Gpr116 Receptor Regulates Distinctive Functions in Pneumocytes and Vascular Endothelium

Abstract

Despite its known expression in both the vascular endothelium and the lung epithelium, until recently the physiological role of the adhesion receptor Gpr116/ADGRF5 has remained elusive. We generated a new mouse model of constitutive Gpr116 inactivation, with a large genetic deletion encompassing exon 4 to exon 21 of the Gpr116 gene. This model allowed us to confirm recent results defining Gpr116 as necessary regulator of surfactant homeostasis. The loss of Gpr116 provokes an early accumulation of surfactant in the lungs, followed by a massive infiltration of macrophages, and eventually progresses into an emphysema-like pathology. Further analysis of this knockout model revealed cerebral vascular leakage, beginning at around 1.5 months of age. Additionally, endothelial-specific deletion of Gpr116 resulted in a significant increase of the brain vascular leakage. Mice devoid of Gpr116 developed an anatomically normal and largely functional vascular network, surprisingly exhibited an attenuated pathological retinal vascular response in a model of oxygen-induced retinopathy. These data suggest that Gpr116 modulates endothelial properties, a previously unappreciated function despite the pan-vascular expression of this receptor. Our results support the key pulmonary function of Gpr116 and describe a new role in the central nervous system vasculature.

Fig 1. Vascular expression and genetic ablation of the Gpr116 gene in mouse.

A. Gpr116 mRNA expression in the published organ-specific EC mRNA dataset [45]. B. Gpr116 mRNA expression assessed by qRT-PCR in EC from 3-weeks-old and 3-months-old ROSA^mT/mG x Tie2-Cre mice. Results are normalized by brain EC expression. Error bars represent SD. (n = 3 mice per genotype). C. Gpr116 mRNA expression in the published brain-specific vascular and EC mRNA dataset [46]. D. Schematic representation of the area targeted by homologous recombination in the Gpr116 locus. Dotted lines indicate the regions of homology in between the Gpr116 locus and the cassette. The dark grey arrow indicates the position of WT primers: both are located in the untranslated region of exon 21, but the area recognized by the forward primer is lost in the mutant allele. The light grey arrow represents the knockout primer, specific for the cassette. Critical Gpr116 domains (SEA, IgG, GAIN and transmembraine, TM) are indicated above the corresponding encoding exons. E. Example of genotyping PCR products on genomic DNA (toe) from Gpr116 WT, heterozygous and knockout littermates. WT primers amplify a 325-bp fragment in the 3´UTR exon 21 of Gpr116 gene representing the wild type allele. The 401 bp band is specific for the mutant allele. F. Example of genotyping PCR products using genomic DNA (toe) from Gpr116 WT, heterozygous and knockout littermates. LacZ primers amplify a 210 bp fragment in LacZ gene present in the insert replacing exon 4 to 21. G. Gpr116 exon 17–18 mRNA expression assessed by qRT-PCR in Gpr116 WT, heterozygous and knockout organs at P4 (n = 3 mice per genotype). H. Gpr116 exon 2–3 mRNA expression assessed by qRT-PCR in Gpr116 WT, heterozygous and knockout organs at P4 (n = 3 mice per genotype). I. mRNA detection by RNAscope in brain cortical capillary vessels from Gpr116 WT (top row), knockout (middle row) and ROSA^mTmG X Tie2-Cre mice (lower row) at 3 weeks. On the left column, note that only the probe signal (red) and the nuclear staining (blue) are visible. On the right column, an endothelial staining (green) is merged to the probe and the nuclear signal: a CD31 antibody staining is on the two upper rows, while Tie-2 Cre mediated GFP is on the lower row. (n = 1 mouse per genotype).

Fig 2. Massive accumulation phenotype in lungs of aged Gpr116 ^-/- mice.

A. Bright field image of the inflated lung from Gpr116 WT, heterozygous and knockout littermates. B. Weights of whole lungs over total body weight from Gpr116 WT, heterozygous and knockout littermates (n≥5 mice per genotype). C. Bright field images of heart from Gpr116 WT, heterozygous and knockout littermates. D. Weights of the heart (left) over total body weight from Gpr116 WT, heterozygous and knockout littermates (n≥5 mice per genotype). E. Bright field images of the spleen from Gpr116 WT, heterozygous and knockout littermates. F. Weights of the spleen (left) over total body weight from Gpr116 WT, heterozygous and knockout littermates (n≥5 mice per genotype). G. BALF collected from Gpr116 WT, heterozygous and knockout littermates (The picture shown is representative of 3 mice for each genotype). H. Quantification of saturated phosphatydilcholine in BALF by ELISA (n = 3 mice per genotype). I. Quantification of protein content in BALF by BCA assay (n = 3 mice per genotype). J. Surfactant proteins detection in BALF by western blot. Molecular weights are indicated on the right. (n = 2 mice per genotype). K. Bright field images of the lung, after hematoxylin and eosin staining. The black arrowheads indicate alveolar macrophages (the image is representative of 4 mice for each genotype). L. Electron microscopy view of Gpr116 wildtype and knockout lungs (n = 2 mice for each genotype). M. Confocal images of lung sections stained with ADRP (white) and nuclear stain (Hoechst, blue). Note that a red autofluorescent signal appears in knockout lungs. (the image shown is representative of 2 mice for each genotype). N. Confocal images of lung sections stained with nuclear marker Hoechst (blue) to show autofluorescent cells accumulated in the alveolar space, either in the green or red channel (the image is representative of 3 mice for each genotype). O. Autofluorescence emission spectrum of macrophages in the old knockout lung, upon 405 nm excitation (the image is representative of 2 mice). P. Detection of autofluorescent cells from Gpr116 knockout lung by FACS (n = 2 mice per genotype)

Fig 3. Retinal vascular patterning in Gpr116 ^-/- mice.

A. Vascular network in P4 retinas. Dashed line indicates the limits of the retina (the picture shown is representative of at least 5 mice for each genotype). B. Quantification of the retinal vascular outgrowth at P4 (n = 5 for WT, n = 12 for heterozygotes and n = 6 for knockout). C. Vascular patterning in P7 retinas from Gpr116 WT, heterozygous and knockout littermates. Isolectin (red), CD31 (green) and Erg (grey) were used to visualize endothelium, and NG2 (green) and ASMA (red) to detect mural cells (the images shown are representative of 3 mice for each genotype). D. Vascular patterning in P7 retinas from Gpr116 ECKO and littermates controls. Isolectin (red) is used to visualize endothelium, and NG2 (green) and smooth muscle actin α (ASMA, blue) to detect mural cells (the images show are representative of 2 mice per genotype). E. Isolectin (red) and FITC-dextran (green) distribution in P21 retinas from Gpr116 WT, heterozygous and knockout littermates. CD31 (green) is used to stain the endothelium, and nuclei are stained with Hoechst (blue) (the images shown are representative of 3 mice per genotype). F. Monolayers formed by isolated endothelial cells from Gpr116 WT, heterozygous and knockout brain. Endothelial cells (CD31) and nuclei (Hoechst) are indicated in green and blue, respectively (the pictures shown are representative of 3 mice for each genotype)

Fig 4. Blood brain barrier breakdown in Gpr116 ^-/- mice.

A. Whole brain images taken after 1kDa cadaverine perfusion (left) and associated quantification of extravasated cadaverine (right) in aged Gpr116 WT, heterozygous and knockout mice (n≥5 mice for each genotype). B. Whole brain images taken 70 kDa tetramethylrhodamine dextran perfusion (left) and quantification of extravasated tracer (right) in Gpr116 WT and heterozygous and Gpr116 ECKO mice (n = 3 for wild type and ECKO, n = 2 for PDGF-B ^ret/ret, n = 1 for uninjected control). C. Confocal images of cerebral cortex from aged Gpr116 WT, heterozygous and knockout mice. Astrocytes (GFAP) appear in green, endothelial cells (CD31) in red (the images are representative of 4 mice per genotype) and associated quantification of perivascular associated astrocytes in aged Gpr116 WT, heterozygous and knockout mice (n = 4 mice for each genotype, 2 sections at least quantified per genotype). D. Whole brain fluorescence images taken after Alexa 555-cadaverine circulation (upper) and quantification of extravasated cadaverine (lower) in 1.5-month-old Gpr116 knockout (n = 3 mice per genotype). E. Whole brain fluorescent images taken after cadaverine circulation (upper) and associated quantification of extravasated cadaverine (lower) in 2-months-old Gpr116 AEC KO (n = 6 mice per genotype). F. Whole brain fluorescent images taken after cadaverine circulation (upper) and quantification of extravasated cadaverine (lower) in 2-months-old Gpr116 ECKO (n = 7 mice per genotype)

Fig 5. Normalized pathological angiogenesis in Gpr116 ^-/- retinas.

A. Confocal images of post-OIR retinas from Gpr116 WT, heterozygous and knockout littermates at P12 (the images shown are representative of 5 mice per genotype). B. Confocal images of post-OIR retinas from Gpr116 WT, heterozygous and knockout littermates at P17 (the images shown are representative of 5 mice per genotype). C. Quantification of the avascular area on the post-OIR retinas from Gpr116 WT, heterozygous and knockout littermates at P12 (n = 5 mice at least per genotype). D. Quantification of the avascular area on the post-OIR retinas from Gpr116 WT, heterozygous and knockout littermates at P17 (n≥7 mice at least per genotype). E. Confocal images of post-OIR tufts (blue arrows) in Gpr116 WT, heterozygous and knockout littermates at P17 (the images shown are representative of 5 mice per genotype)