An in vitro model of dermatophyte invasion of the human hair follicle

Journal of Medical & Veterinary Mycology 1996, 34, 3742 Accepted 18 July 1995 An in vitro model of dermatophyte invasion of the human hair follicle A. RASHID, M. B. HODGINS & M. D. RICHARDSON Regional Mycology Reference Laboratoq, Department of Dermatology, Robertson Building, University of Glasgow, Glasgow GI I 6NU, UK Introduction Affected hair tissues from cases of tinea capitis have been studied using light microscopy, scanning, and transmission electron microscopy [1-3]. However, ultrastructural findings of the parasitic form of Trichophyton mentagrophytes in the hair tissue have been inconclusive and the studies have not used intact hair follicles. The pathological changes of the affected hair structure are poorly understood. The stages by which detached hairs are attacked by keratinophilic fungi are: (i) cuticle lifting, (ii) cortical erosion, (iii) production of penetrating organs, and (iv) colonization of the medulla [4]. The ability to invade hair in vitro is a property of the keratinophilic fungi in general but various species differ in the way this is accomplished. It has been found that the direction of invasion and the pathological role of the fungal elements within the hair apparatus are significantly different between fungi [1]. A novel in vitro model for the study of hair invasion by Trichophyton mentagrophytes was developed. Hair was obtained by microdissection and plucking. Following inoculation of the hair follicle with arthroconidia growth of the fungus was seen on the hair and within the follicle. Growth was observed to begin at the shaft end and to extend along the hair shaft towards the bulb area. In follicles maintained in organ culture the inner root sheath in particular was invaded by T. mentagrophytes and provided a good substrate for fungal growth. Initially, the cuticle formed a barrier to fungal penetration of the hair. After incubation for 4 days, germlings of T. mentagrophytes were seen to penetrate under the cuticle and in between the layers of cuticular cells to invade the cortex. There was no evidence of intracellular growth; fungal elements were seen intercellularly. There were similarities between the findings in this study of the process of hair invasion by dermatophyte fungi and that in the natural disease. Correspondence: Dr M. D. Richardson, Regional Mycology Reference Laboratory, Department of Dermatology, Robertson Building, University of Glasgow, 56 Dumbarton Road, Glasgow Gll 6NU, UK. Tel.: +44 141 330 4010; Fax: +44 141 330 4008; E-mail: goga07@udcf.gla.ac.uk. 1996 ISHAM Keywords tinea capitis, Trichophyton mentagrophytes, hair follicle. Baxter & Mann [5] observed the invasion of human hairs in vitro by three keratinophilic fungi, namely T. mentagrophytes, T. rubrum and T. ajelloi. Differences were seen in the penetration and digestion of cuticular cells by these three species of fungi. The early stage in the colonization of human hair by T. mentagrophytes involves the production of flattened, branched fronds, the 'eroding mycelium', which seems to be common to all species of keratinophilic fungi but the 'perforating organs' are formed only by a restricted number of dermatophytes. Based on studies of hair invasion described by Baxter & Mann [5] and Mercer & Verma [6] there is evidence that the process of hair invasion involves enzymatic breakdown. Hsu & Volz [7] observed complete digestion of human hair resulting from enzymatic lysis by T. terrestre. A specific keratinase has been isolated from T. mentagrophytes which is capable of digesting hair [8,9]. Kunert & Krajci [10] studied the degradation and digestion of human hairs by Microsporum gypseum in vitro and suggested that along with the action of enzymes, mechanical action of the hyphae was also found in the cuticle. Experimental studies of hair penetration by dermatophytes are few and have been restricted to cut or plucked hair [6,11,12], and arthroconidia have not been previously

~l~ Rashid et al. used in experimental studies of hair infection. There are no satisfactory experimental models for detailed study of the invasion of hair follicles by dermatophyte fungi. This study was undertaken to investigate the ability of T. mentagrophytes arthroconidia to morphologically transform on hair follicles cultured in vitro and to observe the process of hair colonization and invasion by germlings of T. rnentagrophytes. Materials and methods Organism and stock cultures One strain of T. mentagrophytes var. interdigitale (strain 126, isolated from a case of tinea pedis, Mycology Reference Laboratory Culture Collection, Glasgow, UK) was selected on the ability to form a homogeneous culture of arthroconidia. Preliminary experiments designed to determine the optimal conditions for arthroconidial production revealed that this was highly variable between strains of T. mentagrophytes. Stock cultures were maintained on 4% glucose-l% peptone