Morphometric comparisons of rat alveolar macrophages, pulmonary interstitial macrophages, and blood monocytes.

Pulmonary interstitial macrophages (IM) account for a substantial fraction of the total pulmonary macrophage (PM) population in the mammalian lung, with the remaining balance of extravascular mononuclear phagocytes being mainly alveolar macrophages (AM). Unlike the AM that can be harvested readily by bronchoalveolar lavage, the lung's IM subpopulation of PM has been characterized less well, primarily because of its relative inaccessibility. Recently we developed a method to isolate viable IM from rat lungs using an Fc gamma receptor affinity technique in conjunction with multiparameter flow cytometry. Using this approach, we undertook the present investigation to characterize quantitatively the structural features of the IM and to compare the morphologic attributes of this subpopulation of PM to those of flow cytometrically sorted AM and blood monocytes (BM). Measured or calculated parameters for each population included mean cellular equivalent circular diameter, cell area and volume, and nuclear, mitochondrial, cytoplasmic, and lysosomal volume densities in each cell type. Lysosomal volume densities were subcategorized further into primary lysosomes, small secondary lysosomes, large secondary lysosomes, lipid droplets, and vacuoles. Additionally, the shape, form, and surface irregularity of the cells and various subcellular components were determined. Comparisons of the size and other structural features of the AM, IM, and BM have indicated that (1) the morphologic phenotypes of these three populations of mononuclear phagocytes distinctly differ from one another, (2) the IM and BM are morphologically and morphometrically more akin to one another than they are to AM, and (3) the IM are more similar to the AM than are the BM. These findings suggest that the IM may represent a transitional stage of maturation between BM and AM. Our findings, however, do not rule out the possibility that at least some of the lung's IM are a discrete, BM-independent population of macrophages.     

Tumor killing by human alveolar macrophages and blood monocytes. Decreased cytotoxicity of human alveolar macrophages

Tumor killing by human alveolar macrophages (AM) might be an important mechanism of pulmonary defense against neoplastic disease. We compared AM and blood monocytes (Mo) for the ability to kill 2 neoplastic targets, A549 human lung adenocarcinoma cells and P815 mastocytoma cells. Blood monocytes were able to kill both targets, whereas AM killed neither. Tumor killing by Mo was spontaneous and was not increased by incubation with lipopolysaccharide. Because the P815 target is highly sensitive to lysis by hydrogen peroxide (H2O2), it afforded the opportunity to compare AM and Mo for the ability to kill tumors by the production of toxic oxygen compounds. Comparable amounts of superoxide anion were produced by AM and Mo after stimulation with phorbol myristate acetate. However, luminol-enhanced chemiluminescence of AM was far less than that of Mo, suggesting that AM could not utilize the myeloperoxidase-H2O2-halide ion system for tumor killing. The addition of exogenous peroxidase to cultures of AM and P815 cells enabled AM to kill this tumor cell. Our results suggest that as Mo mature into AM, their ability to kill tumor cells declines and that AM may be unable to kill H2O2-sensitive tumors because of a loss of myeloperoxidase during maturation.     

Phenotypic characterisation of alveolar macrophages and peripheral blood monocytes in COPD.

Alveolar macrophages (AM) participate actively in the inflammatory response that characterises chronic obstructive pulmonary disease (COPD). The present study investigated potential changes in AM phenotypes in patients with COPD. Using flow cytometry, the surface expression of receptors implicated in phagocytosis (CD44, CD36, CD51, CD61, CD14), antigen-presenting capacity (human leukocyte antigen (HLA)-DR), costimulatory molecules (CD80, CD86, CD40) and complement receptor type 3 were assessed in AM from 18 patients with COPD, 14 smokers with normal lung function and nine nonsmokers. When compared to smokers with normal lung function and nonsmokers, the surface expression of HLA-DR and CD80 was lower in AM of patients with COPD. In addition, these patients had a higher percentage of AM with a low level surface expression of CD44. There did not appear to be any difference in the other receptors studied in AM between the three groups. The expression of all these receptors in peripheral blood monocytes also did not differ between groups. In conclusion, these observations suggest that the cell-mediated immune function of alveolar macrophages can be reduced in chronic obstructive pulmonary disease, and that this is a local rather than a systemic event.     

Carbohydrate differences in HLA-DR molecules synthesized by alveolar macrophages and blood monocytes

Two-dimensional gel electrophoresis was used to compare immunoprecipitated, radiolabeled HLA-DR molecules synthesized by alveolar macrophages (AM) and blood monocytes (BM) of the same healthy, nonsmoking volunteers. No charge differences were noted by isoelectric focusing. Molecular weight analysis of HLA-DR from AM revealed a complex pattern of radioactive spots differing slightly in molecular weight. This pattern was apparent in both the alpha and the beta chains. In contrast, a simpler pattern without the slight molecular weight differences was noted in HLA-DR from BM. The differences were eliminated by labeling AM and BM in the presence of tunicamycin, a specific inhibitor of protein-carbohydrate linkages of the N-glycosidic type. Thus, it appears that neutral, N-linked carbohydrate moieties account for the differences in HLA-DR molecules between AM and BM. The differential glycosylation of HLA-DR molecules in AM and BM may relate to mononuclear phagocyte development as well as antigen-presenting function.     

The binding of monomeric IgG to human blood monocytes and alveolar macrophages

Macrophage receptor sites for IgG are important in the immune clearance of particles both from the blood and lung. We studied the number, affinity, and density of binding sites for monomeric IgG on human blood monocytes and alveolar macrophages. Monocytes and alveolar macrophages had a similar affinity for monomeric IgG at 37 degrees and 4 degrees C. The half-time for dissociation of the IgG-receptor complex was also similar for both cells. However, alveolar macrophages expressed approximately 5-fold more IgG binding sites than monocytes at both 37 degrees and 4 degrees C. Nevertheless, when cell surface area was estimated, these cells expressed a similar density of IgG binding sites (monocytes = 110 +/- 14.8 IgG binding sites/square micron; pulmonary macrophages = 138 +/- 46.9 IgG binding sites/square micron; p greater than 0.50). Gamma interferon increased the number and density of monocyte binding sites for monomeric IgG by 162 +/- 89%. Furthermore, patients with sarcoidosis, a disorder in which gamma interferon is spontaneously elaborated, expressed a similar increase in the number of IgG binding sites per monocyte, from 24,968 +/- 1,361 for normal subjects to 44,860 +/- 6,652 for patients with sarcoidosis. However, alveolar macrophages from 5 patients with sarcoidosis expressed a normal number of IgG binding sites. These data suggest that there is no major alteration in the Fc(IgG) receptor as monocytes differentiate into alveolar macrophages. Both gamma interferon treatment and sarcoidosis are associated with enhanced expression of the Fc(IgG) receptor on monocytes.     

Plasminogen activation by blood monocytes and alveolar macrophages in primary pulmonary hypertension.

The pathophysiology of primary pulmonary hypertension (PPH) remains poorly understood. Vascular wall remodeling and endothelial dysfunction reflected by modifications in plasma fibrinolytic proteins and von Willebrand factor have been well documented in PPH. We hypothesize that endothelial mediators, produced in excess in PPH patients, may stimulate migrating mononuclear cells and thereby modulate alveolar macrophage function; in particular, the plasminogen activation system. Components of the fibrinolytic system were therefore studied in plasma, blood monocytes and alveolar macrophages obtained from bronchoalveolar lavage in 10 patients with PPH and in four controls. Compared with controls, PPH patients had elevated plasma levels of tissue-type plasminogen activator (15.6 +/- 9.9 versus 5.5 +/- 3 ng/ml) and plasminogen activator inhibitor-1 (27.8 +/- 23 versus 16.4 +/- 12 ng/ml). In contrast, binding and activation of plasminogen by single-chain urokinase-type plasminogen activator (scu-PA) at the surface of blood monocytes and alveolar macrophages were not different from those of control values. Dissociation constants (K(d)) for binding of scu-PA and plasminogen to alveolar macrophages were similar in both PPH (4.7 +/- 1.5 and 0.88 +/- 0.3 micromol/l, respectively) and control (6.7 +/- 0.1 and 1.02 +/- 0.12 micromol/l, respectively) groups. These results indicate that in PPH patients the fibrinolytic activity of alveolar macrophages is normal, whereas endothelial fibrinolytic proteins are abnormally elevated in plasma.     

A controlled study of oxygen metabolite release by alveolar macrophages from children with interstitial lung disease

The ability of alveolar macrophages (AM) to release O2 metabolites was studied in 8 children with interstitial lung disease (ILD), and in 11 children without lung parenchyma disorder. AM were collected by bronchoalveolar lavage. The experiments were performed on unstimulated AM and on AM stimulated by phorbol myristate acetate (PMA) or zymosan. Our results indicated that, with or without triggering agent, the amount of O2 metabolites release was a linear function pattern with time. The accumulation of superoxide anion (O2-) and hydrogen peroxide (H2O2) into the extracellular medium differed depending on the triggering agent used: with PMA, the amount of O2- released was threefold the amount of H2O2 detected in the medium, whereas with zymosan the O2- accumulation was tenfold higher than the amount of H2O2 measured. In patients with ILD, a significant increase in the amount of H2O2 release was observed for both unstimulated and stimulated AM (p less than 0.001). In this group, the measurement was repeated after a 2-month steroid treatment: prednisone had markedly improved the clinical, radiologic, and functional status of the patients, and this improvement was in good correlation with the decrease of O2 metabolite production. The amount of H2O2 release in each case was within the range of control values. Evaluation of O2 metabolite release by AM could be a useful parameter in the assessment of the activity of ILD.     

Expression of the CD11/CD18 cell surface adhesion glycoprotein family and MHC class II antigen on blood monocytes and alveolar macrophages in interstitial lung diseases

The expression of molecules of the CD 11/CD18 cell surface adhesion glycoprotein family and HLA/DR antigen was studied on peripheral blood monocytes (PBM) and alveolar macrophages (AM) in bronchoalveolar lavage (BAL) fluid from patients with sarcoidosis, idiopathic pulmonary fibrosis (IPF), and extrinsic allergic alveolitis (EAA). Patients with these interstitial lung diseases showed increased numbers of macrophages in BAL fluid. This was probably caused by an increased influx of PBM to the alveoli since the numbers of cells with a monocytic morphology were also significantly increased in BAL samples from patients with interstitial lung disease, most prominently in IPF and EAA.The increased influx of PBM into the alveoli in patients with interstitial lung diseases was not reflected by an increased expression of the CD11/CD18 leukocyte function antigens on PBM.In healthy volunteers as well as in those with sarcoidosis, IPF, and EAA, the percentages of AM positive for CD11b (the C3bi complement receptor) and CD11c were lower than among PBM. This indicates that the expression of these cell surface adhesion molecules is downregulated during maturation and migration of PBM to the alveoli. The absolute numbers of AM positive for CD11b were increased in BAL fluid of IPF and EAA patients compared to healthy volunteers. EAA patients also showed increased absolute numbers of AM positive for CD11a and CD11c. This differentially increased expression of these leukocyte function antigens on AM suggests the influence of locally produced cytokines. Offprint requests to: H. C. Hoogsteden     

Thalidomide reduces IL-18, IL-8 and TNF-alpha release from alveolar macrophages in interstitial lung disease.

Thalidomide exhibits diverse actions of anti-inflammation, immunomodulation and anti-angiogenesis. The efficacy of thalidomide treatment in sarcoidosis with lupus pernio is thought to be due to inhibition of tumour necrosis factor (TNF)-alpha. The mechanisms that underlie the properties of thalidomide are still unclear in interstitial lung disease. The current authors investigated the potential inhibitory effects of thalidomide at concentrations of 0.1, 0.01 and 0.001 mM on the production of transforming growth factor-beta, TNF-alpha, interleukin (IL)-1beta, IL-6, IL-8, IL-10, IL-12p70, IL-12p40 and IL-18 by alveolar macrophages from bronchoalveolar lavage in patients with sarcoidosis (n = 8), hypersensitivity pneumonitis (HP; n = 8) and idiopathic pulmonary fibrosis (IPF; n = 12). In sarcoidosis and HP patients, thalidomide induced a dose-dependent, partial suppression of lipopolysacchride (LPS)-stimulated TNF-alpha, IL-12p40 and IL-18 release. At the highest thalidomide concentration (0.1 mM), LPS-stimulated IL-8 production was also suppressed. In IPF patients, although spontaneous production of TNF-alpha, IL-12p40, IL-18 and IL-8 was lower than in sarcoidosis and HP patients, with LPS stimulation the cytokines were significantly elevated and also partially inhibited by thalidomide. In conclusion, thalidomide has the potential to improve the therapeutic regimens for sarcoidosis, hypersensitivity pneumonitis and idiopathic pulmonary fibrosis by reducing tumour necrosis factor-alpha, interleukin-12p40, interleukin-18 and interleukin-8 production.     

Dyscoordinate expression of tumor necrosis factor-alpha by human blood monocytes and alveolar macrophages

Previous studies have shown that human alveolar macrophages produce less interleukin-1 (IL-1) in response to lipopolysaccharide (LPS) than do their precursors, blood monocytes. The purpose of this study was to compare the capacities of alveolar macrophages and blood monocytes to synthesize tumor necrosis factor (TNF) in response to LPS. Alveolar macrophages were obtained by bronchoalveolar lavage of healthy nonsmoking subjects, and blood monocytes were obtained by adherence of mononuclear cells to plastic. TNF activity was measured in supernatants and cell lysates as cytotoxicity to L929 fibroblasts (uptake of neutral red at 570 nm). TNF activity of alveolar macrophages stimulated at 10(6) cells/ml with LPS (10 micrograms/ml) for 16 h was 596 +/- 367, and of blood monocytes it was 60 +/- 84 U/ml (mean +/- SD, p less than 0.005). At no concentration of LPS and at no period of stimulation did alveolar macrophages express less TNF activity than did blood monocytes. In concurrent experiments, supernatants of LPS-stimulated alveolar macrophages contained less IL-1 activity than did blood monocytes. Lysates of both cell types contained less than 20% of total TNF activity. The TNF activity of LPS-stimulated alveolar macrophages was neutralized greater than 99% by monoclonal antibody to TNF-alpha; control monoclonal antibody OKT3 had no effect. Next, alveolar macrophages and blood monocytes were biosynthetically labeled with [3H]leucine during incubation with LPS; supernatants were immunoprecipitated with anti-TNF, and precipitates were electrophoresed on polyacrylamide gels. Autoradiographs indicated that immunoreactive TNF was produced by both blood monocytes and alveolar macrophages and that the relative molecular weights were identical (17,000).(ABSTRACT TRUNCATED AT 250 WORDS)     

Pulmonary macrophages: alveolar and interstitial populations

The sizes of the alveolar macrophage (AM) and interstitial macrophage (IM) populations in the lungs of adult Fischer-344 rats were determined during steady state. AM labeled with opsonized erythrocytes during an in situ phagocytic assay were lavaged from excised lungs. The lungs were then dispersed into single-cell suspensions with collagenase and mechanical agitation, and the remaining mononuclear phagocytes were identified following a second labeling step. The size of the AM population was 1.3 X 10(7) cells, or approximately equal to 3% of the total lung cell population. The AM were negative for cytoplasmic myeloperoxidase granules. The size of the IM population was 8 X 10(6) cells, or approximately equal to 2% of the total lung cell population. IM were also negative for myeloperoxidase and, like AM, demonstrated marked Fc gamma R-mediated phagocytic activity. The high cell yields (approximately equal to 4 X 10(8) cells/lung; viability, greater than 85%) and the percentages of type II cells (11%) and ciliated epithelial cells (less than 0.5%) indicated the enzymatic dispersion method resulted in a highly efficient and representative sampling of the lung parenchyma. The collagenase method used in this study to disperse the lung cells into single-cell suspensions, in conjunction with additional cell separation techniques, may be of potential use for isolating enriched populations of IM, as well as other lung cell types, for in vitro study.     

Respiratory syncytial virus infection of human cord and adult blood monocytes and alveolar macrophages

We studied the permissiveness of human leukocytes, blood monocytes, alveolar macrophages, and cord blood monocytes to infection with respiratory syncytial virus (RSV). Specific immunofluorescence was used to determine the percentage of infected leukocytes. The results indicated that monocytes were the most susceptible human leukocyte to in vitro infection with RSV. Polymorphonuclear leukocytes demonstrated no specific fluorescent staining after 24 h of exposure to RSV, whereas peripheral blood nonadherent mononuclear cells demonstrated a low percentage of positive cells, with a mean of 6 +/- 1% SE. In contrast, 37 +/- 5% of monocytes expressed RSV antigen after viral exposure. Exposure of monocytes to lipopolysaccharide (LPS) for 1 h prior to RSV increased the percentage of infected cells to 48 +/- 6% and stimulated their secretion of prostaglandin E2 (PGE2) and alpha tumor necrosis factor (TNF). Intrinsic mononuclear phagocytic factors influencing the permissiveness to RSV were studied by determining infection of adult and cord blood and alveolar mononuclear phagocytes (MP). Alveolar and blood MP simultaneously isolated from adult donors were similarly infected by RSV, which varied with the viral dose. Cord blood MP were more susceptible to RSV infection than were adult MP, 58 +/- 9% infected versus 37 +/- 5%, respectively (p less than 0.05). Treatment with LPS for 1 h prior to RSV exposure did not increase infection of cord blood MP as seen with adult blood MP. However, LPS can induce human monocytes to secrete cytokines with antiviral activity, and our results indicate that both gamma interferon and TNF, independently or in combination, prevented infection of monocytes in a dose-dependent manner.     

Antigen-presenting capacity of alveolar macrophages and monocytes in pulmonary tuberculosis

Using purified protein derivative of tuberculin (PPD) as an antigen, antigen-presenting capacity by monocytes (Mo) and alveolar macrophages (AM) was determined in 17 patients with pulmonary tuberculosis and nine healthy controls. All of the patients and healthy controls were positive for PPD skin test. Although Mo obtained from both the control and tubercular subjects revealed antigen-presenting capacity to autologous blood T-lymphocytes, no significant difference was observed between the two groups. In contrast, AM obtained from the tubercular patients, but not from the controls, showed antigen-presenting capacity to autologous blood T-lymphocytes and to lung T-lymphocytes. No significant difference was shown in HLA-DR antigen expression on AM between the control and tubercular patients. Besides, the exogenous addition of interleukin-1 (IL-1) did not induce antigen-presenting capacity by AM obtained from the controls. These results suggest that neither increased HLA-DR antigen expression on AM nor an increased release of IL-1 from AM is responsible for the enhanced antigen-presenting capacity in tuberculosis.     

Interleukin-10 secretion by alveolar macrophages and monocytes in sarcoidosis

BACKGROUND: Alveolitis and the production of proinflammatory cytokines are known features of sarcoidosis. Because of the usually spontaneous resolution of alveolitis despite local secretion of mediators causing inflammation and granuloma formation, we hypothesized that downmodulating mechanisms such as anti-inflammatory cytokines might be involved in this process. OBJECTIVE: Investigation of the secretion of the macrophage deactivating cytokines interleukin-10 (IL-10) and transforming growth factor-beta (TGF-beta) by alveolar macrophages in untreated sarcoidosis of the lung. METHODS: Fourteen consecutive and untreated patients with pulmonary sarcoidosis and 18 volunteers underwent bronchoscopy. Alveolar macrophages (AM) were obtained by bronchoalveolar lavage and the secretion of IL-10 and TGF-beta was studied. RESULTS: Spontaneous IL-10 production by AM was found in 6 of 14 patients and in 2 of 18 controls. The IL-10 level of lipopolysaccharide-stimulated AM was significantly higher in patients. Monocytes secreted significantly more IL-10 than AM, but there was no difference between sarcoid and control monocytes. No difference was found in the secretion of TGF-beta between patients and controls. CONCLUSION: Increased local secretion of IL-10 - but not TGF-beta - may represent a downmodulating mechanism involved in the spontaneous resolution of alveolitis in sarcoidosis.     

Proliferation activity and bacteriostatic potential of human blood monocytes,macrophages in pleural effusions,ascites, and of alveolar macrophages

Summary Human blood monocytes, macrophages from pleural effusions, ascites, and alveolar macrophages obtained by bronchopulmonary lavage were investigated. The proliferative activity of these cells was determined by the³H-thymidine labeling index in vitro (³H-TDR L.I.). The bacteriostatic capacity was measured by the potential of the cells to block DNA-synthesis of proliferating Escherichia coli after phagocytosis.In most cases³H-TDR L.I. of blood monocytes, macrophages from pleural effusions and ascites was less than 1 %. However, macrophages of some patients with neoplastic diseases exhibited³H-TDR L.I. between 4.0–9.6%. This probably reflected the action of factors, possibly lymphokines, which stimulated macrophage proliferation. In contrast, alveolar macrophages seemed to have almost totally lost their proliferative potential. The bacteriostatic capacity of blood monocytes proved to be significantly lower than that of macrophages. This demonstrates the functional immaturity of blood monocytes. In all types of macrophages investigated the bacteriostatic power was very high. No differences could be detected either between macrophages of different sources or between macrophages of benign, inflammatory, or malignant diseases.     

Chemiluminescence and antibody-dependent, cell-mediated cytotoxicity between human alveolar macrophages and peripheral blood monocytes in smokers, nonsmokers, and lung cancer patients

Some evidence suggests a capability of peripheral blood monocytes to destroy tumor cells, while this ability by human alveolar macrophages, the main defense cells in the alveoli, is still debatable. We measured the chemiluminescence and antibody-dependent, cell-mediated cytotoxicity in the PBMs and HAMs of 12 lung cancer patients and 20 healthy subjects; the latter included ten smokers and ten nonsmokers. The PBMs were prepared by using a Ficoll-Hypaque density gradient, then separated by Petri-dish adherence. The HAMs were taken during the bronchoalveolar lavages. The chemiluminescence in the HAMs of smokers was significantly higher than nonsmokers, (p less than 0.05), which did not occur in the PBMs. Chemiluminescence in HAMs from the lung cancer patients was also significantly higher than the smoker control subjects (p less than 0.01). However, the lung cancer patients had significantly lower ADCC activity than the smokers in the control group (4.52 +/- 2.96 vs 8.27 +/- 2.83 percent; p less than 0.05). The chemiluminescence in the PBMs showed no statistical difference between the lung cancer patients and smoker control, but PBMs of the lung cancer patients had significantly lower ADCC activity than the smoker normal control group. The HAMs from lung cancer patients produced more superoxide anion, for an increased chemiluminescence reaction was noted, although ADCC activity was lower than in smokers, ie, HAMs were ineffective in killing tumors. Environmental factors such as cigarette smoking affect HAM's function by causing an increase of superoxide anion production. The chemiluminescence and ADCC activity in the PBMs does not always correspond to the HAMs findings. These results suggest that PBMs can not accurately reflect or predict the HAMs' function in lung diseases.     

Abnormalities in pathways of alveolar fibrin turnover among patients with interstitial lung disease

Fibrin deposition is prominent in the histopathologic features of chronic interstitial lung disease. Human alveolar macrophages can potentially modulate this process because normal macrophages synthesize and express the initial enzymes of both coagulation and fibrinolytic pathways. In the present study, we examined the cell-associated procoagulant activity of macrophages lavaged from patients with sarcoidosis (n = 14) or idiopathic pulmonary fibrosis (n = 13) and compared the enzyme activities with that of a group of normal volunteers (n = 16). Cells from sarcoid patients had a mean (+/- 1 SD) tissue factor activity of 1,491 +/- 2,160 units/5 X 10(5) cells, as compared with a mean of 480 units (range, 140 to 1,000 units) for normal control subjects. The same cells had a mean plasma Factor VII equivalent of 4.7 ng/10(6) cells, as compared with 0.81 ng/10(6) cells (range, 0.2 to 2.0 ng) for the normal control subjects. The enhanced activity correlated with disease activity as judged by radiographic stage: only patients with Stage II or Stage III disease had consistently elevated procoagulant activity. There was no correlation of procoagulant activity with the percentage of lymphocytes in the alveolar fluid. Cells from patients with idiopathic pulmonary fibrosis also had increased tissue factor (mean, 2,980 +/- 2,619 units) but less consistently elevated Factor VII. There was considerable variation in both procoagulant activity and cell differentials between lavage sites in 10 patients in whom 2 separate lobes were studied concurrently. In addition, we examined the plasminogen activator (PA) activities of lavaged cells and concentrated alveolar fluids.(ABSTRACT TRUNCATED AT 250 WORDS)     

Alveolar macrophages, blood monocytes, and density-fractionated alveolar macrophages differ in their ability to promote lymphocyte proliferation to mitogen and antigen

We compared the relative abilities of alveolar macrophages (AM), blood monocytes (BM), and density-fractionated AM to support antigen- and mitogen-induced proliferation of autologous T cells in normal volunteers. The AM were able to promote the T-cell proliferative response to antigen, but they did so less effectively than did the BM. Density-fractionated AM were heterogeneous in their ability to support T-cell responses. More proliferation occurred with the densest AM than with the least dense AM. In contrast to antigen-induced responses, mitogen-induced responses were supported more effectively by AM and density-fractionated AM than by BM. The ability of AM, density-fractionated AM, and BM to support T-cell responses did not correlate with surface expression of class II major histocompatibility determinants. Addition of purified IL-1 resulted in a partial restoration of T-cell proliferation when low-density AM were used but no augmentation when unfractionated AM were used. This suggests that reduced IL-1 activity may partially explain the decreased ability of low density AM to promote T-cell responses, but that other processes may also contribute to differences in accessory cell function among AM, BM, and density-fractionated AM.