| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Pediatric Endocrinology, Children’s Hospital (M.W.E., M.B.R., B.S.S.), Institute of Brain Research (M.H.D., R.M.), and Department of Neurosurgery (F.D., E.H.G.), University of Tuebingen, D-72076 Tuebingen, Germany
Address all correspondence and requests for reprints to: Dr. Martin W. Elmlinger, Section of Pediatric Endocrinology, Children’s Hospital, University of Tuebingen, Hoppe-Seyler Strasse 1, D-72076 Tuebingen, Germany. E-mail: martin.elmlinger{at}med.uni-tuebingen.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The results suggest cell type-specific accumulation of IGFBP-1, -2, and -3 in human glial tumors of the brain. The increase in IGFBP-2 expression with this malignancy suggests a role of IGFBP-2 in the biology of human gliomas.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In addition, alterations in the expression of IGF-binding proteins (IGFBPs), which modulate the biological actions of the IGFs (7), were found in brain tumors. Elevated concentrations of IGFBP-2 are detected frequently in the cerebrospinal fluid of patients with brain tumors. The most elevated levels of IGFBP-2 were measured in cerebrospinal fluid of highly malignant tumors (8). Furthermore, high levels of IGFBP-1, -2, and -3 were detected in the cyst fluid of a patient with a hypothalamic astrocytoma (9) and in a diverse range of gliomas (10). In vitro experiments revealed that the overexpression of IGFBP-2 in C6 glioma cells resulted in reduced growth potential of these cells (11). Taken together, these data emphasize an important role of IGFBP-2 in the biology of brain tumors.
Therefore, we studied the relationship between clinical features, i.e. malignancy, and the expression of IGFBP-2 in a wide range of human astrocytomas and in the rat C6 glioblastoma model (12). In particular, we investigated the accumulation of IGFBP-1, -2, and -3 in 22 human glioblastoma multiforme, 9 anaplastic astrocytomas, 1 gemistocytic astrocytoma, 5 protoplasmic astrocytomas, and 13 fibrillary astrocytomas by immunohistochemistry. In advance, all tumors were classified histopathologically according to the WHO Malignancy Scale (13). Double labeling experiments with antibodies directed against CD68 (monocytes/macrophages), glial fibrillary acidic protein (GFAP), CD31 (endothelial cells), CD3 (T cells), and human leukocyte antigen (HLA)-DR, -DP, and -DQ (major histocompatibility complex class II) were performed to characterize the cell type of IGFBP+ cells and to study the cell specificity of IGFBP expression. As controls, the accumulation of IGFBP-2 was investigated in normal brain and rat C6 glioblastoma model. To determine the expression of IGFBP-1, -2, and -3 messenger RNA (mRNA) in the gliomas, semiquantitative RT-PCR analysis was performed from total RNA of 9 gliomas of different grades and in 8 cell lines representing the cellular composition of human glioma. The accumulation of IGFBP-1, IGFBP-2, and IGFBP-3 was evaluated quantitatively by determination of labeling scores, which were compared with the grade of malignancy of each tumor.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
): 22 glioblastoma multiforme, 9
anaplastic astrocytomas, 5 protoplasmic astrocytomas, 1 gemistocytic
astrocytoma, and 13 fibrillary astrocytomas. All tumors were resected
at the Department of Neurosurgery (Tuebingen, Germany) or at the
Department of Neurosurgery of the Schildautalklinik (Seesen, Germany)
and characterized with respect to their clinical features as previously
described (14). Frontal cortex specimens of six
neuropathologically unaltered control brains were obtained from
autopsies at the Department of Pathology in Tuebingen (Table 2) and have been described previously
(14).
|
|
Immunohistochemistry
Before immunohistochemistry, the slices were irradiated in a
microwave oven five times for 5 min each time in citrate buffer (2.1 g
sodium citrate/liter, pH 6.0) and incubated in nonspecific porcine
serum 1:10 in 90% Tris-balanced salt solution containing 0.025
M Tris and 0.15 M NaCl, pH 7.5 (TBS), for 15
min. The primary polyclonal rabbit antihuman IGFBP antibodies were all
applied at a dilution of 1:2000 in TBS-BSA (10% BSA in TBS) to the
sections overnight at 4 C. Antibody binding was detected by incubating
the slices with a biotinylated swine antirabbit IgG
F(ab)2 antibody (DAKO Corp.,
Hamburg, Germany) and consecutively with an AB complex (DAKO Corp.), both at a dilution of 1:400 in TBS for 30 min. The
reaction was visualized with diaminobenzidine (Fluka, Buchs,
Switzerland) as a substrate, and slices were consecutively
counterstained with hematoxylin.
