HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Free All Versions of this Article: 96/3/1080    most recent 00771.2003v1 Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Yan, X. Articles by Hsia, C. C. W. Search for Related Content PubMed PubMed Citation Articles by Yan, X. Articles by Hsia, C. C. W.

Retinoic acid induces nonuniform alveolar septal growth after right pneumonectomy

Departments of ¹Internal Medicine, ²Pathology, and ³Cardiovascular and Thoracic Surgery, University of Texas Southwestern Medical Center, Dallas, Texas 75390-9034

Submitted 24 July 2003 ; accepted in final form 20 October 2003

	ABSTRACT

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
To determine whether all-trans retinoic acid (RA) enhances compensatory lung growth in fully mature animals, adult male dogs (n = 4) received 2 mg·kg^-1·day^-1 po RA 4 days/wk beginning the day after right pneumonectomy (R-PNX, 55-58% resection). Litter-matched male R-PNX controls (n = 4) received placebo. After 4 mo, the remaining lung was fixed by tracheal instillation of fixatives at a constant airway pressure for detailed morphometric analysis. After RA treatment compared with placebo, lung volume was slightly but not significantly lower. Volume density of septum to lung was 37% higher because of a 50 and 25% higher volume density of capillary and septal tissue, respectively. Mean septal thickness was 27% higher. Absolute volumes of endothelial cells and capillary blood were 31-37% higher, whereas epithelial and interstitial volumes were not different between groups. Absolute alveolar-capillary surface areas did not differ between groups, and alveolar septal surface-to-volume ratio was 20% lower in RA-treated animals. RA treatment exaggerated interlobar differences in morphometric indexes and caused alveolar capillary morphology to revert to a more immature state. Thus RA treatment during early post-R-PNX adaptation preferentially enhanced alveolar capillary and endothelial cell volumes consistent with formation of new capillaries, but the associated septal distortion precluded a corresponding increase in gas-exchange surface or morphometric estimates of lung diffusing capacity.

compensatory lung growth; dog; lung morphometry; endothelial cell volume; alveolar capillary growth

We have extensively characterized a robust model of postpneumonectomy (PNX) lung growth in adult dogs to define structure-function relationships during compensatory lung growth. In adult dogs, right PNX (55% lung resection) triggers compensatory growth of alveolar tissue in the remaining lung that partially restores septal tissue volumes, alveolar-capillary surface areas, and lung diffusing capacities toward those in two normal lungs. Our aim was to determine the in vivo effects of exogenous RA on post-PNX lung growth and function. Our hypotheses were 1) exogenous RA enhances compensatory alveolar growth in the remaining lung after right PNX and 2) RA-enhanced cellular growth improves lung function. These hypotheses were admittedly simplistic; however, given the enthusiasm for applying RA to patients with chronic lung disease (26), it is important to understand its histological and physiological consequences in a whole animal model where the loss of alveoli is reproducible, the remaining lung is not "injured" in the typical sense, and adaptive changes can be readily quantified. We administered all-trans RA to adult dogs during the first 4 mo after right PNX, the period of the most vigorous compensatory response. This paper reports the analysis of lung structure, while assessment of lung function is reported in the companion paper (6).

	METHODS

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Animal procedures. All procedures were approved by the Institutional Animal Care and Use Committee. A flow chart of the experimental design is shown in Fig. 1. Eight adult litter-matched male foxhounds underwent right PNX at 1 yr of age. The surgical procedure and drug administration are detailed in the companion article (6). Beginning 1 day after right PNX, four animals received all-trans RA (Sigma Chemical, 2 mg/kg po dissolved in vegetable oil), whereas four littermates received placebo (equal volume of oil). The drug or placebo was administered once daily, 4 days/wk over 4 mo. A drug holiday was provided 3 days per week to minimize the induction of drug metabolism. The drug or placebo was mixed with 1 tablespoon of honey and 1 tablespoon of peanut butter and fed into the dog's mouth. This method of administration was kindly suggested by Drs. Donald and Gloria Massaro (Lung Biology Laboratory, Georgetown University). Oral all-trans RA is rapidly absorbed. The chosen dose (2 mg·kg^-1·day^-1) was higher than that previously given intraperitoneally to rats (0.5 mg·kg^-1·day^-1) (29, 43) and orally to emphysema patients in pilot clinical studies (1 mg·kg^-1·day^-1) (26) but well below the level known to cause toxicity in dogs (5-10 mg·kg^-1·day^-1, Aronex Pharmaceuticals, The Woodlands, TX, now Antigenics). Dogs were fed a standard unrestricted diet. After 3 mo of drug administration, resting physiological studies and high-resolution CT scans of the chest were performed at rest under anesthesia, described in the companion paper (6).

View larger version (20K): [in this window] [in a new window]   Fig. 1. Flow chart of experimental protocol. PNX, pneumonectomy; CT, computed tomography.

Terminal procedure. After 4 mo of drug administration, animals were fasted overnight, premedicated with atropine (0.05 mg/kg sc), deeply anesthetized with pentobarbital sodium (25 mg/kg iv), and intubated via a tracheostomy with a cuffed endotracheal tube tied securely to the trachea with umbilical tape. The animal was mechanically ventilated at a tidal volume of 15 ml/kg and a rate of 12 breaths/min. The abdomen was opened via a midline incision. The ventilator was disconnected, and a rent was made through the left hemidiaphragm to allow collapse of the left lung. Then the left lung was reinflated within the thorax by tracheal instillation of 2.5% buffered glutaraldehyde at 25 cmH₂O above the highest point of the sternum. An overdose of pentobarbital sodium (100 mg/kg iv) was administered simultaneously. After the flow of fixatives was stopped, the endotracheal tube was clamped, and after 30 min the lungs and heart were removed en bloc, immersed in 2.5% buffered glutaraldehyde in a plastic bag, floated on a water bath, and stored at 4°C for 1 mo before further processing.

Lung volume measurement. The fixed left lung was divided into upper and lower strata by separately clamping the lobar bronchus to the lower lobe. The upper stratum consisted of the upper and middle lobes, which were often incompletely separated. The lower stratum consisted of the lower lobe. Volume of each intact stratum was measured by the immersion method of Scherle and coworkers (49) with the clamps in place to maintain airway pressure. Each stratum was then sliced serially at 2-cm intervals with an electric knife, with the first cut placed at a random orientation. The face of each section was photographed by a 35-mm Nikon camera using Kodak tungsten color film. A volume estimate of the sectioned lung was obtained from the photographs by point counting using the Cavalieri principle (34), i.e., measuring the slice area by point counting, multiplying by slice thickness, and summing over all slices. A comparison of these methods in our laboratory is detailed elsewhere (52). Only lung volume estimated by the Cavalieri principle, when the tissue was free from tension, was used in subsequent morphometric calculations.

Sampling, tissue processing, and morphometric analysis. A previously established four-level stratified analytical scheme was employed: gross (level 1, about x2), low-power light microscopy (level 2, x275), high-power light microscopy (level 3, x550), and electron microscopy (EM; level 4, x19,000) (47). For level 1, photographs of the 2-cm serial sections were analyzed by point counting using standard test grids to exclude structures larger than 1 mm in diameter, to estimate the volume density of coarse parenchyma per unit lung volume. For level 2, four blocks were sampled per stratum (total 8 blocks per dog) by a systematic random scheme, embedded in glycol methacrylate for thick sections (5 µm) stained with toluidine blue. One section per block was overlaid with a test grid. From a random start, at least 10 nonoverlapping microscopic fields were systematically sampled at x275 magnification. By using point counting, structures between 20 µm and 1 mm in diameter were excluded to estimate the volume density of fine parenchyma per unit volume of coarse parenchyma.

For levels 3 and 4, four blocks were sampled per stratum (8 blocks per dog) by a systematic random scheme, postfixed with 1% osmium tetroxide in 0.1 M cacodylate buffer, treated with 2% uranyl acetate, dehydrated through graded alcohol, and then embedded in Spurr. Each block was sectioned at 1-µm thickness and stained with toluidine blue. One section per block was overlaid with a test grid at x550 magnification. From a random start, at least 20 nonoverlapping microscopic fields per block were systematically imaged (80 images per stratum) to estimate the volume density of alveolar septa per unit volume of fine parenchyma by excluding all structures exceeding 20 µm in diameter (level 3). The fine parenchyma comprised the gas-exchange region and served as reference space for EM analysis (47).

For level 4 analysis, two blocks per stratum were sectioned at 80-nm thickness and mounted on copper grids. Each grid was examined with a JEOL EXII transmission electron microscope at approximately x19,000 magnification. Thirty nonoverlapping EM fields per grid (60 images per stratum) were sampled systematically from a random start, captured with a charge-coupled device camera (Gatan, model C73-0200) and projected onto a Sony high-resolution monitor overlaid with a test grid. Images were also digitized by computer. Septal cells were identified by conventional criteria of typical morphological characteristics (13). The volume densities of epithelium (types I and II), interstitium, and endothelium were estimated by point counting. The alveolar epithelial and capillary surface densities were estimated by intersection counting. At least 300 points or intersections were counted per grid, yielding a coefficient of variation below 10%. The length of test lines (l) that transect the barrier from the epithelial surface to the nearest red cell membrane were measured to calculate harmonic mean thickness of the tissue-plasma barrier (_hb)

(1)

Morphometric estimation of lung diffusing capacity. Diffusing capacities for oxygen (DL_O2) were calculated by a modified version (48) of the morphometric model established by Weibel (46). The model describes the diffusion path from alveolar air to capillary hemoglobin as two serial conductances through the tissue-plasma barrier (Db_O2) and within capillary erythrocytes (De_O2)

(2)

(3)

(4)

SA and Sc are the measured alveolar and capillary surface areas, respectively, and Vc is the pulmonary capillary blood volume. Kb_O2 is the Krogh diffusion coefficient for O₂ in tissue and plasma (5.5 x 10^-10 cm²·s^-1·Torr^-1) (50). _O2 is the rate of O₂ uptake by dog whole blood measured in vitro; we used a value of _O2 = 0.06466 ml O₂·(ml blood·s·Torr)^-1 based on standard equations (42) and average values previously measured in dogs during treadmill exercise, i.e., rectal temperature 40°C, hemoglobin concentration 19.0 g/dl, and mean alveolar O₂ tension of 100 Torr. Estimates of diffusing capacity by this model correlate strongly with that measured by physiological methods at peak exercise in the same animal (42).

Absolute volume and surface area of each structure were calculated by relating the respective volume and surface densities at each level back through the cascade of levels to the volume of the stratum measured by the Cavalieri principle (47). Data were calculated for each stratum separately; then a volume-weighted average for the entire lung was obtained.

Immunohistochemical localization of RA in lung tissue. To ensure that the orally administered drug accumulated in lung tissue and exerted detectable cellular effects, three additional normal adult dogs received oral 2 mg/kg po all-trans RA for 3 or 7 days or 4 days/wk over 90 days. At the end of the treatment period, these animals were anesthetized as described above for PNX animals. The left lung was exposed via a lateral thoracotomy through the fifth intercostal space. The left upper lobe was tied with silk ligature and removed; tissue samples were taken from several different regions of the lobe and fixed in Bouin's solution. Separate samples (50-100 mg wet wt) were snap-frozen in liquid nitrogen for molecular assays (see below). The animal was then euthanized, and the remaining lobes were fixed in situ by tracheal instillation of fixatives as described above.

To demonstrate the accumulation of RA in alveolar tissue, the above lung samples fixed in Bouin's solution were embedded in paraffin and sectioned at 4 µm. Randomly selected sections were deparaffinized in xylene in stages, hydrated through graded alcohols, incubated with 0.1% H₂O₂, and rinsed with PBS buffer. Sections were incubated with a mouse monoclonal antibody against all-trans RA that was purified from a hybridoma cell line (a generous gift of Dr. Maija Zile, Dept. of Food Science and Human Nutrition, University of Michigan) (54). Labeling was detected by immunoperoxidase staining (Vectastain Elite ABC kit, Vector Laboratories) with hematoxylin counterstain. Normal untreated dog lung tissue (10) served as controls.

Immunoblot of surfactant protein-A and pro-surfactant protein-C expression. To demonstrate the biological activity of administered RA within alveolar tissue, we performed immunoblot assays on the above lung samples taken from the left upper lobe of three additional adult dogs given RA for 3, 7, and 90 days compared with lung tissue taken from the left upper lobe of 1-yr-old untreated adult dogs (10) as controls. Tissues were minced on ice and homogenized by Polytron in a buffer containing 300 mM sucrose, 20 mM Tris, pH 8.0, 10 mM HEPES, 5 mM EGTA, 2 mM -mercaptoethanol, and protease inhibitors aprotinin (5 µg/ml), pepstatin A (1.5 µM), leupeptin (100 µM), trypsin inhibitor (5 µg/ml), p-aminobenzamidine (200 µM), and phenylmethylsulfonyl fluoride (1 mM). Homogenates were cleared by centrifugation (15,000 g) and centrifuged at 100,000 g to separate membrane-bound and soluble protein fractions. Total protein content in each fraction was quantified by Bradford assay (Bio-Rad Laboratories, Hercules, CA).

The soluble supernatant was assayed for surfactant protein (SP)-A and pro-SP-C. Proteins were resolved by SDS-PAGE (20 µg/lane), transferred onto polyvinylidene difluoride membranes, blocked in Blotto-Tween solution (5% nonfat dry milk, 0.05% Tween in PBS) for 1 h, and incubated in Blotto-Tween with the primary antibody for at least 2 h. Blots were washed in PBS-0.1% Tween for 30 min, incubated in peroxidase-labeled goat anti-rabbit secondary antibody for 1 h, and then washed again in PBS-Tween for 30 min. Labeled protein was visualized by enhanced chemiluminescence (ECL, Amersham, Piscataway, NJ). The primary antibodies were polyclonal rabbit anti-human SP-A, 1:150 (gift from Carole Mendelson, Dept. of Biochemistry, University of Texas Southwestern Medical Center) and polyclonal rabbit anti-human pro-SP-C (Chemicon International, Temecula, CA) 1:1,000. Two different tissue samples from each animal were analyzed.

Data analysis. Results were analyzed both in absolute values and after normalization by the terminal body weight. Paired data from littermates were compared by paired t-test and/or repeated-measures ANOVA. A difference of P < 0.05 was considered to be significant.

	RESULTS

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
There was no clinical evidence of RA toxicity, and whole body hemoglobin concentration did not change during treatment. We did not follow blood chemistry or liver enzymes in this cohort. In a subsequent cohort given the same RA regimen after left PNX, we monitored these parameters at monthly intervals and found no significant alteration (unpublished observation). Body weights were similar between groups at the onset of the study (placebo: 24.3 ± 1.2 kg, RA: 24.2 ± 2.1 kg, means ± SD). In RA-treated dogs, body weight increased 11% in the first 2 mo and then stabilized, whereas body weight did not change in the placebo group throughout the study period (P < 0.05) (Table 1). Because it was highly unlikely that weight gain in these adult dogs reflected further skeletal growth, we elected to report the absolute measurements without normalization by body weight, so the short-term weight change in one group does not bias data interpretation.

Morphometric results. After RA treatment compared with placebo, absolute volume of the fixed lung was slightly but not significantly smaller, and pulmonary capillary hematocrit was similar (Table 1). The RA-treated lung showed a significantly thickened septum (by 28%) as well as septal crowding with a 37% higher volume density of septum per unit lung volume (both P < 0.05) (Table 2). The thicker septum was the result of a higher volume density of tissue as well as volume density of capillary blood per unit lung volume (Table 2, Fig. 2). Volume densities of individual cell compartments per unit volume of septum was not significantly different between groups, but because of septal crowding the absolute volumes of septal endothelium and capillary blood were significantly elevated in the RA-treated group (Table 3); the increase in endothelial cell volume was statistically significant even after normalization of the data by the terminal body weight (placebo: 0.401 ± 0.043 ml/kg, RA: 0.469 ± 0.039 ml/kg; means ± SE, P < 0.05 by paired t-test).

View larger version (137K): [in this window] [in a new window]   Fig. 2. Representative light (A) and electron (B) micrographs. After retinoic acid (RA) treatment, alveolar septa were thickened associated with capillary engorgement.

Despite the larger septal tissue and blood volume, alveolar surface area per unit volume of septum was significantly smaller by 20%, whereas capillary surface area per unit septal volume was not different in RA-treated lungs compared with placebo. Absolute alveolar surface area was not different in the RA-treated group. Absolute capillary surface area was modestly higher in the RA-treated group compared with placebo, but the difference did not reach statistical significance (P = 0.09). However, the ratio of alveolar epithelial surface-toseptal volume was significantly lower in RA-treated lungs compared with placebo, whereas the ratio of capillary surface-to-septal volume was not different between groups, suggesting either a loss of surface complexity at the air-tissue interface (fewer epithelial folds) or greater convolution of capillaries with respect to alveolar surface.

Morphometric estimates of diffusing capacity. The higher capillary blood volume in the RA group contributed to a significantly (37%) higher erythrocyte conductance for O₂ (De_O2) than in the placebo group. Because the mean harmonic barrier thickness and alveolar-capillary surface areas were not different between groups, estimates of O₂ diffusing capacity of the membrane barrier (Db_O2) was also not different between groups. Overall, DL_O2 was slightly but not significantly elevated in the RA treated group (Table 3). When normalized by the terminal body weight, none of the components of diffusing capacity was significantly different between groups.

Regional response to RA. In the placebo group, estimates of volume and surface areas were not different between the upper and lower strata (Table 4). However, in response to RA, the major structural changes were more pronounced in the lower stratum than in the upper stratum.

Quantification of double septal capillary profiles. An incidental observation was that double alveolar capillary profiles were more frequently encountered in RA-treated lungs compared with placebo. This was intriguing because double capillary profiles are typical of the developing lung and only infrequently observed in the fully mature adult lung. To quantify this observation, we systematically sampled EM grids at x1,000 magnification and counted the number of capillary intercepts with the standard test grid used in level 4 analysis. Each capillary profile that intercepts a test line was classified as a single (only one capillary profile along a given portion of the septum) or double (two separate capillary profiles overlap the same portion of the septum) capillary. Two grids (one from upper and one from lower stratum) were systematically and completely counted per animal, resulting in over 200 total intercepts per animal.

Figure 3 shows the number of double capillaries expressed as a percent of total (double + single) capillary profiles. In RA-treated dogs the percentage of double capillaries was significantly higher compared with placebo (6.65 ± 2.95%, 3.49 ± 1.40%, respectively, n = 4 each, means ± SD, P < 0.001). In contrast, in separate adult dogs after left PNX where compensatory lung growth does not normally occur, the double capillary count was 2.76 ± 0.62% (n = 4), which was significantly lower than in RA-treated animals after right PNX (P < 0.0001) but insignificantly lower than placebo-treated animals after right PNX (P > 0.05).

View larger version (101K): [in this window] [in a new window]   Fig. 3. Relative frequency of double septal capillary profiles was significantly increased in RA-treated lungs.

Localization and cellular effects of RA in lung tissue. In Fig. 4, normal untreated dog lung tissue shows little staining for all-trans RA. Staining increased after oral RA administration, reflecting tissue accumulation of the drug. Labeling for all-trans RA localized to the alveolar septal tips, where new alveolar tissue growth is most likely to originate. Labeling was also evident within alveolar macrophages, which became more abundant after exogenous RA treatment.

View larger version (82K): [in this window] [in a new window]   Fig. 4. Immunolabeling of all-trans RA in lung tissue increased with the duration of RA treatment. RA label (brown) localized to the end tips of alveolar septa and alveolar macrophages, which were more abundant after 90 days (d) of RA treatment.

Figure 5 shows the representative immunoblot in three normal dogs given oral RA for 3, 7, or 90 days compared with three untreated normal adult dogs. In RA-treated animals, SP-A expression was markedly higher at day 3 compared with untreated lungs but diminished with longer treatment. In contrast, pro-SP-C expression did not change appreciably with RA treatment. These qualitative results were consistent in replicate assays, although the small number of animals precludes a quantitative comparison. Nonetheless, these results in Fig. 4 and 5, combined with the above morphometric results, demonstrate that our oral regimen of RA administration elicited a significant biological response in lung tissue.

View larger version (49K): [in this window] [in a new window]   Fig. 5. Immunoblot of surfactant protein (SP)-A and pro-SP-C protein expression in lung tissue from 3 normal dogs receiving RA for 3, 7, and 90 days compared with untreated normal control dog lungs. SP-A protein level was markedly higher after 3 days of RA treatment, but the increase did not persist with longer treatment duration. Expression of pro-SP-C did not change with RA treatment.

	DISCUSSION

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
Summary of results. In the adult dog, right PNX increases mechanical strain of the remaining lung and induces progressive compensatory growth of new alveolar tissue (18). Supplementing the diet with RA during the period of most active post-PNX cellular adaptation selectively enhances endothelial cell and capillary growth without enhancing cell growth of other alveolar cell types. This nonuniform or "dysanaptic" enhancement of septal growth by RA did not significantly increase gas-exchange surfaces or morphometric estimates of lung diffusing capacity during 4 mo of treatment but was associated with a significantly lower alveolar surface area-toseptal volume ratio, suggesting a loss of alveolar surface complexity and/or distorted alveolar geometry. RA treatment also altered septal capillary morphology to that more typical of an immature developing lung. Structural response to RA was more exaggerated in the lower lobe than in the upper or middle lobes. Results show that preferential manipulation of one septal cell type may not optimize overall alveolar structure or function because a favorable microscopic response (enhanced cell proliferation) may lead to unfavorable macroscopic consequences (tissue distortion) that limit the overall benefit of treatment.

Critique of methods. The optimal dose and duration of RA treatment remain undefined. We chose a dose of all-trans RA that elicits definite physiological, cellular, and morphological responses in the lung without significant toxicity. We also chose a reasonable treatment duration that covers the period of most active post-PNX cellular proliferation and growth. It seems unlikely that an even larger dose or longer duration of treatment would yield a higher risk-benefit ratio, given that the incidence of RA toxicity will almost certainly increase. We carefully litter matched the groups to reduce the number of animals needed to demonstrate intergroup differences. Because significant changes in key parameters could be documented with four pairs of animals studied at a single time point with a single-dose regimen, more animal cohorts seemed unnecessary. Because feeding was not restricted, the initial weight gain in RA-treated animals may have been due to an increased appetite or altered metabolism. Whether the data were analyzed in absolute values or after normalization by the terminal body weight did not materially alter the key conclusions of this study.

Compensatory growth of alveolar septum. In adult large animals, intensity of the compensatory response is directly related to the amount of lung resected. Resection of 45% of lung by left PNX is well tolerated by the remaining lung without invoking compensatory growth of new lung tissue (17). On the other hand, resection of 55% of lung by right PNX triggers regenerative growth of new alveolar tissue that partially restores cell volumes, surface areas, and gas-exchange capacities toward normal (18). The major signal for compensatory alveolar growth is mechanical strain experienced by the remaining lung (20), which when exceeding a critical threshold initiates a cascade of molecular and biochemical events to generate a balanced pattern of cell growth. Compensatory alveolar growth after right PNX progresses through two phases spanning several months. In the first phase, cell proliferation and/or hypertrophy increases the volumes of epithelium and endothelium similarly (2.5-fold) but increases the interstitium disproportionately (3.5- to 4.0-fold) (18), leading to a longer harmonic mean tissue-plasma barrier length and a high resistance to diffusion. In this period, gas-exchange surface area, lung diffusing capacity, and aerobic capacity show little compensation (19). Later, septal remodeling consists primarily of thinning and pruning of the interstitium to match that of other tissue compartments, leading to a reduction of harmonic mean barrier length for diffusion, as well as allowing spreading of the increased epithelial and endothelial cell volumes to create a greater surface for gas exchange. Thus alveolar cell growth does not translate immediately into functional enhancement until overall septal architecture normalizes. Uniform growth preserves a normal septal surface-to-volume ratio and optimizes gas-exchange efficiency of the remaining lung.

Nonuniform alveolar septal growth induced by RA. Information regarding in vivo effects of RA on endothelial cell growth is extremely limited, and in vitro data are conflicting. Some studies show that RA suppresses VEGF expression, angiogenesis, and endothelial cell migration (8, 23, 51), actions that are believed to promote cell differentiation and inhibit oncogenesis; however, others report a positive effect of RA on endothelial cells, either directly or via modulation of another angiogenic mediator (12, 25, 44). In this study, RA treatment after PNX primarily enhances growth of alveolar endothelial cells and alters capillary morphology. Selective enhancement of capillary and endothelial growth without epithelial or interstitial growth distorts septal architecture, causing a low septal surface-to-volume ratio and no increase in gas-exchange surface areas. The net effect is less functional improvement than normally expected from a given increase of alveolar septal volume density. In the present animals, lung diffusing capacity, assessed by independent morphometric and physiological methods, either failed to increase or actually decreased during RA supplementation. This is an example of dysanaptic or nonuniform septal growth that illustrates a general but important concept, i.e., an isolated pharmacological agent or growth factor is unlikely to reproduce the entire spectrum of a natural response that requires the coordination of all tissue constituents. Selective manipulation of one or a few cell types leads to distortions of normal structure-function relationships even within a unit as small as an alveolus. Such distortion may ultimately limit global functional compensation and hence the clinical utility of the pharmacological agent.

It is possible that capillary surface area may increase to some extent following the completion of alveolar remodeling after cessation of RA treatment. However, 4 mo of RA treatment post-PNX clearly had little effect on either the epithelium or interstitium, making it very unlikely that epithelial surface or harmonic mean barrier length (key determinants of membrane diffusing capacity) will become augmented after cessation of treatment or after a longer observation period. Even if the capillary surface is able to enlarge further with remodeling, a mismatch between the endothelial and epithelial surfaces will likely persist.

Effect of RA on capillary morphology. Double septal capillary profiles are typical of the developing lung. As the lung matures, the septum thins out and only a single capillary profile is evident along any given portion of the septum, creating a thin side and a thick side to the diffusion barrier. Double capillaries are infrequently observed in the normal adult lung; however, they became nearly twice as frequent in RA-treated animals compared with placebo controls. Thus RA treatment caused the alveolar microvasculature to revert to a more immature morphological state. This unexpected observation directly supports the conclusion that RA treatment distorts alveolar septal architecture. The prevalence of double capillaries is also consistent with the formation of neocapillaries by the process of intussusception (22), which has been observed in the chicken chorioallantoic membrane as well as in the postnatal developing lung. In this process, a tissue pillar grows into the lumen of an existing capillary, eventually dividing it into two capillaries. This is one possible mechanism by which RA-enhanced endothelial cell volume translates into growth of capillaries. This mechanism is also consistent with the lack of a corresponding increase in gas-exchange surface areas after RA treatment, which would occur only if the mature single-capillary morphology is restored.

Regional differences in RA action. That RA exaggerates the interlobar structural differences after R-PNX is also unexpected. Normally pulmonary blood flow is preferentially distributed to the dorsocaudal portions of the dog lung even under isogravitational conditions (1); regional nonuniformity could be exaggerated after PNX as blood flow to the remaining lung doubles, resulting in greater endothelial strain and/or shear in the caudal portions of the remaining lung. Thus one mechanism of RA action may be amplification of strain-related endothelial cell response (7), leading to greater stimulation of capillary endothelial growth in the lower lobe than in the upper or middle lobes.

Molecular markers of RA action. Three adult dogs were given all-trans RA for 3, 7, or 90 days to demonstrate the adequacy of the dose and the biological activity of oral RA administration. RA preferentially localized to the tips of alveolar septa and alveolar macrophages, suggesting that RA might stimulate formation of new septa or, more likely in the adult animal, lengthening of existing septa. Because this is not a systematic investigation of the molecular actions of RA and these few animals (n = 3) do not allow a quantitative comparison, we did not survey all the surfactant proteins nor did we choose to examine the RA receptors. Even so, a distinct elevation of SP-A but not SP-C protein level has emerged from whole lung immunoassays in RA-treated dogs, suggesting an early induction of SP-A expression by RA in vivo that does not persist with longer treatment.

Surfactant proteins are expressed in different patterns during lung growth. During postnatal development, SP-A and pro-SP-C protein levels are inversely related to indexes of cell proliferation. However, 3 wk after right PNX, SP-A is elevated in direct relation to heightened cell proliferative activity, whereas pro-SP-C does not change compared with matched controls (10). SP-D is modestly higher, whereas pro-SP-B does not change significantly during either postnatal or post-PNX lung growth. Reported effects of RA on surfactant protein expression are variable. In human fetal lung explants (33), RA treatment reduces SP-A protein as well as SP-A and -C mRNA levels and increases SP-B mRNA (14, 15), but RA has no effect on SP-A mRNA in human pulmonary adenocarcinoma cells (15). RA interferes with the expression of SP-A, -B, and -C in the developing fetal lung (4) and inhibits dexamethasone-induced increase in SP-A and -B mRNA in vitro (38). In addition, maternal administration of retinyl palmitate decreases SP-A level in fetal rat lung (11). These studies suggest a suppressive effect of RA on surfactant protein metabolism, but other studies suggest a positive effect. Vitamin A deficiency in fetal rats reduces mRNA and protein levels of surfactant proteins in one study (5) but not another (53). Bogue et al. (2) reported that RA causes a dose-dependent increase in SP-A, -B, and -C mRNA levels in fetal rat lung explants. Mechanisms underlying the divergent responses are unclear, but a RA response element consensus sequence has been identified in the rat SP-A gene, and RA-receptor complexes may act on the SP-A gene via this response element (2).

Summary and conclusions. In adult dogs after right PNX (55% resection) where increased mechanical strain normally induces compensatory alveolar growth in the remaining lung, RA supplementation during the most active period of cellular proliferation and/or hypertrophy selectively enhances septal endothelial cell and capillary growth without enhancing the growth of other septal cell types. This nonuniform alveolar septal growth results in distortion of alveolar architecture and prevents a corresponding increase in alveolar surface area or lung diffusing capacity during the treatment period. Beyond the immediate goal of clarifying the therapeutic utility of RA, these results provide a model for understanding the fundamental determinants of compensatory lung growth when a balanced natural response is selectively manipulated. RA-induced distortion of septal architecture and consequently the mechanical interactions among alveolar components cannot be predicted from its in vitro action at a cellular or molecular level but can profoundly influence the outcome by limiting physiological improvement independent of cellular growth, as shown in the companion paper (6).

	ACKNOWLEDGMENTS

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 
We thank Debbie Tuttle Hogg and Richard Hogg for technical assistance in animal studies, Laurie Task and the staff of the Animal Resource Center for animal care, Dr. Maija H. Zile and Dr. Carole R. Mendelson for the gift of antibodies, Drs. Donald Massaro, Gloria Massaro, and David Mangelsdorf for helpful advice regarding RA administration, Jean Wang for assistance with tissue processing, and Dr. Ewald R. Weibel for valuable comments and suggestions.

GRANTS

This study was supported by National Heart, Lung, and Blood Institute Grants R01 HL-45716, HL-62873, HL-40070, and HL-54060.

	FOOTNOTES

 

Address for reprint requests and other correspondence: C. C. W. Hsia, Pulmonary and Critical Care Medicine, Dept. of Internal Medicine, 5323 Harry Hines Blvd., Dallas, TX 75390-9034.

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. Section 1734 solely to indicate this fact.

	REFERENCES

TOP ABSTRACT METHODS RESULTS DISCUSSION ACKNOWLEDGMENTS REFERENCES

 

1. Beck KC and Rehder K. Differences in regional vascular conductances in isolated dog lungs. J Appl Physiol 61: 530-538, 1986.[Abstract/Free Full Text]
2. Bogue CW, Jacobs HC, Dynia DW, Wilson CM, and Gross I. Retinoic acid increases surfactant protein mRNA in fetal rat lung in culture. Am J Physiol Lung Cell Mol Physiol 271: L862-L868, 1996.[Abstract/Free Full Text]
3. Cardoso WV, Mitsialis SA, Brody JS, and Williams MC. Retinoic acid alters the expression of pattern-related genes in the developing rat lung. Dev Dyn 207: 47-59, 1996.[CrossRef][Web of Science][Medline]
4. Cardoso WV, Williams MC, Mitsialis SA, Joyce-Brady M, Rishi AK, and Brody JS. Retinoic acid induces changes in the pattern of airway branching and alters epithelial cell differentiation in the developing lung in vitro. Am J Respir Cell Mol Biol 12: 464-476, 1995.[Abstract]
5. Chailley-Heu B, Chelly N, Lelievre-Pegorier M, Barlier-Mur AM, Merlet-Benichou C, and Bourbon JR. Mild vitamin A deficiency delays fetal lung maturation in the rat. Am J Respir Cell Mol Biol 21: 89-96, 1999.[Abstract/Free Full Text]
6. Dane DM, Yan X, Tamhane RM, Johnson RL Jr, Estrera AS, Hogg DC, Hogg RT, and Hsia CCW. Retinoic acid-induced alveolar cellular growth does not improve function after right pneumonectomy. J Appl Physiol 96: 1090-1096, 2004.[Abstract/Free Full Text]
7. Dekker RJ, van Soest S, Fontijn RD, Salamanca S, de Groot PG, VanBavel E, Pannekoek H, and Horrevoets AJ. Prolonged fluid shear stress induces a distinct set of endothelial cell genes, most specifically lung Kruppel-like factor (KLF2). Blood 100: 1689-1698, 2002.[Abstract/Free Full Text]
8. Diaz BV, Lenoir MC, Ladoux A, Frelin C, Demarchez M, and Michel S. Regulation of vascular endothelial growth factor expression in human keratinocytes by retinoids. J Biol Chem 275: 642-650, 2000.[Abstract/Free Full Text]
9. Federspeil SJ, DiMari SJ, Guerry-Force ML, and Haralson MA. Extracellular matrix biosynthesis by cultured fetal rat lung epithelial cells. II. Effects of acute exposure to epidermal growth factor and retinoic acid on collagen biosynthesis. Lab Invest 63: 455-466, 1990.[Web of Science][Medline]
10. Foster DJ, Yan X, Bellotto DJ, Moe OW, Hagler HK, Estrera AS, and Hsia CCW. Expression of epidermal growth factor and surfactant proteins during postnatal and compensatory lung growth. Am J Physiol Lung Cell Mol Physiol 283: L981-L990, 2002.[Abstract/Free Full Text]
11. Fraslon C and Bourbon JR. Retinoids control surfactant phospholipid biosynthesis in fetal rat lung. Am J Physiol Lung Cell Mol Physiol 266: L705-L712, 1994.[Abstract/Free Full Text]
12. Gaetano C, Catalano A, Illi B, Felici A, Minucci S, Palumbo R, Facchiano F, Mangoni A, Mancarella S, Muhlhauser J, and Capogrossi MC. Retinoids induce fibroblast growth factor-2 production in endothelial cells via retinoic acid receptor alpha activation and stimulate angiogenesis in vitro and in vivo. Circ Res 88: E38-E47, 2001.[Web of Science][Medline]
13. Gehr P, Bachofen M, and Weibel ER. The normal human lung: ultrastructure and morphometric estimation of diffusion capacity. Respir Physiol 32: 121-140, 1978.[CrossRef][Web of Science][Medline]
14. George TN, Miakotina OL, Goss KL, and Snyder JM. Mechanism of all-trans retinoic acid and glucocorticoid regulation of surfactant protein mRNA. Am J Physiol Lung Cell Mol Physiol 274: L560-L566, 1998.[Abstract/Free Full Text]
15. George TN and Snyder JM. Regulation of surfactant protein gene expression by retinoic acid metabolites. Pediatr Res 41: 692-701, 1997.[Web of Science][Medline]
16. Gray T, Koo JS, and Nettesheim P. Regulation of mucous differentiation and mucin gene expression in the tracheobronchial epithelium. Toxicology 160: 35-46, 2001.[CrossRef][Web of Science][Medline]
17. Hsia CCW, Fryder-Doffey F, Stalder-Navarro V, Johnson RL Jr, and Weibel ER. Structural changes underlying compensatory increase of diffusing capacity after left pneumonectomy in adult dogs [Published erratum in J Clin Invest 93: 913, 1994]. J Clin Invest 92: 758-764, 1993.[Web of Science][Medline]
18. Hsia CCW, Herazo LF, Fryder-Doffey F, and Weibel ER. Compensatory lung growth occurs in adult dogs after right pneumonectomy. J Clin Invest 94: 405-412, 1994.[Web of Science][Medline]
19. Hsia CCW, Herazo LF, Ramanathan M, Johnson RL Jr, and Wagner PD. Cardiopulmonary adaptations to pneumonectomy in dogs. II. Ventilation-perfusion relationships and microvascular recruitment. J Appl Physiol 74: 1299-1309, 1993.[Abstract/Free Full Text]
20. Hsia CCW, Wu EY, Wagner E, and Weibel ER. Preventing mediastinal shift after pneumonectomy impairs regenerative alveolar tissue growth. Am J Physiol Lung Cell Mol Physiol 281: L1279-L1287, 2001.[Abstract/Free Full Text]
21. Koo JS, Yoon JH, Gray T, Norford D, Jetten AM, and Nettesheim P. Restoration of the mucous phenotype by retinoic acid in retinoid-deficient human bronchial cell cultures: changes in mucin gene expression. Am J Respir Cell Mol Biol 20: 43-52, 1999.[Abstract/Free Full Text]
22. Kurz H, Burri PH, and Djonov VG. Angiogenesis and vascular remodeling by intussusception: from form to function. News Physiol Sci 18: 65-70, 2003.[Abstract/Free Full Text]
23. Lachgar S, Charveron M, Gall Y, and Bonafe JL. Inhibitory effects of retinoids on vascular endothelial growth factor production by cultured human skin keratinocytes. Dermatology 199, Suppl 1: 25-27, 1999.[Web of Science][Medline]
24. Liebeskind A, Srinivasan S, Kaetzel D, and Bruce M. Retinoic acid stimulates immature lung fibroblast growth via a PDGF-mediated autocrine mechanism. Am J Physiol Lung Cell Mol Physiol 279: L81-L90, 2000.[Abstract/Free Full Text]
25. Maeno T, Tanaka T, Sando Y, Suga T, Maeno Y, Nakagawa J, Hosono T, Sato M, Akiyama H, Kishi S, Nagai R, and Kurabayashi M. Stimulation of vascular endothelial growth factor gene transcription by all-trans retinoic acid through Sp1 and Sp3 sites in human bronchioloalveolar carcinoma cells. Am J Respir Cell Mol Biol 26: 246-253, 2002.[Abstract/Free Full Text]
26. Mao JT, Goldin JG, Dermand J, Ibrahim G, Brown MS, Emerick A, McNitt-Gray MF, Gjertson DW, Estrada F, Tashkin DP, and Roth MD. A pilot study of all-trans retinoic acid for the treatment of human emphysema. Am J Respir Crit Care Med 165: 718-723, 2002.[Abstract/Free Full Text]
27. Marcus R and Coulston AM. Fat-soluble vitamins. In: Goodman & Gilman's The Pharmacological Basis of Therapeutics, edited by Hardman JG and Limbird LE. New York: McGraw-Hill, 1996, p. 1573-1590.
28. Massaro GD and Massaro D. Postnatal treatment with retinoic acid increases the number of pulmonary alveoli in rats. Am J Physiol Lung Cell Mol Physiol 270: L305-L310, 1996.[Abstract/Free Full Text]
29. Massaro GD and Massaro D. Retinoic acid treatment abrogates elastase-induced pulmonary emphysema in rats [published erratum appears in Nat Med 3: 805, 1997]. Nat Med 3: 675-677, 1997.[CrossRef][Web of Science][Medline]
30. Massaro GD and Massaro D. Retinoic acid treatment partially rescues failed septation in rats and in mice. Am J Physiol Lung Cell Mol Physiol 278: L955-L960, 2000.[Abstract/Free Full Text]
31. McGowan S, Jackson SK, Jenkins-Moore M, Dai HH, Chambon P, and Snyder JM. Mice bearing deletions of retinoic acid receptors demonstrate reduced lung elastin and alveolar numbers. Am J Respir Cell Mol Biol 23: 162-167, 2000.[Abstract/Free Full Text]
32. McGowan SE, Doro MM, and Jackson SK. Endogenous retinoids increase perinatal elastin gene expression in rat lung fibroblasts and fetal explants. Am J Physiol Lung Cell Mol Physiol 273: L410-L416, 1997.[Abstract/Free Full Text]
33. Metzler MD and Snyder JM. Retinoic acid differentially regulates expression of surfactant-associated proteins in human fetal lung. Endocrinology 133: 1990-1998, 1993.[Abstract/Free Full Text]
34. Michel RP and Cruz-Orive LM. Application of the Cavalieri principle and vertical sections method to lung: estimation of volume and pleural surface area. J Microsc 150: 117-136, 1988.[Web of Science][Medline]
35. Nabeyrat E, Besnard V, Corroyer S, Cazals V, and Clement A. Retinoic acid-induced proliferation of lung alveolar epithelial cells: relation with the IGF system. Am J Physiol Lung Cell Mol Physiol 275: L71-L79, 1998.[Abstract/Free Full Text]
36. Oberg KC and Carpenter G. EGF-induced PGE2 release is synergistically enhanced in retinoic acid treated fetal rat lung cells. Biochem Biophys Res Commun 162: 1515-1521, 1989.[CrossRef][Web of Science][Medline]
37. Oberg KC, Soderquist AM, and Carpenter G. Accumulation of epidermal growth factor receptors in retinoic acid-treated fetal rat lung cells is due to enhanced receptor synthesis. Mol Endocrinol 2: 959-965, 1988.[Abstract/Free Full Text]
38. Oshika E, Liu S, Singh G, Michalopoulos GK, Shinozuka H, and Katyal SL. Antagonistic effects of dexamethasone and retinoic acid on rat lung morphogenesis. Pediatr Res 43: 315-324, 1998.[Web of Science][Medline]
39. Randell SH, Mercer RR, and Young SL. Postnatal growth of pulmonary acini and alveoli in normal and oxygen-exposed rats studied by serial section reconstructions. Am J Anat 186: 55-68, 1989.[CrossRef][Web of Science][Medline]
40. Rosenthal FS. Investigation of retinoic acid therapy in experimental emphysema in dogs (Abstract). Am J Respir Crit Care Med 165: A230, 2002.
41. Schuger L, Varani J, Mitra R Jr, and Gilbride K. Retinoic acid stimulates mouse lung development by a mechanism involving epithelial-mesenchymal interaction and regulation of epidermal growth factor receptors. Dev Biol 159: 462-473, 1993.[CrossRef][Web of Science][Medline]
42. Takeda S, Hsia CCW, Wagner E, Ramanathan M, Estrera AS, and Weibel ER. Compensatory alveolar growth normalizes gas exchange function in immature dogs after pneumonectomy. J Appl Physiol 86: 1301-1310, 1999.[Abstract/Free Full Text]
43. Tepper J, Pfeiffer J, Aldrich M, Tumas D, Kern J, Hoffman E, McLennan G, and Hyde D. Can retinoic acid ameliorate the physiologic and morphologic effects of elastase instillation in the rat? Chest 117: 242S-244S, 2000.[CrossRef][Medline]
44. Tokuda H, Hatakeyama D, Akamatsu S, Tanabe K, Yoshida M, Shibata T, and Kozawa O. Involvement of MAP kinases in TGF-beta-stimulated vascular endothelial growth factor synthesis in osteoblasts. Arch Biochem Biophys 415: 117-125, 2003.[CrossRef][Web of Science][Medline]
45. Veness-Meehan KA, Bottone FG Jr, and Stiles AD. Effects of retinoic acid on airspace development and lung collagen in hyperoxia-exposed newborn rats. Pediatr Res 48: 434-444, 2000.[Web of Science][Medline]
46. Weibel ER. Morphometric estimation of pulmonary diffusion capacity. I. Model and method. Respir Physiol 11: 54-75, 1970.[CrossRef][Medline]
47. Weibel ER. Morphometric and stereological methods in respiratory physiology, including fixation techniques. In: Techniques in the Life Sciences: Respiratory Physiology, edited by Otis AB. County Clare, Ireland: Elsevier, 1984, p. 1-35.
48. Weibel ER, Federspiel WJ, Fryder-Doffey F, Hsia CCW, König M, Stalder-Navarro V, and Vock R. Morphometric model for pulmonary diffusing capacity. I. Membrane diffusing capacity. Respir Physiol 93: 125-149, 1993.[CrossRef][Web of Science][Medline]
49. Weibel ER, Kistler GS, and Scherle WF. Practical stereological methods for morphometric cytology. J Cell Biol 30: 23-38, 1966.[Abstract/Free Full Text]
50. Weibel ER, Taylor CR, O'Neil JJ, Leith DE, Gehr P, Hoppeler H, Langman V, and Baudinette RV. Maximal oxygen consumption and pulmonary diffusing capacity: a direct comparison of physiologic and morphometric measurements in canids. Respir Physiol 54: 173-188, 1983.[CrossRef][Web of Science][Medline]
51. Weninger W, Rendl M, Mildner M, and Tschachler E. Retinoids downregulate vascular endothelial growth factor/vascular permeability factor production by normal human keratinocytes. J Invest Dermatol 111: 907-911, 1998.[CrossRef][Web of Science][Medline]
52. Yan X, Polo Carbayo JJ, Weibel ER, and Hsia CC. Variation of lung volume after fixation when measured by immersion or Cavalieri method. Am J Physiol Lung Cell Mol Physiol 284: L242-L245, 2003.[Abstract/Free Full Text]
53. Zachman RD and Grummer MA. Effect of maternal/fetal vitamin A deficiency on fetal rat lung surfactant protein expression and the response to prenatal dexamethasone. Pediatr Res 43: 178-183, 1998.[Web of Science][Medline]
54. Zhou HR, Abouzied MM, and Zile MH. Production of a hybridoma cell line secreting retinoic acid-specific monoclonal antibody. J Immunol Methods 138: 211-223, 1991.[CrossRef][Web of Science][Medline]

This article has been cited by other articles:

				C. C. W. Hsia Quantitative morphology of compensatory lung growth Eur. Respir. Rev., December 1, 2006; 15(101): 148 - 156. [Abstract] [Full Text] [PDF]

				M. D. Roth, J. E. Connett, J. M. D'Armiento, R. F. Foronjy, P. J. Friedman, J. G. Goldin, T. A. Louis, J. T. Mao, J. R. Muindi, G. T. O'Connor, et al. Feasibility of retinoids for the treatment of emphysema study. Chest, November 1, 2006; 130(5): 1334 - 1345. [Abstract] [Full Text] [PDF]

				X. Yan, D. J. Bellotto, D. M. Dane, R. G. Elmore, R. L. Johnson Jr., A. S. Estrera, and C. C. W. Hsia Lack of response to all-trans retinoic acid supplementation in adult dogs following left pneumonectomy J Appl Physiol, November 1, 2005; 99(5): 1681 - 1688. [Abstract] [Full Text] [PDF]

				T. D. Le Cras, L. G. Fernandez, P. A. Pastura, and V. E. Laubach Vascular growth and remodeling in compensatory lung growth following right lobectomy J Appl Physiol, March 1, 2005; 98(3): 1140 - 1148. [Abstract] [Full Text] [PDF]

				C. C. W. Hsia Signals and mechanisms of compensatory lung growth J Appl Physiol, November 1, 2004; 97(5): 1992 - 1998. [Abstract] [Full Text] [PDF]

				P. Ravikumar, C. Yilmaz, D. M. Dane, R. L. Johnson Jr., A. S. Estrera, and C. C. W. Hsia Regional lung growth following pneumonectomy assessed by computed tomography J Appl Physiol, October 1, 2004; 97(4): 1567 - 1574. [Abstract] [Full Text] [PDF]

				Mechanisms and Limits of Induced Postnatal Lung Growth Am. J. Respir. Crit. Care Med., August 1, 2004; 170(3): 319 - 343. [Full Text] [PDF]

				D. M. Dane, X. Yan, R. M. Tamhane, R. L. Johnson Jr., A. S. Estrera, D. C. Hogg, R. T. Hogg, and C. C. W. Hsia Retinoic acid-induced alveolar cellular growth does not improve function after right pneumonectomy J Appl Physiol, March 1, 2004; 96(3): 1090 - 1096. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Free All Versions of this Article: 96/3/1080    most recent 00771.2003v1 Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (10) Citing Articles via Google Scholar Google Scholar Articles by Yan, X. Articles by Hsia, C. C. W. Search for Related Content PubMed PubMed Citation Articles by Yan, X. Articles by Hsia, C. C. W.