Hair Cycle-Dependent Changes in Skin Immune Functions: Anagen-Associated Depression of Sensitization for Contact Hypersensitivity in Mice

Udo Hofinann,*t Yoshiki Tokura,* Takafumj Nishijima,* Masahiro Takigawa,* and Ralf Paus"!" " Department of Dermatology, Hamamatsu University School of M edic ine . Hamam atsu. Japan: and j"Departmellt of Dermatology.

Virchow H ospital , Humbo ldt Universitiit Berlin. l3erlill, German y

To assess whether hair follicle cycling influences skin immunity, we examined the association between highly synchronized hair follicle cycling and experi­mental contact hypersensitivity in C57BLl6 mice. Hair cycle synchronization was performed by depila­tion of hair shafts on the back with telogen skin. Mice were sensitized on the lower back skin with picryl chloride between 0 and 25 d, after anagen induction by depilation, and challenged on the earlobes with picryl chloride 5 d later. The magnitude of contact hypersensitivity was significantly decreased in mice sensitized on day 1, was minimal on day 3 (early anagen), and slowly increased thereafter, reaching a level comparable to day 0 on day 25 (telogen). The significantly depressed contact hypersensitivity re­sponse in anagen skin was confirmed in mice with spontaneously developed follicles. Lymph node cells taken from mice sensitized with picryl chloride on days 0, 1, and 3 after depilation were cultured ill vitl'o

Severa l experimen ta l o bse rvatio ns in n'lice a',ld r,ats have suggested a cl ose association between hai r follicle cy­cling and cutaneous immuni ty, as observed in immuno­pa tho logic ph enomena associa ted with human hair diseases (Paus {'I (/1, 1989; R.ook and Dawber. 199 1;

Westgate CI ti l , '1991; Paus cl (/1, 1994b; Maj o r histocompatibility complex (MH C ) class I is imI11un o histochemica lly absent in th e ma trix and inne r root sheath of gro wing hair fo ll icl es . whe reas Qa-2, a nonclassic MHC class I I11 o lecul e, is present in the distal hair bulb durin g the enti re hair cycle (Paus el (/1, 1994a). Altho ug h o nly very discrete , if any, changes in Langerhans cells are notable during the va rio us stages of the murin e hair cycle ' (Hashi zume cl (/1, 1994), the number of dermal MH C class 11 -1- ce ll s increases dramatically in anagen and decreases again in telogen skin.' Of

Manuscript received February 18, 1995; rev ised November 14 . 1995; accepted for publi cation November 20. 1995.

R.eprint reques ts to : Dr. Yoshiki Tok ura, H amarna tsu Unive rsity School of Medicine. 3M)O H anda-cho. I-Iamamatsu 43 1-3 1. J apan .

Abbrev ia tio ns: ll6 , C5713L/6 : H S. contact hypersensiti vity: PC I. picryl chlo ride; TN P-L -EC. trinitrophcnylatcd Lan ge rhans cell- enriched ep i­dermal ce ll s.

1 Eichll1lill e r S. Pall s R.. l3c1l1lc E. Czarnetzki 13M. Hoti nann U: H ai r cycle-associated changes of the skin imlllllne sta tu s: MH class 11- and NLDC '145-il11nllllloreactivity in Illurine skin . .f I" ""sl Derliialol 103:446. 1994 (abstr. ).

in the presence of syngeneic, haptenized, Langerhans cell- enriched epidermal cells. Marked proliferative responses of lymph node cells to haptenized Langer­hans cells were found in mice not only of day 0, but also of days 1 and 3, suggesting that immune T cells exist even in lymph node cells of the low- responsive mice. Flow cytometric analyses demonstrated that the nUlllber of intraepidermal Langerhans cells and their functions, including the expression of major histocompatibility complex class II, GD54, and CD86, and mixed epidermal cell lymphocyte reac­tions, were not changed in skin on days 0, 1, and 3. These findings deillonstrate that contact hypersensi­tivity is induced most effectively via skin with telogen hair follicles and that the depressed response in early anagen skin is not simply due to failure in Langerhans cell ' functions or sensitization of T cells. Key lVoI,ds: C/Ita1/.eOJlS illUlumityldelayed-tylJe Ilypel'sellsitivitylllai'l' growtll. ] I,west DCI'II/atol 106:598-604, 1996

parti cular in te res t are the num eric and m o rpho logic chan ges of "YO-T-cell receptor (TCR)-bearing dendriti c epide rmal T cells (DETC). During anagen , the number of DETC increases highl y significantly in the interfollicuhll' epidermj s and in the suprainfun­dibular portion of the hair bulb (Hashi zume el (/1 , 1994; PaUS e/ (/1, 1994c) , along with a m o rphol ogic change from round o r ang ular to dendritic confi gurati o n (HashizuI11e el (/1, 1994). Finally, hair cycl e-dep endent changes in the transc"iption of genes fo r cyto­kin es , sLI ch as tumor necrosis factor-{3 and transforming growth facto r-{3, have been reported recentl y (Seiberg ei (/1, 1995).

Murine contact hypersensitivi ty (CHS) to hapten is a hjghJy spec ifi c, delayed-type hyperse nsiti vity reaction (Greene el (/1, 1978). Va"ious e pidermal and dermal cells arc in volved in bo th positive and negative regulation of immuno logic sensiti zation , including Lan gcrhans cells. D ETC, keratinocytes, and dermal m acrophages (Streilein, 1989; Tse and Cooper, 1990). Althou gh possessulg stro n g an tigen-presenting capacity, Langerhans cells can be con­verted fro m immunogenic to to lerogenic cells by immunomodula­tion such as u ltravio let B radiation (Simo n cl (/1. 1991b) . D ETC, whose fun ction is not fu ll y understood, are tho ught to downmodu­la te C HS reactions (Sullivan et ti l , 1986; Amornsiripanitchj el (/1. 1988; W elsh and Kripke, 1990; Bergstresser cl (/1, 1993). For example, m o use strains with a high num eri c ratio of DETC to Langerhans ce lls exhjbit lower C H S respo nses than those with a low ratio (Bigby cl (/1 , 1987). In addition, derma l MH C class n+

0022-202X/96/S1D.5 0 • Copyrigh t l 1996 b y T he Society for Investigative Dermatology. Inc.

