Ager-CreER^T2: A New Genetic Tool for Studying Lung Alveolar Development, Homeostasis, and Repair

Abstract

The alveolar region of the lung is composed of two major epithelial cell types: cuboidal alveolar type 2 cells (AT2 cells), which produce surfactant proteins, and large, thin, alveolar type 1 cells (AT1 cells), specialized for efficient gas exchange. AT1 cells cover more than 95% of the alveolar surface and constitute a major barrier to the entry of pathogenic agents. Relatively few genetic tools are available for studying the development of AT1 cells, the function of genes expressed in them, and the effect of specifically killing them in vivo in the adult lung. One distinguishing feature of AT1 cells is the high level of expression of the gene Ager, encoding the advanced glycation endproduct–specific receptor, a member of the immunoglobulin superfamily of cell surface receptors. In this paper, we report the generation of a novel Ager-CreER^T2 allele in which Cre recombinase is inserted into the first coding exon of the endogenous gene. After treatment with tamoxifen the allele enables Ager⁺ progenitor cells to be efficiently lineage labeled during late embryonic development and AT1 cells to be killed in the adult lung using a Rosa26-diphtheria toxin A allele. Significantly, adult mice in which approximately 50% of the AT1 cells are killed survive the loss; repair is associated with increased proliferation of SFTPC⁺ (surfactant protein C–positive) AT2 cells and the upregulation of Ager expression. The Ager-CreER^T2 allele thus expands the repertoire of genetic tools for studying AT1 turnover, physiology, and repair.

The distal region of the mammalian lung is composed of millions of tiny air sacs, known as alveoli, that are made up of epithelial cells, capillary vasculature, stromal cells, and resident immune cells. There are two epithelial cell types: cuboidal alveolar type 2 cells (AT2 cells), specialized for the production, storage, and secretion of surfactant proteins; and large, thin, alveolar type 1 cells (AT1 cells) that are the major sites of gas exchange between air and blood (1–6). Each AT1 cell has an area of around 5,100 μm², and together they cover approximately 95% of the surface of the distal lung. They are connected to each other, and to AT2 cells, by apical tight junctions, and the basal surface lies on a basal lamina, which is fused with that of the closely apposed capillary endothelial cells. Reconstruction of tissue sections shows that individual AT1 cells can have a quite complex “branched” three-dimensional structure, with cytoplasmic “plates” extending between adjacent alveoli (3). The cytoplasm itself contains relatively few specialized organelles apart from microvesicles involved in transcellular transport.

Given that AT1 cells provide the major air–blood barrier of the distal lung, it is important that they are replaced quickly if they are damaged, such as by exposure to toxic agents. New AT1 cells also have to be generated during the compensatory regrowth and alveologenesis that follows the surgical removal of a lung lobe by pneumonectomy (2, 7, 8). In all injury/repair models that have been studied, mature AT1 cells do not self-renew but are replaced by the proliferation, differentiation, and spreading of AT2 cells that, as a population, function as alveolar stem cells (9–11). There is, however, evidence that during remodeling after pneumonectomy, a small percentage of AT1 cells, which may be immature, can give rise to AT2 cells (2, 7).

To better understand the biology, function, and regeneration of AT1 cells, researchers have used a number of strategies to purify the cells from the postnatal lung and to catalog genes preferentially expressed in them compared with other lung epithelial cell types (2, 12). For example, Hopx (HOP homeobox) is a transcription factor expressed in all AT1 cells and podoplanin (also known as T1α), Ager (advanced glycation endproduct receptor), and Igfbp2 (insulin growth factor–binding protein) are membrane-associated proteins preferentially expressed on the cell surface. Ager belongs to the immunoglobulin superfamily of transmembrane receptors and has a broad repertoire of ligands, including proteins that accumulate extracellularly during inflammation, aging, and infection (13). Ager lies upstream of NF-κB as well as mitogen-activated protein kinases, and there is some evidence that it promotes cell adhesion and spreading of AT1 cells, but its precise in vivo function is still not fully understood; several isoforms are generated by proteolytic cleavage and alternative splicing, including a soluble form that may function as a decoy receptor. Ager-null mice are fully viable but show physiological defects related to inflammation and vascular function (13).

Experimental insights into the molecular mechanisms driving lung regeneration and repair in vivo depend to a large degree on the availability of genetic tools for manipulating genes in specific cell types. Given that there may be different subpopulations of AT2 and AT1 cells expressing different amounts or subsets of marker genes (2, 10, 11), it is advantageous for the field to have available a “toolkit” of different reporter and Cre recombinase alleles driven by different endogenous genes. In previous studies on the role of AT1 cells in lung biology, researchers have used Hopx-ERCre (7) and HopX-CreER and Igfbp-CreER alleles (2). In this paper, we report the generation of a new Ager-CreER “knock-in” allele that is very efficient in driving recombination in Ager-expressing cells after administration of tamoxifen (Tmx). The allele can be used to lineage label Ager⁺ progenitor cells in the embryonic lung and to kill adult AT1 cells after recombination of Rosa26-DTA. Surprisingly, even after death of about 50% of the AT1 cells, adult mice are viable, and lung architecture is minimally affected, likely owing to the rapid replacement of AT1 cells by the proliferation and differentiation of AT2 cells.

Methods

Mice

To generate the Ager-CreER^T2 knock-in allele, the previously described strategy was used (14). Briefly, 8 kb of 5′ untranslated region and exons 1–7 was retrieved from a BAC (bacterial artificial chromosome) clone (bMQ174; Source BioScience) and recombined into the vector pL25B. A CreER^T2-polyA cassette and an FRT-flanked neomycin resistance cassette were recombined into the start codon to obtain the targeting construct. The construct was then electroporated into G4 (C57BL/6Ncr × 129S6/SvEvTac) hybrid embryonic stem cells. CreER^T2 recombination into the Ager locus was confirmed using Southern blot analysis. Four correct clones were injected into C57BL/6 blastocysts. To remove the neomycin resistance cassette, mice were bred to 129S4-Gt(ROSA)26Sor^tm2(FLP*)Sor/J. Ager-CreER^T2 mice were maintained on a C57BL/6J background. Rosa26-CAG-lsl-tdTomato (hereafter Rosa26-tdTm) (15) and B6.129P2-Gt(ROSA)26Sor^{tm1(DTA)Lky/J} (Rosa26-DTA; JAX 009669; The Jackson Laboratory) (16, 17) and Ager-H2B:Venus (in which H2B:Venus fusion protein is expressed under control of the endogenous Ager locus [14]) mice were maintained on a C57BL/6J background. All experiments were performed according to institutional animal care and use committee–approved protocols.

Tmx Administration

For embryonic experiments, pregnant female mice were gavaged with one dose of 50 or 100 μg/g body weight Tmx dissolved in corn oil. To label or kill adult AT1 cells, four doses of 200 μg/g body weight Tmx were given every other day, unless otherwise stated.

5-Ethynyl-2′-Deoxyuridine Labeling

Mice were injected intraperitoneally with 50 μg/g body weight 5-ethynyl-2′-deoxyuridine (EdU) dissolved in PBS, and lungs were fixed after 3 hours. EdU incorporation was detected using the Click-iT Plus EdU Alexa Fluor 647 kit (C10640; Thermo Fisher Scientific).

Histology and Immunofluorescence Analysis

Lungs were fixed and processed as described previously (14). For immunofluorescence analysis, tissue sections were blocked in 3% BSA, 10% donkey serum, and 0.1% Triton X-100 for 1 hour at room temperature. Primary antibodies were diluted in block solution and were applied and incubated overnight at 4°C. All fluorophore-conjugated secondary antibodies were diluted at 1:500 and incubated for 1 hour at room temperature. Primary antibodies used were as follows: CD11b (101201, 1:100 dilution; BioLegend), LAMP-3/CD208 (DDX0191, 1:200 dilution; Dendritics), GFP (GFP-1020, 1:500 dilution; Aves Labs), HOPX (sc-398703, 1:50 dilution; Santa Cruz Biotechnology), red fluorescent protein (600401379, 1:250 dilution; Rockland Immunochemicals), receptor for advanced glycation endproducts/AGER (MAB1179, 1:200 dilution; R&D Systems), and SFTPC (surfactant protein C) (ab3786, 1:500 dilution; Merck Millipore; and sc-7706, 1:100 dilution; Santa Cruz Biotechnology). Images were captured using AxioCam imager and LSM 710 and LSM 780 microscopes (Carl Zeiss Microscopy).

Quantification and Statistics

For quantification, one longitudinal section per sample was imaged and analyzed using ImageJ software (NIH) (n ≥ 3 animals/experiment). Prism software (GraphPad Software) was used to plot the quantification values. Statistical analysis was performed using unpaired two-tailed Student’s t tests between groups. Values on graphs are shown as mean ± SD.

qRT-PCR

Total RNA and cDNA were prepared as described previously (14). qPCR was performed using iQ SYBR Green Supermix (Bio-Rad Laboratories), and reactions were run on a StepOnePlus system (Applied Biosystems). Expression levels were normalized to Gapdh. Primer sequences were as follows: Ager forward: 5′-CCATCCAACTCCGAGTCAG-3′; Ager reverse: 5′-CACAGGTCAAGGTCACAGTC-3′; Hopx forward: 5′-GCTGTGCCTCATCGCAG-3′; Hopx reverse: 5′-CAAGCCTTCTGACCGCC-3′.

Results

Generation of Ager-CreER^T2 Mice

The generation of the Ager-CreER^T2 “knock-in” allele, which uses the ATG of the first coding exon of the endogenous gene, is shown in Figures 1A and 1B. Evidence that it is a null allele comes from immunohistochemical analysis of sections of adult wild-type, heterozygous, and homozygous mutant lungs (Figure 1C) and from qPCR analysis of RNA extracted from them (Figure 1D). Previous studies have shown that Ager-null mutant mice are viable (13). Likewise, homozygous Ager-CreER^T2 mice were apparently normal and fully fertile.

Figure 1.

Generation of Ager-CreER^T2 knock-in mice. (A) Schematic diagram of the strategy for generating the Ager-CreER^T2 allele. (B) Correct targeting of embryonic stem cells was confirmed using Southern blot analysis. AflII enzyme was used to generate 4.8-kb wild-type (WT) and 3.5-kb mutant DNA fragments. (C) Immunohistochemistry shows that the AT1 (alveolar type 1) markers AGER (advanced glycation endproduct–specific receptor) and HOPX (HOP homeobox) label Ager-CreER^T2 adult lung tissue. However, AGER is not detected in the homozygous Ager-CreER^T2/CreER^T2 lung. Scale bars = 100 μm. (D) qPCR analysis reveals that Ager transcript levels are significantly reduced in homozygous mouse lung, whereas levels of Hopx transcript are not affected. The mRNA expression levels of Ager-CreER^T2/+ and Ager-CreER^T2/CreER^T2 samples were normalized to Ager^+/+. n = 3 animals. Data shown as mean ± SD. *P < 0.05. HSV-TK = herpes simplex virus thymidine kinase; n.s. = not significant; polyA = polynucleotide adenylyltransferase; UTR = untranslated region.

Ager-CreER^T2 Drives Recombination in Ager⁺ Progenitor Cells of the Embryonic Mouse Lung

Previous studies have shown that Ager is expressed in epithelial progenitor cells in the developing mouse lung (7, 12, 18). Specifically, immunohistochemistry shows that at Embryonic Day (E) 15.5, it is coexpressed with HOPX in a population of cells in the “stalks” of distal tips (7). To test the ability of the Ager-CreER^T2 allele to drive recombination in the embryonic lung, we administered Tmx to pregnant Ager-CreER^T2; Rosa26-tdTm mice at E16.5 and harvested embryonic lungs at E17.5 and E18.5. As shown in Figures 2A and 2B, 18.32 ± 6.70% of AGER⁺ cells are lineage labeled at E17.5, whereas 52.37 ± 4.14% of HOPX⁺ cells are labeled at E18.5. The majority of cells that were lineage labeled with tdTomato are AT1 cells (HOPX⁺ or AGER⁺) (96.16 ± 1.03% at E17.5; 97.78 ± 1.16% at E18.5), and only a few Tomato⁺ cells express SFTPC (3.84 ± 1.03% at E17.5; 2.19 ± 1.13% at E18.5) (Figure 2B).

Figure 2.

Ager-CreER^T2; Rosa26-tdTm lineage labels Ager⁺ epithelial progenitors. (A) Ager-CreER^T2; Rosa26-tdTm embryos were exposed to tamoxifen (50 μg/g) at Embryonic Day (E)16.5, and lungs were fixed and analyzed by immunohistochemistry at E17.5 (n = 3). All panels are images from the same section that was stained with antibodies to RFP (red fluorescent protein), AGER, and SFTPC (surfactant protein C). Of AGER⁺ AT1 cells, 18.32 ± 6.70% were lineage labeled. Of the lineage-positive cells, 96.16 ± 1.03% are AGER⁺, and 3.84 ± 1.03% are SFTPC⁺. (B) A higher dose of tamoxifen (100 μg/g) was given at E16.5, and lungs were fixed and analyzed at E18.5. Of HOPX⁺ AT1 cells, 52.37 ± 4.14% were labeled. At this dose, the majority of Tomato⁺ (tdTomato-positive) cells are still HOPX⁺ (97.78 ± 1.16%). However, a few Tomato⁺ cells are SFTPC⁺ (2.19 ± 1.13%) (arrow). The two left panels are images of the same section stained with antibodies to RFP and HOPX. The far right panel is an adjacent section stained with antibodies to RFP and SFTPC. Scale bars = 100 μm.

Use of Ager-CreER^T2 to Lineage Label AT1 Cells in the Adult Lung

To establish the efficiency and specificity of the Ager-CreER^T2 allele in the adult mouse lung, we treated mice carrying the Rosa26-tdTm allele with different concentrations of Tmx (four doses of 10 μg/g and either one or four doses of 0.2 mg/g) (Figure 3). This resulted in labeling of 9.52 ± 0.98%, 21.85 ± 1.18%, and 56.87 ± 3.63% AT1 cells, respectively, as judged by coexpression of the lineage trace with the marker protein HOPX (Figures 3D and 3F–3H). Only a small percentage of SFTPC⁺ AT2 cells were labeled (1.35 ± 0.35%, 1.52 ± 0.06%, and 1.04 ± 0.57%, respectively). In control lungs not exposed to Tmx, only 0.12 ± 0.03% of AT1 cells were labeled (Figure 3E). Epithelial cells lining the bronchiolar airways and in the mesothelium were not labeled.

Figure 3.

Ager-CreER^T2 efficiently and specifically lineage labels adult AT1 (alveolar type 1) cells using the Rosa26-tdTm reporter. (A) Adult Ager-CreER^T2; Rosa26-tdTm mice (n = 3) were treated with four doses of tamoxifen (Tmx; 0.2 mg/g) to lineage label Ager⁺ cells. After 4 days, Tomato⁺ cells were widely distributed in the alveolar region but not in the bronchioles (br). (B) Higher magnification shows that tdTomato signal colocalizes with AGER. (C) A small percentage (1.04 ± 0.57%) of SFTPC⁺ AT2 cells were also labeled (arrow). The inset shows the boxed region with SFTPC signal only. (D and D′) Immunohistochemistry of the same section with antibodies to RFP/tdTomato and HOPX. Quantification based on colocalization of tdTomato signal with HOPX nuclear staining (arrowheads) shows that 56.87 ± 3.63% of HOPX⁺ AT1 cells were lineage labeled. All non–lineage-labeled HOPX⁺ cells are marked with arrows. (E–H) Lineage labeling with different concentrations of Tmx. (E) Of AT1 cells, 0.12 ± 0.03% were tdTomato⁺ in lungs without Tmx. (F) Four doses of 10 μg/g Tmx labels 9.52 ± 0.98% of AT1 cells. (G) One dose of 0.2 mg/g Tmx labels 21.85 ± 1.18% of AT1 cells. (H) Four doses of 0.2 mg/g Tmx label 56.87 ± 3.63% of AT1 cells. Scale bars = 100 μm in A and D, 20 μm in B, and 50 μm in C and E–H.

Use of Ager-CreER^T2 to Specifically Kill AT1 Cells In Vivo

In previous studies, we have successfully used a combination of Rosa26-DTA and Sftpc-CreER^T2 alleles to specifically ablate approximately 50% of the AT2 cells of the adult lung, with little effect on overall tissue architecture (18). After recombination, the Rosa26-DTA allele results in expression of the A subunit of diphtheria toxin, which kills cells even if they do not express diphtheria toxin surface receptor. To use the same strategy for killing AT1 cells, we bred mice carrying Ager-CreER^T2, Rosa26-tdTm and Rosa26-DTA alleles. As shown in Figure 4, 4 days after the administration of four doses of Tmx (0.2 mg/g), there are approximately 50% less HOPX⁺ AT1 cells in the lung than in lungs of control mice (Figures 4A and 4F), suggesting that approximately 50% of the AT1 cells were killed. Significantly, all of the mice (n = 4) survived the Tmx administration. In agreement with earlier studies in which about half of the AT2 cells were killed with DTA (18), there was no major alteration in tissue architecture associated with loss of AT1 cells, apart from the appearance of more CD11b⁺ cells (monocytes, neutrophils, and macrophages) in the alveoli (Figure 4B). Some lineage-labeled AT1 cells survived the treatment, presumably because they recombined the Rosa26-tdTm but not the Rosa26-DTA allele. Significantly, no EdU incorporation was seen in any of these AT1 cells 4 days after Tmx treatment. EdU incorporation was, however, seen in 21.15 ± 3.98% of the AT2 cells, suggesting that they proliferate in response to loss of AT1 cells (Figures 4D and 4F). In addition, some of the SFTPC⁺ AT2 cells show a more elongated morphology compared with their normal cuboidal shape seen in the control lung (Figure 4D, yellow arrowheads). To further investigate changes in AT2 cells after ablation of AT1 cells, we generated Ager-CreER^T2 mice carrying the Ager-H2b:Venus allele in which an H2b:Venus fusion protein is expressed in the nucleus under the control of the endogenous Ager locus (14). Ager-CreER^T2/H2b:Venus; Rosa26-DTA mice were treated with four doses of Tmx (0.2 mg/g), and their lungs were fixed and analyzed 4 days later. In the Ager-CreER^T2/H2b:Venus lungs, approximately 5% of SFTPC⁺ AT2 cells are Venus⁺. By contrast, approximately 50% of SFTPC⁺ cells are Venus⁺ in Ager-CreER^T2/H2b:Venus; Rosa26-DTA lung (Figure 4E).

Figure 4.

