SP-D counteracts GM-CSF-mediated increase of granuloma formation by alveolar macrophages in lysinuric protein intolerance

Background: Pulmonary alveolar proteinosis (PAP) is a syndrome with multiple etiologies and is often deadly in lysinuric protein intolerance (LPI). At present, PAP is treated by whole lung lavage or with granulocyte/monocyte colony stimulating factor (GM-CSF); however, the effectiveness of GM-CSF in treating LPI associated PAP is uncertain. We hypothesized that GM-CSF and surfactant protein D (SP-D) would enhance the clearance of proteins and dying cells that are typically present in the airways of PAP lungs. Methods: Cells and cell-free supernatant of therapeutic bronchoalveolar lavage fluid (BALF) of a two-year-old patient with LPI were isolated on multiple occasions. Diagnostic BALF samples from an age-matched patient with bronchitis or adult PAP patients were used as controls. SP-D and total protein content of the supernatants were determined by BCA assays and Western blots, respectively. Cholesterol content was determined by a calorimetic assay or Oil Red O staining of cytospin preparations. The cells and surfactant lipids were also analyzed by transmission electron microscopy. Uptake of Alexa-647 conjugated BSA and Dil-labelled apoptotic Jurkat T-cells by BAL cells were studied separately in the presence or absence of SP-D (1 mu g/ml) and/or GM-CSF (10 ng/ml), ex vivo. Specimens were analyzed by light and fluorescence microscopy. Results: Here we show that large amounts of cholesterol, and large numbers of cholesterol crystals, dying cells, and lipid-laden foamy alveolar macrophages were present in the airways of the LPI patient. Although SP-D is present, its bioavailability is low in the airways. SP-D was partially degraded and entrapped in the unusual surfactant lipid tubules with circular lattice, in vivo. We also show that supplementing SP-D and GM-CSF increases the uptake of protein and dying cells by healthy LPI alveolar macrophages, ex vivo. Serendipitously, we found that these cells spontaneously generated granulomas, ex vivo, and GM-CSF treatment drastically increased the number of granulomas whereas SP-D treatment counteracted the adverse effect of GM-CSF. Conclusions: We propose that increased GM-CSF and decreased bioavailability of SP-D may promote granuloma formation in LPI, and GM-CSF may not be suitable for treating PAP in LPI. To improve the lung condition of LPI patients with PAP, it would be useful to explore alternative therapies for increasing dead cell clearance while decreasing cholesterol content in the airways.