Identiﬁcation and developmental expression of mRNAs ... · phoresis and tryptic digests of cockroach (Periplaneta americana) bursicon containing extracts (Honegger et al., 2002),

Available online at www.sciencedirect.com

www.elsevier.com/locate/ygcen

General and Comparative Endocrinology 156 (2008) 113–125

Identification and developmental expression of mRNAsencoding putative insect cuticle hardening hormone, bursicon

in the green shore crab Carcinus maenas

David C. Wilcockson, Simon G. Webster *

School of Biological Sciences, Bangor University, Bangor, Gwynedd LL57 2UW, UK

Received 30 October 2007; revised 29 November 2007; accepted 10 December 2007Available online 25 January 2008

Abstract

Bursicon is the ultimate hormone in insect ecdysis, which is involved in cuticle hardening. Here we show that mRNAs encoding theheterodimeric cystine knot protein bursicon (Burs a, b), are present in crustaceans, suggesting ubiquity of this hormone in arthropods.We firstly report the cloning, sequencing of mRNAs encoding subunits from the water flea, Daphnia arenata and the CNS of the crab,Carcinus maenas, in comparison with insect bursicon subunits. Expression patterns of a and b burs mRNAs were examined by in-situhybridisation (ISH) and quantitative RT-PCR. In the thoracic ganglion, burs a and b mRNAs were completely colocalised in neuronesexpressing crustacean cardioactive peptide (CCAP). However, in the brain and eyestalk, bursicon transcripts were never observed,despite a complex expression pattern of CCAP interneurones. Patterns of expression of burs a and b mRNAs were constitutive duringthe moult cycle of adult crabs, in stark contrast to the situation in insects. Whilst copy numbers of burs b transcripts closely matchedthose of CCAP, those of burs a mRNA were around 3-fold higher than burs b. This pattern was apparent during embryogenesis, wherebursicon transcripts were first observed at around 50% development—the same time as first expression of CCAP mRNA. Transcriptratios (burs a: b) increased during development. Our studies have shown, for the first time, that bursicon mRNAs are expressed in iden-tified neurones in the nervous system of crustaceans. These findings will now promote further investigation into the functions of bursiconduring the moult cycle and development of crustaceans.� 2007 Elsevier Inc. All rights reserved.

Keywords: Arthropods; Bursicon; Crustaceans; Developmental expression; Ecdysis; In-situ hybridisation; Neurohormones; Quantitative RT-PCR

1 Abbreviations used: AK, arginine kinase; AMV-RT, avian myeloblas-tosis virus-reverse transcriptase; CCAP, crustacean cardioactive peptide;CNS, central nervous system; CG, cerebral ganglion; DIG-11-UTP,

1. Introduction

All arthropods periodically replace their exoskeletons inorder to grow and metamorphose; it has long been knownthat this moulting is controlled by several hormones-current research in insects has suggested considerable andunforeseen complexity. At the very least, six or more differ-ent hormones are involved: notwithstanding the wellknown roles of ecdysteroids in directing the synthesis ofthe new cuticle (review by Riddiford, 1989) and the role

0016-6480/$ - see front matter � 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.ygcen.2007.12.003

* Corresponding author. Fax: +44 01248 371644.E-mail address: [email protected] (S.G. Webster).

of prothoracicotropic hormone (PTTH)1 in stimulatingecdysteroid secretion by the prothoracic glands (reviewsby Gilbert et al., 1996, 2002), a series of neuropeptide hor-mones are involved in the insect ecdysis programme. Theseact in a precisely timed series during the penultimate stagesof premoult, which are characterised by a precipitous

digoxygenin-11-uridine-5 0 triphosphate; ES, eyestalk; EST, expressedsequence tag; ISH, in-situ hybridisation; RACE, rapid amplification ofcDNA ends; RT-PCR, reverse transcriptase PCR; TG, thoracic ganglion;UTR, untranslated region.

mailto:[email protected]

114 D.C. Wilcockson, S.G. Webster / General and Comparative Endocrinology 156 (2008) 113–125

decline in ecdysteroid levels (reviews by Ewer and Rey-nolds, 2002; Mesce and Fahrbach, 2002). Until recently,most studies used larvae of the tobacco hornworm, Mand-uca sexta, as a model. Several hormones, includingpre-ecdysis triggering hormone (PETH), ecdysis triggeringhormone (ETH), eclosion hormone (EH), crustacean cardi-oactive peptide (CCAP), and finally the ultimate hormoneinvolved in this cascade- bursicon, are now known to beinvolved in the ecdysis programme (review by Truman,2005). Not unexpectedly, the rich genetic resources avail-able in Drosophila by way of mutants, targeted ablationsvia knock outs and transgenic technologies have nowadded considerably to the lepidopteran models, and suchtechniques have expedited tremendous progress in pursuitof previously intractable questions; these have recentlybeen elegantly reviewed (Ewer, 2005).

As alluded to above, the ultimate hormone in the ecdy-sis cascade in insects has long been known to be bursicon,which was identified as a key hormone involved in tan-ning and melanisation of the insect cuticle more than 40years ago (Fraenkel and Hsiao, 1962; Cottrell, 1962;Fraenkel and Hsiao, 1963, 1965; Fraenkel et al., 1966).At that time, the classical bioassay for this hormoneinvolved injection of newly-eclosed, neck-ligated flies (Sar-cophaga bullata) with extracts of the CNS, and subsequentrecording of tanning (melanisation and sclerotisation) ofthe cuticle (review by Seligman, 1980). Despite numerousattempts at purification and characterisation of bursicon(Kaltenhauser et al., 1995; Kostron et al., 1995, 1999;Honegger et al., 2002), the complete determination ofthe structure of bursicon remained elusive. However,using sequence information obtained from 2 D-gel electro-phoresis and tryptic digests of cockroach (Periplanetaamericana) bursicon containing extracts (Honeggeret al., 2002), and in silico data mining, bursicon hasrecently been identified in Drosophila, as a heterodimericcysteine knot protein comprising the product ofCG13419 (burs or burs a) and CG15284 (pburs, or partnerof burs, burs b) (Luo et al., 2005; Mendive et al., 2005).Subsequently, heterodimeric bursicon molecules have beenidentified from a number of model insects, from concep-tual translation of cDNA and gDNA sequences (VanLoy et al., 2007). The cognate receptor of bursicon,DLGR2 (a GPCR) has also been identified in Drosophila(Baker and Truman, 2002; Luo et al., 2005; Mendiveet al., 2005), for which loss of function mutations inDLGR2 display a phenotype (rickets rk�/�) which is sim-ilar to that induced by bursicon deficiency resulting fromneck ligation.

Bursicon would perhaps be firstly considered as aprototype insect hormone, but since it has been knownfor some years that it is present in crustacean cardioactivepeptide (CCAP) expressing neurones in the insect CNS(Kostron et al., 1996; Honegger et al., 2002; Davis et al.,2003; Dewey et al., 2004) and because these neurones havea surprisingly conserved architecture in all arthropods(review by Dircksen, 1998), it seemed possible that crusta-

ceans might also possess bursicon-like molecules. Further-more, in view of a report that extracts of abdominalganglia from lobsters (Homarus americanus) showed burs-icon activity in the Sarcophaga bioassay (Kostron et al.,1995), we reasoned an attractive hypothesis would be thatbursicon-like molecules might have a wider distributionthan previously suspected in the arthropods. This hypoth-esis also had tentative grounding from our previous obser-vations showing that CCAP release in crustaceans wasassociated with stereotyped ecdysis behaviour in crayfishand crabs (Phlippen et al., 2000). However, our most ger-mane observation was that an early interrogation of theDaphnia arenata EST database revealed a single EST(Wfes0000858) with a remarkable similarity to Drosophilaburs a.

Here we report the cloning and sequencing of mRNAsencoding bursicon subunits (burs a, burs b) in the modelcrustacean Daphnia, followed by cloning, sequencing andexpression profiles of both subunit mRNAs in our crabmodel, Carcinus maenas during the moult cycle of adultsand during embryonic development. We also report thedistribution of these transcripts in the CNS of Carcinusby whole mount in-situ hybridization. We show that inthe thoracic ganglion expression of bursicon mRNAsoccurs exclusively in CCAP neurones, and that expressionof burs a, burs b mRNA in the CNS of Carcinus is consti-tutive during the moult cycle, in contrast to the situation ininsects, perhaps suggesting new roles for this hormone incrustaceans.

2. Materials and methods

2.1. Animals, tissue preparation and RNA extraction

Adult C. maenas were collected using baited traps (Menai Strait,Anglesey, UK) and maintained in a recirculating seawater aquarium sys-tem under ambient conditions with ad libitum feeding. Specimens of D.arenata (Log 50 clone) were a generous gift from Prof. J. Colbourne, Uni-versity of Indiana, Bloomington, IA, USA. Daphnia were cultured in glassjars containing filtered pond water and a mixed phytoplankton culture indirect sunlight.

