| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Functional Genomics and Proteomics Unit (T.J., E.M., M.L., K.V., S.J.H., L.T., L.S.), Department of Biology, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium; and Areawide Pest Management Research (R.J.N.), Southern Plains Agricultural Research Center, United States Department of Agriculture, College Station, Texas 77845
Address all correspondence and requests for reprints to: Tom Janssen, Functional Genomics and Proteomics Unit, Department of Biology, Katholieke Universiteit Leuven, Naamsestraat 59, B-3000 Leuven, Belgium. E-mail: Tom.Janssen{at}bio.kuleuven.be.
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
-amylase). They also act as satiety promoters in the brain, regulating food intake (1, 2, 3). The actions of CCK/gastrin are mediated by two specific G protein-coupled receptors (GPCRs), named CCK1R (previously CCK-AR) and CCK2R (previously CCK-BR) (2, 4). CCK1R is mainly localized in peripheral organs and discrete areas of the brain, whereas CCK2R is primarily expressed in the brain and the stomach. Although CCK1R only shows high specificity for sulfated CCK, CCK2R has the same affinity for both gastrin and CCK, and does not discriminate between sulfated and unsulfated CCK (5, 6, 7). Both CCK/gastrin and their receptors are highly conserved in vertebrates.
Many species of invertebrates, including cockroaches, flies, beetles, moths, locusts, crickets, shrimp, and lobsters, express sulfated neuropeptides called sulfakinins (SKs). Members of the SK family typically contain the C-terminal hexapeptide Y(SO3H)GHMRFamide and exhibit several biological effects. They increase the frequency and amplitude of foregut (8) and hindgut contractions (9, 10, 11, 12), and also increase the frequency of heartbeat in the cockroach (11, 13). Comparable to the satiety effect of CCK observed in mammals, insect SKs can also reduce food intake in locusts (14, 15), cockroaches (8), and crickets (16), and can inhibit carbohydrate feeding in the fly Phormia regina (17, 18). In addition, SK can stimulate the release of the digestive enzyme
-amylase in the beetle Rhynchophorus ferrugineus (19) and in the moth Opisina arenosella (20). The observed distribution of immunoreactivity in the nervous system of insects and in crustaceans suggests that SKs may function as central neurotransmitters or neuromodulators (16, 21, 22, 23). In cockroaches and locusts, the expression near neurohemal release sites implies that they could also be released into the hemolymph and act as hormones (24, 25). The strong similarities in structure and biological activities between SKs and CCK/gastrins has led many researchers to hypothesize that SKs constitute the functional arthropod homologs of the vertebrate CCK/gastrin peptides, and have evolved from a common peptide ancestor (8, 9, 10, 12, 15, 26, 27, 28).
Additional evidence in favor of this hypothesis can be found in the peptide receptors. Johnsen (1) already postulated that the two CCK/gastrin receptors likely evolved from a common ancestor, based on the high degree of mutual similarity. Hewes and Taghert (29) later suggested that the Drosophila GPCRs CG6857 and CG6881 diverged from a common ancestor of the CCK and gastrin family of receptors. One of these receptors, designated DSK-R1, was deorphanized by Kubiak et al. (30) and found to interact with several insect SKs. Unsulfated SK was 3000 times less potent than the sulfated form, indicating that the sulfate moiety is important for receptor activation. This is consistent with the observation for CCK1R activation in mammals. Thus, the CCK/gastrin and SK signaling system represents a good example of coevolution of neuropeptides and their receptors (26).
So far, SK peptides have only been isolated from arthropods. Although CCK/gastrin-like immunoreactivity has been detected in the nervous system of annelids (31), snails (14), and nematodes, no CCK/gastrin-related peptide has been identified so far in these phyla.
Twenty-three years ago, McIntire and Horvitz (32) already reported the presence of CCK-like immunoreactivity in neuronal cells of Caenorhabditis elegans. A year later, also in Ascaris suum, CCK-like immunoreactivity was observed (33). Regrettably, despite many efforts, the identity of the nematode CCK-like peptide(s) has remained a mystery ever since (34, 35, 36, 37).
The C. elegans genome contains two genes that have been identified as potential homologs of the mammalian CCK/gastrin receptors and their invertebrate SK receptor counterparts (30, 38, 39). In this study we have cloned and expressed these receptors, and used them in a reverse pharmacology approach to identify their ligand(s). In this way we could identify two peptides encoded by the neuropeptide-like protein-12 gene (nlp-12) with a high degree of sequence similarity to CCK/gastrin and the SKs. We also identified the nematode nlp-12 expressing cells in vivo, and demonstrate that this novel nematode neuropeptide signaling system has biological functions similar to those of CCK/gastrin in vertebrates and SKs in arthropods.
| Materials and Methods |
|---|
|
|
|---|
Molecular cloning
The open reading frame (ORF) of each receptor gene was obtained by reverse transcriptase-PCR. mRNA, extracted from mixed stage C. elegans N2 (QuickPrep Micro mRNA Purification Kit; Amersham Biosciences, Diegem, Belgium), was used as a template for cDNA synthesis (SuperScript First-Strand Synthesis System for RT-PCR; Invitrogen, Carlsbad, CA). The full-length cDNA was then amplified by PCR using gene-specific oligonucleotide primers (Sigma-Genosys, Haverhill, UK) based on the predicted cDNA sequence (WormBase release 135, www.wormbase.org). All PCR primers used are listed in supplemental Table 2. A partial Kozak sequence (CACC) was also incorporated immediately preceding the authentic initiation codon to optimize initiation of translation. The resulting PCR product of each receptor (isoform) was cloned directly into the eukaryotic expression vector pcDNA3.1D (pcDNA3.1 Directional TOPO Expression Kit; Invitrogen) and sequenced.
