Isolation and characterization of a macrophage-derived high molecular weight protein involved in the regulation of mucus-like glycoconjugate secretion Edith G. Gollub, and Zvi Marom,
PhD,* Satindra K. Goswami, MD* New York, N.Y.
PhD,* Kirk Sperber,
MD,**
Pulmonary macrophages release a variety of mediators that are involved in injlammatory processes and probably are involved in respiratory mucus secretion. Conditioned media obtained from activated pulmonary macrophages were found to contain a protein that functioned as a secretagogue for mucus-like glycoconjugate (MLGC) in an in vitro bioassay. A human macrophage-derived hybridoma cell line, HB-63, exhibited the same properties and was very useful in obtaining large amounts of the protein for purtjication and characterization. With ultrafiltration membranes and gel electrophoresis, the protein isolated from the conditioned media of zymosan- or lipopolysaccharide-treated cells was found to have a molecular weight of approximately 68 kd. The purified protein obtained from hybridoma cells and from pulmonary macrophages exhibited strong biologic activity when it was used to stimulate MLGC secretion, both in human airway explants and in an in vitro human secretory epithelial cancer cell line. The proteins from both sources were found to have similar amino acid compositions. Preliminary results indicate the presence of the 68 kd protein in the bronchoalveolar lavage fluid of a patient with severe chronic bronchitis and mucus hypersecretion. The role of this novel protein in the lungs is, so far speculative. The 68 kd protein may be a useful tool for studying the biosynthesis and. regulation of MLGC secretion and hypersecretion. (J ALLERGY CLIN IMMUNOL 1992;89:696-702 .) Key words: Pulmonary macrophages, hybridoma, mucous glycoprotein secretagogue, 68 kd protein
Pulmonary macrophages play a variety of roles in the respiratory system. They constitute most phagocytic cells in the lower respiratory tract and are important components of inflammatory and immunologic processes in the lung. Since they secrete numerous products, macrophages can interact with, and affect, other cells. For example, through their release of var-
From the Divisions of *Pulmonary and Critical Care Medicine and **Clinical Immunology, Mt. Sinai Medical Center, New York, N.Y. Received July 2, 1991. Revised Oct. 18, 1991. Accepted Oct. 23, 1991. Reprint requests: Edith G. Gollub, PhD, Division of Pulmonary and Critical Care Medicine, Box 1232, Mt. Sinai Medical Center, One Gustave Levy Place, New York, NY 10029. Dr. Edith G. Gollub is a recipient of the New York Lung Association ResearchAward. Dr. Zvi Marom was supportedby National Institutes of Health Grant NHL-37254. 111134525
696
Abbreviations used
HBSS: MEM: SI: MMS: LPS: MLGC: HB-63: PBS: Tw 20: MW: BSA: PD I, PD II: SDS-PAGE:
Hanks’balanced salt solution Minimal essentialmedia Secretory index Macrophagelmonocyte-derived mucus secretagogue Lipopolysaccharide
Mucus-like glycoconjugate Hybridoma cell line No. 63 Phosphate-bufferedsaline Tween 20 Molecular weight Bovine serum albumin Periods I and II Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
ious cytokines, including inhibitory as well as growth factors, lung macrophages take part in both cell damage and tissue repair.
VOLUME NUMBER
Macrophage protein in mucus Feguiation
89 3
Among the many secretoryproducts synthesized by macrophages,a novel peptide, MMS, has been described.‘2. This peptide was found to stimulate MLGC secretion in vitro, both in human airway explants”’and,morerecently,in a secretoryepithelial cell line.‘. 4 Studiesindicatedthat MMS is an acidic proteinwhoseM W is estimatedat approximately2000 daltonsby gel-filtration procedures.’ More recently, a polyclonalantibodywas raisedin rabbitsagainstthe MMS peptideafter its conjugationto crab hemocyanin.’This antibodyhas proved useful in our additional studiesof MMS and in the indentificationof other possibly relatedproteins. In this study, we describethe isolationandcharacterization of a new protein. which cross-reacts with theMMS antibody,functions as an MLGC secretagogue in vitro, and is elaboratednot only by pulmonarymacrophages but also by a macrophage-derived HB-63.5,’ This new macrophage-derived secretagogue may be an important m e d iatorof mucussecretionin vivo and may help to clarify the role of pulmonarymacrophages in inflammatory lung diseasesassociatedwith mucushypersecretion. Use of the hybridomacell line hasfacilitated the isolation and purification of the new protein, and most of the experimentsdescribedin this study were performedwith the hybridoma-derived protein. MATERIAL
AND METHODS
MEM, RPM1 1640 media, and HBSS were from Grand Island Biologicals, Grand Island, NY. Ficoll-Hypaque, protein A-Sepharose Cl-4B, and Q-Sepharose fast-flow anionexchange gel were from Pharmacia/LKB Chemicals, Piscataway, N.J. Zymosan particles, LPSs, BSA, and Tw 20 were from Sigma Chemical Co., St. Louis, MO. Dimiscintscintillation fluid was from National Diagnostics, Manville, N.J. Precast polyacrylamide minigels were from Integrated Separation Systems, Hyde Park, Mass. Nitrocellulose membranes were from Schleicher & Schuell (Keene, N.H.). Ultrafiltration membranes were from Amicon Corp., Danvers, Mass. Centripor concentrator tubes were from Spectrum Medical Industries, Inc., Los Angeles, Calif. Goat antirabbit and rabbit antiguinea pig horseradish-peroxidase conjugates and 4-chloro-1-naphthol were from Bio-Rad Laboratories. Richmond, Calif. ‘*51-labeledprotein A and [ ‘Hlglucosamine were from Amersham Corp., Arlington Heights. Ill. X-ray film was from Eastman Kodak Co.,
Rochester, N.Y. Polyclonalantibodies againstthe 68 kd protein were generated from guinea pigs by Pocono Rabbit Farms, Canadensis, Pa., with gel-purified 68 kd protein, as described later in METHODS.
Isolation, culture, and stimulation pulmonary macrophages
of human
lsolation of pulmonary macrophages has previously been described.’ In brief, the macrophages were isolated from surgical specimans of human lungs, maintained at 22” C in
697
HBSSwithoutCa-- or M g ”. Fragments of parenchyma wereteasedapart,filteredthroughsterilewire mesh,and centrifuged; cell pellets were pooled, layered onto FicollHypaque, and centrifuged. The mononuclear cell preparations were removed from the interface, washed in HBSS media, and resuspendedin CMRL-1066 medium with penicillin G (100 U/ml), streptomycin (100 pg/mI), and 0.25 pgiml of amphotericin B without serum. The concentration of macrophages was adjusted to lo6 cells per milliliter, Cells were allowed to adhere to plastic dishes for 45 minutes and then washed vigorously with serum-free medium to remove nonadherent cells. Cells were maintained m serum-free medium. Stimulation of macrophages was achieved by treatment with LPS (10 kg/ml) for 4 hours or wifh activated zymosan (1 mgiml) for 4 hours as prevmusl> described.l ’ ’
Preparation
of HB-63
The generation, characterization, and growth of HB-63 has been described.’ In brief, hybridomas were generated by fusing an hypoxanthine guanine phosphoribosyl transferase-deficient mutagenized promonocytic line, U937. with macrophages obtained by maturing human monocytes in Teflon bags. Stimulation by LPS and by activated zymosan was achieved as described above for pulmonary macrophages ’
Growth
of the epithelial
cell line
These cells are from a human endometrial adenocarcinoma cell line (Ishikawa)” ’ ’ and are routinely maintained in MEM containing Earle’s salts, 1 5 % fetal calf serum, 100 U/ml of penicillin, 100 kg/ml of streptomycin, and 0.25 p,g/ml of amphotericin B. Cells were plated in plastic dishes (35 mm) at 5 x lo4 cells per plate and incubated for 48 hours. After which, the medium was discarded, and Ihe cells were incubated for 24 hours in fresh serum-free MEM containing antibiotics, as above. At the time of the experiment, cells had reached confluency.