Nervous systems (thoracic ganglia, brain, eyestalk) were dissected fromice-anaesthetised, moult staged Carcinus, and immediately placed inRNAlater (Ambion, Texas) (4 �C overnight, followed by storage at�80 �C), or for tissues used in for in-situ hybridisation (ISH), fixed imme-diately in 4% paraformaldehyde (PFA) in phosphate buffered saline (PBS).After overnight fixation at room temperature (RT), tissues were brieflywashed in PBS then dehydrated in a graded methanol/PBS series. Tissueswere stored in 100% methanol at RT for a maximum of 3 days before ISH.Embryos were removed in batches of 100 from ovigerous female Carcinus,developmentally staged (Chung and Webster, 2004), and stored in RNA-later. For Daphnia samples, quantities (ca 200) of adults were staged(according to ovarian development/presence of embryos or ephippia inthe brood chamber), sampled by individually picking out with a mountedinsect pin and placed directly into liquid nitrogen.

Total RNA was prepared using TRIzol (Invitrogen, Carlsbad, CA,USA), followed by treatment with 2 units of DNase 1 (37 �C, 1 h), fol-lowed by clean up with DNA-free (Ambion) and quantification (ND-1000, NanoDrop Technologies, Wilmington DE, USA). For embryos,and RNA from thoracic ganglia subsequently used in RACE procedures,mRNA was isolated using Dynabeads (Dynal, Oslo, Norway), and storedin 10 mM Tris at �80 �C.

D.C. Wilcockson, S.G. Webster / General and Comparative Endocrinology 156 (2008) 113–125 115

2.2. cDNA synthesis and rapid amplification of cDNA ends

(RACE)

For 3’ RACE, mRNA samples of 1–5 lg total RNA from thoracic gan-glia of premoult (D2–D4) crabs or whole Daphnia were reverse transcribedusing Superscript III RT (Invitrogen) (50 �C, 50 min) using the Gene Racer3’oligo(dT) adapter primer. For 5 0 RACE, mRNA was dephosphorylated,decapped, ligated to a 5 0 RACE RNA oligo (Invitrogen), and reverse tran-scribed using Superscript III using random primers according to the manu-facturers instructions. Following reverse transcriptions, samples weretreated (37 �C, 15 min) with 2 units RNase H. Samples (Carcinus) used forquantitative RT-PCR: thoracic ganglia RNA (200 ng), embryo mRNA(100 embryos), and quantified cRNA standards (1010–103 copies/ll), weresimultaneously reverse transcribed in large batches using total reaction vol-umes of 10 ll, with random primers and AMV-RT (Promega, Southamp-ton, UK). Samples were reverse transcribed for 10 min, 25 �C, 50 min at37 �C, and reactions terminated by denaturation at 95 �C for 5 min.

2.3. Primers

See Table 1 for a complete list of primers used and their identifyingabbreviations, which are referred to below.

2.4. 3 0 and 5 0 RACE PCR of Daphnia arenata cDNA encoding burs

a and burs b

2.4.1. Burs a

3 0 RACE was performed on cDNA prepared from mRNA extractedfrom batches of around 200 D. arenata which contained mature embryosand/or ephippia (as detailed above). The primers used were designed fromsequence information of a putative burs a transcript (WFes0000858) iden-tified in the wFleaBase EST database (http://wfleabase.org). PCR reagentswere: 22.5 ll Megamix Blue (Helena Biosciences, Sunderland, UK),1.25 ll (10 lM) DaphBursF1 primer, 1.25 ll (10 lM) 3 0 GeneRacer pri-mer, 1 ll 3 0 RACE cDNA template. PCR conditions were: 94 �C 4 min,35 cycles of 94 �C 30 s, 55 �C 30 s, 72 �C 45 s, final extension at 72 �Cfor 7 min. The second nested PCR used primer pairs DaphBursF1N and3 0 nested GeneRacer primer, with 0.5 ll of the first PCR reaction as tem-plate, using identical PCR conditions. 5 0 RACE PCR was performed oncDNA prepared from reverse transcription of ligated mRNA (as detailedabove). PCR reagents were 22.5 ll Megamix Blue (Helena Biosciences,Sunderland, UK), 1.25 ll (10 lM) DaphBursR1primer, 1.25 ll (10 lM)5 0 GeneRacer primer, 1 ll 5 0 RACE cDNA template. PCR conditionswere the same as for the 3 0 RACE PCR. The second nested PCR used pri-mer pairs DaphBursR2 and 5

0 nested GeneRacer primer, with 0.5 ll of thefirst PCR reaction as template, using identical PCR conditions.

2.4.2. Burs bUsing cDNA prepared for 3 0 and 5 0 RACE described above, PCR was

performed using primers designed from a putative transcript of Daphniapulex burs b (NCBI_GNO_662053) available from the D. pulex genomeassembly V1.1. Nested 3 0 and 5 0 RACE PCR strategies conditions andreagents were essentially as before, except that a confirmatory PCR involv-ing primer pairs D.p. F and D.p. R was included, to produce sequence over-lap for 3 0 RACE (using primer pairs D.p. 3 0 RACE F1 and 3

0 Gene Racerprimer, and subsequently D.p. 3 0 RACE F2 and 3

0 Gene Racer nested pri-mer). For 5 0RACE primer pairs D.p. 5 0RACE R1 and 5

0GeneRacer primer,and D.p. 5 0 RACE R2 and 5

0 nested GeneRacer primers were used in nestedPCRs as described above. PCR products were electrophoresed on agarosegels, and bands of (expected) sizes were excised and extracted using a Nucle-otrap gel purification kit (Machery Nagel, Düren, Germany).

2.5. Degenerate PCR of Carcinus cDNA encoding burs a

Using available amino acid sequence data available for the Burs a sub-unit for insects and Daphnia, a set of degenerate hybrid primers weredesigned using CODEHOP (Rose et al., 1998). These primers were designed

with bias for Callinectes sapidus global codon usage. PCR reagents were:12.5 ll AmpliTaq Gold Master mix (Roche, Branchburg, NJ, USA), 9 llwater, 1.25ul (100 lM) forward and reverse primers (4F GCVPKIP, 11 RMCRPCTSIE or 8 F ERSCMCCQE, 9 R CMCCRPCTSI) as shown inTable 1, 1 ll cDNA template. Touchdown PCR conditions were: 1 cycleof 94 �C 9 min, 5 cycles of 94 �C 30 s, 63 �C 30 s, 5 cycles of 94 �C 30 s,60 �C 30 s, 25 cycles of 94 �C 30 s, 57 �C 30 s, 72 �C 45 s, final extension at72 �C for 10 min.

2.6. 3 0 and 5 0 RACE PCR of Carcinus cDNA encoding burs a

Using sequence information obtained from degenerate PCR, (asalluded to earlier) gene specific primers (GSP) were designed for 3 0 and5 0 RACE PCR. For 3 0 RACE, nested PCRs were performed as follows:12.5 ll AmpliTaq Gold, 9 ll water, 1.25 ll (10 lM) F1GSP (See Table1), 1.25 ll (10 lM) 3 0 GeneRacer primer, 1 ll 3 0 RACE cDNA template.Touchdown PCR conditions were the same as shown above. The secondnested PCR used primer pairs F2GSP and 3 0 nested GeneRacer primer,and 1 ll of the first PCR reaction mixture as template. PCR conditionswere identical to the first PCR. PCR products were electrophoresed andbands extracted as described earlier.

For 5 0 RACE, nested PCRs were performed as follows: 12.5 ll AmpliT-aq Gold, 9 ll water, 1.25 ll (10 lM) Carc5bursA R1 primer, 1.25 ll (10 lM)5 0 GeneRacer primer, 1 ll 5 0 RACE cDNA template. PCR conditions were:94 �C 9 min, 30 cycles of 94 �C 30 s, 58 �C 30 s, 72 �C 45 s, final extension at72 �C for 7 min. The second nested PCR used primer pairs Carc5bursAR2and 5 0 nested GeneRacer primer, and 1 ll of the first PCR reaction mixtureas template. PCR conditions were identical to the first PCR. Products wereelectrophoresed and bands extracted as described earlier.

2.7. 3 0 and 5 0 RACE PCR of Carcinus cDNA encoding burs b

Gene specific primers were designed from an EST (CMC 02618; Parti-Gene ARTHROPODA database (http://www.nematodes.org/Neglect-edGenomes/ARTHROPODA/index.shtml) encoding a putative burs boriginating from a ‘‘mixed tissue’’ Carcinus cDNA library. For 3 0 RACE,nested PCRs were performed as follows: 22.5 ll Megamix Blue (HelenaBiosciences, Sunderland, UK), 1.25 ll (10 lM) F1 Carcburs B primer,1.25 ll (10 lM) 3 0 GeneRacer primer, 1 ll 3 0 RACE cDNA template.PCR conditions were: 94 �C 4 min, 35 cycles of 94 �C 30 s, 58 �C 30 s,72 �C 1 min, final extension at 72 �C for 7 min. The second nested PCRused primer pairs F1N CarcBurs B and 3

0 nested GeneRacer primer, and0.5 ll of the first PCR reaction as template, using identical PCR condi-tions. The faint band of expected size obtained after electrophoresis wasexcised, gel purified, as detailed earlier and 1 ll of this product was ream-plified using the nested primers and PCR conditions detailed above.

