During embryonic development of the nervous system, neurons compete for a limited supply of survival factors provided by their target tissues. Nerve growth factor (NGF) was the first identified member of structurally related secreted proteins, termed neurotrophins (Levi-Montalcini, 1987). To date, four members of the NGF family of neurotrophic factors, NGF, brain-derived neurotrophic factor (BDNF) (Barde et al., 1982), neurotrophin-3 (NT-3) (Ernfors et al., 1990; Jones and Reichardt, 1990), and neurotrophin-4/5 (NT4/5) (Berkemier et al., 1991; Ip et al., 1992), have been identified in mammals. These molecules exert their effects by binding specific receptor protein tyrosine kinases (the Trk family), and each neurotrophin also interacts with a low-affinity receptor, p75NTR. NGF binds and activates TrkA receptor, BDNF and NT-4 bind TrkB receptor, and NT-3 predominantly binds TrkC receptor, but can also interact with TrkA and TrkB receptors (Barbacid, 1994; Davies et al., 1995; Fariñas et al., 1998). Internalized receptor-ligand complexes are transported to the neuronal soma where they activate a cascade of proto-oncogene signals involved in cell proliferation and differentiation (Bothwell, 1995; Segal and Greenberg, 1996). The phenotypic consequences of null mutations of either the ngf and its high-affinity receptor trkA genes or nt-3 and trkC genes revealed selective loss of small-diameter nociceptive and thermoceptive neurons or large-diameter mechanoceptive and proprioceptive neurons, respectively (Crowley et al., 1994; Ernfors et al., 1994; Fariñas et al., 1994; Smeyne et al., 1994; Reichardt and Fariñas, 1997). Aside from their survival-promoting effects, neurotrophic factors have other important biological activities, including effects on process development, and synaptic plasticity of the nervous system (Huang and Reichardt, 2001).
Recent studies showed that NGF and NT-3 play a major role in axonal growth (Hoyle et al., 1993; Schnell et al., 1994; Zhang et al., 1994; ElShamy et al., 1996; Lentz et al., 1999; Ulupinar et al., 2000). In whole-mount explant cultures of the trigeminal pathway, the peripheral (whisker pad) and the central (brainstem) target tissues of the trigeminal ganglion (TG) are left intact, and TG cells survive and display embryonic age-specific axonal growth patterns. Using such cultures, we previously showed that in the presence of NGF, central trigeminal axons leave the trigeminal tract and grow without branching, whereas NT-3 promotes precocious arborization along the edges of the tract. However, it has been difficult to differentiate between axonal and survival effects of neurotrophins. To circumvent this problem, some investigators took advantage of mice with targeted deletion of apoptotic genes to keep neurons alive in culture without the addition of neurotrophins (White et al., 1998; Patel et al., 2000). In the absence of proapoptotic bax gene, NGF and NT-3 have been shown to promote axon elongation and arborization in dissociated dorsal root ganglion (DRG) cells (Lentz et al., 1999). However, in the TG, a significant proportion of neurons still die in the absence of bax (Middleton et al., 2000). In addition, bax forms homodimer with bcl-2, which has antagonistic action on apoptosis and could modulate the effects of bcl-2 on neuronal differentiation and axonal growth (Hilton et al., 1997; Middleton et al., 1998; Korsmeyer, 1999).
In the present study, we used whole-mount explant cultures of trigeminal pathway to gain insights into class-specific responses of TG cells to exogenous NGF or NT-3 treatments. We prepared the cultures at two different developmental ages, E13 and E15. Following retrogradely labeling of the TG cells by carbocyanine dye DiI or double labeling with TrkA and TrkC receptor antibodies, we could visualize TG cells. We found differences in the soma size distributions and Trk receptor expression of neurons under different neurotrophin treatment conditions.