Bioassay for MLGC secretion Assays were performed as previously debcribed.” ’ Ishikawa cells were cultured in 35 mm plastic dishes containing I .5 ml of MEM for 24 hours with [‘H]glucosamine (1 @i/ml). Labeling medium was discarded. cells were washed once in fresh MEM, and then MEM containing [‘Hlglucosamine was added. Plates were incubated for two 1-hour periods (PD I and PD II). ’ The harvested media from PD I measures the nonmanipulated release of labeled MLGC, whereas PD II is usually the period during which the cells are manipulated. At the start of PD It, fresh MEM plus various preparations of the 68 kd protein were added to the experimental plates. whereas control plates received only media. At the end of the second hour, media were again harvested. The collected media from PDs 1 and 11 were processed for determination of labeled ML&C as follows: 1 ml aliquots were precipitated overnight with an equal volume of 0.1 moliL of potassium acetate in absolute ethanol. The precipitated MLGC was separated hv ccntrifu-
696
Gollub
J. ALLERGY
et al.
gation at 3500 rpm for 15 minutes; the pellets were washed twice with 55% ethanol and dissolved in 250 pl of formic acid. Pour milliliters of scintillation fluid (Dimiscint) was added, and the samples were counted in a scintillation counter. The ratio of the counts per minute of PD II to the counts per minute of PD I was determinedfor both control and experimental cultures to establish an SI for each. Biologic activity kd protein
of fractions
containing
68
Stimulation of secretion by the 68 kd protein was quantified by establishing an SI for the control cultures and comparing it with the SI for the experimentally manipulated cultures. The result was expressed as percent change, namely, X 100 (SI experimental/S1 control - l), as described above. Percent changeswere averagedover experiments and tested for significance with a one-sampleStudent’s t test (t test mean/standarderror). Ap value of CO.05 was consideredstatistically significant. Biologic activity was always compared to the activity of a known secretagogue (positive control), usually carbachol. The stimulatory activity of carbachol at 10 pmol/L (usually 30% to 35% increase over control) is defined as a unit of activity. One unit of activity of the 68 kd protein is that amount that causesa similar increaseof MLGC secretion over control. Thus, data are expressedin both ways, as percent increase above control and as units of activity. Preparation of polyclonal antibody against the 68 kd protein. YM-30 membrane retentates from HB-63 conditioned media were centrifuged in Centripor concentrator tubes containing membraneswith a 50,000 dalton MW cutoff. The concentratedmaterial was further purified by gel electrophoresis and electroelution of the 68 kd bands from 10 gel lanes, as describedin METHODS. After dialysis against distilled water, the material in the bags was concentrated by lyophilization and reelectrophoresedon SDS-polyacrylamide gels. This procedureresulted in the appearanceof a single band at 68 kd on Coomassieblue staining; one lane of the gel containing the same material was electrophoretically transferredto nitrocellulose for immunoblotting with the anti-MMS antibody’and demonstrateda positive reaction for the 68 kd band. Coomassieblue-stainedbandswere sliced from the secondgel and usedas the antigenic material for polyclonal antibody preparation. Antisera from two guinea pigs were combined and tested at dilutions of 1: 50 to 1: 1000 by dot-blot assay with the partially purified 68 kd protein to determine the best dilution for immunoblot assays.Antibodies were also testedagainstBSA and human serum albumen, two proteins with similar MWs. Neither of these proteins demonstratedcross-reactivity with the antibody. The preimmune sera elicited negativereactions in all cases.The antiserawere also testedby Western-blotanalysis (see METHODS below) against the same proteins after SDS-gel electrophoresis (lo%), and the polyclonal antibody reacted only with the 68 kd protein in the partially purified 68 kd fraction and not with either BSA or human serum albumin. Preimmune sera did not react with any of the proteins.
Purification of the 68 kd protein HB83-conditioned media
CLIN. IMMUNOL. MARCH 1992
from
One liter of conditioned media was concentratedby passagethrough a YM-30 ultrafiltration membranein the cold; the flow-through material was frozen at - 2” C for possible use in purification of the low MW protein. The YM-30 retentatematerial (containing proteins with MWs >30 kd) was dialyzed in the cold against2 to 5 L of deionizedwater, lyophilized, resuspendedin 0.05 mol/L of Tris/O.2 mol/L of KCl, pH 7, and then passedthrough an anion-exchange column, Q-Sepharosefast-flow anion-exchangegel. Proteins were eluted by gradient elution with 0.2 to 1 mol/L of KC1/0.05 moliL of Tris buffer, and 0.5 ml fractions were collected. Each fraction was screenedby an immunodot-blot assay, and positive peak fractions (1 to 2 ml) were pooled, dialyzed at 4” C, and concentratedfor additional testing. Protein concentration was determined according to the assay by Lowry et a1.9 Immunodot-blot
assay
A modification of qualitative immunodot-blot assay”was used to assessthe presenceof immunoreactive proteins in various fractions during purification; 2 to 10 yl aliquots were spottedonto nitrocellulose membranes(0.45 pm) and air-dried; the membraneswere then blocked in 5% BSA in PBS or in 0.05% Tw 20 in PBS for 1 hour at room temperature and rinsed briefly with PBS. Incubation initially was with the anti-MMS polyclonal antibody at 1: 200 dilution, and in later assays, with the anti-68 kd polyclonal antibody at 1: 300 in 1% BSAIPBS or 0.05% Tw 20lPBS incubation performed at room temperaturefor 1 hour or overnight in the cold. Membraneswere washedthree times (10 minutes each) with PBS containing 0.05% Tw 20 and incubated for 1 hour either with rabbit antiguinea pig or goat antirabbit-conjugated, horseradish-peroxidasecomplex or with ‘29-labeledprotein A. Membraneswere rinsed with PBS two to three times and then washedfor four lominute periodsas above. 1251-labeled protein A-treatedmembranes were exposed to x-ray film for 8 to 18 hours, and enzyme-treated membranes were developed with the 4chloro- 1-naphtholhydrogenperoxide substrate. Gel electrophoresis and Western-blot analysii Preformed minislab gels, SDS-polyacrylamide gels (10% to 20% gradient), with Laemmli” buffers, were used to analyze the proteins presentin those fractions that demonstrated positive immunoreactivity and positive bioassay activity. Gels were stained with Coomassieblue or were transferred electrophoretically to nitrocellulose to be analyzed by Westernblot. I* Membraneswere blocked and processedas describedfor dot-blotting, except that ‘251-labeled protein A was routinely usedfor processingthe blots. Electroelution of gel slices was performed by placing gel slices in dialysis tubing containing Laemmli” gel-running buffer and subjecting them to 1 A of current in the electrotransfer apparatus,with Laemmli” buffer for approximately 2 to 3 hours. Gel slices were removed, and the eluants were dialyzed against water overnight.
