Insulin-Like Growth Factor-I Inhibits Cell Growth in the A549 Non-Small
Lung Cancer Cell Line
Yuzo Kodama, Robert C. Baxter, and Janet L. Martin
Kolling Institute of Medical Research, University of Sydney, Royal North Shore Hospital, St. Leonards, New South Wales, Australia
Insulin-like growth factors (IGFs) are potent mitogenic and and are believed to regulate cell proliferation in an auto-
antiapoptotic factors for many cell types, including some nor- crine/paracrine way (2–4). The mitogenic actions of IGFs
mal and neoplastic lung cells in vitro. However, in this study
are initiated by interaction with the IGF-I receptor (IGF-IR),
we show that IGF-I, at concentrations of 10 ng/ml or greater,
significantly inhibits DNA synthesis and cell proliferation in a which has high affinity for IGF-I and IGF-II, and low af-
human lung adenocarcinoma cell line, A549. Inhibition of finity for insulin. The cellular events that follow IGF-I
DNA synthesis was completely reversed by an IGF-I receptor– binding and account for its biologic effects are still being
neutralizing antibody, IR-3, indicating that IGF-I receptor ac- elucidated, but signaling pathways that are activated upon
tivation is involved in its inhibitory effect. Attenuation of the IGF-I binding have been defined. IGF-IR is a receptor
p44/42 mitogen-activated protein kinase (MAPK) and phos-
phatidylinositol 3 -kinase (PI 3 -kinase) pathways downstream protein kinase expressed in a wide variety of cell types, in-
of the IGF-I receptor using the inhibitors PD98059 and cluding mesenchymal, epithelial, and hematopoietic cells.
LY294002, respectively, partially reversed IGF-I–induced inhi- The receptor is a transmembrane heterotetramer consist-
bition. Acute (2–60 min) and chronic (24 h) exposure of A549 ing of two subunits and two subunits linked by disul-
cells to 100 ng/ml IGF-I resulted in sustained phosphorylation fide bonds (5). Ligand binding to the receptor results in re-
of Akt/protein kinase B downstream of PI 3 -kinase, whereas
p44/42 MAPK phosphorylation was decreased in response to
ceptor oligomerization, activation of tyrosine kinase, and
chronic exposure to IGF-I. An IGF-I dose-dependent increase intermolecular receptor autophosphorylation, followed by
in the cyclin-dependent kinase inhibitor p21Cip1/WAF1 was also phosphorylation of members of a family of insulin recep-
observed over 24 h of treatment. Collectively, these data sug- tor substrates (IRSs) and the other molecules, including
gest that IGF-I is growth inhibitory to A549 cells, possibly via Shc, Crk, and Grb2 (5). Subsequent to these initial events,
sustained activation of the PI 3 -kinase signaling pathway, and
induction of p21Cip1/WAF1.
two signaling pathways that mediate many effects of IGF-I
are activated: the phosphatidylinositol 3 -kinase (PI 3 -kinase)
Lung cancer is one of the leading causes of cancer mortal- and mitogen-activated protein kinase (MAPK) pathways
ity in both men and women, and its incidence is increasing (5). Phosphorylation of cellular substrates in these signal-
all over the world. The average 5-yr survival rate of lung ing pathways consequently leads to gene activation, DNA
cancer patients is 14% (1), and this result has not im- synthesis, and cell proliferation.
proved for 30 yr. The long-term survival rate of patients It has been reported that in vitro, some lung cancer cell
with non–small cell lung carcinoma remains unsatisfac- lines express IGF mRNAs and produce IGF-I or -II, and
tory, even when they undergo complete and potentially cell growth is stimulated by IGF-I (4). In this study we un-
curative surgery. Therefore, there is an urgent need for expectedly found in A549 cells that IGF-I inhibited DNA
new strategies aimed to improve lung cancer management. synthesis and cell proliferation, and have investigated pos-
It is also essential to know how lung cancer cell growth dif- sible mechanisms by which this occurred.
fers from normal cell development and proliferation.
The insulin-like growth factors (IGFs)-I and -II are im- Materials and Methods
portant mitogens for normal and neoplastic cell types (2, 3)
Media for cell culture, glutamine, bovine serum albumin (BSA),
bovine insulin, hydrocortisone, and epidermal growth factor
(Received in original form February 12, 2002 and in revised form April 15, (EGF) were purchased from Sigma (St. Louis, MO). Cholera en-
2002) terotoxin was purchased from ICN Biomedicals Australia (Seven
Address correspondence to: Dr. Janet Martin, Kolling Institute of Medical
Hills, New South Wales, Australia). Fetal calf serum (FCS) was
Research, Royal North Shore Hospital, St. Leonards, New South Wales purchased from Trace Biosciences (North Ryde, New South
2065, Australia. E-mail: email@example.com Wales, Australia). Tissue culture plasticware was supplied by
Abbreviations: anti-insulin like growth factor receptor I antibody, IR-3;
Nunc (Roskilde, Denmark) and Corning, Inc. (Corning, NY).
bovine serum albumin, BSA; cell-conditioned media, CM; 4,6-diamidino- Receptor grade [long Arg3]IGF-I ([LR3]IGF-I) was obtained
2-phenylindole, DAPI; ethylenediaminetetraacetic acid, EDTA; epidermal from GroPep Pty. Ltd. (Adelaide, South Australia, Australia),
growth factor, EGF; fetal calf serum, FCS; insulin-like growth factor, IGF; and recombinant human IGF-I was donated by Genentech (South
IGF-I receptor, IGF-IR; IGF-binding protein, IGFBP; insulin receptor San Francisco, CA). The MAPK kinase inhibitor PD98059, the
substrate, IRS; [long Arg3]IGF-I, [LR3]IGF-I; mitogen-activated protein PI 3 -kinase inhibitor LY294002, and monoclonal antibody against
kinase, MAPK; phosphate-buffered saline, PBS; phosphatidylinositol 3 - type I IGF receptor, IR-3, were obtained from Calbiochem-
kinase, PI 3 -kinase; phosphatase and tensin homolog deleted on chromo- Novabiochem (Alexandria, New South Wales, Australia). Poly-
some ten, PTEN; protein phosphatase 2A, PP2A; radioimmunoassay, RIA;
clonal antibodies against Thr202/Tyr204 phosphorylated and total
sodium dodecyl sulfate–polyacrylamide gel electrophoresis, SDS-PAGE;
serum-free medium, SFM; Tris-HCl buffered saline, TBS; transforming
p44/42 MAPK, and Ser473 phosphorylated and total Akt/protein
growth factor- , TGF- ; 12-O-tetradecanoylphorbol-13-acetate, TPA. kinase B (Akt) were purchased from Cell Signaling (Beverley,
MA). Mouse monoclonal p21Cip1/WAF1 antibody was purchased
Am. J. Respir. Cell Mol. Biol. Vol. 27, pp. 336–344, 2002
DOI: 10.1165/rcmb.2002-0021OC from Transduction Laboratories (Lexington, KY). Electrophore-
Internet address: www.atsjournals.org sis reagents and protein molecular weight markers were pur-
Kodama, Baxter, and Martin: An Inhibitory Effect of IGF-I in Lung Cancer Cells 337
chased from Amrad Pharmacia Biotech (Sydney, New South using trypsin-EDTA. Aliquots of suspended cells were counted
Wales, Australia) and Bio-Rad Laboratories, Inc. (Richmond, using a hemacytometer.
