Cytokine mRNA Levels in Alopecia Areata Before and After Treatment with the Contact Allergen Diphenylcyclopropenone

Rolf Hoffmann, BIke Wenzel, Andrea Huth, Pieter van der Steen, Monika Schaufele, Hans-Peter Henninger, * and Rudolf Happle Department of Dermatology, Philipp University, Marburg; and "Department of Biochemistry, Albert-Ludwig Universiry, Freiburg, Germany.

Although the nature of the noxious signal and the ana­tomical target in alopecia areata (AA) are still un­known, it has been assumed that CD4+ T lymphocytes surrounding and infiltrating the hair bulb might trigger the hair loss. As these T lymphocytes do not promote cytotoxic activity we hypothesize that AA is triggered by cytokines. Topical immunotherapy with diphenylcyclopropenone (DCP) is at present the most effective approach. If it is true that AA results from a distinct cytokine pattern, we can hypothesize that the beneficial effect of DCP should be mediated by locally secreted cytokines during the contact al­lergy. Using semiquantitative reverse transcription­polymerase chain reaction with RNA extracted from scalp biopsies from patients with AA before and after successful treatment with DCP, and from healthy con­trols we detected aT-cell response with increased steady state nill.NA levels for interferon (IFN)-y, inter­leukin (IL)-lP, and IL-2 in untreated AA of the totalis

Alopecia areata (AA) is a common disease of the hair follicle affecting about 1-2% of the total of new patients attending dermatology clinics. The unpre­dictable course of the disease and the lack of a specific treatment have profound effects on the quality of life

of aff~cted individuals. The events in AA that induce premature telogen or dystrophic anagen growth pattern are not understood. In the past, immunologic studies have provided evidence that CD4+ T lymphocytes surrounding and infiltrating the affected hair bulbs might trigger the hair loss [1,2]. Although the nature of the noxious signal and the anatomical target are still unknown, it has been as­sumed that an autoimmune process is involved [3].

Unequivocal evidence for the existence of antifollicle autoim­mune reactivity is still lacking. and other immunologic findings have likewise attracted much attention. In this regard increased levels of soluble interleukin (IL)-2 receptors [4] and increased ex­pression of adhesion molecules (ICAM-l , ELAM) [5,6] and HLA-

Manuscript received December 22, 1993; accepted for publication May 10,1994.

Reprint requests to: Dr. med. Rolf Hoffmann, Zentrum fiir Hautkrank­heiten, Philipp Universitat, Deutschhausstra(\e 9, 35037 Marburg, Ger­many.

Abbreviations: AA, alopecia areata; DCP, diphenylcyclopropenone; PCR, polymerase chain reaction; RT, reverse transciption.

type. After DCP treatment, the IFN-y expression was reduced but still above the constitutive level found in controls, whereas mRNA expression ofIL-2, IL-8, IL-10, and tumor necrosis factor-a was increased. Our results point towards cytokines involved in the patho­genesis in AA. AT H 1 type cytokine pattern is present in untreated AA, and this is modified by cytokines se­creted during DCP treatment. IL-10 has recently been described as an immunomodulator of the THl response and, therefore, we hypothesize that basal keratinocytes or lesional T cells secrete bioactive IL-10 after DCP application, resulting in an inhibitory effect on le­sional T lymphocytes. This hypothesis would explain the effectiveness of DCP and implies the theoretical possibility of a response to topical or intralesional ap­plication of recombinant IL-10. Key words: alopecia ar­eatajdiphenylcyclopropenone. J Invest Dermatol 103:530-533, 1994

DR/HLA-A13C molecules on affected hair follicles and endothelial cells [7 - 9] have been reported. So far, however, no comprehensive pathophysiologic explanation for these findings is available.

