Sequential immunohistochemical study of depigmenting and repigmenting minigrafts in vitiligo

ARTICLE

Auteur(s) : Marwa ABDALLAH*¹, Mohamed B. ABDEL-NASER*¹, Manal H. MOUSSA², Chalid ASSAF³ Constantin E. ORFANOS³

¹ Department of Dermatology and Venereology, Ain Shams University, Cairo, Egypt
² Department of Histology; Ain Shams University, Cairo, Egypt
³ Department of Dermatology, University Medical Center Benjamin Franklin; The Free University of Berlin, Berlin, Germany

* The first two authors contributed equally to the experiment presented in this paper
This work was conjointly performed at the Free University in Berlin, Germany, and Ain Shams University in Cairo, Egypt.

Article accepted on 11/8/03

Since skin colour is due to the melanin content of keratinocytes rather than of melanocytes, one may anticipate that depigmented patches are clinically manifested a few weeks after the melanocytic damage. By that time, most of the immunological phenomena would have disappeared; therefore, only minimal alterations can still be detected in vitiliginous tissue.

So far, examination of the very early events of melanocytic damage in vitiligo has not been possible. Surgical treatment by transferring viable melanocytes to the depigmented areas may represent a treatment option for selected cases, although Koebner's phenomenon is a limiting factor for surgical procedures in vitiligo [4]. Autologous minigrafting also provides an opportunity to examine the sequence of events of early melanocytic damage in patients not responding to the procedure, and to study the alterations of grafted melanocytes in vitiliginous tissue.
In this investigation we employed minigrafts and compared the sequence of events that occurred to grafted autologous melanocytes as a means to better understand the mechanisms leading to vitiligo.

Patients and methods

Patients

Six patients were included in the study, 4 males and 2 females, age range 24-37 years, Fitzpatrick's skin type IV-V, with stable vitiligo of the non-segmental widespread type. The experimental aim of the study was explained to the patients and their informed consent was taken.

Methods

Minigrafting was performed according to the technique described by Falabella [5] using 2 mm biopsy punches. At least 7 minigrafts were transplanted into each test area, 4 were taken sequentially, and the rest were left to observe the clinical outcome. One of the vitiliginous punches harvested on day 0 (= A) was used as preoperative control. The treated areas were bandaged for 2 weeks. Sequential punch biopsies (2.5 mm in diameter) were taken on days 14 (B), 17 (C), 21 (D) and 28 (E). All specimens were snap frozen in liquid nitrogen and stored at -70°C until used. The period of observation was preset, as the clinical outcome is usually evident by that time.

Immunohistochemical staining

Frozen sections (5 µm thick) were fixed over 30 minutes with acetone and were exposed to the primary monoclonal antibody (MoAb) (30 min) diluted in 0.1 M phosphate buffered saline (PBS). The primary MoAbs were directed against gp100 (HMB-45), CD4 (Leu 3a), CD8 (OKT 8), LFA-1 (IOT 16) and ICAM-1 (RR 1/1). They were diluted 1:50, except HMB-45 (1:150). With the exception of CD4, which was purchased from Novocastra (Newcastle, UK) and ICAM-1 from Serva (Heidelberg, Germany), all other MoAbs were purchased from DAKO (Glostrup, Denmark). As a secondary antibody, anti-mouse IgG (Immunotech, Marseille, France) diluted in 1:100 in PBS was used with the APAAP complex. The incubation steps with the secondary antibody and the APAAP complex were repeated twice in order to intensify the labelling reaction. Naphthol As-BI sodium salt (Sigma, St. Louis, USA) and new fuchsin (Merck, Darmstadt, Germany) were used for visualisation. Lastly, counterstaining with Mayer's hematoxylin solution (Merck) was done. Sections not exposed to the primary MoAbs served as negative controls. The slides were mounted with Kaiser's glycerol gelatine (Merck) and examined by light microscopy. Positive staining appears bright pink or red, whereas negative staining acquires the counterstain as a faint blue colour [6].

Evaluation of immunohistochemistry

Slides were first examined by light microscopy to evaluate positivity of staining and for general evaluation of the specimens. Using digital image analyser (Leica Q 500MC program, Leica Mikroskopie und Systeme GmbH, Wetzlar, Germany) cells were counted per mm skin. As melanocytes are located in the basal epidermal layer, examination was focused on the basal and suprabasal epidermis, dermo-epidermal junction (DEJ) and the papillary dermis. The number of cells stained for ICAM-1 was too high for accurate counting. Counting was done on control sections and on days 17 and 28 postoperatively, as they showed the most striking differences. The number of positive cells was counted for every specimen in at least 4 different mm. Mean and standard deviation (SD) of fields were calculated for responders and non-responders and a comparison was made between the two groups using Student's t-test. The progression of reaction was assessed within the same group of patients using one way ANOVA test; p values < 0.05 were considered significant.

Results

Clinical

Six patients completed the study (4 males and 2 females, mean age 31.1 ± 3.6 years). Patients treated with minigrafts were divided into two groups according to the outcome of the minigrafting procedure: In the responder group (n = 3) the patients showed evidence of increasing pigment spread, whereas, in the non-responder group (n = 3) evidence of early pigment loss was registered.

Melanocytes

Vitiligo specimens taken as control (A) showed negative staining with HMB-45, thus confirming the absence of epidermal melanocytes. In responders, the first specimen taken immediately after removal of the bandage on day 14 (B) showed negative staining. In specimen C (day 17) a few melanocytes were detected and their density gradually increased on day 21 to reach normal numbers on day 28 (18.8 ± 1.5 cells/mm vs. 16.7 ± 3.2 for normal skin, p > 0.05) (Table I). Cells were concentrated in the center of the specimens and tended to decrease towards the periphery. By comparing the cell count of responders using one way ANOVA test, there was a significant increase in melanocyte numbers after grafting (A vs. C, p < 0.001); the difference was found significant between days 0 and 17, (p < 0.001) and 17 and 28 (p < 0.01).

In non-responders, the staining for melanocytes was negative on days 14 and 17 in two cases, whereas, in the third case melanocytes with visible dendrites were found in specimen B (2 ± 1.4 cells/mm) with a consecutive decrease of their number. The reaction was weakly positive showing small dendrites on day 17 to fade completely on day 28. The cells seemed to disappear from the periphery towards the centre (Fig. 1). In addition to the melanocyte findings, epidermal keratinocytes showed marked vacuolisation in all non-responders (Fig. 1, 2b), whereas in responders this vacuolisation was either absent or minimal (Fig. 2c).

Intercellular adhesion molecule-1 (ICAM-1)

Preoperative samples of vitiligo taken as control (A), showed negative staining in the basal epidermal layer in all cases, only endothelial staining was present. Postoperative specimens (B-E) showed positivity in most cases in the basal cell layer involving both melanocytes and keratinocytes. The positivity was focal, more intense at the tips of the rete ridges and sometimes extending to the suprabasal layers. The intensity of the reaction was stronger in the non-responders (Fig. 2).

Leukocytes

In non-responders, there was a significant increase in LFA-1 positive cells after grafting (one way ANOVA p < 0.05) as compared to control specimen in vitiligo (data not shown). It was difficult to differentiate between perivascular and diffusely distributed cells in the papillary dermis. Moreover, a higher number of LFA-1 positive cells was present in non-responders compared to responders in specimens C and E (p < 0.05) (Table I, Fig. 3).
In responders, two cases showed low numbers of infiltrating leukocytes. Only a third patient had a relatively dense infiltrate in specimens C and E. There was no significant difference between pre- and postoperative readings (p > 0.05).

T-helper lymphocytes

Few CD4-positive T-lymphocytes were observed in perivascular and perifollicular areas. There was no significant difference between control vitiligo biopsies (A) and those taken sequentially after minigrafting (B-E). Also no significant difference was found in T-helper counts between responders and non-responders.

Cytotoxic/suppressor T-lymphocytes

Only a few CD8-positive T-lymphocytes were seen in control specimens in the epidermis and the subepidermal areas in all cases. After minigrafting, the non-responders showed a large number of infiltrating cells in the dermo-epidermal junction in specimen C and E, higher than in the controls. Cytotoxic/suppressor T-cells were detected focally where melanocytes reside. There, CD8 positive cells seemed to embrace melanocytes, since they were present at the immediate subepidermal area of the dermis and within the basal and suprabasal epidermal layers. ANOVA test showed a significant difference during the observation (p < 0.01), the difference being detected only between A, before and C, after grafting (p < 0.05).

In the responder group, two patients showed a small amount of infiltrating cells in specimen C (day 17, mean 5.0 and 8.34 cells/mm skin for either patient), while the third patient showed a more intense infiltrate (44.67 cells/mm), resulting in a high total mean for all 3 patients and a non-significant difference compared with the non-responders. However, the number of CD8 positive cells was significantly higher in non-responders compared with responders on day 28 after minigrafting (p < 0.05) (Table I).

Discussion

Autologous minigrafting is a common surgical procedure used for repigmenting stable vitiligo lesions [7, 8]. In the present investigation, three grafted patients showed good response and another three showed loss of pigmentation, i.e. loss of the grafted melanocytes. In responders, the first biopsy taken 14 days after minigrafting failed to show any melanocyte staining, however, melanocytes appeared a few days later and reached normal numbers on day 28. Matsumoto et al. [9] also found only a few DOPA-positive melanocytes 48 hours after grafting human skin to nude athymic mice, thereafter melanocyte numbers gradually increased. In contrast, total absence of melanocytes was found in all specimens of two non-responders. In the third non-responsive patient positive melanocyte staining was seen on day 14, which decreased gradually and disappeared completely two weeks later. Perilesional biopsies taken from active vitiligo patches also showed reduced melanocyte numbers to disappear completely in lesional areas [10, 11].
ICAM-1 staining of basal epidermis was evident in most postoperative biopsies, the intensity being focal and stronger in non-responders as compared to responders. Al-Badri et al. [12] and van den Wijngaard et al. [11] demonstrated the presence of ICAM-1 and HLA-DR expression in perilesional melanocytes around active vitiligo patches.
Interaction between ICAM-1 and its ligand LFA-1 is a necessary first step for contact-dependent immunologic reactions via leukocytes [13, 14]. Its expression on keratinocytes and melanocytes and its upregulation on endothelial cells are induced by several inflammatory cytokines such as IFN-γ, TNF-α, IL-1, IL-6, IL-7 [13, 15, 16]. More recently, Li et al. [17] reported that IgG antimelanocyte antibodies derived from patients with active vitiligo can induce ICAM-1, HLA-DR antigen expression and IL-8 release from cultured melanocytes. These cytokines could have been induced in varying levels in our patients by the inflammatory infiltrate and/or by the neighbouring keratinocytes in response to an antigen specific immune response.
LFA-1 expression was markedly upregulated both in density and in distribution after minigrafting with its density being significantly higher in non-responders than in responders. Previous studies have shown absence of B-lymphocytes from the infiltrate [11, 18] and, recently, scarcity of NK cells was reported [19]. LFA-1 could be attributable mainly to T-lymphocytes and to a lesser extent to monocytes, since the number of LFA-1 positive cells was higher. Le Poole [20] and van den Wijngaard [11] and their co-workers have demonstrated the presence of macrophages in perilesional vitiligo skin, where melanocytes were disappearing. They ascribed their presence to their role in melanocyte destruction, or to a clearing action after cytotoxic T-cell killing, since they expressed the phagocytosis-related markers CD36 and CD68.
CD4 positive lymphocytes were generally sparse before and after minigrafting showing no clear differences in numbers before and after the procedure or between responders and non-responders. Previous studies also showed a small CD4+ infiltrate compared to CD8+ cells both in melanoma-induced and spontaneous vitiligo [10, 11, 18-20]. Since T-helper lymphocytes activate other members of the immune system, the reason for their scarcity is not clearly understood.
In our investigations, CD cytotoxic T-lymphocytes constituted the major amount of the cellular infiltrate, being detected regularly in the regions where melanocytes reside. Their number in non-responders significantly outnumbered that of the responders on day 28. In perilesional skin of actively enlarging patches and in inflammatory vitiligo several previous studies also showed a CD8+ infiltrate. The infiltrating cells were focally detected and seemed to attack melanocytes both from the dermal and the epidermal sites, as evidenced by double immunostaining techniques [10, 11, 18, 20]. This may also provide an explanation for the marked vacuolisation seen in the epidermis of non-responders, since the inflammatory infiltrate leads to increased oxidative stress with accumulation of hydrogen peroxide within the cells [21].
In conclusion, our findings indicate that a cell-mediated immune response is triggered after autologous minigrafting in vitiligo, involving mainly CD8+ lymphocytes and inducing ICAM-1 expression in epidermal cells. This response is mild in responders where the grafted melanocytes survive, but is more severe in non-responders leading to their destruction. By analogy to studies conducted on perilesional vitiligo skin, our findings provide a better understanding of the early changes that occur in vitiligo indicating that an autoimmune reaction destroys the living melanocytes in vitiliginous patches in the early phase of the disease. n

References

1. Kovacs SO. Vitiligo. J Am Acad Dermatol 1998; 38: 647-66.

2. Erf GF, Trejo-Skilli AV, Smyth JR. T cells in regenerating feathers of Smyth line chicken with vitiligo. Clin Immunol Immunopathol 1995; 76: 120-6.

3. Abdel-Naser MB, Krueger-Krasagakes S, Krasagakis K. Gollnick H, Orfanos CE. Further evidence of involvement of both cell mediated and humoral immunity in generalized vitiligo. Pigm Cell Res 1994; 7: 1-8.

4. Falabella R, Armnategui A, Barona MI, Alzate A. The minigrafting test for vitiligo: detection of stable lesions for melanocyte transplantation. J Am Acad Dermatol 1995; 32: 228-32.

5. Falabella R. Repigmentation of stable leukoderma by autologous minigrafting. J Dermatol Surg Oncol 1986; 12: 172-9.

6. Cordell JL, Falini B, Erber WN, Gosh AK, Abdulaziz Z, Mc Donald S, Pulford KA, Stein H, Mason DY: Immunoenzymatic labeling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal antialkaline phosphatase (APAAP complex). J Histochem Cytochem 1984; 32: 219-29.

7. Falabella R. Repigmentation of segmental vitiligo by autologous minigrafting. J Am Acad Dermatol 1983; 9: 514-21.

8. Falabella R. Treatment of localized vitiligo by autologous minigrafting. Arch Dermatol 1988; 124: 1649-55.

9. Matsumoto K, Robb E, Warden G, Nordlund J. The expression of cytokines, growth factors and ICAM-1 in the healing of human cutaneous xenografts on nude mice. J Exp Dermatol 1997; 6: 13-21.

10. Abdel-Naser MB. Evaluation of T cell and T cell subpopulations and immunohistochemical studies in vitiligo with the use of monoclonal antibodies. MD Thesis. Ain Shams Univ 1992.

11. van den Wijngaard R, Wankowicz-Kalinska A, Le Poole C, Tigges B, Westerhof W, Das P. Local immune response in skin of generalized vitiligo patients. Destruction of melanocytes is associated with the prominent presence of CLA + T cells at the perilesional site. Lab Invest 2000; 80: 1299-309.

12. Al-Badri AMT, Foulis AK, Todd PM, Garioch JJ, Gudgeon JE, Stewart DG, Gracie JA, Goudie RB. Abnormal expression of MHC class II and ICAM-1 by melanocytes in vitiligo. J Pathol 1993a; 169: 203-6.

13. Norris D. Cytokine modulation of adhesion molecules in the regulation of immunologic cytotoxicity of epidermal targets. J Invest Dermatol 1990; 95: 111S-20S.

14. Yohn JJ, Critelli M, Lyons MB, Norris DA. Modulation of melanocyte intercellular adhesion molecule-1 by immune cytokines. J Invest Dermatol 1990; 95: 233-7.

15. Kirnbauer R, Charvat B, Schauer E, Kock A, Urbanski A, Forster E, Neuner P, Assmann I, Luger TA, Schwarz T. Modulation of intercellular adhesion molecule-1 expression on human melanocytes and melanoma cells: Evidence of a regulatory role of IL-6, IL-7, TNF-b and UVB light. J Jnvest Dermatol 1992; 98: 320-6.

16. Fabbri M, Bianchi E, Fumagalli L, Pardi R. Regulation of lymphocyte traffic by adhesion molecules. Inflamm Res 1999; 48: 239-46.

17. Li YL, Yu CL, Yu HS. IgG Anti-melanocyte antibodies purified from patients with active vitiligo induce HLA-DR and intercellular adhesion molecule-1 expression and an increase in interleukin-8 release by melanocytes. J Invest Dermatol 2000; 115: 969-73.

18. Al-Badri AMT, Todd PM, Garioch JJ, Gudgeon JE, Stewart DG, Goudie RB. An immunohistochmenical study of cutaneous lymphocytes in vitiligo. J Pathol 1993b; 170: 149-55.

19. Le Gal FA, Avril MF, Bosq J, Lefebvre P, Deschemin JC, Andrieu M, Dore MX, Guillet JG. Direct evidence to support the role of antigen-specific CD8 (+) T cells in melanoma-associated vitiligo. J Invest Dermatol 2001; 117: 1464-70.

20. Le Poole IC, van den Wijngaard RM, Westerhof W, Das PK. Presence of T cells and macrophages in inflammatory vitiligo skin parallels melanocyte disappearance. Am J Pathol 1996; 148: 1219-28.

21. Schallreuter KU, Moore J, Wood JM, Beazley WD, Gaze DC, Tobin DJ, Marshall HS, Panske A, Panzig E, Hibberts NA. In vivo and in vitro evidence for hydrogen peroxide (H2O2) accumulation in the epidermis of patients with vitiligo and its successful removal by a UVB-activated pseudocatalase J Investig Dermatol Symp Proc 1999; 4: 91-6.