For 5 0 RACE nested PCRs were performed as follows: 22.5 ll Mega-mix Blue, 1.25 ll (10 lM) Burs B R1 primer, 1.25 ll (10 lM) 5 0 GeneR-acer primer, 1 ll 5 0 RACE cDNA template. PCR conditions were: 94 �C4 min, 30 cycles 94 �C 30 s, 55 �C 1 min, 72 �C 45 s, final extension at72 �C 7 min. The second nested PCR used primer pairs BursB R2 and5 0nested GeneRacer primer, and 1 ll of the first PCR reaction as template,using identical PCR conditions, except that 35 cycles were performed. ThePCR product was purified on a Montage PCR purification module (Mil-lipore, Bedford, MA, USA).

2.8. Cloning and sequencing of PCR products

Purified PCR products were ligated into a PCR 4-TOPO vector andtransformed (TOP-10F0, Invitrogen) according to the manufacturer’sinstructions. Plasmid DNA from positive clones containing inserts of cor-rect sizes were purified and sequenced.

2.9. In situ hybridisation: Carcinus burs a, burs b, CCAP

DNA templates for making digoxygenin (DIG)-labelled cRNA probesfor in-situ hybridisation were prepared by in vitro ligation of phage pro-moter T7 adaptors to PCR products using a Lig’n Scribe (Ambion) kit.

http://wfleabase.orghttp://www.nematodes.org/NeglectedGenomes/ARTHROPODA/index.shtmlhttp://www.nematodes.org/NeglectedGenomes/ARTHROPODA/index.shtml

Table 1Primers used for bursicon sequence identification, production of cRNA probes and quantitative PCR

Primer name Sequence

Burs a Carcinus Degenerate4F GCVPKPIP CGGCTGCGTGCCCAARSCNATHCC11R MCRPCTSIE TCGATGGAGGTCCAGGGNCKRCACATRC8F ERSCMCCQE GAGCGCTCCTGCATGTGYYGYCARGA9R CMCRPCTSI TGGAGGTGCAGGGGCKRCACATRCANT

Burs a Carcinus 3’ RACEF1GSP CGCCTGCCAGGGTCGATGTACTTCATACF2GSP CAGGGTCGATGTACTTCATACGTGCAGG

Burs a Carcinus 5’ RACECarc5bursA R1 CTGGGCCAGCACAGTGCCTTCCTCCACCarc5bursA R2 CCAGCACAGTGCCTTCCTCCACATCAG

Burs b Carcinus 3’ RACEF1 CarcbursB GCCAGATCTTATGGTGGGAATGTGAGACF1N CarcbursB GGTGTGGAATGTGAGACACTGCC

Burs b Carcinus 5’ RACEBursB R1 CACAGACGGCTGCACTTTGGATACACBursB R2 GGATACACACGCGCCCTCACAT

Burs a Daphnia 3’ RACEDaphBursF1 TTCTACCGGTCATGATGTTCCTACDaphBursF1N CCTACTTGTGGGATTGGTCTC

Burs a Daphnia 5’ RACEDaphBursR1 CCCGTCGCAGTTTTGGTTCDaphBursR2 CGCAGTTTTGGTTCTCCAGGTGCG

Burs b DaphniaD.p. F TGGGTGCCGGTTCGTCTCGD.p. R TTTGCAATCGGCGGGTTCG

Burs b Daphnia 3’ RACED.p. 3’ RACE F1 GGCGATATTGGTGTCGCTAAATGTGAAGD.p. 3’ RACE F2 GGTGTCGCTAAATGTGAAGGATCGTG

Burs b Daphnia 5’ RACED.p. 5’ RACE R1 GGTGATGTGGATTGTCGATGGAAGAGD.p. 5’ RACE R2 GTCGATGGAAGAGTTTCACAAGTCCC

Burs a Carcinus ISHF2GSP CAGGGTCGATGTACTTCATACGTGCAGGCarc5bursA R2 CCAGCACAGTGCCTTCCTCCACATCAG

Burs b Carcinus ISH BursBGSP1F GACACTGCCTTCCACTATCCACATBursBGSP1R TTGCCACCTAACCTAACCACTCC

CCAP Carcinus ISH

CCAP1F GTTGGGACGTACATGGGCTGGTGCCAP1R GTCGGGCTGTGTTTTCTGGTCTTCA

Burs a Carcinus Q-RTPCRCarBursA qPCRF0L CATACCACAGCATCCGCCACCTCCarBursA qPCRR0L TGCCACAGCTTGCTTCCAGATACCCarBursA qPCRF0S GTGGTGGGAGCGGCGGTAACTCarBursA qPCRR0S CTGGCAGGCGAAGGAAGGTATGG

Burs b Carcinus Q-RTPCRCarBursB qPCRF0L GGGCGCGGGACATCACCTTCarBursB qPCRR 0L TACACTGTTGCCGCCTGCTCTTTCCarBursB qPCRR 0S GTTGCCACACCACCACCTCACCAT

AK Carcinus Q-RTPCR

AKSR AAACGGTCACCCTCCTTGAAKLF AAAGGTTTCCTCCACCCTGTAKLR ACTTCCTCGAGCTTGTCACG


Primer pairs were: burs a: F2GSP, Carc5bursA R2, burs b: Burs GSP1 F,1 R, CCAP: CCAP1 F, 1 R. PCR conditions were 5 ll each primer(10 lM), 93 ll Megamix Blue or 50 ll AmpliTaq Gold and 43 ll water,1 ll (ca.50 ng) cDNA (from an RT of 1 lg total RNA, random primersand Superscript III as detailed above). PCR conditions were 94 �C4 min (or 9 min for Amplitaq Gold), followed by 35 cycles of 94 �C 30 s,55 �C 30 s, 72 �C 45 s, final extension 72 �C 7 min. After confirmation ofspecific amplification on agarose gels, PCR products were concentratedon Montage PCR purification modules, and 25% (ca. 50 ng) of the purified

PCR product used for ligation of T7 adaptors. PCR of 2 ll of ligationproduct using forward or reverse GSPs and T7 adapter primers was usedto prepare material for production of antisense cRNA (forward GSP) andsense cRNA (reverse GSP), using PCR procedures shown above. Afterpurification of PCR products, 100–200 ng of DNA was used for produc-tion of run-off transcripts using a MegashortScript kit (Ambion) accord-ing to manufacturers instructions, but transcription procedures weremodified to allow inclusion of DIG-UTP in the reaction: Thus transcrip-tion conditions were: 1 ll each of CTP, GTP, ATP, 0.5 ll UTP (all


75 lM), 2 ll DIG-11UTP (10 lm) (Roche Diagnostics GmbH, Mann-heim, Germany). Following transcription, quality of transcripts werechecked by dotting 1 ll of diluted probe (dilution range 1:1–1:10000) ontoNylon membranes (Hybond-N, GE Healthcare, Amersham, BuckinghamUK). Probes were cross linked to the membranes in a UV cabinet (1 min,max. power), washed (2 · 5 min) in PBST (50 mM phosphate buffered sal-ine, pH 7.4, 0.1% Tween 20), blocked for 30 min in 2% dried milk powderin PBST, and incubated in 1:5000 anti DIG-alkaline phosphatase (sheep,Fab fragments) in PBST (Roche Diagnostics GmbH, Mannheim, Ger-many) for 30–60 min. Membranes were then washed in PBST(3 · 10 min), washed briefly in 0.1 M TRIS, pH 9.5 containing 100 mMNaCl, 50 mM MgCl2 (TMN) including 0.1% Tween 20 (TMNT), andincubated for 30 min (RT, dark) in 1:50 4-nitroblue tetrazolium chlo-ride/5-bromo-4-chloro-3-indolyl-phosphate (NBT/BCIP) (Roche) inTMNT. Reactions were terminated by washing in water. Probe qualitywas deemed satisfactory when a signal was visible at a 1:10000 dilution(approx. 200 pg RNA).

