Dose-Dependent Nonlinear Pharmacokinetics of Ethylene Glycol Metabolites in Pregnant (GD 10) and Nonpregnant Sprague-Dawley Rats following Oral Administration of Ethylene Glycol

L. H. Pottenger*,1, E. W. Carney{dagger} and M. J. Bartels{dagger}

* Toxicology and Environmental Research and Consulting, Dow Europe, SA, Horgen, Switzerland, CH-8810; and {dagger} Toxicology and Environmental Research and Consulting, The Dow Chemical Company, Midland, Michigan 48674

Received January 4, 2001; accepted March 13, 2001


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The kinetics of orally administered ethylene glycol (EG) and its major metabolites, glycolic acid (GA) and oxalic acid (OX), in pregnant (P; gestation day 10 at dosing, GD 10) rats were compared across doses, and between pregnant and nonpregnant (NP) rats. Groups of 4 jugular vein-cannulated female rats were administered 10 (P and NP), 150 (P), 500 (P), 1000 (P), or 2500 (P and NP) mg 13C-labelled EG/kg body weight. Serial blood samples and urine were collected over 24-hr postdosing, and analyzed for EG, GA, and OX using GC/MS techniques. Pharmacokinetic parameters including Cmax, Tmax, AUC, and ßt1/2 were determined for EG and GA. Pregnancy status (GD 10–11) had no impact on the pharmacokinetic parameters investigated. Blood levels of GA were roughly dose-proportional from 10 to 150 mg EG/kg, but increased disproportionately from 500 to 1000 mg EG/kg. EG and GA exhibited dose-dependent urinary elimination at doses >= 500 mg EG/kg, probably due to saturation of metabolic conversion of EG to GA, and of GA to downstream metabolites. The shift to nonlinear kinetics encompassed the NOEL (500 mg EG/kg) and LOEL (1000 mg EG/kg) for developmental toxicity of EG in rats, providing additional evidence for the role of GA in EG developmental toxicity. The peak maternal blood concentration of GA associated with the LOEL for developmental toxicity in the rat was quite high (363 µg/g or 4.8 mM blood). OX was a very minor metabolite in both blood and urine at all dose levels, suggesting that OX is not important for EG developmental toxicity.

Key Words: ethylene glycol; glycolic acid; metabolism; nonlinear pharmacokinetics; pregnant rats; developmental toxicity mechanisms..


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ethylene glycol (EG) is an important industrial chemical, for which the largest industrial use is the production of polyethylene terephthalate for polyester fibers, containers, and films. Due to its solubility and thermal characteristics, it is also used in applications such as heat-transfer fluids, automotive coolants, aircraft and runway deicers, and as a solvent in the manufacture of dyes and inks.

An extensive animal database on EG developmental toxicity has been generated that includes studies using a variety of exposure routes, including oral, dermal, and inhalation (reviewed in Carney, 1994), as well as several mechanistic studies (Carney et al., 1996Go, 1999Go; Khera, 1991Go). The developmental toxicity of EG is highly route-dependent, with high-dose, oral-gavage exposures causing developmental effects in rodents, but routes more relevant to typical human exposures (dermal, drinking water, inhalation) are much less toxic. Furthermore, clear species differences in susceptibility to EG-induced developmental toxicity have been demonstrated. Rodents (rats and mice) demonstrated developmental effects following high-dose bolus administration of EG (Neeper-Bradley et al., 1995Go; Price et al., 1985Go), while no developmental effects were identified in rabbits, even following bolus oral doses that were fatal to more than 40% of the dams (Tyl et al., 1993Go).

Available data indicate that the developmental effects resulting from high-dose exposure of rodents to EG are dependent upon the metabolism of EG (Carney, 1994Go; Carney et al., 1996Go; 1999). In general, metabolism of EG proceeds via the alcohol dehydrogenase/aldehyde dehydrogenase (ADH/ALDH) enzymatic complex, by forming glycolic acid (GA), which can be further biotransformed to oxalic acid and CO2 (Frantz et al., 1996aGo, bGo, cGo; Jacobsen and McMartin, 1986Go; Marshall, 1982Go). Several of these studies demonstrated a shift in elimination of GA, the major urinary metabolite of EG, and accumulation of GA in plasma with increasing doses of EG (Frantz et al., 1996aGo, bGo, cGo; Marshall, 1982Go). GA was further investigated as the potential proximate toxicant for EG developmental toxicity. Recent work has provided clear evidence that administration of GA causes the same spectrum of developmental effects identified following high dose EG, both in vitro (Carney et al., 1996Go) and in vivo (Carney et al., 1999Go; Munley et al., 1999Go). Importantly, GA can cause some developmental effects in the absence of metabolic acidosis, both in vitro (Carney et al., 1996Go) and in vivo (Carney et al., 1999Go), which had been proposed earlier as the primary causative agent (Khera, 1991Go). These recent data indicate that GA is the primary causative agent for the high-dose developmental effects of EG in rodents, with the accompanying metabolic acidosis playing a secondary, exacerbating role at higher dose levels.

The pharmacokinetics of EG metabolism have been studied in different animal models by several investigators. Most recently, Frantz et al. (1996a,b,c) reported on the dose- and route-dependency of EG metabolism in male and nonpregnant female rats and mice. In particular, these authors reported dose-proportional pharmacokinetics for the elimination from blood of orally (gavage) administered EG in female rats and mice over the dose range investigated (10–1000 mg EG/kg bw), with a shift to increasing urinary elimination of EG and its major metabolite GA, with increasing dose. However, their data on blood levels of GA indicated nonlinearity across doses, with a 1000- to 10,000-fold increase in GA blood levels for a 100-fold increase in administered EG dose. This disproportionality for GA was also reflected by an increased urinary elimination of GA with increasing administered EG doses. Thus, administration of high doses of EG via gavage had the potential to result in accumulation of GA, the proximate developmental toxicant. In contrast, dermal exposures of up to 1000 mg/kg showed no such shift in metabolism, with the majority of the EG dose being eliminated as CO2.

