Different extrinsic trophic factors regulate neurite outgrowth and synapse formation between identifiedLymnaea neurons

Different Extrinsic Trophic Factors Regulate NeuriteOutgrowth and Synapse Formation betweenIdentified Lymnaea Neurons

David W. Munno,1,2 Melanie A. Woodin,1 Ken Lukowiak,1 Naweed I. Syed,1

Patsy S. Dickinson2

1 Respiratory and Neuroscience Research Groups, University of Calgary, 3330 Hospital Drive NW,Calgary, Alberta, Canada T2N 4N1

2 Department of Biology, Bowdoin College, Brunswick, Maine 04011

Received 30 November 1999; accepted 21 March 2000

ABSTRACT: The requirement for trophic factorsin neurite outgrowth is well established, though theirrole in synapse formation is yet to be determined. More-over, the issue of whether the trophic factors mediatingneurite outgrowth are also responsible for synapse spec-ification has not yet been resolved. To test whethertrophic factors mediating neurite outgrowth and syn-apse formation between identified neurons are con-served in two molluscan species and whether these de-velopmental processes are differentially regulated bydifferent trophic factors, we used soma–soma and neu-rite–neurite synapses between identifiedLymnaeaneu-rons. We demonstrate here that the trophic factorspresent in Aplysia hemolymph, although sufficient toinduce neurite outgrowth from Lymnaeaneurons, do notpromote specific synapse formation between excitatorypartners. Specifically, the identified presynaptic neuronvisceral dorsal 4 (VD4) and postsynaptic neuron leftpedal dorsal 1 (LPeD1) were either paired in a soma–soma configuration or plated individually to allow neu-ritic contacts. Cells were cultured in either Lymnaeabrain-conditioned medium (CM) or on poly-L-lysinedishes that were pretreated with Aplysia hemolymph(ApHM), but contained only Lymnaeadefined medium(DM; does not promote neurite outgrowth). In ApHM-

coated dishes containing DM,Lymnaeaneurons exhib-ited extensive neurite outgrowth, but appropriate exci-tatory synapses failed to develop between the cells.Instead, inappropriate reciprocal inhibitory synapsesformed between VD4 and LPeD1. Similar inappropriateinhibitory synapses were observed inAplysia hemo-lymph-pretreated dishes that contained dialyzedAplysiahemolymph. These inhibitory synapses were novel andinappropriate, because they do not existin vivo. A re-ceptor tyrosine kinase inhibitor (Lavendustin A)blocked neurite outgrowth induced by both LymnaeaCM and ApHM. However, it did not affect inappropri-ate inhibitory synapse formation between the neurons.These data demonstrate that neurite outgrowth but notinappropriate inhibitory synapse formation involves re-ceptor tyrosine kinases. Together, our data provide di-rect evidence that trophic factors required for neuriteoutgrowth are conserved among two different molluscanspecies, and that neurite extension and synapse specifi-cation between excitatory partners are likely mediatedby different trophic factors. © 2000 John Wiley & Sons, Inc. J

Neurobiol 44: 20–30, 2000

Keywords: trophic factors; neurite outgrowth; synapseformation; in vitro; regeneration

The role of individual trophic factors in neurite out-growth (Ridgeway et al., 1991; Thoenen, 1995; Schu-

man, 1995; Syed et al., 1996; Ghirardi et al., 1996;White, 1998) and synapse formation (Syed et al., 1996;Ghirardi et al., 1996; Feng et al., 1997) has been studiedextensively. It is, however, unclear whether the trophicfactors involved in the above developmental programshave unique and dedicated function or if they can simul-taneously participate in various different cellular pro-cesses. For instance, whether trophic molecules mediat-ing neurite outgrowth from a select group of neurons arealso responsible for synapse specification in the sameneurons has not yet been determined.

Correspondence to:D. Munno, ([email protected]).Contract grant sponsor: Surdna Undergraduate Research Fel-

lowship, Bowdoin College (DWM).Contract grant sponsor: Alberta Heritage Foundation for Med-

ical Research, University of Calgary, (DWM).Contract grant sponsor: National Science Foundation (PSD).Contract grant sponsor: Whitehall Foundation (PSD).Contract grant sponsor: Medical Research Council, Canada

(NIS, KL).© 2000 John Wiley & Sons, Inc.

20

Many extrinsic trophic factors, most falling intothe neurotrophin family of neurotrophic factors, havebeen identified and purified from mammalian sys-tems. The most extensively studied among these isnerve growth factor (NGF), which is involved in celldifferentiation (Fillmore et al., 1992; Welker et al.,1998; Bogenmann et al., 1998; Kane et al., 1998;Lentz et al., 1999), axonal elongation (Katz, 1986;Hu-Tsai et al., 1994; Conner and Varon, 1997), syn-apse formation (Syed et al., 1996), and apoptosis(Casaccia-Bonefil et al., 1998; Kuner and Hertel,1998; Frade and Barde, 1999). Similarly, neurotro-phin-3 (NT-3) has been shown to induce both neuriteoutgrowth (White, 1998) and synaptogenesis (Vi-cario-Abejon et al., 1998), whereas brain-derived neu-rotrophic factor (BDNF) promotes both excitatory andinhibitory synapse formation between cultured hip-pocampal neurons (Vicario-Abejon et al., 1998).Many other neurotrophins have been identified in themammalian nervous system; however, because of thecomplexity of vertebrate nervous system, fundamen-tal information regarding both the general and specificactions of many of these trophic factors is still lack-ing.

