Portal de Periódicos da CAPES

Dimethylfumarate Inhibits Tumor-Necrosis-Factor-Induced CD62E Expression in an NF-κB-Dependent Manner

2001; Elsevier BV; Volume: 117; Issue: 6 Linguagem: Inglês

10.1046/j.0022-202x.2001.01576.x

1523-1747

Rainer de Martin, Ulrich Mrowietz, Wolfgang Wiegrebe, David Jirovský, Robert Loewe, Manuela Pillinger, Marion Gröger, Wolfgang Holnthoner, Klaus Wolff, Peter Petzelbauer,

Cytokine Signaling Pathways and Interactions

Fumaric acid esters are thought to improve psoriasis by altering leukocyte, keratinocyte, and/or endothelial functions. To determine specificity, kinetics, and molecular mechanisms of different fumaric acid esters in their ability to inhibit endothelial cell activation, we analyzed CD62E and CD54 expression in endothelial cells in vivo and in vitro. In lesional skin of psoriatic patients, oral fumaric acid ester treatment resulted in a marked reduction of CD62E but not CD54 expression on dermal microvessels. Using human umbilical vein endothelial cells, dimethylfumarate almost completely inhibited tumor-necrosis-factor-induced CD62E, but not CD54 expression at concentrations ≨70 µM, mimicking the situation in vivo. A 60 min dimethylfumarate preincubation was sufficient to block tumor-necrosis-factor-induced CD62E expression for up to 24 h. In contrast, equimolar concentrations of methylhydrogenfumarate, the hydrolysis product of dimethylfumarate, did not suppress tumor-necrosis-factor-induced CD62E expression. Likewise, all fumaric acid esters other than dimethylfumarate were ineffective. Using CD62E, NF-κB, or AP-1-responsive promoter constructs, dimethylfumarate inhibited tumor-necrosis-factor-induced activation of the CD62E and the NF-κB but not the AP-1 promoter construct. In summary, at a dose range ≨70 µM, dimethylfumarate appeared to be a specific inhibitor of CD62E expression in an NF-κB-dependent manner. Fumaric acid esters are thought to improve psoriasis by altering leukocyte, keratinocyte, and/or endothelial functions. To determine specificity, kinetics, and molecular mechanisms of different fumaric acid esters in their ability to inhibit endothelial cell activation, we analyzed CD62E and CD54 expression in endothelial cells in vivo and in vitro. In lesional skin of psoriatic patients, oral fumaric acid ester treatment resulted in a marked reduction of CD62E but not CD54 expression on dermal microvessels. Using human umbilical vein endothelial cells, dimethylfumarate almost completely inhibited tumor-necrosis-factor-induced CD62E, but not CD54 expression at concentrations ≨70 µM, mimicking the situation in vivo. A 60 min dimethylfumarate preincubation was sufficient to block tumor-necrosis-factor-induced CD62E expression for up to 24 h. In contrast, equimolar concentrations of methylhydrogenfumarate, the hydrolysis product of dimethylfumarate, did not suppress tumor-necrosis-factor-induced CD62E expression. Likewise, all fumaric acid esters other than dimethylfumarate were ineffective. Using CD62E, NF-κB, or AP-1-responsive promoter constructs, dimethylfumarate inhibited tumor-necrosis-factor-induced activation of the CD62E and the NF-κB but not the AP-1 promoter construct. In summary, at a dose range ≨70 µM, dimethylfumarate appeared to be a specific inhibitor of CD62E expression in an NF-κB-dependent manner. calcium bis(monomethylfumarate) dimethylfumarate ethylhydrogenfumarate fumaric acid fumaric acid esters human umbilical vein endothelial cells methylhydrogenfumarate Fumaric acid esters (FAE) have been used empirically in the treatment of psoriasis for many years (reviewed byMrowietz et al., 1999Mrowietz U. Christophers E. Altmeyer P. Treatment of severe psoriasis with fumaric acid esters: scientific background and guidelines for therapeutic use. The German Fumaric Acid Ester Consensus Conference.Br J Dermatol. 1999; 141: 424-429https://doi.org/10.1046/j.1365-2133.1999.03034.xCrossref PubMed Scopus (182) Google Scholar). In the first controlled study, a therapeutic potency was shown for dimethylfumarate (DMF) only and not for salts of ethylhydrogenfumarates (EHF) (Nieboer et al., 1989Nieboer C. de Hoop D. van Loenen A.C. Langendijk P.N. van Dijk E. Systemic therapy with fumaric acid derivates: new possibilities in the treatment of psoriasis.J Am Acad Dermatol. 1989; 20: 601-608Abstract Full Text PDF PubMed Scopus (86) Google Scholar). Kolbach et al used a combination of DMF plus EHF salts and found this mixture superior to DMF alone (Kolbach and Nieboer, 1992Kolbach D.N. Nieboer C. Fumaric acid therapy in psoriasis: results and side effects of 2 years of treatment.J Am Acad Dermatol. 1992; 27: 769-771Abstract Full Text PDF PubMed Scopus (91) Google Scholar). As all subsequent open and controlled clinical studies used this mixture of FAE (Altmeyer et al., 1994Altmeyer P.J. Matthes U. Pawlak F. et al.Antipsoriatic effect of fumaric acid derivatives. Results of a multicenter double-blind study in 100 patients.J Am Acad Dermatol. 1994; 30: 977-981Abstract Full Text PDF PubMed Scopus (257) Google Scholar;Mrowietz et al., 1998Mrowietz U. Christophers E. Altmeyer P. Treatment of psoriasis with fumaric acid esters: results of a prospective multicentre study. German Multicentre Study.Br J Dermatol. 1998; 138: 456-460Crossref PubMed Scopus (170) Google Scholar), questions regarding the most active ingredient(s) of the FAE mixture and the mode of action are still under debate. The situation is further complicated by the fact that, due to its chemical structure, DMF hydrolyzes into methylhydrogenfumarate (MHF) and methanol, but MHF has yet not been tested in a monotherapy regimen. In vitro studies have only partially resolved the question of individual efficacies of different FAE. For example, in lymphocytes, MHF modulated cytokine expression towards a Th2 cytokine profile (de Jong et al., 1996de Jong R. Bezemer A.C. Zomerdijk T.P. Pouw-Kraan T. Ottenhoff T.H. Nibbering P.H. Selective stimulation of T helper 2 cytokine responses by the anti-psoriasis agent monomethylfumarate.Eur J Immunol. 1996; 26: 2067-2074Crossref PubMed Scopus (200) Google Scholar;Asadullah et al., 1997Asadullah K. Schmid H. Friedrich M. Randow F. Volk H.D. Sterry W. Docke W.D. Influence of monomethylfumarate on monocytic cytokine formation – explanation for adverse and therapeutic effects in psoriasis?.Arch Dermatol Res. 1997; 289: 623-630https://doi.org/10.1007/s004030050251Crossref PubMed Scopus (72) Google Scholar). Subsequent work employing a different setup confirmed and extended these data by showing that also DMF switched T cell responses towards a Th2 cytokine profile (Ockenfels et al., 1998Ockenfels H.M. Schultewolter T. Ockenfels G. Funk R. Goos M. The antipsoriatic agent dimethylfumarate immunomodulates T-cell cytokine secretion and inhibits cytokines of the psoriatic cytokine network.Br J Dermatol. 1998; 139: 390-395Crossref PubMed Scopus (139) Google Scholar). Likewise, in keratinocytes, both MHF and DMF inhibited keratinocyte proliferation (Thio et al., 1994Thio H.B. Zomerdijk T.P. Oudshoorn C. Kempenaar J. Nibbering P.H. van der Schroeff J.G. Ponec M. Fumaric acid derivatives evoke a transient increase in intracellular free calcium concentration and inhibit the proliferation of human keratinocytes.Br J Dermatol. 1994; 131: 856-861Crossref PubMed Scopus (40) Google Scholar). Both MHF and DMF inhibited monocyte differentiation into dendritic cells (Zhu and Mrowietz, 2001Zhu K. Mrowietz U. Inihibition of dendritic cell differentiation by fumaric acid esters.J Invest Dermatol. 2001; 116: 203-208Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). In vascular endothelial cells, DMF but not EHF suppressed tumor necrosis factor (TNF) induced endothelial adhesion molecule expression, but MHF has not yet been tested (Vandermeeren et al., 1997Vandermeeren M. Janssens S. Borgers M. Geysen J. Dimethylfumarate is an inhibitor of cytokine-induced E-selectin, VCAM-1, and ICAM-1 expression in human endothelial cells.Biochem Biophys Res Commun. 1997; 234: 19-23https://doi.org/10.1006/bbrc.1997.6570Crossref PubMed Scopus (141) Google Scholar). We therefore analyzed the effect of different FAE on CD62E and CD54 expression in vascular endothelial cells in vivo and in vitro. CD62E and CD54 are inducible endothelial adhesion molecules. They play a very important role in mechanisms governing tethering, rolling, and tight adhesion of leukocytes on endothelial cell surfaces. This cascade of events is required to initiate and perpetuate tissue inflammation (Bevilacqua et al., 1989Bevilacqua M.P. Stengelin S. Gimbrone Jr, M.A. Seed B. Endothelial leukocyte adhesion molecule 1: an inducible receptor for neutrophils related to complement regulatory proteins and lectins.Science. 1989; 243: 1160-1165Crossref PubMed Scopus (1635) Google Scholar;Lawrence et al., 1995Lawrence M.B. Berg E.L. Butcher E.C. Springer T.A. Rolling of lymphocytes and neutrophils on peripheral node addressin and subsequent arrest on ICAM-1 shear flow.Eur J Immunol. 1995; 25: 1025-1031Crossref PubMed Scopus (107) Google Scholar). Here, we show that DMF, but not MHF, at concentrations ≨70 µM, is a specific inhibitor of CD62E expression in an NF-κB-dependent manner. Human umbilical vein endothelial cells (HUVEC) were isolated from umbilical cords and cultured as described previously (Petzelbauer et al., 1993Petzelbauer P. Bender J.R. Wilson J. Pober J.S. Heterogeneity of dermal microvascular endothelial cell antigen expression and cytokine responsiveness in situ and in cell culture.J Immunol. 1993; 151: 5062-5072PubMed Google Scholar). HUVEC were grown in Iscove's modified Dulbecco's medium (IMDM; Life Technologies, Gaithersburg, MD) containing 20% heat-inactivated fetal bovine serum (Life Technologies), 2 mM per l glutamine (Life Technologies), 50 IU per ml penicillin-streptomycin (Life Technologies), and endothelial cell growth supplement with heparin (Promocell, Heidelberg, Germany). Cells were used between passages 2 and 6. TNF-α was purchased from Strathmann Biotech (Hamburg, Germany) and used at a concentration of 10 ng per ml. All FAE were from Fumapharm (Muri, Switzerland). DMF was dissolved in methanol as a 70 mM stock solution. Fumaric acid (FA), calcium bis(monomethylfumarate) (CaMF), MHF, and EHF were prepared as a 70 mM stock solution in IMDM and stored for up to 3 d. Fumaderm (Fumamedica, Herne, Germany) is the only FAE registered for oral treatment of psoriasis. One tablet consists of DMF (120 mg) and a mixture of Ca2+, Mg2+, and Zn2+ salts of EHF (95 mg). Maximal daily dose is six tablets. After informed consent, 5 mm punch biopsies from lesional and nonlesional skin of three patients with severe psoriasis before and during systemic therapy with FAE (Fumaderm) were obtained from the Department of Dermatology of the University of Kiel, Germany. The biopsies were snap frozen in liquid nitrogen and shipped on dry ice. Specimens were cut on a Leica cryostat (Vienna, Austria), air dried, and fixed with acetone for 10 min at 4°C. Then, a three-step immunohistochemistry technique was performed using the ABC Vectastain kit (Vector Laboratories, Burlingame, CA) as described previously (Petzelbauer et al., 1993Petzelbauer P. Bender J.R. Wilson J. Pober J.S. Heterogeneity of dermal microvascular endothelial cell antigen expression and cytokine responsiveness in situ and in cell culture.J Immunol. 1993; 151: 5062-5072PubMed Google Scholar). First-step reagents were CD54 (clone BBIG-I1) and CD62E (clone BBIG-E4) (both R&D Systems, Minneapolis, MN). Bound peroxidase was visualized by 3-amino-9-ethylcarbazole (Sigma-Aldrich, Deisenhofen, Germany). Sections were counterstained with Mayers Hemalaun and examined with an Olympus microscope (Olympus Optical, Hamburg, Germany). Confluent HUVEC were stimulated with TNF for 4 h in the presence or absence of the indicated FAE. Cells were suspended in trypsin/ethylenediamine tetraacetic acid (Life Technologies), washed in phosphate-buffered saline (PBS), and incubated with CD54, CD62E, or mouse IgG1 κ isotype control (MOPC 31c, Sigma-Aldrich) in PBS (each 1 µg per ml) for 30 min on ice. After additional washings and incubation with the fluorescein isothiocyanate conjugated second-step antimouse IgG antibody (Sigma-Aldrich, 1 µg per ml) for 30 min on ice, surface-bound fluorescence was analyzed by FACScan (Becton Dickinson, San Jose, CA). Total fluorescence values were calculated as described previously (Petzelbauer et al., 1993Petzelbauer P. Bender J.R. Wilson J. Pober J.S. Heterogeneity of dermal microvascular endothelial cell antigen expression and cytokine responsiveness in situ and in cell culture.J Immunol. 1993; 151: 5062-5072PubMed Google Scholar); briefly the absolute number of positive cells was multiplied by the mean channel fluorescence corrected for the individual background. Data were pooled from three or five experiments and the mean ± SD was calculated. Total RNA was isolated using the RNEasy Kit from Qiagen (Hilden, Germany). A total of 10 µg RNA was mixed with RNA loading buffer (Sigma-Aldrich) and denatured for 10 min at 65°C prior to separation on a 0.8% agarose gel. RNA was transferred on Nylon membranes and stained with methylene blue. Probes for CD62E and CD54 (a cocktail of six oligonucleotide antisense probes, R&D Systems) were end-labeled with 32P by using T4 polynucleotide kinase. Blots were hybridized with the labeled probes in ExpressHyb buffer (Amersham, Buckinghamshire, U.K.) for 60 min at 42°C, and then washed with 2 × SSC (0.3 M NaCl, 30 mM sodium citrate) plus 0.1% sodium dodecyl sulfate (SDS) first, then in 0.5% SSC plus 0.1% SDS two times, and a last wash with 0.25 × SSC plus 0.1% SDS. The blots were then exposed to Kodak films (Eastman Kodak, Rochester, NY) for 1 or 2 d. Northern signals were determined by densitometry (Kodak electrophoresis documentation and analysis system 120) and normalized to their 28S RNA bands. A triplicate of the consensus AP-1 site (48 bp, 3′-TGTGATGACTCAGGTT) or a triplicate of the consensus NF-κB site (60 bp, 3′-AATCGTGGAATTTCCTCTGA) flanked by SpeI sites were inserted into the SpeI site of the pTK-UBT-luc vector, respectively (de Martin et al., 1993de Martin R. Strasswimmer J. Philipson L. A new luciferase promoter insertion vector for the analysis of weak transcriptional activities.Gene. 1993; 124: 137-138Crossref PubMed Scopus (19) Google Scholar). A 1.3 kb construct of the CD62E promoter spanning from bp -1285 to bp +482 was inserted into the NdeI site of the pMAM neo-luc vector (Clontech) as described previously (Cooper et al., 1996Cooper J.T. Stroka D.M. Brostjan C. Palmetshofer A. Bach F.H. Ferran C. A20 blocks endothelial cell activation through a NF-kappaB-dependent mechanism.J Biol Chem. 1996; 271: 18068-18073Abstract Full Text Full Text PDF PubMed Scopus (215) Google Scholar). Cells were transfected using the standard Ca2+ phosphate precipitation method as described elsewhere (Gille et al., 1996Gille J. Swerlick R.A. Lawley T.J. Caughman S.W. Differential regulation of vascular cell adhesion molecule-1 gene transcription by tumor necrosis factor alpha and interleukin-1 alpha in dermal microvascular endothelial cells.Blood. 1996; 87: 211-217Crossref PubMed Google Scholar). Briefly, HUVEC between passages 3 and 5 were seeded in gelatine-coated six-well plates (Falcon) and grown to 70%-80% confluence. For transfection, 2.5 µg of the respective promoter construct was added per well. To control for transfection efficiency, cotransfection with 500 ng of a pSV-β-galactosidase control vector (Promega, Madison, WI) was performed in each experiment. Two days after transfection, cells were stimulated for 2 h with 10 ng per ml TNF with or without 40 µM DMF. Cells were then trypsinized, pelleted, rinsed, and resuspended in 200 µl reporter lysis buffer (Promega) for 15 min according to the manufacturer's instructions. Luciferase activity was measured with a Berthold AutoLumat LB9507 luminometer using the luciferase assay system (Promega). β-Galactosidase activity was determined by a β-galactosidase enzyme assay system (Promega). Luciferase activities obtained with the respective promoter constructs were normalized according to their β-galactosidase activity. The variation of β-galactosidase activity within individual experiments was < 10%. Skin biopsies of psoriatic lesions and normal skin before and at weeks 1, 3, and 5 during oral treatment with FAE (Fumaderm) were analyzed for CD62E and CD54 expression. To allow a semiquantitative evaluation of the staining intensity with the respective antibodies, we performed serial dilutions of monoclonal antibodies. For CD62E, at a concentration of 0.3 µg per ml, minimal reactivity was seen in untreated lesional skin and CD62E staining was undetectable at days 7 through 35 after the initiation of treatment. At 3 µg per ml, strong reactivity was seen at days zero and 7, and staining was undetectable at days 22 and 35. Staining at an antibody concentration of 1 µg per ml is shown in Figure 1, showing strong CD62E reactivity at day zero, weak reactivity at day 7, and no reactivity thereafter. For CD54, at a concentration of 0.3 µg per ml, all specimens showed an equally weak staining pattern; at 3 µg per ml all specimens uniformly showed strong staining. Also the CD54 expression on keratinocytes and dermal infiltrating cells persisted. Histologically, lesions still showed a psoriatic phenotype. Clinical improvement was seen at day 21. Centers have cleared at day 35 with active margins still present. The effects of different FAE and of FA on TNF-induced CD62E expression in HUVEC are shown in Figure 2(a). DMF completely blocked the TNF-induced CD62E expression at concentrations ≨70 µM compared with the FA control and with other FAE tested. A less than 10% inhibition of CD62E expression by FAE other than DMF was seen at concentrations of 700 µM. Methanol did not inhibit TNF-induced expression of CD62E at any concentration tested. A summary of three independent experiments comparing DMF and MHF effects on CD62E expression is shown in Figure 2(b). Figure 2(c) compares the inhibitory effect of DMF on CD62E with that on CD54 expression. At 25 and 40 µM, DMF inhibited CD62E but not CD54 expression. A minimal reduction of CD54 expression was seen at 140 µM DMF. In order to see a significant inhibition of CD54 expression by DMF, concentrations had to be raised to 700 µM (data not shown). In the next set of experiments, we tested onset and duration of DMF effects. HUVEC were pretreated with DMF for 1, 10, or 60 min. DMF was then removed, and cells were rested for various periods of time up to 24 h and then stimulated with TNF. Following a 1 or 10 min DMF preincubation, subsequent TNF-induced CD62E expression was not inhibited. In contrast, after a 60 min preincubation with DMF subsequent TNF-induced CD62E expression was inhibited for up to 24 h and returned to normal thereafter. Results of a 4 h resting period are shown in Figure 3. CD62E and CD54 mRNA expression was analyzed by northern blot analysis Figure 4. HUVEC were preincubated with DMF or FA and then stimulated with TNF for various times. Due to the different inducibility of the two proteins tested, different time courses were chosen. CD62E mRNA is inducible by TNF within the first 3 h. Therefore, mRNA from HUVEC stimulated with TNF for 1 and 3 h and pretreated with DMF or FA was compared. After 1 h of stimulation with TNF, CD62E mRNA was detectable in HUVEC preincubated with FA. After 3 h of stimulation, the detectable CD62E mRNA signal increased nearly 3-fold. After preincubation with DMF, after 1 or 3 h of TNF stimulation, CD62E mRNA was nearly undetectable, resulting in a dramatically reduced CD62E protein expression detected by FACS analysis. On the other hand, CD54 mRNA is inducible over a more prolonged period of time, reflecting the transcriptional control exerted by different transcription factors. Therefore, HUVEC were stimulated for up to 18 h with TNF. After preincubation with FA, CD54 mRNA was strongly detectable after 3 and 12 h of stimulation; after 18 h the signal intensity slowly decreased. HUVEC preincubated with DMF showed a reduced amount of mRNA after 3 h of incubation, but after 12 or 18 h the detectable amount of mRNA compared with the FA preincubated cells was equal or even slightly increased. HUVEC transfected with the CD62E promoter or constructs containing NF-κB or AP-1 binding sites were treated with TNF with or without 40 µM DMF Figure 5. Cells transfected with the CD62E promoter showed a mean 8-fold increase in luciferase activity following a 2 h TNF stimulation, which was inhibited by DMF (p <0.05). The level of inhibition by 40 µM DMF seen in the CD62E promoter construct roughly corresponded to the DMF-induced inhibition of CD62E protein expression at the same concentration. Similarly, using an NF-κB-responsive construct, the TNF-induced increase in luciferase activity was significantly inhibited by DMF. In contrast, using an AP-1-responsive construct, DMF had no effect on TNF-induced luciferase activity. Recruitment of white blood cells from the circulation into tissues strongly depends on the expression of endothelial adhesion molecules. Many of these molecules are inducible by TNF. Indeed, several inflammatory diseases benefit from an inhibition of TNF-mediated cell activation (Neurath et al., 1996Neurath M.F. Pettersson S. Meyer zum Buschenfelde K.H. Strober W. Local administration of antisense phosphorothioate oligonucleotides to the p65 subunit of NF-kappa B abrogates established experimental colitis in mice.Nat Med. 1996; 2: 998-1004Crossref PubMed Scopus (716) Google Scholar;Paleolog et al., 1996Paleolog E.M. Hunt M. Elliott M.J. Feldmann M. Maini R.N. Woody J.N. Deactivation of vascular endothelium by monoclonal anti-tumor necrosis factor alpha antibody in rheumatoid arthritis.Arthritis Rheum. 1996; 39: 1082-1091Crossref PubMed Scopus (220) Google Scholar). Here, we demonstrate that DMF inhibits TNF-induced endothelial cell activation, as evidenced by reduced endothelial CD62E expression in psoriatic skin in vivo and reduced CD62E mRNA and protein expression in vitro. Likewise, DMF inhibited transcription of a CD62E promoter construct and of an NF-κB reporter construct to a similar extent, but not of an AP-1 construct. This is in line with the fact that CD62E promoter activation mainly depends on NF-κB (Schindler and Baichwal, 1994Schindler U. Baichwal V.R. Three NF-kappa B binding sites in the human E-selectin gene required for maximal tumor necrosis factor alpha-induced expression.Mol Cell Biol. 1994; 14: 5820-5831Crossref PubMed Google Scholar;Read et al., 1997Read M.A. Whitley M.Z. Gupta S. Pierce J.W. Best J. Davis R.J. Collins T. Tumor necrosis factor alpha-induced E-selectin expression is activated by the nuclear factor-kappaB and c-JUN N-terminal kinase/p38 mitogen-activated protein kinase pathways.J Biol Chem. 1997; 272: 2753-2761Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar) and allows the conclusion that DMF inhibited CD62E expression at the transcriptional level in an NF-κB-dependent way. The specificity of DMF to inhibit CD62E expression is dose dependent and seen at concentrations ≨70 µM. At these concentrations DMF did not block CD54 expression. In contrast to CD62E expression, which is mainly under the control of NF-κB, regulation of CD54 expression is more complex (Roebuck and Finnegan, 1999Roebuck K.A. Finnegan A. Regulation of intercellular adhesion molecule-1 (CD54) gene expression.J Leukoc Biol. 1999; 66: 876-888Crossref PubMed Scopus (444) Google Scholar). The CD54 promoter is activated by three major pathways, through NF-κB, AP-1, and through JAK/STAT pathways, which partly interact and partly overlap. TNF activates CD54 gene expression through NF-κB and AP-1, whereas JAK/STAT activation occurs through interferons. In cell culture, CD54 mRNA expression was slightly reduced 3 h following TNF stimulation reflecting the NF-κB-dependent early phase of CD54 induction. At later time points, no inhibition of CD54 mRNA could be detected. In contrast, after 12 and 18 h the amount of detectable CD54 mRNA was slightly higher in the DMF-pretreated cells as determined by relative band densities. It is known that the AP-1-dependent transcription of CD54 starts at a later time point; therefore the mRNA levels at 12 and 18 h of incubation reflect the not inhibited AP-1-dependent gene transcription. CD62E mRNA expression was nearly totally inhibited during the whole time course. Corresponding to the effects of DMF on the mRNA expression of CD54, CD54 protein expression was not inhibited by ≨70 µM DMF. It is thus likely that the reduced NF-κB-dependent transcription is balanced by AP-1. Indeed, transcription of an AP-1 reporter construct was not affected by DMF. Moreover, this finding correlated with our observation in vivo, where CD54 expression remained unchanged throughout a treatment period of 5 wk. DMF readily hydrolyzes into MHF and methanol. Therefore, MHF and not DMF was believed to mediate the biologic response (Nibbering et al., 1993Nibbering P.H. Thio B. Zomerdijk T.P. Bezemer A.C. Beijersbergen R.L. van Furth R. Effects of monomethylfumarate on human granulocytes.J Invest Dermatol. 1993; 101: 37-42Abstract Full Text PDF PubMed Google Scholar). Our experiments using endothelium, however, demonstrated that MHF, at equimolar concentrations to DMF, did not inhibit CD62E expression. Thus, in the context of endothelial cells, MHF is unlikely to be the active metabolite of DMF. We wish to point out that effects of FAE on endothelial cells differ from those on lymphocytes and keratinocytes, where both DMF and MHF appear equally effective (Thio et al., 1994Thio H.B. Zomerdijk T.P. Oudshoorn C. Kempenaar J. Nibbering P.H. van der Schroeff J.G. Ponec M. Fumaric acid derivatives evoke a transient increase in intracellular free calcium concentration and inhibit the proliferation of human keratinocytes.Br J Dermatol. 1994; 131: 856-861Crossref PubMed Scopus (40) Google Scholar;de Jong et al., 1996de Jong R. Bezemer A.C. Zomerdijk T.P. Pouw-Kraan T. Ottenhoff T.H. Nibbering P.H. Selective stimulation of T helper 2 cytokine responses by the anti-psoriasis agent monomethylfumarate.Eur J Immunol. 1996; 26: 2067-2074Crossref PubMed Scopus (200) Google Scholar). The mechanism by which DMF transmits signals to endothelial cells is not yet understood. Among several possibilities, DMF may inhibit the TNF–TNF receptor interaction. This is unlikely, however, because the TNF-induced CD54 expression is not blocked by DMF. Our finding that a 60 min preincubation with DMF followed by washout is sufficient to maintain endothelial cells unresponsive to TNF raises another possibility. DMF may be taken up into cells and lead to a persisting inhibition of signal transduction. Interestingly, DMF is known to interfere with the cellular glutathione metabolism (Miller et al., 1993Miller A.C. Gafner J. Clark E.P. Samid D. Posttranscriptional down-regulation of ras oncogene expression by inhibitors of cellular glutathione.Mol Cell Biol. 1993; 13: 4416-4422Crossref PubMed Scopus (32) Google Scholar;Nelson et al., 1999Nelson K.C. Carlson J.L. Newman M.L. et al.Effect of dietary inducer dimethylfumarate on glutathione in cultured human retinal pigment epithelial cells.Invest Ophthalmol Vis Sci. 1999; 40: 1927-1935PubMed Google Scholar). As glutathione is known to interfere with NF-κB translocation, this could be a possible mechanism for the observed CD62E inhibition by DMF. As another possibility, DMF might inhibit nuclear translocation of NF-κB p50/p65 heterodimers as CD62E expression depends on these heterodimers. Indirect evidence for this assumption comes from recent work by Vandermeeren et al who found that DMF inhibited nuclear entry of NF-κB p50 in dermal fibroblasts (Vandermeeren et al., 2001Vandermeeren M. Janssens S. Wouters H. Borghmans I. Borgers M. Beyaert R. Geysen J. Dimethylfumarate is an inhibitor of cytokine-induced nuclear translocation of NF-κB1, but not RelA in normal human dermal fibroblast cells.J Invest Dermatol. 2001; 116: 124-130https://doi.org/10.1046/j.1523-1747.2001.00211.xAbstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar). Importantly, Vandermeeren et al found translocation of NF-κB p65 not to be affected. Therefore, results obtained in fibroblasts cannot easily be transposed to effects of DMF seen in endothelial cells. In endothelium, as in many other cell types, p50/p65 is shuttled as a complex; thus translocation of both proteins should be affected by DMF. It is unlikely that DMF inhibits CD62E expression by inhibiting translocation of p50 homodimers, because p50 homodimers have the reverse effect - they inhibit p50/p65 DNA binding and transcription (Baer et al., 1998Baer M. Dillner A. Schwartz R.C. Sedon C. Nedospasov S. Johnson P.F. TNFα transcription in macrophages is attenuated by an autocrine factor that preferentially induces NF-κB p50.Mol Cell Biol. 1998; 18: 5678-5689Crossref PubMed Google Scholar). CaMF, EHF, and MHF exhibit an inhibitory effect on endothelial functions only at 700 µM. At this concentration, even interferon-γ-induced HLA-DR expression on endothelial cells is inhibited (data not shown). As HLA-DR expression is independent from NF-κB signaling, it may be assumed that this high concentration of FAE inhibits endothelial cell activation by pathways other than NF-κB. With regard to DMF, specificity for NF-κB-dependent genes is lost at concentrations of approximately 100 µM. We thus conclude that with regard to the inhibition of endothelial activation DMF is the therapeutic active FAE. In endothelium, DMF appears to have a dose-dependent specificity for NF-κB. It thus can be assumed that the therapeutic potency of DMF in psoriasis is due to the combined effect of DMF on functions of endothelial cell, keratinocyte, lymphocyte, and dendritic cell (Thio et al., 1994Thio H.B. Zomerdijk T.P. Oudshoorn C. Kempenaar J. Nibbering P.H. van der Schroeff J.G. Ponec M. Fumaric acid derivatives evoke a transient increase in intracellular free calcium concentration and inhibit the proliferation of human keratinocytes.Br J Dermatol. 1994; 131: 856-861Crossref PubMed Scopus (40) Google Scholar;Vandermeeren et al., 1997Vandermeeren M. Janssens S. Borgers M. Geysen J. Dimethylfumarate is an inhibitor of cytokine-induced E-selectin, VCAM-1, and ICAM-1 expression in human endothelial cells.Biochem Biophys Res Commun. 1997; 234: 19-23https://doi.org/10.1006/bbrc.1997.6570Crossref PubMed Scopus (141) Google Scholar;Ockenfels et al., 1998Ockenfels H.M. Schultewolter T. Ockenfels G. Funk R. Goos M. The antipsoriatic agent dimethylfumarate immunomodulates T-cell cytokine secretion and inhibits cytokines of the psoriatic cytokine network.Br J Dermatol. 1998; 139: 390-395Crossref PubMed Scopus (139) Google Scholar;Zhu and Mrowietz, 2001Zhu K. Mrowietz U. Inihibition of dendritic cell differentiation by fumaric acid esters.J Invest Dermatol. 2001; 116: 203-208Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). This work was in part supported by a research grant from Fumapharm AG, Muri, Switzerland.