LXXLL-related motifs in Dax-1 have target specificity for the orphan nuclear receptors Ad4BP/SF-1 and LRH-1

Abstract

The orphan receptor Ad4BP/SF-1 (NR5A1) is a constitutive activator, and its activity is repressed by another orphan receptor, Dax-1 (NR0B1). In the present study, we investigated the molecular mechanisms underlying this repression by Dax-1. Yeast two-hybrid and transient-transfection assays confirmed the necessity of three LXXLL-related motifs in Dax-1 for interaction with and repression of Ad4BP/SF-1. In vitro pull-down experiments confirmed that Dax-1 interacts with Ad4BP/SF-1 and also with LRH-1 (NR5A2). The target specificity of the LXXLL-related motifs was indicated by the observations that Ad4BP/SF-1, ERalpha (NR3A1), LRH-1, ERR2 (NR3B2), and fly FTZ-F1 (NR5A3) interacted through their ligand binding domains with all the LXXLL-related motifs in Dax-1 whereas HNF4 (NR2A1) and RORalpha (NR1F1) did not. Transcriptional activities of the receptors whose DNA binding domains (DBDs) were replaced by the GAL4 DBD were repressed by Dax-1 to various levels, which correlated with the strength of interaction. Amino acid substitutions revealed that Ad4BP/SF-1 and LRH-1 preferentially interact with L(+1)XXLL-related motifs containing serine, tyrosine, serine, and threonine at positions -2, +2, +3, and +6, respectively. Taken together, our results indicate that the specificities of LXXLL-related motifs in Dax-1 based on their amino acid sequences play an important role in regulation of orphan receptors.

FIG. 1.

Structure of Dax-1 constructs used in repression assays (Fig. 2 and 5C) and yeast two-hybrid assays (Fig. 3). For repression assays, full-length proteins were expressed using the pCMX vector. To be used as bait in yeast two-hybrid assays, proteins were fused at their N termini to GAL4 DBD in the pGBT9 vector. Numbers represent amino acid positions in Dax-1, reading from the N terminus. The sequences of three LXXLL-related motifs in these constructs are shown together with their position numbers (in brackets), and substituted amino acid residues are underlined. Black and white bars represent intact and altered LXXLL-related motifs, respectively.

FIG. 2.

LXXLL-related motifs are necessary for the repressor function of Dax-1. Repressor activities of the Dax-1 constructs were examined against CYP11A1 gene promoter transcription. CV-1 cells were transfected as described in Materials and Methods, except that the cells in each well received 200 ng of the reporter plasmid pS2.3H-LUC, 130 ng of pCMX-βGAL for reference, and 300 ng of MT-CEVα (an expression plasmid of the catalytic subunit of protein kinase A) with or without 40 ng of pCMX-Ad4BP and with increasing amounts (0, 10, 20, 40, and 80 ng) of pCMX-Dax-1 plasmids as indicated. Data are mean and SD.

FIG. 3.

LXXLL-related motifs are necessary for interaction of Dax-1 in yeast. This table shows a summary of the results of yeast two-hybrid assays. The bait and prey plasmids of indicated combinations were transfected into PJ69-4A yeast cells as described in Materials and Methods. + and − indicate yeast that was resistant and sensitive, respectively, to 3-AT.

FIG. 4.

The interaction of Ad4BP/SF-1 and ERα requires the AF-2 core for interaction with LXXLL-related motifs in Dax-1. (A) Schematic structures of bait and prey constructs for the mammalian two-hybrid interaction assays. Numbers represent amino acid positions reading from the N terminus. (B and C) Interaction of LXXLL-related motifs in Dax-1 with ERα (B) and Ad4BP/SF-1 (C) on their LBDs. CV-1 cells were transfected as described in Materials and Methods with 25 ng of bait plasmids per well and 50 ng of prey plasmids per well as effectors, as indicated. The reporter plasmid was UAS_G-TK-LUC. Data are mean and SD.

FIG. 5.

