Published ahead of print on January 27, 2005, doi:10.1165/rcmb.2004-0343OC
© 2005 American Thoracic Society DOI: 10.1165/rcmb.2004-0343OC Normal Lung Development in RAIG1-Deficient Mice Despite Unique Lung Epithelium–Specific ExpressionDivision of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital at Harvard Medical School, Boston, Massachusetts Correspondence and requests for reprints should be addressed to Jingsong Xu, Ph.D., Instructor in Medicine, Harvard Medical School, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115. E-mail: jxu{at}rics.bwh.harvard.edu
RAIG1, 2, and 3 and GPCR5d represent a new subfamily of orphan G protein–coupled receptors. RAIG1 is expressed abundantly and specifically in the lung during development and in adult mice. During lung development, RAIG1 expression is initiated at E14.5 and gradually increases, reaching its highest levels at E18. High levels of expression are maintained in adult lungs. Given its abundant lung-specific expression and role in retinoic acid signaling, we hypothesized that RAIG1 plays a role in epithelial cell differentiation during lung development. To determine RAIG1 function and track endogenous RAIG1 spatial expression, a null allele of Raig1 was generated and the lacZ gene was "knocked-in." Although expression was detected in both proximal and distal epithelium during embryogenesis, it became restricted to type I and type II pneumocytes and the most distal bronchiolar cells in postnatal lungs. This is the first gene known to have this unique epithelial cell expression pattern. Despite this high level of expression, targeted inactivation of Raig1 did not cause significant developmental defects. Epithelial cell differentiation was normal and lung structure was intact. Analysis of other family members demonstrated some overlapping embryonic expression of RAIG3 mRNA that could have led to functional redundancy in the single RAIG1 null mutant mouse.
Key Words: RAIG1 • G protein-coupled receptors • retinoic acid
Retinoic acids (RAs) are a group of pleiotropic signaling molecules that regulate a broad range of physiologic and developmental processes. They elicit biological functions by interacting with two families of nuclear receptors, retinoic acid receptor (RAR) and retinoid X receptor (RXR), each consisting of three isoforms—  , ß, and  —encoded by independent genes. A functional RAR is a heteromeric dimer composed of one isofom from each of these two receptor families (1). Ligand binding is believed to induce translocation of the receptors to the nucleus, where they act as transcription factors for downstream target genes.
Several lines of investigation have highlighted the importance of RA signaling in lung development. First, mice deficient in both RAR RA-induced gene one (RAIG1) may be one of the downstream targets of RA signaling. It was first identified through differential display in an effort to identify genes that are regulated by RA in an oral carcinoma cell line (8). Strikingly, among all adult tissues surveyed, human RAIG1 was predominantly expressed in the lung. RAIG1 is also expressed in several lung cancer cell lines. Based on its seven-transmembrane motif and strong sequence homology with other type 3 G protein–coupled receptors (GPCRs), such as members of the metabotrophic glutamate receptor family, RAIG1 is categorized as a GPCR. Although the ligand for RAIG1 is not known, recently three other closely related GPCRs—RAIG2, RAIG3, and GPCR5 d—have been identified (9–12). Together with RAIG1, they form the C5 subfamily of GPCRs (13, 14). As its name implies, RA may regulate the expression of RAIG1, 2, and 3. However, little is known about the biochemistry of these receptors, their ligands, and the downstream signaling events to which they are coupled. To investigate the role of RAIG1 in lung biology, we cloned the murine RAIG1 gene, and conducted spatial and temporal analysis of its expression in both developing and adult lungs. We also generated "knock-in" and "knock-out" mice to more precisely define RAIG1 expression and function during lung development. Histologic and molecular analysis did not show apparent developmental defects in RAIG1-deficient (RAIG1–/–) lungs. We propose that there might be functional redundancy between RAIG1 and RAIG3 based on overlapping embryonic expression of RAIG3 in the developing lung.
Northern Blot Northern blot was performed as previously described (15). A multi-tissue Northern blot was obtained from Clontech (Palo Alto, CA). The RAIG1 probe contains sequences from codon 219–303 and was generated by polymerase chain reaction (PCR) using IMAGE clone 2,123,135 as a template. The forward primer sequence is TCCTCTCCATTGCCATCTGGGT, and the reverse primer sequence is GGGAGTAGGCTCTGTTCTCC.
