0022-202X/85/8506-0569$02.00/0 THE JOUIINAL OF INVEST IGATIVE 0EHMATOLOGY, 85:569- 572, 1985 Copyright © 1985 by The Williams & Wilkins Co.

Vol. 85, No. 6 Printed in U.S.A.

Expression of HLA-DR by Anagen Hair Follicles in Alopecia Areata

ANDREW G_ MESSENGER, M_B_, M.RC.P . AND STANLEY 8. BLEEHEN, M.R, F .RC_P.

University Department of Dermatology, Royal Hallamshire Hospital, Sheffield, U.K.

The expression of HLA-DR within hair follicles in alopecia areata was studied using an immunoperoxidase method_ Scalp biopsies were taken from 12 patients with alopecia areata and from 6 normal control subjects. Frozen sections were stained with a panel of 4 anti­HLA-DR monoclonal antibodies, Leu 2, Leu 3, Leu 4, and T6 antibodies. The expression of DR in normal hair follicles and in most anagen follicles from nonlesional alopecia skin was confined to dendritic cells which were sparse below the level of the arrector pilorum insertion. Of the 37 anagen follicles examined in lesional skin, 25 displayed staining for DR on epithelial cells in the pre­cortical matrix and presumptive cortex. Six follicles showed DR staining in other epithelial compartments, the lower bulb matrix, inner root sheath, and outer root sheath. Infiltration of the hair bulb matrix by T cells -w-as seen in the majority of follicles where epithelial cells were DR+. The aberrant expression of DR antigens by hair follicle epithelium provides direct evidence that immune mechanisms are operating in the pathogenesis of alopecia areata. In a previous study of alopecia areata -w-e found evidence of cell injury confined to the precor­tical matrix and presumptive cortex in lesional anagen follicles. The relative restriction of epithelial DR expression to the same site suggests that this region of the follicle is of fundamental importance in the disease process.

The cell membrane molecules encoded by the major histo­compatibility complex (MHC) play an important role in regu­lating immune responses [1]. Class I MHC molecules (HLA­A,B,C) are expressed by the majority of nucleated cells. The expression of class II molecules (which include HLA-DR) is restricted to certain cell types: these include B lymphocytes, dendritic cells such as Langerhans cells, activated T cells, macrophages, and some endothelial cells. In the skin DR is also expressed by acrosyringial epithelium (2]. However, it is now well documented that other cell types will exhibit DR expres­sion in a variety of disease states. Epidermal keratinocytes, which are normally DR-, have been reported to express DR in graft-versus-host disease [3], lichen planus [ 4], contact allergic dermatitis (5], and discoid lupus erythematosus (6]. This phe­nomenon has also been observed in other tissues including thyroid epithelium in Graves' disease (7], and pancreatic islet cells in type I diabetes [8] . Immune-mediated reactions are known to be involved or are implicated in the pathogenesis of these disorders and all are characterized histologically by lym­phocytic infiltration.

Alopecia areata is widely believed to be an autoimmune disease. However, much of the evidence for this is circumstan­tial [9]. The only direct evidence that the hair follicle lesion is

Manuscript received J anuary 24, 1985; accepted for publication July 11, 1985.

Reprint requests to: Dr. A. G. Messenge r, University Department of Dermatology, Royal Ha llamshire Hospital, Glossop Road, Sheffield SlO 2JF, U.K.

Abbreviations: MHC: major histocompatibili ty complex PBS: phosphate-buffered saline

immune-mediated is: (i) histologically there is a peribulbar and intrabulbar lymphoid infiltrate comprised mainly of T cells (10]; (ii) a report that autoantibodies reactive with the endo­thelium of perifollicular capillaries could be eluted from circu­lating lymphocytes of patients with alopecia areata [11]. In view of the reports of aberrant DR expression in immune­mediated disorders, and in particular in autoimmune disease, we have studied alopecia areata to see whether a similar phe­nomenon occurs within the hair follicle.

PATIENTS AND METHODS

Alopecia. Areata

Scalp biopsies were taken from 12 patients with alopecia areata. Nine were from lesional areas (3 bald patches, 2 alopecia totalis, 4 hypopigmented ha ir regrowth) and 3 from clinically nonlesional areas (1 uninvolved scalp, 2 areas of pigmented hair regrowth ).

