Research Paper Mediators of Inflammation, 7, 397–407 (1998)

The exposure of the macrophage cell line, J774 to mast cell granules (MCG) led to the formation of altered nuclear transcription factor proteins (NF-kappaBx), which had faster electrophoretic mobility than the p50 homodimer of NF-KB, but retained comparable DNA binding capacity. Antibodies to N-terminal peptides of p50, p52, p65 or c-Rel supershifted only a fraction of NF-kappaBx. Western blot analyses revealed that nuclear p65 and c-Rel were progressively degraded after exposure to MCG, whereas nuclear p50 appeared to be unaffected. In contrast, cytoplasmic p50, p65, c-Rel as well as IkBalpha remained intact after MCG treatment, although p52 was clearly degraded. In comparison to J774 cells, incubation of mouse peritoneal macrophages with MCG resulted in more extensive alterations to NF-KB proteins. The alterations in NF-KB proteins did not affect the expression of inducible nitric oxide synthase (iNOS) or TNF-alpha mRNA inJ774 cells. These data indicate that exposure of J774 cells to MCG leads to generation of altered nuclear p52, p65 and c-Rel, which retain intact N-terminal peptides, specific oligonucleotide binding and transactivating activity. On the other hand, in peritoneal macrophages, MCG induce more extensive modifications to NF-KB proteins with associated inhibition of iNOS or TNF-alpha mRNA expression.


Introduction
Macrophages play a ke y role in mic robic idal and tumoricidal ac tivity as sociated w ith the production of a varie ty of cytokines and inflammatory me diators. 1,2 Mast ce lls induc e hypers ensitivity re actions by sec re ting mediators such as his tamine, prote oglycans, various cytokines, metabolite s of arachidonate and unique prote ases. 3 -6 The c ytoplasmic granule s of mast ce lls are me mbrane -bound organe lle s, w hich contain bioge nic amine s, proteoglyc ans, prote as es and superox ide dismutase. 3 -8 The interaction be tw e en mast c ells and macrophage s, and the up take of mast c ell granule s (MCG 2 ) by macrophage s both in v ivo 9 and in vitro 8,1 0 is w ell docume nte d. Studies in mast ce ll de fic ie nt animals have indic ated the important role of mast cells in modulating inflammatory re spons e. 1 1 In v itro studies demonstrate d that phagocytosis of MCG by macrophage s re sulted in sup pre ssion of macrophage func tions such as Fcg 2a re c eptorme diated phagocytosis, 1 0 PMA-trigge re d superox ide gene ration, 1 2 LPS-induce d nitric ox ide production and tumoric idal ac tivity, 1 3 and LPS-induc ed transcription of mRNA for nitric ox ide synthase (iNOS) and TNF-a . 1 4 The transcription fac tor NF-k B has be en show n to be involved in LPS-induced ac tivation of the iNOS gene in mouse mac rophages. 1 5 Although the me chanisms of MCG re gulation of mac rophage function are unknow n, it is possible that phagocytose d MCG alter the structure and function of NF-k B. We there fore inve stigate d the effect of MCG on NF-k B prote ins and the ex pre ssion of iNOS and TNF-a mRNA in the mouse macrophage c ell line J774, w ith and w ithout LPS stimulation . The data pre se nte d in this paper indicate that ex posure to MCG re sults in transie nt de gradation of multiple transcription fac tors including nucle ar NF-k B proteins in J774 cells. The degrade d NF-k B proteins re tain DNA binding domains and nuclear localizing signal pe ptides re sulting in translocation to nuclei. Although the modified NF-k B prote ins translocate d to the nuclei are of low e r molecular w eight (de signated as NF-k Bx ) the y did not lose the ability to promote transc ription of iNOS and TNF-a mRNA in J774 ce lls. In mouse pe ritone al macrophage s, on the other hand, MCG ex posure re sulte d in more ex te nsive modulation in NF-k B prote ins, and re sulted in the inhibition of iNOS and TNF-a mRNA ex pre ssion.

