Increased Perivascular Norepinephrine Following Intracerebroventricular Infusion of NGF into Adult Rats

EXPERIMENTAL NEUROLOGY ARTICLE NO.

139, 54–60 (1996)

0080

Increased Perivascular Norepinephrine Following Intracerebroventricular Infusion of NGF into Adult Rats L. G. ISAACSON

AND

S. C. BILLIEU

Center for Neuroscience, Department of Zoology, Miami University, Oxford, Ohio 45056

by peripheral target tissues of the adult mouse observed following systemic administration of NGF (3) and the apparent growth response by aged perivascular nerves associated with the middle cerebral artery following the application of NGF to the vessel (1) provide evidence for axonal plasticity by mature sympathetic axons. In addition, mature sympathetic cell bodies in the superior cervical ganglion have been shown to alter dendritic morphology in response to the application of exogenous NGF to perivascular axons (2) as well as following systemic administration of NGF (23). We have previously observed that sympathetic cerebrovascular axons undergo a robust sprouting response following a 2-week intracerebroventricular infusion of NGF (12–15) and that the proliferation of the sprouted axons is dependent upon continued exposure to exogenous NGF (15). In contrast to the NGF-induced axonal growth by sympathetic perivascular axons, sensory perivascular axons associated with the extracerebral blood vessels respond to exogenous NGF by increasing neuropeptide content (14). The sympathetic response by perivascular axons associated with the intradural internal carotid artery (ICA) was first reported by Saffran and Crutcher (24) when, following NGF infusion, they observed an increase in perivascular catecholamine histofluorescence associated with this vessel. The use of the catecholamine histofluorescence technique, however, did not permit determination of whether the increased fluorescence represented an increase in the number of axons or an increase in the concentration of sympathetic neurotransmitter. Our subsequent studies using electron microscopy (EM) revealed that intracerebroventricular administration of NGF elicits a sprouting response by mature, uninjured sympathetic perivascular axons associated with the ICA (12, 13), suggesting that the increased histofluorescence observed following exogenous NGF was due, at least in part, to the growth of sympathetic perivascular axons associated with the ICA. However, in addition to an increase in the number of sympathetic axons, the increased perivascular histofluorescence could also result from increased catecholamine neurotransmitter within individual axons. In order to evaluate the perivascular sympathetic

In the present study, we used high performance liquid chromatography coupled with electrochemical detection to examine perivascular catecholamines associated with the intradural segment of the internal carotid artery following a 2-week in vivo intracerebroventricular infusion of the neurotrophin nerve growth factor (NGF). Following administration of NGF, a significant increase (87.3%) in perivascular norepinephrine (NE; mg/g) was observed when compared with vehicle-infused controls, suggesting that increased sympathetic neurotransmitter accompanies the NGFinduced sprouting response by sympathetic perivascular axons previously observed using electron microscopy (13, 15). The biochemical quantification of perivascular NE in the present study taken together with our previous morphological quantification of perivascular sprouts at the ultrastructural level reveal that the increase in NE is not proportional to the increase in the number of axons. Thus, when compared with controls, the relative amount of norepinephrine per axon apparently is reduced following NGF infusion. The apparent decrease in NE per axon following NGF infusion suggests that, during the 2-week infusion period, exogenous NGF did not stimulate the biosynthesis of perivascular NE beyond that necessary to accommodate the newly sprouted axons. These results extend our morphological findings and provide evidence for plasticity of neurotransmitter expression by adult sympathetic perivascular axons in vivo. In addition, we provide evidence that the increased perivascular catecholamine histofluorescence previously observed following NGF infusion results from an increase in the number of perivascular axons associated with the vessel rather than from an increase in the amount of NE within individual axons. r 1996 Academic Press, Inc.

INTRODUCTION

Results from a number of in vivo studies have shown that mature uninjured sympathetic neurons retain responsivity to neurotrophins such as nerve growth factor (NGF) and retain the capacity for continued growth and remodeling. Both the sympathetic response 0014-4886/96 $18.00 Copyright r 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.

