- Email: [email protected]
Quantitative immunocytochemical characterization of mononuclear phagocytes
CELLULAR
IMMUNOLOGY
105,374-385
Quantitative
(1987)
lmmunocytochemical Characterization of Mononuclear Phagocytes
I. Monoblasts,
Promonocytes, Monocytes, and Alveolar Macrophages’
and Peritoneal
PETERH. NIBBERING,~ETER C.J.LEIJH, ANDRALPHVANFURTH Department of Infectious Diseases, University Hospital, P.O. Box 9600, 2300 RC Leiden. The Netherlands Received November 3, 1986; acceptedNovember 18, 1986 The purpose of the present study was to compare the phenotype of tissue macrophages with that of their precursors in the bone marrow and blood. The phenotype was determined on the basis of the quantitative binding of monoclonal antibodies to cell-surface antigens (antigen F4/80, complement receptor III, Fc receptor II, Ia antigen, common leukocyte antigen, and Mac-2 and Mac-3 antigens) on individual mononuclear phagocytes. Monoclonal antibody binding to cells, detected by the biotin-avidin immunoperoxidase procedure, was quantitated by cytophotometric determination of the amount of enzyme reaction product on cells. The results of this quantitation are expressedasthe median of the specific absorbance per unit of cell-surface area (0.25 pm2) and per cell. Shortly after collection of the mononuclear phagocytes, binding of all monoclonal antibodies except those directed against the common leukocyte and Mac-2 antigens to peritoneal macrophages was enhanced compared with binding to blood monocytes; for alveolar macrophages we found reduced binding of monoclonal antibodies F4/80 and Ml/70 (complement receptor III) and enhanced binding of monoclonal antibodies with specificity for the common leukocyte antigen and Mac-2 and Mac-3 antigens. The results obtained with cultured mononuclear phagocytes show that during the development from monoblast to tissue macrophages, monoclonal antibody binding to the various types of mononuclear phagocyte, expressed per unit of cell-surface area, was not significantly altered except that of M3/38 (Mac2 antigen) to peritoneal macrophages and that of F4/80 and Ml/70 (complement receptor III) to alveolar macrophages. Expressed on a per cell basis, the results show an increase in the binding of all monoclonal antibodies except those directed against the Fc receptor II and Mac-3 antigen during the development from promonocytes to peritoneal macrophages; binding of most monoclonal antibodies to alveolar macrophages was considerably lower than that to blood monocytes. It is concluded that the expression of the various cell-surface antigens alters during mononuclear phagocyte differentitaion. The expression changed also during culture, although distinct patterns of alteration could not be distinguished. 0 1987 Academic press, 1~.
INTRODUCTION Mononuclear phagocytes constitute a cell line originating from the hemopoietic stem cell; the precursors develop in the bone marrow and are transported via the ’ This work was supported by the Netherlands Asthma Foundation.
0008-8749/87 $3.00 Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved
374
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
375
blood to the tissues, where they differentiate into tissue macrophages (1). Mononuclear phagocytes can be identified from a number of features such as characteristic morphology, staining for specific enzymes, adherence to surfaces, phagocytic and bactericidal ability, and such membrane characteristics as Fc and C receptors and Ia antigen (2). Because none of these features is specific for mononuclear phagocytes, monoclonal antibodies characterizing determinants on mononuclear phagocytes were produced (reviews in Refs. (3) and (4)). However, most of the anti-macrophage antibodies are not entirely specific for mononuclear phagocytes either: only the antibody designated F4/80 is restricted to cells belonging to the mononuclear phagocyte system of the mouse (5). During differentiation of bone marrow cells in culture, antigen F4/80 first appeared on nonadherent immature mononuclear phagocytes in 3day-old cultures (6). In these cultures virtually all adherent mononuclear phagocytes expressedantigen F4/80, complement receptor type III (CR III). and Fc receptor type II (FcR II). Despite these observations, quantitative information about the expression of cell-surface antigens by immature mononuclear phagocytes is lacking, and quantitative data on the occurrence of antigen F4/80, CR III. and FcR II on mature murine mononuclear phagocytes are limited (5,7-9). The present study was undertaken to compare the phenotypes of peritoneal and alveolar macrophages with those of their precursors in bone marrow cultures and blood. To define the phenotype of individual mononuclear phagocytes. we used monoclonal antibodies and a quantitative cytochemical method ( 10). MATERIALS AND METHODS Animals Specific pathogen-free (SPF) male Swiss mice (Central Institute for the Breeding of Laboratory Animals, TNO, Zeist. The Netherlands) weighing 20-25 g were used. Preparation and Culture qfkfononuclear Phagocytes Bone marrow’ phagocytes. Bone marrow cells were harvested by flushing one of the mouse femurs with Dulbecco modified Eagle’s medium (DMEM; Grand Island Biologicals, Grand Island, NY) and were cultured (for 7 days) by plating 5 X 1O4cells per Leighton tube with a flying coverslip as described elsewhere (11). The culture medium consisted of DMEM containing 20% heat-inactivated horse serum (Flow Laboratories, Irvine, UK) and 30% embryonic mouse fibroblast-conditioned medium as source of colony-stimulating factor- 1 (CSF- 1). Blood monocytes. Blood was collected by cardiac puncture, and erythrocytes and granulocytes were removed by Ficoll-Isopaque centrifugation (p = 1.077; Pharmacia Inc., Uppsala, Sweden) for 20 min at 440 g. The cells in the fraction rich in monocytes and lymphocytes were washed four times with phosphate-buffered saline (PBS; pH 7.2) containing 0.5 U/ml heparin. Cytocentrifuge preparations were made immediately after collection for the study of blood monocytes. For comparison of blood monocytes with bone marrow phagocytes, 1 X IO6cells of the monocyte-rich FicollIsopaque interphase were cultured at 37°C in Leighton tubes with flying coverslips; the culture medium was either Medium 199 (M 199; Microbiological Assoc., Walkersville, MD) containing 20% heat-inactivated newborn calf serum (NBCS; GIBCO
376
NIBBERING,
LEIJH,
AND
VAN
F’URTH
Europe, Paisly, UK), 2000 U/ml sodium penicillin G (Gist-Brocades, Delft, The Netherlands), and 50 pg/ml streptomycin (Gist-Brocades) or DMEM supplemented with horse serum (GIBCO Europe) and mouse embryonic fibroblast-conditioned medium. After incubation for 2 hr, the nonadherent cells were removed by three washes with PBS. The glass-adherent cells were cultured under the described conditions for 7 days. Tissue macrophages. Peritoneal macrophages were collected from the peritoneal cavity by lavage with 2 ml PBS containing 50 U/ml heparin, as described elsewhere ( 12). Alveolar macrophages were isolated by lavage with 15 ml PBS containing 0.6 mM EDTA, as described elsewhere ( 13). Cytocentrifuge preparations were made for the study of tissue macrophages immediately after collection. Peritoneal and alveolar macrophages (5 X 10’ cells/ml) were cultured as described for blood monocytes. Monoclonal Antibodies The rat anti-mouse monoclonal antibody (Mab)-producing hybrid cell lines 30.G. 12 (common leukocyte antigen) (14) Ml/70 (complement receptor III, CR III) (15) M3/38 (Mac-2 antigen) (16), M3/84 (Mac-3 antigen) (17) and M5/114 (Iab,d.q Ied%k) (18) were obtained from the American Type Culture Collection (Rockville, MD). The hybrid cell line producing 2.4.G.2 Mab (Fc receptor II, FcR II) ( 19) was a gift from Dr. J. Unkeless (Rockefeller University, New York, NY). The cells were cultured at 37°C under 5% CO2 in RPM1 1640 medium containing 20 mM Hepes buffer (Flow Laboratories, Ltd.), 2 mM L-glutamine (Microbiological Assoc.), 10% fetal bovine serum (Flow Laboratories, Ltd.), 1000 U/ml sodium penicillin G (GistBrocades), 20 yg/ml gentamycin (Gist-Brocades), and 5 X 10e5A4 2-mercaptoethanol. The