Vol.
184,
No.
April
15,
1992
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
NEUROTROPHIC
FACTOR-LIKE
73-79
ACTIVITY IN DROSOPHILA
Izumi Hayashi, Martha Perez-Magallanes and John M. Rossi Department of Molecular Genetics, Beckman Research Institute of the City of Hope Duarte, California 91010 Received
February
25,
1992
Summary: The NGF-family of neurotrophic factors are structurally similar peptides with related functional properties. So far, this family of neurotrophic factors has only been identified in the vertebrate nervous system. We have determined that cultured Drosophila embryonic cells produce and secrete into medium, an activity which stimulates neurite outgrowth of embryonic chick sensory ganglia. This Drosophila activity can be blocked by antibodies to mouse NGF, indicating an immunological relationship between the Drosophila factor, mouse NGF and possibly other vertebrate neurotrophic factors. Addition of mouse NGF to Drosophila embryonic cells in culture results in increased cell number and enrichment of the neuronal phenotype, indicating that Drosophila cells have the ability to respond to the vertebrate factor. In addition, poly(A)+RNA extracted from Drosophila contains a single 1.4 Kb band which cross-hybridizes with a mouse NGF cRNA probe. These results indicate that vertebrate neurotrophic factor-like functions may operate in a genetically defined inverQ 1992 Academic Press, Inc. tebrate species. The NGF-family of neurotrophic factors regulate differentiation of the vertebrate nervous system and maintenance of its various functional properties. Nerve growth factor (NGF), the first neurotrophic factor to be discovered, was originally described as an agent which stimulated the neurite outgrowth of chick embryonic sympathetic and sensory ganglia (1). A large body of studies since then has established that NGF is essential for the survival and maintenance of specific populations of differentiated neurons derived from the neural crest (2). NGF has also been shown to have trophic interactions on central cholinergic neurons (3). widening its range of targets on differentiated neuronal cell populations. A less appreciated function of NGF is its ability to stimulate cell division. A mitogenic role for NGF acting on embryonic, mitotically active neuron precursor cells, was described in the early days of NGF history (4). This function has found strong support in recent years by the demonstration of a mitogenic effect of NGF on neural crest-derived cell lines (5), neuro-epithelial stem cells (6) and pluripotent adrenal chromaffii cells (7). The specificity of NGF action on subsets of cells in the nervous system
suggested
the existence of
other neurotrophic factors, which led to the discoveries of brain-derived neurotrophic factor (BDNF, ref.8) and neurotrophin-3 (NT-3, mf.9). All three neurotrophic factors possess neurite outgrowth promoting activity on embryonic chick dorsal root ganglion explants (10). Biological activities so far examined suggest distinct as well as overlapping properties for each neurotrophic factor (11). All three neurotrophic factors sharestructural similarities. Comparison of NGFs among different species (12) also indicate that neurotrophic factors are evolutionarily conserved
73
The significance of
0006-291X/92 $1.50 Copyright 0 1992 by Academic Press, Inc. Ail rights of reproduction in any form reserved.
Vol.
184,
No.
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
these similarities for understanding the primary role of neurotrophic factors in nervous system development is yet to be elucidated. Existence of a molecule(s) functionally or biochemically related to vertebrate neurotrophic factors has not been reported in invertebrates. Here we report that Drosophila possesses neurotrophicfactorlike functional activity. Biochemicaland molecularcharacterizationof this Drosophila
activity will
ultimately allow usto dissectthe fundamentalrole of neurotrophicfactorsin nervoussystemdevelopmentin anorganismamenableto geneticanalysis. Materials and Methods Drosophila embryonic cell culture. Drosophila CantonS embryoniccells werepreparedfor culture by Seecofs method(13). Briefly, the embryoswere collectedfor 2 hourson standardcorn meal-agar platesandallowedto develop for 3 hoursat 25°C. The early gastrulastageembryoswere dechorionatedin 1.5%sodiumhypochlorite/47.5%ethanol,andhomogenizedwith a Dounce homogenizerin Schneider’smodifiedDrosophila medium,supplemented with 18% fetal calf serum and0.2pg/mlbovine insulin. The homogenate waspassedthrough a Nytex meshfilter (50 p) andthe singlecellswereplatedin the abovemedium.After attachment(15 minutesafter plating), the mediumwaschangedto experimentalconditions.The cultureswere maintainedat 25°C. by Analysesof NGF effects in DrosophilaembryoniccelLFin culture. Cell growth wasassessed countingcells 8 hoursafter plating,whenthe majority of the cells haveceasedto divide but have not yet differentiated. Culturesweretrypsinized,stainedwith trypan blue and cellswere countedwith a hemocytometer.To assess differentiation,myotubesandneuronclusterswerescoredfrom terminally differentiatedcultures24 hoursafter plating. Cellswerefixed in 4% pamformaldehydeandcounted through a calibratedreticle in the ocularlensof a phasecontrastmicroscope.Ten fields wereexaminedfor eachsamples.Mouse NGF (2.5s NGF, culture grade)wasobtainedfrom Collaborative Research.Affinity purified sheepanti-mouseNGF antiserum(14) wasusedat 250 rig/ml. Assayfor neurite outgrowth promoting activity. Dorsalroot sensorygangliawere isolatedfrom 8 to 10day old chick embryosand placedon FalconPremariadishesin the presenceof smallquantitiesof freshculture medium(1: 1 mixture of serum-freeRPM1 1640,andSchneider’smodifiedDrosophila mediumsupplementedwith 5% fetal calf serum).After attachmentof the ganglia(about30 minutes), 0.5 ml of experimentalmediumwasaddedto eachdish. The gangliawerecultured at 37°Cin a humidified CO2 incubatorfor 24 hours. The resultswerephotographedusingan OlympusVanox phase contrastmicroscope,after fixation of the gangliain 4% paraformaldehydeDrosophila cultureconditionedmediumwasobtainedasfollows: Gastrulastagecellswerecultured asdescribedabove,except that the cellswere platedat 4 x106 cellsper60 mmdishin fresh Schneider’smodifiedDrosophila mediumsupplemented with 5% fetal calf serum. After 5 days in culture, the mediumwascollected,centrifuged,passedthrougha 0.2 p Gelmanfilter andusedimmediately. Northern analysisof Drosophilapoly(A)+RNA. Drosophila poly(A)+RNA sampleswereanalyzed usinga 32P-labelledcRNA probederivedfrom a mouseNGF cDNA clone, 12E4(15). Poly(A)+RNA wasglyoxalated andelectrophoresed overnight through a 1.5% agarosegel andtransferred onto a nitrocellulosemembrane.Hybridization wasperformedovernight at 60°C in 50% formamide,SXSSC, 50 mM NaPi, pH6.5,0.5X Denhardt’s,0.1%SDS, 1mM EDTA, 250 vg/ml salmon sperm DNA, and 10 pglml Drosophila 3rd instarpoly(A)-RNA. The NGF probewasusedat 106 cpm/mlof hybridization buffer. Washingwasin 2XSSC. 0.1% SDS at room temp.for 30 minutes twice, followed by a final washat 60°C for 20 minutesin 0.1 X SSC, 0.1% SDS. Drosophila 6-tubulin cRNA probewasusedto quantitatepoly(A)+RNA samplesfrom different stagesof development, sinceB-tubulinmRNA is found at constantlevelsthroughoutDrosophila development(16). After strippingthe mouseNGF cRNA probehybridization in 0.1XSSC, 0.1% SDS at 80°C for 30 min, the tubulin probewasusedasdescribedabove. The final washwasat 65°C for 30 minutesin 0.1X SSC and0.1% SDS. 74
Vol.
184,
No.
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Results and Discussion
Drosophila embryonic cells in culture produce neurite outgrowth promoting activity. The classic in vitro neuriteextensionassayof chick embryonic sensoryganglia(1) wasusedto examinewhether Drosophila
cells produce a neurite outgrowth
promoting
activity similar to vertebrate neurotrophic
factors. When dorsalroot sensorygangliafrom chick embryosweregrown in freshculture medium, no neurite outgrowth was observed as shown in Figure 1A. Addition of mouse NGF to this medium
resultedin a characteristicneuritehalo (Figure 1B). The mouseNGF-inducedneurite outgrowth was inhibited by concomitantadditionof anti-mouseNGF antibodies(Figure 1C).
Figure 1. Neurite outgrowth promoting activity of conditioned medium from Drosophila embryonic cells in culture on chick embryonic dorsal root ganglion explants. A: Fresh culture medium (1: 1 mixture of Schneider’s modified Drosophila medium supplemented with 5% fetal calf serum and serum-free RPMI 1640); B: Fresh culture medium + mouse NGF; C: Fresh culture medium + mouse NGF + anti-mouse NGF antibodies; D: 1: 1 mixture of serum-free RPMI 1640 and Drosophila culture conditioned medium; E: Medium as in D, with anti-mouse NGF antibodies. Mouse NGF: 20 rig/ml; anti-mouse NGF antibodies: 250 rig/ml. The bar at the right bottom indicates 0.2 mm. 75
Vol.
184, No. 1, 1992
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Addition of Drosophilaconditionedmedium,obtainedfrom 5day culturesof Drosophila gastrula stageembryonic cells,alsoresultedin stimulationof significantneuriteoutgrowth in the absenceof any mouseNGF (Figure 1D). The neuriteoutgrowth in the presenceof Drosophilaconditionedmedium wasblockedby anti-mouseNGF antibodies(Figure 1E). Theseresultsindicatethat Drosophila cellsin culture producea factor which is functionally similarto vertebrateneurotrophicfactors in stimulatingneuriteoutgrowthof chick embryonic sensoryganglia. In addition,the resultssuggest that the factor producedby Drosophilacells sharesimmunologicalhomologyto mouseNGF, since the neuriteoutgrowthpromotingeffect of the conditionedmediumwasblockedin the presenceof antibodiesto mouseNGF. Neuriteoutgrowth observedin the presence of Drosophila conditioned mediumwasnot dueto increasedspreadingof the connectivetissuecells,sinceno neurite outgrowth wasobservedin the presenceof anti-mouseNGF in spiteof a similarspreadingof connective tissue cellsin suchcultures. The effect of the conditionedmediumin stimulatingneuriteoutgrowth increasedwith the ageof the conditionedmedium. The conditionedmediumobtainedfrom 5 day-old culturesof Drosophilaembryoniccellswas moreeffective thanthosefrom 3-day or l-day cultures (datanot shown). This couldbe dueto the accumulationof the neuriteoutgrowth promotingactivity in the conditionedmediumover prolongedculture periodfor Drosophilaembryoniccells, or it could be due to an increasein the productionof the neurite stimulatingactivity by more differentiated and older cells in the culture.It shouldbe notedthat Drosophilaconditionedmedium,althoughqualitatively similarto mouseNGF, waslesseffective than the purified growthfactor in stimulatingneurite outgrowth under the experimentalconditionsemployed. This maybe duein part to the conditionsfor obtainingDrosophilaconditionedmediumor alternatively, the less-than-optimal effect of the Drosophila conditionedmediumcouldbe dueto the biochemicalnatureof the Drosophilafactor. Examination of thesedifferent possibilitiesis technicaliy not feasiblewithout a partial purification of the Drosophila neuriteoutgrowthpromotingactivity, due to the differencesin culture conditionsbetween the Drosophila embryoniccellsandembryonicchick ganglia. Mouse NGF stimulatesdivision of Drosophila embryonicneuronprecursor cells in cukure. Highly reproducibleprimary culturescanbe obtainedfrom ceilsof early gastrulastageDrosophila embryos(13). In thesecultures,aninitially undifferentiatedpopulationof cellsundergoesrapid multiplication and terminally differentiate into multinucleatedmyotubesanddifferentiatedneurons.The Drosophilaculture systemoffers a uniqueopportunityto examinethe effects of variousepigenetic substances on the early phaseof neuronaldevelopment,whenprecursorcells of the nervoussystem are still mitoticahy competent(17, 18). Addition of mouseNGF to the culture mediumresultedin a 2.4 fold increasein cell numberover the control cultures8 hoursafter plating(Figure 2A). Addition of mouseNGF did not affect the percentageof trypan blue positive, deadcells at 8 hours(about5% in ail cultures), indicatingthat the stimulatoryeffect of mouseNGF on the cell numberis morelikely to be due to an increasein celI division ratherthanan increasein the numberof viable cells. The mouseNGF effect wasconcentration dependent,andwasmost effective at 2 rig/ml(lo-%). Concomitantadditionof mouseNGF andanti-mouseNGF antibodiesabolishedthe NGF effect on the cell numberincrease(Figure2B). Theseresultsindicatethat the increasein cell numberobservedafter NGF additionto Drosophila embryoniccell culturesis due to NGF itself, andnot to possiblecontaminantsin the NGF prepara76
Vol.
184,
No.
1, 1992
BIOCHEMICAL
0
0.2
2
NGF
b.gfd
AND
BIOPHYSICAL
20 NGF Ab
-
+ -
RESEARCH
+ +
COMMUNICATIONS
;
? 20
NGF Ab
-
+ -
+ t
0
+
0.2 NGF
2 @hU
Finme 2. Effect of mouse NGF and anti-mouseNGF antibodieson the growth and differentiation of Drosophila gastrula stage embryonic cells in culture. A: Effectof mouse NGFoncell proliferation. Cellswereplatedat 0.8X 106per35mmculturedish.B: Supression of growth-stimulating effectof mouse NGFby anti-mouse NGF antibodies. Cellswereplatedat0.5X 106per35mmculturedish. C: Supression of cellgrowthby anti-mouse NGFantibodies in highdensitycultures.Thecellswere platedat 2 X 106per35mmculturedish.Notethattheantibodies inhibitedgrowthnotonly in the presence, butalsoin theabsence of mouse NGF. D: Effectof mouse NGFonthedifferentiationof Drosophilaembryoniccells. An average of 10fieldsperplatewereexamined andscoredfor neuron clusters andmyotubes, eachwith randomlychosen differentunitareaof 500pm*. Thisexperiment wasdonein parallelwith theexperiment in A. Forall experiments. barsrepresent themeanof counts fromduplicateplateswith standard error. tion. The resultsalsosuggest that the Drosophilacellsmay possess a receptorwhich canrecognize mouseNGF. Anti-mouseNGF antibodiesalonedid not haveany effect on cell numberin thesecultures. However, when the cellswereculturedat highdensities,the anti-mouseNGF antibodiesnot only suppressed the increasein cell numberobservedin the presenceof mouseNGF, but alsoinhibited the cell numberincreaseby 50% in the controlculturesto which no NGF hadbeenadded(Figure 2C). The effect of mouseNGF on cell growth wasalsolessevident in high density cultures(Figure 2C). Theseresults,in addition to the neuriteoutgrowthpromotingactivity of the Drosophilacell culture conditionedmedium,suggestthe presenceof a neurotrophicfactor in the Drosophilaculture medium. Drosophilacells in culture appearto be producingandrespondingto their own trophic factor whoseeffect canbe mimickedor replacedby mouseNGF. Anti-mouseNGF antibodiesappearto interfere with the action of the Drosophilatrophic factor presentin the high densitycultures. The increasein cell numberobservedin the presenceof mouseNGF seemsto resultfrom the preferentialstimulationof the growth of neuronalprecursorcells. Cellsthat form terminally differentiatedneuronclustersareclonally related,arisingfrom a singleneuronprecursorcell (17). Hence,our of resultswhich showan increasein the numberof neuronclusters per unit areain the presence mouseNGF (Figure 2D), strongly suggestsanunderlyingincreasein the numberof neuronprecursor cells. The numberof myotubesperunit arearemainedconstant. Severalmechanisms are possiblewhich couldresultin the increasein cell numberandthe subsequentenrichmentof the neuronalphenotypeof the Drosophilaculturesin the presenceof mouse NGF. Mouse NGF (or the equivalentDrosophilafactor) may stimulatethe rate of cell divisionor prolongthe duration of the growth phaseof Drosophilaneuronprecursorcells. Alternatively, the 77
Vol.
184,
No.
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Figure 3. Northern analysis of Drosophilcl poly(A)+RNA. Top panel: Drosophila poly(A)+RNA samples from (2) 5hr embryos, (3) 1st (4) 2nd and (5) 3rd instar larva and (6) young adults (within 1 week of hatching), probed with a cRNA probe derived from mouse NGF cDNA clone, 12E4. Numbers represent lanes. Twenty ug of Drosophila poly(A)+RNA samples were loaded in each lane. Lane (1) is male mouse submaxillary gland total RNA (300 ng), used as a standard. Bottom panel: The same filter probed with Drosophila B-tubulin cRNA probe The arrow indicates the position of 16s E. cob ribosomal RNA. The mouse NGF cDNA clone 12E4 (gift from Dr A.Gray, Genentech), originally carried in a pBR322 plasmid vector, was subcloned into a Bluescribe plasmid vector and the cRNA probe was synthesized from the EcoKI-linearized Bluescribe clone using T3 RNA polymerase in the presence of 32P-UTP (New England Nuclear, 3000 Wmol). For the tubulin probe, a 398 bpDrall-EcoRV subfragment of the B-tubulin cDNA pTu56 (gift of Dr. R. Renkawitz-Pohl, Max-Planck Institute, Munich) was subcloned into Bluescript plasmid vector and the cRNA probe was synthesized from the EcoRV-linearized clone.
