SEARCH

SEARCH BY CITATION

Keywords:

  • capsaicin;
  • calcitonin gene-related peptide;
  • PGP 9.5;
  • skin innervation;
  • stereology

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

Innervation is required to preserve several aspects of skin homeostasis. Previous studies in rodents have shown that sciatic nerve transection leads to epidermal thinning and reduced keratinocyte proliferation. As the sciatic nerve is composed of sensory and motor axons, it is not clear whether skin alterations reflect motor or sensory disturbances. In this study, we used neonatal capsaicin treatment to evaluate whether sensory chemical denervation affects keratinocyte proliferation at 1, 3, and 6 months of age. Using design-based stereological methods, we estimated the total length of intraepidermal nerve fibers (IENF) that were of peptidergic type and the number of bromodeoxyuridine-labeled (BrdU+) nuclei in the hind paw glabrous epidermis of control and capsaicin-treated rats. We found that the treatment decreased the total fiber length of IENF immunoreactive for both protein gene product 9.5 (PGP+) and of IENF immunoreactive for calcitonin gene-related peptide (CGRP+). The length of PGP+ fibers decreased by 83%, 81%, and 77% and that of CGRP+ fibers decreased by 48%, 58%, and 58% at 1, 3, and 6 months, respectively. Double-immunofluorescence staining for neural beta III tubulin and CGRP revealed that the majority of the remaining fibers in the epidermis after capsaicin treatment were of peptidergic type. The number of BrdU+ nuclei was similar in both groups. Our findings suggest that IENF present after capsaicin treatment are sufficient to maintain epidermal replacement. Anat Rec, 2010. © 2010 Wiley-Liss, Inc.

The skin is innervated by nonpeptidergic and peptidergic neurons located in sensory ganglia. However, the peptidergic neurons are not exclusively devoted to convey mechanical, chemical, or thermal information to the central nervous system (Kruger, 1996; Roosterman et al., 2006). In the dermis, the peptidergic nerve fibers terminate in close proximity to blood vessels, hair follicles, and immune cells. In the epidermis, peptidergic fibers establish synaptic-like contacts with Langerhans cells, melanocytes, and keratinocytes (Hosoi et al., 1993; Hilliges et al., 1995; Hara et al., 1996). Although the mechanisms are not yet fully understood, peptidergic terminals are involved in the regulation of skin gene expression, skin pigmentation, blood flow, immune response, and wound healing (Hosoi et al., 1993; Hara et al., 1996; Fundin et al., 2002; Smith and Liu, 2002).

A critical process for epidermal homeostasis is the continual replacement of keratinocytes. Besides cutaneous growth factors, it is believed that nerve-derived factors may regulate keratinocyte proliferation. In psoriasis, for instance, the hyperproliferation of keratinocytes has been associated with an increased density of peptidergic fibers (Chan et al., 1997). Furthermore, sensory derived neuropeptides act as mitogens in cultures of cell lines of human melanocytes and keratinocytes (Takahashi et al., 1993; Hara et al., 1996). In addition, it has been found that sciatic nerve transection decreases epidermal thickness and keratinocyte proliferation (Hsieh and Lin, 1999). However, sciatic nerve lesions not only damage sensory innervation but also destroy autonomic and motor axons. Consequently, the epidermal changes might be caused by postural abnormalities or by alterations in the vascular supply (Li et al., 1997; Stankovic et al., 1999). Therefore, experiments that are designed to modify more selectively the intraepidermal nerve fibers (IENF) are needed to better define their role in the regulation of keratinocyte proliferation.

Neonatal capsaicin treatment induces the death of a great number of small sensory neurons that give rise to C and Aδ axons. These axons, especially the C fibers, innervate different layers of the living strata of the epidermis. The consequent reduction in the density of peripheral endings after capsaicin treatment has been proposed to alter the normal epithelial replacement in the cornea and in the skin (Beuerman and Schimmelpfennig, 1980; Maggi et al., 1987). However, the effects of chemical denervation on normal epithelial proliferation have not yet been evaluated. Although it is known that the neonatal capsaicin treatment diminishes the content of substance P and calcitonin gene-related peptide (CGRP) in the sensory ganglia and spinal cord (Carr et al., 1990), it is not well defined how the capsaicin treatment affects the peptidergic innervation of the skin during development. Our work had two goals. First, we wanted to quantify the IENF loss of immunoreactive fibers in capsaicin-treated rats for the neuronal marker protein gene product 9.5 (PGP+) or CGRP+. Second, we wanted to investigate the long-term effect of chemical sensory denervation in keratinocyte proliferation. To do this, we used design-based stereological methods on our histological preparations to measure the total fiber length of IENF and the number of bromodeoxyuridine-labeled nuclei in glabrous skin of the rat. We found that capsaicin severely reduced epidermal fibers, but did not completely eliminate peptidergic fibers. The loss of IENF was associated with a mild, but not significant effect on keratinocyte proliferation.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

Animals

Female Wistar rats were raised in the animal facility of the Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México. Rats were kept in a colony room at 21°C on a 12 hr illumination cycle (lights on at 6:00 am) with free access to food and water. The experimental procedures were approved by the local animal rights committee and carefully followed the guidelines of the Código Bioético para la Investigación con Animales published by the Instituto de Investigaciones Biomédicas. All efforts were made to minimize the number of animals used and their potential suffering.

A single dose of capsaicin was administered to reduce sensory innervation as described elsewhere (Newson et al., 2005). Briefly, capsaicin was dissolved in a vehicle consisting of 10% ethanol, 10% Tween 80, and in 0.9% saline. Within 24–36 hr after birth, rats pups (mean weight = 7.8 g) were anesthetized by hypothermia and injected subcutaneously in the back skin with 40–60 μL of either vehicle or capsaicin solution (50 mg/kg). After the injection, the rats were put into a clear chamber, placed between the fingers of latex gloves filled with 37°C water. As apnea is an acute consequence of capsaicin injection, three puffs of salbutamol aerosol were sprayed into the chamber each time a rat pup was introduced. This procedure helps to alleviate respiratory difficulties. In addition, some rat pups required chest massage to reinstate their normal respiration. Rats were returned to their mother until respiratory difficulties disappeared.

One (N = 8), three (N = 6), and six (N = 6) months after birth, both vehicle and capsaicin-treated rats were intraperitoneally injected with bromodeoxyuridine (BrdU; 50 mg/kg) dissolved in 0.007 N NaOH/0.9%NaCl. Rats were killed 1 hr later by administering an overdose of sodium pentobarbital. The entire plantar hind paw skin was dissected and fixed in Zamboni's fixative [4% paraformaldehyde; 15% (v/v) saturate picric acid, 0.1 M phosphate buffer, pH 7.4] at 4°C for 48 hr and then transferred successively to 20 and 30% sucrose. The plantar hind paws have three pair of pads. The footpads were designated as first, second, and third from the toes to the heel. The second pair of footpads were embeded in Tissue-Tek OCT compound (Sakura Seiki, Tokyo, Japan) and sectioned transversely into 50 μm slices in a cryostat. The sections were individually collected in 96-well plates filled with cryoprotectant solution (25% ethylene glycol and 25% glycerol in a 0.05 M phosphate buffer) and stored at −20°C. A systematic random sampling procedure was used to select the series of sections for neuronal antigens or BrdU immunohistochemistry. To analyze 12–15 sections per animal, a sampling interval of six was used. The first sampled section was selected at random between the first and the sixth section, then each successive section was sampled systematically every sixth section. All samples were coded and the observers were blinded to the coding information.

Immunohistochemistry

Series of sections were put onto gelatinized slides, washed 2× with 0.1 M phospate buffer (PB), rinsed 1× with 0.3% Triton X-100 in PB (PBT), and incubated with 1% H2O2 in PB for 2 hr. After three PB rinses, sections were incubated with Immuno/DNA retriever at 70°C for 30 min (Bio SB, Santa Barbara), and then washed with PB. For BrdU immunodetection, DNA was denaturated with 1 N HCl at 30°C for 1 hr and neutralized with 0.1 M sodium borate buffer. Finally, sections were blocked with 5% normal horse serum in PBT for 1 hr. The sections were incubated overnight with either rabbit anti-PGP 9.5 (1:1000, Chemicon, Temecula), rabbit anti-CGRP (1:10000, Peninsula Labs, San Diego), or mouse anti-BrdU (1:500, Roche Applied Science, Penzberg, Germany). After PB rinses, sections were incubated with the appropriated biotinylated secondary antibodies (Chemicon) for 2 hr and then with the avidin–biotin complex (Vector Laboratories, Burlingame) for 90 min. The immunohistochemistry reaction was made visible by 3–3′-diaminobenzidine/nickel precipitation (Vector Laboratories). After a rinse in water, sections were incubated in 0.05 M sodium bicarbonate buffer pH 9.6 for 10 min and then exposed to DAB enhancing solution (Vector laboratories). Finally, sections were counter-stained with methyl green. High-magnification digital photomicrographs of intraepidermal nerve fibers were taken with a video camera attached to an Olympus BX51 WI microscope. To better display the extent of PGP+ and CGRP+ fibers, multiple images were captured at different focal planes and merged into a composite micrograph using the extended depth of focus function of the Image-Pro Plus 5 software (Media Cybernetics).