Controls

Normal scalp skin was obtained from 6 subjects during t he excision of benign tumors.

Monoclonal Antibodies

The following monoclona l antibodies were used: anti-HLA-DR (Bec­ton Dickinson) DA6.147, DA6.164, and DA6.231, (provided by Dr Keith Guy, MRC Clinical a nd Population Cytogenetics Unit, Edinburgh) . DA6 antibodies recognize "fra mework " antigens on the HLA-DR mol­ecule. DA6.231 probably also recogn izes DC and SB locus products. Leu 2 (T suppressor/cytotoxic) , Leu 3 (T helper/ inducer), and Leu 4 (pan-T cell) were purchased from Becton Dickinson and T6 from Coulter.

The following dilut ions were used: Anti-HLA-DR 1:50; T6 1:25; Leu 2, Leu 3, and Leu 4 1:200. The DA6 antibodies were used undiluted.

Immu.noperoxidase Staining

The biopsies were snap-frozen and 6-J.Lm cryostat sections were mounted on poly-L-lysine coated slides. After fixation in acetone t he sections were incubated sequentially with the primary ant ibody, bio­tinylated ant i mouse IgG, and an avidin -biotin-peroxidase complex. The peroxidase reaction was developed with diaminobenzidine 0.05%, H20 2 0.01% in 0.05 M Tris buffer pH 7.6. The sections were light ly counter­stained, dehydrated, and rnounted. Phosphate-buffered salme (PBS) was used for all the dilut ions and washes between the staining proce­dures. Negative con t rols were carried out on all t he biopsies by replacing t he primary ant ibody with PBS.

RESULTS Controls

Twenty-five anagen follicles were studied in the 6 biopsies from normal control subjects. Within hair follicle epithelium positive staining with anti-HLA-DR, DA6.147 and DA6.164 was confined to dendritic cells. These were very sparse below the level of the arrector pilorum insertion. A similar staining pattern was seen with the T6 antibody. In addition to staining dendritic cells, DA6.231 also produced weak cell membrane staining of isthmus outer root sheath epithelium in 3 of the 6 biopsies. Endothelial cells in perifollicular capillaries were weakly DR+ but the capillary loop within the dermal papilla did not stain.

Alopecia A reata

569

Seventeen anagen follicles were ident ified in the 3 biopsies taken from clinically nonlesional areas of patients with alopecia

. ·~--

570 MESSENGER AND BLEEHEN

areata. Most showed a staining pattern for DR similar to that seen in normal follicles. In addition, some anagen follicles showed a light DR+ peribulbar lymphoid infiltrate and a few DR+ cells within t he dermal papilla. Isolated DR+ cells were seen within t he hair bulb matrix in 2 follicles but there was no apparent staining of follicular epithelium.

T he 9 biopsies from clinically lesional scalp showed the usual histopathologic features of alopecia areata, i.e., an increase in t he proportion of catagen and telogen follicles, a perifollicular lymphoid infilt rate, and anagen follicles in early stages of development. Thirty-seven anagen follicles, in which all com­ponents were visualized (dermal papilla, matrix, inner and outer root sheaths, and cortex) were studied with the DR and T6 antibodies. Concurrent staining with the T-cell antibodies was carried out on 15 anagen follicles from 7 lesional biopsies. The small size of many of the anagen follicles in these biopsies precluded staining with the complete antibody panel in all cases.

The 4 DR antibodies gave identical staining patterns. The results are summarized in Table I. Staining for DR was seen in t he dermal papilla in 33 follicles. In all but one case this was diffusely distributed throughout the papilla, precluding precise identification of its cellular origin. In 25 follicles, staining in t he dermal papilla was accompanied by staining for DR on matrix cells around the upper half of the dermal papilla and on epithelial cells within the presumptive cortex (Fig 1) . This

·often extended throughout the presumptive cortex and was occasionally seen in the keratinized hair shaft. Both cell mem­brane and cytoplasmic sta ining patterns were seen. This dis-

TABLE L Alopecia areata les ional shin: distribution of DR expression by follicular epithelium

