Neuron.

Vol. 8, 983-993,

May,

1992, Copyright

0 1992 by Cell Press

The Neurotrophins BDNF, NT-3, and NGF Display Distinct Patterns of Retrograde Axonal Transport in Peripheral and Central Neurons Peter S. Czeslaw Patricia Ronald Regeneron 777 Old Tarrytown,

DiStefano, Beth Friedman, Radziejewski, Charles Alexander, Boland, Christine M. Schick, M. Lindsay, and Stanley J. Wiegand Pharmaceuticals, Inc. Saw Mill River Road New York 10591-6707

Summary The pattern of retrograde axonal transport of the targetderived neurotrophic molecule, nerve growth factor (NGF), correlates with its trophic actions in adult neurons. We have determined that the NCF-related neurotrophins, brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3), arealso retrogradely transported by distinct populations of peripheral and central nervous system neurons in the adult. All three 1251-labeled neurotrophins are retrogradely transported to sites previously shown to contain neurotrophin-responsive neurons as assessed in vitro, such as dorsal root ganglion and basal forebrain neurons. The patterns of transport also indicate the existence of neuronal populations that selectively transport NT-3 and/or BDNF, but not NCF, such as spinal cord motor neurons, neurons in the entorhinal cortex, thalamus, and neurons within the hippocampus itself. Our observations suggest that neurotrophins are transported by overlapping as well as distinct populations of neurons when injected into a given target field. Retrograde transport may thus be predictive of neuronal types selectively responsive to either BDNF or NT-3 in the adult, as first demonstrated for NGF. Introduction Nerve growth factor (NGF) represents the prototype of target-derived, retrogradely transported trophic molecules. Typically, the trophic activities of such molecules are limited to subsets of peripheral and central nervous system neurons (Snider and Johnson, 1989). Trophic factors initiate their effects by binding to high affinity neuronal receptors. In the case of NGF this is followed by uptake and retrograde transport to the cell body. Although the role of retrograde transport in signal transduction is unclear, the ability of neurons to respond to NGF correlates well with transport. The retrograde transport of exogenous, 1251-labeled NGF has been well documented in sympathetic and neural crest-derived sensory neurons (Hendry et al., 1974; Stockel et al., 1975; Stockel and Thoenen, 1975; Johnson et al., 1978). Similarly, retrograde transport of endogenous NGF has been demonstrated (Korsching and Thoenen, 1983; Palmatier et al., 1984), indicating that neurons do transport NGF under physiological conditions. While these studies have shown that retrograde transport is a hallmark of peripheral

neurons responsive to NGF, the demonstration of selective retrograde transport of NGF within the central nervous system (CNS) has been pivotal in the discovery that central forebrain cholinergic neurons respond to this factor (Schwab et al., 1979; Seiler and Schwab, 1984). Thus, retrograde transport serves both as a means of confirming trophic responses of various neurons and as a powerful tool to reveal neuronal cell types not previously known to respond to a given trophic factor. The fact that NGF maintains the survival of only certain neuronal subtypes suggests that there exist additional molecules with distinct specificities which fulfill the roleof target-derived, retrogradelytransported neurotrophic factors. Indeed, brain-derived neurotrophic factor (BDNF; Barde et al., 1982; Leibrock et al., 1989), neurotrophin-3 (NT-3; Hohn et al., 1990; Maisonpierre et al., IVVOa; Ernfors et al., IVVOa), and neurotrophin-4 (NT-4; Hallbook et al., 1991) have recently been identified as members of an NGF-related family of gene products now termed neurotrophins. From in vitro studies it is clear that all members of the neurotrophin family promote the survival and neurite outgrowth of a variable percentage of neural crestderived sensory neurons in the dorsal root ganglion (DRG), whereas BDNF and NT-3, but not NGF, support placodally derived sensory neurons of the nodose ganglion (Lindsay et al., 1985; Leibrock et al., 1989; Maisonpierre et al., IVVOa; Hohn et al., 1990). Cultured embryonic basal forebrain cholinergic neurons respond to both BDNF and NGF (Alderson et al,, IVVO), whereas cultured retinal ganglion cells (Johnson et al., 1986) and cultured dopaminergic neurons of the embryonic ventral mesencephalon (Hyman et al., 1991) appear to respond to BDNF, but not NGF. Thus, overlapping but distinct neuronal populations respond to the various neurotrophins in vitro. BDNF has been shown to prevent naturally occurring cell death of several sensory neuronal populations in chick embryos (Kalcheim et al., 1987; Hofer and Barde, 1988). In cultured adult sensory neurons, BDNF enhances neurite outgrowth, but is not required for survival (Lindsay, 1988). However, evidence that adult neurons respond to BDNF and NT-3 in vivo is lacking. We report here patterns of retrogradetransport of BDNF and NT-3 in adult rat peripheral and central neurons. These patterns are distinct from, but partially overlap, the pattern of NGF transport. Results Biological Activity of ‘251~Labeled Neurotrophins All three neurotrophins were labeled to high specific activities, and the biological activities of the 1251-labeled neurotrophins were found to be VO%-99% of their unlabeled counterparts, as assessed in the chicken DRG neurite outgrowth assay (Figure 1). The

Nl3lVXl 984

ifi :

6

O.B- A

5-

0.6 -

‘25 I-BDNF

w

q

WWT

LEFT

4-

0.0.

2-

PBS

BDNF

NT-3

NGF

NT-3

NGF

NT-3

NGF

1251-NGF 0,

7. .Ol

. . . . .. .l

.

. ......,

. . . ..“l 1

5 - -

lz51-NT-3

nglml

CONCENTRATION,

: H

B.

10

1251.BDNF

PBS

BDNF

42.4

C.

‘251.NGF

2.0 1.6 1.2

0.8 .Ol

.l

1

10

PBS

BDNF

Figure 2. Quantitation and Pharmacological ‘WLabeled BDNF, NT-3, and NGF Retrograde sory Neurons of the DRG

6 5

-----t-

NT-3 1251.NT-3

z 4

01

.l CONCENTRATION,

Figure

0.4

0.0

nglml

CONCENTRATION,

1. Bioassay

of ‘2il-Labeled

10

1

Characteriratlon of Transport In Sen-

Injections of “Wabeled BDNF (0.52 pmol), NT-3 (0.61 pmol), and NCF (0.99 pmol), premixed with either phosphate-buffered saline or a IOO-fold excess of unlabeled factors (denoted on x-axes), were performed as described in Experimental Procedures. All survival times for these experiments were 18 hr. “jl-BDNF (A); ‘?-NT-3 (B); ‘WNGF (C). Counts per minute obtained for L4 and L5 ganglia were added and expressed as femtomoles of Yneurotrophin transported per L4 plus L5 DRCs. Values are mean + SEM for 3-8 animals. Asterisks indicate p < 0.01 versus phosphate-buffered saline coinjection; A dagger indicates p < 0.01 versus NGF or BDNF coinjection (ANOVA; Newman-Keuls’ multiple comparison analysis).

rig/ml Neurotrophins

Biological activities of ‘Wabeled NCF (top), BDNF (middle), and NT-3 (bottom) and their unlabeled counterparts were determined by addition to chicken DRC explants at 5.0, 1.0, 0.2, and 0.04 ngiml for 18 hr. Mean bioassay scores (y-axis) were determined by blinded observers. ECso values were not different between labeled and unlabeled neurotrophins (p > 0.05). Values represent mean + SD of 4-5 independent determinations.

ECso values for the respective labeled and nonlabeled neurotrophins were not significantly different. These results provide an independent measure of the biological activities of the radiolabeled neurotrophins used for transport studies and demonstrate that the radiolabeling procedure did not significantly disrupt the integrity of the neurotrophin molecules.

Neurotrophin Transport in DRG Neurons Since BDNF, NT-3, and NGF all enhance, to different degrees, the in vitro survival and neurite outgrowth of DRG neurons, we examined their retrograde transport patterns in adult rat spinal sensory neurons following injection into the sciatic nerve. This system was chosen for comparative studies, as NGF is retrogradely transported by neurons of the DRG (Sttickel and Thoenen, 1975; Yip and Johnson, 1983; Richardson and Riopelle, 1984). Additionally, retrogradetransport was examined in motor neurons, whose axons also course through the sciatic nerve. All three 1251-labeled neurotrophins were specifically retrogradely transported from the crush site to the ipsilateral fourth and fifth lumbar (L4 and L5) DRGs (Figure 2). This specificity was assessed by the absence of radio-

Retrograde 985

Axonal

Transport

of Neurotrophms

activity in the contralateral (left) L4 and L5 ganglia, as well as the ability of a IOO-fold excess of homologous, unlabeled neurotrophin to block transport. In heterologous neurotrophin competition studies, BDNF and NT-3 each significantly inhibited the transport of the other (Figures 2A and 2B), but not that of NGF (Figure 2C). From Figure26 it is apparent that a portion of NT-3 transport is not blocked by BDNF or NGF. Coinjection of excess NGF significantly attenuated 1251-NT-3 transport (Figure 28). Although NGF did not significantly block the uptake and retrograde transport of lz51-BDNF, partial inhibition was apparent (Figure 2A). These results suggest that the neurotrophins are transported by DRG neurons via pharmacologically distinct mechanisms. The apparent amount of transport to DRG neurons varied among the three neurotrophins, with the greatest accumulation seen with V-NGF, followed by rLSI-NT-3 and 1251-BDNF. This may reflect differences in the relative numbers of sensory axons capable of transporting the neurotrophins or the avidity with which individual neurons transport these factors. Therefore, to assess the distribution of transported material, L4 and L5 ganglia of injected animals were processed for emulsion autoradiography and examined with dark-field optics. Transport of iodinated BDNF, NT-3, and NGF along the sciatic nerve produced neuronal labeling in the cell bodies of neurons in the DRG (Figures 3A-3C). Neurotrophin transport appeared to be receptor mediated, since unlabeled NT-3 completely blocked the transport of 1251-NT-3 (Figure 3D). Similarly, rZ51-BDNF transport was blocked by unlabeled BDNF (data not shown). The 1251-labeled neurotrophins were shown to label cells of various sizes within the sensory ganglia. Clearly, size-frequency analysis will be required to delineate further cell size distributions of the transported neurotrophins; it will be of interest to determine the extent to which subsets of retrogradely labeled DRG neurons correspond to those that encode distinct sensory modalities, such as nociception or proprioception. Dense fiber tract labeling was prominent with NGF transport compared with BDNF or NT-3 transport. The fact that tract labeling was more pronounced with NGF transport is most likely the result of a greater overall capacity of DRG neurons to transport NGF compared with BDNF or NT-3 (see Figure 2), thus making tract labeling more readily detectable. Neurotrophin Transport in Spinal Motor Neurons Duringearlydevelopment in thechicken and rat, NGF is retrogradely transported by motor neurons; however, NGF is not transported by adult motor neurons (Wayne and Heaton, 1988; Yan et al., 1988). We examined whether BDNF or NT-3 could be retrogradely transported by adult motor neurons when injected into the crushed sciatic nerve. Large neurons in the ventral spinal cord ipsilateral to the side of the injection showed retrograde labeling 18-24 hr after injection of ?-BDNF (Figure 4A). These neurons were

