Download - Animal models for brain tumors: historical perspectives and future directions

J Neurosurg 80:865-876, 1994

Animal models for brain tumors: historical perspectives and future directions

DANIEL L. PETERSON, M.D., PETER J. SnERIOAN, PH.D., AND WILLIS E. BROWN, JR., M.D.

Division of Neurosurgery and Department of Cellular and Structural Biology, The University of Texas Health Science Center, San Antonio, Texas

~," The scientific understanding of the biology of human brain tumors has advanced in large part through the use of animal models. For most of this century, investigators have been evaluating the inciting factors in brain tumor development, and applying this knowledge to direct tumor growth in laboratory animals. Virus-induced, carcinogen-induced, and transplant-based models have been vigorously investigated. As knowledge of the molecular biology of neoplasia has advanced, transgenic technology has been introduced. The authors review the development of animal models for brain tumor, and focus on the role of transgenic models in elucidating the complex process of central nervous system neoplasia.

KEy WORDS �9 animal model . brain neoplasm transgenie mouse

ESPITE tremendous efforts by neurosurgeons, basic scientists, oncologists, radiation therapists, and others, the prognosis for the patient with

a malignant primary brain tumor has changed mini- mally over the last two decades. Improvement in the quality of survival for patients with glioma remains one of the foremost challenges facing this generation of neurosurgeons. As stated by Ausman, et aL, 1 in 1970, "there seems to be little possibility of improving the surgical or radiation therapies of malignant brain tumors." It appeared at that time that the development of new and better animal models of glioma would be needed for in vivo evaluation of potential chemotherapeutic and adjunctive therapies for brain tumors. The extensive contributions involving animal models by investigators such as Bigner and Wilson have directed the biological study and adjunctive therapy of brain tumors over the past 25 years. This review is meant to supplement and update information in this already well-documented field.

The justification and need for accurate, representa- tive animal models of glioma remain today just as great as in 1970, although the advent of brachytherapy and stereotactically coupled focused beam irradiation offer room for some optimism. While countless chemotherapeutic agents have been tested, few have been found useful in the adjunctive therapy of malignant glioma. As our understanding of the molecular genetics of brain tumors expands, the direction of adjunctive therapy will follow in a biologically oriented manner.

The pioneering work of Martuza and colleagues 36 at the Molecular Neurogenetics Unit at the Massachusetts General Hospital, and the work of Oldfield and colleagues 3~ at the National Institute of Neurological Dis- orders and Stroke in the area of retroviral herpes simplex virus-l-thymidine kinase gene transfer-induced ganciclovir sensitivity will lead the therapy of malignant glioma into the nucleus, and into the next century. The elegance and complexity of this work deserve an animal model that will adequately represent the growth characteristics of gliomas. In this article, we outline the development of animal models for brain tumor to date, and speculate on future directions for such animal models.

Requirements for Animal Model of Glioma

A valid animal model for brain tumors must display a variety of specifications: it must be composed of glial-derived neoplastic cells; the growth rate of the tumor and its malignancy characteristics should be pre- dictable and reproducible; the species used should be small and inexpensively maintained so that large numbers may be evaluated for reliable chemotherapeutic testing; the time to tumor induction should be relatively short and the survival time after induction should be standardized; the tumor should have glioma-like intra- parenchymal growth and blood-brain characteristics showing invasion, neovascularity, and no encapsula- tion; there should be no seeding of tumor in the subcutaneous, epidural, or subdural compartments; and the

J. Neurosurg. / Volume 80 / May, 1994 865

D. L. Peterson, R J. Sheridan, and W. E. Brown, Jr.

tumor should be able to grow in culture and must be safe for laboratory personnel. 2~ Most importantly, the therapeutic responsiveness of the model must imitate that of human brain tumors) > Although many avail- able animal models have some utility, none has fulfilled all of these requirements.

Mammalian CNS Tumors

Spontaneous tumors of the central nervous system (CNS) occur in nonhuman mammals at a frequency thought to approximate that of humans. The incidence of CNS tumors in dogs is considered to be about 0.1% to 0.5%. 82 Roughly 60% of these tumors are neuroectodermal, with about 15% being meningeal. The incidence of tumors has been found to be somewhat higher in brachycephalic breeds such as boxers and bulldogs. 75 In contrast, approximately 80% of feline primary CNS tumors are of meningeal origin, s2 Pri- mary CNS tumors appear to be quite rare in nonhuman primates. 122 Interestingly, an autopsy series of 14,000 rhesus monkeys revealed no CNS tumors] ~-3 and another study of 1247 aged Old World monkeys (baboons and Macaque monkeys) found no CNS tumors. 62 Spon- taneous CNS tumors are also exceedingly rare in laboratory rodents. Guerin autopsied 16,500 rats and found 567 tumors, among which only four were primary CNS tumors, and Mawdesley-Thomas and Newman 8~ reported an autopsy series of 41,000 Sprague-Dawley rats with 38 brain tumors, an incidence of 0.09%; approximately one-third were meningiomas, one-third were gliomas, and the remainder were poorly differentiated or sarcomatous. In a study by Fitzgerald, et al.)8 involving autopsy of 7803 Sprague-Dawley rats, the incidence of CNS tumors was 0.44%.

The most frequently arising spontaneous astrocytic animal tumor, the VM murine astrocytoma, was first reported in 1971. 39 These tumors were discovered as an incidental finding during the examination of over 10,000 mouse brains used in experimental scrapie research in an inbred mouse strain. The incidence of astrocytic tumors in this line is only 1.5% by 600 days of age, making it impractical for therapeutic studies. Cell lines derived from tumors in this strain have been well characterized, and have allowed the development of a useful transplantation model.

Carcinogen-Induced Tumors

The experimental production of skin cancer in laboratory animals with coal tar in 1915 initiated the study of chemical structure and its relationship to carcinogenesis. 114 It also led to the evaluation of tissue-specific susceptibility of different carcinogens. The first studies directed at brain oncogenesis involved a subarachnoid injection of a tar emulsion in the rabbit, but no brain tumors developed. The first successful brain tumor induction was reported by Seligman and Shear 114 in 1939. They implanted pellets of methylcholanthrene into the brains of 20 mice; gliomas developed in 11, meningeal fibrosarcomas in two, and seven had no tumors. The growth characteristics of these tumors were sub-

sequently evaluated by subcutaneous passage) 31 Zim- merman ~3" reported on a study by Cohn, who grew fragments of a methylcholanthrene-induced ependymoma in chick eggs. After several passages, the tumors became an amorphous mass of cells with a morphological appearance that bore no resemblance to the original glial neoplasm. Interestingly, the tumor morphology re- turned to that of an ependymoma after subcutaneous transplantation from chick egg to mouse) 3~ This finding brought attention to the influence that the environment of a tumor has on its microscopic structure.

In the 1960% a new class of synthetic carcinogen, the N-nitrosoureas, was developed. The incidence, type, and location of N-nitrosourea-induced tumors was found to vary greatly with dose, age, strain, and physiological state of the animal, such as pregnancy or malnutrition. Druckrey, et at., 31 evaluated the effects of ethylnitrosourea (ENU) on laboratory rats, and found that the oncogenic effect of ENU was much greater in younger rats as compared to older rats. This observation led to the injection of ENU into pregnant rats, which demonstrated the transplacental oncogenicity of ENU. Schmidek, et a l l ~8 performed 36 weekly injections of N-nitrosomethylurea (NNMU) into 108 male rats. Of the 58 animals that survived the injections, 22 developed primary brain tumors between 17 and 43 weeks after termination of NNMU administration. A spectrum of glial tumors resulted, but the low incidence of tumors (22 in 108 animals) invalidated this model for therapeutic studies. Koestner, et at., 67 administered a single intravenous ENU dose of 50 mg/kg to Sprague-Dawley rats on the 20th day of gestation. All 25 offspring developed tumors of the nervous system, with a total of 102 neural and seven non- neural tumors. Classification of the tumors by light and electron microscopy revealed 32 oligodendrogliomas, four astrocytomas, 19 mixed gliomas, five anaplastic gliomas, 14 ependymomas, one meningioma, and 27 neurinomas. Brucher and ErmeP 9 were able to demonstrate induction of neuroblastoma in Wistar rats by transplacental administration of NNMU. The advan- tage of tumors induced transplacentally is that they may be obtained with a single injection; the result is a litter composed of several tumor-bearing animals of the same age. These and similar studies provided reliable and reproducible tumors of neural origin for morphological and biological investigation, xenografts, and colony-formation assays. 129 Disadvantages of the NNMU and ENU animal models include extensive variability in tumor type, location, induction times, and malignancy characteristics.

