Quantitative and morphological differences of nerve fibers between proliferative and mature scars in two- and three-dimensional spaces☆
Yibing Wang1, Xia Li2, Rui Zhang3, Yongqiang Feng1, Yu Liu4

1Department of Burn, Provincial Hospital Affiliated to Shandong University, Jinan  250021, Shandong Province, China
2Department of Burn, Weihai Municipal Hospital, Weihai  264200, Shandong Province, China
3Department of Pathology, Shandong Medical College, Jinan  250002, Shandong Province, China
4Department of Pathology, Provincial Hospital Affiliated to Shandong University, Jinan  250021, Shandong Province, China

Yibing Wang☆, Doctor, Professor, Chief physician, Department of Burn, Provincial Hospital Affiliated to Shandong University, Jinan  250021, Shandong Province, China

Corresponding author: Rui Zhang, Master, Associate professor, Department of Pathology, Shandong Medical College, Jinan  250002, Shandong Province, China; Xia Li, Master, Department of Burn, Weihai Municipal Hospital, Weihai  264200, Shandong Province, China
47772437@qq.com; lixia970@sina.com

Abstract
BACKGROUND: Numerous studies use fluorescent microscopy to obtain two-dimensional op-tical images of the morphology of nerve fibers in hypertrophic scars. In addition, current con-focal microscopy studies have focused on normal, not pathological, cutaneous nerves. However, laser scanning confocal microscopy results in a three-dimensional structure of the nerve fibers.
OBJECTIVE: To observe quantitative and morphological differences in nerve fibers from the proliferative and mature stage in hypertrophic scars using fluorescent and confocal microscopy.
DESIGN, TIME AND SETTING: Neuropathological, comparison study was conducted at the Provincial Hospital Affiliated to Shandong University, China from June 2006 to July 2007.
PARTICIPANTS: Specimens were selected from 30 patients undergoing scar restoration at the Provincial Hospital Affiliated to Shandong University of China at 1 month to 23 years following wound healing. The study comprised 20 males and 10 females. The scars were fibrous lesions, erythematous, tough, confined to skin lesions, did not exhibit ulceration or infection, exhibited telangiectasia, with or without itching and pain, and were not locally treated. Samples were equally assigned to two groups according to course of disease: proliferative group (< 6 months) and mature group (6–24 months). Control samples were collected from full-thickness skin from donor sites (n = 10).
METHODS: Nerve fiber morphology was observed using fluorescent and confocal microscopy following immunofluorescence of the skin specimens. The microscopic images were semi-quantitatively analyzed to acquire a positive area ratio of neurofilament protein-positive nerve fibers.
MAIN OUTCOME MEASURES: Morphology and positive area ratio of neurofilament pro-tein/positive nerve fibers was measured.
RESULTS: The positive area ratio of neurofilament protein-positive nerve fibers was signifi-cantly greater in the proliferative group compared to the normal control group (P < 0.05). Nerve fibers were irregularly distributed and exhibited local swelling, twisting, and disconnection. However, the positive area ratio of neurofilament protein-positive nerve fibers was significantly less in the mature group compared with the normal control group (P < 0.05). The nerve fibers were arranged in an orderly manner, with intact inner and stereoscopic structures similar to normal skin.
CONCLUSION: Compared with mature scars, hypertrophic scars exhibited a greater number of nerve fibers, with more serious pathologies.
Key Words: scar; hypertrophic; neurofilament proteins; innervation; nerve regeneration

INTRODUCTION
  
Previous studies have shown that nerve fibers play an important role in hypertrophic scar formation[1], but the interaction mechanisms remain poorly understood. The thickness of regenerated epidermis depends on innervation, to some extent[2], but keratinocytes of the epidermis are involved in hypertrophic scar formation[3]. In addition, neurogenic inflammation is associated with growth and generation of hypertrophic scars[4]. The present study analyzed the relationship between cutaneous nerve fibers and hypertrophic scars by observing morphological changes and connections.
In the past, silver nitrate and gold chloride were used to stain cutaneous nerves. However, the tiny nerve fibers were not clearly displayed due to poor sensitivity. Recently, immunohistochemistry has been utilized to stain small nerve fibers using many different neural markers, such as S-100, protein gene product 9.5 (PGP9.5), and neurofilament proteins. Neurofilament proteins stain nerve fibers with high specificity and reflect the dynamic process of nerve fiber damage and repair more effectively than other nerve markers. Neurofilament protein is an intermediate filament of the neuronal cytoplasm[5]. It is a specific marker of neuronal perikarya, nerve dendrites, and axons, and it does not cross-react with other intermediate filaments. Therefore, neurofilament protein has been more effective than other neural markers in reflecting the dynamic process of neural damage and repair. In the present study, immunofluorescence was utilized to stain the nerve fibers. Recently, laser scanning confocal microscopy (LSCM) has taken the place of fluorescent microscopy in fluorescent staining of some specimens. LSCM can obtain internal micro-structure fluorescent images by exciting fluorescent probes with ultraviolet or visible light[6]. Cells can be scanned, layer-by-layer, using LSCM, and two-dimensional images of each layer can be reconstructed to show the three-dimensional structure of tissue cells. The experiments served to investigate innervation in various periods of hypertrophic scar formation.