factor may play a role in influencing theDrosophila
gastrula stage embryonic
cells to takea neuronal
pathway, thus increasingthe proportionof cellswhich becomeneuronprecursorcells,aswasrecently suggestedwith BDNF for pluripotentneuralcrest-derivedcells in the chick embryo(19). Thesedifferent possibilitieswill haveto be examinedfurther by detailedanalysesof neurotrophic factor effect(s) on the growth and differentiationpropertiesof the Drosophila
embryoniccells in
culture. A Drosophila mRNA is recognizedby a mouseNGFprobe. In orderto obtainadditionalmolecular evidencefor a Drosophila
factor that resemblevertebrateneurotrophicfactors,Drosophila
poly(A)+RNA wasexaminedby Northernblot analysis,usinga probespecificfor mouseNGF mRNA. The mouseprobehybridized to a singlebandof Drosophila
poly(A)+RNA in all stagesof
developmentexamined(Figure 3). Crosshybridization of a Drosophila mRNA bandwith the mouse NGF probeunderstringentcross-species hybridization conditionssuggests a strongsimilarity in the sequenceof mouseNGF andthe Drosophila
mRNA band. The sizeof the Drosophila
mRNA (1.4
Kb) was alsosimilarto that of themouseBNGFmlWA (1.3 Kb), aswell asto other known vertebrateneurotrophicfactor mlWAs (8,9). In addition,the level of theDrosophila
mRNA appearedto
changeasa function of developmentalstages.Significantly, the mRNA was alreadydetectablein 5 hour embryos,when thereis active multiplicationof neuronprecursorcellsbut no overt detectable neuronaldifferentiation. The Drosophila
mRNA increaseddramaticallyin the larval stages,coinci-
dent with our observationson the neuriteoutgrowthpromotingactivity of the 5 day-old conditioned medium(in vivo equivalentof the latelarval stagecells). ConcludingRemarks. Our studiesdemonstratethe existenceof a Drosophila moleculewith both structuralandfunctional similaritiesto the vertebrateneurotrophicfactors. Many factorsin the past havebeenshownto stimulateneuriteoutgrowthof chick embryonicsensoryganglia andin this regard, the neuriteoutgrowth promotingactivity of the Drosophila conditionedmediumitself is not surprising. However, in our studies,actionsof this factor on Drosophila 78
cellscanbe mimickedby
Vol.
184, No. 1, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
mouseNGF, the mostextensively studiedandreadily availableneurotrophicfactor of the NGF-family of peptides. Much of our presentknowledgeon the actionsandlocalizationsof vertebrateneurotrophicfactors are basedon NGF. Recentstudieson the NGF-family of neurotrophicfactorssuggest,that many of the earlierstudiesattributing the functionalactivity to NGF may requirereevaluation.Theseinclude indicationsthat NGF, BDNF andNT-3 may usethe samereceptorsto mediatetheir effects (20,21), andthe possiblecrossreactivity of anti NGF antibodiesto NT-3 andBDNF (22). In our present studies,we have usedmouseNGF asa point of referencefor the Drosophilaneurotrophicactivity. This doesnot precludepossibilitiesthat the Drosophilafactor may turn out to be moresimilarto the other membersof the vertebrateneurotrophicfactors,or found to bequiteuniquewith limited but significantsimilarity to all vertebrateneurotrophicfactors. The determinationof the natureof the Drosophila molecule,andthe extent of its homologyto vertebrateneurotrophicfactorsmustcome from sequencingstudies. Informationgainedfrom suchstudiesmay alsobring insightinto understandingthe structure-functionrelationshipandthe evolutionary aspectof the vertebrateneurotrophic factors. Acknowledgments The authorswish to thank Dr. H. Thoenenfor his supportin the initial phaseof this work, during a sabbaticalof I. Hayashiin his laboratory. The work wassupportedin part by a grant from NSF (BNS-8616936)andNINCDS (POlNS18858-07Sl). References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Levi-MontaJcini, R. andHamburger,V. (1953). J. Exp. Zool.. 123,233-288. Levi-Montalcini, R. andAngeletti, P. U. (1968). PhysioLRev. 48,534-569. Williams, L. R., et al. (1986). Proc. Natl. Acad. Sci., USA 83,9231-9235. Levi-Montalcini, R. andBooker, B. (1960).Proc. Natl. Acad. Sci., USA 46,373-384. Revoltella, R. P. andButler, R. H. (1980). 1. Cell. Physiol. 104,27-34. Cattaneo,E. andMcKay, R. (1990). Nature347,762-765. LiIlien, L. E. andClaude,P. (1985).Nature 317,632-634. Leibrock, J., eral. (1989). Nature 341,149-152. Maisonpierre,P.C., er al. (199Oa).Science247,1446-1451. Thoenen,H. (1991). TINS, 14,165-170. Maisonpierm,P. C., et al. (199Ob).Neuron 5,501-509. Ullrich, A.,er al. (1983). Nature 303,821-825. Meier, R.,eral. (1986). EMBO J. 5, 1489-1493. Stoekel, K., et al. (1976). J. Neurochem.26,1207-1211. Bialojan, S., et al. (1984). EMBO J. 3,2543-2548. Seecof,R. L. andUnanue,R. L. (1968). Exp. Cell. Rex 50,654-660. Furst, Aand Mahowald,A. P. (1985). Dev. Biol. 112,467-476. Salvaterra,P. M., et al. (1987). J. Neurosci.7, 10-22. Sieber-Blum,M. (1991). Neuron 6,949-955. Klein, R., et al. (1991). Cell 65, 189-197. Cordon-Cardo,C., eral. (1991). Cell 66, 173-183. Acheson,A., et al. (1991). Neuron,7,265-275.
79
184,
No.
April
15,
1992
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS Pages
NEUROTROPHIC
FACTOR-LIKE
73-79
ACTIVITY IN DROSOPHILA
Izumi Hayashi, Martha Perez-Magallanes and John M. Rossi Department of Molecular Genetics, Beckman Research Institute of the City of Hope Duarte, California 91010 Received
February
25,
1992
Summary: The NGF-family of neurotrophic factors are structurally similar peptides with related functional properties. So far, this family of neurotrophic factors has only been identified in the vertebrate nervous system. We have determined that cultured Drosophila embryonic cells produce and secrete into medium, an activity which stimulates neurite outgrowth of embryonic chick sensory ganglia. This Drosophila activity can be blocked by antibodies to mouse NGF, indicating an immunological relationship between the Drosophila factor, mouse NGF and possibly other vertebrate neurotrophic factors. Addition of mouse NGF to Drosophila embryonic cells in culture results in increased cell number and enrichment of the neuronal phenotype, indicating that Drosophila cells have the ability to respond to the vertebrate factor. In addition, poly(A)+RNA extracted from Drosophila contains a single 1.4 Kb band which cross-hybridizes with a mouse NGF cRNA probe. These results indicate that vertebrate neurotrophic factor-like functions may operate in a genetically defined inverQ 1992 Academic Press, Inc. tebrate species. The NGF-family of neurotrophic factors regulate differentiation of the vertebrate nervous system and maintenance of its various functional properties. Nerve growth factor (NGF), the first neurotrophic factor to be discovered, was originally described as an agent which stimulated the neurite outgrowth of chick embryonic sympathetic and sensory ganglia (1). A large body of studies since then has established that NGF is essential for the survival and maintenance of specific populations of differentiated neurons derived from the neural crest (2). NGF has also been shown to have trophic interactions on central cholinergic neurons (3). widening its range of targets on differentiated neuronal cell populations. A less appreciated function of NGF is its ability to stimulate cell division. A mitogenic role for NGF acting on embryonic, mitotically active neuron precursor cells, was described in the early days of NGF history (4). This function has found strong support in recent years by the demonstration of a mitogenic effect of NGF on neural crest-derived cell lines (5), neuro-epithelial stem cells (6) and pluripotent adrenal chromaffii cells (7). The specificity of NGF action on subsets of cells in the nervous system
suggested
the existence of
other neurotrophic factors, which led to the discoveries of brain-derived neurotrophic factor (BDNF, ref.8) and neurotrophin-3 (NT-3, mf.9). All three neurotrophic factors possess neurite outgrowth promoting activity on embryonic chick dorsal root ganglion explants (10). Biological activities so far examined suggest distinct as well as overlapping properties for each neurotrophic factor (11). All three neurotrophic factors sharestructural similarities. Comparison of NGFs among different species (12) also indicate that neurotrophic factors are evolutionarily conserved
73
The significance of
0006-291X/92 $1.50 Copyright 0 1992 by Academic Press, Inc. Ail rights of reproduction in any form reserved.
Vol.
184,
No.
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
these similarities for understanding the primary role of neurotrophic factors in nervous system development is yet to be elucidated. Existence of a molecule(s) functionally or biochemically related to vertebrate neurotrophic factors has not been reported in invertebrates. Here we report that Drosophila possesses neurotrophicfactorlike functional activity. Biochemicaland molecularcharacterizationof this Drosophila
activity will
ultimately allow usto dissectthe fundamentalrole of neurotrophicfactorsin nervoussystemdevelopmentin anorganismamenableto geneticanalysis. Materials and Methods Drosophila embryonic cell culture. Drosophila CantonS embryoniccells werepreparedfor culture by Seecofs method(13). Briefly, the embryoswere collectedfor 2 hourson standardcorn meal-agar platesandallowedto develop for 3 hoursat 25°C. The early gastrulastageembryoswere dechorionatedin 1.5%sodiumhypochlorite/47.5%ethanol,andhomogenizedwith a Dounce homogenizerin Schneider’smodifiedDrosophila medium,supplemented with 18% fetal calf serum and0.2pg/mlbovine insulin. The homogenate waspassedthrough a Nytex meshfilter (50 p) andthe singlecellswereplatedin the abovemedium.After attachment(15 minutesafter plating), the mediumwaschangedto experimentalconditions.The cultureswere maintainedat 25°C. by Analysesof NGF effects in DrosophilaembryoniccelLFin culture. Cell growth wasassessed countingcells 8 hoursafter plating,whenthe majority of the cells haveceasedto divide but have not yet differentiated. Culturesweretrypsinized,stainedwith trypan blue and cellswere countedwith a hemocytometer.To assess differentiation,myotubesandneuronclusterswerescoredfrom terminally differentiatedcultures24 hoursafter plating. Cellswerefixed in 4% pamformaldehydeandcounted through a calibratedreticle in the ocularlensof a phasecontrastmicroscope.Ten fields wereexaminedfor eachsamples.Mouse NGF (2.5s NGF, culture grade)wasobtainedfrom Collaborative Research.Affinity purified sheepanti-mouseNGF antiserum(14) wasusedat 250 rig/ml. Assayfor neurite outgrowth promoting activity. Dorsalroot sensorygangliawere isolatedfrom 8 to 10day old chick embryosand placedon FalconPremariadishesin the presenceof smallquantitiesof freshculture medium(1: 1 mixture of serum-freeRPM1 1640,andSchneider’smodifiedDrosophila mediumsupplementedwith 5% fetal calf serum).After attachmentof the ganglia(about30 minutes), 0.5 ml of experimentalmediumwasaddedto eachdish. The gangliawerecultured at 37°Cin a humidified CO2 incubatorfor 24 hours. The resultswerephotographedusingan OlympusVanox phase contrastmicroscope,after fixation of the gangliain 4% paraformaldehydeDrosophila cultureconditionedmediumwasobtainedasfollows: Gastrulastagecellswerecultured asdescribedabove,except that the cellswere platedat 4 x106 cellsper60 mmdishin fresh Schneider’smodifiedDrosophila mediumsupplemented with 5% fetal calf serum. After 