Abnormal mouse lung alveolarization caused by Smad3 deficiency is a developmental antecedent of centrilobular emphysema

Hui Chen, Jianping Sun, Sue Buckley, Cheng Chen, David Warburton, Xiao-Fan Wang, Wei Shi


Transforming growth factor-β (TGF-β) signaling plays an important regulatory role during lung development and remodeling. Smad3 is a major downstream signal transducer in the TGF-β pathway from the cell membrane to the nucleus. In Smad3 null mutant mice, we have observed retarded lung alveolarization from postnatal day 7 to day 28, and subsequently centrilobular emphysema starting from day 28, as determined by morphometric analysis. In addition to the morphological changes, peripheral lung cell proliferation in Smad3 knockout mice was reduced compared with the wild-type control between postnatal days 7 and 28. Expression of tropoelastin at the mRNA level was also dramatically decreased in Smad3 knockout lungs from postnatal day 28 through adulthood. Furthermore, increased matrix metalloproteinase-9 protein expression and activity were detected in the Smad3 knockout mouse lung tissue and the bronchoalveolar lavage fluid at postnatal day 28 when the centrilobular emphysema pathology was just beginning to appear. Therefore, these results indicate that Smad3 not only has a positive regulatory impact on neonatal lung alveolarization but also potentially plays a protective role against the occurrence of centrilobular emphysema later on in life.

  • transforming growth factor-β
  • matrix metalloproteinase-9

transforming growth factor (TGF)-β is a subfamily of growth factors regulating cell proliferation, differentiation, apoptosis, migration, and syntheses of extracellular matrix (ECM) proteins (26). As studied in cultured cells, extracellular TGF-β binds and activates its cognate receptor complex on the cell surface. The activated receptor kinase subsequently phosphorylates the intracellular receptor-regulatory Smad2 and Smad3 proteins. These phosphorylated Smads then dissociate from the receptors, form complexes with a co-mediator Smad4, and translocate into the nucleus, acting as transcriptional comodulators to induce or repress differential TGF-β target gene expression (2, 14, 24). In addition to the canonical Smads, several Smad-independent pathways have also been reported to transmit TGF-β signals from the cognate cell membrane receptors to the nucleus (8). Although both Smad2 and Smad3 are seemingly homologous TGF-β-specific signaling molecules, nonredundant activities of Smad2 and Smad3 are observed in mice with a specific gene mutation. Null mutation of Smad2 results in early embryonic lethality (29), whereas the knockout of Smad3 results in a viable mouse that has several postnatal abnormal phenotypes (7, 31, 34). It also has been shown that abrogation of Smad3 function attenuates bleomycin-induced pulmonary fibrosis and accelerates wound healing, indicating that Smad3 alone is sufficient to mediate a number of important pathophysiological responses (1, 33).

TGF-β signaling plays an important role in regulating fetal lung morphogenesis, postnatal lung growth, injury repair, and remodeling. Excessive TGF-β signaling may contribute to the etiology of bronchopulmonary dysplasia and pulmonary fibrosis (12, 33). Moreover, abrogation of the active TGF-β ligand regulators, fibrillin, latent TGF-β binding protein 4 (LTBP4), and β6-integrin all cause reduced lung alveolarization and subsequent emphysema in mice accompanied by changed TGF-β activity, suggesting an important regulatory function for TGF-β signaling in postnatal lung growth and emphysema prevention (17, 18, 27). However, there is no direct evidence that abrogation of intracellular TGF-β signaling activity causes abnormal alveogenesis and subsequent emphysema, since fibrillin-1, LTBP4, and β6-integrin are also involved in other key extracellular events, such as the correct formation of microfibrils.

