Immunohistochemical localization of the high-affinity NGF receptor (gp 140-trkA) in the adult human dorsal root and sympathetic ganglia and in the nerves and sensory corpuscles supplying digital skin.код для вставкиСкачать
THE ANATOMICAL RECORD 240:579-588 (1994) Immunohistochemical Localization of the High-Affinity NGF Receptor (gpl40-trkA) in the Adult Human Dorsal Root and Sympathetic Ganglia and in the Nerves and Sensory Corpuscles Supplying Digital Skin J.A. VEGA, E. VAZQUEZ, F.J. NAVES, M.E. DEL VALLE, B. CALZADA, AND J.J. REPRESA Departamento de Morfologia y Biologia Celular, Uniuersidad de Ouiedo, Ouiedo, Spain ABSTRACT Background: Nerve growth factor (NGF)is produced in target tissues of sympathetic and neural-crest derived sensory neurons, including skin, to provide them trophic support. The biological effects of NGF on responsive cells are mediated by specific high-affinity receptors. Recently, a protein tyrosine kinase of = 140 kDa molecular weight, encoded by the proto-oncogene trkA, has been identified as the high-affinity NGF receptor (gpl40-trkA).The present work was undertaken to study the localization of gpl40-trkA-like immunoreactivity (IR) in human peripheral ganglia (sympathetic and dorsal root ganglia), and in glabrous skin. Methods: Lumbar dorsal root ganglia, para- and prevertebral sympathetic ganglia, and digital glabrous skin were studied immunohistochemically using a rabbit anti-gpl40-trkA polyclonal antibody. In order to accurately establish the localization of gpl40-trkA IR, the neurofilament proteins and S-100 protein were studied in parallel in: (1)sensory and sympathetic ganglia, to label neuron cell bodies and satellite or supporting cells, respectively; ( 2 ) human skin, to label axons, Schwann and related cells within nerves and sensory corpuscles. Moreover, a quantitative study (neuron size, intensity of immunostaining) was carried out on sympathetic and dorsal root ganglia neuron cell bodies. Results: A specific gpl40-trkA-like IR was found in: (1)a subpopulation (65%)of primary sensory neuron cell bodies, including most of the largesized ones but also small- and intermediate-sized ones; (2) most of sympathetic neuron cell bodies (82%);(3)the perineurial cell, Schwann cells, and large axons of the nerve trunks supplying digital skin; (4) the lamellar cells of Meissner corpuscles; (5) the central axon, inner-core, outer-core, and capsule of Pacinian corpuscles. In addition, the occurrence of gpl40-trkAlike IR was observed in some non-nervous tissues of the skin, including epidermis (mainly in the basal layer), sweat glands, and arterial blood vessels. Conclusions: Present results provide evidence for the localization of gpl40-trkA-like IR in: (1)nerve cells which are known to be NGF-responsive, and (2) non-nervous cutaneous tissues which are innervated by NGFdependent peripheral neurons. These findings suggest that, in addition to the well-established role of NGF on sensory and sympathetic neurons, this neurotrophin may be able to regulate some other functions on non-nervous cells which are targets for NGF-dependent peripheral neurons. 0 1994 Wiley-Liss, Inc. Key words: Nerve growth factor receptors, gpl40-trkA, Dorsal root ganglia, Sympathetic ganglia, Cutaneous sensory corpuscles, Skin, Immunohistochemistry, Man Nerve growth factor (NGF) is produced in target Of sympathetic and neural-crest derived sensory neurons, skin, to provide them trophic port (Levi-Montalcini, 1987). In addition, there is evi0 1994 WILEY-LISS, INC. Received November 10, 1993; accepted June 21, 1994, Address reprint requests to Jose A. Vega, Departemento de Morfologia y Biologia Celular, Facultad de Medicina, Univ. de Oviedo, Ci Julian Claveria, sin, E-33006 Oviedo, Spain. 580 J.A. VEGA ET AL. dence for NGF-mediated paracrine effects in the skin itself (Yaar et al., 1991). NGF binds two classes of receptors: the high-affinity receptors (Kd = lop1' MI and the low-affinity receptors (Kd = lo-' M) (see Bothwell, 1989). The first one is a protein tyrosine kinase of = 140 kDa of molecular weight encoded by the proto-oncogene trkA (Kaplan et al., 1991; Barbacid et al., 1991; Ross, 1991; Meakin and Shooter, 1992; Chao, 1993). The low-affinity one is a glycoprotein with a molecular weight of = 75 kDa (see Bothwell, 19891, which is able to bind all different members of the NGF family of neurotrophins (Rodriguez-Tebar et al., 1990). The high- and low-affinity NGF receptors are referred to here a s gpl40-trkA and p75, respectively. The distribution of p75 IR in the adult dorsal root and sympathetic ganglia and their targets has been widely studied in avian and mammalian species, including man (Chesa et al., 1988; Sobue et al., 1989; Wyatt et al., 1990). Furthermore, it has been reported that in the cutaneous targets, both nerves and sensory corpuscles express p75 IR (Byers, 1990; Ribeiro da Silva et al., 1991; Yamamoto et al., 1992; Vega et al., 1992, 1993). On the contrary, studies on the localization of gp140trkA in