INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IN NORMAL LUNG, the air-liquid interface at the alveolar surface is covered with a saturated surfactant phospholipid film that reduces surface tension to near zero as alveolar surface area decreases. The surface tension-lowering properties of surfactant prevent alveolar collapse at end expiration and minimize the pressure required to inflate the lung during inspiration (1). Saturated surfactant phospholipids alone do not spontaneously form films in vitro (3) and do not stabilize alveoli when delivered to the surfactant-deficient immature lung (15). However, the addition of surfactant protein (SP)-B to surfactant phospholipids promotes film formation at the air-liquid interface in vitro (3, 16), and intratracheal instillation of SP-B and phospholipid mixtures restores lung compliance and pressure-volume characteristics in prematurely delivered ventilated rabbit lungs (15). Hereditary SP-B deficiency in newborn babies (13) and disruption of the SP-B locus in mice (5) result in postnatal death due to respiratory failure, demonstrating an absolute requirement of SP-B for lung function.

In addition to its role in promoting the surface activity of surfactant, SP-B might play a role in host defense of the lung. The host defense properties of SP-A and SP-D are well established, and the results of recent studies suggest that SP-B may also contribute to this important function. SP-B was reported to decrease endotoxin-induced nitric oxide production by isolated alveolar macrophages consistent with an anti-inflammatory function in the lungs (12). The preliminary studies (6, 16) in mice treated with intratracheal endotoxin or hyperoxia indicated that lung inflammation was less severe in transgenic mice with higher SP-B levels. These results suggest that SP-B may play a protective role in response to challenge. In a separate study (17), intratracheal instillation of vesicles containing a T cell-dependent antigen (trinitrophenyl-keyhole limpet hemocyanin) and SP-B enhanced the systemic immune response in mice depleted of alveolar macrophages. This effect was specific to SP-B and was not detected with antigen vesicles containing SP-A or SP-C. Taken together, these data suggest that SP-B may play a role in both immune and nonimmune host defense in the lung.

Native SP-B is a homodimer in which subunit association is stabilized via covalent linkage through cysteine 48. As a first step toward characterizing the structural basis for SP-B function, we generated transgenic mice expressing a form of SP-B in which the cysteine residue involved in intermolecular disulfide bridge formation was mutated to serine (3). These transgenic mice were crossed with SP-B(+/) mice for several generations to produce offspring that expressed the transgene in the SP-B(/) background (SP-Bmon⁺). The levels of SP-B peptide in transgenic mice were similar to those in wild-type littermates, and although transgenic mice could not form sulfhydryl-dependent dimers, they survived without overt evidence of lung disease. Lung hysteresis in SP-Bmon⁺ transgenic mice was decreased, and the in vitro surface properties of mixtures of mutant SP-B (purified from SP-Bmon⁺ alveolar lavage fluid) and phospholipids were also decreased; however, the latter effect was reversed by increasing the concentration of SP-Bmon (3). Taken together, these results suggest that the intermolecular disulfide bridge was not critical for the surface properties of SP-B. The present study was undertaken to determine whether the loss of the intersubunit disulfide bridge altered the immunomodulatory properties of SP-B.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Mice. Generation of transgenic mice expressing a form of SP-B in which the cysteine residue involved in formation of homodimers (cysteine 248 in the human SP-B preproprotein) was mutated to serine was previously described (3). To express the SP-Bmon transgene in the SP-B(/) background, transgenic mice were first crossed with SP-B(+/) mice. Offspring that carried the SP-Bmon transgene and were heterozygous for endogenous SP-B [SP-B(+/)] were then crossed; from the latter cross, 3 of 16 progeny carried the SP-Bmon transgene in the SP-B(/) background (SP-Bmon⁺ mice). For the present study, SP-Bmon⁺ mice were crossed to generate offspring that were homozygous for the transgene (SP-Bmon⁺⁺). To determine whether offspring were heterozygous or homozygous for the transgene, each animal was bred to a wild-type mouse; the progeny of a homozygous SP-Bmon⁺⁺ parent all carried the transgene, whereas only 50% of progeny from a heterozygous SP-Bmon⁺ parent were transgenic. Mice (7- to 9-wk-old) were used for all studies, and six to eight mice were studied in each group. All procedures were approved by the Institutional Animal Care and Use Committee at the Cincinnati Children's Hospital Research Foundation.

