| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
From the Department of Pathology,*
Ludwig Maximilians
University of Munich, 80337 München, and the Department of
Virology,
Max Planck Institute for
Biochemistry, 82152 Martinsried, Federal Republic of Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
One of the most frequently affected growth control mechanisms known to date in human malignancies is the p53 tumor suppression pathway. Nearly all of the different kinds of human malignancies analyzed thus far were shown to contain alterations in the p53 gene and the p53 regulating pathway in a considerable percentage of cases.5-9 The p53 status of angiosarcomas is only sparsely documented compared to that of other human malignancies. In carcinoma cells p53-dependent growth control is often impaired by missense mutations in one allele and the loss of the other, resulting in the accumulation of defective p53 protein. Alternatively, wild-type-p53 protein (wt-p53) can be functionally inactivated by binding to the cellular murine double minus-2 protein (MDM-2). This protein forms a tight complex with both mutant and wt-p53 and inhibits p53-mediated transactivation. Recent studies showed that the mdm-2 gene amplification has effects similar to p53 mutation. Thus, genetic alterations of either p53 or mdm-2 obviously represent alternative mechanisms for inactivating the same growth suppression pathway and overexpression of MDM-2 can lead to an escape from p53-regulated growth control as well.10,11 In fact, the human mdm-2 gene is frequently altered in human sarcomas and several studies demontrated that sarcomas that maintain wild-type p53 alleles overexpress MDM-2.12-16 To date, angiosarcomas have not been investigated for MDM-2 status. Therefore the aim of the present study was to investigate the role of p53 and MDM-2 in the pathogenesis of angiosarcomas. We present a collection of 19 angiosarcoma cases in comparison to 10 benign vascular control lesions. We evaluated p53 expression by immunohistochemistry, investigated the p53 gene structure in the area of the so-called mutation hot spots (exon 5 to 9) by sequence analysis, and confirmed the sequence data by a denaturing gradient gel electrophoresis assay (DGGE). MDM-2 was studied by immunohistochemistry. The study was completed by investigating the expression of vascular endothelium growth factor (VEGF) because a link between p53 activity and VEGF expression was recently shown.17,18
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The angiosarcoma study group (n = 19) was composed of 11 females and 8 males with an average age of 49 years (range, 25–88 years). Six angiosarcomas (cases 3, 4, 7, 8, 15, and 19) were located in the breast. Five angiosarcomas (cases 2, 5, 6, 10, and 18) were found in the soft tissue of arms and legs, one tumor (case 1) occurred in the head, and one (case 17) in the trunk. Three patients (cases 11, 12, and 13) showed a primary tumor location in the thyroid. One patient with angiosarcomatous infiltration of the lung (case 14) exhibited no other primary tumor localization. One case originated in the heart and great vessels (case 9); another arose in the sternum (case 16). The known facultative predisposing factors for the development of angiosarcoma in the cases studied were chronic lymphedema (cases 2 and 6) and postirradiation status (cases 10 and 13). Thyroid angiosarcomas (cases 11, 12, and 13) are known to show a predilection for mountainous regions of the world (such as the Bavarian alpine region) with iodine deficiency and development of long-standing nodular goiter.19 In none of the cases was there any indication of occupational exposure to thorotrast (thorium dioxide), arsenic solutions, or vinyl chloride, all of which may be associated with the development of angiosarcomas.20-22
For comparison and negative control, 10 specimens of benign vascular
lesions and normal skin were included in the study. The control group
(n = 10) consisted of five females and five
males with an average age of 31 years (range, 10 days to 69 years). For
detailed information regarding patients' age, sex, and tumor
localization, see Tables 1 and 2
.
|
|
Surgical specimens of angiosarcomas, benign vascular lesions, and
normal skin were collected at the Munich University Institute of
Pathology between 1983 and 1996. The material had been fixed in neutral
buffered formalin for 24 to 48 hours and processed routinely in a
low-melting-point paraffin wax (Paraplast, Vogel, Giessen, Germany).
