Corazonin gene expression in the waxmoth Galleria mellonella

Insect Molecular Biology (2001)

10

(4), 341–346

© 2001 Blackwell Science Ltd

341

Blackwell Science, Ltd

Corazonin gene expression in the waxmoth

Galleria mellonella

I. A.

Hansen,

1

F.

Sehnal,

2

S. R.

Meyer

1

and K.

Scheller

1

1

Department of Cell- and Developmental Biology, Biocentre of the University, Würzburg, Germany;

2

Institute of Entomology, Academy of Sciences,

C

eské Bud

e

jovice, Czech Republic

Abstract

We cloned and sequenced a full length cDNA codingfor [Arg7]-corazonin in the greater wax moth

Galleriamellonella

. The deduced corazonin preprohormoneconsists of a nineteen amino acid signal peptide, theactual eleven amino acid corazonin sequence, followedby a Gly serving for amidation, a Lys-Arg processingsite and an eighty amino acid corazonin precursor-related peptide. The data confirm the phylogenetic con-servation of the actual corazonin sequence. The signalpeptide and the precursor-related peptide exhibit asimilar spacing of a few amino acids as detected in thecorazonin preprohormone of

Drosophila melanogaster

.Northern blots and

in situ

hybridization experimentsrevealed that the

G. mellonella

corazonin gene is tissue-specifically expressed in four pairs of lateral neurose-cretory cells in the brains of penultimate and last instarlarvae, as well as of pupae and adults. No corazoninmRNA was detected in other cells of the nervous sys-tem, fat body, gut, and several other organs.

Keywords: corazonin,

Galleria mellonella

, brain, insectneuropeptides,

in situ

hybridization, immunolocalization.

Introduction

Insect neuropeptides are involved in various processessuch as development and reproduction. They also play akey role in the regulation of homeostasis. The peptide[Arg7]-corazonin (pGln-Thr-Phe-Gln-Tyr-Ser-Arg-Gly-Trp-Thr-Asn-amide) was first identified by Veenstra (1989) as acardioactive peptide in the cockroach

Periplaneta americana

.

Search for homologous compounds led to the detection ofcorazonin in two other cockroaches (Veenstra, 1991), twolepidopterans (Veenstra, 1991; Hua

et al.

, 2000), in a cricket(Hua

et al.

, 2000), and at the genomic level in

Drosophilamelanogaster

(Veenstra, 1994). A [His7]-corazonin homo-logue was found in the locust

Schistocerca americana

(Veenstra, 1991) and later, during research on locust pig-mentation, it was isolated as a dark-inducing peptide in

Locusta migratoria

(Tawfik

et al.

, 1999). These data indi-cate that corazonin is a widespread insect neuropeptide.

The corazonin gene was characterized up-to-date only in

Drosophila melanogaster

(Veenstra, 1994). In this reportwe describe the identification of the cDNA and deducedamino acid sequence of a putative preprohormone contain-ing an [Arg7]-corazonin in

Galleria mellonella

. By means ofNorthern blots,

in situ

hybridization, and immunocytochem-istry we could localize the brain cells that express thecorazonin gene and synthesize the neuropeptide.

Results

Structure of the corazonin-cDNA and corazonin protein

By screening an expression library we identified a positiveclone that contained a corazonin cDNA sequence (GmCrzcDNA) and which was therefore subcloned and sequenced(GenBank accession no. AAF87082). The GmCrz cDNAclone (Fig. 1) comprises 948 bp, including an open readingframe beginning with a methionine codon ATG at bp 162.The presence of three purines upstream from the putativeinitiator ATG fulfils the criteria for a translation start site(Kozak, 1991). The translation stop codon TAA is present atposition 501 and is followed by a 3

′

-untranslated sequencecontaining two polyadenylation signals ATTAAA. We pre-sume that the second of them at bp 780 is used. Thededuced primary structure of the clone yields a polypeptideof 110 amino acids with a deduced molecular mass of12 766 Da. Computer analysis of the sequence revealedthe presence of a typical nineteen amino acid signal pep-tide necessary for transmembrane transport and exportfrom the synthesizing cells. The signal peptide is immedi-ately followed by the sequence Gln-Thr-Phe-Gln-Tyr-Ser-Arg-Gly-Trp-Thr-Asn, which shows 100% identity to all known[Arg7]-corazonins (without its C- and N-terminal modifica-tions). Corazonin is known to be C-terminally amidated.

