Abstract
Glucocorticoids (GCs) are commonly used to treat a variety of immune diseases. However, the efficacy of treatment is greatly influenced by an individual variation in sensitivity to GCs, which is caused by differences in the glucocorticoid receptor (GR). The variable receptor profile results from variations in the GR gene, or alternative splicing of the gene coded. We investigated the evolution of the GR gene by comparing genomic GR sequences of vertebrates. Exon length and amino acid sequence are conserved among all classes of vertebrates studied, which indicates strong evolutionary pressure on conservation of this gene. Interestingly, teleostean fishes have two different GR proteins. One of the duplicate fish GR genes has a nine-amino-acid insert in the DNA binding region that results from alternative splicing. The duplicate GR genes and products of alternative splicing in teleostean fishes are differentially expressed in vivo and show different transactivation capacity in vitro. The presence of two GR genes appears to be a result of divergence of receptors rather than of ligands. Teleostean fishes express different, evolutionarily related, functional GR proteins within a single organism. Hereby, teleostean fishes present a model that facilitates investigation of the molecular basis of cortisol resistance and different regulatory functions of cortisol.
Introduction
Glucocorticoids (GCs; cortisol and corticosterone) play a pivotal role in vertebrate physiology through a plethora of control mechanisms. The GC system, with a nuclear glucocorticoid receptor (GR), is found in all vertebrates, consistent with an evolutionary well-conserved stress response. GCs regulate cell growth, bone density, metabolism and cardiovascular system and influence behaviour (Charmandari et al. 2005). Moreover, in vitro studies with human, murine and rat immune cells showed the immunosuppressive effects of elevated GC levels following stressful circumstances. GCs suppress Th1 cellular immunity and mediate a Th2 shift by suppressing production of T helper cells type 1/(Th1; tumour necrosis factor-α, interferon-γ, interleukin (IL)-2 and IL-12) cytokines and inducing production of T helper cells type 1/ (Th2; IL-4, IL-10 and IL-13) cytokines (Elenkov 2004).
For this reason, GCs are used to treat a variety of immune diseases. Local anti-inflammatory properties of GCs make them a first-choice medication for asthma (Walsh 2005) and rheumatoid arthritis (Boers 2004). However, there is considerable individual variation in sensitivity to GCs (Hearing et al. 1999), which may affect the outcome of GC treatment. Mutations and splice variation in the GR gene and subsequent polymorphisms in the protein of humans and other vertebrates may explain the individual variation in sensitivity, rather than alteration in the consensus sequence of the GR-binding site in the DNA of target genes (Brandon et al. 1991, Keightley & Fuller 1994, Lamberts et al. 1996, DeRijk et al. 2002, van Rossum et al. 2002, Bray & Cotton 2003, Stevens et al. 2004b).
Recently it was shown that two teleostean (bony) fishes, the rainbow trout (Oncorhynchus mykiss; Bury et al. 2003) and Burtons’ mouthbrooder (Haplochromis burtoni; Greenwood et al. 2003), each have two different GR genes. The two different receptors encoded by these genes display high amino acid sequence identity, especially in the part of the gene coding for the C-terminal part of the receptor protein. Rainbow trout GR1 and GR2 show 51% overall identity, and 86% identity in the C-terminal part (DNA-binding region, hinge region and ligand-binding region); for Burtons’ mouthbrooder, these percentages are 49 and 80% respectively. Moreover, alternative splicing occurs in these fish GR genes (Takeo et al. 1996, Greenwood et al. 2003), which leads to functional transcripts with different transactivation properties compared to wild-type. Both the duplicate GR genes as well as the splice variants of these GR genes in fish are differentially expressed in tissues and differ in their affinity for cortisol, the single dominant GC in fish (Bury et al. 2003, Greenwood et al. 2003).
We question whether differences among GRs affect the sensitivity to GCs and the potential impact of these differences for biological responses. We address the meaning of duplicate genes and splice variants. Genomic database information and review of the literature give evidence that duplicate GR genes and alternative splicing are common features of all teleostean fishes and set them apart from other vertebrate taxa.
Glucocorticoid receptor
The GR (Hollenberg et al. 1985) structure and function is well established in mammals and conserved among all vertebrate species analysed so far. The GR belongs to the nuclear receptor superfamily (Fuller 1991, Mangelsdorf et al. 1995, Kumar & Thompson 1999, Evans 2005). Its members act as ligand-dependent transcription factors. All receptors in this superfamily, which include those for steroid hormones, thyroid hormones, retinoic acid and vitamin D3, share a similar domain structure, which was first predicted for the GR (Giguere et al. 1986) (Fig. 1). The N-terminal region varies greatly among different members of the superfamily, both in size and composition, and is involved in transactivation of downstream genes (hereafter transactivation). Mutations in this domain decrease transcriptional activity of target genes without affecting ligand affinity (Giguere et al. 1986). The DNA-binding region (see below) is the central domain and binds to glucocorticoid responsive elements (GRE) in promoter regions to initiate transcription of a vast array of GC responsive genes. The amino acid sequence of this region is strictly conserved, both in different members of the superfamily and in virtually all vertebrate species (Fig. 1). The hinge region, involved in conformational changes during receptor–ligand binding, is quite variable in its length and amino acid sequence. The relatively well-conserved ligand-binding region (see below) is situated at the C-terminus. The marked similarity of amino acid sequences in the ligand-binding region explains why multiple receptors can bind the same ligand and, subsequently, elicit a similar response (activation of a downstream gene) in an in vitro transactivation assay. In these assays, a cell lacking endogenous corticoid receptors is co-transfected with a transcription vector with the corticoid receptor to be tested and a reporter plasmid bearing one or multiple GREs in its promoter. After addition of the ligand, the in vitro expressed corticoid receptor protein can bind the ligand and translocate to the nucleus, where it will subsequently bind to the GRE on the promoter of the reporter gene and induce transcription of this gene. These in vitro experiments showed that cortisol enhanced transcriptional activity of the mineralocorticoid receptor (MR) of Burtons’ mouthbrooder (EC50 0.02 nmol) at a lower concentration than that required for transactivation via the GR (EC50 2–5 nmol) in one and the same assay. In rainbow trout, a similar result was found; the MR required a lower concentration of cortisol to induce transcriptional activity and a difference was seen for the duplicate GRs with EC50s of ~100 nmol, ~10 nmol and ~1 nmol, for GR1, GR2 and MR respectively (Greenwood et al. 2003, Sturm et al. 2004).
DNA-binding region
The 3D structure of the DNA-binding region of the human GR was confirmed by nuclear magnetic resonance imaging and it shows two sub-domains (CI and CII), each consisting of a zinc finger (Hard et al. 1990).
The most N-terminally located zinc finger is involved in binding of the receptor to DNA. Site-specific DNA recognition depends on the amino acids glycine, serine and valine in the proximal (P) box that bind to the hormone responsive element (GRE) of target genes (Umesono & Evans 1989, Luisi et al. 1991). The main loop of this N-terminal zinc finger is a short segment of antiparallel β-sheet, whereas the second pair of cysteines is at the start of an α-helix that provides the contact with the major groove of the GRE. This α-helix between Ser459 and Glu469 (in rat) is referred to as the DNA recognition helix (Hard et al. 1990, Luisi et al. 1991). The more C-terminal zinc finger is involved in receptor homodimerisation at the GRE and stabilises binding of the GR to the GRE, which is mediated by amino acids AGRND in the distal (D) box (Luisi et al. 1991, Kumar & Thompson 1999, 2005). The zinc fingers together with the conserved amino acids in the two sub-domains result in the 3D structure required for specific DNA binding and thereby for the downstream effects of GCs. This is demonstrated by insertional mutagenesis, where disruption of the C-x-x-C motif silences transcriptional activity of the receptor without interfering with the ligand-binding ability (Giguere et al. 1986, Wickert & Selbig 2002).
Ligand-binding region and heat shock proteins (Hsps)
To allow binding of the ligand, the receptor conformation has to change to give the ligand access to a hydrophobic cleft in the ligand-binding region (Pratt & Toft 2003). Heat shock proteins Hsp90 and Hsp70 are both essential as chaperones for the opening of this cleft (Morishima et al. 2000b). In formation of the hormone-receptor complex, the chaperone Hsp70 first binds directly to the ligand-binding region, which induces a conformational change in the GR that initiates the opening of the steroid-binding cleft, an ATP-consuming process (Morishima et al. 2000a). Hsp90 binds to this primed GR*Hsp70 complex and completely opens the cleft. As soon as the steroid-binding cleft is opened, cortisol enters and binds to the hydrophobic pocket of the ligand-binding region (Morishima et al. 2000a, Kanelakis et al. 2002, Pratt & Toft 2003). After binding of the hormone with the GR*Hsp70*Hsp90 (GR-hetero) complex, the chaperone complex dissociates. The receptor conformation returns from a labile open structure into a stable compact DNA-binding state that can exist independently of chaperone complexes (Pratt & Toft 1997). Subsequently, the receptor is translocated into the nucleus, where it will homodimerise and bind as a homodimer to the specific DNA motifs (GREs) which in turn will activate transcription of the downstream genes (Schoneveld et al. 2004).
However, translocation of the receptor to the nucleus depends on phosphorylation of the receptor (Ismaili & Garabedian 2004). Furthermore, phosphorylation is involved in the half-life of GR and in downregulation following ligand treatment (Webster et al. 1997). Thus, GR phosphorylation has an essential effect on the ultimate receptor function (Ismaili & Garabedian 2004).
Although details of the formation of the GR heterocomplex are yet to be established in teleostean fishes, we assume, based on sequence conservation of Hsp70, Hsp90 and the GR ligand-binding region, that comparable processes take place in all fish and thus in vertebrates. At present, data on phosphorylation of the teleostean fish GR proteins are not available. However, we will make a prediction based on conserved phosphorylation consensus motifs in alignments of the N-terminal region of fish and mammalian sequences.
Stress and the immune system
The stress response of all vertebrates, including teleostean fishes, is comparable. In fish, stress signals are conveyed from the hypothalamus via corticotrophin-releasing hormone to the pituitary, where adrenocorticotrophic hormone (ACTH) is released into the circulation. ACTH in turn activates the cortisol-producing cells of the head kidney, the functional equivalent of the adrenal gland. In fish under acute stress, cortisol levels can easily increase more than tenfold, from 10–20 ng/ml to over 350 ng/ml (Wendelaar Bonga 1997). Both basal levels as well as increases in cortisol level during stress are in the same range as those in the mammalian stress response (Van Cauter et al. 1996, Hennessy et al. 1997, Piekarzewska et al. 2000).
High cortisol levels significantly affect the immune system (Sapolsky et al. 2000, Franchimont 2004, Charmandari et al. 2005). However, the resulting response depends on the parameter studied and type of immune cell involved. In vitro studies showed that, in common carp, high levels of cortisol that occur during stress rescue neutrophilic granulocytes from apoptosis by neglecting them (Weyts et al. 1998b). Activated B-lymphocytes, however, show increased apoptosis when exposed to high levels of cortisol (Weyts et al. 1998a). Thus, in stressed fish, the innate immune system becomes activated. This is further illustrated by the redistribution of neutrophilic granulocytes from the head kidney into circulation after immersionvaccination (Huising et al. 2003). This stress-induced neutrophilia is considered to improve peripheral surveillance and is also reported in mammals (Dhabhar et al. 1996). Meanwhile, the adaptive immune system becomes temporarily repressed (Franchimont 2004).
Evolution of vertebrate GRs
GR proteins of teleostean fishes and other vertebrates were compared based on genomic sequences and gene structure of representatives of mammals, birds, amphibians and bony fishes. Moreover, the duplicate amino acid sequences of Burtons’ mouthbrooder and rainbow trout GR proteins and other fish species, for which only a single GR protein has been identified to date, were investigated. From this information, we dare to make general predictions about the similarities and, more interestingly, the differences of fish GR sequences compared to other vertebrate GR sequences.
In this review, we obey the present rainbow trout GR nomenclature with subscripts ‘a’ and ‘b’ for splice variants. (Bury et al. 2003). In this nomenclature, GR1 has alternative splicing, in which GR1a has a nine-amino-acid insert and GR1b does not. In GR2, alternative splicing has so far not been reported. Unfortunately, the published nomenclature of Burtons’ mouthbrooder is deviant; GR2b has the nine-amino-acid insert as a result of alternative splicing, and GR2a does not have this insert. The GR1 gene does not show alternative splicing (Greenwood et al. 2003).
Conservation of sequence identity, but two different GR genes
There are over 23 000 species of teleostean fishes compared to ‘only’ 14 000 species of birds and mammals taken together (Nelson 1994). Species of teleostean fishes are vastly different and live in very diverse ecosystems, but still use the same basic regulatory mechanisms.
As mentioned earlier, it has been shown that both rainbow trout as well as Burtons’ mouthbrooder have duplicated GR genes (Bury et al. 2003, Greenwood et al. 2003). It was then investigated whether other teleostean fishes also have duplicate copies of their GR gene. Using the Ensembl genome database, the gene structures of both tetraodon (Tetraodon nigroviridis) and fugu (Takifugu rubripes) GR genes were predicted. Both these teleostean fishes were considered ideal for sequencing as their genomes are extremely compact due to a major lack of intronic DNA and this enabled efficient comparisons with other vertebrate genomes (Hedges & Kumar 2002). As predicted, in the genome of both these species, duplicate copies of the GR gene were found. These two different proteins are encoded on different scaffolds in fugu and on different chromosomes in tetraodon (GR1 on chromosome 7 and GR2 on chromosome 1). The duplicate copies of the GR genes have conserved amino acid sequences, but are distinctly different. Both teleostean fish GR genes are conserved in exon length (Fig. 3) and amino acid sequences (Fig. 1). The zebrafish (Danio rerio) genome also has been sequenced. To date, however, only one zebrafish GR gene has been identified.
A phylogenetic tree was constructed, comparing the GR sequences of different vertebrate species using the neighbour-joining method (Saitou & Nei 1987, Kumar et al. 2004) (Fig. 2). This phylogenetic tree clearly shows a cluster of tetrapod GRs and a cluster of teleost fish GRs. This cluster of teleost GRs is subdivided into two different clades, representing two different GR genes. The duplicate GR genes of teleostean fishes appear to be the result of the genome duplication that occurred early in the evolution of actinopterygii, i.e. before the radiation of the teleostean fishes and after the divergence of tetrapods from the fish lineages, which occurred around 450 million years ago (Hedges & Kumar 2002). Duplicate GR genes are seen in both salmonids (rainbow trout) and percomorphs (fugu, tetraodon and Burtons’ mouthbrooder), which are two lineages of fish species that diverted very early in teleostean fish evolution. Therefore, it is reasonable to assume that the presence of duplicate GR genes throughout fish species relates to the fish-specific genome duplication that occurred approximately 300 million years ago (Hoegg et al. 2004, Nelson 1994, Volff 2005).
A ‘corticoid receptor’ of a sea lamprey (Petromyzon marinus) was included in our analysis. This receptor was named corticoid receptor as it resembled both the vertebrate corticoid receptors, MR and GR. Sea lamprey is an agnathan (jawless) fish of an evolutionary lineage that preceded the teleostean fish lineage. Moreover, it was predicted that evolution of the vertebrate steroid receptors took place from an ancestral gene through serial genome expansions. In this hypothesis, a common corticoid receptor gene was the ancestral entity for both the GR and the MR (Thornton 2001). Indeed, our analysis shows that the lamprey corticoid receptor seems to cluster with the outgroup MR, not with any of the GR genes. Based on this phylogenetic tree, we therefore cannot predict which of the duplicate GR genes is more reminiscent of the ancestral GR.
If the duplication of the GR gene is indeed a result of the fish-specific genome duplication, we predict that many more teleostean fishes will have duplicate copies of their GR gene. In fact, from the genome point of view, all teleostean fishes should have at least duplicate GR genes, but one of the copies could of course be secondarily lost due to mutational events if one copy was sufficient for biological function.
Phosphorylation sites in N-terminal region of fish GR
As mentioned earlier, phosphorylation has significant effects on GR function. By comparing the mouse, human and rat phosphorylation sites, Ismaili and co-workers constructed an instructive scheme of consensus sequences for phosphorylation and the putative protein kinases that are predicted to target certain motifs (Ismaili & Garabedian 2004). By alignment of the GR gene sequence of several vertebrate species (both mammalian and fish), predictions can be made about presence and number of phosphorylation sites in teleostean fishes (Table 1).
The mammals in this alignment show good similarity. Similarity decreased when chicken, African clawed frog and teleostean fishes were compared. The serine residues corresponding to Ser122 and Ser234 in mouse are conserved in mammals, chicken and African clawed frog. The serine corresponding to Ser315 in mouse is conserved throughout all the species but, as there is no consensus motif, it is hard to predict if this serine is indeed part of a true phosphorylation site in all the species.
The alignment suggests that the consensus motif for the casein kinase II, (CKII; S/T(P)-x-x-E/D) is widely present in the AB region of teleostean fishes and found at conserved locations in a series of fish species. Moreover, all fish species examined show one or more mitogen-activated protein kinases motifs (non-polar-x-S/T(P)-P) in their AB domain, yet these motifs do not correspond to conserved locations. Furthermore, the consensus motif for cyclin-dependent kinase (CDK) (S/T(P)-P-x-R/K) could not be found in any of the fish species examined. This indicates that fish exploit a different consensus motif for the CDKs, or that different kinases are used to phosphorylate the GR. Indeed, at the location corresponding to Ser212 in mouse, the fish species have a different consensus motif (S/T(P)-x-x-E/D) than mammals.
Whether the conserved serines and consensus motifs for the different kinases do indeed correspond with phosphorylation sites needs to be confirmed by peptide mapping, mutagenesis and phosphorylation studies.
DNA-binding region in fish GR
Another interesting difference between GR proteins in teleostean fishes and other vertebrates is seen in the DNA-binding region. This C-domain is the best-conserved region of the GR protein (98, 100, 100 and 85% amino acid identity between human GR and mouse, chicken, xenopus and tetraodon GR1a sequences respectively). However, some fish GR1 genes show a 27-nucleotide insert that encodes nine additional amino acids. Apart from this insert, the C-domain of teleostean fishes is almost identical to that in other vertebrates: 96 versus 85% amino acid identity between human GR and tetraodon GR1b and GR1a respectively; Fig. 1). These nine amino acid inserts between the two zinc fingers are also remarkably conserved among teleostean fish species. In rainbow trout (Ducouret et al. 1995, Takeo et al. 1996), Japanese eel (Anguilla japonica) (Todo & Nagahama 1998), Japanese flounder (Paralichthys olivaceus) (Tokuda et al. 1999), Burtons’ mouthbrooder (Greenwood et al. 2003), brown trout (Salmo trutta; AY863149), European sea bass (Dicentrarchus labrax; AY549305) and fugu (T. rubripes) a WRARQNTDG insert is present. A slightly different insert is found in tetraodon (T. nigroviridis): WRARQNTVC. Especially, the D to V substitution is interesting, as aspartic acid is hydrophilic and negatively charged, whereas valine is hydrophobic. Glycine and cysteine are also substantially different with regard to size and biochemical properties.
The nine amino acid inserts appear to be the result of alternative splicing and are encoded by an extra exon. This extra exon is ‘hidden’ in the intron that separates exons 3 and 4 (Fig. 3). This is evident from the genomic sequences of fugu and tetraodon and was confirmed experimentally by the amplification of a genomic DNA fragment in rainbow trout (Lethimonier et al. 2002).
Duplicate GRs; expression and transactivation capacities
The presence of duplicate GR genes which also show this particular alternative splicing gives reason to hypothesise that the resulting duplicate receptor proteins have separated functions. This hypothesis is strengthened as teleostean fishes do not possess the mineralocorticoid aldosterone and use cortisol instead of osmoregulation (Dean et al. 2003, Metz et al. 2003). This adds another important function to the long list of different functions to be regulated by one hormone. To be able to regulate all these different functions with a single hormone, one can assume that differentiation occurred in the hormone receptor.
Differences in receptor protein function could be indicated by differential expression of the duplicate GR genes in different organs. A literature comparison of different species of fish reveals that both genes are expressed in all organs examined. In all these species, GR1 (the gene showing the nine-amino-acid insert) is more strongly expressed than the GR2 gene (Bury et al. 2003, Greenwood et al. 2003). Moreover, most first-reported teleostean fish GR sequences have the nine-amino-acid insert and would therefore be classified as GR1, such as in European sea bass and brown trout (Ducouret et al. 1995, Takeo et al. 1996, Todo & Nagahama 1998, Tokuda et al. 1999). Whether this reflects relative abundance of the GR1 messenger or preferential primer design is unclear.
Separate biological functions of the duplicate GR forms may further result from different affinity for the ligand or from different transactivation properties. Information about ligand binding is scarce, but in rainbow trout it was reported that duplicate GR genes have similar dexamethasone-binding affinities that are comparable to the affinity of human GR for dexamethasone (Ducouret et al. 1995, Ray et al. 1999, Bury et al. 2003). However, the resulting biological effect, as measured with transactivation assays, is quite different for the duplicate GR genes.
In rainbow trout, concentrations of cortisol required to induce activation of downstream genes is significantly different between the duplicate GR proteins. Rainbow trout (rt)GR2 requires lower concentrations of cortisol (EC50 of 0.7 nM) than rtGR1a (with nine-amino-acid insert; EC50 of 46 nM) to induce transactivation (Bury et al. 2003).
Variation in transactivation properties of the receptor after stimulation with a ligand is not the only result of the GR protein involved. Variation in transactivation also depends greatly on the ligand used. In in vitro experiments in rainbow trout, the GR agonist dexamethasone activated the tested downstream gene at lower concentrations than cortisol in the same experiment: EC50 for GR1 stimulated with cortisol was 46 nM while it was 4.3 nM with dexamethasone treatment; for GR2, this was 0.7 nM for cortisol and 0.35 nM for dexamethasone (Bury et al. 2003). In other experiments, it was shown that, with the same concentration of ligand, dexamethasone gave a stronger activation of the reporter gene than cortisol (Takeo et al. 1996, Tokuda et al. 1999, Greenwood et al. 2003).
It is crucial to keep this difference between cortisol and dexamethasone in mind when testing newly found GR genes or GR variants. Indeed, experiments with dexamethasone might slightly (rainbow trout GR2; twofold) or severely (rainbow trout GR1; tenfold) overestimate the cortisol-induced transactivation capacity.
GR splice variants; expression and transactivation capacities
Several different splice variants are described for the GR gene (Oakley et al. 1999, Geng et al. 2005). However, in this review, we focus only on the splice variations in the DNA-binding region of the GR gene.
Again, as with the duplicate genes, it is interesting to hypothesise about functional differences between the two splice variants. The expression level of the GR1 variant with the nine-amino-acid insert exceeds that of the non-spliced form (Takeo et al. 1996, Bury et al. 2003, Greenwood et al. 2003). In Burtons’ mouthbrooder, the variant without the insert was found in all organs tested in rainbow trout; however, it could only be detected in testis (Takeo et al. 1996, Greenwood et al. 2003). As mentioned before, most teleostean fish GR sequences that were found show this nine-amino-acid insert (Ducouret et al. 1995, Takeo et al. 1996, Todo & Nagahama 1998, Tokuda et al. 1999). It remains to be investigated if in all species mentioned a variant of the GR1 gene without the nine-amino-acid insert is also expressed. This variant can easily be overlooked as the difference between the two is so small that specific primers are required to detect them.
Differences in biological functions of the splice variants could also result from different affinity for the ligand or different transactivation properties related to this difference in affinity. As there is no difference between the splice variants in the ligand-binding region, it is unlikely that there is any effect on ligand-binding affinity.
The only differences between the splice variants are found in the DNA-binding region and, indeed, differences in DNA-binding affinity do occur. In an experiment with rainbow trout, two different promoters were used. One luciferase reporter plasmid contained one and the other two GREs from the rat tyrosine aminotransferase promoter. Interestingly, whereas the GR1a (with the nine-amino-acid insert) has a better binding affinity for single GRE than the GR1b, the opposite was true for the double GRE sequence (Lethimonier et al. 2002).
Differences in the final transactivation capacity also occur. Quanta Protein Design computer (Accelrys, Cambridge, UK) predictions of corticoid receptor DNA-binding domain function based on secondary structure analysis showed that insertions of one-, four- or nine-amino-acid residues are possible without destruction of the secondary structure. However, in all models predicted, the glutamic acid Glu469 (in rat) was no longer included in forming the DNA recognition helix. The nine-amino-acid insert seen in teleostean fishes may form a loop that extends to the outside of the protein. Receptor function predicted by homology modelling of this peculiar receptor did not differ from functioning of wild-type receptor (Wickert & Selbig 2002). A Predict Protein software analysis (Cubic, Columbia University, New York, NY, USA) predicts that the nine-amino-acid insert would lengthen the α-helix involved in GRE recognition (Lethimonier et al. 2002). In rainbow trout, it was experimentally shown in vitro that trans-activation after stimulation of rtGR1a (with insert) with dexamethasone or cortisol was increased compared to GR1b (without the insert). These results were obtained by the use of RBCF-1 cells that are derived from goldfish fin tissue and a MSG-cat reporter plasmid (Takeo et al. 1996). In another experiment with rainbow trout GR using Chinese hamster ovary (CHO-K1) cells, it was again shown that the use of a single GRE luciferase reporter plasmid, the GR1a, gave a stronger hormone-independent transactivation than GR1b. However, with a double GRE reporter gene, this difference was abolished (Lethimonier et al. 2002). Moreover, in an in vitro co-transfection of Burtons’ mouthbrooder with a triple GRE in the tat3-luciferase reporter plasmid, GR2b (with nine-amino-acid insert) showed only half maximal activation in response to cortisol compared to GR2a (without insert). These experiments were carried out with CV1b (corticoid receptor-deficient primate renal) cells (Greenwood et al. 2003). Differences between triple, double and single GRE promoters therefore seem to substantially contribute to the results obtained and hamper a good comparison.
Receptor functionality has been shown for the GR splice variant with the nine-amino-acid insert in rainbow trout, Burtons’ mouthbrooder and Japanese flounder (Tokuda et al. 1999). Therefore, it is assumed that other teleost GR proteins, with this particular nine-amino-acid insert (or a slight variation thereof) at exactly the same location in the DNA-binding region, will also be functionally active.
Alternative splicing has been reported in mammalian species as well. Human small lung cell carcinoma exhibits a GC-resistant phenotype, which was thought to be caused by a tri-nucleotide insertion in the GR gene; this insertion translates into an additional amino acid, Arg453 (Ray et al. 1996). This particular human GR splice variant was later named hGRγ (Rivers et al. 1999). Although the insertion found in hGRγ occurs at exactly the same location as the nine-amino-acid insert in the fish GR, it results from a different splice variation process. In the human GRγ, an alternative splice site is used that is located three nucleotides downstream of the original splice site at the 3′ end of coding exon 3. This leads to an insertion of one extra codon in exon 3 (Fig. 3) (Rivers et al. 1999). In teleostean fishes, in contrast, the original splice site of exon 3 is used, and a whole new exon is introduced between exons 3 and 4. The intron in between exons 3 and 4 contains this extra 27 nucleotide exon, which is flanked by common splice sequences AG at the 5′ end and GT at the 3′ end (Fig. 3). Human GRγ is ubiquitously expressed, but only accounts for 5% of all GR transcripts (Rivers et al. 1999, Stevens et al. 2004a)
The addition of an arginine as in the human GRγ or the marmoset monkey (see below) would create a new hydrogen bond and, therefore, allow a new 3-turn to be formed, which could affect affinity of receptor–DNA binding (Wickert & Selbig 2002). Indeed, the human GRγ (Arg453) is half maximally activated compared to wild-type with a similar EC50 in an in vitro co-transfection assay with GR arg453 and pMMTV-luciferase reporter plasmid in COS7 (African green monkey kidney) cells (Ray et al. 1996, Wang & Fondell 2001). Alternative splicing comparable to that in human GRγ is also seen in the cotton-top marmoset (Saguinus oedipus) and thus is of wider occurrence (Brandon et al. 1991).
Collectively, these results show that the specificity of the alternatively spliced GRs for the specific target genes tested seems unaffected, but the differences in affinity of DNA binding or efficiency of dimerisation result in changes in transactivation capacity of the GR1 protein with the nine-amino-acid insert. Results of transactivation comparisons between insert and non-insert forms are likely to be affected by different experimental procedures (Lethimonier et al. 2002, Greenwood et al. 2003). Differences in cell type, reporter plasmid and the use of single or multiple GREs are paramount in the resulting transactivation capacity. It is thus essential to determine how the GR splice variants of different teleostean fish species perform in a standardised transactivation assay.
Variation in the ligand-binding region of fishes and mammals
An important region in the GR gene that can influence cortisol sensitivity is the ligand-binding region. It has been shown that in rat a specific L-x-x-L-L motif (amino acids 547–553) at the N-terminal side of the ligand-binding region is essential for Hsp90 binding. This motif forms a hydrophobic patch that contributes to the stability of the tertiary structure of the ligand-binding region and thereby facilitates steroid binding. Mutations of Leu550 or Leu553 to Serine dramatically reduce steroid binding and biological activity, without altering the binding of Hsp90 (Giannoukos et al. 1999). Interestingly, although Leu553 is conserved throughout all vertebrate species, Leu550 is not found in any of the teleostean species. In fact, all fish GR sequences described to date have a Met at this location and duplicate GR genes also do not differ at this particular location.
Although there is little information about binding affinities of fish GRs, the affinity for dexamethasone of GR1 (Kd 5.5 nM) and GR2 (Kd 3.5 nM) were shown to be similar and comparable to human GR affinity (Kd 4.6) (Ducouret et al. 1995, Ray et al. 1999, Bury et al. 2003). In common carp (Cyprinus carpio) lymphocytes, affinity for dexamethasone was lower (Kd 16 nM) (Weyts et al. 1998c). In this experiment, only one or one class of carp GR was detected, although we now know from mRNA expression data that at least two different GRs, with comparable expression profiles, are present in this fish (E.H Stolte et al. in preparation). The biochemical analysis used to determine affinity, however, has insufficient resolution to confirm this. It is therefore predicted that the binding affinities of the duplicate GRs will be very comparable. Affinity for cortisol of these carp lymphocyte GRs was Kd 3.8 nM (Weyts et al. 1998c). This is comparable to the affinity for cortisol of Coho salmon (Oncorhynchus kisutch) gill GR receptors (Kd 2.2 nM) and Chinook salmon (Oncorhynchus tshawytscha) brain GR receptors (Kd 4.5 nM) (Maule & Schreck 1990, Knoebl et al. 1996).
While the L-x-x-L-L motif in guinea pig (Cavia porcellus) is exactly the same as in rat, it has been shown to exhibit resistance to GCs that results from lower sensitivity for the ligand (Kraft et al. 1979). Using a domain-swap approach, the guinea pig GR ligand-binding region was inserted in the human GR gene, replacing the human ligand-binding region. The chimeric GR gene product showed a decrease in hormone sensitivity of approximately 40-fold (Keightley & Fuller 1994). In further studies, the critical regions and/or residues of the ligand-binding region were identified and targeted for site-directed mutagenesis (Keightley et al. 1998). A similar approach is now used in rainbow trout research (Sturm & Bury 2005). In this case, domain-swap mutants were produced, not between GR genes of different species, but between the two different rainbow trout GR genes, rtGR1 and rtGR2. It was shown that the N-terminal region of rtGR1 had a more potent transactivation function than rtGR2, whereas the opposite was observed for the ligand-binding region. By combinations of rtGR1 and rtGR2 domains, therefore, hypersensitive and hyposensitive recombinant receptors could be obtained (Sturm & Bury 2005).
In this way, both the guinea pig and the rainbow trout provide naturally ‘altered’, but functional GRs (Keightley & Fuller 1994). Domain swaps and site-directed mutagenesis provide the opportunity to investigate the molecular basis of cortisol resistance. GR domain-swap research within one species, however, can only be performed on teleostean fishes.
Perspectives
To fully comprehend the functional implications of the different teleostean GRs, we now search for (immune) cells that differentially express GRs. In situ hybridisation is probably the most powerful tool. To objectively compare the mammalian and the teleostean GRs, they have to be tested in the same in vitro system. This will yield only approximations of fish receptor function as the protein is expressed in a mammalian cell with its mammalian responsive machinery. An important advantage of studies with fish is that we can study receptor function at different biologically relevant ambient temperatures as fishes are poikilotherms. This is especially interesting as it is known from studies with mammals that GR function is critically dependent on Hsps/stress proteins, the expression of which is temperature dependent. However, to perform transactivation studies with a teleostean GR at different temperatures, one would need a cell type that functions properly at low temperature, preferably a fish cell type. The major drawback is that, as yet, no fish cell types/lines comparable to their mammalian counterparts and lacking endogenous corticoid receptors have been described.
Summary and conclusion
Teleostean fishes have duplicate GR proteins where other vertebrates only have one. This likely results from the fish-specific genome duplication event. One of these two GR genes has two different transcripts that are generated by alternative splicing. The most abundant of these two splice variant transcripts has a nine-amino-acid insert in its DNA-binding region. This insertion is the result of transcription of an extra exon, which is located in the intron between exons 3 and 4, a feature that is unique for the GRs of teleostean fishes. Most interestingly, the three different GR forms in fish are differentially expressed in vivo and show different transactivational capacities, but only slightly different affinity for their ligand. As both genes and different splice variants are transcriptionally active, it is suggested that they both play an important and probably different role in the fishes’ physiology.
The splice variation found in the DNA-binding region of the GR gene of teleostean fishes, occurring at exactly the same location as the DNA-binding region splice variation in mammals, but resulting from a different mechanism, is associated with different affinity for DNA binding and different transactivation capacity. Moreover, fish with two different GR genes exhibit intraspecies genetic variation that gives the opportunity to investigate which region of the GR gene and which residues within these regions are involved in ligand affinity and resulting transactivation acitivity. Since transactivation capacity following cortisol binding in different GR gene products does not necessarily correlate with transactivation capacity after dexamethasone binding, it is of paramount importance to check biological activity of both these ligands when a novel GR sequence is discovered. The multiple corticoid receptors in fish should provide a stimulating model to test the hypothesis of ligand exploitation (Thornton 2001). In fish, a major function is ascribed to cortisol, but also significant levels of 11-deoxycortisol and 11-deoxycortico-sterone occur (Sturm et al. 2004) and these ligands bind to GRs (and MRs). The notion of a far more complex signalling by related steroids and a subtle repertoire of adjusted/adjusting receptors may eventually greatly improve our understanding of fish physiology. The same notion may give rise to a better understanding of variation in GC sensitivity in man.
Putative phosphorylation motifs in the glucocorticoid receptors of different speciesa
| Mammals | Chicken | African clawed frog | Teleostean fishes | Concensus mammals | Alternative concensus | Kinases mammals | Alternative kinases | |
|---|---|---|---|---|---|---|---|---|
| aDepicted are the residues phosphorylated in mouse GR and a score if these residues are conserved in mammals, chicken, African clawed frog or teleostean fishes. The consensus motif for a particular phosphorylation site is mentioned for mammalian and teleostean fishes separately. The putative protein kinases expected to target these motifs are also shown separately and are based on mammalian literature. This table was constructed after analysis of a multiple sequence alignment of human (Homo sapiens) GRα, P04150; mouse (Mus musculus) GR, P06537; rat (Rattus norvegicus) GR, NP_036708; guinea pig (Cavia porcellus) GR, P49115; pig (Sus scrofa) GR, AY779185; cow (Bos taurus) GR, AY238475; chicken (Gallus gallus) GR, Q8JHA4; African clawed frog (Xenopus laevis) GR, P49844; zebrafish (Danio rerio) GR, ENSDARESTT00000005443 (Chr.14); rainbow trout (Oncorhynchus mykiss) GR1, P49843, R2, AY4953720; Burton’s mouthbrooder (Haplochromis burtoni) GR1, AF263738, GR2b, AF263740; Japanese flounder (Paralichthys olivaceus) GR, O73673; European sea bass (Dicentrarchus labrax) GR, AY549305; brown trout (Salmo trutta) GR, AY863149; fugu (Takifugu rubripes) GR1, GENSCAN00000003615 (scaffold 1264) and GENSCAN00000029451 (scaffold 4328), GR2, SINFRUG00000143550 (scaffold 59); tetraodon (Tetraodon nigroviridis) GR1, GIDT00024792001 (Chr. 7), GR2, GSTENG00017027001 (Chr. 1) (adapted from Ismaili & Garabedian 2004) | ||||||||
| Mouse | ||||||||
| S122 | Not in guinea pig | Yes | Yes | No | S/T(P)-x-x-E/D | CKII | ||
| S150 | Not in guinea pig | No | No | No | None | |||
| T171 | Only in rat | No | No | No | Non-polar-x-S/T(P) | MAPK/GSK3 | ||
| S212 | All | Yes | Different | Different | S/T(P)-P-x-R/K | CDK | ||
| S220 | Not in pig, not in cow | Yes | No | No | S/T(P)-P-x-R/K | S/T(P)-x-x-E/d | CDK | CKII |
| S234 | All | Yes | Yes | No | Non-polar-x-S/T(P) | MAPK | ||
| S315 | Not in human | Yes | Yes | Yes | None | |||
Percentage amino acid identity of the different receptor regions of several members of the nuclear receptor superfamily of vertebrate species. Length of the boxes represents lengths of particular regions. MR, mineralocorticoid receptor; AR, androgen receptor. AB, N-terminal region; C, DNA-binding region; D, hinge region; E, ligand-binding region.
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06703
Phylogenetic tree; comparison of the amino acid sequences of the vertebrate glucocorticoid receptors (GRs) and corticoid receptor (CR). This tree was generated with MEGA version 3.1 software (Institute of Molecular Evolutionary Genetics, Pennsylvania State University, PA, USA) using the neighbour-joining method. Reliability of this tree was assessed by bootstrapping using 1000 bootstrap replications (values in percentage are indicated at branch nodes). Mineralocorticoid receptor (MR) sequences were used as outgroup. Human (Homo sapiens) GRα, P04150; human GRβ, NP_001018661; human GRγ, NP_001019265; mouse (Mus musculus) GR, P06537; rat (Rattus norvegicus) GR, NP_036708; guinea pig (Cavia porcellus) GR, P49115; pig (Sus scrofa) GR, AY779185; cow (Bos taurus) GR, AY238475; chicken (Gallus gallus) GR, Q8JHA4; African clawed frog (Xenopus laevis) GR, P49844; western clawed frog (Xenopus tropicalis) GR, NM_001016967; zebrafish (Danio rerio) GR, ENSDARESTT00000005443 (Chr.14); rainbow trout (Oncorhynchus mykiss) GR1, P49843, GR2, AY4953720; Burton’s mouthbrooder (Haplochromis burtoni) GR1, AF263738, GR2a, AF263739, GR2b, AF263740; Japanese flounder (Paralichthysolivaceus) GR, O73673; European sea bass (Dicentrarchus labrax) GR, AY549305;browntrout (Salmo trutta) GR, AY863149; sea lamprey (Petromyzon marinus) (ancestral) CR, AY028457; human MR, M16801; mouse MR, XP_356093; rainbow trout MR, AY495584.
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06703
Gene structure of vertebrate GR receptors. Exon lengths are shown in bold and intron lengths in normal font. Splice sequences flanking the exons and nucleotide sequence preceding start codon are shown above blocks representing exons. Untranslated (parts of) exons are depicted in light grey. Elongation of exon 3 by alternative splicing in human GRγ is shown in dark grey. Extra exon, only present in splice variants in fish GR, is depicted in black. Exon 8 in Fugu GR1 could not be found in the Ensembl database. Human (Homo sapiens) GR, ENSG00000113580; mouse (Mus musculus) GR, ENSMUSG00000024431; chicken (Gallus gallus) GR, ENSGALG00000007394; western clawed frog (Xenopus tropicalis) GR, ENSXETT00000003968; fugu (Takifugu rubripes) GR1, GENSCAN00000003615 (scaffold 1264) and GENSCAN00000029451 (scaffold 4328), GR2, SINFRUG00000143550 (scaffold 59); tetraodon (Tetraodon nigroviridis) GR1, GIDT00024792001 (Chr. 7); GR2, GSTENG00017027001 (Chr. 1).
Citation: Journal of Endocrinology 190, 1; 10.1677/joe.1.06703
We would like to acknowledge Sander Kranenbarg and Mark Huising for their critical reading of the manuscript, suggestions and advice. The authors declare that there is no conflict of interest that would prejudice the impartiality of this scientific work.
References
Boers M 2004 Glucocorticoids in rheumatoid arthritis: a senescent research agenda on the brink of rejuvenation? Best Practice and Research Clinical Rheumatology 18 21–29.
Brandon DD, Markwick AJ, Flores M, Dixon K, Albertson BD & Loriaux DL 1991 Genetic variation of the glucocorticoid receptor from a steroid-resistant primate. Journal of Molecular Endocrinology 7 89–96.
Bray PJ & Cotton RG 2003 Variations of the human glucocorticoid receptor gene (NR3C1): pathological and in vitro mutations and polymorphisms. Human Mutation 21 557–568.
Bury NR, Sturm A, Le Rouzic P, Lethimonier C, Ducouret B, Guiguen Y, Robinson-Rechavi M, Laudet V, Rafestin-Oblin ME & Prunet P 2003 Evidence for two distinct functional glucocorticoid receptors in teleost fish. Journal of Molecular Endocrinology 31 141–156.
Charmandari E, Tsigos C & Chrousos G 2005 Endocrinology of the stress response. Annual Review of Physiology 67 259–284.
Dean DB, Whitlow ZW & Borski RJ 2003 Glucocorticoid receptor upregulation during seawater adaptation in a euryhaline teleost, the tilapia (Oreochromis mossambicus). General and Comparative Endocrinology 132 112–118.
DeRijk RH, Schaaf M & de Kloet ER 2002 Glucocorticoid receptor variants: clinical implications. Journal of Steroid Biochemistry and Molecular Biology 81 103–122.
Dhabhar FS, Miller AH, McEwen BS & Spencer RL 1996 Stress-induced changes in blood leukocyte distribution, Role of adrenal steroid hormones. Journal of Immunology 157 1638–1644.
Ducouret B, Tujague M, Ashraf J, Mouchel N, Servel N, Valotaire Y & Thompson EB 1995 Cloning of a teleost fish glucocorticoid receptor shows that it contains a deoxyribonucleic acid-binding domain different from that of mammals. Endocrinology 136 3774–3783.
Elenkov IJ 2004 Glucocorticoids and the Th1/Th2 balance. Annals of the New York Academy of Sciences 1024 138–146.
Evans RM 2005 The nuclear receptor superfamily: a rosetta stone for physiology. Molecular Endocrinology 19 1429–1438.
Franchimont D 2004 Overview of the actions of glucocorticoids on the immune response: a good model to characterize new pathways of immunosuppression for new treatment strategies. Annals of the New York Academy of Sciences 1024 124–137.
Fuller PJ 1991 The steroid receptor superfamily: mechanisms of diversity. FASEB Journal 5 3092–3099.
Geng CD, Pedersen KB, Nunez BS & Vedeckis WV 2005 Human glucocorticoid receptor alpha transcript splice variants with exon 2 deletions: evidence for tissue- and cell type-specific functions. Biochemistry 44 7395–7405.
Giannoukos G, Silverstein AM, Pratt WB & Simons SS Jr. 1999 The seven amino acids (547–553) of rat glucocorticoid receptor required for steroid and hsp90 binding contain a functionally independent LXXLL motif that is critical for steroid binding. Journal of Biological Chemistry 274 36527–36536.
Giguere V, Hollenberg SM, Rosenfeld MG & Evans RM 1986 Functional domains of the human glucocorticoid receptor. Cell 46 645–652.
Greenwood AK, Butler PC, White RB, DeMarco U, Pearce D & Fernald RD 2003 Multiple corticosteroid receptors in a teleost fish: distinct sequences, expression patterns, and transcriptional activities. Endocrinology 144 4226–4236.
Hard T, Kellenbach E, Boelens R, Maler BA, Dahlman K, Freedman LP, Carlstedt-Duke J, Yamamoto KR, Gustafsson JA & Kaptein R 1990 Solution structure of the glucocorticoid receptor DNA-binding domain (in Reports). Science 249 157–160.
Hearing SD, Norman M, Smyth C, Foy C & Dayan CM 1999 Wide variation in lymphocyte steroid sensitivity among healthy human volunteers. Journal of Clinical Endocrinology and Metabolism 84 4149–4154.
Hedges SB & Kumar S 2002 Genomics. Vertebrate genomes compared. Science 297 1283–1285.
Hennessy MB, Davis HN, Williams MT, Mellott C & Douglas CW 1997 Plasma cortisol levels of dogs at a county animal shelter. Physiology and Behavior 62 485–490.
Hoegg S, Brinkmann H, Taylor JS & Meyer A 2004 Phylogenetic timing of the fish-specific genome duplication correlates with the diversification of teleost fish. Journal of Molecular Evolution 59 190–203.
Hollenberg SM, Weinberger C, Ong ES, Cerelli G, Oro A, Lebo R, Thompson EB, Rosenfeld MG & Evans RM 1985 Primary structure and expression of a functional human glucocorticoid receptor cDNA. Nature 318 635–641.
Huising MO, Guichelaar T, Hoek C, Verburg-van Kemenade BM, Flik G, Savelkoul HF & Rombout JH 2003 Increased efficacy of immersion vaccination in fish with hyperosmotic pretreatment. Vaccine 21 4178–4193.
Ismaili N & Garabedian MJ 2004 Modulation of glucocorticoid receptor function via phosphorylation. Annals of the New York Academy of Sciences 1024 86–101.
Kanelakis KC, Shewach DS & Pratt WB 2002 Nucleotide binding states of hsp70 and hsp90 during sequential steps in the process of glucocorticoid receptor.hsp90 heterocomplex assembly. Journal of Biological Chemistry 277 33698–33703.
Keightley MC & Fuller PJ 1994 Unique sequences in the guinea pig glucocorticoid receptor induce constitutive transactivation and decrease steroid sensitivity. Molecular Endocrinology 8 431–439.
Keightley MC, Curtis AJ, Chu S & Fuller PJ 1998 Structural determinants of cortisol resistance in the guinea pig glucocorticoid receptor. Endocrinology 139 2479–2485.
Knoebl I, Fitzpatrick MS & Schreck CB 1996 Characterization of a glucocorticoid receptor in the brains of chinook salmon, Oncorhynchus tshawytscha. General and Comparative Endocrinology 101 195–204.
Kraft N, Hodgson AJ & Funder JW 1979 Glucocorticoid receptor and effector mechanisms: a comparison of the corticosensitive mouse with the corticoresistant guinea pig. Endocrinology 104 344–349.
Kumar R & Thompson EB 1999 The structure of the nuclear hormone receptors. Steroids 64 310–319.
Kumar R & Thompson EB 2005 Gene regulation by the glucocorticoid receptor: structure:function relationship. Journal of Steroid Biochemistry and Molecular Biology 94 383–394.
Kumar S, Tamura K & Nei M 2004 MEGA3: integrated software for molecular evolutionary genetics analysis and sequence alignment. Briefings in Bioinformatics 5 150–163.
Lamberts SW, Huizenga AT, de Lange P, de Jong FH & Koper JW 1996 Clinical aspects of glucocorticoid sensitivity. Steroids 61 157–160.
Lethimonier C, Tujague M, Kern L & Ducouret B 2002 Peptide insertion in the DNA-binding domain of fish glucocorticoid receptor is encoded by an additional exon and confers particular functional properties. Molecular and Cellular Endocrinology 194 107–116.
Luisi BF, Xu WX, Otwinowski Z, Freedman LP, Yamamoto KR & Sigler PB 1991 Crystallographic analysis of the interaction of the glucocorticoid receptor with DNA. Nature 352 497–505.
Mangelsdorf DJ, Thummel C, Beato M, Herrlich P, Schutz G, Umesono K, Blumberg B, Kastner P, Mark M et al.1995 The nuclear receptor superfamily: the second decade. Cell 83 835–839.
Maule AG & Schreck CB 1990 Glucocorticoid receptors in leukocytes and gill of juvenile coho salmon (Oncorhynchus kisutch). General and Comparative Endocrinology 77 448–455.
Metz JR, van den Burg EH, Bonga SE & Flik G 2003 Regulation of branchial Na(+)/K(+)-ATPase in common carp Cyprinus carpio L. acclimated to different temperatures. Journal of Experimental Biology 206 2273–2280.
Morishima Y, Kanelakis KC, Silverstein AM, Dittmar KD, Estrada L & Pratt WB 2000a The Hsp organizer protein hop enhances the rate of but is not essential for glucocorticoid receptor folding by the multiprotein Hsp90-based chaperone system. Journal of Biological Chemistry 275 6894–6900.
Morishima Y, Murphy PJ, Li DP, Sanchez ER & Pratt WB 2000b Stepwise assembly of a glucocorticoid receptor.hsp90 heterocomplex resolves two sequential ATP-dependent events involving first hsp70 and then hsp90 in opening of the steroid binding pocket. Journal of Biological Chemistry 275 18054–18060.
Nelson JS. 1994. Fishes of the world. New York, NY, USA: John Wiley & Sons Inc.
Oakley RH, Jewell CM, Yudt MR, Bofetiado DM & Cidlowski JA 1999 The dominant negative activity of the human glucocorticoid receptor beta isoform, Specificity and mechanisms of action. Journal of Biological Chemistry 274 27857–27866.
Piekarzewska AB, Rosochacki SJ & Sender G 2000 The effect of acute restraint stress on regional brain neurotransmitter levels in stress-susceptible pietrain pigs. Journal of Veterinary Medicine. A, Physiology, Pathology, Clinical Medicine 47 257–269.
Pratt WB & Toft DO 1997 Steroid receptor interactions with heat shock protein and immunophilin chaperones. Endocrine Reviews 18 306–360.
Pratt WB & Toft DO 2003 Regulation of signaling protein function and trafficking by the hsp90/hsp70-based chaperone machinery. Experimental Biology and Medicine (Maywood) 228 111–133.
Ray DW, Davis JR, White A & Clark AJ 1996 Glucocorticoid receptor structure and function in glucocorticoid-resistant small cell lung carcinoma cells. Cancer Research 56 3276–3280.
Ray DW, Suen CS, Brass A, Soden J & White A 1999 Structure/function of the human glucocorticoid receptor: tyrosine 735 is important for transactivation. Molecular Endocrinology 13 1855–1863.
Rivers C, Levy A, Hancock J, Lightman S & Norman M 1999 Insertion of an amino acid in the DNA-binding domain of the glucocorticoid receptor as a result of alternative splicing. Journal of Clinical Endocrinology and Metabolism 84 4283–4286.
Saitou N & Nei M 1987 The neighbor-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4 406–425.
Sapolsky RM, Romero LM & Munck AU 2000 How do glucocorticoids influence stress responses?, Integrating permissive, suppressive, stimulatory, and preparative actions. Endocrine Reviews 21 55–89.
Schoneveld OJ, Gaemers IC & Lamers WH 2004 Mechanisms of glucocorticoid signalling. Biochimica et Biophysica Acta 1680 114–128.
Stevens A, Donn R & Ray D 2004a Regulation of glucocorticoid receptor gamma (GRgamma) by glucocorticoid receptor haplotype and glucocorticoid. Clinical Endocrinology 61 327–331.
Stevens A, Ray DW, Zeggini E, John S, Richards HL, Griffiths CE & Donn R 2004b Glucocorticoid sensitivity is determined by a specific glucocorticoid receptor haplotype. Journal of Clinical Endocrinology and Metabolism 89 892–897.
Sturm A & Bury NR 2005 Duplicate teleost glucocorticoid receptors are not functionally redundant: relationship between structural and functional differences. Comparative Biochemistry and Physiology 141A S208.
Sturm A, Bury N, Dengreville L, Fagart J, Flouriot G, Rafestin-Oblin ME & Prunet P 2004 11-deoxycorticosterone is a potent agonist of the rainbow trout (Oncorhynchus mykiss) mineralocorticoid receptor. Endocrinology 146 47–55.
Takeo J, Hata J, Segawa C, Toyohara H & Yamashita S 1996 Fish glucocorticoid receptor with splicing variants in the DNA binding domain. FEBS Letters 389 244–248.
Thornton JW 2001 Evolution of vertebrate steroid receptors from an ancestral estrogen receptor by ligand exploitation and serial genome expansions. PNAS 98 5671–5676.
Todo T & Nagahama 1998 Structure and function of fish steroid hormone receptors. Kaiyo Monthly 30 96–102.
Tokuda Y, Touhata K, Kinoshita M, Toyohara H, Sakaguchi M, Yokoyame Y, Ichikawa T & Yamashita S 1999 Sequence and expression of a cDNA encoding Japanese flounder glucocorticoid receptor. Fisheries Science 65 466–471.
Umesono K & Evans RM 1989 Determinants of target gene specificity for steroid/thyroid hormone receptors. Cell 57 1139–1146.
Van Cauter E, Leproult R & Kupfer DJ 1996 Effects of gender and age on the levels and circadian rhythmicity of plasma cortisol. Journal of Clinical Endocrinology and Metabolism 81 2468–2473.
van Rossum EF, Koper JW, Huizenga NA, Uitterlinden AG, Janssen JA, Brinkmann AO, Grobbee DE, de Jong FH, van Duyn CM, Pols HA et al.2002 A polymorphism in the glucocorticoid receptor gene, which decreases sensitivity to glucocorticoids in vivo, is associated with low insulin and cholesterol levels. Diabetes 51 3128–3134.
Volff JN 2005 Genome evolution and biodiversity in teleost fish. Heredity 94 280–294.
Walsh GM 2005 Novel therapies for asthma—advances and problems. Current Pharmaceutical Design 11 3027–3038.
Wang Q & Fondell JD 2001 Generation of a mammalian cell line stably expressing a tetracyclin-regulated epitope-tagged human androgen receptor: implications for steroid hormone receptor research. Analytical Biochemistry 289 217–230.
Webster JC, Jewell CM, Bodwell JE, Munck A, Sar M & Cidlowski JA 1997 Mouse glucocorticoid receptor phosphorylation status influences multiple functions of the receptor protein. Journal of Biological Chemistry 272 9287–9293.
Wendelaar Bonga SE 1997 The stress response in fish. Physiological Reviews 77 591–625.
Weyts FA, Flik G, Rombout JH & Verburg-van Kemenade BM 1998a Cortisol induces apoptosis in activated B cells, not in other lymphoid cells of the common carp, Cyprinus carpio L. Developmental and Comparative Immunology 22 551–562.
Weyts FA, Flik G & Verburg-van Kemenade BM 1998b Cortisol inhibits apoptosis in carp neutrophilic granulocytes. Developmental and Comparative Immunology 22 563–572.
Weyts FA, Verburg-van Kemenade BM & Flik G 1998c Characterisation of glucocorticoid receptors in peripheral blood leukocytes of carp, Cyprinus carpio L. General and Comparative Endocrinology 111 1–8.
Wickert L & Selbig J 2002 Structural analysis of the DNA-binding domain of alternatively spliced steroid receptors. Journal of Endocrinology 173 429–436.