Cells
The mouse macrophage-like c ell line, J774, w as derive d from a tumor of a fe male BALB/c mouse and posse sse s charac te ristics typ ic al of mac rophages. 1 6 Proteose p eptone -elicite d mac rophages w e re harveste d from tw o to three month old male C57Bl/6 mic e as de scribe d. 1 3 The c ulture media and other re age nts use d in the cell culture w ere asc ertaine d to contain <0.025 ng of LPS/ml by the Limulus ame bocyte lysate assay, and the ce lls w ere maintain e d in culture as de scribe d. 1 7-20 Mast cell granule preparation Briefly, mast ce lls w ere obtained by lavage of the peritone al c avity of male Sprague Daw le y rats w ith minimum essential medium c ontainin g 15 mM HEPES, penicillin (100 U/ml), streptomycin (100 m g/ml), 10% fetal bovine se rum and 50 m g/ml heparin (HMEM) as desc ribe d. 1 2,13 Ce lls w ere pooled, centrifuged for 15 min at 400 3 g at 25°C, w as hed tw ic e and re suspende d in HMEM. The ce ll suspe nsions (5 to 10 3 10 7 / 2 ml) w ere laye re d ove r 4 ml cushions of 22.5% me trizamide (dens ity of 1.125 g/ml) in HMEM and ce ntrifuge d at 2003 g for 15 min. Mast ce lls in the pellet w ere c ollec te d, w as hed tw ic e and re suspe nde d in HMEM. Mas t c ells isolate d in this manner ex c ee de d 90% in purity and viability. To pre pare MCG, the ce ll suspe nsion (in 2 ml of HMEM) w as sonicated for 15 s, coole d for 30 s on ice and re -sonicate d for 15 s at a pow e r setting of 2.5 w ith a microtip sonicator to re lease granules. 1 2,13 The disrupte d cells w ere incubated for 15 min at 30°C, vortex ed for 1 min, layere d over 2 ml of 0.34 M suc rose and c entrifuged at 50 3 g for 10 min at 4°C to re move de bris. The re sulting supe rnatant w as centrifuge d at 1800 3 g for 20 min at 4°C. The pellet c onsisting of a highly purifie d, homogene ous pre paration of granules w as w ashed and re suspended in HMEM. The quantity of MCG used in each ex pe riment w as ex pre ssed as the e quivale nt of the starting mast c ell number.

Assay for protease activity in MCG
Chymotrypsin-like (chymase) ac tivity w as as sayed by monitoring the hydrolysis of N-suc cinyl-L-phenylalanine-p -nitroaniline at 405 nm 2 1 in a re action mix ture (1 ml) c ontainin g 100 mM Tris-HCl (pH 7.6), 100 m M substrate and MCG sonic ate equivale nt of 2 to 4 3 10 5 mast ce lls. The sp ecificity of the enzymatic ac tivity w as confirme d by the ability of TPCK to inhibit the re ac tion. Trypsin-like (tryptase) activity w as as sayed by measuring the hydrolysis of be nzoyl-L-arginine -p-nitroaniline at 410 nm 2 2 in a re action mix ture (1 ml) c onsisting of 50 m M Tris-HCl (pH 8.2), 0.02 m M CaCl 2 , 100 m M substrate and MCG sonicate e quivale nt to 2 to 4 3 10 5 mast cells. Commercially available trypsin (Sigma, St Louis, MO) se rved as c ontrol. Carbox ype ptidase A ac tivity w as assaye d by monitoring the hydrolysis of hippuryl-L-phenylalanine at 254 nm 2 3 in a re ac tion mix ture (1 ml) c ontainin g 50 mM Tris-HCl (pH 7.5), 1 mM substrate and MCG sonic ate equivale nt to 2 to 4 3 10 5 mast ce lls. In all cases, one unit of e nzyme activity w as defined as the enzyme re quire d for hydrolyzing 1 m mole of substrate pe r min at 25°C.

Antisera
Antise ra directe d against various NF-k B and Ik Ba prote ins w ere raise d in rabbits by immunizing against synthetic pe ptide s coupled to ke yhole limpet hemocyanin (ge ne rous gift from Dr Nancy Ric e of NCI-Fre de rick Cance r Re search and Development Ce nter). The pe ptide sequenc es are as follow s: ADDDPYGTGQMFHLC (#1263, N-te rminus of mouse p50); CADMDFSALLSQISS (#1226, C-te rminus of mouse p65); CEQLSDPFTYGFFKI (#1266, C-te rminus of mouse c -Rel); FQPAGHGQDWAMEGRC (#751, N-te rminus of mouse Ik Ba ); and DEL-PYDDCVFGGQRLTL (#1258, C-te rminus of mouse Ik Ba ). 2 4 These peptide s w e re coup le d to he moc yanin either through the free amino group at the peptide N-te rminus or through an added c yste ine at the N or C te rminus. Each antiserum is capable of immunoprecipitation of the spec ific protein and doe s not c rossre act w ith other family membe rs. Additional antisera dire cte d against N-te rminus of p52 and p65 w ere purchas ed from Santa Cruz (Santa Cruz, CA).

Western blot analysis
The prote ins se parated by a standard SDS-PAGE w ere ele ctrophore tically transferre d to nitrocellulose membrane (0.2 m m pore , Schle icher and Schue ll, Kee ne , NH) using a Trans-Blot SD Se mi-Dry Ele ctrophoretic Transfer Ce ll (Bio-Rad, He rc ule s, CA) in 36 min at 5.5 mA/cm 2 . The me mbrane w as the n immers ed for 10 min at 25°C in TBST (20 mM Tris-HCl, pH 7.6, 137 mM NaCl, 0.1% Tw een 20) c ontainin g 22% non-fat dry milk, w ashe d 3 time s w ith TBST, and inc ubated for 1 h at 25°C w ith rabbit antisera dire cte d against various NF-k B prote ins in the absenc e or pre se nce of corresp onding antigen peptides. The me mbrane s w ere then w ashe d 3 time s w ith TBST and inc ubate d for 30 min at 25°C w ith perox idase-tagge d goat antib ody against rabbit IgG (Sigma). The membrane w as w ashed 5 times w ith TBST and developed by ECL syste m (Amersham, Arlington Heights, IL).