CA). Hybond C Nitrocellulose membrane and Hyperfilm en-
hanced chemiluminescence (ECL) were purchased from Amer- Flow Cytometric Analysis
sham Pharmacia Biotech (Bucks, UK). Nonidet P-40 was pur- Cells were plated at 2 105 per well in 12-well plates for 24 h,
chased from Fluka Chemical Co. (Basel, Switzerland) then incubated with fresh SFM for 48 h before treatment with the
indicated concentrations of IGF-I for 24 h. Media containing de-
Cell Culture tached cells were collected, and adherent cells were dispersed us-
The human lung adenocarcinoma cell line A549 was obtained ing trypsin-EDTA. Cell pellets were rinsed with PBS and fixed
from Dr. Maria Kavillaris at the Children’s Cancer Research In- with 70% ethanol and stored at 20 C. For analysis, pelleted
stitute, New South Wales, Australia. A549 cells were maintained cells were rinsed with PBS, and suspended in 1 ml of fluoro-
in RPMI containing 15 mM Hepes, 10% FCS, and 2 mM gluta- chrome solution (50 g/ml propidium iodide, 1 mg/ml RNase A)
mine. MCF-10A, a phenotypically normal human breast epithe- for at least 1 h in the dark at 4 C. Cell cycle analysis was per-
lial cell line (6) which is growth-stimulated by IGF-I (7), was used formed using a Coulter ELITE flow cytometer (Coulter, Hi-
for comparison with A549 cells. MCF-10A cells were the kind gift aleah, FL). Twenty thousand cells were analyzed for each sam-
of Drs. Robert Pauley and Herbert Soule at the Karmanos Can- ple, and quantitation of cell cycle distribution was performed
cer Institute, Detroit, MI. MCF-10A cells were maintained in using Multicycle software (Phoenix Flow Systems, San Diego,
RPMI containing 15 mM Hepes, 5% horse serum, 10 g/ml bo- CA). Labeled nuclei were gated on light scatter to remove debris,
vine insulin, 20 ng/ml EGF, 100 ng/ml cholera enterotoxin, 0.5 and the percentage of nuclei with a sub-G 1 DNA content was de-
g/ml hydrocortisone, and 2 mM glutamine. Both cell lines were termined.
passaged by trypsinization every 4–5 d. For stimulation experi-
ments, cells were incubated for 48 h in serum-free RPMI medium Apoptosis Assay
containing 15 mM Hepes, 2 mM glutamine, and 1 g/liter BSA Cells were plated at 5 105 per well in six-well plates for 24 h
(SFM) before addition of test reagents in fresh SFM. and were then incubated with SFM for 48 h. Cells were treated
with or without 100 ng/ml IGF-I. After 24 h, cells were rinsed with
[3H]Thymidine Incorporation PBS, fixed in ice-cold methanol for 10 min, and then stained with
For analysis of DNA synthesis, cells were dispensed into 96-well 0.8 g/ml 4,6-diamidino-2-phenylindole (DAPI; Sigma). The per-
plates at a density of 2.0 104 cells/well. Twenty-four hours later centage of apoptotic cells was determined microscopically as cells
cells were changed into SFM for 48 h before addition of treat- with visible nuclear fragmentation.
ments. Spent media were replaced with fresh SFM containing ad-
ditives such as IGF-I and intracellular signaling pathway inhibi- Preparation of Cell Conditioned Media and Lysates for
tors, and incubations were continued for 20 h. [ 3H]Thymidine Protein Analysis
(0.5 Ci/well) was added in 50 l of SFM for a further 4 h incuba- Cells were seeded onto 12-well plates at a density of 2.0 105
tion at 37 C. Monolayers were rinsed once with ice-cold saline cells per well. After 24 h the spent media were replaced with
and fixed with 0.2 ml/well ice-cold methanol:acetic acid (3:1) at SFM, then incubated for 48 h. Cells were treated with or without
4 C for a minimum of 2 h. Cells were solubilized in 0.5% sodium 0.1, 10, or 100 ng/ml IGF-I for times indicated in individual exper-
dodecyl sulfate (SDS), and 200 l of each lysate mixed with scin- iments. Cell-conditioned media (CM) were collected and stored at
tillant (UltimaGold; Packard Biosciences, Groningen, The Neth- 20 C before assay. Cells were then washed once with cold PBS
erlands) before counting for 1 min in a Hewlett-Packard counter. and immediately homogenized in SDS-polyacrylamide gel elec-
trophoresis (SDS-PAGE) sample buffer (62.5 mM Tris-HCl, pH
Cell Proliferation Assays 6.8, 2% SDS, 10% glycerol, 50 mM dithiothreitol, 0.1% bromphe-
Cells were plated onto six-well plates at a density of 1.0 105 nol blue) or PBS-TDS buffer (PBS, 1% Triton X-100, 0.1% SDS,
cells/well for 24 h. Following serum starvation for 48 h, cells were 0.5% sodium deoxycholate, 1 mM sodium orthovanadate, 4 mM
treated with or without IGF-I at the concentration of 0.1, 10, or sodium pyrophosphate, 10 mM sodium fluoride and protease in-
100 ng/ml in 2 ml SFM for 3 d. Media were removed, cells were hibitor cocktail tablet (Complete; Roche Molecular Biochemicals,
washed with phosphate-buffered saline (PBS) twice and dispersed Mannheim, Germany), sonicated for 15 s, and frozen at 80 C.