VOLUME NUMBER
89 3
Macrophage
protein
in mucus regulation
699
3
FIG. 1. Dot blots of immunoreactive secreted proteins present in YM-30 retentate fractions from both macrophages and HB-63 cells; lane A, macrophages; lane B, HB63; row 1, 2 ~1; row 2, 4 pl; row 3, 6 ~1; row 4, 10 ~1.
Inhibition
of biologic
activity
by antibody
The polyclonal antibody preparedagainstthe 68 kd protein was usedto block secretagogue activity of the purified protein in the bioassay.Preincubationof the antibody at 1 to 50. 1 to 100, 1 to 200, and 1 to 400 with the active fraction containing 5 kg of protein was carried out for 16 hours at 4” C. A 50% slurry of protein A-SepharoseCl-4B was addedto the tubes that were then mixed gently for 1 hour in the cold beforecentrifugingthe tubesfor 15 minutes. Supematentswere then tested in the bioassayin parallel with samplesfrom control plates(1) containingonly the 68 kd fraction and (2) containing only the antibody. These sampleswere not treatedwith protein-A-Sepharose. Amino
acid analysis
Purified, lyophilized fractions from HB-63 and macrophageswere hydrolyzed in constantboiling HCl (5.7 N) for 24 hours at 107”C. Amino acid residueswere derivatized with 9-fluorenylmethyl chloroformate and then resolved by reverse-phase high-performanceliquid chromatography.” RESULTS immunodot-blot
reaction
To establish whether macrophages and hybridoma cells, induced with zymosan or LPS, secreted mucus secretagoguesother than the oligopeptide previously described,” ’ various fractions obtained during puri-
FIG. 2. Gel electrophoresis and immunoblot analysis of secreted proteins from HB-63 and macrophages. Lane 2 and Lane 3, Coomassie blue-stained proteins of YM-30 retentates from HE-63 and macrophages, respectively; lane 4 and lane 5, immunoblots of proteins from HB-63 and macrophages, respectively; lane 6 and lane 7, immunoblots of anion-exchange fractions from MB-63 and macrophages, respectively, illustrating immunoreactive band at 68,000 daltons, corresponding to major Coomassie blue-stained band (not illustrated here); lane 7, standard proteins of MWs, 95.5, 68, 55, 43, 29, and 18 4 kd.
fication of the MMS peptide were tested for their cross-reactivity with the polyclonal antibodjj. YM-30 retentates were dot blotted onto nitrocellulose membranes and treated with the MMS polyclonal antibody at 1:200 dilution, as described in METHODS. Positive reactions were found with fractions derived from both macrophage (Fig. 1, lane A)- and bybridoma (Fig. 1, lane @-conditioned media. Culture media alone (MEM) were fractionated and tested by immunodot-blot as a control and elicited a negative response (data not presented). These results indicated that one or more secreted proteins of M W >30,00(1 daltons were present in the supematant media that reacted with the MMS antibody. YM-30 retentate fractions, when they were additionally purified by ion-exchange chromatography (see METHODS), yielded mmunoreactive fractions that eluted at 0.X mol! LA of KC110.05 mol /L of Tris. These fractions were pooled and dialyzed extensively against distilled water containing sodium azide at 4” C before lyophilization and additional analyses. SDS-PAGE and Western-btot
analysis
To identify the protein(s) immunoreactive with the anti-MMS antibody, various fractions from both macrophage and hybridoma preparations were subjected to SDS-PAGE on 10% to 20% gradient minigels, as
700
Gollub et al.
J. ALLERGY
TABLE I. Stimulation of MLGC secretion the 68 kd protein from HB-63 Secretory
system
in human
airway
*Purified 66 kd protein (&plate)
explants
and lshikawa
% increase above control +SEM(n=4)
epithelial
CLIN. IMMUNOL. MARCH 1992
cells by
Biologic activityt (units)
Human airway explants 0
-
2.5 5.0
10.0
-
21 * 2 49 -+ 3 63 +- 4
0.60
22.4 + 3 52.5 ? 4 87.5 2 4
0.64 1.50 2.50
1.42 1.80
Epithelial cells 0 2.5 5.0 10.0
Plates contained 1.5 of media. *Purified by ion-exchangechromatography,as describedin METHODS. tBased on the activity of carbachol, which at 10 WmoliL, stimulates secretion between 30% and 35% above control values (no drugs), and is defined as l-unit of activity (see METHODS).
describedin METHODS. YM-30 retentatesfrom hybridoma (Fig. 2, lane 2)- and macrophage-conditioned media (Fig. 2, lane 3) are revealedto contain several similar Coomassieblue-stainedbands with a major band at approximately 68 kd MW. Ion-exchangefractions from both sourceshad only single Coomassieblue-stainedbandsat 68 kd (datanot presented).Gels were also subjectedto Westernblotting, as described in METHODS, initially with the antiMMS antibody,andlater with the polyclonal antibody to the 68 kd protein, to identify which of the bands were immunoreactive. Both antibodies elicited the sameresults. Lanes4 and 5, Fig. 2, are immunoblots of proteins from YM-30 retentatefractions of HB-63 and macrophages,respectively. In Fig. 2, lanes 6 and 7, are immunoblots of the ion-exhangefractions of HB-63 and macrophages,respectively. The results presentedare for the polyclonal 68 kd antibody. The single immunoreactiveband observed at 68 kd was not found when similar blots were treatedwith preimmune rabbit serum (control for the MMS antibody) or preimmune guinea pig serum (control for the 68 kd antibody). Biologic
activity
of immunoreactive
fractions
It has been demonstratedin previous studies that stimulatedhuman airway explants and Ishikawa cells secreteMLGC into the media.3,4 In addition, it was demonstrated,by gel filtration methodsand treatment with proteoglycan-specificenzymes, that the MLGC from Ishikawa cell are devoid of proteoglycans.4To determinewhetherthe immunoreactive68 kd protein observedin the gels describedabove also contained mucus-secretagogue activity, aliquots of the various
fractions were sterilized by filtration through 0.2 pm membranes,and appropriateamountswere addedto Ishikawa cell cultures or human airway explants to assayfor secretionof 3H-labeledMLGC, as described in METHODS. Results found with the ion-exchange purified material from HB-63-conditioned media are presentedin Table I. At very low protein concentrations, this fraction stimulated MLGC secretion in both cell systems in a dose-relatedmanner. Similar results were obtainedwith the fractions derived from macrophage-conditionedmedia (data not presented). Inhibition
of bioactivity
by antibody
To demonstratethat the 68 kd protein was the protein responsiblefor the biologic activity of the fraction, the polyclonal antibody prepared specifically againstthe 68 kd protein was used to block secretagogueactivity in the bioassay,as describedin METHODS. That secretagogueactivity was blocked by the antibody to different degrees,dependingon the antibody dilution, with little inhibition observed at 1: 400 dilution, is presentedis Table II. The antibody itself appearedto have no effect on the bioactivity. Additional confirmation that inhibition of secretagogueactivity was due to the specificity of the antibody was obtained from the following experiment: The preimmune (day 0) guinea pig antibody was tested, as describedin the experiment above, at dilutions of 1: 50, 1: 100, and 1: 200 in parallel with the 68 kd antibody (1: 100) as the positive control. As observedin TableII, pretreatmentof the 68 kd protein with the preimmune antibody did not inhibit its activity; the antibody also had no mucoregulatoryac-
VOLUME NUMBER
99 3
Macrophage
TABLE il. Inhibition
of 68 kd MLGC secretagogue
activity
by antibody
protein
in mucus reguiatmn
701
to the 68 kd protein
--.--~ Material
tested
68 kd protein only* 68 kd polyclonal Ab only Ab plus 68 kd protein
68 kd protein only* Preimmune (day 0) guinea pig serum only Preimmune serum plus 68 kd protein
Bioactivity (% increase over control)
Antibody dilution
l-50 I-50 l-loo l-200 l-400 l-50 l-50 l-100
% Inhibition
._
45 +- 3 0 0 922 22 If- 3 42 t 3 40 t 3 0 39 t 2 41 ‘-c 2
10 i f)l 1 .’! +
{j ---..
0
l__l_
Ab. Antibody.
Resultsare from four separateexperimentspresentedas the mean 2 SEM. Experimentswere performedas described in METHODS. Ail results were statistically significant. *Ion-exchange-purified68 kd protein was used at a concentrationof 5 wg of protein per assaydish (1.5 ml media,.
tivity of its own. The results of the work described aboveindicatethat (1) the newly isolated68 kd protein is immunologicallyrelated to MMS, (2) the 68 kd protein possessesMLGC secretagogue activity, and (3) the activity canbe blockedby its specificantibody. Aminoacid analysis To determinewhetherthe 68 kd proteinfrom macrophagesandhybridomawerein fact identical, a m ino acid analysesby high-performance liquid chromatography were perfotmed on samplesof each, as described in METHODS. The very close identity of theseproteinsare presentedin TableIII. DISClJSSlON Pulmonarymacrophages arethe most prevalentcell type involved in inflammatorylung diseasesthat are associatedwith mucus hypersecretion.Therole of the pulmonarymacrophage-derived productsin mucussecretion was studied.In a previousstudy,‘,’ we demonstratedthat humanpulmonarymacrophages, when they areactivated,releasea newly synthesizedpeptide (MMS) that causesculturedhumanairway cells and secretoryepithelialcells to releaseincreasedamounts of radiolabeledMLGCs. Thesedata, presentedhere, describea secondnovel protein of approximately68 kd, inducible in human lung macrophages,which stimulatesMLGC secretionfrom both humanairway cells in vitro and secretoryepithelial cells in culture. This proteinis alsopresentin the humanmacrophagederived HB-63, when it is stimulatedby LPS and activatedzymosan. Since the original anti-MMS polyclonal antibody detectedthe 68 kd protein and the new polyclonal
TABLE M l.Comparison of amino acid analyses of the 68 kd secretagogue protein derived from macrophages and HE-63 -ReeldllasllOBB resktues Mecropluge
Aspartic
105
Serine Glutamic Tbreonine Arginine Glycine Alanine Proline Valine Phenylalanine Isoleucine Leucine Lysine Methionine Tyrosine
58 162 60 55 66 108 58
65 48
26 105 a4 ND
ND, Not done.
antibody preparedagainst68 kd protein also crossreactsstronglywith MMS, it is possible that MMS is derivedfrom the68 kd protein,andtherefore,epitopes commonto both the peptideandthe largerproteinare recognizedby both antibodies.Although MMS may be a proteolytic breakdownproductof processing,it is also possible,althoughthis is highly unlikely, that the 68 kd proteinis an aggregateform of the Iow M W peptide. The relationshipbetweenthe two proteins remainsto be defined.(A monoclonalantibodyagainst
702
Gollub
et al.
the 68 kd protein has recently been characterized’4; perhapsit will be useful in elucidatingthis issue.) In preliminary studies we were able to detect, immunologically, the 68 kd protein in the bronchoalveolar fluid from a patient with chronic bronchitis and bronchorrhea(unpublishedresults). The protein was purified and found to be biologically active. The role of this novel 68 kd protein in the lungs is speculative.It may be part of the macrophage-defense arsenalusedfor ridding airwaysof pathogenicbacteria and their products. Alternatively, it may be a component of lung inflammation, in which its activity m ight contributeto the hypersecretionof mucousglycoprotein associatedwith acute and chronic pulmonary diseases.Furthermore, since macrophage-derived productshave generallybeenfound to have diverseactivities,l5for example,interleukin-I andtumor necrosisfactor, it is possible that the 68 kd protein may also have diverse activities that remain to be discovered.The mucus secretagogue function of this protein establishesit as a very useful tool for studies of the regulatory mechanismsinvolved in mucus secretion. We thank Dr. Elaine Schwartz (Department of Dermatology) for the amino acid analyses, and Ms. Rozlyn Richardson for her patience and excellent work in the preparation of the manuscript.
1. Marom Z, ShelhamerJH, Kaliner M. Human pulmonary macrophage-derived mucus secretagogue.J Exp Med 1984; 159:844-60. 2. Marom Z, ShelhamerJH, Kaliner M. Human monocyte-derived mucus secretagogue.J Clin Invest 1985;75:191-8. 3. Goswami SK, Gollub EG, Holinka C, Gurpide E, Marom Z. Characterizationof a mucin-like glycoprotein from an epithelial cell line [Abstract]. Am Rev Respir Dis 1989;141:465.
J. ALLERGY
CLIN. IMMUNOL. M A R C H 1992
4. Goswami S, Kivity S, Marom Z. Erythromycin inhibits respiratory glycoconjugatesecretionfrom human airways in vitro. Am Rev Respir Dis 1990;141:72-8. 5. Sperber K, Goswami S, Gollub EG, Mayer L, Marom Z. Mucus secretagogueproduction by a human macrophagehybridoma. J ALLERGYCLINIMMLJNOL~~~~;~~:~~O-8. 6. SperberK, Batter J, Pizzimenti A, Najfeld J, Mayer L. Identification of subpopulationsof humanmacrophagesthroughthe generation of human macrophagehybridomas. J Immunol Methods 1990;129:31-40. 7. Nishida MK, KasaharaK, Kaneko M, Iwasaki H. Establishment of a new human endometrialadenocarcinomacell line, Ishikawa cells, containingestrogenand progesteronereceptors. Acta Obstet Gynaecol Jpn 1985;37:1103-l 1. 8. Amin DN, Goswami S, Klein T, Maayani S, Marom Z. Functional antagonismbetweenhormone receptor systems:modulation of glycopmtein secretion in secretory epithelial cells. Am J Respir Cell Mol Biol 1991;4:135-9. 9. Lowry OH, RosebroughAJ, Farr AL, Randall RJ. Protein measurementswith folin-phenol reagents. J Biol Chem 1951;193:265-95. 10. Hawkes R, Niday E, Gordon J. A dot-immunobindingassay for monoclonal and other antibodies. Analyt Biochem 1982;119:142-7. 11. Laemmli UK. Cleavageof structural proteins during the assembly of the head of bacteriophageT4. Nature 1970;217: 680-5. 12. Towbin H, StaehelinT, Gordon J. Electrophoretictransfer of proteinsfrom polyacrylamidegels to nitmcellulose sheets:procedure and applications. Proc Nat1 Acad Sci USA 1979;76:4350-4. 13. Schwartz E. Connectivetissue alterationsin the skin of ultraviolet-irradiatedhairlessmice. J Invest Dermatol 1988;91:15861. 14. Sperber K, Gollub E, Goswami SK, Mayer L, Marom Z. Detectionof a novel 68,000 d mucus secretagogue by a monoclonal antibody [Abstract]. J ALLERGYCLIN IMMUNOL 1991;87:287. 15. Nathan CF. Secretoryproductsof macrophages.J Clin Invest 1987;79:319-26.
