Biochemical and Biophysical Research Communications 309 (2003) 485–494

www.elsevier.com/locate/ybbrc

BBRC

Molecular cloning, functional expression, and gene silencing of twoDrosophila receptors for the Drosophila neuropeptide pyrokinin-2q

Carina Rosenkilde,a Giuseppe Cazzamali,a Michael Williamson,a Frank Hauser,a

Leif Søndergaard,b Robert DeLotto,b and Cornelis J.P. Grimmelikhuijzena,*

a Department of Cell Biology, Zoological Institute, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen, Denmarkb Department of Genetics, Institute of Molecular Biology, University of Copenhagen, Øster Farimagsgade 2A, DK-1353 Copenhagen, Denmark

Received 4 August 2003

Abstract

The database of the DrosophilaGenome Project contains the sequences of two genes, CG8784 and CG8795, predicted to code for

two structurally related G protein-coupled receptors. We have cloned these genes and expressed their coding parts in Chinese

hamster ovary cells. We found that both receptors can be activated by low concentrations of the Drosophila neuropeptide pyrokinin-

2 (CG8784, EC50 for pyrokinin-2, 1� 10�9 M; CG8795, EC50 for pyrokinin-2, 5� 10�10 M). The precise role of Drosophila py-

rokinin-2 (SVPFKPRLamide) in Drosophila is unknown, but in other insects, pyrokinins have diverse myotropic actions and are

also initiating sex pheromone biosynthesis and embryonic diapause. Gene silencing, using the RNA-mediated interference tech-

nique, showed that CG8784 gene silencing caused lethality in embryos, whereas CG8795 gene silencing resulted in strongly reduced

viability for both embryos and first instar larvae. In addition to the two Drosophila receptors, we also identified two probable

pyrokinin receptors in the genomic database from the malaria mosquito Anopheles gambiae. The two Drosophila pyrokinin receptors

are, to our knowledge, the first invertebrate pyrokinin receptors to be identified.

� 2003 Elsevier Inc. All rights reserved.

The recent publications of the genomes from the

fruitfly Drosophila melanogaster [1] and the malaria

mosquito Anopheles gambiae [2] represent a break-through in insect research, because they enable us to

identify proteins that play a key role in the physiology or

behavior of insects. Our research group is especially

interested in neuropeptide receptors and their ligands,

because these proteins (and peptides) occupy a high hi-

erarchical position in the physiology of insects and steer

important processes, such as reproduction, develop-

ment, feeding, and behavior.

qThe nucleotide sequences in this paper have been submitted to the

GenBank/EBI Data Bank with Accession Nos. AY277898, AY277899,

BK001383, and BK001384.* Corresponding author. Fax: +45-35321200.

E-mail address: cgrimmelikhuijzen@zi.ku.dk (C.J.P. Grim-

melikhuijzen).

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

0006-291X/$ - see front matter � 2003 Elsevier Inc. All rights reserved.

doi:10.1016/j.bbrc.2003.08.022

The database of the Drosophila Genome Project con-

tains a list of 40–45 genes annotated to code for G pro-

tein-coupled neuropeptide receptors (www.flybase.org)[3]. In many cases, however, these annotations have

turned out to be incorrect, because the predicted intron/

exon organizations were wrong, or because other anno-

tated neighboring genes were also part of the correct re-

ceptor gene [4–9]. Furthermore, the ligands for most

annotated receptor genes are unknown, i.e., they are or-

phan receptors. Therefore, proper cDNA cloning of the

annotated receptor genes, expression of the cDNA incells, and subsequent ligand identification, using the or-

phan receptor or reverse pharmacology strategies [10],

are still necessary processes. In the present paper, we have

cloned the cDNAs corresponding to two related anno-

tated receptor genes, CG8784 and CG8795, expressed

them in Chinese hamster ovary (CHO) cells, and identi-

fied the Drosophila neuropeptide pyrokinin-2 as their

cognate ligand. The two Drosophila pyrokinin-2 recep-tors are, to our knowledge, the first invertebrate pyrok-

inin receptors to be identified.

486 C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494

Materials and methods

Total RNA from D. melanogaster (Canton S) was isolated using

TRIzol reagent (Invitrogen). mRNA was isolated with the lMACS

mRNA Isolation kit (Miltenyi Biotech). cDNA was synthesized using

Thermoscript reverse transcriptase (Invitrogen). PCR was performed

using the Advantage2 PCR enzyme system (Clontech) or the Hercu-

lase Hotstart DNA polymerase (Stratagene). Rapid amplification of

cDNA ends (RACE) was done using the SMART RACE cDNA kit

(Clontech).

