|
|
||||||||
1Department of Internal Medicine, University of Iowa Roy J. and Lucille A. Carver College of Medicine; and 2Department of Otolaryngology, Head and Neck Surgery, University of Iowa Hospitals and Clinics, Iowa City, Iowa
Submitted 30 December 2005 ; accepted in final form 1 February 2006
| ABSTRACT |
|---|
|
|
|---|
, an erbB ligand. In addition, coculturing with human lung fibroblasts tested whether in vivo subepithelial fibroblasts function as an endogenous source of ligands able to activate erbB receptors expressed by the overlying epithelial cells. Epithelia were stained with hematoxylin and eosin and used for morphometric analysis. Trastuzumab treatment decreased the ciliated cell number by 49% and increased the metaplastic, flat cell number by 640%. Heregulin-
treatment increased epithelial height and decreased the number of metaplastic and nonciliated columnar cells, whereas it increased the goblet cell number. We found that normal human lung fibroblasts express transforming growth factor-
, heparin-binding epidermal-like growth factor, epiregulin, heregulin-
, and amphiregulin, all of which are erbB ligands. Cocultures of airway epithelia with primary fibroblasts increased epithelial height comparable to that achieved following heregulin-
treatment. These data show that erbB2 stimulation is required for maintaining epithelial differentiation. Furthermore, the mesenchyme underlying the airway epithelium secretes a variety of erbB ligands that may direct various pathways of epithelial differentiation. trastuzumab; fibroblasts
|
into the apical airway surface liquid and restricts erbB receptors to the basolateral membrane. Compromising epithelial integrity results in immediate receptor-ligand interactions that enhance wound repair. ErbB receptors belong to a family of tyrosine kinase receptors (6) whose prototype is epithelial growth factor receptor (EGFR; also known as erbB1). These receptors are implicated in a variety of cancers and are associated with poor prognoses, which has led to molecular targeting of erbB receptors in anti-cancer therapy (10, 21, 22). Trastuzumab, a humanized anti-erbB2 antibody, accelerates receptor internalization and degradation (5, 9) and has become the standard of care in patients with increased erbB2 breast cancer (3, 8, 11, 12, 19). Importantly, 4.7% of patients experience cardiac dysfunction (27); however, the benefits from trastuzumab outweigh this potential risk (29). Although the mechanism of cardiac toxicity remains unclear, it has been hypothesized that erbB2 may be required for cardiac myocyte regeneration following injury.
This case presentation suggests that blocking erbB2 alters airway epithelial differentiation. If trastuzumab plays a causal role in dedifferentiating the patients airway epithelium, heregulin-
and/or other erbB ligands may preserve differentiation. Thus we hypothesized that a basolateral source of erbB ligands sustains a basal state of differentiation; disruption of signaling by trastuzumab dedifferentiates the epithelium.
| METHODS |
|---|
|
|
|---|
Cocultures of primary human airway epithelia and normal human lung fibroblasts were grown on 6.5-mm clear Transwells (Corning Costar, Corning, NY). One day before seeding, the epithelial sides of Transwells were coated with human placental collagen type IV (Sigma, St. Louis, MO) as previously described (7, 15, 17). The next day, the collagen was aspirated and Transwells were washed twice with phosphate-buffered saline. On the day of seeding, a mixture of 80% Vitrogen (bovine dermal collagen; Cohesion Technologies, Palo Alto, CA), 10% Dulbeccos modified Eagles (DME)/F-12 medium, and 10% 0.1 N NaOH was titrated to a pH of 7.2. Passage 5 normal human lung fibroblasts (Cambrex, East Rutherford, NJ) were added to this working Vitrogen solution at a density of 2.03.8 x 105 fibroblasts per milliliter of gel. The underside of the Transwells was fitted with a plastic insert ring attachment holder (Jims Instruments, Iowa City, IA) onto which aliquots of fibroblasts and Vitrogen mixture were seeded. Inserts were maintained upside down in a 37°C incubator without CO2 for at least 1 h to allow the Vitrogen matrix to gel. Once gelled, Transwells were flipped right side up and primary human airway epithelia were seeded on top. Cocultures were initially maintained in 5% fetal calf serum in DME/F-12. One day postseeding, coculture medium was switched to 2% Ultroser G serum substitute (USG)-2% fetal calf serum (needed for fibroblasts) in DME/F-12 and continuously maintained in this medium; 2% serum added to USG had no appreciable effect on epithelial differentiation.
Primary cultures were treated with recombinant human heregulin-
(R&D Systems) or trastuzumab (Genentech) added to the medium. Control cultures were not treated.
