![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081107123533im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
2004 Progress Report: Chlorotriazine Protein Binding: Biomarkers of Exposure & Susceptibility
EPA Grant Number: R828610Title: Chlorotriazine Protein Binding: Biomarkers of Exposure & Susceptibility
Investigators: Andersen, Melvin E. , Tessari, John D.
Institution: Colorado State University
EPA Project Officer: Deener, Kacee
Project Period: June 1, 2000 through May 31, 2003 (Extended to May 31, 2006)
Project Period Covered by this Report: June 1, 2004 through May 31, 2005
Project Amount: $710,617
RFA: Biomarkers for the Assessment of Exposure and Toxicity in Children (2000)
Research Category: Children's Health , Health Effects
Description:
Objective:The main objective of this research project is to test the hypothesis that binding of chlorotriazines by hemoglobin and hair proteins can be used to evaluate differences in exposure and in individual sensitivity toward chlorotriazines. The specific objectives of this project are to: (1) refine further gas chromatography/mass spectrometry (GC/MS) methods to assess the reactivity of chlorotriazines and metabolites with thiol-containing amino acid residues in hemoglobin; (2) determine if hair binding of sulfhydryl reactive triazines can be used as noninvasive measures of exposure to these triazines; (3) develop physiologically based pharmacokinetic (PBPK) models for juvenile and adult ages utilizing blood protein and hair protein binding (binding will be used to assess tissue exposure to total chlorotriazines in relation to ambient exposure); and (4) use these PBPK models with protein binding measurements to recreate exposure characteristics in laboratory animals and in a limited set of human blood and hair samples.
Progress Summary:Dr. Andersen continues to support this research while he is at the Chemical Industry Institute of Technology-Centers for Health Research (CIIT-CHR) in Research Triangle Park, North Carolina. Dr. Andersen continues to serve as co-advisor for Ms. Tami McMullin, a Ph.D. student who has worked on developing the PBPK models to support our biomarker research efforts with atrazine. Ms. McMullin will defend her Ph.D. dissertation work on Wednesday, June 22, 2005. Dr. Andersen will serve as an outside advisor to Greg Dooley, a Ph.D. student working on determination of hemoglobin adduct formation.
Dr. Andersen will continue his support of the research effort as an adjunct professor and continue to direct the efforts to develop PBPK models to atrazine and related chlorotriazines.
Year 5 of our research has centered on the following main focal areas:
Analytical Method Development
We have developed an analytical method for the determination of atrazine, desethylatrazine 2-chloro-4-amino-6-isopropylamino-s-triazine (De-ethyl atrazine), deisopropylatrazine 2-chloro-4-ethylamino-6-amino-s-triazine (De-isopropyl atrazine), and diaminoochlorotriazine, 2-chloro-4,6-diamino-1,3,5-triazine (DACT) from rodent brain samples by using polymeric mixed mode cation exchange solid-phase extraction followed by chemical derivatization and GC/MS. Estimated method detection and quantitation limits are below 20 ng/g and 40 ng/g, for each compound, respectively. The method was validated at 200 ng/g fortification of each compound according to the needs of a pilot study for which the method was developed. Mean recoveries for all analytes were between 85 and 129 percent, respectively. The analytical method was applied to a pilot animal study to determine time-course concentrations of chlorotriazines in brain samples following a single oral gavage dose with atrazine to female Sprague Dawley (SD) rats.
Adduct Determination Project
Atrazine Adduct. The analysis of protein adducts may provide an effective method for biomonitoring environmental exposure to chemicals such as pesticides. Previous data from our laboratory suggests that the commonly applied herbicide atrazine may covalently bond with hemoglobin in SD rats. In vivo exposures of rats to 300 mg/kg atrazine resulted in significantly greater globin adduct formation than in vitro experiments with control rat whole blood, suggesting that metabolism of atrazine is important for adduct formation. Mass spectrometry data indicates an approximately 110 addition to the beta globin chain, which may correspond to a reaction with a dechlorinated and dealkylated atrazine. Using globin isolated from SD rats exposed to atrazine, we investigated the nature of the covalent binding of metabolized atrazine to the beta chain of globin.
High performance liquid chromatography (HPLC) of rat globin was used as a means to separate the alpha and beta chains. The two alpha subunits were not resolved from each other, but were resolved from the beta subunits. HPLC analysis of the total globin from control rats gives two beta peaks (major and minor), whereas total globin from rats 72 hours following exposure to 300 mg/kg atrazine gives three beta peaks. The extra beta peak from the exposed rats eluted prior to the major and minor beta subunits and was suspected to be a modified beta subunit. Because previous data suggest the adduct is on one of the beta subunits, we collected fractions of the beta subunits for analysis with matrix-assisted laser desorption ionization-time of flight-mass spectrometry (MALDI-TOF-MS) to confirm the mass and purity of the fraction.
