Hormonal Regulation in Insects: Facts, Gaps, and

0 downloads 0 Views 20MB Size Report
Previously, structure prediction and molecular mod- ..... now have free NH2 and COOH termini, respectively, is achieved ..... matched a hemolymph peak of AVP-IDH, but there were ..... is rather at odds with the normal function of MT, where ...... Regulation of insect de- americana, and Blattella germanica) and of a cricket (G.
PHYSIOLOGICAL

REVIEWS

Vol. 77, No. 4, October 1997 Printed in U.S.A.

Hormonal

Regulation in Insects: Facts, Gaps, and Future Directions

GERD GADE,

KLAUS-HUBERT

HOFFMANN,

AND JEFFREY

H. SPRING

Zoology Department, University of Cape Town, Rondebosch, South Africa; Lehrstuhl Tkr6kologie I, Universitdit Bayreuth, Bayreuth, Germany; and Department of Biology, University of Southwestern Louisiana, Lafayette, Louisiana

I. Introduction II. Neuropeptides and Metabolism A. Adipokinetic and hypertrehalosemic neuropeptides B. Diuretic and antidiuretic neuropeptides III. Neuropeptides and Muscle Activity A. Proctolin and cardiostimulatory peptides B. Myokinins C. Sulfakinins D. Pyrokinins/myotropins E. Tachykinins F. Periviscerokinin G. Accessory gland myotropins H. Other myotropins I. Myoinhibitory peptides and FMRFamide-related peptides IV. Epithelial Hormones and Neuropeptides: Reproduction, Growth, A. Pheromone biosynthesis activating neuropeptides B. Juvenile hormones and allatotropins/allatostatins C. Ecdysteroids and PTTH/bombyxin D. Allatinhibin, folliculostimulins, and folliculostatins E. Ecdysis-triggering and eclosion hormones F. Diapause hormones V. Miscellaneous Insect Hormones and Neuropeptides A. Bursicon B. Control of morphological and physiological color change VI. Conclusions and Future Perspectives

and Development

964 964 964 970 977 978 979 980 980 981 981 981 982 982 983 983 986 996 1004 1006 1007 1009 1009 1010 1010

G%de, Gerd, Klaus-Hubert Hoffmann, and Jeffrey H. Spring. Hormonal Regulation in Insects: Facts, Gaps, and Future Directions. Physiol. Rev. 77: 963-1032, 1997. -There are two main classes of hormones in insects: 1) the true hormones produced by epithelial glands and belonging to the ecdysteroids or juvenile hormones and 2) the neuropeptide hormones produced by neurosecretory cells. Members of these classes regulate physiological, developmental, and behavioral events in insects. Detailed accounts are given on isolation, identification, structureactivity relationships, mode of action, biological function, biosynthesis, inactivation, metabolism, and feedback for hormones involved in 1) metabolic regulation such as the adipokinetic/hypertrehalosemic peptides and the diuretic and antidiuretic peptides; 2) stimulation or inhibition of muscle activity such as the myotropic peptides; 3) control of reproduction, growth, and development such as allatotropins, allatostatins, juvenile hormones, ecdysteroids, folliculostimulins and folliculostatins, ecdysis-triggering and eclosion hormones, pheromone biosynthesis activating neuropeptides, and diapause hormones; and 4) regulation of tanning and of color change. Because of the improvements in techniques for isolation and structure elucidation, there has been rapid progress in our knowledge of the chemistry of certain neuropeptide families. With the employment of molecular biological techniques, the genes of some neuropeptides have been successfully characterized. There are, however, areas that are still quite underdeveloped. These are, for example, 1) receptor studies, which are still in their infancy; 2) the hormonal status of certain sequenced peptides is not clarified; and 3) functional studies are lacking even for established hormones. The authors plead for a concerted effort to continue research in this field, which will also advance our knowledge into the use of insect hormones as safer and species-specific molecules for insect pest management.

0031-9333/97

$15.00

Copyright

0 1997 the American

Physiological

Society

963

964

GiiDE,

HOFFMANN,

I. INTRODUCTION Humans have always been intrigued by insects such as butterflies and moths, not only for their general beauty and striking colors, but also because of the spectacular changes that occur during their life cycle. Ecdysis and metamorphosis are events known to everyone from childhood, and it is therefore not surprising that these processes, which are hormonally mediated, have drawn the attention of many insect physiologists and endocrinologists. The Polish scientist Kopec was the first to conclude, from his studies on the control of molt in the gypsy moth, Lymantria dispar, that the brain contained a factor responsible for the metamorphosis of the final larval instar into the pupa. Moreover, this factor had to be released into the hemolymph, since severing the neuronal connections between the brain and body did not inhibit pupation, whereas removal of the brain did (362, 363). This “brain hormone” is known today as prothoracicotropic hormone (PTTH). Chemically, it is a peptide, and much insight has been gained into its action (see sect. rvC). As spectacular as these developmental changes are, they are not the only humorally mediated processes in insects. Indeed, a vast number of physiological, developmental, and behavioral events are regulated by hormones. These include such diverse actions as the control of substrate mobilization, water and ion balance, reproduction and growth, eclosion, diapause, and pheromone production, to name but a few. The majority of the regulating substances are, as we will see, peptides. In contrast to the juvenile hormones and ecdysteroids, which are produced in epithelial hormonal glands, the peptides are synthesized in neurosecretory cells, which are most abundant in the brain but do occur throughout the nervous system. The main neurohemal organs are the paired corpora cardiaca (CC), which are the source of both neurosecretory products from the brain and endogenous neuropeptides, and the perisympathetic organs, which release neurosecretory material from the ventral nerve cord. True hormones are rigorously defined by criteria such as storage in and release from neurohemal sites and transport via the hemolymph. In many cases, these criteria have not yet been conclusively demonstrated, so we more often use the term neuropeptide rather than neurohormone. The last decade has been especially fruitful in the elucidation of primary structures of insect neuropeptides. The main reasons for this are the tremendous improvements in isolation techniques, notably high-performance liquid chromatography (HPLC), and in protein chemical detection methods including a new generation of automated amino acid sequencers and mass spectrometers. Further impetus for the surge in structural information came from a different direction. The use of classical pesticides has been heavily criticized by the environmentally conscious public. so alternative strategies for combating

AND

SPRING

Volume

77

insect pests are being sought. Baculoviruses, natural insect-specific biological control agents, have become a favorable model system for scientists at universities and pharmaceutical companies alike. The action of baculoviruses can be enhanced by incorporating foreign genes into their genomes. Because many insect neuropeptides control key functions in the life cycles of insects, such substances are targeted as likely candidates. First, however, their structures, or at least portions of them, have to be known to identify the corresponding gene using molecular biological techniques. In the present overview, we mainly discuss the neuropeptides and steroids whose structures are known, although, where necessary, we also review the literature on incompletely characterized neuropeptides. II.

NEUROPEPTIDES

A. Adipokinetic Neuropeptides

AND

METABOLISM

and Hypertrehalosemic

The first evidence that intermediary metabolism in insects was under hormonal control was provided by Steele (667) for carbohydrate mobilization in cockroaches and by Beenakkers (17) and Mayer and Candy (426) for lipid breakdown in locusts. When the first adipokinetic hormone (AKH) of the locust, Lom-AKH-I (see Table l), was fully characterized (669), it became clear that the primary structure was strikingly similar to that of the red pigment-concentrating hormone (RPCH) of prawns (PabRPCH; Ref. 140). Furthermore, the peptides were functionally cross-reactive (436). These results were the basis for grouping these and similar peptides, which were structurally related but had diverse functions, into what is now known as AKI-IRPCH family of peptides. More than 30 members of this family have been sequenced to date (180, 181; Table 1). So much information has been accumulated on this peptide family, regarding their physiology, biochemistry, and action, that the interested reader is also referred to the following overviews: extraction, purification and sequencing (167, 180,600), actions and structures (168, 226, 485), role during flight (173, 221, 222, 739), synthesis and inactivation (490)) and evolutionary relationships (178, 197). Although peptide structures are known for many insect species, most of the coherent physiological data have been gathered on the adipokinetic peptides of locusts and hypertrehalosemic peptides of cockroaches. Thus these data are preferentially dealt with, although results from other species are incorporated where necessary. 1. Horrnone source, and precursors

localization,

synthesis,

The CC of insects are the major neurohemal organs of the endocrine svstem. Thev store and release neurohor-

Octobe14 1997

INSECT

HORMONES

manes that are synthesized by neurosecretory cells of the brain. Additionally, they contain intrinsic neurosecretory cells which, in some insects, such as the locusts, are clustered together. In the locust, this region is called the glandular lobe to distinguish it from the storage lobe, in which the axons of the brain’s neurosecretory cells arborize. The intrinsic neurosecretory cells of the CC typically contain a large number of electron-dense secretory granules of ZOO- to 600-nm diameter (544), which are the source of the adipokinetic activity, as shown by differential centrifugation (668). Localization of the AKH peptides in the intrinsic neurosecretory cells was corroborated by immunoeytochemical studies (622). With the use of antibodies highly specific for either Lom-AKH-I or Lom-AKH-II, it was shown that individual neurosecretory cells contain both peptides (255) and that they are even colocalized in the same secretory granules (105). Very recently, in situ hybridization experiments demonstrated that the signals for the mRNA of all three AKH preprohormones of Locusta migratoria are colocalized in the neurosecretory cells ‘ . w The detailed pathways of the biosynthesis of the two adipokinetic hormones from Schistocerca gregaria (LomAKH-I and Scg-AKH-II), including the characterization of the hormone precursors, have been elucidated by direct protein chemistry methods as well as molecular cloning (254, 490, 623). Similar studies have corroborated some of these results for AKH family peptides in other species, such as Schistocerca nitans (480), Manduca sexta (39), Drosophila melanogaster (479), L. migratoria (31), and also for the crustaceans, Care&us maenas (393) and Callinectes sapidus (350). There is a distinct mRNA encoding for each AKH precursor; thus there are two specific mRNAs in S. gregaria/nitans and three in L. migratoria, which are translated into the discrete precursors, the prepro-AKHs. So, in the ease of L. migratoria, three precursors are synthesized. The organization of the precursor is basically the same for all AKH peptides: a signal peptide is followed by the respective AKH sequence, subsequently a Gly residue for amidation and a processing site and, at the COOH terminus, the sequence for another peptide, known as the “tail peptide” or AKH precursor-related peptide (APRP). The function of the APRPs is not yet known. Their length is quite variable; whereas the tail peptides from the Lom-AKH-I, -11, and Scg-AKH-II precursor are 2%mers and form dimers, those from Lom-AKH-III and precursors for M. sexta, D. melanogaster, and the crustaceans are longer, and it is not known whether they form dimers. Although no function has been attributed to the APRPs, they can be coreleased with the AKHs in vitro, and it is speculated that they fulfill a distinct function during flight as well (256). What is unique in the biosynthesis of the AKHs from S. gregaria is that their immediate precursors are not linear prohormones but dimers (490). For example. after cleaving off the signal peptide, the two

965

independently translated monomers of the pro-Lom-AKHI, which consist of the sequence for Lam-AKH-I and t’he Cys-containing tail peptide, are oxidized to a precursor dimer. Thereafter, the precursor is processed to the following products: two monomeric molecules of Lom-AKHI and one dimeric molecule of the APRP. Recently, the processes necessary for prohormone processing have been elucidated in vitro (565). The CC contain endoproteolytic activity which cleaves the peptide at the COOIIterminal side of the Arg residue at the processing site in each chain of the dimer, resulting in a COOH-terminally extended Lam-AKH-I (AKH-Gly Lys-Arg). This is subsequently cleaved by a carboxypeptidase H-like enzyme, removing Arg and then Lys. The amidated AKH is now produced from the Gly-extended AKH by the action of a peptidylglycine-cr-amidating monooxygenase. Apparently, the synthesis of authentic Lam-AKH-I also occurs in the brain of L. migratovia, but it is not known whether cerebral Lom-AKH-I is released or has a possible neuromodulatory role within the central nervous system (CNS) (440). 2. Release of AKITs In locusts, the release of AKH has been demonstrated during flight, and it was shown that only a fraction of the material stored in the CC is released into the hemolymph (67). Both Lom-AKH-I and -11were shown to be released during flight, and it appeared that the release is controlled by octopamine and adenosine 3’,5’-cyclic monophosphate (CAMP) (504; see Ref. 485). Other research groups, however, were able neither to find octopamine-immunoreactive fibers in the CC of L. migratoria (354) nor to demonstrate AKH release by octopamine (505). Recently, axon terminals immunoreactive to locustatachykinin I (LomTK-I) were shown to reside in close contact with the glandular cells of the CC. In addition, release of Lom-AKH-I was initiated by Lom-TK-I in an in vitro system (466). Recently, the crustacean cardioactive peptide (CCAP; see Ref. 654) has been isolated from the brain of the desert locust, Schistocerca gregaria, and shown to release AKH from the corpus cardiacum (721). With regard to other insects, the release of their respective AKIIRPCH family peptides has been reported from blowflies (720, 751) and from M. sexta (779). Once the peptides are released into the hemolymph, they can be transported to their target cells, bind to specific membrane-bound receptors, and exert their biological functions. Before we review these topics, however, it is important to examine the chemical nature of the peptides. 3. Structural

data on AKEURPCH peptides

The CC is the source of these peptides, and compared with most other insect neuropeptides, relatively

966

GiiDE,

TABLE 1. Primary

structures

Lom-AKH-I Phm-AKH Del-CC Cam-HIT H-I Car-n-HrTH-II

Phl-CC Taa-HoTH Hez-HrTH Rom-CC Bld-HrTH

Mas-AKH

Psi-AKH Lia-AKH

II

Locusta migratoria Schistocerca gregaria Phymateus morbillosus Decapo toma lunata Carausius morosus C. morosus Sipyloidea sipylus Extatosoma tiaratum Phymateus leprosus Tabanus atratus Helio this xea Romalea microptera Blaberus discoidalis Nauphoeta cinerea Leucophaea maderae Gromphadorhina portentosa Blat tella gerrnanica Platypleura capensis Munxa trimeni Cacama valavata Diceroprocta semicincta Magicicada sp. Manduca sexta H. xea Bombyx mori Pseudagrion inconspicuum Ischnura senegalensis Libellula auripennis Ceratogomphus pictus Pantala

Emp-AKH Ani-AKH

AND

of the adipokinetic

Species

Peptide

Plc-HrTH-I-/-,

of peptides

HOFFMANN,

SPRING

homnone/red Sequence

PQLNFTPNWGTamide PQLNFTPNWGSamide pQLNFSPNWGNamide pQLTFTPNW*GTamide PQLTFTPNWGTamide

PQLTFTPNWGSamide pQLTFTPGWGYamide pQLTFSSGWGNamide pQVNFTPNWGTamide pQVNFSPGWGTamide

pQVNFSPSWGNamide

pQLTFI’SSWGamide

pQVNFTPGWamide pQVNFTPSWamide

jfavescens

Empusa pennata Sphodromantis sp. Anax imperator Aeshna subpupillata

pQVNFTPNWamide pQVNFSPSWamide

Ano togas ter sieboldii Pea-&AH-I

Grb-AKH

Tern-HrTH

Pab-RPCH

Lom-AKH-II Scg-AKH-II

Periplaneta americana Blatta orientalis Lep tino tarsa decemlineata Trinervitermes trinervoides Mas to termes darwiniensis Gryllus bimaculatus Acheta domesticus Gryllodes sigillatus Romalea microptera Tenebrio moli tor Zophobas rugipes Onymacris plana 0. rugatipennis Physadesmia globosa Polyphaga aegyptiaca Decapotoma lunata Pandalus borealis Cancer magis ter Carcinus maenas Orconectes limosus L. migratoria S. gregaria Schistocerca nitans P. leprosus P. morbillosus Heterodes namaqua Acanthoproctus cerwinus Libanasidus vittatus Anabrus simplex

pQVNFSPNWamide

pQVNFSTGWamide

pQLNFSPNWamide

pQLNFSPGWamide

pQLNFSAGWamide pQLNFSTGWamide

Volume

pigment-concentrating

77

hormone family Reference

No.

633, 669 669 194 179 195 199 165 199 193 310 308 187 250 198 201 201 201, 716 189 189 718 718 556 776 309 300 311 311 170 M. Janssens, R. Kellner, and G. Gade, unpublished data Janssens, Kellner, and Gade, unpublished data 172 172 190 Janssens, Kellner, and Gade, unpublished data Janssens, Kellner, and Gade, unpublished data 15, 599, 630, 753 201 191 392 392 200 91, 758 174 187 202 202 177 177 177 192 179 140 206 206 206 185, 633 185, 633 185 193 194 174 174 174 S.E. Reynolds and D.A. Schooley, unpublished data

October

1997

TABLE

l-continued

Peptide

Mem-CC

Scd-CC-I

Scd-CC-II Ona-CC Lom-AKH-III Miv-CC Poa-HrTH Pea-CAH-II

Taa-AKH Pht-HrTH

INSECT

Species

Melolontha melolontha Geo trupes s tercorosus Pachnoda marginata Pachnoda sinuata Scarabaeus deludens Gareta ni tens Onitis aygulus Onitis pecuarius Scarabaeus deludens Gareta ni tens 0. aygulus 0. pecuarius L. migratoria Microhodoterrnes viator Polyphaga aegyptiaca P. americana Blatta orientalis Leptinotarsa decemlineata Tabanus atratus Phomzia terraenovae Drosophila melanogaster

967

HORMONES

Sequence

pQLNYSPDWamide

pQFNYSPDWamide

pQFNYSPVWamide pQYNFSTGWamide pQLNFT.PWWamide pQINFTPNWamide pQITFTPNWamide pQLTFTPNWamide

pQLTFTPGWamide pQLTFSPDWamide

AKH, adipokinetic hormone; RPCH, red pigment-concentrating hormone; 7 In all species of cicadas, 2 peptides were isolated by high-performance liquid sequence. At the moment, modification on peptide I is not known.

high quantities are stored in these glands. Accordingly, the CC are routinely dissected, extracted in 80% methanol, and sufficiently purified for structural analysis by a single-step procedure on reverse-phase HPLC (184). If, however, whole heads are used at the beginning of the purification and, thus, more contaminating material is introduced, more steps for isolation are necessary (250, 392). Both terminals of the peptides are blocked, and so structure elucidation is achieved mainly by Edman degradation (after deblocking the NH,-terminal pyroglutamate residue) and various mass spectrometric techniques (see Ref. 181). The family is comprised of more than 30 members at present (Table 1). Representative peptides have been found in most insect orders, and the sequence information has been utilized to construct possible phylogenetic trees (166, 178, 197). Whereas the RPCH seems to be common to all crustacean species, insects show a high degree of structural variability. Common characteristics of the family are that the peptides have a length of S- 10 amino acids, are NHZ-terminally blocked by a pyroglutamate, have a COOH-terminal amide block, the amino acids at positions 8 and 9 are Trp and Gly, are mainly uncharged, and contain at least 2 aromatic amino acids (mostly Phe, but some have Tyr at position 4 and Trp at position 8). There are, however, some charged members sequenced from certain Diptera (Pht-HrTH) and scarabaeoid beetle species (Mem-CC, &d-CC-I). Moreover, the recently discovered octapeptides of certain dung beetles (Scd-CC-I and -11, Ona-CC) contain an additional aromatic amino acid (Tyr or Phe) at position 2 (182, 183). In addition to the posttransla-

Reference

No.

