Proc. Nati. Acad. Sci. USAVol. 81, pp. 4505-4509, July 1984Immunology

Identification and characterization of a monoclonal antibody to anantigen expressed on activated macrophages

(lymphokine/hybridoma/cell surface antigens/immunofluorescence/macrophage-mediated cytotoxicity)

THOMAS P. KOESTLER*, DAVID RIEMAN*, KATHY MUIRHEADt, RUSSELL G. GREIG*, AND GEORGE POSTE*t

Departments of *Tumor Biology and tImmunology, Smith Kline & French Laboratories, 1500 Spring Garden Street, Philadelphia, PA 19101; andtDepartment of Pathology and Laboratory Medicine, University of Pennsylvania, Philadelphia, PA 19104

Communicated by Zanvil A. Cohn, March 19, 1984

ABSTRACT A hybridoma clone secreting a monoclonalantibody, designated MA158.2, that reacts with an antigen expressed on lymphokine-treated macrophages was produced byfusion of mouse myeloma cells with rat spleen cells immunizedagainst C57BL/6 peritoneal macrophages rendered tumorici-dal in vitro by incubation with the lymphokine macrophage-activating factor. The specificity of the antibody for activatedmacrophages and lack of reactivity with histologically diversecell types was determined by radioimmune indirect bindingand flow cytometry. MA158.2 antibody binds to mouse perito-neal macrophages elicited by nonspecific inflammatory agentsand to tumoricidal macrophages elicited with Corynebacteriumparvum. Resident peritoneal, splenic, and alveolar macro-phages were only weakly positive. Several macrophage celllines (P388D1, WEH1-231, J774, RAW 264.7), murine fibro-blasts, and neutrophils did not bind detectable amounts ofMA158.2. Radioimmune indirect binding analysis demon-strated that cell suspensions prepared from C57BL/6 mousespleen, thymus, and lymph node as well as polymorphonuclearleukocytes, lymphocytes, and T- and B-cell murine lympho-mas were MA158.2 negative. Expression of the reactive anti-gen on the macrophage cell surface was enhanced 3-fold fol-lowing in vitro activation of elicited macrophages with macro-phage-activating factor and the kinetics of activation to thetumoricidal state paralleled the increased expression of theantigen recognized by MA158.2. MA158.2 is a rat IgG2a anti-body containing a single specific heavy and light chain thatdoes not detect a polymorphic determinant. This monoclonalantibody will be a useful tool for monitoring the efficacy ofagents in activating murine macrophages to the tumoricidalstate and in analyzing the sequence of biochemical events thatculminate in macrophage activation.

Macrophages (MO) are an important effector cell in the rec-ognition and destruction of neoplastic cells (1, 2), and ampli-fication of MO-mediated tumoricidal activity could provide auseful modality for adjuvant therapy of neoplasia. The term"activated" MO is an operational definition and its use isoften extended to MO displaying enhanced phagocytic (3), ormicrobicidal properties (or both) (4, 5) in response to a vane-ty of inflammatory stimuli, but which lack tumoricidal activ-ity. In 'this article the term activated will be restricted tothose MO that display tumoricidal activity against the B16melanoma.MO activation is believed to involve a complex sequence

of phenotypic changes that are acquired in stepwise fashion,resulting in the development of microbicidal activity fol-lowed by what appears to be the ultimate step in activation,expression of tumoricidal capacity (6-11). Many agents ofdiverse biological origin and unrelated chemical structure,including Corynebacterium parvum, endotoxin, muramyl di-

peptide, and soluble mediator(s) (lymphokines) releasedfrom antigen or mitogen-stimulated lymphocytes, have beenused alone and in combination to render MO tumoricidalboth in vitro and in vivo (1, 12, 13). Whether activation bythese agents occurs via a common mechanism and whetherMO activated by different agents express qualitatively orquantitatively similar phenotypic markers} such as the ap-pearance or disappearance of cell surface antigens has notbeen established.The use of monoclonal antibodies as probes to analyze MO

has been directed toward identifying cell surface antigenswhose presence (or absence) correlates closely with a specif-ic stage in monocyte-MO differentiation. Ho and Springer(14) characterized a mononuclear phagocyte antigen termedMac-2, whose expression is elicited only by strong inflam-matory stimuli. In contrast, expression of a MO antigen rec-ognized by another monoclonal antibody, F4/80, is signifi-cantly diminished in activated MO (15).The present studies were undertaken to identify murine

MO cell surface antigens whose diminished or enhancedexpression was correlated with the tumoricidal phenotype.We describe the production of a monoclonal antibody,MA158.2, whose binding specificity is restricted to cells ofthe monocyte-Mo lineage and which reacts with an antigenwhose expression is enhanced on mononuclear cells whenactivated to the tumoricidal state in vivo and in vitro.

