Requirement for cell-to-cell contact for the immunosuppressive activity of human alveolar macrophages.

The ability of alveolar macrophages (AM) obtained by bronchoalveolar lavage of healthy volunteers to suppress T lymphocyte responses to the mitogen phytohemagglutinin (PHA) in vitro was investigated. AM but not monocytes (MN) inhibited responses of peripheral blood mononuclear cells (PBMC) to PHA as measured by incorporation of [H-3]thymidine ([H-3]TdR) and interleukin-2 (IL-2) expression. Supernatants of AM generated for various periods and with various concentrations of cells did not, however, inhibit PBMC responses to PHA. To examine the role of cell contact in the inhibitory activity of AM, AM or MN were added to PBMC in 6-well plates either directly (in co-culture) or separated by a 0.45-mu-m filter. MN did not inhibit PBMC blastogenic responses under either condition. AM at a 1:2 ratio with PBMC inhibited blastogenesis by 75 +/- 11% (mean +/- SD, n = 3, P < 0.01) when cultured directly with PBMC but had no inhibitory effect on blastogenesis when physically separated from target PBMC. AM in coculture with PBMC also inhibited PHA-stimulated IL-2 production by 70% but did not inhibit IL-2 production when AM were separated from PBMC in dual chambers. To assess the role of the cell surface in the inhibitory activity of AM, AM and MN were fixed with 2% paraformaldehyde. Neither fixed nor unfixed MN inhibited PBMC blastogenic responses, but both fixed and unfixed AM inhibited responses similarly (77 to 95%). The inhibitory activity of AM on blastogenic responses of PBMC was not due to cytotoxicity of AM since PBMC routinely excluded trypan blue dye and released similar levels of lactate dehydrogenase activity whether AM or MN were present in culture. These experiments suggested that the cell surface was required for suppression by AM of mitogenic responses to PHA. Because a major portion of cell membranes is lipid, as an initial step in characterization of the suppressive activity, organic extracts of AM prepared with chloroform/methanol/water were tested and found to inhibit the PBMC response to PHA in a manner similar to that of the intact cells. Extracts of AM inhibited the response to a 3- to 10-fold greater extent than did extracts of MN. The extracts of AM differed qualitatively from the extracts of MN in that they contained phosphatidylglycerol. Phosphatidylglycerol (100 to 400-mu-g/ml) also directly inhibited the PBMC response to PHA. The extracts of AM also contained hydrophobic proteins. The inhibitory activity of the extract was unaffected by heat and trypsin but was partially diminished by treatment with pronase. Thus, AM inhibited lymphocyte responses to PHA; the suppressive activity was, in part, due to a cell-associated hydrophobic material possibly comprised of both lipid and protein; and either cell contact or close proximity between lymphocytes and AM was required for mediation of the suppressive activity.