- Email: [email protected]
Experiment and observation on invasion of brain glioma in vivo
Journal of Clinical Neuroscience (2002) 9(6), 668±671 & 2002 Elsevier Science Ltd. All rights reserved. DOI: 10.1054/jocn.2002.1140, available online at http://www.idealibrary.com on
Laboratory studies
Experiment and observation on invasion of brain glioma in vivo Xiang Zhang MD, PHD, Xia Li MD, Jing-wen Wu MD, PHD, Da-kuan Gao MD, Jing-wen Liang MD, Xian-zhen Liu MD Department of Neurosurgery, Xi-Jing Hospital, The Fourth Military Medical University, Xi'an, P.R. China
Summary Research on invasion and metastasis of glioma in vivo was performed by implanting C6 glioma cells transfected with enhanced green fluorescent protein (EGFP) gene into the brain of SD rats. Firstly, C6 glioma cells were transfected with a plasmid vector (pEGFP-N3) containing the EGFP gene. Stable EGFP-expressing clones were isolated and examination for these cells by flow cytometry and electron microscope was done. Secondly, EGFP-expressing cells were stereotactically injected into the brain parenchyma of SD rats to establish xenotransplanted tumor. Four weeks later rats were killed and continuous brain sections were examined using fluorescence microscopy after adjacent sections were examined by immunohistochemistry or routine hematoxylin and eosin staining for the visualization and detection of tumor cell invasion. Xenotransplanted tumor was primarily cultured to determine the storage of EGFP gene in vivo. The results showed that EGFP-transfected C6 glioma cells maintained stable high-level EGFP expression in the central nervous system during their growth in vivo. EGFP fluorescence clearly demarcated the primary tumor margin and readily allowed for the visualization of distant micrometastasis and invasion on the single-cell level. Small locally invasive foci, including those immediately adjacent to the leading invasive edge of the tumor, were virtually undetectable by routine hematoxylin and eosin staining and immunohistochemistry. These results suggested that EGFP-transfected C6 cells can be visualized by fluorescence microscopy after intracranial implantation. This model is an excellent experimental animal model in research on invasion and metastasis of brain glioma in vivo. & 2002 Elsevier Science Ltd. All rights reserved. Keywords: glioma, green fluorescent protein, invasion, flow cytometry
INTRODUCTION
MATERIALS AND METHODS
Knowledge of the biological character of glioma is rapidly expanding day by day. However, there are some shortcomings in the experimental model in vivo which could not permit the visualization of early tumor invasion, especially for the single invasive tumor cell. Identification of small numbers or even individual invading tumor cells against a vast background of normal host CNS cells may be impossible by standard HE staining. The deficiency of a special antibody against glioma cells decreased the sensitivity and specificity of immunohistochemistry.1 Glioma cells transfected with the Lac-Z marker gene have been implanted into the nude mice to detect metastasis and invasion of tumor in vivo, but the visualization of invasive cells requires extensive histological preparation and thus precludes rapid, easy detection of invasion.2 In this study C6 cells were transfected with the enhanced green fluorescent protein (EGFP) gene, and then implanted into brain of SD rats to establish xenotransplanted tumor. The metastasis and invasion of xenotransplanted tumor was examined by fluorescence microscopy and compared with results from HE staining and immunohistochemistry staining. Xenotransplanted tumor was primarily cultured and then examined by FCM to analyze the storage of EGFP gene in vivo.
Experiment materials
Received 30 August 2000 Accepted 13 November 2001 Correspondence to: Professor Xiang Zhang, Director of Department of Neurosurgery, Xi-Jing Hospital, Fourth Military Medical University, West Chang-Le road, No. 15, Xi'an 710032, P.R. China. Tel.: 86-29-3375323; Fax: 86-29-3295533; E-mail address: [email protected] (Xiang Zhang).
668
C6 glioma cell line (Cell Biology Institute of Shanghai); fetal calf serum, RPMI-1640 medium, trypsin (Hyclone); G418, Lipofectamine (Gibco); pEGFP-N3 plasmid (Clontech); mouse GFAP polyclonal antibody against rat (Sigma); SABC immunohistochemistry kit (Wuhan Boster); 10 male SD rats (weighing 250±300 g) were randomly divided into two groups with 5 rats in each group (Experimental group rats had implanted C6 cells/transfected EGFP gene; control group rats had implanted untransfected C6 cells); stereotactic frame (our Institute); low melting point agarose (Gibco).
EGFP gene transfection C6 glioma cells were cultured in RPMI-1640 medium containing 10% fetal calf serum, 100 units/ml penicillin G, and 100 units/ml streptomycin in humid atmosphere at 37 C. Fifty percent confluent cells in 25 ml flask were incubated overnight for gene transfection after fresh medium was changed in experimental group. Fifteen ml lipofectamine and 2 mg plasmid containing the EGFP and neomycin resistance gene respectively were diluted into 100 ml by the serum-free medium. The cells were syringed three times with serum-free medium before transfection. 0.8 ml serum-free medium was added to the DNAlipofectamine complex, and then mixed liquid was dropped onto the cell surface. Five hours later, 2 ml 20% FCSRPMI1640 medium was added into the culture flask. After 48 h of incubation, isolation of stable transfectants expressing EGFP was performed by selection in 600 mg/ml G418 for 30 days. Isolated clones were harvested by trypsin and further amplified using conventional culture method.3 Tumor cells in control group were not transfected by EGFP gene.
Invasion of brain glioma in vivo 669
Examination for C6 cells in vitro C6 cells of experimental and control group were harvested with 0.25% trypsin during the log phase of growth, washed twice in phosphate-buffered saline (PBS). 1 109 cells were resuspended in PBS, and then were mingled with 2 ml 100% alcohol which had been cooled at 4 C. The complex was examined by flow cytometry. Cells were harvested according to above method and examined by electron microscope after they were fixed by 3% glutaraldehyde. Single cell suspension was seeded in the 90 mm dish, and then cells which were syringed twice with PBS and fixed with paraformldehyde for 30 min after their adhesion and extension were examined by fluorescent microscopy. Experimental background fluorescence was established by non-EGFP-transfected C6 cells in the analysis, and the length wave of excited light is 488 nm.
method.5 Cells were passaged with routine methods and divided randomly into experimental group and control group. The experimental group maintained selection in 600 mg/ml G418, while control group was passaged routinely. Two groups of cells passaged to fifth generation during the log growth phase respectively performed fluorescence and cell circle examination by the flow cytometry, and the expression rate of EGFP was compared between these two groups. Statistic assay S) The data was expressed by mean standard deviation (X and the t test was performed by SPSS statistic software. RESULTS Examination of transfected C6 cells in vitro
Establishment of rat xenotransplanted tumor Stable EGFP-expressing C6 cells were passaged according to routine method and harvested with trypsin. Then the cells were washed twice in serum-free RPMI 1640. 3 106 cells were deposited by centrifugalization and then mixed with 30 ml double medium containing 1% low melting point agarose. The mixed cell suspension was placed in 37 C water bath for use. Male SD rats were anesthetized with phenytoin sodium which was administered intraperitoneally at 40 mg/kg. A midline scalp incision was made and the coronal suture exposed. A hole in cranial bones was made 3 mm to the right of midline and 0.5 mm anterior to the coronal suture by dental broach. Using a Hamilton syringe fixed in a small animal stereotactic frame, 10 ml of cell suspension was injected to a depth of 4±5 mm and at a rate of 2 ml per minute. Bone wax was applied over the injected site, and then skin was closed with 7±0 silk suture.4 Four weeks later, the rats were killed and the whole brain was fixed with 4% paraformldehyde and removed. Samples were dehydrated, lucidied, soaked in wax, and then were cut to 5 mm-thick sections with routine histological methods. The xenotransplanted tumor established with untransfected C6 cells by the method as above was used as control, whose occurrence and diameter was determined and compared with those of xenograft established by transfected cells. Sections were not made in these xenografts. Examination of rat xenotransplanted tumor Adjacent sections were dewaxed, stained with HE and immunohistochemistry method. The glial fibrillary acidic protein (GFAP) was used as a tumor marker antigen for immunostaining. Each section was incubated with the mouse polyclonal antibody against rat GFAP at 1:50 dilution and the streptavidin-biotin complex was used to localize the antibody bound to antigen. Concrete operation procedure accords to the operation manual of SABC immunohistochemistry kit producted by Wuhan boster company. For negative controls, immune serum was replaced by PBS. Astrocytoma tissue served as positive controls. Sections were observed directly under fluorescent microscopy using an excitation wavelength of 488 nm and a 530 nm bandpass filter for EGFP detection after these stained sections had been observed under microscope. Primary culture of the xenografts The xenografts established by transfected cells were aseptically removed and primarily cultured by tissue-mass culture & 2002 Elsevier Science Ltd. All rights reserved.