Figure 1. The effect of IGF-I and [LR3]IGF-I
on DNA synthesis in A549 cells. A549 cells (A)
and MCF-10A cells (B) were treated with IGF-I
(open symbols) or [LR3]IGF-I (closed symbols)
at the indicated concentrations for 24 h, and
DNA synthesis was determined by [ 3H]thymi-
dine incorporation as described in MATERIALS
AND METHODS. The data are expressed as per-
centage of the counts incorporated in the ab-
sence of additions (SFM), and points shown are
mean SE of combined data from six experi-
ments for IGF-I or two experiments for [LR3]IGF-I
performed in triplicate. Significant difference
in DNA synthesis compared with control (SFM)
are indicated as *P 0.05, **P 0.005, and †P
0.0001. Statistical significance was determined
by ANOVA and Fisher’s PLSD test.
338 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 27 2002
Determination of Secreted IGFBP-3 TABLE 1
IGFBP-3 concentrations in CM were determined by radioimmu- Effect of IGF-I on DNA synthesis in A549 cells under
noassay (RIA) as previously described (8). Briefly, the assay was different growth conditions
performed in a buffer containing 0.1 M sodium phosphate, pH [3H]thymidine incorporation
6.5, 0.02% sodium azide, and 0.25% bovine albumin. Incubation (% of 24 h control)
mixtures (500 l total volume) consisted of appropriately diluted
samples or standards (50 l), antiserum R100 diluted 1:20,000 Time course 24 h 48 h
(100 l), cross-linked IGFBP-3-IGF-I tracer, 10,000 cpm (100 Control 110.0 1.2 42.4 3.7
l), and assay buffer (250 l), added in that order and vortex IGF-I 61.8 5.3* 26.3 1.4*
mixed. The standard range was 0.1–20 ng of pure IGFBP-3. After % reduction 61.8 61.9
16 h incubation at 22 C, 0.5 l normal rabbit serum and 2 l goat
anti-rabbit immunoglobulin were added and, after a further 30 A549 cells were incubated in SFM for 24 or 48 h as indicated, then treated
with 100 ng/ml IGF-I for 24 h. DNA synthesis was determined by [3H]thymidine
min incubation, 1 ml of ice-cold 6% polyethylene glycol solution incorporation as described in MATERIALS AND METHODS. Data shown are pooled
(PEG 6,000) in 0.15 M sodium chloride was added, and tubes from two experiments carried out in triplicate.
were centrifuged 20 min at 4,200 rpm in a Beckman J-6 centrifuge * P 0.0001 compared to control.
(Beckman, Palo Alto, CA) cooled to 2 C. Supernatants were de-
canted, and the radioactivity in the pellets determined in a
counter. For immunoblotting signaling intermediates, protein content
in PBS-TDS lysates was measured by the method of Bradford us-
SDS-PAGE and Western Blotting ing a Bio-Rad kit (Richmond, CA), mixed with SDS-PAGE sam-
For ligand blotting of secreted IGFBPs, 100 l of CM were mixed ple buffer as described above. All samples were heated at 95 C
with sample buffer (62.5 mM Tris-HCl, pH 6.8, 2% SDS, 10% for 5 min, then fractionated on 12% gels, before transfer to nitro-
glycerol, 0.1% bromphenol blue) and boiled for 5 min, fraction- cellulose membrane. Blots were blocked using 5% skim milk
ated by 12% SDS-PAGE under nonreducing conditions. Proteins powder in TBS containing 0.1% Tween 20 (TBS-T), then probed
were transferred to Hybond C nitrocellulose membrane for 2 h at with the phospho-Thr202/Tyr204 p44/42 antibody (1:2,000 dilution)
150 mA. Blots were incubated in blocking buffer (TBS; 10 mM or p21Cip1/WAF1 antibody (1:500 dilution) in the same buffer, or
Tris-HCl, pH 7.4, containing 9 g/liter NaCl, 10 g/liter BSA, 0.05% phospho- Ser473 Akt antibody (1:2,000 dilution) in TBS-T con-
Nonidet P-40, and 0.2 g/liter sodium azide) at 37 C for 2 h, then taining 5% BSA overnight at 4 C and protein detected using en-
probed with 125I-labeled IGF-I (1 106 cpm) diluted in blocking hanced chemiluminescence (Supersignal ECL; Pierce, Rockford,
buffer, overnight at 22 C. Blots were washed in TBS containing VA). Blots were stripped by submersion in commercial stripping
0.05% Nonidet P-40, before autoradiography using Hyperfilm buffer reagent (Restore Western Blot Stripping Buffer; Pierce)
MP for 3 d at 80 C. for 20 min at 22 C, then reprobed with total p44/42 MAPK anti-
Figure 2. IGF-IR–dependent mechanism of
IGF-I–induced inhibition of DNA synthesis. In
A and B, media conditioned for 24 h by A549
cells in the absence or presence of IGF-I at the
indicated concentrations were analyzed by ligand
blotting with [125I]IGF-I (A) and IGFBP-3 RIA
(B), as described in MATERIALS AND METHODS.
(C) A549 cells were treated for 24 h with the
indicated concentrations of IGFBP-3, and DNA
synthesis was determined as described in MA-
TERIALS AND METHODS. 100% represents the
counts incorporated in the absence of additions
(SFM). Data shown are the mean SE of com-
bined data from two experiments performed in
triplicate. (D) A549 cells were pretreated for
1 h with or without anti–IGF-I receptor antibody
IR-3 (final concentration 10 g/ml), then
IGF-I was added (final concentration 100 ng/ml)
for 24 h before assessment of DNA synthesis as
described in MATERIALS AND METHODS. Results are
expressed as percentage of the counts incorpo-
rated in the absence of additions (SFM). Points
shown are mean SE of pooled data from two
independent experiments, performed in tripli-
cate or quadruplicate wells. In B, significant dif-
ferences compared with control are indicated as
*P 0.005, **P 0.0001; and in D, *P 0.05
compared with untreated control, **P 0.001
compared with IGF-I in the absence of IR-3 (no
other treatments were significantly different from
control). Statistical significance was determined
by ANOVA and Fisher’s PLSD test.
