SEARCH

SEARCH BY CITATION

Keywords:

  • Myf5;
  • MyoD;
  • mouse embryo;
  • respiratory muscles;
  • lung hypoplasia;
  • pneumocyte differentiation

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

In the current study, the role of contractile activity of respiratory muscles in fetal lung growth and cell differentiation was examined using Myf5−/−:MyoD−/− mouse embryos. As previously found, Myf5−/−:MyoD−/− mouse embryos had no respiratory musculature. Consequently, they suffered from pulmonary hypoplasia and died shortly after birth. The hypoplastic lung had decreased proliferation and increased apoptotic index as early as embryonic day 14.5. By contrast, only at the last gestational day, the number of lung cells expressing platelet derived growth factor B and insulin growth factor I was decreased, while the gradient of the thyroid transcription factor 1 was not maintained. Type II pneumocytes had a failure in glycogen utilization and surfactant storage and secretion but were able to synthesize the surfactant-associated proteins. Type I pneumocytes were readily detectable using an early differentiation marker (i.e., Gp38). However, the late differentiation of type I pneumocytes never occurred, as revealed by transmission electron microscopy. Together, our findings suggest that pulmonary distension due to fetal breathing-like movements plays an important role not only in lung growth but also in lung cell differentiation. Developmental Dynamics 233:772–782, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The lung primordium undergoes significant changes after its formation as two ventral buds in the cranial end of the foregut. In addition to cell proliferation and death, occurring during lung development (Scavo et al., 1998; Tseng et al., 2000; Stiles et al., 2001), pulmonary maturation is achieved by biochemical and morphological differentiation of lung cells. During development, fetal lung is subjected to both hormonal and physical factors. Physical forces are produced by the contractile activity of main respiratory muscles causing lung expansion and pulmonary cell stretch. These intermittent contractions that are homologous to adult respiratory activities are called fetal breathing-like movements (FBMs). FBMs are generated by rhythmic neuronal impulses of the respiratory center in the brainstem. They start at the embryonic day (E) 14.5 in mice and 10 weeks gestation in humans (de Vries et al., 1986; Abadie et al., 2000). Mechanical strain due to FBMs seems to play an important role in lung development. For instance, previous studies report hypoplastic lungs in fetal rabbits and sheep with transected spinal cord above the phrenic motor neurons (Wigglesworth and Desai, 1979; Liggins et al., 1981). A large number of data shows that lack of FBMs causes various abnormalities in lung organogenesis in mice (Inanlou et al., 2003; Inanlou and Kablar, 2003, 2004). It has also been shown in human embryos that in utero akinesia, including absent FBMs, leads to pulmonary hypoplasia with high morbidity and mortality rates in the neonatal period (Sandler et al., 1994). Even though the importance of FBMs in the development of the lung is clear, it is incompletely understood whether and how the mechanical stimuli are translated to meaningful instructions for the differentiation of pulmonary cells.

In mice, embryonic lung formation is composed of three main stages: pseudoglandular (E9.5–E16.5), canalicular (E16.5–E17.5), and saccular (E17.5–day 5 after birth; Ten Have-Opbroek, 1981). At the pseudoglandular stage, the primitive lung looks like a gland and is composed of tubular structures called acinar tubules (Ten Have-Opbroek, 1981). The most prominent event at the canalicular stage is the development of pulmonary blood supply and conductive airways (Laudy and Wladimiroff, 2000). The saccular stage is defined by the accelerated increase in the formation of the gas-exchanging part of the lung, and the lung is occupied by smooth-walled large airspaces named saccules (Laudy and Wladimiroff, 2000). Saccules give rise to the formation of the alveoli after birth (Ten Have-Opbroek, 1981).

To study the role of FBMs in lung organogenesis, we used knock-out mouse embryos lacking both Myf5 and MyoD (i.e., double-mutant or amyogenic embryos). Myf5 and MyoD belong to the basic-helix-loop-helix family of transcription factors. Mouse embryos lacking both Myf5 and MyoD express no skeletal muscle-specific mRNA or proteins and have no skeletal muscle fibers, including main respiratory muscles (Rudnicki et al., 1993).

Growth and maturation of the lung were analyzed at E14.5 (i.e., pseudoglandular stage and the beginning of FBMs), E16.5 (i.e., canalicular stage), and E18.5 (i.e., saccular stage). Like any other organ, normal growth of the lung relies upon regulated cell cycle. In our investigation, the influence of FBMs in the regulation of lung cell proliferation, apoptosis, and the expression of mediators that are hypothesized to control cell cycle (e.g., platelet-derived growth factor-B [PDGF-B], its receptor [PDGFR-β], insulin growth factor-I [IGF-I], and thyroid transcription factor-1 [TTF-1]) was studied. In addition, biochemical and morphological differentiation of two main pulmonary epithelial cells (i.e., pneumocytes type I and II), as an indicator of lung maturation, were also examined using immunohistochemistry and transmission electron microscopy (TEM).

Our study shows that, in the absence of FBMs, some mechanochemical signal transduction pathways are affected and both growth and maturation of the fetal lung are disturbed. Abnormal lung growth is probably caused by the decreased cell proliferation and increased cell apoptosis, followed by down-regulated expression of some mediators like PDGF-B, IGF-I, and TTF-1. Moreover, neither type I nor type II pneumocytes seem to be able to complete their differentiation program. Whereas type II pneumocytes failed in glycogen utilization and surfactant storage and secretion, they were able to synthesize the surfactant-associated proteins. Type I pneumocytes were readily detectable using an early differentiation marker (i.e., Gp38), but their late differentiation never occurred, as revealed by TEM. Taken together, our findings suggest that FBMs play an important role in lung growth as well as in the final differentiation steps of type I and II pneumocytes in vivo.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Myf5−/−:MyoD−/− Embryos Die Immediately After Birth and Have Lung Hypoplasia

As previously reported, even though double-mutant pups had pink color initially, they became cyanotic and died shortly after birth. Double-mutants did not move in response to stimulation and lacked spontaneous movements (including respiratory activities) due to the absence of skeletal musculature (Rudnicki et al., 1993). Exposing the thoracic cavity under the stereomicroscope showed that the diaphragm was intact across the abdominal wall in double-mutant pups. However, it lacked skeletal muscles and was composed of thin connective tissue (data not shown). No diaphragmatic herniation was observed. The amount of amniotic fluid appeared to be normal.

Double-mutant embryos were found to suffer from pulmonary hypoplasia. We used two criteria to determine hypoplastic lungs. First, without body weight adjustment, the ratio of wet lung weight to body weight was detected to be less than 4% in double-mutant embryos, indicating pulmonary hypoplasia (Cooney and Thurlbeck, 1985; Seegmiller et al., 1986; Table 1). Second, histological examination of the double-mutant lung showed essentially identical findings as previously reported for mdx:MyoD−/−9th and Myf5−/− embryos (Inanlou and Kablar, 2003, 2005).

Table 1. Comparison of the Body Weight and Lung Weight in the Wild-Type and Myf5−/−:MyoD−/− Embryos at Embryonic Day 18.5
GenotypeWild-typeMyf5−/−: MyoD−/−
  • *

    Statistically significant difference with Student's t-test (P < 0.05).

  • BW, body weight; LW, lung weight.

BW (mg)1213 ± 77953 ± 27
Wet LW (mg)47.7 ± 5.727.3 ± 4.1*
Dry LW (mg)6.3 ± 0.14.0 ± 0.1*
Percentage of wet LW/BW3.9 ± 0.12.8 ± 0.2*

Double-Mutant Lung Tissue Has Disturbances in Cell Proliferation and Death

To understand the underlying factors that may contribute to pulmonary hypoplasia, cell proliferation and death were examined as mechanisms known to be involved in lung growth. Compared with wild-type littermates, the proliferation index was significantly decreased and the apoptotic index was significantly increased at all stages of double-mutant lung development, in both epithelial and mesenchymal compartments (Table 2). In contrast to our previous studies, in which an increase in the apoptotic index was observable only from the canalicular stage (E16.5) of lung development (Inanlou and Kablar, 2003, 2005), the apoptotic index significantly increased in double-mutant embryos as early as E14.5, perfectly coinciding with the initiation of FBMs in control embryos (Abadie et al., 2000). These results indicate that, similarly to the in vitro data (Liu et al., 1999; Liu and Post, 2000), physical forces play an important early role in regulating pulmonary cell cycle in vivo as well.

Table 2. Comparison of PCNA- and TUNEL-Positive Cells Between Wild-Type and Myf5−/−:MyoD−/− Embryos
GenotypeCompartmentPCNA(+)TUNEL(+)□
E14.516.5E18.5E14.5E16.5E18.5□
  • PCNA, proliferating cell nuclear antigen; TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling; WT, wild-type; DM, double mutant.

  • *

    Statistically significant difference with Student's t-test (P < 0.05).

WTEpithelial55 ± 3%50 ± 5%46 ± 7%0.10 ± 0.050.12 ± 0.090.15 ± 0.06
Mesenchymal61 ± 10%57 ± 5%54 ± 3%0.15 ± 0.060.22 ± 0.090.30 ± 0.08
DMEpithelial43 ± 4%*38 ± 1%*29 ± 4%*0.33 ± 0.06*0.70 ± 0.20*1.40 ± 0.30*
Mesenchymal42 ± 3%*40 ± 3%*36 ± 3%*0.70 ± 0.10*1.30 ± 0.25*2.11 ± 0.37*

Moreover, the expression pattern of several mediators that influence cell proliferation and death was studied. These are PDGF-B, its receptor PDGFR-β, and IGF-I (Souza et al., 1994; Liu et al., 1995). The distribution pattern of PDGF-B, its receptor, and IGF-I appeared to be similar at earlier stages of lung development (i.e., E14.5 and E16; data not shown). By contrast, at term, the number of lung cells expressing the growth factors significantly decreased in double-mutant embryos, in both the epithelial and the mesenchymal compartments (Fig. 1; Table 3). These data are in accordance with our previously obtained results using either the diaphragm- or the intercostals-lacking embryos (Inanlou and Kablar, 2003, 2005).

thumbnail image

Figure 1. The expression pattern of platelet-derived growth factor-B (PDGF-B), its receptor PDGFR-β, insulin growth factor-I (IGF-I), and thyroid transcription factor-1 (TTF-1) is altered in the lung of double-mutant embryos. A–H: Compared with wild-type embryos (A,C,E,G), the representative cross-sections immunostained against PDGF-B (A,B), PDGFR-β (C,D), and IGF-I (E,F) show decreased numbers of positive cells (arrowheads in A–F), as well as a loss of TTF-1 proximal-to-distal gradient of expression (i.e., compare G, with no expression of TTF-1, to H, with many TTF-1–positive nuclei in the epithelium, as indicated by the arrowheads) in double-mutant (B,D,F,H) littermates. Original magnification, ×1,000.

Download figure to PowerPoint

Table 3. Percentage of Immunoreactive Cells Against PDGF-B, PDGFR-β and IGF-I in Wild-Type and Myf5−/−:MyoD−/− Embryos at Embryonic Day 18.5
GenotypeCompartmentPDGF-βPDGFR-βIGF-I□
  • PDGF-β, platelet-derived growth factor-beta; PDGFR-β, platelet-derived growth factor-beta receptor; IGF-I, insulin growth factor-I; WT, wild-type; DM, double mutant.

  • *

    Statistically significant difference with Student's t-test (P < 0.05).

WTEpithelial50 ± 4%49 ± 7%39 ± 3%
Mesenchymal54 ± 3%50 ± 6%42 ± 2%
DMEpithelial33 ± 4%*21 ± 2%*32 ± 1%*
Mesenchymal36 ± 4%*27 ± 4%*35 ± 3%*

The other mediator that has a role in lung organogenesis is TTF-1 (Warburton et al., 1999; Zhou et al., 2001). As previously reported for mdx:MyoD−/−9th and Myf5−/−embryos (Inanlou and Kablar, 2003, 2005), in the Myf5−/−:MyoD−/− lung, TTF-1 proximal-to-distal expression gradient was not maintained (Fig. 1).

Together, our observations determine that lack of physical forces due to the absence of FBMs is associated with not only the disturbance in normal pulmonary cell cycle but also in the expression of its regulating mediators.

Type II Pneumocyte Differentiation Is Disturbed in Double-Mutant Embryos

Biochemical and morphological criteria were used to investigate whether, in addition to lung growth, pulmonary cell differentiation is also affected in the absence of FBMs. Lung epithelium is composed of two main cell types: pneumocytes type I and type II (Mallampalli et al., 1997). Earlier studies reveal that immature type II pneumocytes have an abundant amount of cytoplasmic glycogen (Ten Have-Opbroek et al., 1988; Brandsma et al., 1993). Maturation of type II cells has been shown to be concomitant with a marked decrease in the glycogen content and a simultaneous increase in the number of lamellar bodies (Brandsma et al., 1993; Koutsourakis et al., 2001). This glycogen is suggested to be used as a substrate for the surfactant synthesis by type II cells (Bantenburg, 1992), where surfactant is defined as the complex of surfactant-associated proteins and phospholipids.

To study differentiation of type II cells, by assessing intracellular glycogen in the pulmonary epithelial cells, histochemistry was used. The sections were stained with periodic acid Schiff (PAS) at E18.5. The staining showed that, in comparison to wild-type embryos, a significantly greater number of epithelial cells in the underdeveloped respiratory system of double-mutant embryos were positive for cytoplasmic glycogen (Fig. 2; Table 4). This finding reveals that differentiation of type II cells is disturbed in the hypoplastic lungs and suggests that type II pneumocytes are unable to assemble the complex surfactant structure.

thumbnail image

Figure 2. A significantly higher number of double-mutant lung cells contain periodic acid Schiff (PAS) -positive cytoplasmic glycogen. A,B: Paraffin-embedded PAS stained cross-sections of the lung in wild-type (A) and double-mutant (B) embryos at embryonic day (E) 18.5. The number of cells containing cytoplasmic glycogen (i.e., the purple-red staining) is significantly higher in double-mutant embryos (arrowheads in B) compared with the wild-type littermates (arrowheads in A), indicating immaturity of the cells in double-mutant lung. Original magnification, ×1,000.

Download figure to PowerPoint

Table 4. Comparison of the Average Number of PAS-Positive Cells in Wild-Type and Myf5−/−:MyoD−/− Embryos at Embryonic Day 18.5
 Genotype Myf5−/−Wild-type MyoD−/−
  • PAS, periodic acid Schiff.

  • *

    Statistically significant difference with Student's t-test (P < 0.05).