identified within the dorsolateral quadrant of the ventral horn at L4 and L5, precisely where motor neurons projecting through the sciatic nerve are located. These neurons have been shown to be positive for choline acetyltransferase (data not shown). Labeling of motor neurons was also observed following 1251-NT-3 injections in the crushed nerve (Figure 4C), but not with injections of 1251-NGF (Figure 4D). The retrograde transport of lZ51-BDNF in motor neurons appeared to be receptor mediated and not a result of diffusion, as coinjection with excess unlabeled BDNF blocked transport (Figure 48). In addition, no cells were labeled in the contralateral spinal cord (data not shown).Thus, neurotrophinstransportedwithinasingle peripheral nerve (sciatic) show distinct retrograde labeling patterns within the sensory and motor neuron populations whose axons project within that nerve. Transport of 1251-Labeled Neurotrophins in Sympathetic Neurons The ability of NGF to be taken up and retrogradely transported bysympatheticaxonsiswell documented (Stockel and Thoenen, 1975; Johnson et al., 1978). To determine whether other neurotrophins could be transported by sympathetic neurons, ‘*5l-labeled BDNF and NT-3 were injected into the anterior eye chamber. The iris, which is juxtaposed to the anterior eye chamber, is sympathetically innervated by the superior cervical ganglion (KG). While 1251-NGF accumulated to a substantial degree in the ipsilateral SCG (Figure 5A), lz51-BDNF (Figure 5B) and 1251-NT-3 (Figure 5C) retrogradely labeled approximately 5%-7% of the number of cells labeled by NGF (see Figure 5D), indicating that BDNF and NT-3 are transported by a very small population of SCG neurons. Retrograde Transport of 1251-Labeled Neurotrophins in the CNS To investigate whether distinct patterns of neurotrophin retrograde transport are evident in the adult CNS,1251-labeled neurotrophinswereinjectedintothe hippocampus. This site is of particular interest because of its rich endogenous supply of BDNF, NT-3, and NGF (Maisonpierre et al., 199Ob; Ernfors et al., 1990b). As observed in previous studies (Seiler and Schwab, 1984), retrograde transport of 1251-NGF from the dorsal hippocampus to the medial septal nucleus and the vertical limb of the diagonal band of Broca was robust (Figure 6C). 1251-BDNF and 12’l-NT-3 also retrogradely labeled cells within these nuclei (Figures 6A and 6B), but the labeling was not as intense as that with NGF. With all neurotrophins, little transport was seen to the contralateral medial septum and diagonal band. In addition to the anticipated labeling of cells in the medial septum and diagonal band nuclei, a number of other neuronal cell groups in the forebrain that retrogradely transported the radiolabeled neurotrophins were identified (Table 1). As observed in the

Neuron 986

Retrograde 987

Figure

Axonal

Transport

4. Retrograde

Axonal

of Neurotrophins

Transport

of WLabeled

Neurotrophins

in Spinal

Cord

Motor

Neurons

Dark-field photomicrographs of spinal cord sections from the same rats as those injected in Figure 3. lLSI-BDNF showed dense labeling in ventral cord motor neurons (A). Transport of ‘WBDNF was completely blocked by coinjection of a IOO-fold excess of unlabeled BDNF (6). Spinal motor neurons were typically lightly labeled in rats injected with ‘WNT-3 (C). Note the relative sparseness of silver grains in the neuropil adjacent to the labeled motor neuron pool in (A) and (C). In contrast to BDNF and NT-3, NCF showed no specific retrograde transport to ventral spinal cord motor neurons (D). Bar, 100 pm.

peripheral nervous system, the patternsof retrograde transport of the three neurotrophins were not always coincident in the CNS. For example, all three neurotrophins were retrogradely transported by neurons of the supramammillary nucleus of the hypothalamus (Figures 6G-61). However, BDNF, but not NGF, was transported by several additional neuronal populations afferent to the hippocampus, including the reuniens nucleus of the thalamus and the entorhinal cortex. In addition to these extrinsic cell groups, BDNFwaswidelytransportedwithin the hippocampal formation itself. The most extensive intrinsic labeling was found within the hilus of the dentate gyrus. Hilar

Figure

3. Localization

of Transported

Neurotrophins

in DRC

neurons were labeled bilaterally, especially caudal to the level of the injection (Figures6D-6F). Aggregations of silver grains were also consistently present over some neuronal perikarya in the pyramidal cell layer of Ammon’s horn and in thedorsal subiculum ipsilateral to the injection site. NT-3 was also retrogradely transported by most of the cell groups that transported BDNF, although the transport of NT-3 was generally less robust than that of BDNF (Table 1). Intrahippocampal transport of 1251-NGF was not observed. As a measureof the specificity of transport from the hippocampus, it was found that BDNF transport was blocked by coinjection with an excess of unlabeled

Neurons

Dark-field photomicrographs of retrogradely labeled neurons in the lumbar (LS) spinal ganglion. These cells differentially accumulated retrogradely labeled iodinated BDNF, NT-3, and NCF injected into the ipsilateral sciatic nerve (doses as in Figure 2). I>?-BDNF many DRC neurons of varying diameters (A). 7251-NGF labeled numerous, predominantly smaller DRG neurons (B). ‘*51-NT-3 was also transported by DRG cells (C); scattered large neurons in the DRC showed extremely dense labeling. Coinjectian of a IOO-fold excess ‘WNT-3 transport (D). Survival times were 18-24 hr. Bar, 100 pm. of unlabeled NT-3 completely blocked

Neuron 988

Ecf LABELED

Figure

5. Localization

of Retrogradely

Transported

Neurotrophins

rn Sympathetic

‘Wabeled NGF (A), BDNF (B), and NT-3 (C)were injected into the right anterior were immersion fixed, counted (D), and processed for emulsion autoradiography photomicrographs were taken from the right (ipsilateral) XC. Bar, 100 urn. animals.