Double labeling experiments
In double labeling experiments, we first labeled the
characterizing antigen. Briefly, slices were deparaffinized, irradiated
in a microwave oven for antigenic retrieval, and incubated with
nonspecific porcine serum as described above. Then the differentiating
monoclonal mouse antibodies directed against human GFAP
(Roche, Mannheim, Germany), CD68 (DAKO Corp.), CD3 (DAKO Corp.), and CD31 (DAKO Corp.) were added to the slices all at a dilution of 1:100 in
TBS-BSA. The functional status of the cells was analyzed by using
anti-HLA-DR, -DP, -DQ, -DR, -DP, and -DQ (DAKO Corp.;
major histocompatibility class II). Visualization was achieved by
adding rabbit antimouse IgG (DAKO Corp.) diluted at 1:20
in TBS for 30 min and APAAP complex (DAKO Corp.)
diluted 1:80 in TBS for 30 min. Consecutively, we developed with Fast
Blue BB salt chromogen-substrate solution, yielding a blue reaction
product. To avoid antibody cross-reactivity in double labeling
experiments, slices were once more irradiated in a microwave for 20 min
in citrate buffer (15). Complete inhibition of alkaline
phosphatase function was achieved as previously described
(16). Then, IGFBP-1, -2, and -3 were immunolabeled as
described above, using polyclonal antibodies to the respective human
antigens (Mediagnost, Tuebingen, Germany).
Controls
Negative antibody controls included nonspecific isotype-matched
primary antibodies. In each labeling experiment serial sections were
incubated with TBS/BSA instead of the primary antibody to assess
secondary antibody-mediated mislabeling. No labeling of the secondary
antibody was observed in any labeling experiments. After overnight
preabsorption of anti-IGFBP-1 antibody with 0.5 µg/ml of the IGFBP-1
protein purified from human amniotic fluid, the initially observed
labeling pattern (Fig. 1A) was completely
abrogated (not shown). After overnight incubation of anti-IGFBP-2 with
0.5 µg/ml recombinant IGFBP-2 protein (gift from Novartis, Basel,
Switzerland), the initially observed labeling pattern (Fig. 1B) was
completely abrogated. The same effect of preabsorption with IGFBP-2 was
detectable in the rat C6 glioblastoma model (Figs. 1D and 2, A and B).
|
|
Evaluation
Staining results were examined at x200 magnification using an
eye-piece grid. Three independent areas of the tumor tissues and of
areas of infiltrative tumor growth were evaluated for each section and
labeling procedure. Areas of zonal necrosis were not taken into
consideration. Endothelial IGFBP-1 and -3 immunolabeling was evaluated
by counting positively labeled endothelial cells with respect to the
total number of counterstained endothelial cells. IGFBP-2 labeling was
evaluated by counting positively stained cells with respect to the
overall number of counterstained nuclei. Positively stained cells were
counted and assigned a semiquantitative labeling score. The mean
labeling score (MLS) of the three evaluated areas was determined as
follows: 0 = no staining, 1 = up to 2% labeled cells, 2
= 3% up to 20% labeled cells, 3 = 21% up to 50% labeled cells,
and 4 = more than 51% labeled cells. Statistical analysis of
pooled low grade glioma MLS (WHO I and II) vs. pooled high
grade glioma MLS (WHO III and IV) was performed using the Mann-Whitney
U test.
Cell lines
A set of eight commercially available (American Type Culture Collection, Manassas, VA) human cell lines, representing
the cellular composition of human gliomas, was used to study the mRNA
expression of IGFBP-1, -2, and -3. These included four glioma cell
lines (LN-229, T98G, U373NG, and U138MG) (17), the lung
fibroblast cell line CRL246, the histiocytic lymphoma cell line U937
with monocyte/macrophage morphology, and the myeloblastic AML cell line
CCL246 (clone K61).
C6 rat glioblastoma cell culture and transplantation
The rat C6 glioblastoma cell line was obtained from the
American Type Culture Collection and raised in RPMI 1640
medium with Glutamax II (Life Technologies, Inc., Paisley,
UK) containing 10% FCS (Life Technologies, Inc.) and
1.2% penicillin/streptomycin (Fluka, Buchs, Switzerland) at 37 C in
5% CO2. Cells were implanted intracranially as
previously described (12). Briefly, cells were harvested,
and 5 µl cell suspension were injected into the basal ganglia region
of Sprague Dawley rats at a concentration of 4 x
105/µl. After 2 weeks, rats were killed and
perfused with 4% buffered formaldehyde, and tumors were removed
for further analysis. Immunohistochemistry was performed using a rabbit
polyclonal antiserum to rodent IGFBP-2 (from Austral, San Ramon,
CA).