MATERIALS AND METHODS

Animals. Specific pathogen-free inbred C57BL/6, DBA/2,C3H/HeN, and BALB/c mouse strains and Lewis rats wereobtained from Charles River Breeding Laboratories and theDepartment of Laboratory Animal Sciences, Smith Kline &French. Mice were matched for age, sex, and body weightwithin each experiment.

Cells. Peritoneal exudate cells (PEC) were obtained afterintraperitoneal injection of one of'the following reagents: 2.0ml of Brewer's thioglycollate medium, 1.5 ml of 10% prote-ose peptone, 40 ug of Salmonella typhosa, lipopolysaccha-ride (LPS) (Westphal) (all reagents were from Difco), 25 Mgof concanavalin A (Sigma), or 350 jig (0.2 ml) of C. parvum(Wellcome). The time of elicitation was 3 days for proteosepeptone, concanavalin A, and LPS, 5 days for thioglycollate,and 7 days for C. parvum. PEC obtained 8 hr after intraperi-toneal inoculation of thioglycollate broth were used as asource of polymorphonuclear leukocytes (neutrophils).

Resident peritoneal MO (RPMO), PEC, and alveolar MOwere obtained from untreated mice as described (16). Great-er than 95% of the adherent cells were able to phagocytose

Abbreviations: MO, macrophage(s); RPMO, resident peritoneal MO;MAF, MO-activating factor; TPM4, thioglycollate-elicited MO;AcM4, MAF-activated TPM4; PEC, peritoneal exudate cells; IBA,trace indirect radioimmunobinding assay; LPS, lipopolysaccharide;GAR, goat F(ab')2 anti-rat Fab.

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

4505

4506 Immunology: Koestler et al.

carbon particles and showed positive staining for nonspecificesterase (17).Murine peripheral blood monocytes were obtained by car-

diac puncture and harvested in acid-citrate/dextrose, anderythrocytes were lysed by hypotonic treatment (18). Lym-phoid organs such as spleen, thymus, and peripheral lymphnode were obtained from 3- to 4-wk-old mice. For analysis ofcells maintained in suspension, cells were resuspended inFalcon no. 2053 tubes at a density of 5 x 105 per ml in Dul-becco's modified Eagle medium (DME medium) or DMEmedium containing MO-activating factor (MAF) and incu-bated for 24 hr before antibody binding analysis. Cell viabili-ty, as determined by trypan blue dye exclusion, was always>85%.Lymphokine. A lymphokine preparation containing MAF

activity was obtained from cell-free supernatants from cul-tures of normal F344 rat lymphocytes incubated for 48 hrwith immobilized concanavalin A (Pharmacia), as describedelsewhere (19). Control preparations were obtained fromnormal rat lymphocyte cell-free supernatants harvested un-der the same 'conditions from' lymphocyte cultures not ex-posed to concanavalin A.

Cell Lines. The following cell lines were obtained fromAmerican Type Culture Collection: J774.1 (20); P388D1 (20);WEH1-231 (20); RAW-264 (20); SP2/0-Agl4 (21); YAC-1(22); X63-Ag8 (23). All lines were maintained in DME medi-um containing 10% fetal bovine serum (CDME medium).

Antisera. Goat F(ab')2 anti-rat Fab (GAR) and fluoresceinisothiocyanate-corijugated GAR, free of cross-reactivity tomouse immunoglobulin, were obtained from Cappel Labora-tories. Trace labeling ofGAR with Na 251 was carried out byNew England Nuclear. The rat monoclonal anti-mouse tryp-sin-resistant Fc receptor antibody 2.4G2 (24) was obtainedfrom New England Nuclear. Monoclonal antibodies derivedfrom the following hybridomas were obtained from Hybri-tech (San Diego, CA): M1/70 HL (25); M1/42 (26); M1/69HL (26); M1/114 (27); M5/49 (28). The hybridomas M3/38(29) and M3/84 (30) were obtained from American Type Cul-ture Collection. Monoclonal antibody F4/80 (15) was a gen-erous gift from S. Gordon (Univ. of Oxford).Production of Monoclonal Antibodies. PEC elicited by

