Intraventricular infusion of nerve growth factor as the cause of sympathetic fiber sprouting in sensory ganglia

J Neurosurg 91:447 453, 1999 Intraventricular infusion of nerve growth factor as the cause of sympathetic fiber sprouting in sensory ganglia HARING J. W. NAUTA, M.D., JOSEPH C. WEHMAN, B.S., VASSILIS E. KOLIATSOS, M.D., MARYLEE A. TERRELL, B.A., AND KYUNGSOON CHUNG, PH.D. Division of Neurosurgery, Department of Anatomy and Neurosciences, and Marine Biomedical Institute, University of Texas Medical Branch, Galveston, Texas; and Departments of Pathology, Neurology, Neuroscience, and Psychiatry and Behavioral Science, The Johns Hopkins University School of Medicine, Baltimore, Maryland Object. The results of previous clinical trials have indicated that intraventricular infusion of nerve growth factor (NGF) in patients with Alzheimer s disease is frustrated by the appearance of weight loss and diffuse back pain. The present study tested whether NGF induces sympathetic sprouting in sensory ganglia. Such sprouting has been implicated in previous studies as a possible mechanism of sympathetically maintained pain in neuropathic animals. Methods. Nineteen Long Evans rats underwent intraventricular infusion of either artificial cerebrospinal fluid (ACSF; seven animals) or NGF (12 animals). After 14 days of infusion, the sensory ganglia of the trigeminal nerve and the C-2, C-8, T-1, L-4, and L-5 dorsal roots were examined for sympathetic sprouting by using tyrosine hydroxylase immunohistochemical analysis. Conclusions. In the animals receiving NGF, 52 of 144 ganglia showed sympathetic fiber sprouting. In the control animals receiving ACSF, only two of 72 ganglia showed minor sympathetic fiber sprouting. A preferential sprouting of sympathetic fibers was demonstrated at lower lumbar ganglia compared with the cervical and thoracic ganglia. The data presented here demonstrate that in the rat intraventricular NGF infusion caused sympathetic sprouting in dorsal root ganglia (p 0.01). These findings may have importance both for the treatment of Alzheimer s disease and the understanding of neuropathic pain. KEY WORDS nerve growth factor Alzheimer s disease therapy pain N ERVE growth factor (NGF), a neurotrophic factor acting on basal forebrain cholinergic neurons, has been proposed as a possible therapy for Alzheimer s disease. 11,16,32 Such a consideration is based on observations that these neurons degenerate early and consistently in patients with this disease. 43 Basal forebrain cholinergic neurons are known to express NGF receptors, 10 and their survival and phenotypic maintenance have been shown to be influenced by the availability of NGF. 7,9, 17,18,44 A clinical trial in which patients with Alzheimer s disease were given an intraventricular infusion of NGF, however, was frustrated by the occurrence of peripheral pain and weight loss in some patients. 27,28,30 On the other hand, results obtained when using a prolonged intraputaminal infusion of NGF to support adrenal medullary autografts indicated a more pronounced and long-lasting effect on motor behaviors in a patient with Parkinson s disease 29 than no NGF infusion. 22 In this case of intraputaminal NGF administration, there were no reported side effects. Other lines of evidence suggest that NGF may be involved in the development of neuropathic pain, possibly through the mechanism of sympathetic fiber sprouting into sensory ganglia. Clinical experience in humans has shown that a fraction of patients with neuropathic pain experience relief after receiving a pharmacological or surgical sympathetic nerve block. 1,6,12,23,34,42 In animal models of peripheral neuropathy, a peripheral nerve injury induces pain behaviors, 2,14,39 and those pain behaviors can be alleviated by sympathectomy or adrenoreceptor antagonists. 15,25,41 Extensive sympathetic fiber sprouting has been observed in the sensory ganglia of injured nerves 3,4,24,35 and sympathectomy removes those sympathetic fibers from the sensory ganglia. 4 It has been postulated that abnormal sympathetic fiber sprouting in sensory ganglia following peripheral nerve injury is induced by local synthesis of NGF in the sensory ganglia. 38,40 It is not yet clear whether the sympathetic fiber sprouting is an essential step in pain generation or whether it is a coincidental epiphenomenon independent of a more direct pain mechanism linked to NGF 20,21 or other mediators. 8,19,46 However, the consistent association of the development of pain with both sympathetic fiber sprouting in sensory ganglia and local NGF production make sympathetic fiber sprouting a candidate for further study. The hypothesis, tested here in a rat model, was that the delivery of NGF to the ventricular cerebrospinal fluid would induce sympathetic fiber sprouting in sensory ganglia in the absence of peripheral nerve injury. Using an implanted osmotic pump as a delivery system for intra- 447

H. J. W. Nauta, et al. FIG. 1. Photomicrographs showing TH-immunostained lumbar DRGs of rats that received ACSF infusion. In the DRGs of ACSF-infused rats, TH-ir profiles are limited to: 1) fibers that innervate the blood vessels (precapillary arterioles) within the DRG (arrow in A); 2) fibers that supply the blood vessels on the capsule of the DRG (arrows in B); and 3) fibers of the spinal nerve penetrating various distances into the DRG fiber tract from the gray rami communicantes before looping outward and following the spinal nerve (arrows in C). The fourth TH-ir profiles in the DRG are a small number of dopaminergic DRG neurons (arrowhead in B) and their processes. Bars = 50 m. ventricular administration of the neurotrophic factor NGF, the present study demonstrated sympathetic fiber sprouting in sensory ganglia along the neuraxes of the animals. We can infer from the data that a similar sympathetic fiber sprouting may underlie the pain syndrome observed in patients with Alzheimer s disease who are treated with NGF. Materials and Methods Animal Preparation Experiments were performed on adult male Long Evans Hooded rats (250 300 g each) with the approval of the Institutional Animal Care and Use Committee at the University of Texas Medical Branch. Efforts were made to minimize the number of animals used and to minimize animal discomfort. The rats were housed two in a cage with free access to food and water in a 12-hour light/dark cycle. The rats were acclimated for 4 days and randomly assigned into two treatment groups. Surgical Procedure Anesthesia was first induced in the animals by administration of ketamine (60 mg/kg given intramuscularly) and Nembutal (40 mg/ kg given intraperitoneally), after which the rats were each mounted in a small-animal stereotactic unit. The stereotactic locations were modified based on the atlas of Paxinos and Watson. 31 The skull was exposed by performing a midline incision and the lambda and bregma were located as well as the target coordinates (from the bregma: anteroposterior 1 mm; lateral 1.5 mm; height 3.5 mm). A small craniotomy opening was made, through which a cannula (blunted 25-gauge stainless-steel butterfly needle with a right-angle bend located 7 mm from the tip) was then introduced stereotactically into the right ventricle. The base of the cannula was affixed to the skull by using dental cement before releasing the rat from the stereotactic instrument. This cannula was connected to a miniosmotic pump by a short length of vinyl tubing. Each pump was filled with either artificial cerebrospinal fluid (ACSF) alone (controls; seven animals) or recombinant human NGF (12 animals), from one of two sources, in ACSF and primed just before surgery. The ACSF was composed of 3.35 mm KCl, 1.26 mm CaCl 2, 0.59 mm NaH 2 PO 4 H 2 O, 1.16 mm MgCl 2 6H 2 O, 2.16 mm urea, 0.14 M NaCl, 3.37 mm dextrose, and 0.02 M NaHCO 3. The amount of dissolved NGF was matched with the pumping rate to deliver dosages of 20, 50, or 100 g/week during the 2-week infusion period. After mounting each cannula, the pump and tubing were passed into a subcutaneous pocket over the upper thoracic spine and the scalp incision was closed to cover the cannula and cement. The animals were then returned to their cages where they received Cefazolin (25 mg/kg administered intramuscularly) twice per day for 3 days. Histological Study After 14 days of infusion, the animals were anesthetized using sodium pentobarbital (70 mg/kg administered intraperitoneally) and perfused with saline followed by cold fixative containing 4% paraformaldehyde and 0.1% picric acid in 0.1 M phosphate buffer, ph 7.2. The C-2, C-8, T-1, L-4, and L-5 dorsal root ganglia (DRGs) and the trigeminal nerve ganglia were removed bilaterally with the aid of a dissecting microscope and stored in perfusion fixative for 2 to 4 hours. Tissues were then transferred to a 30% sucrose solution in 0.1 M phosphate buffer until equilibrated. Each ganglion was sectioned on a freezing microtome (30 m) and all sections (20 45 sections/ganglion) were processed for tyrosine hydroxylase (TH) immunohistochemical analysis for identification of sympathetic postganglionic fibers. In brief, the tissues were incubated in primary antibodies for TH at a dilution of 1:1200 for 2 to 3 days at 4 C, which was followed by a reaction with biotinylated second antibody and, finally, with avidin-biotin complex reagent. Triton X-100 (0.3%) was used in all reaction solutions to enhance antibody penetration. The immunoreactivity was visualized using the diaminobenzidine method in the presence of H 2 O 2 in 0.1 M phosphate buf- 448

Nerve growth factor and sprouting fer. Tissues were mounted on gelatin-coated slides, air dried, counterstained with neutral red dye, coverslipped, and observed with the aid of a light microscope. The brains were also sectioned on a freezing microtome to verify cannula placement. Anatomical and Statistical Analysis Sympathetic fiber sprouting in the DRG was determined on the basis of TH immunoreactivity and morphological characteristics. The determination of sympathetic fiber sprouting for each ganglion was based on examination of the entire series of serial sections through that ganglion. Any indication of sympathetic fiber sprouting was marked as positive (+), whereas no sprouting was marked as zero (0). Blinded analysis was performed throughout the experiments to avoid biased judgment. One person planted the osmotic pump with a specific dose of NGF or ACSF and randomly assigned a number to each animal. After this, the animal, which was identified only by number, was first given to the person who performed tissue sampling and immunostaining and later to the person who analyzed the slides. After the slide reading was completed, the code was broken and the animal number was matched with the specific agent administered. The statistical analysis of data, the number of animals showing sympathetic fiber sprouting in the DRG, was performed using Fisher s exact test. Sources of Supplies and Equipment The Long Evans Hooded rats were obtained from Harlan Sprague Dawley Co., Houston, TX, and the small-animal stereotactic unit to which they were mounted was purchased from David Kopf Instruments, Tujunga, CA. The Alzet model 2ML4 miniosmotic pump was purchased from Alza Corp., Palo Alto, CA. We selected two sources for the recombinant human NGF; one batch was purchased from Sigma Chemical Co., St. Louis, MO (No. N- 4398 used in eight rats) and the other batch was a generous gift from Dr. Louis E. Burton and Genentech Inc., San Francisco, CA (used in four rats). Primary antibodies for TH were obtained from Pel- Freeze Biologicals, Rogers, AR. Results In the histological sections of the normal rat DRG, the TH-immunoreactive (TH-ir) profiles were sparse within the ganglion (Fig. 1A), but numerous on the capsule of the ganglion (Fig. 1B). These TH-ir fibers can be classified in four general categories. Three of them are sympathetic postganglionic fibers: 1) fibers that innervate the blood vessels (precapillary arterioles) within the DRG (Fig. 1A, arrow); 2) fibers that supply the blood vessels on the capsule of the DRG (Fig. 1B, arrows); and 3) fibers of the spinal nerve penetrating various distances into the DRG fiber tract from the gray rami communicantes before looping outward and following the spinal nerve (Fig. 1C, arrows). The fourth TH-ir profiles in the DRG are a small number of dopaminergic DRG neurons (Fig. 1B, arrowhead) and their processes. The perivascular sympathetic fibers display the classic beads on a string structure and demonstrate spiral arrangement along the vascular walls (Fig. 1A, arrow). On the other hand, looping sympathetic fibers from the gray rami communicantes and dopaminergic sensory fibers display a smooth appearance along the long axes of fibers. Including all TH-ir profiles described earlier, the TH-ir profiles are scarce inside the capsule of the normal DRG and extremely rare in the cellular region of the ganglia (Fig. 1). Many sensory ganglia in the NGF-infused rats demonstrated various amounts of sympathetic fiber sprouting (Fig. 2). Sympathetic fiber sprouting was identified as the presence of TH-ir fiber profiles bearing the typical beaded structure that extended into the DRG cellular region, but did not exhibit normal perivascular confinement. The sprouting appeared as an abnormal extension of normal sympathetic fibers into the cellular zones and fiber tracts of the DRG. Many sprouted fibers could be traced back to one of the three preexisting sympathetic fibers described earlier, the most common being the perivascular sympathetic fibers. In cases showing perivascular sympathetic fiber sprouting, the normal spiraling innervation of the vasculature (Fig. 2A, double arrowheads) developed numerous branching fibers that extended into the surrounding cellular zone (Fig. 2A, arrow). The normal dense sympathetic innervation of the capsule also gave rise to occasional direct invasion through the capsule and sprouting among sensory neuron clusters near the capsule (Fig. 2C and D). In addition, fibers from the gray rami communicantes occasionally gave rise to sprouting fibers that exited from the fiber tract to invade the DRG cellular zone (Fig. 2B). The degree and presence of these forms of sympathetic fiber sprouting were of variable intensity and prevalence throughout the sections of a single ganglion. The typical presentation of a sensory ganglion was of several sections showing a moderate degree of these types of sprouting, whereas most of the other sections appeared normal. The representative sprouting seen in the various ganglia of each animal is presented in Table 1. The 12 animals receiving NGF showed a combined 52 sprouted ganglia (36.1%) of 144 sampled ganglia. Meanwhile, the seven animals receiving only ACSF showed sprouting in only two (2.8%) of 72 ganglia. Furthermore, although only two of seven control animals showed a minor degree of sprouting, the animals that received NGF showed various degrees of sprouting, from minor to extensive, in 11 of 12 animals (p 0.01). Sprouting was most prevalent at the lumbar levels and less so at the cervical and thoracic levels studied. The L-4 and L-5 ganglia combined represented 25 (48.1%) of 52 sprouted ganglia in NGF animals, although they represented only one third of the ganglia sampled. To substantiate further preferential sprouting at the lumbar levels studied, it appeared that the most extensive forms of sprouting, in which numerous sections showed dense sprouting over large portions of the sections, occurred exclusively in the L-4 and L-5 ganglia. The correlation between the frequency of sprouted DRGs and the infusion dose of NGF was not determined because the numbers of specimens for each dose tested were not sufficient for a statistical analysis. The data, however, revealed a tendency for a greater number of sprouted ganglia with a higher dose of NGF: the proportion of the ganglia with sympathetic fiber sprouting was 48% of the total at 100 g/week of NGF and 32% at 20 g/week. Sprouting was observed to a similar extent with each of the two preparations of recombinant human NGF used in these experiments. Discussion In this present study we demonstrate that intraventricular infusion of NGF induces sympathetic fiber sprouting 449

H. J. W. Nauta, et al. FIG. 2. Photomicrographs showing TH-immunostained lumbar DRGs of rats that received NGF infusion. In the DRGs of NGF-infused rats, various degrees of sympathetic fiber sprouting are evident and the sprouted fibers show the typical beads on a string structure (D, enlarged from the rectangle shown in C) and extend into the DRG cellular region, but do not exhibit normal perivascular confinement. Many sprouted fibers (arrows in A C) can be traced back to one of the three preexisting sympathetic fibers: the perivascular fibers (double arrowheads in A); the capsular sympathetic fibers (double arrowheads in C); and the sympathetic fibers in the fiber tracks (double arrowheads in B). Some sprouted fibers are in close apposition to DRG neurons (single arrowhead in B and D). Bars = 50 m. in sensory ganglia of rats without peripheral nerve injury. We infer from the data a possible cause of the diffuse back pain symptoms that developed in patients with Alzheimer s disease treated by intraventricular NGF infusion. The investigators stress that the use of neurotrophic substances to stimulate residual dysfunctional neurons in the central nervous system needs to be reevaluated in light of potential undesirable side effects, as shown in this animal study. Previous work has shown that sympathetic fiber sprouting into the DRGs can be induced by injury to a peripheral nerve in the distribution of the corresponding ganglion. 3,4,24,35 Such sprouting has been correlated to the development of mechanical and thermal hypersensitivities in animal models of peripheral neuropathy, 4,35 and these hypersensitivities return almost to normal after sympathectomy. 15 Based on these data, it has been postulated that sympathetic fiber sprouting in the DRG may be one of the causes of sympathetically maintained pain behaviors in animal models of peripheral neuropathy. The cause of sympathetic fiber sprouting in the DRG is speculated to be due to NGF fluctuations in the DRG after a peripheral nerve injury. Adult sympathetic neurons are regulated by NGF in their morphological structure and survival 33,36 and can respond by fiber outgrowth to supplemental NGF. 13,45 Nerve growth factor administered by chronic intraventricular infusion has been shown to induce sympathetic fiber sprouting in the subpial region of the medulla and spinal cord 45 and the intracranial vasculatures. 13,37 None of these previous studies, however, documented an invasion of abnormally sprouted sympathetic fibers into the neural tissue of the central nervous system. On the other hand, in transgenic mice with an overexpression of NGF in epidermal tissue, extensive sympathetic projections to sensory ganglia have been demonstrated, associated with a mechanical hyperalgesia. 5 In addition, an upregulation of local NGF synthesis has been reported in the rat DRG after peripheral nerve injuries 38,40 concurrent with sympathetic fiber sprouting in the DRG. The effects of NGF infusion on sympathetic fiber sprouting into the spinal and cranial sensory ganglia have not been studied. The present experiments were intended to evaluate whether NGF alone, without peripheral nerve injury, can induce sympathetic fiber sprouting. In the present experiments, NGF was delivered intraventricularly to the cerebrospinal fluid at a site far removed from the sites of ac- 450

Nerve growth factor and sprouting TABLE 1 Sympathetic sprouting in sensory ganglia of rats after intraventricular infusion of NGF or ACSF* Dorsal Root Ganglion Trigeminal Nerve No. of Infusion C-2 C-8 T-1 L-4 L-5 Ganglion Ganglia Animal Agent W/ No. ( g/wk) Rt Lt Rt Lt Rt Lt Rt Lt Rt Lt Rt Lt Growth 1 ACSF 0 0 0 0 0 + 1 2 ACSF 0 0 0 0 0 0 0 3 ACSF 0 0 0 0 0 0 0 0 0 0 0 0 0 4 ACSF 0 0 0 0 0 0 0 0 0 0 0 0 0 5 ACSF 0 0 0 0 0 0 0 0 0 0 0 0 0 6 ACSF 0 0 0 0 0 0 + 0 0 0 0 0 1 7 ACSF 0 0 0 0 0 0 0 0 0 0 0 0 0 8 NGF (100) 0 0 0 0 0 0 + 0 0 0 0 + 2 9 NGF (100) 0 + + 0 + + + + + + + + 10 10 NGF (100) 0 0 0 0 0 0 + + 0 + 0 + 4 11 NGF (100) 0 0 0 + + 0 0 + + + + + 7 12 NGF (50) 0 0 0 0 0 0 0 0 0 0 0 0 0 13 NGF (50) + + + 0 0 0 0 + 0 0 + + 6 14 NGF (20) 0 + 0 0 0 0 0 + 0 0 + 0 3 15 NGF (20) 0 + + 0 + 0 + + + + + + 9 16 NGF (20) 0 0 + 0 0 0 + 0 + 0 0 0 3 17 NGF (20) + 0 0 + 0 0 + + + 0 0 0 5 18 NGF (20) 0 0 0 0 0 0 0 + 0 0 0 0 1 19 NGF (20) 0 0 0 0 0 0 + + 0 0 0 0 2 * Entire sections of each ganglion were examined and graded. Any indication of sympathetic fiber sprouting was marked as positive (+), whereas no sprouting was marked as zero (0); = not sampled. tion under observation. No direct physical trauma to the nerves was expected as a result of the NGF delivery. Even under these experimental conditions, NGF clearly induced sympathetic nerve fiber sprouting into the rat DRG. The increase in the number of TH-ir fibers in the cellular regions of the ganglia was interpreted as sympathetic fiber sprouting in the DRG. Conversely, one could interpret the appearance of new TH-ir fibers as upregulation of TH by preexisting sympathetic fibers due to supplemented NGF. This alternate hypothesis is less likely. As stated earlier, in the naive uninjured DRG there are very few THir profiles viewed within the cellular zone, except for the spiraling sympathetic fibers confined to the surface of blood vessels. In the case of peripheral nerve ligation, the dramatically increased number of TH-ir fibers within the DRG were found to colocalize with GAP-43, a substance that is shown only in developing or regenerating neurons (unpublished data). With the noted increase in local NGF synthesis in such an injured peripheral nerve ganglion, 38,40 it appears more likely that the new TH-ir fibers in the present study are newly sprouted fibers extending from capsular, perivascular, and fiber-tract origin. The determination of positive and negative sprouting in each ganglion was based on the impression of the observer rather than the quantification of fiber numbers. There are several reasons for approaching data analysis this way. The observer has extensive experience in examination of the TH-immunostained normal DRG and is capable of recognizing abnormal patterns. Second, many examples of minor sympathetic fiber sprouting were well localized and, thus, appeared only in a fraction of the DRG sections. Due to this localized sprouting, examination of all serial sections of each ganglion was required for determination of sprouting. Third, the amount of sprouting varied greatly among different sections of one ganglion, among different ganglia from the same animal, and among different animals. With these predisposed restrictions and variations, we conjectured that the number of ganglia considered positive or negative for sympathetic fiber sprouting would represent the best overall indicator of NGF effects without applying laborious counting procedures. It is important to note that the distribution of sympathetic fiber sprouting in the sensory ganglia was not equal at all segmental levels of the neuraxis. Surprisingly, the most commonly observed and vigorous sprouting was in the lumbar ganglia (L-4 and L-5). We are uncertain why sprouting was more prominent in the lumbar area because these ganglia were maximally distant from the site of NGF delivery among the ganglia studied. Possible differences at various segmental levels with respect to 1) the permeability or access to the DRG by the NGF, 2) the expression of NGF receptors in DRG neurons and sympathetic fibers, and/or 3) the availability of preexisting sympathetic axons may underlie this unexpected finding. The findings presented here have implications both for Alzheimer s disease therapy and pain syndromes associated with peripheral nerve injury. Increased local NGF synthesis in the DRG in response to neural injury 38,40 or simple supplementation of exogenous NGF may induce localized sympathetic fiber sprouting in sensory ganglia and cause the development of pain syndromes. However, it remains uncertain whether NGF and the associated sympathetic fiber sprouting play a causal role in pain generation or are merely concomitants of other mechanisms such as peripheral and central sensitization of pain pathways, 21 signaling through mast cells, 46 or phenotype switching of sensory neurons. 26 Artificial elevation of cerebrospinal fluid NGF levels, such as that which occurs in experimen- 451

H. J. W. Nauta, et al. tal intraventricular NGF infusion for Alzheimer s disease, could potentially result in sympathetic fiber sprouting and pain phenomena, and such pain would likely be more diffuse than that found in a peripheral nerve injury. This concept leads to a possible explanation for the diffuse back pain syndrome seen in some patients with Alzheimer s disease who are treated with human NGF. 27 If NGF is to be useful in Alzheimer s disease therapy, its effects on sensory ganglia, as demonstrated here, and on other responsive tissues will need to be taken into account and the delivery system modified so that these effects are minimized. It remains to be shown that a benefit for Alzheimer s disease can be achieved while minimizing these unwanted effects. Conclusions The present study demonstrates that intraventricular NGF infusion caused sympathetic fiber sprouting in the DRGs of rats without peripheral nerve injury. This finding may have an implication for the lower-back pain phenomenon observed in patients with Alzheimer s disease who are treated with NGF, thus suggesting that careful adjustments of the neurotrophic factor delivering system need to be considered to minimize unwanted side effects. References 1. Arnér S: Intravenous phentolamine test: diagnostic and prognostic use in reflex sympathetic dystrophy. Pain 46:17 22, 1991 2. Bennett GJ, Xie YK: An experimental peripheral neuropathy in rat that produces abnormal pain sensation, in Dubner R, Gebhart GF, Bond MR (eds): Proceedings of the Vth World Congress on Pain. Pain Research and Clinical Management, Vol 3. Amsterdam: Elsevier, 1988, pp 129 134 3. Chung K, Kim HJ, Na HS, et al: Abnormalities of sympathetic innervation in the area of an injured peripheral nerve in a rat model of neuropathic pain. Neurosci Lett 162:85 88, 1993 4. Chung K, Lee BH, Yoon YW, et al: Sympathetic sprouting in the dorsal root ganglia of the injured peripheral nerve in a rat neuropathic pain model. J Comp Neurol 376:241 252, 1996 5. Davis BM, Albers KM, Seroogy KB, et al: Overexpression of nerve growth factor in transgenic mice induces novel sympathetic projections to primary sensory neurons. J Comp Neurol 349:464 474, 1994 6. Davis KD, Treede RD, Raja SN, et al: Topical application of clonidine relieves hyperalgesia in patients with sympathetically maintained pain. Pain 47:309 317, 1991 7. Fischer W, Wictorin K, Björklund A, et al: Amelioration of cholinergic neuron atrophy and spatial memory impairment in aged rats by nerve growth factor. Nature 329:65 68, 1987 8. Gonzales R, Sherbourne CD, Goldyne ME, et al: Noradrenaline-induced prostaglandin production by sympathetic postganglionic neurons is mediated by 2-adrenergic receptors. J Neurochem 57:1145 1150, 1991 9. Hefti F: Nerve growth factor promotes survival of septal cholinergic neurons after fimbrial transections. J Neurosci 6: 2155 2162, 1986 10. Hefti F, Hartikka J, Eckenstein F, et al: Nerve growth factor increases choline acetyltransferase but not survival or fiber outgrowth of cultured fetal septal cholinergic neurons. Neuroscience 14:55 68, 1985 11. Hefti F, Will B: Nerve growth factor is a neurotrophic factor for forebrain cholinergic neurons: implications for Alzheimer s disease. J Neural Transm Park Dis Dement Sect 24: 309 315, 1987 12. Hoffert MJ, Greenberg RP, Wolskee PJ, et al: Abnormal and collateral innervations of sympathetic and peripheral sensory fields associated with a case of causalgia. Pain 20:1 12, 1984 13. Isaacson LG, Saffran BN, Crutcher KA: Nerve growth factorinduced sprouting of mature, uninjured sympathetic axons. J Comp Neurol 326:327 336, 1992 14. Kim SH, Chung JM: An experimental model for peripheral neuropathy produced by segmental spinal nerve ligation in the rat. Pain 50:355 363, 1992 15. Kim SH, Chung JM: Sympathectomy alleviates mechanical allodynia in an experimental animal model for neuropathy in the rat. Neurosci Lett 134:131 134, 1991 16. Koliatsos VE: Biological therapies for Alzheimer s disease: focus on trophic factors. Crit Rev Neurobiol 10:205 238, 1996 17. Koliatsos VE, Applegate MD, Knüsel B, et al: Recombinant human nerve growth factor prevents retrograde degeneration of axotomized basal forebrain cholinergic neurons in the rat. Exp Neurol 112:161 173, 1991 18. Kromer LF: Nerve growth factor treatment after brain injury prevents neuronal death. Science 235:214 216, 1987 19. Levine JD, Taiwo YO, Collins SD, et al: Noradrenaline hyperalgesia is mediated through interaction with sympathetic postganglionic neurone terminals rather than activation of primary afferent nociceptors. Nature 323:158 160, 1986 20. Lewin GR, Ritter AM, Mendell LM: Nerve growth factor-induced hyperalgesia in the neonatal and adult rat. J Neurosci 13: 2136 2148, 1993 21. Lewin GR, Rueff A, Mendell LM: Peripheral and central mechanisms of NGF-induced hyperalgesia. Eur J Neurosci 6: 1903 1912, 1994 22. Lindvall O, Backlund EO, Farde L, et al: Transplantation in Parkinson s disease: two cases of adrenal medullary grafts to the putamen. Ann Neurol 22:457 468, 1987 23. Loh L, Nathan PW, Schott GD, et al: Effects of regional guanethidine infusion in certain painful states. J Neurol Neurosurg Psychiatry 43:446 451, 1980 24. McLachlan EM, Jänig W, Devor M, et al: Peripheral nerve injury triggers noradrenergic sprouting within dorsal root ganglia. Nature 363:543 546, 1993 25. Neil A, Attal N, Guilbaud G: Effects of guanethidine on sensitization to natural stimuli and self-mutilating behaviour in rats with a peripheral neuropathy. Brain Res 565:237 246, 1991 26. Neumann S, Doubell TP, Leslie T, et al: Inflammatory pain hypersensitivity mediated by phenotypic switch in myelinated primary sensory neurons. Nature 384:360 364, 1996 27. Nordberg A, Almkvist O, Amberla K, et al: Responders and non-responders to tacrine, ondansetron and NGF treatment in Alzheimer patients as evaluated by positron emission tomography and APOE genotype, in Iqbal K, Winblad B, Nishimura T, et al (eds): Alzheimer s Disease: Biology, Diagnosis and Therapeutics. Chichester: John Wiley & Sons, 1997, pp 647 654 28. Olson L: NGF and the treatment of Alzheimer s disease. Exp Neurol 124:5 15, 1993 29. Olson L, Backlund EO, Ebendal T, et al: Intraputaminal infusion of nerve growth factor to support adrenal medullary autografts in Parkinson s disease. One-year follow-up of first clinical trial. Arch Neurol 48:373 381, 1991 30. Olson L, Nordberg A, von Holst H, et al: Nerve growth factor affects 11 C-nicotine binding, blood flow, EEG, and verbal episodic memory in an Alzheimer patient (case report). J Neural Transm Park Dis Dement Sect 4:79 95, 1992 31. Paxinos G, Watson C: The Rat Brain in Stereotaxic Coordinates, ed 2. San Diego: Academic Press, 1986 32. Phelps CH, Gage FH, Growdon JH, et al: Potential use of nerve growth factor to treat Alzheimer s disease. Neurobiol Aging 10:205 207, 1989 452

Nerve growth factor and sprouting 33. Purves D, Snider WD, Voyvodic JT: Trophic regulation of nerve cell morphology and innervation in the autonomic nervous system. Nature 336:123 128, 1988 34. Raja SN, Treede RD, Davis KD, et al: Systemic alpha-adrenergic blockade with phentolamine: a diagnostic test for sympathetically maintained pain. Anesthesiology 74:691 698, 1991 35. Ramer MS, Bisby MA: Rapid sprouting of sympathetic axons in dorsal root ganglia of rats with a chronic constriction injury. Pain 70:237 244, 1997 36. Ruit KG, Osborne PA, Schmidt RE, et al: Nerve growth factor regulates sympathetic ganglion cell morphology and survival in the adult mouse. J Neurosci 10:2412 2419, 1990 37. Saffran BN, Woo JE, Mobley WC, et al: Intraventricular NGF infusion in the mature rat brain enhances sympathetic innervation of cerebrovascular targets but fails to elicit sympathetic ingrowth. Brain Res 492:245 254, 1989 38. Sebert ME, Shooter EM: Expression of mrna for neurotrophic factors and their receptors in the rat dorsal root ganglion and sciatic nerve following nerve injury. J Neurosci Res 36: 357 367, 1993 39. Seltzer Z, Dubner R, Shir Y: A novel behavioral model of neuropathic pain disorders produced in rats by partial sciatic nerve injury. Pain 43:205 218, 1990 40. Shen H, Chung JM, Chung K: Expression of neurotrophin mrnas in the dorsal root ganglion after spinal nerve injury. Mol Brain Res 64:186 192, 1999 41. Shir Y, Seltzer Z: Effects of sympathectomy in a model of causalgiform pain produced by partial sciatic nerve injury in rats. Pain 45:309 320, 1991 42. Wahren LK, Torebjörk E, Nyström B: Quantitative sensory testing before and after regional guanethidine block in patients with neuralgia in the hand. Pain 46:23 30, 1991 43. Wilcock GK, Esiri MM, Bowen DM, et al: The nucleus basalis in Alzheimer s disease: cell counts and cortical biochemistry. Neuropathol Appl Neurobiol 9:175 179, 1983 44. Williams LR, Varon S, Peterson GM, et al: Continuous infusion of nerve growth factor prevents basal forebrain neuronal death after fimbria fornix transection. Proc Natl Acad Sci USA 83: 9231 9235, 1986 45. Winkler J, Ramirez GA, Kuhn G, et al: Reversible Schwann cell hyperplasia and sprouting of sensory and sympathetic neurites after intraventricular administration of nerve growth factor. Ann Neurol 41:82 93, 1997 46. Woolf CJ, Ma QP, Allchorne A, et al: Peripheral cell types contributing to the hyperalgesic action of nerve growth factor in inflammation. J Neurosci 16:2716 2723, 1996 Manuscript received January 18, 1999. Accepted in final form May 10, 1999. This work was supported by a grant from the Moody Foundation to Dr. Nauta and National Institutes of Health Grant No. NS 35057 to Dr. Chung. Address reprint requests to: Kyungsoon Chung, Ph.D., Marine Biomedical Institute, University of Texas Medical Branch, Galveston, Texas 77555 1069. email: kchung@utmb.edu. 453