Semiquantitative RT-PCR
Semiquantitative RT-PCR was performed to assess the levels of
IGFBP-1, IGFBP-2, IGFBP-3, IGF-I, and IGF-II mRNA (results for IGFs
not shown) in tissue samples from nine representative glioma tissues of
different grades and in six glioma cell lines. RT-PCR was performed
according to a protocol described previously (18). The
gliomas had been characterized with respect to their grade of
malignancy (13). The cell lines were described above. PCR
products stained with ethidium bromide, were analyzed by an imaging
system and quantitated by a specific software (Aida, Raytest,
Straubenhardt, Germany). Glyceraldehyde phosphate dehydrogenase (GAPDH)
served as an internal standard of mRNA expression. The sequences of
IGFBP-1-specific PCR primers were GAGAGCACGGAGATAA CTGAGG for the sense
strand and TTGGTGACATGGAGAGCCTTCG for the antisense strand; the size of
the amplicon was 131 bp. The primers for IGFBP-2 PCR have been
published (18); the amplicon was 121 bp. Primers for
IGFBP-3 PCR were TAGTGAGTCGGAGGAAGACC (sense) and
GAGAAGTTCTGGGTATCTGTGC (antisense); the amplicon had a size of 192 bp.
Northern blot analysis (not shown) was performed to confirm the
presence of IGFBP-2 mRNA in the total RNA from glioma tissue.
Statistical methods
Data were analyzed using standard statistical methods, including
Mann-Whitney U test. Significance was assigned a value of
P < 0.05 for differences between two sets of data.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). IGFBP-1 immunoreactivity was
observed predominantly in endothelial (Fig. 1A) and
macrophage/microglial cells. In neuropathologically unaltered control
brains, IGFBP-1, -2, and -3 were not expressed (Table 2). Occasionally,
faint IGFBP-1, -2, and -3 immunoreactivity was observed in singular
endothelial cells and neurons. In one patient, however, we observed
prominent IGFBP-2 immunoreactivity in the plexus choroideus endothelial
cells. Furthermore, no staining of IGFBPs was achieved, when antiserums
were preabsorbed with the respective IGFBP, as a control. The effect of
preabsorption with IGFBP-2 on IGFBP-2 immunostaining in rat C6
glioblastoma is shown in Fig. 2. As in the slices from human tumors,
positive staining of IGFBP was completely abrogated by preabsorption of
the antiserum with the respective antigen. Using the Mann Whitney U test, we detected significantly (P = 0.023) more IGFBP-1+ endothelial cells in low grade gliomas (MLS = 2.105; SEM = 0.1857) than in high grade gliomas (MLS = 1.387; SEM = 0.1715). There were no significant differences (P = 0.586) in the number of IGFBP-1-immunoreactive macrophages/microglial cells in low (MLS = 1.0; SEM = 0.076) vs. high (MLS = 1.097; SEM = 0.097) grade gliomas.
IGFBP-2 immunoreactivity was detected in glioma,
macrophages/microglial, and scattered endothelial cells. It is of note
that IGFBP-2+ cells frequently accumulated in the
immediate vicinity of areas of focal necrosis (Fig. 1B). Accordingly,
we calculated a positive correlation of IGFBP-2+
cells with the grade of malignancy of the tumors. The number of
IGFBP-2+ cells was significantly
(P = 0.0006) lower in low grade gliomas (MLS =
1.0; SEM = 0.171) than in high grade gliomas
(MLS = 2.129; SEM = 0.201). Double labeling
experiments confirmed the coexpression of IGFBP-2 in
GFAP+ (Fig. 1E), in CD68+
(Fig. 1F) and in HLA-DR, -DP, and -DQ+ cells.
Accumulation of IGFBP-2+ cells in and adjacent to
areas of necrosis was confirmed in the rat C6 glioma model (Fig.
1D).
IGFBP-3 immunoreactivity was distributed similarly to IGFBP-1
immunoreactivity, predominantly in endothelial cells (Fig. 1C) and in
scattered macrophages/microglial cells. However, no significant
differences (P = 0.080) were calculated in the amount
of IGFBP-3+ endothelial cells in low (MLS =
1.789; SEM = 0.224) vs. high (MLS
= 1.258; SEM = 0.202) grade gliomas. There were
no significant (P = 0.82) differences in the number of
IGFBP-3+ macrophages in low grade gliomas
(MLS = 1.053; SEM = 0.093) compared with
high grade gliomas (MLS = 1.097; SEM =
0.054).