598

VOL. 106. NO.4 APR.lL 1996

sensitization

anagen IIlduction ~_ bYdePilatIO/~~

~ day 0-25

~ day 0* telogen day 0 day I anagen I

day 3 anagen 1I-f11 day S anagen I V day 8 anagen V-VI day 12 aJiagen VI day 17- 19 calagen day 2S telogen

chal lenge

5 days after sensi ti zalion

Figure 1. Protocol of sensitization via skin in various stages of the m.u:ine hair cycle. *Mice not depilated (tclogen skin) . .

celh may be important for the induction of CBS (Tse and Cooper,

1990; Meunier e/ ai, 1993). Even keratinocytes ,]S nonprofessional

epidermal immunocompetent cells are capable of modulating CHS by releasing various cytokines that ~ Itcr the function of Langer hans

cells and DETC (Kupper, 1990). These observations encouraged us to investigate hair cycle-associated changes in murine CHS as an

example of functional consequences that migllt be associated with

highly synchronized hair cyclipg. To this end, we use d the well-defined depilation-induced hair

cycle ofC57BL/6 (B6) mice (Paus ef ai, 1989, 1990, 1994a,c) . B6 mice with aU back-skin follicles in different, defined hair cycle

stages were sensitized with a hapten, and the magnitude of CHS responses was compared. We provide evidence for !~air-cycle

dependency of the development and magnitude of CHS, in that CHS is depresse d when J11i~e are sensitized during synchronized

early anageJl.

MATERJALS AND METHODS

Animals B6 and BALB/c mi ce obtained from Japan SLC (Hamamatsu. Japan) were l11aintained in a conventional anilnal f.1ciljty of Halnanlatsu Uni versity School of Medicin e . Male mice, 7-8 wk old unless otherwise mentioned, were used in aU experiments.

Syncl),ronization of Hair Cycle by Depilation Anagen was induced by d epilation of hair shafts on the backs of B6 mice with all follicl es in te1ogen, as described previously (Argyris. 1967) . Briefly. anesthetized mice with all back- skin follicles in telogen were painted with a warm wax/rosin mixture (1 :1), which was peeled off after hardening. thus depilating all back-skin hair shafts. Telogen skin was identifi ed by its pink co lor. Although immediately after depilation, the skin ma y go through a short phase of a d.iscrete wound-healing response (Paus el aI, 1990), this tcchnigue predictably induces the development of very homogeneous anagen follicle pop ulations , which arc morphologic,Illy and functiona lly indistinguishable from spontaneously developing anagen follicles (Argyris, 1967; Paus eI (/1, 1990; Slominski el aI, 199:1). . .

Sensitization to Picryl Chloride (PCl) and Elicitation of CHS Contrary to conventional sensitization protocols for C HS studies, B6 m ice were sensitized with PC l (2,4,6-trinitrochlorobenzene; Tokyo Kasei Co., Tokyo. Japan) by painting once with 50 /-LI of 0 .5%, 1 %, 2'%. and 5% PCI in eth' \11ol:acetone (3:1) onto the clipped lower back (approximately 1.5 cm' ) . Five days after sensitization, 20 /-LI ofl 'X, PCI in oliv.e o il was painted on both sides of each earlobe. Ear thickness was measured with a dial thickness gauge (Peacock, Tokyo, Japan) before and 24 h after challenge . For each nlouse, the ear s\ve lHng response \vas expressed as the ll1 can in crc lnent in thickness above the basa l control value from both earlobes.

ProtocolofExperil11ents As shown in Fig 1 , the back skin of B6 mice with ali hair foUicles in tclogen was depilated on day O. The groups of mice were sensitized with 1°1., or 5% PC I onto the depilated lower back between days 0 and 25. and challenged with 1 % PCI 5 d later. The indicated stages of induced hair cycle were confirmed previously by microscopic analysis of longitudinal skin sections (Paus et ai, 1990, 1994a ,c; Slominski el al. 1991) and corresponded to the full hair cycle as c haracterIzed previously (Straik ci

I-lAIR CYCLING AND CO NTA CT HYPER.SENS IT IVITY 599

aI, 1961). When mice were immunized o n the depilated skin. haiJ' shafts that had regrown after depilation were carefull y shaved before sensiti zation without injuring the epidermis. For con tro l purposes , the CHS response of these dep ilation-induced 'lIlagen mice was compared with that of 12-week­old mice that had entered anagen VI spontan eously and with undcpiJated telogen mice. Two independent series of experiments. in which each group consisted of four to five mice. were performed ; both experimental series gave highl y comparable results. T he results we.re therefo re expressed by pooling the data from both series of experim ents (n = 8-1 0 mice per groupf

Culture Mediul11 We used R.PMI-1640 (Gibco Laboratories, Grand Isl:lIld , NY) supplemented with 10'X, hea t-inactivated feta l bovine serum, 2 mM L-glutamine. 5 X 10- 5 M 2-me rcap toethaJlol, 100 U / ml penicillin , and 100 /-Lg/ ml streptomycin (Tokura el nl. 1994a).

III Vit,'o Assay for 11111uune Ly-.nph Node Cell Response Three.B6 mice in each group were sensiti zed with 5% PCI onto the lower back O. 1. or 3 d after d ep ilation. Five days later, I)'mph node cells were co llected from the inguinal and axi ll ary ·rcf,';ons.

Epidermal cell suspensions were obtained from naive mouse earlobes with trypsin, as described previously (Token'a ('/ aI, 199 1) . Freshly isolated single-cell suspensions o f epidennal ce ll s were cu ltured in plastic dishes for I d in the culture m edium . Langerhans cell- enriched epidermal cells were prepared b y separating nonadherent ce lls with Ficoll-Hypaqlle (specific gravity: 1.083 ; Sigma C hemica l Co., St. Louis, MO) (Tokura el al. 1994b). The percentages of Langerhans cells and DETC in the Langerhans \=e11-enriched epidermai cell fi'action were 10 .1 % and 4.5%, respectively. as assessed by flow cytom etric analys is, as described below. '

The trinitrophenyl coupling of Langer hans cell- e nri ched ep idermal ce lls (TNP-LC-EC) was performed by incubation ,v;th 2,4,6-trinitrobenzcne sulfonic ac id (TokY9 Kasei Co .), as re ported previously (Tokura el al. 19943).

Immune lymph !lode cell s (3 X 10'/150 /-LI) were incubated in the culture m edium in the presence or abse nce of TNP-LC-EC (1 X 10

5 or 5 X 104 / we ll) or concanavalin A (Con A) (final concen tratio n: 5 /-Lg/ ml ; Sib'm a C hemica l Co .). Indomethacin (Sigma C hemical Co.) was added to the culture at a final concentra tion of 1 /-Lg/ ml. Triplicate cultures were Illajntaincd in 96-,vcli Inicro tes t plates (Corning Glass Works. Cornin g, NY) at 37°C under 5% CO, in air for 72 h. Methyl-tritiated thymidine eH-TdR) (Amersham IntI. , Arlington . IL; 1 /-LCi/ well) was present for the last 16 h of culture. The cells we re collected on glass fiber filters using a cell harvester (Cambridge Technologies, Watertown, MA). and their radioac­tivity upt::tke ,vas ITIcasured in a scintiUation counte r.

Mixed Epidermal Cell Lymphocyte Reactions Single-cell suspen­sions of spleen cells were prepared from naive DALB/c mice by teasing spleens in R.PMI-1640 , and e rythrocytes were lysed in 0.83% ammonium chlo ride in Tris buffe r (pH 7.6). Epidermal cell s were freshl y isolated with trypsin from the dorsa l skin ofBG mice unde pilated (d .. y 0) o r depilated 1 or 3 d before. Spleen cells (2 X 105 / well, responders) were cultured for 72 h with epidermal cells (2 X 105 or 5 X 104 / well, stimulators) , and 3H-TdR (1 /-LCi / we ll) was added to each well 16 h before termination of culture.

