- Email: [email protected]
A biologically based toxicokinetic model for pyrene in rainbow trout
TOXICOLOGY
AND
APPLIED
A Biologically
PHARMACOLOGY
Based Toxicokinetic F. C. P.
Environmental
Toxicology
(1991)
110,390-402
LAW,
Model for Pyrene in Rainbow
S. ABEDINI,
AND C. J. KENNEDY
Program, Department qf Biological Burnaby, British Columbia, Canada
Received
November
Trout’
19. 1990; accepted
Sciences. Simon V5A IS6
Fraser
University,
June 11, 1991
Biologically Based Toxicokinetic Model for Pyrene in Rainbow Trout. LAW, F. C. P., ABEDINI, KENNEDY, C. J. (1991). Toxicol. Appl. Pharmacol. 110,390-402. A biologically based toxicokinetic model was developed to stimulate the metabolic disposition of pyrene in trout with an average body weight of 450 g and dosed with a single bolus injection of the chemical (10 mg/ kg). The model consists of a membrane-limited muscle compartment and six flow-limited compartments including the gills, liver, gut, kidney, carcass, and blood. The compartments are rep resented by mass balance equations including terms for the binding of pyrene to tissue and blood proteins, biotransformation, penetration rate into the muscle, blood flow rate, tissue mass, etc. The model also provides for nonsaturable and saturable clearances of pyrene by the liver and kidney. Michaelis-Menten constants for pyrene metabolism (K,, V,,,,,) were determined from in vitro experiments using isolated liver cells. Renal clearance of pyrene was very close to the glomerulus filtration rate of trout. Solution of the system of equations yielded the time courses of pyrene concentration in the tissues. Predicted concentrations of pyrene in the gills, liver, gut, kidney, muscle, and blood were consistent with experimental observations for at least 6 days. The model was validated by comparing the model predicted and experimental results of trout weighing 285 g and dosed with a single intraarterial dose (3 mg/kg) of pyrene. The predicted pyrene concentrations also were in adequate agreement with the empirical data. o 1991 Academic Press. hc. A
S.,
AND
The polycyclic aromatic hydrocarbons (PAHs) are pollutants released into the environment as a result of incomplete combustion of organic matter (Hase and Hites, 1978). Therefore, PAHs have been found in sewage, lake, and ground waters (Bomeff, 1977) and in marine sediments and food organisms (Malins et al., 1985). There have been many reported studies on the accumulation of PAHs by aquatic organisms (Varanasi and Malins, 1977; Neff and Anderson, 198 1). Waterborne PAHs and those associated with sediments are taken up rapidly by fish (Lee et al., 1972; Anderson et al., 1974); equilibrium concentrations of PAHs are often reached in fish tissue in 24 hr or less after ex’ Presented in part at the 10th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Toronto. Ontario. October 2%November 2, 1989. 0041-008X/91
$3.00
Copyright 0 199 1 by Academic Press. Inc. All rights of reproduction in any form reserved.
posure. In contrast, the uptake of PAHs from food and sediments are much slower. In most fish species, more than half of the orally administered PAHs are unabsorbed but associated with the digestive tract or its contents (Whittle et al., 1977). Pyrene is one of the major PAHs found in fish taken from polluted waters (Krahn et al., 1987). Although pyrene is not considered a potent carcinogen, it probably is a cocarcinogen (Weinstein and Troll, 1977). Pyrene is absorbed rapidly but eliminated slowly by trout; the uptake and elimination of pyrene by trout can be described adequately by a classical, three-compartment toxicokinetic model (Kennedy and Law, 1990). Biologically (physiologically) based toxicokinetic models (BBTM) have been used to reduce the uncertainties in the risk assessment process (NRC, 1987; U.S. EPA, 1987). BBTM 390
PYRENE TOXICOIUNETIC
also is a tool to increase understanding of interspecies differences in the biological fates of chemicals. However, very few BBTM, with the exception of the flow-limited models of methotrexate in sting rays (Dasyatidae sabina) (Zaharko et al., 1972) and phenol red in dogfish shark (Squalus acanthias) (Bungay et al., 1976), have been reported in the aquatic species. In the course of this work, Nichols et al. (1989) have reported a BBTM for three chlorinated ethanes in rainbow trout (Oncorhynthus mykiss). In view of the wide distribution of PAHs in the aquatic environment, it was decided that a BBTM for pyrene in trout should facilitate the use of fish in pharrnacokinetic studies of drugs, environmental pollutants, and other xenobiotics. The model also would permit a better evaluation of exposure and health risk assessment of humans who use PAH-contaminated fish as food. The work reported here describes the development of a BBTM of pyrene in trout. MATERIALS
AND
MODEL
391
IN TROUT
Feed obtained from Moore-Clark Co. (Vancouver, British Columbia). They were not fed for 3 days prior to an experiment. After a trout was anesthetized with MS 222, the dorsal aorta was cannulated according to the procedure of Smith and Bell (1964). The fish was allowed to recover from surgery for 24 hr in a darkened Plexiglas box (20 X 20 X 80 cm) supplied with continuously flowing water at 10°C.
Intraarterial
Administration
of Pyrene
Unlabeled pyrene was dissolved in Tween 80 before being mixed with 0.2 ml of modified Hank’s medium (Moon et al., 1985). The solution was injected as a bolus through the cannula into the dorsal aorta of trout. This was followed by the injection of 0.2 ml of fresh Hank’s medium through the cannula to ensure that the entire dose entered the circulatory system. Little or no pyrene was found to adsorb onto the cannula. Two different doses of pyrene were studied: trout with an average body weight (BW) of 285 g were given 3 mg/kg pyrene; trout with an average BW of 450 g were dosed with 10 mg/kg pyrene. At different times after pyrene administration, three trout were removed from the tank, euthenized, and frozen at -5°C until analysis. The intraarterial route was used to develop the BBTM instead of the oral or branchial route since the intraarterial route did not involve complicated absorption processes.
METHODS Unchanged
Pyrene
Concentration in Tissues
Chemicals
[4,5,9,10-‘4C]Pyrene (sp act 56 mCi/mmol) was purchased from Amersham Canada Ltd. (Oakville, Ontario). Unlabeled pyrene was obtained from Aldrich Chemical Co. (Milwaukee, WI). The 14C-labeled pyrene was purified by thin-layer chromatography (TLC) on silica-gel plates using bexanes as the developing solvent. The radiochemical purity of the “‘C-labeled pyrene, determined by TLC, exceeded 99%. The unlabeled pyrene was puriiied by repeated recrystallization from methanol. The purity of pyrene was determined by high-pressure liquid chromatography (HPLC) and exceeded 99%. Polyoxyethylene sorbitan monooleate (Tween 80) and ethyl-N-aminobenzate methane sulfonic acid (MS 222) were purchased from Sigma Chemical Co. (St. Louis, MO).
Fish
Rainbow trout (0. mykiss) weighing between 200-500 g were obtained from Spring Valley Trout Farms, British Columbia. The trout were kept in flowing, dechlorinated water at 10 +_ 2°C in holding tanks for at least 3 weeks before use. The fish were fed daily with New Age Fish
After the fish were thawed, we removed the liver, kidney, gut, gills, spleen, skin, white muscle, and blood from the carcass. Fat was trimmed off from the gut which included the esophagus, stomach, and large and small intestines. Gut contents were also discarded. The organs and tissues were rinsed with distilled water, blotted dry, and weighed. A 25% w/v tissue homogenate was prepared in distilled water. An aliquot of the homogenate (0.2-0.3 ml) was pipetted into a centrifuge tube and mixed with 0.5 ml 0. I N sulfuric acid and 0.5 ml distilled water. The mixture was vortexed and extracted three times with 3-ml portions of methylene chloride. The methylene chloride extracts were combined and passed through a glass column containing 5 g of florosil. The column was washed with 10 ml of fresh methylene chloride. The eluants were collected, combined, and evaporated down to dryness. The residues were redissolved in hexanes containing benzanthracene (1 pg/ml) as an internal standard. The hexane solution (20 ~1) was injected directly into a Hewlett-Packard liquid chromatograph (Model 1090, Avondale, PA) equipped with an ODS-Hypersil column (Hewlett-Packard Co., 5 pm, 100 X 4.6 mm i.d.) and a diode-array uv detector which had a setpoint at 230 nm. Column temperature was 40°C. Column eluant (methanol:water/73:27) had a flow
392
LAW, ABEDINI,
rate of 1.5 ml/min. The extraction method was found to be very efficient in removing pyrene from the various fish tissues; percentage recoveries + SD of pyrene from liver, kidney, gill, gut, white muscle, and blood were 78.9 + 9.3, 75.7 + 12.3, 71.4 f 6.6, 80.5 + 6.0, 88.4 + 8.1, and 85.3 + 6.8, respectively.
Urinary
Excretion
of ‘T-Labeled
Pyrene
In a separate experiment, three trout were anaesthetized with MS 222 and the dorsal aorta was cannulated as described above. A polyethylene cannula also was inserted into the urino-genital papilla for urine collection (Holmes and Stainer, 1966). After the trout were injected intraarterially with a single dose (10 mg/kg) of “‘C-labeled pyrene. mine was collected as 24-hr samples for 6 days. The volumes of urine samples were measured. The radioactivities in the urine were determined by liquid scintillation counting.
Pyrene
Metabolism
by Isolated
Hepatocytes
Liver cells were isolated from trout according to the procedures of French et al. (198 I), Moon et al. (1985), and Walsh (1986). Briefly, fish were anesthetized with a solution of 0.2 g/liter MS 222 and 0.2 g/liter NaHCO, until opercular movement ceased. A ventral incision was made in the abdomen of the fish and the hepatic portal vein cut. The intestinal vein was cannulated with a 23gauge stainless-steelneedle which was sutured in place with silk sutures. The liver was perfused with well-oxygenated Medium A (Moon et al., 1985) supplied by a peristalic pump for 10 min to clear the liver of blood. The flow rate was adjusted to approximately 2 ml/min/g liver to guarantee fully aerobic conditions. The liver was perfused for an additional 30-45 min with Medium B (Moon et al.. 1985) which contained collagenase and hyaluronidase. After the gall bladder was removed, the liver was chopped with a razor blade. The minced liver was filtered through a 253- pm and then a 73-pm nylon mesh. The cells in the filtrate were collected by centrifugation (50g) for 2 min in a Sorval RC-5B super speed centrifuge (DuPont, Wilmington, DE). The supematant was decanted and the cells pellet resuspended in ice-cold Medium C (Moon et al., 1985). Cell viability was tested by the trypan blue exclusion test. The viability ofthe isolated hepatocytes exceeded 95%. Incubation of pyrene with the hepatocytes was performed within 3 hr after cell preparation; an aliquot (1.9 ml) of the cell suspension (3 X lo6 cells/ml) was preincubated for I5 min in a liquid scintillation vial under an atmosphere of air containing 0.25% CO,. A solution of “C-labeled pyrene (0. I ml) was added to the liver cells (0.5, 2.5, 5.0, 25, or 50 /IM final concentration). The incubation mixture was gassed for the duration of the incubation. At the conclusion of the incubation, the cells
AND KENNEDY were isolated by centrifugation and homogenized with a Polytron homogenizer (Brinkman Co., Rexdale, Ontario). The homogenate was extracted by methylene chloride as described above. Radioactivities in the methylene chloride extract and the remaining aqueous layer were determined separately by liquid scintillation counting. The methylene chloride extract was also analyzed by HPLC for unchanged pyrene as described before. Since [‘%Z]pyrene was metabolized by the hepatocytes mainly to water-soluble metabelite(s), the radiolabel counts in the remaining aqueous layer represented the amount of pyrene metabolites formed. In vitro Michaelis-Menten constants of pyrene metabolism by the hepatocytes were determined from the Lineweaver-Burk plot of the incubation data.
Biological
Simulation
Model
The pyrene model (Fig. 1) contained 7 compartments: blood, gills, liver, gut, kidney, muscle and carcass. The gut compartment included the highly perfused organs such as the esophagus, stomach, intestine, pyloric caeca, spleen, ventricle, and red muscle. The carcass was comprised of remaining tissues such as skin, skull, branchial basket fins, and eyes.All tissues with the exception of the muscle were depicted as blood-flow-limited compartments. The muscle was membrane-limited since it met the criterion that Q/ H 1, where Q is organ blood flow and H is mass transfer coefficient (Dedrick and Bischoff, 1968). Elimination of
-
KIDNEY
‘QK
. KK I
*
MUSCLE
-:ARCASS -
QM
QC
1. Schematic diagram of the biologically based toxicokinetic model used to simulate the metabolic disposition of pyrene in trout. The symbols and parameters used to describe the model are included in Table 1 and the Ap pendix. FIG.
PYRENE TOXICOKINETIC pyrene was modeled by hepatic and renal clearances since pyrene was excreted by trout mainly as water-soluble metabolite(s) in the bile and urine (Kennedy and Law, 1990). Metabolism was assumed to occur only in the liver and was described by a combination of linear metabolism and Michaelis-Menten metabolism. Renal clearance was a firstorder rate process. It was obtained by trial and error and was found to be very close to the glomerulus filtration rate (GFR) of trout. Mass balance differential equations for the model and the definition for the algebraic terms are given in the Appendix.
Model
Parameters
Three types of parameters are needed to implement the BBTM; physiological parameters of trout, tissue:blood partition coefficient of pyrene, and the clearances of pyrene from trout. The data, which were obtained by experiment or from the literature are summarized in Table 1. Physiological parameters. The liver, gut, muscle, gills, kidney, and blood were assigned volumes equal to 1.16, 8.52,46.5, 3.90,0.8, and 4.11% of body weight. The sum of these tissue volumes accounted for 65.0% of total BW, the remaining 35.0% was the carcass. Extracellular and intracellular fluid spaces of muscle were 0.78 and 45.72% of total BW, respectively.Cardiac output (CO) was assigned a value of 19.73 ml/min/kg. It was assumed that the gills received the entire CO. Blood flow distribution to the liver, gut, muscle, and kidney was, respectively, 2.75, 15.39, 39.77, and 10.22% of CO. Clearances. Total body clearance was obtained by dividing the administered dose with AUC, the total area under the blood pyrene concentration-time curve. AUC was shown to be 3888 &min/ml (Kennedy and Law, 1990). Rate constants of Michaelis-Menten metabolism by the liver, I’,,,, and K, , were derived from the isolated hepatocyte experiment (see above). Renal clearance (KK) was assigned a value of 169 ml/kg/day. First-order metabolic clearance (KL) of pyrene by a 450-g trout was estimated by the difference between total body clearance and the sum of renal and saturable hepatic clearances. The V,,,,,. KK. and KL values for the 285-g trout were scaled from the following allometric equations (Travis, 1987): V,,,,, = I.‘,,,,,, X KK = KK, X KL = KL
BW’O.” BW’O.”
c X BW’-O.“,
(1)
(2) (3)
where Vmaxc,Kk;, and KL, are the scaling coefficients for the maximum velocity of Michaelis-Menten metabolism, renal clearance, and first-order metabolic clearance, respectively. They were determined from trout with an average BW of 450 g. Gargas et al. (1986) have suggested that the metabolic rate VW, can be scaled using body
MODEL
IN TROUT
393
TABLE 1 KINETIC CONSTANTSAND MODEL PARAMETERS USED IN THE BIOLOGICALLY BASED TOXICOKINETIC MODEL FOR PYRENE
Trout Parameter
Abbreviations”
Vohnne (% body wt) Gills VGL Liver VL Gut VGT Kidney VK W. Muscle Extracellular VEM Intracellular VIM Carcass vc Blood VB Blood Flow (% cardiacoutput) Gills Liver QL-QGT Gut Kidney g W. Muscle QM Carcass QC Cardiac output (ml/min) QT Partition coefficients Gills/blood RGL Liver/blood RL Gut/blood RGT Kidney/blood RK W. Muscle/blood RM Carcass/blood RC Mass transfercoefficient H Muscle Biochemical constants (&mitt/g liver) ~nlax (a/ml) Kn (ml/mitt) KL (ml/min) Iw Dosem/kg)
450 g
285 g
3.90 1.16 8.52 0.80 46.50 0.78 45.72 35.0 4.11
3.90 1.16 8.52 0.80 46.50 0.78 45.72 35.0 4.11
100 2.15 15.39 10.23 39.77 31.86
100 2.75 15.39 10.23 39.77 31.86
8.875
5.622
2.15 37.0 1.90 15.3 0.20 7.75
2.75 37.0 1.90 15.3 0.20 7.75
0.035
0.035
0.255 3.05 0.40 0.06 10
0.185 3.05 0.56 0.04 3
a See Appendix for definitions of abbreviations for biochemical constants.
weight. Ramsey and Andersen (1984) have proposed that I’,,,, is proportional to the surface area. Tissue:bloodpartition coeficients. Two different methods were used to estimate tissue:blood partition coefficients: (a) In vitro tissue binding studies. These studies were carried out in dialysis cells which consisted of a dialysis bag (prepared from Spectra/par molecularporous membrane tubing, 18 mm X 50 ft, Spectrum Medical Industries, Inc. Los Angeles, CA) and a glass measuring cylinder. About 0.5 ml ofthe tissue homogenate (25% w/v) or whole blood (25% v/v) prepared in Tris-HCl buffer (0.1 M, pH 7.4) was pipetted into the dialysis bag and dialyzed against the
394
LAW, ABEDINI,
same buffer for I day at 4°C before use. The tissue homogenate or blood was then dialyzed against three different initial concentrations (0.1, 0.25, and 0.5 mM) of [%]pyrene solution in Tris-HCl buffer (0.1 M, pH 7.4) at 4°C for an additional 2 days. Preliminary studies using different dialysis durations (1, 2, and 3 days) suggested that equilibrium was reached by most fish tissues in 2 days. At the conclusion of the dialysis, radioactivities in the tissue and buffer were determined separately by liquid scintillation counting. The sum of buffer and tissue radioactivity was found to account for ~95% of the radioactivity initially added to the dialysis cells. Little or no pyrene was metabolized during dialysis since nearly all the radioactivities in the buffer and tissues were extractable by methylene chloride. The unbound fraction of [‘%Z]pyrene in each tissue was determined by dividing the chemical concentration in buffer by the concentration in the tissue homogenate as described by Lin et al. (1982). (b) In vivo experimental data. Initial estimates of tissue:blood partition coefficients also were determined from in vivo experimental data according to the method of Dedrick et al. (1973). Briefly. trout were injected intraarterially with a solution of unlabeled pyrene (10 mg/kg) as described before. At different times after pyrene administration (8. 16. 24, and 48 hr), three trout were euthanized and their organs removed. Unchanged pyrene concentrations in the tissues were determined by HPLC analyses of the methylene chloride extracts of tissue homogenates. A Freundlich-type plot (log-log) of tissue concentration vetsus blood concentration was prepared for each tissue. The tissue:blood partition coefficients were determined from the best straight line with a unit slope drawn through the data. These estimates were adjusted after comparing the experimental data of pyrene-treated trout (10 mg/kg) with the computer simulation. Mass transfer coejicient of muscle. The mass transfer coefficient of muscle (H) was estimated from the in vivo experimental data of muscle and blood. A detailed description of the procedure and the rationale of the methodology has been described by Dedrick and Bischoff ( 1968). Briefly, an initial estimate of H was determined from the slope of a straight line by plotting [VM(ACM/At)] versus [CBV-CM/RM)]. The estimate was adjusted after comparing the experimental data of muscle from pyrene-treated trout (10 mg/kg) with the computer simulation.
Computer
Simulation
The differential and algebraic equations describing the movement of pyrene through the trout were formulated as a computer program (see Appendix). The set of differential and algebraic equations was solved numerically with the aid of a Fortran-based software package, CSMP (IBM, 1976). The resultant solutions gave the predicted pyrene concentrations in the tissues and percentage dose excreted in the urine.
AND KENNEDY Determination
of Radioactivity
Radioactivity was determined by a Beckman LS-8000 liquid scintillation counter. A correction of quenching was made using the external standard technique. The radioactivity in the urine sample was determined directly by liquid scintillation counting after the addition of Biofluor. Tissue homogenates were digested with a 1-ml solution of Protosol and ethanol (1:2) at 60°C for 0.5 hr, decolorized with 30% hydrogen peroxide, and counted in a liquid scintillation counter after the addition of Bioflour.
RESULTS
AND DISCUSSION
We have presented a BBTM (Fig. 1) for predicting the concentrations of pyrene in seven tissue compartments of trout following a single intraarterial dose. The model was developed and validated with trout weighing 450 and 280 g, respectively. The model simulates most experimental data closely for the two administered doses (10 and 3 mg/kg). Predicted and observed pyrene concentration in the liver, gut, muscle, kidney, gills, and blood are shown in Figs. 2a-2f. The curves are the simulations based on the model illustrated in Fig. 1. The points represent the experimental data. We have selected 10 mg/kg as the upper dose limit of the study since pyrene solubility is limited by the volume of the injected solution. The intent of the dose-dose comparison is to delineate the rate-limiting or saturable process involved in the metabolic disposition of pyrene in trout. However, a threefold difference in the dose apparently is insufficient to detect if there are pyrene-related effects on the clearance process. The concentration profiles between the gut and gill compartments (Figs. 2b and 2c) are exceedingly similar; therefore, they could be modeled as a single compartment. Although the adipose tissue of the pyrene-treated trout contains a high level of the chemical (data not shown), fat is not modeled as a separate compartment since the volume of fat in trout is about 1% of the body weight (Gingerich et al., 1987). Moreover, the amounts of fat found in trout vary with individual fish. Groves (1970)
PYRENE TOXICOKINETIC
MODEL
395
IN TROUT
4 160
ak 0.01
4
c--v++ 0
20
40
60
100
120
140
160
FIG.2. Time courses of pyrene concentrations in trout tissues. Each point represents the experimental data from the tissue of one trout following an intraarterial injection of pyrene (0 10 mg/kg; A 3 mg/kg). Solid lines represent simulation using the model. a. liver; b, gill; c, gut; d, kidney: e, muscle; f, blood.
has suggested that fat is more a function of the nutritional history of the fish. After adjustment with the BW, the physiological and biochemical parameters used in modeling are exactly the same for both doses (Table 1). The physiological parameters also are consistent with the reported values of trout (Stevens, 1968; Daxboeck, 1981; Barron et al., 1987; and Gingerich et al., 1987). For example, the assigned white muscle volume was 46.5% BW (Table 1). This agrees with the reported value of Gingerich et al. (1987), al-
though it is less than the 66% BW assumed by Stevens (1968). The total blood volume (4.11% BW) is identical to that of Gingerich et al. (1987) but is slightly less than the 5.58 and 6.08% reported by Mill&m and Wood (1982) and Daxboeck (198 I), respectively. The liver volume is 1.16% BW (Table 1). This is very close to the liver volumes reported by Daxboeck (198 l), Gingerich et al. (1987), and Barron et al. (1987). The kidney volume is 0.8% BW; it is consistent with those reported by Barron et al. (1987) and Gingerich et al.
