Ancient peptidergic neurons regulate ciliary swimming and settlement in Platynereis dumerilii

Background: Neuronal antibodies that show immunoreactivity across a broad range of species are important toolsfor comparative neuroanatomy. Nonetheless, the current antibody repertoire for non-model invertebrates is limited.Currently, only antibodies against the neuropeptide RFamide and the monoamine transmitter serotonin areextensively used. These antibodies label respective neuron-populations and their axons and dendrites in a largenumber of species across various animal phyla.Results: Several other neuropeptides also have a broad phyletic distribution among invertebrates, includingDLamides, FVamides, FLamides, GWamides and RYamides. These neuropeptides show strong conservation of thetwo carboxy-terminal amino acids and are α-amidated at their C-termini. We generated and affinity-purified specificpolyclonal antibodies against each of these conserved amidated dipeptide motifs. We thoroughly tested antibodyreactivity and specificity both by peptide pre-incubation experiments and by showing a close correlation betweenthe immunostaining signals and mRNA expression patterns of the respective precursor genes in the annelidPlatynereis. We also demonstrated the usefulness of these antibodies by performing immunostainings on a broadrange of invertebrate species, including cnidarians, annelids, molluscs, a bryozoan, and a crustacean. In all species,the antibodies label distinct neuronal populations and their axonal projections. In the ciliated larvae of cnidarians,annelids, molluscs and bryozoans, a subset of antibodies reveal peptidergic innervation of locomotor cilia.Conclusions: We developed five specific cross-species-reactive antibodies recognizing conserved two-amino-acidamidated neuropeptide epitopes. These antibodies allow specific labelling of peptidergic neurons and theirprojections in a broad range of invertebrates. Our comparative survey across several marine phyla demonstrates abroad occurrence of peptidergic innervation of larval ciliary bands, suggesting a general role of theseneuropeptides in the regulation of ciliary swimming. BackgroundAntibodies that show specific immunoreactivity across abroad range of species are valuable tools for comparativeneuroanatomy in non-model organisms. For example,antibodies against serotonin commonly label cell bodiesand their projections, allowing comparative studies ofneurodevelopment and neuroanatomy across diversespecies and phyla [1]. Another commonly used antibodyis that against FMRFamide, a neuropeptide first disco-vered in molluscs [2,3]. Similar RFamide neuropeptideswere later found to be widespread among eumetazoans[4-6]. A pioneering work reported the development ofantibodies against the conserved amidated dipeptidemotif RFamide [7]. This RFamide and other FMRFamideantibodies have been extensively used in invertebrateneuroanatomy, owing to the broad distribution ofRFamide-like peptides [8]. The RFamide antibody labelsdistinct neuronal subsets and their projections, and canbe applied as a neuronal marker to increase morpho-logical resolution in complex adult tissues [9], or toreveal aspects of nervous system development andorganization, allowing the clarification of phylogeneticrelationships within phyla [10-12] or the study ofnervous system evolution between related groups [13].* Correspondence: [email protected] Planck Institute for Developmental Biology, Spemannstrasse 35,Tübingen 72076, Germany © 2012 Conzelmann and Jékely; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of theCreative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use,distribution, and reproduction in any medium, provided the original work is properly cited.Conzelmann and Jékely EvoDevo 2012, 3:23http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 49Neuropeptides are signalling molecules that are trans-lated as precursor molecules, typically consisting of anN-terminal signal peptide and multiple copies of similarpeptide motifs, flanked by dibasic cleavage sites (Lys andArg residues). The precursor is cleaved and often furthermodified to yield shorter active neuropeptides [14,15].α-amidation is the most common post-translationalmodification, where a C-terminal glycine is enzymaticallyconverted into an amide group. This modification protectsthe small peptides from degradation and is critical forreceptor binding [16-18]. Amidation is also thought to con-fer high immunogenic potential to short neuropeptides[19-21] and antibodies raised against amidated peptides arehighly specific for the amidated peptide moiety [21].Changes in hydrogen bonding capability caused by theamide group may lead to the improved receptor bindingand increased immunogenicity of C-terminally amidatedpeptides [22].The C-terminal residues in amidated neuropeptidesare often highly conserved across different species andeven phyla [23]. We reasoned that, like the RFamideantibodies, other dipeptide antibodies could also poten-tially be used as neuronal markers across a wide rangeof species. Here we report the development of specificneuronal antibodies against the amidated dipeptidemotifs of five conserved neuropeptides, DLamide, FVamide,FLamide, GWamide and RYamide. We show that theseantibodies recognize specific subsets of neurons and theirprojections in cnidarian, annelid, mollusc, bryozoan andcrustacean larvae. Furthermore, our antibody stainingsreveal that the neuropeptidergic innervation of locomotorcilia is a general feature of ciliated larvae. MethodsGeneration of polyclonal neuropeptide antibodiesThe amidated peptides, coupled to an adjuvant (lipoad-juvant Pam3) via an N-terminal cysteine (CRYamide,CGWamide, CFVamide, CFLamide, CDLamide), wereused to immunize rabbits. Sera were affinity-purified onthe respective peptide epitopes using a SulfoLink resin(Thermo Scientific, Rockford, USA) that allows thecoupling of cysteine containing peptides via a disulphidebond. After coupling of 1 mg peptide epitope to 2 mlresin in Coupling Buffer (CB; 50 mM TRIS pH 8.5,5 mM EDTA), the resin was washed three times with10 ml CB. Excess reactive sites were blocked by incubat-ing the resin in 2 ml 50 mM cysteine for 45 min, fol-lowed by three washes with 1 M NaCl and three washeswith 25 ml phosphate buffered saline (PBS). Next, 25 mlserum was applied to the resin and this was incubatedovernight to allow antibody binding. After flow-throughof the serum, the resin was washed five times with 25 mlPBS followed by a wash with 15 ml 0.5 M NaCl/PBS andagain twice with 10 ml PBS. The antibodies were elutedand fractionated with eight times 1 ml of 100 mM glycinepH 2.7, eight times 1 ml of 100 mM glycine pH 2.3 andeight times 1 ml of 100 mM glycine pH 2.0. The fractionswere neutralized by directly collecting them in an ad-equate volume (about 40, 75 and 95 μl for the differentpH solutions) of 1 M TRIS–HCl pH 9.5. The protein con-centration of each fraction was determined, and the firsttwo fractions of the pH 2.7 peak (usually fractions 2 and3) were discarded, since these contained the lowest affinityantibodies. The peak fractions and the end-of-peak frac-tions were pooled, and concentrated, if necessary, usingVivaspin centrifugation tubes with a molecular weightcut-off of 10 kDa (Sartorius, Göttingen, Germany). Anti-bodies were stored in 50% glycerol at −20°C for mid-term(up to 1 year), and −80°C for long-term storage. A detailedprotocol is available [24]. ImmunohistochemistryFor immunostainings, larvae were fixed in 4% formalde-hyde in PTW (PBS + 0.1% Tween-20) for 2 h and storedin 100% methanol at −20°C until use. After stepwiserehydration to PTW, samples were permeabilized withproteinase-K treatment (100 μg/ml in PTW, for 1 to3 min). To stop proteinase-K activity, larvae were rinsedwith glycine buffer (5 μg/ml in PTW) and post-fixed in 4%formaldehyde in PTW for 20 min followed by two 5 min-washes in PTW and two 5 minwashes in THT (0.1 MTRIS–HCl pH 8.5 + 0.1% Tween-20). Larvae and anti-bodies were blocked in 5% sheep serum in THT for 1 h.Primary antibodies were used at a final concentration of1 μg/ml for rabbit neuropeptide antibodies and 0.5 μg/mlfor mouse anti-acetylated tubulin antibody (Sigma, SaintLouis, USA) and incubated overnight at 6°C. Weaklybound primary antibodies were removed by two 10 minwashes in 1 M NaCl in THT, followed by five 30 minwashes in THT. Larvae were incubated overnight at 6°Cin the dark in 1 μg/ml anti-rabbit Alexa Fluor® 647 anti-body (Invitrogen, Carlsbad, CA, USA) and in 0.5 μg/mlanti-mouse FITC antibody (Jackson Immuno Research,West Grove, PA, USA) and then washed six times for30 min with THT-buffer, and mounted in 87% glycerol in-cluding 2.5 mg/ml of the anti-photobleaching reagent 1,4-diazabicyclo[2.2.2]octane (Sigma, St. Louis, MO, USA).Pecten larvae were additionally treated with 4% parafor-maldehyde in PBS with 50 μM EDTA pH 8.0 for 1 h to de-calcify their shells before the immunostaining procedure(performed as described previously). For cnidarian larvae,we also used a mouse anti-tyrosylated tubulin antibody(Sigma, Saint Louis, USA) at 1 μg/ml. For immunostainingwith multiple rabbit primary antibodies in the same sam-ple, antibodies were directly labelled with a fluorophoreusing the Zenon® Tricolour Rabbit IgG Labelling Kit(Invitrogen, Carlsbad, CA, USA) and used in combin-ation with mouse anti-acetylated tubulin antibody.Conzelmann and Jékely EvoDevo 2012, 3:23Page 2 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 50For blocking experiments, we pre-incubated the anti-bodies in 5 mM of the respective full-length Platynereispeptides (YYGFNNDLamide, AHRFVamide, AKYFLamide,VFRYamide, RGWamide) for 2 h before immunostainings. Microscopy and image processingImages were taken on an Olympus Fluoview-1000confocal microscope (Olympus Deutschland GmbH,Hamburg, Germany) using a 60× water-immersionobjective and the appropriate laser lines to capturefluorescent signals. Signals from RNA in situ hybridizations(nitro blue tetrazolium chloride/5-Bromo-4-cloro-3-indolylphosphate precipitate) were imaged with reflection confocalmicroscopy as described [25]. Images were processed withImaris 6.4 (BitPlane Inc., Saint Paul, USA) and ImageJ 1.45software [26]. All image stacks are available [24]. Bioinformatic toolsNeuropeptide prediction was performed using NeuroPred[27], N-terminal signal peptides were predicted usingSignalP 4.0 Server [28]. For multiple sequence alignments,we used ClustalW [29]. The GenBank accession numberfor the Platynereis RGWamide neuropeptide precursor:JX412226. ResultsGeneration of specific antibodies against amidateddipeptide epitopes of neuropeptidesWe set out to develop antibodies against the conservedC-amidated dipeptides DLa, FVa, FLa, GWa and RYa(‘a’ = ‘amide’) neuropeptides that are conserved acrossphyla (Figure 1) [23,25-36]. We have recently shown inthe marine annelid model Platynereisdumerilii that theprecursor mRNAs for these neuropeptides are expressedin largely non-overlapping subsets of neurons in thelarval episphere. None of these neuropeptides co-expresseswith FMRFamide in Platynereis [23], suggesting that anti-bodies against their conserved amidated dipeptides couldalso