Main

The proliferation of fibroblast, myofibroblast, and smooth muscle cell precursors is pathognomonic of bronchopulmonary dysplasia (1), a syndrome of high frequency, morbidity, and mortality among premature infants (2). These differentiated cells arise from the pulmonary mesenchyme, and arresting the proliferation of fetal lung mesenchymal cells may reduce the morbidity and mortality of this disorder. Fetal lung mesenchymal cell mitogenesis is regulated by peptide growth factors that bind to cognate receptors. These receptors initiate a series of intracellular signaling events converging on Ras activation of the MAP kinase cascade. Because inflammatory cell signaling induces mesenchymal cell proliferation in the preterm but not the mature lung (3), we postulate that receptor signaling pathways are developmentally modulated through differential expression of signaling intermediates.

The Shc proteins are possible candidates for developmental mitogenic regulation. Although best characterized as substrates of EGF receptor tyrosine kinases (4), Shc proteins are tyrosine phosphorylated by most classes of cell-surface receptors, including G-protein coupled receptors (5,6), T-cell receptors (7), B-cell receptors (8), and integrins (9). Shc proteins thus funnel a wide array of signaling processes into the Ras protooncogene, which then initiates the MAP kinase amplification cascade. Shc proteins are also constitutively phosphorylated in tumor cell lines (10), consistent with their putative function as intermediates of mitogenic signaling.

The Shc proteins are expressed as three known variants with differential effects on Ras function. Shc 52 is a ubiquitously expressed and well-characterized facilitator of Ras activation. Shc 52 contains two phosphotyrosine binding sites: an SH2 domain at the carboxyl terminus (11), and a unique domain (termed the PTB domain) at the amino terminus (12). Upon binding to receptor tyrosine kinases, Shc 52 is activated by phosphorylation of tyrosine residues in the central proline-rich collagen homology domain (13). Two of these tyrosine residues become binding sites for Grb2 adapter proteins (8) which then bind Sos, a cytoplasmic GTP exchange protein (8). The resulting complex translocates to the plasma membrane where Ras is encountered (14) and activated. Ras then activates the Erk1/Erk2 MAP kinase cascade (15).

In contrast, Shc 66 has been characterized as a dominant negative form of Shc 52 (16). Shc 66 is an alternative transcription product and differs from Shc 52 only in the addition of 109 amino acids to the amino terminus (17). Although both isoforms bind Grb2 upon phosphorylation by receptor tyrosine kinases (16,18,19), Shc 66 does not induce Ras activation, MAP kinase activation, c-fos activation, or transformation of NIH-3T3 cells (18). Shc 66 accelerates MAP kinase inactivation after EGF stimulation and may compete with Shc 52 for a limited pool of Grb2 (16).

Shc 46 differs from Shc 52 only in the deletion of 45 amino acids at the amino terminus (11), and is an alternative translation product of the Shc 52 mRNA (11). Expression of Shc 52 and Shc 46 vary across cell types in a reciprocal fashion (18), and some proteins have been shown to bind Shc 66 and Shc 52 but not Shc 46 (20). Neither isoform-specific regulation nor isoform-specific function of Shc 46 have been well defined.

Because Shc 66 and Shc 52 possess opposing mitogenic signaling functions, it is likely that Ras activation by Shc signaling is regulated by the relative expression of these isoforms. Shc is an intermediate in many signaling pathways implicated in mammalian morphogenesis, including insulin (21), nerve growth factor (22), fibroblast growth factor (23), and EGF (24). Changes in Shc isoform expression would modulate Ras activation by the entire repertoire of signaling processes acting through Shc, thus regulating growth factor mitogenesis in an essentially global fashion. We hypothesized that the expression of Shc isoforms are temporally and spatially restricted during mouse lung morphogenesis. To address this postulate, we quantitated Shc protein and mRNA expression during mouse lung morphogenesis and assessed Shc 66 localization during gestation. We found that Shc 66 is transcriptionally downregulated during the peripartum period. Within the fetal lung, Shc 66 is strongly localized to epithelial and mesenchymal cells contacting the basement membrane of large airways.

