Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Intrauterine tracheal obstruction (TO) is one of the most potent stimuli of fetal lung growth known to date (1). Although described more than three decades ago (7), the subject of intrauterine TO has gained renewed interest recently, as this intervention was found to provide lung growth and improve lung function in animal models of pulmonary hypoplasia (8). The potential therapeutic effects of intrauterine TO have been confirmed in the first clinical trials for severe forms of lung hypoplasia associated with congenital diaphragmatic hernia (12, 13).

The mechanisms regulating TO-induced intrauterine lung growth remain only partially understood. It is believed that mechanical stretch, secondary to increased intratracheal pressure (4, 14, 15), is a critical stimulatory force after TO, analogous to its pivotal role in normal fetal lung development and compensatory lung growth (16, 17). In vitro studies have demonstrated that mechanical stretch can induce cell division in fetal lung explants (18), cultured lung fibroblasts (19), and cultured fetal alveolar type II cells (20), and it is conceivable that mechanical factors play a major role in the post-TO growth spurt as well. In addition to mechanical stretch, TO may be regulated by soluble growth factors present in the lung liquid (4). Whether these soluble factors, presumably autocrine/paracrine growth factors produced by epithelial and mesenchymal cells, are primary or secondary effectors (secreted in response to increased mechanical distension), remains to be elucidated. Finally, the maternal/fetal surgical stress caused by TO may result in changes in the level of glucocorticoids, thyrotropin-releasing hormone, and catecholamines, analogous to the events described during childbirth (21).

Important drawbacks to widespread clinical application of intrauterine TO include the observation that prolonged occlusion in animals causes qualitative and quantitative deficiencies of the type II cell population (5, 6, 22, 23). Type II cells are of critical clinical importance because they produce and secrete surfactant, are progenitors of the gas-exchanging type I cells, and are involved in ion transport in the alveolus (24). In addition, a wide range of responses to TO, ranging from failure of growth to exaggerated growth and hydrops, has been observed in the few clinical reports (12, 13). This varied response may be due to a lack of standardization in these human fetuses with congenital diaphragmatic hernia and, in particular, to the variation in gestational age.

The aim of the present study is to analyze the effects of timing and duration of TO on the growth kinetics of lungs and type II cells. Fetal rabbits underwent TO at 24 and 27 d gestational age (DGA), corresponding to late pseudoglandular and late canalicular/early terminal sac stages, respectively (16 to 20 wk versus 24 to 28 wk in humans). We have previously reported the time course of lung and type II cell growth during the first 5 d after early TO (24 DGA) (6). In the present study, we extended the studied time interval after early TO from 5 to 6 d, and compared the growth kinetics following early TO with those induced by late TO, performed at 27 DGA. The growth kinetics of lungs and type II pneumocytes were studied by a combination of 5'-bromo-2'-deoxyuridine (BrdU) pulse labeling and stereologic volumetry, as previously described (6). The findings of this study indicate that the effects of intrauterine TO on fetal lung growth are determined by the lung maturity at the time of the intervention.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Tracheal Occlusion

Time-mated New Zealand white rabbits were obtained at 20 to 22 DGA (term = 31 DGA) (Millbrook Farm, Amherst, MA). They were housed in separate cages under standard laboratory conditions, allowed free access to water and chow, and acclimated to their new environments. At 24 or 27 DGA, following a 12-h fast, each doe was premedicated with xylazine (5 mg/kg) and ketamine (20 mg/kg) intramuscularly. Cefazolin sodium (100 mg/kg) and medroxyprogesterone acetate (10 mg) were given intramuscularly at separate sites. The abdomen was shaved. The rabbits were placed in supine position, intubated, and maintained under halothane anesthesia (1 to 3% in oxygen).