Basic morphological diagnosis and classification of the cases were
carried out according to the criteria of Enzinger and
Weiss4
as well as to the AFIP and WHO
criteria corresponding to the affected organs
(breast,23,24
thyroid,19
lung,25
heart and great
vessels,26,27
and bone).28
Figure 1, A through D
shows the
basic morphology of these cases. Angiosarcomas of Kaposi's type were
not included in this study. A panel of vascular markers (antibodies to
FVIII (Dako, Copenhagen, Denmark), CD34 (Serotec, Oxford, UK), CD31
(Dako), and Ulex Europeus-Antigen (Dako)) were used for
immunohistochemical confirmation of the diagnosis. Other neoplasms that
may legitimately affect the differential diagnosis in a given case were
ruled out using additional antibodies (Cytokeratin-Antigen (Dako),
Smooth Mus-cle Aktin-Antigen (Boehringer, Mannheim, Germany),
Desmin-Antigen (Dako), and S100-Antigen (Dako)). From the panel
of vascular markers the CD 31 antigen, also termed platelet-endothelial
cell adhesion molecule-1, seemed to be the most sensitive and specific
antigen for endothelial differentiation. According to the literature,
all studied benign and malignant vascular tumors expressed this
membrane protein, whereas more than 100 soft-tissue tumors of
nonvascular origin did not.29
For this reason
cases which met the basic morphological criteria and clearly expressed
this sensitive vascular marker were accepted as angiosarcomas
and included in this study. Histopathological tumor grading was
carried out according to Coindre.30
|
Immunohistochemistry
Consecutive 3-µm sections were cut and mounted on sialinized slides (Superfrost Plus, Menzel-Gläser, Braunsischweig, Germany). Sections were dewaxed in xylene and rehydrated. For antigen retrieval, sections were immersed in 10 mmol/L citrate buffer, pH 6.0 (p53, MDM-2), or Target Retrival Solution (TRS, Dako S1700) (VEGF), heated three times in a microwave oven (800 W) for a total of 30 minutes and allowed to cool in the buffer for 20 minutes. Immunohistochemical studies were performed with a sensitive standard immunohistochemical streptavidin-biotin-peroxidase technique31,32 using a commercially available staining kit (Universal Dako LSAB-1 KIT/K 0681 Dako, Copenhagen, Denmark). Sections were immersed 2 x 5 minutes in Tris-buffered saline, pH 7.6, after which the endogenous peroxidase was blocked using 7.5% hydrogen peroxide for 10 minutes before the slides were rinsed and returned to Tris buffer for 2 x 5 minutes. Sections were incubated with Blocking Reagent (LSAB-1 KIT, Dako) for 10 minutes and then coated with primary monoclonal mouse antibody for 60 minutes. The DAKO-p53 antibody, DO-7, recognizes an epitope in the N terminus of the human p53 protein between amino acids 19 and 26. This p53 antibody was used in a final working concentration of 8 µg/ml. The MDM-2 antibody (IF2, Oncogene Research Products, Cambridge, MA), which recognizes an epitope in the amino terminal portion of the 491 AA human MDM-2 protein, was used in a final working concentration of 10 µg/ml. The purified polyclonal rabbit VEGF antibody (Dianova Calbiochem, PC 37), raised against a peptide from the N-terminus region of VEGF, was used in a final working concentration of 10 µg/ml. After incubation with the primary antibody the sections were washed for 2 x 5 minutes in Tris buffer. They were then coated with Link Antibody (LSAB-1 KIT, Dako) for 30 minutes, rinsed, immersed in Tris buffer for 2 x 5 minutes, and incubated with horseradish peroxidase-labeled streptavidin-biotin complex (LSAB, Dako) for 30 minutes. The sections were stained by 3-amino-9-ethylcarbarol (Sigma, Deisenhofen, Germany) for 15 minutes, rinsed, and counterstained with Mayer's hemalaun. To ensure specificity and for a control of background staining, controls were included in all staining runs. The primary antibody was replaced with bovine serum albumin in these controls and no immunohistochemical staining was observed.
Analysis and Quantification of the Immunohistochemical Results
Scoring was done according to studies of Remmele and coworkers.33 This method has been shown to be valid in routine morphology (estrogen and progesterone receptor analysis on breast cancer). The results were organized in six categories based on total percentage of cells staining positively (PP), as follows: Group 0 = negative; Group 1 = 1–20%; Group 2 = 21–40%; Group 3 = 41–60%; Group 4 = 61–80%; Group 5 = 81–100%. The results were further scored for staining intensity (SI) in 4 categories (groups 0–3): negative, light, intermediate, or intense staining.
The documentation of immunohistochemical staining was evaluated independently by two of the authors (CZ, MR). The interobserver variability was low. Questionable cases were reviewed simultaneously by the observers and a final score agreed by discussion.
The results clearly show several cases in the benign control group with
a definite negative immunoreaction as well as angiosarcomas with
intense staining, as documented in Figure 2
. Because we could refer to
negative and positive controls inside our study group there was no need
for further external positive controls. Benign control cases and
angiosarcomas were immunohistochemically investigated together and all
cases were prepared according to the same standardized method,
permitting consistent analysis of staining patterns and intensity.
|
The p53 immunohistochemical cutoff values are not well established and are the subject of controversial discussion. In some investigations, cases have been considered to be immunoreactive irrespective of the number of stained cells, whereas in other studies different cutoff values have been used.6 Most immunohistochemical studies on the p53 status of malignancies evaluated cases as p53-positive if they showed a nuclear immunoreaction in either more than 10% or more than 30% of the tumor cells. Against this background we chose as a cutoff value 40% of the cells with a staining intensity grade of at least intermediate, both in controls and sarcoma samples. Only cases above this cutoff value were considered to be p53-positive, ie, to show a definite p53 protein accumulation.
Cutoff Value for MDM-2 Positivity
Most studies using Southern blotting techniques demonstrated that MDM-2 protein overexpression was due to mdm-2 gene amplification. However, MDM-2 dysregulation has also been shown to be associated with transcriptional and/or translational deregulation.11 Therefore, proof of an increased MDM-2 protein amount by immunohistochemistry should be the best way to reflect all these different ways of dysregulation. No definite cutoff values for the evaluation of immunohistochemical MDM-2 positivity have been established. Only tumors with an intermediate or stronger nuclear MDM-2 immunoreaction in more than 40% of the tumor cells were graded MDM-2-positive. This cutoff point is validated by the MDM-2 staining results of the benign control group, in which the number of MDM-2-positive endothelial cells and the MDM-2 staining intensity were clearly lower.
p53 Gene Sequence Analysis
Histopathological examination of hematoxylin- and eosin-stained
sections of the tumors enabled us to select representative areas
consisting predominantly of tumor from which DNA was extracted after
microdissection. Ten 8-µm-thick paraffin sections were cut and placed
into an Eppendorf tube using sterile toothpicks. The microtome was
cleaned with xylene and the blade was exchanged for a new one before
trimming the next case. Sections were dewaxed in xylene, rehydrated in
graded ethanol, and genomic DNA was purified according to a standard
extraction method: tumor samples were digested with proteinase K (20
µg/ml; Sigma) in 200 µl digestion buffer (10 mmol/L Tris-HCl, pH
8.0; 50 mmol/L KCL; 1,5 mmol/L MgCL2; 0,5% Tween 20) at 55°C for 48
hours with constant shaking. DNA was extracted first with
phenol/chloroform and then with chloroform/isoamylethanol. Cycle
sequencing was performed according to standard protocols using
dye-labeled terminators (ABI Prism Dye Terminator Cycle Sequencing
Ready Reaction Kit/AmpliTaq DNA polymerase, Applied Biosystems GmbH,
Weiter-stadt, Germany) for performing enzymatic extension
reactions and sets of PCR primers according to published sequences of
exons 5–9 of the p53 gene (Table 3)
.
Automatic sequencing was performed on an ABI Prism 310 Genetic Analyzer
(Applied Biosystems).
|
For control and confirmation of the p53 gene sequence analysis a GC-clamped DGGE assay34,35 of the p53 exons 5–8 was done according to Beck36 (primers: exons 5–7;36 exon 837).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The DO-7 antibody recognizes an epitope in the N terminus of the human p53 protein (amino acids 19 to 26), which is expressed on wild-type as well as on mutant p53 protein.
