Conditionally induced RAGE expression by proximal airway epithelial cells in transgenic mice causes lung inflammation

Two transgenic lines of mice were generated and mated to create conditional doxycycline (dox)-inducible mice that overexpress RAGE (CCSP-RAGE TG) [24]. One mouse line specifically included a tetracycline-inducible RAGE construct and another utilized the club cell (Clara) secretory protein (CCSP) promoter successfully used to specifically target proximal airway epithelium [25]. At post-natal (PN) day 20, mice were weaned, genotyped as similarly outlined [26], and continuously fed dox (625 mg/kg; Harlan Teklad, Madison, WI) until their sacrifice date at PN 60. Single or non-transgenic mice used as controls were also fed dox. On the day animals were sacrificed, en bloc lungs were inflation-fixed with 4% paraformaldehyde for histology [27], lavaged for the assessment of bronchoalveolar lavage fluid (BALF) [28], or resected prior to the isolation of total protein/RNA [29]. Mice were housed and utilized in accordance with protocols approved by the IACUC at Brigham Young University.

In order to assess whether RAGE was effectively increased in the airways of CCSP-RAGE TG mice, quantitative real time RT-PCR (qPCR) and immunoblotting were performed for RAGE using primers, antibodies, and experimental conditions already described in detail [27]. Lungs from control and CCSP-RAGE TG mice were inflation fixed, processed, and sectioned as previously outlined [30]. Slides were stained with hemotoxylin and eosin (H&E, Thermo Scientific, Pittsburg, PA) using standard techniques. Immunohistochemical localization of RAGE was performed as summarized [7]. CCSP immunohistochemistry was performed using a rabbit polyclonal IgG at a concentration of 1:500 (Seven Hills BioReagents, Cincinnati, OH) and staining for FoxJ1 was performed with a mouse monoclonal IgG at 1:500 (Seven Hills BioReagents). A TUNEL TdT-FragEL DNA Fragmentation Detection Kit (Calbiochem, Rockland, MA) was used to evaluate apoptosis wherein TUNEL positive cells were counted by blinded individuals in high power fields prior to normalization to counts observed in control animals [31].

CCSP-RAGE TG and control mice were sacrificed and BALF was removed as described [27]. Supernatants were assayed for total protein with a bicinchoninic acid (BCA) total protein kit (Thermo Scientific). Total numbers of pelleted cells were counted with a hemocytometer and cells stained with a modified Wright-Giemsa stain (Diff-Quik; Baxter, McGaw Park, IL) were subjected to a blinded manual differential cell count in which 200 cells were counted per slide, and the percent of total cells was determined. Counting was performed in triplicate and the average was obtained.

Total RNA from CCSP-RAGE TG and control lungs (n =3 per group) was isolated using the Absolutely RNA Kit (Stratagene, Santa Clara, CA) and treated with DNAse. RNA was quantified and 1 μg of each sample was converted to cDNA. Cytokines were assessed using the Mouse Inflammatory cDNA Plate Array (Signosis, Sunnyvale, CA) and the cytokine concentrations were internally normalized to 18 s RNA as outlined in the instructions. Immunoblotting for TNF-α, IL-7, and IL-14 was also performed in order to confirm differential expression in CCSP-RAGE TG mouse lungs compared to controls. Blotting was performed using antibodies for TNF-α (sc-52746, Santa Cruz Biotechnology, Santa Cruz, CA), IL-7 (sc-7921, Santa Cruz Biotechnology), and IL-14 (sc-80994 Santa Cruz Biotechnology) with 20 μg protein precisely quantified using BCA quantification and using standard methods already outlined [32]. Membranes were incubated with appropriate secondary antibodies, detected with ECL-plus (Amersham, Piscataway, NJ) and developed. Band densitometry utilized digitized images and the Un-Scan-It software package (Silk Scientific, Orem, UT).

Values are expressed as mean ± SD obtained from at least three separate experiments in each group. Data were assessed by one- or two-way analysis of variance (ANOVA). When ANOVA indicated significant differences, the Student t-test was used with Bonferroni correction for multiple comparisons. Results presented are representative, and those with p-values <0.05 were considered significant.

