Juvenile hormone resistance: ¡ no PASaran !

The control of growth, differentiation, metamorphosis, and reproduction in insects is a complex interplay of hormones, with the steroid hormone ecdysone and the sesquiterpenoid juvenile hormone (JH) playing major roles. Molting is controlled by ecdysone, and the type of molt is controlled by JH. The presence of JH maintains the insect in the juvenile (larval) state, hence its name (1), and the relative absence of JH allows metamorphosis to proceed. With the discovery of insect JH came the idea that application of JH or its agonists at an inopportune time, when the insect required the absence of JH for metamorphosis—the “Achilles’ heel” of insect development—would disrupt this complex endocrine machinery. Analogs of JH, developed since the 1970s and euphemistically called “insect growth regulators,” are now among the most environmentally friendly types of insecticides. These chemicals, including methoprene, pyriproxyfen, and fenoxycarb, proved to be excellent mimics of the natural hormone, but, paradoxically, the molecular mode of action of JH has remained very much a mystery (2, 3). Now, studies on the insect’s response to these powerful chemicals—resistance—provide a direct avenue into the mode of action of JH itself.

Ashok et al. (4) report in the current issue of the Proceedings the identity of the Methoprene-tolerant gene of Drosophila melanogaster as a member of the bHLH-PAS family of transcriptional regulators. This significant advance raises many questions, and its implications will be discussed here.

Selection for methoprene-resistant mutants in D. melanogaster led to the definition of a genetic locus, Methoprene-tolerant or Met (5). Flies carrying the Met alleles are up to 100-fold resistant to structurally diverse JH analogs and to JH itself. The Met flies are not only resistant to the lethal effects of JH, but also to all the more subtle morphogenetic signs of hormone excess. For instance, treatment with methoprene produces defects that mimic the phenotype of a subset of mutations in the Broad Complex, an ecdysone-regulated locus essential in metamorphosis; Met flies are protected from these defects (6). Whereas early studies on JH analog resistance in mosquitoes and flies implicated cytochrome P450-mediated detoxification as the principal means of resistance to these compounds, studies of the Met gene clearly implicated a target site resistance: no differential penetration or metabolism can be shown in Met flies, and the mutation is expressed autonomously in genetic mosaics. Met alleles can be obtained by a variety of means, including x-ray treatment, chemical mutagenesis, or P-element insertion (transposon tagging) mutagenesis. The latter technique allowed the molecular cloning of the Met gene. Ashok et al. (4) provide evidence that resistant Met flies transformed with a genomic fragment of the wild-type Met⁺ gene or with a Met⁺ cDNA driven by a heat shock promoter regain susceptibility to methoprene treatment. A molecular analysis is consistent with earlier genetic evidence that the Met alleles are amorphic, or functional null mutants.

The evidence for a role of the Met gene in JH action is thus very strong, but the identity of the MET protein proved to be a surprise. The discovery of a relationship to transcription factors was in itself predictable, but the membership of MET in the family of bHLH-PAS proteins was not expected. These proteins are characterized by a basic helix–loop–helix domain (bHLH) found in a number of transcription factors, and a PAS domain (7), named after the initial members of the family, products of the period gene (Per) and the single-minded (Sim) gene of Drosophila, and the vertebrate aryl hydrocarbon receptor (AhR) and Ah receptor nuclear translocator (ARNT). Whereas Per and Sim have known dimerization partners but no known small ligand, the AhR and ARNT proteins, in the presence of agonist ligands, form a functional dimer that up-regulates a battery of genes whose products metabolize various foreign chemicals (8). The known ligands of the AhR are a number of polycyclic aromatic hydrocarbons, with dioxin (TCDD) as the prototype. Endogenous ligands of the AhR are not known, but are suspected to be eicosanoids and/or indole-like metabolites. The family of bHLH-PAS transcription factors is rapidly expanding (9). The MET protein of 716 amino acids is most closely related to recently discovered members of this family from humans, expressed in brain and muscle, to a Drosophila homolog of the vertebrate Ah receptor nuclear translocator (dARNT) and to ARNT itself.