substantially increase the number of neurons that canbe labelled in other species.Rabbits were immunized with the short amidated pep-tides extended with an N-terminal cysteine to allow coup-ling to a carrier during the immunization procedure. Wealso used the cysteine residue to couple the peptides to aresin and to affinity purify the antibodies from the respec-tive sera. We employed a high stringency affinity purifica-tion protocol including high salt washes and low pHelution to obtain high-affinity antibody fractions.Next, we tested the reactivity of the affinity purifiedneuropeptide antibodies in whole mount immunostainingson Platynereis larvae. We found labelling for all antibodiesin a subset of neurons and their axons in the larvalepisphere (Figure 2A-D). To test the specificity of ourantibodies, we pre-incubated them in the synthetic ami-dated full-length Platynereis peptides. This treatment ledto a complete block of the signal for the anti-DLa, anti-FVa and anti-RYa antibodies (Figure 3A,C,E) and a strongreduction in signal intensity for the anti-FLa and anti-GWa antibodies (Figure 3B,D). These results indicate thatthe antibodies bind to the respective peptides and this pre-vents further binding to epitopes in the tissue. The speci-ficity of the antibodies is further supported by theclose correlation between the cell body positionsrevealed by immunostaining and the expression pat-terns of the respective precursors (Figure 3 A-E, bot-tom panels, asterisks) as shown by whole-mount RNA Figure 1 Two amino-acid amidated motifs are conserved in neuropeptides across phyla. Multiple sequence alignment of matureneuropeptides with a conserved C-terminus for DLamides (A), FVamides (B), FLamides (D), GWamides (D) and RYamides (E) from annelids(Platynereis, Capitella, Helobdella), molluscs (Lottia), platyhelminthes (Schmidtea, Macrostomum), nematodes (Caenorhabditis), arthropods (Cherax,Cancer, Drosophila, Apis) and cnidarians (Podocoryne, Hydractinia). The conserved C-termini that were used for antibody production arehighlighted in black. RPCH, red pigment concentrating hormone; AKH, adipokinetic hormone.Conzelmann and Jékely EvoDevo 2012, 3:23Page 3 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 51Figure 2 Correspondence of antibody signals with the respective precursor mRNA expression in Platynereis. (A-E) Immunostaining (red)with the DLamide (A), FVamide (B), FLamide (C), GWamide (D) and RYamide (E) antibodies counterstained for acetylated tubulin (acTub, white) inthe upper panels. Bottom panels show immunostainings for the respective antibodies in white. (A0-E0) mRNA in situ hybridization (red)counterstained for acetylated tubulin (acTub, black) for the DLamide (A0), FVamide (B0), FLamide (C0), GWamide (D0) and RYamide (E0)neuropeptide precursors. With the exception of GWamide, all precursorsin situ were described in [23] and are shown here for comparison only.All images are anterior views of Platynereis larvae 48 h post fertilization (hpf). Asterisks indicate cells that show a spatial correspondence with themRNA in situ hybridization signals in A0-E0. Scale bars: 50 μm. Figure 3 Blocking of immunostaining signals with peptide pre-incubation. (A-E) Regular immunostaining (upper panels), and stainings withantibodies that were pre-incubated with the corresponding synthetic Platynereis full length neuropeptide (bottom panels) for DLamide (A),FVamide (B), FLamide (C), GWamide (D) and RYamide (E), all shown in red. Samples were counterstained for acetylated tubulin (acTub, cyan). Allimages are anterior views of Platynereis larvae 72 h post fertilization (hpf). Scale bar: 50 μm.Conzelmann and Jékely EvoDevo 2012, 3:23Page 4 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 52in situ hybridization (Figure 3 A0-E0). The recentlydescribed antibodies raised against full length Platy-nereis DLa, FVa, FLa and RYa peptides also show verysimilar neuronal signals [23].Overall, our specificity tests in Platynereis demonstratethat the antibodies raised against amidated dipeptidemotifs are remarkably specific and can be used to obtainhigh-quality tissue stainings. To test the utility of ourantibody collection as cross-species-reactive neuronalmarkers, we performed immunostainings on a variety ofmarine larvae from different species and phyla. DLamide immunoreactivity in annelidsDLa neuropeptides have been described from the errantannelid (Errantia) Platynereis and the sedentary annelids(Sedentaria) Capitella and Helobdella [23,36] (Figure 1A).Since errant and sedentary annelids encompass most ofannelid diversity [37], DLa neuropeptides are potentiallywidely distributed among annelids. To test whether ourDLa antibody could be used as a pan-annelid nervous sys-tem marker, we also tested its reactivity in Capitella. InCapitella larvae, we found staining in neurons of theapical organ. These neurons have a flask-shaped morph-ology typical of sensory cells and project to the larvalciliary band (Figure 4A, arrow) in a similar fashion to thatobserved for Platynereis larvae (compare with Figure 2A).We also observed strong staining in the ventral nerve cordin older Capitella larvae (Figure 4B). The specific reactivityof the DLa antibody in both errant and sedentary annelidspecies demonstrates its usefulness as a pan-annelid neur-onal marker. FVamide and FLamide immunoreactivity in annelids andmolluscsFVa and FLa neuropeptides have been described inannelids, molluscs and platyhelminths. In the annelidsPlatynereis and Capitella, there is one FVa neuropeptideprecursor, whereas there are three different precursorsin the mollusc Lottia gigantea (Figure 1B,C) [35]. FLapeptides are either encoded by a separate precursor geneand expressed in distinct subsets of cells, as in annelids,or co-occur on the same precursor together with FVapeptides, as in molluscs. Regardless of the number ofprecursor genes, the conserved FVa and FLa epitopescould allow the labelling of all FVa and FLa expressingneurons in annelids and molluscs. We tested the reacti-vity of both antibodies on Capitella larvae and on larvaeof the bivalve mollusc Pecten maximus and the nudi-branch mollusc Phestilla sibogae (for morphologicaldetails see [38]). In Capitella, we found FVa immuno-reactivity in apical organ neurons with projections to theciliary band (Figure 5A, arrow, compare with Figure 2B),and also in the ventral nerve cord (Figure 5B). The FLaantibody labels neurons in the brain and in the ventralnerve cord in older stages of Capitella (Figure 5F). InPecten veliger larvae, the FVa antibody labels a smallnumber of neurons in the cerebral and visceral ganglia,some of which project to the ciliated velum (Figure 5C).In Phestilla, both antibodies show strong staining in thecerebropleural ganglion between the eyes (Figure 5D,G).The FVa antibody also labels two nerve fibres in the cili-ated foot (Figure 5E, arrows). GWamide immunoreactivity in annelids, molluscs andcrustaceansGWa neuropeptides are present in annelids, molluscs(APGWa), platyhelminths, crustaceans (as red pigmentconcentrating hormone, RPCH) and insects (as adipoki-netic hormone, AKH, Figure 1D). Although the sequencesimilarity is limited, the annelid and mollusc GWa precur-sors are the likely lophotrochozoan orthologues of arthro-pod RPCH and AKH neuropeptide precursors [39]. Wetested our GWa antibody in the annelid Capitella, themolluscs Pecten and Phestilla and a nauplius larvafrom a cirripede crustacean (for morphological details,see [40]) collected from a plankton sample (Figure 6).Like Platynereis, Capitella larvae show staining in a smallnumber of neurons in the apical organ (Figure 6A, comparewith Figure 2D) and in the ventral nerve cord of olderlarvae (Figure 6B), with no ciliary innervation. In Pecten Figure 4 DLamide immunoreactivity in Capitella larvae. (A) Anterior view of an early Capitella larva and (B) ventral view of a late Capitellalarva stained with the DLamide antibody (red) counterstained for acetylated tubulin (acTub, white). Scale bars: 50 μm. cbn, ciliary band nerve.Conzelmann and Jékely EvoDevo 2012, 3:23Page 5 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 53veliger larvae, the GWa antibody labels a small number ofneurons and their projections (Figure 6C). In Phestilla, wefound staining in the cerebropleural ganglion between theeyes (Figure 6D). In the crustacean larvae, the antibodylabels two cerebral neurons that project to the protocereb-ral neuropil and a pair of neurons on either side of thelabrum (Figure 6E). RYamide immunoreactivity in cnidarian, annelid,bryozoan, mollusc and crustacean larvaeRYa neuropeptides have been described in a number ofmarine phyla, including cnidarians, annelids, molluscs,platyhelminthes and crustaceans. They are also presentin terrestrial invertebrates, such as nematodes andinsects (Figure 1E). In cnidarians, platyhelminthes andnematodes, RYa peptides co-occur with RFa peptides onthe same precursor [33,34,41], whereas in most otherphyla they originate from a distinct precursor. Given thebroad phyletic distribution of RYa peptides and theobservation that they often derive from distinct precur-sors expressed in different cells than RFa [23], the RYaantibody could have a great value for comparativeneuroanatomical studies. To explore the potential of theRYa antibody, we tested its reactivity in cnidarians, anne-lids, molluscs, bryozoans and crustaceans.In Capitella larvae, we found RYa staining in indivi-dual sensory neurons in the apical organ. As with DLaand FVa, these neurons project to the ciliary band nerveFigure 5 FVamide and FLamide immunoreactivity in annelid and mollusc larvae. Immunostainings with the FVamide antibody (red)counterstained for acetylated tubulin (acTub, white) in (A) an early Capitella larva, anterior view, (B) a late Capitella larva, ventral view, (C) a Pectenveliger larva, lateral view, and (D, E) a Phestilla larva, ventral (D) and dorsal (E) views. Immunostainings with the FLamide antibody (red)counterstained for acetylated tubulin (acTub, white) in (F) a late Capitella larva, ventral view, and (G) a Phestilla larva, ventral view. Arrows in (C)and (E) point at projections that run along the ciliated velum of Pecten and the ciliated foot of Phestilla. Scale bars: 50 μm. cbn, ciliary band nerve;cg, cerebral ganglion; cpg, cerebropleural ganglion; DIC, differential interference contrast; vg, visceral ganglion.Conzelmann and Jékely EvoDevo 2012, 3:23Page 6 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 54(Figure 7A, arrow, compare with Figure 2E), suggestinga role for RYa neuropeptides in regulating ciliary activityin Capitella. We also observed a strong staining in theventral nerve cord in older Capitella larvae (Figure 7B).In larvae of the bryozoan Cryptosula species (for mor-phological details see [42]), we detected strong RYaimmunoreactivity in the nerve nodule and in the lateralnerves projecting to the coronal ciliary band (Figure 7C,arrows) and several axons that embrace the pyriformorgan. In Pecten larvae, we detected two pairs of neu-rons and their projections along the ciliated velum(Figure 7D, arrow). Using primary antibodies directlypre-labelled with different fluorophores, we also co-stained