Numerof Number of follicles showing DR

staining of ep ithe lial cells Biopsy anagen

folli cles Matrix studjed PC iRS ORS

L u 1 7 2 6 6 2 0 2 7 0 3 3 0 0 3 2 0 1 1 0 1 4 2 0 2 2 0 0 5 4 3 2 2 3 0 6 3 0 1 1 0 0 7 5 0 3 3 0 0 8 5 0 5 5 0 0 9 2 0 2 2 0 0

Total 37 5 25 25 5 1

U = upper; L = lower; PC = presumptive cortex; IRS = inner root s heath; ORS = outer root sheath.

Vol. 85, No.6

tribution of epithelial DR expression was seen in some or all of the anagen follicles in all the biopsy specimens from lesional areas. In the majority of follicles its restriction to this region of the hair follicle was very striking, even when staining was intense (Figs 2, 3). In 2 biopsies a total of 5 anagen follicles showed staining of matrix cells in the lower hair bulb and of the inner root sheath epithelium. Cell membrane DR staining of outer root sheath epithelium was also seen in several follicles in 1 biopsy but hair bulb structures were examined in only 2 of these follicles .

The peribulbar lymphoid infiltrate was strongly DR+. Leu 4+ cells comprised between 50-70% of cells in the infiltrate. All of the DR+ dermal papillae studied were infiltrated by Leu 4+ cells. Analysis ofT-cell subsets in the peribulbar and papillary infiltrates revealed a Leu 3:Leu 2 ratio consistently between 3:1 and 4:1. Invasion of the outer root sheath by T cells in similar subset proportions was a frequent finding in the pres­ence of a peribulbar infiltrate, but in only 1 biopsy was t his associated with DR staining of outer root sheath cells. Analysis

r

. ' ,,

..

FIG 2. Alopecia areata: anagen foJ.Iicle stained for HLA-DR. There is intense staining of the suprapapilla r matrix a nd presumptive cortex. There is a lso diffuse sta ining in t he dermal papilla. The lower bulb matrix, inner root sheath, and outer root sheath a re DR-. Staining of keratinized inner root sheath is with toluidine blue. Bar= 100 I'm·

FIG 3. Alopecia areata: a nagen follicle sta ined for HLA-DR. There is positive staining restricted to the presumptive cortex. Bar= 100 I'm·

FIG 1. Alopecia areata: transverse sections t hrough anagen follicle stained for HLA-DR. The levels of section are s hown on t he accompanying p hotomicrograph. a, Section through lower bulb. Mat rix cells are DR- . b, Section t hrough upper bulb. The dermal papilla is demarcated by the dotted line. There is diffuse s taining in t he dermal papilla. Matrix ce lls around the derma l papilla a re DR+. Peripherally located ce lls (which will form the inner root sheath) are DR- . c, Section at supra bulbar level. Centrally located cells of the presumpt ive cortex a re DR+. The inner root sheath is DR- . Bar = 40 11m.

Dec. 1985 ALOPECIA AREATA: HLA-DR EXPRESSION IN HAIR FOLLICLES 571

of T -cell subsets was carried out on 10 anagen follicles showing DR staining of upper bulb matrix and presumptive cortical epithelium. Leu 2+ and Leu 3+ cells were found infiltrating matrix epithelium in 8 of these, although their distribution was not identical to that of epithelial DR expression. The number of T cells within the matrix was too low to permit accurate quantification of subset ratios ( < 6/section) although the pro­portion of Leu 2+ cells appeared to be greater than in the peri bulbar and papillary infiltrates. A T -cell matrix infiltrate was also seen in 2 of the 5 anagen follicles where epithelial cells were DR-. We were unable to correlate epithelial DR expression with the predominance of any particular T-cell phenotype.

In 2 biopsies 5-10% of cells within the peribulbar infiltrates stained with T6. A single anagen follicle from each of these biopsies showed an increase in the number of intrabulbar T6+ cells.

DISCUSSION

Matrix cells in the lower part of the hair bulb mainly give rise to the inner root sheath while matrix cells around the upper half of the dermal papilla predominantly go to form the cortex (precortical matrix) [12,13). The presumptive cortex contains cells undergoing early differentiation which are des­tined to form the hair cortex. Indications of cortical differen­tiation include: (i) the accumulation of wispy cytoplasmic ker­atin filaments [14]; (ii) the ability to bind antibodies raised against hair keratin [15] ; (iii) pigment transfer. Although there is no distinct anatomic division between the matrix and the presumptive cortex, these features can be seen within 2-3 cell layers above the dermal papilla and at a lower level in periph­erally located cells.

In a histologic study of alopecia areata, Thies reported de­generative changes in the suprapapillary matrix in anagen follicles [16]. At electron microscopy, we found that lesional anagen follicles in alopecia areata showed evidence of nonspe­cific cell injury in the precortical matrix region and the pre­sumptive cortex [17] . Cell injury was not seen in the lower bulb matrix nor in other differentiating compartments such as t he inner root sheath. The expression of HLA-DR by follicular epithelium was also largely restricted to the precortical matrix and presumptive cortex. This appeared to be a marker of active disease as DR expression was not seen in follicles from nonle­sional areas nor in follicles from control subjects. However, alopecia areata causes an alteration in the dynamics of the hair cycle and we cannot entirely exclude the possibility that DR expression occurs as a constitutive phenomenon at certain stages in the hair cycle not represented in our control biopsies.

In keeping with previous studies [10) the peribulbar and papillary infiltrates were predominantly T cell in nature. The expression of DR by almost all cells in the infiltrate indicates that T cells are activated [18). Soluble lymphokines released by activated T cells, notably gamma interferon, are potent inducers of DR expression in a variety of cell types [19-21], including epidermal keratinocytes [22) . T cells were present within the dermal papilla in all the follicles studied showing epithelial DR expression and, in,the majority, T cells were al$0 found infiltrating the matrix. Thus DR expression by follicular epithelium in alopecia areata may be secondary to the release of inductive lymphokines generated during a T cell-mediated immune reaction. An alternative explanation for aberrant DR expression was suggested by Bottazzo eta! following the obser­vation of DR+ thyrocytes in Graves' disease [23]. They proposed that DR expression by epithelial cells could be induced by a variety of nonspecific stimuli, such as viral infections, in ge­netically predisposed individuals. Under certain circumstances this would allow the presentation of autoantigens, normally secluded from immune surveillance, to helper T cells and give rise to an autoimmune response. The same group have since reported that DR+ thyrocytes are able to act as antigen-pre-

senting cells providing antigen processing is not required (24]. The restriction, in many cases, of DR expression to the cortical region of t he hair bulb is difficult to explain simply in terms of a secondary response to soluble factors unless these cells are more sensitive to the inductive effect than other cells. The hypothesis that increased DR expression is a primary event in t he evolution of an autoimmune response would be compatible with the relative restriction of DR expression and cell injury to the same site within the hair bulb in alopecia areata. Against this, however, is the absence of DR expression in follicles from nonlesional areas, particularly in follicles that have previously been disease-affected, i.e., regrowing pigmented hair follicles.

Whether DR expression by follicular epithelium in alopecia areata is a primary event in the pathogenesis or merely an epiphenomenon of uncertain functional significance, two con­clusions can be drawn from this study. Firstly, this finding provides direct evidence that immune mechanisms are operat­ing in the pathogenesis of the hair follicle lesion. Secondly, the finding that maximal DR expression is concentrated in the same epithelial compartment as cell injury raises the possibility that the precortical matrix/presumptive cortex forms the pri­mary target for the disease process within t he anagen hair follicle.

We are grateful to Dr R. E. Church and Dr C. I. Harrington for allowing us to study patients under their care, and to Dr N. Rooney for supplying some of the antibodies.

