Animal Cell Substrates: Back to the Future

1950s

It all started in the 1950s with a question about what cell substrate should be used to develop a live adenovirus vaccine (1). Essentially there were two choices: HeLa cells, derived from a human carcinoma, in which the adenovirus grew to high titers, and primary monkey kidney cells, in which the virus did not grow well. The government committee that reviewed these cell substrate options determined that HeLa cells were not acceptable as a substrate for the vaccine and that normal cells should be used for vaccine production. However, they did not define “normal.”

The group involved in the adenovirus vaccine development effort included Maurice Hilleman, who is now recognized as a major figure in vaccine development in the twentieth century. Hilleman and his group pursued adenovirus vaccine development, and eventually began a Phase I clinical study of a vaccine grown in what they thought were normal human embryonic cells. In retrospect, the cells used to produce the vaccine were anything but normal and were, in fact, most likely a HeLa cell contaminant. The Phase I clinical study was limited to 11 people, one of whom was Hilleman himself. An 11-year follow-up of these 11 human subjects showed that they were all alive and well, despite the fact that they had almost certainly received a vaccine produced in HeLa cells.

The next significant development in cell substrates was the use of primary monkey kidney cells for the manufacture of the Salk vaccine (inactivated polio vaccine, IPV). Subsequent to the use of primary monkey kidney cells, other primary cells (e.g., chicken embryo fibroblasts, duck embryo fibroblasts, hamster kidney, and dog kidney) began to be used for vaccine research and development. Essentially by default, normality was equated with primary cells, and they were used widely for vaccine development.

1960s

The next major event in the cell substrate arena was the development of human diploid cells (HDCs) (2). The safety issues pertaining to HDCs illustrate the “what if” problem, which was highlighted by Albert Sabin who, when considering the acceptability of HDCs as a substrate for vaccine production, said, “There is no full characterization for any cell line because for everything we do, there is always that hypothetical something for which we cannot test” (3). If one is dealing with a theoretical agent, there obviously are no tests for it. This creates the untenable position of being unable to demonstrate its absence, even using the best technology available.

The debate over HDCs continued for a number of years, but eventually there was a gradual acceptance of HDCs for vaccine production. This acceptance was based on a significant amount of characterization data that showed that HDCs had a normal chromosomal constitution, a finite lifespan, and were unable to form tumors in animal models. Another factor crucial for the acceptance of HDCs was the willingness on the part of regulatory agencies and their advisors to consider benefits as well as risks and to treat the hypothetical human oncogenic agent as a remote theoretical risk. The acceptance of HDCs occurred first in Europe, but eventually in the U.S., and then elsewhere in the world. As a consequence of the HDC experience, the characterization of cells became a central feature in the evaluation of all the new cell types that followed HDCs.

There is a sidebar to the HDC issues of the late 1960s and early 1970s that should be highlighted. At the time, the U.S. regulatory agency responsible for biological products was the Division of Biologics Standards (DBS), which was part of the National Institutes of Health (NIH). There was the perception that DBS was unwilling to accept HDCs and move toward licensing any products, such as polio vaccines, produced in them. As a result, there were complaints to Congress. Eventually, political pressure was placed on the NIH, which included a Senate hearing, and a political decision was made to transfer the regulatory authority for biological products from NIH to the Food and Drug Administration (FDA) (4). That is why biological products are currently regulated by the FDA. The relevance of this sidebar to history is that cell substrate issues can be very sensitive and important not only from a safety standpoint, but because novel cell substrates may permit technological advances to current manufacturing capabilities that make it feasible to develop and introduce important new products. This was the case in the 1970s, and it continues to be true in the 2000s.

1970s

The next milestone was in the 1970s when interferon (IFN) was being studied as an interesting molecule with therapeutic potential. Unfortunately, it was in short supply, because at that point in time it was derived from primary human lymphocytes, which placed a considerable limitation on the amount that could be generated. It was recognized that the IFN supply problem could be solved by the use of human lymphoma cells (Namalwa) for IFN production. However, the problem with the Namalwa cells was that they contain an integrated Epstein Bar virus and undefined cancer-related DNA sequences. Nevertheless, after considerable discussion, there was agreement on the part of regulatory agencies in the U.S. and Europe to allow the use of Namalwa cells for the production of IFN. The decision was based on the fact that IFN was not a replicating agent, and that advances in purification and validation demonstrated, to the extent technology would allow, that virus, viral sequences, and cellular DNA were undetectable (5).

At the time, cellular DNA was emerging as a major safety issue, which Maurice Hilleman suggested could be easily solved by the use of DNase to inactivate potential biological activity of contaminating cellular DNA. While logical, unfortunately, as is often the case when one recommends a new approach to solve an old problem, the suggestion was met with a great deal of resistance. The argument against DNase hinged on the potential problems associated with adding yet another element to the manufacturing process, and as a result DNase was abandoned until relatively recently.

1980s

The 1980s ushered in the era of recombinant DNA technology and monoclonal antibodies. Along with it, cell substrate issues emerged once again because the cell substrates being used were tumorigenic continuous cell lines (CCLs) such as Chinese hamster ovary (CHO). CCLs were selected because, in contrast to HDCs, they were easy to transfect, could be grown to high densities, and express high levels of product. DNA derived from CCLs and potential viruses in CCLs, both of which might contaminate the final products, were the main foci of concern. Transforming proteins also were considered, but they were dismissed as trivial for solid scientific reasons (6).