Nuclear protein extraction
Nucle ar prote in ex trac ts w ere p re pare d by the me thod of Dignam e t a l., 2 5 as described. 17 -2 0 Brie fly, ce lls (2.5 3 10 6 ) w ere w ashed and lysed w ith a Dounce homoge nize r at 4°C in 0.5 ml of buffer I (10 mM HEPES-KOH, 10 mM KCl, and 1.5 mM MgCl 2 , pH 7.9) c ontainin g 0.5% Nonide t P-40. All subse que nt proce dures w e re carrie d out at 4°C. The lysate w as ce ntrifuge d for 5 min at 1000g . The pellet obtaine d w as w ashed tw ic e w ith buffe r I containin g 0.5% Nonodiet P-40, and c entrifuged at 10,000 3 g for 5 min to c ollec t the nuclear pe lle t. Nuclear prote ins w ere ex tracte d from the pellet for 10 min w ith 40 m l of buffer II (20 mM HEPES, 420 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, and 25% glycerol, pH 7.9). After vigorous mix ing, the nucle ar suspens ion w as c entrifuge d (10,000 3 g for 5 min) and the re sultant supe rnatant w as diluted w ith 60 m l of dilution buffer (20 mM HEPES, 50 mM KCl, 0.2 mM EDTA, and 20% glyce rol, pH 7.9). The follow ing re agents w e re adde d to all buffers just be fore use: 0.5 mM dithiothreitol, 0.5 mM phe nylmethylsulfonyl fluoride , 1 m g/ml pepstatin A, and 10 m g/ml each of aprotinin, le upeptin (all from Sigma) and soybean trypsin inhibitor (Boe hringe r-Mannheim, Indianapolis, IN).

Electrophoretic mobility shift assay (EMSA)
The NF-k B-spe cific oligonucleotide, c ontainin g tw o tandemly arranged NF-k B sites (underline d) of the HIV-1 enhanc er (59 -ATCAGGGACTTTCCGCTGGG-GACTTTCCG-39 ), 17 the ISRE-spec ific oligonucleotide (59 -ACTGTCAATATTTCACTTTCATAATGGAAA-39 ), 26 an oligonucle otide c ontainin g a H2TF1 binding site of H2K b promoter (59 -GCCCTAGGGCTGGGGATTCCCC-ATCTCCACA-39 ), an unre late d oligonucleotide lacking a k B site (59 -AGGATGGGAGTGTGATATATATCCTT-GAT-39 ) 17 and their re spe ctive c omplimentary oligonucleotide s w e re synthesized in the Biote chnology Support Facility of the Univers ity of Kansas Me dic al Ce nte r. 17 ,1 8 The oligonucleotides spe cific for AP-1, SP-1, OCT-1 and CREB and their re spective complimentary strands w e re purchased from Promega (Madison, WI). Each oligonucle otide and its complime ntary strand w ere anne aled and the n e nd-labelled using g -[ 32 P]ATP (6000 Ci/mmol; New England Nuc le ar, Boston, MA) and T4-polynucleotide kinase (Amersham), as desc ribed. 17 -20 Nuclear protein ex trac ts (1 m g) w ith 3 m g poly (dI-dC).poly (dI-dC) (Pharmacia) and ~2 ng (100,000-400,000 cp m) of e ndlabelled DNA (w hich w as added last) w e re mix e d in 30 m l of EMSA buffe r and incubated for 30 min at 25°C after mix ing. The EMSA buffe r for NF-k B consis te d of 10 mM Tris-HCl buffer, p H 7.5, c ontainin g 40 mM NaCl, 1 mM EDTA, 1 mM b -me rc aptoe thanol, 4% glyce rol, 0.1% NP-40 and 1 m g/m l bovine serum albumin. Follow ing this initial binding re ac tion, 20 m l of the mix ture w e re elec trop hore se d at 15 V/cm for 1-1.5 h at 25°C through a nativ e 6% polyacrylamide gel, w hich w as pre pared in 45 mM Tris-borate buffer containin g 1 mM EDTA. Ge ls w ere the n proce sse d for autoradiography.

UV crosslinking
Briefly, bromodeox yuridine substituted HIV NF-k B oligonucleotide w as made by the prime r ex te nsion me thod 18 using HIV-e nhance r te mplate and the prime r, 59 -ATCAGGGAC-39 , Kle now fragment of DNA polymerase I (25 units, Gibco BRL) and [a -3 2 P]-dCTP (6000 Ci/mmol, Amersham). Nucle ar ex tract (10 m g) inc ubate d w ith 1 3 10 6 cpm of the probe w as subje cte d to EMSA. The w et EMSA gel w as UV-irradiate d at a distance of 5 cm w ith a trans-illuminator (312 nm, 8000 m W/cm 2 ) for 30 min at 0°C and autoradiographe d at 4°C. The bands corresponding to NF-k B prote ins w e re cut out, crushe d and e quilibrated for 16 h at 4°C in 100 mM Tris-HCl, pH 6.8, 50 mM DTT and 2% SDS. The sample s w ere the n mic rofuged and the re sultant supe rnatant s w ere subjecte d to e lectrophoresis on a SDS-10% polyacrylamide gel.
Northern blot analyses of nitric oxide synthase and TNF-a mRNA expression J774 ce lls (1 3 10 7 ) w e re incubated w ith or w ithout MCG for 3 h and w ere activate d w ith LPS (100 ng/ml) for 6 h. Total c ellular RNA w as ex tracte d using the guanidinium thioc yanate proc edure . 27 RNA (10 m g) w as e lectrophore sed in a 1% agarose ge l containin g 2.2 M formaldehyde and transferred to a Nytran nylon me mbrane w hich had be en p re hybridized w ith salmon spe rm DNA. The membranes w ere then inc ubate d for 20 h at 42°C w ith 32 P-labelled cDNA probe s specific for iNOS or TNF-a labe lled by the random hex ame r priming me thod using [a -32 P]dCTP. The me mbrane s w e re the n w ashed tw ic e at 25°C in 0.1% SDS and 2x SSPE for 30 min and tw ice at 42°C in 0.1% SDS and 0.1x SSPE. The me mbrane w as autoradiograp hed at -80°C w ith intensifying sc re ens . After stripp ing, the same membranes w ere re hybridize d w ithb -actin 3 2 P-labelled c DNA probe to control for RNA loading.

Effect of MCG on constitutive and LPS-activated NF-k B in J774 cells
The re sults pre sented in Fig. 1 show that the ex posure of J774 ce lls to LPS substantially inc re ased the nuclear levels of NF-k B1 (p50 homodime r) and B2 (p50/p65 he te rodimer) and induc ed a third c ompone nt, NF-k B3 (p65/c -Rel hete rodimer) (lane 5 vs 1). The ex posure of J774 c ells to MCG for a total of 1.5 h at the J774 ce ll to mast ce ll ratio of 500 to 1 (lane 2) or 50 to 1 (lane 3) did not affect the c onstitutiv e ly ex pre ssed le ve ls of NF-k B1 and B2. The cells ex pose d at a J774 c ell and mast c ell ratio of 5 to 1 (lane 4) had de cre ase d B1 and B2 le ve ls and c ontaine d a low level of a ne w specie s of nuclear p rote ins (designate d as NF-k Bx ) w hich bound to NF-k B oligonucle otide and had faste r ele ctrophore tic mobility than NF-k B1. The ex posure of MCG-pre treate d c ells to LPS also c ause d substantial activation of NF-k B p rote ins (lanes 6-8). The inte nsity of NF-k Bx gradually incre ase d as the c oncentration of MCG w as incre ase d (lanes 6 -8), w here as the bands corresp onding to NF-k B2 and B3 dec re ased or disappeared. The binding of LPS-induce d NF-k B p rote ins (including NF-k Bx ) pre sent in both MCG-tre ated and untreate d c ells to NF-k B probe w as spec ific , as the pre senc e of the 100 -fold ex cess of competitor H2K b (lane 9 and 11), but not unrelated oligonucleotides (lane 10 and 12), complete ly blocke d the binding. Based on the se re sults, the ratio of J774 cells to mast ce ll e quivale nts of 5 to 1 w as routine ly use d.

Detection of the degradation of NF-k B proteins by supershift analysis
Furthe r charac te rization of NF-k Bx prote ins activate d in MCG-pre tre ated J774 ce lls w as carried out by supe rshift analysis using antib odies spec ific for the N-te rminal pe ptide of p50, p52, p65 or c-Rel. These antib odie s supershifte d various NF-k B prote ins from LPS-tre ated J774 ce lls (Fig. 2 top) as ex pec te d from pre vious studies. 18 The pre incubation of nuclear ex tract from MCG and LPS-treate d ce lls w ith the same antib odie s re sulted in the c re ation of the supershifte d bands indic ated by arrow s (Fig. 2 bottom). How ever, FIG. 1. Effect of MCG exposure on NF-k B proteins in J774 cells. J774 cells were incubated for 1h at 37°C without MCG (lanes 1, 5, 9 and 10) or with MCG (lanes 2 to 4, 6 to 8, 11 and 12) at J774 cells to mast cell ratios of 500 to 1 (lanes 2 and 6), 50 to 1 (lanes 3 and 7) or 5 to 1 (lanes 4, 8, 11 and 12). They were further incubated for 30min at 37°C without (lanes 1 to 4) or with LPS (10ng/ml) (lanes 5 to 12). The nuclear proteins extracted from the cells were incubated with [ 32 P]-labeled HIV NF-k B motifcontaining oligonucleotide in the absence (lanes 1 to 8) or presence of 100-fold excess of either competitor H2K b (lanes 9 and 11) or unrelated (lanes 10 and 12) oligonucleotide, and were subjected to EMSA. The autoradiograph shown is the representative of three separate experiments, which gave very similar results.
FIG. 2. Supershift analysis of NF-k Bx proteins generated following phagocytosis of MCG. J774 cells were incubated for 3h at 37°C without MCG (top) or with MCG (bottom) at a macrophage to mast cell ratio of 5 to 1. They were then exposed to LPS (10ng/ml) for 30 min at 37°C. The nuclear proteins extracted from the cells at the end of incubation period were incubated at 25°C for 30 min with antibodies against the N-terminals of p50, p52, p65 or c-Rel or with control non-immune rabbit serum, and then subjected to EMSA as in Fig. 1. Autoradiographs were prepared by exposing to X-ray films for 14.5h. Arrows indicate the supershifted bands. Similar experiments carried out using antibodies against the C-terminal peptides of p65 or c-Rel did not show any supershifted band (Data not shown).
the majority of NF-k Bx prote ins w ere not supe rshifte d w ith any antibodies.
The approx imate size of NF-k Bx prote ins pre se nt in the nucle ar ex trac ts of the MCG-ex posed c ells w as then estimated by UV-crosslinking me thod. The autoradiograph of the gel (Fig. 3) re veale d that UVcrosslinked NF-k Bx prote ins migrated during SDS-PAGE primarily as tw o bands of p 45 -50 and p55 -65 (lane Bx , indicate d by arrow s). UV-crosslinked-NFk B1 (lane B1), k B2 (lane B2) and k B3 (lane B3) prote ins c onsiste d, as pre viously re porte d, 1 8 of p55, p55 plus tw o closely moving p75 and p80, and p75 -p80 prote ins, re spec tively. Various amounts of high molec ular w e ight prote ins se en in NF-k B2 and k B3 prote ins probably re pre sent aggre gated forms w hich are delete d by ex posure to MCG.

MCG degradation of other transcription factors
A possibility that the ex posure of J774 cells to MCG also affec ts transc ription factors other than NF-k B w as ex amined. Nuclear ex tracts from the ce lls w ere analyzed by EMSA for NF-k B, AP-1, CREB, OCT-1, SP-1 and ISRE using radiolab elled oligonucleotide specific for each transc ription fac tor. Results (Fig. 4) show ed that the constitutive le vels of transcription fac tors othe r than NF-k B w ere not markedly affec te d by the ex posure of the ce lls to LPS and that the ex posure of the ce lls to MCG re sulted in disappearanc e (AP-1) or substantial changes in elec trophore tic mobilitie s (CREB and SP-1) of these nuc le ar associate d transcription fac tors. In contrast, the le ve ls of OCT-1 and ISRE pre sent in the nuclear ex trac t re maine d unaffe cte d by the treatment of c ells w ith either LPS and/or MCG.

Kinetics of the effect of MCG on NF-k B degradation in J774 cells
The question of w hethe r MCG-trigge re d degradation of NF-k B prote ins is transie nt w as nex t investigate d.
To this e nd, five matched groups of J774 c ells w ere cultured for 24 h be fore ex posure to LPS. During this pre -inc ubation period, J774 c ells w ere ex posed to MCG for 24, 6, 3 or 0 h. One half of e ach group of Nuclear prote ins as w e ll as cytosolic prote ins w ere se parately subjecte d to w este rn analysis using antise ra spec ific for the N-or C-te rminus of p50, p52, p65, c -Rel and Ik Ba . Re sults of the EMSA (Fig. 5) show e d that NF-k B prote ins c onstitutiv ely ex pre ssed in the nuclei (lane 1) w e re substantially inc re as ed upon LPS stimulation as thre e discernible bands, NF-k B1, k B2 and k B3, re spective ly (lane 6), as note d above. Tre atme nt of the ce lls for 30 min w ith MCG alone (lane 5) caused a disapp earanc e of NF-k B3 and an incre ase in the inte nsity of the bands corresponding to NF-k B1 and k Bx . Tre atment of the se cells w ith LPS substantially inc re as ed the levels of NF-k B2, k B1 and k Bx (lane 10). The nucle ar proteins obtain ed from the c ells w hich w ere incubated for 6.5 or 3.5 h w ith MCG alone (lanes 3, 4) contained low le vels of NF-k B2 and k B1 and a substantial le vel of NF-k Bx , but no NF-k B3. Tre atme nt of the se ce lls w ith LPS re sulte d in the marked inc re ase of the le ve ls of these NF-k B proteins othe r than NFk B3 (lanes 8,9). The ce lls incubated for 24.5 h w ith MCG alone (lane 2) contained NF-k B1 and k B2, but ne ithe r NF-k B3 nor NF-k Bx . Tre atme nt of the se cells w ith LPS (lane 7) substantially enhanc ed the le ve ls of both NF-k B2 and k B1, but did not c ause the activation of NF-k B3 or k Bx . These data thus suggeste d that phagoc ytosis of MCG le d to a gradual loss of NF-k B3 and NF-k B2 and a progressive gene ration of NF-k Bx in J774 c ells by 6.5 h. Judged by the re sults w ith c ells w hich w e re incubated for 24.5 h w ith MCG alone or MCG plus LPS, the de gradation of NF-k B proteins w as transie nt but at 24.5 h after the initiation of phagocytosis, re c overy w as not c omple te . The samples in lanes 2 and 7 containe d no NF-k B3 and possibly a re duced level of NF-k B2 (c omp are lane 7 to lane 6) after LPS stimulation.

The effect of MCG on NF-k B proteins in mouse peritoneal macrophages
The e ffec t of MCG on the NF-k B p rofile in control and LPS-activate d murine peritone al mac rop hages w as also ex amined. Four matched samp le s of mac rophage s w e re cultured for 6 h p rior to LPS addition. During this pre -incubation pe riod, macrophages w ere ex pose d to MCG for 6, 3 or 0 h. One half of each group of MCG-treate d ce lls w as the n ex posed for 30 min to LPS (10 ng/ml), w he re as the othe r half se rved as c ontrol. Thus, all groups w ere c ulture d for a total of 6.5 h w ith variable time of ex posure to MCG. Nucle ar prote ins w ere then ex trac te d from each group of cells and w ere subjec te d to EMSA using radiolabe lle d NF-k B oligonucle otide as ligand.
The re sults (Fig. 6) show e d that all NF-k B prote ins cons titutiv ely ex pre ssed in the nuclei w e re marke dly inc re as ed upon LPS stimulation. Tre atment of the cells w ith MCG in the absenc e of LPS c ause d in virtual  10). They were cultured without MCG (lanes 1 and 6) or with MCG for 24h (lanes 2 and 7), 6h (lanes 3 and 8), 3h (lanes 4 and 9) or 0h (lanes 5 and 10) at a macrophage to mast cell ratio of 5 to 1. One of each matched group of MCG-untreated (lane 6) or -treated cells (lanes 7 to 10) was then exposed for 30min to LPS (10ng/ml), whereas the other half was not. Thus, all groups of the cells were cultured for a total of 24.5h. After washing the cells, nuclear proteins were extracted from each group of cells and were subjected to EMSA using radiolabelled NF-k B oligonucleotide as ligand as described in the Experimental section. An additional experiment gave a similar result to that shown.
disapp earanc e of nucle ar NF-k B1, B2 and B3, w ith the appearanc e of a k Bx population at all times studie d. As ex pe cte d, treatme nt of mac rophages w ith LPS re sulte d in the marke d incre ase of the le vels of the NFk B prote ins. Signific antly highe r amounts of nuclear NF-k B proteins including NF-k Bx w ere note d in macrophage s treated w ith MCG and LPS simultaneously. The nuclear NF-k B1, B2 and B3 w ere absent by 3.5 h w ith a progre ssive loss of k Bx at 6.5 h in LPSstimulated ce lls.
Location and mechanism of NF-k B protein degradation following MCG exposure The que stion of w hether the progressive degradatio n of NF-k B prote ins follow ing MCG phagoc ytosis occ urs in nuclei or in cytosol w as inve stigate d by w este rn analysis of nuclear and c ytosolic prote ins ex tracte d from J774 ce lls.
Ide ntific ation of nucle ar frac tions w ith anti-N-te rminus antib odie s (Fig. 7a) show ed that p hagocytosis of MCG for 0.5, 3.5 and 6. We ste rn blot analysis using antib odies against the C-te rminus of p 50, p52, p65 or c -Rel (Fig. 7b) show e d quantitativ e changes to nuclear p50, p52, p65 or c -Rel in re sp onse to LPS ac tivation in both c ontrol and MCG-treated cells. LPS-induced transloc ation and MCG-induce d breakdow n of these NFk -B prote ins are clearly evide nt. How e ve r, the antib odies against C-te rminus of p65 or c-Re l failed to re c ognize smalle r molecules w hich w ere re adily re cognize d by antibodie s against N-te rminus of p65 or c-Re l, sugge sting that MCG c ause d the loss of antigenic de te rminants in the C-te rminal re gions of p65 and c-Rel.
The que stion of w he the r phagocytosis of MCG causes prote olytic de gradation of c ytosolic NF-k B and/or Ik Ba prote ins w as also investigated by Weste rn analysis. The cytosolic fractions w ere pre pared from J774 ce lls, w hich w ere treated w ith MCG alone, or MCG plus LPS under identical condition s as utilize d for re cove ry of nuclear frac tions. We ste rn analyses carried out using anti-N-te rminus or anti C-te rminus antib odie s re veale d that the cytosolic p50, p52, p65, c-Re l and Ik Ba w e re minimally affe cte d by MCG tre atment (data not show n). LPS tre atment of c ontrol or MCG-treated ce lls re duc ed the amounts of p50, p52, p65, and c-Rel, most like ly due to nuclear translocation of the se prote ins. The le ve ls of cytosolic Ik Ba w ere also re duced follow ing LPS treatme nt, most likely due to LPS-induce d proteolytic degradation. 24 ,2 6 Extracellular degradation of cytosolic NF-k B proteins by MCG Rat MCG contain se veral prote ase s, including chymase and carbox ype ptidase A. 2 1,2 3 ,2 8 , 2 9 In orde r to ex amine w hether MCG-prote ase s could ac t on isolated c ytosolic NF-k B/Ik Ba proteins, the cytosolic fractions of J774 ce lls w ere pre pared, inc ubate d for 1 h at 37°C, w ith or w ithout MCG, at a J774 ce ll and mast c ell ratio of 5 to 1, and w e re then subjec te d to Weste rn analysis using antib odies direc te d against the N-te rminus of p50, p52, p65, c -Rel, or Ik Ba . The re sults (Fig. 8) show e d that the inc ubation of the isolated cytosolic fraction w ith MCG c aused substantial loss of all c omp one nts of NF-k B/Ik Ba p rote ins, suggesting their degradation by MCG-de rive d prote ase s.
The observe d ge ne ration of NF-k Bx p rote ins following phagocytosis of MCG c ould be due to proteolysis of NF-k B prote ins by prote ase s know n to be pre se nt in MCG. 2  production, 13 and iNOS and TNF-a mRNA ex pre ssion. 1 4 The virtual disappearanc e of nuc le ar NF-k B prote ins in peritone al macrophages ex posed to MCG for 6 h (Fig. 6) imp lic ated this trancription fac tor in the MCG e ffec t. In the pre se nt study, w e also investigate d w hether or not the transie nt and low le vel degradation of NF-k B prote ins in J774 ce lls, c omp ared to peritone al macrophage s, alte re d the ex pre ssion of iNOS and TNFa mRNA in J774 ce lls. Northe rn blot analysis (Fig 9) show that J774 cells w hich w e re pre -treate d w ith MCG for 3 h and the n stimulate d w ith LPS for 6 h, ex pre ssed the same amounts of iNOS and TNF-a mRNA as LPSstimulated c ontrol ce lls. LPS-induced NO production by J774 ce lls w as also not affe cte d by MCG treatme nt (data not show n). TNF-a mRNA le vels w ere also unchanged w hen analyzed after 30 min of LPS ac tivation of J774 ce lls treate d w ith MCG for 0, 3, 6 and 24 h (data not show n).

Discussion
The data p re sented in Figures 1-6 collective ly show that nucle ar NF-k B prote ins (partic ularly p65 and c-Re l) in J774 ce lls (Figs 1-5) and in mouse pe ritoneal macrophage s (Fig. 6) are proteolytically modifie d upon ex posure of the phagoc ytes to MCG at a macrophage to mast c ell ratio of 5 to 1. The alte rations to nucle ar NF-k B proteins in mouse pe ritone al macrophage s is ex te ns ive w hen compared to J774 cells and is in agre eme nt w ith the pre viously obse rved inhibition of iNOS and TNF-a mRNA ex pre ssion. 13,14 The re sults of Weste rn blotting ex pe riments (Fig. 6) clearly show that nuclear p52, p65 and c-Re l, in either their cons titutiv e or LPS-ac tivated forms, undergo p ro-gre ssive de gradation for up to 6.5 h afte r the initiatio n of MCG p hagocytosis , as se en by the de cre ase in the amounts of intac t forms and by the incre ase d gene ration of fragme nts smalle r than the pare nt molec ule s. De gradation of these prote ins is transie nt w ith the max imal effect noted at 3-6 h and some re sidual effects w e re noted e ve n at 24 h afte r the addition of MCG. Nucle ar p50 appears to be most re sis tant to degradation by MCG c ompare d to the othe r NF-k B prote ins. Ik Ba prote in in the cytosol of J774 c ells re mains intac t follow ing phagocytosis of MCG, but is degrade d upon ex posure of the c ells to LPS, as pre viously note d. 3 0 Mas t c ell granules used in the se ex perime nts w ere found to c ontain at le as t tw o differe nt active p rote as es, carbox ype ptidase A and chymase. If the contents of MCG are re leased into c ytosol follow ing phagoc ytosis, the se prote as es are c ap able of de grading c ytop lasmic prote ins such as the NF-k B/Ik Ba complex (Fig. 8). How ever, We ste rn analysis (Figs. 7) clearly show ed that phagocytosis of MCG leads to more ex te nsive proteolytic degradation of nucle ar NFk B proteins than c ytosolic NF-k B/Ik Ba proteins. This could be due to the prote c tion of cytosolic NFk B/Ik Ba prote ins from prote olytic dige stion, be cause these e nzyme s are c ontain ed w ithin the phagolysosome follow ing p hagocytosis of MCG by J774 ce lls. A possibility ex is ts, how ever, that proteolytic enzyme s in MCG w hich leak out into the c ytop lasmic c omp artme nt from the phagolysosome may degrade NF-k B prote ins c omplex ed w ith intact Ik Ba in c ytosol. Subse quent LPS tre atment activates c ellular TPCKse nsitive prote ase w hich degrades Ik Ba 3 0 Prote olytically alte re d and LPS-ac tivated NF-k B prote ins then translocate to nuclei and bind to NF-k B motifs, if NFk B prote ins re tain the ir nuclear loc alizing signal pep tide s and DNA-binding domains. Alternativ ely, the association of NF-k B, Ik Ba and the Ik Ba -associate d PKAc prote ins 31 in the cytosol may prote ct the compone nts from prote olysis by MCG enzymes if pre sent in the cytoplasm. The observation that NF-k B prote ins translocated to the nucleus constitutiv ely or by LPS stimulation and are transie ntly degraded by ex posure of the ce lls to MCG, sugge sts that c omp one nts of the MCG may also be translocate d to the nuclei. The re sults of Weste rn blot analyse s are cons is te nt w ith the idea of intra-nuclear degradation of NF-k B prote ins. This intriguing question and identification of the possible me chanisms by w hich MCG induc es alte rations to the nuclear NF-k B p rote ins re quires furthe r investigation. Another MCG compone nt, sup erox ide dismutase 3 ,7 may also participate in the re gulation of NF-k B. How e ve r, the pre senc e of superox ide dismutase does not ex plain the app earance of NF-k Bx , be c ause the changes in the patte rn of NF-k B in mac rophages ove r-ex pre ssing supe rox ide dismutase 32 are totally diffe re nt from those note d afte r phagoc ytosis of MCG.
Effe ct o f m a s t ce ll g ra nu le s o n NF-k B in m a cro ph a g e s FIG. 9. Effect of MCG exposure on LPS-induced iNOS and TNF-a mRNA expression in J774 cells. J774 cells were pretreated with MCG at a J774 cell to mast cell ratio of 5 to 1 for 3h and were then stimulated with LPS for 6h. Total RNA extracted at the end of the incubation were electrophoresed on 1% agarose gel, transferred to a Nytran nylon membrane, and sequentially hybridized with specific 32 P-labelled cDNA probes.
