effects of the neurotrophins nerve growth factor, neurotrophin-3, and brain-derived neurotrophic...

Effects of the Neurotrophins Nerve Growth Factor,Neurotrophin-3, and Brain-Derived NeurotrophicFactor (BDNF) on Neurite Growth from AdultSensory Neurons in Compartmented Cultures

K. Kimpinski,1 R. B. Campenot,2 K. Mearow1

1 Division of Basic Medical Sciences, Faculty of Medicine, Memorial University of Newfoundland,St. John’s, Newfoundland A1B 3V6, Canada

2 Department of Anatomy and Cell Biology, University of Alberta, Edmonton, Alberta, Canada

Received 29 January 1997; accepted 15 May 1997

quantitated. There was no increased neurite extensionABSTRACT: We used compartmented culturesinto NGF plus NT3 compartments, while the combina-to study the regulation of adult sensory neurite growthtion of BDNF plus NGF resulted in an inhibition ofby neurotrophins. We examined the effects of the neu-neurite extension compared with NGF alone. We thenrotrophins nerve growth factor (NGF), neurotrophin-investigated whether the regrowth of neurites that had3 (NT3), and BDNF on distal neurite elongation fromoriginally grown into NGF subsequent to in vitro axo-adult rat dorsal root ganglion (DRG) neurons. Neu-tomy still required NGF. The results demonstratedrons were plated in the center compartments of three-that unlike adult sensory nerve regeneration in vivo,chambered dishes in the absence of neurotrophin, andthe in vitro regrowth did require NGF, and neitherneurite extension into the distal (side) compartmentsBDNF nor NT3 was able to substitute for NGF. Sincecontaining NGF, BDNF, or NT3 was quantitated. Ini-the initial growth from neurons after dissociation

tial proximal neurite growth did not require any of (which is also a regenerative response) did not requirethe neurotrophins, while subsequent elongation into NGF, it would appear that neuritic growth and re-distal compartments required NGF. After neurites growth of adult DRG neurons in vitro includes bothhad extended into NGF-containing distal compart- NGF-independent and NGF-dependent components.ments, removal of NGF by treatment with anti-NGF The compartmented culture system provides a uniqueresulted in the cessation of growth with minimal neu- model to further study aspects of this differential reg-rite retraction. In contrast to the effects of NGF, no ulation of neurite growth. q 1997 John Wiley & Sons, Inc.distal neurite elongation was observed into compart- J Neurobiol 33: 395–410, 1997ments with BDNF or NT3. To examine possible addi- Keywords: adult dorsal root ganglion neurons; NGF;tive influences, neurite extension into compartments BDNF; NT3; neurite growth; axotomy; compart-containing BDNF plus NGF or NT3 plus NGF was mented cultures

INTRODUCTION (NT) family of molecules. These factors have beenshown to influence the differentiation, development,

Nerve growth factor (NGF), BDNF, and neuro- and survival of a wide variety of neurons in both thetrophin-3 (NT3) are members of the neurotrophin central nervous system (CNS) and the peripheral

nervous system (PNS). Many studies have providedCorrespondence to: K. M. Mearow information pointing to the importance of the NTsContract grant sponsor: Canadian NeuroScience Network in the differentiation, early development and earlyContract grant sponsor: Natural Sciences and Engineering

postnatal survival of the various classes of sensoryResearch Council of Canadaq 1997 John Wiley & Sons, Inc. CCC 0022-3034/97/040395-16 neurons (reviewed in Davies, 1996; Kalcheim,

395

1864/ 8p29$$1864 08-18-97 19:34:49 nbioa W: Neurobio

396 Kimpinski, Campenot, and Mearow

1996; Ockel et al., 1996). While postnatal and adult ate the feasibility of using adult DRG neurons inthe compartmented culture system and to examinesensory neurons do not require NTs for their contin-

ued survival, the continued maintenance of particu- the influence of NGF, NT3, and BDNF on thegrowth of neurites from both neonatal and adultlar mature phenotypic characteristics ( including

peptide content and physiologic responsiveness) neurons in such cultures. The results of these experi-ments have indicated that adult DRG neurons adaptseems to at least be responsive to, if not dependent

upon, NTs such as NGF and NT3 (Mulderry, 1994; readily to the compartmented system and provide auseful model with which to examine both local andWright and Snider, 1996; reviewed in Lindsay,

1992). retrograde effects of NTs. In these cultures (inwhich the neuronal cell bodies were plated in theNeurotrophins also play a major role in the regu-

lation of neuronal growth, and particularly in absence of NTs and thus only the distal axons hadaccess to the NTs), NGF alone was able to elicitgrowth or regeneration as a response to nerve injury.

In the mature PNS, many types of neurons possess distal neurite elongation from the DRG neurons.Once neurites had extended into the distal compart-the capability both to sprout new axon collaterals

when presented with the appropriate stimulus and ments, further extension could be stopped by anti-NGF treatment. Furthermore, following axotomy atto regenerate axons that have been severed. In adult

dorsal root ganglion (DRG) neurons, NGF is re- a time when the neuritic growth had stabilized, NGFwas also required for neurite regeneration. Neitherquired for collateral sprouting of intact neurons, yet

has no observable effects on the regeneration of BDNF nor NT3 alone was able to elicit growthof neurites out of the central compartments. Thethose neurons (Diamond et al., 1992a,b; Mearow et

al., 1994; Mearow and Kril, 1995). Although these combination of BDNF plus NGF resulted in inhibi-tion of neuritic extension, while the combination ofstudies have provided evidence concerning the role

of NGF in the regulation of axonal growth, the in NT3 plus NGF had little effect over and above thatof NGF alone. The results are discussed with respectvivo system cannot be easily manipulated to investi-

gate the underlying mechanisms involved in collat- to the role of the NTs and the similarities and differ-ences between the growth processes exhibited byeral sprouting as compared to regeneration. The

compartmented culture model (Campenot, 1977, sensory neurons in vivo and in vitro.1992) provides an attractive system with which tofurther investigate the role of the NTs in regulatingadult neuronal outgrowth. In this model, the neu- MATERIAL AND METHODSronal cell bodies and the proximal portions of theirneurites remain in a center compartment, while the Culture Proceduresdistal portions of the axons extend into side com-

Dorsal root ganglia were dissected from postnatal day 1partments that are effectively isolated from the cen-or adult (100–200 g) Sprague–Dawley rats (supplied byter chamber. In this way, local effects of NTs orMemorial University Vivarium). Ganglia were removedother agents on either the distal neurite terminals orfrom all spinal levels and the roots were trimmed andthe cell bodies can be examined. This, unlike usualthen chemically dissociated by separate incubations in

culture systems, is perhaps a closer approximation 0.25% collagenase (GIBCO-BRL, Burlington, Ontario,of the in vivo situation, especially in the case of the Canada), 0.25% trypsin (GIBCO-BRL), and finally, 0.52DRG neurons whose cell bodies and terminals are mg/mL soybean trypsin inhibitor (SBTI), 0.04 mg/mLseparated by substantial distances and are thus not DNase, and 3 mg/mL bovine serum albumin (BSA)likely exposed to similar environments. (Sigma, St. Louis, MO). Adult ganglia were treated with

collagenase for 1.5 h at 377C, followed by trypsin for 30Studies on the influence of NTs in vitro havemin, and finally, with the SBTI-DNAse for 5 min. Neona-generally used either embryonic or neonatal tissues;tal ganglia were dissociated with 30 min collagenase, 15while these neurons tend to be easier to establishmin trypsin, and SBTI-DNas for 5 min. The ganglia wereand maintain in culture, one complicating factor isthen mechanically dissociated by trituration with a flame-that neuronal survival is often dependent on thenarrowed pasteur pipette in L15 medium with 2% fetalpresence of a particular NT. In this regard, the usecalf serum (FCS) (GIBCO-BRL). The resulting cell sus-

of adult DRG neurons presents a useful model for pension in L15–air medium supplemented with 2% FCS,differentiating between the survival and growth-in- 100 U/mL penicillin/streptomycin (GIBCO-BRL), andfluencing effects of the NTs, since these neurons 3 g/100 mL glucose (Sigma) was centrifuged (1000 rpm)survive independently of added NTs (Lindsay, through a cushion of 15% BSA (Fraction V; Sigma) in1988; but see Acheson et al., 1994). L15 medium; this procedure eliminates much of the cellu-

lar debris and results in a neuronally enriched pellet. TheIn the present study, our objectives were to evalu-