Kodama, Baxter, and Martin: An Inhibitory Effect of IGF-I in Lung Cancer Cells 339
Figure 3. IGF-I inhibition of cell pro-
liferation but no induction of apoptosis
in A549 cells. Cells were treated with
or without IGF-I at the concentrations
shown for three days (A) or at 100 ng/ml
for 24 h (B and D) as described in MA-
TERIALS AND METHODS. (A) Cells were
counted by hemacytometer after trypsiniza-
tion. Data are shown as the increase in
cell number over the 3-d test period. (B)
Cells were analyzed by flow cytometry,
and the percentage of cells in the pre-
G1 peak was determined; the data are
summarized in C. (D) DAPI-stained
cells were scored for nuclear fragmenta-
tion (arrows) as a morphologic marker
of apoptosis, and the percentage of
apoptotic cells was determined and
summarized in E. In A, points shown
are mean SE of combined data (de-
termined in triplicate wells) from three
experiments; in C, values shown are
means of three independent data sets
SE; in E, values shown are means of two
data sets SE (A) Significant de-
creases in cell number compared with
control (SFM) are indicated as *P
0.001 and **P 0.0001. (C) Significant
differences are indicated as *P 0.02
and **P 0.001. (E) Significant differ-
ences are indicated as *P 0.0001.
Statistical significance was determined
by ANOVA and Fisher’s PLSD test.
body (1:2,000 dilution) or Akt antibody (1:2,000 dilution) as ap- used for the purpose of comparison because it shows a
propriate and ECL as before. Band densities on developed films more typical response to IGF-I (7). IGF-I has previously
were measured using National Institutes of Health (NIH) image been reported to be mitogenic for A549 cells when added
software (version 1.61). either immediately after the removal of serum-containing
medium, or 24 h later (2, 3, 9, 10). We evaluated whether
Statistical Analysis IGF-I could inhibit DNA synthesis at different time-points,
Individual experiments were conducted in triplicate or quadru- 24 and 48 h after the removal of serum, to compare the ef-
plicate wells; experiments were performed at least three times fect of IGF-I in A549 cells with previous reports. As
unless indicated otherwise. Data were analyzed by ANOVA and shown in Table 1, IGF-I inhibited thymidine incorporation
Fisher’s protected least significant difference test using the Stat- by 60% with no difference between the different incu-
View program for Macintosh (SAS Institute, Inc., Cary, NC); dif-
ferences were considered significant where P 0.05.
A549 cells secrete various IGFBPs, which may affect
responsiveness to IGFs (3, 10). We therefore evaluated
Results whether IGFBPs were involved in the IGF-I–induced inhi-
In the A549 cell line IGF-I had a biphasic action on DNA bition of DNA synthesis or whether the IGF-IR was in-
synthesis measured by thymidine incorporation, which was volved, using the IGF-I analog [LR3]IGF-I, which can
significantly stimulated at up to 1 ng/ml IGF-I (P 0.005) bind to the IGF-IR but has greatly reduced affinity for
and unexpectedly inhibited at more than 10 ng/ml (P IGFBPs. Similarly to IGF-I, [LR3]IGF-I at 0.1 ng/ml sig-
0.0001; Figure 1A). Similar results were seen in more than nificantly stimulated DNA synthesis (P 0.05) and inhib-
10 experiments. In contrast, IGF-I stimulated thymidine ited DNA synthesis at concentrations of 3 ng/ml or
incorporation 5- to 8-fold in a dose-dependent manner in greater (P 0.0001) (Figure 1A), but with 3- to 5-fold
MCF-10A cells (P 0.0001; Figure 1B). This cell line was greater potency than IGF-I. This differential effect be-
340 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 27 2002
tween IGF-I and [LR3]IGF-I suggests that the inhibition agents suggests that both MAPK and PI 3 -kinase path-
does not require interaction between IGF-I and IGFBPs, ways downstream of IGF-IR activation may be involved in
but that it maybe limited by endogenous IGFBPs. We ex- the suppression of DNA synthesis induced by IGF-I.
amined the secretion of IGFBPs induced by IGF-I. Media We then evaluated intracellular signaling pathways in
conditioned by A549 cells treated with the indicated con- A549 cells in the presence or absence of IGF-I at 100 ng/
centrations of IGF-I were analyzed by IGF ligand blot. As ml for the indicated times (up to 60 min) to assess the ef-
shown in Figure 2A, IGFBP-3 was readily detected and in- fect of acute exposure to IGF-I, or at the indicated concen-
creased in a dose-dependent manner by IGF-I. A 30-kD trations for 24 h (chronic exposure to IGF-I). In A549
IGFBP was also detected but not affected by IGF-I treat- cells, IGF-I significantly stimulated Akt with sustained ac-
ment. We confirmed a significant increase of IGFBP-3 tivation up to 60 min (Figures 5A and 5C). MAPK showed
with IGF-I treatment by RIA (Figure 2B). However, ex- variable activation (Figure 5B), not significant as deter-
ogenous IGFBP-3 had no inhibitory effect on DNA syn- mined by quantitating signal intensity in two separate ex-
thesis in A549 cells (Figure 2C), suggesting that IGFBP-3 periments (Figure 5D). In contrast to A549 cells, MCF-
is not responsible for the IGF-I–induced suppression of 10A cells showed significant activation of both Akt and
DNA synthesis. MAPK by IGF-I at 4 min, declining at 8 min for Akt or at
To confirm involvement of the IGF-IR in the inhibitory 60 min for MAPK (Figures 5A–5D). Measured after 24 h
effect of IGF-I, we evaluated whether IR-3 could abolish exposure to IGF-I, phosphorylation of Akt remained stim-
the suppression of DNA synthesis induced by IGF-I. Cells ulated by 100 ng/ml IGF-I in A549 cells but not in MCF-
were treated with 10 g/ml IR-3 in the presence or ab- 10A cells (Figures 6A and B). In contrast, MAPK activation
sence of 100 ng/ml IGF-I. IGF-I–induced suppression of was significantly decreased in a dose-dependent manner
DNA synthesis in A549 cells was completely reversed by by IGF-I (P 0.005 for A549 cells, P 0.02 for MCF-10A
IR-3 (P 0.001; Figure 2D), suggesting that the inhibi- cells; Figures 6C and D).