SUBJECTS AND METHODS

Design
Neuropathological, comparison study.
Time and setting
Experiments were performed at the Provincial Hospital Affiliated to Shandong University, China from June 2006 to July 2007.
Subjects
All subjects were selected from patients undergoing scar restoration procedures at the Provincial Hospital Affiliated to Shandong University in China 1 month to 23 years following wound healing. The subjects comprised 20 males and 10 females, aged 14–51 years. Inclusion of experimental subjects[7]: raised fibrous lesions, erythematous, tough quality, confined to skin lesion, without ulceration and infection, exhibiting telangiectasia, with or without itching and pain, and without local treatment. The patients exhibited normal liver function and other laboratory test results were normal. They exhibited no skin diseases or other chronic diseases. Samples were equally assigned to two groups according to course of disease: proliferative (< 6 months) and mature (6–24 months). Control samples were collected from full-thickness donor sites, respectively (n = 10).
According to the Administrative Regulations on Medical Institution published by the State Council of China[8], the patients were informed of the pilot project. The study obtained permission from the Medical Ethics Committee of Provincial Hospital Affiliated to Shandong University. Written informed consents were obtained prior to experimentations.
Methods
Reagents and devices

Specimen grouping and intervention
Each specimen was collected from the center area of the scar tissue and sectioned into 0.5 cm × 0.5 cm-1.0 cm × 1.0 cm blocks. The tissues were stored in liquid nitrogen until further use[9].
Each tissue block was divided into two sets of slices. One set was cryosectioned (10-μm thickness), underwent immunofluorescent staining, and was analyzed using a fluorescent microscope. The other 30-μm thick tissue set underwent immunofluorescent staining and was analyzed by LSCM[10]. Negative controls were used in each group.
Staining and observation
The tissue sections (10-μm thick) from experimental and normal control groups were fixed in pure acetone at 4 °C for 10 minutes and immersed in 3% H2O2 at room temperature for 10 minutes. The sections were then rinsed three times with phosphate-buffered saline (PBS) for 5 minutes, followed by 10% goat serum at room temperature for 20 minutes. Following removal of serum without rinsing, 100 mL mouse anti-human neurofilament protein monoclonal antibody (1: 500 dilution) was added to the samples and incubated overnight, followed by a PBS rinse step. The samples were then incubated with FITC-labeled goat anti-mouse IgG (1: 250 dilution) in the dark at 37 °C for 30 minutes, rinsed three times in PBS, and the coverslips were coated with buffered glycerol (0.5 mol/L pH 9.0 of carbonate buffer solution mixed with glycerol of equal volume) for subsequent observation by fluorescence microscopy.
The 30-μm thick sections were stained using identical procedures and were observed with LSCM. The incubation time of the primary and second antibodies was extended to 72 hours and 1 hour, respectively, and the rinsing time was extended to 30 minutes[11]. In the negative control group, the primary antibody was replaced by PBS.
In both methods, positive immunofluorescent staining exhibited bright green fluorescent excitation at 488 nm.
The LSCM procedure was as follows: (1) the observation region was selected; (2) sequential scanning: the Z-axis was selected, and the entire wavelength was scanned at an interval of 0.5 μm; (3) three-dimensional reconstruction: all, or part, of the optical slices were selected and sequentially overlapped to form three-dimensional images[12].
Data collection and analysis
Morphological analysis software (Leica Qwin Pro V3.3.1 Wetzlar, Germany) was used to perform semi-quantitative analysis of the images obtained by fluorescence microscopy (200 × magnification). The positive area ratio (area of positive nerve fibers/total area) was calculated. In addition, morphological changes revealed by LSCM  (480 × magnification) were described in detail. Experimental results were analyzed using a blind method.
Main outcome measures
Morphology and positive area ratios of neurofilament protein-positive nerve fibers were measured.
Statistical analysis 
Data were expressed by Mean ± SD and analyzed using statistical software SPSS 11.5 (SPSS, IL, Chicago, USA). One-way analysis of variance and least-significant- difference t-test were respectively applied for multi-group and intergroup comparison. A value of P < 0.05 was considered statistically significant.