Gene Targeting and Genotyping
ß-Galactosidase Staining
In Situ Hybridization
Morphology
Real-Time and Conventional RT-PCR
Mouse RAIG1 mRNA Expression To identify a mouse cDNA for RAIG1, the previously published cDNA sequence for human RAIG1 was used to query the Genebank database. An Expressed Sequence Tag (EST) (IMAGE clone # 2,123,135) was more homologous to previously published human RAIG1 cDNA than any other cDNA sequences in the database. It encodes a protein of 356 amino acids and shares 75% homology with human RAIG1, suggesting that it is the murine ortholog of the human RAIG1 gene. A 250-bp probe was generated by PCR and used for detecting mRNA expression in adult tissues on a multitissue Northern blot. Similar to human RAIG1, mouse RAIG1 was predominantly expressed in the lung (Figure 1A). However, in the mouse, no appreciable expression was detected in the kidney and gut, two sites of low-level expression of human RAIG1 (8). Also in contrast to humans, mice only have one 2.4-kb RAIG1 mRNA, whereas the human gene, through alternative use of polyadenylation sites, generates two isoforms of mRNA that are 2.4 and 4.8 kb.
Consistent with the tissue expression pattern, mouse RAIG1 is expressed in TMCC and MLE15 cells, which are immortalized Clara cells and Type II pneumocytes, respectively (18, 19) (Figure 1B). However, expression in neither cell line was inducible by RA, probably due to the high basal level of RAIG1 expression. (It was shown previously that RA inducibility was only manifested in cell lines that have low basal levels of RAIG1 expression [8].) The temporal pattern of mouse RAIG1 expression was also examined by Northern blot analysis (Figure 1C). In developing lung RAIG1, mRNA was first detected at E14.5. The level of expression gradually increased from E14.5 to E18.5, when it reached its highest levels. Strong expression of RAIG1 persisted postnatally. The onset of mouse RAIG1 expression in embryonic lung coincides with the onset of lung differentiation, suggesting a role for RAIG1 in lung maturation or differentiation. In situ hybridization demonstrated strong expression in the lung parenchyma at E18.5 (Figure 1D). The signal was also detected in the epithelial layer of the large bronchi. However, no RAIG1 expression was detected in the pulmonary vasculature. A more detailed assessment of cellular specificity at this and other stages of development was achieved using ß-galactosidase "knock-in" mice (see below).
Gene Targeting of the Mouse Raig 1 Gene To understand the in vivo expression and biological functions of mouse RAIG1, we generated a mutant allele of the Raig1 gene through homologous recombination in embryonic stem (ES) cells (Figure 2). A reporter lacZ gene was "knocked into" the translation start site in exon II to recapitulate the endogenous Raig1 expression and facilitate expression analysis. Using a probe located 3' to the region that was deleted in the targeted allele, Northern blot analysis failed to show any detectable RAIG1 mRNA (Figure 2D) in homozygous mutant lungs. This suggests that the transcription of the Raig1 gene was disrupted, and that a null allele was generated.
LacZ "Knock-In" Mice Demonstrate Lung Epithelium–Specific Expression of RAIG1 Northern blot analysis showed that the lacZ "knock-in" expressed a transcript of 4.4 kb which, in adult mice, was only present in the lung (data not shown). Thus, lacZ expression in the "knock-in" has a tissue distribution similar to the endogenous Raig1 expression. During embryonic development, lacZ expression was initiated at E15.5, again temporally similar to that of endogenous Raig1 (data not shown). The production of ß-galactosidase as revealed by X-gal histochemical staining was slightly delayed and was not detectable until E16.5 (Figures 3A and 3B). At this stage, lacZ activity was limited to the epithelium but was broadly expressed in both terminal buds and future conducting airway epithelium. At E18.5, a similar expression persisted (Figure 3C) with no expression in either the mesenchyme or vascular tissue (arrowhead in Figure 3C). In addition to the strong expression in the lung parenchyma, weak expression was also detected in the trachea and the esophagus (Figure 3D). These patterns are virtually identical to the endogenous Raig1 expression, as revealed through radioactive in situ hybridization analysis (Figure 1D). Dynamic changes in lacZ staining patterns occurred postnatally. At postnatal day 10 (P10), while lacZ continued to be detecteded in the parenchyma, expression was lost in the large airways, particularly in the trachea epithelium (Figure 3E). In adult mice, there was no expression of lacZ in the conducting airways except for a small group of bronchiolar cells at the transition from terminal bronchiolus to alveolus (arrowheads in Figure 3F). In the respiratory epithelium, both type II cells at the corner of alveoli and flattened type I cells stained blue (Figure 3F), identifying RAIG1 as a unique epithelial gene expressed by both types of respiratory epithelial cells.
Phenotype of RAIG1 Null Mutant Mice Offspring from RAIG1 heterozygote matings were born in normal Mendelian ratios. The RAIG1–/– offspring were grossly phenotypically indistinguishable from their heterozygous and wild-type littermates. Adult RAIG1–/– mice appeared in good health and seemed to have a normal lifespan. Thus, RAIG1 is not essential for general health and survival under normal (nonstressed) living conditions. Histologically, lungs of 3-mo-old RAIG1–/– mice had normal airspace size based on mean linear intercept analysis, suggesting no significant defects in alveogenesis (Figures 4A and 4B). Nor did a similar analysis detect any airspace changes at 9 mo, indicating no destructive tissue damage associated with RAIG1 deficiency in the lung parenchyma. In situ hybridization revealed similar distribution and number of cells that express SP-C (Figures 4C and 4D). Type II cells in the Raig1-null mutant also produced normal lamellar bodies based on transmission electron microscopy (Figures 4E and 4F). At the molecular level, RAIG–/– lungs expressed normal levels of SP-A, SP-B, SP-C, SP-D, CC10, T1a, and Aquaporin5 based on Northern blot analysis (Figure 4G). Together these data suggest no apparent developmental abnormality in the histogenesis of Clara cells, Type I cells, or Type II pneumocytes. Phospholipid assay showed similar phospholipid content in the BAL of both wild-type and null mutant lungs (data not shown), providing additional evidence for normal surfactant production. We also compared embryonic expression of SP-C and SP-B mRNA in knockout mice and their wild-type littermates (Figure 5). Normal temporal expression of SP-C and SP-B at E15.5–17.5 suggested no delay of epithelial differentiation during long development.
Expression of Mouse RAIG2, 3, and GPCR5 d in the Lung The lack of phenotypic defect in RAIG1 –/– mice, despite the strikingly strong lung-specific expression, prompted us to explore potential functional redundancy with closely related family members, RAIG2, 3, and GPRC5d. The expression patterns of these three genes in adult tissues and developing lungs were interrogated with either Northern blot analysis and/or real-time PCR assays (Figures 6 and 7). Among a panel of 16 adult tissues examined, mouse RAIG2 was predominantly expressed in cerebral cortex and cerebellum (Figure 7A). Abundant expression was also detected in testis. This is intriguing in that human RAIG2 was not expressed at high levels in testis according to previously published real-time PCR data (12). Expression is relatively low in all stages of lung development compared with the levels in cortex, cerebellum, or testis.
RAIG3 messenger RNA was detected in many tissues on a multitissue Northern blot analysis, with highest levels in liver and kidney (Figure 6A). Expression in adult lung was relatively weak but readily detectable (Figures 6A and 6B). Northern blots also showed stronger expression at E12.5, 14.5, 16.5, and 18.5, as well as at P0 and Day 10, than in the adult mouse lung. Conventional RT-PCR using a forward primer located on exon 2 and a reverse primer located on exon3 also detected RAIG3 transcript in embryonic and adult lungs (Figure 6C). The RAIG3 transcript was also detected by both assays in MLE15 and TMCC cells (Figures 6B and 6C), suggesting its expression in the epithelial compartment. There did not seem to be any compensatory increase of RAIG3 mRNA expression in RAIG1–/– lungs at E16.5 on a Northern blot analysis (data not shown). GPRC5d messenger RNA was previously shown to be present in multiple tissues using conventional qualitative RT-PCR assay (9). However, the levels of expression in these tissues are unknown. Using real-time PCR, we were able to assess the tissue expression more quantitatively. Our data revealed predominant expression of mouse GPRC5d in adult skin with little expression in either embryonic or adult lungs (Figures 7C and 7D). The expression in 14 other adult tissues was unremarkable. Thus, GPRC5d is unlikely to play important roles in lung biology. Together these data suggest that RAIG3 but not RAIG2 or GPRC5d may compensate for the function of RAIG1 in RAIG1-deficient mice, during both lung development and postnatal homeostasis.
In this report we cloned the murine RAIG1 gene and found that it is prominently expressed both in epithelial cells during lung development as well as in adult lungs. Its expression pattern is unique in that it is specifically expressed in both type I and type II alveolar epithelial cells. RAIG1 lacZ "knock-in" mice confirmed this expression. Nevertheless, RAIG1 null mutant "knock-out" mice did not show any obvious structural or molecular developmental defects. We also observed RAIG3 expression in the lung, which might mask the importance of RAIG1 during lung development.
A unique feature of RAIG1 is its coexpression in both type I and type II pneumocytes. Although many genes have specific patterns of expression in the distal respiratory epithelium, they are expressed either in type I or type II cells but not in both. For example, SPC and SPA are expressed in type II but not type I cells (20, 21), whereas T1 Although RAIG1 was originally cloned as an RA-responsive gene, its expression was not inducible in several cell lines, including the two mouse cell lines used in this study. It is interesting that all these cell lines have high baseline expression of RAIG1. We speculate that there exist negative regulators of RAIG1 expression, and that RA signaling can release the transcriptional suppression by these regulators. Some tumor cell lines may have lost the function of these repressors due to loss of chromosomes or other genomic alterations, which renders high expression of RAIG1 and the lack of inducibility by RA. The unique pattern of RAIG1 expression and its lung tissue specificity also represent a useful tool for targeting genes to the distal epithelium through transgenic and/or "knock-in" approaches, particularly if type I cell expression is desired. The heightened transcriptional activity of the RAIG1 locus as revealed here by Northern blot analysis further supports its utility in genetic engineering. There are a large number of G protein–coupled receptors present in the mammalian genome that serve a broad range of physiologic functions. For example, G protein signaling is critical in regulating ionic fluxes in lung epithelium. Activation of the ß-adrenergic receptor accelerates the clearance of alveolar fluid by increasing the expression and activity of epithelial Na+ channels (ENAC) and Na, K-ATPase. Nucleotide receptor P2Y2, which has been shown to be expressed in airway epithelium (25), is likely to promote Cl– secretion by mobilizing intracellular Ca2+ and stimulating chloride channels, such as Ca2+-regulated Cl– channel (CRCC) and cystic fibrosis transmembrane regulator (CFTR) (26). The unique expression of RAIG1 in respiratory epithelium may implicate other physiological functions mediated by G protein in the lung epithelium. The elucidation of these functions may require the identification of its ligands and a greater biochemical understanding of its downstream signaling. We hypothesize a potential role of RAIG1 in regulating epithelial differentiation during lung development, and functional redundancy between RAIG1 and RAIG3 based on their temporal expression pattern. First, embryonic RAIG3 expression overlaps with that of RAIG1 from E14.5 to E18.5. Second, the expression of RAIG3 in TMCC and MLE15 cell lines suggests that it is expressed in the same epithelial cells that also express RAIG1. Lastly, although expression of RAIG3 is low in adult lung it is easily detectable with conventional RT-PCR assay and may be enough to compensate for the loss of RAIG1 function. Therefore, we expect abnormal lung development or/and lung homeostasis in mice that are deficient in both RAIG1 and RAIG3. Moreover, subtle structural changes in RAIG1–/– mice might be present and brought out by future lung injury models where lung homeostasis is perturbed.
The authors thank Mary Bauman and the Washington University Embryonic Stem Cell Core facility for performing ES cell transfection, Ron McCarthy for pronuclei microinjection, and Marilyn Levy for EM analysis.
This work was supported by NIH grant R01 HL62859. Conflict of Interest Statement: J.X. has no declared conflicts of interest; J.T. has no declared conflicts of interest; S.D.S. has participated in Advisory Boards for Boehringer Ingelheim, GlaxoSmithKline, Millenium, Pfizer, Wyeth, and ICOS for the past three years and has performed research with his laboratory in collaboration with Pfizer, Arriva, ONO, and Taisho. No personal income was obtained. Received in original form November 18, 2004 Received in final form January 5, 2005
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
