Evaluating the sensitivity of newborn rats and newborn hamsters to oncogenic DNA.

To evaluate the risk of residual cellular DNA in vaccines manufactured in tumorigenic cell lines, we have been establishing in vivo assays to quantify the oncogenic activity of DNA. We had generated three oncogene-expression plasmids: pMSV-T24-H-ras, which expresses activated H-ras; pMSV-c-myc, which expresses c-myc; and pMSV-T24-H-ras/MSV-c-myc, which expresses both oncogenes. Tumors were induced in mice by pMSV-T24-H-ras plus pMSV-c-myc or by pMSV-T24-H-ras/MSV-c-myc. Because newborn hamsters and newborn rats have been recommended for oncogenicity testing of the DNA from tumorigenic mammalian cell-substrates used for vaccine production, we evaluated their sensitivity. Newborn hamsters and rats were inoculated with different doses of pMSV-T24-H-ras/MSV-c-myc to determine their sensitivity to tumor induction and with the single-oncogene-expression plasmids to determine whether single oncogenes could induce tumors. Newborn rats were more sensitive than newborn hamsters, and activated H-ras but not c-myc induced tumors in newborns of both rodent species. DNA from four cell lines established from tumors induced by pMSV-T24-H-ras/MSV-c-myc was inoculated into newborn rats. Because no tumors were induced by this cellular DNA, which should be optimal as it contains both oncogenes linked and present in several copies, we conclude that available in vivo models are not sensitive enough to detect the oncogenicity of cellular DNA.

Responsiveness to basement membrane extract as a possible trait for tumorigenicity characterization.

Immortalized cell lines used to produce vaccines are expected to be described in terms of their tumorigenicity. However, current in vivo tumorigenicity assays can be time-consuming and results can be equivocal, especially for weakly tumorigenic cells. Basement membrane extract (BME) derived from the Engelbreth-Holm-Swarm mouse tumor, such as Matrigel and Cultrex, consists of laminin, collagen IV, entactin, heparan sulfate, and proteoglycans, as well as biologically active peptides and growth factors. For nearly three decades, BME has been used in cancer research to enhance tumorigenicity assays (both tumor "take" as well as tumor growth are substantially improved). We assessed the feasibility of using BME to facilitate the evaluation of vaccine cell substrate tumorigenicity. Vero cells (WHO 10-87) were serially passaged and banked at every ten passages beginning with p140; for the present study, low-passage Vero cells (Vero LP, originating from cells banked at p140) and high-passage Vero cells (Vero HP, originating from cells banked at p250) were used. In addition, Vero TPX2 and Vero NM1, cell lines established from tumors formed in nude mice by Vero HP cells, as well as other cell lines relevant to vaccine production (HeLa, MDCK, 293, and ARPE-19), were assessed. Female adult athymic nude mice were injected subcutaneously with cells in the absence or presence of BME. We observed that the tumorigenicity of ARPE-19 cells as well as Vero cells below passage 258 (Vero LP and Vero HP; previously characterized as non-tumorigenic or weakly tumorigenic, respectively) was not enhanced by BME. In contrast, BME shortened the latency and decreased the tumor-producing cell dose of HeLa, 293, and MDCK cells as well as the tumorigenic Vero derivatives TPX2 and NM1. Thus, responsiveness to BME may reflect the status of the neoplastic process and possibly serve as a useful trait for better defining the tumorigenic phenotype of cells.

Development of a micro-neutralization assay for ebolaviruses using a replication-competent vesicular stomatitis hybrid virus and a quantitative PCR readout.

Development of vaccines against highly pathogenic viruses that could also be used as agents of bioterrorism is both a public health issue and a national security priority. Methods that can quantify neutralizing antibodies will likely be crucial in demonstrating vaccine effectiveness, as most licensed viral vaccines are effective due to their capacity to elicit neutralizing antibodies. Assays to determine whether antibodies are neutralizing traditionally involve infectious virus, and the assay most commonly used is the plaque-reduction neutralization test (PRNT). However, when the virus is highly pathogenic, this assay must be done under the appropriate level of containment; for tier one select agents, such as Ebola virus (EBOV), it is performed under Biological Safety Level 4 (BSL-4) conditions. Developing high-throughput neutralization assays for these viruses that can be done in standard BSL-2 laboratories should facilitate vaccine development. Our approach is to use a replication-competent hybrid virus whose genome carries the envelope gene from the pathogenic virus on the genetic backbone of a non-pathogenic virus, such as vesicular stomatitis virus (VSV). We have generated hybrid VSVs carrying the envelope genes for several species of ebolavirus. The readout for infectivity is a one-step reverse transcriptase quantitative PCR (RT-qPCR), an approach that we have used for other viruses that allows robustness and adaptability to automation. Using this method, we have shown that neutralization can be assessed within 6-16h after infection. Importantly, the titers obtained in our assay with two characterized antibodies were in agreement with titers obtained in other assays. Finally, although in this paper we describe the VSV platform to quantify neutralizing antibodies to ebolaviruses, the platform should be directly applicable to any virus whose envelope is compatible with VSV biology.

Assessment of viral replication in eggs and HA protein yield of pre-pandemic H5N1 candidate vaccine viruses.

H5N1 infection and the potential for spread from human to human continue to pose a severe public health concern. Since vaccination remains the most effective way to prevent a potential H5N1 pandemic, the World Health Organization (WHO) Collaborating Centers (CCs) and Essential Regulatory Laboratories (ERLs) engineered and developed a panel of H5N1 pre-pandemic vaccine viruses for pandemic vaccine preparedness as well as production of antigen potency testing reagents (reference antigen and reference anti-serum) for vaccine standardization. To develop a strategy utilizing a number of biochemical methods for the characterization of the viral growth properties and protein yield in eggs, we have selected eight H5N1 pre-pandemic viruses and determined the viral Egg Infectious Dose 50 (EID50), total protein yield, hemagglutinin (HA) to nucleoprotein (NP) ratios (HA:NP), and HA1 content of each virus. Our results showed that all the tested H5N1 vaccine viruses grew to high titers in eggs. The total viral protein yield varies within a narrow range, whereas there were greater differences in the HA:NP protein ratios among the eight viruses. The RP-HPLC based HA1 content analysis demonstrated that the viruses A/Anhui/1/2010, A/Hubei/1/2005, and A/goose/Guiyang/337/2006 contained higher HA contents than other five viruses including A/Vietnam/1203/2003. Our approach for analyzing virus growth and protein yield will allow us identify optimal vaccine virus in a timely manner. In addition, we successfully purified the HA proteins of H5N1 vaccine viruses by optimizing bromelain cleavage conditions. Our studies on the HA protein purification may improve the quality control of the production of influenza vaccine test reagent.

Increased hemagglutinin content in a reassortant 2009 pandemic H1N1 influenza virus with chimeric neuraminidase containing donor A/Puerto Rico/8/34 virus transmembrane and stalk domains.

The glycoproteins, heamagglutinin (HA) and neuraminidase (NA) of influenza virus confer host protective immune responses during vaccination, which is the most effective approach for preventing influenza-associated morbidity and mortality. Since the functional balance between the HA and NA proteins may affect viral receptor binding and replication, a pandemic influenza A virus (H1N1 pdm09), strain A/Texas/05/2009, was optimized to elevate its HA antigen content by modifying the NA gene. In this study, we have constructed two 2:6 reassortant viruses between pdmH1N1 (A/Texas/05/2009) and A/Puerto Rico/8/34 (PR8), in which the NA gene of A/Texas/05/2009 was modified to contain part of the NA gene from PR8. One chimeric NA virus has the PR8 transmembrane (TM) region (HNtm 2:6) and the other contains both the PR8 NA TM and stem regions (HNst 2:6). Using quantitative reverse phase-HPLC (RP-HPLC) analysis, we observed that the HNst2:6 virus contains a higher HA1 content than HN2:6 wild type. In addition, this mutant virus displays a higher HA1 to nucleoprotein (NP) ratio, based on gel electrophoresis densitometry analysis. Furthermore, the neuraminidase activity of purified HNst 2:6 virus is approximately 30% lower than that of HN2:6 virus, which is suggestive of a lower incorporation of NA into the viral envelope. Therefore, we propose that the reduction of NA packaging in the virion may lead to a compensatory increase of HA. Such an improvement in HA yield is possibly beneficial to H1N1 pdm09 vaccine production.