Notwithstanding the fact that several inverte-brate preparations offer simpler nervous systems,the identity of individual growth factors presenteither in the brain-conditioned medium (CM, essen-tial for neurite outgrowth; see- Wong et al., 1984;Ridgeway et al., 1991) or hemolymph (Schacherand Proshansky, 1983; Ghirardi et al., 1996; Martinet al., 1997; Schacher et al., 1999) has not yet beenfully deduced. A notable exception is a cystine-richneurotrophic factor (CRNF), which was recentlypurified from Lymnaea CM (Fainzilber et al.,1996). CRNF promotes neurite outgrowth from aselect group ofLymnaeaneurons (Fainzilber et al.,1996), though its role in synapse formation has notyet been determined.

Before they can establish synaptic contacts, neu-rons must extend axonal and dendritic processes to-ward their potential target cells, which are located atsome distance from their somata. Because the neuriteoutgrowth that precedes synapse formation is contin-gent on extrinsic trophic factors, a direct involvementof growth factors in synapse formation, independentof neurite outgrowth, cannot be studied directly. Toobtain synapses in the absence of neurite outgrowth,therefore, neurons from leech (Fuchs et al., 1981,1982; Nicholls et al., 1990),Helisoma (Haydon,1988),Aplysia (Klein, 1994), andLymnaea(Feng etal., 1997; Woodin et al., 1999) have been cultured ina soma–soma configuration. Because these synapsesform in the absence of neurite outgrowth, the role of

extrinsic trophic factors in synapse formation can beinvestigated directly. InLymnaea,soma–soma syn-apses are both target-cell–specific and electrophysi-ologically comparable to synapses seenin vivo(Woodin et al., 1999). At the electron-microscopiclevel, the soma–soma synapse appears structurallysimilar to the in vivo synapse (Feng et al., 1997).Taking advantage of the soma–soma model, Feng etal. (1997) showed that the formation of appropriate(i.e., identical to the synapse observedin vivo) inhib-itory synapses between the neurons visceral dorsal 4(VD4) and right pedal dorsal 1 (RPeD1) did notrequire extrinsic trophic factors. Synapse formationbetween the paired neurons occurred in defined me-dium (DM), which contains no trophic factors. How-ever, altered gene activity and new protein synthesiswere required. In contrast, formation of an appropriateexcitatory synapse between the presynaptic cell VD4and postsynaptic cell left pedal dorsal 1 (LPeD1) didrequire extrinsic trophic factors derived from CM(Woodin et al., 1999). Thus, although appropriateinhibitory synapse formation does not require extrin-sic trophic factors, excitatory synapses fail to developin the absence of such trophic factors. The trophicfactor–induced excitatory synapse formation wasshown to be mediated via receptor tyrosine kinases(Hamakawa et al., 1999; Woodin et al., 1999).

Because the medium containing trophic factors(CM) used in all of the above studies is known tosupport neurite outgrowth (Ridgeway et al., 1991)as well as synaptogenesis, these studies do notprovide a clear answer as to whether synapse for-mation and neurite outgrowth require the same ordifferent trophic factor(s). To extend our under-standing of the role played by extrinsic trophicfactors in neurite outgrowth and synaptogenesis andof the mechanisms through which they actin vitro,we asked whetherAplysiahemolymph contains theextrinsic trophic factors required for both neuriteoutgrowth and excitatory synapse formation be-tween identifiedLymnaeaneurons. We demonstratehere that althoughAplysiahemolymph contains theextrinsic trophic factor(s) necessary to support neu-rite outgrowth of Lymnaeaneurons, it does notinduce appropriate excitatory synaptogenesis. Inaddition, the inappropriate synapse formation thattook place inAplysiahemolymph was independentof receptor tyrosine kinase activity. This study,therefore, provides direct evidence that neurite out-growth and synaptogenesis are either regulated bydifferent extrinsic trophic factors or involve differ-ent families of receptor tyrosine kinases.

Neurite Outgrowth and Synapse Formation 21

METHODS AND MATERIALS

Animals

Laboratory raised stocks ofLymnaea stagnaliswere main-tained at room temperature in well aerated pond water andwere fed lettuce and fish food. Animals used in all experi-ments had shell lengths of 10–20 mm (approximately 2–4months old).

Cell Culture

Cell isolation and culture procedures were adapted fromthose described previously (Ridgeway et al., 1991; Syed etal., 1996). Animals were anesthetized in a 10% Listerinesolution in normalLymnaeasaline (in mM: 51.3 NaCl, 1.7KCl, 4.0 CaCl2, and 1.5 MgCl2 buffered with HEPES to pH7.9) for 10–15 min and then dissected in normal salineunder sterile culture conditions (Syed et al., 1999). Thecentral ring ganglia were isolated and washed several timeswith antibiotic saline (gentamicin, 50mg/mL). Ganglia weretreated with trypsin (2 mg/mL; Sigma, St. Louis, MO) for24 min and then trypsin inhibitor (2 mg/mL; Sigma) for 10min, each dissolved in 50% Liebowitz L-15 (Gibco, GrandIsland, NY; special order) defined medium (DM). DM wasprepared with added inorganic salts (same concentration assaline) and 20mg/mL gentamicin (Syed et al., 1999). Afterenzyme treatment, the central ring ganglia and buccal gan-glia were pinned down on a dissection dish containing highosmolarity DM (DM with added 20 mM D-glucose). Theouter and inner connective tissue were removed using fineforceps, and neurons were isolated and extracted using aGilmont syringe and fire-polished glass pipettes (50 to90-mm tip diameter).