No data have been published on the potential effect of pregnancy on EG metabolism. The physiology of pregnancy results in major changes in physiological parameters that could have considerable impact on pharmacokinetics and metabolism of EG. Examples include overall weight gain and increase in organ size, particularly liver; increase in total body water and blood volume; increased blood flow to uterus; decreased overall plasma protein level but increases in specific plasma proteins; altered hormone blood levels; decreased GI motility and increased GI transit time (Miller, 1983Go). These parameters do not offer a comprehensive list of pregnancy-related physiological changes, but do represent a few that could easily affect the pharmacokinetics and metabolism of EG. Any pregnancy-related changes in EG metabolism and pharmacokinetics would be expected to have impact on the formation and elimination of the proximate developmental toxicant, GA, and could be critical to understanding the species-specific developmental toxicity of EG. This study investigated the effect of pregnancy (GD 10) on the pharmacokinetics and metabolism of EG in female Sprague-Dawley rats, comparing blood concentration-time profiles and urinary excretion of parent and metabolites across doses between pregnant and nonpregnant rats. GD 10 was chosen based on prior in vivo (Khera, 1991Go) and whole-embryo culture studies (Carney et al., 1996Go) showing this period to be highly sensitive to EG toxicity. This study was conducted in compliance with the requirements of Good Laboratory Practices (EPA-TSCA, 1989Go; OECD, 1982Go).


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Chemicals.
Ethylene glycol (EG; 13C2-labelled) was obtained from Isotec Inc. (Miamisburg, OH; purity 96.7% via 1H-NMR). Pentafluorobenzoyl chloride and N-(tert-butyldimethylsilyl)-N-methyltrifluoroacetamide (MTBSTFA) were obtained from Aldrich Chemical Company (Milwaukee, WI). All other compounds and solvents were reagent grade or better.

Test animals.
Adult female (nonpregnant and time-mated) Sprague-Dawley rats were purchased from Hilltop Lab Animals, Scottsdale, Pennsylvania. Rats were shipped on GD 8 and arrived at our laboratory on GD 9. All rats had an in-dwelling jugular vein cannula implanted at the supplier's facility on GD 6 (or the same calendar day for the nonpregnant rats). The rats were allowed to recover for 2 days prior to being shipped. The cannula was exteriorized under light methoxyflurane anesthesia upon arrival at the laboratory. Similar surgical procedures in pregnant animals have been used in the past in this laboratory, and no untoward effects on pregnancy were observed (Carney et al., 1999Go). Upon receipt the animals were examined by a veterinarian and found to be in good health. The animals were acclimated to the laboratory environment for 1 day prior to dose administration. The rooms in which the animals were housed had a 12-h photocycle and are designed to maintain adequate environmental temperature, relative humidity, and airflow for the rat. Municipal drinking water and Purina Certified Rodent Chow #5002 (Purina Mills, Inc., St. Louis, MO) were provided ad libitum during the predosing period, except that on the day prior to dosing, a uniform amount of chow was fed (approximately 15 gram/rat). Also, food was completely withdrawn approximately 2 h prior to the administration of the test material and was returned about 4 h postdosing. These steps were intended to minimize between-animal variation in absorption of test material, yet limit any deleterious effects of feed restriction on the pregnant animal. Rats were selected from those with patent jugular vein cannulae on GD 10 and then were randomly assigned to treatment groups using a computerized procedure based on animal body weight. Rats were identified by a uniquely numbered metal eartag.

Dose administration.
13C-labelled EG, as an aqueous solution, was administered to 5 groups of 4 time-mated female Sprague-Dawley rats by gavage, with blunted feeding needles, at the following dose levels: 2500, 1000, 500, 150, and 10 mg EG/kg body weight. 13C-labelled EG was also administered, in the same fashion, to 2 groups of nonpregnant female rats (2500 mg EG/kg, n = 5; 10 mg EG/kg, n = 4). The target dose volume for all animal groups was 5 ml/kg. The experiment was divided into 2 replicates of approximately 14–15 rats each, due to the number of rats involved. Immediately following dosing, the animals were placed in glass Roth-type metabolism cages for blood sampling and for the separation and collection of urine.

Specimen Collection
Urine.
All urine voided during the study was collected in dry ice-cooled traps at 12-h intervals and the cage was rinsed with a minimal volume (< 10 ml) of deionized water. Analyses were done on the combined urine specimen and cage rinse.

Whole blood.
Approximately 0.2 ml blood/rat were collected at 0 (predose), 1, 3, 6, 9, 12, 18, and 24 hours after administration of test material. Following collection of each blood sample, approximately 0.2 ml of heparinized saline was slowly injected to flush the cannula and provide a "heparin lock."

Terminal sacrifice.
After collection of the last blood sample, the animals were euthanized via carbon dioxide inhalation. The uterus of each rat then was examined for implantation sites to determine pregnancy status. Blood and urine samples from any time-mated rats that were found to be nonpregnant during necropsy were analyzed and included in the comparative nonpregnant group, as appropriate.

