The complement receptor 3 (CR3; CD11b/CD18) is present exclusively on leukocytes, particularly on NK cells, monocytes and polymorphonuclear neutrophils. Approximately 10% of peripheral T lymphocytes and, as we found now mainly CD8+ cells, expressed CD11b. Upon stimulation, however, expression of CD11b was up‐regulated also on CD4+ cells. Stimulation of T cells either bycross‐linked anti‐CD3 and IL‐2 or by mononuclear cells and mitogen yielded up to 28% CD11b+ T cells. The majority of CD11b+ T cells also expressed CD56. T cell lines established from healthy donors were also found to express CR3. When restimulated up to 90% of cells became positive for CD11b making those cells an ideal tool for studying the functional role of CD11b. Antibodies to CD11b and bona fide ligands for the complement receptor inhibited the anti‐CD3‐induced T cell proliferation and as well as IL‐2 release . In contrast, proliferation of a CD11b– T cell line was not inhibited. Taken together, our data indicate an activation‐dependent expression of the complement receptor on T cells and suggest a regulatory function.

Polymorphonuclear neutrophils (PMNs) produce an abundance of bactericidal and cytotoxic molecules consistent with their role as first-line defense against bacterial infection. PMNs, however, also cause efficient cellular cytotoxicity when targeted through Fc receptors to appropriate antibody-coated target cells. Although this so-called antibody-dependent cellular cytotoxicity (ADCC) was described many years ago, the mechanism of killing is still elusive. We now have found that PMNs contain perforin and granzyme B, the 2 molecules known as the cytotoxic entity of natural killer cells and of cytotoxic T lymphocytes as well. Lysates of PMNs were lytic for chicken erythrocytes in a time-, temperature-, and Ca 2؉ -dependent manner. Moreover, apoptosis of Jurkat cells was induced, consistent with the observation that the PMN lysates contain enzymatically active granzyme B. Taken together, our data provide evidence for the presence of perforin and granzyme B within the cytotoxic arsenal of PMNs.

SUMMARYUpon cultivation with interferon-g (IFN-g ) and granulocyte/macrophage-colony stimulating factor (GM-CSF) polymorphonuclear neutrophils (PMN) acquire characteristics of dendritic cells, including expression of major histocompatibility complex (MHC) class II antigens, of the co-stimulatory antigens CD80, CD86 and of CD83, the latter considered to be specific for dendritic cells. Dendritic-like PMN were also able to present to T cells antigens in a MHC class II-restricted manner. To assess whether dendritic-like PMN are also generated in vivo , cells of patients with acute bacterial infections and of patients with chronic inflammatory diseases (primary vasculitis) were tested. During acute infection up to 80% of PMN acquired CD83, but remained negative for MHC class II, CD80 or CD86. PMN of patients with primary vasculitis expressed MHC class II antigens, CD80 and CD86, but not CD83, indicating that up-regulation of MHC class II and of CD83 are not necessarily linked to each other. Indeed, parallel studies with PMN of healthy donors showed that while IFN-g and granulocyte/macrophage colony stimulating factor (GM-CSF) induced both, MHC class II and CD83, tumour necrosis factor (TNF)-a selectively induced de novo synthesis of CD83. The function of CD83 on PMN is still elusive. A participation in the MHC class II-restricted antigen presentation could be ruled out, consistent with the segregation of MHC class II and CD83 expression. Regardless, however, of its function, CD83 expression could serve as a marker to differentiate between acute and chronic inflammation.

In polymorphonuclear neutrophils (PMN) CD14, one of the receptors for lipopolysaccharides (LPS) is stored intracellularly as a preformed protein, with only few receptors expressed on the surface. We now report that in patients with severe bacterial infections, CD14 expression is profoundly upregulated, as is CD64 (FcgammaRI), the high-affinity receptor for IgG, whereas CD16 (FcgammaRIII) was partly lost from the surface. To further analyze regulation of these receptors, PMN of healthy donors were exposed to low doses of LPS. By brief exposure (10-120 min) to LPS, CD14 was transferred to the surface in a cytochalasin B-sensitive manner, as were CD16 and CD64. Prolonged culture (up to 48 h) resulted in a further upregulation of CD14, sustained expression of CD64, and profound decline of CD16, yielding a similar pattern of receptor expression as seen in the patients. Subsequent studies revealed that LPS induced de novo synthesis of CD14: the increase of surface expression could be inhibited by cycloheximide and by interfering with a known LPS-induced signaling event, the translocation of NFkappaB. Moreover, an up to 10-fold increase of specific mRNA was seen, as was incorporation into CD14 of 35S-methionine. The de novo synthesis prolonged expression of CD14, whereas the CD16 expression declined, generating a PMN phenotype characteristic for severe infection and indicative of escape from apoptosis of a PMN subpopulation.

The pathogenesis of posttraumatic osteomyelitis, one of the major complications after orthopedic surgery, is not yet understood. Formation of bacterial biofilms on the implant is presumed, conferring resistance to antibiotic therapy and probably also to the host defense mechanisms. In that context, the polymorphonuclear neutrophils (PMN) having infiltrated the infected site were recovered and characterized phenotypically and functionally. Loss of CD62L and upregulation of CD14 were seen, as was expression of CD83. Expression of the latter is dependent on de novo protein synthesis and thus is indicative of an extended life span and a transdifferentiation of the PMN at the infected site. The infiltrated PMN had lost their chemotactic activity, whereas the capacity to produce superoxides was preserved and in some patients even enhanced. In vitro experiments done in parallel showed that long-term culture with interferon-gamma resulted in similar alterations of PMN: loss of chemotactic activity, whereas other functions of PMN, such generation of superoxides and phagocytosis of opsonized bacteria, were preserved or even enhanced. The loss of the migratory capacity of PMN having already emigrated from the blood vessel to the infected site is not expected to affect the host defense negatively. Assuming, however, that bacteria are organized as a biofilm and that infiltration into this biofilm is required for phagocytosis of the bacteria, our data could to some extent explain why despite being activated, the PMN are not able to control the infection. By releasing their cytotoxic, proteolytic, and collagenolytic potential, PMN might instead contribute to tissue destruction and eventually to osteolysis.