An Oily Mess

"You may not know our name, but we're an integral part of your everyday life" - Rhone-Poulenc Inc.

In 1995, the very first genetically engineered cotton crop entered our food supply. The resulting product, BXN® Cotton, was designed by Calgene Incorporated, in partnership with Rhone-Poulenc, a French chemical company. Rhone-Poulenc provided the bromoxynil, also known by the trade name Buctril®, while Calgene Incorporated provided the genetic technology to create cotton seeds that would be resistant to its herbicide. The new technology allows the crop to be oversprayed with bromoxynil, killing surrounding weeds while sparing the cotton plants. In principle, any resulting contamination of the cotton plant with bromoxynil residues is immaterial since the main product, cotton fabric, is not intended for consumption. But cotton plants not only provide bolls for clothing, but also cottonseed oil, hulls, and cotton gin used as animal feed. Following bromoxynil treatment, these cotton commodities will inevitably carry higher residues levels of bromoxynil than before genetic technology permitted overspraying of a previously sensitive crop. Calgene and Rhone-Poulenc argue that BXN®Cotton use one herbicide instead of many, which eliminates excessive costs to farmers who are used to buying a myriad of expensive herbicides in a season.

Bromoxynil is in the same class of herbicides as 2-4D, and 2,4,5T and like them, is used to control broad leaf weeds. BXN®Cotton is cottonseed created to be resistant to bromoxynil due to the incorporation of a bacterial gene into the cotton genome. The introduced bacterial gene acts by hydrolyzing nitrile groups into a carboxylic acid, 3,5-dibromo-4-hydroxybenzoic acid (DBHA) (U.S. EPA, 1997c). DBHA is a metabolite or product of bromoxynil nitrilization. Under non-engineered circumstances, bromoxynil acts by inhibiting photosynthesis in plants; but by introducing the nitrilase gene into specific crops the new gene acts to detoxify the bromoxynil, allowing the engineered crop to continue to live. Bromoxynil is a phenoxy herbicide. This class of herbicides regulates plant growth primarily by inhibiting photosynthesis.

Toxicity:
Bromoxynil octanoate, the active ingredient, is converted into bromomoxynil phenol when metabolized in mammals and in the environment. In "bridging studies," both of these compounds have been found to be toxicologically equivalent in the mammalian body (U.S. EPA, 1997a). Unlike many other herbicides which are water soluble, bromoxynil is soluble in fat, allowing it to bioaccumulate and concentrate within the fatty tissues of mammals. Also, the metabolite DBHA that occurs as a result of the newly introduced gene in BXN®Cotton seed has potentially equal toxicity to the parent compound, bromoxynil (U.S. EPA, 1997d). The compound has a relatively low oral acute LD50 (lethal dose for 50% of the population) of 779 mg/kg. Dermal acute toxicity of bromoxynil is greater than 2000 mg/kg (Hazardous Substances Data Base, 1992). Bromoxynil is a developmental toxicant which induces structural malformations in mammals when the chemical is administered orally or dermally (U.S. EPA, 1996). The most sensitive indicator of developmental toxicity is the induction of supernumerary ribs in low doses, 5mg/kg/day (US EPA, 1996). Other forms of developmental toxicity observed at higher dose levels include; increased overall incidences of minor anomalies; defects in the spine, sternabrae, and skull; and, reduced fetal weights (U.S. EPA, 1996). Dietary exposure in rats caused an increase in developmental disorders in fetuses. Also, at sub-lethal doses, female rats showed severe clinical signs of toxicity including panting, salivation, and vomiting (U.S. EPA, 1997b). The compound produced birth defects in rats at low oral doses. In studies conducted using rabbits, birth defects included changes in bone formation in the skull and hydrocephaly (Hazardous Substance Data Base, 1992).

Bromoxynil phenol is classified as a Group C- possible human carcinogen. Studies have shown that bromoxynil phenol resulted in an increased incidence of hepatocellular tumors ( both benign and malignant) in both female and male mice (US EPA, 1997). Two studies have supported evidence of this increase in tumor activity. A 1982 study observed increases in liver tumors at doses of 30 parts per million (ppm) and 100 ppm dose groups (U.S. EPA, 1996). In 1994 Rhone-Poulenc sponsored a carcinogenicity study that detected increases in adenomas (benign tumors of glandular origin) at dietary levels of 75 ppm and 300 ppm, and carcinomas (malignant tumors of cellular membrane origin) at 20 ppm and 300 ppm.

Regulatory Information:
On May 5, 1995, the Environmental Protection Agency (EPA) granted Rhone-Poulenc conditional registration for the use of bromoxynil for BXN®Cotton. This conditional registration expired on April 1, 1997. Along with this conditional registration, the EPA also established a temporary tolerance- a maximum permissible limit for the residues of bromoxynil in or on cottonseed and cotton gin trash (animal feed). The temporary tolerance was set at .04 parts per million.