tory effect of IGF-I at 100 ng/ml on DNA synthesis in Because sustained activation of Akt has previously been
A549 cells is mediated through the IGF-IR. Taken together, reported to lead to induction of the cell cycle kinase inhibi-
these data indicate that intracellular signaling via the IGF-IR tor p21Cip1/WAF1 (11), we examined this in A549 cells treated
rather than extracellular modulation by any IGFBPs may with the indicated concentrations of IGF-I. As shown in
be involved in the inhibitory effect of IGF-I in A549 cells. Figure 7, p21Cip1/WAF1 was increased in a dose-dependent
We then evaluated cell growth in A549 cells treated with manner by 24 h treatment with 100 ng/ml IGF-I, suggesting
various concentrations of IGF-I for 3 d. The accumulation a possible explanation for the decrease in DNA synthesis.
of cells over this period was significantly inhibited by IGF-I
at concentrations of 10 and 100 ng/ml (P 0.001; Figure Discussion
3A); a similar effect was seen after 5 d treatment with the In recent years the IGF axis has been shown to have im-
highest dose of IGF-I (data not shown). We investigated portant influences on cancer biology, cancer risk, and car-
whether induction of apoptosis could contribute to the inhi-
bition of cell proliferation. The fragmentation of DNA
characteristic of apoptosis results in a hypodiploid DNA
content that can be visualized as a pre-G1 peak on a DNA
cell cycle histogram. Figure 3B shows the percentage of
cells in the pre-G1 peak of both control and IGF-I–treated
A549 cells. The pre-G1 peak was significantly decreased in
a dose-dependent manner by IGF-I (P 0.02), indicating
that IGF-I, although causing a decrease in cell number, was
also inhibitory to apoptosis (Figure 3C). We also analyzed
changes in nuclear morphology indicative of apoptosis us-
ing the cell-permeable DNA dye DAPI and scoring for the
presence of nuclear fragmentation (Figure 3D). A signifi-
cant decrease in apoptosis in IGF-I–treated cells compared
with controls was observed (P 0.0001; Figure 3E), con-
firming the results of the flow cytometric analysis.
To examine whether intracellular signaling through the
IGF-IR is involved in the IGF-I inhibitory effect, we used Figure 4. Effect of signaling pathway inhibitors on DNA synthe-
inhibitors of signaling intermediates of two key pathways sis in A549 cells treated with IGF-I. A549 cells were treated for
downstream of the IGF-IR, the PI 3 -kinase pathway and 24 h with or without 100 ng/ml IGF-I in the presence or absence
the p44/42 MAPK pathway. Interestingly, IGF-I–inhibited of either LY294002 (to block PI 3 -kinase) or PD98059 (to block
DNA synthesis in A549 cells could partly be reversed by p44/42 MAPK) at 1 M concentration, and DNA synthesis was
determined as described in MATERIALS AND METHODS. The data
either LY294002 (1 M) or PD98059 (1 M), which inhibit
are expressed as percentage of the counts incorporated in the ab-
the PI 3 -kinase and p44/42 MAPK pathways, respectively, sence of additions (SFM). Points shown are mean SE of com-
although the effects of these agents were modest (P 0.05; bined data (determined in triplicate wells), with the PD98059 ex-
Figure 4). Higher concentrations of each agent were also periment performed two times and the LY294002 experiment
tested, but were strongly inhibitory to basal thymidine incor- performed three times. *P 0.05 and **P 0.001 compared
poration. The partial reversal of IGF-I inhibition by these with 100 ng/ml IGF-I, by ANOVA and Fisher’s PLSD.
Kodama, Baxter, and Martin: An Inhibitory Effect of IGF-I in Lung Cancer Cells 341
Figure 5. Effect of acute exposure to IGF-I on phosphorylation of Akt and p44/42 MAPK. In A and B, cells were plated at 2.0 105 cells/well
in 12-well plates in RPMI medium with 15 mM Hepes, 2 mM glutamine, and 10% FCS or 5% horse serum. After 24 h, the spent media were
changed to serum-free medium with 0.1% BSA and 2 mM glutamine and incubated for 48 h. Cells were treated with or without IGF-I at 100
ng/ml for the indicated times (up to 60 min) and homogenized as described in MATERIALS AND METHODS. Fifty microliters of each lysate
was resolved by SDS-PAGE and immunoblotted with anti–phospho-Akt (Ser473) antibody, or anti-Akt antibody, and anti–phospho-p44/42
MAPK (Thr202/Tyr204) antibody, or anti-p44/42 antibody as indicated. Data are representative of two experiments. In C and D, band densi-
ties were measured by NIH image software, and the ratio of phosphorylated to total was then calculated from two experiments and ex-
pressed as a percentage of maximal ratio. Data shown are the mean SE of combined data from two experiments performed in duplicate for
both cells. (C) *P 0.003 and **P 0.0001 compared with 0 min control; †P 0.001 compared with 4 min. (D) *P 0.0005 compared
with 0 min control, by ANOVA and Fisher’s PLSD.
cinogenesis (12). Population studies are providing evi- the suppression of DNA synthesis in A549 cells similarly
dence for a possible relationship between circulating levels to IGF-I, but with greater apparent potency. This suggests
of both IGF-I and IGFBP-3 and the risk of several com- that an IGF–IGFBP interaction is not involved in the inhi-
mon cancers, including premenopausal breast cancer, co- bition of DNA synthesis in A549 cells treated with IGF-I,
lon cancer, prostate cancer, and lung cancer (13). More- and indeed endogenous IGFBPs may act to limit this ef-
over, in vitro studies also indicate that IGF-I acts as a fect. We also found that exogenous IGFBP-3 had no effect
mitogenic factor in many normal and neoplastic cells. In this on DNA synthesis. It is therefore unlikely that IGFBP-3 is
study we have described the unexpected action of IGF-I as involved in the inhibitory action of IGF-I in A549 cells, al-
an inhibitor of DNA synthesis and cell proliferation in the though we cannot exclude a contribution from endogenous
human lung adenocarcinoma cell line A549. IGFs interact IGFBP-3. Because IGF-I and [LR3]IGF-I activate the IGF-IR
with a family of soluble IGFBPs, of which six distinct similarly, our data are consistent with a mechanism involv-
members (IGFBP-1 to IGFBP-6) have been identified and ing signaling via the IGF-IR. This is supported by the dem-
cloned (14). IGFBP-3 appears to regulate cell prolifera- onstration that IGF-IR antibody, IR3 could completely
tion as a negative mediator in some cancer cell lines such reverse the IGF-I–induced suppression of DNA synthesis
as breast and lung cancer either in vitro or in vivo (15, 16). in A549 cells. A recent study has shown that exposure to
The expression and the production of IGFBPs has been IGF-I can reduce the sensitivity of this cell line to IGF ac-
evaluated in many lung cancer cell lines, including A549 tions, by downregulating the IGF-IR (18). However, our
cells (3, 10), which secrete various IGFBPs, predominantly results indicate that the IGF-IR was still capable of effec-
IGFBP-3. We have previously shown that IGFBP-3 can tive—albeit inhibitory—signaling after 24 h exposure to
act as an inhibitory factor in breast cancer cells and im- IGF-I.
mortalized breast epithelial cells as a result of inhibition of In direct contrast to our observation of decreased DNA
DNA synthesis as well as induction of apoptosis (7, 17). In synthesis caused by IGF-I in A549 cells, the same cell line
the present study, IGFBP-3 was upregulated by IGF-I, an was reported to show increased thymidine uptake and cell
effect also observed by others in A549 cells (18). There- proliferation when stimulated by IGF-I after 24 h under
fore, we hypothesized that this protein could be involved serum-free conditions, and this effect was abolished by
in the suppression of DNA synthesis in A549 cell treated IR-3 (2, 3, 9, 10). Although most of our studies were per-
with IGF-I. formed after a 48-h serum-free period, we saw identical re-
The IGF-I analog, [LR3]IGF-I, which has extremely sults after only 24 h. The difference between our results
low affinity for IGFBPs but binds to the IGF-IR, induced and those previously published cannot, therefore, be ex-
342 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 27 2002
Figure 6. Effect of chronic exposure to IGF-I on signaling pathways. In A and B, cells were treated with or without IGF-I at the concen-
trations shown for 24 h and homogenized as described in MATERIALS AND METHODS. One hundred micrograms of proteins from cell ex-
tracts was resolved by SDS-PAGE and immunoblotted with antibodies against phospho-Akt antibody, or total Akt antibody, and phos-
pho-p44/42 MAPK, or total MAPK as indicated. In C and D, band densities were measured by NIH image software, and the ratio of
phosphorylated to total was then calculated and expressed as a percentage of maximal ratio. Data shown are the mean SE from com-
bined data. (C) *P 0.0001 compared with control. (D) *P 0.02, **P 0.005, †P 0.0005, and ††P 0.0001 compared with control,
by ANOVA and Fisher’s PLSD.
plained at present, but may result from other changes in two growth factors are not identical in their biologic ef-
cell culture conditions, or different subclones of A549 cells fects even though they bind to the same receptor.
in different laboratories. IGF-I at 20 ng/ml (inhibitory in The phorbol ester 12-O-tetradecanoylphorbol-13-acetate
our study) has recently been reported to have a biphasic (TPA) has also been reported as another negative regula-
effect on colon carcinoma cell lines, causing transient growth tor in various cell lines including A549 cells (21, 22), al-
stimulation followed by growth inhibition, believed to be though it acts as a mitogenic factor in many cell lines and
mediated through upregulation of p27kip1 (19). Some other activates MAPK. The duration of MAPK activity may be
growth factors, normally regarded as stimulating cell pro- important in determining a cellular response; for example,
liferation, are known to inhibit the growth of certain cell prolonged activation of the MAPK pathway may play a
lines. Siegfried reported that EGF was inhibitory to lung role in cell proliferation as well as in differentiation, de-
tumor cells including A549 cells (20). Interestingly, the pending on the cell type. However, in our study acute ex-
EGF receptor ligand, transforming growth factor- (TGF- posure of A549 cells to IGF-I had no consistent effect on
) stimulated rather than inhibited growth in A549 cells as MAPK activation, and 24 h exposure actually decreased
well as primary lung tumor cells. The opposite response of MAPK phosphorylation significantly in A549 cells, which
lung cancer cells to EGF and TGF- suggests that these were growth-inhibited, and even more so in MCF-10A
cells, which were growth-stimulated. It is therefore unclear
how regulation of MAPK phosphorylation by IGF-I re-
lates to the inhibition of DNA synthesis in these cells.
Nevertheless, the MAPK kinase inhibitor PD98059 repro-
ducibly caused a partial reversal of the inhibitory effect,
and the possibility remains that inhibition of the MAPK
pathway by IGF-I in A549 cells contributed to their de-
creased DNA synthesis.
The PI 3 -kinase inhibitor LY294002 also partially re-
versed the inhibitory effect of IGF-I on thymidine incor-
Figure 7. The effect of IGF-I on induction of p21 Cip1/WAF1. Cells
were exposed to IGF-I at the concentrations shown for 24 h, then poration. Akt phosphorylation, which is PI 3 -kinase–
lysated and immunoblotting with p21 Cip1/WAF1 antibody. Total dependent, is reported to be constitutively activated in
Akt data were used as a loading control as shown in Figure 6E, as most non–small cell lung cancer lines, although only weakly
described in MATERIALS AND METHODS. Representative data for in A549 (23). In the present study, the phosphorylation of
one of three experiments with similar results are shown. Akt increased within 4 min of exposure to IGF-I, and re-
Kodama, Baxter, and Martin: An Inhibitory Effect of IGF-I in Lung Cancer Cells 343
mained elevated at 60 min in A549 cells, in contrast to the cellular functions (i.e., both proliferative and antiprolifera-
transient response seen in MCF-10A cells. Even after 24 h tive effects) under different conditions. The outcome of
exposure to 100 ng/ml IGF-I, Akt phosphorylation re- IGF-I stimulation may depend not only on specific pheno-
mained high in A549 cells. This sustained phosphorylation typic features of A549 cells, but also on variations of the
could indicate that the PI 3 -kinase pathway upstream of Akt microenvironment in which the IGF-I signal is generated.
remains activated, or that the dephosphorylation of Akt it- This concept may have important implications for under-
self does not occur appropriately. The phosphatidylinositol standing the clinical effects of biologic response modifiers
phosphatase, PTEN, acts to disrupt signaling through PI and other therapeutic agents when used in the manage-
3 -kinase, and a loss of PTEN activity would be expected to ment of lung cancer. Insulin has been reported to inhibit
cause sustained Akt activation. However, a survey of PTEN cell growth in A549 cells as well as primary lung carcinoma
mutations in non-small lung cancer cell lines found that the cells, but not normal bronchial epithelial cells (20). This
great majority—23 out of 25, including A549—lacked PTEN suggests that A549 cells may resemble primary lung can-
mutations or deletions, although 40% of small cell lung cers in their response to agents such as IGF-I. Therefore,
cancer lines had intragenic PTEN deletions (24). Dephos- further investigation of the mechanism of inhibition of DNA
phorylation of Akt by the protein phosphatase PP2A is synthesis by IGF-I in this lung cancer line may shed new
also believed to play a major role in regulating Akt activity light on the aberrant growth regulation of lung cancer cells.
