Interleukin-4 induces apoptosis in cultured human follicular keratinocytes, but not in dermal papilla cells

ARTICLE

Hair growth is a highly regulated, rhythmical biological process in which each hair follicle goes through successive periods of growth (anagen) and spontaneous bulb regression (catagen), followed by a resting phase (telogen). This cycling process implies that a large portion of follicular keratinocytes (FK) is subject to apoptotic phenomena, whereas another portion remains and responds to the signals which reinitiate the next cycle [1]. Based on morphological studies of apoptosis in the human scalp [2, 3] and standard biochemical analysis of catagen formation in mice [4, 5], apoptosis of follicular keratinocytes seems to play a major role in proximal hair bulb regression observed during catagen. Yet, the exact control of follicular keratinocyte apoptosis remains obscure.

Cytokines and growth factors are increasingly believed to play an important role in controlling hair growth processes. Recently, whole-organ culture methods showed that hair follicle growth in vitro is inhibited by pro-inflammatory cytokines such as interleukin-1alpha (IL-1alpha), interleukin-1beta (IL-1beta) and tumor necrosis factor alpha (TNFalpha) [6-8], however, these molecules, also known as Th1 cytokines, are unlikely to regulate hair growth under physiological conditions. It would appear that cytokines involved in natural hair follicle cycling should induce apoptosis without provoking inflammation. Furthermore, we assume that anti-inflammatory agents could take part in hair follicle cycling. Among them, transforming growth factor (TGF) beta is expressed in late anagen, immediately before catagen [9] and interleukin-4 [10] (IL-4), a pleiotropic Th2 cytokine, is secreted e.g. by mononuclear cells [11], which are found around the hair follicle during late anagen [12], indicating a possible interaction between the hair follicle and its surrounding cells during catagen formation.

In this study we aimed to identify cytokines and growth factors which may participate in physiological catagen formation. We analyzed the effects of cytokines and growth factors of the Th1 type known to inhibit hair follicle growth on cultured human follicular keratinocytes and dermal papilla cells (DPC), in comparison to IL-4 (Th2). DPC are believed to be a long term survivor cell population, since no apoptotic phenomena could be detected in DPC during hair follicle cycling in previous studies in vivo [2-5]. However, staurosporine, a protein kinase C (PKC) inhibitor, was shown to induce apoptosis in DPC in vitro [13]. Thus, staurosporine was used as a positive apoptotic control in DPC and its effect on FK was explored.

For further understanding, since apoptotic phenomena are seemingly controlled by an interplay between anti- and pro-apoptotic proteins such as bcl-2 and bax [14, 15], we investigated the expression of bcl-2 and bax RNA in human cultured hair follicle cell populations under culture conditions with and without incubation with IL-4.

Materials and methods

Chemicals

Cytokines: basic fibroblast growth factor (bFGF), epidermal growth factor (EGF), insulin-like growth factor (IGF 1), interleukin (IL)-1alpha, IL-1beta, IL-2, IL-6, IL-8, IL-10, IL-12, beta nerve growth factor (betaNGF), TGFbeta1 and vascular endothelial growth factor (VEGF) were obtained from R&D Systems, Minneapolis, MN, USA; IL-4 and TNFalpha from TEBU Pro Tech Inc., Rocky Hill, NJ, USA; interferon gamma (IFNgamma) from Sigma, St. Louis, MO, USA and staurosporine from Sigma, Deisenhofen, Germany.

Isolation and culture of human follicular keratinocytes

FK were established from 30-50 plucked human terminal hair follicles from the occipital and frontal region of healthy volunteers. The distal hair shafts were subsequently cultivated in serum containing culture medium (DMEM/HAM's F12 1:1, supplemented with 10% FCS, 100 IU penicillin/streptomycin, 10 ng/ml epidermal growth factor (EGF), 5 mug/ml keratinocyte growth factor, 0.08 mug/ml cholera toxin, 0.4 mug/ml hydrocortisone; all from Biochrom), as described by Detmar et al. [16]. When keratinocyte outgrowth was first observed (Fig. 1A), the serum-containing medium was switched to serum-free keratinocyte medium containing 0.1-0.2 ng/ml recombinant EGF and 25 mug/ml bovine pituitary extract (Gibco, Grand Island, NY, USA). When keratinocyte outgrowth ranged up to a diameter of 1 cm, FK primary cultures were subcultured using 0.1% trypsin, 0.02% ethylenediaminetetraacetic acid solution and propagated to first passage in serum-free keratinocyte medium in 2.8 cm² (1 x 10³ cells/well, 24-well, Nunclon^TM, Nunc, Wiesbaden, Germany) or 0.38 cm² culture dishes (1 x 10² cells/well, chamber slides^TM, Nunc, Wiesbaden, Germany) in a humidified 5% CO₂ atmosphere. Sixty to 70% confluent cells were treated with cytokines, growth factors or staurosporine in fresh serum-free keratinocyte medium.

Isolation and culture of human dermal papilla cells

DPC were isolated from occipital scalp fragments of healthy patients undergoing face lifting, by microdissection [17, 36] and cultivated in Dulbecco's modified Eagle's medium (DMEM) containing 20% fetal calf serum (FCS), 100 IU/ml penicillin/streptomycin, 0.4 mM L-glutamine and 50 ng/ml amphotericin B (all from Biochrom, Berlin, Germany) in a humidified atmosphere with 5% CO₂ at 37 °C (Fig. 1B). After the first 5 days, the medium was changed every third day. Confluent cultures were subcultured in a 1:3 split after 8 weeks using 0.1% trypsin, 0.02% ethylenediaminetetraacetic acid solution and propagated in culture medium as described above containing 20% FCS. For experiments DPC (1 x 10³ cells/well) were seeded in 2.8 cm² culture dishes in culture medium containing 10% FCS. At a confluence of 60-70% fresh medium supplemented with cytokines, growth factors or staurosporine was added.

Incubation with cytokines, growth factors and staurosporine

FK were used only in the first passage, whereas DPC were used from passages 2 to 6. For cell death detection ELISA, FK and DPC (1 x 10³ cells/well) were grown in 2.8 cm² culture dishes (Nunc, Wiesbaden, Germany) as described above. TUNEL experiments were performed with FK grown on 0.38 cm² culture dishes (chamber slides^TM, Nunc, Wiesbaden, Germany) at a concentration of 1 x 10² cells/well. After cells had been grown to 60-70% confluence, cytokines, growth factors or staurosporine were added to each well in fresh medium at the concentrations shown in Tables I and II and left for 24 hrs before quantifying apoptosis and necrosis. IL-4 was tested on FK also for 3 and 18 hrs to measure apoptosis and necrosis. Concentrations were taken according to previous investigations on cytokines and staurosporine [6-8, 13, 18, 19, 30, 41, 43, 44].

Detection of apoptosis

After incubation with tested agents the enrichment of mono- and oligonucleosomes in the cytoplasm of apoptotic cells due to DNA degradation was detected by a photometric enzyme immunoassay measuring cytoplasmic histone-associated DNA fragments (cell death detection ELISA^PLUS®, Boehringer Mannheim, Mannheim, Germany) according to the manufacturer's protocol. Absorbance was measured at A_{405 nm} - A_{490 nm} and was calculated as the mean ± SEM of at least 3 treated wells (n >= 3) of containing FK or DPC from one donor and compared to the mean ± SEM of medium control. Mean absorbance values of each treatment (cytokine, growth factor or staurosporine) were compared to medium control and significance was calculated using student's t-test. The specific enrichment of apoptotic cells versus control was demonstrated as enrichment factor according to the formula given by the manufacturer:

absorbance at the sample (dying/dead cells) / absorbance of the corresponding control (cells without treatment)
= enrichment factor (EF)

In addition, apoptotic cells were reevaluated using an established TUNEL kit (ApopTag, Oncor, Gaithersburg, MD, USA). They were pretreated with the same panel of cytokines on chamber slides^TM and subsequently fixed with formalin (1%) for 10 min at room temperature, washed in two changes of PBS and post-fixed in ethanol/acetic acid (2:1) for 5 min at - 20° C. Slides were washed twice with PBS as described above, covered with equilibration buffer for 5 min, and incubated with TdT solution for 1 hr at 37° C in a humidified chamber. The reaction was terminated with stop/wash buffer (30 min at 37 °C). Digoxigenin-dUTP-labeled DNA compounds were detected by anti-digoxigenin fluorescein-isothiocyanate (FTIC)-conjugated F(ab)₂ fragments. Dilutions used for TUNEL staining were adopted from the manufacturer's manual, and negative controls for the TUNEL were made by omitting TdT.

Detection of necrosis

After incubating each cytokine, growth factor or staurosporine for 24 hrs the necrotic effect was analyzed based on the release of lactate dehydrogenase (LDH) activity into the supernatant by damaged DPC or FK (cytotoxicity detection kit, Boehringer Mannheim, Mannheim, Germany). Cells were centrifuged (2,000 rpm, 10 min); then 100 mul of the supernatant were transferred to an multititerplate, and 100 mul of the incubation mix were added according to the manufacturer's protocol. Absorbance was measured at 490 nm after 10 and 20 min.

Total RNA isolation

To isolate RNA, 3 x 10⁶ cells were grown in 75 cm² culture dishes. RNA extraction was performed with the use of TRIzol reagent (Gibco, Grand Island, NY, USA) using the following steps: phenol-chloroform-isopropanol-ethanol extraction with additional ethanol precipitation with 0.3 M sodium acetate. The pellet was air-dried, reconstituted in RNAse-free water (USB, Cleveland, OH, USA), photometrically measured and stored at - 20° C.

Reverse transcription-polymerase chain reaction (RT-PCR) of FK and DPC

The RNAs were converted into cDNA with the T-primed first-strand ready-to-go kit from Pharmacia Biotech (Uppsala, Sweden) according to their instructions. PCR was performed with gene-specific, intron-spanning primers at a final concentration of 1 mM dNTPs, 10 muM each of primers and 2.5 Units of Taq DNA polymerase (Perkin Elmer, Heidelberg, Germany) in a total volume of 50 mul. Thermal cycling was carried out as follows: one cycle of 94° C for 1 min followed by 30 cycles of 94° C for 30 sec., 60° C for 1 min, 72° C for 1 min and an additional extension period at 72° C for 5 min. An aliquot of 10 mul of each reaction was separated by electrophoresis using a 1% agarose gel and visualized with ethidium bromide.

The following primer pairs were purchased from TibMolbiol (Berlin, Germany) as we designed. For each set, the upper strand (5') is listed first:

bcl-2 5'-TGG CTC AGA TAG GCA CCC AG-3'

5'-ACG GTG GTG GAG GAG CTC TT-3'

bax 5'-TTC TTC CAG ATG GTG AGC GAG-3'

5'-CGA GTG GCA GCT GAC ATG TTT-3'

beta actin 5'-AGC CTC GCC TTT GCCGA-3'

5'-CTG GTG CCT GGG GCG-3'

Statistical analysis

Paired t-tests were used to compare differences between treated and untreated cells of healthy donors. P < 0.05 was considered statistically significant using a two-tailed hypothesis. Results are expressed as mean ± SEM.

Results

Detection of cell death in FK and DPC

In our experiments, IL-4 significantly induced dose-dependent apoptosis in FK (p < 0.01) at concentrations of 100 ng/ml (Fig. 2). The enrichment of apoptotic cells after treatment with IL-4 for 24 hrs was about 4.69 fold higher versus control. Mean enrichment factor of IL-4 treated FK was 2.8 in all tested individuals (n = 4, Table I). Lower concentrations (10 and 50 ng/ml) did not show a significant effect after 24 hrs treatment (Fig. 2A). These findings were confirmed by positive TUNEL staining of condensed nuclei (Fig. 3B, arrow), indicating that DNA fragmentation had occurred after 24 hrs treatment with 100 ng/ml IL-4, as compared to negative control (Fig. 3A). The induction of apoptosis in FK was time-dependent (Fig. 2B); IL-4 treatment (100 ng/ml) revealed significant results after 24 hrs incubation, but no significant apoptosis was demonstrated after 3 and 18 hrs. Furthermore, IL-4 induced apoptosis in all tested primary FK cultures (n = 4) obtained from four healthy donors (Fig. 4). Other cytokines and growth factors such as IL-1alpha, IL-1beta, IFNgamma, TGFbeta1, and TNFalpha did not induce significant apoptosis in FK (Table I) and no apoptotic nuclei were seen in the TUNEL testing (data not shown). None of the cytokines and growth factors tested on FK induced LDH release into the supernatant analyzed by cytotoxicity ELISA, thus a necrotic effect was excluded.

The PKC inhibitor staurosporine incubated at a concentration of 0.01 and 0.1 muM revealed dose-dependent apoptotic signals in FK after 24 hrs incubation (Table I) and dose-dependent necrosis was found (data not shown). In contrast to FK, none of the selected cytokines and growth factors tested on DPC showed enrichment of apoptotic cells (EF, see Table II). Staurosporine was used as a positive control. Again, 24 hrs treatment of DPC with staurosporine was dose-dependently followed by apoptotic signals in the cell death detection ELISA (Table II) and necrosis in the cytotoxicity detection kit (data not shown).

Reverse transcription-polymerase chain reaction analysis

In a separate series of experiments, polymerase chain reaction was used to assess bcl-2 and bax mRNA expression in cultured FK and DPC both before and after treatment with IL-4. Our findings showed that both FK and DPC expressed bcl-2 and bax under culture conditions, whereas, treatment with IL-4 (100 ng/ml, 24 hrs) did not exert a significant effect on the expression of bcl-2 and bax versus medium control measured by densitometric analysis (Fig. 5). The expression of the housekeeping gene beta-actin was used as an internal control.

Discussion

The ability of FK to undergo apoptosis may play an important role in bulb regression (catagen), however, the mechanisms regulating this process are largely unknown. Since previous investigations had suggested that IL-1alpha, IL-1beta and TNFalpha completely abrogate hair follicle growth [6, 7, 18, 19], we aimed to explore whether the growth-inhibiting effect of these pro-inflammatory cytokines is consistent with induction of apoptosis in FK. Also IFNgamma and TGFbeta, which seemingly do not affect human hair growth [7, 8, 30], were included in this investigation. Interestingly, cytokines of main interest for influencing catagen development differ between hair follicles derived from human donors and mice. In mice, injection of TGFbeta induced catagen formation of hair follicles [44]. Human derived hair follicles incubated with TGFbeta responded with only partial inhibition of growth, furthermore, incubation with TGFbeta did not induce any morphological changes or catagen formation in human hair follicles [7]. These different results in human and mice after treatment with TGFbeta imply that data obtained from mice should be confirmed in human cell cultures.

In our experiments, the tested pro-inflammatory, mainly Th1 cytokines [37] and also TGFbeta1, failed to induce apoptosis or necrosis in cultured human FK at concentration levels tested. We assume that the inhibitory effect of these Th1 cytokines on hair growth is not mediated via apoptosis in FK; the exact mechanisms, therefore, by which IL-1alpha, IL-1beta, IFNgamma, TGFbeta1 and TNFalpha may influence human hair follicle cycling, as suggested by previous authors, remain to be elucidated. In contrast, we found that IL-4, an anti-inflammatory Th2 cytokine [10, 43], was able to induce apoptosis in cultured human FK at a concentration of 100 ng/ml. This finding supports our hypothesis, that cytokines involved in natural hair follicle cycling should induce apoptosis without provoking inflammation. Interestingly, the IL-4 receptor was found recently within whole hair follicles in human skin [42]. Thus, it seems possible that the anti-inflammatory cytokine IL-4 may influence hair follicle regression directly by inducing apoptosis in FK.

An inhibitory effect of IL-4 has been previously shown in human cancer cells in vitro [41, 43]. Patients with alopecia areata responding poorly to topical immunotherapy showed markedly increased levels of IL-4 [46], indicating that resistance to topical immunotherapy might by associated to the predominance of IL-4 induced apoptotis in FK.

Moreover, local application of tacrolimus (FK 506) showed inhibition of IL-4 production followed by protection from chemotherapy-induced alopecia [45]. Since a considerable number of mononuclear cells occur near the hair follicle during late anagen [12], cell to cell communications may take place between mesenchymal and ectodermal tissues during the hair cycle being responsible for bulb regression. IL-4 is secreted by mononuclear cells such as monocytes or CD8⁺ Th2 like T cells [10, 11]. Also, a decrease of CD8⁺ T cells was observed during hair regrowth in mice treated with topical immunotherapy [39]. A decrease of local IL-4 production by CD8⁺ T cells could release FK from an apoptotic stimulus and subsequently may provide for hair regrowth.

In contrast to FK, cultured DPC thought as the non-cycling portion of the hair follicle in vivo [2-5], revealed no apoptotic behavior in vitro under the influence of the tested cytokines and growth factors (Table II). Only staurosporine, a potent PKC inhibitor, used as a positive control in DPC [13] revealed after simultaneous analysis of supernatant concurrent and dose-dependent apoptotic and necrotic effects. This effect of staurosporine on FK was consistent with our results in DPC. Overall, IL-4 and other cytokines reportedly produced by DPC and its surrounding cells during hair follicle cycling in vivo [21-24, 32, 35, 36, 38] do not influence DPC apoptosis, further supporting the hypothesis that DPC are an extraordinarily resistant, long-living cell population in contrast to FK in vivo [3, 4]. Staurosporine, an unphysiological enzyme inhibitor, was able to induce apoptosis in FK and DPC serving as a positive control in both cell types.

Since the apoptotic behavior of different hair follicle cell populations has been correlated to the detection of anti- and pro-apoptotic proteins like bcl-2 and bax by immunohistochemical analysis [4], we studied their expression in FK and DPC under culture conditions by RT-PCR. Interestingly, the expression of bcl-2 and bax did not differ before and after 24 hrs incubation of FK and DPC with IL-4, indicating that gene expression of these molecules may not be essential for IL-4-controlled bulb regression as previously indicated by null mutation experiments [40]. Thus, bcl-2 and bax are seemingly not involved in IL-4 induced apoptosis in FK and the non-apoptotic behavior of DPC. The signaling pathways of IL-4 in FK should be further investigated.

CONCLUSION

Acknowledgements

The work was supported by a research grant from Merck, Sharp and Dohme, USA.

Article accepted on 27/5/02

REFERENCES

1. Cotsarelis G, Sun TT, Lavker R. Label-retaining cells reside in the bulge area of pilosebaceous unit: implications for follicular stem cells, hair cycle and skin carcinogenesis. Cell 1990; 61: 1329-37.

2. Olson RL, Everett MA. Epidermal apoptosis: cell deletion by phagocytosis. J Cutan Pathol 1975; 2: 53-7.

3. Weedon D, Strutton G. Apoptosis as the mechanism of the involution of hair follicles in catagen transformation. Acta Derm Venereol 1981; 61: 335-9.

4. Lindner G, Botchkarev VA, Botchkareva NV, et al. Analysis of apoptosis during hair follicle regression (catagen). Am J Pathol 1997; 151: 1601-15.

5. Matsuo K, Mori O, Hashimoto T. Apoptosis in murine hair follicles during catagen regression. Arch Dermatol Res 1998; 290: 133-6.

6. Jindo T, Imai R, Tsuboi R, et al. The effect of various cytokines on hair growth of mouse vibrissae in organ culture. J Dermatol Sci 1994; 7 (suppl.): 73S-8S.

7. Hoffmann R, Eicheler W, Huth A, et al. Cytokines and growth factors influence hair growth in vitro. Possible implications for the pathogenesis and treatment of alopecia areata. Arch Dermatol Res 1996; 288: 153-6.

8. Philpott M, Sanders D, Bowen J, Kealey T. Effects of interleukins, colony stimulating factor and tumor necrosis factor on human hair follicle growth in vitro: a possible role for interleukin-1 and tumor necrosis factor-alpha in alopecia areata. Br J Dermatol 1996; 135: 942-8.

9. Seiberg M, Marthinus J, Stenn KS. Changes of expression of apoptosis-assoicated genes in skin mark early catagen. J Invest Dermatol 1995; 104: 78-82.

10. Goerdt S, Orfanos CE. Other functions, other genes: alternative activation of antigen-presenting cells. Immunity 1999; 10: 137-42.

11. Clerici M, Fusi ML, Ruzzante S, et al. Type 1 and type 2 cytokines in HIV-infection: a possible role in apoptosis and disease progression. Ann Med 1997; 29: 185-8.

12. Westgate G, Craggs R, Gibson W. Immune privilige in hair growth. J Invest Dermatol 1991; 97: 156-9.

13. Ferraris C, Cooklis M, Polakowska R, Haake A. Induction of apoptosis through the PKC pathway in cultured dermal papilla fibroblasts. Exp Cell Res 1997; 234: 37-46.

14. Hockenbery D, Zoltan NO, Xiao-Ming Y, et al. Bcl-2 functions in an antioxidant pathway to prevent apoptosis. Cell 1993; 75: 241-51.

15. Carson DA, Ribeiro JM. Apoptosis and disease. Lancet 1993; 341: 1251-4.

16. Detmar M, Schaart FM, Blume U, Orfanos CE. Culture of hair matrix and follicular keratinocytes. J Invest Dermatol 1993; 101 (suppl.): 130S-4S.

17. Jahoda C, Horne K, Oliver R. Induction of hair growth by implantation of cultured dermal papilla cells. Nature 1984; 311: 560-2.

18. Hoffmann R, Eicheler W, Wenzel E, Happle R. Interleukin-1beta-induced inhibiton of hair growth is mediated by cyclic AMP. J Invest Dermatol 1997; 108: 40-2.

19. Xiong Y, Harmon CS. Interleukin-1 beta is differentially expressed by human dermal papilla cells in response to PKC activation and is a potent inhibitor of human hair follicle growth in organ culture. J Int Cyt Res 1997; 17: 151-7.

20. Stenn K, Eilertsen K. Molecular basis of hair growth control. J Invest Dermatol 1996; 107: 669-70.

21. Ahmed AA, Nordlind K, Schultzberg M, et al. Immunohistochemical studies of proinflammatory cytokines and their receptors in hair follicles of normal human skin. Acta Derm Venereol 1996; 76: 348-52.

22. Chiu HC, Chang CH, Chen JS, Jee SH. Human hair follicle dermal papilla cell, dermal sheath cell and interstitial dermal fibroblast characteristics. J Formos Med Assoc 1996; 95: 667-74.

23. Danilenko DM, Ring BD, Pierce GF. Growth factors and cytokines in hair follicle development and cycling: recent insights from animal models and the potentials for clinical therapy. Mol Med Today 1996; 2: 460-7.

24. Jahoda CAB. Cellular and developmental aspects of androgenetic alopecia. Exp Dermatol 1998; 7: 235-48.

25. Kock A, Schwartz T, Kirnbauer R, et al. Human keratinocytes are a source of tumor necrosis factor alpha: evidence for synthesis upon stimulation. J Exp Med 1990; 172: 1609-14.

26. Lachgar S, Charveron M, Gall Y, et al. Vascular endothelial cells: targets for studying the activity of hair follcile cell-produced VEGF. Cell Biol Toxicol 1996; 12: 331-4.

27. Little J, Redwood K, Granger S, Jenkins G. In vivo cytokine and receptor gene expression during the rat hair growth cycle. Analysis by semiquantitaive RT-PCR. Exp Dermatol 1996; 5: 202-12.

28. Mahé YF, Buan B, Billoni N, et al. Pro-inflammatory cytokine cascade in human plucked hair. Skin Pharmacol 1996; 9: 366-75.

29. Panaretto BA. Gene expression of potential morphogens during hair follicle and tooth formation. Reprod Fert Dev 1993; 5: 345-60.

30. Paus R, Foitzik K, Welker P, et al. Transforming growth factor-beta receptor type I and type II expression during murine hair follicle development and cycling. J Invest Dermatol 1997; 109: 518-26.

31. Philpott M, Sanders D, Westgate G, Kealey T. Human hair growth in vitro: a model for the study of hair follicle biology. J Dermatol Sci 1994; 7 (suppl.): 55S-72S.

32. Takata Y, Seki S, Dobashi H, et al. Inhibtion of IL-12 synthesis of pe-ripheral blood mononuclear cells (PBMC) stimulated with a bacterial superantigen by pooled human immunoglobulin: implications for its effect on Kawasaki disease (KD). Clin Exp Immunol 1998; 114: 311-9.

33. Terenghi G, Mann D, Kopelman PG, Anand P. TrkA and trkC expression is increased in human diabetic skin. Neurosci Lett 1997; 228: 33-6.

34. Westgate G, Little J, Philpott M, et al. A new model of hair growth regulation. 19th IFSCC Congress, Sydney 1996; 3: 99-113.

35. Blume-Peytavi U. Die Kontrolle des Haarzyklus: Funktion und Bedeutung der dermalen Papille. Z Hautkr 1996; 71: 410-5.

36. Kozlowska U, Blume-Peytavi U, Kodelja V, et al. Expression of vascular endothelial growth factor (VEGF) in various compartments of the human hair follicle. Arch Dermatol Res 1998; 290: 661-8.

37. Kozlowska U, Blume-Peytavi U, Kodelja V, et al. Vascular endothelial growth factor induced by proinflammatory cytokines (interleukin-1alpha, beta) in cells of the human pilosebaceous unit. Dermatology 1998; 196: 89-92.

38. Messenger AG. The control of hair growth: an overview. J Invest Dermatol 1993; 101 (suppl.): 4S-9S.

39. Freyschmidt-Paul P, Sundberg JP, Happle R, et al. Successful treatment of alopecia areata-like hair loss with the contact sensitizer squaric acid dibutylester (SADBE) in C3H/HeJ mice. J Invest Dermatol 1999; 113: 61-8.

40. Veis D, Sorenson JCM, Shutter JR, et al. Bcl-2 defficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented hair. Cell 1993; 75: 229-40.

41. Puri RK, Siegel JP. Interleukin-4 and cancer therapy. Cancer Invest 1993; 11: 479-86.

42. Varricchio F, Husain SR, Leland P, et al. Interleukin-4 receptor expression in vivo on human AIDS-related kaposi's sarcoma. Oncol Res 1997; 9: 495-503.

43. Opal SM, DePalo V. Anti-inflammatory cytokines. Chest 2000; 117: 1162-72.

44. Foitzik K, Lindner G, Mueller-Roever S, et al. Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo. FASEB J 2000; 14: 752-60.

45. De Tran QH, Guay E, Chartier S, et al. Tracolimus in dermatology. J Cutan Med Surg 2001; 5: 329-35.

46. Yoshino T, Asada H, Ando Y, et al. Impaired responses of peripheral blood mononuclear cells to stimulants in alopecia areata patients with a poor response to topical immunotherapy. Br J Dermatol 2001; 145: 415-21.