Received 16 January 2001; accepted after revision 19 March 2001.Correspondence: Prof. Dr Klaus Scheller, Biozentrum der Universität, AmHubland, D-97074 Würzburg, Germany. Tel.: + 49-931-888-4266; fax:+ 49-931-888-4252; e-mail: [email protected]

IMB272.fm Page 341 Friday, July 6, 2001 8:37 AM

342

I. A. Hansen, F. Sehnal, S. R. Meyer and K. Scheller

© 2001 Blackwell Science Ltd,

Insect Molecular Biology

,

10

, 341–346

The glycine residue providing the amide group is presentat the C-terminus of the corazonin peptide and is directlyfollowed by the basic residues Lys-Arg, which generallyfunction as an active peptide precursor cleavage site. Thetotal sequence ends up with an eighty residue C-terminalpeptide, the corazonin precursor related peptide (CPRP),which exhibits no similarities to any peptide published in thedata banks.

Localization of cells expressing the corazonin gene and co-localization with the corazonin neuropeptide

The tissue-specific appearance of the corazonin-mRNAwas investigated by Nothern blot analysis. From varioustissues tested only the brain RNA generated a signalwhen probed with specific antisense corazonin-mRNA(Fig. 2). We detected one single band corresponding to aRNA of about 950 nucleotides. This is in good agreementwith the size of the sequenced cDNA (948 bp).

To detect cells that express corazonin-mRNA wedissected brains and hybridized the whole mounts withthe same antisense RNA probe that was used for theNorthern blots. In all our brain preparations, a limitedgroup of four cells, localized in the lateral region of eachbrain hemisphere, showed a clear hybridization signal(Fig. 3a,b). A hybridization signal could not be detected,neither in any other part of the entire central nervoussystem nor in the

corpora allata-cardiaca

complex (notshown).

We next investigated the localization of the corazoninpeptide, using a polyclonal corazonin antibody (Veenstra,1991). Inspection of the whole mounts revealed four positiveperikarya with extending axons in each brain hemisphere(Fig. 3c).

A combination of the

in situ

hybridization with the immun-ostaining technique proved that the immunoreactive cellswere identical with those transcribing the GmCrz-mRNA

Figure 1. cDNA and deduced amino acid sequence of G. mellonella corazonin – putative translation start (bp 162) and polyadenylation signals (bp 529 and 780) are typed in bold. Horizontal arrows indicate signal peptide (SP), the actual corazonin, and the precursor-related peptide (CPRP) in the deduced translation product. The corazonin sequence is double-underlined, the amidation site (Gly) and adjacent cleavage site (Lys-Arg) are single-underlined.

Figure 2. Northern blot showing the tissue-specific occurrence of corazonin mRNA in G. mellonella brains. Total RNA samples (20 µg each) from various tissues were probed with a DIG-labelled corazonin-antisense mRNA. RNA was derived from: (1) integument, (2) Malphighian tubules, (3) gut, (4) fat body, (5) salivary gland, (6) brain. All tissues were dissected from the last instar larvae. The lower panel shows the 18 s-RNA stained with ethidiumbromide (loading control).


Galleria mellonella

[Arg7]-corazonin

343



,

10

, 341–346

(Fig. 3d). Although, the background fluorescence in thesewhole mounts was higher than in those that were justimmunostained, the four cells could be unambiguouslyidentified.

Together, our data demonstrate that the corazonin geneis expressed exclusively in two pairs of neurones in thelateral region of each brain hemisphere and that these cellsare the only source of the corazonin preprohormone pep-tide in

G. mellonella

.