The MALDI-TOF-MS analysis of the two peaks from the control rats showed the first beta peak with a mass of 15,914 and the second peak with a mass of 15,869. The same analysis of the three peaks from the exposed rats identified the second and third eluting beta peaks that had similar mass as the controls, but the first eluting peak had a mass of 15,993. The mass differences on the first and third peaks was 124 plus or minus 32, suggesting that the first peak is the same subunit as the third, with an adduct similar to original data of a 110 mass addition.
We investigated the nature of the first eluting peak by collecting HPLC area data from all rats in the experiment at doses of 0 mg/kg, 10 mg/kg, 30 mg/kg, 100 mg/kg, and 300 mg/kg at 72 hour post-exposure and from rats exposed to 300 mg/kg at 0 hours, 24 hours, 48 hours, 72 hours, 10 days, 1 month, and 2 months post-exposure. By taking the area of each of the three beta peaks and standardizing by percentage of the total area of these peaks, we were able to identify two significant trends. The percent of the total of the first eluting peak increases with dose as the percent of the total of the third eluting peak decreases with dose. The second eluting peak does not change with dose. The time-course data suggest shows the first eluting peak increases up to a maximum at 10 days and then decreases to 0 hour levels after 2 months. The third eluting peak shows an inverse pattern decreasing from 0 hour to 10 days and returning to 0 hour levels after 2 months. This data also suggest that the first eluting peak is a modified version of the third eluting beta peak that disappears as red blood cells are turned over after two months.
Tryptic digestion of the first and third eluting beta peaks with MALDI-TOF-MS-MS and Mascot protein library search analyses were used to identify the beta peaks and compare the mass of each digest peptide to locate the amino acid with the adduct. The MALDI-TOF-MS and Mascot analysis of the two peaks from the control rats showed the first beta peak with a mass of 15,914 as the minor beta subunit and the second peak with a mass of 15,869 as the major beta subunit. The same analysis of the three peaks from the exposed rats identified the second and third eluting beta peaks as the minor and major beta subunits, respectively, with similar mass. The first eluting peak had a mass of 15,993 and was identified as 83 percent similar to the major beta subunit. When the two digest spectra were compared the third eluting peak had a peptide with a mass of 1,340 that was not found in the first eluting peak digest. This peptide corresponded to amino acids 121-132, which contains Cys 125 that has been demonstrated in the literature to form covalent adducts with other xenobiotics, such as stryrene, benzene oxide, and acylamide. Another significant difference in the two digest spectra was the first eluting peak digest contained a significant peak at 1449.9 that was not found in the third eluting peak digest. This peak at 1449.9 is thought to be the 1,340 peptide with a 110 mass addition from the adduct. Tandem mass spectrometry (MS-MS) of this 1449.9 peptide allowed us to sequence and identify the specific amino acid with the mass addition. The amino acid sequence of the 1449.9 peak was the same as the sequence determined with MS-MS for the 1,340 peptide except that Cys 125 of the 1449.9 peak was shown to have the 110 mass addition. The Cys 125 of the 1,340 peak was a normal cystiene residue. The 110 mass additions to the Cys 125 is hypothesized to be the product of a dechlorinated, dealkylated atrazine reaction with Cys 125, as this dechlorinated diaminotrazine has a mass of 110.
These results show that atrazine metabolism in SD rats yields a reactive intermediate capable of forming a covalent adduct with sulfhydryl functional groups found on cystiene residues. This adduct has a mass of 110 and is found on Cys 125 of the major beta subunit. The structure of this adduct is not known, but based on the mass, the only reasonably anticipated structure is a dechlorinated diaminotrazine. Although the exact metabolic pathway leading to adduct formation is not known, exposure of SD rat to atrazine will form an adduct that is reproducibly detected and could provide an analytical model for detection of atrazine adducts in other macromolecules with sulfhydral functional groups.
Gel Electrophoresis Experiments. The objective of these experiments was to separate and isolate the alpha and beta subunits of SD rat globin and human globin.
Globin was extracted from blood collected from rats and humans by the same method. Red blood cells were separated from plasma following centrifugation of blood and lysed with cold distilled water. Heme was removed from the globin with acidification, and the globin was precipitated with ethyl acetate and the pellet dried. Purified globin was then dissolved in MO water at a concentration of 25 mM for separation with polyacrylamide gel electrophoresis (PAGE).
The separation of the subunits was performed using precast 12 percent Tris-HCl polyacrylamide gels (BioRad). Globin solutions were mixed with equal volumes of a bromophenol blue/glycerol/sodium dodecyl sulphate treatment buffer and heated in boiling water for 90 seconds. This solution was loaded onto the gel and run at 120 volts for 70 minutes. The gels were then stained with a 20 percent MeOH, 0.2 percent Coomassie Blue, and 0.5 percent acetic acid solution for 20 minutes. The gels were then destained with a 30 percent MeOH solution for at least 2 hours. The result of the gel for human globin is displayed in Figure 1.
Figure 1. Human Globin Run on a 12% Tris-HCl Polyacrylamide Gel at 120 V for 70 Minutes
The molecular weights of the alpha and beta subunits of human globin are 15,851 and 15,110 respectively. Following distaining, the bands corresponding to the alpha and beta subunits were excised with a razor blade. The same PAGE procedure was done with rat globin, but the results were quite different. For some reason, the subunit bands were not separated as with the human globin (Figure 2).
Figure 2. SD Rat Globin Run on a 12% Tris-HCl Polyacrylamide Gel at 120 V for 70 Minutes
Two procedures were used to extract the subunits from the gel bands of human globin: sonication with acetonitrile and electroelution. The sonication procedure entailed mincing the band in an Eppendorf tube with a 50 percent acetonitrile/0.1 percent trifluoroacetic acid solution and sonicating for 20 minutes. This was repeated three times and the solutions pooled. The solution was evaporated to 20 μl with a Speed Vac for analysis with MALDI-TOF-MS. The electroelution procedure utilized an electroelutor tube (BioWorld) to extract the subunits from the gel. The gel bands were inserted in the tube that was filled with a Tris-base buffer. The tube was placed in a horizontal electrophoresis tank and run at 120 V for 30 minutes. The theory of this procedure is that the protein will be pulled from the gel and stick to a dialysis membrane on the side of the tube. Reversal of the charge will pull the protein off the membrane and into the inner solution of the tube. Following the electroelution, the inner solution was incubated at room temperature with a proprietary MS buffer (BioWorld). The solution was then mixed with a 50 percent trichloroacetic acid solution, incubated at 4°C, and centrifuged at 14,000 RPM. The supernatant was decanted, the pellet redissolved in acetone, incubated overnight at -20°C, and centrifuged. The pellet was then air dried for MALDI-TOF-MS analysis.
MALDI-TOF-MS analysis of the extracted subunits showed that both methods were unsuccessful in extracting isolated subunits of human globin. Very low concentrations of either human or rat globin subunit were extracted from the gels and interferences from the buffer salts and Coomassie blue stain may have lead to the poor results in the MS data. The beta extracts did show low levels of beta subunit, but the alpha subunit also was present. This indicated that visual separation was achieved with this method for human globin, but extracting isolated subunits from the gel was not successful. The molecular weights of the subunits may be too close to separate with PAGE. The use of a razor blade to separate the two bands in the gel also may not have been as precise as needed. Based on the poor results in the MS data, other alternative methods for separation if the subunits was investigated with HPLC.
Human Blood Experiment
Whole human blood was incubated with 90 ppm DACT in dimethyl sulfoxide for 48 hours at 37°C. At 48 hours, the blood was centrifuged and plasma pipetted off the red blood cells. Globin was precipitated from the red blood cells and analyzed with HPLC. Based on previous experiments with rat globin, we looked for an additional beta subunit peak that would correspond to a modified subunit. We did not see any additional peak in the chromatograph, indicating that no covalent modification is occurring between DACT and human hemoglobin as seen with rat hemoglobin.
Pharmacokinetic Models/Neuroendocrine Projects
A series of pharmacokinetic models were developed to describe in vitro and in vivo kinetic data on atrazine and its metabolites. The time-course concentrations of atrazine and the chlorinated metabolites in plasma and brain were regulated by dose-dependent and sequential absorption of compound from gut, oxidative metabolism in the liver and intestine, reactivity with hemoglobin in red blood cells and with plasma proteins, systemic clearance by glutathione S-transferase mediated glutathione conjugation, and urinary elimination. These processes resulted in minimal concentrations of atrazine and retention of the mono-dealkylated metabolites and DACT in plasma and brain. DACT was the major chlorotriazine present in tissue, representing more than 95 percent of total chlorotriazine area under the concentration curve after dosing with atrazine.
In evaluating the neuroendocrine mode of action of atrazine and its metabolites, we determined that atrazine and DACT suppress the luteinizing hormone (LH) surge by mechanisms other than altering binding of estrogen to its cognate receptors in the hypothalamus. Moreover, pituitary responsiveness was altered in animals treated with concentrations of DACT that suppressed the estradiol/progesterone-induced LH surge. The high degree of reactivity of DACT with sulfhydryl residues in rat hemoglobin indicate that tissue reactivity of DACT should be considered as a possible mode of action instead of a direct interaction, either inhibition or activation, with cellular receptor molecules. This research has improved our understanding of the mechanisms by which chlorotriazines alter the LH surge and the factors that control chlorotriazine tissue dose under conditions where neuroendocrine responses are observed. As such, these studies should assist low-dose, and to some extent, interspecies extrapolations of the dose-response behavior of chlorotriazines. This research also identified an additional mode of action and kinetic studies that will be required to create a complete model that describes the kinetic and biological effects to support low dose and interspecies extrapolation of risks in humans posed by atrazine.
Upon reviewing the time-course results, it was evident that the concentration of substrates and products changed over time and that there were characteristics of inhibition occurring in these incubations. Therefore, basic Michaelis-Menten enzyme kinetics would not adequately produce Vmax and Km for this study. A more complex kinetic analysis, building on the Michaelis-Menten equation, was necessary to generate kinetic values. A kinetic analysis was completed to address changes in parent compound and chlorinated metabolites over time and multi-substrate inhibition. A kinetic analysis using Berkeley-Madonna allowed us to quantitatively evaluate the changes in substrate and product concentration over time while accounting for competitive inhibition. To accomplish this, three models were generated and used to determine kinetic constants.
Future studies include experiments to obtain kinetic parameters from desisopropyl-atrazine and desethyl-atrazine metabolism to DACT.
Future Activities:The results of these studies provide a strong basis for the objectives put forth by the research plan. Atrazine exposure can and does lead to the formation of hemoglobin adducts in rats, and these adducts meet several requirements for a biomarker of exposure. First, they are sensitive and reflect internal dose, as shown by the significant association between dose level and percent of modified globin. Second, they provide a measure of toxic effect, because the formation of this adduct may be positively correlated with dose-dependent LH surge suppression in SD rats dosed with atrazine at the same levels. Finally, the adducts appear to be chemically stable, and to some extent, they reflect the individual susceptibility of the SD rat strain to the endocrine disrupting toxic effects of atrazine. This research is showing that atrazine-induced hemoglobin conjugates behave differently between humans and rats, and between in vivo versus in vitro treatment in rats. These discoveries raise questions as to what mechanisms are involved for adducts to form, as well as exactly what the reactive metabolite is. It will be important to determine the exact mechanism of adduction, as this study indicates that conjugate formation relies on metabolic activation. One of the long-term goals of this project is to use hemoglobin biomarkers as a measure of tissue exposure to active compounds. They could also be used to determine individual susceptibility, because those who have the capacity to produce the reactive intermediate may be more vulnerable to atrazine toxicity. Understanding the reaction of proteins with atrazine would bring us closer to that goal, as well as to understanding the exact mode of action by which atrazine produces its toxic effects.
Our work will be instrumental in improving our understanding of risks of these herbicides to children. Its value has to be measured in relation to two phases: (1) development of accurate tools to assess both exposure and potential susceptibility to triazine herbicides in children; and (2) use of these tools with specific populations of children who may be at higher risks. Currently, methods for assessing exposure in children are based on a series of assumptions regarding uptake and metabolic rate differences in children without methods to accurately assess the validity of these assumptions. With the triazines, metabolite identification in urine and/or salivary or urinary analysis of atrazine lacks the sensitivity for use as anything other than a monitor for acute, high exposures. By completing the analytical methods, we will have broadly integrated biomarkers of exposure and susceptibility that can be applied to different juvenile populations. The PBPK model will permit calculation of expected triazine binding in various populations. Study design criteria for biomonitoring in children and workers can be established, partially at least, on the basis of these calculations.
Journal Articles:No journal articles submitted with this report: View all 29 publications for this project
Supplemental Keywords:exposure, risk assessment, health effects, susceptibility,
chemicals, atrazine, DACT, chlorotriazine, hemoglobin, hair proteins, adducts,
pharmacokinetic biomarkers, PBPK models,
,
Toxics, Scientific Discipline, Health, RFA, Susceptibility/Sensitive Population/Genetic Susceptibility, Biology, genetic susceptability, Health Risk Assessment, Children's Health, Biochemistry, pesticides, Environmental Chemistry, health effects, chlorotriazine, endocrine disruptors, harmful environmental agents, susceptibility, agricultural community, metabolites, pesticide exposure, triazine herbicides, biological markers, pharmacokinetc model, growth & development, tissue reactivity, chlorotriazine protein binding, Human Health Risk Assessment, protein binding, atrazine
Relevant Websites:
Progress and Final Reports:
2000 Progress Report
2001 Progress Report
2002 Progress Report
2003 Progress Report
Original Abstract
Final Report