From the genomic sequence of the gene CG8784 (www.flybase.org)

the following primers were designed and used for PCR: sense 50-GG

CGAGTGCGCAATTTTTGCCTGTCG-30 and 50-CTTCACGGA

CGCCATGTGCAT-30 (corresponding to nucleotide positions )210 to)185 and 525 to 545 of Fig. 1) and antisense 50-GCCCATGATGC

ACATGGCGTCCG-30 and 50-GCGTTGTTGTTTACACTGCAAC

TGG-30 (corresponding to nucleotide positions 530 to 552 and 1319 to

1343 of Fig. 1). 30-RACE was done with the sense primer 50-CCT

TCAACGATTACTTCCGGATACTTGATTACACG-30 and the

nested sense primer 50-CGGGGTGCTCTACTTTCTCTCCACC-30

(corresponding to nucleotide positions 1082 to 1116 and 1119 to 1143

of Fig. 1). From the genomic sequence of the gene CG8795

(www.flybase.org) the following primers were designed and used for

PCR: sense 50-GCCTAAGTCGAAATTGATGAGCC-30 and 50-GGA

CCTCTATAACCTCTGGCACCC-30 (corresponding to nucleotide

positions )169 to )147 and 354 to 377 of Fig. 2) and antisense

50-CGTTCGACTGTGAACGCGGTAATG-30 and 50-CGTTGTCGT

TCCATTGCCACTGC-30 (corresponding to nucleotide positions 459

to 482 and 1238 to 1260 of Fig. 2). For 30-RACE the sense primer

50-TTCCCAGGCGATGTTACGATGTAAACC-30 (corresponding to

nucleotide positions 776 to 802 of Fig. 2) was used. All PCR products

were cloned into pCR4-TOPO (Invitrogen) using the TOPO TA

cloning kit (Invitrogen) and sequenced.

The entire coding region of CG8784 was amplified using the sense

primer 50-AAGATGTTGCAAGGCGTCGCCATCACC-30 (the un-

derlined part corresponds to nucleotide positions 1 to 24 of Fig. 1) and

the antisense primer 50-TCACTGGGATTTGGGTGAGCCATA

GGT-30 (corresponding to nucleotide positions 1951 to 1977 of Fig. 1).

The entire coding region of CG8795 was amplified using the sense

primer 50-AAGATGCTGCCCACTAACAGTTCCGGTG-30 (the un-

derlined part corresponds to nucleotide positions 1 to 25 of Fig. 2) and

the antisense primer 50-TTAAAAGGCGGCCCGCTCTTCAGTG

TCG-30 (corresponding to nucleotide positions 1761 to 1788 of Fig. 2).

The PCR products were cloned into the expression vector pCR3.1

(Invitrogen) using the TA Cloning Kit (Invitrogen) and sequenced.

CHO cells stably expressing the human G-protein G-16 (CHO/G16)

were grown and transfected as described previously [11,12]. The bio-

luminescence assay was performed as described in [11,12].

To synthesize dsRNA, the following PCR-primers containing a 27

nucleotide long T7 RNA polymerase promoter sequence at their 50

ends were constructed: for CG8784 the sense primer 50-TAATA

CGACTCACTATAGGGAGACCACCATTTGGCTGGCGGCCTT

CC-30 (corresponding to nucleotide positions 693 to 712 of Fig. 1) and

the antisense primer 50-TAATACGACTCACTATAGGGAGACC

ACGCGTTGTTGTTTACACTGCA-30 (corresponding to nucleotide

positions 1324 to 1343 of Fig. 1), for CG8795 the sense primer 50-TAA

TACGACTCACTATAGGGAGACCACCGGAAACGGCGGCCA

ATGCG-30 (corresponding to nucleotide positions 428 to 447 of Fig. 2)

and the antisense primer 50-TAATACGACTCACTATAGGGAG

ACCACGGCAGATCCATTTAGCCCAG-30 (corresponding to

nucleotide positions 1214 to 1233 of Fig. 2). As negative control, a

700 bp fragment of a Hydra magnipapillata gene with a, so far, un-

known function was amplified using the sense primer 50-TAATACGA

CTCACTATAGGGAGACCACGCGTTCGGGGGGTGGTATGG-

30 and the antisense primer 50-TAATACGACTCACTATAGGGA

GACCACGGACCAGGGATCGAACTTGC-30. One microgram of

each PCR product was used as template for RNA synthesis using the

mCAP RNA capping kit (Stratagene). The dsRNA was phenol–chlo-

roform extracted, precipitated, and dissolved at 5 lM in 0.1mM

sodium phosphate (pH 7.8), 5mM KCl.

One-hour-old embryos were harvested, dechorionized using 2%

sodium hypochlorite, positioned on a glass slide, desiccated for 6min

at 24% relative humidity in a box containing Drierite (Fluka), and

covered with Voltalef 10S oil. dsRNA was injected into the posterior

end of the embryos using glass electrodes. The glass slides were placed

on apple juice-agar plates supplemented with yeast. After two days the

number of hatched larvae was counted. The larvae were further ob-

served through their larval stages and finally the number of emerging

pupae was counted.

Software programs, data processing, and peptide syntheses were as

in [8,13]. The Drosophila pyrokinins-1 and -2 were synthesized by

Genemed Synthesis (San Fransisco).

Results

The Drosophila Genome Project database contains

the sequences of the genes CG8784 and CG8795, which

were annotated to code for two related G protein-cou-

pled neuropeptide receptors (www.flybase.org) [3]. We

designed primers based on the annotated exons of the

two genes and performed PCR and RACE to obtain the

corresponding cDNAs. Fig. 1 shows the cDNA of geneCG8784. It is 2514 nucleotides long and contains two

overlapping, putative polyadenylation sites in its un-

translated 30-region and several stop codons preceding

the start codon in its untranslated 50-region. The cDNA

codes for a protein of 658 amino acid residues (Fig. 1).

There are seven transmembrane helices, suggesting that

the protein is a G protein-coupled receptor. The extra-

cellular N terminus of the protein contains four putativeN-glycosylation sites.

A comparison of the cDNA of Fig. 1 with the ge-

nomic sequence of the annotated gene CG8784 reveals

six exons and five introns (Table 1) and shows, there-

fore, that the original prediction of the intron/exon or-

ganization by the Drosophila Genome Project was

correct. This comparison also reveals deletions of 8

nucleotides and 15 other nucleotide differences betweenthe genomic sequence of the database and the cDNA.

These differences lead to a deletion of two Gly residues

and to three different amino acid residues in the protein,

of which one is a conserved residue (see GenBank

Accession No. AY277898).

Fig. 2 shows the cDNA of gene CG8795, which is

2228 nucleotides long. It contains a polyadenylation site

in its untranslated 30-region and several stop codonspreceding the translation start codon in its untranslated

50-region. The cDNA codes for a protein of 595 amino

acid residues, which contains seven transmembrane

helices, again suggesting that it is a G protein-coupled

receptor. As with the first protein (Fig. 1), this protein

contains four putative glycosylation sites in its extra-

cellular N terminus (Fig. 2).

Fig. 1. cDNA and deduced amino acid sequence of the gene CG8784. Nucleotides are numbered from the 50- to the 30-end, the amino acid residues

are numbered from the first ATG codon in the open reading frame. The stop codons in the 50-untranslated region are underlined, the two putative,

overlapping polyadenylation sites in the 30-untranslated region are underlined twice. The translation stop codon is indicated by an asterisk. The

introns are indicated by arrows and numbered 1–5, the exon nucleotides bordering the introns are highlighted in grey. The transmembrane domains

are indicated by boxes and numbered TM I–VII. The potential glycosylation sites (following the NXS/T consensus sequence) are indicated by filled

triangles.

C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494 487

Fig. 2. cDNA and deduced amino acid sequence of the gene CG8795. The figure is presented in the same way as Fig. 1.

488 C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494

A comparison of the cDNA of Fig. 2 with the ge-

nomic sequence of gene CG8795 reveals the existence of

seven exons and six introns (Table 2), which was also

predicted by the Drosophila Genome Project (www.fly-

base.org). This comparison also showed 10 nucleotide

differences, which lead to two different non-conserved

Table 1

Intron/exon boundaries of the CG8784 gene

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

1 G gtgagat.. 60 ..ctgcag GC 1

Gly Gly

2 CG gtgagt.. 1254 ..tttgtag G 2

Arg Arg

3 ACG gtggg.. 71 ..cttag ATG 3

Thr Met

4 G gtaggtg.. 1055 ..ttgcag TA 1

Val Val

5 AAG gtaag.. 195 ..ttaag ATC 3

Lys Ile

Table 2

Intron/exon boundaries of the CG8795 gene

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

1 G gtgagtg.. 247 ..tcttag GA 1

Gly Gly

2 AG gtaagg.. 79 ..gatacag G 2

Arg Arg

3 ACG gtgcg.. 317 ..tgtag ATG 3

Thr Met

4 G gtaagca.. 58 ..tcccag TG 1

Val Val

5 AAG gtgag.. 156 ..tttag GTA 3

Lys Val

6 ACG gtgag.. 245 ..tgcag GAC 3

Thr Asp

C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494 489

amino acid residues (see GenBank Accession No.