54

NGF-INDUCED INCREASE IN PERIVASCULAR NOREPINEPHRINE

neurotransmitter response following a 2-week intracerebroventricular infusion of NGF, the present study was carried out using high performance liquid chromatography with electrochemical detection (HPLC-ECD) to identify and quantify perivascular catecholamines associated with the ICA and to document any changes in perivascular catecholamine concentrations that accompany the morphological sprouting response elicited by in vivo infusion of NGF. We report that perivascular NE levels associated with the ICA are significantly increased by infusion of NGF, providing evidence for neurotransmitter plasticity in the adult nervous system. Yet, when taken together with previous EM data, there is an apparent decrease in the relative amount of NE per axon following a 2-week NGF exposure, suggesting that NGF infusion does not stimulate the biosynthesis of perivascular NE except to accommodate the newly sprouted axons. Some of the findings reported in the present study have been published in abstract form (19). METHODS

Animals and Presentation of Data Young female adult Sprague–Dawley rats (6–8 weeks of age) were purchased from Harlan Laboratories (Indianapolis, IN) and maintained on 12-h light–dark cycle. Food and water were available ad libetum. Rats received a 14-day intracerebroventricular infusion of NGF (n 5 9) or cytochrome C (VEH; n 5 8) or received no infusion (n 5 8). Following HPLC-ECD analysis, catecholamine measurements were obtained from the right and left intradural internal carotid arteries (ICA), expressed as µg/g protein, and these two values were averaged to obtain a mean from each animal. The means obtained from each treatment were compared using ANOVA for ranked data (26). Values obtained from ELISA were subjected to one-way ANOVA. In both comparisons, differences among the groups were detected using Duncan’s multiple comparison test (26). In addition, a regression analysis was carried out to examine the relationship between cerebellar NGF values and perivascular NE values obtained from the ICA of the same animal. Osmotic Minipump Infusion Pump infusions were carried out using a modification of the protocol first introduced by Williams et al. (30) and described previously (12–15, 24, 25). Briefly, animals were anesthetized with an intramuscular injection of Ketamine (80 mg/kg): Rompun (14 mg/kg) mixture and placed in a stereotaxic frame. A 27-gauge cannula was lowered into the right lateral ventricle (1.0 mm lateral to bregma to 4.5 mm ventral to the brain surface) and cemented into place using dental acrylic.

55

Polypropylene tubing was used to connect the cannula to an osmotic minipump (Alzet 2002) located under the skin in the back. The pump contained approximately 200 µl of cytochrome C (Sigma; 100 µg/ml) or NGF (100 µg/ml) purified from male mouse submandibular gland. Because approximately 50 µl infusate was found in the pump at the end of the 2-week infusion period, it was estimated that 15 µg NGF or VEH (i.e., 1.07 µg/day) entered the ventricular system. Measurement of Catecholamines Following the 2-week infusion period, animals were sacrificed using a Harvard guillotine apparatus. The brain was rapidly removed from the skull and placed in ice-cold saline for a 5-min period to rinse arterial blood from the vessels. A 2-mm segment of the intradural ICA from each side was clipped from the base of the brain using dissecting scissors and immediately placed in a 1.5-ml polypropylene centrifuge tube. The entire tube was then dropped into liquid nitrogen to ‘‘snap freeze’’ the vessel. Individual vessels were stored at 280°C until analysis, at which time 125 µl mobile phase (see below) was added to the tube containing the vessel. The sample was sonicated (Kontes model KT-50) for 20 s and then centrifuged for 5 min at 4°C in an Eppendorf microfuge 5415C. The supernatant was collected for HPLC-ECD injection and the pellet was resolubilized in 100 µl NaOH, frozen at 280°C, and reserved for quantification of protein. An 80-µl aliquot of supernatant was injected into a C-18 reverse phase analytical column (3 µm, 100 3 3.2 mm, Phase-II ODS, Bioanalytical Systems) via a Waters refrigerated WISP (Model 712). The analytical column was protected by a precolumn filter. Catecholamines were measured using an electrochemical detector with glassy carbon (Model LC-4B, Bioanalytical Systems) set at 10.80V relative to an internal Ag/AgCl reference electrode. The mobile phase, delivered at 1.0 ml/min, consisted of 1.0% methanol, 0.21 mM EDTA, 50 mM NaH2PO4, and 2.47 mM heptane sulfonic acid, pH 5 3.6 (adjusted with with phosphoric acid). All detection was integrated using Maxima 810 software installed on a Gateway 386/33C computer. Measurements were calculated from standard curves generated with known amounts of catecholamines run at the same time as the samples. Recovery ranged from 70 to 90% and was calculated from samples to which a known amount of DHBA (750 pg) had been added. The lower limit of sensitivity was approximately 10 pg for norepinephrine (NE) and the NE metabolite dihydroxyphenylglycol (DOPEG) and was 25 pg for the dopamine metabolite dihydroxyphenylacetic acid (DOPAC). In instances where DOPAC detection was below lower limit of sensitivity, the value was expressed as one-half the lower limit of detection. The protein content of each vessel was determined

56

ISAACSON AND BILLIEU

using the micro Lowry technique (17). Briefly, a 500-µl aliquot of 1.0% copper sulfate was added to the protein pellet recovered following centrifugation. After 20 min, 500 µl of 0.2 N Folin reagent (Sigma) was added. Sixty minutes later, samples were read using a spectrophotometer set at 578 nm and compared to known concentrations of bovine serum albumin (Sigma). Successful Delivery of NGF The success of the pump placement and infusion was monitored by the addition of 0.1% solution of the fluorescent dye bisbenzimide to the infusate as described previously (12–15, 24, 25). The number of animals per treatment stated above represents the actual number of animals from which HPLC-ECD data were collected and reflects the number of successful infusions as determined by examination of BIS labeling of the right lateral and third ventricles. As an additional means to monitor the success of NGF entry into the ventricular system following chronic pump implantation, cerebellar tissue from selected noninfused controls (n 5 5), VEH-infused (n 5 4), and NGF-infused (n 5 6) animals were processed using ELISA techniques as previously described (24, 25). NGF measurements from cerebellar samples were calculated from standard curves generated with known dilutions of NGF.

FIG. 1. Comparison of mean (6SEM) concentration of perivascular catecholamines observed in association with the intradural segment of the ICA from rats receiving a 2-week infusion of NGF (NGF) or vehicle (VEH) into the lateral ventricle or receiving no infusion (CONT). Perivascular norepinephrine (NE) was significantly increased in the NGF-infused treatment group when compared with VEH and CONT groups (P , 0.05). No significant changes were observed among the three treatments in the concentration of the NE metabolite dihydroxyphenylglycol (DOPEG) or the dopamine metabolite dihydroxyphenylacetic acid (DOPAC).

RESULTS

Perivascular Catecholamines Associated with ICA Using the present chromatographic methods, measurable concentrations of norepinephrine (NE), the NE metabolite dihydroxyphenylglycol (DOPEG), and the dopamine metabolite dihydroxyphenylacetic acid (DOPAC) were routinely found in association with individual samples of the extracerebral ICA and these three compounds are the focus of the present analysis. Dopamine was detected at measurable levels in approximately 8% of the vessels. In control animals, NE and DOPEG were present at measurable concentrations in every sample of ICA examined (Fig. 1). The mean value for norepinephrine, the predominant catecholamine associated with the ICA, was 33.9 6 3.5 µg/g protein (Fig. 1), whereas the mean concentration for DOPEG was 22.0 6 5.7. In addition, a detectable concentration of DOPAC was observed in 50% of these samples and, when present, measured a mean of 5.5 6 2.5 µg/g. Because chronic VEH pump implantation was believed to elicit a perivascular axonal response in a recent study (1), the values for perivascular catecholamines from VEH and CONT groups were listed separately for comparison (Fig. 1) and were analyzed separately along with the NGF-infused treatment group.

The mean values obtained for perivascular NE and DOPEG following VEH infusion were not significantly different from that obtained from noninfused controls (Fig. 1). DOPAC was present in 33% of the vessel samples from the VEH-infused treatment and the mean concentration was not significantly altered by VEH infusion. The 2-week intracerebroventricular infusion of NGF resulted in a significant increase in perivascular NE associated with the ICA when compared with both noninfused and VEH treatment groups (P , 0.05). Following infusion of NGF, the mean value for perivascular NE was increased by 87.3% when compared with the VEH group (Fig. 1). The concentration of DOPAC could be measured in 17% of the samples from NGFinfused animals. The mean values for DOPEG or DOPAC in the NGF-infused group were not significantly different when compared with VEH or unoperated controls (Fig. 1). Measurement of NGF in Cerebellar Tissues The mean value for cerebellar NGF from NGFinfused cases was increased 76 and 31% when compared with NGF values from the noninfused and VEH treatment groups, respectively (Table 1). The increase in NGF was not found to be statistically significant at

NGF-INDUCED INCREASE IN PERIVASCULAR NOREPINEPHRINE

TABLE 1 TRT

NGF (ng/g)

NE (µg/g)

CONT VEH NGF

0.29 6 .07 0.39 6 .03 0.51 6 .10

28.0 6 1.7 44.5 6 5.5 79.5 6 12.3*

Note. Mean cerebellar NGF 6 SEM (ng/g) obtained from ELISA and the mean perivascular NE 6 SEM (µg/g protein) obtained from HPLC-ECD analysis of vessels from the same animals following a 2-week infusion of NGF or vehicle (VEH) into the lateral ventricle or following no infusion (CONT). No significant differences were observed in NGF levels among the three treatment groups (see text for details), whereas a significant increase in perivascular NE was observed in the NGF-infused treatment group (P , 0.05) when compared with VEH and CONT groups. Note that the mean perivascular NE values differ from those in Fig. 1 because NE measurements reported in this table reflect only the perivascular NE measurements from cases in which cerebellar NGF also was measured.

P , 0.05 due to high variability within the NGF-infused group. For example, the cerebellar tissues from two of the six NGF-infused cases exhibited no increase in NGF, though, in both of these cases, we observed proper pump placement, BIS labeling of the ventricular system (see methods), and a significant elevation in perivascular NE associated with the ICA. We conclude that these two cases received successful NGF infusion because when proper pump placement and BIS labeling of the ventricular system following NGF infusion were observed in the past, increased perivascular catecholamine histofluorescence at the light microscopic level was observed (24, 25, Isaacson, personal observation) and a morphological sprouting response at the electron microscopic level (12, 13, 15) as well as increased immunoreactivity using antibodies to DBH (unpublished observation) and CGRP (14) have been documented. Nevertheless, a significant and direct relationship between the level of NGF as determined by ELISA and NE concentration as determined using HPLC-ECD was observed (P , 0.05; r 2 5 0.37) (Table 1). Though the relationship between NGF and NE was significant, the relatively low correlation factor likely reflects the variability in NGF measurements within the NGFinfused treatment. DISCUSSION

In previous studies, we observed that a 2-week intracerebroventricular infusion of NGF elicits growth of perivascular sympathetic axons (13) and that these sprouted axons proliferate and disappear as a function of exogenous NGF exposure (15). Our present findings suggest that changes in perivascular NE accompany the neuronal growth and that perivascular NE is increased by intracranial infusion of NGF. These results are consistent with a response by mature sympathetic axons in vivo following exogenous NGF (3, 13, 23–25).

57

The biochemical quantification of perivascular catecholamines together with our previous quantitative morphological findings at the ultrastructural level (12, 13, 15) provide the opportunity to integrate our biochemical and morphological data regarding the NGFinduced sympathetic response. Following quantification at the electron microscopic level, the mean number of perivascular axons associated with the intradural ICA from noninfused and VEH-infused groups was 629 and 711 axons, respectively (15). Following a 2-week intracerebroventricular infusion of NGF, the number of perivascular axons was significantly increased to 1827 axons (15). When these morphological data are taken together with the biochemical measurements obtained from HPLC analysis, a ratio of the relative amount of NE per axon can be calculated and indices of 0.054 and 0.053 in the noninfused and VEH groups, respectively, are obtained. Following in vivo administration of NGF, this index is calculated as 0.039, suggesting a decrease in the relative amount of NE per axon during the 2-week infusion period. The apparent reduction in NE per axon in the sprouted cases may reflect the immature stage of development of the newly formed axonal sprouts. Using electron microscopy, we observe that the newly sprouted axons are strikingly similar to the normal appearance of peripheral axons during development (22) and it has been reported that neuronal growth typically precedes the expression of transmitter phenotype (5). In the present study, though exogenous NGF stimulated axonal growth, the axonal sprouts formed during the 2-week infusion period may not possess neurotransmitter levels comparable to mature sympathetic axons. Our findings confirm previous histochemical studies that suggested a sympathetic response by the mature nervous system following in vivo exposure to exogenous NGF. For example, Bjerre et al. (3) observed an increase in catecholamine histofluorescence associated with peripheral organs following NGF administration to the mouse. In addition, a dramatic increase in catecholamine histofluorescence by perivascular axons associated with the extracerebral blood vessels was noted following intracerebroventricular infusion of NGF (24, 25). Though increased perivascular fluorescence was observed following NGF infusion, the use of catecholamine histofluorescence did not allow for the determination of whether the increased fluorescence represented an increase in the number of perivascular axons or a relative increase in the concentration of neurotransmitter per axon. The present results suggest that the increased perivascular catecholamine histofluorescence observed previously following NGF infusion may not result from an increase in the amount of NE within individual axons but instead results primarily from an increase in the number of perivascular axons.

58

ISAACSON AND BILLIEU

Neuronal Activity by Sympathetic Perivascular Axons Measurements of perivascular NE associated with the ICA were fairly consistent in both the noninfused control and VEH groups and these consistent measurements may reflect a steady state measurement of perivascular NE for the ICA in the young adult. In other parts of the nervous system, neurotransmitter content is a stable property of a given group of neurons and can be used as an index of the number of neuronal elements present in a given tissue (10). Within a region, any change in the constant level of neurotransmitter is interpreted by some researchers as a change in axonal number (28). The finding that exogenous NGF does not increase the amount of sympathetic neurotransmitter per axon may reflect a steady state phenomenon associated with the ICA and our evidence that the increased sympathetic neurotransmitter is primarily a consequence of axonal growth is consistent with the interpretations of other researchers using other neuronal systems (10, 28). Because of the steady state of many neurotransmitters within neuronal systems, the concentration of a given transmitter generally has little value as a measure of transmitter dynamics. The examination of metabolites (i.e., the NE metabolite DOPEG), however, can provide information concerning neurotransmitter release and/or neuronal activity. Monoamine metabolites apparently result from neurotransmitter reuptake and inactivation following release and frequently are used as indicators of neuronal activity (10, 28). In the present study, no change in perivascular DOPEG was observed following NGF infusion though perivascular NE was significantly increased, suggesting that the infusion of NGF over a 2-week period did not alter NE release by perivascular axons. The finding that exogenous NGF infusion did not alter neurotransmitter release provides additional evidence that the newly formed axonal sprouts are indeed in an immature stage of development. However, the issue of whether neurotransmitter release is altered by exogenous NGF will not be resolved until direct measurements of NE turnover are carried out. NGF Measurements Using Cerebellar Tissues Following ELISA, the increase in cerebellar NGF in the NGF-infused group was expected, yet two of the six NGF-infused cases (these values were included in the mean) did not exhibit elevated NGF levels even though our light microscopic observations provided evidence of NGF entry into the ventricular system and elevated perivascular NE was associated with the ICA. It is difficult to explain why NGF levels would not be elevated in every case where we observe BIS labeling of the ventricular system. Previous studies using a similar infusion paradigm have found elevated NGF in

association with the cerebellum following NGF pump implantation (24, 25). In our procedure we use a smaller diameter cannula (27- vs 25-gauge needle) connected to the osmotic pump. The use of the smaller cannula is advantageous because it results in minimal tissue damage but it also results in a slower rate of infusion and thus allows for greater penetration of infusate into the perivascular spaces associated with the intraparenchymal vessels, an alternate route for substances delivered to the ventricular system (21). It is possible that in some instances (i.e., the two NGFinfused cases in the present study), a large proportion of infusate reached the extracerebral arteries via the perivascular space and bypassed the fourth ventricle and the cerebellar cisternae. Also, the brief saline rinse prior to vessel removal (see methods) may result in less NGF associated with the cerebellar tissues. Effect of Pump Implantation on Perivascular NE In the present study, the infusion of cytochrome C had no significant effect on the concentration of perivascular catecholamines associated with the ICA of the young adult. This is in agreement with previous findings of Andrews and Cowen (1) where VEH pump implantation did not affect perivascular density in the young adult. However, in the same study, a 2-week infusion of cytochrome C onto the middle cerebral artery of the aged rat resulted in an increased density of immunopositive fibers when an antibody directed against the general neuronal marker PGP9.5 was utilized (1). This VEH-induced response in the aged rat was attributed to a possible immune or inflammatory reaction induced by invasive surgical procedures. Using HPLC methods, we did not observe any changes in perivascular NE that could be attributed to VEH pump implantation. Dopamine Associated with Extracerebral Blood Vessels To our knowledge, the present findings from noninfused control vessels represent the first report of perivascular catecholamines associated with a single vessel of the extracerebral vasculature. Individual segments of ICA were examined in the present analysis and the dopamine that we observed in our samples was typically below our detection limits. Though we did not observe detectable levels of dopamine in our samples, others have reported that dopamine is associated with the innervation of the peripheral vasculature. Head et al. (11) described a relatively large proportion (40%) of dopamine per total catecholamines in association with the intraparenchymal arterioles and venules. In addition, there are reports of dopaminergic innervation of cerebral arterioles in local brain regions that utilize dopamine as a neurotransmitter (8, 9) and of the renal blood vessels (6). However, evidence of a dopaminergic

NGF-INDUCED INCREASE IN PERIVASCULAR NOREPINEPHRINE

innervation of the extracerebral blood vessels has not been substantiated (7) and our findings suggest that any perivascular dopamine associated with the intradural ICA is present solely for the biosynthesis of NE.

REFERENCES 1.

2.

Conclusions Our results suggest that, following intracranial NGF infusion, an increase in sympathetic neurotransmitter concentration accompanies the axonal growth previously observed at the ultrastructural level. The finding that perivascular NE is increased following exposure to exogenous NGF provides evidence for neurotransmitter plasticity within the adult nervous system. Yet, the apparent decrease in the relative amount of NE per axon during a period of robust axonal growth suggests that the infusion of NGF does not stimulate the biosynthesis of perivascular NE during the 2-week infusion period beyond that necessary to accommodate the newly formed axons. These findings imply that the sprouted axons, still in an immature stage of development, do not express a neurotransmitter phenotype similar to fully developed preexisting axons. Though our findings cannot provide precise information regarding the release or activity of the sympathetic axons following exogenous NGF, the finding that DOPEG concentrations were unchanged whereas NE was significantly increased suggests that NGF infusion did not result in increased activity by sympathetic perivascular axons associated with the intradural ICA. These results extend our previously obtained morphological findings and are consistent with the accumulating evidence for neuronal plasticity in the adult and the ability of uninjured neurons to undergo continued remodeling and growth in the adult. The observed neuronal plasticity following delivery of exogenous growth factors to the brain may be relevant due to the continued interest in the use of exogenous NGF as a therapeutic agent for the treatment of Alzheimer’s or other neurodegenerative disorders (4, 16, 18, 20, 27, 29). It will be important to examine in more detail the specificity and functional consequences of the sympathetic response to NGF infusion.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

ACKNOWLEDGMENTS 15. The authors thank Phyllis Callahan for her help in data presentation and interpretation, to Jim Oris for his assistance in HPLC-ECD methodology and technique and statistical analyses, and to John Bailer who kindly provided statistical advice. Lesley Hickman provided technical assistance. We thank Keith A. Crutcher for supplying the NGF used in this study, for the ELISA analysis of cerebellar samples, and for his helpful comments on an early draft of this manuscript. This work was supported by National Institutes of Health NS 17131 (K.A.C.) and NS 32876 (L.G.I.) and OBOR Research Challenge Program (L.G.I.).

59

16.

17.

ANDREWS, T. J., AND T. COWEN. 1994. In vivo infusion of NGF induces the organotypic regrowth of perivascular nerves following their atrophy in aged rats. J. Neurosci. 14: 3048–3058. ANDREWS, T. J., AND T. COWEN. 1994. Nerve growth factor enhances the dendritic arborization of sympathetic ganglion cells undergoing atrophy in aged rats. J. Neurocytol. 23: 234– 241. BJERRE, B., A. BJORKLUND, W. MOBLEY, AND E. ROSENGREN. 1975. Short- and long-term effects of nerve growth factor on the sympathetic nervous system in the adult mouse. Brain. Res. 94: 263–277. BUCKHOLTZ, N. S. 1994. Alzheimer’s Disease drug development and testing at National Institute on Aging. Psychopharmacol. Bull. 30: 15–17. COCHARD, P., M. GOLDSTEIN, AND I. B. BLACK. 1979. Initial development of the noradrenergic phenotype in autonomic neuroblasts of the rat embryo in vivo. Dev. Biol. 71: 100–114. DINERSTEIN, R. J., J. VANNICE, R. C. HENDERSON, L. J. ROTH, L. I. GOLDBERG, AND P. C. HOFFMANN. 1979. Histofluorescence techniques provide evidence for dopamine containing neuron elements in canine kidney. Science 205: 497–499. EDVINSSON, L., E. T. MACKENZIE, AND J. MCCULLOCH, EDS. 1993. Cerebral Blood Flow and Metabolism, pp. 231–252. Raven Press, New York. FAVARD, C., A. SIMON, AND J. NGUYEN-LEGROS. 1989. Close relationship between dopaminergic varicosities and small blood vessel walls in rat and monkey retina. In Neurotransmission and Cerebrovascular Function (J. Seylaz and E. T. MacKenzie, Eds.), pp. 159–162. Elsevier, Amsterdam. FELTEN, D. L., AND K. A. CRUTCHER. 1979. Neuronal-vascular relationships in the raphe nuclei, locus coeruleus and substantia nigra in primates. Am. J. Anat. 155: 467–482. FINLAY, J. M., AND M. J. ZIGMOND. 1995. A critical analysis of neurochemical methods for monitoring transmitter dynamics in the brain. In Psychopharmacology: The Fourth Generation of Progress (F. E. Bloom and D. J. Kupfer, Eds.), pp. 29–40. Raven Press, New York. HEAD, R. J., J. T. HJELLE, B. JARROTT, B. BERKOWITZ, G. CARDINALE AND S. SPECTOR. 1980. Isolated brain microvessels: Preparation, morphology, histamine, and catecholamine contents. Blood Vessels 17: 173–186. ISAACSON, L. G., AND K. A. CRUTCHER. 1990. Intracerebral infusion of NGF induces hyperinnervation of cerebral blood vessels. Neurobiol. Aging 11: 51–55. ISAACSON, L. G., B. N. SAFFRAN, AND K. A. CRUTCHER. 1992. Nerve growth factor-induced sprouting of mature, uninjured sympathetic axons. J. Comp. Neurol. 326: 327–336. ISAACSON, L. G., D. ONDRIS, AND K. A. CRUTCHER. 1995. The plasticity of sensory cerebrovascular axons following intracranial infusion of nerve growth factor. J. Comp. Neurol. 361: 451–460. ISAACSON, L. G., AND K. A. CRUTCHER. 1995. The duration of sprouted cerebrovascular axons following intracranial infusion of nerve growth factor. Exp. Neurol. 131: 174–179. LAPCHAK, P., AND F. HEFTI. 1992. Pharmacological actions of recombinant human nerve growth factor on lesioned central cholinergic neurons: Rationale for the treatment of Alzheimer’s disease. In Cholinergic Basis for Alzheimer’s Therapy. (R. Becker and E. Giacobini, Eds.), pp. 397–403. Birkhauser. LOWRY, O. H., N. J. ROSEBROUGH, A. L. FARR, AND R. J. RANDALL. 1951. Protein measurement with the folin phenol reagent. J. Biol. Chem. 193: 265–275.

60 18. 19.

20.

21.

22. 23.

24.

ISAACSON AND BILLIEU OLSON, L. 1993. NGF and the treatment of Alzheimer’s disease. Exp. Neurol. 124: 5–15. ONDRIS, D., C. BILLIEU, K. SCHWENK, S. PAGE, AND L. G. ISAACSON. 1994. Nerve Growth Factor-induced plasticity of cerebrovascular innervation. Soc. Neurosci. Abstr. 20: 1093. PHELPS, C. H., F. H. GAGE, J. H. GROWDON, F. HEFTI, R. HARBAUGH, M. V. JOHNSTON, Z. S. KHACHATURIAN, W. C. MOBLEY, D. PRICE, M. RASKIND, J. SIMPKINS, L. THAL, AND J. WOODCOCK. 1989. Potential use of nerve growth factor to treat Alzheimer’s disease. Neurobiol. Aging 10: 205–207. RENNELS, M. L., T. F. GREGORY, O. R. BLAUMANIS, K. FUJIMOTO, AND P. A. GRADY. 1985. Evidence for a ‘perivascular’ fluid circulation in the mammalian central nervous system, provided by the rapid distribution of tracer protein throughout the brain from subarachnoid space. Brain Res. 326: 47–63. RUBIN, E. 1985. Development of the rat superior cervical ganglion: Ganglion cell maturation. J. Neurosci. 5: 673–684. RUIT, K. G., P. A. OSBORNE, R. E. SCHMIDT, E. M. JOHNSON, JR., AND W. D. SNIDER. 1990. Nerve growth factor regulates sympathetic ganglion cell morphology and survival in the adult mouse. J. Neurosci. 10: 2412–2419. SAFFRAN, B. N., J. E. WOO, W. C. MOBLEY, AND K. A. CRUTCHER. 1989. Intraventricular NGF infusion in the mature rat brain enhances sympathetic innervation of cerebrovascular targets

25.

26. 27.

28.

29. 30.

but fails to elicit sympathetic ingrowth. Brain Res. 492: 245– 254. SAFFRAN, B. N., AND K. A. CRUTCHER. 1990. NGF-induced remodeling of uninjured mature axon collaterals. Brain Res. 525: 11–20. SAS INSTITUTE, INC. 1985. SAS User’s Guide: Statistics; 5th Edition. 956 pp. Cary, N.C.: SAS Institute, Inc. SEIGER, A., A. NORDBERG, H. VON HOLST, L. BACKMAN, T. EBENDAL, I. ALAFUZOFF, K. AMBERLA, P. HARTVIG, A. HERLITZ, A. LILJA, H. LUNDQVIST, B. LANGSTROM, B. MEYERSON, A. PERSSON, M. VIITANEN, B. WINBLAD, AND L. OLSON. 1993. Intracranial infusion of purified nerve growth factor to an Alzheimer patient: The first attempt of a possible future treatment strategy. Behav. Brain Res. 57: 255–261. TIAN, Y., K. J. LOOKINGLAND, AND K. E. MOORE. 1991. Contribution of noradrenergic neurons to 3,4-dihydroxyphenylacetic acid concentrations in the regions of the rat brain containing incertohypothalamic dopaminergic neurons. Brain Res. 555: 135–140. TUSZYNSKI, M. H., AND F. H. GAGE. 1994. Neurotrophic factor and diseases of the nervous system. Annals Neurol. Exp. 35: S9–S12. WILLIAMS, L. R., H. L. VAHLSING, T. LINDAMOOD, S. VARON, F. H. GAGE, AND M. MANTHORPE. 1987. A small-gauge cannula device for continuous infusion of exogenous agents into the brain. Exp. Neurol. 95: 743–754.