Prepara~ion of CnIture Snpernatants From Keratinocytes Epi­dermal cell suspensions were prepa red with trypsin trom the dorsal skin of B6 mice that was undepilated (day 0) or depilated 1 o r 3 d before. The cells (5 X 106 / ml) were c ultured in supplemented RPMJ-1640 in a 24-well plate (Corning Glass Works) at 37°C under 5% CO, in air. Seventy-twO ho urs after incubation, the supernatants were harvcsr~d.

Flow Cytol11etric Analysis The following monoclonal antibodies (MoAbs) were used: phycoerythrin-labeled anti-Illouse I_A" (Becton Dick­inso n , Mo untain Vicw, CAl, fluorescein iso thiocyanate (FlTC)-labe1ed anti-mouse pan yo chain of T-ce ll receptor (TCR) (PharMingen , San Diego, CAl, FlTC-labeled anti-mouse intercellular adhesion molecule-l (CD54; Phat:Mingcn) , FlTC-labe led anti-Illo use B7-2 (CD86; PharMin­gen) , and a;lti -1110USe Fcyll / lIl receptor (PharMingen). FITC-Iabeled anti­mouse <r{3 TCR (PharMingen) and FITC-Iabeled rat 19G (PharM1I1gen) served as contr.ols . All MoAbs were used at 1-5 /-Lg/ 1 06 cell s. and incubation was perfollllcd for 30 min at 4°C, fo llowed by twO washes in phospha te­buffered saline supple mented with 5% feta l bovine serum and 0.02% sodIUm azid~ . Epidennal ce lls were first incubated with anti-mouse Fcyll / trJ receptor M oAb fo r 5 min to prevent binding of the subsequent reagents to Fe recep tors. Nonspec ifi c srains were adcguate same-class 19s for speCIfi c MoAbs, such as a{3 TCR for CD54 and rat tgG for B7-2 . Fluorescent profi les were generated using a FACScan (Becton Dickinson Imln~noc)'­tometry Systems, Mountain View. C Al. Dead cells were Idcntlned by propidium iodide uptake. To anal yze the expression of CD 54 and CD86 on Langerhans cell s. we double- sta ined ep ide rmal ce lls with phycoe,-ythrin-

600 HOFMANN ET AL.

12

\I --- 1% PCI

E 10 ••••••••• 5 % PCI

u '( 9

0

>< 8

.., 7 '" c:: 6 8. '" ~ 5 bI)

. s 4 dj ~

r··· .. . ** .. .... .···1 ! Nega'ive

••• •• Control ···········1·········································· .................. ......... .

'" .... 2 '" UJ

o I 3 5 8 12 25

Days post depilation

Figure 2. CHS responses during the induced murine hair cycle. Seven to 8-week-old B6 mice with tclogen hairs were undepilated (day 0) or depilated and sensitized with 1% or 5% picryl chloride (PCI) on the indicated days after depilation. Negative control mice were challenged without sensitization. Values were obtained from eight or ten mke, except for days 17, 18, and 19 (catagen groups, n = 23 mice), which were combined because of no significant differences among them. Ve,1icnl bars represent so. ' p < 0.05, " p < 0.005, compared with the day-O group.

labeled anti-I-Ab and FlTC-labeled anti-C054 or -C086 MoAb. I_Ab + cells were ga ted, and the intensities of these adhesion molecules were monitored.

Statistical Analysis Student's t test (unpaired) was used to determine statisticitl differences between means in the ear swelling response and 3H_TdR incorporation.

R ESULTS

Successful Induction of CHS to PCI by Painting With Varying Concentrations of PCI onto the Telogen Lower Back Skin Before testing the association between CHS and hair follicle cycling, we examined the efficacy of sen sitizatio n w ith varying concentrations of PCl onto the lower back skin with

Figure 3. Skin reaction to the application of 5% PCl in mice sensitized on days 0, 1, 3, and 5 after depilation (from Icft to J'igllt) . T he day-O animal was not depilated (telogen).

THE JOU ltNAL OF INVESTIGATIVE DERMATOLOGY

telogen fo llicl es, corresponding to the depilation site. Because thi is a n unconventional sensitization site for CHS, it was necessary to dem.o nstrate that challenge with PCI would still e licit a CHS response of th e car. Significant CHS respon ses we re indeed induc­ible in mice with telogen skin by application of 0.5- 5% PCI in a dose-dependent manner (data not shown) . T hus, we used 1'!!., and 5% PCI for sensitization in th e following experiments.

Depression of CHS Responses Induced via Skin in Depila ­tion-Induced Early Anagen Hair-Cycle Stage T he back skin ofB6 mi ce with telogen h air follicles was depilated, and groups of mice w e re sensitized with 1 % PCI applied o nto the depilated lower back between day 0 and da y 25 after anagen induction and cha ll enged w ith 1% PCI 5 d late r (see Fig 1). T h e indi cated stages of th e induced hair cycle had been assessed previously by micro­scopi c morphometry (Paus e/ aI, 1990, 1994a,c; Slomin ski et al. 1991; Handjiski e/ aI, 1994; Hashizum e e/ aI, 1994) and corre­sponded to the full h air cycle as characterized previously (StraiJe er ai, 1961) .

T h e magnitude of CHS induced with 1% PC I was significantly decreased in mice sensitized on d ay 1, was minimal on day 3 (early anagen), and slowly increased again thereafter, reaching a level comparable to day 0 on day 25, i.e., only during the subsequent telogen skin (Fig 2) . In mice sensitized with 5% PCI, a similar pattern of CBS responses was found between days ° and 5 after depilation. T h o ugh maxim al suppression of CHS was seen when mice had been sensitized during earl y anagen, a ll time points of sensitization during the entire anagen phase were associated with a substantially decreased CHS resp o nse after e licitation on the ear­lobes. It is noteworth y that the percentage suppressio n o n day 3 relative to day 0 was as high as 95% and 75% in mice sensi ti zed witll 5%, and 10

;') PC l, respectively; that is, CHS practically could not be induced b y 5% PCl when ea rly anagen (II-Ill) mice were sensitized. The CHS values of negative contro l mice that were challenged without sensitizatio n were unchan ged during the h air cycle.

In relation to the magnitude ofCHS, each experimenta l g roup of mice showed different skin reactions to applied PCI. Whereas mice on day ° exhibited a stron g reaction, i.e., crust formation, the skin reaction of day-3 and day-5 mice was subtle , and that of day-1 mice was intermediate (Fig 3) . Th us, the degree of sensitization reaction through ea rl y an agen skin was parallel to th e magnitude of the CHS response .

Depression of CHS Responses Induced via Spontaneous Anagen Skin T h e C H S response evoked via skin of spontalle-

VOL. 106, NO.4 APR.IL 1996

9

R

......... 7 E u

"( 0 6

>< Q) 5 v:: t:: 0 Q.. v::

4 ~ O/j

.5 "0 :I ~ v::

~ 2 t..Ll

Telogen Induced anagen

Spontaneous Negative anagen control

Cycle stage

Figure 4. Comparison ofCHS between mice with spontaneous and depilation-induced anagen. Telogen, 8-week-old mice with telogen skin; induced anagen, 9-week-old mice on day 12 after depilation; sponta­neous anagen, 12-week-old mice with anagcn skin; negative control, 8-week-old mice challenged without sensitization. Each group consisted of eight to ten mice. Vertical bars represent SD. ' p < 0.01, compared with telogen group.

ously developed anagen was examined in comparison with depila­tion-induced anagen skin. Twelve-week-old mice with spontane­ously developed hair follicles in the later stages of anagen (VI), as identified by their black skin color (Paus et ai, 1990; Slominski et ai, 1991), were used as the "spontaneous anagen" group . Mice on day 12 after depilation (anagen Vr) were used as the "induced anagen" group because of its histologic resemblance to spontaneous anagen (Su:aile et ai, 1961; Argyris, 1967; Paus et ai , 1990) . As shown in Fig 4, mice with spontaneous anagen showed a level ofCHS depression that was comparable to that in the induced-anagen group, thus confirming that the results obtained from the experiments using the induced hair cycle do not reRect abnormal skin responses associated with depilation.

In Vitro Proliferative Responses to TNP-LC-EC of hnmune Lyttlph Node Cells From. Mice Sensitized via Skin ofVari­ous Hair-Cycle Stages We compared the magnitude of the proliferative T -cell response in mice that were sensitized with PCl v ia skin of early-anagen hair follicles (Fig 5). Mice on days 0, 1, and 3 a fter depilation were sensitized with PCI, and lymph node cells were collected 5 d later and cultured with TNP-LC-EC or Con A. All three groups of lymph node cells responded well to TNP-LC­Ee, indicating that lymph node cells in mice sensitized via early­anagen skin as well as telogen skin contained a substantial number of primed T cells. The degree of responses of the day-l and day-3 groups induced by 5 X 104 TNP-LC-EC, however, was signifI­cantly lower than that of the day-O group, as monitored by the ab solute number of 3H-TdR incorporation. The proliferation of sensitized lymph node cells to Con A also was significantly higher

HAIR C YC LING AND C ONTACT HYPERSENSITIV1TY 601

Day 0

TNp· LC-EC 5.

Day 1 TNP-LC·EC 5x 10 •••

Day 3

0 10 20 30 40 50

Day 0

Con A

Day 1

Con A

Day 3

Con A

0 100 200 J H-TdR ineorporalion ± SO (epmx 10 .

3)

Figure 5. Proliferative responses of lymph node cells (LNC) from mice sensitized via skin in different hair stages. LNC (3 X lOS/well) from mice sensitized on days 0, I, and 3 post-depilation were cultured for 72 h in the presence or absence of trinitrophenylated Langerhans cell­em-jehed epidermal cells (TNP-LC-EC) (1 X 105 or 5 X 10'/well) or concanavalin A (Con A) (5 J.Lg/ ml). The mean cpm ::!: SD of the TNP-LC-EC group was 1,120 :': 293 (1 X 105) and 169 ::!: 16 (5 X 104

), and that of day-O LNC + untreated LC-EC was 8,596 ::!: 1,790. ' p < 0.01 , compared with the corresponding no-addition group; " p < 0.05, com­pared with the TNP-LC-EC (5 X 10' ) group on day 0; "'p < 0.05, compared with the COil A-addition group on day O.

in lymph node cells derived from telogen mice than from early­anagen mice. Accepting that the addition of Con A induces T-cell proliferation, lymph node cells from telogen mice therefore con­tained T-cell populations that proliferated more readily in response to nonspecific stimuli.

Numeric Changes of Langerhans Cells and DETC and Alteration of Expression of MHC Class II and Costimula­tory Molecules on Langerhans Cells in Early-Attagen Skin Previous studies using immunohistochemical teclmiques indicated that no remarkable changes take place in the intraepidennal number of Langerhans cells and DETC during days 0 to 3 after depilation (Hashizume et ai, 1994; Paus et ai, 1994c). To confirm tlus observation and to clarifY tile mechanisms for early anagen­associated CHS depression, we obtained epidermal cell suspensio~s from tbe skin of B6 mice and subjected them to flow cytometnc analysis . As summarized in Table I, the percentages of both Langerhans cells and DETC were decreased on day 3. If no one takes into account, however, the anagen-associated increment in the number of epidermal keracinocytes on day 3 (Paus e( ai, 1990; Slominski et ai, 1991), the numbers of these immunocompetent cells were not substantially changed. Tlus suggests that th~ ob­served CHS depression is not simply derived from a decrease 111 the numbers of intraepidennal Langerhans cells or DETC.

Alternatively, it is possihle that functional changes occur in Langerhans cells during days 0 to 3. Because the expression of MHC class II and costimulatory molecules is correlated with the a.ntigen-presenting functions of Langerhans cells (Simon et ai,

602 HOFMANN ET At. THE JOU RNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Number of Langerhans Cells (LC) and Dendritic Epidermal T Cells (DETC) and Expression of Major Histocompatibility Complex Class II and Costimulatory Molecules on Langerhans Cells in Telogen Vel'sus

Early-Anagen Skin"

Total Number of EC/Mouse

( X 107)

Percentage of Cells (Absolute Number, Xl 0s)

Intensity of Surface Molecule Expression on LC (Log Mean Lntensity)

Day 0 1.6 ± 0.2

Day 1 1.5 ± 0.2

Day 3 3.1 ± 0.3

LC

5.3 ± 0.1 (8.4 ± 0.9) 5.7 ± 0.1

(8.5 ± 0.9) 3.7 ± 0.1" (11 ± 1.3)

DETC

10.7 ± 0.2 (17 ± 2.0) 7.2 ± 0.1 (11 ± 1.2) 4.7 ± 0.1" (15 ± 1.6)

CD54 CD86

196 ± 7 59 ± 4 67 ± 5

192 ± 5 58 ± 4 69 ± 4

190 ± 4 69 ± 4 69 ± 3

/I Epidermal ce ll (EC) slIspen sion s were prepared from the sam e nrca of undcpUatcd (tc logc n, day 0) or depilated (carly 311agcll. day "' o r 3) skin ofBG mice 0 11 th e illdic..::ucd day after depilation. After cou ll ting of the total e pide rmal cell number. they were subje cted to flow c ytomctric analys is. The percentages ofL"ngcrhans cells and DETC were examined with anti-I-A h and :lI1 ti-')'15 TCR M oAb. respectively. After double-staining w ith anti-I-All ;1nd anti-CDS4 o r :lt1ti-CD86 M oAb. the J_Ah+ cells wcre gated. and thc intensity of surf.1ce molecules on Langcrhans cells was monitored. The v;,lues represen t the m ean :t SO of triplicate determinations .

j, p < 0.05 , compared with the percentages of L",gerhans ce lls and DETC on day O.

1991 a; Symington cl aI, 1993), the expression leve ls of I-A, CD54, and CD86 on Langerhans cells were examined in epidermal cells at day 0, 1 , and 3 after depilation . As seen in Table I , the levels of these three molecules were not decreased in day-3 epidermal cells compared with the day-O group, suggesting that the depression of CHS does not stem from functional changes of Langerhans ce lls re lated to the expression of these indices.

Proliferative Capacity and Allogeneic Stimulatory Ability of Langer hans Cells in Early-Anagen Skin Mixed epidermal cell lymphocyte reactions were performed to examine furth e r whether functional changes of Langerhans cells occur at the early- anage n stage. At the same time, the proliferative capaci ty of keratinocytes durin g early anagen was monitored . Epiderma l cell s taken from B6 mice on days 0, 1, and 3 after depi lation were cultured fOl- 72 h , and their DNA syn thesis was measured. Table II shows that the basic proliferative ability of day-3 epidermal cell s was significantly higher than that of the day-O and day-1 groups, confirming the resultant increment of epidermal cell numbe r .in day-3 skin. This is in line with the well-recognized, substantia l increase in epidermal thickness during the anagen stage of the murine hajr cycle (Argyris, 1967; Paus el ai, 1990, 1994a-c; Slominski el (/1, 1991). By co- cul turing these B6 epidermal cells (H-zi') with spleen cells from BALB/c mice (H_2d

), their allogeneic stimu latory ability was also tested. There were no substantial differences among the three groups of epidermal cells used as stimulators, as assessed by ~cpm of th e response, except that day-3 epidermal cells (4 X 104

) showed a margin al decrease in their all ogen eic stimulatory ability as compared with the sam e cell number of the day-O group .

Suppressive Effect of Supernatant From Epidermal Cells of Day 0 and Day 1, Not Day 3, on the Proliferative Response of Immune Lymph Node Cells to TNP-LC-EC Pel-immune lymph node cells were cultured with TNP-LC-EC from naive mice in the presence of cnlture supernatants from day-O, -1 , and -3 epidermal cells at 1:4 and 1 :16 final dilutions. As shown in Table III, epidermal cell supernatants of day 0 and day 1, bnt not day 3, at a 1:4 dilution significantly but marginally suppressed the prolif­eration of immune lymph node cells stimulated with TNP-LC-EC. N o signifi cant change in the lymph node cell responsiveness was induced by the addition of e ither of these supernatants at 1:16 . T hese data suggest that the day-3 supernatant did not contain an immunosuppressive factor(s) for primed lymph node cell prolifer­ation.

DISCUSSION

Our study demonstrates that the development of CHS to PCI is influen ced by both depilation-induced and spontaneous hajr cycling in skin to which PC I is applied durin g the sensitization phase . Our

data indica te that CI-IS is indu ced most effectively VIa skin with telogen hair follicles and that the entire anagen stage of the murine h air cycle is associated with substantial C HS depression . Maximal CHS depression occurred when CHS was induced in ea rl y-anagen skin; this was associated with minimal cutaneous reactions to PCI at the sensitized site as compared with those in telogcn mice.

By ;'1 IIiIYO analyses, prolifera tive responses oflymph node cells to haptenized Langerhans cells were marked not only in telogen n1.i ce, but a lso in early-anagen mice, although the proliferative response of the latter group was marginally lower than that of the form er group. Thjs suggests that immune T cells do exist even in these low-responsive mice and that, therefore, the depressed CHS re­spon se in early-anagen mice is not simply du e to a failure in

Table II. Spontaneous Epidermal Cell (EC) Proliferation and Mixed Epidermal Cell Lymphocyte Reaction"

Group

EC alone Day 0

2 X 10' 4 X 10.1

Day 1 2 X 10' 4 X 10'1

Day 3 2 X 10' 4 X 10'

SC + EC SC alone Day 0

2 X 10' 4 X 10.1

Day 1 2 X 105

4 X 10' Day 3

2 X 105

4 X 10'

JH_TdK Incorporation ± SD (cpm)

927 ± 99 343 ± 65

873 ± 135 346 ± 45

2,696 ± 54" 2,484 ± 345"

2,732 ± 220

4,698 ± 20 1 3,708 ± 402

3,629 ± 7"1 3,487 ± 352

5,936 ± 619 4,955 ± 218

~cpm

3,77 1 3,365

3,542 3,14 1

3,240 2,471

( I Epidcnn'l l cell suspensions were obtained from tclogcn B6 mice (d;,)' 0) and from mice depil ated ·1 or 3 d before. The percentages of Langerhan s cells, as assessed by Aow cytotJ1etric analysis usin g 'lI1ti-I_A b MoAb, were as follows: day 0 , 5.3%: day 1. 5.7%; day 3. 3.7'Yo. T he indicated numbe rs of epidermal cells were cultured for 72 h. and [heir =' H-TdR incorporation was measured. At the sam e time, nuxcd epidermal cell lymphocyte reac tions were performed using the sa m e B6 cpidcrlll.Ll cc Us and IJALB/c sp leen cell s (sq . ~cpm represents (cpm of [SC + EC] - cpm of [Ee alone]) .

b p < 0.01, compared with da y-O and day-·I groups at the correspondiJlg cell numbe r.

VOL. 106, NO. 4 APRIL 1996

Table III. Effects of Culture Supernatants From Epidermal Cells (EC) of Depilated Skin on Picryl

Chloride-Immune Lymph Node Cells (LNC) Stimulated With Trinitrophenylated Langerhans Cell-Enriched

Epidermal Cells (TNP-LC-EC)"

EC Supernatant Added

LNC alone LNC + TNP-

LC-EC Oay-O supernatant Oay-l supernatant Oay-3 supernatant

3H-TdR Incorporation ::':: SO (cpm) at Final Dilution of Culture Supernatant

1:4 1 :16

5,661 ::':: 1,076

44,630 ::':: 1,847 34,938 ::':: 2,254" 45,534 ::':: 5 ,250' 36,048 ::':: 4,520" 46,752 ::': 2 ,860' 44,214 ::':: 738" 50,288 ::':: 3,101'

a Culture supernatants from epidermal cells obtained from tclogcll mice (day 0) and from mice depilated .1 or 3 d before were added to the Pel-immune lyluph node cells stimulated with TNP-LC-EC. The mean ::': SD of day-a LNC + untreated LC-EC was 7,580 :!: 824.

" P < 0.05. cornparcd with LNC + TNP-LC-EC group. e StatisticaHy not significant.

sensitization of T cells. Surprisingly, the response of lymph node cells to a nonspecific T-cell mitogen, Con A, varies greatly between mice with all back-skin follicles in telogen and those in early anagen, with telogen mice showing a significantly enhanced pro­liferation response. This suggests that even extracutaneous immune functions in mice are affected by the predominating hair-cycle stage in a large skin territory (back skin). Tlus astonishing concept is suggested by recent evidence that murine splenocyte proliferation fluctuates significantly in a hair cycle-dependent lmumer (Slomin­ski et "/, in press).

Several possible mechanisms for depressed sensitization through early-anagen skin can be envisioned on the basis of previously characterized immunologic changes associated with hair cycling in mice. First, the decreased sensitization may be due to numeric and functional changes of immunocompetent cells, i.e., Langerhans cells, DETC, and demlal macrophages. Our previous immunohis­tochernical studies1 (Hasluzume e/ "/, 1994), however, and the present flow cytometric analysis demonstrate that the absolute number of intraepidermal Langerhans cells does not change sub­stantially during induced hair cycling. Ye t, beca use the total number of keratinocytes in early-anagen epidermis increases as a result of anagen-associated epidermal proliferation, the percentage of Langerhans cells decreases at this time. This might be relevant for the depression of CHS when mice are sensitized via ea.d y­anagen epidernus. Standard indices for assessing Langerhans cell functions, including the expression of MHC class II and of the costimulatory molecules CD54 (Simon ef "I, 1991a) and CD86 (Symington e/ aI, 1993; Kawamura and Fume, 1995) and mixed epidermal ceIJ lymphocyte reactions ascribed to these mol ecules (Sin"1on et a/, 1991a), were not altered among day-O, -1, and -3 skin. Thus, it appears unlikely that anagen-associated changes in standard Langerhans cell functions explain the observed suppression of CHS.

In contrast to Langerhans cells, the number of intraepidermal and intrafo1Jjcular DETC changes significantly during the murine hair cycle; during early anagen development, the DETC number in­creases mainly in the interfollicular epidermis (Hashizume et ,,/, 1994; Paus e/ a/, 1994c). Given the proposed immunosuppressive role of DETC for the development of CHS (SuIJivan et "/, 1986; Bigby et a/, 1987; Amornsiripanitchi e/ a/, 1988; Welsh and Kripke, 1990; Bergstresser e/ "/, 1993), the downregulation of CHS in early- anagen skin may be related to the increased number of DETC. There is a time lag, however, be tween the day of the maximal depression of CHS (day 3) and the peak in the number of DETC (days 8-10) (Hashizume et ,, /,1994; Paus et a/, 1994c), and the flow cytometric analysis revealed no substantial change in the number ofDETC in skin on days 0, 1, and 3 . This indicates that the

HAIR CYCLING AND CONTACT HYPERSENSIT IVITY 603

depression ofCHS induction cannot be explained solely by numeric changes of DETC. Also, the number of class n+ macrophages in the dermis increases as early as day 4 after depilation. J Though this seems to parallel the observed depression of CHS, it remains unclear whether these populations of macro phages serve as positive enforcers (Welsh and Kripke, 1990) or as counterregulatory cells of CHS.

Second, it is possible that hair cycling influences the production ofkeratinocyte-det;ved cytokines capable of modulating the induc­tion ofCHS, such as interieukin-10 (Enk et ai, 1994; Li et ,,/,1994) interleukin-1 a (Robe rtson et "/, 1987), and tumor necrosis factor-a (Kur1moto and Streilein, 1992). The absence of a strong skin reaction at the PCI sensitization site on day 3 may reflect a low or excessive production of inflanU11atory cytokines by keratinocytes (Gran stein et "I, 1986) that are either il11munopotentiating or immunosuppressive during early anagen, with resultant unsuccess­ful induction of CHS. In this context, we have found a strong anagen-associated increase in the cutaneous expression of the pro-opiomelanocortin gene and its product /3-endorphin in mice (Slominski et " /, 1992), and have checked the presence of the immunosuppressive pro-opiomelanocortin product, adrenocortico­tropic hormone, in human scalp hair follicles in anagen, but not in telogen (Slominski et "/, 1993b). Given the generallyinununosup­pressive properties of pro-opiomelanocortin products such as adre­nocorticotropic hormone and a -melanocyte-stimulating hormone (Slominski e/ ,,/, 1993a), it is reasonable to speculate that pro­opiomelanocortin products operate as hair cycle-dependent de­pressors of CHS responses . In this study, we could not detect immunosuppressive activity of day-3 epidermal cell culture super­natant for immune lymph node cell proliferation to hapteluzed Langerhans cells, whereas day-O and day-1 ep idermal cell superna­tants inhibited the lymph node cell proliferation. T lus raises the possibility that although the immunosuppressive factor is no longer present at day 3, its long-lasting action leads to the depression of CHS 'It day 3. This fact together with the finding that the relative number of Langerhans cells in early-anagen skin is lower than in control skin might contribute to the observed immunosuppression.

Third, hair cycle-dependent fluctuations in mast cell functions and the potential role of mast cells in hair groWtll regulation (Maurer e/ aI, in press) may be relevant to understanding the early anagen-associated d epression of CHS. For example, we have shown recently that early anagen in mice is associated with substantial mast cell degranulation a.nd a dramatic decline in lustochemically detectable skin mast cells (Paus e/ ,,/, 1994d). The precise role of mast cells in the sensitization, elicitation, and la te-phase reaction of CHS responses is still obscure, and much conflicting evidence has been reported; several studies have sug­gested a role for mast cells in the efferent, though not the afferent phase (Askenase ,md Van Loveren , 1983; Galli and Dvorak, 1984) . Certainly, it is reasonable to ask whether the massive degranulation of m ast cells seen in the phase of induced early anagen (pallS ef ai ,

1994d), wluch m ay be accompanied by substantial release of tumor necrosis f.'lctor-a, has a downregulatory effect on the sensitization phase of murin e CHS.

That the depression through early-anagen skin does not stem simply fi'om failure in immunization ofT cells (see above) raises the possibility that suppressor T cells or T helper 2 (Th2) cells are preferentia.1ly induced by hapten application duriug this hair cycl.e phase. It is known that immunomodulating agents such as ultravl­olet B irradiation convert Langerhans cell functions to preferentially activate Th2 cells (Simon et "/, 1990) and stimulate keratinocytes to produce interleu.kin-l0 (Ullrich, 1994), w luch in turn suppresses delayed-type hypersensitivity-mediating T hl cells (Fiorentino e/ "/,

1989). Also, the possibility d eserves consideration that the discrete wounding response associated with the early stage of depilation­induced anagen (Paus e/ "i, 1990; Argyris, 1967; Slominski et a/, 1991; PallS, Homlann, e/ aI, 1994) contribu tes to the maximal depression of CHS induced via day-3 skin. Whether such alter­ations in antigen-presenting and immunoregulatory cell functions as well as in keratinocyte cytokine production actually do occur

604 HOFMANN ET AL.

during early anagen is an intriguing issue to be clarified in future studies.

The present find in gs also provide an important insight for the conventional murin e experimental system of CHS. Mice with ages from 7 to 11 weeks are usualJy used for CHS experiments because of the e mpiri cally well -appreciated e ffectiv eness of sensitization in mi ce of this age group (Ashe rson and Ptak, 1968) . In murine spontaneous hair cycling, the first and second waves of anagen de ve lopment occur around the fourth and 11 tb w eeks after birth , respectively (Dry, 1926) . Therefore, 7- 10-week-old mjce have telogcn skin, through which we have shown CHS to be induced m ost e ffecti vely. It seems that the conventional usage of this age range is based not only on a fully developed immune system at this time, but also on inadvertent usage of the best hair-cycle stage for sensitiza tio n in mice of this age. I t should be kept in mind that the C H S response is closely associated with syn chron.ized hair cycling patte rn s in the skin where sensitization is pe rformed . Taken together, the current data strongly suggest that previously noted hajr cycl e-associated flu ctuations in a number phenomenologic skin immune indices (Westgate el ai, 1991 ; Slominski el ai, 1992; Hashizume ef ai, 1994; Paus el ai, 1994a-d; Seiberg et ai, 1995; Maurer el ai, in press) are fun ctionall y re levant. T his strongly encomages o ne to dissect further the as yet obscure interdependen­cies of hair follicl e cycljng and the skin immune stages (Pans and Link, 1988; PallS el ai, 1994b,e) .

T"is '"ll rk lI'ns sllpp0I1ed ill pm1 by n gralllfrolll Coslll el% gJ' FOlllldalioll , japall, 10

MT; by a g ralll fro III Delllse"e Forsc/llwgsgellleillscl"lji (DFG Pa 34513-2) 10 RP;

alld hJ' a Irmlel gralll frOIll Boehrillger IlIgel"eilll Fl/l llis 10 UH. T l,e nrbliee ali(I

sllpport oj Pro! Beale M. Heu z is gratiflllly aekllollliedged.

REFERENCES

Amornsi ~i panitchi S, Barnes LM , Nordlund)), Trinkle LS, Rheins LA: Immune studies in the depigmcntcd C57BIl Ler-vit/vit m ice : an apparent isolated loss of contact hypersensitivity.] 1111111111101140:3438 - 3445, 1988

Argyris TS: Hair growth induced by damage. A dv Bioi Skill 9:339 -34 5. 1967 Asherson GL. Ptak W : Contact and delayed hypersensitivity in the mouse. I. Active

sensitization and passive tra_nsfer. Immrmolog), 15:405-416. 1968 Askenase PW. Van Loveren H : Delayed-type hypersensitivi ty: ac ti vation of mast cells

by antigen-specific T-cell factors initiates the cascade of cellul ar interactions. hlllllllllol Today 4:259 -264, 1983

Bergstresser PRo Cruz PO, T akashima A: Dendritic epidennal T cells: lessons from mice to humans.] I'lilest Oerlllalol 100:80S-835, 1993

Bigby M, Kwan T , Sy M-S: R atio of Langerhans cell s to Thy-1 + dendritic epidermal celts in murine epidermi s influen ces the intensity of contact hypersensitivity. J hlllesl Derlllalol 89:495-499, 1987

Dry EW: The coat of the mouse (Mus musculum).] Geller 16:287-340. 1926 Enk AH . 5aloga J, Becker D. Mohamadzadeh M . KnopJ: Induction o f hapten-specific

tolerance by interleukin 10 in vivo. ] Exp Mcd 179 :1 397-1402. 1994 Fiorentino OF, Bond MW, Mosmann TR: Two types of mouse helper cells. IV. Th2

clones secrete a factor that inhibits cytokine production by Thl clones .] Exp Med 170:2081-2095, 1989

Galli 5) , Dvorak AM: What do m ast cells have to do with delayed-rype hypersensi­tivity? Lab hlllw 50:365-368, 1984

Granstein RD. Margolis R.. Mizel SB. Saunder DN : In vivo inflammatory activity of cpidcnnal cell-derived thymocyte activating fa ctor and recombinant interleukin 1 in the mouse.] Clill hllles' 77:1020-1027,1986

Greene MJ , Sugimoto M. Benacerraf B: Mechani sms of regulation of cell-mediated immune response . I. Effect of dle route of immunization with T NP-collplcd syngeneic cells o n the induction and suppression o f contac t sensitivity to picryl chloride . ] [11111111110/120:1604-1611 . 1978

Hnndjiski B, Eichmiiller S, Hofmann U , CZ;II·nctzki BM, Palls R..: Alkaline phosphatase activity and localiza tion during the murin e hair cycle. Br] Del"lllalo/131 :303-31 O. 1994

Hashizume I-I, T okura Y . Takigawa M: Increased number of dendritic epidermal T cells associated with induced imagen phnsc o f hnir cycles. J Demrntol Sci 8: 119-124 ,1 994

Kawnmura T. Funte M: Comparative analysis of 07-1 and D7-2 cxpression in Langerhans cells: differential regulation by T helper type 1 and T helper rype 2 cytokincs. Ell r] /"'"111"01 25: 19 13-1917. 1995

Kupper TS: Immune and inflammatory processes in cutnneous tissues: mech nn.isl11s and speculations. ] Clill III lies' 86: 1783-1 789. 1990

Kurillloto I. Streil c.in JW: cis-Urocani c acid suppression of contact hypersensiti vity induction is mediated IIi" tumor necrosis fa ctor-a. J 1"",,"'101 148:3072-3078. 1992

Li L. Elliot JF, MOS1l1ann T R.: IL-t 0 inhibits cytokinc produ ction . vascular leakage. and

THE JOURNAL OF IN VEST IGATIVE D ER.MATOLOGY

swelling: during T helper t cell-induced delayed-type hypersensiti vity. J 111"'''11101 153:3967-3972. 1994

Maure r M. Paus R. Czarnctzki 13M: Mast cells as modul a'·o rs of hair follid e cycHng. Exl' Oe,."lO,ol (in press)

Me unier L. Gonzalez-Ramos A, Cooper KD: Heterogeneous populations of class II MI-I C ·j. cells in human dernw l cell suspension s: ide ntificatio n of a slllall subset responsibl e for potent derm al antigen-presenting cell activity \-vi th features analogous to Langerhans cells. ] 111111111,1111 151 :4067-4 080. 1993

POlUS R~ Link lli: T he psoriatic epiderm;ll les ion and an;lgcll hair g row th m ay share rJ1 C

same " switch-on" mechanism. l/(/Ie) Ui,,1 M .. d 6 1:467-476.1988 P41U S R . Eichrni.illcr S. H ofi1l 'JIlIl U. Cza rnctzki OM . R o binson P: Expression of

classica l and li o n-classical MH C ci:ISS I ;mti gc ll s in murin e h Ol ir fo llicles. B,. J 0"l"IlIal,,/13 1 :1 77-183 .1 994a

Paus R. Handjiski B. Czam etzki BM . Eichllliill cr S: Biologic des Haa rrollikcls. 1-la,"(/r::1 45:808-825. 1994b

Palls P..... I-Iofm anll U , EichmuUcr S. CZ;lrn etzki UM : Distrihuti on and challging density of gnmma-dclta T cells in murine skin d uring: the induced hair cycle . BrJ Dcmratol 130:28 1-289. 1994c

POlUS R. Maure r M. Slo minski A. Czarnct"zki 13M: Ma st cell involve me nt in murine hair growth. Ow Uirol 163 :230-240. 1994d

Paus R . Slominski A. Czarnctzki OM : Is alopecia arcata nn auto imlllulle-response agOli nst melanogenesis- related proteins. ex posed hy abn oTl"llill MI-I C class I-ex­pression in the :magen hair bulb' \,(/1 .. ./ U;"I M .. d 66:5 4·1-554. '1 994e

P Ol US It. S t C l111 KS. Link Il...E: T he indu ctio n of al1:1gen hai r gruwth in ("dogc n mouse skin by cyclosporin A administra tio n . Lab 11I" est 60:3(.5-369. 1989

P olU S R . SWlln KS . Link R.E: Tc logcll sk ill con(":lin s an inhibito r of hair growth . BrJ Oerlllal<" 122:777-784 . ·1990

Robertson B. Gahring L. Newton R , Dayncs R: IIr IJil IO administration o f intcrleukin 1 to 110 rl11:11 rnice depresses their capaci ty to elicit COll t:lct hypcrsensitivity respo nscs: prostaglandins arc involved in thi s 1110d ifH;ation of imrnllllC fun ction. ) I'""'sl OCI"lII alol 88:380 - 387 . .1 987

I,-oak A. Dawber 11.. (cds.). Diseases (!j" III e Hair alld Sca l" . Blackwell. Oxford. 199 1 Sci lberg M. Marthinuss J. SWill) KS: C han ges in expressio n of apoptosis-associated

genes in skin mark early illlngen . J l'IllCS( Dcr-IIIIIt,,1 104:78 - 82, 1995 Simo n ) C, Cruz PD. Bergstresser PRo Tige laa r R E: Low dose ul traviolet B-irradiated

Langerhans ce lls prefcrcntial.1y nctivate CD4 + cells of the T h elpe r 2 subset. ] 1111111111101 145:2087- 2091, J 990

Sil110n J C, C ruz PO Jr. T igelaa.r RE, Sontheill1c r RD. Bergstresser PR: Adhesion 11l0 lecuJ cs CD lla, CD1 8, and ICAM-l o n human epidermnl Langcrhans cells serve a functio nnl role in the activation of aUoreactive T cel1 s. J f' lllesr Denll(!{ol 96 :1 48 -151 . 199 1a

Simon ) C, T igelaar RE, Bergstresser 1'11... Edclbaul11 0, Cruz I'D: Ultravio let B radiation converts Langerhalls cells fro m immunogenic to tolerogcn ic antigeu­presenting cells. ] 1111111111101 146:485-491, 199 1b

Slo minski A. C;:;oodman-Sni tkoff G, Maures M , Paus R : H air cycle-associated changes in splenocyte proliferation .) /JIIIW Oem.a'ol (let.) (in press)

Slominski A, Palls R, Constantino R : Differential expression and activity of melan o­genesis-related protein s during induced hair growdl in mice. J I" ves( Dc""acol 96: 172-1 79. 1991

Slominski A. Paus R. Mazurkicwicz J E: Pro-opiomclanocortin expression in the skin during induced hai.r growth in mice. Experilll elltia 48:50-54 , 1992

Slo l11.inski A, Paus R... Wortsman J: O n the potential ro]e of proopiomclanocortin ill skin physio logy and pathology. Mol Cell Elldo,lillol 93:C1-C6. 1993a

Slominski A. Wortsman J. Mazurkiewicz J E. Matsuoka L, Dietrich J . Lawrence K. Gorbani A, Paus R: Detection of proopiomclanocor tin-derived nntigcns in normal and pathologic human skin.] Lab Clill lvled 122:658 - 666. 1993b

Strnilc W E, Chase ;J-f.B, Arsenault C : Growth and di(fercntiatio ll of hnir follides between periods of activity and quiescence.] EXI} 200/148:205- 221. 1961

Streil ein)W: Skin-associated lympho id tissue. In : Norris DA (cd.). [IIlIIlIlIIe Mcc/,allisllls ill C lllall eo lls Disease. Marce l Dekker Inc .. N ew York. 1989. pp 73-94

Sullivan S. Bergstresser PR, Tigclaar RE, Streilein jW: Induction and regulation of contact hypersensitivity by resident, bone marrow-derived, dendritic epidermal cells: Langerhans cells and Thy-I + cpidermal cells. ] [111111.11101 137:2460-2467. 1986

Symington FW, Brady W , Lisley PS: Expression and function of B7 on human epidermal Langerhans ce ll s. ] 1111111111101 150:1286-1295, 1993

Tokura Y, Satoh T, Yamada M, T akigawa M: Genetic contro l of contac t pho tosell­sitivity to tctrachlorosa1icylanilidc. II. Igh complex controls the SenSlOVlty induced by photohapten-modi_fied spl een cell s but no t epidermal cells. Cdl 1111"'"110/1 35:195-207 , 1991

Tokura Y. Yagi I-I. I-Iashizume H , Yagi) , Furukawa F. Takigawa M: Accessory cell ability of Langerhnns cells for slIperalltigcn is resistant to lIltraviol et-B light. Plloloc/.ell. P/IO/obi,,1 60:147-1 53. 1994a

Tokura Y. Yagi) . O'Malley M, Lewis )M , Taki gawa M. Edelson I'-L. Tigelaar RE: Supcn1l1tigell_ic stnphylococcal cxotoxins induce T -ceU proliferation in the pres­ence of Langerhans cells or class II-bearing kcratinocytcs and stimulate kerati­nocytes to produce T-cell-activating cytokincs. ] fllI'csl D erlllalol 102:31-38 , 1994 b

Tse Y, Coopcr KD: C utaneous derlllalla+ cells arc capable of initi ating delayed type hypersensitivity responses. ) flmw D emwloI 94: 267-272, 1990

UUdch SE: Mcchanjsl11 involvcd in the systemic suppress ion of antigen-presenting cell fun ction by UV irradiation: kcratillocyte-derived IL-10 modulates alltigcn~

presenting cell function of splcnic adherent cells. ] 1111111111101 152:34 10-3416, 199-1 Welsh EA. Kripk e ML: Murin e Th y-1 + dendritic epiderm:11 cell s induce immuno logic

to lerance ill "i,,,,. ) 111'1111111011 44 :883-891.1990 W estgate GE. Craggs Rl, Gibson WT: I mill II lie privilege in hair growth. J 1II'lfS'

Domlrl",1 97:4 17-420, 199 1