In ten of 19 angiosarcomas (53%) nuclear staining of p53 was observed
in more than 40% of the total number of malignant endothelial cells.
Staining intensity was intermediate or high in these cases (Table 1
,
Figure 2, A and B
). In five sarcomas 30% of the tumor cells, and in
four cases approximately 10% of the cells were positive.
Cytoplasmic p53 staining was never observed. The grade of p53
positivity did not correlate with histological tumor grading, tumor
localization, or the patient's age (Table 1)
.
Among the benign control lesions, nuclear p53 positivity graded
intermediate in 50% of cells was observed in a hemangioma of infancy
(case C6). One granulation tissue (case C3) exhibited an intermediate
p53 staining reaction in 30% of the proliferating capillary
endothelial cells, whereas in all other cases only a few (cases C2, C4,
C5, C7) or no (cases C1, C8, C9, C10; Figure 2G
) p53-positive
endothelial cells were observed (Table 2)
.
Sequence Analysis of the p53 Gene
All angiosarcoma cases were investigated for p53 mutation in the hot spot regions of mutation (exons 5 to 9) using automatic DNA cycle sequencing. Cases 1 and 8 revealed missense mutations in the DNA-binding domain (codon 102 to 292) of the p53 gene.
Transversion was found in exon 6, codon 197 (GTG to GGG, case 1) and in exon 8, codon 274 (GTT to CTT, case 8). The observed mutations were tumor-specific because they were absent in DNA from normal cells of the patients.
The remaining 17 angiosarcoma cases exhibited p53 wild-type sequence.
DGGE Assay
The GC-clamped DGGE assay also revealed p53 mutation in cases 1 and 8. With this sensitive method no further p53 mutations were found in the angiosarcoma group. The DGGE assay clearly confirmed the data of the p53 sequence analysis.
Detection of MDM-2 Protein
In 13 of 19 angiosarcomas (68%) the number of MDM-2-positive
tumor cells exceeded 40% of the total number of malignant cells. This
grading was based on cells with clear nuclear intermediate or intense
staining (Figure 2C)
.
In four cases MDM-2 was positive in 30%, and in two cases in
10% of the malignant endothelial cells (Table 1)
. In addition, a
cytoplasmic MDM-2 immunoreaction was observed in five cases (cases 2,
5, 7, 11, and 16). In these cases nearly all tumor cells exhibited
cytoplasmic staining (Figure 2D)
.
Again, no correlation between MDM-2 staining and histological tumor
grading, tumor localization, or the patient's age was observed (Table 1)
.
The benign control lesions revealed clearly lower levels of MDM-2
protein (Table 2)
. Only in a capillary hemangioma (case C5) were 50%
of the cells MDM-2-positive, but in this case staining intensity was
very low. A hypertrophic granulation tissue (case C4) exhibited MDM-2
staining intensity graded intermediate in 30% of the proliferating
endothelial cells. Either very faint MDM-2 immunoreaction in a
few cells or no MDM-2 staining was observed in all other control cases.
A cytoplasmic MDM-2 immunoreaction was not observed in the control
group.
Detection of VEGF Protein
In 15 of 19 investigated angiosarcomas (79%) nearly all tumor
cells (81–100%) revealed a uniform VEGF positivity. Staining
intensity was either high (in 8 cases) or intermediate (in 7 cases)
(Table 1
, Figure 2, E and F
). In three angiosarcomas 70% (case 13),
50% (case 14), and 30% (case 4) of the tumor cells were found
to be positive for VEGF protein. One sarcoma (case 15) showed no VEGF
staining. In these experiments no correlation between VEGF staining
and, histological tumor grading, tumor localization, or the patient's
age was observed (Table 1)
.
In the benign control group, high VEGF expression similar to that seen
in the angiosarcomas was observed in a juvenile hemangioma (case C6).
All other cases exhibited clearly lower numbers of positive cells and
less intense staining (Table 2)
.