AY277899).

We stably expressed the cDNAs of the two orphan

receptors in CHO cells that also were stably expressing

the a subunit of the promiscuous G protein G-16 [11].

Furthermore, these cells were transiently transfected

with DNA coding for the protein apoaequorin. Three

hours before the assay (see below), coelenterazine wasadded to the culture medium. Activation of the heter-

ologously expressed receptors in the pretreated CHO

cells would result in an IP3/Ca2þ-mediated biolumines-

cence response that could easily be measured and

quantified [11,12].

We tested a peptide library, consisting of 23 Dro-

sophila or other invertebrate neuropeptides, and seven

monoamines on the transfected CHO cells and foundthat both receptors were strongly activated by low

concentrations of Drosophila pyrokinin-2 (Drm-PK-2;

SVPFKPRLamide) [14]. Of both receptors, receptor

CG8795 had the highest affinity for Drm-PK-2 (EC50,

5� 10�10 M), whereas the affinity of receptor CG8784 to

Drm-PK-2 was somewhat less (EC50, 1� 10�9 M)

(Fig. 3). The insect pyrokinins are myotropic neuro-

peptides that have a wide range of actions. Structurally,they are characterized by the C-terminal sequence

FXPRLamide [14–19]. There exists a second pyrokinin

in Drosophila, Drm-PK-1 (TGPSASSGLWFGPRLa-

mide) [19], which, however, only weakly activated the

two receptors (Table 3). There occur other neuropep-

tides in Drosophila that have the C-terminal sequence

PRLamide in common with Drm-PK-1 and -2, but that

are not pyrokinins. Two of these peptides, hug-c [14]

and Drosophila ecdysis triggering hormone-1 (Drm-

ETH-1) [20] did also activate the two receptors, albeit

with lower affinities (Table 3). All other Drosophila

neuropeptides tested and also the seven monoamines did

not activate the two receptors (Fig. 3 and Table 3).

Amino acid sequence alignments showed that the two

Drm-PK-2 receptors are clearly structurally related

(Fig. 4). Furthermore, the two receptors are also struc-

turally related with a third G protein-coupled receptor

encoded by the Drosophila gene CG9918 (www.fly-

base.org) [3]. Comparison of the two Drm-PK-2 recep-tors with the proteins found in the genomic database

from the malaria mosquito Anopheles gambiae [2] re-

vealed two proteins that also strongly resembled the

Drm-PK-2 receptors (Fig. 4). Comparison of the geno-

mic organizations of the five above-mentioned proteins

showed that the two Drm-PK-2 receptors had five in-

trons in common (with identical intron phasings). Four

of these introns are also shared by one of the two pu-tative Anopheles receptors (APR-2), while the other

putative Anopheles receptor (APR-1) and the putative

Drosophila CG9918 receptor share two introns with the

Fig. 3. Bioluminescence responses of non-transfected CHO/G-16 cells, and of two permanently transfected cell lines, expressing either the CG8784 or

CG8795 receptor protein. 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), and 10–15 s (white) after addition of the peptides are shown. (A) Responses of CHO/G-16 cells to

2.5� 10�7 M Drm-PK-2. (B) Responses of CHO/G-16/CG8784 cells to 2.5� 10�7 M Drm-PK-2. (C) Responses of CHO/G-16/CG8784 cells to

2.5� 10�7 M hug-c. (D) Dose–response curve of the effects of Drm-PK-2 and hug-c on CHO/G-16/CG8784 cells. The EC50 for Drm-PK-2 is

1� 10�9 M, that for hug-c is 3� 10�8 M. (E) Responses of CHO/G-16 cells to 2.5� 10�7 M Drm-PK-2. (F) Responses of CHO/G-16/CG8795 cells to

2.5� 10�7 M Drm-PK-2. (G) Responses of CHO/G-16/CG8795 cells to 2.5� 10�7 M hug-c. (H) Dose–response curves of the effects of Drm-PK-2

and hug-c on the CHO/G-16/CG8795 cells. The EC50 for Drm-PK-2 is 5� 10�10 M, that for hug-c is 7� 10�9 M. In addition to Drm-PK-2 and hug-c,the two receptors are also activated, but much less potently, by Drm-ETH-1 (both receptors EC50, 2� 10�7 M), but not by Drm-ETH-2. Drosophila

capa-3 (also called myotropin, or Drm-PK-1) only slightly activates the CG8795 receptor, but is without effect (tested up to 10�5 M) on the CG8784

receptor. The following peptides did not activate the receptor (tested up to 10�6 or 10�5 M): crustacean cardioactive peptide; capa-1 and -2; co-

razonin; Drosophila adipokinetic hormone; Drosophila tachykinin-3; Drosophila short neuropeptide F-1; Drm-ETH-2; Drosophila myosuppressin;

Drosophila pigment dispersing hormone; drostatins-A4, -B2, and -C; FMRFamide; perisulfakinin; Heliothis zea hypertrehalosemic neuropeptide;

leucomyosuppressin; leucokinin-III; and proctolin. For peptide structures see [16–18]. The following amines did not activate the receptors (tested up

to 10�5 M): adrenaline; dopamine; histamine; noradrenaline; octopamine; serotonin; and tyramine.

Table 3

Amino acid sequences of some structurally related insect neuropeptides and their potencies to activate the CG8784 and CG8795 receptors

Name Structure Species EC50 (M) with the

CG8784 receptor

EC50 (M) with the

CG8795 receptor

Drm-PK-2 SVPFKPRLamide D. melanogaster 1� 10�9 5� 10�10

Hug-c pQLQSNGEPAYRVRTPRLamide D. melanogaster 3� 10�8 7� 10�9

Drm-ETH-1 DDSSPGFFLKITKNVPRLamide D. melanogaster 2� 10�7 2� 10�7

Drm-ETH-2 GENFAIKNLKTIPRIamide D. melanogaster NA NA

Drm-PK-1 (Drm-myotropin, capa-3) TGPSASSGLWFGPRLamide D. melanogaster NA >5� 10�6

Capa-1 GANMGLYAFPRVamide D. melanogaster NA NA

Capa-2 ASGLVAFPRVamide D. melanogaster NA NA

Leucopyrokinin (Lem-PK) pQTSFTPRLamide L. maderae 6� 10�8 6� 10�9

NA, not active in concentrations up to 10�5 M.

490 C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494

two Drm-PK-2 receptor genes (Fig. 4). All these com-

parisons clearly indicate that the five receptors are both

evolutionarily and structurally closely related. The gene

structures of the two Drm-PK-2 receptors, and one of

the putative Anopheles receptors (APR-2) are strikingly

similar, strongly suggesting that the Anopheles protein

Fig. 4. Amino acid sequence comparisons between the two Drm-PK-2 receptors (gene products of CG8784 and CG8795), two putative Anopheles

pyrokinin receptors (APR-1 with GenBank Accession No. BK001383; APR-2 with Accession No. BK001384), and the gene product of the annotated

Drosophila gene CG9918. Amino acid residues that are identical between the CG8784 gene product and at least one of the other proteins are

highlighted in grey. The seven membrane spanning helices are indicated by TM I–VII.

C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494 491

also is a pyrokinin receptor. The same, however, might

also be true for APR-1 and the CG9918 receptor.

To understand the function and importance of the

two Drm-PK-2 receptors, we created their loss offunction phenotypes by the RNA-mediated interference

(RNAi) gene silencing technique [21]. We used two pipet

diameter sizes, by which we either injected a smaller

(�0.5 fmol) or larger (�2.5 fmol) amount of double

stranded RNA (dsRNA) into fertilized Drosophila eggs.

Injection of the smaller amount of dsRNA corre-

sponding to the CG8784 gene gave 19� 6% hatching of

the embryos (49� 2% in the controls). From the hatched

animals 21� 1% survived the first instar larval stage

(69� 3% in the controls) (Table 4). After injection of thelarger amount of dsRNA, the success of hatching was

reduced to 0� 0%, showing that the CG8784 loss of

function phenotype is embryonic lethal. For the CG8795

gene, the small amount of injected dsRNA gave 25� 7%

hatching of the embryos and from these hatched animals

29� 6% survived the first instar larval stage. For the

Table 4

Embryonic and first instar larval survival rate after injection of embryos with two doses (small, 0.5 fmol; and large, 2.5 fmol) of dsRNA from

different sources

Injected Hatchings (%) First instar larval survival (%) Number of eggs injected (times repeated)

Hydra dsRNA (small) 49� 2 69� 3 30 (2)

Hydra dsRNA (large) 50� 2 63� 13 30 (2)

CG8784 dsRNA (small) 19� 6 21� 1 40 (3)

CG8784 dsRNA (large) 0� 0 0� 0 30 (3)

CG8795 dsRNA (small) 25� 7 29� 6 20 (3)

CG8795 dsRNA (large) 8� 1 11� 19 30 (3)

The SEM is given. All receptor dsRNA values (lines 3–6) are significantly different (0:001 < P < 0:05) from the controls (lines 1–2).

492 C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494

larger amount, these numbers were 8� 1% and

11� 19%. The loss of function phenotype for the

CG8795 gene, therefore, is reduced embryonic and lar-

val survival. All the above-mentioned numbers were

significantly different from the ones obtained after in-

jection of embryos with unrelated (non-Drosophila)dsRNA (Table 4). Unfortunately, we were unable to see

consistent anatomical defects in the embryos or larvae

that could cause the lethality or reduced viability after

silencing of the two genes.

Discussion

In the present study we have cloned two Drosophila

receptors that could be activated by very low concen-

trations of Drm-PK-2 (SVPFKPRLamide). The EC50 of

CG8784 to Drm-PK-2 was 1� 10�9 M, whereas that of

CG8795 was 5� 10�10 M (Fig. 3). Because these EC50

values are so low, the two receptors are likely to be the

physiologically relevant Drm-PK-2 receptors. The insect

pyrokinins, to which Drm-PK-2 belongs, are charac-terized by the C-terminal FXPRLamide sequence

[16–18]. There exists another pyrokinin in Drosophila,

Drm-PK-1 (TGPSASSGLWFGPRLamide), also called

Drm-myotropin or capa-3 [17,19], which, however, does

not, or only weakly, activate the two receptors (Table 3).

In contrast, the Drosophila neuropeptide hug-c [14],

which is not a pyrokinin, but which has the C-terminal

sequence PRLamide in common with Drm-PK-2, is stillable to activate the two receptors, albeit at 14–30 times

higher concentrations (Fig. 3 and Table 3). The same is

true for Drm-ETH-1, although 200–400 times higher

concentrations are needed here. This last peptide,

therefore, might not be a physiologically relevant ligand

for the two receptors, also because another receptor

exists in Drosophila that has a much higher affinity to

Drm-ETH-1 than the two Drm-PK receptors [6]. Theremaining question, therefore, is in how far hug-c is a

cognate ligand for the two receptors. In this context it is

important to realize that hug-c and Drm-PK-2 are both

contained as one copy within the same prohormone, and

that the peptidergic neurons, producing the prohor-

mone, are probably releasing the two peptides in equi-

molar amounts [14]. If both peptides would be equally

stable (peptidase-resistant) after their release, then the

two Drm-PK-2 receptors would already be fully acti-

vated by Drm-PK-2, before hug-c could reach concen-

trations high enough to activate them (Figs. 3D and H).

However, it is possible that under certain conditions(tissue surroundings or transport in the hemolymph)

hug-c is more stable than Drm-PK-2, because it con-

tains an N-terminal <Glu group, which protects it

against unspecific aminopeptidases (Table 3). In these

cases hug-c could eventually also become a ligand for

the two Drm-PK-2 receptors. This situation would be

very interesting, because a non-pyrokinin neuropeptide

would then be the natural ligand for a pyrokinin re-ceptor. Such a finding would question the pyrokinin

nomenclature.

The CG8784 and CG8795 receptor cDNAs have re-

cently also been cloned by Park et al. [22], although only

their coding regions were determined. Moreover, these

authors characterized the two receptors after expression

in frog oocytes. They found that the CG8784 receptor

could only be activated by very high (non-physiological)concentrations of Drm-PK-2 and hug-c (both peptides

having EC50 values higher than 10�5 M), making it very

questionable what the cognate ligand was [22]. Fur-

thermore, they found that the CG8795 receptor was

more sensitive to Drm-PK-2 and hug-c, but also here,

their EC50 values could not be determined, because of

rapid receptor desensitization in the oocyte expression

system. This example illustrates that proper receptorligand identification is very much dependent upon the

choice of the expression system—some insect receptors

are better and more functionally expressed in the frog

oocyte system, others do better in mammalian cell lines.

We feel that the critical factor in these expression

systems is probably the G protein.

The insect pyrokinins are important neuropeptides

with very diverse functions. The first pyrokinin wasisolated from the cockroach Leucophea maderae and this

peptide (leucopyrokinin, see Table 3) obtained its name,

because of its myostimulatory (“kinin”-like) action on

the cockroach hindgut and a special structural feature,

namely the presence of a pyroglutamate group at its N

terminus [18,23]. Since then, numerous other pyrokinins

C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494 493

(all characterized by the FXPRLamide C-terminal se-quence) have been isolated from other insect groups,

and even from crustaceans, where they, again, stimulate

visceral muscle contractions, but also initiate sex pher-

omone biosynthesis, induce cuticle coloration, or lead to

embryonic diapause [15,24–29]. In most insects, several

pyrokinins occur. From the American cockroach, e.g.,

six such peptides have been isolated [30]. As mentioned

earlier, at least two pyrokinins exist in Drosophila, Drm-PK-1 and -2, each encoded by a different gene (Table 3)

[14,17,19]. The exact functions of Drm-PK-1 and -2 are

unknown. It has been found that Drm-PK-2 (at 10�8–

10�7 M) increases heartbeat in semi-isolated Drosophila

heart preparations and that overexpression of the Drm-

PK-2 gene (also encoding hug-c) disrupts normal larval

ecdysis [14]. This last finding, however, could be the

consequence of high Drm-PK-2 and hug-c concentra-tions, which could interfere with one of the other pep-

tide systems (e.g., Drm-ETH, see Table 3). We have

used another, and in our view, a more specific strategy

and down-regulated the two Drm-PK-2 receptors, using

the RNAi technique [21]. Gene silencing of the Drm-

PK-2 receptor CG8784 causes embryonic lethality,

whereas silencing of the CG8795 gene significantly

lowers both embryonic and first instar larval survival(Table 4). These findings, therefore, strongly suggest

that Drm-PK-2 is essential for Drosophila embryonic

development.

Insect researchers have paid much attention to PBAN

(pheromone biosynthesis activating neuropeptide),

which belongs to a group of structurally related pyr-

okinins isolated from a variety of moths (Lepidoptera),

and which action is to initiate sex pheromone biosyn-thesis in adult, sexually mature female lepidopterans,

thereby attracting remote males for mating [24,25]. The

PBAN prohormone has been cloned from several le-

pidopterans and it contains, in addition to one copy of

PBAN, one copy of a pyrokinin diapause hormone

(inducing diapause in lepidopteran embryos), and three

other pyrokinins having other physiological functions

[31–35]. Many attempts have been made to characterizethe PBAN receptors and to synthesize PBAN mimetics,

in order to develop new insecticides and interfere with

mating and other physiological processes in lepidopter-

ans [29,36–38]. The molecular identification of the first

insect (and first invertebrate) pyrokinin receptors in

Drosophila may now pave the way to clone all other

pyrokinin receptors in other model insects, including the

PBAN receptors in lepidopterans. That this is feasible isillustrated by our finding of probable pyrokinin recep-

tors in the malaria mosquito Anopheles (Fig. 4). Fur-

thermore, we have found that one of the cockroach

pyrokinins (leucopyrokinin) also activates the two

Drosophila receptors (Table 3), showing that the cock-

roach receptors also must resemble their counterparts in

Drosophila. The identification of pyrokinin receptors in

Drosophila and other model insects will certainly furtherour understanding of important processes in insects,

such as reproduction, development, and diapause. In

addition, the availability of cloned pyrokinin receptors

functionally expressed in cell lines (grown in 96- or 384-

well plates in cell culture), will enable researchers to

screen large chemical libraries for the presence of high-

affinity antagonists. This work might finally lead to the

development of new, insect-specific and environmentallysafe insecticides.

Acknowledgments

We thank Dr. S. Rees and Dr. J. Stables (Glaxo Wellcome, Ste-

venage, UK) for supplying cell line CHO/G-16, Dr. G€oosta Nachman

(University of Copenhagen) for discussing the statistics of Table 4,

Birgitte Paulsen for typing the manuscript, and the Lundbeck Foun-

dation and Fabrikant Vilhelm Pedersen og Hustrus Mindelegat (the

latter was granted on recommendation from the Novo Nordisk

Foundation) for financial support.

References

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

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

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

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

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

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

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

628–633.

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

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

924–931.

[7] G. Cazzamali, F. Hauser, S. Kobberup, M. Williamson, C.J.P.

Grimmelikhuijzen, Biochem. Biophys. Res. Commun. 303 (2003)

146–152.

[8] K. Egerod, E. Reynisson, F. Hauser, G. Cazzamali, M. William-

son, C.J.P. Grimmelikhuijzen, Proc. Natl. Acad. Sci. USA 100

(2003) 9808–9813.

[9] K. Egerod, E. Reynisson, F. Hauser, M. Williamson, G. Cazza-

mali, C.J.P. Grimmelikhuijzen, Biochem. Biophys. Res. Commun.

306 (2003) 437–442.

[10] O. Civelli, H.-P. Nothacker, Y. Saito, Z. Wang, S.H. Lin, R.K.

Reinscheid, Trends Neurosci. 24 (2001) 230–237.

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

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

[12] 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.

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

USA 99 (2002) 12073–12078.

[14] 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.

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

18 (1997) 145–156.

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

963–1032.

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

[18] D.R. N€aassel, Prog. Neurobiol. 68 (2002) 1–84.

494 C. Rosenkilde et al. / Biochemical and Biophysical Research Communications 309 (2003) 485–494

[19] L. Kean, W. Cazenave, L. Costes, K.E. Broderick, S. Graham,

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

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

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

133–138.

[21] S.M. Hammond, A.A. Caudy, G.J. Hannon, Nat. Rev. Genet. 2

(2001) 110–119.

[22] Y. Park, Y.J. Kim, M.E. Adams, Proc. Natl. Acad. Sci. USA 99

(2002) 11423–11428.

[23] G.M. Holman, B.J. Cook, R.J. Nachman, Comp. Biochem.

Physiol. C 85 (1986) 219–224.

[24] A.K. Raina, H. Jaffe, T.G. Kempe, P. Keim, R.W. Blacher, H.M.

Fales, C.T. Riley, J.A. Klun, R.L. Ridgway, D.K. Hayes, Science

244 (1989) 796–798.

[25] A. Kitamura, H. Nagasawa, H. Kataoka, T. Inoue, S. Matsum-

oto, T. Ando, A. Suzuki, Biochem. Biophys. Res. Commun. 163

(1989) 520–526.

[26] K. Imai, T. Konno, Y. Nakazawa, T. Komiya, M. Isobe, K.

Koga, T. Goto, T. Yaginuma, K. Sakakibara, K. Hasegawa, O.

Yamashita, Proc. Jpn. Acad. Ser. B 67 (1991) 98–101.

[27] S. Matsumoto, A. Kitamura, H. Nagasawa, H. Kataoka, C.

Orikasa, T. Mitsui, A. Suzuki, J. Insect Physiol. 36 (1990) 427–432.

[28] P. Torfs, J. Nieto, A. Cerstiaens, D. Boon, G. Baggerman, C.

Poulos, E. Waelkens, R. Derua, J. Calderon, A. De Loof, L.

Schoofs, Eur. J. Biochem. 268 (2001) 149–154.

[29] A. Rafaeli, Int. Rev. Cytol. 213 (2002) 49–91.

[30] R. Predel, M. Eckert, J. Comp. Neurol. 419 (2000) 352–

363.

[31] M.T.B. Davis, V.N. Vakharia, J. Henry, T.G. Kempe, A.K.

Raina, Proc. Natl. Acad. Sci. USA 89 (1992) 142–146.

[32] T. Kawano, H. Kataoka, H. Nagasawa, A. Isogai, A.

Suzuki, Biochem. Biophys. Res. Commun. 189 (1992) 221–

226.

[33] Y. Sato, M. Oguchi, N. Menjo, K. Imai, H. Saito, M. Ikeda, M.

Isobe, O. Yamashita, Proc. Natl. Acad. Sci. USA 90 (1993) 3251–

3255.

[34] P.W.K. Ma, D.C. Knipple, W.L. Roelofs, Proc. Natl. Acad. Sci.

USA 91 (1994) 6506–6510.

[35] L. Duportets, C. Gadenne, F. Couillaud, Peptides 20 (1999)

899–905.

[36] M. Altstein, O. Ben-Aziz, S. Daniel, I. Zeltser, C. Gilon, Peptides

22 (2001) 1379–1389.

[37] M. Altstein, Biopolymers 60 (2001) 460–473.

[38] P.E. Teal, R.J. Nachman, Peptides 23 (2002) 801–806.