Proc. Natl. Acad. Sci. USAVol. 90, pp. 2456-2460, March 1993Biochemistry

Callatostatins: Neuropeptides from the blowfly Calliphora vomitoriawith sequence homology to cockroach allatostatinsHANNE DUVE*, ANDERS H. JOHNSENt, ALAN G. SCOTT*, CAO G. Yut, KOICHIRO J. YAGIt,STEPHEN S. TOBEt, AND ALAN THORPE*§*School of Biological Sciences, Queen Mary and Westfield College, University of London, Mile End Road, London El 4NS, United Kingdom; tDepartment ofClinical Biochemistry, Rigshospitalet, University of Copenhagen, DK-2100 Copenhagen 0, Denmark; and tDepartment of Zoology, University of Toronto,ON, Canada M5S lAl

Communicated by Hans H. Ussing, November 25, 1992

ABSTRACT Five neuropeptides with C-terminal aminoacid sequence homology to cockroach allatostatins have beenidentified in the blowfly Calliphora vomitoria. Three have thesame pentapeptide C-terminal amino acid sequence as alla-tostatin 1 ofthe cockroachDiplopterapunctata. A hexadecapep-tide designated callatostatin 1, isolated from thoracic ganglia,brains, and heads, has the sequence Asp-Pro-Leu-Asn-Glu-Glu-Arg-Arg-Ala-Asn-Arg-Tyr-Gly-Phe-Gly-Leu-NH2. Calla-tostatins 2 and 3 have been isolated from heads and thoracicganglia, respectively; they comprise the last 14 and 8 residuesof callatostatin 1. Callatostatin 4, isolated from thoracic gan-glia, has the sequence Xaa-Arg-Pro-Tyr-Ser-Phe-Gly-Leu-NH2, where Xaa is either Asp or Asn. This peptide, with aserine substitution for glycine at position 5, has a C-terminalpentapeptide sequence identical to that of allatostatins 3 and 4ofD. punctata. Callatostatin 5, with the sequence Gly-Pro-Pro-Tyr-Asp-Phe-Gly-Met-NH2, was identified from whole flies.All five peptides inhibit juvenile hormone production by thecorpora allata of D. punctata in vitro. Caliatostatin 5 was themost potent allatostatin so far tested in this species, withmaximum inhibition occurring at 1 nM. In contrast, none of thecaliatostatins or the allatostatins showed allatostatic activity inmature female C. vomitoria when tested at concentrations of100 to 0.1 ,uM. In accordance with these results, immunore-activity to an antiserum directed against the common C ter-minus of callatostatin 1 and aliatostatin 1 was observed in thecorpora allata of D. punctata but not in the corpus allatum ofC. vomitoria, despite its presence in neurons of the brain.Neurons in the thoracic ganglion of C. vomitoria that areimmunoreactive against this antiserum project to the hindgut,rectum, rectal papillae, and oviduct, suggestive of a functiondifferent from that of a true allatostatin.

Neuropeptides capable of inhibiting the production ofjuve-nile hormone by the corpus allatum of insects have beentermed allatostatins (1). To date, members of this class ofneuropeptides have been identified in only two species. In thecockroach Diploptera punctata, five allatostatins ranging insize from 8 to 18 amino acid residues have been characterized(2-4). These neuropeptides show C-terminal similarity,which suggests that they are members of a family, possiblyoriginating from a single gene. From the lepidopteran Man-duca sexta, an N-terminally blocked allatostatin with anamino acid sequence different from those of the cockroachhas recently been identified (5).We report here on the identification, structure, localiza-

tion, and biological activity of five neuropeptides from theblowfly Calliphora vomitoria with sequence homology to thecockroach allatostatins.