Adult DRG Neurons in Compartmented Cultures 397

resulting cellular material was suspended in complete Axotomies were performed on day-7 cultures aftermeasurements of neurite extension had been recorded. AL15–air medium supplemented with 1.5% methylcellu-

lose and neurotrophins when required. jet of sterile water was projected at the most proximalportions of neurites growing into the side compartmentsCompartmented cultures were constructed as de-

scribed by Campenot (1992), with a modification in the using a syringe and a 23-gauge needle; this procedure wasrepeated four times in total and resulted in compartmentsculture substrate: 35-mm culture dishes were coated with

polyornithine (10 mg/mL) and laminin (5 mg/mL). Paral- totally devoid of neurites.lel tracks in the substrate were made by scratching theculture dish surface with a pin-rake (Tyler Research In-struments, Edmonton, Alberta, Canada). Subsequently, a

Measurementsthree-chambered Teflon divider (Tyler Research Instru-ments) was then seated on top of the tracks as described Cultures were examined using the 116 objective of aelsewhere (Campenot, 1992). Zeiss inverted microscope equipped with an ocular mi-

Dissociated neurons were plated into the center com- crometer. Growth measurements were taken as neuritepartments and maintained in L15–air medium containing extension from the proximal edge of the grease barrier2% FCS, penicillin/streptomycin plus 1.5% methylcellu- of the side compartment to the most distal neurite pro-lose, and 20 mM cytosine arabinoside (Sigma). For neo- cesses in a given track. Of the total 20 tracks in a sidenatal cultures, 2.5S NGF (Cedarlane Labs. Hornby, On- compartment, only 14 were measured because the cul-tario, Canada) was added to a final concentration of 10 tures consistently showed little or no neurite growth inng/mL. Adult neurons did not require NGF in the center the top or bottom three lanes. Data from the 14 lanes ofcompartments for survival. Cultures had neurites emerg- each culture were pooled with equivalent cultures at theing into the side compartments after 3 days, and after 5 same time point. The total number of cultures examineddays neurites were well established; at this point, NGF at each time point ranged from 10 to 15. Data were ana-was removed from the center compartments. Compart- lyzed using two-way analysis of variance with MINITABments containing only neurites received varying concen- statistical software. Error bars in all figures indicate stan-trations of neurotrophins (NTs), ranging from 0 to 100 dard errors of the mean (SEM); where error bars are notng/mL of either NGF, BDNF, or NT3 (Cedarlane Labs; seen, they do not exceed the width of the symbol. In allAmgen, Thousand Oaks, CA). In some situations anti- cases, significance was considered to be p õ 0.05 and isNGF immunoglobulin Gs (IgGs) (Cedarlane Labs, Hor- indicated with an asterisk.nby, Ontario, Canada) were added at a concentration of Neuritic density measurements were performed as de-100–200 ng/mL to neutralize NGF. Medium was scribed below.changed every 3–5 days and cultures were maintainedin an air atmosphere at 377C.

The anti-NGF antibody has been previously character-ized (Coughlin and Collins, 1985; Diamond et al., Immunocytochemistry1992a,b; van der Zee et al., 1995). It is cross-reactive

The medium was carefully aspirated from the cultures,with NGF and NT3 on enzyme-linked immunosorbentwhich were then washed with phosphate-buffered salineassays (J. Diamond, personal communication) and ap-(PBS) and fixed for 10 min in 4% formaldehyde in PBS.pears to inhibit neurite production by neonatal mouseAfter washing in PBS, the cultures were incubated in aDRG neurons in response to both NGF and NT3 (vanblocking solution (10% normal goat serum, 0.1% Tritonder Zee et al., 1995). However, it should be noted thatX in PBS) for 1 h. Subsequently the cultures were ex-the concentrations of the anti-NGF IgG preparation em-posed to the primary antibodies for overnight incubationployed by van der Zee et al. (1995) are two orders ofat 47C.magnitude higher than those we used in our experiments

The primary antibodies employed are as follows: for(10 mg/mL vs. 100–200 ng/mL), and higher than thosep75, the monoclonal MC192 (Oncogene Sciences, Cam-required to completely block NGF’s activity. This anti-bridge, MA; 1:100 dilution); TrkA, polyclonal SC-118,NGF antibody has been previously shown to neutralizenonreactive with TrkB or TrkC (Santa Cruz Biotechnol-the activity of purified mouse NGF (200 ng/mL) half-ogy, Santa Cruz, CA; 1:500 dilution); TrkB, polyclonalmaximally at 100 ng/mL and completely at 500 ng/mLSC-12, nonreactive with TrkA or TrkC, or truncated TrkB(Coughlin and Collins, 1985).(Santa Cruz Biotechnology; 1:500 dilution); TrkC, poly-clonal SC-117, nonreactive with TrkA or TrkB (SantaCruz Biotechnology; 1:500 dilution). The cultures werewashed and exposed to the biotinylated-secondary anti-Treatmentsbody for 1 h at room temperature; visualization of stainingwas completed using a Vectastain Kit (Vector Labs, Bur-Experimental manipulations included either the cell bod-

ies or the neurites being exposed to differing concentra- lingame, CA) and diaminobenzidine (DAB). Alterna-tively, fluorescein isothiocyanate (FITC)-tagged second-tions of NGF, BDNF, or NT3 (0–100 ng/mL of complete

medium). ary antibodies were used. Cultures were then coverslipped



under glycerol, examined, and photographed either with Influence of NGF on DRG Neuritea Leitz photomicrocopic system or a Hund Wilovert-S Growth in Compartmented Culturesinverted microscope. Using a computer-based imagingsystem (Northern Exposure; EMPIX Imaging, Missis- In initial experiments, neurons were plated in thesauga, Ontario, Canada), density measurements were center compartment in the presence of 10 ng/mLmade on cultures where neurites had extended into NGF. In these cultures, there was no neurite exten-NGF, subsequent to immunocytochemical staining for

sion of adult or neonatal neurons into side compart-p75 and visualization using horseradish peroxidase–avi-

ments that contained equivalent (10 ng/mL) or lessdin and DAB.NGF. However, if the distal compartments con-tained 100 ng/mL NGF, then there was substantialextension of the neonatal neurites but only minimalgrowth of the adult neurites [Fig. 1(a,b)] (see be-RESULTSlow). Subsequently, all cultures were plated in theabsence of NGF in the center compartments. In such

Culture Characteristics cultures, there was extensive growth of both theneonatal and adult neurites into distal NGF-con-In these experiments, we examined the influence oftaining chambers [Fig. 1(c,d)] . While the adultthe presence or absence of various neurotrophins onneurons did extend neurites in the absence of NGFthe survival and neurite growth of both neonatalin the center compartment, there was never any sig-and adult DRG neurons. Dissociated neurons werenificant elongation into distal compartments that didalways plated in the center compartments of thenot contain NGF.three-chambered Teflon dividers, essentially as pre-

The total extensions that both the neonatal andviously described (Campenot, 1992); in our cul-adult neurites reached at the end of the experimentstures, the medium in the center or side compart-(9 days in vitro) were similar for all concentrationsments was identical except for the addition ofof NGF. The rate of neurite extension over the 9-NT(s) . The measurements of neurite extensionday period ranged from 106 to 136 mm/day forwere taken from the point at which the neuritesthe adults and 105–157 mm/day for the neonatalemerged from under the grease barrier to the mostneurons. A significant dose-dependent effect of thedistal point in the lanes being examined; it shouldNGF was observed on the density of the neurites,be noted that at this point the neurites had alreadywith greatly increased neuritic density in the cul-extended in excess of 0.5 mm, since the width oftures containing the higher concentrations of NGFthe barrier itself was 0.5 mm.(Fig. 2; see Table 1). Thus, while the overall dis-Within 24 h of plating, neuritic growth was ob-tance that the neurites grew was not that differentserved in the center compartments. However, therebetween conditions, the effect of increasing NGFwas a lag phase before neurites were observed ex-concentrations was either to increase the branchingtending into the side compartments. Generally, inof the neurites resulting in an increased density ormost of the cultures examined, it was possible toto increase the number of neurons that extendedview neuritic processes just about to emerge intoneurites into the distal compartments. Backlabelingthe distal compartments, and in certain experimentalexperiments did not provide consistent results withconditions (see below) where elongation into therespect to this issue; that is, there was no clear evi-distal compartment did not occur, one could stilldence of differences in the number of neurons la-observe neurites at the edge of the compartmentbeled with the differing concentrations of NGF.barrier. This suggests that under the differing condi-

A significant observation was that while the adulttions where elongation did not proceed, the neuronsneurons survived in the absence of NGF, the neona-or neurites still had an equivalent opportunity total neurons were able to survive plating in the ab-extend into the distal compartments given the appro-sence of NGF only if there was a distal compartmentpriate stimulus.with NGF available. When the neonatal neuronsWe also observed that in the case of the adultwere plated in the absence of NGF and not providedneurons, none of the NTs were required for survival,with an NGF-containing distal compartment, by dayas expected. With respect to the neonatal neurons,5 of culture, the percentage of surviving neuronsNGF was required for their survival; while we didwasõ20% (data not shown). The neonatal culturesnot carry out extensive analyses, BDNF and NT3were employed as a way of assessing optimal condi-did not appear to have significant effects on thetions for the adult neurons. Because of the low sur-neonatal survival and could not replace NGF in this

regard. vival in the absence of NGF and our primary interest



Figure 1 Nerve growth factor (NGF) is required for distal neurite growth of sensory neuronsin compartmented cultures. Dissociated sensory neurons were plated in the center compartmentsof three-chambered Teflon dividers in either the presence of NGF [(A) neonatal neurons or(B) adult neurons,] or the absence of NGF [(C) neonatal neurons or (D) adult neurons];the distal compartments contained medium with no NGF or varying concentrations of NGF.Measurements of growth were taken as total neurite extension as described in Methods. (A)Neonatal sensory neurons plated in compartments with NGF (10 ng/mL) displayed no growthinto distal compartments containing 10 ng/mL NGF, but did extend into distal compartmentswith 100 ng/mL NGF. (B) Adult sensory neurons plated in compartments with NGF (10 ng/mL) displayed little growth into the NGF-containing distal compartments over the 7-day cultureperiod. In contrast, neurons plated in the absence of NGF [(C) neonatal, (D) adult] showedextensive distal neurite growth into the side chambers containing NGF; no growth was seenin the absence of NGF. Each point represents the mean of 7–10 cultures { S.E.M.

to the adult cultures, neither BDNF nor NT3 in anyin adult neuronal growth, subsequent experimentsof the concentrations used in these experiments wasfocused on adult cultures.able to support neurite growth into the distal com-partments (Figs. 3 and 4).

To ascertain whether these cultures were healthyInfluence of BDNF and NT3and capable of extending processes into the sideon DRG Neurite Growth incompartments, the following manipulation was car-Compartmented Culturesried out. At day 7 of culture, the medium in the

The influence of BDNF and NT3 on eliciting neurite distal compartments (originally containing BDNFgrowth was also investigated. Cultures were pre- or NT3) was replaced with medium containing 25pared as above, plating the neurons in the absence ng/mL NGF. Following a 1-day lag, all the culturesof any NT in the center compartment. The distal displayed relatively rapid neurite growth into thecompartments contained 0, 1, 10, or 100 ng/mL of NGF-containing compartments [Fig. 3(a,b)] , indi-

cating that there was nothing intrinsically impairedBDNF, or 1, 10, or 100 ng/mL of NT3. With respect



Figure 2 Density of neuritic growth increases with increasing NGF concentrations in the sidecompartments. Distal neurites were labelled with Fluorogold (2% in sucrose/HBBS) for 2–4h; the cultures were then washed with HBSS and the medium was replenished and incubatedfor a further 24 h. The cultures were subsequently fixed and labeled neurites were visualizedwith ultraviolet illumination. The photomicrographs are representative examples of neuritesgrowing into distal compartments containing (A) 100 ng/mL, (B) 10 ng/mL, or (C) 1 ng/mLNGF. In some cultures at day 7, anti-NGF was added to the medium; (D) and (E) arerepresentative examples of cultures treated with anti-NGF for 5 days and labeled with Fluoro-gold. Calibration bar Å 100 mm.

in these cultures. It is also important to note that in BDNF Plus NGF in Distal CompartmentsResults in an Inhibition of Extensionthese cultures, in the absence of any NT in the side

compartment for 7 days (i.e., the 0 ng/mL BDNF To assess the possibility that the combination ofand NT3 conditions) , NGF was still able to elicit BDNF plus NGF in the distal compartments mightneurite elongation at essentially the same rate of have an additive influence on neurite extension, cul-growth (250–300 mm/day). tures were set up with 100 ng/mL BDNF with either

To rule out the possibility that lack of NTs was 10 ng/mL NGF or 100 ng/mL NGF. As suggesteddeleterious to any specific neuronal population, cul- by the studies of Barker and Shooter (1994), ourtures were plated with various combinations of NTs expectation was that BDNF would interfere with thein the center and distal compartments. Thus, neu- action of NGF, likely by competing for p75 low-affin-rons were plated with 1 ng/mL BDNF in the center ity receptor-binding sites. The data from these experi-compartment and provided with distal compartments are presented in Figure 4(a), where it is appar-ments containing 100 ng/mL BDNF or NGF, or 10 ent that the effect of adding BDNF to NGF was tong/mL NT3; 1 ng/mL NT3 in the center and 100 inhibit neurite extension. With higher concentrationsng/mL BDNF or NGF, or 10 ng/mL NT3 in the of BDNF (1 mg/mL; data not shown), the extensionsides, and finally, 1 ng/mL NGF in the center and was almost completely inhibited. These results suggest100 ng/mL BDNF or NGF, or 10 ng/mL NT3 in the importance of the appropriate interactions of boththe sides. The results of this experiment confirmed p75 and TrkA in mediating the actions of NGF.the previous observations that neuronal growth oc-

NT3 Plus NGF in Distal Compartmentscurred only into distal compartments containingIs Not AdditiveNGF; there was no significant outgrowth into any

of the compartments containing NT3 or BDNF or In the case of NT3, we were interested in determin-ing whether the addition of NT3 to NGF wouldno NT (Table 2).



Neurons Plated in the Presence ofAnti-NGF Exhibit Normal NeuriteExtension into NGF-ContainingDistal Compartments

Our observations indicated that there was significantneuritic growth in the center compartments in theabsence of added NGF. To examine whether theremight be some endogenous NGF being producedby the small number of nonneuronal cells that wereoften found in the cultures, the effect of platingneurons in the presence of anti-NGF was assessed.

