Molecular cloning, characterization, and differential expression pattern of mouse lung surfactant convertase.

We recently reported the purification and partial amino acid sequence of "surfactant convertase," a 72-kDa glycoprotein involved in the extracellular metabolism of lung surfactant (S. Krishnasamy, N. J. Gross, A. L. Teng, R. M. Schultz, and R. Dhand. Biochem. Biophys. Res. Commun. 235: 180-184, 1997). We report here the isolation of a cDNA clone encoding putative convertase from a mouse lung cDNA library. The cDNA spans a 1,836-bp sequence, with an open reading frame encoding 536 amino acid residues in the mature protein and an 18-amino acid signal peptide at the NH2 terminus. The deduced amino acid sequence matches the four partial amino acid sequences (68 residues) that were previously obtained from the purified protein. The deduced amino acid sequence contains an 18-amino acid residue signal peptide, a serine active site consensus sequence, a histidine consensus sequence, five potential N-linked glycosylation sites, and a COOH-terminal secretory-type sequence His-Thr-Glu-His-Lys. Primer-extension analysis revealed that transcription starts 29 nucleotides upstream from the start codon. Northern blot analysis of RNA isolated from various mouse organs showed that convertase is expressed in lung, kidney, and liver as a 1,800-nucleotide-long transcript. The nucleotide and amino acid sequences of putative convertase are 98% homologous with mouse liver carboxylesterase. It thus may be the first member of the carboxylesterase family (EC 3.1.1.1) to be expressed in lung parenchyma and the first with a known physiological function.

Protein-lipid interactions and enzyme requirements for light subtype generation on cycling reconstituted surfactant.

Surfactant convertase is required for conversion of heavy density (H) natural surfactant to light density (L) subtype during cycling in vitro, a technique that reproduces surfactant metabolism. To study mechanisms of H to L conversion, we prepared liposomes of dipalmitoylphosphatidylcholine (DPPC) and phosphatidylglycerol (PG), or the phospholipids (PL) in combination with either surfactant protein A (SP-A), surfactant protein B (SP-B), or both SP-A and SP-B. Phospholipids alone showed time-dependent conversion from heavy to light subtype on cycling in the absence of convertase, which was decreased by adding SP-B, but not SP-A, to phospholipids (p < 0.01 for PL+SP-B, or PL+SP-A+SP-B vs. PL, or PL+SP-A). The ultrastructure, surface activity, buoyant density, and L subtype generation on cycling PL+SP-A+SP-B with partially purified convertase or with phospholipase D were similar to those of natural TM. In conclusion, a reconstituted surfactant mimics the behavior of natural surfactant on cycling, and reveals that interaction of SP-B with phospholipids decreases L subtype generation. In addition, esterase/ phospholipase D activity is required for conversion of heavy to light subtype on cycling.

Lung "surfactant convertase" is a member of the carboxylesterase family.

The extracellular conversion of lung surfactant from tubular myelin to the small vesicular form has previously been shown to require a serine-active enzyme called "surfactant convertase." In the present study, a 72kD serine-active enzyme previously identified in mouse lung alveolar lavage and having convertase activity was partially sequenced. Sixty-eight residues obtained from amino acid sequencing of this protein show that it is a new member of the mouse carboxylesterase family (EC 3.1.1.1). The 72kD lung protein also has esterase activity. A commercial esterase of the same family was able to reproduce surfactant convertase bioactivity in vitro, unlike several serine proteinases previously tested. We conclude that surfactant convertase is a carboxylesterase which mediates a biochemical step in the extracellular metabolism of surfactant.

Isolation, cloning and characterization of a putative type-1 astrocyte cell line.

We have established a permanent cell line (1H91) of putative type-1 astrocyte precursor cells that were clonally derived from a single cell isolated from E16 mouse cerebellum. Epidermal growth factor (EGF) and transforming growth factor (TGF alpha) are strong mitogens for 1H91 cells (ED50 of 9.02 + 1.74 ng/ml and 15.98 +/- 2.34 ng/ml, respectively), while basic fibroblast growth factor (bFGF) is only weakly mitogenic and platelet derived growth factor (PDGF) has no mitogenic activity. In the proliferative state, the 1H91 cells are immunohistochemically positive for nestin and vimentin, and negative for A2B5, CNPase, neurofilament (NF), and neuron specific enolase (NSE). The majority of EGF-treated 1H91 cells are not immunoreactive for glial acid fibrillary protein (GFAP). In the presence of 5 ng/ml bFGF, 1H91 cells become non-mitotic and develop a morphology consistent with a fibrous astrocyte. In contrast to the proliferating cultures, the bFGF treated cultures were strongly immunoreactive for GFAP, only mildly immunoreactive for nestin and vimentin, and negative for A2B5, CNPase, NF, and NSE. Type-1 astrocytes are known to proliferate in response to EGF, and are immunohistochemically GFAP positive, A2B5 negative, and CNPase negative [38]. However, type-1 astrocytes only develop a fibrous morphology during the process of reactive gliosis [31]. Since EGF is a strong mitogen for 1H91 cells, and these cells may be differentiated into GFAP positive, A2B5 negative, CNPase negative astrocytes, we conclude that 1H91 cells conform to a type-1 astrocyte precursor phenotype. In addition, the fibrous morphology of the bFGF treated 1H91 cells suggests that these cells follow the process of reactive gliosis. Therefore, the 1H91 clonal cell line may provide an in vitro model for studying the underlying cellular mechanisms of the type-1 astrocyte in reactive gliosis.