Double transgenic offspring (CCSP-RAGE TG) from CCSP-rtTA and TetO-RAGE transgenic mice were obtained and dox-mediated up-regulation of RAGE commenced on PN20, a period that coincided with the completion of lung morphogenesis. In order to assess the effectiveness of conditional RAGE up-regulation, qPCR and immunoblotting experiments were conduced using total mRNA and whole lung lysates, respectively. qPCR revealed that CCSP-RAGE TG mice fed dox from PN20 through PN60 expressed significantly more RAGE mRNA than age matched single or non-transgenic controls (Figure 1A). Lungs exposed to dox for 40 days were also evaluated by immunoblotting and the results demonstrated that RAGE protein was up-regulated when compared to control mouse lungs (Figure 1B). Immunostaining for RAGE was necessary in order to localize RAGE overexpression. Immunohistochemical staining for RAGE revealed persistent basal expression in the distal lung compartment (Figure 1C and D, arrowheads). However, punctate RAGE up-regulation by airway epithelial cells (Figure 1D, arrows) was observed in CCSP-RAGE TG mice while RAGE expression was undetectable in airway epithelial cells of control lungs (Figure 1C, arrow).

Body weights of CCSP-RAGE TG animals did not significantly differ following dox administration when compared to dox-exposed control animals (not shown). Lung weights were also indistinguishable between the two groups of mice (not shown). Classic H&E staining of lung samples was completed and there were no appreciable histological differences between lungs from control animals (Figure 2A) and samples procured from CCSP-RAGE TG mice (Figure 2B) after 40 days of up-regulation. In particular, mean cord lengths and average airway wall thickness were measured [26] and these histological evaluations revealed insignificant differences in proximal airway number and appearance. Furthermore, parenchymal regions of the peripheral lung, including the alveolar compartment, were also normal in terms of size and appearance. In order to qualitatively evaluate the proximal airways, club (Clara) cells were characterized by immunostaining for CCSP. CCSP expression was not different when comparisons were made between control lung sections (Figure 2C) and CCSP-RAGE TG sections (Figure 2D). Ciliated pulmonary epithelium, the other predominant cell population in the conducting airways, was also immunohistochemically evaluated by staining for FoxJ1, a nuclear transcription factor that identifies ciliated pulmonary epithelial cells [33]. Similar to immunohistochemistry for CCSP, FoxJ1 abundance was not different when considering CCSP-RAGE TG lung sections and controls (not shown). Specialized stains for total collagen (Picro-sirius red) and proteoglycans (Periodic acid-Schiff) to identify mucus abundance revealed no differences between the groups (not shown). Even though cell-specific analyses did not suggest abnormal quantities, TUNEL staining was conduced to test whether turnover was affected. Representative staining identified sporadic apoptotic cells in lung parenchyma (not shown); however, counts revealed that despite a trend toward increased apoptosis in CCSP-RAGE TG mouse lungs, there was not a significant increase in cell death when comparing the two groups (Figure 3).

There was not a net increase in total protein abundance in BALF samples following 40 days of RAGE up-regulation in the proximal lung (Figure 4A). In order to assess the potential for leukocyte extravasation, total cell quantities in BALF samples were obtained prior to their morphological identification. The total number of cells in BALF from RAGE TG mice was significantly elevated when compared to controls (Figure 4B). When cells were anatomically identified, the percentage of polymorphonucleocytes (PMNs) was significantly elevated (Figure 4C), suggesting diapedesis of these cells is mediated, at least in part, by RAGE signaling. Eosinophils also trended higher in RAGE TG mice compared to controls; however, the average increase was just beyond significance (p =0.06). Altered leukocyte quantities provided the rationale for the subsequent evaluation of inflammatory cytokines implicated as participants in leukocyte chemoattraction. Characterization of mRNA isolated from mouse lung samples revealed that RAGE TG mice had significantly increased levels of TNF-α, IL-7, and IL-14 compared to control animals (Figure 5). We also discovered marginally increased expression of important Th2 related cytokines including IL-4, IL-6, and IL-13; however, such elevations in expression were not significant. In order to correlate mRNA expression with protein levels, immunoblotting for TNF-α, IL-7, and IL-14 was also completed. Compared to lung lysates from control animals, TNF-α, IL-7, and IL-14 were each up-regulated in lungs from CCSP-RAGE TG mice (Figure 6).