Figure 3 Neither BDNF nor NT3 supported neuriticgrowth into the side compartments. Adult neurons werecenter-plated in the absence of added NTs and providedwith distal compartments containing BDNF (A) or NT3(B). There was no neurite growth into the side chamberscontaining BDNF or NT3. On day 7 in vitro, the mediumin the side chambers was replaced with medium con-taining 25 ng/mL NGF. Replacing the BDNF or NT3with NGF resulted in rapid growth of neurites into theside chambers. Each point represents the mean { S.E.M.of seven cultures.

potentiate the neuritic growth response. Our initialexperiments were carried out with concentrations ofNT3 that should activate TrkC receptors primarily,

Figure 4 The effects of combining NGF with BDNFalthough at higher concentrations NT3 can also actor NT3 in the distal compartments. (A) Neurons werevia the TrkA receptor (Rodriguez-Tebar et al.,plated in the absence of neurotrophin and provided with

1992; Belliveau et al., 1997). We carried out further distal compartments containing either NGF (10 or 100experiments using 100 ng/mL NT3 and found that ng/mL), BDNF (100 ng/mL), or a combination ofeven at this higher concentration, there was no distal BDNF (100 ng/mL) with 10 or 100 ng/mL NGF. Thereneurite extension [Fig. 4(b)] . To investigate the was no extension into compartments containing BDNFpossibility that NT3 might potentiate the effects of alone (solid triangles); extension proceeded normally

into the NGF-containing compartments. The combinationNGF, neurons were provided with compartmentsof BDNF with NGF resulted in an inhibition of the distalcontaining 100 ng/mL NT3 plus either 10 ng/mLextension, although it was not blocked totally (open cir-or 100 ng/mL NGF, and the amount of distal exten-cles and squares) . (B) Neurons were plated in the ab-sion was quantitated. The results are presented insence of neurotrophins and provided with distal compart-Figure 4(b), where it is apparent that NT3 does notments containing either NGF (10 or 100 ng/mL), NT3act to increase the distal neurite extension; there(100 ng/mL), or a combination of NT3 (100 ng/mL)

were no obvious increases in neuritic density in with 10 or 100 ng/mL NGF. There was no extension intothese cultures, either. Thus, while NT3 can activate compartments containing NT3 alone (solid triangles) .TrkA, it also competes for p75, with the end result There was no further increase in extension in compart-conceivably being that the possible positive influ- ments with both NT3 and NGF compared with the NGF-ences of NT3 are offset by an effective decrease in only compartments. Each point represents the mean

{ S.E.M. of four cultures.the ratio of p75:TrkA available for NGF.



Table 1 Neurite Densities from Neurons Culturedwith 0, 1, 10, and 100 ng/mL NGF in DistalCompartments

Distal Compartment Density per Lane(ng/mL NGF) (mm2 1 103)

Control 0.01 38.47 { 14.6 mm2 *10 94.18 { 10.1 mm2 *100 140.50 { 14.4 mm2 *

Neuritic densities are taken as the average area covered bythe neurites growing across lanes in distal compartments. Valuesare expressed as the average density for single lanes. Control Åno NGF added to side compartments. Density was measured onday 7 in vitro and represents the mean { S.E.M. of 10 to 15lanes from two to three cultures.

* Differences are significant (p õ 0.001).

Neurons were exposed to 200 ng/mL anti-NGFIgGs at the time of plating and for the course of theexperiment. The presence of anti-NGF in the centralcompartments had little effect on initial proximalneurite growth. There was no obvious difference inthe amount of initial proximal neurite outgrowth,

Figure 5 Anti-NGF applied to cell bodies did not blockalthough we could not quantitate this initial out-distal neurite extension, but did if applied to the distal

growth because the density of the plated neurons neurites. (A) Neurons were plated in the center compart-made it difficult to determine the identified neuronal ments in the presence of 200 ng/mL anti-NGF (affinity-source of given neuritic processes. Furthermore, the purified IgGs); sister cultures were plated in the usualaddition of anti-NGF to the central compartment medium in presence of control IgGs. Distal neurite exten-did not significantly influence distal extension into sion was measured from days 5 to 9 after plating; that

there was no significant difference in the amount of neu-NGF-containing side compartments, as shown inrite extension. (B) Removal of NGF with anti-NGF treat-Figure 5(a) .ment halted neurite extension but did not result in neuriteretraction over a period of 5 days. Adult neurons wereInfluence of NGF Removal in allowed to send neurites into side compartments with

Established Cultures NGF (10 or 100 ng/mL) for 7 days. On day 7, the NGFwas removed from half the cultures and replaced withAs the neurons required NGF to extend into theanti-NGF (200 ng/mL affinity-purified IgG); the re-distal compartments, we then examined whether amaining cultures were maintained with their respectiveconcentrations of NGF. Growth measurements were then

Table 2 Neurite Extension from Neurons Cultured recorded for 5 more days. There is no further neuritewith Neurotrophins in the Center Compartment extension in the anti-NGF treated cultures; the open

squares are the anti-NGF treated cultures that were ini-Centre Side Compartment (100 ng/mL)tially grown with 10 ng/mL NGF, and the open circlesCompartmentrepresent the anti-NGF treated cultures originally ex-(10 ng/mL) NT3 BDNF NGFposed to 100 ng/mL NGF. Sister cultures maintained withNGF for the same 5 days show the normal rate of neuriteControl 0 0 / (969 { 210 mm)

NT3 0 0 / (701 { 139 mm) extension (solid squares and circles) . Each time pointrepresents the mean { S.E.M. for five cultures.BDNF 0 0 / (725 { 28 mm)

NGF 0 0 / (537 { 29 mm)

The addition of neurotrophins (NGF, BDNF, and NT3) tothe neuronal cell body compartment did not result in increased continued supply of NGF was necessary for furthergrowth into distal compartments, nor did it induce any neurite extension. In these experiments, the adult culturesgrowth into distal compartments containing neurotrophins other

were allowed to grow neurites into the distal com-than NGF. Control Å no neurotrophin. Extension was measuredpartments containing the various concentrations ofas described in Methods on day 7 in vitro and represents the

mean { S.E.M. of four to six compartments. NGF. At 7 days in vitro, the medium containing



NGF was removed and replaced with medium con- point, neurites were already established in the distalcompartments containing NGF [Fig. 6(a)] . Thesetaining anti-NGF (200 ng/mL affinity-purified

IgGs), and the cultures were assessed for neurite results indicating a requirement for NGF for theregeneration are different from the in vivo situation,growth or regression over the next 5 days. The effect

of the NGF removal ( i.e., the anti-NGF) was to where NGF has been shown to have little influence(Diamond et al., 1992b).block any further neurite growth [Fig. 5(b)] . There

was a small amount of neurite regression, primarilyof very fine processes that had extended out from Immunocytochemical Assessmentneurite bundles [Fig. 2(d)] . In contrast, the sister of the Compartmented Culturescultures which were maintained in NGF displayedcontinual outgrowth over the same period of time Immunocytochemistry using antibodies directed

against p75 and TrkA, -B, and -C was performed[Fig. 5(b)] , thus demonstrating the requirement ofa continual source of NGF for neuritic growth. to provide some assessment of the neuronal compo-

sition of the cultures. As expected, the staining forTrkA-IR and p75-IR was found in a majorityEffects of NGF on Neurite(ú50%) of the neurons in the cultures; fewer ofRegeneration Following Axotomy inthe cultured neurons displayed TrkC-IR or TrkB-Compartmented CulturesIR. Figure 7 presents representative photomicro-graphs of the respective immunoreactive neurons inOne of our interests in pursuing these experiments

was to determine whether this in vitro system would the center compartments paired with the corre-sponding distal neurite compartments. The neuritesprovide an adequate model with which to investi-

gate the differential role of NGF observed in the in the distal compartments were positively stainedfor p75 and TrkA (Fig. 7); no staining was observedregulation of collateral sprouting and regeneration

of adult sensory neurons in vivo. While neurite with either the TrkB or TrkC antibodies.growth from dissociated cells in culture is necessar-ily a regenerative response initially, it was possiblethat once the neurites extended into the distal NGF- DISCUSSIONcontaining compartments, the growth would stabi-lize and could be used to approximate the in vivo The aim of the present study was to investigate

the regulation of neurite growth from adult sensorymodel. One of the advantages of using the compart-mented cultures is that the distal growth can be neurons by the NTs NGF, BDNF, and NT3. Our

findings demonstrate that NGF was the only NTaxotomized in a relatively precise fashion, and theregrowth or regeneration of those neurites further able to elicit elongation of distal neurites in com-

partmented cultures of both neonatal and adult neu-examined. In these experiments, cultures of bothneonatal and adult DRG neurons were plated in the rons. There was no neuritic elongation into compart-

ments containing either BDNF or NT3. NGF pro-center compartments in the absence of NGF andallowed to grow for 7 days into distal compartments duced maximal neurite growth when applied solely

to the distal neurite compartment, and subsequentcontaining NGF. On day 7, the neurites in the distalchambers were axotomized using a stream of dis- growth was halted when the NGF was effectively

removed or neutralized using anti-NGF. In addition,tilled water, and the regrowth of the neurites wasevaluated under differing conditions. Axotomized neurite growth was affected by NGF concentrations

with the highest concentration of NGF resulting inneurites rapidly regrew back into compartmentscontaining NGF [Fig. 6(a,b)] . However, in the a much greater density of neuritic growth, which

was most likely due to both increased numbers andpresence of anti-NGF in the distal compartments,there was no regrowth of the axotomized neurons, increased branching of neurites. Our results show

that NGF was the only NT of those tested whichthus indicating a requirement of NGF for this regen-eration [Fig. 6(c)] . Neither BDNF nor NT3 (100 produced distal neurite growth in compartmented

cultures, and thus provide a model for further studyng/mL) was able to support the regeneration of theaxotomized neurons [Fig. 6(d)] . There was little of NGF-dependent neuronal responses.

The neonatal neuron cultures were initially em-difference between the neonatal and adult neuronsin terms of the requirement of NGF for regenera- ployed as a way of assessing optimal conditions for

the adult neurons, and also because of the differ-tion; however, the axotomized neonatal neuronswere able to regenerate processes more quickly than ences in requirements of NGF for survival, it was

conceivable that neonatal neurons might exhibitthe adults, such that at the day-1 examination time



Figure 6 Nerve growth factor was required for regrowth following in vitro axotomy of bothneonatal and adult neurons. Neonatal (A) and adult (B) neurons were grown in compartmentedcultures for 7 days. On day 7, distal neurites in side compartments containing NGF (1, 10, or100 ng/mL) were axotomized with a jet of distilled water as described in Methods, and theNGF-containing medium was replenished. As can be seen in (A) and (B), neurons rapidlyregrew neurites into these compartments, provided NGF was present. (C) Anti-NGF blockedthe regrowth of neurites. As above, after day 7 of culture, the neurites of adult neurons wereaxotomized, but the NGF-containing medium was replaced with anti-NGF (200 ng/mL IgG).There was no significant regrowth into the compartments in the absence of NGF. (D) NeitherBDNF nor NT3 could substitute for NGF in promoting regrowth of neurites after axotomy. Asabove, after 7 days of culture and neurite extension into NGF-containing medium, axotomywas carried out; the NGF-containing medium was replaced with that containing either BDNF(plus anti-NGF) or NT3 (plus anti-NGF). There was no significant regrowth into either BDNF-or NT3-containing (100 ng/mL) compartments. Each point represents the mean { S.E.M. forsix to 10 cultures.

some differences compared to adult neurons. Find- response of the adult PNS to injury and NTs, wesubsequently focused primarily on the adult DRGings of the present study indicated that neonatal

neurons had the same characteristic growth re- cultures.sponses as adult neurons in compartmented cultures;growth rates, neurite elongation, and responses after Control of Axon Growth by NGFaxotomy of the neonatal neurons were comparableto those of the adult DRG neurons. The major differ- Adult DRG neurons do not require NGF or other

NTs for neurite elaboration in culture, although theence between these cultures was the fact that tosurvive in vitro, the neonatal neurons required ac- presence of these factor results in increased neuritic

growth (Lindsay, 1988; Jiang et al., 1995; Mohiud-cess to NGF. NGF could be removed from the cellbody compartment and the neurons would survive din et al., 1995; Smith and Skene, 1997).

In our cultures of adult DRG neurons, the neu-only if the distal neurite compartment was providedwith NGF. Because of our primary interest in the ronal cell bodies did not require exposure to NGF



Figure 7 Immunocytochemical staining for p75 and TrkA. Cultures were fixed and processedfor immunocytochemistry as described in Methods. (A,B) p75 immunoreactivity was visualizedusing an FITC-tagged secondary antibody and photographed with fluorescence optics. (A)Neurons in the cell body compartment; (B) neurites in the distal compartment. (C,D) TrkAimmunoreactivity was visualized using an HRP-tagged secondary antibody and diaminobenzi-dine, and photographed with transmitted light optics. (C) Neurons in the center compartment;(D) neurites in the distal compartment.

for either initial neurite growth or subsequent exten- growth into the distal chambers. The fact that theNGF in the center compartment also inhibited thesion into the distal NGF-containing compartments.

This is demonstrated by the findings that the neu- distal extension into NGF-containing compartmentsmay indicate that the neurons growing in the pres-rons were routinely plated in the absence of NGF,

and further, if the neurons were plated in the presence of NGF have exhausted their potential to growinto the side compartments. Alternatively, this mayence of anti-NGF, there was little obvious alteration

in the amount of proximal neurite growth in the be a result of altering the neuronal response to in-jury. A recent in vivo study suggests that provisioncenter compartment or subsequent distal extension

into NGF-containing compartments [Fig. 5(a)] . In- of exogenous NGF via intrathecal infusion impairsperipheral axonal regeneration, perhaps by reducingterestingly, if NGF was provided to the cell body

compartment, there was little growth of the adult the cell body response to injury (Gold, 1997). Inthat model, provision of NGF to the neurons resultsneurites into the distal NGF-containing compart-

ments (Fig. 1) . The ability of the neonatal neurons in a significant delay in both the onset and maximaldistance of axonal extension subsequent to a periph-to extend into the higher concentration of NGF may

reflect differences in the NGF sensitivity of devel- eral nerve crush. It is conceivable that in the presentexperiments, NGF may have a similar influence.oping neurons.

Nerve growth factor appeared to influence the Dorsal root ganglion neurons appear to have anintrinsic capability for two distinct forms of neuritegrowth of DRG neurites in a local manner, such that

exposure of the cell bodies to NGF, while eliciting growth in vitro: an initial relatively short arborizingform of growth followed by a switch to neuriteincreased growth locally, did not in itself result in



elongation (Smith and Skene, 1997). These authors tures were p75-IR and TrkA-IR. One possibility toaccount for the fact that neither BDNF nor NT3suggest that most DRG neurons are constitutively

competent to undergo arborizing growth, but very supported neurite extension into the distal compart-ments is that there were insufficient numbers oflimited longer neurite extension; this arborization

occurs in the absence of NGF or other factors, but neurons expressing TrkB or TrkC in our culturesfor BDNF or NT3 to elicit significant neurite elon-can be influenced by NGF or density of plating. The

subsequent switch to elongation requires a transcrip- gation. However, we carried out immunocytochem-istry using the commercially available antibodiestion-dependent transition, which may be triggered

by interruption of a signal from the periphery, as for TrkC and TrkB, and our preliminary observa-tions indicate that õ10–20% of the neurons werewould occur after nerve injury or explantation into

culture (Smith and Skene, 1997). Their results sup- TrkC or Trk-IR. Similar results were observed withDRG explants in culture (Edstrom et al., 1996).port the idea that many adult DRG neurons in vivo

are essentially maintained in an arborization-com- Explants of adult mouse lumbar DRGs withattached segment of nerve or root cultured in gelspetent state, which would be exemplified by the

ability of certain neurons to maintain or increase of extracellular matrix material displayed smallamounts of neurite growth in the absence of NTs.their terminal fields in response to alterations in

NGF [i.e., to undergo collateral sprouting in re- However, the addition of NGF substantially in-creased both the amount and length of such neurites.sponse to increased NGF levels (Diamond et al.,

1992a; Mearow et al., 1994)] . On the other hand, While a much smaller effect of NT3 was observed,BDNF had no apparent influence upon the neuriteSmith and Skene suggested that competence for

elongation is normally suppressed in intact neurons growth from such explants; the results were sug-gested to be attributable to the relatively low per-by signals conveyed from the peripheral targets, and

can be induced by disruption of ongoing signals, centage of TrkB-expressing neurons in the lumbarDRGs (Edstrom et al., 1996).such as occurs with nerve injury. The nature of the

signal(s) is unclear, but is not likely to be simply While we have evidence that the cultures do con-tain BDNF- and NT3-responsive neurons, and thata loss of transport of NGF, for example, since prior

studies have demonstrated that neither NGF nor exposure of the cell bodies and proximal axons tothese NTs results in enhanced growth (data notanti-NGF treatment in vivo affects the regeneration

of damaged adult DRG neurons (Diamond et al., shown), it may be that the neurons do not extendsufficiently long neurites to reach the distal com-1992b; Mearow et al., 1994; Mearow and Kril,

1995). Our results (discussed above) also point to partments, or that the NTs must act on cell bodyreceptors to get a response. For example, sympa-the possibility that exposure of the neuronal cell

bodies to NGF suppresses the elongation compe- thetic neurons are capable of different responsesdepending upon whether the cell bodies, distal neu-tence and places the neurons in the arborization

state. rites, or both are exposed to trophic factors (Tomaet al., 1997). Miller et al. (1994) showed that exo-genous NGF applied to the terminal fields canBDNF and NT3 Do Not Support Distalupregulate TrkA or p75 mRNAs independently andNeurite Extensionalso promote differential expression of the receptorson the terminals versus cell bodies such that thereOur results demonstrate that neither BDNF nor NT3

was effective in promoting distal neurite extension. is an increase in the ratio of p75:Trk on the terminalneurites.While it is clear that the DRGs are composed of a

heterogeneous population of neurons, including We were also interested in determining whetherBDNF or NT3 could affect the response of the neu-those that express receptors for NGF, BDNF, and

NT3, only NGF elicited distal neurite growth or rons to NGF, and thus carried out experiments inwhich neurons plated in the absence of the NTselongation, presumably from the TrkA-expressing

neurons. In the adult DRGs, TrkA-expressing neu- were provided with distal compartments containingBDNF plus NGF, or NT3 plus NGF. The biologicalrons comprise approximately 40–50% of the total,

while TrkB and TrkC-expressing cells account for activities of the NTs are mediated by binding totheir cognate high-affinity Trk receptors: NGF toanywhere from 5 to 30 and 10 to 20%, depending

upon the spinal level (McMahon et al., 1994; TrkA; BDNF/NT4 to TrkB; NT3 to TrkC and, athigher concentrations, also to TrkA (reviewed inWright and Snider, 1995; Molliver et al., 1995). In

line with these studies, our observations indicated Barbacid, 1994). All the NTs also bind with loweraffinity to the p75 receptor, and while the functionthat a majority (ú50%) of the neurons in the cul-



of p75 is less clear, proposed roles include enhanc- Differences between In Vivo andIn Vitro Growth Responsesing the binding of the NTs to the Trks and modulat-

ing Trk signaling (Barker and Shooter, 1994; Hant-As discussed above, adult sensory neurons in vitrozopoulos et al., 1994; reviewed in Chao, 1994, andhave been shown to exhibit two forms of axonalCarter and Lewin, 1997). p75 has been postulatedgrowth that in some respects could be consideredto be involved in NGF-mediated signaling by beingsimilar to in vivo growth responses. Thus, the arbo-able increase the binding of NGF to TrkA by in-rizing form of growth, which is suggested as thecreasing the local availability of NGF (Barker anddefault state of many adult DRG neurons (SmithShooter, 1994). Using excess BDNF to block bind-and Skene, 1997), may be similar to the in vivoing of NGF to p75 (since BDNF will compete withsprouting of the NGF-responsive population of no-NGF for p75, but not TrkA), the cotreatment ofciceptive neurons, and elongation may correspondPC12 cells with BDNF and NGF resulted in a de-to regeneration.creased activation of TrkA. This diminished activa-

Collateral sprouting of intact adult DRG neuronstion was indicated by crosslinking experiments torequires NGF; the only DRG neurons that have beenbe due to a reduction in binding of NGF to TrkAshown to sprout in the adult mammal have been the(Barker and Shooter, 1994). In our experiments,nociceptive, NGF-responsive neurons (Diamond etthe combination of BDNF with NGF resulted inal., 1987, 1992a). Anti-NGF treatment halts thethe inhibition of neurite extension, and in similarsprouting, while systemic NGF results in a hyperin-experiments using a monoclonal antibody againstnervation of the terminal fields (Diamond et al.,p75, MC192, preliminary results have also shown1992a,b) . In contrast, there is no observable sprout-a decrease in neurite extension (Kimpinksi anding from non-nociceptive afferents—that is, theMearow, in preparation). These results are consis-non-NGF responsive neurons, including those thattent with a significant role of the p75 receptor inexpress TrkB and TrkC receptors (Diamond et al.,

the regulation of this neurite growth response by1987, 1992a,b; Doucette and Diamond, 1988).

NGF. The ratio of p75:Trk receptor has been postu-However, corresponding in vivo experiments testing

lated as being an important factor in the responsethe efficacy of BDNF or NT3 on influencing either

of cells to NGF, and by altering the ability of NGF sprouting or regeneration have been difficult to per-to bind to p75, one is effectively altering this ratio, form because of the lack of sufficient quantities ofand thus the response of the neurons to NGF (e.g., either blocking antisera or the NTs themselves.Benedetti et al., 1993; Miller et al., 1994). A similar sprouting response can be approxi-

We also examined whether NT3 would have sim- mated in the compartmented culture system once theilar effects to BDNF or would potentiate the re- neurites have extended into the distal compartmentssponse of NGF, perhaps through its actions on (e.g., Campenot, 1987). It is possible that some ofTrkA. In our initial experiments, 1 or 10 ng/mL of the mechanisms through which NGF elicits growthNT3 was employed to more specifically activate of distal neurites in compartmented cultures areTrkC receptors; at higher concentrations NT3 has analogous to those exerted by the neurotrophin dur-been shown to also activate the TrkA receptors in ing collateral sprouting in vivo (but see below).sympathetic neurons (e.g., Belliveau et al., 1997), NGF is supplied only in the distal compartments inalthough it does not do the same in PC12 cells (Ip our cultures, a manner which mimics the in vivoet al., 1993). However, distal neurite extension did situation to a certain extent, and the action of NGFnot occur in any concentration of NT3. Our results in both cases primarily influences the growth re-[Fig. 4(B)] demonstrate that the combination of sponse, as opposed to a survival effect. Each re-NT3 plus NGF did not effect an increased amount sponse is reliant on NGF and is halted or abolishedof neurite extension, nor were there any noticeable by the removal of the neurotrophin (Diamond et al.,alterations in neuritic density. While NT3 can acti- 1992a; Mearow and Kril, 1995), indicating thatvate TrkA, it also interacts more efficiently with both collateral sprouting and growth of distal neu-p75 than NGF (Rodriguez-Tebar et al., 1992) and rites in the compartmented cultures require a con-could thus interfere with the binding of NGF in a stant supply of NGF.similar fashion as BDNF. To account for the lack In contrast, the regeneration of peripheral sen-of effect of the combination of NT3 plus NGF, it sory afferents in vivo occurs in the absence of NGFis conceivable that the effects of NT3 on TrkA are (Rich et al., 1984; Diamond et al., 1992b), althoughoffset by potential inhibitory influences via binding NGF has been shown to locally promote growth of

the central processes of DRG neurons (Oudega andto p75.



Hagg, 1996; Tuszinsky et al., 1994). In other exper- growth. BDNF and NT3 did not elicit such neuritegrowth, nor did they appear to be able to positivelyiments, however, intrathecal infusion of NGF at the

time of peripheral nerve injury has been demon- influence NGF-dependent growth. Use of the com-partmented system affords a unique model to studystrated to delay the onset and rate of regeneration

(Gold, 1997). aspects of the regulation of adult neurite growththat are often difficult to pursue in vivo, as well asThe in vivo model predicts that regeneration is

NGF independent and that it continues to be inde- providing a paradigm for specific investigation ofnociceptive afferents.pendent of NGF even when regenerating within the

target tissue (Diamond et al., 1992b). In the presentexperiments, regeneration in vitro seems to have This work was supported by grants from the Canadianboth NGF-independent and NGF-dependent compo- NeuroScience Network (KM and RBC) and the Naturalnents. Thus, the initial growth of proximal neurites Sciences and Engineering Research Council of Canadain the adult cultures (i.e., in the center compart- (KM). RBC is a Heritage Medical Scientist of the Albertaments) which is necessarily a regenerative response Heritage Foundation for Medical Research.occurred in the absence of any added neurotrophicfactors and was unaffected by treatment with anti-NGF. The subsequent and significant elongation and

REFERENCESincreases in the complexity of terminal branchingpatterns, however, required the presence of NGF.The regrowth after in vitro axotomy also required ACHESON, A., IP, N. Y., SQUINTO, S. P., and LINDSAY,

R. M. (1993). BDNF antisense oligonucleotides elicitthe presence of NGF, a result which is at varianceselective neuronal death in cultures of adult rat sensorywith the predictions of the in vivo model. The rea-neurons: evidence that a neurotrophin autocrine loopsons for the difference between the in vivo and insustains neuronal survival. Nature 374:450–453.vitro results are not clear, but could involve differ-

AVERILL, S., MCMAHON, S. B., CLARY, D. O., REI-ential interactions with the growth substratum andCHARDT, L. F., and PRIESTLY, J. V. (1995). Immunocy-NGF or other NTs. In addition to its trophic effects,tochemical localization of trkA receptors in chemically

NGF has long been suggested to exhibit tropic in- defined subgroups of adult sensory neurons. Eur. J.fluences on growing neurites, particularly with re- Neurosci. 7:1484–1494.spect to extension of growth cones, for example BARBACID, M. (1994). The trk family of neurotrophin(Campenot, 1995). Our observations in vitro sug- receptors. J. Neurobiol. 25:1386–1403.gest that this may be a possibility. Some limited BARKER, P. A., and SHOOTER, E. M. (1994). Disruption

of NGF binding to the low affinity neurotrophin recep-regrowth was observed in the absence of NGF in thetor p75LNTR reduces NGF binding to TrkA on PC12culture medium, perhaps owing to the interactioncells. Neuron 13:203–215.of neurites with residual substratum-bound NGF.

BELLIVEAU, D., KRIVKO, I., KOHN, J., LACHANCE, C., POZ-However, when anti-NGF was added to the mediumNIAK, C., RUSAKOV, D., KAPLAN, D., and MILLER, F. D.in these distal compartments, no regeneration of(1997). NGF and NT3 both activate TrkA on sympa-neurites was observed. In in vivo experiments, whilethetic neurons but differentially regulate survival and

neither NGF nor anti-NGF affects the regeneration neuritogenesis. J. Cell Biol. 136:375–388.of the sensory fibers (Diamond et al., 1992b), the BENEDETTI, M., LEVI, A., and CHAO, M. V. (1993). Dif-regrowing fibers were able to grow along a favor- ferential expression of nerve growth factor receptorsable substratum, i.e., the degenerating nerve sheaths leads to altered binding affinity and neurotrophin re-rich in various extracellular matrix molecules that sponsiveness. Proc. Natl. Acad. Sci. 90:7859–7863.

CAMPENOT, R. B. (1977). Local control of neurite devel-can influence nerve growth (cf., Reichardt, 1993;opment by nerve growth factor. Proc. Natl. Acad. Sci.Gavazzi and Cowen, 1996). Furthermore, this re-74:4516–4519.growth could very well have been influenced by

CAMPENOT, R. B. (1982). Development of sympatheticfactors other than NGF, such as the insulinlikeneurons in compartmentalized cultures: I. Local controlgrowth factors (Glazner et al., 1993; Zhuang et al.,of neurite growth by nerve growth factor. Dev. Biol.1996) or fibroblast growth factors (Danielson et al.,93:1–12.

1988; Eckenstein et al., 1991). CAMPENOT, R. B. (1987). Local promotion of neuriteIn conclusion, we have shown that NGF alone sprouting in cultured sympathetic neurons by nerve

was able to sustain growth or elongation of distal growth factor. Dev. Brain Res. 37:293–301.neurites in compartmented cultures of both adult CAMPENOT, R. B. (1992). Compartmented culture analy-and neonatal DRG neurons. The neurons exhibited sis of nerve growth. In: Cell–Cell Interactions: A Prac-

tical Approach. B. Stevenson, D. Paul, and W. Gallin,both NGF-independent and -dependent forms of



Eds. Oxford University Press, New York, pp. 275– mouse dorsal root ganglia in vitro. Neurosci. Lett.121:125–128.298.

IP, N. Y., STITT, T. N., TAPLEY, P., KLEIN, R., GLASS, J.,CAMPENOT, R. B. (1995). NGF and the local control ofFANDL, J., GREENE, L. A., BARBACID, M., and YANCO-nerve terminal growth. J. Neurobiol. 25:599–611.POULOS, G. (1993). Similarities and differences in theCARTER, B. D., and LEWIN, G. R. (1997). Neurotrophinsway neurotrophins interact with the Trk receptors inlive or let die: does p75NTR decide? Neuron 18:187–neuronal and non-neuronal cells. Neuron 10:137–149.190.

JIANG, Z.-G., SMITH, R. A., and NEILSON, M. (1995). TheCHAO, M. V. (1994). The p75 neurotrophin receptor. J.effects of NGF on neurite outgrowth from culturedNeurobiol. 25:1373–1385.adult and aged mouse sensory neurons. Dev. Brain Res.COUGHLIN, M. D., and COLLINS, M. B. (1985). Nerve85:212–219.growth factor independent development of embryonic

KALCHEIM, C. (1996). The role of neurotrophins in themouse sympathetic neurons in dissociated cell culture.development of neural-crest cells that become sensoryDev. Biol. 110:392–401.ganglia. Phil. Trans. R. Soc. Lond. B. 351:375–381.DANIELSON, N., PETTMAN, B., VAHLSING, H. L., MAN-

LINDSAY, R. M. (1988). Nerve growth factors (NGF,THORPE, M., and VARON, S. (1988). Fibroblast growth

BDNF) enhance axonal regeneration but are not re-factor effects on peripheral nerve regeneration in a sili-quired for survival of adult sensory neurons. J. Neu-cone chamber model. J. Neurosci. Res. 20:320–330.rosci. 8:2394–2405.

DAVIES, A. M. (1996). The neurotrophic hypothesis:LINDSAY, R. M. (1992). The role of neurotrophic factors

where does it stand? Phil. Trans. R. Soc. Lond. B.in the functional maintenance of mature sensory neu-

351:389–394.rons. In: Sensory Neurons: Diversity, Development and

DIAMOND, J., COUGHLIN, M., and HOLMES, M. (1992a). Plasticity. S. A. Scott, Ed. Oxford University Press,Endogenous NGF and nerve impulses regulate the col- New York, pp. 404–420.lateral sprouting of sensory axons in the skin of the MCMAHON, S. B., ARMANINI, M. P., LING, L. H., andadult rat. J. Neurosci. 12:1454–1466. PHILLIPS, H. S. (1994). Expression and coexpression

DIAMOND, J., HOLMES, M., FOERSTER, A., and COUGHLIN, of Trk receptors in subpopulations of adult primaryM. (1992b). Sensory nerves in adult rats regenerate sensory neurons projecting to identified targets. Neuronand restore sensory function to the skin independently 12:1161–1171.of endogenous NGF. J. Neurosci. 12:1467–1476. MEAROW, K. M., KRIL, Y., GLOSTER, A. S., and DIA-

ECKENSTEIN, F. P., SHIPLEY, G. D., and NISHI, R. (1991). MOND, J. (1994). Expression of NGF receptor andAcidic and basic fibroblast growth factors in the ner- GAP-43 mRNAs in DRG neurons during collateralvous system: distribution and differential alteration of sprouting or regeneration of dorsal cutaneous nerves.levels after injury of central versus peripheral nerve. J. Neurobiol. 25:127–142.J. Neurosci. 11:412–419. MEAROW, K. M., and KRIL, Y. (1995). Anti-NGF treat-

EDSTROM, A., EKSTROM, P. A. R., and TONGE, D. (1996). ment blocks the upregulation of NGF receptor mRNAAxonal outgrowth and neuronal apoptosis in cultured expression associated with collateral sprouting of ratadult mouse dorsal root ganglion preparations: effects dorsal root ganglion axons. Neurosci. Lett. 184:55–58.of neurotrophins, of inhibition of neurotrophin actions MILLER, F. D., SPEELMAN, A., MATHEW, T. C., FABIAN,and of prior axotomy. Neuroscience 75:1165–1174. J., CHANG, E., POZNIAK, C., and TOMA, J. G. (1994).

GAVAZZI, I., and COWEN, T. (1996). Can the neurotrophic Nerve growth factor derived from terminals selectivelyhypothesis explain degeneration and loss of plasticity increases the ratio of p75 to trkA NGF receptors onin mature and ageing autonomic neurons? J. Autonom. mature sympathetic neurons. Dev. Biol. 161:206–217.Nerv. Sys. 58:1–10. MOHIUDDIN, L., FERNANDEZ, K., TOMLINSON, D. R., and

GLAZNER, G., LUPIEN, S., MILLER, J. A., and ISHII, D. N. FERNYHOUGH, P. (1995). Nerve growth factor and neu-(1993). Insulin-like growth factor-II increases the rate rotrophin-3 enhance neurite outgrowth and upregulateof sciatic nerve regeneration in rats. Neuroscience the levels of messenger RNA for growth associated54:791–797. protein GAP-43 and Ta1 a-tubulin in cultured adult

GOLD, B. (1997). Axonal regeneration of sensory nerves rat sensory neurons. Neurosci. Lett. 185:20–23.is delayed by continuous intrathecal infusion of nerve MOLLIVER, D. C., RADEKE, M. J., FEINSTEIN, S. C., andgrowth factor. Neuroscience 76:1153–1158. SNIDER, W. D. (1995). Presence of absence of TrkA

HANTZOPOULOS, P. A., SURI, C., GLASS, D. J., GOLDFARB, protein distinguishes subsets of small sensory neuronsM. P., and YANCOPOULOS, G. D. (1994). The low af- with unique cytochemical characteristics and dorsalfinity NGF receptor, p75, can collaborate with each horn projections. J. Comp. Neurol. 361:404–416.of the trks to potentiate functional responses to the MULDERRY, P. K. (1994). Neuropeptide expression byneurotrophins. Neuron 13:187–201. ewborn and adult rat sensory neurons in culture: effects

of NGF and other neurotrophic factors. NeuroscienceHORIE, H., BANDO, Y., CHI, H., and TAKENAKA, T.59:673–688.(1991). NGF enhances neurite regeneration from

nerve-transected terminals of yound adult and aged OCKEL, M., VON SHACK, D., SCHROPEL, A., DECHANT,



G., LEWIN, G., and BARDE, Y.-A. (1996). Roles of NT3 TOMA, J. G., ROGERS, D., SENGER, D. L., CAMPENOT,R. B., and MILLER, F. D. (1997). Spatial regulation ofduring early development of the peripheral nervous

system. Phil. Trans. R. Soc. Lond. B. 351:383–387. neuronal gene expression in response to nerve growthfactor. Dev. Biol. 184:1–9.OUDEGA, M., and HAGG, T. (1996). Nerve growth factor

promotes regeneration of sensory axons into adult rat VAN DER ZEE, C. E. E. M., RASHID, K., LE, K., MOORE,K., STANISZ, J., DIAMOND, J., RACINE, R. J., and FAHN-spinal cord. Exp. Neurol. 140:218–229.

REICHARDT, L. F. (1992). Adhesive interactions that reg- ESTOCK, M. (1995). Intraventricular administration ofantibodies to nerve growth factor retards kindling andulate neuronal behaviour. In: Sensory Neurons: Diver-

sity, Development and Plasticity. S. A. Scott, Ed. Ox- blocks mossy fibre sprouting in adult rats. J. Neurosci.15:5316–5323.ford University Press, New York, pp. 215–241.

RICH, K. M., YIP, K. K., OSBORNE, P. A., SCHMIDT, R. E., WRIGHT, D. E., and SNIDER, W. D. (1995). Neurotrophinreceptor expression defines distinct subpopulations ofand JOHNSON, E. M. (1984). Role of nerve growth

factor in the adult dorsal root ganglia neuron and its neurons in rat dorsal root ganglion. J. Comp. Neurol.351:329–338.response to injury. J. Comp. Neurol. 230:110–118.

RODRIGUEZ-TEBAR, A., DECHANT, G., GOTZ, R., and YASUDA, T., SOBUE, G., ITO, T., MITSUMA, T., and TAKA-

HASHI, A. (1990). Nerve growth factor enhances neu-BARDE, Y. A. (1992). Binding of neurotrophin-3 to itsneuronal receptors and interactions with nerve growth rite arborization of adult sensory neurons; a study in

single-cell culture. Brain Res. 524:54–63.factor and brain-derived neurotrophic factor. EMBO J.11:917–922. ZHUANG, H. X., SYNDER, C. K., PU, S. F., and ISHII, D. N.

(1996). Insulin-like growth factors rescue or arrest dia-SMITH, D. S., and SKENE, J. H. P. (1997). A transcrip-tion-dependent switch controls the competence of adult betic neuropathy—effects on hyperalgesia and im-

paired nerve regeneration in rats. Exp. Neurol.neurons for distinct modes of axon growth. J. Neurosci.17:646–658. 140:198–205.


effects of the neurotrophins nerve growth factor, neurotrophin-3, and brain-derived neurotrophic...

Documents