Isolated cells were plated onto poly-L-lysine (MW95,000) pretreated glass coverslips attached to plastic dishes(Falcon 3001, Lincoln Park, NJ) in one of the following fourconditions: (1) DM alone, (2) brain-conditioned mediumfrom Lymnaea(CM); (3) dishes pretreated withAplysiahemolymph (ApHM) and filled with 2 mL ofLymnaeaDM,and (4) dishes pretreated withAplysia hemolymph andcontaining dialyzedAplysia hemolymph. To prepare CM,gentamicin-treated ganglia were incubated in DM containedin Sigma-Cote–treated glass Petri dishes for 72 h (in ahumidified chamber). The CM was frozen at220°C untiluse (up to 6 months). ApHM was drawn from the foot usingsterile hypodermic needles and syringes and kept frozen at220°C until use. To prepare ApHM dishes, hemolymphwas thawed, added to the poly-L-lysine dishes (2 mL/dish)and left for 20 min, during which time the dishes wereshaken periodically. ApHM was washed off, and the disheswere rinsed several times with normalLymnaeasaline. Priorto neuronal plating,LymnaeaDM was added to ApHM-pretreated dishes.Aplysiahemolymph was also dialyzed tothe osmolarity ofLymnaeaDM. Specifically,Aplysia he-molymph was added to Spectrapor dialysis tubing (MWcutoff 6000–8000; Spectrum Medical Industries, Los An-geles) inLymnaeasaline for 24 h with constant stirring. The

dialyzed Aplysia hemolymph was then diluted 1:1 withLymnaeaDM. Prior to plating, the 1:1 DM/dialyzedAplysiahemolymph was added to ApHM-pretreated dishes.

For experiments involving receptor tyrosine kinaseblockers, Lavendustin A (LavA; Sigma) and its inactiveform, Lavendustin B (LavB; Sigma) were used. These com-pounds were dissolved in 0.1% DMSO (which had no effecton neurite sprouting in control experiments;n 5 12) andadded to the dishes at a final concentration of 10mM, 1–2h after plating.

Sprouting Assay

Cells were left undisturbed overnight (at least 12 h) and thenviewed and photographed under a Zeiss (Axiovert 135;Thornwood, NY) microscope with an Olympus 35-mmcamera (Lake Success, NY). As was the case with previousstudies onHelisoma(Bulloch and Hauser, 1990) andLym-naea(Ridgeway et al., 1991), neurons were considered tohave sprouted if they possessed one or more neurites. 2soma diameters in length and they exhibited at least oneactive growth cone. Occasionally, neurons exhibited exten-sive processes that lacked growth cones and closely resem-bled “veils”; these cells were considered nonsprouted.

Electrophysiology

Conventional electrophysiological techniques were used torecord simultaneous intracellular activity from pairs of neu-rons as described previously (Feng et al., 1997). Specifi-cally, neurons were viewed under a Zeiss (Axiovert 135)inverted microscope and impaled using Narishige (Tokyo,Japan) micromanipulators (MM 202 and MM 204). Glassmicroelectrodes were filled with a saturated solution ofpotassium sulfate (resistance 20–40 MV). Intracellular sig-nals were amplified (IR-283; Neurodata, New York, NY)displayed on a storage oscilloscope (Fluka 2000, Buchs,Switzerland) and recorded on a chart recorder (TA 2405;Gould, Cleveland, OH). Neurons were considered to haveformed a chemical synapse if action potentials in the pre-synaptic cell evoked 1:1 excitatory postsynaptic potentials(EPSPs) or if bursts of action potentials induced compoundinhibitory postsynaptic potentials [IPSPs; 1:1 IPSPs arerarely seen (Woodin et al., 1999)] in a target neuron with aconstant delay (50–100 ms) and amplitude (1–10 mV)characteristic of PSPs inLymnaeaboth in vivo and in vitro(Syed et al., 1990, 1992).

RESULTS

Extrinsic Trophic Factors Are Requiredfor Neurite Outgrowth From LymnaeaNeurons

The neurons of the pedal A cluster (PeA) (Fig. 1),which consists of approximately 30 morphologically,

22 Munno et al.

electrophysiologically, and functionally identical cellsin each pedal ganglion, were used for all neuriteoutgrowth assays (Ridgeway et al., 1991). Because ofthe large numbers of PeA cluster cells and the relativeease with which they can be extracted from the intactganglia, these cells have been used extensively instudies involving growth assays (Ridgeway et al.,1991; Fainzilber et al., 1996).

PeA neurons were cultured in DM, CM, or ApHMfor 12–24 h, and the ability of these media to promoteneurite outgrowth was examined. The CM providedPeA neurons with sufficient trophic support to induceneurite outgrowth (92%,n 5 107;Fig. 2), as it did inprevious experiments (Ridgeway et al., 1991). In ad-dition, ApHM-coated dishes containingLymnaeaDMpromoted robust neurite outgrowth from PeA clusterneurons (86%,n 5 55). The neurite outgrowth pat-terns of neurons cultured in CM and ApHM werecomparable; neurons under both of these experimentalconditions sprouted neurites with robust arborizations(Fig. 2). In contrast, PeA neurons failed to exhibitneurite outgrowth in DM alone (0%,n 5 95). There-fore, both CM and ApHM are sufficient to elicitneurite extension from PeA neurons.

CM-Derived Extrinsic Trophic FactorsSupport Excitatory Synaptogenesisbetween Identified Lymnaea Neurons

Because PeA cluster neurons form only electricalsynapsesin vivo, we used the neurons VD4 andLPeD1 for all experiments involving chemical syn-

apses.In vivo, the presynaptic neuron VD4 forms aone-way excitatory synapse with LPeD1 (Fig. 1).Appropriate synapse formation in CM has beenshown to occur when neurons are paired in a neurite–neurite (Syed et al., 1996) and soma–soma (Woodin etal., 1999) configuration.

To test whether appropriate synaptogenesis is con-tingent on the same extrinsic trophic factor(s) thatpromoted neurite outgrowth, we attempted to recon-struct the excitatory synapse between VD4 andLPeD1. VD4 and LPeD1 were paired in either soma–soma [Fig. 3(A)] or neurite–neurite configurations[Fig. 3(B)] in DM, CM, or ApHM. After 18–24 h inculture, simultaneous intracellular recordings weremade from pre- and postsynaptic cells. Appropriateexcitatory synapses formed in the presence of CM,regardless of whether neurons were paired in a soma–soma (n 5 16) or neurite–neurite (n 5 10) config-uration. Single action potentials [Fig. 3(C)] in VD4either produced unitary EPSPs or induced action po-tentials in LPeD1 (Fig. 3D).

Aplysia Hemolymph Does not SupportAppropriate Excitatory SynapseFormation between Lymnaea Neurons

When VD4–LPeD1 pairs were cultured on growthpermissive, ApHM-pretreated dishes containing DM,the appropriate excitatory synapse did not form. In-stead, an inappropriate reciprocal inhibitory synapse,similar to that seen in DM alone (n 5 14), developedbetween the paired neurons (n 5 16; Fig. 4). Inap-propriate, reciprocal inhibitory synapses formed inApHM regardless of whether the neurons were pairedin a soma–soma (n 5 9) or neurite–neurite (n 5 7)configuration (Fig. 4). Despite the fact that neuronswere able to extend neurites and establish synapticconnections at neuritic contacts, cell pairs were un-able to form an appropriate excitatory synapse in anycase.

Aplysia neurons cultured onAplysia hemolymph-pretreated dishes required up to 5 days to form stablesynapses (Schacher and Proshansky, 1982). There-fore, we attempted to determine if increasing the timein culture would allow the paired neurons to form anappropriate excitatory synapse. After an additional 18to 24 h in culture (for a total of 36 h in culture), pairedcells still failed to form excitatory chemical synapses.Instead, they became electrically coupled [n 5 7;Fig. 4(C)], as evidenced by the fact that both depo-larizing and hyperpolarizing current pulses passedreadily between the cells.

Figure 1 (A) Schematic depicting theLymnaeacentralring ganglia. The pedal A neurons indicated here were usedfor all neurite outgrowth experiments. The identified neu-rons VD4 and LPeD1 were used for synaptogenesis exper-iments. (B) Diagram showing a one-way excitatory synapticconnection between neurons VD4 and LPeD1in vivo.

Neurite Outgrowth and Synapse Formation 23

Soluble Trophic Factors in AplysiaHemolymph Do not Support ExcitatorySynapse Formation between LymnaeaNeurons

In the above experiments, neurons were cultured ondishes that contained only substrate-bound factorsfrom Aplysia hemolymph. To determine if solublefactors fromAplysia hemolymph contained the tro-phic factors necessary for appropriate excitatory syn-apse formation, VD4–LPeD1 were cultured in asoma–soma configuration on ApHM-pretreated dishescontaining dialyzedAplysiahemolymph (see Materi-als and Methods). We found that the soluble factorsfrom Aplysiahemolymph were also unable to supportappropriate excitatory synapse formation (n 5 9; Fig.5) between VD4 and LPeD1. Instead only inappro-

priate inhibitory synapses (n 5 4) or no synapses (n5 5) formed.

Receptor Tyrosine Kinase InhibitorsBlocked Neurite Outgrowth and theFormation of Appropriate Excitatory butnot Inappropriate Inhibitory Synapsesbetween Lymnaea Neurons

To determine whether the formation of inappropriateinhibitory synapses was mediated by receptor tyrosinekinases, as has been shown for the formation of ap-propriate excitatory synapses (Woodin et al., 1999;Hamakawa et al., 1999), we tested the effect of re-ceptor tyrosine kinase blockers on synaptogenesis.VD4–LPeD1 pairs were thus cultured in a soma–

Figure 2 CM and ApHM promote neurite outgrowth from identified neurons. Neurons wereconsidered to have sprouted if they possessed at least one neurite with an active growth cone greaterthan one soma diameter in length. Neurons sprouted between 78 and 92% of the time in CM andApHM (n 5 91 and 55, respectively), but did not sprout in any case in DM (n 5 107). Scale bar5 15 mm.

24 Munno et al.

soma configuration in (1) CM1 LavA (n 5 13); (2)CM 1 LavB (n 5 7); (3) ApHM 1 LavA (n 5 6);and (4) ApHM 1 LavB (n 5 5). LavA treatmentcompletely blocked the formation of appropriate ex-

citatory synapses between VD4 and LPeD1 pairs cul-tured in CM in 46% of the cases [Fig. 6(A)]. In theremaining instances, the EPSP amplitude recordedfrom LPeD1 was markedly smaller than that recorded

Figure 3 CM promotes appropriate excitatory synapse formation. Specific excitatory synapsesbetween VD4 and LPeD1 somata reestablished in conditioned medium when juxtaposed in asoma–soma (A) or neurite–neurite (B) configuration. Simultaneous intracellular recordings weremade after 12–18 h in culture. A one-way excitatory synapse reestablished between the presynapticneuron, VD4, and its postsynaptic partner, LPeD1. Specifically, single action potentials in VD4(C,D) produced either unitary excitatory postsynaptic potentials (C) or induced action potentials (D)in LPeD1, whereas action potentials from LPeD1 did not induce a response from VD4 (E). Therecordings shown above were from neurons cultured in a soma–soma configuration. As can be seenfrom the traces, there was no electrical component to the synapse. The postsynaptic response ofLPeD1 was identical in neurons cultured in a neurite–neurite configuration. Scale bar5 time(C,D,E) 200 ms; (C) LPeD1, 10 mV; VD4, 40 mV; (D) LPeD1, VD4, 40 mV; (E) LPeD1, 40 mV;VD4, 10 mV.

Neurite Outgrowth and Synapse Formation 25

in control conditions. Current was injected into VD4to produce a single action potential (10 injections atapproximately 30-s intervals) and the resulting EPSPsin LPeD1 were recorded. The EPSP amplitude inLPeD1 from pairs cultured in CM1 LavA (2.8 mV,n 5 7) was significantly smaller than that of neuronscultured in CM alone [10.05 mV,n 5 6, ANOVA, p5 .001, Fig. 6(B)]. Consistent with earlier studies(Hamakawa et al., 1999), LavB treatment had noeffect on the formation of excitatory synapses be-tween VD4–LPeD1 pairs cultured in CM (n 5 7).

When VD4–LPeD1 pairs were plated in a soma–soma configuration on ApHM-coated dishes contain-ing DM and LavA (n 5 6) or LavB (n 5 6), theinappropriate inhibitory synapse still formed [Fig.7(D)]. Although neither LavA nor LavB affected theformation of inappropriate inhibitory synapses be-tween VD4–LPeD1 pairs, LavA inhibited neurite out-growth from 70% of neurons cultured on ApHM-

Figure 4 Inappropriate inhibitory synapses develop be-tween soma–soma and neurite–neurite paired neurons inApHM. Reciprocal inhibitory synapses between VD4 andLPeD1 somata were formed in ApHM pretreated dishesfilled with DM when juxtaposed in a soma–soma or neurite–neurite configuration. Simultaneous intracellular recordingswere made after 12–18 h in culture. Specifically, a train ofaction potentials in LPeD1 (A) or VD4 (B,C) producedcompound inhibitory postsynaptic potentials in the VD4 (A)or LPeD1 (B,C), respectively. VD4–LPeD1 pairs (n 5 7)were also tested after 36 h after culture to see if increasingthe time in culture allowed the appropriate excitatory syn-apse to establish in ApHM-treated dishes. By this time,however, the neurons had become electrically coupled. Cur-rent was passed into VD4, and we recorded the electronicpotential in LPeD1. Recordings shown are all from neuronscultured in a neurite–neurite configuration. The reciprocalinhibition was identical in neurons cultured in a soma–somaconfiguration. Scale bar5 time (A,B) 2 s; (C,D) 1 s; (A)LPeD1, VD4, 40 mV; (B) LPeD1, 40 mV; VD4, 80 mV; (C)LPeD1, 40 mV; VD4, 5 mV; (D) LPeD1, VD4, 40 mV.

Figure 5 Soluble factors fromAplysiahemolymph did notpromote appropriate excitatory synapse formation. VD4–LPeD1 pairs cultured on ApHM-coated dishes containing1:1 DM/dialyzedAplysiahemolymph either were not ableto form a synapse (A,B) or formed an inappropriate inhib-itory synapse (C). Trains of action potentials in VD4 eitherinduced no response (A,B) or induced a compound IPSPthat inhibited spontaneous firing in LPeD1. Scale bar5 time, 200 ms; (A,C) LPeD1, VD4, 40 mV; (B) LPeD1, 10mV; VD4, 40 mV.

26 Munno et al.

coated dishes containing DM [n 5 70, Fig. 7(A,C)].Neurons cultured on ApHM-coated dishes containingDM and LavA did not sprout neurites but still main-tained a healthy resting potential (between255 and265 mV). The addition of LavB to ApHM-coateddishes containing DM did not affect the percent ofsprouting [90% sprouting;n 5 10, Fig. 7(B,C)].

In summary, these results demonstrate that both

CM and ApHM contain sufficient trophic factor(s) tosupport neurite outgrowth, which is mediated via re-ceptor tyrosine kinases. CM, but not ApHM, wassufficient to induce appropriate excitatory synapseformation. Because ApHM was able to support neu-rite outgrowth but not appropriate excitatory synapseformation, these experiments provide the first directevidence that the extrinsic molluscan trophic factorsrequired for neurite outgrowth differ from those thatare involved in excitatory synapse formation. In ad-dition, because the receptor tyrosine kinase blockerLavA did not affect inappropriate inhibitory synapseformation, we suggest that the formation of suchsynapses does not require receptor tyrosine kinasesfrom the same family as those that regulate neuriteoutgrowth.

DISCUSSION

The results presented here demonstrate that extrinsictrophic factor(s) required for neurite outgrowth in twomolluscan species are indeed conserved. Consistentwith this conclusion are data from an earlier study inwhich ApHM was shown to promote neurite out-growth fromHelix neurons (Ghirardi et al., 1996). Asimilar conservation of trophic factor activity wasreported in another study, which showed that verte-brate trophic molecules such as NGF (Ridgway et al.,1991) and CNTF (Syed et al., 1996) also promoteneurite outgrowth from selectLymnaeaneurons. Be-cause the identity of any given trophic molecule inmost invertebrates has not yet been deduced, theirability to induce neurite outgrowth from vertebrateneurons has not yet been tested. From the abovestudies on invertebrate neurons, it nevertheless seemssafe to infer that the extrinsic trophic factors mediat-ing neurite outgrowth are likely conserved acrossvertebrate and invertebrate species. Further supportfor this notion stems from an earlier study in which aCRNF purified from LymnaeaCM was shown tointeract with the vertebrate p75 receptor (Fainzilber etal., 1996).

Trophic factor induced neurite outgrowth frommost vertebrate neurons involves a family of receptortyrosine kinases (Schelessinger and Ullrich, 1992).Consistent with the roles of receptor tyrosine kinasesin vertebrates, the data presented in this study dem-onstrated that both CM- and ApHM-induced neuriteoutgrowth from Lymnaeaneurons was blocked byLavA, a specific inhibitor of receptor tyrosine kinase,but not by its inactive analogue (LavB). Although thespecificity of these compounds in theLymnaeamodelhas not yet been fully determined, aLymnaeahomo-

Figure 6 Receptor tyrosine kinase inhibitors blocked ex-citatory synapse formation in CM. Pairs of neurons culturedin CM containing the receptor tyrosine kinase blocker Lav-endustin A (LavA) failed to form an appropriate excitatorysynapse in 46% of the cases (n 5 13) and instead did notform any synapse. In the remainder of the pairs cultured inCM-containing LavA, LPeD1 displayed a significantly re-duced EPSP amplitude to single action potentials from VD4(2.8 mV,n 5 7) compared to the amplitude of EPSPs fromLPeD1 neurons in control conditions (10.05 mV,n 5 7).The addition of LavB, the inactive form of LavA, had noeffect on the formation of appropriate excitatory synapseformation between LPeD1 pairs. Scale bar5 time, 2 s;VD4, 40 mV; (B) 5 mV; (C) LPeD1, 2.5 mV.

Neurite Outgrowth and Synapse Formation 27

logue, termedLymnaeatrack or (LTrK), of the mam-malian Trk receptor has recently been identified andcharacterized (van Kesteren et al., 1998). These stud-ies suggest that receptor tyrosine kinases are indeedconserved from invertebrates to vertebrates and aremost likely involved in trophic factor–induced neuriteoutgrowth.

Despite its neurite promoting activity, ApHMfailed to induce specific excitatory synapse formationbetweenLymnaeaneurons. These data are in contrastwith an earlier study in which excitatory synapses

formed between identifiedHelix neurons cultured onApHM-coated dishes (Ghirardi et al., 1996). Becausethe specificity of the synaptic responses (electricalversus chemical) and synapse formation was not rig-orously investigated in theHelix study, it is difficult toascertain whether the observed differences betweenthe two studies are species specific or reflect cellculture artifacts.

ApHM (50% diluted with DM) not only inducesneurite outgrowth fromAplysia neurons in culture,but also supports appropriate excitatory and inhibitory

Figure 7 Receptor tyrosine kinase blockers inhibited neurite outgrowth but did not preventinappropriate inhibitory synapse formation in ApHM. Only 30% of PeA neurons cultured on ApHMdishes containing DM and the receptor tyrosine kinase blocker Lavendustin A (LavA; A,n 5 70)displayed neurite outgrowth. However, the percent of neurons cultured on ApHM coated dishescontaining DM and LavB (B,n 5 10), the inactive form of LavA, that were able to sprout neuriteswas comparable to that of neurons cultured on ApHM dishes containing DM alone (C,n 5 55).VD4–LPeD1 pairs cultured on ApHM dishes containing DM and LavA still formed an inappro-priate inhibitory synapse (D). Trains of action potentials in VD4-induced compound IPSPs inLPeD1 that inhibited spontaneous firing. Scale bar5 (A,B) 15 mm; (D) time, 200 ms; LPeD1, VD4,40 mV.

28 Munno et al.

synapse formationin vitro (Kleinfeld et al., 1990). Inthe present study, because the substrate bound com-ponents of the ApHM were used, it could be arguedthat the factor(s) responsible for specific excitatorysynapse formation may be soluble. This notion, how-ever, is inconsistent with our experiments on dialyzedApHM, which also failed to induce excitatory synapseformation betweenLymnaeaneurons. Together, thesedata suggest that the trophic factors required for ex-citatory synapse formation inLymnaeaare either notconserved inAplysia or that the specific factors re-quired for the formation of an excitatory synapse areunique to any given cell pair. We argue that, unlikebrain–conditioned medium fromLymnaea,ApHMlacks specific trophic factors that would be required topromote excitatory synapse formation between vari-ous different cell types. In support of this argument,we present earlier studies on culturedAplysianeuronsin which appropriate excitatory synapses betweenidentified neurons failed to develop; instead novel,inappropriate inhibitory synapses were consistentlyobserved between pre- and postsynaptic cells (Klein-feld et al., 1990). Similarly, inappropriate inhibitorysynapses were observed between leech neurons thatwere paired in serum (Arechiga et al., 1986; Chiquetand Nicholls, 1987; Nicholls et al., 1990). These dataclearly demonstrate that trophic factors required forneurite outgrowth may be “blood-bound,” whereasthe trophic activity essential for excitatory synapseformation may reside only within nervous tissue.

This restricted localization of specific synapse-pro-moting molecules within the central ring ganglia sug-gests that such factors are either (1) produced in rela-tively small quantities, (2) vulnerable to enzymaticdegradation, or (3) not secreted indiscriminately to otherparts of the animal body. Because trophic factors spe-cifically regulate synaptic plasticity in an activity-depen-dent manner, we reason that their release must also belocally restricted. This reasoning is consistent with ear-lier studies on cultured vertebrate neurons in whichexcitatory and inhibitory synapse formation was shownto be differentially regulated by NT-3 and BDNF.

Recent studies from our laboratory have shown that,in the absence of CM, identifiedLymnaeaneuronspaired in a soma–soma configuration make novel inap-propriate synapses, which are not normally observed inthe intact ganglia (Woodin et al., 1999; Hamakawa et al.,1999). In this study, we have demonstrated that inappro-priate inhibitory synapses seen in the soma–soma con-figuration are not an artifact of this cell culture paradigm;rather, they directly support our hypothesis that neuriteoutgrowth and excitatory synapse formation are differ-entially regulated by different trophic factors. Our ex-periments in which LavA failed to block inappropriate

inhibitory synapse formation between neurons supportone of two hypotheses. First, these novel synapses maybe a “default” of the developmental program (seeWoodin et al., 1999) and thus do not require extrinsictrophic factors. Alternatively, the cell–cell signaling thatmediates inappropriate inhibitory synapse differentiationmay involve a family of receptor tyrosine kinases dif-ferent from the one that regulates neurite outgrowth.

In conclusion, this study provides direct evidencethat trophic factors mediating neurite outgrowth areconserved between two different molluscan speciesand that this neuronal sprouting is mediated via re-ceptor tyrosine kinases. Moreover, we demonstratedthat neurite outgrowth and excitatory synapse forma-tion require different extrinsic trophic factors and thatthe molecules mediating neurite outgrowth alone arenot sufficient for the formation of appropriate excita-tory synapses. Finally, in the absence of specific CM-derived factors (i.e., in ApHM),Lymnaeaneuronsmake inappropriate inhibitory synapses, which areindependent of receptor tyrosine kinase activity.

The authors thank Wali Zaidi for excellent technicalsupport.

REFERENCES

Arechiga II, Chiquet M, Kuffler DP, Nicholls JG. 1986.Formation of specific connections in culture by identifiedleech neurons containing serotonin, acetylcholine andpeptide transmitters. J Exp Biol 126:15–31.

Bogenmann E, Peterson S, Maekawa K, Matsushima H.1998. Regulation of NGF responsiveness in human neu-roblastoma. Oncogene 17:2367–2376.

Bulloch AGM, Hauser GC. 1990. Sprouting by isolatedHelisomaneurons: enhancement by glutamate. Int J DevNeurosci 8:391–398.

Casaccia-Bonnefil P, Kong H, Chao MV. 1998. Neurotro-phins: the biological paradox of survival factors elicitingapoptosis. Cell Death Differ 5:357–364.

Chiquet M, Nicholls JG. 1987. Neurite outgrowth and syn-apse formation by identified leech neurons in culture. JExp Biol 132:191–206.

Conner JM, Varon S. 1997. Developmental profile of NGFimmunoreactivity in the rat brain: a possible role of NGF inthe establishment of cholinergic terminal fields in the hip-pocampus and cortex. Brain Res Dev Brain Res 101:67–79.

Fainzilber M, Smit AB, Syed NI, Wildering WC, Hermannvan der Schors RC, Jimenez C, Li KW, van Minnen J,Bulloch AG, Ibanez CF, Geraerts WP. 1996. CRNF, amolluscan neurotrophic factor that interacts with the p75neurotrophin receptor. Science 274:1540–1543.

Feng ZP, Klumperman J, Lukowiak K, Syed NI. 1997.Invitro synaptogenesis between the somata of identifiedLymnaeaneurons requires protein synthesis but not ex-

Neurite Outgrowth and Synapse Formation 29

trinsic growth factors or substrate adhesion molecules.J Neurosci 17:7839–7849.

Fillmore HL, Mainardi CL, Hasty KA. 1992. Differentiationof PC12 cells with nerve growth is associated with in-duction of transin synthesis and release. J Neurosci Res31:662–669.

Frade JM, Barde YA. 1999. Genetic evidence for cell deathmediated by nerve growth factor and the neurotrophinreceptor p75 in the developing mouse retina and spinalcord. Development 126:683–690.

Fuchs PA, Nicholls JG, Ready DF. 1981. Membrane prop-erties and selective connections of identified leech neu-rons in culture. J Physiol (Lond) 316:203–223.

Fuchs PA, Henderson LP, Nicholls JG. 1982. Chemicaltransmission between Ritzius cells and sensory neuronsof leech in culture. J Physiol (Lond) 323:195–210.

Ghirardi M, Casadio A, Santarelli I, Montarolo PG. 1996.Aplysiahemolymph promotes neurite outgrowth and syn-aptogenesis of identifiedHelix neurons in cell culture.Invert Neurosci 2:41–49.

Hamakawa T, Woodin MA, Bjorgum MC, Painter SD,Takasaki M, Lukowiak K, Nagle GT, Syed NI. 1999.Excitatory synaptogenesis between identifiedLymnaeaneurons requires trophic factors and is mediated by re-ceptor tyrosine kinases. J Neurosci 19:9306–9312.

Haydon PG. 1988. The formation of chemical synapsesbetween cell-cultured neuronal somata. J Neurosci8:1032–1038.

Hu-Tsai M, Winter J, Emson PC, Woolf CJ. 1994. Neuriteoutgrowth and GAP-43 mRNA expression in culturedadult rat dorsal root ganglion neurons: effects of NGF orprior peripheral axotomy. J Neurosci Res 39:634–645.

Kane MD, Vanden Heuvel JP, Isom GE, Schwarz RD.1998. Differential expression of group I metabotropicglutamate receptors (mGluRs) in the rat pheochromocy-toma cell line PC12: role of nerve growth factor and rats.Neurosci Lett 252:1–4.

Katz MJ. 1986. Quantitative effects of NGF on growth ofembryonic frog axons. Brain Res 366:211–216.

Klein M. 1994. Synaptic augmentation by 5-HT at restedAplysia sensorimotor synapses: independence of actionpotential prolongation. Neuron 13:159–166.

Kleinfeld D, Parsons TD, Raccuia-Behling F, Salzberg BM,Obaid AL. 1990. Foreign connections are formedin vitroby Aplysia californicainterneurone L10 and itsin vivofollowers and non-followers. J Exp Biol 154:237–255.

Kuner P, Hertel C. 1998. NGF induces apoptosis in a humanneuroblastoma cell line expressing the neurotrophin re-ceptor p75NTR. J Neurosci Res 54:465–474.

Lentz SI, Knudson CM, Korsmeyer SJ, Snider WD. 1999.Neurotrophins support the development of diverse sen-sory axon morphologies. J Neurosci 19:1038–1048.

Martin KC, Casadio A, Zhu H, Yaping E, Rose JC, Chen M,Bailey CH, Kandel ER. 1997. Synapse-specific, long-term facilitation ofAplysiasensory to motor synapses: afunction for local protein synthesis in memory storage.Cell 91:927–938.

Nicholls JG, Liu Y, Payton BW, Kuffler DP. 1990. The

specificity of synapse formation by identified leech neu-rons in culture. J Exp Biol 153:141–154.

Ridgeway RL, Syed NI, Lukowiak K, Bulloch AGM. 1991.Nerve growth factor (NGF) induces sprouting of specificneurons of the snail,Lymnaea stagnalis.J Neurobiology22:377–390.

Schuman EM. 1997. Synapse specificity and long-term in-formation storage. Neuron 18:339–342.

Schacher S, Proshansky E. 1983. Neurite regeneration byAplysianeurons in dissociated cell culture: modulation byAplysia hemolymph and the presence of initial axonalsegment. J Neurosci 3:2403–2413.

Schacher S, Wu F, Pankyo JD, Zhong-Yi S, Wang D. 1999.Expression and branch-specific export of mRNA are reg-ulated by synapse formation and interaction with specificpostsynaptic targets. J Neurosci 19:6338–6347.

Schlessinger J, Ullrich A. 1992. Growth factor signaling byreceptor tyrosine kinases. Neuron 9:383–391.

Syed NI, Bulloch AGM, Lukowiak K. 1990.In vitro recon-struction of the respiratory central pattern generator(CPG) of the molluscLymnaea.Science 250:282–285.

Syed NI, Lukowiak K, Bulloch AGM. 1992. Specificinvitro synaptogenesis between identifiedLymnaeaandHe-lisomaneurons. Neuroreport 3:793–796.

Syed NI, Richardson P, Bulloch AGM. 1996. Ciliary neu-rotrophic factor, unlike nerve growth factor, supportsneurite outgrowth but not synapse formation by adultLymnaeaneurons. J Neurobiol 29:293–303.

Syed NI, Zaidi H, Lovell P. 1999.In vitro reconstruction ofneuronal circuits: a simple model system approach. In:Windhurst V, Johansson H, editors. Modern techniques inneuroscience research. Berlin: Springer-Verlag. p. 361–377.

Thoenen H. 1995. Neurotrophins and neuronal plasticity.Science 270:593–598.

van Kesteren RE, Fainzilber M, Hauser G, van Minnen J,Vreugdenhil E, Smit AB, Ibanez CF, Geraerts WP, Bul-loch AG. 1998. Early evolutionary origin of the neuro-trophin receptor family. EMBO J 17:2534–2542.

Vicario-Abejon C, Collin C, McKay RDG, Segal M. 1998.Neurotrophins induce formation of functional excitatoryand inhibitory synapses between cultured hippocampalneurons. J Neurosci 18:7256–7271.

Welker P, Grabbe J, Grutzkau A, Henz BM. 1998. Effectsof nerve growth factor (NGF) and other fibroblast-de-rived growth factors on immature human mast cells(HMC-1). Immunology 94:310–317.

White DM. 1998. Contribution of neurotrophin-3 to theneuropeptide Y-induced increase in neurite outgrowth ofrat dorsal root ganglion cells. Neuroscience 86:257–263.

Wong RG, Barker DI, Kater SB, Bodnar DA. 1984. Nervegrowth-promoting factor produced in culture media con-ditioned by specific CNS tissues of the snailHelisoma.Brain Res 292:81–91.

Woodin MA, Hamakawa T, Takasaki M, Lukowiak K, SyedNI. 1999. Trophic factor-induced plasticity of synapticconnections between identifiedLymnaeaneurons. Learn-ing Memory 6:307–316.

30 Munno et al.