Stable EGFP-expressing cells could be readily visualized by fluorescence microscopy equipped with a 488 nm excitation filter and a 530 nm emission filter for EGFP detection after selected by G418 (Fig. 1A). The analysis by FCM showed that the rate of EGFP expression of transfected C6 cells during the log phase of growth was 95% and a small subset of EGFPnegative cells which did not give out fluorscence represented quiescent cells (G0/G1) (Fig. 2B). There were no evident differences in cell cycle and ultrastructure between these two groups. Examination of xenografts Rats were killed 4 weeks after tumor cells were implanted. All of the rats had grossly visible tumors averaging 3.5 0.014 mm in diameter while they were killed and the rate of tumorigenesis was 100%. There was no evident difference in the rate of tumorigenesis and diameter of xenograft compared with control group (100% for the rate of tumorigenesis, and 3.6 0.016 mm for averaging diameter of control group, P > 0.05, t 1.526). HE and immunohistochemistry staining could demarcate the xenografts, but these two kinds of methods failed to efficiently identify small numbers or even individual invading tumor cells against background of normal host CNS cells. Under fluorescence microscopy, the tumors were found to be highly fluorescent throughout and the margin between the tumor bulk and normal brain tissue was easy to demarcate, thus demonstrating continued stable, highlevel EGFP expression in vivo during tumor growth intracerebrally (Fig. 1B). Detection of locally invasive tumor cells as well as distant metastases down to the single-cell level was readily accomplished using fluorescence microscopy and the most invasive distance is 9.5 mm determined by fluorescence microscope (Fig. 1C). Analysis of EGFP expression for xenografts FCM analysis of the cells recovered from the tumor extracts showed that the rate of EGFP expression in the experimental and control group respectively were 93% and 92%, which revealed that there was not loss of EGFP gene in the growth of stable transfected cells in vivo (Fig. 2C, D). DISCUSSION Brain glioma represents 40±60% of primary brain tumors. Although operation, radiotherapy, chemotherapy and immunotherapy have been used for the patients with glioma, the Journal of Clinical Neuroscience (2002) 9(6), 668±671
670 Zhang et al.
A
B
C
Fig. 1 Examination for the expression of EGFP with the fluorescent microscopy. (A) transfected C6 cells in vitro 500; (B) tumor bulk area 400; (C) single invasive glioma cell, 9.5 mm to the center of transplanted tumor 650.
prognosis of these patients is bad because of invasive growth of the glioma in vivo.6 A significant task encountered in studies of glioma invasion has been to establish an experimental in vivo model that could clearly facilitate the detection and subsequent analysis of tumor invasion.7 HE and immunohistochemistry staining are common methods to detect the tumor cell morphology. The sensitivity and specificity of HE is much worse and identification of individual invasive tumor cells against other cells may be impossible by HE under microscope with low magnification. Immunohistochemistry is a method extensively used, but lack of a specific antibody against glioma cells limited its application in detection of glioma invasion. GFAP is involved in the construction of the skeleton of glioma cells but GFAP also exist in other cells derived from astrocytes. Marking the glioma cells with antibody against GFAP reduced the specificity.8 Tumor cells labeled by exotic fluorescein, such as fast blue, were used to detect xenotransplanted tumor in previous study, but it is possible that fluorescein infiltrates out of tumor cells and fluorescence intensity decreases following the passage of cells, which influences the examination.9A recent study suggested that FB staining reduced cell adhesion and proliferative activity and also had a significant inhibitory effect on cell migration, so xenografts formed by FB staining cells failed to imitate the growth of tumor in vivo.10 Luciferase is also used as reporter gene, but it is complicated to detect fluorescence because exotic fluorescein is needed.11 Transfection of the Lac-Z gene into human tumor cells has been extensively used Journal of Clinical Neuroscience (2002) 9(6), 668±671
in previous studies for the detection of micrometastasis and invasion. However, detection of Lac-Z requires extensive histological preparation and may result in a high background because of endogenous b-galactosidase activity in certain host cells. As Lac-Z detection is accomplished only after the fixation of cells, the failure to directly examine the viable cells limited the usage of Lac-Z system. Recently found EGFP which emits green fluorescence when excited by wavelengths corresponding to 488 nm is a kind of protein with low molecular weight and stable chemical property. It has been used to label protein to detect their localization within the cell.12 More recently, investigation has also shown that many kinds of tumor cells transfected with EGFP mutants, mediated by all kinds of vectors in vitro, stably expressed EGFP.13,14 It is unknown whether transfected glioma cells could maintain stable EGFP expression under the nonselective conditions of the CNS and thus allow the easy detection of glioma invasion and micrometastasis. In our studies C6 cells were transfected with a plasmid vector (pEGFPN3) containing the EGFP gene and neomycin resistance gene in vitro and stable EGFP-expressing cells were isolated by G418. Relevant studies showed that there were no obvious differences in cell morphology, ultrastructure, proliferative rate and cell cycle between positive transfected cells and untransfected cells. By stereotactical injecting EGFP-expressing cells intracerebrally into rats, bulk xenotransplanted tumor could form and tumorigenesis had no evident changes compared with control group. Stable expression of EGFP by & 2002 Elsevier Science Ltd. All rights reserved.
Invasion of brain glioma in vivo 671
Fig. 2 FCM analysis of the EGFP expression rate of C6 cells. (A) the rate of control C6 cells is 0; (B) the rate of transfected C6 cells is 95%; (C) the rate of primarily cultured xenografts cells experimental group is 93%; (D) the rate of primarily cultured xenografts cells control group is 92%.
transfected cells was able to clearly visualize distant tumor invasion and metastasis down to the single-cell level by fluorescence microscopy. Furthermore interesting genes for research can be inserted into the multiple clonal site of pEGFP-N3 vector, which provides conditions for the research on genes related to tumor invasion and metastasis. Our study suggested that tumor invasion and micrometastases, especially small numbers, or even individual invading tumor cells, can be detected in the xenotransplanted tumor established by implanting glioma cells transfected with EGFP gene into brain of SD rats. This method is superior to routine H&E, immunohistochemistry, and standard Lac-Z detection methods. Thus, EGFP tracking can be used as a preclinical experimental model by which potentially relevant therapeutic targets can be identified, measurement of the effect on invasion and metastasis by novel therapeutic agents can be conducted, and further insight into the mechanism of glioma invasion may be gained.
3. 4.
5.
6. 7.
8.
9.
10.
11.
REFERENCES 1.
2.
Weizsaecker M, Deen D, Rosenblum M et al. The 9L rat brain tumor: description and application of an animal model. J Neurol 1981; 224: 183±192. Lampson L, Lampson M, Dunne A et al. Exploiting the lacZ reporter gene for quantitative analysis of disseminated tumor growth within the brain: the use of the lacZ gene products as a tumor antigen for evaluation of antigenic modulation and to facilitate image analysis of tumor growth in situ. Cancer Res 1993; 53: 176±180.
& 2002 Elsevier Science Ltd. All rights reserved.
12.
13. 14.
Brash DE, Ponten J. Skin precancer. Cancer Surv 1998; 32: 69±113. Peoc'h M, Le Duc G, Trayaud A et al. Quantification and distribution of neovascularization following microinjection of C6 glioma cells in rat brain. Anticancer Res 1999; 19: 3025±3030. Zhang XY, Pettengell R, Nasiri N et al. Characteristics and growth patterns of human peritoneal mesothelial cells: comparison between advanced epithelial ovarian cancer and non-ovarian cancer sources. J Soc Gynecol Investig 1999; 6: 333±340. Zhang Xiang. The Oncology of Nervous System. The military medical press, 1999; Beijing, 1±50. Pilkington GJ, Bjerkvig R, De Ridder L et al. In vitro and in vivo models for the study of brain tumour invasion. Anticancer Res 1997; 17: 4107±4109. Luider TM, Kros JM, Sillevis PA et al. Glial fibrillary acidic protein and its fragments discriminate astrocytoma from oligodendroglioma. Electrophoresis 1999; 20: 1087±1091. Soderqvist H, Imreh G, Kihlmark M et al. Intracellular distribution of an integral nuclear pore membrane protein fused to green fluorescent protein: localization of a targeting domain. Eur J Biochem 1997; 250: 808±813. Vince GH, Bouterfa H, Goldbrunner R et al. Fast blue, a fluorescent tracer in glioma cell culture, affects cell proliferation and motility. Neurosci Lett 1997; 233: 148±150. Kurebayashi J, Mcleskey SW, Johnson MD et al. Quantitative demonstration of spontaneous metastasis by MCF-7 human breast cancer cells cotransfected with fibroblast growth factor 4 and lac-Z. Cancer Res 1993; 53: 2178±2187. Lin WC, Pretlow TP, Pretlow TG et al. Bacterial lac-z gene as a highly sensitive marker to detect micrometastasis formation during tumor progression. Cancer Res 1990; 50: 2808±2817. Chishima T, Yang M, Miyagi Y et al. Governing step of metastasis visualized in vitro. Proc Natl Acad Sci USA 1997; 94: 11573±11576. Perry A, Parisi JE, Kurtin PJ et al. Metastatic adenocarcinoma to the brain: an immunohistochemical approach. Hum Pathol 1997; 28: 938±943.
Journal of Clinical Neuroscience (2002) 9(6), 668±671
Laboratory studies
Experiment and observation on invasion of brain glioma in vivo Xiang Zhang MD, PHD, Xia Li MD, Jing-wen Wu MD, PHD, Da-kuan Gao MD, Jing-wen Liang MD, Xian-zhen Liu MD Department of Neurosurgery, Xi-Jing Hospital, The Fourth Military Medical University, Xi'an, P.R. China
Summary Research on invasion and metastasis of glioma in vivo was performed by implanting C6 glioma cells transfected with enhanced green fluorescent protein (EGFP) gene into the brain of SD rats. Firstly, C6 glioma cells were transfected with a plasmid vector (pEGFP-N3) containing the EGFP gene. Stable EGFP-expressing clones were isolated and examination for these cells by flow cytometry and electron microscope was done. Secondly, EGFP-expressing cells were stereotactically injected into the brain parenchyma of SD rats to establish xenotransplanted tumor. Four weeks later rats were killed and continuous brain sections were examined using fluorescence microscopy after adjacent sections were examined by immunohistochemistry or routine hematoxylin and eosin staining for the visualization and detection of tumor cell invasion. Xenotransplanted tumor was primarily cultured to determine the storage of EGFP gene in vivo. The results showed that EGFP-transfected C6 glioma cells maintained stable high-level EGFP expression in the central nervous system during their growth in vivo. EGFP fluorescence clearly demarcated the primary tumor margin and readily allowed for the visualization of distant micrometastasis and invasion on the single-cell level. Small locally invasive foci, including those immediately adjacent to the leading invasive edge of the tumor, were virtually undetectable by routine hematoxylin and eosin staining and immunohistochemistry. These results suggested that EGFP-transfected C6 cells can be visualized by fluorescence microscopy after intracranial implantation. This model is an excellent experimental animal model in research on invasion and metastasis of brain glioma in vivo. & 2002 Elsevier Science Ltd. All rights reserved. Keywords: glioma, green fluorescent protein, invasion, flow cytometry
INTRODUCTION
MATERIALS AND METHODS
Knowledge of the biological character of glioma is rapidly expanding day by day. However, there are some shortcomings in the experimental model in vivo which could not permit the visualization of early tumor invasion, especially for the single invasive tumor cell. Identification of small numbers or even individual invading tumor cells against a vast background of normal host CNS cells may be impossible by standard HE staining. The deficiency of a special antibody against glioma cells decreased the sensitivity and specificity of immunohistochemistry.1 Glioma cells transfected with the Lac-Z marker gene have been implanted into the nude mice to detect metastasis and invasion of tumor in vivo, but the visualization of invasive cells requires extensive histological preparation and thus precludes rapid, easy detection of invasion.2 In this study C6 cells were transfected with the enhanced green fluorescent protein (EGFP) gene, and then implanted into brain of SD rats to establish xenotransplanted tumor. The metastasis and invasion of xenotransplanted tumor was examined by fluorescence microscopy and compared with results from HE staining and immunohistochemistry staining. Xenotransplanted tumor was primarily cultured and then examined by FCM to analyze the storage of EGFP gene in vivo.
Experiment materials
Received 30 August 2000 Accepted 13 November 2001 Correspondence to: Professor Xiang Zhang, Director of Department of Neurosurgery, Xi-Jing Hospital, Fourth Military Medical University, West Chang-Le road, No. 15, Xi'an 710032, P.R. China. Tel.: 86-29-3375323; Fax: 86-29-3295533; E-mail address: [email protected] (Xiang Zhang).
668
C6 glioma cell line (Cell Biology Institute of Shanghai); fetal calf serum, RPMI-1640 medium, trypsin (Hyclone); G418, Lipofectamine (Gibco); pEGFP-N3 plasmid (Clontech); mouse GFAP polyclonal antibody against rat (Sigma); SABC immunohistochemistry kit (Wuhan Boster); 10 male SD rats (weighing 250±300 g) were randomly divided into two groups with 5 rats in each group (Experimental group rats had implanted C6 cells/transfected EGFP gene; control group rats had implanted untransfected C6 cells); stereotactic frame (our Institute); low melting point agarose (Gibco).
EGFP gene transfection C6 glioma cells were cultured in RPMI-1640 medium containing 10% fetal calf serum, 100 units/ml penicillin G, and 100 units/ml streptomycin in humid atmosphere at 37 C. Fifty percent confluent cells in 25 ml flask were incubated overnight for gene transfection after fresh medium was changed in experimental group. Fifteen ml lipofectamine and 2 mg plasmid containing the EGFP and neomycin resistance gene respectively were diluted into 100 ml by the serum-free medium. The cells were syringed three times with serum-free medium before transfection. 0.8 ml serum-free medium was added to the DNAlipofectamine complex, and then mixed liquid was dropped onto the cell surface. Five hours later, 2 ml 20% FCSRPMI1640 medium was added into the culture flask. After 48 h of incubation, isolation of stable transfectants expressing EGFP was performed by selection in 600 mg/ml G418 for 30 days. Isolated clones were harvested by trypsin and further amplified using conventional culture method.3 Tumor cells in control group were not transfected by EGFP gene.
Invasion of brain glioma in vivo 669
Examination for C6 cells in vitro C6 cells of experimental and control group were harvested with 0.25% trypsin during the log phase of growth, washed twice in phosphate-buffered saline (PBS). 1 109 cells were resuspended in PBS, and then were mingled with 2 ml 100% alcohol which had been cooled at 4 C. The complex was examined by flow cytometry. Cells were harvested according to above method and examined by electron microscope after they were fixed by 3% glutaraldehyde. Single cell suspension was seeded in the 90 mm dish, and then cells which were syringed twice with PBS and fixed with paraformldehyde for 30 min after their adhesion and extension were examined by fluorescent microscopy. Experimental background fluorescence was established by non-EGFP-transfected C6 cells in the analysis, and the length wave of excited light is 488 nm.
method.5 Cells were passaged with routine methods and divided randomly into experimental group and control group. The experimental group maintained selection in 600 mg/ml G418, while control group was passaged routinely. Two groups of cells passaged to fifth generation during the log growth phase respectively performed fluorescence and cell circle examination by the flow cytometry, and the expression rate of EGFP was compared between these two groups. Statistic assay S) The data was expressed by mean standard deviation (X and the t test was performed by SPSS statistic software. RESULTS Examination of transfected C6 cells in vitro
Establishment of rat xenotransplanted tumor Stable EGFP-expressing C6 cells were passaged according to routine method and harvested with trypsin. Then the cells were washed twice in serum-free RPMI 1640. 3 106 cells were deposited by centrifugalization and then mixed with 30 ml double medium containing 1% low melting point agarose. The mixed cell suspension was placed in 37 C water bath for use. Male SD rats were anesthetized with phenytoin sodium which was administered intraperitoneally at 40 mg/kg. A midline scalp incision was made and the coronal suture exposed. A hole in cranial bones was made 3 mm to the right of midline and 0.5 mm anterior to the coronal suture by dental broach. Using a Hamilton syringe fixed in a small animal stereotactic frame, 10 ml of cell suspension was injected to a depth of 4±5 mm and at a rate of 2 ml per minute. Bone wax was applied over the injected site, and then skin was closed with 7±0 silk suture.4 Four weeks later, the rats were killed and the whole brain was fixed with 4% paraformldehyde and removed. Samples were dehydrated, lucidied, soaked in wax, and then were cut to 5 mm-thick sections with routine histological methods. The xenotransplanted tumor established with untransfected C6 cells by the method as above was used as control, whose occurrence and diameter was determined and compared with those of xenograft established by transfected cells. Sections were not made in these xenografts. Examination of rat xenotransplanted tumor Adjacent sections were dewaxed, stained with HE and immunohistochemistry method. The glial fibrillary acidic protein (GFAP) was used as a tumor marker antigen for immunostaining. Each section was incubated with the mouse polyclonal antibody against rat GFAP at 1:50 dilution and the streptavidin-biotin complex was used to localize the antibody bound to antigen. Concrete operation procedure accords to the operation manual of SABC immunohistochemistry kit producted by Wuhan boster company. For negative controls, immune serum was replaced by PBS. Astrocytoma tissue served as positive controls. Sections were observed directly under fluorescent microscopy using an excitation wavelength of 488 nm and a 530 nm bandpass filter for EGFP detection after these stained sections had been observed under microscope. Primary culture of the xenografts The xenografts established by transfected cells were aseptically removed and primarily cultured by tissue-mass culture & 2002 Elsevier Science Ltd. All rights reserved.
Stable EGFP-expressing cells could be readily visualized by fluorescence microscopy equipped with a 488 nm excitation filter and a 530 nm emission filter for EGFP detection after selected by G418 (Fig. 1A). The analysis by FCM showed that the rate of EGFP expression of transfected C6 cells during the log phase of growth was 95% and a small subset of EGFPnegative cells which did not give out fluorscence represented quiescent cells (G0/G1) (Fig. 2B). There were no evident differences in cell cycle and ultrastructure between these two groups. Examination of xenografts Rats were killed 4 weeks after tumor cells were implanted. All of the rats had grossly visible tumors averaging 3.5 0.014 mm in diameter while they were killed and the rate of tumorigenesis was 100%. There was no evident difference in the rate of tumorigenesis and diameter of xenograft compared with control group (100% for the rate of tumorigenesis, and 3.6 0.016 mm for averaging diameter of control group, P > 0.05, t 1.526). HE and immunohistochemistry staining could demarcate the xenografts, but these two kinds of methods failed to efficiently identify small numbers or even individual invading tumor cells against background of normal host CNS cells. Under fluorescence microscopy, the tumors were found to be highly fluorescent throughout and the margin between the tumor bulk and normal brain tissue was easy to demarcate, thus demonstrating continued stable, highlevel EGFP expression in vivo during tumor growth intracerebrally (Fig. 1B). Detection of locally invasive tumor cells as well as distant metastases down to the single-cell level was readily accomplished using fluorescence microscopy and the most invasive distance is 9.5 mm determined by fluorescence microscope (Fig. 1C). Analysis of EGFP expression for xenografts FCM analysis of the cells recovered from the tumor extracts showed that the rate of EGFP expression in the experimental and control group respectively were 93% and 92%, which revealed that there was not loss of EGFP gene in the growth of stable transfected cells in vivo (Fig. 2C, D). DISCUSSION Brain glioma represents 40±60% of primary brain tumors. Although operation, radiotherapy, chemotherapy and immunotherapy have been used for the patients with glioma, the Journal of Clinical Neuroscience (2002) 9(6), 668±671
670 Zhang et al.
A
B
C
Fig. 1 Examination for the expression of EGFP with the fluorescent microscopy. (A) transfected C6 cells in vitro 500; (B) tumor bulk area 400; (C) single invasive glioma cell, 9.5 mm to the center of transplanted tumor 650.
prognosis of these patients is bad because of invasive growth of the glioma in vivo.6 A significant task encountered in studies of glioma invasion has been to establish an experimental in vivo model that could clearly facilitate the detection and subsequent analysis of tumor invasion.7 HE and immunohistochemistry staining are common methods to detect the tumor cell morphology. The sensitivity and specificity of HE is much worse and identification of individual invasive tumor cells against other cells may be impossible by HE under microscope with low magnification. Immunohistochemistry is a method extensively used, but lack of a specific antibody against glioma cells limited its application in detection of glioma invasion. GFAP is involved in the construction of the skeleton of glioma cells but GFAP also exist in other cells derived from astrocytes. Marking the glioma cells with antibody against GFAP reduced the specificity.8 Tumor cells labeled by exotic fluorescein, such as fast blue, were used to detect xenotransplanted tumor in previous study, but it is possible that fluorescein infiltrates out of tumor cells and fluorescence intensity decreases following the passage of cells, which influences the examination.9A recent study suggested that FB staining reduced cell adhesion and proliferative activity and also had a significant inhibitory effect on cell migration, so xenografts formed by FB staining cells failed to imitate the growth of tumor in vivo.10 Luciferase is also used as reporter gene, but it is complicated to detect fluorescence because exotic fluorescein is needed.11 Transfection of the Lac-Z gene into human tumor cells has been extensively used Journal of Clinical Neuroscience (2002) 9(6), 668±671
in previous studies for the detection of micrometastasis and invasion. However, detection of Lac-Z requires extensive histological preparation and may result in a high background because of endogenous b-galactosidase activity in certain host cells. As Lac-Z detection is accomplished only after the fixation of cells, the failure to directly examine the viable cells limited the usage of Lac-Z system. Recently found EGFP which emits green fluorescence when excited by wavelengths corresponding to 488 nm is a kind of protein with low molecular weight and stable chemical property. It has been used to label protein to detect their localization within the cell.12 More recently, investigation has also shown that many kinds of tumor cells transfected with EGFP mutants, mediated by all kinds of vectors in vitro, stably expressed EGFP.13,14 It is unknown whether transfected glioma cells could maintain stable EGFP expression under the nonselective conditions of the CNS and thus allow the easy detection of glioma invasion and micrometastasis. In our studies C6 cells were transfected with a plasmid vector (pEGFPN3) containing the EGFP gene and neomycin resistance gene in vitro and stable EGFP-expressing cells were isolated by G418. Relevant studies showed that there were no obvious differences in cell morphology, ultrastructure, proliferative rate and cell cycle between positive transfected cells and untransfected cells. By stereotactical injecting EGFP-expressing cells intracerebrally into rats, bulk xenotransplanted tumor could form and tumorigenesis had no evident changes compared with control group. Stable expression of EGFP by & 2002 Elsevier Science Ltd. All rights reserved.
Invasion of brain glioma in vivo 671
Fig. 2 FCM analysis of the EGFP expression rate of C6 cells. (A) the rate of control C6 cells is 0; (B) the rate of transfected C6 cells is 95%; (C) the rate of primarily cultured xenografts cells experimental group is 93%; (D) the rate of primarily cultured xenografts cells control group is 92%.
transfected cells was able to clearly visualize distant tumor invasion and metastasis down to the single-cell level by fluorescence microscopy. Furthermore interesting genes for research can be inserted into the multiple clonal site of pEGFP-N3 vector, which provides conditions for the research on genes related to tumor invasion and metastasis. Our study suggested that tumor invasion and micrometastases, especially small numbers, or even individual invading tumor cells, can be detected in the xenotransplanted tumor established by implanting glioma cells transfected with EGFP gene into brain of SD rats. This method is superior to routine H&E, immunohistochemistry, and standard Lac-Z detection methods. Thus, EGFP tracking can be used as a preclinical experimental model by which potentially relevant therapeutic targets can be identified, measurement of the effect on invasion and metastasis by novel therapeutic agents can be conducted, and further insight into the mechanism of glioma invasion may be gained.
3. 4.
5.
6. 7.
8.
9.
10.
11.
REFERENCES 1.
2.
Weizsaecker M, Deen D, Rosenblum M et al. The 9L rat brain tumor: description and application of an animal model. J Neurol 1981; 224: 183±192. Lampson L, Lampson M, Dunne A et al. Exploiting the lacZ reporter gene for quantitative analysis of disseminated tumor growth within the brain: the use of the lacZ gene products as a tumor antigen for evaluation of antigenic modulation and to facilitate image analysis of tumor growth in situ. Cancer Res 1993; 53: 176±180.
& 2002 Elsevier Science Ltd. All rights reserved.
12.
13. 14.
Brash DE, Ponten J. Skin precancer. Cancer Surv 1998; 32: 69±113. Peoc'h M, Le Duc G, Trayaud A et al. Quantification and distribution of neovascularization following microinjection of C6 glioma cells in rat brain. Anticancer Res 1999; 19: 3025±3030. Zhang XY, Pettengell R, Nasiri N et al. Characteristics and growth patterns of human peritoneal mesothelial cells: comparison between advanced epithelial ovarian cancer and non-ovarian cancer sources. J Soc Gynecol Investig 1999; 6: 333±340. Zhang Xiang. The Oncology of Nervous System. The military medical press, 1999; Beijing, 1±50. Pilkington GJ, Bjerkvig R, De Ridder L et al. In vitro and in vivo models for the study of brain tumour invasion. Anticancer Res 1997; 17: 4107±4109. Luider TM, Kros JM, Sillevis PA et al. Glial fibrillary acidic protein and its fragments discriminate astrocytoma from oligodendroglioma. Electrophoresis 1999; 20: 1087±1091. Soderqvist H, Imreh G, Kihlmark M et al. Intracellular distribution of an integral nuclear pore membrane protein fused to green fluorescent protein: localization of a targeting domain. Eur J Biochem 1997; 250: 808±813. Vince GH, Bouterfa H, Goldbrunner R et al. Fast blue, a fluorescent tracer in glioma cell culture, affects cell proliferation and motility. Neurosci Lett 1997; 233: 148±150. Kurebayashi J, Mcleskey SW, Johnson MD et al. Quantitative demonstration of spontaneous metastasis by MCF-7 human breast cancer cells cotransfected with fibroblast growth factor 4 and lac-Z. Cancer Res 1993; 53: 2178±2187. Lin WC, Pretlow TP, Pretlow TG et al. Bacterial lac-z gene as a highly sensitive marker to detect micrometastasis formation during tumor progression. Cancer Res 1990; 50: 2808±2817. Chishima T, Yang M, Miyagi Y et al. Governing step of metastasis visualized in vitro. Proc Natl Acad Sci USA 1997; 94: 11573±11576. Perry A, Parisi JE, Kurtin PJ et al. Metastatic adenocarcinoma to the brain: an immunohistochemical approach. Hum Pathol 1997; 28: 938±943.
Journal of Clinical Neuroscience (2002) 9(6), 668±671