- Email: [email protected]
Altered proliferation and differentiation of human epidermis in cases of skin fibrosis after radiotherapy
Int. J. Radiation Oncology Biol. Phys., Vol. 53, No. 2, pp. 385–393, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/02/$–see front matter
PII S0360-3016(01)02732-8
CLINICAL INVESTIGATION
Normal Tissue
ALTERED PROLIFERATION AND DIFFERENTIATION OF HUMAN EPIDERMIS IN CASES OF SKIN FIBROSIS AFTER RADIOTHERAPY VIRGINIE SIVAN, PH.D.,* MARIE-CATHERINE VOZENIN-BROTONS, PH.D.,† YVES TRICAUD,‡ JEAN-LOUIS LEFAIX, PH.D.,§ JEAN-MARC COSSET, M.D.,㛳 BERNARD DUBRAY, M.D.,㛳 AND MICHE` LE T. MARTIN, PH.D.* *Service de Ge´nomique Fonctionnelle, CEA, DSV, DRR, Evry, France; †Laboratoire de Radiosensibilite´, IPSN, IGR, Villejuif, France; ‡ Laboratoire de Radiobiologie et d’Etude du Ge´nome, CEA, DSV, DRR, Gif sur Yvette, France; §Laboratoire de Radiotoxicologie, CEA, DSV, DRR, Bruye`re le Chaˆtel, France; 㛳De´partement d’Oncologie/Radiothe´rapie, Institut Curie, Paris, France Purpose: To characterize, at the histopathologic and molecular levels, the irradiated epidermis in cases of human skin fibrosis induced by radiotherapy. Methods and Materials: Surgical samples were obtained from 6 patients who had developed cutaneous fibronecrotic lesions from 7 months to 27 years after irradiation. The proliferation and differentiation status of the irradiated epidermis was characterized with specific markers using immunohistochemical methods. Results: All samples presented with hyperplasia of the epidermis associated with local inflammation. The scar epidermis exhibited an increased expression of proliferating cell nuclear antigen, which revealed hyperproliferation of keratinocytes. Furthermore, an abnormal differentiation was found, characterized by the expression of K6 and K16, and by alterations in protein amounts and localization of cytokeratins, involucrin, and transforming growth factor-1. Conclusion: These results demonstrate that late damage of irradiated skin is not only characterized by fibrosis in the dermis but also by hyperplasia in the epidermis. This hyperplasia was due to both hyperproliferation and abnormal differentiation of keratinocytes. © 2002 Elsevier Science Inc. Hyperplastic epidermis, Radiation fibrosis, Proliferation, Cytokeratin, TGF-1.
tissue by a wound healing process. After activation by cytokines and growth factors, fibroblasts differentiate into myofibroblasts, proliferate, and deposit extracellular matrix components. The constitution and remodeling of this scar tissue take place during the first years after irradiation and are principally controlled by myofibroblasts. In normal wound healing, myofibroblast activation is a transient process, but in fibrosis, the chronic activation of myofibroblasts leads to a long-term deposition and remodeling of the scar tissue. Several years after the end of the treatment, these processes decrease, and the very late phase of fibrosis is characterized by atrophy of the skin, associated with telangiectasia. Nevertheless, this lesion can be potentially reactivated when exposed to any mechanical or chemical trauma, which can lead to late necrosis of the skin. Because of these chronic cellular activations, radiation-induced fibrosis can be considered a perpetual wound.
INTRODUCTION Radiation-induced fibrosis is one of the most common sequelae occurring after therapeutic or accidental irradiation of the skin and may constitute a problem in clinical practice (1). The early effects occurring during radiotherapy (RT) within the first 6 weeks are characterized by depilation and erythema, followed by dry and moist desquamation of the epidermis that can lead to either healing of the lesion or radiation necrosis. During the next few months or years, fibrosis may develop, exhibiting the typical stages of skin reaction. Schematically, fibrosis development is divided into 3 phases. During the first months after the end of the treatment, the initiation phase is characterized by an acute inflammatory reaction in which endothelial cells seem to play a significant role. The vascular permeability induces the extravasation of serum proteins, which attract inflammatory and mesenchymal cells that initiate the restoration of the damaged
Acknowledgments—We warmly thank Dr. X. Sastre (Curie Institute, France) for his help and comments, and Drs. S. Delanian (St. Louis Hospital, France) and D. A. Lawrence (LREG, CEA) for their careful reading of the manuscript. Received Jun 6, 2001, and in revised form Oct 17, 2001. Accepted for publication Oct 23, 2001.
Reprint requests to: Miche`le T. Martin, Ph.D., Service de Ge´nomique Fonctionnelle, 2 Rue Gaston Cre´mieux, CP 22, 91057 Evry cedex, France. Tel: 01 60 87 34 91; Fax: 01 60 87 34 98; E-mail: [email protected] Supported by CEE research Grant FI4P-CT95-0029, by the Comite´ de Radioprotection d’Electricite´ de France and by thesis grants from CIFRE/Oxykin Therapeutics and from the Ligue Contre le Cancer. 385
386
I. J. Radiation Oncology
● Biology ● Physics
Fibrosis is a dynamic process implicating several cell types, including inflammatory, endothelial, and epithelial cells, as well as fibroblasts (2, 3). Several studies have been devoted to the radiation-induced fibrosis of the human dermis and thus to the role of fibroblasts and myofibroblasts. On the contrary, the scar epidermis, overlaying the fibrotic dermis, has been poorly characterized and, except in the work of Benyahia and Magdelena (4), has never been studied at the molecular level. Nevertheless, epithelial cells have been shown to contribute to the fibrotic process in several organs. For example, during the development of radiation-induced pulmonary fibrosis in mice, epithelial cells play a significant role by growth factor secretion and by intercommunication with mesenchymal cells (2). Many studies have reported the importance of interactions between keratinocytes and fibroblasts during normal and pathologic situations. For example, keratinocytes can regulate the collagen synthesis of dermal fibroblasts (5), and, conversely, fibroblasts can control the growth of keratinocytes (6). Moreover, activated keratinocytes are a source of specific proinflammatory cytokines and are involved in inflammation and wound healing processes (7). Recently, we described the hypertrophic morphology of the human scar epidermis located around fibronecrotic lesions induced by accidental irradiation (8). In this study, we assessed the damage induced in the epidermis by irradiation in cases of RT for breast cancer or of accidental irradiation. We also further characterized the fibrosis by the study of keratinocyte differentiation and proliferation. Concerning cell differentiation, keratin, involucrin, and integrin are major markers of keratinocyte maturation. Cytokeratins, members of the intermediate filament family, contribute to the constitution of the keratinocyte cytoskeleton and provide the structural resilience of the epidermis. In normal epidermis, proliferative cells are located in the basal layer where they strongly express K5 and K14 keratins (9). In this layer, around 30% of the keratinocytes express the proliferating cell nuclear antigen (PCNA) marker (10). PCNA is a 36-kD molecule that functions as a cofactor for DNA polymerase ␦ both in the S phase and in DNA synthesis associated with DNA repair. It is thus a good marker to discriminate proliferative cells from those entering the differentiation process. By withdrawing from the cell cycle and migrating upward, keratinocytes differentiate and exhibit K1 and K10 expression (9). During wound healing, the disruption of epidermis maturation is associated with a specific expression of K6 and K16 keratins, which are absent from normal epidermis (11). Involucrin and integrins constitute other widely used markers of differentiation. Involucrin is a cytoplasmic protein precursor of the cornified envelope and its expression in the epidermis is related to the terminal differentiation status of the keratinocyte. 1-Integrin, which constitutes the -subunit of several extracellular matrix receptors, is specifically ex-
Volume 53, Number 2, 2002
Table 1. Clinical description of healthy and irradiated patients and localization of the surgical biopsies Pt. No.
Age (y)/Gender
Irradiation type
Time after Biopsy site irradiation
1 2 3 4 5
64/F 72/F 62/F 77/F 21/M
RT RT RT RT Accidental
27 y 22 y 24 y 22 y 18 mo
6 Normal skin Normal skin Normal skin
23/M 17/F 64/F 21/M
Accidental
7 mo
Breast Breast Breast Breast Inguinal region Back Breast Abdominal Inguinal region
Abbreviations: Pt. No. ⫽ patient number; F ⫽ Female; RT ⫽ radiotherapy; M ⫽ male.
pressed in keratinocytes from the basal layer to allow epidermis adhesion to the basal membrane (12). Keratinocytes are also known to secrete several cytokines, which are involved in the maturation of the epidermis. For instance, transforming growth factor (TGF)-1 is involved in the initiation of keratinocyte differentiation by blocking their proliferation (13). This cytokine is implicated in both normal and abnormal processes. During wound healing, TGF-1 induces cell migration and matrix secretion by keratinocytes (7); it has also been described in fibrosis as a molecular switch for the fibrotic program (14). Moreover, after acute ␥-irradiation, we showed that TGF-1 expression was induced in the skin within hours (15) and that it remained high during the subsequent stages that led to the development of fibrosis (14). The aim of the present study was to characterize, at the histopathologic and molecular levels, the human scar epidermis induced by ionizing radiation. We studied surgical samples from 6 patients who developed cutaneous fibronecrotic lesions, from 7 months to 27 years after irradiation, and we compared the lesions resulting from therapeutic RT (4 patients) with those resulting from accidental irradiation (2 patients). We found that all these samples presented with a hyperplastic epidermis, and we established the proliferation and differentiation status of this scar epidermis.
