Future research and speculation

632 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 of JHs, JH mimics and anti-JHs, and the results from these experiments have been equivocal. It was reported at this symposium that no JH could be detected in eggs, larvae, nymphs and adults of the hard tick, Dermacentor variabilis, by the Galleria wax test bioassay; no JH could be detected in the hemolymph of adult D. varia- bilis and the adult soft tick, Ornithodoros parkeri, by GCMS; and no JH, MF or farnesol was synthesized by the synganglion or any other tissues of nymphs and adults of these two species Roe et al., 2000; Neese et al., 2000. These results are not the final word on the subject of JH in ticks but do question whether the com- mon insect JHs have a role in tick development and per- haps in chelicerates in general. Certainly, the processes of metamorphosis andor reproduction are regulated by a hormone, but the structure of this hormone and its origin remain conjectural.

4. Future research and speculation

4.1. Introduction Concurrent with our expanding knowledge of the gen- eral field of endocrinology has come the revelation that what was accepted as a central dogma is no longer true. In the case of steroid hormones, we now know that some can act at the level of the cell membrane in addition to acting as classic nuclear hormones at the level of the genome Christ et al., 1999. We predict that the same will be true of JH because of the patency studies of Davey this symposium and the accumulated evidence indicating that JH can act at the level of the genome. Indeed, it is known that membrane-oriented target sys- tems can influence gene regulation since kinases traverse the cytoplasm and enter the nucleus. It may be of interest to this readership and perhaps recall memories to Professor Applebaum, who organized JH VII, that Applebaum and Gilbert 1972 authored a paper showing that 20-hydroxyecdysone stimulates adenylate cyclase in lepidopteran pupal wing epidermis. This study together with those of Leenders et al. 1970, who showed that Drosophila salivary glands previously exposed to 20-hydroxyecdysone 20E contained increased quantities of endogenous cAMP, were cer- tainly among the first studies to indicate the multivalent action of steroid hormones. We note these early studies for two reasons. First, it is a way to acknowledge Shalom Applebaum both as a scientist and friend and for hosting this wonderful meeting in Jerusalem, and second, to emphasize an important fact that is now known to occur commonly in biological systems. The papers mentioned above on the stimulation of adenylate cyclase by 20E have not been cited more than a dozen times in the last twenty-five years, probably because mammalian endoc- rinologists have only recently read papers on insect endocrinology and, of course, because it was only disco- vered relatively recently that progesterone can act at the level of the cell membrane of the amphibian oocyte. What can we learn about this experience as far as JH research is concerned? First, discoveries that seem so important and earth shattering to insect researchers are usually not applauded by individuals working on higher organisms. They forget the ground-breaking work of Ulrich Clever, Keilin, etc. and, in most cases, do not read the insect literature. It is therefore important that this community of scholars working on JH acknowledge the excellent research conducted by their peers and work toward having the best of the projects funded by extra- mural agencies. Secondly, we must continue to publish some of our most excellent papers in journals dedicated to insect research, rather than publishing our best work in JBC, PNAS, etc. It is important to demonstrate that insects and their molecules are wonderful tools for eluci- dating basic developmental biological questions and that insect hormones still have promise as insect growth regulators. There are many problems to solve in the future, some of which have been raised in the preceding portion of this paper, but two emerge as quite basic: the JH receptor and the question of elucidating the nature and, perhaps, the function of JH “metabolites”. 