Organization of the corazonin gene

To gain some insights into the structure of the corazoningene and to assess the number of gene copies, genomicDNA of single animals was digested with three different restric-tion enzymes and analysed by Southern hybridizationusing the complete GmCrz-antisense-mRNA as a probe

(Fig. 4). Digestion with

Sty

I generated a single band whichwas in the same position in all animals analysed. Similarresults were obtained with

Eco

RI, however digestion with

Hin

dIII evinced a restriction site polymorphism. In threeof four analysed individuals we detected two hybridizingbands (Fig. 4, lanes h 1–3) whereas in the fourth animalthe lower one was missing (lane h 4). From this result wemight conclude that the corazonin gene is present as asingle copy gene in

Galleria mellonella

and probably con-tains an intron with a

Hin

dIII restriction site polymorphism.

Discussion

A cDNA encoding [Arg7]-corazonin preprohormone wasidentified in an expression cDNA library prepared from thebrains of

G. mellonella

larvae (Fig. 1). The corresponding

Figure 3. G. mellonella brain neurones producing corazonin. In situ hybridization demonstrating four neurones containing corazonin-mRNA in the region of the lateral neurosecretory cells in each brain hemisphere and absence of additional cells in the nerve cord (a,b). Immunolocalization of the corazonin peptide containing cells in the lateral protocerebrum (c). Combined in situ hybridization and immunohistochemistry proving the identity of the corazonin-mRNA expressing cells with the corazonin peptide containing cells (d). All pictures show whole mounts of brains dissected from the last instar larvae. Bars indicate 50 µm.

Figure 4. Southern blot, demonstrating a slight polymorphism of the corazonin gene in different individuals. DNA from single last instar larvae (four parallel experiments) was digested with StyI (s 1–4), HindIII (h 1–4), and EcoRI (e 1–4), electrophoresed, transferred to membranes, and hybridized with a DIG-labelled GmCrz-antisense RNA probe.


344

I. A. Hansen, F. Sehnal, S. R. Meyer and K. Scheller



,

10

, 341–346

corazonin

gene appears to be present in the genome ina single copy and probably includes an intron exhibitingheterogeneity in the

Hin

dIII cleavage site (Fig. 4). ThecDNA is longer in the 3

′

and 5

′

non-translated regions,as well as in the coding region, than the correspondingand presumably transcribed part of the

corazonin

genein

D. melanogaster

, the only other species from which thegene is known (Veenstra, 1994).

Corazonin

transcripts ofboth species seem to contain two polyadenylation signalsATTAAA (only one was mentioned by Veenstra, 1994).

The deduced corazonin preprohormones of

G. mellon-ella

and

D. melanogaster

are structurally similar. Each con-sists of a putative signal peptide of nineteen amino acids,the actual [Arg7]-corazonin (eleven amino acids), a Gly pro-viding amidation, a Lys-Arg cleavage site, and a corazoninprecursor-related peptide (CPRP) of eighty amino acids in

G. mellonella

and thirty-nine in

D. melanogaster.

The simil-arities further include conserved positions of a few aminoacids in both the signal peptide and the CPRB (Fig. 5).Data on more species are needed to appreciate the signi-ficance of these conservations.

D. melanogaster

CPRBcontains an Arg-Arg dipeptide that is often used as aproteolytic processing site of preprohormones, whereas asingle Arg, which is conserved in

G. mellonella

CPRB, israrely used (Veenstra, 2000). We regard CPRB cleavageunlikely in either species because the presence of Leu in+1 position and in

G. mellonella

of Lys in –5 position,does not comply to the rules characterizing the pro-cessing sites (Rholam

et al.

, 1995).The results of gene analyses (Veenstra, 1994; present

data) and peptide isolations (Veenstra, 1989, 1991; Predel

et al.

, 1994; Hua

et al.

, 2000) show that [Arg7]-corazonin isconserved in insects ranging from primitive Polyneopterasuch as cockroaches to the most advanced Holometabolasuch as moths and flies. By contrast, the sequences ofsignal peptide and CPRP have obviously rapidly diversifiedin insect phylogeny, as seen from the comparison of

G.mellonella

and

D. melanogaster

(Fig. 5). A similar situationis known for genes of the AKH/RPCH family and it hasbeen suggested that the sole function of the precursor-related peptides is to give sufficient length to the prepro-hormone for proper processing (Linck

et al.

