Growth and development of the fetal lungs is critically dependent on the degree to which the lungs are expanded by liquid; increases in fetal lung expansion accelerate lung growth, whereas reductions in lung expansion cause lung growth to cease. The mechanisms mediating expansion-induced lung growth are unknown but likely include alterations in the expression of genes that regulate lung cell proliferation. Our aim was to isolate and identify genes that are up- or downregulated by increased fetal lung expansion. In chronically catheterized fetal sheep at 126 days gestational age (GA), the left lung was expanded for 36 h, while the right lung remained at a control level of expansion. Subtraction hybridization was used to isolate genes differentially expressed between the left and right lungs. Screening of ∼6,000 clones identified 1,138 and 118 cDNA fragments that were up- and downregulated by increased lung expansion, respectively. Northern blot analyses in separate groups of control fetuses and fetuses exposed to increased lung expansion were used to verify differential expression. Increased fetal lung expansion upregulated heat shock protein 47, thrombospondin-1, TROP2, tropoelastin, and tubulin-α3 in fetal lung tissue by ∼200–300%; connective tissue growth factor and cysteine-rich angiogenic inducer 61 were increased by 20–30%. Genes downregulated by increased fetal lung expansion included CCSP-related protein-1, elongation factor-1α and vitamin D3 upregulated protein 1. We conclude that an increase in fetal lung expansion differentially regulates the expression of numerous genes in lung tissue, many of which have important putative roles in lung development, while the functions of others are currently unknown.
- lung development
- subtraction hybridization
- gene array
respiratory insufficiency is the major cause of neonatal morbidity and mortality and is primarily a consequence of preterm birth or pulmonary hypoplasia. Infants born very preterm are born before their lungs have had sufficient time to develop and usually require respiratory support after birth. However, this commonly causes lung injury and is closely associated with the development of chronic lung disease (CLD) (4). Despite major improvements in neonatal intensive care, the incidence of CLD is increasing (4), indicating that there is a pressing need to improve the therapeutic options available to these infants. However, the development of new therapeutic strategies can only be achieved by first identifying the factors that regulate lung growth and development, particularly those characteristics that determine its gas exchange potential.
The degree to which the fetal lungs are expanded by liquid is a critical determinant of fetal lung growth and development (18). Before birth, the lungs are filled with a liquid that is secreted across the pulmonary epithelium into the lung lumen and leaves the lungs via the trachea. The fetal upper airway restricts the efflux of lung liquid, which promotes lung liquid accumulation within the future airways, resulting in the formation of an internal distending pressure of 1–2 mmHg (46). As a result, the fetal lungs are maintained in a constantly distended state, which provides a stretch stimulus that is critical for their growth and development (18). Further increases in the basal degree of lung expansion accelerate lung growth, structural development, and epithelial cell differentiation, whereas lung growth and structural development ceases if the fetal lungs are deflated (2, 34, 38, 40). Although the relationship between the degree of fetal lung expansion and lung development is well established, the mechanisms involved are unknown. We hypothesized that an increase in fetal lung expansion causes a “stretch” stimulus that alters the expression of genes that accelerate lung growth, extracellular matrix (ECM) synthesis and deposition (25, 39), and alveolar epithelial cell trans-differentiation (10, 11).
Increases in fetal lung expansion, caused by obstruction of the fetal trachea, are a potent stimulus for fetal lung growth (2, 20, 34, 38, 39). The lung growth response follows a specific time course and closely parallels the increase in lung expansion following tracheal obstruction (TO) (20). In fetal sheep during the early alveolar stage of lung development, TO stimulates an increase in the rate of DNA accumulation over the first 7 days of obstruction, after which time DNA accumulates at control rates. However, lung DNA synthesis rates are greatest, at ∼800% above control levels, at 2 days of TO (39), indicating that the mechanisms by which stretch activates cell proliferation would be most active just before 48 h of TO. Thus, in the current study, our aim was to identify genes that mediate stretch-induced changes in fetal lung development at 36 h of increased lung expansion, using a subtraction hybridization gene analysis technique. To reduce the number of false-positive genes identified, we have modified a unique animal model (34) in which the basal degree of expansion in the left and right lungs of the same fetus can be controlled separately (13). This allows genes expressed in expanded and control lung tissue to be compared from the same animal and eliminates differences between animals (13); differential expression was subsequently confirmed in separate groups of fetuses. This analysis has allowed us to successfully identify over 80 genes that are differentially expressed in response to an increase in fetal lung expansion. Identifying these genes is a critical first step toward understanding the factors that regulate the alveolar stage of lung development.
All procedures involving the use of animals were approved by the Monash University Animal Ethics Committee. Surgery was performed on pregnant Merino × Border Leicester sheep at 119 days of gestational age (d GA; term ∼147 days, n = 3), as described previously (13). Anesthesia was induced by an intravenous injection of thiopental sodium (1 g) and maintained by continuous inhalation of 1.5% halothane in O2-N2O (50:50 vol/vol). Catheters were inserted into the fetal carotid artery, jugular vein, and amniotic sac to monitor fetal well being. Two catheters were also inserted into the fetal trachea, one directed toward the lungs and the other directed toward the larynx (19). Another catheter was inserted into the left main bronchus, via the trachea, and the left main bronchus was ligated around this catheter (13, 34). All catheters were exteriorized from the ewe and the tracheal catheters, as well as the catheter inserted into the left main bronchus, were connected to produce a continuous, externalized tracheal loop. This enabled liquid to flow into and out of both the left and right lungs unimpeded during the following 7-day recovery period; this recovery period ensured that any changes in gene expression we detected were not associated with the surgical manipulation.
At 126d GA, which is during the saccular/early alveolar stage of lung development, the left main bronchus was obstructed for ∼36 h by obstruction of the left main bronchus catheter to increase expansion of the left lung; normal flow of lung liquid to and from the right lung was maintained. An increase in intraluminal pressure recorded from the left lung verified that the left bronchus was obstructed (13, 39); the pressure recorded from the right lung remained at control levels throughout the experimental period. At the conclusion of the 36-h experimental period, the left lung and the right lung were drained of lung liquid and the ewe and fetus were humanely killed with an overdose of pentobarbital sodium (6.5 g iv). The fetus was weighed and the lungs were removed, and the left and right lungs were separated and weighed. Portions of both the left and right lungs were frozen in liquid nitrogen and stored at −70°C.
A separate group of control fetuses (n = 5) and fetuses exposed to 36 h of increased lung expansion (n = 5) were used for Northern blot analyses to verify differential gene expression. These fetuses underwent surgery at ∼118d GA, and tracheal catheters, but not the left bronchus catheter, were inserted to control liquid flow from the lungs as described above. At ∼129d GA, in the experimental fetuses only, the tracheal catheter was obstructed for 36 h, causing an increase in lung expansion (36 h TO) in both lungs of these fetuses. At the end of the experiment, the lungs of all fetuses were drained of liquid and the ewes and fetuses were humanely killed with an overdose of pentobarbital sodium (6.5 g iv). At autopsy, each fetus was weighed, the lungs were removed and weighed, and the left bronchus was ligated. The left lung was removed distal to this ligature, and portions of the left lung were frozen in liquid nitrogen and stored at −70°C for biochemical analysis. The right lung was pressure fixed at 20 cmH2O in 4% paraformaldehyde and postfixed for future light and electron microscopy.