Hyperoxia. Wild-type, SP-Bmon⁺⁺, and SP-Bmon⁺ mice were exposed either to 95% O₂ or to room air. Pressure-volume curves were measured after 3 days of exposure to 95% O₂ or room air. Mice were sedated with pentobarbital sodium (100 mg/kg ip) and placed in a box containing 100% O₂ to ensure complete collapse of the alveoli by O₂ absorption after spontaneous breathing stopped. The mice were killed by exsanguination, and the trachea was cannulated and connected by a syringe to a pressure sensor (mouse pulmonary testing system, TSS, Cincinnati, OH) via a three-way connector. After the diaphragm was opened, the lungs were inflated in 75-µl increments every 10 s to a maximum pressure of 36 cmH₂O and similarly deflated. Lung volume per kilogram body weight was determined at intervals of 5 cmH₂O and at maximum lung volume during inflation and deflation (8). Five 1-ml aliquots of 0.9% NaCl were flushed into the lungs and withdrawn by syringe three times for each aliquot. The lavaged lung tissue was removed and homogenized in 2 ml of 0.9% NaCl. Total protein in alveolar lavages was estimated by the method of Lowry et al. (10).

Intratracheal endotoxin administration. Wild-type, SP-Bmon⁺⁺, and SP-Bmon⁺ mice were anesthetized with isoflurane (2-3%) and orally intubated with a 25-gauge animal-feeding needle. Each mouse received 80 µl of saline as control or saline containing 10 µg of Escherichia coli lipopolysaccharide (serotype O55: B5, Sigma Chemical, St. Louis, MO). A nonmanipulated mouse group was also included in these studies. At 3 and 16 h after intratracheal injection (IT), mice were deeply anesthetized with pentobarbital sodium (100 mg/kg ip) and were killed by exsanguination. Alveolar lavage fluid and postlavage lung homogenates were prepared as described above in Mice.

Cell counts and differentials. An aliquot of alveolar lavage fluid was centrifuged at 500 g for 10 min to separate cells from supernatant. The supernatant was saved and frozen for subsequent cytokine measurements, and the pelleted cells were resuspended in a small amount of PBS. Cells were stained with trypan blue and counted on a hemocytometer. Differential cell counts were performed on cytospin preparations after staining with Diff-Quik (Scientific Products, McGaw Park, IL). Cell numbers are given per kilogram of body weight.

Cytokine production. Lung homogenates were centrifuged at 800 g, and the supernatants were stored at 20°C. Tumor necrosis factor (TNF)-, interleukin (IL)-1, IL-6, and macrophage inflammatory protein (MIP)-2 were quantitated in supernatants by using quantitative murine sandwich ELISA kits (R&D Systems, Minneapolis, MN) according to the manufacturer's directions. All plates were read on a microplate reader (THERMO max, Molecular Devices, Menlo Park, CA) and analyzed with the use of a computer-assisted analysis program (Softmax Pro version 3, Molecular Devices).

Apoptotic cells. Apoptotic cells in alveolar lavage were detected by annexin V and propidium iodide staining (Pharmingen) (19). Cells were washed and immediately analyzed on a fluorescence-activated cell sorter (FACS) caliber flow cytometer (Beckton Dickenson, Mountain View, CA).

Statistical analysis. Two-group comparisons were carried out by unpaired Student's t-tests. Comparison among wild-type, SP-Bmon⁺⁺, and SP-Bmon⁺ mice were by ANOVA with Tukey's tests used for post hoc analyses. Results were expressed as means ± SE. Significance was accepted at the 5% level.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The total volume of BALF recovered from each genotype was similar with an average of 4.35 ± 0.03 ml. Saturated phosphatidylcholine content (µmol/kg body wt) in BALF was similar among the three groups (Fig. 1A). Under nonreducing electrophoretic conditions, SP-B was detected as a homodimer [relative mobility (Mr) of ~16k] in BALF from wild-type mice (Fig. 1B), whereas SP-B migrated as a monomer (Mr of ~8k) in SP-Bmon⁺ and SP-Bmon⁺⁺ mice. Because the nonreduced monomer is much more immunoreactive than the homodimer, it is not possible to directly compare SP-B levels in wild-type mice and SP-Bmon transgenic mice by Western blotting.

View larger version (30K): [in this window] [in a new window]   Fig. 1.   A: saturated phosphatidylcholine (Sat PC; µmol/kg body wt) in bronchoalveolar lavage fluid (BALF). Sat PC levels in BALF were similar among the 3 groups. B: 10 ml of total BALF were subjected to SDS-PAGE and Western blotting with mature surfactant protein (SP)-B antiserum under nonreducing electrophoretic conditions. SP-Bmon+ mice (lanes 8-11) and SP-Bmon++ mice (lanes 4-7) expressed only the monomeric form of SP-B. SP-B was detected as a dimer in wild-type mice (lanes 1-3). Expression of SP-Bmon was higher in SP-Bmon++ mice than in SP-Bmon+ mice.

Hyperoxia. Wild-type, SP-Bmon⁺⁺, and SP-Bmon⁺ mice all tolerated 3 days of exposure to 95% O₂. Body weights were similar for all groups, with a mean of 21.8 ± 0.4 g, and body weight loss was similar (0.8 ± 0.1 g) for all groups after 3 days of exposure to hyperoxia.

Cells and protein in alveolar lavage fluid. Total cell numbers in alveolar lavage from wild-type mice were similar to SP-Bmon⁺⁺ mice and slightly lower than in SP-Bmon⁺ mice (P < 0.05) in air. Alveolar cells were increased (P < 0.05) by hyperoxic exposure (Fig. 2A) in wild-type and SP-Bmon⁺ mice groups. Cell differentiation analyses indicated that >93% of the alveolar cells were monocytes in all groups. Monocyte numbers tended to increase after hyperoxic stress but were not statistically different among groups. Protein in alveolar lavage fluid was expressed as milligrams per kilogram body weight by using body weight before oxygen exposure for this calculation since body weight tightly correlates with lung size (14). Protein in alveolar lavage was increased in all three genotypes after hyperoxia, indicating that lung injury occurred in each group (Fig. 2B).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   A: total cell numbers in alveolar lavage fluid from wild-type, SP-Bmon++, and SP-Bmon+ mice exposed to air or 95% O2 (O2). Total alveolar cell numbers were increased by hyperoxia in wild-type and SP-Bmon+ mice groups. B: total protein in alveolar lavage fluid was increased in all genotype groups after hyperoxia. * P < 0.05 vs. air.

Cytokines. Cytokines TNF-, IL-6, IL-1, and MIP-2 in alveolar lavage and in the supernatant of homogenized lung tissue were analyzed. IL-6 levels in alveolar lavage samples and in tissues are shown in Fig. 3. Although IL-6 in alveolar lavage samples was increased 7-fold (wild type) and 4-fold (SP-Bmon⁺⁺) after hyperoxia, IL-6 was increased over 50-fold in alveolar lavage sample from SP-Bmon⁺ mice. IL-6 in lung tissue was also significantly higher (P < 0.001) in SP-Bmon⁺ relative to other groups after hyperoxic stress. IL-1 and MIP-2 in alveolar lavage samples were very low, and significant differences among groups before or after hyperoxia were only detected in SP-Bmon⁺ (P < 0.01; data not shown). IL-1 and MIP-2 levels in lung tissue are shown in Fig. 4. IL-1 in wild-type and SP-Bmon⁺⁺ lungs were similarly increased after hyperoxic stress. IL-1 and MIP-2 were significantly higher (P < 0.001) in SP-Bmon⁺ compared with other groups. SP-Bmon⁺ mice were more susceptible to hyperoxia-induced lung injury than the two other genotype groups, and the degree of lung inflammation in SP-Bmon⁺⁺ and wild-type mice was similar.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Cytokine interleukin (IL)-6 levels in alveolar lavage fluid (A), and lung tissue homogenate (B) after alveolar lavage. IL-6 in alveolar lavage fluid was increased by hyperoxia in all genotype groups. The increase in IL-6 in SP-Bmon+ mice was significantly higher in both alveolar lavage and lung tissue compared with other groups. * P < 0.05 vs. air; tP < 0.001 vs. other.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   IL-1 (A) and microphage inflammatory protein (MIP)-2 (B) levels in lung tissue homogenate. IL-1 and MIP-2 were significantly higher in SP-Bmon+ compared with other groups. * P < 0.05 vs. air; tP < 0.001 vs. other.

Pressure-volume curve. To test whether hyperoxia differentially altered lung function in the three groups, pressure-volume curves were performed after 3 days of hyperoxia and compared with mice kept in air (Fig. 5). Pressure-volume curves in air for SP-Bmon⁺ mice lung had significantly decreased hysteresis area, as previously shown (3), with lower volumes at 10 and 15 cmH₂O on the deflation limb (P < 0.05) compared with wild-type mice and SP-Bmon⁺⁺ mice. Pressure-volume curves after hyperoxia were very similar in all the groups. The changes in lung volumes after 3 days of hyperoxia were larger in wild-type mice and SP-Bmon⁺⁺ mice than in SP-Bmon⁺ because baseline pressure-volume mechanics in air were already altered in the latter group of mice.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Lung pressure-volume curve of mice in air () and after 3 days of exposure to 95% oxygen (). Pressure-volume curves in air for wild-type (A) and SP-Bmon++ (B) mice were similar, and both groups were similarly decreased by hyperoxia. Pressure-volume curve for SP-Bmon+ mice (C) in air had a significantly decreased hysteresis area than other groups. The changes in lung volume in SP-Bmon+ mice after hyperoxia were only shown at 5 cmH2O on deflation limb because the pressure-volume curve was already altered in air. * P < 0.05 vs. air.

Intratracheal endotoxin administration. For each genotype, there were four treatment groups, two of which were controls. One control was a nonmanipulated group, and the other was treated with saline IT and analyzed 3 h later. No lung inflammation was detected in any of the genotype groups treated with saline IT, and results were similar to nonmanipulated groups. For the 10-µg endotoxin IT group, lung inflammation was studied at 3 and 16 h after injection.

Alveolar inflammatory cells. Three hours after endotoxin IT, alveolar cell numbers were significantly increased in the SP-Bmon⁺ group relative to the wild type (Fig. 6A). Sixteen hours after endotoxin IT, cell numbers were increased five- to sevenfold in all genotype groups, with higher numbers of cells in the SP-Bmon⁺⁺ and SP-Bmon⁺ groups than in the wild-type group (P < 0.05). Cell differentiation analyses (Fig. 6B) showed a significant increase in neutrophils at both 3 and 16 h after endotoxin injection, but no differences were detected among the three genotype groups.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Total cell numbers (A) and neutrophils (B) in alveolar lavage samples from nonmanipulated control group (none), saline intratracheal injection (IT) control group (saline), 3 h after 10-µg endotoxin IT group (3h Endo), and 16 h after endotoxin IT group (16h Endo). Alveolar cell number was higher in SP-Bmon+ mice group than in wild-type mice 3 h after endotoxin IT and increased in all genotype groups 16 h after endotoxin IT compared with control groups. Alveolar cell numbers were higher in SP-Bmon++ and SP-Bmon+ mice than in wild-type mice 16 h after endotoxin. Neutrophils in alveolar lavage samples were similarly increased by endotoxin IT for all genotype groups. * P < 0.05 vs. wild-type group.

Cytokines in lung tissue. TNF-, MIP-2, IL-6, and IL-1 levels in lung tissue homogenates for each group are shown in Figs. 7 and 8. Cytokine levels were very low or undetectable in control groups; cytokine levels in endotoxin-treated groups were maximal at 3 h posttreatment and decreased at 16 h posttreatment. In the SP-Bmon⁺⁺ group, MIP-2 was significantly higher (P < 0.05) than wild type at 3 and 16 h after endotoxin. TNF- at 16 h and IL-1 at 3 h in the SP-Bmon⁺⁺ mice after endotoxin were higher than the in the wild-type group. IL-6 in the SP-Bmon⁺⁺ mice was not significantly different from wild-type mice at both 3 and 6 h. Sixteen hours after endotoxin IT, TNF-, MIP-2, IL-6, and IL-1 in the SP-Bmon⁺ group were significantly higher (P < 0.05) than in the wild-type group. IL-6 in lung tissue in the SP-Bmon⁺ group was higher than in the SP-Bmon⁺⁺ and wild-type mice. SP-Bmon⁺ group tended to show slower recovery from elevated cytokines than the wild-type and SP-Bmon⁺⁺ groups after endotoxin injection.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Tumor necrosis factor (TNF)- (A) and MIP-2 (B) levels in lung tissue. Both cytokines were elevated 3 h after endotoxin IT and decreased after 16 h in all genotype groups. At 16 h after endotoxin, levels of TNF- and MIP-2 were higher in SP-Bmon+ mice than in wild-type and SP-Bmon++ mice. * P < 0.05 vs. wild-type mice;  P < 0.0001 vs. SP-Bmon++ mice.

View larger version (15K): [in this window] [in a new window]   Fig. 8.   IL-6 (A) and IL-1 (B) levels in lung tissue and IL-1 level in alveolar lavage sample (C). Both cytokines were increased by endotoxin in all genotype groups. SP-Bmon+ showed higher levels of IL-6 and IL-1 in lung tissue 3 and 16 h after endotoxin IT relative to wild-type mice. IL-6 level was highest in SP-Bmon+ mice lung tissue. At 16 h after endotoxin IT, IL-1 in alveolar lavage sample in SP-Bmon++ and SP-Bmon+ mice was higher than in wild-type mice. * P < 0.01 vs. wild-type mice;  P < 0.001 vs. SP-Bmon++ mice.

Cytokines in alveolar lavage. In alveolar lavage, both TNF- and MIP-2 levels were not detectable or were <10 pg/ml in the nonmanipulation and saline control groups. Three hours after endotoxin, both cytokines were elevated to similar levels in the three genotype groups (TNF-: 3,319 ± 128 pg/ml; MIP-2: 2,132 ± 112 pg/ml). By 16 h after endotoxin IT, TNF- and MIP-2 levels were decreased similarly in all 3 genotype groups (TNF-: 939 ± 23 pg/ml; MIP-2: 148 ± 18 pg/ml). The mean IL-6 level in alveolar lavage did not show any differences among different genotypes (nonmanipulated and saline group: 40 ± 20 pg/ml; 3-h endotoxin groups: 447 ± 99 pg/ml; 16-h endotoxin groups: 337 ± 30 pg/ml). IL-1 levels in alveolar lavage for all genotypes in the nonmanipulation and saline groups were 6.9 ± 1.1 pg/ml and were increased twofold for all genotypes 3 h after endotoxin IT (mean of 17.3 ± 1.3 pg/ml). After 16 h of endotoxin IT, alveolar IL-1 did not change from the 3-h level for wild-type mice, whereas a 2.3-fold increase (P < 0.05) was detected in the SP-Bmon⁺⁺ group (Fig. 8C). IL-1 was increased 3.7-fold (P < 0.001) in SP-Bmon⁺ mice relative to wild-type mice.

Apoptotic cells. Prolonged recovery from elevated cytokines after intratracheal endotoxin injection was consistently detected in the SP-Bmon⁺ group. To determine whether there were any changes in apoptotic cells after endotoxin IT among the different genotypes, the percentage of apoptotic cells in alveolar lavage was estimated (Fig. 9). As shown in Fig. 6, most of the cells in alveolar lavage were neutrophils. After endotoxin injection, the percentage of apoptotic cells was significantly reduced (P < 0.05) in the SP-Bmon⁺ group.

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Percent apoptotic cells in alveolar lavage was significantly lower in the SP-Bmon+ group than in the SP-Bmon++ group and the wild-type group at 3 and 16 h after endotoxin IT. * P < 0.001 vs. wild type mice;  P < 0.01 vs. SP-Bmon++.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The complete loss of SP-B expression in mice results in lethal neonatal respiratory distress syndrome (RDS), indicating that SP-B is absolutely required for postnatal lung function (5). Transgenic mice that expressed abundant SP-B proprotein but failed to process the precursor to the active peptide also died of neonatal RDS, indicating that formation of the mature peptide is critical for SP-B function (4). The mature peptide is always detected as a disulfide-linked homodimer in mice, suggesting that formation of an intermolecular disulfide bridge is essential for SP-B function. This hypothesis was tested by generating transgenic mice (SP-Bmon⁺) in which endogenous SP-B was replaced with a peptide that could not form the intermolecular disulfide bridge (3). Surprisingly, SP-Bmon⁺ transgenic mice survived and demonstrated relatively small changes in pressure-volume mechanics compared with wild-type littermates (3). The results of in vitro surface activity measurements suggested that increasing the concentration of SP-Bmon peptide might improve lung function in transgenic mice. The present study confirmed this prediction and also demonstrated that the inflammatory response associated with lung injury was significantly muted in mice with elevated expression of the SP-Bmon transgene.

Transgenic mice expressing lower levels of the SP-B transgene (SP-Bmon⁺) were significantly more susceptible to hyperoxic-induced lung inflammation than wild-type mice, whereas the increased concentration of SP-Bmon in SP-Bmon⁺⁺ transgenic mice resulted in significantly less severe lung inflammation, similar to that detected in wild-type mice. The severity of lung inflammation induced by endotoxin IT was greater than that induced by hyperoxia, and more lung inflammation was detected in both the SP-Bmon⁺ and SP-Bmon⁺⁺ transgenic lines than in wild-type mice. Susceptibility to endotoxin-induced lung inflammation was similar for both transgenic lines 3 h after endotoxin IT; however, 16 h after endotoxin IT, recovery from lung injury was delayed in the SP-Bmon⁺ group compared with the SP-Bmon⁺⁺ and wild-type groups. Prolonged recovery from lung inflammation in SP-Bmon⁺ was accompanied by a significantly lower percentage of apoptotic cells in alveolar lavage fluid. Apoptosis, or programmed cell death, is the process that deletes cells from a population in a deliberate manner. Through apoptosis, cells activated by endotoxin-induced inflammation are eliminated to avoid a sustained inflammatory response (7). It is likely that delayed elimination of cells via apoptosis contributed to a sustained inflammatory response in SP-Bmon⁺ mice. Overall, in two different models of lung injury, elevated levels of the SP-Bmon transgene were associated with decreased inflammation.

Although the role of SP-A in modulating inflammatory responses has been extensively investigated (20), only a few studies have examined the immunomodulatory potential of SP-B. Prolonged exposure to endotoxin was associated with decreased levels of SP-B and abnormal surfactant function (9). The results of preliminary experiments (6) indicated that IT endotoxin resulted in more severe inflammation in SP-B(+/) mice relative to wild-type littermates. Furthermore, transgenic mice that overexpressed wild-type mature SP-B peptide were significantly more resistant to endotoxin-induced lung inflammation than SP-B(+/+) or SP-B(+/) mice. The results of the present study are also consistent with a protective role for SP-B; however, whether the protective effect of SP-B is due to a direct anti-inflammatory action or an indirect effect arising from increased resistance to lung injury remains unclear. Evidence for a direct, protective effect of SP-B was suggested by the results of a study in which SP-B inhibited lipopolysaccharide-induced nitric oxide formation by isolated alveolar macrophages in a dose-dependent manner (12). The anti-inflammatory effect was not reproduced by SP-A, SP-C, or surfactant lipids. SP-B may therefore specifically interact with alveolar macrophages to dampen the inflammatory response via a peptide structure that does not require formation of an intersubunit disulfide bridge.

The mechanism underlying the concentration-dependent, anti-inflammatory effect of SP-Bmon may be related to the formation of noncovalently linked SP-B homodimers. Circular dichroism and mass spectrometry of SP-B isolated from SP-Bmon⁺ mice indicated that the peptide was largely monomeric at concentrations of <1 µM (21). However, at concentrations of >2 µM, the peptide formed noncovalent dimers. Molecular modeling of the native SP-B dimer identified hydrogen bonding/ion pairing between glutamine 51 in one subunit and arginine 52 in the other subunit (22). These observations suggest that intersubunit hydrogen bonding/ion pairing is sufficient for homodimer formation and that the intersubunit disulfide bridge simply stabilizes the dimer structure. The importance of the dimer structure for SP-B function is supported by the detection of noncovalently linked SP-B homodimers in SP-Bmon⁺ transgenic mice (21) and the results of a recent study, which showed that the dimeric form of a truncated, synthetic SP-B peptide had significantly better surface properties than the monomeric form (18). Whether intersubunit pairing of glutamine 51 with arginine 52 is absolutely required for dimer formation and the larger issue of whether dimer formation is essential for SP-B function will require ablation of this noncovalent bridge.