Page 135 - The Toxicology of Fishes

Toxicokinetics in Fishes 115

methods is often used to characterize uptake efficiency for these compounds. The diet also represents
an important route of uptake for lipophilic environmental contaminants (log K > 5) (Bruggeman et
ow
al., 1984). This is due in part to the fact that these compounds accumulate to high levels in prey items.
High log K compounds also tend to bind to dissolved and particulate organic material in water, reducing
ow
their availability for branchial uptake. Similar considerations may also apply to several metalloids and
organometallic compounds, including selenium, cesium, methylmercury, and tributyltin.
Considerable effort has been expended to identify factors that control dietary uptake of hydrophobic
compounds, and recent modeling work reflects this emphasis. Within the intestine, uptake of dietary
lipid and a reduction of meal volume increase chemical activity in the gut contents above that of the
meal, potentially resulting in chemical uptake by fish even when chemical concentrations in their prey
are at or near equilibrium with those in water (see Mechanism of Biomagnification section below). It is
important, however, to note that even low rates of metabolism could reduce the extent of chemical
accumulation substantially (Clark et al., 1990; de Wolf et al., 1992; Endicott and Cook, 1994; Nichols
et al., 2004b). Chemicals in the diet are particularly susceptible to metabolic clearance because of the
potential for presystemic metabolism in tissues of both the gut and liver (Van Veld, 1990).
To date, very few PBTK models incorporating a dietary uptake description have been developed for
mammals, and only two have been published for fish. Most of the mammalian models have treated
dietary uptake as a first-order absorptive process. If absorption is assumed to be 100% efficient, the dose
can be modeled as a mass of compound and it is not necessary to define a gut compartment. A second
approach also employs a first-order absorptive rate constant but provides for the possibility that absorption
is less than 100% efficient. In this second approach, a gut lumen compartment is defined and the
consumption of food and egestion of feces are modeled as bulk flow rates or periodic events. In both
of these approaches, the value of the first-order rate constant is generally obtained by fitting model
simulations to measured (usually blood or plasma) chemical concentrations.
At the other end of the spectrum in terms of complexity, Bungay et al. (1981) developed a highly
detailed model for dietary uptake of chlordecone in the rat. The gut portion of this model consists of
six compartments corresponding to the stomach, small intestine (three compartments), cecum, and lower
intestine. Diffusion limitations on chemical flux were assumed to exist at both the tissue–blood and
tissue–gut lumen interfaces, and 12 different rate constants were fitted by modeling to measured residues
in tissues and gut contents.
Nichols et al. (1998) used two approaches to model the accumulation of TCDD in feeding studies
with brook trout. The first approach was based on the assumption that a chemical equilibrium is
established between fecal material contained within the terminal colon (C ) and venous blood draining
fec
the GIT (C ):
vgt
C = C P (3.105)
fec
vgt fb
where P is the feces–blood equilibrium chemical partition coefficient. Termed the fecal partitioning
fb
submodel, this approach is similar to that proposed by Barber et al. (1991) to model PCB accumulation
by lake trout, the principal difference being that an equilibrium is established with venous blood exiting
the gut and not with the entire fish.
From mass-balance considerations, it follows that:
( Q food C food) −( Q C fec) = ( QC vgt) −( QC art) (3.106)
fec
gt
gt
where Q food is the feeding rate, C food is the TCDD concentration in the diet, Q is the fecal egestion
fec
rate, Q is the blood flow rate to the lower intestines, and C is the chemical concentration in arterial
art
gt
blood. Substituting Equation 3.105 into Equation 3.106 and solving for C yields:
vgt
C vgt = ( Q C art + Q food C food) ( Q gt + Q P fb) (3.107)
gt
fec
Venous blood draining the tissues of the lower intestine was then assumed to mix with arterial blood
flowing to the liver.