DR framework antigens and 63D3 antigens on human peripheral blood monocytes and alveolar macrophages

CLINICAL

IMMUNOLOGY

AND

IMMUNOPATHOLOGY

29, 119-128 (1983)

DR Framework Antigens and 6303 Antigens on Human Peripheral Blood Monocytes and Alveolar Macrophages MARYELLENMOORE,MILTON

D. ROSSMAN,ALAN STEVEN D. DOUGLASS

D. SCHREIBER,AND

Departments of Microbiology, Medicine, and Pediatrics, Universio of Pennsylvania School of Medicine, Division of Allergy-Immunology-Pulmonology, The Children’s Hospital of Philadelphia, Joseph Stokes, Jr. Research Institute, and HematologylOncology Section, Hospital of the University of Pennsylvania, Philadelphia, Pennsylvania 19104

The expression of surface antigens on human peripheral blood monocytes and alveolar macrophages was compared using the monoclonal antibodies 63D3, which react with monocytes, and OKIa, which reacts with DR framework antigens. Fluorescence-activated cell-sorter analysis revealed that 71.5 +- 4.6% of peripheral blood monocytes reacted with 63D3 and showed a uniform distribution of binding. Alveolar macrophages also reacted with 63D3 and displayed uniform binding. Furthermore, the relative fluorescence intensity was very similar to that of monocytes. These data suggest that 63D3 antigen expression is similar on peripheral blood monocytes and alveolar macrophages and may be a stable marker in the differentiation of monocytes to macrophages. Analysis showed that 45.8 f 4.9% of peripheral blood monocytes reacted with OKIa and showed a nonuniform distribution of binding, while 74 Z? 8.4% of alveolar macrophages reacted with OKIa and exhibited uniform binding. The relative fluorescence intensity of alveolar macrophages which were reacted with OKIa was significantly greater than that of blood monocytes (P < 0.001). Size analysis suggested that alveolar macrophages express approximately five times more DR antigens per unit surface area than do peripheral blood monocytes. The expression of DR antigens on the alveolar macrophage surface suggests an important role for macrophages in antigen presentation in the lung.

INTRODUCTION

Characterization of the differences between the circulating blood monocyte and the tissue macrophage has been hampered by the lack of a definitive means of identifying mononuclear phagocytes. Monoclonal antibodies (Ab)2 are valuable probes for the identification of these cells and the possible delineation of stages in mononuclear phagocyte differentiation. Mononuclear phagocyte surface antigens (Ag) have been identified with monoclonal Abs. Ml/70 Ab reacts with the MAC-I Ag expressed on mouse macrophages, granulocytes, and peritoneal macrophages and it cross-reacts with human monocytes, granulocytes, and null cells (1). In addition, surface Ags of human monocytes have been defined by monoclonal Abs MO-I-MO-~ (2), OKMl (3), and 63D3 (4). Ml/70 expression is reported to increase during monocyte maturation (l), while MO-1 and MO-~ may represent ’ To whom correspondence should be addressed: Division of Allergy-Immunology-Pulmonology, The Children’s Hospital of Philadelphia, Philadelphia, Penn. 19104. 2 Abbreviations used: Ab, Antibody; Ag, antigen; PBM, peripheral blood monocytes; AM, alveolar macrophages; FGAM, fluoresceinated goat anti-mouse antibody; PBS/BSA/NaN,, phosphate-buffered saline with 2% bovine serum albumin and 0.02% sodium azide. 119 0090-1229183 $1.50 Copyri&l All Mhts

0 1983 by Academic Press, Inc. of reproduction in any form reserved

120

MOORE

ET AL.

distinct stages of late monocyte-granulocyte maturation (5). The expression of 63D3 on different human mononuclear phagocyte populations has not been studied, probably due to the difficulty in obtaining macrophages from humans. In this paper we demonstrate that 63D3 Ag is expressed in similar distribution on human peripheral blood monocytes (PBM) and alveolar macrophages (AM). Assessment of Ia surface Ags on mononuclear phagocytes may further define differences between these cells and aid in understanding the interactions of mononuclear phagocytes with other cells. Different macrophage populations in the mouse bear different amounts of Ia (6). Most peritoneal exudate cells are Iat while most spleen cells are Ia-. In humans, values ranging from 30 to 100% of mononuclear phagocytes expressing DR Ags have been reported (7-9). Quantitative differences in DR expressed per cell have been observed. PBM and peritoneal cells have widely varying amounts of DR Ags per cell (9), while AM that differed in source and state of activation showed little heterogeneity of DR Ag expression (10). We compared the expression of surface DR Ags on human PBM and AM. Forty-six percent PBM and 74% AM are DR+ and AM express approximately five times more DR Ags per unit surface area than PBM. MATERIALS

AND METHODS

Preparation of cells. Mononuclear cells were obtained from Ficoll-Hypaque density-gradient centrifugation of peripheral blood from 16 healthy adult donors and suspended in Medium 199 (Grand Island Biological Co., Grand Island, N.Y.) with 20% heat-inactivated horse serum (GIBCO). PBM were isolated by adherence to microexudate-coated flasks as reported previously (11). Briefly, the mononuclear cell suspension was incubated for 45 min at 37°C in tissue-culture flasks (Corning Glassware, Corning, N.Y.) in which baby hamster kidney cells had been grown to confluence and subsequently removed. The nonadherent cells were decanted, the flask was rinsed twice, and the cells were then incubated with 5 mM EDTA for 15 min at 37°C. The adherent cells were poured from the flask, centrifuged, and resuspended in minimal essential medium without phenol red (GIBCO) with 5% heat-inactivated fetal calf serum (Rehatuin, Armour Pharmaceutical, Kansas City, Kans.). The resulting cell preparation was %-98% PB-M as shown by nonspecific esterase staining, Fc receptors, morphology, and latex phagocytosis. The alveolar cells were obtained by segmental bronchopulmonary lavage of nine normal nonsmoking donors as described previously (12). For these studies between 300 and 500 ml of lavage fluid was introduced into the lung in 50-ml aliquots. Of the lavage fluid, 72 & 3.7% was recovered, and 31.9 + 2.8 x lo6 cells were recovered in this fluid. Viability by trypan blue exclusion was 83 it 3.3%. Cytocentrifuge preparations were stained with Diff-Quik (Harleco, Philadelphia, Pa.) and nonspecific esterase. The mean differential count was macrophages, 89.6 + 1.9%; lymphocytes, 9.3 + 1.7%; neutrophils, 0.9 ? 0.27%; and eosinophils, 0.2 ? 0.07%. Zrnmunofluorescence. 63D3 (Bethesda Research Laboratories, Inc., Gaithersburg, Md.) is a mouse monoclonal antibody that recognizes human monocytes and does not react with human peripheral blood T or B lymphocytes. It has weak reactivity with human granulocytes as well as platelets (4). OKIa (Ortho Diag-

SURFACE

ANTIGENS

ON

MONOCYTES

AND

ALVEOLAR

MACROPHAGES

121

nostic Systems, Raritan, N.J.) is a mouse monoclonal anti-human DR framework antibody that precipitates two major bands of 34,000 and 29,000 Da from labeled human B-lymphoblastoid lines. OKIa identifies B lymphocytes, activated T lymphocytes, and some monocytes (13). Fluorescein isothiocyanate-conjugated goat anti-mouse IgG (FGAM) was purchased from Tago, Inc. (Burlingame, Calif.). All antibodies were diluted in phosphate-buffered saline with 2% bovine serum albumin (Sigma Chemical Co., St. Louis, MO.) and 0.02% sodium azide (PBS/BSA/ NaN3). The concentrations of the antibodies used were determined in preliminary experiments to be saturating concentrations and were 63D3, 1.25 &ml; OKIa, 5 cl,g/ml; and FGAM, 42.5 l&ml. PBM and AM were aliquoted into 12 x 75mm polypropylene tubes (Falcon Plastics, Cockeysville, Md.) at a concentration of 1.0 x lo6 cells in 0.1 ml. The cells were incubated in 0.1 ml of the first antibody for 15-20 min, washed twice, incubated in 0.1 ml of the fluorescent antibody for 15-20 min, washed twice, and resuspended to a final volume of 0.5 ml. The incubations and centrifugations were carried out at 4°C and PBS/BSA/NaN3 was used for washing. The cells were analyzed by flow cytometry on a FACS IV cell sorter (BectonDickinson, Gaithersburg, Md.) with a linear amplifier. Controls included cells incubated in PBS/BSA/NaN3 and then with FGAM to monitor nonspecific staining with the fluoresceinated antibody. PBM and AM from several different donors were incubated with mouse antibodies of the same subclasses as 63D3 and OKIa, IgGI, and IgGZt,, respectively, but which did not react specifically with these cells. The average percentage reactivity of PBM and AM with antibodies of irrelevant specificity was 4% or less. Occasionally, cells were fixed in 2.5% paraformaldehyde after the second incubation and were kept at 4°C until they were analyzed. Preliminary studies demonstrated that paraformaldehyde fixation did not alter analysis by flow cytometry. Cell-sorter analysis. Fluorescence-labeled cells generate a signal which is linearly related to the number of fluorescein molecules present on the surface of a cell (14). When a saturating concentration of fluorescent antibody is used, the fluorescent signal is proportional to the number of binding sites for that cell. This signal then reflects the relative quantitative expression of antigens detected by the antibody. To condense the data obtained for a large number of individual cells, we calculated a mean fluorescence intensity per cell in arbitrary units averaged over the cell population under study. This calculation, developed by Smith and Ault (9), is 256 c F=l

m=N’

nfF

[II

with m as the mean intensity of fluorescence of a population of cells stained with a given antibody, N as the total number of cells analyzed (usually 20,000), F as the fluorescent channel number (from 1 to 256), and rrf as the number of cells with fluorescence equal to F. In each experiment, cells were stained for both the specific antigen in question and for nonspecific (control) uptake of the fluoresceinated antibody, and a relative fluorescence intensity index (R), which is pro-

122

MOORE

ET

AL.

portional to the amount of surface antigen per cell averaged over the group of cells, was calculated as

I21 with m, as the mean fluorescence intensity of the control-stained cells, m, as the mean fluorescence intensity of the cells stained for the antigen in question (experimental), and g as the fluorescence gain. The percentage reactivity of a population of cells in the experimental cell population was calculated as 256 c

% positive

=

256 9

c

9

F= 2(Fcon)

F=XF,,,)

N experimental

N control

x 100,

L31

_

with F as the fluorescence channel number, nf as the number of cells in the cell population with fluorescence equal to F, N as the total number of cells analyzed, and Fcon as the fluorescence channel number equal to the mean fluorescence intensity of the control population. The forward scatter of individual cells is proportional to cell radius (15). Thus, similar to fluorescence intensity, a mean forward scatter intensity was calculated as

with u as the mean forward scatter intensity, S as the forward scatter channel, n, as the number of cells in a given scatter channel S, and N as the total number of cells. For the analysis of purified monocytes, only those cells with forward-angle vs right-angle scatter distribution typical of monocytes (16) were analyzed. For alveolar macrophages, the cells with right-angle scatter greater than that of lymphocytes were analyzed. Sorting for this population of large cells results-in 98% macrophages by nonspecific esterase stain and by Wright’s stain. P values were calculated using Student’s independent c test. RESULTS

The pattern of binding of human PBM that were reacted with 63D3 (PBM 63D3) was different from the pattern of binding of human PBM that were reacted with OKIa (OKIa PBM). 63D3 PBM showed a uniform distribution of binding, whereas OKIa PBM showed a nonuniform distribution of binding (Fig. 1). A significantly greater percentage of PBM reacted with 63D3 than with OKIa (Table 1). Also, the relative intensity of lkorescence (R) of 63D3 PRM was significantly greater than that of OKIa PBM. In contrast to PRM, human AM were observed to have a uniform distribution of binding both to 63D3 and OKIa (Pigs. 2 and 3). The intense autofluorescence exhibited by AM resulted in a significant overlap be-

CHANNEL

NUMBER

63D3+FGM

+

FIG. 1. Binding of peripheral blood monocytes with 63D3 and OKIa monoclonal antibodies. Human PBM were incubated with buffer + FGAM (A), 63D3 + FGAM, (0). or OKIa + FGAM (a) and analyzed on a FACS IV cell sorter. Fluorescence gain, FG = 1.

FLDORBSCENCE

OKIe

+

+ FGAM

FcAN

-e

124

MOORE REACTIVITY

OF HUMAN

ET AL.

TABLE 1 PBM AND AM

WITH

63D3

AND

OKIa

PBM

63D3

Rb

%

R

71.5 iz 4.6c

17.1 2 1.7 (21) 9.4 2 1.3 (13) 0.003

NDd

15.1 k 3.1 (9) 262.1 ” 71.1’ (4) 0.001

45.8 it 4.9

(12) P(63D3 vs OKIa)

---

%”

(16) OKIa

AM

0.001

74.0 IL 8.4 (4) -

0 Percentage of the cells that react specifically with the monoclonal antibody 63D3 or OKIa. b R is the relative fluorescence intensity of the cell population analyzed by the cell sorter: R = mean fluorescence intensity (experimental - control)/gain. c Values are expressed as the mean t SE with the number of samples studied in parentheses. d ND, not done (see text). e P < 0.001 compared to OKIa PBM. All P values are calculated using Student’s independent r test.

tween AM control and 63D3-binding curves (Fig. 2). Because of this overlap, direct calculations of the percentage of AM binding 63D3 were not possibk. However, R values could be used to compare AM reactivity with 63D3 and OKla. OKIa-labeled AM (OKIa AM) had a relative intensity of fluorescence that was significantly greater than 63D3-labeled AM (63D3 AM) (able 1). A cqmparison between PBM and AM antibody reactivity showed that the R values for 63D3 PBM and 63D3 AM were very similar. A significantly greater percentage of AM reacted with OKIa than PBM, and AM exhibited a 30-fold greater intensity of fluorescence with OKIa than did PBM (Table I). Since the forward scatter is proportional to the cell radius, the relative amount of monoclonal Ab bound to PBM or AM can be related to their surface area. The average of PBM mean scatter intensities (a) from 12 donors was 149.2 f 6.6. The average for two AM donors was significantly greater (345.0, P < 0.001). Since the mean amount of surface Ag per cell is proportional to the R value and the surface area is proportional to the square of u (or radius squared), then the mean amount of antibody bound per unit surface area is proportional to Rh2. For OKIa AM this value is five times greater than that for OKIa PBM. Thus, the difference in PBM and.AM binding to OKIa is not simply due to cell size, but to a greater amount of Ab binding per unit surface area by AM. DISCUSSION

A monoclonal antibody, 63D3, which was raised against human PBM has been reported to react with a 200,000-Da single-chain polypeptide present on most PBM and on a small proportion of granulocytes (4). We-have found that an average of 72% PBM show reactivity with 63D3 (Table 1). The intense autofluorescence of AM does not allow for a good separation of the control curve from the 63D3binding curve (Fig. 2). The binding profiles show, however, that the intensity of fluorescence of the 63D3 AM population is greater than that of corresponding

SURFACE

ANTIGENS

ON

ttt

MONOCYTES

AND

ALVEOLAR

MACROPHAGES

125

150

OK18 + FGAM

+

FLUORESCENCE CHANNEL NLMBER

100

FGM

--c

200

250

Frc. 3. Binding of human AM with OKLa. Human AM were incubated with buffer ( + ); buffer + FGAM (A); or OKla + FGAM (m) and analyzed on a FACS IV cell sorter. Fluorescence gain. FG = 4.

50

AUTOFLUORESCENCE

+

SURFACE

ANTIGENS

ON

MONOCYTES

AND

ALVEOLAR

MACROPHAGES

127

portions of the control curve, indicating that AM are binding 63D3. R value analysis of 63D3 PBM and 63D3 AM (Table 1) indicated no difference between the relative intensities of fluorescence of these populations with 63D3. Thus, PBM and AM appear to express the same amount of 63D3 Ag. Differentiation of the circulating blood monocyte to AM is not accompanied by an increase in expression of 63D3 Ag despite a substantial increase in cell size. Furthermore, 63D3 shows uniform distribution of binding and thus similar expression of 63D3 Ag among PBM and AM cell populations. This suggests that 63D3 may be a stable marker in the differentiation of monocytes to macrophages and may be useful in assessing the modulation of other surface markers. Previous studies report from 30 to 100% human PBM as DR+ (7-9). Utilizing the monoclonal Ab OKIa, our studies indicate that 46% PBM express surface DR Ags and 74% AM are DR + , which is similar to the 72-92% HLA-DR + reported for AM (17). A major difference was found in the amount of DR Ag expressed on PBM and AM. The relative fluorescence intensity (R) values for PBM and AM were significantly different (P < 0.001) (Table 1). Since R is proportional to the mean amount of surface Ag per cell, AM expressed more DR Ags on their surface than PBM. This was not due to differences in cell size between PBM and AM, since AM have approximately five times more DR Ags per unit surface area than PBM. There are also differences in the cell-sorter binding profiles of OKIa with PBM and AM. PBM show a nonuniform distribution of binding (Fig. 1); most cells show little fluorescence (hence, having few OKIabinding sites), a small percentage of cells show intense fluorescence (having many binding sites), and some cells show fluorescence between these extremes. Smith and Ault also observed that PBM and peritoneal washing cells express DR in widely varying amounts per cell (9). In contrast to PBM, the AM population shows a uniform distribution of binding with OKIa (i.e., individual AM cells appear to express the same amount of DR per unit surface area) (Fig. 3). The evidence that the differences in fluorescence intensity between individual cells in the AM population are due to differences in cell size is supported by analysis of forward scatter vs intensity of fluorescence on the cell sorter. This finding is consistent with the observation that despite differences in source and state of activation of AM, single-cell analysis revealed no heterogeneity of DR Ag expression (10). Expression of Ia Ags on monocytes and macrophages has been studied in mouse and human cell populations. Studies of Ia Ag expression on macrophages cultured from mouse bone marrow cells differ in their conclusions. One study reported that few macrophages from bone marrow cells cultured in lung-conditioned medium expressed Ia while growing or quiescent, but T-lymphokine-stimulated ceils synthesize and express Ia (18). The authors conclude that growth and differentiation stimuli may have opposing effects on Ia induction. Results of another study show mouse bone marrow cells cultured in lung-conditioned medium express Ia during growth and after reaching the stationary phase, whereas cells cultured in L-ceil-conditioned medium express less Ia during growth and show a decrease in Ia expression upon reaching the stationary phase (19). These authors concluded that lung cell-conditioned medium favored the appearance and persis-

128

MOORE

ET AL.

tence of Ia + macrophages without further maturation. Studies of DR expression on cultured human peripheral blood monocytes show that there was an increase in surface DR over the first 12 hr and a slight decline over the next 3 days (9). Increase in surface DR was independent of phagocytosis or immune stimuli and required metabolic energy and protein synthesis. Our data suggest that DR Ag expression is increased in macrophages as compared to monocytes and that this expression is homogeneous among the AM population. The fact that circulating blood monocytes have different amounts of DR Ags on their surfaces may be attributable to different stages of maturation. Since the presence of DR Ags on macrophages has been correlated with Ag-presenting ability, the alveolar macrophage may play an important role in the local immunity of the lung by acting as an antigen-presenting cell. ACKNOWLEDGMENTS We thank Dr. Alan Piccard for technical assistance with the cell sorter and computer. We are also indebted to Carolyn S. Cody for her excellent technical help and valuable discussions. This work was supported in part by the National Institutes of Health, Department of Human Services Grant 1 RO1 HL-27068. 1 PO1 NS17752, CA-15236, and the Thrasher Research Fund.

REFERENCES 1. Springer, T., Galfre, G., Secher, D. S., and Milstein, C., Eur. J. Immunol. 9, 301, 1979. 2. Todd, R. F., and Schlossman, S. F., Blood 59, 775, 1982. 3. Breard, J., Reinherz, E. I., Kung, P. C., Goldstein, G., and Schlossman, S. F., J. Immunol. 124, 1943, 1980. 4. Ugolini, V., Nunez, G., Smith, G. R., Stastny, P., and Capra, J. D., Proc. Nur. Acad. Sci. UsA 77, 6764, 1980. 5. Todd, R. F., Nadler, L. M., and Schlossman, S. F.. J. Immunol. 126, 1435, 1981. 6. Cowing, C., Schwartz, B. D., and Dickler, H. B., J. Immunol. 120, 378, 1978. 7. Albrechtson, D., Stand. J. Immunol. 6, 907, 1977. 8. Ng, A. K., Zhang, Y. H., Russo, C., and Ferrone, S., Cell Immunol. 66, 78, 1982. 9. Smith, B. R., and Auit, K. A., J. Immunol. 127, 2020, 1981. 10. Mason, R., Austyn, J., Brodsky, F., and Gordon, S., Amer. Rev. Respir. Dis. 125, 586, 1982. 11. Ackerman, S. K., and Douglas, S. D., J. Immunol. 120, 1372, 1978. 12. Rossman, M. D., Dauber, J. H., Cardillo, M. E., and R. P. Daniele, Amer. Rev. Respir. Dis. 125, 366, 1982. 13. Reinhertz, E. L., Kung, P. C., Pesando. J. M., Ritz, J., Goldstein, G., and Schlossman, S. F.. J. Exp. Med. 150, 1472, 1979. 14. Loken, M. R., and Herzenberg, L. A., Ann. N. Y. Acad. Sci. 254, 163, 1975. 15. Mullaney, P. F., and Dean, P. N., Biophys. J. 10, 764, 1970. 16. Loken, M. R., Sweet, R. G., and Herzenberg, L. A., J. Histochem. Cytochem. 24, 284. 1976. 17. Toews, G. B., Vial, W., Nunez, G., Stastny, I?, and Lipscomb, M. F., Amer. Rev. Resp. Dis. (Abstr.) 125, 84, 1982. 18. Lu, C. Y., and Unanue, E. R., Infect. Immun. 36, 169, 1982. 19. Lee, K. C.. and Wong, M., J. Immunol. 125, 86, 1980. Received January 12, 1983; accepted February 23, 1983