Molecular identification of the first insect ecdysis triggering hormone receptors

Molecular identification of the first insect ecdysis triggeringhormone receptorsq

Annette Iversen, Giuseppe Cazzamali, Michael Williamson,Frank Hauser, and Cornelis J.P. Grimmelikhuijzen*

Department of Cell Biology, Zoological Institute, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen, Denmark

Received 8 November 2002

Abstract

The Drosophila Genome Project website (www.flybase.org) contains an annotated gene sequence (CG5911), coding for a G

protein-coupled receptor. We cloned the cDNA corresponding to this sequence and found that the gene has not been correctly

predicted. The corrected gene CG5911 has five introns and six exons (1–6). Alternative splicing yields two cDNAs called A (con-

taining exons 1–5) and B (containing exons 1–4, 6). We expressed these splicing variants in Chinese hamster ovary cells and found

that the corrected CG5911-A and -B cDNAs coded for two different G protein-coupled receptors that could be activated by low

concentrations of Drosophila ecdysis triggering hormones-1 and -2. Ecdysis (cuticle shedding) is an important behaviour, allowing

growth and metamorphosis in insects and other arthropods. Our paper is the first report on the molecular identification of ecdysis

triggering hormone receptors from insects.

� 2002 Elsevier Science (USA). All rights reserved.

Insects are ecologically and economically important

animals, because more than 70% of all flowering plants

depend on insects for their pollination, and insects can

be vectors for serious diseases such as malaria, ele-

phantiasis, sleeping disease, and yellow fever. Despite

the importance of insects, however, our knowledge oftheir biology is still incomplete. This will certainly

change after the recent completion of the sequencing of

the Drosophila genome (www.flybase.org; [1]), which has

made it possible to identify and characterize all proteins

in an insect and, thereby, to understand its whole bio-

chemistry and biology. We ourselves are particularly

interested in neuropeptide receptors and their ligands,

because these proteins and peptides occupy a high hi-erarchic position in the physiology of an insect and steer

central processes such as reproduction, development,

and feeding [2,3].

The website of the Drosophila Genome Project con-

sortium contains a list of 40–45 potential neuropeptide

receptor genes (www.flybase.org; [4]). Most of these re-

ceptor genes, however, have been identified by computer

programs and their structures have often not been cor-

rectly predicted [5,6]. Furthermore, the ligands for mostof these annotated receptors are unknown, i.e., they are

orphan receptors, and we do not know their functions.

Therefore, proper cDNA cloning of these receptors,

functional expression in cells, and identification of their

cognate ligands are still necessary.

In the present paper we have investigated the an-

notated neuropeptide receptor gene CG5911, estab-

lished its proper (corrected) intron/exon organization,and identified the Drosophila ecdysis triggering hor-

mones-1 and -2 (ETH-1 and -2) as the cognate receptor

ligands. This is the first report on the molecular iden-

tification of an insect ecdysis triggering hormone (ETH)

receptor.

Materials and methods

Oligonucleotide primers were designed on the basis of the proposed

exons of gene CG5911(www.flybase.org). The primers used to clone

Biochemical and Biophysical Research Communications 299 (2002) 924–931

www.academicpress.com

BBRC

qThe nucleotide sequences reported in this paper have been

submitted to the GenBank Data Bank with Accession Nos. AF505863

and AF505864.* Corresponding author. Fax: +45-35-32-12-00.

E-mail address: [email protected] (C.J.P. Grim-

melikhuijzen).

URL: http://www.zi.ku.dk/cellbiology/

0006-291X/02/$ - see front matter � 2002 Elsevier Science (USA). All rights reserved.

PII: S0006 -291X(02 )02798 -5

CG5911-A were: sense, 50-GCCACGAGATGTGCAAGGCTGTG-30

and antisense, 50-CCAGGAAGCTTTTCTCGACATTTTGTTGG-30

(corresponding to nucleotide positions 242–264 and 1345–1373 of Fig.