Does the homology to ARNT and AhR imply that MET is a JH receptor protein? Not necessarily. Mutant mammalian cells “resistant” to the toxicity of aryl hydrocarbons include both AhR mutants and ARNT mutants, and the genetic and biochemical evidence does not resolve the question whether MET functions as a homodimer (as ARNT may; ref. 10) or as a heterodimer. It thus will be necessary to express MET in a heterologous system and test its ligand binding activity toward JH and its analogs, as well as to find a presumptive binding partner for MET. dARNT, the product of the tango gene, is known to dimerize with Sim and with Trh, the product of the trachealess gene, and the list of Drosophila bHLH-PAS proteins is likely to increase.

The genetic evidence for a role of the MET protein in JH action at the molecular level contrasts with biochemical evidence presented recently for a role of the Drosophila Ultraspiracle protein (USP) as JH receptor protein (11). USP is a member of the nuclear receptor superfamily, an insect homolog of the retinoid X receptor. As a dimerization partner of the ecdysone receptor, USP forms a functional transcription stimulator of ecdysone-responsive genes. Jones and Sharp (11) show that JH binds to USP at relatively high concentrations, modifying its conformation and inducing USP-dependent transcription. Moreover, JH can promote homodimerization of USP in a two-hybrid system. Thus JH-dependent homodimerization is potentially competing with heterodimerization of USP with the ecdysone receptor, and one might envisage a model for the interference of JH with ecdysone-regulated genes. Such a competition of the ecdysone receptor for USP already has been observed with DHR38 and Seven-Up, two other Drosophila members of the nuclear receptor superfamily (12, 13). However, the current evidence for a JH receptor role of USP is not entirely convincing. USP does not bind the active analog methoprene (14), but binds JH acid, not known to be biologically active in Drosophila. Because several members of the nuclear receptor superfamily are known to bind sesquiterpenoids, for instance LXRα whose constitutive transcriptional activity is dependent on endogenous metabolites of mevalonate (15), binding of JH to USP may reflect the presence of an ancient and conserved terpenoid binding site. Retinoid X receptor binds methoprene carboxylic acid (14), and FXR, the “farnesoid X receptor,” binds JH (16), in each case foreign chemical ligands. Moreover, USP orthologs in other insect species, present in several isoforms, have diverged rapidly in their ligand binding domain (e.g., ref. 17), an observation that needs to be reconciled with a conserved function as JH receptor. Nonetheless, patterns of USP expression are not inconsistent with a role as JH receptor.

Is there a JH receptor then, and what is it? Perhaps one difficulty is that the work on MET and the work on USP both were done on Drosophila, an insect in which the role of JH in metamorphosis is not as clear as it is in other insects (2). Perhaps the gonadotropic role of JH (more ancestral) and the metamorphic role of JH (more recent) involve different transduction machineries; a membrane receptor for JH has even been suggested. Perhaps JH operates through somewhat redundant MET-like receptors, as the viability of the Met null mutants would suggest, and extensive gene duplication in the bHLH-PAS family has indeed occurred (9). Or perhaps MET functions not (only) in homotypic dimerization, as the AhR/ARNT model suggests, but binds heterotypically, to another class of transcription factors. In this respect, it should be noted that steroid receptor coactivators, such as SRC-1, TIF2, RAC-3, and p/CIP, involved in nuclear hormone receptor transcriptional activity are indeed members of the bHLH-PAS family (18–20). MET thus may turn out to be a “receptor coactivator” for either a JH receptor such as USP and/or the ecdysone receptor. The LXXLL signature motif identified in the vertebrate transcriptional coactivators and needed for binding to nuclear receptors (19, 21, 22) is seen in the MET sequence, and its potential function now may be tested experimentally. Directed mutagenesis of Met also may lead to dominant negative MET variants that might help elucidate the role of the Met gene. Ultimately, however, evidence for the elusive nature of the JH receptor will have to come from other insect species. Cloning of Met orthologs and a biochemical characterization of MET proteins will benefit from insect models in which the metamorphic and gonadotropic actions of JH are easier to study.

Cloning of the Met gene also should please those who do not see Drosophila only as a model eukaryote, but rather as a model insect pest. In the never-ending battle against pesticide resistance, Met is perhaps the first example of a predicted case where silencing of an insecticide target gene leads to resistance (23). The fitness deficit of Met (24) and its lack of full dominance would not prevent Met-type resistance from appearing in field populations of major agricultural pests treated with JH analogs, such as the whitefly (25). But advance knowledge of a potential resistance mechanism should allow monitoring and management of this type of resistance. The farmer will detect pests deficient in this PAS protein and echo the cry of resistance ¡ no PASaran !