thioglycollate (TPMO) from female C57BL/6 mice were cul-tured for 2 hr at 107 cells per'100-mm tissue culture dish (Fal-con no. 3003). Nonadherent cells were removed by washingwith DME medium and the adherent monolayers (>95%MO) were treated with a 1:5 dilution of a MAF preparationdetermined previously to induce maximal tumoricidal activi-ty. Following treatment for 24'hr, the MAF-activated TPM4(AcM4) were washed three times with DME medium andharvested with a rubber policeman. Cell viability, as deter-mined by trypan blue dye exclusion, was between 40% and60%. Lewis rats were primed by intraperitoneal and subcuta-neous injection of 2 x 107 AcMO in complete Freund's adju-vant (GIBCO). Two weeks later they were inoculated in asimilar fashion but by using incomplete adjuvant. Spleenswere removed 3 days after a final inoculation of AcM4 ad-ministered intravenously and fused with the'SP2/0-Agl4 my-eloma line (21) following a previously described procedure(31).

Characterization of MA158.2. Ouchterlony analysis wasperformed by using subclass-specific antisera (Miles). Toobtain biosynthetically labeled antibody for chain analysis,hybridoma 158.2 was cultured for 8 hr with 200 ,Ci (1 Ci =37 GBq) of [35S]methionine (New England Nuclear) in 1 mlof methionine-free DME medium supplemented with 5% dia-lyzed fetal bovine serum. Culture supernatant was then ex-amined by one-dimensional (32) and two-dimensional Na-DodSO4/polyacrylamide gel electrophoresis according tothe method of Van Blerkom (33) as adapted by Greig et al.(34). Karyotypic analysis was performed by Hazelton Labo-

ratories America (Vienna, VA). Staphylococcus aureus(Cowan 1) protein A (IgGsorb) was obtained from The En-zyme Center (Boston, MA).

Cell Surface Marker Analysis. For initial screening andanalysis of monoclonal antibody binding to cells, a trace ra-dioimmune binding assay (IBA) was used (35). Binding of ratmonoclonal antibody to cells was detected by 1251-labeledGAR (125_-GAR). IBA was carried out on adherent cells (2 x105 per well) in 96-well plastic tissue culture dishes (Falconno. 3042 Microtest II) or for cells in suspension in polypro-pylene tubes. Cells in DME medium supplemented with 1%bovine serum albumin/10 mM NaN3 (DME medium/albu-min) were first incubated with saturating amounts of mono-clonal antibody for 60 min at 40C, washed three times withDME medium/albumin, and then incubated for an additional60 min at 40C with I251-GAR. The cells were washed threetimes with DME medium/albumin and the radioactivitybound was counted in a gamma counter. For flow cytome-try, cells in suspension were prepared as above and labeledwith fluorescein isothiocyanate-GAR and analyzed on anEPICS V cell sorter (Coulter) by using appropriate scatterwindows to gate out cellular debris and cell aggregates andpropidium iodide uptake to gate out'dead cells (36).MO-Mediated Cytotoxicity Assay. MO-mediated cytotoxi-

city was assessed in an in vitro 72-hr assay employing B16melanoma target cells labeled for 18 hr with 5-[125 ]iodo-2'-deoxyuridine (37). Effector cells (TPMO) were treated for 24hr in vitro with a final dilution (1:5) of either MAF or super-natant from control lymphocytes. Percentage cytotoxicitywas calculated as described (37).

RESULTS

Production and Characterization of Antibody MA158.2. Togenerate a monoclonal antibody that would be restricted inspecificity for murine M+, Lewis rats were hyperimmunizedwith a purified population of TPM4 activated to the tumori-cidal phenotype with MAF (AcMO) and supernatants from>1500 hybridoma -clones were screened for antibody thatbound preferentially to AcM4 as detected by an IBA. Onehybridoma clone (158.2) secreting monoclonal antibody(MA158.2) that reacted perferentially with AcMO was identi-fied and isolated. Analysis of reduced hybridoma 158.2 su-pernatant on one- and two-dimensional NaDodSO4/PAGEfollowed by Coomassie blue staining or fluorographic analy-sis' of [35S]methionine metabolically labeled MA158.2 re-vealed a single heavy and light chain. Gel diffusion analysiswith subclass-specific antisera revealed that MA158.2 is ofthe rat IgG2a isotype. Consistent with this subclass designa-tion is the observation that MA158.2 does not absorb to S.aureus (Cowan 1) protein and is not directly cytotoxicagainst mouse MO in the presence of rabbit, mouse, or ratserum as a source of complement (results not shown).