Expression of mRNA of the respective IGFBPs was confirmed in a
representative set of gliomas by RT-PCR and Northern blot (not shown)
using total RNA from nine tumors of different grades (grades 1–4; Fig. 3). Of the three IGFBPs, IGFBP-1 reached
higher mRNA levels relative to GAPDH mRNA in the gliomas, than IGFBP-2
and -3. Malignancies of higher grade (grades 3 and 4) tended to express
higher IGFBP-2 mRNA levels than tumors of lower grade (grades 1 and
2). No differences in IGFBP-1 and IGFBP-3 mRNA expression were seen
between tumors of high and low grades of malignancy (Fig. 3).
|
). The
mRNA of IGFBP-1 and IGFBP-3, of which significant protein expression
was shown histochemically only in the endothelial cells, were weakly
expressed in all of the cell lines.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
IGFBP-1 expression in endothelial cells has been described previously, but no correlation with malignancy has been reported to date (19, 20). The correlation of IGFBP-1 expression with malignancy in glial tumors of the brain is of note, because IGFBP-1 inhibits IGF-I mediated proliferation of a wide range of cell types, including breast cancer, mouse whole brain, arterial smooth muscle, and prostate cancer (21, 22, 23, 24). Moreover, IGFBP-1 has been demonstrated to affect apoptosis and cell adhesion in breast cancer cells by signal transduction through integrins (25). Increased expression of IGFBP-1 in low vs. high grade glioma therefore suggests the inactivation of a potential mechanism of antiangiogenesis in high grade human astrocytomas. Accordingly, high expression of IGFBP-1 mRNA was detected here in the gliomas, but not in the cell lines, as no endothelial cell line was included.
IGFBP-2 expression has been demonstrated to be involved in the development of the brain (26, 27). Furthermore, differential induction of IGFBP-2 expression in the brain has been associated with a variety of pathological conditions, including hypoxia, regeneration, and trauma (28, 29, 30). In meningeomas, plexus papillomas, and glioblastomas, a high IGF-II/IGFBP-2 mRNA ratio has been suggested to constitute a sign of biologically aggressive behavior that can influence treatment strategies (31, 32, 33). In brain tumors, IGFBP-2 accumulation has been localized to astrocytes and choroid plexus.
Our results provide further information on the cellular distribution of IGFBP-2-expressing cells in human gliomas. The immunohistochemistry of the tumors in vivo is in accordance with analysis of IGFBP-2 mRNA expression in the tumor cell lines, i.e. high expression in glioma and macrophage-like cell lines and low expression in the fibroblast cell line. As the IGFBP-2+ cells accumulated in the immediate vicinity of focal necrosis of the tumor, high expression of IGFBP-2 may indicate the presence of a highly malignant glioma. This is in accordance with the earlier finding of elevated concentrations of IGFBP-2 in the cerebrospinal fluid of patients with malignant brain tumors (8).
Furthermore, the present data add considerable knowledge about the involvement of IGFBP-2 in human glioma pathophysiology. In addition to the influence of IGFBP-2 in the IGF-dependent proliferation of glioma cells (11), IGF-independent effects of IGFBP-2 in the regulation of apoptosis and cell adhesion, i.e. metastasis, are likely in brain tumors. In future studies it has to be investigated whether IGFBP-2 affects tumorigenesis through RGD-specific binding to integrins, as it was seen for IGFBP-1 in breast cancer (25).
IGFBP-3 expression in endothelial cells is cytokine sensible and can be
triggered by IGF-I and inhibited by transforming growth factor-ß
(34). Most importantly, proapoptotic effects of IGFBP-3
have been described in glioblastoma multiforme (35). In
addition, IGFBP-3 obviously plays a role in the modulation of breast
cancer cell proliferation by tumor necrosis factor-
(36). However, clinical studies revealed that high levels
of IGFBP-3 are associated with unfavorable prognostic features of
breast cancer (37). Uniform and restricted cellular
IGFBP-3 expression in astrocytomas of all malignancies therefore
defines this peptide as potential antineoplastic agent in these
neoplasias.