Fixed, methanol dehydrated tissues were rehydrated (100, 66, 33, 0%methanol/PBS, 10 min each) and incubated in 200 lg/ml Proteinase K(Roche) in PBS for 20–30 min. Tissues were then washed in PBST(PBS, 0.1% Tween 20) (2·, then 3 · 5 min) and post-fixed for 1 h in4% PFA. Tissues were rinsed (3 · 10 min, PBST) and prehybridised within-situ hybridisation solution (DakoCytomation, Carpinteria, CA, USA)for 30 min at 50 �C. Hybridisation was performed in fresh hyb. solutionat 1 ng/ll DIG-labelled run-off (antisense or sense) for 18 h at 50 �C.Post-hybridisation stringent washes were performed as follows: 2· SSC50% formamide 50 �C 10 min, 2· 0.2· SSC 10 min, 50 �C, PBST/0.2·SSC (33, 66, 100%) 5 min per wash, RT. To remove unhybridisedDIG-cRNA, ganglia were then washed in 1· TE containing 0.5 M NaCl(TNE) 3· 10 min, and incubated (20 min, RT) in 20 lg/ml RNase A inTNE. Ganglia were then blocked with 1% BSA in PBST for 2 h, andincubated (overnight, RT) in anti-DIG alkaline phosphatase (Roche),1:5000 in 0.1% BSA in PBST. Ganglia were subsequently washed inPBST (5, 10, 2· 15, 2· 30, 1· 60 min), washed in TMNT, briefly washedin TMN, and developed in NBT/BCIP. Colour development was moni-tored microscopically and terminated in distilled water. After visualisa-tion, tissues were further carefully dissected to remove remainingperineural sheath material, and where necessary, tissues were removedwhich overlaid strongly hybridizing cells, which were easily visible undera dissecting microscope. Preparations were mounted in 80% glycerol/PBS. Digitally acquired microscopic images were manipulated usingAdobe Photoshop 7.0 and CorelDraw 8.0.

2.10. Quantitative RT-PCR

Single-species cRNA templates for qPCR were prepared by PCR, liga-tion to T7 promoter adapters (Lig’n Scribe (Ambion), and in vitro tran-scription with T7 RNA polymerase (MEGAshortscript, Ambion),essentially as previously described (Chung and Webster, 2003). A 287 bpburs a template was prepared with primer pairs CarBursA qPCRF 0L, Car-

Fig. 1. Alignments and comparisons of bursicon a and b subunit precursor pmaenas, conceptually translated from cDNA sequences. Putative signal peptidare boxed in grey.

BursA qPCRR 0L. A 336 bp burs b template was prepared with primerpairs CarBursB qPCRF 0L, CarBursB qPCRR 0L. For the control gene(arginine kinase, AK), primer pairs AKLF, AKLR were used (450 bp tem-plate). Run off transcripts were purified on Urea (6 M)–polyacrylamide(10%) gels, and bands of the expected size excised and eluted overnightin Probe Elution Solution (Ambion). Eluted material was ethanol precip-itated, rapidly dried, resuspended in TE, quantified (Abs. @ 260 nm,Nanodrop ND-1000) and converted to copy number (Avogadro’s con-stant/moles). Standards were diluted to 1 · 1011 copies per ll in TE andstored in silanised PCR tubes at �80 �C.

Quantitative PCR was performed on an Applied Biosystems 7500 ther-mocycler using Sensimix (dT) SYBR green PCR Mastermix (Quantace,Watford, UK). Reaction volumes were 25 ll, with 200 nM of one of the fol-lowing primer pairs: CarBursA qPCRF’S/R’S, CarBursB qPCRF’L/R’S,AKLF/SR. Each reaction contained 1 ll of cDNA diluted to containaround 20 ng cDNA (thoracic ganglia), cDNA from 10 embryo equiva-lents, or 1010–103 copies per ll of standards. Assays were performed induplicate for standards and unknowns in 96-well plates (MicroAmp,Applied Biosystems). Cycling conditions were: Initiation 95 �C, 10 min fol-lowed by 40 cycles of 95 �C, 15 s, 60 �C, 60 s. PCR efficiencies were deter-mined using the formula E = �1+10 (�1/slope) where ‘slope’ refers to thegradient of the line plotted from the Ct value/log copies RNA. In all casesPCR efficiencies were >90%. For thoracic ganglia material, copy numbersof burs a and burs b, copy number were normalised per 106 copies AK. Forembryo material, copy numbers were calculated per embryo equivalent.

3. Results

3.1. Characterization of cDNAs encoding burs a, bprecursors

Using primers designed from available ESTs of D. are-nata for burs a, and from the D. pulex genome databasefor burs b, a strategy of 3 0 and 5 0 RACE identified full-length sequences for both subunits, which have been pre-dicted from sequencing the D. pulex genome. Nevertheless,a few differences in sequence were noted in the UTRs forboth subunits, in that the 5 0 UTR of burs a was slightlytruncated, and the entire 3 0 UTR was absent. For burs b,there was a small truncation at the 5 0 UTR. It should bestressed at this point, that the nucleotide sequences ofmRNAs encoding both bursicon subunits was identical tothose predicted from the current D. pulex genome release.The amino acid sequences of both subunits, including sig-nal peptides, identified using Signal P 3.0 (http://www.cbs.dtu.dk/services/SignalP/) are shown in Fig. 1.

eptides in the water flea, Daphnia arenata and green shore crab, Carcinuse sequences are underlined, and amino acids identical in both crustaceans

http://www.cbs.dtu.dk/services/SignalP/http://www.cbs.dtu.dk/services/SignalP/


Given this sequence information, which established thatboth putative bursicon subunit mRNAs were present in aso-called ‘‘primitive’’ crustacean we then moved on to theprimary aim of this project, to identify the correspondingfull-length cDNAs encoding Burs a and Burs b in an‘‘advanced’’ crustacean, the green shore crab C. maenas.Using a strategy of fully degenerate PCR to identify bursa, and using existing EST information of a putative bursb clone from Carcinus to design gene specific primers,together with sequence information obtained fromsequencing burs b in D. arenata, we identified mRNAsencoding both subunits of bursicon in our crab model.Sequences have been submitted to EMBLGenBank dat-

Fig. 2. Sequence alignments of bursicon a and b subunit proteins from crustaceintroduced in the C-terminal region of the crustacean bursicon b subunit, tonumbered. Accession numbers or dbEST identifiers (underlined, from http://wlobster) http://www.aphidbase.org for aphid) are as follows: Bursicon a: Drmellifera, AM420631; Tribolium castaneum, DQ138189; Acyrthosiphon pisum, Ab: Homarus americanus, HOC01591; Drosophila melanogaster, AY823257;castaneum, DQ138190; Acyrthosiphon pisum, APD13715; Manduca sexta, DQ

abases: Accession Nos. D. arenata burs a, EU139431; D.arenata burs b, EU139430; C. maenas burs a, EU139428;C. maenas burs b, EU139429. Conceptually translatedsequences of precursors and mature peptides are shownon Fig. 1. For D. arenata mRNAs encoding Burs a andBurs b subunits consisting of signal peptides of 14 and 20residues, followed by coding sequences for mature peptidesconsisting of 118 (Mr 12969.12 Da) and 118 residues(12821.50 Da) respectively, determined with Cys-reducedresidues (S-H). For C. maenas mRNAs encoding Burs aand Burs b subunits were identified consisting of ORFsencoding conceptually translated peptides including signalpeptides of 27 and 23 residues for Burs a and Burs b,

ans and insects. Identical residues are boxed in grey. Single gaps have beenmaximise sequence similarity with insect peptides. Cysteine residues areww.nematodes.org/NeglectedGenomes/ARTHROPODA/index.shtml forosophila melanogaster, AY672905; Anopheles gambiae, AY735443; ApisPD05277; Manduca sexta, DQ09449; Bombyx mori, BN 000691. BursiconAnopheles gambiae, AY823259; Apis mellifera, AM420632; Tribolium

291147; Bombyx mori, BN000690.

http://www.nematodes.orghttp://www.aphidbase.org


respectively, and mature peptides for Burs a of 121 residues(Mr 13368.48 Da) and for Burs b 115 amino acids (Mr12725.34 Da).

3.2. Expression of burs a, burs b and CCAP in the CNS ofadult Carcinus by in situ hybridisation and RT-PCR

Whole mount in-situ hybridisation using DIG labelledcRNA was performed on thoracic, cerebral and eyestalkganglia of C. maenas, from moult staged crabs. For the

Fig. 3. Expression of bursicon a, b and crustacean cardioactive peptide (CCAmount of fused thoracic ganglion (comprising sub-oesophageal ganglion (sog),DIG- burs b cRNA probe. (b) Thoracic ganglion labelled with antisense DIG-Ca cRNA probe. (d–f) Views of serially iterated pairs of cells in the thoracic gangof large (55–60 lm) and closely associated small (20 lm) cells corresponding toArrows in (d) point to small cells in each neuromere. (g–i) Detail of the sub-oe(a–c). Arrows point to small cells (20 lm) which were closely adjacent to thre(20 lm) anteriorly. (j–l) Detail of the abdominal ganglion cells which hybridizdescribed in (a–c). (m–n) High power detail of serially iterated large and small ceor antisense DIG-burs b (n). (o) Overview of eyestalk ganglia hybridized withdetailed in (p–r). (p) Group (11–12) of small cells (10–12 lm) anterior to the Xinterna and medulla terminalis. (r) Group (6) of small (10–12 lm) cells posterpairs) of small (10 lm) cells hybridizing with antisense DIG-CCAP cRNA probposition. In all cases, for every tissue examined, control (sense) DIG- labelled cRc), 500 lm (o), 200 lm (g, h, j–l), 100 lm (d–f, i, s), 50 lm (m, n, q, t), 25 lm

entire (fused) thoracic ganglion (TG), very striking patternsof cell specific gene expression for burs a, b and CCAP wereobserved which were strongly suggestive of complete colo-calisation in all TG neurones as shown in Fig. 3. Antisenseprobes gave impressive, strong and clear positive hybridisa-tion signals, which could always be seen by low powermicroscopical examination during visualisation with NBT/BCIP. In all cases sense probes showed no specific hybridisa-tion (results not shown). Overall, expression of all threetranscripts was surprisingly similar in the TG (Fig. 3a–c).

