Carrier Mediated Delivery of NGF:Alterations in Basal Forebrain Neuronsin Aged Rats Revealed Using Antibodies

Against Low and High AffinityNGF Receptors

CRISTINA BACKMAN,1 GREGORY M. ROSE,3,4 RAYMOND T. BARTUS,6

BARRY J. HOFFER,2,3 ELLIOTT J. MUFSON,5 AND ANN-CHARLOTTE GRANHOLM1,3*1Department of Basic Science and Neuroscience Program, University of Colorado HSC,

Denver, Colorado 802622Department of Psychiatry, University of Colorado HSC, Denver, Colorado 80262

3Department of Pharmacology, University of Colorado HSC, Denver, Colorado 802624Medical Research Service, Veterans Affairs Medical Center, Denver, Colorado 802205Department of Neurological Sciences, Rush Presbyterian–St. Luke’s Medical Center,

Chicago, Illinois 606126Alkermes Incorporated, Cambridge, Massachusetts 02139-4136

ABSTRACTThe distribution of low and high affinity nerve growth factor (NGF) receptors was

investigated in the basal forebrain during aging and NGF treatment. A peripheral administra-tion model for NGF was utilized. NGF was conjugated to a transferrin receptor antibody(OX-26-NGF), and this conjugate was injected into the tail vein of aged Fischer 344 male rats(24 months) twice weekly for 5 weeks (equivalent to 50 µg of NGF/injection). Controls wereinjected with a non-conjugated mixture of OX-26 and NGF. The aged rats treated withconjugate showed a significant increase in cell size of p75- and trkA-immunoreactive neuronsin the medial septal nucleus and vertical limb of the diagonal band as compared to controls. Asignificant increase in cell size of trkA-immunoreactive neurons was also observed in thehorizontal limb of the diagonal band in rats treated with conjugate. Rats treated withconjugate also showed a significant increase in overall staining density for p75 and trkAantibodies in the medial septal nucleus as compared to controls. A significant increase instaining density of p75-immunoreactive structures was also observed in the vertical andhorizontal limbs of the diagonal band. Therefore, treatment with OX-26-NGF conjugate hasregional effects on both the low and high affinity NGF receptors in terms of cell body size andstaining density in the basal forebrain of aged rats. The current findings support the idea thatthis delivery system might be useful in therapeutic approaches involving the delivery ofneurotrophic factors and other large molecules into the brain. J. Comp. Neurol. 387:1–11,1997. r 1997 Wiley-Liss, Inc.

Indexing terms: septal nucleus; neurotrophic factors; aging; plasticity

Nerve growth factor (NGF) is a well characterizedneurotrophic substance found in both the peripheral andcentral nervous system (CNS; Levi-Montalcini, 1982;Gnahn et al., 1983; Thoenen and Edgar, 1985). Numerousinvestigations have revealed that this trophic factor sup-ports the maintenance, neurite outgrowth, and phenotypicexpression of basal forebrain cholinergic neurons in theCNS (Hefti and Will, 1987; Ebendal, 1992). The cholinergicneurons of the basal forebrain provide innervation to thehippocampus, olfactory bulb, cortex, and amygdala (Mesu-

lam et al., 1983; Wainer and Mesulam, 1990). Evidencefrom studies in both humans and animals suggestthat these cholinergic cell groups located within the

Grant sponsor: USPHS; Grant numbers: AG12122, AG04418, AG10755,MH49661.

*Correspondence to: Dr. Ann-Charlotte Granholm, Dept. Basic Science,Box C286, Univ. Colorado, Health Sci. Ctr., 4200 East Ninth Ave., Denver,CO 80262. E-mail: granholm@UCHSC.edu

Received 19 November 1996; Revised 6 May 1997; Accepted 7 May 1997

THE JOURNAL OF COMPARATIVE NEUROLOGY 387:1–11 (1997)

r 1997 WILEY-LISS, INC.

medial septal nucleus (Ch1), nucleus of the diagonal bandof Broca (Ch2), nucleus of the horizontal band of Broca(Ch3), and the nucleus basalis of Meynert (Ch4; Mesulamet al., 1983), play an important role in memory function(Hepler et al., 1985; Miyamoto et al., 1987). This systemhas been the focus of intense research in light of thepossibility that NGF may be of potential therapeuticsignificance in alleviating cholinergic neuron atrophy andcognitive dysfunction in neurodegenerative disorders thataffect the cholinergic neuronal population, such as demen-tia of the Alzheimer’s type (SDAT).

It is well known that the basal forebrain cholinergicsystem also undergoes degenerative changes during nor-mal aging in both humans and animals (Gilad et al., 1987;Altavista et al., 1988, 1990; Fisher et al., 1989). Theshrinkage and loss of acetyl cholinetransferase (ChAT)and p75-positive neurons in the basal forebrain has beencorrelated with the extent of spatial memory impairmentin aged rats (Koh et al., 1989; Fisher et al., 1989, 1992;Armstrong et al., 1993). A link between age-related impair-ments in cognitive dysfunction, NGF, and the atrophy ofbasal forebrain cholinergic neurons was suggested by theobservation that chronic intraventricular infusion of NGFcan reduce cholinergic neuron atrophy and improve spatialmemory retention in memory impaired aged rats (Fischeret al., 1987, 1992). It has also been shown that intracranialNGF administration can ameliorate age-related choliner-gic hypofunction by significantly increasing extracellularacetylcholine (ACh) levels in both parietal cortex andhippocampus (Scali et al., 1994). The studies summarizedabove suggest that NGF may reverse the degenerationprocess of the cholinergic system during aging. However,the detailed mechanisms for this cell rescue have not yetbeen fully established.

NGF exerts its actions on cholinergic neurons throughspecific receptors (Ebendal et al., 1992; Lapchak et al.,1993). Cholinergic neurons in the basal forebrain expressboth low affinity (p75; Batchelor et al., 1989) and highaffinity (trkA; Steininger et al., 1993; Sobreviela et al.,1994) receptors for NGF and exhibit selective uptake andretrograde transport of labeled NGF from their corticaltarget regions (Schwab et al., 1979; Seiler and Schwab,1984). Virtually all trkA-immunoreactive neurons withinthe septal/diagonal band complex (.95%) co-express p75(Sobreviela et al., 1994). The p75 receptor also binds othermembers of the neurotrophin family, such as brain-derivedneurotrophic factor (BDNF), neurotrophin-3 (NT3), andneurotrophin-4/5 (NT-4/5; Bothwell, 1991; Ebendal, 1992).It has been suggested that the p75 receptor may play afunctional role in the neurotrophic activity of NGF, appar-ently by sequestering NGF to the binding sites (Hemp-stead et al., 1989). However, the signaling mechanismmore likely involves the activity of the high affinity NGFreceptor trkA . TrkA is a receptor tyrosine kinase of the trkgene family and, in contrast to p75, is highly specific forNGF (Kaplan et al., 1994; Klein et al., 1991; Chao, 1992).NGF binding to trkA results in dimerization and autophos-phorylation of tyrosine residues with resultant signaling(Battleman et al., 1993). Although it is generally agreedthat trkA participates in high affinity NGF binding andsignal transduction, it has been proposed that the p75receptor is capable of modulating the neurotrophic re-sponse by facilitating the activation of trkA (Verdiet al., 1994).

Previous studies have shown that both low and highaffinity NGF receptors are altered during aging. A reduc-tion in cell numbers, cell size, and optical density ofp75-immunoreactive neurons has been shown in the basalforebrain of aged rats as compared to young animals (Kohet al., 1988, 1989; Hefti et al., 1989; Markram and Segal,1990). TrkA mRNA levels are also significantly reducedduring aging in the septal area (Cooper et al., 1994;Parhad et al., 1995). However, to date, studies of NGFeffects on the aged forebrain cholinergic system have beenmostly concentrated on cholinergic markers and low affin-ity NGF receptors. Studies have not yet been performed, toour knowledge, on trkA receptor immunohistochemistryduring aging. Therefore, in the present study, we exam-ined the effects of NGF on aged trkA-immunoreactiveneurons. Furthermore, in order to compare the effects ofexogenous NGF on both receptor types, we performed aparallel examination of p75 vs. trkA-immunoreactive struc-tures in the basal forebrain of aged rats.

A major concern in potential clinical applications of NGFto treat neurodegeneration is the means of administrationof this growth factor. The NGF molecule is too large tocross the blood-brain barrier (BBB) in any significantamounts (Friden et al., 1993; Granholm et al., 1994;Kordower et al., 1994), and all currently available deliverymethods require intracranial administration. Even if thesemethods have been proven to be effective in rescuing thecholinergic neurons in different animal models (Mobley etal.,1985; Kromer, 1987), they all require intracranial sur-gery which could cause negative side effects, such asinfection and changes in intracerebral pressure. There-fore, a non-invasive delivery system of NGF into the brainwould be advantageous. In the present investigation weutilized a non-invasive peripheral NGF delivery system toinvestigate the possibility of transporting a growth factoracross the blood-brain barrier to ameliorate the degenera-tive changes in cholinergic neurons during aging. We havepreviously demonstrated that NGF can be covalentlylinked to rat (OX26) or monkey (AK30) anti-transferrinreceptor antibodies and that intravenous injection of theconjugate leads to the transport of biologically active NGFacross the BBB (Friden et al., 1993; Kordower et al., 1993).In a previous study, we have shown that aged rats, whichwere behaviorally impaired in a spatial learning task,significantly improved their learning skills after receivingtreatment with the OX-26-NGF conjugate (Backman et al.,1996). The present study is an extension of this previouswork, wherein we investigate the in situ morphologicaleffects of the OX-26-NGF conjugate in the different fore-brain cholinergic nuclei of aged rats. The specific questionsaddressed in this study were: 1) Does treatment withOX-26-NGF promote the maintenance of neurons contain-ing the low and high affinity NGF receptors in the basalforebrain of aged rats? 2) Does treatment with OX-26-NGFinduce regional changes in the basal forebrain cholinergicnuclei of aged rats? 3) Is the low and high affinity NGFreceptor immunostaining of individual neurons affected bytreatment with an OX-26-NGF conjugate?

MATERIALS AND METHODS

Intravenous injections

Aged male Fischer 344 rats (24 months old) were testedfor spatial learning ability in the Morris water maze.Based upon their performance, compared to a group of

2 C. BACKMAN ET AL.

3-month-old animals, the aged rats were divided intolearning impaired and learning unimpaired groups. Thelearning impaired rats were randomly divided into anexperimental group (n 5 4) and a control group (n 5 5).The behavioral results of this study have been publishedelsewhere (Backman at al., 1996). The experimental group(OX-26-NGF conjugate group) received intravenous injec-tions of OX-26-NGF conjugate (50 µg NGF equivalent doseper injection), while the control group received intrave-nous injections of a co-mixture of unconjugated OX-26 andNGF (OX-26/NGF co-mixture group) at the same dosereceived by the experimental group. An additional controlgroup consisted of five untreated age-matched male Fischer344 rats. Tail vein injections were performed twice weekly,at the same time for both groups, during a period of 5weeks. After this period of time the conjugate and co-mixture treated animals were retested in the water maze(see Backman et al., 1996). All the animals were thenhumanely killed for immunohistochemical analysis. Allanimal procedures followed the NIH guidelines and wereapproved by the local animal care and use committee.

Immunohistochemistry

The animals were anesthetized with chloral hydrate(300 mg/kg i.p.) and perfused transcardially with salinefollowed by paraformaldehyde (PF; 4%) in phosphatebuffer (PB; 0.1 M; pH 7.2). The brains were dissected,postfixed in PF overnight, and transferred to 30% sucrosein 0.1M PB for at least 16 hours. Serial cryostat sections(30 microns thick) were obtained through the area of thebasal forebrain which contained the basal forebrain cholin-ergic nuclei. Every sixth section through this region wascollected for immunohistochemical staining. In order tocontrol for staining variability, specimens from each of thethree experimental groups were included in every batchand reacted together in a net well tray under the sameconditions. Immunohistochemistry for the trkA receptorwas carried out according to a previously reported proce-dure (Sobreviela et al., 1994; the production and specificityof the trkA antibody [RTA] has also been described previ-ously by Clary et al., 1994). Briefly, sections were rinsed inTris-buffered saline solution (TBS), incubated in a TBSsolution containing 0.1 M sodium periodate to inhibitendogenous peroxidase activity and preincubated for 1hour with 10% normal goat serum (NGS) and 2% bovineserum albumin (BSA) in a TBS solution containing 0.25%TritonX-100 (TBS-Tx). After this, sections were incubatedwith the primary antibody directed aginst the high affinityNGF receptor (RTA, dilution 1:10,000) for 24 hours at roomtemperature in TBS containing 0.4% Tx and 3% blockingserum. Thereafter, the sections were rinsed in TBS andincubated with biotinylated secondary IgG antibodies (1:500, goat anti-rabbit) for 1 hour. Sections were thenincubated for 1 hour in an avidin-biotin complex (‘‘Elitekit’’ Vector Laboratories, Burlingame, CA; dilution 1:200)and washed in a 0.2 M sodium acetate, 1.0 M imidazolebuffer (pH 7.4). Finally, sections were developed with ametal intensified diaminobenzidine (DAB) reaction (2.5%nickel II sulfate, 0.05% 3,3’ diaminobenzidine, and 0.005%H2O2 mixed in this buffer). All sections were rinsed 3 x 10minutes in TBS after each step, unless otherwise indicated.

To visualize the low affinity NGF (p75) receptor, freefloating sections were rinsed in 0.1 M PBS and treatedwith 0.3% H2O2 to inhibit residual endogenous peroxidase.Sections were then preincubated for an hour with normal

goat serum (3%) and bovine serum albumin (2 g/100 ml) in0.1 M PBS to block background staining. After this, thesections were incubated for 48 hours with primary antibod-ies directed against monoclonal 192 IgG antibody againstlow-affinity p75 NGF receptors (Boehringer Mannheim,Indianapolis, IN; dilution 1:500 in 0.1 M PBS with TritonX-100, NGS and BSA; see Backman et al., 1995). Thereaf-ter, the sections were washed in PBS and incubated withbiotinylated horse secondary IgG antibodies directedagainst mouse IgG. In the last step, the sections wereincubated with the ‘‘ABC’’ Elite substrate (Vector), washedwith 0.1 M PB and developed with a metal intensifieddiaminobenzidine (DAB) reaction (20 ml 0.1 M PBS, 0.01 gDAB, 100 µl 8% nickel ammonium sulfate and 68 µl 3%H2O2). All sections were rinsed 3 X 10 minutes in 0.1 MPBS after each step, unless otherwise indicated. Sectionswere mounted on subbed slides, dehydrated, cover slipped,and examined using light microscopy. Sections where theprimary antibody was omitted were included as controls ineach experimental series. When the words ‘‘-immunoreac-tive’’ or ‘‘-positive’’ are used in the text, this always refers to‘‘-like immunoreactivity,’’ since no direct evidence for loca-tion of molecules can be obtained with immunohistochemi-cal techniques.

Image analysis

Image analysis to determine average cell size and over-all staining density in each group reacted with the p75 andtrkA antibodies was achieved with a scale bar in themicroscope in conjunction with a Cohu analog/digital videocamera (4990 series), a Scion frame grabber card, aMacIntosh Quadra 840AV computer, and the NIH ‘‘Image’’software package (written by Waine Rasband). Images fordetermination of cell sizes were acquired with a 20xobjective. Overall staining density measurements weredone with a 10x objective. The sections were digitized asunaltered images, meaning that neither filters nor digitalenhancements were applied. Cell size and overall stainingdensity measurements were performed on five sectionsthrough the basal forebrain for each antibody. Specificanatomical landmarks were used in the selection of thesections to include all the cholinergic nuclei (Ch1–Ch4) inthe analysis. Ch1–Ch3 were measured starting at thegenu of the corpus callosum rostrally, to the level of thecrossing of the anterior commissure caudally. Ch4 wasdefined as the area beginning rostrally at the level of thecrossing of the anterior commisure and extending caudallyto the level of the dorsal hippocampus. Mean cell size andoverall staining density values were determined for eachregion in each rat. A visible nucleus and two main pro-cesses were used as a criterion for a cell body. Opticaldensity measurements were performed by using a scale,such that 0 represented a white image, and 256 repre-sented a black image. With this scale, the optical density ofp75 and trkA immunoreactivity was determined in definedareas. The areas measured were defined by tracing alongthe cell bodies in the outskirts of each region. To control forstaining variability between groups, an average back-ground value, which was taken in three different areas ofthe image without immunoreactivity, was substracted fromeach measured area. Therefore, overall staining measure-ments are quantitative although they are based on relativeintensity units. Relative gray scale units due to variabilitybetween animals have been appropriately controlled be-tween specimens as indicated above. The mean values for

NGF AND ITS RECEPTORS DURING AGING 3

each region in each rat were analyzed by one-way analysesof variance followed by pairwise comparisons between thethree groups via the Tukey-Kramer method. The thresholdfor statistical significance was P , 0.05.

RESULTS

p75-immunoreactive neurons

General characteristics. Alarge number of p75 immu-noreactive (p75-ir) neurons was observed within the basalforebrain in all the groups. p75-ir cell bodies and fibers inall the aged animals appeared dark brown throughout thebasal forebrain when reacted with nickel intensification.The intracellular staining distribution was similar be-tween the aged controls and the aged rats treated withconjugate. p75 staining appeared to be located along theplasma membrane of cell bodies. In addition, some granu-lar accumulations of staining could be observed in thecytoplasm. Immunoreactive fibers were also evenly stainedalong the entire surface. Both smooth and varicose fiberswere observed in the p75 stained sections, suggesting thatthe low affinity NGF receptor may be expressed on thesurface of projecting axons as well as on dendrites (seeFig. 1A).

Cell size. As previously reported (Backman et al.,1996), aged rats treated with the OX-26-NGF conjugateshowed a significant increase in p75-ir cell size in Ch1, ascompared to the age-matched animals treated with theOX-26/NGF co-mixture (Fig. 2A,B). We now extend thisfinding by presenting data from Ch2–Ch4. Image analysisfor cell sizes in Ch2 showed a significant difference be-tween the conjugate and old untreated group (*P , 0.05;see Figs. 3A–C and 5). Ch3 did not show a significantdifference between the groups (see Figs. 4A–C and 5).However, we observed a trend towards larger cell sizevalues in the conjugate treated group as compared tocontrols (see Fig. 5). The average cell sizes in Ch4 werealmost equal between the different groups (see Fig. 5).These results suggest that the area most affected byOX-26-NGF treatment is Ch1, followed by Ch2 and Ch3. In

terms of cell size of p75-ir cells, Ch4 did not appear to beaffected by conjugate treatment.

Staining density. A clear difference in the density ofp75-immunoreactive staining was apparent in the areaoccupied by cell bodies between the experimental andcontrol groups in Ch1 (Fig. 2A–C), Ch2 (Fig. 3A–C), andCh3 (Fig. 4A–C). Quantitative measurements of opticaldensity showed that aged rats treated with OX-26-NGFconjugate exhibited a significant increase in overall p75-immunoreactive staining density in Ch1 and Ch2 ascompared to aged control animals. A significant differencewas also found in Ch3 between the conjugate treatedanimals and the aged untreated control group (see Fig. 6).This difference was not statistically significant in Ch4.However, we observed a small trend in Ch4 towards higherstaining density values in conjugate treated animals ascompared to controls. Thus, as was the case for cell size,overall staining density was most affected by the conjugatetreatment in Ch1, followed by Ch2 and Ch3. The increasein staining density observed in these areas could be due toan increase in the number of stained neurons and neuro-nal cell size, as well as to an increase in the number ofimmunoreactive processes. However, it is clear from theimmunohistochemical staining in Figures 2, 3, and 4, thatthe increase in overall staining density is to a great extentdue to an increase in the density of p75 immunoreactivenerve processes.

TrkA immunoreactive neurons

General characteristics. TrkA-immunoreactive cellbodies in the aged animals appeared black throughoutthe basal forebrain when reacted with nickel intensifica-tion. Similar to the p75 antibody, the distribution ofstaining within individual neurons did not seem to differbetween the different groups, even though the total num-ber of fibers appeared to be increased in the conjugategroup. TrkA appeared to be located mainly in the cyto-plasm of immunoreactive cell bodies. At a higher magnifi-cation dense granular black staining deposits could beobserved within the cell soma of trkA immunoreactiveneurons (Fig. 1B). As with the p75 antiboby, immunoreac-

Fig. 1. Low affinity nerve growth factor (NGF) receptor (p75; A) and high affinity NGF receptor (trkA;B) immunoreactive neurons in the basal forebrain of aged rats treated with an NGF conjugate(OX-26-NGF). Notice the difference in fiber staining between the p75 and trkA antibodies. Scale bar 520 µm.

4 C. BACKMAN ET AL.

tive fibers were found between the cell bodies in thedifferent cholinergic nuclei. However, trkA immunoreac-tive fibers showed a segmental staining along the neuritelength, with numerous regularly spaced varicosities (seeFig. 1B). Therefore, there was a clear difference in stainingdistribution between the p75 and trkA antibodies. Whilethe p75 receptor appeared to be evenly expressed along theentire cell membrane in the cell body and neurites, the

trkA receptor appeared to be primarily expressed in thecytoplasm of cell bodies and in the varicosities of neurites.Furthermore, the expression pattern for the trkA antibodysuggests that mostly axons were labeled.

Cell size. Sections stained with the high affinity trkAreceptor antibody showed cell size changes similar to whatwas observed with p75 staining after treatment with theOX-26-NGF conjugate. A significant increase in cell size

Fig. 2. p75- (A–C) and trkA- (D–F) immunoreactive neurons in themedial septal nucleus (Ch1). A,D: Aged rats treated with co-mixture.B,E: Aged rats treated with conjugate. C,F: Untreated age-matchedcontrols. Note the increase in cell size in the animals treated with theconjugate as compared to the controls and also the vacuolarization and

shrunken appearance of the aged control neurons (A,D). An increase infiber density is also observed in the sections from animals treated withthe conjugate and processed for the p75 antibody (B). Scale bar 550 µm.

NGF AND ITS RECEPTORS DURING AGING 5

was observed in Ch1 of rats treated with the OX-26-NGFconjugate as compared to control groups (see Figs. 2D–Fand 7). A significant increase in cell size was also observedin Ch2 (see Figs. 3D–F and 7) and Ch3 in rats treated withthe OX-26-NGF conjugate as compared to controls (seeFigs. 4D–F and 7). However, as was also seen with the p75antibody, the average cell size for neurons stained with the

trkA antibody did not show any significant difference inCh4 between the different groups.

Staining density. A significant increase in overalltrkA immunoreactive (trkA-ir) staining density was ob-served in Ch1 of animals treated with the OX-26-NGFconjugate as compared to aged controls (see Fig. 8).However, in contrast to the p75 receptor, we did not

Fig. 3. p75- (A–C) and trkA- (D–F) immunoreactive neurons in thenucleus of the diagonal band of Broca (Ch2). A,D: Aged rats treatedwith co-mixture. B,E: Aged rats treated with conjugate. C,F: Un-treated age-matched controls. The average size for neurons was larger

in the animals treated with the conjugate as compared to the controls.Analysis for the fiber density in p75 stained sections showed anincreased amount of fibers in the animals treated with conjugate.Scale bar 5 50 µm.

6 C. BACKMAN ET AL.

observe any significant difference in staining density inCh2 or Ch3 between the different groups. Taken together,cell size and overall staining density measurements forboth antibodies suggest that the most affected area byOX-26-NGF is Ch1 followed by Ch2 and Ch3. TrkA-irneurons in Ch4 did not appear to be affected by conjugatetreatment, similar to the results obtained with the p75

antibody. Mean staining density measurements for thetrkA antibody are shown in Figure 8.

DISCUSSION

We have shown in a previous study (Backman et al.,1995) that chronic treatment with an NGF conjugate can

Fig. 4. p75- (A,C) and trkA- (D,F) immunoreactive neurons in thenucleus of the horizontal band of Broca (Ch3). A,D: Aged rats treatedwith co-mixture. B,E: Aged rats treated with conjugate. C,F: Un-treated age-matched controls. The animals treated with the conjugate

showed an increase in cell body size as compared to the controls. Thefiber plexus was also more dense in p75 stained sections from animalstreated with the conjugate. Scale bar 5 50 µm.

NGF AND ITS RECEPTORS DURING AGING 7

upregulate cholinergic markers in aged septal forebrainintraocular transplants. Furthermore, we have shown thatthe same conjugate can attenuate mnemonic dysfunction inaged rats and that this effect on memory is coupled with aneffect on p75-ir neurons in Ch1 (Backman et al., 1996).However, most investigations to date have focused on theeffects of NGF on the low affinity (p75) NGF receptor. Inthe present study, we analyzed and compared the effects

of an NGF conjugate (OX-26-NGF) on trkA and p75receptors during the aging process. The close interactionbetween the p75 and trkA receptors requires an elucida-tion of the NGF effects on the trkA receptor in order to fullyunderstand how this growth factor promotes cholinergiccell survival in the basal forebrain of aged animals.Furthermore, the parallel comparison between the tworeceptor types allowed us to determine if both p75 andtrkA in aged rats show a similar response to treatmentwith the OX-26-NGF conjugate in the basal forebrain.Unfortunately, changes in the innervation of target areassuch as the neocortex and hippocampus could not beaddressed in this investigation, since previous investiga-tors have shown that neither the p75 nor trkA antibodiesare reliable markers for target site innervation patterns(Sobreviela et al., 1994).

Fig. 5. Bar graph illustrating average cell sizes in control andexperimental groups for the low affinity NGF receptor or p75 antibodyin Ch2, Ch3, and Ch4. White bars represent the co-mixture-treatedgroup (n 5 5 rats), black bars indicate the conjugate-treated group(n 5 4 rats), and dotted bars represent the aged non-treated group(n 5 5 rats). The error bars represent standard error of the mean.Average values for p75 cell size in Ch1 have been presented in aprevious publication (Backman et al., 1996). The average cell size inCh1 was significantly increased in the group treated with the conju-gate as compared to the control groups (Backman et al., 1996). We alsofound a significant increase in cell size in Ch2 between the conjugategroup and old untreated group (*P , 0.05). The same trend wasobserved in Ch3. However, average values in this area were notsignificantly different between the groups (P . 0.05, ANOVA withTukey-Kramer a posteriori analysis).