The gravid bicornuate uterus was exposed through a midline abdominal incision and the ovarian-end fetuses assigned to either TO or control (C) groups. TO was limited to two alternate fetuses located at the ovarian end of the uterus. After fetal orientation was determined by gentle palpation, a purse-stringed antimesometrial hysterotomy was made over the fetal head, using electrocautery to incise all layers of the uterine wall. The fetal head and neck were exposed, a midline cervical incision was made, and the trachea was identified. In TO fetuses, a mini-hemoclip was placed across the trachea and the fetus returned to the amniotic cavity. C fetuses consisted of sham-operated contralateral and adjacent littermates who underwent a hysterotomy and neck dissection. The uterine wall was closed in one layer using a running, noninterlocking 5-0 polyglycolic acid suture (Vicryl, Ethicon; Johnson & Johnson, Cincinnati, OH). Maternal laparotomies were closed in two layers using 3-0 Vicryl (fascia) and 4-0 chromic (skin) sutures.

Animal Death and Tissue Processing

The animals were killed at time intervals ranging from 8 h to 6 d after occlusion (25 to 30 DGA for animals occluded at 24 DGA, and 27.3 to 31 DGA for fetuses occluded at 27 DGA). At 2 h before death, the animals were weighed and intraperitoneally injected with BrdU (Sigma Chemical Co., St. Louis, MO) dissolved in phosphate-buffered saline (PBS) at a dose of 50 mg/kg. The does were killed (Beuthanasia-D Special; Shering-Plough Animal Health Corp., Kenilworth, NJ) (0.2 ml/kg, intracardiac). The midline incision was opened, exposing the uterus; fetuses were killed in utero by cardiac puncture before the first breath. All procedures and protocols were approved by the Rhode Island Hospital Animal Care and Use Committee.

The weights of the fetuses and of both lungs were recorded. The left lung was then weighed separately and fixed in toto by immersion in freshly prepared 4% paraformaldehyde in PBS, pH 7.4. After overnight fixation at 21°C, the volume of the left lung was estimated by the volume displacement method (25). The lung was dehydrated in graded ethanol solutions and embedded in paraffin. Sections, 4 µm thick, were prepared and stained with hematoxylin and eosin (H&E).

Stereologic Volumetry of Lungs

Morphometry of the various lung compartments was performed using standard stereologic volumetric techniques as described previously (6, 26). Randomly sampled H&E-stained tissue sections from the left lung were analyzed using a computerized image analysis system (Olympus BX-40 microscope; Olympus America, Melville, NY) interfaced via a CCD video camera (KP-161; Hitachi, Norcross, GA) to a Power Macintosh 7100/80AV (Apple Corp., Cupertino, CA) equipped with software for image analysis (Image NIH 1.59 for Macintosh; National Institutes of Health, Bethesda, MD). Data derived from measurements of the left lung were extrapolated to both lungs.

The critical data set and hierarchical equations, obtained by examining the lung at increasing levels of magnification, comprised:

Volume of lung (V[lu]). The volume of fixed lung was determined according to the Archimedes principle (25). Because the lungs of experimental and control animals were processed identically, no correction was made for dehydration and embedding.

Volume of parenchyma: V(pa) = A_A(pa/lu) × V(lu). Subsequent steps in the structural hierarchy involved point-counting methods based on computer-assisted image analysis. The number of fields to be examined for each type of measurement was determined by testing the reproducibility of results in a pilot study. The parenchymal areal density (A_A[pa/lu]) was estimated by dividing the number of points falling on parenchyma (lung excluding large-sized bronchi and blood vessels) by the number of points falling on the entire lung (original magnification ×10). The total parenchymal volume was calculated by multiplying A_A(pa/lu) by V(lu).

Volume of air-exchanging parenchyma: V(ae) = A_A(ae/ pa) × V(pa). The areal density of air-exchanging parenchyma (A_A[ae/pa]) was estimated by studying random fields of peripheral lung parenchyma and dividing the number of points falling on air-exchanging parenchyma (peripheral lung parenchyma excluding air space) by the number of points falling on the entire field (tissue and air space) (original magnification ×100). The total of volume of air-exchanging parenchyma was calculated by multiplying A_A(ae/pa) by V(pa).