culture supernatants were harvested every 3-4 days. The culture supernatant of the F4/80 Mab-producing hybrid cell line (5) was a gift from Dr. S. Gordon (Sir William Dunn School of Pathology, Oxford University, Oxford, UK). Quantitation of the Binding of Monoclonal Antibodies to Mononuclear Phagocytes The method used to quantitate the binding of monoclonal antibodies to individual mononuclear phagocytes and the size of these cells has been described elsewhere ( 10). In short, after fixation of the cells with 0.05% glutaraldehyde (Polysciences, Warrington, CN) in PBS at 4°C for 30 min, they were first incubated for 1 min with 0.01 N HCl (pH 2) to inactivate endogenous peroxidase and then for 10 min at room temperature with 20% normal goat serum to block nonspecific binding sites, and finally were overlayered with a saturating concentration of Mab for 30 min. Supematant of the plasmacytoma cell line P3X 163 A98 653 (P3) (20) served as control. After the removal of excess Mab, monoclonal antibody binding was detected with 2.5 yg/ml biotinylated rabbit anti-rat IgG (Vector Laboratories, Burlingame, CA) in 5% normal mouse serum and 5 pg/ml horseradish peroxidase-conjugated avidin-D (Vector Laboratories), the peroxidase being visualized by incubation for 60 min with 0.6 mg/ml 3’3’-diaminobenzidine-tetrahydrochloride (DAB, Merck, Darmstadt, FGR) in Tris-HCl buffer (pH 7.6) supplemented with 0.0 1%H202. For identification of mononuclear phagocytes on a morphological basis in heterogeneous preparations, the cells were counterstained with 10 pg/ml ethidiumbromide (2,7 diamino-lo-ethyl-9-phenyl-phenanthridiumbromide; Calbiochem-Behring, La
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
377
Jolla, CA) in distilled water for 1 min. The amount of peroxidase reaction product present on the mononuclear phagocytes was determined by measuring the absorbance at 466 nm with a microscope photometer (Zeiss, Oberkochen. FRG) interfaced to PDP 1 I/ 10 microcomputer (Digital Equipment Corporation, Maynard, MA). Simultaneous assessment of the integrated absorbance (A,,,) of the peroxidase reaction product and the ethidiumbromide fluorescence of the nucleus was performed with the microscope photometer provided with an epi-illuminator III RS (Zeiss). Scanning of the cells with photometric sampling at 0.5~brn intervals and computation of the mean .4,,, per 0.25~pm’ projected cell-surface area with the ARRAYSCAN program of the HIDACSYS package have been described elsewhere (2 1). Monoclonal antibody binding to cells in each population of mononuclear phagocytes was determined on 100 randomly selected cells in three to four experiments. The nonspecific binding of antibodies and the autooxidation of DAB were assessed in mononuclear phagocytes exposed to P3-plasmacytoma supernatant instead of Mab and otherwise treated as described above. The mean nonspecific A,,,‘s per 0.25pm’ cell surface were normally distributed. Cells expressing a certain cell-surface antigen were distinguished with 95% confidence limits from those lacking the antigen by setting the threshold A,,,, per 0.25~pm’ cell surface at the mean nonspecific A,,, per 0.25-pm’ cell surface plus twice the SD: i.e.. cells with an observed A,,, per 0.25~pm’ cell surface that was higher than the threshold A,,, were considered to express the cell antigen. The mean specific A,,, per 0.25~pm’ cell surface of the mononuclear phagocytes expressing a certain cell-surface antigen was defined as the mean observed A,,,, per 0.25~pm’ cell surface minus this threshold A,,,. The mean specific A,,,‘s per 0.35-pm’ cell surface were used to determine the phenotype of the mononuclear phagocytes. For comparison, the results were expressed on a per cell basis. The specific A,,, per cell was calculated from the observed A,,, per cell. the mean nonspecific A,,, per cell, and the SD, as described above. The sizes of individual mononuclear phagocytes were estimated with the microscope photometer by determining the projected cell-surface area. The results are expressed as the mean values of 200 randomly selected cells per population of mononuclear phagocytes. Statisticd
und~vis
Cell size distributions and distributions ofthe mean specific A,,,‘s per 0.25~pm’ cell surface of mononuclear phagocytes were analyzed for normality on the basis of tests of skewness and kurtosis (22). P < 0.0 1 marked departure from normality. The significance of differences in the cell size and mean specific Aint’s per 0.25pm’ cell surface of the various populations of mononuclear phagocytes was estimated by the Mann-Whitney U test (23). P < 0.0 1 was considered significant. RESULTS All results are expressed as the percentages of mononuclear phagocytes expressing the various cell-surface antigens and the median of the mean specific A,,t’~ per 0.25pm’ projected cell-surface area of 100 mononuclear phagocytes and their range. Frequency histograms showed that the cell-surface antigens were not normally distrib-
378
NIBBERING, LEIJH, AND VAN FURTH TABLE 1 Size of Mononuclear Phagocytes, as Determined by Cytophotometry” Mononuclear phagocytes
Monoblasts Promonocytes Macrophagelike cells Blood monocytes Peritoneal macrophages Alveolar macrophages
Shortly after collection *
After 7 days of culture’
Size
Range
Size
pm2
elm’
c(m2
ndd nd nd 146 222 161
nd nd nd 61-321 128-427 85-263
128 131 316 216 208 140
Range d 79-185 64-193 118-753 8 l-673 90-886 62-284
a The results are expressedas the median and range of the projected cell-surface area of 200 mononuclear phagocytes in three to five experiments. * The cells were sedimented on microscope slides shortly after collection. ‘Culture was performed for 7 days on glasscoverslips. d Not done.
uted on either freshly collected or cultured mononuclear phagocytes (results not shown). Phenotype of Bone Marrow Phagocytes
Mononuclear phagocytes in 7-day cultures of bone marrow cells were identified on the basis of morphological features and localization within individual mononuclear phagocyte colonies. The following types of mononuclear phagocyte were distinguished (11): monoblasts, i.e., small, round cells with a nuclear-to-cytoplasmic ratio higher than 1; promonocytes, i.e., small, slightly stretched cells with one pseudopod and a nuclear-to-cytoplasmic ratio of about 1; and macrophagelike cells, i.e., large elongated cells with two or more pseudopods or, sometimes, large round cells without a pseudopod and a nuclear-to-cytoplasmic ratio lower than 1. No difference in size between monoblasts and promonocytes were found, but the marcrophagelike cells were significantly larger (P < 0.001) (Table 1). Frequency histograms of the cell sizes showed that monoblasts (P > 0.1) and promonocytes (P > 0.05) were normally distributed whereas the macrophagelike cells were not (results not shown). The antigen F4/80, CR III (Mab M1/70), and Mac-2 (Mab M3/38) and Mac-3 (Mab M3/84) antigens were detected on virtually all monoblasts, promonocytes, and macrophagelike cells (Fig. 1). FcR II (Mab 2.4.G.2) was clearly expressed by almost all monoblasts and promonocytes but was considerably (P < 0.00 1) lessby the macrophagelike cells. Ia antigen (Mab M5/ 114) was weakly expressed by about two-thirds of the monoblasts and about 40% of the other bone marrow mononuclear phagocytes. Common leukocyte antigen (Mab 30.G. 12) was clearly expressed by virtually all monoblasts and considerably (P < 0.001) more strongly by promonocytes and macrophagelike cells (Fig. 1).
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
379
SPECFK AINT PER 0.25 pm2 CELL SURFACE
I
1
::
F4/80
i
2.4 G2
1
: I
/
t ;;
30 G 12
I
/
FIG. 1. Phenotypes of bone marrow mononuclear phagocytes. Bone marrow cells were cultured in the presence of CSF- I for 7 days on coverslips. Expression of cell-surface antigens by the mononuclear phagocytes was quantitated by photometric determination of the amount of peroxidase reaction product. The results are expressed as percentages of cells expressing the various cell-surface antigens and the medianspecific A,., per 0.25~pm’ projected cell-surface area and the range for 100 mononuclear phagocytes from three to four experiments.
Phenotype qf Blood A4onocyte.s Comparison of the cell sizes of freshly collected monocytes and monocytes cultured for 7 days showed that the latter cells were significantly larger (P < 0.00 1) (Table 1). The sizes of the cells in both groups were not normally distributed (results not shown). The binding of monoclonal antibodies to monocytes is shown in Fig. 2. After culture, the percentages of monocytes expressing antigen F4/80, CR III (Mab M1/70), and Mac-2 antigen (Mab M3/38) were lower than for recently collected monocytes, whereas the percentage ofblood monocytes expressing FcR II (Mab 2.4.G.2.) Ia antigen (Mab M5/ 114) and Mac-3 antigen (Mab M3/84) increased during culture (Fig. 2). The percentage of monocytes expressing common leukocyte antigen (Mab 30.G. 12) did not alter after culture. Expression of antigen F4/80 and CR III on blood monocytes was significantly (P < 0.001) higher and FcR II significantly (P < 0.001) less than that by cultured monocytes, and common leukocyte antigen was clearly and Ia, Mac-2, and Mac-3 antigens were weakly expressed by both groups.
I
:
:
: I
I
,
4
4
-I
,
I I
I
M3/84
U3/38
30.G.12
MS/114
FIG. 2. Phenotype of blood monocytes and peritoneal and alveolar macrophages shortly after collection and after 7 days of culture. The results are expressed as percentages of cells expressing a given cell-surface antigen and the median-specific Aimt per 0.2%pm’ cell-surface area and range for 100 mononuclear phagocytes in three to four experiments.
I:
I
t :
t
t
I :
I
t
w
w
I
2.C.G.2.
Ml/70
CULTURED MONONUCLEAR PHAGOCYTES SPECIFIC A/NT PER 0.25 pm2 CELL SURFACE
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
381
For investigation of the effect of CSF- 1 on the phenotype of cultured mononuclear phagocytes, blood monocytes were cultured for 7 days in the presence and absence of this factor. Comparison of the phenotypes of both groups of monocytes revealed no significant differences. For example, the specific mean Aint per 0.25~pm2 cell surface, which reflects the expression of antigen F4/80, amounted to 0.109 (range O.OOS0.33 1) for monocytes cultured with CSF-1 and that for monocytes without this factor to 0.105 (range 0.005-0.457). This result and those for the other cell-surface antigens (results not shown) showed no effect of CSF- 1 on the phenotype of blood monocytes.
Phenotype cfPeritoneu1 Macrophages No difference in size between cultured and noncultured peritoneal macrophages was observed (Table I). Frequency histograms showed that the sizes of the cells in both groups of peritoneal macrophages were not normally distributed (results not shown). The binding of monoclonal antibodies to peritoneal macrophages is shown in Fig. 2. The percentage of peritoneal macrophages expressing the various cell-surface antigens decreased during culture. Expression of antigen F4/80 was intense by both groups of cells and that of CR III (Mab M l/70) by noncultured peritoneal macrophages was strong, whereas that of cultured peritoneal macrophages was significantly (P < 0.00 1) weaker. FcR II (Mab 2.4.G.2.) and Ia antigen (Mab M5/ 114) were weakly expressed by both groups. Common leukocyte antigen (Mab 30.G. 12) was clearly expressed by noncultured and significantly (P < 0.00 1) higher by cultured peritoneal macrophages. Expression of Mac-2 antigen (Mab M3/38) by noncultured was weak and that by cultured peritoneal macrophages was significantly (P < 0.001) higher. The reverse was the case for the Mac-3 antigen (Mab M3/84) (P < 0.00 1). Control experiments showed that CSF- 1 does not affect the phenotype of cultured peritoneal macrophages (results not shown). No effect of either Ficoll-Isopaque density centrifugation or EDTA on the phenotype of peritoneal macrophages was seen, and the same holds for sedimentation of cells on the microscope slides and adherence of cells to glass during incubation for 1 hr (results not shown).
Comparison of the size of cultured and noncultured alveolar macrophages showed no significant difference (Table 1). Frequency histograms of the cell sizes in both groups showed that the distributions were not normal (results not shown). The binding of monoclonal antibodies to alveolar macrophages is shown in Fig. 2. After 7 days of culture, the percentage of alveolar macrophages expressing antigen F4/80. CR III (Mab M1/70), FcR II (Mab 2.4.G.2.). and Mac-3 antigen (Mab M3/ 84) dropped. The percentages of macrophages expressing Ia antigen (Mab M5/ 114). common leukocyte antigen (Mab 30.G. 12). and Mac-2 antigen (Mab M3/38) were not altered after culture. Antigen F4/80 was readily expressed by noncultured alveolar macrophages and considerably less so (P < 0.00 1) by cultured alveolar macrophages. CR III, FcR II, and Ia antigen were weakly expressed by both groups of cells. Expression of the common leukocyte, Mac-2, and Mac-3 antigens by alveolar macrophages was intense, and was considerably (P < 0.001) weaker for similar cells after culture.
382
NIBBERING, LEIJH, AND VAN FIJRTH SPECIFIC AINT PER CELL
1
I!
30.G.12
I I
I I
I!
I ::
I f
M313.9
FIG. 3. Binding of monoclonal antibodies to bone marrow mononuclear phagocytes. The results are expressed as percentages of cells expressing the various cell antigens and the median-specific Ai,, per cell and range for 100 mononuclear phagocytes in three to four experiments.
DISCUSSION The present results concerning cultured mononuclear phagocytes show that the expression of all cell-surface antigens except Mac-2 and Mac-3 antigens, does not alter during the development from monoblast/promonocyte to blood monocytes; upon differentiation of blood monocytes into peritoneal macrophages the expression of the F4/80 and Mac-2 antigens increases and differentiation of monocytes into alveolar macrophages is accompanied by a decreasein the expression of antigen F4/ 80, CR III, and FcR II. This conclusion is based on the results of the quantitation of the expression of cell-surface antigen per 0.25~pm2cell-surface area. The method we used to quantitate the expression of cell-surface antigens by cells is based on Mabs, peroxidase immunocytochemistry, and photometric determination of the amount of peroxidase reaction product. This method is reproducible, sensitive, and specific (10). The Aint of the peroxidase reaction product proved to be proportional to the amount of cell-surface antigen immobilized on nitrocellulose paper (24) (P. H. Nibbering et al., Submitted for publication). The possibility that the procedures used for the preparation ofthe cells, e.g., Ficoll-Isopaque, EDTA, or CSF1, and sedimentation and adherence to glass,have an effect on the expression of cell-
MEMBRANE ANTIGENS IN MONONUCLEAR
383
PHAGOCYTE DIFFERENTIATION
TABLE 2 Binding of Monoclonal Antibodies to Blood Monocytes and Peritoneal and Alveolar Macrophages Shortly after Collection and after 7 Days of Culture” Peritoneal
Blood monocytes Shortly after collection F4/80 Ml/70 2.4.G.2 MS/II4 3O.G 12 M3/3R M3184
92 85 I0 16 98 48 22
112(4460) 102(8-327) 7(3-?I) I’(:-160) 70(4-245) 28 (2-162) I, (l-43)
After 7 days of culture 93 110(2-418) XX 64(5-244) 50 79 (3-408) 34 46(11-181) 99 116(21-322) 44 74(13-213) 47 70(29-167)
Shonly after collection 100 99 8X 40 9X 40 IO0
196(29-567) 187(30--166) 45(1-1601 52(6-89) X3(9-158) l4(1-91) lll(?l-331)
y The results are expressed as percentages of cells rxpressmg the various range for 100 mononuclear phagocytes m three to four expenmcnfc
Alveolar
macrophages After 7 days of culture X6 93 40 19 99 91 39
93(1-6551 6X(1-674) 43(6-248) 43(6-227) 137(32&543) 177(25-Rl9) 49(5-367)
cell-surface
anttgens
Shonly after collection v3 5?(3-1631 74 25(I-281) 71 21 (2-56) 17 19(11-307) 100 250(22-668) I(X) 164(42&635) IO0 I61 (32-585) and the median
macrophages After 7 days of culture X5 x0 IX 35 100 100 95
24 (l-93) 56 (3-2601 I7 (3-32) 13(1L34) 117(39-195) 54(2-143) 39(1-X8)
specific A,,, per cell and
surface antigens by cells, had to be considered. Control experiments showed that none of these procedures has an effect on the phenotype of mononuclear phagocytes; the low concentration of glutaraldehyde used for fixation of the cells has no effect either (25). Therefore, it is justifiable to compare the phenotypes of the various populations of mononuclear phagocytes. The expression by peritoneal macrophages of cell antigens, defined by Mabs F4/80, M1/70, and 2.4.G.2 and assessedby the present quantitative method, is consistent with the results obtained with a radioimmunobinding assay(7). Furthermore, the percentagesof mononuclear phagocytes expressing complement and Fc receptors, as assessedby Mabs and rosette assays(26) are similar. Analysis of cell-surface characteristics in populations of mononuclear phagocytes by radioimmunobinding assays,fluorescence activated cell-sorter analysis (FACS), and enzyme-linked immunosorbent assayshas the disadvantage that the results represent average values for cells expressing a certain cell-surface antigen and cells lacking the antigen. This drawback was overcome by quantitating the expression of cellsurface antigens by individual mononuclear phagocytes. Expression of the results per unit cell-surface area is an important advantage offered by the present method, becausethe results are independent of the cell size. Expression of cell-surface antigens by the monoblasts, promonocytes, and macrophagelike cells in bone marrow cultures, expressedper 0.25pm* cell-surface area, was similar. Calculation of the specific Ai”l’s per cell showed that the expression of all cell-surface antigens except Ia antigen by macrophagelike cells was considerably higher than that for monoblasts and promonocytes (Fig. 3) which is in agreement with the results reported by others (6). Distributions of the various cell-surface antigens on mononuclear phagocytes, expressed on a per cell basis, were not normal. Comparison the results obtained with cultured mononuclear phagocytes and expressed on a per cell basis showed that expression of all cell-surface antigens (except Ia antigen) by cultured blood monocytes was considerably higher than by monoblasts and promonocytes (Fig. 3 and Table 2). Comparison of the phenotype of cultured blood monocytes and cultured peritoneal macrophagesrevealed enhanced expression of Mac-2 antigen and reduced expression of Fc receptor II and Mac-3 antigen by peritoneal macrophages (Table 2). Cultured
384
NIBBERING, LEIJH, AND VAN FURTH
alveolar macrophages expressantigen F4/80, FcR II, and Ia antigen significantly less than blood monocytes (Table 2). In sum, expression of the various cell-surface antigens on a per cell basis showed that the development from monoblast/promonocyte to blood monocytes is accompanied by an increase in both the size of the cells and the expression of the various cell-surface antigens; when blood monocytes differentiate into tissue macrophages these cell-surface characteristics either remain the same (peritoneal macrophages) or diminish (alveolar macrophages). Mononuclear phagocytes studied shortly after collection showed a slightly different picture of antigen expression on a per cell basis. Compared to the phenotypes of blood monocytes, we found enhanced expression of antigen F4/80, CR III, FcR II, Ia, and Mac-3 antigens by peritoneal macrophages and reduced expression of antigen F4/80 and CR III and enhanced expression of common leukocyte antigen and the Mac-2 and Mac-3 antigens by alveolar macrophages (Table 2). These results agree with the few available quantitative data on the expression of cell-surface antigens by circulating monocytes and peritoneal and alveolar macrophages obtained by radioimmunobinding assaysand FACS analysis (5, 7-9). Since there are very few monoblasts and promonocytes in freshly isolated bone marrow (1 I), the phenotypes of these mononuclear phagocytes were not studied. From the results of the present study it is evident that both cell size and cell antigen expression alter during culture. Unfortunately, our observations show no specific patterns of alteration in cell-surface antigen expression during culture, and the reason for the alterations in the phenotypes of mononuclear phagocytes during culture in vitro is not known either. Although the changes in the phenotype observed during culture might not be physiological, in vitro differentiation and aging of the cells could be responsible for the observed alterations in the morphology and phenotype of mononuclear phagocytes. REFERENCES 1. Van Furth, R., In “Mononuclear Phagocytes: Functional Aspects” (R. van Furth, Ed.), pp. I-30. Nijhoff, The Hague/Boston/London, 1980. 2. Van Furth, R., In “Methods for Studying Mononuclear Phagocytes” (D. 0. Adams, P. J. Edelson and H. Koren, Eds.), pp. 243-252. Academic Press,New York, 1981. 3. Springer, T. A., In “Methods for Studying Mononuclear Phagocytes” (D. 0. Adams, P. J. Edelson, and H. Koren, Eds.), pp. 305-3 13. Academic Press,New York, 1981. 4. Todd, R. F., and Schlossman, S. F., In “Immunology of the Reticuloendothelial System: A Comprehensive Treatise” (J. A. Bellanti and H. B. Herscowitz, Eds.), Vol. 6, pp. 87-l 12. Plenum, New York, 1984. 5. Austyn, J. M., and Gordon, S., Eur. J. Immunol. 11,805, 1981. 6. Hirsch, S., Austyn, J. M., and Gordon, S., J. Exp. Med. 154,7 15, 1981. 7. Ezekowitz, R. A. B., and Gordon, S., J. Exp. Med. 155, 1623, 1982. 8. Ho, M. K., and Springer, T. A., J. Immunol. 128,228 1, 1982. 9. Nussenzweig, M. C., Steinman, R. M., Unkeless, J. C., Witmer, M. D., Gutchinov, B., and Cohn, Z. A., J. Exp. Med. 154, 168, 1981. 10. Nibbering, P. H., Leijh, P. C. J., and van Furth, R., J. Histochem. Cytochem. 33,453, 1985. 1I. Goud, Th. J. L. M., Schotte, C., and van Furth, R., J. Exp. Med. 142, I 180, 1975. 12. VanFurth,R.,andCohn,Z.A., J. Exp.Med. 128,415,1975. 13. Blusse van Oud Alblas, A., and van Furth, R., J. Exp. Med. 149, 1504, 1979. 14. Ledbetter, J. A., and Herzenberg, L. A., Immunol. Rev. 47,63, 1979. 15. Springer, T. A., Galfre, G., Secher, D. S., and Milstein, C., Eur. J. Immunol. 9, 30 1, 1979. 16. Ho, M. K., and Springer, T. A., J. Immunol. 128, 1221, 1982.
MEMBRANE 17. 18. IY. 20. ?I. 2’. 73. ‘4. 25. 26.
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
385
Ho. M. K., and Springer, T. A.. J. Biol. ~‘/~on. 258,636. 1983. Bhattacharya. A., Dorf, M. E.. and Springer. T. A.. J. Iwrmurzol. 127.2488. 198 I. Unkeless. J. C.,J. Ill-p Med. 150, 580. 1979. Kearny. J. F.. Radbruch, A., Liesegang, B., and Rajewsky, K., J. Immwol. 123, 1548. 1979. Van der Ploeg. M., Van den Broek. K.. Smeulders. A. W. M., Vossepoel. A. M.. and van Duijn. P.. lZr.rtochcr?~i~rr~ 54, 273. 1977. Snedecor. G. W.. and Cochran, W. G.. “Statistical Methods.” Iowa State Uni\. Press. Cambriss. Ames. IA. 1973. Siegel. S. “Nonparametric Statistics.” McGraw-HIII. International Book Co.. Tohyo. Japan 1956. NIbbering. P. H.. Marijnen. J. G. J.. Raap, 4. K.. Leijh. P. C. J.. and van Furth. R.. Micfoc,llrr)lI.strr. 84, 538. lY86. Leenen. P. J. M.. Jansen. .4. M. A. C.. and ban Ewijk. W.. In “Techniques tn Immunoc>tochemist~” (G. R. Bullock and P. Petrusz, Eds.). Vol. 3. pp. l-24. Academic Press. London. 1985. Diesselhoff-den Dulk. M. M. C.. and \an Furth. R.. b7 “Methods for Studying Mononuclear Phagoc)tcs” CD 0. .Adams. P. J. Edelson. and H. Korcn. Eds.). pp. 253-272. Academic Press. New \‘ork. IYXI.
IMMUNOLOGY
105,374-385
Quantitative
(1987)
lmmunocytochemical Characterization of Mononuclear Phagocytes
I. Monoblasts,
Promonocytes, Monocytes, and Alveolar Macrophages’
and Peritoneal
PETERH. NIBBERING,~ETER C.J.LEIJH, ANDRALPHVANFURTH Department of Infectious Diseases, University Hospital, P.O. Box 9600, 2300 RC Leiden. The Netherlands Received November 3, 1986; acceptedNovember 18, 1986 The purpose of the present study was to compare the phenotype of tissue macrophages with that of their precursors in the bone marrow and blood. The phenotype was determined on the basis of the quantitative binding of monoclonal antibodies to cell-surface antigens (antigen F4/80, complement receptor III, Fc receptor II, Ia antigen, common leukocyte antigen, and Mac-2 and Mac-3 antigens) on individual mononuclear phagocytes. Monoclonal antibody binding to cells, detected by the biotin-avidin immunoperoxidase procedure, was quantitated by cytophotometric determination of the amount of enzyme reaction product on cells. The results of this quantitation are expressedasthe median of the specific absorbance per unit of cell-surface area (0.25 pm2) and per cell. Shortly after collection of the mononuclear phagocytes, binding of all monoclonal antibodies except those directed against the common leukocyte and Mac-2 antigens to peritoneal macrophages was enhanced compared with binding to blood monocytes; for alveolar macrophages we found reduced binding of monoclonal antibodies F4/80 and Ml/70 (complement receptor III) and enhanced binding of monoclonal antibodies with specificity for the common leukocyte antigen and Mac-2 and Mac-3 antigens. The results obtained with cultured mononuclear phagocytes show that during the development from monoblast to tissue macrophages, monoclonal antibody binding to the various types of mononuclear phagocyte, expressed per unit of cell-surface area, was not significantly altered except that of M3/38 (Mac2 antigen) to peritoneal macrophages and that of F4/80 and Ml/70 (complement receptor III) to alveolar macrophages. Expressed on a per cell basis, the results show an increase in the binding of all monoclonal antibodies except those directed against the Fc receptor II and Mac-3 antigen during the development from promonocytes to peritoneal macrophages; binding of most monoclonal antibodies to alveolar macrophages was considerably lower than that to blood monocytes. It is concluded that the expression of the various cell-surface antigens alters during mononuclear phagocyte differentitaion. The expression changed also during culture, although distinct patterns of alteration could not be distinguished. 0 1987 Academic press, 1~.
INTRODUCTION Mononuclear phagocytes constitute a cell line originating from the hemopoietic stem cell; the precursors develop in the bone marrow and are transported via the ’ This work was supported by the Netherlands Asthma Foundation.
0008-8749/87 $3.00 Copyright 0 1987 by Academic Press, Inc. All rights of reproduction in any form reserved
374
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
375
blood to the tissues, where they differentiate into tissue macrophages (1). Mononuclear phagocytes can be identified from a number of features such as characteristic morphology, staining for specific enzymes, adherence to surfaces, phagocytic and bactericidal ability, and such membrane characteristics as Fc and C receptors and Ia antigen (2). Because none of these features is specific for mononuclear phagocytes, monoclonal antibodies characterizing determinants on mononuclear phagocytes were produced (reviews in Refs. (3) and (4)). However, most of the anti-macrophage antibodies are not entirely specific for mononuclear phagocytes either: only the antibody designated F4/80 is restricted to cells belonging to the mononuclear phagocyte system of the mouse (5). During differentiation of bone marrow cells in culture, antigen F4/80 first appeared on nonadherent immature mononuclear phagocytes in 3day-old cultures (6). In these cultures virtually all adherent mononuclear phagocytes expressedantigen F4/80, complement receptor type III (CR III). and Fc receptor type II (FcR II). Despite these observations, quantitative information about the expression of cell-surface antigens by immature mononuclear phagocytes is lacking, and quantitative data on the occurrence of antigen F4/80, CR III. and FcR II on mature murine mononuclear phagocytes are limited (5,7-9). The present study was undertaken to compare the phenotypes of peritoneal and alveolar macrophages with those of their precursors in bone marrow cultures and blood. To define the phenotype of individual mononuclear phagocytes. we used monoclonal antibodies and a quantitative cytochemical method ( 10). MATERIALS AND METHODS Animals Specific pathogen-free (SPF) male Swiss mice (Central Institute for the Breeding of Laboratory Animals, TNO, Zeist. The Netherlands) weighing 20-25 g were used. Preparation and Culture qfkfononuclear Phagocytes Bone marrow’ phagocytes. Bone marrow cells were harvested by flushing one of the mouse femurs with Dulbecco modified Eagle’s medium (DMEM; Grand Island Biologicals, Grand Island, NY) and were cultured (for 7 days) by plating 5 X 1O4cells per Leighton tube with a flying coverslip as described elsewhere (11). The culture medium consisted of DMEM containing 20% heat-inactivated horse serum (Flow Laboratories, Irvine, UK) and 30% embryonic mouse fibroblast-conditioned medium as source of colony-stimulating factor- 1 (CSF- 1). Blood monocytes. Blood was collected by cardiac puncture, and erythrocytes and granulocytes were removed by Ficoll-Isopaque centrifugation (p = 1.077; Pharmacia Inc., Uppsala, Sweden) for 20 min at 440 g. The cells in the fraction rich in monocytes and lymphocytes were washed four times with phosphate-buffered saline (PBS; pH 7.2) containing 0.5 U/ml heparin. Cytocentrifuge preparations were made immediately after collection for the study of blood monocytes. For comparison of blood monocytes with bone marrow phagocytes, 1 X IO6cells of the monocyte-rich FicollIsopaque interphase were cultured at 37°C in Leighton tubes with flying coverslips; the culture medium was either Medium 199 (M 199; Microbiological Assoc., Walkersville, MD) containing 20% heat-inactivated newborn calf serum (NBCS; GIBCO
376
NIBBERING,
LEIJH,
AND
VAN
F’URTH
Europe, Paisly, UK), 2000 U/ml sodium penicillin G (Gist-Brocades, Delft, The Netherlands), and 50 pg/ml streptomycin (Gist-Brocades) or DMEM supplemented with horse serum (GIBCO Europe) and mouse embryonic fibroblast-conditioned medium. After incubation for 2 hr, the nonadherent cells were removed by three washes with PBS. The glass-adherent cells were cultured under the described conditions for 7 days. Tissue macrophages. Peritoneal macrophages were collected from the peritoneal cavity by lavage with 2 ml PBS containing 50 U/ml heparin, as described elsewhere ( 12). Alveolar macrophages were isolated by lavage with 15 ml PBS containing 0.6 mM EDTA, as described elsewhere ( 13). Cytocentrifuge preparations were made for the study of tissue macrophages immediately after collection. Peritoneal and alveolar macrophages (5 X 10’ cells/ml) were cultured as described for blood monocytes. Monoclonal Antibodies The rat anti-mouse monoclonal antibody (Mab)-producing hybrid cell lines 30.G. 12 (common leukocyte antigen) (14) Ml/70 (complement receptor III, CR III) (15) M3/38 (Mac-2 antigen) (16), M3/84 (Mac-3 antigen) (17) and M5/114 (Iab,d.q Ied%k) (18) were obtained from the American Type Culture Collection (Rockville, MD). The hybrid cell line producing 2.4.G.2 Mab (Fc receptor II, FcR II) ( 19) was a gift from Dr. J. Unkeless (Rockefeller University, New York, NY). The cells were cultured at 37°C under 5% CO2 in RPM1 1640 medium containing 20 mM Hepes buffer (Flow Laboratories, Ltd.), 2 mM L-glutamine (Microbiological Assoc.), 10% fetal bovine serum (Flow Laboratories, Ltd.), 1000 U/ml sodium penicillin G (GistBrocades), 20 yg/ml gentamycin (Gist-Brocades), and 5 X 10e5A4 2-mercaptoethanol. The culture supernatants were harvested every 