4.2. JH metabolites It has been known for several decades that JH is degraded to JH acid and then perhaps to JH acid diol see previous discussion. Recent studies from the Bhaskaran laboratory Ismail et al., 1998 indicate that the major yolk protein in Manduca, as well as its mRNA, are detectable in the prepupal stage, and its production is enhanced by methoprene. The ability of the cells to respond to methoprene is acquired after the ecdysteroid- initiated commitment to metamorphosis. They have shown further that the acquisition of competence requires in addition to ecdysteroid, prior exposure to JH II acid, as shown by the injection of JH II acid and 20E into isolated abdomens. Methoprene acid also induces competence to respond to methoprene. Their analyses confirmed that vitellogenic transcripts are present in fat body only if the isolated larval abdomens were pre- treated with both ecdysteroid, JH acid or methoprene acid. They conclude that JH acid is not simply a metab- olite of JH, but is a hormone in its own right. This is of real interest since L.G. has proposed for several years that ecdysone metabolites may have morphogenetic roles of their own and may not be simple degradation pro- ducts, as is now proposed for the JH acid. We have known for quite a while that JH acid can be secreted by the corpora allata of Manduca at specific stages Sparagana et al., 1984; Janzen et al., 1991, and Nimi and Sakurai 1997 have shown that at some criti- 633 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 cal stages the hemolymph titer of Bombyx JH acid sur- passes that of the JH titer. Analogous results were found by the Bhaskaran group studying the production of pupal proteins by the Verson’s gland Ismail et al., 2000. Treatment in vitro with JH acid or methoprene acid plus a very low dose of RH5992 induced competence in the gland whereas RH5992 alone, methoprene acid alone, or methoprene plus RH5992 did not. Thus, these glands after treatment with the correct hormone mixture pro- duced pupal proteins. Similar results were found with the crochet epidermis. One other interesting fact is that JH acid acts in the presence of basal levels of ecdys- teroid, i.e. 2 of the molt-inducing dose. Data from the Gilbert lab indicate that the prothoracic glands always make a small amount of ecdysteroid and that there are two very distinct peaks in fifth stadium Manduca, the first of which is termed the “commitment” peak which presumably elicits the metamorphic process in the absence of JH see Henrich et al., 1999. Studies must now be done using JH acid. Finally, the Bhaskaran group, collaborating with J. Willis, explored whether juvenoid acids could play a role in the acquisition of metamorphic competence in abdominal rings of first sta- dium Drosophila personal communication. The mark- ers they used were the alcohol dehydrogenase ADH transcript, which is initiated at the distal adult pro- moter, and the broad complex, a transcription factor that first appears in response to ecdysteroids early in the final larval stadium. RT-PCR was used to identify transcripts. They found that juvenoid acid 0.5 µ gml plus ecdys- teroid 0.1 µ gml induced both the broad complex and the adult promoter of ADH. The juvenoid acids alone, juvenoid alone or ecdysteroid alone were ineffective. Juvenoid plus juvenoid acid inhibited the response. Thus, in the Manduca fat body, the Manduca Verson’s gland and Drosophila first instars, juvenoid acid plus a low dose of ecdysteroid induced metamorphic com- petency. Juvenoid acids are, therefore, concluded to be metamorphic hormones. Just recently, Richards et al. 1999 conducted a series of experiments on Drosophila that provided insight to the term “competence” at the molecular level. Their data indicate that the response of the salivary gland to ecdy- sone depends on developmental age, as might be expected from much earlier studies, and that individual loci appeared to evolve constantly in their capacity to respond to the hormone. Whether the Diptera and Lepi- doptera share mechanisms for achieving competency is conjectural, although neither JH or JH acid was utilized in the present studies. The use of various mutants has been quite helpful and as suggested by Richards et al. 1999, “the present challenge is to understand the mol- ecular mechanisms underlying the ‘acquisition of com- petence’”. The use of Drosophila mutants should be of great help in this regard, and it would be of real interest to utilize JH, JH acid, JHB 3 , etc. in some of these experi- ments. 4.3. Receptors For hormones to function, they must interface with the target cell via specific receptors, which for peptides, proteins and neurotransmitters are in the cell membrane, while for steroid hormones, thyroid hormones, etc., the receptor-ligand complex is in the nucleus. The mem- brane receptors transduce the hormonal signal generating intracellular second messengers, including calcium influx, while nuclear receptor-ligand complexes interact with the genome. The complexity of membrane-receptor interactions is exemplified by recent work on protho- racicotropic hormone stimulation of ecdysteroidogenesis in the prothoracic gland of Manduca Fig. 3. In some cases, the initial transduction could be at the level of the membrane, but the product of that receptor-ligand interaction can indeed migrate to the nucleus to exert its effect in a manner similar to nuclear receptor-ligand complexes. Thus, there are actually three cellular mech- anisms, two