In conclusion, the association of elevated IGFBP-2 protein and mRNA with the malignancy and the cellular distribution indicates a possible role for IGFBP-2 in the processes of malignant transformation, tumor necrosis, and metastasis of brain tumors. The mechanism by which IGFBP-2 acts at the cellular level, however, remains to be elucidated. Furthermore, the measurement of IGFBP-2 in the serum and liquor by immunoassay may be of value in the diagnosis of a malignant brain tumor.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received September 5, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
antisense oligonucleotide. Mol Pharmacol 55:396–402 on MCF-7 breast cancer cells is
associated with increased insulin-like growth factor binding protein-3
accumulation. Int J Oncol 13:865–869[Medline]
This article has been cited by other articles:
 |
|
 |
|
  E. K. Rowinsky, H. Youssoufian, J. R. Tonra, P. Solomon, D. Burtrum, and D. L. Ludwig IMC-A12, a Human IgG1 Monoclonal Antibody to the Insulin-Like Growth Factor I Receptor Clin. Cancer Res., September 15, 2007; 13(18): 5549s - 5555s. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Fukushima, T. Tezuka, T. Shimomura, S. Nakano, and H. Kataoka Silencing of Insulin-like Growth Factor-binding Protein-2 in Human Glioblastoma Cells Reduces Both Invasiveness and Expression of Progression-associated Gene CD24 J. Biol. Chem., June 22, 2007; 282(25): 18634 - 18644. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. W Frommer, K. Reichenmiller, B. S Schutt, A. Hoeflich, M. B Ranke, G. Dodt, and M. W Elmlinger IGF-independent effects of IGFBP-2 on the human breast cancer cell line Hs578T. J. Mol. Endocrinol., August 1, 2006; 37(1): 13 - 23. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. K. Wang, L. Hu, G. N. Fuller, and W. Zhang An Interaction between Insulin-like Growth Factor-binding Protein 2 (IGFBP2) and Integrin {alpha}5 Is Essential for IGFBP2-induced Cell Mobility J. Biol. Chem., May 19, 2006; 281(20): 14085 - 14091. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Bredel, C. Bredel, D. Juric, G. R. Harsh, H. Vogel, L. D. Recht, and B. I. Sikic Functional Network Analysis Reveals Extended Gliomagenesis Pathway Maps and Three Novel MYC-Interacting Genes in Human Gliomas Cancer Res., October 1, 2005; 65(19): 8679 - 8689. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. C. Russo, B. S. Schutt, E. Andaloro, S. I. Ymer, A. Hoeflich, M. B. Ranke, L. A. Bach, and G. A. Werther Insulin-Like Growth Factor Binding Protein-2 Binding to Extracellular Matrix Plays a Critical Role in Neuroblastoma Cell Proliferation, Migration, and Invasion Endocrinology, October 1, 2005; 146(10): 4445 - 4455. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Vorwerk, K. Mohnike, H. Wex, F.-W. Rohl, M. Zimmermann, W. F. Blum, and U. Mittler Insulin-Like Growth Factor Binding Protein-2 at Diagnosis of Childhood Acute Lymphoblastic Leukemia and the Prediction of Relapse Risk J. Clin. Endocrinol. Metab., May 1, 2005; 90(5): 3022 - 3027. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Godard, G. Getz, M. Delorenzi, P. Farmer, H. Kobayashi, I. Desbaillets, M. Nozaki, A.-C. Diserens, M.-F. Hamou, P.-Y. Dietrich, et al. Classification of Human Astrocytic Gliomas on the Basis of Gene Expression: A Correlated Group of Genes with Angiogenic Activity Emerges As a Strong Predictor of Subtypes Cancer Res., October 15, 2003; 63(20): 6613 - 6625. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Wang, H. Wang, W. Shen, H. Huang, L. Hu, L. Ramdas, Y.-H. Zhou, W. S-L. Liao, G. N. Fuller, and W. Zhang Insulin-like Growth Factor Binding Protein 2 Enhances Glioblastoma Invasion by Activating Invasion-enhancing Genes Cancer Res., August 1, 2003; 63(15): 4315 - 4321. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Solomon, M. Grueterich, D.-Q. Li, D. Meller, S.-B. Lee, and S. C. G. Tseng Overexpression of Insulin-like Growth Factor-Binding Protein-2 in Pterygium Body Fibroblasts Invest. Ophthalmol. Vis. Sci., February 1, 2003; 44(2): 573 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Hoeflich, R. Reisinger, H. Lahm, W. Kiess, W. F. Blum, H. J. Kolb, M. M. Weber, and E. Wolf Insulin-like Growth Factor-binding Protein 2 in Tumorigenesis: Protector or Promoter? Cancer Res., December 1, 2001; 61(24): 8601 - 8610. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