The data of Fig. 5 show that the e ffec ts of MCG phagoc ytosis on NF-k B proteins are transie nt. The ele ctron mic roscopic studies pre viously demonstrate d that the majority of phagocytosed MCG are eliminate d from pe ritone al mac rophages through a phagolysosomal pathw ay w ithin 45 min after the initiation of phagoc ytosis. 8 How e ve r, our data demonstrate that pre dominant NF-k B spe cie s in the nuclear ex tract from J774 ce lls at 6.5 h after phagoc ytosis of MCG are the prote olytic ally modified NF-k Bx prote ins. This is p ossible if the prote as es derived from MCG afte r phagoc ytosis re main ac tive and are associate d w ith J774 cells at 6 h after granules are no longer intac t in J774 cells. Alternativ ely, NF-k B prote ins fragmented follow ing phagoc ytosis of MCG may re main for at le ast 6.5 h and may not be re plac ed by ne w ly synthe size d NF-k B prote ins, if the cons titutiv e synthesis of NF-k B is inhibite d by the oc cupation of NF-k B site s w ith p65 or c-Re l w hich re tains DNA binding domains but have lost C-te rminal re gions. The finding that NF-k B prote ins ex trac te d from the c ells at 24.5 h afte r phagocytosis are mostly intac t forms of p50, p52, p65 and c-Re l could be ex plained by inactiviation of p hagocytose d MCG-derived prote as es and the synthe sis of ne w NF-k B proteins.
The prote olysis of NF-k B proteins is limited to sites distal to the nuclear loc alizing signal peptide sequence s of these proteins, le aving DNA-binding domains intact, as evident by the ability of NF-k Bx to spec ifically bind to NF-k B oligonucleotide (Fig. 1). Thus, the prote olytic ally alte re d p65 and c -Rel re tains DNA binding and transac tivating domains in spite of prote olytic degradation from their C-te rminal ends (Fig. 7). This may be due to the ac tion of carboxypeptidase A derived from MCG. It is intere sting to note that p50 w hich doe s not posse ss a transactivating domain but posse sse s both nuclear loc alizing signal pe ptide and DNA-binding domain is re latively unaffec te d by phagocytosis of MCG. The lack of inhibition of iNOS and TNF-a me sse nger RNA ex pre ssion in MCG-ex p ose d J774 c ells may be ex plaine d by the transie nt nature of NF-k B de gradation and re te ntion of the DNA binding and transac tivating domains. The ex te nsive prote olytic modifications to NF-k B prote ins in p eritone al macrophages by MCG (Fig. 5 vs Fig. 6) as c ompared to the re lative ly moderate alterations in J774 c ells may ex plain the inhibition of iNOS and TNF-a ge ne ex pre ssion in the forme r and the lack of inhibition in the latte r. Although the underlying mechanism for the diffe re nce in MCG effect in these tw o c ell type s is unknow n, the association be tw e en the magnitude of p rote olytic alterations to NF-k B and the func tional changes are inte re sting. It is note w orthy that in spite of the ability of MCG prote ase s to degrade all NF-k B proteins w hen inc ubate d w ith c ytosol pre parations (Fig. 7), MCG minimally affe cte d the NF-k B prote ins in the c ytosol of intac t ce lls after the ir phagoc ytosis. This suggests that in intac t cells the cytosolic prote ins are prote c te d from MCG ac tion. The MCG p rote as es may also activate other unidentifie d prote ase s in the host ce ll w hich function at the nuclear level to modify translocated NF-k B prote ins.
Mas t cell de granulation is know n to ge ne rate and re lease a numbe r of factors that dire ctly stimulate a varie ty of ce lls such as e pithe lial c ells, 33 fibroblasts 3 4 and endothe lial c ells. 3 5 This is the first e videnc e w hich demonstrate s that the uptake of MCG by cells leads to alteration s in the struc ture and possibly the functions of transc rip tion fac tors of the NF-k B/Re l family w hich are important re gulators of inflammation, c ell prolife ration and apoptosis. The ex planation for the variable func tional re sponses of pe ritoneal macrophage s and J774 ce lls to MCG re quire s further study.