METHODS AND MATERIALS Patients Excess surgical tissue from resected skin was obtained from 6 patients presenting with cutaneous fibronecrotic lesions resulting from therapeutic or accidental irradiation (Table 1). The RT patients (Patients 1– 4) exhibited cutaneous fibronecrotic lesions 22 to 27 years after breast cancer RT. All these patients developed skin necrosis without healing that required surgical resection. The irradiation pro-
Altered proliferation and differentiation in skin fibrosis after RT
Table 2. List of antibodies used in the immunohistochemical study Antibody
Target
Dilution
Source
Mouse monoclonal
K6
1:500
Mouse monoclonal
K16
1:500
Mouse monoclonal
K10
1:500
Mouse monoclonal
K14
1:500
Mouse monoclonal
Involucrin
1:5000
Mouse monoclonal
Integrin-1
1:500
Mouse monoclonal
PCNA
1:500
Rabbit polyclonal
TGF-1
1:500
Progen Biotechnik GmbH Novocastra Laboratories Ltd. Progen Biotechnik GmbH Novocastra Laboratories Ltd. Novocastra Laboratories Inc. Santa Cruz Biotechnology Inc. Dako, Glostrup, Denmark Santa Cruz Biotechnology, Santa Cruz, CA
Abbreviations: K ⫽ keratin; PCNA ⫽ proliferating cell nuclear antigen; TGF ⫽ transforming growth factor.
tocols could be obtained only for Patients 1 and 4. Patient 1 underwent daily RT with cobalt and betatron and received a total dose of 85 Gy. Patient 4 also underwent daily RT with cobalt and received a total dose of 85 Gy. Patients 5 and 6 developed, respectively, fibrotic and fibronecrotic skin lesions 7 and 18 months after accidental exposure to a 137Ce source (16). These lesions were surgically removed at the Curie Institute (Paris), and the samples were either fixed in formalin for immunohistochemistry studies or frozen in liquid nitrogen for protein extraction. Three biopsies of normal skin were obtained: one biopsy from a breast plastic surgery specimen of a 17-year-old girl and two biopsies from the abdominal skin and from the nonaltered margin of the surgical wound, in the case of Patients 1 (64 years old) and 5 (21 years old), respectively. Histopathologic evaluation and immunohistochemistry The samples were fixed in 10% formalin-buffered solution and embedded in paraffin wax. Serial 5-m sections were cut, dewaxed, and either stained with hematoxylin-eosin-safranin or treated for immunohistochemical detection. To unmask antigens, a chemical and heat treatment was performed by incubating sections in a 0.01 M sodium citrate buffer (pH 6) at 95°C for 5 min. The sections were then incubated in a 0.3% H2O2–phosphate-buffered saline (PBS) solution for 10 min to quench endogenous peroxidases and in a 3% bovine serum albumin–PBS solution for 30 min to block nonspecific sites. Sections were incubated overnight at 4°C in a PBS solution containing the primary antibody. All the antibodies used in this immunohistochemical study are listed in Table 2. The primary antibodies were revealed by the LSAB 2 Kit (Dako, Glostrup,
● V. SIVAN et al.
387
Denmark). In brief, the slides were submitted to sequential 10-min incubations with biotinylated link antibody and peroxidase-labeled streptavidin followed by a DAB⫹ (Dako) development. Protein extraction The epidermis of normal skin from Patient 1 and of fibronecrotic skin from Patients 1 and 6 were crushed to powder in liquid nitrogen in a freezer/mill (Spex 6700). Every sample was homogenized in 50 mM Tris-HCl buffer (pH 7.4) containing 600 mM KCl, 5 mM ethylenediaminetetraacetic acid, 5 mM ethylene glycol bis(-aminoethyl ether-tetraacetic acid (EGTA), 1% triton X-100, and protease inhibitors (cocktail tablets, Roche Diagnostics). After homogenization in a tissue grinding tube (Dounce), samples were centrifuged at 10,000 rpm for 30 min. To determine the quantity of extracted proteins in the supernatant, the DC protein assay (Bio-Rad, Hercules) was used. Western blot analysis Total proteins (30 g) from the epidermis were fractionated on a 10% sodium dodecyl sulfate-polyacrylamide gel. After blotting, the membrane was incubated overnight at 4°C with 3% dry milk in PBS (pH 7) to block nonspecific sites. Next, the membrane was incubated for 1 h at room temperature with the mouse monoclonal antibody anti-PCNA (Dako) at dilution 1:200 or with the monoclonal antibody anti-glyceraldehyde-3phosphate dehydrogenase (GAPDH, Biodesign International) at dilution 1/1000 in 3% milk/0.2% Tween-20/PBS buffer. The membrane was washed with phosphate-buffered saline/Tween (PBST) and incubated for 1 h with peroxidase-linked antimouse antibody (Amersham Pharmacia, Buckinghamshire, UK) at dilution 1:1000 before development with the enhanced chemiluminescence reaction (ECL Kit, Amersham Pharmacia). The blot was scanned and the intensity of the bands quantified with the program Molecular Analyst (data not shown). RESULTS Morphologic changes in fibrotic skin The examination of the fibronecrotic skin biopsies from the 6 patients after either RT or accidental irradiation (Table 1) revealed common characteristics. The dermis was hyalinized and exhibited disorganized collagen and extracellular matrix fibers, accumulation of myofibroblasts, and excessive neovascularization. An inflammatory reaction was generalized in the skin, with polymorphonuclear and lymphocyte infiltration in both the papillary and the reticular dermis, associated with mild edema and vascular fibrinoid tissular necrosis (Fig. 1Ab). Skin biopsies from the 6 patients exhibited heterogeneity within the lesions, with areas of normal-like skin structures adjacent to strongly disorganized zones. The epidermis overlaying the fibrotic dermis, herein referred to as scar epidermis, was significantly modified. In the biopsies of the 6 patients after either RT or accidental
388
I. J. Radiation Oncology
● Biology ● Physics
Volume 53, Number 2, 2002
Fig. 1
Altered proliferation and differentiation in skin fibrosis after RT
● V. SIVAN et al.
389
Table 3. Summary of results of immunohistochemical stainings performed on normal and scar epidermis of all patients, using differentiation and proliferation markers Scar epidermis Hyperplastic areas
Normal epidermis
Nonhyperplastic areas
Marker
b
s
g
h
b
s
g
h
b
s
g
h
K6 K16 K10 K14 Integrin-1 Involucrin PCNA TGF-1
⫺ ⫺ ⫺ ⫹⫹ ⫹ ⫺ ⫹⫹ ⫺
⫺ ⫺ ⫹⫹ ⫺ ⫺ ⫺ ⫺ ⫹
⫺ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫹
⫺ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫺
⫺ ⫺ ⫺ ⫹⫹ ⫹ ⫺ ⫹⫹ ⫹⫹
⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫺ ⫹⫹ ⫹ ⫹⫹
⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫺ ⫹⫹ ⫹ ⫹⫹
⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫺
⫺ ⫺ ⫺ ⫹⫹ ⫹ ⫺ ⫹⫹ ⫹
⫺ ⫹ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫹ ⫹
⫹ ⫹ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫹
⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫺
Abbreviations: b ⫽ basal layer; s ⫽ spinous layer; g ⫽ granular layer; h ⫽ horny layer; other abbreviations as in Table 2. Symbols: ⫺ ⫽ absence of staining; ⫹ ⫽ weak staining; ⫹⫹ ⫽ strong staining.
irradiation, the epidermis exhibited heterogeneous thickness, with a majority of hyperplastic areas, adjacent to normal ones. Acanthosis, defined as hyperplasia of the living cell layers of the epidermis, was observed for all patients (Fig. 1Ab). Keratosis, defined as hyperplasia of the stratum corneum, was observed only in Patient 5. No atrophic area was observed in any of the 6 patients. The hyperplastic epidermis was at least 5-fold thicker than normal (250 m vs. 50 m) (Fig. 1B). This thickness was due to an increased number of cellular layers; the normal epidermis exhibited 3– 6 layers of keratinocytes, whereas the number of cellular layers in the scar epidermis was 10 –30. The abnormal thickness of the scar epidermis was also related to a modification of the keratinocyte morphology (Fig. 1C). In the suprabasal layers of the scar epidermis, we observed either hypertrophic or abnormally small keratinocytes. The diameter of the hypertrophic cells was 13–18 m, corresponding to twice the diameter of normal cells. They exhibited either a rounded shape with hypertrophic cytoplasm and a round nucleus or a spindle shape oriented toward the horny layer with elongated cytoplasm and an oval nucleus. We also observed abnormally small keratinocytes 3 m in diameter, with a polygonal shape similar to that observed in the basal layer of the normal epidermis. The scar epidermis exhibited a significant widening of the intercellular space (Fig. 1C) compared with normal epidermis, in which a close arrangement of keratinocytes maintains the skin barrier. This widening suggested alterations in the network of intercellular junction proteins.
Keratinocyte activation: cytokeratin K6 and K16 expression K6 and K16 were studied because they are specific markers of keratinocytes exhibiting activated proliferation. In the normal epidermis, no expression of these proteins was observed; however, these cytokeratins were strongly expressed in the scar epidermis (Fig. 2). In the basal layer, K6 staining was either present or absent, depending on the area, and K16 was never expressed. In the suprabasal layers, K6 and K16 shared the same general pattern of expression, with staining observed either in the spinous layers or in all the suprabasal layers. However, no specific correlation could be established between this heterogeneous staining and the variation of the epidermis thickness or keratinocyte morphology. Abnormal keratinocyte differentiation: cytokeratins K10 and K14, 1-integrin and involucrin expression Because K6 and K16 revealed an activated phenotype of the scar keratinocytes, the maturation of the epidermis was further studied with several differentiation markers. For 2 proteins, K10 and 1-integrin, no difference was observed between the normal and scar epidermis (Table 3). K10 was expressed only in the suprabasal layers, and 1-integrin expression was restricted to the basal layer. In contrast, the expression of cytokeratin K14 and involucrin was altered in scar epidermis (Table 3). In the normal epidermis, K14 was restricted to the basal layer, but in the scar epidermis, its expression was detected
Fig. 1. (A) Morphology of normal skin and fibrosis. (a) Hematoxylin-eosin-safranin staining of normal skin. (b) Hematoxylin-eosin-safranin staining of a fibrotic dermis associated with an hyperplastic epidermis. Original magnification ⫻10. (B) Comparison of normal and scar epidermis thickness. The thickness was measured for 3 normal and 6 fibrotic skin biopsies. The histograms present the results of 10 measurements by biopsy. The measures were taken at random across each normal biopsy. For the fibrosis biopsies, the measurements were taken only in the hyperplastic epidermis, because the epidermis thickness was heterogeneous with the normal and hyperplastic areas (Student’s t test, p ⬍0.05). (C) Abnormal morphologies of scar keratinocytes. Black arrows show the abnormal widening of intercellular space. (a) Hypertrophic keratinocytes with a rounded shape. (b) Elongated keratinocytes with a spindle shape. Original magnification ⫻20.
390
I. J. Radiation Oncology
● Biology ● Physics
Volume 53, Number 2, 2002
Fig. 2. Alterations of cell differentiation and proliferation in the scar epidermis from Patient 5 for the following proteins (original magnification ⫻20): K6 in (a) normal and (b) scar epidermis; K16 in (c) normal and (d) scar epidermis; TGF-1 in (e) normal and (f) scar epidermis; and PCNA in (g) normal and (h) scar epidermis. These results were representative of all the patients studied, because the epidermis of the other 5 patients exhibited the same pattern of protein expression.
Altered proliferation and differentiation in skin fibrosis after RT
● V. SIVAN et al.
391
Fig. 3. Differential expression of PCNA protein in normal and scar epidermis. Western blot analysis was carried out with proteins extracted from two normal (breast and abdominal skin) and two scar epidermis from Patient 5 (accidental lesion) and Patient 1 (therapeutic lesion). GAPDH was used as a control of protein loading.
from the basal up to the suprabasal layers. For each biopsy, we observed a heterogeneous staining, either confined to the basal and spinous layers or localized in the whole epidermis. Involucrin expression in the normal epidermis was detected in the granular layer, but it extended up to the spinous layer in the scar epidermis. This abnormal pattern of involucrin expression was associated with the thickness of the scar epidermis. In the less hypertrophic areas, involucrin was localized in the granular layer, as observed in the normal epidermis. However, when the epidermis exhibited abnormal thickness, the staining extended up to the spinous layer. TGF-1 expression Because TGF-1 is involved in both epidermis differentiation and the fibrotic process, we checked its expression in the scar epidermis. Interestingly, the immunohistochemical study revealed alterations of TGF-1 expression. In the normal epidermis, TGF-1 was restricted to the suprabasal layers, for which the staining was weak and fuzzy. By contrast, in the scar epidermis, TGF-1 expression was strong, with a punctiform staining, which was observed in the upper layers but also in the basal layer (Fig. 2). Abnormal keratinocyte proliferation: PCNA expression To assess whether the alterations of epidermis differentiation were associated with alterations in cell proliferation, we studied the expression of the proliferation marker, PCNA. Significant differences in PCNA expression between normal and scar epidermis were revealed by immunoperoxidase staining (Fig. 2). In normal epidermis, PCNA expression was restricted to the basal layer where only the proliferative keratinocytes were stained. In the scar epidermis, the staining was heterogeneous. Depending on the area, the expression of PCNA was restricted to the basal layer or extended up to the spinous layer or up to the horny layer. The differential expression of PCNA between hyperplastic
and normal epidermis was confirmed by Western blot analysis (Fig. 3). PCNA protein expression was at least 4-fold stronger in the scar epidermis than in the normal one. These results show that scar epidermis exhibits cell alterations at both the differentiation and the proliferation levels.
DISCUSSION Because skin fibrosis is mainly a mesenchymal disease, little is known about the concomitant epidermal alterations and the possible role of keratinocytes in the fibrotic process. Benyahia and Magdelena (4) studied superficial skin fibrosis resulting from therapeutic irradiation and described molecular alterations in both the dermis and the epidermis. They showed that the scar epidermis exhibited an increased expression of epidermal growth factor receptor and TGF-1 and suggested altered interactions between the epidermal and dermal skin compartments. More recently, we described a fibronecrotic lesion of the skin, resulting from accidental ␥-irradiation and showed that the scar epidermis exhibited a hyperplastic phenotype, as well as an abnormal expression of tumor necrosis factor-␣ and TGF-1 (8). In the present work, we characterized the proliferation and differentiation status of the epidermis of irradiated patients with specific markers. We demonstrated that the scar epidermis exhibits a hyperplastic morphology and an increased expression of PCNA, both of which reveal cell hyperproliferation. Furthermore, we described an abnormal differentiation characterized by the expression of K6 and K16, a general disorganization of the cytokeratins and involucrin, and a paradoxical overexpression of TGF-1. In the literature, histologic studies of superficial radiation-induced fibrosis described both hyperplastic (8) and atrophic (17) epidermis, depending on the stage of the fibrotic development. In our study, no epidermis atrophy
392
I. J. Radiation Oncology
● Biology ● Physics
was found. Interestingly, hyperplasia was observed in all patients, regardless of the length of time after irradiation (7 months to 27 years) or the type of irradiation (accidental or therapeutic). We propose that the observed increase of the epidermis thickness was due to the active status of the lesions, which were all characterized by strong inflammation, and the proliferation of myofibroblasts within the fibrotic tissue, suggesting strong interactions between dermal and epidermal cells. Conversely, atrophic epidermis is probably associated with quiescent phases, when inflammation is reduced and cell density decreases in the fibrotic tissue. To go further into the understanding of the proliferation status of scar epidermis, we investigated PCNA expression and found abnormal expression of PCNA from the basal to the suprabasal layers that was surprisingly accompanied by a marked expression of TGF-1 in the whole scar epidermis. TGF-1 has been described as a strong inhibitor of epithelial cell proliferation (18). In normal epidermis, TGF-1 is absent from the basal layer where cells proliferate. Its expression in the suprabasal cells blocks the proliferation of keratinocytes, which then differentiate. In this study, the expression of TGF-1 in the basal layer was inconsistent with the hyperproliferative phenotype described for the scar epidermis. Our results could be explained by a loss of keratinocyte response to TGF-1 through alterations of the signal transduction pathway, such as has been described during transformation of keratinocytes in carcinogenesis (19). Our group recently described such a process for myofibroblasts from a porcine model of radiation fibrosis (20). The TGF-1 signal transduction pathway is currently under investigation in our group to assess this hypothesis for scar keratinocytes. TGF-1 is involved in normal processes such as epidermis maturation, but it also participates in both the initiation and the development of fibrosis. In human fibrotic lesions, induced by either RT or accidental irradiation, overexpression of TGF-1 protein has been described in situ in both dermis and epidermis and in vitro in cultured fibroblasts and keratinocytes (4, 8). Moreover, several studies have focused on its role in the dermis where it regulates myofibroblast proliferation and matrix component expression (14). We propose that scar keratinocytes are involved in the fibrotic process by interacting with the dermis cells, because TGF-1 expressed by keratinocytes could specifically target myofibroblasts in a paracrine way and then take part in the initiation and development of fibrosis. The results of the present study demonstrate the hyperproliferative status of the scar epidermis, but the differentiation also appeared disorganized. Typical expression of keratin K6 and K16 was found in the scar epidermis and thus demonstrated the activated phenotype of keratinocytes undergoing altered differentiation (21). To investigate this abnormal differentiation, we compared keratin, involucrin, and integrin expression in scar and normal epidermis. We found that the scar epidermis exhibited an altered K14 and
Volume 53, Number 2, 2002
involucrin expression but that K10 and 1-integrin expression remained normal. These features are typical of abnormal differentiation and could be compared with those found in several pathologic epidermis disorders. Expression of K6 and K16 in scar epidermis from fibrosis resembles that observed in psoriasis (22), squamous carcinoma (23), and hypertrophic scar (24). The scar epidermis also exhibited altered expression of K14 and involucrin, similar to that observed in psoriasis. The promoters of K5 and K14 have been shown to be activated by TGF-1 (25). Because TGF-1 was overexpressed in the scar epidermis, it could be implicated in the abnormal expression of K14 in the suprabasal layers. By contrast, 1-integrin expression remained the same in normal and scar epidermis, confined to the basal layer, but in psoriasis or normal wound healing, its expression would abnormally extend to the suprabasal layers (26). Moreover, in psoriasis, the increase of K16 is associated with a decrease in K1/K10 expression. We found no variation in K10 expression in the scar epidermis, similar to that found in squamous carcinoma (22) and hypertrophic scar (23). In conclusion, scar epidermis from fibrosis shares several, but not all, features of epidermal diseases, thus exhibiting a specific abnormal pattern of protein expression. Surprisingly, the abnormal differentiation observed in the scar epidermis from late radiation fibrosis exhibited large similarities with the alterations described in epidermis that has just been irradiated. Liu et al. (27) showed that after 1–3 weeks of daily X-ray irradiation, the leg epidermis of mice exhibited altered expression of K6, K10, K14, involucrin, and 1-integrin, similar to the expression we found in delayed radiation sequelae. It seems that, whatever the time after irradiation, the epidermis develops the same activated keratinocyte phenotype even if the cellular microenvironment is different. This phenomenon could be the consequence of irreversible alterations of the keratinocytes after irradiation, which could be reactivated even several years after treatment. However, it could also be related to inflammation, which is a common feature of both early and late responses of the skin to irradiation. The development of skin fibrosis after ionizing irradiation is the consequence of complex processes involving several cell types, cell– cell interactions, and cytokines, which have been extensively studied in the dermis. By contrast, the epidermis was poorly characterized at the cellular and molecular levels. The results of our work show that the epidermis overlaying superficial RT-induced fibrosis is strongly disorganized and presents characteristic features of abnormal proliferation and differentiation. These results open new areas for fibrosis investigation. Moreover, because major progress has recently been obtained in the treatment of superficial radiation-induced fibrosis (28 –30), we propose the keratinocyte as a new potential therapeutic target to downregulate the fibrotic process.
Altered proliferation and differentiation in skin fibrosis after RT
● V. SIVAN et al.
393
REFERENCES 1. Archambeau JO, Pezner R, Wasserman T. Pathophysiology of irradiated skin and breast. Int J Radiat Oncol Biol Phys 1995;31:1171–1185. 2. Rodemann HP, Bamberg M. Cellular basis of radiation-induced fibrosis. Radiother Oncol 1995;35:83–90. 3. Hill RP, Rodemann HP, Hendry JH, et al. Normal tissue radiobiology: From the laboratory to the clinic. Int J Radiat Oncol Biol Phys 2001;49:353–365. 4. Benyahia B, Magdelena H. Immunohistochemical characterization of human gamma-irradiated skin. Bull Cancer Radiother 1993;80:126 –134. 5. Warren L, Garner MD. Epidermal regulation of dermal fibroblast activity. Plast Reconstr Surg 1998;102:135–139. 6. Rubin JS, Bottaro DP, Chedid M, et al. Keratinocyte growth factor as a cytokine that mediates mesenchymal-epithelial interaction. EXS 1995;74:191–214. 7. Moulin V. Growth factors in skin wound healing. Eur J Cell Biol 1995;68:1–7. 8. Vozenin-Brotons MC, Gault N, Sivan V, et al. Histopathological and cellular studies of a case of cutaneous radiation syndrome after accidental chronic exposure to a cesium source. Radiat Res 1999;152:332–337. 9. Fuchs E. Keratins and the skin. Annu Rev Cell Dev Biol 1995;11:123–153. 10. Furukawa F, Imamura S, Fujita M, et al. Immunohistochemical localization of proliferating cell nuclear antigen/cyclin in human skin. Arch Dermatol Res 1992;284:86 –91. 11. McGowan K, Coulombe PA. Subcellular biochemistry, intermediate filaments. New York: Plenum Press; 1998. 12. Watt FM, Jones PH. Expression and function of the keratinocyte integrins. Dev Suppl 1993;49:185–192. 13. Fuchs E, Byrne C. The epidermis: Rising to the surface. Curr Opin Genet Dev 1994;4:725–736. 14. Martin M, Lefaix J, Delanian S. TGF-1 and radiation fibrosis: A master switch and a specific therapeutic target? Int J Radiat Oncol Biol Phys 2000;47:277–290. 15. Martin M, Vozenin MC, Gault N, et al. Coactivation of AP-1 activity and TGF-beta1 gene expression in the stress response of normal skin cells to ionizing radiation. Oncogene 1997;15: 981–989. 16. Gottlober P, Bezold G, Weber L, et al. The radiation accident in Georgia: Clinical appearance and diagnosis of cutaneous radiation syndrome. J Am Acad Dermatol 2000;42:453– 458. 17. Hopewell JW. The skin: Its structure and response to ionizing radiation. Int J Radiat Biol 1990;57:751–773.
18. Moses HL, Yang EY, Pietenpol JA. TGF-beta stimulation and inhibition of cell proliferation: New mechanistic insights. Cell 1990;63:245–247. 19. Reiss M. Transforming growth factor-beta and cancer: A love-hate relationship? Oncol Res 1997;9:447– 457. 20. Reisdorf P, Lawrence D, Sivan V, et al. Alteration of transforming growth factor-1 response involves down-regulation of Smad3 signaling in myofibroblasts from skin fibrosis. Am J Pathol 2001;159:263–272. 21. Coulombe PA. Towards a molecular definition of keratinocyte activation after acute injury to stratified epithelia. Biochem Biophys Res Commun 1997;236:231–238. 22. Leigh IM, Navsaria H, Purkis PE, et al. Keratins (K16 and K17) as markers of keratinocyte hyperproliferation in psoriasis in vivo and in vitro. Br J Dermatol 1995;133:501–511. 23. Stoler A, Kopan R, Duvic M, et al. Use of monospecific antisera and cRNA probes to localize the major changes in keratin expression during normal and abnormal epidermal differentiation. J Cell Biol 1988;107:427– 446. 24. Machesney M, Tidman N, Waseem A, et al. Activated keratinocytes in the epidermis of hypertrophic scars. Am J Pathol 1998;152:1133–1141. 25. Jiang CK, Tomic-Canic M, Lucas DJ, et al. TGF beta promotes the basal phenotype of epidermal keratinocytes: Transcriptional induction of K5 and K14 keratin genes. Growth Factors 1995;12:87–97. 26. Hertle MD, Kubler MD, Leigh IM, et al. Aberrant integrin expression during epidermal wound healing and in psoriatic epidermis. J Clin Invest 1992;89:1892–1901. 27. Liu K, Kasper M, Trott KR. Changes in keratinocyte differentiation during accelerated repopulation of the irradiated mouse epidermis. Int J Radiat Oncol Biol Phys 1996;69:763– 769. 28. Lefaix JL, Delanian S, Leplat JJ, et al. Successful treatment of radiation-induced fibrosis using Cu/Zn-SOD and Mn-SOD: An experimental study. Int J Radiat Oncol Biol Phys 1996; 35:305–312. 29. Vozenin-Brotons MC, Sivan V, Gault N, et al. Antifibrotic action of Cu/Zn SOD is mediated by TGF-beta1 repression and phenotypic reversion of myofibroblasts. Free Radic Biol Med 2001;30:30 – 42. 30. Delanian S, Balla-Mekias S, Lefaix JL. Striking regression of chronic radiotherapy damage in a clinical trial of combined pentoxifylline and tocopherol. J Clin Oncol 1999;17:3283– 3290.
PII S0360-3016(01)02732-8
CLINICAL INVESTIGATION
Normal Tissue
ALTERED PROLIFERATION AND DIFFERENTIATION OF HUMAN EPIDERMIS IN CASES OF SKIN FIBROSIS AFTER RADIOTHERAPY VIRGINIE SIVAN, PH.D.,* MARIE-CATHERINE VOZENIN-BROTONS, PH.D.,† YVES TRICAUD,‡ JEAN-LOUIS LEFAIX, PH.D.,§ JEAN-MARC COSSET, M.D.,㛳 BERNARD DUBRAY, M.D.,㛳 AND MICHE` LE T. MARTIN, PH.D.* *Service de Ge´nomique Fonctionnelle, CEA, DSV, DRR, Evry, France; †Laboratoire de Radiosensibilite´, IPSN, IGR, Villejuif, France; ‡ Laboratoire de Radiobiologie et d’Etude du Ge´nome, CEA, DSV, DRR, Gif sur Yvette, France; §Laboratoire de Radiotoxicologie, CEA, DSV, DRR, Bruye`re le Chaˆtel, France; 㛳De´partement d’Oncologie/Radiothe´rapie, Institut Curie, Paris, France Purpose: To characterize, at the histopathologic and molecular levels, the irradiated epidermis in cases of human skin fibrosis induced by radiotherapy. Methods and Materials: Surgical samples were obtained from 6 patients who had developed cutaneous fibronecrotic lesions from 7 months to 27 years after irradiation. The proliferation and differentiation status of the irradiated epidermis was characterized with specific markers using immunohistochemical methods. Results: All samples presented with hyperplasia of the epidermis associated with local inflammation. The scar epidermis exhibited an increased expression of proliferating cell nuclear antigen, which revealed hyperproliferation of keratinocytes. Furthermore, an abnormal differentiation was found, characterized by the expression of K6 and K16, and by alterations in protein amounts and localization of cytokeratins, involucrin, and transforming growth factor-1. Conclusion: These results demonstrate that late damage of irradiated skin is not only characterized by fibrosis in the dermis but also by hyperplasia in the epidermis. This hyperplasia was due to both hyperproliferation and abnormal differentiation of keratinocytes. © 2002 Elsevier Science Inc. Hyperplastic epidermis, Radiation fibrosis, Proliferation, Cytokeratin, TGF-1.
tissue by a wound healing process. After activation by cytokines and growth factors, fibroblasts differentiate into myofibroblasts, proliferate, and deposit extracellular matrix components. The constitution and remodeling of this scar tissue take place during the first years after irradiation and are principally controlled by myofibroblasts. In normal wound healing, myofibroblast activation is a transient process, but in fibrosis, the chronic activation of myofibroblasts leads to a long-term deposition and remodeling of the scar tissue. Several years after the end of the treatment, these processes decrease, and the very late phase of fibrosis is characterized by atrophy of the skin, associated with telangiectasia. Nevertheless, this lesion can be potentially reactivated when exposed to any mechanical or chemical trauma, which can lead to late necrosis of the skin. Because of these chronic cellular activations, radiation-induced fibrosis can be considered a perpetual wound.
INTRODUCTION Radiation-induced fibrosis is one of the most common sequelae occurring after therapeutic or accidental irradiation of the skin and may constitute a problem in clinical practice (1). The early effects occurring during radiotherapy (RT) within the first 6 weeks are characterized by depilation and erythema, followed by dry and moist desquamation of the epidermis that can lead to either healing of the lesion or radiation necrosis. During the next few months or years, fibrosis may develop, exhibiting the typical stages of skin reaction. Schematically, fibrosis development is divided into 3 phases. During the first months after the end of the treatment, the initiation phase is characterized by an acute inflammatory reaction in which endothelial cells seem to play a significant role. The vascular permeability induces the extravasation of serum proteins, which attract inflammatory and mesenchymal cells that initiate the restoration of the damaged
Acknowledgments—We warmly thank Dr. X. Sastre (Curie Institute, France) for his help and comments, and Drs. S. Delanian (St. Louis Hospital, France) and D. A. Lawrence (LREG, CEA) for their careful reading of the manuscript. Received Jun 6, 2001, and in revised form Oct 17, 2001. Accepted for publication Oct 23, 2001.
Reprint requests to: Miche`le T. Martin, Ph.D., Service de Ge´nomique Fonctionnelle, 2 Rue Gaston Cre´mieux, CP 22, 91057 Evry cedex, France. Tel: 01 60 87 34 91; Fax: 01 60 87 34 98; E-mail: [email protected] Supported by CEE research Grant FI4P-CT95-0029, by the Comite´ de Radioprotection d’Electricite´ de France and by thesis grants from CIFRE/Oxykin Therapeutics and from the Ligue Contre le Cancer. 385
386
I. J. Radiation Oncology
● Biology ● Physics
Fibrosis is a dynamic process implicating several cell types, including inflammatory, endothelial, and epithelial cells, as well as fibroblasts (2, 3). Several studies have been devoted to the radiation-induced fibrosis of the human dermis and thus to the role of fibroblasts and myofibroblasts. On the contrary, the scar epidermis, overlaying the fibrotic dermis, has been poorly characterized and, except in the work of Benyahia and Magdelena (4), has never been studied at the molecular level. Nevertheless, epithelial cells have been shown to contribute to the fibrotic process in several organs. For example, during the development of radiation-induced pulmonary fibrosis in mice, epithelial cells play a significant role by growth factor secretion and by intercommunication with mesenchymal cells (2). Many studies have reported the importance of interactions between keratinocytes and fibroblasts during normal and pathologic situations. For example, keratinocytes can regulate the collagen synthesis of dermal fibroblasts (5), and, conversely, fibroblasts can control the growth of keratinocytes (6). Moreover, activated keratinocytes are a source of specific proinflammatory cytokines and are involved in inflammation and wound healing processes (7). Recently, we described the hypertrophic morphology of the human scar epidermis located around fibronecrotic lesions induced by accidental irradiation (8). In this study, we assessed the damage induced in the epidermis by irradiation in cases of RT for breast cancer or of accidental irradiation. We also further characterized the fibrosis by the study of keratinocyte differentiation and proliferation. Concerning cell differentiation, keratin, involucrin, and integrin are major markers of keratinocyte maturation. Cytokeratins, members of the intermediate filament family, contribute to the constitution of the keratinocyte cytoskeleton and provide the structural resilience of the epidermis. In normal epidermis, proliferative cells are located in the basal layer where they strongly express K5 and K14 keratins (9). In this layer, around 30% of the keratinocytes express the proliferating cell nuclear antigen (PCNA) marker (10). PCNA is a 36-kD molecule that functions as a cofactor for DNA polymerase ␦ both in the S phase and in DNA synthesis associated with DNA repair. It is thus a good marker to discriminate proliferative cells from those entering the differentiation process. By withdrawing from the cell cycle and migrating upward, keratinocytes differentiate and exhibit K1 and K10 expression (9). During wound healing, the disruption of epidermis maturation is associated with a specific expression of K6 and K16 keratins, which are absent from normal epidermis (11). Involucrin and integrins constitute other widely used markers of differentiation. Involucrin is a cytoplasmic protein precursor of the cornified envelope and its expression in the epidermis is related to the terminal differentiation status of the keratinocyte. 1-Integrin, which constitutes the -subunit of several extracellular matrix receptors, is specifically ex-
Volume 53, Number 2, 2002
Table 1. Clinical description of healthy and irradiated patients and localization of the surgical biopsies Pt. No.
Age (y)/Gender
Irradiation type
Time after Biopsy site irradiation
1 2 3 4 5
64/F 72/F 62/F 77/F 21/M
RT RT RT RT Accidental
27 y 22 y 24 y 22 y 18 mo
6 Normal skin Normal skin Normal skin
23/M 17/F 64/F 21/M
Accidental
7 mo
Breast Breast Breast Breast Inguinal region Back Breast Abdominal Inguinal region
Abbreviations: Pt. No. ⫽ patient number; F ⫽ Female; RT ⫽ radiotherapy; M ⫽ male.
pressed in keratinocytes from the basal layer to allow epidermis adhesion to the basal membrane (12). Keratinocytes are also known to secrete several cytokines, which are involved in the maturation of the epidermis. For instance, transforming growth factor (TGF)-1 is involved in the initiation of keratinocyte differentiation by blocking their proliferation (13). This cytokine is implicated in both normal and abnormal processes. During wound healing, TGF-1 induces cell migration and matrix secretion by keratinocytes (7); it has also been described in fibrosis as a molecular switch for the fibrotic program (14). Moreover, after acute ␥-irradiation, we showed that TGF-1 expression was induced in the skin within hours (15) and that it remained high during the subsequent stages that led to the development of fibrosis (14). The aim of the present study was to characterize, at the histopathologic and molecular levels, the human scar epidermis induced by ionizing radiation. We studied surgical samples from 6 patients who developed cutaneous fibronecrotic lesions, from 7 months to 27 years after irradiation, and we compared the lesions resulting from therapeutic RT (4 patients) with those resulting from accidental irradiation (2 patients). We found that all these samples presented with a hyperplastic epidermis, and we established the proliferation and differentiation status of this scar epidermis.
METHODS AND MATERIALS Patients Excess surgical tissue from resected skin was obtained from 6 patients presenting with cutaneous fibronecrotic lesions resulting from therapeutic or accidental irradiation (Table 1). The RT patients (Patients 1– 4) exhibited cutaneous fibronecrotic lesions 22 to 27 years after breast cancer RT. All these patients developed skin necrosis without healing that required surgical resection. The irradiation pro-
Altered proliferation and differentiation in skin fibrosis after RT
Table 2. List of antibodies used in the immunohistochemical study Antibody
Target
Dilution
Source
Mouse monoclonal
K6
1:500
Mouse monoclonal
K16
1:500
Mouse monoclonal
K10
1:500
Mouse monoclonal
K14
1:500
Mouse monoclonal
Involucrin
1:5000
Mouse monoclonal
Integrin-1
1:500
Mouse monoclonal
PCNA
1:500
Rabbit polyclonal
TGF-1
1:500
Progen Biotechnik GmbH Novocastra Laboratories Ltd. Progen Biotechnik GmbH Novocastra Laboratories Ltd. Novocastra Laboratories Inc. Santa Cruz Biotechnology Inc. Dako, Glostrup, Denmark Santa Cruz Biotechnology, Santa Cruz, CA
Abbreviations: K ⫽ keratin; PCNA ⫽ proliferating cell nuclear antigen; TGF ⫽ transforming growth factor.
tocols could be obtained only for Patients 1 and 4. Patient 1 underwent daily RT with cobalt and betatron and received a total dose of 85 Gy. Patient 4 also underwent daily RT with cobalt and received a total dose of 85 Gy. Patients 5 and 6 developed, respectively, fibrotic and fibronecrotic skin lesions 7 and 18 months after accidental exposure to a 137Ce source (16). These lesions were surgically removed at the Curie Institute (Paris), and the samples were either fixed in formalin for immunohistochemistry studies or frozen in liquid nitrogen for protein extraction. Three biopsies of normal skin were obtained: one biopsy from a breast plastic surgery specimen of a 17-year-old girl and two biopsies from the abdominal skin and from the nonaltered margin of the surgical wound, in the case of Patients 1 (64 years old) and 5 (21 years old), respectively. Histopathologic evaluation and immunohistochemistry The samples were fixed in 10% formalin-buffered solution and embedded in paraffin wax. Serial 5-m sections were cut, dewaxed, and either stained with hematoxylin-eosin-safranin or treated for immunohistochemical detection. To unmask antigens, a chemical and heat treatment was performed by incubating sections in a 0.01 M sodium citrate buffer (pH 6) at 95°C for 5 min. The sections were then incubated in a 0.3% H2O2–phosphate-buffered saline (PBS) solution for 10 min to quench endogenous peroxidases and in a 3% bovine serum albumin–PBS solution for 30 min to block nonspecific sites. Sections were incubated overnight at 4°C in a PBS solution containing the primary antibody. All the antibodies used in this immunohistochemical study are listed in Table 2. The primary antibodies were revealed by the LSAB 2 Kit (Dako, Glostrup,
● V. SIVAN et al.
387
Denmark). In brief, the slides were submitted to sequential 10-min incubations with biotinylated link antibody and peroxidase-labeled streptavidin followed by a DAB⫹ (Dako) development. Protein extraction The epidermis of normal skin from Patient 1 and of fibronecrotic skin from Patients 1 and 6 were crushed to powder in liquid nitrogen in a freezer/mill (Spex 6700). Every sample was homogenized in 50 mM Tris-HCl buffer (pH 7.4) containing 600 mM KCl, 5 mM ethylenediaminetetraacetic acid, 5 mM ethylene glycol bis(-aminoethyl ether-tetraacetic acid (EGTA), 1% triton X-100, and protease inhibitors (cocktail tablets, Roche Diagnostics). After homogenization in a tissue grinding tube (Dounce), samples were centrifuged at 10,000 rpm for 30 min. To determine the quantity of extracted proteins in the supernatant, the DC protein assay (Bio-Rad, Hercules) was used. Western blot analysis Total proteins (30 g) from the epidermis were fractionated on a 10% sodium dodecyl sulfate-polyacrylamide gel. After blotting, the membrane was incubated overnight at 4°C with 3% dry milk in PBS (pH 7) to block nonspecific sites. Next, the membrane was incubated for 1 h at room temperature with the mouse monoclonal antibody anti-PCNA (Dako) at dilution 1:200 or with the monoclonal antibody anti-glyceraldehyde-3phosphate dehydrogenase (GAPDH, Biodesign International) at dilution 1/1000 in 3% milk/0.2% Tween-20/PBS buffer. The membrane was washed with phosphate-buffered saline/Tween (PBST) and incubated for 1 h with peroxidase-linked antimouse antibody (Amersham Pharmacia, Buckinghamshire, UK) at dilution 1:1000 before development with the enhanced chemiluminescence reaction (ECL Kit, Amersham Pharmacia). The blot was scanned and the intensity of the bands quantified with the program Molecular Analyst (data not shown). RESULTS Morphologic changes in fibrotic skin The examination of the fibronecrotic skin biopsies from the 6 patients after either RT or accidental irradiation (Table 1) revealed common characteristics. The dermis was hyalinized and exhibited disorganized collagen and extracellular matrix fibers, accumulation of myofibroblasts, and excessive neovascularization. An inflammatory reaction was generalized in the skin, with polymorphonuclear and lymphocyte infiltration in both the papillary and the reticular dermis, associated with mild edema and vascular fibrinoid tissular necrosis (Fig. 1Ab). Skin biopsies from the 6 patients exhibited heterogeneity within the lesions, with areas of normal-like skin structures adjacent to strongly disorganized zones. The epidermis overlaying the fibrotic dermis, herein referred to as scar epidermis, was significantly modified. In the biopsies of the 6 patients after either RT or accidental
388
I. J. Radiation Oncology
● Biology ● Physics
Volume 53, Number 2, 2002
Fig. 1
Altered proliferation and differentiation in skin fibrosis after RT
● V. SIVAN et al.
389
Table 3. Summary of results of immunohistochemical stainings performed on normal and scar epidermis of all patients, using differentiation and proliferation markers Scar epidermis Hyperplastic areas
Normal epidermis
Nonhyperplastic areas
Marker
b
s
g
h
b
s
g
h
b
s
g
h
K6 K16 K10 K14 Integrin-1 Involucrin PCNA TGF-1
⫺ ⫺ ⫺ ⫹⫹ ⫹ ⫺ ⫹⫹ ⫺
⫺ ⫺ ⫹⫹ ⫺ ⫺ ⫺ ⫺ ⫹
⫺ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫹
⫺ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫺
⫺ ⫺ ⫺ ⫹⫹ ⫹ ⫺ ⫹⫹ ⫹⫹
⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫺ ⫹⫹ ⫹ ⫹⫹
⫹⫹ ⫹⫹ ⫹⫹ ⫹⫹ ⫺ ⫹⫹ ⫹ ⫹⫹
⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫺
⫺ ⫺ ⫺ ⫹⫹ ⫹ ⫺ ⫹⫹ ⫹
⫺ ⫹ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫹ ⫹
⫹ ⫹ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫹
⫹ ⫺ ⫹⫹ ⫺ ⫺ ⫹⫹ ⫺ ⫺
Abbreviations: b ⫽ basal layer; s ⫽ spinous layer; g ⫽ granular layer; h ⫽ horny layer; other abbreviations as in Table 2. Symbols: ⫺ ⫽ absence of staining; ⫹ ⫽ weak staining; ⫹⫹ ⫽ strong staining.
irradiation, the epidermis exhibited heterogeneous thickness, with a majority of hyperplastic areas, adjacent to normal ones. Acanthosis, defined as hyperplasia of the living cell layers of the epidermis, was observed for all patients (Fig. 1Ab). Keratosis, defined as hyperplasia of the stratum corneum, was observed only in Patient 5. No atrophic area was observed in any of the 6 patients. The hyperplastic epidermis was at least 5-fold thicker than normal (250 m vs. 50 m) (Fig. 1B). This thickness was due to an increased number of cellular layers; the normal epidermis exhibited 3– 6 layers of keratinocytes, whereas the number of cellular layers in the scar epidermis was 10 –30. The abnormal thickness of the scar epidermis was also related to a modification of the keratinocyte morphology (Fig. 1C). In the suprabasal layers of the scar epidermis, we observed either hypertrophic or abnormally small keratinocytes. The diameter of the hypertrophic cells was 13–18 m, corresponding to twice the diameter of normal cells. They exhibited either a rounded shape with hypertrophic cytoplasm and a round nucleus or a spindle shape oriented toward the horny layer with elongated cytoplasm and an oval nucleus. We also observed abnormally small keratinocytes 3 m in diameter, with a polygonal shape similar to that observed in the basal layer of the normal epidermis. The scar epidermis exhibited a significant widening of the intercellular space (Fig. 1C) compared with normal epidermis, in which a close arrangement of keratinocytes maintains the skin barrier. This widening suggested alterations in the network of intercellular junction proteins.
Keratinocyte activation: cytokeratin K6 and K16 expression K6 and K16 were studied because they are specific markers of keratinocytes exhibiting activated proliferation. In the normal epidermis, no expression of these proteins was observed; however, these cytokeratins were strongly expressed in the scar epidermis (Fig. 2). In the basal layer, K6 staining was either present or absent, depending on the area, and K16 was never expressed. In the suprabasal layers, K6 and K16 shared the same general pattern of expression, with staining observed either in the spinous layers or in all the suprabasal layers. However, no specific correlation could be established between this heterogeneous staining and the variation of the epidermis thickness or keratinocyte morphology. Abnormal keratinocyte differentiation: cytokeratins K10 and K14, 1-integrin and involucrin expression Because K6 and K16 revealed an activated phenotype of the scar keratinocytes, the maturation of the epidermis was further studied with several differentiation markers. For 2 proteins, K10 and 1-integrin, no difference was observed between the normal and scar epidermis (Table 3). K10 was expressed only in the suprabasal layers, and 1-integrin expression was restricted to the basal layer. In contrast, the expression of cytokeratin K14 and involucrin was altered in scar epidermis (Table 3). In the normal epidermis, K14 was restricted to the basal layer, but in the scar epidermis, its expression was detected
Fig. 1. (A) Morphology of normal skin and fibrosis. (a) Hematoxylin-eosin-safranin staining of normal skin. (b) Hematoxylin-eosin-safranin staining of a fibrotic dermis associated with an hyperplastic epidermis. Original magnification ⫻10. (B) Comparison of normal and scar epidermis thickness. The thickness was measured for 3 normal and 6 fibrotic skin biopsies. The histograms present the results of 10 measurements by biopsy. The measures were taken at random across each normal biopsy. For the fibrosis biopsies, the measurements were taken only in the hyperplastic epidermis, because the epidermis thickness was heterogeneous with the normal and hyperplastic areas (Student’s t test, p ⬍0.05). (C) Abnormal morphologies of scar keratinocytes. Black arrows show the abnormal widening of intercellular space. (a) Hypertrophic keratinocytes with a rounded shape. (b) Elongated keratinocytes with a spindle shape. Original magnification ⫻20.
390
I. J. Radiation Oncology
● Biology ● Physics
Volume 53, Number 2, 2002
Fig. 2. Alterations of cell differentiation and proliferation in the scar epidermis from Patient 5 for the following proteins (original magnification ⫻20): K6 in (a) normal and (b) scar epidermis; K16 in (c) normal and (d) scar epidermis; TGF-1 in (e) normal and (f) scar epidermis; and PCNA in (g) normal and (h) scar epidermis. These results were representative of all the patients studied, because the epidermis of the other 5 patients exhibited the same pattern of protein expression.
Altered proliferation and differentiation in skin fibrosis after RT
● V. SIVAN et al.
391
Fig. 3. Differential expression of PCNA protein in normal and scar epidermis. Western blot analysis was carried out with proteins extracted from two normal (breast and abdominal skin) and two scar epidermis from Patient 5 (accidental lesion) and Patient 1 (therapeutic lesion). GAPDH was used as a control of protein loading.
from the basal up to the suprabasal layers. For each biopsy, we observed a heterogeneous staining, either confined to the basal and spinous layers or localized in the whole epidermis. Involucrin expression in the normal epidermis was detected in the granular layer, but it extended up to the spinous layer in the scar epidermis. This abnormal pattern of involucrin expression was associated with the thickness of the scar epidermis. In the less hypertrophic areas, involucrin was localized in the granular layer, as observed in the normal epidermis. However, when the epidermis exhibited abnormal thickness, the staining extended up to the spinous layer. TGF-1 expression Because TGF-1 is involved in both epidermis differentiation and the fibrotic process, we checked its expression in the scar epidermis. Interestingly, the immunohistochemical study revealed alterations of TGF-1 expression. In the normal epidermis, TGF-1 was restricted to the suprabasal layers, for which the staining was weak and fuzzy. By contrast, in the scar epidermis, TGF-1 expression was strong, with a punctiform staining, which was observed in the upper layers but also in the basal layer (Fig. 2). Abnormal keratinocyte proliferation: PCNA expression To assess whether the alterations of epidermis differentiation were associated with alterations in cell proliferation, we studied the expression of the proliferation marker, PCNA. Significant differences in PCNA expression between normal and scar epidermis were revealed by immunoperoxidase staining (Fig. 2). In normal epidermis, PCNA expression was restricted to the basal layer where only the proliferative keratinocytes were stained. In the scar epidermis, the staining was heterogeneous. Depending on the area, the expression of PCNA was restricted to the basal layer or extended up to the spinous layer or up to the horny layer. The differential expression of PCNA between hyperplastic
and normal epidermis was confirmed by Western blot analysis (Fig. 3). PCNA protein expression was at least 4-fold stronger in the scar epidermis than in the normal one. These results show that scar epidermis exhibits cell alterations at both the differentiation and the proliferation levels.
DISCUSSION Because skin fibrosis is mainly a mesenchymal disease, little is known about the concomitant epidermal alterations and the possible role of keratinocytes in the fibrotic process. Benyahia and Magdelena (4) studied superficial skin fibrosis resulting from therapeutic irradiation and described molecular alterations in both the dermis and the epidermis. They showed that the scar epidermis exhibited an increased expression of epidermal growth factor receptor and TGF-1 and suggested altered interactions between the epidermal and dermal skin compartments. More recently, we described a fibronecrotic lesion of the skin, resulting from accidental ␥-irradiation and showed that the scar epidermis exhibited a hyperplastic phenotype, as well as an abnormal expression of tumor necrosis factor-␣ and TGF-1 (8). In the present work, we characterized the proliferation and differentiation status of the epidermis of irradiated patients with specific markers. We demonstrated that the scar epidermis exhibits a hyperplastic morphology and an increased expression of PCNA, both of which reveal cell hyperproliferation. Furthermore, we described an abnormal differentiation characterized by the expression of K6 and K16, a general disorganization of the cytokeratins and involucrin, and a paradoxical overexpression of TGF-1. In the literature, histologic studies of superficial radiation-induced fibrosis described both hyperplastic (8) and atrophic (17) epidermis, depending on the stage of the fibrotic development. In our study, no epidermis atrophy
392
I. J. Radiation Oncology
● Biology ● Physics
was found. Interestingly, hyperplasia was observed in all patients, regardless of the length of time after irradiation (7 months to 27 years) or the type of irradiation (accidental or therapeutic). We propose that the observed increase of the epidermis thickness was due to the active status of the lesions, which were all characterized by strong inflammation, and the proliferation of myofibroblasts within the fibrotic tissue, suggesting strong interactions between dermal and epidermal cells. Conversely, atrophic epidermis is probably associated with quiescent phases, when inflammation is reduced and cell density decreases in the fibrotic tissue. To go further into the understanding of the proliferation status of scar epidermis, we investigated PCNA expression and found abnormal expression of PCNA from the basal to the suprabasal layers that was surprisingly accompanied by a marked expression of TGF-1 in the whole scar epidermis. TGF-1 has been described as a strong inhibitor of epithelial cell proliferation (18). In normal epidermis, TGF-1 is absent from the basal layer where cells proliferate. Its expression in the suprabasal cells blocks the proliferation of keratinocytes, which then differentiate. In this study, the expression of TGF-1 in the basal layer was inconsistent with the hyperproliferative phenotype described for the scar epidermis. Our results could be explained by a loss of keratinocyte response to TGF-1 through alterations of the signal transduction pathway, such as has been described during transformation of keratinocytes in carcinogenesis (19). Our group recently described such a process for myofibroblasts from a porcine model of radiation fibrosis (20). The TGF-1 signal transduction pathway is currently under investigation in our group to assess this hypothesis for scar keratinocytes. TGF-1 is involved in normal processes such as epidermis maturation, but it also participates in both the initiation and the development of fibrosis. In human fibrotic lesions, induced by either RT or accidental irradiation, overexpression of TGF-1 protein has been described in situ in both dermis and epidermis and in vitro in cultured fibroblasts and keratinocytes (4, 8). Moreover, several studies have focused on its role in the dermis where it regulates myofibroblast proliferation and matrix component expression (14). We propose that scar keratinocytes are involved in the fibrotic process by interacting with the dermis cells, because TGF-1 expressed by keratinocytes could specifically target myofibroblasts in a paracrine way and then take part in the initiation and development of fibrosis. The results of the present study demonstrate the hyperproliferative status of the scar epidermis, but the differentiation also appeared disorganized. Typical expression of keratin K6 and K16 was found in the scar epidermis and thus demonstrated the activated phenotype of keratinocytes undergoing altered differentiation (21). To investigate this abnormal differentiation, we compared keratin, involucrin, and integrin expression in scar and normal epidermis. We found that the scar epidermis exhibited an altered K14 and
Volume 53, Number 2, 2002
involucrin expression but that K10 and 1-integrin expression remained normal. These features are typical of abnormal differentiation and could be compared with those found in several pathologic epidermis disorders. Expression of K6 and K16 in scar epidermis from fibrosis resembles that observed in psoriasis (22), squamous carcinoma (23), and hypertrophic scar (24). The scar epidermis also exhibited altered expression of K14 and involucrin, similar to that observed in psoriasis. The promoters of K5 and K14 have been shown to be activated by TGF-1 (25). Because TGF-1 was overexpressed in the scar epidermis, it could be implicated in the abnormal expression of K14 in the suprabasal layers. By contrast, 1-integrin expression remained the same in normal and scar epidermis, confined to the basal layer, but in psoriasis or normal wound healing, its expression would abnormally extend to the suprabasal layers (26). Moreover, in psoriasis, the increase of K16 is associated with a decrease in K1/K10 expression. We found no variation in K10 expression in the scar epidermis, similar to that found in squamous carcinoma (22) and hypertrophic scar (23). In conclusion, scar epidermis from fibrosis shares several, but not all, features of epidermal diseases, thus exhibiting a specific abnormal pattern of protein expression. Surprisingly, the abnormal differentiation observed in the scar epidermis from late radiation fibrosis exhibited large similarities with the alterations described in epidermis that has just been irradiated. Liu et al. (27) showed that after 1–3 weeks of daily X-ray irradiation, the leg epidermis of mice exhibited altered expression of K6, K10, K14, involucrin, and 1-integrin, similar to the expression we found in delayed radiation sequelae. It seems that, whatever the time after irradiation, the epidermis develops the same activated keratinocyte phenotype even if the cellular microenvironment is different. This phenomenon could be the consequence of irreversible alterations of the keratinocytes after irradiation, which could be reactivated even several years after treatment. However, it could also be related to inflammation, which is a common feature of both early and late responses of the skin to irradiation. The development of skin fibrosis after ionizing irradiation is the consequence of complex processes involving several cell types, cell– cell interactions, and cytokines, which have been extensively studied in the dermis. By contrast, the epidermis was poorly characterized at the cellular and molecular levels. The results of our work show that the epidermis overlaying superficial RT-induced fibrosis is strongly disorganized and presents characteristic features of abnormal proliferation and differentiation. These results open new areas for fibrosis investigation. Moreover, because major progress has recently been obtained in the treatment of superficial radiation-induced fibrosis (28 –30), we propose the keratinocyte as a new potential therapeutic target to downregulate the fibrotic process.
Altered proliferation and differentiation in skin fibrosis after RT
● V. SIVAN et al.
393
REFERENCES 1. Archambeau JO, Pezner R, Wasserman T. Pathophysiology of irradiated skin and breast. Int J Radiat Oncol Biol Phys 1995;31:1171–1185. 2. Rodemann HP, Bamberg M. Cellular basis of radiation-induced fibrosis. Radiother Oncol 1995;35:83–90. 3. Hill RP, Rodemann HP, Hendry JH, et al. Normal tissue radiobiology: From the laboratory to the clinic. Int J Radiat Oncol Biol Phys 2001;49:353–365. 4. Benyahia B, Magdelena H. Immunohistochemical characterization of human gamma-irradiated skin. Bull Cancer Radiother 1993;80:126 –134. 5. Warren L, Garner MD. Epidermal regulation of dermal fibroblast activity. Plast Reconstr Surg 1998;102:135–139. 6. Rubin JS, Bottaro DP, Chedid M, et al. Keratinocyte growth factor as a cytokine that mediates mesenchymal-epithelial interaction. EXS 1995;74:191–214. 7. Moulin V. Growth factors in skin wound healing. Eur J Cell Biol 1995;68:1–7. 8. Vozenin-Brotons MC, Gault N, Sivan V, et al. Histopathological and cellular studies of a case of cutaneous radiation syndrome after accidental chronic exposure to a cesium source. Radiat Res 1999;152:332–337. 9. Fuchs E. Keratins and the skin. Annu Rev Cell Dev Biol 1995;11:123–153. 10. Furukawa F, Imamura S, Fujita M, et al. Immunohistochemical localization of proliferating cell nuclear antigen/cyclin in human skin. Arch Dermatol Res 1992;284:86 –91. 11. McGowan K, Coulombe PA. Subcellular biochemistry, intermediate filaments. New York: Plenum Press; 1998. 12. Watt FM, Jones PH. Expression and function of the keratinocyte integrins. Dev Suppl 1993;49:185–192. 13. Fuchs E, Byrne C. The epidermis: Rising to the surface. Curr Opin Genet Dev 1994;4:725–736. 14. Martin M, Lefaix J, Delanian S. TGF-1 and radiation fibrosis: A master switch and a specific therapeutic target? Int J Radiat Oncol Biol Phys 2000;47:277–290. 15. Martin M, Vozenin MC, Gault N, et al. Coactivation of AP-1 activity and TGF-beta1 gene expression in the stress response of normal skin cells to ionizing radiation. Oncogene 1997;15: 981–989. 16. Gottlober P, Bezold G, Weber L, et al. The radiation accident in Georgia: Clinical appearance and diagnosis of cutaneous radiation syndrome. J Am Acad Dermatol 2000;42:453– 458. 17. Hopewell JW. The skin: Its structure and response to ionizing radiation. Int J Radiat Biol 1990;57:751–773.
18. Moses HL, Yang EY, Pietenpol JA. TGF-beta stimulation and inhibition of cell proliferation: New mechanistic insights. Cell 1990;63:245–247. 19. Reiss M. Transforming growth factor-beta and cancer: A love-hate relationship? Oncol Res 1997;9:447– 457. 20. Reisdorf P, Lawrence D, Sivan V, et al. Alteration of transforming growth factor-1 response involves down-regulation of Smad3 signaling in myofibroblasts from skin fibrosis. Am J Pathol 2001;159:263–272. 21. Coulombe PA. Towards a molecular definition of keratinocyte activation after acute injury to stratified epithelia. Biochem Biophys Res Commun 1997;236:231–238. 22. Leigh IM, Navsaria H, Purkis PE, et al. Keratins (K16 and K17) as markers of keratinocyte hyperproliferation in psoriasis in vivo and in vitro. Br J Dermatol 1995;133:501–511. 23. Stoler A, Kopan R, Duvic M, et al. Use of monospecific antisera and cRNA probes to localize the major changes in keratin expression during normal and abnormal epidermal differentiation. J Cell Biol 1988;107:427– 446. 24. Machesney M, Tidman N, Waseem A, et al. Activated keratinocytes in the epidermis of hypertrophic scars. Am J Pathol 1998;152:1133–1141. 25. Jiang CK, Tomic-Canic M, Lucas DJ, et al. TGF beta promotes the basal phenotype of epidermal keratinocytes: Transcriptional induction of K5 and K14 keratin genes. Growth Factors 1995;12:87–97. 26. Hertle MD, Kubler MD, Leigh IM, et al. Aberrant integrin expression during epidermal wound healing and in psoriatic epidermis. J Clin Invest 1992;89:1892–1901. 27. Liu K, Kasper M, Trott KR. Changes in keratinocyte differentiation during accelerated repopulation of the irradiated mouse epidermis. Int J Radiat Oncol Biol Phys 1996;69:763– 769. 28. Lefaix JL, Delanian S, Leplat JJ, et al. Successful treatment of radiation-induced fibrosis using Cu/Zn-SOD and Mn-SOD: An experimental study. Int J Radiat Oncol Biol Phys 1996; 35:305–312. 29. Vozenin-Brotons MC, Sivan V, Gault N, et al. Antifibrotic action of Cu/Zn SOD is mediated by TGF-beta1 repression and phenotypic reversion of myofibroblasts. Free Radic Biol Med 2001;30:30 – 42. 30. Delanian S, Balla-Mekias S, Lefaix JL. Striking regression of chronic radiotherapy damage in a clinical trial of combined pentoxifylline and tocopherol. J Clin Oncol 1999;17:3283– 3290.