Pneumocystis carinii Glycoprotein A Stimulates Interleukin‐8 Production and Inflammatory Cell Activation in Alveolar Macrophages and Cultured Monocytes

Pneumocystis carinii pneumonia (PCP) remains a major cause of morbidity and mortality in patients with the Acquired Immune Deficiency Syndrome (AIDS) as well as other immunocomprornised patients. The pathogenesis of PCP is poorly understood, but emerging evidence suggests that inflammation plays a role in the production of cellular injury. The success of adjunctive corticosteroids for PCP, the identification of bronchoalveolar lavage (BAL) neutrophilia as a marker of more severe disease, and the demonstration of inflammatory mediators in BAL fluid of patients with PCP are all evidence that despite the lack of cellular infiltration seen in biopsies of patients with PCP, inflammatory mechanisms are at work [10,14,17]. Recently, in an effort to identify neutrophil chemotactic factors responsible for BAL neutrophilia, we demonstrated that interleukin8 (ILS) was elevated in BAL fluid of patients with PCP and correlated with the severity of the disease [14]. A subsequent study corroborated this, and suggested that IL8 production was a target of corticosteroids' ameliorative effect [2]. Specific triggering mechanisms for IL8 production and inflammation in PCP, however, are not known. Recent work suggests that glycoprotein A (gpA), an abundant, immunogenic surface antigen of P. carinii , may be the trigger for IL8 production and thus ultimately for neutrophil accumulation and lung injury in PCP. gpA is a high-mannose glycoprotein believed to be involved in pathogen-host cell interactions, including binding to alveolar macrophages and alveolar epithelial cells [4,5,7,11,13,20]. These interactions appear to involve mannoseand fibronectin-mediated mechanisms [4,13,19,21]. Finally, passive immunoprophylaxis with anti-gpA antibody protects against or ameliorates experimental PCP, further highlighting gpA's importance in pathogenesis [6]. In work described here, we directly tested the ability of gpA to stimulate ILS production and activate inflammatory cells. We used careful controls and highly purified gpA in order to allow observed effects to be attributed to gpA rather than to contaminating proteins or endotoxin (LPS). MATERIALS AND METHODS. Rat-derived P. carinii gpA (rgpA) was prepared from the lungs of corticosteroid-treated rats by homogenization of infected rat lungs followed by filtration and density gradient centrifugation on Ficoll-Hypaque as previously described [ 151. Human-derived gpA (hgpA) was prepared similarly, but without the density gradient step; the yield of hgpA is much lower. After treatment of cell pellets with lyticase to release cellsurface components, gpA was prepared using HPLC with sequential size exclusion and ion exchange columns, followed b y concentration and desalting [ 161. Contaminating LPS was removed from all gpA preparations by triton X-114 extraction [I]. This step is critical to attributing modulatory effects of gpA to the protein, rather than to contaminating LPS. The detergent extraction used here reduces LPS contamination of gpA by greater than 1000-fold. Absence of LPS from gpA preparations was confirmed using the Limulus amebacyte lysate assay. Material prepared as detailed here produces a single band by SDS-polyacrylamide gel electrophoresis 1161 and, combined with the detergent extraction, represents the most exhaustive purification of gpA commonly in use, producing material less likely to contain contaminating host or organism components than preparations used in other published studies [9.21,23]. Two cultured cell lines as well as human peripheral blood monocytes (prepared by elutriation), human neutrophils (collected by density gradient centrifugation and dextran sedimentation), rat alveolar macrophages (collected by lavage of dissected lungs), and human alveolar macrophages (collected by bronchoalveolar lavage from patients undergoing diagnostic bronchoscopy), were tested for responsiveness to gpA. U937 and Mono Mac-6 cell lines are derived from human leukemic monoblasts; the latter behaves physiologically like normal monocytehacrophages in response to many activating stimuli [22]. For assays of IL8 production (in human cells only), cells were washed in RPMI media supplemented with penicillin and streptomycin, then cultured for 1 hour on 6-well tissue culture plates to allow macrophages to adhere. After rinsing with fresh media to remove non-adherent cells, cells were treated with gpA and modulators, then cultured overnight. Media from these experiments was analyzed for IL8 by a commercially available ELISA (R & D Systems). Human neutrophils were used in preliminary studies of respiratory burst induction in response to human-derived gpA. Respiratory burst induction was assayed as production of superoxides using a standard spectrophotometric method [3]. Chemotaxis was assayed using a double-we11 chemotaxis chamber with the wells separated by a polycarbonate membrane containing 8pm pores. For comparisons of gpA effects to controls, a Student's t-test for paired data was used, with significance defined as a p value less than 0.05. RESULTS AND DISCUSSION. Human peripheral blood monocytes and the two monocyte/macrophage cell lines tested produced IL8 at baseline and significantly increased IL8 production in response to gpA. The ILS response to hgpA in normal monocytes was doseand time-dependent, and was blocked by co-incubation with a rabbit polyclonal antibody to gpA (Fig. 1); pre-immune seum from the same animal had no effect (data not shown). Peak response to gpA varied with different preparations.