BDNF, and transported pal afferents

1L51-cytochrome c was not by either intrinsic or extrinsic (data not shown).

retrogradely hippocam-

Discussion Receptor Mechanisms of Transport The present results demonstrate tive, 1251-labeled BDNF and NT-3 transported by an unexpectedly ronal cell types in the peripheral

that biologically accan be retrogradely wide variety of neuand central nervous

mNF NEUROTROPHIN

NT-3 INJECTED

Neurons

eye chamber, and animals were killed 16 hr later. SCCs as described in Experimental Procedures. Dark-field In (D), values are expressed as mean f SEM; n = 4-12

systems. Results from heterologous neurotrophin competition studies in the DRG suggest that common as well as distinct receptor mechanisms may be involved in neurotrophin transport. Recent reports have demonstrated that the proto-oncogenes trk (Klein et al., 1991; Hempstead et al., 1991), trkB (Squint0 et al., 1991; Soppet et al., 1991), and trkC (Lamballe et al., 1991) are distinct, biologically relevant receptors for NGF, BDNF, and NT-3, respectively, when expressed in fibroblasts or COS cells. While trk, trkB, and trkC appear to be neurotrophin specific in their pharma-

Retrograde 989

Figure

Axonal

Transport

6. Retrograde

Axonal

of Neurotrophins

Transport

of ‘2FI-Labeled

Neurotrophins

in the CNS

Dark-field photomicrographs of retrogradely labeled cells in the medial septum/diagonal band (MS/DE%, A-C), hippocampus (Hip, D-F), and supramammillary nucleus of the hypothalamus (SUM, G-l) following intrahippocampal injections of ‘*Y-labeled BDNF, NT-3, and NCF. Injected amounts ranged from 160 to 820 fmol of trophic factor, comprising a total of 0.4-0.8 PCi. Labeled cells were invariably identified in the medial septum/diagonal band following the injection of each neurotrophin (A-C). Cells and nerve fibers were most heavily labeled following injections of NGF (C). Labeling of nerve fibers was not apparent in the medial septum/diagonal band following injections of BDNF (A) or NT-3 (B). (D)-(F) illustrate the pattern of labeling in the hippocampal formation, approximately 1.5 mm caudal to the center of the injections. Numerous heavily labeled cells were identified in the dentate hiluslCA4 region (Hil) following an injection of BDNF (D) and to a lesser extent of NT-3 (E). Following injection of BDNF, but not NT-3, labeled cells were also present in the dorsal subiculum (Sub, arrows) and within the most medial part of the CA3 pyramidal cell layer (open arrow). Retrograde labeling of neurons within the hippocampal formation was never observed following injection of NCF (F). Many heavily labeled neurons were present in the supramammilary nucleus of animals injected with BDNF (C), NT-3 (H), and NGF (I). Bars: 200 pm, MS/D6 (A-C); 500 vrn, Hip (D-F); 100 pm, SUM (C-l).

NT-3 shows some degree of promiscuity in it is recognized by trk and trkB, as well as by trkC. Assuming that trk proteins mediate high affinity binding, uptake, and retrograde transport in vivo, this may explain why NT-3 partially blocks BDNF transport to the DRG and why NGF significantly attenuates the transport of NT-3. It will be of interest to determine whether cells transporting the individual neurotrophins contain mRNA and protein for their cognate trk

cology,

that

receptor molecules and whether a single neuron might contain more than one trk. In addition, the question arises as to what role the low affinity NGF receptor (LNGFR) plays in mediating transport and subsequent responsiveness of adult neurons to the various neurotrophins. It is known that BDNF, NT-3, NGF, and NT-4 bind to the LNGFR with comparable affinities (Rodriguez-T6bar et al., 1990; Squint0 et al., 1991; HallbiiGk et al., 1991) and that the LNGFR is itself

Neuron 990

Table 1. Retrograde Neuronal Labeling in the Rat Forebrain following lntrahippocampal Injection of ‘251-labeled Neurotrophic Factors Brain

Area

Basal forebrain Medial septum Diagonal band (v) Diagonal band (h) Basal nut. meynert Hippocampus HiluKA4 CA1 CA2 CA3 Dentate gyrus Subiculum Other Supramam. nut. N. reuniens thal. Entorhinal cortex

NCF

BDNF

NT-3

++++ ++++ ++ +

+++ +++ +

++ ++ -

-

++ -

-

++++ f + ++

ND -

ND +

ND -

++ -

+++ f +

++-f + -

*

Plus signs represent relative numbers of retrogradely labeled cells present in theareas indicated. (+ + + +) many labeled cells; (+) a few labeled cells; (+) areas where labeling was noted in some, but not all injected animals. A minus sign indicates that no labeled cells were observed. ND (not determined) indicates cell groups not amenable to evaluation, given their location in relation to the injection site.

retrogradely transported in peripheral and central NGF-responsive neurons (Johnson et al., 1987). The fact that NGF exerts partial inhibition of 1251-BDNF transport, for example, may represent a subtle interaction of neurotrophins with the LNGFR. Overlapping but Distinct Patterns of Neurotrophin Transport in Peripheral and Central Neurons The observation that BDNF and NT-3 (but not NGF) are retrogradely transported to the ventral spinal cord suggests that these neurotrophins may play a role in the biology of adult motor neurons. It is known that, early in development, motor neurons contain NCFbinding sites (Raivich et al., 1985, 1987), are immunopositive for LNGFR (Yan and Johnson, 1988), and can retrogradely transport 1251-NGF (Wayne and Heaton, 1988; Yan et al., 1988), although there is no known physiological effect of NCF on developing motor neurons (Oppenheim et al., 1982). In contrast, BDNF and NT-3 have recently been shown to stimulate choline acetyltransferase levels in cultures of purified rat embryo spinal neurons (Wong et al., submitted), although neither factor promotes survival of motor neurons. It is of importance to determine whether BDNF and/or NT-3 subserve a maintenance or rescue function during regeneration and degeneration of adult motor neurons. As an additional consideration, BDNF and NT-3 may be mimicking an as yet unidentified neurotrophin that utilizes the same high affinity binding and uptake sites. In contrast to transport patterns observed in spinal motor neurons, sympathetic neurons transport very little BDNF and NT-3, compared with NGF. Nonethe-

less, neurons are retrogradely labeled with these two factors. The fact that NT-3 was initially described as having slight survival effects on SCG explants (Hohn et al., 1990; Maisonpierre et al., 1990a) and that trkB mRNA has been localized in sympathetic neurons by in situ hybridization (Klein et al., 1989) suggests that a population of SCG neurons may be responsive to NT-3 and/or BDNF. Results from intrahippocampal injectionsof labeled neurotrophins show that, as in the periphery, the neurotrophins show overlapping (medial septumldiagonal band, supramammillary nucleus) yet distinct (hilarICA4 neurons, entorhinal cortex) patterns of neuronal transport within the CNS. The overlapping transport patterns of BDNF and NGF in medial septal neurons are consistent with findings showing survival effects of these two neurotrophins on cultured embryonic septal neurons (Alderson et al., 1990). As with transport studies in the DRG, it will be of interest to determine whether neurons within the medial septum that transport NGF and BDNF show distinct trkltrkB localization patterns and whether these cell types are neurochemically distinct (choline acetyltransferase, LNGFR colocalization). The question as to whether these or any cells within the nervous system express more than one of the trkfamily members must also be addressed. That transport processes may be taking place locally within the hippocampus suggests that BDNF and NT-3 may serve roles distinct from that of a classic targetderived trophic factor. Recent studies showing rapid changes in BDNF and NGF mRNA levels within the hippocampus following electrical or pharmacological stimulation (Gall and Isackson, 1989; Zafra et al., 1990; lsackson et al., 1991) suggest that neurotrophins may be involved in multiple local events, such as neuronal sprouting, synaptogenesis, long-term potentiation, and modulation of the neurochemical phenotype of adult neurons. If retrograde transport is requisite for the biological actions of the neurotrophins, then transport of ‘L51-labeled neurotrophins represents a powerful tool to trace specific neurotrophin-responsive neural systems, exposing populations of adult neurons that respond to neurotrophins which are not amenable to study using tissue culture techniques (Thoenen, 1991). Understanding the neuronal specificities of each neurotrophin will bring insight to the specific role of each member of the family in the processes of development, regeneration, and degeneration of the nervous system. This may provide clues for the therapeutic application of neurotrophic molecules. Experimental

Procedures

Purification of Neurotrophins NGF was purified from male mouse submaxillary gland\ a\ described (Bocchini and Angeletti, 1969). Recombinant human BDNF and NT-3 were prepared from the conditioned media ot Chinese hamster ovary (CHO) cells expressing the individual

Retrograde

Axonal

Transport

of Neurotrophins

991

proteins. The coding regions of the BDNF and NT-3 cDNAs (designated pCRhBDNFand pCBhNT-3, respectively)weresubcloned into the Xhol site of the pCDM8 expression vector as described (Maisonpierre et al., 1990a) and stably cotransfected into dhfrDC44 CHO cells using the ~410 mutant of the dihydrofolate reductase expression vector as described (Ciudad et al., 1988). Cells were initially selected in IO nM methotrexate; amplification was achieved at concentrations up to 1 uM methotrexate. Subclones of the BDNF- and NT-3-producing cell lines were maintained in supplemented Dulbecco’s modified Eagle’s medium containing 10 nM methotrexate for 6 days prior to harvesting for neurotrophin purification. Media from BDNF- or NT-5expressing CHO cells were passed over an S-Sepharose Fast Flow (Pharmacia) cation exchange column and washed with 0.01 M sodium phosphate, 0.15 M NaCl (pH 7.4), followed by 0.02 M bicine, 0.15 M NaCl (pH 8.5). Neurotrophins were eluted with a 0.15-0.60 M NaCl gradient in 0.02 M bicine buffer (pH 8.5). The concentrated effluent wasapplied toa column of Sephacryl S-100 HR (Pharmacia) and eluted in 0.04 M sodium phosphate, 0.04 M NaCl (pH 7.5). NT-3 was further purified by high-pressure liquid chromatography, usingaVydacC,columnwith aO%-66% acetonitrile gradient mobile phase in trifluoroacetic acid, and exchanged into 0.01 M sodium phosphate, 0.15 M NaCl (pH 7.4). The final purity was 95%; contaminants consisted of NT-3 molecules lacking 5 aa from the amino terminus. No high-pressure liquid chromatographystepwas used for BDNF purification.The final BDNF purity was 90%; contaminants consisted, in part, of BDNF molecules lacking 6 aa from the amino terminus. Purification yields of BDNF and NT-3 ranged from 38% to 60% of the starting conditioned media, which contained between 1 and 10 mg of neurotrophin per liter. Neurotrophin purity was determined by SDS-polyacrylamide gel electrophoresis followed by staining with Coomassie (brilliant) blue and by N-terminal sequencing. BDNF, NT-3, and NCF were tested for biological activity by determining their ability to stimulate survival and neurite outgrowth in embryonic day 8 chicken DRG explants as described (Maisonpierre et al,, 1990a). Briefly, ganglia were dissected and placed on 35 mm plastic culture dishes coated with a collagen matrix containing minimal essential medium buffered with NaHCO,. After allowing the collagen matrix to set, F-14 medium (containing 5% horse serum) was added to the ganglia along with various neurotrophins to be tested. Gangliawere scored for neurite outgrowth 24 hr later using a O-5+ scale, where 0 = no neurite formation and 5+ = maximal neurite outgrowth. Scores between 3 and 4, for example, were given a value of 3+, or 3.5. lodination of Neurotrophins BDNF, NT-3, NCF, and cytochrome c (Sigma) were iodinated by a modification of the lactoperoxidase method (Marcha!onis, 1969). Briefly, 1 mCi of Na’2il (New England Nuclear) was added with 1.2 Kg of lactoperoxidase (Sigma), 85 uM H20,, and 6-10 ug of neurotrophin or cytochrome c at pH 6.0 for 12 min. The reaction was stopped by addition of 0.1 M Nal, 0.1 M sodium phosphate, 1 .O M NaCl (pH 7.5). The reaction was diluted I:1 with 2% bovine serum albumin (Boehringer Mannheim) in phosphate-buffered saline. Solutions were dialyzed to eliminate free (unincorporated) l.‘il, Percent incorporation (ranging from 70% to 90%) was determined by thin layer chromatography. Specific activities were calculated based on a molecular weight of 26,000 for BDNF, NT-3, and NGF and 13,000 for cytochrome c. The specific activrties of BDNF, NT-3, NCF, and cytochrome c were 2095-4609 cpmi fmol, 4044-7283 cpmlfmol, 2817-3506 cpmlfmol, and 1382-2057 cpmlfmol, respectively. Labeled neurotrophins were assayed for broactivity using the chicken DRC explant assay (see above). Animal Treatments For sciatic nerve studies adult male Sprague-Dawley rats (Taconic Farms; 250-300 g; n = 61) were anesthetized with a mixture of pentobarbital (35.2 mglkg) and chloral hydrate (170 mg/kg), and the right sciatic nervewas exposed. A crush lesion was made 4 mm distal to the tendon of the obturator internus muscle, and 2 ul of ‘251-labeled neurotrophin, containing phosphate-buffered

saline or a 108fold excess of unlabeled neurotrophins, was injected into the crush site with a Hamilton syringe. Wounds were sutured, and the animals were allowed to recover for 18-24 hr. Rats were killed, and the DRGs were dissected, placed in 4% paraformaldehyde, and counted in a gamma counter for 1 min. Differences in mean transport values were analyzed by analysis of variance (ANOVA) followed by Newman-Keuls’test for multiple comparisons. Differences between means were evaluated for significance at the 0.05 and 0.01 level. In other experiments, animals were reanesthetized and perfused transcardially with heparinized saline followed by a buffered 4% paraformaldehyde solution. DRGs and spinal cords were dissected and counted for 1 min in fixative. For sympathetic neuron transport studies, ‘~+labeled neurotrophins were injected into the anterior eye chamber as described (Johnson et al., 1978). After 16 hr, SCCs were removed, counted in 4% paraformaldehyde, and processed as for DRGs. For intrahippocampal injections, animals (n = 17) were anesthetized (as above) and a solution containing 1151-labeled trophic factor or cytochrome c (0.2-0.4 ~1) was injected stereotaxically into the hippocampus byway of a borosilicate glass micropipette attached to a pressure microinjection apparatus. The coordinates, from Bregma, were AP -3.8, ML t2.2, DV -3.4. Approximately24 hr later, the animals were perfused as described above. Fifteen of the 17 injections were centered, as intended, in the dentate gyruslhilar region of the dorsal hippocampus with variable involvement of the adjacent medial parts of CA3 and/or the suprapyramidal layers of CAI. All animal use in this study was conducted in compliance with approved institutional animal care and use protocols and according to NIH guidelines (Guide for the Care and Use of Laboratory Animals, NIH publication no. 86-23, 1985). Tissue Sectioning and Emulsion Autoradiography Fixed DRCs and spinal cords were equilibrated with buffered sucrose, frozen in methyl butane, sectioned in a cryostat (IO urn for DRC and SCC, and 20 urn for spinal cord), and then mounted onto microscope slides. The brains from perfused animals were removed and equilibrated in buffered sucrose. Frozen sections (25 urn thick) were cut in the coronal plane and mounted. Slides were then processed for emulsion autoradiography (using Kodak NTB-2 emulsion) following established procedures (Cowan et al., 1972). Exposure times ranged from 1 to 3 weeks; however, comparable exposure times were used for any individual region. After being developed, tissues were counterstained through the emulsion (NTB-2, Kodak) with thionin. Acknowledgments We thank Dr. James Miller and Robert Rosenfeld of Amgen Inc., Thousand Oaks, CA, for purified BDNF. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertrsement” in accordance with 18 USC Section 1734 solely to indicate this fact. Received

December

5, 1991;

revised

March

6, 1992.

References Alderson, R. F., Alterman, A. L., Barde, Y.-A., and Lindsay, R. M. (1990). Brain-derived neurotrophic factor increases survival and differentiated functions of rat septal cholinergic neurons. Neuron 5, 297-306. Barde, Y.-A., Edgar, D., and Thoenen, H. (1982). Purification of a new neurotrophic factor from mammalran brain. EMBO J. 1,549553.

Bocchini, V., and Angeletti, P. U. (1969). The nerve growth factor: purification as a 30,000-molecular-weight protein. Proc. Natl. Acad. Sci. USA 64, 787-794. Ciudad,

C. J., Urlaub,

G., and Chasin,

L. A. (1988). Deletion

analy-

NWWJll 992

sis of the Chinese hamster dihydrofolate moter. J. Biol. Chem. 263, 16274-16282. Cowan, W. M., Gottlieb, D. I., Hendrickson,A. Woolsey, T. A. (1972). The autoradiographic axonal connections in the central nervous 21-51.

reductase

gene

pro-

E., Price, J. L., and demonstration of system. Brain Res. 37,

Ernfors, P., Ib&iez, C. F., Ebendal, T., Olson, L., and Persson, H. (1990a). Molecular cloning and neurotrophic activities of a protein with structural similarities to P-NGF: developmental and topographical expression in the brain. Proc. Natl. Acad. Sci. USA 87, 5454-5458.

Leibrock, J., Lottspeich, F., Hohn, A., Hofer, M., Hengerer, B., Masiakowski, P., Thoenen, H., and Barde, Y.-A. (1989). Molecular cloning and expression of brain-derived neurotrophicfactor. Nature 347, 149-152. Lindsay, R. M. (1988). Nervegrowth factors (NGF, BDNF) axonal regeneration but are not required for survival sensory neurons. J. Neurosci. 8, 2394-2405.

enhance of adult

Lindsay, R. M., Thoenen, H., and Barde, Y.-A. (1985). Placode and neural crest-derived sensory neurons are responsive at early developmental stages to brain-derived neurotrophic factor. Dev. Biol. 772, 319-328.

Ernfors, P., Wetmore, C., Olson, L., and Persson, H. (1990b). Identification of cells in rat brain and peripheral tissues expressing mRNA for members of the nerve growth factor family. Neuron 5, 511-526.

Maisonpierre, P. C., Belluscio, M. E., Lindsay, R. M., and Neurotrophin-3: a neurotrophic Science 247, 1446-1451.

Gall, C. M., and Isackson, P. J. (1989). Limbic seizures neuronal production of messenger RNA for nervegrowth Science 245, 758-761.

Maisonpierre, P. C., Belluscio, L., Friedman, B., Alderson, R. F., Wiegand, S. J., Furth, M. E., Lindsay, R. M., and Yancopoulos, G. D. (1990b). NT-3, BDNF, and NGF in thedeveloping rat nervous system: parallel as well as reciprocal patterns of expression. Neuron 5, 501-509.

increase factor.

Hallb66k, F., IbBAez, C. F., and Persson, H. (1991). Evolutionary studies of the nerve growth factor family reveal a novel member abundantly expressed in Xenopus ovary. Neuron 6, 845-858. Hempstead, B. L., Martin-Zanca, Chao, M. V. (1991). High-affinity sionofthetrkproto-oncogeneand Nature 350, 678-683. Hendry, I. A., StGckel, The retrograde axonal Res. 68, 103-121. Hofer, factor

M. M., and Barde, prevents neuronal

D., Kaplan, D., Parada, L. F., and NGF binding requires coexpresthelow-affinity NCFreceptor.

K., Thoenen, H., and Iversen, L. L. (1974). transport of nerve growth factor. Brain Y.-A. (1988). Brain-derived death in viva. Nature

neurotrophic 337, 261-262.

Hohn, A., Leibrock, J., Bailey, K., and Barde, Y.-A. (1990). Identification and characterization of a novel member of the nerve growth factor/brain-derived neurotrophic factor family. Nature 344, 339-341. Hyman, C., Hofer, M., Barde, Y.-A., Juhasz, M., Yancopoulos, G. D., Squinto, S. P., and Lindsay, R. M. (1991). BDNF is a neurotrophic factor for dopaminergic neurones of the substantia nigra. Nature 350, 230-232. Isackson, P. J., Huntsman, M. M., Murray, (1991). BDNF mRNA expression is increased after limbic seizures: temporal patterns from NGF. Neuron 6, 937-948.

K. D., and Gall, C. M. in adult rat forebrain of induction distinct

Johnson, E. M., Andres, R. Y., and Bradshaw, R. A. (1978). Characterization of the retrograde transport of nerve growth factor (NCF) using high specific activity [‘251]NCF. Brain Res. 750, 319331. Johnson, E. M., Taniuchi, M., Clark, H. B., Springer, 1. E., Koh, S., Tayrien, M. W., and Loy, R. (1987). Demonstration of the retrograde transport of nervegrowth factor receptor in the peripheral and central nervous system. J. Neurosci. 7, 923-929. Johnson, J. E., Barde, Y.-A., Schwab, M., and Thoenen, H. (1986). Brain-derived neurotrophic factor supports the survival of cultured rat retinal ganglion cells. J. Neurosci. 6, 3031-3038. Kalcheim, C., Barde, Y.-A., Thoenen, (1987). In viva effect of brain-derived survival of developing dorsal root 2871-2873.

H., and Le Douarin, N. M. neurotrophic factor on the ganglion cells. EMBO J. 6,

Klein, R., Parada, L. F., Coulier, F., and Barbacid, M. (1989). trkB, a novel tyrosine kinase receptor expressed during mouse neural development. EMBO J. 8, 3701-3709. Klein, R., Jing, S., Nanduri, V., O’Rourke, (1991). The trk proto-oncogene encodes growth factor. Cell 65, 189-197.

E., and Barbacid, M. a receptor for nerve

Korsching, S., and Thoenen, H. (1983). Quantitative demonstration of the retrograde axonal transport of endogenous nerve growth factor. Neurosci. Lett. 39, l-4. Lamballe, F., Klein, R., and Barbacid, M. (1991). t&C, a new member of the trk family of tyrosine protein kinases, is a receptor for neurotrophin-3. Cell 66, 967-979.

L., Squinto, S., Ip, N. Y., Furth, Yancopoulos, G. D. (1990a). factor related to NGF and BDNF.

Marchalonis, J. J. (1969). An enzymic method for the trace lodination of immunoglobulins and other proteins. Biochem. J. 773, 299-305. Oppenheim, R. W., Maderdrut, J. L., and Wells, D. J. (1982). Cell death of motoneurons in the chick embryo spinal cord. VI. Reduction of naturally occurring cell death in the thoracolumbar column of Terni by nerve growth factor. J. Comp. Neural. 270, 174-189. Palmatier, M. A., Hartman, B. K., and Johnson, E. M. (1984). Demonstration of retrogradely transported endogenous nerve growth factor in axons of sympathetic neurons. J. Neurosci. 4, 751-756. Raivich, C., Zimmermann, A., and Sutter, A. (1985). 1 he spatial and temporal pattern of PNCF receptor expression in the developing chick embryo. EMBO J. 4, 637-644. Raivich, C., Zimmermann, A., and Sutter, A. (1987). Nerve growth factor (NCF) receptor expression in chicken cranial development. J. Comp. Neural. 256, 229-245. Richardson, P. M., and Riopelle, R. J. (1984). Uptake of nerve growth factor along peripheral and spinal axons of primary sensory neurons. J. Neurosci. 4, 1683-1689. Rodriguez-Tebar, A., Dechant, C., and Barde, Y.-A. (1990). Binding of brain-derived neurotrophic factor to the nerve growth tactor receptor. Neuron 4, 487-492. Schwab, M. E., Otten, U., Agid, Y., and Thoenen, H. (1979). Nerve growth factor (NGF) In the rat CNS: absence of specitic retrograde axonal transport and tyrosine hydroxylase induction in locus coeruleus and substantia nigra. Brain Res. 768, 473-483. Seiler, M., and Schwab, M. E. (1984). Specific of nerve growth factor from the neocortex the rat. Brain Res. 300, 33-39.

retrogradetransport to nucleus basalis

Snider, W. D., and cules. Ann. Neural.

Neurotrophlc

Johnson, E. M. (1989). 26, 489-506.

in

mole-

Soppet, D., Escandon, E., Maragos, J., Middlemas, D. S., Reid, S. W., Blair, J., Burton, L. E., Stanton, B. R., Kaplan, D. R., Hunter, T., Nikolics, K., and Parada, L. F. (1991). The neurotrophic factors brain-derived neurotrophic factor and neurotrophrn-3 are Iigands for the trkB tyrosine kinase receptor. Cell 65, 895-903. Squinto, S. P., Stitt, T. N., Aldrich, T. H., Davis, S., Bianco. 5. M., Radziejewski, C., Glass, D. I., Masiakowski, P.. Furth, M. E., Valenzuela, D. M., DiStefano, P. S., and Yancopoulos, G. D. (1991). trkB encodes a functional receptor for brain-derived neurotrophic factor and neurotrophin-3 but not nerve growth factor. Cell 65, 885-893. Stiickel, K., and Thoenen, of nerve growth factor: Brain Res. 85, 337-341. St(ickel, K., Schwab, retrograde transport

H. (1975). Retrograde axonal transport specificity and biological importance.

M. E., and Thoenen, H. (1975). Speciflcityot of nerve growth factor (NCF) in sensory

Retrograde 993

neurons: I-14.

Axonal

Transport

a biochemical

of Neurotrophins

and morphological

Thoenen, H. (1991). The changing Trends Neurosci. 14, 165-170. Wayne, D. B., and Heaton, NCF by early chick embryo 127, 220-223.

scene

study.

Brain

of neurotrophic

M. B. (1988). Retrograde spinal cord motoneurons.

Res. 89, factors.

transport of Dev. Biol.

Yan, Q., and Johnson, E. M. (1988). An immunohistorhemical study of the nerve growth factor (NCF) receptor in developing rats. J. Neurosci. 8, 3481-3498. Yan,Q., Snider, W. D., Pinzone, J. J., and Johnson, E.M., Jr. (1988). Retrograde transport of nerve growth factor (NGF) in motoneurons of developing rats: assessment of potential neurotrophic effects. Neuron 1, 335343. Yip, H. K., and Johnson, E. M. (1983). Retrograde transport of nerve growth factor in lesioned goldfish retinal ganglion ceils. 1. Neurosci. 3, 2172-2182. Zafra, F., Hengerer, B., Leibrock, J., Thoenen, H., and Lindholm, D. (1990). Activity dependent regulation of BDNF and NCF mRNAs in the rat hippocampus is mediated by non-NMDA glutamate receptors. EMBO J. 9, 3545-3550.