In AA the hair follicles are not destroyed but rather "switched off" in stage IV [10] anagen although T lymphocytes invade the hair follicles and apparently interact with the matrix cells. Such interaction may occur either by direct cell-to-cell contact and cyto­toxicity, or indirectly by lymphokines that might affect the differ­entiation and growth of follicular epithelial cells. As the T lympho­cytes obviously do not promote cytotoxic activity we hypothesize that AA is triggered by soluble mediators. The presence of such mediators (e.g., interferon [IFN]-y, tumor necrosis factor [TNF]-a) would explain increased adhesion molecule expression on affected hair follicles and the non-destructive nature of the disease.

At present, topical immunotherapy with a potent contact allergen such as diphenylcyclopropenone (DCP) appears to be superior to all other modes of treatment [11-14]. Why an allergic contact derma­titis should reverse the pathogenic process and induce hair regrowth is still unknown. According to present knowledge, the initiation and maintenance of an allergic contact dermatitis involves a cascade of events in which cytokines are supposed to playa crucial role [15] . In particular, IL-1P, TNF-a, IFN-y, IL-2, IL-8, and IL-IO have been demonstrated to be preferentially expressed in allergic contact dermatitis [16-19].

If it is true that AA is the result of a distinct cytokine pattern, we

0022-202X/94/S07.00 Copyright © 1994 by The Society for Investigative Dermatology, Inc.

530

VOL. 103, NO.4 OCTOBER 1994

Table I. List of Primer Pairs Used for Cytokine mRNA Amplification, with Data Regarding the Annealing

Temperature (TmOC), the Size of the PCR Products, and the PCR Cycles Within a Linear Range for Amplification

Primer Tm(oq bp Cycles

fJ-actin 60 838 22 IFN-y 63 427 36 IL-lfJ 60 802 28 IL-2 63 458 38 IL-4 60 465 40 IL-6 60 628 33 IL-8 65 289 31 IL-I0 65 432 33 TNF-a 60 444 28

can infer that the beneficial effect of contact allergy is mediated by locally secreted cytokines neutralizing this noxious pattern. To find an answer to this hypothesis, we performed semiquantitative polymerase chain reaction (PCR) for a panel of cytokines in scalp biopsies obtained from healthy controls and patients with AA be­fore and after successful treatment with DCP. A differential expres­sion of distinct cytokine patterns, implying the possibility of a novel therapeutic approach, will be described.

MATERIALS AND METHODS

Patients With informed consent, excisional scalp biopsies were taken. As controls, scalp biopsies obtained from five patients (31 -65 years of age, mean 41 years) undergoing elective neurosurgery were used. They had no inflammatory skin disease, no systemic disease, and no immunosuppressive therapy. Biopsies were taken from 12 patients (25 - 55 years of age, mean 36 years) with alopecia areala of the totalis type. After skin biopsy, sensitization was induced by painting an area of25 cm2 of the patient scalp with 2% DCP. Subsequently, elicitation and maintenance of a contract dermatitis were accomplished by weekly applications with DCP concentrations varying from 1 % to 0.0000001 %. When a response in the form of hair regrowth after unilateral treatment was noted (in 10 patients, approximately 8-18 weeks after the initiation of the contact dermatitis), additional skin biopsies were taken 24-36 h after the last DCP application.

Chemicals Guanidinium thiocyanate, Tris-borate, glycerol , bromo­phenol bluc, Tris-HCI, and NaCINa2H2 ethylencdiaminetetraacetic acid (EDT A} were purchased from Sigma (Dcisenhofcn, Germany), and phenol, isopropanol, and chloroform from Merck (Darmstadt, Gcrmany). The taq polymerasc and the kit for reverse transcription (RT kit) camc from Phar­macia (Uppsala, Sweden). Ethidium bromide and the DNA-VIII molecular weight marker came from Boehringer Mannheim (Manuheim, Germany). Primers for PCR amplification of fJ-actin, IFN-y, IL-l/l, IL-2, IL-4, IL-6, IL-8, and TNF-a were bought from C lontech (Palo Alto, CAl. Primers for IL-I0 were purchased from Stratagene (Heidelberg, Gcrmany). These primer pairs arc all intron-spanning to avoid amplification of contaminating genomic DNA.

Extraction of Total RNA From Excisional Scalp Biopsies Total RNA from scalp skin was isolated according to Chomczynski and Sacchi (20). Twenty milligrams wet weight scalp skin was frozen in liquid nitro­gen, minced, resuspended in guanidinium thiocyanate followed by addition of acidified phcnol/ch loroform. After centrifugation, the RNA was precipi­tated from the aqueous phase with isopropanol. Individual RNA samples from different paticnts were equalized with regard to their RNA concentra­tion by ultraviolet (UV) spectroscopy. The RNA was dissolved in RNasc­free sterile water. Approximately 10,ug total RNA was isolatcd from each preparation.

PCR Analysis of Cytokine mRNA and p-Actin mRNA About 1 ,ug of total RN A from biopsies was reverse transcribed with random hexamer primers and moloney mnrine reverse transcriptase tumor virus. Each primer pair was tested on total RNA and subsequent PCR amplification without prior reverse transcription. No PCR products werc obtained in controls. For each primer pair a three-temperature-stcp PCR cycle program waS carried out (2 min annealing, 2 min extension at 72°C, and 1 min denaturation at 94 0 C; Table I). This PCR setup was optimizcd to obtain only onc PCR product of the expected length (Table I) . Standard curves for the amount of RNA and cycle numbers were establishcd as described [21J. The specificity of each PCR product was demonstrated by comparison of expected PCR

CYTOKlNE mRNA LEVELS IN ALOPECIA AREATA 531

product length with the actual Icngth of the amplified DNA visualized by gel electrophoresis, comparison with the given positive controls provided by CIon tech, and by restriction analysis of the amplified DNA fragments. Semiquantitative analysis of rhe PCR amplification products was achieved by high-performance liquid chromatography (HPLC) [21]. The specific PCR products were recognized by their elution time in comparison to a DNA standard. Before calculation of rhe cytokine mRNA expressions, the probes were equalized for /l-actin mRNA as an internal standard. This was achieved by coall1pl ification of the /l-actin Il1RNA to each cytokinc mRNA followed by HPLC analysis. The mean peak area for each cytokine in 10 patients/five controls was calculated. Constitutive cytokine expression levels in thc controls were set to zero.

HPLC Operating Conditions As shown previously [21 ], semiquantita­tive analysis of PCR products by HPLC allows a precise calculation of the amount of PCR-amplified transcripts. Opcrating conditions and buffers were as described [21].

Gel Electrophoresis For documcntation of the different cytokine PCR products, thc resulting PCR products were precipitated with ice-cold eth­anol and ccntrifuged for 20 min . The remaining DNA was dissolved in20 ,ul Tris-borate-EDTA buffer and loaded in an alkaline 1 % agarose gel [22]. The gels were stained with 0.1 ,ug/1l11 ethidium bromide in TBE bufFer and visualized under UV light (data not shown).

RESULTS

The PCR primers for the different cytokines were shown to specifi­cally amplify the expected cytokine mRNA. All primer pairs used resulted in one PCR product of the expected length and expected fragments after restriction analysis. Using semiquantitative PCR technology by HPLC analysis of PCR products, a reliable quantifi­cation through integration of the relevant peak area is possible; a typical chromatogram of amplified /l-actin cDNA is shown in Fig 1. The peak areas of the corresponding PCR products are propor­tional to the amount of DNA loaded onto the analytical column, allowing precise quantitation. In Fig 1, a typical HPLC run for an 838-base pair (bp) PCR product for /l-actin is shown. The PCR amplification profile of each primer pair was estimated empirically by analysis of increasing PCR with equal amounts of reverse tran-

E 0.02 c:

0 838 bp '" ~ " u c:

'" Primers -€ 0

1: « 0.01

5 10 15

Relention lime [min]

Fignre 1. Example of semiquantitative HPLC analysis ofRT-PCR product. Approximately 1 ,ug of total RNA was subjected to RT-PCR/ HPLC for /l-actin mRNA in scalp biopsies. The chromatogram shows a typical HPLC run of the completed PCR. Besides dNTPs and primers, a single peak corresponding to the 838-bp reverse-transcribed and amplified /l-actin mRNA is found. Quantification of the PC R products is done through peak area integration. The peak areas are proportional to the amount of DNA loaded onto the analytical column. As the elution time depends on the overall charges of DNA molecules, PCR products of difFer­ent length can be differentiated.

