Biosafety for Large-scale Operations Using Recombinant DNA Technology

September 1, 2005
Richard Fink, SM (NRM)

Enda Moran, Ph.D.

BioPharm International, BioPharm International-09-01-2005, Volume 18, Issue 9

Current best practices of containment reduce the risks associated with biotech development.

Over the last 30 years, recombinant DNA (rDNA) technology has evolved from an exciting new "ground-breaking" set of techniques into an established science. Industry and public concerns regarding the safety of rDNA technology developed in parallel with the emergence of the science. In the beginning of the rDNA era, the possibility of recombinant organisms becoming pathogenic or otherwise causing harm to humans or the environment was an unfamiliar and previously unassessed risk. Now, prokaryotic and eukaryotic expression systems engineered using rDNA methods are routinely used for the production of a wide range of products for medical, diagnostic, and other uses. The use of biological and physical containment methods, resulting from risk assessment experiments, and a history of safe use of recombinant organisms to date, have alleviated initial concerns regarding the safety of this technology.

In this paper, we review the evolution of biosafety practices and detail current industry regulations and approaches for treatment of waste from production operations using rDNA-derived expression systems. The primary focus centers on a possible hazard associated with rDNA technology that has not been extensively considered to date — the potential for chromosomal DNA from recombinant organisms to cause harm to humans or the environment via horizontal gene transfer to bacteria. As we prepared to start-up our new biopharmaceuticals manufacturing plant in Ireland in 2004, a key question for us was "Does the DNA from inactivated CHO cells that we routinely deliver to our site waste hold tanks, and subsequently to the municipal waste treatment facility, pose any threat to the environment?" The scientific information and thought processes we engaged in assessing this risk are described here.


In 1979, Dr. Rollin Hotchkiss wrote to the US NIH Recombinant DNA Advisory Committee: "I did in 1950, after some deliberation, perform the first drug resistance DNA transformations, and in 1964 and 1965 took part in early warning against indiscriminant 'transformations' that were then being imagined." The extent of scientists' apprehensions regarding this new field of rDNA technology increased and at the June 1973 Gordon Research Conference on nucleic acids in New Hampton, NH, the US NIH was asked to study the issues of safety for laboratory workers. Concern was so great that a number of notable scientists called for a moratorium on rDNA research. These events led directly to the 1975 Asilomar Conference and the creation of NIH oversight of rDNA research in the US. Shortly afterwards, Canada, the United Kingdom, and other countries instituted similar regulations and oversight bodies. The initial regulations were fairly restrictive, requiring moderate to high physical and biological containment. Risk assessment experiments were initiated to quantify risks based on data rather than supposition.

Risk Assessment Experiments

One of the worries early in the history of rDNA technology was that a harmless microorganism could become pathogenic. A hybrid Shigella flexneri/E. coli K-12 was created by cloning a number of Shigella surface antigens into the K-12. Volunteers ingested this recombinant either in a single 3x1010 dose or three 3x1010 doses, each one week apart. None of the volunteers became ill and the K-12 was undetectable after five days for those receiving a single dose and within 13 days for those receiving multiple doses.1 This and other experiments that followed led to the conclusion that a non-pathogenic organism cannot become pathogenic from the insertion of a single well-known protein not involved in pathogenicity. Moreover, even multiple pathogen genes will not necessarily transform a non-pathogen into a pathogen.2,3

Another concern was that an individual or animal would become colonized with a recombinant organism producing a hormone or other pharmacologically active substance. It was calculated that even if one's intestines were completely colonized by E. coli K-12, producing insulin, human growth hormone, or interferon and these proteins were not digested, the quantities produced would be too low to have any effect. This theoretical exercise was then experimentally substantiated in animal experiments.2

A final concern centers on the possible escape of a recombinant organism and subsequent disturbance of the environment. Adverse environmental impact would require escape, transport to a suitable niche, expression of the recombinant protein, successful multiplication in the face of competing indigenous organisms, and then spread, resulting in damage due to expression of the rDNA gene.2,4,5 While escape is possible, especially at scales (>10 L) governed by Good Large Scale Practices (GLSPs), the subsequent steps are much less likely to occur. First, the rDNA organisms are leaving a man-made world optimized for growth (i.e., temperature control, nutrients, etc.) and are entering the natural world, which has a poorer nutritional status, and would very likely encounter a sudden temperature shift of 10 degree to 15 degree C. While organisms can adapt to these conditions, it takes time. During that time, they are being preyed upon and out-maneuvered for available nutrients. Unless the rDNA enables the organism to use a substrate that the native population cannot metabolize, or enables the organism to reproduce at a higher rate, it is unlikely that the organism will survive.


Due to many experiments of the type noted earlier, the lack of injuries to laboratory workers, and the lack of evidence of harm,3,6 the restrictive requirements for working with rDNA have been somewhat relaxed. Certain experiments that required P-2 containment (now called Containment Level 2/Biosafety Level 2 or BL-2) at the beginning of the rDNA era are now rated as exempt from regulation. One "organism" category affected by this change was mammalian cell lines such as Chinese Hamster Ovary (CHO) cells. It was readily noted that mammalian cells in culture "had no capacity for propagation outside the laboratory."7 It is now recognized that insertion of a well-characterized gene into organisms with a history of safe use, such as CHO cells, does "not raise any safety considerations beyond those that might be posed by the [gene] products themselves."6 It is noteworthy that after approximately 30 years of rDNA research and industrial use, no harm to the environment, no infections in laboratory workers, and no deaths due to exposure to rDNA organisms have been documented.


European Community member states implement Directive 98/81/EC through their own legislation governing the contained use of genetically modified microorganisms (GMMs). As might be expected, there is much similarity between the regulation requirements of member states covering topics such as risk assessment, containment and control, disposal, etc. There are clear edicts on expectations regarding biowaste treatment before disposal. For example, the Genetically Modified Organisms (Contained Use) Regulations of the UK (2000)8 and Ireland (2001)9 both require that GMM-containing waste generated from Containment Levels 2-4 activities is inactivated by validated means before disposal. The UK regulations state that Containment Level 1 activities (e.g., in Europe, these involve CHO cell lines) also must use validated methods of inactivation, while in this category Irish regulations state it is optional (though expectations are that biopharmaceutical companies will inactivate waste from large-scale production operations).

The well-characterized nature of CHO cell lineages and the over two decades of safe use of these cell lines in large-scale production operations have exempted their categorization in the US from the containment levels BL1-LS to BL3-LS.10 The categorization of CHO cell lines under the umbrella of GLSP has resulted in decontamination/inactivation requirements being largely particular to individual state and local government legislation. Hence, certain companies will inactivate CHO cell waste prior to disposal; others will not. Since CHO and other mammalian cells cannot persist in the environment, the only risk factor might be the existence of their chromosomal DNA and subsequent uptake by competent bacterial cells. Table 1 compares inactivation practices for large-scale operations using mammalian cells in certain states in the US and Europe.

Table 1. Examples of Large-scale Biowaste Treatments Prior to Disposal in Biotechnology Companies Currently Operating in the US and Europe as of April 2004. All companies use mammalian cell technology (CHO or NS0 cell lines) for the production of therapeutic proteins. Three different companies in west and east coast states of the US are represented. Information gathered through personal experiences and personal communications to the authors. 3=method used, x =method not used.

Companies implementing inactivation methods are generally expected to use validated procedures. Validation demonstrates that the inactivation method is suitable for purpose and removes the need to monitor for the GMM in waste streams after an inactivation event during routine operations. For certain classes of operation, usually Containment Level 2 and above, the capability of monitoring for the presence of GMM outside of the contained process is expected under UK and Irish guidelines previously mentioned. The need for this is dependent on practical feasibility and risk assessment outcome and is usually considered on a case-by-case basis. However, for well-characterized Group/Class 1 GMMs such as CHO cell lines used under Containment Level 1 (in Europe), monitoring for the GMM would be expected only under exceptional circumstances, e.g., if there was a serious breach of large-scale process containment.


Each year, the decay of pollen, leaves, fruit, and animals results in thousands of tons of DNA being released into the environment,11 with measurable levels being reported in soils,12 marine waters, freshwater lakes, rivers, and ponds.13 Given that eukaryotic organisms do not exchange genes via conjugation (transposons, plasmids) or transduction (bacteriophage) the only route available for the natural transfer of eukaryotic genes to prokaryotes is via direct gene uptake. For a transfer to take place there must be a release of functional DNA, persistence of the DNA, and uptake by a competent cell.14,15 For the foreign gene to have an environmental effect, it must be expressed and translated. A number of factors should be considered when questioning if eukaryotic genes in the natural environment can transfer to prokaryotic organisms. These factors include the fate of the DNA in the environment and its stability relative to the time required for a competent cell (a cell capable of taking up DNA) to take in the genetic material. Other factors include the fate of the DNA in the prokaryotic cell, i.e., it could be enzymatically digested, circularized into a plasmid, or integrated into the genome and, finally, whether it would become a functional, expressed gene (Figure 1).

Figure 1. Possible Fates of DNA after Transformation of a Competent Bacterial Cell

DNA, while abundant in the environment, is continuously under attack by physical shearing forces, chemical modifications, and microbially secreted nucleases. It is often not available for transformation because it binds to solids and other matter(Figure 2). Most bacteria in the environment are not normally in a competent state, and those that are have low transformation efficiencies.13

Figure 2. Fate of Eukaryotic DNA Released into the Soil Environment

Under laboratory conditions optimized for transformation using bacterial plasmids, transformation of two ubiquitous soil bacteria was completed in 45-120 minutes.13,14 In the natural environment, completion of transformation would probably take longer. The estimated half-life of DNA in wastewater is 1 to 13.8 minutes. Thus, the probability of CHO cell DNA surviving long enough in wastewater to transform bacteria at ambient temperatures is very low. There are barriers to expression of the transforming DNA if HGT does occur, thus preventing potential environmental impact. These include the failure of DNA stabilization mechanisms within the cell (as cited earlier; integration into the bacterial genome by homologous recombination or circularization to plasmid), the failure of accurate expression of the gene, and post-translational modification of the gene product.

The principal barrier to successful transformation and expression is the lack of homology between the foreign DNA and the competent cell's DNA. An exponential decrease in recombination events with an increase in divergence of DNA sequence has been observed with enterobacteria and Bacillus spp.16 No gene transfer has been found to occur from genetically modified plants to soil bacteria despite the detection of the transgene in the soil.11,15,17,18 Extensive review of a variety of bacterial genomes has revealed only six genes that are tentatively believed to have been transferred via HGT from eukaryotes to bacteria. Thus, despite the great amount of DNA that prokaryotes are exposed to and the great amount of time that prokaryotes and eukaryotes have coexisted, the incorporation of eukaryotic DNA into bacteria, as best can be determined, has been a very rare event.19,20, 21,22

Recombinant CHO cell lines are used in the biotechnology industry for the production of many important biopharmaceutical and diagnostic products. Is there a significant risk that naturally occurring prokaryotic organisms in the waste stream or in the waste treatment plant will be transformed when exposed to cell culture waste containing viable CHO cells or their DNA? Consider, for example, a large-scale operation using CHO cells to produce an IgG antibody. A risk assessment of harm or damage resulting from a HGT event would consider the following:

  • The genetic information for any eukaryotic product is already in the environment due to death and decay of eukaryotic organisms. It is likely that if this DNA were capable of transforming a prokaryotic organism, it already would have occurred.

  • Many eukaryotic proteins such as IgG must be glycosylated and intricately folded to become biologically active, and prokaryotic organisms lack the ability to do this. Thus, even if transformation occurred with a complete coding gene, only an inactive form of these proteins would be produced.

  • The barriers associated with HGT and progression to transcription and translation of the genetic information as described earlier would have to be bypassed. In this hypothetical case, the possibility of homology between the IgG genetic information and the bacterial chromosomal or plasmid DNA is negligible. This lack of homology would apply to a wide range of eukaryotic genetic information.

  • For stable carriage of eukaryotic information in a prokaryote, the genetic element would have to assist a bacterium in its competition with other organisms in the environment. As noted previously, this is very unlikely to occur.

The answer to the question of risk posed earlier is that while we cannot definitively discount all risk, the probabilities of damage due to horizontal transfer of eukaryotic DNA to prokaryotes is vanishingly small and is, thus, a negligible risk.


While there is some evidence that there has been HGT from eukaryotes to prokaryotes, this has been a very rare event. There are significant barriers to the transfer of eukaryotic DNA to prokaryotic organisms including lack of homology with the recipient's DNA complement and the degradation of DNA in the environment, particularly the rapid degradation in wastewater. Therefore, the probability of genetic transfer of the free recombinant CHO cell gene sequences in process wastewater to bacteria is extremely remote due to (a) the rapid degradation of DNA in wastewater, (b) the low percentage of competent bacteria in the environment, (c) the low transformation efficiency of the competent bacteria that could be present and, (d) the lack of homology between bacterial host DNA and the mammalian cell DNA. The possibility that there would be an adverse environmental impact is more remote because even if the extremely unlikely transformation event occurred, there would have to be an expression of protein that would have to confer a selective advantage. Without the selective pressure, the transferred gene would be lost due to random mutation or deletion.23,24,25

Thus, we conclude that the current best practice of containment as specified by the US NIH and the EU is effective in containing the risks involved in biotechnology development and large-scale production operations using CHO cells or other cell lines in the Containment Level 1 or GLSP categories. Additionally, eukaryotic cell-derived DNA in biotechnology facility wastewater poses no known or new risk that warrants monitoring for the DNA itself or for transformation of bacteria in wastewater.

All rights reserved. Reprinted with permission from the American Biological Safety Association (ABSA), Mundelein, IL. Originally published in Applied Biosafety: Journal of the American Biological Safety Association, 10(1), 2005.

Enda Moran, Ph.D. is associate director, process development Wyeth Medica Ireland, The Wyeth BioPharma Campus at Grange Castle, Grange Castle International Business Park, Clondalkin, Dublin 22, Rep. of Ireland.

Richard Fink, SM (NRM) is biosafety officer, Wyeth BioPharma, 1 Burtt Rd., Andover, MA 01810, USA.


1. Levine MM, Formal SB, Gemski Jr. P, DuPont HL, Hornick RB, Snyder MJ. Studies with a new generation of oral attenuated Shigella vaccine: Escherichia coli bearing surface antigens of Shigella flexneri. Journal of Infectious Diseases. 1977;136:577-582.

2. Winkler KC. Safety and risks of handling organisms with recombinant DNA. II. Risk assessment of small scale recombinant DNA procedures in laboratories. Chimicaoggi. 1988;April:49-53.

3. WHO. WHO Meeting, Dublin. Health impact of biotechnology. Report on a working group. 1983.

4. Liberman DF, Israeli E, Fink R. Risk assessment of biological hazards in the biotechnology industry. Occupational Medicine State of the Art Reviews. 1991;6:285-299.

5. Liberman DF, Wolfe LB, Fink R, Gilman E. Biological considerations for environmental release of transgenic organisms and plants. In M. A. Levine & E. Israeli (Eds.), Engineered organisms in environmental settings. Boca Raton, FL: CRC Press, Inc. 1996;41-63.

6. OECD. Recombinant DNA safety considerations. 1986;Paris: Organisation for Economic Co-Operation and Development.

7. National Institutes of Health. NIH guidelines for the use of recombinant DNA. Washington, DC: U.S. Department of Health and Human Services. 1980;Nov. 21.

8. Health & Safety Executive, UK. Genetically modified organisms (contained use) Regulations, Health & Safety Executive, UK. 2000

9. Dublin Stationery Office. Genetically modified organisms (contained use) regulations. Statutory Instruments S.I,. No. 73. 2001

10. National Institutes of Health. NIH guidelines for research involving recombinant DNA molecules. Washington, DC: U.S. Department of Health and Human Services. 2002;April.

11. Dale PJ, Clarke B, Fontes EMG. Potential for the environmental impact of transgenic crops. Nature Biotechnology. 2002;20:567-574.

12. Paget E, Lebrun M, Freyssinet G, Simonet P. The fate of recombinant plant DNA in soil. European Journal of Soil Biology. 1998;34:81-88.

13. Lorenz MG, Wackernagel W. Bacterial gene transfer by natural genetic transformation in the environment. Microbiology Review. 1994;58:563-602.

14. Bruns S, Reipschlager MG, Lorenz G, Wackernagel W. Characterization of natural transformation of the soil bacteria Pseudomonas stutzeri and Acinetobacter calcoaceticus by chromosomal and plasmid DNA. In M. J. Gauthier (Ed.), Gene transfers and environment. London: Springer-Verlag. 1992;115-126.

15. Kay E, Vogel TM, Bertolla F, Nalin R, Simonet P. In situ transfer of antibiotic resistance genes from transgenic (transplastomic) tobacco plants to bacteria. Applied Environmental Microbiolgy. 2002;68:3345-3351.

16. Nielsen KM, Bones AM, Smalla K, van Elsas JD. Horizontal gene transfer from transgenic plants to terrestrial bacteria—A rare event? FEMS Microbiology Review. 1998;2:79-103.

17. Gebhard F, Smalla K. Monitoring field releases of genetically modified sugar beets for persistence of transgenic plant DNA and horizontal gene transfer. FEMS Microbiol. Ecol. 1999;28:261-272.

18. de Vries J, Heine M, Harms K, Wackernagel W. Spread of recombinant DNA by roots and pollen of transgenic potato plants, identified by highly specific biomonitoring using natural transformation of an Acinetobacter sp. Appied Environmental Microbiology. 2003;69:4455-4462.

19. Doolittle RF, Feng DF, Anderson KL, Alberro MR. A naturally occurring horizontal gene transfer form a eukaryote to a prokaryote. J. Mol. Evol. 1990;31:383-388.

20. Gogarten P, Doolittle WF, Lawrence G. Prokaryotic evolution in light of gene transfer. Mol. Bio. Evol. 2002;19:2226-2238.

21. Syvanen M. Horizontal gene transfer: Evidence and possible consequences. Annual Review of Genetics. 1994;28:237-261.

22. Smith MW, Feng DB, Doolittle RF. Evolution by acquisition: The case for horizontal gene transfers. Trends in Biochemical Science. 1992;17:489-493.

23. Kurtland CG, Canback B, Berg, OG. Horizontal gene transfer: A critical view. PNAS. 2003;100:658-662.

24. Lawrence JG, Ochman, H Molecular archaeology of the Escherichia coli genome. PNAS. 1998;95:9413-9417.

25. Ochman H, Lawrence JG, Grolsman EA. Lateral gene transfer and the nature of bacterial innovation. Nature. 2000;405:299-304.