396
LAW,
ABEDINI,
AND
KENNEDY
4 160
,“oo
1 u 0
20
40
60
FIG.
Z-Continued
, I 100
L
\
120
140
4 160
PYRENE
TOXICOKINETIC
RG.
MODEL
IN
TROUT
397
2-Continued
(1987). The gill volume (3.9% BW) was taken from Stevens (1968). It is slightly higher than the 2.0 and 2.78% reported by Barron et al. (1987) and Gingerich et al. (198 l), respectively. The CO is 19.73 ml/min/kg. This is consistent with the CO reported for trout at a similar water temperature: 18.5 ml/min/kg (Davis and Cameron, 197 1) and 17.6 ml/min/ kg (Kiceniuk and Jones, 1977). Barron et al. (1987) have shown that the CO of trout declines linearly with decreasing acclimation temperature. If the regression relationship proposed by Barron et al. (1987) were used to calculate the CO, a much higher value (26.6 ml/min/kg) would be obtained, resulting in a lower model-predicted pyrene concentration in the blood compartment. The blood flow distribution to the various compartments of the BBTM also are consistent with the literature values (Daxboeck, 1981; Barron et al., 1987). For example, Barron et al. (1987) have shown that the liver, kidney, and white muscle of trout receive about 1.95, 12.8, and 38.9% of CO, respectively. In contrast, Daxboeck (198 1) has reported that the liver, kidney, and white muscle receive about 4.5-8.73, 4.5-9.26, and 36.449.0% of co. The model-predicted cumulative urinary excretion of 14C also agrees quite well with the
experimental data over the 6-day observation period (Fig. 3). The predicted cumulative urinary excretion at Day 6 also is consistent with the experimental value reported previously by Kennedy and Law (1990). The kinetic constants of pyrene metabolism (K,, I’,,,,,) (Table 1) were calculated from the Michaelis-Menten constants of two separate isolated hepatocyte experiments with different trout. Figure 4 shows that [14C]pyrene was metabolized by the hepatocytes linearly for at least 30 min. The V,,, values obtained from the two studies were 0.00 13 and 0.0 113 nmol/ min/ 1O6cells; the corresponding K,,, were 13.3 and 16.8 pM, respectively (Fig. 5). Previous studies indicated that each gram of liver yielded about 0.69-3.28 X 10’ cells (Parker et al., 1981) or 1.3 X lo8 cells (Selglen, 1973). To convert the Michaelis-Menten constants obtained from the hepatocyte studies to that of the intact liver, we assume each gram of trout liver yields about 2.0 X 10’ cells (average of 0.69 X 10’ and 3.28 X 10’). The calculated Vmaxand Km values of pyrene metabolism by the intact liver of a 450-g trout are 0.255 r.Lg/ min/g liver and 3.05 pg/ml, respectively (Table 1). The ultimate goal of the BBTM is to predict the biological fates of pyrene without resorting to prior in vivo experiments. However, several
398
LAW, ABEDINI,
AND KENNEDY
FIG. 3. Cumulative amount of [‘4C]pyrene equivalent collected in the urine of trout. The points are experimental data for a single intraarterial injection of 10 mg/kg into the dorsal aorta of three different trout. Solid lines represent model simulation of ‘%Zaccumulation in the urine.
of the parameters used in the model cannot be determined a priori with a great deal of confidence. These include the tissue:blood partition coefficients, mass transfer coefficient of muscle, extracellular and intracellular spaces of muscle, and renal clearance of pyrene by trout: (a) The tissue:blood partition coefficients used in the model were determined from the in vivo experimental data (Table 1).
5 a 0
20
10
rime
30
These differ greatly from those determined by the binding studies (1.94,2.56, 1.26,2.49, and 1.34 for the gills, liver, gut, kidney, and white muscle, respectively). A possible explanation for the discrepancy of results may be that the conditions for the binding experiment still are not optimal for equilibrium to be achieved. The appropriateness of using the in vivo tissue:
I 40
(min)
FIG. 4. Time course of “%-labeled pyrene metabolism by the isolated hepatocytes of trout. Displayed are the typical results of one of two experiments.
FIG. 5. A Lineweaver-Burk plot of “‘C-labeled pyrene metabolism by the isolated hepatocytes. Displayed are the typical results of one of two experiments.
PYRENE TOXICOKINETIC
blood partition coefficients in model simulations is further supported by the value of the apparent volume of distribution (I’,). The V, of trout calculated by summing the R X V products of Table 1 is 3.62 liters/kg. This is very close to the Vd (2.57 liters/kg) determined by the equation, dose/(AUC X X,), where X, (0.001 min-1) and AUC (3888 pg*min/ml) (Kennedy and Law, 1990) are, respectively, the slope of the terminal exponential phase and the area under the curve of the triexponential blood concentration-time curve of trout following an intraarterial injection of pyrene. (b) The muscle is a membrane-limited compartment in the model. Initially, the muscle compartment was modeled as a flow-limited compartment. However, a flow-limited muscle compartment overpredicted pyrene concentrations at early times. Moreover, the predicted peak concentration of pyrene occurred at a time much earlier than that of the experimental data. If the muscle flow rate were lowered arbitrarily by lo%, the simulations would agree more closely with the experimental data. This suggests that the transfer of pyrene from blood into muscle is slower than would be predicted on the basis of published blood flow. The slow uptake of pyrene by the muscle cells probably is due to low cell membrane permeabilities or maldistribution (shunt) of blood flow. Further studies will be required to clarify the mechanism(s) of pyrene uptake by the muscle cells of trout. Randall (1970) has shown that white muscle, constituting the bulk of trout body, is poorly vascularized and is used only in burst activity of short duration. The slow release of lactate from trout muscle has been explained by poor vascularization (Randall, 1970) and the diffusion resistance of the muscle cell membrane (Neumann et al., 1983). It is very likely that similar mechanisms are involved in the uptake of pyrene by trout muscle cells. (c) Initial estimates of the extracellular and intracellular fluid spaces of white muscle were taken directly from Milligan and Wood (1982): they are 9.5 and 90.5% of total white muscle water content, respectively. These authors determined total
MODEL
IN TROUT
399
muscle water content by drying duplicate muscle samples at 85 “C to a constant weight. Muscle intracellular fluid volume was estimated as the difference between the total water content of muscle and the calculated extracellular fluid volume (Milligan and Wood, 1982). Since the total muscle water volume is determined by weight lost and the reported extracellular and intracellular spaces of muscle (Milligan and Wood, 1982) have not been confirmed by other studies, these estimates were adjusted after comparing the experimental data of pyrene-treated trout ( 10 mg/kg) with the computer simulation. The adjusted extracellular and intracellular fluid spaces of white muscle were 1.7 and 98.3%, respectively (Table 1). (d) The model assumes the excretion of pyrene by trout urine is a linear process involving filtration only. However, little is known about the renal clearance of pyrene in animals or fish. Our recent publication provides the only study on the time course of total 14C in the urine of rats after intravenous injection of [14C]pyrene (Withey et al., 199 1). However, the 14C excretion data do not allow the determination of renal clearance of pyrene by the rat. Therefore, renal clearance of pyrene by trout was estimated by trial and error. It is interesting to note that the renal clearance (Table 1) is very close to the GFR of trout. Several GFR of trout maintained at 10°C water temperature have been reported, they are: 0.117 ml/min/kg (Holmes and McBean, 1963), 0.073 ml/min/kg (Elger and Hentschel, 1983), and 0.086 ml/min/kg (Elger et al., 1986). The GFR reported by Holmes and McBean ( 1963) is identical to that of the renal clearance of pyrene by trout. Although the BBTM has been developed in trout following a single intraarterial injection of pyrene, it could be used in different animal species and dosing scenarios (Ramsey and Andersen, 1984). Indeed, we have successfully applied this model to predict the time course of pyrene concentrations in the blood of trout after the branchial route of exposure (Law et al., 1989). Therefore, it is possible that this model will be useful in predicting body bur-
400
LAW, ABEDINI,
dens of PAHs in trout exposed to PAH-contaminated water and/or sediments. APPENDIX Mass Balance D$erential and Algebraic Equations for the BBTM of Pyrene in Trout Gills VGL(dCGL/dt)
AND KENNEDY
rene in carcass; CK, concentration of pyrene in kidney tissue; CBA, concentration of pyrene in arterial blood; CBV, concentration of pyrene in venous blood, M, dose of pyrene; g(t), injection function; t, time; V,, , maximal velocity of saturable metabolism by liver; K,,,, Michaelis-Menten constant; KL, linear metabolism by liver; KK, renal clearance. See Table 1 for definition of other abbreviations.
= QT(CBV - CGL/RGL)
ACKNOWLEDGMENTS
Liver VL(dcL/df)
= CBA(QL - QGT) + QGT(CGT/RGT) - QL(CL/RL)
- KL(CL/RL)
- l’,,,,,(CL/RL)/(K,
+ CL/RL)
Muscle VEM(dCEM/dt)
This study was supported by an operating grant from the Natural Sciences and Engineering Research Council of Canada. We thank Dr. S. Kloster, Simon Fraser University Computer Services, for his assistance in using the CSMP for toxicokinetic model development and express our gratitude to Dr. A. P. Farrell, Department of BioSciences,Simon Fraser University, for his help in obtaining physiological information on trout.