PhD,* Satindra K. Goswami, MD* New York, N.Y.
PhD,* Kirk Sperber,
MD,**
Pulmonary macrophages release a variety of mediators that are involved in injlammatory processes and probably are involved in respiratory mucus secretion. Conditioned media obtained from activated pulmonary macrophages were found to contain a protein that functioned as a secretagogue for mucus-like glycoconjugate (MLGC) in an in vitro bioassay. A human macrophage-derived hybridoma cell line, HB-63, exhibited the same properties and was very useful in obtaining large amounts of the protein for purtjication and characterization. With ultrafiltration membranes and gel electrophoresis, the protein isolated from the conditioned media of zymosan- or lipopolysaccharide-treated cells was found to have a molecular weight of approximately 68 kd. The purified protein obtained from hybridoma cells and from pulmonary macrophages exhibited strong biologic activity when it was used to stimulate MLGC secretion, both in human airway explants and in an in vitro human secretory epithelial cancer cell line. The proteins from both sources were found to have similar amino acid compositions. Preliminary results indicate the presence of the 68 kd protein in the bronchoalveolar lavage fluid of a patient with severe chronic bronchitis and mucus hypersecretion. The role of this novel protein in the lungs is, so far speculative. The 68 kd protein may be a useful tool for studying the biosynthesis and. regulation of MLGC secretion and hypersecretion. (J ALLERGY CLIN IMMUNOL 1992;89:696-702 .) Key words: Pulmonary macrophages, hybridoma, mucous glycoprotein secretagogue, 68 kd protein
Pulmonary macrophages play a variety of roles in the respiratory system. They constitute most phagocytic cells in the lower respiratory tract and are important components of inflammatory and immunologic processes in the lung. Since they secrete numerous products, macrophages can interact with, and affect, other cells. For example, through their release of var-
From the Divisions of *Pulmonary and Critical Care Medicine and **Clinical Immunology, Mt. Sinai Medical Center, New York, N.Y. Received July 2, 1991. Revised Oct. 18, 1991. Accepted Oct. 23, 1991. Reprint requests: Edith G. Gollub, PhD, Division of Pulmonary and Critical Care Medicine, Box 1232, Mt. Sinai Medical Center, One Gustave Levy Place, New York, NY 10029. Dr. Edith G. Gollub is a recipient of the New York Lung Association ResearchAward. Dr. Zvi Marom was supportedby National Institutes of Health Grant NHL-37254. 111134525
696
Abbreviations used
HBSS: MEM: SI: MMS: LPS: MLGC: HB-63: PBS: Tw 20: MW: BSA: PD I, PD II: SDS-PAGE:
Hanks’balanced salt solution Minimal essentialmedia Secretory index Macrophagelmonocyte-derived mucus secretagogue Lipopolysaccharide
Mucus-like glycoconjugate Hybridoma cell line No. 63 Phosphate-bufferedsaline Tween 20 Molecular weight Bovine serum albumin Periods I and II Sodium dodecyl sulfate-polyacrylamide gel electrophoresis
ious cytokines, including inhibitory as well as growth factors, lung macrophages take part in both cell damage and tissue repair.
VOLUME NUMBER
Macrophage protein in mucus Feguiation
89 3
Among the many secretoryproducts synthesized by macrophages,a novel peptide, MMS, has been described.‘2. This peptide was found to stimulate MLGC secretion in vitro, both in human airway explants”’and,morerecently,in a secretoryepithelial cell line.‘. 4 Studiesindicatedthat MMS is an acidic proteinwhoseM W is estimatedat approximately2000 daltonsby gel-filtration procedures.’ More recently, a polyclonalantibodywas raisedin rabbitsagainstthe MMS peptideafter its conjugationto crab hemocyanin.’This antibodyhas proved useful in our additional studiesof MMS and in the indentificationof other possibly relatedproteins. In this study, we describethe isolationandcharacterization of a new protein. which cross-reacts with theMMS antibody,functions as an MLGC secretagogue in vitro, and is elaboratednot only by pulmonarymacrophages but also by a macrophage-derived HB-63.5,’ This new macrophage-derived secretagogue may be an important m e d iatorof mucussecretionin vivo and may help to clarify the role of pulmonarymacrophages in inflammatory lung diseasesassociatedwith mucushypersecretion. Use of the hybridomacell line hasfacilitated the isolation and purification of the new protein, and most of the experimentsdescribedin this study were performedwith the hybridoma-derived protein. MATERIAL
AND METHODS
MEM, RPM1 1640 media, and HBSS were from Grand Island Biologicals, Grand Island, NY. Ficoll-Hypaque, protein A-Sepharose Cl-4B, and Q-Sepharose fast-flow anionexchange gel were from Pharmacia/LKB Chemicals, Piscataway, N.J. Zymosan particles, LPSs, BSA, and Tw 20 were from Sigma Chemical Co., St. Louis, MO. Dimiscintscintillation fluid was from National Diagnostics, Manville, N.J. Precast polyacrylamide minigels were from Integrated Separation Systems, Hyde Park, Mass. Nitrocellulose membranes were from Schleicher & Schuell (Keene, N.H.). Ultrafiltration membranes were from Amicon Corp., Danvers, Mass. Centripor concentrator tubes were from Spectrum Medical Industries, Inc., Los Angeles, Calif. Goat antirabbit and rabbit antiguinea pig horseradish-peroxidase conjugates and 4-chloro-1-naphthol were from Bio-Rad Laboratories. Richmond, Calif. ‘*51-labeledprotein A and [ ‘Hlglucosamine were from Amersham Corp., Arlington Heights. Ill. X-ray film was from Eastman Kodak Co.,
Rochester, N.Y. Polyclonalantibodies againstthe 68 kd protein were generated from guinea pigs by Pocono Rabbit Farms, Canadensis, Pa., with gel-purified 68 kd protein, as described later in METHODS.
Isolation, culture, and stimulation pulmonary macrophages
of human
lsolation of pulmonary macrophages has previously been described.’ In brief, the macrophages were isolated from surgical specimans of human lungs, maintained at 22” C in
697
HBSSwithoutCa-- or M g ”. Fragments of parenchyma wereteasedapart,filteredthroughsterilewire mesh,and centrifuged; cell pellets were pooled, layered onto FicollHypaque, and centrifuged. The mononuclear cell preparations were removed from the interface, washed in HBSS media, and resuspendedin CMRL-1066 medium with penicillin G (100 U/ml), streptomycin (100 pg/mI), and 0.25 pgiml of amphotericin B without serum. The concentration of macrophages was adjusted to lo6 cells per milliliter, Cells were allowed to adhere to plastic dishes for 45 minutes and then washed vigorously with serum-free medium to remove nonadherent cells. Cells were maintained m serum-free medium. Stimulation of macrophages was achieved by treatment with LPS (10 kg/ml) for 4 hours or wifh activated zymosan (1 mgiml) for 4 hours as prevmusl> described.l ’ ’
Preparation
of HB-63
The generation, characterization, and growth of HB-63 has been described.’ In brief, hybridomas were generated by fusing an hypoxanthine guanine phosphoribosyl transferase-deficient mutagenized promonocytic line, U937. with macrophages obtained by maturing human monocytes in Teflon bags. Stimulation by LPS and by activated zymosan was achieved as described above for pulmonary macrophages ’
Growth
of the epithelial
cell line
These cells are from a human endometrial adenocarcinoma cell line (Ishikawa)” ’ ’ and are routinely maintained in MEM containing Earle’s salts, 1 5 % fetal calf serum, 100 U/ml of penicillin, 100 kg/ml of streptomycin, and 0.25 p,g/ml of amphotericin B. Cells were plated in plastic dishes (35 mm) at 5 x lo4 cells per plate and incubated for 48 hours. After which, the medium was discarded, and Ihe cells were incubated for 24 hours in fresh serum-free MEM containing antibiotics, as above. At the time of the experiment, cells had reached confluency.