, 1993).We demonstrate that corazonin is produced exclusively

in four pairs of brain neurones (Figs 2 and 3). Previousimmunohistochemical studies by Veenstra & Davis (1993)and Cantera

et al.

(1994) detected a higher number ofcorazonin-positive neurones in the central nervous sys-tems of the cockroach

Periplaneta americana

and the

blowfly

Phormia terraenovae

, respectively. The most prom-inent immunoreactive neurones, however, were found in thelateral brain regions, similar to

G. mellonella

. Since no

in situ

hybridizations were done in

P. americana

and

P. terraenovae

,it cannot be excluded that the weak reaction in some othercells was unspecific in these insects.

From our whole mount

in situ

hybridization experiments(not shown) it is clear that the corazonin gene is expressedin the penultimate and last instar larvae, pupae, and adultsof

G. mellonella

. In all stages tested the four pairs of neu-rones were exclusively stained (not shown). This mightindicate that corazonin performs a function common to allthese developmental stages. However, it cannot be excludedthat the peptide could have different functions in differentstages even in the same species. In comparison with otherneuropeptides of similar size, such as the adipokinetichormones (Gäde

et al.

, 1997), the structure of corazonin ishighly conserved during evolution. Both corazonin effectsthat have been demonstrated so far seem to be species-specific. The acceleration of heartbeat was shown in three,but could not be detected in six other cockroach species(Veenstra, 1989; Predel

et al.

, 1994), and only high cora-zonin doses elicited a modest response in the heart of

Manduca sexta

(mentioned in Cantera

et al.

, 1994). In testswith the

in situ

heart preparations of

G. mellonella

prepupae,we also obtained only a negligible and transient stimulationof heartbeat frequency with [Arg7]-corazonin concentra-tions higher than 10

–6

M (our unpublished data). The othercorazonin effect, i.e. the induction of dark body pigmenta-tion, was demonstrated in locusts but could not be elicitedin the silkworm and a cricket (Hua

et al.

, 2000). We alsofailed to affect body pigmentation in both a normal and alight-pigmented strain of

G. mellonella

by injecting caterpil-lars with 1 nmol [Arg7]-corazonin per 100 mg body weight(our unpublished data). Our attempts to alter the productionof juvenile hormones and ecdysteroids by exposing explanted

corpora allata

and prothoracic glands, respectively, from10

–9

M to 10

–5

M [Arg7]-corazonin in a culture also failed.We have to conclude that the function of corazonin remainselusive.

Experimental procedures

Insects and tissues

Two strains of the waxmoth,

Galleria mellonella

(Lepidoptera:Pyralidae), were used, a standard laboratory strain that has beenmaintained in our laboratories for decades, and a stock of lightlycoloured larvae that is used by commercial producers of bait for

Figure 5. Comparison of the corazonin preprohormones of Galleria mellonella (Gm) and Drosophila melanogaster (Dm, sequence taken from Veenstra, 1994).


Galleria mellonella

[Arg7]-corazonin

345



,

10

, 341–346

fishing. The cultures were kept at 32

°

C and larvae were fed asemi-artificial diet (Sehnal, 1966). For some experiments, larvaewithin 12 h of their last instar were chilled on melting ice for 2 h toactivate their

corpora allata

(Bogu

s

& Scheller, 1991). They arereferred to as ‘chilled’ larvae and their age is given in hours afterchilling. Organs were dissected in insect saline (130 m

M

NaCl,5 m

M

KCl, 1 m

M

MgCl

2

, 0.5 m

M

CaCl

2

) from animals anaesthetizedfor 10 min in water.

cDNA analysis

Total RNA was extracted from 100 brains of chilled larvae (2–20 hafter chilling) using the TriFast Kit (Peqlab, Erlangen, Germany).cDNA was generated and amplified with the SMART III cDNALibrary Construction Kit (Clontech Laboratories, Palo Alto,California). The library was cloned in a pTriplEx2 vector andscreened with hybridoma supernatants AF 43 and AF 143 that hadbeen prepared in search for an allatotropic factor (Bogu

s

&Scheller, 1996). The screening was performed according tostandard protocols (Sambrook

et al.

, 1989). Six positive clones wereused for

in vivo

excision. The plasmids were grown in XL1-bluecells (Strategene, La Jolla, California) and the inserts sequencedon a Perkin-Elmer 310 sequencer. One clone proved to contain acorazonin sequence (GmCrz cDNA). It was analysed in detail andused as a template for the preparation of a digoxygenine (DIG)-labelled antisense-RNA probe.

Genome analysis by Southern blotting

Genomic DNA was isolated from single larvae, each homogenizedin 500

µ

l ice cold 10 m

M

Tris, pH 7, containing 60 m

M

NaCl, 10 m

M

EDTA, 150 m

M

spermine, 15 m

M

spermidine, and 5% sucrose. Thehomogenate was mixed at room temperature with an equal volumeof 0.2

M

Tris, pH 9, 30 m

M

EDTA and 2% SDS. Ten microlitresof proteinase K (20 mg/ml) were added, the sample incubatedfor 4 h at 50

°

C, and the DNA subsequently extracted twice withphenol/chloroform/isoamylalcohol (25 : 24 : 1) and three timeswith chloroform. The final aqueous phase was precipitatedwith ethanol, washed three times with 70% ethanol, air driedand resuspended in water. One animal yielded 200–400

µ

ggenomic DNA. Aliquots of 30

µ

g DNA were digested withselected restriction enzymes, electrophoresed on a 0.8% agarosegel, and transferred to a Nylon membrane by capillary transfer(Sambrook

et al.

, 1989). As a hybridization probe we used aDIG-labelled antisense RNA, synthesized from linearizedGmCrz-cDNA as a template using the DIG-RNA-labelling Kit(T7) from Roche Molecular Biochemicals (Mannheim, Germany).The DIG-labelled RNA hybridization probe was precipitated byaddition of two volumes of ethanol containing 0.4

M

LiCl, andresuspended in water. This probe was also used for Northernblotting and

in situ

hybridization.

In situ

hybridization and immunostaining

Brains or entire central nervous systems were dissected in insectsaline and promptly fixed in MEMFA (0.1

M

MOPS, 2 m

M

EGTA,1 m

M MgSO4, 3.7% formaldehyde) for 2 h at room temperature.The tissues were dehydrated with methanol and stored at –20 °Cuntil used for whole mount in situ hybridization performed asdescribed by Harland (1991). Some preparations were clearedin benzylalcohol /benzylbenzoat (2 : 1) for 10 min before beingmounted for microscopic analysis.

Organs selected for immunostaining were fixed overnight at 4 °Cwith 4% paraformaldehyde in PBS. After multiple washings with0.5× PAT (PBS, 1% albumin, 0.5% Triton X100), the samples weretreated with 3% goat serum in 0.5× PAT for 2 h at room temperature,and then washed three times with 0.5× PAT. Subsequent incubationwith corazonin antiserum (diluted 1 : 2000) lasted 2 days at 4 °C(Veenstra, 1991). After extensive washing with 0.5× PAT, the tissueswere incubated overnight at 4 °C with a Cy2 (cyanine 2-OSubisfunctional)-conjugated affinity purified goat anti-rabbit IgG (1 : 50;Rockland, Gilbertsville, USA). After thoroughly washing, the sam-ples were ready for observations. The specificity of the corazoninimmunoreaction was verified by omitting the primary antibody.

Some preparations processed for in situ hybridization were re-hydrated at the end of the procedure. The tissues were then takenfor immunostaining, performed as described above. All observa-tions were done under a Leica fluorescent microscope and thepictures were taken with a Pixera CCD camera.

Acknowledgements

We are indebted to Dr Jan Veenstra for the gift of corazoninantibodies, and to Mr P. L. Gualandi (Bagnacavallo, Italy) forsupplying us with the lightly coloured G. mellonella strain.We thank Anneliese Striewe-Conz for competent technicalassistance. This research was supported by a German-Czech collaborative grant CZE-00–005 provided byDeutsches Zentrum für Luft- und Raumfahrt, Bonn, and theCzech Ministry of Education, Youth, and Sports.

References

Bogus, M. and Scheller, K. (1991) Activation of juvenile hormonesynthesis in vitro by larval brains of Galleria mellonella. Zool JbPhysiol 95: 197–208.

Bogus, M. and Scheller, K. (1996) Allatotropin released by thebrain controls larval molting in Galleria mellonella by affectingjuvenile hormone synthesis. Int J Dev Biol 40: 205–210.

Cantera, R., Veenstra, J.A. and Nässel, D.R. (1994) Postembryonicdevelopment of corazonin-containing neurons and neurose-cretory cells in the blowfly, Phormia terranovae. J Comp Neurol350: 559–572.

Gäde, G., Hoffmann, K.H. and Spring, J.H. (1997) Hormonalregulation in insects: facts, gaps, and future directions. PhysiolRev 77: 963–1032.

Harland, R.M. (1991) In situ hybridization and improved whole-mount methods for Xenopus embryos. Meth Cell Biol 36:685–695.

Hua, Y.-J., Ishibashi, J., Saito, H., Tawfik, A.I., Sakakibara, M.,Tanaka, Y., Derua, R., Waelkens, E., De Baggerman, G., Loof, A.,Schoofs, L. and Tanaka, S. (2000) Identification of [Arg7] cora-zonin in the silkworm, Bombyx mori and the cricket, Gryllusbimaculatus, as a factor inducing dark color in an albino strainof the locust, Locusta migratoria. J Insect Physiol 46: 853–859.

Kozak, M. (1991) An analysis of vertebrate mRNA sequences: inti-mation of translational control. J Cell Biol 115: 887–903.

Linck, B., Klein, J.M., Mangerich, S., Keller, R. and Weidemann, W.M.(1993) Molecular cloning of crustacean red pigment concen-trating hormone precursor. Biochem Biophys Res Commun195: 807–813.


346 I. A. Hansen, F. Sehnal, S. R. Meyer and K. Scheller

© 2001 Blackwell Science Ltd, Insect Molecular Biology, 10, 341–346

Predel, R., Agricola, H., Linde, D., Wollweber, L., Veenstra, J.A.and Penzlin, H. (1994) The insect neuropeptide corazonin:physiological and immunocytochemical studies in Blattariae.Zoology 98: 35–49.

Rholam, M., Brakch, N., Germain, D., Thomas, D.Y., Fahy, C.,Boussetta, H., Boileau, G. and Cohen, P. (1995) Role of aminoacid sequences flanking dibasic cleavage sites in precursorproteolytic processing. The importance of the first residue C-terminal of the cleavage site. Eur J Biochem 227: 707–714.

Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning,a Laboratory Manual. Cold Spring Laboratory Press, New York.

Sehnal, F. (1966) Kritisches Studium der Bionomie undBiometrik der in verschiedenen Lebensbedingungen gezüchtetenWachsmotte, Galleria mellonella. Z wiss Zool 174: 53–82.

Tawfik, A.I., De Tanaka, S., Loof, A., Schoofs, L., Baggerman, G.,Waelkens, E., Derua, R., Milner, Y., Yerushalmi, Y. and Pener,M.P. (1999) Identification of the gregarization-associated dark-

pigmentotropin in locusts through an albino mutant. Proc NatlAcad Sci USA 96: 7083–7087.

Veenstra, J.A. (1989) Isolation and structure of corazonin, acardioactive peptide from the American cockroach. FEBSLett 250: 231–234.

Veenstra, J.A. (1991) Presence of corazonin in three insect spe-cies, and isolation and identification of [His7] corazonin fromSchistocerca americana. Peptides 12: 1285–1289.

Veenstra, J.A. (1994) Isolation and structure of the Droso-phila corazonin gene. Biochem Biophys Res Comm 204:292–296.

Veenstra, J.A. (2000) Mono- and dibasic proteolytic cleavage sitesin insect neuroendocrine peptide precursors. Arch Insect Bio-chem Physiol 43: 49–63.

Veenstra, J.A. and Davis, N.T. (1993) Localization of corazonin inthe nervous system of the cockroach Periplaneta americana.Cell Tissue Res 274: 57–64.


Documents

Corazonin gene expression in the waxmoth Galleria mellonella