Wolfgang Bäumer* and Manfred Kietzmann
Department of Pharmacology, Toxicology, and Pharmacy, University of Veterinary Medicine Hannover, Foundation, D-30559 Hannover, Germany
Author for correspondence; telephone: ++49 (0)511 9538732, fax: ++49 (0)511 9538581, e-mail: wolfgang.baeumer@tiho-hannover.de
Running heading: CsA and cilomilast affect keratinocytes
Abstract
The calcineurin inhibitor cyclosporin A and the phosphodiesterase 4 inhibitor cilomilast exhibit potent immunomodulatory properties which makes them interesting therapeutics for the treatment of skin disorders like canine and human atopic dermatitis. Cyclosporin A and phosphodiesterase 4 inhibitors have already demonstrated clinical efficacy in the therapy of canine and human atopic dermatitis. The direct impact on keratinocytes, especially canine keratinocytes, is less obvious. Thus, an investigation was carried out to ascertain as to whether cyclosporin A and cilomilast modulate keratinocyte proliferation and secretion of
proinflammatory mediators. Cyclosporin A inhibited canine and murine keratinocyte
proliferation, whereas cilomilast did not affect the proliferation. Cyclosporin A and cilomilast reduced the lipopolysaccharide induced prostaglandin E2 synthesis in canine keratinocytes.
Both immunomodulators also inhibited the production of the murine CXC chemokine KC and CCL2 in the murine keratinocyte cell line MSC-P5. The two immunomodulators also
significantly reduced the interferon γ induced production of interferon induced protein 10 in human HaCat cells. Thus, cyclosporin A and cilomilast directly modulate keratinocyte functions, which might contribute to the anti-inflammatory and immunomodulatory action of these compounds in the treatment of allergic skin diseases.
Introduction
For decades now, the calcineurin inhibitor cyclosporin A (CsA) has been used for its
immunomodulatory properties in humans after organ transplantation. In human dermatology, CsA is used for the treatment of psoriasis and severe atopic dermatitis (AD). In canine dermatology, CsA has been implemented for several indications and is registered for the treatment of canine AD (Atopica®)1. Some effects of CsA have been described on human or murine keratinocytes. CsA reduces the proliferation of human and murine keratinocytes in a dose dependent manner2;3.
For canine keratinocytes, however, such studies are not obtainable. The influence of CsA on keratinocyte cytokine expression and secretion seems to be rather controversial: one study demonstrated effects on cytokine gene expression4, whereas another study could only observe effects in psoriatic lesions, but not in cultured keratinocytes5. Two further studies did not find inhibitory effects of CsA on chemokine or cytokine secretion in keratinocytes6;7. The
transcription factor nuclear factor of activated T cells (NFAT), which is modulated by CsA, has recently been described to be functionally active in keratinocytes8. This study therefore set out to re-evaluate the impact of CsA on activated keratinocytes.
The highly selective phosphodiesterase 4 (PDE4) inhibitor cilomilast, currently under
investigation for the treatment of chronic obstructive pulmonary disease in humans9, has been tested for its anti-inflammatory/immunomodulatory potential in allergic skin diseases10;11. In past studies cilomilast was seen to significantly reduce the allergic inflammatory response in murine models of allergic contact dermatitis, accompanied by a vast reduction of cytokine concentrations (e.g. interleukin (IL-)1β IL-4, IL-6 and macrophage inflammatory protein 2) and inflammatory cell influx in challenged skin10;12. Further, it was also demonstrated that dendritic cell function (migration through skin and cytokine secretion) was modulated by cilomilast13.
The above cited studies indicate that cilomilast might be suitable for the treatment of allergic skin diseases such as human and canine AD. No data are available concerning direct effects of the PDE4 inhibitor cilomilast on keratinocyte functions. It has been demonstrated, however, that human keratinocytes express high amounts of PDE4 and might therefore be a target for PDE4 inhibitors such as cilomilast14.
Materials and methods
Canine keratinocytes were isolated and cultured as previously described 15 with slight modifications. Briefly, healthy skin was obtained from dogs undergoing surgery, cut into pieces (5 mm2) and incubated in William´s E medium (Sigma, Deisenhofen, Germany) with 10 mg/ml trypsin (Sigma) for 24 h. The epidermis was stripped off the dermis and epidermis was incubated with trypsin/EDTA (0.05%/0.02%, Biochrom, Berlin, Germany) at room temperature for 30 min. (softly stirred suspension). The suspension was sieved, centrifuged at 1,000 g for 10 min. and the keratinocytes were seeded into culture flasks containing
William´s E medium supplemented with 2 mmol/l glutamine, 20% foetal calf serum (FCS), 10 ng/ml Epidermal Growth Factor (EGF), 100 µg/ml streptomycin and 100 U/ml penicillin (Biochrom, Berlin, Germany). For sub-cultivation (passage 5 to 10 were used for the
described experiments), foetal calf serum was reduced to 10 %. A possible contamination with fibroblasts was excluded by microscopic evaluation.
The murine keratinocyte cell line MSC-P5 (Cell Lines Service, Eppenheim, Germany) has been used e.g. for cancer research studies16. MSC-P5 cells were cultured in RPMI 1640 medium supplemented with 10% FCS, 100 µg/ml streptomycin and 100 U/ml penicillin (Biochrom).
The human keratinocyte cell line HaCat was a kind gift from Dr. N.E. Fusenig,
German Cancer Research Centre, Heidelberg, Germany. HaCat cells were cultured in DMEM medium, being supplemented with 10% FCS, 100 µg/ml streptomycin and 100 U/ml
penicillin (Biochrom).
Proliferation studies
Keratinocytes were seeded into 96 well plates (10 000 cell/well). After 4 h these cells were treated with cilomilast (elbion, Radebeul, Germany) and CsA (elbion) with concentrations as indicated (1- 10 µmol/l) or vehicle (DMSO, 0.1 %). 48 h after treatment the proliferation was analysed by BrDU incorporation (Amersham, Freiburg, Germany). It was determined in pilot studies that maximum proliferation is achieved 48 h after seeding. Cells were labelled with BrDU 16 h before fixation and the ELISA was performed according to manufacturer`s protocol.
In order to exclude an artefact due to loss of cell viability, confluent grown cells were
incubated with cilomilast and CsA (1-10 µmol/l) for 48 h and cell viability was determined by CellTiter© AQueous One Solution cell proliferation assay (Promega, Mannheim, Germany).
Three independent experiments were performed.
Stimulation with lipopolysaccharide (LPS), peptidoglycan (PGN) and interferon γ (IFNγ)
Canine and murine keratinocytes cultured from passage 5 to 10 were seeded in 96-well-plates and grown until confluence occurred. Cells were pre-incubated with cilomilast and CsA (1-10 µmol/l) for 90 min., and then stimulated with 100 µg/ml LPS (E. coli; O111, B4, Sigma).
Supernatants were harvested 48 h after stimulation. Prostaglandin E2 (PGE2) was measured with ELISA (R&D systems, Wiesbaden, Germany) according to manufacturer´s protocol.
Additionally, murine keratinocytes were also stimulated with PGN (10 to 50 µg/ml, Sigma).
The stimuli were also added 90 min. after incubation with cilomilast or CsA. The supernatants were harvested 24 h after stimulation. CCL2 (murine monocyte chemo-attracting protein 1, MCP-1) and the CXC chemokine KC were measured by means of ELISA (R&D systems, Wiesbaden, Germany) according to manufacturer´s protocol.
HaCat cells were treated with 50 ng/ml rhIFNγ (Biozol, Eching, Germany) 90 min. after incubation with cilomilast or CsA. 24 h after stimulation supernatants were taken for determining IP-10 (R&D systems). Every stimulation was performed at least in three independent experiments (n = 4 to 6 samples were taken for each group).
Statistical evaluation
Figures are presented as mean (± SD). Statistically significant differences in
cytokine/chemokine concentrations between the drug treatments and controls were assessed by a one-way ANOVA followed by a post-hoc test (Dunnett’s test).
Results
Proliferation studies
In order to determine the effects of CsA and cilomilast on the growth of keratinocytes, canine and murine keratinocytes were incubated for 48 h with various doses of CsA and cilomilast (0.1 – 10 µmol/l). Cilomilast had only minor effects on keratinocyte proliferation (decrease from 2,631 to 2,445 (untreated control – 10 µmol/l cilomilast) for murine cells and from 607 to 553 (untreated control – 10 µmol/l cilomilast) in canine cells. The murine keratinocyte line MSC-P5 as well as canine keratinocytes were growth inhibited in a dose dependent manner by CsA (decrease from 2,079 to 1,505 (untreated control – 10 µmol/l CsA) in murine cells and from 553 to 302 (untreated control – 10 µmol/l CsA) in canine cells (Fig. 1)). Cell
cells. Viability was not affected by CsA or cilomilast at the indicated doses after incubation for 48 h (Table 1).
Effect of CsA and cilomilast on LPS induced PGE2 synthesis
Murine and canine keratinocytes were incubated with the toll-like receptor ligand 4; LPS induced an increase of PGE2 synthesis, measured 48 h after stimulation. CsA as well as cilomilast could reduce the LPS induced PGE2 synthesis in canine (Fig. 2) and murine (not shown) keratinocytes. The CsA dose dependently reduced the PGE2 synthesis. It failed, however, to reduce this significantly, obviously due to the variability of the positive control.
Cilomilast showed already maximal effect at 1 µmol/l.
Effect of CsA and cilomilast on PGN induced KC and CCL2 synthesis
PGN induced a dose dependent increase in KC secretion in murine keratinocytes (Fig. 3A).
Cilomilast (Fig. 3B) and CsA (not shown) reduced the KC secretion in a dose dependent manner. The maximal inhibitory effect was 30 % for cilomilast (50 µmol/l) and 35 % for CsA (10 µmol/l).
A recent study has revealed that high concentrations of PDE4 inhibitors enhance KC secretion in endothelial cells17. Therefore, it seemed necessary to measure KC secretion after
incubation with cilomilast without the following PGN stimulation of keratinocytes. There was no enhanced KC production after cilomilast treatment compared to untreated controls
(control: 56 ± 20 pg/ml, cilomilast, 10 µmol/l: 42 ± 17 pg/ml).
The PGN induced CCL2 secretion was subject to larger variations (even in aliquots of the same samples measured for KC secretion). Therefore, a clear dose response was not obtainable as had been the case for KC secretion. Nevertheless, the highest tested
concentrations of cilomilast and CsA significantly inhibited the PGN induced CCL2 secretion in murine keratinocytes (Fig. 4).
Effect of CsA and cilomilast on IFNγ induced IP-10 synthesis in HaCaT keratinocytes IFN γ induced an ample increase in IP-10 synthesis 24 h after stimulation. CsA as well as cilomilast reduced the IFN γ-induced IP-10 secretion in a dose dependent manner. The maximum inhibition varied from 30 to 50 % between the experiments (Fig. 5). CsA, or cilomilast administered to non-stimulated HaCat cells did not induce IP-10 at the tested concentrations (control: 26 ± 17 pg/ml, cilomilast, 10 µmol/l: 17 ± 15 pg/ml, CsA, 10 µmol/l:
27 ± 7 pg/ml).
Discussion
The calcineurin inhibitor CsA has been proven to be effective for the treatment of human and canine AD1;18. There are also promising clinical studies for PDE4 inhibitors, which
demonstrate efficacy for the treatment of human19 and canine20 AD. The PDE4 inhibitor arofylline, for example, was tested in dogs with AD and improved clinical symptoms as well as pruritus to the same extent as prednisone20.
However, the influence of CsA or PDE4 inhibitors on keratinocyte function, especially the canine keratinocyte function has not yet been studied extensively. It has already been demonstrated that PDE4 is expressed in keratinocytes and that the expression can be
upregulated by forskolin14. Nonetheless, an effect on cytokine and chemokine secretion has not been demonstrated so far.
In the past keratinocytes were merely considered as passive targets of immunological attack by infiltrating T cells. Recent studies clearly point out an active participation of keratinocytes in the cutaneous immune response. Particularly the recent advances in toll-like receptor research reveal that keratinocytes respond to various stimuli with secretion of cytokines, chemokines and adhesion molecules21. Keratinocytes form the outer barrier against microbial antigens and allergens.
Thus, the first contact with allergens as well as topically administered drugs takes place through the keratinocytes. Additionally, cytokines like tumour necrosis factor α (TNFα) or IFNγ, secreted by skin infiltrating immune cells results in keratinocyte stimulation and secretion of inflammatory mediators. Therefore, keratinocytes can transmit both positive and negative signals to cells of innate and adaptive immunity.
Apart from the effect of CsA and cilomilast on inflammatory mediators secreted by activated keratinocytes, a further central keratinocyte function, proliferation, was studied. The
antiproliferative effect of CsA has already been demonstrated for human and murine
keratinocytes and is discussed as being one additional benefit for the treatment of psoriasis in humans4. Although epidermal hyperproliferation is much more pronounced in psoriasis compared to atopic dermatitis22, there is also a hyperproliferation in human atopic dermatitis patients compared to healthy controls23.
Epidermal hyperplasia is also described for canine atopic dermatitis, although there are conflicting results concerning the magnitude of epidermal hyperplasia24. As a clinical improvement of atopic dermatitis is associated with a reduction of epidermal
the treatment of human atopic dermatitis and it can be speculated that it may also be
supportive for the treatment of canine atopic dermatitis. This antiproliferative effect of CsA was also demonstrated in vivo in murine models of epidermal hyperplasia26,27. In contrast to CsA, cilomilast had no significant impact on keratinocyte proliferation at the tested doses.
The first set of experiments concerning keratinocyte activation was performed with the toll-like receptor 4 agonist lipopolysaccharide (LPS), as keratinocytes are described as expressing functional toll-like receptor 4 and CD1428. However, high concentrations of LPS (100 µg/ml) were necessary to obtain an increased PGE2 synthesis. These concentrations were about 100 to 1,000 times higher than those needed for the stimulation of other cells, like dendritic cells13.
Hence, apart from LPS, we used peptidoglycan (PGN) from Staphylococcus aureus as a proinflammatory stimulus for keratinocytes. This stimulus has some relevance with regard to AD, as increased numbers of S. aureus are found in over 90 % of human AD lesions29. There is also a high prevalence of recurrent staphylococcal infections in dogs with atopic
dermatitis30.
The conclusion was reached that coagulase-positive staphylococci, like S. intermedius are also a trigger factor of canine atopic dermatitis since the numbers of S. intermedius organisms as well as their adherence to keratinocytes appear to decrease with remission of AD in dogs31. It is thought that staphylococcal infections exacerbate or maintain skin inflammation in atopic dermatitis in secreting superantigens, which stimulate marked activation of T cells and macrophages29.
Nevertheless, recently, it could be demonstrated, that PGN, which predominates the cell wall of staphylococci activates nuclear factor-κB and induces IL-8 production from keratinocytes via toll-like receptor 2 abundantly21;32. Furthermore, it was demonstrated that PGN also activates mitogen-activated protein kinases. Apart from IL-8, keratinocytes secrete IL-6, monocyte chemo-attractant protein 1, TNFα, CCL5 and granulocyte macrophage colony-stimulating factor upon stimulation with PGN.
All these cytokines and chemokines are involved in skin disorders32 and most of them are involved in the pathophysiology of AD33. Here, we could demonstrate that CsA and cilomilast had an impact on PGN-induced CCL2 and KC secretion. This might be an additional
explanation for their anti-inflammatory action in allergic skin diseases like AD or allergic contact dermatitis and, in the case of CsA, in psoriasis.
IFNγ is a typical Th1 cytokine and allergic diseases are thought to be predominatly Th2 related. However, in chronic lesions of AD a mixed population of T-helper cells can be found
in humans34 and dogs35. The inhibitory effect of CsA and cilomilast on IFNγ induced cytokine secretion (IP-10) might therefore be beneficial in chronic affected skin lesions of AD. It has already been demonstrated in HaCat cells, that CsA can inhibit TNFα secretion36. Therefore, we could confirm the anti-inflammatory effect of CsA for a further pro-inflammatory
mediator, IP-10, in keratinocytes.
A crucial point to bear in mind is the concentration of CsA and cilomilast, which can be achieved in keratinocytes in vivo. For cilomilast, high micromolar concentrations can only be achieved by topically administering the affected skin. As PDE4 inhibitors administered via the systemic route have dose limiting side effects such as nausea and emesis in humans9 and dogs20, the topical route may be anyway preferable for the treatment of AD.
For CsA the concentration in the skin is up to ten times higher than blood concentrations in laboratory animals and humans1. Won et al. (1994) were able to demonstrate that CsA inhibits keratinocyte cytokine gene expression and keratinocyte proliferation in the same dose range as was observed in our study4 in the case of cytokine and chemokine secretion.
In man skin tissue concentrations up to 2.4 µmol/l CsA have been determined. CsA has been described to be more potent in long term treatment. It was demonstrated that after five days exposure with 2 µmol/l CsA had the same inhibitory potency on IL-1 gene expression as 2 day exposure with 10 to 15 µmol/l CsA4. Therefore, the concentrations tested in our study can be of clinical relevance for the treatment of canine and human AD.
The discrepancy of described inhibitory effects of CsA on cytokine and chemokine expression and secretion in keratinocytes4;36, or the lack of it6;7, might be due to culture conditions, like incubation times, and due to the selected cells. Marionnet et al. (1997) could demonstrate that in immortalised keratinocytes the IL-1 and TNFα expression was affected in a different manner when treated with CsA as compared to normal human keratinocytes36.
In conclusion, our data provide evidence that CsA and cilomilast directly modulate
keratinocyte function. This might contribute to the anti-inflammatory and immunomodulatory action of these compounds in the treatment of allergic skin diseases.
Acknowledgements
The technical support of Victoria Garder, Grazyna Ludwig and Caroline Schultz is greatly acknowledged. We thank Frances C. Sherwood-Brock and Jennifer Offinger for their helpful comments on the English.
References
1. Guaguere E, Steffan J, Olivry T. Cyclosporin A: a new drug in the field of canine dermatology. Veterinary Dermatology 2004; 15: 61-74.
2. Furue M, Gaspari AA, Katz SI. The effect of cyclosporin A on epidermal cells. II.
Cyclosporin A inhibits proliferation of normal and transformed keratinocytes. Journal of Investigative Dermatology 1988; 90: 796-800.
3. Fisher GJ, Duell EA, Nickoloff BJ et al. Levels of cyclosporin in epidermis of treated psoriasis patients differentially inhibit growth of keratinocytes cultured in serum free versus serum containing media. Journal of Investigative Dermatology 1988; 91: 142-6.
4. Won YH, Sauder DN, McKenzie RC. Cyclosporin A inhibits keratinocyte cytokine gene expression. British Journal of Dermatology 1994; 130: 312-9.
5. Elder JT, Hammerberg C, Cooper KD et al. Cyclosporin A rapidly inhibits epidermal cytokine expression in psoriasis lesions, but not in cytokine-stimulated cultured keratinocytes. Journal of Investigative Dermatology 1993; 101: 761-6.
6. Kaplan A, Matsue H, Shibaki A et al. The effects of cyclosporin A and FK506 on proliferation and IL-8 production of cultured human keratinocytes. Journal of Dermatological Science 1995; 10: 130-8.
7. Friccius H, Pohla H, Adibzadeh M et al. The effects of the antifungal azoles
itraconazole, fluconazole, ketoconazole and miconazole on cytokine gene expression in human lymphoid cells. International Journal of Immunopharmacology 1992; 14:
791-9.
8. Al Daraji WI, Grant KR, Ryan K et al. Localization of calcineurin/NFAT in human skin and psoriasis and inhibition of calcineurin/NFAT activation in human
keratinocytes by cyclosporin A. Journal of Investigative Dermatology 2002; 118: 779-88.
9. Giembycz MA. Development status of second generation PDE4 inhibitors for asthma and COPD: the story so far. Monaldi Archives for Chest Disease 2002; 57: 48-64.
10. Baumer W, Gorr G, Hoppmann J et al. Effects of the phosphodiesterase 4 inhibitors SB 207499 and AWD 12-281 on the inflammatory reaction in a model of allergic dermatitis. European Journal of Pharmacology 2002; 446: 195-200.
11. Bäumer W, Seegers U, Braun M et al. TARC and RANTES, but not CTACK, are induced in two models of allergic contact dermatitis. Effects of cilomilast and
diflorasone diacetate on T-cell-attracting chemokines. British Journal of Dermatology 2004; 151: 823-30.
12. Bäumer W, Gorr G, Hoppmann J et al. AWD 12-281, a highly selective phosphodiesterase 4 inhibitor, is effective in the prevention and treatment of inflammatory reactions in a model of allergic dermatitis. Journal of Pharmacy and Pharmacology 2003; 55: 1107-14.
13. Bäumer W, Sülzle B, Weigt H et al. Effects of the immunomodulatory drugs tacrolimus, rapamycin and cilomilast on dendritic cell function in a rodent model of
allergic contact dermatitis. In: Hillier,A, Foster,AP, Kwochka,KW, eds. Advances in Dermatology, Vol. 5. Oxford: Blackwell Publishing, 2005: 89-96.
14. Chujor CS, Hammerschmid F, Lam C. Cyclic nucleotide phosphodiesterase 4
subtypes are differentially expressed by primary keratinocytes and human epidermoid cell lines. Journal of Investigative Dermatology 1998; 110: 287-91.
15. Köhler HB, Huchzermeyer B, Martin M et al. TNF-alpha dependent NF-kappa B activation in cultured canine keratinocytes is partly mediated by reactive oxygen species. Veterinary Dermatology 2001; 12: 129-37.
16. Scholz K, Furstenberger G, Muller-Decker K et al. Differential expression of prostaglandin-H synthase isoenzymes in normal and activated keratinocytes in vivo and in vitro. Biochemical Journal 1995; 309 ( Pt 1): 263-9.
17. McCluskie K, Klein U, Linnevers C et al. Phosphodiesterase type 4 inhibitors cause proinflammatory effects in vivo. Journal of Pharmacology and Experimental
Therapeutics 2006; 319: 468-76.
18. Akdis CA, Akdis M, Bieber T et al. Diagnosis and treatment of atopic dermatitis in children and adults: European Academy of Allergology and Clinical
Immunology/American Academy of Allergy, Asthma and Immunology/PRACTALL Consensus Report. Journal of Allergy and Clinical Immunology 2006; 118: 152-69.
19. Hanifin JM, Chan SC, Cheng JB et al. Type 4 phosphodiesterase inhibitors have clinical and in vitro anti-inflammatory effects in atopic dermatitis. Journal of
19. Hanifin JM, Chan SC, Cheng JB et al. Type 4 phosphodiesterase inhibitors have clinical and in vitro anti-inflammatory effects in atopic dermatitis. Journal of