REFERENCES 1. Dausset J: The major histocompatibility complex in man. Past,

present and future concepts. Science 213:1469_-1474, 1981 2. Murphy GF, Shepard RS, Harrist T J .. Bronsteu~ BR, Bhan. AK:

Ultrastructural documentation of HLA-DR antigen reactivity 111 normal human acrosyringial epithelium. J Invest Dermatol 81 :181- 183, 1983

3. Lampert !A, Suitters AJ Chisholm PM. Expression of Ia antigen on epidermal keratino~ytes in graft-versus-host disease. Nature 293:149-150, 1981

4. Tjernlund UM: !a-like antigens in lichen planus. Acta Derm Ve-nereal (Stockh) 60:309-315, 1980 . . .

5. MacKie RM, Turbitt ML: Quantitation of dendntic cells 111 nor~al and abnormal human epidermis using monoclonal antibodies directed against Ia and HTA antigens. J Invest. Dermatol 81:216-220, 1983

6. Lampert !A: Expression ofHLA-DR (!a-like) antigen on epidermal keratinocytes in human dermatoses. Clin Exp Immunol 57:93-100, 1984

7. Hanafusa T, Chiovato L, Doniach D, Pujol-Borrell R, ~ussell RCG, Bottazzo GF: Aberrant expression of HLA·D~ ant.Ig_en on thy­rocytes in Graves' disease: relevance for aut01mmumty. Lancet 2:1111-1115, 1983 .

8. Bottazzo G, Dean BM: Evidence of the expressiOn of class II (HLA­DR) and increased presentation of class I (HLA-A,B,C) mole­cules in pancreatic islets in type I (insultn-dependent) diabetes. Diabetologia 27:259A, 1984 .

9. Editorial. Alopecia areata-an autoimmune disease? Lancet 1:1335- 1336, 1984 . .

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

11. Nunzi E Hamerlinck F Cormane RH: Immunopathological studies on alo'pecia areata. Arch Dermatol Res 269:1-11, 1980

12. Kligman AM: The human hair cycle. J Invest Dermatol 33:307-31~ 1959 . .

13. Epstein WL, Maibach HI: Cell proliferatiOn and movement m human hair bulbs, Advances in Biology of Skm, vol IX, Hair Growth. Edited by W Montagna, RL Dobson. New York, Per­gamon Press, 1967, pp 83-97

14. Birbeck MSC Mercer EH: The electron microscopy of the human hair follicl~. I. Introduction and the hair cortex. J Biophysic Biochem Cytol 3:203-213, 1957

15. Cotton DWK, Kirkham N, Young BJJ: Immunoperoxidase anti­keratin staining of epidermal and pilar cysts. Br J Dermatol 111:63-68, 1984

16. Thies W: Vergleichende histologische Untersuchungen bei Alopecia areata und narbig-atrophisierenden Alopecien. Arch Klin Exp Dermatol 227:541-549, 1966

17. Messenger AG, Bleehen SS: Alopecia areata: light and electron microscopic pathology of the regrowing white hair. Br J Derma to! 110:155-162, 1984

572 MESSENGER AND BLEEHEN

18. Evans RL, Faldetta TJ, Humphreys RE, Pratt DM, Yunis EJ, Schlossman SF: Peripheral human T cells sensitized in mixed leukocyt"! culture synthesize and express !a-like antigens. J Exp Med 148:1440- 1445, 1978

19. Steeg PS, Moore RN, Oppenheim JJ : Regulation of murine mac­rophage !a-antigen expression by products of activated spleen cells. J Exp Med 152:1734- 1744, 1980

20. King DP, Jones PP: Induction of Ia and H -2 antigens on a mac­rophage cell line by immune interferon. J Immunol131:315- 318, 1983

21. Basham TY, Merigan TC: Recombinant interferon-')' increases

Vol. 85, No.6

HLA-DR synthesis and expression. J Immunol 130:1492- 1494, 1983

22. Basham TY, Nickoloff BJ, Merigan TC, Morhenn VB: Recombi­nant gamma interferon induces HLA-DR expression on cultured human keratinocytes. J Invest Dermatol 83:88-90, 1984

23. Bottazzo GF, Pujol-Borrell R, Hanafusa T, Feldmann M: Role of aberrant HLA-DR expression and antigen presentation in in­duction of endocrine autoimmunity. Lancet 2:1115-1119, 1983

24 . Londei M, Lamb JR, Bottazzo GF, Feldmann M: Epithelial cells expressing aberrant MHC class II determinants can present antigen to cloned human T cells. Nature 312:639-641, 1984