171 171 196 196 182 182 183 183 182 182 183 183 494 392 192 599, 630, 753 201 191 310 203 601

CC, corpora cardiaca. * There is a hexose substituted on the Trp. chromatography. Edman degradation sequencing yielded the same

tional modifications at the terminals, the hypertrehalosemic hormone I of the stick insect, Carausius mo~osus (Cam-HrTH-I) is glycosylated (195). The site is not the usual Ser/Thr (0-glycosylation) or Asn (N-glycosylation), but Trp. As in human ribonuclease (275), the hexose is very likely C-glycosidically linked to the C-2 atom of the indole ring of Trp (G. Gade, R. Kellner, R. Kaufmann and M. Kalbitzer, unpublished data). Another, as yet unidentified, modification occurs in a hypertrehalosemic peptide of cicadas (189, 516, 718). Previously, structure prediction and molecular modeling studies on a number of members of the AKIURPCH family have indicated that in many of them there is a potential to adopt a P-turn conformation (670, 741). Recent studies employing circular dichroism (CD) spectroscopy on six members of the family, including Lom-AKHI and -11, have revealed that none shows a clear ordered conformation in aqueous solution at pH 7.5 and room temperature (0. Cusinato, A. Drake, G. Gade, and G. Goldsworthy, unpublished data). At low temperature in ethanediol-water (2: l), however, CD spectra are consistent with a left-handed extended helix conformation. For those peptides that have been shown by hyperlipemic bioassay to be very active in the locust, interaction with sodium dodecyl sulfate (SDS) micelles, which are good membrane mimics, induced the formation of a P-turn. Nuclear magnetic spectroscopy (NMR) studies on the octapeptide Emp-AKH, using dimethyl sulfoxide as solvent, led to a model for the secondary structure consisting of a /?-sheet-type conformation for residues l-5 and a pturn for residues 5-8 (781).

968 4. St?uctzcre-act,ifuity

GiiDE,

and recepto/r-binding

HOFFMANN,

studies

The mode of act,ion of peptide hormones is characterized by binding of the peptide ligand to a specific cell membrane-bound receptor molecule. Binding to the extracellular portion of that receptor induces a conformational change in the receptor molecule causing further changes in membrane proteins to transduce the extracellular signal into an intracellular message; often G proteins, adenylate cyclase, and the second messenger CAMP are involved. None of the receptor proteins of the AKH/RPCH family peptides is known yet, and indeed, there is only one report that looks at specific binding of an AKH ligand. Binding of the M. sexta adipokinetic hormone (Mas-AKH), which activates glycogen phosphorylase in larval fat body, to purified membrane fractions of fat cells, was followed using stringent criteria (778). This study was possible, since a labeled AKH with a high specific activity was available; the authors used a tritiated ligand (at position 4), which was as biologically active as the unmodified Mas-AKH. All other reports do not directly analyze binding but investigate this problem indirectly by studying the effect of various ligands (either bioanalogs or synthetic analogs) on biological activity. Such structure-activity studies have been performed on either lipid mobilization in locusts and M. sexta (150, 169, 176, 224, 225, 670), carbohydrate mobilization in cockroaches (147, 164, 169, 175, 186, 249), phosphorylase activation in J& sexta (776), or increasing of heart rate in Periplaneta americana (14). Generalizing, it appears that receptors of those insects containing only one endogenous peptide, like M. sexta and Blaberus discoidalis, are more selective. Most of the bioanalogs tested in t,hese species exhibited little activity. The receptors of those species containing two (P. americana) or three (,Y migra toria) endogenous peptides showed maximal activity with the majority of bioanalogs, and the half-maximal response for certain bioanalogs was 50-fold higher than for the native peptides. Such data may suggest that the latter species have a receptor subtype for each of their endogenous peptides and, therefore, exhibit a broader spectrum of binding overall. It would be interesting indeed to test this hypothesis conducting proper receptor binding assays such as those discussed for M. sexta above. The novel Ona-CC peptide containing a Tyr residue at position 2 may be a good candidate for iodination and use in receptor binding assays for locusts and cockroaches. For further discussion of t’he multiple receptor hypothesis, see section VI. A study using single amino acid replacement analogs for the endogenous octapeptide Pea-CAH-I and measuring carbohydrate mobilization in P. americana quite convincingly demonstrated the absolute requirement for the aromatic amino acid side chains at positions 4 (Phe) and 8 (Trp) and the great importance of the blocked terminals

AND

SPRING

Volume

77

(pGlu at position 1 and the amide at the COOH terminus) for interaction with the receptor (186). In addition, replacements at positions 6 and 7 were not crucial in PeaCAH-I, which is in keeping with the previous results on secondary structure. In a P-turn encompassing residues 5-8, the corner residues 6 and 7 would not directly interact with the receptor. The turn would primarily be present to orient the NHZ-terminal pentapeptide residues and the COOH-terminal Trp-amide for correct interaction with the receptor. 5. Second messenger systems physiological actions

and

Many insects are well known for their impressive flight performance. These may be intermittent in nature, as in cockroaches or flies, or of long duration, as in locusts and migrating butterflies and moths. In any case, flight muscles have a high energy demand for their contraction, and it is thought that peptides of the AKIURPCH family are responsible for supplying them with fuel stored mainly in the fat body. Fuels for flight can be carbohydrates (as in cockroaches, bees, and flies), lipids (as in locusts and certain Lepidoptera), proline (as in tsetse flies and certain beetles), or mixtures of these fuels (173, 739). Although lipids are the main fuel for long-distance flight in locusts, carbohydrates are used during the initial phase of flight and still contribute substantially during the later phase (18). Mobilization of both substrates is controlled by AKHs. Prerequisites for mobilization of glycogen and triacylglycerol stores in the fat body are activation of the enzymes glycogen phosphorylase and triacylglycerol lipase. Activation of phosphorylase by low doses of synthetic Lom-AKH-I and -11 in vivo (162, 224) and by all three Lom-AKHs in vitro (722) has been shown in L. migratoria, but lipase activation is still mainly unresolved (641, 684). Which other steps between receptor binding and enzyme activation are involved in the signal transduction pathway? The first evidence that CAMP and Ca”’ were involved in the AKH-induced mobilization of lipids from fat body of locusts came from in vitro (641) and in vivo studies (161, 188). Exogenous dibutyryl CAMP mimics AKH action in vivo (188) and in vitro (641), and the levels of CAMP increase in the fat body after injection of synthetic Lom-AKHI (161) and natural Lom-AKH-I and -11(224). Synthetic LomAKH-I, -11, and -111have recently been shown to stimulate glycogenolysis in locust fat body by the formation of CAMP in an in vitro system (722). This and accompanying studies (709, 723) strongly suggested that AKHs in locust fat body activate adenylate cyclase via receptors coupled to a stimulatory G protein and that extracellular Ca” is essential for the formation of CAMP. However, not only is the influx of extracellular Ca’+ into the fat body increased by AKHs, but also the efflux of cytosolic Ca’+ from the fat body. This intracellular Ca2+ mobilization may be under the control

October

1997

INSECT

969

HORMONES

of inositol 1,4,5trisphosphate (IP,) (501, 708). It appears that for AKH action on phosphorylase and (possibly) lipase, multiple second messenger systems are necessary. It may well be that the different AKHs, which are known to vary in their activity in these systems, bind to different receptor subtypes and involve different second messengers (see sect. VI). Signal transduction for the hypertrehalosemic action of these peptides in cockroaches has also been extensively studied and is not reviewed in detail here. The reader is referred to an overview by Keeley et al. (332). Briefly, there are some major differences to the action of AKHs in locusts. First, CAMP does not appear to be involved, since this nucleotide is not increased by Pea-CAHI or -11 and Bld-HrTH in isolated fat bodies of the respective cockroaches (387, 488). There is, however, a very minor increase of CAMP in vivo upon Pea-CAH-I and -11 injection (163). Second, intracellular Ca”+ appears to play a more important role as second messenger for the hypertrehalosemic signal than extracellular Ca”’ (334). Third, the inositol polyphosphate-diacylglycerol transduction cascade does not seem to be involved in mediating the Bld-HrTH signal (387). With respect to the physiological actions of the AKIU RPCH family peptides, it is now well accepted that they are multifunctional and pleiotropic. The “classical” effect of the AKHs in locusts is lipid mobilization and, after a lipase is activated, triacylglycerols in the fat body are broken down to monoacylglycerols, which are subsequently reacylated to form stereospecific sn-1,2-diacylglycerols, and these are released from the fat body into the hemolymph (18). In the hemolyrnph, AKHs are responsible for an overall increase in the lipid-carrying capacity. For this, the predominant species of lipoprotein in resting locust hemolymph, high-density lipophorin, is loaded with the lipids released from the fat body and simultaneously associates with an apoprotein, apolipophorin III. The overall result is the creation of a larger but less dense particle, low-density lipophorin (see Ref. 319). The direct action of AKH on the utilization of fuel at the flight muscles has been previously discussed (221, 222). The pleiotropy of the AKHs in locusts is exemplified by their other, quite diverse functions. Protein synthesis is inhibited in migratory locusts in vivo (57, 58) and in isolated fat bodies (7). In the cricket, Acheta domesticus, a similar phenomenon is also observed upon injection of the cricket’s native Grb-AKH (91). Moreover, in S. gregaria, fatty acid synthesis is inhibited by AKH (220). Recent studies on L. migratoria have confirmed this effect using a newly developed more convenient and rapid assay method, such as inhibition of [14C]acetate uptake into synthesized lipids by small fat body pieces (385) or dispersed fat body cells (386). Finally, AKHs also inhibit RNA synthesis in vitro in locust fat body (352). In the latter two bioassavs, Lom-AKH-III is the most potent of the peptides.

In these assays, CAMP does not appear to be the second messenger, but neither is there direct evidence for a role of the inositol phosphate system (223). The main action of hypertrehalosemic hormones is to stimulate fat body glycogenolysis and make the resulting products available as precursors for the biosynthesis of hemolymph carbohydrate, which is mainly trehalose. In addition to this classical role, however, a number of other actions of hypertrehalosemic peptides are known. In fact, stimulation of the frequency of the heart beat in P. americana by CC extract and isolated fractions (702) was discovered before the hypertrehalosemic effect. As a result, the heart bioassay was used successfully to isolate and sequence the peptides Pea-CAH-I and -11 (15, 599). Similarly, isolation of the same peptides was achieved by monitoring increase in muscle tone and contraction frequency of the spontaneous activity of the isolated leg of a locust (491). Furthermore, in the cockroach, B. discoidalis, Keeley et al. (332) found that the endogenous peptide BldHrTH stimulates the biosynthesis of mitochondrial cytoheme a + b. The mechanism is not known yet, but gene expression is likely involved. In another case, it is known that Bld-HrTH induces gene expression for a cytochrome P-450 enzyme. In addition, this peptide stimulates the rate of protein synthesis in the fat body of adult female cockroaches, and it is thought that it acts in collaboration with juvenile hormone (see Ref. 332). It has been suggested that proline metabolism in the Colorado potato beetle is under hormonal control, but no endogenous peptide was known at that time (737). Recently, however, experiments on the fruit beetle, Pachnoda sinuata (L. Auerswald and G. Gade, unpublished data), and dune beetles of the genera Onitis and Scarabaeus (182,183) have shown that exogenous Mem-CC and Ona-CC-I and -11, respectively, increase the concentration of proline in the hemolymph. Proline is the most important flight substrate in these beetles and, concomitant with this hyperprolinemic effect, there is a decrease in the hemolymph alanine concentration. 6. Inactivation

and metabolism

The neuropeptide-mediated signal eventually has to be terminated by enzymatic degradation. Because members of the AKHRPCH family have blocked terminals that provide protection against normal exopeptidases, the initial enzymatic breakdown has to be by an endopeptidase. Such a membrane-bound endopeptidase has, for example, been found in the CNS of S. gregaria, cleaving the Asn3/ Phe4 bond of Lom-AKH-I (296). Both endogenous peptides of this locust, Lom-AKH-I and Scg-AKH-II, are also cleaved at this bond on the external surface of fat body cells by an endopeptidase that has characteristics similar to that of mammalian endopeptidase 24.11 (564). The fragments that cleavage generated are now susceptible to degrada-

970

GiiDE,

HOFFMANN,

tion by exopeptidase. In the desert locust, an aminopeptidase in the hemolymph is responsible for further degradation of the COOH-terminal fragments of both peptides. In the hemolymph of M. sexta, however, the endopeptidase is apparently a metalloprotease that very likely cleaves at the COOH-terminal side of Thr5 (149). In the cockroach, P. americana, homogenates of the fat body cleaved PeaCAH-II by the action of an endopeptidase (634). Pea-CAHI was removed by Malpighian tubules of P. americana from the incubation mixture in vitro, and the peptide was destroyed by a neutral metaloendopeptidase (16). Uptake and breakdown by isolated Malpighian tubules is also reported for S. gregaria (631, 632). The first step of proteolytic degradation appears to be by a postproline cleaving enzyme for Lom-AKH-I, whereas Scg-AKH-II, which does not contain the Pro” residue, is broken down by an endopeptidase similar to chymotrypsin, which cleaves between Phe” and Ser”. Further breakdown of the fragments which now have free NH2 and COOH termini, respectively, is achieved by exopeptidases of the aminopeptidase and carboxypeptidase A or B type, which have been demonstrated to be present in homogenates of Malpighian tubules. B. Diuretic

and Antidiuretic

Neuropeptides

As has been previously discussed, one of the difficulties in addressing the functions of putative hormones is that of pleiotropism. Furthermore, one can only detect the biological activity for which one is assaying, which may or may not be the primary function of the hormone. In recent years, this situation has been exacerbated with the discovery of dozens of new peptides, resulting in the phenomenon of “reverse endocrinology” in which we have many neuropeptides of known structure, but unknown function (371), or if the functions are known, they may not be the primary ones for the peptides. The problem becomes more acute in the case of the diuretic (DP) and antidiuretic peptides (ADP). Most bioassays for diuretic activity involve an in vitro preparation of the Malpighian tubules (MT) or hindgut, which may or may not reflect actual diuretic activity in the whole animal (reviewed by Refs. 513, 644). Indeed, Nicolson (473) has suggested that many diuretics may in fact simply act to increase the filtration rate of the hemolymph and may better be termed “clearance factors” than diuretics. The MT assays are further complicated by the fact that many insects have tubules with as many as four segments (252), some of which are secretory, some absorptive, and in the most extreme case, the New Zealand glowworm, Arachnocampa luminosa, one segment is a light-producing, organ (234). At the risk of some ambiguity, but for the sake of simplicity, in this review we refer to any factor that increases tubule secretion, inhibits hindgut reabsorption,

AND

SPRING

Volume

77

or increases whole animal water loss as a DP, and any factor that inhibits tubule function, promotes hindgut reabsorption, or inhibits whole animal water loss as an ADP. In general, the DPs, most of which have been bioassayed on MT, fall into two families: those with varying degrees of homology to the vertebrate corticotropin-releasing factor (CRF-DPs; Ref. 329) and the smaller kinins, first isolated as myotropins (251,280; see sect. 1~23). There are, of course, exceptions. Proux and co-workers (528, 530) isolated a DP related to the arginine vasopressin (AVP) family of peptides from the suboesophageal and thoracic ganglia of L. migratoria. Davies et al. (97) reported that the cardioacceleratory peptide CAPzb acts as a DP in D. melanogaster, and Maddrell and co-workers (412, 413) have shown that the biogenic amine serotonin, in addition to its known pharmacological effects on tubule secretion, may act as a true diuretic factor in the bloodsucking bug Rhodnius prolixus (13). 1. Diuretic

pep tides

To understand the ways in which DPs can increase fluid transport by the MT, some background on the formation of the primary urine is necessary. Insects have an open circulatory system so that the blood and the extracelluar fluid are intermixed to form the hemolymph. The tracheal system obviates the need for the hemolymph to function in gas transport, so circulation rates are highly variable and there is normally little or no blood pressure. For these reasons, primary urine formation is by secretion rather than filtration. The basics of MT secretion have been known for more than 40 years. The blind-ended MT secrete a cation-rich (usually predominantly K’) solution that is essentially isosmotic with the hemolyrnph (557). Water and hemolymph solutes follow passively, with some active transport of potentially toxic substances (405, 407). O’Donnell and co-workers (483,484) showed that in R. prolixus, even under conditions of maximum urine production where the equivalent of the entire hemolymph volume is excreted every 30 s (407), a gradient of 3.3 mosM (- 1% of hemolymph osmotic pressure) was sufficient to drive fluid transport using transcellular mechanisms alone. Until recently, the biggest unknown regarding MT secretion was the presence of the “common cation transporter” (405), which was presumed to be present on the apical membrane and which could actively transport either Na’ or K’. With the recent work of Wieczorek and others (349, 627, 744-746), it has become clear that the driving force for MT secretion is in fact a vacuolar-type Hf-ATPase located on the apical membrane. Sodium and K’ transport are thus secondarily active, involving either a differentially selective or, more likely, two discrete antiports for Na+/H’ and K+/H’ exchange. Moreover, this antiport mechanism clears up two previously unanswerable

October

1997

INSECT

HORMONES

questions, namely, 1) why do unstimulated tubules transport Kf as the predominant cation, when this acts counter to ionic homeostasis of the hemolymph, and 2) why is there a shift in the urine Na’/K’ after stimulation, even in non-blood-feeding insects (e.g., Refs. 648, 475)? It would appear that the concentration of K’ available to the apical antiport, normally at least lo-fold greater than that of Na+, can account for the Kf-rich transport in unstimulated MT, rather than dietary influence, as was previously proposed (65, 512). Normal entry of ions into MT cells is via a Na’K’-Xl cotransporter on the basolateral membrane. The cotransporter from M. sexta has recently been cloned and its primary structure determined (567). In addition, each ion species is also presumed to have an independent entry pathway into the cells (409). Intracellular Na’ concentration is normally on the order of 12 mM compared with 140 mM for Kf (236), so any increase in the rate of ion uptake by this pathway could be expected to exert a disproportionate effect on the Na’/H+ antiport, even in phytophagous insects. In blood-feeding, and perhaps some other insects, increasing the permeability of the basolatera1 membrane to Na+ would produce both increased fluid transport and a shift in the urine Na+/K+, as was observed by Maddrell and Overton (410) when they blocked the action of the Na’-Kf-ATPase in R. prolixus tubules. It has been demonstrated in both Aedes (24, 503) and D. mezanogaster (481) that the rate of fluid transport by the tubules is also regulated by a Cl- shunt pathway, via the stellate cells. Increasing the permeability of the tubules to this anion acts to increase fluid transport, even when cation transport is unaltered (24). Coast et al. (85) have proposed a similar mechanism in A. domesticus, although the absence of stellate cells renders the location of the shunt somewhat problematic. Finally, it is theoretically possible to alter fluid transport by changing the permeability of the MT to water, as happens in the collecting duct of the vertebrate kidney. The permeability of the MT to water is already high (406), however, and does not appear to be altered in response to putative hormones or second messengers. All of the known diuretic factors are peptidergic and thus operate via second messengers. Given the multiplicity of pathways by which fluid transport might be regulated, it is not surprising that nearly every known second messenger system has been implicated in the control of MT function. In every insect species tested, CAMP increases MT secretion (reviewed in Refs. 24, 474), and in those species for which data are available, CAMP preferentially increases Na’ transport. In Aedes, CAMP acts by increasing the basolateral Na’ conductance of the primary cells (597, 750), mimicking the mosquito natriuretic peptide (MNP), a member of the CRF-DP family. In this species, CAMP is believed to act by stimulating the bumetanide-sensitive Na’-K’-Xl cotransporter located on the basolateral membrane [253). In the tsetse flv, CAMP mav