Bioassay for MLGC secretion Assays were performed as previously debcribed.” ’ Ishikawa cells were cultured in 35 mm plastic dishes containing I .5 ml of MEM for 24 hours with [‘H]glucosamine (1 @i/ml). Labeling medium was discarded. cells were washed once in fresh MEM, and then MEM containing [‘Hlglucosamine was added. Plates were incubated for two 1-hour periods (PD I and PD II). ’ The harvested media from PD I measures the nonmanipulated release of labeled MLGC, whereas PD II is usually the period during which the cells are manipulated. At the start of PD It, fresh MEM plus various preparations of the 68 kd protein were added to the experimental plates. whereas control plates received only media. At the end of the second hour, media were again harvested. The collected media from PDs 1 and 11 were processed for determination of labeled ML&C as follows: 1 ml aliquots were precipitated overnight with an equal volume of 0.1 moliL of potassium acetate in absolute ethanol. The precipitated MLGC was separated hv ccntrifu-
696
Gollub
J. ALLERGY
et al.
gation at 3500 rpm for 15 minutes; the pellets were washed twice with 55% ethanol and dissolved in 250 pl of formic acid. Pour milliliters of scintillation fluid (Dimiscint) was added, and the samples were counted in a scintillation counter. The ratio of the counts per minute of PD II to the counts per minute of PD I was determinedfor both control and experimental cultures to establish an SI for each. Biologic activity kd protein
of fractions
containing
68
Stimulation of secretion by the 68 kd protein was quantified by establishing an SI for the control cultures and comparing it with the SI for the experimentally manipulated cultures. The result was expressed as percent change, namely, X 100 (SI experimental/S1 control - l), as described above. Percent changeswere averagedover experiments and tested for significance with a one-sampleStudent’s t test (t test mean/standarderror). Ap value of CO.05 was consideredstatistically significant. Biologic activity was always compared to the activity of a known secretagogue (positive control), usually carbachol. The stimulatory activity of carbachol at 10 pmol/L (usually 30% to 35% increase over control) is defined as a unit of activity. One unit of activity of the 68 kd protein is that amount that causesa similar increaseof MLGC secretion over control. Thus, data are expressedin both ways, as percent increase above control and as units of activity. Preparation of polyclonal antibody against the 68 kd protein. YM-30 membrane retentates from HB-63 conditioned media were centrifuged in Centripor concentrator tubes containing membraneswith a 50,000 dalton MW cutoff. The concentratedmaterial was further purified by gel electrophoresis and electroelution of the 68 kd bands from 10 gel lanes, as describedin METHODS. After dialysis against distilled water, the material in the bags was concentrated by lyophilization and reelectrophoresedon SDS-polyacrylamide gels. This procedureresulted in the appearanceof a single band at 68 kd on Coomassieblue staining; one lane of the gel containing the same material was electrophoretically transferredto nitrocellulose for immunoblotting with the anti-MMS antibody’and demonstrateda positive reaction for the 68 kd band. Coomassieblue-stainedbandswere sliced from the secondgel and usedas the antigenic material for polyclonal antibody preparation. Antisera from two guinea pigs were combined and tested at dilutions of 1: 50 to 1: 1000 by dot-blot assay with the partially purified 68 kd protein to determine the best dilution for immunoblot assays.Antibodies were also testedagainstBSA and human serum albumen, two proteins with similar MWs. Neither of these proteins demonstratedcross-reactivity with the antibody. The preimmune sera elicited negativereactions in all cases.The antiserawere also testedby Western-blotanalysis (see METHODS below) against the same proteins after SDS-gel electrophoresis (lo%), and the polyclonal antibody reacted only with the 68 kd protein in the partially purified 68 kd fraction and not with either BSA or human serum albumin. Preimmune sera did not react with any of the proteins.
Purification of the 68 kd protein HB83-conditioned media
CLIN. IMMUNOL. MARCH 1992
from
One liter of conditioned media was concentratedby passagethrough a YM-30 ultrafiltration membranein the cold; the flow-through material was frozen at - 2” C for possible use in purification of the low MW protein. The YM-30 retentatematerial (containing proteins with MWs >30 kd) was dialyzed in the cold against2 to 5 L of deionizedwater, lyophilized, resuspendedin 0.05 mol/L of Tris/O.2 mol/L of KCl, pH 7, and then passedthrough an anion-exchange column, Q-Sepharosefast-flow anion-exchangegel. Proteins were eluted by gradient elution with 0.2 to 1 mol/L of KC1/0.05 moliL of Tris buffer, and 0.5 ml fractions were collected. Each fraction was screenedby an immunodot-blot assay, and positive peak fractions (1 to 2 ml) were pooled, dialyzed at 4” C, and concentratedfor additional testing. Protein concentration was determined according to the assay by Lowry et a1.9 Immunodot-blot
assay
A modification of qualitative immunodot-blot assay”was used to assessthe presenceof immunoreactive proteins in various fractions during purification; 2 to 10 yl aliquots were spottedonto nitrocellulose membranes(0.45 pm) and air-dried; the membraneswere then blocked in 5% BSA in PBS or in 0.05% Tw 20 in PBS for 1 hour at room temperature and rinsed briefly with PBS. Incubation initially was with the anti-MMS polyclonal antibody at 1: 200 dilution, and in later assays, with the anti-68 kd polyclonal antibody at 1: 300 in 1% BSAIPBS or 0.05% Tw 20lPBS incubation performed at room temperaturefor 1 hour or overnight in the cold. Membraneswere washedthree times (10 minutes each) with PBS containing 0.05% Tw 20 and incubated for 1 hour either with rabbit antiguinea pig or goat antirabbit-conjugated, horseradish-peroxidasecomplex or with ‘29-labeledprotein A. Membraneswere rinsed with PBS two to three times and then washedfor four lominute periodsas above. 1251-labeled protein A-treatedmembranes were exposed to x-ray film for 8 to 18 hours, and enzyme-treated membranes were developed with the 4chloro- 1-naphtholhydrogenperoxide substrate. Gel electrophoresis and Western-blot analysii Preformed minislab gels, SDS-polyacrylamide gels (10% to 20% gradient), with Laemmli” buffers, were used to analyze the proteins presentin those fractions that demonstrated positive immunoreactivity and positive bioassay activity. Gels were stained with Coomassieblue or were transferred electrophoretically to nitrocellulose to be analyzed by Westernblot. I* Membraneswere blocked and processedas describedfor dot-blotting, except that ‘251-labeled protein A was routinely usedfor processingthe blots. Electroelution of gel slices was performed by placing gel slices in dialysis tubing containing Laemmli” gel-running buffer and subjecting them to 1 A of current in the electrotransfer apparatus,with Laemmli” buffer for approximately 2 to 3 hours. Gel slices were removed, and the eluants were dialyzed against water overnight.
