Immunobiol., vol. 176, pp. 432-439 (1988)
Department of Infectious Diseases, University Hospital of Leiden, Leiden, The Netherlands
Quantitative Immunocytochemical Characterization of Four Murine Macrophage-Like Cell Lines PETER
H. NIBBERING and RALPH VAN FURTH
Received August 18, 1987· Accepted November 2, 1987
Abstract The aim of the present study was to obtain quantitative data on the expression of seven cellsurface antigens by the murine macrophage-like cell lines WEHI-3, P388-D1, ]774.1, and PU5-1.8, and to compare these data with those obtained for various mononuclear phagocytes. Binding of monoclonal antibodies F4/80, M1I70, 2.4.G.2., 30.G.12, M3/38, M3/84, and 59AD2.2 to cells from these four cell lines was detected by the biotin-avidin immunoperoxidase staining method and quantified by cytophotometry. The results are expressed as the percentage cells expressing a given antigen and the mean specific integrated absorbance per 0.25 11m2 cell-surface area. The results revealed that the phenotypes of the four macrophage-like cell lines differ markedly. Expression of the cell antigens by the cells of the various cell lines did not follow a normal distribution. Although the cell lines divide continuously and have certain characteristics in common with mature mononuclear phagocytes, none of the macrophage-like cell lines accurately resembles any of the mononuclear phagocyte populations. The strongest correlations for membrane were found between on the one hand WEHI-3 and P388-D1 cells and mono blasts and promonocytes (P > 0.011) on the other. P388-Dl, ]774.1 and PUS-1.8 cells expressed four of the six antigens to the same extent as skin macrophages (P > 0.012). With respect to expression of antigens recognized by antibodies 2.4.G.2., 30.G.12, M3/38, and M3/ 84 PUS-1.8 cells resembled activated peritoneal macrophages (P > 0.031). It is concluded that WEHI-3 is the most immature cell line, followed by the P388-D1 cell line, while ]774.1, and PU5-1.8 are the most mature cell lines.
Introduction Murine macrophage-like cell lines, which have many characteristics in common with normal mononuclear phagocytes (1-5), are used for various types of investigations. The popularity of these cell lines lies in their convenience as a source of pure macrophages. The cells of the macrophagelike cell lines are considered to be arrested in a particular stage of differentiation (5). Because these cells divide continuously, some of the cell lines could therefore resemble immature, dividing mononuclear phagocytes, i.e., Abbreviations: PBS = phosphate-buffered saline; A;nt = integrated absorbance; SD = standard deviation; RIBA = radio-immunobinding assay; FACS = fluorescence-activated cell sorter; ELISA = enzyme-linked immunosorbent assay.
Antigens on Macrophage-Like Cell Lines' 433
mono blasts and promonocytes, whereas on the basis of various characteristics (1-5), others could resemble macrophages. The development of a quantitative immunocytochemical method for assessment of the expression of antigens by individual cells allowed us to study changes in the expression of cell-surface antigens during mononuclear phagocyte differentiation (6, 7). The results showed that the phenotypes of mononuclear phagocytes from bone marrow, blood, and various tissues differed and that the expression of cell-surface antigens changed during culture. The purpose of the present study was to quantify the expression of cell-surface antigens by four of the most frequently used murine macrophage-like cell lines and to compare these results with those which were obtained from other mononuclear phagocyte populations (6-8).
Materials and Methods Culture of macrophage-like cell lines
The cell lines WEHI-3 (9), P388-Dl (10), }774.1 (11), and PU5-1.8 (9) were all kindly donated by Dr. P. Ralph (Memorial Sloan Kettering Cancer Center, Rye, NY, U.S.A.). The culture medium consisted of RPMI 1640 medium containing 20 mM HEPES buffer (Flow Laboratories Ltd., Irvine, U.K.), 2.0 mM L-glutamine (M. A. Bioproducts, Walkersville, MD, U.S.A.), 10 % fetal calf serum (Flow Laboratories Ltd.), 1,000 U/ml sodium-penicillin G (Gist-Brocades NV, Delft, The Netherlands), and 50 [lg/ml streptomycin (Gist-Brocades NV). About 1 x 105 cells/ml were suspended in culture medium and incubated in plastic tissue-culture flasks (Falcon Labware, Becton Dickinson and Co., Oxnard, CA, U.S.A.) in a humidified atmosphere (5 % CO 2 in air) at 37°C for 5 days. The cells were collected after agitation by repeated pipetting, washed, and resuspended in phosphate-buffered saline (PBS, pH 7.4) to a concentration of 1.5 x 106 cells/m!; subsequently, cytocentrifuge preparations were made. Quantitation of monoclonal antibody binding to macrophage-like cell lines
The method for quantitation of the binding of monoclonal antibodies to individual cells has been described in detail elsewhere (12). Briefly, binding of monoclonal antibodies to cells is detected by the biotin-avidin immunoperoxidase method. The amount of final reaction product, i.e. the integrated absorbance (Aint), was measured by cytophotometry. The phenotype is defined by the binding of monoclonal antibodies F4/80 (13), M1I70 (complement receptor III, CR III, 14), 2.4.G.2. (Fc receptor II, FcR II, 15), 30.G.12 (common leukocyte antigen, 16), M3/38 (Mac-2 antigen, 17), M3/84 (Mac-3 antigen, 18), and 59AD2.2 (Thy-l antigen, 16) to cells of each cell line. For each cell line, the binding of monoclonal antibodies to 100 randomly selected cells was assessed in 3-4 experiments and expressed as the median and range of the mean specific Aint per 0.25 [lm 2 cell-surface area. A cell was considered to express a given antigen when the mean Aint per 0.25 [lm2 cell surface was more than the average of the mean non-specific Aint per 0.25 [lm2 cell surface for mononuclear phagocytes incubated with plasmacytoma supernatant instead of monoclonal antibody plus twice the SD. Expression of the various antigens by macrophage-like cells was classified, on the basis of the specific mean Aints per 0.25 [lm2 cell-surface area, as weak, moderate, or intense (7). Statistical analysis
Distributions of the mean specific Aint per 0.25 11m2 cell surface for the cells of the four cell lines were analyzed for normality with the Fisher exact test (19); P < 0.01 marked departure
434 . PETER H. NIBBERING and RALPH VAN FURTH from normality. The significance of differences in the mean specific Aim per 0.25 ~m2 cell surface between the four macrophage-like cell lines and the various populations of murine mononuclear phagocytes was estimated by the Mann-Whitney U -test (20). P <0.01 was considered significant.
Phenotype of macrophage-like cell lines The phenotypes of the four cell lines are shown in Figure 1. Frequency histograms of the mean specific Aints per 0.25 Ilm 2 cell surface did not show a normal distribution (results not shown). Antigen F4/80, which is present exclusively on mononuclear phagocytes (13), was weakly expressed by a small percentage of PU5-1.8 cells and about 50 % of the WEHI-3 cells; a significantly more intense (P < 0.001) expression was noted for 80 % of the ]774.1 and P388-Dl cells. The CR III (antibody M1/70) was weakly expressed by 76-90 % of the PU5-1.8, WEHI-3, and ]774.1 cells and by almost all P388-Dl cells. The FcR II (antibody 2.4.G.2.) was weakly expressed by about 75 % of the PU5-1.8 cells; a significantly more intense (P < 0.001) expression was exhibited by almost all cells of the WEHI-3 and ]774.1 cell lines. Virtually all P388-Dl cells expressed this receptor intensely (P < 0.001). Almost all cells of the four cell lines expressed the common leukocyte antigen (antibody 30.G.12) moderately. Weak expression of the Mac-2 antigen (antibody M3/38) was exhibited by 60-75 % of the WEHI-3 and ]774.1 cells, and significantly more intense (P < 0.001) expression by about 90 % of the P388-Dl and SPECIFIC AINT PER O.25lJm2 CELL-SURFACE AREA mLLNE
a::::J rr=:::J 1I
00 01 Dc::::::::::J 0
0c:::::J 1Bu=J 95
M3I18 100 100 84 100
c:::=::::J1 CJc:::=::J 01
Figure 1. Phenotypes of the macrophage-like cell lines WEHI-3, P3SS-Dl, ]774.1, and PUS-l.S. Monoclonal antibody binding to cells was analyzed by photometric determination of the amount of peroxidase reaction product. The results are expressed as the percentage of cells expressing a given cell-surface antigen and the median (vertical crossbar) and range (horizontal bar) of the mean specific Aint per 0.25 ~m2 cell surface for the antibody-binding cells found in 3-4 experiments.
Antigens on Macrophage-Like Cell Lines' 435
PU5-1.8 cells. The Mac-3 antigen (antibody M3/84) was expressed moderately by all PU5-1.8 cells and significantly more intensely (P < 0.001) by virtually all cells of the other three macrophage-like cell lines. Expression of the Thy-1 antigen (antibody 59AD2.2) was intense by all WEHI-3 cells (median specific mean Aint per 0.25 flm 2 cell surface was 0.247 with a range of 0.013-0.496) and moderate by 93 % of the ]774.1 cells (median specific mean Aim per 0.25 flm 2 cell surface was 0.122 with a range of 0.014-0.382); PU5-1.8 and P388-D1 cells did not express this antigen at all.
Discussion From the present results, it is concluded that the cells of the four murine macrophage-like cell lines, i.e., WEHI-3, P388-D1, ]774.1, and PU5-1.8, differ from one another in expression of cell-surface antigens. The quantitative results of this study are not consistent with the results obtained with either radioimmunobinding assays (RIB A) (2) and fluorescence-activated cell sorter (FACS) analysis (2) or FACS analysis (5) and enzyme-linked immunosorbent assays (ELISA) (5). Other characteristics of the cell lines, such as specific enzymes and functional activities, have also been reported by a number of laboratories to differ (1, 3-5). These differences might be due to the assay methods used. For instance, analysis of the expression of antigens by cells by RIBA, F ACS analysis, and ELISA has the disadvantage that the results are expressed as the average of the values for both cells that express the antigen and negative cells. To avoid this problem, we quantified the expression of antigens by individual cells. Comparison of the present results, expressed as the mean (and SD) of the values for all cells instead of only the mean (and SD) for the antigen-expressing cells, with the results obtained with the above-mentioned methods (2, 5) also showed dissimilarities, indicating that other explanations have to be considered. It is conceivable that variant clones of the cell lines were selected during the culture procedures employed by the various laboratories. This seems to be the case, because when the membrane characteristics of five clones of the J774 cell line were analyzed quantitatively in one laboratory, the results differed considerably (5). Therefore, macrophage-like cell lines should be checked routinely for changes in their characteristics. To determine whether the macrophage-like cell lines represent cells in a certain stage of differentiation, we compared the phenotypes of the macrophage-like cell lines statistically with those of monoblasts, promonocytes, monocytes, various resident tissue macrophages, and alveolar and peritoneal macrophages from mice injected with a sterile stimulus or live BCG. The results, summarized in Figure 2 and Table 1 (from 6-8, and unpublished observations), show that WEHI-3 cells expressed CR III, FcR II, common leukocyte antigen and Mac-3 antigen to about the same extent as
436 . PETER H. NIBBERING and RALPH VAN FURTH
7 7 7 0 1 2 7 1 2 2 0 1 7 0 0 1 2 7 0 0 0 1 0
MA(R(PHAGE -LIKE CEll LINES WEH 1-1
macrqJhage-llke (!!lis ~ood mcnocytI!s
(!!lis mcnary tiSSUl! ~ a1vedar macrqJhages
killed &6 a1wo1ar ~ IIYI! &6 aIwoIar IIIiItJ'OIiIagIs NBCS p!!IituneallTlilCTOlilalJl hVl! &6 p!!IituneallTlilCTOlilalJl
• I •
Figure 2. Comparison of the specific A;n" per 0.25 [tm 2 cell-surface area for four murine macrophage-like cell lines and other mononuclear phagocytes. Differences were analyzed with the Mann-Whitney U-test. P < 0.01 was considered significant. The intensity of staining of macrophage-like cell lines is more than (.), about the same as (mllil), or less than (0) that of the mononuclear phagocyte.
(P > 0.011) and antigen F4/80 and Mac-2 antigen to a significantly (P < 0.001) lesser degree than monoblasts and promonocytes. Comparison of the percentage cells expressing the various antigens revealed no differences, except that a considerably lower percentage of WEHI-3 cells than monoblasts and promonocytes express antigen F4/80 and Mac-2 antigen. WEHI-3 cells are therefore poor representatives of blood monocytes and resident and exudate tissue macrophages. P388-D1 cells and monoblasts, promonocytes and skin macrophages expressed antigen F4/80, CR III, common leukocyte antigen and Mac-2 antigen to about the same degree as (P > 0.012) but FcR II and Mac-3 antigen to a significantly (P < 0.001) lesser degree than P388-D1 cells. The percentages of immature mononuclear phagocytes, skin macrophages, and P388-Dl cells expressing the various antigens were about the same, except that a considerable lower percentage of skin macrophages expressed CR III and FcR II. P388-D1 cells do not resemble either blood monocytes or resident and exudate macrophages very well. Expression of cell antigens by]774.l cells correlates poorly with that by mono blasts, promonocytes, blood monocytes, and almost all resident and exudate macrophages. Expression of antigen F4/80, CR III, common leukocyte antigen, and Mac-3 antigen by ]774.1 cells and skin macrophages
Antigens on Macrophage- Like Cell Lines . 437 Table 1. Comparison of the percentage of antigen-expressing cells of the macrophage-like cell lines and mononuclear phagocytes mononuclear phagocyte
days in culture
F4/80 M1/70 2.4.G.2. 30.G.12 M3/38 M3/84 monoblasts promonocytes macrophage-like cells blood monocytes
spleen macrophages Kupffer cells pulmonary tissue macrophages alveolar macrophages
skin macrophages peritoneal macrophages
killed BCG' alveolar macrophages live BCG b alveolar macrophages NBcse peritoneal macrophages live BCG b peritoneal macrophages WEHI -3 cell line P388-Dl cell line J774.1 cell line PU5-1.8 cell line a
7 7 7 0 2 7 1 2 2 0 1 7 0 0 2 7 0 0 0 1 0
87 100 99 93
92 95 85 89 50 73 94 100 77 73 100 100 100 85 40 30 95 94 79 43 76 84 18
93 88 97 87 98 81 83 94 52 63 86 60
71 56 99 100 100 86 67 35 96 99
96 90 97 76 85
98 99 94 19 84 62 55
97 100 100 98 100 98 100 98
32 38 90 85 21 28 97 97 63 31 3 20 38 46 21
91 100 90 100 70 100 100 94 97 100 100 87 94 100
99 98 98 67 70 63 39 85 56 63 100 100 97 82 60 5 20 83 100 90 52 62 99
84 100 90 90
59 87 74 90
96 95 78
94 97 100 34 67 49 46 87 22 81 100 69 87
71 100 11 24 29 100 83 99 32 99 100 100 84 100
Twenty-one days after intravenous injection of 6 X 106 heat-killed BCG. Twenty-one days after intravenous injection of 6 x 106 live BCG. Twenty-one hours after intraperitoneal injection of 1 ml new-born calf serum (NBCS).
was about the same (P > 0.019), but expression of FcR II and Mac-2 antigen by]774.1 cells was significantly (P < 0.001) different from than that by skin macrophages. The percentage of skin macrophages expressing CR III and FcR II were considerably lower than the percentage of ]774.1 cells. PUS-loS cells are poor representatives of monoblasts, promonocytes, blood monocytes, and most of the resident and exudate macrophages. However, no significant (P> 0.012) difference between the expression of antigen F4/S0, FcR II, common leukocyte antigen, and Mac-2 and Mac-3 antigens between PUS-loS cells and skin macrophages was observed; PUS1.8 cells expressed significantly (P < 0.001) less CR III than skin macrophages. The percentage of PUS-loS cells expressing antigen F4/S0 were considerably lower and the percentage expressing CR III and FcR II
438 . PETER H. NIBBERING and RALPH VAN FURTH
considerably higher than that which was found for skin macrophages. Furthermore, PUS-loS cells expressed FcR II, common leukocyte antigen, and Mac-2 and Mac-3 antigens to about the same degree as (P> 0.031) and antigen F4/S0 and CR III to a significantly lesser extent (P < 0.001) than activated peritoneal macrophages. The percentage of PUS-loS cells expressing antigen F4/S0 were lower and that expressing FcR II higher than found for activated peritoneal macrophages. Interestingly, WEHI-3 and ]774.1 cells expressed the Thy-l antigen, known to be present on hemopoietic stem cells and possibly early immature mononuclear phagocytes (5, 21, 22). However, monoblasts and promonocytes do not express this antigen (unpublished observations). On the basis of the results of the present study and other characteristics of both these cell lines (1-5) and murine mononuclear phagocytes (6, 7,23, 24), we classified the cell lines in a maturation sequence: WEHI-3 seems to be the most immature cell line followed by the P3SS-Dl cell line; ]774.1 and PUS-loS are the most mature cell lines. Acknowledgements This study was supported by the Netherlands Asthma Foundation. The authors are indebted to Mrs. E. M. LOOVENS for secretarial assistance. The authors thank Dr. P. C. J. LEIJH for his critical reading of the manuscript.
References 1. VAN FURTH, R., M. VAN SCHADEWIJK-NIEUWSTAD, 1. ELZENGA-CLAASEN, C. CORNELISSE, and P. NIBBERING. 1985. Morphological, cytochemical, functional, and proliferative characteristics of four murine macrophage-like cell lines. Cell. Immunol. 90: 339. 2. RALPH, P., M. K. Ho, P. B. LITCOFSKY, and T. A. SPRINGER. 1983. Expression and induction in vitro of macrophage differentiation antigens on murine cell lines. J. Immunol. 130: 108. 3. RALPH, P. 1981. Continuous cell lines with properties of mononuclear phagocytes. In: Methods for Studying Mononuclear Phagocytes. D. O. ADAMS, P. J. EDELSON, and H. S. KOREN (Eds.). Academic Press, New York, USA, pp. 155-173. 4. VAN LOOVEREN, H., J. HILGERS, J. M. DE BAKKER, R. A. DE WEGER, P. BREDEROO, and W. DEN OTTER. 1983. Macrophage-like cell lines: endogenous peroxidase activity, cellsurface antigens and colony stimulating factor production. J. Reticuloendothel. Soc. 33: 221. 5. LEENEN, P. J. M., A. M. A. C. JANSEN, and W. VAN EWIJK. 1986. Murine macrophage cell lines can be ordered in a linear differentiation sequence. Differentiation 32: 157. 6. NIBBERING, P. H., P. C. J. LEIJH, and R. VAN FURTH. 1987. Quantitative immunocytochemical characterization of mononuclear phagocytes. 1. Monoblasts, promonocytes, monocytes, and peritoneal and alveolar macrophages. Cell. Immunol. 105: 374. 7. NIBBERING, P. H., P. C. J. LEIJH, and R. VAN FURTH. 1987. Quantitative immunocytochemical characterization of mononuclear phagocytes. II. Monocytes and tissue macrophages. Immunol. 62:71. 8. NIBBERING, P. H., W. SLUITER, G. A. VAN DER HEIDE, and R. VAN FURTH. 1987. Immunocytochemical characterization of the cellular responses to BCG. Submitted for publication.
Antigens on Macrophage-Like Cell Lines' 439 9. RALPH, P., M. A. S. MOORE, and K. NILSON. 1976. Lysozyme synthesis by established human and murine histiocytic lymphoma cell lines. J. Exp. Med. 143: 1528. 10. KOREN, H. S., B. S. HANDWERGER, J. R. WUNDERLICH. 1975. Identification of macrophage-like characteristics in a cultured murine tumor line. J. Immunol. 114: 894. 11. RALPH, P., J. PRICHARD, and M. COHN. 1975. Reticulum cell sarcoma: an effector cell in antibody-dependent cell-mediated immunity. J. Immunol. 114: 898. 12. NIBBERING, P. H., P. C. J. LEIJH, and R. VAN FURTH. 1985. Cytochemical method to quantitate the binding of monoclonal antibodies to individual cells. J. Histochem. Cytochem. 33: 453. 13. AUSTYN, J. M., and S. GORDON. 1981. F4/80, a monoclonal antibody directed against the mouse macrophage. Eur. J. Immunol. 11: 805. 14. SPRINGER, T. A., G. GALFRE, D. S. SECHER, and C. MILSTEIN. 1979. Mac-l: a differentiation antigen identified by a monoclonal antibody. Eur. J. Immunol. 9: 301. 15. UNKELESS, J. C. 1979. Characterization of a monoclonal antibody directed against mouse macrophage and lymphocyte Fe receptors. J. Exp. Med. 150: 580. 16. LEDBETTER, J. A., and 1. A. HERZENBERG. 1979. Xenogeneic monoclonal antibodies to mouse lymphoid differentiation antigens. Immunol. Rev. 47: 63. 17. Ho, M. K., T. A. SPRINGER. 1982. Mac-2, a novel 32,000 Mr macrophage subpopulationspecific antigen defined by monoclonal antibody. J. Immunol. 128: 1221. 18. Ho, M. K., T. A. SPRINGER. 1983. Tissue distribution, structural characterization, and biosynthesis of Mac-3, a macrophage surface glycoprotein exhibiting molecular weight heterogeneity. J. BioI. Chern. 258: 636. 19. SNEDECOR, G. W., W. G. COCHRAN. 1973. Statistical methods. The Iowa State Press, Cambriss, Amis, Iowa, USA. 20. SIEGEL, S. 1956. Nonparametric Statistics. McGraw-Hill, International Book Company, Tokyo, Japan. 21. BASCH, R. S., and J. W. BERMAN. 1982. Thy-l determinants are present on many murine hematopoietic cells other than T cells. Eur. J. Immunol. 12: 359. 22. BOSWELL, H. S., P. M. WADE Jr., and P. J. QUESENBERRY. 1984. Thy-l antigen expression by murine high-proliferative capacity hematopoietic progenitor cells. I. Relation between sensitivity to depletion by Thy-l antibody and stem cell generation potential. J. Immunol. 133: 2940. 23. VAN FURTH, R., P. H. NIBBERING, J. T. VAN DISSEL, and M. M. C. DiESSELHOFF.DEN DULK. 1985. The characterization, origin, and kinetics of skin macrophages during inflammation. J. Invest. Dermatol. 85: 398. 24. VAN FURTH, R. 1987. Development and distribution of mononuclear phagocytes in the normal steady state and inflammation. In: Basic Principles and Clinical Correlates. J. I. GALLIN,J. M. GOLDSTEIN, and R. SNIJDERMAN (Eds.). Raven Press, New York (in press). Dr. RALPH VAN FURTH, University Hospital, Department of Infectious Diseases, Building 1, C5-P, P.O. Box 9600, 2300 RC Leiden, The Netherlands