BrdU Immunohistochemistry

The proliferation marker BrdU is a thymidine analogue that is incorporated in cells in the S phase of the cell cycle. For immunohistochemical detection of BrdU-labeled nuclei, lung sections were treated with hydrogen peroxide, denatured in 2 N HCl, washed in borate buffer (0.1 M, pH 8.0), and incubated with monoclonal anti-BrdU antibody (Boehringer, Ingelheim, Germany). Bound antibody was detected using the ABC immunoperoxidase system (27), followed by 3,3'-diaminobenzidine tetrahydrochloride (DAB) treatment and a light hematoxylin counterstain. The BrdU labeling of peripheral and central air-exchanging parenchyma was evaluated separately. Peripheral (subpleural) lung parenchyma was defined as air-exchanging parenchyma located within one ×40 high-power field of the pleura. The total number of nuclei in each of these compartments, and the number of BrdU-positive nuclei therein, were assessed semiautomatically using a computerized image-analysis system. Results were expressed as the percentage of cells positive for BrdU (BrdU labeling index [BrdU-LI]). For each animal, we counted at least 5,000 cells per air-exchanging zone, derived from at least 10 randomly selected fields. Controls for specificity consisted of omission of the primary antibody.

Surfactant Immunohistochemistry and Type II Pneumocyte Morphometry

Random sections of left lung were processed for avidin- biotin complex (ABC) immunoperoxidase staining (27) using a polyclonal goat antirabbit surfactant protein (SP)- A antiserum, kindly provided by Prof. Dr. K. Sueishi (28) (First Department of Pathology, Kyushu University, Fukuoka, Japan), that reacts with type II and Clara cells. After treatment with DAB (Sigma), sections were lightly counterstained with hematoxylin, cleared, and mounted. Controls for specificity consisted of incubation with nonimmune goat serum and omission of the primary antibody.

Immunohistochemical anti-SP-A staining of type II and Clara cells produced a higher optical density than that of the background, allowing the SP-A-positive area to be evaluated semiautomatically. The SP-A-positive areal density (A_A[pnII/ae]), representing the SP-A immunoreactive area per unit area of air-exchanging parenchyma, was determined by dividing the points falling on SP-A-immunoreactive cells by the points falling on air-exchanging parenchyma (original magnification ×200). To avoid inclusion of SP-A-positive Clara cells in the measurements, morphometric analysis was limited to the air-exchanging parenchyma, with exclusion of bronchi and bronchioles. At least 50 randomly selected fields were analyzed per lung. For each section, the light intensity was standardized by threshold calibration using surfactant-negative interstitial tissue as standard.

Surfactant and BrdU Immunohistochemical Double Labeling

The proliferative rate of type II cells was determined by double immunostaining for BrdU and surfactant. Hereto, sections were denatured in 2 N HCl, washed in borate buffer (0.1 M, pH 8.0), and incubated with monoclonal anti-BrdU antibody (Boehringer), followed by detection using the ABC immunoperoxidase system (27) as described previously. After treatment with DAB, the sections were incubated with polyclonal antirabbit SP-A antiserum. Bound anti-SP-A antibody was visualized using the ABC-alkaline phosphatase system (Dako, Glostrup, Denmark) and naphthol-AS-MX-phosphate/fast red TR salt (Sigma) as chromogen. To validate the results of the double immunostaining procedure, sequential sections were stained independently for BrdU and SP-A using the same chromogens as in the double-staining procedure. The enzyme reactions yielded brown reaction products in the nucleus (BrdU) and red deposits in the cytoplasm (SP-A). SP-A-positive and double BrdU-positive and SP-A-positive cells were counted manually (×400 original magnification). Results were expressed as the percentage of type II pneumocytes positive for BrdU (Pn II BrdU-LI). At least 2,000 type II cells from 10 randomly selected microscope fields were counted per lung. Analysis was limited to the central parenchyma, excluding the subpleural zones.

Data Analysis

Statistical analysis was performed using unpaired Student's t test for comparison of C and TO groups at each time point. P < 0.05 was considered statistically significant. Values are expressed as means ± standard deviation (SD) or, where appropriate, as means ± standard error of the mean (SEM). Statview software (Abacus, Berkeley, CA) was used for all statistical work.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Body and Organ Weights

The body weights of fetuses that had undergone TO at 24 DGA (TO-24 fetuses) were comparable with those of sham- operated control littermates (C-24 fetuses) at all time points studied (Figure 1A). Similarly, experimental and C fetuses operated on at 27 DGA (TO-27 and C-27, respectively) had equivalent body weights (Figure 1B). Whereas the body weights of C-24 and TO-24 fetuses showed a steady increase with gestation, there was an apparent stagnation of growth in C-27 and TO-27 fetuses around 29 DGA. This growth retardation was transient, because by 30 DGA the body weights of C-27 and TO-27 fetuses were again in the same range as those of the 24-DGA group.

