Functional domains on elephant shark GR and other vertebrate GRs.
In Figure 2, we compare the functional domains of elephant shark GR to corresponding domains in selected vertebrate GRs (human, chicken, Xenopus, zebrafish) and MRs (human, zebrafish and elephant shark). Elephant shark GR and human GR have 98% and 67% identity in DBD and LBD, respectively. Elephant shark GR and MR have 95% and 61% identity in DBD and LBD, respectively. This strong conservation of the DBD and LBD contrasts with the low sequence identity between the NTD in elephant shark GR and the NTD in other GRs and MRs. The NTD of elephant shark GR has only 21% sequence identity with elephant shark MR.
Effect of corticosteroids and progesterone on transcriptional activation of full-length elephant shark GR and MR.
We screened a panel of steroids (cortisol, corticosterone, 11-deoxycorticosterone, aldosterone, progesterone, 19norProgesterone) at 10-8 M for activation of full-length elephant shark GR (Figure 3A). For comparison, activation of full-length elephant shark MR by these steroids is shown in Figure 3B. Cortisol, corticosterone, 11-deoxycorticosterone, and aldosterone activated full-length elephant shark GR (Figure 3A) and MR (Figure 3B). However, fold-activation of full-length elephant shark GR by corticosteroids at 10-8 M was over 10-fold higher than for full-length elephant shark MR. For example, 10-8 M corticosterone activated full-length GR by 225-fold and full-length MR by 8-fold. Interestingly, neither progesterone nor 19norProgesterone activated elephant shark GR, while these steroids activated elephant shark MR, as previously reported (41).
Effect of corticosteroids and progesterone on transcriptional activation of truncated elephant shark GR and MR.
We investigated the role of the NTD in the response to corticosteroids and progestins at 10-8 M, of truncated elephant shark GR and MR, in which the NTD was deleted (Figure 3C, D). Figure 3C shows that activation of truncated elephant shark GR by 10-8 M corticosterone decreased by over 90%, compared to full-length elephant shark GR (Figure 3A), and there was no significant activation of truncated GR by 10-8 M of either cortisol or 11-deoxycorticosterone, both of which activated full-length GR. This indicates that NTD in elephant shark GR contains an AF1 domain. Progesterone did not activate truncated elephant shark GR. In contrast, at 10-8 M, corticosteroids and progesterone had similar levels of activation of truncated elephant shark MR (Figure 3D) and full-length MR (Figure 3B), indicating that there is a major difference between NTD activation of transcription of elephant shark GR and MR.
To investigate the effect of the elephant shark DBD on activation of elephant shark GR and MR, we replaced the DBD with GAL4 DBD, which has been used for analysis of transcription of the GR and MR (27,28,43,46). As shown in Figure 3E, there was only a low level of activation by 10 nM corticosterone and aldosterone of GAL4 DBD GR hinge-LBD, and no activation by 11-deoxycorticosterone and cortisol. In contrast, GAL4 DBD MR LBD is activated by aldosterone, 11-deoxycorticosterone and other corticosteroids (41).
Concentration-dependent activation by corticosteroids and progesterone of full-length and truncated elephant shark GR and MR.
To gain a quantitative measure of NTD activation of elephant shark GR, we determined the concentration dependence of transcriptional activation by corticosteroids and progesterone of full-length and truncated elephant shark GR (Figure 4A, B) and for comparison of elephant shark MR (Figure 4C, D). This data was used to calculate a half maximal response (EC50) for steroid activation of elephant shark GR and MR (Tables 1 and 2). Corticosterone and cortisol, two physiological glucocorticoids in mammals, had EC50s of 7.9 nM and 35 nM, respectively, for full-length elephant shark GR (Table 1). For comparison, the EC50s of corticosterone and cortisol are 0.61 nM and 1.6 nM, respectively, for full-length elephant shark MR (Table 2) (41). However, aldosterone and 11-deoxycorticosterone, two physiological mineralocorticoids in mammals, had EC50s of 24 nM and 28 nM, respectively, for elephant shark GR. Activation by these steroids of elephant shark GR indicates that it retains some properties of its MR paralog. For comparison, the EC50s of aldosterone and 11-deoxycorticosterone are 0.14 nM and 0.1 nM, respectively, for full-length elephant shark MR (Table 2) (41).
The EC50 of corticosterone for truncated elephant shark GR was 9.4 nM. The EC50s of cortisol and aldosterone for truncated elephant shark GR were 66 nM and 65 nM respectively. In contrast, the EC50s of corticosterone and aldosterone for GAL4-DBD+GR-hinge-LBD were 25 nM and 71 nM, respectively, and cortisol and 11-deoxycorticosterone did not activate GAL4-DBD+GR-hinge-LBD. The higher EC50 for corticosterone and the loss of activation by cortisol and 11-deoxycorticosterone, when elephant shark DBD is replaced with GAL4-DBD indicates that the DBD is important in transcriptional activation of elephant shark GR.
All four corticosteroids, progesterone and 19norProgesterone had EC50s for truncated MR (DBD+LBD) that were either similar to or lower than that of full-length MR (Table 2). EC50s for Gal4-DBD+MR-hinge-LBD were either similar to or lower than that of full-length MR.
The dependence of the level of cortisol and 11-deoxycorticosterone activation on the NTD in elephant shark GR (Figures 4A, C) is similar to that of human GR, in which deletion of the NTD reduces cortisol and 11-deoxycorticosterone fold-activation by over 90% (19,27). The diminished activation of GAL4-DBD-elephant shark GR-hinge-LBD (Figure 4E) indicates that the DBD also is important in fold-activation of elephant shark GR. These results indicate that allosteric signaling between the NTD and DBD-LBD in elephant shark GR is critical for its response to corticosteroids. This contrasts to elephant shark MR, in which truncated MR and full-length elephant shark MR have similar levels of activation by corticosteroids and progesterone (Figures 4B, D). Together this indicates that an activation function in the GR NTD and higher EC50s for some corticosteroids that also activated elephant shark MR evolved early in the divergence of the GR from its MR kin.
Table 1. EC50 values for steroid activation of full-length and truncated elephant shark GR
|
Cortisol
|
Corticosterone
|
Aldo
|
DOC
|
Prog
|
19norProg
|
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
Elephant shark GR-full
|
35 nM
|
7.9 nM
|
24nM
|
28 nM
|
-
|
-
|
95% confidence intervals
|
27-46 nM
|
6.1-10.0 nM
|
19.0-31 nM
|
19-42 nM
|
-
|
-
|
% Relative Induction
|
88 %
|
100 %
|
92 %
|
53 %
|
-
|
-
|
Elephant shark GR-DBD-hinge-LBD
|
66 nM
|
9.4 nM
|
65 nM
|
17 nM
|
-
|
-
|
95% confidence intervals
|
55-80 nM
|
7.5-12 nM
|
44-95 nM
|
8.6-34 nM
|
-
|
-
|
% Relative Induction
|
112 %
|
100 %
|
109 %
|
30 %
|
-
|
-
|
Elephant shark Gal4-DBD-GR-hinge-LBD
|
-
|
25 nM
|
71 nM
|
-
|
-
|
-
|
95% confidence intervals
|
-
|
14-45 nM
|
31-165 nM
|
-
|
-
|
-
|
% Relative Induction
|
-
|
100 %
|
81 %
|
-
|
-
|
-
|
Skate GR
[Gal4-DBD-GR-hinge-LBD]#
|
139 nM
|
58 nM
|
11 nM
|
306 nM
|
N.D.
|
N.D.
|
Elephant shark Chimera
MR-NTD+GR-DBD-hinge-LBD
|
93 nM
|
13 nM
|
43 nM
|
60 nM
|
-
|
-
|
95% confidence intervals
|
58-150 nM
|
8.0-21 nM
|
28-67 nM
|
28-128 nM
|
-
|
-
|
% Relative Induction
|
108 %
|
100 %
|
88 %
|
20 %
|
-
|
-
|
# Values obtained from Mol Biol Evol. 12, 2643-2652 (2008)
Percent relative induction for GR is compared to maximal response to corticosterone.
- = no activation. N.D. = not determined
Aldo = aldosterone, DOC =11-deoxycorticosterone, Prog = progesterone, 19norProg = 19norProgesterone
Table 2. EC50 values for steroid activation of full-length and truncated elephant shark MR
|
Cortisol
|
Corticosterone
|
Aldo
|
DOC
|
Prog
|
19norProg
|
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
EC50 (M)
|
Elephant shark MR-full
|
1.6 nM
|
0.61 nM
|
0.14 nM
|
0.1 nM
|
0.45 nM
|
0.11 nM
|
95% confidence intervals
|
0.92-2.8 nM
|
0.34-1.1 nM
|
0.069-0.26 nM
|
0.54-0.2 nM
|
0.21-0.94 nM
|
0.06-0.22 nM
|
% Relative Induction
|
118 %
|
88 %
|
100 %
|
86 %
|
38 %
|
79 %
|
Elephant shark MR-DBD-hinge-LBD
|
1.1nM
|
0.58 nM
|
0.26 nM
|
0.1 nM
|
0.63 nM
|
0.2 nM
|
95% confidence intervals
|
0.67-1.7 nM
|
0.39-0.86 nM
|
0.15-0.45 nM
|
0.05-0.16 nM
|
0.44-0.9 nM
|
0.13-0.31 nM
|
% Relative Induction
|
117 %
|
111 %
|
100 %
|
77 %
|
78 %
|
98 %
|
Elephant shark Gal4-DBD-MR-hinge-LBD*
|
0.19 nM
|
0.1 nM
|
0.04 nM
|
0.02 nM
|
0.48 nM
|
0.02 nM
|
95% confidence intervals
|
0.08-0.2 nM
|
0.08-0.1 nM
|
0.02-0.05 nM
|
0.01-0.03 nM
|
0.39-0.57 nM
|
0.01-0.02 nM
|
% Relative Induction
|
79 %
|
90 %
|
100 %
|
81 %
|
40 %
|
98 %
|
Skate MR
[Gal4-DBD-MR-hinge-LBD]#
|
1.04 nM
|
0.09 nM
|
0.07 nM
|
0.03 nM
|
N.D.
|
N.D.
|
Elephant shark Chimera
GR-NTD+MR-DBD-hinge-LBD
|
0.3 nM
|
0.1 nM
|
0.05 nM
|
0.03 nM
|
0.2 nM
|
0.04 nM
|
95% confidence intervals
|
0.2-0.44 nM
|
0.09-0.16 nM
|
0.04-0.07 nM
|
0.02-0.04 nM
|
0.17-0.27 nM
|
0.03-0.06 nM
|
% Relative Induction
|
100 %
|
101 %
|
100 %
|
97 %
|
81 %
|
95 %
|
*Values obtained from Sci Signal. 12, eaar2668 (2019)
# Values obtained from Mol Biol Evol. 12, 2643-2652 (2008)
Percent relative induction for MR is compared to maximal response to aldosterone.
N.D. = not determined.
Aldo = aldosterone, DOC =11-deoxycorticosterone, Prog = progesterone, 19norProg = 19norProgesterone
Progesterone binds to, but does not activate, full-length elephant shark GR.
Neither progesterone nor 19norProgesterone activated transcription by elephant shark GR (Figure 3A), although these steroids are transcriptional activators of elephant shark MR (Figure 3B) (41). The lack of progesterone activation of elephant shark GR could be due to decreased affinity of progesterone for elephant shark GR or to loss of activation of elephant shark GR after binding progesterone. We find that progesterone and 19norProgesterone inhibit activation of elephant shark GR by 10 nM corticosterone (Figure 5A) indicating that elephant shark GR recognizes progesterone. A parallel study showed that neither progesterone nor 19norProgesterone inhibited activation of human GR by 10 nM cortisol (Figure 5B). This indicates that the loss of activation by progesterone preceded the loss of progesterone binding to the GR, which occurred later in vertebrate evolution.
Corticosteroid activation of chimeras in which the NTD is swapped between elephant shark GR and MR.
To investigate further the origins of NTD activation of transcription by the GR, we studied corticosteroid activation of chimeras of elephant shark GR and MR, in which the GR NTD was fused to MR DBD-hinge-LBD and the MR NTD was fused to GR DBD-hinge-LBD (Figure 6). We find that despite low sequence similarity between the NTDs in elephant shark GR and MR, their NTDs had dominant effects on transcription of the DBD-hinge-LBD in each chimera (Figure 6, Tables 1 and 2). Thus, in the GR NTD-MR DBD-hinge-LBD chimera (GR NTD fused to MR DBD-hinge-LBD), cortisol and corticosterone increased activation by over 30-fold, compared to full length elephant shark MR (Figures 3B, 4B and 6B). Moreover, the GR NTD-MR DBD-hinge-LBD chimera also had a higher level of activation in the presence of progesterone and 19norProgestrone. In addition, the EC50s for corticosteroids and progestins were lower in the GR NTD-MR DBD-hinge-LBD chimera indicating that the NTD affects the affinity of steroids for the chimera, as well as the level of transcriptional activation (Table 2).
In contrast, the MR NTD reduced activation by corticosterone and other corticosteroids of the MR NTD-GR DBD-hinge-LBD chimera by over 90% compared to full-length GR (Figures 3A, 4A and 6A, Table 1). These data indicate that increased activation by the NTD in elephant shark GR evolved soon after it diverged from the MR.
RU486 is a glucocorticoid antagonist of full-length elephant shark GR.
To gain another measure of the similarities and differences between elephant shark GR and human GR, we investigated the effect of RU486 on transcriptional activation of elephant shark GR. RU486, originally developed as a progesterone antagonist (47,48) also is used as an antagonist for glucocorticoid activation of human GR (47), although at high concentrations (>100nM), RU486 is a weak GR agonist. We found that RU486 inhibited activation by 2 nM dexamethasone of elephant shark and human GR (Figure 7A, B). At 0.9 nM, RU486 inhibited dexamethasone activation of human GR by 50% (Figure 7A). At 2.2 nM, RU486 inhibited dexamethasone activation of elephant shark GR by 50% (Figure 7B).
Concentration dependent activation by RU486 of elephant shark GR and human GR, shown in Figure 7C and D, was used to calculate EC50 values for RU486. The EC50 of RU486 for elephant shark GR is 8.7 nM and for human GR is 0.55 nM. Maximal transcriptional activation by RU486 of elephant shark GR (Figure 7D) is considerably lower than for activation of human GR (Figure 7C). Nevertheless, although RU486 is not as active towards elephant shark GR as towards human GR, the responses of elephant shark GR to RU486 mimic those of human GR.