532 HOFFMANN ET AL

scribed RNA. PCR cycles ultimately chosen for the PCR were in the linear range for each primer pair (Table I).

Using our experimental approach we detected increased steady state mRNA levels for IL-1P, IL-2, and IFN-y in comparison to the constitutive mRNA expression of these cytokines in intact scalp from healthy donors (Fig 2) . Transcripts for IL-4 were found nei­ther in patients nor in healthy donors. Furthermore, transcripts for IL-6, IL-8, IL-10, and TNF-a mRNA were present in all samples, but the expression pattern in scalp specimens from patients with AA was nearly similar to those detected in scalp biopsies from healthy donors. As the effectivity for PCR amplification is not necessarily equal for different primer pairs, a quantitative comparison of differ­ent primers is not possible. Thus, we cannot say which cytokine is maximally expressed in AA.

Subsequently, one-half of the scalp from patients with AA was painted with DCP, as described in Materials and Methods. After uni­lateral hair regrowth was noted, additional scalp biopsies were taken 24 - 36 h after DCP application. The experimental design and se­miquantitative analysis of the different cytokines was performed in the same way as described above. Compared to control scalp, in successfully DCP-treated AA we found increased mRNA steady state levels for IL-2, IL-8, IL-10, and TNF-a (Fig 3). Again no transcripts for IL-4 were found, whereas IL-6 mRNA expression was low. IFN-y and IL-1P transcripts were increased when com­pared to controls, but compared to untreated AA, the overall ex­pression levels were 60% and 15% reduced for IFN-y and IL-1P, respectively (compare Figs 2 and 3).

DISCUSSION

To explain the pathogenesis of AA, it has been assumed that T lymphocytes surrounding and infiltrating the hai r bulb trigger an autoimmune process, directed against some so far unidentified growth-controlling factor(s) or their receptor(s) [23]. As affected hair roots are rather "switched off" than destroyed, the inhibitory effect on the hair cycle in AA can be best explained by soluble mediators released from lesional T lymphocytes.

Here we provide experimental evidence that cytokines partici­pate in the pathogenesis of AA. We detected increased lesional levels of IFN-y, IL-2, and IL-1P transcripts in scalp biopsies from patients with untreated AA, corresponding to a cytokine pattern of the T H 1 T -helper type. This pattern has been shown to mediate

1500

to ~ ..: 1000 to

'" a. c: to

'" ~ 500

0

o Controls

~ Alopecia (/rt' tlfil untreated

C; <- en. « z I !::: ..J en.

0

'" .... '" 00

..J ..J ..J ..J d

Figure 2. Cytokine mRNA expression in alopecia areata as compared to healthy controls. One microgram total RNA aliquots from patients and controls were reverse transcribed and subjected to PCR amplification for the indicated mediators. The expression of p-actin mRNA was always coampli­fied to serve as an internal standard. Expression levels for the various cyto­kines in the different skin samples were calculated by equalizing the samples to their p-actin mRNA expression, and by setting the constitutive cytokine mRNA expression in healthy control scalp to zero. Presented data are mean values (error bars, ± SD) of the expression levels from 10 patients and five controls. Transcripts for IL-4 were not detected.

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

o Controls

1500 _ Alopecia areaW DCP·trcatcd

to <> ~ 1000

.><: to .., a. c: to .., ~ 500

o nIll r r r rII r r r 00

d

Figure 3. Cytokine mRNA expression in alopecia areata after treat­ment with DCP. The experimental design was the same as described for Fig 2. Biopsies were taken after unilateral hair regrowth was noted after DCP treatment. Presented data are mean values (error bars, ± SD) of the expression levels from 10 patients and five controls.

delayed-type hypersensitivity responses [24]. Following this line of thought IL-2 would promote activation and clonal expansion of lesional T lymphocytes . Furthermore, our findings offer an expla­nation for the aberrant expression of molecules such as ICAM-l and HLA-DR on affected hair follicles in AA. Our results confirm previous studies demonstrating the presence of IFN-y and IL-2 within the peribulbar T-cell infiltrate [25].