Beginning in 1995, the Environmental Protection Agency regulated the safety of human exposure to bromoxynil by establishing tolerance levels for residues in raw commodities, like cotton oil, animal feed and meat from livestock fed cotton by-products. (A tolerance is the legal limit for a pesticide chemical residue in or on a food) A permanent tolerance was not established for bromoxynil on cotton due to data gaps in toxicity testing uncovered by the EPA while reviewing studies submitted by Rhone-Poulenc. The EPA believed that the most significant data gap involved the absence of a method and residue data for the metabolite 3,5-dibromo-4-hydroxybenzoic acid (DBHA) in cottonseed and cotton gin trash (U.S. EPA, 1995). This is particularly important as this metabolite accounts for most of the toxic residue found in the BXN®Cotton July 25, 1996, Rhone-Poulenc requested an extension of the time limited tolerance for bromoxynil on transgenic cotton. Rhone-Poulenc requested that the tolerance be kept at .04 ppm even though it lacked conclusive data on the metabolite DBHA. Rhone-Poulenc acted pre-emptively to ensure the acceptability of its transgenic seed. If the tolerance was kept low (at 0.04ppm), they correctly reasoned that bromoxynil could not be used and farmers would not buy the transgenic variety. Were cotton growers to purchase the conventional varieties that are less expensive, Rhone-Poulenc would lose an entire growing season and perhaps the "brand loyalty" they needed to ensure future reliance by cotton growers on their new transgenic variety. (Most farmers begin buying their seed for the next season during the winter prior to planting.)

After reviewing the newly submitted information regarding the metabolite DBHA, the EPA issued a Notice of Proposed Rulemaking on May 2, 1997 for establishment of tolerances for residues of bromoxynil and the metabolite DBHA on cotton commodities. The proposed tolerances are limited to one year, expiring January 1, 1998 and use will be limited to a maximum of 500,000 acres or 3% of the cotton acreage planted. The EPA's proposed tolerances for; undelinted cottonseed were set at 7ppm; for cotton gin byproducts at 50ppm; and for cotton hulls at 21 ppm (U.S. EPA, 1997e). Remakably, the tolerances proposed by the EPA are greater than those proposed by Rhone-Poulenc. EPA's decision appears to be motivated as much by practical considerations as by toxicity concerns. The increase accounts for the needed tolerance for existing residues of DBHA and the parent compound in cottonseed and cotton gin byproducts. The raised tolerances take dietary risk into account. According to the EPA, the carcinogenic risk from bromoxynil is negligible within the meaning of standard risk assessment. In their view, the lifetime upper-bound carcinogenic risk is well within the 1 in a million risk standard. If true, this standard also technically meets the requirements of the new Food Quality Protection Act (FQPA). This act amended Section 408 of the Federal Food, Drug, and Cosmetic Act (FFDCA) which authorizes the establishment of tolerances, exemptions from the requirement of a tolerance, modifications in tolerances, and revocations of tolerances for residues of pesticide chemicals in or on raw agricultural commodities and processed foods. The FQPA created a new subsection to section 408 which adds that the EPA must determine if a tolerance is 'safe'. 'Safe' in this instance is defined as having 'reasonable certainty that no harm will result from aggregate exposure to the pesticide chemical residue, including all dietary exposures and all other exposures for which there is reliable information (104th Congress of the U.S., 1996).

Consequences
We take issue with this conclusion for bromoxynil on three counts: first, the true food source risk from cotton is 1.5 in a million, above the regulatory threshold. Second, it fails to factor in possible bioaccumulation if cattle are left "on feed" for protracted periods ; and third, it neglects interactive effects that might enhance bromoxynil's reproductive toxicity. Such effects could occur, for instance, if a person were concurrently exposed to other teratogens during pregnancy. At issue is the fact that farmers are losing their sovereignty while the public is ingesting more harmful residues without providing reasonable disclosure or warnings. Given the financial incentives, Calgene, Rhone-Poulenc, Stoneville Pedigreed (the seed distributor), and Monsanto, owners of the former and latter, are busy lobbying for the distribution of their products. Meanwhile, farmers across the cottonbelt sit on their bags of transgenic seed waiting for the Environmental Protection Agency to decide the fate of their newly purchased BXN®Cotton.

Consider how we reached this untenable end-game: Rhone-Poulenc pressed its seed into production before it finished its toxicity testing. Consumers of cottonseed oil or bromoxynil contaminated meat from livestock fed cotton "trash" are at high risk from getting unmonitored residues. This is clearly an instance of putting efficiency at making cotton above health and safety. Does it portend similar genetic "quick fixes" in the future?