Targeted disruption of the surfactant protein (SP) D (SP-D) gene caused a marked pulmonary lipoidosis characterized by increased alveolar lung phospholipids, demonstrating a previously unexpected role for SP-D in surfactant homeostasis. In the present study, we tested whether the local production of SP-D in the lung influenced surfactant content in SP-D-deficient [SP-D(−/−)] and SP-D wild-type [SP-D(+/+)] mice. Rat SP-D (rSP-D) was expressed under control of the human SP-C promoter, producing rSP-D, SP-D(+/+) transgenic mice. SP-D content in bronchoalveolar lavage fluid was increased 30- to 50-fold in the rSP-D, SP-D(+/+) mice compared with the SP-D(+/+) parental strain. Lung morphology, phospholipid content, and surfactant protein mRNAs were unaltered by the increased concentration of SP-D. Likewise, the production of endogenous mouse SP-D mRNA was not perturbed by the SP-D transgene. rSP-D, SP-D(+/+) mice were bred to SP-D(−/−) mice to assess whether lung-selective expression of SP-D might correct lipid homeostasis abnormalities in the SP-D(−/−) mice. Selective expression of SP-D in the respiratory epithelium had no adverse effects on lung function, correcting surfactant phospholipid content and decreasing phosphatidylcholine incorporation significantly. SP-D regulates surfactant lipid homeostasis, functioning locally to inhibit surfactant phospholipid incorporation in the lung parenchyma and maintaining alveolar phospholipid content in the alveolus. Marked increases in biologically active tissue and alveolar SP-D do not alter lung morphology, macrophage abundance or structure, or surfactant accumulation.
- pulmonary surfactant
- alveolar epithelium
- alveolar macrophage
- surfactant protein D
- surfactant protein A
- saturated phosphatidylcholine
the accumulation of alveolar surfactant is regulated at the levels of synthesis, secretion, recycling, and degradation by diverse physiological and humoral stimuli (7). The mechanisms regulating surfactant homeostasis are not completely understood, but it is generally believed that surfactant protein (SP) A, SP-B, SP-C, and SP-D play important roles (3, 17). Contrary to expectations, the generation of mice with targeted deletions of theSP-A gene demonstrated that SP-A was not essential for either normal surfactant accumulation or function (13). SP-D was not previously believed to have a significant role in either surfactant function or homeostasis. Therefore, the observation that targeted deletions of SP-D increased saturated phosphatidylcholine (Sat PC) in the alveolus and lung parenchyma was unexpected (2, 15). Morphological consequences of SP-D deletion included the presence of abnormally large alveolar type II cell lamellar bodies, increased peribronchial lymphoid tissue, and enlargement of alveolar air spaces. Alveolar macrophages accumulated progressively, exceeding six times the normal value by 10 wk of age. Alveolar macrophages in SP-D-deficient [SP-D(−/−)] mice also became larger with advancing age, with increased phospholipid and neutral lipid content (2, 15). Although the mechanism underlying the increased phospholipid content in SP-D(−/−) mice remains unclear, accumulation of alveolar and intracellular Sat PC occurs in the presence of only modest changes in the incorporation of [3H]choline by lung tissue, indicating that alterations in both the synthesis and degradation of phospholipids might contribute to the abnormalities seen in SP-D(−/−) mice (2).
Although SP-D is expressed abundantly in the respiratory epithelium, SP-D is also expressed in various tissues including the gastrointestinal tract, pancreas, and heart (6, 20). It is therefore unclear whether the observed abnormalities in surfactant homeostasis of SP-D(−/−) mice are mediated by the direct effects of SP-D on lung metabolism or indirectly by influencing metabolism in other organs. In normal humans, the normal range of alveolar SP-D abundance varies from 0.7 to 49 ng/nM phospholipid (9). Such differences have no known physiological significance. An apparent inverse relationship between alveolar phospholipid and SP-D and SP-A after acute endotoxin exposure (19) is consistent with a potential physiological role for SP-D in decreasing alveolar surfactant abundance. However, a 50% reduction in the abundance of SP-D does not alter surfactant accumulation or macrophage morphology (15).
In the present study, we sought to determine whether SP-D functioned locally within the lung to regulate surfactant pool size, alveolar macrophage abundance, and morphology and whether the increased expression of SP-D altered lung phospholipid and metabolism or regulated endogenous SP-D expression in the mouse. To increase SP-D concentration in the lung, a construct with the human SP-C promoter (14) and a rat SP-D (rSP-D) cDNA clone (23) was developed to produce transgenic mice [rSP-D, SP-D(+/+)] that expressed SP-D in respiratory epithelial cells. We also bred rSP-D, SP-D(+/+) mice to SP-D(−/−) mice to produce mice expressing rSP-D only in the lung.
The entire rSP-D coding sequence, including the polyadenylation signal, was excised with EcoR I from a previously reported cDNA clone (23) and cloned into a unique EcoR I site in an expression construct between the human SP-C promoter and SV40 polyadenylation sequence (27). One correctly oriented clone was expanded and purified with a Qiagen Plasmid Purification Kit (Qiagen, Hilden, Germany). The construct was then linearized by double digestion with Not I and Nde I and separated from the plasmid sequences on an agarose gel with a DEAE cellulose membrane (NA45, Schleicher & Schuell, Keene, NH) as previously described (22).
Fertilized eggs were obtained from the mating of C57BL/6 × C3H male and female mice. Pronuclei were injected with 4 μg/ml of linearized vector in a volume of 1–2 pl. Ova surviving the microinjection were inoculated into pseudopregnant B6/C3 hybrids obtained from Taconic (Germantown, NY) as previously described (8). Mice were weaned at 3 wk of age, and DNA from tail clips was evaluated for the presence of the transgene by DNA blot analysis after digestion with EcoR I, with the entire rat SP-D cDNA as a probe and hybridization conditions as previously described (23). The transgene was identified by the presence of a 1.3-kb hybridization band corresponding to rat cDNA. This probe also hybridized to a single 8-kb mouse SP-D restriction fragment. Relative abundance of mouse and rat DNAs was estimated by comparing the density of bands with a PhosphorImager (Molecular Dynamics, Sunnyvale, CA). To generate a stable genotype, mice transmitting the transgene in a manner consistent with single autosomal integration with a consistent copy number in successive generations were selected for further study. One strain with approximately two to three copies per genome was used for breeding experiments.
SP-D(−/−) mice were developed as previously described (15) and have been outbred to National Institutes of Health Swiss Black mice four times, with SP-D(−/−) mice recovered after each outbreeding by crossbreeding the resulting SP-D(+/−) progeny. For breeding experiments with SP-D(−/−) mice, an rSP-D, SP-D(+/+) male was bred to three SP-D(−/−) females. The resulting SP-D(+/−) mice that also carried the rSP-Dtransgene were crossed with SP-D(−/−) mice to generate rSP-D, SP-D(−/−) founders. rSP-D, SP-D(−/−) mice were subsequently crossed with SP-D(−/−) mice, yielding SP-D(−/−) and rSP-D, SP-D(−/−) littermates in equal numbers for further study.
SP Gene Expression
To assess mRNA accumulation for SP-A, SP-B, SP-C, and SP-D, RNA blots were probed with rat cDNA clones as previously described (12). All blots were counterprobed with glyceraldehyde-3-phosphate dehydrogenase to verify equal RNA loading. Blots were scanned with a PhosphorImager from Molecular Dynamics. Signal intensity was compared between groups of four mice for each SP, and each experiment was repeated once for a total of eight mice per analysis. Western blots were performed as previously described (15) with an antibody raised in rabbits against purified rSP-D.
To assess endogenous mouse SP-D mRNA accumulation in SP-D(+/+) and rSP-D, SP-D(+/+) mice, we used a competitive PCR (1). Primers specific for mouse or, alternatively, rat mRNA were designed as follows: mouse sense, 5′-CGTCTAGAGGTTGCCTTCTCC-3′, and antisense, 3′-GGCTCAGACCTGTATGTTGCCA-5′; and rat sense, 5′-CTAGAGGCTGCCTTTTCTCG-3′, and antisense, 3′-GGTATAGCTCAGACCTCTATG-5′. Although the sequences of the rat and mouse cDNAs are 92% identical, because of differences in the 3′-untranslated region, the above primer pairs are species specific. Competitors for PCRs were developed by conducting site-directed mutagenesis, deleting an internal Pvu II site in the mouse SP-D cDNA clone and a Pst I site in the rat cDNA clone. After experimental determination of an appropriate dilution of competitor to cDNA to achieve linearity, four 1:2 dilutions of mimic were combined with the master mix containing cDNA obtained from transgenic and normal mouse lung RNA by RT and amplified. Amplifications were done at an annealing temperature of 63°C with primer concentrations of 20 μM each and Life Technologies Platinum PCR SuperMix. On completion of 25 cycles, samples were precipitated with ethanol and subjected to digestion for 2 h with a 10-fold excess of Pvu II for mouse and Pst I for rat amplicons. The efficiency of RT was verified by comparing the intensity with that of amplified β-actin. A fraction of the sample was resolved on a Bio-Rad 10% TBE Ready Gel at 150 V for 1.5 h. The gel was stained with SYBR Green I (FMC BioProducts, Rockland, ME) and scanned on a Molecular Dynamics PhosphorImaging System. Relative abundance was calculated as previously described (1, 24) by comparison of the signal intensity ratios of unknown concentrations of mimic cDNA with the ratios obtained from a known input of cloned cDNA. The results are expressed as picograms of mRNA per microgram of total RNA input. To determine whether accumulation of SP-A varied depending on genotype, bronchoalveolar lavage (BAL) fluid, lung homogenate, and macrophage pellets were all subjected to SP-A ELISA as previously described (18).
The 8-wk-old mice were given a weight-adjusted intraperitoneal injection of 0.03 μCi/g body weight of [3H]choline chloride (American Radiolabeled Chemicals, St. Louis, MO) for the rSP-D, SP-D(+/+) study and 0.07 μCi/g body weight of [3H]choline chloride for the rSP-D(−/−) comparison study. The animals were deeply anesthetized 8 or 24 h later with intraperitoneal pentobarbital sodium. The lungs were processed as previously described (13, 15). In brief, a 20-gauge catheter was tied into the trachea, and an extensive alveolar wash was recovered for each animal. The lungs were homogenized in saline. Sat PC was recovered by chromatography with neutral alumina columns from chloroform-methanol (2:1) extracts of alveolar washes and lung homogenates (13, 15). Isolated Sat PC was divided for radioactivity measurement and phosphorous assay (10). All values are given as means ± SE.
Lungs from three mice of 10–12 wk of age of each genotype, SP-D(+/+), rSP-D, SP-D(+/+), SP-D(−/−), and rSP-D, SP-D(−/−), were perfused free of blood with PBS, inflation fixed with Formalin infused at 20 cmH2O, embedded in paraffin, sectioned, and stained with hematoxylin and eosin. Representative sections that were not identified as to genotype were viewed by four individuals, three of whom had formal training as pathologists, and were scored as normal or abnormal. Abnormalities were described by each individual, and descriptions were compared between individuals. Immunostaining for SP-D was performed as previously described (6). Macrophages from BAL fluid were counted with a hemacytometer. Morphological evaluations of BAL fluid macrophages at the light microscopy level were done with cytospin preparations and LeukoStat stain (Fisher Scientific). Electron microscopy evaluation of BAL fluid macrophages was performed as previously described after fixation of a low-speed pellet of alveolar macrophages in 0.1 M sodium cacodylate, pH 7.2, and 2.5% glutaraldehyde (16). Immunostaining was done as previously described (15, 23).
Differences between groups were evaluated with unpaired two-tailedt-tests, with P < 0.05 considered significant. For assessment of linkage, distribution of genotypes was evaluated with χ2 analysis, with P < 0.05 considered significant.
Fifteen independent rSP-D founders were identified by tail clip DNA blot analysis with rat cDNA as a probe. The copy number of the transgene in founders varied between 2 and 10 copies per genome as estimated by comparison of hybridization band intensity. Two males and two females with four to eight copies per genome were bred to a C57BL/6/C3H hybrid strain. Germ-line transmission was evident for all founders. Further breeding was into a C57/BL/6 background. Transgene transmission was assessed by PCR analysis with rSP-D-specific cDNA primers, and copy number was assessed by DNA blot. Figure 1 is a representation of a DNA blot demonstrating integration of the transgene in the specific strain used for all systematic studies of SP-D transgene expression and subsequent breeding experiments. With an EcoR I digest, a full-length construct hybridization band is apparent at 1.3 kb and another more intense band is apparent at ∼1.6 kb, indicating the integration of at least two and possibly three copies of the rat cDNA.
To confirm expression, RNA blot analysis was conducted with a mouse SP-D cDNA clone as a probe. The intensity of the RNA hybridization signal for the transgene appeared to be greater than that for the normal mouse and was identifiable as transgenic in origin because of a 150-bp size difference (Fig. 2 A). With scanning densitometry of the blots probed with mouse cDNA, a band signal of approximately twice that of the normal mouse was obtained for the transgenic band, indicating at least a threefold increase in total SP-D mRNA in rSP-D, SP-D(+/+) mice. The hybridization intensity of rat mRNA is much greater when rat cDNA is used as a probe, indicating the difficulty of precise quantification of similar but not identical mRNA species with a slightly dissimilar probe. SP-D in the lung homogenate also appeared to be much greater than that in the normal mouse (Fig.2 B). ELISA of lung homogenate and BAL fluid confirmed increased SP-D (Fig. 3). There was no extrapulmonarySP-D transgene expression identified by RT-PCR analysis of stomach, intestine, pancreas, liver, and heart RNAs with rat-specific SP-D primers (data not shown). After four sequential outbreedings of one lineage, the transgene copy number stabilized, with approximately two to three copies per genome. This line was used for all further experiments. For experiments assessing phenotype, mice were heterozygous for the transgene locus. RNA blot and protein blot analyses from these lines are shown in Fig. 2, A and B. Immunohistochemical staining of acid alcohol-fixed sections demonstrated SP-D expression in the alveolar epithelium but more intense immunostaining in the small-airway epithelium. rSP-D mice demonstrated staining that was more intense than that in normal mice (Fig. 4).
Analysis of rSP-D Mice
Lung structure. Three transgenic mice with SP-D expression of at least 20-fold of the normal value by comparison of Western blots from transgenic and normal littermates were studied at 10–12 wk of age. The lungs were removed from the thorax and inflation fixed to a pressure of 20 cmH2O with Formalin and examined after routine hematoxylin and eosin staining (Fig.5). There were no consistent abnormalities identified between rSP-D, SP-D(+/+) and SP-D(+/+) mice when they were subjected to blinded analysis by four investigators. The number of alveolar macrophages recovered from BAL fluid was identical between normal mice and littermates expressing the transgene (1.45 ± 1.4 × 106 and 2.0 ± 1.1 × 106 cells in normal and transgenic mice, respectively; P = 0.26;n = 11 each), and there were no structural abnormalities of alveolar macrophages evaluated by both light (Fig. 5) and electron microscopy (data not shown). Differential counts were identical between the normal and overexpressing mice (data not shown).
SP gene expression. RNA blot analysis was conducted to estimate SP-A, SP-B, and SP-C mRNA abundance. There were no differences in mRNA for SP-A, SP-B, or SP-C between normal and rSP-D mice (Fig.6). To determine whether expression of therSP-D transgene reduced the accumulation of steady-state mRNA for endogenous mouse SP-D, primer pairs were designed that were specific for rat or, alternatively, mouse SP-D cDNA. Specificity was confirmed with cDNA clones for both rSP-D and mouse SP-D and also by performing RT-PCR with RNA from mouse and rat lungs as templates. In mice with an ∼20-fold increase in SP-D as assessed by Western blot and ELISA, no detectable decrease in endogenous mouse SP-D mRNA was observed [5.4 ± 3.33 × 10−2 and 5.98 ± 2.98 × 10−2 pg/μg of RNA for SP-D(+/+) and rSP-D, SP-D(+/+) mice, respectively; P = 0.85; n = 6 each]. Thus expression of the transgenic rSP-D did not decrease the expression of the endogenous SP-D gene. Concentrations of SP-A were also measured in lung lavage fluid and homogenates by ELISA, revealing no differences between wild-type [SP-D(+/+)] and rSP-D mice (data not shown).
Surfactant Sat PC Measurement
Twenty-four SP-D(+/+) and nineteen rSP-D, SP-D(+/+) mice were used for surfactant measurements. There were no significant differences in body weight between the groups. Alveolar, lung, and total (alveolar plus lung tissue) pool sizes of Sat PC were not different between SP-D(+/+) and rSP-D, SP-D(+/+) mice (Fig.7 A). Likewise, there were no differences in the incorporation of [3H]choline into Sat PC between the normal andrSP-D transgenic mice 8 (Fig. 7 B) and 24 (Fig.7 C) h after injection.
Generation of Rescued rSP-D, SP-D(−/−) Mice
To express the rat transgene in the SP-D(−/−) genotype, three rSP-D, SP-D(+/+) males were bred to three SP-D(−/−) females and backcrossed to generate the following genotypes: 4 SP-D(−/−), rSP-D+ (13%); 13 SP-D(+/−), rSP-D+ (43%); 1 SP-D(+/−), rSP-D−(3%); and 12 SP-D(−/−), rSP-D− (40%). With the assumption of single autosomal integration of SP-D-rSP-D, the distribution of genotypes suggests linkage disequilibrium and implies physical linkage between the rat transgene and the mouse SP-Dlocus (P ≤ 0.001 by χ2 analysis). Three backcrosses to the SP-D(−/−) genotype were performed with 21 pups, 9 SP-D(−/−) and 12 rSP-D, SP-D(−/−), also suggesting single autosomal integration of the rSP-Dtransgene.
Expression of rSP-D in SP-D(−/−) Mice
Although rSP-D and mouse SP-D are nearly identical, it was possible that rSP-D might not be biologically active in mice. Furthermore, because the endogenous SP-D is expressed in various tissues, it was unclear whether pulmonary abnormalities in SP-D(−/−) mice were directly related to the local actions of SP-D in the lung. Therefore, we determined whether rSP-D would correct lung abnormalities in SP-D(−/−) mice. Lung histology, including macrophage morphology, normalized in the corrected rSP-D, SP-D(−/−) mice (Fig. 5, C, D, G, and H). Additionally, alveolar and lung homogenate Sat PC values were similar in SP-D(+/+) and rSP-D, SP-D(−/−) mice and significantly lower in SP-D(−/−) mice (Fig.8 A). Therefore, expression of rSP-D in SP-D(−/−) mice corrected Sat PC to that seen in SP-D(+/+) mice. Incorporation of [3H]choline into Sat PC was significantly reduced in rSP-D, SP-D(−/−) mice compared with that in their SP-D(−/−) littermates at 8 h (Fig. 8 B). Therefore, the rat transgene was biologically active, correcting the abnormalities observed in the lungs of SP-D(−/−) mice.
The distal respiratory epithelial cell-specific expression of rSP-D in the mouse had no effect on lung morphology, surfactant phospholipid content, metabolism, or SP mRNAs in vivo. However, the local expression of rSP-D in the lungs of SP-D gene-targeted mice fully corrected the increased surfactant phospholipid content and reduced the incorporation of [3H]choline into lung phospholipids. SP-D functions locally to regulate lung phospholipid synthesis and content. Likewise, the local production of SP-D corrected abnormalities in alveolar macrophage morphology and mononuclear cell infiltrates seen in the SP-D(−/−) mice.
Although SP-D mRNA has been detected in various tissues including the lung, heart, stomach, and kidney (20), the present findings demonstrate that cell-selective expression of recombinant SP-D restored the abnormalities in histology and lung surfactant concentrations seen in the SP-D(−/−) mice. In the normal lung, SP-D mRNA and protein are detected primarily in respiratory epithelial cells, including nonciliated respiratory epithelial cells or Clara cells, and alveolar type II cells in rat (9, 26), human (4), and mouse (28) lungs. In the present study, the 3.7-kb human SP-C promoter was used to express the rat SP-D transgene. As previously described (14), gene expression with the construct was confined to the lung, likely representing expression in both peripheral bronchiolar and alveolar type II epithelial cells as observed repeatedly with thisSP-C gene promoter. Marked increases in SP-D abundance were demonstrated with antibodies that recognize both rSP-D and mouse SP-D but with a relatively higher affinity for rSP-D. The extent to which SP-D is elevated is difficult to quantify precisely, but when this antibody was used to probe equal loads of recombinant mouse SP-D or, alternatively, rSP-D on an immunoblot, the signal intensity for mouse SP-D was ∼50% of that for rSP-D. Therefore, the observed increases in SP-D in rSP-D(+/+) mice are likely to be at least 20-fold. Such markedly increased concentrations of SP-D in rSP-D, SP-D(+/+) mice did not cause abnormalities in lung histology or surfactant homeostasis. Although the equivalence of rSP-D and mouse SP-D has not been established firmly, the present finding that expression of rSP-D in SP-D(−/−) mice corrected the abnormalities in surfactant homeostasis supports the biological activity of rSP-D in the mouse model. The amino acid sequences of the mouse SP-D and rSP-D molecules share ∼95% identity (National Center for Biotechnology Information), providing further support for the likely activity of recombinant SP-D across species. Furthermore, the finding that the local expression of SP-D corrected pulmonary surfactant abnormalities supports the concept that the effects of SP-D are mediated by cell signaling within the lung rather than by possible effects of SP-D on metabolism in other tissues.
Surfactant phospholipid synthesis, as assessed by metabolic labeling studies, is increased in the lungs of SP-D(−/−) mice (15). In the present work, the rate of [3H]choline incorporation into Sat PC seen in the SP-D gene-targeted mice was significantly reduced by expression of rSP-D cDNA in the lung, demonstrating that SP-D regulates surfactant homeostasis, at least in part, by inhibiting synthesis of Sat PC by type II epithelial cells. The consistent and progressive alterations in macrophage morphology seen in the SP-D(−/−) mice in vivo (2, 15) and in previous in vitro experiments (25, 29) demonstrating the effects of SP-D on alveolar macrophage function support the possibility that the local effects of SP-D on surfactant concentrations are mediated by a SP-D-dependent change in alveolar macrophage function. It is increasingly clear that alveolar macrophages play a critical role in surfactant phospholipid and protein catabolism in the lung. Marked accumulations of both SPs and lipids are associated with abnormalities in alveolar macrophage function in granulocyte-macrophage colony-stimulating factor (GM-CSF)-deficient [GM-CSF(−/−)] and the GM-CSF β-chain receptor-deficient mice in vivo (5, 11, 21). Metabolic clearance of both SPs and lipids are markedly decreased in the lungs of the GM-CSF(−/−) mice, with no accompanying changes in their synthesis (11). The present findings demonstrating the local effects of SP-D on surfactant homeostasis also implicate the role of SP-D on type II cell metabolism but do not clarify whether SP-D also influences alveolar macrophage function. It is therefore unclear whether the lipid accumulations and foamy macrophages seen in SP-D(−/−) mice are secondary to the increased phospholipid content in the lung or are related to primary abnormalities in macrophage function.
Surfactant lipid content was markedly increased in SP-D(−/−) mice, comparable to the increased concentrations of phospholipids observed in the GM-CSF(−/−) mice with alveolar proteinosis. However, unlike the GM-CSF(−/−) mice, SPs did not accumulate in the lungs of SP-D(−/−) mice and were not influenced by increased expression of SP-D in the lung. Paradoxically, SP-A mRNA and protein concentrations were decreased in the lungs of SP-D(−/−) mice (15). In the present study, mouse SP-A mRNA was not altered by expression of rSP-D in SP-D(+/+) mice. High levels of expression of rSP-D in the wild-type mice did not decrease endogenous mouse SP-D mRNA in vivo. Thus although SP-D restores normal surfactant lipid homeostasis, endogenous SP-Dgene expression was not directly regulated locally by the increased content of SP-D in the lung. Although it is possible that the failure of rSP-D to decrease endogenous mouse SP-D expression is related to changes in protein structure between the species, close sequence conservation in rSP-D and mouse SP-D makes this explanation unlikely. The present findings support a model in which SP-D expression is controlled by factors distinct from the level of SP-D in the alveolus.
In summary, the present findings demonstrate that the local expression of SP-D corrects surfactant lipid content and reduces choline incorporation into Sat PC within the lung, supporting the hypothesis that SP-D regulates the metabolic functions of type II epithelial cells. It remains unclear whether SP-D also has a direct effect on macrophage function. SP-D appears to play a novel role in the regulation of lipid homeostasis, distinct from that seen in other forms of alveolar proteinosis in which both lipids and SPs accumulate, the latter effects being mediated primarily by abnormalities in alveolar macrophage-dependent clearance. The present findings support the hypothesis that SP-D plays a primary role in the regulation of surfactant metabolism by type II epithelial cells in the lung.
This work was funded by National Heart, Lung, and Blood Institute Grants HL-41320 (to J. H. Fisher), HL-41496 (to J. A. Whitsett), HL-28623 (to J. A. Whitsett and T. R. Korfhagen), and HL-61612 (to F. X. McCormack); National Institute of Child Health and Human Development Grant HD-11932 (to M. Ikegami); and National Institute of Environmental Health Sciences Grant P30-ES-06639 (to Y.-S. Ho).
Address for reprint requests and other correspondence: J. H. Fisher, Denver Health Medical Center, 777 Bannock St., Denver, CO 80204 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society