= QM(CBA - CEM) - H(CEM - CIM/RM) VIM(dCIM/dt)
= H(CEM - CIM/RM)
CTM = [CIM(VIM)
+ (CEM)VEM]/(VIM
REFERENCES + VEM)
Gut VGT(dCGT/dt)
= QGT(CBA - CGT/RGT)
Carcass VC(dcC/dt)
= QC(CBA - CC/RC)
Kidney VK(dCK/dt)
= QK(CBA - CK/RK) - KK(CK/RK)
Arterial Blood VB(dCBA/dt)
= QT(CGL/RGL
- CBA) + Mg(t)
Venous Blood VB(&BV/df)
= QL(CL/RL)
+ QK(CK/RK)
+ (QM)(CEM)
+ QC(CC/RC) - (QT)(CBV)
Abbreviations used: CGL, concentration of pyrene in gill tissue; CL, concentration of pyrene in liver tissue; CTM, concentration of pyrene in muscle tissue; CEM, concentration of pyrene in extracellular space fluid of muscle; CIM, concentration of pyrene in intracellular space fluid of muscle; CGT, concentration of pyrene in gut tissue; CC, concentration of py-
ANDERSON, J. W., NEFF, J. M., Cox, B. A., TATEM, H. E., AND HIGHTOWER, G. M. (1974). The effects of oil on estuarine animals: Toxicity, uptake, depuration and respiration. In Pollution and Physiology of Marine Organisms (F. J. Vember& Ed.), pp. 285-3 10. Academic Press, New York. BARRON,M. G., TARR, B. D., AND HAYTON, W. L. ( 1987). Temperature-dependence of cardiac output and regional blood flow in rainbow trout, Salmo gairdneri Richardson. J. Fish Biol. 31, 735-744. BORNEFF,J. (1977). Fate of carcinogens in aquatic environments. In Fate of Pollutants in the Air and Water Environment, Part 2 (I. H. Suffet, Ed.), pp. 393-408. Wiley, New York. BUNGAY, P. M., DEDRICK, R. L., AND GUARINO. A. M. ( 1976). Pharmacokinetic modeling of the dogfish shark (Squalus acanthias): Distribution and urinary and biiary excretion of phenol red and its glucuronide. J. Pharmacokinet.
Bipoharm.
4, 377-388.
CLEWELL, H. J.. AND ANDERSEN, M. E. (1985). Risk assessment extrapolations and physiological modeling. Toxicol. Indust. Health 1, 111-134. DAVIS, J. C., AND CAMERON, J. N. (1971). Water flow and gas exchange at the gills of rainbow trout, Salmo gairdneri. J. Exp. Biol. 54, l- 18. DAXBOECK, C. ( I98 1). A study of the cardiovascular system of the rainbow trout (Salmo gairdneri) at rest and during swimming exercise. Ph.D. Thesis, University of British Columbia, Vancouver, British Columbia.
PYRENE TOXICOIUNETIC DEDRICK, R. L., AND BISCHOFF,K. B. (1968). Pharmacokinetics in applications of the artificial kidney. Chem. Eng. Prog. Symp. Ser. 64,32-43. DEDRICK, R. L., ZAHARKO, D. S., AND LUTZ, R. J. ( 1973). Transport and binding of methotrexate in vivo.J. Pharm. Sci. 62, 882-890. ELGER, B., AND HENTSCHEL, H. (1983). Effect of adrenergic blockade with bretylium and phentolamine on glomerular filtration rate in the rainbow trout, sulmo gairdneri rich, adapting to saline water. Comp. B&hem. Physiol. 75C, 253-258. ELGER, B., NEIJKIRCH, M., AND HENTSCHEL, H. (1986). The kidney of rainbow trout, Salmo gairdneri Richardson in the acute phase of viral haemorrhagic kpticaemiia; in vivo experiments on the renal excretion of fluid, electrolytes and protein. J. Fish Diseases 9,381-392. FRENCH, C. J., MOMMSEN, T. P., AND HOCHACHKA, P. W. ( 198 1). Amino acid utilization on isolated hepatocytes from rainbow trout. Eur. J. Biochem. 113, 311-317. GARGAS, M. L., CLEWELL, H. J., AND ANDERSEN,M. E. (1986). Metabolism of inhaled dihalomethanes in vivo: Differentiation of kinetic constants for two independent pathways. Toxicol. Appl. Pharmacol. 82,2 1 l-223. GINGERICH,W. H., PITYER, R. A., AND RACH, J. J. (1987). Estimates of plasma, packed cell and total blood volume in tissues of the rainbow trout (Salmo gairdnen). Comp. Biochem. Physiol. 87A(2), 25 1-256. GROVES,T. D. D. (1970). Body composition changes during growth in young sockeye (Oncorhynchus nerka) in fresh water. J. Fish. Res. Board Can. 27, 929-942. HASE, A.. AND HITES, R. A. (1978). On the origin ofpolycyclic aromatic hydrocarbons in the aqueous environment. In Identification and Anal.vsis oforganic Pollutants in Water (L. H. Keith, Ed.), Ann Arbor Science, Ann Arbor, MI. HOLMES, W. N., AND MCBEAN, R. L. (1963). Studies on the glomerular filtration rate of rainbow trout (Salmo gairdneri). J. Exp. Biol. 40, 335-341. HOLMES, W. N., AND STAINER, I. M. (1966). Studies on the renal excretion of electrolytes by the trout (Salmo gairdneri). J. Exp. Biol. 44, 33-46. International Business Machines Corporation (IBM) (1976). Continuous System Modeling Program III (CSMP III). Program Reference Manual 3rd ed. (Armond: Internal Business Machines). KENNEDY, C. J., AND LAW, F. C. P. (1990). Toxicokinetics of selected polycyclic aromatic hydrocarbons in rainbow trout following different routes of exposure. Environ. To,xico/. Chem. 9, 133-139. KICENIUK, J. W., AND JONES,D. R. (1977). The oxygen transport system in trout (Sulmo gairdneri) during sustained exercise. J. Exp. Biol. 69, 247-260. KRAHN, M. M.. BURROWS,D. G., MACLEOD, W. D.. JR., AND MALINS, D. C. (1987). Arch. Environ, Contam. Toxicol. 16, 5 1 l-522.
MODEL
IN TROUT
401
LAW, F. C. P., KENNEDY, C. J., AND ABEDINI, S. (1989). A physiologically-based toxicokinetic model (PBTM) for pyrene and fluorene in rainbow trout (Salmo gaiodnert]. Abstract Proceeding of the 10th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Toronto, Ontario, Canada. LEE, R. F., SAUERHEBER,R., AND DEBBS, G. H. (1972). Uptake, metabolism and discharge of polycyclic aromatic hydrocarbons by marine fish. Mar. Biol. 17,20 l208. LIN, J. H., SUGIYAMA. Y., AWAZU, S., AND HANANO, M. (1982). In vitro and in vivo evaluation of the tissue-toblood partition coefficient for physiological pharmacokinetic models. J. Pharmacokinet. Biopharm. 10,637647. MALINS, D. C., KRAHN, M. M., MYERS, M. S., RHODES, L. D., BROWN, D. W., KRONE, C. A., MCCAIN, B. B., AND CHAN, S.-L. (1985). Toxic chemicals in sediments and biota from a creosote-polluted harbar: Relationships with hepatic neoplasms and other hepatic lesions in English sole (Parophrys vetulus). Carcinogenesis 6, 14631469. MILLIGAN, C. L., AND WOOD, C. M. (1982). Disturbances in haematology, fluid volume distribution and circulatory function associated with low environmental pH in the rainbow trout, Salmo gairdneri. J. Exp. Biol. 99, 397-415. MOON, T. W., WALSH, P. J., AND MOMMSEN, T. P. (1985). Fish hepatocytes: A model metabolic system. Can. J. Fish. Aquat. Sci. 42, 1772-1782. National Research Council (NRC) (1987). Pharmacokinetics in Risk Assessment Drinking Water and Health, Vol. 8. National Academy Press, Washington, DC. NEFF, J. M., AND ANDERSON, J. W. ( 198 1). Response of Marine Animals to Petroleum and Spect$c Petroleum Hydrocarbons. Applied Science, London, England. NEUMANN, P., HOLETON, G. F.. AND HEISLER.N. (1983). Cardiac output and regional blood flow in gills and muscles after exhaustive exercise in rainbow trout (Salmo Gairdneri). J. Exp. Biol. 105, I-14. NICHOLS, J. W., MCKIM. J. M., LIEN, G. J., AND HOFFMAN, A. D. (1989). Physiologically based modeling of chemical uptake and disposition in fish: Studies with rainbow trout (Oncorhynchus mykiss) exposed to three chlorinated ethanes. Abstract Proceeding of the 10th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Toronto, Ontario. Canada. PARKER, R. S.. MORRISSEY, M. T., MOLDEUS, P., AND SELIVONCHICK, D. P. (1981). The use of isolated hepatocytes from rainbow trout (Sulmo guirdneri) in the metabolism of acetaminophen. Comp. Biochem. Physiol. 70B, 63 l-633. RAMSEY, J. C., AND ANDERSEN, M. E. (1984). A physiologically based description of the inhalation pharmacokinetics of styrene in rats and humans. To,xicol. .4ppl. Pharmacol. 73, 159-175.
402
LAW, ABEDINI,
RANDALL, D. J. (1970). The circulatory system. In Fish Physiology (W. S. Hoar and D. J. Randall, Eds.), Vol. 4, pp. 162. Academic Press, New York. SE~LEN, P. 0. (1973). Preparation of rat liver cells. III. Enzymatic requirements for tissue dispersion. Exp. Cell. Res. 82, 39 l-398. SMITH, L. S., AND BELL, G. R. (1964). A technique for prolonged blood sampling in free-swimming salmon. J. Fish. Res. Board Can. 21, 71 l-717. STEVENS, E. D. (1968). The effect of exercise on the distribution of blood to various organs in rainbow trout. Comp. Biochem. Physiol. 25, 6 15-625. TRAVIS, C. C. (1987). Interspecies and dose-route extrapolations. In Drinking Water and Health. Pharmacokineties in Risk Assessment, Vol. 8, pp. 208-220. National Academy Press. Washington, DC. U.S. Environmental Protection Agency (EPA) (1987). Update of the Health Assessment Document and Addendum for Dichloromethane (Methylene Chloride): Pharmacokinetics, Mechanisms of Action, and Epidemiology. EPA 600/&87/03OA.
AND KENNEDY VARANASI, U., AND MALINS, D. C. (1977). In Effects of Petroleum on Arctic and Subarctic and Organisms, Vol. 2, Biological
Marine Effects
Environments
(D. C. Matins, Ed.), pp. 175-270. Academic Press, New York. WALSH, P. J. (1986). Ionic requirements for intracellular pH regulation in rainbow trout hepatocytes. Am. J. Physiol. 250, R24-R29. WEINSTEIN, I. B., AND TROLL. W. (1977). National Cancer Institute workshop on tumor promotion and cofactors in carcinogenesis. Cancer Res. 37, 3461-3463. WITHEY, J. R., LAW, F. C. P., AND ENDRENYI, L. (1991). Pharmacokinetics and bioavailability of pyrene in the rat. J. Toxicol. Environ. Health. 32, 429-448. WHITTLE, B. G., MURRAY, J., MACKIE, P. R., HARDY, R., AND FARMER, J. (1977). Fate of hydrocarbons in fish. Rapp. P.-V. Reun. Cons. Int. Explor. Mer. 171, 139-142. ZAHARKO. D. S., DEDRICK, R. L., AND OLIVERIO, V. T. (1972). Prediction of the distribution of methotrexate in the sting rays Dasyatidae sabina and sayi by use of a model developed in mice. Camp. Biochem. Physiol. 42A, 183-194.
AND
APPLIED
A Biologically
PHARMACOLOGY
Based Toxicokinetic F. C. P.
Environmental
Toxicology
(1991)
110,390-402
LAW,
Model for Pyrene in Rainbow
S. ABEDINI,
AND C. J. KENNEDY
Program, Department qf Biological Burnaby, British Columbia, Canada
Received
November
Trout’
19. 1990; accepted
Sciences. Simon V5A IS6
Fraser
University,
June 11, 1991
Biologically Based Toxicokinetic Model for Pyrene in Rainbow Trout. LAW, F. C. P., ABEDINI, KENNEDY, C. J. (1991). Toxicol. Appl. Pharmacol. 110,390-402. A biologically based toxicokinetic model was developed to stimulate the metabolic disposition of pyrene in trout with an average body weight of 450 g and dosed with a single bolus injection of the chemical (10 mg/ kg). The model consists of a membrane-limited muscle compartment and six flow-limited compartments including the gills, liver, gut, kidney, carcass, and blood. The compartments are rep resented by mass balance equations including terms for the binding of pyrene to tissue and blood proteins, biotransformation, penetration rate into the muscle, blood flow rate, tissue mass, etc. The model also provides for nonsaturable and saturable clearances of pyrene by the liver and kidney. Michaelis-Menten constants for pyrene metabolism (K,, V,,,,,) were determined from in vitro experiments using isolated liver cells. Renal clearance of pyrene was very close to the glomerulus filtration rate of trout. Solution of the system of equations yielded the time courses of pyrene concentration in the tissues. Predicted concentrations of pyrene in the gills, liver, gut, kidney, muscle, and blood were consistent with experimental observations for at least 6 days. The model was validated by comparing the model predicted and experimental results of trout weighing 285 g and dosed with a single intraarterial dose (3 mg/kg) of pyrene. The predicted pyrene concentrations also were in adequate agreement with the empirical data. o 1991 Academic Press. hc. A
S.,
AND
The polycyclic aromatic hydrocarbons (PAHs) are pollutants released into the environment as a result of incomplete combustion of organic matter (Hase and Hites, 1978). Therefore, PAHs have been found in sewage, lake, and ground waters (Bomeff, 1977) and in marine sediments and food organisms (Malins et al., 1985). There have been many reported studies on the accumulation of PAHs by aquatic organisms (Varanasi and Malins, 1977; Neff and Anderson, 198 1). Waterborne PAHs and those associated with sediments are taken up rapidly by fish (Lee et al., 1972; Anderson et al., 1974); equilibrium concentrations of PAHs are often reached in fish tissue in 24 hr or less after ex’ Presented in part at the 10th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Toronto. Ontario. October 2%November 2, 1989. 0041-008X/91
$3.00
Copyright 0 199 1 by Academic Press. Inc. All rights of reproduction in any form reserved.
posure. In contrast, the uptake of PAHs from food and sediments are much slower. In most fish species, more than half of the orally administered PAHs are unabsorbed but associated with the digestive tract or its contents (Whittle et al., 1977). Pyrene is one of the major PAHs found in fish taken from polluted waters (Krahn et al., 1987). Although pyrene is not considered a potent carcinogen, it probably is a cocarcinogen (Weinstein and Troll, 1977). Pyrene is absorbed rapidly but eliminated slowly by trout; the uptake and elimination of pyrene by trout can be described adequately by a classical, three-compartment toxicokinetic model (Kennedy and Law, 1990). Biologically (physiologically) based toxicokinetic models (BBTM) have been used to reduce the uncertainties in the risk assessment process (NRC, 1987; U.S. EPA, 1987). BBTM 390
PYRENE TOXICOIUNETIC
also is a tool to increase understanding of interspecies differences in the biological fates of chemicals. However, very few BBTM, with the exception of the flow-limited models of methotrexate in sting rays (Dasyatidae sabina) (Zaharko et al., 1972) and phenol red in dogfish shark (Squalus acanthias) (Bungay et al., 1976), have been reported in the aquatic species. In the course of this work, Nichols et al. (1989) have reported a BBTM for three chlorinated ethanes in rainbow trout (Oncorhynthus mykiss). In view of the wide distribution of PAHs in the aquatic environment, it was decided that a BBTM for pyrene in trout should facilitate the use of fish in pharrnacokinetic studies of drugs, environmental pollutants, and other xenobiotics. The model also would permit a better evaluation of exposure and health risk assessment of humans who use PAH-contaminated fish as food. The work reported here describes the development of a BBTM of pyrene in trout. MATERIALS
AND
MODEL
391
IN TROUT
Feed obtained from Moore-Clark Co. (Vancouver, British Columbia). They were not fed for 3 days prior to an experiment. After a trout was anesthetized with MS 222, the dorsal aorta was cannulated according to the procedure of Smith and Bell (1964). The fish was allowed to recover from surgery for 24 hr in a darkened Plexiglas box (20 X 20 X 80 cm) supplied with continuously flowing water at 10°C.
Intraarterial
Administration
of Pyrene
Unlabeled pyrene was dissolved in Tween 80 before being mixed with 0.2 ml of modified Hank’s medium (Moon et al., 1985). The solution was injected as a bolus through the cannula into the dorsal aorta of trout. This was followed by the injection of 0.2 ml of fresh Hank’s medium through the cannula to ensure that the entire dose entered the circulatory system. Little or no pyrene was found to adsorb onto the cannula. Two different doses of pyrene were studied: trout with an average body weight (BW) of 285 g were given 3 mg/kg pyrene; trout with an average BW of 450 g were dosed with 10 mg/kg pyrene. At different times after pyrene administration, three trout were removed from the tank, euthenized, and frozen at -5°C until analysis. The intraarterial route was used to develop the BBTM instead of the oral or branchial route since the intraarterial route did not involve complicated absorption processes.
METHODS Unchanged
Pyrene
Concentration in Tissues
Chemicals
[4,5,9,10-‘4C]Pyrene (sp act 56 mCi/mmol) was purchased from Amersham Canada Ltd. (Oakville, Ontario). Unlabeled pyrene was obtained from Aldrich Chemical Co. (Milwaukee, WI). The 14C-labeled pyrene was purified by thin-layer chromatography (TLC) on silica-gel plates using bexanes as the developing solvent. The radiochemical purity of the “‘C-labeled pyrene, determined by TLC, exceeded 99%. The unlabeled pyrene was puriiied by repeated recrystallization from methanol. The purity of pyrene was determined by high-pressure liquid chromatography (HPLC) and exceeded 99%. Polyoxyethylene sorbitan monooleate (Tween 80) and ethyl-N-aminobenzate methane sulfonic acid (MS 222) were purchased from Sigma Chemical Co. (St. Louis, MO).
Fish
Rainbow trout (0. mykiss) weighing between 200-500 g were obtained from Spring Valley Trout Farms, British Columbia. The trout were kept in flowing, dechlorinated water at 10 +_ 2°C in holding tanks for at least 3 weeks before use. The fish were fed daily with New Age Fish
After the fish were thawed, we removed the liver, kidney, gut, gills, spleen, skin, white muscle, and blood from the carcass. Fat was trimmed off from the gut which included the esophagus, stomach, and large and small intestines. Gut contents were also discarded. The organs and tissues were rinsed with distilled water, blotted dry, and weighed. A 25% w/v tissue homogenate was prepared in distilled water. An aliquot of the homogenate (0.2-0.3 ml) was pipetted into a centrifuge tube and mixed with 0.5 ml 0. I N sulfuric acid and 0.5 ml distilled water. The mixture was vortexed and extracted three times with 3-ml portions of methylene chloride. The methylene chloride extracts were combined and passed through a glass column containing 5 g of florosil. The column was washed with 10 ml of fresh methylene chloride. The eluants were collected, combined, and evaporated down to dryness. The residues were redissolved in hexanes containing benzanthracene (1 pg/ml) as an internal standard. The hexane solution (20 ~1) was injected directly into a Hewlett-Packard liquid chromatograph (Model 1090, Avondale, PA) equipped with an ODS-Hypersil column (Hewlett-Packard Co., 5 pm, 100 X 4.6 mm i.d.) and a diode-array uv detector which had a setpoint at 230 nm. Column temperature was 40°C. Column eluant (methanol:water/73:27) had a flow
392
LAW, ABEDINI,
rate of 1.5 ml/min. The extraction method was found to be very efficient in removing pyrene from the various fish tissues; percentage recoveries + SD of pyrene from liver, kidney, gill, gut, white muscle, and blood were 78.9 + 9.3, 75.7 + 12.3, 71.4 f 6.6, 80.5 + 6.0, 88.4 + 8.1, and 85.3 + 6.8, respectively.
Urinary
Excretion
of ‘T-Labeled
Pyrene
In a separate experiment, three trout were anaesthetized with MS 222 and the dorsal aorta was cannulated as described above. A polyethylene cannula also was inserted into the urino-genital papilla for urine collection (Holmes and Stainer, 1966). After the trout were injected intraarterially with a single dose (10 mg/kg) of “‘C-labeled pyrene. mine was collected as 24-hr samples for 6 days. The volumes of urine samples were measured. The radioactivities in the urine were determined by liquid scintillation counting.
Pyrene
Metabolism
by Isolated
Hepatocytes
Liver cells were isolated from trout according to the procedures of French et al. (198 I), Moon et al. (1985), and Walsh (1986). Briefly, fish were anesthetized with a solution of 0.2 g/liter MS 222 and 0.2 g/liter NaHCO, until opercular movement ceased. A ventral incision was made in the abdomen of the fish and the hepatic portal vein cut. The intestinal vein was cannulated with a 23gauge stainless-steelneedle which was sutured in place with silk sutures. The liver was perfused with well-oxygenated Medium A (Moon et al., 1985) supplied by a peristalic pump for 10 min to clear the liver of blood. The flow rate was adjusted to approximately 2 ml/min/g liver to guarantee fully aerobic conditions. The liver was perfused for an additional 30-45 min with Medium B (Moon et al.. 1985) which contained collagenase and hyaluronidase. After the gall bladder was removed, the liver was chopped with a razor blade. The minced liver was filtered through a 253- pm and then a 73-pm nylon mesh. The cells in the filtrate were collected by centrifugation (50g) for 2 min in a Sorval RC-5B super speed centrifuge (DuPont, Wilmington, DE). The supematant was decanted and the cells pellet resuspended in ice-cold Medium C (Moon et al., 1985). Cell viability was tested by the trypan blue exclusion test. The viability ofthe isolated hepatocytes exceeded 95%. Incubation of pyrene with the hepatocytes was performed within 3 hr after cell preparation; an aliquot (1.9 ml) of the cell suspension (3 X lo6 cells/ml) was preincubated for I5 min in a liquid scintillation vial under an atmosphere of air containing 0.25% CO,. A solution of “C-labeled pyrene (0. I ml) was added to the liver cells (0.5, 2.5, 5.0, 25, or 50 /IM final concentration). The incubation mixture was gassed for the duration of the incubation. At the conclusion of the incubation, the cells
AND KENNEDY were isolated by centrifugation and homogenized with a Polytron homogenizer (Brinkman Co., Rexdale, Ontario). The homogenate was extracted by methylene chloride as described above. Radioactivities in the methylene chloride extract and the remaining aqueous layer were determined separately by liquid scintillation counting. The methylene chloride extract was also analyzed by HPLC for unchanged pyrene as described before. Since [‘%Z]pyrene was metabolized by the hepatocytes mainly to water-soluble metabelite(s), the radiolabel counts in the remaining aqueous layer represented the amount of pyrene metabolites formed. In vitro Michaelis-Menten constants of pyrene metabolism by the hepatocytes were determined from the Lineweaver-Burk plot of the incubation data.
Biological
Simulation
Model
The pyrene model (Fig. 1) contained 7 compartments: blood, gills, liver, gut, kidney, muscle and carcass. The gut compartment included the highly perfused organs such as the esophagus, stomach, intestine, pyloric caeca, spleen, ventricle, and red muscle. The carcass was comprised of remaining tissues such as skin, skull, branchial basket fins, and eyes.All tissues with the exception of the muscle were depicted as blood-flow-limited compartments. The muscle was membrane-limited since it met the criterion that Q/ H 1, where Q is organ blood flow and H is mass transfer coefficient (Dedrick and Bischoff, 1968). Elimination of
-
KIDNEY
‘QK
. KK I
*
MUSCLE
-:ARCASS -
QM
QC
1. Schematic diagram of the biologically based toxicokinetic model used to simulate the metabolic disposition of pyrene in trout. The symbols and parameters used to describe the model are included in Table 1 and the Ap pendix. FIG.
PYRENE TOXICOKINETIC pyrene was modeled by hepatic and renal clearances since pyrene was excreted by trout mainly as water-soluble metabolite(s) in the bile and urine (Kennedy and Law, 1990). Metabolism was assumed to occur only in the liver and was described by a combination of linear metabolism and Michaelis-Menten metabolism. Renal clearance was a firstorder rate process. It was obtained by trial and error and was found to be very close to the glomerulus filtration rate (GFR) of trout. Mass balance differential equations for the model and the definition for the algebraic terms are given in the Appendix.
Model
Parameters
Three types of parameters are needed to implement the BBTM; physiological parameters of trout, tissue:blood partition coefficient of pyrene, and the clearances of pyrene from trout. The data, which were obtained by experiment or from the literature are summarized in Table 1. Physiological parameters. The liver, gut, muscle, gills, kidney, and blood were assigned volumes equal to 1.16, 8.52,46.5, 3.90,0.8, and 4.11% of body weight. The sum of these tissue volumes accounted for 65.0% of total BW, the remaining 35.0% was the carcass. Extracellular and intracellular fluid spaces of muscle were 0.78 and 45.72% of total BW, respectively.Cardiac output (CO) was assigned a value of 19.73 ml/min/kg. It was assumed that the gills received the entire CO. Blood flow distribution to the liver, gut, muscle, and kidney was, respectively, 2.75, 15.39, 39.77, and 10.22% of CO. Clearances. Total body clearance was obtained by dividing the administered dose with AUC, the total area under the blood pyrene concentration-time curve. AUC was shown to be 3888 &min/ml (Kennedy and Law, 1990). Rate constants of Michaelis-Menten metabolism by the liver, I’,,,, and K, , were derived from the isolated hepatocyte experiment (see above). Renal clearance (KK) was assigned a value of 169 ml/kg/day. First-order metabolic clearance (KL) of pyrene by a 450-g trout was estimated by the difference between total body clearance and the sum of renal and saturable hepatic clearances. The V,,,,,. KK. and KL values for the 285-g trout were scaled from the following allometric equations (Travis, 1987): V,,,,, = I.‘,,,,,, X KK = KK, X KL = KL
BW’O.” BW’O.”
c X BW’-O.“,
(1)
(2) (3)
where Vmaxc,Kk;, and KL, are the scaling coefficients for the maximum velocity of Michaelis-Menten metabolism, renal clearance, and first-order metabolic clearance, respectively. They were determined from trout with an average BW of 450 g. Gargas et al. (1986) have suggested that the metabolic rate VW, can be scaled using body
MODEL
IN TROUT
393
TABLE 1 KINETIC CONSTANTSAND MODEL PARAMETERS USED IN THE BIOLOGICALLY BASED TOXICOKINETIC MODEL FOR PYRENE
Trout Parameter
Abbreviations”
Vohnne (% body wt) Gills VGL Liver VL Gut VGT Kidney VK W. Muscle Extracellular VEM Intracellular VIM Carcass vc Blood VB Blood Flow (% cardiacoutput) Gills Liver QL-QGT Gut Kidney g W. Muscle QM Carcass QC Cardiac output (ml/min) QT Partition coefficients Gills/blood RGL Liver/blood RL Gut/blood RGT Kidney/blood RK W. Muscle/blood RM Carcass/blood RC Mass transfercoefficient H Muscle Biochemical constants (&mitt/g liver) ~nlax (a/ml) Kn (ml/mitt) KL (ml/min) Iw Dosem/kg)
450 g
285 g
3.90 1.16 8.52 0.80 46.50 0.78 45.72 35.0 4.11
3.90 1.16 8.52 0.80 46.50 0.78 45.72 35.0 4.11
100 2.15 15.39 10.23 39.77 31.86
100 2.75 15.39 10.23 39.77 31.86
8.875
5.622
2.15 37.0 1.90 15.3 0.20 7.75
2.75 37.0 1.90 15.3 0.20 7.75
0.035
0.035
0.255 3.05 0.40 0.06 10
0.185 3.05 0.56 0.04 3
a See Appendix for definitions of abbreviations for biochemical constants.
weight. Ramsey and Andersen (1984) have proposed that I’,,,, is proportional to the surface area. Tissue:bloodpartition coeficients. Two different methods were used to estimate tissue:blood partition coefficients: (a) In vitro tissue binding studies. These studies were carried out in dialysis cells which consisted of a dialysis bag (prepared from Spectra/par molecularporous membrane tubing, 18 mm X 50 ft, Spectrum Medical Industries, Inc. Los Angeles, CA) and a glass measuring cylinder. About 0.5 ml ofthe tissue homogenate (25% w/v) or whole blood (25% v/v) prepared in Tris-HCl buffer (0.1 M, pH 7.4) was pipetted into the dialysis bag and dialyzed against the
394
LAW, ABEDINI,
same buffer for I day at 4°C before use. The tissue homogenate or blood was then dialyzed against three different initial concentrations (0.1, 0.25, and 0.5 mM) of [%]pyrene solution in Tris-HCl buffer (0.1 M, pH 7.4) at 4°C for an additional 2 days. Preliminary studies using different dialysis durations (1, 2, and 3 days) suggested that equilibrium was reached by most fish tissues in 2 days. At the conclusion of the dialysis, radioactivities in the tissue and buffer were determined separately by liquid scintillation counting. The sum of buffer and tissue radioactivity was found to account for ~95% of the radioactivity initially added to the dialysis cells. Little or no pyrene was metabolized during dialysis since nearly all the radioactivities in the buffer and tissues were extractable by methylene chloride. The unbound fraction of [‘%Z]pyrene in each tissue was determined by dividing the chemical concentration in buffer by the concentration in the tissue homogenate as described by Lin et al. (1982). (b) In vivo experimental data. Initial estimates of tissue:blood partition coefficients also were determined from in vivo experimental data according to the method of Dedrick et al. (1973). Briefly. trout were injected intraarterially with a solution of unlabeled pyrene (10 mg/kg) as described before. At different times after pyrene administration (8. 16. 24, and 48 hr), three trout were euthanized and their organs removed. Unchanged pyrene concentrations in the tissues were determined by HPLC analyses of the methylene chloride extracts of tissue homogenates. A Freundlich-type plot (log-log) of tissue concentration vetsus blood concentration was prepared for each tissue. The tissue:blood partition coefficients were determined from the best straight line with a unit slope drawn through the data. These estimates were adjusted after comparing the experimental data of pyrene-treated trout (10 mg/kg) with the computer simulation. Mass transfer coejicient of muscle. The mass transfer coefficient of muscle (H) was estimated from the in vivo experimental data of muscle and blood. A detailed description of the procedure and the rationale of the methodology has been described by Dedrick and Bischoff ( 1968). Briefly, an initial estimate of H was determined from the slope of a straight line by plotting [VM(ACM/At)] versus [CBV-CM/RM)]. The estimate was adjusted after comparing the experimental data of muscle from pyrene-treated trout (10 mg/kg) with the computer simulation.
Computer
Simulation
The differential and algebraic equations describing the movement of pyrene through the trout were formulated as a computer program (see Appendix). The set of differential and algebraic equations was solved numerically with the aid of a Fortran-based software package, CSMP (IBM, 1976). The resultant solutions gave the predicted pyrene concentrations in the tissues and percentage dose excreted in the urine.
AND KENNEDY Determination
of Radioactivity
Radioactivity was determined by a Beckman LS-8000 liquid scintillation counter. A correction of quenching was made using the external standard technique. The radioactivity in the urine sample was determined directly by liquid scintillation counting after the addition of Biofluor. Tissue homogenates were digested with a 1-ml solution of Protosol and ethanol (1:2) at 60°C for 0.5 hr, decolorized with 30% hydrogen peroxide, and counted in a liquid scintillation counter after the addition of Bioflour.
RESULTS
AND DISCUSSION
We have presented a BBTM (Fig. 1) for predicting the concentrations of pyrene in seven tissue compartments of trout following a single intraarterial dose. The model was developed and validated with trout weighing 450 and 280 g, respectively. The model simulates most experimental data closely for the two administered doses (10 and 3 mg/kg). Predicted and observed pyrene concentration in the liver, gut, muscle, kidney, gills, and blood are shown in Figs. 2a-2f. The curves are the simulations based on the model illustrated in Fig. 1. The points represent the experimental data. We have selected 10 mg/kg as the upper dose limit of the study since pyrene solubility is limited by the volume of the injected solution. The intent of the dose-dose comparison is to delineate the rate-limiting or saturable process involved in the metabolic disposition of pyrene in trout. However, a threefold difference in the dose apparently is insufficient to detect if there are pyrene-related effects on the clearance process. The concentration profiles between the gut and gill compartments (Figs. 2b and 2c) are exceedingly similar; therefore, they could be modeled as a single compartment. Although the adipose tissue of the pyrene-treated trout contains a high level of the chemical (data not shown), fat is not modeled as a separate compartment since the volume of fat in trout is about 1% of the body weight (Gingerich et al., 1987). Moreover, the amounts of fat found in trout vary with individual fish. Groves (1970)
PYRENE TOXICOKINETIC
MODEL
395
IN TROUT
4 160
ak 0.01
4
c--v++ 0
20
40
60
100
120
140
160
FIG.2. Time courses of pyrene concentrations in trout tissues. Each point represents the experimental data from the tissue of one trout following an intraarterial injection of pyrene (0 10 mg/kg; A 3 mg/kg). Solid lines represent simulation using the model. a. liver; b, gill; c, gut; d, kidney: e, muscle; f, blood.
has suggested that fat is more a function of the nutritional history of the fish. After adjustment with the BW, the physiological and biochemical parameters used in modeling are exactly the same for both doses (Table 1). The physiological parameters also are consistent with the reported values of trout (Stevens, 1968; Daxboeck, 1981; Barron et al., 1987; and Gingerich et al., 1987). For example, the assigned white muscle volume was 46.5% BW (Table 1). This agrees with the reported value of Gingerich et al. (1987), al-
though it is less than the 66% BW assumed by Stevens (1968). The total blood volume (4.11% BW) is identical to that of Gingerich et al. (1987) but is slightly less than the 5.58 and 6.08% reported by Mill&m and Wood (1982) and Daxboeck (198 I), respectively. The liver volume is 1.16% BW (Table 1). This is very close to the liver volumes reported by Daxboeck (198 l), Gingerich et al. (1987), and Barron et al. (1987). The kidney volume is 0.8% BW; it is consistent with those reported by Barron et al. (1987) and Gingerich et al.
396
LAW,
ABEDINI,
AND
KENNEDY
4 160
,“oo
1 u 0
20
40
60
FIG.
Z-Continued
, I 100
L
\
120
140
4 160
PYRENE
TOXICOKINETIC
RG.
MODEL
IN
TROUT
397
2-Continued
(1987). The gill volume (3.9% BW) was taken from Stevens (1968). It is slightly higher than the 2.0 and 2.78% reported by Barron et al. (1987) and Gingerich et al. (198 l), respectively. The CO is 19.73 ml/min/kg. This is consistent with the CO reported for trout at a similar water temperature: 18.5 ml/min/kg (Davis and Cameron, 197 1) and 17.6 ml/min/ kg (Kiceniuk and Jones, 1977). Barron et al. (1987) have shown that the CO of trout declines linearly with decreasing acclimation temperature. If the regression relationship proposed by Barron et al. (1987) were used to calculate the CO, a much higher value (26.6 ml/min/kg) would be obtained, resulting in a lower model-predicted pyrene concentration in the blood compartment. The blood flow distribution to the various compartments of the BBTM also are consistent with the literature values (Daxboeck, 1981; Barron et al., 1987). For example, Barron et al. (1987) have shown that the liver, kidney, and white muscle of trout receive about 1.95, 12.8, and 38.9% of CO, respectively. In contrast, Daxboeck (198 1) has reported that the liver, kidney, and white muscle receive about 4.5-8.73, 4.5-9.26, and 36.449.0% of co. The model-predicted cumulative urinary excretion of 14C also agrees quite well with the
experimental data over the 6-day observation period (Fig. 3). The predicted cumulative urinary excretion at Day 6 also is consistent with the experimental value reported previously by Kennedy and Law (1990). The kinetic constants of pyrene metabolism (K,, I’,,,,,) (Table 1) were calculated from the Michaelis-Menten constants of two separate isolated hepatocyte experiments with different trout. Figure 4 shows that [14C]pyrene was metabolized by the hepatocytes linearly for at least 30 min. The V,,, values obtained from the two studies were 0.00 13 and 0.0 113 nmol/ min/ 1O6cells; the corresponding K,,, were 13.3 and 16.8 pM, respectively (Fig. 5). Previous studies indicated that each gram of liver yielded about 0.69-3.28 X 10’ cells (Parker et al., 1981) or 1.3 X lo8 cells (Selglen, 1973). To convert the Michaelis-Menten constants obtained from the hepatocyte studies to that of the intact liver, we assume each gram of trout liver yields about 2.0 X 10’ cells (average of 0.69 X 10’ and 3.28 X 10’). The calculated Vmaxand Km values of pyrene metabolism by the intact liver of a 450-g trout are 0.255 r.Lg/ min/g liver and 3.05 pg/ml, respectively (Table 1). The ultimate goal of the BBTM is to predict the biological fates of pyrene without resorting to prior in vivo experiments. However, several
398
LAW, ABEDINI,
AND KENNEDY
FIG. 3. Cumulative amount of [‘4C]pyrene equivalent collected in the urine of trout. The points are experimental data for a single intraarterial injection of 10 mg/kg into the dorsal aorta of three different trout. Solid lines represent model simulation of ‘%Zaccumulation in the urine.
of the parameters used in the model cannot be determined a priori with a great deal of confidence. These include the tissue:blood partition coefficients, mass transfer coefficient of muscle, extracellular and intracellular spaces of muscle, and renal clearance of pyrene by trout: (a) The tissue:blood partition coefficients used in the model were determined from the in vivo experimental data (Table 1).
5 a 0
20
10
rime
30
These differ greatly from those determined by the binding studies (1.94,2.56, 1.26,2.49, and 1.34 for the gills, liver, gut, kidney, and white muscle, respectively). A possible explanation for the discrepancy of results may be that the conditions for the binding experiment still are not optimal for equilibrium to be achieved. The appropriateness of using the in vivo tissue:
I 40
(min)
FIG. 4. Time course of “%-labeled pyrene metabolism by the isolated hepatocytes of trout. Displayed are the typical results of one of two experiments.
FIG. 5. A Lineweaver-Burk plot of “‘C-labeled pyrene metabolism by the isolated hepatocytes. Displayed are the typical results of one of two experiments.
PYRENE TOXICOKINETIC
blood partition coefficients in model simulations is further supported by the value of the apparent volume of distribution (I’,). The V, of trout calculated by summing the R X V products of Table 1 is 3.62 liters/kg. This is very close to the Vd (2.57 liters/kg) determined by the equation, dose/(AUC X X,), where X, (0.001 min-1) and AUC (3888 pg*min/ml) (Kennedy and Law, 1990) are, respectively, the slope of the terminal exponential phase and the area under the curve of the triexponential blood concentration-time curve of trout following an intraarterial injection of pyrene. (b) The muscle is a membrane-limited compartment in the model. Initially, the muscle compartment was modeled as a flow-limited compartment. However, a flow-limited muscle compartment overpredicted pyrene concentrations at early times. Moreover, the predicted peak concentration of pyrene occurred at a time much earlier than that of the experimental data. If the muscle flow rate were lowered arbitrarily by lo%, the simulations would agree more closely with the experimental data. This suggests that the transfer of pyrene from blood into muscle is slower than would be predicted on the basis of published blood flow. The slow uptake of pyrene by the muscle cells probably is due to low cell membrane permeabilities or maldistribution (shunt) of blood flow. Further studies will be required to clarify the mechanism(s) of pyrene uptake by the muscle cells of trout. Randall (1970) has shown that white muscle, constituting the bulk of trout body, is poorly vascularized and is used only in burst activity of short duration. The slow release of lactate from trout muscle has been explained by poor vascularization (Randall, 1970) and the diffusion resistance of the muscle cell membrane (Neumann et al., 1983). It is very likely that similar mechanisms are involved in the uptake of pyrene by trout muscle cells. (c) Initial estimates of the extracellular and intracellular fluid spaces of white muscle were taken directly from Milligan and Wood (1982): they are 9.5 and 90.5% of total white muscle water content, respectively. These authors determined total
MODEL
IN TROUT
399
muscle water content by drying duplicate muscle samples at 85 “C to a constant weight. Muscle intracellular fluid volume was estimated as the difference between the total water content of muscle and the calculated extracellular fluid volume (Milligan and Wood, 1982). Since the total muscle water volume is determined by weight lost and the reported extracellular and intracellular spaces of muscle (Milligan and Wood, 1982) have not been confirmed by other studies, these estimates were adjusted after comparing the experimental data of pyrene-treated trout ( 10 mg/kg) with the computer simulation. The adjusted extracellular and intracellular fluid spaces of white muscle were 1.7 and 98.3%, respectively (Table 1). (d) The model assumes the excretion of pyrene by trout urine is a linear process involving filtration only. However, little is known about the renal clearance of pyrene in animals or fish. Our recent publication provides the only study on the time course of total 14C in the urine of rats after intravenous injection of [14C]pyrene (Withey et al., 199 1). However, the 14C excretion data do not allow the determination of renal clearance of pyrene by the rat. Therefore, renal clearance of pyrene by trout was estimated by trial and error. It is interesting to note that the renal clearance (Table 1) is very close to the GFR of trout. Several GFR of trout maintained at 10°C water temperature have been reported, they are: 0.117 ml/min/kg (Holmes and McBean, 1963), 0.073 ml/min/kg (Elger and Hentschel, 1983), and 0.086 ml/min/kg (Elger et al., 1986). The GFR reported by Holmes and McBean ( 1963) is identical to that of the renal clearance of pyrene by trout. Although the BBTM has been developed in trout following a single intraarterial injection of pyrene, it could be used in different animal species and dosing scenarios (Ramsey and Andersen, 1984). Indeed, we have successfully applied this model to predict the time course of pyrene concentrations in the blood of trout after the branchial route of exposure (Law et al., 1989). Therefore, it is possible that this model will be useful in predicting body bur-
400
LAW, ABEDINI,
dens of PAHs in trout exposed to PAH-contaminated water and/or sediments. APPENDIX Mass Balance D$erential and Algebraic Equations for the BBTM of Pyrene in Trout Gills VGL(dCGL/dt)
AND KENNEDY
rene in carcass; CK, concentration of pyrene in kidney tissue; CBA, concentration of pyrene in arterial blood; CBV, concentration of pyrene in venous blood, M, dose of pyrene; g(t), injection function; t, time; V,, , maximal velocity of saturable metabolism by liver; K,,,, Michaelis-Menten constant; KL, linear metabolism by liver; KK, renal clearance. See Table 1 for definition of other abbreviations.
= QT(CBV - CGL/RGL)
ACKNOWLEDGMENTS
Liver VL(dcL/df)
= CBA(QL - QGT) + QGT(CGT/RGT) - QL(CL/RL)
- KL(CL/RL)
- l’,,,,,(CL/RL)/(K,
+ CL/RL)
Muscle VEM(dCEM/dt)
This study was supported by an operating grant from the Natural Sciences and Engineering Research Council of Canada. We thank Dr. S. Kloster, Simon Fraser University Computer Services, for his assistance in using the CSMP for toxicokinetic model development and express our gratitude to Dr. A. P. Farrell, Department of BioSciences,Simon Fraser University, for his help in obtaining physiological information on trout.
= QM(CBA - CEM) - H(CEM - CIM/RM) VIM(dCIM/dt)
= H(CEM - CIM/RM)
CTM = [CIM(VIM)
+ (CEM)VEM]/(VIM
REFERENCES + VEM)
Gut VGT(dCGT/dt)
= QGT(CBA - CGT/RGT)
Carcass VC(dcC/dt)
= QC(CBA - CC/RC)
Kidney VK(dCK/dt)
= QK(CBA - CK/RK) - KK(CK/RK)
Arterial Blood VB(dCBA/dt)
= QT(CGL/RGL
- CBA) + Mg(t)
Venous Blood VB(&BV/df)
= QL(CL/RL)
+ QK(CK/RK)
+ (QM)(CEM)
+ QC(CC/RC) - (QT)(CBV)
Abbreviations used: CGL, concentration of pyrene in gill tissue; CL, concentration of pyrene in liver tissue; CTM, concentration of pyrene in muscle tissue; CEM, concentration of pyrene in extracellular space fluid of muscle; CIM, concentration of pyrene in intracellular space fluid of muscle; CGT, concentration of pyrene in gut tissue; CC, concentration of py-
ANDERSON, J. W., NEFF, J. M., Cox, B. A., TATEM, H. E., AND HIGHTOWER, G. M. (1974). The effects of oil on estuarine animals: Toxicity, uptake, depuration and respiration. In Pollution and Physiology of Marine Organisms (F. J. Vember& Ed.), pp. 285-3 10. Academic Press, New York. BARRON,M. G., TARR, B. D., AND HAYTON, W. L. ( 1987). Temperature-dependence of cardiac output and regional blood flow in rainbow trout, Salmo gairdneri Richardson. J. Fish Biol. 31, 735-744. BORNEFF,J. (1977). Fate of carcinogens in aquatic environments. In Fate of Pollutants in the Air and Water Environment, Part 2 (I. H. Suffet, Ed.), pp. 393-408. Wiley, New York. BUNGAY, P. M., DEDRICK, R. L., AND GUARINO. A. M. ( 1976). Pharmacokinetic modeling of the dogfish shark (Squalus acanthias): Distribution and urinary and biiary excretion of phenol red and its glucuronide. J. Pharmacokinet.
Bipoharm.
4, 377-388.
CLEWELL, H. J.. AND ANDERSEN, M. E. (1985). Risk assessment extrapolations and physiological modeling. Toxicol. Indust. Health 1, 111-134. DAVIS, J. C., AND CAMERON, J. N. (1971). Water flow and gas exchange at the gills of rainbow trout, Salmo gairdneri. J. Exp. Biol. 54, l- 18. DAXBOECK, C. ( I98 1). A study of the cardiovascular system of the rainbow trout (Salmo gairdneri) at rest and during swimming exercise. Ph.D. Thesis, University of British Columbia, Vancouver, British Columbia.
PYRENE TOXICOIUNETIC DEDRICK, R. L., AND BISCHOFF,K. B. (1968). Pharmacokinetics in applications of the artificial kidney. Chem. Eng. Prog. Symp. Ser. 64,32-43. DEDRICK, R. L., ZAHARKO, D. S., AND LUTZ, R. J. ( 1973). Transport and binding of methotrexate in vivo.J. Pharm. Sci. 62, 882-890. ELGER, B., AND HENTSCHEL, H. (1983). Effect of adrenergic blockade with bretylium and phentolamine on glomerular filtration rate in the rainbow trout, sulmo gairdneri rich, adapting to saline water. Comp. B&hem. Physiol. 75C, 253-258. ELGER, B., NEIJKIRCH, M., AND HENTSCHEL, H. (1986). The kidney of rainbow trout, Salmo gairdneri Richardson in the acute phase of viral haemorrhagic kpticaemiia; in vivo experiments on the renal excretion of fluid, electrolytes and protein. J. Fish Diseases 9,381-392. FRENCH, C. J., MOMMSEN, T. P., AND HOCHACHKA, P. W. ( 198 1). Amino acid utilization on isolated hepatocytes from rainbow trout. Eur. J. Biochem. 113, 311-317. GARGAS, M. L., CLEWELL, H. J., AND ANDERSEN,M. E. (1986). Metabolism of inhaled dihalomethanes in vivo: Differentiation of kinetic constants for two independent pathways. Toxicol. Appl. Pharmacol. 82,2 1 l-223. GINGERICH,W. H., PITYER, R. A., AND RACH, J. J. (1987). Estimates of plasma, packed cell and total blood volume in tissues of the rainbow trout (Salmo gairdnen). Comp. Biochem. Physiol. 87A(2), 25 1-256. GROVES,T. D. D. (1970). Body composition changes during growth in young sockeye (Oncorhynchus nerka) in fresh water. J. Fish. Res. Board Can. 27, 929-942. HASE, A.. AND HITES, R. A. (1978). On the origin ofpolycyclic aromatic hydrocarbons in the aqueous environment. In Identification and Anal.vsis oforganic Pollutants in Water (L. H. Keith, Ed.), Ann Arbor Science, Ann Arbor, MI. HOLMES, W. N., AND MCBEAN, R. L. (1963). Studies on the glomerular filtration rate of rainbow trout (Salmo gairdneri). J. Exp. Biol. 40, 335-341. HOLMES, W. N., AND STAINER, I. M. (1966). Studies on the renal excretion of electrolytes by the trout (Salmo gairdneri). J. Exp. Biol. 44, 33-46. International Business Machines Corporation (IBM) (1976). Continuous System Modeling Program III (CSMP III). Program Reference Manual 3rd ed. (Armond: Internal Business Machines). KENNEDY, C. J., AND LAW, F. C. P. (1990). Toxicokinetics of selected polycyclic aromatic hydrocarbons in rainbow trout following different routes of exposure. Environ. To,xico/. Chem. 9, 133-139. KICENIUK, J. W., AND JONES,D. R. (1977). The oxygen transport system in trout (Sulmo gairdneri) during sustained exercise. J. Exp. Biol. 69, 247-260. KRAHN, M. M.. BURROWS,D. G., MACLEOD, W. D.. JR., AND MALINS, D. C. (1987). Arch. Environ, Contam. Toxicol. 16, 5 1 l-522.
MODEL
IN TROUT
401
LAW, F. C. P., KENNEDY, C. J., AND ABEDINI, S. (1989). A physiologically-based toxicokinetic model (PBTM) for pyrene and fluorene in rainbow trout (Salmo gaiodnert]. Abstract Proceeding of the 10th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Toronto, Ontario, Canada. LEE, R. F., SAUERHEBER,R., AND DEBBS, G. H. (1972). Uptake, metabolism and discharge of polycyclic aromatic hydrocarbons by marine fish. Mar. Biol. 17,20 l208. LIN, J. H., SUGIYAMA. Y., AWAZU, S., AND HANANO, M. (1982). In vitro and in vivo evaluation of the tissue-toblood partition coefficient for physiological pharmacokinetic models. J. Pharmacokinet. Biopharm. 10,637647. MALINS, D. C., KRAHN, M. M., MYERS, M. S., RHODES, L. D., BROWN, D. W., KRONE, C. A., MCCAIN, B. B., AND CHAN, S.-L. (1985). Toxic chemicals in sediments and biota from a creosote-polluted harbar: Relationships with hepatic neoplasms and other hepatic lesions in English sole (Parophrys vetulus). Carcinogenesis 6, 14631469. MILLIGAN, C. L., AND WOOD, C. M. (1982). Disturbances in haematology, fluid volume distribution and circulatory function associated with low environmental pH in the rainbow trout, Salmo gairdneri. J. Exp. Biol. 99, 397-415. MOON, T. W., WALSH, P. J., AND MOMMSEN, T. P. (1985). Fish hepatocytes: A model metabolic system. Can. J. Fish. Aquat. Sci. 42, 1772-1782. National Research Council (NRC) (1987). Pharmacokinetics in Risk Assessment Drinking Water and Health, Vol. 8. National Academy Press, Washington, DC. NEFF, J. M., AND ANDERSON, J. W. ( 198 1). Response of Marine Animals to Petroleum and Spect$c Petroleum Hydrocarbons. Applied Science, London, England. NEUMANN, P., HOLETON, G. F.. AND HEISLER.N. (1983). Cardiac output and regional blood flow in gills and muscles after exhaustive exercise in rainbow trout (Salmo Gairdneri). J. Exp. Biol. 105, I-14. NICHOLS, J. W., MCKIM. J. M., LIEN, G. J., AND HOFFMAN, A. D. (1989). Physiologically based modeling of chemical uptake and disposition in fish: Studies with rainbow trout (Oncorhynchus mykiss) exposed to three chlorinated ethanes. Abstract Proceeding of the 10th Annual Meeting of the Society of Environmental Toxicology and Chemistry, Toronto, Ontario. Canada. PARKER, R. S.. MORRISSEY, M. T., MOLDEUS, P., AND SELIVONCHICK, D. P. (1981). The use of isolated hepatocytes from rainbow trout (Sulmo guirdneri) in the metabolism of acetaminophen. Comp. Biochem. Physiol. 70B, 63 l-633. RAMSEY, J. C., AND ANDERSEN, M. E. (1984). A physiologically based description of the inhalation pharmacokinetics of styrene in rats and humans. To,xicol. .4ppl. Pharmacol. 73, 159-175.
402
LAW, ABEDINI,
RANDALL, D. J. (1970). The circulatory system. In Fish Physiology (W. S. Hoar and D. J. Randall, Eds.), Vol. 4, pp. 162. Academic Press, New York. SE~LEN, P. 0. (1973). Preparation of rat liver cells. III. Enzymatic requirements for tissue dispersion. Exp. Cell. Res. 82, 39 l-398. SMITH, L. S., AND BELL, G. R. (1964). A technique for prolonged blood sampling in free-swimming salmon. J. Fish. Res. Board Can. 21, 71 l-717. STEVENS, E. D. (1968). The effect of exercise on the distribution of blood to various organs in rainbow trout. Comp. Biochem. Physiol. 25, 6 15-625. TRAVIS, C. C. (1987). Interspecies and dose-route extrapolations. In Drinking Water and Health. Pharmacokineties in Risk Assessment, Vol. 8, pp. 208-220. National Academy Press. Washington, DC. U.S. Environmental Protection Agency (EPA) (1987). Update of the Health Assessment Document and Addendum for Dichloromethane (Methylene Chloride): Pharmacokinetics, Mechanisms of Action, and Epidemiology. EPA 600/&87/03OA.
AND KENNEDY VARANASI, U., AND MALINS, D. C. (1977). In Effects of Petroleum on Arctic and Subarctic and Organisms, Vol. 2, Biological
Marine Effects
Environments
(D. C. Matins, Ed.), pp. 175-270. Academic Press, New York. WALSH, P. J. (1986). Ionic requirements for intracellular pH regulation in rainbow trout hepatocytes. Am. J. Physiol. 250, R24-R29. WEINSTEIN, I. B., AND TROLL. W. (1977). National Cancer Institute workshop on tumor promotion and cofactors in carcinogenesis. Cancer Res. 37, 3461-3463. WITHEY, J. R., LAW, F. C. P., AND ENDRENYI, L. (1991). Pharmacokinetics and bioavailability of pyrene in the rat. J. Toxicol. Environ. Health. 32, 429-448. WHITTLE, B. G., MURRAY, J., MACKIE, P. R., HARDY, R., AND FARMER, J. (1977). Fate of hydrocarbons in fish. Rapp. P.-V. Reun. Cons. Int. Explor. Mer. 171, 139-142. ZAHARKO. D. S., DEDRICK, R. L., AND OLIVERIO, V. T. (1972). Prediction of the distribution of methotrexate in the sting rays Dasyatidae sabina and sayi by use of a model developed in mice. Camp. Biochem. Physiol. 42A, 183-194.