We then sought to demonstrate that Shc 66 participates in fetal lung mesenchymal cell signaling by demonstrating its phosphorylation in response to EGF stimulation. Primary fetal mouse lung mesenchymal cells were isolated by tryptic digestion and differential adhesion. Cultured primary cells were then exposed to EGF for up to 15 min and assessed for tyrosine-phosphorylated Shc isoform content. We found that all three Shc isoforms are tyrosine phosphorylated by EGF stimulation, suggesting that Shc 66 likewise participates in growth factor signaling. This finding, together with previous characterization of Shc 66 as a dominant negative form of Shc 52, suggests that growth factor mitogenic signaling is modulated during lung development through differential expression of Shc 66.

METHODS

Western analysis. Fetuses were dissected from timed-gestation Swiss-Webster mice at 12 to 22 d postconceptual age (term = 18.5 d) and the lungs were isolated as previously described (25). Specimens from three litters were obtained at each time point; for d 12 through 14, multiple lungs from the same litter were pooled for each sample. Throughout this report, specimens obtained on the last day of gestation before parturition are designated as 18 d postconceptual age, whereas tissues obtained immediately after delivery are assigned a postconceptual age of 18.5 d. The lungs were homogenized using sonication and stored at -70°C. Homogenates were cleared by centrifugation at 14 000 × g for 20 min, protein concentrations determined by the bicinchonoic acid method, and the protein contents equalized. Samples were subjected to 7.5% SDS-PAGE, transferred to Immobilon-P (Millipore, Bedford, MA), blocked overnight, and probed with a rabbit antipan Shc polyclonal antibody (Transduction Laboratories, Lexington, KY). Specific binding was detected by chemiluminescence substrate (Pierce, Rockford, IL) using autoradiography film and quantitated by densitometry.

The anti-pan Shc antibody recognizes the C-terminal domain of human Shc 52/46 from amino acid 359 to 473, and detects three major proteins on Western analysis. This antibody was used to screen a λgt-11 15-d fetal mouse cDNA expression library (Clontech, Palo Alto, CA). Seven clones were isolated and partially sequenced. Two were identified as the Shc 66 cDNA described by Okada et al. (16). The remaining five clones were identified as the Shc 52/46 cDNA.

Quantitation of mRNA concentration by reverse transcription and competitive polymerase chain reaction. To assess Shc isoform mRNA levels, lungs were obtained from fetal Swiss-Webster mice at 12 to 22 d postconception (term = 18.5 d). Specimens from three different litters were obtained at each time point; for d 12 through 14, multiple lungs from the same litter were pooled for each sample. Three independent reverse transcription products were thus evaluated for each data point. Total RNA was extracted by guanidinium isothiocyanate and reverse transcribed as previously described (26).

Shc 66 cDNA concentrations were quantitated by competitive PCR (27) using primers designed to amplify bases 172 to 477 of Shc 66. A competitor template was produced by PCR using composite primers against v-erbB (see diagram in Figure 1A). Solid areas represent portions recognized by the Shc 66-specific primers, whereas hatched areas represent portions specific to v-erbB. PCR amplification of v-ErbB with these composite primers yielded a 430-base competitor template recognized by the Shc 66-specific primers. Shc 66 mRNA expression is estimated by adding a known and constant quantity of competitor template to each reverse-transcribed RNA sample, amplifying the two templates together in a single PCR reaction, and quantitating the two products. A standard curve was constructed using 2 fg of competitor template and 0, 0.625, 1.25, 2.5, 5, 10, 20, 40, 80, and 160 fg of Shc 66 target product (Fig. 1B,lanes 1-10). The target product/competitor product ratio is plotted against the target concentration as a solid line in Figure 3C (see "Results" section). The broken line of Figure 1C represents a similar standard curve in which 2 fg of the competitor DNA is coamplified with 10, 20, 40, 80, and 160 fg of reverse-transcribed total RNA from 18-d fetal mouse. The two lines are superimposable, demonstrating that competitive PCR quantitation is not confounded by contaminating DNA species. The standard curve was highly predictive and regressed to the equation: Equation 1 Shc 66 mRNA expression was quantitated by coamplifying 2 fg of competitor DNA with cDNA reverse transcribed from a 40-ng total RNA sample, determining the ratio of the products and calculating the mRNA concentration as its cDNA equivalent. Reactions were performed with 35 cycles of denaturation at 94°C for 2 min, annealing at 62°C for 2 min, and extension at 72°C for 2 min in a Robocycler 96 (Stratagene, La Jolla, CA) using a reaction mixture of 10 mM Tris (pH 9.4), 50 mM KCl, 2 mM MgCl2, 0.01% gelatin, 0.1% Triton X-100, 10 pmol each primer, and 0.5 U DNA polymerase in a 50 µl volume. PCR products were separated by 3% agarose electrophoresis, stained with 5 µg/mL ethidium bromide, photographed under UV illumination, and the band intensities quantitated by densitometry.

Figure 1
figure 1

Shc 66 competitive PCR assay. Shc 66 competitor template is synthesized by PCR amplification of v-erbB using sequences recognized by Shc 66-specific primers (Fig. 5A, solid regions) annealed back-to-back with primers specific for a 430-base portion of v-erbB (hatched regions). In figure 5B, 2 fg of the competitor template were coamplified with 0, 0.625, 1.25, 2.5, 5, 10, 20, 40, 80, and 160 fg of Shc 66 cDNA (lanes 1- 10). With increasing cDNA concentration, the Shc 66-specific target product increases relative to that of the competitor. Product concentrations are quantitated by densitometry and the ratio plotted against target template concentration as a solid line in Figure 5C. The broken line represents a similar standard curve in which 2 fg of competitor DNA is coamplified with 10, 20, 40, 80, and 160 fg of reverse-transcribed total RNA from an 18-d fetal mouse. The two lines are superimposable, demonstrating that competitive PCR quantitation is not confounded by contaminating species.

Full size image
Figure 3
figure 3

Shc 66 mRNA decreases with maturation. Competitive PCR of reverse-transcribed lung RNA from fetal and neonatal mice at 16 to 25 d postconception is shown in A. Specimens obtained on the last day of gestation before delivery are designated as 18 d postconception, whereas pups obtained immediately after birth are identified as 18.5 d postconception. Relative to the competitor, Shc 66 band intensity decreases after birth (term = 18.5 d gestation). Band intensity is quantitated by densitometry in B. For each specimen, the Shc 66 target/competitor product ratio was normalized against the β-actin target/competitor product ratio. From 18 to 21 d postconception, Shc 66 cDNA levels dropped by 82.6% (p = 0.00845). In specimens obtained immediately before and after birth (18 and 18.5 d postconception, respectively), Shc 66 cDNA levels decreased by 51.6% (p = 0.0204). Bars represent SD. Three samples were assessed at each time point.

Full size image

Similarly, β-actin mRNA expression was quantitated as an internal standard for each reverse transcription sample. This assay also regressed to a highly predictive calibration curve: Equation 2 Shc 46 and Shc 52 are translated from the same mRNA. To specifically amplify this splice variant, primers were designed to straddle the junctions between exons 1 and 2a, and between exons 2a and 3 (18). This strategy yielded the expected single 214-base target product. PCR conditions were identical to those described above except for an annealing temperature of 68°C and addition of 5% dimethylsulfoxide and 1.0 M GC-Melt (Clontech) to the buffer. A 250-base competitor template was synthesized and a standard curve determined, which regressed to: Equation 3 Shc 52/46 mRNA concentrations were normalized against β-actin mRNA content for each sample.

Immunohistochemistry. An isoform-specific polyclonal Shc 66 antibody was prepared using a synthetic peptide corresponding to amino acids 26 to 42. The peptide was conjugated to keyhole limpet hemocyanin and used to immunize New Zealand White rabbits. The antibody was affinity purified using the immunizing peptide immobilized to Sephadex. Isoform specificity was confirmed by Western analysis of proteins immunoprecipitated from 16 d gestation lung cell total lysates by the Shc 66 isoform-specific antibody and by the anti-pan Shc antibody described previously (see Fig. 5).

Figure 5
figure 5

Western analysis demonstrating specificity of anti-Shc 66 antibody. Lanes 1 and 2, primary mesenchymal cell lysates immunoprecipitated with anti-Shc 66 antibody. Lanes 3 and 4, lysates immunoprecipitated with anti-pan Shc antibody. Blot was probed with anti-Shc 66 antibody.

Full size image

The Shc 66-specific antibody was used to immunohistochemically localize Shc 66 expression in sectioned lungs obtained from mice starting at 14 d postconception through 24 d postconception. Three individuals from different litters were used at each time point. Specimens were fixed in Carnoy's solution, transferred to ethanol, sectioned, and mounted. Before immunostaining, sections were deparaffinized, rehydrated in an ethanol gradient, and exposed to 8 min of radiation in 0.01 M citrate buffer pH 6.0 using a Maytag CME-300 commercial microwave oven at full power (28). Sections were quenched in 3% H2O2 in 90% methanol, blocked in normal goat serum, and incubated successively with 2 µg/mL polyclonal rabbit anti-Shc 66 and 10 µg/mL biotinylated goat anti-rabbit IgG antibody. The antigen was visualized with streptavidin-horseradish peroxidase and aminoethyl carbazole (Zymed, South San Francisco, CA). Specificity was confirmed by staining identically processed, known positive sections using 2 µg/mL anti-Shc antibody preabsorbed overnight with 10:1 wt/wt immunizing peptide and demonstrating total lack of staining.

Shc isoform tyrosine phosphorylation. Preliminary experiments were performed in 3T3-Swiss cells to establish the time course of Shc phosphorylation. 3T3-Swiss cells were plated at a density of 2.5 × 104 cells/cm2 in 25 cm2 flasks. The cultures were rested for 24 h, changed to serum-free Delbucco's modified Eagle's medium (DMEM) overnight, and stimulated with 20 ng/mL EGF in DMEM for 20 sec, 40 sec, 1 min, 2 min, 5 min, 10 min, and 15 min. The cells were then rinsed with iced phosphate-buffered saline and lysed in radioimmuno-precipitation assay buffer (1% NP-40, 0.1% SDS, 0.5% deoxycholate, 150 mM NaCl, 10 mM Tris pH 7.4, 1 mg/mL pefabloc, 10 µg/mL E-64, 1 mM ethylenediaminetetraacetic acid, 1 mM ethyleneglycoltetraacetic acid, 1.0 mM NaVO4, 0.2 mM phenylmethylsulfonyl fluoride). Lysates were cleared by centrifugation at 14 000 × g for 20 min, the protein concentrations determined by bicinchonoic acid method, and the samples normalized by protein content. Shc proteins were then immunoprecipitated with rabbit anti-Shc antibody (Transduction Laboratories, Lexington, KY) and Protein A/G agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA). Beads were washed in radioimmuno-precipitation assay buffer, and resuspended in 1% SDS loading buffer, and the precipitates released by boiling. Precipitates were subjected to 7.5% SDS-PAGE, transferred to nitrocellulose, blocked with 3% BSA, and probed with recombinant anti-phosphotyrosine antibody (Transduction Laboratories, Lexington, KY). Specific binding was detected by chemiluminescence and phosphoprotein contents quantitated by densitometry. Three trials were performed.

Shc isoform tyrosine phosphorylation was then assessed in primary fetal lung mesenchymal cells isolated from timed gestation mice at 16 d postconception by tryptic digestion and differential adhesion to tissue culture plastic as previously described (25). Fetal lungs were isolated from timed-gestation Swiss-Webster mice and rinsed in Ca2+- and Mg2+-free Hanks' Balanced Salt Solution (HBSS). All media were supplemented with 100 units/mL penicillin G, 100 Fg/mL gentamicin, and 0.25 fg/mL amphotericin B. Fetal lungs from four to six litters were pooled for each experiment and dissociated with 0.1% trypsin and 0.001% DNAase in DMEM. The cells were filtered through a 10 fm nylon mesh and plated onto 75 cm2 tissue culture flasks for 1 h, after which the nonadherent cells were discarded. Adherent cells were detached with 0.05% trypsin, resuspended in DMEM with 10% fetal bovine serum, and plated at a density of 2.5 × 104 cells/cm2 in 25-cm2 flasks. Cells were rested for 24 h and starved overnight as previously described before stimulation with 20 ng/mL EGF for 1 min. Shc tyrosine phosphorylation was then assessed by Shc immunoprecipitation and phosphotyrosine Western analysis as previously described. Three trials were performed with essentially identical results.

Statistical analyses. Statistical comparisons were performed on the largest differences within a comparison group and repeated on successively smaller differences until comparisons were no longer significant. Significance was assessed by two-tailed t test assuming equal variances, with significance accepted at p = 0.05. Comparisons were performed using SPSS for Windows Release 6.1.2 (SPSS, Chicago, IL) and Excel 5.0a (Microsoft, Redmond, WA).

RESULTS

Western analysis. Western analyses were performed on a single blot to minimize differences in technique. Sample volumes were adjusted to equalize protein content. The concentration of Shc 52 gradually decreased between 14 and 20 d postconception (p = 0.08, term = 18.5 d) (Fig. 2, A and B). Shc 66 and Shc 46 were each expressed at approximately half the concentration of Shc 52 until d 18.5, after which the relative expression of these two isoforms decreased. Specimens obtained on the last day of gestation before parturition are designated as 18 d postconceptual age, whereas tissues obtained immediately after delivery are assigned at postconceptual age of 18.5 d.

Figure 2
figure 2

Relative Shc 66 and Shc 46 protein expression are differentially downregulated in the peripartum period. (A) Western analysis of whole lung lysates from fetal and newborn mice of 14 to 20 d PC. Each data point was performed in triplicate and obtained from a single blot. Specimens obtained immediately after birth are designated 18.5 d PC. Shc 66 and Shc 46 protein expression decrease rapidly after parturition, whereas Shc 52 protein expression decreases gradually throughout the study period. (B) Densitometric quantitation of Shc isoform expression. Shc 66 tends to decrease with advancing gestation (p = 0.08). Shc 66 and 46 are expressed at approximately half the level of Shc 52 until parturition, when expression of these isoforms drops disproportionately. (C) Shc 66 protein expression normalized against Shc 52 content for each specimen. From d 17 to 22 PC, relative Shc 66 expression decreases by 84.9% (p = 0.004), and from d 18.5 to 22 PC, relative Shc 66 expression decreases by 71.7% (p = 0.028). All other comparisons were statistically insignificant. (D) Shc 46 protein expression normalized against Shc 52 content for each specimen. From d 17 to 20 PC, relative Shc 42 expression decreased by 83.3% (p = 0.0002); from d 18.5 to 20 PC, relative Shc 46 expression decreased by 68.7% (p = 0.002); and from d 17 to 18.5 PC, relative Shc 46 expression dropped by 46.7% (p = 0.001). All other comparisons were statistically insignificant. Bars represent SD.

Full size image

To determine whether the ratio of Shc 66 to Shc 52 is downregulated, Shc 66 protein concentrations were normalized against those of Shc 52 within each sample (Fig. 2C). From postconceptual d 17 to 20, relative Shc 66 expression decreased by 84.9% (p = 0.004), and from postconceptual d 18.5 to 20, relative Shc 66 expression decreased by 71.7% (p = 0.028). All other comparisons were statistically insignificant. These data indicate that Shc 66 expression relative to Shc 52 decreases significantly in the peripartum period.

Similarly, Shc 46 expression was normalized against Shc 52 content (Fig. 2D). From postconceptual d 17 to 20, relative Shc 42 expression decreased by 83.3% (p = 0.0002); from postconceptual d 18 to 20, relative Shc 46 expression decreased by 68.7% (p = 0.002); and from postconceptual d 17 to 18.5, relative Shc 46 expression dropped by 46.7% (p = 0.001). All other comparisons were statistically insignificant. These data indicate that Shc 46 expression relative to Shc 52 decreases significantly in the peripartum period.

Competitive PCR. Shc 66, Shc 52/46, and β-actin mRNA concentrations were quantitated in fetal mouse lungs from 16 to 25 d postconceptual age, and the Shc 66 and Shc 52/46 mRNA concentrations were normalized against the β-actin mRNA content within each sample. The concentration of Shc 66 mRNA dramatically decreases after birth, parallel to the drop in Shc 66 protein expression observed by Western analysis (Fig. 3A). From 18 to 21 d postconception, Shc 66 cDNA levels dropped by 82.6% (p = 0.008) (Fig. 3B). The rate of decrease in Shc 66 mRNA expression is most dramatic in the peripartum period. In mouse pups obtained on the d 18 of gestation immediately before and after birth (18 and 18.5 d postconception, respectively), Shc 66 cDNA levels fell by 51.6% (p = 0.02).

By comparison, Shc 52/46 cDNA levels peaked at d 15, then diminished gradually from d 16 onward. The expression of Shc 52/46 cDNA parallels the combined expression of Shc 52 and Shc 46 proteins during this period (Fig. 4).

Figure 4
figure 4

Shc 52/46 mRNA expression parallels combined Shc 52 and Shc 46 protein expression. Shc 52/46 mRNA expression peaks on the 15th postconceptual day, then decreases gradually in parallel with a decrease in combined Shc 46 and Shc 52 protein expression. Protein expression was calculated from Western blot data of Figure 1. Shc 52 and Shc 46 are alternative translational products of the Shc 52/46 mRNA, which was quantitated by competitive RT-PCR amplification of the exon specific to this mRNA. Shc 52/46 mRNA expression decreases significantly between d 15 and 19 (p = 0.04).

Full size image

Immunohistochemistry. To demonstrate specificity of the Shc 66 isoform-specific antibody, primary fetal mouse lung mesenchymal cell lysates were immunoprecipitated with anti-Shc 66 antibody and anti-pan Shc antibody. Immunoprecipitates were then subjected to Western analysis using the anti-Shc 66 antibody. A single major band is present in all lanes, demonstrating specificity for Shc 66 (Fig. 5).

The anti-Shc 66 antibody was used to immunolocalize Shc 66 in fetal mouse lungs at different gestational ages. Sections from each gestational age were mounted on the same slide, assuring identical processing and staining conditions. Two sets of these multispecimen slides were produced, incorporating samples from three individuals from different litters at each time point. Results were identical between individual specimens at each time point.

At 14 d gestation, Shc 66 is expressed in the cytoplasm and perinuclear areas of most lung cells (Fig. 6A). Localization is strongest and most evident in peribronchial smooth muscle precursors and in the columnar epithelia of intermediate-size airways (inset). Periarterial smooth muscle expresses Shc 66 at relatively high levels at all ages examined. At 15 d (Figure 6B), Shc 66 expression becomes both stronger and more restricted, and is most evident in the nuclei of a single line of mesenchymal cells adjoining the parenchymal face of the basement membrane investing the large primordial airways (inset). The clusters of the overlying airway epithelial cells also exhibit some Shc 66 staining. Scattered stained nuclei are evident throughout the parenchyma, among both mesenchymal and epithelial populations. The parenchymal expression is similar to that of the 14-d lung.

Figure 6
figure 6

Immunolocalization of Shc 66 in developing lung. At 14 d gestation, Shc 66 is expressed in the cytoplasm and perinuclear areas of most lung cells (A). Staining is most evident in peribronchial smooth muscle precursors and columnar epithelia of intermediate-size airways (inset). Periarterial smooth muscle expresses Shc 66 at all ages examined. At 15 (B) and 16 d (C) gestation, perinuclear staining becomes both stronger and more restricted to the bronchial smooth muscle precursors and overlying epithelia (inset). By 17 d gestation, cells with perinuclear staining have become less common and are concentrated in the large airways (inset) (D). Scattered cells with perinuclear staining are present toward the periphery of the lung. Expression is dramatically decreased immediately after parturition (18.5 d postconception) (E). Cells expressing perinuclear Shc 66 are rare and restricted to the pleural margins. Lung tissue at 15 d gestation was not stained by anti-Shc 66 antibody preabsorbed with the immunizing peptide (F).

Full size image

The expression of Shc 66 in the 16-d lung (Fig. 6C) continues to be widespread, with generalized parenchymal staining and strong bronchial and periarterial perinuclear staining. Both epithelial and mesenchymal populations express Shc 66 at high levels. The pattern of Shc 66 expression is similar to that observed in the 15-d lung. By d 17 of gestation, the line of peribronchial cells with strong perinuclear Shc 66 staining has become discontinuous and patchy. Shc 66 expression, however, continues to be most obvious near the large airways (Fig. 6D). Shc 66 is still highly expressed in the lung parenchyma, with cells exhibiting both generalized cytosolic staining and occasional perinuclear localization. Perinuclear expression of Shc 66 is more common toward the periphery of the lung.

Lungs obtained from mouse pups immediately after parturition (18.5 d postconception) have dramatically decreased Shc 66 protein expression (Fig. 6E). The peribronchial mesenchymal cells now express little Shc 66. Parenchymal expression of Shc 66 has also largely disappeared, and cells expressing perinuclear Shc 66 are rare and mostly restricted to the pleural margins. At 20 d gestation (2 d postpartum) (not shown), Shc 66 expression is quite rare and restricted to the pleural margins.

Lung tissue at 15 d gestation was not stained by anti-Shc 66 antibody preabsorbed with the immunizing peptide (Fig. 6F). The concentration of the antibody was unchanged, and the sections were adjacent and processed identically to the previously described specimens. Preabsorbed antibody at this concentration did not react with specimens from any gestational age.

Shc isoform tyrosine phosphorylation. The time course of Shc isoform tyrosine phosphorylation was first assessed in 3T3-Swiss cells. This is a line of contact-inhibited fetal lung mesenchymal cells with mitogenic responses very similar to those of primary fetal lung mesenchymal cells (unpublished findings). Phosphorylation of all three Shc isoforms begins within 20 s of EGF stimulation and peaks by 40 s (Fig. 7, B and C). Shc tyrosine phosphorylation diminished rapidly afterward, falling to half-maximal values by 2 min of stimulation. These results suggest that all three Shc isoforms participate in EGF signaling and that peak tyrosine phosphorylation occurs between 20 s and 2 min after initiation of EGF stimulation. Shc isoform phosphorylation was then assessed in primary mesenchymal cells isolated from the lungs of fetal mice at 16 d gestation. All three Shc isoforms are phosphorylated by 1 min exposure to 20 ng/mL EGF (Fig. 7A), consistent with results obtained from 3T3-Swiss cells.

Figure 7
figure 7

All three Shc isoforms are tyrosine phosphorylated by EGF stimulation in fetal mesenchymal cells. (A) EGF stimulation of primary cells isolated by tryptic digestion from fetal mouse lungs at 16 d gestation. Cells were stimulated for 1 min with 20 ng/mL EGF, lysed, the Shc isoforms immunoprecipitated, and precipitates subjected to phosphotyrosine Western analysis. All three Shc isoforms are phosphorylated, suggesting participation in EGF signaling. (B) Time course of Shc isoform phosphorylation in 3T3-Swiss fetal murine mesenchymal cell line. Cells were stimulated with 20 ng EGF for up to 15 min as noted before lysis and assessment of Shc tyrosine phosphorylation as noted previously. Densitometric quantitation (C) demonstrates prolonged peak Shc isoform phosphorylation occurs between 20 sec and 2 min. Both experiments were repeated three times.

Full size image

DISCUSSION

These data indicate that the pulmonary expression of Shc 66 relative to Shc 52 is strongly attenuated during the peripartum period, concomitant with pulmonary maturation and adaptation to air breathing. Shc 66 has been characterized as a domain antagonist of Shc 52 (18,16), and it is likely that the mitogenicity of Shc signaling depends on the stoichiometry of Shc isoform expression. Decreased relative expression of pulmonary Shc 66 after parturition suggests that Ras activation and MAP kinase signaling become increasingly responsive to tyrosine kinase signaling through Shc. Shc 46 is also downregulated after birth; however, because Shc 46 has not yet been demonstrated to have a distinct role in mitogenic signaling, the significance of this finding is unclear.

In the earliest specimens examined, Shc 66 is widely distributed throughout the developing lung. Beginning at d 15 of gestation, the population of cells expressing Shc 66 becomes increasingly restricted to the peribronchial smooth muscle precursors and adjacent epithelia in apposition to the basement membrane. Shc 66 is concentrated in the perinuclear areas of these cells, where the Shc isoforms accumulate after activation by receptor tyrosine kinases (14). Shc phosphorylation by the laminin-binding α6β4 integrin is associated with keratinocyte proliferation (29), and we speculate that Shc 66 is phosphorylated in response to the laminin component of the basement membrane. Mesenchymal cells adjoining the basement membrane express more Shc 66 than neighboring cells not in contact with the membrane, and we suggest that expression as well as activation of Shc isoforms is regulated by extracellular matrix.

Shc 66 protein expression diminishes in tandem with a decrease in Shc 66 mRNA levels. Shc 66 mRNA is transcribed from the same gene as the Shc 52/46 mRNA, and the two mRNAs differ by the excision of a single intron. Shc 66 downregulation may therefore result from increased Shc RNA editing efficiency during lung maturation. However, the expression of Shc 52/46 mRNA or protein products do not increase as Shc 66 expression decreases, as would be expected from modulation of RNA editing efficiency alone. We therefore suggest that primary Shc gene transcription is also regulated. Alternatively, the two mRNAs may be initiated from separate and independently regulated initiation sites.

The mechanisms underlying Shc 46 expression regulation may also be of interest. Shc 46 protein concentrations parallel Shc 66 protein expression, but because Shc 46 is an alternative translational product (11), its regulation presumably differs from that of Shc 66. Shc 46 expression is not significantly increased by transfection with Shc 66 mRNA (18), and Shc 52/46 cDNA levels parallel the combined expression of Shc 52 and Shc 46 proteins. Our data indicate that the mechanisms regulating Shc 66 RNA editing and translation from the Shc 46 initiation site are somehow synchronized. Again, because specific roles for Shc 46 have not yet been established, the implications of Shc 46 expression regulation are still unclear.

All three Shc isoforms are tyrosine phosphorylated by EGF stimulation, enabling their association with Grb2 (16,18,19). Because Shc 66 is believed to compete with Shc 52 for limited intracellular pools of Grb2, Shc 66 phosphorylation is consistent with its putative dominant negative function. The phosphorylation kinetics of the three Shc isoforms are similar, suggesting that the mechanism of Shc dephosphorylation is not isoforms dependent.

We conclude that the three Shc isoforms are differentially regulated in the developing murine lung. The relative expression of Shc 66 and Shc 46 decrease during lung maturation. Shc 66 is spatially as well as temporally restricted in the developing lung, and its expression may be regulated by the extracellular matrix. Shc 46 presumably is regulated through separate mechanisms, but its temporal expression parallels that of Shc 66. The expression of both isoforms are significantly attenuated at parturition, coincident with the physiologic adaptation of the lung to extrauterine existence. The coordinated downregulation of the Shc isoforms in the perinatal lung suggests a distinct developmentally specific role for each Shc isoform, and we speculate that the disappearance of Shc 46 and Shc 66 may be important in the pulmonary cellular adaptation to extrauterine existence.