ErbB Transmembrane Tyrosine Kinase Receptors Are Expressed by Sensory and Motor Neurons Projecting into Sciatic Nerve

Abstract

Adult spinal cord motor and dorsal root ganglion (DRG) sensory neurons express multiple neuregulin-1 (NRG-1) isoforms that act as axon-associated factors promoting neuromuscular junction formation and Schwann cell proliferation and differentiation. NRG-1 isoforms are also expressed by muscle and Schwann cells, suggesting that motor and sensory neurons are themselves acted on by NRG-1 isoforms produced by their peripheral targets. To test this hypothesis, we examined the expression of the NRG-1 receptor sub-units erbB2, erbB3, and erbB4 in rat lumbar DRG and spinal cord. All three erbB receptors are expressed in these tissues. Sciatic nerve transection, an injury that induces Schwann cell expression of NRG-1, alters erbB expression in DRG and cord. Virtually all DRG neurons are erbB2- and erbB3-immunoreactive, with erbB4 also detectable in many neurons. In spinal cord white matter, erbB2 and erbB4 antibodies produce dense punctate staining, whereas the erbB3 antibody primarily labels glial cell bodies. Spinal cord dorsal and ventral horn neurons, including α-motor neurons, exhibit erbB2, erbB3, and erbB4 immunoreactivity. Spinal cord ventral horn also contains a population of small erbB3+/S100β+/GFAP– cells (GFAP-negative astrocytes or oligodendrocytes). We conclude that sensory and motor neurons projecting into sciatic nerve express multiple erbB receptors and are potentially NRG-1 responsive.

Keywords

neuregulin glial growth factor heregulin regeneration acetylcholine receptor-inducing activity

Embryonic and neonatal dorsal root ganglion (DRG) sensory and spinal cord motor neurons express multiple members of the neuregulin-1 (NRG-1) family of growth and differentiation factors (Chen et al. 1994; Corfas et al. 1995; Ho et al. 1995). These NRG-1 proteins exert important effects on developing neuromuscular junctions and Schwann cells. At developing neuromuscular junctions, NRG-1 stimulates the expression of synapse-associated proteins such as acetylcholine receptor subunits and sodium channels that are necessary for the formation of functional synapses (Fischbach and Rosen 1997). Axon-associated NRG-1 also regulates the survival, proliferation, and differentiation of Schwann cell precursors (Dong et al. 1995). In neonatal animals, NRG-1 “matches” the number of Schwann cells and axons by promoting the proliferation, migration, and survival of axon-associated committed immature Schwann cells (Topilko et al. 1996; Garratt et al. 2000). Based on these observations, the concept has arisen that NRG-1 acts developmentally as a neuronally produced protein that is transported anterogradely along axons, ultimately acting to regulate the phenotype of peripheral targets such as skeletal muscle and Schwann cells.

Several lines of evidence indicate that NRG-1 continues to play an important role in regulating the phenotype of adult skeletal muscle and Schwann cells. Mice heterozygous for a null mutation of the NRG-1 locus develop a myasthenic phenotype and decreased postsynaptic acetylcholine receptor densities, demonstrating that NRG-1 is necessary for the maintenance of nerve–muscle synapses throughout life (Sandrock et al. 1997). Furthermore, NRG-1 expression is induced in injured adult rat sciatic nerve coincident with the onset of Schwann cell mitogenesis (Carroll et al. 1997) and NRG-1 triggers the demyelination of mature differentiated Schwann cells in vitro (Zanazzi et al. 2001), suggesting that NRG-1 promotes Schwann cell proliferation and dedifferentiation in injured nerve. Consistent with this hypothesis, transgenic mice over-expressing the NRG-1 isoform glial growth factor (GGF) β3 develop demyelinative hypertrophic neuropathies, prominent Schwann cell hyperplasia, and malignant peripheral nerve sheath tumors (Huijbregts et al. 2003). Interestingly, the NRG-1 proteins eliciting these effects in adult skeletal muscle and Schwann cells may be derived from multiple sources. Adult DRG sensory and spinal cord motor neurons continue to express NRG-1 [primarily members of the sensory and motor neuron-derived factor (SMDF) NRG-1 subfamily] (Bermingham-McDonogh et al. 1997), indicating that axon-derived NRG-1 is potentially available to the peripheral targets of these neurons. However, adult skeletal muscle and Schwann cells also express NRG-1 isoforms (predominantly members of the GGF NRG-1 subfamily) (Carroll et al. 1997; Rimer et al. 1998), suggesting that NRG-1 may act in an autocrine or paracrine fashion on these cell types.

An alternative possibility is that NRG-1 derived from skeletal muscle or Schwann cells targets associated sensory and motor neurons. This postulate is consistent with observations that other adult neurons express the erbB receptors necessary for NRG-1 responsiveness (Gerecke et al. 2001) and that NRG-1 acts on at least some of these neurons to induce myriad effects such as increased neurotransmitter receptor expression (Ozaki et al. 1997; Yang et al. 1998; Rieff et al. 1999), modulation of ion channel function (Subramony and Dryer 1997; Cameron et al. 2001), and promotion of neurite outgrowth (Bermingham-McDonogh et al. 1996). It is therefore reasonable to hypothesize that NRG-1 produced by Schwann cells or skeletal muscle may act on the sensory and motor neurons innervating these cells to elicit as yet unknown effects. If this hypothesis is correct, then these neurons should express one or more of the erbB membrane tyrosine kinases mediating NRG-1 actions. To test this postulate, we have examined the expression of erbB2, erbB3, and erbB4 by neurons in adult rat lumbar DRG and spinal cord. Because Schwann cell expression of NRG-1 potentially targeting DRG sensory and spinal cord motor neurons is induced after peripheral nerve injury, it is possible that neuronal expression of the erbB kinases mediating NRG-1 responses may also be altered after axotomy. Consequently, our studies included an examination of erbB receptor expression in DRG sensory and spinal cord motor neurons after sciatic nerve transection.

Materials and Methods

Animals and Tissue Preparation

The Institutional Animal Care and Use Committee of the University of Alabama at Birmingham approved protocols for these experiments. Rats were maintained in accordance with the guidelines of the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Adult (200–300 g) male Sprague-Dawley rats were obtained from Charles River Laboratories (Wilmington, MA). Animals were maintained at constant temperature and humidity and kept on a 12-hr day/12-hr night schedule. Rats were anesthesized by inhalation of methoxyfluorane (Metofane). For animals receiving axotomies, the sciatic nerve was exposed in the midgluteal region, transected at the sciatic notch, and reflected caudally to prevent regeneration. Wounds were sutured shut and animals allowed to recover. Water and food were provided ad libitum until animals were sacrificed. Each experimental group used for immunoblotting and immunohistochemical analyses contained a minimum of three animals.

Northern Blotting Analysis

Total cellular RNA isolated from tissues by the technique of Chomczynski and Sacchi (1987) was blotted onto nylon membranes and hybridized to ³²P-labeled probes following our previously described methodology (Carroll et al. 1997). The erbB3 cDNA probe used for Northern blotting is a fragment from clone pSLC138 that corresponds to the portion of the mRNA beginning with sequences encoding sub-domain III of the extracellular ligand-binding domain and extending to the 3′ untranslated region.