Topical immunotherapy is effective in AA, but many questions regarding this treatment are still unanswered. The fact that similar results can be obtained with contact allergens of different structure indicates that the therapeutic effect is not due to a specific pharma­cologic mechanism but to an immune response common to different allergens [26] . An initial step to provide evidence that the therapeu­tic effect may be ascribed to topical "immunomodulation" was achieved by studies in which the ratio of peribulbar CD4+ /CD8+ T cells was shown to decrease from 4: 1 to nearly 1 : 1 [27]. Moreover, the aberrant expression of HLA-ABC and HLA-DR molecules on affected hair follicles was shown to be down-modulated after treat­ment with DCP [28], compatible with the concept of antigenic competition [29].

As we detected increased mRNA-levels for IL-2, IL-8, IL-I0, and TNF-a in DCP-treated AA, we suggest that the beneficial effect of contact allergens are due to immunomodulation by pro-inflamma­tory cytokines released by keratinocytes or lesional T cells. Our findin gs are supported by several reports concerning the mediators present during initiation and amplification of a contact allergy, demonstrating a preferential expression ofTNF-a, IFN-y, IL-1P, IL-8, and IL-10 in lesional skin [1 6- 19]'

At present we cannot define the mediator(s) responsible for the therapeutic response. However, IFN-y has been shown to inhibit IL-4 -mediated T H2-type responses [30], whereas IL-4 was shown to reduce T H 1-type responses [31]. More recently IL-10 was charac­terized by its ability to inhibit T H 1 cytokine production [32]. Several lines of evidence, however, show a more pleiotropic function of IL-IO as a modulator of immune responses [33,34]. Several docu­mented immunologic effects after DCP treatment of AA may be ascribed to the presence of IL-1 ° in basal keratinocytes [35] and in their vicinity. W e observed a reduced expression level of IFN-y in lesional AA scalp after treatment with DCP. We do not know whether this is of clinical significance, but IL-I0 has been shown to inhibit IFN-y-mediated responses [36] and this may explain the decreased expression of molecules such as ICAM-1 and HLA-DR in DCP-treated AA, although we cannot exclude an immunomodu­lating action of IL-8 and TNF-a as well. At present we cannot say what happens in unsuccessfully DCP-treated AA but it has been

VOL. 103, NO.4 OCTOBER 1994

demonstrated [5] that iCAM-l expression is still present in lesional hair bulbs; this might reflect a different immune response after DCP treatment in those patients.

In conclusion, we provide evidence that a cytokine pattern of the T H 1 type participates in the pathogenesis of AA. From the detection of IL-I0 transcripts in DCP-treated AA we infer that IL-IO inhibits the T H 1 response. We cannot say whether this suppressive effect is directed against lesional T cells or against antigen-presenting cells (APC), but possibly IL-IO-dependent inhibition of APC function is a crucial mechanism [37] . Theoretically, application of IL-10 should provide a therapeutic approach for AA. As a potent immune response modifier ofT H 1 T -cells, recombinant IL-l 0 might be ben­eficial when applied topically. If so, this approach would be more specific than application of a potent contact allergen.

This work was supported by a grallt frOI1J the Delltsche Forscnllllgsgel1Jeinschajt (Ho 1598/1-1), BOIJIJ, Gerl/Jatly.

REFERENCES

1. Perret C, Wiesner-Menzel L, Happle R: Immunohistochemical analysis ofT-cell subsets in the peri bulbar and intrabulbar inliltrates of alopecia arcata. Acta Dermatol Vellereol (StacH) 64:26-30, 1984

2 . Baadsgaard 0, Lindskov R, C lemmensen 0]./11 sillllymphocyte subsets in alope­cia arcata before and during treatment with a contact allergen. eli" Exp Derma­tol l 2:260 - 264, 1987

3 . Bystryn J C, Tamesis J: Immunologic aspects of hair loss. ] ltwest D ermatol 96:88S - 89S, 1991

4. Valsecchi R,lmberti G, Martino D, Cainelli T: Alopecia arcata and interleukin-2 receptor. Dermatology 184:126-128,1992

5 . Shapiro J, Tan J, Ho V, Abbott F, Tron V: Treatment of chronic severe alopecia areata with topical diphenylcyclopropenone and 5% minoxidil: a clinical and immunopathologic evaluation.] Am Acad Damatol 29:729-735,1993

6. Nickoloff BJ, Griffiths CEM: Aberrant intercellular adhesion molecule-l (ICAM-l) expression by hair-follicle epithelial cells and endothelial leukocyte adhesion molecule-l (ELAM-l) by vascular cells arc import.1nt adhesion-mole­cule alterations in alopecia arcata.] ltlVest DermatoI96:9 1S -92S, 1991

7. Hamm H, Klemmer S, Kreuzer I, Steijlen PM, Happle R, Brocker E: HLA-DR and HLA-DQ antigen expression of anagen and telogen hair bulbs in long­stand ing alopecia arcata. Arch DermaJoI Res 280:179-181,1988

8. Khoury EL, Price VH, Greenspan JS: HLA-DR expression by hair follicle kerati­nocytes in alopecia arcata: evidence that it is secondary to the lymphoid infiltra­tion.] 1,lVest Dermato!90:193-200, 1988

9. Messenger AG, B1eehen SS, Slater DN, Rooney N : Expression of HLA-DR in hair follicles in alopecia areata . L.,ICet2:287 -288,1984 .

10. Messenger AG, Slater DN, B1eehen SS: Alopecia arcata: alterations in the halT growth cycle and correlation with the follicular pathology. Br] Dennatol 114:337 -347,1986

11. Happle R, Kalveram KJ, Buchner U , Echternacht-Happle K, Gtiggelmann .W, Summer KH: Contact allergy as a therapeutic tool for alopecia areata: appilca­tion of squaric acid dibutylester. Dermatologica 161 :289 - 297, 1980

12. Perret CM, Happle R. Treatment of alopecia areata. In: Orfanos CE, Happle R (eds.). Hair alld Hair Diseases. Springer-Verlag, Berlin, Germany, 1990, pp 571 - 587

13. Steen PHM van der, Baar HJM van, Perret CM, Happle R: Treatment of alopecia arcata with diphenylcyclopropenone.] Am Aead Dermotol 24:253-257, 1991

14. Steen PHM van der, Boezeman JBM, Happle R: Topical immunotherapy for alopecia arcata: re-evaluation of 139 cases after an additional follow-up penod of 19 months. Dermatology 184:198-201,1992

CYTOKINE mRNA LEVELS IN ALOPECIA AREATA 533

15. Cher DJ, Mosmann TR: Two types of murine helper T cell clones. II. Delayed­type-hypersensitivity in mediated Thl clones. ] IlIImllllOl 138:3688-3694, 1987

16. Griffiths CEM, Barker JNWN, Kunkel S, N ickoloffBJ: Modulation ofleucocyte adhesion molecules, a T-cell chemotaxin (IL-8) and a regulatory cytokine (TNF-a) in allergic contact dermatitis (rhus dermatitis). Br ] Dermatol 124:519-526, 1991

17. Enk AH, Katz S: Early molecular eventS in the induction phase of contact sensitiv­ity. Proc Nael Acad Sci USA 89:1398- 1402, 1992

18. Piguet PF, Grau GE, Hauser C , Vassalli 1': Tumor necrosis factor is a critical mediator in hapten-induced irritant and CQnt.1.ct hypersensitivity reactions. } Exp Med 173:673-679, 1991

19. Wei L, Muller M, SauratJH, Hauser C: Lymphocyte profiles in contact sensitivity induced by dinitroAuorobenzene and tolerance by dinitrothiocyanobenzene. Arch Derlllatol Res 284:427-431,1993

20. Chomczynski P, Sacchi N: Single-step method of RNA isolation by acid guanidi­nium thiocyanate-phenol-chloroform extraction. A llal Binehem 162:156-159, 1987

21. Henninger HP, Hoffmann R, Grewe M, Schulze-Specking A, Decker K: PuriJi­cation and quantitative analysis of nucleic acids by anion-exchange high-per­fonnance liquid chromatography. Bioi Chelll H oppe-Seyler 274:625-634,1993

22. Maniatis T, Sambrock J, Fritsch EF: Molewlar C!ollillg, a Laboratory Manllal, 2nd Ed., Cold Spring Harbor Laboratory Press, New York, 1989, pp 7.39-7.41

23. Baadsgaard 0: Alopecia arcata: an immunologic disease? ] ltwest D erlllatol 96:89S-90S, 1991

24. Fong TAT, Mosmann TR: The role of IFN-y in delayed-type hypersensitivity mediated by Thl clones.] 1111111111101143:2887 -2893,1989

25. Gollnick H, Orfanos CE: Alopecia arcata: pathogenesis and clinical picture. In: Orfanos EC, Happle R (cds.). Hair alld Hair Diseases. Springer-Verlag, Berlin, 1990, pp 529 - 569

26. Happle R, Echternacht K: Induction of hair growth in alopecia areat., with D.N.C.B. La'ICet 2:1002-1003, 1977

27. Happle R, Klein HM, Macher E: Topical immunotherapy changes the composi­tion of the peribulbar infiltrate in alopecia arcata. Arch Dennato! Res 278:214-218,1986

28. Brocker Ell, Echternacht-Happle K, Hanlln H, Happle R: Abnonnal expression of class I and class II major histocompatibility antigens in alopecia arcata: modulation by topical immunotherapy. ] Illvest DerlllatoI 88:564-568, 1987

29. Happle R: Antigenic competition as a therapeutic concept for alopecia arcata. A rell Dermatol Res 267:109-1 14, 1980

30. Gajewski TF, Fitch FW: Anti-proliferative efFects oflFN-y in murine regulation. ] ImmllnoI140:4245 - 4252, 1988

31. Mosmann TR, Cherwinski H, Bond MW, Giedlin MA, Coffman RL: Two types of murine T cell clone.] TIllIllIltIOI1 36:2348-2357, 1986

32. Fiorentino DF, Bond M W, Mosmann TR: Two types of mouse T helper cflllV. Th2 clones secrete a factor that inhibits cytokine production by Th 1 clones.] Exp M ed 170:2081 - 2095, 1989

33. Mosmann TR, Schumacher JH, Fiorentino DF, Leverah J, Moore KW, Bond MW: Isolation of monoclonal antibodies specific for IL-4, IL-5, IL-6, and a new Th2-specific cytokinc (lL-l0, cytokine synthesis inhibitory factor) by using a solid-phase radioimmunoadsorbent assay.] Il1ltn.llloI14S:2938 -2945, 1990

34. Waal Malefyt de R, Abrams J, Bennet B, Figdor C, Vries JE: Interleukin 10 (IL- tO) inhibits cytokine synthesis by human monocytes: An autoregulatory role of1L-10 produced by monocytes.] ExpMed 174:1209- 1220,1991

35. Enk AH, Katz S: Identification and induction of kcratinocyte-derived lL-l0.] Imm'lIIoI149:92 - 95 , 1992

36. Fiorentino DF, ZI'Jtnik A, Vieira P, Mosmann TR, Howard M, Moore KW, O'GarTa A: IL-l0 acts on the antigen-presenting cell to inhibit cytokine pro­duction by Thl celis.] Tmm'lIIoI146:3444-3451, 1991

37. Enk AH, Angeloni VL, Udey MC, Katz SI: Inhibition of Langer hans cell anti gen­presenting function by IL-l 0.] Itnm'lIIo!l 5 1 :2390 - 2398, 1993