The sea lamprey (Petromyzon marinus), belongs to an ancient group of jawless vertebrates known as cyclostomes, which last shared a common ancestor with vertebrates about 550 million years ago (1–3). As an outgroup to the jawed vertebrates, lampreys are important for studying early events in the evolution of vertebrates (1–8). Lampreys and hagfish, the other extant cyclostome lineage, contain a corticoid receptor (CR), which is the common ancestor to both the mineralocorticoid receptor (MR) and the glucocorticoid receptor (GR) (4, 9). The MR and GR first appear as separate steroid receptors in sharks and chimeras (4, 9–13). The CR, MR and GR belong to the nuclear receptor family of transcription factors, which also contains the progesterone receptor, estrogen receptor and androgen receptor (14–18).
Early studies from Thornton’s laboratory provided important insights into corticosteroid activation of lamprey CR (4). These studies reported that several corticosteroids (Fig. 1), including aldosterone, cortisol, corticosterone,11-deoxycorticosterone and 11-deoxycortisol activated lamprey CR. Although aldosterone, the physiological mineralocorticoid in humans and other terrestrial vertebrates (13, 19–25), is the strongest activator of lamprey CR (4), neither lampreys nor hagfish synthesize aldosterone (4). Later studies with lamprey revealed that 11-deoxycortisol and 11-deoxycorticosterone (Fig. 1) are the circulating corticosteroids in sea lamprey (26–28).
Until recently, due to complexities in sequencing and assembly of the lamprey’s highly repetitive and GC rich genome (3, 33, 34), DNA encoding 344 amino acids at the amino terminus of lamprey CR was present on a separate contiguous sequence that was not joined with the rest of the lamprey CR sequence and therefore not retrieved with BLAST searches of GenBank. The recent sequencing of the sea lamprey germline genome (35) provided contiguous DNA for two CR isoforms, CR1 and CR2, that encode the previously unassembled 344 amino acids at the amino terminus. The sequences of lamprey CR1 and CR2 reveal that like other nuclear receptors, lamprey CR is a multi-domain protein consisting of an N-terminal domain (NTD), a central DNA-binding domain (DBD), a hinge domain and a C-terminal ligand-binding domain (LBD) (4, 14, 36, 37) (Fig. 2). The DBD and LBD in lamprey CR are conserved in vertebrate MRs and GRs (Fig. 2) (9, 38), while their sequences in the NTD and hinge domains have diverged (Fig. 2). The only sequence difference between lamprey CR1 and CR2 is a four amino acid insertion in the DBD in CR1 is not present in the DBD in CR2 or in other GRs and MRs.
With the full sequences of the two lamprey CR isoforms in hand, we investigated four questions about this receptor that shares common ancestry with gnathostome GRs and MRs (9, 13, 38). First, what is the response of lamprey CR to a panel of physiological corticosteroids (aldosterone, cortisol, corticosterone, 11-deoxycorticosterone and 11-deoxycortisol) (Fig. 1) for vertebrate GRs and MRs, and, second, how does this response compare to the response to these steroids by elephant shark GR and MR (12, 39)? Comparison of corticosteroid activation of lamprey CR with elephant shark GR and MR can provide insights into the evolution of corticosteroid specificity in the GR and MR.
Third, what is the role, if any, of the NTD in transcriptional activation of lamprey CR? The NTD on human GR contains an activation function 1 (AF1) domain that is very important in GR activation by steroids (40–48). The NTD on human MR also contains an AF1 domain, although it is a much weaker activator of the MR (41, 49–51) compared to the AF1 on human GR. The NTD on elephant shark GR also contains a strong AF1 (12). The low sequence identity (less than 20%) between the NTD in lamprey CR and in elephant shark MR and GR raises the question: Is there AF1 activity in lamprey CR or did a strong AF1 evolve after CR duplication and divergence to form vertebrate GR and MR?
Fourth, what is the role of the MMTV (52, 53) and TAT3 (54) promoters in transcriptional activation of lamprey CR? That is, does lamprey CR have different responses to corticosteroids in cells co-transfected with MMTV or TAT3 promoters, as we found for cells co-transfected with either MMTV or TAT3 and either elephant shark MR (55) or GR (12). Comparison of activation of lamprey CR in cells with either MMTV or TAT3 with that of elephant shark GR and MR could indicate whether lamprey CR was closer to the GR or to the MR, and thus shed light on the evolution of the GR and MR from their common ancestor.
In this report, we used two metrics for evaluating activation by steroids of lamprey CR and other receptors. The first metric was the half maximal response (EC50) to various steroids, and the second metric was the strength (fold-activation) of transcription. Combined, these two metrics provide insights into the relevance of a steroid as a physiological ligand for the CR.
Our initial experiments focused on lamprey CR1 because RNA-Seq analysis indicates that CR1 is more highly expressed than CR2 (greater than 99%) in lamprey tissue. However, CR1 and CR2 have similar EC50s for corticosteroids. We find that the EC50s for activation by 11-deoxycorticosterone and 11-deoxycortisol, the two circulating corticosteroids in lampreys (26–28), of full-length lamprey CR1 in HEK293 cells with MMTV were 0.16 nM and 1.5 nM, respectively. These are the lowest EC50s for CR1 among the corticosteroids that we studied. Aldosterone, cortisol and corticosterone had EC50s from 2 nM to 9.9 nM for activation of CR1 in cells with MMTV. For truncated CR1, which lacks the NTD, the EC50 of 11-deoxycorticosterone for lamprey CR1 in HEK293 cells with MMTV was 0.4 nM, while EC50s of the other corticosteroids for lamprey CR1 increased from 3 to 6-fold.
Comparison of corticosteroid activation of CR with that of elephant shark MR and GR reveals that full-length and truncated elephant shark MR and the CR have similar EC50s for 11-deoxycorticosterone, 11-deoxycortisol and other corticosteroids, in contrast to full-length and truncated elephant shark GR, which has a negligible response to 11-deoxycortisol and weak responses to other corticosteroids (12), indicating that, based on steroid specificity, elephant shark MR is a closer to CR1 and CR2 than is elephant shark GR, which has diverged more from its common ancestor with the MR.
Interestingly, we found differences between the effect of the MMTV and TAT3 promoters on fold-activation of transcription of full-length CR and of truncated CR. Unexpectedly, fold-activation of full-length CR1 to corticosteroids was about 3 to 4-fold higher in cells with the MMTV promoter than in cells transfected with the TAT3 promoter. Removal of the NTD on CR1 decreased fold-activation by corticosteroids of truncated CR1 in cells with MMTV by about 70% indicating that there is an activation function in the NTD. In contrast, compared to full-length CR1, transcriptional activation by corticosteroids of truncated CR1 in cells with TAT3 increased by about 6-fold, indicating that the CR1 NTD represses steroid activation in the presence of TAT3. These data indicate that allosteric regulation by the NTD evolved before the evolution of the GR and MR from their common ancestor in cartilaginous fishes, with divergence of specificity for various corticosteroids evolving in elephant shark GR and MR (12).