The role of homodimers in surfactant protein B function in vivo.

Surfactant protein B (SP-B) is detected in the airways as a sulfhydryl-dependent dimer (M(r) approximately 16,000). To test the hypothesis that formation of homodimers is critical for SP-B function, the cysteine residue reported to be involved in SP-B dimerization was mutated to serine (Cys(248) --> Ser) and the mutated protein was targeted to the distal respiratory epithelium of transgenic mice. Transgenic lines which demonstrated appropriate processing, sorting, and secretion of human SP-B monomer were crossed with SP-B +/- mice to achieve expression of human monomer in the absence of endogenous SP-B dimer (hSP-B(mon), mSP-B-/-). In two of three transgenic lines, hSP-B(mon), mSP-B-/- mice had normal lung structure, complete processing of SP-C proprotein, well formed lamellar bodies, and normal longevity. Pulmonary function studies revealed an altered hysteresis curve for hSP-B(mon), mSP-B-/- mice relative to wild type mice. Large aggregate surfactant fractions from hSP-B(mon), mSP-B-/- mice resulted in higher minimum surface tension in vitro compared with surfactant from wild type mice. Surfactant lipids supplemented with 2% hSP-B monomer resulted in slower adsorption and higher surface tension than surfactant with 2% hSP-B dimer. Taken together, these data indicate a role for SP-B dimer in surface tension reduction in the alveolus.

table respiratory distress syndrome at birth and death shortly thereafter (2). Similarly, mice in which both SP-B alleles were disrupted (SP-BϪ/Ϫ) rapidly developed lethal, neonatal respiratory distress syndrome (3). Pulmonary compliance was dramatically decreased in SP-BϪ/Ϫ neonates and intracellular packaging of surfactant phospholipids was disrupted, resulting in a virtual absence of lamellar bodies, the intracellular storage granule for surfactant. Intracellular processing of the surfactant protein C proprotein was disrupted, leading to accumulation of an SP-C processing intermediate (M r ϳ 8,000) and reduced levels of SP-C mature peptide (M r ϳ 3,500) in the alveolar surfactant pool. Collectively, these observations indicate that SP-B plays a critical role in the maintenance of pulmonary surfactant homeostasis.
SP-B is synthesized as a 381-amino acid preproprotein in Type II cells and Clara cells of the distal respiratory epithelium. Ablation of SP-B expression in Clara cells does not affect lung funtion whereas ablation of SP-B expression in Type II cells results in lethal, neonatal respiratory dysfunction (4). In Type II cells, SP-B proprotein is routed to the multivesicular body where amino-and carboxyl-terminal peptides are sequentially cleaved to produce the 79-amino acid mature SP-B peptide (5). The mature peptide shares sequence homology with saposin proteins A-D, placing it in the saposin-like protein (SAPLIP) family. Members of the SAPLIP family share a common secondary structure characterized by three conserved intramolecular cysteine bridges (6). SP-B differs from other members of this family in that it is more hydrophobic and forms sulfhydryl-dependent homodimers (7). The SP-B proprotein, M r ϳ 42,000, and its major processing intermediate, M r ϳ 25,000, are never detected in oligomeric form, consistent with dimerization of the SP-B mature peptide following proprotein processing. In humans and mice, the mature peptide is detected exclusively as a homodimer (M r ϳ 16,000), suggesting that dimerization may be important for pulmonary surfactant structure and/or function. The current study was undertaken to test the hypothesis that formation of SP-B homodimers is critical for SP-B function in transgenic mice.

MATERIALS AND METHODS
DNA Constructs-To generate a dimerization-deficient human SP-B construct (hSP-B mon ), site-directed mutagenesis was employed using a PCR-ligation-PCR protocol to substitute serine for the cysteine residue involved in dimerization of the mature peptide (Cys 248 3 Ser) (8). Primers were chosen which would amplify a 1.6-kb fragment of human SP-B cDNA that included the entire 381-amino acid preproprotein sequence (5Ј-ATTACCGTCGACCTCGAGAAGCTGCAGAGGTGCCAT-G-3Ј and 5Ј-CAGGAATTCGATATCAATCTTTGAGTGCATGGGGGTG-GG-3Ј). Internal primers used in the early round of PCR resulted in the substitution of serine (TCA) for cysteine (TGC (upstream primer 5Ј-C-TGGCTGAGCGCTACTCCGTCATC-3Ј and downstream primer 5Ј-TG-ACTGGCAGATGCCGCCCGCCACCA-3Ј)). This construct was cloned into the mammalian expression vector PCI-neo (Promega, Madison, WI) and subjected to bidirectional sequencing to verify the fidelity of the PCR product throughout the SP-B coding region and to confirm the mutation.
* This work was supported by National Institutes of Health Grants HL36055 and HL56285 (to T. E. W.), HL38859 (to J. A. W.), and HL61646 (to M. I.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ This work was done in partial fulfillment for the doctoral degree in Molecular and Developmental Biology and the Physician Scientist Training Program at the University of Cincinnati.
Targeted Expression of Monomeric SP-B to the Distal Respiratory Epithelium of Transgenic Mice-To generate mice that expressed the mutant peptide in a spatial and temporal manner similar to endogenous SP-B, hSP-B mon was cloned into the pUC-hSP-C expression vector (9). This vector contained the 3.7-kb human SP-C promoter fragment, which drives cell-specific expression in the distal respiratory epithelium, as well as a 400-base pair fragment containing the SV40 small t intron and polyadenylation signals. The transgene was microinjected into fertilized FVB/N oocytes by the Children's Hospital Transgenic Core facility and founders were identified by transgene-specific PCR and confirmed by Southern analysis using a 32 P-radiolabeled probe that recognized the SV40 small t intron (10). Transgenic animals were crossed with hemizygous SP-B gene targeted mice (mSP-Bϩ/Ϫ) for four consecutive generations to generate lines with single transgene insertion sites in the SP-Bϩ/Ϫ background.
Surfactant Protein Expression in Transgenic Mice-To identify transgenic lines that expressed monomeric human SP-B, lung tissue from 6-week-old progeny of each founder was isolated and homogenized in phosphate-buffered saline with 1 volume % protease inhibitor mixture (Sigma). Protein concentrations were determined by bicinchoninic acid protein assay (11) and equal amounts of total lung protein were analyzed by SDS-PAGE in the absence of 2-mercaptoethanol to preserve sulfhydryl-dependent oligomerization. Gels were electrophoretically transferred to nitrocellulose and Western blotting was performed using antibodies directed against the carboxyl terminus of pro-SP-B, mature SP-B, the amino terminus of pro-SP-C (12,13), and recombinant mature SP-C (Byk Gulden, Konstanz, Germany).
Expression of Monomeric SP-B in the SP-B Null Background-To generate animals that expressed monomeric human SP-B in the SP-BϪ/Ϫ background, transgenic animals (hSP-B mon , mSP-Bϩ/ϩ) were first crossed with SP-Bϩ/Ϫ mice. Transgenic offspring with a single copy of the endogenous SP-B allele (hSP-B mon , mSP-Bϩ/Ϫ) were subsequently crossed with SP-B hemizygous animals to generate offspring that contained the monomer transgene in the absence of endogenous SP-B (hSP-B mon , mSP-BϪ/Ϫ). The SP-BϪ/Ϫ genotype was confirmed by amplification of a 438-base pair PCR product corresponding to the neomycin resistance gene (upstream primer 5Ј-CACAACAGACAATCG-GCT-3Ј and downstream primer 5Ј-CAGTTCGGCTGGCGCGAG-3Ј) and the absence of PCR product for murine SP-B (10). Six to eight hSP-B mon , mSP-BϪ/Ϫ mice from three independent lines (A, D, and E) were killed and surfactant protein expression assessed by Western analysis. To determine SP-B concentration in the air spaces (i.e. secreted SP-B), surfactant was isolated by bronchoalveolar lavage from two groups of five wild type and five hSP-B mon , mSP-BϪ/Ϫ mice (14). The surfactant pellets were extracted with chloroform/methanol/water (3:2:0.75) and the organic phase was recovered for isolation of SP-B by gel filtration over Sephadex LH-60 in chloroform, methanol, 0.01 M HCl, 19:19:2 (v/v/v) (15).
Lung Morphology in Transgenic Animals-To assess lung morphology, lungs from four to six adult hSP-B mon , mSP-BϪ/Ϫ mice from lines A, D, and E 1 were inflation fixed for immunostaining and light microscopy as described (16). Immunostaining for surfactant proteins was performed with antisera directed against the carboxyl terminus of pro-SP-B, mature SP-B, and the amino terminus of pro-SP-C. To assess the ultrastructural characteristics of Type II cells, lungs from four adult hSP-B mon , mSP-BϪ/Ϫ mice from transgenic lines A and D were fixed, embedded, and viewed by transmission electron microscopy as described (4).
Surfactant Protein mRNA Expression-To determine relative expression levels of human SP-B mRNA, S1 nuclease mapping was performed similarly to that described (17). S1 probes specific for murine cytoplasmic ␤-actin, murine SP-C, murine SP-B, and human SP-B were radiolabeled with [␥-32 P]ATP (17,18). Three micrograms of total lung RNA was hybridized with 0.01 pmol of each probe at 55°C overnight, followed by S1 nuclease (Life Technologies, Inc., Gaithersburg, MD) digestion at room temperature for 1 h. Protected fragments were separated in a 6% polyacrylamide, 8 M urea gel that was then dried and quantitated by PhosphorImaging (Molecular Dynamics, Sunnyvale, CA). All PhosphorImage data were analyzed in ImageQuant (Molecular Dynamics) and SP-B RNA levels were normalized to cytoplasmic ␤-actin.
Analysis of Lung Function-To test lung function in vivo, pressurevolume curves were obtained for six wild type, seven hSP-B mon , mSP-Bϩ/Ϫ, and five hSP-B mon , mSP-BϪ/Ϫ 5-week-old mice from the D transgenic line. Mice were killed with a lethal intraperitoneal injection of sodium pentobarbital and allowed to expire in a 100% oxygen chamber. The chest wall was completely opened, the trachea was cannulated, and connected to a syringe and pressure sensor (X-ducer; Motorola, Phoenix, AZ). The lungs were inflated in 75-or 100-l increments to a maximum pressure of 30 cm H 2 O and deflated to negative pressure. The hysteresis ratio was calculated as the area bound by the inflation and deflation curves divided by the total area bound by the maximum and minimum values for pressure and volume area (software provided by Huvard Research and Consulting, Chesterfield, VA).
Surfactant Protein Production in GM-CSFϪ/Ϫ Mice-In order to produce larger quantities of human SP-B monomer or dimer in vivo, transgenic lines were crossed with GM-CSFϪ/Ϫ mice, which have increased levels of surfactant proteins in alveolar lavage fluid (17). Mice from transgenic line E (hSP-B mon , mSP-BϪ/Ϫ) were bred with GM-CSF null mice (GMϪ/Ϫ) to generate animals with single copies of the transgene, murine SP-B, and GM-CSF (hSP-B mon , mSP-Bϩ/Ϫ, GMϩ/Ϫ). These animals were subsequently crossed to generate mice that produced only SP-B monomer in the GM-CSF null background (hSP-B mon , mSP-BϪ/Ϫ, GMϪ/Ϫ). Similar crosses were performed with a transgenic line that expressed human SP-B dimer in the absence of murine SP-B to generate mice that produced increased quantities of the human peptide (hSP-B dimer , mSP-BϪ/Ϫ, GMϪ/Ϫ) (10). A PCR specific for the endogenous GM-CSF allele (5Ј-TGCCCAAGCCCAGGAGAGC-3Ј and 5Ј-CCACCGAGAAGCAAGCAACCA-3Ј) was used to identify GM-CSFϪ/Ϫ animals. SP-B protein was isolated as described above.
In Vitro Surface Tension Measurements-To assess the surface tension reducing properties of surfactant, large aggregate surfactant was isolated from bronchoalveolar lavage of wild type and hSP-B mon , mSP-BϪ/Ϫ mice from the D transgenic line and diluted to a final concentration of saturated phosphatidylcholine equal to 10 mol/ml, as described (14). A Wilhelmy balance with a platinum dipping stick was used to estimate minimum surface tensions achieved during five consecutive cycles of compression and expansion. The surface properties of SP-B monomer and dimer isolated from GM-CSFϪ/Ϫ mice were assessed by captive bubble surfactometry as described (19). Samples containing hSP-B mon /DPPC/POPG, hSP-B dimer /DPPC/POPG, and DPPC/POPG at concentrations of 25 g/l were placed at the air-liquid interface.

Characterization of Human SP-B Monomer Transgenic
Mice-SP-B is synthesized as a preproprotein which is proteolytically cleaved to generate the 79-amino acid mature peptide. The mature peptide is detected exclusively as a homodimer in mouse bronchoalveolar lavage fluid. In order to identify the function of SP-B dimerization, transgenic mouse lines were generated in which the cysteine residue proposed to be responsible for mediating dimerization (7) was mutated to serine. Expression of the SP-B monomer was targeted to the distal respiratory epithelium using the 3.7-kb human SP-C promoter (Fig. 1A). Seven of 13 (54%) offspring from fertilized oocyte injections were transgene positive, as identified by both PCR and Southern blot analyses of tail DNA (not shown). Under nonreducing electrophoretic conditions, the SP-B homodimer (M r ϳ 16,000) was detected in all animals whereas the SP-B monomer (M r ϳ 8,000) was detected only in transgenic animals (Fig. 1B). This result confirmed that cysteine 248 was essential for SP-B homodimer formation. Monomeric SP-B protein was detected in offspring from three of the seven transgenic lines (A, D, and E). These mice (hSP-B mon , mSP-Bϩ/ϩ) survived without any overt evidence of respiratory pathophysiology and had normal body weight, lung weight, reproductive function, longevity, and lung structure (not shown), indicating that expression of SP-B monomer in the wild type SP-B background did not significantly alter lung function.
SP-BϪ/Ϫ mice die of acute respiratory distress syndrome shortly after birth (3). To determine if SP-B monomer was able to restore lung function in SP-BϪ/Ϫ mice, transgenic lines expressing monomer protein were bred with hemizygous animals (SP-Bϩ/Ϫ) in order to achieve expression of SP-B monomer in the null background (hSP-B mon , mSP-BϪ/Ϫ). Adult offspring that exclusively expressed SP-B monomer were detected for three independent transgenic lines, indicating that dimerization of SP-B was not required for survival in the neonatal period. Western blotting with SP-B-specific antiserum confirmed the complete absence of SP-B dimer in hSP-B mon , mSP-BϪ/Ϫ mice (Fig. 2). To determine if there was increased mortality in animals expressing only SP-B monomer, a series of 10 crosses was performed in which a rescued male from transgenic line D (hSP-B mon , mSP-BϪ/Ϫ) was crossed with hemizygous females (mSP-Bϩ/Ϫ). The expected Mendelian inheritance pattern for mice alive at 6 weeks of age was 1:2:0 for rescued (hSP-B mon , mSP-BϪ/Ϫ): hemizygous (mSP-BϮ): null (mSP-BϪ/Ϫ) animals. The observed frequencies were not significantly different from the expected frequencies (Table I).
Overall, replacement of endogenous SP-B with SP-B monomer resulted in complete reversal of the neonatal lethal SP-BϪ/Ϫ phenotype.
Lung Structure and Surfactant Protein Expression in hSP-B mon , mSP-BϪ/Ϫ Mice-SP-B null mice (mSP-BϪ/Ϫ) have poor lung compliance, resulting in collapsed airways and alveoli (20). To test the ability of SP-B monomer to restore normal lung structure in mSP-BϪ/Ϫ mice, histological sections from hSP-B mon , mSP-BϪ/Ϫ lungs were examined. Lung architecture in mice from transgenic lines D and E 1 was histologically indistinguishable from wild type littermates (not shown). Immunostaining for SP-B in these animals revealed that the human SP-B monomer was appropriately expressed in Type II cells and non-ciliated bronchiolar epithelial cells, recapitulating the endogenous expression pattern for SP-B (not shown). Ultrastructure analyses revealed that animals which expressed only SP-B monomer (hSP-B mon , mSP-BϪ/Ϫ) formed lamellar bodies, in contrast to SP-BϪ/Ϫ mice in which mature lamellar bodies were never detected (not shown) (3). In addition, levels of mature SP-C peptide, which were dramatically reduced in SP-BϪ/Ϫ mice, were normal in hSP-B mon , mSP-BϪ/Ϫ mice (Fig. 3). Furthermore, the SP-C processing intermediate which accumulated in SP-BϪ/Ϫ mice was not detected in hSP-B mon , mSP-BϪ/Ϫ mice, consistent with restoration of SP-C proprotein processing (Fig. 3).
The level of SP-B peptide in surfactant has been shown to be important in lung function (21,22). The amount of monomeric SP-B protein in bronchoalveolar lavage fluid from transgenic line D hSP-B mon , mSP-BϪ/Ϫ mice relative to wild type mice was estimated from amino acid composition analyses following organic extraction (23). In two separate experiments, bronchoalveolar lavage from wild type mice contained 0.36 and 0.25 g of SP-B per mouse, whereas hSP-B mon , mSP-BϪ/Ϫ mice contained 0.42 and 0.25 g of SP-B per mouse. This result suggested that monomeric SP-B was able to function effectively when present at levels comparable to endogenous SP-B levels in wild type mice.
Consistent with the hypothesis that SP-B protein levels are critical for lung function, transgenic line E, which displayed variable hSP-B expression levels, had decreased survival. Offspring resulting from crosses with a transgenic male (E 1 ) expressed high levels of SP-B monomer protein and survived in the mSP-BϪ/Ϫ background, whereas offspring from a transgenic female (E 2 ) had low levels of monomer protein in the SP-B null background and never survived the neonatal period (Fig. 4). This was demonstrated for five consecutive generations of animals in the SP-BϪ/Ϫ background, and was therefore not likely the result of independent integration sites. The most likely explanation for this finding is that the transgene had inserted into a site subject to genomic imprinting (24,25). S1 nuclease analyses confirmed that levels of hSP-B mRNA in transgenic line E 2 (hSP-B mRNA/␤-actin mRNA ϭ 0.5) were  3, 7, and 8).

TABLE I SP-B monomer reversed the SP-BϪ/Ϫ neonatal lethality
A series of crosses was set up between mice that expressed the human monomer in the null background (hSP-B mon , mSP-BϪ/Ϫ) and SP-Bϩ/Ϫ animals. The frequency of rescued offspring (hSP-B mon , mSP-BϪ/Ϫ) did not differ significantly from that predicted by Mendelian inheritance, which indicated that human monomer was able to compensate for endogenous mSP-B. It is therefore likely that expression levels of SP-B monomer in transgenic line E 2 were insufficient to restore normal lung function in mSP-BϪ/Ϫ mice.
Surfactant Function in hSP-B mon , mSP-BϪ/Ϫ Mice-To determine if the SP-B monomer was as effective as its dimer in reducing surface tension in the lung, pressure-volume curves were generated for wild type, hSP-B mon , mSP-Bϩ/Ϫ, and hSP-B mon , mSP-BϪ/Ϫ mice (Fig. 5A). The hysteresis area was significantly decreased in hSP-B mon , mSP-BϪ/Ϫ mice when compared with wild type and hSP-B mon , mSP-Bϩ/Ϫ mice (Fig. 5B).
To test the hypothesis that surfactant from hSP-B mon , mSP-BϪ/Ϫ mice had altered surface properties, the ability to reduce surface tension in vitro was tested using a Wilhelmy balance. Minimum surface tensions for five consecutive cycles were obtained for large aggregate surfactant fractions isolated from wild type and hSP-B mon , mSP-BϪ/Ϫ mice (Fig. 6). The minimum surface tension achieved with wild type surfactant was less than 10 milliNewtons/m, consistent with previous studies (26). Surfactant from hSP-B mon , mSP-BϪ/Ϫ mice generated higher surface tension upon compression.
The effect of sulfhydryl-dependent dimerization of SP-B on the ability of surfactant phospholipids to adsorb to an air-liquid interface and reduce surface tension was directly tested (Fig. 7) by recombining purified SP-B monomer or dimer with a defined, synthetic phopholipid mixture. Data in Fig. 7A demonstrate that minimum surface tension was attained rapidly for all surfactant mixtures in the absence of cycling. However, surfactant lipids with 2% SP-B monomer (w/w) had a significantly slower rate of adsorption than did surfactant with 2% SP-B dimer (Fig. 7A). After multiple cycles of compression and expansion, minimum surface tension achieved with monomer surfactant was significantly higher than surfactant with SP-B dimer (Fig. 7B). These results indicate that, on a mass basis, SP-B monomer is not as effective as wild-type SP-B at reducing surface tension at an air-liquid interface. However, when the concentration of SP-B monomer was doubled relative to the concentration of SP-B dimer, a similar reduction in surface tension was achieved. Gel filtration of monomeric SP-B (approximately 10 M) over Sephadex LH-60 in chloroform, methanol, 0.1 M HCl (19:19:2, by volume) showed that the molecule migrated with an apparent molecular mass of 9 kDa, indicating a monomeric quaternary structure in this concentration range. Overall, the results of in vivo and in vitro studies indicate that dimerization of SP-B is not absolutely required for lung function, but that SP-B dimerization contributes to optimal surfactant function.

DISCUSSION
The SP-B mature peptide is detected exclusively as a homodimer in humans and mice. It is the only member of the SAPLIP family that forms disulfide-dependent dimers, suggesting that dimerization of SP-B may be critical for its role in surface tension reduction at the alveolar air-liquid interface. To test this hypothesis, transgenic mouse lines were generated that expressed SP-B monomer in the distal respiratory epithelium. Under oxidizing conditions that maintained cysteine bridge integrity, transgenic protein was detected as a monomer, M r ϳ 8,000. This result confirmed the prediction from in vitro analyses that SP-B cysteine residue 248 participates in an FIG. 4. SP-B monomer expression levels in line E 2 are insufficient to restore normal lung function to hSP-B mon , mSP-B؊/؊ mice. Four micrograms of lung homogenate protein from 6-week-old mice were subjected to SDS-PAGE under nonreducing conditions and analyzed by Western blotting with anti-mature SP-B antibody. hSP-B monomer (arrow) was detected in lines D and E. However, line E was subject to genomic imprinting, which resulted in lower expression levels when the transgene was inherited maternally (E 2 ) than when inherited paternally (E 1 ). No hSP-B mon , mSP-BϪ/Ϫ offspring survived from crosses with transgenic females from the E line, whereas surviving offspring were readily identified from crosses involving transgenic males from the E line.

FIG. 5. Lung hysteresis is decreased in hSP-B mon , mSP-B؊/؊ mice.
A, representative pressure-volume curves are shown for 5-weekold wild type, hSP-B mon , mSP-Bϩ/Ϫ, and hSP-B mon , mSP-BϪ/Ϫ mice. Narrowing of the pressure-volume curve in hSP-B mon , mSP-BϪ/Ϫ mice was consistently observed. B, graphic depiction of the altered hysteresis in hSP-B mon , mSP-BϪ/Ϫ mice. Hysteresis area was significantly reduced in hSP-B mon , mSP-BϪ/Ϫ mice when compared with wild type and hSP-B mon , mSP-Bϩ/Ϫ mice (p Ͻ .01 by ANOVA, Tukey Compromise post-hoc test, values shown are mean Ϯ S.D.). intermolecular disulfide bridge and mediates SP-B homodimerization (7). No overt phenotype was observed in hSP-B mon , mSP-Bϩ/ϩ mice, consistent with previous studies demonstrating that expression of human SP-B did not alter lung structure or function in wild type mice (4, 10).
Expression of SP-B monomer protein in SP-BϪ/Ϫ mice completely reversed the neonatal lethality associated with SP-B deficiency. Normal longevity was restored in these mice, indicating that the formation of SP-B dimer was not essential for survival in unchallenged animals. The role of SP-B dimerization in lung function was assessed by generation of pressurevolume curves. The hysteresis area of lungs containing only SP-B monomer was significantly decreased when compared with lungs containing equivalent amounts of SP-B dimer on a weight basis. This indicates that there was a loss of function in the absence of SP-B dimer that could not be compensated for by the presence of SP-B monomer. These data indicate that SP-B dimer is important in establishing normal lung hysteresis, likely due to an important role in surfactant function.
To further investigate the role of SP-B dimerization in surfactant film formation and surface tension reduction, the surface properties of large aggregate surfactant fractions isolated from hSP-B mon , mSP-BϪ/Ϫ, and wild type mice were tested in vitro. As predicted from pressure-volume studies, surfactant from hSP-B mon , mSP-BϪ/Ϫ mice was less able to reduce surface tension forces compared with wild type surfactant. Two explanations are possible: SP-B dimers contribute directly to the surface tension reducing properties of pulmonary surfactant, or SP-B dimerization is critical in the Type II cell for formation of an optimal surfactant protein-lipid complex. To distinguish between these possibilities, SP-B monomer and dimer were isolated from mice that produced large quantities of human SP-B dimer (hSP-B dimer , mSP-BϪ/Ϫ, GM-CSFϪ/Ϫ) or human SP-B monomer (hSP-B mon , mSP-BϪ/Ϫ, GM-CSFϪ/Ϫ) and used to reconstitute synthetic surfactant preparations for in vitro analyses. Surfactant from GM-CSFϪ/Ϫ mice has been shown to have surface properties similar to equivalent amounts of wild type surfactant, providing a useful model for generating large quantities of functional surfactant proteins and lipids (14). SP-B levels for both transgenic lines in the GM-CSFϪ/Ϫ background were increased more than 100-fold when compared with wild type mice (data not shown). When equal masses (2% w/w) of monomer or dimer were combined with a DPPC/POPG mixture, the monomer preparation exhibited surface properties that were not significantly different from lipid alone. Since SP-B dimer consists of two covalently linked mature peptides, there were twice as many molecules of SP-B monomer than SP-B dimer. Increasing the concentration of SP-B monomer 2-fold (four times the molar quantity) resulted in normal surface activity. These data suggest that increased levels of SP-B monomer can compensate for diminished protein function.
The surface tension reducing properties of surfactant are very dependent upon concentration (26). The current study utilized low concentrations of surfactants (25 g/l) in order to better identify differences between the surface activities of SP-B dimer and SP-B monomer. At this dilution, the surfactant containing 2% SP-B monomer had relatively poor surface activity. However, it should be noted that at such low concentrations, even surfactant containing SP-B dimer did not result in surface tensions approximating zero, which are characteristic of higher surfactant concentrations.
Despite decreased surface properties in vitro, the SP-B monomer completely reversed the neonatal lethality in SP-BϪ/Ϫ mice. One explanation for this outcome is that SP-B is present in concentrations in excess of that required for optimal lung function. Consistent with this hypothesis, SP-B hemizygous mice survive and have only subtle changes in lung function; similarly, human carriers of a single null SP-B allele demonstrate no evidence of lung disease (27). In the hSP-B mon , mSP-BϪ/Ϫ mice, SP-B monomer was present in similar or slightly higher levels compared with wild type mice. The results of in vitro analyses suggest that a decrease in monomer levels to 50% of wild type levels may result in a significant decrease in lung function. An alternative explanation for normal survival of hSP-B mon , mSP-BϪ/Ϫ mice is that the role of SP-B in surface tension reduction is partially redundant with a similar role for SP-C (28). In SP-B null mice, SP-C mature peptide levels were greatly decreased; whereas in hSP-B mon , mSP-BϪ/Ϫ mice, SP-C mature peptide was restored to levels comparable to wild type mice. Therefore, mature SP-C protein may be critical for lung function in hSP-B mon , mSP-BϪ/Ϫ mice.
Incomplete SP-C processing in humans with hereditary SP-B deficiency and mSP-BϪ/Ϫ mice resulted in a relative deficiency of mature SP-C peptide in the airways. Lamellar body formation in Type II cells was also disrupted, indicating that SP-B plays an important role in the intracellular processing and packaging of pulmonary surfactant. The mechanisms of lamellar body formation and processing of surfactant proteins B and C in the multivesicular body are unknown, although it is likely that SP-B plays a pivotal role. In vitro studies demonstrated that the SP-B mature peptide is both fusogenic and lytic, and these properties may contribute to lamellar body formation (29). SP-B dimerization occurs concomitantly with protein processing in the multivesicular body, suggesting that SP-B dimer may be necessary for packaging of surfactant phospholipids during the transition from MVB to lamellar body. However, this study demonstrated that lamellar bodies formed in the absence of SP-B dimers and processing of SP-C to the mature peptide was complete in hSP-B mon , mSP-BϪ/Ϫ mice. Since the SP-B monomer was sufficient for correction of both lamellar body formation and SP-C proprotein processing, it is likely that the primary function of SP-B dimerization is related to surface tension reduction.
Previous work in this laboratory suggested that SP-B must be expressed in all Type II cells to restore normal lung structure and function (4). In the current study, one transgenic line was established that demonstrated expression of human SP-B monomer in a subset of Type II cells (A line, data not shown). In hSP-B mon , mSP-BϪ/Ϫ mice from this line, there were two populations of Type II cells, one with normal lamellar bodies and one with abnormal lipid inclusions that resembled those found in SP-BϪ/Ϫ mice. The SP-C processing intermediate that accumulates in the absence of SP-B was detected in lungs from hSP-B mon , mSP-BϪ/Ϫ mice in this transgenic line. Lungs of adult hSP-B mon , mSP-BϪ/Ϫ offspring from this line had altered lung structure, with greatly widened air spaces. Taken together, these data support the hypothesis that SP-B production is required in all Type II cells, and that this requirement cannot be met through the presence of exogenous (i.e. airway) surfactant. A model has been proposed in which alveoli containing Type II cells deficient in SP-B have a tendency to collapse due to surfactant with poor surface activity (4).
In summary, expression of SP-B monomer reversed the neonatal lethality and restored SP-C processing and lamellar body formation in SP-BϪ/Ϫ mice. Monomer rescued mice revealed no overt phenotype, bred successfully, and experienced normal longevity. These results indicate that formation of SP-B dimers is not essential for surfactant function. However, hSP-B monomer was associated with altered lung hysteresis in vivo and altered surface properties in vitro. Taken together, these studies indicate a role for SP-B dimer in establishing normal lung hysteresis, and suggest that SP-B dimerization may be required for optimal lung function.