Virus-Induced Thmors

The role of viruses in the induction of human neural tumors has been vigorously investigated. Although in- tact virus particles have been found in human brain tumors, z9 an etiological relationship has never been established. With regard to animal models, a host of CNS tumors have been induced by oncogenic viruses. The observation that viruses that are nontumorigenic in their permissive host can become tumorigenic in a non-

866 J. Neurosurg. / Volume 80 / May, 1994

Animal models for brain tumors

permissive host (when inoculated at high titers) has led to many models of CNS neoplasia. Virus-induced tumors have a number of advantages: the inoculation pro- duces minimal trauma to brain tissue; an autochthonous tumor develops with a natural blood supply; the tumors can be cultured for in vitro studies; and these tumors may be predictably induced. There are disadvantages, however: the viruses can be difficult to maintain in the laboratory; they may be harmful to man; and the resultant tumors vary greatly in histopathology, growth kinetics, and location.

Ribonucleic Acid Viruses

The most commonly used viruses in brain tumor induction are retroviruses, particularly the avian sarcoma viruses. 2~ As early as 1936, V~izquez-L6pez t24 induced intracerebral sarcoma in adult chickens with intracerebral inoculation of Rous sarcoma virus (RSV)-I, an avian sarcoma virus first discovered in 1911J ~ The type of tumor induced by viruses depends largely on the site of inoculation and the age of the animal at the time of inoculation. A superficial cerebral or vermian location of inoculation favors sarcomas, whereas inoculation in the subependymal plate region favors glioma development. 27 Neonatal animals are much more susceptible to tumor development and progression than adult animals. 26

An early model of experimental brain tumors was developed in the dog using intracerebral injection of the Schmidt-Ruppin strain of the RSV. 7,~~176 This system was so refined by Bigner, et al., 6 that a 0.01-ml concentrated suspension of virus particles was able to produce 100% tumor induction in newborn beagles. Tumors derived from various glial precursors and intracranial sarcomas resulted. When tested against chemotherapeutic agents, the RSV beagle model was responsive to BCNU (carmustine) and CCNU (lomus- tine), but not cyclophosphamide, which correlates with clinical experience. 29 This work also revealed a substantial amount of information regarding ultrastruc- tural aspects of brain tumors, including identification and characterization of blood-brain barrier defects, en- dothelial disruption, and perivascular tumor spreadJ 25 No virions were found in mammalian hosts, demonstrating that biologically detectable replication of the RSV need not be present in virus-induced tumors, and that viral oncogenesis can occur independent of par- ticle replication. This tumor model is typical of a nonpermissive host system, and in every case the RSV genome could be rescued by inoculation of mammalian tumors into the permissive avian host.

An alternative method of viral brain tumor induction was presented by Tabuchi, e t a l . , 119 in 1985. They in- fected fibrohlasts derived from chick embryos with the Schmidt-Ruppin strain of RSV, then inoculated the fibroblasts into the right frontal lobe of Japanese monkeys (Macaca fascata). Tumors were induced after a latency period of 15 to 67 days in 73% of the experimental animals. It was concluded that the inoculated chick embryo fibroblasts regressed early after implantation and infection of brain cells. This conclusion was

supported by two experimental findings: 1) RSV particles were not seen under electron microscopy, a finding consistent with incorporation into a nonpermissive host (in this case, the monkey); and 2) chromosomal analysis of these tumors subsequently grown in culture revealed a diploid number of 42, indicating that the origin of the tumor cells was from the monkey rather than the chick. Unfortunately, 66% of the tumors examined immunohistochemically proved to be sarcomas.

Deoxyribonucleic Acid Viruses

Human adenovirus 12 has been used to induce tumors in mice, rats, and hamsters after intracerebral and intraorbital inoculation. 8s Neuroblastoma and retinoblastoma are the primary tumor types induced, and per- manent cell lines have been established from both of these tumors. The survival of animals bearing tumors induced by this virus is variable, and these models have proven most useful in cytokinetic studies. Simian adenovirus (SA7) in the hamster induces choroid plexus papillomas and medulloblastomas; the latter have been maintained in culture. 24,~5 The chicken-embryo-lethal- orphan (CELO) virus is an endogenous adenovirus of chicken eggs, whose infection is inapparent in its permissive host. Mancini and colleagues 76 inoculated intracerebrally a 0.02-ml suspension of purified CELO virus into newborn Syrian hamsters. All 17 animals that survived the inoculation developed CNS sympto- tautology and tumors arising from ependymal sur- faces. These tumors were initially thought to be papillary ependymomas, but choroid plexus papillomas seem a more likely histopathology. 2~

Some animal brain tumor models have arisen from apparently unrelated research. Papovaviruses isolated from brain tissue of humans afflicted with progressive multifocal leukoencephalopathy (PML) have been shown to be tumorigenic when inoculated intracerebrally in newborn Syrian hamsters. 91,9z Tumor types induced include cerebellar medulloblastomas, thalamic gliomas, intraventricular papillary ependymomas, a meningeal tumor, and a pineocytoma. London, et al. ]3 inoculated JC polyomavirus derived from human cases of PML into four Colombian owl monkeys; primary brain tumors developed at 16 and 25 months in two of the animals. One tumor was a high-grade glioma, resembling human glioblastoma, and the other con- tained a mixture of neoplastic glial and neuronal cell types. The simian vacuolating virus SV40 was isolated from monkeys and is not oncogenic in monkeys, its permissive host. 118 It has been shown to produce dose- dependent tumorigenesis when inoculated intracerebrally in newborn Syrian hamsters; 32,33 these tumors were interpreted histopathologically as choroid plexus papillomas or ependymomas. Inoculation of SV40 did not induce tumors in young dogs, rats, cats, or rabbits. The transforming genes from SV40, namely, the large T- and small t-antigens, have been sequenced and cloned. The large T-antigen gene has been used extort-

J. Neurosurg. / Volume80 / May, 1994 867

D. L. Peterson, E J. Sheridan, and W. E. Brown, Jr.

sively to effect malignant transformation of cells in culture 117 as well as to direct oncogenesis in vivo, as will be discussed later.

Models Based on Cell and Tissue Transplantat ion

Transplantation of tumor cells or tissue into recipient animals has generated numerous brain tumor models. As discussed by Bullard and Bigner, 2~ there are essentially two types of transplant models: syngeneic tumors induced by viruses or chemicals and maintained via cell culture or subcutaneous passage for prolonged periods, and heterotransplantation models in which tumor from another species (or a different strain of the same species) is transplanted into either an immunoincompetent recipient or an immunologically privileged site, such as the cheek pouch, brain, or the anterior chamber of the eye? 8

Transplantation techniques have evolved tremen- dously. Early workers used a needle and stylet to inject minced tumor fragments freehand. This imprecise and nonquantitative method was improved upon by Wilson, et aL, ~27 who achieved a low mortality rate using stereotactic implantation of quantitatively standardized cell suspensions from a cell culture. Further improve- ments in implantation technique, including better stereotactic localization, concentrated cell suspensions (104 to 107 cells/10/xl), improved injection needles, flushing of the operative field, and augmentation of the cell suspension with 0.5% to 1.0% agar, have resulted in a 90% to 100% yield of intracerebral tumor growth and have reduced the incidence of extracranial and spinal metastases to 0% to 5%. 66

Syngeneic Transplantation Models

In 1941, Zimmerman and Arnold TM transplanted methylcholanthrene-induced ependymoblastoma intracerebrally in mice and obtained 90% to 100% tumor take. From frozen stocks of this original tumor, Aus- man, et al., 1 derived several stable cell lines, including the extensively studied ependymoblastoma A. They transplanted minced tissue intracerebrally from subcu- taneously passaged ependymoblastoma and demonstrated quantitative differences in tumor sensitivity to different drugs, measuring increased survival compared with control tumor-bearing animals. This work greatly stimulated the study of chemotherapeutic agents from brain tumor in animal models. While screening of chemotherapeutic agents against ependymoblastoma was initially promising, 45 its popularity has declined. Two chemotherapeutic agents, CCNU and BCNU, were found to have cure rates in animal models of 80% to 90% and 30% to 40%, 4~ respectively, and procarbazine was similarly effective; however, these drugs rarely produce cures in man. This tumor was shown to be exquisitely sensitive to dianhydro- galactitol in experimental studies, 69 while results with this drug in human studies have been disappointing. The prevailing opinion is that this model overpredicts the activity of certain agents. 7~ Furthermore, human ependymoblastomas account for less than 1.0% of all

intracranial tumors, and the histology and kinetics of murine ependymoblastoma are not comparable to human tumors.

The C6 line of tumor cells has been used extensively in transplantation-based glioma studies. The origin of this cell line remains somewhat in question. It was induced in rats through transplacental exposure to N-nitrosourea; however, the particular strain of rat is not precisely documented. The cell line is originally men- tioned by Benda, et al.,4 but the strain is not described. Although Dr. Benda is deceased, a coauthor recalls that the rat strain was not inbred (G Sato, personal com- munication, 1993). The conventional wisdom is that the strain of origin of the C6 cell line is randomly bred Wistar rats. 5,112

The C6 cell line demonstrates astrocytic morphology and has been widely studied in neurobiology because of the presence of glial cell markers such as glial fibrillary acidic protein (GFAP) and S-100 protein. It has been used extensively in vitro, and numerous investigators have published in vivo transplantation models using C6 cells as techniques have become more refined. 3v Kaye, et aL, 64 showed that 106 C6 cells sus- pended in 1.0% agarose will grow intracranially as a xenograft model in adult mice of several strains. They used a monoclonal antibody to C6 to show conclusively that microscopic invasion of tumor cells occurred into surrounding brain. Numerous investigators have reported models relying on C6 cells grown intracerebrally in adult rats, and it appears that C6 cells grown in adult Wistar rats show more homology to human glioma (with parenchymal invasion, neovascularity, and necrosis) than C6 cells grown in other strains, including Sprague-Dawley and Long-Evans rats, where they grow as sharply demarcated, encapsulated tumors.25,37,107

The 9L gliosarcoma cell line was developed from tumors induced in Wistar and Fischer rats injected with NNMU. Several clonal tumor cell lines originated and were reported to have glial morphology with stable growth patterns and biological characteristics? ,u2 Sar- comatous change occurred with serial passage. 3 While it may not represent glioma as well as other lines, the 9L cell line has been used extensively in screening tests in vitro, including the colony formation assay, and has been used as a model for spinal axis tumor dissemi- nation, ms Additionally, the 9L cell line has been used extensively to evaluate the effect of various agents on the blood-brain barrier. This cell line has lost its ability to invade the parenchyma of the brain, and therefore grows mainly in the Virchow-Robin spaces. This growth pattern may be due to the inability of the cells to effect neovascularity, to produce enzymes that re- model the extracellular matrix, or to produce a solid matrix. Schwartz, et aL, 113 used these characteristics of 9L cells as control features to evaluate the in vitro and in vivo effects of a variety of retrovirally transfected oncogenes (ras, neu, and sis) on the 9L line. The 9L/ras and 9L/neu cells produced more rapidly growing, large, solid, highly vascularized tumors in vivo than did non- transformed 9L cells. The 9L/sis cells produced tumors

868 d. Neurosurg. / Volume 80 / May, 1994


intermediate between control and 9L/ras and 9L/neu cells. Transfection of these oncogenes did not alter the growth characteristics of 9L cells in vitro, as compared with controls; this implies that the tumor progression was not merely a result of accelerated cellular metabolism. The authors suggested that the reason for tumor progression is that the oncogene-transfected cells are capable of enhanced angiogenesis or increased responsiveness to local growth factors. The question of whether these oncogenes enabled the 9L cells to invade the parenchyma was cleverly addressed. The 9L/oncogene cells were cotransfected with the prokaryotic .gene for/3-galactosidase, an enzyme in which activity is marked by blue staining in the presence of the appropriate substrate X-gal (5-bromo-4-chloro-3-indoyl- /3-galactoside). The 9L cells that expressed the foreign oncogene in addition to the /3-galactosidase activity were then co-injected with C6 cells. The resultant tumors showed blue staining in the center, with a non- staining infiltrating surround, and demonstrated that the expression of these oncogenes did not alter the non- infiltrative growth pattern of the 9L cells.

The BT4A cell line was initiated in BD IX fetal rat brains by transplacental exposure to ENU; the resultant tumor cell line was then stabilized by serial passage in these syngeneic animals. This cell line provides 100% tumor take in BD IX rats and has the properties of single-cell invasion, the presence of S-100 protein, and histological characteristics in common with differentiated oligodendroglioma, ss Additionally, because the BT4A cell line shows low immunogenicity and moderate sensitivity to chemotherapy, it is a promising line for further studies. 83 Other similar lines include the RG22~ and the T9 s rat glioma.

Bradford and colleagues 13-~5 have worked extensively with the spontaneous VM murine astrocytoma. From an initial cell line described by Serano, et al., 115 they isolated and characterized the 497-P(1) cell line. These cells stained for GFAP and other glial markers in vitro, and have been shown to be highly tumorigenic following implantation, with the tumors developing necrosis, pseudopallisading, and local invasion. Median survival time of the VM mouse model following transplantation was 12.5 days. Bradford, et al., TM assessed the sensitivity of VM 497-P(1) intracerebral tumors to BCNU, CCNU, vincristine, and procarbazine. While treatment with BCNU and CCNU showed substantial increases in life expectancy (142.5% and 132.5%, respectively), no cures were obtained unlike the methylcholanthrene-induced ependymoblastoma model. Vin- cristine produced a moderate response, and procarbazine was ineffective. The results with procarbazine in this model do not parallel its activity against human gliomas, a discrepancy seen in many tumor models. Wilson 126 suggested in 1978 that the failure of tumors in some animal models, particularly rat models, to re- spond to procarbazine was due to differences in the metabolism of the drug in that animal. The ineffective- ness of procarbazine against the murine 497-P(1) tumor model is thought to be due to inherent cellular re- sistance of the 497-P(1) cells.

Heterologous Transplantation Models

I-teterologous transplantation of human brain tumors was essentially unsuccessful until the advent of the nude mouse, a mutant that is genetically thymus- deficient and therefore immunoincompetent. Early workers showed minimal success with transplantation of brain tumors into the immunologicallyprivileged anterior chamber of the eye in guinea pigs 4~ or into immunologically suppressed animals such as corti- sone-treated rabbits. 48 The nude mouse was first reported by Pantelouris 96 in 1968, and since then a plethora of human tumors have been transplanted in athymic mice, with variable success. Tumors grown in this way frequently retain many of the biological characteristics of the original biopsy specimens including histopathology, biochemical markers, and hormone production. 63

In 1977, Rana, et al., t~ injected one human glioblastoma into three athymic mice. In each mouse, the tumor grew and developed the morphological traits of the original specimen: multinucleated giant cells and pseudopallisading around areas of necrosis. Subsequently, several authors have reported results with human brain tumor xenografts in nude mice. Bradley, et al.,17 successfully grew subcutaneous xenografts from 11 (44%) of 25 adult gliomas, and other authors have reported higher success rates. Shapiro, et al., 116 placed tissue fragments from seven anaplastic gliomas into nude mice. All of the tumors grew initially, but only two resulted in stable tumor cell lines; both of the serially transplantable cell lines were identified histopathologically as gliosarcomas. Jones, et al., ~3 reported a 94% success rate with the transplantation of anaplastic astrocytomas into nude mice; 16 of 17 xenografts produced a stable tumor cell line. Latency periods from transplantation to initial tumor growth varied from 24 to 364 days. Morphologically, most of these tumors showed cellularity greater than that of the original specimens, and virtually all of them showed a general absence of vascular proliferation. The vascular tissue in these tumors is considered to be of mouse origin, and the difference in species is thought to account for this lack of neovascularity. Regarding the biological characteristics of these transplanted tumors, cellularity, growth rate, and necrosis were either "retained, changed or deleted in an unpredictable way. ''63 Some tumors exhibited increased GFAP expression, some maintained GFAP expression similar to the original tumor, and some lost GFAP expression with heterotransplantation. Evaluation of the chromosomal content of malignant human glioma xenografts in nude mice gen- erally showed karyotypes similar to the original tissue, although some minor changes were seen. 1~ Schold and colleagues 22'1~ have extensively characterized the therapeutic responses of human glioma lines in athymic mice, and have found variable correlations with efficacy in clinical trials. The morphological, genetic, and biological differences in heterotransplanted tumors may be due to normal tumor evolution or augmented selection of certain cell subpopulations in an artificial environment. As Bullard and Bigner 2~ state, "(the) na-

Z Neurosurg. / Volume 80 / May, 1994 869

D. L. Peterson, P. J. Sheridan, and W. E. Brown, Jr.

ture of what determines tumorigenicity in the nude mouse is not clearly dcfincd and may reprcscnt factors not pertinent to human tumors."

Other tumor types grown in nude mice include me- . 4 1 , ~ 2 5 4 - , " . ' : ' l " dulloblastoma �9 and clanlopharyngxoma.- Fried-

man, et al., 42 reported on a variety of medulloblastoma cell lines grown both intracerebrally and subcutane- ously in nude mice, and showed prolonged median survival times when tumor-bearing mice were treated with melphalan, cyclophosphamide, iphosphamide, and thiotepa, while CCNU and busulfan were ineffective. Studies such as these bear strong implications for clinical trials.

While transplant models have become the mainstay of in rive brain tumor research, it is important to realize that they have several inherent limitationsJ ~2 Human brain tumors are extremely heterogeneous, and clonal cell lines do not represent this tendency. Tumor cell lines maintained in culture can, with serial passage, show variability in mitotic rates and transformation potential, and frequently demonstrate sarcomatous degeneration. Direct transplantation from animal to animal somewhat minimizes this variability, although work with the nude mouse has shown that serial subcutaneous passage commonly results in altered tumor morphologies. Variability persists in even the most carefully controlled clonal transplant studies. Unfor- tunately, despite stereotactic implantation and injection of consistent numbers of cells, survival time may be influenced by errors in technique, such as injection of dead or insufficient cells or seeding of extracranial sites. For these reasons, the search for newer and better models continues.

Transgenic Models

The ability to introduce specific genes into the germ lines of mammals is one of the major recent techno- logical advances in molecular biology. Jaenisch and Mintz, 6a in 1974, first reported experiments aimed at inserting foreign genes into mammals; they injected SV40 deoxyribonucleic acid (DNA) directly into the blastocyst cavity of developing mouse embryos. The resultant animals showed a mosaic pattern of SV40 DNA integration; however, the germ line did not demonstrate SV40 DNA. Other marginally successful el- forts at genome transfer experiments included embryo cell mixing and nucleus transfer. 4~i,57,sl Subsequently, infection of mouse embryos with retroviruses encoding foreign genes became possible ~9 and the standard means of introducing foreign genes into embryos has become direct microinjection.

Microinjection has been made possible by the tremendous advances in DNA cloning technology, It is the most widely used and most successful means of generating animals with integrated foreign genes, termed "transgenic animals." By far the most commonly utilized species for transgenic experiments is the mouse, although transgenic pigs, rabbits, sheep, and rats have been produced. 5~ The highly detailed techniques of transgene microinjection have been published previously, 52,5~,93,97 but a brief review is war-

ranted. Fertilized e,,os~,~,, from timed malings are harvested from the oviducts of pregnant females, and briefly subjected to hyaluronidase to remove cumulus cells, yielding single-cell embryos at the pronuclear stage of development. Using micromanipulators, the male pronucleus is injected with roughly 500 linear copies of purified DNA coding for the gene of interest. These eggs are then transferred microsurgically into the oviducts of females made pseudopregnant by mating with vasectomized males. The resultant offspring are born after a 19-day gestation. The foreign DNA is usu- ally incorporated into the chromosome at random breaks, with multiple copies arranged in a head-to-tail fashion. If the foreign gone has undergone stable chromosomal integration at the one-cell stage, the transgenic DNA will be present in every cell at the adult mouse. These animals, referred to as "transgenic founders," are identified by screening for the presence of the foreign gone in DNA from a segment of the tail. Transgenic founders may then be bred to give rise to progeny that will inherit the transgene in a stable, men- delian fashion.

The success rate in developing transgenic animals is relatively poor. Microinjected transgenes integrate into the genome at random chromosomal breaks and, in the best of circumstances, the rate of incorporation in one-cell embryos approaches 20%. Only one-half of all transgenes successfully incorporated are expressed; those that are unexpressed are likely incorporated into translationally silent areas of genomic DNA. Occa- sionally, the chromosomal position of transgene insertion may result in expression of the transgene in unexpected tissues. Because of these factors, even in the best-designed experiment, only about 5% of offspring bear the transgene and express it in an appropriate tissue-specific fashion, s2

Early experiments with transgenic animals found that the crucial problem in their utility was the extremely low or variable expression of the transgene products; this stimulated investigation into the biology of gene expression. It was soon realized that inclusion of prokaryotic vector sequences, necessary for clonal expansion of the foreign gene of interest in bacteria, inhibited transgene expression. 54 The importance of DNA elements that regulate gone expression, namely, the promoter and enhancer sequences, came under a great deal of scrutiny. It was found that the foreign DNA insert must contain a strong promoter sequence, which optimizes the potential for expression of the transgene. Tissue-specific enhancers are DNA sequences whose presence near a gone of interest greatly enhances the expression of that gene. Enhancers may be upstream, downstream, or located in intervening nontranscribed segments of the gene (introns), and may be relatively short, such as the 72 base pair enhancer in the SV40 genome, s6 or substantially longer. They may be located immediately flanking the gene of interest or, as in the case of albumen, an element located 10 kb upstream of the promoter is required for liver-specific expression. 99 Appropriate promoter and enhancer sequences are required in the transgenic

870 J. Neurosurg. / Volume 80 / Ma~; 1994


construct in order to enable tissue- or cell-specific expression of the transgenc. One of the great remaining mysteries of molecular biology is the role and function of introns, the nontranscribed intervening sequences present in eukaryotic genes. Transgenic studies have shown that the incorporation of genomic sequences, including introns, gives a much higher expression of transgene DNA than does complementary DNA which is composed of coding sequences only and is derived from messenger ribonucleic acid. 74>J5

The types of transgenic models for CNS tumors have evolved over the last 10 or so years, and malay of the published models have arisen somewhat serendipi- tously. The first reported models utilized the expression of the SV40 early region genes, including the large T- and small t-antigens. Surprisingly, when the SV40 large T-antigen gene, regulated by its own promoter and enhancer sequences, is placed in transgenic mice, tumors of the choroid plexus reproducibly develop.~<79.~4 The large T-antigen was subsequently shown to induce transformation of cells in a tissue-specific manner, as directed by tissue-specific transcriptional control elements. Using promoters for the genes for insulin, s~ elastase, 9~ and albumin, 93 T-antigen expression has been used to direct tissue-specific tumor development in the pancreas sl and liver. The large T-antigen expression under the control of the promoter for the murine tyrosi- nase gene, necessary for melanin processing, induces ocular and cutaneous melanomas in transgenic miceJ a

In the CNS, a variety of neural cell types have been transformed in transgenic experiments using T-antigen expression. O'Brien, et al., s~ and Marcus, et al., TM utilized the SV40 large T-antigen fused to the promoter for the/3-subunit of luteinizing hormone and generated pituitary adenomas. A single founder, however, developed bilateral ocular malignancies with the histopathological appearance of retinoblastoma, and these tumors proved to be heritable in 90% of heterozygous transgene-bearing offspring. Retinoblastoma is the most common form of ocular malignancy in childhood and has both hereditary and nonhereditary forms. In all retinoblastoma cases examined, both alleles of the retinoblastoma gone, located on chromosome 13g14, are mutated. 71 The retinoblastoma protein product (pl05- RB) has been designated an anti-oncogene, and binds a variety of viral oncoproteins, including the SV40 T- antigen. This binding of the retinoblastoma gene product is thought to release the cell cycle from negative control, resulting in malignancy. The development of ocular malignancy in this model is surprising, but many investigators have found unexpected patterns of expression of the transgenesJ 2,93 We postulate that some random site of chromosomal integration resulted in placing the T-antigen under stronger, retinal-specific regulatory control, thus initiating retinal-specific tumorigenesis. Interestingly, 27% of these animals developed midline pineal region neoplasms with a pathological appearance consistent with neuroblastoma.

The unpredictable in vivo response of T-antigen expression under nonspecific control led to other fortu- itous discoveries. Gotz and colleagues 47,tz~ have devel-

opcd a model of transgenic mice expressing the large T-antigen under the control of the Moloney murine sarcoma virus cnhancer and the SV40 promoter. Midline brain neoplasms uniformly grew in these animals by 4.5 months of age. When evaluated histologically, the neoplasms were characterized as primitive neuroectodermal tumors, composed of small cells with scanty cytoplasm. To better assess if these midline neoplasms correlate with human primitive neuroectodermal tumors, they performed extensive immunocytochemical analysis. The presence of S-antigen and rhodopsin in these tumors showed them to be comparable to human pineal tumors and to a certain type of medulloblastoma displaying photoreceptor-like features. ~ The absence of GFAP activity in the neoplastic cells and the dis- tribution of neuron-specific enolase and neurofilaments also resemble the pattern in human medulloblastomas. The reason for the selective effectiveness of this transgene in pineal tissue is unknown.

Another surprising finding was reported by Perraud, et al., ~)~ who characterized the promoter sequence for the human cystic fibrosis (CF) transmembrane conductance regulator (CFTR), the gene mutated in CE To assess the in vivo potential for oncogenesis linked to tissue-specific activity of this promoter, they fused it with the SV40 T-antigen. The resulting transgenic mice developed a typical phenotype, with a bulging cranium and hydrocephalus due to intraventricular ependymomas. All animals died at a young age, suggesting that the hydrocephalus was present in utero. No pathological lesions were found in the lungs or pancreas, where human CF is primarily manifest. Likewise, T-antigen expression was not identified in lung or pancreatic tissue using Northern blot analysis. These findings imply that CFTR is important in the brain development of the mouse, but several variables prevent strong con- clusions. The human CFTR promoter may be more ac- tive in ependymal cells in the mouse than in the lung or pancreas, and may lack the regulatory sequences required for expression in the usual organs. In addition, ependymal precursors may be inherently more sensitive to SV40 T-antigen transformation. The facts that CFTR is a CI- channel and that ependymal cells are involved in ion and protein exchange do, however, sug- gest a role for the CFTR protein in these cells.

The postmitotic nature of mature, differentiated neurons has impeded neuronal study in vitro. Transgenic methodology has enabled the development of immortalized cell lines of neural origin. Mellon, et aL, ~4 reported neurally targeted transgenic tumorigenesis using the T-antigen fused to the promoter region of the gene for gonadotropin-releasing hormone (GnRH). Two of nine transgenic founders developed anterior hypothalamic tumors. One tumor was successfully cultured into a clonal differentiated GnRH-secreting neuronal cell line that is immortal. These cells express neuronal markers including neuron-specific enolase, the synaptic membrane marker VAMP-2, and neuron- specific neurofilament proteins. They lack glial markers such as GFAP, myelin basic protein, and myelin proteolipid protein. The neurosecretory function of

J. Neurosurg. / Volume 80 / May, 1994 87i

D. L. Peterson, P. J. Sheridan, and W. E. Brown, Jr.

these cells appears limited to GnRH; its release is stimulated by depolarization and norepinephrine and inhibited by tetrodotoxin, demonstrating the presence of fast Na + channels. Morphologically, these cells ex- hibit neuronal characteristics, with spontaneous exten- sion of neurites that end in growth cones or contact distant cells. Unfortunately, the gonads and accessory sex organs in these animals were underdeveloped, and they were infertile, suggesting that transgene expression interfered with sexual maturation.

Cell lines of the pituitary gonadotroph lineage have also been derived by targeted oncogenesis in transgenic mice. Windle, et aL, 12s developed a construct coding for 1.8 kb of the 5' flanking region of the gene for the human glycoprotein a-subunit linked to the SV40 T- antigen. More than one-half of the animals developed anterior pituitary tumors, and several stable cell lines were derived from these tumors. Windle, etal., antici- pated that the resulting cell lines would represent gonadotrophs or thyrotrophs; interestingly, the tumors and cultured cells expressed only the a-subunit and did not express any of the/3-subunits. Several of the cell lines expressed the a-subunit in response to GnRH, but not thyrotropin-releasing hormone. It is thought that, as these animals develop with T-antigen expression in stem cells, the cells destined to become gonadotrophs or thyrotrophs are transformed before complete differentiation. The authors postulated that r expression occurs only in completely differentiated anterior pituitary cells. This correlates with the previous observation that a-subunit expression occurs in development before luteinizing hormone-/3 expression. It is interest- ing that a common type of pituitary tumor secretes only the a-subunit and is presumed to develop long after full differentiation of the anterior pituitary. This model therefore provides evidence for the concept of transformation-mediated dedifferentiation.

The regulatory sequences of the phenylethanolamine N-methyltransferase gene were utilized by Hammang, et aL, 49 to direct T-antigen-related neurosecretory tumorigenesis. These animals developed bilateral retinal tumors as well as adrenal pheochromocytomas by 12 weeks of age. A line of immortalized retinal neurons was derived from this model that demonstrate retinal neuron-specific markers, as well as neurofilaments. 2 To date, there are no reports of T-antigen-transformed de- paminergic neurons, but one could postulate the value of such a line in the experimental study of Parkinson's disease.

The SV40 antigen-related models have provided a great deal of insight into regulatory element specificity and tumorigenesis in general; however, molecular genetic analysis of brain tumors provides a much more complicated story. An ever-increasing number of proto-oncogenes have been identified in brain tumors. Amplification of oncogenes, s alterations in ploidy, chromosomal deletions, 61 substitutions, double min- utes, 9 and tumor suppressor gene mutations are among the complex genetic alterations seen. Some of the specific oncogenes amplified in malignant gliomas are neu, N-myc, 43 C-myc, 3s gli, al and C-erb-~3 (epidermal

growth factor receptor). 11,34,56 Mutations of the tumor suppressor p53 are well documented, 44 and other tumor suppressor genes are postulated to exist on chromosomes 10 and 22. 35,61.121

Intuitively, models expressing tumor-associated oncogenes in a tissue-specific fashion may provide more insight into the complex process of oncogenesis than the T-antigen models. Many transgenic models have used tissue-specific expression of tumor-related oncogenes to assess their role in tumorigenesis. The work of Pattengale, et al.,97 in the field of mammary tumorigenesis is the most well known; they constructed transgenes of the neu, ras, and myc oncogenes driven by the murine mammary tumor virus (MuMTV) promoter, and developed highly reproducible murine mammary tumors. Using different lineages derived from each transgenic construct, they demonstrated that the genotype would predict the phenotype of the resultant mammary tumor. 23 Histopathologically, the tumors could be divided into three categories; small cell, large cell, and intermediate cell. The tumors for the MulVlTV/ras lineage were all small cell, the MuMTV/myc tumors were mainly composed of large cells, and the cells of the MuMTV/neu line were of intermediate size.

Unfortunately, such extensive work using glial- directed expression of oncogenes has not yet been reported; however, glial-specific regulatory sequences have been identified and utilized in transgenic studies. Mucke, et aL,87 reported a transgenic mouse model that shed light on the astrocytic response to focal trauma. The murine GFAP promoter was fused to the bacterial reporter gene/3-galactosidase. The/3-galactosidase activity can be readily detected in tissue sections and has no harmful effects in vivo. This model showed baseline GFAP activity in the hippocampus, glial limitans, and along white-matter tracts including the optic nerves, and GFAP expression was proven to be induced within 1 hour of focal injury (needlestick in right frontal lobe). Our laboratory is currently developing transgenic lines that express the temperature-sensitive mutant p53 tumor suppressor gene N-myc and the epidermal growth factor receptor gene under the control of the GFAP promoter. Myelin basic protein (MBP) regulatory sequences have been studied in transgenic models, and two groups have published models demonstrating correction of the phenotype of the shiverer mouse. 65,1~ The homozygous shiverer (shi/shi) mouse is an auto- somal recessive mutant characterized by the almost total lack of CNS myelin and all MBP's. By 2 weeks of age, the animals show the neurological dysfunction of shivering, develop tonic convulsions, and die at an early age. Readhead, et aL, TM and Kimura, etaL, 65 have published models demonstrating that introduction of the gene for MBP in homozygous shiverer embryos results in substantial resolution of their neurological phenotype, although MBP production was only 25% to 50% of that in wild-type mice.

The only published transgenic model for glial neoplasia was reported by Hayes, et al. 53 In this excellent study, the transgene construct was composed of the neu oncogene under the control of the MBP transcriptional

872 J. Neurosurg. / Volume 80 / May, 1994


regulatory elements, with the goal of obtaining transformed oligodendrocytes. The neu oncogene is a membrane protein that belongs to the tyrosine kinase family of oncogenes (which includes the epidermal growth factor receptor homolog, C-erb-[3). Of note, the neu (.~acogene was originally isolated and cloned from a series of rat glioblastoma cell lines induced by ENU? 12 The MBP/neu animals displayed a low incidence of tumor development, but tumors from three animals were examined histologically and immunocytochemi- cally. The pathological appearance of the tumors was that of a high-grade glioma, with necrosis, giant cells, and fine eosinophilic processes. None of the tmnors exhibited the classic histopathological features of oligodendrogliomas. One tumor provided a stable cell line in culture; these cells, when cultured in variable media concentrations, may resemble either type 2 astrocytes or oligodendrocytes that result from the differentiation of the O-2A bipotential progenitor cell. The O-2A cells differentiate into oligodendrocytes constitutively, and into astrocytes in the presence of high serum concentrations. 72,m~ The MBP/neu cells will express oligodendrocyte-specific antigens and appear morphologically similar to oligodendrocytes when cultured in media that favors oligodendrocyte survival. When cultured in high serum concentrations, they express abun- dant GFAP as well as MBP and do not react with oligodendrocyte-specific antibodies. The retained MBP expression in these cells implies that they were transformed farther along the oligodendrocyte lineage than at the level of the O-2A progenitor, These cells hold promise for further studies in that they express sig- nificant amounts of MBE We speculate that, by modu- lating the transforming effecl of neu, these cells may be able 1o myelinate axons in culture.

The transgenic models described here have not yet provided opportunities for the evaluation of new adjunctive therapies, such as new chemotherapeutic agents, hyperthermia, or alternative forms of irradiation. They do, however, provide a great deal of insight into the processes of differentiation and oncogenesis. It is evident that one problem with the current models is that stem cells are transformed at an early stage of development, prior to differentiation. Transformation of cells in the adult animal would enhance the utility of the model in therapeutic screening. Future efforts will be directed at identifying tissue-specific regulatory sequences inducible by exogenous substances or con- ditions. An example of an inducible promoter is that for the uniquitous protein metallothionein, whose promoter can be induced by metals or glucocorticoids. The identification of a glial-specific inducible promoter would allow more controlled oncogenesis. Another problem with transgenic models is that levels of transgene expression vary greatly from line to line, and even among siblings. This may be due to the site of chromosomal insertion, DNA methylation, heterologous re- combination, or other variables that remain unknown. The anatomical site of tumor development is variable, as is survival after tumor development. While a truly valid transgenic model for brain tumors remains far off,

the number of discoveries and breakthroughs to date, serendipitous or not, mandate continued efforts in this endeavor.

References

1. Ausman JI, Shapiro WR, Rail DP: Studies on the chemotherapy of experimental brain tumors: development of an experimental model. Cancer Res 30:2394-2400, 1970

2. Bactgc EE, Behringer RR, Messing A, et al: Transgenic mice express the human phenylethanolamine N-methyl- transfcrase gene in adrenal medulla and retina. Proe Natl Acad Sei USA 85:3648-3652, 1988

3. Barker M, Hoshino T, Gurcay O, et al: Development of an animal brain tumor animal model and its response to therapy with 1,3,bis-(2-chlomethyl)-l-nitrosourea. Can- cer Res 33:976-989, 1973

4. Benda P, Lightbody J, Sate G, et al: Differentiated rat glial strain in tissue culture. Science 161:370-371, 1968

5. Benda P, Someda K, Messer J, et al: Morphological and immunochemical studies of rat glial tumors and clonal strains propagated in culture. J Neurosurg 34:310-323, 1971

6. Bigner DD, Kvedar JP, Vick NA, et al: An improved method of preparing Schmidt-Ruppin Rous sarcoma virus for brain tumor induction in mammals. Neurology 20:413, 1970 (Abstract)

7. Bigner DD, Odom GL, Mahaley MS Jr, et al: Brain tumors induced in dogs by the Schmidt-Ruppin strain of Rous sarcoma virus. Neuropathological and immunological observations. J Neuropathoi Exp Neurol 28:648-680, 1969

8. Bigner SH, Burger PC, Wend AJ, et al: Gene amplification in malignant human gliomas: clinical and histopathologic aspects. J Neuropathol Exp Neurol 47:191-205, 1988

9. Bigner SH, Mark J, Burger PC, et al: Specific chromosomal abnormalities in malignant human gliomas. Cancer Res 48:405-411, 1988

10. Bigner SH, Schold SC, Friedman HS, et al: Chromosomal composition of malignant human gliomas through serial subcutaneous transplantation in athymic mice. Cancer Genet Cytogenet 40:111-120, 1989

11. Bigner SH, Wong AJ, Mark J, et al: Relationship between gene amplification and chromosomal deviations in malignant human gliomas. Cancer Genet Cytogenet 29: 165-170, 1987

12. Botteri FM, van der Punen H, Wong DF, et al: Unexpected thymic hyperplasia in transgenic mice harboring a neuronal promoter fused with simian virus 40 large T antigen. Mol Cell Biol 7:3178-3184, 1987

13. Bradford R, Darling JL, Sier N, et al: The VM model of glioma: preparation of multicellular tumour spheroids (MTS) and their response to chemotherapy. J Neurooncol 9:105-114, 1990

14. Bradford R, Darling JL, Thomas DGT: The chemotherapeutic response of a murine (VM) model of human glioma. Br J Cancer 61:46-50, 1990

15. Bradford R, Darling JL, Thomas DGT: The development of an animal model of glioma for use in experimental neuro-oncology. Br J Neurosnrg 3:197-210, 1989

16. Bradl M, Klein-Szanto A, Porter S, et al: Malignant mela- noma in transgenic mice. Proe Nail Aead Sci USA 88: 164-168, 1991

17. Bradley NJ, Bloom HJG, Davies AJS, et al: Growth of human gliomas in immune-deficient mice: a possible role for pre-clinical therapy studies. Br J Cancer 38:263-272, 1978

18. Brinstcr RL, Chen HY, Messing A, et al: Transgenic mice harboring SV40 T-antigen genes develop characteristic brain tumors. Cell 37:367-379, 1984

J. Neurosurg. / Volume80 /May, 1994 873

D. L. Peterson, E J. Sheridan, and W. E. Brown, Jr.

19. Brucher JM, Ermel AE: Ccntral neuroblastoma induced by transplacental administration of methylnitrosourea in Wis- tar-R rats. An electron microscopic study. J Neurol 208: 1-16, 1974

20. Bullard DE, Bigner DD: Animal models and virus induction of tumors, in Thomas DGT, Graham DI (eds): Brain Tnmours: Scientific Basis, Clinical Investigation and Current Therapy. Boston: Butterworths, 1980, pp 51-84

21. Bullard DE, Bigner DD: Heterotransplantalion of human craniopharyngiomas in athymic "nude" mice. Neurosur- gery 4:308-314, 1979

22. Bullard DE, Schold SC Jr, Bigner SH, et at: Growth and chemotherapeutic response in athymic mice of tumors arising from human glioma-dervied cell lines. J Neuro- pathol Exp Neurol 40:410-427, 1981

23. Cardiff RD, Sinn E, Muller W, et al: Transgenic oncogene mice. Tumor phenotype predicts genotype. Am J Pathol 139:495-501, 1991

24. Chen TF, Mora EC, Morley J Jr: Cultivation of medulloblastoma cells derived from simian adenovirus SA7- induced hamster brain tumor. Cancer Res 35:3566-3570, 1975

25. Cohen JD, Robins HI, Javid MJ, et al: Intracranial C6 glioma model in adult Wistar-Furth rats. J Neurooncol 8: 95-96, 1990 (Letter)

26. Copeland DD, Bigner DD: Influence of age at inoculation on avian oncornavirus-induced tumor incidence, tumor morphology, and postinoculation survival in F344 rats. Cancer Res 37:1657-1661, 1977

27. Copeland DD, Bigner DD: The role of the subependymal plate in avian sarcoma virus brain tumor induction. Com- parison of incipient tumors in neonatal and adult rats. Acta Neuropathol 38:1-6, 1977

28. Crafts D, Wilson CB: Animal models for brain tumors. Natl Cancer Inst Monogr 46:11-17, 1977

29. Cuatico W, Cho JR, Spiegelman S: Molecular evidence of a viral etiology of human CNS tumors. Acta Neurochir 35:149-160, 1976

30. Culver KW, Ram Z, Wallbridge S, et al: In vivo gene transfer with retroviral vector-producer ceils for treatment of experimental brain tumors. Science 256:1550-1552, 1992

31. Druckrey H, Ivankovic S, Preussmann R: Teratogenic and carcinogenic effects in the offspring after single injection of ethylnitrosourea to pregnant rats. Nature 210: 1378-1379, 1966

32. Duffell D, Hinz R, Nelson E: Neoplasms in hamsters induced by simian virus 40. Light and electron microscopic observations. Am J Pathol 45:59-73, 1964

33. Edcly BE: Simian virus 40 (SV-40): an oncogenic virus. Prog Exp Tumor Res 4:1-26, 1964

34. Ekstrand AJ, James CD, Cavenee WK, et al: Genes for epidermal growth factor receptor, transforming growth factor alpha, and epidermal growth factor and their expression in human gliomas in vivo. Cancer Res 51: 2164-2172, 1991

35. Englehard HH III, Butler AB IV, Bauer KD: Quantifica- tion of the c-myc oncoprotein in human glioblastomas cells and tumor tissue. J Neurosurg 71:224-232, 1989

36. Ezzeddine ZD, Martuza RL, Platika D, et al: Selective kill- ing of glioma cells in culture and in vivo by retrovirus transfer of the herpes simplex virus thymidine kinase gene. New Biol 3:608-614, 1991

37. Farrell CL, Stewart PA, Del Maestro RF: A new glioma model in the rat: the C6 spheroid implantation technique permeability and vascular characterization. J Neurooncol 4:403-415, 1987

38. Fitzgcrald JE, Schardein JL, Kurtz SM: Spontaneous tumors of the nervous system in albino rats. J Nail Cancer lnst 52:265-273, 1974

39. Fraser H: Astroeytomas in an inbred mouse strain. J Pathol 1tl3:266-270, 1971

40. Frceman D, Zimmerman HM: Experimental brain tumors. V. Behavior in intraocular transplants. Cancer Res 4: 273=278, 1944