of which have been seen in insect systems. In the nucleus, the steroid hormone-receptor complex binds to specific sequences in the DNA hormone response elements, initiating or inhibiting the transcrip- tion of specific genes. There is a compelling body of evidence demonstrating that ecdysteroids act through such a mechanism and evidence to suggest that JH may function similarly. There is very little known about JH receptors, but pro- gress in the ecdysone receptor EcR field has been more dramatic, and what we have learned about EcR may help in our thinking about the JH receptor. The ecdysone receptors are members of the nuclear receptor super- family and are defined by a 66–68 amino acid sequence, the DNA binding domain, which contains two cysteine- cysteine zinc fingers, necessary for the recognition of, and binding of, the receptor to a hormone response element see Henrich et al., 1999 for review. Other characteristics of these receptors are an amino terminal domain, through which contact is made with other tran- scriptional factors the transactivating domain, and a hormone binding domain on the carboxy terminal side of the DNA binding domain. The EcR’s, as with many other nuclear receptors, exist as multiple isoforms as a result of differences in the transactivating domain. It is believed that the structural differences, although not major, confer unique nuclear properties on each isoform. The ecdysone receptor of Drosophila has been identified as a heterodimeric complex consisting of EcR and its retinoid-like partner, ultraspiracle USP, and it is thought that the complex is stabilized by 20E see Hen- rich et al., 1999. The absolute differences in the capacity of various cell types to respond to ecdysteroids, as well as temporal differences in the capacity of the 634 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 Fig. 3. A model for the action of prothoracicotropic hormone PTTH on cells of the prothoracic glands: R, PTTH receptor; a, b and g, G- protein subunits; CaM, calmodulin; A.C., adenylyl cyclase; PKA, pro- tein kinase A; PI-3K, phosphatidylinositol-3-OH kinase; S6 k, S6 kin- ase p70 ; S6, phosphorylated ribosomal protein S6; C, cholesterol; 7dC, 7-dehydrocholesterol; 3dE, 3-dehydrocholesterol; E, ecdysone; 20E, 20-hydroxyecdysone, EcR, ecdysone receptor. E is converted to 20E by an ecdysone 20-monoxygenase in peripheral, “target” tissues. It is of interest that the Drosophila S6 kinase appears to be a regulator of cell size Montagne et al., 1999 as well as an indirect stimulator of translational events see Henrich et al., 1999. same tissue, implies that more than just the presence of EcR and USP is involved in the response. These differ- ences may possibly be accounted for in part by the fact that the EcR locus of Drosophila encodes at least three isoforms, A, B1, and B2, which are produced by combi- nations of alternative promoters and splicing Talbot et al., 1993 and which have unique N-terminal domains linked to common DNA and ligand-binding domains. In Manduca, homologs of the Drosophila EcR-A and B1 isoforms have been cloned Fujiwara et al., 1995; Jindra et al., 1997. EcR-B1 is hypothesized to play a role dur- ing the commitment period early in each larval molt and in the pre-differentiative phase of the larval molt when it coordinates various molecular events during develop- ment such as cell proliferation, inactivation of genes and mechanisms related to the synthesis of new cuticle Riddiford, 1996. It would be of interest to study the relationship between JH acid and the activity of EcR-B1. In contrast to EcR, the USP locus appears to encode mainly a single transcript in Drosophila Henrich et al., 1994, although there were reports of truncated USP mRNAs in adult female ovaries and early embryos of Drosophila Khoury-Christianson et al., 1992; Henrich et al., 1994. Jindra et al. 1997 have identified two USP mRNAs in the epidermis of Manduca, the deduced iso- forms having distinct differences at the N-terminal region. These data and those from Drosophila indicate that ecdysteroids can elicit myriad effects in different insect cells, tissues and organs through combinatorial regulation, but which isoforms of EcR and USP form specific heterodimers at precise times to elicit a parti- cular effect is not known with certainty. Studies on the prothoracic gland of Manduca utilizing specific immunoprecipitation and Western blot analyses indicated that the EcR complex does indeed include EcR and USP but also the immunophilin FKBP46 Song et al., 1997. All three gene products localize within the nucleus of the prothoracic gland cells, and the data sug- gested that the hemolymph ecdysteroid titer, of which 20-hydroxyecdysone 20E is the major component, modulates the expression of both EcR and USP to achi- eve feedback regulation. Further