Fig. 6. Bar graph illustrating average staining density values incontrol and experimental groups for the low affinity NGF receptor (p75antibody). White bars represent the co-mixture-treated group (n 5 5rats), black bars indicate the conjugate- treated group (n 5 4 rats), anddotted bars represent the aged non-treated group (n 5 5 rats). Theerror bars show standard error of the mean. ANOVA with Tukey-Kramer a posteriori analysis was used for statistical analysis. Wefound a significant increase in staining density in Ch1 (*P , 0.05conjugate versus co-mixture, **P , 0.01 conjugate vs. aged controls),Ch2 (**P , 0.01 conjugate versus co-mixture, *P , 0.05 conjugate vs.aged untreated controls), and Ch3 (*P , 0.05 conjugate vs. ageduntreated controls) of conjugate-treated animals. No significant differ-ences were found in Ch4.

Fig. 7. Bar graph illustrating average cell sizes in control andexperimental groups for the high affinity NGF receptor (trkA anti-body) in Ch1-4. White bars represent the co-mixture-treated-group(n 5 5 rats), black bars indicate the conjugate-treated group (n 5 4rats), and dotted bars represent the aged non-treated group (n 5 5rats). The error bars represent standard error of the mean. We found asignificant difference in cell size between the conjugate and agedcontrol groups in Ch1 (***P , 0.001 conjugate vs. co-mixture and ageduntreated controls), Ch2 and Ch3 (*P , 0.05 conjugate vs. co-mixtureand aged untreated controls). No significant differences were foundin Ch4.

Fig. 8. Bar graph showing average staining density measurementsin control and experimental groups for the high affinity NGF receptor(trkA antibody) in Ch1–Ch4. White bars represent the co-mixture-treated group (n 5 5 rats), black bars indicate the conjugate-treatedgroup (n 5 4 rats), and dotted bars represent the aged control group(n 5 5 rats). The error bars show standard error of the mean. ANOVAwith Tukey-Kramer a posteriori analysis was used for statisticalanalysis. We found a significant increase in staining density in Ch1 ofconjugate-treated animals (*P , 0.05 conjugate versus co-mixture andaged untreated controls). No significant differences between thegroups were found in Ch2, Ch3, and Ch4.

8 C. BACKMAN ET AL.

We found a specific regional increase of p75 and trkAimmunoreactivity in the basal forebrain of aged ratstreated with the conjugate as compared to age-matchedcontrols. The increased immunoreactivity in the experimen-tal group was quite similar for both receptor types in termsof regional effects, cell size, and staining density measure-ments. Furthermore, we compared the staining distribu-tion at the cellular level between the low and high affinityNGF receptors in the different groups. We observed a cleardifference in the staining distribution between the p75 andtrkAreceptors, suggesting that these receptors are commit-ted to different functions. Despite their obvious differencein biological functions, the NGF receptor types appear tohave a similar response to NGF during the aging process.Based on previous studies using young controls (Sobre-viela et al., 1994; Backman et al., 1996), we conclude fromthe present study that treatment with the OX-26-NGFconjugate can partially reverse changes seen in both p75and trkA receptors during the aging process.

Previous studies have shown that intracranial treat-ment with NGF can partially reverse the atrophy of basalforebrain cholinergic neurons and improve spatial learn-ing in aged rats (Fischer et al., 1987, 1991, 1994;Markowska et al., 1994; Scali et al., 1994). It has beenshown that NGF may up-regulate the expression of the lowaffinity NGF receptor, p75, in the basal forebrain of agedrats, suggesting that the loss of p75 immunoreactivity inaged rats may be due to a reduced availability of target-derived NGF (Chen and Gage, 1995). Furthermore, exog-enous NGF can stimulate the retrograde transport of thehigh affinity NGF receptor, trkA, in the peripheral nervoussystem (Ehlers et al., 1995), and up-regulate trkA mRNAlevels in the brain (Holtzman et al., 1992; Li et al., 1995).These findings suggest that NGF may play a regulatoryrole in the expression and function of both its low and highaffinity receptors. Therefore, in this study we analyzed theeffects of an NGF conjugate on the basal forebrain of agedrats as compared to age-matched controls untreated ortreated with a co-mixture of OX-26 and NGF. Recentfindings from SDAT brains and aged rodents have shownthat NGF levels are unaltered, or elevated, in target areas(Crutcher et al., 1991; Alberch et al., 1991). However, itwas demonstrated that NGF levels are significantly de-creased in forebrain cholinergic nuclei (Mufson et al.,1994), and Markram and Segal (1990) have found that theloss of p75 labeling in aged rats is also not uniform in basalforebrain. The largest labeling loss was seen in Ch1,followed by Ch2, areas that project to the hippocampus. Alesser degree of age-related p75 staining reduction wasseen in Ch3, which projects to olfactory and cortical areas.These regional differences in the basal forebrain duringaging support our findings in the present study.

We found that the area most affected by treatment withOX-26-NGF conjugate was Ch1. A significant increase incell size and overall staining density was observed in thisarea for both the p75 and trkAantibodies in animals treatedwith the conjugate. Therefore, it appears that cholinergicneurons in Ch1, which have been shown to manifest thegreatest reduction in p75 staining during aging (Markramand Segal, 1990), also show the largest response to treat-ment with the NGF conjugate. Ch2 and Ch3 showed amore moderate response to treatment with the OX-26-NGFconjugate. In these areas a significant increase in overallstaining density was observed with the p75 antibody in theconjugate treated animals, as compared to controls. Wealso found a significant increase in cell size in the conjugate

treated group as compared to the old untreated controls inCh2. TrkA immunoreactive sections showed a significantincrease in cell size in the conjugate treated group in bothCh2 and Ch3. Staining density measurements providesdata on alterations in all the stained structures within anarea including cell bodies and processes. However, theoverall staining density measurements together with theimmunohistochemical data presented here suggest thatfiber staining density is increased in Ch1 for both the p75and trkA antibodies. This increase in fiber density can beclearly observed in the sections corresponding to themedial septal nucleus or Ch1 for both antibodies in theanimals treated with the conjugate as compared to con-trols (see Fig. 1A–F). An increase in overall stainingdensity was also observed in areas Ch2 and Ch3 for thep75 antibody (see Figs. 3A,B,C and 4A,B,C). However, wedid not find any significant changes between the groups intrkA overall staining density in Ch2 or Ch3. This discrep-ancy may be due to differences in the fiber staining patternbetween the receptor antibodies since trkA does not appearto stain all neuronal processes (see below). A possibleexplanation for the regional differences in OX-26-NGFeffects between Ch1-3 and Ch4 could be a difference invascular distribution or permeability between these differ-ent regions. Preliminary permeability studies in our lab(Curtis et al., 1995) indicate that there might be regionaldifferences in permeability and/or vascularization betweenthe different nuclei in the forebrain. Further studies needto be performed to determine the involvement of suchfactors using the OX-26-NGF delivery system.

The fact that both the density of p75- and trkA-irneurites were increased in conjugate treated rats does notreveal whether this was the result of neurite sprouting orjust increased levels of receptors within existing neurites.Future studies using growth cone-specific markers such asgap43 might reveal whether actual sprouting had oc-curred. In addition to the marked effects seen in neuronalprocesses, other morphological parameters were affectedby conjugate treatment. The vacuolated and shrunkenappearance of numerous neurons in the aged controlanimals also seemed to be reduced by treatment withconjugate (see Fig. 2A–F). These results suggests thatprocesses have been initiated at the transcriptional levelto restore the function of atrophied cholinergic neurons inanimals treated with conjugate, possibly by up-regulatingthe number of receptors transported to nerve terminals.The induction of p75 and trkA expression by NGF suggeststhat neurotrophin-mediated up-regulation of the receptorsis an important feature of NGF actions and that neuro-trophins may regulate the activity of responsive neuronsthrough increasing endogenous levels of their receptors.Other investigators have also demonstrated that exog-enous NGF can upregulate the expression of p75 (Ma et al.,1992; Mearow et al., 1995), but this correlation betweenNGF and receptor levels has not been shown, to ourknowledge, for the high affinity trk receptors. Further-more, the finding that morphological parameters for thetrkA receptor can be improved during the aging process bytreatment with an NGF conjugate is of importance sincethis receptor is critical for mediation of the biological effectof NGF by initiating autophosphorylation and secondmessenger signalling pathways in the cell (Kaplan andStephens, 1994; Chao and Hempstead, 1995). While thefindings in this study are in agreement with the findings ofothers with respect to the effects of NGF on the p75

NGF AND ITS RECEPTORS DURING AGING 9

receptor, a significant difference between those studies andour findings are that: 1) NGF was delivered to the brainusing a non-invasive systemic delivery system, and 2) wealso analyzed the effects of NGF treatment on its highaffinity trkA receptor.

Recent work in patients with SDAT has suggested thateven though there is a decrease in number of p75 labeledneurons in the basal forebrain during aging and dementia(Mufson et al., 1995), the remaining neurons appear tocontain normal levels of low affinity NGF receptor. Thesame appears to be true for the aged rats in the presentstudy, in that remaining neurons did contain both trkAand p75 staining, even though significantly fewer fiberscould be seen. These findings suggest that the cytoplasmicproduction and the anterograde transport of these recep-tors to the end terminals has not been completely dis-rupted during the aging process. Furthermore, we observeda clear difference in the staining distribution between thelow and high affinity NGF receptors. The p75 appeared tobe located along the plasma membrane with granularaccumulations in the cytoplasm of immunoreactive cellbodies. In contrast, the profile of trkA immunoreactiveneurons seemed to be characterized by dense granularaccumulations in the cytoplasm of positive cell bodies. Thecell membrane did not seem to be clearly stained. We alsoobserved a striking difference in the neurite stainingdistribution between the low and high affinity NGF recep-tors. p75 immunoreactive neurites were evenly stainedalong the entire surface and both smooth and varicosefibers were observed in p75 stained sections. In contrast,trkA neurites showed a segmental staining along theneurite length, with accumulation of staining at the vari-cosities; no smooth fibers seemed to be stained. A potentialmechanism that could account for these differences instaining distribution is that the p75 receptor may act toconcentrate NGF locally in the microenvironment sur-rounding cell surface trkA receptors. The p75 receptor maytherefore act to enhance the ability of trkA to respond toNGF by increasing the likelihood of a stable NGF-trkAinteraction. It is possible that there is a requirement of aneven distribution of p75 receptors along the entire neuriteto optimize the local concentration of NGF. In contrast, thetrkA receptor might be mainly localized at the transportvesicle sites at the varicosities, where it can bind NGF, andthen be internalized as a ligand-receptor complex (Ross etal., 1994). Furthermore, it has been shown that the p75receptor binds to other neurotrophic factors such as BDNF,NT-3 and NT4/5 (Bothwell, 1991; Ebendal,1992), and thatcholinergic neurons are also sensitive to some of theseneurotrophins. For example, BDNF has been shown toprotect Ch1 neurons following fimbria-fornix lesions, andin vitro positive effects of NT-3 and BDNF have also beendescribed in basal forebrain cultures (Friedman et al.,1993; Morse et al., 1993). Therefore, the expression of p75receptors along the entire membrane surface both in thecell bodies and neurites may be necessary for the interac-tion of this receptor with different members of the neuro-trophin family. It is highly likely that the p75 and trkAimmunoreactive cell bodies are the same cholinergic neu-rons of the septal nucleus, since previous studies haveshown a high correlation between these stainings (Sobre-viela et al., 1994). Thus, the increase in cell size describedhere must reflect a true hypertrophy of these neurons, andnot simply a change in staining characteristics within thecell, since both receptors responded similarly.

In conclusion, we have compared the staining distribu-tion for the high and low affinity NGF receptors in agedcontrols vs. NGF conjugate-treated animals. We observeda significant increase in overall staining density and cellsize in specific regions of the basal forebrain in agedanimals treated with the conjugate as compared to controlanimals. Therefore, it is suggested that age-related changesin forebrain cholinergic neurons can be reversed using anon-invasive delivery system of NGF. This study, in con-junction with previous investigations, indicates that periph-erally administered NGF can be delivered into the brain,by utilizing the iron-transferrin transport mechanism, toameliorate the detrimental effects of aging on forebraincholinergic neurons.

ACKNOWLEDGMENTS

We thank Dr. Gary Zerbe for expert assistance with thestatistics. We gratefully acknowledge Genentech for supply-ing the NGF used in this study.

LITERATURE CITED

Altavista, M.C., A.R. Bentivoglio, P. Crociani, P. Rossi, and A. Albanese(1988) Age-dependent loss of cholinergic neurons in the basal ganglia ofrats. Brain Res. 455:177–181.

Altavista, M.C., P. Rossi, A.R. Bentivoglio, P. Crociani, and A. Albanese(1990) Ageing is associated with diffuse loss of forebrain cholinergicneurons. Brain Res. 508:51–59.

Alberch J., E. Perez-Navarro, E. Arenas, and J. Marsal (1991) Involvementof nerve growth factor and its receptor in the regulation of thecholinergic function in aged rats. J. Neurochem. 57:1483–1487.

Armstrong, D.M., R. Sheffield, G. Buzsaki, K.S. Chen, L.B. Hersh, B.G.Nearing, and F.H. Gage (1993) Morphological alterations of cholineacetyltransferase-positive neurons in the basal forebrain of aged behav-iorally characterized Fischer 344 rats. Neurobiol. Aging 14:457–470.

Backman, C, P.T. Biddle, P.M. Ebendal, P.M. Friden, G.A. Gerhardt, M.A.Henry, L. Mackerlova, S. Soderstrom, I. Stromberg, L. Walus, B.J.Hoffer, and A.-C. Granholm (1995) Effects of transferrin receptorantibody-NGF conjugate on young and aged septal transplants in oculo.Exp. Neurol. 132:1–15.

Backman, C., G.M. Rose, B.J. Hoffer, M.A. Henry, R.T. Bartus, P. Friden,and A-C. Granholm (1996) Systemic administration of a nerve growthfactor conjugate reverses age-related cognitive dysfunction and pre-vents cholinergic neuron atrophy. J. Neurosci. 16:5437–5442.

Battleman, D.S, A.I. Geller, and M.V. Chao (1993) HSV-1 vector-mediatedgene transfer of the human nerve growth factor receptor p75hNGFRdefines high affinity NGF binding. J. Neurosci. 13:941–951.

Batchelor, P.E., D.M. Armstrong, S.N. Baker, and F.H. Gage (1989) Nervegrowth factor receptor and choline acetyltransferase colocalization inneurons within the basal forebrain: Response to fimbria fornix transec-tion. J. Comp. Neurol. 284:187–204.

Bothwell, M. (1991) Keeping track of neurotrophin receptors. Cell 65:915–918.

Bothwell M. (1995) Functional interactions of neurotrophins and neuro-trophin receptors. Annu. Rev. Neurosci. 18:223–253.

Chao M.V. (1992) Neurotrophin receptors: A window into neuronal differen-tiation. Neuron 9:583–593.

Chao, M.V., and B.L. Hempstead (1995) p75 and trk: A two-receptor system.TINS 18:321–326.

Chen, K.S., and F.H. Gage (1995) Somatic gene transfer of NGF to the agedbrain: Behavioral and morphological amelioration. J. Neurosci. 15:2819–2825.

Clary, D.O., G. Weskamp, L.R. Austin, and L.F. Reichardt (1994) Trk-Acrosslinking mimics neuronal responses to nerve growth factor. Mol.Biol. Cell. 5:549–563.

Cooper, J.D., D. Lindholm, and M.V. Sofroniew (1994) Reduced transport of(125I) nerve growth factor by cholinergic neurons and down-regulatedTrkA expression in the medial septum of aged rats. Neuroscience62:625–629.

Crutcher, K.A., S.A. Scott, S. Liang, W.V. Everson, and J. Weingartner(1993) Detection of NGF-like activity in human brain tissue: increasedlevels in Alzheimer’s disease. J. Neurosci. 13:2540–2550.

10 C. BACKMAN ET AL.

Curtis, M., M.A. Henry, B.J. Hoffer, R.T. Bartus, and A-C. Granholm (1995)Light and electron microscopic localization of OX-26-immunoreactivityin brain. Soc Neurosci. Abstract 21:549.

Ebendal, T. (1992) Function and evolution in the NGF family and itsreceptors. J. Neurosci. Res. 32:461–470.

Ehlers M.D., D.R. Kaplan, D.L. Price, and V.E. Koliatsos (1995) NGF-stimulated transport of trkA in the mammalian nervous system. J. CellBiol. 130:149–156.

Fisher, W., K. Wictorin, A. Bjorklund, L. R. Williams, S. Varon, and F. H.Gage (1987) Amelioration of cholinergic neuron atrophy and spatialmemory impairment in aged rats by NGF. Nature (London). 329:53–61.

Fischer, W., F.H. Gage, and A. Bjorklund (1989) Degenerative changes inforebrain cholinergic nuclei correlate with cognitive impairments inaged rats. Eur. J. Neurosci. 1:34–45.

Fischer, W., A. Bjorklund, K. Chen, and F.H. Gage (1991) NGF improvesspatial memory in aged rodents as a function af age. J. Neurosci.11:1889–1906.

Fisher, W., K.S. Chen, F.H. Gage, and A. Bjorklund (1992) Progressivedecline in spatial learning and integrity of forebrain cholinergic neu-rons in rats during aging. Neurobiol. Aging 13:9–23.

Fischer, W., A. Sirevaag, S.J. Wiegand, R.M. Lindsay, and A. Bjorklund(1994) Reversal of spatial memory impairments in aged rats by nervegrowth factor and neurotrophins 3 and 4/5 but not brain-derivedneurotrophic factor. Proc. Natl. Acad. Sci. U S A 91:8607–8611.

Friden, P.M., L.R. Walus, P. Watson, S.R. Doctrow, J.W. Kozarich, C.Backman, H. Bergman, B. Hoffer, F. Bloom, and A.-C. Granholm (1993)Blood-brain barrier penetration and in vivo activity of an NGF conju-gate. Science 259:373–377.

Friedman W.J., C.F. Ibanez, F. Hallbook, H. Persson, L.D. Cain, C.F.Dreyfus, and I.B. Black (1993) Differential actions of neurotrophins inthe locus coeruleus and basal forebrain. Exp. Neurol. 119:72–78.

Gilad, G.M., J.M. Rabey, Y. Tizabi, and V.H. Gilad (1987) Age-dependentloss and compensatory changes of septohippocampal cholinergic neu-rons in two rat strains differing in longevity and response to stress.Brain Res. 436:311–322.

Gnahn, H., F. Hefti, R. Hevermann, M. Schwab, and H. Thoenen (1983)NGF-mediated increase of choline acetyltransferase (ChAT) in theneonatal forebrain: Evidence for a physiological role of NGF in thebrain? Brain Res. 9:45–52.

Granholm, A.-C., C. Backman, F. Bloom, T. Ebendal, G.A. Gerhardt, B.Hoffer, L. Mackerlova, L. Olson, S. Soderstrom, L.R. Walus, and P.M.Friden (1994) NGF and anti-transferrin receptor antibody conjugate:Short and long term effects on survival of cholinergic neurons inintraocular septal transplants. J. Pharmacol. Exp. Ther. 268:448–459.

Hefti, F., and D.C. Mash (1989) Localization of nerve growth factorreceptors in the normal human brain and in Alzheimer’s disease.Neurobiol. Aging 10:75–87.

Hempstead, B.I., L.S. Schleiter, and M.V. Chao (1989) Expression offunctional nerve growth factor receptors after gene transfer. Science243:373–375.

Hepler, D.J., G.L. Wenk, B. Cribbs, D.S. Olton, and J.T. Coyle (1985) Memoryimpairments following basal forebrain lesions. Brain Res. 346:8–14.

Holtzman, D.M., Y. Li, L.F. Parada, S. Kinsman, C-K. Chen, J.S. Valleta, J.Zhou, J.B. Long, and W.C. Mobley (1992) p140trk mRNA marksNGF-responsive forebrain neurons: Evidence that trk gene expressionis induced by NGF. Neuron 9:465–478.

Kaplan, D.R., and R.M. Stephens (1994) Neurotrophin signal transductionby the trk receptor. J. Neurobiol. 25:1404–1417.

Klein, R., S. Jing, V. Nanduri, E. O’Rourke, and M. Barbacid (1991) The trkproto-oncogene encodes a receptor for nerve growth factor. Cell 65:189–197.

Koh, S., and R. Loy (1988) Age-related loss of NGF sensitivity in rat basalforebrain neurons. Brain Res. 440:396–401.

Koh, S., P. Chang, T.J. Collier, and R. Loy (1989) Loss of NGF receptorimmunoreactivity in the basal forebrain neurons of aged rats: correla-tion with spatial memory impairment. Brain Res. 498:397–404.

Kordower J.H., E.J. Mufson, A.-C. Granholm, B. Hoffer, and P.M. Friden(1993) Delivery of trophic factors to the primate brain. Exp. Neurol.124:21–30.

Kordower J.H., V. Charles, R. Bayer, R.T. Bartus, S. Putney, L.R. Walus,and P. Friden (1994) Intravenous administration of a transferrinreceptor antibody-nerve growth factor conjugate prevents the degenera-tion of cholinergic striatal neurons in a model of Huntington disease.Proc. Natl. Acad. Sci. U S A 91:9077–9080.

Kromer, L.F. (1987) Nerve growth factor treatment after brain injuryprevents neuronal death. Science 235:214–216.

Lapchak, P.A., D.M. Araujo, S. Carswell, and F. Hefti (1993) Distribution of(125I) nerve growth factor in the rat brain following a single intraven-tricular injection: Correlation with the topographical distribution oftrkA messenger RNA-expressing cells. Neuroscience 54:445–460.

Levi-Montalcini, R. (1982) Developmental neurobiology and the naturalhistory of nerve growth factor. Annu. Rev. Neurosci. 5:341–362.

Li, Y., D.M. Holtzman, L.F. Kromer, D.R. Kaplan, J. Chua-Couzens, D.O.Clary, B. Knusel, and W.C. Mobley (1995) Regulation of trkA and Chatexpression in developing rat basal forebrain: Evidence that bothexogenous and endogenous NGF regulate differentiation of cholinergicneurons. J. Neurosci. 15:2888–2905.

Ma, Y., R.B. Campenot, and F.D. Miller (1992) Concentration-dependentregulation of neuronal gene expression by nerve growth factor. J. CellBiol. 117:135–141.

Markowska, A.L., V.E. Koliastos, S.J. Breckler, D.L. Price, and D.S. Olton(1994) Human nerve growth factor improves spatial memory in agedbut not in young rats. J. Neurosci. 14:4815–4824.

Markram, H., and M. Segal (1990) Regional changes in NGF receptorimmunohistochemical labeling in the septum of the aged rat. Neurobiol.Aging 11:481–484.

Mearow, K.M., and Y. Kril (1995) Anti-NGF treatment blocks the upregula-tion of NGF receptor mRNA expression associated with collateral sproutingof rat dorsal root ganglion neurons. Neurosci. Letters 184:55–58.

Mesulam, M.-M., E.J. Mufson, B.H. Wainer, and A.I. Levey (1983) Centralcholinergic pathways in the rat: an overview based on an alternativenomenclature (Ch1–Ch6). Neuroscience 10:1185–1204.

Miyamoto, M., M.J. Kato, S. Narumi, and A. Nagoaka (1987) Characteris-tics of memory impairment following lesioning of the basal forebrainand medial septal nucleus in rats. Brain Res. 419:19–31.

Mobley, W.C., J.L. Rutkowski, G.I. Tennekon, K. Buchanan, and M.V.Johnston (1985) Choline acetyltransferase in striatum of neonatal ratsincreased by nerve growth factor. Science 229:284–287.

Morse, J.K., S.J. Wiegand, K. Anderson, Y. You, N. Cai, J. Carnahan, J.Miller, P.S. DiStefano, C.A. Altar, R.M. Lindsay, and R.F. Alderson(1993) Brain-derived neurotrophic factor (BDNF) prevents the degenera-tion of medial septal cholinergic neurons following fimbria transection.J. Neurosci. 13:4146–4156.

Mufson, E.J., J.M. Connor, J.S. Varon, and J.H. Kordower (1994) Nervegrowth factor immunoreactivity in the basal forebrain and hippocampus inprimates and Alzheimer’s disease. J. Comp. Neurol. 341:507–519.

Mufson, E.J., J.M. Conner, and J.H. Kordower (1995) Nerve growth factorin Alzheimer’s disease: Defective retrograde transport to nucleusbasalis. NeuroReport 6:1063–1066.

Parhad, L.M., J.N. Scott, L.A. Cellars, J.S. Bains, C.A. Krekoski, and A.W.Clark (1995) Axonal atrophy in aging is associated with a decline inneurofilament gene expression. J. Neurosci. Res. 41:355–366.

Ross, A.H., M.B. Lachyankar, D.K. Poluha, and R. Loy (1994) Axonal trans-port of the trkA high-affinity NGF receptor. Prog. Brain Res. 103:15–21.

Scali, C., F. Casamenti, M. Pazzagli, L. Bartollini, and G. Pepeu (1994)Nerve growth factor increases extracellular acetylcholine levels in theparietal cortex and hippocampus of aged rats and restores objectrecognition. Neurosci. Lett. 170:117–120.

Seiler, M., and M.E. Schwab (1984) Specific retrograde transport of nervegrowth factor receptor (NGF) from neocortex to nucleus basalis in therat. Brain Res. 300:33–39.

Schwab, M.E., U. Otten, Y. Agid, and H. Thoenen (1979) Nerve growthfactor (NGF) in the rat CNS: Absence of specific retrograde axonaltransport and tyrosine hydroxylase induction in locus coeruleus andsubstantia nigra. Brain Res. 168:473–483.

Sobreviela, T., D.O. Clary, L.F. Reichardt, M.M. Brandabur, J.H. Kordower,and E.J. Mufson (1994) TrkA-immunoreactive profiles in the centralnervous system: Colocalization with neurons containing p75 nervegrowth factor receptor, choline acetyltransferase, and serotonin. J.Comp. Neurol. 350:587–611.

Steininger, T.L., B.H. Wainer, R. Klein, M. Barbacid, and H.C. Palfrey(1993) High affinity nerve growth factor receptor (trk) immunoreactiv-ity is localized in cholinergic neurons of the basal forebrain andstriatum in the adult brain. Brain Res. 612:330–335.

Thoenen, H., and D. Edgard (1985) Neurotrophic factors. Science 229:238–242.

Verdi, J.M., S.J. Birren, C.F. Ibanez, H. Persson, D.F. Kaplan, M. Benedetti,M.V. Chao, and D.J. Anderson (1994) p75LNGFR regulates Trk signaltransduction and NGF-induced neuronal differentiation in MAH cells.Neuron 12:733–745.

Wainer, B.H., and M.-M. Mesulam (1990) Ascending cholinergic pathwaysin the rat brain. In: M. Steriade and D. Biesold (eds): Brain CholinergicSystems. New York: Oxford UP, pp. 65–119.

NGF AND ITS RECEPTORS DURING AGING 11