Page 374 - The Toxicology of Fishes

354 The Toxicology of Fishes

gene. Also, neofunctionalization may occur in which duplication of one daughter gene with the ancestral
function occurs while the other acquires new functions (He and Zhang, 2005). It seems likely that
subfunctionalization or neofunctionalization of NR paralogs may have been involved in the generation
of fish variability. Additionally, from a molecular perspective, multiplicity of nuclear receptors may be
an important factor that contributes to both signal diversification and specification (Gronemeyer et al.,
2004). In this capacity, gene paralogs may impart novel functions that suit the unique physiological
demands of inhabiting an aquatic environment or specific reproductive strategies. As discussed below,
additional differences are noted between teleost and mammalian NRs, including structural changes
resulting in differential ligand-binding characteristics (peroxisome proliferator-activated receptor [PPAR]
alpha and gamma), absence of particular NRs, constitutive androstane receptor (CAR), or quantitative
differences in ligand binding (ER).
The NR1I Subfamily
The NR1I subfamily has received a great deal of attention due to an essential role in regulating phase
I and II genes involved in xenobiotic metabolism. For many years, the mechanisms governing cytochrome
P450 gene induction following xenobiotic exposure remained elusive. AhR–CYP1A interactions were
well described, but the processes by which induction of CYP2, CYP3, and CYP4 families occurred had
yet to be determined (Hahn et al., 2005). With the discovery of the nuclear receptors CAR and pregnane
X receptor (PXR), our mechanistic understanding of CYP regulation was greatly enhanced. Both CAR
and PXR are low-affinity (high substrate concentration) xenosensors in mammals, capable of regulating
genes associated with the metabolism, transport, and elimination of exogenous substrates. Human and
rodent PXR can be activated by a variety of compounds known to induce hepatic P450 enzymes, including
prescription drugs, steroids, and bile acids, and by several suspected endocrine-disrupting compounds
(Hurst and Waxman, 2004; Masuyama et al., 2002; Moore et al., 2002). This broad substrate promiscuity
of PXR is due to a 50- to 60-amino-acid insert between helix 1 and helix 3. The position of this insert
was confirmed by x-ray examination, which revealed an unusually large ligand-binding pocket capable
of accommodating a wide range of lipophilic ligands (Moore et al., 2002). Significant species differences
arose from structural changes in the LBD resulting in an array of ligand-binding and transactivation
specificities. PXR was initially identified as a candidate xenobiotic receptor based on its association with
the induction of the hepatic P450 enzymes CYP2B and CYP3A (Savas et al., 1999; Waxman, 1999; Xie
and Evans, 2001). Subsequent work proved that PXR mediates metabolism and the elimination of harmful
hepatotoxic compounds by the concerted action of the oxidative phase I CYP enzymes, phase II
conjugating enzymes, and drug transporters (Willson and Kliewer, 2002). To date, PXR is involved in
the regulation of transcription targets for phases I, II, and III, including CYP2B, CYP3A, UGT1A1,
MDR1, CYP24, and 5-AAS (Reschly and Krasowski, 2006). The development of PXR transgenic and
knockout mice, the use of microarrays, and being able to screen mammalian genomes for putative PXR
response elements have helped identify numerous PXR genes and their targets, including cell growth
and differentiation and heme biosynthesis, among others, thus raising the possibility of a broader
physiological role for PXR (Hartley et al., 2004).
The molecular characteristics of CAR and PXR have been compared in mammals. CAR differs from
PXR in ligand-binding affinities, cytoplasmic localization, basal activity, and transrepression by specific
ligands, including androgen steroids (Kakizaki et al., 2003). CAR does not contain the helix 1–3 insert
as PXR does, with the result that a smaller ligand-binding pocket in CAR restricts the diversity of
suitable ligands. The action of CAR as a xenobiotic receptor was confirmed by several studies examining
CAR-dependent gene transcription (Ueda et al., 2002). Unique to the NR1I nuclear receptor family,
CAR can undergo ligand-independent activation following treatment with phenobarbital (PB). The
manner in which this is achieved may involve PB initiating a phosphorylation-dependent signal cascade
that leads to translocation of CAR from the cytoplasm to the nucleus without direct PB binding (Kodama
and Negishi, 2006). Additionally, CAR and PXR share overlapping transcriptional targets through
recognition of similar DNA response elements. This cross-talk is hypothesized to operate as a metabolic
safety net between receptors (Maglich et al., 2002); however, given the striking differences in their
pharmacologic profiles, these receptors may have evolved to serve distinct physiological roles (Moore
et al., 2002).