Specimen analysis.
Samples of whole blood and urine were analyzed via gas chromatography-mass spectrometry (GC/MS) to quantitate parent 13C2-EG, and the metabolites 13C2-GA and 13C2-OX. Samples were derivatized as described below to increase analyte volatility and detector response. The use of a stable isotope-labeled form of EG test material (13C2-), resulted in the formation of 13C2-labeled metabolites (13C2-GA and 13C2-OX). Analysis of blood and urine samples via mass spectral detection allowed for the determination of 13C2-EG and the 13C2-labelled metabolites independent of the presence of background, and perhaps varying levels of unlabelled EG, GA, and OX. Unlabeled GA and OX served as internal standards in the quantitation of 13C2-labelled GA and OX. These unlabeled analogs were added to the blood and urine samples at concentrations of approximately 100 µg/ml or 1000 µg/ml, respectively, which was approximately 50x background levels of these compounds. Deuterium-labelled EG (D4-EG) was utilized as an internal standard in the quantitation of 13C2-EG. Quantitation limits for 13C2-EG, 13C2-GA, and 13C2-OX in blood were 0.1, 2.1, and 4.9 µg/g blood, respectively. Quantitation limits for 13C2-EG, 13C2-GA, and 13C2-OX in urine were 1.0, 0.94, and 2.0 µg/g urine, respectively.

Aliquots of each blood sample from the time-course experiment (approximately 0.1 g) were added to 0.9 ml of an internal standard solution (0.9 ml 1N HCl containing 11.5 µg/ml D4-EG) in a 4-ml glass vial. An additional 0.1 ml of a second internal standard solution (approximately 100 µg/ml GA and OX) was also added to each sample. Samples were treated with 0.5–1 g NaCl and extracted with methyl-t-butyl ether (MTBE) containing 0.5% tri-n-octylphosphine oxide (2 x 2 ml). The combined MTBE extracts (containing GA and OX) were evaporated to dryness (nitrogen stream), reconstituted in 0.9 ml toluene and derivatized with 100 µl N-methyl-N-(t-butyldimethylsilyl)trifluoroacetamide (60°C x 1 h). The derivatized sample was transferred to a 2-ml glass vial for analysis of 13C2-GA and 13C2-OX. A second aliquot of each blood sample (approx. 0.1 g) was prepared as above for analysis of parent 13C2-EG, without NaCl, to minimize chromatographic interferences. The aqueous phase remaining after MTBE extraction (containing EG) was treated with 300 µl 5N NaOH and derivatized with 20 µl pentafluorobenzoyl chloride and 1 ml toluene (vortex-mixed at 45°C x 30 min). The toluene layer was transferred to a 2-ml glass vial for analysis of 13C2-EG.

Aliquots (0.1 g) of urine samples were added to 0.8 ml 1N HCl, and fortified with 5 µl or 100 µl of internal standard solution (approx. 1000 µg/ml GA, OX and D4-EG) and extracted with MTBE containing 0.5% tri-n-octylphosphine oxide (2 x 2 ml). Further treatment of the combined MTBE extracts (containing GA and OX) was as described above for the blood samples. A second aliquot (0.1 g) of each urine sample was fortified with 100 µl of internal standard solution (approx. 1000 µg/ml GA, OX and D4-EG), treated with 50 µl 5N NaOH and derivatized with 50 µl pentafluorobenzoyl chloride and 2 ml toluene (vortex-mixed at 45°C x 30 min). The toluene layer was transferred to a 2-ml glass vial for analysis of 13C2-EG.

GC-MS analyses were performed on a Finnigan TSQ-700, SSQ-710, or Hewlett Packard 5989X mass spectrometer (Finnigan MAT, San Jose, CA; Hewlett Packard, Avondale, PA), equipped with a Hewlett Packard 5890 gas chromatograph and a 7673A autosampler. Separations were achieved using a J&W DB-5 fused silica capillary column (J&W Scientific, Folsom, CA) (GA/OX: 30 m x 0.25 mm id x 0.25 µm film; EG: 30 m x 0.32 mm id x 1 µm film); helium carrier gas (10 psig) at a flow rate of approximately 0.5 ml/min; gas chromatograph oven temperature program for GA/OX: 100°C (0.5 min initial hold) to 280°C at 15°/min, then to 300°C at 25°/min with injector and capillary transfer line at 200°C and 250°C, respectively; gas chromatograph oven temperature program for EG: 100°C (0.5 min initial hold) to 280°C at 20°/min with injector and capillary transfer line at 250°C; 1-µl autosampler injection (GA/OX: 40 ml split; EG: 0.1 min splitless). The mass spectrometer conditions for GA/OX were electron impact ionization (EI): ion source temperature, 150°C; ionizing current, 0.4 mA; electron energy, 70 eV. The mass spectrometer conditions for EG were negative-ion chemical ionization (NCI): ion source temperature, 150°C; ionizing current, 0.4 mA; electron energy, 70 eV. Quantitation of the t-butyldimethylsilyl derivatives of GA, 13C2-GA, OX and 13C2-OX was achieved by selected ion monitoring (m/z 247, 249, 261 and 263 @ 70 msec/ion/scan). Quantitation of the pentafluorobenzoyl ester derivatives of 13C2-EG and D4-EG was achieved by selected ion monitoring (m/z 452 and 454 @ 75 msec/ion/scan).

Statistics and data analysis.
Descriptive statistics were used (i.e., mean ± SD). Results were generally expressed as percentage of administered dose and/or as µg of parent EG or metabolite. Certain pharmacokinetic parameters were estimated for blood data, including Cmax and AUC for parent material and GA (Gibaldi and Perrier, 1982Go), time to reach maximum blood levels (Tmax), and half-life of elimination (ßt1/2). The reported pharmacokinetic (PK) values were estimated using a commercially available computer modeling program (PK Solutions v2.02, Summit Research Services, Ashland, OH).