Lung emphysema is classified as a major component of chronic obstructive pulmonary disease (COPD). The pathology of emphysema comprises “abnormal, permanent enlargement of air spaces distal to the terminal bronchioles accompanied by destruction of their walls without obvious fibrosis” (25). These pathological changes result in dramatic loss of functional gas-exchange surfaces and compliance of the lung, exacerbating failure of respiratory function in COPD patients. Cigarette smoke is considered to be a major exogenous factor causing emphysema. Moreover, genetic factors also contribute to the susceptibility to irreversible destruction and dilation of the terminal air spaces. The mechanisms of emphysema pathogenesis are not completely defined. However, imbalanced protease/antiprotease activities are thought to be one of the major mechanisms contributing to a large number of emphysema cases (3). Changes in protease expression and activities are regulated by many factors, such as mechanical stress, free oxide radicals, cytokines, and growth factors as well as extracellular proteins. Abnormal lung development and maturation may result in perturbation of these factors, subsequently causing imbalance of extracellular protease activities, and finally results in pulmonary emphysema.

In the present study, we therefore compared the postnatal lung alveolarization and growth of Smad3 knockout vs. wild-type control mice by morphology, cell biology, and molecular biology to elucidate the direct effect of TGF-β signaling in normal alveogenesis and subsequent emphysema pathogenesis. We found that loss of Smad3 function caused reduced pulmonary cell proliferation with retarded alveogenesis in neonatal mouse lung, which was followed by centrilobular emphysema that continued to progress during adulthood.


Smad3 null mutant mice.

Smad3 null mutant mice were always bred from C57BL/6 mice heterozygous for a targeted disruption of exon 1 of the Smad3 gene (7). The genotypes of both wild-type and Smad3-deficient mice were determined by PCR analysis on tail DNA, as previously described (33). Mice were kept under specific pathogen-free conditions in the University of Southern California (USC) animal facility until use. Animal protocols used herein were approved by the USC Institutional Animal Care and Use Committee and were in accordance with National Institutes of Health guidelines for animal welfare. Phenotypic comparison was always carried out between wild-type and Smad3 knockout littermates.


The mouse lungs were inflated with 4% paraformaldehyde fixation solution through cannulated trachea under a fixed water pressure (25 cm for adult and 20 cm for neonates), and the excised lungs with ligated trachea were submerged in fixative overnight at 4°C, followed by routine processing and paraffin embedding. Tissue sections with 5-μm thickness were prepared and stained with hematoxylin and eosin (H&E) for histological examination. Bronchoalveolar lavage (BAL) fluid was also collected from the postnatal lung as published previously (33).

Elastin was stained as reported previously (23). Briefly, the deparaffinized sections are first treated in potassium permanganate for 30 min, followed by a rinse in a 5% oxalic acid solution, and then stained in resorcin-fuchsin solution overnight and finally in van Gieson's solution for 1 min.

Morphometric analysis of lung sections.

Five sections from the same lobes of each sample were randomly chosen at ∼250-μm intervals and stained with H&E. The mean linear intercept (MLI) was then measured according to established methods (9, 28). Briefly, an image of each examined section was digitally captured at ×40 magnification. The horizontal and vertical lines of a rectangle grid at 0.9-mm intervals were then used to count alveolar surface intersections. The MLI was calculated by the equation: the sum of the length of all counting lines divided by the total number of counted intercepts of alveolar septa. In addition, peripheral lung structures were also analyzed by radial alveolar counts using a method published previously (6, 10). Briefly, microscopic images of H&E-stained lung sections at ×100 magnification were acquired using a computer-assisted digital camera. The respiratory bronchiole with a wall partially lined by columnar epithelium was used as a starting point to draw a perpendicular line to the nearest connective tissue septum. The number of alveoli cut by this line was then counted. At least 15 counts were performed in each sample. All these morphometric data were analyzed with Student's t-test to compare the differences between mean values and were considered significant if P < 0.05.

Bromodeoxyuridine labeling and cell proliferation in vivo.

Bromodeoxyuridine (BrdU) was incorporated into DNA in place of thymidine during DNA synthesis in mitotic cells, following the method described previously (22). Briefly, BrdU (0.1 mg/g body wt) was intraperitoneally injected into the mice 3 h before lung isolation. The fixed lungs were then immunostained using a primary antibody against BrdU and a peroxidase-conjugated secondary antibody (Zymed). The peroxidase reaction products were brown colored with diaminobenzidine substrate. Normal serum was used in parallel slides as negative controls.

Tissue total RNA isolation and quantitative real-time RT-PCR.

Total RNA from lung tissues was isolated using RNeasy kit (Qiagen). The quality of isolated RNA was checked by gel electrophoresis before the reverse transcription reaction in which iScript reverse transcriptase (Bio-Rad, Hercules, CA) was used.

Real-time PCR was run using the iCycler iQ detection system (Bio-Rad) for intercalating SYBR Green I dye, which binds nonspecifically to double-stranded DNA. The intensity of accumulated fluorescence bound to the amplified DNA target reflects the level of cDNA template in the sample. The reaction was assembled according to the manufacturer's recommendations. Briefly, 25 μl of reaction mixture contained 1× iQ SYBR Green Supermix (Bio-Rad), 0.3 μM forward and reverse primers, and the cDNA template from the sample. The PCR conditions were 3 min at 95°C, followed by 40 cycles of 30 s at 95°C, 30 s at 60°C, and 30-s extension at 72°C. Fluorescent data are specified for collection during the extension step. A standard curve for the desired gene was established for the quantification purpose. GAPDH was used as internal control. The relative cDNA ratio was calculated using the value of threshold cycles (19). Primers for tropoelastin were TGC CAA AGC TGC TGC TAA GGC T and AGT CCA AAG CCA GGT CTT GCT G.

Western blot.

Detection of lung tissue proteins has been previously described (32). Briefly, fresh lung tissues were lysed on ice in radioimmunoprecipitation assay buffer containing 1 mM PMSF, 0.2 U/ml aprotinin, and 1 mM sodium orthovanadate. Protein concentration was measured by the Bradford method. Total tissue lysate proteins (40 μg) were loaded in each lane for SDS-PAGE, and the separated proteins were transferred to Immobilon-P membrane (Millipore). The membrane was then blocked for nonspecific binding overnight at 4°C by incubating in 5% fat-free dry milk in TBST (10 mM Tris·HCl, pH 7.4, 150 mM NaCl, and 0.1% Tween 20). Primary antibody was diluted in blocking buffer and incubated with the membrane for 1 h at room temperature. After being washed in TBST (10 min, 4×), horseradish peroxidase-conjugated secondary antibody was incubated for 45 min at room temperature. The antibody-detected protein bands were finally visualized by ECL reagent (Amersham). All antibodies of proliferating cell nuclear antigen (PCNA; Zymed), matrix metalloproteinase (MMP)-9, MMP-12 (Chemicon), and caspase-3 (Cell Signaling Technology) were purchased commercially. The intensities of protein bands were quantified by Scion Image software (Scion, Frederick, MD) and normalized by the loading control.

Measurement of MMP activities by SDS-PAGE zymography.

Equivalent volumes of BAL fluid were collected from each lung. After centrifugation at 300 g for 10 min at 4°C to pellet cells, the cell-free supernatants were concentrated ∼10-fold using Centricon 10 concentration units (Millipore, Chicago, IL) according to the manufacturer's instructions. Denatured concentrates of BAL (corrected for concentration factor) and lung tissue lysates (corrected for protein) were electrophoresed through commercially prepared gelatin (1 mg/ml; Novex, San Diego, CA) under nonreducing conditions. Conditioned medium from NIH/3T3 cells, which contained both MMP-9 and MMP-2, was included as control. The gels were renatured in 2.5% (vol/vol) Triton X-100 in water and at room temperature for 30 min and developed using Novex developer according to the manufacturer's instructions (30 min at room temperature, then overnight at 37°C). The gels were then stained for 30 min in hot 0.5% Coomassie brilliant blue R, destained, and dried at room temperature. Areas of protease activity showed up as clear bands that were quantified by Scion Image software.

Data presentation and statistical analysis.

All experiments were repeated at least three times, with similar results obtained within repetitive experiments. All data were expressed as means ± SD. Student's t-test was used for comparison of statistical difference among experimental groups with different Smad3 genotypes, and P values <0.05 were considered significant.


Smad3 deficiency causes reduced pulmonary alveolarization and subsequent development of centrilobular emphysema in postnatal mouse lung.

To determine the role of TGF-β-Smad3 signaling in postnatal lung growth and maturation, lung morphology of Smad3 knockout mice was compared with their wild-type littermates at various stages of growth (Figs. 1 and 2). Alveolarization starts at postnatal day 5 in mouse lung, when secondary alveolar crests develop and extend to make new secondary septa, resulting in an increase of alveolar units, accompanied by decreased mean alveolar size. In Smad3 knockout lung, no significant changes of alveolar size in the newborn mice (postnatal day 1) were observed (Fig. 1) compared with the wild-type littermates. However, slightly larger terminal air spaces were detected in Smad3 knockout lungs at postnatal day 7, when alveolarization had begun, which may have resulted from attenuated growth of septal structures. This difference persisted and became more severe as alveolarization proceeded. Furthermore, the most strikingly uneven enlargement of air spaces in peripheral lung of the Smad3 knockout mice occurred toward the end of alveolarization (postnatal day 28), after which the Smad3 null mutant lungs consistently exhibited characteristics of centrilobular emphysema. As the Smad3 knockout mice became older, enlarged air spaces occupied most of the lung. This centrilobular emphysema phenotype was not seen in wild-type mouse lung at the same age (4.5 mo). These changes in peripheral pulmonary alveolar size were also quantified by morphometric measurement of MLI (Fig. 2). No significant difference of MLI was detected on the first day of postnatal life between Smad3 knockout and wild-type controls. However, a 25% increase of MLI was detected in the Smad3 knockout mouse lung vs. wild-type at an early alveolarization stage (postnatal day 7). The difference further increased to 58% by the end of alveolarization (postnatal day 28), when typical centrilobular emphysema pathology was observed. In adult mice (4.5 mo), further enlargement of terminal air spaces, indicated by a 69% increase in MLI in Smad3 knockout lung, was detected. To exclude artifactual MLI changes caused by varied lung inflation and collapse, the radial alveolar count was also measured. Consistent with MLI changes, the radial alveolar count was significantly reduced starting from postnatal day 7, but not in newborn stage postnatal day 1 (Fig. 2). The number of peripheral alveoli in Smad3 knockout lung was further reduced from postnatal day 28 to 4.5 mo, possibly resulting from destruction of preformed alveolar walls. Therefore, Smad3 deficiency resulted in attenuated alveolarization in early life and subsequently in development of centrilobular emphysema pathology in later life.

Fig. 1.

Lung morphological changes of Smad3 knockout (KO) mice at different postnatal ages. The hematoxylin and eosin-stained tissue sections were obtained from the top lobe of the right lung at the indicated ages: postnatal days 1, 7, and 28 (P1, P7, and P28) and postnatal 4.5 mo. WT, wild-type mice. Images are representative of 3–5 individual mice of each genotype at different ages.

Fig. 2.

Morphometric analyses of Smad3 KO lung by mean linear intercept (MLI) and radial alveolar count. A: diameters (shown by bidirectional arrows) of the pulmonary alveoli in Smad3 KO mouse lung at P7 were significantly increased compared with WT controls. *Focal enlargement of terminal air spaces with adjacent normal-sized alveoli in Smad3 KO lung was apparent by P28. B: quantitative analyses of the pulmonary alveolar sizes at different ages, as measured by MLI. C: comparison of the numbers of peripheral alveoli using radial alveolar count. 4.5M, postnatal 4.5 mo. *P < 0.05. Figure is representative of 3–5 individual mice of each genotype at different ages.

Reduced peripheral lung cell proliferation during alveolarization of Smad3 knockout lung.