Double Immunofluorescence

Hind paw sections were put onto gelatinized slides, washed 2× with PB, rinsed 1× with PBT, and incubated with Immuno/DNA retriever at 70°C for 30 min. and then washed with PB. Then, sections were blocked with 5% normal horse serum in PBT for 1 hr. The sections were incubated overnight with mouse anti-neural III beta-tubulin (1:500, Promega, Madison) and rabbit anti-CGRP (1:7500, Chemicon). After PB rinses, sections were incubated with biotinylated goat anti-mouse IgG (1:500, Chemicon), which was used to amplify the signal of beta III tubulin signal. Following PB rinses, sections were incubated for 2 hr at room temperature with a mixture of donkey anti-rabbit IgG conjugated to Alexa 594 (1:500 Molecular Probes, Eugene) and donkey anti-goat IgG conjugated to Alexa 488 (1:500, Molecular Probes). Finally, sections were counter-stained with DAPI and coverslipped with Dako Fluorescence Mounting Medium (Dako, Carpinteria). High-magnification digital photomicrographs of intraepidermal nerve fibers were taken with a Hamamatsu C9100CCD camera attached to an Olympus BX51 WI microscope equipped with a disk scanning unit. Composite micrographs were generated as above.

Quantification of Fiber Length

PGP+ or CGRP+ fiber total length was quantified by using the space balls method for design-based stereology (Mouton et al., 2002). Space balls are based on the principle that the number of probe-object intersections is proportional to the total length of linear objects in the tissue. Data were obtained with an Olympus BX51 WI microscope equipped with a 4× and 60× (water immersion, 1.2 NA) objectives, a motorized xyz stage control, and the StereoInvestigator Software 8 (MBF Bioscience, Williston). Briefly, the boundaries of the epidermis were traced, excluding the stratum corneum. The virtual hemispheres generated by the software were systematically and randomly placed through the epidermal tracing. Intersections of immunoreactive fibers with hemispheres were marked while focusing through the z-axis of each sample site. Only those fibers within the epidermis and clearly crossing dermal-epidermal border were considered for quantification (Lauria et al., 2005). The stereological parameters were as follows: sampling grid, 225 × 155 μm; radius of the hemisphere, 13 μm; and guard zones, 10% of section thickness. The mean section thickness was 32.8 ± 2 (Mean ± SD), measured every fifth sampling site. The fiber length was calculated by the equation L = 2 * (ΣQ) * V/a * 1/ssf. Where (ΣQ) is the sum of the number of intersections of IENF and hemispheres, V is the volume associated with grid parameters and section thickness, a is the surface area of the sphere, and ssf is the serial section fraction and refers to the periodicity of sampling. The results were expressed as total fiber length as described previously (Stocks et al., 1996).

Quantification of Keratinocyte Proliferation

The number of BrdU+ nuclei in the epidermis was estimated by the optical fractionator procedure for design-based stereology. Data were obtained with a Nikon Labophot-2 microscope equipped with 10× and 100× (oil immersion, 1.4 NA) objectives, a motorized x-y-z stage control, and interfaced with StereoInvestigator. BrdU+ nuclei were counted at uniformly random sampled sites within the epidermal tracing. The size of the grid was X = 318 μm and Y = 104 μm with a counting frame measuring 36 × 30 and a height of 20 μm. The guard zone was 10% of the mean section thickness of 25.1 ± 2.5, which was measured at every sampling site. The total number of BrdU+ nuclei was calculated by using the formula: N = ΣQ * t/h*1/asf *1/ssf, where ΣQ is the number of particles counted, t the mean section thickness, h the counting frame height, and asf is the relation of the area of counting frame to the area of sampling grid.

