U.S. flag An official website of the United States government
  1. Home
  2. Vaccines, Blood & Biologics
  3. Science & Research (Biologics)
  4. Development of Assays of Defined Sensitivity for the Regulatory Management of Novel Cell Substrates
  1. Science & Research (Biologics)

Development of Assays of Defined Sensitivity for the Regulatory Management of Novel Cell Substrates


Andrew M. Lewis, MD

Andrew M. Lewis, MD

Office of Vaccines Research and Review
Division of Viral Products
Laboratory of DNA Viruses



Andrew Lewis is a Principal Investigator in the Laboratory of DNA Viruses. He received his M.D. degree from Duke University in 1961 and worked as a scientist at the National Institute of Allergy and Infectious Diseases from 1963 to 1994.  His work at the NIH focused on virology, the discovery of non-defective adenovirus-SV40 hybrids, and the ability of DNA viruses such as SV40 to induce neoplastic transformation of cells in tissue culture and cause tumors in rodents.  Due to the concerns over the possible risks associated with the non-defective adeno-SV40 hybrid viruses, he was an active participant in the Asilomar Conference in 1975 that focused on the possible risks posed by laboratory experimentation involving recombinant DNA.  He joined CBER, FDA in the Division of Viral Products in 1995 to focus on safety issues associated with the use of transformed cells as cell substrates for vaccine manufacture.  At CBER, Dr. Lewis first worked on the possible association between SV40 contamination of early polio vaccines and subsequent tumor development in humans, which had implications for the safety of continuous cell lines for viral vaccine manufacture.  He became Chief of the Laboratory or DNA Viruses in the late 1990s.  He and others initiated a program to study the tumorigenic potential of transformed cells used during the production of new vaccines. These data were presented at the FDA advisory committee meeting in 2000. This committee encouraged Dr. Lewis and his lab to pursue this project further to identify and characterize the basic biological processes that allow VERO cells to develop the capacity to form tumors.  The lab has continued these studies for the past 20 years and has developed what may represent a hypothetical model of neoplasia in tissue culture.

General Overview

Vaccines are an essential public health tool for controlling viral diseases. Viral vaccines are produced in living cells; when cells are used during vaccine manufacture, they are called cell substrates. The development of safe and effective viral vaccines requires that these cell substrates and the other materials used in the production of vaccines must be carefully characterized.

There are a variety of types of cell substrates, including cells from embryonated eggs and cells from mammals as well as  other species grown in culture. Some cells used as cell substrates are immortalized (transformed into cancer-like cells), that is, they continually multiply so the culture never dies out. Cells from cultures of transformed cells have the potential to form tumors in animals (tumorigenic cell substrates).

A major challenge to the safety of vaccines manufactured in transformed/tumorigenic cell substrates is contamination with infectious agents. Problems include cancer-causing viruses and genetic material (DNA) that might encode infectious agents or small microRNAs (miRNA) that might trigger neoplastic activity in the vaccine recipient. However, transformed cell substrates are important for the development of vaccines for HIV/AIDS, vaccines against annual and pandemic influenza, vaccines to protect against the new viruses that are producing epidemics (e.g., Ebola, Zika, SARS-COV-2), and vaccines to protect against viral agents of bioterrorism. FDA reviewers must evaluate the safety issues posed by all cell substrate reagents used in the manufacture of viral vaccines.

Regulatory evaluation of transformed cell substrates could be improved by a better understanding of the processes involved in cell transformation. We have developed a systematic approach to neoplastic transformation of cells in tissue culture, creating cell lines across the spectrum of neoplasia from normal cells to tumorigenic cells. We have carefully characterized these cells for their tumorigenicity in animals, and applied miRNA analysis, chromosome analysis, methylation analysis, and bioinformatics to study these models. In this way, our laboratory is beginning to understand the fundamentals of mammalian cell transformation, how transformed cells develop the ability to form tumors, and how these tumors actually develop. In collaboration with the Peden laboratory, we are developing ways to examine the possible risks that might be associated with the DNA from tumor-forming cell substrates, including: 1) the possibility of transferring cancer-causing activity (DNA) to vaccines, 2) the possibility of transferring infectious microorganisms to vaccines, and 3) the possibility that enzymes added during manufacturing may cause harmful DNA degradation/breakdown.

Scientific Overview

The current projects underway include: 1) understanding the process of neoplastic transformation in cells in tissue culture, and using this understanding to characterize the tumorigenic phenotypes expressed by neoplastic cell substrates, including the VERO line of African green monkey kidney cells and Madin-Darby canine kidney (MDCK) cells; and 2) evaluating the oncogenicity/infectivity of DNA from neoplastic cell substrates;

Determining whether transformed cell substrates are tumorigenic requires injecting cells into immune-incompetent mice. The cell doses selected represent dose-response assays that includes both tumor-forming and non-tumor forming cell doses. Such assays quantitatively define fundamental traits that characterize the neoplastic cell tumorigenic phenotype (cell-dose trait and tumor-latency trait). Using cell lines characterized by dose-response assays, we are studying neoplastic processes leading to the development of the tumorigenic phenotype and evaluating mechanisms of tumor formation in animals. Understanding the nature of cell doses required for tumor formation has been a major challenge for decades. We hypothesized that tumorigenic cell doses represent the ratio of tumor-forming stem cells to stromal/accessory cells required for tumor formation in animals. We are therefore examining miRNA expression across a spectrum of cell lines that exhibit 4000-fold differences in numbers of cells required for tumor formation, and results are providing insights into mechanisms associated with such differences.

These studies now provide a model for the systematic study of neoplastic transformation in tissue culture and the cell-culture reagents needed to study the details of this process. These reagents have allowed us to study chromosome alterations and patterns of miRNA expression associated with the neoplastic processes that occur during the passage of mammalian cells in tissue culture. Chromosome changes are fundamental to the process of neoplastic transformation. We have also found that patterns of expression of miRNAs are basic components of this process. Based these findings, 6 – 8 miRNAs appear to be biomarkers of the VERO cell tumorigenic phenotype. This observation may contribute to the regulatory management of neoplastic cell substrates by reducing the need for animal models to study tumorigenicity. Association of these biomarkers with VERO-cell tumorigenicity has been expanded by developing another tumorigenic line of African green monkey kidney (AGMK1-9T7) cells. These data have allowed us to identify some type of yet-to-be defined renal stem cells as the origin of AGMK cell lines. Based on these findings, we have developed a comprehensive hypothesis of tissue culture transformation.

With the Peden laboratory, we are evaluating the oncogenic activity posed by DNA from neoplastic cell substrates and the role of the murine immune system in tumorigenicity/metastases. Activated H-ras and c-myc oncogenes are oncogenic in mice when injected together in different plasmids or when combined in the same plasmid. Less than a nanogram of plasmid DNA containing both oncogenes can induce tumors. These plasmid-mouse models allow the study of possible oncogenic activity associated with the DNA of neoplastic cell substrates, as well as evaluation of the impact of DNA degradation on the removal of oncogenic activity and the infectivity of cell DNA-containing retroviral genomes.


  1. PLoS One 2023 Dec 7;18(12):e0293406
    GLI1+ perivascular, renal, progenitor cells: The likely source of spontaneous neoplasia that created the AGMK1-9T7 cell line.
    Lewis AM Jr, Foseh G, Tu W, Peden K, Akue A, KuKuruga M, Rotroff D, Lewis G, Mazo I, Bauer SR
  2. Biologicals 2023 Nov;84:101724
    Evaluating the sensitivity of newborn rats and newborn hamsters to oncogenic DNA.
    Sheng-Fowler L, Tu W, Phy K, Macauley J, Lanning L, Lewis AM Jr, Peden K
  3. PLoS One 2022 Oct 24;17(10):e0275394
    The AGMK1-9T7 cell model of neoplasia: evolution of DNA copy-number aberrations and miRNA expression during transition from normal to metastatic cancer cells.
    Lewis AM Jr, Thomas R, Breen M, Peden K, Teferedegne B, Foseh G, Motsinger-Reif A, Rotroff D, Lewis G
  4. Vaccine X 2019 Apr 11;(1):100004
    Responsiveness to basement membrane extract as a possible trait for tumorigenicity characterization.
    Murata H, Omeir R, Tu W, Lanning L, Phy K, Foseh G, Lewis AM Jr, Peden K
  5. Vaccine 2017 Oct 4;35(41):5503-9
    Assessment of potential miRNA biomarkers of VERO-cell tumorigenicity in a new line (AGMK1-9T7) of African green monkey kidney cells.
    Teferedegne B, Rotroff DM, Macauley J, Foseh G, Lewis G, Motsinger-Rief A, Lewis AM Jr
  6. Chromosome Res 2015 Dec;23(4):663-80
    A novel canine kidney cell line model for the evaluation of neoplastic development: karyotype evolution associated with spontaneous immortalization and tumorigenicity.
    Omeir R, Thomas R, Teferedegne B, Williams C, Foseh G, Macauley J, Brinster L, Beren J, Peden K, Breen M, Lewis AM Jr


Back to Top