RESULTS

General subject data (Table 1)
 

Observation results of neurofilament protein-positive nerve fibers using two-dimensional imaging and fluorescent microscopy
Nerve fibers presented as well-distributed bright green strips with regular morphology. Thick nerve tracts were rarely observed in the normal control group     (Figure 1A).
Nerve fibers of varying length and thickness, with richer branches that twisted and wound, formed a more prominent network. In most specimens of the proliferative group, clusters of nerve fibers were closely bunched together (Figure 1B).
In the mature group, nerve fibers, which were shorter in length and equal thickness, were regularly arranged with little connections among them (Figure 1C).
The positive area ratio of nerve fibers in the proliferative group (2.22 ± 0.35) % was significantly greater than the normal control group [(0.62 ± 0.23) %, P < 0.05]. The positive area ratio of nerve fibers in the mature group [(0.34 ± 0.20) %] was significantly less than the normal control group (P < 0.05).
Observation results of neurofilament protein-positive nerve fibers using three-dimensional imaging and LSCM
In normal skin, nerve fibers with intact and regular inner structure were common. A complete stereoscopic structure, without any pathological changes, was revealed following three-dimensional reconstruction (Figures 2A, 3A).
Sequential scanning and three-dimensional reconstruction revealed an incomplete inner structure with localized vertical or horizontal fractures, as well as damaged or collapsed structures, during the proliferative stage. Irregular swelling and distortion was also revealed. Pathological changes of varying degrees were observed at different levels and angles (Figures 2B, 3B).
In the mature group, a relatively complete inner structure, without obvious pathological changes, was observed during sequential scanning. The nerve fibers exhibited increasing continuity with few branches and equal thickness, which was similar to the normal control group (Figures 2C, 3C).
 

DISCUSSION

Hypertrophic scars are a common result of pathological healing. The number of nerve fibers significantly increases during early stages and gradually decreases with scar maturation. Regardless of whether the scar is at an early or late stage of maturation, more intensive innervation exists in hypertrophic scars than in normal skin[14]. However, results have shown[15] that within seven months post-burn, hypertrophic scars exhibit a lower intensity of innervation than both normal skin and non-hypertrophic scars. To study changes of nerve fibers more comprehensively, the present study utilized the positive area ratio as the evaluation index, because quantitative and morphological changes occurred in the hypertrophic scar. Results showed that the positive area ratio of nerve fibers in the proliferative group was significantly greater than in normal skin, but the mature group exhibited the opposite. These results suggested that the high density of nerve fibers could be associated with hyperplasia of scars. 
LSCM reveals minimal changes that are not detectable by ordinary light microscopy, which is due to of the 3-dimensional image of the axon[16]. Because a laser serves as the light source, LSCM provides more opportunities for observation, and this is due to reduced fluorescence quenching that typically occurs when sections are imaged for extended and repeated periods of time. In the present study, LSCM revealed pathological changes, such as distortion, fracture, and collapse in nerve fibers during the proliferative stage, which improved to normal levels during the mature stage.
Cutaneous nerves are commonly accompanied by arteries and veins. The skin of the torso and pate, with exception of extremities, is well innervated and exhibits a strong correlation between nerves and accompanying vessels[17]. Changes in capillary vessels in scars impact on regeneration of nerve fibers, because the accompanying vessels serve as nutrient vessels. Studies have shown that hypertrophic scars and keloids exhibit increased microcirculation compared with normal skin and mature scar tissue[18]. In addition, the number of vessels increases during the proliferative stage[19], although the blood supply is not less compared with normal skin, even in the case of blockage in some vessels. This phenomenon was consistent with altered innervation observed in the present study. The increased number of neurofilament protein-positive nerve fibers and branches could compensate for fragmentation and stenosis of some nerve fibers, which could account for greater innervation of hypertrophic scar during the proliferative stage compared with normal tissue.
The interaction between nerve fibers and hypertrophic scars remains controversial. A previous study has demonstrated that scars formed in areas close to nerve fibers affect nerve function in cases of nerve entrapment syndrome and nerve trauma[20]. The nervous tissue scar was similar to other scar tissue and was composed of excessive fibroblasts and collagen-based extracellular matrix[21]. Nerve fibers are affected not only by the exterior of the scar outside but also by the interior. The traumatic neuroma is a tangle of neural fibers and connective tissue, which develops following nerve injury. Microscopically, the outer layer of the neuroma, formed by a fibrous connective tissue, is continuous with the perineurium of the intact nerve sheath. The present study attempted to provide a clear and visual morphological explanation for the effect of nerves on scarring.
Bias and limitations of this study are as follows: hypertrophic scars are influenced by many factors. However, the present study did not consider age, gender, race, or injury causes. Furthermore, skin from a person without scars was not used; instead, a donor site from the scar patient was used, due to ethical restrictions. This could affect accuracy of our results.

REFERENCES

[1] Zhang LQ, Laato M. Innervation of normal and hypertrophic human scars and experimental wounds in the rat. Ann Chir Gynaecol. 2001;90(215):29-32.

[2] Hsieh ST, Lin WM. Modulation of keratinocyte proliferation by skin innervation. J Invest Dermatol. 1999;113(4):579-586.
[3] Hakvoort TE, Altun V, Ramrattan RS, et al. Epidermal participation in post-burn hypertrophic scar development. Virchows Arch. 1999;434(3):221-226.
[4] Akaishi S, Ogawa R, Akimoto M, et al. The relationship between keloid growth pattern and stretching tension: visual analysis using the finite element method. Ann Plast Surg. 2008;60(4):445-451.
[5] Yu F, Yuan SJ, Sui G, et al. Neuronal apoptosis and neurofilament protein expression in the lateral geniculate body of cats following acute optic nerve injuries. Neural Regen Res. 2007;2(10): 577-581.
[6] Wu MY, Wu XY, Zheng RM, et al. Observation of immunohistochemical staining specimens with laser scanning confocal microscope. Linchuang yu Shiyan Binglixue Zazhi. 2001;17(6):533-534.
[7] Li HY, Lu KH, Guo SZ. New Edition of Study on Scars. Xi’an: the Fourth Military Medical University Press. 2003:20-21.
[8] State Council of the People's Republic of China. Administrative Regulations on Medical Institution. 1994-09-01.
[9] Zhang KP, Zhou Y, Long F. Making method of frozen section of different tissues. Jiangxi Yixueyuan Xuebao. 2006;46(4):13.
[10] Besne I, Descombes C, Breton L. Effect of age and anatomical site on density of sensory innervation in human epidermis. Arch Dermatol. 2002;138(11):1445-1450.
[11] Xue JW, Yu L, Sun L, et al. Effect of verapamil on the morphology of fibroblast existed in nerve’s scar tissue and cultured in vitro. Zhongguo Laonianxue Zazhi. 2008;11(28):2108-2109.
[12] Deng P, Xu XH, Chen YW, et al. Observation of axons in brainstem with laser scanning confocal microscope. Zhonghua Wuli Yixue yu Kangfu Zazhi. 2000;22(1):23-24.
[13] Reilly DM, Ferdinando D, Johnston C, et al. The epidermal nerve fibre network: characterization of nerve fibres in human skin by confocal microscopy and assessment of racial variations. Br J Dermatol. 1997;137(2):163-170.
[14] Cheng B, Fu XB, Sheng ZY, et al. Biologic significance of neurofilament protein expression in scar tissue. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue. 2003;15(12):715-717.
[15] Altun V, Hakvoort TE, van Zuijlen PP, et al. Nerve outgrowth and neuropeptide expression during the remodeling of human burn wound scars. A 7-month follow-up study of 22 patients. Burns. 2001;27(7): 687-689.
[16] Smith GJ, Bagnell CR, Bakewell WE, et al. Application of confocal scanning laser microscopy in experimental pathology. J Electron Microsc Tech. 1991;18(1):38-49.
[17] Taylor GI, Gianoutsos MP, Morris SF. The neurovascular territories of the skin and muscles anatomic study and clinical implications. Plast Reconstr Surg. 1994;94(1):1-5.
[18] Burd A, Huang L. Hypertrophic response and keloid diathesis: two very different forms of scar. Plast Reconstr Surg. 2005;116(7): 150e-157e.
[19] Polo M, Ko F, Busillo F, et al. The 1997 Moyer Award. Cytokine production in patients with hypertrophic burn scars. J Burn Care Rehabil. 1997;18(6):477-482.
[20] Kroeber MW, Diao E, Hida S, et al. Peripheral nerve lengthening by controlled isolated distraction: a new animal model. J Orthop Res. 2001;19(1):70-77 .
[21] Diegelmann RF, Peterkofsky B. Inhibition of collagen secrection from bone and cultured fibroblasts by microtubular disruptive drugs. Proc Natl Acad Sci U S A. 1972;69(4):892-896.
 (Edited by Li HM/Qiu Y/Song LP)