EXPERIMENTAL PROCEDURESInsects and Tissue Extraction. Adult C. vomitoria main-

tained at 25°C and 65% relative humidity were fed a diet ofsugar, water, and protein (either Ovaltine or liver). Sourcesof material and extraction procedures were as follows; dis-sected thoracic ganglia (n = 3625), 80% methanol/0.1 MHC1/0.1% 2-mercaptoethanol (6); dissected brains (n =

1000), boiling 0.9% NaCl (2); heads (n = 20,000), 87%methanol/5% acetic acid/8% water followed by acetoneprecipitation (7); whole flies (n = 10,000), 1 M acetic acid/0.02 M HC1/0.1% 2-mercaptoethanol (8).Chromatography. A summary of the chromatographic pro-

cedures for the purification of callatostatins from heads,thoracic ganglia, and whole flies is given in Fig. 1. For brains,the protocol of the first three HPLC steps has been publishedpreviously (9). A final step added to the procedure wasidentical to that given in the protocol for the purification ofthe peptides from whole flies (Fig. 1). The initial purificationof callatostatin 1 from thoracic ganglia was as described in astudy of the Calliphora -Phe-Met-Arg-Phe-NH2 peptides (thecalliFMRFamides; ref. 6).Radioimmunoassay. Column eluants (10-200 ,ul) of each

fraction were monitored by means ofa RIA using a polyclonalantiserum specific for callatostatin 1 and 125I-labeled syn-thetic callatostatin 1 (Core Facility, Insect Biotech Canada,Kingston, ON). The basic details of the RIA protocol areidentical to those described earlier for YGGFMRF (8). Theassay requires the carboxyamidated hexapeptide C-terminalamino acid sequence of callatostatin 1 for full recognition.Thus, allatostatin 1 from the cockroach was detected at alevel of only -10%, despite the fact that the C-terminalpentapeptide sequence is identical to that of callatostatin 1.With allatostatins 2-4, where substitutions with respect toallatostatin 1 occur at the fourth residue from the C terminus,recognition in the assay was <1%. Callatostatin 5, in whichthe C-terminal leucine residue is replaced by methionine, andnonamidated callatostatin 3 were not detected by this RIA.Amino Acid Sequence Analysis. The amino acid sequences

of the purified peptides (5-50 pmol) were determined with anautomatic protein sequencer (model 475A, Applied Biosys-tems) equipped with an on-line HPLC system for the detec-tion of the amino acid phenylthiohydantoin derivatives,which were separated on a C18-DB column (5-7943, Supelco).All chemicals and solvents were sequence or HPLC grade(Applied Biosystems).Mass Spectrometry. A portion (5-50 pmol) of each of the

purified peptides was analyzed by using a Bio-Ion 20 plasmadesorption time-of-flight mass spectrometer (Applied Biosys-tems). The dried samples were dissolved in 10 ,ul of 0.1%trifluoroacetic acid in 20% acetonitrile and two 5-,lI aliquots

Abbreviations: JHB3, juvenile hormone III bisepoxide; JH III,juvenile hormone III.

5To whom reprint requests should be addressed.

2456

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 90 (1993) 2457

were applied to aluminized Mylar foil (coated with nitrocel-lulose) and evaporated. The spectra were recorded for 1-6 x106 primary ions. The method has an accuracy of 0.1%. Forcallatostatin 1, the number of free carboxyl groups wasdetermined by subjecting an aliquot to methylation andremeasuring the mass. The method requires a 2-hr reaction ofthe peptide with 2 M HCl in methanol, obtained by additionof acetyl chloride to dry methanol (10).

Peptide Synthesis. Peptides were synthesized on an AppliedBiosystems model 430A automatic solid-phase peptide syn-thesizer (Insect Biotech Canada). They were purified on apreparative C18 reversed-phase column (Vydac).

Assays for Aliatostatic Activity. The corpus ailatum of C.vomitoria produces juvenile hormone III bisepoxide (JHB3).Details of the radiochemical assay used to measure the ratesof biosynthesis and release of this compound have beenpublished previously (9), as have details of the assay used tomeasure the rates of production ofjuvenile hormone III (JHIII) in D. punctata (2, 11-13).Immunocytochemistry. For immunocytochemistry, aque-

ous Bouin's fixative was used (14, 15). Paraffin sections (6,um) were obtained from the brain, corpus allatum, andthoracic ganglion of 7-day-old female C. vomitoria flies fedsugar, meat, and water, as well as from the brain and corporaallata of 2-day-old virgins or 7-day-old mated D. punctata.They were immunostained with two antisera specific for thecommon C terminus of callatostatin 1 and allatostatin 1 byusing the peroxidase-antiperoxidase technique (16). Forwhole mounts, the indirect immunofluorescence techniquewas used (17). Controls included liquid-phase absorption ofthe antisera with callatostatin 1 or allatostatin 1. Resultsshowed conclusively that both peptides (10 nM) were able to

abolish the immunostaining of the two antisera (diluted1:1000).