LXXLL-related motifs in Dax-1 interact with LRH-1, ERR2, and fly FTZ-F1. (A) In vitro interaction of MBP-Ad4BP and MBP-LRH-1 with fluorescence-labeled GFP-Dax-1[WT] and GFP-Dax-1[M123]. GFP was used not as a source of fluorescence but as a sequence that allows efficient expression (see Materials and Methods), because for unknown reasons, pCMX-Dax-1[WT] and pCMX-Dax-1[M123] failed to express the proteins in vitro. “10% input” represents one-tenth of the amount of labeled protein in a reaction mixture. (B) Interaction of LXXLL-related motifs in Dax-1 with orphan receptors on their LBDs. CV-1 cells were transfected as described in Materials and Methods with 25 ng of bait plasmids per well and 50 ng of prey plasmids per well as effectors as indicated. Data are mean and SD. (C) Repression by Dax-1 against transcription activated by orphan receptors whose DBDs were replaced by GAL4 DBD. EPC cells were transfected as described in Materials and Methods with 25 ng of pCMX-GAL4-LBD plasmids per well, with or without 50 ng of pCMX-Dax-1[WT] per well as effectors in the presence or absence of β-E₂ as indicated. Bars represent fold repression by Dax-1. Relative luciferase activities are expressed numerically. In these experiments, the reporter plasmid was UAS_G-TK-LUC.

FIG. 6.

Critical residues on DAXpep140-156 for interaction with Ad4BP/SF-1 and LRH-1. (A) Amino acid sequences of GAL4-TIFpep684-700, GAL4-DAXpep140-156, and related constructs that contain an amino acid substitution(s). Residues characteristic of GAL4-TIFpep684-700 and GAL4-DAXpep140-156 are indicated by black and gray boxes, respectively. For reference, the first leucine in the core sequence was defined as position +1. Protein expression in these bait constructs was confirmed by Western blot analysis with anti-GAL4 DBD monoclonal antibody RK5C1 (Santa Cruz Biotech, Santa Cruz, Calif.) as shown together with the result of the prey constructs of Ad4BP/SF-1, LRH-1, and ERα probed with anti-VP16 AD monoclonal antibody 1-21 (Santa Cruz Biotech). (B) Interaction of GAL4-TIFpep684-700 with orphan receptors and ERα on their LBDs. (C) Interaction of Y147H/S148R with Ad4BP/SF-1, LRH-1, and ERα on their LBDs. (D to F) Effects of amino acid substitutions on interaction with Ad4BP/SF-1 (D), LRH-1 (E), and ERα (F) on their LBDs. In these panels, GAL4-DAXpep140-156 and GAL4-mutDAXpep140-156 are abbreviated as “DAX” and “mutDAX,” respectively. (G) Effect of substitutions in S144K/T151Q on interaction with Ad4BP/SF-1, LRH-1, and ERα. In these experiments, CV-1 cells were transfected as described in Materials and Methods with 25 ng of bait plasmids and 50 ng of prey plasmids as effectors as indicated. The reporter plasmid was UAS_G-TK-LUC. Data shown in panels B to G are mean and SD.

FIG. 7.

Cofactors for Ad4BP/SF-1, LRH-1, and FTZ-F1 contain LXXLL or LXXLL-related motifs with S, Y, S, and/or T at positions −2, +2, +3, and +6, respectively. Six amino acid alignments show sequences of the LXXLL or LXXLL-related motifs (many of these are still putative) in Dax-1, FTZ, and SHP. Black boxes in the alignments highlight serine, tyrosine, serine, and threonine at positions −2, +2, +3, and +6, respectively. Solid bars in the schematic diagram indicate the positions of the LXXLL or LXXLL-related motifs, and numbers indicate which alignment (1 to 6) corresponds to the bar. Sequence comparison of Dax-1 and SHP suggested that the positions of the LXXLL-related motifs in the LBD shown in alignments 5 and 6 are essentially identical (data not shown). All sequences were retrieved from the GenBank database. The sequences of pig, chicken, zebrafish, and salmon SHP have putatively been obtained by conceptual translation from expressed sequence tag sequences retrieved by homology. Accession numbers are as follows: for Dax-1, 5016090 (human), AAK01643 (chimpanzee), AAK01644 (orangutan), AAK01645 (gibbon), AAK01647 (rhesus monkey), AAK01646 (marmoset), 1870692 (pig), AAF89613 (horse), 1345127 (mouse), 1620449 (rat), 4100303 (wallaby), AAF19395 (chicken), AAD55095 (alligator), and BAB85864 (wrinkled frog); for FTZ, 123366 (Drosophila melanogaster); for SHP, BAB68530 (human), AW325605 (pig expressed sequence tag), 1374949 (mouse), 1871465 (rat), BG712451 (chicken expressed sequence tag), AA495245 (zebrafish expressed sequence tag), and BG935659 (salmon expressed sequence tag).