Normalisation of hair follicle morphology in C3H/HeJ alopecia areata mice after treatment with squaric acid dibutylester

ARTICLE

Alopecia areata is a non-scarring, reversible disease of the hair follicle ranging from patchy hair loss to the total loss of scalp hair (alopecia totalis). In some patients all body hair is lost (alopecia universalis) [1]. Although the cause of alopecia areata is unknown, an autoimmune attack on anagen hair follicles is suspected, as suggested by the presence of autoantibodies specifically directed to the hair follicle [2-6]. However, the most common diagnostic feature seen in human alopecia areata biopsies is a pronounced CD4⁺/CD8⁺ T lymphocyte infiltrate within and around dystrophic anagen hair follicles [7, 8].

The C3H/HeJ mouse model for alopecia areata exhibits clinical, morphological, and immunological features similar to the human disease [9]. These mice also exhibit a very similar anti-hair follicles antibody response to that observed in human alopecia areata [10]. Hair loss in this model is also associated with peri- and intrafollicular inflammatory cell infiltrates consisting largely of CD8⁺ and CD4⁺ T lymphocytes, and other leukocytes including granulocytes and macrophages [9, 11, 12]. A direct role for T lymphocytes in alopecia areata has recently been supported by the observation that alopecia areata-derived human T cells, primed by exposure to hair follicle antigens, were capable of inducing hair loss in alopecic scalp grafted onto severe combined immunodeficient (Prkdc^scid/Prkdc^scid) mice [13].

Current treatments for alopecia areata are not thought to change or reverse the ultimate course of the disease [14]. Therapeutic results can also be difficult to interpret because of spontaneous remissions and relapses. Topical immunotherapy with contact sensitisers is considered the most effective therapeutic modality [14, 15]. The contact sensitisers used are all effective, but their properties are wide-ranging. Dinitrochlorobenzene is mutagenic in the Ames test [16] and is thus no longer used. Squaric acid dibutylester (SADBE) or diphencyprone are not mutagenic and further advantages include their infrequent use in industry, absence in the environment, and the fact that they do not cross react with other contact allergens [17]. Induction of hair regrowth in alopecia areata by SABDE has been noted by many authors with varying rates of response [18-21]. We have recently described, immuno-phenotypically, hair regrowth in C3H/HeJ mice with alopecia areata-like hair loss after contact sensitization with SADBE [22]. Here we extend these findings, using high resolution light microscopy (HRLM) and transmission electron microscopy (TEM) with accompanying histomorphometric analysis, to assess specifically and in depth the morphological status of the hair follicles and its associated immune cell infiltrate after treatment with SADBE.

Material and methods

Animals

This study included eleven C3H/HeJ mice (10 female, 1 male) with alopecia areata-like hair loss (SADBE-treated animals), a female C3H/HeJ mouse with normal hair growth (negative control) and 3 female alopecic, untreated, C3H/HeJ mice (positive controls). Of the 11 experimental mice, 3 had patchy alopecia areata and 8 had extensive/diffuse alopecia totalis involving the entire dorsal and ventral surfaces.

Treatment

Following contact sensitisation with 2% SADBE on a 1 cm² area, the left side of the dorsal skin of all 11 experimental mice was treated weekly with topical applications of 0.1-1.0% SADBE dissolved in acetone. The right side of the dorsal skin remained untreated and served as a control, in order to avoid misinterpretation of a possible spontaneous remission. Mice were individually housed and the unilateral distribution of the resultant eczema indicated that the mice did not lick off the SADBE or transfer it to the opposite side mechanically. The concentration of the contact sensitiser was chosen individually until a moderately severe contact dermatitis [22] was induced in all mice. Clinical details are shown in Table I. Treatment was continued until unilateral hair regrowth was observed, but in no case longer than 22 weeks.

Tissue

Full thickness skin specimens (0.5 cm²) of mid-dorsal skin were taken symmetrically from both the treated (with full hair regrowth) and untreated sides of each experimental mouse and from the unaffected control mouse. Skin was also taken from the mid-ventral surface in the alopecic untreated mice. All tissue was immersed immediately in half strength Karnovsky's fixative [23], and processed for conventional transmission electron microscopy (TEM) and high resolution light microscopy (HRLM) as previously described [24]. For HRLM, semi-thin sections (0.5-1 mum) were stained with toluidine blue in di-sodium tetraborate buffer, examined by light microscopy, and photographed. Representative sections (longitudinal and horizontal) were cut from 3-6 blocks from each experimental (treated and untreated sides) and control mouse; total number of blocks examined: 39 untreated and 35 treated experimental mouse skin specimens, and 3 negative and 3 positive control skin ­ total of 80 blocks.

Morphometric analysis of SADBE-treated and untreated skin

Skin of C3H/HeJ mice was analysed by HRLM for a range of tissue parameters including; epidermal thickness, dystrophic anagen, follicular inflammatory cell infiltrates, mast cells in the dermis, and stage in hair growth cycle. These parameters were assessed and quantified in representative 2 mm-long, full-thickness tissue sections from all mice using the x 4 objective for complete sectional overview (i.e. x 40 at viewer's eyepiece) and using the x 25 objective for full contiguous counting of mast cells (i.e. x 250 at viewer's eyepiece). Stage of hair growth cycle is optimally assessed only if all hair follicles are cut longitudinally so that the full structure of the most proximal hair follicle is apparent. However, in the absence of consistent optimally-orientated hair follicles (due in part to disease-associated dystrophy), hair follicle cycle stage was assessed additionally as follows; in addition to commonly accepted hair growth cycle-specific hair follicle morphologies and epidermal thickness [25], the cycle stage of hair follicles lacking observable bulbs in semi-thin sections was assessed on the basis of how deep the hair follicles extended into the dermis. Those extending deep into the subcutaneous fat layer were determined to be in anagen/early catagen while telogen follicles were located high above the subcutaneous fat layer. Vellus hair follicles, also located high in the dermis, were identified morphologically and excluded from analysis. Parameter counts were analysed using the nonparametric Wilcoxon Signed Ranks Test given that values were not normally distributed.

Transmission Electron Microscopy

Ultrathin sections (~100 nm) were cut and counter-stained with 2% aqueous uranyl acetate and Reynolds' lead citrate [24]. Grids were examined from 3 blocks per mouse (selected from optimally oriented HRLM sections) from both experimental and control tissues (total of 57 blocks) and photographed using the JOEL 100 CX (Tokyo, Japan) electron microscope.

Results

The histopathology and ultrastructure of untreated C3H/HeJ alopecia areata skin exhibited characteristic features of alopecic skin including; perifollicular leukocyte infiltrates, consisting of lymphocytes, granulocytes, and macrophages. Moderate anagen dystrophy and melanin clumping were also observed. Unaffected tissue exhibited normal hair follicle structure.

Induction of hair regrowth in SADBE-treated alopecia areata mice

Clinical details of each SADBE-treated mouse are summarised in Table I. Two mice were excluded from the histomorphometric analyses; one that failed to regrow hair and exhibited no histopathological change in hair follicles in the treated skin despite the induction of a severe eczematous reaction with subsequent scarring and another that re-grew hair bilaterally. The treatment of alopecic C3H/HeJ mice with the contact sensitiser SADBE resulted in hair re-growth in 9 of 11 mice studied, with unilateral hair regrowth on the treated side 6-10 weeks after the start of treatment. Contact dermatitis was not induced on vehicle-treated skin [22] and correlated with no hair regrowth or evidence of anagen induction, i.e. hair follicle melanogenesis.

Morphological/morphometric analysis of hair follicles in mice responding to SADBE treatment

Hair follicle dystrophy was reduced after SADBE treatment

Dystrophic anagen hair follicles were common in untreated skin and often presented as highly kinked structures (Fig. 1, Table II). The targets of dystrophy were primarily the hair follicle inner root sheath and hair shaft. The frequency of dystrophic anagen hair follicles was significantly decreased (p < 0.01) in treated skin compared to untreated skin (Table II). Moreover, cytoplasmic vacuolation, apparently due to organelle degeneration, was commonly observed in the Huxley's layer of the inner root sheath and in the outer root sheath of untreated skin. The frequency of vacuolation was also reduced after treatment.

The perifollicular and intrafollicular infiltrates were reduced in SADBE-treated skin

Untreated skin was characterised by a marked perifollicular infiltrate consisting of mononuclear and polymorphonuclear leukocytes (PMNs) (Fig. 2a). In this study, PMNs are defined as white blood cells with multi-lobed nuclei and cytoplasmic granules and include neutrophils and eosinophils. Inflammatory cells were most commonly observed from the supra-bulbar to infundibular regions of the hair follicle and less commonly within and around the proximal hair follicle bulb. Granulocytes were identified as either eosinophils or neutrophils on the basis of their characteristic granules (Fig. 2b) while other leukocytes exhibited morphological features of macrophages and lymphocytes. Infiltration of mononuclear cells into the hair follicle was also evident (Fig. 2c). This follicular infiltrate was much reduced around anagen hair follicles in SADBE-treated skin (p < 0.05) and, in some cases, almost totally absent along the entire length of the hair follicle (Fig. 2d and Table II).

PMNs in perifollicular blood vessels were more frequently observed in untreated skin than in treated skin, indicative of leukocyte trafficking to alopecic skin. In many cases sectioned profiles of perifollicular blood vessels contained many more granulocytes than erythrocytes (Fig. 2e). Notably, in addition to their intravascular distribution, leukocytes were also marginated [i.e. attached to the wall of blood vessels] and present in the tissue immediately surrounding such vessels. By contrast, perifollicular blood vessels in SADBE-treated alopecia areata skin contained fewer PMNs.

Mast cells were distributed in low numbers throughout the dermis and were associated with the pilosebaceous unit in untreated skin. By contrast, a significant increase (p < 0.01) in mast cell numbers was observed after SADBE treatment, particularly in the upper dermis (Fig. 2f), compared with untreated skin (Table II). PMNs were interspersed with the numerous mast cells in this region.

Normalisation of the hair follicle pigmentary unit after SADBE-induced hair regrowth

Pigmentary abnormalities detected in untreated skin included; melanin clumping (clustering of multiple pigment granules) (Fig. 3a) and melanin incontinence (pigment observed in dermal papilla or ectopically in the perifollicular dermis) (Fig. 3b). Clumping was largely restricted to the hair bulb matrix around the apex of the dermal papilla, but some melanin clumps were also observed within the dermal papilla itself. SADBE-induced regrowth was associated with a marked reduction in these features, although residual effects were detected in some cases. In treated skin, melanin granules were occasionally observed within macrophages in the papillary dermis, either scattered or in association with mast cells. In addition, some pigmentary changes were also detected in the epidermis. While the epidermis of untreated and control skin was invariably amelanotic, melanin granules and occasional active melanocytes were observed in the epidermis of SADBE-treated skin (Figs. 4a and 4b). In one treated mouse, regrowing hair included some that was unpigmented (data not shown).

Hair follicles in full anagen were more common after SADBE treatment

The number of full anagen hair follicles, with hair bulbs located in the subcutaneous fat, was significantly higher (p < 0.05) in the treated skin when compared with untreated skin (Table II). This correlated with increased thickening of the intrafollicular epidermis, due to increased cell layers ­ a feature of anagen skin. Entry of telogen hair follicles into anagen or recovery of dystrophic anagen hair follicles was further evidenced by the resumption of cell proliferation in hair follicle bulbs as determined by the presence of cells in mitosis (data not shown).

Discussion

This study further explores the effects of SADBE treatment on hair follicle structure associated with successful initiation of hair regrowth in the C3H/HeJ alopecia areata mouse model. The therapeutic effect of SADBE derives from its induction of an allergic contact dermatitis and an apparent normalisation of hair follicle status by positively affecting the immunological milieu [15]. Anagen hair follicles are primarily targeted in alopecia areata [9, 26]. Thus, it is likely that relevant antigens are expressed only during this phase of the hair growth cycle [4] and injury may precipitate these hair follicles prematurely into catagen and telogen [26]. Cytotoxic damage by the peri-/intrafollicular infiltrate may adversely affect the differentiation of hair follicle cells, especially the anagen-specific inner root sheath keratinocytes that comprise the bulk of matrix-derived keratinocytes that mould the emerging hair fiber.