MATERIALS AND METHODS
Preparation of Whole-Mount Cultures
Institutional Animal Care and Use Committees of both Osmangazi University Faculty of Medicine and Louisiana State University Health Sciences Center approved experimental procedures used in this study. Day of sperm positivity following overnight mating was designed as E0 and seven litters of embryos were used for each developmental stage. E13 and E15 embryos (Sprague-Dawley) were removed by cesarean section following euthanasia of the dam by intraperitoneal injection of a lethal dose of sodium pentobarbital (50 mg/kg body weight). Embryos were placed in sterile petri dishes containing ice-cold Gey's balanced salt solution (Gibco) supplemented with D-galactose (6.4 mg/l). All of the dissections were performed in this solution by using a stereomicroscope with dark-field illumination.
First, heads of the embryos were cut at the level of upper cervical spinal cord, then connective tissues and meninges around the brain were removed. Following a transverse cut through the pontine flexure, both hemispheres and the midbrain were taken out. To prepare the intact whole-mount cultures, trigeminal ganglia on both sides were carefully dissected with their peripheral and central projections and target fields intact. The hindbrain and maxillary processes (whisker pads) were trimmed from the rest of the head. Since the infraorbital nerve crosses into the maxillary processes just below the eye, the eyecups were left in place (Fig. 1).
Whole-mount explants were then transferred on microporous Millicell membranes (Millipore) with the ventral (meningeal) side down and the ventricular side up. In control cases, cultures were grown in serum-free medium (Collazo et al., 1992), and in experimental cases, 50 ng/ml NGF or NT-3 (Collaborative Research, and Regeneron) were added into this medium. For each condition, a total of 10 bilateral whole-mount cultures were prepared from E13 and 10 from E15 embryos. They were grown at 33°C, in a humidified CO2 incubator, for 3 days and then fixed with 4% buffered paraformaldehyde (pH 7.4, 0.1 M).
The TG neurons were retrogradely labeled by inserting a tiny fragment of DiI (1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate; Molecular Probes) into the central trigeminal tract (Fig. 1) using a stereomicroscope. Cultures were kept at 37°C for 1 more week, allowing diffusion of the lipophilic tracer. To convert fluorescent dye to a permanent label, cultures were placed in a petri dish containing 0.15% diaminobenzidine (DAB) in 0.1 M Tris buffer (pH 8.2), and epifluoresced with a rhodamine filter set (Sandell and Masland, 1988). Cultures were rinsed, mounted on subbed slides, and coverslipped with gelatin and glycerol. Images were taken with PM10SP automatic photomicrographic system attached to an Olympus CH40 microscope. Negatives were then scanned and transferred to a power PC computer. Adobe Photoshop program was used for adjustment of the brightness and contrast of the images and their photomontages.
A separate series of whole mounts (kept in vitro for 3 days) was processed for TrkA and TrkC double immunocytochemistry. Cultures were fixed with 4% paraformaldehyde in phosphate buffer, cryoprotected with 30% sucrose, and sectioned at a thickness of 25 μm using a cryostat. Following three rinses in phosphate-buffered saline (PBS; pH 7.4), sections were incubated in 10% normal donkey serum containing 0.3% Triton X-100 for 30 min at room temperature. Sections were then incubated in the mixture of rabbit anti-TrkA antibody (1:1,000) and goat anti-TrkC antibody (1:200) diluted in blocking solution overnight at 4°C. Both antibodies are gifts from L. Reichardt (University of California at San Francisco). The next day, they were rinsed and incubated in CY-3-conjugated donkey antirabbit and FITC-conjugated donkey antigoat (1:200; Chemicon) secondary antibodies for 2 hr at room temperature in the dark. The slides were then washed with PBS and coverslipped with Fluormount (Sigma). Control sections were processed as above except that the primary antibody was left out. Double-labeled images were photographed using a confocal microscope (Nikon TE-300 with Radians 2000 laser scanning system; Bio-Rad) to acquire Z-series of multiple focal planes at Z-spacing 0.68 μm under high-power (Nikon fluor 40× and 60×) oil objectives.
Morphometric Analyses of Neurons
For quantitative measurements of the neurons, systematic random sampling method was applied by using an unbiased counting frame to prevent edge effect. Neurons were examined with 40× magnification and their somas were drawn on transparent sheets with the aid of a drawing tube. The longest axis of the soma was measured as the major (a) diameter, and crossing axis of the major diameter from its midpoint at right angle was measured as the minor (b) diameter. These measurements were then used to estimate the mean diameter (d) of each profile by using the relationship d = √(a · b) (Bedi et al., 1980). A stage micrometer was used as a magnification standard for measurements and photomicrographic documentations of images. For each condition, 100 neurons were measured from E13 and 184 neurons were measured from E15 cultures.
Statistical differences between the mean diameters of neurons were determined by using one-way analysis of variance test and posthoc comparisons were done with the Tukey test. The data presented here represent means and SEMs. Soma size distributions of neurons cultured in different conditions were compared by using Kolmogorov-Simirnov test.
In a separate quantitative analysis, we sampled four nonoverlapping quadrants from each TG (from three different animals) double-immunostained with TrkA and TrkC antibodies. Confocal microscopic images taken at 40× oil immersion were used to count TrkA- and TrkC-labeled neurons with the use of Metaview Image Analysis Program (Universal Imaging). Ratios of TrkA- and TrkC-expressing neurons were determined for cultures maintained in SFM, in NGF, or NT-3 at E13 and E15 time points. Comparisons between these ratios were done using one-way analysis of variance test and posthoc comparisons were done with the Tukey test. The significance level was set at P < 0.05 level and the error bars in the graphs represent one standard deviation.
As illustrated in Figure 1, morphological characteristics of whisker pad or brainstem explants were not compromised after 3 days in culture, but they also did not follow their in vivo course of differentiation. For instance, in the maxillary processes of E13 embryos, developing whisker follicles were discernible, but at the end of culture period, they never attained an appearance like the ones prepared from E15 embryos (compare E13 and E15 cultures in Fig. 1). In E15 whisker pad explants, five rows of whisker follicles (marked by asterisks) develop and even rudimentary hairs form within the follicle cores but do not attain differentiation similar to that seen in E18 embryos. The brainstem also retains its organotypic characteristics and the central axons of the TG cells bifurcate upon entry into the hindbrain to form the ascending and descending components of the trigeminal tract (data not shown). In vivo, the central axons of TG neurons begin collateralization and arborization by E17, but in whole-mount cultures derived from E13 and E15 rats, these axons remain unbranched even after 3 days in culture (Ulupinar et al., 2000). Thus, in whole-mount trigeminal pathway cultures, morphological features of the explant tissues largely maintain their characteristics of the day they were dissected out.
Placing DiI crystals into the central trigeminal axons led to retrograde labeling of TG neurons (Fig. 1). Examination of labeled cells at higher magnification revealed that the cells are of different sizes and shapes (Figs. 2 and 3). Although most ganglion cells retained their bipolar morphology, pear-, spindle-, spherical-, or trapezoid-shape neurons were also present. In both E13 (Fig. 2) and E15 (Fig. 3) cultures, small- and large-diameter neurons were present following different neurotrophin treatments. In E13 cultures grown in the presence of NGF (Fig. 2B), there were noticeably more large cells in comparison to controls (Fig. 2A) and NT-3-treated cases (Fig. 2C). In E15 cultures, on the other hand, the soma sizes of the TG neurons treated with NGF (Fig. 3B) were smaller than control (Fig. 3A) or NT-3-treated (Fig. 3C) cases. At this age, most of the neurons had pseudounipolar morphologies.
The mean soma diameter of E13 TG neurons was 14.895 ± 0.21 μm in cultures grown in SFM (Table 1). In the presence of NT-3, the mean diameter of the TG neurons (15.293 ± 0.20 μm) was not significantly different from those grown in SFM. In contrast, the mean diameter of the neurons grown in the presence of NGF (16.546 ± 0.21 μm) was significantly larger than those grown in SFM or NT-3. Interestingly, at E15, the mean diameters of the neurons treated with NGF were significantly smaller (16.151 ± 0.16 μm) than those cultured in either SFM (18.089 ± 0.19 μm) or NT-3-supplementd SFM (17.920 ± 0.17 μm). At this age, again, there was no significant difference between the mean diameter of neurons treated with NT-3 or SFM.