agar. Production and preparation of arthroconidia Arthroconidia were prepared as pure suspensions of separated cells from 10-day-old cultures grown as surface lawns on 4% glucose-l% peptone agar in an atmosphere of 10% CO2, 90% air at 28 C in an incubation module (Flow Laboratories, Irvine, UK), and flushed through daily with 10% CO2, 90% air. Surface growth was harvested from the culture plate with a scalpel blade, suspended in phosphate-buffered saline (PBS), ph 7'2 and agitated for 5 min. The suspension was filtered through column chromatography grade glass wool to remove chains of arthroconidia, washed three times by agitation in PBS followed by separation at 300 g for 3 min, and adjusted to a concentration of 5 x 104/ml in PBS. This procedure resulted in a uniform arthroconidial suspension of single cells. No obvious damage to arthroconidia had occurred. Viability of singlet arthroconidia was determined by incubation in Sabouraud glucose broth for 16 h at 37 C. An arthroconidium was considered to have germinated when a visible germ tube had developed. The percentage germination was 97.00 + 1.89 (SEM). Microdissection and maintenance of hair follicles in organ culture Using a stereomicroscope, anagen hair follicles were microdissected from fresh skin pieces of normal human scalp after swabbing and removal by normal surgical procedures. Scalp skin fi'om two normal adult females undergoing cosmetic surgery was used. Hair follicles were dissected at the dermosubcutaneous fat interface and were maintained for 4 days at 37 C in Dulbecco's modified Eagle's medium supplemented with 2% w:v L-glutamine and gentamycin (50 tg ml-1). Macroscopic observation revealed continued viability as shown by increased growth of maintained follicles with the appearance of a hair shaft. Terminal hairs were also plucked from the scalp, and anagen roots cut with a scalpel just distal to the adherent outer root sheath. Some plucked hairs were dissected to remove either the outer root sheath only or both the outer root sheath and the inner root sheath. Microscopical examination of uninoculated hair follicles did not reveal colonizing yeasts. Inoculation of hair follicles Dissected human terminal hair follicles were washed twice in carbon dioxide-independent tissue culture medium (Gibco BRL, Paisley, UK; based on Eagle's minimal essential medium) and then placed onto sterile polycarbonate filter membranes (Nucleopore) floating on 2 ml of CO2-independent tissue culture medium. Plucked anagen hair follicles were maintained in a similar way. A 5/A inoculum of a suspension containing 5 x 104 arthroconidia per ml in PBS was inoculated onto the hair shaft end Of the follicles. The petri dishes containing hair follicles were incubated for varying lengths of time at 28 C. Hair follicles set up as above, and ones from the same donors, but without arthroconidia inoculated on them were used as controls. Assessment of hair invasion and antifungat activity Fungal growth was assessed by gross examination. Intact specimens were examined daily using a dissecting microscope and photographed. For light microscopy hairs were inoculated as above and then examined by phase contrast microscopy. For Toluidine blue staining hairs were fixed in gluteraldehyde, dehydrated and embedded in araldite resin as for transmission electron microscopy. Sections 1 flltl in thickness were cut and then stained with Toluidine blue. Hair follicles were also fixed and processed for scanning electron microscopy using standard methods. Results Organ culture of hair follicles On gross examination microdissected hair follicles which had been cultured showed an increase in length, confirmed by the appearance of a keratinized hair shaft. On scanning electron microscopy plucked follicles showed a well-formed shaft, with its sheath and the bulb area intact,

Dermatophyte invasion of hair follicle a b Fig. 1 Light microscopic appearance of: (a) a freshly isolated hair follicle and (b) after 96 h in culture ( 250). the cuticular cells being flat and overlapping with the free edge of the cells. Microdissected hair follicles showed an increase in length, which was confirmed with the appearance of a keratinized hair shaft. At 4 days dissected hair follicles showed a fully-formed hair shaft emerging from the connective tissue sheath (Fig. 1). Gross assessment of fungal growth on hair fol#cles On follicles inoculated with arthroconidia, hyphal growth was visible by 3 days of incubation at 28 C. Growth started at the shaft end and gradually extended down towards the bulb end. On plucked hair follicles growth of the fungus was seen to be inhibited on the outer root sheath and sparing the bulb area. After 7 days incubation, the keratogenous zone of the hair shaft was seen to be digested and it was observed that some hair had totally disappeared under the thick mycelial mat formed with only pigment remaining. These changes were more marked in plucked follicles compared with dissected follicles. Growth was more rapid on the dissected hair follicles which were covered by mycelium early compared 1996 ISHAM, Journal of Medical & Veterinary Mycology 34, 37~12 with the plucked hair follicles. It was observed that fungal growth was maximal on the shaft of the follicle compared with the other components. Arthroconidial germination was visible on inoculated follicles by 16 h of incubation at 28 C. Germination (100%) of arthroconidia was observed by 40h. With longer incubation the whole hair follicle was covered by a dense mat of mycelium. With the plucked hair follicles, in the initial stages, growth of germlings was inhibited on the outer root sheath. It was also noticed that there was lack of fungal growth around the hair bulb which was seen to be swollen. Germlings were seen growing along and encircling the hair shaft. The hyphae were seen very prominently on the keratogenous zone. The fungus was seen destroying this zone while the rest of the hair was relatively intact. Invasion of the inner root sheath Longitudinal sections of uninoculated dissected hair stained with Toluidine blue showed clearly the different layers. Sections of hair after 24 h incubation showed that the arthroconidia had germinated and germlings were growing on the inner root sheath with evidence of penetration of this layer. At 40 h of incubation there was extensive growth of fungal elements on the inner root sheath which was fully invaded by the fungus. The cuticle was not penetrated by the fungus at this stage, although fungal elements were seen entering through distal and proximal ends of the follicle. Plucked follicles with the outer root sheath removed showed abundant growth. Where the outer and inner root sheaths both had been dissected, extensive invasion of the follicles was seen with most of the structure replaced by the fungal mycelium. Invasion of the hair shaft After incubation for 4 days, invasion of the shaft was seen from the distal and proximal ends. The inner root sheath was broken up and fully invaded by fungal elements which were seen to penetrate in between the cuticular cells. No invasion of the cortex was observed through the cuticular layers (Fig. 2). Around 6 days of incubation invasion of the cortex was seen by fungal elements penetrating through inner root sheath and cuticular cells, and from the distal and proximal ends. Fungal elements were seen growing in the cortex between the fibrils and with time, replacing most of the structure (Fig. 3). Following inoculation of arthroconidia at the distal end of the follicle, germination of arthroconidia was seen by scanning electron microscopy and germlings of T. mentagrophytes began to grow abundantly along the hair surface, closely encircling the hair shaft towards the proximal end, with no apparent damage to the hair in the early

Rashid et al. Fig. 2 Toluidine blue-stained section of a dissected hair follicle showing extensive growth of fungus in the inner root sheath after incubation for 4 d at 28 C. At the level of the cuticular cells there is evidence of inhibition of fungal invasion of the hair cortex (arrow) although a few fungal elements are seen within the cuticle. Fig. 3 Toluidine blue-stained section of dissected hair showing extensive growth of fungus after incubation for 4 d at 28 C. Note the fungal elements breaking up the inner root sheath and growing within the hair cortex (arrow). stages of infection. The direction of growth on the hair surface was against the direction in which the cuticular scales overlap. In addition, some of the hyphae seemed to penetrate beneath the free edge of cuticular cells which in the early stages remain firmly adherent, but were seen to be slightly raised later on. Hairs infected with the fungus showed large flaps of cuticle raised from the underlying cortex (Fig. 4). Extensively invaded hair showed hollowing out of the hair shaft with abundant arthroconidial formation. Around 7 days of incubation most of the hair was covered Fig. 4 Scanning electron micrograph of a plucked hair showing growth of fungal hyphae on the hair shaft after incubation for 40 h at 28 C. Note the flaps of raised cuticle ( 800). Fig. 5 Scanning electron micrograph of an early germ tube originating from an arthroconidium lying on the cuticular scales of a plucked hair. Incubation at 28 C for 16 h ( 20 000). with mycelium, with the hyphae growing in all directions. Dissected hair follicles showed rapid and extensive growth of fungus and were fully covered with growth by 4 days of incubation. In the initial stages of hair invasion, separation of the cuticular cells caused by the germlings was seen but without evidence of damage to these cells (Fig. 5). Later on the cuticular cells had become markedly detached from the underlying cortex with germlings penetrating in between them with evidence of damage to the endocuticle of cuticular cells (Fig. 6). Fungal elements were seen penetrating the central part of the hair cortex (Fig. 6).

Dermatophyte invasion of hair follicle Fig. 6 Transmission electron micrograph of fungal elements penetrating in between and separating layers of the cuticular layers of a hair shaft. Note the appearance of slight damage to the endocuticle (arrow; x 15 000). Most of the filaments of the mycelium were aligned parallel to the long axis of the cortical cells. Arthroconidia were not observed in the cortex. Discussion In this study hair invasion was observed in terminal hair follicles isolated by microdissection and maintained in tissue culture in the absence of serum. The conventional methods available for the isolation of intact hair follicles from humans are either to pluck hair from the skin using mechanical force or to microdissect follicles from biopsy tissues. Hair follicles can also be isolated by shearing, a technique that recovers the whole follicle intact and with no evidence of tissue damage. Although the follicles remain metabolically viable for several days in culture, they do not grow to produce a new hair [13]. In contrast, by careful microdissection it is possible to obtain undamaged human hair follicles which grow in vitro at an in vivo rate. Philpott et al. [14] have shown that it is possible to maintain isolated human hair follicles in serum-free medium in vitro for up to 10 days during which time the dermal papilla remains elongated and more significantly, the hair follicles continue to produce a keratinized hair shaft with the hair root sheaths growing with the hair shafts. Hair follicles used in the present study showed an increase in their length of the keratinized shaft indicating that they were viable. The advantages of using dissected hair follicles is that the most proximal part of the hair with its intact sheaths can be seen and the effect of the connective tissue sheaths on fungal growth can be assessed. However, use of plucked hairs permits direct application of fungal conidia to the follicular root sheaths and proximal hair shaft. The two methods therefore permit complementary experimental approaches. There were similarities between the findings in this study of the process of hair invasion by dermatophyte fungi and that in the natural disease. As the experimental system described here with dissected hair is very close to the in vivo situation hair roots culture in this way can serve as a useful model for studying the morphological transition of arthroconidia and for the detailed study of the process of hair invasion by different dermatophyte fungi. It is known that dermatophytes are keratinophilic fungi and the ability to invade hair in vitro is a property of the keratinophilic fungi in general, such as Anixiopsis fulvescens and A. stericoraria. Vanbreusenghem [15] described that the majority of dermatophytes are able to grow on hair and invade it, although the growth of dermatophytes on hair in vitro was quite different from that in vivo. Differences in the structure of the keratinaceous tissue is possibly a reason for the irregular growth of dermatophytes in the skin, nail and hair. Dermatophytes attack the keratinous tissue as a parasitic form, while it has been shown that T. mentagrophytes is capable of causing extensive breakdown of keratin in the saprophytic stages. Baxter & Mann [5] examined the pattern of invasion of cut human hair in vitro by three dermatophytes (T. mentagrophytes, T. rubrum, and T. ajelloi) and found variations in the keratinolytic ability of these species. T. mentagrophytes was the most keratinolytic and was seen to breakdown the hair keratin. Large numbers of granules and mitochondria were observed in T. mentagrophytes which were not observed in the other species and the authors suggested that this could be due to the higher enzymatic activity of T. mentagrophytes. Mercer & Verma [6] look at the invasion of sterile cut human hair by T. mentagrophytes in vitro. They also found that the process of hair invasion involved an enzymatic breakdown of keratin bundles, with complete absence of the keratin bundles as seen in the present study. T. mentagrophytes was seen to digest the endocuticle, thus detaching the cells from the underlying cortex. Similar findings were observed in the present study where evidence of damage to the endocuticular layer was seen and the cuticular cells had detached from the underlying cortex. In both of these studies the dermatophyte was initially seen growing intercellularly followed by intracellular growth. In the present study there was no evidence of intracellular growth as most of the fungal elements were seen growing intercellularly. In this study, germlings of T. mentagrophytes were seen to penetrate under the cuticle and in between the layers of cuticular cells. Daniels [16] observed on the surface of emergent hair, frond-like mycelium which appeared to be responsible for cortical erosion and suggested that it might also be the agent of cuticle lifting. It was suggested

Rashid et al. that M. canis was capable of digesting human hair keratin enzymatically and keratinases have been isolated from different dermatophytes. It has been shown that keratinases of T. mentagrophytes are capable of digesting hair [8]. The stages of hair invasion by dermatophytes [4] were all recognized in the present study but with some degree of overlap. Penetrating organs were not seen as they are a feature of saprophytic growth of T. mentagrophytes on hair. Kunert & Krajci [10] studied the process of hair invasion by M. gypseum in vitro and found that the process of keratin degradation had features of enzymatic breakdown, with some evidence of mechanical effect of the hyphae on the cuticular cells. Shelley et al. [17] suggested that the cuticle acts as a barrier against fungal invasion and is not vulnerable to the keratolytic enzymes. The pattern of invasion of human hair by non-dermatophytic fungi shows many similar features. Chrysosporium tropicum is a soil-inhabiting fungus which when inoculated onto autoclaved hair formed perforating hyphae and showed all the characteristics suggestive of enzyme digestion [18]. In contrast, the lipophilic yeast Malassezia fitrjur, when seen in a case of scalp hair infection, appeared to form nodules adhering to the hair shaft but with no evidence of cuticular invasion [19]. In the present study germlings were seen penetrating under and in between the cuticular cells and it is possible that this may be due to a mechanical action of the germlings, although there was evidence of enzymatic damage to the cuticular cells. It appears that the process of hair invasion by dermatophytes is a combination of both mechanical and chemical forces, one augmenting the other. The inhibition of growth by the outer root sheath and the preferential invasion of the inner root sheath deserve further study. Acknowledgements We thank the staff of the Department of Ophthalmology for their assistance with the scanning electron microscopy and to Mrs Rita Mackie, Department of Dermatology, for her transmission electron microscopy skills. References 10kuda C, Ito M, Sato Y, Oka F. Fungus invasion of human hair tissue in tinea capitis caused by Microsporum eanis. Light and electron microscopic study. Arch Dermatol Res 1989; 281: 238M6. 20kuda C, Ito M, Sato Y. Trichaphyton rubrum invasion of human hair apparatus in tinea capitis and tinea barbae: light and electron microscopic study. Arch Dermatol Res 1991; 283: 233 9. 3 Tosti A, Villardita S, Fazzini ML, Scalici A. Contribution to the knowledge of dermatophyte invasion of hair. J Invest Dermatol 1970; 55:123 34. 4 English MP. The saprophytic growth of keratinophilic fungi on keratin. Sabouraudia 1963; 2:115-30. 5 Baxter M, Mann PR. Electron microscopic studies of the invasion of human hair in vitro by three keratinophilic fungi. Sabouraudia 1969; 7: 33-7. 6 Mercer EH, Verma BS. Hair digested by Trichophyton mentagrophytes. Arch Dermatol 1963; 87:357 60. 7 Hsu YC, Volz PA. Penetration of Trichophyton terrestre in human hair. Mycopathologia 1975; 55:179 83. 8 Yu R J, Harmon SR, Blank F. Hair digestion by a keratinase of Trichophyton mentagrophytes. J Invest Dermatol 1969; 53: 166 71. 9 Yu R J, Ragot J, Blank F. Keratinases: hydrolysis of keratinous substrates by three enzymes of Trichophyton mentagrophytes. Experientia 1972; 28: 1512-13. 10 Kunert J, Krajci D. An electron microscopy study of keratin degradation by the fungus Microsporum gypseum in vitro. Mykosen 1981; 24: 485-96. 11 Kaaman T, Forslind B. Ultrastructural studies on experimental hair infections in vitro caused by Trichophyton mentagrophytes and Trichophyton rubrum. Acta Dermato-Venerol (Stoek) 1985; 65: 536~9. 12 Raubitschek F, Evron R. Experimental invasion of hair by dermatophytes. Arch Dermatol 1963; 88: 837-45. 13 Green MR, Clay CS, Gibson WT, et al. Rapid isolation in large numbers of intact, viable, individual hair follicles from skin: biochemical and ultrastructural characterization. J Invest Dermatol 1986; 87:768 70. 14 Philpott MP, Westgate GE, Kealey T. An in vitro model for the study of human hair growth, in: Stenn AG, Messenger HP, Baden HP, eds. The Molecular and Structural Biology of Hair vol. 642. New York: Annals of the New York Academy of Sciences, 1991: 148-64. 15 Vanbreuseghem R, Keratin digestion by dermatophytes: a specific diagnostic method. Mycologia 1952; 44: 17(~82. 16 Daniels G. The digestion of human hair keratin by Microsporum canis Bodin. J Gen Microbial 1953; 8:289 94. 17 Shelley WB, Shelley ED, Burmeister V. The infected hairs of tinea capitis due to Microsporum canis. Demonstration of uniqueness of the hair cuticle by scanning electron microscopy. J Am Acad Dermatol 1987; 16: 354-61. 18 Fusconi A, Filipello-Marchisio V. Ultrastructural aspects of the demolition of human hair in vitro by Chrysosporium tropicum Carmichael. Mycoses 1991; 34: 153-65. 19 Lopes JO, Alves SH, Benevenga JP, Encarnacao CS. Nodular infection of the hair caused by MolasseziaJurJur. Myeopathologia 1994; 125: 149-52.