Morphometry. Hematoxylin and eosin-stained epithelial sections were chosen at random, and photographs of different regions were randomly selected for analysis. Periodic acid-Schiff (PAS)-stained sections for some epithelial cultures were also analyzed, and the number of goblet cells counted in PAS- vs. hematoxylin and eosin-stained sections correlated. However, not all cultures were PAS stained. The photographs were enlarged uniformly, overlaid with transparent grids, and used for morphometric analysis (31). The volume densities of the filter and tissue were determined using point counting; the length of the filter was measured using an eyepiece measuring grid. Two independent observers blinded to experimental conditions made all morphometric measurements. Two donors were analyzed per condition, and six Transwells per donor were treated; four to six sections per Transwell were evaluated. The height of the airway epithelia was normalized to the height of the filter to correct for possible variability in the sectioning of the epithelia (the filters are 20 µm thick). Airway epithelial cell number was normalized per length of the filter and area of the epithelia. We trained two independent observers to identify the following cell types: basal, ciliated, goblet, and columnar nonciliated. Dysplastic cells were identified as those cells that did not resemble any of the above-mentioned cell types. The correlation coefficients between observers for height, area, and cell density were 0.99, 0.99, and 0.98, respectively. The correlation coefficients for basal, ciliated, goblet, and nonciliated columnar cells were 0.90, 0.87, 0.97, and 0.99, respectively.
Western blot analysis. Lysate was made by incubation in lysis buffer (50 mM Tris·HCl, pH 7.5, 150 mM NaCl, 5 mM EDTA, 2 mM Na3V04, 10 mM sodium pyrophosphate, 100 mM NaF, PMSF, leupeptin, pepstatin, and aprotonin) with 1% Triton X-100 at 4°C. Lysate was collected and homogenized, membranes were pelleted, and the soluble fraction was incubated with antibody for erbB2 (DAKO). Immunoprecipitated protein was complexed with protein G-Sepharose (Pierce), and the complex was pelleted and washed. Protein was eluted by incubation in 2x sample buffer (4% SDS, 100 mM DTT, 20% glycerol, 0.005% bromphenol blue, and 0.065 M Tris, pH 6.8), boiled, separated by SDS-PAGE, and transferred to polyvinylidene difluoride (PVDF) membrane (Immobilon-P; Millipore). PVDF membrane was blocked in 5% BSA, washed, and incubated with anti-phosphotyrosine antibody (Upstate Biotech). After washes, incubation in horseradish peroxidase-conjugated secondary antibody (1:10,000; Amersham Pharmacia) allowed for bound antibody detection with SuperSignal solution (Pierce) and exposure to film (X-Omat AR; Kodak Scientific Imaging Film).
RT-PCR of ErbB ligands.
Fibroblasts embedded in Vitrogen matrix were pulled away from Transwells and incubated in 1% collagenase (Sigma) for 30 min at 37°C to digest the matrix. Fibroblasts were then harvested by centrifugation and washed with phosphate-buffered saline. RNA isolation was performed using the RNAqueous-4PCR kit (Ambion, Austin, TX) per manufacturers instructions. RNA was reverse transcribed with Thermoscript RT (Invitrogen, Carlsbad, CA). Controls included the omission of RT. The generated cDNA served as the template in PCR reactions using erbB ligand-specific forward and reverse primers for betacellulin, TGF-
, heparin-binding epithelial growth factor (HB-EGF), epiregulin, heregulin-
, and amphiregulin. The forward and reverse primers used were as follows: betacellulin, GAA TAT GTC CCT GGG TGT GG and CTA CAA GGC AGG ACA CC; TGF-
, TAG GCA TTT CAG GCC AAA TC and GCA TTT CCT CAC ATA AGG AGT TT; HB-EGF, CTC TCC CTG CCA AGT CTC AG and TGA ACC AGG TTT GGA AAT ACA; epiregulin, CCA AGG ACG GAA AAT GCT TA and AAA GGG CAT AGT GCT TGC AT; heregulin-
, ATA TCC ACC ACT GGG ACA AGC CAT and TAT CAC GGG TGG AGA CAT TTC CGA; and amphiregulin, TGG ATT GGA CCT CAA TGA CA and CGT TCA CCG AAA TAT TCT TGC. After heat denaturation, DNA was amplified in an Eppendorf Mastercycler thermocycler (Westbury, NY) as follows: 94°C for 30 s, 56°C for 30 s, and 72°C for 60 s for 40 cycles, followed by a 72°C postdwell for 2 min. Amplified DNA was separated by agarose gel electrophoresis and imaged (Polaroid, Cambridge, MA) under UV light.
| RESULTS |
|---|
|
|
|---|
|
stimulates human airway epithelia differentiation.
Heregulin-
and erbB2 are important in epithelial differentiation and wound repair (30). The current data further suggest that when erbB2 is blocked by trastuzumab, the epithelium dedifferentiates. Whether trastuzumab treatment alters erbB2 receptor mass or leads to loss of receptor signaling without changing receptor mass is not clear from these studies. These data suggested that an erbB ligand, heregulin-
, might have the opposite effect of trastuzumab, which is to stimulate epithelial differentiation. To quantitatively examine the effects of heregulin-
on human airway epithelia, we added increasing concentrations of recombinant human heregulin-
to the basolateral medium. Figure 3A shows examples of hematoxylin and eosin-stained sections. In contrast to trastuzumab, heregulin-
treatment increased epithelial height, a characteristic of well-differentiated airway epithelia (24), and decreased the number of cells per area (Fig. 3, B and C). Interestingly, heregulin-
treatment appeared to make the cilia longer; however, this effect was not quantified.
|
stimulates differentiation of airway epithelial cell types.
We asked whether heregulin-
treatment had an effect on the cell types present in the epithelia. We found that the percentage of ciliated and basal cells was not affected (Fig. 3, D and G). However, there was a significant increase in the number of goblet cells and a decrease in the number of nonciliated columnar cells (Fig. 3, E and F). Interestingly, the number of metaplastic cells decreased at the highest dose of heregulin-
tested (Fig. 3H).
Lung fibroblasts express erbB ligands and stimulate human airway epithelial differentiation.
Because the airway epithelium dedifferentiates when exposed to trastuzumab, erbB receptor stimulation may be required to maintain a basal level of differentiation in the epithelium. In vitro, erbB ligands may leak from the apical side of the epithelia or some ligands may be provided by the USG in the medium. However, in vivo, subepithelial lung fibroblasts are a likely source of ligand. Their localization basal to the epithelia would be optimal for direct and constitutive epithelial erbB receptor stimulation. Therefore, pulmonary fibroblasts were assayed for expression of erbB ligands by RT-PCR. Human lung fibroblasts express TGF-
, HB-EGF, epiregulin, heregulin-
, and amphiregulin, all erbB ligands (16) (Fig. 4A). Betacellulin was the only ligand tested that was not expressed. We then asked whether pulmonary fibroblasts could alter the differentiation of human airway epithelia in a manner similar to that of basolateral recombinant heregulin-
treatment. Qualitative changes are shown in epithelial morphology in cocultures (Fig. 4C) compared with untreated primary cultures of epithelia grown in isolation (Fig. 4B). These changes resemble those of adding recombinant heregulin-
to epithelial cultures in isolation (Fig. 3A). Epithelia grown in isolation reach a height of 33.7 µm with a standard deviation of 3.7 µm. However, when grown in the presence of pulmonary fibroblasts, epithelial height increased to 43.2 ± 8.8 µm, comparable to the epithelial height achieved after basolateral treatment with 3 nM heregulin-
(45.1 ± 4.9 µm). Figure 4D demonstrates Western blot analysis of erbB2 receptor phosphorylation. In primary epithelial cultures, erbB2 is not phosphorylated. As expected, however, basolateral treatment of primary epithelial cultures with recombinant heregulin-
as well as epithelia grown in coculture with fibroblasts both resulted in erbB2 receptor phosphorylation. In addition, when epithelial-fibroblast cocultures were treated with trastuzumab, erbB receptor phosphorylation was inhibited (data not shown).
|
| DISCUSSION |
|---|
|
|
|---|
ErbB receptors belong to a family of four related tyrosine kinase receptors, the prototype of which is EGFR, also known as erbB1. The extracellular domain of these receptors can bind ligands, resulting in a conformational change allowing for receptor homo- or heterodimer formation. Differential ligand affinity for monomers and the resulting receptor dimer combinations allow for a complex and promiscuous signaling capacity. Two of these monomers are uniquely handicapped: erbB2 is an orphan receptor (ligand for erbB2 has not yet been identified), and erbB3 has a dead kinase. However, these two monomers can conveniently come together and complement each others limitations. ErbB3 binds heregulin, resulting in dimerization with erbB2, which, once engaged, autophosphorylates and transphosphorylates erbB3. ErbB2 also can dimerize with erbB1 and erbB4, resulting in complex, coordinated, and specific signaling cascades in response to all known erbB ligands (EGF, amphiregulin, TGF-
, betacellulin, HB-EGF, and epiregulin).
In an in vitro model of human airway epithelia, incubation with trastuzumab led to epithelial dedifferentiation. The number of flat, metaplastic cells increased, a finding that closely resembled the patients airway biopsies. Although not tested in this study, trastuzumab treatment also may result in cell death, a possibility that cannot be excluded. Incubation with heregulin-
increased epithelial cell height and decreased the number of metaplastic cells. Interestingly, heregulin-
treatment also increased the number of goblet cells, whereas trastuzumab did not alter the goblet cell population. These data suggest that increasing concentrations of heregulin-
induce transdifferentiation of nonciliated columnar cells into goblet cells. Transdifferentiation is well documented (4). In addition, this paradox suggests that whereas heregulin-
stimulates epithelial differentiation, increasing goblet cell numbers, other ligands acting via erbB2-containing heterodimers must regulate other differentiation lineages. Alternatively, altering erbB2 activation may result in Clara cell and goblet cell differentiation, rather than global epithelial dedifferentiation; the present study cannot distinguish between the two possibilities. Interestingly, whereas our data suggest that erbB2 activation stimulates both cellular proliferation and differentiation, others have demonstrated that activation of erbB1 leads to goblet cell hyperplasia without altering the total number of epithelial cells (28). These data suggest a role for erbB1 in differentiation but not in proliferation. Similarly, Shimizu et al. (25) found that inhaling endotoxin led to an increase in the goblet cell population without changing the total cell number. Together, these data suggest that multiple ligands activate a variety of erbB receptors and that this complexity likely maintains both epithelial cell numbers and differentiation status, factors critical for normal airway barrier function. This also is consistent with our data showing that pulmonary fibroblasts express all but one of the erbB ligands and that human airway epithelia express all four erbB receptors (1, 2, 20, 30). Importantly, Western blot analysis (Fig. 4D) demonstrated that epithelial-fibroblast cocultures resulted in erbB2 receptor activation. Such interactions between epithelia and subepithelial cells are not new. In fact, epithelial/mesenchymal communications have been widely studied, and their roles in development and carcinogenesis have been recognized for decades (13, 14, 18).
Despite being cultured in the absence of fibroblast, the in vitro model of human airway epithelia appears morphologically differentiated. One explanation for this is that apical ligand may constitutively diffuse basolaterally at low levels. Moreover, normal turnover of cells may allow airway surface liquid ligands to diffuse into the basolateral compartment and stimulate cellular differentiation in a precise temporal and localized fashion. Finally, an additional artificial source of ligand may be the serum substitute (Ultroser G) present in the medium. Ultroser G is a largely undefined serum substitute that Widdicombe and colleagues (23, 32) have shown to promote airway epithelial differentiation.
The data presented suggest a central but complex role for erbB receptors and ligands for epithelial differentiation. They further suggest that mesenchymal-epithelial communication modulates this process. Our initial studies focused on only one ligand, heregulin-
, and serve as proof of a principle from which to model future studies aimed at unraveling this complexity.
| GRANTS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
N. T. N. Trinh, A. Prive, L. Kheir, J.-C. Bourret, T. Hijazi, M. G. Amraei, J. Noel, and E. Brochiero Involvement of KATP and KvLQT1 K+ channels in EGF-stimulated alveolar epithelial cell repair processes Am J Physiol Lung Cell Mol Physiol, October 1, 2007; 293(4): L870 - L882. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. M. Myerburg, J. D. Latoche, E. E. McKenna, L. P. Stabile, J. S. Siegfried, C. A. Feghali-Bostwick, and J. M. Pilewski Hepatocyte growth factor and other fibroblast secretions modulate the phenotype of human bronchial epithelial cells Am J Physiol Lung Cell Mol Physiol, June 1, 2007; 292(6): L1352 - L1360. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. LeSimple, I. van Seuningen, M.-P. Buisine, M.-C. Copin, M. Hinz, W. Hoffmann, R. Hajj, S. L. Brody, C. Coraux, and E. Puchelle Trefoil Factor Family 3 Peptide Promotes Human Airway Epithelial Ciliated Cell Differentiation Am. J. Respir. Cell Mol. Biol., March 1, 2007; 36(3): 296 - 303. [Abstract] [Full Text] [PDF] |
||||
![]() |
E. Puchelle, J.-M. Zahm, J.-M. Tournier, and C. Coraux Airway Epithelial Repair, Regeneration, and Remodeling after Injury in Chronic Obstructive Pulmonary Disease Proceedings of the ATS, November 1, 2006; 3(8): 726 - 733. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |