CELLULAR IMMUNOLOGY 98, 57-67 (1986)

P-Glucuronidase Release from Human Monocytes Induced with Aggregated lmmunoglobulins of Different Classes’

NICHOLAS R. FERRERI, LUCINDA BECK, AND HANS L. SPIEGELBERG

Department of Immunology Research Institute of Scripps Clinic, La Jolla, California 92037

Received August 28, 1985; accepted October 15. 1985

We studied &ducuronidase release from human monocytes induced with aggregated immu- noglobulins of the nine different human classes and subclasses. Release was induced in a time and dose-dependent manner by all aggregated IgG subclasses. Aggregated IgA 1 caused a greater Q-glucuronidase release than aggregated IgM, IgD, and IgE, but the difference was not statistically significant. Release of Pglucuronidase was not phagocytosis dependent since inhibition of phago- cytosis by cytochalasin B or dihydrocytochalasin B did not diminish enzyme release. On the contrary, cells incubated with cytochalasin B prior to addition of aggregated IgG released ap proximately twice as much enzyme compared to untreated controls. Enzyme release induced by latex particles, a non-Fc receptor mechanism, was decreased by cytochalasin B. Monomeric IgGl , IgG2, IgG3, and IgG4 inhibited aggregated IgGl enzyme release in a dose-dependent manner. The ability of monomeric IgG to inhibit /3glucuronidase release correlated with previous reports describing the binding affinities of monomeric IgG to monocytes, i.e., IgG2 was relatively ineffective compared to the other subclasses. Monomeric IgA, IgE, and pentameric IgM were unable to diminish b&-induced enzyme release. The dam indicate that normal peripheral blood monocytes express predominantly Fc receptors for IgG and that all four IgG subclasses induce the release of the ly%Ximal enyme &ghCUrOnidaSe. 0 1986 Academic p, Inc.

INTRODUCTION

Monocytes and macrophages have been shown to release lysosomal enzymes fol- lowing interaction with antigen-antibody complexes (l-3). This process may contribute importantly to the ability of monocytes to participate in inflammatory reactions and reflects a general mechanism for lysosomal enzyme release from neutrophils and eo- sinophils as well (4-6). Phagocytosis need not occur to initiate Fc receptor mediated lysosomal enzyme release. Recently, eosinophils were shown to release peroxidase in response to anaphylactic immunoglobulins, IgG2a or IgE, but not to IgGl, IgG2b, IgG2c, or IgM (6). Release of peroxidase did not require phagocytosis and was only induced via receptors for IgG2a and/or IgE.

It is generally accepted that monocytes have receptors for the Fc fragment of IgG and that crosslinking of these receptors by IgG aggregates causes lysosomal enzyme release. Whether monocytes and macrophages have Fc receptors for other immuno-

i This is publication No. 3997 IMM from the Research Institute of Scripps Clinic, La Jolla, Calif. 92037. The work was supported by grants from the USPHS AI-10386, General Clinical Research Grant RR 00833, and Biomedical Research Support Program RR0 5514. Dr. Ferreri is supported by Training Grant GM 07437.

57

OOO8-8749/86 $3.00 Copyright 0 1986 by Academic Press, Inc. All ri&s of rcprodu~on in any form reurved.

58 FERRERI, BECK, AND SPIEGELBERG

globulin classes is less clear. Several investigators reported that human monocytes have Fc receptors for IgA and IgM (7, 8). Other laboratories could not demonstrate IgA Fc receptors on monocytes and macrophages either by functional (9) or rosette assays and binding of radiolabeled immunoglobulins to cells (10, 11). In past studies using human neutrophils, release of &$.tcuronidase induced by aggregated myeloma proteins of different classes correlated with the presence of immunoglobulin class specific Fc receptors (4). Therefore, we determined whether the release of &glucuron- idase from human peripheral blood monocytes stimulated with insoluble immune complexes of different classes and subclasses correlated to expression of Fc receptors. We also studied whether phagocytosis of the immune complexes was required for lysosomal enzyme release. For these experiments cytochalasin B or dihydrocytochalasin B, fungal metabolites known to inhibit phagocytosis (12), were added to cells before incubation with the aggregated immunoglobulins.

MATERIALS AND METHODS

Cell isolation and identification. Human venous blood from healthy donors was collected into l/10 vol of 0.2 M disodium EDTA, pH 7.2. The blood was centrifuged for 15 min at 8OOg, the buffy coat collected and diluted with RPMI- 1640 (GIBCO, Grand Island, N.Y.), and mononuclear cells (PBM) were isolated by Ficoll-Hypaque centrifugation. T-Cell depleted PBM (TCD/PBM) were prepared using neuraminidase (Behring Diagnostics, La Jolla, Calif.)-treated sheep erythrocytes (EN) as previously described (13). Briefly, EN were mixed with PBM for 15 min at 37°C centrifuged at low speed, and incubated for 60 min at 4°C. Following resuspension, cells were sep- arated by centrifugation on a Ficoll-Hypaque cushion. The interface cells (TCD/ PBM) were harvested to RPMI- 1640 containing 0.25% BSA (RPMI/BSA) and washed three times. T cells (EN-positive) recovered from the pellet after EN were lysed with 0.83% NH&l and washed three times with RPMI/BSA. Peripheral blood lymphocytes (PBL) were prepared from PBM depleted of monocytes by treatment with colloidal iron followed by Ficoll-Hypaque centrifugation (14). Neutrophils (PMN) were prepared as previously described (15). Viability assessed by trypan blue exclusion was greater than 98% for all cell preparations tested, release of lactate dehydrogenase (LDH) was 2-3% following incubation with immunoglobulin aggregates.

The number of monocytes present in each cell preparation was determined by nonspecific esterase staining (16). T and B cells were identified by rosette formation with EN and ox erythrocytes (Eo) coated with glutaraldehyde-coupled complexes of rabbit Fab’-anti-E0 and immunoadsorbent-purified goat F(ab’)2 anti-Ig antibodies, re- spectively ( 17). Neutrophils were identified by Wright-Giemsa staining. The following represent typical values for the different cell preparations: PBM, 65-70% T cells, 15-20% monocytes, lo-15% B and null cells, 2-4% basophils; TCD/PBM, 50-65% monocytes, 20-30% B and 20-30% null cells, 2-4% basophils, l-3% T cells; T cells 80-90% T cells, l-3% B cells, l-3% monocytes; PBL > 98% lymphocytes, < 1% mono- cytes; PMN > 95%, 3-5% eosinophils.

Human myeloma proteins. Myeloma proteins K,IgG 1 (Er) K&G 1 (Ma), X,&G2 (Ke), K,IgG2 (Lu), K,IgG3 (Fi), X,IgG4 (He), ~,1gAl (As), X&Al (Pu), X&42 (Ta), ~,Igh4

(An), ~,1gM (Vi), X,IgD (Di) and X,IgE (PS) were isolated as previously described (11). Rabbit IgG antibodies to bovine serum albumin (BSA) were purified from pooled

antisera using BSA coupled to Sepharose4B (Pharmacia Fine Chemicals, Piscataway,

B-GLUCURONIDASE RELEASE FROM MONOCYTES 59

N.J.). The antibodies were eluted from the column with 3.5 M NaSCN, dialyzed against PBS, and concentrated by pressure dialysis.

Aggregation of irnmunoglobulins with bis-diazotized-benzidine (BDB). Immuno- globulins were aggregated with BDB according to Ishizaka et al. (18). Briefly, S mg of immunoglobulin was dialyzed against borate buffer, pH 8.5. The protein was aggregated with BDB for 1 S min at 0°C then dialyzed at 4°C against two changes of phosphate- buffered saline (PBS). The aggregates were vigorously pipetted before addition to pro- duce as uniform a suspension as possible.

Aggregation of immunoglobulins with purified rabbit IgG anti-BSA. Insoluble im- mune complexes were formed by incubating BSA with purified antibodies in antibody excess. After a I-hr incubation at 37°C complexes were incubated overnight at 4°C and washed three times with PBS. An aliquot was solubilized with 0.1 N NaOH and the concentration determined by optical density at 280 nm.

Incubation of cells with insoluble immune complexes. Two million esterase positive cells were resuspended in Earle’s balanced salt solution (GIBCO, Grand Island, N.Y.) without phenol red containing 0.25% BSA (EBSS/BSA) and incubated with immune complexes in a volume of 200 ~1 for various times. Cells were preincubated with inhibitors (cytochalasin B, dihydrocytochalasin B monomeric immunoglobulins) for S or 10 min, respectively, at room temperature before addition of the immune com- plexes. In some experiments, cells were incubated with latex particles at a particle to cell ratio of SO:1 (2.02 Km, Dow Chemical Co., Indianapolis, Ind.) prepared as pre- viously described (3). Following additions, cells were pelleted (5 min, 10°C 100s) to bring them in contact with insoluble immune complexes and then gently resuspended before incubation at 37°C. The reaction was terminated by placing tubes in an ice bath for 5 min. Cells were removed by centrifugation and the supematants assayed for ,&glucuronidase or LDH activity. fi-Glucuronidase activity was determined by a calorimetric assay using phenolphthalein glucuronic acid (Sigma Chemical Co., St. Louis, MO.) as substrate ( 19) and measuring absorbance at 540 nm. LDH activity was determined in a kinetic assay using pyruvate as substrate (Sigma Chemical Co., St. Louis, MO.) by measuring absorbance at 340 nm. Total enzyme content was obtained by lysing cell suspensions by three cycles of freeze-thaw using liquid nitrogen. Results are expressed as percentages of the total enzyme release, assays were done in triplicate.

Histamine bioassay. Histamine released into the supemate was detected by guinea pig ileum assay. Strips (lo-15 mm) of ileum were suspended in a 2-ml organ bath at 37°C with oxygenated Tyrode’s buffer. Responsiveness was determined using histamine and acetylcholine (Sigma Chemical Co., St. Louis, MO.) as standards. Pretreatment of tissue with an Hi-receptor antagonist, pyrilamine ( 1 X lob5 M) abolished histamine- induced contractions without affecting the response to acetylcholine. Supemates from PBM incubated with BDB-aggregated IgE or buffer, as described above, were tested for the presence of histamine by contraction of guinea pig ileum in the absence but not the presence of pyrilamine. Tissue viability following pyrilamine was assessed pretreatment by contractions to acetylcholine; these contractions were not different from those obtained before pretreatment.

RESULTS Release of P-Glucuronidase from PBA4, PBL, TCD/PBM, and Purified T C&

The percentage enzyme release from untreated cell controls which averaged 3 +- 1% (n = 15) of the total @-glucuronidase was subtracted from the percentage release of

60 FERRERI, BECK, AND SPIEGELBERG

cells incubated with aggregated immunoglobulins. Release of LDH was between 2 and 4% of the total LDH for all conditions and stimuli tested. To compare the &glucuronidase release from PBM and PBL, cells were incubated with different IgG immune complexes. PBM incubated with insoluble BSA-rabbit IgG anti-BSA immune complexes, BDB-aggregated normal human IgG or myeloma IgG 1 all released 8- 12% of total /3-glucuronidase (Table I). In contrast, PBL preparations, which contained less than 1% monocytes, did not release significant amounts of enzyme following incubation with the three types of IgG immune complexes. BGlucuronidase release induced by insoluble BSA-rabbit IgG anti-BSA immune complexes was compared in PBM, TCD/PBM preparations and purified T cells. When incubated with rabbit IgG immune complexes, &&tcuronidase release increased twofold in TCD/PBM compared to PBM preparation; viz., 12 + 5% vs 24 f 6% (mean + SD, n = 4). Purified T cells incubated with the same complexes did not release significant amounts of enzyme (2 f 1%). Similar results were obtained with normal human IgG or an IgGl human myeloma protein aggregated with BDB. These results indicated that monocytes were the cell type responsible for /3-glucuronidase release induced by aggregated IgG and thus permitted the use of PBM in subsequent experiments.

Dose-Response Stimulation of P-Glucuronidase Release from PBM by Aggregated Human Myeloma Proteins

The human myeloma proteins, K,IgGl (Ma) and, to a smaller extent, X&Al (Pu) aggregated with BDB released &glucuronidase in a dose-dependent, saturable manner (Fig. 1). Maximal release was induced by approximately 10 pg/ml insoluble IgG and IgA 1. Insoluble aggregates of IgM and IgE did not release enzyme in a dose-dependent manner. Although the percentage enzyme release by the IgAl myeloma protein (Pu) was higher than that of IgA 1 protein (As) and the IgE and IgM proteins, the differences were not statistically significant (Table 2). However, since the enzyme was released in a dose-dependent manner only by an IgA myeloma protein capable of forming rosettes with monocytes [20-30% IgA (Pu) vs 2-5% IgA (As)], release may have been induced in an Fc receptor related manner. To ensure a sufficient dose of aggregated protein, kinetic studies and testing of different myeloma proteins were performed at a dose of 250 pg/ml of immune aggregates.

TABLE 1

@-Glucuronidase Release from Human Mononuclear Cells (PBM) and Peripheral Blood Lymphocytes (PBL) Induced by Aggregated IgG

5% j3-glucuronidase release’

PBM PBL

Insoluble immune complexes BSA-rabbit IgG anti-BSA

Bis-diazotized benzidine (BDB) aggregated immunoglobulins Normal IgG (Cohn II) Myeloma, K&G 1

12 + 5 (4) 2 f 1 (3)

8 + 2 (2) 0 (1) 11 * 3 (3) 1 + 1 (3)

n PBM (2 X lo6 esterase’) or PBL (7 X 106) were incubated with insoluble IS; immune complexes (250 &ml) for 30 min at 37% Values represent mean + SD (number of experiments).

@GLUCURONIDASE RELEASE PROM MONGCYTES 61

FIG. 1. Dose response of &glucuronidase release from PBM induced by BDB-aggregated human myeloma proteins. Cells (2 X lo6 esterase+) in Earle’s balanced salt solution (without phenol red) + 0.25% BSA were incubated for 30 min at 37°C with insoluble immune complexes. Values represent mean -+ SD (n = 4). Proteins are identifted by abbreviation of patient’s name.

Kinetics of P-Glucuronidase Release Induced by Insoluble BDB-Aggregated Myeloma Proteins

Aggregated human myeloma proteins, qIgG1 (Ma) and X&Al (Pu), and BSA- rabbit IgG anti-BSA immune complexes induced rapid release of /3-glucuronidase (Figs. 2 and 5). Significant release was observed by 5 min for IgG and was maximal after 30 min. Enzyme release did not increase significantly up to 60 min. Aggregated qIgA 1 (As) and IgE myeloma proteins induced approximately 2% release, and aggre- gated IgM released less than 1% (Fig. 2). However, because of background variation, the release by rc,IgAl (As), IgM, and IgE was not statistically significant.

Eflects ofD$erent Immunoglobulin Classes and Subclasses on P-Glucuronidase Release from PBM and Neutrophils; Histamine Release from PBM

Release of &glucuronidase from PBM induced by aggregated myeloma proteins of the nine different human classes and subclasses is summarized in Fig. 3. Each point

TABLE 2

&Glucuronidase Release from Human Mononuclear Cells (PBM) and Neutrophils (PMN) by Aggregated Myeloma Proteins

Bisdiazotized benzidine (BDB) aggregated myeloma proteins

% P-glucuronidase release”

PBM PMN

&Cl (Ma) 12 k 4 (5)b 16 f 2 (4)b dgA1 (Ad 3 + 1 (4)’ 7 k 2 (4) udl (W 5 * 2 (4)’ 14 f 2 (4)b d&f (Vi) 2 f 2 (3)’ 3 + 2 (3) UgE (W 2 * 2 (3)’ 4 + 1 (3)

’ PBM (2 X lo6 esterase+) or PMN (2 X 106) were incubated for 30 min at 37°C with BDB-aggregated myeloma proteins of different classes. Values represent mean -t SD (number of experiments). Proteins are identified by abbreviation of patient’s name.

’ The dam were analyzed to test for group differences by a one-way analysis of variance; IgG released significantly (P < 0.001) more &glucuronidase than the other Ig classes.

’ Multiple comparisons of group means did not reveal a statistically sign&ant difference (P > 0.05).

62 PERRERI, BECK, AND SPIEGELBERG

Minutes

FIG. 2. Kinetics of &ghtcuronidase release from PBM induced by 250 g/ml BDB-aggregated human myeloma proteins. Cells (2 X lo6 esterase’) in EBSS + 0.25% BSA were incubated for the indicated times at 37°C Representative figure from five experiments.

represents release from individual normal donors. The figure also includes experiments involving different myeloma proteins of the same class or subclass. Significant fi-glucuronidase release was only obtained with IgG myeloma proteins. There was no significant difference between IgG proteins of the four different subclasses or another K,IgGl (Er) myeloma protein.

The BDB-aggregated myeloma proteins were also tested for their ability to release &glucuronidase from neutrophils. Both IgGl and IgA 1 proteins induced a significant percentage of &$tcuronidase release over background (Table 2). Aggregated IgG my- eloma proteins induced similar amounts of enzyme release from PBM and PMN. IgA preparations induced two to three times more enzyme from PMN compared to PBM. Aggregated IgM and IgE myeloma proteins were inactive in both PBM and PMN.

Because aggregated IgE preparations did not release j3glucuronidase from mono- cytes, we determined their capacity to release histamine from the 1 to 2% basophils

IgGl lgG2

n

: .

1 l .

l .

*.

u\a.El

.

: .

Ke

1963 lgG4

.

.

.

T-

IgAl lgA2

.

. .

IN

. .

0 -L

He As,Pu Kr All.“,

IgO

. : .

-iii-

IoE

. i ps

FIG. 3. @Glucuronidase release from PBM induced by 250 &ml BDB-aggregated human myeloma proteins of different classes. Cells (2 X lo6 esterase’) in EBSS + 0.25% BSA were incubated for 30 min at 37°C with aggregated myeloma proteins of the nine different human classes and subclasses. Bach point represents an experiment from individual normal donors.

,&GLUCURONIDASE RELEASE FROM MONOCYTES 63

present in PBM preparations. Supernates from PBM incubated with BDB-aggregated IgE induced contractions of guinea pig ileum at a 1:50 dilution. Contraction was immediate in onset and was completely inhibited by pretreatment of the tissue with pyrilamine (1 X lO-5 M), an Hi-receptor antagonist (Fig. 4). The tissue remained viable as shown by the identical contraction induced by acetylcholine before and after pyrilamine pretreatment. Contraction of the ileum induced by acetylcholine following pyrilamine pretreatment also indicated the specificity of receptor blockade.

Eflects of Cytochalasin B on &Glucuronidase Releasefrom PBM Induced by Aggregated Myeloma Proteins

To assess whether phagocytosis of the immune complexes was necessary for enzyme release, the effects of cytochalasin B and dihydrocytochalasin B on &lucuronidase release were investigated following incubation with immune complexes or latex par- ticles. The kinetics of enzyme release induced by insoluble BSA-rabbit IgG anti-BSA immune complexes (Fig. 5) was similar to that previously shown for BDB-aggregated human IgG myeloma proteins (Fig. 2). When cells were pretreated with 5 pg/ml cy- tochalasin B, a dose that inhibits phagocytosis by 85-90’70, enzyme release increased approximately twofold for rabbit immune complexes and BDB-aggregated myeloma proteins compared to untreated controls. Cytochalasin B potentiated enzyme release induced by IgG at all time points tested without altering the general kinetics of release. Similar results were obtained using dihydrocytochalasin B. As expected, pretreatment with cytochalasin B decreased enzyme release from cells incubated with latex particles, an insoluble non-Fc receptor stimulus, compared to untreated controls (Fig. 5). Cy- tochalasin B-treated cells were also challenged with aggregated IgA, IgM, and IgE immune complexes. Release of &+lucuronidase was not increased from cytochalasin- treated cells incubated with K&AI (As), IgM, and IgE. In contrast, compared to con- trols, cytochalasin B caused approximately a twofold increase in &&curonidase release from PBM incubated with X&Al (Pu).

t t Acetylcholme PBM

5x11S6M Supernate

T $ Pyrdamw 1 )I I Od5M

t

9

+ i(i + ll -i-- t Hwamme Acetylchohne Pt

1.5~10% 5 x 1 ml Supernate

FIG. 4. Histamine release from PBM induced by BDB-aggregated IgE. See legend Fig. 1 for incubation conditions. Supemates from cells incubated with aggregated IgE were tested for histamine by bioassay using guinea pig ileum in the absence or presence of pyrilamine.

64 PERRERI, BECK, AND SPIEGELBERG

FIG. 5. Effects of cytochalasin B (CB) and dihydrocytochalasin B (DHCB) on @-glucuronidase release induced by 250 &ml IgG immune complexes or latex particles. Cells (2 X ld esterase’) were pretreated with CB (5 &ml) for 5 min at room temperature before incubation with insoluble IgG immune complexes or latex particles. Values represent mean f SD (n = 4).

Inhibition of &G-Induced P-Glucuronidase Release by Monomeric IgG

PBM were incubated with monomeric K,IgGl (Ma), ~,1gG2 (Lu), ~,1gG3 (Fi), ~,1gG4 (Ho) or pentameric K&M (An) prior to challenge with BDB-aggregated ~,1gGl (Ma) to determine whether receptor occupancy by unaggregated immunoglobulins could block the effects of aggregated IgG. Unaggregated myeloma proteins of all classes and subclasses did not induce significant &lucuronidase release (l-2%). Monomeric K,IgGl (Ma) inhibited enzyme release induced by 10 @g/ml BDB-aggregated rc,IgGl (Ma) in a dose-dependent manner (Fig. 6). Approximately 20, 50, and 60% inhibition was achieved with 0.25, 0.50, 1, and 2 mg/ml, respectively, of monomeric K,IgG (Ma). Preincubation with ~,1gG3 (Fi) and ~,1gG4 (Ho) inhibited fl-glucuronidase in a manner similar to that of K,IgGl (Ma). In contrast, K,IgG2 (Lu) was a less effective inhibitor of enzyme release compared to the other IgG subclasses. Approximately four times more K,IgG2 (Lu) was required for the same inhibition of enzyme release achieved by the other IgG subclasses. Control experiments using K&M (An), ~,1gAl (As), and

F’lG. 6. Inhibition by monomeric IgG, IgA, IgE, and pentameric IgM of fl-ghrcuronidase release from PBM induced with BDBaggregated K&G (Ma). Cells (2 X 106 esters&) were preincubated with monomeric IgG of each subclass, IgM, monomeric IgA, and IgE prior to incubation for 30 min at 37°C with aggregated w&G1 (Ma). Values represent mean + SD (n = 4).

&GLUCURONIDASE RELEASE FROM MONOCYTES 65

X,IgE (PS) were done to ensure that the inhibitory effects of monomeric IgG were not due to nonspecific protein effects. Only slight inhibition, at high doses, was observed by preincubation with IgM indicating that the effects of monomeric IgG were specific.

DISCUSSION

Incubation of monocytes with aggregated normal IgG and IgG myeloma proteins of all four subclasses induced release of the lysosomal enzyme &$.icuronidase from these cells. In contrast, the fl-glucuronidase release induced by aggregated myeloma proteins of the Ig classes, IgA, IgM, IgE, and IgD, was not statistically significant. In agreement with previous reports (3, 2 l), the release of fi-glucuronidase was mediated by specific Fc receptors. In addition, release was independent of phagocytosis and was inhibited by monomeric IgG but not monomeric myeloma proteins of classes other than IgG. Most likely, IgG of all four subclasses reacted with the same Fc receptor, because the release of &glucuronidase induced with IgGl was inhibited with myeloma proteins of all four subclasses. IgG2 was the least effective inhibitor, which is in agree- ment with the relatively low affinity of monomeric IgG2 to FcyR (11). The release of /Lglucuronidase from monocytes of normal subjects parallels the ability of these cells to bind radiolabeled Ig (11) and form rosettes with erythrocytes coated with Ig of different classes (13). It appears from these studies that normal monocytes express predominantly IgG Fc receptors.

Fanger et al. (7) reported that 27 to 40% of normal monocytes form specific rosettes with erythrocytes sensitized with dimeric rabbit secretory IgA antibodies. In contrast, monomeric IgA antibodies were ineffective in sensitizing erythrocytes for detecting IgA rosetting monocytes. This suggests that the proposed IgA Fc receptors on monocytes are either of low affinity or low in number. The disparate results obtained with ~,1gAl

(As) and h&Al (Pu) support the presence of low-affinity Fc receptors for IgA on normal monocytes. These two myeloma proteins have previously been shown to differ in their ability to bind to monocytes (11). In addition, striking differences in the number of FccuR+ monocytes were obtained depending on the myeloma protein used (unpublished observations). Thus, it appears that the IgA Fc receptors on monocytes are difficult to demonstrate by rosette assays and by a functional assay such as Fc receptor mediated release of lysosomal enzymes.

We previously described a lo-20% subpopulation of normal monocytes that formed rosettes with&$ coated erythrocytes (13). Therefore, the aggregated myeloma proteins may have induced @-glucuronidase release from the FctR+ subpopulations. However, because of assay sensitivity and the variation of nonspecifically released &glucuron- idase, the release of enzyme from the 10 to 20% FccR+ monocyte subpopulations could not be reliably detected. Joseph et~ul. showed that IgE induced &lucuronidase release from pulmonary macrophages of allergic patients, indicating that under certain conditions IgE can induce the enzyme release from the monocyte related macrophages (1). Although IgE did not induce &lucuronidase from normal peripheral blood monocytes, it is conceivable that cells isolated from different anatomical locations, such as mucosal macrophages, could express large numbers of FCC receptors that induce measurable amounts of /3glucuronidase release. Similarly, pathological conditions such as allergic disease, might result in preferential expression of Fc receptors for a specific isotype. This was shown in severely atopic individuals who have increased percentages of peripheral blood monocytes expressing Fc receptors specific for IgE

66 FERRERI, BECK, AND SPIEGELBERG

compared to nonatopic individuals (22). The function of these receptors is currently under investigation.

The biological activity of IgA and IgE aggregates was assessed using human neu- trophils and guinea pig ileum, respectively. The IgA aggregates were generally as ef- fective as IgG aggregates in inducing ,&glucuronidase release from neutrophils, which have previously been shown to bind radiolabeled IgA (11) and express functional FccuR (4). Similarly, the BDB-aggregated IgE myeloma protein induced histamine release from basophils present in the mononuclear cell preparations. Therefore, the FccuR+ or FccR+ subpopulations that are present in normal monocyte preparations were either too small or did not express sufficient numbers of FcR to functionally interact with aggregated K&AI (As) and IgE.

Phagocytosis of the IgG aggregates was not required for the release of P-glucuronidase. Preincubation of the monocytes with cytochalasin B or dihydrocytochalasin B, potent inhibitors of phagocytosis (12), did not inhibit the ,!3glucuronidase release induced by aggregated IgG; on the contrary, it resulted in a twofold increase. As expected, cyto- chalasin B decreased &lucuronidase release induced by phagocytosis of latex particles. The increase of /3glucuronidase release resulting from incubation with cytochalasin B is not understood; however, it has also been observed by others (3,23,24). Perhaps the cytochalasin B-induced alteration of the cytoskeleton inhibits a normally occurring down-regulatory mechanism of Fc receptor-mediated lysosomal enzyme release. In- teraction of Fc receptors with IgG aggregates may result in internalization or shedding of FcR, thereby preventing further FcR interactions. Cytochalasin B may inhibit this FcR turnover, allowing a more prolonged Fc receptor-IgG interaction ultimately re- sulting in greater &+ucuronidase release.

The monocyte was the cell type responsible for the /3-glucuronidase release from PBM preparations containing monocytes and lymphocytes, a finding which was also made by Keeling and Henson (3). Lymphocyte preparations that contained less than 1% monocytes did not release significant amounts of enzyme, although the lymphocytes contained approximately 25 to 40% of the total /Qlucuronidase present in PBM prep- arations. Either the FcyR+ lymphocyte subpopulation was too small or, more likely, stimulation of FcrR on FcrR+ B, T, and null cells did not induce P-glucuronidase release. Because the lymphocytes contained &$ucuronidase but did not release it, the actual percentage of enzyme release from monocytes in the PBM was higher than the reported lo- 12% because the enzyme content of the lymphocytes was included in the 100% value. Therefore, it is estimated that on a per cell basis, monocytes contained approximately three to five times as much enzyme compared to lymphocytes. Ac- cordingly, approximately 25-30% of the enzyme content of the monocyte was released after challenge of TCD/PBM with aggregated IgG.

Monocytes play an important role in inflammatory reactions and in killing of mi- crobes and tumor cells (25, 26). In part, these functions are initiated by interaction with immune complexes and are effected by soluble mediators (24). The fact that only aggregated IgG induced statistically significant &lucuronidase release from monocytes suggests that of the five major antibody classes, IgG antibodies are the predominant class that induces lysosomal enzyme release from normal peripheral blood monocytes. However, as mentioned above, Fc receptors to other classes may be expressed on monocytes under pathological conditions such as the IgE receptors in allergic disease (22). Whether these Fc receptors induce the release of the same or different mediators as IgG receptors remains to be investigated.

/3-GLUCURONIDASE RELEASE FROM MONOCYTES 67

ACKNOWLEDGMENTS

We thank Dr. T. E. Hugh and M. S. Kawahara for assistance with the histamine assay and Mrs. Margaret Stone for preparing the manuscript.

REFERENCES

1. Joseph, M., Tonnel, A. B., Torpies, G., Capron, A., Amoux, B., and Benveniste, J., J. Clin. Invest. 71, 221, 1983.

2. Joseph, M., Tonnel, A. B., Capron, A., and Voisin, C., Clin. Exp. Zmmunol. 40,416, 1980. 3. Keeling, P. J., and Henson, P. M., J. Zmmunol. 128, 563, 1982. 4. Henson, P. M., Johnson, H. B., and Spiegelberg, H. L., J. Zmmunol. 109, 1182, 1972. 5. Lucisano, Y. M., and Mantovani, B., J. Zmmunol. 132,2015, 1984. 6. Khalife, J., Capron, M., Grzych, J. M., Bazin, H., and Capron, A., J. Zmmunol. 134, 1968, 1985. I. Fanger, M. W., Shen, L., Pugh, J., and Bemier, G. M., Proc. Natl. Acad. Sci. USA 77, 3640, 1980. 8. Haegert, D. G., Clin. Exp. Zmmunol. 35, 484, 1979. 9. Reynolds, H. Y., Kazmierowski, J. A., and Newball, H. H., J. Clin. Invest. 56, 376, 1975.

10. Spiegelberg, H. L., In “Advances in Immunopathology”, p. 123, Elsevier North Holland, Inc., 1981. 11. Lawrence, D. A., Weigle, W. O., and Spiegelberg, H. L., J. Clin. Invest. 55, 368, 1975. 12. Axline, S. G., and Reaven, E. P., J. Cell Biol. 62, 647, 1974. 13. Melewicx, F. M., and Spiegelberg, H. L., J. Zmmunol. 125, 1026, 1980. 14. Perlmann, H., Perlmann, P., Pape, G. R., and Halladen, G., &and. J. Zmmunol. 57(Suppl. 5), 1976. 15. Henson, P. M., J. Zmmunol. 107, 1535, 1971. 16. Tucker, S. B., Pierre, R. V., and Jordan, R. E., J. Zmmunol. Methods 14, 261, 1917.

17. Spiegelberg, H. L., and Dainer, P. M., Clin. Exp. Zmmunol. 35, 286, 1979. 18. Ishizaka, T., Ishizaka, K., Salmon, S., and Fudenberg, H., J. Zmmunol. 99,82, 1967. 19. Talalay, P., Fishmann, W. H., and Huggins, C., J. Biol. Chem. 166,757, 1946. 20. Wroblewski, F., and LaDue, J. S., Proc. Sot. Exp. Biol. Med. 90, 210, 1955. 21. Musson, R. A,, Shafran, H., and Henson, P. M., J. Reticuloendothel. Sot. 28, 249, 1980. 22. Melewicz, F. M., Zeiger, R. S., Mellon, M. H., O’Connor, R. D., and Spiegelberg, H. L., J. Zmmunol.

126, 1592, 1981. 23. Rothlein, R., and Kim, Y. B., J. Zmmunol. 129, 1859, 1982. 24. Temple, A., Loewi, G., Davies, P., and Howard, A., Immunology 24,655, 1973. 25. Bast, R. C., Cleveland, R. P., Littman, B. H., Zbar, B., and Rapp, H. J., Cell. Zmmunoi. 10,248, 1974. 26. Nathan, C., and Cohn, Z., J. Exp. Med. 152, 198, 1980. 21. Takemura, R., and Werb, Z., Amer. J. Physiol. 246, Cl, 1984.