In this study we have identified that these infiltrating immune cells include granulocytes and macrophages in addition to lymphocytes. Some dendritic cells have also been observed in the skin of these mice [11]. A CD4⁺/CD8⁺ T cell ratio of 1:3 has been reported [9] and this decreases to 1:1 after CD8⁺ T cell numbers are reduced following SADBE therapy [22]. While the perifollicular infiltrate in the treated skin was reduced overall, some eosinophils remained. It is possible that these are induced by the allergic contact dermatitis on the treated side [27]. However, Elston et al. suggested eosinophils may be diagnostic for alopecia areata after they found these cells in 38 of 71 human cases of alopecia areata [28]. Eosinophils were rare in negative control, non-alopecic mouse skin, whereas these cells were observed in untreated skin of mice with alopecia areata and so may not be fully related to treatment. Eosinophils are unlikely to be involved directly in alopecia areata given that engraftment of alopecia areata skin onto Prkdc^scid/Prkdc^scid mice resulted in hair regrowth despite perifollicular granulocytic infiltrates [12]. The significant increase in the number of mast cells in the upper dermis is likely to be associated with the SADBE-induced allergic dermatitis.

One of the more striking, enigmatic, features of human alopecia areata is its associated pigmentary anomalies including; preferential targeting of pigmented hair and relative sparing of white hair, initial regrowth of unpigmented hair, melanocyte degeneration and melanin incontinence and clumping [29-31]. Regrowth of unpigmented hair was also a finding in this study, although this was restricted to a minority of mice and is likely to be a sequela to injury. Melanin incontinence and clumping observed in this study are likely to be related to melanocyte degeneration and death [31]. An interesting finding of this study was the observation of melanin granules and rare melanocytes in the epidermis of the treated C3H/HeJ skin. As the untreated alopecic skin was invariably amelanotic, this observation suggests that SADBE therapy may activate dormant melanocytes already present in the epidermis or induce the migration of follicular melanocytes. Mouse pelage epidermis is considered to be devoid of active melanocytes [32], except in mice having specific mutations such as adrenocortical dysplasia (acd) [33]. The allergic contact dermatitis induced by SADBE may have stimulated melanocytes from the upper outer root sheath to migrate into the epidermis or caused their differentiation in situ. Regardless of the mechanism involved, the presence of melanin granules within epidermal keratinocytes in the SADBE-treated skin indicate a functional epidermal-melanin unit [34].

Chronic, stable alopecia areata totalis in humans may be associated with lower numbers of anagen hair follicles and a much reduced peri/intrafollicular leukocyte infiltrate. This suggests that reduction of the infiltrate itself may not be sufficient to "release" hair follicles back into full anagen again. Many authors have found an increased telogen: anagen ratio in human alopecia areata skin [26, 35]. Although the numbers of pilar units may remain unchanged in alopecia areata, telogen germinal units do indeed appear to be significantly increased in both active and stationary phases of alopecia areata [36]. While we cannot assume the same for mice since hair density and cycles are very different, the present study similarly demonstrated that the number of hair follicles in anagen was significantly greater (p < 0.05) in SADBE-treated skin compared to contra-lateral untreated skin.

The mechanism behind the ability of SADBE to induce hair regrowth is unknown. The hypothetical phenomenon of "antigenic competition" has been proposed, whereby an immune reaction to a given antigen (e.g. in epidermis) may inhibit the development of the immune response to another unrelated antigen (e.g. in hair follicle) [37]. Other, more likely, explanations include a tolerization of alopecia areata-specific T cells and the induction of a non-permissive milieu for these cells [15, 38] or induction by SADBE treatment alters trafficking/homing of leukocytes in the affected skin. It has recently been proposed that contact sensitisers alter the alopecia areata-associated cytokine profile by the induction of an antagonistic cytokine profile and that this non-specifically mediates a beneficial effect on alopecia areata [15, 38].

This study supports the use of potent contact allergens such as SADBE, for the treatment of alopecia areata and provides histological evidence for the apparent normalisation of hair follicles that correlated with a good clinical response. The similarity of SADBE-induced hair growth response in C3H/HeJ alopecia areata mice and human alopecia areata further validates the use of this animal model in the study of therapeutic interventions in alopecia areata.

Article accepted on 16/6/00

CONCLUSION

Acknowledgements

This study was supported by the Department of Biomedical Sciences, University of Bradford Ph.D. Studentship Fund (DJT), by grants from the Deutsche Forschungsgemeinschaft (HO 1598/1-3 to RH), National Alopecia Areata Foundation (to JPS) and National Institute of Health (AR 43801 to JPS). We would also like to acknowledge the expert technical assistance of Ms. Kathleen A. Silva at the Jackson Laboratory, Bar Harbor, Maine, USA, and Ms. Stephanie Metz at the Department of Dermatology, Philipp University, Marburg, Germany.

REFERENCES

1. Gollnick H, Orfanos CE. Alopecia areata: Pathological and clinical picture. In: Hair and Hair Diseases (Orfanos CE, Happle R, eds). Berlin: Springer-Verlag, 1990; 529-69.

2. McElwee KJ, Tobin DJ, Bystryn JC, King LE Jr, Sundberg JP. Alopecia areata: an autoimmune disease? Exp Dermatol 1999; 8: 371-9.

3. Tobin DJ, Orentreich N, Bystryn JC. Autoantibodies to hair follicles in normal individuals. Arch Dermatol 1994; 130: 395-6.

4. Tobin DJ, Orentreich N, Fenton DA, Bystryn JC. Antibodies to hair follicles in alopecia areata. J Invest Dermatol 1994; 102: 721-24.

5. Tobin DJ, Bystryn JC. Alopecia areata is associated with antibodies to hair follicle-specific antigens located predominantly in the proliferative region of hair follicles. In: Hair Research for the Next Millennium (Van Neste DJJ, Randall VA, eds). Amsterdam: Elsevier, 1996; 237-41.

6. Tobin DJ, Hann SK, Song MS, Bystryn JC. Hair follicle structures targeted by antibodies in patients with alopecia areata. Arch Dermatol 1997; 133: 57-61.

7. Perret C, Wiesner-Menzel L, Happle R. Immunohistochemical analysis of T-cell subsets in the peribulbar and intrabulbar infiltrates of alopecia areata. Acta Derm Venereol (Stockh) 1984; 64: 26-30.

8. Ranki A, Kianto U, Kanerva L, Tolvanen E, Johansson E. Immunohistochemical and electron microscopic characterisation of the cellular infiltrate in alopecia (areata, totalis, universalis). J Invest Dermatol 1984; 83: 7-11.

9. Tobin DJ, Sundberg JP, King LE Jr., Boggess D, Bystryn JC. Antibodies to hair follicles in C3H/HeJ mice with alopecia areata-like hair loss. J Invest Dermatol 1997; 109: 329-33.

10. Sundberg JP, Cordy WR, King LE Jr. Alopecia areata in aging C3H/HeJ mice. J Invest Dermatol 1994; 102: 847-56.

11. McElwee KJ, Boggess D, Olivry T, Oliver RF, Whiting D, Tobin DJ, Bystryn JC, King LE Jr, Sundberg JP. Comparison of alopecia areata in human and non human mammalian species. Pathobiol 1998; 407: 1-26.

12. McElwee KJ, Boggess D, King LE Jr, Sundberg JP. Experimental induction of alopecia areata-like hair loss in C3H/HeJ mice using full-thickness skin grafts. J Invest Dermatol 1998; 111: 797-803.

13. Gilhar A, Ullmann Y, Berkutzli T, Assy B, Kalish RS. Autoimmune hair loss (alopecia areata) transferred by T lymphocytes to human scalp explants on SCID mice. J Clin Invest 1998; 101: 62-7.

14. Shapiro J, Price V. Hair regrowth. Therapeutic agents. Dermatol Clin 1998; 16: 341-56.

15. Hoffmann R, Happle R. Topical immunotherapy in alopecia areata: What, How and Why? Dermatol Clin 1996; 14: 739-44.

16. Summer KH, Göggelmann W. 1-chloro-2,4-dinitrobenzene depletes glutathione in rat skin and is mutagenic in Salmonella typhimurium. Mutat Res 1980; 77: 91-3.