2) and the primers to clone CG5911-B were: sense, as previously

mentioned and antisense, 50-GGCCCAGACAAGTGAGCAGCAG-30

(corresponding to nucleotide positions 997–1018 of Fig. 3). A mix of

cDNA from second and third instar larvae, pupae, and adult Dro-

sophila melanogaster was used as a template. The PCR parameters

were: 94 �C for 3 min, then touchdown PCR for 5 cycles, 94 �C for 30 s,

(for the -A splicing variant) 70 �C (for the -B splicing variant) 71 �C for

45 s, decreasing 2 �C for (-A) during 2 and (-B) during 3 cycles and

68 �C for 3min followed by 30 cycles of 94 �C for 30 s, (-A) 60 �C (-B)

61 �C for 45 s, 68 �C for 3min, and a final extension step of 68 �Cfor 10min. The reactions were carried out using the Advantage2

PCR enzyme system (Clontech). 50-RACE was made with the anti-

sense primer: 50-GGCCACCCACAGAATGGGGCTCGTAAAG-30

followed by the nested antisense primer, 50-GAACGGCACAGCCTT

GCACATCTCGTGG-30 (corresponding to nucleotide positions 426–

453 and 243–270 of Fig. 2) using the Herculase hotstart enzyme system

(Stratagene). The PCR programme was: 94 �C for 3min, then touch-

down PCR for 10 cycles, 94 �C for 30 s, 75 �C for 45 s, decreasing

1 �C each cycle and 72 �C for 3min followed by 40 cycles of 94 �C for

30 s, 65 �C for 45 s, 72 �C for 3min, and a final extension step of 72 �Cfor 10min. For the 30-RACE, the following primers were used for

CG5911-A: sense, 50-GGGGTTTCAAGCGGCTTTGTCAGG-30 fol-

lowed by the nested sense primer: 50-GGTGCAGCGAGGATATCAG

TCGC-30 (corresponding to nucleotide positions 908–931 and 1109–

1131 of Fig. 2) and for CG5911-B: sense, 50-GGCTCCTCACAGGTG

CCCAGCACCAAGG-30 followed by the nested sense primer: 50-GA

GGATGTCGAGGGTCTGGGCATTGCCGG-30 (corresponding to

nucleotide positions 691–718 and 859–887 of Fig. 3). The PCR

parameters were: 94 �C for 3min, then touchdown PCR for 5 cycles,

94 �C for 30 s, (-A) 70 �C (-B) 79 �C for 45 s, decreasing 2 �C for 2 cycles

and 68 �C for 3min followed by 30 cycles of 94 �C for 30 s, (-A)

60 �C (-B) 69 �C for 45 s, 68 �C for 3min, and a final extension step of

68 �C for 10min using the Advantage2 PCR enzyme system (Clon-

tech). The SMART RACE cDNA kit (Clontech) was used for the

RACE reactions. The vector pCR4-TOPO (Invitrogen) was used for

the cloning.

DNA sequence compilation as well as nucleotide and amino acid

sequence alignments were performed using the Lasergene DNA Soft-

ware package (DNASTAR). The prediction server TMHMM (v. 2.0)

from the Center for Biological sequence analysis, BioCentrum-DTU

(www.cbs.dtu.dk), was used for analysis of the secondary structure of

the receptor proteins.

Chinese hamster ovary (CHO) cells were grown as described

earlier [7]. To amplify the coding region of the D. melanogaster ETH

receptor gene, the following primers were used for CG5911-A: sense,

50-GAGATGCTGCCACAGATTCCATC-30 and antisense, 50-TTAG

AGGGTATTTTCGGTCAGATTTACTA-30 (corresponding to nu-

cleotide positions 1–20 and 1388–1416 of Fig. 2) and for CG5911-B:

