Page 351 - The Toxicology of Fishes

Liver Toxicity 331

In this chapter, we seek to: (1) briefly review key aspects of liver morphology and physiology of fish,
emphasizing those features that differ among fish species and between fish and their mammalian
counterparts, and emphasizing why and how these features are relevant to understanding the liver toxicity
of fish; (2) review the literature on liver pathology in toxicant-exposed fish and provide insight into
major pathological and pathophysiological response patterns of fish liver; and (3) identify specific
information gaps and provide suggestions for future directions.

Liver Structure

Fundamental aspects of fish liver anatomy, especially those important in the interpretation of toxicity,
were recently reviewed (Hinton and Couch, 1998; Hinton et al., 2001), and the interested reader is
referred to these detailed descriptions. In an extensive treatise on the biology and pathobiology of the
human liver by Arias et al. (1988), six specific principles were listed that are required and govern structure
and function of a liver. We paraphrase these as follows: (1) exocrine and metabolic functions expressed
in the same cell (the hepatocyte); (2) a double blood supply (arterial from the dorsal aorta and afferent
venous from the hepatic portal vein); (3) a specific architectural arrangement of single cells and cell
masses that facilitates exchange between blood and hepatocytes; (4) features of perihepatocellular or
perisinusoidal spaces (of Disse), such as the absence of basement membrane, thereby maximizing the
exchange between blood and hepatocyte; (5) separation of the lumen of biliary from blood spaces,
thereby compartmentalizing excretory from other hepatocytic functions; and (6) specific biochemical
activities of hepatocytic receptors, enzymes, organelles, and cell membranes that regulate and enable
specific functions.
As is demonstrated in subsequent sections of this chapter, fish hepatocytes fulfill requirement 1
above. In addition, we review and provide new evidence for the dual blood supply (requirement 2).
The specific architectural pattern of cell masses is very similar in embryos and larval equivalent livers
of mammals and fishes; however, with respect to requirement 3, mammalian adult liver differs from
livers of fishes, reptiles, amphibians, and birds. Electron microscopic observations reveal fenestrae in
sinusoidal endothelium of mammals and fishes, and both lack a basement membrane to facilitate
exchange (requirement 4). Fish liver meets requirement 5, and these functional characteristics are
reviewed in the subsequent section.
Despite of the many structural similarities between the livers of fishes and those of mammals, certain
features of hepatic gross and microscopic anatomy of fishes are recognized as different (Hampton et al.,
1985, 1988, 1989). Some of these include the number of lobes; presence or absence of lobules and portal
tracts; architectural arrangement of hepatocytes; the absence in many fish species of Kupffer cells; and
the distinctive and common presence of macrophage aggregates in fishes (Table 7.1).
Coverage of liver anatomy in fish is not complete without mention of the exocrine pancreatic cells
that occupy the periadventitia of portal veins within the body of the liver. Certain species contain abundant
pancreatic cells (varying in amount), thus leading to the term hepatopancreas. The recent observations and
three-dimensional reconstructions of the tilapia (Oreochromis niloticus) should be consulted (Figueiredo-
Fernandes et al., 2007).
The above structural similarities and differences represent more than a simple academic exercise in
comparative liver anatomy; rather, they are important in interpretation of toxic responses within livers
of fishes. In vitro studies with isolated primary hepatocyte preparations (Rabergh and Lipsky, 1997) or
precision slices of trout liver (Singh et al., 1996) have clearly shown that exposure is followed by the
presence of cytosolic enzymes in culture medium, validating the toxic nature of these responses.
Architectural (the absence of lobules) and physiological (lack of metabolic zonation; see below) differ-
ences in fish vs. rodent livers apparently lead to different histological patterns of toxicity. Assuming that
the absence of a mosaic pattern of response, such as centrolobular necrosis following carbon tetrachloride
exposure of rodents (see below), indicates that fish are not capable of responding to such agents is
erroneous; instead, other patterns of morphologic response are likely in fish (Droy, 1988).