P) transcripts in the central nervous system of Carcinus maenas. (a) Wholethoracic ganglion (tg) and abdominal ganglion (ag)) labelled with antisenseCAP cRNA probe. (c) Thoracic ganglion labelled with antisense DIG-burslion corresponding to the sequence of probes described for (a–c). Five pairs

each thoracic neuromere were seen on each side of the thoracic ganglion.sophageal ganglion, corresponding to the sequence of probes described ine pairs of large cells (55–60 lm) posteriorly and two pairs of smaller cellsed with all three probes, corresponding to the probe sequence previouslyll pairs in the thoracic ganglion, hybridizing with antisense DIG-burs a (m)antisense DIG-CCAP cRNA probe. Boxes show position of cell groups

-organ). (q) Pair of large (40 lm) cells at the junction between the medullaior to the X-organ. (s) Ventral view of brain showing two groups (11–12e. Box (t) shows position of a pair of large cells (25 lm) in a dorsal midlineNA probes showed no specific hybridization signals. Scale bars: 1 mm (a–

(p, r).


Within the TG, 5 dorsal pairs of large (55–60 lm) and smallperikarya (20 lm) corresponding to thoracic neuromeres 1–5 were always observed (Fig. 3d–f, m, and n), whilst in thesub-oesophageal ganglion, three pairs of large (55–60 lm)and small (20 lm) dorsal perikarya were observed proxi-mally, and two pairs of small (20 lm) distally positionedperikarya (Fig. 3g–i). For the fused abdominal ganglion(AG), the precise number of perikarya was difficult to deter-mine with certainty, due to the somewhat aggressive proce-dures used in ISH, and resultant deformation/digestion oftissues. Nevertheless, in general two pairs of prominentanterior perikarya (60 lm) appeared to be followed by ninepairs of posteriorly located cells of similar diameter (Fig. 3,j–l). For the remainder of the CNS, burs a and b transcriptswere never observed, despite the observation that CCAPtranscripts were present in a variety of cells in the eyestalkganglia and brain: For the brain (CG), CCAP transcriptswere noted in a pair of large (25 lm) dorsal midline cells(Fig. 3t), and 12 pairs of small (10–12 lm) ventral cells inthe protocerebrum (Fig. 3s). In the eyestalk, two large(40 lm) cells were always seen at the posterior margin ofthe medulla interna (Fig. 3q), and small (10–12 lm diame-ter), n = 6 were observed posteriorly and anterio-dorsally(n = 11–12) to the X-organ (Fig. 3p and r).

To get a broad picture of burs a, b and CCAP, expres-sion, end-point PCR was performed in ES, CG and TG(Fig. 4). CCAP expression was seen in all three tissues,and particularly the TG, as expected. However, for bursa- and burs b, high level expression was only seen in theTG, in accord with the in-situ hybridisation experiments.Nevertheless, low level expression of burs b was alwaysclearly detectable in the ES. This was not due to gDNAcontamination or carry over, since RT—experiments werealways negative (results not shown). PCR of cDNA fromother tissues (midgut gland, heart, ovary testis) did notamplify burs a, b or CCAP (results not shown).

3.3. Quantitative RT-PCR: Expression of burs a and burs b,in the thoracic ganglia of C. maenas during the moult cycle,

and embryogenesis

Quantitative RT-PCR was performed on RNAextracted from thoracic ganglia of moult staged male and

Fig. 4. Expression of CCAP, burs a, b transcripts in the nervous system ofCarcinus maenas. RNA (100 ng) from premoult crabs was reversetranscribed and amplified using primer pairs used for in-situ probeproduction. PCR product sizes were; CCAP 430 bp, burs a 223 bp, burs b401 bp. Size ladder (pUC 19 Sau3A digest) shown on left.

female C. maenas. The normalising gene used was argininekinase (AK), which we have previously shown via Q-RT-PCR to exhibit small changes in expression (a 2-fold reduc-tion during premoult) during the moult cycle in eyestalktissues (Chung and Webster, 2003). The results obtained(Fig. 5), indicate that both transcripts are reasonably abun-dant and undergo quite modest changes in abundance dur-ing the moult cycle. Nevertheless, for burs a, levels ofexpression in intermoult (C4) were significantly higher thanlate postmoult (C1–3) or early premoult (D0), (P < 0.05;ANOVA, Tukey–Kramer multiple comparisons). Whilstlevels of intermoult burs b expression were also higher thanthose of late postmoult or early premoult, differences werenot significant between different moult stages (P = 0.28;one-way ANOVA). Expression levels of burs a were about3-fold higher than burs b. This was not accountable due todifferences in PCR efficiency (burs a, slope �3.57, 91% effi-ciency, b slope �3.47, 94% efficiency) or to quantification

Fig. 5. Expression of burs a (upper panel) and burs b (lower panel) mRNAin thoracic ganglia of Carcinus maenas during the moult cycle. Moultstages were determined according to Drach and Tchernigovtzeff (1967); A-B: postmoult, C1-4: intermoult, D0-4: premoult. Measurements of bursiconcopy numbers were normalised to the expression of a housekeeping gene,arginine kinase (AK). Error bars represent +1 SEM. Thoracic ganglia (4–13) were used at each moult stage. All standards and unknowns wereassayed in duplicate.


errors since crossing threshold (Ct) values for both quanti-fied standards were almost identical (Ct burs a: bursb = 0.99 at 108 copies, Ct burs a:burs b = 1.03 at 104

copies).During embryogenesis, significant expression of burs a

and burs b was first seen during early eye development(eye smear, 50% development) according to the develop-mental stages outlined for this species (Chung and Webster,2004). Levels increased dramatically during embryogenesis,particularly between the eye anlage and mid-eye stages. Forburs b, levels were somewhat reduced just before hatching,

Fig. 6. Expression of burs a (upper panel) and burs b (lower panel) mRNAduring embryonic development of Carcinus maenas. Developmental stageswere as previously defined (Chung and Webster, 2004). Independentsamples (4-14) were taken at each developmental stage. Data are expressedas copies of mRNA per embryo equivalent. Error bars represent +1SEM.All standards and unknowns were assayed in duplicate. During earlyembryonic development (yolk-limb bud stages), low, but significantnumbers of transcripts were measured and means are shown in brackets.Inset in lower panel shows ratios of burs a/burs b mRNA duringembryonic development.

but this was statistically non-significant (ANOVA, Tukey–Kramer multiple comparisons, P > 0.05), (Fig. 6). The ratioof burs a: burs b increased during development. When firstexpressed, both transcript ratios were similar, but burs awas around 3-fold more abundant relative to burs b duringlater stages of development (Fig. 6, inset), and just prior tohatching, the burs a: burs b ratio increased further.

4. Discussion

In this study we have identified cDNAs (mRNAs)encoding both subunits of a putative bursicon molecule incrustaceans. The recently completed Daphnia GenomeProject allowed us to firstly confirm the existence of bursa and burs b mRNAs from Daphnia arenata, which hadbeen predicted from ESTs and shotgun sequencing strate-gies of the D. pulex genome (D. pulex, D. arenata, D. puli-caria, D. melanica are considered to be part of the samespecies complex (Colbourne and Hebert, 1996)). Exceptinga truncation in the 3 0 UTR for D. arenata burs a (undoubt-edly due to hybridisation of a poly A rich region of the 3 0

end immediately adjacent to the stop codon, with the 3 0

RACE adapter primer during reverse transcription), anda slightly truncated 5 0 UTR for burs a and to a lesser extentburs b, the coding regions for burs a and burs b wereentirely identical to the nucleotide and protein sequencespredicted from the current D. pulex genome release(V1.1). Since the most important aim of this investigationwas to identify bursicon transcripts in a decapod crusta-cean, the sequences obtained from D. arenata, in conjunc-tion with a sequence from an EST encoding a putative bursb obtained from a ‘‘mixed-tissue’’ cDNA library of C. mae-nas were used to identify full-length cDNAs derived fromCarcinus nervous tissue mRNA. The conceptual sequencesof both crustacean bursicon subunits, including putativesignal peptides are shown on Fig. 1. From these, it is clearfor both crustaceans, that there is a striking degree ofsequence conservation. When compared to unambiguousfull-length sequences of insect bursicon subunits (Fig. 2)it is once more immediately obvious that all are very highlyrelated, not only with regard to identical positions of allCys residues-which is important bearing in mind that theCys in position 6 has been proposed to form the singledisulphide bridge between the two monomers (Luo et al.,2005), but also from the viewpoint of our parsimoniousinterpretation of sequence similarity, in which only identi-cal residues are highlighted; this clearly identifies bothgroups of peptides in the crustaceans and insects. For theCys-containing core region of Burs a, sequence identitywith the prototype Drosophila sequence is 76% and 74%respectively for Daphnia and Carcinus sequences. For theBurs b subunit identities with the Drosophila sequence are59% (Daphnia and Carcinus) and 62% (Homarus). A nota-ble feature of crustacean Burs b is a deletion of D/E, 19 res-idues from the C-terminus. Since this is found in all threespecies of crustaceans so far examined it seems possiblethat this is a feature of crustacean Burs b subunits. Whilst


for Burs b species specific differences seem to be concen-trated on the N-terminus, and to a lesser extent from resi-dues 16–31 proximal to the C-terminus, for Burs a there isconsiderable variation at both the N and C-termini.

Although it is well established that bursicon subunits arecystine knot proteins, related to growth factors such asbone morphogenetic proteins (BMPs) and their antagonists(Luo et al., 2005), transforming growth factor beta (TGFb), glycoprotein hormone subunits and mucins (review byVitt et al., 2001) it is notable that none of the these lattermolecules are well conserved enough to be considered closerelatives of bursicon-like molecules with the notable excep-tion of two putative transcripts recently identified from thegenome of the (deuterostome) sea urchin Stongylocentrotuspurpuratus, whose subunits share about 45% similarity (butnot identity) with arthropod bursicon (Van Loy et al.,2007). This observation is intriguing, particularly sincerelated molecules have not been identified from genome-wide screens of the model ‘‘ecdysozoan’’ Caenorhabditiselegans. Given the wide distribution of bursicon-like mole-cules in arthropods and likely the echinoderms, the succinctstatement by these authors, encouraging comparative func-tional studies is entirely justified.

To examine qualitative and quantitative expression pat-terns of burs a and burs b transcripts, and their distributionwithin the CNS of Carcinus, we used whole mount in-situhybridisation (ISH) and quantitative RT-PCR. In-situhybridisations of thoracic ganglia (which are fused tosub-oesophageal and abdominal ganglia in crabs) showedthat both transcripts were localised in the same neurones(as would be expected). Comparison with ISH for CCAPshowed that for the thoracic ganglia, all neurones express-ing burs a and burs b transcripts apparently also expressedCCAP. Whilst it is arguable that a more discriminatoryapproach would have been to raise antisera directedagainst bursicon subunits in conjunction with double label-ling for CCAP (Dircksen and Keller, 1988), it seems verylikely from the results obtained for thoracic ganglia in-situhybridisations that all three transcripts occur in the sameneurones. The pattern of distribution of bursicon tran-scripts in the TG is particularly striking when comparedto the detailed neuroanatomy of CCAP containing neuro-nes in the TG of Carcinus (Dircksen and Keller, 1988; Dir-cksen 1998). In particular, the segmentally iterated pairs oflarge, 55–60 lm (cnc-type-1) and small, 20 lm (cdn-type-2)(nomenclature: Dircksen, 1998), CCAP expressing neuro-nes (perikarya) are identical in terms of burs a, b hybridisa-tion patterns and anatomical position as are the pairedsegmentally iterated neurones in the sub-oesophageal gan-glion. For the abdominal ganglion, only large (cdc-type-1)neurones gave similar hybridisation patterns to all threeprobes, and it is notable that in this ganglion, small (cdn-type-2) CCAP containing neurones appear to be absent.This feature has also been noted by Dircksen (1998) bywhole mount immunohistochemistry for CCAP on Carci-nus abdominal ganglia. Additionally, the projection pat-terns of these neurones, which are suggestive of a

neuroendocrine function (projection to the pericardialorgans) has been noted by this author. Striking similaritiesare also seen with regard to colocalisation of CCAP andbursicon in insects, for example the 24–27 neurones inthe abdominal ganglion of M. sexta, show bursicon bioac-tivity (Taghert and Truman, 1982) were later shown to con-tain CCAP (Davis et al., 2003). Similarly bursicon activityin CCAP neurones has been shown in crickets, Gryllusbimaculatus (Kostron et al., 1996), cockroaches, P. ameri-cana (Honegger et al., 2002) and Drosophila melanogaster(Dewey et al., 2004). Antisera raised against bursicon aand b subunits co-label subsets of four bilateral neuronesin third instar Drosophila larvae (Luo et al., 2005), yet theseauthors note that some neurones in the suboesophagealand posterior abdominal neuromeres express only Burs abut not Burs b in Drosophila, and they suggest other asyet unknown roles for Burs a.

Nevertheless, not all CCAP cells contain bursicon ininsects. For example, two pairs of CCAP-ir neurones inthe brain and in the first thoracic neuromere do not expressbursicon in Drosophila (Dewey et al., 2004). Furthermore,in abdominal ganglia of P. americana, a pair of smallcdn-type-2 neurones seem to be immunopositive only forCCAP, in contrast to the adjacent pair of large cnc-type-1 neurones, which are immunopositive for both Burs band CCAP (Luo et al., 2005). It seems possible that theperikarya which exclusively express CCAP might beinvolved in control of the motor network regulating theecdysis motor pattern. However, CCAP KO flies exhibitprofound defects in cuticle tanning, wing extension andhave low survival to adulthood (Park et al., 2003), consis-tent with the hypothesis that most CCAP neurones co-express bursicon. In contrast, it has been shown (Honeggeret al., 2002) that brain neurones that showed bursiconlabelling (via an antiserum raised against a dodecamer ofa tryptic fragment of cockroach Burs a) did not colocalisewith CCAP in P. americana.

In the present study, some neurones in the brain andeyestalk were identified via ISH, which only expressedCCAP. These are most likely interneurones (Dircksen,1998). Whilst low levels of expression of burs b wereobserved in eyestalk tissues via RT-PCR, we neverobserved any neurones that hybridised with bursiconprobed by wholemount ISH in eyestalk or brains. Clearlydefinitive studies to unequivocally determine colocalisationof CCAP with bursicon are now needed, using immunohis-tochemical approaches.

Our studies on expression of bursicon transcripts in thethoracic ganglia of Carcinus, during the moult cycleshowed that both burs a and burs b transcripts were quiteabundant, and that burs a mRNA was around 3-fold moreabundant than burs b. Whilst this difference might beaccounted for by consistent systematic errors in run-offtranscript quantification of standard cRNAs, or differencesin RT or PCR efficiencies (a 3-fold difference in quantifica-tion would only be observed by a single-cycle difference inCt value), because we observed a Ct correlation of 0.99–


1.03 between (un)quantified transcripts of both burs a andburs b and consistently high (91–94%) PCR efficiencies, wecan be confident that the observed differences indeed reflectsignificant differences in steady-state transcription profilesof mRNAs encoding both transcripts. However, notwith-standing any differences in ratios of steady-state transcrip-tion of both subunit mRNAs, the most surprisingobservation was that mRNA levels of bursicon transcriptswas somewhat invariant during the moult cycle. This mightalso be pertinent in view of our previous studies on CCAPexpression in the thoracic ganglia of Carcinus, using quan-titative RT-PCR (Chung et al., 2006), where little evidencefor upregulation of CCAP mRNA during the moult cyclewas found, notwithstanding a dramatic release of this pep-tide from the pericardial organs just before ecdysis (Phlip-pen et al., 2000). Levels of both bursicon transcripts werehighest during intermoult, (significantly so for burs a). Thissituation contrasts vividly with that of Drosophila, wherelevels of burs a and burs b expression decline dramaticallybetween pharate and post-eclosion adults (Luo et al.,2005)—as would be expected given our current understand-ing of the mode of action of this hormone in insects (Ewerand Reynolds, 2002). Additionally, we never saw obviouschanges in bursicon transcript expression in thoracic gan-glia taken from crabs during premoult, ecdysis, or postmo-ult. Comparing transcriptional levels for CCAP, burs a andburs b, mRNA copy number, normalised against AKshows that for CCAP, steady-state copy number was about2–4 · 104 copies per 106 AK (data from Chung et al., 2006).This compares well with levels of burs b which were verysimilar (2–6·104 copies per 106 AK). As noted earlier, bursa transcript numbers were around 3-fold higher (5–15 · 104

copies per 106 AK). Despite these rather small differences, areasonable assumption would be that transcription ofmRNAs encoding all three proteins was rather similar inthe TG, and that expression of these is constitutive in adultcrabs, irrespective of moult cycle stage.

Since we have previously reported changes in expressionpatterns of CCAP during embryogenesis in Carcinus, wewere interested in comparing expression patterns of burstranscripts during this process, particularly in view of ourobservations in adults suggesting that burs a was more abun-dantly expressed compared to burs b. As would be expected,levels of both bursicon transcripts increased during embyo-genesis, and both were detected during eye smear/anlageformation (Chung and Webster, 2004), i.e. at about 50%development. As with adults, burs a transcripts were moreabundant than burs b, and were approximately 6-fold moreabundant just before hatching. Determination of bursicontranscripts during the hatching (zoea 1) moult remain tobe determined, as are measurements during the megalopa-first crab metamorphosis, which will be interesting in com-parison with analogous metamorphic moults in insects.However, whilst our results clearly show that bursiconexpression mirrors that of CCAP, it is interesting to notethat bursicon expression is 10–20-fold lower than that ofCCAP during embryonic development. Since the material

used for measurement of both CCAP and bursicon expres-sion by quantititative PCR originated from identical poolsof embryo cDNA that were used in our study of CCAPexpression (Chung et al., 2006), it is likely that this repre-sents a real difference in the ratios of CCAP and bursiconexpression during embryogenesis.

An intriguing dilemma regarding the roles of CCAP andbursicon during the ecdysis program of insects has beennoted by Honegger et al. (2002): ‘‘How are these two neu-ropeptides released such that they have sequentialactions?’’ In Manduca, and most likely in other insects, itis clear that CCAP is released at ecdysis, and activatesthe ecdysis motor program that results in stereotypedbehaviour associated with shedding of the old cuticle(Gammie and Truman, 1997), notwithstanding a plethoraof other roles. However, CCAP-KO Drosophila that arelacking in both CCAP and bursicon express normal eclo-sion behaviour (Park et al., 2003), although in such fliesheartbeat seems to be abnormal (Dulcis et al., 2005). Wehave previously shown that ecdysis in crayfish and crabs(Orconectes and Carcinus) is accompanied by a massiverelease of CCAP into the haemolymph (Phlippen et al.,2000), that by analogy to insects might be expected to ini-tiate ecdysial behaviour programmes. Yet, since CCAP andbursicon are expressed in the same neurones in the thoracicganglia, which project neurones to the pericardial organs(the principal site of CCAP and hence bursicon release),it would be expected that release of both hormones wouldoccur simultaneously. Yet, release of bursicon—the ulti-mate player in the ecdysis programme—should intuitivelyoccur after CCAP release. Since we have shown that allneurones expressing CCAP mRNA in the TG probablyco-express both bursicon subunit mRNAs, and consideringthe neuroanatomy of the TG, whereby all cdc-type-1 neu-rones project axons to the pericardial organs (Dircksen,1998), it seems reasonable to suggest that both neurohor-mones are simultaneously released into the haemolymphjust before moulting. Obviously, this remains to be investi-gated, and it must be emphasised that it is critically impor-tant not to make comparative assumptions regarding theputative roles of bursicon in cuticle tanning and sclerotisa-tion in crustaceans- particularly in view of our observationssuggesting that expression is constitutive throughout theentire moult cycle in adult crustaceans, which are com-pletely different from expression profiles of these hormonesin the pupal/adult moult in holometabolous insects. Addi-tionally, the possible roles of the cdn-type-2 CCAP neuro-nes might be relevant. These neurones do not project axonsto the pericardial organs, and are most likely interneurones(Dircksen, 1998). The surge in CCAP in the haemolymphduring ecdysis is due to release from PO endings innervatedby cdc-type 1 neurones, which could directly lead toincreases in cardiac output. This may be essential duringecdysis, particularly when the considerable expansion oftissues in postmoult is taken into account, which shouldbe accompanied by increases in tissue perfusion via simul-taneous increases in cardiac output. It may also be possible


that the cdn-type 2 interneurones contribute to initiation ofstereotyped motor behaviours-indeed such a scenariomight separate the actions of CCAP and bursicon in thesense of local (CNS) vs. central (haemolymph) release dur-ing the ecdysis program. In this context, a recent reportdescribing an elegant functional genetic dissection of neu-ronal networks in adult Drosophila, via enhancer trapping(Luan et al., 2006) is particularly important. 14 CCAP/bursicon-expressing neurones (NCCAP-c929) in the abdom-inal ganglion lie within the expression pattern of the enhan-cer-trap line c929-Gal4 (Hewes et al., 2003). Targetedsuppression of excitability of this group of cells viaCCAP-Gal4 expression of two K+ channel constructsblocked bursicon release and consequent tanning and wingexpansion. Yet a second group of CCAP cells (N CCAP-R),mostly seen in the abdominal ganglion, (which lie outsidethe expression pattern of this enhancer-trap), seem to beinvolved in regulating bursicon secretion from NCCAP-c929. The model of regulation of bursicon release fromCCAP neurones proposed by these authors is very interest-ing and intriguing, and might possibly address the dilemmaregarding temporal separation of CCAP and bursiconrelease during the ecdysis programme.

Clearly there is still much to learn about the hormonecascades during ecdysis in arthropods, and particularlycrustaceans. Finding the biological roles of bursicon duringecdysis (and during the moult cycle), determining the pre-cise spatial and temporal patterns of release, and possibleparticipation of other peptides in crustaceans is the nextchallenge; completion of the series of endocrine cascadesand comparison with the much better known systems in(genetically tractable) insects, will yield fascinating insightsinto the common and divergent endocrine mechanismsinvolved in control of ecdysis in both arthropod groups.

Acknowledgments

We are indebted to Prof. J. Colbourne, University ofIndiana, Bloomington, USA, for his gift of Daphnia arenat-a. Sequence data for Daphnia were produced by the USDepartment of Energy Joint Genome Institute http://www.jgi.doe.gov/in collaboration with the DaphniaGenomics Consortium http://daphnia.cgb.indiana.edu.We also thank Prof. Heinrich Dircksen (University ofStockholm, Sweden) for insightful comments and many li-vely discussions. This work was funded by the Biotechnol-ogy and Biological Sciences Research Council, GrantReference No. BB/E023126/1.

References

Baker, J.D., Truman, J.D., 2002. Mutations in the Drosophila glycopro-tein receptor, rickets, eliminate neuropeptide-induced tanning andselectively block a stereotyped behavioural program. J. Exp. Biol. 205,2555–2565.

Chung, J.S., Webster, S.G., 2004. Expression and release patterns ofneuropeptides during embryonic development and hatching in thegreen crab, Carcinus maenas. Development 131, 4751–4761.

Chung, J.S., Webster, S.G., 2003. Moult cycle-related changes inbiological activity of moult-inhibiting hormone (MIH) and crustaceanhyperglycaemic hormone (CHH) in the crab, Carcinus maenas: fromtarget to transcript. Eur. J. Biochem. 270, 3280–3288.

Chung, J.S., Wilcockson, D.C., Zmora, N., Zohar, Y., Dircksen, H.,Webster, S.G., 2006. Identification and developmental expression ofmRNAs encoding crustacean cardioactive peptide (CCAP) in decapodcrustaceans. J. Exp. Biol. 209, 3862–3872.

Colbourne, J.K., Hebert, P.D., 1996. The systematics of North AmericanDaphnia (Crustacea: Anomopoda) a molecular phylogenetic approach.Philos. Trans. R. Soc. Lond. B Biol. Sci. 351, 349–361.

Cottrell, C.B., 1962. The imaginal ecdysis of blowflies. J. Exp. Biol. 39,413–431.

Davis, N.T., Homberg, U., Dircksen, H., Levine, R.B., Hildebrand, J.G.,2003. Crustacean cardioactive peptide-immunoreactive neurons in thehawkmoth Manduca sexta and changes in their immunoreactivityduring postembryonic development. J. Comp. Neurol. 338, 612–627.

Dewey, E.M., McNabb, S.L., Ewer, J., Kuo, G.R., Takanishi, C.L.,Truman, J.W., Honegger, H.W., 2004. Identification of the geneencoding bursicon, an insect neuropeptide responsible for cuticlesclerotization and wing spreading. Curr. Biol. 14, 1208–1213.

Dircksen, H., 1998. Conserved crustacean cardioactive peptide (CCAP)neuronal networks and functions in arthropod evolution. In: Coast,G.M., Webster, S.G. (Eds.), Recent Advances in Arthropod Endocri-nology, vol. 65. Cambridge University Press, Cambridge, pp. 302–333.

Dircksen, H., Keller, R., 1988. Immunocytochemical localization ofCCAP, a novel crustacean cardioactive peptide, in the nervous systemof the shore crab, Carcinus maenas L. Cell Tissue Res. 254, 347–360.

Drach, P., Tchernigovtzeff, C., 1967. Sur la méthode de détermination desstades d’intermue et son application générale aux Crustacés. Vie etmilieu ser. A Biol. 18, 595–610.

Dulcis, D., Levine, R., Ewer, J., 2005. Role of the neuropeptide CCAP inDrosophila cardiac function. J. Neurobiol. 64, 259–274.

Ewer, J., 2005. Behavioral actions of neuropeptides in invertebrates:insights from Drosophila. Horm. Behav. 48, 418–429.

Ewer, J., Reynolds, S., 2002. Neuropeptide control of molting in insects.In: Pfaff, D.W., Arnold, A.P., Fahrbach, S.E., Etgen, A.M., Rubin,R.T. (Eds.), Hormones, Brain and Behavior. Academic Press, SanDiego CA, pp. 1–92.

Fraenkel, G., Hsiao, C., 1962. Hormonal and nervous control of tanningin the fly. Science 138, 27–29.

Fraenkel, G., Hsiao, C., 1963. Tanning in the adult fly: a new function ofneurosecretion in the brain. Science 141, 1057–1058.

Fraenkel, G., Hsiao, C., 1965. Bursicon, a hormone which mediatedtanning of the cuticle in the adult fly and other insects. J. InsectPhysiol. 11, 513–556.

Fraenkel, G., Hsiao, C., Seligman, M., 1966. Properties of bursicon: aninsect protein that controls cuticular tanning. Science 151, 91–93.

Gammie, S.C., Truman, J.W., 1997. Neuropeptide hierarchies and theactivation of sequential motor behaviors in the hawkmoth, Manducasexta. J. Neurosci. 17, 4389–4397.

Gilbert, L.I., Rybczynski, R., Tobe, S., 1996. Endocrine cascades in insectmetamorphosis. In: Gilbert, L.I., Tata, J., Atkinson, P. (Eds.),Metamorphosis: Post-embryonic Reprogramming of Gene Expressionin Amphibian and Insect Cells. Academic Press, San Diego, CA, pp.59–107.

Gilbert, L.I., Rybczynski, R., Warren, J.T., 2002. Control and biochem-ical nature of the ecdysteroidogenic pathway. Ann. Rev. Entomol. 47,883–916.

Hewes, R.S., Park, D., Gauthier, S.A., Schaefer, A.M., Taghert, P.H.,2003. The bHLH protein Dimmed controls neuroendocrine celldifferentiation in Drosophila. Development 130, 1771–1781.

Honegger, H.W., Market, D., Pierce, L.A., Dewey, E.M., Kostron, B.,Wilson, M., Choi, D., Klukas, K.A., Mesce, K.A., 2002. Cellularlocalization of bursicon using antisera against partial peptide sequences ofthis insect cuticle-sclerotizing hormone. J. Comp. Neurol. 452, 163–177.

Kaltenhauser, U., Kellerman, J., Andersson, K., Lottspeich, F., Honeg-ger, H.W., 1995. Purification and partial characterization of bursicon,

http://www.jgi.doe.gov/inhttp://www.jgi.doe.gov/inhttp://www.daphnia.cgb.indiana.edu


a cuticle sclerotizing neuropeptide in insects, from Tenebrio molitor.Insect Biochem. Mol. Biol. 25, 525–533.

Kostron, B., Markquardt, K., Kaltenhauser, U., Honegger, H.W., 1995.Bursicon, the cuticle sclerotizing hormone: comparison of its molecularmass in different insects. J. Insect Physiol. 41, 1045–1053.

Kostron, B., Kaltenhauser, U., Seibel, B., Bräunig, P., Honegger, H.W.,1996. Localization of bursicon in CCAP-immunoreactive cells in thethoracic ganglia of the cricket Gryllus bimaculatus. J. Exp. Biol. 199,367–377.

Kostron, B., Market, D., Kellerman, J., Carter, C.E., Honegger, H.W.,1999. Antisera against Periplaneta americana Cu, Zn-superoxidedismutase (SOD): separation of the neurohormone bursicon fromSOD, and immunodetection of SOD in the central nervous system.Insect Biochem. Mol. Biol. 29, 861–871.

Luan, H., Lemon, W.C., Peabody, N.C., Pohl, J.B., Zelensky, P.K.,Wang, D., Nitabach, M.N., Holmes, T.C., White, B.H., 2006.Functional dissection of a neuronal network required for cuticletanning and wing expansion in Drosophila. J. Neurosci. 26, 573–584.

Luo, C-W., Dewey, E.M., Sudo, S., Ewer, J., Hsu, S.Y., Honegger, H.W.,Hsueh, A.J.W., 2005. Bursicon, the insect cuticle-hardening hormone,is a heterodimeric cystine knot protein that activates G protein-coupled receptor LGR2. Proc. Natl. Acad. Sci. USA 102, 2820–2825.

Mendive, F.M., Van Loy, T., Claeysen, S., Poels, J., Williamson, M.,Hauser, F., Grimmelikhuijzen, C.J.P., Vassart, G., Vanden Broeck, J.,2005. Drosophila molting neurohormone bursicon is a heterodimer andthe natural agonist of the orphan receptor DLGR2. FEBS Lett. 579,2171–2176.

Mesce, K.A., Fahrbach, S.E., 2002. Integration of endocrine signals thatregulate insect ecdysis. Front. Neuroendocrinol. 23, 179–199.

Park, J.H., Schroeder, A.J., Helfrich-Förster, C., Jackson, F.R., Ewer, J.,2003. Targeted ablation of CCAP neuropeptide-containing neurons ofDrosophila causes specific defects in execution and circadian timing ofecdysis behaviour. Development 130, 2645–2656.

Phlippen, M.K., Webster, S.G., Chung, J.S., Dircksen, H., 2000. Ecdysisof decapod crustaceans is associated with a dramatic release ofcrustacean cardioactive peptide into the haemolymph. J. Exp. Biol.203, 521–536.

Riddiford, L.M., 1989. The epidermis as a model system for ecdysteroidaction. In: Ecdysone: From Chemistry to Mode of Action. Thieme,Stuttgart, pp. 407–413.

Rose, T.M., Schultz, E.R., Henikoff, J.G., Pietrokovski, S., McCallum,C.M., Henikoff, S., 1998. Consensus-degenerate hybrid oligonucleotideprimers for amplification of distantly-related sequences. Nucl. AcidsRes. 26, 1628–1635.

Seligman, I.M., 1980. Bursicon. In: Miller, T.A. (Ed.), NeurohormonalTechniques in Insects. Springer-Verlag, Berlin, pp. 137–153.

Taghert, P.H., Truman, J.W., 1982. Identification of the bursicon-containing neurones in abdominal ganglia of the tobacco hornwormManduca sexta. J. Exp. Biol. 98, 385–401.

Truman, J.W., 2005. Hormonal control of insect ecdysis: endocrinecascades for coordinating behaviour with physiology. Vitam. Horm.73, 1–30.

Van Loy, T., Van Hiel, M.B., Vandermissen, H.P., Poels, J., Mendive, F.,Vassart, G., Vanden Broeck, J., 2007. Evolutionary conservation ofbursicon in the animal kingdom. Gen. Comp. Endocrinol. 153, 59–63.

Vitt, U.A., Hsu, S.Y., Hsueh, A.J., 2001. Evolution and classification ofcystine knot-containing hormones and related extracellular signallingmolecules. Mol. Endocrinol. 15, 681–694.

Identification and developmental expression of mRNAs encoding putative insect cuticle hardening hormone, bursicon in the green shore crab Carcinus maenasIntroductionMaterials and methodsAnimals, tissue preparation and RNA extractioncDNA synthesis and rapid amplification of cDNA ends (RACE)Primers3 prime and 5 prime RACE PCR of Daphnia arenata cDNA encoding burs alpha and burs beta Burs alpha Burs beta

Degenerate PCR of Carcinus cDNA encoding burs alpha 3 prime and 5 prime RACE PCR of Carcinus cDNA encoding burs alpha 3 prime and 5 prime RACE PCR of Carcinus cDNA encoding burs beta Cloning and sequencing of PCR productsIn situ hybridisation: Carcinus burs alpha , burs beta , CCAPQuantitative RT-PCR

ResultsCharacterization of cDNAs encoding burs alpha , beta precursorsExpression of burs alpha , burs beta and CCAP in the CNS of adult Carcinus by in blank situ hybridisation and RT-PCRQuantitative RT-PCR: Expression of burs alpha and burs beta , in the thoracic ganglia of C. maenas during the moult cycle, and embryogenesis

DiscussionAcknowledgmentsReferences

Documents

Identiﬁcation and developmental expression of mRNAs ... · phoresis and tryptic digests of cockroach (Periplaneta americana) bursicon containing extracts (Honegger et al., 2002),