The issues of viruses and viral safety were addressed initially by characterization studies that attempted to identify what type of viruses might be in the cell substrates. Additionally, manufacturing processes were configured to remove or clear any potential viruses that might be in the cell substrate.

DNA was addressed in the early days by setting limits on the amount of DNA in the final product. Interestingly, after more than 25 years, there is still discussion about what is an appropriate limit for cellular DNA in products. While current WHO guidance is 10 ng per dose, there is growing recognition that many factors such as the size of the DNA fragments should be taken into account, and that the acceptable amount of DNA in a product should be decided on a case-by-case basis.

2000s

In contemporary times, several interesting additions to types of cell substrate are being explored and used in product research and development. For example, HeLa cells are now being used as a substrate for human immunodeficiency virus (HIV) vaccines. This illustrates the meaning of “back to the future” in the title of this paper, since cell substrate issues began in the 1950s with HeLa cells. Additionally, cell lines of human embryonic origin transformed with adenovirus genes, such as PER.C6 and 293-ORF6, are being used for production of influenza and HIV vaccines. Other new cell substrates being used for vaccine production include Madin-Darby Canine Kidney (MDCK) derived from dog kidney for influenza vaccines and insect cells for human papillomavirus vaccine.

Lessons from the Past

Looking back at the history of cell substrates, one might reasonably ask if we have learned anything. One of the major lessons learned is that cell substrates have been, and probably will continue to be, a focus of attention, anxieties, and concern with respect to the safety of products derived from them. Product safety and the acceptability of a cell substrate for production of a given product are the two interlinked issues that have dominated our attention for the past 50 years. With regard to safety, there are two components of concern: transmissible agents and elements. Transmissible agents may cause infectious diseases such as HIV. There are also agents such as human T-cell leukemia virus type 1 that can cause cancer and agents such as prions that are associated with encephalopathies. The single most important transmissible element is DNA, which may contain specific genes such as activated oncogenes (e.g., H-ras).

The acceptability of a cell substrate at any given point in time depends on the state of knowledge and our understanding of the risks. If we look back to the 1950s when HeLa cells were initially being considered, the general state of knowledge of cancer and human tumor cells was relatively primitive. We have come a very long way since then, and our understanding of risks is significantly better than it was in the 1950s. An example of progress is that our understanding of the risks associated with HeLa cells and how those risks may be mitigated has allowed for their use as a cell substrate for an experimental HIV vaccine.

Along with a better understanding of risk, we now possess a greater ability to characterize cells. To the extent that cells can be thoroughly characterized, we are able to understand risks inherent to the cells. With the risks defined, and the elements associated with the risk identified, it may be possible to mitigate risks during manufacturing and to reduce if not eliminate them.

Another important point that we have learned is that the starting point for assuring a safe product is the characterization of the raw materials, including the cell substrate. The point here is that if the risks in the raw materials have not been defined, then the manufacturing process may not be capable of minimizing or eliminating them.

The final point in terms of what has been learned is that there has been an enhanced willingness to accept what is being termed a defined level of risk. Increasingly, regulatory authorities have been looking at the acceptability issue in terms of a defined level of risk. In the case of vaccines, that risk may be one in a million or one in 10 million. In the case of the presence of cellular DNA, the concern is related to the risk of activated oncogenes. Based on experimental data as well as on theoretical calculations, grams of cellular DNA would be required to provide sufficient activated oncogenes to cause a biological effect in an animal model. No one, to my knowledge, has demonstrated the development of a tumor from cellular DNA derived from human or animal tumor cells.

Guidelines

Over the years, a variety of cell substrate guidelines have been developed by national and international organizations. One of those is the World Health Organization (WHO). The most recent WHO document, TRS 878, Annex 1, was issued in 1998. Due to advances in knowledge, the WHO decided to update this guidance, and appointed a study group in 2006 to revise the document. A draft revision is expected to be considered by the Expert Committee on Biological Standardization in late 2010.

Summary

In my opinion, it is fair to say that the risks associated with cell substrates today are essentially the same as those that were identified over 50 years ago, namely, transmissible agents and transmissible elements. The concept of acceptability of cells for the production of products is being replaced with assessing risk and benefit on a case-by-case basis, taking into consideration the data that are generated on the characterization of a cell substrate and how the manufacturing process addresses the risks that have been defined. In addition, when one looks at risk and benefit, the seriousness of the disease and the availability of other therapies are two other important components that regulatory agencies are increasingly taking into consideration.

So after 50 years, we are really back where we started: Is HeLa acceptable or not? However, this time it is not just HeLa cells; it is a whole variety of cell types including other human tumor cells that are being considered as substrates. The difference between now and the 1950s is that the outcome of those considerations is dramatically different. This turn of events is best summarized by T. S. Elliot: “And the end of all our exploring will be to arrive where we started, and know the place for the first time” (7).

Conflict of Interest Declaration

The author declares that he has no competing interests.