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Actual concentrations of EG in the various dose solutions ranged from 95–102% of target. Administered dosage of EG ranged from 92–100% of target across all dose groups. At dosing, individual animal body weights ranged from 211 to 264 g across all dose groups. There were no remarkable changes in behavior or demeanor recorded during the daily animal observations. At terminal sacrifice, the pregnancy status of the time-mated rats was determined, and the number of implantations/animal ranged from 7–22 for the pregnant rats.

Blood Concentration-Time Course of EG and Metabolites
Individual blood samples were analyzed for parent EG, GA, and OX concentrations. The mean values for the concentration-time courses of EG, GA, and OX are presented in Table 1Go, and Figures 1 and 2GoGo depict the concentration-time course profiles for EG and GA, respectively. The pharmacokinetic parameters, estimated using the blood data, are presented in Table 2Go. No pharmacokinetic parameters, including area-under-the-curve (AUC) data, were calculated for the OX concentration-time curve, as the blood levels of 13C2-OX were either at or near the quantitation limit of 4.9 µg/g blood in all samples analyzed.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Concentration-Time Course for Mean Blood Levels of Ethylene Glycol (EG), Glycolic Acid (GA), and Oxalic Acid (OA) following Oral Administration to Female S-D Rats
 


View larger version (21K):
[in this window]
[in a new window]
 
FIG. 1. Concentration-time profile of unchanged parent ethylene glycol (EG) in the blood of pregnant or nonpregnant female S-D rats following oral administration of EG. Symbols represent the mean ± SD from 5 rats. LOQ = 0.1 µg EG/g blood.

 


View larger version (15K):
[in this window]
[in a new window]
 
FIG. 2. Concentration-time profile of glycolic acid (GA) in the blood of pregnant or nonpregnant female S-D rats following oral administration of EG. Symbols represent the mean ± SD from 5 rats. GA was not quantifiable at any time points for both of the 10 mg EG/kg dose groups. LOQ = 2.1 µg GA/g blood.

 

View this table:
[in this window]
[in a new window]
 
TABLE 2 Pharmacokinetic Parameters Estimated for EG and GA following Oral Administration of EG to Female Rats
 
Comparison of the data collected for pregnant (P) and nonpregnant (NP) groups at both the high (2500 mg EG/kg) and low (10 mg EG/kg) dose clearly demonstrated that pregnancy at this stage (GD 10–11) did not have any significant impact on the blood concentration-time profiles of EG, GA, or OX. As shown in Figures 1 and 2GoGo, the time-courses for the mean EG and GA blood levels of P and NP rats administered either 2500 or 10 mg EG/kg were superimposable. Thus, the pharmacokinetic parameters summarized in Table 2Go did not differ significantly between the P and NP dose groups.

Time course of parent EG.
Parent EG was not detectable in any pretreatment blood samples. The Tmax for EG for all dose groups occurred as expected at 1 h postdosing, the first sampling time. The blood concentration of EG decreased in a linear fashion following the 1 h Cmax. Blood EG was no longer detectable for the low dose groups (10 mg EG/kg) by 12 h postdosing, and for the 150 and 500 mg EG/kg dose groups by 24 h postdosing, although parent EG was detected at 24 h postdosing for the 2500 and 1000 mg EG/kg dose groups.

Comparison of Cmax values across dose levels demonstrated linearity in the dose-response for parent EG for all dose groups except the 2500 mg EG/kg. The linearity of the dose-response for parent EG was also supported by comparison of AUC values for EG, where the AUC increased across dose levels proportionately with dose for all the groups, except the high dose. The 1000 and 2500 mg EG/kg doses differed by 2.5-fold, while the respective AUC and Cmax values demonstrated about a 4-fold difference. This disproportionate increase may indicate that blood clearance of EG was initially saturated at the 2500 mg EG/kg dose level. Such initial saturation could be due to either saturation of metabolic clearance of EG from blood (i.e., saturation of formation of GA from EG) or to saturation of renal elimination of EG from blood. The estimated t1/2 of elimination of EG from blood was short, less than 2 h for all dose levels, indicating that, overall, parent EG was rapidly cleared from blood. Thus, any initial saturation in blood levels at the high dose did not last long enough to alter its half-life of elimination from blood.

Time course of GA in blood.
GA, recognized as an endogenously present compound (Jolivet et al., 1985Go; Poore et al., 1997Go), has been shown to be present in human plasma and urine at concentrations of approximately 0.1 and 20 µg/g, respectively (Hoffman et al., 1989Go; Tanaka et al., 1980Go). Analysis of control blood samples from untreated rats, in the absence of added GA internal standard, afforded true concentrations of endogenous, background GA of up to 2.1 µg/g rat blood (data not shown).

Blood levels of the 13C2-GA, derived from the test material, increased to a peak at 3 h postdosing, except for the 10 mg EG/kg dose groups, which demonstrated nondetectable levels of 13C2-GA over the entire 24-h time course. For all dose groups, blood levels of GA decreased by 24 h postdosing to undetectable levels. Elimination of GA from blood appeared biphasic until it reached background levels, with the ß phase for each curve defined as follows: 9–18 h (2500 mg EG/kg), 6–18 h (1000 mg EG/kg), 6–12 h (500 mg EG/kg), and 3–9 h (150 mg EG/kg). Estimation of the ß t1/2 of elimination of GA from blood resulted in similar values across dose levels (1.1–1.9 h), based on these biphasic elimination curves. No pharmacokinetic parameters were estimated for GA data from the 10 mg EG/kg dose groups, as those samples were below LOQ levels.

Examination of GA blood levels as a function of dose demonstrated nonlinear, dose-dependent kinetics. Figure 3Go depicts graphically the Cmax blood levels determined for GA, and compares them with estimated examples of what the Cmax levels would have been for a linear relationship between Cmax and administered dose (GA Ex Cmax). At the 10 and 150 mg EG/kg dose levels, GA blood levels appeared to be roughly proportional to dose, given that GA levels were < 2.1 µg/g for the 10 mg EG/kg dose, as compared to 20.6 µg/g for the 150 mg EG/kg dose (15-fold higher dose). A marked shift in GA blood kinetics was observed as the dose increased from 150 to 500 mg EG/kg. Over this 3.3-fold dose interval, the Cmax for GA increased by a factor of 6.4, which was a clear disproportionate increase. This disproportionality with dose persisted to a somewhat lesser degree between the 500 and 1000 mg EG/kg dose levels, where the 2-fold increase in dose brought about a 2.8-fold increase in GA Cmax.



View larger version (12K):
[in this window]
[in a new window]
 
FIG. 3. Disproportionate dose-response relationship between orally administered dose of ethylene glycol (EG) and resulting blood Cmax and AUC values quantified for its metabolite, glycolic acid (GA), in pregnant S-D rats. Experimental data are found in Table 2Go. Values for a linear response were estimated based on the 150 mg EG/kg dose measured values for GA, and are shown as dotted lines (- - - -) for comparison.

 
A similar shift, with a disproportionate increase, was found for the AUC values obtained from the GA blood concentration-time courses, particularly between 150 to 500 mg EG/kg. This shift is depicted graphically in Figure 3Go, where the actual AUC values for GA are compared with estimated values for a linear relationship between AUC and administered dose (GA Ex AUC). Estimated AUC for GA increased by 7.6- and 2.9-fold for a 3.3- and a 2-fold increase in administered dose, for the 150–500 and 500-1000 mg EG/kg dose group intervals, respectively. Thus, elimination of GA from blood demonstrated saturation at dose levels of 500 mg EG/kg and above.

Comparison of GA Cmax values for the 2 top dose levels showed a less than dose-proportionate increase, with only a 25% increase for a 2.5-fold increase in dose. In fact, the GA Cmax for 1000 mg EG/kg was basically equivalent to the GA Cmax for the 2500 mg EG/kg dose level. However, AUC values for GA were dose-proportionate for this dose interval of 1000 to 2500 mg EG/kg, indicating that over time, a similar fraction of administered dose of EG was converted to GA for both 1000 and 2500 mg EG/kg.

Time course of OX in blood.
OX, another endogenously present compound (Poore et al., 1997Go; Ribaya-Mercado and Gershoff, 1984Go), has been found in human urine at concentrations of approximately 20 µg/g (Tanaka et al., 1980Go). Analysis of control blood samples from untreated rats, in the absence of added OX internal standard, afforded true concentrations of endogenous, unlabelled OX of up to 4.9 µg/g rat blood (data not shown).

The concentrations of blood 13C2-OX, derived from the test material, varied between undetectable and about 2 times the limit of quantitation (4.9 µg OX/g blood) over the 24-h collection period. There was no obvious pattern to OX blood levels; while the 2500 mg EG/kg dose groups had the most time points with detectable OX levels, the highest mean OX value was obtained with a sample from the 18-h 1000 mg EG/kg dose group (9.2 ± 5.3). Based on these data, it does not appear that OX accumulates in P or NP female rats following oral gavage administration of EG. Given what appeared as mostly trace levels of OX, no additional analysis of the OX data was indicated or feasible. The lack of any obvious dose-response relationship in OX blood levels over such a large dose range (250-fold increase) suggests that OX does not play a major role in the expression of developmental toxicity of EG in rats.

Urinary Profiles and Interval Excretion of EG, GA, and OX
Urinary 13C-labeled-EG, -GA, and -OX levels were quantified over 12-h intervals, and the results are summarized in Table 3Go, both as total amount excreted (µg) and as fraction of the administered dose. Again, there were no substantial differences between the P and NP dose groups, at comparable dose levels. Therefore, gestation has no impact on urinary elimination of orally administered EG to female rats at this stage of pregnancy (GD 10–11).


View this table:
[in this window]
[in a new window]
 
TABLE 3 Urinary Interval Elimination of EG, GA, and OX following Oral Administration of EG to Female S-D Rats
 
Overall urinary elimination of EG and its metabolites demonstrated dose-dependency, with the high dose groups (2500 mg EG/kg) eliminating almost 70% of administered dose in urine, compared with about 16% of administered dose eliminated via urine by the low dose groups (10 mg EG/kg).

Quantitation of individual metabolites in urine demonstrated that the shift in urinary elimination was mainly due to increased urinary GA and EG, and not to increased elimination of OX (Table 3Go). Parent EG was always detectable in urine. The fraction of administered dose represented by urinary EG increased with increasing dose, reaching a maximum of about 42% of administered dose at the 500 mg EG/kg dose group, above which urinary EG remained around 40%. Therefore, the 3 highest dose groups all resulted in elimination of a similar fraction of administered dose as urinary EG. This suggests that renal elimination of EG was never saturated, even at the highest dose level. This supports the hypothesis that the initial saturation in EG blood levels was due to saturation of metabolic conversion of EG to GA and not to saturation of renal elimination of EG.

The increase in urinary EG elimination probably corresponded in large part with saturation of metabolism of EG to GA, resulting in an increased amount of EG available for renal elimination. In fact, the first collection intervals for the 500 and 1000 mg EG/kg dose groups showed 2–3 times as much urinary EG as GA. Then, once the metabolic conversion was no longer saturated and the excess EG had cleared (i.e., during the 12–24 h collection interval), the ratio of urinary GA/EG was roughly 1, while the overall percentage of the administered dose eliminated in the second interval was considerably decreased.