Table 1. Multiple comparisons of the mean diameters of E13 and E15 trigeminal neurons grown in cultures containing serum-free medium, nerve growth factor, or neurotrophin-3
Mean Diameter (μm) ± SEM
The mean difference is significant at the 0.05 level.
In E13 cultures, number of the TG neurons grown in the presence of NT-3 (59%) or SFM (55%) peaked between the 13 and the 16 μm soma diameter range (Fig. 4), whereas in the presence of NGF, distribution of soma sizes peaked at two values. While 43% of the neurons were in the 13–16 μm range, 43% of them were in the 16–19 μm range. It is worth noting that 12% of the NGF-treated neurons had soma diameters larger than 19 μm that was rarely seen under other culture conditions. Thus, there was a highly significant difference between the soma size distributions of neurons grown in NGF and SFM alone (P < 0.001) or in cultures supplemented with either NGF or NT-3 (P < 0.01; Fig. 4). At this age, the percentage of small (10–12 μm) TG neurons was 10% lower, but 7% higher for large (16–19 μm) neurons in NT-3-treated cases compared to controls. Yet their cell size distributions (like their mean soma diameters) were not significantly different from the control cases (Fig. 4, Table 1).
In cultures prepared from E15 embryos, size-frequency histogram of the TG neurons treated with NT-3 was virtually identical to that of the control cultures and there was no significant difference between their cell size distributions (Fig. 4). Only 1% of the neurons were in the 10–13 μm range, 16% of them were in the 13–16 μm, and 50% of them were in the 16–19 μm diameter range in both conditions. Between the range of 22 and 25 μm diameters, even a higher percentage of neurons were present in control cultures (7%) than those NT-3-treated ones (3%). Again, the most striking differences in the distributions of soma diameters of E15 TG neurons were observed in cultures supplemented with NGF (Fig. 4). In these cultures, the number of neurons peaked between 13 and 16 μm diameters with an obvious shift toward smaller soma sizes in their histograms. There was a modest increase (39%) between 16 and 19 μm, but only 10% of the neurons had diameters larger than 19 μm.
Different classes of cells expressed TrkA and TrkC in both control and experimental cultures; very few cells coexpressed both receptors (Fig. 5, merged panels). This observation is similar to that seen in the in vivo development of the mouse TG with approximately 9% of TrkA and TrkC double-labeled neurons at E11.5 and 5% at E12.5 (Huang et al., 1999). In control cultures, all of the small cells were TrkA-positive and all of the large cells were TrkC-positive (Figs. 5 and 6). At both developmental time points, NGF- or NT-3-treated cultures also had TrkA- and TrkC-positive cells, and coexpression of Trk receptors did not appear to change (Figs. 5 and 6). NT-3-treated cultures did not show any notable difference in the size of TrkA- and TrkC-expressing cells in comparison to controls. Interestingly, in NGF-treated E13 cultures, many of the large neurons were TrkA-positive, whereas only small-size cells were TrkA-positive in E15 cultures. These qualitative observations provide important clues to shifts in cell size distributions in NGF-treated cases in terms of phenotypic differentiation of TG neurons.
Our quantitative analyses of the TrkA- and TrkC-expressing neurons revealed the following. In E13 control cultures, TrkA:TrkC ratio was 2.39, indicating that at this time point there are more TrkA-positive cells in the ganglion. In NGF- and NT-3-treated cultures, the ratios of TrkA-positive cells were also twice more than those of TrkC-positive cells (2.49 and 2.0, respectively), similar to control cultures. There were no significant differences between these ratios at E13 (Fig. 7). In E15 cultures, TrkA-positive neurons were 2.5 times more than TrkC-positive cells, again similar to that seen at E13. In neurotrophin-treated cultures, on the other hand, there were highly significant (P < 0.001) changes in this ratio between these two groups. In NGF-treated cultures, TrkA:TrkC ratio was 3.5, while in NT-3-treated cultures this ratio was 1.9 (Fig. 7). Thus, in accordance with cell size measurements, these results reveal that NGF promotes survival/maintenance of small-size TrkA-expressing neurons and NT-3 large-diameter TrkC-expressing neurons at E15.
The rodent TG cells are divided into two morphological classes: large light cell bodies (type A cells) with thick myelinated (Aα/β) fibers, and smaller dark cell bodies (type B cells) with thin myelinated (Aδ) or unmyelinated (type C) fibers (Kai-Kai, 1989; Lawson, 1992; Waite and Tracey, 1994). Functionally different classes of these primary sensory neurons express specific peptide markers and neurotrophin receptors (Arumae et al., 1993; Sugimoto et al., 1997; Ichikawa et al., 2002; Lazarov, 2002). Both in vitro and in vivo studies show that small- to medium-diameter nociceptive neurons expressing TrkA receptors, and calcitonin gene-related peptide (CGRP) and substance P are NGF-dependent, whereas larger-diameter neurons expressing TrkC receptors, and calcium-binding proteins parvalbumin and calbindinD-28k are NT-3-dependent (Crowley et al., 1994; Ernfors et al., 1994; Fariñas et al., 1994; Smeyne et al., 1994; Davies, 1997). Signals that induce the development and maintenance of distinct phenotypes of sensory neurons in the TG (and in dorsal root ganglia) are largely unknown.
Autoradiographic studies show that two distinct subpopulations of TG neurons are generated in albino rats over a 3-day period, just after the midpoint of gestation (Forbes and Welt, 1981). Within this time period, larger neurons are generated 1 day prior to smaller neurons with a peak on E12. Recent neuronal birth-dating experiments in mice with BrdU labeling and antibodies to each of the neurotrophin receptors revealed that cells expressing different Trk receptors are generated at different embryonic stages (Huang et al., 1999). While TrkC-expressing large-diameter sensory neurons are generated before E11.5, TrkA-expressing small-diameter neurons are generated between E11.5–13.5. These time points correspond to E13.5 and E13.5–15.5 in the rat (Erzurumlu and Killackey, 1983; Stainier and Gilbert, 1990, 1991; Kaufman and Bard, 1999). Here we examined the effects of exogenous NGF and NT-3 on differentiation of small, large, TrkA-, and TrkC-positive neurons with these time points and neurotrophic factor dependence of sensory neurons in mind.
We found that in both E13 and E15 cultures, mean diameters and size-frequency histograms of the TG neurons treated with NT-3 were not significantly different from those maintained in SFM. In both groups, there was a heterogeneous population of neurons consisting of smaller- and larger-diameter cells. In addition, small-diameter cells were always TrkA-positive and large-diameter cells TrkC-positive, and very few cells expressed both receptors. These findings indicate that under our culture conditions, NT-3 treatment did not selectively promote survival of a specific subpopulation of neurons. Although at E13, the proportion of TrkA- and TrkC-expressing trigeminal neurons were similar, at E15, the number of TrkC-expressing neurons increased in NT-3-treated cultures compared to control or NGF-treated cultures. Previous studies showed that both endogenous and exogenous NT-3 could increase the level of TrkC mRNA in E13 mouse trigeminal neurons, but not in sympathetic neurons (Wyatt et al., 1999). NT-3-induced upregulation of TrkC expression could be attributed to instructive role of this neurotrophin in trigeminal ganglion development. Previously, we showed that in whole-mount trigeminal pathway cultures, NT-3 induces precocious arborization of central trigeminal axons (Ulupinar et al., 2000). Thus, this neurotrophic factor has an instructive role on axonal responses of both E13 and E15 TrkA- and TrkC-expressing TG neurons. In addition, it promotes an increase in the number of TrkC-expressing cells at E15.
In NGF-treated cultures, soma sizes of trigeminal neurons were remarkably different between the two developmental ages studied. At E13, the mean diameter of trigeminal neurons treated with NGF was significantly larger than their counterparts cultured in SFM or NT-3. The size-frequency histograms of these neurons showed a significant shift to larger soma diameters. Immunocytochemistry using TrkA and TrkC antibodies further corroborated these results and revealed the presence of large-diameter TrkA-positive neurons in NGF-treated E13 cultures along with large-diameter TrkC-labeled neurons. This neuronal hypertrophy is consistent with one of the earliest described trophic effects of NGF in embryonic chick DRG and sympathetic neurons (Levi-Montalcini, 1987). Even in maturity, when peripheral neurons no longer require NGF for survival, administration of NGF increases cell soma cross-sectional area in sensory neurons and dendritic branching in sympathetic neurons (Ruit et al., 1990; Kimpinski and Mearow, 2001). Moreover, injection of NGF leads to mechano and thermal hypersensitivity in neonatal and adult rats (Lewin et al., 1993; Lewin and Mendell, 1994; Woolf, 1996). In transgenic mice that overexpress NGF in the skin, levels of p75NTR, TrkA, and TrkC mRNA expression increase (Goodness et al., 1997; Kitzman et al., 1998), and there is hypertrophy of TrkA neurons (Davis et al., 1997; Goodness et al., 1997). Our present results corroborate and extend these findings such that not only peripheral overexpression of NGF but also exogenous supply of NGF to trigeminal neurons can lead to hypertrophy of TrkA-positive neurons at E13, but not at later ages.
We found that soma sizes of neurons were significantly smaller in E15 cultures treated with NGF than those of control or NT-3-treated cultures. Presumably, in cultures grown in the presence of SFM or NT-3, the majority of smaller neurons are eliminated, but exogenous addition of NGF into the culture medium rescues their survival. The reasons for these differential effects of NGF might be related to temporal changes in neurotrophin dependencies and Trk receptor expression patterns of the trigeminal neurons. In cultures prepared from E10 mice, the majority of trigeminal neurons survive with either BDNF or NT-3 and very few respond to NGF, but at later stages, neurons switch their neurotrophin dependence to NGF (Buchman and Davies, 1993; Paul and Davies, 1995; Davies, 1997). In NT-3 mutant embryos, neuronal loss emerges after E10.5 and is complete by E13.5 (Wilkinson et al., 1996), whereas in TrkA mutant embryos, neuronal death peaks between E13 and E14 (Piñón et al., 1996). This is consistent with the later onset of NGF dependency.
The cellular mechanisms underlying time-dependent survival effects of neurotrophins and their action on phenotypic differentiation are emerging. In response to NGF binding, TrkA tyrosine kinase undergoes autophosphorylation and initiates a signal transduction cascade, including Ras/phosphatidylinositol (PI)-3-kinase/Akt-induced suppression of apoptotic proteins, and MEK (mitogen-activated protein kinase)/MAPK activation of antiapoptotic proteins, to stimulate survival (Klesse and Parada, 1999; Kaplan and Miller, 2000). In the adult central nervous system, transgenic activation of Ras-MEK/MAPK alone, without changes in the activity of PI3 kinase, its target kinase (Akt), or in the expression of antiapoptotic proteins, induces pronounced neuronal hypertrophy without altering the total number of neurons (Heumann et al., 2000). In contrast, a specific inhibitor of MEK, PD98059, causes sustained inhibition of NGF-stimulated neurite outgrowth from adult mouse DRG explants (Sjogreen et al., 2000). Surprisingly, outgrowth stimulation of these neurons by NT-3 is markedly enhanced by PD98059 and also by U0126, another inhibitor of MAPK (Wiklund et al., 2002). These results show that there are important differences between NGF and NT-3 signaling pathways involving positive and negative control mechanisms by MAPK activation, respectively. While Ras is necessary and sufficient for NGF-stimulated neuronal growth in the adult sensory neurons, the Ras effectors Raf and Akt induce distinct morphologies in embryonic sensory neurons (Markus et al., 2002). Raf-1 activation causes axon elongation, but Akt activation increases axon caliber and branching. Thus, different Trk signaling pathways mediate differentiation of distinct morphological characteristics of developing neurons. It remains to be determined whether these pathways mediate NGF-induced morphological differentiation of developing rodent trigeminal neurons.
We thank Dr. F. Yücel for discussion of the morphometric analyses and Dr. F. Sahin for the help with statistical analysis. Supported by Osmangazi University (OGU) Commission of Scientific Research Project grant 200011021 (to E.U.) and National Institutes of Health/National Institute of Dental and Craniofacial Research (NIDCR) grant DE07734 (to R.S.E.).