222
P.F. Landrum and s.w. Fisher
that uses lipoprotein carriers. If the passive diffusion model is true, in the strictest
sense, then the mode of transport is diffusion created by a fugacity gradient and
the site of magnification is the GIT (Gobas et aI., 1993b). If, however, the second
model is true, then the mode of accumulation is lipid co-assimilation and the site
of magnification is in the organism's (or predator's) tissues.
Gobas et al. (l993b) evaluated these two models experimentally in goldfish
(Carassius auratus) by using chlorinated benzenes and PCBs. In this experiment,
groups of goldfish were fed identical concentrations of contaminant in food, with
lipid levels ranging from 0 to 13.5%. If the contaminants were transferred from
food to fish strictly via lipid co-assimilation, then contaminant uptake should be
very low in a diet deficient in lipid because lipids are the primary transport vehicle
for the contaminants. By contrast, contaminant uptake in high-lipid diets should
be significantly greater than from low-lipid diets if lipid co-assimilation predominates. If passive diffusion is responsible for contaminant transfer, then a dramatic
change in dietary lipid content should occasion only a small alteration in the
amount of contaminant taken up, assuming that fecal egestion rates do not change
between treatments.
The results of the experiment suggested that although diffusion resulting from
increased fugacity caused by digestion was the main force driving accumulation,
lipid co-assimilation also played a role depending on the hydrophobicity of the
contaminant. For the lower log Kow contaminants, there was no difference in
uptake efficiencies from low- versus high-lipid diets. That is, diffusion appeared
to be the driving force behind accumulation, and magnification was taking place
in the GIT. However, for hexa-, octa-, and decachlorobiphenyl, the uptake efficiency was significantly higher from low-fat food than from high-fat food due to
greater digestibility (Gobas et aI., 1993b). This indicated that the amount of
transfer was tied to dietary lipid levels as suggested by the lipid co-assimilation
model, albeit in the direction opposite that which was initially predicted. This was
attributed to the higher digestibility of the low-lipid food. Still, the fact that
contaminant uptake was associated with dietary lipid levels was viewed as evidence for the lipid co-assimilation model.
In fact, experimental results exist to support both models. For instance, a
tenfold increase in dietary triglycerides had no effect on benzo[a)pyrene (BaP),
3-methy1cholanthrene, or aminostilbene accumulation from food in rats (Laher et
aI., 1984; Kamp and Neumann, 1975). This lends credence to the passive diffusion
model. Evidence in support of the lipid co-assimilation model has also been
adduced, particularly for extremely hydrophobic contaminants such as BaP (Vetter et aI., 1985; Rees et aI., 1971).
Given the contradictory experimental results, Gobas et al. (l993b) proposed a
model that includes aspects of both the diffusion and lipid co-assimilation mechanisms. In this view, contaminants remain associated with dietary lipids as they
move from food into contact with intestinal mucosa. This association is particularly significant in the case of extremely hydrophobic contaminants for which
disassociation into the aqueous milieu of the gut is thermodynamically and kinetically unfavorable. Thus, attaching to lipid material during the transport avoids
P.F. Landrum and s.w. Fisher
that uses lipoprotein carriers. If the passive diffusion model is true, in the strictest
sense, then the mode of transport is diffusion created by a fugacity gradient and
the site of magnification is the GIT (Gobas et aI., 1993b). If, however, the second
model is true, then the mode of accumulation is lipid co-assimilation and the site
of magnification is in the organism's (or predator's) tissues.
Gobas et al. (l993b) evaluated these two models experimentally in goldfish
(Carassius auratus) by using chlorinated benzenes and PCBs. In this experiment,
groups of goldfish were fed identical concentrations of contaminant in food, with
lipid levels ranging from 0 to 13.5%. If the contaminants were transferred from
food to fish strictly via lipid co-assimilation, then contaminant uptake should be
very low in a diet deficient in lipid because lipids are the primary transport vehicle
for the contaminants. By contrast, contaminant uptake in high-lipid diets should
be significantly greater than from low-lipid diets if lipid co-assimilation predominates. If passive diffusion is responsible for contaminant transfer, then a dramatic
change in dietary lipid content should occasion only a small alteration in the
amount of contaminant taken up, assuming that fecal egestion rates do not change
between treatments.
The results of the experiment suggested that although diffusion resulting from
increased fugacity caused by digestion was the main force driving accumulation,
lipid co-assimilation also played a role depending on the hydrophobicity of the
contaminant. For the lower log Kow contaminants, there was no difference in
uptake efficiencies from low- versus high-lipid diets. That is, diffusion appeared
to be the driving force behind accumulation, and magnification was taking place
in the GIT. However, for hexa-, octa-, and decachlorobiphenyl, the uptake efficiency was significantly higher from low-fat food than from high-fat food due to
greater digestibility (Gobas et aI., 1993b). This indicated that the amount of
transfer was tied to dietary lipid levels as suggested by the lipid co-assimilation
model, albeit in the direction opposite that which was initially predicted. This was
attributed to the higher digestibility of the low-lipid food. Still, the fact that
contaminant uptake was associated with dietary lipid levels was viewed as evidence for the lipid co-assimilation model.
In fact, experimental results exist to support both models. For instance, a
tenfold increase in dietary triglycerides had no effect on benzo[a)pyrene (BaP),
3-methy1cholanthrene, or aminostilbene accumulation from food in rats (Laher et
aI., 1984; Kamp and Neumann, 1975). This lends credence to the passive diffusion
model. Evidence in support of the lipid co-assimilation model has also been
adduced, particularly for extremely hydrophobic contaminants such as BaP (Vetter et aI., 1985; Rees et aI., 1971).
Given the contradictory experimental results, Gobas et al. (l993b) proposed a
model that includes aspects of both the diffusion and lipid co-assimilation mechanisms. In this view, contaminants remain associated with dietary lipids as they
move from food into contact with intestinal mucosa. This association is particularly significant in the case of extremely hydrophobic contaminants for which
disassociation into the aqueous milieu of the gut is thermodynamically and kinetically unfavorable. Thus, attaching to lipid material during the transport avoids