Secondary septal formation during alveolarization is a complicated process that requires coordination of outgrowth of epithelial cells, maturation of capillary networks, correct orientation of alveolar myofibroblasts, and appropriate deposition of the elastic interstitial matrix. To determine the cellular mechanisms of attenuated alveolarization in Smad3 knockout mouse lung, cell proliferation in mouse lung with different Smad3 genotypes was compared at different postnatal ages using PCNA as a molecular marker for cellular DNA synthesis. Compared with wild-type control, PCNA protein expression in Smad3 knockout lungs was significantly reduced by 19% at postnatal day 7 and 49% at postnatal day 28 (Fig. 3, A and B), indicating reduced cell proliferation during alveolarization. Furthermore, reduced pulmonary cell proliferation at postnatal day 28 was confirmed by BrdU incorporation 3 h before sampling (Fig. 3C). Proliferating epithelial cells in proximal bronchioles were detected in the lungs of both wild-type and Smad3 knockout mice at similar levels. However, cell proliferation in peripheral air spaces was significantly reduced in Smad3 null mutant lung compared with wild-type control, suggesting that reduced cell proliferation in the peripheral airways might result in less alveolarization in Smad3 knockout lung.

Fig. 3.

Reduced peripheral lung cell proliferation in Smad3 KO mice. A: decreased expression of proliferating cell nuclear antigen (PCNA), but not apoptotic protein caspase-3, in lung tissue lysates as detected by Western blot. GAPDH detection was used as the loading control. B: relative changes of PCNA protein expression in Smad3 KO mouse lung were compared with WT control. Bands in A were quantified using Scion Image software and normalized by protein loading control. Data are expressed as the percentage of WT control for each stage. Comparisons were performed in 3 pairs of Smad3 KO vs. WT control lungs at each stage. *P < 0.05. C: bromodeoxyuridine (BrdU) labeling of the proliferating lung cells at P28. The BrdU-labeled cells (brown color) in peripheral lungs were shown by arrows, and the BrdU positively stained cells in proximal large airways (arrowheads) were shown (insets) at the same magnification.

Apoptosis was also compared between Smad3 knockout and wild-type control lungs. The apoptotic protein caspase-3, as detected by Western blot (Fig. 3A), and cell nuclear DNA fragmentation, as detected by TdT-mediated dUTP nick end labeling staining (data not shown), were not different between wild-type and Smad3 knockout lungs.

Decreased tropoelastin expression in Smad3 knockout lung.

Reduced alveolarization and disruption of lung acinar structures could also be caused by abnormal synthesis and/or degradation of ECM proteins, particularly elastin fiber that provides structural recoil properties essential for respiratory function. Lack of elastin synthesis or accelerated degradation of elastin is one of the common causes of emphysema (30). Therefore, we examined the expression of tropoelastin, the monomeric form of elastin, and elastin distribution in the Smad3 knockout mouse lungs to assess the pathogenesis of abnormal alveolarization and emphysema. With the use of quantitative real-time PCR methods, significant reduction (5- to 6.7-fold) of tropoelastin expression in Smad3 knockout lungs was only detected from the late alveolarization stage (postnatal day 28) through adulthood compared with the gene expression of wild-type control lungs (Fig. 4A). Moreover, elastin fibers were reduced in number as a result of reduction of alveolar septae but appeared to be structurally normal in the remaining alveolar walls of the enlarged air space in Smad3 knockout mouse lungs at the age of 4.5 mo (Fig. 4B). No significant changes of type I procollagen α1-chain expression were observed between Smad3 knockout and wild-type lungs (data not included).

Fig. 4.

Reduced tropoelastin gene expression in Smad3 KO mouse lung. A: gene expression of tropoelastin at the mRNA level was measured by real-time quantitative RT-PCR. Tropoelastin mRNA level in Smad3 KO mouse lung was reduced 6.7-fold at P28 and 5-fold at 4.5 mo. The level of tropoelastin mRNA in WT tissues was arbitrarily defined as 100%. *P < 0.05. B: elastin staining of lung tissues at the age of 4.5 mo. Elastin fibers (black color, shown by arrows) were detected in the alveolar walls of Smad3 KO mouse lung, with the similar length and thickness as seen in WT control. However, the number of positively stained elastin fibers in Smad3 null mutant mouse lung was significantly reduced. Figure is representative of 3 individual mice of each genotype at specified ages.

Increased MMP-9 expression and activity in postnatal day 28 lung of Smad3 knockout mice.

Increases of extracellular proteases and/or decreases of protease inhibitors have been implicated as the direct cause for centrilobular emphysema in which excessive degradation of ECM proteins occurs. For example, increased MMP-12, MMP-2, and MMP-9 have been shown to be associated with emphysema (4, 11, 20). We therefore compared these MMPs between Smad3 knockout and wild-type control mouse lungs. MMP-9 protein expression in Smad3 knockout lung tissue was dramatically increased at postnatal day 28 compared with wild-type control (156 ± 4%, Fig. 5). There was no significant change in MMP-12 protein level. Furthermore, MMP-2 and MMP-9 protease activities were measured by gelatin zymography (Fig. 6). Consistent with the changes at protein level, MMP-9 activity was dramatically increased in Smad3 lung tissue (294 ± 15%, Fig. 6) and BAL fluid only at postnatal day 28, the time at which a typical centrilobular emphysematous histopathology was first observed. Therefore, MMP-9 might be one of the important proteases involved in the subsequent centrilobular emphysema caused by Smad3 deficiency. On the other hand, no excessive inflammation was detected in the Smad3 knockout lungs at age postnatal day 28 or adult, as measured by the number and profile of inflammatory cells in BAL fluids (data not shown and Ref. 33), which suggests that the increase of MMP-9 expression and activity may result from upregulation of this gene in lung parenchymal cells, not from infiltrated inflammatory cells.

Fig. 5.

A: protein expression of matrix metalloproteinase (MMP)-9 and MMP-12 in Smad3 KO and WT mouse lung tissues was detected by Western blot. GAPDH detection was used as protein loading control. B: relative changes of MMP-9 protein expression were compared in 3 pairs of Smad3 KO vs. WT control lungs at each stage. MMP-9 protein bands in A were quantified using Scion Image software and normalized by protein loading control. Data are expressed as the percentage of WT control for each stage. *P < 0.05.

Fig. 6.

A: MMP-2 and MMP-9 activities in lung tissues and bronchoalveolar lavage fluid (BALF) of the Smad3 KO and WT control mice were analyzed by gelatin zymography. The supernatant of BALF was concentrated in the same scale among different samples and loaded with equal volume. B: the relative change of MMP-9 protease activity in lung tissue lysate was quantified using Scion Image software. Data are expressed as the percentage of the MMP-9 activity in WT control for each stage. Comparisons were performed in 3 pairs of Smad3 KO vs. WT control lungs at each stage. *P < 0.05.


It is well known that TGF-β plays an important role in normal lung development and remodeling as well as injury repair and pulmonary inflammation. In Smad3-deficient mice, pulmonary alveogenesis seems retarded, as shown by mean size of alveoli and radial alveolar count at the early alveolarization stage (postnatal day 7). Several biological processes determine the rate of pulmonary septation, including cell proliferation, cell apoptosis, and ECM deposition. We found that overall, peripheral pulmonary parenchymal cell proliferation was attenuated in the absence of Smad3 as the earliest event when alveolarization started (postnatal day 7), whereas the proximal airway epithelial cells had the same growth rate. No difference in cell apoptosis between Smad3 knockout and wild-type control lungs was detected. Therefore, decreased cell proliferation of peripheral lung may account for the attenuated septal formation and hence alveolarization in Smad3 knockout mice, resulting in abnormally larger alveoli in Smad3 knockout mice compared with wild-type control at early alveolarization stage.

Furthermore, by the end of alveolarization (postnatal day 28), Smad3 knockout lung displayed the onset of typical centrilobular emphysema pathology in which focal alveolar enlargement mainly occurred around respiratory bronchioles. The size of alveoli (MLI) in Smad3 knockout lung was increased by 58% compared with the wild-type lung at the same age. This phenotype persisted and worsened into adulthood (4.5 mo). Thus the retarded alveolarization of Smad3-deficient lung had developed into the typical pathology of emphysema by the age of 4 wk.

The potential role of TGF-β in the pathogenesis of abnormal alveolarization and subsequent emphysema has also been suggested by several other studies. In mice with a fibrillin-1 hypomorphic allele or null mutant genotype, progressive distal air space enlargement was observed immediately after birth due to impairment of distal alveolar septation, followed by development of typical emphysematous changes in later life (18). However, increased TGF-β signaling activity was reported to accompany the emphysematous morphology in this case. Furthermore, abrogation of LTBP4 caused profound defects in elastin fiber structure and emphysema with reduced TGF-β signaling activities in lung tissues (27). Both fibrillin-1 and LTBP4 are integral components of the microfibrils that form elastic fibers with tropoelastin cores in addition to regulating TGF-β ligand activation. Deficiency of these proteins may therefore directly interfere with the formation of normal functional elastic fibers, which are essential for normal alveolar formation and function. It is possible that changes in TGF-β activity are merely coincident with changed elastic fiber function in the above scenarios. Moreover, Morris et al. (17) reported that loss of integrin αvβ6-mediated TGF-β activation resulted in MMP-12-dependent emphysema in β6-integrin null mutant mice. Therefore, direct evidence is required to determine whether the TGF-β-specific pathway itself is involved in promoting pulmonary alveogenesis and preventing emphysematous pathology and whether the TGF-β-Smad canonical pathway is responsible for these functions. Herein, our data strongly suggest a potential role of TGF-β-Smad3 signaling in promoting mouse postnatal alveolarization as well as preventing emphysema pathogenesis in later life.

Centrilobular emphysema usually results from excessive destruction of alveolar walls due to disturbed homeostasis of cell growth and ECM protein metabolism. The appropriate balance of proteinases and antiproteinases is a key factor responsible for maintaining alveolar wall integrity. MMPs are important proteases involved in emphysema (21). Increased gelatinase (MMP-2 and MMP-9) and macrophage elastase (MMP-12) were reported to be associated with emphysema (4, 11, 20). MMP-9, which is produced from many different cell sources, is able to degrade a number of key ECM components, including elastin and type IV collagen, and also can activate other proproteases. A polymorphyism in the promoter of MMP-9 has been associated with an increased incidence of emphysema among smokers in Japan (16). The diversity of changes in MMPs in emphysematous lungs suggests that there might be important differences in susceptibility between individuals and even within the same individual at different times.

Therefore, we have examined the expression and/or activity of MMP-2, MMP-9, MMP-12, and other MMPs at different time points. Interestingly, increased expression and activity of only MMP-9 were detected in Smad3 knockout lung tissues as well as in BAL fluid at the particular postnatal day 28 time point when the centrilobular emphysema pathology was first detected, suggesting that MMP-9 is a possible candidate responsible for the transition from retarded secondary septa formation to excessive destruction of preformed alveolar walls. However, the changes of MMP-9 level and activity did not last indefinitely, since no significant change of MMP-9 protein level and activity was detected in the lung tissue at the age of 4.5 mo. The molecular mechanisms of MMP-9 upregulation in Smad3 knockout lung at postnatal day 28 are still unknown. Retarded alveolarization may result in reduced lung compliance and hence more mechanical stress toward the end of the alveolarization stage. MMP-9 expression may thus be upregulated either directly or indirectly. Increased MMP-9 may also trigger excessive degradation of ECM proteins, particularly elastin, resulting in further destruction of already imperfectly formed alveolar wall structures around respiratory bronchioles, and this may further reduce the compliance of the lung. Thus, respiratory function deteriorates much more rapidly than in wild-type control, and the lung proceeds into an MMP-9-independent, progressive emphysema.

Reduced expression of tropoelastin was also detected in Smad3 knockout lung from the end of alveolarization (postnatal day 28) through adulthood, but not in early postnatal ages (postnatal days 1 and 7). Consistently, fewer elastin fibers, but with normal structure, were detected in adult Smad3 knockout lung by special histological staining. The molecular mechanisms of this gene downregulation within a specific time course remain unclear. Tropoelastin is a core component of elastin fibers. Reduction of tropoelastin expression may change the quantity and quality of functional elastin fibers that form a key component of alveolar wall structural proteins, which are required to maintain normal lung compliance. Previous studies showed that TGF-β1 increased the expression of elastin by stabilization of tropoelastin mRNA through several signaling pathways, including TGF-β-specific Smads (13, 15). However, abrogation of one downstream Smad (Smad3) activity alone may not be sufficient in reducing tropoelastin expression in early postnatal ages. The excessive mechanical and other stresses to the abnormal lung of Smad3 knockout mice during breathing may result in perturbation of the other TGF-β pathways, finally leading to reduction of tropoelastin expression. Therefore, the changes in tropoelastin gene expression of Smad3 knockout lung are possibly a secondary consequence of abnormal alveolarization, since tropoelastin expression is originally normal at the beginning of the alveolarization process. This downregulation of tropoelastin expression probably further contributes to subsequent emphysematous pathology.

During the preparation of this manuscript, Bonniaud et al. (5) reported that Smad3 null mice develop emphysema-like air space enlargement at the age of 8 wk. However, an abnormal alveolarization phenotype per se was not described. In that paper, no significant changes of alveolar size before the age of 8 wk were reported even though “incomplete alveogenesis with some enlarged terminal bronchioles” was observed. The apparent discrepancy between the latter data and ours could be caused by several factors. Different Smad3 knockout strategies were used, with exon 1 deletion of Smad3 genomic DNA in our study (7) and exon 8 deletion in Bonniaud and colleagues' study (5, 31), even though both knockout mice were on a C57BL strain background. Also, different inflation pressures of lung sample preparation and morphometric measurements of sample analyses were used. In addition, we measured the dynamic changes of MMP-2, MMP-9, and MMP-12 using Western blot and/or zymography at the protein level. Most importantly, we have observed that decreased cell proliferation and increased alveolar size first occurred in Smad3 knockout lung at postnatal day 7, an early time point in alveolarization. These changes support our conclusion that Smad3 plays an important positive role in neonatal lung alveogenesis. Therefore, we postulate herein that the emphysematous changes in Smad3-deficient mice are a consequence of abnormal postnatal alveolarization.

In summary, we have shown that loss of Smad3 gene expression attenuated early mouse lung alveolarization by reducing peripheral lung cell proliferation. Toward the end of alveolarization, Smad3 knockout lung also displayed typical and progressive centrilobular emphysema pathology, which was concomitant with increased MMP-9 protein expression and activity, as well as decreased tropoelastin expression (Fig. 7). These data suggest that not only is Smad3-mediated TGF-β signaling essential for normal alveogenesis in the postnatal lung but also that Smad3 deficiency may predispose the lung to subsequent emphysema. Currently, there is no effective therapeutic treatment for emphysema, so prevention and interruption of the disease progression at an early stage would be more effective, realistic, and economical. Our current studies suggest that TGF-β-Smad3 signaling is essential for normal postnatal lung growth and maturation, and hence, that Smad3 could function as a susceptibility gene for the occurrence of pulmonary emphysema in later life.

Fig. 7.

Speculative role of defective alveogenesis in pulmonary emphysema of Smad3-deficient mice. TGF, transforming growth factor.


This study was supported by National Heart, Lung, and Blood Institute Grants HL-68597 and HL-61286 (to W. Shi) and HL-44977, HL-44060, HL-60231, and HL-75773 (to D. Warburton), and by an American Heart Association Grant-in-Aid (to W. Shi).


We thank Dr. Dallas M. Hyde for advice on morphometric methods.


  • * H. Chen and J. Sun contributed equally to this work.

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract