USDA Animal Plant Health Inspection Service (APHIS)
Environmental Assessment for Field Test of Permit of
Ventria Bioscience rice genetically engineered to express human lactoferrin
USDA/APHIS Docket No. 05-006-1
Consumers Union* appreciates the opportunity to comment on the USDA/APHIS Environmental Assessment (EA) for Ventria’s plant-made pharmaceutical (PMP) rice variety that has been genetically engineered to express human lactoferrin. We restrict our comments to issues associated with the human (and animal) health. For the reasons laid out below, we believe that the EA is inadequate and contains internal inconsistencies. Consequently, we urge USDA/APHIS to deny Ventria’s permit application and to not issue any further permits for field trial of any of Ventria’s PMP rice varieties before a full regulatory review has been completed by FDA and EPA on the human health and ecological implications, respectively, of this PMP rice variety. In our view, USDA should not grant permits for any outdoor trials of PMPs. However, it certainly should not do so before FDA has adequately addressed any safety issues associated with the PMP. We also believe that USDA/APHIS should not regard this EA as final until the potential impacts on non-target species—particularly birds (especially waterfowl) and rodents—have been investigated in more scientific detail.
We have found a number of serious deficiencies and potential inconsistencies in Ventria’s EA vis-à-vis the potential human health impact of their PMP rice variety expressing human lactoferrin. Many of the deficiencies relate to issues associated with allergenicity. Allergenicity is one of the key potential human health impacts associated with genetically engineered proteins. It is important to assess any risk that recombinant human lactoferrin (rhLf), a protein normally found in milk (among other body fluids), but which is being introduced here into rice, might pose, especially to people with milk allergies. Although there is no single test for allergenicity, there has been global agreement on the types of tests that should be done (FAO/WHO 2001). In particular, characteristics of allergens include: that they have amino acid sequence similarity to known allergens, they are glycoproteins, they have resistance to digestion, and display heat stability.
Ventria FDA Submission Contradicts EA on Sequence Homology
The EA states that Ventria’s studies conclude that “no [amino acid] sequence homology of the recombinant lactoferrin to known toxicants, allergens, or proteins likely to harm non-target organisms. . . . The SWISS-PROT AND TrEMBL databases were utilized and no amino acid sequence homology was found between lactoferrin and known allergens.” italics added. This statement appears to directly contradicts statements Ventria made in their November 24, 2003 submission to FDA: “A search against a database of known allergens, identified in the SWISS-PROT and TrEMBL databases, also using the techniques described by Gendel, found 52% amino acid sequence homology between human lactoferrin and chick ovotransferrin (conalbumin), which is a known allergen” italics added (Bethell, 2003: 46). Ventria then seeks to dismiss this finding by arguing that human lactoferrin can’t be an allergen because i) it’s found in human breast milk and ii) it has never been identified as an allergen. But the fact remains that Ventria did find sequence similarity between lactoferrin and a known human allergen.
We are very concerned that Ventria’s conclusions in these two documents appear to directly contradict each other. We believe the EA should be revised to incorporate the information submitted to FDA and that the implications of this information in terms of allergy risk should be fully addressed rather than dismissed as Ventria does. We are especially concerned that risks to people with milk allergies be addressed. A prominent French allergy specialist, Dr. Jean-Michel Wal, found that 41 out of 92 milk-allergic patients had detectable IgE antibodies to bovine lactoferrin and concluded that bovine lactoferrin may be an important allergen (Wal, 1998). Bovine and human lactoferrin share 68% amino acid identity (Groenink et al., 1999).
Both Codex Alimentarius ((CAG/GL 45-2003) and a global expert meeting on allergenicity (FAO/WHO, 2001) have proposed that any global sequence homology between a transgenic protein and known allergen that exceeds 35% would be considered significant and should trigger further research. The sequence similarity between rhLf and the egg allergen conalbumin and bovine lactoferrin is 52% and 68%, respectively. Both figures greatly exceed the 35% threshold suggest by both Codex Alimentarius and FAO/WHO. These data clearly show that rhLf has significant sequence homology with two known allergens. These findings should clearly trigger further research into the potential allergenicity of rhLf.
In order to assess sequence similarity to known allergens, Ventria must be required to provide the complete amino acid sequence information for their rhLf. The exact amino acid sequence of the plant-derived lactoferrin has not been determined. When comparing the physical and chemical properties of recombinant human lactoferrin (rhLf) and purified human lactoferrin (hLf), the EA notes that “N-terminal sequence were found to be identical”; on this basis APHIS concludes that amino acid sequence of rhLf and hLf are identical. However, the EPA’s Scientific Advisory Panel has specifically stated that this approach is inadequate and has recommended full-length amino acid sequencing of plant-produced recombinant proteins, noting that one or two point mutations can modify allergenic property: “sequence similarity of full length amino acid sequence (highly undesirable is the sequence analysis of 10-15 N and/or C-terminal amino acids and up to three short internal protein sequences). For example, two isoforms of -lactoglobulin i.e. genetic variants which differed only by two point mutations on residues 64 and 118 (D and V versus G and A) showed modified allergic properties [Wall, 1998]” italics added (SAP, 2000, pg. 14). Furthermore, the process of genetic engineering is not exact, so that during the insertion process, small point mutations may occur in the inserted DNA that could lead to amino acid substitutions in the plant-produced rhLf.
Further, under molecular characterization, the EA states that Ventria’s data demonstrate “that there are approximately 6 copies of the lactoferrin coding sequence integrated into the rice genome.” However, in the submission to FDA, Ventria states that “it is estimated there are 8 copies of the complete 3156 bp chimeric lactoferrin expression cassette” (Bethell, 2003: 20). This apparent discrepancy between data submitted to USDA/APHIS and data submitted to FDA should be addressed and resolved before the EA is considered final.
EA Contradicts Itself on Resistance to Digestion
The EA suggests that rhLf is rapidly digested and so wouldn’t be a problem (one of the characteristics of known allergens is stability to digestion). In the in vitro pepsin digestion experiment, “Ventria’s rice-derived lactoferrin (rhLf) is equivalent to native human lactoferrin (hLf) which is rapidly degraded (<30 sec) in gastric fluid.” This rapid digestion (<30 seconds) clearly suggests that rhLf doesn’t survive digestion. However, we note that Ventria proposes to extract rhLf from the rice and use it “as supplements in yogurts, meal replacement and performance beverages, bars (for example granola bars), and in nutritional supplement drinks.” If rhLf is degraded as quickly as Ventria suggests, then it wouldn’t be orally active. Yet, since rhLf is supposed to have beneficial (antimicrobial) activity, then it would have to be orally active. If it isn’t orally active, then there would be no real benefit to putting it into the suggested products. In addition, Ventria refers to a study where chicks fed the transgenic rice had improved health and growth rates compared to controls. This study clearly shows that the rhLF (and recombinant human lysozyme) in the rice was orally active, suggesting that it does survive digestion. In addition, this chick study also noted that previous studies found that lysozyme and lactoferrin “are highly resistant to hydrolysis by acids and proteases and to digestion in the gastrointestinal tract” italics added (Humphrey et al., 2002: 1214). Indeed, one of the cited studies is titled, “Functional fragments of ingested lactoferrin are resistant to proteolytic degradation in the gastrointestinal tract of adult rats” (Kuwata et al., 2001). Finally, a study done by Dr. Bo Lonnerdal using Ventria’s rhLf also demonstrated that that rhLf is resistant to digestion in an in vitro digestive system (Lonnerdal, 2002). Dr. Lonnerdal’s system consisted of exposing the recombinant human milk proteins to low pH and pepsin for 30 minutes at 37 C, then adjusting the pH to pH 7, adding pancreatin (mixture of pancreatic enzymes), and then incubating for 30 to 60 minutes. Dr. Lonnerdal found that for “all three proteins [recombinant versions of human lactoferrin, lysozyme and -1-antitrypsin] we studied, activities remained after treatment” (Lonnerdal, 2002: 220S). So, the data do appear to show that rhLf does survive gastric digestion. We are very concerned that this EA contains these internal inconsistencies and believe they should be addressed before this document is considered final. In addition if, as appears to be the case, rhLf is orally active and effects growth of birds, this has important environmental implications. What are the consequences if this rice is consumed by wildlife? What might be the effect on ecological relationships? Could there be dramatic population increases in pest birds or mammals, such as Canada geese and starlings or mice, due to their increased health? These issues must be explored. EA Contradicted by Omitted Studies on Heat Stability
The EA also suggests that rhLf is not heat stable, noting that after using a commercial rice cooker for 20 minutes, “lactoferrin protein could no longer be detected.” However, a heat stability study done by Dr. Lonnerdal, published in the Journal of the American College of Nutrition, found rhLf to be heat stable. Dr. Lonnerdal did the heat stability studies as part of research into the use of rhLf in infant formula and baby foods. As Dr. Lonnerdal noted: “If recombinant human milk proteins are to be added to infant formula or baby foods, some degree of processing may be involved. We therefore exposed the recombinant proteins [human lactoferrin, lysozyme and -1-antitrypsin], both in pure form in solution and as added to infant formula, to various heat treatments, ranging from 78-100 C for 8 seconds up to 30 minutes. Except for the most severe treatment, 100 C for 5 minutes, which partially inactivated both recombinant and native human milk proteins, these proteins maintained activities similar to those of the native proteins” (Lonnerdal, 2002: 220S). Thus, rhLf does appear to be heat stable. The implications of this information in terms of allergy risk should be thoroughly addressed.
The EA states that the “only biochemical difference detected between the recombinant lactoferrin and the purified human lactoferrin were differences in glycosylation patterns.” The EA dismisses this difference by noting that “Ventria has demonstrated that the digestibility between rhLf and hLf are similar, showing that the stability of the protein has not been affected by glycosylation.” Since most known human allergens are glycoproteins, the fact that rhLf is glycosylated differently than the native hLf is potentially very significant. The different glycosylation patterns between rhLF and hLf could have a direct impact on the potential allergenicity. The fact that there may be no real change in digestibility is not comforting because there is clear evidence that hLf is resistant to digestion. The implications of theses findings in terms of allergy risk should be fully discussed and addressed.
Conclusions on Allergenicity
In sum, in terms of allergenicity, rhLf does have a number of traits—sequence similarity to known allergens, heat stability, resistance to digestion, being a glycoprotein—in common with known allergens. The fact that bovine lactoferrin may be an allergen is also of concern. Although Ventria, in their submission to FDA, argues that rhLf can’t be an allergen because it’s found in breast milk, their reasoning is spurious. By this reasoning, any human-derived drug couldn’t cause allergic reactions because the drug is normally found in the human body. However, it is well known that protein drugs derived from humans and produced via genetic engineering, can provoke an immune system response to the recombinant human protein and sometimes to the native human version. Thus, some patients receiving recombinant blood clotting Factor VIII and the multiple sclerosis drug beta-interferon develop antibodies to the recombinant proteins, which reduces the drug’s potency (Freese et al., 2003). In another case, Amgen’s recombinant megakaryocyte growth and development factor (MGDF) was discontinued in clinical trials after some patients developed antibodies to both the recombinant MGDF as well as the natural version of MGDF, which resulted in bleeding. Such immune system reactions to recombinant human proteins were unexpected as the recombinant and human versions of the proteins were identical in amino acid sequence and glycosylation patterns. Even though the scientists couldn’t detect differences between the recombinant and human versions of the proteins, the immune systems of the affected patients were able to distinguish between them. If these human protein drugs, produced using tightly-controlled fermentation using mammalian cell culture, could result in such unexpected immune system effects, why should we think that something similar could not happen with rice-produced rhLf compared to hLf?
Finally, since rhLf does have a number of traits—sequence similarity to known allergens, heat stability, resistance to digestion, being a glycoprotein—in common with known allergens and since rhLf has a different glycosylation pattern than hLf, there still remains a question about the potential allergenicity of rhLf. Perhaps that is why noted food allergy expert Dr. Steven Taylor of the University of Nebraska stated that FDA’s regulations “will have to be rethought before rice-grown lactoferrin, and other human proteins made by genetically modified organisms, can be approved for production” (Pearson, 2002: ).
If rhLf is a potential allergen, then the issue of possible mixing with or contamination of conventional food rice becomes a serious concern. These possibilities need to be much more fully evaluated in the EA.
Other Potential Human Health Effects
Recombinant human lactoferrin could have other adverse health effects on humans besides potential allergenicity. Lactoferrin is found in numerous body secretions and has antibacterial, antimycotic, antineoplastic and anti-inflammatory activities. Lactoferrin is part of the body’s “iron withholding defense system,” which serves to find and bind to toxic quantities of iron. Since many microbes need iron, lactoferrin serves to starve those microbes of the needed iron. Work has shown that lactoferrin can suppress bacteria such as Listeria monocytogenes. Other work shows that lactoferrin and lysozyme act synergistically to protect against various microbes, including E. coli, Salmonella typhimurium and Vibrio cholerae; individually there’s no negative effect on the bacteria (Ellison and Giehl, 1991).
While rhLf’s iron-binding properties make it a good defense against various microbes, there are microbes that can feed off the iron in lactoferrin and thereby possibly become more of a problem. Microbes such as the Trichomonas protozoa (Weinberg, 1999), Helicobacter pylori (Dhaenens et al., 1997) as well as members of the bacterial family Neissariaceae (which contains numerous venereal disease bacteria) (Vogel et al., 1997) can obtain iron from human lactoferrin. Since H. pylori is the major etiologic agent of chronic gastritis and a component of the etiology of gastric ulcers and carcinomas, supplying rhLf to people with H. pylori infections could make the infections worse rather than better.
Another potential adverse effect of rhLf is that is that it could induce an antibody response in some people. Antibodies to natural lactoferrin have been found in patients with autoimmune diseases such as systemic lupus erythematosis, rheumatoid arthritis and primary scerosing cholangitis (Skogh and Peen, 1993; Afeltra et al., 1996). Since rice-produced rhLf is glycosylated differently than native hLf, its allergic potential could be increased. The finding of antibody responses to many of the human protein drugs produced via genetic engineering shows that an immune response to a recombinant version of a human protein can happen.
As a result of the potentials for stimulating certain pathogens (by being a source of iron) and for provoking an allergic response, some scientists have advised a bit of caution in promoting the use of rhLf for disease reduction. Dr. Eugene Weinberg, of Indiana University, wrote a review article on the potential uses of rhLf in treating various medical conditions in which he cautioned, “an adverse response to the protein might occur if it were to stimulate antibody production or if it were to provide iron to the invading pathogen. . . . Precaution is needed, however, to avoid antigenic sensitization as well as introduction of the protein to tissues that may be infected with specific protozoa or bacteria that utilize Lf in their acquisition of host iron” (Weinberg, 2001).
Conclusion on Potential Human Health Effects
In sum, as we have discussed, there are serious inadequacies in this EA with respect to potential human health effect. These inadequacies must be fully addressed in any EA prior to granting a permit.
In our view, USDA should not grant permits for any outdoor trials of PMPs. However, it certainly should not do so before FDA has adequately addressed any safety issues associated with the PMP. In this regard, we note that Ventria submitted a “Food and Feed Assessment Summary” to FDA on November 23, 2003 as part of the voluntary consultation process for genetically engineered plants. To date, FDA has not issued a letter saying that they have no further concern, suggesting that FDA has further questions on this compound.
RhLF May Have adverse effects on non-target birds and mammals
The finding that certain human pathogenic bacteria and protozoa can be increased by the presence of rhLf also raises the possibility that other bacteria and protozoa that are pathogenic for various animal species, particularly birds, could also be increased by the presence of rhLf. USDA/APHIS argues that there will be no adverse impact on bird species that eat the rice, based on the single chick feeding study that found beneficial effects. That study did not involve chicks that were infected with any pathogens that may be able to obtain iron from rhLf. Thus, one cannot conclude from this single chick study that there would be no adverse effects on birds. USDA/APHIS should determine if there are any pathogenic microbes that infect birds that have the ability to obtain iron from lactoferrin and determine what would happen if they consumed Ventria’s rhLf-producing rice. Since the rhLF levels are very high in Ventria’s PMP rice, any iron-scavenging pathogens could significantly increase their numbers. Since rodents and other mammals could also consume the rice before or after harvest (rice spilled during harvesting), they could be adversely affected as well. It’s also possible that the potential protective effects of rhLf could be greater than the negative effects on birds and rodents. In either case, the possibility exists that there could be a disruption in the ecosystem as a result of consumption of this rice by birds or rodents. Thus, the EA is inadequate on this point. We therefore believe that USDA/APHIS should not regard this EA as final until the potential impacts on non-target species—particularly birds (especially waterfowl) and rodents—have been investigated in more scientific detail.
Afeltra, A., Sebastani, G.D., Galeazzi, M., Caccavo, D., Ferri, G.M., Marcolongo, R. and L. Bonomo. 1996. Antineutrophil cytoplasmic antibodies in synovial fluid and in serum of patients with rheumatoid arthritis and other types of synovitis. Journal of Rheumatology, 23: 10-15.
Bethell, D.R. 2003. Safety, Compositional and Nutritional Aspects of LF164 Rice Transformational Event. BNF: 082. Food and Feed Assessment Summary Voluntarily submitted to FDA on November 23, 2003.
Dhaenens, L., F. Szczebara, and MO Husson. 1997. Identification, characterization, and immunogenicity of the lactoferrin-binding protein from Helicobacter pylori. Infection and Immunity, 65(2): 514-518.
Ellison, R.T. III, and T.J. Giehl. 1991. Killing of gram-negative bacteria by lactoferrin and lysozyme. Journal of Clinical Investigation 88: 1080-1091.
FAO/WHO. 2001. Evaluation of Allergenicity of Genetically Modified Foods. Report of a Joint FAO/WHO Expert Consultation on Allergenicity of Foods Derived from Biotechnology, January 22-25, 2001. Rome, Italy.
Freese, B., Hansen, M. and D. Gurian-Sherman. 2004. Pharmaceutical Rice in California: Potential Risks to Consumers, the Environment and the California Rice Industry. July, 2004. http://www.foe.org/safefood/pharmrice.pdf
Groenink, J., Walgreen-Weterings, E., van’t Hof, W., Veerman, E.C.I. and A.V. Nieuw Amerongen. 1999. Cationic amphiphathic peptides, derived from bovine and human lactoferrins, with antimicrobial activity against oral pathogens. FEMS Microbiology Letters, 179: 217-222.
Humphrey, B.D., Huang, N. and K.C. Klasing. 2002. Rice expressing lactoferrin and lysozyme has antibiotic-like properties when fed to chicks. Journal of Nutrition, 132: 1214-1218.
Kuwata, H., Yamauchi, K., Teraguchi, S., Ushida, Y., Shimokawa, Y., Toida, T. and H. Hayawasa. 2001. Functional fragments of ingested lactoferrin are resistant to proteolytic degradation in the gastrointestinal tract of adult rats. Journal of Nutrition, 131: 2121-2127.
Lonnerdal, B. 2002. Expression of human milk proteins in plants. Journal of the American College of Nutrition, 21(3): 218S-221S.
Pearson, H. 2002. Milk in rice could curdle. Nature, April 26, 2002 http://www.nature.com/news/2002/020422/full/020422-19.html
SAP 2000. Session II – Mammalian Toxicity Assessment Guidelines for Protein Plant Pesticides. FIFRA Scientific Advisory Panel Report No. 2000-03B, September 28, 2000.
Skogh, T. and E. Peen. 1993. Lactoferrin, anti-lactoferrin antibodies and inflammatory disease. Pp. 533-538 in: Gross, W.L. (Ed). ANCA-associated Vasculitides: Immunological and Clinical Aspects. Plenum, New York.
Vogel, L. Geluk, F., Janse, H., Dankert, J. and L. van Alphen. 1997. Human lactoferrin receptor activity in non-encapsulated Haemophilus influenzae. FEMS Microbiology Letters, 156: 165-170.
Wal, J.-M. 1998. Strategies for assessment and identification of allergenicity in (novel) foods. International Dairy Journal, 8: 413-423.
Weinberg, E.D. 1999. The role of iron in protozoan and fungal infectious diseases. Journal of Eukaryote Microbiology, 46: 231-238.
———————-2001. Human lactoferrin: a novel therapeutic with broad spectrum potential. Journal of Pharmacy and Pharmacology, 53(10): 1303-1310.
* Consumers Union is a nonprofit membership organization chartered in 1936 under the laws of the State of New York to provide consumers with information, education and counsel about goods, services, health, and personal finances; and to initiate and cooperate with individual and group efforts to maintain and enhance the quality of life for consumers. Consumers Union’s income is derived solely from the sale of Consumer Reports, its other publications and from noncommercial contributions, grants and fees. In addition to reports on Consumers Union’s own product testing, Consumer Reports, with approximately 4.5 million paid circulation, regularly carries articles on health, product safety, marketplace economics and legislative, judicial and regulatory actions which affect consumer welfare. Consumers Union’s publications carry no advertising and receive no commercial support.