Generation and characterization of rgs5 mutant mice

Abstract

Regulators of G-protein signaling (RGS) are involved in a wide variety of functions, including olfaction, vision, and cell migration. RGS5 has a perivascular expression pattern and was recently identified as a marker for brain pericytes. This suggests a role for RGS5 in vascular development and pericyte biology. We have created a mouse line which lacks the rgs5 gene and replaced it with a green fluorescent protein (GFP) reporter (rgs5(GFP/GFP)). The mice are viable and fertile and display no obvious developmental defects, and the vasculature appears to develop normally with proper pericyte coverage. Also, no differences were observed in the vasculature under pathological conditions, such as tumor growth and oxygen-induced retinopathy. The GFP expression in pericytes of rgs5(GFP) mice allows detection and sorting of these cells, thereby providing a valuable novel tool for pericyte research.

FIG. 1.

Generation, verification, and genotyping strategy of rgs5^GFP mice. (a) Schematic representation of rgs5 gene targeting strategy achieved by the use of VelociGene technology. Top, wild-type rgs5 allele with exons 1 to 5 (boxes; coding sequences have dark shading; 5′ and 3′ untranslated regions have light shading). Bottom, targeting vector and recombined locus. The genomic sequence of rgs5 from part of exons 2 to 5 has been replaced in frame by the coding sequence of GFP and a loxP-flanked neomycin gene. Primers used for genotyping (see the text for details) are represented by arrowheads. (b) PCR genotyping with primers for wild-type and recombinant alleles. A 262-bp fragment was amplified from the wild-type allele, and a 363-bp fragment was amplified from the mutant allele. (c) RT-PCR on cDNA generated from brain RNA from 5-week-old animals. No band was observed for the mutant mice. (d) Quantitative real-time PCR results on rgs5 transcript in P6 brain RNA from wild-type (n = 3), heterozygous (n = 2), or knockout animals (n = 4). Wild-type samples were used as calibrators for comparative expression results. (e and f) Immunohistochemical staining of wild-type and mutant P20 retinas with antibodies against CD31 (red) and GFP (green). (g and h) Immunohistochemical staining of brain tissue from rgs5^GFP/GFP mice with antibody against CD31 (red) or GFP (green) (g) or against PDGFR-β (red) or GFP (green) (h). GFP expression colocalizes with PDGFR-β and appears predominantly in the cell bodies of pericytes (arrowheads). (i and j) In situ hybridization on E18.5 embryo sections from wild-type animals using rgs5 antisense probe as detailed in the text. (i) Strong expression is seen in mesangial cells (arrow) of the kidney. Arrowheads indicate Bowman's space surrounding the kidney glomerulus. (j) Pericytes of brain vessels express rgs5 (arrowheads). (k to p) Immunohistochemical stainings of rgs5^GFP/GFP mouse organs. Kidney glomeruli (k), kidney cortex (l), brain (m), heart (n), intestine (o), and pancreas (p) are shown, with antibody against CD31 (red) or GFP (green). Scale bar, 20 μm (e to h and k to p) or 25 μm (i and j). For inserts (l to p), original magnification, ×2.

FIG. 2.

rgs5 mRNA expression and analysis of tissue architecture and kidney function. (a to d) Hematoxylin-eosin staining of P30 brains (a and b) or kidneys (c and d) of wild-type and rgs5 mutant animals. Tissue architecture appears to be normal for the rgs5^GFP/GFP mice. Scale bar, 50 μm. (e) Sodium dodecyl sulfate-polyacrylamide gel on urine samples from rgs5^+/+ (lanes 1 to 3) or rgs5^GFP/GFP (lanes 4 to 6) mice. Lane 7, urine from proteinuric Fat2 mutant mouse (positive control). The arrow indicates the molecular mass of albumin (∼67 kDa). No albuminuria was detected in tissue from rgs5^GFP/GFP mice.

FIG. 3.

Pericyte quantification, pericyte distribution, and quantitative PCR analyses. (a and b) Representative images of layer 1 used for pericyte quantification from rgs5^+/+-XlacZ and rgs5^GFP/GFP-XlacZ retinas stained with isolectin (red) and β-galactosidase (black). Arrowheads indicate β-galactosidase-positive pericyte nuclei. Scale bar, 50 μm. (c) Quantitative real-time PCR results on rgs4 and rgs16 (left) and the pericyte markers PDGFR-β and Kir6.1 (right). Wild-type samples were used as calibrators for comparative expression results. Neither rgs4 nor rgs16 mRNA seems to be upregulated in order to compensate for the loss of rgs5. Note the 50% decrease in rgs4 mRNA expression for mutant mice compared to results for littermate controls. No differences were observed in the expression of pericyte markers (PDGFR-β and Kir6.1) between the control and mutant mice (n = 3 for all samples). (d and e) Immunohistochemical staining of wild-type and mutant P30 retinas with an antibody against NG2 (green). Scale bar, 100 μm.

FIG. 4.

Tumor growth and analyses of tumor vessels. (a) Tumor growth curve of T241 cells inoculated on control and rgs5^GFP/GFP mice (n = 5 for each genotype and time point). (b to d) Immunohistochemical stainings of T241 tumors from control and mutant animals. Tumors from wild-type (b) or mutant (c) mice stained with antibody against CD31 (red), αSMA (green), or NG2 (blue). Both tumors appear to have similar pericyte coverage and expression of the analyzed markers. (d) Tumor from mutant mouse stained with antibody against CD31 (red) or GFP (green). Arrowheads indicate GFP-positive pericytes. Scale bar, 20 μm (b to d).

FIG. 5.

Pericyte distribution in normal retina and in oxygen-induced retinopathy. IHC on flat-mounted retinas from control or rgs5^GFP/GFP mice. (a to d) Retinas from mice subjected to oxygen-induced retinopathy. NG2-immunoreactive pericytes are in green. No differences were observed in neovascularization or in the amount and distribution of NG2-positive pericytes between the control and mutant mice. Scale bar, 100 μm (a and b) or 500 μm (c and d).

FIG. 6.

MAP in rgs5 transgenic mice. MAP was determined by tail cuff plethysmography in unanesthetized control and rgs5 null mice. rgs5^GFP/GFP mice are hypotensive relative to littermate wild-type (rgs5^+/+) mice (n = 8 animals; *, P < 0.05 by unpaired Student's t test [two tail, equal variance]).