Research Paper Mediators of Inflammation, 7, 355–361 (1998)

Previous studies have shown that mast cell granules (MCG) inhibit numerous macrophage functions including tumour cytotoxicity, superoxide and nitric oxide (NO) production, and FCgamma2a receptor-mediated phagocytosis. In this study, the effect of MCG on macrophage TNF alpha and nitric oxide synthase (iNOS) mRNA expression, and the production and fate of TNF alpha were examined. Upon activation with LPS+IFN gamma, macrophages expressed both TNF alpha and iNOS mRNA and produced both TNF alpha and NO. Co-incubation of LPS+IFN gamma-activated macrophages with MCG resulted in dose-dependent inhibition of iNOS mRNA expression. TNF alpha production in the activated macrophages was decreased by MCG, which was associated with a reduction in TNF alpha mRNA expression. MCG were also capable of degrading both macrophage-generated and recombinant TNF alpha. The direct effect of MCG on TNF alpha was partially reversed by a mixture of protease inhibitors. These results demonstrate that MCG decrease the production of NO and TNF alpha by inhibiting macrophage iNOS and TNF alpha gene expression. Furthermore, MCG post-transcriptionally alter TNF alpha levels via proteolytic degradation.


Introduction
Macrophages play a major role in microbicidal and tumoricidal ac tivity, antige n pre se ntation, and the production of a varie ty of cytokines and inflammatory me diato rs. 1,2 Mast c ells, on the othe r hand, play a ke y role in hype rsensitivity re ac tions by se cre ting mediators such as histamine, prote oglycans, various cytokines, metabolite s of arachidonate and unique prote as es. 3 -6 The re cruitment of macrophages to sites of mast c ell degranulation and the subsequent phagocytosis of granule s in vivo w as first re ported by Faw ce tt. 7 We have p re viously show n that mast ce ll granule s (MCG) inte rac t w ith rodent macrophage s and dow nre gulate superox ide production, 8 -1 0 tumour c ell killing and NO produc tion. 11 The se re ports c onfirm that mast c ell-mac rophage c ommunic ation has a re gulatory effect on host defenc e me chanisms and inflammation. A role for mast c ells in tumour grow th and me tastasis is evident by the pre senc e of an inc re as ed numbe r of mast ce lls at the periphery of certain tumours 12 -1 4 and by the angiogenic ac tivity of seve ral mast ce ll produc ts. 1 5,1 6 Similarly, mac rophage products such as NO and TNFa are implicated in tumour grow th and inflammation. The fac t that incre ase d numbe r of mast c ells are found in the periphery of tumours and MCG interaction inhibited mac rophage-me diated tumour c ell killing prompte d us to ex amine the mechanism of the effe ct of MCG on mac rophage NO and TNFa production. The foc us of this study w as, there fore , to ex amine the effect of MCG on mRNA ex pre ssion of iNOS and TNFa in mac rophages. The re sults de monstrate that MCG inte rac t w ith macrophage s, inhibit iNOS and TNFa mRNA ex pre ssion, and degrade TNFa .
Harvesting and culture of murine peritoneal macrophages Thre e days prior to harve sting peritone al ex udates, each mouse w as inje cte d intra peritone ally w ith 1.5 ml of a ste rile solution of prote ose pep tone (10% w /v). The pe ritoneal c avity of each mouse w as lavage d tw ice w ith 2.5 ml of minimum essential me dium containin g 15 mM HEPES, 100 units ml of penic illin, 100 m g/ml streptomyc in, 10% fetal bovine se rum (HMEM) and 5 units/ml of heparin. The poole d ce ll suspe nsion w as se dime nte d by c entrifugation at 250 3 g for 10 min, w as he d tw ice and re suspended in HMEM w ithout he parin. Aliquots of the ce ll suspension containin g 2 3 10 5 c ells w ere seeded in each w ell of a 96-w ell c ulture plate . Non-adhe re nt c ells w ere re move d by w as hing afte r 2 -4 h of incubation at 37°C. Mac rophages w ere activate d w ith LPS (100 ng/ ml) + IFNg (10 units/ml) in all ex perime nts, sinc e this conc entrations have be e n found to induc e optimum activation.

Isolation of mast cells
The method e mployed for the isolation of mast cells has be en pre viously described. 11 ,1 7 In brief, mast c ells w ere colle cte d by lavage of the pe ritone al and thoracic c avities of adult rats w ith 50 ml HMEM containin g 5 units/ml heparin. The lavaged ce lls from all animals w ere p ooled, c entrifuged at 250 3 g for 10 min at room te mperature , and w as hed tw ice w ith HMEM. Tw o ml of the ce ll suspension c ontainin g 6-8 3 10 7 c ells w e re gently laye re d on a 3 ml c ushion of 22.5% (w /v) metrizamide (dens ity 1.125 g/ml) in HMEM in a 15 ml c entrifuge tube, and c entrifuged at 200 3 g for 15 min at room te mpe rature. Mast cells w ere sedimented at the bottom of the conical tube w hile other c ells (pre dominantly macrophage s) c ollec te d at the inte rface . The mast c ell frac tions w ere collecte d, w ashed tw ice , and re suspe nded in HMEM w ithout heparin. Purity and viability of mast c ells isolated by this proce dure ex c eeded 95%.

Preparation of MCG and MCG sonicate
Under ste rile c onditions at 0 -4°C, MCG w e re prepared from me trizamide-purified mast ce lls by controlle d sonication and suc rose gradient c entrifugation. Mas t c ells w ere suspe nde d in 1 ml of HMEM and sonicate d tw ice for 20 s w ith a microtip sonic ator (Sonifier Ce ll Disruptor, model W140) at a p ow er se tting of 2.5 and te mpe rature of 4°C. 9,1 8 The disrup te d c ells w ere incubated at 30°C for 15 min and mix e d vigorously for 1 min. The sonic ate w as layere d on 2 ml of 0.34 M sucrose and centrifuged at 50 3 g for 10 min at 4°C. The granules at the interfac e w ere collecte d and sedimented by centrifugation at 1800 3 g for 20 min at 4°C. The re sulting pellet c onsisting of a homogene ous pre paration of MCG w as w ashe d tw ice and re suspe nded in the culture me dium. The re c overy of granule s isolated by this proce dure range d from 60% to 80% based on the his tamine content of the starting mast cells.
MCG-sonicate w as pre pared by sonic ating purifie d MCG in HMEM thre e time s for 30 s at max imum pow e r. The quantity of MCG and MCG-sonic ate use d in each ex pe riment w as ex pre sse d as the e quivale nt of the starting mast c ell numbe r.

Assay of TNFa
The conc entration of TNFa in the c ulture me dium w as assayed by both ELISA and bioassay. The sens itivity of ELISA is 31 pg/ml using the protocol re comme nde d by the manufacture r. For the bioassay, the targe t L929 c ells in RPMI-1640 me dium w ere see de d (4.2 3 10 4 pe r w ell) in a 96-w e ll culture plate . After 18 h of culture, me dium w as re placed w ith 5 m g/ml of dac tinomyc in in RPMI. Tw o to 4 h after inc ubation w ith dac tinomyc in, se rially dilute d sample s and TNFa standard s (2-250 pg/ml) w ere added and inc ubate d for an additional 18-24 h. The me dium w as then re move d, and 1 mg/ml of MTT dye in RPMI (w ithout se rum and p henol re d) w as added. After 2-3 h of inc ubation, the me dium w as aspirated c omp le te ly, and 100 m l of 2-propanol w as adde d to each w ell. The plate w as agitated for 10 min on an orbital shaker and re ad on a mic rotitre plate re ade r at 595 nm.

Analysis of iNOS and TNFa mRNA expression
Mouse peritone al c ells w hich c ontain ed > 85% mac rophage s w ere see de d in a 10 c m Pe tri dish and inc ubate d for 4 h to allow for macrophage adhere nc e. After w ashing thre e times w ith HMEM, the ce lls w e re tre ate d w ith MCG and activate d w ith LPS + IFNg . At se le cte d time p oints afte r activation, culture me dium w as re moved, and the total cellular RNA w as ex tracte d using an RNA ex trac tion kit. Total RNA (10 m g) w as ele ctrophore sed on 0.8% agarose-formaldehyde gel and the n transferred to Nytran nylon me mbrane. After 2 h of p re hybridization, the membrane w as hybridize d w ith 3 2 P-labelled cDNA probe spe cific for murine iNOS, TNFa or b -actin. The probe s w ere labe lle d by random hex amer priming me thod using [a -32 P] dCTP as p re viously described. 19 The membrane w as then w ashe d four time s and autoradiographed on a Kodak X-OMAT-AR film. A scanner w as use d to dete rmine the density of the mRNA band and the re lative de nsity of each band w as normalized to b -actin.

Treatment of MCG with protease inhibitors
MCG w ere incubated w ith a mix ture of p rote as e inhibitor s containin g PMSF 2.5 mM, pepstatin A 5 m g/ml, trypsin inhibitor 50 m g/ml, leupe ptin 50 m g/ml, and aprotinin 50 m g/ml, at 37°C for 2 h. The prote ase inhibitor-treate d MCG w ere then evaluate d for the ir e ffec ts on macrophages.

Statistical analysis
Whe ne ver applic able, the data w e re analysed by onew ay analysis of variance w ith subsequent Stude nt-New man-Kue ls' te st. All re sults w ere ex pre ssed as me ans ± SEM and P< 0.05 w as c onside re d signific ant.

Effects of MCG on macrophage iNOS mRNA expression
Earlie r w ork demonstrated that MCG inhibit macrophage NO produc tion. 11 To inve stigate the mechanism, the effe ct of MCG on macrophage iNOS mRNA ex pre ssion w as analyse d at se le cte d mast ce ll-tomacrophage ratios. The addition of MCG simultaneously w ith LPS+IFNg to macrophage monolayer inhibited iNOS mRNA ex pre ssion in a dose-depe nde nt manner (Fig. 1). Max imal inhibition w as note d at a MCG dose e quivalent to a mast c ell-to-mac rophage ratio of 1:5 and the inhibition w as e vident e ve n at a ratio of 1:40. When mac rophages w ere pre -inc ubate d w ith MCG for 3 h prior to activation, the inhibitory effect of MCG de cre ase d compared w ith simultaneous addition of MCG and the ac tivato rs. Furthermore , if mac rophages w e re incubated w ith MCG for more than 6 h p rior to ac tivation, there w as no inhibition of the ex pre ssion of iNOS transcript (Fig. 2).

Effects of MCG on TNFa production
Whe n ac tivated w ith LPS+ IFNg , macrophages generated large amounts of TNFa . Addition of MCG or MCG-S at the time of LPS+ IFNg stimulation re sulte d in dose-depende nt decre as e in TNFa le ve ls and more than a 90% dec re as e w as note d at a MCG dose equivale nt to a mast c ell-to-mac rop hage ratio of 1.5 to 2 (Fig. 3). The MCG-induc ed de cre ase in TNFa levels w as se en w he n dete rmine d by ELISA (Fig. 3A) and by bioassay (Fig. 3B). Although the assayed value s differe d for the same sample s depe nding on the assay me thod employed, the de cre ase of TNFa le vels w as e vide nt. There w as no statistic ally signific ant differenc e be tw ee n the effe cts of MCG and MCG-S. There fore , in later ex pe rime nts only the intact MCG w ere used. Unactivate d mac rophages or those ex pose d only to MCG or MCG-S did not produc e dete ctable levels of TNFa . The production of TNFa by mac rophages w as e vide nt as early as 1.5 h afte r activation. The c umulative TNFa level inc re ased for up to 24 h w hile the rate of produc tion appe are d to de cre ase (Fig. 4). The pre senc e of MCG de cre ase d the le ve l of TNFa by approx imately 40% at 1.5 and 6 h in contrast to a 95% decre as e at 24 h. This e ffec t w as also evident w hen TNFa w as as sayed by bioassay (data not show n).

Effects of MCG on TNFa mRNA expression
In order to as sess the e ffec t of MCG on TNFa mRNA ex pre ssion by ac tivated mac rophages, the transc ripts w ere analysed at 2, 6 and 24 h of inc ubation after activation. The re sult demonstrate s that TNFa mRNA w as rapidly ex pre sse d to a max imal le ve l w ithin 2 h after activation and then dec line d the re after (Fig. 5). Addition of MCG to LPS+ IFNg -stimulated mac rophage s inhibited TNFa mRNA ex pre ssion by 60% at a mast c ell-to-mac rophage ratio of 1: 3.

Effect of MCG proteases on TNFa degradation
Rat mast c ell granules contain a varie ty of prote as es including tryptase, chymase , and c arbox ype pti-Mediators of Inflammation · Vol 7 · 1998 dase . 20 -2 3 To inve stigate if TNFa is susce ptible to these enzymes, c onditione d media w ith know n levels of TNFa from LPS+ IFNg -ac tivated macrophage s or re c ombinant mouse TNFa w e re incubated w ith MCG for 24 h at 37°C. The TNFa le ve ls, assaye d by ELISA, indicate that both re c ombinant and mac rophagese cre te d TNFa w e re degrade d by MCG (Fig. 6). This effect w as partially re verse d by the p re -treatme nt of MCG w ith a mix ture of protease inhibitors containin g PMSF, pe pstatin A, trypsin inhibitor, le upe ptin and aprotinin (Fig. 7). The complete abrogation of TNFa degradation by MCG could not be achie ve d even after ex te nding the pre -treatment of MCG w ith the se prote ase inhibitors to 24 h at 4°C.

Discussion
The p re se nt study de monstrates that incubation of macrophage s w ith MCG during activation w ith LPS+ INFg re sults in the inhibition of iNOS and TNFa mRNA ex pre ssion w ith a conse que nt dec re ase in NO and TNFa p roduction. This corroborates our pre vious re port docume nting MCG inhibition of NO production and tumour c ell killing by LPS+ INFg -activate d macrophage s. 11 The MCG effe ct w as evident e ve n at an estimated mast ce ll to macrophage ratio of 1:40, a re lationship like ly to oc cur in v ivo . The inhibitory effect of MCG on macrophage iNOS mRNA ex p re ssion see ms transie nt be cause this e ffec t disappears w hen the macrophages w ere p re -incubated w ith MCG for 6 h or more . The same phe nomenon w as observe d w ith the inhibito ry effect of MCG on TNFa mRNA ex pre ssion (data not show n). Our earlier studies have ruled out the possibilities that MCG may inactivate LPS or IFNg . Further inve stigation is ne e de d to clarify the me chanism by w hich MCG inhibit macrophage iNOS and TNFa mRNA ex pre ssion in a transie nt manner.
The ex pre ssion of TNFa by activate d mac rophages is of p artic ular significance since the gene s enc oding iNOS and TNFa c an be c oordinate ly re gulated. In v ivo studies have suggeste d a role for TNFa in the positive re gulation of iNOS follow ing LPS injec tion. Anti-TNFa antib odie s or soluble TNFa re c eptor antagonists partially block LPS-induc ed pulmonary iNOS activity or hep atic iNOS mRNA ex pre ssion re spec tively. 2 4,2 5 Furthe rmore , the involvement of TNFa in both the induc tion and mainte nance of iNOS mRNA in mac rophage s w as re cently re porte d using an in v ivo mouse mode l. 26 The se studies indic ate that TNFa is an autoc rine re gulator of iNOS ex pre ssion. The TNFa mRNA ex p re ssion be gins as early as 0.5 h, 2 7 w here as iNOS mRNA c annot be de te c te d until 8 h after activation. 28 Our data also show that TNFa mRNA ex pre ssion by macrophages is optimal at 2 h after activation and then progressive ly dec re ases. Based on those data w e conclude that the depletion of mac rophage -se cre te d TNFa by MCG c ould c ontribute to inhibition of LPS-induce d iNOS mRNA ex pre ssion. It is clear that the MCG to macrophage ratio that comple te ly inhibits the ex pre ssion of iNOS mRNA only partially inhibits the ex pre ssion of TNFa mRNA. We do not yet have an ex planation for this phenomenon. It is possible that the iNOS ex pre ssion ne e ds a finite amount of TNFa to trigge r the gene .
Rodent mast ce lls c ontain histamine, serotonin, and prote as es including serine p rote as es, ne utral Re g u la tio n o f m a cro p ha g e NO a nd TNFa pro du ctio n by m a s t ce lls prote ase s, and carbox ypeptidase A. 2 0 -23 His tamine and serotonin at c oncentration s pre sent in mast ce lls (20 m g and 2 m g per million c ells re spe ctively) may induc e immunomodulatory e ffec ts. For instanc e, histamine up-re gulates macrophage synthesis of IL-1. 29 How e ver, histamine at conc entrations in the range of 10 -6 -10 -3 M (e qualling and ex cee ding that pre se nt in MCG use d in this study) faile d to affe ct TNFa and NO produc tion by mac rophages (data not show n). The se re sults are in agre eme nt w ith our earlier re port of MCG inhibition of mac rophage superox ide production, w hich show ed that unlike MCG, histamine and se rotonin w ere ineffe ctive in modulating the re spiratory burst. 9 The e ffec t of serotonin on macrophage NO and TNFa w as not te ste d in this study. It is also note w orthy that histamine and serotonin are short-lived and their effects are rapidly lost afte r mast cell degranulation. He parin is an important constituent of MCG and murine mac rophages possess heparin re c eptors. 30 A pre vious re port has show n that c ommercial he parin and MCG are c ap able of inhibiting Fcg 2a re ce ptor me diated phagoc ytosis in macrophage cell line. 3 1 Although comme rcial heparin differs from rat MCGhe parin, it is possible that MCG-hep arin may have some modulatory e ffec ts on macrophage iNOS and TNFa ex pre ssion. Our re sults pre se nte d here show that both MCG and MCG-sonic ate are capable of degrading TNFa w ith loss of immunoreactivity and biological activity, and that both se cre te d and re c ombinant TNFa are susc eptible. This indic ates that the decre as ed le vels of TNFa c ould be partially due to the prote olytic effe cts of MCG-p rote as es, in addition to the effect of MCG inhibition of macrophage TNFa mRNA ex pre ssion. How e ver, this p rote olytic effe ct w as only partially re ve rsed by treatme nt of MCG w ith a mix ture of protease inhibitors. The failure of prote ase inhibitors to comple te ly abrogate the MCG effe ct may be due to inc omp le te inhibition of the prote ase activity or due to the abse nce of specific inhibitor s for certain prote ase s.
Mas t ce ll degranulation gene rates the re lease of a varie ty of inflammatory molecules that dire ctly stimulate many ce ll types including epithelial cells, 3 2 fibroblasts 33 and endothelial ce lls. 17 This study is the first e videnc e that mast c ell granule s inhibit mac rophage TNFa and iNOS mRNA ex p re ssion. We also found that the mast cell proteases degrade re combinant and macrophage sec re te d TNFa . The p re -and post-transcriptional re gulation of TNFa produc tion in LPS-activate d mac rophages by MCG may c ontribute to the inhibition of iNOS mRNA ex pre ssion.