Lung tissue from one fetus exposed to the 36-h left main bronchus obstruction was used for this analysis. Total RNA was extracted separately from both the left and the right lobes of the lung using a modified guanidine thiocyanate method (7), and 2 μg of poly(A+) RNA were subsequently isolated for each lung (Dynabeads mRNA purification kit; Dynal Biotech, Oslo, Norway). “Forward” and “reverse” subtraction hybridization was performed using the PCR-Select cDNA Subtraction kit (Clontech, Palo Alto, CA), following the recommended protocol, to isolate genes that are upregulated and downregulated by increased lung expansion. In the forward subtraction, the cDNA from the left lung served as the “tester” cDNA, and the “driver” cDNA was derived from the right lung. In the forward subtraction, upregulated genes (i.e., more abundant genes) in the left lung (overexpanded) compared with the right lung (control) were isolated. Conversely, in the reverse subtraction, the tester cDNA and the driver cDNA were the cDNA from the right and left lungs, respectively. In the reverse subtraction, genes that are downregulated by increased lung expansion (i.e., more abundant in the right control than the left expanded lung) were isolated. All reagents apart from the cDNA templates were supplied in the PCR-Select kit. Briefly, cDNA was synthesized from the mRNA that was isolated from both lungs. This cDNA was then digested with Rsa I to produce shorter, blunt-ended cDNA fragments. An aliquot of the digested cDNA was ligated to adaptor 1, and another aliquot of digested cDNA was ligated to adaptor 2R using T4 DNA ligase. An excess amount of Rsa I-digested driver cDNA was added separately to each tester cDNA ligated to either adaptor 1 or adaptor 2R. The samples were heat denatured at 98°C for 1.5 min and allowed to anneal at 68°C for 8 h. These two samples were then combined, and fresh denatured driver cDNA was added. The reaction was incubated at 68°C overnight to allow the samples to anneal. Primary PCR was then performed on an aliquot of the subtracted cDNA using PCR primer 1, and secondary PCR was performed using nested PCR primers 1 and 2R. This procedure enables only the differentially expressed fragments to be exponentially amplified; all other fragments fail to be amplified or are amplified only in a linear fashion. The final result is two pools of cDNA fragments highly enriched for up- and downregulated cDNA fragments, respectively.
The isolated double-stranded cDNA fragments from the secondary PCR were ligated into the pGEM-T easy vector (Promega, Madison, WI) and transformed into Electromax DH5α-E cells (Invitrogen, Carlsbad, CA). Three thousand seventy-two bacterial colonies, each containing a subtracted cDNA fragment from each lung, and therefore a differentially expressed gene, were screened for false positives by spotting them out robotically (BioRobotics BioGrid; Australian Genome Research Facility, Melbourne), in duplicate, onto identical nylon filters. Each pair of identical filters, containing either the upregulated or downregulated genes, was hybridized separately with 33P-labeled probes made from the subtracted cDNA isolated from the left lung and right lung, according to the manufacturer's instructions for the PCR-Select Differential Screening kit (Clontech, Palo Alto, CA). Briefly, 20–90 ng of subtracted cDNA (i.e., upregulated or downregulated cDNA fragments) were used as a template to generate the 33P-labeled cDNA probes. The membranes were hybridized to each 33P-labeled probe, separately, overnight at 72°C with continuous agitation. The next day, the membranes were washed four times in 2× SSC, 0.5% SDS, and twice in 0.2× SSC, 0.5% SDS, at 68°C for 20 min/wash. The membranes were sealed in plastic and exposed to a phosphor screen, which was scanned using a phosphorimager (Storm 860; Molecular Dynamics, Sunnyvale, CA) to obtain a digital image.
Candidate genes were selected by comparison of the spot intensity of the hybridization between each pair of membranes (ArrayGauge; Fuji Photo Film, Tokyo, Japan). Clones that exhibited a fivefold or greater intensity on the blot containing upregulated clones and hybridized with upregulated genes (Fig. 1A), compared with an identical blot hybridized with downregulated genes (Fig. 1B), represented genes likely to be upregulated by increased lung expansion. Clones that exhibited a fivefold or greater intensity on the blot containing downregulated clones and hybridized with downregulated genes (Fig. 1D), compared with an identical blot hybridized with upregulated genes (Fig. 1C), represented genes likely to be downregulated by increased lung expansion. Genes that hybridized at a similar intensity were likely false positives and were eliminated from further investigation. The clones that were selected as candidate genes were sequenced and then identified using the nucleotide-nucleotide basic local alignment search tool (BLAST) search engine of the GenBank database (http://www.ncbi.nlm.nih.gov).
Northern Blot Analysis
Differential expression of the up- and downregulated genes was verified by Northern blot analysis, as previously described (28), using separate groups of animals. Briefly, total RNA was extracted from the lung tissue of control fetuses and fetuses exposed to 36 h of increased lung expansion (n = 5 for each group). Twenty micrograms of RNA from each fetus were denatured, loaded into separate wells, and electrophoresed on a 1% agarose gel containing 2.2 M formaldehyde. The RNA was transferred by capillary action to a nylon membrane (Duralon; Stratagene, Cedar Creek, TX) and then cross-linked to the membrane by ultraviolet light (Hoeffer UVC500).
The membranes were hybridized with 32P-labeled cDNA probes generated from differentially expressed clones. After overnight hybridization at 42°C, the membranes were washed, sealed in air-tight bags, and exposed to a storage phosphor screen (Kodak S0230, Molecular Dynamics) at room temperature for 2–4 days. A digital image was obtained using a phosphorimager (Storm 860, Molecular Dynamics) before the membranes were stripped and rehybridized with an ovine-specific 32P-labeled cDNA probe for 18S rRNA, to adjust for minor loading differences between lanes. The relative mRNA levels of each gene were analyzed by measuring the density of each band (ImageQuaNT software, Molecular Dynamics). The density of each band was then expressed as a proportion of the density of the 18S rRNA band for that lane and as a percentage of the mean of the control values on that blot.
Results are presented as means ± SE. Student's unpaired t-test was used to test for differences in gene expression between control fetuses and fetuses exposed to 36 h of increased lung expansion. A P value <0.05 was taken to be statistically significant.
We screened 3,072 clones containing potentially upregulated cDNA fragments. Using a minimum fivefold difference in spot intensity (Fig. 1), 1,138 cDNA fragments were selected as likely true positives, i.e., genes that were upregulated by 36 h of increased lung expansion. Sequencing 112 of the highest-intensity clones demonstrated that 91 of those cDNA fragments arose from 60 genes present in GenBank, whereas 22 clones did not match with significant homology to genes in the GenBank database (see Table 1). Of the genes present in GenBank, 49 encoded proteins with known or putative function; 15 are involved in protein processing, 12 are signaling proteins, 10 are structural proteins, 6 have roles in cell proliferation, 3 are involved in cellular metabolism, and 3 are transcription factors.
Verification of differential expression.
The mRNA levels of seven of the upregulated genes with known function were verified in separate groups of control fetuses and fetuses exposed to 36 h of TO (Fig. 2). The mRNA levels of connective tissue growth factor (CTGF) (controls 100.0 ± 2.1% vs. 36-h TO 124.4 ± 8.4%; P < 0.05) and cysteine-rich angiogenic inducer, 61 (CYR61) (controls 100.0 ± 6.0% vs. 36-h TO 130.0 ± 5.2%; P < 0.01), were increased after 36 h of increased lung expansion. Heat shock protein 47 (HSP47) expression also increased from 100.0 ± 14.6% in control fetuses to 229.5 ± 31.2% in fetuses exposed to 36 h of increased lung expansion (P < 0.05). Thrombospondin-1 (TSP-1) expression increased to 347.5 ± 73.6% of control values (100.0 ± 14.0%) following 36 h of increased lung expansion (P < 0.05). Similarly, trophoblast antigen 2 (TROP2) expression increased from 100.0 ± 18.3 to 214.6 ± 12.7% (P < 0.005), and tropoelastin (TPE) mRNA levels increased from 100.0 ± 14.0% to 262.6 ± 46.0% (P < 0.05) in response to increased lung expansion. Tubulin-α3 expression was also upregulated by 36 h of increased lung expansion (controls 100.0 ± 14.9% vs. 36-h TO 281.8 ± 33.0%; P < 0.005).
We screened 3,072 potentially downregulated cDNA fragments. Using a minimum fivefold difference in spot intensity as the selection criteria (Fig. 1), 181 of the cDNA fragments were designated likely true positives, i.e., genes downregulated by 36 h of increased lung expansion. After sequencing, these cDNA fragments were found to arise from 22 different genes with known or putative function (see Table 2). Thirteen of the downregulated genes had functions related to the control of protein translation or posttranslational modifications, four regulate cellular metabolism, two are involved in cell proliferation/differentiation, two have putative roles in host defense, and one is thought to regulate cell structure. There were also four cDNA fragments (arising from 23 clones) that were not homologous with genes in the GenBank database and could not be identified. There was a high degree of redundancy among the downregulated clones such that ∼60% of the clones encoded elongation factor-1α (EF-1α; 94 cDNA fragments) and vitamin D3 upregulated protein-1 (VDUP1; 15 cDNA fragments). This was likely due to the Rsa I digestion of the full-length cDNA, generating small cDNA fragments that are independently amplified during the PCR procedure, resulting in multiple copies, and different fragments, of the same gene being cloned and sequenced. The high proportion of EF-1α clones may also be because it is the second most abundant protein in mammalian cells (8).
Verification of differential expression
The mRNA levels of three of the downregulated genes were verified in separate groups of animals at 0- and 36-h TO (Fig. 3). Although Clara cell secretory protein-related protein-1 (CCSP-RP1) mRNA levels at 36-h TO were 41.7 ± 8.8% of control levels (100.0 ± 40.3%), this was not significant due to the large variation in control fetuses. Similarly, EF-1α and VDUP1 mRNA levels tended to be reduced from 100.0 ± 7.1% and 100.0 ± 23.4% in control fetuses to 84.0 ± 3.0% (P = 0.08) and 51.8 ± 7.4% (P = 0.1), respectively, in fetuses exposed to 36 h of TO. After 2 days of TO, however, fetal lung mRNA levels for all three of these proteins were significantly (P < 0.05) reduced [CCSP-RP1 100.0 ± 10.5% vs. 65.3 ± 6.7%; EF-1α 100.0 ± 3.7% vs. 60.9 ± 2.2%; and VDUP1 100.0 ± 8.3% vs. 37.2 ± 3.7% (9)].
It is well established that the degree to which the fetal lungs are expanded with liquid is a major determinant of fetal lung growth and development (20, 34, 39). Although the mechanisms involved are unknown, it is possible that the stretch stimulus is transmitted throughout the lung via the ECM, causing a change in cell shape. This, in turn, may either directly affect gene transcription (32, 45) or initiate a signaling cascade via activation of ECM receptors (21, 22) or by activation of stretch-sensitive ion channels (29). In either case, it is likely that an alteration in gene expression occurs, leading to increased cell proliferation, alveolar epithelial cell differentiation, and structural maturation of the lung. In this study, we have shown that an increase in fetal lung expansion induces differential expression of over 80 genes. Because many of these genes have known roles in cell growth, differentiation, and ECM remodeling, it is likely that they play an important role in stimulating and coordinating the normal growth and development of the lung during the alveolar stage. We also isolated immediate early response genes, including early growth response 1 (EGR1), CTGF, and CYR61, indicating that the stimulus and technique we used were very effective in identifying factors that may initiate the response to increased lung expansion.
The animal model we developed to perform this analysis enabled us to compare genes expressed in expanded and control lung tissue from the same animal (13), thereby avoiding the problem of detecting genes simply due to normal genetic variation between individuals. It also allowed us to separate the timing of the experiment from the time of surgery to avoid detecting surgery-related alterations in gene expression. We have previously used differential display RT-PCR to identify genes altered by increased fetal lung expansion (13), but only a few genes were identified (13), as that technique favors the isolation of noncoding sequences (3′-UTR) that are often not present in gene databases. Thus, in this study, we used subtraction hybridization, which isolates cDNA sequences more likely to contain coding regions, thereby increasing the chance of gene identification.
Many of the upregulated genes we have identified have previously described roles in cell proliferation, angiogenesis, or ECM remodeling and are, therefore, likely candidates for potential regulators of those processes during normal lung development. For example, the upregulated gene EGR1 is a transcription factor that induces the expression of cell cycle proteins such as thymidine kinase and cyclin D1, growth factors such as vascular endothelial growth factor (VEGF), and ECM molecules like fibronectin and collagen III (12, 35). CTGF and CYR61 are immediate early response genes in the CCN gene family (36) that encode matricellular proteins that link the ECM to cell adhesion molecules such as integrins. In the lung, they are secreted by endothelial, epithelial, and airway smooth muscle cells, and in other tissues they have diverse functions, including roles in fibroblast proliferation, cell differentiation, ECM production, and angiogenesis (26). Surprisingly, CTGF and CYR61 have not previously been implicated in the regulation of lung development, although CTGF knockout mice die within minutes of birth (24). Mechanical stretch is a key regulator of CYR61 (via activation of EGR1) (16) and CTGF (26) gene expression, supporting their identification as genes upregulated by mechanical stimuli like increased lung expansion. TSP-1 is a secreted ECM protein with a variety of functions, mediated in part by the large variety of proteins to which it binds (1). Depending on the cell type and ligand, TSP-1 either stimulates or inhibits angiogenesis, cell proliferation, and cell migration; for example, TSP-1 is anti-angiogenic and anti-proliferative in endothelial cells (3, 15) but stimulates proliferation of venous endothelial cells (5) and smooth muscle cells (30). We have previously shown that endothelial cells, epithelial cells, and fibroblasts all proliferate in response to 2 days of TO (40), which may be due to the increase in EGR1, TSP-1, CTGF and/or CYR61 expression. Further studies are needed to elucidate the role of these genes in fetal lung development.
We isolated several other genes that also have roles in cell hypertrophy and hyperplasia, which are key responses to an increase in fetal lung expansion. VDUP1 inhibits cell proliferation via its interaction with thioredoxin and/or cyclin A2 (17), and downregulation of a growth suppressor like VDUP1 could be permissive in allowing fetal lung growth to proceed at an accelerated rate. Indeed, we have recently shown that VDUP1 has a highly significant and inverse relationship with cell proliferation rates during normal, accelerated, and delayed fetal lung growth (9), suggesting that VDUP1 may be a critical regulator of fetal lung cell proliferation. At 2 days of TO, there is a greater increase in protein content than DNA content of the lung, suggesting that cell hypertrophy precedes cell hyperplasia (39). It was surprising, therefore, to find a reduction in mRNA levels of some of the protein production machinery such as ribosomal proteins and regulators of protein translation (EF-1α and eIF4A) at 36 h of TO. It is possible that these genes were upregulated at an earlier time point to accelerate transcription and translation and that the subsequent downregulation of these genes may act to limit cell hypertrophy. Indeed, we have shown that there is no further increase in lung protein content after 2 days of TO (39).
Deposition and remodeling of the ECM are key features of lung maturation and are closely associated with alveolarization, reduced blood-gas diffusion distances, and increased lung tissue compliance. For example, the deposition of elastin is intimately involved in alveolar formation (33, 48), collagen I is critical for providing tensile strength, collagen III aggregates at the entrance rings of alveoli, providing flexibility, and collagen IV is a critical component of basement membranes that compartmentalizes the tissue. Numerous genes involved in the synthesis of ECM components and ECM remodeling such as HSP47 and CTGF were upregulated by increased fetal lung expansion. HSP47 transports the mature form of procollagen to the Golgi apparatus, where it is secreted from the cell and incorporated into the ECM (37). An increase in HSP47 expression in vitro is associated with an increase in collagen production and accumulation in the ECM (43), whereas mice lacking HSP47 die during embryonic development due to a failure of mature type I and IV collagen deposition within the ECM (37). CTGF also stimulates the production of collagens I and IV and fibronectin (36). In keeping with the roles of HSP47 and CTGF, we also isolated fibronectin and collagen types III and IV as genes upregulated by increased lung expansion, which is consistent with the increase in collagen accumulation induced by TO (39). Similarly, the increase in expression of the elastin precursor TPE at 36 h of TO confirms our previous finding of an ∼250% increase in TPE expression in response to 2 days of TO (25). The increase in TPE expression is consistent with the increase in alveolar number that occurs during normal lung development and following an increase in fetal lung expansion (2, 40). The upregulation of these genes clearly indicates that structural development of the lung is enhanced by increased fetal lung expansion and suggests that HSP47 and CTGF may mediate the effect of increased lung expansion on collagen deposition.
Increases in fetal lung expansion also induce alveolar epithelial cell (AEC) trans-differentiation, favoring an increase in the proportion of type I AECs and a decrease in the proportion of type II AECs. The reduction in the proportion of type II AECs is closely paralleled by a reduction in the mRNA levels of the type II AEC marker, surfactant protein C (SP-C) (10, 11, 28, 41). The isolation of SP-C as a downregulated gene in the current study confirms the validity of this technique to identify differentially expressed genes. Recently, fatty acid-binding protein, glutathione transferase and secretoglobin, in addition to SP-C, were all identified as markers of type II cells (14). Their isolation as downregulated genes in the current study, therefore, is consistent with the known effects of increased lung expansion on type II cells and their markers.
It has been proposed that cells exist in a state of isometric tension generated by intracellular contractile filaments and that externally applied forces imposed on this preexisting force equilibrium cause changes in intracellular structural fiber alignment until the force equilibrium is reestablished (6, 23, 31). This may lead to alterations in cell shape, cell proliferation, differentiation, migration, and/or synthesis of ECM components. Microtubules are key components of the cytoskeleton and form a trans-membranous structural continuum with the ECM (47), placing them in an ideal position to sense and transduce a mechanical stimulus. They are composed of α- and β-tubulins and have diverse functions including cell division (47) and membrane receptor regulation (27). The increase in expression of tubulin-α3 and -α1 at 36 h of TO is consistent with the known effect of stretch on the synthesis of cytoskeletal elements (6, 31), and we speculate that the increase in tubulin-α3 and -α1 mRNA levels is a cellular response to an increase in tension. Numerous signal transduction pathways are activated by physical forces (23) and are likely involved in both transducing the initial stimulus and mediating the subsequent growth response. This study has isolated a variety of genes, including adenylate cyclase 6, calmodulin-like 4, and TROP2, that may act as mechanotransduction mediators. Adenylate cyclase 6 catalyzes cAMP formation, which would increase the potential for activation of cAMP-signaling pathways. TROP2 is a cell surface glycoprotein that may act as a receptor that transduces calcium signals (42). Previous studies have suggested that calcium is an important second messenger in mechanotransduction in the fetal lung. For example, blockade of calcium channels with nifedipine causes hypoplasia in lung explants (44), and gadolinium, a stretch-sensitive calcium channel blocker, has been shown to inhibit lung cell proliferation (29). We have also previously identified the genes encoding calmodulin (CALM-1, -2 and -3), a mediator of calcium signaling, as genes that are upregulated by increased lung expansion (Ref. 13 and unpublished data), further supporting a role for calcium in the response to increased lung expansion.
Although none of the downregulated genes examined by Northern blots was significantly reduced at 36 h of TO, they were all significantly reduced by 2 days of TO. The latter finding confirms their initial identification as downregulated genes and highlights the sensitivity of the subtraction technique. Although the inability to detect a significant difference at 36 h of TO was predominantly due to the large variability in control values, it is also likely that small differences in lung distensibility and/or lung liquid accumulation rates will alter the time to threshold of gene activation/inactivation in different fetuses. It is also of interest that, using subtraction hybridization, we did not identify any of the common growth factors thought to mediate fetal lung development. This supports our previous findings in which we have been unable to detect an increase in the mRNA levels encoding platelet-derived growth factor (PDGF), insulin-like growth factor-II (IGF-II), transforming growth factor-β1 (TGFβ1), and the IGF type I receptor at 36 h, 2 days, or 4 days of TO (Wallace MJ, unpublished data); these periods coincide with the peak proliferative phase induced by TO (39). Similarly, the levels of active p38, p42, and p44 MAPK proteins are not changed, indicating that growth factors utilizing the MAPK signal transduction pathway are unlikely to be involved in the growth response (Wallace MJ, unpublished data). However, we found that one isoform of VEGF was increased by 36 h of TO, and it is possible that one of the upregulated cDNA fragments isolated, but not sequenced yet, encodes VEGF.
In this study, we have identified over 80 genes that are differentially expressed in response to an increase in fetal lung expansion, with only a few of these (TPE and SP-C) having been identified previously. Many of the identified genes have well-established roles in cell proliferation, differentiation, and ECM deposition or remodeling, but the roles of many of these genes (e.g., VDUP1, CTGF, and CYR61) in lung growth and development have not been previously explored. Although further studies are needed to verify the roles of these genes in expansion-induced lung growth, we speculate that EGR1, CTGF, and CYR61 may be among the first genes to respond and may initiate many of the growth and structural changes that occur following an increase in fetal lung expansion. We also suggest that alterations in many of the other genes may contribute to the coordinated increase in lung growth and structural maturation of the gas-exchanging regions that occur in response to an increase in fetal lung expansion.
The experiments described in this manuscript were funded by the National Health and Medical Research Council of Australia.
We are indebted to Alex Satragno for surgical assistance and to Alison Thiel for expert technical assistance.
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: M. Wallace, Dept. of Physiology, PO Box 13F, Monash Univ., VIC 3800, Australia (e-mail:)
↵* F. Sozo and M. J. Wallace contributed equally to this study.
- Copyright © 2006 the American Physiological Society