearly completely inhib-
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
2274
LUCAS ET AL
ited. As indicated in Fig 7, DFO treatment resulted in an
almost 90% reduction in the level of activity seen at the 40hour time point.
It has been shown in several cell systems that the ~33'~''
kinase forms active complexes primarily with two cyclins,
cyclin E and cyclin A,?X.?'~.~~.4h~X.~?.53..55
and that the cyclin Eassociated activity precedes cyclin A-associated activity in
the GI phase of the cell ~ y ~ l e . ~ " The
~ " ~effects
~ ' ~ 'of
~ DFO
on cyclin E- and A-associated HI histone kinase activities
were, therefore, examined, as shown in Fig 8. Cyclin Eassociated activity was first detected in mid-GI phase (about
20 hours): a larger increase, although not as great as that
seen for overall ~ 3 3 " ' ~activity,
'
was seen after S-phase entry.
As for initial overall p33"Ik2 activity, the initial appearance
of cyclin E-associated activity was refractory to DFO. Even
at 30 hours after activation, inhibition of cyclin E-associated
activity was modest: the level of cyclin E-associated activity
A
M
1
2
3
4
5
B
C
1
detected in S phase was also onlypartially(about
40%)
inhibited by DFO. In contrast, the appearance of cyclin Aassociated activity (Fig 8) was completely inhibited by DFO,
consistent with observations that the drug prevents the appearance of cyclin A protein. Itis also shown that a low
level of cyclin A-associated activity appears atabout 30
hours after stimulation (in the absence of DFO) and increases
more dramatically as cells enter S phase. Thus, p33'"'' activity seen before about 30 hours appears to be due primarily
to cyclin E / ~ 3 3 ' ~ ~and
' , its synthesis and activity are largely
resistant to the effects of DFO. Beginning at about 30 hours.
~ 3 3 " ' ~activity
'
is due in some part to cyclin A/p33"'k2. As
cyclin A synthesis is completely inhibited, this component
of ~33'~''activity does not appear and cell cycle progression
is halted. However, the possibility that the modest decrease
seen in GI phase cyclin E-associated activity also plays
some role in the DFO-induced arrest is not precluded by the
present data.
Eflects of DFO on occumulntion of cyclin E and cvclin A
mRNAs. To further analyze the level at which DFO exerted
its inhibitory effect on cyclin A protein accumulation, the
levels of cyclin A and cyclin E mRNAs were compared. As
shown in Fig 9, cyclin E mRNA was easily detected at 24
and 48 hours after T-cell activation. As expected based on
the observation that cyclin E protein accumulation proceeded
fairly normally during this period (Fig 4), DFOhad little
effect on cyclin E mRNA levels. Cyclin A mRNA levels,
which could also bereadily detected atboth 24 and 48
hours after activation, appeared only slightly depressed in
p33cdk2degradation
Fig 5. Effect of DFO on the accumulation, rate of synthesis, and
rate of degradationof the p33- protein in activated humanT cells.
(A) Accumulation of the
~ 3 3 ' ~protein
"
in T cells in the presence and
absence of DFO were compared. T cells were activated with PDB/I
in the absence (A, lanes 1 and 3) or presence of DFO (A, lanes 2 and
4) or of DFO and excess iron (A, lane 51 and harvested at 30 (A, lanes
1 and 2) or 40(A, lanes 3 through 5) hours after incubation. The
~ 3 3protein
' ~ ~ was isolated by immunoprecipitation. After resolution
of proteins by gelelectrophoresis, the ~ 3 3 protein
~ ~ " wasvisualized
by immunoblotting.(B) The rates of synthesis of the ~ 3 3 protein
~ "
in the presence and absence of DFO were compared. T cells were
activated with PDB/l in the absence (B, lanes 1, 3,4, and 6) or presence (B, lanes 2 and 5) of DFO. Cultures were metabolically labeled
with 35S-methioninefrom 18 t o 25 (B, lanes 1 through 31 or 26 t o 33
(B, lanes 4 through 61 hours of incubation. The ~ 3 3 protein
~ ~ " was
prepared by immunoprecipitation IB, lanes 2,3,5, and 6) with specific
rabbit polyclonal antibody and detected by fluorography
after resolution of protein by gel electrophoresis. Control radiolabeled samples
were incubatedwith rabbit antiserum from
a nonimmunized animal
instead of the specific a n t i - ~ 3 3 ~antiserum
~"
and were similarly processed (B, lanes 1 and 4) t o facilitate identification of the ~ 3 3 ' ~ "
protein band, which is indicatedin the figure. (C) Thedegradation of
in the presence and absence of DFO were comthe ~ 3 3 ' ~protein
~'
pared. T cells were activated with PDB/I in the absence (C, lanes 1
through 3) or presence (C, lanes 4 through 61 of DFO and metabolically labeled with "S-methionine at 42 t o 45 hours of incubation.
Samples were harvested immediately after labeling (C, lanes 1 and
41 and at 2 (C, lanes 2 and 5) and 4 (C, lanes 3 and 6) hours after
removal of the"S-methionine from the cultures by
extensive washing. The ~ 3 3 ' ~protein,
"
indicated in the figure, was isolated by immunoprecipitation and visualized by fluorography after resolution of
proteins by gelelectrophoresis, as described above. Protein extracts
prepared from equal numbers of cells were loaded in the lanes.
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CELL CYCLE ARREST BY IRON DEPLETION
2275
1
100
100
l
PA
z
50
3
DFO
PA
DFO
Fig 6. Effect of DFOon ~ 3 4 and
~ p33"associated
H1 histone kinase activities in human T cells. T cells were activated with PDB/I either
and p33proteins were isolated by immunoprecipitain the presence or absence of DFO
and harvested at 55 hours of incubation.The p34tion and assessed for H1 histone kinase activity, as described in Materials and Methods. The activity associated with immunoprecipitates
prepared from DFO-treated cells is shown graphically, as compared with those prepared from the control activated cells. Background levels
were determined by determining kinase activity associated with duplicate samples prepared without primary specific antibody. The insets
show the radiolabeled H1 histone band, after gel electrophoresis and autoradiography, phosphorylated bykinase prepared from control
activated cells (lanes 1) or from cells stimulated in the presence of DFO (lanes 2). Lanes 3 show the background levels of radioactivity caused
by performing the reaction with samples prepared from cells activated in the absence of DFO, but processed without addition of the specific
anti-p34- or anti-p330d" antibodies. The values given by these samples were subtracted from the experimental data.
the presence of DFO (Fig 9). despite the fact that no cyclin
A protein is present in DFO-arrested T cells (Fig 4). At 48
hours after stimulation, there appeared to be slightly more
cyclin E and cyclin A mRNA in the absence of DFO (lane
4). However, as demonstrated in the bottom panel of Fig 9,
which shows the ribosomal RNAs present in the gel, it appears that somewhat more overall RNA was applied to the
lane showing RNA from cells activated in the absence of
DFO (lane 4).
bers of the D family,"' have hampered convincing demonstration of the actual functions of the ~ 3 3 ' ~kinase.
~'
Furthermore, the discovery of numerous other members of
the cyclin-dependent kinase and cyclin families in animal
cells23-2"
suggests a possible level of complexity beyond simple models that describe the cell cycle in terms of a single
linear chain of events in which major transitions are performed by single kinase/cyclin complexes. Even the model
in which the ~ 3 4 ' ~ 'kinase
'
regulates two major transitions
in yeast is complicated by the recent identification of a multiDISCUSSION
tude of different cyclins in yeast, each of which shows its
own particular cell cycle-dependent pattern of synthesis and
Recent analysis of the cell cycle of mammalian cells has
association with the ~34'~''(or p34cK2x)protein."'"'
shown that major cell cycle transitions are regulated at least
In the present study, the effects of DFO on proliferation of
in part by a family of serine-protein kinases, called cyclinnormal humanT cells were definedin terms of the molecules
dependent kinases, and the cyclin molecules with which they
interact.".'?.2S.?h One member of this family, ~ 3 4 ' ~(or
' ~ involved in GI-phase progression and S-phase entry. Inhibition of cell cycle progression with DFO arrested the cells at
~ 3 4 ' ~ ~complexed
'),
with cyclins of the B family, plays a
a point in G1 phase before the appearance of cyclin A.
role in regulating entry into mitosis.'R"' Much evidence suggests that a second member of the kinase family, ~33'~'*, However, the G1 phase-associated synthesis of cyclin E and
p33""' and their assembly into a complex with enzymatic
complexed with cyclins of the A and/or E families, regulates
activity proceeded almost normally. These results suggest
earlier, important cell cycle regulatory events: committment
that the p3Ydk'/cyclin A complex is needed for completion
to cell division, occurring in late GI phase (at a putative
of the GI phase of the cell cycle. However, the possibilities
start or restriction point), the GUS phase transition, and,
perhaps, even passage through S phase."~'9~"~4"4x~''~s7 How- that the early cyclin E-associated activity produced in the
presence of DFO, although little affected, is still insufficient
ever, the fact that no in vivo substrates for these cdkkyclin
for completion of G I -phase or that still other, as yet unidenticomplexes have been definitively identified (although several have been proposed), combined with the apparent profied, cell cycle regulatory events are also inhibited have not
miscuity of the enzyme in combining with cyclins of many
been excluded. For example, whether or notDFO has an
families including not only A and E cyclins, but also memeffect on the synthesis and/or activities of low-molecular-
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
2276
LUCAS ET AL
~
i
3 Activity
3
~
~
10
rc
0
x
2b
S
0'
3b
4b
T i m after PA (hours)
Fig 7. Effect of DFO on the time-course of appearance of p33"du
kinase activity in humanT cells. T cells were activatedwith PDB/l in
the absence ( W or presence(U)of DFO,and samples were hervested
at 0,20,30, and 40 hours of incubation. The p330dulcyclin complexes
were isolated by immunoprecipitation and assessed for H1 histone
kinase activity, as describedin Materials and Methods and as shown
in Fig 6. The w
tts are expressed as percent of maximal kinase
activity, with 100% being the actlv** present in control cells (those
activated in the absence of DFO) at 40 hours after treatment with
PDB/I. Activity was determined as the radioactivity (cpm) incorporated into H1 histone duringa 10-minute reaction using an immunoprecipitate from 2 x 10' cells. The scale on the y-axis is broken becauseof the approximate10-foldincrease in activity that occurs
between 30 and 40 hours of incubation with PDBII. The numbers
within the open boxer showthe percent inhibitionof p33- activFty
in the DFO-treated samples, as compared with the untreated activated cells.
weight proteins such as p21 and p27, which can inhibit the
activity of cyclin-dependent kinase^,^'.^' remains to be determined. As the synthesis of DNA is blocked, because of
inhibition of ribonucleotide reductase, many events associated with normal progression through S phase are also
blocked. These events include the complete S phase-associated increases in (1) cyclin E and p33- protein levels, (2)
cyclin E-associated kinase activity, and (3) complete hyperphosphorylation of the pllORbprotein. As shown in Figs 4,
7, and 8, all of these events are dampened. It is suggested,
however, that these events are subsequent or secondary to
the late GI-phase block described here and in previous rep o r t ~ . " - As
~ ~ events associated with S-phase progression,
including the dramatic increase in p33'"%yclin E activity,
were depressed by DFO, a role for this latter complex in S
phase progression is not ruled out by the data. In fact, the
dramatic increase seen as cells enter S phase make such a
role likely and supports a model in which multiple cdklcyclin
complexes function at particular phases of the cell cycle.
The exact mechanism by which iron depletion results in
the lack of appearance of cyclin A protein remains to be
determined. As we showed previously, cyclin A synthesis is
initiated before S-phase entry innormalhuman
T cells.4"
Initiation of its synthesis is independent of S-phase entry,
and it is accumulated to aberrantly high levels in cells that
are blocked at the GUS interface by treatment with aphidicolin. We have, therefore, begun to examine events leading
to production of cyclin A, and the effects of DFO on these
events. Surprisingly, itwas found that levels of cyclin A
mRNA were little affected by the presence of DFO. Because
the rate of synthesis of the ~33'"' protein (Fig 5) and the
kinetics of accumulation of cyclin E protein (Fig 4) are
proceeding nearlynormally during DFO arrest, whenno
detectable cyclin A protein is accumulated (Fig 4), the possibility arises that DFO is preferentially inhibiting the translation of certain mRNAs. Roles for ironin the translation
of proteins have been demonstrated and discussed previo u s l ~ . Alternatively,
~"~~
DFO may be inhibiting the appearance of cyclin A mRNA in the cytoplasm. A possible role
for iron in mRNA transport has also been recently noted.74
Lastly, it is possible that the stability of any cyclin A protein
thatis synthesized is greatly reduced in the presence of
the drug. Allof these possibilities are amenable to direct
experimental analysis, as is nowin progress. Finally, it
should be noted that synthesis of the ~ 3 4 " ~
protein
'
is also
prevented by treatment with DFO. However, since this protein is not activated as a kinase until near the G2M-phase
border in the first cycle of normal T cells induced to enter
the cell cycle:'" it is unlikely that inhibition of its synthesis
is responsible for the arrest in late G1 phase observed here.
Nowthattheblock
to cyclin A production in DFO-arrested T cells has been identified, more incisive molecular
analysis of the phenomenon, both in normal T cells and in
cultured cell lines, will be possible. Such studies will include
determination of the mechanism by which cyclin A absence
affects subsequent events in T-cell cycle progression. To this
end, we have begun an examination of how the absence of
cyclin A protein in DFO-arrested cells affects the forms of
the E2F transcription factor that are present. E2F, which can
be found in complexes with cyclin A and cyclin E, appears
to be needed in a free form for activation of transcription of
genes necessary for S phase (reviewed by N e ~ i n s ~Gel
~ ) shift
.
analysis has yielded the expected result that DFO inhibits the
appearance of the cyclin A-containing form of E2F complex
(Terada et al, unpublished results, February 1994). However,
the effects of this on the relative amounts of other forms of
E2F andon transcription of E2F-regulated genes require
further study.
Since our initial reports on the inhibition of accumulation
of ~ 3 4 in"normal
~ ~ ~T cell^'^.'^ and neuroblastoma cells,"
the effects of another iron chelator, mimosine, on HI histone
kinases has been described.76In contrast with results presented here, it was noted that the addition of mimosine to
human fibroblasts entering the cell cycle resulted in essentially complete inhibition of total cdk2 activity. However, it
should be noted that, whereas our studies used the minimal
dose (10 pnoVL) of DFO needed for effective inhibition of
S-phase entry, the latter study used a concentration of mimosine of 300 pmolL. Determination of whether or not the
observed differences seen are due to different concentrations
of agents, to different cell types, or to actual differences in
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
2211
BY IRON DEPLETION
CELLCYCLEARREST
"t-c
Cyclin A-osoodated Kinase
PA
0 P/I + WO
Cydin E-associated Kinase
PA
0 PA + DFO
(II
c
3
30
a0
Time after PA (hours)
-t 0
3b
Time after P/I (hours)
40
Fig 8. Effect ofDFO on thetime-course ofappearance of the cyclinE- and cyclinA-associated H1 histone kinase activities in human T cells.
T cells were activatedwith PDB/I in the absence )1. or presence IO) of DFO and harvested at the timesindicated. Cyclinlcdk complexes were
isolated by immunoprecipitation with antibody specific for either cyclin E or cyclin A and then assessed for H1 histone kinase activity, as
described in Materials andMethods. Activity is shown
as percent of maximal activity, which is taken
as the value associated with thesamples
isolated at 40 hours after stimulation, in the absence of DFO. Because different specific antibodies are used for isolation of complexes using
antibodies t o cyclin A, cyclin E, or the kinases (as in Figs 6 and 7). direct comparisons of absolute values are not made. Note that some
different time points wereused for each analysis, because cyclin E-associated activity can first be detectedat about 15 t o 20 hours, whereas
cyclin A and its associated activity are not seen until later (about 30 hours). The numbers within the open boxes in the left panel show the
percent inhibition of the cyclin
E-associated activity in the DFO-treated samples, as compared with theuntreated activatedcells; because the
percent inhibition of cyclin A-associated activity was nearly complete at all time pointsmeasured, numbers are not shown in the rightpanel.
the mechanisms of action of the drugs will likely require
direct comparisons of the two systems. However, despite the
outcome of such studies, it is important, in view of the
current and potential clinical usefulness of DFO, to determine the mechanism(s) of action of the drug not only in
normal human T cells, but also in other normal and transformed cell types. Finally, it is noteworthy that the effects
described in the present report were all observed using a
dose of 10 pmol/L, a concentration that can be achieved
in humans in clinical application^,'^"^ as noted above. As
reported by Ajchenbaum et al," treatment of T cells with
100 pmol/L DFO resulted in dramatic inhibitory effects on
expression of several cell cycle regulatory molecules, including cdk2 and, to a lesser extent, cyclin E. However, it is
likely that this excess of chelating agent has other more
nonspecific and toxic effects on sensitive cells. Consistent
with the findings of Ajchenbaum et al,77we had previously
demonstrated"." that 100 pmol/L DFO, butnot 10 pmoV
L DFO, will cause an almost complete inhibition of the
accumulation and phosphorylation of the p1 IORb tumor suppressor protein.
Several recent studies suggest that cell cycle models developed solely from analysis of continuously growing, transformed
cell lines may not adequately reflect mechanisms involved in
normal cell growth. This possibility was identified in several
earlier s t u d i e ~ , " ~that
~ ~ ~showed
'~
that the relative levelsof certain molecules, such as some cdks and cyclins, varied greatly
among various cell lines and did not reflect
m e levels in n o d
cells of origin. More recently, it was noted that certain complexes, comprised of cdks, cyclins, proliferating cell nuclear
antigen (a component
of DNA polymerase S;PCNA), and other
s m a l l molecules, were altered in composition or completely disrupted in normal fibroblastic cells after virus-induced transformatiOn.61.79 Fwhermore, our recent comparison of a growth-controlledB-lymphoblastoidcelllinewith
n o d lymphocytes
raised the possibility that the mechanisms regulating the orderly
activation of appropriate cdks was fundamentally altered in the
cell line.35 In brief, it appeared that both the
cdc2- and c&encoded kinases were activated to a high level before S-phase
entryinthefirstcycle
ofcellsexiting
a resting (CO) state.
Synthesis of the p3Pk4protein also appeared to be aberrantly
regulated in the cell lir1e.3~ Collectively, these observations, combined with a growing body of evidence showing abnormal cyclin
production in a variety of tumors and tumor-derived cells,26,31-33
suggest the possibility that agents suchas DFO that target pathways involved in the synthesis and/or regulation of these molecules might showuseful, differential effects on normal and transformed cells, a possibility now being explored in further studies.'"
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
LUCAS ET AL
2278
Lane
1
2
3
4
5
-28s
Reed for providing reagents (plasmids and antibodies) and suggestions concerning their use during the course of this work. We also
thank Drs Paul Seligman, Jaime Modiano, and Richard Franklin for
helpful discussions andthe personnel of the Children's Hospital
Blood Donor Center (Denver, CO) for providing us withhuman
blood.
REFERENCES
-1 8s
-28s
-
*
:yclfn A
TIME:
DFO:
Lane
-18s
0
24
- -
24
+
48
-
48
+
1 2 3 4 5
Fig 9. Effect of DFO on accumulation of mRNAs encoding cyclin
E and cyclin A in activated human T cells. T cells were stimulated
with PDB/I in the absence (lanesl,2, and 4) or presence (lanes 3 and
51 of DFO (10 pmollL) and harvested at 0 (lanes 11, 24 (lanes 2 and
3). or 48 (lanes 4 and 5) hours after activation. RNA was extracted,
and mRNA levels for the cyclin E and cyclin A transcripts were determined by Northern blotting, as described in Materials and Methods.
Analysis for cyclin E shows a single major transcript, whereas cyclin
A mRNAin human cells appears as a doublet, as indicated by the
arrows. The inset at t h e bottom shows the 18s and 28s ribosomal
bands after ethidium bromide-staining of the gel. After probing for
cyclin A transcripts, t h e blot was stripped of cyclin A probe and
retained for several weeks to permit decay of any residual radioactive
material. I t was then probed for cyclin E transcripts.
As noted above, drugs that can exert profound antiproliferative effects, such as DFO, are used in many clinical situations.
Often, these agents inhibit cell growth at particular points in
thecellcycle,andattemptshave
beenmade to definethe
molecular basis for such inhibitory effects. Emerging concepts
andpreciseknowledgeoftheidentitiesandmechanismsof
action of the molecules regulating major cell cycle transitions
should permit identification of the targets of such drugs and,
as a result, their more efficacious use in clinical applications.
ACKNOWLEDGMENT
We are grateful to Drs Giulio Draetta, Steven Elledge, Steven
Hanks, Tony Hunter, Matthew Meyerson, David Morgan, and Steven
I . Modell B: Advances in the use of iron-chelating agents for the
treatment of iron overload. Prog Hematol 11:267, 1979
2. Robotham JL, LietmanPS: Acute iron poisoning: A review.
Am J Dis Child 1342375, 1980
3. Modell B, LetskyEA,FlynnDM,
Peto R, Weatherall DJ:
Survival and desferrioxamine in thalassaemia major. BrMed J
284: IO81, I982
4. Cazzola M, Bergamaschi G, Dezza L, Arosio P Manipulations
of cellular iron metabolism for modulating normal and malignant
cell proliferation: Achievements and progress. Blood75:1903,
1990
5. Bierer BE, Nathan DG: The effect of desferrithiocin, an oral
iron chelator, on T-cell function. Blood 76:2052, 1990
6. Kemp JD: The role of iron and ironbinding proteins in lymphocyte physiology and pathology. J Clin lmmunol 13:81, 1993
7. Lederman HM, Cohen A, Lee JWW, Freedman MH, Gelfand
EW: Deferoxamine: A reversible S-phase inhibitor of human lymphocyte proliferation. Blood 64:748, 1984
8. Blatt J, Stitely S: Antineuroblastoma activity of desferoxamine
in human cell lines. Cancer Res 47:1749, 1987
9. Becton DL, Bryles P: Deferoxamine inhibition of human neuroblastoma viability and proliferation. Cancer Res 48:7189. 1988
IO. Brodie C, Siriwardana G, Lucas J. Schleicher R, Terada N,
Szepesi A, Gelfand E, Seligman P: Neuroblastoma sensitivity to
growth inhibition by deferrioxamine: Evidence for a block in G1
phase of the cell cycle. Cancer Res 533968, 1993
I I. Terada N, Lucas JJ, Gelfand EW: Differential regulation of
the tumor suppressor molecules, retinoblastoma susceptibility gene
product (Rb) and p53, during cell cycle progression of normal human
T cells. J lmmunol 147:698, 1991
12. Terada N. Or R, Szepesi A, Lucas JJ. Gelfand EW: Definition
of the roles for iron and essential fatty acids in cell cycle progression
of normal human T lymphocytes. Exp Cell Res 204:260, 1993
13. Lucas JJ, Terada N. Szepesi A, Gelfand EW: Regulation of
synthesis of ~34'~''and its homologues and their relationship to
pl IORh phosphorylation during cell cycle progression of normal human T cells. J Immunol 148:1804, 1992
14. Summers MR. Jacobs A, Tudway D,Perera P, Ricketts C:
Studies in desferrioxamine and ferrioxamine metabolism in normal
and iron-loaded subjects. Br J Haematol 42:547, 1979
15. Allain P, Mauras Y, Chaleil D, Simon P, Ang K. Cam G,
LeMignon L, Simon M: Pharmacokinetics and renal elimination of
desferrioxamine and ferrioxamine in healthy subjects and patients
with haemochromatosis. Br J Clin Pharmacol 24:207, 1987
16. Reichard P, Ehrenberg A: Ribonucleotide reductase-A radical enzyme. Science 221514, 1983
17. Lalande M: A reversible arrest point in the late GI phase of
the mammalian cell cycle. Exp Cell Res 186:332, 1990
18. lkegami S, Taguchi T, Ohashi M: Aphidicolin prevents mitotic cell division by interefering with the activity of DNA [email protected] 275:458, 1978
19. Huberman JA: New views of the biochemistry of eukaryotic
DNA replication revealed by aphidicolin, anunusual inhibitor of
DNA polymerase a. Cell 23547, 1981
20. Kumagai N. Benedict SH, Mills GB, Gelfand EW: Comparison of phorbol esterkalcium ionophore and phytohemagglutinininduced signalling in human T lymphocytes: Demonstration of in-
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
CELLCYCLE ARREST BY IRON DEPLETION
2279
cdk2 and cyclin Ncdk2 kinases associate with p107 and E2F in a
terleukin 2-independent transferrin receptor gene expression. J Imtemporally distinct manner. Genes Dev 6:1874, 1992
munol 140:37, 1988
41. Laemmli UK: Cleavage of structural proteins during the as21. Reed SI: The role of p34 kinases in the G1 to S-phase transisembly of the head of bacteriophage T4. Nature 227:680, 1970
tion. Annu Rev Cell Biol 8:529, 1992
42. Harlow E, Lane D: Antibodies: A Laboratory Manual. Cold
22. Shen CJ: Mammalian G1 cyclins. Cell 73:1059, 1993
Spring Harbor, NY, Cold Spring Harbor Laboratory Press, 1988
23. Meyerson M, Faha B, Su L-K, Harlow E, Tsai L-H: The
43. Howe PH, Draetta G , Leof EB: Transforming growth factor
cyclin-dependent kinase family. Cold Spring Harb Symp Quant Biol
@linhibition of ~ 3 4 ' ~phosphorylation
'
and histone H1 kinase activ56:177, 1991
ity is associated with GUS-phase growth arrest. Mol Cell Biol
24. Meyerson M, Enders GH, Wu C-L, Su L-K, Gorka C, Nelson
11:1185, 1991
C, Harlow E, Tsai L-H: A family of human cdc2-related protein
44. Takase K, Sawai M, Yamamoto K, Yata J, Takasaki Y, Tekinases. EMBO J 11:2909, 1992
raoka H, Tsukada K: Reversible GI arrest induced by dimethyl
25. Pines J: Cyclins and cyclin-dependent kinases: Take your
sulfoxide in human lymphoid cell lines: Kinetics of the arrest and
partners. TIBS 18:195, 1993
expression of the cell cycle marker proliferating cell nuclear antigen
26. Motokura T, Arnold A Cyclins and oncogenesis. Biochim
in Raji cells. Cell Growth Differ 3515, 1992
Biophys Acta 1155:63, 1993
45. Modiano SF, Domenico J, Szepesi A, Lucas JJ, Gelfand EW:
27. Furukawa Y, Piwnica-Worms H, Emst TJ, Kanakura Y, GrifDifferential requirements for interleukin-2 distinguish the expression
fin JD: cdc2 gene expression at the G1 to S transition in human T
and activity of the cyclin-dependent kinases CDK4 and CDK2 in
lymphocytes. Science 250:805, 1990
human T cells. J Biol Chem 269:32972, 1994
28. Elledge SJ, Richman R, Hall FL, Williams RT, Lodgson N.
46. Tsai L-H, Harlow E, Meyerson M: Isolation of the human
Harper JW: CDK2 encodes a 33-kDa cyclin A-associated protein
cdk2 gene that encodes the cyclin A- and adenovirus E l A-associated
kinase and is expressed before CDC2 in the cell cycle. Roc Natl
p33 kinase. Nature 353:174, 1991
Acad Sci USA 89:2907, 1992
47. Rosenblatt J, Gu Y, Morgan DO: Human cyclin-dependent
28a. Lucas JJ, Szepesi A, Domenico J, Tordai A, Terada N, Gelkinase 2 is activated during the S and G2 phases of the cell cycle
fand EW: Differential regulation of the synthesis and activity of the
and associates with cyclin A. Proc Natl Acad Sci USA 89:2824,
major cyclin-dependent kinases, p34""*, p3YdU, and p3edk4during
1992
cell cycle entry and progression in normal human T lymphocytes. J
48. Pagano M, Pepperkok R, Verde F, Ansorge W, Draetta G:
Cell Physiol (in press)
29. Kung AL, Sherwood SW, Schimke RT: Cell line-specific
Cyclin A is required at two points in the human cell cycle. EMBO
differences in the control of cell cycle progression in the absence
J11:961, 1993
49. Terada N, Lucas JJ, Szepesi A, Franklin RA, Domenico J,
of mitosis. Proc Natl Acad Sci USA 87:9553, 1990
Gelfand EW: Rapamycin blocks cell cycle progression of activated
30. Xiong Y, Connolly T, Futcher B, Beach D: Human D-type
cyclins. Cell 65:691, 1991
T cells prior to events characteristic of the middle to late G1 phase
of the cycle. J Cell Physiol 154:7, 1993
3 1. Buckley MF, Sweeney KJE, Hamilton JA, Sini RL, Manning
50. Furukawa Y, DeCaprio JA, Freedman A, Kanakura Y, NakaDL, Nicholson RI, deFazio A, Watts CKW, Musgrove EA, Sutherland RI,: Expression and amplification of cyclin genes in human
mura M, Ernst TI, Livingston DM, Griffin JD: Expression and state
breast cancer. Oncogene 8:2127, 1993
of phosphorylation of the retinoblastoma susceptibility gene product
32. Leach FS, Elledge SJ, Shen CJ, Willson JKV, Markowitz S,
in cycling and noncycling human hematopoietic cells. Proc Natl
Kinzler KW, Vogelstein B: Amplification of cyclin genes in coloAcad Sci USA 87:2770, 1990
rectal carcinomas. Cancer Res 53:1986, 1993
51. DeCaprio JA, Furukawa Y, Ajchenbaum F, Griffin JD, Liv33. Palmer0 I, Holder A, Sinclair A J , Dickson C, Peters G:
ingston DM: The retinoblastoma-susceptibility gene product beCyclins D l and D2 are differentially expressed in humanBcomes phosphorylated in multiple stages during cell cycle entry and
lymphoid cell lines. Oncogene 8:1049, 1993
progression. Proc Natl Acad Sci USA 89:1795, 1992
34. Kumagai N, Benedict SH, Mills GB, Gelfand EW: Require52. Koff A, Cross F, Fisher A, Schumacher J, Leguellec K, Phiments for the simultaneous presence of phorbol esters and calcium
lippe M, Roberts JM: Human cyclin E, a new cyclin that interacts
ionophores in the expression of human T lymphocyte proliferationwith two members of the CDC2 gene family. Cell 66:1217, 1991
related genes. J Immunol 139:1393, 1987
53. KoffA, Giordano A, Desai D, Yamashita K, Harper JW,
35. Takase K, Terada N, Teraoka H, Gelfand EW, Lucas JJ:
Elledge S, Nishimoto T, Morgan DO, Franza BR, Roberts JM: ForRelease from GO/Gl arrest induced by dimethyl sulfoxide in human
mation and activation of a cyclin E-cdk2 complex during the GI
lymphoid cells: Regulation of synthesis and activation of the ~ 3 3 " ~ ~ 'phase of the human cell cycle. Science 257:1689, 1992
and ~34'~''kinases. Cell Growth Differ 5:1051, 1994
54. Dou Q-P, Levin AH, Zhao S, Pardee AB: Cyclin E and cyclin
36. Sawami H, Terada N, Franklin RA, Okawa H, Uchiyama T,
Aas candidates for the restriction point proteins. Cancer Res
Lucas JJ, Gelfand EW: Signal transduction by interleukin 2 in human
53:1493, 1993
T cells: Activation of tyrosine and ribosomal S6 kinases and cell55. Faha B, Harlow E, Lees E: The adenovirus EIA-associated
cycle regulatory genes. J Cell Physiol 151:367, 1992
kinase consists of cyclin E - ~ 3 3 " ~and
~ ' cyclin A - ~ 3 3 ' ~ ~JVirol
'.
37. Elledge SJ, Spottswood MR: Signal transduction by interleu67:2456, 1993
kin 2 in human T cells: Activation of tyrosine and ribosomal S6
56. Pagano M, Pepperkok R, Lukas J, Baldin V, Ansorge W,
kinases and cell-cycle regulatory genes. EMBO J 10:2653, 1991
Bartek J, Draetta G: Regulation of the cell cycle by the cdk2 protein
38. Hanks SK Homology probing: Identification of cDNA clones
kinase in cultured human fibroblasts. J Cell Biol 121:101, 1993
encoding members of the protein-serine kinase family. Proc Natl
57. Tsai L-H, Lees E, Faha B, Harlow E, Riabowol K: The cdk2
Acad Sci USA 84:388, 1987
kinase is required for the G1-to-S transition in mammalian cells.
39. Pines J, Hunter T: Human cyclin A is adenovirus EIA-associOncogene 8:1593, 1993
ated protein p60 and behaves differently from cyclin B. Nature
58. Pines J, Hunter T: p34'"' kinase: The S and M kinase. New
346:760, 1990
Biol 2:389, 1990
40. Lees E, Faha B, Dulic V, Reed SI, Harlow E: Cyclin E/
59. Forsburg SL, Nurse P: Cell cycle regulation in the yeasts
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
2280
Saccharomycis cerevisiae and Saccharomyces pombe. Annu Rev
Cell Biol 7:227, 1991
60. Maller JL: Mitotic control. Cum Opin Cell Biol 3:269, 1991
61. Xiong Y, Zhang H, Beach D: D type cyclins associate with
multiple protein kinases and the DNA replication and repair factor
PCNA. Cell 71:505, 1992
62. Richardson HE, Stueland CS, Thomas J, Russel P, Reed SI:
Human cDNAs encoding homologs of the small p34cd"z8'cd'2-asso~iated protein of Saccharomyces cerevisiae and Schizosaccharomyces
pombe. Genes Dev 4:1332, 1990
63. Bueno A, Russell P Two fission yeast B-type cyclins, cig2
and cdcl3, have different functions in mitosis. MolCellBiol
13:2286, 1993
64. Grandin N, Reed SI: Differential functions and expression of
Saccharomyces cerevisiae B-type cyclins in mitosis and meiosis.
MolCellBiol 13:2113, 1993
65. Cross FR, Blake CM: The yeast Cln3 protein is an unstable
activator of Cdc28. Mol Cell Biol 13:3266, 1993
66. Schwob E, Nasmyth K: CLBS and CLB6,a new pair of B
cyclins involved in DNA replication in Saccharomyces cerevisiae.
Genes Dev 7: 1 160, 1993
67. Tyers M, Futcher B: Far1 and Fus3 link the mating pheromone signal transduction pathway to three G,-phase Cdc28 kinase
complexes. Mol Cell Biol 13:5659, 1993
68. Elledge SJ, Harper JW: Cdk inhibitors: On the threshold of
checkpoints and development. Curr Opin Cell Biol 65347, 1994
69. Xiong Y, Hannon GJ, Zhang GJ, Casso D, Kobayashi R,
Beach D: p21 is a universal inhibitor of cyclin kinases. Nature
366:701, 1993
70. Polyak K, Kat0 J-Y, Solomon MJ, Sherr CJ, Massague J,
Roberts JM, KoffA: p27kip1, a cyclin-cdk inhibitor, links trans-
LUCAS ET AL
forming growth factor-@and contact inhibition to cell cycle arrest.
Genes Dev 8:9, 1994
71. Constable A, Quick S, Gray NK, Hentze MW:Modulation
of the RNA-binding activity of a regulatory protein by iron in vitro:
Switching between enzymatic and genetic function? Proc Natl Acad
Sci USA 89:4554, 1992
72. Abbruzzese A, Hanauske-Abel HM. Park MH, Henke S, Folk
JE: The active site of deoxyhypusyl hydroxylase: Use of catecholpeptides and their component chelator and peptide moieties as
molecular probes. Biochim Biophys Acta 1077:159, 199 I
73. Menick WC: Mechanism and regulation of eukaryotic protein
synthesis. Microbiol Rev 56:291, 1992
74. Bevic D, Klier H, Holter W, Tschachler E, Valent P, Lottspeich F, Baumrucker T, Hauber J: Induced gene expression of the
hypusine-containing protein eukaryotic initiation factor 5A in activatedhuman T lymphocytes. ProcNatlAcadSciUSA
91:10829,
1994
75. Nevins JR: E2F: A link betweentheRb tumor suppressor
protein and viral oncoproteins. Science 258:424, 1992
76. Feldman ST, Schonthal A: Negative regulation of histone HI
kinase expression by mimosine, a plant amino acid. Cancer Res
54:494, 1994
77. Ajchenbaum F, Ando K, DeCaprio JA, Griffin JD: Independent
regulation of human D-type cyclin gene expression during
GI phase
in primary human T lymphocytes. J Biol Chem 268:4113, 1993
78. Crissman HA, Gadbois D, Tobey RA, Bradbury EM: Transformed mammalian cells are deficient in kinase-mediated control of
progression through the G1 phase of the cell cycle. Proc Natl Acad
Sci USA 88:7580, 1991
79. Xiong Y, Zhang H, Beach D: Subunit rearrangement of the
cyclin-dependent kinases is associated with cellular transformation.
Genes Dev 7: 1572, I993
From www.bloodjournal.org by guest on June 16, 2017. For personal use only.
1995 86: 2268-2280
Effects of iron-depletion on cell cycle progression in normal human T
lymphocytes: selective inhibition of the appearance of the cyclin Aassociated component of the p33cdk2 kinase
JJ Lucas, A Szepesi, J Domenico, K Takase, A Tordai, N Terada and EW Gelfand
Updated information and services can be found at:
http://www.bloodjournal.org/content/86/6/2268.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.