sense, as previously mentioned and antisense, 50-TCAGATCTTGC

TGGCGCGTCG-30 (corresponding to nucleotide positions 1366–

1386 of Fig. 3). The same template was used as before. The PCR

programme was: 94 �C for 3min, then touchdown PCR for 6 cycles,

94 �C for 30 s, 68 �C for 45 s, decreasing 2 �C for 2 cycles and 72 �C for

3min followed by 30 cycles of 94 �C for 30 s, 56 �C for 45 s, 72 �C for

3min, and a final extension step of 72 �C for 10min using the Her-

culase enzyme system and then cloned into the vector pCR3.1 (In-

vitrogen). The insertions were fully sequenced and mutations were

Fig. 1. Organization of the corrected Drosophila gene CG5911 and its two splicing variants. The exons (italic numbers) are presented as bars, being

broad for the coding and small for the noncoding regions. The introns (roman numbers) are presented as lines and are not drawn to scale. (A)

Splicing variant CG5911-A. (B) Splicing variant CG5911-B. (C) The corrected gene CG5911.

Table 1

Intron/exon boundaries of the corrected Drosophila gene CG5911

Intron 50-Donor Intron size (bp) 30-Acceptor Intron phase

1 CTG gtaagtggt. . . 10756 . . .ccgtttcag CCA –

2 ATG gtgagtcct. . . 447 . . .ccgttacag GTG 3

Met Val

3 T gtgagtggc. . . 73 . . .ctatcccag GC 1

Cys Cys

4 AG gtaagcgat. . . 1932 . . .cccgagcag C 2

Ser Ser

4+5+5 AG gtaagcgat. . . 4132 . . .ttcttgcag T 2

Ser Ser

The intron and exon positions are explained in Fig. 1. Exon 5 can either be an exon (and be part of the receptor CG5911-A cDNA) or an intron

(when the receptor CG5911-B cDNA is formed). In this last situation, a large intron is created consisting of intron 4 plus exon 5 plus intron 5 (4 +5

+5).

A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931 925

corrected using the QuikChange XL Site-Directed Mutagenesis kit

(Stratagene) generating pCR3.1/CG5911-A/-B, which were transfected

into CHO cells [7]. The bioluminescence assay was performed as

described by [7,8].

Northern blots were carried out as in [9]. The cDNA probes used

corresponded to nucleotide positions 908–1416 (Fig. 2) or 481–1350

(Fig. 3). Peptides were obtained from Bachem (Bubendorf, Switzer-

land) or GeneMed Synthesis (San Francisco).

Fig. 2. cDNA and deduced amino acid sequence of the splicing variant CG5911-A. Nucleotides are numbered from 50- to 30-end and the amino acid

residues are numbered starting with the first ATG codon in the open reading frame. The four introns are indicated by arrows and the exon nu-

cleotides, bordering these introns, are highlighted in grey. The seven membrane spanning domains are boxed and labelled TM I–VII. The translation

termination codon is indicated by an asterisk. In-frame stop codons in the 50-noncoding region are underlined. The putative polyadenylation signal in

the 30-noncoding region is underlined twice.

926 A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931

Results

The website of the Drosophila Genome Project con-

sortium (www.flybase.org) contains the annotated geneCG5911, which is assumed to code for a G protein-

coupled receptor. We constructed primers based on the

predicted exons of this gene, performed PCR, and found

that the first predicted exon contained an intron

(thereby yielding two exons) and that another predicted

Table 3

Nucleotide differences between the cDNA of Fig. 3 and the corre-

sponding genomic sequences from the Berkley ‘‘Drosophila Genome

Project’’

Position of

the nucleotide

in the cDNA

Type of

nucleotide in

the gene

Type of

nucleotide in

the cDNA

Change in

amino acid

455 C G Thr! Ser

462 C T Ser! Ser

564 C T Phe! Phe

732 A G Gln! Gln

735 A G Gln! Gln

798 C T Ser! Ser

930 A G Leu! Leu

936 A G Ser! Ser

954 C G Leu! Leu

1158 A C Arg! Arg

1224 T C Gly! Gly

1248 G C Ser! Ser

1296 G A Glu! Glu

1347 A C Thr! Thr

Table 2

Nucleotide differences between the cDNA of Fig. 2 and the corre-

sponding genomic sequences from the Berkley ‘‘Drosophila Genome

Project’’

Position of the

nucleotide in

the cDNA

Type of

nucleotide

in the gene

Type of

nucleotide in

the cDNA

Change in

amino acid

)805 A G –

)773 C T –

)668 A C –

309 C T Ile! Ile

939 G T Gly! Gly

955 T C Leu! Leu

963 A G Thr! Thr

1337 C G Ala! Gly

1807 G T –

1905–1911 – TAGGCCG –

1993 A G –

2057–2058 CATACC CC –

2151–2152 ATGTAC AC –

Fig. 3. cDNA and deduced amino acid sequence of the splicing variant CG5911-B (the presentation is the same as in Fig. 2). Only the last exon (exon

6) is given in this figure, which is following exon 4 (see Fig. 1). The two upper lines (until intron 4) are identical to that shown in Fig. 2.

A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931 927

exon was an intron. Based on these and numerous otherPCRs and 50- and 30-RACEs, we concluded that the

corrected gene CG5911 has six exons (designated 1–6)

and five introns (Fig. 1, Table 1). We also found that the

gene gives rise to two splicing variants, which we named

CG5911-A, containing exons 1–5 and CG5911-B, con-

taining exons 1–4, and 6 (Fig. 1).

The complete cDNA of the CG5911-A splicing var-

iant is shown in Fig. 2. It has a polyadenylation site atits 30-end and several in-frame stop codons preceding the

first ATG (start) codon in its untranslated 50-region. The

cDNA codes for a protein of 471 amino acid residues.

This protein has seven transmembrane domains, sug-

gesting that it is a G protein-coupled receptor.

Fig. 3 shows the exon no. 6 (following exon 4, see Fig.1) of the CG5911-B splicing variant (the exons 1–4 of

this variant, preceding intron 4, are the same as in Fig.

2). Also, this CG5911-B cDNA has a polyadenylation

signal in its 30 untranslated region. The complete cDNA

codes for a protein of 461 amino acid residues long,

which, again, has seven transmembrane domains, sug-

gesting that it is a G protein-coupled receptor.

Comparison of the cDNAs from Figs. 2 and 3 withthe corresponding genomic sequences from the Dro-

sophila Genome Project database (>www.flybase.org)

revealed a small number of nucleotide differences. Only

in two cases do these differences lead to a change in

amino acid residues (Tables 2 and 3).

Fig. 4. Bioluminescence responses of nontransfected CHO cells and of two permanently transfected and cloned cell lines expressing either the

CG5911-A or -B receptor splicing variants. The SEMs are given as vertical bars, which are sometimes smaller than the symbols. In these cases only

the symbols are given. The responses 0–5 s (black), 5–10 s (grey), or 10–15 s (white) after addition of the peptides are given. (A) Responses of

nontransfected cells to 10�5 M Drm-ETH-1. (B) Responses of nontransfected cells to 10�5 M Drm-ETH-2. (C) Responses of CHO cells, expressing

CG5911-A, to 10�5 M Drm-ETH-1. (D) Responses of CHO cells, expressing CG5911-A, to 10�5 M Drm-ETH-2. (E) Dose–response curves of Drm-

ETH-1 and -2 for the activation of CG5911-A (EC50 of Drm-ETH-1 is 2:0� 10�7 M; EC50 of Drm-ETH-2 is 1:8� 10�6 M). (F) Same as (A). (G)

same as (B). (H) Responses of CHO cells, expressing CG5911-B to 10�5 M Drm-ETH-1 (I) Responses of CHO cells, expressing CG5911-B to 10�5 M

Drm-ETH-2. (J) Dose–response curves of Drm-ETH-1 and -2 for the activation of CG5911-B (EC50 of Drm-ETH-1 is 3:7� 10�8 M; EC50 for Drm-

ETH-2 is 1:6� 10�7 M). The following peptides did not activate the two receptors in concentrations up to 10�6 or 10�5 M: crustacean cardioactive

peptide; corazonin; drostatins-A4, -B2, -C; Drosophila capa-1, -2, and -3; Drosophila tachykinin-3; Drosophila small neuropeptide F-1; Drosophila

adipokinetic hormone; Drosophila myosuppressin; Drosophila hug c; Drosophila pigment dispersing hormone; Drosophila pyrokinin-2; FMRFamide;

leucokinin III; leucomyosuppressin; leucopyrokinin; proctolin. For peptide structures see [2,3,17–22].

928 A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931

We stably transfected Chinese hamster ovary (CHO)cells with DNA coding for either the CG5911-A or -B

receptor. These CHO cells were also stably expressing

the promiscuous G protein, G-16 [8]. Two days before

the bioassay, we transiently transfected these cells with

DNA coding for the protein apoaquorin and 3 h before

the assay, we added coelentrazine to the cell culture

medium. Addition of receptor ligands to these pre-

treated cells would lead to an IP3=Ca2þ-mediated bio-

luminescence response that could easily be measured

[5–8,10].

We tested a peptide library, consisting of Drosophila

and other invertebrate hormonal peptides, on the pre-

treated cells, expressing either the CG5911-A or -B re-

ceptor. Of all peptides tested, only the Drosophila ETH-1

and -2, which play a central role in molting (cuticle

shedding) of Drosophila [11], elicited a bioluminescence

response (the EC50 of ETH-1 was 2:0� 10�7 M with

CG5911-A and 3:7� 10�8 M with CG5911-B; the EC50

of ETH-2 was 1:8� 10�6 M with CG5911-A and 1:7�10�7 M with CG5911-B; Fig. 4). These results identified

the two receptors as Drosophila ETH receptors.

We carried out Northern blots of various develop-

mental stages of Drosophila, using probes specific for

each of the two receptor splicing variants, and found

that the two receptors were differently expressed.

CG5911-A is mainly expressed in larvae, whereas

CG5911-B is mainly expressed in embryos and pupae(Fig. 5). The sizes of the transcripts corresponded well

with the sizes of our cloned cDNAs, but the CG5911-B

probe also hybridized with a second, much larger tran-

script from adult flies, of which the origin was unknown

(Fig. 5B).

Searching of the databases from GenBank and the

malaria mosquito Anopheles gambiae Genome Project

[12] revealed the existence of a putative Anopheles Gprotein-coupled receptor (agCG48355) that showed a

remarkable sequence identity with the two Drosophila

ETH receptors (75% identical and 86% similar amino

acid residues between CG5911-A and the corresponding

Anopheles receptor) (Fig. 6). In addition to the common

structural features of the receptor proteins, their genes

are also similarly organized in that they have all three

introns in the coding region at identical positions andwith identical intron phasings (Fig. 6). Furthermore,

Fig. 6. Amino acid sequence comparison between the Drosophila ETH receptor splicing variant CG5911-A (DER-A), -B (DER-B) and two putative

splicing variants from the putative Anopheles ETH receptor (AER-A and -B) encoded by the gene agCG48355. Spaces are introduced to optimize

alignment. Amino acid residues that are identical among all four splicing variants are highlighted in grey. The seven membrane spanning domains

(for DER-A) are indicated by TM I–VII. Common introns in the genes are indicated by vertical boxes. Intron 4 in CG5911 is located in TM IV and is

followed by either exon 5 (DER-A) or exon 6 (DER-B). Note that the Anopheles receptor gene has two exons that are very similar to the exons 5 and

6 (Fig. 1) of the Drosophila receptor gene, suggesting that Anopheles, like Drosophila, has two splicing variants.

Fig. 5. Northern blots of different developmental stages from Dro-

sophila. Each vertical lane contained 2.5lg mRNA from embryos,

larvae, pupae, adult body (thorax/abdomen), and adult head (mixed

sexes). The numbers at the right give estimated mRNA sizes in kilo-

base. (A) Hybridization with a cDNA probe specific for CG5911-A.

(B) The blot of A was stripped and hybridized with a cDNA probe

specific for CG5911-B. (C) The blot of B was stripped and hybridized

with a cDNA probe specific for ribosomal protein 49. This blot gives

the loading efficiency.

A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931 929

there are equivalents for CG5911 exon 5 and exon 6(Fig. 1) in the Anopheles gene, suggesting that this gene

also creates two splicing variants (Fig. 6). All these data

indicate that the Anopheles receptor is a mosquito ETH

receptor.

Finally, Fig. 6 shows that the regions of 5911-A and -B

that correspond to exons 5 and 6 also have numerous

sequence identities, suggesting that these exons have

arisen by exon duplication.

Discussion

Insects and other arthropods have an external skele-

ton (cuticle) that they need to exchange during growth

or metamorphosis. This shedding of the old cuticle is

called molting or ecdysis. Because of the very largenumbers of insects living on earth, molting is one of the

most commonly performed behaviours on our planet

[13]. Molting is initiated and regulated by a hormonal

peptide, ecdysis triggering hormone (ETH) produced by

small organs, the epitracheal glands, which are situated

near the openings of the trachea of insects [13–15]. The

first insect ETH was isolated from the moth Manduca

sexta [14]. Subsequently, an ETH preprohormone wascloned from Drosophila, which contained two Drosoph-

ila ETH peptides, Drm-ETH-1 (DDSSPGFFLKITKN

VPRLamide) and Drm-ETH-2 (GENFAIKNLKTIP

RIamide) [11]. In the present paper, we have cloned two

receptor splicing variants (CG5911-A and -B) that,

when expressed in CHO cells, could be activated by low

concentrations of Drm-ETH-1 and -2 (Fig. 4). Drm-

ETH-1 was the most potent peptide in activating the tworeceptors (EC50 for CG5911-A, 2:0� 10�7 M; EC50 for

CG5911-B, 3:7� 10�8 M). These EC50 values compare

very well with the physiological ETH concentrations in

the haemolymph, at which ETH receptors are expected

to act, which for M. sexta is around 1:8� 10�7 M [14].

They are also in agreement with ecdysis bioassays in

Drosophila, where EC50 values of around 5� 10�8 M

can be estimated for Drm-ETH-1 [11]. These findingssuggest that Drm-ETH-1 is the cognate ligand for the

two receptors under physiological conditions. Whether

Drm-ETH-2 under these conditions also acts as a ligand

for the two receptors is uncertain, since this peptide has

a 10–20 times lower potency than Drm-ETH-1 to acti-

vate the two receptors (Fig. 4). Because Drm-ETH-1

and -2 are both present as one copy on the Drm-ETH

prohormone [11], it can be expected that they are bothreleased into the haemolymph and that they arrive at a

similar concentration at their target organs (among

them the central nervous system that initiates ecdysis

behaviour [14,15]). Under these conditions Drm-ETH-1

will already have activated the two receptors before the

Drm-ETH-2 reaches a concentration high enough for

activation (Fig. 4).

High affinity receptors specific for ETH have not beencloned from any insect (or invertebrate) before. Our pa-

per, therefore, is the first report on the molecular identi-

fication of insect ETH receptors. These results will enable

other researchers to clone ETH receptors from other

model insects used in molting research, such as M. sexta

and Bombyx mori [13–16]. That this is a feasible option is

illustrated by our finding that the malaria mosquito

Anopheles gambiae contains a receptor that is closely re-lated to the Drosophila ETH receptors (both with respect

to amino acid sequence and gene structure) and that most

likely is anAnophelesETH receptor (Fig. 6). This work on

other model insects, together with the more genetically

oriented work in Drosophila (e.g., knock-outs), will cer-

tainly lead to a better understanding of molting, which is

such a basic process for insects and other arthropods.

Acknowledgments

We thank Drs. S. Rees and J. Stables (Glaxo Wellcome, Stevenage,

UK) for supplying cell line CHO/G16, Birgitte Paulsen for typing the

manuscript, and Lundbeck Foundation, the Danish Natural Science

Research Council (equipment grant), and Novo Nordisk Foundation

for financial support.

References

[1] M.D. Adams et al., Science 287 (2000) 2185–2195.

[2] G. G€aade, K.H. Hoffmann, J.H. Spring, Physiol. Rev. 77 (1997)

963–1032.

[3] L. Schoofs, D. Veelaert, J. Vanden Broeck, A. De Loof, Peptides

18 (1997) 145–156.

[4] R.S. Hewes, P.H. Taghert, Genome Res. 11 (2001) 1126–1142.

[5] G. Cazzamali, N.P.E. Saxild, C.J.P. Grimmelikhuijzen, Biochem.

Biophys. Res. Commun. 298 (2002) 31–36.

[6] A. Iversen, G. Cazzamali, M. Williamson, F. Hauser, C.J.P.

Grimmelikhuijzen, Biochem. Biophys. Res. Commun. 299 (2002)

628–633.

[7] F. Staubli, T.J.D. Jørgensen, G. Cazzamali, M. Williamson, C.

Lenz, L. Søndergaard, P. Roepstorff, C.J.P. Grimmelikhuijzen,

Proc. Natl. Acad. Sci. USA 99 (2002) 3446–3451.

[8] J. Stables, A. Green, F. Marshall, N. Fraser, E. Knight, M. Sautel,

G. Milligan, M. Lee, S. Rees, Anal. Biochem. 252 (1997) 115–126.

[9] F. Hauser, L. Søndergaard, C.J.P. Grimmelikhuijzen, Biochem.

Biophys. Res. Commun. 249 (1998) 822–828.

[10] G. Cazzamali, C.J.P. Grimmelikhuijzen, Proc. Natl. Acad. Sci.

USA 99 (2002) 12073–12078.

[11] Y. Park, D. Zitnan, S.S. Gill, M.E. Adams, FEBS Lett. 463 (1999)

133–138.

[12] R.A. Holt et al., Science 298 (2002) 129–149.

[13] K.A. Mesce, S.E. Fahrbach, Front. Neuroendocrinol. 23 (2002)

179–199.

[14] D. Zitnan, T.G. Kingan, J.L. Hermesman, M.E. Adams, Science

271 (1996) 88–91.

[15] J.W. Truman, Science 271 (1996) 40–41.

[16] D. Zitnan, L. Hollar, I. Spalovska, P. Takac, I. Zitnanova, S.S.

Gill, M.E. Adams, J. Exp. Biol. 205 (2002) 3459–3473.

[17] L. Kean, W. Cazanave, L. Costes, K.E. Broderick, S. Graham,

V.P. Pollock, S.A. Davies, J.A. Veenstra, J.A.T. Dow, Am. J.

Physiol. Regul. Integr. Comp. Physiol. 282 (2002) R1297–R1307.

930 A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931

[18] J. Vanden Broeck, Peptides 22 (2001) 241–254.

[19] C. Lenz, M. Williamson, C.J.P. Grimmelikhuijzen, Biochem.

Biophys. Res. Commun. 273 (2000) 1126–1131.

[20] M. Williamson, C. Lenz, �AA.M.E. Winther, D.R. N€aassel, C.J.P.

Grimmelikhuijzen, Biochem. Biophys. Res. Commun. 281 (2001)

544–550.

[21] M. Williamson, C. Lenz, �AA.M.E. Winther, D.R. N€aassel, C.J.P.Grimmelikhuijzen, Biochem. Biophys. Res. Commun. 282 (2001)

124–130.

[22] X. Meng, G. Wahlstr€oom, T. Immonen, M. Kolmer, M. Tirronen,

R. Predel, N. Kalkkinen, T.I. Heino, H. Sariola, C. Roos, Mech.

Dev. 117 (2002) 5–13.

A. Iversen et al. / Biochemical and Biophysical Research Communications 299 (2002) 924–931 931