Edited by: Maria Caterina Mione, University of Trento, Italy
Reviewed by: Poongodi Geetha-Loganathan, SUNY Oswego, United States; Liping Xiao, UCONN Health, United States
This article was submitted to Stem Cell Research, a section of the journal Frontiers in Genetics
†These authors have contributed equally to this work
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
This study demonstrates that FGF10/FGFR2b signaling on distal epithelial progenitor cells, via ß-catenin/EP300, controls, through a comprehensive set of developmental genes, morphogenesis, and differentiation. Fibroblast growth factor (FGF) 10 signaling through FGF receptor 2b (FGFR2b) is mandatory during early lung development as the deletion of either the ligand or the receptor leads to lung agenesis. However, this drastic phenotype previously hampered characterization of the primary biological activities, immediate downstream targets and mechanisms of action. Through the use of a dominant negative transgenic mouse model (
In mice, the first morphological evidence of lung development is seen at embryonic day (E) 9.5 with the budding of the ventral foregut endoderm, forming the tracheal primordium ventrally and the esophagus dorsally. Concomitantly, distal to the tracheal primordium, two primary lung buds form, initiating the early stages of pseudoglandular development (E9.5-E12.5) [for reviews on early lung development, see (Warburton et al.,
Branching morphogenesis and epithelial differentiation depend on poorly understood cross-talk among a number of signaling pathways, involving fibroblast growth factors (FGF), sonic hedgehog (SHH), bone morphogenic proteins (BMP), and wingless/integrase 1 (WNT) ligands (El Agha and Bellusci,
Almost twenty years after the discovery of FGF10 as a key growth factor regulating branching morphogenesis, the primary targets and biological activities controlled by FGF10 are still unclear (El Agha and Bellusci,
Genetically modified mouse strains, based on the reverse tetracycline transactivator (rtTA) system, do exist to conditionally inhibit FGF ligand activity at the protein level, and employ a dominant negative soluble form of FGFR2b. Soluble FGFR2b is a hybrid protein, where the extracellular part of the receptor, responsible for binding FGF ligands, is fused to the heavy chain of mouse immunoglobulin (Celli et al.,
Little research has explicitly used soluble FGFR2b to study the role of FGF signaling on early lung development, while the papers that do exist only touch on the question tangentially. For example, Hokuto et al. (
This study demonstrates that FGF10/FGFR2b signaling on distal epithelial progenitor cells via ß-catenin/EP300 controls, through a comprehensive set of developmental genes, cell adhesion, and differentiation. Altogether, our results clarify the role of FGF10 on tip epithelial progenitor cells during development. Our transcriptomic approach has also provided a valuable dataset for future mechanistic studies aiming to characterize the role of newly found players in FGF10 signaling. Such knowledge will be instrumental to better understanding the role of FGF10 signaling at later stages of lung development, as well as during the repair process after injury.
First, at different stages during embryonic lung development, we monitored by qPCR the expression of
Expression of genes encoding the main FGFR2b ligands during early lung development and impact of FGFR2b ligand inactivation on branching morphogenesis
Therefore, our results indicate that
We performed
In conclusion, our detailed analysis reveals that subtle branching defects were already apparent 3 h after exposure to Dox. The major impact of inhibiting FGF10 activity was on the epithelium, where a complete arrest in budding, and a transient retraction of the epithelium (which correlated with an increase in the distance between the mesothelium and the distal tip epithelium), was observed.
Next, we analyzed the branching defects at the cellular level using 3D-reconstructions of serial confocal images of distal epithelial buds in control and experimental lungs. These lungs were isolated 9 h following a Dox-IP to pregnant females carrying E12.5 embryos, and were whole-mount stained with CDH1 (E-cadherin) antibody. Figure
Inhibition of FGF10 activity for 9 h in E12.5 lungs leads to collapse of the epithelial bud associated with cell rearrangements and altered cell-cell adhesion
Analysis of proliferation and cell death in control and experimental lungs
We also analyzed the appearance of epithelial tip cells in control and experimental lungs by transmission electron microscopy (TEM; Figure
Our results demonstrate that inhibition of FGF10 signaling leads to impaired distal bud morphology, including collapsed bud lumens and thicker epithelial layers. It is likely that this phenotype is primarily a result of epithelial disorganization caused by adhesion and rearrangement defects.
Based on the previous results, we selected E12.5 as the ideal time point to determine the specific transcriptomic targets of FGF10
Identification of early FGF10 target genes by a gene array approach
Figure
The Late 4 cluster displays supplementary genes linked to epithelial differentiation, such as
In order to gain more insight into the genes identified in each of the four groups, we made use of the online expression-profiling database “Genepaint.org” to identify the expression domain of the genes found in the different groups (Figures S3–S10). In addition, we carried out a gene array between isolated distal tip epithelial and mesenchymal cells of E12.5 wild type lungs (
We then compared the differential expression of the Early 4 and Late 4 genes (FGF10 signature genes), determined from the epithelium vs. mesenchyme gene array, with their expression after FGF10 inhibition (Figures
A similar analysis was performed for the Late 3 and Late 1 groups (Figures
Next, we evaluated the expression of the major transcription factors expressed in the lung at E12.5. A previous report indicated that out of 1100 transcription factors analyzed (covering 90% of all transcription factors encoded in the mouse genome), only 62 exhibited localized expression in the epithelium and/or mesenchyme of the developing lung (Herriges et al.,
Identification of the transcription factors regulated by FGF10 and impacts on epithelial differentiation
Next, we examined the impact of attenuated FGFR2b signaling on the differentiation of the multipotent epithelial progenitor cells. Figure
Until recently, the close examination of epithelial tip cell differentiation was limited, as only a few signature genes denoting differentiation status were known to be expressed in those cells. However, this limitation has been overcome after a paradigm-shifting paper published by Treutlein et al. (
Our KEGG analysis indicated that WNT signaling was also significantly regulated by FGF10 inhibition (Figure
FGF10 activity is primarily mediated through ß-catenin/EP300
Given that WNT/ß-catenin signaling is vital for proper branching morphogenesis and cellular differentiation (for a review see De Langhe and Reynolds,
Figures
We also compared, by qPCR, the expression levels of a number of genes found in the “FGF10 gene signature” between experimental and control lung explants (
In the attempt to produce results more comparable to the 9 h
IQ1 treatment for 9 h phenocopies 9 h FGF10 inhibition
In summary, our data suggest that a large proportion of the regulation by FGF10 signaling of epithelial branching morphogenesis and differentiation is mediated specifically through ß-catenin/EP300 transcriptional activity.
In this paper, we report the impacts on the epithelial tip cells of E12.5 lungs by blocking FGFR2b ligands, primarily FGF10. Both
Graphical summary: FGF10 signaling leads to phosphorylation of Y734 on FGFR2 associated with Sh3bp4 recruitment (Francavilla et al.,
One of the major limitations of our
To control for the potential secondary effects of our model, we validated the location of FGF10's primary targets by a gene array comparing expression of genes in the epithelium vs. mesenchyme of E12.5 wild type lungs (Figure
Furthermore, we assessed the well-established FGF10-SHH regulatory feedback loop during lung development as a means of validating our array (Figures
From our gene array data, we identified an “FGF10 gene signature.” These genes, primarily enriched in the epithelium, show decreased expression shortly after FGF10 inhibition, and continue to decrease during inhibition; therefore, these genes likely represent primary targets of FGF10, and are potential key mediators of FGF10/FGFR2b signaling.
We also found that FGF10 regulates many previously identified lung specific transcription factors (Herriges et al.,
We propose that the comprehensive set of target genes and transcription factors identified in our study is a valuable resource for future investigations on early lung branching morphogenesis and differentiation.
Sustained SOX9 expression in the tip epithelium of the developing lung has been associated with epithelial stem cell self-renewal. The current model predicts that individual tip cells, under the influence of FGF10, are prone to remain in the tip domain; as these cells divide, some of the daughter cells acquire bronchial progenitor characteristics associated with the exit from the tip domain.
The transcription factor SOX9 has been extensively studied during early lung development (Perl et al.,
As SOX9 is lost in the distal epithelial cells of experimental lungs, the expression of SOX2 in these cells increases, further suggesting these cells are losing their multipotency, and are adopting a proximal fate. This idea is supported by the evidence, at this stage, of a loss of the AT2 signature in the multipotent progenitors upon FGF10 inhibition.
Taken together, our data suggest that the loss of SOX9, downstream of FGF10 signaling, affects the morphogenesis and multipotent potential of distal epithelial cells.
The importance of ß-catenin signaling during pseudoglandular branching morphogenesis has been extensively studied (reviewed in De Langhe and Reynolds,
An intriguing possibility concerning available cytoplasmic ß-catenin is the relationship between ß-catenin and CTNND2, which regulate E-cadherin stability at the adherens junction complex of epithelial cells. While ß-catenin stabilizes E-cadherin to promote cell adhesion, CTNND2 leads to E-cadherin destabilization, the release of ß-catenin to the cytoplasm, and cell motility (Lu et al.,
Our data support the importance of the proposed CTNND2/CDH1/CTNNB1 regulatory axis. Contained in the FGF10 gene signature is the drastic reduction of
In the nucleus, ß-catenin associates with a number of transcriptional co-activators to regulate gene expression via the TCF/LEF family of DNA-bound transcription factors (reviewed in Hoppler and Kavanagh,
In summary, the
In conclusion, we have carried out a comprehensive analysis of FGF10/FGFR2b signaling on epithelial tip progenitor cells during E12.5 mouse lung development. This analysis revealed a new “FGF10 transcriptomic signature” which will be instrumental in designing new mechanistic studies concerning the role of FGF10 in alveolar epithelium formation during development, as well as maintenance during homeostasis and repair after injury.
Further information and requests for resources and reagents should be directed to, and will be fulfilled by, the corresponding author, Dr. Saverio Bellusci (
Animal experiments were performed at Children's Hospital Los Angeles under the research protocols (31–08 and 31–11) approved by the Animal Research Committee and in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The approval identification for Children's Hospital Los Angeles is AAALAC A3276-01. Harvesting organs and tissues from wild type and mutant mice following euthanasia using pentobarbital was approved at Justus Liebig University Giessen by the federal authorities for animal research of the Regierungspraesidium Giessen, Hessen, Germany (Approved Protocol GI 20/10 Nr. G 84/2016).
All mice used to generate experimental and control embryos were housed in a specific pathogen free (SPF) environment with free access to food and water. Up to five females were housed together, while males were housed singly. Females between 9 and 12 weeks of age were used to generate embryos.
Timed-pregnant females were used to conduct
Embryonic lungs used for
Embryonic lungs were dissected and cultured on 13 mm Whatman Track-Etch polycarbonate membranes, with 8.0 μm pores (Merck, Darmstadt, Germany) positioned atop DMEM culture medium in a 24-well culture dish [medium contained: Dulbecco's Modified Eagle Medium (1x DMEM), supplemented with D-Glucose, L-Glutamine, HEPES, Pyruvate, and Phenol red (Gibco, Paisley, UK), 10% fetal bovine serum (FBS), 1% penicillin (10,000 units/ml)-streptomycin (10 mg/ml)]. Lungs were incubated at 5% CO2 and 37°C for ~45 min to allow them to settle. At the beginning of the experiment (
Brightfield images of lungs from
To assess the expression of genes in the epithelial and mesenchymal compartments of distal E12.5 lung tips, C57BL/6 wild type embryos were used. Embryonic lungs were isolated in culture medium, and the distal epithelial buds, along with the surrounding mesenchyme and mesothelium, were carefully dissected with fine-tipped pincers. The dissected tips were then immediately transferred to 500 μl undiluted dispase (Corning, Amsterdam, The Netherlands) where they were incubated for 20–30 min on ice. The dispase-digested samples were then transferred to pure FBS, and incubated for 15 min on ice, thus blocking the enzymatic activity of dispase. Using tungsten microdissection needles, the epithelium was gently separated from the surrounding mesenchyme. Separated tissues were then prepared for total RNA extraction and microarray analysis.
Freshly dissected E12.5 lungs were washed in sterile PBS (2 × 5 min), fixed in 4% PFA for 20 min on ice, and then washed again (3 × 5 min). Lungs were dehydrated by successive washes in a graded ethanol series (30, 50, 70, 100, 100%) for 5 min each, and stored in 100% ethanol at −20°C.
To embed the lungs in paraffin, they were first washed in Xylol (2 × 5 min, or until clear), incubated for 1 h at 60°C in a 1:1 Xylol/paraffin mixture, washed in pure paraffin (3 × 20 min) at 60°C, and then stored in pure paraffin overnight at 60°C. Lungs were then embedded in paraffin blocks and sectioned to a thickness of 4 or 5 μm. Sections were placed in a 40°C water bath for ~30 min, and then placed on glass slides and incubated at 37°C overnight.
Before antibody staining, sections were first washed with gentle shaking in Xylol (3 × 10 min), and then in serial dilutions of ethanol (100, 70, 50, and 30%) for 3 min each, and finally in distilled water for 5 min. For each stain an antigen retrieval step was performed, which involved incubating the slides in 75–90°C citrate-based antigen unmasking solution (pH 6.0; Vector Laboratories, Peterborough, UK) for 15 min and then cooling on ice for ~30 min. Sections were then washed with PBST (1x PBS + 0.1% TWEEN20; 3 × 5 min). Blocking solution (1x PBS + 3% BSA + 0.4% TritonX) was then added atop each section for 1 h at room temperature. Primary antibodies were added to incubation buffer (1x PBS + 1.5% BSA + 0.2% TritonX) and samples were incubated overnight at 4°C (anti-SOX2, anti-Phospho-ßCatenin (Ser552), and anti-LEF1 were added at 1:100 dilution; anti-CDH1, anti-LAMA1, and anti-SOX9 were added at 1:200 dilution; see Table
Sections were imaged on a Leica DM 5500B upright fluorescent microscope system, with a DFC 360FX camera, and Leica Application Suite Advanced Fluorescence imaging software. Signal intensity was optimized to either a control or experimental sample for an experiment, and the acquisition and intensity values were similarly applied to each sample in that experiment, thus ensuring valid comparisons.
To assess the morphology of intact distal lung buds in control and experimental embryos, whole mount immunofluorescence followed by confocal imaging was performed.
E12.5 lungs were dissected and fixed in 4% PFA for 20 min on ice. Samples were washed in PBS + 1% TritonX (3 × 10 min), and incubated in blocking buffer (1x PBS + 1% TritonX + 10% FBS) for 1.5 h at room temperature, followed by two washes in blocking buffer. Samples were then incubated for 2 h with FITC-conjugated anti-CDH1 (Dilution: 1:200) diluted in 1/4 blocking buffer and PBS, at 4°C in the dark. Lungs were then washed in PBS (3 × 10 min), and transferred to custom made imaging dishes (composed of a 35,0/10 mm glass bottom cell culture dish (Greiner Bio-One, Frickenhausen, Germany) and a 10,0/1 mm rubber washer fixed to the middle of the dish with a suitable adhesive, thus creating an ideal well to mount and image the sample). ProLong Gold antifade reagent with DAPI was added to each well and covered with a glass coverslip.
Z-stacks of distal lung buds were obtained on a Leica TCS SP5 confocal microscopy system using Leica Application Suite X software. For each bud, the first optical section of the z-stack was acquired at the basal edge of the epithelium. Z-stack images were taken at 0.5 μm increments through the bud, until imaging was no longer possible due to complete loss of signal intensity. Compensation of intensity loss through the bud was obtained using the linear compensation by AOTF option. 3-D reconstructions and movies were created using Leica Application Suite X software.
To identify the effects of FGFR2b ligand inhibition on the ultrastructure of distal epithelial and mesenchymal cells, control and experimental E12.5 lungs were prepared for transmission electron microscopy.
The chest cavities of freshly harvested E12.5 embryos were gently opened by incising from the lower abdomen, through the sternum, to just under the chin using a pair of fine dissection scissors. An incision was made along the diaphragm from the midline to the spine. The embryos were then immediately placed, ventral side up, in an immersion fixative solution containing 4% PFA + 2% sucrose + 0.05% calcium chloride + 1x PIPES buffer (0.1M, pH 7.4; Sigma-Aldrich, Taufkirchen, Germany) in a 50 ml Falcon tube, such that each sample was immersed in 5X the volume of fixative in its own tube. Tubes were placed on ice and gently shaken for 2 h, after which the fixative was removed and replaced by 4% PFA + 0.05% glutaraldehyde (GA; Sigma-Aldrich, Taufkirchen, Germany), 2% sucrose, 0.05% calcium chloride + 1X PIPES buffer (0.1 M, pH 7.4). Samples remained in this fixative overnight at 4°C.
The next morning, the samples were processed for routine transmission electron microscopy. Briefly, the fixed lungs were dissected and placed into molten agar, which was allowed to harden before the samples were cut longitudinally in half. Each half was fixed for 30 min in 1.5% GA fixative containing 2% sucrose + 0.05% calcium chloride + 1X PIPES buffer (0.1 M, pH 7.4). The fixative solution was then removed and samples were washed with 1X PIPES (3 × 5 min). Samples were then incubated for 1 h at room temperature in reduced osmium fixation solution containing 0.15% sodium hexacyanoferrate(II) and 2% reduced osmium, then washed very briefly with distilled water, and dehydrated via washes in a graded ethanol series (70, 80, 90, 100%), 3 times 10 min each step. Samples were then embedded by immersion in propylene oxide (3 × 5 min), in 1:1 propylene oxide:Agar 100 epoxy resin (1 × 30 min) following the manufacturer's instructions to produce blocks of medium hardness (Agar Scientific, Essex, UK), and finally in pure Agar 100 epoxy resin in a desiccation chamber at room temperature overnight.
The Agar 100 resin-penetrated lungs were then flat embedded into fresh Agar 100 resin and polymerized at 60°C for at least 2 days, or until complete polymerization was achieved. Ultrathin sections were then prepared and micrographs were obtained using a Zeiss LEO 906 transmission electron microscope equipped with a TRS slow-scan 2K CCD camera and ImageSP software.
DNA was isolated from the tails and hind limbs of E12.5 embryos. Gene-specific primers were used to detect the presence of
Whole embryonic lungs or separated epithelium and mesenchyme used for total RNA isolation were first put in 700 μl QIAzol Lysis Reagent (Qiagen, Hilden, Germany). For tissue disruption and homogenization, the samples were transferred to gentleMACS M Tubes and homogenized in a gentleMACS Dissociator (Miltenyi Biotec) for 1 min. Total RNA was then isolated using the miRNeasy Mini Kit (Qiagen, Hilden, Germany), and eluted in 30 μl RNase-free water. RNA amount and purity was assessed with a NanoDrop 2000c (Thermo Scientific). Up to 1 μg of total RNA for each sample was then reverse transcribed using the QuantiTect Reverse Transcription Kit (Qiagen, Hilden, Germany).
Primers were designed to amplify specific mature mRNAs using NCBI's primer-BLAST option (
Differential gene expression was investigated using microarray analysis. Depending on the amount of RNA isolated per sample in an experiment, one of two possible microarray protocols was used. For RNA concentrations >50 ng/μl, the T7-protocol was followed. In this protocol, purified total RNA was amplified and Cy3-labeled using the LIRAK kit (Agilent Technologies, Waldbronn, Germany) following the kit instructions. Per reaction, 200 ng of total RNA was used. The Cy3-labeled aRNA was hybridized overnight to 8 × 60 K 60mer oligonucleotide spotted microarray slides (Agilent Technologies, design ID 028005).
For experiments where samples yielded <50 ng/μl of RNA, the SPIA-protocol was utilized. In this protocol, purified total RNA was amplified using the Ovation PicoSL WTA System V2 kit (NuGEN Technologies, Leek, The Netherlands). Per sample, 2 μg amplified cDNA was Cy-labeled using the SureTag DNA labeling kit (Agilent Technologies). The Cy3-labeled aRNA was hybridized overnight to 8 × 60 K 60mer oligonucleotide spotted microarray slides (Agilent Technologies, design ID 074809).
For each protocol, hybridization, and subsequent washing and drying of the slides were performed following the Agilent hybridization protocol. The dried slides were scanned at 2 μm/pixel resolution using the InnoScan is900 (Innopsys). Image analysis was performed with Mapix 6.5.0 software, and calculated values for all spots were saved as GenePix results files. Stored data were evaluated using the R software (version 3.3.2) and the limma package (version 3.30.13) from BioConductor. Gene annotation was supplemented by NCBI gene IDs via biomaRt (last accessed 08–03–2018).
Proliferation in E12.5 lungs was assessed using the Click-iT EdU Imaging Kit (Invitrogen, Schwerte, Germany). 5-ethynyl-2′-deoxyuridine (EdU), a nucleoside analog of thymidine incorporated into DNA during DNA synthesis, was injected (i.p.) 2 h before pregnant females were sacrificed (Dosage: 0.005 mg EdU / g mouse weight). Embryonic lungs were harvested, paraffin embedded, sectioned and placed on glass slides. Sections were then deparaffinised and stained for EdU according to the manufacturer's protocol.
Apoptosis was assessed using the TdT-mediated dUTP Nick-End Labeling (TUNEL) assay. The assay was performed using the DeadEnd Fluorometric TUNEL System (Promega, Mannheim, Germany). E12.5 lungs were harvested, paraffin embedded, sectioned and placed on glass slides. The assay was performed according to the manufacturer's protocol.
To assess the expression patterns of genes in early stage embryonic lungs (E14.5), the online database genepaint.org was used (last accessed 01–08–2018). Each of the genes significantly regulated in our
Using still images of lungs, mesothelium and airways were traced in Adobe Illustrator CS6 (version 16.0.4) to create skeletal outlines. These outlines were exported and lengths and areas were quantified either using MetaMorph (version 1.5.0) or FIJI (version 2.0.0-rc-68/1.52g) software.
Significance was determined by unpaired two-tailed Student's
ΔCt and ΔΔCt values were calculated according to the following formulas:
Note, this equation accounts for the fact that Ct is proportional to the –log of gene expression. ΔCt is therefore positively related to the expression of the gene of interest.
Unpaired two-tailed Student's
Mean spot signals were background corrected with an offset of 1 using the NormExp procedure on the negative control spots. The logarithms of the background-corrected values were quantile-normalized. The normalized values were then averaged for replicate spots per array. From different probes addressing the same NCBI gene ID, the probe showing the maximum average signal intensity over the samples was used in subsequent analyses. Genes were ranked for differential expression using an unpaired two-tailed Student's
SB, MJ, JZ, PM, and CC: concept and design; MJ, SaD, AL, SoD, JW, GC, and RM: acquisition of data; SB, MJ, JW, EB-V, and DA: analysis and interpretation; MJ, SaD, SB, and JZ: drafting and editing of the manuscript. All authors read and approved the final manuscript.
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to thank Kerstin Goth and Jana Rostkovius for managing and genotyping the mice used for experiments. We also thank Dr. Elie El Agha for the critical reading of this manuscript.
The Supplementary Material for this article can be found online at:
Experimental validation of the
Transmission electron microscopy: Compared to controls, experimental lungs (DoxIP + 9 h) show reduced numbers and stunted microvilli (see black arrows, a–d), opened tight junctions (see black arrows in e and g; white asterisks in f and h), and flattened Golgi with increased staining (see black arrows, i–l). epith. = epithelium.
Genes and Expression pattern found in the Early 1 cluster
Genes and Expression pattern found in the Early 2 cluster
Genes and Expression pattern found in the Early 3 cluster
Genes and Expression pattern found in the Early 4 cluster
Genes and Expression pattern found in the Late 1 cluster
Genes and Expression pattern found in the Late 2 cluster
Genes and Expression pattern found in the Late 3 cluster
Genes and Expression pattern found in the Late 4 cluster
Relative level of expression of the genes of interest in epithelium and mesenchyme of WT E12.5 lungs and regulation of gene expression upon FGF10 inhibition
Primary antibodies.
Primer sequences for qRT-PCR.
Longitudinal section of control bud.
Longitudinal section of experimental bud (9 h Dox-IP).
Cross section of control bud.
Cross section of experimental bud (9 h Dox-IP).