Neurotrophic activities of trk receptors ... - Wiley Online Library

4 downloads 2384 Views 613KB Size Report
1 Department of Biological Chemistry, Molecular Neurobiology Group, Weizmann Institute of. Science, Herzl St. No. ... transmembrane domains of Ltrk undergo ligand-in- ... imately 600 million years ago. .... 20 min with serum-free medium and.
Neurotrophic Activities of trk Receptors Conserved over 600 Million Years of Evolution Gad Beck,1 David W. Munno,2 Zehava Levy,1 Helga M. Dissel,3 Jan van-Minnen,3 Naweed I. Syed,2 Mike Fainzilber1 1

Department of Biological Chemistry, Molecular Neurobiology Group, Weizmann Institute of Science, Herzl St. No. 1, 76100 Rehovot, Israel

2

Faculty of Medicine, Neuroscience Research Group, University of Calgary, Calgary, Alberta, Canada T2N 4N1

3

Department of Molecular and Cellular Neurobiology, Research Institute Neurosciences, Vrije Universiteit, De Boelelaan 1087, 1081 HV Amsterdam, The Netherlands

Received 21 May 2003; accepted 8 September 2003

ABSTRACT:

The trk family of receptor tyrosine kinases is crucial for neuronal survival in the vertebrate nervous system, however both C. elegans and Drosophila lack genes encoding trks or their ligands. The only invertebrate representative of this gene family identified to date is Ltrk from the mollusk Lymnaea. Did trophic functions of trk receptors originate early in evolution, or were they an innovation of the vertebrates? Here we show that the Ltrk gene conserves a similar exon/intron order as mammalian trk genes in the region encoding defined extracellular motifs, including one exon encoding a putative variant immunoglobulin-like domain. Chimeric receptors containing the intracellular and transmembrane domains of Ltrk undergo ligand-in-

duced autophosphorylation followed by MAP kinase activation in transfected cells. The chimeras are internalized similarly to TrkA in PC12 cells, and their stimulation leads to differentiation and neurite extension. Knock-down of endogenous Ltrk expression compromises outgrowth and survival of Lymnaea neurons cultured in CNS-conditioned medium. Thus, Ltrk is required for neuronal survival, suggesting that trophic activities of the trk receptor family originated before the divergence of molluscan and vertebrate lineages approximately 600 million years ago. © 2004 Wiley Periodicals, Inc.

INTRODUCTION

binding and signaling through the trk family of receptor tyrosine kinases (RTK) (Huang and Reichardt, 2001). Gene families with fundamental roles in nervous system development and maintenance are generally highly conserved throughout evolution (Chisholm and TessierLavigne, 1999; Hallbook, 1999). It is therefore puzzling that neurotrophin homologs are yet to be found outside the vertebrate lineage (Barde, 1994; Chao, 2000; Jaaro et al., 2001), and that the only known invertebrate ligand that binds a neurotrophin receptor is not sequence related to neurotrophins (Fainzilber et al., 1996). The only invertebrate representative of the trks identified to date is Ltrk from the mollusk Lymnaea(van Kesteren et al., 1998). The sequence of Ltrk conserves most of the

Neurotrophins control many aspects of neuronal survival and differentiation in the vertebrate nervous system by

Correspondence to: M. Fainzilber (mike.fainzilber@weizmann. ac.il) Contract grant sponsor: EU; contract grant number: HPRN-CT2002-00263. Contract grant sponsor: The Dutch NWO/ALW. Contract grant sponsor: The Canadian Institute of Health Research (CIHR). © 2004 Wiley Periodicals, Inc. DOI 10.1002/neu.10329

12

J Neurobiol 60: 12–20, 2004

Keywords: NGF; neurotrophin; neuronal survival; trophic; trk; evolution

Evolutionary Conservation of trk Functions

characteristic features of the trk family, and its expression is confined to the central nervous system and associated endocrine tissues. However, neither an endogenous ligand nor the functional significance of this receptor have yet been determined. A Drosophila RTK named Dtrk and originally thought to represent a distant trk family member (Pulido et al., 1992) has since been reclassified as a homolog of Lemon, a Hydra RTK (Miller and Steele, 2000). Most recent data suggest that this fly RTK, now renamed off-track, is actually involved in the signaling of axon guidance molecules of the semaphorin family (Winberg et al., 2001). Moreover, exhaustive scans of the fully sequenced genomes of C. elegans and Drosophila have not identified bona fide trk genes (Chao, 2000; Jaaro et al., 2001; Hidalgo, 2002). This deficiency might reflect a relative paucity of trophic interactions in the development of the nervous system in these phyla, or could suggest that molecules other than neurotrophins and trks control survival and determine cell numbers in the brains of insects and nematode worms. Evidence for the latter possibility has recently been obtained in genetic studies in Drosophila. Two independent studies have shown that the neuregulin homolog Vein maintains survival of a subpopulation of longitudinal glia (Hidalgo et al., 2001), while the TGF␣ homolog Spitz maintains survival of midline glia (Bergmann et al., 2002). In both cases, the trophic ligands suffice to maintain survival of only part of the target glial population. Thus, two different members of the EGF family are gliatrophins in Drosophila. EGF family ligands are found throughout the invertebrates, including LIN-3 in C. elegans (Liu et al., 1999) and L-EGF in Lymnaea (Hamakawa et al., 1999). Although neuronal survival promoting activities of these molecules have not yet been reported, L-EGF was shown to support neurite outgrowth (Hermann et al., 2000). The accumulating evidence thus suggests that trophic mechanisms arose early in nervous system evolution (Beck and Fainzilber, 2002; Hidalgo, 2002), but does not indicate if gene families apart from EGFs had this role in early evolution. Specifically, did trks and their ligands have trophic roles from inception of the gene family, or were they adapted to such roles in the vertebrate lineage? Here we address this issue by examining the genomic sequence, signaling capacity, and functional effects of Ltrk.

13

prepared by random priming on templates encoding the Ltrk tyrosine kinase (TK) domain or the Ltrk extracellular domain (ECD). Plaques were plated at a density of 3 ⫻ 104 per 150-mm dish, transferred to N⫹ hybond, and hybridized at 37°C in dextran sulfate. Phage DNA from selected clones was digested with SalI, followed by MboI or TasI and subcloned to pBlueScript. In addition, a set of PCRs were performed on the phage clones or on Lymnaea genomic DNA using Ltrk primers, and products were cloned to pGEM-T-Easy (Promega, Madison, WI). Sequences were obtained on ABI automated sequencers, edited to discard vector or phage arms, and assembled into contigs using Sequencher 4.1 (GeneCodes, Ann Arbor, MI). The final assembled contig sequences were submitted to Genbank under accession numbers AY137388-AY137391. Exon order and boundaries were determined by comparison to the cDNA and by prediction using NetGene2 (Pertea et al., 2001). Putative beta sheets encoded in Ltrk exon 6 were predicted by PROF (Rost and Sander, 2000). A structural model of this exon was generated using 3D-PSSM (Kelley et al., 2000).

Construction of Chimeras An AflII site was engineered into pBlueScript-Ltrk 5⬘ to the transmembrane (TM) region. The site was then used for replacement of the Ltrk ECD by a PCR generated product encoding the ECD of hPDGFR-␤. This chimera was named hPL (human PDGFR extracellular/Ltrk TM and intracellular). A second chimera, named LtrkC, was constructed by replacing the Ig-like domain encoding region of Ltrk between the engineered AflII site and an endogenous EcoRI site (nucleotides 1171–1178 in the Ltrk cDNA) with the two immunoglobulinlike domains of TrkC. Integrity of the chimeras was confirmed by sequencing and in vitro translation. The chimeras were than subcloned into pCDNA3 (Invitrogen, Carlsbad, CA) for mammalian cell expression under the CMV promoter.

Autophosphorylation and MAPK Phosphorylation Assays Transient transfections in COS cells were carried out with DEAE-dextran. Forty-eight hours post-transfection, cells were washed 3 ⫻ 20 min with serum-free medium and exposed to ligands, PDGF-BB for hPL-transfected cells or NT3 for LtrkC-transfected cells. Cells were then lysed in ice-cold NP40 lysis buffer. For autophosphorylation assay lysates were immunoprecipitated with anti-hPDGFR-␤ (R&D Systems, Minneapolis, MN), followed by Western blot with anti-phosphotyrosine (Santa-Cruz, Santa Cruz, CA). MAPK phosphorylation was detected with anti-phospho-MAPK (Sigma, St. Louis, MO).

METHODS Cloning, Sequencing, and Analyses of Ltrk Genomic Clones Low stringency screening was performed on a Lymnaea genomic library in EMBL3 (Bogerd et al., 1993). Probes were

Immunocytochemistry Plasmids were transiently transfected to PC12 cells by electroporation using an ECM 830 Electro Square Porator™ (BTX, Holliston, MA). Staining was done with anti-

14

Beck et al.

hPDGFR-␤ (R&D Systems) or anti-TrkA RTA (kindly provided by Dr. Louis Reichardt) followed by Cy2 conjugated secondary antibodies, and visualized by confocal microscopy. For comparison of internalization, PC12-6.24 TrkA overexpressor cells were transfected with hPL and plated. Twenty-four hours later, cells were incubated with antiTrkA RTA (1:250), anti-PDGFR-␤ (1:100), and 100 ng/ml PDGF-BB in DMEM HEPES supplemented with 1 mg/ml BSA and 1% donkey serum for 1 h at 4°C; washed and incubated 30 min at 4°C with anti-rabbit-cy5, anti-mousecy2, and PDGF-BB; and then washed and incubated for 15 min at 37°C in the presence of PDGF-BB. Cells were fixed and mounted as described above.

Neuronal Culture Neurons were isolated from the central ring ganglia of laboratory bred Lymnaea stagnalis and maintained in cell culture in defined medium (DM) or conditioned medium (CM) as previously described (Munno et al., 2000).

Antisense and dsRNA Knock-Down Experiments Antisense (GACCTCGCATGATCC) and mismatch control (GACCTGGCAAGATCG, three mismatches underlined) oligonucleotides were synthesized based on the Ltrk cDNA sequence from position ⫺5 (counting from the ATG initiation codon). Double stranded RNA (dsRNA) was transcribed using T7 RNA Polymerase on a template encoding the Ltrk ECD flanked on both sides with a T7 promoter sequence. Antisense or control oligonucleotides were applied in culture to a final concentration of 10 ␮M, while dsRNA was applied to a final concentration of 10 pM. Cultured neurons were incubated in a dark humidified chamber for 48 h, and scored for neurite sprouting under an inverted microscope. Neurons were considered to have sprouted if they contained at least one neurite with distinct growth cones and filopodia that was at least two soma diameters in length.

Neuronal Viability Assays Resting membrane potential was measured as previously described (Munno et al., 2000). The mechanism of cell death was further assessed by acridine orange stain (Abrams et al., 1993). After 5 days in culture, neurons were washed with sterile Lymnaea saline and stained with acridine orange (5 ␮g/ml in saline) for 30 min in a dark chamber. Cells were viewed after three saline washes of 15 min each, in the dark.

Statistics All parametric data are presented as mean ⫾ S.E. Statistical analysis was carried out using GB-Stat (Dynamic Microsystems, Silver Spring, MD). Differences between mean values from each experimental group were tested using repeated measures ANOVA.

RESULTS Predicted Ig-Like Domain in Ltrk The first cloned Ltrk cDNA did not appear to retain any obviously conserved immunoglobulin-like (Ig) domain (van Kesteren et al., 1998). This was puzzling, since the second Ig domain is thought to be the main neurotrophin-binding site in mammalian trks (Wiesmann and de Vos, 2001). We therefore searched for other Ltrk transcripts that might incorporate an Ig domain, however, exhaustive RT-PCR and low stringency screens did not reveal any additional Lymnaea trk homologs or alternatively spliced Ltrk transcripts. Cloning and sequencing of three clones from a genomic library, combined with direct PCR on genomic DNA, revealed 12 kb of the Ltrk gene, including almost complete coverage of the region between the CR and the transmembrane domain exons (Fig. 1). Exon sequences, size, and order were well conserved in this region, with the main exception being the lack of an Ig-C2 domain exon. A comparison of exon 6 in Ltrk against a library of hidden Markov models in the superfamily database (Gough et al., 2001) indicated that it encodes an Ig-like class I domain composed of seven putative beta sheets. Structural modeling of this region using a method designed to recognize remote sequence homologies encoding structurally related domains (Kelley et al., 2000) revealed strong similarity between the Ltrk-Iglike domain model and the published structure of the corresponding domain in human TrkA (Wiesmann et al., 1999) (Fig. 2). Thus, in addition to the previously identified cysteine and leucine rich motifs, the already known Ltrk transcript retains a single Ig-like domain, strengthening the possibility that it may be a functional receptor for a neurotrophin-like ligand. Two additional small exons were identified in the Ltrk gene between exon 6 and the exon encoding the transmembrane domain (Fig. 1). Of these, putative exon 8 has not been found in known Ltrk cDNAs (van Kesteren et al., 1998; and data not shown) and encodes a short sequence stretch with stop codons in all three frames. Splice variants containing this exon should therefore encode a secreted protein corresponding to the extracellular domain of Ltrk, analogous to those previously reported for mammalian FGFR1 (Werner et al., 1992) and molluscan AchBp (Smit et al., 2001). However RT-PCR experiments on adult Lymnaea CNS have so far failed to identify transcripts containing this exon (data not shown).

Evolutionary Conservation of trk Functions

Figure 1 The Ltrk genomic locus. (A) Southern blots to identify RsaI digested phage clones containing Ltrk-encoding tyrosine kinase (TK) or extracellular domain (ECD) sequences. (B) Exon/intron map of the ECD-encoding region in the Ltrk locus compared to that of human TrkB (Stoilov et al., 2002). Note that despite the overall exon conservation, the Ltrk gene lacks an exon encoding for an Ig-C2 domain, and the intron lengths are much shorter than in its mammalian counterpart. The inverted brackets indicate a single unsequenced gap. The lines above the schematic delineate the sequencing strategy to obtain complete coverage of this region of interest: dashed lines, subclones of phage DNA; continuous lines, PCR on genomic DNA; dotted lines, both PCR and shotgun cloning sequences.

Figure 2 A predicted class I Ig-like domain in Ltrk. (A) Sequence alignment of the predicted class I Ig-like domain encoded by exon 6 in the Ltrk gene, compared with the corresponding domain in human TrkC. Boxed regions and arrows indicate predicted beta sheets. (B) Ribbon diagram representations of a model of the predicted Ig-like domain in Ltrk compared to the structure of the corresponding domain in human TrkA. Although there is only 18% primary sequence identity between Ltrk and TrkA in this domain, the conservation of most of the structure-determining elements imparts a prediction confidence p ⬍ 0.05 for the Ltrk domain model.

15

16

Beck et al.

Ltrk Intracellular Domain Retains a Capacity for Signaling and Internalization Ltrk was previously shown to bind mammalian NT-3, however signaling responses to NT-3 were not observed (van Kesteren et al., 1998). This could be due to reduced kinase activity of Ltrk as compared to mammalian family members, or alternatively could suggest that NT-3 binds Ltrk in a conformation that does not enable productive dimerization and phosphorylation of the kinase domains. In order to address

this question, we created two chimeric constructs termed hPL and LtrkC. The hPL chimera encodes for the extracellular domain of the human PDGF receptor beta (Thompson et al., 1997) fused with the transmembrane and intracellular domains of Ltrk. The LtrkC chimera consists of Ltrk with an insertion of both TrkC Ig domains instead of the single Ltrk Ig-like region [Fig. 3(A)]. Transfection of the chimeras into COS and PC12 cells revealed ligand-induced autophosphorylation of the Ltrk kinase domain that was both dose- and time-dependent [Fig. 3(B); and data not shown]. We then examined the effects of NT-3 by comparing MAP kinase activation in Ltrk or LtrkC transfected cells. As shown in Figure 3(C), NT-3 stimulated robust activation of erk1 and erk2 in LtrkC cells, whereas no such response was observed with Ltrk. Finally, hPL-transfected PC12 cells responded to PDGF-BB by differentiation and neuritogenesis [Fig. 3(D)], ranging from 20% (⫾5%) differentiation at 20 ng/ml PDGF-BB to 83% (⫾12%) at 150 ng/ml PDGF-BB. Taken together, these results indicate that the Ltrk intracellular domain can function as a signaling kinase to induce differentiation of PC12 cells, similarly to the well-characterized role of endogenous TrkA in these cells. NGF-induced differentiation of PC12 cells requires internalization of TrkA (Zhang et al., 2000). We therefore examined internalization of the hPL chimera in transfected PC12 cells, and observed ligand-dependent internalization of the receptor upon PDGF-BB application [Fig. 4(A)]. The endosomal targeting of internalized hPL was determined by com-

Figure 3 The Ltrk intracellular domain is a functional signaling kinase. (A) Schematic of Ltrk chimeric constructs. hPL denotes a chimera of the extracellular domain of the human PDGF receptor-beta fused to the transmembrane and intracellular domains of Ltrk; LtrkC is composed of the TrkC immunoglobulin domains inserted between CR2 motif and the transmembrane domain of Ltrk. C, cysteine-rich region; L, leucinerich motif; Ig, immunoglobulin-like domain; TK, tyrosine kinase domain. (B) hPL was heterologously expressed in COS cells, and challenged with 150 ng/ml of PDGF-BB for the indicated times (upper row). In subsequent experiments, cells were challenged with different concentrations of PDGF-BB for 5 min (lower row). Cell lysates were immunoprecipitated with anti-hPDGFR-␤ and hPL was visualized on Western blot with anti-phosphotyrosine (Santa Cruz). (C) LtrkC or Ltrk were heterologously expressed in COS cells and challenged with 100 ng/ml of NT3 for the indicated times. MAP kinase phosphorylation was observed by Western blot with anti-phosphoerk1/2. (D) PC12 cells transfected with hPL differentiate and extend neurites after 48 h of stimulation with 100 ng/ml PDGF-BB.

Evolutionary Conservation of trk Functions

17

turned to primary culture of identified Lymnaea Pedal A (LPA) neurons, which require CNS-conditioned medium (CM) to extend neurites (Syed et al., 1990; Munno et al., 2000). We first verified Ltrk expression by in situ hybridization on LPA neurons in culture, and were able to observe down-regulation of the signal to background levels in the presence of specific dsRNA or antisense oligonucleotides to Ltrk [Fig. 5(A)]. Down-regulation of Ltrk expression strongly inhibited the sprouting response of these neurons after 48 h culture in CM [Fig. 5(B,C)], whereas similar

Figure 4 Ligand induced internalization of an Ltrk chimera is similar to TrkA. (A) Internalization of hPL in transfected PC12 cells after application of 100 ng/ml PDGF-BB for the indicated times followed by immunostaining. (B) Internalization of hPL (green) compared with that of TrkA (red) in transfected PC12-6.24 cells 30 min after application of 100 ng/ml PDGF-BB in combination with the TrkA- ligand-mimicking RTA antibody. The graph shows fluorescence intensity levels along the line drawn through the three vesicles present in the boxed region of the overlay panel. The presence of TrkA, hPL, or both, was quantified for individual vesicles by measuring the fluorescence intensity of each dye (Cy2 for hPL and Cy5 for TrkA). (C) Different angles (indicated in degrees in upper right corners) of a confocal three dimensional reconstruction of two vesicles. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

parison to TrkA in TrkA-overexpressing PC12 cells. As shown in Figures 4(B) and (C), the majority of internalized hPL co-localized with internalized TrkA in these PC12 cells. Counts of hPL positive vesicles in 22 cells revealed co-localizing TrkA in 64.3% (⫾15%) of the vesicle population. Thus, in addition to transmitting a trk-like differentiation signal in PC12 cells, the transmembrane and intracellular domains of Ltrk contain sorting signals for internalization and trafficking in a similar manner as mammalian trks.

Functions of Ltrk in Lymnaea Neurons In order to determine if Ltrk can transmit trophic signals in a physiologically relevant context, we

Figure 5 Inhibition of neurite outgrowth after knockdown of Ltrk expression. (A) In situ hybridization for Ltrk on cultured Lymnaea Pedal A (LPA) neurons incubated with 10 pM dsRNA for Ltrk (right), or a vector control dsRNA (left). (B) Lack of sprouting of Ltrk-dsRNA treated neurons compared to controls after 48 h in culture. (C) Quantification of the effects of 10 pM dsRNA or 10 ␮M antisense (AS) oligonucleotides on outgrowth of LPA neurons after 48 h in culture. The number of cells is shown as a percentage of the total (average ⫾ S.D.). *Indicates statistically significant difference from the CM control (p ⬍ 0.01). CM, conditioned medium; DM, defined medium; L, Ltrk; I, molluscan insulin receptor; C, control vector; MM, mismatch control antisense oligonucleotide.

18

Beck et al.

Ltrk is required for the transmittal of both survival and neurite outgrowth signals from a soluble extracellular ligand in Lymnaea LPA neurons.

DISCUSSION

Figure 6 Reduction of neuronal survival after knockdown of Ltrk expression. (A) Membrane potential measurements, and (B) acridine orange staining of representative control and Ltrk antisense treated cells, after 5 days in culture. DNA aggregates (arrows) in the Ltrk antisense treated cells stained with acridine orange are indicative of apoptosis. (C) Quantification of the effects of 10 pM dsRNA or 10 ␮M antisense oligonucleotides on the survival of LPA neurons in culture. Statistical parameters and key to treatments are as described in the legend to Figure 5. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

knock-down treatments for the molluscan insulin receptor (Roovers et al., 1995) had no effect [Fig. 5(C)]. Culture of LPA neurons for 5 days or longer in nonconditioned medium, or in CM with Ltrk antisense or dsRNA treatments, resulted in a loss of capacity to maintain resting membrane potential, which is indicative of neuronal death [Fig. 6(A)]. Staining LPA cells with acridine orange, an established indicator of apoptosis in invertebrate systems (Abrams et al., 1993), confirmed that the neuronal cell death was by apoptosis [Fig. 6(B)]. Quantification of the complete range of Ltrk and control knock-down treatments confirmed that neuronal survival was severely compromised specifically in cultures treated with dsRNA or antisense for Ltrk [Fig. 6(C)]. Thus,

This study provides direct evidence that the single known invertebrate member of the trk gene family retains the canonical functional capacities of trk receptors and, moreover, is required for trophic signaling responses in Lymnaea neurons. The first cloned Ltrk cDNA did not appear to retain any obviously conserved immunoglobulin-like (Ig) domain (van Kesteren et al., 1998). This was puzzling, since the second Ig domain is thought to be the main neurotrophin-binding site in mammalian trks (Wiesmann and de Vos, 2001). We therefore investigated if the cDNA represented a splice variant lacking an Ig domain, indeed splice variants of mammalian trks lacking segments of the extracellular domain have been described (e.g., Ninkina et al., 1997). However, exhaustive PCR and low stringency screens failed to uncover additional Ltrk transcripts in Lymnaea, and the data now obtained from sequencing of the Ltrk genomic locus (Fig. 1) shows that an Ig-C2 domain is completely lacking in this single molluscan trk. However, the exponential increase in sequence databases in the 5 years since the original publication of the Ltrk cDNA sequence has significantly enlarged the known boundaries of the Ig structural superfamily, enabling the prediction of a Class I Ig-like domain from homology scans and structural modeling of the sequence encoded by exon 6 in the Ltrk gene (Fig. 2). This exon is expressed and included in the original Ltrk cDNA published by van Kesteren et al. (1998). These predictions, and the analogous position of the Ig-like domain in the protein suggest that it might provide a binding site for a neurotrophin-like ligand. Nevertheless, it should be noted that the LRM can contribute to neurotrophin binding in mammalian trks (Windisch et al., 1995; Ninkina et al., 1997) and may provide the binding domain for a molluscan ligand. Such binding might also be influenced by the presence or absence of co-receptors, as has been shown with p75 in mammalian systems (Esposito et al., 2001; Zaccaro et al., 2001). Whatever the physiological ligand for Ltrk in Lymnaea, our data now show that the receptor is a functional tyrosine kinase, retaining both signaling and internalization motifs that function in the context of the model PC12 cell line (Figs. 3, 4). Indeed, critical functional motifs for trk signaling are conserved in the Ltrk intracellular domain, including binding sites for Shc and for PLC␥ (van Kesteren et al., 1998). The

Evolutionary Conservation of trk Functions

differentiation and neurite extension response to LtrkICD activation in PC12 cells [Fig. 3(D)] further supports the notion that Ltrk is functionally close to mammalian trk receptors. Furthermore, Ltrk and TrkA intracellular internalization motifs seem to act in a largely overlapping manner in PC12 cells (Fig. 4). This similarity further supports functional conservation of Ltrk, since internalization and retrograde transport are a major aspect of trk signaling in mammalian systems (Ginty and Segal, 2002). Perhaps the most striking aspect of our results is that endogenous Ltrk mediates both outgrowth and survival signals in Lymnaea neurons (Figs. 5, 6). The absolute dependence of LPA neurons on Ltrk signaling for trophic support is specific, as knock-down of the insulin receptor tyrosine kinase in these same neurons did not affect their survival (Fig. 6). Apoptotic death of molluscan neurons has not been reported thus far in vivo (Croll, 2000; Lim et al., 2002), and Lymnaea neurons can survive without exogenous trophic support in culture (Ridgway, 1991; Fainzilber et al., 1996; Syed et al., 1996). This survival may be due to the existence of autocrine Ltrk-activating trophic loops, as was previously reported for mammalian DRG neurons (Acheson et al., 1995). The recent validation of the dsRNA knock-down approach in vivo in Lymnaea (Korneev et al., 2002) should enable future studies on the possible role of Ltrk in regulating neuronal cell numbers during development of the snail. If trks will also be found in the vastly more complex brains of cephalopod mollusks (Moreno et al., 1998), comparative studies will advance our understanding of the molecular mechanisms that underlay the evolution of brain complexity. To summarize, this study provides direct evidence that the single known invertebrate member of the trk gene family retains the canonical survival-promoting role of trk receptors. The data suggest that trophic roles of the trk receptor family predated the evolutionary divergence of molluscan and vertebrate lineages, approximately 600 million years ago. We thank Ralph Bradshaw and Erik Foehr for reagents and advice in generating Ltrk/PDGFR chimeras, and Louis Reichardt for the TrkA antibody. M.F. is the incumbent of the Daniel Koshland Sr. Career Development Chair at the Weizmann Institute, and N.I.S. is an Alberta Heritage Foundation for Medical Research Scientist and a CIHR Investigator.

REFERENCES Abrams JM, White K, Fessler LI, Steller H. 1993. Programmed cell death during Drosophila embryogenesis. Development 117:29 – 43.

19

Acheson A, Conover JC, Fandl JP, DeChiara TM, Russell M, Thadani A, Squinto SP, Yancopoulos GD, Lindsay RM. 1995. A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature 374:450 – 453. Barde YA. 1994. Neurotrophic factors: an evolutionary perspective. J Neurobiol 25:1329 –1333. Beck G, Fainzilber M. 2002. Genetic models meet trophic mechanisms: EGF family members are gliatrophins in Drosophila. Neuron 33:673– 675. Bergmann A, Tugentman M, Shilo BZ, Steller H. 2002. Regulation of cell number by MAPK-dependent control of apoptosis: a mechanism for trophic survival signaling. Dev Cell 2:1–20. Bogerd J, Van Kesteren RE, Van Heerikhuizen H, Geraerts WP, Veenstra J, Smit AB, Joosse J. 1993. Alternative splicing generates diversity of VD1/RPD2 alpha peptides in the central nervous system of Lymnaea stagnalis. Cell Mol Neurobiol 13:123–136. Chao MV. 2000. Trophic factors: an evolutionary cul-desac or door into higher neuronal function? J Neurosci Res 59:353–355. Chisholm A, Tessier-Lavigne M. 1999. Conservation and divergence of axon guidance mechanisms. Curr Opin Neurobiol 9:603– 615. Croll RP. 2000. Insights into early molluscan neuronal development through studies of transmitter phenotypes in embryonic pond snails. Microsc Res Tech 49:570 –578. Esposito D, Patel P, Stephens RM, Perez P, Chao MV, Kaplan DR, Hempstead BL. 2001. The cytoplasmic and transmembrane domains of the p75 and Trk A receptors regulate high affinity binding to nerve growth factor. J Biol Chem 276:32687–32695. Fainzilber M, Smit AB, Syed NI, Wildering WC, Hermann PM, van der Schors RC, Jimenez C, Li KW, van-Minnen MJ, Bulloch AG, Ibanez CF, Geraerts WP. 1996. CRNF, a molluscan neurotrophic factor that interacts with the p75 neurotrophin receptor. Science 274:1540 –1543. Ginty DD, Segal RA. 2002. Retrograde neurotrophin signaling: Trk-ing along the axon. Curr Opin Neurobiol 12:268–274. Gough J, Karplus K, Hughey R, Chothia C. 2001. Assignment of homology to genome sequences using a library of hidden Markov models that represent all proteins of known structure. J Mol Biol 313:903–919. Hallbook F. 1999. Evolution of the vertebrate neurotrophin and Trk receptor gene families. Curr Opin Neurobiol 9:616 – 621. Hamakawa T, Woodin MA, Bjorgum MC, Painter SD, Takasaki M, Lukowiak K, Nagle GT, Syed NI. 1999. Excitatory synaptogenesis between identified Lymnaea neurons requires extrinsic trophic factors and is mediated by receptor tyrosine kinases. J Neurosci 19:9306 –9312. Hermann PM, van Kesteren RE, Wildering WC, Painter SD, Reno JM, Smith JS, Kumar SB, Geraerts WP, Ericsson LH, Smit AB, Bulloch AG, Nagle GT. 2000. Neurotrophic actions of a novel molluscan epidermal growth factor. J Neurosci 20:6355– 6364. Hidalgo A. 2002. Interactive nervous system development:

20

Beck et al.

control of cell survival in Drosophila. Trends Neurosci 25:365–370. Hidalgo A, Kinrade EF, Georgiou M. 2001. The Drosophila neuregulin vein maintains glial survival during axon guidance in the CNS. Dev Cell 1:679 – 690. Huang EJ, Reichardt LF. 2001. Neurotrophins: roles in neuronal development and function. Annu Rev Neurosci 24:677–736. Jaaro H, Beck G, Conticello SG, Fainzilber M. 2001. Evolving better brains: a need for neurotrophins? Trends Neurosci 24:79 – 85. Kelley LA, MacCallum RM, Sternberg MJ. 2000. Enhanced genome annotation using structural profiles in the program 3D-PSSM. J Mol Biol 299:499 –520. Korneev SA, Kemenes I, Straub V, Staras K, Korneeva EI, Kemenes G, Benjamin PR, O’Shea M. 2002. Suppression of nitric oxide (NO)-dependent behavior by doublestranded RNA-mediated silencing of a neuronal NO synthase gene. J Neurosci 22:RC227. Lim CS, Lee JC, Kim SD, Chang DJ, Kaang BK. 2002. Hydrogen peroxide-induced cell death in cultured Aplysia sensory neurons. Brain Res 941:137–145. Liu J, Tzou P, Hill RJ, Sternberg PW. 1999. Structural requirements for the tissue-specific and tissue-general functions of the Caenorhabditis elegans epidermal growth factor LIN-3. Genetics 153:1257–1269. Miller MA, Steele RE. 2000. Lemon encodes an unusual receptor protein-tyrosine kinase expressed during gametogenesis in Hydra. Dev Biol 224:286 –298. Moreno H, Nadal M, Leznik E, Sugimori M, Lax I, Schlessinger J, Llinas R. 1998. Nerve growth factor acutely reduces chemical transmission by means of postsynaptic TrkA-like receptors in squid giant synapse. Proc Natl Acad Sci USA 95:14997–15002. Munno DW, Woodin MA, Lukowiak K, Syed NI, Dickinson PS. 2000. Different extrinsic trophic factors regulate neurite outgrowth and synapse formation between identified Lymnaea neurons. J Neurobiol 44:20 –30. Ninkina N, Grashchuck M, Buchman VL, Davies AM. 1997. TrkB variants with deletions in the leucine-rich motifs of the extracellular domain. J Biol Chem 272: 13019 –13025. Pertea M, Lin X, Salzberg SL. 2001. GeneSplicer: a new computational method for splice site prediction. Nucleic Acids Res 29:1185–1190. Pulido D, Campuzano S, Koda T, Modolell J, Barbacid M. 1992. Dtrk, a Drosophila gene related to the trk family of neurotrophin receptors, encodes a novel class of neural cell adhesion molecule. EMBO J 11:391– 404. Ridgway R, Syed NI, Lukowiak K, Bulloch AG. 1991. Nerve growth factor (NGF) induces sprouting of specific neurons of the snail, Lymnaea stagnalis. J Neurobiol 22:377–390. Roovers E, Vincent ME, van Kesteren E, Geraerts WP, Planta RJ, Vreugdenhil E, van Heerikhuizen H. 1995. Characterization of a putative molluscan insulin-related peptide receptor. Gene 162:181–188.

Rost B, Sander C. 2000. Third generation prediction of secondary structures. Methods Mol Biol 143:71–95. Smit AB, Syed NI, Schaap D, van Minnen J, Klumperman J, Kits KS, Lodder H, van der Schors RC, van Elk R, Sorgedrager B, Brejc K, Sixma TK, Geraerts WP. 2001. A glia-derived acetylcholine-binding protein that modulates synaptic transmission. Nature 411:261–268. Stoilov P, Castren E, Stamm S. 2002. Analysis of the human TrkB gene genomic organization reveals novel TrkB isoforms, unusual gene length, and splicing mechanism. Biochem Biophys Res Commun 290:1054 –1065. Syed N, Richardson P, Bulloch A. 1996. Ciliary neurotrophic factor, unlike nerve growth factor, supports neurite outgrowth but not synapse formation by adult Lymnaea neurons. J Neurobiol 29:293–303. Syed NI, Bulloch AG, Lukowiak K. 1990. In vitro reconstruction of the respiratory central pattern generator of the mollusk Lymnaea. Science 250:282–285. Thompson LM, Raffioni S, Wasmuth JJ, Bradshaw RA. 1997. Chimeras of the native form or achondroplasia mutant (G375C) of human fibroblast growth factor receptor 3 induce ligand- dependent differentiation of PC12 cells. Mol Cell Biol 17:4169 – 4177. van Kesteren RE, Fainzilber M, Hauser G, van Minnen J, Vreugdenhil E, Smit AB, Ibanez CF, Geraerts WP, Bulloch AG. 1998. Early evolutionary origin of the neurotrophin receptor family. EMBO J 17:2534 –2542. Werner S, Duan DS, de Vries C, Peters KG, Johnson DE, Williams LT. 1992. Differential splicing in the extracellular region of fibroblast growth factor receptor 1 generates receptor variants with different ligand-binding specificities. Mol Cell Biol 12:82– 88. Wiesmann C, de Vos AM. 2001. Nerve growth factor: structure and function. Cell Mol Life Sci 58:748 –759. Wiesmann C, Ultsch MH, Bass SH, de Vos AM. 1999. Crystal structure of nerve growth factor in complex with the ligand-binding domain of the TrkA receptor. Nature 401:184 –188. Winberg ML, Tamagnone L, Bai J, Comoglio PM, Montell D, Goodman CS. 2001. The transmembrane protein offtrack associates with plexins and functions downstream of semaphorin signaling during axon guidance. Neuron 32:53– 62. Windisch JM, Marksteiner R, Lang ME, Auer B, Schneider R. 1995. Brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin-4 bind to a single leucine-rich motif of TrkB. Biochemistry 34:11256 –11263. Zaccaro MC, Ivanisevic L, Perez P, Meakin SO, Saragovi HU. 2001. p75 co-receptors regulate ligand-dependent and ligand-independent Trk receptor activation, in part by altering Trk docking subdomains. J Biol Chem 276:31023–31029. Zhang Y, Moheban DB, Conway BR, Bhattacharyya A, Segal RA. 2000. Cell surface Trk receptors mediate NGF-induced survival while internalized receptors regulate NGF-induced differentiation. J Neurosci 20:5671–5678.