DEVELOPMENTAL DYNAMICS 212:482–494 (1998) Mesenchyme Specifies Epithelial Differentiation in Reciprocal Recombinants of Embryonic Lung and Trachea JOHN M. SHANNON,1,2* LARRY D. NIELSEN,1 SARAH A. GEBB,1 AND SCOTT H. RANDELL3 of Medicine, National Jewish Medical and Research Center, Denver, Colorado 2Division of Pulmonary Sciences and Critical Care Medicine, University of Colorado Health Sciences Center, Denver, Colorado 3Cystic Fibrosis/Pulmonary Research and Treatment Center, University of North Carolina School of Medicine, Chapel Hill, North Carolina 1Department ABSTRACT Normal lung morphogenesis and cytodifferentiation require interactions between epithelium and mesenchyme. We have previously shown that distal lung mesenchyme (LgM) is capable of reprogramming tracheal epithelium (TrE) from day 13–14 rat fetuses to branch in a lung-like pattern and express a distal lung epithelial phenotype. In the present study, we have assessed the effects of tracheal mesenchyme (TrM) on branching and cytodifferentiation of distal lung epithelium (LgE). Tracheae and distal lung tips from day 13 rat fetuses were separated into purified epithelial and mesenchymal components, then recombined as homotypic (LgM 1 LgE or TrM 1 TrE) or heterotypic (LgM 1 TrE or TrM 1 LgE) recombinants and cultured for 5 days; unseparated lung tips and tracheae served as controls. Control lung tips, LgM 1 LgE, and LgM 1 TrE recombinants all branched in an identical pattern. Epithelial cells, including those from the induced TrE, contained abundant glycogen deposits and lamellar bodies, and expressed surfactant protein C (SP-C) mRNA. Trachea controls, and both TrM 1 TrE, and TrM 1 LgE recombinants did not branch, but instead formed cysts. The epithelium contained ciliated and mucous secretory cells; importantly, no cells containing lamellar bodies were observed, nor was SP-C mRNA detected. Mucin immunostaining showed copious production of mucous in both LgE and TrE when recombined with TrM. These results demonstrate that epithelial differentiation in the recombinants appears to be wholly dependent on the type of mesenchyme used, and that the entire respiratory epithelium has significant plasticity in eventual phenotype at this stage in development. Dev. Dyn. 1998;212:482–494. r 1998 Wiley-Liss, Inc. Key words: lung; alveolar epithelium; trachea; epithelial-mesenchymal interactions; surfactant proteins INTRODUCTION Development of the rodent lung begins at midgestation as paired evaginations of the foregut endoderm into mesenchyme derived from the splanchnic mesor 1998 WILEY-LISS, INC. derm. Repetitive terminal and lateral branching of the epithelium gives rise to the pulmonary tree, a process commonly referred to as branching morphogenesis. Concomitant with patterning in the developing lung is the differentiation of several epithelial cell types, e.g., alveolar type II cells, nonciliated bronchiolar (Clara) cells, ciliated cells, goblet cells, mucous cells, that will play specific functional roles postnatally. Differentiation of these epithelial cell types within the lung occurs in a spatially precise manner, such that proximal and distal cell types arise in areas significant to their postnatal function. Interactions between epithelium and mesenchyme have been shown to be critical for the morphogenesis of many different organs (Grobstein, 1967), including the lung. Studies in both avian (Dameron, 1961) and mammalian (Masters, 1976; Sampaolo and Sampaolo, 1959; Spooner and Wessells, 1970; Taderera, 1967) species have demonstrated that an interaction of the presumptive lung epithelium with lung mesenchyme is absolutely required for normal branching morphogenesis to proceed. These epithelial-mesenchymal interactions driving lung morphogenesis and epithelial differentiation most likely involve a complex interplay of hormones, growth factors, and extracellular matrix molecules (for reviews see Cardoso, 1995; Minoo and King, 1994; Shannon and Deterding, 1997). The inductive potency of lung mesenchyme is most strikingly demonstrated by its ability to induce early tracheal epithelium that has been denuded of its own mesenchyme to branch in a lung-like pattern (Alescio and Cassini, 1962). We extended these observations by showing that lung branching morphogenesis induced in tracheal epithelium by lung mesenchyme is accompanied by a reprogramming of the tracheal epithelial cells to express an alveolar type II cell phenotype (Shannon, 1994), including the expression of the type II cellspecific marker surfactant protein C (SP-C). The specific requirement for distal lung mesenchyme to support branching morphogenesis of the pseudoglan- Grant sponsor: National Heart, Lung, and Blood Institute of the NIH; Grant numbers: HL45011 and HL58345. *Correspondence to: John M. Shannon, Ph.D., Department of Medicine, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206. E-mail: firstname.lastname@example.org Received 26 February 1998; Accepted 3 April 1998 MESENCHYMAL SPECIFICATION OF RESPIRATORY EPITHELIAL DIFFERENTIATION dular stage lung epithelium is further emphasized in experiments done by Wessells (1970), who observed that grafting tracheal mesenchyme onto distal lung epithelium that had been denuded of its own mesenchyme inhibited any further branching. Since these experiments only examined the effects of tracheal mesenchyme on lung epithelial branching, the differentiated phenotype of the epithelial cells was unknown. It is possible that although the epithelium adopts a morphogenetic pattern specified by tracheal mesenchyme, lung epithelial cytodifferentiation persists. Such a separation of tissue patterning and epithelial differentiation has been observed in recombinants of salivary gland mesenchyme and mammary gland epithelium (Sakakura et al., 1976). A second possibility is that lung epithelium associated with tracheal mesenchyme does not receive the appropriate signals to sustain lung differentiation, and therefore dedifferentiates. A third possibility is that tracheal mesenchyme produces signals specifying tracheal patterning and differentiation, and reprograms lung epithelium to express a tracheal epithelial phenotype. In the present study, we have constructed reciprocal tissue recombinants between day 13 fetal rat lung and trachea and evaluated them for markers of lung and trachea epithelial differentiation to distinguish among these possibilities. RESULTS Culture of intact distal lung tips or bisected tracheal rudiments for 5 days gave results essentially identical to those previously reported (Shannon, 1994). The cut edge of the distal lung tips healed within 24 hours, such that the epithelium formed a dilated cyst within the mesenchyme (Fig. 1B). Incipient branching was seen as early as 48 hours after explanting, was clearly established by 72 hours (Fig. 1C), and continued until harvest at 120 hours (Fig. 1D). In contrast, tracheal explants did not branch. After the cut ends of the epithelium had healed (Fig. 1F), the epithelium enlarged within the mesenchyme by 72 hours (Fig. 1G), but did not branch by 120 hours (Fig. 1H), although the epithelium appeared dilated. Epithelial patterning within the tissue recombinants was dictated by the type of mesenchyme that was used. This effect was so pronounced that lung and tracheal epithelium appeared grossly identical as soon as 24 hours after recombination with the same type of mesenchyme. When recombined with lung mesenchyme, both lung and tracheal epithelium formed a dilated cyst by 24 hours (Fig. 1J and N). Lung epithelium recombined with lung mesenchyme showed some branching by 48 hours, which was even more apparent by 72 hours (Fig. 1K). Recombination of lung mesenchyme with tracheal epithelium also induced branching; as in the homotypic lung recombinants, this was readily apparent by 72 hours (Fig. 1O). Branching of both lung and tracheal epithelium in these recombinants initially appeared to lag behind intact lung tip controls, but appeared identical by day 5 of culture (Fig. 1D, L, and P). 483 When recombined with tracheal mesenchyme, both tracheal and lung epithelium initially formed a tight mass of cells (Fig. 1R and V). These epithelia appeared essentially unchanged at 72 hours (Fig. 1S and W). On day 5 of culture, both the tracheal and lung epithelium appeared dilated, similar to what was seen in intact tracheal explants, but did not branch (Fig. 1T and X). Thus, branching of the epithelium from the distal lung tip, where active branching morphogenesis occurs, was either suppressed or uninduced when associated with tracheal mesenchyme. Sections of intact lung tips, homotypic lung mesenchyme (LgM) 1 lung epithelium (LgE) recombinants, and heterotypic LgM 1 tracheal epithelium (TrE) recombinants cultured for 5 days appeared similar by light microscopy. Branching of both lung and tracheal epithelium in response to distal lung mesenchyme resulted in a large lumen in the center of the explants and acinar structures in the explant periphery (Figs. 2A–C, 9). Cells constituting the central (proximal) area appeared uniformly columnar, while those in the peripheral (distal) region were cuboidal to cuboidal-columnar (Figs. 2A–C). In explants of intact lung tips, as well as both types of recombinants, the cells exhibited significant accumulations of glycogen, which in vivo serves as a substrate for surfactant phospholipid biosynthesis in the maturing type II cell during late gestation. Glycogen deposition was enhanced by addition of dexamethasone to the medium. The identity of these areas as glycogen was confirmed by the abolition of periodic acid-Schiff (PAS) staining by pretreating the sections with diastase (data not shown). Darkly stained inclusions were observed in many of the epithelial cells as well as the lumina (Figs. 2A–C, arrows) of the most distal acinar structures. Ultrastructural examination of cultured lung tips and both homotypic and heterotypic LgM-containing recombinants revealed that the epithelial cells in the distal acinar structures exhibited an ultrastructure typical of late gestation alveolar type II cells (Fig. 3A–C). The cells contained many osmiophilic lamellar bodies, which are the storage organelles of pulmonary surfactant. Lamellar bodies were also present in a number of the lumina, demonstrating that the type II cells not only synthesized pulmonary surfactant, but also secreted it. These ultrastructural observations are consistent with the reprogramming of tracheal epithelium to express a distal lung epithelial phenotype. In contrast to tissues containing LgM, explants and recombinants containing tracheal mesenchyme (TrM) showed a strikingly different morphology. When viewed by light microscopy, sections of intact trachea cultured for 5 days revealed a multilayered epithelium of primarily columnar cells lining a large central lumen (Fig. 4A). The basal portion of the epithelium sat on a dense stroma that contained multiple layers of mesenchymal cells. Many of the epithelial cells contained glycogen, and ciliated cells were readily observed (Fig. 4A, arrows). Homotypic TrM 1 TrE recombinants appeared 484 SHANNON ET AL. Fig. 1. Time course study of reciprocal recombinants made from isolated day 13 fetal distal lung mesenchyme (LgM), distal lung epithelium (LgE), tracheal mesenchyme (TrM), and tracheal epithelium (TrE). Intact distal lung tips (Lung) and tracheae served as controls. Arrows (I, M, Q, U) indicate epithelial rudiments surrounded by mesenchyme at the time of recombinations. All panels are shown at the same magnification (original magnification 5 563). This figure demonstrates that epithelial morphoge- netic pattern is specified by the type of mesenchyme with which it is associated. Epithelia recombined with LgM (I–L and M–P) show a pattern of branching identical to that seen in cultured intact distal lung tips (A–D). In contrast, TrM does not support branching when recombined with either tracheal (Q–T) or lung (U–X) epithelium, instead inducing the formation of an epithelial cyst identical to that seen in cultured intact trachea (E–H). MESENCHYMAL SPECIFICATION OF RESPIRATORY EPITHELIAL DIFFERENTIATION essentially identical to intact tracheal explants (Fig. 4B). Heterotypic TrM 1 LgE recombinants also appeared similar to intact tracheal explant controls, with the exception that a greater number of epithelial cells in these recombinants appeared to contain significant 485 glycogen deposits (Fig. 4C). Concentrations of cartilage precursor cells were observed in the mesenchyme of intact trachea and both homotypic and heterotypic TrM-containing recombinants by the end of the culture period (data not shown). No cartilage precursors, however, were observed in LgM 1 TrE recombinants. This suggests that LgE produces a factor(s) that reciprocally supports cartilage formation in TrM, and that this factor acts permissively. The ultrastructure of both homotypic and heterotypic recombinants made with TrM appeared essentially identical to that seen in cultured intact tracheal explants (Fig. 5A–C). A portion of the epithelial cell population consisted of columnar cells containing glycogen; these cells exhibited no apparent differentiated phenotype. A significant percentage of the epithelial cells, however, had differentiated into the readily identifiable ciliated and mucous secretory cells. The ultrastructural evidence supported the conclusion that the LgE had been reprogrammed to express a tracheal epithelial cell phenotype. This interpretation was further supported by the fact that we never observed osmiophilic lamellar bodies or tubular myelin in TrM 1 LgE recombinants, even when they were cultured in the presence of dexamethasone. The differentiated status of the epithelial cells in the cultured explants and recombinants was evaluated by their ability to express mRNAs for the surfactant proteins. Poly A1 RNA was isolated from approximately equal amounts of tissue, reverse-transcribed, and amplified using intron-spanning primers for SP-A, SP-B, SP-C and b-actin. Because of the small amount of starting tissue, no attempt was made to quantitate the amount of mRNA in each sample. The results are shown in Figure 6. Explants of lung tips showed strong signals for all three surfactant proteins (Fig. 6, lane 4), as did both homotypic (Fig. 6, lane 5) and heterotypic (Fig. 6, lane 6) recombinations made using LgM. Notable in these results is the induction of SP-C mRNA in TrE by LgM, confirming our earlier observation that LgM grafted onto a stretch of trachea denuded of its own mesenchyme resulted in expression of both SP-C mRNA and immunodetectable SP-C pro-protein (Shannon, 1994, 1996). Intact tracheal explants (Fig. 6, lane 3), as well as heterotypic recombinants (Fig. 6, lane 8) made with TrM expressed mRNAs for SP-A and SP-B. Interestingly, the level of SP-B expression in TrM 1 LgE Fig. 2. Representative light microscopy of intact day 13 lung tips (A), LgM 1 LgE recombinants (B) and LgM 1 TrE recombinants (C) that have been cultured for 5 days. Dexamethasone (1027M) was present for the final 24 hrs in this experiment. The epithelium has undergone significant branching and acini have formed under these conditions. The areas shown here are cross-sections through the most distal regions of all three tissue types. Note that the induced epithelia exhibit a similar morphology. The constituent cuboidal epithelial cells contain many darkly-stained inclusions and large amounts of pale-staining material, which was shown by histochemistry to be glycogen. Acinar lumina also contain material (arrows) that was shown by ultrastructural examination to be secreted lamellar bodies. All panels are at the same magnification (original magnification 5 5253). 486 SHANNON ET AL. Fig. 3. Electron microscopy of epithelial cells from the distal regions of intact lung tips (A), LgM 1 LgE recombinants (B), and LgM 1 TrE recombinants (C) that have been cultured as in Figure 2. Note the similarity in morphology of epithelial cells from all three tissues. Extensive areas of cytoplasmic glycogen are seen in most cells, as are numerous osmiophilic lamellar bodies. TrE recombined with LgM has assumed a morphology indistinguishable from that seen in late-gestation distal lung epithelium. Original magnifications: A 5 59503; B 5 61603; C 5 67203. recombinants appeared greater than that seen in either intact tracheal explants or TrM 1 TrE recombinants, but below that seen in explants or recombinants containing LgM. Expression of SP-C was not detected Fig. 4. Representative light microscopy of intact day 13 tracheae (A), TrM 1 TrE recombinants (B), and TrM 1 LgE recombinants (C) that have been cultured as in Figure 2. The epithelium appears similar in all three tissues. Cells containing glycogen are seen, as well as cells with apical cilia (arrows). Note the lack of inclusions in LgE when recombined with TrM. All panels are at the same magnification (original magnification 5 5253). in either intact explants (Fig. 6, lane 3) or homotypic (Fig. 6, lane 7) or heterotypic (Fig. 6, lane 8) recombinants that contained TrM, even though distal tip lung MESENCHYMAL SPECIFICATION OF RESPIRATORY EPITHELIAL DIFFERENTIATION Fig. 5. Electron microscopy of intact day 13 trachea (A), TrM 1 TrE recombinants (B), and TrM 1 LgE recombinants (C) that have been cultured as in Figure 2. The appearance of the constituent epithelial cells in all three types of tissues is essentially identical. Ciliated cells and mucous-containing cells are readily observed. Lamellar bodies are notably absent in TrM 1 LgE recombinants. Original magnifications: A 5 70003; B 5 67203; C 5 64403. epithelium is positive for SP-C mRNA at the time of isolation (Wert et al., 1993). Although we did not perform quantitative polymerase chain reaction (PCR), amplification with intron-spanning b-actin primers showed that similar amounts of mRNA were present in 487 Fig. 6. Reverse transcription polymerase chain reaction (RT-PCR) identification of surfactant protein mRNA expression in reciprocal recombinants, intact tracheae, and intact lung tips cultured for 5 days. Uncultured day 13 tracheae and lung tips served as controls. Poly A1 RNA was isolated from approximately equal amounts of tissue, treated with RNasefree DNase, reverse-transcribed using an oligo(dT) primer, and subjected to PCR using specific primers for rat surfactant proteins (SP-) A, B, and C. PCR products were electrophoresed on an agarose gel, transferred to nylon membranes by Southern blotting, then probed with 32P-labeled full-length cDNAs for SP-A, SP-B, and SP-C. Lanes: 1, uncultured day 13 tracheae; 2, uncultured day 13 lung tips; 3, intact tracheae cultured for 5 days; 4, intact lung tips cultured for 5 days; 5, LgM 1 LgE recombinants cultured for 5 days; 6, LgM 1 TrE recombinants cultured for 5 days; 7, TrM 1 TrE recombinants cultured for 5 days; 8, TrM 1 LgE recombinants cultured for 5 days. SP-A mRNA is present in all samples, although the level of expression is highest in cultured tissues that contain LgM. SP-B mRNA is not detected in day 13 trachea (lane 1) or cultured TrM 1 TrE recombinants (lane 7), and is faintly detectable in cultured intact trachea (lane 3). TrM 1 LgE recombinants (lane 8) express SP-B mRNA, although not at the level seen in cultures of either intact lung tips, LgM 1 LgE recombinants, or LgM 1 TrE recombinants (lanes 4–6). SP-C mRNA is not detectable in any tissues that contain TrM, notably TrM 1 LgE recombinants (lane 8), where SP-C expression has been extinguished. Day 13 lung tips (lane 2) are positive for SP-C mRNA expression, which increases dramatically after 5 days culture (lane 4). Both LgM 1 LgE and LgM 1 TrE recombinants are strongly positive for SP-C mRNA expression, indicating induction of a gene specific to the distal lung in tracheal epithelium. Amplification using intron-spanning b-actin primers demonstrates that there is no genomic DNA contamination of the samples. each sample, and that it was not contaminated with genomic DNA. We next examined the spatial expression of respiratory epithelial markers, using SP-C as a marker of distal lung epithelial differentiation (Glasser et al., 1991; Wert et al., 1993) and CC10, the 10-kDa Clara cell 488 SHANNON ET AL. secretory protein, as a marker of the proximal lung epithelium. Expression of SP-C mRNA in intact lung tips cultured for 5 days was confined to the most distal aspects of the explants (Fig. 7A, D). Although some small proximal epithelial tubules expressed SP-C transcripts, the most proximal epithelial cells constituting the large central lumen of the explants were invariably SP-C-negative. In contrast, these proximal epithelial cells exhibited strong CC10 mRNA expression (Fig. 7G, J). Cells positive for CC10 mRNA extended into the distal epithelium, which appeared more weakly positive than the proximal epithelium. Examination of consecutive sections, however, clearly showed that some cells in the distal epithelium expressed both SP-C and CC10 (Fig. 7D, J, arrows). The spatial patterns of expression of SP-C and CC10 mRNAs in both homotypic (Fig. 7B, E, H, K) and heterotypic (Fig. 7C, F, I, L) recombinants made with LgM were identical to those seen in cultured intact lung tips. Confirming the results obtained with reverse transcriptase (RT)-PCR, in situ hybridization demonstrated that tissues containing TrM did not express SP-C mRNA, even after prolonged exposure of the autoradiograms. This was true for cultured intact tracheal rudiments (Fig. 8A, D), as well as homotypic (Fig. 8B, E) and heterotypic (Fig. 8C, F) recombinants constructed with TrM. Thus, SP-C mRNA expression was extinguished when LgE was associated with TrM. Expression of CC10 mRNA, however, was strong in virtually all of the epithelial cells in cultured intact tracheal explants (Fig. 8G, J), and in both homotypic (Fig. 8H, K) and heterotypic (Fig. 8I, L) recombinants. Differentiation to a tracheal phenotype was further evaluated by immunostaining with monoclonal antibody RTE 11, which reacts with all rat tracheobronchial mucins. As can be seen in Figure 9, intact tracheal explants, as well as both homotypic and heterotypic recombinants made with TrM, showed strong immunoreactivity in the central lumen, indicating mucin synthesis and secretion. RTE 11 immunostaining in homotypic and heterotypic recombinants made with LgM was identical to that seen in cultured intact distal lung tips: some immunoreactivity was present in the proximal lumina of the explants, but the most distal epithelial cells and lumina were negative. DISCUSSION The data presented here demonstrate that mesenchyme not only dictates the pattern of respiratory epithelial morphogenesis, but also plays a preeminent role in specifying the eventual phenotype of the epithelium. The entire day 13 respiratory epithelium, from the trachea to the distal lung tips, has the potential to express either a proximal or distal epithelial phenotype, depending on the type of mesenchyme with which it is associated. Thus, both the tracheal and distal lung epithelial phenotypes appear to require a specific inducing factor(s) provided by tracheal and distal lung mesenchyme, respectively. This is further emphasized by our observations that purified day 13 lung (Deterd- ing and Shannon, 1995) and day 13 tracheal (Shannon, 1996) epithelium cease proliferation and differentiation when cultured in the absence of mesenchyme. We and others (Alescio and Cassini, 1962; Wessells, 1970) have demonstrated that mesenchyme isolated from the distal tips of glandular stage lung can induce tracheal epithelium of the same age to branch in a lung-like pattern. The present and a previous (Shannon, 1994) study has extended these observations to demonstrate that these changes in branching morphogenesis are accompanied by changes in epithelial differentiation. Under the influence of distal lung mesenchyme, tracheal epithelium was reprogrammed to express both morphological and biochemical markers of alveolar type II cell differentiation. Notable among these are the presence of cytoplasmic and luminal osmiophilic lamellar bodies, which are the storage organelles for pulmonary surfactant, and the expression SP-C mRNA. Furthermore, SP-C expression was restricted to the distal epithelium, as occurs in vivo, indicating that parameters dictating spatial organization in the developing lung were conserved in the recombinants. Because the tracheal epithelium has been induced to express a new phenotype, we consider this to be an instructive interaction. The potential for tracheal epithelium to be reprogrammed by lung mesenchyme does not persist indefinitely, however, since we have observed that day 16 tracheal epithelium was not competent to respond to the inductive influence of day 13 lung mesenchyme (Shannon, 1994). The mechanism by which this restriction occurs is unknown, although our present results would suggest it occurs via signals from the mesenchyme. Our present results show for the first time that tracheal mesenchyme is able to instructively reprogram distal lung epithelium to express a tracheal epithelial cytodifferentiation. Wessells (1970) observed that tracheal mesenchyme grafted onto the distal tip epithelium of glandular stage mouse lungs from which lung mesenchyme had been removed caused cessation of epithelial branching. These results suggested that branching was not a property intrinsic to the lung epithelium, but instead required specific interaction with lung mesenchyme. Our data confirm these observations on the effect of tracheal mesenchyme on branching morphogenesis, and demonstrate further that the distal lung epithelial cells have undergone a change in differentiated phenotype, as indicated by the morphological differentiation of ciliated cells and mucous secretory cells, which are not seen in the distal lung epithelium. Our observations indicate that LgM and TrM are fundamentally different in their ability to induce epithelial morphogenesis and differentiation, even though both share the same general spatial origin within the embryo. Unequivocal characterization of the tracheal epithelial phenotype is hindered by a lack of definitive markers of differentiation among the different epithelial cell types. TrM 1 LgE recombinants did, however, produce immunodetectable tracheobronchial mucins. MESENCHYMAL SPECIFICATION OF RESPIRATORY EPITHELIAL DIFFERENTIATION 489 Fig. 7. In situ hybridization of cultured intact lung tips, LgM 1 LgE recombinants, and LgM 1 TrE recombinants. Consecutive sections were hybridized with 33P-labeled antisense probes for SP-C (A–F) and the 10 kDa Clara cell secretory protein, CC10 (G–L). Exposure times of the autoradiographs are identical for each probe. Darkfield images were photographed and printed identically. Cultured intact lung tips (A,D) and LgM 1 LgE recombinants (B,E) express SP-C mRNA only in the most distal epithelial cells. CC10 mRNA is expressed throughout the epithe- lium, although the signal appears stronger in the proximal epithelium. LgM 1 TrE recombinants exhibit patterns of expression of both SP-C (C,F) and CC10 mRNAs (I,L) in the branched epithelium that are identical to those seen in tissues containing LgE. Some cells in the distal epithelium appear to express both SP-C and CC10 mRNAs (arrows). Control preparations using radiolabeled sense probes show no hybridization signal (not shown). All panels are at the same magnification (original magnification 5 1323). Although mucins are not completely specific markers of tracheal differentiation, the distribution of strongly immunopositive cells and the copious amount of mucin present in the lumina were essentially identical among intact tracheal explants and TrM-containing homotypic and heterotypic recombinants. That the changes we observed in TrM 1 LgE recombinants represent a change in the developmental fate of lung epithelium is further supported by the lack of lamellar bodies and the extinguishing of SP-C mRNA expression. The mechanism by which SP-C mRNA is suppressed in the presence of TrM is unknown. Tracheal mesenchyme may produce some factor(s) that actively inhibits SP-C expression. An alternative and equally plausible possi- 490 SHANNON ET AL. Fig. 8. In situ hybridization of cultured intact tracheae, TrM 1 TrE recombinants, and TrM 1 LgM recombinants. Consecutive sections were hybridized with antisense probes for SP-C (A–F) and CC10 (G–L). Exposure of the autoradiographs was identical for each probe, and samples were photographed and printed identically. SP-C mRNA is not detectable in any epithelium (arrows) that is associated with TrM, even though LgE is SP-C mRNA positive at the time of isolation. In contrast, CC10 mRNA is strongly expressed throughout the epithelium in all TrM-containing tissues. Control preparations using radiolabeled sense probes show no hybridization signal (not shown). All panels are at the same magnification (original magnification 5 1323). bility is that tracheal mesenchyme does not produce a factor(s) whose continuous presence is required for sustaining distal lung epithelial differentiation. Although our data demonstrate that several key characteristics of distal lung epithelium have been altered by tracheal mesenchyme, the apparent increased amounts of glycogen and SP-B mRNA in TrM 1 LgE versus TrM 1 TrE recombinants suggest the possibility of an intermediate phenotype, which might occur if the lung epithelium is already partially restricted to a lung developmental fate on day 13. As a whole, however, our data support the concept that tracheal mesenchyme induces lung epithelium to initiate a new program of gene expression resulting in tracheal differentiation. The molecular nature of the inductive cues provided by mesenchyme are undoubtedly complex, as evidenced by observations documenting the roles of growth factors, hormones, and extracellular matrix molecules in MESENCHYMAL SPECIFICATION OF RESPIRATORY EPITHELIAL DIFFERENTIATION 491 Fig. 9. Mucin immunostaining of cultured tissues. Intact tracheae (A), distal lung tips (B), TrM 1 TrE recombinants (C), LgM 1 LgE recombinants (D), TrM 1 LgE recombinants (E), and LgM 1 TrE recombinants (F) were cultured for 5 days, fixed, and processed for immunostaining with monoclonal antibody RTE 11, which reacts against mucin. LgMcontaining tissues exhibit some positive luminal reactivity that is confined to the proximal portion of the explant. Note that the most distal epithelial acini are negative. Cultured intact trachea, as well as both types of TrM-containing recombinants, exhibit strong mucin reactivity in epithelial cells and the lumina. All panels are at the same magnification (original magnification 5 2653). specifying respiratory epithelial phenotype (Minoo and King, 1994; Hilfer, 1996; Shannon and Deterding, 1997). Observations from experiments in which epithelial and mesenchymal rudiments were cultured with a filter interposed between them demonstrate that the active factor(s) are diffusible over a short distance (Taderera, 1967; Shannon, 1996). Although a number of growth factors have been shown to be present in the developing lung, recent experiments suggest that the fibroblast growth factor (FGF) family may play a key role in mediating epithelial-mesenchymal interactions in the developing lung. FGF1, FGF7, and FGF10 have all been shown to stimulate proliferation and differentiation of embryonic rat or mouse lung epithelium cultured in the absence of mesenchyme (Deterding and Shannon, 1995; Nogawa and Ito, 1995; Deterding et al., 1996; Cardoso et al., 1997; Bellusci et al., 1997). Antisense oligonucleotides to FGF7 have been shown to inhibit branching morphogenesis in embryonic rat lung explants (Post et al., 1997). Furthermore, when a dominant negative form of the FGF7 receptor is targeted to the lung with the SP-C promoter, lung branch- 492 SHANNON ET AL. ing does not occur at all (Peters et al., 1994). These observations, combined with the fact that FGF7 (Mason et al., 1994; Finch et al., 1995; Post et al., 1997) and FGF10 (Bellusci et al., 1997) are produced in the mesenchymal cells subtending the lung epithelium, suggest a primary role for these growth factors in mediating epithelial-mesenchymal interactions in lung development. It should be noted, however, that transgenic mice null for FGF7 expression exhibit no abnormalities in lung development (Guo et al., 1996). This suggests other members of the FGF family, perhaps FGF10, can serve as functionally redundant molecules. In summary, we have demonstrated that epithelium from the entire respiratory tree of the early pseudoglandular stage rat remains competent to respond to induction by heterotypic respiratory mesenchyme. Precisely when and, more importantly, how these epithelia become restricted in their developmental fates has yet to be elucidated. We believe that the ability to reprogram both tracheal and lung epithelium to express new phenotypes will prove useful in determining the factors involved in both the induction and restriction of respiratory tissues. EXPERIMENTAL PROCEDURES Animals Timed-pregnant Sprague-Dawley rats were obtained from Charles Rivers Laboratories (Raleigh, NC); the day on which a sperm-positive vaginal plug was detected was considered day 0 of gestation. Pregnant dams were sacrificed on day 13 of gestation, then the fetuses were removed and weighed. Fetuses weighing 80–90 mg, whose lungs are at the pseudoglandular stage of development, were used in these experiments. Tissue Separation, Recombination, and Culture Fetuses were decapitated and transferred to a Maximov depression slide containing ice-cold Hank’s balanced salt solution (HBSS; GIBCO/BRL, Gaithersburg, MD) supplemented with 100 U/ml penicillin, 100 µg/ml streptomycin, 2.5 µg/ml amphotercin B (all from GIBCO/ BRL), and 10 µg/ml gentamicin sulfate (Sigma Chemical Co., St. Louis, MO). The lungs and trachea were dissected intact using Moria microsurgery knives (Fine Science Tools Inc., Foster City, CA). Separation of lung epithelium and mesenchyme was performed as previously described (Shannon, 1994; Deterding and Shannon, 1995). Briefly, distal lung tips were cut off, then incubated with dispase (Collaborative Biomedical Products, Bedford, MA) for 30 minutes at 37°C. The epithelium and mesenchyme were separated using tungsten needles (Fine Science Tools Inc.), washed four times with HBSS containing 10% fetal bovine serum (FBS; GIBCO/BRL) to remove any residual enzyme. Mesenchymal rudiments were transferred to Waymouths 752/1 medium (GIBCO) containing 10% FBS (hereafter culture medium) and stored on ice until recombination. Epithelial rudiments were given a second incubation in dispase for 10 minutes to remove any possible adherent mesenchymal cells, washed as above, and stored on ice in culture medium; this method has been shown to effectively eliminate contamination by mesenchymal cells (Deterding and Shannon, 1995). The trachea was removed from the lung/trachea complex just above the bifurcation and bisected. Epithelial and mesenchymal tracheal rudiments were isolated by incubating the tracheal halves in HBSS containing 0.05% collagenase (CLS IV; Worthington Biochemicals, Freehold, NJ) and 1% FBS for 45 minutes at 37°C. The tracheae were washed twice with HBSS 1 10% FBS, treated briefly with 1–2 drops of a 1-mg/ml solution of DNase I (Sigma) in HBSS, and separated into epithelial and mesenchymal components with tungsten needles. Mesenchymal rudiments were washed twice with HBSS 1 FBS, then placed on ice in culture medium. Epithelial rudiments were treated with dispase for 10 minutes, washed, and placed on ice in culture medium. Epithelial and mesenchymal tissues were recombined on the surface of a semisolid medium consisting of 0.5% agarose (Sigma) in culture medium. The appropriate epithelial and mesenchymal rudiments for each experimental condition were transferred to 35-mm dishes containing 2 ml of semisolid medium and excess liquid medium removed with a flame-drawn Pasteur pipette. Epithelial and mesenchymal rudiments were teased into close apposition using microsurgery knives. Typically 3–4 lung mesenchymes, or 2–3 tracheal mesenchymes, were used per epithelial rudiment in each recombinant. After the positioning of the epithelial and mesenchymal rudiments, a moat was created in the agarose around each recombinant using a flame-drawn Pasteur pipette; this prevented liquid from flooding the rudiments and causing them to drift apart during the initial culture period. Excess fluid was then removed from around the recombinants and the tissues repositioned, if necessary. After culture overnight, the recombinants were transferred to 35-mm dishes containing fresh semisolid medium. A few drops of culture medium was added to each dish to maintain the recombinants at an air-liquid interface. Cultures were maintained for a total of 5 days, whereupon they were processed as described below. Four different recombinants were constructed: TrM 1 TrE; TrM 1 LgE; LgM 1 LgE; and LgM 1 TrE. Intact lung tips and bisected tracheae were cultured as controls. The ability of glucocorticoids to accelerate maturation of the surfactant system in the developing distal lung epithelium has long been appreciated (for review see Odom and Ballard, 1997). Therefore in some experiments dexamethasone (1027M; Sigma) was added to the culture medium of all groups for the final 24 hours to maximize the possibility of distal epithelial maturation within the 5-day culture period. This allowed us to better assess the ability of LgM and TrM to induce and suppress the distal lung epithelial phenotype, respectively. MESENCHYMAL SPECIFICATION OF RESPIRATORY EPITHELIAL DIFFERENTIATION Isolation of mRNA and Reverse Transcription Polyadenylated RNA was isolated from recombinants and control cultures using a Micro-Fast Track kit (Invitrogen, San Diego, CA). Approximately equal amounts of tissue from each condition were used for each isolation. The mRNA was treated with 35 U of RNase-free DNase (Boehringer Mannheim, Indianapolis, IN) in the presence of 40 U of RNasin (Promega Biotech, Madison, WI), then ethanol precipitated with 5 µg mussel glycogen present as a carrier. The mRNA was reverse-transcribed in a final volume of 20 µl using a cDNA Cycle kit (Invitrogen) and immediately used for PCR. PCR Amplification and Analysis of Products Oligonucleotide primers were synthesized on an Applied Biosystems Model 381A DNA synthesizer (Foster City, CA) in the Molecular Resource Center at the National Jewish Medical and Research Center. The primers were designed from the published sequences for rat SP-A (Fisher et al., 1988), rat SP-B (Emrie et al., 1989), and rat SP-C (Fisher et al., 1989). The sequences for the SP-A primers were (1) 58-AGTCCTCAGCTTGCAAGGATC-38 coding sense and corresponding to bases 422–442, and (2) 58-CGTTCTCCTCAGGAGTCCTCG-38 coding antisense and corresponding to bases 549–569; the predicted size of the SP-A product was 148 base pairs (bp). The sequences of the SP-B primers were (1) 58-GAGCAGTTTGTGGAACAGCAC-38 coding sense and corresponding to bases 997–1017, and (2) 5-TGGTCCTTTGGTACAGGTTGC-38 coding antisense and corresponding to bases 1152–1172; the predicted SP-B product size was 176 bp. The sequences of the SP-C primers were (1) 58-CATACTGAGATGGTCCTTGAG-38 coding sense and corresponding to bases 202–222, and (2) 58-TCTGGAGCCATCTTCATGATG-38 coding antisense and corresponding to bases 381–401; the predicted SP-C product size was 200 bp. The sequences for the b-actin primers were (1) 58-GTATGGAATCCTGTGGCATCC-38 coding sense and corresponding to bases 2 683–2703 of the genomic DNA sequence, and (2) 58TACGCAGCTCAGTAACAGTCC-38 coding antisense and corresponding to bases 3135–3155 of the genomic DNA sequence. The predicted size of the b-actin product amplified from cDNA was 349 bp, whereas the size of the product amplified from contaminating genomic DNA was 473 bp. Four microliters of each reverse-transcribed cDNA were used for amplification of SP-A, SP-B, and SP-C, and 2 µl for amplification of b-actin. The final 50 µl reaction mixture contained 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl2, 200 µM each of dATP, dCTP, dGTP, and dTTP, 0.01% gelatin, 0.5 µM of the appropriate upstream and downstream primers, and 2.5 U AmpiTaq DNA polymerase (Perkin-Elmer Cetus, Norwalk, CT). Reaction mixtures were overlaid with silicone oil and amplified for 30 cycles. The amplification profile consisted of denaturation at 94°C for 1 minute, annealing at 58°C for 2 minutes, and extension 493 at 72°C for 3 minutes; the reaction was terminated by cooling to 4°C. After amplification the reaction products were electrophoresed through a 1% agarose (Sigma) gel, then transferred to Nytran (Schleicher and Schuell, Keene, NH) for Southern blot analysis by standard methods. Southern blots were probed with full length cDNAs for SP-A, SP-B, and SP-C that were labeled to high specific activity with 32P-dCTP (Amersham, Arlington Heights, IL) using a Redi-prime kit (Amersham). In Situ Hybridization Tissues and recombinants for in situ hybridization were enrobed in a solution of neutralized rat tail collagen (Emerman and Pitelka, 1977) prior to fixation. Samples were transferred in a small volume of medium to a piece of Parafilm (American National Can, Neena, WI), then covered with 200–300 µl of cold collagen solution. The explants were mixed into the collagen and centered at the bottom of the drop. After gelation at 37°C oven for 15 minutes, the enrobed tissues were inverted into a 35-mm dish containing 2 ml of 4% paraformaldehyde in RNase-free phosphate-buffered saline (PBS; pH 7.4) and fixed overnight at 4°C. After fixation, the samples were washed once in PBS, dehydrated, and embedded in paraffin. The enrobement procedure not only expedited locating the tissues when sectioning, but also allowed several samples to be evaluated in one section. In situ hybridization was performed on 4-µm sections as previously described (Deterding and Shannon, 1995), with the exception that 33P-UTP (2,000–4,000 Ci/mmol; NEN Life Science Products, Boston, MA) was used in the transcription of RNA probes. Slides were dipped in NTB-2 nuclear track emulsion (Eastman Kodak Co, Rochester, NY) and developed after an exposure period appropriate for each probe. Hybridization with radiolabeled sense RNA probes was done as a control in all experiments. As a marker of the distal lung epithelial phenotype, we examined the expression of SP-C, which was transcribed from a full length rat cDNA (Fisher et al., 1989). As a marker of the proximal respiratory epithelial cell phenotype, we examined the expression of the 10-kDa Clara cell secretory protein (CC10) using a 275-bp fragment of the rat cDNA (generously provided by Dr. Arun Rishi). Electron Microscopy Samples were fixed in 2% glutaraldehyde-1% paraformaldehyde, postfixed in 1.5% osmium tetroxide, stained en bloc with uranyl acetate, and embedded in Lufts 3:7 mixture of LX-112 (Ladd Research Industries, Burlington, VT). Sections 1- to 1.5-µm-thick were stained with Mallory’s azure II-methylene blue. Ultrathin sections were stained with uranyl acetate and lead citrate and examined in a Phillips 400 electron microscope. 494 SHANNON ET AL. Immunocytochemistry and Histochemistry Samples for immunocytochemistry were fixed with Omni-Fix II and embedded in paraffin. As a marker of the tracheal and proximal airway phenotype, sections were stained as previously described with the mouse monoclonal antibody RTE 11, which reacts with mucin (Randell et al., 1993). The presence of glycogen was detected in paraformaldehyde fixed tissues by comparing PAS staining in sequential sections that were or were not pretreated with diastase (Sheehan and Hrapchak, 1980). ACKNOWLEDGMENTS This work was performed in the Lord and Taylor Laboratory for Lung Biochemistry and the Anna Perahia Adatto Clinical Research Center. The authors gratefully acknowledge Lynn Cunningham for performing the histology and histochemistry, Janet Leiber for performing the electron microscopy, and Leigh Landskroner and Barry Silverstein for assistance in preparing the illustrations. We also thank Kathy Ryan Morgan for excellent assistance in preparing the manuscript. This work was supported by the National Heart, Lung, and Blood Institute of the NIH, grants HL45011 (J.M.S.) and HL58345 (S.H.R.). REFERENCES Alescio T, Cassini A. Induction in vitro of tracheal buds by pulmonary mesenchyme grafted on tracheal epithelium. J. Exp. Zool. 1962;150: 83–94. Bellusci S, Grindley J, Emoto H, Itoh N, Hogan BL. Fibroblast growth factor 10 (FGF10) and branching morphogenesis in the embryonic mouse lung. Development 1997;124:4867–4878. Cardoso WV. Transcription factors and pattern formation in the developing lung. Am. J. Physiol. 1995;269:L429–L442. Cardoso WV, Itoh A, Nogawa H, Mason I, Brody JS. FGF-1 and FGF-7 induce distinct patterns of growth and differentiation in embryonic lung epithelium. Dev. Dyn. 1997;208:398–405. Dameron F. Etude de la morphogenese de la bronche de l’embryon de Poulet associee a differents mesenchymes en culture in vitro. Comput. Rend. Acad. Sci. 1961;262:1642–1645. Deterding RR, Shannon JM. Proliferation and differentiation of fetal rat pulmonary epithelium in the absence of mesenchyme. J. Clin. Invest. 1995;95:2963–2972. Deterding RR, Jacoby CR, Shannon JM. Acidic fibroblast growth factor and keratinocyte growth factor stimulate fetal rat pulmonary epithelial growth. Am. J. Physiol. 1996;271:L495–L505. Emerman JT, Pitelka DR. Maintenance and induction of morphological differentiation in dissociated mammary epithelium on floating collagen membranes. In Vitro 1977;13:316–328. Emrie PA, Shannon JM, Mason RJ, Fisher JH. cDNA and deduced amino acid sequence for the rat hydrophobic pulmonary surfactantassociated protein, SP-B. Biochim. Biophys. Acta. 1989;994:215– 221. Finch PW, Cunha GR, Rubin JS, Wong J, Ron D. Pattern of keratinocyte growth factor and keratinocyte growth factor receptor expression during mouse fetal development suggests a role in mediating morphogenetic mesenchymal-epithelial interactions. Dev. Dyn. 1995; 203:223–240. Fisher JH, Emrie PA, Shannon JM, Sano K, Hattler B, Mason RJ. Rat pulmonary surfactant protein A is expressed as two differently sized mRNA species which arise from differential polyadenylation of one transcript. Biochim. Biophys. Acta. 1988;950:338–345. Fisher JH, Shannon JM, Hoffman T, Mason RJ. Nucleotide and deduced amino acid sequence of the hydrophobic surfactant protein SP-C from the rat: Expression in alveolar type II cells and homology with SP-C from other species. Biochim. Biophys. Acta. 1989;995:225– 230. Glasser SW, Korfhagen TR, Wert SE, Bruno MD, McWilliams KM, Vorbroker DK, Whitsett JA. Genetic element from human surfactant protein SP-C gene confers bronchiolar-alveolar cell specificity in transgenic mice. Am. J. Physiol. 1991;261:L349–L356. Grobstein C. Mechanisms of organogenetic tissue interaction. Nat. Cancer. Inst. Monogr. 1967;26:279–299. Guo L, Degentstein L, Fuchs E. Keratinocyte growth factor is required for hair development but not for wound healing. Genes Dev. 1996;10:165–175. Hilfer SR. Morphogenesis of the lung: Control of embryonic and fetal branching. Annu. Rev. Physiol. 1996;58:93–113. Mason IJ, Fuller-Pace F, Smith R, Dickson C. FGF-7 (keratinocyte growth factor) expression during mouse development suggests roles in myogenesis, forebrain regionalisation and epithelial-mesenchymal interactions. Mech. Dev. 1994;45:15–30. Masters JRW. Epithelial-mesenchymal interaction during lung development: The effect of mesenchymal mass. Dev. Biol. 1976;51:98–108. Minoo P, King RJ. Epithelial-mesenchymal interactions in lung development. Annu. Rev. Physiol. 1994;56:13–45. Nogawa H, Ito T. Branching morphogenesis of embryonic mouse lung epithelium in mesenchyme-free culture. Development 1995;121: 1015–1022. Odom MW, Ballard PL. Developmental and hormonal regulation of the surfactant-system. In: McDonald JA ed. Lung Growth and Development, Vol. 100. New York: Marcel Dekker, 1997;495–575. Peters K, Werner S, Liao X, Wert S, Whitsett JA, Williams L. Targeted expression of a dominant negative FGF receptor blocks branching morphogenesis and epithelial differentiation of the mouse lung. EMBO J. 1994;13:3296–3301. Post M, Souza P, Liu J, Tseu I, Wang J, Kuliszewski M, Tanswell K. Keratinocyte growth factor and its receptor are involved in regulating early lung branching. Development 1996;122:3107–3115. Randell SH, Shimizu T, Bakewell W, Ramaekers FCS, Nettesheim P. Phenotypic marker expression during fetal and neonatal differentiation of rat tracheal epithelial cells. Am. J. Respir. Cell. Mol. Biol. 1993;8:546–555. Sakakura T, Nishizuka Y, Dawe CJ. Mesenchyme-dependent morphogenesis and epithelium-specific cytodifferention in mouse mammary gland. Science 1976;194:1439–1441. Sampaolo G, Sampaolo L. Observations histologiques sur le poumon de foetus de Cobaye, cultive in vitro. Comput. Rend. Assoc. Anat. 1959;45:707–714. Shannon JM. Induction of alveolar type II cell differentiation in fetal tracheal epithelium by grafted distal lung mesenchyme. Dev. Biol. 1994;166:600–614. Shannon JM. Epithelial-mesenchymal interactions in lung development. J. Jpn. Med. Soc. Biol. 1996;26:47–65. Shannon JM, Deterding RR. Epithelial-mesenchymal interactions in lung development. In: McDonald JA, ed. Lung Growth and Development, Vol. 100. New York: Marcel Dekker, 1997;81–118. Sheehan DC, Hrapchak BB. Theory and Practice of Histotechnology. St. Louis: Mosby, 1980. Spooner BS, Wessells NK. Mammalian lung development: Interactions in primordium formation and bronchial morphogenesis. J. Exp. Zool. 1970;175:445–454. Taderera JT. Control of lung differentiation in vitro. Dev. Biol. 1967; 16:489–512. Wert SE, Glasser SW, Korfhagen TR, Whitsett JA. Transcriptional elements from the human SP-C gene direct expression in the primordial respiratory epithelium of transgenic mice. Dev. Biol. 1993;156:426–443. Wessells NK. Mammalian lung development: Interactions in formation and morphogenesis of tracheal buds. J. Exp. Zool. 1970;175:455– 466.