Average number79 ± 13314 ± 33*

To further investigate the differentiation of type II cells, their ability to synthesize surfactant-associated proteins (SPs) was analyzed by immunohistochemistry. The four distinctly isolated SPs are SP-A, -B, -C, and -D (Whitsett and Sever, 1997). SP-C is considered to be a specific biochemical indicator of type II cell differentiation (Phelps and Floros, 1991; Wohlford-Lenane et al., 1992). Immunohistochemistry was used against SP-A, -B, -C, and -D at different stages of lung development to evaluate maturation of type II cells.

At E14.5, we could not observe any immunostaining in the lung of both control and double-mutant embryos (data not shown). At E16.5, SPs were expressed weakly on the apical borders of pulmonary epithelial cells, with no observable difference in the distribution pattern between wild-type and double-mutant embryos (data not shown). As previously suggested, the restriction of immunohistochemistry to the apical cell borders at this stage could be attributed to the higher aggregation of rough endoplasmic reticulum in the periphery of cytoplasm (Ten Have-Opbroek et al., 1988). At E18.5, the staining became even more prominent, but the expression pattern was still similar between normal and hypoplastic lungs (Fig. 3). These data are in accordance to what was observed previously in our two studies (Inanlou and Kablar, 2003, 2005). Together, our data show that type II cells differentiate to the extent of being capable of synthesizing surfactant-associated proteins normally (Van Tuyl et al., 2003) but cannot finalize their differentiation program, as revealed by their inability to use the glycogen as a substate for the surfactant synthesis (Bantenburg, 1992).

thumbnail image

Figure 3. The expression pattern of surfactant-associated protein (SP) -C is indistinguishable between wild-type and double-mutant embryos. A,B: Similar immunostaining pattern against SP-C is observable in wild-type (arrowheads in A) and double-mutant (arrowheads in B) embryos in paraffin-embedded cross-sections of the lung at embryonic day (E) 18.5, indicating the ability of type II cells to synthesize surfactant-associated protein C. Original magnification, ×400.

Download figure to PowerPoint

To further elucidate our biochemical findings, the morphology of type II cells was examined using TEM. In type II pneumocytes, the surfactant is stored in the form of cytoplasmic lamellar bodies that are composed of layers of membranes surrounding a nonlamellar proteinaceous central core (Askin and Kuhn, 1971; Gil and Reiss, 1973; Williams, 1977). The TEM analysis, followed by the morphometric analysis, showed that, similarly to the results of previous studies (Nagai et al., 1988; Brandsma et al., 1993), the average size of the nuclei was decreased and most of the cytoplasm was occupied by glycogen in type II cells of the hypoplastic lung. Moreover, we found that the number of cytoplasmic lamellar bodies per cell was lower, the lamellar bodies had different morphology, and they were smaller in size in the hypoplastic lungs (Fig. 4A,B; Table 5).

thumbnail image

Figure 4. Differentiation of type II cells is disturbed in double-mutant embryos. A–H: Electron photomicrographs of type II cells (A–D), intra-alveolar lamellar bodies (E,F), and myelin tubules (G,H) in wild-type (A,C,E,G) and double-mutant (B,D,F,H) embryos at embryonic day 18.5. A,B: The nucleus (n) is smaller (B) and the cytoplasm contains higher amount of glycogen (Gly in B) in double-mutant embryos compared with the wild-type littermates (A). A–D: In both normal and hypoplastic lungs, intracellular lamellar bodies are ovoid and surrounded by a limiting membrane (arrowheads). C,D: However, the lamellar bodies are smaller, and they are composed of less-compact membranes in double-mutant embryos (arrowheads and Table 5). E,F: After their release, the lamellar bodies seem to lose the limiting membrane and become circular (arrowheads). Compared with the regularly arranged and compact membranes in wild-type embryos (arrowhead in E), loosely organized, disarranged, and thick membranes in double-mutant embryos (arrowhead in F) indicate that secreted lamellar bodies are defected in the hypoplastic lungs. The proteinaceous nonlamellar central core seems to be similar between genotypes (short arrow in E and F). H: In addition to being loosely arranged, myelin tubules are hard to detect in double-mutant embryos (arrowhead). G: However, they are well-organized and easily observable in wild-type embryos (arrowhead). Original magnifications, ×11,000 in A,B, ×22,000 in C,D, ×25,000 in E,F, ×40,000 in G,H.

Download figure to PowerPoint

Table 5. Comparison of the Average Size of the Nuclei, Average Number and Size of Lamellar Bodies in Wild-Type and Myf5−/−:MyoD−/− Embryos at E18.5
GenotypeWild-typeMyf5−/−: MyoD−/−
  • *

    Statistically significant difference with Student's t-test (P < 0.05).

Average size of the nuclei20.4 ± 1.6 μ215.2 ± 0.7 μ2*
Number of lamellar bodies per cell7 ± 23 ± 1*
Average size of lamellar bodies1.1 ± 0.2 μ20.4 ± 0.1 μ2*

The surfactant is secreted into the pulmonary air spaces in the form of intra-alveolar lamellar bodies that have similar composition as cytoplasmic lamellar bodies (Askin and Kuhn, 1971; Gil and Reiss, 1973; Williams, 1977). The TEM analysis of intra-alveolar lamellar bodies also showed morphological differences between the control and the double-mutant lungs. Even though the proteinaceous homogenous nonlamellar central core looked alike in the normal and hypoplastic lungs, the lamellae were loose, disorganized, less compact but thicker in the double-mutant embryos (Fig. 4E,F).

After secretion, to accomplish its function, lamellar body material forms a phospholipid monolayer upon alveolar surfaces (Williams, 1977). The exact sequence of events for this rearrangement is not completely understood. However, transformed structures of lamellar bodies, called myelin tubules, are suggested to play an intermediate role (Williams, 1977). Compared with normal lungs, myelin tubules were hard to find in double-mutant embryos (Fig. 4G,H). In addition, their structure was loose and disorganized in the hypoplastic lungs (Fig. 4G,H).

Collectively, our data suggest that type II pneumocytes do not undergo final differentiation steps leading to the ability for surfactant assembly (i.e., glycogen utilization failure), storage (i.e., cytoplasmic lamellar bodies' disorganization), and secretion (i.e., intra-alveolar lamellar body and myelin figures disorganization).

Type I Pneumocyte Differentiation Is Disturbed in Double-Mutant Embryos

Pneumocytes type I are another important epithelial lung cells studied. Differentiating type I cells are reported to express a surface antigen of an unknown function, called Gp38 (Dobbs et al., 1988; Williams, 2003; Ramirez et al., 2003). Using immunohistochemistry, Gp38 protein was used as a specific marker to study early differentiation of pneumocytes type I in the normal and hypoplastic lungs. At E12.5, epithelial cells of the primitive esophagus expressed Gp38, but we could not detect any expression in the embryonic lungs (Fig. 5A,B). By E14.5, epithelium of acinar tubules was recognized to express Gp38 protein (Fig. 5C,D). With advancing gestational age, the intensity of Gp38 increased in the lung and decreased in the esophagus of the embryos. At E16.5, the epithelium of dilated acinar tubules was positive for Gp38, but no staining was recognized in the growing bronchioles (Fig. 5E,F). Epithelial cells of the respiratory system showed the highest level of staining at term. Even though at the beginning the staining was restricted to the apical cell membrane, by later stages of development it was also detectable at the lateral membrane of some cells (Fig. 5G,H). Basal membrane and cytoplasm of the cells showed no immunoreactivity. The expression pattern of Gp38 looked similar at all stages of lung development between double-mutant and wild-type embryos. The immunolabeled cells were always juxtaposed. Together, these data indicate that early specification and differentiation of type I pneumocytes are not affected by the absence of FBMs in double-mutant embryos.

thumbnail image

Figure 5. The expression pattern of Gp38 in the lung is indistinguishable between wild-type and double-mutant embryos during development. A–H: Paraffin-embedded cross-sections of the lung in wild-type (A,C,E,G) and double-mutant (B,D,F,H) embryos at embryonic day (E) 12.5 (A,B), E14.5 (C,D), E16.5 (E,F), and E18.5 (G,H) immunolabeled against Gp38. A,B: At E12.5, the immunostaining is detected in the epithelium of the primordial esophagus (arrowhead in A,B). However, no positive cell can be found in the developing lungs, indicating that biochemical differentiation of type I cells has not been started. C,D: By E14.5, the epithelium of acinar tubules in both wild-type and double-mutant embryos express Gp38 similarly (arrowheads). E,F: At 16.5, immunostaining is observed in the epithelium of dilated acinar tubules in the lung of wild-type and double-mutant embryos (arrowheads); however, the epithelium of the growing bronchioles is negative for Gp38 (short arrows). G,H: Even though by term (E18.5) lung development is arrested in the canalicular stage in double-mutant embryos, the expression pattern of Gp38 appears to be indistinguishable between normal and hypoplastic lungs (arrowheads). The staining is restricted to the apical and lateral membrane of the pulmonary epithelial cells, and no staining is detectable at the basal membranes (arrowheads in insets in G,H). Original magnifications, ×200 in A,B, ×400 in C–H, ×1,000 in insets in G,H.

Download figure to PowerPoint

To analyze later steps of functional and morphological differentiation of type I cells in double-mutant embryos, TEM was used. Well-known morphological markers of type I cells are a flattened nucleus and an extended cytoplasm containing numerous, well-defined and small cytoplasmic vesicles (Williams, 1990). In fact, in wild-type embryos, differentiating pneumocytes type I were observed lining the prospective alveoli. The cells appeared to be squamous with extended cytoplasm. We also detected the vesicles in their cytoplasm, indicating that they are differentiating pneumocytes type I (Fig. 6). Surprisingly, no similar cells were distinguished in the epithelium of hypoplastic lungs. The pulmonary epithelium in double-mutant embryos was composed of immature type II pneumocytes, as described above (i.e., containing lamellar bodies), as well as another type of cuboidal (but never flattened) cells that did not fit the description of either type II or type I pneumocytes (e.g., some cytoplasmic vesicles could be found in those cuboidal cells, but these vesicles did not seem to be well-defined; Fig. 6). Together, our data indicate that, even though immunohistochemistry confirmed the initiation of biochemical differentiation of type I cells in the lung of double-mutant embryos, the TEM studies revealed that these cells were not able to finalize their differentiation program and give rise to morphologically mature type I pneumocytes.

thumbnail image

Figure 6. Type I pneumocytes are not morphologically recognizable in double-mutant embryos. A,B: Electron photomicrographs of a differentiated type I cell in wild-type (A) and undifferentiated pulmonary alveolar cell in double-mutant (B) embryos at embryonic day 18.5. A: The flattened nucleus, extended cytoplasm, and cytoplasmic vesicles (arrowheads) characterize differentiating type I pneumocytes. Differentiation of type I cells appears to be arrested in double-mutant embryos. B: The pulmonary epithelium of the hypoplastic lungs is composed of cuboidal cells. The occasional vesicle is not well-defined in the undifferentiated cuboidal cells of double-mutant embryos (arrowhead). Original magnifications, ×11,000 in A,B.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The aim of this study was to determine whether FBMs in vivo, in addition to having a role in lung growth, can also influence lung epithelial cell differentiation. It is well established that the developing lung is affected by both biochemical and mechanical factors (Riley et al., 1990). The important effect of FBMs on lung growth due to pulmonary distension and pulmonary cell stretch has been shown in previous studies (Harding, 1991; Kitterman, 1996). However, it is not entirely certain whether, during in utero lung development, mechanical stimuli due to FBMs are translated into biochemical instructions for pulmonary cells differentiation. To address this question, we used double-mutant embryos completely lacking FBMs. Compared with wild-type littermates, double-mutant embryos were found to (1) suffer from pulmonary hypoplasia, where the lungs were arrested in the canalicular stage of development, as previously reported for other types of genetically altered mice (Tseng et al., 2000; Inanlou and Kablar, 2003, 2005); (2) have in their smaller lungs decreased cell proliferation and increased cell apoptosis, followed by down-regulated expression of some cell cycle mediators such as PDGF-B, IGF-I, and TTF-1; (3) have deflated lungs, in part due to the type II pneumocytes failure in glycogen utilization, surfactant storage, and surfactant secretion (despite the ability of type II cells to synthesize the surfactant-associated proteins); (4) have nonfunctional lungs in part due to type I pneumocytes inability to complete their differentiation program. Taken together, our findings suggest that FBMs play an important role in type I and II pneumocytes differentiation in vivo.

Even though Myf5 and MyoD are expressed in the skeletal muscles (Kablar and Rudnicki, 2000), none of them is expressed in the normal lung tissue (Inanlou and Kablar, 2003, 2005). Therefore, the absent expression of Myf5 and MyoD cannot explain the revealed pulmonary hypoplasia in double-mutant embryos. In addition, an intact diaphragm composed of connective tissues without herniation into the thorax and a sufficient amount of amniotic fluid rules out visceral herniation or oligohydramnios as the causes of pulmonary hypoplasia. Conclusively, absence of FBMs appears to be the main cause of the underdeveloped lung in double-mutant embryos.

Normal growth of any organ is dependent on the cell proliferation and death. We hypothesized that the absent FBMs may lead to disturbed pulmonary cell proliferation and death, and eventually to lung hypoplasia in double-mutant embryos. Our analyses showed decreased proliferation and increased apoptosis index in the pulmonary cells of the epithelial and mesenchymal compartments of the hypoplastic lungs, starting from the earliest stages of lung development. Disturbance in cell cycle has been reported previously as a mechanism capable of producing lung hypoplasia (Tseng et al., 2000; Inanlou and Kablar, 2003, 2005). In contrast to earlier studies showing an increased apoptosis index at the canalicular stage, here, in double-mutant embryos, the apoptosis was provoked earlier at the pseudoglandular stage. This finding can be explained by the total absence of FBMs in double-mutant embryos due to the complete lack of skeletal musculature, compared with the presence of some FBMs in the experimental embryos used in the previous investigations (i.e., the diaphragm is present in Myf5−/− embryos, whereas the intercostals are present in the mdx:MyoD−/−9th embryos; Inanlou and Kablar, 2003, 2005). It also indicates that pulmonary cell proliferation pathways are more susceptible to the insufficiency or absence of FBMs than apoptosis pathways.

PDGFs, IGF-I, and TTF-1 play a role in lung growth by modulating lung cell cycle during embryonic development (Baxter, 1988; Souza et al., 1994; Reynolds et al., 2003). Mechanical stretch has been shown to up-regulate the expression of PDGF-B and IGF-I (Liu et al., 1995, 1999; Joe et al., 1997). Similar to our previous study (Inanlou and Kablar, 2005), indistinguishable expression of PDGF-B, PDGFR-β, IGF-I, and TTF-1 at E14.5 and E16.5 could not explain disturbed pulmonary cell proliferation and death at the earlier stages of lung development, leaving the identity of the factors that might be involved elusive. However, attenuated expression of the growth factors at E18.5 might be responsible for changes in cell proliferation and apoptosis at the saccular stage of the double-mutant lung development. Whether at later stages of lung development cell cycle regulation exclusively relies on the normal expression of PDGFs, IGF-I, TTF-1, or some other factors and their combinations has yet to be determined.

Moreover, that Myf5−/− and Myf5−/−:MyoD−/− lung phenotypes are very similar (the only difference is that an increased apoptotic index is observable as early as E14.5 in Myf5−/−:MyoD−/− lung only) but their muscle phenotypes are drastically different (Myf5−/−:MyoD−/− have a complete ablation of skeletal musculature and consequently of all the growth factors normally secreted by the muscle) suggests the double-mutants' lung phenotype may be the consequence of not only the absence of FBMs but also of the absence of muscle-secreted growth factors.

In contrast to some studies (Alcorn et al., 1980; Tseng et al., 2000), our results suggest that, in addition to lung growth, differentiation of pulmonary cells is also influenced by FBMs (Nagai et al., 1988). It is established that maturation of type II pneumocytes is associated with a remarkable decrease in cytoplasmic glycogen and a simultaneous increase in the number of lamellar bodies (Chi, 1985; Ten Have-Opbroek et al., 1990), which means that the glycogen is used as a substrate in the process of surfactant assembly (i.e., surfactant consists of surfactant-associated proteins and phospholipids). Studies on the developing rat lung show that glycogen supplies substrates for the biosynthesis of surfactant phospholipids and plays a significant role in the prenatal formation of multi-lamellar bodies (Ten Have-Opbroek et al., 1990; Batenburg, 1992). Higher number of PAS-positive cells in double-mutant embryos indicated that most of the type II epithelial cells were still immature, and this conclusion was supported by TEM analysis. However, immunohistochemistry against SPs showed an indistinguishable distribution pattern in normal and hypoplastic lungs. This observation indicates that type II cells differentiate to some extent in the hypoplastic lungs, allowing for the surfactant-associated protein synthesis to occur. It may also suggest that the expression of SPs relies more on intrinsic and hormonal factors than mechanical forces, as previously suggested (Inanlou and Kablar, 2005). In fact, normal expression of epithelial differentiation markers in our previous (Inanlou and Kablar, 2003, 2005) and current experiments are in accordance with a recently performed elegant study using lung allografts under the renal capsule (Vu et al., 2003). It would be interesting to perform TEM analysis using these allografts to evaluate further steps of pneumocytes differentiation.

Higher amount of cytoplasmic glycogen in type II cells of the hypoplastic lungs in PAS-stained slides and TEM analyses indicated that the formation of cytoplasmic lamellar bodies (i.e., the storage of surfactant) might be disturbed as well. Indeed, similar to previous findings (Nagai et al., 1988; Benachi et al., 1999), we showed that the number of cytoplasmic lamellar bodies per cell was lower and they had smaller size in double-mutant embryos. Even though cytoplasmic lamellar bodies are ovoid organelles surrounded by a limiting membrane (Douglas et al., 1975), they tend to be circular after secretion, suggesting that the limiting membrane is dissolved in the alveolar fluid (Williams, 1977). Compared with wild-type embryos, loosely organized phospholipid membranes that were less compact and thicker, indicated that secreted lamellar bodies and myelin tubules were immature in double-mutant embryos (Weaver et al., 2002). Together, our data suggest that both storage and secretion of surfactant by type II pneumocytes are affected in the complete absence of FBMs.

Likewise, differentiation of type I cells appears to be affected by the lack of FBMs. In the lung, Gp38, a 40- to 42-kDa protein is exclusively expressed in pneumocytes type I (Dobbs et al., 1988). Even though the molecular function of this protein is still unknown, the expression of Gp38 is required for early differentiation of lung type I cells (Ramirez et al., 2003). By the beginning of FBMs (i.e., E14.5), a weak expression of Gp38 was detectable in the pulmonary epithelial cells of the developing lungs and the intensity of staining increased with the advancing gestational age. This finding supports the idea that, from very early stages of lung development, some of the epithelial cells are not only determined to become peripheral alveolar cells but also partially differentiated (Williams and Dobbs, 1990). Negative staining in the epithelium of the airways at term and positive staining of almost all of the epithelial cells of the acinar tubules at the earlier stages of lung development suggest that, through unknown mechanisms, the expression of Gp38 may be suppressed in the prospective epithelium of conductive system. However, expression of the protein not only persists in the differentiating type I cells but it is also increased. Nevertheless, the Gp38 expression did not appear to indicate the final morphological differentiation of pneumocytes type I, as revealed by the complete absence of morphologically identifiable type I pneumocytes in our TEM analysis. Previous in vitro studies show that mechanical distension up-regulates RTI40, a rat homologous of Gp38 (Dobbs and Gutierrez, 2001). However, the indistinguishable distribution pattern of Gp38 in the normal and hypoplastic lungs in our study suggests that FBMs do not influence Gp38 expression in vivo. This evidence shows that, in addition to Gp38, other factors play a role in the differentiation of type I cells. The lack of morphologically differentiated type I cells in double-mutant embryos shows that these cells are unable to undertake complete differentiation in the absence of FBMs. It is completely unknown what factors are influenced by the FBMs to finalize type I cell differentiation.

In fact, we are currently comparing the lung of E18.5 Myf5−/−:MyoD−/− embryos to the lung of normal control embryos using the cDNA microarray analysis. By way of this approach, it is possible to perform molecular comparisons between the mutant and the control tissues. Molecules that are not present in the mutant lung are assumed to be specific for the lacking pneumocyte types. This approach will allow identification of genes that are involved in the final differentiation steps of type I pneumocytes.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Animal Care and Interbreeding

To generate mice lacking both Myf5 and MyoD, homozygous MyoD−/− mice were bred with Myf5+/− knock-out heterozygous mice, as previously described (Rudnicki et al., 1993). The resulted heterozygous offspring for both Myf5 and MyoD (i.e., Myf5+/−: MyoD+/−) from this cross were intercrossed to obtain embryos of nine different genotypes, including Myf5−/−:MyoD−/− with a frequency of 1:16. The pregnant females were killed by cervical dislocation at the required embryonic day, and the embryos were collected by Caesarean section. The embryos were genotyped by PCR genotyping (Henneberger et al., 2000) using fetal genomic DNA. Care of animals was in accordance with the institutional guidelines.

Light Microscopy Tissue Preparation

Wild-type (four embryos per stage, n = 4) and double-mutant (three embryos per stage, n = 3) embryos were decapitated immediately after Caesarean section. Formation of the rib cage and skeletal musculature was not detected in double-mutant embryos under the stereomicroscope. The lungs were taken out after exposing the intrathoracic cavities. Lung fixation was performed in 4% paraformaldehyde for 2 hr using vacuum conditions to facilitate tissue infiltration. The lungs were then dehydrated and embedded in paraffin. Serial 4-μm-thick sections were cut with a rotary microtome for immunohistochemistry and tissue staining (i.e., hematoxylin–eosin [H&E] and PAS).

TEM Tissue Preparation

The lungs of two wild-type and two double-mutant embryos at E18.5 were prepared for electron microscopy studies. In brief, after fixation of the lung tissues in 2.5% glutaraldehyde in buffer (0.1 M sodium cacodylate PH 7.3), 1% osmium tetroxide in buffer, and 0.5% uranyl acetate (each step for 2 hr), dehydration was done by acetone. The lung tissues then were embedded in Epon araldite. The 100-nm-thick sections were cut with an ultramicrotome and were placed on copper mesh 300 grids. Staining was performed with 2% aqueous uranyl acetate and lead citrate.

Immunohistochemistry

Immunohistochemistry was performed as previously described (Inanlou and Kablar, 2003, 2005) with mouse monoclonal antibody (Dako) diluted 1:500 against proliferating cell nuclear antigen (PCNA); rabbit polyclonal antibody (Santa Cruz) diluted 1:50 against PDGF-B and PDGFR-β; goat polyclonal antibody (R&D Systems) diluted 1:50 against IGF-I; mouse monoclonal TTF-1 antibody (Neomarkers) diluted 1:50; goat polyclonal antibodies (Santa Cruz) diluted 1:100 against SP-A, -B, -C, and -D; and 8.1.1 (an anti murine-glycoprotein 38 or Gp38, also known as T1α, OTS-8, or PA2.26) hamster monoclonal antibody (Developmental Studies Hybridoma Bank) diluted 1:100 against pneumocyte type I surface antigen. To detect the apoptotic cells, terminal deoxynucleotidyl transferase mediated dUTP nick end labeling (TUNEL) method was used as previously described (Inanlou and Kablar, 2005).

Morphometry and Statistical Analysis

The epithelial and mesenchymal lung cells with the positive reaction against PCNA, TUNEL (i.e., proliferation and apoptosis index, respectively), PDGF-B, PDGFR-β, and IGF-I were counted among 1,000 randomly selected cells in the epithelial and mesenchymal compartments in serial sections at the magnification of ×400, as previously described (Inanlou and Kablar, 2003, 2005). The number of PAS-stained cells was counted under the magnification of ×400 in 10 randomly chosen microscopic fields with the area of 9 mm2 per field. The number of independent experiments was four for the control (n = 4) and three for the mutant (n = 3) embryos in all experiments. The statistical analysis was done using SPSS (version 11) software. Data were presented as means ± standard deviation (SD). Statistical significance was considered at P < 0.05 in all experiments. The values were compared by Student's t-test.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Anne C. Belliveau and Mary Anne Trevors for expert technical assistance. This work was supported by operating grants from Natural Sciences and Engineering Research Council of Canada (NSERC), The Lung Association of Nova Scotia (LANS), and infrastructure grants from Canada Foundation for Innovation (CFI) and Dalhousie Medical research Foundation (DMRF) to B.K. M.R.I. is a recipient of a Nova Scotia Health Research Foundation (NSHRF) fellowship for this project.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES