Assays to detect oncogenic agents in cell substrates

Investigators

Abstract

Currently, our research program consists of 4 projects that are being pursued using, where possible, a complementary approach. Each of these projects will be summarized independently. The development of in vivo and in vitro assays of defined sensitivity to detect oncogenic agents in cell substrates used for the production of vaccines and other biologicals. Contamination with adventitious agent represents a major safety issue during the development/manufacture of viral vaccines and advancing methods of adventitious agent detection, especially unknown agents, in cell substrates is a continuing challenge. Thus, detection techniques that can be use to assure at, defined levels per vaccine dose, the likely absence of oncogenic agents represent important advances in managing safety issues associated with viral vaccine development. PERT assays can be used quantitatively to search cell substrates for retroviruses. However, current assays for detecting known or unsuspected oncogenic DNA viruses, such as PCR amplification with virus-specific primers and newborn animal inoculation, represent techniques that are either been difficult to quantify (standard PCR) or have not been assessed for their level of sensitivity (animal assays). The development of Real Time Quantitative PCR technology has provided an assay that has the potential to become the method of choice to evaluate novel cell substrates for known oncogenic or infectious agents. The inoculation of newborn mice, rats, or hamsters with extracts or culture fluids from cell substrates is one of the few assays available for the detection of unknown oncogenic agents. For regulatory purposes, the problem with PCR is it inability to detect unknown viruses and the problem with animal assays, especially assays in mice, is that the efficiency (number of infectious/tumor-inducing doses/ml) with which they can detect viruses has not been well established. Developing TaqMan assays with defined sensitivity for specific viruses requires developing primers and probes that can be shown to be highly specific for the virus to be detected and assessing their ability to detect the genome of this virus under the conditions needed to evaluate cell substrates. Evaluating newborn animals for their ability to detect oncogenic viruses is more complex and depends on the sensitivity of the injected animals and the oncogenic activity of the putative virus. As we believe that the TaqMan assay and animal assays can be used in a complementary fashion to evaluate novel cell substrates, the aim of this project is to define the relative sensitivities of both the TaqMan assays and assays in immunocompromised animals for their threshold limits to detect occult oncogenic agents. Because of their wide distribution in human tissues including neoplastic tissues, their propensity to establish latent infections, and the public health challenge posed by SV40, the polyomavirus that contaminated of early poliovaccines, we are using polyomaviruses as models for developing TagMan assays to detect adventitious agents. For these assays we have developed 4 primer/probe sets for BKV, JCV and SV40 and have shown that each reacts with its viral DNA over a range of 8-9 logs. Ten of twelve primer./probe sets could detect 1-10 copies of their respective DNA per reaction, while 2/12 detected 10-100 copies. Only I primer/probe set (JCV JL4) cross-reacted with the DNA from another (SV40) polyomavirus at levels of 10-100 copies per reaction. When used to evaluate their specificity in blinded samples containing mixed samples of the DNA of all 3 viruses in concentrations ranging from 10-10000 copies per reaction, the 2 primer/probe sets representing each virus reacted only with the DNA of their respective viruses. The data from these assays accurately predicted the amount of the respective viral DNA present in the different samples (variance 4-40%). We are planning to use these primer/probe sets to estimate the number of copies of polyomavirus DNA present in a variety of different tissues and fluids. For this study, we are comparing the oncogenicity of serial concentrations (10e7-10e4) of SV40 (highly oncogenic in hamsters but nononcogenic in mice), adenovirus 12 oncogenic in both hamsters and mice, and adenoviruses 2 and 5 (nononcogenic in all species tested but capable of transforming cells from hamsters, mice, rats, and humans into tumor cells in vitro) for their capacity to induce tumors in T-cell deficient adult and newborn nude mice, T-cell and B-cell deficient adult beige-nude-xid mice, and T-cell, natural-killer cell deficient, adult CD3epsilon transgenic mice. Animals from each strain were inoculated SC with 10e7 pfu/mouse were observed for tumor development for up to 8 months. Ad12 induced tumors in both adult and newborn nude mice. The tumor incidence in newborns was 63% compared to 44% in adults and tumors developed in newborns at 10e4 and 10e5 pfu/mouse but only in adult nude mice injected with 10e6 and 10e7 pfu/mouse. In addition, time to tumor appearance in adults ranged between 13-24 weeks while the range in newborns was 8-16 weeks. Ad12 induced tumors in only 36% of adult beige nude-xid mice at doses of 10e6 pfu/mouse. No tumors developed in adult nude mice, beige-nude-xid mice, or CD3 epsilon mice injected with either Ad2, Ad5 or SV40. The CD 3 epsilon mice suffered a high level of mortality due to pneumocystis carnii infection and the results of this part of the experiment are not considered valid. These results suggest that the newborn nude mouse is the most sensitive of these animal models to detect oncogenic adenoviruses and possibly other types of oncogenic DNA viruses. Work is underway to verify these results. Evaluation of VERO cell evolution from a nontumorigenic to a tumorigenic phenotype. VERO cells, a continuous line of African green monkey kidney cells that became neoplastic by spontaneous transformation events, represent the prototypic neoplastic cell line that is being used to develop and manufacture viral vaccines. While usually non-tumorigenic at passage levels up to 160, these cells have the capacity to evolve into cells that can form tumors upon continuous tissue culture passage. The question (posed by the Vaccines and Related Biological Products Advisory Committee in 2000) is whether the instability in the VERO cell phenotype poses a safety concern for vaccines developed in these cells. We have developed high passage VERO cells and have shown them to be capable of forming tumors in athymic nude. These initial mouse experiments are being repeated and we are in the early stages of evaluating the properties of the tumorigenic, high passage, VERO cells and the VERO cells derived from nude mouse tumors in an effort to determine whether a multi-step process is associated with their evolution to a tumorigenic phenotype. Documenting the multi-step nature of the evolution of the VERO cell phenotype from non-tumorigenic to tumorigenic should reduce the concerns about the risks posed by the potential for tumorigenic activity of these cells. Developing animal models to evaluate the oncogenic activity of cell substrate DNA. For over 40 years, the proposed use of immortalized, neoplastic cells (which includes transformed cells and tumorigenic cells) for the manufacture of viral vaccines has raised serious concerns about product safety. One of these concerns is due to the contamination of vaccines manufactured in such cells with residual cell-substrate DNA. DNA from neoplastic cell substrates has two activities that have been considered to be potential risks to vaccine recipients. The first is the potential risk for oncogenic activity, and the second is the potential risk for infectivity from retrovirus provirus genomes or the genomes of other whole viruses that may be present in these types of cells. In this project, we (Laboratory of DNA Viruses in collaboration with Dr. Keith Peden and Li Sheng, Laboratory of Retroviruses and Drs. Don Blair, Steve Hughes, and John Coffin, NCL, and Frank Sistare, CDER) are developing animal models that can be used for regulatory purposes to estimate and evaluate the risks of oncogenic activity posed by cell-substrate DNA. Plasmids containing the cellular oncogenes H-ras and c-myc have been developed and shown to be functional by their capacity to transform NIH 3T3 cell in vitro. When injected in to two strains of fully immunocompetent adult and newborn mice, these plasmids have produced tumors. Cell lines have been established from these tumors and both H-ras and c-myc oncogene DNA have been found in the cells from these tumors. Experiments are underway to confirm and expand these results, to develop a database that can be used to estimate the potential risk of oncogenic activity by cell substrate DNA that carries dominant activated oncogenes, and to develop methods to evaluate these risks for DNA from all types of cell substrates. Assessing the possibility that SV40 is circulating in humans as a consequence of exposure to SV40 contamination of early poliovaccines. SV40 contamination of early poliovaccines manufactured between 1955 and 1960 continues to present challenges to the regulatory process. Millions of individuals were exposed to these contaminated vaccines during polio immunizations before 1963. In the past ten years, the genome of SV40 and, in some cases infectious SV40 virions, have been detected in certain types of human tumors in individuals who were not exposed to SV40 contaminated poliovaccines, suggesting that SV40 is circulating in humans. These results have become extremely controversial and the subject of international meetings, reviews by the Institute of Medicine, of the National Academy of Sciences and hearings before congressional committees. At the present time, it is unclear as to whether and how SV40 might be carried in the human population. Two types of studies are underway in the Laboratory of DNA Viruses to address the issue of whether/how SV40 might be carried by humans. . First, we are supporting work by Dr. Rosina Girones, University of Barcelona, to evaluate environmental samples including urban sewage and waste water samples for polyomaviruses including SV40. Work from this group has found, by surveying 52 samples collected geographical areas ranging from South Africa to Sweden and the United States, BKV in 46/51, JCV in 50/51. None of the samples was positive for SV40 DNA at levels of 10 genomes/4ml of sample. The same techniques detected SV40 in excreta from a colony of cynomolgus monkeys housed in the United Kingdom. Work is underway to attempt to detect SV40 and other polyomaviruses in similar samples from northern India, the home of the rhesus monkey, which is the natural host for SV40 and the source of the virus present in early poliovaccines. In our laboratory, we have found cross- reactivity between BKV and SV40 antibodies in plaque neutralization assays. Currently, we are attempting to confirm these results using other types of serological assays and are exploring possible interactions between SV40 and BKV in mixed infections in tissue culture. As human are infected with BKV during the first few years of life, any interactions between BKV and SV40 could have implications for the spread of SV40 among infected humans. We are also in the process of recovering SV40 from adenoviruses adapted to grow in rhesus monkey kidney cells in 1955-1956. The genomes of these SV40 virions will be sequenced. These sequence data should greatly expand our knowledge of the SV40 strains that were present in rhesus monkeys during the early stages of polio and adenovirus vaccine development, and they should serve as a basis for comparing SV40 human isolates with any new isolates obtained from feral monkeys and identifying any changes that have occurred in the SV40 genome as a result of human passage or host selection over the past 50 years.