(25). Although PP2A mutations are reported in lung cancer
(26), the PP2A status of A549 cells has not been described. Acknowledgments: The authors thank Dr. Malcolm A. King (Clinical Immunol-
ogy, Royal North Shore Hospital, Sydney) for flow cytometric analysis, and Pro-
Because non-small lung cancer cell lines with constitutively fessor Norbert Berend (Department of Respiratory Medicine, Royal North
active Akt/PKB are resistant to chemotherapy and irradia- Shore Hospital, Sydney) for generous advice and discussions. This work was sup-
tion (23), A549 cells with sustained active Akt/PKB caused ported in part by grant No. 107244 from the National Health and Medical Re-
search Council, Australia (J.L.M., R.C.B.) and a scholarship from the Graduates’
by treatment with IGF-I might have resistance to chemo- Association of Juntendo University School of Medicine (Tokyo, Japan) (Y.K.).
therapy and irradiation, though this remains to be proven.
Although Akt phosphorylation may be associated with
growth-stimulatory signaling through PI 3 -kinase, sus- References
tained Akt upregulation has also been reported to induce 1. Lippman, S. M., and M. R. Spitz. 2001. Lung cancer chemoprevention: an
the cell cycle inhibitor p21Cip1/WAF1 (11), which at high con- integrated approach. J. Clin. Oncol. 19:74s–82s.
2. Ankrapp, D. P., and D. R. Bevan. 1993. Insulin-like growth factor-I and hu-
centration binds to, and inhibits, cyclin–cyclin-dependent man lung fibroblast-derived insulin-like growth factor-I stimulate the pro-
kinase complexes (27). Akt has recently been shown to liferation of human lung carcinoma cells in vitro. Cancer Res. 53:3399–3404.
phosphorylate p21Cip1/WAF1, resulting in enhanced protein 3. Favoni, R. E., A. de Cupis, F. Ravera, C. Cantoni, P. Pirani, A. Ardizzoni,
D. Noonan, and R. Biassoni. 1994. Expression and function of the insulin-
stability (28). We found a strong upregulation of p21Cip1/WAF1 like growth factor I system in human non-small-cell lung cancer and nor-
in A549 cells after 24 h exposure to 100 ng/ml IGF-I, con- mal lung cell lines. Int. J. Cancer 56:858–866.
comitant with the sustained increase in Akt phosphoryla- 4. Quinn, K. A., A. M. Treston, E. J. Unsworth, M. J. Miller, M. Vos, C. Grim-
ley, J. Battey, J. L. Mulshine, and F. Cuttitta. 1996. Insulin-like growth fac-
tion. Induction of p21Cip1/WAF1 by IGF-I has been described tor expression in human cancer cell lines. J. Biol. Chem. 271:11477–11483.
previously, in breast cancer cells, although in that situation 5. LeRoith, D., H. Werner, D. Beitner-Johnson, and C. T. Roberts, Jr. 1995.
Molecular and cellular aspects of the insulin-like growth factor I receptor.
it was paradoxically associated with increased, rather than Endocr. Rev. 16:143–163.
inhibited, cell cycle activity (29). This is believed to reflect 6. Soule, H. D., T. M. Maloney, S. R. Wolman, W. D. Peterson, Jr., R. Brenz,
a stimulatory role of p21Cip1/WAF1 at low concentrations in C. M. McGrath, J. Russo, R. J. Pauley, R. F. Jones, and S. C. Brooks. 1990.
Isolation and characterization of a spontaneously immortalized human
the assembly of cell cycle kinase complexes (27). It is pos- breast epithelial cell line, MCF-10. Cancer Res. 50:6075–6086.
sible that this effect accounts for the slight stimulation of 7. Martin, J. L., and R. C. Baxter. 1999. Oncogenic ras causes resistance to the
DNA synthesis observed at low IGF-I concentrations in growth inhibitor insulin-like growth factor binding protein-3 (IGFBP-3) in
breast cancer cells. J. Biol. Chem. 274:16407–16411.
A549 cells. In other situations where p21Cip1/WAF1 upregu- 8. Baxter, R. C., and J. L. Martin. 1986. Radioimmunoassay of growth hormone-
lation by growth factors has been described, the effect on dependent insulinlike growth factor binding protein in human plasma. J.
Clin. Invest. 78:1504–1512.
cell proliferation has been inhibitory (30, 31). The increase 9. Price, W. A., B. M. Moats-Staats, and A. D. Stiles. 1995. Insulin-like growth
in p21Cip1/WAF1 caused by IGF-I that we observed in A549 factor-I (IGF-I) regulates IGFBP-3 and IGFBP-4 by multiple mechanisms
cells may therefore account for their inhibited DNA syn- in A549 human adenocarcinoma cells. Am. J. Respir. Cell Mol. Biol. 13:
thesis and proliferation. 10. Noll, K., B. R. Wegmann, K. Havemann, and G. Jaques. 1996. Insulin-like
Despite its possible role in p21Cip1/WAF1 induction in growth factors stimulate the release of insulin-like growth factor-binding
A549 cells, Akt is known as a potent cell survival factor, protein-3 (IGFBP-3) and degradation of IGFBP-4 in nonsmall cell lung
cancer cell lines. J. Clin. Endocrinol. Metab. 81:2653–2662.
inhibiting the pro-apoptotic proteins Bax and Bad, as well 11. Lawlor, M. A., and P. Rotwein. 2000. Insulin-like growth factor-mediated
as other important apoptosis effectors (32). Consistent muscle cell survival: central roles for Akt and cyclin-dependent kinase in-
hibitor p21. Mol. Cell. Biol. 20:8983–8995.
with this, IGF-I was inhibitory to apoptosis in A549 cells, 12. Pollak, M. 2000. Insulin-like growth factor physiology and cancer risk. Eur.
as judged by flow cytometric quantitation of cells with a J. Cancer 36:1224–1228.
pre-G1 DNA content and by morphologic change with 13. Wu, X., H. Yu, C. I. Amos, W. K. Hong, and M. R. Spitz. 2000. Joint effect
of insulin-like growth factors and mutagen sensitivity in lung cancer risk. J.
DAPI staining. Nevertheless, the increase in A549 cell Natl. Cancer Inst. 92:737–743.
number over 3 d of exposure was reduced by exposure to 14. Shimasaki, S., and N. Ling. 1991. Identification and molecular characteriza-
IGF-I. This suggests that the marked IGF-I–induced block- tion of insulin-like growth factor binding proteins (IGFBP-1, -2, -3, -4, -5
and -6). Prog. Growth Factor Res. 3:243–266.
ade of cell cycle activity caused a net decrease in cell pro- 15. Huynh, H., X. Yang, and M. Pollak. 1996. Estradiol and antiestrogens regu-
liferation despite the concomitant inhibition of cell death. late a growth inhibitory insulin-like growth factor binding protein 3 auto-
crine loop in human breast cancer cells. J. Biol. Chem. 271:1016–1021.
In conclusion, our data indicate that IGF-I may be a 16. Hochscheid, R., G. Jaques, and B. Wegmann. 2000. Transfection of human
pleiotropic effector capable of inducing mutually exclusive insulin-like growth factor-binding protein 3 gene inhibits cell growth and
344 AMERICAN JOURNAL OF RESPIRATORY CELL AND MOLECULAR BIOLOGY VOL. 27 2002
tumorigenicity: a cell culture model for lung cancer. J. Endocrinol. 166: the PTEN/MMAC1/TEP1 gene in human lung cancers. Genes Chromo-
553–563. somes Cancer 22:152–156.
17. Butt, A. J., S. M. Firth, M. A. King, and R. C. Baxter. 2000. Insulin-like growth 25. Millward, T. A., S. Zolnierowicz, and B. A. Hemmings. 1999. Regulation of
factor-binding protein-3 modulates expression of Bax and Bcl-2 and potentiates protein kinase cascades by protein phosphatase 2A. Trends Biochem. Sci.
p53-independent radiation-induced apoptosis in human breast cancer cells. J. 24:186–191.
Biol. Chem. 275:39174–39181. 26. Wang, S. S., E. D. Esplin, J. L. Li, L. Huang, A. Gazdar, J. Minna, and G. A.
18. Bostedt, K. T., C. Schmid, C. Ghirlanda-Keller, R. Olie, K. H. Winterhalter, and Evans. 1998. Alterations of the PPP2R1B gene in human lung and colon
J. Zapf. 2001. Insulin-like growth factor (IGF) I down-regulates type 1 IGF re- cancer. Science 282:284–287.
ceptor (IGF 1R) and reduces the IGF I response in A549 non-small-cell lung can- 27. LaBaer, J., M. D. Garrett, L. F. Stevenson, J. M. Slingerland, C. Sandhu,
cer and Saos-2/B-10 osteoblastic osteosarcoma cells. Exp. Cell Res. 271:368–377. H. S. Chou, A. Fattaey, and E. Harlow. 1997. New functional activities for
19. Ewton, D. Z., S. Kansra, S. Lim, and E. Friedman. 2002. Insulin-like growth the p21 family of CDK inhibitors. Genes Dev. 11:847–862.
factor-I has a biphasic effect on colon carcinoma cells through transient in- 28. Li, Y., D. Dowbenko, and L. Lasky. 2001. Akt/PKB phosphorylation of
activation of forkhead1, initially mitogenic, then mediating growth arrest p21Cip/WAF1 enhances protein stability of p21Cip/WAF1 and promotes
and differentiation. Int. J. Cancer 98:665–673. cell survival. J. Biol. Chem. 277:11352–11361.
20. Siegfried, J. M. 1987. Detection of human lung epithelial cell growth factors 29. Dupont, J., M. Karas, and D. LeRoith. 2000. The potentiation of estrogen
produced by a lung carcinoma cell line: use in culture of primary solid lung on insulin-like growth factor I action in MCF-7 human breast cancer cells
tumors. Cancer Res. 47:2903–2910. includes cell cycle components. J. Biol. Chem. 275:35893–35901.
21. Bradshaw, T. D., A. Gescher, and G. R. Pettit. 1991. The effect of fetal calf 30. Johnson, M. R., C. Valentine, C. Basilico, and A. Mansukhani. 1998. FGF
serum on growth arrest caused by activators of protein kinase C. Int. J. signaling activates STAT1 and p21 and inhibits the estrogen response and
Cancer 47:929–932. proliferation of MCF-7 cells. Oncogene 16:2647–2656.
22. Alblas, J., R. Slager-Davidov, P. H. Steenbergh, J. S. Sussenbach, and B. 31. Johannessen, L. E., S. L. Knardal, and I. H. Madshus. 1999. Epidermal
van der Burg. 1998. The role of MAP kinase in TPA-mediated cell cycle growth factor increases the level of the cyclin-dependent kinase (CDK) in-
arrest of human breast cancer cells. Oncogene 16:131–139. hibitor p21/CIP1 (CDK-interacting protein 1) in A431 cells by increasing
23. Brognard, J., A. S. Clark, Y. Ni, and P. A. Dennis. 2001. Akt/protein kinase the half-lives of the p21/CIP1 transcript and the p21/CIP1 protein. Bio-
B is constitutively active in non-small cell lung cancer cells and promotes chem. J. 337:599–606.
cellular survival and resistance to chemotherapy and radiation. Cancer 32. Brazil, D. P., and B. A. Hemmings. 2001. Ten years of protein kinase B sig-
Res. 61:3986–3997. nalling: a hard Akt to follow. Trends Biochem. Sci 26:657–664.
24. Kohno, T., M. Takahashi, R. Manda, and J. Yokota. 1998. Inactivation of