Phenoxy acid herbicides cause peroxisome proliferation in

Phenoxy acid herbicides cause peroxisome proliferation in Chinese hamsters. 70-73. An increase in either the size or amount of peroxisomes was obtained in the liver cells of Chinese hamsters after the animals were exposed to the phenoxy herbicides 2,4-di chlorophenoxyacetic acid (2,4-D) or 4-chloro-2-methylphenoxyacetic acid I(MCPA). At the dose level studied, 2,4-D was found to be more potent than MCPA in increasing the number of peroxisomes. A phenoxy acid derivative, clofibrate, one of the peroxi some proliferators known to possess carcinogenic properties in rodents, appeared to be still more potent in inducing peroxisome proliferation than either of the herbicides studied. Further investigations are warranted to clarify the significance of peroxisome proliferation to the toxicity of phenoxy herbicides.

Increased risk of soft tissue sarcomas has been reported among workers exposed to phenoxy acid herbicides (3,4,5,7). The mutagenic activity of phenoxy herbicides has been, however, low or nonexisting in the studies performed (11). Recently, Reddy & Azarnoff {IO) suggested a novel class of chemical carcinogens not possessing mutagenic activity in bacterial mutagenesis assays but acting via the excessive production of hydrogen peroxide (H202) and causing proliferation of peroxisomes. All the hypolipidemic peroxisome proliferators, including the widely used drug clofibrate, have also produced liver tumors in rats {6, 10). Furthermore, an overall excess of deaths from a variety of cancers, mainly gastrointestinal, has been reported for men receiving clofibrate therapy (6). Since clofibrate is closely related to phenoxy acids chemically, it was of interest to determine whether the commonly used phenoxy herbicides 2,4-dichlorophenoxyacetic acid (2,4-D) and 4-chloro-2-Reprint requests to: Dr H Vainio, Institute of Occupational Health, Haartmaninkatu 1, SF-00290 Helsinki 29, Finland. 0355-3140/82/010070-04 methylphenoxyacetic acid (MCPA) could also act as hepatic peroxisome proliferators.

Materials and methods
The test animals used in the study were 12-to 20-week-old Chinese hamsters weighing 34-40 g. They were bred in our laboratory from an outbred colony which originates from the Shell Research Centre (Sittingbourne, Kent, UK).
The phenoxy acid compounds considered were 2,4-D and MCPA, kindly provided by the manufacturer, Kemira Oy (Finland). Both treatment solutions consisted of commercial herbicide products wearing the trade names "Vesakontuho tasku" (solution containing 2,4-D) and "Vesakontuho MCPA" (solution containing MCPA). The amounts of 2,4-D and MCPA in the herbicide solutions are 550 g/kg and 500 g/kg, respectively, and both chemicals contain only one effective compound.
A single dose of 100 mg/kg of the phenoxy acid compound dissolved in physiological saline was given daily for 9 d by gavage to three animals in each exposure group (the dose volumes ranged from 0.11 to 0.14 ml, depending on the weight of the animal). The three control animals were treated similarly with saline only. The animals were killed on the 10th day of the experiment, 24 h after the last dose.
In the positive control experiment, the same single dose of 100 mglkg of clofibrate {Klofiran®, Remeda, Finland) dissolved in olive oil was injected subcutaneously into three animals daily for 7 d (the dose volumes ranged from 0.27 to 0.35 ml). The corresponding control animals were injected with olive oil only. The animals were killed on the eighth day of the experiment, 24 h after the last dose.
Two biopsies from each animal (three controls and three animals from each exposed group) were taken from the medial lobe immediately after the killing and fixed by immersion in 2 0/0 phosphatebuffered (pH 7.4) glutaraldehyde for 2.5 h.
Minced liver samples were washed for 0.5 h in pho3phate buffer (pH 7.4). The specimens were then incubated for 90 min at 38°C in a catalase localization medium [alkaline diaminobenzidine medium (8») and washed in phosphate buffer (pH 7.4) overnight. The specimens were postfixed in 1 Ofo phosphate-buffered (pH 7.4) osmium tetroxide, dehydrated, and embedded in Epon. Ultrathin sections were stained with lead citrate and examined with a JEM-I00 CX electron microscope.
The areas of peroxisomes were determined from 12 animals {3 per group) in electron micrographs {6,600 X) projected on a drawing board (final enlargement 100,000 X). The outlines of the peroxisomes were drawn and used for later morphometric evaluation. The areas of 330 peroxisomes (between 80 and 90 in each group) were analyzed. For the calculation of the frequency of peroxisomes per unit area of cytoplasm, the peroxisomes of 40 hepatocytes, taken by random sampling, were counted {10 per group, total number of peroxisomes 5,754) in electron micrographs with an enlargement of 3,000 X.
The areas of peroxisomes and hepatocytes were examined with the aid of a semiautomatic measuring device (Hipad digitizer and ABC-80 data processing unit). The statistical difference from the control group was calculated with Cochran's approximate t-test (2).

Results
The mean size of the peroxisomes of the control rats was 1.38 flm 2 (table 1). The 2,4-D treatment increased the mean area slightly but not significantly to 1.46 flm 2 • A significant increase in size was seen after MCPA and clofibrate treatment, to 2.02 and 2.37 flffi2, respectively.
The mean frequency of peroxisomes per 100 flm 2 of cytoplasm was 17.7 in the controls (table 2). MCPA treatment increased the frequency slightly but not ... P < 0.001, NS = not significant. ... p < 0.001, NS = not significant. A slight increase in the smooth, and decrease in the rough, endoplasmic reticulum was seen after clofibrate treatment, the reticulum being normal in the other groups. Mitochondria were normal in all cases.

Discussion
Little is known about the possible mechanism by which exposure to phenoxy acid herbicide::; may be associated with increased risk of soft-tissue sarcoma and malignant lymphoma, as suggested on the basis of epidemiologic studies done in Sweden (3,4,5). Although in many cases mutagenic and carcinogenic effects are experimentally correlated (1), the phenoxy acid herbicides do not appear to have direct interaction with DNA (deoxyribonucleic acid). Phenoxy acids are not known to bind to DNA, neither have they pos-::;essed mutagenic activity in the Salmonella/microsome assay (11).
The present data show that the phenoxy herbicides 2,4-D and MCPA are able to increase either the size or amount of peroxisomes in the liver cells of Chinese hamsters. Thus the possibility exists that phenoxy acid herbicide::; could act indirectly by a mechanism similar to the one suggested recently for tumor initiation by known peroxisome proliferators. This novel class of chemical carcinogens includes, eg, clofibrate, a phenoxy acid derivative, and other related hypolipidemic drugs, as well as phthalate ester di(2ethylhexyl)phthalate (8,10). Their mechanism of action has been suggested to occur via excessive production and/or breakdown of H 2 0 2 • Thus the initiation of neoplastic transformation would be the result of continued production of DNA-damaging oxygen radicals as a result of the persistent proliferation of peroxisomes.
In summary, the data obtained suggest Received for publication: 6 October 1981 that the phenoxy herbicides 2,4-D and MCPA act as peroxisome proliferators. Taking the existing evidence into account, further 'studies should be carried out to clarify the potential significance of this peroxisome proliferation in the toxicity of phenoxy herbicides.