Correlation between the Results
In 14 of 19 angiosarcoma cases comparably high numbers of both p53- and MDM-2-positive tumor cells were observed. In two tumors (cases 8 and 16) the number of p53-positive cells was twofold higher than that of MDM-2-positive cells, whereas in three angiosarcomas (cases 13, 14, and 19) the number of p53-positive tumor cells was clearly lower compared to the number of MDM-2-positive cells. All sarcomas with more than 40% p53-positive and/or MDM-2-positive cells were also clearly positive for VEGF. In three sarcomas (cases 5, 6, and 7) fewer cells expressed p53/MDM-2 than VEGF. Two angiosarcomas of the breast (cases 4 and 15) did not show increased p53 value or significant staining reaction for MDM-2 and VEGF. Both sarcomas with p53 mutation (cases 1 and 8) showed the highest levels of p53 and VEGF immunoreaction.
Overall, we demonstrated that in 74% (14/19) of the angiosarcomas investigated in this study the expression of p53, MDM-2, and VEGF is clearly increased in comparison to control tissues.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Based on studies of cultured human fibroblasts from cancer-prone
Li-Fraumeni patients, Bouck and Dameron recently suggested a new
function of wt-p53 in the regulation of
angiogenesis.39,40
They found that wt-p53 is a
transcriptional activator of thrombospondin-1 (TSP-1) expression. TSP-1
is present in normal resting endothelial cells and absent in actively
forming endothelial cell sprouts and has been shown to be a potent
inhibitor of endothelial cell migration and mitogenesis. In the
presence of angiogenic stimuli TSP-1 maintains the differentiated,
quiescent phenotype of endothelial cells and inhibits their conversion
to a migratory, invasive phenotype. Induction of TSP-1 expression in
transformed endothelial cells restores a normal phenotype and
suppresses the tumorigenicity of the cells.41-45
Wt-p53 not only activates the expression of the important
antiangiogenic TSP-1 but also down-regulates the promoter activity of
the strongly angiogenic VEGF in a dose-dependent
manner.17
Therefore, the observed deregulation of
p53/MDM-2 expression may affect differentiation and phenotype of
endothelial cells by modulation of the TSP-1/VEGF balance. The high
expression of VEGF in our angiosarcomas and the clearly lower degree of
p53, MDM-2, and VEGF immunoreaction in the benign control group
supports this hypothesis (Figure 3)
.
|
Given the present state of knowledge about the pathogenesis of angiosarcomas, a correlative analysis aimed at prognostic or diagnostic factors was not a primary goal of our study. Investigation of very rare entities often cannot achieve reliable and confident statistical results because of small case numbers. Our angiosarcoma collection of 19 cases is one of the largest studied. However, an analysis of survival times correlated to our immunohistochemical results and matched with parameters like age and sex would need to be based on significantly higher case numbers to yield results with a high degree of confidence.
Taken together, the results of the present study show that more than two-thirds of the angiosarcomas analyzed exhibited a dysfunction of the p53/MDM-2 pathway, although there were several differences in tumor localization, histological tumor grading, age, and promotion events. A dysfunction of the p53/MDM-2 pathway may influence the development of angiosarcomas not just by effects on growth and apoptosis control but also by up-regulation of VEGF. Additional oncogenes and tumor suppressor genes certainly play a role in the genesis of malignant endothelial transformation, but our data suggest that functional impairment of the p53/MDM-2 pathway may be a key to the initiation and/or progression of a high percentage of angiosarcomas.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
This study contains parts of MR's doctoral thesis.
Accepted for publication August 6, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  J.-C. M. Walkes, W. R. Smythe, and M. J. Reardon Cardiac Neoplasms Card. Surg. Adult, January 1, 2008; 3(2008): 1479 - 1510. [Full Text]
|
|
|
|
 |
|
 F Noack, L Balleisen, P Valent, and H-P Horny Angiosarcoma of bone marrow with unusual expression of chymase: diagnosis in a trephine biopsy specimen J. Clin. Pathol., October 1, 2007; 60(10): 1183 - 1183. [Full Text] [PDF]
|
|
|
|
 |
|
 G. Zhou, Y. Zhai, Y. Cui, X. Zhang, X. Dong, H. Yang, Y. He, K. Yao, H. Zhang, L. Zhi, et al. MDM2 Promoter SNP309 Is Associated with Risk of Occurrence and Advanced Lymph Node Metastasis of Nasopharyngeal Carcinoma in Chinese Population Clin. Cancer Res., May 1, 2007; 13(9): 2627 - 2633. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Perry, J. Banyard, E. R. McLaughlin, R. Watnick, A. Sohn, D. N. Brindley, T. Obata, L. C. Cantley, C. Cohen, and J. L. Arbiser AKT1 Overexpression in Endothelial Cells Leads to the Development of Cutaneous Vascular Malformations In Vivo Arch Dermatol, April 1, 2007; 143(4): 504 - 506. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. LaRusch, M. W. Jackson, J. D. Dunbar, R. S. Warren, D. B. Donner, and L. D. Mayo Nutlin3 Blocks Vascular Endothelial Growth Factor Induction by Preventing the Interaction between Hypoxia Inducible Factor 1{alpha} and Hdm2 Cancer Res., January 15, 2007; 67(2): 450 - 454. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-Y. Yang, C. S. Zong, W. Xia, Y. Wei, M. Ali-Seyed, Z. Li, K. Broglio, D. A. Berry, and M.-C. Hung MDM2 Promotes Cell Motility and Invasiveness by Regulating E-Cadherin Degradation. Mol. Cell. Biol., October 1, 2006; 26(19): 7269 - 7282. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Kaluzova, S. Kaluz, M. I. Lerman, and E. J. Stanbridge DNA Damage Is a Prerequisite for p53-Mediated Proteasomal Degradation of HIF-1{alpha} in Hypoxic Cells and Downregulation of the Hypoxia Marker Carbonic Anhydrase IX Mol. Cell. Biol., July 1, 2004; 24(13): 5757 - 5766. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Reardon and W. R. Smythe Cardiac Neoplasms Card. Surg. Adult, January 1, 2003; 2(2003): 1373 - 1400. [Full Text]
|
|
|
|
|
|
M. G. Backlund, S. L. Trasti, D. C. Backlund, V. L. Cressman, V. Godfrey, and B. H. Koller Impact of Ionizing Radiation and Genetic Background on Mammary Tumorigenesis in p53-deficient Mice Cancer Res., September 1, 2001; 61(17): 6577 - 6582. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Zietz, U. Rumpler, M. Sturzl, and U. Lohrs Inverse Relation of Fas-Ligand and Tumor-Infiltrating Lymphocytes in Angiosarcoma : Indications of Apoptotic Tumor Counterattack Am. J. Pathol., September 1, 2001; 159(3): 963 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. K. Linderholm, T. Lindahl, L. Holmberg, S. Klaar, J. Lennerstrand, R. Henriksson, and J. Bergh The Expression of Vascular Endothelial Growth Factor Correlates with Mutant p53 and Poor Prognosis in Human Breast Cancer Cancer Res., March 1, 2001; 61(5): 2256 - 2260. [Abstract] [Full Text]
|
|
|
|
|
|
L. Zhang, D. Yu, M. Hu, S. Xiong, A. Lang, L. M. Ellis, and R. E. Pollock Wild-Type p53 Suppresses Angiogenesis in Human Leiomyosarcoma and Synovial Sarcoma by Transcriptional Suppression of Vascular Endothelial Growth Factor Expression Cancer Res., July 1, 2000; 60(13): 3655 - 3661. [Abstract] [Full Text]
|
|
|
|
|
|
J. L. Arbiser, H. Larsson, L. Claesson-Welsh, X. Bai, K. LaMontagne, S. W. Weiss, S. Soker, E. Flynn, and L. F. Brown Overexpression of VEGF 121 in Immortalized Endothelial Cells Causes Conversion to Slowly Growing Angiosarcoma and High Level Expression of the VEGF Receptors VEGFR-1 and VEGFR-2 in Vivo Am. J. Pathol., April 1, 2000; 156(4): 1469 - 1476. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