Pecten larvae for RYa and FVa. We observed co-labelling only in a subset of neurons, suggesting thatthese cells co-express RYa and FVa neuropeptides(Figure 7E, asterisks). This result demonstrates thatthese antibodies can also be used in combination in asingle specimen. In Phestilla larvae, we detected a strongRYa signal in the apical organ, the cerebropleural gan-glion between the eyes and in the pedal ganglion, as wellas in nerves connecting these ganglia (Figure 7F), andalso in projections running to the ciliated foot(Figure 7G, arrows). In the nauplius larvae, we foundRYa in a group of cerebral neurons that have a flask-shaped morphology and in various neurons surroundingthe labrum (Figure 7H). In the cnidarians Aurelia andClava, we detected RYa staining in sensory neurons ofthe ciliated planula larvae, mainly located at the aboralpole (Figure 7I-K). These neurons have sensory morph-ology with apical sensory dendrites projecting to the sur-face of the ciliated neuroectoderm (Figure 7K). Thebasal neuronal projections run along the basal side ofthe ciliated neuroectoderm and terminate in the anteriorplexus. These results show that the RYa antibody is aneuronal marker widely applicable across several in-vertebrate phyla. It should be noted that the RYa anti-body may cross-react with invertebrate neuropeptidesbelonging to the NPF/NPY (short neuropeptide F,NPF; short neuropeptide Y, NPY) family, that some-times have a C-terminal RYa, such as in Apis melliferaand Bombyx mori NPFs [43]. DiscussionAmidated dipeptide epitopes allow the generation ofspecific antibodiesWe have shown, using a variety of species, that ourantibodies against amidated dipeptides can be used tolabel distinct subsets of peptidergic neurons (Figure 8,confocal stacks are available [24]). Our finding thatseveral different amidated dipeptides could be usedto generate specific antibodies broadens our under-standing of the immunogenic potential of peptidesequences. It is interesting to note that the company wecontacted for antibody production initially warned us notto carry out the project, arguing that ‘It is considered thatup to 5aa peptides are not immunogenic at all.’ The strongimmunogenicity of these peptides must be due to theirFigure 6 GWamide immunoreactivity in annelid, mollusc and crustacean larvae. Immunostainings with the GWamide antibody (red)counterstained for acetylated tubulin (acTub, white) in (A) an early Capitella larva, anterior view, (B) a late Capitella larva, ventral view, (C) a Pectenveliger larva, lateral view, (D) a Phestilla larva, ventral view, (E) and a crustacean larva, ventral view and(E0) a close-up confocal scan. Scale bars:50 μm. cpg, cerebropleural ganglion; DIC, differential interference contrast; la, labrum; pc-Np, protocerebral neuropil.Conzelmann and Jékely EvoDevo 2012, 3:23Page 7 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 55C-terminal amidation, in the context of the two residues.It is not the amide group that is recognized alone, since allpeptides have it, yet we see no cross-reactivity. The im-portance of amidation, and not the two amino acids alone,is supported by the observation that such dipeptide motifscan be found in thousands of other proteins (for example,61,717 DL, 32,582 FV, 9,459 GW and 18,792 RY in theCapitella predicted proteome), yet we do not see strongbackground staining in our immunostainings on many dif-ferent invertebrate species.Figure 7 RYamide immunoreactivity in annelid, bryozoan, mollusc, crustacean and cnidarian larvae. Immunostainings with the RYamideantibody (red) counterstained for acetylated tubulin (acTub, white) in (A) an early Capitella larva, anterior view, (B) a late Capitella larva, ventralview, (C) a Cryptosula larva, anterior view, (D) a Pecten veliger larva, anterior view, (E) a Pecten veliger larva counterstained with the anti-FVamideantibody (green), anterior view, (F) a Phestilla larva, ventral and (G) dorsal view, (H) a crustacean larva, ventral view and(H0) a close-up confocalscan. Immunostainings with the RYamide antibody (red) counterstained for acetylated or tyrosylated tubulin (acTub, white) in (I) a Clava planulalarva, lateral view, (J) an Aurelia planula larva, lateral view, and (K) close-up of a sensory neuron of Aurelia. Arrows in (C) point at axons thatproject to the ciliary band of Cryptosula. Arrows in (D) and (G) point at projections to the ciliated velum of Pecten and the ciliated foot of Phestilla.Asterisks in (E) indicate neurons that are co-labelled for RYa and FVa. Scale bars: (A-J) 50 μm, (K) 10 μm. ao, apical organ; cbn, ciliary band nerve;cg, cerebral ganglion; cpg, cerebropleural ganglion; DIC, differential interference contrast; la, labrum; pg, pedal ganglion; po, pyriform organ; vg,visceral ganglion.Conzelmann and Jékely EvoDevo 2012, 3:23Page 8 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 56Overall, our data argue that the antibodies stronglyand specifically bind the amidated peptides we used forimmunization. First, the stringent affinity purificationprotocol we employed together with the peptide-blocking experiments indicates that the antibodiesstrongly bind to the short amidated peptides. Sec-ond, the specific neuronal stainings in tissues corre-sponding to the expression patterns of the precursorgenes in Platynereis show that the antibodies specif-ically bind to the respective peptides.The strategy we employed to generate specific cross-species antibodies could also be applied to other con-served neuropeptides present in diverse taxa. With theincreasing sampling of metazoan genomes and transcrip-tomes, and the accumulation of data from understudiedgroups (for example, hemichordates, platyhelminths,priapulids), we will have the chance to identify furtherconserved peptide motifs. Further sampling will alsoallow the identification of other taxonomic groups inwhich the antibodies described here could be used asneuronal markers. Given the brevity of the sequences,reactivity to multiple neuropeptide families with the sameamidated termini cannot be excluded. For the proper inter-pretation of staining patterns, it is therefore also importantto study mRNA expression and to scrutinize available tran-scriptomic and genomic resources. Importantly, our resultsFigure 8 Summary diagrams of peptidergic cells in various species. Summary diagrams showing the position of peptidergic cells labelledwith our cross-species reactive neuropeptide antibodies for Platynereis, anterior view (A), Capitella, anterior view (B), Pecten, lateral view (C),Phestilla, ventral view (D), Cryptosula, anterior view (E), Aurelia, lateral view (F) and the crustacean larva, ventral view (G). co-Cn, circumoralconnective; crn, coronal ring nerve; dcc, dorsal branch of the circum-oesophageal connectives; ln, lateral nerve; m, mouth; mb, mandible;Pc-Np protocerebral neuropil; po, pyriform organ; vcc, ventral branch of the circum-oesophageal connectives.Conzelmann and Jékely EvoDevo 2012, 3:23Page 9 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 57show that these antibodies are not widely cross-reactiveand do not recognize other amidated peptides. A singleamino acid change seems to be sufficient to prevent anti-body binding, since the DLa, FLa and FVa antibodies allrecognize different cells.Finally, C-terminal amidation is commonly used forimmunization for peptides that derive from an internalpart of the protein, to keep the peptide closer to its nat-ural state. Our results caution that such an unnaturalterminal amide in internal peptide sequences may triggeran undesired immune response, and potentially causecross-reactivity to naturally occurring amidated peptides. Cross-species antibodies suggest that theneuropeptidergic control of cilia is widespread in marinelarvaeWith the DLa, FVa and RYa antibodies, we commonlyobserve projections to larval ciliary bands. We have re-cently shown that the neurons expressing these neuro-peptides also innervate the ciliary band in Platynereislarvae, and that these peptides regulate the activity ofcilia. All three peptides increase the beating frequency ofcilia and inhibit ciliary arrests, thereby influencing theswimming depth of planktonic Platynereis larvae [23]. InCapitella larvae, all three neuropeptides are present inthe ciliary band nerve. In Pecten, we found FVa and RYaimmunoreactivity in nerves running along the ciliatedvelum, and in Phestilla in projections in the ciliated foot.RYa neurons also seem to innervate locomotor cilia inbryozoan and cnidarian larvae. This suggests that thesepeptides may also regulate ciliary activity in these larvae,indicating a general role for neuropeptides in the regula-tion of ciliary locomotion in marine invertebrate larvae[44]. ConclusionsWe developed specific cross-species reactive antibodies thatrecognize the conserved neuropeptide motifs DLamide,FVamide, FLamide, GWamide and RYamide. These anti-bodies can be used in a wide range of marine invertebrates,including annelids, molluscs, bryozoans and cnidarians.Further genomic and transcriptomic sampling couldidentify other animal groups where these peptide motifsare conserved and where our antibodies could also beemployed. Our work also highlights the antigenic po-tential of very short amidated peptide motifs. The on-going sampling of neuropeptide diversity will allow thedevelopment of other similar antibodies, to enrich fur-ther the comparative neurobiology toolbox. Our sam-pling across diverse marine larvae demonstrates thebroad utility of these antibodies, and also indicates thatthe neuropeptidergic regulation of ciliary locomotionmay be a general feature of marine ciliated larvae.AbbreviationsacTub: Acetylated tubulin; AKH: Adipokinetic hormone; ao: Apical organ;cbn: Ciliaryband nerve; cg: Cerebral ganglion; co-Cn: Circumoral connective;cpg: Cerebropleural ganglion; crn: Coronal ring nerve; Dcc: Dorsal branch ofthe circum-oesophageal connectives; DIC: Differential interference contrast;Hpf: Hours post fertilization; la: Labrum; ln: Lateral nerve; m: Mouth;mb: Mandible NPF; F: Short neuropeptide; NPY: Short neuropeptide;Pc-Np: Protocerebral neuropil; pg: Pedal ganglion; po: Pyriform organ;RPCH: Red pigment concentrating hormone; tyrTub: Tyrosylated tubulin;vcc: Ventral branch of the circum-oesophageal connectives; vg: Visceralganglion. Competing interestsA patent application has been submitted. Authors’ contributionMC carried out antibody purification, tissue stainings and imaging and wrotethe paper. GJ conceived the study, participated in its design and wrote thepaper. Both authors read and approved the final manuscript. AcknowledgementWe thank Harald Hausen for our Capitella stock and Andreas Bick for liveAurelia. We thank Elizabeth A. Williams for collecting Clava larvae and forcomments on the manuscript, Nadine Randel and Michael G. Hadfield forPhestilla larvae, and Mechtild Seyboldt for collecting Pecten larvae. The Clavawork was supported by an ASSEMBLE grant 227799. The research leading tothese results received funding from the European Research Council underthe European Union’s Seventh Framework Programme (FP7/2007-2013)/European Research Council Grant Agreement 260821. Received: 26 April 2012 Accepted: 23 July 2012Published: 1 October 2012 References1. Hay-Schmidt A: The evolution of the serotonergic nervous system. ProcBiol Sci 2000, 267:1071–1079.2. Frontali N, Williams L, Welsh JH: Heart excitatory and inhibitorysubstances in molluscan ganglia. Comp Biochem Physiol 1967, 22:833–841.3. Price DA, Greenberg MJ: Structure of a molluscan cardioexcitatoryneuropeptide. Science 1977, 197:670–671.4. Dockray GJ, Reeve JR, Shively J, Gayton RJ, Barnard CS: A novel activepentapeptide from chicken brain identified by antibodies toFMRFamide. Nature 1983, 305:328–330.5. Dockray GJ, Vaillant C, Williams RG: New vertebrate brain-gut peptiderelated to a molluscan neuropeptide and an opioid peptide. Nature 1981,293:656–657.6. Walker RJ, Papaioannou S, Holden-Dye L: A review of FMRFamideandRFamide-like peptides in metazoa. Invertebr Neurosci 2009, 9:111–153.7. Grimmelikhuijzen CJ: Antisera to the sequence Arg-Phe-amide visualizeneuronal centralization in hydroid polyps. Cell Tissue Res 1985, 241:171–182.8. Walker RJ: Neuroactive peptides with an RFamide or Famide carboxylterminal. Comp Biochem Physiol C, Comp Pharmacol Toxicol 1992, 102:213–222.9. Heuer CM, Loesel R: Immunofluorescence analysis of the internal brainanatomy of Nereis diversicolor (Polychaeta, Annelida). Cell Tissue Res 2008,331:713–724.10. Brinkmann N, Wanninger A: Neurogenesis suggests independentevolution of opercula in serpulid polychaetes. BMC Evol Biol 2009, 9:270.11. Raikova OI, Reuter M, Jondelius U, Gustafsson MK: The brain of theNemertodermatida (Platyhelminthes) as revealed by anti-5HT andanti-FMRFamide immunostainings. Tissue Cell 2000, 32:358–365.12. Stegner MEJ, Richter S: Morphology of the brain in Hutchinsoniellamacracantha (Cephalocarida, Crustacea). Arthropod Struct Dev 2011,40:221–243.13. Kristof A, Wollesen T, Wanninger A: Segmental mode of neural patterningin Sipuncula. CurrBiol 2008, 18:1129–1132.14. Chun JY, Korner J, Kreiner T, Scheller RH, Axel R: The function anddifferential sorting of a family of aplysia prohormone processingenzymes. Neuron 1994, 12:831–844.15. Hook V, Funkelstein L, Lu D, Bark S, Wegrzyn J, Hwang S-R: Proteases forprocessing proneuropeptides into peptide neurotransmitters andhormones. Annu Rev Pharmacol Toxicol 2008, 48:393–423.Conzelmann and Jékely EvoDevo 2012, 3:23Page 10 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 5816. Edison AS, Espinoza E, Zachariah C: Conformational ensembles: the role ofneuropeptide structures in receptor binding. J Neurosci 1999, 19:6318–6326.17. Eipper BA, Stoffers DA, Mains RE: The biosynthesis of neuropeptides:peptide α-amidation. Annu Rev Neurosci 1992, 15:57–85.18. Han S-K, Dong X, Hwang J-I, Zylka MJ, Anderson DJ, Simon MI: Orphan Gprotein-coupled receptors MrgA1 and MrgC11 are distinctively activatedby RF-amide-related peptides through theGαq/11 pathway. P Natl AcadSci USA 2002, 99:14740–14745.19. Minth CD, Qiu H, Akil H, Watson SJ, Dixon JE: Two precursors ofmelanin-concentrating hormone: DNA sequence analysis and in situimmunochemical localization. P Natl Acad Sci USA 1989, 86:4292–4296.20. Nahon JL, Presse F, Bittencourt JC, Sawchenko PE, Vale W: The ratmelanin-concentrating hormone messenger ribonucleic acid encodesmultiple putative neuropeptides coexpressed in the dorsolateralhypothalamus. Endocrinology 1989, 125:2056–2065.21. Maillère B, Hervé M: The specificity of antibodies raised against a T cellpeptide is influenced by peptide amidation. Mol Immunol 1997,34:1003–1009.22. In Y, Ono H, Ishida T: Structural studies on C-amidated amino acids andpeptides: function of amide group in molecular association in crystalstructures of Val-Gly-NH2, Ser-Phe-NH2, Gly-Tyr-NH2 and Pro-Tyr-NH2hydrochloride salts. Chem Pharm Bull 2002, 50:571–577.23. Conzelmann M, Offenburger S-L, Asadulina A, Keller T, Münch TA, Jékely G:Neuropeptides regulate swimming depth of Platynereis larvae. Proc NatlAcad Sci U S A 2011, 108:E1174–E1183.24. Jékely G: Max Planck Institute for Developmental Biology; Jékely Lab.http://jekely-lab.tuebingen.mpg.de/.25. Jékely G, Arendt D: Cellular resolution expression profiling using confocaldetection of NBT/BCIP precipitate by reflection microscopy. Biotechniques2007, 42:751–755.26. Schneider CA, Rasband WS, Eliceiri KW: NIH Image to ImageJ: 25 years ofimage analysis. Nat Methods 2012, 9:671–675.27. Southey BR, Amare A, Zimmerman TA, Rodriguez-Zas SL, Sweedler JV:NeuroPred: a tool to predict cleavage sites in neuropeptide precursorsand provide the masses of the resulting peptides. Nucleic Acids Res 2006,34:W267–72.28. Petersen TN, Brunak S, von Heijne G, Nielsen H: SignalP 4.0: discriminatingsignal peptides from transmembrane regions. Nat Methods 2011, 8:785–786.29. Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, Higgins DG, ThompsonJD: Multiple sequence alignment with the Clustal series of programs.Nucleic Acids Res 2003, 31:3497–3500.30. Christie AE, Cashman CR, Brennan HR, Ma M, Sousa GL, Li L, Stemmler EA,Dickinson PS: Identification of putative crustacean neuropeptides usingin silico analyses of publicly accessible expressed sequence tags. GenComp Endocrinol 2008, 156:246–264.31. Christie AE, Stemmler EA, Dickinson PS: Crustacean neuropeptides. Cell MolLife Sci 2010, 67:4135–4169.32. Gäde G, Hoffmann KH, Spring JH: Hormonal regulation in insects: facts,gaps, and future directions. Physiol Rev 1997, 77:963–1032.33. Gajewski M, Schmutzler C, Plickert G: Structure of neuropeptide precursorsin Cnidaria. Ann N Y Acad Sci 1998, 839:311–315.34. Mcveigh P, Mair GR, Atkinson L, Ladurner P, Zamanian M, Novozhilova E,Marks NJ, Day TA, Maule AG: Discovery of multiple neuropeptide familiesin the phylum Platyhelminthes. Int J Parasitol 2009, 39:1243–1252.35. Veenstra JA: Neurohormones and neuropeptides encoded by thegenome of Lottia gigantea, with reference to other molluscs and insects.Gen Comp Endocrinol 2010, 167:86–103.36. Veenstra JA: Neuropeptide evolution: neurohormones and neuropeptidespredicted from the genomes of Capitella teleta and Helobdella robusta.Gen Comp Endocrinol 2011, 171:160–175.37. Struck TH, Paul C, Hill N, Hartmann S, Hösel C, Kube M, Lieb B, Meyer A,Tiedemann R, Purschke G, Bleidorn C: Phylogenomic analyses unravelannelid evolution. Nature 2011, 471:95.38. Croll RP: Development of embryonic and larval cells containingserotonin, catecholamines, and FMRFamide-related peptides in thegastropod mollusc Phestilla sibogae. Biol Bull 2006,211:232–247.39. Martínez-Pérez F, Becerra A, Valdés J, Zinker S, Aréchiga H: A possiblemolecular ancestor for mollusc APGWamide, insect adipokinetichormone, and crustacean red pigment concentrating hormone. J MolEvol2002, 54:703–714.40. Semmler H, Wanninger A, Høeg JT, Scholtz G: Immunocytochemicalstudies on the naupliar nervous system of Balanus improvisus (Crustacea,Cirripedia, Thecostraca). Arthropod Struct Dev 2008, 37:383–395.41. Li C, Kim K, Nelson LS: FMRFamide-related neuropeptide gene family inCaenorhabditis elegans. Brain Res 1999, 848:26–34.42. Santagata S: Evolutionary and structural diversification of the larvalnervous system among marine bryozoans. Biol Bull 2008, 215:3.43. Nässel DR, Wegener C: A comparative review of short and longneuropeptide F signaling in invertebrates: any similarities to vertebrateneuropeptide Y signaling? Peptides 2011, 32:1335–1355.44. Jékely G: Origin and early evolution of neural circuits for the control ofciliary locomotion. P R Soc B 2011, 278:914–922. doi:10.1186/2041-9139-3-23Cite this article as: Conzelmann and Jékely: Antibodies againstconserved amidated neuropeptide epitopes enrich the comparativeneurobiology toolbox. EvoDevo 2012 3:23. Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission • Thorough peer review • No space constraints or color figure charges • Immediate publication on acceptance • Inclusion in PubMed, CAS, Scopus and Google Scholar • Research which is freely available for redistribution Submit your manuscript atwww.biomedcentral.com/submitConzelmann and Jékely EvoDevo 2012, 3:23Page 11 of 11http://www.evodevojournal.com/content/3/1/23 Appendix : Publ icat ions 59Publication 3: Conserved MIP receptor-ligand pairregulates Platynereis larval settlement; PNAS (2013) Conzelmann, Williams et al. Appendix : Publ icat ions 60Conserved MIP receptor–ligand pair regulatesPlatynereis larval settlement Markus Conzelmann, Elizabeth A. Williams, Sorin Tunaru, Nadine Randel, Réza Shahidi, Albina Asadulina,JürgenBerger, Stefan Offermanns, and Gáspár Jékely Max Planck Institute for Developmental Biology, 72076 Tuebingen, Germany; and Max Planck Institute for Heart and Lung Research, 61231 BadNauheim, Germany Edited by Cornelia Bargmann, The Rockefeller University, New York, NY, and approved March 14, 2013 (received for review November 21, 2012) Life-cycle transitions connecting larval and juvenile stages in meta-zoans are orchestrated by neuroendocrine signals including neuro-peptides and hormones. In marine invertebrate life cycles, whichoften consist of planktonic larval and benthic adult stages, settle-ment of the free-swimming larva to the sea floor in response toenvironmental cues is a key life cycle transition. Settlement isregulated by a specialized sensory–neurosecretory system, the lar-val apical organ. The neuroendocrine mechanisms through whichthe apical organ transduces environmental cues into behavioralresponses during settlement are not yet understood. Here we showthat myoinhibitory peptide (MIP)/allatostatin-B, a pleiotropic neu-ropeptide widespread among protostomes, regulates larval settle-ment in the marine annelid Platynereis dumerilii. MIP is expressedin chemosensory–neurosecretory cells in the annelid larval apicalorgan and signals to its receptor, an orthologue of the Drosophilasex peptide receptor, expressed in neighboring apical organ cells.We demonstrate by morpholino-mediated knockdown that MIPsignals via this receptor to trigger settlement. These results reveala role for a conservedMIP receptor–ligand pair in regulating marineannelid settlement. M life cycles show great diversity in larval, juvenile,and adult forms, as well as in the timing and ecologicalcontext of the transitions between these forms. In many animalspecies, neuroendocrine signals involving hormones and neuro-peptides regulate life cycle transitions (1–3). Environmental cuesare often important instructors of the timing of life cycle tran-sitions (4), and can affect behavioral, physiological, or morpho-logical change via neuroendocrine signaling (5).Marine invertebrate larval settlement is a prime example ofthe strong link between environmental cues and the timing oflife-cycle transitions. Marine invertebrate life cycles often consistof a free-swimming (i.e., pelagic) larval stage that settles to theocean floor and metamorphoses into a bottom-dwelling (i.e.,benthic) juvenile (6–8). In many invertebrate larvae, a pelagic–benthic transition is induced by chemical cues from the envi-ronment (9, 10). Larval settlement commonly includes the ces-sation of swimming and the appearance of substrate exploratorybehavior, including crawling on or attachment to the substrate(11–14). In diverse ciliated marine larvae (15), the apical organ,an anterior cluster of larval sensory neurons (16) with a strongneurosecretory character (17–20), has been implicated in thedetection of cues for the initiation of larval settlement (21).Although molecular markers of the apical organ have been de-scribed (22–24), our knowledge of the neuroendocrine mecha-nisms with which apical organ cells transmit signals to initiatelarval settlement behavior is incomplete.Here, we identify a conserved myoinhibitory peptide (MIP)/allatostatin-B receptor–ligand pair as a regulator of larval set-tlement behavior in the marine polychaete annelid Platynereisdumerilii. MIPs are pleiotropic neuropeptides (25) first describedin insects as inhibitors of muscle contractions (26, 27). In someinsect species, MIPs modulate juvenile hormone (28) or ecdysonesynthesis (29), and are also referred to as allatostatin-B or pro-thoracicostatic peptide (30, 31). These peptides are known tosignal via a G protein-coupled receptor, the sex peptide receptor(SPR) (31, 32). MIPs show sequence similarity to cnidarianGLWamides and belong to an ancestral eumetazoan Wamidefamily that also includes mollusk APGWamides and other pep-tides (33). We found that Platynereis MIP is expressed in chemo-sensory–neurosecretory cells in the apical organ and triggers larvalsettlement behavior by signaling via an SPR orthologue expressedin adjacent apical organ cells. Our results identify a conservedneuropeptide receptor–ligand pair in the apical organ, which maytransduce environmental signals to initiate settlement in a pelagic–benthic life cycle. ResultsMIP Triggers Larval Settlement in Platynereis. Platynereis has a pe-lagic–benthic life cycle with a freely swimming ciliated larval stagethat spends as long as several days in the plankton before tran-sitioning to a benthic lifestyle (34, 35). Searching for regulators ofannelid settlement, we identified a neuropeptide, an orthologueof arthropod MIP (Fig. S1 A and B), that efficiently triggers set-tlement of Platynereis larvae. The peptide is derived from a pre-cursor protein that shows high sequence similarity to the MIPpreprotein from the distantly related polychaete Capitella teleta(36, 37) and to MIP precursors from arthropods (also called alla-tostatin-B or prothoracicostatic peptide) (30). The predicted ma-ture protostome MIP peptides share a conservedW(X5–8)Wsequence motif and are amidated (Fig. S1B). We did not identifyany MIP orthologue in deuterostomes.Treatment of free-swimming Platynereis larvae in a verticalswimming assay (19) with synthetic PlatynereisMIP peptide rapidlyinduced downward vertical movement in trochophore and necto-chaete larval stages (Fig. 1A, Fig. S2, andMovies S1, S2, and S3), aneffect that could be reversed by washout (Fig. 1A). Different ver-sions of MIPs derived from the same Platynereis precursor protein,including a nonamidated MIP, also rapidly triggered larval sinking(Fig. S2 A–E). We focused on MIP7 in subsequent experimentsbecause this peptide closely matches the consensus sequence de-rived from all Platynereis MIPs (Fig. S1 C and D). Substitution ofthe last (W10A), but not the first (W2A), tryptophan residue toalanine rendered MIP7 inactive in the vertical swimming assay Author contributions: M.C., E.A.W., S.O., and G.J. designed research; M.C., E.A.W., S.T.,N.R., R.S., J.B., and G.J. performed research; M.C. and E.A.W. contributed new reagents/analytic tools; M.C., E.A.W., S.T., N.R., A.A., S.O., and G.J. analyzed data; and M.C., E.A.W.,and G.J. wrote the paper. Conflict of interest statement: A patent application on the potential use of MIP/allatos-tatin-B has been submitted. This article is a PNAS Direct Submission. Freely available online through the PNAS open access option. Data deposition: The sequences reported in this paper have been deposited in the GenBankdatabase (accession nos. JX513876, JX513877, and JX513878). M.C. and E.A.W. contributed equally to this work.To whom correspondence should be addressed. E-mail: [email protected]. This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1220285110/-/DCSupplemental. www.pnas.org/cgi/doi/10.1073/pnas.1220285110PNAS Early Edition | 1 of 6NEUROSCIENCE Appendix : Publ icat ions 61(Fig. 1A). To understand the mechanism of downward movement,we analyzed the effects ofMIP7 treatment on ciliary activity. MIP7did not significantly alter the beat frequency of cilia (Fig. S2G), buttriggered long and frequent ciliary arrests in a concentration-de-pendent manner (Fig. 1B). Close-up videos of 2-d-old larvae trea-ted with MIP7 in the vertical column showed that the larvae weresinking with their anterior pointed upward (Movie S2). Larvae atage 3 d showed downward movement with their anterior pointeddownward, implying more complex behavioral effects in olderlarvae (Movie S3).After MIP7-treated larvae reached the bottom of the culturedish, they showed sustained exploratory crawling behavior, withfrequent touching of the apical side to the substrate (Fig. 1 C andD and Movies S4 and S5). Such substrate exploratory behaviorcould also be observed at low frequency for control larvae, with latenectochaete stages [5–6 d postfertilization (dpf)] showing a greatertendency to contact the substrate.MIP7 treatment strongly inducedsustained substrate contact between 1 and 6 d of development (Fig.1C). Another Platynereis neuropeptide, AKYFLamide, which haspreviously been shown to trigger larval sinking (19), was inactive inthe crawling assay even after 90 min of incubation (Fig. 1D).Overall,MIP treatment can rapidly induce two distinct behaviors inPlatynereis larvae: (i) inhibition of cilia and cessation of swimmingand (ii) crawling on the substrate, both considered hallmarks ofmarine invertebrate larval settlement (11–14). These results iden-tify Platynereis MIP as a settlement-inducing neuropeptide. MIP Is Expressed in Neurosecretory Annelid Apical Organ. Whole-mount in situ hybridization on Platynereis larvae revealed MIPmRNA expression in sensory cells of the apical organ from 20 hpostfertilization (hpf) on, as well as two pairs of cells in the trunkfrom 48 hpf on. The apical organ expression was observed in anincreasing number of cells with age (Fig. 2 A–D and Fig. S3B–H). Immunostaining with a specific MIP antibody (Fig. S3A)showed that the axonal projections of these cells terminate in theapical nerve plexus (Fig. 2 E and F), a region of strong neuro-secretory activity (18, 19). By using the Platynereis MIP antibodyor an antibody against the conserved C-amidated dipeptideVWamide that is strongly conserved in mollusks and annelids(Figs. S1B and S3A), we observed similar immunolabeling insensory cells in the larval apical organ and in the nerve plexus ofCapitella (Fig. 1 G and H and Fig. S3I).To confirm the neurosecretory nature of the MIP-expressingcells, we analyzed MIP coexpression with known neuroendocrinemarkers by using image registration of in situ hybridization scans toan average anatomical reference template (19, 38, 39). We scan-ned, registered, and averaged at least five individual larvae pergene fromwhole-mount in situ hybridization samples.We analyzedthe neurosecretory marker prohormone convertase prohormoneconvertase 2 (phc2) and the neuroendocrine transcription factorsorthopedia (otp) (18) and dimmed (dimm) (Fig. S4 A–D). In Dro-sophila, dimm directs the differentiation of neuroendocrine neu-rons and is coexpressed with MIP in the median brain (40, 41). Invertebrates, otp is required for the terminal differentiation of hy-pothalamic neuropeptidergic neurons (42). Image registrationrevealed that the average MIP signal colocalized with phc2, andpartially overlapped with otp and dimmaverage gene expressionpatterns. Additionally, otp and dimm colocalized broadly in thePlatynereis apical organ neurons (Fig. S4 E–J). These resultsindicate that the MIP neurons are part of the neurosecretoryapical organ. Platynereis MIP Neurons Have Dual Chemosensory–NeurosecretoryFunction. To investigate the sensory modality of the MIP neurons,we performed dye-filling experiments on live 30-h-old Platynereislarvae. In nematodes, dye-filling is the exclusive property of che-mosensory neurons (43). Upon incubation of the larvae with fluo-rescent MitoTracker dye, we observed labeling of several flask-shaped neurons in the larval episphere, with long apical microvilliat the tip of their dendrites, including two prominent apical organcells (Fig. 3 A and B). Laser ablation of these two apical organcells, followed by subsequent fixation and MIP immunostaining,identified the cells as the two median MIP sensory neurons(Fig. 3A–D, arrowheads). The dye-filling results together with thecharacteristic microvillar morphology of theMIP neurons suggestthat these cells likely have a chemosensory function.To further characterize the morphology of the MIP neurons inPlatynereis at an ultrastructural level, we traced the entire volumeof two MIP cells (Figs. 2 D and E and 3C, arrows) by using serial-sectioning transmission EM (TEM). In a dataset of 664 ultrathinsections (50 nm), we identified these two ventral MIP neuronsbased on the position of their cell bodies relative to large adjacentsecretory gland cells and the characteristic shape and positionof their dendrites and axons (Fig. S5 and Movies S6 and S7).Both identified neurons have a sensory dendrite with a cilium andlong apical microvillar extensions surrounding the central cilium,characteristic of annelid chemosensory neurons (44) (Fig. 3E).These extensions run at the basal side of the cuticle in a subcuticularFig. 1. MIP triggers larval settlement behavior in Platynereis. (A) Angularhistograms of the displacement vectors of swimming tracks for larvae trea-ted with DMSO (control) and larvae treated with the indicated peptide [n >100 larvae (55–60 hpf) each]. (B) Percentage of time in which cilia are closedin DMSO-treated larvae (control) vs. larvae exposed to MIP7 [n > 10 larvae(48–50 hpf) each]. (C) Percentage of 1-, 2-, 5-, and 6-dpf larvae that showedsustained substrate contact after 2 min exposure to DMSO (control), 20 μMMIP7 (for 1 dpf and 2 dpf), or 50 μM MIP7 (for 5 dpf and 6 dpf; n = 4 rep-etitions with >30 larvae). (D) Percentage of larvae that showed sustainedsubstrate contact after 90 min exposure to DMSO (control), AKYFLamide(FLa), and MIP7 [n = 10 repetitions with >30 larvae (50 hpf) each]. In A, theP values of an χ test comparing the number of upward and downwardswimming larvae are indicated as *P < 0.05 and ***P < 0.001. Data in B–Dare shown as mean ± SEM. P values of an unpaired t test are indicated as *P <0.05, **P < 0.01, and ***P < 0.001, with “ns” indicating P > 0.05. 2 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1220285110Conzelmann et al. Appendix : Publ icat ions 62space that is permeable to seawater. The axons project to thedense apical nerve plexus of the larva, where they branch ex-tensively (Fig. 3 F and G and Movie S7). We observed that thebranched axons are full of large dense-cored vesicles (Fig. 3H),but are completely devoid of synapses that are typical for cellscontaining classical neurotransmitters (45), indicating that theseMIP chemosensory neurons signal exclusively via the releaseof neuropeptides. AnnelidMIP Peptides Signal via ConservedG Protein-CoupledReceptor.In insects, MIP peptides signal via a G protein-coupled receptor,the SPR (31, 32). We identified the orthologues of insect SPRsin Platynereis and Capitella (Fig. S6) and tested whether thesereceptors could be activated by annelid MIP peptides. Ligandstimulation in CHO cells expressing the respective Platynereis orCapitella receptor, a bioluminescent Ca reporter, and a pro-miscuous G protein (46, 47) showed that annelid MIP peptidesare potent agonists for these receptors. Activation assays withincreasing concentrations of ligand showed that Platynereis MIPactivated the Platynereis receptor in the nanomolar range (EC50of 10 nM; Fig. 4A). The activation was specific, as 12 otherFig. 2. MIP is expressed in apical organ neurons in Platynereis and Capitella. (A) SEM image of a 48-hpf Platynereis larva in ventral view. (B) Whole-mountmRNA in situ hybridization for Platynereis MIP (red) in a 48-hpf larva in ventral view. (C) SEM image of a 48-hpf Platynereis larva. (D) Whole-mount mRNA insitu hybridization for MIP (red) in a 48-hpf Platynereis larva. (E and F) MIP immunostaining on a 48-hpf Platynereis larva, color-coded for depth. (G and H)MIP immunostaining on a Capitella larva, color-coded for depth. In B, D, F, and H, larvae are counterstained for acetylated α-tubulin (white, anti-acTub).(C–H) Anterior views. (D and E) Arrowheads indicate two prominent apical organ cells, and arrows indicate ventral neurons that have been reconstructed byTEM (compare with Fig. 3). (Scale bars: 50 μm.) Fig. 3. Platynereis MIP neurons are chemosensory and neurosecretory. (A and B) Chemosensory neurons with long microvilli in the apical organ filled withMitoTracker dye in a 37-hpf Platynereis larva. Arrowheads indicate two prominently labeled apical organ cells, one of which was laser-ablated. (C) MIPimmunostaining on a 52-hpf nonablated larva, color-coded for depth. (D) MIP immunostaining in a 52-hpf larva in which the MitoTracker-filled apical organcell on the right side of the body (left side of image) was ablated. (E) Apical sensory ending of a ventral MIP neuron (blue) in a TEM image. Arrow points at thebasal body of the sensory cilium, and arrowheads indicate apical microvilli. The red dashed line indicates the border of the cuticle. (F) TEM image of theneurosecretory projections of the two ventral MIP neurons filled with dense-cored vesicles (arrowhead). (G) Volume reconstruction from serial TEM sections ina 72-hpf larva of the two ventral MIP neurons indicated with arrows in C and D. The apical microvilli are shown in yellow and red. (H) Close-up of the areaindicated by the dashed box in G shows a reconstruction from serial TEM images of all dense-cored vesicles in the two ventral MIP neurons. (C and D)Arrowheads indicate two MIP neurons, one of which has been ablated in D; arrows indicate ventral neurons that have been reconstructed by TEM. (A–D)Anterior views. (Scale bars: A and B, 20 μm; C and D, 25 μm; E and F, 1 μm; G, 10 μm.) Conzelmann et al.PNAS Early Edition | 3 of 6NEUROSCIENCE Appendix : Publ icat ions 63Platynereis neuropeptides (19, 45) did not activate the MIP re-ceptor (Fig. 4B). Consistently, Capitella MIP activated theCapitella SPR orthologue (Fig. 4D). Substitution of the last, butnot the first, tryptophan residue with alanine in the Platynereisand the Capitella synthetic peptides resulted in a loss of activity(Fig. 4 C and D), indicating that the conserved C-terminal tryp-tophan residue is crucial for receptor activation. This result isconsistent with our observations in the vertical swimming assays(compare with Fig. 1A). We also found that MIP peptides fromthe two annelid species cross-activated the receptor from theother species (Fig. 4 C and D).By using mRNA in situ hybridization, we found that the MIP-receptor is expressed in cells of the Platynereis apical organ, in-terspersed between the MIP-expressing neurons with little over-lap (Fig. 4 E and F), as revealed by image registration. AverageMIP-receptor expression also colocalizes with phc2, otp, and dimmaverage gene expression, indicating that these cells are also partof a neurosecretory apical organ system (Fig. S7).We next asked if the observed effects of synthetic MIP peptideson Platynereis larval settlement were caused by signaling via theMIP receptor. To test this, we knocked down the Platynereis re-ceptor by microinjecting two different translation-blocking mor-pholinos (MOs) and a control mismatch MO into fertilizedPlatynereis eggs. In MIP-receptor–knockdown larvae, we no longerobserved an effect of MIP peptide on ciliary closures in 2-d-oldlarvae, showing that MIP triggers settlement behavior by signalingvia the MIP receptor (Fig. 4G). DiscussionMIP Orchestrates Platynereis Larval Settlement Behavior via Neu-roendocrine Signaling in Apical Organ.Understanding the molecularmachinery that operates within larval sensory structures is essentialto our understanding of how the environment shapes larval be-havior and development. The behavior evoked by synthetic MIP inPlatynereis larvae mimics the settlement behavior described forother marine larvae upon encountering a natural inductive set-tlement cue (11–14), suggesting that MIP activates a behavioralprogram for settlement. In laboratory culture, in the apparentabsence of a cue, Platynereis larvae gradually transition froma pelagic to a benthic lifestyle over a period of approximately 6 d(35). Our results suggest that, were favorable chemical cues en-countered, larvae would be competent to transition to a benthiclifestyle at an earlier stage. This is supported by the early expres-sion of the MIP precursor in sensory cells of the apical organ andthe strong settlement-inducing effects of MIP from early larvalstages on. In early larvae, MIP is exclusively expressed in the apicalorgan (Fig. S3B); therefore, by adding exogenous peptide, welikely mimic the endogenous release of MIP from these apicalsensory cells following chemosensory stimulation. A direct dem-onstration of this model will require the identification of the cur-rently unknown settlement-inducing chemicals for Platynereis andtheir chemoreceptors in the MIP cells, in combination with geneticmanipulations to show a link among environmental chemical cues,chemoreceptors, and settlement.The integrative nature of the settlement behaviors inducedby MIP (sinking followed by substrate exploration) is characteris-tic of neuroendocrine signaling, which can act simultaneouslyon a number of neurons and genes within an organism (48, 49). InPlatynereis, several other neuropeptides are expressed in sensoryneurons in the apical organ and regulate larval swimming depthduring the pelagic phase of the life cycle (19).MIP is unique amongthese neuropeptides in that it can trigger rapid sinking and sub-strate exploratory behavior. The expression of the MIP precursorbroadens and MIP effects become more complex with age, im-plying an elaboration of MIP targets during development.The chemosensory MIP cells in the apical organ are alsoneurosecretory, as they have high concentrations of dense corevesicles and express the neurosecretory cell markers phc2, otp,and dimm. These cells could directly release MIP upon sensorystimulation. The expression of the MIP-receptor in apical organcells adjacent to the MIP cells indicates that there is peptidergicparacrine signaling between the MIP sensory neurons and thereceptor-expressing cells. The settlement-inducing effects of MIPwere blocked by the MO-mediated knockdown of the receptor,indicating that the MIP receptor is required for the orchestrationFig. 4. AnnelidMIP peptides signal via aG protein-coupled receptor. (A) Dose–response curve of the Platynereis (Pdu) MIP receptor treated with varyingconcentrations of Platynereis MIP7. (B) Activation of the Platynereis MIP re-ceptor by Platynereis MIP7 exclusive of other Platynereis neuropeptides. Allpeptides were used at 1 μMconcentration. (C) Activation of the PlatynereisMIPreceptor by Ala-substituted Platynereis MIP7 and Capitella (Cte) MIP1. Allpeptides were used at 1 μM concentration. (D) Activation of the Capitella MIPreceptor by Ala-substituted Capitella MIP1 and Platynereis MIP7. All peptideswere used at 1 μM concentration. (E) Whole-mount in situ hybridization forPlatynereis MIP-receptor (red) in a 48-hpf Platynereis larva, counterstained foracetylated α-tubulin (white, anti-acTub), in anterior view. (F) Average expres-sion patterns ofMIP (red) and dimm (cyan) obtained by image registration. (G)Percentage of time inwhich cilia are closed in uninjected Platynereis larvae, andinMIP-receptor-mismatch (MM),MIP-receptor-start1 (start1), andMIP-receptor-start2 (start2) MO-injected larvae, exposed to 5 μM MIP7. In B–D, the positivecontrol represents activation of a human G-protein coupled receptor 109awith100 μM nicotinic acid. The values are shown as mean ± SEM (n = 3 measure-ments). InG, P values of an unpaired t test are indicated: ***P < 0.001 and “ns”represents P > 0.05; n> 10 larvae (48 hpf). (Scale bars: E, 50 μm; F, 20 μm.) 4 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1220285110Conzelmann et al. Appendix : Publ icat ions 64of peptide-induced settlement behavior. The molecular finger-print of the MIP-receptor–expressing neurons shows that thesecells are also neurosecretory and may release downstream sig-naling molecules in a neuroendocrine cascade. Wamide Signaling: Ancestral Component of Eumetazoan AnteriorNervous System?The conservation of MIP and its receptor inPlatynereis andCapitella indicates thatMIP signaling is widespreadamong annelids. MIP is also present in other Lophotrochozoans(Fig. S1) (50, 51). As for insects and annelids, the MIP receptor inthe sea hare Aplysia californica is also an SPR orthologue (32).In annelids and insects, a prominent domain of MIP expressionis in the most anterior neurosecretory region of the developingbrain (30), an area demarcated by the expression of the homeoboxgene six3 (20). TheMIP neurons in annelids and insect also expressdimm and phc2. These similarities identify the MIP neurons asa conserved neurosecretory cell type in the anterior protostomebrain. Although MIP seems to have been lost along the deutero-stome lineage, the coexpression of dimm and otp in the Platynereisapical organ also supports a link among insect, annelid, and ver-tebrate neuroendocrine centers (18, 52).Protostome MIPs belong to a diverse and ancient Wamideneuropeptide family that also includes cnidarian GLWamides(33), inducers of cnidarian larval metamorphosis (53–55). In cni-darians, GLWamides are expressed in sensory cells at the anteriorpole of the larva (8, 56). This anterior territory shares a conservedregulatory signature with apical organs in many bilaterian ciliatedlarvae (24, 57).MIP andLWamidemay thus represent a conservedeffector gene in the neurosecretory anterior brain of cnidariansand protostomes. The role of MIP in annelid settlement andLWamide in cnidarian metamorphosis suggests that one of theancestral functions of Wamides may have been the regulation ofa life cycle transition. MethodsGene Identification. Platynereis genes were identified from expressed se-quence tag (EST) sequences generated from a full-length normalized cDNAlibrary from mixed larval stages. Capitella genes were identified at the JointGenome Institute Genome Portal (58). Behavioral Assays. Platynereis larvae were obtained from an in-housebreeding culture, following a previous publication (59). Vertical larvalswimming and ciliary beating assays were performed and analyzed as pre-viously described (19).The substrate–contact assay was carried out in Nunclon 24-well tissueculture dishes or in customized quartz cuvettes. We increased the peptideconcentration for 5-dpf and 6-dpf larvae from 20 μM to 50 μM because, inlater stages, penetration of peptides could be impaired as a result ofa thickening of the outer larval cuticle. After peptide application, sustainedcontact of the larva with the bottom of the tissue-culture dish was scored. Alarva was scored to show sustained substrate contact behavior if it was incontact with the bottom of the dish for at least 5 s (75 frames in a 15 frames-per-second movie). Antibodies and Tissue Staining. Rabbit antisera against Platynereis MIP7 andagainst the conserved C-terminal VWamide motif were affinity-purified byusing the respective peptides coupled to a SulfoLink resin (Thermo Scientific)via an N-terminal Cys (CAWNKNSMRVWamide and CVWamide) as previouslydescribed (60). In situ hybridization was performed as previously described (39). Dye Filling and Laser Ablation. MitoTracker red FM (50 μg; special packaging;Invitrogen) was freshly dissolved in 100 μL DMSO. The solution was added to30-hpf larvae at 1:500 dilution and incubated for 1 h for optimal dye filling.Single larvae were mounted on a glass slide with two layers of adhesive tapeon both sides in 20 μL natural seawater and covered with a coverslip toimmobilize them. Dye-filled neurons were ablated on an Olympus FV1000confocal microscope equipped with a 355 nm-pulsed laser (Teem Photonics)coupled via air and controlled by the SIM Scanner (Olympus). The ablatedlarvae were recovered, fixed at 48 or 52 hpf, and processed for anti-MIP immunostaining. Light Microscopy and Image Registration. Confocal imaging was performed aspreviously described (19). Image registration to a 48-hpf whole-body nuclearreference template and colocalization analysis were performed as previouslydescribed (39). TEM. Platynereis larvae (72 hpf) were frozen using a high-pressure freezer(BAL-TEC HPM 010; Balzers) and quickly transferred to liquid nitrogen.Frozen samples were treated with a substitution medium containing 2%(wt/vol) osmium tetroxide in acetone and 0.5% uranyl acetate in a cryosub-stitution unit (EM AFS-2; Leica). The samples were cryosubstituted throughgradually rising temperatures and embedded in Epon. Serial sections werecut at 50 nm on a Reichert Jung Ultracut E microtome and collected onsingle-slotted copper grids (NOTSCH-NUM 2 × 1 mm; Science Service) withFormvar support film, contrasted with uranyl acetate and Reynolds leadcitrate, and carbon-coated to stabilize the film. Imaging of one specimen(Platynereis-72-HT-9-3) was performed at a pixel size of 3.87 nm on a TECNAISpirit transmission electron microscope (FEI) equipped with an UltraScan4000 4 × 4 k camera by using Digital Micrograph (Gatan) Stitching andalignment were done by using TrakEM2 (61). All structures were segmentedmanually as area lists, which were exported into 3Dviewer and Imaris. Receptor Deorphanization. Platynereis and Capitella SPR orthologues werecloned into a pcDNA3.1+ vector (Invitrogen) with HindIII and NotI. The Platynereisreceptor was PCR-amplified from larval cDNA by using the primers5′-ACAATAAAGCTTGCCACCATGATGGAAGTAAGCTATTCAAATGGAAATG(including HindIII site and Kozak consensus) and 5′-ACAATAGCGGCC-GCTTAAATATTTGTAGTTTTAGTCGTGTGATCG (including NotI site), andthe Capitella receptor clone used was a synthetic construct (GenScript). CHO-K1cells stably expressing a calcium-sensitive bioluminescent fusion protein weretransfected, and receptor activation was measured. Measurements were per-formed by using a fluorescent plate reader. The area under each calcium tran-sient (measured for 1 min) was calculated by using Ascent software (ThermoElectron) and expressed as integrated luminescence units (relative units). MO Injection. For microinjections, fertilized Platynereis eggs developing at16 °C were rinsed approximately 1 h after fertilization with sterile filteredseawater in a 100-μm sieve to remove the egg jelly, followed by treatmentwith 70 μg/mL proteinase K for 1 min to soften the vitellin envelope. Injectionswere carried out by using Femtotip II needles with a FemtoJet microinjector(Eppendorf) on a Zeiss Axiovert 40 CL inverted microscope equipped witha Luigs and Neumann micromanipulator. The temperature of the developingzygotes was maintained at 16 °C throughout injection by using a Luigs andNeumann Badcontroller V cooling system and a Cyclo 2 water pump (Roth).Two translation-blocking MOs and one corresponding 5-bp mismatchcontrol MO were designed to target the Platynereis MIP-receptor gene. MOswith the following sequences were purchased from Gene Tools: MIP-receptor-start1, TCCATCATTTTGAATGTTGAATGCA; MIP-receptor-mismatch,TCCATGATTTTCAATCTTCAATCCA; and MIP-receptor-start2, GTCAATGAGGTC-ACAAACATCCAAC. Nucleotides complementary to the start codon (ATG) areunderlined; nucleotides altered in mismatch control MOs are denoted byboldface italics.MOs were diluted in water with 12 μg/μL fluorescein dextran (Mr of 10,000dalton; Invitrogen) as a fluorescent tracer. MOs (0.6 mM) were injected withan injection pressure of 600 hPa for 0.1 s and a compensation pressure of 35hPa. Injected zygotes were kept in Nunclon six-well plates in 10 mL filteredseawater, and their development was monitored daily. Injected larvae at 48hpf were used for ciliary resting measurements in the presence of 5 μMsynthetic MIP7 peptide or DMSO as control. ACKNOWLEDGMENTS. We thank Harald Hausen for Capitella larvae; DetlevArendt, Raju Tomer, Heidi Roebert, and Nicola Kegel for advice on Platynereismicroinjection; and Aurora Panzera for help with microscopy. This work wassupported by Max Planck Society Sequencing Grant M.IF.A.ENTW8050 (toG.J.). The research leading to these results was supported by the EuropeanResearch Council under European Union Seventh Framework ProgrammeFP7/2007–2013 and European Research Council Grant Agreement 260821. 1. Truman JW, Riddiford LM (1999) The origins of insect metamorphosis. Nature401(6752):447–452.2. Heyland A, Moroz LL (2006) Signaling mechanisms underlying metamorphic tran-sitions in animals. Integr Comp Biol 46(6):743–759. Conzelmann et al.PNAS Early Edition | 5 of 6NEUROSCIENCE Appendix : Publ icat ions 653. Laudet V (2011) The origins and evolution of vertebrate metamorphosis. Curr Biol21(18):R726–R737.4. Gilbert SF (2012) Ecological developmental biology: Environmental signals for normalanimal development. Evol Dev 14(1):20–28.5. Denver RJ (2009) Stress hormones mediate environment-genotype interactions duringamphibian development. Gen Comp Endocrinol 164(1):20–31.6. Rieger RM (1994) The biphasic life cycle—a central theme of metazoan evolution. AmZool 34:484–491.7. Hadfield MG (2000) Why and how marine-invertebrate larvae metamorphose so fast.Semin Cell Dev Biol 11(6):437–443.8. Hodin J (2006) Expanding networks: Signaling components in and a hypothesis for theevolution of metamorphosis. Integr Comp Biol 46(6):719–742.9. Pawlik J (1992) Chemical ecology of the settlement of benthic marine invertebrates.Oceanogr Mar Biol 30:273–335.10. Zimmer RK, Butman CA (2000) Chemical signaling processes in the marine environ-ment. Biol Bull 198(2):168–187.11. Qian P (1999) Larval settlement of polychaetes. Hydrobiologia 402:239–253.12. Hadfield MG, Koehl MAR (2004) Rapid behavioral responses of an invertebrate larvato dissolved settlement cue. Biol Bull 207(1):28–43.13. Walters LJ, Miron G, Bourget E (1999) Endoscopic observations of invertebrate larvalsubstratum exploration and settlement. Mar Ecol Prog Ser 182:95–108.14. Maldonado M, Young CM (1996) Effects of physical factors on larval behavior, set-tlement and recruitment of four tropical demosponges. Mar Ecol Prog Ser 138:169–180.15. Young CM, ed. (2002) Atlas of Marine Invertebrate Larvae (Academic, San Francisco).1st Ed.16. Nielsen C (2005) Larval and adult brains. Evol Dev 7(5):483–489.17. Page LR, Parries SC (2000) Comparative study of the apical ganglion in planktotrophiccaenogastropod larvae: Ultrastructure and immunoreactivity to serotonin. J CompNeurol 418(4):383–401.18. Tessmar-Raible K, et al. (2007) Conserved sensory-neurosecretory cell types in annelidand fish forebrain: Insights into hypothalamus evolution. Cell 129(7):1389–1400.19. Conzelmann M, et al. (2011) Neuropeptides regulate swimming depth of Platynereislarvae. Proc Natl Acad Sci USA 108(46):E1174–E1183.20. Steinmetz PR, et al. (2010) Six3 demarcates the anterior-most developing brain regionin bilaterian animals. Evodevo 1(1):14.21. Hadfield MG, Meleshkevitch EA, Boudko DY (2000) The apical sensory organ ofa gastropod veliger is a receptor for settlement cues. Biol Bull 198(1):67–76.22. Rentzsch F, Fritzenwanker JH, Scholz CB, Technau U (2008) FGF signalling controlsformation of the apical sensory organ in the cnidarian Nematostella vectensis. De-velopment 135(10):1761–1769.23. Jackson DJ, et al. (2010) Developmental expression of COE across the Metazoa sup-ports a conserved role in neuronal cell-type specification and mesodermal de-velopment. Dev Genes Evol 220(7-8):221–234.24. Santagata S, Resh C, Hejnol A, Martindale MQ, Passamaneck YJ (2012) Developmentof the larval anterior neurogenic domains of Terebratalia transversa (Brachiopoda)provides insights into the diversification of larval apical organs and the spiraliannervous system. Evodevo 3:3.25. Nässel DR, Winther AME (2010) Drosophila neuropeptides in regulation of physiologyand behavior. Prog Neurobiol 92(1):42–104.26. Schoofs L, Holman GM, Hayes TK, Nachman RJ, De Loof A (1991) Isolation, identifi-cation and synthesis of locustamyoinhibiting peptide (LOM-MIP), a novel biologicallyactive neuropeptide from Locusta migratoria. Regul Pept 36(1):111–119.27. Schoofs L, Veelaert D, Broeck JV, De Loof A (1996) Immunocytochemical distributionof locustamyoinhibiting peptide (Lom-MIP) in the nervous system of Locusta mi-gratoria. Regul Pept 63(2-3):171–179.28. Lorenz MW, Kellner R, Hoffmann KH (1995) A family of neuropeptides that inhibitjuvenile hormone biosynthesis in the cricket, Gryllus bimaculatus. J Biol Chem 270(36):21103–21108.29. Hua YJ, et al. (1999) Identification of a prothoracicostatic peptide in the larval brainof the silkworm, Bombyx mori. J Biol Chem 274(44):31169–31173.30. Williamson M, Lenz C, Winther AM, Nässel DR, Grimmelikhuijzen CJ (2001) Molecularcloning, genomic organization, and expression of a B-type (cricket-type) allatostatinpreprohormone from Drosophila melanogaster. Biochem Biophys Res Commun281(2):544–550.31. Yamanaka N, et al. (2010) Bombyx prothoracicostatic peptides activate the sex pep-tide receptor to regulate ecdysteroid biosynthesis. Proc Natl Acad Sci USA 107(5):2060–2065.32. Kim Y-J, et al. (2010) MIPs are ancestral ligands for the sex peptide receptor. Proc NatlAcad Sci USA 107(14):6520–6525.33. Liu F, Baggerman G, Schoofs L, Wets G (2008) The construction of a bioactive peptidedatabase in Metazoa. J Proteome Res 7(9):4119–4131.34. Fischer A, Dorresteijn A (2004) The polychaete Platynereis dumerilii (Annelida): Alaboratory animal with spiralian cleavage, lifelong segment proliferation and a mixedbenthic/pelagic life cycle. Bioessays 26(3):314–325.35. Fischer AH, Henrich T, Arendt D (2010) The normal development of Platynereisdumerilii (Nereididae, Annelida). Front Zool 7:31.36. Struck TH, et al. (2011) Phylogenomic analyses unravel annelid evolution. Nature471(7336):95–98.37. Veenstra JA (2011) Neuropeptide evolution: Neurohormones and neuropeptidespredicted from the genomes of Capitella teleta and Helobdella robusta. Gen CompEndocrinol 171(2):160–175.38. Tomer R, Denes AS, Tessmar-Raible K, Arendt D (2010) Profiling by image registrationreveals common origin of annelid mushroom bodies and vertebrate pallium. Cell142(5):800–809.39. Asadulina A, Panzera A, Veraszto C, Liebig C, Jékely G (2012) Whole-body gene ex-pression pattern registration in Platynereis larvae. EvoDevo 3:27.40. Park D, et al. (2008) The Drosophila basic helix-loop-helix protein DIMMED directlyactivates PHM, a gene encoding a neuropeptide-amidating enzyme. Mol Cell Biol28(1):410–421.41. Hamanaka Y, et al. (2010) Transcriptional orchestration of the regulated secretorypathway in neurons by the bHLH protein DIMM. Curr Biol 20(1):9–18.42. Acampora D, et al. (1999) Progressive impairment of developing neuroendocrine celllineages in the hypothalamus of mice lacking the Orthopedia gene. Genes Dev 13(21):2787–2800.43. Perkins LA, Hedgecock EM, Thomson JN, Culotti JG (1986) Mutant sensory cilia in thenematode Caenorhabditis elegans. Dev Biol 117(2):456–487.44. Purschke G (2005) Sense organs in polychaetes (Annelida). Hydrobiologia 535:53–78.45. Jékely G, et al. (2008) Mechanism of phototaxis in marine zooplankton. Nature456(7220):395–399.46. Baubet V, et al. (2000) Chimeric green fluorescent protein-aequorin as bioluminescentCa2+ reporters at the single-cell level. Proc Natl Acad Sci USA 97(13):7260–7265.47. Tunaru S, Lättig J, Kero J, Krause G, Offermanns S (2005) Characterization of deter-minants of ligand binding to the nicotinic acid receptor GPR109A (HM74A/PUMA-G).Mol Pharmacol 68(5):1271–1280.48. Gilbert SF, Epel D (2009) Ecological Developmental Biology (Sinauer, Sunderland,MA).49. Kim Y-J, Zit!nan D, Galizia CG, Cho K-H, Adams ME (2006) A command chemicaltriggers an innate behavior by sequential activation of multiple peptidergic ensem-bles. Curr Biol 16(14):1395–1407.50. Moroz LL, et al. (2006) Neuronal transcriptome of Aplysia: Neuronal compartmentsand circuitry. Cell 127(7):1453–1467.51. Veenstra JA (2010) Neurohormones and neuropeptides encoded by the genome ofLottia gigantea, with reference to other mollusks and insects. Gen Comp Endocrinol167(1):86–103.52. Wirmer A, Bradler S, Heinrich R (2012) Homology of insect corpora allata and ver-tebrate adenohypophysis? Arthropod Struct Dev 41(5):409–417.53. Leitz T, Morand K, Mann M (1994) Metamorphosin A: A novel peptide controllingdevelopment of the lower metazoan Hydractinia echinata (Coelenterata, Hydrozoa).Dev Biol 163(2):440–446.54. Gajewski M, Leitz T, Schloßherr J, Plickert G (1996) LWamides from Cnidaria constitutea novel family of neuropeptides with morphogenetic activity. Rouxs Arch Dev Biol205:232–242.55. Erwin PM, Szmant AM (2010) Settlement induction of Acropora palmata planulae bya GLW-amide neuropeptide. Coral Reefs 29:929–939.56. Piraino S, et al. (2011) Complex neural architecture in the diploblastic larva of Clavamulticornis (Hydrozoa, Cnidaria). J Comp Neurol 519(10):1931–1951.57. Sinigaglia C, Busengdal H, Leclère L, Technau U, Rentzsch F (2013) The bilaterian headpatterning gene six3/6 controls aboral domain development in a cnidarian. PLoS Biol11(2):e1001488.58. Grigoriev IV, et al. (2012) The genome portal of the Department of Energy JointGenome Institute. Nucleic Acids Res 40(database issue):D26–D32.59. Hauenschild C, Fischer A (1969) Platynereis Dumerilii. Mikroskopische Anatomie,Fortpflanzung, Entwicklung (Gustav Fischer, Stuttgart).60. ConzelmannM, Jékely G (2012) Antibodies against conserved amidated neuropeptideepitopes enrich the comparative neurobiology toolbox. Evodevo 3(1):23.61. Cardona A, et al. (2010) An integrated microand macroarchitectural analysis of theDrosophila brain by computer-assisted serial section electron microscopy. PLoS Biol8(10):8. 6 of 6 | www.pnas.org/cgi/doi/10.1073/pnas.1220285110Conzelmann et al. Appendix : Publ icat ions 66Supporting Information Conzelmann et al. 10.1073/pnas.1220285110 Fig. S1. Annelid myoinhibitory peptide (MIP) neuropeptide precursors. (A) Schematics of MIP precursor proteins for Platynereis dumerilii and Capitella teleta.N-terminal signal peptides (teal) and the predicted peptides (gray) flanked by basic cleavage sites (red) are shown. The predicted mature MIP sequences andtheir numbering as used in the text are listed. (B) Multiple alignment of arthropod, mollusk, and annelid MIP and cnidarian GLWamide sequences show themonoor dibasic cleavage sites (underlined in red), the amidation signature glycine, the conserved tryptophan (red), and aliphatic residues (teal). The peptidesliberated following protease cleavages are indicated between the arrowheads. (C) Multiple alignment of predicted mature PlatynereisMIPs. The C-terminal Glyis expected to be converted into an amide group. (D) MIP sequence logo created from an alignment of all PlatynereisMIPs. MIP7 is the peptide with the highestsequence similarity to the consensus and was therefore used for most behavioral assays, receptor deorphanization, and antibody design. Conzelmann et al. www.pnas.org/cgi/content/short/12202851101 of 10 Appendix : Publ icat ions 67Fig. S2. Different synthetic MIPs trigger larval settlement in Platynereis without significantly affecting ciliary beat frequency. (A–E) Angular histograms of thedisplacement vectors of swimming tracks for 55 to 60 h postfertilization (hpf) control larvae (gray) and larvae in the presence of the indicated peptides (red). (F)Angular histograms of the displacement vectors of swimming tracks for 30-hpf control larvae (gray) and larvae in the presence of 5 μM MIP7. (A–F) P values ofan χ test comparing the number of larvae swimming upward and downward are indicated: ***P < 0.001; n > 100 larvae. (G) Ciliary beat frequency in controllarvae vs. larvae exposed to increasing concentrations of MIP7. Data are shown as mean ± SEM. P values of an unpaired t test are indicated: “ns” indicates P >0.05; n > 15 larvae (50 hpf) each. Conzelmann et al. www.pnas.org/cgi/content/short/12202851102 of 10 Appendix : Publ icat ions 68Fig. S3. MIP immunostaining and RNA in situ hybridization in Platynereis and Capitella larvae. (A) Platynereis MIP antibody and cross-species MIP antibodyhave specific affinity for their epitopes. MIP immunostaining on 48-hpf larvae with an antibody raised against CAWNKNSMRVWamide; MIP immunostaining(red) following preincubation of the antibody with 5 mM AWNKNSMRVWamide for 2 h (blocked); immunostaining with a cross-species–reactive MIP antibodyraised against CVWamide (red); and immunostaining with a cross-species–reactive MIP antibody (red) following preincubation with 5 mM VWamide for 2 h(blocked). Images were acquired by using the same confocal microscopy and image processing parameters. (B–I) Spatial expression of MIP throughout anneliddevelopment; developmental stages are indicated. (B, C, E, and G) Whole-mount in situ hybridization for the MIP precursor mRNA on different Platynereislarval stages. (D, F, and H) Immunostaining with an MIP antibody on different Platynereis stages. (I) Immunostaining with a cross-species MIP antibody againstthe conserved dipeptide VWamide on a Capitella larva (ventral view). All larvae are counterstained for acetylated α-tubulin (white, gray, or cyan). (A–H)Anterior views. (Scale bars: 50 μm.) Conzelmann et al. www.pnas.org/cgi/content/short/12202851103 of 10 Appendix : Publ icat ions 69Fig. S4. Gene expression profiling by image registration of MIP-expressing neurons in 48-hpf Platynereis larvae. (A–D) Average expression patterns of MIPprecursor RNA and the neuroendocrine genes prohormone convertase 2 (phc2), dimmed (dimm), and orthopedia (otp) projected onto a common whole-bodyreference template. Only the larval episphere is shown. Acetylated tubulin signal was also projected onto the reference. (E–J) Pairwise colocalization of theaverage expression patterns of the indicated genes in the reference template. The overlap in the average gene expression 3D image stacks is shown in white.All images are anterior views. (Scale bars: 50 μm.) Conzelmann et al. www.pnas.org/cgi/content/short/12202851104 of 10 Appendix : Publ icat ions 70Fig. S5. Identification of two ventral MIP neurons in a serial transmission EM (TEM) dataset. (A) Dendrite cross-section of the two ventral MIP neurons at thelevel of the secretory gland cells. Arrows point at the dendrites (compare with Movie S6). (B) Reconstruction of the two ventral MIP neurons (blue and green) inrelation to the three gland cells. The apical dendritic microvilli are shown in yellow and red (compare with Movie S7). (C) Close-up view of the two ventral MIPneurons labeled by MIP immunostaining (red) on a 72-hpf Platynereis larva, counterstained for acetylated α-tubulin (white). Arrows point at the long dendritesthat project apically between the gland cells (GC). (Scale bar: 5 μm.) Fig. S6. Phylogenetic tree of MIP/sex peptide receptors. Neighbor-joining phylogenetic tree of MIP/sex peptide receptors. The receptors that have beendeorphaned previously (1, 2) and in the present study are underlined. The bootstrap values at selected nodes are indicated. 1. Yamanaka N, et al. (2010) Bombyx prothoracicostatic peptides activate the sex peptide receptor to regulate ecdysteroid biosynthesis. Proc Natl Acad Sci USA 107(5):2060–2065.2. Kim Y-J, et al. (2010) MIPs are ancestral ligands for the sex peptide receptor. Proc Natl Acad Sci USA 107(14):6520–6525. Conzelmann et al. www.pnas.org/cgi/content/short/12202851105 of 10 Appendix : Publ icat ions 71Fig. S7. Gene expression profiling by image registration of MIP-receptor–expressing neurons in 48-hpf Platynereis larvae. (A) Average expression pattern ofPlatynereis MIP receptor projected onto a common whole-body reference template. Only the larval episphere is shown. Acetylated tubulin signal was alsoprojected onto the reference. (B–D) Colocalization of the Platynereis MIP receptor with the neuroendocrine markers phc2, otp, and dimm. The overlap in theaverage gene expression 3D image stacks is shown in white. All images are anterior views. (Scale bar: 50 μm.) Movie S1. MIP treatment triggers sinking of Platynereis larvae. Time-lapse recording of 57-hpf untreated larvae (Left) and larvae in the presence of 5 μMMIP7(Right) in vertical columns (up in the column is at the top of the movie). (Left) Background-subtracted raw videos. (Right) Larval tracks are shown in white, andred dots mark the position of the larvae. Three hundred frames of a 15-frames-per-second movie are shown (sped up 2×).