British Journal of Dermatology (1990) 123, 717-724.
Expression of extracellular matrix in hair follicle mesenchyme in alopecia areata AJ.G.MCDONAGH, LISA CAWOOD AND A.G.MESSENGER Department of Dermatology, Royal Hallamshire Hospital, Giossop Road, Sheffield Sio 2JF, U.K. Accepted for publication 16 June 1990
SUMMARY
Human hair follicle mesenchyme contains a distinctive extracellular matrix which varies in volume and composition in concert with the hair growth cycle. T o investigate the possibility that mesenchymal function is disturbed in alopecia areata we have studied the expression of extracellular matrix constituents in scalp biopsies from 14 patients by immunohistochemical staining for basement membrane proteins, proteoglycans and interstitial collagens. The staining patterns in follicles from non-lesional scalp were normal. Miniature anagen follicles from bald patches also showed relatively normal expression of basement membrane proteins and proteoglycans. However, in some large anagen follicles from lesional sites, there was loss of the normal staining pattern for chondroitin-6-sulphate in the dermal papilla. In lesional eatagen follicles, the glassy membranes showed marked convolution and thickening. These structures stained strongly for laminin and type IV collagen but only weakly for interstitial collagens.
Histological and ultrastructural studies in alopecia areata have demonstrated injury to cells in the precortical matrix region of lesional anagen follicles.''^ These cells also frequently display aberrant expression of class I and class II antigens of the major histocompatibility complex (MHC).' * This appears to be a late event and may result from cytokine release by infiltrating immune cells.^ These observations led us to suggest that cells undergoing early cortical differentiation form the primary disease target within the hair follicle.' An alternative hypothesis is that epithelial changes in the hair follicle are a consequence of dysfunction of the dermal papilla. This is supported by the results of morphometric and autoradiographic studies.''^ The dynamics of the hair-growth cycle are disturbed in alopecia areata,*^ and it is known from experimental evidence in the rat that the dermal papilla is of crucial importance in controlling growth and differentiation of hair bulb epithelium."*"^" Relatively little is known of the molecular processes involved in signalling between the mesenchyme-derived dermal papilla and the epithelial matrix of the hair follicle. However, there is evidence to suggest that the extracellular matrix within the mesenchymal compartment Correspondence; Dr A.J.G.McDonagh. 717
7i8 A.J.G.McDonagh et al. plays a significant role in hair growth control. The extracellular matrix of the dermal papilla differs in content from that of the interfollicular dermis, being rich in basement membrane proteins and chondroitin proteoglycans. Moreover, in normal follicles, the volume and constitution of this extracellular matrix shows regular variation with the phase of the hairgrowth cycle. During anagen, the dermal papilla cells are separated by an abundant extracellular matrix. The volume of the extracellular matrix diminishes during catagen and is minimal during tclogen when the dermal papilla consists of a tightly packed ball of cells in close apposition with the secondary epithelial germ. Production of extracellular matrix resumes when the follicle reenters anagen. To investigate correlations of the pathophysiology of alopecia arcata with extracellular matrix expression, we performed an immunohistochemical study using a panel of antibodies against basement membrane proteins, proteoglycans and interstitial coUagens on biopsies from patients with various stages of the disease. METHODS
Materials Details of the antibodies used for immunohistochemieal staining are summarized in Table i. Biotinylated secondary antibodies and avidin-biotin-peroxidase complex were purchased from Vector Laboratories and chondroitin ABC lyase from ICN Biomedicals. The chondroitin and dermatan antibodies recognize epitopes revealed by prior digestion of glyeosaminoglycans with chondroitin ABC lyasc'' The antibody to dermatan sulphate also recognizes chondroitin-4sulphate. Tissue sections
Fourteen scalp biopsies were obtained with informed consent from patients with alopecia areata (Table 2). Normal scalp for control purposes was obtained from two patients during routine excision of benign tumours. Biopsy specimens were snap frozen, embedded in OCT medium and stored at — 70 C until required. Sections 6 /(m thick were cut on a eryostat, mounted on poly-1-lysine coated slides, air-dried and fixed in acetone. Immunohistochemistry
Tissue sections were washed in phosphate-buffered saline (PBS) and those for proteoglycan staining were pre-incubated at 37 C with chondroitin ABC lyase (02 U/ml) for 30 min. Serial
TABLE I. Antibodies used for immunohistochemical staining Specificity
Species
Type
Tvpe I collagen Type III collagen Type IV collagen Laminin Unsulphated chondroitin (1B5) Chondroitin-6-sulphatf (3B3) Dermatan sulphate (9A2)
Goat Goat Mouse Rabbit Mouse Mouse Mouse
Monoclonal Monoclonal Monoclonal Polyclonal Monoclonal Monockina! Monoclonal
Dilution : 100 : 100 : 100 : 100 : 100 : 100 : 100
Supplier Sera-Lab Ltd Sera-Lab Ltd ICN Biomedicals Unilever research ICN Biomedicals ICN Biomedicals ICN Biomedicals
Hair follicle extracellular matrix in A A
719
TABLE 2. Clinical features of patients included in tbe study Age 26 32 22
27 26 56 57 60 19 39 21
38 39 46
Sex
Duration
Biopsy Site
F F M M F F M M F M M F F F
4 years 12 years 10 months 8 months 4 years 51 years 9 months 18 months 13 years 3 years 20 years 6 months 9 months 15 years
Occipital Parietal Temporal Occipital Occipital Occipital Occipital Occipital Temporal Occipital Occipital Occipital Parietal Temporal
Type Non-lesional Non-lesional Pigmented regrowth Pigmented regrowth Edge of active patch Edge of active patcb Edge of active patch Edge of active patcb Bald area Bald area Non-pigmented regrowtb Non-pigmented regrowtb Non-pigmcnted regrowtb Non-pigmented regrowth
incubation was then performed with blocking serum, primary antibody, biotinylated secondary antibody and avidin-biotin-peroxidase complex. All incubations were for 30 min with three 5 min washes in PBS containing i",, bovine serum albumin between each stage. The peroxidasc reaction was developed with oos",, diaminobenzidine and hydrogen peroxide 0 01",, in 005 M Tris buffer at pH 76. The sections were lightly counterstained with haematoxylin, dehydrated and mounted. Controls included replacing the primary antibody with PBS or irrelevant antibodies and omission of chondroitin ABC lyase digestion where appropriate. RESULTS
Scalp from normal subjects, non-lesional alopecia areata and pigmented hair regrowth
Staining patterns in these biopsies were identical to those previously noted in normal scalp.'' In anagen follicles, there was staining throughout the dermal papilla with all the antibodies. The antibodies to collagen type I, collagen type III and dermatan sulphate also stained in the connective tissue sheath and throughout the interfollicular dermis. The antibodies to nonsulphated chondroitin and chondroitin-6-sulphate stained in the lower part of the connective tissue sheath, while those to type IV collagen and laminin stained the follicular basement membranes. During follicular regression in catagen, expression of proteoglycans declined with loss of non-sulphated chondroitin and chondroitin-6-sulphate preceding that of dermatan sulphate. By the telogen phase, the volume of matrix in the dermal papilla was greatly diminished. Expression of proteoglycan molecules was not apparent and that of basement membrane components was minimal at this stage. Typical patterns of staining for type I collagen, laminin and chondroitin-6-sulphate in anagen follicles are shown in Figure i. Margins of extending lesions
Biopsies from the margins of extending areas of alopecia areata contained a high proportion of catagen follicles. In one biopsy where surviving anagen follicles showed a heavy peribulbar
720
A.J.G.McDonagh et al.
I. Staining of non-lesional anagen hair bulbs, (a) Type I collagen showing staining throughout tbe dermal papilla and connective tissue sheath, (b) Laminin showing staining in tbe basement membranes oi the outer root sheath and blood vessels and tbraugbout tbe dermal papilla, (c) Chondroitin6-sulpbate showing staining of the matrix of the dermal papilla and connective tissue sheath. Magnification x 142.
lymphoid infiltrate, there was loss of staining for chondroitin-6-sulphate in the upper half of the dermal papilla and the lower connective tissue sheath. In some catagen follicles, a greatly thickened and convoluted glassy membrane was noted. The glassy membrane showed intense staining for collagen type IV and laminin but stained only weakly with the antibodies to type I and type III collagen (Figs 2a and b). Staining for chondroitin-6-sulphate was restricted to a small rim at the outer margin of the glassy membrane (Fig. 2c). In some follicles, remnants of the glassy membrane were left beneath the hair bulb as the level of this structure rose in the dermis with the approach of telogen (Fig. 2d). Staining for collagen types I, III and IV and laminin was maintained in the dermal papilla during catagen but staining for non-sulphated chondroitin and chondroitin-6-sulphate was greatly diminished. Bald areas and non-pigmented regrowth
In biopsies from established bald patches, numerous miniature anagen follicles were seen. The dermal papillae in these miniature follicles showed relatively normal patterns of staining with all the antibodies including chondroitin-6-sulphate (Fig. 3a). Some anagen follicles in biopsies from areas of non-pigmented regrowth showed patchy staining or absence of staining from the upper pole of the dermal papilla for chondroitin-6sulphate (Fig. 3b). Staining of dermal papillae in these follicles with other antibodies showed no gross abnormalities. However, some abnormalities of follicular anatomy were clearly demon-
Hair follicle extracellular matrix in AA
FIGURE 2. Stainiii.'.-ii K-i.mal catagen follicle, (a; Type IV atllagcn showing intense staining of tbe glassy membrane (thick arrow) and dermal papilla (thin arrow). (b)Type III collagen showing dense staining of the dermal connective tissues and dermal papilla while the glassy membrane (arrow) is virtually unstained, (c) Cbondroitin-6-sulphate showing staining confined to a narrow rim (thick arrow) at the periphery of tbe glassy membrane and very weak in the dermal papilla (thin arrow), (d) Cbondroitin-6sulphate staining of lesional late catagen. telogen follicle showing unstained remnants of the glassy membrane (thick arrow) beneath the dermal papilla (thin arrow). The papilla is only weakly positive. Magnification: x 198.
721
722
AJ.G.McDonagh et al. U
FIGURE 3. (a) Miniature anagen follicle from bald scalp showing normal expression of chondroitin-6sulphate in the dermal papilla. Magnification: x 198. (b) Lesional anagen follicle from area of nonpigmented regrowth showing absence from the upper pole of the dermal papilla of chondroitin-6sulphatc. The papilla is delineated by the dotted line. Magnification: x 320. (c) Lesional follicle from area of non-pigmented regrowth showing relatively large dermal papilla withfibroustracts extending into the dermis beneath (type IV collagen stain). Magnification: x 80.
Hair follicle extracellular matrix in A A
723
strated, including fibrous tracts streaming beneath follicles situated unusually high in the dermis (Fig. 3c). All the controls were negative. DISCUSSION
The expression of basement membrane proteins and protcoglycans during hair follicle development and in the hair-growth cycle has been studied in the rat. Laminin and type IV collagen are present throughout the hair cycle' ^' though they are most abundant in association with anagcn follicles. Expression of fibronectin, heparan-sulphate proteoglycan and chondroitin-sulphate proteoglycan all show correlation with the phase of the hair growth cycle.'^"'^ This is most marked for chondroitin-sulphate proteoglycan and it has been suggested that this basement membrane component may be involved in the dermo-epidermal interactions leading to formation and growth of skin appendages.'"^ Moreover, in chick embryo tissue culture experiments, disturbed proteoglycan metabolism has been found to result in morphogenetic abnormalities of feather follicles."' Expression of proteoglycans during the human hair cycle has recently been studied in normal scalp with antibody techniques similar to those we have used.'^ Similar findings of a cyclical pattern of expression of chondroitin proteoglycans have emerged; staining for non-sulphated chondroitin and chondroitin-6-suIphate declines in the dermal papilla and connective tissue sheath early in catagen and becomes undetectable by telogen. These changes appear to occur earlier than the decline in expression of other proteoglycans or basement membrane proteins. Involvement of proteoglycans in human hair-growth regulation had previously been inferred from observations in diseases such as pretibial myxoedema and the Hunter/Hurler syndrome in which disturbed proteoglycan metabolism is associated with hypertrichosis.' '"^ In newly involved areas of alopecia areata, anagen follicles are precipitated abruptly and prematurely into catagen. The unusual patterns of staining for chondroitin-6-sulphate in some anagen follicles from lesional sites may indicate impending catagen conversion. It was apparent that these changes were only in follicles where entry into catagen was likely to be imminent. However, loss of chondroitin-6-sulphate staining is not usually observed before the occurrence of early morphological changes of catagen. It is possible that the abnormal appearances in these lesional anagen follicles was due to degradation of extracellular matrix by infiltrating inflammatory cells. Nevertheless, we cannot exclude the possibility that these changes indicate primary dysfunction of the hair follicle mesenchymc with resulting premature entry of follicles into catagen. The grossly thickened and convoluted appearance of the glassy membrane in lesional catagen follicles has, to our knowledge, not been previously reported. We have not encountered similar appearances in catagen follicles from normal subjects. The role of the glassy membrane in the evolution of catagen is unknown and the significance of this observation is not clear. Nevertheless, these findings provide confirmation that the glassy membrane is composed of basement membrane proteins. Chondroitin-6-sulphate has been regarded as a basement membrane component' '^ and has recently been suggested to be important in the pathogenesis of dystrophic epidermolysis bullosa."" It is interesting, therefore, that staining for this proteoglycan did not co-distribute with laminin and type IV collagen in catagen follicles. In conclusion, the changes we have observed in alopecia areata may result from dynamic disturbances of the hair growth cycle. However, the loss of the normal staining pattern for chondroitin-6-sulphate in some lesional anagen follicles suggests that the changes which occur
724 A.J.G.McDonagh el al in the dermal papilla during normal catagen are occurring prematurely. This observation warrants further study and should be amenable to investigation by in vhro techniques. ACKNOWLEDGMENT
This work was supported by a research grant from the Special Trustees of the former United Sheffield Hospitals. REFERENCES 1 Thies W. Vergleichende hisiologische Untersuchungen bei Alopecia areata und narbig-atrophisierenden Alopecien. Arch Klin Exp Dermatol 1966; 227: 541-9. 2 Messenger AG, Bleehen SS. Alopecia areata: light and electron microscopic pathology of the regrowing white hair. Br J Dermatol 1984; I i o : 155-62. 3 Messenger AG, Bleehen SS. Expression of HLA-DR by anagen hair follicles in alopecia areata. J Invest Dermatol 1985; 85: 569-724 Briitker EB, Echternacht-Happle K, Hamm H, Happle R. Abnormal expression of class I and class II major histocompatibilily antigens in alopecia areata: modulation by topical immunotherapy. J Invest Dermatol 1987; 88: 564-8. 5 Khoury EL, Price VH, Greenspan JS. HLA-DR expression by hair follicle keratinocytes in alopecia areata: evidence that it is secondary lo the lymphoid infiltration. _7 Ifivest Dermatol 1988; 90: 193-200. 6 Van Scott EJ, Ekel T M . Geometric relationships between Che matrix of the hair bulb and its dermal papilla in normal and alopecie scalp. J Invest Dermatol 1958; 31: 281-77 Pierard GE, de la Brassinne M. Cellular activity in the dermis surrounding the hair bulb in alopecia areata. J Cutan Pathol 1975; 2: 240-5. 8 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: 337-479 Oliver RF. The induction of hair follicle formation in the adult hooded rat by vibrissa dermal papiUa^.J Embryol Exp Morphol 1970; 23: 219-36. 10 Oliver RF. The experimental induction of whisker growth in the hooded rat by impiantation of dermal papillae. J Embryol Exp Morphol 1967; 18; 43-51. 11 Couchman JR, CatersenB, Christner JB, Baker JR. Mappingby monoclonal antibody detection of glycosaminoglycans in connective tissues. Nature 1984; 307: 650-2. 12 Westgate GE, Lewis LP, Messenger AG, Gibson WT. Distribution of proteoglycans during the human hair cycle. 7 Invest Dermatol 1989; 92: 540A. 13 Couchman JR, Gibson WT. Expression of basement membrane components through morphological changes in the hair growth cycle. Dev Biol 1985; 108: 290-80. 14 Couchman JR, Rat hair foliicie dermal papillae have an extracellular matrix containing basement membrane components. J Invest Dermatol 1986; 87: 762-7. 15 Couchman JR, King JL, McCarthy KJ. Distribution of two basement membrane proteoglycans through hair follicle development and the hair growth cycle in the rat. J Invest Dermatol 1990; 94: 65-70. 16 Goetinck PF, Carlone DL. Altered proteoglycan synthesis disrupts feather pattern formation in chick embryonic skin. Dev Biol 1988; 127: 179-86. 17 Roberts SOB, Weismann K. The skin in systemic disease. In: Textbook of Dermatology (Rook A, Ebling FJG, Wilkinson DS et al., eds), 4th edn. Oxford: Blackwell Scientific Publications, 1986; 2343-74. 18 McKusick VA,Neufeld EF. The mucopolysaccharide storage diseases. In: The Metabolic Basis of Inherited Disease (Stanbury JB, Wyngaarden JB, Freilrickson DS el al., eds), 5th edn. New York: McGraw-Hill, 1983; 751-77, 19 Fine J-D, Couchman JR, Chondroitin-6-sulfate-containing proteoglycan: a new component of skin dermoepidermal junction, J Invest Dermatol 1988; 90; 283-8, 20 Fine J-D, Couchman JR, Chondroitin-6-sulfate proteogiycan but not heparan sulfate proteoglycan is abnormally expressed in skin basement membrane from patients with dominant and recessive dystrophic epidermolysis bullosa. J Invest Dermatol 1989; 92: 611-16,
Expression of extracellular matrix in hair follicle mesenchyme in alopecia areata AJ.G.MCDONAGH, LISA CAWOOD AND A.G.MESSENGER Department of Dermatology, Royal Hallamshire Hospital, Giossop Road, Sheffield Sio 2JF, U.K. Accepted for publication 16 June 1990
SUMMARY
Human hair follicle mesenchyme contains a distinctive extracellular matrix which varies in volume and composition in concert with the hair growth cycle. T o investigate the possibility that mesenchymal function is disturbed in alopecia areata we have studied the expression of extracellular matrix constituents in scalp biopsies from 14 patients by immunohistochemical staining for basement membrane proteins, proteoglycans and interstitial collagens. The staining patterns in follicles from non-lesional scalp were normal. Miniature anagen follicles from bald patches also showed relatively normal expression of basement membrane proteins and proteoglycans. However, in some large anagen follicles from lesional sites, there was loss of the normal staining pattern for chondroitin-6-sulphate in the dermal papilla. In lesional eatagen follicles, the glassy membranes showed marked convolution and thickening. These structures stained strongly for laminin and type IV collagen but only weakly for interstitial collagens.
Histological and ultrastructural studies in alopecia areata have demonstrated injury to cells in the precortical matrix region of lesional anagen follicles.''^ These cells also frequently display aberrant expression of class I and class II antigens of the major histocompatibility complex (MHC).' * This appears to be a late event and may result from cytokine release by infiltrating immune cells.^ These observations led us to suggest that cells undergoing early cortical differentiation form the primary disease target within the hair follicle.' An alternative hypothesis is that epithelial changes in the hair follicle are a consequence of dysfunction of the dermal papilla. This is supported by the results of morphometric and autoradiographic studies.''^ The dynamics of the hair-growth cycle are disturbed in alopecia areata,*^ and it is known from experimental evidence in the rat that the dermal papilla is of crucial importance in controlling growth and differentiation of hair bulb epithelium."*"^" Relatively little is known of the molecular processes involved in signalling between the mesenchyme-derived dermal papilla and the epithelial matrix of the hair follicle. However, there is evidence to suggest that the extracellular matrix within the mesenchymal compartment Correspondence; Dr A.J.G.McDonagh. 717
7i8 A.J.G.McDonagh et al. plays a significant role in hair growth control. The extracellular matrix of the dermal papilla differs in content from that of the interfollicular dermis, being rich in basement membrane proteins and chondroitin proteoglycans. Moreover, in normal follicles, the volume and constitution of this extracellular matrix shows regular variation with the phase of the hairgrowth cycle. During anagen, the dermal papilla cells are separated by an abundant extracellular matrix. The volume of the extracellular matrix diminishes during catagen and is minimal during tclogen when the dermal papilla consists of a tightly packed ball of cells in close apposition with the secondary epithelial germ. Production of extracellular matrix resumes when the follicle reenters anagen. To investigate correlations of the pathophysiology of alopecia arcata with extracellular matrix expression, we performed an immunohistochemical study using a panel of antibodies against basement membrane proteins, proteoglycans and interstitial coUagens on biopsies from patients with various stages of the disease. METHODS
Materials Details of the antibodies used for immunohistochemieal staining are summarized in Table i. Biotinylated secondary antibodies and avidin-biotin-peroxidase complex were purchased from Vector Laboratories and chondroitin ABC lyase from ICN Biomedicals. The chondroitin and dermatan antibodies recognize epitopes revealed by prior digestion of glyeosaminoglycans with chondroitin ABC lyasc'' The antibody to dermatan sulphate also recognizes chondroitin-4sulphate. Tissue sections
Fourteen scalp biopsies were obtained with informed consent from patients with alopecia areata (Table 2). Normal scalp for control purposes was obtained from two patients during routine excision of benign tumours. Biopsy specimens were snap frozen, embedded in OCT medium and stored at — 70 C until required. Sections 6 /(m thick were cut on a eryostat, mounted on poly-1-lysine coated slides, air-dried and fixed in acetone. Immunohistochemistry
Tissue sections were washed in phosphate-buffered saline (PBS) and those for proteoglycan staining were pre-incubated at 37 C with chondroitin ABC lyase (02 U/ml) for 30 min. Serial
TABLE I. Antibodies used for immunohistochemical staining Specificity
Species
Type
Tvpe I collagen Type III collagen Type IV collagen Laminin Unsulphated chondroitin (1B5) Chondroitin-6-sulphatf (3B3) Dermatan sulphate (9A2)
Goat Goat Mouse Rabbit Mouse Mouse Mouse
Monoclonal Monoclonal Monoclonal Polyclonal Monoclonal Monockina! Monoclonal
Dilution : 100 : 100 : 100 : 100 : 100 : 100 : 100
Supplier Sera-Lab Ltd Sera-Lab Ltd ICN Biomedicals Unilever research ICN Biomedicals ICN Biomedicals ICN Biomedicals
Hair follicle extracellular matrix in A A
719
TABLE 2. Clinical features of patients included in tbe study Age 26 32 22
27 26 56 57 60 19 39 21
38 39 46
Sex
Duration
Biopsy Site
F F M M F F M M F M M F F F
4 years 12 years 10 months 8 months 4 years 51 years 9 months 18 months 13 years 3 years 20 years 6 months 9 months 15 years
Occipital Parietal Temporal Occipital Occipital Occipital Occipital Occipital Temporal Occipital Occipital Occipital Parietal Temporal
Type Non-lesional Non-lesional Pigmented regrowth Pigmented regrowth Edge of active patch Edge of active patcb Edge of active patch Edge of active patcb Bald area Bald area Non-pigmented regrowtb Non-pigmented regrowtb Non-pigmcnted regrowtb Non-pigmented regrowth
incubation was then performed with blocking serum, primary antibody, biotinylated secondary antibody and avidin-biotin-peroxidase complex. All incubations were for 30 min with three 5 min washes in PBS containing i",, bovine serum albumin between each stage. The peroxidasc reaction was developed with oos",, diaminobenzidine and hydrogen peroxide 0 01",, in 005 M Tris buffer at pH 76. The sections were lightly counterstained with haematoxylin, dehydrated and mounted. Controls included replacing the primary antibody with PBS or irrelevant antibodies and omission of chondroitin ABC lyase digestion where appropriate. RESULTS
Scalp from normal subjects, non-lesional alopecia areata and pigmented hair regrowth
Staining patterns in these biopsies were identical to those previously noted in normal scalp.'' In anagen follicles, there was staining throughout the dermal papilla with all the antibodies. The antibodies to collagen type I, collagen type III and dermatan sulphate also stained in the connective tissue sheath and throughout the interfollicular dermis. The antibodies to nonsulphated chondroitin and chondroitin-6-sulphate stained in the lower part of the connective tissue sheath, while those to type IV collagen and laminin stained the follicular basement membranes. During follicular regression in catagen, expression of proteoglycans declined with loss of non-sulphated chondroitin and chondroitin-6-sulphate preceding that of dermatan sulphate. By the telogen phase, the volume of matrix in the dermal papilla was greatly diminished. Expression of proteoglycan molecules was not apparent and that of basement membrane components was minimal at this stage. Typical patterns of staining for type I collagen, laminin and chondroitin-6-sulphate in anagen follicles are shown in Figure i. Margins of extending lesions
Biopsies from the margins of extending areas of alopecia areata contained a high proportion of catagen follicles. In one biopsy where surviving anagen follicles showed a heavy peribulbar
720
A.J.G.McDonagh et al.
I. Staining of non-lesional anagen hair bulbs, (a) Type I collagen showing staining throughout tbe dermal papilla and connective tissue sheath, (b) Laminin showing staining in tbe basement membranes oi the outer root sheath and blood vessels and tbraugbout tbe dermal papilla, (c) Chondroitin6-sulpbate showing staining of the matrix of the dermal papilla and connective tissue sheath. Magnification x 142.
lymphoid infiltrate, there was loss of staining for chondroitin-6-sulphate in the upper half of the dermal papilla and the lower connective tissue sheath. In some catagen follicles, a greatly thickened and convoluted glassy membrane was noted. The glassy membrane showed intense staining for collagen type IV and laminin but stained only weakly with the antibodies to type I and type III collagen (Figs 2a and b). Staining for chondroitin-6-sulphate was restricted to a small rim at the outer margin of the glassy membrane (Fig. 2c). In some follicles, remnants of the glassy membrane were left beneath the hair bulb as the level of this structure rose in the dermis with the approach of telogen (Fig. 2d). Staining for collagen types I, III and IV and laminin was maintained in the dermal papilla during catagen but staining for non-sulphated chondroitin and chondroitin-6-sulphate was greatly diminished. Bald areas and non-pigmented regrowth
In biopsies from established bald patches, numerous miniature anagen follicles were seen. The dermal papillae in these miniature follicles showed relatively normal patterns of staining with all the antibodies including chondroitin-6-sulphate (Fig. 3a). Some anagen follicles in biopsies from areas of non-pigmented regrowth showed patchy staining or absence of staining from the upper pole of the dermal papilla for chondroitin-6sulphate (Fig. 3b). Staining of dermal papillae in these follicles with other antibodies showed no gross abnormalities. However, some abnormalities of follicular anatomy were clearly demon-
Hair follicle extracellular matrix in AA
FIGURE 2. Stainiii.'.-ii K-i.mal catagen follicle, (a; Type IV atllagcn showing intense staining of tbe glassy membrane (thick arrow) and dermal papilla (thin arrow). (b)Type III collagen showing dense staining of the dermal connective tissues and dermal papilla while the glassy membrane (arrow) is virtually unstained, (c) Cbondroitin-6-sulphate showing staining confined to a narrow rim (thick arrow) at the periphery of tbe glassy membrane and very weak in the dermal papilla (thin arrow), (d) Cbondroitin-6sulphate staining of lesional late catagen. telogen follicle showing unstained remnants of the glassy membrane (thick arrow) beneath the dermal papilla (thin arrow). The papilla is only weakly positive. Magnification: x 198.
721
722
AJ.G.McDonagh et al. U
FIGURE 3. (a) Miniature anagen follicle from bald scalp showing normal expression of chondroitin-6sulphate in the dermal papilla. Magnification: x 198. (b) Lesional anagen follicle from area of nonpigmented regrowth showing absence from the upper pole of the dermal papilla of chondroitin-6sulphatc. The papilla is delineated by the dotted line. Magnification: x 320. (c) Lesional follicle from area of non-pigmented regrowth showing relatively large dermal papilla withfibroustracts extending into the dermis beneath (type IV collagen stain). Magnification: x 80.
Hair follicle extracellular matrix in A A
723
strated, including fibrous tracts streaming beneath follicles situated unusually high in the dermis (Fig. 3c). All the controls were negative. DISCUSSION
The expression of basement membrane proteins and protcoglycans during hair follicle development and in the hair-growth cycle has been studied in the rat. Laminin and type IV collagen are present throughout the hair cycle' ^' though they are most abundant in association with anagcn follicles. Expression of fibronectin, heparan-sulphate proteoglycan and chondroitin-sulphate proteoglycan all show correlation with the phase of the hair growth cycle.'^"'^ This is most marked for chondroitin-sulphate proteoglycan and it has been suggested that this basement membrane component may be involved in the dermo-epidermal interactions leading to formation and growth of skin appendages.'"^ Moreover, in chick embryo tissue culture experiments, disturbed proteoglycan metabolism has been found to result in morphogenetic abnormalities of feather follicles."' Expression of proteoglycans during the human hair cycle has recently been studied in normal scalp with antibody techniques similar to those we have used.'^ Similar findings of a cyclical pattern of expression of chondroitin proteoglycans have emerged; staining for non-sulphated chondroitin and chondroitin-6-suIphate declines in the dermal papilla and connective tissue sheath early in catagen and becomes undetectable by telogen. These changes appear to occur earlier than the decline in expression of other proteoglycans or basement membrane proteins. Involvement of proteoglycans in human hair-growth regulation had previously been inferred from observations in diseases such as pretibial myxoedema and the Hunter/Hurler syndrome in which disturbed proteoglycan metabolism is associated with hypertrichosis.' '"^ In newly involved areas of alopecia areata, anagen follicles are precipitated abruptly and prematurely into catagen. The unusual patterns of staining for chondroitin-6-sulphate in some anagen follicles from lesional sites may indicate impending catagen conversion. It was apparent that these changes were only in follicles where entry into catagen was likely to be imminent. However, loss of chondroitin-6-sulphate staining is not usually observed before the occurrence of early morphological changes of catagen. It is possible that the abnormal appearances in these lesional anagen follicles was due to degradation of extracellular matrix by infiltrating inflammatory cells. Nevertheless, we cannot exclude the possibility that these changes indicate primary dysfunction of the hair follicle mesenchymc with resulting premature entry of follicles into catagen. The grossly thickened and convoluted appearance of the glassy membrane in lesional catagen follicles has, to our knowledge, not been previously reported. We have not encountered similar appearances in catagen follicles from normal subjects. The role of the glassy membrane in the evolution of catagen is unknown and the significance of this observation is not clear. Nevertheless, these findings provide confirmation that the glassy membrane is composed of basement membrane proteins. Chondroitin-6-sulphate has been regarded as a basement membrane component' '^ and has recently been suggested to be important in the pathogenesis of dystrophic epidermolysis bullosa."" It is interesting, therefore, that staining for this proteoglycan did not co-distribute with laminin and type IV collagen in catagen follicles. In conclusion, the changes we have observed in alopecia areata may result from dynamic disturbances of the hair growth cycle. However, the loss of the normal staining pattern for chondroitin-6-sulphate in some lesional anagen follicles suggests that the changes which occur
724 A.J.G.McDonagh el al in the dermal papilla during normal catagen are occurring prematurely. This observation warrants further study and should be amenable to investigation by in vhro techniques. ACKNOWLEDGMENT
This work was supported by a research grant from the Special Trustees of the former United Sheffield Hospitals. REFERENCES 1 Thies W. Vergleichende hisiologische Untersuchungen bei Alopecia areata und narbig-atrophisierenden Alopecien. Arch Klin Exp Dermatol 1966; 227: 541-9. 2 Messenger AG, Bleehen SS. Alopecia areata: light and electron microscopic pathology of the regrowing white hair. Br J Dermatol 1984; I i o : 155-62. 3 Messenger AG, Bleehen SS. Expression of HLA-DR by anagen hair follicles in alopecia areata. J Invest Dermatol 1985; 85: 569-724 Briitker EB, Echternacht-Happle K, Hamm H, Happle R. Abnormal expression of class I and class II major histocompatibilily antigens in alopecia areata: modulation by topical immunotherapy. J Invest Dermatol 1987; 88: 564-8. 5 Khoury EL, Price VH, Greenspan JS. HLA-DR expression by hair follicle keratinocytes in alopecia areata: evidence that it is secondary lo the lymphoid infiltration. _7 Ifivest Dermatol 1988; 90: 193-200. 6 Van Scott EJ, Ekel T M . Geometric relationships between Che matrix of the hair bulb and its dermal papilla in normal and alopecie scalp. J Invest Dermatol 1958; 31: 281-77 Pierard GE, de la Brassinne M. Cellular activity in the dermis surrounding the hair bulb in alopecia areata. J Cutan Pathol 1975; 2: 240-5. 8 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: 337-479 Oliver RF. The induction of hair follicle formation in the adult hooded rat by vibrissa dermal papiUa^.J Embryol Exp Morphol 1970; 23: 219-36. 10 Oliver RF. The experimental induction of whisker growth in the hooded rat by impiantation of dermal papillae. J Embryol Exp Morphol 1967; 18; 43-51. 11 Couchman JR, CatersenB, Christner JB, Baker JR. Mappingby monoclonal antibody detection of glycosaminoglycans in connective tissues. Nature 1984; 307: 650-2. 12 Westgate GE, Lewis LP, Messenger AG, Gibson WT. Distribution of proteoglycans during the human hair cycle. 7 Invest Dermatol 1989; 92: 540A. 13 Couchman JR, Gibson WT. Expression of basement membrane components through morphological changes in the hair growth cycle. Dev Biol 1985; 108: 290-80. 14 Couchman JR, Rat hair foliicie dermal papillae have an extracellular matrix containing basement membrane components. J Invest Dermatol 1986; 87: 762-7. 15 Couchman JR, King JL, McCarthy KJ. Distribution of two basement membrane proteoglycans through hair follicle development and the hair growth cycle in the rat. J Invest Dermatol 1990; 94: 65-70. 16 Goetinck PF, Carlone DL. Altered proteoglycan synthesis disrupts feather pattern formation in chick embryonic skin. Dev Biol 1988; 127: 179-86. 17 Roberts SOB, Weismann K. The skin in systemic disease. In: Textbook of Dermatology (Rook A, Ebling FJG, Wilkinson DS et al., eds), 4th edn. Oxford: Blackwell Scientific Publications, 1986; 2343-74. 18 McKusick VA,Neufeld EF. The mucopolysaccharide storage diseases. In: The Metabolic Basis of Inherited Disease (Stanbury JB, Wyngaarden JB, Freilrickson DS el al., eds), 5th edn. New York: McGraw-Hill, 1983; 751-77, 19 Fine J-D, Couchman JR, Chondroitin-6-sulfate-containing proteoglycan: a new component of skin dermoepidermal junction, J Invest Dermatol 1988; 90; 283-8, 20 Fine J-D, Couchman JR, Chondroitin-6-sulfate proteogiycan but not heparan sulfate proteoglycan is abnormally expressed in skin basement membrane from patients with dominant and recessive dystrophic epidermolysis bullosa. J Invest Dermatol 1989; 92: 611-16,