Expression of MA158.2 Antigen on Primary Mouse Cellsand Established Cell Lines. Culture supernatants from 158.2and M1/70, a hybridoma which secretes monoclonal anti-body that binds to the iC3b receptor on murine and humangranulocytes (38), were used to analyze the specificity ofeach antibody toward various lymphoid cells as detected byIBA. Dilutions of hybridoma supernatants used were in anti-body excess, as previously determined employing AcM4 astargets. The specificity of MA158.2 binding was restricted tomurine cells of the mononuclear phagocyte series (Table 1)and this was confirmed in several experiments in which rep-licate cell populations, pretreated with MAF, were analyzedfor the expression of MA158.2 antigen by immunofluores-cence (results not shown). Nylon wool-purified splenic Tcells, blood leukocytes, and suspensions of cells from lymphnode and thymus failed to bind detectable levels ofMA158.2. In a panel of established murine cell lines of di-

Proc. NatL Acad Sci. USA 81 (1984)

Proc. Natl. Acad Sci USA 81 (1984) 4507

Table 1. Expression of MA158.2-reactive antigen onprimary cells

125I-GAR bound,cpm x 10-2*

Cell type % M+* MA158.2 M1/70 HL

RPM4 (mouse) 80-90 6 ± 2 73 ± 3TPM4 (mouse) 95-98 20 ± 2 67 ± 3AcM4 (mouse) 95-98 78 ± 3 70 ± 2AcM4 (rat) 95-98 3 ± 1 NDAlveolar MO (mouse) 95-98 10 ± 1 30 ± 3Splenic MO (mouse) 60-70 9 ± 1 NDNeutrophil (mouse) 5-20 3 ± 1 48 ± 2

Live cells were assayed by IBA. Briefly, adherent monolayers (2x 105) were incubated with hybridoma culture supernatants (50 p11per well) for 60 min at 40C, washed three times with DME medium/albumin, and then stained with 125I-GAR (6 x 105 cpm, 50 kd perwell). ND, not determined.*The percentage of adherent cells representing MO was determinedby their ability to phagocytose carbon particles and upon positivestaining for nonspecific esterase as described (17).*Means of triplicates ± SEM are given. Background (180 ± 110cpm) was determined on AcM0 labeled with 10 jig of normal ratIgG per ml.

verse biological origin, including WEH1-231, P388D1, RAW-264, J774.1, B16, YAC, X63-Ag8, and SP2/0-Agl4, all failedto bind MA158.2 (results not shown). The antigen detectedby MA158.2 was not strain or sex specific and was found onTPMO and AcMO from mice carrying H-2d, H-2b, and H-2khaplotypes, respectively (Table 2).

Expression of MA158.2 Antigen on PEC Induced by Vari-ous Agents. To determine whether MA158.2 reacted with anantigen on a specific subpopulation(s) of MO(s), PEC elicit-ed by inflammatory agents, bacteria, and other stimuli wereexamined for reactivity with MA158.2 by IBA (Fig. 1). PECelicited with thioglycollate, proteose peptone, concanavalinA, C. parvum, and LPS all expressed MA158.2-reactive anti-gen, though the absolute level of expression was dependenton the nature of the elicitation agent. PEC elicited by thio-glycollate, proteose peptone, concanavalin A, and LPS ex-posed to MAF for 24 hr resulted in an enhanced expressionof the 158.2 antigen. In contrast, PEC elicited with C. par-vum failed to display enhanced antigen expression followingsimilar treatment. To determine if the enhanced antigenexpression observed in MAF-treated PEC was attributableto a highly reactive MO subpopulation rather than a uniformincrease in antigen expression, flow cytometry was em-ployed. There was a significant increase (=4-fold) in themean cell fluorescence of TPM4 treated with MAF as com-pared to the TPM4 treated with a control lymphokine prepa-

Table 2. Sex and strain specificity of MA158.2125IGAR bound,

Mouse cpm x 10-2

strain Haplotype Sex TPMO AcMO Ratio*

C57BL/6 H-2b d 20 ± 1 79 ± 3 3.9Y 19 2 78 ± 1 4.1

DBA/2 H-2d d 19 2 72 ± 2 3.8Y 19 1 75 ±3 3.9

BALB/c H-2d d 20 1 74 ± 1 3.7Y 20 ± 2 75 ±1 3.7

C3H/HeN H-2k d 17 ± 1 71 ± 3 4.2Y 16 ±+2 62 ± 1 3.9

Live cells were assayed by IBA as adherent monolayers as de-scribed in the legend for Table 1. Cell cultures were treated witheither MAF or control lymphokine for 24 hr before analysis by IBA.Results are expressed as means of triplicates ± SEM.*Ratio of AcM4 to TPM4.

76

6

5-

TPM4 Peptone Con A C. parvum LPS

FIG. 1. Analysis of MA158.2 binding to peritoneal MO elicitedby different agents. Adherent cell monolayers were treated with ei-ther MAF (i) or control (ol) lymphokine for 24 hr, washed threetimes with CDME medium and then assayed by IBA for reactivitywith MA158.2 as described in the legend for Table 2. Results areexpressed as means of triplicates ± SEM.

ration (Fig. 2). As a control, AcMO were treated with irrele-vant antibody (normal rat IgG2a).

Quantitative Binding Analysis of Different Anti-Mononucle-ar Cell Monoclonal Antibodies to Elicited and Activated Peri-toneal Mo. To determine whether MA158.2 binds to a previ-ously recognized Mo antigen, a series of rat monoclonalantibodies characterized by other investigators as havingspecificity for determinants found on mononuclear cellswere examined by IBA for their pattern of reactivity againstTPM4 and AcM4 (Table 3). Expression of the MA158.2-re-active antigen was enhanced 3-fold following in vitro activa-tion of TPM4O with MAF, whereas none of the other mono-clonal antibodies examined displayed any significant in-crease in binding concomitant with activation. In contrast toMA158.2, monoclonal antibody F4/80 displayed decreasedbinding to AcMO, consistent with previously published find-

x

6

Clo:

10 100Relative fluorescence intensity

FIG. 2. Immunofluorescent flow cytometry analysis of TPMOstained with MA158.2 plus fluorescein isothiocyanate-conjugatedGAR. Control treated for 24 hr with MAF and labeled with 25 pAg ofnormal rat IgG2a per ml (.), treated for 24 hr with a lymphokinepreparation devoid of MAF activity (------), and treated for 24 hrwith MAF ( ). Forward and right-angle scatter gates were set toexclude lymphocytes and dead cells from analysis by using propi-dium iodide as a viability stain. Fluorescence intensity is plotted ona logarithmic scale and relative cell number is on a linear scale.

Immunology: Koestler et aL

4508 Immunology: Koestler et al.

Table 3. Reactivity of various anti-MO monoclonal antibodieswith elicited and MAF-activated peritoneal MO

125I-GAR bound,

Hybridoma Ref. Designation Ig class cpm X 10-2TPM4 AcM4

158.2 MA158.2 IgG2a 20 ± 2 74 ± 2M1/70HL 25 MAC-1 IgG2b 70 ± 2 60 ± 1M3/38 29 MAC-2 IgG2a 72 ± 2 75 ± 2M3/84 30 MAC-3 IgG1 20 ± 1 20 ± 1F4/80 15 (F4/80) IgG2b 56 ± 2 36 ± 22.4G2 24 Fc receptor IgGl 46 ± 2 40 ± 2M1/42 26 H-2K IgG2a 43 ± 2 33 ± 1M1/69HL 26 Heat-stable IgG2b 49 ± 2 43 ± 2

antigenM1/114 27 Ia IgG2b 67 ± 1 65 ± 3M5/49 28 Thy-1 IgG2a 6 ± 1 3 ± 1

ings (15). Several reports have indicated that MO possesstrypsin-sensitive Fc receptors for the IgG2a subclass (24,39). To exclude the possibility that the enhanced binding ofMA158.2 (IgG2a) to AcM4 is due to nonspecific binding tothese MO Fc receptors, AcM4 were treated with trypsin (1mg/ml ofDME medium) for 30 min at 37°C prior to the IBA.Trypsin treatment had no effect on MA158.2 binding toAcM4 (results not shown). The lack of binding of MA158.2to mouse neutrophils or to rat AcMO provided further evi-dence against nonspecific binding to Fc receptors (Table 1).Furthermore, monoclonal antibodies to Mac-2, H-2 mono-

typic, and Thy-1 antigens, also of the rat IgG2a isotype, donot display augmented binding to MAF-activated MO (Table3).

Functional Correlation of MA158.2-Reactive Antigen. Todetermine if MA158.2 could serve as a useful probe for moni-toring MO activation in vitro, replicate MO cultures were ex-

amined to correlate the kinetics of MAF-induced expressionof MA158.2-reactive antigen with development of tu-moricidal activity. Acquisition of tumoricidal capacityfollowing treatment with MAF parallels increased expres-sion of antigen recognized by MA158.2 (Fig. 3). Maximalexpression of tumoricidal capacity and MA158.2-reactiveantigen was essentially complete following treatment withMAF for 24 hr.

DISCUSSION

Several lines of evidence suggest that MA158.2 described inthis paper recognizes a MO surface determinant not identi-fied hitherto and that an enhanced or an elevated expressionof the 158.2 antigen is related to development of the tumori-cidal phenotype.

First, MA158.2 binds to an antigen only detected on Mo.An unexpected result was the finding that MA158.2 does notdisplay measurable binding to a panel of established mono-

cytic cell lines. This may reflect the transformed nature ofthese lines and their antiquity in culture. Second, TPMOtreated with MAF displayed a significant increase (3-fold) inbinding of MA158.2. In contrast, under identical experimen-tal conditions, no alterations in antibody binding were de-tected by using a panel of monoclonal antibodies reportedpreviously to react selectively with murine Mo. Third, thekinetics of enhanced expression of the antigen after treat-ment with MAF parallels the evolution of tumoricidal capac-ity. Previous studies from this laboratory have demonstratedthat maximal expression of tumoricidal properties, as deter-mined in a 72-hr cytotoxicity assay using B16 melanomacells as targets, requires 24 hr of exposure to lymphokinepreparations containing MAF (40). Consistent with this ob-servation is the finding that peak expression of MA158.2-reactive antigen also occurs at 24 hr.

0

x

6a,

'0

04I>t0c)._,o

Activation time, hr

FIG. 3. Correlation of the expression of MA158.2-reactive anti-gen with acquisition of the tumoricidal phenotype. TPM4 were ob-tained and adherent cell monolayers were treated with a 1:5 dilutionof MAF for the times indicated, washed three times with CDMEmedium, fixed in 1% paraformaldehyde for 15 min at 22°C, and as-sayed by IBA as described in the legend for Table 2. Replicate cul-tures were treated in the same manner, except after washing cellswere refed with CDME medium (100 ,ul per well) and maintained inculture for a total time of 24 hr. After a total culture time of 24 hr,the cytotoxic potential of the MAF- and control-treated cultureswere determined'by the addition of 5000 B16 melanoma target cellsprelabeled with 5-[1251]iodo-2'-deoxyuridine at a macrophage/targetcell ratio of 20:1. The percentage cytotoxicity as compared with thecontrol-treated M40 was determined at 72 hr. Results are expressedas means of quadruplicates + SEM.

Recently, Taniyama and Tokunaga (41) have describedthree monoclonal antibodies that recognize antigens associ-ated with different stages ofMO activation. WE15 displayedbroad specificity of binding to MO in various stages of acti-vation. MM9 recognized RPM4, TPM4, and proteose pep-tone-elicited MO, whereas ACM 1 specificity was restrictedto MO elicited with pyran and C. parvum. However, the re-activity of these monoclonal antibodies against MO renderedtumoricidal by MAF was not reported. The specificity andpattern of binding of MA158.2 demonstrated here stronglysuggest that this is a unique antibody whose specificity isdistinct from other anti-MO monoclonal antibodies [Mac-1,-2, and -3 (29, 30), F4/80 (15), 2.4G2 (24), M1/114 (27)].Immunoprecipitation of the antigen recognized by MA158.2has so far been unsuccessful.

M4f populations are not homogeneous but comprise ofsubpopulations that may display distinct phenotypic markersand respond differently to challenge with agents such asMAF and LPS (42). The usefulness of MA158.2 as a tool fordistinguishing among M4 populations elicited with differentagents is illustrated by the present data showing that M4elicited with different agents display different quantitativelevels in the expression of MA158.2-reactive antigen beforeand after treatment with MAF. Although MO elicited with C.parvum did not display enhanced binding of MA158.2 fol-lowing treatment with MAF, these MO express tumoricidalproperties against the B16 melanoma in the absence ofMAFtreatment. Furthermore, C. parvum MO are significantlysmaller than M4 'elicited with thioglycollate, proteose pep-tone, concanavalin A, and LPS, which may account for thelower level of expression of MA158.2-reactive antigen (43).The determination of epitope density of 158.2 antigen on MOelicited with different agents has thus far been unsuccessfuldue to the loss of binding properties of MA158.2 followingiodination (unpublished' observation). -In light of the well-documented heterogeneity in biochemical (11, 44), morpho-logical (44-46), and, more importantly, functional character-

Proc. NatL Acad Sci. USA 81 (1984)

Proc. Natl. Acad. Sci USA 81 (1984) 4509

istics (3-5, 47-50), MO elicited by diverse agents can beviewed as distinct cell subpopulations that are capable of re-sponding differently to activation stimuli. However, it is notyet clear whether MO elicited by various agents are differen-tiating along distinct maturation pathways or are induced todistinct phenotypes along the same pathway.

It has been reported that lymphokine preparations con-taining MAF do not render MO fully cytolytic (51) and thatMAF instead "primes" MO to become more responsive to asecond signal(s) such as LPS, which triggers final expressionof tumoricidal properties. The "priming" activity in prepara-tions of MAF has also been reported to be mimicked bycloned murine interferon-y (52). The precise contribution ofMAF, interferon-)y, or LPS in the induction of tumoricidalproperties by the lymphokine preparations used in the pres-ent experiments is unknown. Nevertheless, we have foundthat enhanced expression of the antigen recognized byMA158.2 offers a reliable in vitro marker for development ofthe tumoricidal phenotype and MA158.2 will be useful in dis-secting the mechanism of MO activation provoked by differ-ent combinations of interferon-y, LPS, and lymphokinepreparations containing MAF.

We thank Marcy Hughes and Jo Ann Cavallaro for typing thismanuscript. Flow cell cytometry expertise was provided by Dr. PaulHoran and Tom Schmitt and their support is greatly appreciated. Weare also grateful to Drs. Nabil Hanna and Deborah Trainer for sug-gestions and critical reviews of the manuscript.

1. Fidler, I. J. & Hanna, M. G. (1981) in Fundamental Mecha-nisms in Human Cancer Immunology, eds. Saunders, J. P.,Daniels, J. C., Serrou, B., Rosenfeld, C. & Denney, C. B. (El-sevier/North-Holland, Amsterdam), pp. 425-438.

2. Fidler, I. J. (1978) Isr. J. Med. Sci. 14, 177-191.3. Cohn, Z. A. (1978) J. Immunol. 121, 813-816.4. North, R. J. (1978) J. Immunol. 121, 806-808.5. Nathan, C. F., Nogueira, N., Juangbhanich, J., Ellis, J. &

Cohn, Z. A. (1979) J. Exp. Med. 149, 1056-1068.6. Fidler, I. J. & Raz, A. (1981) Lymphokines 3, 345-363.7. Adams, D. (1982) Immunol. Today 3, 285-287.8. Hibbs, J. B., Taintor, R. R., Chapman, H. A. & Weinberg,

J. B. (1977) Science 197, 279-282.9. Ruco, L. P. & Meltzer, M. S. (1978) J. Immunol. 121, 2035-

2042.10. Meltzer, M. S. (1981) Lymphokines 3, 319-343.11. Johnson, W. J., Marino, P. A., Schreiber, R. D. & Adams,

D. 0. (1983) J. Immunol. 131, 1038-1043.12. Schultz, R. M. (1980) Adv. Pharmacol. Chemother. 17, 157-

192.13. Chirigos, M. A., Mitchell, M., Mastrangelo, M. J. & Krim,

M., eds. (1981) Modulation of Cellular Immunity in Cancer byImmune Modifiers (Raven, New York).

14. Ho, M. K. & Springer, T. A. (1982) J. Immunol. 128, 1221-1228.

15. Ezekowitz, R. A. B., Austyn, J., Stahl, P. D. & Gordon, S.(1981) J. Exp. Med. 154, 60-76.

16. Poste, G., Bucana, C., Raz, A., Bugelski, P., Kirsh, R. &Fidler, I. J (1982) Cancer Res. 42, 1412-1422.

17. Tucker, S. B., Pierre, R. V. & Jordon, R. E. (1977) J. Immu-nol. Methods 14, 267-269.

18. Ford, W. L. (1979) in Handbook of Experimental Immunolo-

gy, ed. Weis, D. M. (Blackwell Scientific Publications, Ox-ford), Vol. 2, p. 23.

19. Fidler, I. J., Darnell, J. H. & Budmen, M. B. (1976) J. Immu-nol. 117, 666-673.

20. Ralph, P. (1981) Lymphokines 4, 175-195.21. Russel, W. C., Newman, C. & Williams, D. H. (1979) Nature

(London) 253, 461-463.22. Cikes, M., Friberg, S. & Klein, G. (1973) J. Natl. Cancer Inst.

50, 347-354.23. Littlefield, J. W. (1964) Science 145, 709-712.24. Unkeless, J. C. (1979) J. Exp. Med. 150, 580-596.25. Springer, T. A., Galfre, G., Secher, D. S. & Milstein, C.

(1979) Eur. J. Immunol. 9, 301-306.26. Springer, T. A. (1980) in Monoclonal Antibodies, eds. Ken-

nett, R. H., McKearn, T. J & Bechtol, K. B. (Plenum, NewYork), pp. 185-217.

27. Bhattacharya, A., Dorf, M. & Springer, T. A. (1981) J. Immu-nol. 127, 2488-2495.

28. Davignon, D., Martz, E., Reynolds, T., Kurzinger, K. &Springer, T. A. (1981) J. Immunol. 127, 590-595.

29. Ho, M. K. & Springer, T. A. (1982) J. Immunol. 128, 1221-1228.

30. Ho, M. K. & Springer, T. A. (1983) J. Biol. Chem. 258, 636-642.

31. Kohler, G. & Milstein, C. (1975) Nature (London) 256, 495-497.

32. O'Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021.33. Van Blerkom, J. (1978) in Methods in Mammalian Reproduc-

tion (Academic, New York), pp. 68-109.34. Greig, R. G., Caltabiano, L., Reid, R., Feild, J. & Poste, G.

(1983) Cancer Res. 43, 6057-6065.35. Austin, J. M. & Gordon, S. (1981) Eur. J. Immunol. 11, 805-

815.36. Muirhead, K. & Horan, P. (1984) in Advances in Cell Culture,

ed. Maramorosch, K. (Academic, New York), Vol. 3, in press.37. Raz, A., Folger, W. E. & Fidler, I. J. (1979) Cancer Immunol.

Immunother. 7, 157-163.38. Ho, M. K. (1983) J. Biol. Chem. 258, 2766-2769.39. Unkeless, J. C., Kaplan, G., Plutner, H. & Cohn, Z. (1979)

Proc. NatI. Acad. Sci. USA 76, 1400-1404.40. Fidler, I. J., Raz, A., Fogler, W. E., Hoyer, L. C. & Poste, G.

(1981) Cancer Res. 41, 495-504.41. Taniyama, T. & Tokunaga, T. (1983) J. Immunol. 131, 1032-

1037.42. Fogler, W. E., Talmadge, J. E. & Fidler, I. J. (1983) J. Reticu-

loendothel. Soc. 33, 165-174.43. Miller, G. A. & Morahan, P. S. (1982) J. Reticuloendothel.

Soc. 32, 111-124.44. Karnovsky, M. L. & Lazdins, J. (1978) J. Immunol. 121, 809-

812.45. Wolman, M. & Eldar, T. (1981) Cell Mol. Biol. 27, 543-549.46. Robertson, T. A., Papadimitriou, J. M., Walters, M. & Wol-

man, M. (1977) J. Pathol. 123, 157-164.47. Lee, K. C. & Berry, D. (1977) J. Immunol. 118, 1530-1540.48. Stuart, A. E. (1977) in The Macrophage and Cancer, eds.

McBride, J. K. & Stuart, A. (Univ. of Edinburgh, MedicalSchool, Edinburgh, Scotland), p. 1.

49. Sun, D. & Lohmann-Matthes, M. L. (1982) Eur. J. Immunol.12, 134-140.

50. Sorg, C. (1982) Mol. Immunol. 19, 1275-1278.51. Pace, J. L. & Russell, S. W. (1981) J. Immunol. 127, 1863-

1867.52. Pace, J. L., Russell, S. W., Schreiber, R. D., Altman, A. &

Katz, D. H. (1983) Proc. Natl. Acad. Sci. USA 80, 3782-3786.

Immunology: Koestler et aL