VOLUME NUMBER
89 3
Macrophage
protein
in mucus regulation
699
3
FIG. 1. Dot blots of immunoreactive secreted proteins present in YM-30 retentate fractions from both macrophages and HB-63 cells; lane A, macrophages; lane B, HB63; row 1, 2 ~1; row 2, 4 pl; row 3, 6 ~1; row 4, 10 ~1.
Inhibition
of biologic
activity
by antibody
The polyclonal antibody preparedagainstthe 68 kd protein was usedto block secretagogue activity of the purified protein in the bioassay.Preincubationof the antibody at 1 to 50. 1 to 100, 1 to 200, and 1 to 400 with the active fraction containing 5 kg of protein was carried out for 16 hours at 4” C. A 50% slurry of protein A-SepharoseCl-4B was addedto the tubes that were then mixed gently for 1 hour in the cold beforecentrifugingthe tubesfor 15 minutes. Supematentswere then tested in the bioassayin parallel with samplesfrom control plates(1) containingonly the 68 kd fraction and (2) containing only the antibody. These sampleswere not treatedwith protein-A-Sepharose. Amino
acid analysis
Purified, lyophilized fractions from HB-63 and macrophageswere hydrolyzed in constantboiling HCl (5.7 N) for 24 hours at 107”C. Amino acid residueswere derivatized with 9-fluorenylmethyl chloroformate and then resolved by reverse-phase high-performanceliquid chromatography.” RESULTS immunodot-blot
reaction
To establish whether macrophages and hybridoma cells, induced with zymosan or LPS, secreted mucus secretagoguesother than the oligopeptide previously described,” ’ various fractions obtained during puri-
FIG. 2. Gel electrophoresis and immunoblot analysis of secreted proteins from HB-63 and macrophages. Lane 2 and Lane 3, Coomassie blue-stained proteins of YM-30 retentates from HE-63 and macrophages, respectively; lane 4 and lane 5, immunoblots of proteins from HB-63 and macrophages, respectively; lane 6 and lane 7, immunoblots of anion-exchange fractions from MB-63 and macrophages, respectively, illustrating immunoreactive band at 68,000 daltons, corresponding to major Coomassie blue-stained band (not illustrated here); lane 7, standard proteins of MWs, 95.5, 68, 55, 43, 29, and 18 4 kd.
fication of the MMS peptide were tested for their cross-reactivity with the polyclonal antibodjj. YM-30 retentates were dot blotted onto nitrocellulose membranes and treated with the MMS polyclonal antibody at 1:200 dilution, as described in METHODS. Positive reactions were found with fractions derived from both macrophage (Fig. 1, lane A)- and bybridoma (Fig. 1, lane @-conditioned media. Culture media alone (MEM) were fractionated and tested by immunodot-blot as a control and elicited a negative response (data not presented). These results indicated that one or more secreted proteins of M W >30,00(1 daltons were present in the supematant media that reacted with the MMS antibody. YM-30 retentate fractions, when they were additionally purified by ion-exchange chromatography (see METHODS), yielded mmunoreactive fractions that eluted at 0.X mol! LA of KC110.05 mol /L of Tris. These fractions were pooled and dialyzed extensively against distilled water containing sodium azide at 4” C before lyophilization and additional analyses. SDS-PAGE and Western-btot
analysis
To identify the protein(s) immunoreactive with the anti-MMS antibody, various fractions from both macrophage and hybridoma preparations were subjected to SDS-PAGE on 10% to 20% gradient minigels, as
700
Gollub et al.
J. ALLERGY
TABLE I. Stimulation of MLGC secretion the 68 kd protein from HB-63 Secretory
system
in human
airway
*Purified 66 kd protein (&plate)
explants
and lshikawa
% increase above control +SEM(n=4)
epithelial
CLIN. IMMUNOL. MARCH 1992
cells by
Biologic activityt (units)
Human airway explants 0
-
2.5 5.0
10.0
-
21 * 2 49 -+ 3 63 +- 4
0.60
22.4 + 3 52.5 ? 4 87.5 2 4
0.64 1.50 2.50
1.42 1.80
Epithelial cells 0 2.5 5.0 10.0
Plates contained 1.5 of media. *Purified by ion-exchangechromatography,as describedin METHODS. tBased on the activity of carbachol, which at 10 WmoliL, stimulates secretion between 30% and 35% above control values (no drugs), and is defined as l-unit of activity (see METHODS).
describedin METHODS. YM-30 retentatesfrom hybridoma (Fig. 2, lane 2)- and macrophage-conditioned media (Fig. 2, lane 3) are revealedto contain several similar Coomassieblue-stainedbands with a major band at approximately 68 kd MW. Ion-exchangefractions from both sourceshad only single Coomassieblue-stainedbandsat 68 kd (datanot presented).Gels were also subjectedto Westernblotting, as described in METHODS, initially with the antiMMS antibody,andlater with the polyclonal antibody to the 68 kd protein, to identify which of the bands were immunoreactive. Both antibodies elicited the sameresults. Lanes4 and 5, Fig. 2, are immunoblots of proteins from YM-30 retentatefractions of HB-63 and macrophages,respectively. In Fig. 2, lanes 6 and 7, are immunoblots of the ion-exhangefractions of HB-63 and macrophages,respectively. The results presentedare for the polyclonal 68 kd antibody. The single immunoreactiveband observed at 68 kd was not found when similar blots were treatedwith preimmune rabbit serum (control for the MMS antibody) or preimmune guinea pig serum (control for the 68 kd antibody). Biologic
activity
of immunoreactive
fractions
It has been demonstratedin previous studies that stimulatedhuman airway explants and Ishikawa cells secreteMLGC into the media.3,4 In addition, it was demonstrated,by gel filtration methodsand treatment with proteoglycan-specificenzymes, that the MLGC from Ishikawa cell are devoid of proteoglycans.4To determinewhetherthe immunoreactive68 kd protein observedin the gels describedabove also contained mucus-secretagogue activity, aliquots of the various
fractions were sterilized by filtration through 0.2 pm membranes,and appropriateamountswere addedto Ishikawa cell cultures or human airway explants to assayfor secretionof 3H-labeledMLGC, as described in METHODS. Results found with the ion-exchange purified material from HB-63-conditioned media are presentedin Table I. At very low protein concentrations, this fraction stimulated MLGC secretion in both cell systems in a dose-relatedmanner. Similar results were obtainedwith the fractions derived from macrophage-conditionedmedia (data not presented). Inhibition
of bioactivity
by antibody
To demonstratethat the 68 kd protein was the protein responsiblefor the biologic activity of the fraction, the polyclonal antibody prepared specifically againstthe 68 kd protein was used to block secretagogueactivity in the bioassay,as describedin METHODS. That secretagogueactivity was blocked by the antibody to different degrees,dependingon the antibody dilution, with little inhibition observed at 1: 400 dilution, is presentedis Table II. The antibody itself appearedto have no effect on the bioactivity. Additional confirmation that inhibition of secretagogueactivity was due to the specificity of the antibody was obtained from the following experiment: The preimmune (day 0) guinea pig antibody was tested, as describedin the experiment above, at dilutions of 1: 50, 1: 100, and 1: 200 in parallel with the 68 kd antibody (1: 100) as the positive control. As observedin TableII, pretreatmentof the 68 kd protein with the preimmune antibody did not inhibit its activity; the antibody also had no mucoregulatoryac-
VOLUME NUMBER
99 3
Macrophage
TABLE il. Inhibition
of 68 kd MLGC secretagogue
activity
by antibody
protein
in mucus reguiatmn
701
to the 68 kd protein
--.--~ Material
tested
68 kd protein only* 68 kd polyclonal Ab only Ab plus 68 kd protein
68 kd protein only* Preimmune (day 0) guinea pig serum only Preimmune serum plus 68 kd protein
Bioactivity (% increase over control)
Antibody dilution
l-50 I-50 l-loo l-200 l-400 l-50 l-50 l-100
% Inhibition
._
45 +- 3 0 0 922 22 If- 3 42 t 3 40 t 3 0 39 t 2 41 ‘-c 2
10 i f)l 1 .’! +
{j ---..
0
l__l_
Ab. Antibody.
Resultsare from four separateexperimentspresentedas the mean 2 SEM. Experimentswere performedas described in METHODS. Ail results were statistically significant. *Ion-exchange-purified68 kd protein was used at a concentrationof 5 wg of protein per assaydish (1.5 ml media,.
tivity of its own. The results of the work described aboveindicatethat (1) the newly isolated68 kd protein is immunologicallyrelated to MMS, (2) the 68 kd protein possessesMLGC secretagogue activity, and (3) the activity canbe blockedby its specificantibody. Aminoacid analysis To determinewhetherthe 68 kd proteinfrom macrophagesandhybridomawerein fact identical, a m ino acid analysesby high-performance liquid chromatography were perfotmed on samplesof each, as described in METHODS. The very close identity of theseproteinsare presentedin TableIII. DISClJSSlON Pulmonarymacrophages arethe most prevalentcell type involved in inflammatorylung diseasesthat are associatedwith mucus hypersecretion.Therole of the pulmonarymacrophage-derived productsin mucussecretion was studied.In a previousstudy,‘,’ we demonstratedthat humanpulmonarymacrophages, when they areactivated,releasea newly synthesizedpeptide (MMS) that causesculturedhumanairway cells and secretoryepithelialcells to releaseincreasedamounts of radiolabeledMLGCs. Thesedata, presentedhere, describea secondnovel protein of approximately68 kd, inducible in human lung macrophages,which stimulatesMLGC secretionfrom both humanairway cells in vitro and secretoryepithelial cells in culture. This proteinis alsopresentin the humanmacrophagederived HB-63, when it is stimulatedby LPS and activatedzymosan. Since the original anti-MMS polyclonal antibody detectedthe 68 kd protein and the new polyclonal
TABLE M l.Comparison of amino acid analyses of the 68 kd secretagogue protein derived from macrophages and HE-63 -ReeldllasllOBB resktues Mecropluge
Aspartic
105
Serine Glutamic Tbreonine Arginine Glycine Alanine Proline Valine Phenylalanine Isoleucine Leucine Lysine Methionine Tyrosine
58 162 60 55 66 108 58
65 48
26 105 a4 ND
ND, Not done.
antibody preparedagainst68 kd protein also crossreactsstronglywith MMS, it is possible that MMS is derivedfrom the68 kd protein,andtherefore,epitopes commonto both the peptideandthe largerproteinare recognizedby both antibodies.Although MMS may be a proteolytic breakdownproductof processing,it is also possible,althoughthis is highly unlikely, that the 68 kd proteinis an aggregateform of the Iow M W peptide. The relationshipbetweenthe two proteins remainsto be defined.(A monoclonalantibodyagainst
702
Gollub
et al.
the 68 kd protein has recently been characterized’4; perhapsit will be useful in elucidatingthis issue.) In preliminary studies we were able to detect, immunologically, the 68 kd protein in the bronchoalveolar fluid from a patient with chronic bronchitis and bronchorrhea(unpublishedresults). The protein was purified and found to be biologically active. The role of this novel 68 kd protein in the lungs is speculative.It may be part of the macrophage-defense arsenalusedfor ridding airwaysof pathogenicbacteria and their products. Alternatively, it may be a component of lung inflammation, in which its activity m ight contributeto the hypersecretionof mucousglycoprotein associatedwith acute and chronic pulmonary diseases.Furthermore, since macrophage-derived productshave generallybeenfound to have diverseactivities,l5for example,interleukin-I andtumor necrosisfactor, it is possible that the 68 kd protein may also have diverse activities that remain to be discovered.The mucus secretagogue function of this protein establishesit as a very useful tool for studies of the regulatory mechanismsinvolved in mucus secretion. We thank Dr. Elaine Schwartz (Department of Dermatology) for the amino acid analyses, and Ms. Rozlyn Richardson for her patience and excellent work in the preparation of the manuscript.
1. Marom Z, ShelhamerJH, Kaliner M. Human pulmonary macrophage-derived mucus secretagogue.J Exp Med 1984; 159:844-60. 2. Marom Z, ShelhamerJH, Kaliner M. Human monocyte-derived mucus secretagogue.J Clin Invest 1985;75:191-8. 3. Goswami SK, Gollub EG, Holinka C, Gurpide E, Marom Z. Characterizationof a mucin-like glycoprotein from an epithelial cell line [Abstract]. Am Rev Respir Dis 1989;141:465.
J. ALLERGY
CLIN. IMMUNOL. M A R C H 1992
4. Goswami S, Kivity S, Marom Z. Erythromycin inhibits respiratory glycoconjugatesecretionfrom human airways in vitro. Am Rev Respir Dis 1990;141:72-8. 5. Sperber K, Goswami S, Gollub EG, Mayer L, Marom Z. Mucus secretagogueproduction by a human macrophagehybridoma. J ALLERGYCLINIMMLJNOL~~~~;~~:~~O-8. 6. SperberK, Batter J, Pizzimenti A, Najfeld J, Mayer L. Identification of subpopulationsof humanmacrophagesthroughthe generation of human macrophagehybridomas. J Immunol Methods 1990;129:31-40. 7. Nishida MK, KasaharaK, Kaneko M, Iwasaki H. Establishment of a new human endometrialadenocarcinomacell line, Ishikawa cells, containingestrogenand progesteronereceptors. Acta Obstet Gynaecol Jpn 1985;37:1103-l 1. 8. Amin DN, Goswami S, Klein T, Maayani S, Marom Z. Functional antagonismbetweenhormone receptor systems:modulation of glycopmtein secretion in secretory epithelial cells. Am J Respir Cell Mol Biol 1991;4:135-9. 9. Lowry OH, RosebroughAJ, Farr AL, Randall RJ. Protein measurementswith folin-phenol reagents. J Biol Chem 1951;193:265-95. 10. Hawkes R, Niday E, Gordon J. A dot-immunobindingassay for monoclonal and other antibodies. Analyt Biochem 1982;119:142-7. 11. Laemmli UK. Cleavageof structural proteins during the assembly of the head of bacteriophageT4. Nature 1970;217: 680-5. 12. Towbin H, StaehelinT, Gordon J. Electrophoretictransfer of proteinsfrom polyacrylamidegels to nitmcellulose sheets:procedure and applications. Proc Nat1 Acad Sci USA 1979;76:4350-4. 13. Schwartz E. Connectivetissue alterationsin the skin of ultraviolet-irradiatedhairlessmice. J Invest Dermatol 1988;91:15861. 14. Sperber K, Gollub E, Goswami SK, Mayer L, Marom Z. Detectionof a novel 68,000 d mucus secretagogue by a monoclonal antibody [Abstract]. J ALLERGYCLIN IMMUNOL 1991;87:287. 15. Nathan CF. Secretoryproductsof macrophages.J Clin Invest 1987;79:319-26.