Statistics

Data are expressed as Mean ± SEM. Statistical analysis was performed using Origin Pro 8. Statistical differences between groups were determined by a two-tailed Student's t test. Multiple comparisons were performed by using one-way ANOVA test followed by a Holm-Sidack post hoc test. Any difference with P < 0.05 was considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

PGP+ IENF in Normal and Capsaicin-Treated Rats

The anatomical changes of the epidermal innervation in glabrous skin at 1 (1 M), 3 (3 M), and 6 months (6 M) after neonatal capsaicin treatment were analyzed with the neuronal marker PGP. Despite the fact that the general innervation pattern was preserved following capsaicin treatment, there were obvious differences in the origin, termination layer, and morphology of the IENF. In control rats, IENF originated from two types of bundles in the dermis. The majority of the IENF arose from the subepidermal plexus as fine single fibers (Fig. 1A,C,E). The rest of the IENF came from axonal bundles running perpendicular to the epidermis. In general, the distribution of the IENF immunoreactive for PGP was heterogeneous throughout the epidermis. In some epidermal regions, we observed the PGP+ fibers to be grouped in high density patches of nerve endings. Within these patches, the nerve endings were separated one from the other for a short distance. Most of the PGP+ fibers terminated either in stratum spinosum (SS) or granulosum (SG). Some of these fibers coursed straight through the stratum basalis (SB) and the lower layers of the SS and then meandered in the upper layers of the SS and SG (Fig. 1A,C,E). Less frequently, PGP+ fibers with a straight trajectory terminated either in the SS or in the SB. Although the IENF patch clustering was similar at 3 M and 6 M, these groups of terminals were more distant from each other. At 3 M, there was an increase of the number of very thin and strikingly beaded fibers (Fig. 1C). In addition, some fibers became more branched and some of them were tightly or loosely intertwined. At 6 M, the number of very thin and beaded fibers tends to decrease (Fig. 1E).

thumbnail image

Figure 1. Neonatal capsaicin treatment decreased the number of intraepidermal nerve fibers. Glabrous skin sections from control and capsaicin-treated rats of 1 M (A and B), 3 M (C and D), and 6 M (E and F) were immunostained for protein gene product 9.5 (PGP, arrows). Note that capsaicin-treated rats showed aberrant thick fibers that penetrate to epidermis (F, arrowhead) and that the subepidermal plexus (asterisk) has fewer fibers in capsaicin-treated rats. Scale bar = 30 μm.

Download figure to PowerPoint

In capsaicin-treated rats, there was a clear-cut diminution of PGP+ epidermal fibers at all the analyzed ages. The fibers separating from subepidermal plexus were scarce and almost all of the IENF had straight trajectories (Fig. 1B,D,F). The majority of the IENF were thick and terminated in the SS, and only a few penetrated into the SG. As in the epidermis of controls rats, the distribution of the IENF was also heterogeneous. It is noteworthy that patches of fibers in capsaicin-treated rats were composed of two or three fibers, which resulted in greater separation between terminals (Fig. 1B,D). Overall, IENF in capsaicin-treated rats had a straight morphology and were poorly ramified. The total length of IENF was reduced 83%, 81%, and 77% in the epidermis of capsaicin-treated rats of 1 M, 3 M, and 6 M, respectively (Fig. 2A). We did not observe differences in the number and intensity of PGP immunoreactive dendritic cells in both groups of animals.

thumbnail image

Figure 2. Total fiber length of PGP+ and CGRP+ fibers was reduced in the glabrous epidermis of capsaicin-treated rats. The mean total length of intraepidermal axons was estimated with the stereological probe of space balls. The length of immunoreactive PGP (A) and CGRP (B) fibers is significantly reduced at 1 M, 3 M, and 6 M after capsaicin treatment (*P < 0.05).

Download figure to PowerPoint

CGRP+ IENF in Normal and Capsaicin-Treated Rats

In normal rats, CGRP+ fibers were observed in dermal papilla and in the central zone of the rete pegs. In 1 M and 3 M, fibers were commonly found in groups of two to three intertwined terminals, whereas at 6 M, the axon terminals were frequently observed as individual fibers (Fig. 3A,C,E). CGRP+ fibers predominantly ended in the SG, although some fibers terminated in the SS, especially at 6 M (Fig. 3E). CGRP+ terminals meandered through keratinocyte layers and exhibited branching points. After reaching the SG, some terminals formed loops that returned to lower strata (Fig. 3A,C).

thumbnail image

Figure 3. CGRP fiber location in the epidermis is preserved following neonatal capsaicin treatment. Glabrous skin sections from control and capsaicin-treated rats of 1 M (A and B), 3 M (C and D), and 6 M (E and F) were immunostained for CGRP. In control rats (A, C, and E), CGRP+ fibers meander through keratinocyte layers and some fibers show terminal loop configuration (arrowhead). Although the CGRP fibers appear at regular intervals after capsaicin treatment, these nerve endings are less complex and frequently appear alone (B, D, and F). Also, in the treated rats we observed the presence of knob-like termination, especially at 6 M. Scale bar = 30 μm.

