Introduction

Top Abstract Introduction Materials and methods Results Discussion References

In the rat lung, interstitial fibroblasts undergo a period of rapid proliferation from postnatal Days 4 to 13 (1). During this time, secondary septa emerge from the walls of existing air spaces as the lung undergoes the transformation from the saccular to the alveolar stage (2). Although the newly formed alveoli continue to increase, few new alveoli develop beyond postnatal Day 13 (3). Immediately after completion of this period of rapid alveolarization, the alveolar walls become thinner and the total number of interstitial lung fibroblasts decreases 20% from 17.5 × 10⁶ on Day 13 to 14.0 × 10⁶ on Day 21 (1). The present study was undertaken to determine whether postnatal rat lung fibroblasts undergo apoptosis beyond Day 13 and, if so, the time course over which this process occurs.

In the course of normal organ and tissue development, cells are often produced at rates that exceed the needs of the mature organ. Apoptosis (programmed cell death) is an integral part of development and serves to remove the unneeded cells in many organ systems, most notably the nervous and immune systems. Approximately half of all neurons produced will eventually die, presumably the result of limited access to the neurotrophic factors essential for neuronal survival (4). Apoptosis is also involved in the deletion of interdigital webs (5, 6), in palatal fusion (7), and in the development of the intestinal mucosa (8).

The morphologic changes seen in a cell undergoing apoptosis are distinct from the changes associated with necrosis (9). In the apoptotic cell, nuclear chromatin is compacted into sharply circumscribed masses of uniform density, whereas in the necrotic cell chromatin aggregates are poorly defined. Necrosis is characterized by the loss of membrane integrity, swelling and subsequent lysis of the cell, and leakage of lysosomal contents, resulting in an inflammatory response. In contrast, membrane blebbing is seen in apoptotic cells. These blebs then separate and the plasma membrane seals, resulting in the formation of membrane-bound apoptotic bodies.

Apoptosis is an active process that requires the coordinated regulation of specific genes and is often dependent on RNA and protein synthesis. Distinct signaling pathways that lead to apoptosis have recently been delineated and numerous proteins have been identified as either inducers or inhibitors of the apoptotic process. Members of the bcl-2 gene family are among those proteins found to regulate apoptosis. Some (bcl-2, bcl-x) suppress while others (BAX, BAD) promote apoptosis (10, 11). The ratio of bcl-2 to BAX gene products is thought to be a determining factor. BAX and bcl-2 interact to form homodimers and heterodimers, the latter being inactive (12). Repression of apoptosis has been found to occur in a hematopoietic cell line when more than half of BAX is heterodimerized with bcl-2 (13).

The present study was conducted to determine whether apoptosis plays a role in the observed decrease in the number of interstitial rat lung fibroblasts from postnatal Days 13 to 21 (1). To address this question, we have used morphologic techniques to detect structural changes in lung fibroblasts, and flow cytometry to quantitate chromatin condensation. DNA fragmentation was assessed in freshly isolated fibroblasts by flow cytometry and in situ by the terminal transferase deoxyuridine triphosphate (dUTP)- digoxygenin nick end-labeling (TUNEL) assay. Changes in levels of the messenger RNAs (mRNAs) for apoptosis-associated genes, bcl-2, BAX, and c-myc were measured by reverse transcriptase-polymerase chain reaction (RT- PCR). Our results indicate that apoptosis plays a key role in the decrease in interstitial fibroblast number that occurs after alveolar septal formation.

	Materials and Methods

Top Abstract Introduction Materials and methods Results Discussion References

Isolation of Rat Lung Fibroblasts

The offspring of timed-pregnant Sprague-Dawley dams (Harlan Sprague Dawley, Inc., Indianapolis, IN) were used in these experiments. Each litter consisted of 10 to 12 pups. The day of birth was designated as Day 0. Pups were killed on postnatal Days 5, 12, 14 to 19, and 23 with a lethal intraperitoneal injection of sodium pentobarbital. The lungs were trimmed to remove major airways, and rinsed with calcium- and magnesium-free Hanks' balanced salt solution (HBSS). Pooled lung tissue from 3 to 5 pups at each postnatal age were then minced into 1 to 2-mm³ pieces and incubated in HBSS containing 0.3 mg/ml type IV collagenase and 0.5 mg/ml trypsin in a shaking water bath maintained at 37°C. At each of six 10-min intervals, the minced tissue was passed through a 25-ml pipette a total of 10 times to dissociate the cells. The cells in suspension were then transferred to an equal volume of cold (4°C) complete media containing 1:1 (vol/vol) Dulbecco's modified Eagle's medium/Ham's F-12, 10% fetal bovine serum, penicillin (10,000 U/100 ml), streptomycin (10,000 µg/100 ml), and glutamine (29.2 mg/100 ml). Tissue-culture reagents and enzymes were purchased from GIBCO BRL Life Technologies (Grand Island, NY). Fresh type IV collagenase/ trypsin was added to the remaining lung homogenate, which was then returned to the shaking water bath. At the end of the 1-h digestion period, the dissociated cells were pelleted by centrifugation (1,200 rpm at 4°C for 10 min), plated in complete media, and incubated at 37°C in 5% CO₂. After 1 h, nonadherent cells were aspirated and the flasks were rinsed thoroughly with HBSS. The remaining adherent fibroblasts were either frozen at 70°C for subsequent RNA extraction or removed from the flask by trypsinization for subsequent flow cytometric analysis. We have found this differential adherence protocol to result in > 99% fibroblasts at 24 h as evidenced by their morphologic appearance when viewed at the light microscopic level and by immunohistochemical staining for vimentin.

Morphologic Assays for Apoptosis

In situ detection of DNA fragmentation in lung tissue. We evaluated tissue sections from frozen, paraffin-embedded, and JB-4-embedded (Polysciences, Warrington, PA) lungs obtained from rat pups ranging in age from 10 to 29 d. Frozen lung sections (8 to 10 µm) were fixed in 10% neutral buffered formalin for 10 min and postfixed in ethanol/acetic acid (2:1) for 5 min. Other lungs were inflation-fixed in situ with either 10% neutral buffered formalin or 4% paraformaldehyde, and dehydrated. The formalin-fixed lungs were embedded in JB-4 (glycol methacrylate) (Polysciences), and sectioned at 2 µm; the paraformaldehyde-fixed lungs were embedded in paraffin and sectioned at 10 µm. The paraffin-embedded and glycol methacrylate-embedded lung sections were rehydrated and treated with proteinase K (20 µg/ml) for 15 min at room temperature. For frozen, paraffin-embedded, and glycol methacrylate-embedded lung tissue sections, the Oncor ApopTag kit (Oncor, Gaithersburg, MD) was used to label the 3'-OH ends of fragmented DNA with dUTP-digoxigenin in the presence of terminal deoxynucleotidyl transferase (TdT) in a humidified chamber at 37°C for 1.5 h. Each slide contained four serial sections. Two were treated with TdT in reaction buffer and the other two sections, which served as the negative control, were incubated with reaction buffer alone. The slides were incubated in peroxidase-conjugated antidigoxigenin for 30 min at room temperature. Diaminobenzidine (DAB) (Vector Laboratories, Burlingame, CA) was used as the chromagen. Tissue sections were counterstained with methyl green, dehydrated in xylene, and mounted with Permount. The slides were viewed with a Nikon Diaphot microscope, and Kodacolor film (ASA 100) (Eastman Kodak, Rochester, NY) was used for the photographs.

Detection of apoptotic bodies in isolated lung cells. Rat pups were killed and lungs were removed and digested with trypsin/collagenase as described previously. Small aliquots of cell suspensions from the whole-lung homogenate were obtained after the first 10 to 30 min of enzymatic digestion. The cell suspension was applied to gelatin-coated slides, air-dried for 5 min, fixed in methanol at 20°C for 5 min, and stored at 20°C. Before immunostaining, the slides were warmed to room temperature and postfixed for 15 min in 4% paraformaldehyde at 4°C. The slides were then rinsed in two changes of cold phosphate-buffered saline (PBS), permeabilized for 5 min in 0.2% Triton X-100 (Sigma, St. Louis, MO) in PBS, rinsed in two changes of PBS, and blocked in 5% normal goat serum (Sigma) for 30 min at room temperature. After rinsing with PBS, the slides were incubated for 3 h at room temperature with a 1:40 dilution of monoclonal mouse anti-pig vimentin (immunoglobulin G1 isotype from Sigma) in PBS containing 1% bovine serum albumin and 0.1% Triton X-100 (PBS+). This antibody cross-reacts with both rat and human vimentin and does not react with other intermediate filament proteins (14). The slides were then rinsed, and incubated at room temperature for 30 min with Cy2-conjugated goat antimouse serum (Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) diluted 1:200 (6.5 µg/ml) with PBS+. After two rinses in PBS+, the cells were then blocked with 5% normal rabbit serum (Sigma) for 30 min, rinsed in PBS+, and incubated with a 1:50 dilution of rabbit antihuman factor VIII (Dako, Carpinteria, CA) for 2 h. The slides were then rinsed in two changes of PBS+, incubated with Cy3-conjugated goat antirabbit serum (Jackson ImmunoResearch) (1:800) for 30 min, and rinsed in two changes of PBS+. Finally, the slides were stained with Hoechst 33342 (Molecular Probes, Eugene, OR) at a concentration of 1 µg/ ml for 10 min, rinsed with PBS, coverslipped with glycerol/ PBS (1:2), and viewed using a Nikon Diaphot 300 microscope equipped with filter sets for fluorescence (Chroma fluorescein isothiocyanate [FITC] for Cy2 and Nikon 4'-6-diamidino-2-phenylindole [DAPI] filter cube for Hoechst).

Flow Cytometric Analysis of Chromatin Condensation and DNA Fragmentation

Hoechst stain. Fibroblasts were isolated from the lungs of 16-d (n = 3) and 17-d (n = 3) rats, plated for 1 h, released by trypsinization, washed with 1 ml PBS, transferred to polypropylene tubes (Falcon #2063; Becton Dickinson, Lincoln Park, NJ), and centrifuged at 1,200 rpm for 5 min. The cell pellet was resuspended in 1 ml PBS containing 5 µg/ml Hoechst 33342, and incubated for 30 min at 37°C in the dark. The stained cells were then stored on ice and analyzed within 30 min on a FACStar Plus (Becton Dickinson, San Jose, CA) equipped with a 5-W argon laser tuned to ultraviolet (UV) (100 mW). Red fluorescence was collected as FL2 area through a 660/ 20 band-pass filter, whereas blue fluorescence was collected as FL1 area through two LP 400 long-pass filters. Blue and red emissions were split using an LP 460 long-pass filter. The UV beam was also used to determine the forward- and side-scatter parameters. Side scatter was collected through an SP 375 short-pass filter. The data were analyzed using Cell Quest software (Becton Dickinson, San Jose, CA). All cells were included in the analysis. Viable (R1) and apoptotic (R2) regions were drawn based on the viable region of the 16-d sample. These same regions were used again for the 17-d sample.

BODIPY-dUTP (15). Fibroblasts freshly isolated from the lungs of 16 to 19-d pups were fixed in 1.0% formaldehyde in HBSS at 4°C for 15 min, centrifuged, resuspended in 70% ethanol, and stored at 20°C for up to 4 d. The cells were then rinsed twice with saline and resuspended in a 50-µl reaction mix containing 1.0 µl TdT (Boehringer Mannheim, Indianapolis, IN), 10 µl 5× TdT buffer, 0.25 µl BODIPY-FL-14-dUTP (Molecular Probes), and 38.75 µl water. The cells were incubated in the reaction mix for 5 h at 37°C, rinsed twice with 15 mM ethylenediaminetetraacetic acid (pH 8.0), rinsed once with 0.1% Triton X-100 in PBS, and resuspended in 0.5 ml of a solution containing 2.5 µg/ml propidium iodide (Sigma) and 0.1% ribonuclease (RNase; Sigma). After being incubated for 20 min at 37°C, the cells were immediately cooled to 4°C and evaluated by flow cytometry using FAC-SCAN (Becton Dickinson). Cell aggregates and doublets were excluded. DNA content was collected at FL2 area and BODIPY fluorescence was collected through FL1. Data analysis was performed using CellQuest software (Becton Dickinson).

Extraction and Reverse Transcription of Total RNA Obtained from Freshly Isolated Fibroblasts

Freshly isolated fibroblasts from the lungs of 5 to 23-d-old rat pups were homogenized with a Polytron homogenizer (Brinkman Instruments, Westbury, NY). Total RNA was extracted with TRI-REAGENT-LS according to the manufacturer's recommended protocol (Molecular Research Center Inc., Cincinnati, OH). Glycogen (100 µg) was used as a carrier. RT-PCR was performed with a Gene-Amp RNA PCR kit (Perkin-Elmer, Norwalk, CT). The protocol recommended by the manufacturer was modified by adding 0.2 U of RNase-free deoxyribonuclease (DNase) I (Worthington, Freehold, NJ) to remove contaminating DNA before the addition of the RT (16). The final reaction volume (20 µl) contained the following: 0.4 µg input RNA; 0.2 U of DNase I; 20 U of RNAse inhibitor; 5 mM MgCl₂; 1 mM CaCl₂; 1 mM each of dGTP, dATP, dTTP, and dCTP; 5 µM random hexamers; 1× PCR buffer II (10 mM Tris-HCl, pH 8.3, 50 mM KCl); and 1 µl of 50 U/µl Moloney murine leukemia virus (added after the DNase treatment). The initial reaction (19 µl) mix was first incubated in the cycler (2400 Perkin-Elmer Cetus) for 30 min at 37°C to remove genomic DNA, then heated at 75°C for 5 min to inactivate the DNase I and cooled to 4°C. Next, 50 U (1 µl) of Moloney murine leukemia virus RT and 20 U RNase inhibitor were added to the reaction mix, which was then incubated at room temperature for 10 min, followed by a 30-min incubation at 42°C. The reverse transcription reaction was terminated by heating at 90°C for 5 min and cooling at 4°C for 5 min. Parallel no RT reactions in which 1 µl of diethylpyrocarbonate-treated water was substituted for 1 µl of RT were run for each sample and primer pair combination.

PCR

The PCR amplification stock solution was prepared on ice and aliquotted into separate tubes. Each 49 µl of stock contained 2 mM MgCl₂; 10 mM Tris-HCl (pH 8.3); 50 mM KCl; 0.04 mM dCTP; 0.08 mM each of dGTP, dATP, and dTTP; and 10 µCi alpha [³²P]-dCTP (3,000 Ci/mmol), 1.25 U AmpliTaq polymerase, and 0.5 µM of each primer. After the addition of 1.0 µl of sample complementary DNA (cDNA), the tubes were placed in a cycler heated to 95°C (2400 Perkin-Elmer Cetus) and maintained at 95°C for 1 min 45 s to inactivate RT. The heated cover on the 2400 Perkin-Elmer cycler prevented refluxing and condensation during thermal cycling, thus avoiding the need to overlay tubes with mineral oil. The samples were then denatured at 94°C for 30 s, annealed at 60°C for 30 s, and extended at 72°C for 1.5 min. The radiolabeled PCR products were resolved by electrophoresis on 12.5% polyacrylamide gels. The dried gels were quantitated with a PhosphorImager (Molecular Dynamics, Sunnyville, CA). Autoradiographic films of the gels were also quantitated with a BioImager (Millipore, Ann Arbor, MI). The identities of the amplified cDNAs were confirmed on an automated sequencer (Perkin-Elmer, Foster City, CA) located in the Macromolecular Structure Analysis Facility at the University of Kentucky (Lexington, KY).

PCR primers. The bcl-2 5' forward primer sequence was CTGTACGGCCCCAGCATGCG, the 3' reverse-compliment primer sequence was GCTTTGTTTCATGGTACATC, and the amplified product was 231 bases (17). The BAX 5' forward primer sequence was TGGAGCTGCAGAGGATGATT, the 3' reverse-compliment primer sequence was AAGTTGCCATCAGCAAACAT, and the amplified product was 100 bases (17). The c-myc 5' forward primer was TACTCACCAGCACAATTATG, the 3' reverse-compliment primer sequence was TTGAACGGACAGGATGTAGG, and the amplified product was 315 bases (18). The cyclophilin 5' forward primer sequence was ATGGTCAACCCCACCGTGTT, the 3' reverse-compliment primer sequence was GCGTGTGAAGTCACCACCCT, and the amplified product was 205 bases (19).

Statistical Analysis

The significance of the differences between apoptotic and control samples was determined using the Kolmogorov- Smirnov test, a nonparametric equivalent of the paired t test. Differences were considered statistically significant when P < 0.05.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

Analysis of DNA Fragmentation by Flow Cytometry

Both the 1974 study by Kauffman and colleagues (1) and recent work by these investigators (20, and P. H. Burri, personal communication) strongly suggested that apoptosis plays a role in lung remodeling after septation. We first used a flow cytometric approach to identify the time of onset of the apoptotic process. Margination and condensation of chromatin, one of the earliest microscopically visible features of apoptosis, is readily detectable by flow cytometry. Chromatin condensation has been shown by other investigators to be correlated with a red shift of the emission spectrum of Hoechst 33342, a DNA-binding vital dye (21- 23). Chromatin condensation can be quantitated independently from DNA content by analyzing the ratio of red to blue fluorescence; a high ratio of red to blue fluorescence is indicative of apoptosis (23). In Figure 1, an upward shift of the red/blue ratio is seen in the fibroblasts from 17-d-old pups when compared with the dot matrix plot of fibroblasts from 16-d-old pups, indicating an increase in the number of apoptotic cells at this postnatal age.

View larger version (27K): [in this window] [in a new window]   Figure 1.   Freshly isolated 16- and 17-d rat lung fibroblasts were stained with Hoechst 33342 (5 µg/ml) for 30 min at 37°C and analyzed for red fluorescence (y-axis) and blue fluorescence (x-axis) by flow cytometry. The viable (R1) and apoptotic regions were drawn on the basis of the viable region of the 16-d sample. The same regions were then used for the 17-d sample. Cells with an increased red fluorescence intensity were considered apoptotic. An upward shift of the red fluorescence is seen in fibroblasts from 17-d pups (n = 3) when compared with the fibroblasts from 16-d pups (n = 3).

A flow cytometric technique was also used to evaluate the percentage of lung fibroblasts that contained DNA strand breaks. We used TdT to attach deoxynucleotides to free 3'-OH ends of the DNA fragments generated by endonuclease digestion of genomic DNA during apoptosis. To quantitate DNA strand breaks by flow cytometry, we labeled the 3'-OH termini with dUTP-conjugated BODIPY in a single-step procedure (15). By using a nucleotide already conjugated with the label (BODIPY), we were able to decrease the number of centrifugation steps, thus minimizing the potential for cell loss. BODIPY has several advantages over fluorescein. This fluorochrome is smaller and, unlike fluorescein, BODIPY does not have a negative charge. When dUTP-conjugated BODIPY is used in conjunction with propidium iodide, a correlation between DNA strand breaks and cell-cycle phase distribution is obtained.

Quantitation of DNA strand breaks labeled with dUTP-conjugated BODIPY in individual fibroblasts demonstrated the presence of two populations of fibroblasts. As shown in the contour plot in Figure 2, unlabeled control fibroblasts are seen in region R1, which was defined by measuring BODIPY fluorescence in a population of first-passage 5-d fibroblasts. The fibroblasts seen in region R2 were labeled with dUTP-conjugated BODIPY, indicative of increased numbers of DNA strand breaks associated with apoptosis, compared with the DNA in fibroblasts in R1 which is stained less intensely. In this experiment, the percentage of fibroblasts in R2 was 47.3 on Day 16, 60.5 on Day 17, and 32.9 on Day 18. This assay was conducted a total of three times. The mean values for BODIPY-positive cells were as follows: Day 16 = 40.6% (n = 2), Day 17 = 51.4 ± 13.4% (n = 3), Day 18 = 36.9 ± 8.6% (n = 3), and Day 19 = 13.8 ± 5.4% (n = 3).

View larger version (37K): [in this window] [in a new window]   Figure 2.   Two-dimensional frequency contour plots of green BODIPY fluorescence (y-axis) indicative of DNA strand breaks versus red propidium iodide fluorescence (x-axis), a measure of DNA content in freshly isolated fibroblasts from 16-, 17-, and 18-d rat lungs. Fibroblasts at each postnatal age were obtained from the pooled lung samples from three or four rat pups; all pups were from the same litter. Region 1 was defined by a population of fibroblasts from 5-d rat lungs (n = 2). The cells in Region 2 were considered positive for BODIPY. The percentages of cells seen in the apoptotic region R2 were as follows: Day 16, R2 = 47%; Day 17, R2 = 60%; and Day 18, R2 = 33%. Although the majority of cells that underwent apoptosis were in the G0/G1 phase of the cell cycle, many cells also underwent apoptosis while in the G2/M phase. Similar results were obtained in each of three separate experiments.

Morphologic Evidence of Apoptosis during Postnatal Lung Development

Lung sections from rats ranging in age from 10 to 29 d were examined for evidence of apoptotic cells. DNA fragmentation was visualized in situ in cells labeled using the TUNEL assay. An evaluation of frozen lung tissue sections indicated that immunoperoxidase-positive cells were most abundant from postnatal Days 17 to 19. Because of the relatively poor resolution seen with frozen sections, we then evaluated both paraffin-embedded and glycol methacrylate-embedded tissue. Although the DAB stain was somewhat less intense in the tissue embedded in JB-4, the preservation of lung architecture was significantly better than that seen in the paraffin sections. An additional advantage of the JB-4-embedded tissue was that thinner sections could be cut, thus facilitating the identification of DAB-positive cells. The complete absence of DAB labeling in the negative control sections confirmed the specificity of the reaction. Furthermore, alveolar macrophages observed in the airspaces were consistently DAB-negative, thus confirming that endogenous peroxidases were quenched by the hydrogen peroxide treatment.

We examined a total of 50 high-power fields (×1,000) for each slide and counted the number of DAB-positive cells per field, excluding those fields that contained major airways or vessels. The data are expressed as the mean number of DAB-positive cells per field. In the lungs of the 10-d pups, the youngest age examined, the distribution of DAB-positive cells was not uniform (Figure 3a). Apoptotic cells were seen more frequently in the alveoli in peripheral than in proximal lung regions. The number of DAB-positive cells per field ranged from 1 to 14, and the mean value was 4.6 per field. Apoptotic cells were observed less frequently in the lungs of 11-d pups. A mean value of 1.6 DAB-positive cells was seen per field. The incidence of apoptotic cells was even lower in the 16-d lungs, where the mean value was 0.8 DAB-positive cells per field. In contrast, lungs sections from the 17-d pups contained an average of 20 DAB-positive cells per field (Figures 3b and 3c). The number of apoptotic cells decreased substantially by Day 29, at which time we found 1.7 DAB-positive cells per field. With the exception of the Day-29 lungs, labeled cells were seen with greater frequency in peripheral alveoli. DAB-positive cells were often observed in the interstitium. Although many of these cells appeared to be fibroblasts, it is possible that other cells types, such as epithelial and endothelial cells, were labeled as well.

View larger version (132K): [in this window] [in a new window]   View larger version (128K): [in this window] [in a new window]   View larger version (125K): [in this window] [in a new window]   Figure 3.   Thin sections of glycol methacrylate-embedded lung tissue from 10- and 17-d rats were labeled with a modification of the TUNEL assay and counterstained with methyl green. Apoptotic (DAB-positive) nuclei are golden brown; normal nuclei are green. Red blood cells are stained light beige. (a) A representative field from a 10-d rat lung containing several apoptotic nuclei (arrows). Bar = 25 µm. (b) In contrast, numerous DAB-positive nuclei (arrows) are seen in the 17-d lung. Bar = 25 µm. (c) Lower-magnification view of a 17-d lung. Representative apoptotic nuclei are identified (small arrowheads). Bar = 25 µm.

Apoptotic bodies are detectable for a relatively short period of time, less than 1 h (24). We were concerned that the enzymatic digestion and isolation of fibroblasts by differential adherence, in addition to the time required for this process (2.5 h), could significantly diminish the number of nuclei containing apoptotic bodies that were observed. Therefore, we examined suspensions of pooled lung homogenate from two or three pups at 16, 17, 18, and 20 d. Lungs were digested with trypsin/collagenase for only 10 to 30 min. Slides containing a small droplet of a cell suspension obtained from this brief enzymatic digestion of whole-lung homogenate were labeled first with antivimentin, an intermediate filament protein, to identify fibroblasts (25); second, with anti-factor VIII to label endothelial cells; and then stained with Hoechst 33342 to identify cellular fragmentation into apoptotic bodies. The slides were then examined under fluorescence microscopy, and apoptotic bodies that stained positively for vimentin but negative for factor VIII were noted. A minimum of 100 cells were counted at each postnatal age. The percentage of vimentin-positive, factor VIII-negative cells containing apoptotic bodies were as follows: Day 16 = 8%, Day 17 = 40%, Day 18 = 18.9%, and Day 20 = 10%.

Assessment of bcl-2 and BAX Gene Expression by RT-PCR

The expression of selected genes associated with the apoptotic process was evaluated by using RT-PCR assay (26). RNA was extracted from fibroblasts obtained from the lungs of rat pups that were 5 to 23 d old. To remove genomic DNA, total RNA (0.4 µg per sample) was first treated with DNase I. The DNase was subsequently heat-denatured. The RNA was then reverse-transcribed and amplified using primer pairs specific for two members of the bcl-2 gene family (BAX and bcl-2), c-myc, and cyclophilin, a constitutively expressed gene (27). Parallel reactions that did not contain RT were run for each of the samples to verify the absence of genomic DNA. The PCR products were separated by polyacrylamide gel electrophoresis (PAGE), and the gels were stained with ethidium bromide to identify the molecular weight standards. The gels were then dried and exposed to radiographic film. Initial experiments were conducted to optimize the RT-PCR reaction by determining the relationship of signal strength to the amount of input RNA and to the PCR cycle number. Each of the genes of interest was then amplified at a cycle number determined to be within the range of linearity.

BAX and bcl-2 mRNA values were normalized to cyclophilin and expressed as a percentage of Day-5 values (Figure 4). There was an increase in bcl-2 mRNA on Days 12 to 15, then a sharp decrease on Days 16 and 17. By Day 18, bcl-2 mRNA had returned to 5-d levels. BAX mRNA increased sharply from Day 15 to Day 16, by which time BAX mRNA levels were 3.5-fold higher than Day 5. BAX mRNA levels decreased gradually thereafter, but on Day 23 were still 1.8-fold greater than 5-d values. A comparison of the post-septation changes in bcl-2 and BAX mRNA levels in three separate experiments, each of which included pups from 2 to 3 litters, indicated a statistically significant decrease in bcl-2 mRNA (P < 0.001) and an increase in BAX mRNA (P < 0.001) from Day 5-6 to Day 17. The mean values for the percentage decrease in bcl-2 mRNA and the percentage increase in BAX mRNA were 73.5 ± 11.4% and 243.0 ± 102%, respectively. The signal for c-myc, a gene associated with cell proliferation (28, 29), decreased 77% from Day 14 to Day 16 and remained unchanged from Day 16 through Day 19 (data not shown).

View larger version (16K): [in this window] [in a new window]   Figure 4.   Bcl-2 message expression is downregulated and BAX message expression is upregulated in fibroblasts isolated from 16-d rat lung when compared with mRNA values from 5-d fibroblasts. Total RNA (0.4 µg) was amplified by RT-PCR using primer pairs specific for bcl-2 (24 cycles), BAX (22 cycles), c-myc (20 cycles), and cyclophilin (16 cycles). The PCR products were separated by PAGE and the radioactivity of the dried gels was quantitated by PhosphorImager analysis. These experiments were repeated three times with similar results.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

The early report by Kauffman and colleagues (1) of a 20% decrease in the number of interstitial fibroblasts in the rat lung from postnatal Days 13 to 21 strongly suggested a role for apoptosis. Focusing on this postnatal age range, we have shown that apoptosis does in fact occur in lung fibroblasts after alveolarization, and that the incidence peaks from postnatal Days 17 to 19. These observations establish a key role for the apoptotic process in the thinning of the alveolar wall that occurs after alveolarization.

Flow cytometry facilitated the rapid assessment of both chromatin condensation and DNA fragmentation in 10⁴ cells per assay. Chromatin condensation, measured as an increase in the intensity of red fluorescence in cells stained with Hoechst 33342, was initially shown by Belloc and coworkers to differentiate between apoptotic and normal polymorphonuclear cells (23). Their results were validated in sorted cells subsequently analyzed for evidence of DNA laddering by gel electrophoresis. In the present study, the percentage of cells in which chromatin condensation was detected increased from 6.7% in the 16-d fibroblasts to 30.6% in 17-d fibroblasts.

Flow cytometry was also used to detect DNA degradation in fibroblasts treated with BODIPY-conjugated dUTP to 3'-OH end-label DNA fragments. The quantitation of BODIPY-positive fibroblasts from 16- to 19-d rats provided further evidence that these cells undergo apoptosis after the completion of alveolar septation. The gradual decline in the percentage of BODIPY-labeled fibroblasts from 51.4% on Day 17 to 13.8% on Day 19 was consistent with a decrease in fibroblast apoptosis over this time period. The actual percentages of apoptotic cells seen in this assay may be somewhat higher than would be encountered in vivo because the processes that normally clear apoptotic cells are unlikely to be operative in vitro.

Although others have shown an increased incidence of apoptosis in fetal lung explants maintained in culture for 1 h (30), we found that apoptosis was not induced as a result of our fibroblast isolation technique. DNA fragmentation was also observed in situ by labeling the 3'-OH ends of DNA with digoxigenin-nucleotide, using a modification of the original TUNEL assay (31). These results confirmed those obtained by flow cytometry. DAB-positive cells were abundant in lungs from 17-d-old rats, but were seen infrequently in lungs from younger (10- to 16-d) or older (29-d) rats.

RT-PCR analysis of the expression of members of the bcl-2 gene family also documented changes in gene expression that were indicative of apoptosis. This assay was conducted on cells subjected to a 1-h enzymatic digestion and might have resulted in the loss of late apoptotic cells, thus underestimating the total apoptotic population. On Day 16, lung fibroblast bcl-2 mRNA levels decreased and BAX mRNA levels increased. The gene products of bcl-2 and BAX interact to form homodimers and heterodimers. Although bcl-2 and BAX heterodimers are inactive, when BAX is in excess and BAX homodimers predominate, cells are likely to undergo apoptosis (12). The observed decrease in bcl-2 mRNA at the same time that BAX mRNA is upregulated suggests an increase in the concentration of BAX homodimers, permitting apoptosis to occur (13).

Although c-myc has been shown to play a role in the induction of apoptosis in a number of cell systems, upregulation of this proto-oncogene does not appear to be required for apoptosis to occur in the postnatal rat lung fibroblast (32). c-myc mRNA, present in significant concentrations throughout the cell cycle in proliferating fibroblasts (33), was found in the present study to decrease from Days 14 to 16 as the period of rapid alveolar formation and fibroblast proliferation ceased. c-myc can induce, but is not required for, apoptosis. This effect is both cell type- and stimulus-specific (28). The results of this study suggest that c-myc does not play a role in the induction of apoptosis in the postnatal rat lung fibroblast.

Using similar techniques, other investigators have recently shown that apoptosis occurs in the late-fetal rat lung at a rate of 0.6 to 1.0% (29, 34) and in the early postnatal rat lung at a somewhat higher rate, 9 to 12% (34). We too observed apoptotic cells in situ on postnatal Day 10, lending support to the theory that apoptosis is an ongoing process in the immature lung. However, we have found that the number of lung fibroblasts that undergo apoptosis after alveolarization was 4- to 5-fold higher than the number of apoptotic cells observed either at birth or during alveolarization, a stage that is characterized by a high index of proliferation.

The reduction, due to apoptosis, in the number of fibroblasts in the interstitium of the developing lung is likely to play a critical role in lung maturation, the final process of which involves the transition of the alveolar wall from a double to a single capillary network layer. During alveolarization, the inner walls of the gas exchange area of the lung are composed of two superimposed sheets of capillaries. The alveoli are formed as low ridges (consisting of one of these capillary sheets) and emerge from the primitive saccular walls, subdividing the saccules into smaller alveoli. These ridges form the new alveolar walls, or secondary septa, which consist of an inner layer of connective tissue surrounded on both sides by sheets of capillary networks. Once the new alveoli have been formed, the two capillary layers in the secondary septa undergo extensive remodeling to form the single-layered capillary network seen in the mature lung. To accomplish this, the connective tissue layer thins markedly, resulting in a decrease in volume that brings the two capillary layers closer together. Interference with the apoptotic process would be expected to affect the lung maturation process adversely.

The nature of the stimulus that induces apoptosis in lung fibroblasts after rapid alveolarization remains to be determined. It is also unclear why some fibroblasts, but not others, are targeted for programmed cell death. The immature lung contains two subsets of fibroblasts that differ with respect to their intracellular lipid content. The percentage of fibroblasts containing lipid droplets has been shown to decrease beyond Day 10 (35, and our unpublished observations), suggesting that apoptosis may be selective for this population. Ongoing studies in our laboratory are currently addressing this question.