41. Friedman HS, Bigner SH, McComb RD, etal: A model for human medulloblastoma. Growth, morphology, and chromosomal analysis in vitro and in athymic mice. J Neuropathol Exp Neurol 42:485-503, 1983

42. Friedman HS, Colvin OM, Skapek SX, et al: Experimental chemotherapy of human medulloblastoma cell lines and transplantable xenografts with bifunctional alkylating agents. Cancer Res 48:4189-4195, 1988

43. Fujimoto M, Sheridan PJ, Sharp D, et al: Proto-oncogene analysis in brain tumors. J Neurosurg 7(1:91~3--915, 1989

44. Fults D, Brockmeyer D, Tullous MW, et al: p53 mutation and loss of heterozygosity on chromosomes 17 and 10 during human astrocytoma progression. Cancer Res 52: 674-679, 1992

45. Geran R, Congleton GF, Dudeck LE, et al: A mouse ependymoblastoma as an experimental model for screening potential antineoplastic drugs. Cancer Chemother Rep 4 (Suppl 87):53-60, 1974

46. Gordon JE, Ruddle FH: Gone transfer into mouse embryos: production of ~ransgenic mice by pronuclear injection. Methods Enzymol 101:411-433, 1983

47. Gotz W, Theuring F, Schachenmayr W, et al: Midline brain tumors in MSV-SV 40-transgenic mice originate from the pineal organ. Acta Neuropathul 83:308-314, 1992

48. Greene HSN, Arnold H: The homologous and heterologous transplantation of brain and brain tumors. J Neu- rnsurg 2:315-331, 1945

49. Hammang JP, Beatge EE, Behringer RR, etal: Immor- talized retinal neurons derived from SV40 T-antigen- induced tumors in transgenic mice. Neuron 4:775-782, 1990

50. Hammer RE, Pursel VG, Rexroad CE Jr, et al: Production of transgenic rabbits, sheep and pigs by microinjection. Nature 315:680-683, 1985 (Letter)

51. Hanahan D: Heritable formation of pancreatic f3-cell tumours in transgenic mice expressing recombinant insulin- simian virus 40 oncogenes. Nature 315:115-122, 1985

52. Hanahan D: Transgenic mice as probes into complex sys- tems. Science 246:1265-1275, 1989

53. Hayes C, Kelly D, Murayama S, et al: Expression of the neu oncogene under the transcriptional control of the myelin basic protein gone in transgenic mice: generation of transformed glial cells. J Neurosci Res 31:175-187, 1992

54. He XM, Wikstrand CJ, Friedman HS, et al: Differentiation charcteristics of newly established medulloblastoma cell lines (D384 Med, D425 Med, and D458 Med) and their transplantable xenografts. Lab Invest 64:833-842, 1991

55. Hiirter T, Mennel HD: Experimental brain tumors and edema in rats. I. Histology and cytology of tumors. Acta Neuropathol 55:105-111, 1981

56. Hurtt MR, Moossy J, Donovan-Peluso M, et al: Ampli- fication of epidermal growth factor receptor gene in gliomas: histopathology and prognosis. J Neuropathol Exp Neurol 51:84-90, 1992

57. lllmensee K, Hoppe PC: Nuclear transplantation in mus musculus: developmental potential of nuclei from preim- plantation embryos. Cell 23:9-18, 1981

58. Jaenisch R: Transgenic animals. Science 240:1468-1474, 1988

874 J. Neurosurg. / Volume 80 / May, I994


59. Jaenisch R, Breindl M, Harbers K, et al: Retroviruses and insertional mutagenesis. Cold Spring Harb Syrup Quant Biol 50:439-445, 1985

60. Jaenisch R, Mintz B: Simian virus 40 DNA sequences in DNA of healthy adult mice derived from primplantation blastocysts injected with viral DNA. Proc Natl Acad Sei USA 71:1250-1254, 1974

61. James CD, Carlbom E, Dumanski JP, et al: Clonal genomic alterations in glioma malignancy stages. Cancer Res 48: 5546-5551, 1988

62. J~inisch W, Schreiber D: Experimental Tumors of the Central Nervous System. Bigner DD, Swenberg JA (eds). 1st English edition. Kalamazoo, Mich: Upjohn, 1977, pp 10-15

63. Jones TR, Bigner SH, Schold SC Jr, et al: Anaplastic human gliomas grown in athymic mice. Morphology and glial fibrillary acidic protein expression. Am J Patho1105: 316-327, 1981

64. Kayc All, Morstyn G, Gardner I, et al: Development of xenograft glioma model in mouse brain. Cancer Res 46: 1367-1373, 1986

65. Kimura M, Sato M, Akatsuka A, et al: Restoration of myelin formation by a single type of myelin basic protein in transgenic shivercr mice. Proc Natl Acad Sci USA 86: 5661-5665, 1989

66. Kobayashi N, Allen N, Clendenon NR, et al: An improved rat brain tumor model. J Neurosurg 53:808-815, 1980

67. Koestner A, Swenberg JA, Wechsler W: Transplacental production with ethylnitrosourea of neoplasms of the nervous system in Sprague-Dawley rats. Am J Pathol 63: 37-50, 1971

68. Korf HW, G6tz W, Herken R, et al: S-antigen and rod- opsin immunoreactions in midline brain neoplasms of transgenic mice: similarities to pineal cell tumors and certain medulloblastomas in man. J Neuropathol Exp Neu- rol 49:424-437, 1990

69. Levin VA, Freeman-Dove MA, Maroten CE: Dianhydro- galactitol (NSC- 132313): pharmacokinetics in normal and tumor-bearing rat brain and antitumor activity against three intracerebral rodent tumors. J Natl Cancer Inst 56: 535339, 1976

70. LevinVA, Wilson CB: Correlations between experimental chemotherapy in 1he murine glioma and effectiveness of clinical therapy regimens. Cancer Chemother Pharma- col 1:41-48, 1978

71. Levine AJ, Momand J: Tumor supressor genes: the p53 and retinoblastoma sensitivity genes and gene products. Bioehim Biophys Aeta 1032:119-136, 1990

72. Lillien LE, Raft MC: Analysis of the cell-cell interactions that control type-2 astrocyte development in vitro. Neuron 4:525-534, 1990

73. London WT, Houff SA, Madden DL, et al: Brain tumors in owl monkeys inoculated with human polyomavirus (JC virus). Science 201:1246-1249, 1978

74. Lozano G, Levinc AJ: Tissue-specific expression of p53 in transgenic mice is regulated by intron sequences. Mol Carcinogen 4:3-9, 1991

75. Luginbuhl H: Tumors of the CNS in animals. Acta Neu- ropathol 1:9-18, 1962

76. Mancini LO, Yates VJ, Jasty V, et al: Ependymomas induced in hamsters inoculated with an avian adenovirus (CELO). Nature 222:190-191, 1969

77. Manuelidis EE: The fate of serial heterologously transplanted glioblastoma multiforme in the eye of the guinea pig. Am J Pathol 48:65-89, 1966

78. Marcus DM, Carpenter JL, O'Brien JM, et al: Primitive

neuroectodermal tumor of the midbrain in a murine model of retinoblastoma. Invest Ophthalmol Vis Sci 32: 293-301, t991

79. Marks JR, Lin J, Hinds P, et al: Cellular gene expression in papillomas of the choroid plexus from transgenic mice that express the simian virus 40 large T antigen. J Virol 63:790-797, 1989

80. Mawdesley-Thomas LE, Newman AJ: Some observations on spontaneously occurring tumours of the central nervous system of Sprague-Dawley derived rats. J Pathol 112: 107-116, 1974

81. McGrath J, Solter D: Nuclear transplantation in the mouse embryo by microsurgery and cell fusion. Science 220: 1300-1302, 1983

82. McG~ath JT: Intracranial pathology of the dog. J Neu- ropathol Exp Neurol 21:327-328, 1962

83. Mella O, Bjerkvig R, Schem BC, et al: A cerebral glioma model for experimental therapy and/n vivo invasion studies in syngeneic BD IX rats. J Nenrooncol 9:93-104,1990

84. Mellon PL, Windle JJ, Goldsmith PC, et al: Immortaliza- tion of hypothalamic GnRH neurons by genetically targeted tumorigenesis. Neuron 5:1-10, 1990

85. Merkow LP, Slifkin M, Pardo M, et al: Pathogenesis of oncogenic simian adenoviruses. VIII. The histopathology and ultrastructure of simian adenovirus 7-induced intracranial neoplasms. Exp Mol Pathol 12:264-274, 1970

86. Messing A, Chen HY, Palmiter RD, et al: Peripheral neuropathies, hepatocellular carcinomas and islet cell adenomas in transgenic mice. Nature 316:461-463, 1985 (Letter)

87. Mucke L, Oldstone MBA, Morris JC, et al: Rapid acti- vation of astrocyte-specific expression of GFAP-lacZ transgene by focal injury. New Biol 3:465-474, 1991

88. Mukai N, Kobayashi S: Primary brain and spinal cord tumors induced by human adenovirus type 12 in hamsters. J Neuropathol Exp Neurol 32:523-541, 1973

89. O'Brien JM, Marcus DM, Bernands R, et al: Trilateral retinoblastoma in transgenic mice. Trans Am Ophthalmol Soc 87:301-326, 1989

90. Ornitz DM, Palmiter RD, Messing A, et al: Elastase I promoter directs expression of human growth hormone and SV40 T antigen genes to pancreatic acinar cells in transgenic mice. Cold Spring Harb Symp Quant Biol 50: 399-409, 1985

91. Padgett BL, Walker DL, ZuRhein GM: Differential neuro- oncogenicity of strains of JC virus, a human polyoma virus, in newborn Syrian hamsters. Cancer Res 37:718-720, 1977

92. Padgett BL, Walker DL, ZuRhein GM: JC papovavirus in progressive multifocal leukoencephalopathy. J Infect Dis 133:686-690, 1976

93. Palmiter RD, Brinster RL: Germ-line transformation of mice. Annu Rev Genet 20:465-499, 1986

94. Palmiter RD, Chen HY, Messing A, et al: SV40 enhancer and large-T antigen are instrumental in development of choroid plexus tumours in transgenic mice. Nature 316: 457-460, 1985 (Letter)

95. Palmiter RD, Sandgren EP, Avarbock MR, etal: Heter- ologous introns can enhance expression of transgenes in mice. Proc Natl Acad Sci USA 88:478-482, 1991

96. Pantelouris EM: Absence of thymus in a mouse mutant. Nature 217:370-371, 1968

97. Pattengale PK, Stewart TA, Leder A, et al: Animal models of human disease. Pathology and molecular biology of spontaneous neoplasms occurring in transgenie mice car- rying and expressing activated cellular oncogenes. Am J Pathol 135:39-61, 1989

J. Neurosurg. / Volume 80 / May, 1994 875

D. L. Pcterson, P. J. Sheridan, and W. E. Brown, Jr.

98. Perraud F, Yoshimura K, Louis B, ctal: The promoter of the human cystic fibrosis transmcmbranc conductance regulator gene directing SV40 T antigen expression induces malignant proliferation r ependymal cells in transgenie mice. Oncogene 7:993-997, I992

99. Pinkert CA, Ornitz DM, Brinster RL, ct al: An albumen enhancer located 10 kb upstream functions along with its promoter to direct efficient, liver-specific expression in transgenic mice. Genes Dev 1:268-276, 1987

100. Rabotti GF, Raine WA: Brain tumors induced in hamsters inoculated intracerebrally at birth with Rous sarcoma virus. Nature 204:898-899, 1964

101. Raft MC: Glial cell diversification in the rat optic nerve. Science 243:1450-1455, 1989

102. Rama B, Spoerri O, Holzgraefe M, etal: Current brain tumor models with particular consideration of the transplantation techiques. Outline of literature and personal primary results. Acta Neurochir 79:35~41, 1986

103. Rana MW, Pinkerton H, Thornton H, etal: Heterotrans- plantation of human glioblastoma multiforme and meningioma to nude mice. Proe Sue Exp Biol Med 155:85-88, 1977

104. Readhead C, Popko B, Takahashi N, etal: Expression of the myelin basic protein gone in transgenic shiverer mice: correction of the dysmyelinating phenotype. Cell 48: 703-712, 1987

105. Rewers AB, Redgate ES, Deutsch M, etal: A new rat brain tumor model: glioma disseminated via the cerebral spinal fluid pathways. J Neuroonenl 8:213-219, 1990

106. Rous P: A sarcoma of the fowl transmissible by an agent separable from the tumor cells. J Exp Meal 13:397-410, 1911

107. San-Galli F, Vrignaud P, Robert J, et al: Assessment of the experimental model of transplanted C6 glioblastoma in Wistar rats. J Neurooncol 7:299-304, 1989

108. Schmidek HH, Nielsen SL, Schiller AL, etal: Morpho- logical studies of rat brain tumors induced by N-nitrosomethylurea. J Neurosurg 34:335-340, 1971

109. Schold SC Jr, Bigner DD: Treatment of five subcutaneous human glioma tumor lines in athymic mice with carmustine, procarbazine, and mithramycin. Cancer Treat Rep 67:811-819, 1983

110. Schold SC Jr, Friedman HS, Bigner DD: Therapeutic pro- file of the human glioma line D-54 MG in athymic mice. Cancer Treat Rep 71:849-850, 1987

111. Schold SC Jr, Friedman HS, Bjornsson TD, etal: Treat- ment of human glioma and medulloblastoma tumor lines in athymic mice and diaziquone-based drug combinations. Cancer Res 44:2352-2357, 1984

112, Schubert D, Heinemann S, Carlisle W, etal: Clonal cell lines from the rat central nervous system. Nature 249: 224-227, 1974

113. Schwartz MS, Morris J, Sarid J: Overexpression of oncogene products can cause tumor progression without parenchymal infiltration in the rat brain. Cancer Res 51: 3595-3601, 1991

114. Seligman AM, Shear M J: Studies in carcinogenesis. VIII. Experimental production of brain tumors in mice with methylcholanthrene. Am J Cancer 37:364-395, 1939

115. Serano RD, Pegram CN, Bigner DD: Tumorigenic cell culture lines from a spontaneous VM/Dk murine astrocytoma (SMA). Aeta Neuropathol 51:53-64, 1980

116. Shapiro WR, Basler GA, Chernik NL, etal: Human brain tumor transplantation into nude mice. J Natl Caneer lust 62:447-453, 1979

117. Shein HM: Neoplastic transformation of hamster astrocytes in vitro by simian virus 40 and polyoma virus. Sci- ence 159:1476-1477, 1968

] 18. Shih C, Padhy LC, Murray M, etal: Transforming genes of carcinomas and neuroblastomas introduced into mouse fibroblasts. Nature 290:261-264, 1981 (Letter)

119. Tabuchi K, Nishimoto A, Matsumoto K, et al: Establish- menl of a brain-tumor model in adult monkeys. J Neu- rosurg 63:912-916, 1985

] 20. Theuring F, Gotz W, Bailing R, etal: Tumorigenesis and eye abnormalities in transgenic mice expressing MSV- SV4t} large T-antigen. Oneogene 5:225-232, 1990

121. U HS, Kelley PY, Hatton JD, etal: Proto-oncogene abnormalities and their relationship to tumorigenicity in some human glioblastomas..1 Neurosurg 71:83-90, 1989

122. Unterharnscheindt F: lntracerebrales angioma cavernosum bei einem Affen (Macacus rhesus). Acta Neuropathol 3:295-296, 1964

123. Unterharnscheindt F, Nagler FP: Syringomyelia bei cinch Affen (Macacus mulatta). Acta Neurnpathol 4:583-568, 1965

124. Vfizquez-L6pez E: On the growth of Rous sarcoma inoculated into the brain. Am J Cancer 26:29-55, 1936

125. Vick NA, Bigner DD: Some structural aspects of dogbrain tumors induced with the Schmidt-Ruppin strain of the Rous sarcoma virus. Prog Exp Tumor Res 17:59-73, 1972

126. Wilson CB: Brain tumor models for experimental therapy, in Laerum OD, Bigner DD, Rajewsky MF (eds): Biology of Brain Tumors. Geneva: UICC, 1978, pp 185-199

127. Wilson CB, Norrell H Jr, Barker M: Intrathecal injection of methotrexate (NSC-740) in transplanted brain tumors. Cancer Chemother Rep 51:1-6, 1967

128. Windle JJ, Weiner RI, Mellon PL: Cell lines of the pituitary gonadotrope lineage derived by targeted oncogenesis in transgenic mice. Mol Endoerinol 4:597-603, 1990

129. Yoshida J, Cravioto H: Nitrosourea-induced brain tumors: an in vivo and in vitro tumor model system. J Natl Cancer Inst 61:365-370, 1978

130. Zimmerman HM: The nature of gliomas as revealed by animal experimentation. Am J Pathol 31:1-29, 1955

131. Zimmerman HM, Arnold H: Experimental brain tumors. I. Tumors produced with methylcholanthrene. Cancer Res 1:91 ~)-938, 1941

Manuscript received December 10, 1992. Accepted in final form December 29, 1993. Address reprint requests to: Willis E. Brown, Jr., M.D., Di-

vision of Neurosurgery, The University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, Texas 78284-7843.

876 J. Neurosurg. / Volume 80 / May, I994

Download - Animal models for brain tumors: historical perspectives and future directions

Top Related