Download figure to PowerPoint

In capsaicin-treated rats, CGRP+ fibers were also distributed at regular intervals, but in most cases fibers were unbranched and frequently terminated in the SS. Just a few of the CGRP+ fibers reached the lower layers of the SG (Fig. 3B,D,F). CGRP+ fibers in capsaicin-treated rats tended to be thicker and less beaded than in controls. At 6 M, thick knob-like terminations were evident in the SS (Fig. 3F). Capsaicin reduced by 48%, 58%, and 58% the total length of CGRP+ fibers at 1 M, 3 M, and 6 M, respectively (Fig. 2B).

In control rats, the length of CGRP+ fibers represented 53%, 49%, and 54% of the length of PGP+ fibers at 1 M, 3 M, and 6 M, respectively. In capsaicin-treated rats, the CGRP+ fiber length was significantly higher (65%, t test, P < 0.05) than that of the PGP+ fibers at 1 M, but the length of PGP+ and CGRP+ fibers was similar at 3 M and 6 M.

Double Labeling for βIII Tubulin and CGRP

To judge whether the predominant fibers in the epidermis of capsaicin-treated rats were of peptidergic type we performed a double labeling experiment for β III tubulin and CGRP. In preliminary experiments, we confirmed that PGP immunoreactivity always colocalized with β III tubulin immunoreactivity (Supporting Information Fig. 1). In the epidermis of control rats we observed that the most abundant type of fibers was the single labeled for β III tubulin at 1 M, 3 M, and 6 M. The CGRP+ fibers in these animals were also immunoreactive for β III tubulin (Fig. 4A,C,E). In contrast, the capsaicin-treated rats exhibited some CGRP+ fibers that were not immunoreactive for β III tubulin (Fig. 4B). This phenomenon was very common at 1 M, but barely seen at 3 M and 6 M. An additional difference between the control and treated rats was that only a few fibers of the total were single labeled for β III tubulin in the treated rats (Fig. 4D). Hence, the most abundant type of IENF in the treated rats was the double labeled ones (Fig. 4B,D,F).

thumbnail image

Figure 4. The epidermis of the capsaicin-treated rats predominantly contains peptidergic fibers. Skin sections from control (A, C, and E) and capsaicin-treated rats (B, D, and F) of 1 M (A and B), 3 M (C and D), and 6 M (E and F) were inmunostained for neural beta III tubulin (βTub) and for CGRP. In control rats, the most abundant type of fibers was immunopositve for βTub and immunonegative for CGRP (green arrows), whereas in treated rats the most common type of fibers was double labeled for βTub and CGRP (yellow arrows). Note that in treated rats there were fibers single labeled for CGRP (red arrows). Images at the third column represent merged images of the first two images in each row. Counterstaining with DAPI is shown in blue. Scale bar = 50 μm.

Download figure to PowerPoint

Evaluating Keratinocyte Proliferation

IENF have been implicated in the maintenance of keratinocyte proliferation. Thus, the total number of BrdU+ nuclei was estimated to determine whether the reduction of epidermal innervation by capsaicin could provoke alterations of keratinocyte replacement. The number of BrdU+ nuclei was found to be similar between control and capsaicin-treated rats at 1 M, 3 M, and 6 M. The thickness of epidermis and number of layers were equivalent in both control and capsaicin-treated rats (Fig. 5). In addition, the number of BrdU+ nuclei tended to decrease with age in both experimental groups (Fig. 6).

thumbnail image

Figure 5. Bromodeoxyuridine labeling in rat glabrous skin. Skin sections from control (A, C, and E) and capsaicin-treated (B, D, and F) of 1 M (A and B), 3 M (C and D), and 6 M (E and F) were inmunostained for bromodeoxyuridine (BrdU). In both groups, the BrdU nuclei were mainly observed in basal stratum of epidermis and there were no appreciable differences in epidermal thickness. The inset in F shows a magnification of BrdU+ nuclei. Scale bar = 150 μm. Inset scale bar = 10 μm.