Cell culture and transfections
Chinese hamster ovary (CHO) cells (CHO-K1), stably overexpressing the mitochondrially targeted apo-aequorin (mtAEQ) and the human G
16 subunit, were used for the Ca2+ measurements and cultured in Hams F12 medium (BioWhittaker Inc., Walkersville, MD/Cambrex Corp., East Rutherford, NJ) containing 10% fetal bovine serum, 100 UI/ml penicillin/streptomycin, 250 µg/ml Zeocin (InvivoGen, San Diego, CA), and 2.5 µg/ml Fungizone (Amphoterin B; Invitrogen). CHO cells stably overexpressing the mtAEQ and lacking the human G
16 subunit were used for dose-response analysis. Cell lines were split every 3 d (1:10) and grown at 37 C in a humidified atmosphere of 5% CO2 in air. CHO/mtAEQ/G
16 or CHO/mtAEQ cells were transiently transfected with the receptor cDNA constructs using the FuGene 6 reagent (Roche, Brussels, Belgium), according to the manufacturers instructions.
Bioluminescence assay
Intracellular calcium was monitored as previously described (40). Briefly, cells expressing the receptor were collected 2 d after transfection in BSA medium (DMEM/HAMs F12 with 15 mM HEPES, without phenol red, supplemented with 0.1% BSA) and loaded with 5 µM coelenterazine h (Invitrogen) for 4 h to reconstitute the holo-enzyme aequorin. After a 10-fold dilution, cells (25,000 per well) were exposed to potential peptide ligands, i.e. HPLC fractions (one twentieth) or synthetic peptides reconstituted in BSA medium. The calcium response was recorded for 30 sec on a Microlumat Plus, LB96V microplate luminometer (EG&G Berthold, Bad Wildbad, Germany) in triplicate. After 30 sec, Triton X-100 (0.1%) was added to the same well as a positive control and as a measure of the total cell Ca2+ response. BSA medium was used as a negative control, and 1 µM ATP was used to check the functional response. Cells transfected with an empty pcDNA3.1D vector were used as a negative control. EC50 values were calculated from dose-response curves, constructed using a computerized nonlinear regression analysis, with a sigmoidal dose-response equation (Sigmaplot 8.0; Systat Software, Inc., San Jose, CA).
Peptides
Based on in silico predictions and in-house peptidomics data (41), a library of 156 synthetic C. elegans peptides was composed and custom synthesized by Sigma-Genosys. See supplemental Table 1 (published as supplemental data on The Endocrine Societys Journals Online web site at http://endo.endojournals.org) for the entire list. All sulfated peptides were synthesized in-house via solid-phase chemistry using Fmoc methodology and Rink Amide resin on an ABI 433A peptide synthesizer according to previously described methods (42). The purified peptides were sulfated by exposure to concentrated sulfuric acid according to previously described procedures (10, 43). The identity of the sulfated peptide analogs was confirmed by matrix-assisted laser desorption ionization (MALDI) time-of-flight (TOF) mass spectrometry (MS). All peptides were initially tested at a concentration of 10 µM. Receptor-activating peptides were purified further using reversed-phase HPLC and quantified with the bicinchoninic acid protein assay (44).
Peptide extraction and HPLC fractionation
Peptide extraction from four 650 ml liquid cultures (Fernbach flasks) of mixed stage wild-type N2 worms was performed as previously described (Janssen, T., S. J. Husson, M. Lindemans, K. Verstraelen, E. Meelkop, L. Temmerman, R. Rademakers, I. Mertens, M. Nitabach, G. Jansen, and L. Schoofs, submitted for publication). The resulting aqueous phase was desalted by solid-phase extraction (MegabondElute; Varian, St-Katelijne-Waver, Belgium). The peptides eluted by 0–60% CH3CN in 0.1% trifluoroacetic acid (TFA) were further fractionated by reversed-phase HPLC on a Delta-pack C18 [2x (25 x 100 mm)] column (Waters, Zellik, Belgium) with a solvent flow rate of 12 ml/min. After injection, a wash step of 10 min using 0.1% aqueous TFA was initiated, followed by a linear gradient over 80 min to a final concentration of 60% CH3CN in 0.1% aqueous TFA. Twelve-milliliter fractions were collected automatically every minute starting from the beginning of the gradient.
MALDI-TOF MS and electrospray ionization (ESI)-double-quadrupole-TOF tandem mass spectrometry (MS/MS)
An aliquot (100 µl) of each HPLC fraction was lyophilized, resuspended in 4 µl water/CH3CN/formic acid (50/49.9/0.1) and subjected to MALDI-TOF MS as previously described (46). The fraction(s) containing the NLP-12 peptides was subjected to ESI double-quadrupole TOF MS/MS as previously described (27) to identify the posttranslational modifications of the endogenous peptides (Waters-Micromass Ltd., Manchester, UK).
Dot blot analysis
A 10-fold dilution series (1–0.01 mg/ml) of each synthetic peptide was lyophilized, redissolved in H2O, applied (2 µl) onto a nitrocellulose membrane, and dried at 120 C for 30 min. The membrane was then blocked with 3% milk powder in Tris-buffered saline (TBS) [10 mM Tris, 6 mM NaCl, and 0.05% Triton X-100 (pH 7.6)]. After 1 h at room temperature, the membrane was incubated overnight in TBS with 1:500 of a polyclonal anti-CCK antibody [raised in rabbit against synthetic sulfated Homo sapiens CCK (26, 27, 28, 29, 30, 31, 32, 33)] (CCK-8; DY[SO3H]MGWMDFamide; Sigma-Aldrich, Bornem, Belgium). Afterwards, the membrane was washed three times with TBS, followed by a 1-h incubation with 1:500 horseradish peroxidase-conjugated goat antirabbit IgG (Dako, Heverlee, Belgium). After three more washes with 10 mM Tris buffer (pH 7.6), detection was performed by adding 0.125% diaminobenzidine and 1:10,000 H2O2 (30%). Several non-NLP-12-like C. elegans peptides were used as a negative control, and synthetic sulfated and nonsulfated H. sapiens CCK-8 (Sigma-Aldrich) peptides were used as a positive control.
Analysis of the expression pattern of nlp-12 transgenes
To investigate the cell-specific expression of nlp-12 in vivo, we created C. elegans transgenes using an mCherry reporter construct that was generated by overlap extension PCR (47). PCR fragments encompassing the putative promoter region (376 bp between nlp-12 and the adjacent gene emr-1) and the ORF of nlp-12 extending into the second exon (176 bp) were fused to a PCR fragment containing mCherry (amplified from the pRSET-B mCherry vector, kindly provided by Roger Y. Tsien, University of California, San Diego, CA). Fusion PCR products were purified using the Qiaquick PCR Purification Kit (QIAGEN, Venlo, The Netherlands). Wild-type animals were microinjected with 50 ng/µl mCherry reporter construct and 30 ng/µl elt-2::GFP as a selection marker. Five independent transgenic lines were generated, and, except for variation in expression levels, all lines showed the same expression pattern. NLP-12::mCherry expression was visualized using the LSM510 multiphoton confocal microscope (Zeiss, Zaventem, Belgium), and cells were identified using a combination of their position and morphology (48). All PCR primers used are listed in supplemental Table 2.
Nile Red assay and quantification of fat content
Nile Red powder (N-1142; Invitrogen) was dissolved in acetone at 500 µg/ml. Immediately before the assay, Nile Red solution was diluted in 1x PBS and added on top of 9-cm NGM plates seeded with OP50 bacteria, to a final concentration of 0.05 µg/ml. Plates were then left at room temperature overnight before use. C. elegans larvae were synchronized by hypochlorite treatment, followed by overnight starvation on empty NGM plates, and then placed on Nile Red plates as starved L1 stage larvae. After 45-h incubation at 20 C, L4 stage-stained well-fed animals were washed three times and then dispensed (10 L4 per well) in V-bottom 96 wells (Greiner, Wemmel, Belgium) using the COPAS Biosorter (Union Biometrica, Geel, Belgium). After anesthetization with 5 mM NaN3 and centrifugation, the fluorescent intensity (corresponding to the fat content) of each well (10 worms) was measured from below using a GENius Plus plate reader (Tecan, Mechelen, Belgium) with excitation and emission wavelengths of 485 and 595 nm, respectively. All strains tested were cultivated and analyzed the same way and measured in octuplicate (eight wells of 10 L4 animals). Two independent experiments were performed.
Determination of pharyngeal pumping and defecation rates
For pharyngeal pumping rate, L4 well-fed animals raised at 20 C were recorded with a video camera, slowed down, and the number of pharyngeal contractions in each 30-sec interval was noted. The defecation cycle of L4 well-fed animals raised at 20 C was observed manually. For each animal (n = 10), the duration of five consecutive cycles was noted.
Cockroach hindgut assay
Leucophaea maderae cockroaches were taken from stock colonies maintained at 27 C and fed dry dog food ad libitum. Hindguts isolated from the central nervous system were dissected, immersed in saline, and prepared for recording of myogenic activity as previously described (49).
Amylase assay
L4 synchronized animals were subjected to 100 µM BODIPY FL conjugated DQ corn starch for 35 min in the presence of OP50 food (EnzChek Ultra Amylase Assay Kit; Invitrogen). Afterwards, the animals were anesthetized with 5 mM NaN3 and washed three times with M9 solution to eliminate background. Animals were then dispensed (20 L4 per well) in V-bottom 96 wells using the COPAS Biosorter, and the fluorescence was measured from below using a GENius Plus plate reader with excitation and emission wavelengths of 485 and 535 nm, respectively. All strains tested were cultivated and analyzed the same way and measured in quadruplicate (four wells of 20 L4 animals). Two independent experiments were performed.
Statistical analysis
The results of every quantitative assay were first checked for normality and homogeneity of variance (SAS 8.1; SAS Institute Inc., Cary, NC). The statistical significance was then determined using the Students t test (Sigmaplot 8.0).
| Results |
|---|
|
|
|---|
Molecular cloning of C. elegans CCK receptors
Sequence-specific primers were used to amplify the ORF of T23B3.4 and the two predicted splice isoforms encoded by Y39A3B.5 (WormBase release 135) by RT-PCR. The ORF of splice isoform Y39A3B.5a (CE21646) could not be amplified. However, amplification of the second splice isoform (Y39A3B.5b,
CE38985) resulted in two separate PCR products. The resulting PCR product(s) of T23B3.4 (1237 bp) and Y39A3B.5b (1200 and 1395 bp, respectively) were cloned directly into the eukaryotic expression vector pcDNA3.1D and sequenced. The sequence of T23B3.4 differs from the WormBase prediction (www.wormbase.org) because intron 5 encoded for an exon. This changes the reading frame of T23B3.4 and results in a stop codon at nucleotide position 682. The resulting protein is 168 amino acids shorter than the predicted one (227 AA vs. 395) and lacks the transmembrane domains needed to be catalogued as a functional 7TM receptor. We repeated the amplification (and cloning) starting from fresh mRNA, but with the same result, and so we choose not to pursue this gene any further. Recently, Y39A3B.5 was predicted to encode four instead of two isoforms [Y39A3B.5a (CE21646), b (CE38985), c (CE38443), and d (CE38986)]. The two sequences that we amplified from Y39A3B.5b both corresponded to a combination of the recently predicted isoforms of Y39A3B.5. The sequence of 1200 bp consists of the first eight exons of isoform c, followed by the last two exons of isoform b, and will be referred to as isoform Y39A3B.5c/b (Fig. 1
). The second sequence (1395 bp) consists of the first eight exons of isoform c, followed by the last four exons of isoform d, and will be referred to as isoform Y39A3B.5c/d. The deducted proteins of Y39A3B.5c/b and Y39A3B.5c/d, composed of 399 and 464 amino acids, respectively, both contain seven putative transmembrane domains (www.ch.embnet.org/software/TMPRED_form.html) and belong to the rhodopsin class of GPCRs. The two splice isoforms only differ from each other at the carboxy-terminal region starting from amino acid position 387 (results are shown in supplemental Fig. 1
, which is published as supplemental data on The Endocrine Societys Journals Online web site at http://endo.endojournals.org). Because McKay et al. (38) recently referred to the C. elegans CCK-like receptors as Ce_CKR-1 (T23B3.4) and CKR-2 (Y39A3B.5), we will hereafter refer to Y39A3B.5c/b as CKR-2a and Y39A3B.5c/d as CKR-2b. The sequence data have been submitted to the GenBank database under accession numbers EU346943 and EU346944, respectively.
|
|
16 and apo-aequorin) transiently expressing CKR-2a or b, the cells were challenged with a library of 156 synthetic C. elegans peptides, consisting of neuropeptides belonging to the established FLP and NLP families of peptides (supplemental Table 1). Five peptides were able to activate both receptors in the calcium bioluminescence assay (data not shown): FLP-1a (SADPNFLRF amide), NLP-14a (ALDGLDGSGFGFD), NLP-13c (pQPSYDRDIMSFamide), NLP-12a (DYR PLQFamide), and NLP-12b (DGYRPLQF amide). This activity was not seen in cells transfected with the empty pcDNA3.1D vector.
Dose-response tests reveal that the CKR-2 receptors can only be activated in a dose-dependent way by peptides NLP-12a (DYRPLQFamide) and NLP-12b (DGYRPLQFamide) (Fig. 3
). CKR-2a and CKR-2b both show a nanomolar affinity for these peptides. NLP-12b is the most potent ligand of CKR-2a (EC50 = 14.71 ± 0.04 nM) (Fig. 3A
), and CKR-2b shows the highest affinity for NLP-12a (EC50 = 56.75 ± 0.05 nM) (Fig. 3B
). FLP-1a, NLP-14a, and NLP-13c could only activate the receptors at high concentrations (
250 nM for NLP-13c, and
10 µM for FLP-1a and NLP-14a). To verify whether the unsulfated NLP-12 peptides indeed constitute the endogenous cognate ligands for these receptors, we also challenged the receptor-expressing cells with HPLC fractions (one fifteenth) of a whole body peptide extract of mixed stage C. elegans. Only fractions 36 and 37, which contain both NLP-12 peptides (as verified by MALDI-TOF MS), were able to elicit a calcium response (data not shown). The overall gene structure of nlp-12 contains three exons and two introns, and this is consistent with the exon/intron organization typical of all CCK-like genes (1).
|
Pharmacological characterization of CKR-2a and CKR-2b
To analyze the G protein signaling pathway, the receptors were tested for their ability to elicit a Ca2+ response in the bioluminescence assay, using CHO/mtAEQ cells expressing CKR-2a or b, but lacking G
16. Addition of NLP-12a or b resulted in a strong calcium response for both receptors, indicating that they signal through a G
q type of G
subunit in vivo.
The Drosophila DSK-R1 and most vertebrate CCK-B receptors show a more than 500-fold higher affinity for sulfated than unsulfated SK/CCK peptides (6, 30, 53). To verify whether this was also the case for CKR-2a, b, we tested and compared sulfated and unsulfated peptide analogs of NLP-12a and NLP-12b at different concentrations (Fig. 4
). Interestingly, both receptors displayed a slightly lower affinity for the sulfated peptides. In addition, a sulfated NLP-13c analog (pQPSY[SO3H]DRDIMSFamide) was tested, but with a similar result (data not shown). These data suggest that the native NLP-12 peptide ligands are likely to be unsulfated, and that the sulfate moiety is not essential and even slightly impedes CKR-2 receptor activation in C. elegans. The sulfated and unsulfated insect SKs and the human CCK-8 peptides were also tested in the bioluminescence assay at different concentrations. None of the C. elegans receptors could be activated by the insect SKs. However, the human CCK-8 peptides were able to mount a calcium response on the CKR-2 receptors. Interestingly, when we tested CCK-8 on CHO cells transfected with the empty pcDNA3.1D vector, we also got a strong response (the response is shown in supplemental Fig. 3
). The sulfated CCK-8 is much more potent than the unsulfated form (supplemental Fig. 3
), and the response could be inhibited by pretreatment of the cells with the CCKR antagonist proglumide (5 µM). This indicates that, in contrast with general knowledge and literature, an endogenous and functional CCK receptor could be present in CHO cells. The finding that human CCK-8 is unable to activate the C. elegans CK receptors is consistent with the findings for the Drosophila SK receptor (30).
|
|
|
Especially the tyrosine residue (Y) at C-terminal position 6 or 7 seems to be conserved perfectly during evolution, except for Trichostrongylus colubriformis CKI, B. malayi CKI/II, and Wuchereria bancrofti CKII/III. This predominant tyrosine (sulfated or unsulfated in vivo) is mostly preceded by an Asp residue or an Asp-Asp/Asp-Gly motif and followed by an Arg or Gly residue. In case of the vertebrate gastrins, the tyrosine residue is preceded by a Glu-Ala or Ala-Ala motif. The consistent C-terminal motif seems to be PLQFamide in all nematode CK (NLP-12)-like peptides, HM/LRFamide in all arthropod SKs, and WMDFamide in all vertebrate CCK/gastrins and C. intestinalis cionin. Methionine and leucine are very similar hydrophobic amino acids that are conserved between all clades at C-terminal position 3. A Met-Leu substitution in SKs or CCK has been used previously to improve chemical stability and did not influence the functional activity of these peptides (43, 60). The amino acid at position 2 from the C terminus [Gln (Q), Arg (R) or Asp (D)] is only weakly conserved.
Analysis of CK (NLP-12) expression pattern
Previous reports on nlp-12 expression are a little confusing and not in agreement with each other. In addition, the cells identities were not specified (54, 61). WormBase indicates that Li et al. (54) found nlp-12 expression in head neurons, whereas the actual paper only reports expression in tail neuron(s). Nathoo et al. (61) only report expression in one tail neuron but used a transcriptional fusion construct and left out the potential downstream regulatory sequences. Li et al. (54) did use a translational fusion construct but only included the initiator methionine, assuming that no intronic regulatory sequences were present. By using these methods, certain cells expressing nlp-12 could have been missed. To clarify this we examined the spatial expression of nlp-12 in C. elegans using a translational reporter fusion construct, including the potential intronic regulatory sequences. Five different transgenic lines containing the translational nlp-12::mCherry fusion construct were created and visualized using an LSM510 multiphoton confocal microscope (Zeiss). Cells were identified using a combination of their position and morphology. Fluorescent signals from NLP-12::mCherry could be observed in hermaphrodites throughout post-embryonic life. All transgenic lines showed consistent NLP-12::mCherry expression in one tail neuron with its cell body located in the dorso-rectal ganglion, which could be identified as the ring interneuron DVA. The fluorescence was also visible in the anteriorly directed process of DVA, running along the ventral nerve cord, toward and around the circumpharyngeal nerve ring (Fig. 7
, B and C). The expression of nlp-12 in the DVA tail neuron could be confirmed by immunostaining with an antibody directed against NLP-12a (Mousley, A., and A. Maule, unpublished data). They also stained Panagrellus redivivus and observed a similar staining pattern (approximate one tail neuron). In addition, expression was observed in all six coelomocytes (ccs), two ventral and posterior to the junction of the pharynx and intestine (Fig. 7A
). Another pair of ventral ccs is located near the vulva (Fig. 7B
), and the last pair is located more dorsally near the posterior bend of the gonad (Fig. 7C
). The fluorescent signals were clearly visible in the prominent cytoplasmic vesicles typical for these ccs. The cc expression could not be confirmed with the immunological method (Mousley, A., and A. Maule, unpublished data). For clarity, we also created 3-dimensional movies from confocal Z-stack projections of nlp-12::mCherry transgenic animals (see supplemental movies).
|
Isolated cockroach hindgut assay
SKs are known to display myotropic properties, increasing the frequency and amplitude of cockroach hindgut contractions (26), and also CCKs have been attributed a role in intestinal motility (1, 2). Therefore, we tested several concentrations (10–5 to 10–9 M) of the sulfated and unsulfated C. elegans CK peptides in a cockroach hindgut myotropic bioassay. SKs from P. americana and Drosophila melanogaster were used as a positive control. However, none of the Ce_CK peptides displayed any myotropic activity at any of the tested concentrations (data not shown).
Determination of pharyngeal pumping and defecation rates
Comparable to the satiety effect of CCKs in mammals (3, 62, 63), the SKs significantly reduce food intake in flies, locusts, crickets, and cockroaches (8, 14, 15, 16, 17, 64). To address the situation in C. elegans, we analyzed the pharyngeal pumping rate (as a measure of food intake) and defecation rate of L4 staged NLP-12 (CK) and CKR-2 mutants, and compared it with wild-type N2 nematodes. No significant difference in pharyngeal pumping rate, or in defecation rate, could be observed (Fig. 8
, A and B). These data indicate that the food intake and defecation rhythm are not affected by CK/CKR-2 signaling in C. elegans.
|
NLP-12 signaling through CKR-2 receptors disrupts fat storage in C. elegans
To analyze the fat storage, the intestinal fat content of L4 stage mutants and wild-type animals was visualized using the lipophilic vital dye Nile Red. nlp-12(ok335) and ckr-2(tm3082) animals both display an increased fat content relative to wild-type worms (Fig. 8C
). The Nile Red fluorescence in animals lacking NLP-12 (CK) is 86% higher than in wild-type animals [nlp-12(ok335) 1918 ± 164 relative fluorescent units (RFUs) per worm, n = 131; N2 1033 ± 84 RFUs per worm, n = 136) and 74% higher in animals lacking CKR-2 (1798 ± 134 RFUs per worm, n = 185]. A lack of CKR-2 or its NLP-12 ligands seems to result in an increase in fat storage.
| Discussion |
|---|
|
|
|---|
Our pharmacological results indicate that both CK receptors signal through a G
q type of G
subunit in vivo, which is in agreement with the Drosophila SK receptor and mammalian CCK receptors in which CCK-B signaling activates phospholipase C, triggering an increase in inositol 1,4,5-trisphosphate production (2, 30). Both C. elegans CK peptides were recognized by a human antibody raised against sulfated CCK-8 in a dot blot analysis. In addition, Drosophila DSK-II and P. americana SK-I could be detected in this way. Therefore, the CCK-like immunoreactivity observed in nematodes, flies, and cockroaches in the past can be attributed to the occurrence of CK-like peptides and SK peptides, respectively. Several nonrelated C. elegans RFamide peptides were used as a control and did not react with the CCK-8 antibody, indicating that CCK-like immunoreactivity was probably not caused by cross-reactivity with RFamide peptides.
Interestingly, the C. elegans CK receptors both show a slightly higher affinity for the unsulfated than the sulfated CK peptides. In addition, the unsulfated CK peptides displayed a higher immunoreactivity toward the CCK-8 antibody than the sulfated ones. The natural CK peptides also perfectly coeluted with the synthetic unsulfated peptides. Although the Asp-Tyr motif is a potential recognition site for sulfation (78), the present data suggest that the native CK (NLP-12) peptide ligands are likely to be unsulfated, and that the sulfate moiety is not essential and might even slightly impede CKR-2 receptor activation in C. elegans. In the case of DSK-1R and CCK1R, the sulfate moiety is essential for receptor activation (6, 7, 30). However, CCK2R does also preferentially bind to the unsulfated CCK/gastrin. In locusts and cockroaches, the sulfate moiety is required for food intake inhibition activity (8, 15, 26, 79) and myotropic activity, although its position within the sequence is not critical (26, 79). A growing number of studies now indicate that the unsulfated CCK/gastrins can also exert specific biological activities (80, 81, 82, 83, 84, 85). In fact, half of the endogenous gastrins are nonsulfated (2, 86, 87, 88, 89). Recently, unsulfated droSKs were also found to inhibit crop and anterior midgut contractions in Drosophila (90). The C. elegans CK peptides are highly conserved within at least three clades of nematodes, and are structurally very similar to the arthropod SKs and the vertebrate CCK/gastrin peptide family. If the nematode CK peptides are representative of the ancestral members of the CCK/gastrin peptide family, this would suggest that the common ancestral peptide was probably unsulfated. With this point of view, the sulfation of the tyrosyl residue could have occurred later in evolution, after the divergence of the nematode clade, and this both in protostomes and deuterostomes.
Additional evidence favoring the hypothesis of CCK/gastrin signal conservation in nematodes can be found in the peptide receptors. In general, our phylogenetic analysis demonstrates that all CCK/SK receptors share a very high degree of sequence similarity (>60%), not only within clusters, but also in between (Fig. 2
). Johnsen (1) already postulated that both mammalian CCK/gastrin receptors likely evolved from a common ancestor, based on the high degree of mutual similarity. The two Drosophila CCKR-like GPCRs CG6857 and CG6881 are also highly similar and encoded by two loosely linked genes (
30 kb apart). Therefore, Hewes and Taghert (29) suggested that they likely occurred through a gene duplication event originating from a common ancestor of the CCK and gastrin family of receptors. The two C. elegans CK receptors represent splice isoforms encoded by the same gene. Because growing evidence suggests that nematodes diverged from the evolutionary tree before the arthropods (91, 92, 93), this makes perfect sense. One common ancestral gene, encoding a single CCK receptor, could have evolved into two different splice isoforms in nematodes. With the divergence of insects, the single gene had already become two separate, but highly similar and loosely linked genes (only 30 kb apart). Moreover, in the vertebrate lineage, the highly similar receptor genes have already become completely separated (different chromosomes). Therefore, the duplication event probably occurred before the phylogenetic divergence of protostomes and deuterostomes.
Thus, the CCK signaling system could represent a good example of coevolution of neuropeptides and their receptors (26). In this respect it is interesting that these ancestral molecules in nematodes contain a neutral residue (Gln) in the C terminus that later diverged in two different directions in terms of molecular character. In insects this residue changed to a positively charged, basic residue (Arg), whereas in vertebrates it changed to the opposite polarity: a negatively charged, acidic residue (Asp). This position is critical for activity in both insects and vertebrate systems. The connectedness of the vertebrate CCK/gastrin family (negative amino acid in a critical position) with the invertebrate SK families (positive residue in the same critical residue) is strengthened by the neutral ancestral form in nematodes.
Expression analysis revealed that the C. elegans CK (NLP-12) peptides are expressed in a single tail neuron, identified as the interneuron DVA. The only function linked to DVA so far is the integration of mechanosensory information, and providing input to both the anterior and posterior touch circuits (94, 95). Typical to DVA are the large, vesicle-filled varicosities in the DVA process in the nerve ring. These vesicle-filled varicosities also contain the NLP-12 peptides as could be confirmed by immunostaining with an NLP-12a antibody. Also in the sour paste nematode P. redivivus a similar staining pattern in a single tail neuron and the vesicle-filled varicosities in the process of the nerve ring was observed (Mousley, A. and A. Maule, personal communication). We believe that these varicosities could represent the nematode equivalent of the neurohemal organs found in insects (i.e. corpora cardiaca) and vertebrates (i.e. pituitary gland). The staining of these vesicle-filled varicosities can easily be mistaken for separate head neurons and could explain the contradicting expression reports in WormBase (e.g. NLP-12 expression in head neurons). It could also explain the mass spectrometric results of Yew et al. (56), which reported the presence of mass peaks corresponding to NLP-12-like peptides in A. suum head neurons. However, it does not give a likely explanation for the presence of nlp-12 transcripts in A. suum head tissue (55). Previously, McVeigh et al. (55) localized the NLP-12 of Trichostrongylus colubriformis to a single tail neuron by in situ hybridization. Therefore, the expression of NLP-12 in a single tail neuron seems to be conserved across variant nematode species and is also in agreement with the neuronal expression of the arthropod SKs (21, 22, 23, 57).
In addition, the NLP-12::mCherry fluorescence was also observed in six cells in the pseudocoelomic cavity called ccs. Because of their ability to take up a variety of molecules from the body cavity fluid, these cells have been suggested to act as scavenger cells (96) and allow for a primitive immune surveillance function (97). In C. elegans, ccs have taken up fluid-phase markers such as India ink, rhodamine-dextran, fluorescein isothiocyanate-BSA, fluorescein isothiocyanate-lipopolysaccharide, and green fluorescent protein (GFP)-tagged yolk particles from the pseudocoelom (97, 98). We cannot exclude the possibility that the presence of NLP-12::mCherry observed in the cc vesicles could be due to the endocytosis of NLP-12::mCherry from the pseudocoelom. This could explain the absence of cc staining observed with immunological methods (Mousley, A., and A. Maule, personal communication) and would require the NLP-12::mCherry product to be secreted into the pseudocoelom (by DVA). In turn, this would suggest a potential "hormonal"/neuromodulatory role for CKI and CKII, and would be consistent with a potential neurohemal role for the vesicle-filled varicosities of DVA.
Previously, McKay et al. (38) used a translational GFP fusion construct to analyze the expression pattern of the CKR-2 receptor(s), but, regrettably, no detectable GFP expression could be observed in any of the CKR-2::GFP transgenic lines.
When tested in the cockroach hindgut assay, none of the CK peptides displayed any myotropic activity at any of the tested concentrations. Most SKs are known to have a myotropic effect on cockroach hindgut contractions (9, 10, 11, 12, 26), but in contrast, they do not have an in vitro effect on gut contraction or intestinal movement in blowflies (21, 99) or moths (24).
In addition, CCK/gastrins have been attributed a role in intestinal motility, but they are not active in the cockroach hindgut assay unless their C-terminal Asp residue is replaced with Arg (26, 100). The sulfate residue also seems to be a strict requirement for myotropic activity in the cockroach assay (26, 79). Therefore, the sequence differences (e.g. C-terminal Gln instead of Arg) and the unsulfated form of CKI and CKII could explain the lack of myotropic activity in the insect assay. However, on isolated muscle preparations of A. suum, NLP-12 has displayed myotropic activities (55). The actions of NLP-12 on both ventral and dorsal muscle strips were not congruent with the results of Reinitz et al. (101), who found that NLP-12a, on injection into the pseudocoele of live adult female A. suum, caused ventral coiling behavior, suggesting an activity specific for ventral muscle (55, 102). To explain this they hypothesized the possibility of a zone-specific localization of NLP-12 receptor subtypes on body wall muscle. The discovery of two receptor splice isoforms in C. elegans strengthens this possibility. In addition, nlp-12 knockdown (RNAi) animals and nlp-12 mutants show aldicarb resistance (103), which adds to the suggestion of a role in locomotor control for nlp-12. A basic analysis of the nlp-12 and ckr-2 mutants locomotor behavior did not reveal any obvious abnormalities. A more detailed analysis will be required to characterize fully the myotropic role of CK signaling in nematodes. Future experiments documenting the CKR-2 expression pattern and possible ligand colocalization will be necessary to substantiate the functional understanding of this system.
The amylase activity/secretion in ckr-2 mutants was significantly lower compared with wild-type animals. This is indicative of a role for CKR-2 signaling in the stimulation of digestive enzyme secretion, which is in agreement with the function of its arthropod and vertebrate counterparts (2, 3, 19, 20). Interestingly, a lack of NLP-12 in C. elegans did not affect the in vivo amylase activity/secretion. This could be attributed to the cumulative effect of non-NLP-12 peptides (i.e. FLP-1a, NLP-14a and NLP-13c), which individually are only active at high concentrations or by the cumulative effect of other endogenous receptor activating substances not picked up in the receptor-screening assay.
nlp-12 and ckr-2 mutant animals both display a highly increased fat content compared with wild-type worms, suggesting a role of CK signaling in the control of nematode fat storage. Recently, Clerc et al. (4) reported an important contributing role for CCK2R in the regulation of food intake in mice. They found that CCK2R deficient mice (CCK2R–/–) developed obesity that was associated with hyperphagia. The obesity was linked to an increased fat deposition resulting from hypertrophy (4). The lack of CCK2R in mice and CKR-2 in C. elegans results in a similar phenotype with regard to fat deposition. However, in worms we could not directly associate the fat increase with hyperphagia. C. elegans feeds through the powerful pumping action of the pharyngeal tube that results in bacterial ingestion, concentration, grinding, and forcing of the crushed suspension into the intestinal lumen. The lack of CK and CKR-2 did not seem to affect the pharyngeal pumping rate and defecation rate in C. elegans. This suggests that CK signaling does not affect the rate of food intake, which seems to contradict the satiety actions of SKs in insects and of CCK in vertebrates.
Only recently, McKay et al. (38) described the conservation of the fat regulatory function of the exopeptidase tripeptidyl peptidase II (TPPII) between mice and C. elegans. In mammals, TPPII has a proposed antisatiety role affected by degradation of CCK-8 (104). McKay et al. (38) showed that a lack of TPPII results in a decreased fat content in both mice and worms. However, this seems to occur independently of feeding behavior, and, therefore, they suggest that TPPII possibly regulates worm fat storage independently of central satiety control. They also used the pharyngeal pumping rate as a measure of food intake.
Still, we believe that we cannot exclude hyperphagia as the cause of fat increase and exclude a role of CK signaling in the control of satiety in nematodes because the pharyngeal pumping rate of worms is modulated by food availability (105) and only indicative of the speed of food intake, and not of meal size. Based on their results in the blowfly P. regina, Downer et al. (17) already suggested that SKs do not influence whether the insect eats or not but rather affect the size of the meal once the insect starts feeding.
At least our data suggest that also in C. elegans CK has an important role in the regulation of fat storage that is mediated specifically through the CKR-2 receptor(s). Whether this regulation is linked to the degrading actions of TPPII on CK, the association with hyperphagia, or acts completely independent of central satiety control remains to be determined.
Conclusions
Together, our results indicate that this newly identified neuropeptide signaling system in C. elegans constitutes not only the structural, but also the functional homolog of the CCK/gastrin signaling system of vertebrates and the SK signaling system of arthropods. In terms of evolution, C. elegans CKI and CKII would represent the oldest members of the CCK/gastrin family of peptides known to date.
C. elegans has great potential as a model to uncover the processes that govern metabolic physiology (45, 69, 106). As obesity and its alimentary diseases are rapidly becoming the industrial epidemic of the 21st century, the study of CK signaling in C. elegans could prove very helpful in the search for new targets for and the development of novel metabotherapeutics.
| Acknowledgments |
|---|
| Footnotes |
|---|
Disclosure Statement: The authors have nothing to declare.
First Published Online March 13, 2008
Abbreviations: BLAST, Basic local alignment search tool; cc, coelomocyte; CCK, cholecystokinin; CHO, Chinese hamster ovary; ESI, electrospray ionization; FLP, FMRFamide-like peptide; GFP, green fluorescent protein; GPCR, G protein-coupled receptor; MALDI, matrix-assisted laser desorption ionization; MS, mass spectrometry; MS/MS, tandem mass spectrometry; mtAEQ, mitochondrially targeted apo-aequorin; NGM, nematode growth medium; NLP, neuropeptide-like protein; ORF, open reading frame; RFU, relative fluorescent unit; SK, sulfakinin; TBS, Tris-buffered saline; TFA, trifluoroacetic acid; TOF, time-of-flight; TPPII, tripeptidyl peptidase II.
Received December 20, 2007.
Accepted for publication March 3, 2008.
| References |
|---|
|
|
|---|
-amylase in two invertebrates: the scallop Pecten maximus and insect Rhynchophorus ferrugineus. In: Beckwith BE, Saria A, Chronwall BM, Sandman CA, Strand FL (eds.) Neuropeptides in development and aging. Ann NY Acad Sci 814:335–338[CrossRef]This article has been cited by other articles:
![]() |
M. Lindemans, F. Liu, T. Janssen, S. J. Husson, I. Mertens, G. Gade, and L. Schoofs Adipokinetic hormone signaling through the gonadotropin-releasing hormone receptor modulates egg-laying in Caenorhabditis elegans PNAS, February 3, 2009; 106(5): 1642 - 1647. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. H. Greeley Jr. Evidence for Intelligent Design in Gastrointestinal Endocrinology: Identification of Novel Cholecystokinin/Gastrin-Like Peptides in the Nematode Caenorhabditis elegans Endocrinology, June 1, 2008; 149(6): 3184 - 3186. [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |