Airway branching morphogenesis in three dimensional culture
© Franzdóttir et al. 2010
Received: 29 June 2010
Accepted: 25 November 2010
Published: 25 November 2010
Skip to main content
© Franzdóttir et al. 2010
Received: 29 June 2010
Accepted: 25 November 2010
Published: 25 November 2010
Lungs develop from the fetal digestive tract where epithelium invades the vascular rich stroma in a process called branching morphogenesis. In organogenesis, endothelial cells have been shown to be important for morphogenesis and the maintenance of organ structure. The aim of this study was to recapitulate human lung morphogenesis in vitro by establishing a three dimensional (3D) co-culture model where lung epithelial cells were cultured in endothelial-rich stroma.
We used a human bronchial epithelial cell line (VA10) recently developed in our laboratory. This cell line cell line maintains a predominant basal cell phenotype, expressing p63 and other basal markers such as cytokeratin-5 and -14. Here, we cultured VA10 with human umbilical vein endothelial cells (HUVECs), to mimic the close interaction between these cell types during lung development. Morphogenesis and differentiation was monitored by phase contrast microscopy, immunostainings and confocal imaging.
We found that in co-culture with endothelial cells, the VA10 cells generated bronchioalveolar like structures, suggesting that lung epithelial branching is facilitated by the presence of endothelial cells. The VA10 derived epithelial structures display various complex patterns of branching and show partial alveolar type-II differentiation with pro-Surfactant-C expression. The epithelial origin of the branching VA10 colonies was confirmed by immunostaining. These bronchioalveolar-like structures were polarized with respect to integrin expression at the cell-matrix interface. The endothelial-induced branching was mediated by soluble factors. Furthermore, fibroblast growth factor receptor-2 (FGFR-2) and sprouty-2 were expressed at the growing tips of the branching structures and the branching was inhibited by the FGFR-small molecule inhibitor SU5402.
In this study we show that a human lung epithelial cell line can be induced by endothelial cells to form branching bronchioalveolar-like structures in 3-D culture. This novel model of human airway morphogenesis can be used to study critical events in human lung development and suggests a supportive role for the endothelium in promoting branching of airway epithelium.
Lung development and critical aspects of pulmonary epithelial differentiation is mostly studied through the use of animal models. Due to a lack of good experimental in vitro models, much less is known about development and stem cell biology in human lungs. While many different human airway epithelial cell lines capture the phenotypic traits of the proximal airways such as trachea and large bronchi [2–4], there is lack of cell lines that mimic normal histological features of the lung, such as branching morphogenesis of the distal airways. Furthermore, there are inherent differences in the cellular composition of the airway epithelium between rodents and humans. In the rodent, basal cells, candidate airway epithelial stem cells, are confined to the trachea, while in the human lung basal cells are present throughout the upper airways, and all the way down to small bronchioles [5–7]. This supports the importance of generating models of human airway development and differentiation to study the cell biology of the human lung including epithelial stromal interactions and branching morphogenesis.
Although many human airway epithelial cell lines have been established, most of them have not been defined with respect to their cellular origin and lack critical characterization in terms of expression of differentiation markers. The most cited airway epithelial cell line, A549, is derived from a human bronchioalveolar carcinoma . Despite its origin in malignant tissue it has been widely used to study lung biology. The human bronchial cell lines 16HBE14o-, Calu-3, and BEAS-2B have been successfully applied to study drug transport, metabolism, and drug delivery due to their ability to form tight junctions (TJ) . The Calu-3  and 16HBE14o cell lines have been identified as the most differentiated cell lines available and have been used to study bronchial epithelial integrity including barrier function and the activity of tight junctions complexes .
In order to mimic the airway epithelial lining, primary human bronchial epithelial cells have been studied under various conditions. When primary human epithelial cells are cultured at the air-liquid interface using serum containing differentiation media, they undergo terminal squamous differentiation instead of forming a pseudostratified polarized and ciliated epithelial layer . However, under the same conditions fibroblasts and fibroblast secretions have been shown to stimulate the formation of a pseudostratified ciliated epithelium . This highlights the importance of the bi-directional communication between the epithelial and stromal cellular compartments. Recently, human alveolar type II cells were shown to form cysts in 3D culture through a novel mechanism of epithelial morphogenesis relying on aggregation and rearrangement . In this model of terminal airway cyst formation using Matrigel based 3-D culture conditions, no branching morphogenesis occurred.
Most studies on epithelial-mesenchymal interactions have focused on fibroblasts and components of the extracellular matrix. Less is known about the role of the vascular endothelium and its interaction with epithelial cells. Mouse culture models, e.g. lung tissue explants have though shown a critical role for the interaction between the vascular and epithelial compartments during lung development . Interestingly, it has been shown that endothelium-derived factors are necessary for distal lung morphogenesis and function in mice, including the formation of alveoli  and the maintenance of alveolar integrity . Furthermore, in vitro models suggest that endothelial cells support airway epithelial tight junction formation . Recent data from other organs such as the liver, pancreas, brain and bone marrow indicate that organ specific endothelial cells are of major importance for fate control of stem cells as well as for organogenesis and tissue maintenance . Better understanding of the heterotypic crosstalk between the epithelium and the surrounding stroma is important to understand such important events as lung development, lung carcinogenesis and repair of the airway epithelium following injury.
In this paper, we describe a 3D epithelial culture model using a recently described human basal-like airway epithelial cell line (VA10) cultured in reconstituted basement membrane (rBM) . VA10 cells form a pseudostratified layer in air-liquid culture and have been used to study airway epithelial defense mechanisms, including tight junction function and the production of antimicrobial peptides [17–19]. In contrast, when cultured in rBM matrix, a condition more favorable for distal lung morphogenesis, VA10 cells form spheres with a clear apical-basal polarity and tight intercellular junctions as shown by the expression of β4-Integrin on the outer (basal) surface, and claudin-1 laterally, but lack branching morphogenesis . VA10 has a phenotype similar to human airway epithelial basal cells, including the expression of p63 and cytokeratin 14 . Airway epithelial basal cells both self-renew and generate luminal daughter-cells in a sphere-forming assay, and have thus been suggested to be candidate airway epithelial stem cells . This suggests that VA10 might be an ideal cell line to use in developmental and morphogenesis studies of human lung differentiation. To mimic heterotypic cellular interactions we have co-cultured the VA10 cells with vascular endothelial cells in a 3D environment, and under these conditions we are able to show marked branching morphogenesis and a differentiation profile of the epithelial cells towards an alveolar epithelial phenotype. Such branching morphogenesis is novel in a human in vitro model. An in vitro cellular model of lung development and differentiation can serve as an important platform to study critical events, such as cellular interaction and cell signaling in lung morphogenesis. Furthermore, our data support a critical interaction between the vascular and epithelial compartments during epithelial branching morphogenesis and differentiation.
The bronchial epithelial cell line VA10 was previously established by retroviral transduction of primary bronchial epithelial cells with E6 and E7 viral oncogenes (). The cells were cultured in bronchial epithelial growth medium, BEGM (Lonza, Walkersville, MD) supplemented with 50 IU/ml penicillin and 50 μg/ml streptomycin (Gibco, Burlington, Canada).
The human lung adenocarcinoma derived alveolar epithelial cell line A549 (American Type Culture Collection, Rockville MA) was cultured in DMEM-Ham's-F12 basal medium supplemented with 10% fetal bovine serum (FBS), 50 IU/ml penicillin and 50 μg/ml streptomycin (Gibco).
Primary human umbilical vein endothelial cells were cultured on T75 tissue culture flasks (Becton Dickinson) on endothelial medium EGM-2 supplemented with 50 UI/ml penicillin, 50 μg/ml streptomycin (Gibco) and 30% FBS for first passage cells or 5-10% FBS after first passage. Endothelial cells were only used up to passage 8.
For 3D culture experiments growth factor reduced reconstituted basement membrane (Matrigel, BD Biosciences, Bedford, MA) was used. Cells were seeded into 300 μl of Matrigel in its liquid state, plated into 24-well culture dishes and allowed to gelatinize at 37°C for 30 minutes before adding 1 ml of culture medium. For co-culture experiments 2.5 × 106 HUVEC cells and 500-1000 VA10 (or A549) cells were seeded into the Matrigel to ensure clonal growth and reduce cell aggregation of epithelial cells. No clumping of cells was observed before seeding. In this experimental setup the endothelial cells remain non-proliferating while marked proliferation is seen in the epithelial cells. For direct staining of branching structures in Matrigel we used 8-well chamber slides with 100 μl Matrigel, 8.3 × 105 HUVEC cells and 333 VA10 cells. For transwell (TW) experiments, HUVEC cells were seeded onto 0,4 μm polyester filter inserts (Corning, MA) and allowed to reach 70% confluence before onset of experiment. 333 VA10 cells were seeded into 100 μl Matrigel in 24-well plates and allowed to gelatinize before the TW filters and culture medium were added to the wells. For FGFR inhibition, the pan-FGFR inhibitor SU5402 was diluted to the indicated concentration from a 100 mM DMSO stock solution in the culture medium. The medium was replaced three times a week. The corresponding concentration of DMSO was used in the control medium.
Immunofluorescent staining of 3D gels was carried out in the culture chamber as previously described . The gels were washed twice with PBS and fixed in methanol (10 minutes at -20°C). For some primary antibodies we used double acetone fixation with ice cold acetone for ten minutes at 4°C followed by drying and repeated acetone treatment. After fixation the gels were washed three times for 15 minutes at room temperature with 100 mM Glycine in PBS, and blocked with 10% goat serum in IF-buffer (0.2% Triton X-100; 0.1% BSA and 0.05% Tween-20 in PBS). To block unspecific binding to the mouse-derived rBM gel, the IF-buffer was supplemented with 1% goat anti mouse immunoglobulin G for 20 minutes. Primary antibodies were incubated overnight at 4°C, followed by three 25 minute washes in 10% Goat Serum in PBS. The secondary antibodies were incubated for 2 hours at RT or over night at 4°C. After PBS rinsing, nuclear staining was performed with TO-PRO-3(r) (Invitrogen, Carlsbad, CA) for 30 minutes followed by 3 × 15 minute washes. The chambers were removed after staining and the samples were embedded in Fluoromount-G (Southern Biotech, Birmingham, AL) for microscopic analysis.
For freeze sections, the gels were flash-frozen in n-hexan and transferred to -80°C. The gels were mounted in Tissue-Tek(r) O.C.T. compound (Sakura Finetek, Zoeterwoude, Netherlands) and sectioned in a cryostat (5-20 μm sections). The samples were dried and then fixed and stained as above, with antibody incubations for 30 minutes at RT.
The following antibodies were used: Rabbit polyclonal antibodies to pro-SPC (AB3786) (Abcam, Cambridge, MA). Mouse monoclonal antibodies to: CD31 (JC/70A), Cytokeratin 17 (E3) (DAKO, Glostrup, Denmark); Cytokeratin 14 (LL02), FGFR2 (Abcam, Cambridge, MA); β4-Integrin (3E1) (Chemicon, Temecula, CA); p63 (7JUL), TTF-1 (SPT24) (Novocastra Laboratories, Newcastle upon Tyne, UK); E-Cadherin (HECD1) (Zymed, South San Francisco, CA).
Isotype specific secondary antibody conjugates Alexa fluor(r) (Alexa fluor, 488 (green), 546 (red), Invitrogen) were used for immunofluorescence experiments with TO-PRO-3(r) (Invitrogen) for nucleic acid staining. Antibody incubations were carried out for 30 minutes at room temperature with secondary antibodies and nucleic stain in darkness. Specimens were rinsed twice for 5 minutes at room temperature between antibody and nucleic stain incubations. For maximum preservation of the fluorescent signal from the samples after staining, specimens were mounted using Fluoromount-G (Southern Biotech, Birmingham, AL) and images visualized with confocal microscope.
Immunofluorescence was visualized and captured using laser scanning Zeiss LSM 5 Pascal Confocal Microscope (Carl Zeiss AG, Munich, Germany). Bright-field and phase-contrast images of Matrigel cultures were captured using a Leica DFC320 digital camera attached to a Leitz Fluovert microscope (Wetzlar, Germany)
We also tested if the widely used human alveolar epithelial cell line A549 could generate branching colonies in rBM. When cultured alone, A549 cells formed irregular colonies and no phenotypic changes occurred in co-culture with endothelial cells, indicating lack of ability to generate branching structures (figure 1D). Furthermore, the endothelial cells did not increase the number of colonies or degree of branching (figure 2B).
To study the nature of the endothelial-induced branching, we set up a transwell co-culture system where the endothelial cells were cultured on top of the filter situated above the 3D gels containing VA10 cells. When endothelial cells were present on the filters a marked stimulation was seen in the total number of colonies seen in the 3D culture, and again the branching phenotype was dependent upon their presence (Figure 2E). This suggests soluble, endothelial derived factors as the mediators of inducing the branching morphogenic phenotype. We also removed the endothelial filters at a time before the earliest signs of branching (day 6), and in this setup, the VA10 cells still showed marked branching suggesting the endothelial cells are inducing a phenotypic switch at an early stage in the 3D colonies (Figure 2E).
In vitro models are important supplements to animal models for studying cellular interactions and basic developmental processes, repair and carcinogenesis. In this paper we describe a novel in vitro model of branching morphogenesis in the human lung, using an airway epithelial cell line in co-culture with endothelial cells, which captures critical aspects of distal lung morphogenesis. We found that in this system branching morphogenesis is depended on an interaction between airway epithelia and the vascular endothelium.
The stroma is known to play a major role in organogenesis and maintenance of tissue structure and function in epithelial based organs such as the mammary gland , prostate and lung[14, 30]. As the lung initially develops from the fetal foregut it is highly dependent on crosstalk between the endodermal cells and the underlying stroma. The outgrowth and branching of epithelial cells is stimulated by various stromal-derived growth factors, especially fibroblast growth factors (FGFs) acting through FGF-receptors on the invading epithelial cells . Studies on mouse embryos and fetal lung tissue explants have shown the stroma to play a critical role in the invasion and branching of the airway epithelium during mouse lung development [32, 33]; however, these studies have not directly identified the cellular components of the stroma that mediates these effects. Our model using a bronchial-derived epithelial cell line with a basal-like phenotype suggests that endothelial cells could be critical mediators in stimulating epithelial invasion and branching behavior through secretion of soluble factors.
Branching morphogenesis is one of the key developmental processes during lung development. A recent seminal paper studying branching morphogenesis of the mouse lung indicated that this process is based on a pattern of three simple branching modes, repeated in different order throughout lung development. The authors proposed that these simple branching modes were controlled genetically through master regulators . They suggested that the sprouty gene family might be among the developmental switches or regulators of this process, specifically sprouty-2. In our model, both FGFR2 and Sprouty-2 are highly expressed at the growing buds of the branching structures. In the mouse embryonic lung, Sprouty-2 appears to be dynamically expressed in the peripheral endoderm and down-regulated in the clefts between new branches. Furthermore, over-expression of Sprouty-2 in the peripheral airway epithelium in vivo results in diminished branching . The ability of VA10 cells to form bronchioalveolar-like structures in co-culture with endothelial cells opens the possibility to study important aspects of human lung morphogenesis, such as the spatial and temporal function of FGF signaling and Sprouty, in a well controlled in vitro system. Furthermore, initial data supports a role for FGFR signaling in our model, although we cannot conclude that the main endothelial derived branching factor is an FGF growth factor given the abundance of FGF in both serum and the growth medium.
In the adult lung, the stroma changes substantially from the proximal conducting part to the distal alveolar part. In the larger bronchi, cartilage tissue supports the epithelial tissue. Fibroblasts, smooth muscle cells, and large vessels are also prominent in the proximal part. In the distal bronchioalveolar zone the cartilage has disappeared and microvessels are prominent, especially surrounding the alveoli. Recent data from various organs such as the liver, pancreas, brain and bone marrow indicate the importance of endothelial cells in fate control of stem cells and for organogenesis and tissue maintenance . Lammert et al. have shown that endothelial cells are important for both pancreas and liver organogenesis . Similarly, endothelial cells were shown to be vital components of the stem cell niche in the nervous system , and in hematopoiesis . It is thus becoming evident that endothelial cells are important regulators of stem cells in many organs and a crucial component for cell fate decisions and tissue morphogenesis.
Much less is known about somatic stem cells of the lung compared to many other epithelial organs, such as gut, breast and prostate. One of the reasons is the histological difference between the proximal and distal part of the lung and the cellular complexity found within the airway epithelium. Candidate stem cell types include basal cells, Clara cells and alveolar type II cells. Of these, basal cells are postulated to be the most pluripotent candidate stem cells . Thus, the basal cell characteristics of VA10 make it an interesting candidate for testing stem cell properties. Indeed, this cell line can form a pseudostratified layer with mature TJs under air-liquid interphase conditions , suggesting that it differentiates significantly depending on environmental conditions. In addition we show that the VA10 cell line forms branching epithelial structures reminiscent of distal bronchioalveolar-like structures with partial alveolar type II differentiation, yet retaining a basal cell phenotype.
In summary, we have described a novel cell culture model that captures critical aspects of human airway branching morphogenesis. This model provides a unique tool to study and characterize various processes of human lung development, morphogenesis and to address the heterotypic interaction between epithelial and stromal components during airway differentiation and tubular branching morphogenesis. Furthermore, our results identify endothelial cells as potential inducers of distal airway epithelial differentiation.
fetal bovine serum
fibroblast growth factor receptor-2
fibroblast growth factors
human umbilical vein endothelial cells
reconstituted basement membrane
Thyroid transcription factor 1
Grant support was provided by the Icelandic Research Council, Landspitali University Hospital Research Fund, University of Iceland Research Fund, Science and Technology Policy Council-Thematic program in postgenomic biomedicine. European Science Foundation (EuroCORES program, EuroSTELLS).
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.