Download figure to PowerPoint

thumbnail image

Figure 6. Capsaicin denervation did not affect keratinocyte proliferation. The number of BrdU+ nuclei in the central pads of the hind paw was estimated using the stereological probe of the optical fractionator.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

Capsaicin-sensitive sensory neurons have been proposed to maintain epithelial integrity (Beuerman and Schimmelpfennig, 1980; Maggi et al., 1987). However, information is lacking concerning the effect of neonatal capsaicin treatment on the time course of development of epidermal innervation and of keratinocyte proliferation. In this study, we have found that although capsaicin treatment severely reduced IENF number, a significant proportion of peptidergic fibers was retained. At the same time, the epidermis of capsaicin-treated rats presented normal characteristics. Overall, the epidermal innervation in the capsaicin-treated rats was morphologically simpler than in controls and there were changes in the epidermal stratum where the IENF terminate. Both observations may imply that there are alterations in both the afferent and efferent functions of the sensory nerves.

The rodent hind paw skin has been frequently used as an experimental model to study nerve degeneration and regeneration (Navarro et al., 1995; Ma and Bisby, 2000; Yen et al., 2006). However, there is little published quantitative data about the development of epidermal innervation in this skin region. Our report provides insights into the normal development of epidermal innervation in the glabrous skin and how it is altered following capsaicin treatment. Our data show that although the footpad increases in size as an animal matures, the IENF length was substantially unchanged in control rats at 1 M, 3 M, and 6 M. This suggests that the supply of IENF remains relatively constant, but is redistributed through the epidermis as the animal ages. Indeed, we observed that the clustering organization of PGP+ fibers is essentially constant at the three analyzed ages, but the patches of IENF appear more dispersed at 3 M and 6 M. With regard to fiber plasticity, the IENF seem to undergo continual remodeling as suggested by the morphological diversity of the IENF observed in the three analyzed ages. In agreement with this notion, the sensory endings in the mouse cornea showed substantial changes in the arborization and location of nerve terminals over a 1-month-period (Harris and Purves, 1989). These findings would explain why we observed the appearance of finely beaded fibers at 3 M, and then their decline at 6 M. In contrast, the IENF of capsaicin-treated rats may possess limited plastic properties. Remarkably, a relatively high number of thick nerve fibers were observed terminating in the epidermis of treated rats. On the basis of fiber appearance, these sensory terminals could be a subset of C fibers with an altered morphology or perhaps Aδ fibers that do not contain GAP43 [a molecule associated with nerve remodeling (Albrecht et al., 2006; Pare et al., 2007)]. In this regard, the perineural application of capsaicin in adult rats reduces the number of epidermal fibers containing GAP43 (Dux et al., 1999). Thus, the constant low levels of epidermal innervation in the treated rats may reflect a limited sprouting ability of the PGP+ or CGRP+ fibers that remain. Although the direct relation between peripheral arborization and specific sensory modalities is beginning to be understood, it is plausible that less branched arbor configuration of IENF in treated rats, may alter fiber excitability and circumscribe its area of activation (MacIver and Tanelian, 1993; Woodbury and Koerber, 2007). We note that the detection of noxious stimuli is impaired by capsaicin injection in human skin (Simone et al., 1998). Moreover, additional sensory alterations may come from the fact that ample zones of epidermis were devoid of innervation in treated rats. Recently, it has been shown that a lesion of sensory nerves lead to aberrant sprouting of sympathetic fibers into the upper dermis of the rat (Yen et al., 2006). We observed such aberrant dopamine-β-hydorxylase imunoreactive fibers in the dermal-epidermal border and also noted some fibers penetrating the epidermis in capsaicin rats of ages 3 M and 6 M (Supporting Information Fig. 2). Whether the presence of nonvascular sympathetic fibers in the upper dermis of capsaicin-treated rats is relevant to sensory physiology remains to be investigated.

Recently, it has been documented that stereological estimation of intraepidermal fiber length is positively correlated with the number of fibers per millimeter in the same sections (McArthur et al., 1998). Considering that neuronal capsaicin sensitivity is given by the expression of the transient receptor potential V1 (TRPV1) (Hiura, 2000), our findings suggest that a considerable number of intraepidermal nerve fibers in the glabrous epidermis of the rat presumably contain this receptor. The TRPV1 is present in small neurons of the dorsal root ganglia of nonpeptidergic and peptidergic type (Guo et al., 1999; Hwang et al., 2005; Price and Flores, 2007). We note here that recently, it has been shown that nonpeptidergic fibers are the most abundant type of fibers within the epidermis in glabrous and hairy skin of the rat (Taylor et al., 2009). Thus, it is plausible that a subpopulation of nonpeptidergic fibers may be among the fibers that were more affected by capsaicin treatment. To go along with this, our results also indicate that a subpopulation of peptidergic fibers is also affected by the capsaicin treatment. In this regard, almost half of the CGRP+ fibers in epidermis could be associated with capsaicin-sensitive neurons. This estimate seems reasonable since 59% of the neurons innervating the skin of the L6 dermatome exhibited colocalization of TRPV1 with CGRP (Hwang et al., 2005). A fraction of these cutaneous neurons may project to the epidermis and innervate there an ample territory by way of distal axonal branching. This would not necessarily imply that all terminals contain the TRPV1 in the epidermis. Indeed, there is a discrepancy in the proportion of IENF immunoreactive for TRPV1 in human and primate skin (Petersen et al., 2002; Lauria et al., 2006; Pare et al., 2007). One possible explanation of these disparate results is that some peptidergic or nonpeptidergic fibers in the epidermis come from capsaicin-sensitive neurons in which TRPV1 is not transported to the same peripheral ending or that these endings contain undetectable levels of this receptor (Guo et al., 1999). Further studies in rodents are needed to determine the exact proportions of peptidergic and nonpetidergic intraepidermal endings colocalizing with TRPV1. Finally, the existence of TRPV1 expression has been reported in various non-neuronal tissues such as keratinocytes (Stander et al., 2004). This finding raises the possibility that neonatal capsaicin treatment may have a cytotoxic effect or may alter the physiology of the keratinocytes. However, it has been shown that TRPV1 is not functional in human keratinocytes probably due to the presence of TRPV1b, which is a dominant negative splice variant of TRPV1. Furthermore, cytotoxic effects on keratinocytes were only observable with high doses of vanilloids and these effects were attributable to a TRPV1-independent pathway (Pecze et al., 2008).

An additional point worthy of mention is that at 1 M the CGRP+ fiber length was greater than the PGP+ fiber length in capsaicin-treated rats. This observation suggests that PGP may be absent or downregulated in a subset of unmyelinated axons after capsaicin treatment. Confirming this, in double labeling experiments we observed that some fibers could be detected with βIII tubulin but not with PGP especially at 1 M in treated rats (Supporting Information Fig. 1). In addition, some CGRP+ fibers did not label for βIII tubulin. In this regard, it has been observed that after sciatic nerve constriction in rats and in skin samples of a human patient with herpetic neuralgia, there are some unmyelinated fibers that display faint or no immunoreactivity for PGP (Ma and Bisby, 2000; Petersen et al., 2002). This phenomenon seems to be temporal and particular to a degenerative condition. The fact that PGP+ and CGRP+ fiber length is similar in 3 M and 6 M in treated rats may indicate that the levels of PGP return to detectable levels in certain fibers. We point out that we cannot rule out the possibility that a percentage of fibers could not be detected with our present analysis. Additional studies are required to determine why the synthesis or transport of PGP and βIII tubulin is altered during nerve regeneration and degeneration and whether this reflects some sort of functional alteration of the nerve ending.

