Primary structure Secondary and tertiary structure

Insect Biochemistry and Molecular Biology 30 2000 671–679 www.elsevier.comlocateibmb Considerations on the structural evidence of a ligand-binding function of ultraspiracle, an insect homolog of vertebrate RXR Grace Jones a, , Davy Jones b a School of Biological Sciences, University of Kentucky, Lexington, KY 40506, USA b Graduate Center for Toxicology, Chandler Medical Center, University of Kentucky, Lexington, KY 40506, USA Received 31 October 1999; received in revised form 31 December 1999; accepted 25 January 2000 Abstract This analysis considers the structural evidence of a ligand-binding function of the nuclear receptor ultraspiracle USP. The positions and nature of residues in the ligand-binding domain of USP from six higher insects is evaluated in comparison to the function of conserved residues vertebrate receptors that have been co-crystallized with ligand. USP appears to conserve residues that in vertebrate receptors 1 form the hydrophobic ligand-binding pocket, 2 contact oxygen-containing moieties on ligands, such as hydroxyl, keto and carboxyl groups, and 3 in response to ligand-binding conformationally change to form a multi-helix hydrophobic groove for recruitment of transcriptional co-activators. These structural features are consistent with the recent report that USP can bind the epoxymethylfarnesoates juvenile hormones and thereupon is induced to change conformation.  2000 Published by Elsevier Science Ltd. All rights reserved. Keywords: Ultraspiracle; USP; Nuclear receptor; RXR; Juvenile hormone During the last decade, there has perhaps been no problem in insect biochemistry more exasperating than the molecular identification of juvenile hormone recep- tors Jones, 1995; Riddiford, 1996; Feyereisen, 1998. A number of potential candidates for nuclear JH receptors have been offered with various levels of supporting evi- dence, but there is not yet consensus on whether the field has actually arrived at its long sought destination Jones and Sharp, 1997; Palli et al., 1994; Ashok et al., 1998; Harmon et al., 1995. The structural relationship between juvenile hormone and retinoic acid prompts the consideration of whether a retinoid-type of receptor in insects may serve as a juv- enile hormone receptor. The last several years have seen an increasing number of reports on the primary structure of invertebrate homologs of vertebrate RXR. The ligand- binding domain of jellyfish, crustacean, tick and grass- hopper homologs have been shown to have closer sequence identity to the vertebrate RXR than to the USP Corresponding author. Tel.: + 1-606-257-2105; fax: + 1-606-323- 1059. E-mail addresses: gjonespop.uky.edu G. Jones, djones- pop.uky.edu D. Jones. 0965-174800 - see front matter  2000 Published by Elsevier Science Ltd. All rights reserved. PII: S 0 9 6 5 - 1 7 4 8 0 0 0 0 0 3 8 - 2 of higher insects such as Diptera and Lepidoptera Chung et al., 1998; Guo et al., 1998; Kostrouch et al., 1998; Hayward et al., 1999. While the jellyfish RXR does bind 9-cis retinoic acid RA, the tick RXR did not support 9-cis RA activation of transcription under the cell line transfection system used, and the binding properties of the grasshopper and crustacean RXRs have yet not been reported. The USP of higher insects has been considered to be an orphan receptor Thummel, 1995; Buszczak and Segraves, 1998, and hypotheses have been advanced that the ligand-binding function of higher insect USP has been lost Kapitskaya et al., 1996; Hayward et al., 1999. However, recently Jones and Sharp 1997 reported that D. melanogaster USP dmUSP can bind to JH III and JH III bisepoxide, which are natural JHs of that insect. The occasion of the VII International Symposium on Juvenile Hormones pro- vides an appropriate occasion to consider what structural evidence exists that USP of higher insects is a nuclear receptor that binds one or more endogenous ligands.

1. Primary structure

The full primary structure of USP as encoded in cDNA has been reported for at least six species of Dip- 672 G. Jones, D. Jones Insect Biochemistry and Molecular Biology 30 2000 671–679 tera and Lepidoptera e.g., Henrich et al., 1990; Oro et al., 1990; Shea et al., 1990; Tzertzinis et al., 1994; Kapit- skaya et al., 1996; Jindra et al., 1997; Perera et al., 1998; Vogtli et al., 1999. The sequence identity between these USPs versus the vertebrate and invertebrate RXRs is in the range of 40–50, while that between invertebrate RXRs and vertebrate RXRs is around 70. However, these statistics in and of themselves do not provide a basis for inferring whether USPs bind ligands and what those ligands might be. For example, vertebrate RAR and RXR both bind the same ligand 9-cis RA, yet the percent identity between them is only around 27 Mangelsdorf et al., 1990. The relevant indicators are whether the residues that have been retained identically or with conservative substitution are those residues necessary to preserve the secondary structures for basic architecture of the ligand-binding domain, and those necessary to preserve a functional ligand-binding pocket within the ligand-binding domain. Thus, in the analyses below, we do not confer importance to a residue in USP merely because it is conserved in USPs, but rather also because of its location in USP relative to the functional location of ligand- or co-activator-associated residues conserved in other nuclear receptors.

2. Secondary and tertiary structure

The crystal structure for unliganded human RXR α hRXR α has been reported, including the residues part- icipating in the various secondary structures Bourguet et al., 1995. It is possible to gauge the accuracy of algorithms for prediction of secondary structure by com- paring the consensus predictions of several methods for hRXR α against the secondary structure arrangement actually observed in crystallized hRXR α . Using the vari- ous secondary structure prediction alogithms available at the http:pbil.ibcp.frNPSA web site, we performed such an analysis. Fig. 1 shows the consensus locations of pre- dicted secondary structures for hRXR α vs those actually observed in the crystal structure, for the regions of the ligand-binding domain that contain corresponding resi- dues known to line the ligand-binding pocket of hRAR γ . In general, the analyses reasonably predict for hRXR α the locations of α -helices. We then applied those analyti- cal methods to Drosophila melanogaster USP dmUSP. The predictions of α -helical secondary structure for dmUSP generally parallel the predictions for hRXR α Fig. 1. These data on the predicted arrangement of sec- ondary structures in dmUSP support the hypothesis that the general tertiary architecture of the ligand-binding domain LBD of dmUSP is similar to that of the LBD for hRXR α . The remaining analyses below are rested on this hypothesis that the general tertiary architecture of secondary structures of the LBD of dmUSP is similar to that of the LBD for hRXR α .

3. Receptor residues contacting oxygen-containing moieties