Urinary GA was quantifiable at all intervals, except the lowest dose levels for the 12–24-h interval. The 150 mg EG/kg and 10 mg EG/kg dose levels resulted in similar percentages of administered dose eliminated as urinary GA, representing only about 1% of the administered dose, clearly a minor fraction. However, the fraction of the administered dose eliminated as urinary GA then increased with increasing dose above 150 mg EG/kg, until it was present in roughly equal proportions with EG, as mentioned above. Comparison across dose levels demonstrated a disproportionate increase in urinary GA starting with the 500 mg EG/kg dose group, where the percentage of administered dose eliminated as urinary GA increased about 11-fold compared with 150 mg EG/kg, over a dose range of only 3.3-fold. The fraction of administered dose represented by urinary GA increased for the 1000 mg EG/kg and again for the 2500 mg EG/kg dose groups, up to about 20% and 33% of the administered dose, respectively. The shift in urinary GA levels corresponded with the shift in blood GA levels discussed earlier, indicating that renal elimination of GA was never saturated.

Urinary OX represented a constant fraction of administered dose across all dose levels. This resulted in a larger total amount of urinary OX with increasing dose, despite the continued endogenous/background blood concentrations of OX, as discussed above.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The data presented here clearly demonstrate that the pharmacokinetic parameters for EG and its metabolites reported here were not affected by GD 10–11 stage pregnancy in Sprague-Dawley rats. This signifies that the rich database on EG pharmacokinetics and metabolism, collected in NP rats, would be applicable to P rats (at least up to GD 11), and can serve, for example, in developing physiologically-based pharmacokinetic models to describe metabolism of EG in both pregnant and nongravid rats.

Overall, the mean EG blood values presented here for NP and P rats are in agreement with values previously published by Frantz et al. (1996a) for NP female rats, based on comparison of data for the 10 and 1000 mg EG/kg dose levels that were common to both studies. In addition, the EG data presented here demonstrated linearity in EG elimination from blood across a wide dose range (10–1000 mg EG/kg). Frantz et al. (1996a) also found that parent EG decreased in a linear fashion across dose levels following oral gavage administration to NP female rats, and reported ß-phase elimination t1/2 values in the same range as those reported here.

These data demonstrated the expected nonlinear dose dependency of GA kinetics, similar to that reported in previous studies with NP rats. For example, Frantz et al. (1996a,b) also found evidence suggesting a shift in GA blood kinetics for high oral bolus doses of EG, although the precise location of this shift along the dose response curve was not determined. They reported the shift to lie between 10 and 400 mg EG/kg. The current data narrow that range to between 150 and 500 mg EG/kg, comparable to the reported NOEL (500 mg EG/kg) for developmental toxicity (Carney et al., 1996Go, 1999Go; Neeper-Bradley et al., 1995Go). This shift in metabolism led to a disproportionate increase in blood Cmax and AUC of GA, and resulted in an increased urinary elimination of GA, both of which have been shown previously in NP rats (Frantz et al., 1996aGo, bGo; Marshall, 1982Go).

However, the additionally less-than-dose-proportionate increase (only 25%) in GA Cmax for 1000 vs. 2500 mg EG/kg, coupled with dose-proportionality in those GA blood AUC values, have not been shown previously. GA blood levels and the resulting GA blood AUC depend on the combined effects of how much GA is formed enzymatically from EG and how fast the GA formed is eliminated from blood. The shift from greater-than- to less-than-dose-proportionate increase demonstrated for GA Cmax values suggests that, either there is not as much GA forming, or that what is formed is eliminated more rapidly. The concomitant dose-proportionate increase in GA AUC does not support any increase in elimination of GA, so would point towards the former possibility of a probable saturation of enzymatic formation of GA. Thus, these data may indicate that metabolic conversion of EG to GA was approaching saturation at the EG blood levels obtained following administration of 2500 mg EG/kg, and that GA was being formed at a maximum rate between 1000 and 2500 mg EG/kg. In support of this hypothesis, as mentioned earlier, the EG Cmax values showed a greater than dose-proportionate increase from 1000 to 2500 mg EG/kg, with about a 4-fold increase in Cmax vs. a 2.5-fold one in administered dose. This suggests that the apparent initial saturation of EG elimination from blood, seen at the 2500 mg EG/kg dose, was most probably due to saturation of the metabolic conversion of EG to GA, rather than to saturation of renal elimination of EG. The conversion of EG to GA by alcohol/aldehyde dehydrogenase (ADH/ALDH) is one of 2 known rate-limiting steps in the EG metabolic pathway. The other major rate-limiting step is the further oxidation of GA.

Overall urinary elimination (percentage of the administered dose) was not affected by pregnancy at GD 10–11, and P and NP rats demonstrated similar urinary elimination kinetics for EG and its metabolites for the 10 and 2500 mg EG/kg doses, respectively. However, overall urinary elimination did demonstrate dose-dependency, with the high dose groups (2500 mg EG/kg) eliminating almost 70% of the administered dose in urine, compared with about 16% of the administered dose eliminated via urine by the low dose groups (10 mg EG/kg). This is in agreement with previously published data demonstrating a shift in disposition of 14C-EG-derived radioactivity from exhalation of 14CO2 at low doses to urinary elimination of radioactivity at high doses following intravenous and oral administration of 14C-EG to female rats (Frantz et al., 1996bGo; Marshall, 1982Go).

The shift in urinary GA levels corresponded with the shift in blood GA levels discussed earlier, indicating that renal elimination of GA was not saturated. In fact, on a wt/wt basis, urinary elimination of GA increased about 3.5- and 4-fold over a 2- and 2.5-fold increase in administered dose, between 500 and 1000 mg EG/kg, and between 1000 and 2500 mg EG/kg, respectively. This suggests that the saturation in elimination of GA from blood was due to saturation of downstream metabolism of GA, and not to saturation of renal elimination of GA. Data from Frantz et al. (1996b), demonstrating a shift in the disposition of 14C-labeled EG, from formation of 14CO2 at a low dose level (presumably representing downstream metabolism of the GA), to increased urinary radioactivity at a high dose level, supports the conclusion of no saturation of renal elimination of GA. Marshall (1982) showed that GA in urine accounted for approximately 2% of a 20 or 200 mg EG/kg bolus dose of EG, but approximately 20% of a 1000 or 2000 mg EG/kg dose.

OX blood levels remained consistently at or below quantitation levels in blood, even with increasing doses of EG. These data are consistent with a slower conversion of GA to OX than metabolism of OX to downstream products such as CO2. In contrast, the fraction of dose eliminated via the urine as OX remained constant across doses, resulting in increasing amounts of urinary OX eliminated with increased dose levels of EG. Renal formation of OX from EG and/or GA is one possible explanation of the differences between blood and urinary OX levels. If intrarenally formed OX were then eliminated via urine without any reabsorption, there would not be any reflection of its formation in systemic blood OX levels. There are no data on intrarenal formation and elimination of OX from EG. In any case, OX was a very minor metabolite in both blood and urine at all dose levels, suggesting that it is not a major factor in EG developmental toxicity. This conclusion is further supported by data from a preliminary study in which levels of OX were measured in the exocoelomic fluid of GD 10 rat conceptuses following gavage exposure to EG. Exocoelomic fluid OX averaged approximately 16.4 µg/g (0.13 mM) after a 2500 mg EG/kg dose, and was nondetectable (LOD = 5 µg/g) following a 500 mg EG/kg dose (Carney et al., 1998Go). A rat whole embryo culture study with OX indicated only minor growth inhibitory effects at a concentration of 126 µg/ml (1 mM) oxalate (Klug and Jaeckh, 1999Go), suggesting that embryonic OX levels in vivo are too low to cause significant toxicity to the embryo.

Consistent with GA being the proximate developmental toxicant, the location of the shift in GA metabolism along the dose response spectrum corresponds quite well with the dose response for developmental toxicity in the rat, as previously discussed (NOEL = 500 mg EG/kg; LOEL = 1000 mg EG/kg; Carney et al., 1996, 1999; Neeper-Bradley et al., 1995). The mean Cmax GA blood levels from the 2500 and 1000 mg EG/kg dose groups were 452 and 363 µg GA/g blood, respectively, indicating that very high maternal blood levels of GA are required to cause developmental toxicity. These values correspond with about 5.9 and 4.8 mM GA, respectively, falling between the concentrations resulting in abnormal embryos (12.5 mM GA) and normal embryos (2.5 mM) from whole embryo culture studies in vitro (Carney et al., 1996Go; Klug and Jackh, 1999). Recent evidence also has shown that levels of GA in rat exocoelomic fluid are almost 2-fold higher than in maternal rat blood, such that the critical level of GA exposure to the embryo may be closer to 10–12 mM following EG doses at or above 1000 mg/kg (Carney et al., 1998Go). To put these GA levels into perspective, clinical studies of human EG intoxications associated with intentional ingestion showed peak GA blood levels of 7–29 mM (Jacobsen et al., 1984Go). However, blood GA levels associated with normal handling and use of EG would not be expected to even approach these values, based on the slower dose-rate for dermal and inhalation exposures and the consequently low likelihood of saturating EG metabolizing enzymes.

In summary, the salient findings of this study were: (1) the shift in GA kinetics previously seen in NP rats also occurs in P (GD 10–11) rats; (2) the onset of this shift occurs at a gavage dose level of 150–500 mg EG/kg; (3) the correspondence of the shift in GA kinetics with the NOEL and LOEL for developmental toxicity adds further evidence supporting GA as the proximate developmental toxicant; and (4) developmental toxicity appears to require maternal blood GA levels in the millimolar range. Achievement of such high GA levels seems plausible only for high dose, oral bolus exposures to EG, but is extremely unlikely for typical human occupational and ambient exposures.


    ACKNOWLEDGMENTS
 
Expert technical assistance in conducting this study was provided by A. Liberacki, A. Wardynski, K. Gibson, C. Thornton, J. Whalen, J. Ormand, B. Kropscott, J. Hammond, and F. Lee. Expert analytical support was provided by D. McNett, D. Markham, and K. Engle. We are also grateful for the veterinary expertise of J. Lacher. Financial support was provided by The Monoethylene Glycol Sector Group of CEFIC, and the Ethylene Glycol Panel of The American Chemistry Council.


    NOTES
 
1 To whom correspondence should be addressed at Dow Europe, SA, Bachtobelstrasse, 3, CH-8810, HORGEN, Switzerland. Fax: +41-1-728-2965 E-mail: lpottenger{at}dow.com. Back

Portions of these data were presented at Tox Forum, July 2000.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Carney, E. W. (1994). An integrated perspective on the developmental toxicity of ethylene glycol. Reprod. Toxicol. 8, 99–113.[ISI][Medline]

Carney, E. W., Freshour, N. L., Dittenber, D. L., and Dryzga M. D. (1999). Ethylene glycol developmental toxicity: Unraveling the roles of glycolic acid and metabolic acidosis. Toxicol. Sci. 50, 117–126.[Abstract]

Carney, E. W., Liberacki, A. B., Bartels, M. J., and Breslin, W. J. (1996). Identification of proximate toxicant for ethylene glycol developmental toxicity using rat whole embryo culture. Teratology 53, 38–46.[ISI][Medline]

Carney, E. W., Pottenger, L. H., Bartels, M. J., Jaeckh, R., and Quast, J. F (1998). Comparative pharmacokinetics and metabolism of ethylene glycol in pregnant rats and rabbits. Toxicol. Lett. 95 (Suppl. 1), P4C27.