IENF have been proposed to modulate keratinocyte proliferation (Hsieh and Lin, 1999). To explore this notion further we injected rat pups with a single dose of capsaicin to reduce the epidermal nerve supply. We found that despite the significant decrease in epidermal innervation, epidermal thickness and keratinocyte proliferation were not altered in capsaicin-treated rats. One possible explanation is that normal keratinocyte proliferation in treated rats resulted from adequate mechanical load and motor performance which are altered following sciatic nerve lesion (Li et al., 1997; Stankovic et al., 1999). The normal keratinocyte proliferation could be modulated by signals derived from the remaining nonpeptidergic C fibers which can be activated by ATP released from keratinocytes after mechanical stimulation (Koizumi et al., 2004; Taylor et al., 2009). However, epidermal thinning is also produced following denervation of skin regions that do not support body weight (Burgess et al., 1974; English, 1977; Nurse et al., 1984). This data suggests additional mechanisms that could act alone or in concert in the capsaicin-treated rats. For instance, it is possible that the diminishment of nerve-derived signals may be compensated in non-injury conditions by increasing the availability of local growth factors. An additional alternative is that also the remaining peptidergic fibers are involved in providing the trophic support that epidermis requires for normal homeostasis. A recent study has shown that CGRP is one of the factors that stimulates keratinocyte proliferation in vivo after ultraviolet irradiation (Seike et al., 2002). On the basis of this data and our observation that peptidergic fibers were the predominant type of fibers after capsaicin treatment, it is possible that CGRP availability in treated rats contributes to maintain adequate keratinocyte replacement in non-injury conditions. Interestingly, capsaicin-treated rats develop spontaneous skin wounds in the head region and show delayed wound healing (Maggi et al., 1987; Smith and Liu, 2002). Moreover, wound repair is accelerated by exogenous administration of CGRP (Engin, 1998). In this context, the restricted lateral expanse of IENF in treated rats probably complicates fiber stimulation, leading to a reduction of nerve-derived signals release during conditions where skin demands high rates of cell proliferation. Therefore, the efferent role of IENF on keratinocyte replacement might be best evaluated with respect to damaged tissue conditions.

We conclude that neonatal capsaicin treatment decreases the density of both nonpeptidergic and peptidergic fibers in the epidermis. However, the loss of peptidergic fibers was less severe. It seems that the remaining epidermal innervation is sufficient to support functional requirements for epithelial replacement in non-injury conditions.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

The authors thank Patricia Padilla for their technical and administrative assistance and Georgina Díaz Herrera and Adolfo Herrera Juárez for providing animal care. They are also grateful to Dr. Edmund Glaser for helpful criticisms and careful editing. They also thank MicroBrightField Inc. of Williston, VT, USA, for donating one of the systems used to carry out the stereological analyses. Additional funding came from Programa de Apoyos Integrales para la Formación de Doctores en Ciencias, CONACyT (53194). E.M.M. and B.T.M. were fellows from CONACyT. E.M.M. was also sponsored by Dirección General de Estudios de Posgrado-UNAM and IMPULSA-02.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. LITERATURE CITED
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
AR_21237_sm_suppinfoFig1.tif4867KSupplemental Fig. 1. Colocalization of the pan-neuronal markers βIII tubulin and PGP in the glabrous skin of the rat. Skin sections from control (A, C, E) and capsaicin-treated rats (B, D, F) of 1M (A, B), 3M (C, D), and 6M (E, F) were inmunostained for neural beta III tubulin (βTub) and for PGP. In both groups of rats all the PGP+ fibers colocalized with βTub + fibers (yellow fibers). However, some βTub+ fibers were not labeled for PGP (B2a, E1;green arrows). The lack or faint staining for PGP seem to be in the distal part of the axons because all nerve fibers seem double labeled in the nerve plexus at the level of lower dermis (B3c). Even in controls rats there were intraepidermal nerve fibers that showed faint immunoreactivity for PGP (C2, yellow arrowhead). We observed in both groups that anti-βTub stained the upper strata of the epidermis. Images at the third column represent merged images of the first two images in each row. Counterstaining with DAPI is shown in blue. Scale bar = 50 μm
AR_21237_sm_suppinfoFig2.tif1080KSupplemental Fig. 2. Dopamine-β-hydroxylase immunoreactive (DBH+) fibers in the upper dermis of capsaicin-treated rats. In the skin of control rats, DBH+ fibers were easily observed around the blood vessels (A) and the sweet glands at 1M (not shown) and 3M and 6M (not shown). At 1M and 3M, DBH+ fibers were not observed near the dermal-epidermal border. In capsaicin-treated rats, DBH+ fibers were rarely observed in the upper dermis at 1M (not shown). While at 3M, treated rats exhibited numerous DBH+ fibers near the dermal-epidermal border (B, arrows). Additionally, some DBH+ fibers were observed to penetrate the epidermis (B, arrowhead). At 6M, DBH+ fibers (arrows) were observed in both the control (C) and treated rats (D). However, in treated rats the DBH+ fibers were thicker, and more numerous and more uniformly distributed than in the control rats. Intraepidermal DBH+ fibers were also present at 6M in treated rats. At 6M many of the DBH+ fibers in the control rats displayed a transverse trajectory with respect to dermal-epidermal border rather than a parallel distribution as in the treated rats. Scale bar = 100 μm.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.