RNase Protection Assays

The template used to produce a ³²P-labeled riboprobe for RNase protection assays was pSLC112 (encoding a portion of the erbB4 autophosphorylation domain) (Carroll et al. 1997). ³²P-labeled riboprobe transcribed from linearized template with T7 RNA polymerase was purified from 5% acrylamide gels containing 8 M urea. After initial optimization of RNase concentrations and hybridization conditions, RNase protection assays were performed with an RPA II kit according to the manufacturer's recommendations (Ambion; Austin, TX). Protected fragments were resolved on 5% acrylamide gels containing 8 M urea. Gels were dried and exposed to Kodak XAR-5 film at −80C to visualize reaction products. Size standards used in these experiments were pBluescript KS (+) digested with Hpa II and ³²P-end-labeled with DNA polymerase I (Klenow fragment).

Antisera and Immunohistochemical Reagents

Rabbit polyclonal antibodies and corresponding antigenic peptides for the carboxy terminal domains of erbB2/c-neu (sc-284), erbB3 (sc-285), and erbB4 (sc-283) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). A second rabbit polyclonal anti-erbB3 antiserum (E-38530) recognizing an epitope in extracellular domain II was purchased from Transduction Laboratories (Lexington, KY). The anti-glial fibrillary acidic protein (GFAP; mouse monoclonal G-A-5) antibody was from Sigma Chemical (St Louis, MO). Mouse MAbs recognizing NeuN (MAB377) and S-100β (MAB079) were obtained from Chemicon International (Temecula, CA). Non-immune rabbit and mouse IgGs were obtained from Santa Cruz Biotechnology and Jackson Immunoresearch Laboratories (West Grove, PA), respectively. Horseradish peroxidase (HRP)-conjugated donkey anti-rabbit secondary antibody was from Santa Cruz Biotechnology. Texas Red-conjugated secondary antibodies were obtained from Jackson Immunoresearch Laboratories. Tyramide signal amplification reagents, including streptavidin-horseradish peroxidase, streptavidin-fluorescein, biotinyl tyramide, amplification diluent, and blocking reagent were purchased from Perkin-Elmer Life Science Products (Renaissance TSA-Indirect kit; Boston, MA). DAB-peroxidase substrate (SK-4100) was obtained from Vector Laboratories (Burlingame, CA).

Immunoblotting

Lysates for immunoblotting were prepared by homogenizing tissue (each specimen being a pool of tissue from a minimum of three animals) in 19 volumes of ice-cold HES buffer [20 mM HEPES (pH 7.4)/1 mM EDTA/250 mM sucrose] containing 2 μg/ml aprotinin and 2 mM phenylmethylsulfonyl fluoride. Protein concentrations were determined using a modified Lowry method. Samples were resolved on 8% SDS-polyacrylamide gels and then immunoblotted and probed as previously described (Carroll et al. 1997; Gerecke et al. 2001). Immunoreactive species were identified by enhanced chemiluminescence.

To verify equivalent loading, residual protein in blotted gels was Coomassie-stained. Stained gels were then scanned and the amount of stained protein between the 79- and 225-kD size markers (Biorad Kaleidoscope markers; this range encompasses the sizes of the erbB receptors and their proteolytic cleavage products) determined using ImagePro Plus software (version 4.1; Media Cybernetics, Silver Springs, MD). The levels of erbB-immunoreactive species were then likewise determined and normalized to the measured levels of Coomassie-stained protein.

Immunohistochemical Staining

For preparation of tissues, animals were anesthetized by methoxyfluorane inhalation and transcardially perfused with 0.85% saline, followed by freshly prepared 4% paraformaldehyde in PBS. Tissues of interest were dissected and post-fixed overnight at 4C in 4% paraformaldehyde in PBS. For preparation of frozen sections, tissues were then rinsed three times with ice-cold PBS and transferred to 30% sucrose in PBS. After 24–36 hr at 4C, tissues were embedded in OCT. For paraffin sections, tissues fixed overnight in 4% paraformaldehyde were rinsed with PBS as described above and then transferred to 70% ethanol. Tissues were then processed through graded alcohols and xylenes and paraffin-embedded.

Single-label IHC for detection of erbB receptor subunits in paraformaldehyde-fixed, sucrose-cryoprotected sections was performed on 10-μm cryostat sections as we have previously described (Carroll et al. 1997). Briefly, sections were incubated with primary antibody overnight at 4C. The next morning, slides were washed three times with PBS/Tween and then incubated for 2 hr at room temperature with Cy3-conjugated or fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit (1:500 dilution in TNB) secondary antibody. After three more washes with PBS/Tween, sections were mounted in 1:1 PBS:glycerol and coverslipped. Sections were examined by epifluorescent microscopy, using cubes for Cy3 or FITC (Leica; Wetzlar, Germany).

Single-label IHC for detection of erbB receptor subunits in paraffin sections was performed on 4–5-μm sections using our previously described methods (Gerecke et al. 2001; Chaudhury et al. 2003). We have demonstrated that the distribution of erbB immunoreactivity detected using this methodology coincides with the expression of erbB mRNAs detected by in situ hybridization (Gerecke et al. 2001). Briefly, paraffin sections were deparaffinized in xylenes and rehydrated by passage through graded alcohols. Slides were gently boiled in 10 mM citrate buffer (pH 6.0) for 15 min, allowed to cool at RT for 1 hr, and then rehydrated in PBS for 15 min. Nonspecific binding was blocked by incubating sections at RT for 15 min in TNB blocking buffer [0.1 M Tris-HCl, 0.15 M NaCl, 5% blocking reagent (pH 7.5; Renaissance TSA-Indirect kit, NEN Life Science Products, Boston MA)]. Primary antibodies were used at the following dilutions in TNB blocking buffer: erbB2 1:100; erbB3 1:500; and erbB4 1:200. Sections were incubated with primary antibodies at 4C overnight. The next morning, sections were rinsed three times with PBS containing 0.05% Tween-20 (PBS/Tween) and then incubated at RT for 2 hr with HRP-conjugated secondary antibody (diluted 1:500 in TNB). Sections were washed three times with PBS/Tween at RT, followed by incubation with biotinyl tyramide (diluted 1:50 in Amplification Diluent; Renaissance TSA-Indirect kit) for 10 min at RT. Sections were again washed three times at RT in PBS/Tween and then incubated at RT for 1 hr with HRP-conjugated streptavidin (1:100 dilution in TNB). After three more rinses in PBS/Tween, immunoreactivity was detected by peroxidase-mediated diaminobenzidine deposition. Immunostained sections were mounted in Permount and cover-slipped for light microscopic examination.

For double-label IHC experiments, initial specimen processing and staining were performed as described above. ErbB receptor-specific rabbit polyclonal antibodies were used at the same concentrations as above in combination with mouse MAbs recognizing S100β (1:1000 dilution), GFAP (1:1000 dilution), or NeuN (1:150 dilution). Signals were detected using Texas Red- and FITC-labeled reagents as previously described (Gerecke et al. 2001). Briefly, sections were incubated with both primary antibodies overnight at 4C. The next morning, slides were washed three times with PBS/Tween and then incubated for 2 hr at RT with a combination of Texas Red-conjugated goat anti-mouse (1:200 dilution in TNB) and HRP-conjugated donkey anti-rabbit (1:500 dilution in TNB) secondary antibodies. After three washes with PBS/Tween, sections were incubated for 10 min with biotinyl tyramide (1:50 dilution in Amplification Diluent). Sections were washed three times with PBS/Tween and then incubated for 45 min with FITC-conjugated streptavidin. After three more washes with PBS/Tween, sections were mounted in 1:1 PBS:glycerol and coverslipped. Sections were examined by epifluorescent microscopy, using cubes for Texas Red and FITC (Leica; Wetzlar, Germany). In these experiments, the specificity of observed staining was confirmed by replacing the primary antibodies with non-immune rabbit and mouse IgGs.

Confocal immunofluorescent images were obtained using a liquid-cooled CCD high-resolution monochromatic digital camera (model CH250; Photometrics, Tucson, AZ) coupled to a Leitz Orthoplan microscope. It was confirmed that this optical setup resulted in no fluorescent bleed-through between channels. Images were processed using IPLab Spectrum software (Scanalytics; Fairfax, VA).

Results

ErbB2, ErbB3, and ErbB4 Are Differentially Expressed in Rat Lumbar DRG and Spinal Cord

Because NRG-1 binds directly to erbB3 and erbB4, at least one of these receptors must be expressed by a cell if it is to respond to NRG-1. To determine whether erbB3 and/or erbB4 mRNAs are expressed in lumbar spinal cord or DRG, total cellular RNA from these tissues was blotted and hybridized to erbB3 and erbB4 cDNA probes. ErbB3 transcripts were readily detectable in both spinal cord and DRG, being evident as a major 5.4-kb transcript, with lesser amounts of a 4.2kb species also present (Figure 1A, Lanes 1 and 5); these mRNAs co-migrated with erbB3 transcripts detected in cerebrum, midbrain, and brainstem (Figure 1A, Lanes 2–4), three CNS regions in which we have previously demonstrated erbB3 (Gerecke et al. 2001). We did not detect erbB4 transcripts in Northern blots of lumbar spinal cord and DRG RNA (data not shown). However, with RNase protection assays, a technique that has greater sensitivity, erbB4 transcripts were evident in both spinal cord and DRG (Figure 1B, Lanes 2 and 9). The specificity of the protected species identified in these experiments was confirmed by the detection of protected species of the same size in several brain regions (Figure 1B, Lanes 1 and 5–8) and the absence of a similar protected species in JS1 cells (Figure 1B, Lane 4), which do not express erbB4 (Frohnert et al. 2003), and in yeast tRNA (data not shown). PCR analyses also confirmed the expression of erbB3 and erbB4 mRNAs in lumbar spinal cord and DRG, and showed that transcripts for erbB2, a co-receptor for erbB3 and erbB4, are also present in these tissues (data not shown).

To assess erbB receptor expression in DRG in the uninjured state and after axon transection, the sciatic nerve of adult male Sprague-Dawley rats was surgically transected and the L4, L5, and L6 DRG, which contain the neurons projecting into the sciatic nerve, were collected from these animals. Lysates were prepared from non-injured and postaxotomy ganglia, immunoblotted, and probed with antibodies recognizing the carboxy terminal (autophosphorylation) domains of erbB2 (amino acids 1169–1186), erbB3 (amino acids 1307–1323), and erbB4 (amino acids 1291–1308). In initial experiments examining tissues collected from uninjured rats and animals 1 day after sciatic transaction, the erbB2 and erbB3 antibodies were found to label a major immunoreactive species of the expected 185-kD size in uninjured DRG lysates (Figures 2A and 2B). Comparison of the signals obtained in these two specimens suggested that the expression of erbB2 and erbB3 was slightly reduced 1 day after sciatic transection (Figures 2A and 2B). Examining the expression of erbB2 and erbB3 in lumbar DRG at later postaxotomy intervals (3, 5, 7, 10, 18, or 30 days after axotomy), we found that the expression of these kinases continued to be modestly reduced up to 30 days after axotomy. ErbB4 expression was also evident in lumbar DRG lysates. The anti-erbB4 antiserum labeled a major protein of the expected 185-kD size in DRG lysates as well as a smaller (~80-kD) band (Figure 2C, arrow) whose size is consistent with a previously described proteolytic cleavage product of erbB4 (Zhou and Carpenter 2000). In contrast to erbB2 and erbB3, erbB4 expression was modestly increased by sciatic nerve transaction, with an increase in the levels of the 185-kD species becoming evident 3 days after axotomy and an increase in the levels of the 80-kD erbB4-immunoreactive protein seen by 10 days after nerve injury.

Figure 1

(A) Northern blot analysis of erbB3 mRNA in lumbar spinal cord (Sp Cord), cerebellum, midbrain, brainstem, and lumbar dorsal root ganglia (DRG). A 5.4-kb species (indicated by an arrow to the left of the autoradiograph) is the major erbB3 transcript detected in each RNA (10 μg total RNA loaded per lane). Lower levels of a 4.2-kb species (also indicated by an arrow to the left of the autoradiograph) were also detected in each of these RNAs; this smaller mRNA may represent an alternatively spliced mRNA previously found to encode only the extracellular domain of erbB3 (Katoh et al. 1993). (B) RNase protection assays performed to detect erbB4 mRNA in 10 μg of total cellular RNA isolated from whole brain (Brain), lumbar dorsal root ganglia (DRG), peritoneal macrophages (Macrophage), JS1 malignant peripheral nerve sheath cells (JS1), brainstem, cerebellum, midbrain, cerebrum, and lumbar spinal cord (Sp Cord). ErbB4 transcripts, when present, are evident as a 550-base protected fragment (indicated by an arrow to the left of the autoradiograph).

To examine the expression of erbB receptors in uninjured and postaxotomy spinal cord, the lumbar enlargement of the spinal cord (the portion of the cord containing the motor neurons projecting into the sciatic nerve) was also collected from the animals described above. Lysates of these tissues were immunoblotted and probed with anti-erbB2, -erbB3, and -erbB4 antibodies. A major erbB2-like species of the expected 185-kD size was detectable in lumbar spinal cord (Figure 3A, left and right panels). A postaxotomy increase in erbB2 immunoreactivity was evident in this tissue (Figure 3A, right panel), with the highest levels observed 7 and 10 days after axotomy and decreasing thereafter. A major protein of the expected 185-kD size was also detected in lumbar spinal cord with the erbB3 antibody (Figure 3B), with expression of this erbB3-immunoreactive species also increasing after sciatic nerve transection. As in the DRG, the erbB4 antibody labeled 185- and 80-kD proteins in lumbar spinal cord lysates (Figure 3C). This antibody also detected a 148-kD erbB4-immunoreactive species in these lysates. The mass of this latter species is similar to that of erbB4 protein before modification by glycosylation (146,958 daltons), suggesting that it may be minimally glycosylated erbB4; alternatively, this 148-kD erbB4-like protein may represent a distinct proteolytic product. Expression of erbB4 was also increased after sciatic nerve transection, with the highest levels of the 185-kD protein observed 3 to 5 days after nerve injury and decreasing thereafter (Figure 3C). Densitometric analyses indicated that the level of expression of the smaller erbB4-immunoreactive proteins paralleled that of the 185-kD species.

Figure 2

Immunoblotting analyses of erbB2, erbB3, and erbB4 receptor expression in rat lumbar DRG. Western blots of tissue lysates (45 μg protein loaded/lane) prepared from lumbar (L4, L5, and L6) DRG collected from uninjured rats and animals 1 day after axotomy were probed with antibodies specific for each of the erbB membrane tyrosine kinases (indicated below each panel). Major erbB2 (A)-, erbB3 (B)-, and erbB4 (C)-like proteins of the expected 185-kD size are readily detectable in these lumbar DRG lysates and show variable alterations in their levels of expression after axotomy. Arrows at left of each panel indicate the position of each expected immunoreactive species and its size (in kD). Arrowhead in C indicates the position of putative erbB4 proteolytic cleavage products. The full length of the blots is presented in this figure to demonstrate that a 185-kD species is the major protein detected by each antibody. Bar graphs to the right of each panel represent the levels of each immunoreactive species (determined by densitometry) normalized to the protein levels measured in each lane as described in Materials and Methods.

Immunoreactivity for ErbB2, ErbB3, and ErbB4 Is Variably Associated with Neurons in Adult Lumbar DRG

Although immunoblotting analyses demonstrated that erbB receptors are expressed in lumbar DRG and spinal cord, it was unclear whether or not these membrane tyrosine kinases are expressed by neurons or other cell types in these tissues. To establish the identity of the erbB-expressing cell type(s) in DRG and spinal cord, adult lumbar (L4, L5, L6) DRG and spinal cord was collected either from non-injured Sprague-Dawley rats or from animals in which the sciatic nerve had been surgically transected 5 days earlier. This time was chosen because it represents a point at which spinal cord expression of all three erbB kinases is increased. In initial experiments, paraformal-dehyde-fixed, sucrose cryoprotected, and paraformal-dehyde-fixed paraffin sections of these tissues were stained with antibodies recognizing the autophos-phorylation domains of erbB2 and erbB4. Comparison of the staining patterns produced by these antibodies in frozen and paraffin sections showed that these patterns were virtually identical. Immunoreactivity was not detected when the primary antibodies were replaced with non-immune rabbit immunoglobulin IgG. As was seen in our previous studies of erbB immunoreactivity in peripheral nerve (Carroll et al. 1997) and brain (Gerecke et al. 2001; Chaudhury et al. 2003) tissues, staining was blocked when the erbB2 and erbB4 primary antibodies were preabsorbed with their corresponding immunizing peptides but not after preincubation with unrelated peptides (data not shown).

Figure 3

ErbB2 (A), erbB3 (B), and erbB4 (C) expression in adult rat lumbar spinal cord collected from control animals (Uninjured) and rats in which the sciatic nerve has been surgically transected (time after axotomy, in days, indicated above each lane). Western blots of these tissue lysates (45 μg protein loaded/lane) were probed with antibodies specific for each erbB membrane tyrosine kinase. Arrows at left of each panel indicate the position of each expected immunoreactive species and its size (in kD). Bar graphs to the right of each panel represent the levels of the 185-kD immunoreactive species (determined by densitometry) normalized to the protein levels measured in each lane as described in Materials and Methods. The middle panel in A is a long exposure of an uninjured spinal cord lysate probed for erbB2 to better demonstrate that erbB2 is present in this tissue.

We also stained frozen and paraffin sections of lumbar DRG and spinal cord with the rabbit polyclonal antibody recognizing the carboxy terminus of erbB3. As with the erbB2 and erbB4 antibodies, the erbB3 antibody produced similar patterns of staining in frozen and paraffin sections that were not present when the primary antibody was replaced with non-immune IgG or was preincubated with the immunizing peptide. However, the erbB3 antibody produced relatively weak immunostaining in paraffin sections. We therefore further evaluated erbB3 immunoreactivity in lumbar DRG and spinal cord by staining sections with a second antibody that recognizes an epitope in extracellular domain II of erbB3. In previous studies, we have found that this erbB3 antibody selectively stains neurons in adult rat brain that contain erbB3 mRNA detectable by in situ hybridization (Gerecke et al. 2001). Comparing the staining patterns of the two erbB3 antibodies, we found that both produced highly similar staining patterns. The sole exception was that the anti-erbB3 extracellular domain antibody produced nuclear staining in some neurons that was not evident when these cells were stained with the anti-erbB3 carboxyl terminal domain antibody. Given the highly similar staining patterns of the two erbB3 antibodies and the superior morphology provided by the use of paraffin-embedded tissues, we used the erbB2 and erbB4 carboxy terminal antibodies and the anti-erbB3 extracellular domain antibody to stain paraffin sections in subsequent experiments.

To determine whether erbB kinases are expressed by DRG neurons, double-label IHC was performed on sections of these ganglia using rabbit polyclonal antibodies specific for each NRG-1 receptor subunit (erbB2, erbB3, and erbB4) in combination with a mouse MAb recognizing the neuronal marker NeuN, which strongly labels the nucleus of DRG neurons. Consistent with our previous observations in sciatic nerve (Carroll et al. 1997), erbB2 and erbB3 immunoreactivity was observed in association with ganglion-associated Schwann cells. However, a comparison of the distribution of immunoreactivity for NeuN and erbB2 demonstrated that an erbB2-like antigen is also associated with the cell bodies of the overwhelming majority of the neurons in lumbar DRG (Figures 4A–4C). Likewise, the cell bodies and nuclei of most DRG neurons were labeled by the erbB3 antibody (Figures 4D–4F), a pattern similar to what we have previously observed in specific neuron populations in the hippocampus and cerebellum (Gerecke et al. 2001). In contrast, only scattered neurons marked for erbB4 (Figures 4G–4I), with a portion of this staining evident in association with the neuronal soma. However, the strongest erbB4 immunoreactivity was associated with neuronal nuclei, consistent with previous observations indicating that the carboxyl terminal proteolytic fragment of erbB4 can be internalized to nuclei (Ni et al. 2001). No differences in the distribution of erbB2, erbB3, or erbB4 immunoreactivity were noted in comparisons of postaxotomy and non-injured DRG. An examination of the sizes of the erbB-immunoreactive DRG neurons, including those labeled by the erbB4 antibody, showed that these cells included small, medium, and large cells, indicating that erbB immunoreactivity was not confined to a particular size class of neurons. We conclude that the vast majority of sensory neurons in both non-injured and postaxotomy DRG express the erbB membrane tyrosine kinases erbB2 and erbB3, with many also expressing erbB4. Given their expression of erbB kinases, DRG neurons are potentially NRG-1-responsive.

Figure 4

ErbB receptor immunoreactivity is variably expressed by neurons in rat DRG. Double-label IHC was performed on sections of paraffin-embedded adult rat lumbar (L4, L5, and L6) DRG using antibodies specific for each of the erbB membrane tyrosine kinases [A,D,G; erbB antibodies indicated at left of each row (FITC, green)] and the neuron-specific nuclear protein NeuN [B,E,H (Cy3, red)]. Merging of the images for erbB2 (C), erbB3 (F), and erbB4 (I) with images of NeuN immunoreactivity in the same field demonstrates that neurons in the DRG are variably immunoreactive for each erbB receptor. Bars = 20 μm.

ErbB Receptors Are Expressed Throughout Adult Rat Spinal Cord

To establish which anatomic regions in adult rat lumbar spinal cord express erbB2, erbB3, and erbB4, and to determine whether there are any differences in the distribution of these membrane tyrosine kinases, we performed IHC on paraffin sections of spinal cord collected from non-injured animals and from rats 5 days after sciatic nerve transection. Immunoreactivity for erbB2 (Figure 5A), erbB3 (Figure 5B), and erbB4 (Figure 5C) is present at multiple locations in adult spinal cord. However, each of these tyrosine kinases has a distinct pattern of expression in the cord. ErbB2 immunoreactivity is prominent in spinal cord white matter (Figure 5D), being evident as punctate neuropil staining. The erbB2 antibody also labeled neuronal processes (Figure 5D, arrows) and glial cell bodies (Figure 5D, arrowheads). In contrast, the erbB3 antibody primarily labeled glial cell bodies in the white matter (Figure 5E, arrowheads). Staining for erbB4 is similar to that for erbB2, with erbB4 immunoreactivity seen both in association with glial cell bodies (Figure 5F, arrowheads) and as punctate neuropil staining.

In the gray matter of the dorsal horn, erbB2-like proteins are detectable as dense punctate immunoreactivity (Figure 5G). Again, the pattern of erbB3 labeling in the dorsal horn gray matter is distinct from that of erbB2, being evident as intense labeling of neuronal and glial cell bodies set against a background of lighter punctate staining (Figure 5H). The erbB4 antibody stains the soma of many small neurons within the dorsal horn gray matter and produces intense punctate labeling of the neuropil (Figure 5I). It is therefore apparent that these three erbB kinases also have distinct distributions in the gray matter of the dorsal horn.

Examining the gray matter of the spinal cord ventral horn, we found that, although distinct patterns of erbB immunoreactivity were again evident, all three erbB kinases were present in large motor neurons. The erbB2 antibody labeled a dense meshwork of neuronal processes in the ventral horn gray matter in addition to the cell bodies of large motor neurons (Figure 5J, arrows). Intense erbB3 immunoreactivity is present in association with the soma of large motor neurons (Figure 5K, arrows) as well as the bodies of smaller possibly glial cells in this region (Figure 5K, arrowheads). Light punctate erbB3 staining is also seen in the neuropil surrounding these labeled cell bodies. The distribution of erbB4 immunoreactivity in the ventral horn is similar to that of erbB3, except that the erbB4 antibody produces little labeling of smaller cells in this region (Figure 5L). No differences among the expression of erbB2, erbB3, and erbB4 in non-injured and 5-day postaxotomy cord were evident at this level of examination. We conclude that the large motor neurons of the spinal cord ventral horn express erbB2-, erbB3-, and erbB4-like immunoreactive proteins.

Figure 5

ErbB receptors are differentially expressed in adult rat spinal cord. (A–C) Low-power views of cross-sections of lumbar spinal cord immunostained with antibodies specific for erbB2 (A), erbB3 (B), and erbB4 (C) demonstrate that each of these membrane tyrosine kinases is expressed with distinct distributions in this tissue. Bars = 400 μm. (D–F) Higher-power views compared with A–C of the expression of erbB2 (D), erbB3 (E), and erbB4 (F) in spinal cord ventral white matter. Arrows indicate erbB2-immunoreactive neuron processes. Arrowheads indicate glial cell bodies labeled by each erbB antibody. (G–I) Higher-power views of the distribution of erbB2 (G), erbB3 (H), and erbB4 (I) receptor subunits in the dorsal horn of adult rat spinal cord. (J–L) Higher-power views of erbB2 (J), erbB3 (K), and erbB4 (L) immunoreactivity in the ventral horn of adult rat spinal cord. Arrows indicate some of the presumptive motor neurons in this spinal cord region. Arrowheads in K indicate some of the smaller erbB3-immunoreactive cells in the ventral horn. Bars = 50 μm.

ErbB3 Is Expressed by a Subset of Glia in Adult Lumbar Spinal Cord

Although it was evident that large motor neurons in the spinal cord express erbB receptors, the identity of the other smaller cells in the ventral horn that express an erbB3-like protein was unclear. One possibility was that these cells were macroglia, a class of cells that express erbB receptors in normal brain (Canoll et al. 1996; Vartanian et al. 1997; Ma et al. 1999; Gerecke et al. 2001) and whose expression of specific erbB receptors is increased when located within or adjacent to regions of brain injury (Cannella et al. 1999; Erlich et al. 2000; Tokita et al. 2001) or disease (Chaudhury et al. 2003). To test the hypothesis that these smaller erbB3-immunoreactive cells were macroglia and to determine whether their expression of erbB3 receptors was altered adjacent to motor neurons injured by sciatic axotomy, we performed a series of double-label IHC experiments using an anti-erbB3 antiserum in combination with antibodies that recognize glial markers. We first compared the distribution of erbB3 immunoreactivity with that of S100β, a marker for both astrocytes and oligodendrocytes. The anti-erbB3 antibody again labeled large motor neurons and a population of smaller cells (Figure 6A). About half of the small erbB3-immunoreactive cells in the ventral horn were also S100β-positive (Figures 6A–6C, arrows), indicating that at least some of the small erbB3-immunoreactive cells in the spinal cord ventral horn represent S100β-positive macroglia. These macroglia were distinct from the large erbB2-, erbB3-, and erbB4-positive cells in the ventral horn, consistent with our identification of these cells as large motor neurons.

Figure 6

ErbB3 immunoreactivity in the ventral horn of adult rat lumbar spinal cord is associated, in part, with S100β-immunoreactive glia. Double-label IHC was performed on sections of paraffin-embedded adult rat lumbar spinal cord nerve using antibodies recognizing erbB3 [A; FITC (green)] and the glial marker S100β [B; Cy3 (red)]. Merging of the images of immunoreactivity for erbB3 and S100β (C) shows some overlap representing co-localization [yellow (green + red) represents co-localization for erbB3 (green) and S100β (red) immunoreactivity]. Arrows indicate some of the cells in which erbB3 and S100β immunoreactivity co-localize. Bars = 20 μm.

To determine whether the erbB3-immunoreactive macroglia in the ventral horn are astrocytes, we compared the distribution of erbB3 staining to that of the astrocyte marker GFAP. GFAP labeling was associated with cells with stellate processes morphologically consistent with astrocytes (Figures 7A–7C). However, GFAP immunoreactivity was clearly distinct from erbB3 staining in the ventral horn. We conclude that glial expression of these erbB kinases in the ventral horn is not found in GFAP-positive astrocytes and hence must be associated with other glial cell types, such as GFAP-negative astrocytes (Walz and Lang 1998) or oligodendrocytes. Comparing the ventral horns ipsi- and contralateral to the site of sciatic axotomy, we found no evidence of alterations in the expression of erbB3 by S100β- or GFAP-positive cells.

Discussion

To test the hypothesis that adult sensory and motor neurons express the erbB receptors necessary for NRG-1 responsiveness, we examined the expression of erbB2, erbB3, and erbB4 in non-injured and postaxotomy lumbar DRG and the lumbar enlargement of the spinal cord. We found that all three of these erbB kinases were expressed in DRG and spinal cord, both normally and after nerve injury. IHC analyses showed that a portion of the erbB receptor expression in these tissues was associated with sensory and motor neurons, with much of the remaining erbB immunoreactivity being found in macroglia. However, the pattern of erbB expression in lumbar DRG sensory and spinal cord motor neurons was variable and complex, being differentially expressed in each neuron population and even in distinct intracellular compartments in these cells. These observations have important implications for NRG-1 actions on sensory and motor neurons.

ErbB2, erbB3, and erbB4 mRNA and protein were readily detected in lumbar DRG both in non-injured rats and after sciatic nerve transection. Although our IHC analyses indicate that a portion of the erbB2 and erbB3 protein in DRG is derived from ganglionic glia, we found that non-injured and axotomized DRG neurons also express erbB2, erbB3, and/or erbB4. By immunoblotting analysis, DRG expression of erbB2 and erbB3 was modestly decreased after sciatic nerve transaction, while this same injury was associated with an increase in the levels of intraganglionic erbB4 protein. The erbB receptors therefore resemble several other growth factor receptors [e.g., receptors for leukemia inhibitory factor (Gardiner et al. 2002) and GDNF family members (Bennett et al. 2000)] in that their expression in DRG neurons is altered after sciatic nerve transection. The continued expression of erbB receptors up to 30 days after axotomy indicates that DRG sensory neurons regenerating into injured sciatic nerve remain potentially responsive to NRG-1 throughout a period (3 to 30 days after axotomy) when NRG-1 mRNA and protein accumulate to high levels in the damaged nerve (Carroll et al. 1997). However, the differences we have observed in postaxotomy regulation of DRG erbB2, erbB3, and erbB4 expression suggest that these receptors mediate distinctly important functions in the ganglionic response to injury. Defining the function of individual erbB kinases in DRG neurons will be critically important for furthering our understanding of these observations.

Figure 7

ErbB3 immunoreactivity in the ventral horn of adult rat lumbar spinal cord is not associated with GFAP-positive astrocytes. Double-label IHC was performed on paraffin-embedded sections of adult rat lumbar spinal cord using antibodies recognizing erbB3 [A; FITC (green)] in combination with an anti-GFAP antibody [B; (Cy3, red)]. Merging of the images for erbB3 and GFAP (C) demonstrates that there is little overlap between immunoreactivity for erbB3 and GFAP. Bars = 20 μm.

There were other clear differences in the expression of specific erbB receptors in normal and postaxotomy lumbar DRG. Much like the receptors for other growth factors, such as the neurotrophins (Carroll et al. 1992; Mu et al. 1993; Molliver and Snider 1997) and members of the GDNF family (Bennett et al. 1998), erbB kinases are differentially expressed by DRG neurons, with erbB2 and erbB3 present in the majority of these cells and erbB4 expression confined to a major sub-population. Nevertheless, expression of all three erbB receptors is evident in small, medium, and large DRG neurons, suggesting that the function of at least some nociceptive, proprioceptive, and mechanoceptive neurons can be modified by NRG-1 action. Given the pattern of erbB expression in DRG neurons, it is possible that the dense punctate erbB immunoreactivity we have observed in the superficial dorsal horn of the spinal cord reflects, in part, erbB protein associated with the central projections of nociceptive DRG neurons and that the less intense punctate erbB immunoreactivity in deeper Rexed laminae is similarly associated with central projections from proprioceptive and mechanoceptive neurons. It is also apparent that at least erbB4 is differentially distributed in the intracellular compartments of DRG neurons. In this study we have shown that erbB4 immunoreactivity is associated with the soma of many DRG neurons. In contrast, we previously demonstrated that erbB4 protein is undetectable in sciatic nerve (Carroll et al. 1997). Because the sciatic nerve contains the axon projections from DRG, it is clear that erbB4 must be differentially distributed in the peripheral projections, soma, and central projections of DRG neurons that express this kinase. Given the distinct signaling capabilities of each of the erbB receptors (see below), it is likely that these differences in the expression of erbB kinases mark populations of DRG neurons that differ in their responses to NRG-1. NRG-1 actions on these cells may differ, depending on whether stimulation occurs in the peripheral projections, the soma, or the central projections of a sensory neuron.

ErbB2, erbB3, and erbB4 were also detectable in lumbar spinal cord, and the expression of these receptors was increased after surgical transection of the sciatic nerve. The increases in erbB expression observed were relatively modest and it is likely that this, together with the diffuse expression of erbB receptors in the cord, explains why it was difficult to establish a precise site or cell type specifically associated with higher levels of erbB protein. Our IHC analyses demonstrated that these erbB kinases were expressed by a variety of cell types in lumbar spinal cord, including large motor neurons in the ventral horn. In contrast to DRG sensory neurons, the cell bodies of spinal cord motor neurons were uniformly and intensely immunoreactive for erbB2, erbB3, and erbB4. Because spinal cord motor neuron expression of erbB receptors is also evident after sciatic nerve transection, we conclude that these neurons, like the DRG neurons, are potentially responsive to the NRG-1 proteins whose expression is induced in injured nerve. The expression of erbB4 by ventral horn motor neurons, considered together with the absence of detectable erbB4 in sciatic nerve, indicates that erbB4 expression is also differentially distributed in the peripheral projections, soma, and central projections of these cells.

In addition to a protein migrating at the expected size of mature erbB4 (185 kD), we observed the presence of an 80-kD erbB4-immunoreactive species in both DRG and spinal cord. The size of this immunoreactive species is consistent with that of a proteolytic fragment consisting of the transmembrane and carboxy terminal domains of erbB4 that remains associated with the cell after cleavage of the mature protein by the metalloprotease TACE/ADAM17 (Carpenter 2003). Subsequent cleavage by 7-secretase releases this fragment into the cytoplasm and allows it to be translocated into the cell nucleus (Ni et al. 2001). Our observation of erbB4 in the nuclei of DRG neurons is therefore consistent with this signaling mechanism. Interestingly, we did not find erbB4 immunoreactivity in the nucleus of α-motor neurons. Because the 80-kD fragment was present in spinal cord, this suggests that the processes involved in translocating the proteolytic fragment into the nucleus of these neurons were not activated under our experimental conditions. Alternatively, the 80-kD fragment observed in our immunoblotting analyses may be derived from cells other than α-motor neurons.

We also found that an anti-erbB3 extracellular domain antibody, but not an antibody recognizing the erbB3 cytoplasmic domain, labeled the nuclei of DRG sensory and spinal cord motor neurons. This raises the interesting question of whether a proteolytic fragment derived from the extracellular domain of erbB3 can also be translocated to the neuron nucleus to mediate presumed signaling functions. However, because the antigen against which the erbB3 extracellular domain antibody was raised was not available to us, we cannot exclude the possibility that the nuclear staining observed with this antibody is nonspecific.

The variable expression of erbB kinases among the sensory and motor neurons examined in this study is significant because each of these erbB receptor subtypes has distinct functional characteristics. Although important for NRG-1 signaling, erbB2 does not directly bind NRG-1 (Sliwkowski et al. 1994). NRG-1 actions are instead mediated when this factor binds to erbB3 or erbB4, with erbB2 activation occurring when this kinase subsequently heterodimerizes with either of these occupied receptors (Carraway and Cantley 1994). Furthermore, although either erbB3 or erbB4 may bind NRGs, these receptors have distinct characteristics. ErbB3 binds NRG-1 with an affinity an order of magnitude lower than that of erbB4 and, unlike erbB4, has little endogenous tyrosine kinase activity because two residues critical for autophosphorylation are “mutated” in this receptor (Carraway and Cantley 1994). ErbB3 also differs from erbB2 and erbB4 in its ability to recruit some signaling molecules (e.g., phosphatidylinositol-3 kinase) to the active complex (Carraway and Cantley 1994). These functional differences are further emphasized by the fact that mice with a targeted mutation of the erbB4 locus (Gassman et al. 1995) have a nervous system phenotype distinct from that of mice with null mutations of the NRG-1 (Meyer and Birchmeier 1995) or erbB2 (Lee et al. 1995) loci. It is therefore apparent that the choice of whether NRG-1 activates erbB3 or erbB4 has profound biological consequences. Therefore, the expression of different erbB kinases by a neuron or in specific compartments in the neuron (e.g., cell bodies vs processes) is likely to be a critical determinant of the responses elicited by NRG-1 in these cells.

Transection of a peripheral nerve is accompanied by several morphological and biochemical alterations in affected neurons and in the glia surrounding the soma of these neurons. These alterations include increased glial expression of some growth factors and their receptors [e.g., platelet-derived growth factor (Hermanson et al. 1995)], an observation that led us to examine erbB expression in the glia of the spinal cord ventral horn. We have found that erbB3 is widely expressed by a population of S100β-positive macroglia in this region. Interestingly, these erbB3+/S100β + macroglia are not immunoreactive for GFAP, suggesting that they represent either GFAP-negative astrocytes or oligodendrocytes. These findings are similar to those we have previously reported for macroglia in the gray matter of the brain (Gerecke et al. 2001). There was, however, no evidence of alterations in the intensity or the distribution of glial erbB expression in the ventral horn after sciatic axotomy. We conclude that alterations in the expression of erbB3 by ventral horn macroglia are likely of minimal importance to the successful regeneration of motor neuron axons. The population of erbB3+/S100β–non-neuronal cells identified in our studies are probably microglia, a cell type we have previously found to express this receptor (Gerecke et al. 2001; Chaudhury et al. 2003).

In conclusion, the NRG-1 receptors erbB2, erbB3, and erbB4 are variably expressed by lumbar DRG sensory and spinal cord motor neurons. The neuronal expression of these kinases indicates that NRG-1 proteins derived from skeletal muscle or Schwann cells in injured nerve segments may act directly on these sensory and motor neurons to elicit as yet unknown effects. Examining the responses elicited by NRG-1 in DRG sensory and spinal cord motor neurons will be of great interest and may provide important insight into the role played by these growth factors in modulating the phenotype of motor and sensory neurons in the normal and injured peripheral nervous system.

Footnotes

Acknowledgements

Supported by grants 1 R01 NS37514 (the National Institute of Neurological Disorders and Stroke) and 1 P50 AG16582 (the National Institute on Aging).

We thank K.A. Roth for his helpful comments on this manuscript.

References

Bennett

Boucher

Armanini

Poulsen

Michael

Priestley

Phillips

(2000) The glial cell line-derived neurotrophic factor family receptor components are differentially regulated within sensory neurons after nerve injury. J Neurosci 20:427–437.

Bennett

DLH

Michael

Ramachandran

Munson

Averill

Yan

McMahon

(1998) A distinct subgroup of small DRG cells express GDNF receptor components and GDNF is protective for these neurons after nerve injury. J Neurosci 18:3059–3072.

Bermingham-McDonogh

McCabe

Reh

(1996) Effects of GGF/neuregulins on neuronal survival and neurite outgrowth correlate with erbB2/neu expression in developing rat retina. Development 122:1427–1438.

Bermingham-McDonogh

Y-T

Marchionni

Scherer

(1997) Neuregulin expression in PNS neurons: isoforms and regulation by target interactions. Mol Cell Neurosci 10:184–195.

Cameron

Dryer

(2001) β-Neuregulin-1 is required for the in vivo development of functional Ca²⁺-activated K⁺ channels in parasympathetic neurons. Proc Natl Acad Sci USA 98:2832–2836.

Cannella

Pitt

Marchionni

Raine

(1999) Neuregulin and erbB receptor expression in normal and diseased human white matter. J Neuroimmunol 100:233–242.

Canoll

Musacchio

Hardy

Reynolds

Marchionni

Salzer

(1996) GGF/neuregulin is a neuronal signal that promotes the proliferation and survival and inhibits the differentiation of oligodendrocyte progenitors. Neuron 17:229–243.

Carpenter

(2003) ErbB-4: mechanism of action and biology. Exp Cell Res 284:66–77.

Carraway

KLI

Cantley

(1994) A neu acquaintance for erbB3 and erbB4: a role for receptor heterodimerization in growth signaling. Cell 78:5–8.

10.

Carroll

Miller

Frohnert

Kim

Corbett

(1997) Expression of neuregulins and their putative receptors, ErbB2 and ErbB3, is induced during Wallerian degeneration. J Neurosci 17:1642–1659.

11.

Carroll

Silos-Santiago

Frese

Ruit

Milbrandt

Snider

(1992) Dorsal root ganglion neurons expressing trk are selectively sensitive to NGF deprivation in utero . Neuron 9:779–788.

12.

Chaudhury

Gerecke

Wyss

Morgan

Gordon

Carroll

(2003) Neuregulin-1 and erbB4 immunoreactivity is associated with neuritic plaques in Alzheimer disease brain and in a transgenic model of Alzheimer disease. J Neuropathol Exp Neurol 62:42–54.

13.

Chen

Bermingham-McDonogh

Danehy

Jr Nolan

Scherer

Lucas

Gwynne

(1994) Expression of multiple neuregulin transcripts in postnatal rat brains. J Comp Neurol 349:389–400.

14.

Chomczynski

Sacchi

(1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156–159.

15.

Corfas

Rosen

Aratake

Krauss

Fischbach

(1995) Differential expression of ARIA isoforms in the rat brain. Neuron 14:103–115.

16.

Dong

Brennan

Liu

Yarden

Lefkowitz

Mirsky

(1995) Neu differentiation factor is a neuron-glia signal and regulates survival, proliferation and maturation of rat Schwann cell precursors. Neuron 15:585–596.

17.

Erlich

Shohami

Pinkas-Kramarski

(2000) Closed head injury induces up-regulation of erbB-4 receptor at the site of injury. Mol Cell Neurosci 16:597–608.

18.

Fischbach

Rosen

(1997) ARIA: a neuromuscular junction neuregulin. Annu Rev Neurosci 20:429–458.

19.

Frohnert

Stonecypher

Carroll

(2003) Constitutive activation of the neuregulin-1/erbB receptor signaling pathway is essential for the proliferation of a neoplastic Schwann cell line. Glia 43:104–118.

20.

Gardiner

Cafferty

Slack

Thompson

(2002) Expression of gp130 and leukaemia inhibitory factor receptor subunits in adult rat sensory neurones: regulation by nerve injury. J Neurochem 83:100–109.

21.

Garratt

Voiculescu

Topilko

Charnay

Birchmeier

(2000) A dual role of erbB2 in myelination and in expansion of the Schwann cell precursor pool. J Cell Biol 148:1035–1046.

22.

Gassman

Casagranda

Orioli

Simon

Lai

Klein

Lemke

(1995) Aberrant neural and cardiac development in mice lacking the erbB4 neuregulin receptor. Nature 378:390–394.

23.

Gerecke

Wyss

Karavanova

Buonanno

Carroll

(2001) ErbB transmembrane tyrosine kinase receptors are differentially expressed throughout the adult rat central nervous system. J Comp Neurol 433:86–100.

24.

Hermanson

Olsson

Westermark

Funa

(1995) PDGF and its receptors following facial nerve axotomy in rats: expression in neurons and surrounding glia. Exp Brain Res 102:415–422.

25.

W-H

Armanini

Nuijens

Phillips

Osheroff

(1995) Sensory and motor neuron-derived factor: a novel heregulin variant highly expressed in sensory and motor neurons. J Biol Chem 270:14523–14532.

26.

Huijbregts

RPH

Roth

Schmidt

Carroll

(2003) Hypertrophic neuropathies and malignant peripheral nerve sheath tumors in transgenic mice overexpressing glial growth factor β3 in myelinating Schwann cells. J Neurosci 23:7269–7280.

27.

Katoh

Yazaki

Sugimura

Terada

(1993) c-erbB3 gene encodes secreted as well as transmembrane receptor tyrosine kinase. Biochem Biophys Res Commun 192:1189–1197.

28.

Lee

K-F

Simon

Chen

Bates

Hung

M-C

Hauser

(1995) Requirement for neuregulin receptor erbB2 in neural and cardiac development. Nature 378:394–398.

29.

Hill

Creswick

Costa

Cornea

Lioubin

Plowman

(1999) Neuregulins signaling via a glial erbB-2-erbB-4 receptor complex contribute to the neuroendocrine control of mammalian sexual development. J Neurosci 19:9913–9927.

30.

Meyer

Birchmeier

(1995) Multiple essential functions of neuregulin in development. Nature 378:386–390.

31.

Molliver

Snider

(1997) Nerve growth factor receptor TrkA is down-regulated during postnatal development by a subset of dorsal root ganglion neurons. J Comp Neurol 381:428–438.

32.

Silos-Santiago

Carroll

Snider

(1993) Neurotrophin receptor genes are expressed in distinct patterns in developing dorsal root ganglia. J Neurosci 13:4029–4041.

33.

C-Y

Murphy

Golde

Carpenter

(2001) γ-Secretase cleavage and nuclear localization of erbB-4 receptor tyrosine kinase. Science 294:2179–2181.

34.

Ozaki

Sasner

Yano

Buonanno

(1997) Neuregulin-β induces expression of an NMDA-receptor subunit. Nature 390:691–694.

35.

Rieff

Raetzman

Sapp

Yeh

Siegal

Corfas

(1999) Neuregulin induces GABA_A receptor subunit expression and neurite outgrowth in cerebellar granule cells. J Neurosci 19:10757–10766.

36.

Rimer

Cohen

Lomo

Burden

McMahan

(1998) Neuregulins and erbB receptors at neuromuscular junctions and at agrin-induced postsynaptic-like apparatus in skeletal muscle. Mol Cell Neurosci 12:1–15.

37.

Sandrock

Jr Dryer

Rosen

Gozani

Kramer

Theill

Fischbach

(1997) Maintenance of acetylcholine receptor number by neuregulins at the neuromuscular junction in vivo. Science 276:599–603.

38.

Sliwkowski

Schaefer

Akita

Lofgren

Fitzpatrick

Nuijens

Fendly

(1994) Coexpression of erbB2 and erbB3 proteins reconstitutes a high affinity receptor for heregulin. J Biol Chem 269:14661–14665.

39.

Subramony

Dryer

(1997) Neuregulins stimulate the functional expression of Ca²⁺-activated K⁺ channels in developing chicken parasympathetic neurons. Proc Natl Acad Sci USA 94:5934–5938.

40.

Tokita

Keino

Matsui

Aono

Ishiguro

Higashiyama

Oohira

(2001) Regulation of neuregulin expression in the injured rat brain and cultured astrocytes. J Neurosci 21:1257–1264.

41.

Topilko

Murphy

Charnay

(1996) Embryonic development of Schwann cells: multiple roles for neuregulins along the pathway. Mol Cell Neurosci 8:71–75.

42.

Vartanian

Goodearl

Viehover

Fischbach

(1997) Axonal neuregulin signals cells of the oligodendrocyte lineage through activation of HER4 and Schwann cells through HER2 and HER3. J Cell Biol 137:211–220.

43.

Walz

Lang

(1998) Immunocytochemical evidence for a distinct GFAP-negative subpopulation of astrocytes in the adult rat hippocampus. Neurosci Lett 257:127–130.

44.

Yang

Kuo

Devay

Role

(1998) A cysteine-rich isoform of neuregulin controls the level of expression of neuronal nicotinic receptor channels during synaptogenesis. Neuron 20:255–270.

45.

Zanazzi

Einheber

Westreich

Hannocks

M-J

Bedell-Hogan

Marchionni

Salzer

(2001) Glial growth factor/neuregulin inhibits Schwann cell myelination and induces demyelination. J Cell Biol 152:1289–1299.

46.

Zhou

Carpenter

(2000) Heregulin-dependent trafficking and cleavage of erbB-4. J Biol Chem 275:34737–34743.