SEARCH

SEARCH BY CITATION

Keywords:

  • neuregulin;
  • ErbB;
  • morphogenesis;
  • submandibular gland;
  • mouse

Abstract

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

Dimerization and activation of ErbB receptors by their ligands play crucial roles in organogenesis. Epithelial morphogenesis of embryonic mouse submandibular gland (SMG) has been shown to depend on intraepithelial signaling mediated by the epidermal growth factor (EGF) family of molecules and the EGF receptor (ErbB1). Here, we report on the neuregulin (NRG) -1 protein and its receptors ErbB2 and ErbB3 in the developing SMG. The expression of these molecules was demonstrated by reverse transcriptase-polymerase chain reaction and Western blot analysis. Immunofluorescence microscopy showed that the two ErbB receptors as well as ErbB1 were expressed mainly in the epithelium, whereas NRG-1 was exclusively found in the mesenchyme. Epithelial morphogenesis was retarded by anti–NRG-1 neutralizing antibody and promoted by recombinant NRG-1 protein. We suggest that, in the developing SMG, both mesenchyme-derived NRG molecules and epithelium-derived EGF molecules regulate ErbB signaling in the epithelium to participate in tissue morphogenesis. Developmental Dynamics 230:591–596, 2004. © 2004 Wiley-Liss, Inc.


INTRODUCTION

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

Branching morphogenesis of the epithelium is a fundamental process in the development of many organs, such as exocrine glands, the kidneys, and the lungs. This process is dependent on reciprocal interactions with the adjacent mesenchyme, which involve molecular signaling mediated by various growth factors and their receptors (Hogan, 1999). The submandibular gland (SMG) of embryonic mice has been studied as a model system to understand the mechanisms of epithelial branching morphogenesis and epithelial–mesenchymal interaction (Hieda and Nakanishi, 1997; Melnick and Jaskoll, 2000). One of the best-characterized signaling systems that participate in regulation of SMG epithelial morphogenesis is the epidermal growth factor (EGF) family of ligands and their receptor, the tyrosine kinase-containing EGFR (also called ErbB1). It has been shown that the branching morphogenesis is retarded by an EGFR inhibitor and is promoted by EGF through activation of the EGFR (Kashimata and Gresik, 1997; Kashimata et al., 2000). Signaling from the EGFR in the embryonic SMG triggers stimulatory actions by means of the mitogen-activated protein kinases ERK-1/2 and phosphatidylinositol-3-kinase (PI3K) but inhibitory ones by means of protein kinase C (PKC) isozymes (Kashimata et al., 2000; Koyama et al., 2003). Furthermore, another member of the EGF family, heparin-binding EGF-like growth factor (HB-EGF), and its processing by metalloproteinases are involved in SMG epithelial morphogenesis (Umeda et al., 2001). Because both EGFR and its ligands, including EGF, HB-EGF, and transforming growth factor-α, are expressed predominantly in the epithelium of the developing SMG (Gresik et al., 1997; Jaskoll and Melnick, 1999; Umeda et al., 2001), it is likely that EGFR/ErbB1 and its ligands there mediate intraepithelial signaling to regulate tissue morphogenesis.

EGFR/ErbB1 is a member of the ErbB family of receptor tyrosine kinases, which includes three other members, i.e., ErbB2/Neu/HER2, ErbB3/HER3, and ErbB4/HER4 (Olayioye et al., 2000). They all have an extracellular ligand-binding domain, a single membrane-spanning region, and a cytoplasmic protein tyrosine kinase domain. ErbB receptors are activated by several ligands belonging to the EGF and neuregulin (NRG) families (Adlkofer and Lai, 2000; Olayioye et al., 2000; Buonanno and Fischbach, 2001; Fall, 2003). The EGF family members bind to EGFR/ErbB1, although some members also bind to ErbB4. Both NRG-1 and NRG-2 complex with ErbB3 and ErbB4, whereas NRG-3 and NRG-4 bind ErbB4 but not ErbB3. The ligand binding leads to the formation of homo- and heterodimers of ErbB receptors, stimulating their intrinsic tyrosine kinase activity. Although no direct ligand for ErbB2 has yet been identified, ErbB2 serves as the preferred heterodimerization partner for all other ErbB receptors and potentiates ErbB receptor signaling. The analysis of genetically modified mice has shown the importance of signaling mediated by ErbB receptors and their ligands in embryogenesis and in the tissue morphogenesis of many organs (Adlkofer and Lai, 2000; Olayioye et al., 2000; Buonanno and Fischbach, 2001; Citri et al., 2003; Fall, 2003).

In the present study, we sought to determine the role of signaling mediated by NRGs and their receptors in the development of the embryonic mouse SMG. We found that, in the developing SMG, ErbB2 and ErbB3 as well as ErbB1 were expressed mainly in the epithelium, whereas NRG-1 was detected in the mesenchyme. SMG epithelial morphogenesis in culture was blocked by neutralizing antibody against NRG-1 and promoted by recombinant NRG-1 protein. These results, together with those reported previously, suggest that, in the developing SMG, both epithelium-derived EGF molecules and mesenchyme-derived NRG molecules regulate ErbB signaling in the epithelium to participate in morphogenesis of the tissue.

RESULTS AND DISCUSSION

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

Expression of ErbB Receptors in Developing SMG Epithelium

The expression of the ErbB receptors (ErbB1∼4) in the developing SMG was examined by reverse transcriptase-polymerase chain reaction (RT-PCR) analysis using specific primer sequences. At all of the developmental stages examined (E12–E15), the mRNA expression of all four ErbB receptors was detected, although the level of ErbB4 mRNA was rather low (Fig. 1A). Immunofluorescence microscopy was then performed on tissue sections from the E14 SMG to determine the expression site of ErbB receptors. The expression of ErbB1, ErbB2, and ErbB3 was clearly detected in the SMG rudiments (Fig. 1B). ErbB1 was expressed only in the epithelial tissue and the staining intensity appeared stronger in the stalk cells than in the lobular cells, as reported previously (Gresik et al., 1997; Kashimata et al., 2000). Expression of ErbB2 and ErbB3 was also detected mainly in the epithelium but also in certain mesenchymal cells. The anti-ErbB4 antibody we used gave no positive staining (data not shown). These results indicate that ErbB2 and ErbB3 as well as ErbB1are expressed mainly in epithelium of the developing SMG.

thumbnail image

Figure 1. Expression of ErbB receptors in the developing submandibular gland (SMG). A: Reverse transcriptase-polymerase chain reaction (RT-PCR) analysis detected obvious expression of mRNA for ErbB1, ErbB2, and ErbB3 in SMG rudiments at embryonic day (E) 12–E15; whereas that of the mRNA for ErbB4 was weak. PCR products (526 bp for ErbB1, 248 bp for ErbB2, 220 bp for ErbB3, and 202 bp for ErbB4) were fractionated on 2% agarose gels. B: Immunofluorescence microscopy of E14 SMG rudiments revealed the predominant expression of ErbB1, ErbB2, and ErbB3 in epithelium. Certain mesenchymal cells near the epithelium also expressed ErbB2 and ErbB3 (arrows). Scale bar = 50 μm in B.

Download figure to PowerPoint

Identification of Type I NRG-1 Expressed in Developing SMG

We next investigated the expression of the neuregulin (NRG) family of ligands for ErbB receptors in the developing SMG. Here, we focused on NRG-1, which binds ErbB3 and is the best-characterized member of the family (Adlkofer and Lai, 2000; Buonanno and Fischbach, 2001; Fall, 2003). RT-PCR analysis was conducted to examine the expression of NRG-1 (Fig. 2). Alternative splicing in different parts of the NRG-1 transcript give rise to multiple protein isoforms, which can be classified into type I (ARIA/NDF/HRG), type II (GGF), and type III (SMDF) forms (Lemke, 1996; Fischbach and Rosen, 1997; Fall, 2003). We designed a primer pair corresponding to the sequences conserved in type I and type II forms, but not in type III form, of NRG-1 to amplify the coding region of the immunoglobulin (Ig)-like through the epidermal growth factor (EGF)-like domains of the molecules. This primer pair was expected to yield a 531-bp product for the type I form and a 430-bp product for the type II one due to the presence in type I, but not in type II form, of a glycosylation-rich segment between the Ig-like and EGF-like domains. We found that, at all of the stages of SMG development examined (E12–E15), a single product of 531 bp was clearly amplified; whereas the 430-bp product was not detectable (Fig. 2A), indicating the predominant expression of type I NRG-1 in the developing SMG. Splicing of NRG-1 also gives rise to variants at the C-terminal part of the EGF-like domain, yielding α and β forms; and NRG-1 transcripts encode either secreted or transmembrane forms. In fact, a PCR product encoding a transmembrane form was obtained (data not shown). These analyses identified the expression in developing mouse SMG of a transmembrane form of type I NRG-1α, whose amino acid sequence exhibited ∼97% identity with that of human and rat NRGs.

thumbnail image

Figure 2. Expression of neuregulin (NRG) -1 in the developing submandibular gland (SMG). A: Reverse transcriptase-polymerase chain reaction analysis was conducted by using a primer pair that could amplify mRNAs for type I and type II NRG-1s. At all developmental stages examined (embryonic day [E] 12–E15), the mRNA for type I NRG-1 (531-bp product) was clearly detected, whereas that for type II (430-bp product) was not detectable. B: Western blots with anti–NRG-1 antibody recognizing the extracellular region detected proteins with apparent molecular masses of approximately 75 kDa, 55 kDa, 45 kDa, and 27 kDa in the developing SMG (arrows). C: Immunofluorescence microscopy of E14 SMG. The anti–NRG-1 antibody exclusively stained the mesenchyme, whereas the epithelium was totally negative. Scale bar = 50 μm in C.

Download figure to PowerPoint

Mesenchyme-Specific Expression of NRG-1 Molecules in Developing SMG

We next investigated the expression of NRG-1 molecules in the developing SMG by Western blot analysis. Anti–NRG-1 antibody recognizing the extracellular region of both the α- and β-isoforms clearly detected proteins with apparent molecular masses of approximately 75 kDa, 55 kDa, 45 kDa, and 27 kDa (Fig. 2B). The level of the 27-kDa and 55-kDa proteins was significantly increased at the later stages of SMG development. Judging from the molecular masses, we suspect the 27-kDa protein to be a secreted form of NRG-1 and those with higher molecular masses to be the transmembrane forms probably with distinct cytoplasmic tails (Lemke, 1996; Fischbach and Rosen, 1997; Fall, 2003). It remains to be determined whether the 27-kDa protein detected was the translational product of mRNA encoding the secreted form or the proteolytic product of the larger, transmembrane forms (Loeb et al., 1998, 1999; Montero et al., 2000; Shirakabe et al., 2001; Wang et al., 2001). These results demonstrate the expression of both secreted and transmembrane forms of NRG-1 in the developing SMG.

To determine the expression site of NRG-1 molecules in the developing gland, we performed immunofluorescence microscopy on tissue sections of E14 SMG by using the same anti–NRG-1 antibody as used for the Western blotting. We found that the anti–NRG-1 antibody exclusively stained the mesenchymal cells on their surfaces and the basal lamina region (Fig. 2C). It should be noted that, despite the expression of transmembrane forms of NRG-1 in the developing SMG, the epithelial tissue of the SMG rudiments was completely negative for NRG-1 molecules. These results suggest that NRG-1 is expressed predominantly by mesenchymal cells in the developing SMG.

Involvement of NRG-1 in SMG Epithelial Morphogenesis

The role of NRG-1 in SMG development was investigated in an organ culture system. As shown in Figure 3, anti–NRG-1α neutralizing antibody inhibited epithelial growth and branching morphogenesis of cultured E13 SMG rudiments. Normal antibody had no apparent effect on the tissue morphogenesis. Culturing SMG rudiments with the anti–NRG-1α antibody resulted in a 35% decrease in branching compared with normal antibody treatment: the ratio of the number of lobules at 48 hr to the number of lobules at 0 hr was 6.7 ± 1.5 for the untreated glands, 6.9 ± 1.0 for the normal antibody-treated ones, and 4.5 ± 1.1 for the anti–NRG antibody-treated ones (n = 7; P < 0.01). On the other hand, recombinant NRG-1α EGF-like domain protein, which can mimic the biological effects of the complete NRG-1 molecule, significantly promoted SMG epithelial branching morphogenesis (Fig. 4). The ratio of the number of lobules at 48 hr to the number of lobules at 0 hr was 5.3 ± 0.7 and 7.7 ± 1.9 for the untreated and treated glands, respectively (n = 6; P < 0.01), indicating a 45% increase in branching by NRG-1α treatment. A similar extent of promotion of epithelial branching was observed at the higher protein concentration of 80 ng/ml (data not shown).

thumbnail image

Figure 3. Effects of anti-neuregulin (NRG) -1α neutralizing antibody on submandibular gland (SMG) epithelial morphogenesis. Embryonic day (E) 13 SMG rudiments were cultured for 48 hr in medium with no antibody added, normal goat IgG, or goat anti–NRG-1α neutralizing antibody. Antibodies were added at a concentration of 400 μg/ml. A: Light photomicrographs of cultured glands. B: The number of lobules expressed as a ratio of the number of lobules at 48 hr to the number of lobules at 0 hr. The branching ratio was significantly lower in the explants treated with anti–NRG-1a antibody than in the untreated and normal goat IgG-treated ones (n = 7; P < 0.01).

Download figure to PowerPoint

thumbnail image

Figure 4. Effects of recombinant neuregulin (NRG) -1 protein on submandibular gland (SMG) epithelial morphogenesis. Embryonic day (E) 13 SMG rudiments were cultured for 48 hr in the absence (control) or the presence of recombinant human NRG-1α epidermal growth factor–like domain protein (20 ng/ml). A: Light photomicrographs of cultured glands. B: The number of lobules expressed as a ratio of the number of lobules at 48 hr to the number of lobules at 0 hr. The branching ratio was significantly higher in the treated explants than in controls (n = 6; P < 0.01).

Download figure to PowerPoint

In summary, our data show that NRG-1 and ErbB2/3 receptors expressed by mesenchyme and epithelium, respectively, are involved in epithelial branching morphogenesis of the developing SMG. It should be noted that, although ErbB3 can bind NRG-1, this receptor is defective in tyrosine kinase activity due to substitutions in critical residues in its kinase domain (Guy et al., 1994) and that NRG-1 binding to ErbB3 activates ErbB2, for which no definite ligand has yet been discovered, as a result of receptor dimerization, leading to tyrosine phosphorylation of both molecules (Alimandi et al., 1995; Pinkas-Kramarski et al., 1996; Citri et al., 2003). Accordingly, it is probable that activation of the ErbB2/3 heterodimer in the epithelium by mesenchyme-derived NRG-1 participates in SMG epithelial morphogenesis. ErbB2 is the preferred heterodimerization partner for all other ErbB family members (Tzahar et al., 1996; Graus-Porta et al., 1997) and plays a role in the potentiation of ErbB receptor signaling (Beerli et al., 1995; Graus-Porta et al., 1997). It has been demonstrated that ErbB1 activation by the EGF family members EGF and HB-EGF participate in SMG epithelial morphogenesis (Kashimata and Gresik, 1997; Kashimata et al., 2000; Umeda et al., 2001). In contrast to NRG-1, the EGF family of ligands is expressed in the epithelium of developing SMG (Jaskoll and Melnick, 1999; Umeda et al., 2001). Epithelial morphogenesis of the developing SMG may involve the intricate regulation of the dimerization of ErbB2 with ErbB1 regulated by epithelium-derived EGF family molecules and with ErbB3 by mesenchyme-derived NRG-1.

EXPERIMENTAL PROCEDURES

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

Materials and Antibodies

Recombinant human NRG-1α EGF-like domain protein was obtained from R&D Systems (Minneapolis, MN). Antibodies used were the following: rabbit anti-EGFR (1005), rabbit anti-ErbB2 (C-18), and rabbit anti-ErbB3 (C-17), from Santa Cruz Biotechnology (Santa Cruz, CA); mouse anti-NRG-1 (clone 7D5), from Lab Vision (Fremont, CA); and goat anti-human NRG-1α neutralizing antibody raised against the EGF-like domain and normal goat IgG, from R&D Systems. Biotinylated goat anti-mouse IgG and goat anti-rabbit IgG, and peroxidase–streptavidin were from Vector (Burlingame, CA).

Mouse Embryos and Organ Culture

SMG rudiments were dissected from embryos of ddY strain mice (Nihon SLC, Hamamatsu, Japan) in Hanks' balanced salt solution. The day of discovery of the vaginal plug was designated as embryonic day 0 (E0). Organ rudiments were cultured on Millipore filters at liquid–air interface in medium 199 containing 10% fetal calf serum, as described previously (Umeda et al., 2001).

Western Blots

SMG rudiments were homogenized in lysis buffer: 20 mM Tris, pH 7.4, containing 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid, 1 mM ethyleneglycoltetraacetic acid, 1% Triton X-100, 1 mM Na3VO4, 1 mM β-glycerophosphate, 1 μg/ml leupeptin, and 1 mM phenylmethyl sulfonyl fluoride. Homogenates were centrifuged at 15,000 rpm for 20 min at 4°C, and the supernatants were recovered for determination of protein content. Aliquots were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis and electroblotted onto Hybond-P polyvinylidene difluoride membranes (Amersham Biosciences, Tokyo, Japan). Specific proteins on the membranes were detected by incubating with specific primary antibodies for 3 hr, followed by species-specific secondary antibodies complexed with horseradish peroxidase and staining by NBT/H2O2, as previously described (Kashimata et al., 2000).

RT-PCR

Isolation of total RNA from SMG rudiments and reverse transcription to cDNA was performed as described previously (Umeda et al., 2001). In the PCR reactions for ErbB1, ErbB2, ErbB3, and ErbB4 expression, the following primers were used: 5′-TGTTGCTCATAAAGACGTTT-3′ (forward) and 5′-ATGCGACCCTCAGGGACCGC-3′ (reverse) for ErbB1 (GenBank/EMBL/DDBJ accession no. AF124513); 5′-CTTTGGTTACCCCCACTGCC-3′ (forward) and 5′-GGGGGAGCTGGTCGATGCTG-3′ (reverse) for ErbB2 (accession no. L47239); 5′- GAGCGGGGTGACGGGAGTAA-3′ (forward) and 5′-GGGTCGCGAACAGTTCTCCC-3′ (reverse) for ErbB3 (accession no. L47240); and 5′- CATCTCAGCCGTTGCACCCT-3′ (forward) and 5′-TGCTGAGGAATATTTGGTCC-3′ (reverse) for ErbB4 (accession no. L47241). For analysis of NRG-1 expression, mouse EST clones encoding the proteins (accession nos. AA798243, AI197081, AW045376, BB595402, BE648742, BE983721, and BE984041) were picked up, and two sets of primer pairs were prepared: 5′-GCACTTGCACAAGTATCTTGAGGG-3′ (forward) and 5′-CCTCCCAGATTGAAAGAGATGAAA-3′ (reverse), which cover the immunoglobulin-like through epidermal growth factor-like domains, and 5′-TCTGGCATGCCTGAGGAAGCT-3′ (forward) and 5′-CCCTCAAGATACTTGTGCAAGTGC-3′ (reverse), which cover the EGF-like domain through a portion of the cytoplasmic domain. The reaction mixtures were heated at 94°C for 5 min and then subjected to 35 cycles for amplification. Cycle parameters were as follows: ErbB1 and ErbB4, 94°C for 1 min, 61°C for 1 min, and 72°C for 1 min; and ErbB2, ErbB3, and NRG-1, 94°C for 1 min, 67°C for 1 min, and 72°C for 1 min. Aliquots of PCR products were subjected to agarose-gel electrophoresis and visualized by ethidium bromide staining. The identity of the PCR products was verified by restriction enzyme digestion and sequencing.

Immunofluorescence Microscopy

SMG rudiments were embedded in optimal cutting temperature (O.C.T.) compound and frozen in liquid nitrogen. Tissue sections of 8-μm thickness were cut in a cryostat (Leica Instruments GmbH, Nussloch, Germany). For detection of ErbB molecules, tissue sections were fixed in methanol at −20°C for 10 min and treated with 0.5% Triton X-100 in phosphate buffered saline (PBS) for 10 min; and for that of NRG-1, sections were fixed in 4% paraformaldehyde in PBS for 20 min. Then, the sections were incubated with primary antibodies, washed in PBS, and incubated with fluorescein isothiocyanate–conjugated secondary antibodies. The sections were examined under an Olympus BX-50 epifluorescence microscope.

Statistical Analysis

Branching differences between untreated and antibody- or peptide-treated glands were quantitated by calculating the ratio of the number of lobules at 48 hr after culture to that at the onset of culture (mean ± SD). Statistical significance was determined by using an unpaired t-test.

Acknowledgements

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

We thank Dr. M. Kashimata (Asahi University) for advice on Western blotting and Dr. H. Inoue (Osaka University) for advice on sequence analysis.

REFERENCES

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