17. Mitchell AJ, Krull EA. Alopecia areata, pathogenesis and treatment. J Am Acad Dermatol 1984; 11: 763-75.

18. Happle R, Kalveram KJ, Buchner U, Echternacht-Happle K, Göggelmann W, Summer KH. Contact allergy as a therapeutic tool for alopecia areata: application of squaric acid dibutylester. Dermatologica 1980a; 161: 289-97.

19. Case PC, Mitchell AJ, Swanson NA, van der Veen EE, Ellis CN, Headington JT. Topical therapy of alopecia areata with squaric acid dibutylester. J Am Acad Dermatol 1984; 3: 447-51.

20. Caserio RJ. Treatment of alopecia areata with squaric acid dibutylester. Arch Dermatol 1987; 123: 1036-41.

21. Micali G, Cicero RL, Nasca MR, Sapuppo A. Treatment of alopecia areata with squaric acid dibutylester. Int J Dermatol 1996; 35: 52-6.

22. Freyschmidt-Paul P, Sundberg JP, Happle R, McElwee KJ, Metz S, Boggess D, Hoffmann R. Successful treatment of alopecia areata-like hair loss with the contact sensitizer squaric acid dibutylester (SADBE) in C3H/HeJ mice. J Invest Dermatol 1999; 113: 61-8.

23. Karnovsky MJ. Formaldehyde-glutaraldehyde fixative of high osmolarity for use in electron microscopy. J Cell Biol 1965; 27: 137-8.

24. Tobin DJ. Morphological analysis of hair follicles in alopecia areata. Micro Res & Tech 1997; 38: 443-51.

25. Handjiski B, Eichmüller S, Hofmann U, Czarnetzki BM, Paus R. Alkaline phosphatase activity and localisation during the murine hair cycle. Br J Dermatol 1994; 131: 303-10.

26. Messenger AG, Slater DN, Bleehen SS. Alopecia areata: alterations in the hair growth cycle and correlation with the follicular pathology. Br J Dermatol 1986; 114: 337-47.

27. Wardlow AJ, Moqubel R, Kay B. Eosinophils: Biology and role in disease. Adv Immunol 1995; 60: 151-66.

28. Elston DM, McCollough ML, Bergfeld WF, Liranzo MO, Heibel M. Eosinophils in fibrous tracts and near hair bulbs: A helpful diagnostic feature of alopecia areata. J Am Acad Dermatol 1997; 37: 101-6.

29. Messenger AG, Bleehen SS. Alopecia areata: light and electron microscopic pathology of the regrowing white hair. Br J Dermatol 1984; 110: 155-62.

30. Tobin DJ, Fenton DA, Kendall MD. Ultrastructural observations on the hair bulb melanocytes and melanosomes in acute alopecia areata. J Invest Dermatol 1990; 94: 803-7.

31. Tobin DJ, Fenton DA, Kendall MD. Cell degeneration in alopecia areata: An ultrastructural study. Am J Dermatopathol 1991; 13: 248-56.

32. Reynolds J. The epidermal melanocytes of mice. J Anat 1954; 88: 45-58.

33. Beamer WG, Sweet HO, Bronson RT, Shire JG, Orth DN, Davisson MT. Adrenocortical dysplasia: a mouse model system for adrenocortical insufficiency. Endocrinol 1994; 141: 33-43.

34. Fitzpatrick TB, Breathnach AS. Das epidermale Melanin-Inherit-System. Dermatol Wockenschr 1963; 147: 481-9.

35. Whiting DA. Histopathology of alopecia areata in horizontal sections of scalp biopsies. J Invest Dermatol 1995; 104: 26s-7.

36. Headington JT, Mitchell A, Swanson N. New histopathological findings in alopecia areata studied in transverse section. J Invest Dermatol 1981; 76: 325.

37. Happle R. Antigenic competition as a therapeutic concept for alopecia areata. Arch Dermatol Res 1980b; 267: 109-14.

38. Hoffmann R, Wenzel E, Huth A, van der Steen P, Schäufele M, Konig A, Happle R. Growth factor mRNA levels in alopecia areata before and after treatment with the contact allergen diphenylcyclopropenone. Acta Derm Venereol 1996; 76: 17-20.