the peripheral nervous system are scarce, specially in adult human tissues. Using in situ hybridization analysis, trkA mRNA has been found in the developing and adult sympathetic and dorsal root ganglia of rat, as well a s in these ganglia of developing human specimens (Martin-Zanca et al., 1990; Verge et al., 1992; Wetmore et al., 1992; Ernfors et al., 1993). However, no data are yet available on trkA mRNA neither in the adult human sensory and sympathetic neurons, nor in their cutaneous targets. Furthermore, the distribution of the trkA-encoded protein, namely gp140trkA, has not been studied in these locations; which is a n important issue, since is well established that mRNA levels do not accurately reflect the protein turnover. In the present study we addressed the localization of gpl40-trkA IR in the adult human dorsal root ganglia, sympathetic ganglia, peripheral nerves, and digital glabrous skin. To ascertain what cell types displayed gpl40-trkA-like IR, neurofilament proteins (NFP) and S-100 protein (SlOOP)were studied in parallel, to label neuron cells bodies, and their peripheral processes (Lawson and Waddell, 1991), as well a s satellite cells and Schwann cells, respectively (Vega et al., 1991).The distribution of both antigens in human sensory corpuscles has been described in previous reports (Haro et al., 1991; Vega et al., 1993). MATERIALS AND METHODS Human lumbar dorsal root and sympathetic ganglia, as well as samples of digital skin, were obtained from adult subjects (age range 21 to 56 years). The lumbar dorsal root ganglia (n = 11)were obtained from 5 subjects who died by traffic or industrial accidents, during the process of removing the organs for transplantation. The sympathetic ganglia (n = 13) were obtained from 6 subjects during surgery; 10 pieces were lumbar paravertebral, and 3 were prevertebral. The samples of digital skin (n = 8) were obtained from 6 subjects who suffered traumatic amputation of one or more hand fingers and collected within 6 h after lesion. All pieces were dissected, washed in a cold 0.9% saline solution, fixed for 24-48 h in 10% buffered formaldehyde (pH 7.41, dehydrated, and routinely embedded in paraffin. lmmunohistochemistry Sections 10 pm thick were obtained, mounted on gelatin-coated slides, and processed for indirect peroxidase-antiperoxidase immunohistochemistry a s follows. Deparaffined and rehydrated sections were washed in 0.05 M Tris-HC1 (TBS, pH 7.4) containing 1%Triton X-100 and 1%bovine serum albumin. Thereafter, the endogenous peroxidase activity and the non-specific binding were blocked. Then, sections were incubated overnight in a humid chamber at 4°C with a rabbit polyclonal antibody against the human gpl40-trk (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 1:lOO in PBS 0.1 M, pH 7.4. This anti-trk antibody (also known a s trkA) is a n affinity-purified rabbit polyclonal antibody raised against a peptide corresponding to the residues 754-790 mapping within the carboxyl terminal domain of the predicted human gp140 encoded protein. This antibody is specific for trkA, hence, shows no trkB nor trkC cross-reactivity. Representative sections of dorsal root and sympathetic ganglia and skin were incubated with mouse monoclonal antibodies anti-NFP 200 kDa (clone NE14; Boehringer-Mannhein, Germany), or anti-S100P (clone 15E2E2;Boehringer-Mannhein) both diluted to 5 pgiml. After incubation with the primary antibodies the sections were washed in PBS and incubated for 1 h a t room temperature with peroxidase-labelled sheep antirabbit IgG (Boehringer-Mannhein; diluted to 5 U/ml) or sheep anti-mouse IgG (Amersham, UK; diluted 1:lOO).Finally, the sections were washed twice in PBS and incubated with 0.05% 3.3'-diaminobenzidinei 0.01% H202for 5-10 min. For control purposes, incubation with the primary antibody was omitted or a non-immune rabbit or mouse IgG were used instead of the primary antibody. Sections were examined and photographed in a light microscope. Quantitative Studies The neuron size, as well as the intensity of gp140trkA-like immunostaining developed in the nerve cell body profiles of both sympathetic and dorsal root ganglia were evaluated using a n automatic image analysis system (MIP, Servicio de Analisis de Imagenes, University of Oviedo) as described early (Dubovy et al., 1990). Briefly, measurements were made on 3 sections of 5 sympathetic ganglia and of 5 dorsal root ganglia. Sections were 100 pm apart to avoid counting the same neuron twice. Five randomly selected fields per section were evaluated, and the number of analyzed neurons was 20-35 per section (60-105 neurons per ganglion). The following measurements were made: (1) neuron size, evaluating the mean diameter (values are expressed in pm) of nerve cell body profiles with a n apparent nucleus. Neurons were divided into three groups: small- (<25 pm), intermediate- (26-50 pm), and large-sized ( > 5 1 pm); results are expressed a s the percentage of labelled neurons in each established subclass. ( 2 ) intensity of immunostaining, expressing the values in arbitrary units of grey-levels ranging from 1 (black) to 256 (white). Neurons were divided into four groups (64 grey-level units width each), referred here gpl40-trkA IN THE HUMAN NERVOUS SYSTEM 581 Fig. 1. Sections of dorsal root ganglia showing distribution of S-100 protein IR (A), gpl4O-trkA-like IR (B,D,E), and neurofilament proteins IR (C). In control sections (F) no positive IR was observed. As can be shown gpl40-trkA-like IR labels primary sensory neuron cell bodies and satellite cells (E, arrowheads). Moreover, a strong immunostaining was present in the intraganglionic blood vessels (B and D, arrows). Arrows in (C) denote intragangiionic axons. Scale bar in (A) = 25 pm for A-C and F; in (D) 40 pm for D and E. TABLE 1. Percentage of primary sensory and sympathetic neurons displaying gpl4O-trkA-likeIR, and intensity of gpl40-trkA-likeimmunostaining into the different neuron-size subclasses RESULTS Dorsal Root Ganglia Neuron size (urn) < 25 (32%) Dorsal root ganglia (cell n = 342) 25-50 (56%) >51(12%) Sympathetic ganglia < 25 (21%) (cell n = 428) 25-50 (69%) >51(10%) Intensity of IR % o f 5% of un- (greylevel IR neu- labeled rons neurons units) 13 41 11 19 55 8 19 15 1 2 14 2 1 8 6 k 16 146 2 13 9629 135 i 7 12659 131 2 6 a s strong-, high-, intermediate-, and low-immunoreactivity. Nerve cell bodies with grey-level values lower than 240 were considered unreactive. Results are expressed as mean S.D. values within each established neuron-size classes. * The distribution of the measured nerve cell profiles into the three pre-established size-classes was 32% of small-, 56% of intermediate-, and 12% of large-sized cells. I n the analyzed lumbar dorsal root ganglia, specific gpl40-trhA-like IR was found in about 65% of nerve cell bodies (Fig. 1; Table 1).Most of the largesized perikarya (11%) displayed a high or intermediate gpl40-trhA-like IR (96 5 9 grey-level units), whereas only a subset of the intermediate- (41%) and the smallsized (13%) resulted labelled. The intensity of IR was 146 5 13 grey-level U for the intermediate-sized neurons, and 186 ? 16 grey-level U for the small-sized nerve cells (Fig. lB,D,E; see Table 1).Thus, the intensity of immunostaining was dependent on the neuron size. On the other hand, although differences in the intensity of immunostaining were encountered among subjects, the pattern of gpl40-trkA-like distribution remained unchanged. Moreover, we have observed that satellite cells, even that sheathing neurons lacking of immunoreactivity, showed gp140-trhA-like IR (Fig. lB,D,E). In comparing sections processed for gp140- 582 J.A. VEGA ET AL. Fig. 2. Localization of S-100 protein IR (A), gpl40-trkA-like IR (B,D,E),and neurofilament proteins IR (C) in the paravertebral lumbar sympathetic ganglia. gpl40-trkA-like IR was present in both neuronal profiles and satellite cells (D,E, arrowhead). Furthermore, blood vessels supplying sympathetic ganglia regularly resulted labelled (B, arrows). Under control conditions no positive IR was found and dark areas corresponded with intraneuronal storage of lipophuscin (F). Scale bar in (A) = 25 pm for A-C and F; in (D) 20 pm for D and E. trkA, NFP and SlOOP (Fig. lA,B,C) the localization of gpl40-trkA-like IR in the neuron cell bodies and satellite cells can be observed. Interestingly, blood vessels supplying dorsal root ganglia showed the highest intensity of immunostaining (Fig. lB,D). 2B) and SlOOP IR (Fig. 2A) confirmed our observations on the localization of gpl40-trkA-like IR. Digital Skin gpl40-trkA-like IR was observed in both neuronal and non-neuronal cutaneous structures. In the deep Sympathetic Ganglia nerve trunks (Fig. 3B) a strong gpl40-trkA-like IR was The distribution of the measured nerve cell profiles present in the perineurium, Schwann cells, and in a into the three pre-established size-classes was 21% of subset of axons, primarily those of large caliber. Moresmall-, 69% of intermediate-, and 10% of large-sized over, a faint gpl40-trkA-like IR was observed in the cells. Most of the neuronal perikarya (82%) of the pre endoneurial fibroblasts. These localizations were conand paravertebral sympathetic ganglia expressed firmed comparing adjacent sections processed for gpl40-trkA-like IR in their cytoplasm with a rather gpl40-trkA with that for SlOOP (Fig. 3A) and NFP homogeneous intensity of immunostaining which was (Fig. 3C) demonstration. independent of the neuron size (Fig. 2). The immunoDifferent kinds of sensory corpuscles showing gp140reactive neurons included 19% of small-, 55%of inter- trhA-like IR were found in the digital skin. The lamelmediate-, and 8% of large-sized cells (see Table 1). lar cells of Meissner corpuscles placed in the dermal Moreover, satellite cells displaying positive IR were papillae showed intense gpl40-trkA-like IR (Fig. 4). also found (Fig. 2D,E), the intensity of IR being stron- Moreover, immunostaining in the central axon of ger than that observed in the neuronal somata. On the Meissner corpuscles was not evident. With regard to other hand, and a s i t was described for the dorsal root the Pacinian corpuscles, all the three main systems ganglia, intraganglionic blood vessels were intensely (i.e., the central axon, inner-core, and outer-corelcaplabelled (data not shown). The comparison of the dis- sule) were immunolabelled (Fig. 5A,B). Results on Patribution for NFP IR (Fig. 2 0 , gpl4O-trkA-like IR (Fig. cini-like corpuscles and on simple lamellar corpuscles Fig. 3. Adjacent sections of a deep cutaneous nerve incubated with antibodies for S-100 protein (A),gpl40-trkA (B),and neurofilament proteins (C).The gpl40-trkA-like IR was found labelling perineurium (p), Schwann cells and axons (arrows). The endoneurial fibroblasts also were labelled. Scale bar = 25 pm. Fig. 4. Meissner’s corpuscles displaying gpl4O-trkA-like IR. The occurrence of IR within central axons cannot be assured, while it was evident in the lamellar cells. e: epidermis. Scale bar = 25 Fm. Fig. 5. Localization of gpl4O-trkA-like IR in Pacinian corpuscles (A,B) and Pacini-like corpuscles ( C ) . All the corpuscular structures had immunolabelling with the intensity of IR being stronger in the inner-core lamellae. Inset in (B) corresponds to an approximate section of the Pacinian corpuscle processed for detection of neurofilament proteins. Arrows indicate the axons; c: capsule; ic or arrowheads: inner-core; oc: outer-core. Scale bar = 25 pm 584 J.A. VEGA ET AL. were similar, with a prominent gpl40-trkA-like IR within the central axon and inner-core lamellar (Fig. 5C). Localization of gpl4O-trkA-like IR on sensory corpuscles was confirmed by comparing sections processed for gpl40-trkA-like detection with that for NFP (inset in Fig. 5B) and SlOOP (data not shown; see Vega et al., 1993). Regarding non-nervous tissues, we observed the presence of gpl40-trkA-like IR in the epidermis, sweat glands, and blood vessels. In the epidermis, the stronger IR was in the basal layer, while the suprabasal layers showed a very faint gpl40-trkA-like IR (Fig. 6). Whether or not Merkel cells show IR cannot be ensured since no specific techniques to label Merkel cells were done. Furthermore, gpl40-trkA-like IR cells showing a dendritic-like morphology were found at the level of the suprabasal layer (Fig. 6A) or within the basal layer (Fig. 6A,B). They could be identified as Langerhans cells and/or melanocytes. In addition to the epidermis, other epithelial structures within the dermis displayed gpl40-trkA-like IR. We identified them as sweat glands (Fig. 7), and the IR intensity was stronger in the ductal part than in the acinar one. At the light microscope level it is not possible to ascertain whether IR labelled myoepithelial cells. Finally, gpl40-trkA-like IR was identified in the walls of both large and small arteries (Fig. 81, but not in the veins (Fig. 8B). The labelling was mainly found in the smooth muscle cells of the tunica media, and occasionally also in the adventitia a s nerve-like profiles. Apparently, no positive immunostaining was present in the endothelial cells. Under control conditions, where the primary antibody was omitted or a non-immune rabbit or mouse IgG were used instead of the primary antibody, no specific IR was observed for any of the assessed antibodies (Figs. l F , 2F, and 6C). DISCUSSION This paper reports for the first time the distribution of gpl40-trkA-like IR in the adult human dorsal root and sympathetic ganglia, in the peripheral nerves, as well a s in their cutaneous targets. These included sensory corpuscles, epidermis, sweat glands, and blood vessels. The exact nature of the gpl40-trkA-like IR cells in the nervous structures was established comparing the pattern of gpl40-trkA-like IR with that of markers for neurons (NFP), or satellite cells (Sloop) whose distribution is well known in the peripheral ganglia and nerves (Lawson and Waddell, 1991; Vega et al., 1991) as well a s in the sensory corpuscles (Vega et al., 1993). The neurons of the sympathetic and dorsal root ganglia (see Levi-Montalcini and Calissano, 1986; Barde, 1989; Verge et al., 19931, as well as neuronal subpopulations of some cranial sensory ganglia, are under NGF control during development (Davies and Lindsay, 1985; Represa and Bernd, 1989). These neurons express both low- and high-affinity NGF receptors during development (Bernd and Represa, 1989; Wyatt et al., 1990) and also in post-natal periods (Ernfors et al., 1993). However, whether or not NGF is essential for maintenance of mature peripheral neurons is unclear (Lindsay, 1988; Gold et al., 1991). Data concerning the presence of NGF receptors in adult human sympathetic and dorsal root ganglia are restricted to the low-affin- ity receptors, thus the p75 protein (Sobue e t al., 1989). Present results demonstrate that most of postganglionic sympathetic neurons (82%) and a wide subset of sensory dorsal root ganglia neurons (65%) express gpl40-trkA-like IR, thus suggesting a n active role of NGF in these mature and differentiated nerve cells (Lindsay, 1988). We have constantly observed gpl40-trkA-like IR in nerve trunks supplying digital skin, labelling perineurial cells, endoneurial fibroblasts, Schwann cells, and large axons. The pattern of gpl40-trkA-like IR in these localizations matches the pattern of distribution of p75 IR in human nerve trunks (see Vega et al., 1992). However, our findings are in disagreement with those reported by Yamamoto et al. (1993) who could not observe expression of gpl40-trkA mRNA in cultured Schwann cells. Specific or technical differences may account for these discrepancies (i.e., levels of mRNA might be undetectable while levels of protein might not be). Both Meissner and Pacinian corpuscles expressed gpl4O-trkA-like IR with a n identical localization to that reported for p75 (Vega et al., 1993). At the light microscope level, nevertheless, whether immunolabelling occurs within the cytoplasm or in the membrane of lamellar cells cannot be ensured. Results obtained in sensory corpuscles may be considered together with that on nerve trunks, due to the fact that the axon, lamellar cells or inner-core and outer-corelcapsule of sensory corpuscles are continuous with the axon, Schwann cells, and perineurial cells, respectively, of the nerve fibers (Malinovsky, 1986; Munger and Ide, 1988). However, i t might also be possible that NGF receptors present in sensory corpuscles play a specific and distinct role in these structures (Ritter et al., 1991; Lewin e t al., 1992; Lewin and Mendell, 1993), although this issue remain controversial (Diamond e t al., 1992a,b). As above described, gpl40-trkA-like IR was mainly found in intermediate- and large-sized primary sensory neurons and, probably, in their parent axons in cutaneous nerves. Interestingly, the axons entering Meissner and Pacinian corpuscles represent the specialized ending of A@and A6 nerve fibers (see Dalsgaard, 1988; Fyffe, 1992) originated from large- and medium-sized sensory neurons (Lawson and Waddell, 1991). Taken together these observations suggest that expression of gpl40-trkA could be related with non-nociceptive mechanoreception. Several studies have suggested a n involvement of NGF in the control of different functions in non-neuronal tissues, other than the classical neurotrophic effects. These tissues include keratinocytes, dendritic cells, sweat glands, and blood vessels of the human skin, since all of them express p75 IR (see Chesa et al., 1988; Ernfors et al., 1991; Vega e t al., 1992). We have now demonstrated the occurrence of gpl40-trkA-like IR in the same non-nervous cutaneous structures, in a pattern which closely resembles that of p75 protein. However, whether gpl40-trkA-like IR in the basal keratinocytes is related with the so-called trkE (Di Marco et al., 1993) must be clarified in future studies. gpl40-trkA-like IR was observed in the cutaneous blood vessels, primarily in the tunica media. No data are so far available on this topic. However, the distribution of gpl40-trkA-like IR showed here basically matches the localization of p75 IR in different blood gpl40-trkA IN THE HUMAN NERVOUS SYSTEM Fig. 6. In the epidermis most cells the basal layer were gpl40-trkAlike IR positive. Within the basal layer and In the suprabasal layers of the epidermis, cells with a dendritic morphology and displaying gpl40-trkA-like IR were observed (A,B, arrowheads). e: epidermis. Scale bar = 25 pm. Fig. 7. Sweat glands showing gpl40-trkA-like IR which was strong in the ductal portions (arrows)and moderate in the acini. Scale bar = 25 pm. 585 Fig. 8. Large (A)and small (B)cutaneous arteries displaying gp140trkA-like IR. In the large ones IR was primarily localized in the tunica media while in the small ones the IR occupied all wall layers. The veins did not display IR. a: adventitia; ar: artery; tm: tunica media; v: vein. Scale bar = 100 pm in A, and 25 pm in B. 586 J.A. VEGA ET AL. corpuscles, as well a s for the epithelial cells forming the sweat glands. As can be observed in Table 2, most cells producing NGF also express both types of NGF receptors. Theregpl40-trkANGF like IR p75-IR2 secretion3 fore, our results strongly suggest t h a t NGF produced and released in human skin may (1)act on NGF-reNerve bundles sponsive sensory and sympathetic neurons, being con+++ +++ ? Perineurium ducted to them throughout axonic retrograde trans++ ++ ? Endoneurial fibroblasts port; (2) have a major role in regulating functions in +++ +++ Yes Schwann cells some non-neuronal cutaneous cells working in a n au++ +I No Axons tocrine or paracrine manner, as is suggested for keraSensory corpuscles tinocytes and dendritic cells (Di Marco et al., 1991; ++ + +I+ ? Capsule* Yaar et al., 1991). In this way Schecterson and Both++ + +I+ ? Outer-core* well (1992) have suggested a role for NFG in neuron to +++ +++ ? Schwann-related cells target tissue signaling. Moreover, since non-neuronal ++ +INo Axon cutaneous cells displaying gpl40-trkA-like IR also exEpidermis press p75 IR, a collaboration between both types of +++ ++ Yes Basal keratinocytes NGF receptors could be hypothesized for the skin. How+I+ Yes Suprabasal layers ? ++ ? Merkel cells ever, co-expression of both NGF receptors to mediate +++ +++ ? Dendritic cells biological responses of NGF in other systems remains controversial (Glass et al., 1991; Hempstead et al., Dermis Fibroblasts 1991; Ibanez et al., 1992; Lee et al., 1992). Further Muscle cells studies will be necessary to clarify the role of NGF in the human skin cells which express NGF receptors, Sweat glands ++ +++ ? Acini and the functional and clinical relevance of the present +++ +++ ? Excretory ducts results remain to be elucidated. TABLE 2. Distribution of gpl4O-trkA-likeIR and p75 IR in human digital skin, and cells synthesizing and secreting NGF' - Blood vessels Endothelium Muscular layer Adventitia - ++ - - + + +I+ No Yes ? '+ + + indicates strong, + + moderate, and + weak I R - indicates no I R ? = data not done; * indicates Pacinian corpuscles. 'From Chesa et al. (1988) and Vega et al. 11992, 1993). 3From Rohrer et al. (1988), Creedon and Tuttle (19911, Di Marco et al. (1991). Yaar et al. (1991), Thoenen et al. (19921, Yamamoto et al. (1993). vessels (see Del Valle e t al., 19921, including the cutaneous ones (Vega et al., 1992). Our observations suggest that vascular smooth muscle cells might be responsive to NGF. In supporting this opinion, Messina and Bell (1991) and Zettler et al. (1991) demonstrated that NGF may be involved in the regulation of blood vessel pressure, probably throughout a sympathetic way (Isaacson et al., 1990) since NGF acts as a n inhibitory neuromodulator of adrenergic transmission in the rat mesenteric artery (Ueyama et al., 1991). NGF is synthesized and released by different cells present in the developing and adult human skin (see Table 2) including keratinocytes (Rohrer e t al., 1988; Wyatt et al., 1990, Di Marco et al., 1991; Yaar et al., 1991), fribroblasts, Schwann cells, and smooth muscle cells (Creedon and Tuttle, 1991; Thoenen et al., 1992; De Vellis, 1993; Reynolds and Woolf, 1993). In this order, the tunica media of elastic arteries (MacGrogah et al., 1992; Scarisbrick et al., 1993) contain high mRNA levels for neurotrophins, including NGF. Although no data are so far available on the synthesis and release of NGF from perineurial cells and their corpuscular derivatives, a presumptive NGF production by these cells could be assumed (see Yoshida and Gage. 1992: Thoenen et al., 1992), since they have a fibroblastic' origin (Halata et al., 1990). The Same may apply for the Schwann-related cells of the Meissner and Pacinian ACKNOWLEDGMENTS Present study was supported by grants from FISS (91/0361) and from the University of Oviedo (DF-92/63, DF-93 218-57, and TA-93 218-35). E.V. was recipient of a grant fellowship from DGICYT (92/0261). Technical assistance of Mr. J.A. Garcia-Sanchez is gratefully acknowledged. LITERATURE CITED Barbacid, M., F. Lambelle, D. Pulido, and R. Klein 1991 The trk family of tyrosine protein kinase receptors. Biochem. Biophys. Acta, 1072tl15-127. Barde, Y.-A. 1989 Trophic factors and neuronal survival. Neuron, 2t1525-1534. Bernd, P., and J . Represa 1989 Characterization and localization of nerve growth factor receptors in the embryonic otic vesicle and cochleovestibular ganglion. Dev. Biol., 134t11-20. Bothwell, M. 1989 Characterization of nerve growth factor receptor. In: Nerve Growth Factors. R.A. Rush, ed. John Wiley & Sons, New York, pp. 299-313. Byers, M.R. 1990 Segregation of NGF receptor in sensory receptors, nerves and local cells of teeth and periodontium demonstrated by ME immunocytochemistry. J . Neurocytol., 19t765-775. Chao, M.V. 1993 Neurotrophin receptors: A window into neuronal differentiation. Neuron, 9:583-593. Chesa, P.G., W.J. Retting, T.M. Thompson, L.J. Old, and R. Melamed 1988 Immunohistochemical analysis of nerve growth factor expression in normal and malignant human tissues. J. Histochem. Cytochem., 36t383-389. Creedon, D., and J.B. Tuttle 1991 Nerve growth factor synthesis in vascular smooth muscle. Hypertension, 18t730-741. Dalsgaard, C.-J. 1988 The sensory system. In: Handbook of Chemical Neuroanatomy. Vol. 6. Peripheral Nervous System. A. Bjorklund, T. Hokfelt, and C. Owman, eds. Elsevier Science Publishers B.V., Amsterdam, pp. 599-636. Davies, A.M., and R.M. Lindsay 1985 The cranial sensory ganglia in culture: Differences in the response of placode-derived and neural crest-derived neurons to nerve growth factor. Dev. Biol., IIlr6272 Del Valle, M.E., J.C. ALvarez-Mendez, B. Calzada, and J.A. Vega 1992 Nerve growth factor receptor IR in non-nervous structures of the adult rat brain. Eur. J . Histochem., 36t435-444. De Vellis. J. 1993 Sumortine cells. Central and DeriDheral. In: Neuroregeneration. A:Gori<ed., Raven Press, New kork, pp. 61-75. Diamond, J . , A. Foerster, M. Holmes, and M. Coughlin 1992a Sensory gpl40-trkA IN THE HUMAN NERVOUS SYSTEM 587 nerves in adult rats regenerate and restore sensory function to Lindsay, R.M. 1988 Nerve growth factors (NGF and BDNF) enhance the skin independently of endogenous NGF. J. Neurosci., 12: axonal regeneration but are not required for survival of adult 1467-1476. sensory neurons. J. Neurosci., 8t2394-2405. Diamond, J., M. Holmes, and M. Coughlin 199213 Endogenous NGF MacGrogah, D., J.P. Saint-Andre, and E. Dicou 1992 Expression of nerve growth factor and nerve growth factor receptors genes in and nerve impulses regulate the collateral sprouting of sensory axons in the skin of the adult rat. J. Neurosci., 12t1454-1466. human tissues and in prostatic adenocarcinoma cell lines. J. NeuDi Marco, E., P.C. Marchisio, S. Bondanza, A.T. Franzi, R. Cancedda, rochem., 59t1381-1391. and M. De Luca 1991 Growth-regulated synthesis and secretion Malinovsky, L. 1986 Mechanoreceptors and free nerve endings. In: of biologically active nerve growth factor by human keratiBiology of the Integument. Vol. 2. Vertebrates. J. Bereiter-Hahn, nocytes. J. Biol. Chem., 266t21718-21722. A.G. Matoltsy, and K.S. Richards, eds. Springer-Verlag, BerlinDi Marco, E., N. Cutuli, L. Guerra, R. Cancedda, and M. De Luca 1993 Heidelberg, pp. 535-560. Molecular cloning of trkE, a novel trk-related putative tyrosine Martin-Zanca, D., M. Barbacid, and L.F. Parada 1990 Expression of kinase receptor isolated from normal human keratynocytes and the trk proto-oncogene is restricted to the sensory cranial and widely expressed by normal human tissues. J. Biol. Chem., 268: spinal ganglia of neural crest origin in mouse development. 24290-24295. Genes Dev., 4t683-694. Dubovy, P., I. Svizenska, and J.A. Vega 1990 Non-specific cholinest- Meakin, S.O., and E.M. Shooter 1992 The nerve growth factor family erase activity in mouse suinal ganglia. The usefulness of histoof receptors. Trends Neurosci., 15t323-331. chemical stuhy and image analysis Tor simple characterization of Messina, A., and C. Bell 1991 Are genetically hypertensive rats defineuron subclasses. Cell. Mol. Biol., 36.23-40. cient in nerve growth factor? NeuroReport, 2.45-48. Ernfors, P., C. Wetmore, M. Eriksdotter-Nilsson, M. Bygdeman, I. Munger, B.L., and C. Ide 1988 The structure and function of cutaneStromberg, L. Olson, and H. Persson 1991 The nerve growth facous sensory receptors. Arch. Histol. Cytol., 51:l-34. tor receptor gene is expressed in both neuronal and non-neuronal Represa, J., and P. Bernd 1989 Nerve growth factor and serum differentially regulate development of the embryonic otic vesicle tissues in the human fetus. Int. J. Dev. Neurosci., 9t57-66. and cochleovestibular ganglia in vitro. Dev. Biol., 134t21-29. Ernfors, P., C.M. Rosario, J.P. Merlio, G. Grant, H. Aldskogius, and H. Reynolds, M.L., and C.J. Woolf 1993 Reciprocal Schwann cell-axon Persson 1993 Expression of mRNA for neurotrophin receptors in interaction. Curr. Opin. Neurobiol., 3t683-693. the dorsal root ganglion and spinal cord during development and Ribeiro da Silva, A,, R.L. Kenigsberg, and A.C. Cuello 1991 Light and following peripheral or central axotomy. Mol. Brain Res., 17t217electron microscopic distribution of nerve growth factor-like IR in 226. the skin of the rat lower lip. Neuroscience, 43t631-646. Fyffe, R.W. 1992 Laminar organization of primary afferent terminations in the mammalian spinal cord. In: Sensory Neurons. Diver- Ritter, A.M., G.R. Lewin, N.E. Kremer, and L.M. Mendell 1991 Requirement for nerve growth factor in the development of myelisity, Development, and Plasticity. S.A. Scott, ed. Oxford Univernated nociceptors in vivo. Nature, 350t500-503. sity Press, New York, pp. 131-139. Glass, D.J., S.H. Nye, P. Hantzopoulos, M.J. Macchi, S.P. Squinto, M. Rodriguez-Tebar, A., G. Dechant, R. Gotz, and Y.-A. Barde 1992 Binding of neurotrophin-3 to its neuronal receptors and interactions Goldfarb, and G.D. Yancopoulos 1991 trkB mediates BDNFINT-3 with nerve growth factor and brain-derived neurotrophic factor. dependent survival and proliferation in fibroblasts lacking the EMBO J., 11t917-922. low affinity NGF receptor. Cell, 66:405-413. Gold, B.G., W.C. Mobely, and S.F. Matheson 1991 Regulation of ax- Rohrer, H., R. Heumann and H. Thoenen 1988 The synthesis of nerve onal calibers, neurofilament content, and nuclear localization in growth factor (NGF) in developing skin is independent of innermature sensory neurons by nerve growth factor. J. Neurosci., vation. Dev. Biol., 128t240-244. llt943-955. Ross, A.H. 1991 Identification of tyrosine kinase trk as a nerve growth Halata, Z., M. Grim, and B. Christ 1990 Origin of spinal cord factor receptor. Cell Regul., 2t685-690. meninges, sheaths of peripheral nerves, and cutaneous receptors Scarisbrick, LA., E.G. Jones, and P.J. Isackson 1993 Coexpression of including Merkel cells. Anat. Embryol., 182:529-537. mRNAs for NGF, BDNF, and NT-3 in the cardiovascular system Haro, J.J., J.A. Vega, M.E. del Valle, B. Calzada, D. Zaccheo, and L. of the pre-and postnatal rat. J. Neurosci., 13r875-893. Malinovsky 1991 Immunohistochemical study of sensory nerve Schecterson, L.C., and M. Bothwell 1992 Novel roles for neurotroformations in human glabrous skin. Eur. J. Morphol., 29.271phins are suggested by BDNF and NT-3 mRNA expression in 284. developing neurons. Neuron, 9t449-463. Hempstead, B.L., D. Martin-Zanca, D.R. Kaplan, L.F. Parada, and Sobue, G., T. Yasuda, T. Mitsuma, and D. Pleasure 1989 Nerve growth M.V. Chao 1991 High-affinity NGF binding requires coexpression factor receptor IR in the neuronal perikarya of human sensory of the trk protoncogene and the low-affinity NGF receptor. Naand sympathetic nerve ganglia. Neurology, 39t937-941. ture, 350t678-683. Thoenen, H., C. Bandtlow, R. Heumann, D. Lindholm, M. Meyer, and Ibanez, C.F., T. Ebendal, G. Barbany, J. Murray-Rust, T.L. Blundell, H. Rohrer 1992 Nerve growth factor: Cellular localization and and H. Persson 1992 Disruption of the low affinity receptor-bindregulation of synthesis. Cell. Mol. Neurobiol., 8r35-40. Ueyama, T., T. Hano, M. Hamada, I. Nishio, and Y. Masuyama 1991 ing site in NGF allows neuronal survival and differentiation by binding to the trk gene product. Cell, 69t329-341. New role of nerve growth factor: An inhibitory neuromodulator of Isaacson, L.G., B.N. Saffran, and K.A. Crutcher 1990 Intracerebral adrenergic transmission. Brain Res., 559t293-296. NGF infusion induces hyperinnervation of cerebral blood vessels. Vega, J.A., M.E. del Valle, B. Calzada, and J.C. Alvarez-Mendez 1991 Neurobiol. Aging, Ilt51-55. Immunohistochemical localization of S-100 protein subunits (a and B) in dorsal root ganglia of the rat. Cell. Mol. Biol., 37t173Kaplan, D.R., B.L. Hempstead, D. Martin-Zanca, M.V. Chao, and L.F. 181. Parada 1991 The trk proto-oncogene product: A signal transducing receptor for nerve growth factor. Science, 252t554-558. Vega, J.A., J.J. Haro, A. de Lamo, M. Ordieres, M.E. del Valle, and B. Klein, R., S. Jing, V. Nanduri, E. ORourke, and M. Barbacid 1991 Calzada 1992 Distribution of low affinity nerve growth factor The trk proto-oncogene encodes a receptor for nerve growth facreceptor (NGFr) immunoreactivity in the human digital skin. tor. Cell, 65t189-197. Eur. J. Dermatol., 2:509-516. Lawson, S.N., and P.J. Waddell 1991 Soma neurofilament IR is re- Vega, J.A., M.E. del Valle, J.J. Haro, B. Calzada, S. Suarez-Garnacho, lated to cell size and fiber conduction velocity in rat primary and L. Malinovsky 1993 Nerve growth factor receptor immunosensory neurons. J. Physiol, 435.41-63. reactivity in Meissner and Pacinian corpuscles of the human digLee, K.-F., E. Li, L.J. Huber, S.C. Landis, A.H. Sharpe, M.V. Chao, ital skin. Anat. Rec., 236t730-736. and R. Heanisch 1992 Targeted regulation of the gene encoding Verge, V.M., J.P. Merlio, J. Grondin, P. Ernfors, H. Persson, R.J. the low affinity nerve growth factor receptor p75 leads to deficits Riopelle, T. Hokfelt, and P.M. Richardson 1992 Colocalization of in the peripheral sensory nervous system. Cell, 69t737-749. NGF binding sites, trk mRNA, and low-affinity NGF receptor Levi-Montalcini, R. 1987 The nerve growth factor 35 years later. ScimRNA in primary sensory neurons. J. Neurosci., 12t4011-4022. ence, 237t1154-1162. Wetmore, C., M. Bygdeman, and L. Olson 1992 High and low affinity Levi-Montalcini, R., and P. Calissano 1986 Nerve growth factor as a NGF receptor mRNAs in human foetal spinal cord and ganglia. paradigm for other polypeptide growth factors. Trends Neurosci., NeuroReport, 3t689-692. 9,473-477. Wyatt, S., E.M. Shooter, and A.M. Davies 1990 Expression of NGF Lewin, G.R. and L.M. Mendell 1993 Nerve growth factor and nocireceptor gene in sensory neurons and their cutaneous targets ception. Trends Neurosci., 16t353-359. prior and during innervation. Neuron, 2r421-427. Lewin, G.R., A.M. Ritter, and L.M. Mendell 1992 On the role of nerve Yaar, M., K. Grossman, M. Eller, and B.A. Gilchrest 1991 Evidence growth factor in the development of myelinated nociceptors. J. for nerve growth factor-mediated paracrine effects in human epiNeurosci., 12r1896-1905. dermis. J. Cell Biol., 115t821-828. 588 J.A. VEGA ET AL Yamamoto, M., H. Kondo, and S. Iseki 1992 Nerve growth factor receptor (NGFR)-likeimmunoreactivity in the perineurial cell in normal and sectioned peripheral nerves of rats. Anat. Rec., 233: 301-308. Yamamoto, M., G. Sobue, M. Li, Y. Arakawa, T. Mitsuma, and K. Kimata 1993 Nerve growth factor (NGF), brain derived neurotrophic factor (BDNF) and low-affinity nerve growth factor receptor (LNGFR) mRNA levels in cultured r a t Schwann cells; differ- ential time- and dose-dependent regulation by CAMP. Neurosci. Lett., 152:37-40. Yoshida, K., and F.H. Gage 1992 Cooperative regulation of nerve growth factor synthesis and secretion i n fibroblasts and astrocytes by fibroblastic growth factor and other cytokynes. Brain Res., 569:14-25. Zettler, C., R. Head, and R.A. Rush 1991 Chronic nerve growth factor treatment of normotensive rats. Brain Res., 533t251-262.