EPA-TSCA (1989). US Environmental Protection Agency-TSCA GLPS. Title 40 CFR, Part 792-Toxic Substances Control Act (TSCA); Good Laboratory Practice Standards, Final Rule.

Frantz, S. W., Beskitt, J. L, Grosse, C. M., Tallant, M. J., Dietz, F. K., and Ballantyne, B. (1996a). Pharmacokinetics of ethylene glycol. I. Plasma disposition after single intravenous, peroral, or percutaneous doses in female Sprague-Dawley rats and CD-1 mice. Drug Metab. Dispos. 24, 911–921.[Abstract]

Frantz, S. W., Beskitt, J. L., Grosse, C. M., Tallant, M. J., Dietz, F. K., and Ballantyne B. (1996b). Pharmacokinetics of ethylene glycol. II. Tissue distribution, dose-dependent elimination, and identification of urinary metabolites following single intravenous, peroral or percutaneous doses in the female Sprague-Dawley rat and CD-1 mouse. Xenobiotica 26, 1195–1220.[ISI][Medline]

Frantz, S. W., Beskitt, J. L., Grosse, C. M., Tallant, M. J., Zourelias, L. A., and Ballantyne, B. (1996c). Pharmacokinetics of ethylene glycol. III. Plasma disposition and metabolic fate after single increasing intravenous, peroral or percutaneous doses in male Sprague-Dawley rats. Xenobiotica 26, 515–539.[ISI][Medline]

Gibaldi, M., and Perrier, D. (1982). Pharmacokinetics, 2nd ed. Marcel Dekker, Inc., New York.

Hoffman, G., Aramaki, S., Blum-Hoffman, E., and Sweetman, L. (1989). Quantitative analysis for organic acids in biological samples: Batch isolation followed by gas chromatographic-mass spectrometric analysis. Clin. Chem. 35, 587–595.[Abstract/Free Full Text]

Jacobsen, D., and McMartin, K. E. (1986). Methanol and ethylene glycol poisonings: Mechanism of toxicity, clinical course, diagnosis and treatment. Med. Toxicol. 1, 309–334.[ISI]

Jacobsen, D., Ovrebo, S., Ostborg, J., and Serjersted, O. M. (1984). Glycolate causes the acidosis in ethylene glycol poisoning and is effectively removed by hemodialysis. Acta. Med. Scand. 216, 409–416.[ISI][Medline]

Jolivet, P., Gans, P., and Triantophylides, C. (1985). Determination of glycolic acid level in higher plants during photorespiration by stable isotope dilution mass spectrometry with double-labelling experiments. Anal. Biochem. 147, 86–91.[ISI][Medline]

Khera, K. S. (1991). Chemically induced alterations in maternal homeostasis and histology of conceptus: Their etiologic significance in rat fetal anomalies. Teratology 44, 259–297.[ISI][Medline]

Klug, S., and Jaeckh, R. (1999). Effects of ethylene glycol and metabolites on in vitro development of rat embryos during organogenesis. German Soc. Toxicol.

Marshall, T. C. (1982). Dose-dependent disposition of ethylene glycol in the rat after intravenous administration. J. Toxicol. Environ. Health. 10, 397–409.[ISI][Medline]

Miller, R. K. (1983). Perinatal toxicology: Its recognition and fundamentals. In Reproductive Toxicology (D. R. Mattison, Ed.), pp. 205–244. Alan R. Liss, Inc., New York.

Munley, S. M., Kennedy, G. L., and Hurtt, M. E. (1999). Developmental toxicity study of glycolic acid in rats. Drug Chem. Toxicol. 22, 569–582.[ISI][Medline]

Neeper-Bradley, T. L., Tyl, R. W., Fisher, L. C., Kubena, M. F., Vrbanic, M. A., and Losco, P. E. (1995). Determination of a no-observed-effect-level for developmental toxicity of ethylene glycol administered by gavage to CD rats and CD-1 mice. Fundam. Appl. Toxicol. 27, 121–130.[ISI][Medline]

OECD (1982). Organisation for Economic Co-Operation and Development-Principles of Good Laboratory Practice, The OECD Principles for GLP as specified by the European Economic Community, Council Directive 87/18 EEC.

Poore, R. E., Hurst, C. H., Assimos, D. G., and Holmes, R. P. (1997). Pathways of hepatic oxalate synthesis and their regulation. Am. J. Physiol. 272, C289–294.[Abstract/Free Full Text]

Price, C. J., Kimmel, C. A., Tyl, R. W., and Marr, M. C. (1985). The developmental toxicity of ethylene glycol in rats and mice. Toxicol. Appl. Pharmacol. 81, 113–127.[ISI][Medline]

Ribaya-Mercado, J. D., and Gershoff, S. M. (1984). Effects of sugars and vitamin B-6 deficiency on oxalate synthesis in rats. J. Nutr. 114, 1447–1453.[ISI][Medline]

Tanaka, K., West-Dull, A., Hine., D. G., Lynn, T. B., and Lowe, T. (1980). Gas-chromatographic method of analysis for urinary organic acids. II. Description of the procedure, and its application to diagnosis of patients with organic acidurias. Clin. Chem. 26, 1847–1853.[Abstract/Free Full Text]

Tyl, R. W., Price, C. J., Marr, M. C., Myers, C. B., Seely, J. C., Heindel, J. J., and Schwetz, B. A. (1993). Developmental toxicity evaluation of ethylene glycol by gavage in New Zealand White rabbits. Fundam. Appl. Toxicol. 21, 402–412.[ISI]