5 days in culture, the mediumwascollected,centrifuged,passedthrougha 0.2 p Gelmanfilter andusedimmediately. Northern analysisof Drosophilapoly(A)+RNA. Drosophila poly(A)+RNA sampleswereanalyzed usinga 32P-labelledcRNA probederivedfrom a mouseNGF cDNA clone, 12E4(15). Poly(A)+RNA wasglyoxalated andelectrophoresed overnight through a 1.5% agarosegel andtransferred onto a nitrocellulosemembrane.Hybridization wasperformedovernight at 60°C in 50% formamide,SXSSC, 50 mM NaPi, pH6.5,0.5X Denhardt’s,0.1%SDS, 1mM EDTA, 250 vg/ml salmon sperm DNA, and 10 pglml Drosophila 3rd instarpoly(A)-RNA. The NGF probewasusedat 106 cpm/mlof hybridization buffer. Washingwasin 2XSSC. 0.1% SDS at room temp.for 30 minutes twice, followed by a final washat 60°C for 20 minutesin 0.1 X SSC, 0.1% SDS. Drosophila 6-tubulin cRNA probewasusedto quantitatepoly(A)+RNA samplesfrom different stagesof development, sinceB-tubulinmRNA is found at constantlevelsthroughoutDrosophila development(16). After strippingthe mouseNGF cRNA probehybridization in 0.1XSSC, 0.1% SDS at 80°C for 30 min, the tubulin probewasusedasdescribedabove. The final washwasat 65°C for 30 minutesin 0.1X SSC and0.1% SDS. 74
Vol.
184,
No.
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Results and Discussion
Drosophila embryonic cells in culture produce neurite outgrowth promoting activity. The classic in vitro neuriteextensionassayof chick embryonic sensoryganglia(1) wasusedto examinewhether Drosophila
cells produce a neurite outgrowth
promoting
activity similar to vertebrate neurotrophic
factors. When dorsalroot sensorygangliafrom chick embryosweregrown in freshculture medium, no neurite outgrowth was observed as shown in Figure 1A. Addition of mouse NGF to this medium
resultedin a characteristicneuritehalo (Figure 1B). The mouseNGF-inducedneurite outgrowth was inhibited by concomitantadditionof anti-mouseNGF antibodies(Figure 1C).
Figure 1. Neurite outgrowth promoting activity of conditioned medium from Drosophila embryonic cells in culture on chick embryonic dorsal root ganglion explants. A: Fresh culture medium (1: 1 mixture of Schneider’s modified Drosophila medium supplemented with 5% fetal calf serum and serum-free RPMI 1640); B: Fresh culture medium + mouse NGF; C: Fresh culture medium + mouse NGF + anti-mouse NGF antibodies; D: 1: 1 mixture of serum-free RPMI 1640 and Drosophila culture conditioned medium; E: Medium as in D, with anti-mouse NGF antibodies. Mouse NGF: 20 rig/ml; anti-mouse NGF antibodies: 250 rig/ml. The bar at the right bottom indicates 0.2 mm. 75
Vol.
184, No. 1, 1992
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Addition of Drosophilaconditionedmedium,obtainedfrom 5day culturesof Drosophila gastrula stageembryonic cells,alsoresultedin stimulationof significantneuriteoutgrowth in the absenceof any mouseNGF (Figure 1D). The neuriteoutgrowth in the presenceof Drosophilaconditionedmedium wasblockedby anti-mouseNGF antibodies(Figure 1E). Theseresultsindicatethat Drosophila cellsin culture producea factor which is functionally similarto vertebrateneurotrophicfactors in stimulatingneuriteoutgrowthof chick embryonic sensoryganglia. In addition,the resultssuggest that the factor producedby Drosophilacells sharesimmunologicalhomologyto mouseNGF, since the neuriteoutgrowthpromotingeffect of the conditionedmediumwasblockedin the presenceof antibodiesto mouseNGF. Neuriteoutgrowth observedin the presence of Drosophila conditioned mediumwasnot dueto increasedspreadingof the connectivetissuecells,sinceno neurite outgrowth wasobservedin the presenceof anti-mouseNGF in spiteof a similarspreadingof connective tissue cellsin suchcultures. The effect of the conditionedmediumin stimulatingneuriteoutgrowth increasedwith the ageof the conditionedmedium. The conditionedmediumobtainedfrom 5 day-old culturesof Drosophilaembryoniccellswas moreeffective thanthosefrom 3-day or l-day cultures (datanot shown). This couldbe dueto the accumulationof the neuriteoutgrowth promotingactivity in the conditionedmediumover prolongedculture periodfor Drosophilaembryoniccells, or it could be due to an increasein the productionof the neurite stimulatingactivity by more differentiated and older cells in the culture.It shouldbe notedthat Drosophilaconditionedmedium,althoughqualitatively similarto mouseNGF, waslesseffective than the purified growthfactor in stimulatingneurite outgrowth under the experimentalconditionsemployed. This maybe duein part to the conditionsfor obtainingDrosophilaconditionedmediumor alternatively, the less-than-optimal effect of the Drosophila conditionedmediumcouldbe dueto the biochemicalnatureof the Drosophilafactor. Examination of thesedifferent possibilitiesis technicaliy not feasiblewithout a partial purification of the Drosophila neuriteoutgrowthpromotingactivity, due to the differencesin culture conditionsbetween the Drosophila embryoniccellsandembryonicchick ganglia. Mouse NGF stimulatesdivision of Drosophila embryonicneuronprecursor cells in cukure. Highly reproducibleprimary culturescanbe obtainedfrom ceilsof early gastrulastageDrosophila embryos(13). In thesecultures,aninitially undifferentiatedpopulationof cellsundergoesrapid multiplication and terminally differentiate into multinucleatedmyotubesanddifferentiatedneurons.The Drosophilaculture systemoffers a uniqueopportunityto examinethe effects of variousepigenetic substances on the early phaseof neuronaldevelopment,whenprecursorcells of the nervoussystem are still mitoticahy competent(17, 18). Addition of mouseNGF to the culture mediumresultedin a 2.4 fold increasein cell numberover the control cultures8 hoursafter plating(Figure 2A). Addition of mouseNGF did not affect the percentageof trypan blue positive, deadcells at 8 hours(about5% in ail cultures), indicatingthat the stimulatoryeffect of mouseNGF on the cell numberis morelikely to be due to an increasein celI division ratherthanan increasein the numberof viable cells. The mouseNGF effect wasconcentration dependent,andwasmost effective at 2 rig/ml(lo-%). Concomitantadditionof mouseNGF andanti-mouseNGF antibodiesabolishedthe NGF effect on the cell numberincrease(Figure2B). Theseresultsindicatethat the increasein cell numberobservedafter NGF additionto Drosophila embryoniccell culturesis due to NGF itself, andnot to possiblecontaminantsin the NGF prepara76
Vol.
184,
No.
1, 1992
BIOCHEMICAL
0
0.2
2
NGF
b.gfd
AND
BIOPHYSICAL
20 NGF Ab
-
+ -
RESEARCH
+ +
COMMUNICATIONS
;
? 20
NGF Ab
-
+ -
+ t
0
+
0.2 NGF
2 @hU
Finme 2. Effect of mouse NGF and anti-mouseNGF antibodieson the growth and differentiation of Drosophila gastrula stage embryonic cells in culture. A: Effectof mouse NGFoncell proliferation. Cellswereplatedat 0.8X 106per35mmculturedish.B: Supression of growth-stimulating effectof mouse NGFby anti-mouse NGF antibodies. Cellswereplatedat0.5X 106per35mmculturedish. C: Supression of cellgrowthby anti-mouse NGFantibodies in highdensitycultures.Thecellswere platedat 2 X 106per35mmculturedish.Notethattheantibodies inhibitedgrowthnotonly in the presence, butalsoin theabsence of mouse NGF. D: Effectof mouse NGFonthedifferentiationof Drosophilaembryoniccells. An average of 10fieldsperplatewereexamined andscoredfor neuron clusters andmyotubes, eachwith randomlychosen differentunitareaof 500pm*. Thisexperiment wasdonein parallelwith theexperiment in A. Forall experiments. barsrepresent themeanof counts fromduplicateplateswith standard error. tion. The resultsalsosuggest that the Drosophilacellsmay possess a receptorwhich canrecognize mouseNGF. Anti-mouseNGF antibodiesalonedid not haveany effect on cell numberin thesecultures. However, when the cellswereculturedat highdensities,the anti-mouseNGF antibodiesnot only suppressed the increasein cell numberobservedin the presenceof mouseNGF, but alsoinhibited the cell numberincreaseby 50% in the controlculturesto which no NGF hadbeenadded(Figure 2C). The effect of mouseNGF on cell growth wasalsolessevident in high density cultures(Figure 2C). Theseresults,in addition to the neuriteoutgrowthpromotingactivity of the Drosophilacell culture conditionedmedium,suggestthe presenceof a neurotrophicfactor in the Drosophilaculture medium. Drosophilacells in culture appearto be producingandrespondingto their own trophic factor whoseeffect canbe mimickedor replacedby mouseNGF. Anti-mouseNGF antibodiesappearto interfere with the action of the Drosophilatrophic factor presentin the high densitycultures. The increasein cell numberobservedin the presenceof mouseNGF seemsto resultfrom the preferentialstimulationof the growth of neuronalprecursorcells. Cellsthat form terminally differentiatedneuronclustersareclonally related,arisingfrom a singleneuronprecursorcell (17). Hence,our of resultswhich showan increasein the numberof neuronclusters per unit areain the presence mouseNGF (Figure 2D), strongly suggestsanunderlyingincreasein the numberof neuronprecursor cells. The numberof myotubesperunit arearemainedconstant. Severalmechanisms are possiblewhich couldresultin the increasein cell numberandthe subsequentenrichmentof the neuronalphenotypeof the Drosophilaculturesin the presenceof mouse NGF. Mouse NGF (or the equivalentDrosophilafactor) may stimulatethe rate of cell divisionor prolongthe duration of the growth phaseof Drosophilaneuronprecursorcells. Alternatively, the 77
Vol.
184,
No.
1, 1992
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
Figure 3. Northern analysis of Drosophilcl poly(A)+RNA. Top panel: Drosophila poly(A)+RNA samples from (2) 5hr embryos, (3) 1st (4) 2nd and (5) 3rd instar larva and (6) young adults (within 1 week of hatching), probed with a cRNA probe derived from mouse NGF cDNA clone, 12E4. Numbers represent lanes. Twenty ug of Drosophila poly(A)+RNA samples were loaded in each lane. Lane (1) is male mouse submaxillary gland total RNA (300 ng), used as a standard. Bottom panel: The same filter probed with Drosophila B-tubulin cRNA probe The arrow indicates the position of 16s E. cob ribosomal RNA. The mouse NGF cDNA clone 12E4 (gift from Dr A.Gray, Genentech), originally carried in a pBR322 plasmid vector, was subcloned into a Bluescribe plasmid vector and the cRNA probe was synthesized from the EcoKI-linearized Bluescribe clone using T3 RNA polymerase in the presence of 32P-UTP (New England Nuclear, 3000 Wmol). For the tubulin probe, a 398 bpDrall-EcoRV subfragment of the B-tubulin cDNA pTu56 (gift of Dr. R. Renkawitz-Pohl, Max-Planck Institute, Munich) was subcloned into Bluescript plasmid vector and the cRNA probe was synthesized from the EcoRV-linearized clone.
factor may play a role in influencing theDrosophila
gastrula stage embryonic
cells to takea neuronal
pathway, thus increasingthe proportionof cellswhich becomeneuronprecursorcells,aswasrecently suggestedwith BDNF for pluripotentneuralcrest-derivedcells in the chick embryo(19). Thesedifferent possibilitieswill haveto be examinedfurther by detailedanalysesof neurotrophic factor effect(s) on the growth and differentiationpropertiesof the Drosophila
embryoniccells in
culture. A Drosophila mRNA is recognizedby a mouseNGFprobe. In orderto obtainadditionalmolecular evidencefor a Drosophila
factor that resemblevertebrateneurotrophicfactors,Drosophila
poly(A)+RNA wasexaminedby Northernblot analysis,usinga probespecificfor mouseNGF mRNA. The mouseprobehybridized to a singlebandof Drosophila
poly(A)+RNA in all stagesof
developmentexamined(Figure 3). Crosshybridization of a Drosophila mRNA bandwith the mouse NGF probeunderstringentcross-species hybridization conditionssuggests a strongsimilarity in the sequenceof mouseNGF andthe Drosophila
mRNA band. The sizeof the Drosophila
mRNA (1.4
Kb) was alsosimilarto that of themouseBNGFmlWA (1.3 Kb), aswell asto other known vertebrateneurotrophicfactor mlWAs (8,9). In addition,the level of theDrosophila
mRNA appearedto
changeasa function of developmentalstages.Significantly, the mRNA was alreadydetectablein 5 hour embryos,when thereis active multiplicationof neuronprecursorcellsbut no overt detectable neuronaldifferentiation. The Drosophila
mRNA increaseddramaticallyin the larval stages,coinci-
dent with our observationson the neuriteoutgrowthpromotingactivity of the 5 day-old conditioned medium(in vivo equivalentof the latelarval stagecells). ConcludingRemarks. Our studiesdemonstratethe existenceof a Drosophila moleculewith both structuralandfunctional similaritiesto the vertebrateneurotrophicfactors. Many factorsin the past havebeenshownto stimulateneuriteoutgrowthof chick embryonicsensoryganglia andin this regard, the neuriteoutgrowth promotingactivity of the Drosophila conditionedmediumitself is not surprising. However, in our studies,actionsof this factor on Drosophila 78
cellscanbe mimickedby
Vol.
184, No. 1, 1992
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
mouseNGF, the mostextensively studiedandreadily availableneurotrophicfactor of the NGF-family of peptides. Much of our presentknowledgeon the actionsandlocalizationsof vertebrateneurotrophicfactors are basedon NGF. Recentstudieson the NGF-family of neurotrophicfactorssuggest,that many of the earlierstudiesattributing the functionalactivity to NGF may requirereevaluation.Theseinclude indicationsthat NGF, BDNF andNT-3 may usethe samereceptorsto mediatetheir effects (20,21), andthe possiblecrossreactivity of anti NGF antibodiesto NT-3 andBDNF (22). In our present studies,we have usedmouseNGF asa point of referencefor the Drosophilaneurotrophicactivity. This doesnot precludepossibilitiesthat the Drosophilafactor may turn out to be moresimilarto the other membersof the vertebrateneurotrophicfactors,or found to bequiteuniquewith limited but significantsimilarity to all vertebrateneurotrophicfactors. The determinationof the natureof the Drosophila molecule,andthe extent of its homologyto vertebrateneurotrophicfactorsmustcome from sequencingstudies. Informationgainedfrom suchstudiesmay alsobring insightinto understandingthe structure-functionrelationshipandthe evolutionary aspectof the vertebrateneurotrophic factors. Acknowledgments The authorswish to thank Dr. H. Thoenenfor his supportin the initial phaseof this work, during a sabbaticalof I. Hayashiin his laboratory. The work wassupportedin part by a grant from NSF (BNS-8616936)andNINCDS (POlNS18858-07Sl). References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
Levi-MontaJcini, R. andHamburger,V. (1953). J. Exp. Zool.. 123,233-288. Levi-Montalcini, R. andAngeletti, P. U. (1968). PhysioLRev. 48,534-569. Williams, L. R., et al. (1986). Proc. Natl. Acad. Sci., USA 83,9231-9235. Levi-Montalcini, R. andBooker, B. (1960).Proc. Natl. Acad. Sci., USA 46,373-384. Revoltella, R. P. andButler, R. H. (1980). 1. Cell. Physiol. 104,27-34. Cattaneo,E. andMcKay, R. (1990). Nature347,762-765. LiIlien, L. E. andClaude,P. (1985).Nature 317,632-634. Leibrock, J., eral. (1989). Nature 341,149-152. Maisonpierre,P.C., er al. (199Oa).Science247,1446-1451. Thoenen,H. (1991). TINS, 14,165-170. Maisonpierm,P. C., et al. (199Ob).Neuron 5,501-509. Ullrich, A.,er al. (1983). Nature 303,821-825. Meier, R.,eral. (1986). EMBO J. 5, 1489-1493. Stoekel, K., et al. (1976). J. Neurochem.26,1207-1211. Bialojan, S., et al. (1984). EMBO J. 3,2543-2548. Seecof,R. L. andUnanue,R. L. (1968). Exp. Cell. Rex 50,654-660. Furst, Aand Mahowald,A. P. (1985). Dev. Biol. 112,467-476. Salvaterra,P. M., et al. (1987). J. Neurosci.7, 10-22. Sieber-Blum,M. (1991). Neuron 6,949-955. Klein, R., et al. (1991). Cell 65, 189-197. Cordon-Cardo,C., eral. (1991). Cell 66, 173-183. Acheson,A., et al. (1991). Neuron,7,265-275.
79
