Identification of neuropeptide-like protein gene families in Caenorhabditis elegans and other species

Abstract

Neuropeptides play critical roles in synaptic signaling in all nervous systems. Unlike classical neurotransmitters, peptidergic neurotransmitters are encoded as preproproteins that are posttranslationally processed to yield bioactive neuropeptides. To identify novel peptidergic neurotransmitters, the Caenorhabditis elegans genome was searched for predicted proteins with the structural hallmarks of neuropeptide preproproteins. Thirty-two C. elegans neuropeptide-like protein (nlp) genes were identified. The nlp genes define at least 11 families of putative neuropeptides with unique motifs; similar expressed sequence tags were identified in other invertebrate species for all 11 families. Six of these families are defined by putative bioactive motifs (FAFA, GGxYamide, MRxamide, LQFamide, LxDxamide, and GGARAF); the remaining five families are related to allatostatin, myomodulin, buccalin/drosulfakinin, orcokinin, and APGWamide neuropeptides (MGL/Famide, FRPamide, MSFamide, GFxGF, and YGGWamide families, respectively). Most C. elegans nlp gene expression is in neurons. The C. elegans nlp genes and similar genes encoding putative neuropeptides in other species are likely to play diverse roles in nervous system function.

Chemical signaling via neurotransmitters is critical for synaptic transmission of information between neurons. Neuropeptides are the most varied and numerous type of neurotransmitters. Invertebrate neuropeptides are thought primarily to modulate synaptic function of classical small-molecule neurotransmitters by means of seven transmembrane domain receptors. However, the recent identification of a FMRFamide-gated sodium channel from Helix lucorum suggests that they may also act as fast transmitters (1). In mammals, neuropeptides and their receptors are implicated in behaviors including feeding and sleep (2–5). Despite their clear roles in synaptic signaling and behavior, neuropeptide functions are still not understood.

Biochemical isolation of neuropeptides has been relatively successful in several invertebrate systems, including Lymnaea stagnalis, Drosophila melanogaster, and Aplysia californica (6–8), and has led to the identification of several invertebrate neuropeptide families. In the nematode Caenorhabditis elegans, 23 FMRFamide-related proteins (FaRP) neuropeptide genes, designated flp-1 to flp-23 (FMRFamide-like proteins), have been identified (9). Only flp-1 has been characterized at the functional level. Animals lacking flp-1 have abnormal behavior, including uncoordinated movement and hyperactivity (10). The only other C. elegans non-flp neuropeptide genes that have been identified are the 37 insulin-like genes (11, 12).

Dense-core synaptic vesicles are prevalent in presynaptic terminals of C. elegans neurons that are neither FaRP immunoreactive nor catecholaminergic (13), suggesting that non-FaRP neuropeptides are present. Additionally, about 130 genes encoding putative neuropeptide receptors were identified in the C. elegans genome (14). This large number of receptors is much higher than the number of flp-encoded FaRPs (9) and is reminiscent of the large number of putative “orphan” neuropeptide receptors in vertebrates. This finding suggests that several more families of neuropeptides are as yet undiscovered.

The relative paucity of non-flp neuropeptide genes previously identified in C. elegans (by either genetic or biochemical techniques) led to the suggestion that FaRPs could be responsible for the majority of neuropeptide signaling in this animal (10). Clearly, neuropeptide signaling is implicated in multiple C. elegans behaviors, including defecation and social behavior (15, 16). Described herein are 32 non-flp putative neuropeptide genes that are expressed primarily in C. elegans neurons. These genes encode 134 unique and 151 total putative neuropeptides. Related candidate neuropeptides are found in other species, suggesting that these genes are functionally conserved.

Materials and Methods

Structural searches used the patternfind program (http://www.isrec.isb-sib.ch/software/PATFND_form.html). blast searches (17, 18) yielded neuropeptide-like protein (nlp) homologs in other species. Signal peptide cleavage sites were defined by using SIGNALP 2.0 (http://www.cbs.dtu.dk/services/SignalP/) (19). Predicted NLPs and peptides were aligned by using CLUSTALW 1.8 (http://searchlauncher.bcm.tmc.edu/multi-align/multi-align.html) (20). nlp gene families were defined by using (primarily) motifs, phylogenetic tree construction (megalign, Dnastar, Madison, WI), and overall predicted peptide homology. A motif is defined as at least three of five identical amino acids at the N or C terminus of at least two predicted peptides within a nlp gene and/or family. Motifs also were used for additional searching with patternfind.

A mixed cDNA library (M. Vidal, Dana-Farber Cancer Institute, Boston) was screened for nlp-1 through nlp-21 by PCR amplification. Primers overlapped nlp gene predicted start and termination sites. Amplified products were directly cloned (pCR2.1-TOPO, Invitrogen). Twelve nlp∷gfp reporter constructs were made by insertion of the nlp promoter into a promoterless green fluorescent protein (GFP)-expressing vector, pPD95.67. The sequences and subcloning sites are: nlp-1: NsiI/XbaI into PstI/XbaI; nlp-3: NheI/NsiI into XbaI/PstI; nlp-16: BamHI/NsiI into BamHI/PstI; nlp-2, nlp-25 through nlp-32: BamHI/XhoI into SalI/BamHI. For the remaining nlp genes, the putative regulatory region was amplified by PCR (reaction 1). The gfp coding regions from pPD95.67 (21) were amplified by using a nlp-gene specific primer that also contained GFP vector 5′ sequence and by using a 3′ GFP vector primer (GFP-C, AAGGGCCCGTACGGCCGACTAGTAGG) in an independent PCR (reaction 2) (22). The two amplified products contained a 20- to 30-bp region of overlap, enabling amplification (reaction 3) of the full-length nlp∷gfp fragment (template DNA from reactions 1 and 2) by using a primer from the nlp promoter and from the GFP vector (GFP-2C, GGAAACAGTTATGTTTGGTATATTGGG). The cellular expression pattern for nlp-1 through nlp-3 fragments was indistinguishable from the cellular expression pattern for nlp-1 through nlp-3 subcloned constructs. PCR primers are listed in Table 1, which is published as supporting information on the PNAS web site, www.pnas.org; in general the upstream regulatory regions were ≈2.5 kb or extended to the next predicted gene.

Transgenic lines were generated by coinjection of the nlp∷gfp construct (50–70 ng/μl) and lin-15 rescue construct (pJM24,100 ng/μl) into lin-15(n765ts) animals (23, 24). At least two independent transgenic lines were analyzed for each nlp gene; 5–10 animals were scored per line. 1,1′-Dioctadecyl-3,3,3′,3′-tetramethylindodicarbocyanine perchlorate (Molecular Probes) was used to stain amphid and phasmid neurons (25) to facilitate identification.

Results

The C. elegans genomic sequence (26) was scanned for homologs of previously characterized invertebrate neuropeptides to identify C. elegans nlp genes. nlp genes are defined as those encoding putative neuropeptides that do not end in RFG. Two different search paradigms were used: similarity-based searching and pattern-based searching.

Similarity-Based Searching Yields nlp-1 and nlp-2.

Roughly 600 neuropeptides from the GenBank protein database were used to search the C. elegans proteome by using the fasta program (27) at genestream (http://www2.igh.cnrs.fr/bin/fasta-guess.cgi). This search, based on similarity of predicted C. elegans proteins to previously characterized neuropeptides, yielded only two non-FaRP genes. nlp-1 and nlp-2 encode putative bioactive peptides with modest similarity to Aplysia californica buccalin and myomodulin, respectively (28, 29).

Pattern-Based Searching Yields nlp-3 Through nlp-32.

Bioactive peptides encoded by a single invertebrate neuropeptide gene are often highly related. For example, the C. elegans flp-1 gene encodes seven putative neuropeptides that are highly related to each other (30). Because of the frequent sequence similarity among bioactive peptides encoded within a given invertebrate neuropeptide gene and the presence of characteristic endoproteolytic cleavage sites (usually KR), pattern-based searching strategies were used to identify additional neuropeptide genes. The patternfind program at the Swiss Institute for Experimental Cancer Research was used to search for putative C. elegans neuropeptide proteins based on structural criteria. Search patterns were permutations of [KR]-[KR]-x(3,20)-[KR]-[KR]-x(3,20)-[KR]-[KR]. Permutations included increasing the number of amino acids between predicted cleavage sites, modulating the number of peptide repeats, and testing single amino acid endoproteolytic sites. Additional related genes/expressed sequence tags (ESTs) were subsequently identified at the National Center for Biotechnology Information by using blast (17, 18). Not only were nlp-1 and nlp-2 reidentified, but an additional 30 non-FaRP genes were found (nlp-3 through nlp-32, see Fig. 1). Only nlp-19 was not previously predicted as a gene by the C. elegans Genome Sequencing Consortium (26).

Figure 1.

nlp genes encode putative neuropeptides. nlp genes are grouped into families (shaded); the family motif is listed in the first column along with the nlp gene name, ESTs, or cDNA designations. Only one yk EST (C. elegans EST project) and/or pHA cDNA clone (this paper) is listed even if multiple clones were identified. The second column indicates the chromosome, location (in map units), and cosmid designation. Predicted peptides are listed for each nlp gene in the third column. Peptides are predicted based on similarity within a nlp gene and putative mono- or dibasic endopeptidase cleavage. Putative peptides are listed sequentially for each nlp gene but identical putative peptides encoded by a nlp gene are listed only once; the number of times the putative peptide is repeated is indicated (2×, 3×). Predicted posttranslational modifications are not shown for individual predicted peptides. The last predicted nlp-21 peptide (parentheses) is differentially spliced. Cells expressing GFP from nlp gene reporter constructs are in the fourth column. Neurons/cells scored unequivocally by position and counterstaining: ASH, ASI, ASJ, ADL, ASK, AWB, PHA, PHB, HSN, BDU, spermatheca, vulval muscles, rectal gland cell, pharyngeal neurons, and hypoderm. Other classes of neurons were counted (head, tail, and lateral) but not always identified. Approximately 90% of cells expressing GFP in this figure are neurons. VNC, ventral nerve cord neurons; RVG, retrovesicular ganglion neurons. The last column lists the species and accession number for nlp gene family members in other species.

For final selection as a putative C. elegans neuropeptide gene the following criteria were required. (i) The predicted preproprotein fits the patternfind search criterion (above) and has at least 3/5 identical amino acids at the N or C terminus of the predicted peptides in addition to the flanking dibasic (or monobasic) cleavage site. Intragenic peptide homology of the nlp genes with multiple predicted neuropeptides is as high as 100% for many nlp genes. (ii) The putative preproprotein must be less than 400 aa in length. More than 98% of previously characterized neuropeptide preproproteins and half of C. elegans predicted proteins meet this criterion. (iii) The predicted preproprotein does not have a predicted biological activity by genetic analysis nor by homology; the objective was identification of new neuropeptide genes. (iv) The predicted protein must have a putative signal peptide. Twenty nine percent of the roughly 19,000 predicted C. elegans proteins have putative signal peptides. The nlp and flp genes were the only predicted C. elegans genes identified in our searching that met all of these criteria. The following exceptions should be noted: nlp-22 contains only one putative peptide, and the putative signal peptide (19) for nlp-4 is unclear.

Some nlp genes are grouped in clusters. nlp-22 is located adjacent to nlp-2 and nlp-23, which encode similar putative neuropeptides. Five nlp genes, nlp-27 through nlp-31, which encode homologous putative peptides, are located on cosmid B0213. Two nlp genes, nlp-13 and nlp-9, which encode putative peptides unrelated by sequence similarity, are located on cosmid E03D2. nlp-25 and nlp-26 are located on cosmid Y43F8C and encode putative peptides with slight similarity. The functional or evolutionary significance of this clustering is unclear.

Posttranslational modification of neuropeptides frequently is required for biological activity. All flp genes encode neuropeptides ending with a C-terminal RFG (31), which is presumably converted to RFamide in vivo by posttranslational chemical modification (32). The C-terminal glycine of various nlp predicted peptides is also likely removed, leaving a C-terminal amide in the putative active peptide, but we have not directly assessed this nor other posttranslational modifications.

nlp Gene Families Are Defined by Conserved Motifs.

Based on (i) the conserved motifs in predicted nlp neuropeptides and (ii) similar ESTs/neuropeptides in other species, families of related neuropeptide genes were constructed. Experimental evidence suggests that the level of homology over an entire neuropeptide is less important than the presence of a specific, conserved motif within the bioactive neuropeptide (33–36). Therefore, nlp genes were initially grouped into families based on 3- to 5-aa motifs found in predicted nlp neuropeptides. A motif is defined as at least three of five identical amino acids at the N or C terminus of at least two predicted peptides of a nlp gene.

Neuropeptides or ESTs from other species that encode similar putative peptides were identified by using blast. The similarity rarely extends beyond the predicted peptide and cleavage sites. Although ESTs are usually incompletely sequenced, these all have the structural hallmarks of neuropeptide preproprotein genes: endopeptidase sites and (usually) potential signal sequences. ESTs and previously characterized neuropeptides from other species were examined for the presence of nlp family motifs. The resulting conserved motifs and family assignments are briefly described below and enumerated in Fig. 1. Functional analysis will be required to determine the biological significance of family assignments and motifs.

nlp gene families are presented below in order of decreasing confidence that they encode neuropeptides. For this purpose, confidence is based solely on homology to previously characterized neuropeptides. nlp gene families most likely to encode neuropeptides based on homology are listed first. Next, nlp genes that share C-terminal motifs with previously characterized neuropeptides are listed. Then, nlp genes whose putative neuropeptides share motifs, putative cleavage sites, and (usually) signal peptides with ESTs in other species are listed. Finally, we list nlp genes that lack homologs in other species but fit the structural criterion for a putative neuropeptide gene.

GFxGF Family.

nlp-3, nlp-8, nlp-14, and nlp-15 contain putative peptides with various similarities. nlp-14 and nlp-15 putative peptides contain a C-terminal GFxGF (or GFGF) motif. They are up to 69% identical to orcokinin, a myotropic neuropeptide from Orconectes limosus (crayfish) abdominal nerves (37), which also contains a GFGF motif. Removal of the C-terminal FGFN or the last phenylalanine of orcokinin causes a loss of biological activity, suggesting that this motif is functionally significant (50). This motif is also found in orcokinins from Carcinus maenas (shore crab) and Procambarus clarkii (red swamp crayfish) (38). Many candidate peptides containing the GFxGF motif are encoded by ESTs in multiple species.

nlp-3 and nlp-8 putative peptides do not contain GFxGF motifs but are very similar to specific putative peptides encoded by nlp-14 and nlp-15. The final putative peptide encoded by nlp-3 (YFDSLAGQSLG) is 70% identical to part of a nlp-15 putative peptide (AFDSLAGQGFTGFE). Also, nlp-3-related putative peptides encoded by a Globodera rostochiensis (potato cyst nematode) EST do not contain a GFxGF moiety.

nlp-8 and nlp-15 both encode putative peptides starting with AFD that diverge at their C termini, and nlp-8 putative peptides are up to 46% identical to nlp-15 putative peptides. ESTs from Meloidogyne incognita and Meloidogyne javanica (root-knot nematodes) also encode putative peptides with this N-terminal motif and various other similarities. The relative functional significance of these various motifs found in ESTs nlp-14, nlp-15, nlp-3, and nlp-8 remains unclear.

FRPamide Family.

nlp-2, nlp-22, and nlp-23 encode putative peptides ending in FRPG. Again, the C-terminal glycine is a likely target for amidation in vivo. nlp-2 putative peptides are up to 60% identical to A. californica and up to 42% identical to Lymnaea stagnalis (great pond snail) myomodulin peptides (8, 39). Myomodulin peptides from these species share N-terminal homology with C. elegans FRPamide family members, but lack the FRPamide of the C. elegans putative peptides. Putative neuropeptides containing an FRPG are found within ESTs from other organisms including Toxocara canis (Tc-huf-316, AA874711), which is similar to nlp-2 (40).

MSFamide Family.

nlp-1, nlp-7, and nlp-13 putative peptides contain MSFG and related C-terminal motifs (including MAFG). The C-terminal glycine is a likely target for amidation; Drosophila melanogaster drosulfakinin (fruit fly) also ends in SFG and is amidated in vivo (41). nlp-1 putative peptides are up to 46% identical to A. californica (sea hare) buccalin, a neuropeptide neurotransmitter that modulates acetylcholine-induced muscle contraction (42). Multiple ESTs with SFG and/or AFG motifs are found in other species.

MGL/Famide Family.

nlp-5 and nlp-6 contain putative peptides ending in MGLG and MGFG. The C-terminal glycine is a likely target for amidation. Allatostatin (43), a Blaberus craniifer (cockroach) neuropeptide, is up to 43% identical to nlp-5. Other allatostatins, helicostatin, Aplysia buccalin, and a putative peptide encoded by an Ancylostoma caninum EST (dog hookworm) also contain putative peptides with a C-terminal GLG motif; a putative peptide encoded by a B. malayi EST is up to 50% identical to the C-terminal end of a nlp-5 putative peptide.

YGGWamide Family.

nlp-24, nlp-25, and nlp-27 through nlp-32 encode putative peptides sharing YGGWG and YGGYG motifs. Interestingly, the APGWamide neuropeptide (U85585) from A. californica has a related C-terminal motif (44). ESTs encoding similar putative peptides are found in insects and nematodes. Four ESTs from the nematode Pristionchus pacificus encode similar putative peptides and appear to arise from different genes.

GGARAF Family.

nlp-9 and nlp-21 encode putative peptides containing an N-terminal GGARAF motif. The conservation of this N-terminal motif is particularly striking in comparison to the lack of C-terminal conservation between nlp-9 and nlp-2 putative peptides—even within the same gene. Putative neuropeptides beginning with this motif are encoded by ESTs from other nematodes.

FAFA Family.

nlp-18 and nlp-20 encode putative peptides ending in FAFA and related motifs. Candidate preproproteins that have the structural hallmarks of neuropeptides and containing multiple FAFA putative peptides are predicted in ESTs from other nematode species. Perusal of the genomic sequence suggests that nlp-20 may be in an operon with a neutral endopeptidase, F45E4.7, which could be involved in nlp-20 processing.

GGxYamide, LxDxamide, LQFamide, and MRxamide Families.

Eight nlp genes (nlp-4, nlp-10, nlp-11, nlp-12, nlp-16, nlp-17, nlp-19, and nlp-26) are not found in multigene families within C. elegans and are not clearly related to well-characterized neuropeptides. Based on motifs found in ESTs from other species, four of these genes can be assigned to gene families. nlp-10 is the C. elegans representative of the GGxYamide family, nlp-11 represents the LxDxamide family, nlp-12 belongs to the LQFamide family, and nlp-17 represents the MRxGamide family. Interestingly, a peptide encoded by nlp-12 was shown to modulate locomotion (45) and shown to have potent myoexcitatory activity in A. suum (N. Marks, personal communication). The remaining nlp genes that lack related ESTs in other species (nlp-4, nlp-16, nlp-19, and nlp-26) may also define additional neuropeptide gene families.

Identification of nlp Gene cDNAs.

The C. elegans and the National Center for Biotechnology Information databases contain ESTs corresponding to 15 nlp genes: nlp-9, nlp-12 through nlp-17, nlp-20, nlp-21, nlp-24, nlp-26, nlp-27, and nlp-29 through nlp-31. A mixed stage library was screened for cDNAs corresponding to nlp-1 through nlp-21. cDNA clones were identified for six additional nlp genes: nlp-3, nlp-5, nlp-7, nlp-8, nlp-10, and nlp-18. In total, 21 of 32 nlp genes have either a C. elegans cDNA or EST. Expression was detected in microarray experiments for seven of the 11 remaining nlp genes (J. Gaudet and S. E. Mango, personal communication). Of these 11 nlp genes, 10 have significant similarity to ESTs in other species, suggesting that they are also expressed genes.

Three flp genes are alternatively spliced (10). So far, only one nlp gene, nlp-21, is known to be alternatively spliced. The putative neuropeptides encoded by alternative transcripts of nlp-21 differ by one putative peptide in a terminal exon: one nlp-21 transcript encodes nine putative peptides whereas the second transcript encodes eight. However, primers for PCR screening of the cDNA library were based on software predictions of intron/exon structure, possibly excluding identification of alternative transcripts for other nlp genes.

Many nlp Genes Are Expressed in Neurons and Secretory Cells.

To address the cellular expression pattern of nlp genes, reporter constructs were generated by using putative 5′ regulatory sequences to drive expression of the GFP gene in transgenic animals (46). Thirty two nlp constructs were analyzed; many transgenic lines had complex neuronal expression patterns. (see Fig. 2, which is published as supporting information on the PNAS web site, www.pnas.org). Transgenic animals had no detectable GFP expression for six genes (nlp-4, nlp-17, nlp-22, nlp-25, nlp-28, and nlp-32) although three of these (nlp-2, nlp-22, and nlp-28) are detectable in microarray experiments (J. Gaudet and S. E. Mango, personal communication). Five nlp gene reporter constructs (nlp-7, nlp-9, nlp-14, nlp-15, and nlp-21) are expressed in neurons located in the ventral nerve cord, which is directly involved in locomotion. Eight reporter constructs (nlp-3, nlp-8, nlp-13, nlp-18, nlp-19, nlp-20, nlp-21, and nlp-24) are expressed in pharyngeal neurons that modulate pharyngeal pumping of food. Expression of GFP in the intestine was also seen for many nlp genes. A putative role for C. elegans intestinal neuropeptides in defection was previously proposed (16) and the vertebrate gut contains neuropeptides. Embryonic expression (precomma stage) was noted for two nlp reporter constructs, suggesting that nlp-21 and nlp-31 may function in development, as previously demonstrated for PACAP in vertebrates (47). Transgenes are rarely expressed in the C. elegans germ line; it is, therefore, unclear whether any nlp genes are expressed in the germ line. An additional neuroendocrine role is suggested for nine genes whose reporter constructs showed expression in somatic gonad tissues, including spermatheca, somatic gonad, uterine muscles, and secretory cells. The predominantly hypodermal expression of YGGWamide family members was striking in comparison to other nlp gene families. Members of this family may play a significant role in non-neuronal signaling or, despite similarity to the APGWamide neuropeptide from Aplysia, these genes may not encode neuropeptides.

Discussion

Using a pattern-based searching strategy, 32 previously uncharacterized putative neuropeptide genes were identified in C. elegans. An additional 51 ESTs encoding putative neuropeptides were identified in various invertebrate species. Our results suggest that the diversity of invertebrate neuropeptides is greater than previously assumed.

We predict that 92 C. elegans genes may encode neuropeptides, including 23 flp genes (9), 37 insulin-related genes (12), and 32 nlp genes. Many of the nlp genes are likely to encode neuropeptide preproproteins because they fit the following criteria. First, nlp genes are structurally related to previously characterized neuropeptide genes; most have a predicted signal peptide region, are of an appropriate length (<400 aa), and have acceptable sites of endoproteolytic cleavage, which flank the conserved motifs within peptides. Second, most nlp genes encode related putative neuropeptides, a common feature of invertebrate neuropeptide genes. Third, many nlp putative peptides show sequence similarity and/or share putative bioactive motifs with bona fide neuropeptides. Eighteen of 32 nlp genes are in families containing previously characterized neuropeptides. Consistent with their putative role as neuropeptides, nlp genes are expressed predominantly in neurons and endocrine tissues.

Pattern-based searching was significantly more productive than similarity searching, indicating the necessity of developing tools that fully use genomic information. However, the pattern-based searching described herein only detects putative neuropeptide genes encoding multiple, related bioactive putative peptides. Individual genes encoding multiple, unrelated bioactive peptides or genes encoding just one novel bioactive peptide were not identified except by similarity. For example, a Brugia malayi EST (filarial parasitic nematode, AI834180), which encodes multiple putative neuropeptides ending in FLHFG, was identified. But, the C. elegans insulin-related genes (11, 12), which do not contain repeated, similar peptides, were not reidentified.

Most nlp gene homologs are found in nematodes. C. elegans predicted proteins were searched for neuropeptide genes based on specific patterns and motifs, then related proteins/ESTs in other species were identified based on overall protein similarity. The EST database was not searched by using patterns for novel neuropeptides. Therefore, the relative lack of nlp gene homologs in non-nematode invertebrates results from either the scarcity of predicted proteins in the database from other species, evolutionary divergence, and/or a paucity of related, repeated-motif neuropeptides in other species. The latter is likely for D. melanogaster. Pattern-based searching of D. melanogaster predicted proteins was relatively unsuccessful; only previously identified/annotated FMRFamide genes were identified in the fruit fly.

No clear homolog of a nlp putative peptide was identified in vertebrates. Previously characterized vertebrate neuropeptide genes do not encode highly related peptides, which further complicates identification. Interestingly, the last nlp-8 predicted peptide contains a C-terminal region (4/5 C-terminal amino acids) of sequence identity with human substance P. More sensitive searching techniques might identify new vertebrate neuropeptides with similarity to nlp putative neuropeptides. Bioactive RFamide family neuropeptides were previously identified in vertebrates by using immunological techniques (48, 49); a similar approach based on nlp gene family motifs might be fruitful.

Neuropeptides are integral to behavior and nervous systems in animals. Using data from the C. elegans genome sequencing project, 32 previously uncharacterized putative neuropeptide genes with homologs in other species were identified. Further characterization of the nlp genes is likely to provide a greater understanding of mechanisms involved in neuropeptide function in development and behavior.

Abbreviations

FaRP

FMRFamide-related protein

nlp

neuropeptide-like protein

flp

FMRFamide-like protein

GFP

green fluorescent protein

EST

expressed sequence tag