971

also increase the Cl- conductance of the primary cells to enhance salt and fluid transport (298), although this is clearly not the ease in Aedes (502), A. domesticus (83), or D. melanogaster (97). In the latter two species, CAMP may act to open a serosal Na’ channel unrelated to the Na’-Kf-Xl cotransporter. The regulation of anion conductance appears to be regulated by intracellular calcium concentrations ([Ca”‘]i) in both A. domesticus (85) and D. melanogaster (97, 112, 481); in Aedes, it is known only that the Cl- shunt is not affected by CAMP (24). The mechanisms must be different; in D. melanogaster, as in Aedes, Cl- movement into the tubule is via a shunt pathway presumed to be in the stellate cells. In opposition to the stellate cell shunt, Beyenbach and co-workers (25, 732) have recently provided evidence for a Ca”+-mediated increase in a Cl--sensitive paracellular pathway that increases the permeability of the MT to such extracellular space markers as sucrose and inulin. In contrast, A. domesticus lacks stellate cells so the Cl- shunt must somehow operate within the primary cells. In the ant, Formica, which has morphologically homogenous MT, the low-resistance (shunt) Cl- transport pathway appears to be transcellular and mediated at least in part by the Na’-Kf2Cll cotransporter (106). Interestingly, both serotonin (5hydroxytryptamine, 5-HT) and CAMP stimulate fluid transport in larval A. aegypti independently of the Naf-K+-2Cl cotransporter (77). In D. melanogaster, the nitric oxide/ guanosine 3’,5’-cyclic monophosphate (cGMP) pathway has been shown to increase tubule secretion independently of CAMP or [Ca”‘], , but the ionic events involved are unknown (97, 112). There have been some contradictory data on the role of Ca”’ in MT secretion. Clark and Spring (78, 646) found that increasing [Ca”‘]i in A. domesticus with the ionophore A-23187 slowed or stopped MT secretion. They were using whole MT/ampulla preparations (648), however, which have four discrete regions, two of which are presumed to be resorptive (252, 651), so the results could be explained by increased reabsorption rather than inhibited secretion. There is a family of DPs that show their greatest struct,ural relationship to the vertebrate CRF/urotensin/ sauvagine family of peptides and are therefore known as the CRF-like diuretic peptides (CRF-DP; Ref. 620; Table 2). The first of these was isolated from 10,000 trimmed heads of pharate adults of M. sex&, and called Mas-DH (also Mas-IDH; in this article, Mas-DPl; Ref. 324). In the ensuing 6 years, a second CRF-DP has been isolated and sequenced from M. sexta (Mas-DP2; Ref. 26) as well as new CRF-DPs from A. domesticus (Acd-DP; Ref. 329), L. migratoyia (Lom-DP; Refs. 331, 390), I? americana (PeaDP; Ref. 330), one common to both Musca domestica and Stomoxys calcitrans (Mud-DP; Ref. 79), and Tenebrio molitor (Tern-DP; Ref. 160). The latter is unique in having a nonamidated COOH terminus that is more hydrophilic than the other CRF-DPs (160). Bevenbach (24) reports

972

GiiDE,

TABLE 2. Primary Peptide

Mas-DP-1 Mas-DP-2 Acd-DP Lom-DP

structures

of the diuretic

DP, diuretic transport peptide;

peptide, CRF, corticotropin-releasing CHH, crustacean hyperglycemic

Tern-DP CRF AVP-IDH AVP Scg-ITP

and antidiuretic

SPRING

peptides,

Volume

including

selected noninsect

Sequence

Cama-CHH

Mud-DP

AND

Species

Manduca sexta Ad? sexta Acheta domesticus Locus ta migratoria Periplaneta americana Musca domestica Stomoxus calci trans Tenebrio moli tor Rat L. migratoria Vertebrates Schistocerca gregaria Carcinus maenas

Pea-DP

HOFFMANN,

77

peptides Reference

No.

RMPSLSIDLPMSVLRQKLSLEKERKVHALRAAANRNFLNDIamide SFSVNPAVDILQHRYMEKVAQNNRNFLNRVamide TGAQSLSIVAPLDVLRQRLMNELNRRRMRELQGSRIQQNRQLLTSIamide MGMGPSLSIVNPMDVLRQRLLLEIARRRLRDAEEQIKANKDFLQQIamide

324 27 329 331, 390

TGSGPSLSIVNPLDVLRQRLLLELARRRMRQSQDQIQANREILQTIamide

330

NKPSLSIVNPLDVLRQRLLLEIARRQMKENTRQVELNRAILKNVamide

79

SPTISITAPIDVLRKTWEQERARKQMVKNREFLNSLN EEPPISLDLTFHLLREVLEMARAEQLAQQAHSNRKLMEIIamide CLITNCPRGamide CYFQNCPRGamide SFFDIQCKGVYDKSIFARLDRICEDCYNLFREPQLHSLCRSDCFKSPYFKGCL QALLLIDEEEKFNQMVEILamide pQIYDTSCKGVYDRALFNDLEHVCDDCYNLYRTSYVASACRSNCYSNLVFRQC MDDLLMMDEEDQYARKVQMVamide

160 587 529

factor; hormone.

AVP, arginine

that MNP has been isolated from Aedes and is also a member of this family; however, structural data are as yet unavailable. The CRF-DPs range in size from 30 to 46 amino acids, and sequence homology is relatively low, -50% in most cases, although Coast et al. (85) suggest that nucleotide homology may be much higher, perhaps 7080%. In all cases, gaps are required in the sequences to show maximal homology with the others of the family. The mode of action of the CRF-DPs appears to be relatively uniform across species, and there are varying degrees of interspecific bioactivity (83). In most cases, interspecific activity is relatively low, so, for example, Mas-DPl produces only 60% of the maximal response in A. domesticus MT (83), as might be expected given the relatively low sequence homology between peptides (85). The exception is M. sexta, in which all the known CRFDPs (except Tern-DP) are roughly equipotent when tested on larval MT (9). The CRF-DPs activate an adenylate cyclase in the primary cells of the MT, elevating CAMP levels in these cells (8, 9, 85). As indicated above, the action of CAMP in the primary cells is to increase Na+ conductance across the basolateral membrane. As Na’ enters the cells from the hemolymph, it must temporarily exceed the capacity of the ubiquitous Na’-Kf-ATPase to extrude it, thus increasing intracellular Na+ concentration. This presents the basal Na’/H+ antiport with a higher concentration of substrate, thereby increasing cation transport, with water and other solutes following passively (24). In the tsetse fly, CAMP appears to transiently increase MT cell volume by increasing basolateral permeability to both Na’ and Cl-, with cell volume returning to normal as the transcellular movment of fluid increases (299). It is also worth noting that ultrastructural studies show that CAMP also affects mitochondrial location and action. In R. prolixus,

vasopressin;

AVP-IDH,

427, 428 335

AVP-like

insect

diuretic

hormone;

ITP, ion-

Bradley and Satir (40,41) showed that mitochondria move into the microvilli of the brush border in response to serotonin (which acts via CAMP; Ref. 435), and Spring and Felgenhauer (647) have shown both intracellular movement and changes in electron density and cristae structure of mitochondria after CAMP treatment in A. domesticus. It is not unreasonable to assume, therefore, that CAMP (and the CRF-DPs) also increases the activity of the apical Hf-ATPase through increased ATP concentrations, further stimulating cation movement across the apical membrane. There is recent electrophysiological evidence that in D. melanogaster, CAMP can act to stimulate the H’ATPase directly (481), although this has not yet been demonstrated in other species. There are relatively few studies of the action of the CRF-DPs in vivo. In addition to increasing MT secretion and CAMP production in vitro, Lom-DP increased the rate of amaranth clearance and temporarily reduced hemolymph volume in L. migratoria. Both in vitro and in vivo responses were blocked by a polyclonal antiserum to Lom-DP (509; reviewed by Ref. 81). Diuretic peptides have frequently been championed as candidates for environmentally friendly pesticides. A recombinant Bombyx mori baculovirus expressing Mas-DPl was shown to have an enhanced lethal effect when injected into larval B. mori, with the increased mortality associated with a significant reduction in hemolymph volume (414). Last instar H. virescens injected with pharmacological doses (500 pmol/ insect) of Mas-DPl reduced both water excretion and food consumption by 70%. Extremely high doses (~500 nM) also inhibited MT secretion in vitro (333). Members of the second major family of DPs were initially isolated on the basis of their myotropic activity on the hindgut of L. maderae (282), and named the insect

October

1997

INSECT

HORMONES

kinins. There are currently more than 20 kinins (Table 3), 8 from L. maderae (leucokinins; Refs. 276, 277, 280, ZSl), 5 from A. domesticus (achetakinins; Ref. 283), 1 from L. migratoria (locustakinin; Ref. 614), 3 from Culex (culekinin depolarizing peptides; Refs. 85,248), 3 from A. aegypti (aedeskinins or aedes leucokinins; Ref. 715), and 3 from Helicoverpa (helicokinins; Ref. 29). These peptides are small, 6-13 residues, and have a highly conserved COOHterminal pentapeptide sequence, FX,X,WGamide, where X1 is S, H, N, or Y, and X2 is S or P. In addition to their myotropic action, the kinins may be involved in the release of digestive enzymes into the gut (G. M. Holman, personal communication). Their most important action, however, appears to be in their diuretic action on the MT. The effect of the kinins on MT secretion is both independent of and additive to the action of the CRF-DPs. Using their elegant perfused tubule preparation, Beyenbath and co-workers (24, 251, 503) have shown that several of the kinins, but specifically leucokinin-VIII, act to depolarize A. aegypti MT. A low-resistance Cl- pathway transforms the MT from an electrically “tight” to a “leaky” epithelium. The change in Cl- conductivity does not occur within the primary cells and is presumed to occur in the stellate cells found in the MT of A. aegypti and D. melanogaster. The architecture of the stellate cells is such that direct intracellular microelectrode measurements are impossible. The change in conductivity is called the Clshunt because it bypasses the primary cells. In D. melanogasteq which has MT tubules similar in structure to A. aegypti, the leucokinins also stimulate fluid transport, mediated by [Ca”‘], , and activating a similar shunt pathway (112, 113, 481). It appears that by increasing the permeability of the tubules to anions, the driving force for cation transport is reduced, enhancing overall fluid transport. As discussed above, there is still some controversy over whether the shunt pathway is located within the stellate cells (111, 481) or is a low-resistance, Cl--selective paracellular pathway (25, 732). The two systems are independent and additive (24, 84, 112), although how they are integrated in the intact insect is unknown. The situation in A. domesticus is less clear. Malpighian tubules from A. domesticus are divided into three discrete regions, each of which consists of a single cell type (252,649). Each region responds differently to secretagogues, including the achetakinins (341, 650). Coast et al. (81, 85) propose that the achetakinins increase [Ca”‘]i and activate a C-kinase, which acts in concert to increase Cl- conductance in A. domesticus tubules. This would not be a shunt pathway, since it acts to alter the electrical properties of the primary cells, but the net effect should be similar to that observed in those species that do utilize a Cl- shunt. Spring and Kim (650) reported similar results for the mid MT of A. domesticus, although the achetakinins had a uniformly inhibitory effect on the distal MT. In both cases, distal and mid, CAMP either added to the neu-

973

tral or stimulatory effects of the kinins (mid) or overcame the inhibitory effects (distal), again demonstrating that different pathways are involved. Very recent electrophysiological evidence suggests that the unstimulated dist#al MT are already highly permeable to Cl (M. J. O’Donnell and J. H. Spring, unpublished data). In A. domes ticus, achetakinin-like immunoreactivity is concentrated almost exclusively in the retrocerebral complex and showed a Ca2+-dependent release in highK+ saline in vitro. Hemolymph titers of achetakinin-like immunoreactive material were determined to be 2.8 nM and increased IO-fold after 48 h of starvation (74, 81). Proux and co-workers (528, 530, 621) first isolated and sequenced a peptide with diuretic activity from 51,000 ganglia of L. migratoria. During their isolation procedures, they discovered two fractions, designated F1 and F2, with identical amino acid sequences and molecular masses of 700 Da and 1.47 kDa, respectively. Both were immunoreactive for antisera to mammalian AVP, and the active fraction, F2, proved to be the antiparallel dimer of F1. This peptide, named AVP-like insect diuretic hormone (AVP-IDH) for its immunoreactivity, was a milestone in the field inasmuch as it was the first DP to be isolated, sequenced, and synthesized. From the beginning, however, there were indications that AVP-IDH represent,ed only part of the story. Picquot and Proux (517) conducted a series of in vivo experiments in which they monitored the AVP-IDH concentration in the hemolymph using radioimmunoassay (RIA) and simultaneously monitored urine output. There was indeed one peak of urine output that matched a hemolymph peak of AVP-IDH, but there were other peaks of diuretic activity for which there was no corresponding peak of AVP-IDH. In examining the data, one might speculate that AVP-IDH was released into the hemolymph at subjective dawn rather than being released in response to water loading due to food ingestion. This, of course, begs the question of whether the diuretic effect of AVP-IDH is actually its primary function. More confusing in this regard, the in vitro bioassay Proux used depends on the maintenance of fluid secretlion of L. migratoria MT, i.e., control MT show a steady decrease in urine production, whereas “stimulated” MT do not; however, at no time is the secretion rate elevated above the initial rate (531,619). To do a direct comparison between AVP-IDH and Lom-DP, Coast et al. (86) utilized their standard Ramsay (557) preparation with L. ~@-cx toria MT and tested both AVP-IDH (F,) and Lam-DP at physiologically relevant concentrations (lo--” M). Lom-DP increased urine output fivefold, whereas AVP-IDH had no effect. At 5 x lo-* M, Lom-DP increased secretion by >1,500 pl/min, whereas AVP-IDH had no effect on secretion even at concentrations of lo- M. Proux and Herault (529) had also shown that AVP-IDH elevated MT CAMP concentrations, which would imply that it acts via the same pathway as the CRF-DPs. Again, using parallel ex-

974 TABLE

GiiDE,

3. Primary

structures

Peptide

Section

IIIA: proctolin and cardioacceleratory peptides

of the neuropeptides

Proctolin Pea-corazonin Scg-corazonin Cama-CCAP

IIIC: sulfakinins

(CAP,,)

Leucokinin I Leucokinin II Leucokinin III Leucokinin IV Leucokinin V Leucokinin VI Leucokinin VII Leucokinin VIII Achetakinin I Achetakinin II Achetakinin III Achetakinin IV Achetakinin V Culekinin I Culekinin II Cule kinin III Aedeskinin I Aedeskinin II Aedeskinin III Helicokinin I Helicokinin II Helicokinin III Locustakinin Lem-SK-I Lem-SK-II Lom-SK Pea-SK Neb-SK-I

Neb-SK-II

IID : pyrokinins/ myotropins

II@: tachykinins

Drm-SK-II Gastrin CCK-8 Lem-PK Lom-PK-I Lom-PK-II Lom-MT-I Lom-MT-II Lom-MT-III Lom-MT-IV Hez-MT-I Hez-MT-II Born-cu-SGNP Born-/?-SGNP Born-y-SGNP Lom-TK-I Lom-TK-II Lom-TK-III Lom-TK-IV Lom-TK-V Cav-TK-I Cav-TK-II Cus-TK-I Cus-TK-II

affecting

AND

SPRING

P. americana P. americana S. gregaria L. migratoria M. sexta T. molitor S. eridania M. sexta L. maderae L. maderae L. maderae L. maderae L. maderae L. maderae L. maderae L. maderae A. domesticus A. domesticus A. domesticus A. domesticus A. domesticus C. salinarius C. salinarius C. salinarius A. aegypti A. aegypti A. aegypti H. xea H. xea H. xea L. migratoria L. maderae L. maderae P. americana L. migratoria P. americana N. bullata D. melanogaster C. vomi toria L. cuprina N. bullata C. vomi toria L. cuprina D. melanogas ter Human Sheep L. maderae L. migratoria L. migratoria L. migratoria L. migratoria L. migratoria L. migratoria H. xea H. xea B. mori B. mori B. mori L. migratoria L. migra toria L. migratoria L. migratoria L. migra toria C. vomi toria C. vomi toria C. salinarius C. salinarius

Volume

77

muscle activity

Species

CW2b

IB: myokinins

HOFFMANN,

Sequence

RYLPT pQTFQYSRGWTNamide pQTFQYSHGWYNamide PFCNAFTGCamide

pQLYAFPAVamide DPAFNSWGamide DPGFSSWGamide DQGFNSWGamide DASFHSWGamide GSGFSSWGamide pQSSFHSWGamide DPAFSSWGamide GADFYSWGamide SGADFYPWGamide AYFSPWGamide ALPFSSWGamide NFKFNPWGamide AFHSWGamide NPFHSWGamide NNANVFYPWGamide WKYVSKQFFSWGamide NSKYVSKQKFYSWGamide NPFHAWGamide NNPNVFYPWGamide YFSPWGamide VRFSPWGamide KVKFSAWGamide AFSSWGamide EQFEDY(S03H)GHMRFamide pQSDDY(S03H)GHMRFamide pQSDDYGHMRFamide pQLASDDY(S03H)GHMRFamide EQFDDY(S03H)GHMRFamide FDDY(SOBH)GHMRFamide

??EEQFDDY(S03H)GHMRFamide GGEEQFDDY(SO,H)GHMRFamide GGEEQFDDY(?)GHMRFamide GGDDQFDDY(S03H)GHMRFamide pQGPWLEEEEEAY(S03H)GWMDFamide DY(S03H)MGWMDFamide pQTSFTPRLamide pQDSGDGWPQQPFVPRLamide pQSVPTFTPRLamide GAVPAAQFSPRLamide EGDFTPRLamide RQQPFVPRLamide RLHQNGMPFSPRLamide MEFTPRLamide TMNFSPRLamide IIFTPKLamide SVAKPQTHESLEFIPRLamide TMSFSPRLamide GPSGFYGVRamide APLSGFYGVRamide APQAGFYGVRamide APSLGFHGVRamide ?PSWFYGVRamide APTAFYGVRamide GLGNNAFVGVRamide APSGFMGMRamide APWGFTGMRamide

Reference

No.

657 712 714 654 388 159 159 290 276 276 277 277 280 280 281 281 283 283 283 283 283 85 85 85 715 715 715 29 29 29 614 450 447 713 607 713 146 471 124 124 146 124 124 472 23 108 279 609 612 611 606 610 610 98 98 328, 596 328, 596 328, 596 605 605 604 604 616 401 401 80 80

October 1997 TABLE

975

HORMONES

3-Continued Section

periviscerokinin IIIG : accessory gland myotropins IIIF:

other myotropins

IIIH:

IHI:

INSECT

myoinhibitory peptides and FaRPs

Peptide

Pea-VK Lom-AG-MT-I Lom-AG-MT-II Mud-AG-MT Mas-MG-MT-I Mas-MG-MT-II Led-OVM Lom-MIP Mas-MIP-I Mas-MIP-II Lom-MIH Lem-MS Scg-FLRFamide Lom-MS Mas-FLRFamide-I Mas-FLRFamide-II Mas-FLRFamide-III Neb-MS Lom-FaRP-I Lom-FaRP-II Lom-FaRP-III Aea-HP-I Aea-HP-II Cav-FMRFamide-I* Drm-FMRFamide-I* Drm-FMRFamide-Ia*

Species

P. L. L. M. M. M. L. L. M. M. L. L.

americana migratoria migratoria domes tica sexta sexta decemlineata migratoria sexta sexta migratoria maderae

S. gregaria L. L. M. M. M. N. D. L. L. L. A. A. C. L. D. D.

migratoria migratoria sexta sexta sexta bullata melanogaster migratoria migratoria migratoria aegypti aegypti vomitoria cuprina melanogaster virilis

Sequence

GASGLIPVMRNamide GFKNVALSTARGFamide AHRFAAEDFGALDTA LLNALPLDALSSLTGamide AGPYTamide DIPPRamide IAYKPEamide AWQDLNAGWamide AWQDLNSAWamide GWQDLNSAWamide pQ?Y?KQSAFNAVSamide pQDVDHVFLRFamide PDVDHVFLRFamide ADVGHVFLRFamide pEDWHSFLRFamide GNSFLRFamide DPSFLRFamide TDVDHVFLRFamide GQERNFLRFamide A??RNFIRFamide AFIRFamide PQRPHypSLKTRFamide TRFamide SVQDNFIRFamide SVKQDFMHFamide SLKQDFMHFamide

Reference

No.

525 499 498 724 769 768 642 608 28 28 617 278 588 613 506, 613 345 344 344 145 470 380 380 380 419 419 120 115 676 676

A question mark in a sequence indicates that these residues were not identified. * There are currently 26 known Cav-FMRFamides (CalliFMRFamides) and 14 of the closely related Drm-FMRFamides (droFMRFamides). Example structures for the FMRFamide-I variations (following numbering system of Ref. 180) from 4 Dipteran species are given for comparative purposes. For sequences of all these peptides, see Reference 180.

periments, Coast et al. (86) found no stimulator-y effect on CAMP levels with concentrations of AVP-IDH ranging from lo-l1 to 10e7M. In contrast, Lom-DP produced significant increases in CAMP levels at lo-’ M with a peak response at 5 X low8 M. Clearly then, AVP-IDH is not a primary messenger for L. migratoria MT, and its true role in homeostasis remains in question. For the present, AVPIDH remains the only AVP-like peptide isolated and sequenced, although immunoreactive material has been found, primarily in the subesophageal ganglia (SOG), in four orders of insects (527). The biogenic amine 5-HT has long been known to be a potent in vitro stimulant of MT secretion, particularly in the blood-sucking bug R. prolixus (412, 413). Initially considered to be a pharmacological agent, recent evidence suggests that 5-HT acts as a second DP in R. prolixus, promoting diuresis by elevating intracellular levels of CAMP in the MT (13). Serotonin acts synergistically with the presumptive DP found in the mesothoracic ganglionic mass, and blocking the release of 5-HT with the neurotoxin 5,7-dihydroxytryptamine (5,7-DHT) delays or completely blocks the normal postprandial diuresis of fifth-instar R. prolixus (408). Serotonin is known to stimulate the NaCl-rich fluid absorption by the midgut (135),

the plasticity of the abdominal cuticle (574), and KC1reabsorption by the lower tubule in R. prolixus (411), so the possibility is raised that 5HT is the primary messenger in the suite of responses that accompany feeding with the magnitude of the diuretic response being modified by the presumptive DP. A series of low-molecular-weight cardioacceleratory peptides (CAP) has been isolated from 1M. sexta (69% 700). About 1 kDa in size, they fall into two classes, CAP, and CAPZ. The first one identified was identical to CCAP, and all the CAP, peptides show strong sequence homology to CCAP, whereas the CAP2 peptides do not, although all are cardioacceleratory (Table 3). A novel octapeptide, CAPZbappears also to be present in D. melanogaster and stimulates MT secretion in vitro via the cGMP pathway (97). The stimulatory action of CAPZbis additive to that of CAMP but not cGMP, and leucokinins IV and VI are additive to both CAMP and cGMP, suggesting that, in D. melanogaster at least, there are three discrete stimulatory pathways for the MT. Both CAMP and cGMP are presumed to affect ion transport by the primary cells, whereas the leucokinins act, via Ca2+, on the Cl- shunt (97). Extracts of the thoracic ganglion stimulate fluid secretion to a greater extent than either CAMP or cGMP (112, 113), so

976

G;inE,

HOFFMANN,

it appears that two DPs, one unknown, but likely of the CRF-DP family, and CAPZI, act on cation transport, and presumably kinins regulate the shunt pathway. This is the most complex system for tubule regulation known to date; however, D. rnelcrnogaster MT are unusual in a number of ways, including their prodigious capacity for fluid transport and their longevity in vitro (112). Reagan (566) has isolated the receptor for Mas-DPl by expression cloning in COS-7 cells. The receptor appears to have the seven transmembrane domains common to other G protein-coupled receptors and is therefore presumed to act via CAMP. Reagan et al. (566, 569) also showed that Mas-DP2, Acd-DP, Lom-DP, and Pea-DP bind to these receptors and stimulate them to induce CAMP production by the COS-7 cells. This adenylate cyclase activat,ion is in accord with what is known about the action of the CRF-DPs (see above) but is somewhat surprising given the 44-63% homology among the CRF-DPs (2633% for the much shorter Mas-DP2; Ref. 9). Sequence homologv between the cloned Mas-DPl receptor and the cloned human CRF receptor (69) is 31%, which would be consistent with the 20-30% homology of the CRF-DPs to their vertebrate counterparts (82). More recently, using degenerate nucleotides from the M. sex& receptor, Reagan (568) isolated and cloned the CRF-DP receptor from A. domestiC2ls. The Acd-DP receptor has 441 residues, with 7 putative transmembrane domains, and shows 53 and 38% sequence homology with recept,ors for Mas-DPl and human CRF, respectively. When the Acd-DP receptor was expressed in COS-7 cells, it bound Acd-DP and activated adenylate cyclase. In contrast to the Mas-DPl receptor, which is activated by all the CRF-DPs, the Acd-DP receptor is 10:’ to 10’ times more sensitive to Acd-DP than to any of the other peptides, when assayed by CAMP production. Also, unlike the MasDPl receptor, the Acd-DP receptor has a relatively long extracellular NH,-t)erminal domain that contains a putative signal sequence, which may account for its increased binding specificity. Chung et al. (75) recently used an iodinated analog of achetakinin-2 to investigate the properties of the kinin binding sites on membranes isolated from A. domesticus MT. They found that binding was rapid, reversible, and specific, and saturation studies showed that there is a single class of binding sites. Kinin concentrations required to displace 50% of the bound radiolabeled ligand were three to five orders of magnitude greater than the concentrations required for half-maximal stimulation of MT fluid secret,ion, implying that loo-fold. Oxidation of the Met residues to methionine sulfoxide markedly enhanced pheromonotropic activity, presumably due to the increased stability of the sulfoxides in the hemolymph (457). Recently, a pseudotetrapeptide analog in which the phenyl ring of the Phe side chain in the original FTPRLamide core pentapeptide is replaced with the hydrophobic cagelike o-carborane moiety (2o-carboranyl-ethanoyl-TPRLamide) was found to be lofold more potent eliciting pheromone production in H. virescens than the much larger 33-mer Hez-PBAN (454). Nachman et al. (453) demonstrated that a type I p-turn in the COOH terminus was important for receptor recognition when testing a cyclic analog of the COOH-terminal hexapeptide of PBAN. Similar results involving NMR techniques have been reported using the hexapeptide COOH-terminal fragment of PBAN (731) or the complete PBAN molecule (76). An internal pentapeptide fragment of Hez-PBAN, which was amidated at its COOH terminus (YRQDPamide), showed high activity at a dosage of 1 pmol but was inactive at 100 and 1,000 pmol (551). The different pharmacological profile displayed by this fragment was interpreted as suggesting

October

INSECT

1997

that two (555).

different

3. Disttibution

types

of receptors

of PBAN and PBAN-like

may be present

activity

Antisera prepared against colloidally adsorbed synthetic PBAN revealed the presence of three clusters of cells along the ventral midline of the SOG (343, 404). These clusters occupy the presumptive mandibular, maxillary, and labial neuromeres and contain 4, 1% 14, and 4 immunoreactive cells, respectively. Cells from the labial cluster project axons to the CC and aorta, reinforcing the idea that PBAN acts as a neurohormone (404, 555). In addition, immunoreactivity can be detected in a pair of cell bodies in each thoracic and abdominal ganglion, and two pairs of axons originating from cells in the SOG extend the entire length of the ventral nerve cord, ending in the terminal abdominal ganglion (TAG; 404). Using a highly specific antiserum to Hez-PBAN in an ELISA, Gazit et al. (207) showed detectable levels of PBAN-like immunoreactivity in head extracts of fourth-instar larvae of Helie this peltigera. Levels of immunoreactivity were essentially identical in both sexes and were -5 pmol/insect in 3- and T-/-day-old moths. Immunoreactivity was also detected in three other Noctuid moth species, Helicoverpa armigera, Corwutiplusia circumcflexa, and Spodoptera littoralis. More recently, using a Western-blot technique, Jacquin-Joly and Descoins (307) were able to detect PBAN-like immunoreactivity in the moths H. xea, Mamestra brassicae, S. littoralis, Spodoptera latifascia, Spodoptera descoinsi, and Eldana saccharina and in the cabbage white butterfly Pieris brassicae, which does not use pheromones for mate attraction. This indicates that not only are the PBANs widely distributed among the Lepidoptera, but that they may also be pleiotropic, having important functions other than the regulation of pheromone biosynthesis. This would be consistent with the myotropic effects discussed above. Using a competitive binding assay for Hez-PBAN, Rafaeli et al. (546) showed that in H. amzigera, higher levels of immunoreactivity were observed in brain-SOG complexes, CC, and thoracic and abdominal ganglia during the fourth to fifth hour after scotophase, compared with mid-photophase. In contrast, the TAG accumulated high levels of PBAN during the photophase, and these dropped during scotophase, suggesting that the TAG may be the hormone release site. No PBAN could be detected in the hemolymph, however. Q. Regulation

of pheromone

synthesis

by PBAN

The effects of PBAN on the enzymes involved in pheromone biosynthesis have been followed in only a few species. It appears that in Argyrotaenia velutinana (679), H. xea (315), and M. brassicae (306), PBAN controls biosynthesis by regulating a step in, or before, fatty acid biosynthesis. In contrast, it has been proposed that PBAN

985

HORMONES

activates a S-11 desaturase in Chrysodeixis chalcites (3), whereas in S. littoralis (134, 416), B. mori (6, 495), and l%aumetopoea pityocampa (133), it appears to act on the reduction of fatty acyl moieties. More recently, using radiolabeled precursors and specific inhibitors, Ozawa and Matsumoto (496) have shown that PBAN regulates the synthetic step catalyzed by acyl CoA reductase and that palmitoyl CoA could be the substrate of the reductase. 5. Signal

transduction

Since the first bioassays, which showed that PBAN could increase pheromone production when injected into the hemocoel(553), PBAN has been assumed to act as a hormone, i.e., to be transported via the hemolymph (549). As such, PBAN would have to act via a second messenger system, such as CAMP, cGMP, or Ca”%almodulin. The presence of PBAN in the TAG and its accumulation during photophase suggested that direct innervation may be important in regulating pheromone synthesis (546); however, hemolymph transport of PBAN was supported by the studies that showed that severing the ventral nerve cord or removing the TAG did not block the action of PBAN (552). The primary difficulty in showing that PBAN is a true neurohormone was that it was not possible to detect PBAN activity in the hemolymph reproducibly (343, 546, 681). Initially, PBAN activity was believed simply to be CAMP mediated, since CAMP analogs can stimulate pheromone synthesis (143,315, 547). More recent studies, however, have shown that signal transduction is much more complex, involving both stimulatory and inhibitory signals. Matsumoto et al. (424) found that in S. litura, CAMP analogs, cGMP analogs, adenylate cyclase activators, and C-kinase activators, tested individually, failed to stimulate pheromone production, whereas a calcium ionophore increased biosynthesis. They interpret these data to mean that Ca”%almodulin and phosphoprotein phosphatase are involved in signal transduction. The stimulatory action of calcium ionophore and the inhibition of PBAN activity by a calcium inhibitor were confirmed in another species, Ostrinia nubilalis (403). In H. armigera, several second messenger systems seem to be involved since, although PBAN activity is also calcium dependent in this species, at high concentrations (20 pmol), a calcium independent pheromonotropic response was observed (639). Further experiments suggested that C-kinase activity is calcium dependent and that adenylate cyclase activity is also calcium dependent, since CAMP levels increased in pheromone glands in response to calcium ionophore. Finally, the biogenic amine octopamine can inhibit both pheromone biosynthesis and CAMP production by the pheromone glands (545). These authors suggest that because the females of H. arrnigera usually undergo a period of

986

GADE,

HOFFMANN,

flight at the onset of scotophase, before pheromone signaling, the corelease of AKH and octopamine to provide fuel for flight (487) would inhibit pheromone production until the female settles for mating. Clearly, this is a complex system deserving of further study.

B. Juvenile Hormones AllatotropinsAllatostatins 1. Chemistry

and

and biosynthesis

AND

SPRING

Volume

0

0

of juvenile

/

/

77

JH 0

JH I

hormones /

0 41MeJH I The juvenile hormones (JH) are unique humoral agents, sesquiterpenes, which play a role in almost every aspect of insect development and reproduction, including embryogenesis, larval molting, metamorphosis, caste determination in the social insects, vitellogenin synthesis JHII 0 ’ and ovarian development, phase determination in locusts and aphids, larval and adult diapause regulation, color, polymorphism, and various aspects of metabolism associated with these functions (476). / The morphogenetic function of JH (neotenin, status 0 JH III quo hormone) was first detected by observing that larval development was accelerated when the CA were removed. The CA are endocrine glands in the posterior region of the head, which are closely associated with the stomatogastric nervous system. The CA are of ectodermal origin, arising early in embryogenic development (687). The characteristic shape of the CA is ovoid to round. The size of the glands is frequently about the diameter of the aorta but differs with age, sex, polymorphism, and activity / JHB, 0 cycle of the glands. In most insects, the CA receive nerves from both the brain (NCA I) and the SOG (NCA II). The CA FIG. 1. Juvenile hormone homologs. JH O-III, juvenile hormone are surrounded by a continuous noncellular basal lamina. O-III; 4 MeJH I, 4-methyl JH I; JHB3, JH III bisepoxide; MF, methyl Only one type of glandular cell occurs in the CA. The farnesoate. presence of gap junctions suggests that the cells are coupled, and the prominence of smooth endoplasmic reticulum is characteristic of cell types that produce cholesterol 1). The hormone soon proved to be one of a series of or terpenoids in large amounts. The hormones produced naturally occurring juvenile hormones, all with similar are not stored in the gland, although glands retain hor- chemical structures. Juvenile hormone II (C,,JH) was submone in proportion to the rate of biosynthesis. The ques- sequently identified by Meyer et al. (430) and JHIII (C16JH) tion of how the hormones leave the CA is still unresolved. by Judy et al. (314). Juvenile hormone III is the principal Segments of endoplasmic reticulum frequently appear in form of JH found in the Orthopteroidea, Coleoptera, Hemclose proximity to the plasma membrane of CA cells, sug- iptera, and Hymenoptera. The Lepidoptera appear to be unique in possessing a mixture of JHI and JHII. The CA gesting the release of JH may depend on such juxtapositions. Because carrier proteins for JH are known to exist of the tobacco hornwortn moth, M. sexta, however, secrete JHIII in addition to JHII (476). Two additional forms in the hemolymph (129), it seems likely that these may of JH, named JHO and 4-methylJH1, have been found in facilitate movement of JH away from the extracellular the eggs of M. sexta (Fig. 1). The CA of certain male spaces of the gland. The chemical structure of a JH was first determined Lepidopteran species secrete JH acids (JHAI, JHAII, by Roller et al. (591) using MS analysis, NMR, and micro- JHAIII, IsoJHAII), which are later methylated in the male accessory glands (12, 264). Juvenile hormone acids are derivatization techniques. The hormone (JHI, &JH) proved to be an unusual sesquiterpenoid with an epoxide also secreted by CA of prepupal stages of M. sexta and group near one end and a methyl ester on the other (Fig. are methylated in the developing imaginal disks. To this

/LAMAo~3MF

October

1997

INSECT

HORMONES

group of JHs may tentatively be added methyl farnesoate (MF), an acyclic sesquiterpenoid ester that is closely related to JHIII in structure (Fig. 1). Very high levels of MF (and JHIII) have been detected in embryos of the cockroach Nauphoeta cinerea and are produced in vitro by CA dissected from embryos; however, MF seems to be unable to substitute for the hormonal role of JHIII in vivo. The role of MF as a hormone or prohormone is perhaps more feasible in crustaceans than in insects (382). Stereochemical investigations have shown that natural JHIII has the same lOR-configuration as JHI. The lOR,llS absolute configuration of JHII has never been rigorously established. Recently, methyl (2Q-6,7; lO,ll-bisepoxyfarnesoate (JHB3; JHIII bisepoxide) has been identified in the higher Diptera (Cyclorrhapha) as the major in vitro JH product of larval ring gland and of adult CC-CA complexes (580; Fig. 1). A variety of biochemical and physiological studies support the hormonal significance of JHB3 (natural configuration 2E, 6S, 7S, 1OR). The CA of adult female locusts, L. migratoria, synthesize and release JHIII lO,ll-diol together with JHIII (205). Virtually all of our knowledge of the biosynthesis of the JH carbon skeleton comes from in vitro studies. In 1970, Roller and Dahm (590) demonstrated JHI biosynthesis by brain-CC-CA complexes from Lepidopteran, Hyalophora cecropia, male pupae incubated for a long term in Grace’s medium. Subsequently, it was discovered that the S-methyl moiety of methionine is the in vivo source of the methyl ester group of the JHs (429). This led to general use of S-methyl radiolabeled methionine to monitor JH biosynthesis in vitro. Tobe and Pratt (686) demonstrated the efficacy of medium 199 as an incubation vehicle and revealed that incubation times of 10 min are sufficient for the recovery of detectable quantities of radiolabeled biosynthesized JH. This radiochemical assay or the slightly modified rapid partition assay (141) has proven invaluable in studies of pathways, precursor and substrate utilization, and enzymatic activities in JH biosynthesis (687). In general, it can be said that the identities of the JHs produced in vitro mirror those found in viva. There is also an excellent correlation between rates of JH biosynthesis in vitro and JH titer in hemolymph and whole bodies. The synthesis of JHIII, the homolog lacking branched side chains, is now known to follow the pathway described for the initial steps in cholesterol synthesis (228). The potential precursors for the carbon skeleton are two-carbon (C-2) units that have arisen from the metabolism of glucose, leucine, isoleucine, and threonine as well as the immediate precursors of acetate (44). Three C-2 units, in the form of acetyl CoA, undergo enzymatic condensation to yield a C-6 intermediate, 3-hydroxy 3-methylglutaryl-CoA (HMG-CoA). HMG-CoA is reduced to mevalonate by the enzyme HMG-CoA reduc-

987

tase, which uses NADPH as an electron donor. This enzyme is well known to play a central role in mammalian steroid biosynthesis, but there is no clear evidence that HMG-CoA reductase is involved in regulatory levels of JH. Fluvastatin, a synthetic HMG-CoA reductase inhibitor, inhibited JH biosynthesis by CA of the migratory locust, L. migratoria migratorioides, in vitro and in vivo. JH production in vitro can be completely restored by addition of mevalonic acid la&one to the culture medium (102). The next step in JH biosynthesis is the conversion of mevalonate to 3-isopentenyl pyrophosphate (IPP), which isomerizes to 3,3’-dimethylallyl pyrophosphate (DMAPP). The last step in the synthesis of the carbon skeleton is the condensation of two units of IPP and one unit of DMAPP to form the basic farnesyl pyrophosphate unit. Farnesyl pyrophosphate is then converted to farnesol. Farnesol is oxidized to farnesal via a dehydrogenase that requires NAD and then further oxidized to farnesoic acid by another dehydrogenase also requiring NAD. The terminal steps in biosynthesis are methyl ester formation at C-l and epoxidation at the C-lo/C-l 1 position. 10,l I-Epoxidase (monooxygenase) is usually linked to cytochrome P-450 and uses NADPH as cofactor. The sequence of the terminal two steps in JH biosynthesis may be reversed in the Lepidoptera, i.e., farnesoic acid to epoxyfarnesoic acid to JHIII, when compared with the situation in most other orders (farnesoic acid -+ methyl farnesoate + JHIII). No single enzymatic step has been definitely shown to be rate limiting in the biosynthetic pathway, and there is evidence to suggest that more than one step may be involved in overall rate limitation (687). The major points of rate limitation seem to occur before or at the formation of mevalonate (see sect. II&~). Branched precursors to JHO, JHI, and JHII originate from propionyl-CoA. The homoisoprenoid units that give rise to the branched chain are formed from condensation of one propionyl-CoA (from isoleucine or valine) and two acetyl-CoA units. This results in the synthesis of a homomevalonate intermediate. Thus JHO is composed of three homomevalonate units, JHI is composed of two homomevalonate and one mevalonate units, and JHII is composed of one homomevalonate and two mevalonate units (228). Biosynthesis of JHB3 in D. melanogaster seems to occur by epoxidation of farnesoic acid, at the 6,7 and 10,ll sites to 6,7; 10,l l-bisepoxyfarnesoic acid (FABE), followed by terminal methylation of FABE (439). Male accessory glands (MAG) of the mosquito A. aegypti, incubated with [ 14C]acetate, synthesized radiolabeled JHIII, JHB3, and methyl farnesoate, which indicates that accessory glands of the male mosquito have a complete biosynthetic pathway. In A. aegypti, ovaries also seem to synthesize [“HI-, [ 14C]JHIII, and JHIII bisepoxide-like molecules from L-[methyl-3H]methionine and [14C]acetate (34, 36).

988 2. Modes of action

GiiDE,

of juvenile

HOFFMANN,

hormones

Juvenile hormone derived its name from the fact that it blocks metamorphosis of nymphs into imagoes or the development of pupae into adult insects. As a developmental hormone, JH controls switches between alternative pathways of development at several points in the life cycle. In ascribing a role to JH in the control of such manifold developmental switches, many authors have assumed that different concentrations of JH are responsible for specifying the different pathways. Most of the experimental work on the hormonal control of molting and metamorphosis gave results consistent with the hypothesis that metamorphosis in holometabolous insects is caused by a successive lowering of the JH titer. Holometabolous insects continue as larvae at high concentrations of JH. Pupation occurs when JH declines to an intermediate or low level, and adults are formed in the absence of JH. It is now clear, however, that JH may not act in a concentration-dependent manner, but acts during discrete critical periods, the JH-sensitive periods (476). In general, if JH is present during a critical period, no developmental switch occurs, and the current developmental state is maintained. If JH is absent during a critical period, gene expression changes, and new developmental processes begin that launch the insect on a new developmental pathway. The actual titer of JH during the JH-sensitive period is immaterial as long as it is above or below a certain threshold. In hemimetabolous metamorphosis, there is one JHsensitive period during each larval instar. During the JHsensitive periods, JH titers are always above threshold. During the last larval instar, the CA are turned off and JH falls below threshold during the JH-sensitive period (267). At that point, the switch from larval to adult commitment is made. In holometabolous metamorphosis, the absence of JH during a sensitive period while the animal is a final instar larva results in the development of pupal characters, whereas absence of JH during the pupal JH-sensitive period provokes adult development. A brief peak of JH just before pupation coincides with a JH-sensitive period during which the pupal-adult switchover of the imaginal disks is controlled. Thus there are two successive JHsensitive periods during the last larval instar. Normal metamorphosis requires the absence of JH during the first JH-sensitive period, which controls a larval-pupal developmental switch of the general integument, and the presence of JH during the second JH-sensitive period, which stabilizes the pupal-determined state of imaginal disk structures and prevents their commitment to adult development. Associated with the pupal-adult development are changes in the nervous system, deposition of an adulttype cuticle, reconstruction of the fat body, development of flight muscles, degeneration of intersegmental muscles, and, most importantly, acquisition of the ability to repro-

AND

SPRING

Volume

77

duce. All of these development changes are blocked by JH (368). The effect of JH on integumental cells is autonomous, which means independent of the developmental fate of neighboring cells. If JH is topically applied to selected regions of the cuticle, only epidermal cells underlying this region of the cuticle fail to metamorphose. The Galleria JH bioassay, which involves inflicting a small injury on a pupal tergite and covering it with JH-containing wax, is based on this principle of cell-autonomous action of JH (for details, see Ref. 94). The question revolves around the probable molecular mechanism of JH control of metamorphosis. It is tempting to believe that JH acts in a similar manner to that of the steroid hormones by inducing specific transcriptions through a distinct interaction with the genome. A JH analog has recently been shown to repress transcription of genes for certain proteins in a larval Lepidopteran (313). A central role in the mechanism of modulation of cellular activity by JH must be played by the JH receptor. Although JH binding proteins from various insects have been intensively studied, it is difficult to know which of these represent true receptors (760). A nuclear JH-binding protein from M. sexta larval epidermis, which has been cloned and sequenced as a potential receptor, does not belong to the known nuclear receptor superfamily, thus raising speculation as to its function (586). Another nuclear JH binding protein (1M, 31,000) has recently been isolated from male reproductive tissue of Melanoplus sanguinipes (304). The use of photoaffinity analogs of JH will further help to characterize JH binding proteins and JH receptors (526). Very recently, Engelmann (131) isolated and purified a JH receptor from another typical target tissue for JH, the adult fat body of the cockroach L. maderae. Developmentally, the JH receptor first appears in fat bodies of last instar nymphs at a time when such animals can be induced to make vitellogenins (see below). The molecular mass of this high-affinity JH-binding protein (dissociation constant = 1-2 nM) is -64 kDa. Effects of JH in insects depend not only on the presence and concentration of the hormone in the hemolymph and tissues, but also on the sensitivity of the tissues to the hormone. Sensitivity may change when new JH receptors are being synthesized (94). The type of response to JH in the diverse cell types is irreversibly fixed during embryonic determination. The gonadotropic effect of JH was observed for the first time by Wigglesworth (747) in R. prolixus. His experiments showed that the CA, which cease to function before metamorphosis, resume JH production in the adult, and the hormone then stimulates egg maturation in females and development of accessory glands in males. In contrast to the premetamorphic actions of JH, which prevent the expression of genes required for metamorphosis, the postmetamorphic activities involve the stimulation of gene expression or cell functions. In females, JH affects egg

October

1997

INSECT

HORMONES

production at different levels. Juvenile hormone regulates vitellogenin (Vg) production and/or uptake into the growing oocytes. Juvenile hormone regulates oogenesis and egg production in apterygotes, and because they do not show overt morphogenetic changes during ontogeny, it was suggested that the gonadotropic function is its primary role (368). Vitellogenin is produced by the fat bodies in most insect species studied, but in D. melanogaster and other Dipterans, the ovarian follicles make it as well. Juvenile hormone directs the tissue-specific transcription of the Vg genes (130). Again, JH is thought to bind to specific receptors in the cytosol and/or nuclei of the target tissue. Subsequently, this hormone-receptor complex is recognized by specific DNA sequences near the Vg gene(s), and transcription is initiated. The effects of JH on vitellogenin synthesis in the fat body are not universal. Juvenile hortnone has no effect on Vg synthesis in some short-lined Lepidopterans. Similarly, in the honey bee Apis mellifera, JH has no effect on Vg synthesis. In D. melanogaster and other Dipterans that produce eggs more or less continuously, the role of JH in Vg synthesis is also not definitive. Ecdysteroids seem to intervene in regulation of Vg synthesis in these species (237). Finally, in Gryllus bimaculatus, a field cricket, JH does not have an all-embracing control of Vg production, although it does exert a marked quantitative effect on Vg synthesis. Allatectomized females still laid a low number of eggs; however, administration of the JH analogs hydroprene, methoprene, and fenoxycarb to allatectomized females restored Vg synthesis and egg production (270). Regulation of Vg uptake from the hemolymph through the ovarian follicular epithelium for deposition of vitellins (Vn) in the oocytes has been found in all insect species studied so far (96). In the American cockroach, P. americana, Vg uptake was blocked by allatectomy, and application of JH reestablished the Vg uptake. In R. prolixus, there is convincing evidence that the primary action of JH on the follicle cells is exerted at the level of the plasma membrane and is mediated by a system that involves Na+-Kf-ATPase and protein kinase C (628). After the application of JH, follicle cells undergo rapid reversible shrinkage resulting in the formation of large intercellular spaces, a condition termed patency (265). Current evidence clearly points to Vg uptake being a receptormediated process. A third organ that depends on JH in many adult insects is the male accessory reproductive gland (MARG) (42,303,519). Dependence of MARG function (production of proteins of the spermatophore as well as certain secretions that are transferred in copulation and affect the reproductive behavior and physiology of females) on JH has been found in many species. Studies on MARG in D. melanogaster suggest that JH may act at the membrane level (764), suggesting the use of a membrane receptor as

989

in the R. prolixus ovarian follicle (759). In MARG of L. migratoria, the presence of intracellular JH binding protein suggests a direct action of JH within the gland that may be modulated by an enzyme of JH catabolism, the JH esterase (42). The following effects of JH are probably based on the action modes described above, but existence of additional mechanisms cannot be excluded. In social insects, such as bees, wasps, and termites, different morphotypes occur simultaneously in a colony. Caste determination in social insects represents stimulation of specific developmental pathways by JH, but interactions with neurohormones and ecdysteroids are crucial. The notion that JH induces queen development in honeybees (A. mellifera) has become textbook knowledge. However, recent data provoked doubt concerning the role of JH as the major “queen-making” agent. In the honeybee, a reaction chain is suggested by which an exogenous trophogenic stimulus first stimulates JH synthesis by the CA during the early fifth instar. Subsequently, the high JH titer may turn on an enhanced ecdysteroid production by the prothoracic glands. An increased ecdysteroid titer early in the prepupal phase then probably acts directly at the DNA level, regulating transcription of genes which shift development toward the queen pathway (541). Polymorphic forms occur not only in social insects but also in other insect groups. In locusts, aphids, and Lepidopterans, polymorphism is associated with seasonal and/or environmental factors. Locusts, armyworms, and cutworms, for example, may be solitary or gregarious. Solitary locusts are characterized by green coloration, large ovaries, and short wings, whereas gregarious locusts are brown or yellow and have small ovaries and large wings. Solitary locust nymphs have higher JH titers than gregarious forms, and it was repeatedly claimed that JH constitutes the primary intrinsic causal factor in phase changes. Recent experiments on L. migratoria (507) revealed that JH deficiency induces disappearance of the green color; however, the green color is not an obligate characteristic of the solitary phase. The control of the green color by JH does not mean that JH also controls all or even most other features of locust phase polymorphism. Aphids form one of the few groups of insects that commonly alternate between parthenogenetic and sexual reproduction. It appears that JH has a role both in determining the parthenogenetic aphid morph and in the regulation of ovarian development in that morph (242). During their life cycle, many insects exhibit a period of developmental arrest, called diapause, to survive unfavorable environmental conditions. Diapause occurs at different development stages in different insect species. Larval diapause is often produced by a short photoperiod that induces a high JH titer in the body, and this prevents metamorphosis. Imaginal diapause also takes place under short photoperiods, but in this ease, CA activity is inhib-

990

GADE,

HOFFMANN,

ited, leading to a depression of the JH titer in the body. This makes gonadal development impossible, so imaginal diapause is associated above all with the inhibition of reproduction (94). Studies by Khan (340) gave clear evidence for a nervous CA inhibition and a negative feedback regulation of JH via the brain in adult diapause control. The nervous system seems to be affected by JH in two ways. Indirect effects follow from the morphogenetic action of JH on the development of the nervous system. Direct responses are manifested as changes in the production of neurohormones and alterations of behavior. The influence of JHIII on the phototactical response of female A. domesticus to the male’s calling song could provide a model system where the regulation of a rather complex behavioral response, essential for reproductive success, is accomplished by hormonal regulation of the response properties of identified auditory neurons (396,671). Cayre et al. (61) recently demonstrated that JH affects adult insect brain structure through neurogenesis. 3. Control of JH synthesis neuropep tides

by allatoregulating

Insect development and reproduction exhibit two main characteristics: a great variety of successive and fundamentally distinct processes, and a strict dependence on environmental factors (269). Regulation of insect development and reproduction involves numerous sensory receptors, nerve transmission, and integration in the brain, which regulates, in part through its neurohormones, synthesis of the two groups of insect developmental hormones, the ecdysteroids and the JHs. The ability of the CA to synthesize JH may be controlled by stimulatory and inhibitory signals that reach the glands via the hemolymph or via nervous connections (228). In many insect species, there is strong evidence that an important control mechanism involves brain neurosecretory cells that project to each CA and exert a paracrine influence on gland cells (288). In L. migratoria migratorioides, two types of brain neurons innervate the CA. Thirteen cells in each pars lateralis of the brain innervate the ipsilateral CA, whereas four cells innervate both glands. Simultaneous electrical stimulation of all neurons innervating one CA always inhibited JH production, while their transection led to a rapid progressive increase in JH biosynthesis in CA from mature females. Thus there is a strong neurally mediated inhibition of the CA at certain phases of the vitellogenic cycle. Candidate neuropeptides that stimulate (allatotropin) and inhibit (allatostatin) JH production, respectively, have recently been isolated from the brains of several insect species. The first studies on allatotropic factors were carried out with larvae of the tobacco hornworm, M. sexta. Brain extracts from larvae seemed to contain a factor that stimulated JH svnthesis (229). -0 \ / In 1989.I Kataoka et al. (323) \- -/

AND

SPRING

Volume

77

isolated and identified a tridecapeptide (Table 5) from heads of pharate adult tobacco hornworms that activated JH synthesis in adult animals but had no effect on CA activity of larvae or pupae. Furthermore, CA from the beetle 7’. molitor, the locust S. gregaria, and the cockroach P. americana were not activated by the synthetic allatotropin, whereas the CA of the noctoid moth Helie this virescens were stimulated, suggesting order specificity. Studies on NHz-terminal truncated fragments of the synthetic peptide suggested that the amino acids 6-13 are the biologically active core. A stimulatory mode of control of JH synthesis has also been observed in larvae of the wax moth Galleria mellonella and in adult locusts (L. migratoria) and crickets (G. bimaculatus). Brains of last instar larvae of G. mellonella were shown to contain a 20kDa polypeptide that stimulates JI-I synthesis in vitro (32). In L. migratoria, there is evidence for one or more allatotropins (0.7-2 kDa) in female and male brain-CC complexes (389). Corpora allata from young male G. bimaculatus responded in a stimulatory way to SOG extracts from vitellogenic females (397). So far, however, none of these allatotropic factors has been identified. The search for neuropeptides with allatostatic activity resulted in the isolation of a family of allatostatins (Tyr/Phe-Xaa-Phe-Gly-Leu/Ile-NHZ) from brain extracts of various cockroach species (Diplop tera punctata, P. americana, and Blattella germanica) and of a cricket (G. bimaculatus; Table 5). With the use of HPLC purification and sequence analysis, the chemical nature of seven allatotastins from D. punctata (Dip-AST; Refs. 522, 755, 756), two allatostatins from P. americana (Pea-AST; Ref. 735), four allatostatic neuropeptides from B. gerrnanica (BlgAST; Ref. 21), and two allatostatins from G. bimaculatus (398) have been identified. In 23. gerrnanica, two of these peptides are identical to Dip-ASTl and Dip-AST5, respectively. In D. punctata and P. americana, molecular cloning has led to the isolation of cDNA that encodes a precursor polypeptide containing 13 and 14 potential allatostatic sequences, respectively, including those formerly identified through conventional purification techniques (109, 664). The allatostatin precursor from D. punctata is 370 amino acids long and shows -70% identity with the one from P. americana, which is 379 amino acids long. The coding regions of the two allatostatin genes are remarkably similar in structure and organization. Southern blot analyses indicated the presence of a single copy of the gene per haploid genome in both cockroaches. Recent data suggest that the principle of precursor organization and peptide conservation as established for these cockroaches extends to numerous other cockroach species. Eight neuropeptides with COOH-terminal amino acid sequence similarity to cockroach allatostatins have been identified in the blowfly C. vomitoria by isolation and sequencing methods. Four of these peptides end COOHterminally in Tyr-Xaa-Phe-Gly-Leu-NH, and were named

October

1997

INSECT

TABLE 5. Primary

structures

Peptide

and common

Species

Cav-AST

l*

Calliphora

Cav-AST

2*

C. vomi toria

Cav-AST

3*

C. vomi toria

Cav-AST

4*

C. vomi toria

Cav-AST

5*

C. vomi toria

Cav-AST Cav-AST Cav-AST Dip-AST

6* 7* 8* 1

C. vomi toria C. vomi toria C. vomi toria Diploptera punctata,

vomi toria

Dip-AST

2*

Dip-AST

3

Dip-AST Dip-AST

4* 5*

Dip-AST

6

Dip-AST

7*

Periplaneta americana Blattella gerrnanica D. punt ta ta, P. americana D. punctata, P. americana D. punctata D. punctata B. gerrnanica D. punctata, P. americana D. punctata

Dip-AST

8*

D. punctata

Dip-AST

9*

D. punctata

Dip-AST Dip-AST Dip-AST Dip-AST

10 1 I* 12 13

D. D. D. D.

synonyms Other

Names

991

HORMONES

for the allatostatin Used

in Literature

Leu-callatostatin 1, callatostatin CAST 1 Leu-callatostatin 2, callatostatin CAST 2 Leu-callatostatin 3, callatostatin CAST 3 Leu-callatostatin 4, callatostatin CAST 4 Met-cahatostatin, callatostatin CAST 5 [ Hyp3] Met-callatostatin [Hyp2]Met-callatostatin des Gly-Pro Met-callatostatin Dipstatin 1, DAST 1, Pea-AST peastatin 1 Big-AST 1,” BLAST 1

family

of peptides Sequence

DPLNEERRANRYGFGL-NH2

119

2,

LNEERRANRYGFGL-NH2

119

3,

ANRYGFGL-NH2

119

4,

NRPYSFGL-NH2

119

5,

GPPYDFGM-NH2

119

1,

GPXYDFGM-NH2 GXPYDFGM-NH2 PYDFGM-NH2 LYDFGL-NH2

117 118 118 21, 109, 664

AYSYVSEYKRLPVYNFGLNH2

109, 522, 664

SKMYGFGL-NH2

109, 664

DGRMYSFGL-NH2 DRLYSFGL-NH2

109, 755 21, 109, 756

ARPYSFGLNH2 APSGAQRLYGFGL-NH2

109 664 109, 523, 756

GGSLYSFGL-NH2

109, 756

GDGRLYAFGL-NH2

109, 756

PVNSGRSSGSRFNFGL-NH2 YPQEHRFSFGL-NH2 PFNFGL-NH2 IPMYDFGI-NH2

109 109, 755 109 109 664 664 664 664, 735 664 664, 735 664 664 664 664 21 21 705; H. Kaatz, personal communication 705; H. Kaatz, personal communication 705; H. Kaatz, personal communication 398 398 399 399 399 399 366 323

Pea-AST 4 Pea-AST 5 Pea-AST 7* Pea-AST 8 Pea-AST 9* Pea-AST 10 Pea-AST 11 Pea-AST 12 Pea-AST 13 Blg-AST 3* Blg-AST 4* Apm-AST l*

punctata punctata punctata punctata, P. americana P. americana P. americana P. americana P. americana P. americana P. americana P. americana P. americana P. americana B. gerrnanica B. germanica Apis mellifera

Apm-AST

2*

A. mellifera

Apm-AST

3*

A. mellifera

GRQPYSFGL-NH2

Grb-AST Grb-AST Grb-AST Grb-AST Grb-AST Grb-AST Mas-AST* Mas-AT*

Al* A2* B l* B2* B3* B4*

Gryllus bimaculatus G. bimaculatus G. bimaculatus G. bimaculatus G. bimaculatus G. bimaculatus Manduca sexta M. sexta

AQHQYSFGL-NH2 AGGRQYGFGL-NH2 GWQDLNGGW-NH2 GWRDLNGGW-NH2 AWRDLSGGW-NH2 AWERFHGSW-NH2 pEVRFRQCYFNPISCF-OH2 GFLNVQMMTARGF-NH2

* Isolated

from

tissues

and purified

I

techniques.

SGNDGRLYSFGL-NH2 DRMYSFGL-NH2 SPSGMQRLYGFGL-NH2 GGSMYSFGL-NH2 ADGRLYAFGLNH2 PVSSARQTGSRFNFGL-NH2 SPQGHRFSFGL-NH2 SLHYAFGL-NH2 PYNFGL-NH2 AGSDGRLYSFGL-NH2 APSSAQRLYGFGL-NH2 YSGARYPVYSFGL-NH2

AYTYVSEYKRLPVYNFGLNH2

Mas-AS

bY conventional

No.

1,

Dipstatin 2, DAST 2, Dip-AST V (5), ASB2, Pea-AST 2, peastatin 2 Dipstatin 3, DAST 3, Pea-AST 3, peastatin 3 Dipstatin 4, DAST 4, Dip-AST VII (7) Dipstatin 5, DAST 5, Dip-AST IV (4), Dip-AS 4, Blg-AST 2,* BLAST 2 Dipstatin 6, DAST 6, Pea-AST 6, peastatin 6 Dipstatin 7, DAST 7, Dip-AST I (l), Dip-As 1, Dip-A 1, ASAL Dipstatin 8, DAST 8, Dip-AST III (3), Dip-AS 3 Dipstatin 9, DAST 9, Dip-AST II (a), Dip-AS 2 Dipstatin 10, DAST 10 Dipstatin 11, DAST 11, Dip-AST VI (6) Dipstatin 12, DAST 12 i Dipstatin 13, DAST 13, Pea-AST 14, peastatin 14 Peastatin 4 Peastatin 5 Peastatin 7, PEAST 1, Pea-AST I (1) Peastatin 8 Peastatin 9, PEAST 2, Pea-AST II (2) Peastatin 10 Peastatin 11 Peastatin 12 Peastatin 13 BLAST 3 BLAST 4

Apm-allatostatin

Reference

X, hydroxyproline;

AST, allatostatin;

AT

ahatotropin.

992

G;I1DE,

HOFFMANN,

AND

SPRING

Volume

77

Leu-callatostatins. The fifth peptide has a COOH-terminal allatostatin-like peptides in several insect species, it is Met instead of Leu (Met-callatostatin; Ref. 119). Bioassays now clear that these allatostatins represent a unique famhave revealed that these callatostatins from the blowfly ily of neuropeptides that probably serve several functions have potent allatostatic activity in the cockroaches, D. in insects (see below). Although the COOH terminus is punctata, P. americana, and B. germanica. However, highly conserved, there is considerable variation in the these peptides do not inhibit synthesis of JHIII bisepoxide NH2-terminal region of the molecules (the “address” sein the fly itself. Another three posttranslationally modified quence; Ref. 664). This variability provides differences in Met-callatostatins have been identified from head extracts relative potency (defined as EDso values, i.e., 50% of maxiof C. vomitoria. [Hyp”]Met-callatostatin and [Hyp2]Metmal inhibition) of the peptides in bioassays, including callatostatin are hydroxylated analogs of Met-callatotheir effect on JH biosynthesis in vitro. The synthetic pepstatin; the third one represents a truncated hexapeptide, tides are effective (EDso) at concentrations of 10-l’ to des Gly-Pro Met-callatostatin. The prohormone gene enlop7 M and act quickly (within hours), and the inhibiting coding the Leu-callatostatin peptides has been isolated effects are fully reversible (665). In most cases, however, recently (126). It is 180 amino acids long and contains 2 the suppression of JH synthesis is incomplete (maximum blocks of tandemly arranged Leu-callatostatin peptides inhibition of 70-90%) even at high doses. The allatostatins with 5 copies of the COOH-terminal sequence -YXFGL. appear to exert their effect early in the JH biosynthetic An allatostatin that is structurally unrelated to the pathway, at least before the conversion of farnesoic acid cockroach allatostatins has been purified and characterto methyl farnesoate (755). Sutherland and Feyereisen ized from heads of pharate adults of M. sexta (366). The (673) speculated that the transport of C-2 units out of l&residue peptide (Table 5) strongly inhibits in vitro JH mitochondria was a likely target for regulation. However, biosynthesis by CA from fifth stadium larvae and from Wang et al. (730) recently demonstrated that two allatoadult females of the moth. It also inhibits the CA of adult statins, Dip-AST5 and Met-callatostatin, are able to inhibit females of another Lepidopteran species, H. virescens, JH synthesis as a result of inhibition of the activity of but has no effect on the activity of CA from adult females terminal enzymes (methyltransferase and/or epoxidase). of the beetle T. molitor, the grasshopper Melanoplus sanThe results also suggest the existence of an alternative guinipes, or the cockroach P. americana. Bendena et al. pathway for JHIII biosynthesis in D. punctata. The pre(22) have recently isolated the gene specifying an allatodominant pathway is farnesoic acid to MF, then to JHIII, statin precursor from the moth Pseudaletia unipunctata. whereas the other, representing -5- 10% of total JH synContained within this precursor is an allatostatic peptide thesis, is farnesoic acid to JHIII acid, then to JHIII. that is identical in sequence to that from M. sexta. A The physiological state of the CA may condition their similar precursor organization and peptide sequence has response to inhibition by allatostatins. For example, there been deduced from a gene sequence isolated from the are major changes in the sensitivity of CA from D. puncDipteran D. melanogaster. tata toward Dip-AST7 during the female reproductive cyRecent isolations of allatostatic neuropeptides from cle. Glands from recently molted adult females showed the desert locust S. gregaria (710), the mosquito A. ae- only 30-40% inhibition by Dip-AST7 at lo-” M, declining to gypti (705), and the honeybee A. mellifera (316; Table < 10% on day 5, at the peak of spontaneous JH synthesis in 5) generated peptides with the common COOH terminus vitro. A dramatic increase to -90% Dip-AST7 sensitivity identical to the cockroach allatostatins. AI1 these peptides at low8 M was observed by day 6, i.e., in postvitellogenic inhibited JH synthesis in D. punctata CA but, so far, have females just before choriogenesis (521). Corpus allatum not been shown to be effective in the source insects. sensitivity to this peptide then wanes during postvitelloThe search for neuropeptides with allatostatic activgenesis and oogenesis (661). Significant differences in the ity in the cricket, G. bimaculatus, led to the isolation of pattern of CA sensitivity are observed with Dip-AST2. The another four nonapeptides with sequence similarity at the CA are more sensitive to Dip-ASTZ during vitellogenesis COOH terminus (Gly-Xaa-Trp-NH,) and a common amino and remain more sensitive in gravid females. The most acid at position 2 (Trp; Table 5). These peptides have been notable change in sensitivity to Dip-ASTZ, however, is designated G. bimaculatus allatostatic neuropeptides seen during the day 5 to 6 transition. In contrast, CA of G. Bl-4 (Grb-AST Bl-4) in accordance with their biological bimaculatus are highly sensitive to both types of cricket activity (399). The primary structures of the four nonapepallatostatic neuropeptides (Grb-AST Al -2; Grb-AST Bl4) when they are at the height of activity during middletides show homology to the locustamyoinhibiting peptide late vitellogenesis. Changes in sensitivity to allatostatins (Lom-MIP), which suppresses the spontaneous contractions of the hindgut and oviduct of L. migratoria and of seem to be peptide specific and species specific, and they are not restricted to the adult stage. The sensitivity of the the hindgut of a cockroach, L. maderae. The allatostatic effect of Grb-AST Bl-4 was also evident on CA from a CA to allatostatins can be experimentally manipulated. related cricket species, the house cricket, A. domesticus. The increase in allatostatin sensitivity of day 6 CA of D. With the discovery of at least 36 different Diploptera punctata results from the interactions of the CA with their

October

1997

INSECT

HORMONES

endocrine milieu. Thus a humoral factor is postulated to be responsible for this transition (704). Although changing titers of allatostatins have been detected in the hemolymph of D. punctata (754, 772), few studies have been done on effects of allatostatins after injection into the hemolymph. Injections of Dip-AST5 and Dip-AST7 into day Q virgin P. americana were effective in bringing about a lowering of total body JHIII levels 12 h postinjection. Similar injections of Dip-AST7 into midcycle mated females produced no apparent effect, but injection of Dip-AST5 again resulted in a reduction of endogenous JHIII levels (736). Concentrations of injected allatostatins were rather high (100 pg/insect) because they are likely to be degraded rapidly (664). Injections of DipAST7, Dip-ASTZ, and Met-callatostatin into mated females of D. punctata on days l-3 after ecdysis at 12-h intervals lowered the length of basal oocytes and/or JH synthesis by CA in vitro when compared with water-injected controls (754). The presence of allatostatins in the hemolymph suggests a humoral pathway for their action in addition to a nervous pathway from the brain. Structure-activity studies showed that allatostatins lacking the COOH-terminal amide have no detectable inhibition of JH biosynthesis. Studies using truncated allatostatins and modified allatostatins (analogs) demonstrated that the conserved COOH-terminal pentapeptide region is critical for potency (i.e., signal transmission to target tissues). This portion of the molecule is likely to interact directly and to be recognized by the potential receptor(s) for allatostatins. The interaction may be modified by the address sequences of the NH2-terminal region. The diversity in the structure of the address regions may represent differences in the degree of binding to allatostatin receptor subtypes, all of which recognize the message sequence (664). With the use of analogs of a small member of the Tyr-Xaa-Phe-Gly-Leu-NH, allatostatin family, Dip-AST5, it could be demonstrated that Tyr”, Phe’, and Leu’ are the most important amino acid residues for the inhibition of JH synthesis. These COOH-terminal structural elements may be oriented in three-dimensional space by a confirmation like a type II p-turn centered around the Phe’-Gly’ position (247). The occurrence of multiple allatostatic peptides has prompted the search for their receptors. The presence of multiple allatostatin species might a priori suggest the existence of individual receptors for each species of molecule. It remains to be determined, however, whether each allatostatin species is associated with a different receptor/ receptor subtype (542, 664). To date, two allatostatinbinding proteins (putative receptors) for Dip-AST7 have been tentatively identified in brain and CA of D. punctata (92, 93). Yu et al. (771) have developed both an in vitro binding assay and a photoaffinity labeling assay to characterize partially putative receptors for allatostatins. The data suggest that the brain membrane has two binding

993

sites for Dip-AST7, but only one for Dip-AST5. Interestingly, the dose-response curve for JH inhibition by DipAST7 shows a biphasic relationship, whereas for DipAST5, the relationship is linear. The array of allatostatins and the likely occurrence of multiple receptor subtypes raises the possibility of multiple signal transduction mechanisms for these peptides in terms of inhibition of JH synthesis (542, 664). Most neuropeptide receptors interact with G proteins to exert their effects via second messengers. In D. punctata, correlations exist between allatostatin action of brain extracts and elevated CAMP levels within the CA. Pharma.eologitally elevated CAMP levels inhibited JH release. Synthetic Dip-allatostatins, however, did not elicit changes in the levels of cyclic nucleotides in isolated CA. Extracellular Ca”+ directly influences rates of JH synthesis. In it/s absence, JH production ceases, whereas elevated concentrations in the medium stimulate JH synthesis in vitro in a dose-dependent manner. Increased K+ concentrations inhibit synthesis almost completely. It’ is assumed that the intracellular Ca”+ concentrations of the gland are regulated in part by voltage-dependent Ca”’ channels in concert with appropriate outward conductances, such as Ca2+-dependent Kf channels. Treatment of CA from mated females of D. punctata with the Ca2+-mobilizing drug thapsigargin resulted in a significant stimulation of JH biosynthesis (543). Further experiments demonstrated that diacylglycerol is an intermediate in the allatostatininduced inhibition of JH production and that IPs is involved in the modulation of JH synthesis at specific developmental times. Such interactions between the two second messengers, diacylglycerol and IPs, could result in either reduced sensitivity to allatostatins at stages in which high CA activity is required (vitellogenesis) or in enhanced sensitivity to allatostatins at times of reduced gland activity. The biogenic amine octopamine enhanced the release of JH from CA of L. migratoria in vivo and elevated the CAMP content of the glands. In D. punctata and G. bimaculatus, octopamine may act as a natural neuromodulator of JH production by regulating ion channels in CA cells, as well as release of allatostatins from the terminal nerves within the CA (683), thereby inhibiting JH synthesis (757). Treatment of the CA of honey bee worker larvae with octopamine and serotonin caused a dose-dependent stimulation of JH production, whereas dopamine and norepinephrine were ineffective (540). The activity of the biogenic amines seems to be mediated by CAMP. Other factors that influence allatostatin release were demonstrated by measuring hemolymph titer of allatostatins after experimental treatment of male D. punctata (662). Males treated with the JH analog 7S-hydroprene showed a strong reduction in JH synthesis, but a less pronounced reduction occurred in animals in which nerves between the brain and the CA were severed. Hemo-

994

GiiDE,

HOFFMANN,

lymph of such treated males showed an increased concentration of allatostatins. The results suggest that JH synthesis by the CA is normally inhibited as a result of feedback from JH titer by delivery of allatostatin directly to the CA, but when nerves to the CA are severed, allatostatins are released into the hemolymph and inhibit JH synthesis, albeit less effectively. When ovarioles were implanted into hydroprene-treated males with denervated CA, hemolymph allatostatins were markedly reduced compared with controls that were similarly treated but not implanted with an ovary. These results show that growing ovaries inhibit the release of allatostatins that may result in increased JH synthesis in vitellogenic females. In G. bimaculatus, implantation of vitellogenic ovaries into adult males of various ages also stimulated JH biosynthesis within 24 h (268). Mature ovaries of G. bimaculatus and other insect species, however, may also contain a factor that directly inhibits JH synthesis by the CA. Such a direct allatostatic effect of an ovary-derived factor was first reported for vitellogenic locust females, L. migratoria (204). Ferenz and Aden (138) have elucidated the peptide nature of this ovarian allatostatic factor, but it has not been identified so far. -4. Mu1 tifunc tional role of alla tos ta tins Brain neurosecretory cells producing allatostatins were identified immunohistochemically with antibodies produced against synthetic peptides in M. sexta; in the cockroaches D. punctata, P. americana (antibodies against D. punctata allatostatin) and B. gerrnanica (antiserum against Leu-allatostatin); and in the crickets G. bimaculatus and A. domesticus (antibody against D. punctata-allatostatin; 664). Polyclonal activity against Dip-AST2 was used to immunolocalize allatostatin-like peptides in the CNS of the locusts S. gregaria and L. migratoria and of the fleshfly Neobellaria bullata (711). From the description of D. punctata allatostatin-immunoreactive cells in the brain that project to the CC-CA, it is clear that the lateral group of neurosecretory cells carries allatostatins to the CA via the NCC2. Four large medial cells in the ventral pars intercerebralis are intensely allatostatin immunoreactive. These cells, however, do not project to the NCCl (connection to the CC) but rather are interneurons with collaterals that branch in the anterior protocerebral neuropil. Allatostatin immunoreactivity is also found in many other somata and axon pathways in the brain, especially in the tritocerebrum. A similar distribution appears to exist in other cockroaches, in locusts, and in crickets. In accordance with the fact that neither callatostatins nor cockroach allatostatin showed allatostatic activity in mature female blowflies, C. vomi toria, immunoreactivity to an antiserum directed against the Phe-Gly-Leu-NH, terminus could not be observed in the CA of C. vomitoria, despite its presence in neurons

AND

SPRING

Volume

77

of the brain. In the fleshfly IV. bullata, there is abundant immunoreactivity in cell bodies of the brain, but also none in the CC-CA complexes. Enzyme-linked immunoassay and immunocytochemistry revealed that allatostatins are located not only in the insect brain, but also in several peripheral tissues, including the gut (573). For example, allatostatin-like immunoreactivity was found in nerve fibers of the stomatogastric nervous system as well as in intrinsic endocrine cells of the midgut of D. punctata. In situ hybridization with a 33%bp fragment of an allatostatin gene that encoded six cockroach allatostatins confirmed that an allatostatin gene is expressed in the intrinsic endocrine cells of the midgut. Northern analysis showed the presence of a specific S. gregaria allatostatin precursor transcript in the midgut of desert locusts (707). Allatostatins may thus represent an example of insect “brain-gut peptides,” and their function may not be restricted to the regulation of JH biosynthesis. Nerves to the antenna1 pulsatile organ muscle and the proctodeal nerves to the muscles of hindgut showed strong immunoreactivity, so attention was directed to these organs with respect to whether allatostatins would affect the activity of the muscles (664). The antenna1 pulsatile organ is an accessory heart that aids circulation of hemolymph into the antennae. Its rhythm can be increased by proctolin and inhibited by octopamine, but the antenna1 heart nerves also exert neural control over the muscle. Application of allatostatin alone or with proctolin did not alter the contraction of the cockroach antenna1 heart muscle. The action of allatostatins on the antenna1 heart muscle is probably to modulate the action of another peptide, most likely pro&Olin. Very recently, Ude and Agricola (701) demonstrated strong allatostatin immunoreactivity in the lateral and segmental heart nerves of P. americana. In the lateral heart nerve, allatostatin immunoreactivity is colocalized with FMRFamide immunoreactivity. In contrast to the inability of allatostatins to alter spontaneous contractions of antenna1 heart muscle, D. punctata hindgut muscle reacted to Dip-AST5 and DipAST7 with a dose-dependent decrease in frequency and amplitude of spontaneous contraction (378). It was also shown that when applied simultaneously with proctolin, the two allatostatins antagonize the stimulatory effect of proctolin on the hindgut muscle. Recently, it could be shown that all 13 Dip-allatostatins inhibit both myogenic and proctolin-induced contractions of the hindgut of D. punctata (377). Dip-allatostatins, however, had no effect on either myogenic or proctolin-induced contractions of the oviduct of I,. migratoria. To investigate whether a regulatory control over the contractility of the gut might be a function of the allatostatin group of peptides common to cockroaches as well as blowflies (callatostatins), Duve et al. (125) studied the effects of blowfly callatostatins on gut mvoactivitv in the cockroach L. maderae. The results

October

1997

INSECT

show that the spontaneous contractile activities of the foregut, but not the hindgut, are inhibited by allatostatins. The existence of the B-allatostatins in G. bimaculatus (Grb-AST Bl-4) with their sequence homology to the L. migratoria myoinhibiting peptide (Lom-MIP) may support the multifunctional role of allatostatic neuropeptides in insects. Duve et al. (116) speculated that one of the reasons for the large number of members of allatostatins (callatostatins) appears to be in the provision of an integrated form of gut mobility control, with different peptides controlling specific regions of the gut. Very recently, Stay et al. (663) have detected allatostatin-containing granules in hemocytes of D. puntata. Immunocytological evidence is accumulating to suggest that both moth and cockroach allatostatin-like materials also occur in other classes of arthropods as well as in other phyla (Hydrozoa, Cestoda, Trematoda, Nematoda, Oligochaeta, Gastropoda, Cephalopoda; Ref. 637). The identity of the compounds responsible for this immunoreactivity is not yet known. 5. Juvenile

hormone

995

HORMONES

degradation

Juvenile hormone titer is regulated by a balance between the rate of its synthesis and degradation. Degradation reduces the levels of JH in specific tissues and may ensure that JH titers are maintained at a low level even when JH synthesis is not totally halted. For example, the low activity of the CA in the latter half of the final larval stadium of G. bimaculatus is associated with very high levels of JH esterase activity in the hemolymph (267). In addition to ester hydrolysis, several other potential routes of JH metabolism exist: epoxide hydration, oxidation, and conjugation. The primary route of JH degradation in the hemolymph is ester hydrolysis, whereas the other routes are possible in the tissues. Recent studies on JH degradation in Lepidopteran larvae have led to the identification of a cDNA fragment encoding JH esterase in H. virescens (240). In Trichoplusia ni, the JH esterase gene encodes a transcript of 2.8 kb, the abundance of which is strongly induced by JH (602). The catabolic activity of the hemolymph, in terms of JH esterase activity, is often greater than the in vitro rate of JH synthesis, and at least in these species, this argues against a role for the JH esterase in JH titer regulation. In general, however, high hemolymph JH esterase activity is associated with declining or low JH levels, and the appearance of hemolymph JH esterase activity at specific times during development and reproduction in a number of insect species has provided at least correlative evidence for a possible functional role for JH esterase in the regulation of JH titer. The timing of the appearance of JH esterase activity appears to be regulated by factors from the head and by JH itself (589). Several additional factors can affect JH

esterase activity including photoperiod, nutrition, parasitism, and stress. In the hemolymph, JH binds to JH-binding proteins (JH BP). Juvenile hormone binding proteins appear to be crucial for maintaining and distributing JH in hemolymph. Low-molecular-mass JH BP are predominantly found in Lepidoptera, whereas high-molecular-mass JH BP are present in other insect orders. In L. migratoria, a ratio of -2,OOO:l has been calculated between bound and free JH. Thus almost all JH III in the hemolymph is associated with JH BP, and this binding is probably important for JH III protection against enzymatic degradation by hemolymph esterases. From the hemolymph of D. punctata, a high-molecular-mass, high-affinity JH BP was identified as a lipophorin (342). Its titer exhibited some changes during the first reproductive cycle of adult females that were not coincident with those in total hemolymph protein concentration. Exogenous JH BP did not affect the biosynthesis and release of JH from the CA in vitro, but plasma JH esterase activity was inhibited by exogenous JH BP in a dose-dependent manner. The lipophorin inhibits JH esterase activity specifically by decreasing the amount of JH available for hydrolysis but does not inhibit the enzyme directly. In adult L. maderae, a form of JH BP protects the enzyme from degradation by direct inhibition of the enzyme as well as by competition for substrate. In D. punctata, the lipophorin is sequestered by developing oocytes before the onset of vitellogenesis and could be related to lipid (and JH) delivery to the ovary at that time. Debernard and Couilland and co-workers (90, 101) studied the catabolism of “H-labeled lOR-JH III in adult female L. migratoria. This in vivo study revealed that JH III acid diol and an unknown compound were the major end products of JH III degradation. When injection of labeled JH tracer occurred 1 day after allatectomy (at this time low or zero JH titer is expected), the disappearance of the labeled JH was similar to that in sham-operated controls. This suggests that the rate of JH catabolism has been considerably decreased after allatectomy. However, additional experiments are required to determine the nature of these changes. 6. Juvenile

hormone-like

effects of retinoic

acid

Retinoids, i.e., retinal, retinol, and retinoic acid, are relatively abundant in the insect body. Both retinoids and JH belong to the sesquiterpenes. In addition to their role in vision, retinoids are known as active morphogenetic agents. The first observation of a JH-mimetic effect is from Slama (635). Retinoid-binding proteins and receptor proteins have recently been discovered in larval epidermis of M. sexta. The receptors belong to the “superfamily” of steroid hormone receptors. When applied to last instar larvae of bugs, Pyrrhocork apterw, Dysdercus cingulatus, and to freshly

996

GiiDE,

HOFFMANN,

molted pupae of T. molitor, retinoic acid exerted a juvenilizing effect. The acid applied to the eggs caused embryonic malformations and further exhibited the capacity to restore reproduction in allatectomized adults of P. apterus and D. cingulatus (468). Retinoic acid stimulated vitellogenin synthesis to the same extent as JH and its analogs. It seems that retinoic acid has the capacity to replenish JH similarly as it was described for juvenoids or even better.

C. Ecdysteroids 1. Chemistry

and PTTH/Bombyxin and bsios yn thesis of ecd ys teroids

Ecdysone was the first invertebrate hormone available in pure form (5Oa). Soon more ecdysone-like substances were found and collectively called ecdysteroids. Ecdysteroids are molting hormones in mandibulate arthropods (63), but they have also been found in other invertebrate animal phyla, where their function may be other than molting (154). Ecdysone is a classical steroid hormone (Fig. 2). It has the usual four-ring nucleus of steroids and the full side chain of cholesterol. Characteristic features of ecdysone are a keto group, found in ring B conjugated to a double bound, as well as five hydroxyl groups (360). The hormonal control includes the regulation of many physiological and biological processes related to molting in embryos, larvae, and nymphs as well as to reproduction in the imago. A closer view as to the identity of the molting hormone has shown that ZO-hydroxyecdysone (ZOE), not ecdysone, is the true molting hormone, because it alone induces molting in larval stages of arthropods. Some insects, e.g., the Heteroptera, use a higher homolog of ecdysone, the 24-methyl-20E or makisterone A, as an alternative molting hormone (137). Honey bees (Hymenoptera) and some flies (Diptera) have also been shown to contain makisterone A as their major ecdysteroid (136). In insect oocytes and embryos, 2deoxyecdysone and 26-hydroxyecdysone may have hormonal functions of their own (Fig. 2). A recent review by Rees (571) quotes 61 different ecdysteroids found in animals (zooecdysteroids), but not all of them are hormonally active. The ecdysone handbook by Lafont and Wilson (374) contains 196 compounds representing either naturally occurring zoo- or phytoecdysteroids, biosynthetic intermediates, or compounds isolated from metabolic studies (see sect. 1vC2). The may protect plants against herbi“phytoecdysteroids” vores, by acting as toxic substances and/or as feeding deterrents (372). Polyphagous insects seem to have developed detoxification mechanisms that efficiently protect them from ingested ecdysteroids, even at high dosages. After feeding of ecdysone or 20E, crickets (G. bimaculatus) efficiently convert these hormones into 14-deoxyec-

AND

SPRING

Volume

77

dysteroids (272; Fig. 2). The 14-dehydroxylation is most probably caused by bacteria from the gut lumen. In larval insects, the prothoracic glands or their homologs (ring gland in Diptera, ventral gland of some primitive insects) are the site of ecdysone biosynthesis. After a variety of stimuli an ecdysiotropic hormone (PTTH) is released from the brain (see sect. 1vc5) and acts as a trigger for the prothoracic gland to secrete ecdysone into the hemolymph. Ecdysone is released from the prothoracic glands in a protein-bound form by exocytosis. There is no storage of hormone within the glands. Ecdysone is then converted to 20E by an ecdysone 20-monooxygenase in particularly in fat body, Malpighian tubules, and midgut. The biosynthesis of ecdysone begins with cholesterol. Because arthropods cannot synthesize cholesterol de novo, they derive it from dietary sources. This poses no problem for zoophagous insects, since cholesterol is the major sterol in animals. Most phytophagous insects can convert plant sterols, such as sitosterol, stigmasterol, and campesterol, to cholesterol by dealkylation (360). Phytophagous insects that are unable to perform this reaction sequence produce makisterone A as principle molting hormone. Recently, it was shown in M. sexta (733) and other Lepidopteran species (346) that the major ecdysteroid released from the prothoracic glands incubated in vitro is not ecdysone, but a mixture of 2-dehydroecdysone (2DE) and 3-dehydroecdysone (3DE), which is rapidly reduced to ecdysone in the presence of a hemolymph 3-dehydroecdysteroid 3@-reductase. The Lepidopteran species tested showed the most potent enzyme activity, although activity was demonstrated in members of other orders (Blattaria, Caelifera, Diptera). Three-dehydroecdysteroid 3P-reductase was recently purified from B. mori larval hemolymph (478). Western blot analysis showed that the enzyme was abundantly present in the hemolymph, but not in other tissues examined. In prothoracic glands of last instar larvae of M. se&a, changes in ecdysteroidogenic capacity are associated with a change in glandular total protein content, a change in the types of proteins synthesized, and a temporally restricted pulse of DNA synthesis, the latter being a possible indicant of cell proliferation (384). The biosynthetic pathway from cholesterol to ecdysone is not completely known (572). Only the first step (conversion of cholesterol to 7-dehydrocholesterol) and some of the last steps can be considered as established. The formation of the A/B &s-ring junction apparently occurs early in the biosynthetic pathway, and 2,22,25-trideoxyecdysone (5P-ketodiol) is a likely intermediate in the biosynthesis of ecdysone. The involvement of 3-0x0a4 compounds as intermediates in ecdysteroid biosynthesis has been shown recently by Blais et al. (30) for the Yorgan of a crustacean species. The 3-0x0-A4 compounds readily underwent a 5@-reduction yielding 5/?-diketol. Indirect evidence has suggested that the transformation of 7-dehydrocholesterol to 5&ketodiol occurs in the mito-

October

1997

INSECT

997

HORMONES OH

OH 9”’

H

:

20E

0

FIG. 2. Structural formulas of some major zooecdysteroids. E, ecdysone; 20E, 20-hydroxyecdysone; 2dE, 2-deoxyecdysone; 3DE, 3-dehydroecdysone; MaA, makisterone A (24Me20E); 14dE, 14deoxyecdysone.

OH

OH i I OH

H

0

chondria (235). Three consecutive hydroxylations convert 5P-ketodiol to ecdysone. The most important sequence in hydroxylation is at C-25, at C-22, and finally at C-Z. Other sequences appear to exist, suggesting multiple pathways of ecdysone biosynthesis, which may differ between insect species. The enzymes catalyzing the final hydroxylation steps have been characterized in some detail (235, 320). They all exhibit properties consistent with classical cytochrome P-450 enzymes. The enzyme that mediates the conversion of cholesterol to 7-dehydrocholesterol is also a cytochrome P-450 monooxygenase. In spite of the undeniable role of prothoracic glands in ecdysteroid biosynthesis of preadult insects (i.e., when ecdysone steroids are involved in molt control), the possible existence of other molting hormone sources has been periodically considered (2 12). Prothoracic glands develop and become functional during late embryogenesis, are active in larval stages, and degenerate in many insect species after the last larval molt or during adult ecdysis. It is now well established that the ovary is capable of synthesizing

14dE

ecdysteroids during late pupal and adult stages, i.e., when reproduction is concerned. Ovarian follicle cells of insects produce ecdysteroids de novo (227, 274, 775), but the fate and function of these compounds may greatly differ according to species (371,592; see sect. 1vC3). For a tissue to be defined as ecdysiosynthetic, it should be established that a sterol precursor is converted by the tissue to give ecdysone or a closely related ecdysteroid (primary ecdysteroid source). In addition to prothoracic glands and ovaries, some other primary ecdysteroid sources exist (103). Gelman, Loeb, and co-workers (211,394) have shown that incubation in vitro of testes of the Lepidopterans H. virescens, Ostrinia nubilalis, and L. dispar from mid to late last-larval instar and pupae release large amounts of ecdysteroids into the medium. Testes from last-instar larvae of S. littoralis also seem to synthesize ecdysteroids (312). Tissues in isolated larval abdomens of B. mori, Mames tra brassicae, M. domes ticus, T. moli tor, Lep tinotarsa decemlineata, and P. americana can convert [‘4C]cholestoro1 to ecdysone and 20E (570). Abdominal

998

GiiDE,

HOFFMANN,

oenocytes may be source of ecdysteroids in T. molitor, 23. mori, and M. domesticus. The first evidence suggesting that integument could be an alternative source of ecdysteroids was given by Cassier et al. (60) in the last larval instar of L. migratoria. Pupal wings of T. molitor, which consist of almost pure epidermis, are a very active ecdysteroid source (103). Recently, it has been observed that several cell lines of epidermal origin could produce ecdysteroids in vitro. These results suggest that the epidermis is capable of being a primary source of ecdysteroids, while retaining the ability to respond to these hormones (see sect. rvC3). Thus ecdysteroids seem to have an autocrine function within the epidermis. The ability of integument to secrete ecdysteroids does not appear to be restricted to molting stages and to an autocrine function. For example, in adult vitellogenic females of G. bimaculatus, during incubation of pieces of the abdominal integument together with the adjacent segmental fat body, a net production of ecdysteroids was observed, similar to that in the ovary (738). The results were corroborated by following the in vitro incorporation of 14C label from cholesterol and 3H label from 5P-ketodiol, respectively, into ecdysone (274). In male adult migratory grasshoppers, M. sanguinipes, the synthetic activity of the integument-fat body complex is -2.5-fold greater than that of the testis (and -7 times that of the MARGs) and may be responsible for ecdysteroids found in the hemolymph, given the parallel changes in integumental ecdysteroid production and hemolymph ecdysteroid titer (214). In last instar larvae of G. bimaculatus, the prothoracic glands reach their maximal ecdysteroid release on day 5. This peak is followed by a maximal release of hormones by the ovary as well as by the abdominal integument on day 6. The highest titer of ecdysteroids in hemolymph is found on day 7, 2 days before imaginal molt (213). These results suggest a hemolymph ecdysteroid composition resulting from at least three hormone sources with temporally different maxima in hormone synthesis and release. In female penultimate instar larvae, however, the prothoracic gland seems to be the only source of hemolyrnph ecdysteroids. 2. Ecdysteroid

metabolism

Many insect tissues are able to metabolize ecdysteroids to a certain extent, the most effective being the fat body, Malpighian tubules (MT) and gut, and the ovary. Ecdysteroid metabolism shows a tissue-, species-, and stage-dependent heterogeneity. Six different types of enzymatic reactions can take place: hydroxylation preferably at C-20 and C-26; oxidation at C-2, C-3, and C-26; reduction at C-3; side-chain cleavage (C-20/22); conjugate formation at C-2, C-3, C-22, C-25, and C-26; and conjugate hvdrolvsis (C-22. C-26: Ref. 373). According to our present

AND

SPRING

Volume

77

knowledge, ZO-hydroxylation, 26-hydroxylation, the formation of ecdysonoic acids, and the reversible formation of 3-dehydroecdysteroids seem to be widespread. Threeepimerization and 14-dehydroxylation were found in the gut of some insect species. It is obvious that different reactions may occur simultaneously on the same molecule. The formation of eedysteroid conjugates is especially heterogeneous (572). Ecdysteroid conjugates can be divided into different classes according to their different polarities. A relatively high concentration of polar conjugated ecdysteroids was initially reported in eggs of B. mori. Ecdysteroids in locust (L. migratoria, S. gregaria) ovaries/eggs also occur almost exclusively as polar conjugates, and they were identified as 22-phosphate derivatives. The cytosolic phosphotransferase from S. gregaria follicle cells has been characterized. In M. sexta, the major conjugate in newly laid eggs is 26-hydroxyecdysone 26phosphate, whereas newly laid eggs of S. littoralis contain primarily 2-deoxyecdysone 22-phosphate. The formation of sulfate esters was conclusively demonstrated in the mosquito, Aedes togoi (629); glucosides were also described from insects (338, 518). Quite different types of conjugates were originally found in tick ovaries and eggs, namely, highly apolar ecdysteroid long-chain 22-fatty-acyl esters (linoleate, palmitate, oleate esters). Meanwhile, analogous ecdysteroid fatty-acyl esters have been found in newly laid eggs of the house cricket A. domesticus (742) and in the ovary of the field cricket G. bimaculatus (49, 266, 682). Interestingly, significant levels of ecdysteroid conjugates do not occur in newly laid eggs of G. bimaculatus (132). Novel ecdysteroid conjugates are increasingly formed during embryogenesis of A. domesticus. These metabolites are double conjugates consisting of ecdysone or 20E esterified to a fatty-acyl group at C-22 and to a neutral polar or negatively charged group possibly at C-25 (743). In adult D. melanogaster females, the ovary showed only a low metabolic activity (230). The gut had the highest capacities of any tissue to synthesize ecdysone 22-fatty acyl esters in vitro, and after injection of [3H]ecdysone, the gut contained large amounts of radiolabeled highly apolar conjugates. However, the main product in vitro was 3DE, as it was in the abdominal carcasses. The hemolymph contained all the different metabolites found in the whole fly, and all metabolites formed can be excreted (231). For the interpretation of the data of the conversion of radiolabeled ecdysone or 20E in vitro and in vivo, however, it should be kept in mind that they do not necessarily reflect the metabolism of endogenous ecdysteroids. In addition, injected and ingested radiolabeled ecdysteroids often showed quite different metabolic patterns (682, 774). What are the functions of these many metabolic reactions? Hydroxylation of ecdysone at C-20 activates the molecule, at least in larvae. In insect embryos where ecdysteroids mav control embrvonic molting (see below). hv-

October

1&x

INSECT

HORMONES

droxylation at C-26 may have the same role. The relative activity of metabolites can be scored by a bioassay, e.g., the “Calliphora test” as used for the first isolation of ecdysone. Most metabolic reactions inactivate the hormone, i.e., they lead to ecdysteroids/ecdysteroid conjugates for storage or excretion. Locust embryos undergo several embryonic molts and cycles of cuticulogenesis, coincident with peak titers of free ecdysteroids. Support for the hypothesis that the polar storage conjugates of S. gregaria eggs, after enzyrnic hydrolysis, could provide a source of active hormone in embryogenesis, at least before differentiation of the prothoracic glands, was furnished by demonstration of enzymatic hydrolysis of conjugates in a cell-free system from embryos (297). Activity of the hydrolytic enzyme increases around the same time that the free ecdysteroid titer increases (598). In later stages of embryonic development, polar ecdysteroids (e.g., ecdysteroid acids, phosphate esters of ecdysteroid acetate) occurred that were quite different from the conjugates in newly laid eggs and were presumed to be inactivation products. Such irreversible inactivation of ecdysteroids could be important in stages of development where excretion is not possible, e.g., eggs and pupae. The highly apolar ecdysteroid conjugates in newly laid eggs of A. domesticus or P. americana also undergo hydrolysis during embryogenesis, releasing free hormone, which is then inactivated by formation of different conjugates. There is some evidence that ecdysteroid conjugates could be a source of active hormone also at stages of development in addition to embryogenesis. For example, pupae of the blowfly S. peregrina at the beginning of adult development can release ZOE from uncharacterized polar conjugates (438). Variation in the rate of ecdysteroid secretion during postembryonic development appears to be an important factor in controlling the hormone titer. Ecdysone and 20E may be excreted as such or as metabolites via the gut or MT. Ecdysteroid-metabolizing enzymes in the fat body, midgut, and MT, in addition to inactivating the hormones, may facilitate excretion (572). Hormone inactivation by the gut may also be important in protecting the insect from ingested phytoecdysteroids. In summary, biosynthesis, metabolism, and excretion together control the hormone concentration in the hemolymph. Rates of all three processes may vary with development in a coordinated fashion. During increased rates of biosynthesis, metabolic reactions and excretion of ecdysteroids are often reduced. 3. Biological

activity

of ecdysteroids

Three major processes are under the control of the ecdysteroids: molting, growth, and gametogenesis (360). The molt and the physiological events associated with molting are controlled bv ecdvsteroids during both em-

999

bryogenesis and larval development. During larval development, remarkable changes are seen in ecdysteroid titer, which indicate a temporal correlation between molts and significant increases in ecdysteroid concentrations in hemolymph (666). In holometabolous (e.g., flies, moths, and beetles) as well as hemimetabolous insects (e.g., locusts, crickets, and cockroaches), the ecdysteroid titer shows a sharp increase at the onset of each larval molt. In general, 20E titers rise before apolysis (separation of the old cuticle from the hypodermis), reach a maximum at, or slightly after, apolysis (when the new epicuticle is being secreted), and fall to low or undetectable levels by the time of ecdysis (726). Recently, it was shown that ecdysteroid titers undergo circadian changes (4). The circadian rhythm of ecdysteroid titers is regulated by a clock that probably is located within the prothoracic glands (95). This finding, together with reports of a daily rhythm of PTTH release from exised brains of the bug R. prolixus, which continues throughout most of the 21 days of larvaladult development, raises significant challenges to several features of our current understanding of the hormonal control of insect development (706). In hemimetabolous insects, the larval-adult molt likewise displays a single massive ecdysteroid peak. This pattern is also seen in the pupal molt of B. mori. Most Lepidoptera, however, exhibit two ecdysteroid peaks in the last larval instar. During the first, smaller rise in concentration, the hormone consists of both ecdysone and 20E, in a ratio of 1:l (94). The function of the hormone is to evoke changes in larval behavior; the larva ceases to feed and searches for a place suitable for pupation. Another important function during this minor peak is reprogramming of the larval epidermis from larval synthesis to pupal synthesis. This peak is apparently lacking in B. mori. The second ecdysteroid peak is much higher, and the ratio of ecdysone to 20E is 1:5. This second peak is related to the larval-pupal molt and probably analogous to the single large peak of earlier larval instars. After pupation, there is another high and very broad peak occupying the period of adult development. If a pupal diapause occurs, the increase in ecdysteroids is delayed until diapause is terminated and adult development commences. The number of larval ecdyses in insects is genetically fixed, but a few additional larval ecdyses can be induced by artificial treatment such as starvation, chilling, or hormone treatment. For example, in B. mori, seven additional larval ecdyses could be induced by supplementing the diet with ecdysone, but not with 20E (678). Ecdysteroids reappear in the adults in association with reproduction (see sect. 1vC1). The physiological significance of the ovarian ecdysteroids varies considerably, depending on the insect species. In many cases, ovarian ecdysteroids are taken up and stored as conjugates in the maturing eggs. During embryogenesis, free ecdysteroids are released from conjugates and will trigger embrvonic

1.000

G;?;DE,

HOFFMANN,

molts (see above). However, ecdysteroid peaks during early embryogenesis have also been observed in insects that do not show embryonic molts. Hagedorn (237) suggested that synthesis of ecdysteroids by the ovaries could be a primitive insect feature and that perhaps the most ancient function of ovarian ecdysteroids was to supply the embryo with molting hormone precursors. Belles et al. (20) proposed that the induction of choriogenesis by 20E, as shown for various cockroach species, could also be among the more primitive functions of ovarian ecdysteroids. Ovarian ecdysteroids may also be involved in oocy-te maturation by triggering reinitiation of meiosis, as shown in L. miglmtoria, P. americana, and G. bimaculatus (381). In the ease of maternal ecdysteroid transfer into the eggs, only a small portion of the ovarian ecdysteroids will be released into the hemolymph (110). The physiological role of the extraovarian ecdysteroids is still controversial. Repeated injection of ma.kisterone A into the hemolymph of adult female milkweed bugs, Oncopeltus fasciatus, causes a decrease in female fertility, correlated with a dose-dependent suppression of the Vgs in the hemolymph. When females of G. bimaculatus were reared at a suboptima1 temperature of ZO”C, oviposition rate was low. Repeated injections of 2-5 rug 20E into these females caused an increase in ovarian fresh weight and stimulated oviposition rate (19). Surgical removal of the CA of young last instar nymphs of G. bimaculatus resulted in molting to morphologically normal adults, and the adult females laid a low number of viable eggs, even in the absence of JH (271). There is good reason to believe that in G. bimaculaPUS ecdysteroids may account for that portion of vitellogenins and number of eggs that are produced and laid by allatect)omized females. Injection of 20E into allatectomized females resulted in increased hemolyrnph Vg titers, although oviposition rate was not affected. ZO-Hydroxyecdysone, therefore, seems to promote fat body Vg synthesis and secretion into the hemolymph but does not enhance Vg uptake into the oocytes. A combined treatment of allatectomized females with 20E and a JH analog resulted in Vg titers and oviposition rates close to that of intact controls (270). In most Dipterans, with subtle differences exist,ing among species, the ovaries secrete substantial amounts of ecdysteroids into the hemolymph, then 20E stimulates the fat body to synthesize Vg (38). Few data are available on ecdysteroid levels in adult male insects. Generally, there is a clear difference in ecdysteroid titer between females and males with a continuous, low hormone level present in males. Only in some species were high concentrations of ecdysteroids detected in the testes. The role of ecdysteroids in adult males is suggested to be associated with spermatogenesis, and perhaps also with functional transitions in the fat body. Repeated pulses of ecdysteroids in premature animals affect both cell cycles in the testes and differentiation of

AND

SPRING

Volume

77

the germ cells (241). In L. migratoria, MARGs become competent to terminal differentiation at a critical period at days 2-3 of the final larval instar, when there is a drop in JH and a transient rise in ecdysteroids. High ecdysteroid activity in adult males at the day of molting may be necessary to complete maturation of the reproductive apparatus and to promote spermatogenesis. In adult male crickets, G. bimaculatus, high ecdysteroid titers may be necessary at the onset of MARG growth and spermatophore production shortly after ecdysis, whereas a drop in ecdysteroid level may enhance release of spermatophores (273). To continue high spermatophore production rates during adulthood, further peaks in ecdysteroid concentration are necessary. Similar results were recently demonstrated for adult males of the migratory grasshopper, M. sanguinipes (302). The testes of adult M. sanguinipes showed a small peak of ecdysteroids on day 2, followed by a decrease to undetectable levels on day 3, the latter coinciding with the movement of sperm into the seminal vesicles. Immediately after copulation, there was a lo-fold increase in testis ecdysteroids compared with those of unmated males. This increase might result from de novo synthesis in the testes and could be significant in terms of spermatogenesis, because in M. sanguinipes, males copulate almost on a daily bases throughout adult life. In some Lepidopteran males (e.g., L. dispar), the hemolymph ecdysteroid titer decreases -1 day before the first release of sperm from the testes. Loeb et al. (394) speculated that in adult male H. virescens, testicular ecdysteroids mediate bursts of sperm release and sexual behavior. That the role of ecdysteroids in adult males is associated with spermatocystic maturation has also been shown in a number of other insects. Very recently, Rachinsky and Engels (541) demonstrated that ecdysteroids strongly influence the pattern of protein synthesis in larval honeybee (A. mellifera) ovaries. These data provoked doubt concerning the role of JH as the major “queen-making” agent in caste development. In the honeybee, a reaction chain is suggested by which an exogenous trophogenic stimulus is first translated into allatotropic signals stimulating the CA to produce JH during the early fifth instar. Subsequently, the high JH titer will turn on an enhanced ecdysteroid production by the prothoracic glands. This increased ecdysteroid titer early in the prepupal phase then probably acts directly at the DNA level, regulating transcription of genes which shift development toward the queen pathway. 4. Modes of action

of ecdysteroids

Steroid hormones act by modification of transcription of specific genes. This dogma applies also to ecdysteroids. Ecdysteroid enters the target cell and migrates to the nucleus where it binds to its receptor molecule. The ecdysteroid receptor complex then interacts with DNA to

October

1997

INSECT

1001

HORMONES

induce new RNA transcripts (360). As a model for studying the mechanisms of action of ecdysteroids, insect epidermis (585), Dipteran salivary glands (391), D. melanogaster imaginal discs (773), and several D. melanogaster cell lines (508) have been used. Direct evidence for the steroid nuclear receptors was obtained by autoradiography and immunohistochemistry (326). Ecdysteroid receptors are site-specific, DNA-binding proteins of -100 kDa. They occur in target cells at an extremely low concentration of - l,OOO/eell(360). The ecdysteroid receptor from D. melanogaster (DmEcR) was purified and its gene closed and sequenced (353). The DmEcR shares similarity with other members of the nuclear steroid receptor superfamily. The cloning of cDNA from Chironomus tentans and A. aegypti encoding homologs (CtEcR, AaEcR) of DmEcR was reported (73, 293). Using cDNA for the DmEcR, Palli et al. (500) isolated a cDNA encoding a member of the steroid hormone receptor superfamily from the tobacco hornworm, M. sexta. The DmEcR functions as an obligate heterodimer with another nuclear receptor, DmUSP. In B. mori, BmEcR and BmCFl are the functional counterparts of DmEcR and DmUSP, respectively, despite considerable sequence divergence between the B. mori and D. melanogaster proteins (675). The AeEcR has five domains similar to those in DmEcR: the regulatory domain A/B, the DNA-binding domain C, the hinge domain D, the ligand-binding domain E, and the F region located at the COOH terminus of the receptor. Talbot et al. (688) have shown three isoforms of the DmEcR that differ in their A/B regions. Ecdysteroid binding to the receptor is characterized by high specificity and high affinity. The receptor concentration is apparently developmentally regulated. Fluctuations in the receptor level correspond well to fluctuations in the ecdysteroid titer, indicating that the number of receptors at any particular stage of insect development may be regulated by the presence of the hormone (100). A well-known example for the regulation of gene expression by ecdysteroids is the ecdysteroid-dependent melanization of the insect cuticle. In M. sexta larvae, the decline of the ecdysteroid titer shortly before the larval molt regulates the expression of the dopa decarboxylase (DDC) gene and leads to synthesis of DDC and subsequent melanin synthesis (263). Recent results by Koch (351) suggest that Lepidopteran wing coloration in the scales is regulated by ZOE in a similar way as in M. sexta cuticle. In the wing color pattern, however, not only a time-dependent but also the pattern-specific DDC expression must be regulated in differential response to the ecdysteroid titer. Ecdysteroid receptors are preferred target sites for ecdysteroid agonists and antagonists. Wing et al. (752) have found that certain substituted hydrazines (e.g., RH5849, RH-5992) can act as nonsteroidal ecdysteroid agonists. These compounds bind to ecdysteroid receptors

with moderate affinity and elicit all known biological effects of ecdysteroids (400, 444, 636). The search for antiecdysteroids became successful when the ability of certain plant brassinosteroids to compete with ecdysteroids for the binding site of ecdysteroid receptors was recognized (326). When applied to intact insects, brassinosteroids inhibited molting. Ecdysteroid agonists and antagonists may become interesting tools for the study of ecdysteroid effects and open the prospect for finding safer insecticides (584). Over the years, more effects of ecdysteroids have been reported that cannot be explained by the model of ecdysteroid-induced alterations of transcription. They indicate that ecdysteroids may also have extramolecular activities. In these cases, the plasma membrane is the apparent target, but details of this mechanism remain to be established (326). 5. Ecdysiotropic

and ecdysiostatic

neuropeptides

Biosynthesis of ecdysteroids in prothoracic glands of insect larvae is under the control of a neuropeptide, PTTH (formerly known as insect brain hormone or activating hormone). Most research on the chemistry of PTTH has been conducted with the Lepidopterans B. mori and J& sexta (734). Prothoracicotropic hormone exists in multiple molecular forms which, based on apparent molecular masses, fall into two major groups: big PTTHs (Z-29 kDa) and small PTTHs (4-7 kDa). In the case of PTTH from M. sexta brains, both molecular forms are active in a PTTH bioassay (in vitro and in vivo). This contrasts with the results attained with PTTH from adult B. mori brains. Here the small PTTH was active only in another Lepidopteran, Samia Cynthia (PTTH-S; S denoting S. Cynthia). The big PTTH was active in B. mori and has been termed PTTH-B. Because PTTH-S does not appear to have prothoracicotropic activity in 23. mori, it has been renamed bombyxin. Studies using SDS-polyaerylamide gel electrophoresis (PAGE) indicated that the big PTTH is a dimeric molecule in which identical or nearly identical subunits are linked by a disulfide bond. In 1991, the primary structure of the big B. mori PTTH (PTTH-B) was clarified to a great extent by using pure PTTH prepared from 3 X lo6 B. mori heads (301, 322, 455). Peptide sequencing revealed that the PTTH monomer consists of at least 104 amino acid residues, although the residue at position 41 could not be identified. On the other hand, cDNA for B. mori PTTH has been cloned and its nucleotide sequence determined (327). The amino acid sequence deduced from the nucleotide sequence suggested that the PTTH monomer consists of 109 amino acid residues, the first 104 of which were precisely identical to the sequence of the peptide shown above (Fig. 3). Amino acid 41 has now been shown to be Asn, indicating that a carbohydrate moiety is attached to this site (the

1002 GNIQVENQAI N%TQQPTCRP LKYRWVAESH

GiiDE,

PDPPCTCKYK PYICKESLYS PVSVACLCTR

KEIEDLGENS ITILKRRETK DYQLRYNNN

HOFFMANN,

VPRFIETRNC SQESLEIPNE

AND

SPRING

Volume

77

hemolymph was low and almost constant during the fourth and early fifth instars of B. mori. In males, the titer rose abruptly 3 days after the beginning of wandering phase. One day after pupation, it rose again steeply to FIG. 3. Amino acid sequence of Bombyx prothoracicotropic hormone (Born-PTTH, PTTH-B). # Carbohydrate moiety linked. reach a maximal level that lasted until the midst of the pupal stage. The titer decreased thereafter and increased again at adult emergence. Females showed a similar patglycosylation is not essential for biological activity). The tern of titer fluctuation, but during pupal-adult developcoding region of the PTTH cDNA encodes a signal peptide, ment, titers were two to three times higher than in males. a Z-kDa peptide, a 6-kDa peptide, and the PTTH subunit Nucleotide sequence analysis of the genomic DNA in this order from the V-end. The PTTH is first synthesized encoding bombyxin revealed that the overall organization as a large precursor protein consisting of 224 amino acids of the preprobombyxin gene is the same as that of preproand then cleaved to liberate the PTTH subunit. The PTTH insulin genes. Bombyxin, therefore, might be biosynthesubunit contains seven Cys residues, and disulfide linksized in the same manner as are insulins. The bombyxin ages will be formed between (1 bond) and within PTTH genome contains multiple copies of the bombyxin gene. subunits (3 bonds), before or after the proteolytic cleav- This contrasts strongly with the vertebrate insulin genes, age to generate a mature PTTH. Southern hybridization which exist as one or two copies per haploid genome. analysis showed that the B. mori haploid genome conLess is known about the chemistry of M. sexta PTTH. tained a single copy of the PTTH gene. The PTTH appears In M. sexta, PTTH again appears to exist in two different to share a common ancestral gene with the vertebrate size groups, a big PTTH with different variants of -25.5 growth factors, P-nerve growth factor, transforming kDa and a bombyxin-like heterogenous PTTH of -7 kDa. growth factor-P& and platelet-derived growth factor BB Both forms directly stimulate prothoracic glands of M. (477). Immunohistochemistry of the B. mori brain-retrosexta in vitro. With the use of immunoaffinity chromatogcerebral complex using a monoclonal antibody recogniz- raphy and SDS-PAGE, the big PTTH has been purified. Its ing PTTH indicated that the PTTH is produced in two native form appears to be a dimer with monomers of 16.5 pairs of dorsolateral neurosecretory cells in the brain and kDa. Some physical data reveal that the M. sexta PTTH transferred to the CA, where it may be liberated into the is an asymmetrical, acidic, homodimeric peptide with inhemolymph (432). tra- and intermolecular disulfide bonds (442). Four HPLCFive molecular species of bombyxin (bombyxin I-V), separated fragments of this peptide were sequenced and which differ only slightly from one another, have been exhibited no sequence similarity with B. mori PTTH. In isolated and characterized. These molecules can induce agreement with this finding, the B. mori PTTH (which is adult development in a brainless pupa of S. Cynthia and also present inM. sexta; Ref. 232) had no PTTH bioactivity in M. sexta. One sequence fragment of the M. sexta PTTH also stimulate in vitro production of ecdysone in prothoracic glands of S. Cynthia, but their intrinsic function in shows 70% similarity to the vertebrate cellular retinoid B. mori is still unknown. So far, the amino acid sequences of bombyxins II and IV have been completely determined A (Fig. 4; Ref. 458). Bombyxin is a heterodimer; the A chain A chain -II: GIVDECCLRP CSVDVLLSYC consists of 20 amino acids and the B chain of 28 amino -IV: -V _____ IQ- TL---AT-acid residues. Bombyxins have -40% sequence similarity