| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Molecular Oncology (M.S., N.F.) and Pathology (F.V.P., G.F.), Genentech, Inc., South San Francisco, California 94080; Institut National de la Santé et de la Recherche Médicale, Unité 435 (M.S.), and Centre Hospitalier Universitaire Ponchaillou Hospital (N.R.-L.), Rennes, France 35042; and Department of Growth and Reproduction, Copenhagen University Hospital (E.R.-D.M.), Copenhagen, Denmark 2100
Address all correspondence and requests for reprints to: Dr. Napoleone Ferrara, Genentech, Inc., 1 DNA Way, South San Francisco, California 94080. E-mail: nf{at}gene.com.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Whereas broad spectrum angiogenic factors such as VEGF are known to play a critical role in tumorigenesis, the discovery of tissue-specific angiogenic factors may allow the development of angiogenesis inhibitors that will be specific for some types of tumors (12). This new concept is particularly relevant for testicular germ cell tumors. The incidence of these tumors has increased over the last decades in many occidental countries to become the most frequent malignancy in 25- to 35-yr-old men (13, 14, 15), and therefore, more research efforts are warranted.
Angiogenesis mechanisms may play an important role in testicular carcinogenesis, although fast growing undifferentiated germ cell-derived tumors, such as embryonal carcinoma, are frequently necrotic because of the insufficient growth of capillary vessels. VEGF is expressed by several different testicular cancer types, especially teratomas (5). To date, it is unknown whether EG-VEGF is expressed in testicular somatic or germ cell tumors and whether it plays a role in supporting angiogenesis and tumor growth.
In the present study we analyzed the expression of EG-VEGF mRNA and protein in normal fetal and adult testes and in a series of 48 testicular neoplasms derived from different cell types, including the preinvasive carcinoma in situ (CIS). For this purpose we used a panel of monoclonal antibodies against recombinant human EG-VEGF protein, developed and validated in our laboratory. We report for the first time that the expression of EG-VEGF protein in human testis is restricted to Leydig cells and Leydig-cell derived tumors. The pattern of expression is consistent with a role for EG-VEGF in promoting the interstitial angiogenesis to support the endocrine function of Leydig cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Normal testis specimens were obtained from patients who underwent orchidectomy to treat prostate tumors at the Centre Hospitalier Universitaire Ponchaillou Hospital (Rennes, France). The clinical status of the patients revealed no reproductive abnormalities or testicular infections. Forty-eight specimens of testicular tumors were obtained from patients with testicular cancer who underwent surgery at one of hospitals in France (Centre Hospitalier Universitaire Ponchaillou Hospital) or Denmark (Copenhagen University Hospital). The series included 21 seminoma specimens, 15 nonseminomatous germ cell tumor (NSGCT) specimens with various histological components, seven CIS specimens, four Leydig cell tumors, and one Sertoli cell tumor. Sections of fixed gonads from 12- to 40-wk-old fetuses were obtained from the Pathology Department of Ponchaillou Hospital. These tissue specimens were isolated after induced or spontaneous abortions when routine autopsies found no signs of disease or developmental disorders. Local ethics committees approved the use of these samples.
Real-time quantitative PCR
For real-time quantitative RT-PCR, 100 ng total RNA were assayed in triplicate with TaqMan kit reagents (PerkinElmer, Wellesley, MA) and an ABI PRISM 7700 Sequence Detector (PerkinElmer). Oligonucleotides and probes used were as follows: human EG-VEGF forward, 5'-CCGGCAGCCACAAGGTC; human EG-VEGF reverse, 5'-TGGGCAAGCAAGGACAGG; human EG-VEGF probe, 6-carboxyfluorescein (FAM)-5'-CCTTCTTCAGGAAACGCAAGCACCAC-3'-tetramethylrhodamine (TAMRA); human glyceraldehyde-3-phosphate dehydrogenase (GAPDH) forward, 5'-TGGGCTACACTGAGCACCAG; human GAPDH reverse 5'-CAGCGTCAAAGGTGGAGGAG; probe, FAM-5'-TGGTCTCCTCTGACTTCAACAGCGACAC-3'-TAMRA; human VEGF forward, AATGACGAGGGCCTGGAGT; human VEGF reverse, TTGATCCGCATAATCTGCATG; human VEGF probe, FAM-5'-TGTGCCCACTGAGGAGTCCAACATCA-3'-TAMRA.
In situ hybridization
All tissues were fixed in 4% buffered formalin and paraffin-embedded. Sections 5 µm thick were deparaffinized, deproteinated in 20 µg/ml proteinase K for 15 min at 37 C, and further processed for in situ hybridization as previously described (16). [33P]UTP-labeled sense and antisense riboprobes were hybridized at 55 C overnight, followed by a high stringency wash at 55 C in 0.1x standard saline citrate for 2 h. Before dipping the slides in photographic emulsion, the dry glass slides were exposed for 3 days at room temperature to Kodak BioMax MR autoradiographic film (Eastman Kodak Co., Rochester, NY). The slides were dipped in NTB2 nuclear track emulsion (Eastman Kodak Co.), exposed in sealed plastic slide boxes containing desiccant for 2–4 wk at 4 C, developed, and counterstained with hematoxylin and eosin (H&E). The EG-VEGF and VEGF probes were prepared as previously described (16). The PCR primers described below were designed to amplify fragments of EG-VEGF and VEGF genes. Upper and lower primers had 27 nucleotide extensions appended to the 5' ends encoding T7 RNA polymerase and T3 RNA polymerase promoters, respectively, for generation of sense and antisense transcripts: EG-VEGF probe: length, 760 nucleotides; upper primer, 5'-GCTTCGAGGGCTGCGGATGT-3'; lower primer, 5'-TGCCTTGGGGTGACTGTCTGC-3'; and VEGF probe: length, 604 nucleotides; upper primer, 5'-GGGCCTCCGAAACCATGAACT-3'; lower primer, 5'-TCCTCCTGCCCGGCTCAC-3'.
Production of monoclonal antibodies
Mice were repeatedly injected i.p. with recombinant human EG-VEGF protein. Animals were killed, and fusions were carried out following a routine protocol. Screening of hybridoma supernatants was performed by ELISA. The expected size of the protein recognized by antibodies was confirmed by Western blots.
Immunohistochemistry
Immunohistochemistry was performed on 5-µm paraffin sections of human tissue. Briefly, tissue sections were deparaffinized in xylene and alcohols, then rehydrated. Antigen retrieval was performed using 0.1% trypsin (S2012, DakoCytomation, Carpinteria, CA) at 37 C for 30 min (for factor VIII detection) or using target retrieval solution (S1700, DakoCytomation) for 20 min at 99 C (for EG-VEGF detection). Tissue sections were allowed to cool for 20 min, washed with PBS, and were blocked for endogenous peroxidase, avidin, and biotin. Tissue sections were then blocked in 10% normal goat serum (for factor VIII detection) or in 10% horse serum (for EG-VEGF detection) in 3% BSA/PBS. Tissue sections were incubated for 30 min at room temperature with 1.425 µg/ml factor VIII-related antigen (A0082, DakoCytomation) or in 10 µg/ml mouse monoclonal anti-EG-VEGF (clone 1A1) antibody for 60 min at room temperature. For secondary detection, tissue sections were incubated with 7.5 µg/ml biotinylated goat antirabbit (BA-1000, Vector Laboratories, Inc., Burlingame, CA; for factor VIII detection) or in 2.5 µg/ml biotinylated horse antimouse (BA-2001, Vector Laboratories, Inc., for EG-VEGF detection) for 30 min at room temperature. Finally, tissue sections were incubated in avidin-biotin-peroxidase complex (PK-6100, Vector Laboratories, Inc.) for 30 min at room temperature. After each incubation, the tissue sections were gently washed three times in PBS. After the final incubation and wash, sections were developed with diaminobenzidine, counterstained, and mounted.
Measurements of vessel density
Vessel surface density was measured as follows: digital images of stained sections were collected using a x10 objective. The pixels corresponding to stained vessels were selected in Photoshop software; contaminating (nonvessel) stray pixels were eliminated; vessel outlines were smoothed. Single pixel vessel borders were then drawn, and the aggregate pixel area (equivalent to vessel perimeter length in pixels) was measured. The net area of the entire image was noted. Vessel perimeter lengths and image area were converted to micrometers and square micrometers, respectively, and used to calculate the vessel surface density (vessel perimeter/image area), reported in units of microns–1.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To further characterize the site of expression of EG-VEGF in the adult testis, we performed a series of experiments to detect both EG-VEGF transcripts and EG-VEGF proteins in the same panel of testis sections. As we have previously shown (7), the expression of EG-VEGF mRNA was detected by in situ hybridization in cell clusters within the interstitial space of testis. These cell clusters presented all features expected for Leydig cells (Fig. 1B
). Moreover, monoclonal antibodies raised against the recombinant EG-VEGF protein were tested for the ability to recognize the EG-VEGF protein by Western blot (data not shown). Then, a series of six monoclonal antibodies was tested in immunohistochemical experiments on frozen adult testis sections (data not shown). One monoclonal antibody, 1A1, presenting the strongest staining, was then tested on paraffin-embedded testis sections (Fig. 1). No staining was observed with IgG isotype control (Fig. 1D) or when the 1A1 antibody was preabsorbed with recombinant EG-VEGF protein (Fig. 1E). With 1A1 antibody, a strong and specific perinuclear staining, as expected for a secreted protein, was observed within cell clusters in interstitial spaces (Fig. 1, F and G and inset), whereas the seminiferous tubules were not labeled.
|
To investigate the testicular localization of EG-VEGF at different developmental stages, paraffin-embedded testis sections from 12-, 13-, 14-, 16-, 18-, 19-, 22-, 24-, 26-, 28-, 32-, 34-, and 40-wk-old human fetuses were used to visualize both EG-VEGF mRNA by in situ hybridization and EG-VEGF protein using the 1A1 antibody by immunohistochemistry. Although no EG-VEGF transcripts were observed in the embryonic gonad at 12 and 13 wk (Fig. 2B), EG-VEGF-positive cells were observed after 14 wk (Fig. 2B), and the number of cells presenting a specific signal increased thereafter, as shown at 34 wk (Fig. 2F). In the same way, using an immunohistochemical approach, no EG-VEGF staining was observed in the embryonic gonad before 13 wk (Fig. 2G), but EG-VEGF-positive cells were detected in the gonads of the 14-wk-old fetus (Fig. 3B) and thereafter (data not shown).
|
|
We used three independent approaches to investigate the presence of EG-VEGF transcripts or proteins in a series of testicular tumors. To measure EG-VEGF mRNA, TaqMan analysis was performed using frozen samples from seven normal adult testes, five seminomas, four testes with CIS, four NSGCT, and one Leydig cell tumor (Fig. 3
). EG-VEGF mRNA was expressed in each adult normal testis at different levels, whereas neither seminoma nor NSGCT expressed EG-VEGF transcript. In contrast, testes containing CIS cells and Leydig cell tumors strongly or moderately expressed EG-VEGF mRNA. To define the cellular localization of EG-VEGF transcripts in testicular tumors, particularly in testes with CIS and Leydig cell tumors, we performed in situ hybridization in eight seminomas, nine NSGCT, seven testis specimens with CIS cells, and four Leydig cell tumor specimens. As expected, no staining was observed in seminoma or NSGCT (Fig. 4, B and D), whereas specific staining was present in testes containing CIS cells, localized in the Leydig cell cluster (Fig. 4F). Moreover, very strong staining was detected in Leydig cell tumors (Fig. 4H).
|
), or in the 15 NSGCT (Fig. 5B). In contrast, immunostaining corresponding to Leydig cell clusters was observed in all seven CIS specimens (Fig. 5C), and more interestingly, a strong immunoreaction was detected in all four Leydig cell tumors (Fig. 5D).
|
Endothelial immunostaining within the testicular tumors using an anti-von Willebrand factor antibody was performed to quantify the vascular density in four seminoma specimens (one shown in Fig. 6A) that presented a negative EG-VEGF expression pattern and in four Leydig cell tumors (one shown in Fig. 6B) that were positive for EG-VEGF. Our data showed that von Willebrand factor reactivity in Leydig cell tumors or seminoma was confined to vascular endothelium. Leydig cell tumors as well as seminoma samples showed relatively uniformly sized and spaced small caliber vessels, approximately 8–10 µm in diameter. The sizes of the biggest blood vessels were similar in seminomas and Leydig cell tumors, but the number of microvessels was 3.2-fold higher in Leydig cell tumors than in seminoma (P < 0.005; Fig. 6C). Finally, to determine whether another angiogenic factor, VEGF, might potentially trigger angiogenesis in synergy with EG-VEGF within seminomas and Leydig cell tumor, we measured the EG-VEGF and VEGF mRNA expression by real-time PCR in a panel of three seminomas and one Leydig cell tumor as well as in normal testis that was used as a positive control (Fig. 7A). The TaqMan analysis showed that EG-VEGF and VEGF mRNAs were expressed in three specimens of normal testis as expected (7, 17). In contrast, VEGF mRNA, but not EG-VEGF, was expressed in three seminomas, whereas EG-VEGF, but, surprisingly, very weakly VEGF, was expressed in the Leydig cell tumor. Indeed, EG-VEGF mRNA was approximately 30-fold (25) more abundant than VEGF mRNA (computed tomography value, 22.2 vs. 27.2). To confirm these finding, we performed in situ hybridization in four Leydig cell tumors to compare EG-VEGF and VEGF mRNA expression. We found that EG-VEGF expression was consistently (four of four) and significantly higher than VEGF expression in all samples studied (Figs. 7B and 8).
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In the present study we demonstrate EG-VEGF expression in Leydig cells of human fetal testes starting from 14 wk gestation. These observations are consistent with the fetal production of another protein, steroidogenic acute regulatory protein (StAR), which has been extensively studied in testis. StAR is involved in testosterone production (18), and the expression of StAR as well as testosterone begins at 14 wk gestation (19). The 14 wk point is therefore a crucial time for human testis development; the onset of EG-VEGF expression during the same period argues that EG-VEGF-mediated angiogenesis at this time may be critical for normal endocrine function. Angiogenesis, dependent on EG-VEGF secretion, may permit the efficient transport of newly secreted testosterone from the testis to other target tissues and may allow the transport of steroidogenic substances and regulatory hormones, such as gonadotropins, from the periphery toward the testis. The overall result of EG-VEGF-mediated angiogenesis may be efficient hormone transport and regulation.
The factors regulating the expression of EG-VEGF are still incompletely characterized. The restricted localization of EG-VEGF expression to Leydig cells, the fact that EG-VEGF is expressed at a crucial time in fetal testicular development, and the presence of a binding site for the transcription factor, steroidogenic factor-1, in the promoter region upstream of the EG-VEGF gene (20) all argue in favor of the hypothesis that hormonal factors regulate EG-VEGF. We found that the amount of EG-VEGF transcripts in whole normal testes was variable from patient to patient (ranging from 20–240), possibly the result of a difference in the hormonal status of these patients.
The best candidate for hormonal regulation in EG-VEGF secretion is LH, which is well known to regulate genes involved in the biosynthesis of testosterone (21, 22, 23). ACTH, which was identified recently as a direct stimulator of testosterone production in fetal testes (24), is another candidate. However, the roles of these hormones as well as that of the transcription factor steroidogenic factor-1 in EG-VEGF induction remain to be established experimentally.
Because angiogenesis is a prerequisite for cancer growth (2), and EG-VEGF is able to induce the proliferation and migration of capillary endothelial cells (7), we investigated EG-VEGF expression in different types of testicular tumors. EG-VEGF expression, as assessed by mRNA as well as protein expression, was found to be restricted to Leydig cell tumors; no expression was observed in neoplasms derived from germ cells or in Sertoli cell tumors. For the CIS specimens, we found high levels of EG-VEGF transcripts compared with normal testis (20-fold more in two of four specimens). The in situ hybridization and immunohistochemical investigation demonstrated clearly that it was not the CIS itself that expressed EG-VEGF, but the Leydig cells around the tubules containing CIS. It has been previously reported that higher LH levels are found in men with CIS of the testis (25). This observation is in agreement with our hypothesis that LH can induce EG-VEGF secretion. It has also been demonstrated that StAR expression preferentially expressed in Leydig cell tumors relative to other testicular tumors (18), and that Leydig cell tumors secreted testosterone and were regulated by LH (26, 27). However, we cannot exclude the possibility that the high levels of EG-VEGF transcripts in testis containing CIS cells may be due to the reduced number of cells in the tissue, which may result in proportionately higher expression of EG-VEGF in these samples. In contrast to EG-VEGF, VEGF has been shown to be expressed in seminomas and NSGCT (5). EG-VEGF expression is strictly limited to Leydig cell neoplasms, confirming a strong relationship between EG-VEGF expression and potential hormone regulation and suggesting that EG-VEGF plays a role in angiogenesis of this tumor. Interestingly, few studies reported that VEGF expression was correlated with microvascular density within testis tumors (5, 28). To investigate the hypothesis of a potential role for EG-VEGF in tumor angiogenesis, we examined microvessel density in seminoma compared with that in Leydig cell tumors. We found that the microvessel count was significantly higher in Leydig cell tumors than in seminoma. Furthermore, although VEGF was expressed by seminomas as expected (5), it was not expressed at significant levels by Leydig cell tumors. These findings suggest that EG-VEGF, but not (or very weakly) VEGF, is responsible for angiogenesis of Leydig cell tumors. Leydig cell tumors are relatively rare; therefore, little is known about the hormone regulation of their growth. Nevertheless, it is well established that certain activating mutations of the LH receptor gene cause Leydig cell tumors (29), confirming the major role of LH signaling pathway in Leydig cell tumor progression.
Taken together, our results suggest that EG-VEGF plays an important role in testis development during the first stages of testicular endocrine activity, and that EG-VEGF secretion can promote Leydig cell tumor growth by increasing angiogenesis. Additional studies will be required to verify this hypothesis, including gain/loss of function approaches.
|
Footnotes |
|---|
Received November 21, 2003.
Accepted May 10, 2004.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-hydroxylase/C17–20 lyase in Leydig cells. Biol Reprod 42:399–404[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  R. A. L. Bayne, T. Forster, S. T. G. Burgess, M. Craigon, M. J. Walton, D. T. Baird, P. Ghazal, and R. A. Anderson Molecular Profiling of the Human Testis Reveals Stringent Pathway-Specific Regulation of RNA Expression Following Gonadotropin Suppression and Progestogen Treatment J Androl, July 1, 2008; 29(4): 389 - 403. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Korsisaari, J. Ross, X. Wu, M. Kowanetz, N. Pal, L. Hall, J. Eastham-Anderson, W. F. Forrest, N. Van Bruggen, F. V. Peale, et al. Blocking Vascular Endothelial Growth Factor-A Inhibits the Growth of Pituitary Adenomas and Lowers Serum Prolactin Level in a Mouse Model of Multiple Endocrine Neoplasia Type 1 Clin. Cancer Res., January 1, 2008; 14(1): 249 - 258. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. K. Malik, M. E. Baldwin, F. Peale, G. Fuh, W.-C. Liang, H. Lowman, G. Meng, N. Ferrara, and H.-P. Gerber Redundant roles of VEGF-B and PlGF during selective VEGF-A blockade in mice Blood, January 15, 2006; 107(2): 550 - 557. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W.-C. Liang, X. Wu, F. V. Peale, C. V. Lee, Y. G. Meng, J. Gutierrez, L. Fu, A. K. Malik, H.-P. Gerber, N. Ferrara, et al. Cross-species Vascular Endothelial Growth Factor (VEGF)-blocking Antibodies Completely Inhibit the Growth of Human Tumor Xenografts and Measure the Contribution of Stromal VEGF J. Biol. Chem., January 13, 2006; 281(2): 951 - 961. [Abstract] [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 |
) and on a panel of specimens from testicular tumors, including five seminomas (Sem1–5), four testis containing positive CIS tubules (CIS1–4; 
), four nonseminomatous germ cell tumors (NST1–4), and one Leydig cell tumor (Leyd T1; 
). The TaqMan analysis revealed no EG-VEGF expression in seminomas and nonseminomatous germ cell tumors, but EG-VEGF expression was present in normal testes, Leydig cell tumor, and testes containing seminiferous tubules with CIS. The values corresponded to the ratio of EG-VEGF mRNA to GAPDH mRNA and were the means of triplicate determinations.