studies indicated that there are two isoforms of USP in the Manduca protho- racic gland, one of which is associated with the EcR complex and which can exist in either a phosphorylated or dephosphorylated form Song and Gilbert, 1998. It was shown subsequently that 20E is responsible for initi- ating the translational expression that results in the phos- phorylation of an isoform p47 in the prothoracic gland. The data indicate that p47 forms a functional complex with EcR and that the ligand-complex interaction results in the down regulation of ecdysteroidogenesis as well as the inhibition of prothoracicotropic hormone-stimulated ecdysteroidogenesis. The second isoform of USP can also exist in a phosphorylated or dephosphorylated state, and it is this second isoform found in the prothoracic gland cells, but not associated with the EcR complex, that may perhaps be the JH receptor see below. It is of interest that the three isoforms of EcR found in the Manduca prothoracic gland consists of two proteins, one of which can exist in the phosphorylated or dephos- phorylated state. Further, preliminary studies indicate that the dephosphorylated state of the USP associated with the complex is the active form since there is about a five-fold increase in specific binding ponasterone A when the dephosphorylated isoform p47 is utilized. Bender et al. 1997 found that mutation of EcR-B1 uncoupled metamorphosis and caused embryonic lethal mutations that mapped to common sequences encoding the DNA and ligand binding domains. In the Drosophila larval salivary gland, the loss of a functional EcR-B1 635 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 prevents the activation of some 20E-induced genes. The bulk of their evidence suggests that EcR-B1 is necessary for normal metamorphic development of tissues, and further studies by this group revealed that studies of the various isoforms using the genetics of Drosophila can be a very profitable approach to hormone action and the role of receptors. It will not be long before the same is done for USP and its isoforms. In addition to USP and EcR, there are other proteins associated with the complex that have postulated roles as molecular chaperones, trafficking agents, etc. in anal- ogous vertebrate steroid hormone systems. To under- score the complexity of the receptor field and to demon- strate why the unequivocal characterization of the JH receptor remains an ethereal challenge, we have included a scheme depicting how the EcR complex matures and becomes an “active” complex capable of eliciting spe- cific transcriptional events Fig. 4. The Gilbert labora- tory has identified the hsp70, FKBP46, HIP the hsp 70- interacting protein, and HOP the hsp 90 organizing protein in the EcR complex of the prothoracic glands unpublished information. Fig. 4 contains factual infor- mation, although a portion is certainly theoretical. It has not been possible to show the presence of hsp 90 for example, that particular protein being a constituent of the steroid hormone receptor complex of mammals. It should be mentioned that Arbeitman and Hogness Fig. 4. A model for the development and activation of the ecdysone receptor EcR. 70 and 90, HSP70 and HSP90; FKBP 46, an immunophilin; USP, ultraspiracle; Repr, repressor; coact, coactivator; Hip and Hop defined on figure. From Rybczynski, Song and Gilbert, unpublished information. personal communication have identified analogous proteins in the EcR complex of Drosophila using a com- pletely different experimental paradigm from the immu- noprecipitation analysis conducted with Manduca. It is of interest that most laboratories working in this field study EcR or USP and neglect the five or seven other proteins that are part of the complex. It would not be surprising to find that the JH receptor is a heterodimer that may also have many constitutive proteins that act as chaperones. What do we actually know about the juv- enile hormone receptor at this time? 4.4. JH receptor The mechanism of action of the JHs has been a Rubric’s cube since the time of Wigglesworth’s classical experiment defining the hormone and identifying its source. Individuals have investigated the JH receptor in both larvae and reproductive adults, at which stage, of course, JH acts as a gonadotropic hormone. In the latter case, studies have been done on the cockroach Englemann laboratory and locust Wyatt laboratory utilizing the synthesis of vitellogenins as an end point Englemann’s work is summarized in his article in this issue. The Wyatt group utilized locust fat body nuclear protein extracts and DNA upstream from a JH inducible gene to examine a specific transcript elicited by JH e.g. 636 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 Zhang et al., 1996. The data in general suggest that the DNA element they found may be a JH response element and that the JH-elicited protein to which it is bound may be a transcription factor activating the initiation of JH target gene transcription. Thus, it could be a JH receptor. It would be of interest to determine the presence of USP in both the locust and cockroach system. Analogous investigation of immature forms of insects has been dif- ficult and virtually stagnant, and new approaches must be taken with this challenging but interesting molecule. Perhaps JH stimulates the production of JH receptors, analogous to 20E activation of EcR, and similar to JH regulation of the expression of its own transport protein in black mutant Manduca larvae as measured by the in vivo abundance of the mRNA transcript Orth et al., 1999. Why not stimulate JH receptor “activation” by applying large doses of JH and then searching for the receptor? There has been one “breakthrough” by Jones and Sharp 1997. As discussed by Jones and Jones this symposium; Jones et al., personal communication, JH and retinoic acid possess many structural similarities. They have similar alpha helices, a similar tertiary struc- ture, and salt bridges ligand-binding domain showing a high degree of conservation. They are also similar in the residue important to the conformation of the ligand- binding pocket, which also suggests similar secondary and tertiary structures. Retinoic acid regulates gene expression in some systems by stabilizing the heterodi- merization of a retinoic acid receptor RAR. Orphan receptors RXRs can bind several isoforms of retinoic acid but their actual ligand is not known with certainty. Perhaps the JH receptor is analogous, and one must search for a partner since USP may be as promiscuous as RXR. Jones and Sharp 1997, using a technique not employed previously at this level with insects, measured the conformational change in USP, the binding pocket of which binds both JH III and JHB 3 . The data were certainly intriguing and suggested further work along those lines to prove definitively that USP with or with- out a partner was the JH receptor. At this symposium, Jones noted her time-dependent studies of JH III-induced conformational change, but 10 27 M JH III was needed to demonstrate JH binding. It was of interest that JH III and methoprene follow exactly the same binding kinetics with USP, and hopefully, further studies by Jones and others may lead to the elucidation of the JH receptor. There is, of course, the criticism that USP could not possibly be the JH receptor since the K d of JH:USP bind- ing was in the order of 600 nmoles, and since it is a central dogma of endocrinology, particularly of ver- tebrate endocrinology, that a K d of 10 29 M is an upper limit for a true hormone receptor. However, as will be seen below, there are vertebrate receptors that bind with a K d in the micromolar range. Before turning to that phenomenon, we should mention the very interesting research of the Wilson laboratory this symposium on the methoprene Met gene in Drosophila, in their attempt to characterize a possible JH receptor. Using a genetic approach, as was done with EcR, they have cloned the gene from Met-resistant flies Met tolerant and believe that the resistance is due to decreased target site sensitivity. A number of Met genes have been cloned and after germ-line transformation, the Wilson labora- tory has shown homology with other transcription fac- tors. Their studies by Northern and immunocytochemical analyses produced data that are in line with a close relationship between the Met gene and the JH receptor. Whether the Met gene product forms a heterodimer is not known at this time. This is an excellent example of the use of genetics in solving developmental and, by extension, endocrinological problems. The one drawback of utilizing Drosophila in the study of JH action is the fact that the epidermis of higher flies does not develop, nor behave, like that of the Lepidoptera and many other orders of insects. That is, there is really no concrete evi- dence for an effect of JH on the larval or pupal epider- mis, in contrast to the situation in the Lepidoptera. The epidermis is derived from abdominal histoblasts in higher flies, rather than by division of epidermal cells as in earlier developmental and metamorphic stages of moths. As a result, there is no wax test type of effect and obtaining supernumerary molts is a rare and non- reproducible end point. Therefore, there is no absolutely convincing evidence for a role of JH in Drosophila development It is, of course, true that JH plays a role in the reproduction of adult females but morphogenetic actions in flies have not been shown unequivocally. We look forward with anticipation to further studies by the Wilson laboratory on the Met gene. These will surely help elucidate the nature of JH receptor activation, if indeed there is a role for JH in the earliest stages of Drosophila development. Another problem with Droso- phila JH studies concerns the finding from one of our laboratories L.G. showing that the corpora allata of Drosophila synthesize the juvenile hormone bis-epoxide JHB 3 , as well as JH III. The assay of JHB 3 on vitellog- enesis in the adult demonstrated that it is as effective as, if not more effective than, JH III. As discussed above, however, there is very little available in terms of mor- phogenetic action that can be bioassayed in Drosophila. It should be noted that JHB 3 has been found to be the predominant JH in several other species of higher Dip- tera. The exact role of JHB 3 relative to JH III is certainly worthy of greater research efforts. We would be remiss in not mentioning a very recent paper from the Bryant laboratory Kawamura et al., 1999, reporting that when the conditioned medium from Drosophila imaginal disk cell cultures was fractionated, a family of so-called imaginal disk growth factors was obtained, the first polypeptide growth factors to be reported from invertebrates. It appears that these growth 637 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 factors cooperate with insulin to stimulate the prolifer- ation, polarization and motility of imaginal disk cells. This growth factor family in Drosophila includes a mini- mum of five members, three of which are encoded by three genes in a tight cluster. Although structurally related to chitinases, they do not have enzymatic activity and are thought to have evolved from chitinases, but acquired new growth promoting functions. They are expressed most strongly in the embryonic yolk cells and in the fat body of the embryo and larva. Further studies suggest that these growth factors are secreted and trans- ported to target tissues via the hemolymph. The genes are also expressed in embryonic epithelium in associ- ation with invagination movements, so the proteins may have local as well as systemic functions. This is a new area in insect endocrinology, and it is possible that the complexity and number of such growth factors could rival that of the present situation with mammals. What does this or EcR have to do with JH? It is our premise that peptides, sterols and isoprenoids interact and that we cannot study one without considering the others. 4.5. FXR It is possible that the study of a mammalian nuclear receptor, FXR farnesoid X receptor may provide one means of proceeding with an investigation of the puta- tive JH receptor. All of the work discussed below on this subject is through the courtesy of Dr. C. Weinberger et al. unpublished information concerning research over the last several years at the National Institute of Environmental Sciences. In mammalian systems, there is a multivalent feedback regulation of 3-hydroxy-3- methylglutaryl coenzyme A reductase HMG CoA reductase activity in cells when mevalonate-derived sterols and non-sterols suppress isoprenoid synthesis and cell growth. Weinberger and his colleagues have shown that bile acids, 3 a-hydroxyecdysteroids and oxysterols, etc. are endogenous FXR affectors. The Weinberger group identified FXR activators using a transactivation assay in CHO cells by transfecting plas- mid DNAs expressing FXR, RXR, and a farnesol responsive CAT reporter gene. In essence, they studied FXR-dependent transcription. It is of interest that JH and methoprene affected FXR as did farnesol, although methoprene acid induced RXR. Could there be a link between this result and that described previously in which JH acid or methoprene acid appear to be essential for a change in competency according to the Bhaskaran laboratory? One metabolite of farnesol is, of course, the insect JH and it shows good affinity to the mammalian FXR. FXR transcriptional activity is also affected by oxidative products of the juvenile hormone antagonist precocene. [When precocenes were first described by Bill Bowers at the first JH Conference, the audience rippled with excitement. However, prospects for the use of precocenes as insect growth regulators were damp- ened by the later recognition of their toxicity to liver and kidney, and their possible conversion to chemically reactive epoxides.] FXR, therefore, is activated by metabolites of farnesyldiphosphate FDP which is a biosynthetic intermediate for all cellular isoprenoids and prenylated proteins. It is of interest that a key enzyme in the synthesis of insect isoprenoids, farnesyldiphosphate synthase FPS, has been cloned and identified recently in a number of tissues, and was in high concentration in the corpora allata of the moth Agrotis ipsilon Castillo- Garcia and Couillaud, 1999. The nucleotide sequence revealed that the cDNA encoded a polypeptide with an M r of 47,140 and that it shares regions homologous to the FPS of other organisms but does show some species specificity. This is the first FPS to be characterized in insects, and although not all tissues expressed FPS tran- scriptional activity, several did. Only further work will demonstrate whether the FPS in the corpora allata differs from the FPS in tissues not capable of JH biosynthesis. Further studies along these lines with the various key enzymes important in isoprenoid biosynthesis will surely help in elucidating the complete biosynthetic regulatory scheme for JH and perhaps identify enzymatic reactions that have the potential for inhibition or stimulation by endogenous molecules, e.g. allatostatins and allatotrop- ins. Farnesoids are also JH agonists and arise from FDP via a metabolic shunt that may serve to excrete surplus isoprenoid precursors in insects as shown years ago see Schooley and Baker, 1985. It is hypothesized that the substances that bind and “activate” FXR may establish an intracellular FXR activation potential in tissues where FXR is expressed. How these substances act to inhibit HMG CoA reductase enzyme activity is conjectural. One suggestion from the Weinberger group is that cAMP kin- ase PKA-mediated phosphorylation may be a means of controlling the enzyme’s activity, as may the addition of ubiquitin or the eliciting of a higher concentration of calcium-regulated proteases. Is the binding of JH to FXR followed by specific transcriptional events related to the insect JH receptor? At this time, we do not know. Another possibility is that FXR effectors may act as inducers of the USP gene product, the hypothesized JH receptor Jones and Sharp, 1997. Is it possible that FXR could form a heterodimer with USP, analogous to the situation with EcR? The finding that FXR responded to JH agonists sug- gests that this mammalian receptor may be related to the insect JH receptor. As an aside, Table 2 shows how well the JH bioassays wax test from thirty-five years ago agree with the current FXR binding and transcription activation assays. Using the FXR cDNA provided by Dr. Weinberger, several laboratories are now screening a variety of insect cDNA libraries to determine if a high stringency clone can be found and expressed, and the gene product tested for JH binding activity. If such 638 L.I. Gilbert et al. Insect Biochemistry and Molecular Biology 30 2000 617–644 Table 2 Comparison of juvenile hormone activities and FXR effector activities of different isoprenoids c Isoprenoid Units of JH FXR activation b activitygram a cecropia oil 1000 n.t. phytol 32 + isophytol n.t. all-trans farnesol 140 ++++ farnesal 32 + farnesyl acetate 5.4 +++ farnesenic farnesoic acid 7.8 ++ hexahydrofarnesol n.t. nerolidol 8.9 ++++ linalool 0.08 geraniol geranyl linalool 0.14 n.t. solanesol 0.05 n.t. a From Schneiderman and Gilbert, 1964. b From Weinberger et al. personal communication. c n.t. = not tested; zero FXR effector activities indicate that the FXR-dependent transcriptional induction was less than 2-fold when tested at doses below cytotoxicity. Juvenile hormone activity was quantified by the Galleria wax test and the A. polyphemus pupal assay. This table is from Weinberger et al. personal communication. exists, it would be another addition to many historical findings in biology and biochemistry that reveal that insects utilized molecules and pathways hundreds of millions of years before mammals evolved and that the roles, but not the basic structures, of some of these mol- ecules may have changed during evolution. It would also be of interest to test whether effectors of FXR can act as inducers of USP, the putative JH receptor. The FXR story is an interesting but bizarre tale. For example, the K d for the binding of a variety of sub- stances, including that of JH to FXR, is in the micromo- lar range. As noted previously, we are accustomed to dealing with active receptors that have binding affinities K d in the nanomolar or picomolar range and in the minds of some, a K d in the micromolar range would be evidence that a compound is not a receptor. The retinoic acid receptor also has a much higher K d than that nor- mally associated with a receptor, and perhaps it is time that we begin to rethink the question. Since the structural basis for the ability of FXR to bind a whole host of molecules is not yet known, the answer to the question of how isoprenoid metabolites stimulate FXR-dependent transcription cannot be answered. Indeed, as pointed out by Weinberger et al., direct binding has not yet been confirmed, and the possibility exists that each of the compounds that stimulates FXR-dependent transcription does so indirectly through the stimulation of synthesis of another ligand. Perhaps in the near future, we will know if insects yield a compound or compounds that show close sequence homology to the mammalian FXR.

5. Conclusions