Specific in vivo ablation of AT1 cells. (A) Immunofluorescence reveals lineage-labeled HOPX⁺ AT1 cells in Ager-CreER^T2; Rosa26-tdTm (control) and Ager-CreER^T2; Rosa26-tdTm/Rosa26-DTA (DTA) (n = 4) lungs 4 days after Tmx treatment. Quantification in F shows that HOPX⁺ cells make up 8.35 ± 0.66% of total DAPI⁺ cells in the control lung, whereas this proportion is reduced to 3.78 ± 0.26% four days after Tmx treatment. By 14 days, the proportion is restored to normal (6.75 ± 1.15% in control and 6.53 ± 0.46% in DTA lungs). (B) Immunofluorescence shows increased numbers of CD11b⁺ cells after ablation of about half of the AT1 cells (4 d after last dose of Tmx). (C) Histochemistry also shows more immune cells present in the DTA lung at this time, but no grossly abnormal phenotype. (D and F) Control and DTA mice were exposed to 5-ethynyl-2′-deoxyuridine (EdU) for 3 hours before being killed, 4 days after last Tmx treatment. Boxed region in upper panels is shown at higher magnification below. In Ager-CreER^T2; Rosa26-tdTm control lungs (left panels), only about 1.53 ± 0.45% of AT2 cells proliferate (EdU⁺). By contrast, the number significantly increases to 21.15 ± 3.98% in Ager-CreER^T2; Rosa26-tdTm/Rosa26-DTA lungs (white arrowheads). SFTPC⁺ cells showing elongated morphology are labeled with yellow arrowheads. (E) Analysis of lungs from Ager-CreER^T2/H2b:Venus and Ager-CreER^T2/H2b:Venus; Rosa26-DTA mice 4 days after Tmx treatment reveals that 50.17 ± 2.13% of SFTPC⁺ cells express Ager-H2B:Venus following loss of AT1 cells, compared with 5.70 ± 1.87% in control. SFTPC⁺ Ager-H2B:Venus⁺ cells are labeled with white arrowheads. Scale bars = 100 μm in A and C and 50 μm in B, D, and E. Data shown are mean ± SD. *P < 0.05; **P < 0.01.

Discussion

In the present paper, we report the generation of a novel Ager-CreER^T2 “knock-in” allele that augments the repertoire of genetic tools available for studying AT1 development, function, and regeneration in the mouse lung. We have shown in the present study that this allele is efficient in lineage labeling Ager⁺ progenitor cells in the developing lung during the very earliest stages of alveologenesis (E16.5) (Figure 2). This result suggests that the allele can, in the future, be useful for following the fate of individual Ager⁺ cells, using low doses of Tmx and Rosa-Confetti lineage-tracing tools.

Previous studies have indicated that AT2 cells in the postnatal lung express low levels of Ager transcript, as judged by RNA-sequencing analysis of isolated cells (2, 12). However, we found that in the adult mouse, the Ager-CreER^T2 allele labels only very few AT2 cells, even after four doses of Tmx (Figure 3). This suggests that the Ager expression level in most AT2 cells is too low to drive sufficient amounts of CreER to support recombination of Rosa26-tdTm. Although it is possible that recombination will occur with other Rosa26 reporter alleles, we conclude that the Ager-CreER allele will likely be a useful tool for studying gene function specifically in AT1 cells of the adult lung. This is backed up by the finding that recombination of the Rosa26-tdTm allele is low in the absence of Tmx (Figure 3E), indicating that expression of Ager-CreER^T2 is not leaky.

A significant finding in this study is that no obvious pathology such as fibrosis was seen after 4 days or 2 weeks in the lungs of Ager-CreER^T2; Rosa26-tdTm/Rosa26-DTA lungs after ablating approximately 50% of the AT1 cells. This result suggests that the surviving AT1 cells and/or AT2 cells are able to rapidly spread to cover denuded basal lamina while AT2 cells proliferate and differentiate into AT1 cells to complete the repair. In our studies, we found evidence that approximately 21% of the AT2 cells proliferate after ablation of AT1 cells, whereas more of them upregulate expression of Ager, as judged by expression of an H2B:Venus fusion protein inserted into the Ager endogenous locus. In the future, it will be interesting to ask whether a specific subpopulation of AT2 cells preferentially responds to the loss of AT1 cells, as suggested by studies using other injury/repair models in the adult mouse (10, 11). In addition, the Ager-CreER^T2 allele will be a useful tool for testing the function of genes specifically expressed in AT1 cells, such as their potential roles in providing a major barrier between the alveolus and the vasculature, supporting the integrity of closely associated endothelial cells, and contributing to the niche of AT2 stem cells.