View larger version (12K): [in this window] [in a new window]   Figure 1.   Body weight of (A) C-24 and TL-24 fetuses and (B) C-27 and TL-27 fetuses. Values represent means ± SD of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2.

Lung weight normalized for body weight (LW/BW) was remarkably constant in C fetuses (C-24 as well as C-27) and ranged between 2.3 and 2.7% (Figure 2). In TO-24 fetuses, LW/BW showed a mild (15%) increase between 26 and 27 DGA, followed by a more pronounced (50%) increase between 28 and 29 DGA (Figure 2A). In TO-27 fetuses, a rapid moderate (35%) increase of LW/BW was noted between 28 and 29 DGA (Figure 2B). At 30 and 31 DGA, LW/BW of TO-24 and TO-27 fetuses was more than twice that of C-24 and C-27 fetuses, respectively.

View larger version (13K): [in this window] [in a new window]   Figure 2.   LW/BW of (A) C-24 and TL-24 fetuses and (B) C-27 and TL-27 fetuses. Values represent means ± SD of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2. *P < 0.05, P < 0.01 versus C.

Lung Morphology

Lungs at 24 DGA were in the late pseudoglandular stage of development, characterized by glycogen-rich peripheral buds and rosettes showing only mild degrees of elongation and tortuosity (Figure 3A). At 27 DGA, the lungs displayed more advanced branching of the airways, flattening of the epithelial lining of the potential air spaces, and diminished interstitial cellularity, corresponding to the late canalicular/early terminal sac stage of fetal lung development (Figure 3B). In both TO-24 and TO-27 groups, TO resulted in accelerated architectural maturation that was evident from 28 DGA on. Occluded lungs after this time point showed enhanced distention of the air spaces, thinning and decreased cellularity of the interalveolar septa, and flattening of the epithelium compared with C lungs (Figures 3C and 3D). The morphologic appearance of bronchial and vascular structures was comparable between TO and C groups.

View larger version (155K): [in this window] [in a new window]   View larger version (134K): [in this window] [in a new window]   View larger version (139K): [in this window] [in a new window]   View larger version (106K): [in this window] [in a new window]   Figure 3.   Lung morphology. (A) C lung at 24 DGA, late pseudoglandular stage. (B) C lung at 27 DGA, late canalicular/early terminal sac stage. (C ) C lung at 30 DGA, late terminal sac stage. (D) TO-24 lung at 30 DGA shows attenuated and relatively acellular alveolar septa with crest formation, corresponding to the early alveolar stage of development. Similar findings were seen in TO-27 lungs at 30 and 31 DGA. (H&E, original magnification ×200).

Stereologic Volumetry of the Lung

Computer-assisted stereologic volumetry was applied to determine the relative contributions of the various lung compartments (specifically, peripheral parenchyma versus future air space) to the observed changes in lung size. During the time interval studied, the total V(lu) of TO-24 and TO-27 fetuses increased by factors of six and four, respectively, which is significantly more than the 2-fold increase seen in control fetuses (Tables 1 and 2). V(lu) of TO-24 fetuses was significantly larger than that of C-24 fetuses from 28 DGA on. V(lu) of TO-27 fetuses showed a more rapid increase and was already 40% larger than C-27 from the second day after occlusion on (Table 2). The A_A(pa/ lu), representing the peripheral parenchymal fraction not containing large-sized bronchi and vascular structures, was similar in all experimental groups and ranged from 94 to 98% at the various time points. The total V(pa), which is the product of V(lu) and A_A(pa/lu), therefore paralleled V(lu) and was also significantly increased after 4 d of TO in TO-24 fetuses, and after 2 d in TO-27 fetuses (Tables 1 and 2).

The A_A(ae/pa), representing the peripheral parenchymal tissue fraction, showed a progressive decrease with gestation in C fetuses between 24 and 28 DGA (Tables 1 and 2). Around 28 DGA a plateau was reached, whereafter A_A(ae/pa) remained more or less constant around 55%. In TO fetuses, in contrast, A_A(ae/pa) continued to decrease progressively after 28 DGA.

These results can also be expressed in terms of ASF, a more traditional parameter reflecting the degree of air-space distension. The ASF of TO-24 fetuses was comparable with that of C-24 fetuses until 3 d after TO (Figure 4A). At 28 DGA, the ASF of TO-24 fetuses was 25% larger than that of C-24 fetuses, and this difference was maintained during the remainder of the study period. In TO-27 fetuses, the ASF was significantly larger than in C-27 fetuses from as early as 2 d after TO (Figure 4B). From 29 DGA on, and persisting throughout the remainder of gestation, the ASF of TO-27 fetuses was 30% larger than that of C-27 fetuses.

View larger version (12K): [in this window] [in a new window]   Figure 4.   ASF of (A) C-24 and TL-24 lungs and (B) C-27 and TL-27 lungs. Values represent means ± SEM of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2. *P < 0.05, P < 0.01, P < 0.001 versus C.

The total V(ae), which takes into account both A_A(ae/ pa) and V(pa) and represents the volume of lung tissue, showed a steady slow increase in C fetuses between 25 and 31 DGA (Figure 5). In TO-24 fetuses, V(ae) was similar to that of C-24 fetuses until 28 DGA. Between 28 and 29 DGA, V(ae) of TO-24 fetuses showed a dramatic 2-fold increase, indicating true tissue growth (Figure 5A). V(ae) of TO-27 fetuses showed a similar but more gradual increase between 29 and 31 DGA (Figure 5B). Interestingly, in both TO-24 and TO-27 models, the increase of V(pa) and ASF preceded the increase of V(ae) by at least 1 d, indicating that passive distension of the air spaces occurred before tissue growth.

View larger version (12K): [in this window] [in a new window]   Figure 5.   V(ae) of (A) C-24 and TL-24 lungs and (B) C-27 and TL-27 lungs. Values represent means ± SEM of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2. *P < 0.05, P < 0.01 versus C.

Stereologic volumetry of the various lung compartments allowed assessment of the relative contributions of distension and tissue growth to the observed increase of V(pa). In TO-24 fetuses, approximately 40% of the increase in V(pa) observed between 25 and 30 DGA was attributable to an increase in V(ae) (tissue growth) (682/ 1,746 µl) and approximately 60% to passive distension of the air spaces, similar to the proportions seen in C-24 fetuses. In comparison, the contribution of passive distension to the final lung growth was slightly higher in TO-27 fetuses (66%).

BrdU Pulse Labeling

In addition to determination of V(ae), the time course of tissue growth and cell proliferation was monitored by BrdU pulse labeling. BrdU is a thymidine analogue that specifically labels proliferating cells in the S-phase of the cell cycle. In C-24 fetuses, the BrdU-LI (percentage of BrdU- labeled cells) of air-exchanging parenchyma remained relatively constant during 25 through 27 DGA, followed by a steady decrease during late gestation (Figure 6A). A similar trend was noted in C-27 fetuses (Figure 6B), although the absolute values were lower than in C-24 lungs. TO at 24 DGA was found to result in an initial suppression of cell proliferation until 27 DGA, followed by a marked increase of the proliferative rate after 28 DGA (4 d after occlusion) (Figure 6A). At 29 DGA, the proliferative rate of TO-24 lungs was five times higher than that of C-24 lungs (P < 0.001) (Figure 7). Occlusion at 27 DGA did not result in an initial decrease of cell proliferation (Figure 6B). From the first postocclusion day on, the proliferative rate of TO-27 lungs tended to be higher than that of C-27 lungs. As in TO-24 fetuses, the proliferative rate of TO-27 lungs was 5-fold higher than C-27 lungs at 29 DGA, although the absolute values were lower than in TO-24 (Figure 6B).

View larger version (13K): [in this window] [in a new window]   Figure 6.   BrdU-LI of (A) C-24 and TL-24 lungs and (B) C-27 and TL-27 lungs. Values represent means ± SEM of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2. *P < 0.05, P < 0.01 versus C.

View larger version (125K): [in this window] [in a new window]   View larger version (111K): [in this window] [in a new window]   Figure 7.   BrdU immunohistochemistry. (A) C-24 lung and (B) TO-24 lung, both at 29 DGA. (BrdU labeling by modified ABC method, hematoxylin counterstain, original magnification ×200.)

Several observations were made with respect to the spatial distribution of proliferative activity. Most of the proliferative activity in C lungs was seen in the interstitium, whereas TO lungs showed many BrdU-labeled nuclei along the luminal, epithelial aspect of the alveolar septa (Figure 7). Furthermore, the proliferative rate of peripheral air-exchanging lung parenchyma (close to pleura) was consistently higher than that of central parenchyma in all experimental groups (not shown).

Surfactant Immunohistochemistry

To monitor the effect of TO on the type II cell population, the cells were identified using anti-SP-A immunohistochemistry. SP-A immunoreactivity was first detected in rare, scattered type II and Clara cells at 25 DGA and could be seen consistently in C-24 and TO-24 fetuses from 26 DGA on. At about the same time, reaction product started to coat the alveolar walls. The lungs of TO-24 animals at postocclusion Days 4 and 5 appeared to have fewer SP-A-immunoreactive cells per area than did C-24 littermates (Figure 8). Intraluminal SP-A-positive secretions were seen from 26 to 27 DGA on, and were more prevalent in C-24 animals than in TO-24 animals (Figure 8). Surfactant-positive cells appeared overall slightly less prevalent in the 27 DGA group, and, in this group, no obvious morphologic differences in immunoreactivity were detected between C-27 and TO-27 lungs.

View larger version (115K): [in this window] [in a new window]   View larger version (101K): [in this window] [in a new window]   Figure 8.   SP-A immunohistochemistry. (A) C-24 lung and (B) TO-24 lung, both at 29 DGA. (SP-A immunostaining by ABC method, hematoxylin counterstain, original magnification ×200.)

These qualitative observations were confirmed by computer-assisted determination of the areal density of SP-A- positive cells (A_A[pnII/ae]). In C-24 fetuses, A_A(pnII/ae) progressively increased with increasing gestation (Figure 9). In the TO-24 fetuses, A_A(pnII/ae) followed a course parallel with C-24 fetuses until 28 DGA, whereafter an abrupt decrease was noted to only 50% of the control values. A_A(pnII/ae) of C-27 and TO-27 fetuses were similar at all time points studied (Figure 9B). In C-27 fetuses, the values of A_A(pnII/ae) were initially significantly lower than those in C-24 fetuses (only 25% of C-24 at 28 DGA). However, during the following 2 gestational d, A_A(pnII/ ae) of C-27 and TO-27 lungs dramatically increased to approach the projected values of C-24 fetuses.

View larger version (11K): [in this window] [in a new window]   Figure 9.   Surfactant-positive areal density. Areal density of SP-A- positive cells of (A) C-24 and TL-24 lungs and (B) C-27 and TL-27 lungs. Values represent means ± SEM of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2. P < 0.01 versus C.

BrdU-SP-A Double Labeling

To determine the effect of TO on the proliferative activity of type II cells, the BrdU-LI of SP-A-positive cells was determined using immunohistochemical double-labeling techniques. The BrdU-LI of SP-A-positive type II cells (Pn II BrdU-LI, percentage of proliferating type II cells) showed a rapid decrease between 27 and 30 DGA in C-24 and C-27 fetuses (Figure 10). TO at 24 DGA was followed by a significant increase of type II cell proliferation that peaked at 28 DGA, thus paradoxically coinciding with the time of decreased SP-A-positive cell density (Figures 10A and 11). At 28 and 29 DGA, the proliferative rate of type II cells in TO-24 lungs was more than four times higher than that of C-24 lungs (Figure 11). TO at 27 DGA yielded no significant alterations in type II cell proliferation at any time point (Figure 10B).

View larger version (11K): [in this window] [in a new window]   Figure 10.   BrdU-LI of SP-A-positive cells of (A) C-24 and TL-24 lungs and (B) C-27 and TL-27 lungs. Values represent means ± SEM of control (CTR) and tracheal-ligated (TL) animals. Number of animals studied: see Tables 1 and 2. *P < 0.05, P < 0.01 versus C.

View larger version (102K): [in this window] [in a new window]   View larger version (102K): [in this window] [in a new window]   Figure 11.   BrdU and SP-A double labeling. (A) C-24 lung at 28 DGA shows numerous type II cells (red cytoplasm) lining the future air spaces. BrdU-labeled nuclei (brown) are limited to the interstitium. (B) TL-24 lung at 28 DGA shows proliferative activity in a large proportion of type II cells (arrows). (BrdU and SP-A labeling by modified ABC method, hematoxylin counterstain, original magnification ×400.)

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Fetal TO is a potent stimulus of accelerated fetal lung growth (1). In this study, we investigated whether the patterns of fetal lung growth after intrauterine TO are influenced by the developmental stage of the lungs at the time of intervention. Fetal rabbits underwent either "early" TO (at 24 DGA) or "late" TO (at 27 DGA), corresponding to late pseudoglandular and early terminal sac stages, respectively. Lung growth was assessed by a combination of several complementary techniques, including determination of LW/BW, stereologic volumetry, and BrdU pulse labeling.

The lung-growth kinetics were strikingly different in the two models. Early TO was followed by a 3-d stagnation of growth, during which cell proliferation was significantly decreased compared with C. Only after 3 d (from 28 DGA on) was there an increase in cell proliferation which, in association with marked air-space distension, resulted in doubling of LW/BW by 30 DGA. In contrast, late TO was followed by an immediate and sustained increase in cell proliferation as compared with C. The cell proliferation noted after late TO was less pronounced than after early TO but, with slightly more pronounced air-space distension, led to doubling of LW/BW in only 3 d.

In both models, the first dynamic event noted was mechanical distension, which seemingly triggered cell proliferation and consequent tissue growth. This mechanical distension, monitored by determination of the ASF of peripheral air-exchanging parenchyma, occurred 4 d after early TO (28 DGA) but less than 2 d after late TO (29 DGA). The 3-d delay in development of mechanical distension after early TO is conceivably related to the undifferentiated state of precanalicular lung epithelium. Although the exact time course of lung liquid secretion in fetal rabbits has not been fully described, luminal lung liquid production in mammalian lungs has been shown to increase progressively after midgestation (29) and, in fetal rabbits, results in high lung liquid content at 28 DGA (30). We speculate that maturation of the liquid-secreting cells is a critical and essential first step in mediating TO-induced lung growth. According to this hypothetical scheme, secretion of lung liquid in occluded lungs by secretorily active cells will create mechanical distension which, in turn, will induce tissue proliferation through a variety of mechanotransductive processes.

It is unclear why mechanical distension after early TO resulted in much more prominent, albeit delayed, cell proliferation than after late TO. The amplitude of mechanical distension, as far as can be judged from determination of ASF, was roughly similar in the two models (25% increase of ASF after early T0; 30% increase after late TO). The differential growth response of the early and late TO models may be related to the difference in biochemical maturity of the mechanotransductive signaling pathways. Although poorly understood, the mitogenic effects of mechanical stretch have been shown to involve numerous regulators and mediators, including immediate early genes, multiple secondary messenger pathways, proto-oncogenes, hormonal influences, and polypeptide growth factors (16). Many of the autocrine/paracrine growth factors implicated in mechanotransduction, including epidermal growth factor and keratinocyte growth factor, are known to have a specific temporal pattern of expression during fetal lung development (32) that may contribute to the age-specific lung-growth response to mechanical stimuli. Alternatively, lung cells "primed" by the 3-d stagnation after early TO may have undergone intrinsic changes that altered their responsiveness to stretch.

Interestingly, TO was found to have stage-specific effects on overall fetal growth. Late TO was associated with temporary fetal-growth retardation, as evidenced by stagnation of fetal growth and a relative decrease of lung-cell proliferation around 29 DGA. A similar growth retardation was not seen after early TO. We speculate that the growth stagnation caused by late TO may be caused by surgical stress, and that the differential effects of late and early TO on fetal growth may be related to differences in hormonal maturity. It remains to be determined whether the stage-dependent effects of surgical intervention on fetal growth also apply to humans.

In addition to the kinetics of overall lung growth, this study was aimed at investigating the effects of TO timing on the population kinetics of type II pneumocytes. To this end, we employed a combination of SP-A immunohistochemistry, computer-assisted morphometry, and BrdU labeling. Again, early and late TO had very different effects on type II cell-growth kinetics. Late TO caused no significant alterations in type II cell proliferation or type II cell density. Early TO, in contrast, resulted in a marked increase of cell proliferation, paradoxically associated with a dramatic depletion of type II cells after 4 d of occlusion.

The increased type II cell proliferation observed after early TO is in agreement with in vitro studies, which have demonstrated increased [³H]thymidine incorporation in cultured fetal type II cells exposed to stretch (20). However, the mechanisms of type II cell depletion occurring several days after early TO, reminiscent of the type II cell loss described after prolonged TO in sheep (5), are likely more complex. Theoretically, reduction of the type II cell population may be due to one or several mechanisms, including increased cell death or accelerated differentiation of type II cells to type I cells. We have recently demonstrated that the type II cell loss induced by prolonged TO is, at least in part, attributable to increased type II cell apoptosis (35). A study of apoptotic activity by terminal deoxynucleotidyl transferase-mediated end labeling combined with SP-A immunohistochemistry revealed significantly higher rates of type II cell apoptosis in TO lungs from 28 DGA on.

We postulate that the marked type II cell depletion after early TO is also partly attributable to accelerated terminal differentiation of type II cells into type I cells. Although controversial, experimental evidence suggests that deformation (elongation) of type II cells in vitro induces loss of type II cell phenotype (lamellar bodies, SP expression) and transdifferentiation to type I-like cells (36, 37). Although mechanical strain applied to a purified cell population in vitro is not entirely comparable with mechanical distension of whole fetal lungs in vivo, the relatively abrupt mechanical distension of fetal lungs after early TO may similarly induce transdifferentiation of type II cells to type I or type I-like cells. Again, "priming" of the type II cells during the 3-d lag phase after early TO may have rendered these cells more susceptible to the effects of mechanical stretch. Indeed, apparently similar degrees of mechanical distension applied only 1 d later than in the early TO model (29 DGA versus 28 DGA) did not result in markedly increased type II cell proliferation and depletion after late TO. Finally, the dramatic decline in type II cell density may be a result of chronic elevation in intratracheal pressure. It is possible that TO at 28 DGA, only 3 d before term, did not allow enough time for type II depletion to occur.

In summary, this study demonstrates that the growth response of fetal rabbit lungs and surfactant system to intrauterine TO depends on the lung maturity at the time of TO, and thus establishes the importance of timing, rather than duration, of occlusion as critical determinant of post-TO lung growth. The observations in this study have important implications for the timing of clinical fetal interventions aimed at enhancing fetal lung growth. Although the stage-specific effects of TO on hypoplastic lungs remain to be elucidated, we speculate that the efficiency of TO in inducing accelerated lung growth will depend on the functional maturity of the lung epithelium. It is thus conceivable that hypoplastic lungs, characterized by biochemical immaturity of the alveolar epithelium, will not respond to TO to the same extent as do fully developed, secretorily active lungs. We therefore propose that the timing of clinical interventions should take into account not only the gestational age of the fetus, but also the estimated functional maturity of the lungs in the specific clinical context.

With this study we have also confirmed that TO in the fetal rabbit provides a valid in vivo model for studying the various mechanical and autocrine/paracrine mediators of TO-induced fetal lung growth, as well as the role of lung maturity in this process. Studies in the larger fetal sheep model, which would allow dissection of mechanical versus autocrine/paracrine effects (4), as well as in vitro studies will be instrumental in clarifying the mechanisms and age-dependency of TO-induced lung growth.