RESULTS

Peptide Identification. A total of five peptides, designatedcallatostatins 1-5, were isolated in this study (Fig. 2). Inaddition, partial amino acid sequences and/or molecularweights were recorded for two other different callatostatin-immunoreactive peptides (not further discussed).The hexadecapeptide callatostatin 1 (Fig. 2) was initially

identified during the purification of a series of N-terminallyextended -FMRFamide peptides (the calliFMRFamides)from thoracic ganglia of C. vomitoria [22.2% CH3CN, step 5(6)]. In the present study, the same peptide was also obtainedfrom extracts of heads and brains monitored with the sub-sequently developed RIA for callatostatin 1 (Figs. 1 and 3).After methylation of an aliquot of callatostatin 1 obtainedfrom thoracic ganglia, a molecular weight of 1949.1 wasrecorded. This compares with an expected molecular weightof 1948.2 with free carboxyl groups on the three acidic aminoacids, Asp', Glu5, and Glu6, but not the C-terminal Leu16. Theconclusion both from this experiment and from the results ofthe RIA, which is specific for the amidated C terminus, is thatcallatostatin 1 is amidated.

Also obtained from heads and identified by means ofcallatostatin 1 RIA was the truncated tetradecapeptide cal-latostatin 2 (Fig. 2). The truncated octapeptide, callatostatin3, also immunoreactive in the callatostatin 1 RIA, wasobtained from thoracic ganglia. Insufficient amounts of cal-latostatin 2 or 3 were available for methylation. However,from the results of RIA compared with semiquantitative

HeadsCH30H/CH3COOH

Waters CiR. 300X7 8mm, 10pm, 125ACH3CN/T A1-3%/omin, 1i5ml/min

37 2-4322% 46 4-5766%

Waters C1, 300X7 8mm, 10pm, 125ACH3CN/Np, OAc16/3omin-3/ 1 5l/min

38 2-4488% 45 8-4822% 50 8-5540/o 56 2-6022%

BioRad Hi-Pore C4, 250X4 6mm, 5pm, 300ACH3CN/TFA05%/min, 1 5ml/min

18 40/o 19 4-2000/o

Waters Protein-Pak 125, 300X7 8mm, 10pm, 125nmCH3CN/TFA0-50/o/min, 1 5ml/min

750/o 740-7300/o

Vydac C1, 150x2 1mm, 5pm, 300ACH3CN/4TA050/./min, 0 2ml/min

2180/o 2220/o 2255% 2344%

#2 #1

Thoracic gangliaCH30H/HCI

Waters Cj . 300X3 9mm, 10pm, 120ACH3CN/TFA1-70/o/min, 1 5ml/min

24-290/o

Waters Phenyl, 300X3 9mm, 10pm, 120ACH3CN/TFA0 50/o/min, 1 5ml/min

21 4-2470/o 24 8-2600/o

BioRad Hi-Pore C4, 250X4 6mm, 5pm, 300ACH3CN/TFA0-50/o/min, 1i5ml/min

182-1970/o

Waters Protein-Pak 125, 300X7 8mm, 10pm, 12 5nmCH3CN/TFA0-50/o/min, 15ml/min

778-76-60/o* '

VydacC0B, 150X2-1 mm, 5pm, 300ACH3CN/TFA05/mvn, 0 2ml/min

18 80/o 19 00/o 19 40/o

Vydac Phenyl, 150x2 5mm, 5pm, 300ACHsCN/TFA01.50/amin, 0 72m/8min

16 50% 17 0%O/ 18 60%

#3 #4

Whole fliesCH3COOH/HCI

Sephadex G 50 FCH3COOH, 5%0

Kav=0-62

Waters CiR, 300'7 8mm, 10pm, 125ACH3CN/TFA1 3%6/min, 1i5mi/min

4400%

BioRad Hi-Pore C4. 250X4 6mm, 5pjm, 300ACH3CN/TFA0 50/o/min, 1i5mi/min

22 1-23/6%

Waters Phenyl, 300-39mm, 10, 120ACH3CN/TFA1°0A/min, 1 5ml/min

208-2240/o

BioRad Hi-Pore C4, 250-4 6mm, 5pm, 300ACH3CN/NH4OAc050//min, 1-5ml/min

2000%

Vydac C8, 150.21mm, 5pm, 300ACH3CN/TFA050/m/rnin, 0-2ml/min

2200/o

#5

FIG. 1. Summary of chromatographic procedures for the isolation of callatostatins from C. vomitoria. Trifluoroacetic acid (TFA) was usedat 0.1%, and ammonium acetate (NH4OAc) was used at 10 mM, pH 6.5. *, Purification of this material (steps 3-5, ref. 6) yielded callatostatin1.

Biochemistry: Duve et al.

Proc. Natl. Acad. Sci. USA 90 (1993)

Amino acid sequence

1.Asp-Pro-Leu-Asn-Glu-Glu-Arg- Arg-Ala-Asn-Arg-Tyr-Gly-Phe-Gly-Leu-NH22. Leu-Asn-Glu-Glu-Arg-Ar-Ala-Asn-Arg-Tyr-Gly4PheGly-Leu-NH23. Ala-Asn-Arg-Tyr-Gly-Phe-Gly-Leu-NH2

4 ~~~~~~~~~(Asp*4. (ASs frg-Pro-Tyr6e0Phe-Gly-Leu-NH2

5. Gly-Pro- Pro-Tyr-Asp-Phe-Gly-Met-NH2*

Mrmeasured expected

1903 9 1906 1

1694 5 1693-9

896 8 896 0

953-0 953*0952.0

883 5 883 0

FIG. 2. Amino acid sequences of callatostatins 1-5 isolated from C. vomitoria. Reference should be made to Fig. 1 and text for details ofthe sources of the peptides. Notes: For callatostatin 4, the first residue was difficult to determine. However, only Asp or Asn fit the recordedmolecular weight. *, Amidation of callatostatin 5 is suggested by the intense biological activity shown by this peptide in the amidated form asopposed to a complete lack of activity of the free acid (see text and Fig. 4). ( ), Tentative assignment from sequence analysis; *, assignmentbased on the recorded molecular weight and expected sequence similarity.

results from sequence analyses, we conclude that thesepeptides are C-terminally amidated.The octapeptide callatostatin 4, where the residue at posi-

tion 1 is either asparagine or aspartic acid, was also obtainedfrom thoracic ganglia. This peptide was only weakly immu-noreactive in the callatostatin 1 RIA, and we cannot be certainthat it is carboxyamidated. However, the fact that the C-ter-minal pentapeptide sequence of this peptide is homologouswith those of the cockroach allatostatins 3 and 4 makes it islikely that the posttranslational amidation has been conserved.

Callatostatin 5 was obtained from whole flies (Figs. 1 and2) (8). This peptide was the major sequence of a doublet in aprominent UV peak which showed weak immunoreactivity inRIA against YGGFMRF. The minor sequence of this pair(incompletely identified) was probably the cause of theYGGFMRF immunoreactivity.

Assays for Aliatostatic Activity. Callatostatins 1-4, synthe-sized in the amidated form, and callatostatin 5 in bothamidated and free acid forms, were tested for allatostaticactivity in both C. vomitoria and D. punctata according topublished methods (9). Callatostatins 1-5, at selected con-centrations from 100 to 0.1 ,uM, were inactive on the corpus

D

EEcmC%J

c

CD0cb

(a)

.0Ec00.0

4:

34

24

14

41

5

i a......::,......,::::

..::::::::::::

. ]30..

10

0

I0o

co

._

L-

.0c

r-

z

I

14 19 23

HR.....b

v 14 19 23

Fraction

FIG. 3. Profiles of Vydac C18 HPLC step 5 in the isolation ofcallatostatins from heads. (a) Optical density at 214 nm (mAU,absorbance milliunits). (b) Callatostatin 1 RIA. Fraction 14 yieldedcallatostatin 1. We were unable to complete the identification of thepeptide giving rise to the immunoreactivity in fraction 15.

allatum of C. vomitoria (6- to 10-day-old female flies in whichthe first gonadotrophic cycle was advanced or complete). Therate of production of JHB3 remained unchanged over a 3-hrtest period relative to controls. In contrast, JH III biosyn-thesis in the cockroach was inhibited with each of thecallatostatins, the most potent peptide being callatostatin 5(IC50 = 0.1 nM) (Fig. 4). Callatostatin 5 in the free acid formwas inactive on the corpora allata of C. vomitoria even at 100,uM and showed only a slight inhibition on the corpora allataof D. punctata at 100 nM.

Immunocytochemistry. Immunocytochemical studies haveidentified callatostatin 1 and allatostatin 1 immunoreactivityin perikarya in the brain-suboesophageal ganglion of both C.vomitoria and D. punctata, as well as in axons and arboriza-tions in the neuropil (Fig. 5). The patterns of immunostainingwere identical with these two antisera and the results shownhere were obtained with callatostatin 1 antiserum. In D.punctata, (Fig. SE), but not in C. vomitoria (Fig. SA),callatostatin 1-immunoreactive nerve terminals were ob-served in the corpora allata. Immunoreactive perikarya andarborizations in the neuropil were also observed in thethoracic ganglion of C. vomitoria (Fig. SB). A group ofneurons in the abdominal ganglion projected axons posteri-

-Log peptide (M)

FIG. 4. Dose-responses for inhibition of JH III release fromcorpora allata ofvirgin (day 2) D. punctata by synthetic callatostatins1-5. Percentage inhibition of groups of treated glands relative tountreated controls is shown (3-hr incubations). Each point forcallatostatin treatment is the mean of 5-16 measurements; verticalerror bars show standard error of the mean. o, Callatostatin 1; *,callatostatin 2; v, callatostatin 3; v, callatostatin 4 (synthesized asAM-Arg-Pro-Tyr-Ser-Phe-Gly-Leu-NH2); o, callatostatin 5; *, cal-latostatin 5 (free acid).

2458 Biochemistry: Duve et al.

Proc. Natl. Acad. Sci. USA 90 (1993) 2459

C

pa

postt

FIG. 5. Drawings and paraffin sections of tissues of C. vomitoria (A-C) and D. punctata (D and E), immunostained with theperoxidase-antiperoxidase technique using a callatostatin 1 antiserum that recognizes the carboxyamidated C-terminal pentapeptide commonto both species. (A) Longitudinal section of corpus allatum of a 7-day-old meat-fed female C. vomitoria, showing absence of immunoreactivity.Arrows indicate immunoreactive material in cardiac recurrent nerve bypassing the corpus allatum (x400). (B) Whole mount of thoracic ganglionof C. vomitoria in dorsal view, showing the immunoreactive perikarya and axons projecting to the hindgut along the abdominal nerve.Immunoreactive material in peripheral nervous system on the surface of the metathoracic nerve bundles to the legs is indicated by arrows. (C)Camera lucida drawing of longitudinal part of hind gut, showing immunoreactive axons. (Graticule = 100 Am.) (D) Frontal section through thebrain of 7-day-old mated female D. punctata at the level of lateral neurosecretory cells, showing immunoreactive cell bodies, nerve fibers andarborizations in neuropil (x400). (E) Longitudinal section ofone of the corpora allata ofD. punctata, showing immunoreactivity in arborizationsbetween the glandular cells of a 2-day-old virgin (x500). Abbreviations: ab g, abdominal ganglion; ca, corpus allatum; cc, corpus cardiacum;cm, cardiac recurrent nerve; il, ileum; Inc, lateral neurosecretory cells; m ab n, median abdominal nerve; pa, rectal papilla; pns, peripheralnervous system; post, posterior; pr, proventriculus; re, rectum; rp, rectal pouch; rv, rectal valve.

orly to the hindgut, rectum, rectal papillae, and oviduct (Fig.5 B and C). In D. punctata, lateral neurosecretory cells of thebrain showed strong callatostatin 1 immunoreactivity in7-day-old mated females (Fig. 5D) but weak immunoreactiv-ity in 2-day-old virgins. In contrast, immunoreactive cells ofthe median neurosecretory group appeared the same inanimals of both ages (data not shown).

DISCUSSIONIn this study we have identified five peptides from the blowflyC. vomitoria with sequence homology to the allatostatins ofthe cockroach D. punctata (2-4). The peptides have beendesignated callatostatins to take into account (a) the speciesof origin [also in line with the recent calliFMRFamide des-ignation (6)], (b) the apparent sequence homology with thecockroach allatostatins, and (c) their ability to inhibit JH IIIproduction in the cockroach. Callatostatins 1-5 were isolatedfrom thoracic ganglia, brains, and heads, as well as fromwhole flies. The C-terminal pentapeptide sequence -Tyr-Gly-Phe-Gly-Leu-NH2 of callatostatins 1-3 is identical with thatof allatostatin 1 from D. punctata. Callatostatin 4 shares withallatostatins 3 and 4 the substitution of serine for glycine atposition 4 from the C terminus. Callatostatin 5 is differentfrom any of the cockroach allatostatins identified to date in

having methionine instead of leucine at the C terminus andaspartic acid at position 4 from the C terminus. We wereunable to show whether the methionine is amidated, althoughbioactivity studies with both free acid (negative) and ami-dated form (strongly positive) suggest that it is. The residueat position 6 from the C terminus in allatostatins 1-4 isleucine, whereas in callatostatins 1-3 it is arginine and incallatostatins 4 and 5 it is proline. This residue appears to bevariable even in the cockroach, since ASB2 (allatostatin 5)has valine in this position (4).An interesting feature of callatostatin 1 is the presence of

a pair of arginine residues at positions 7 and 8. A similarpotential dibasic cleavage site (-Lys-Arg-) occurs at positions9 and 10 of the cockroach octadecapeptide allatostatin 5 (4).In C. vomitoria, we have isolated the octapeptide (callatosta-tin 3) that would result from such prohormone processing.However, this sequence may also be coded for separately onthe callatostatin gene. Furthermore, we have identified cal-latostatin 2, which lacks the first two N-terminal residues,Asp-Pro-. This peptide could originate from posttranslationalcleavage of the prohormone, but it could also represent adistinct amino acid sequence determined by the gene, or anartifact of the extraction procedure (18).

All callatostatins have the ability to inhibit JH III releasewhen tested on corpora allata of 2-day-old virgins of D. punc-

Biochemistry: Duve et al.

Proc. Natl. Acad. Sci. USA 90 (1993)

tata. The most potent peptide so far tested in this species is, infact, callatostatin 5, not yet identified in D. punctata itself. Ofthe other peptides, callatostatins 1 and 2 were more potent thanthe shorter callatostatin 3, a finding in agreement with struc-ture-activity studies of the cockroach allatostatins (4).Although peptides with allatostatic activity have recently

been shown to exist in partially purified brain extracts of C.vomitoria (9), none of the callatostatins 1-5 identified in thepresent study (and none of the cockroach allatostatins 1-5)was capable of inhibiting the production of JHB3 from thecorpus allatum of C. vomitoria under our assay conditions.This suggests either that the callatostatins are important inthe inhibition of JHB3 production during other stages of thelife cycle (e.g., in larval stages or during adult diapause) orthat they serve a different function. Evidence for the latterhas come from immunocytochemical studies of the centralnervous system and retrocerebral complex of C. vomitoriausing a C-terminal-specific callatostatin 1 antiserum. Axonsof the callatostatin 1-immunoreactive neurons in the medianneurosecretory cell group do not appear to project to thecorpus allatum. Furthermore, the lateral neurosecretorycells, another known source of innervation of the corpusallatum, show no immunoreactivity to the callatostatin anti-serum. In contrast, the corpora allata of the cockroach areinfiltrated by callatostatin 1-immunoreactive material, pre-sumably within nerve terminals of axons originating from themedian or lateral neurosecretory cells. Of particular interestis the fact that the callatostatin 1-immunoreactive neurons ofthe thoracic ganglion of the blowfly directly innervate thehindgut, rectum, rectal papillae, and oviduct. These neuronsdo not appear to transport any of their material to the dorsalneural sheath for release into the hemolymph. Thus, itappears unlikely that the callatostatins so far identified affectthe corpus allatum, even indirectly.

In conclusion, it is clear that despite a conservation ofamino acid sequence within the "allatostatins" in two widelydivergent groups of insects, a conservation offunction of thepeptides-i.e., as inhibitors of juvenile hormone-is notapparent.

We are grateful to Alan Kastrup for expert technical assistance.This study was supported by the Agricultural and Food ResearchCouncil of Great Britain (AG68/022) (A.T.), the Novo Foundation(A.H.J.), and the Natural Sciences and Engineering Research Coun-cil of Canada and Insect Biotech Canada, a Network of Centres ofExcellence funded by the Government of Canada (S.S.T.).

1. Tobe, S. S. (1980) in InsectBiology ofthe Future: VBW80, eds.Locke, M. & Smith, D. R. (Academic, New York), pp. 345-367.

2. Woodhead, A. P., Stay, B., Seidel, S. L., Khan, M. A. &Tobe, S. S. (1989) Proc. Natl. Acad. Sci. USA 86, 5997-6001.

3. Pratt, G. E., Farnsworth, D. E., Siegel, N. R., Fok, K. F. &Feyereisen, R. (1989) Biochem. Biophys. Res. Commun. 163,1243-1247.

4. Pratt, G. E., Farnsworth, D. E., Fok, K. F., Siegel, N. R.,McCormack, A. L., Shabanowitz, J., Hunt, D. F. & Feyerei-sen, R. (1991) Proc. Natl. Acad. Sci. USA 88, 2412-2416.

5. Kramer, S. J., Toschi, A., Miller, C. A., Kataoka, H., Quistad,G. B., Li, J. P., Carney, R. L. & Schooley, D. A. (1991) Proc.Nat!. Acad. Sci. USA 88, 9458-9462.

6. Duve, H., Johnsen, A. H., Sewell, J. C., Scott, A. G.,Orchard, I., Rehfeld, J. F. & Thorpe, A. (1992) Proc. Natl.Acad. Sci. USA 89, 2326-2330.

7. Kay, I., Wheeler, C. H., Coast, G. M., Totty, N. F., Cusinato,O., Patel, M. & Goldsworthy, G. J. (1991) Biol. Chem. Hoppe-Seyler 372, 505-512.

8. Duve, H., Sewell, J. C., Scott, A. G. & Thorpe, A. (1991)Regul. Pept. 35, 145-159.

9. Duve, H., Thorpe, A., Yagi, K. J., Yu, C. G. & Tobe, S. S.(1992) J. Insect Physiol. 38, 575-585.

10. Talbo, G., H0jrup, P., Rahbek-Nielsen, H., Andersen, S. 0. &Roepstorff, P. (1991) Eur. J. Biochem. 195, 495-504.

11. Tobe, S. S. & Pratt, G. E. (1974) Biochem. J. 144, 107-113.12. Pratt, G. E. & Tobe, S. S. (1974) Life Sci. 14, 575-586.13. Tobe, S. S. & Clarke, N. (1985) Insect Biochem. 15, 175-179.14. Duve, H. & Thorpe, A. (1990) Cell Tissue Res. 259, 147-157.15. Duve, H., Thorpe, A. & Tobe, S. S. (1991) Cell Tissue Res. 263,

285-291.16. Stemberger, L. A. (1974) Immunocytochemistry (Prentice-

Hall, Englewood Cliffs, NJ).17. Tsang, P. W. & Orchard, I. (1991)J. Comp. Neurol. 311, 17-32.18. Conlon, J. M., Arnold-Reed, D. E. & Balment, R. J. (1990)

Peptides 11, 891-895.

2460 Biochemistry: Duve et al.