3-4 days. The culture supernatant of the F4/80 Mab-producing hybrid cell line (5) was a gift from Dr. S. Gordon (Sir William Dunn School of Pathology, Oxford University, Oxford, UK). Quantitation of the Binding of Monoclonal Antibodies to Mononuclear Phagocytes The method used to quantitate the binding of monoclonal antibodies to individual mononuclear phagocytes and the size of these cells has been described elsewhere ( 10). In short, after fixation of the cells with 0.05% glutaraldehyde (Polysciences, Warrington, CN) in PBS at 4°C for 30 min, they were first incubated for 1 min with 0.01 N HCl (pH 2) to inactivate endogenous peroxidase and then for 10 min at room temperature with 20% normal goat serum to block nonspecific binding sites, and finally were overlayered with a saturating concentration of Mab for 30 min. Supematant of the plasmacytoma cell line P3X 163 A98 653 (P3) (20) served as control. After the removal of excess Mab, monoclonal antibody binding was detected with 2.5 yg/ml biotinylated rabbit anti-rat IgG (Vector Laboratories, Burlingame, CA) in 5% normal mouse serum and 5 pg/ml horseradish peroxidase-conjugated avidin-D (Vector Laboratories), the peroxidase being visualized by incubation for 60 min with 0.6 mg/ml 3’3’-diaminobenzidine-tetrahydrochloride (DAB, Merck, Darmstadt, FGR) in Tris-HCl buffer (pH 7.6) supplemented with 0.0 1%H202. For identification of mononuclear phagocytes on a morphological basis in heterogeneous preparations, the cells were counterstained with 10 pg/ml ethidiumbromide (2,7 diamino-lo-ethyl-9-phenyl-phenanthridiumbromide; Calbiochem-Behring, La
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
377
Jolla, CA) in distilled water for 1 min. The amount of peroxidase reaction product present on the mononuclear phagocytes was determined by measuring the absorbance at 466 nm with a microscope photometer (Zeiss, Oberkochen. FRG) interfaced to PDP 1 I/ 10 microcomputer (Digital Equipment Corporation, Maynard, MA). Simultaneous assessment of the integrated absorbance (A,,,) of the peroxidase reaction product and the ethidiumbromide fluorescence of the nucleus was performed with the microscope photometer provided with an epi-illuminator III RS (Zeiss). Scanning of the cells with photometric sampling at 0.5~brn intervals and computation of the mean .4,,, per 0.25~pm’ projected cell-surface area with the ARRAYSCAN program of the HIDACSYS package have been described elsewhere (2 1). Monoclonal antibody binding to cells in each population of mononuclear phagocytes was determined on 100 randomly selected cells in three to four experiments. The nonspecific binding of antibodies and the autooxidation of DAB were assessed in mononuclear phagocytes exposed to P3-plasmacytoma supernatant instead of Mab and otherwise treated as described above. The mean nonspecific A,,,‘s per 0.25pm’ cell surface were normally distributed. Cells expressing a certain cell-surface antigen were distinguished with 95% confidence limits from those lacking the antigen by setting the threshold A,,,, per 0.25~pm’ cell surface at the mean nonspecific A,,, per 0.25-pm’ cell surface plus twice the SD: i.e.. cells with an observed A,,, per 0.25~pm’ cell surface that was higher than the threshold A,,, were considered to express the cell antigen. The mean specific A,,, per 0.25~pm’ cell surface of the mononuclear phagocytes expressing a certain cell-surface antigen was defined as the mean observed A,,,, per 0.25~pm’ cell surface minus this threshold A,,,. The mean specific A,,,‘s per 0.35-pm’ cell surface were used to determine the phenotype of the mononuclear phagocytes. For comparison, the results were expressed on a per cell basis. The specific A,,, per cell was calculated from the observed A,,, per cell. the mean nonspecific A,,, per cell, and the SD, as described above. The sizes of individual mononuclear phagocytes were estimated with the microscope photometer by determining the projected cell-surface area. The results are expressed as the mean values of 200 randomly selected cells per population of mononuclear phagocytes. Statisticd
und~vis
Cell size distributions and distributions ofthe mean specific A,,,‘s per 0.25~pm’ cell surface of mononuclear phagocytes were analyzed for normality on the basis of tests of skewness and kurtosis (22). P < 0.0 1 marked departure from normality. The significance of differences in the cell size and mean specific Aint’s per 0.25pm’ cell surface of the various populations of mononuclear phagocytes was estimated by the Mann-Whitney U test (23). P < 0.0 1 was considered significant. RESULTS All results are expressed as the percentages of mononuclear phagocytes expressing the various cell-surface antigens and the median of the mean specific A,,t’~ per 0.25pm’ projected cell-surface area of 100 mononuclear phagocytes and their range. Frequency histograms showed that the cell-surface antigens were not normally distrib-
378
NIBBERING, LEIJH, AND VAN FURTH TABLE 1 Size of Mononuclear Phagocytes, as Determined by Cytophotometry” Mononuclear phagocytes
Monoblasts Promonocytes Macrophagelike cells Blood monocytes Peritoneal macrophages Alveolar macrophages
Shortly after collection *
After 7 days of culture’
Size
Range
Size
pm2
elm’
c(m2
ndd nd nd 146 222 161
nd nd nd 61-321 128-427 85-263
128 131 316 216 208 140
Range d 79-185 64-193 118-753 8 l-673 90-886 62-284
a The results are expressedas the median and range of the projected cell-surface area of 200 mononuclear phagocytes in three to five experiments. * The cells were sedimented on microscope slides shortly after collection. ‘Culture was performed for 7 days on glasscoverslips. d Not done.
uted on either freshly collected or cultured mononuclear phagocytes (results not shown). Phenotype of Bone Marrow Phagocytes
Mononuclear phagocytes in 7-day cultures of bone marrow cells were identified on the basis of morphological features and localization within individual mononuclear phagocyte colonies. The following types of mononuclear phagocyte were distinguished (11): monoblasts, i.e., small, round cells with a nuclear-to-cytoplasmic ratio higher than 1; promonocytes, i.e., small, slightly stretched cells with one pseudopod and a nuclear-to-cytoplasmic ratio of about 1; and macrophagelike cells, i.e., large elongated cells with two or more pseudopods or, sometimes, large round cells without a pseudopod and a nuclear-to-cytoplasmic ratio lower than 1. No difference in size between monoblasts and promonocytes were found, but the marcrophagelike cells were significantly larger (P < 0.001) (Table 1). Frequency histograms of the cell sizes showed that monoblasts (P > 0.1) and promonocytes (P > 0.05) were normally distributed whereas the macrophagelike cells were not (results not shown). The antigen F4/80, CR III (Mab M1/70), and Mac-2 (Mab M3/38) and Mac-3 (Mab M3/84) antigens were detected on virtually all monoblasts, promonocytes, and macrophagelike cells (Fig. 1). FcR II (Mab 2.4.G.2) was clearly expressed by almost all monoblasts and promonocytes but was considerably (P < 0.00 1) lessby the macrophagelike cells. Ia antigen (Mab M5/ 114) was weakly expressed by about two-thirds of the monoblasts and about 40% of the other bone marrow mononuclear phagocytes. Common leukocyte antigen (Mab 30.G. 12) was clearly expressed by virtually all monoblasts and considerably (P < 0.001) more strongly by promonocytes and macrophagelike cells (Fig. 1).
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
379
SPECFK AINT PER 0.25 pm2 CELL SURFACE
I
1
::
F4/80
i
2.4 G2
1
: I
/
t ;;
30 G 12
I
/
FIG. 1. Phenotypes of bone marrow mononuclear phagocytes. Bone marrow cells were cultured in the presence of CSF- I for 7 days on coverslips. Expression of cell-surface antigens by the mononuclear phagocytes was quantitated by photometric determination of the amount of peroxidase reaction product. The results are expressed as percentages of cells expressing the various cell-surface antigens and the medianspecific A,., per 0.25~pm’ projected cell-surface area and the range for 100 mononuclear phagocytes from three to four experiments.
Phenotype qf Blood A4onocyte.s Comparison of the cell sizes of freshly collected monocytes and monocytes cultured for 7 days showed that the latter cells were significantly larger (P < 0.00 1) (Table 1). The sizes of the cells in both groups were not normally distributed (results not shown). The binding of monoclonal antibodies to monocytes is shown in Fig. 2. After culture, the percentages of monocytes expressing antigen F4/80, CR III (Mab M1/70), and Mac-2 antigen (Mab M3/38) were lower than for recently collected monocytes, whereas the percentage ofblood monocytes expressing FcR II (Mab 2.4.G.2.) Ia antigen (Mab M5/ 114) and Mac-3 antigen (Mab M3/84) increased during culture (Fig. 2). The percentage of monocytes expressing common leukocyte antigen (Mab 30.G. 12) did not alter after culture. Expression of antigen F4/80 and CR III on blood monocytes was significantly (P < 0.001) higher and FcR II significantly (P < 0.001) less than that by cultured monocytes, and common leukocyte antigen was clearly and Ia, Mac-2, and Mac-3 antigens were weakly expressed by both groups.
I
:
:
: I
I
,
4
4
-I
,
I I
I
M3/84
U3/38
30.G.12
MS/114
FIG. 2. Phenotype of blood monocytes and peritoneal and alveolar macrophages shortly after collection and after 7 days of culture. The results are expressed as percentages of cells expressing a given cell-surface antigen and the median-specific Aimt per 0.2%pm’ cell-surface area and range for 100 mononuclear phagocytes in three to four experiments.
I:
I
t :
t
t
I :
I
t
w
w
I
2.C.G.2.
Ml/70
CULTURED MONONUCLEAR PHAGOCYTES SPECIFIC A/NT PER 0.25 pm2 CELL SURFACE
MEMBRANE
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
381
For investigation of the effect of CSF- 1 on the phenotype of cultured mononuclear phagocytes, blood monocytes were cultured for 7 days in the presence and absence of this factor. Comparison of the phenotypes of both groups of monocytes revealed no significant differences. For example, the specific mean Aint per 0.25~pm2 cell surface, which reflects the expression of antigen F4/80, amounted to 0.109 (range O.OOS0.33 1) for monocytes cultured with CSF-1 and that for monocytes without this factor to 0.105 (range 0.005-0.457). This result and those for the other cell-surface antigens (results not shown) showed no effect of CSF- 1 on the phenotype of blood monocytes.
Phenotype cfPeritoneu1 Macrophages No difference in size between cultured and noncultured peritoneal macrophages was observed (Table I). Frequency histograms showed that the sizes of the cells in both groups of peritoneal macrophages were not normally distributed (results not shown). The binding of monoclonal antibodies to peritoneal macrophages is shown in Fig. 2. The percentage of peritoneal macrophages expressing the various cell-surface antigens decreased during culture. Expression of antigen F4/80 was intense by both groups of cells and that of CR III (Mab M l/70) by noncultured peritoneal macrophages was strong, whereas that of cultured peritoneal macrophages was significantly (P < 0.00 1) weaker. FcR II (Mab 2.4.G.2.) and Ia antigen (Mab M5/ 114) were weakly expressed by both groups. Common leukocyte antigen (Mab 30.G. 12) was clearly expressed by noncultured and significantly (P < 0.00 1) higher by cultured peritoneal macrophages. Expression of Mac-2 antigen (Mab M3/38) by noncultured was weak and that by cultured peritoneal macrophages was significantly (P < 0.001) higher. The reverse was the case for the Mac-3 antigen (Mab M3/84) (P < 0.00 1). Control experiments showed that CSF- 1 does not affect the phenotype of cultured peritoneal macrophages (results not shown). No effect of either Ficoll-Isopaque density centrifugation or EDTA on the phenotype of peritoneal macrophages was seen, and the same holds for sedimentation of cells on the microscope slides and adherence of cells to glass during incubation for 1 hr (results not shown).
Comparison of the size of cultured and noncultured alveolar macrophages showed no significant difference (Table 1). Frequency histograms of the cell sizes in both groups showed that the distributions were not normal (results not shown). The binding of monoclonal antibodies to alveolar macrophages is shown in Fig. 2. After 7 days of culture, the percentage of alveolar macrophages expressing antigen F4/80. CR III (Mab M1/70), FcR II (Mab 2.4.G.2.). and Mac-3 antigen (Mab M3/ 84) dropped. The percentages of macrophages expressing Ia antigen (Mab M5/ 114). common leukocyte antigen (Mab 30.G. 12). and Mac-2 antigen (Mab M3/38) were not altered after culture. Antigen F4/80 was readily expressed by noncultured alveolar macrophages and considerably less so (P < 0.00 1) by cultured alveolar macrophages. CR III, FcR II, and Ia antigen were weakly expressed by both groups of cells. Expression of the common leukocyte, Mac-2, and Mac-3 antigens by alveolar macrophages was intense, and was considerably (P < 0.001) weaker for similar cells after culture.
382
NIBBERING, LEIJH, AND VAN FIJRTH SPECIFIC AINT PER CELL
1
I!
30.G.12
I I
I I
I!
I ::
I f
M313.9
FIG. 3. Binding of monoclonal antibodies to bone marrow mononuclear phagocytes. The results are expressed as percentages of cells expressing the various cell antigens and the median-specific Ai,, per cell and range for 100 mononuclear phagocytes in three to four experiments.
DISCUSSION The present results concerning cultured mononuclear phagocytes show that the expression of all cell-surface antigens except Mac-2 and Mac-3 antigens, does not alter during the development from monoblast/promonocyte to blood monocytes; upon differentiation of blood monocytes into peritoneal macrophages the expression of the F4/80 and Mac-2 antigens increases and differentiation of monocytes into alveolar macrophages is accompanied by a decreasein the expression of antigen F4/ 80, CR III, and FcR II. This conclusion is based on the results of the quantitation of the expression of cell-surface antigen per 0.25~pm2cell-surface area. The method we used to quantitate the expression of cell-surface antigens by cells is based on Mabs, peroxidase immunocytochemistry, and photometric determination of the amount of peroxidase reaction product. This method is reproducible, sensitive, and specific (10). The Aint of the peroxidase reaction product proved to be proportional to the amount of cell-surface antigen immobilized on nitrocellulose paper (24) (P. H. Nibbering et al., Submitted for publication). The possibility that the procedures used for the preparation ofthe cells, e.g., Ficoll-Isopaque, EDTA, or CSF1, and sedimentation and adherence to glass,have an effect on the expression of cell-
MEMBRANE ANTIGENS IN MONONUCLEAR
383
PHAGOCYTE DIFFERENTIATION
TABLE 2 Binding of Monoclonal Antibodies to Blood Monocytes and Peritoneal and Alveolar Macrophages Shortly after Collection and after 7 Days of Culture” Peritoneal
Blood monocytes Shortly after collection F4/80 Ml/70 2.4.G.2 MS/II4 3O.G 12 M3/3R M3184
92 85 I0 16 98 48 22
112(4460) 102(8-327) 7(3-?I) I’(:-160) 70(4-245) 28 (2-162) I, (l-43)
After 7 days of culture 93 110(2-418) XX 64(5-244) 50 79 (3-408) 34 46(11-181) 99 116(21-322) 44 74(13-213) 47 70(29-167)
Shonly after collection 100 99 8X 40 9X 40 IO0
196(29-567) 187(30--166) 45(1-1601 52(6-89) X3(9-158) l4(1-91) lll(?l-331)
y The results are expressed as percentages of cells rxpressmg the various range for 100 mononuclear phagocytes m three to four expenmcnfc
Alveolar
macrophages After 7 days of culture X6 93 40 19 99 91 39
93(1-6551 6X(1-674) 43(6-248) 43(6-227) 137(32&543) 177(25-Rl9) 49(5-367)
cell-surface
anttgens
Shonly after collection v3 5?(3-1631 74 25(I-281) 71 21 (2-56) 17 19(11-307) 100 250(22-668) I(X) 164(42&635) IO0 I61 (32-585) and the median
macrophages After 7 days of culture X5 x0 IX 35 100 100 95
24 (l-93) 56 (3-2601 I7 (3-32) 13(1L34) 117(39-195) 54(2-143) 39(1-X8)
specific A,,, per cell and
surface antigens by cells, had to be considered. Control experiments showed that none of these procedures has an effect on the phenotype of mononuclear phagocytes; the low concentration of glutaraldehyde used for fixation of the cells has no effect either (25). Therefore, it is justifiable to compare the phenotypes of the various populations of mononuclear phagocytes. The expression by peritoneal macrophages of cell antigens, defined by Mabs F4/80, M1/70, and 2.4.G.2 and assessedby the present quantitative method, is consistent with the results obtained with a radioimmunobinding assay(7). Furthermore, the percentagesof mononuclear phagocytes expressing complement and Fc receptors, as assessedby Mabs and rosette assays(26) are similar. Analysis of cell-surface characteristics in populations of mononuclear phagocytes by radioimmunobinding assays,fluorescence activated cell-sorter analysis (FACS), and enzyme-linked immunosorbent assayshas the disadvantage that the results represent average values for cells expressing a certain cell-surface antigen and cells lacking the antigen. This drawback was overcome by quantitating the expression of cellsurface antigens by individual mononuclear phagocytes. Expression of the results per unit cell-surface area is an important advantage offered by the present method, becausethe results are independent of the cell size. Expression of cell-surface antigens by the monoblasts, promonocytes, and macrophagelike cells in bone marrow cultures, expressedper 0.25pm* cell-surface area, was similar. Calculation of the specific Ai”l’s per cell showed that the expression of all cell-surface antigens except Ia antigen by macrophagelike cells was considerably higher than that for monoblasts and promonocytes (Fig. 3) which is in agreement with the results reported by others (6). Distributions of the various cell-surface antigens on mononuclear phagocytes, expressed on a per cell basis, were not normal. Comparison the results obtained with cultured mononuclear phagocytes and expressed on a per cell basis showed that expression of all cell-surface antigens (except Ia antigen) by cultured blood monocytes was considerably higher than by monoblasts and promonocytes (Fig. 3 and Table 2). Comparison of the phenotype of cultured blood monocytes and cultured peritoneal macrophagesrevealed enhanced expression of Mac-2 antigen and reduced expression of Fc receptor II and Mac-3 antigen by peritoneal macrophages (Table 2). Cultured
384
NIBBERING, LEIJH, AND VAN FURTH
alveolar macrophages expressantigen F4/80, FcR II, and Ia antigen significantly less than blood monocytes (Table 2). In sum, expression of the various cell-surface antigens on a per cell basis showed that the development from monoblast/promonocyte to blood monocytes is accompanied by an increase in both the size of the cells and the expression of the various cell-surface antigens; when blood monocytes differentiate into tissue macrophages these cell-surface characteristics either remain the same (peritoneal macrophages) or diminish (alveolar macrophages). Mononuclear phagocytes studied shortly after collection showed a slightly different picture of antigen expression on a per cell basis. Compared to the phenotypes of blood monocytes, we found enhanced expression of antigen F4/80, CR III, FcR II, Ia, and Mac-3 antigens by peritoneal macrophages and reduced expression of antigen F4/80 and CR III and enhanced expression of common leukocyte antigen and the Mac-2 and Mac-3 antigens by alveolar macrophages (Table 2). These results agree with the few available quantitative data on the expression of cell-surface antigens by circulating monocytes and peritoneal and alveolar macrophages obtained by radioimmunobinding assaysand FACS analysis (5, 7-9). Since there are very few monoblasts and promonocytes in freshly isolated bone marrow (1 I), the phenotypes of these mononuclear phagocytes were not studied. From the results of the present study it is evident that both cell size and cell antigen expression alter during culture. Unfortunately, our observations show no specific patterns of alteration in cell-surface antigen expression during culture, and the reason for the alterations in the phenotypes of mononuclear phagocytes during culture in vitro is not known either. Although the changes in the phenotype observed during culture might not be physiological, in vitro differentiation and aging of the cells could be responsible for the observed alterations in the morphology and phenotype of mononuclear phagocytes. REFERENCES 1. Van Furth, R., In “Mononuclear Phagocytes: Functional Aspects” (R. van Furth, Ed.), pp. I-30. Nijhoff, The Hague/Boston/London, 1980. 2. Van Furth, R., In “Methods for Studying Mononuclear Phagocytes” (D. 0. Adams, P. J. Edelson and H. Koren, Eds.), pp. 243-252. Academic Press,New York, 1981. 3. Springer, T. A., In “Methods for Studying Mononuclear Phagocytes” (D. 0. Adams, P. J. Edelson, and H. Koren, Eds.), pp. 305-3 13. Academic Press,New York, 1981. 4. Todd, R. F., and Schlossman, S. F., In “Immunology of the Reticuloendothelial System: A Comprehensive Treatise” (J. A. Bellanti and H. B. Herscowitz, Eds.), Vol. 6, pp. 87-l 12. Plenum, New York, 1984. 5. Austyn, J. M., and Gordon, S., Eur. J. Immunol. 11,805, 1981. 6. Hirsch, S., Austyn, J. M., and Gordon, S., J. Exp. Med. 154,7 15, 1981. 7. Ezekowitz, R. A. B., and Gordon, S., J. Exp. Med. 155, 1623, 1982. 8. Ho, M. K., and Springer, T. A., J. Immunol. 128,228 1, 1982. 9. Nussenzweig, M. C., Steinman, R. M., Unkeless, J. C., Witmer, M. D., Gutchinov, B., and Cohn, Z. A., J. Exp. Med. 154, 168, 1981. 10. Nibbering, P. H., Leijh, P. C. J., and van Furth, R., J. Histochem. Cytochem. 33,453, 1985. 1I. Goud, Th. J. L. M., Schotte, C., and van Furth, R., J. Exp. Med. 142, I 180, 1975. 12. VanFurth,R.,andCohn,Z.A., J. Exp.Med. 128,415,1975. 13. Blusse van Oud Alblas, A., and van Furth, R., J. Exp. Med. 149, 1504, 1979. 14. Ledbetter, J. A., and Herzenberg, L. A., Immunol. Rev. 47,63, 1979. 15. Springer, T. A., Galfre, G., Secher, D. S., and Milstein, C., Eur. J. Immunol. 9, 30 1, 1979. 16. Ho, M. K., and Springer, T. A., J. Immunol. 128, 1221, 1982.
MEMBRANE 17. 18. IY. 20. ?I. 2’. 73. ‘4. 25. 26.
ANTIGENS
IN MONONUCLEAR
PHAGOCYTE
DIFFERENTIATION
385
Ho. M. K., and Springer, T. A.. J. Biol. ~‘/~on. 258,636. 1983. Bhattacharya. A., Dorf, M. E.. and Springer. T. A.. J. Iwrmurzol. 127.2488. 198 I. Unkeless. J. C.,J. Ill-p Med. 150, 580. 1979. Kearny. J. F.. Radbruch, A., Liesegang, B., and Rajewsky, K., J. Immwol. 123, 1548. 1979. Van der Ploeg. M., Van den Broek. K.. Smeulders. A. W. M., Vossepoel. A. M.. and van Duijn. P.. lZr.rtochcr?~i~rr~ 54, 273. 1977. Snedecor. G. W.. and Cochran, W. G.. “Statistical Methods.” Iowa State Uni\. Press. Cambriss. Ames. IA. 1973. Siegel. S. “Nonparametric Statistics.” McGraw-HIII. International Book Co.. Tohyo. Japan 1956. NIbbering. P. H.. Marijnen. J. G. J.. Raap, 4. K.. Leijh. P. C. J.. and van Furth. R.. Micfoc,llrr)lI.strr. 84, 538. lY86. Leenen. P. J. M.. Jansen. .4. M. A. C.. and ban Ewijk. W.. In “Techniques tn Immunoc>tochemist~” (G. R. Bullock and P. Petrusz, Eds.). Vol. 3. pp. l-24. Academic Press. London. 1985. Diesselhoff-den Dulk. M. M. C.. and \an Furth. R.. b7 “Methods for Studying Mononuclear Phagoc)tcs” CD 0. .Adams. P. J. Edelson. and H. Korcn. Eds.). pp. 253-272. Academic Press. New \‘ork. IYXI.