Rofe AM, Philcox JC, Haynes DR, Whitehouse MW, Coyle P. Changes in plasma zinc, copper, iron, and hepatic metallothionein in adjuvant- induced arthritis ...
Research Paper Mediators of Inflammation, 8, 253–259 (1999)
IN additio n to its w ell-e s tablis h ed e ffe ct o n T cells, cyclos p orin A (CsA) also in h ib its in fla m m atory cytokin e p r oduction by m acr op h age s . How eve r, little is kn ow n about th e m ech an is m o f actio n of CsA on m acr op h age cytokin e p ro duction . We m e as ur e d th e effe ct o f CsA o n bas al and p h o rbo l-m yris tate-ace tate (PMA)-s tim ulate d pr o duction o f in terle ukin -6 usin g th e h um an m on ocyte ce ll lin e U937 dif fe r e ntiated w ith dim eth ylsulfox ide (DMSO). In te rleukin -6 leve ls w er e m e as ure d in s up er n atan t an d cell lys ates usin g s p ecific en zym e-lin ke d im m uno s orben t as s ays . We fo un d th at CsA de cr e as e s n ot on ly IL-6 r ele as e but also cyto kin e s ynth e s is . Th e con ce ntratio n of CsA us ed did n ot affe ct eith e r cell viability o r pr olife ratio n. Th r ee p os s ib ilitie s m ay be advan ced to ex p lain th e CsA-due de cr e as e in IL-6 p r oductio n by m acro p h age s : (a ) in h ibitio n of th e s yn th es is of an early co m m on r egulatory p r otein , (b) in h ibitio n o f cytokin e ge ne tran s crip tio n, or (c ) m odulation of p os t-tran s crip tion al eve nts . Th e fir s t p os s ib ility w as tes ted by m eas ur in g th e effe ct of cycloh ex im ide on th e ex p e rim en tal s ys te m durin g th e fir s t 3 h o urs of cultur e. Alth ough cycloh ex im ide decre as e d to tal cytokin e s yn th es is , th e p atter n of cytokin e m odulation by CsA p ers is te d. Th es e da ta s ugge s t th at CsA-m edia ted m acr op h age cytokin e in h ibitio n is n ot m edia ted by an early com m o n r egulato ry p r otein . To furth er ex p lore th e in h ib itio n m ech anis m , w e m e as ur e d IL-6 m RNA leve ls by Nor th er n blot. IL-6 m RNA le vels w e r e un affe cted by CsA bo th in r es tin g an d PMA-stim ulated cells. We conclude th at in h um an m acr op h ages CsA dim in is h es IL-6 p r oductio n at p os t-tran s crip tio nal leve l.
Cyclosporin A decreases human macrophage interleukin-6 synthesis at post-transcriptional level 2 Juan E. Losa Garc´õ a1,CA, Ana M. Rodr´õ guez Lopez ´ , 3 Mar´õ a Rosa Mart´õ n de Cabo , Fernando Mateos Rodr´õ guez 4, Jes us ´ Perez ´ Losada 5, 5 Rogelio Gonzalez ´ Sarmiento , 5 Antonio Jim enez ´ Lopez ´ and Jose´ Luis Perez ´ Arellano6
1
Area de Medicina. Fundaci on ´ Hospital de Alcorcon, ´ Departmento de Fisiolog´õ a. Universidad de Salamanca, 3Centro de Salud Miguel Servet. Alcorcon, ´ 4Hospital General de Albacete, 5 Departmento de Medicina. Universidad de Salamanca, and 6Departamento de Ciencias Cl´õ nicas. Universidad de Las Palmas, Spain. 2
CA
Corresponding Author Internal Medic ine Unit, Fundaci´o n Hospital de Alcorc o´ n, Budape st 1, 28922 Alcorc o´ n, Madrid, Spain Te l: (+349) 16219513 Fax : (+349) 16219901 Email: je losa@fhalc orc on.es
Key w or ds : Cyclosporin A, Interleukin-6, Macrophage , U937.
Introduction Cyc losporin A (CsA) is an immunomodulator y substanc e use d in the pre ve ntio n of trans plant re je ction and in the tre atme nt of autoimmune dis ease.1 Classically, CsA has be en re porte d to ex e rt its immunomodulatory ac tion through its effect on T lymphocyte s, mainly helpe r T lymphoc yte s, by inhibiting c alcine urin depe nde nt inte rleukin-2 synthesis .2,3 Rece ntly, how e ver, it has bee n suggeste d that the inhibition of T lymphocyte ac tivation by CsA doe s not appropriate ly ac count for all the e ffec ts observe d follow ing in vivo adminis tratio n of the drug.4 Ac cording ly, despite w ide spre ad use of the drug, the sp ecific me chanis m of ac tion of CsA re mains to be fully elucidate d.5 It is also possible that its e ffec ts c ould be due in part to the ac tion of CsA on othe r cells.6 In this se nse , it should be note d that CsA binds spe cific ally, re versibly and in a time -and te mperature -depende nt fashion to all human leucocyte s. How e ve r, granulocyte s and c ells of the mononuclear phagoc yte syste m
(MPS) some time s bind more CsA than lymphocyte s, probably bec ause the drug is rapidly inte rnaliz ed.2,7 Ce lls of the MPS offe r an attrac tive possibility to study the effects of this drug because the y partic ipate in a crucial manne r in antige n pre se ntatio n and, in many c ase s, are the final effe ctors of the immune syste m. Many re fe re nce s in the lite rature re port the capacity of CsA to alte r the ac tivitie s of MPS ce lls. In gene ral, the capacitie s, func tions and ac tions of macrophage s re late d to non-spe cific de fe nse such as che motax is , phagocytosis , e nzyme re le as e and re spirator y burst, can be said to be more re sistant to CsA.8,9 By c ontras t, those re late d to immunore gulation or acc essory func tions , such as monokine production, the ex pre ssion of his toc ompatibility antigens and antige n pre sentation, are more susc eptible to the ac tion of CsA.8,9 Mac rophages synthe size and secre te IL-1b , TNF-a , IL-6 and IL-8.10 These monokine s are know n mediators of the inflammator y re sponse 10 and are the re fore also know n as inflammatory cytokine s. It has be en
ISSN 0962-9351 print/ISSN 1466-1861 online/99/040253-07 © 1999 Taylor & Francis Ltd
253
J. E. Lo s a Ga rc´õ a e t al.
re porte d that CsA inhibits the produc tion of IL-1b , TNF-a and IL-6 in diffe re nt c ells of the MPS.11,12 How e ver, the me chanism through w hich this oc curs is not w ell know n. 13 In the light of the fore going, the aim of the pre se nt study w as to e valuate and de te rmine the me chanism of ac tion of CsA on the produc tion of inte rleukin-6 by macrophage s. To do so, the human diffe re ntiate d U937 monoc yte line w as employed.
Methods Cell line U937 c ells w ere kindly supplie d by Dr J. Olmos of the Hospital Unive rs itario Marqu´e s de Valde cilla, Santande r (Spain). The cell line w as ke pt at 37°C in a humidifie d atmosphere w ith 5% CO2 in culture w ith complete medium c ontaining RPMI 1640 (Sigma®), 100 U/mL of p enic illin (Sigma®), 100 U/mL of stre ptomyc ine (Sigma®), 2 mM L-glutamine (Sigma®) and 10% fetal c alf se rum (FCS) (Gibc o®) in ste rile c ulture flasks (Nunclon®) at a conc entratio n betw e en 0.75 3 105 and 5 3 105 ce lls/mL. Ce ll viability w as gre ate r than 90% and the duplic atio n time w as be tw een 24 and 48 hours. Ce lls w e re re gularly scre ene d for lipopolysac charide (LPS), bac te ria, mycoplasma and fungal c ontamination and found to be ne gative . Ce lls w e re diffe rentiate d by c ulture ove r 4 days in complete me dium containing 1.3% dime thylsulfox ide (DMSO). Cell cultures 5 3 105 ce lls togethe r w ith diffe re nt stimuli and/or CsA at the follow ing conc entratio n w ere adde d to each w ell of the culture dishes (Costar®) w ith comple te me dium: Lipopolysaccharide from E. co li (Sigma®) at 100 m g/mL and 10 m g/mL; phorbol12-myristate -13-ace tate (PMA) (Sigma®) at 10–4M and 10–5M; human gamma-inte rfe ron (Sigma®) at 1000 U/mL and 100 U/mL and CsA (Sandoz®) at thre e non-tox ic c onc entratio ns in vivo : 200, 20 and 2 ng/mL. Ce lls w e re kept at 37°C in a humifie d atmosphere w ith 5% CO2 for 18 hours, afte r w hich the supernatants w e re c entrifuge d at 500 g for 10 min to se dime nt c ells in suspens ion and the n store d at –70°C until late r de te rmination. Afte r the supernatant had bee n re move d, 400 m L of a 0.25 M sucrose –0.02 EDTA solution w as adde d to each w e ll. Afte r the bottom of each w ell had bee n sc rape d vigorous ly, the conte nt w as as pirate d and adde d to the pre vious ce ll pellet, thus re covering all the ce lls. The cells w ere the n subjec te d to heat disinte gration by re p eate d fre ezing –thaw ing (6 c ycle s of 10 min fre ezing and 15 min thaw ing) and w e re store d at –70°C until dete rmination . 254
Mediators of Inflammation · Vol 8 · 1999
To study the inhibitio n of early prote in synthe sis, cyclohex imide (Sigma®) at a concentration of 1 m g/mL w as adde d during the first 3 hours of cell culture . Cytokine determination The inflammatory cytokine IL-6 w as dete rmine d by dire ct double-sandw ich enzyme-linke d immunosorbent assay (ELISA) of antib odie s in the supe rnatants and c ell lysate s. BIOTRAK® (Ame rsham, UK) c ommercial kit spec ific for human IL-6 w as use d; this show e d no c ros s-re activity w ith othe r cytokine s. It had a coeffic ie nt of variation less than 10%, and a sens itivity limit of 0.35 pg/mL. All sample s w e re measure d in duplic ate . Re sults w ere ex pre ssed in pic ograms/106 ce lls. It should be note d that in ELISA assays none of the compone nts of the c ulture medium (RPMI 1640, antib iotic s, glutamine or fe tal c alf se rum) show e d immunore ac tivity w ith the c ytokine studie d. IL-6 mRNA expression Trans cription of the IL-6 ge ne w as studie d by Northern blot. Diffe re ntiate d U937 c ytoplasmic RNA w as isolate d by 150 mM NaCI, 1 mM MgCI2, 10 mM TrisHCI pH 7.4, 0.5% Nonide t P40 and 250–1000 U/mL RNAs ing (Promega®) lysis, follow ed by tw o succ essive ultrac e ntrifug atio ns w ith Tris-buffe re d phe nol and 10% SDS and phenol, re sp ective ly. RNA w as pre c ipitate d ove rnight w ith cold ethanol and 3M sodium ac etate pH 7.2, and re dissolved in Tris-EDTA pH 7 buffe r. RNA w as quantifie d by dete rmining optic al density at a w ave length of 260 nm. The quality of the RNA w as che cke d by the demons tration of tw o bands corre sponding to ribosomal RNA (18s and 28s) and the absenc e of de gradation of the RNA by staining w ith ethidium bromide in the ele ctrophore sis of a 1% agarose minige l loade d w ith 1.5 m g of RNA. RNA w as denature d by inc ubatio n for 1 h at 50°C in a solution of 1 M glyox al, 50% DMSO and 10 mM NaH2PO4 . Follow ing this , 5 m g of RNA w e re mix e d w ith elec trophore sis sample buffer (50% glyc erol, 49.6% 10 mM NaH2PO4 p H 7.0, 0.4% bromophe nol blue ). Afte r loading sample s, e le ctrop hore sis (35V) w as c arrie d out in 1% agarose -10 mM NaH2PO4pH 7 gels using 10 mM NaH2PO4pH 7 buffe r. Follow ing ele ctrophore sis, the RNA w as trans ferre d ove rnight by c apillarity from the ge l to a nitro cellulose membrane using 203 SSC. This membrane w as drie d w ith Whatman pape r and the RNA w as c ros s-linke d to the me mbrane by ex posure to UV light. Then, the me mbrane w as pre hybridize d for 2 h at 65°C in a solution containing 6.4% dex tran, 33 SSC, 13 Denhart’s, 0.1% SDS and 250 m g/mL he at-de nature d salmon spe rm DNA. IL-6 cDNA and G3PDH cDNA probes comme rc ialize d by the Ame ric an Type Culture Colle ction (ATCC) w ere labelled w ith 32P using
Cy c lo spo rin a n d m a c roph a g e IL-6 sy n th e sis
the c ommercial ‘Re ady To Go’ kit (Pharmac ia Biote ch®) follow ing the protoc ol indic ate d by the manufac ture r. Hybridizatio n (18 h, 65°C) w as carrie d out using the same pre hybridiz atio n solution containing he at-de nature d 32P-labe lle d probes. Follow ing succ essive w ashings in 0.1 SDS-0, 1 SSC at room te mperature and 65°C, re spe ctive ly, me mbrane s w ere blotte d dry and use d for autoradio graphy w ith XR film (Cronex 105, Dupont®) at –70°C ove r 6 days . Hybridiz atio n w ith the G3PDH probe w as use d as an RNA loading c ontro l for blots.
Results Differentiation of the U937 line The data (not show n) c onfirming that suitable diffe rentiation of the monocyte ce ll line tow ard mac rophage cells had be en achie ve d w ere as follow s: a dec re ase in the ce llular prolife ratio n rate w ithout viability being affe c te d; an inc re ase in adhe re nc e to the culture supports and among c ells; an inc re as ed c ytoplasm, the disappearanc e of nuc lear polylobulation and a de cre ase in nuc le ar atyp ias and the number of nuc leoli; an inc re ase in the c onte nt of mac rop hage e nzyme s (non-spe cific este ras es); and an inc re ase in cytoplasmic RNA conte nt. d
d
d
d
d
Determination of type of stimulus and most effective concentration on interleukin-6 secretion In orde r to study the most suitable typ e of stimulus for induc ing the se cre tion of IL-6 and the most e ffec tive conc entratio n, U937 cells w e re used unde r both diffe re ntiate d and undiffe re ntiate d c onditio ns. Thre e classic stimuli w ere studie d at tw o diffe re nt conc entratio ns: lipopolysaccharide (LPS), phorbol myris tate acetate (PMA) and gamma-inte rfe ron (g -IFN). The me an value s of the re sults of the study carrie d out in quadrup lic ate are show n in Fig. 1. Ove rall, the data obtaine d indic ate that the most pote nt stimulus of cytokine sec re tion w as PMA at a conce ntratio n of 10–5M. IL-6 sec re tion w as gre ate r by diffe re ntiate d U937 ce lls than by undiffe re ntiate d one s. Effect of CsA on the synthesis of interleukin-6 The re sults of the effect of diffe re nt conc entratio ns of CsA on IL-6 se cre tion unde r both basal and stimulate d conditio ns are show n in Table 1. These data show the me an value s obtaine d in four diffe re nt as says. As may be se en, both in undiffe re ntiate d and diffe re ntiate d U937 ce lls, at the highe st conce ntratio n use d (200 ng/ mL) CsA dec re ased stimulate d se cre tion of IL-6. Cs A can be said to decre ase the sec re tion of this cytokine more than 50%. The effect of CsA on basal IL-6 se cre tion w as much low e r.
FIG. 1. Effect of different stimuli on IL-6 secretion by U937 cells. The most potent stimulus of cytokine secretion was PMA at a concentration of 10–5M. Definition of abbreviations: IFN, interferon; IL-6, interleukin-6; LPS, lipolysaccharide; PMA, phorbol myristate acetate. Note the logaritmic scale. Mediators of Inflammation · Vol 8 · 1999
255
J. E. Lo sa Ga rc ´õ a e t al. Table 1. Effect of CsA on basal and stimulated IL-6 secretion by U937 cells (pg/106 cells). Both in undifferentiated and differentiated U937 cells, at the highest concentration used (200 ng/mL) CsA decreased stimulated secretion of IL-6. Definition of abbreviations: CsA, cyclosporin A; PMA, phorbol myristate acetate; SEM, standard error of mean. Mean
SEM
Undifferentiated Basal + CsA (2 ng mL–1 ) + CsA (20 ng mL–1) + CsA (200 ng mL–1 ) + PMA 10–5 M + CsA (200 ng mL–1 )
38,4 37,6 34,0 25,2 1705,1 731,4
7,9 13,1 12,7 4,2 313,5 66,0
DMSO differentiated Basal + CsA (2 ng mL–1 ) + CsA (20 ng mL–1) + CsA (200 ng mL–1 ) + PMA 10–5 M + CsA (200 ng mL–1 )
90,4 99,7 91,4 67,1 6343,7 2035,1
27,6 33,3 34,2 20,6 3135,6 1072,6
Using U937 ce lls diffe re ntiate d w ith 1.3% DMSO, the intrac ellular conte nt of IL-6 w as me as ure d afte r stimulatio n w ith phorbol myristate in tw o diffe re nt assays. At 200 ng/mL, CsA dec re as ed the intrac ellular conte nt of IL-6 from 151,0 ± 97,2 to 100,0 ± 61,3 pg/ 106 cells. So, CsA can be said to decre as e the production of cytokine s sinc e both the intra-and ex trac e llular c onte nts w ere affe c te d. Effect of CsA on cell viability and cell proliferation capacity The firs t and simple st possibility conside re d w as that CsA w ould affe ct ce ll viability and/or p rolife ration and he nce that the dec re ase in c ytokine produc tion w ould be due to a dec re as e in these c apacitie s. According ly, a duplicate study w as pe rforme d in w hich the prolife ratio n capac ity and c ell viability of a U937 line subjecte d to the same stimuli as those use d in the study w ere evaluate d. Table 2 show s the re sults of this study. As seen, CsA did not affe ct either viability or prolife ratio n capac ity, either in undiffe re ntiate d U937
ce lls or in ce lls diffe re ntiate d w ith DMSO. There fore the re sults on c ytokine se cre tion canno t be inte rpre te d in the se te rms. In this study, the e ffec ts of se veral stimulato ry substanc es w ere also assessed. Both PMA and LPS, but not gamma-inte rfe ron, w ere seen to decre as e ce ll viability. Effect of cycloheximide on CsA inhibition of IL-6 production To study the mechanism by w hich CsA dec re as es IL-6 production, thre e w orking hypothe ses w ere addre ssed: that CsA affe cts the synthe sis of an early prote in, re spons ible for the de cre ase d le vels of the cytokine ; that it ex erts its ac tion at trans criptional level, and that its e ffec t is mediate d at post-trans criptional le ve l. The first mechanism proposed w as evaluate d by tre atment w ith c yclohex imide (an inhibitor of prote in synthe sis) of diffe re ntiate d U937 c ells during the first 3 hours of culture . Prio r to this study it w as asce rtaine d that cyclohex imide does not affe ct ce ll viability (data not show n). The e ffec t of cyclohex imide on the sec re tion of IL-6 is detaile d in Fig. 2. As may be see n, although ove rall this drug de cre ased both bas al and stimulate d cytokine se cre tion in diffe re ntiate d ce lls, w hich w as ex pe cte d, the inhibitory e ffec t of CsA persis te d in all case s. It is the re fore unlike ly that the me chanism of action of CsA w ould de pend on inhibition of the synthe sis of an e arly p rote in. Effect of CsA on expression of the IL-6 gene Study of the action of Cs A on the ex pre ssion of IL-6 mRNA by Northe rn blot w as carrie d out on the diffe re ntiate d c ell line unde r both basal and stimulate d conditio ns. Tw o as says w ere performe d and, in each, Northe rn blot w as imple mente d in duplic ate . Figure 3 show s the re sults of one of the assays . As may be seen, CsA did not affe c t the le ve ls of IL-6 mRNA unde r eithe r basal or stimulate d conditions .
Table 2. Effect of CsA and different stimuli on cell proliferation and viability. In this study we started out with a viable cell concentration of 250,000/mL, obtaining the data after 18 hours of culture. CsA did not affect either the viability or proliferation capacity of either differentiated or undifferentiated cells. Additionally, both LPS and PMA, but not interferon, decreased cell viability. Definition of abbreviations: CsA, cyclosporin A; LPS, lipopolysaccharide; PMA, phorbol myristate acetate Undifferentiated U937 cell line
Basal conditions PMA 10–5 M LPS (100 m g/mL) g -Interferon (1000 U/mL) CsA (200 ng/ML) CsA (200 ng/mL) + PMA 10–5 M 256
Differentiated U937 cell line
Viable cells/mL
Viability (%)
Viable cells/mL
Viability (%)
370.000 240.000 335.000 400.000 385.000 240.000
91 84 88 94 94 77
317.000 128.000 280.000 285.000 300.000 115.000
95 76 88 92 93 81
Mediators of Inflammation · Vol 8 · 1999
Cy c lo spo rin a n d m a c roph a g e IL-6 sy n th e sis
FIG. 2. Effect of cycloheximide on the action of CsA in U937 cells. The effect of cycloheximide was evaluated in differentiated cells. Although overall cycloheximide decreased both basal and stimulated cytokine secretion, the inhibitory effect of CsA persisted in all cases. Definition of abbreviations; CsA, cyclosporin A; IL-6, interleukin-6; PMA, phorbol myristate acetate.
Discussion Differentiation of the U937 cell line Study of the me chanism of ac tion of CsA on inflammatory c ytokine produc tion by human mac rophages could have bee n carrie d out in tissue mac rophages obtaine d ex vivo . How eve r, the limite d numbe r of ce lls availab le for the diffe re nt ex pe riments, together
w ith the diffic ulty invo lve d in obtaining healthy control ce lls, w ithin-group variab ility and, above all, the influe nce of the ac tion of the drug on othe r cells – espe cially lymphocyte s – counse lle d agains t this. There fore , to study the mechanis m of action of CsA on cytokine p roduc tion in mac rophages and to have available a suffic ie nt number of c ells, the possibility of using a monocyte c ell line – name ly, U937 c ells – w as implemente d. This had the additio nal advantag e of enabling study of a complete ly pure populatio n; that is, complete ly ex cluding the pre senc e of lymphocyte s, w hich might alte r the re sults obtaine d. The lite rature contains re fere nce s to use of the U937 ce ll line in the e valuatio n of diffe re nt macrophage func tions.14 This human c ell line grow s c ontinuou sly in suspe nsion and w as initially obtaine d from a pleural effusion in a patie nt w ith histiocytic lymphoma.15 Its phenotype c orre sponds to immature monocyte cells arre ste d in a diffe re ntiatio n state close to the mye lomonocytic ste m ce ll.15 It should be re calle d that macrophage s are highly diffe re ntiate d c ells and there fore have low prolife ratio n c apacity w hile c ell line s show a strong duplication rate . Thus, the first ste p in this part of the w ork w as to diffe re ntiate the human U937 monoc yte c ell line . In the pre senc e of se ve ral substanc e s U937 cells unde rgo a process of diffe re ntiatio n and the re fore acquire morphological and func tional charac te ristic s similar to those of mac rophages.16 Sinc e one of the inte re sts in the pre sent w ork w as to study the e ffec t of PMA on cytokine se cre tion, DMSO w as chosen as a
FIG. 3. Effect of CsA on IL-6 mRNA expression. Study of the action of CsA on IL-6 mRNA expression by Northern blot was carried out on the differentiated cell line under both basal and stimulated conditions. As seen, CsA did not affect IL-6 mRNA levels under either basal or stimulated conditions. Definition of abbreviations: CsA, cyclosporin A; GPDH, glyceraldehyde phosphate dehydrogenase; IL-6 interleukin-6; PMA, phorbol myristate acetate. Mediators of Inflammation · Vol 8 · 1999
257
J. E. Lo sa Ga rc ´õ a e t al.
diffe re ntiating agent be cause it displays diffe re ntiating capacity on the U937 c ell line .17 Diffe re ntiation w as e valuate d by kine tic , morphological, c ytochemic al and mole cular te chniques. The re sults obtaine d, w hich w ere similar to those re porte d by other authors,15,17 allow ed us to conclude that the protoc ol used is inde ed able to diffe re ntiate this cell line .
therapeutic c onc entrations in v ivo . Additionally, CsA decre as ed the intrac e llular conc entratio n of IL-6. According ly, in the diffe re ntiate d U937 ce ll line CsA can be said to dec re ase the produc tion of inflammatory c ytokine s. The effect of CsA on the U937 ce ll line is not w ell documente d. In a re vie w of the lite rature no re fere nc es to the se phenome na w ere found.
Choice of cellular stimulus DMSO-diffe re ntiate d U937 c ells re tain their capacity to be stimulate d late r on w ith diffe re nt agents,17 among them PMA.18 It should be note d here that diffe re ntiation and activatio n are not inc ompatible conc epts and that it is in fac t possible to e mploy se veral substanc es 16 to ac tivate ce lls diffe re ntiate d w ith other age nts. Here , w e e valuate d the effec t of thre e know n macrophage stimulating agents and observe d that for this purpose the most suitable agent w as PMA at a conce ntratio n of 10–5 M. Although ac cording to the lite rature low e r conc entratio ns of this substanc e are usually used,15,17 in each ex pe rime ntal syste m it is ne cessary to obtain e vide nce of the most e ffec tive conc entratio n in the induc tion of a give n effe ct. There fore in the pre se nt study the afore mentio ne d conce ntration w as use d. Although at this c onc entratio n PMA dire ctly decre ase s c ellular viability, from a quantitativ e point of vie w the inc re as e that it induc es in cytokine se cre tion canno t be attrib ute d to the eliminatio n of cytokine s by dead ce lls. Production of interleukin-6 by U937 cells Onc e the re quire d ex perime ntal conditions had be en establis hed, IL-6 sec re tion by U937 ce lls w as studie d. The se cre tion of IL-6 w as stronger in diffe re ntiate d than in undiffe re ntiate d c ells. Rec ently it has be en re porte d that the CD4 19 and CD2320 surfac e molecules pre sent in diffe re ntiate d MPS c ells are invo lve d in the se cre tion of IL-6 by U937 cells. Although none of the m use d a protoc ol ide ntic al to the one use d he re , diffe re nt studie s have documente d the produc tion of diffe re nt inflammatory c ytokine s by these c ells using bioassays,21 immunoassays 22–25 or molec ular biology te chnique s.21,24 Has s e t a l.22 did not de te c t IL6 secre tion, possibly due to the stimulatio n c onditions, and the value s described by Jiang e t a l.23 for IL6 are low er than our ow n, although the y use d diffe re nt protoc ols. Effect of CsA on interleukin-6 production Both in undiffe re ntiate d and diffe re ntiate d U937 ce lls, at the highest conce ntration use d use d (200 ng/mL) CsA dec re ased stimulate d se cre tion of IL-6, more than 50%, a value similar to that re porte d by Moutbarrik e t a l.26 In human monocyte s at 258
Mediators of Inflammation · Vol 8 · 1999
Mechanism of action of CsA on the production of interleukin-6 From the theore tic al point of vie w, the effe ct of CsA on the produc tion of inte rle ukin-6 could be due to a decre as e in cell viability and prolife ration and henc e to a de cre ase in the number of producing c ells, to the alte ration of an e arly prote in that w ould re gulate the gene ex pre ssion of cytokine s, to a dire ct action on gene trans cription, or to a modific atio n of posttrans c riptional events. The data obtaine d here allow us to ex clude a dire ct effect of CsA on ce ll prolife ration and viability. We c an also ex clude its ac tion at e arly le ve l sinc e inc ubatio n w ith c yclohex imide during the first hours of inc ubation did not affe c t the inhibitory effe ct of CsA. The ex pe cte d decre ase in IL-6 produc tion by U937 c ells due to the ac tion of cyclohex imide is c onsiste nt w ith the obse rvatio ns of othe r authors 27 w ho have re porte d that this drug inhibits stimulatio n of IL-1b synthe sis in the U937 cell line . Finally, this immunomodulator does not alte r IL-6 mRNA ex pre ssion, allow ing us to conclude that the effect of Cs A on IL-6 produc tion occurs at posttrans c riptional level. Re sults similar to those obtaine d by us have bee n publishe d by Kato e t a l.28 w ho observe d that CsA sc arc e ly affe c ts the ex pre ssion of the IL-6 gene in human periphe ral blood monocyte s. Like w ise , in a murine mast ce ll line Nair e t a l.29 re porte d that CsA doe s not alte r IL-6 mRNA ex pre ssion although it does affe ct IL-3 mRNA. It w ould thus appear that in the human MPS CsA has little effect on the trans c ription of monokine gene s sinc e the mRNA levels of IL-1,30 TNF30 and IL-8 31 are not affe cte d by CsA e ithe r. The ex act mechanism by w hich CsA ex erts its effect on cells of the MPS is not know n although it has bee n suggeste d that its e ffec t could be re late d to an inhibition of p rote in se cre tion/synthesis .32,33 If, as comme nte d earlie r, it is true that CsA inhibits the se cre tion of inflammator y cytokine s and does not affe c t the trans cription of the ir gene s, its action w ould appear to oc cur at post-trans c riptional le vel. In vie w of the de cre ase in immunogenic prote in, CsA may ac t at some ste p betw ee n mRNA and synthe sis of the prote in. Although this as pec t w as not studie d ex plic itly in the pre se nt w ork, the ide a that CsA might act on intrac e llular membrane s invo lve d in prote in synthe sis c ould be ente rtaine d. At least thre e
Cy c lo spo rin a n d m a c roph a g e IL-6 sy n th e sis
theore tic al possibilitie s could be advanc ed to account for this. Firs t, one could be de aling w ith a non-spec ific inte rac tion, ow ing to the lipophilic nature of this immunomodulator. In this sense , it has be en demonstrate d that othe r lipophilic substanc e s or lipid emulsions are able to inhibit c ytokine produc tion by macrophage s through this mechanism.34 Sec ond, although there is no conse nsus about the possibility, 35,36 CsA might inhibit prote in kinas e C activity in macrophage s. Thus, it has re ce ntly been observe d that inhibition of this e nzyme dec re ases TNF-a produc tion in LPS-stimulate d macrophage s.37 Finally, another possibility w ould be that one w ould be dealing w ith a spe cific ac tion derive d from the inte rac tion of CsA w ith c ytokine s associate d w ith intrac ellular membrane s.38 Furthe r studie s are e vide ntly re quire d to clarify all the se aspec ts. It should be note d that some authors have situate d the site of action of CsA on monokine se cre tion at the level of re le as e sinc e the y obse rve d a dec re as e in the se cre tion of TNF-a w ith normal ex pre ssion of the ge ne and normal intrac ellular cytokine synthe sis.13 How e ve r, the se studie s w ere c arrie d out in mouse MPS cells w hile those use d here w e re from a human sourc e. Acc ording ly, it may be c onc luded that at therapeutic and non-tox ic conc entratio ns in v ivo , CsA decre as es the produc tion of inte rleukin-6 by human macrophage s at post-trans criptional le ve l.
References 1. Fathman CG, Mye rs BD. Cyclos porine therapy for autoimmune diseas e. N Eng l J Med 1992; 326: 1693–1695. 2. He ss AD, Colombani PM, Esa AH. Cyclos porine and the immune re sponse: basic aspe cts . Crit Re v Im m u no l 1986; 6: 123–149. 3. Schre iber SL, Crabtre e GR. The mechanisnm of action of cyclosp orin A and FK506. Im m un o l To d a y 1992; 13: 136–142. 4. Re n K, Van Liew JB, Noble B. The effect of c yclosporin A on dise ase progre ssion in prolife rative immune comple x glomerulonephritis. Clin Im m u no l Im m un o pa th o l 1993; 66: 107–113. 5. He ss AD. Mechanisms of action of cyclos porine : considerations for the tre atme nt of autoimmune dise ases. Clin Im m un o l Im m un o p a th o l 1993; 68: 220–228. 6. Chapman I, Mazzoni L. Mechanisms of inhibition by cyclosp orin A on pulmonary leukoc yte accumulatio n. Tre n ds Pha rm a c o l Sc i 1994; 15: 99–101. 7. Ryffel B, Willard-Gallo KE, Tammi K, Loken MR. Quantitative fluore sce nce analys is of c yclosporine binding to human le ukocyte s. Tra n s p la n ta tio n 1984; 37: 276–280. 8. Mrow ie tz U, Christophers E. Effects of cyclos porine A tre atme nt on psoriasis. I: Influenc e of low -dose c yclosporine on human monocyte function in vitro. Tra n s pla nt Pro c 1988; 20(Suppl 4): 53–57. 9. Losa JE, Mate os F, Jim´e nez A, P´e re z JL. Action of cyclosp orin A on mononucle ar phagocytes. J Inv e s t Allerg o l Clin Im m u n o l 1996; 6: 222–231. 10. Losa JE, Mateos F, Mart´õn MR, e t a l. Evaluation of inflammatory cytokine se cre tion by human alve olar mac rophage s. Me dia t In fla m m 1999; 8: 43–51. 11. Tipton DA, Pabst MJ, Dabbous MK. Inte rleukin-1 beta- and tumor nec rosis factor-alpha-inde pe ndent monoc yte stimulation of fibroblast collage nase ac tivity. J Ce ll Bio ch e m 1990; 44: 253–264. 12. Rofe AM, Philc ox JC, Hayne s DR, Whitehouse MW, Coyle P. Changes in plas ma zinc, coppe r, iron, and he patic metallothionein in adjuvantinduced arthritis tre ate d w ith cyclos porin. Bio l Tra c e Elem Re s 1992; 34: 237–248. 13. Nguye n DT, Eskandai MK, DeForge LE, e t a l. Cyclosporin a modulation of tumor necros is factor gene ex pre ssion and e ffects in vitro and in vivo. J Im m u no l 1990; 144: 3822–3828. 14. Ke icho N, Saw ada S, Kitamura K, Yots umoto H, Takaku F. Effects of an immuno-suppressant, FK506, on interle ukin 1 alpha productio n by
15. 16. 17.
18.
19. 20. 21.
22. 23. 24.
25. 26. 27. 28. 29. 30.
31.
32. 33. 34.
35. 36.
37.
38.
human macropha ges and a macropha ge-like ce ll line, U937. Ce ll Im m u no l 1991; 132: 285–294. Minta JO, Pambrun L. In vitro induction of cytologic and functional differe ntiation of the immature human monocyte like ce ll line U-937 w ith phorbol myristate ace tate . Am J Pa th o l 1985; 119: 111–126. ¨ Oberg F, Botling J, Nilsson K. Mac rophage s and the cytokine ne tw ork. Tra n s p la n t Pro c 1993; 25: 2044–2047. Sakano T, Fujie A, Hamasaki T, Harada Y, Taniguchi H, Ue da K. Intrace llular Ca2+ mobilization in immature and more mature U937 induce d to differe ntiate by dimethyl sulfox ide or phorbol myristate ace tate . Ce ll Im m u no l 1988; 111: 390–397. Rzigalinski BA, Rose nthal MD. Effec ts of DMSO-induced differentiation on arachidonate mobilization in the human histiocytic lymphoma ce ll line U937: re sponsiveness to sub-mic romolar c alcium ionophore A23187 and phorbol este rs. Bio ch im Bio phy s Acta 1994; 1223: 219–225. Taimi M, Dornand J, Nicolas M, Marti J, Fave ro J. Involve ment of CD4 in interleukin-6 sec retion by U937 monoc ytic ce lls stimulate d w ith the lectin jac alin. J Le u k o c Bio l 1994; 55: 214–20. Paul-Eugene N, Amirand C, Ouaaz F, e t a l. Bioche mical and functional alterations induce d by CD23 ligation in the human promonocytic ce ll line U937. Im m u n o lo g y 1993; 80: 424–430. Taimi M, De facque H, Commes T, e t a l. Effec t of retinoic acid and vitamin D on the ex pre ssion of interle ukin-1 beta, tumour nec rosis fac tor-alpha and inte rleukin-6 in the human monocytic c ell line U937. Im m u n o lo g y 1993; 79: 229–235. Has s R, Lonnemann G, Mannel D, e t a l. Regulation of TNF-alpha, IL-1 and IL-6 synthe sis in differe ntiating human monoblastoid leukemic U937 ce lls. Le u k Re s 1991; 15: 327–339. Jiang WG, Puntis MC, Hallett MB. U937 ce lls stimulated w ith opsonise d zymozan particle s provide a conve nient laboratory source of tumour nec rosis factor alpha. J Im m u n o l Meth o ds 1992; 152: 201–207. Bosco MC, Gus ella GL, Espinoza-Delgado I, Longo DL, Vare sio L. Interferon-gamma upre gulate s interleukin-8 gene ex pre ssion in human monocytic c ells by a posttranscriptional mechanism. Blo o d 1994; 83: 537–542. Srivastava MD, Srivastava R, Srivas tava BI. Constitutive productio n of interleukin-8 (IL-8) by normal and malignant human B-ce lls and othe r ce ll type s. Le u k Re s 1993; 17: 1063–1069. Moutabarrik A, Takahara S, Namiki M, e t a l. Effec t of FK 506 and cyclosp orine on the ex pre ssion of IL-6 and its re ce ptor on stimulated monocyte s. Tra ns pla nt Pro c 1993; 25: 2320–2321. Be rge r AE, Carte r DB, Hanke y SO, McEw an RN. Cytokine re gulation of the inte rleukin-1 rec eptor antagonist prote in in U937 ce lls. Eu r J Im m u no l 1993; 23: 39–45. Kato K, Yokoi T, Takano N, e t a l. De tec tion by in situ hybridization and phe notypic charac terization of ce lls ex pre ssing IL-6 mRNA in human stimulated blood. J Im m un o l 1990; 144: 1317–1322. Nair APK, Hanh S, Banholze r R, Jirsch HH, Moroni C. Cyclos porin A inhibits grow th of autocrine tumour ce ll lines by destabilizing inte rleukin-3 mRNA. Na tu re 1994; 369: 239–242. Grane lli-Piperno A, Ke ane M, Steinman RM. Evidence that cyclos porine inhibits ce ll-mediated immunity primarily at the leve l of the T lymphocyte rathe r than the ac ce ssory c ell. Tra ns pla nta tio n 1988; 46 (2 Suppl): 53S–60S. Mrow ie tz U, Sticherling M, Mielke V, Schrode r JM, Christophe rs E. Neutrophil-ac tivating pe ptide 1/interleukin 8 mRNA ex pre ssion and prote in secre tion by human monoc yte s: effec t of c yclosporin A. Cy to k in e 1991; 3: 322–326. Roth J, Goebe ler M, Erpe nstein U, Sorg C. Differential re gulation of the macropha ge-specific surface antige n RM3/1 by cyclos porine , azathioprine, and dex ame thas one. Tra n s pla nta tio n 1994; 57: 127–133. Pigatto PD, Be rsani L, Altomare G, Mozzanica N, Finzi A. Effec ts of cyclosp orin A on psoariatic monoc yte functions and secre tion. Acta Derm Ven e re o l Su p pl Sto c hk 1989; 146: 165–167. Langadale LA, Maier RV, Wilson L, Pohlman TH, Williams JG, Rice CL. Liposome-e ncapsulated he moglobin inhibits tumor nec rosis factor re leas e from rabbit alve olar macropha ges by a posttransc riptional mechanism. J Le u ko c Bio l 1992; 52: 679–686. Chiara MD, Be doya F, Sobrino F. Cyclosporin A inhibits phorbol e ste rinduced ac tivation of supe rox ide productio n in reside nt mous e pe ritoneal macropha ges. Bio ch e m J 1989; 264: 21–26. We nzel-Seifert K, Schachte le C, Hummel R, Grunbaum L, Seifert R. Evidenc e that inhibition of phorbol e ste r-induc ed superox ide anion formation by cyclos porin A in phagocyte s is not mediated by direct inhibition of prote in kinase C. Bio ch e m Pha rm a co l 1994; 48: 859–864. Tschaikow sky K. Prote in kinase C inhibitors suppre ss LPS-induced TNF productio n in alve olar mac rophage s and in w hole blood: the role of enc apsulation into liposome s. Bio c him Bio p hy s Acta 1994; 1222: 113–121. Fruman DA, Burakoff SJ, Biere r BE. Immunophilins in prote in folding and immunosuppre ssion. FASEB J 1994; 8: 391–400.
Received 30 August 99; accepted 24 September 99 Mediators of Inflammation · Vol 8 · 1999
259
