An efficient inducible model for the control of gene expression in renin cells

Abstract

Fate mapping and genetic manipulation of renin cells have relied on either noninducible Cre lines that can introduce the developmental effects of gene deletion or bacterial artificial chromosome transgene-based inducible models that may be prone to spurious and/or ectopic gene expression. To circumvent these problems, we generated an inducible mouse model in which CreERT2 is under the control of the endogenous Akr1b7 gene, an independent marker of renin cells that is expressed in a few extrarenal tissues. We confirmed the proper expression of Cre using Akr1b7^CreERT2/+;R26R^mTmG/+ mice in which Akr1b7⁺/renin⁺ cells become green fluorescent protein (GFP)⁺ upon tamoxifen administration. In embryos and neonates, GFP was found in juxtaglomerular cells, along the arterioles, and in the mesangium, and in adults, GFP was present mainly in juxtaglomerular cells. In mice treated with captopril and a low-salt diet to induce recruitment of renin cells, GFP extended along the afferent arterioles and in the mesangium. We generated Akr1b7^CreERT2/+;Ren1^cFl/−;R26R^mTmG/+ mice to conditionally delete renin in adult mice and found a marked reduction in kidney renin mRNA and protein and mean arterial pressure in mutant animals. When subjected to a homeostatic threat, mutant mice were unable to recruit renin⁺ cells. Most importantly, these mice developed concentric vascular hypertrophy ruling out potential developmental effects on the vasculature due to the lack of renin. We conclude that Akr1b7^CreERT2 mice constitute an excellent model for the fate mapping of renin cells and for the spatial and temporal control of gene expression in renin cells.

NEW & NOTEWORTHY Fate mapping and genetic manipulation are important tools to study the identity of renin cells. Here, we report on a novel Cre mouse model, Akr1b7^CreERT2, for the spatial and temporal regulation of gene expression in renin cells. Cre is properly expressed in renin cells during development and in the adult under basal conditions and under physiological stress. Moreover, renin can be efficiently deleted in the adult, leading to the development of concentric vascular hypertrophy.

INTRODUCTION

Renin cells are essential for survival. In the adult unstressed mammal, renin cells are located in the afferent arterioles near the glomeruli, thus their name juxtaglomerular (JG) cells. JG cells synthesize and release renin, an enzyme-hormone crucial for the regulation of blood pressure and fluid electrolyte homeostasis (1). In addition, renin cells have important roles in tissue morphogenesis, kidney vascular development, tissue repair and regeneration, and innate immune defense (2–5). Renin cells exhibit a high degree of plasticity. In response to homeostatic stress, cells of the renin lineage, such as vascular smooth muscle cells, mesangial cells, and pericytes, reexpress renin in a process known as recruitment (6). This response is usually sufficient to restore homeostasis. However, under conditions of chronic and persistent activation of the renin cell program, such as the permanent inhibition of the renin-angiotensin system (RAS), renin cells contribute to the development of concentric vascular hypertrophy (7–15). The genetic and epigenetic regulatory mechanisms that control the remarkable plasticity of renin cells are beginning to be unraveled, although the complete repertoire of signaling mechanisms remains unidentified. To understand the plasticity of renin cells in vivo, it is necessary to have a model that allows the temporal and spatial control of gene regulatory networks without developmental effects that may occur using constitutive Cre lines.

Gene targeting in renin cells has relied either on noninducible Cre recombinase or bacterial artificial chromosome (BAC) transgene-based inducible mouse lines (6, 16, 17). In particular, the noninducible knockin Ren1^{d Cre} line developed in our laboratory (6) has been extensively used to delete specific genes in renin lineage cells and to label cells with fluorescent markers for lineage tracing and cell isolation, providing significant information on renin cell identity and function (18, 19). However, noninducible Cre recombinase mouse models can introduce confounding developmental effects of gene deletion in the adult animal, and transgenic lines are prone to spurious and/or ectopic gene expression. To circumvent these problems, an efficient, specific, and temporally and spatially conditional model is necessary.

Temporal control of gene expression can be regulated using a modified Cre recombinase fused to a mutated ligand-binding domain of the estrogen receptor (CreERT2) (19). This modified domain binds with high affinity to the synthetic ER ligand tamoxifen but not endogenous estrogen. Upon tamoxifen administration, Cre is translocated into the nucleus and catalyzes the recombination at loxP-flanked alleles (20). For spatial control of gene expression, the CreERT2 cassette is placed under the control of a cell-specific gene. The aldo-keto-reductase Akr1b7 is an independent marker of renin cells that is coexpressed with renin under different developmental, physiological, and pathological conditions (21, 22). It has been suggested that the function of Akr1b7 is to remove harmful aldehydes resulting from the high synthetic activity of renin cells (21, 22). During embryonic life, Akr1b7 and renin are expressed in the large intrarenal arteries, along arterioles, and in JG and mesangial cells, whereas in the adult, both genes become confined to the JG area (21, 22). In addition, under physiological stress, Akr1b7 is coexpressed with renin when renin cell descendants (smooth muscle cells, mesangial cells, and pericytes) reacquire the renin phenotype (22). Using an independent promoter, such as Akr1b7, to drive the expression of CreERT2 in renin cells would enable gene manipulation in renin cells, regardless of the renin gene status, such as in the context of Ren1 deletion. Based on this information, we hypothesized that Akr1b7 was an excellent candidate for fate tracking and the temporal and spatial control of genes involved in the development and regulation of renin cells.

Here, we report the characterization of a novel Cre mouse model, Akr1b7^CreERT2, for the spatial and temporal regulation of gene expression in renin cells. We show that Akr1b7^CreERT2 is expressed in renin cells during development and in adults under both basal conditions and physiological stress. In addition, renin can be efficiently deleted in the adult using Akr1b7^CreERT2 mice, leading to a characteristic vascular pathology-concentric vascular hypertrophy, independent of renin deletion during development.

METHODS

Generation of Akr1b7^CreERT2 Mice

Akr1b7^CreERT2 mice were generated by inserting a P2A-CreERT2 knockin cassette immediately upstream of the TGA stop codon of the mouse Akr1b7 gene followed by the endogenous 3′-untranslated region (UTR) using a homologous recombination-based technique (Ingenious Targeting Laboratory, Ronkonkoma, NY; Fig. 1A). In brief, a targeting vector was constructed by subcloning a 9.1-kb region from an Akr1b7 positively identified C57BL/6 fosmid clone (WI1-1577M19) into the 2.45 kb-iTL cloning vector derived from the pSP72 vector (Promega; Madison, WI). The region was designed so the long homology arm (LA) extended 4.9 kb upstream of the P2A-CreERT2 knockin cassette, and the short homology arm (SA) extended 1.9 kb 3′ to a LoxP-FRT-flanked Neomycin cassette that was inserted downstream of the 3′-UTR. The entire P2A-CreERT2 knockin cassette was confirmed by sequencing. Ten micrograms of the 15.8 kb targeting vector were linearized with NotI and transfected by electroporation into FLP C57BL/6 (BF1) embryonic stem cells (ESCs) for homologous recombination. After selection with G418 antibiotic, surviving clones were expanded for PCR analysis to identify recombinant ESC clones. We confirmed the correct integration of the Akr1b7^CreERT2 knockin allele by PCR analysis of expanded ESC clones using the following primers: NEOGT: 5′-GTC CGT GTC GCGA AGT TCC TAT ACT TTC-3′ and A1: 5′-ACT AGT CAC TTT CTC ATG CAA GGG G-3′, product size: 2.04 kb (Fig. 1, A and B). The Neo cassette in the targeting vector was removed during ES clone expansion. We confirmed the retention of the P2A-CreERT2 cassette in the ESC clones by PCR using the following primers: SQ1: 5′-TGA TAG CTC GCA GGC CTT CCT TC-3′ and FN2A: 5′-AAC TTC GCG ACA CGG ACA CAA TCC-3′, product size: 2.61 kb (Fig. 1, A and B). We determined the copy number of the targeting vector in ESC clones by real-time PCR to confirm a single integration (Fig. 1C). Probe A was designed at the knockin insertion site and therefore annealed only to the wild-type (WT) allele. One copy indicated that the clone was correctly targeted. Probe B was designed in the region of the LA and annealed to both WT and targeted alleles. Clones with two copies were considered as having a single integration. Targeted iTL BF1 (C57BL/6 FLP) ESCs were microinjected into Balb/c blastocysts. The resulting chimeras with a high percentage black coat color were mated to C57BL/6N WT mice to generate germline Neo-deleted mice. Akr1b7^CreERT2 mice were crossed with C57BL/6J mice or reporter mice on a C57L/6J background for characterization.

Figure 1.

Generation of Akr1b7^CreERT2 mice. A: schematic diagram of the targeting strategy for the generation of the Akr1b7^CreERT2 knockin allele. A, top: map of the mouse Akr1b7 gene in chromosome 6 showing the intron/exon structure. Exons are indicated by numbered blue boxes. A, middle: schematic of the targeting vector. Homology arms are indicated by red bars. The long homology arm extended 4.9 kb upstream of the P2A-CreERT2 knockin cassette and the short homology arm extended 1.9 kb 3′ to the Neo cassette. The P2A-CreERT2 knockin cassette was inserted immediately upstream of the TGA stop codon of the Akr1b7 gene, followed by the endogenous 3′-untranslated region (UTR). An FRT-flanked Neo selection cassette was inserted downstream of the 3′-UTR. A, bottom: knockin allele after FRT-directed Neo deletion in ESCs. B, top: PCR analysis of two expanded clones (clone 183 and clone 184) confirmed correct integration of the Akr1b7^CreERT2 knockin allele. Wild-type (WT) DNA was used as a negative control. DNA from an individual clone (before expansion) was used as a positive control (+). No DNA was used as a blank (−). Primers: NEOGT/A1. Product size: 2.0 kb. B, bottom: PCR analysis of expanded ESC clones confirmed retention of the P2A-CreERT2 cassette after expansion of clones. Primers: SQ1/FN2A. Product size: 2.6 kb. C: analysis of copy number of targeting vector in ESC clones by real-time PCR. Probe A was designed at the knockin insertion site and therefore annealed only to the WT allele. One copy indicated that the clone was correctly targeted. The WT sample, indicated as BF1, had two copies. Probe B was designed in the region of the long homology arm and annealed to both WT and targeted alleles. Clones with two copies were considered as having a single integration. Clones 183 and 184 were correctly targeted and carried a single copy of the targeting vector sequence. ESCs, embryonic stem cells.

Genotyping of Akr1b7^CreERT2 Mice

Genotyping of the mice was conducted by PCR of DNA from tail biopsies using primers SQ1: 5′-TGA TAG CTC GCA GGC CTT CCT TC-3′/CRE-ERT SCSQ1: 5′-ATC TTC AGG TTC TGC GGG AAA CCA-3′ to detect the Akr1b7^CreERT2 allele (product size: 538 bp) and PNDEL1: 5′-CAG AGA AAC GAA GGG CAA CG-3′/PNDEL2: 5′-GCT CCC TGA TGC CTC CTC TA-3′ (product size 838 bp) to detect the WT allele (Fig. 1A).

Animals

All animals used in this study were maintained in the C57BL/6 background. We generated the following mouse lines: 1) Akr1b7^CreERT2/+;R26R^mTmG/+ animals by crossing Akr1b7^CreERT2 mice with the R26R^mTmG Cre reporter mouse (Strain No. 007676, The Jackson Laboratory, Bar Harbor, ME) (23). Experiments were conducted at embryonic day (E)18.5, postnatal day (P)5, and in females and males at 1–3 mo of age. 2) Akr1b7^CreERT2/+;R26R^tdTomato/+ animals by crossing Akr1b7^CreERT2 mice with the R26R^tdTomato Cre reporter (Strain No. 007909, also known as Ai9, The Jackson Laboratory). Female and male mice were studied at 3 mo of age. 3) Akr1b7^CreERT2/+;Ren1^cYFP;R26R^tdTomato/+ animals by crossing Akr1b7^CreERT2 mice with Ren1^cYFP;R26R^tdTomato reporter mice. In this line, the Ren1^cYFP transgene labels cells actively expressing renin as we have previously shown (24). Female and male mice were studied at 1–2 mo of age. 4) Akr1b7^CreERT2/+;Ren1^cFl/−;R26R^mTmG/+ (Renin cKO^GFP) by crossing Akr1b7^{CreERT2/CreERT2;} Ren1^c+/− to Ren1^{cFl/Fl (Homo)}; R26R^mTmG/mTmG to both delete renin in adult animals and simultaneously label renin cells with green fluorescent protein (GFP). The Ren1^c null (Ren1^c−/−) mouse line was generated by Takahashi et al. (9). The Ren1^c floxed mouse were generated by Xu et al. (25). This crossing strategy yielded Akr1b7^CreERT2/+;Ren1^cFl/−;R26R^mTmG/+ mutant mice and Akr1b7^CreERT2/+;Ren1^cFl/+;R26R^mTmG/+ littermate controls. Female and male mice were studied at ages of 2–5 mo for short-term (15 days) renin deletion and 6 mo for long-term (5 mo) renin deletion.

All animals were housed in the University of Virginia’s vivarium facilities equipped with controlled temperature and humidity conditions on a 12:12-h dark-light cycle. All procedures were performed per the National Institutes of Health Guide for the Care and Use of Laboratory Animals (https://grants.nih.gov/grants/olaw/guide-for-the-care-and-use-of-laboratory-animals.pdf) and approved by the University of Virginia Animal Care and Use Committee. Euthanasia was performed in adult mice by anesthesia administration [tribromoethanol (300 mg/kg ip) or isoflurane] followed by cervical dislocation and in neonates (P5) by decapitation.

Tamoxifen Treatment

For embryonic studies, we crossed Akr1b7^CreERT2/+;R26R^+/+ males to Akr1b7^+/+;R26R^mTmG/mTmG females. Pregnant mice received two consecutive intraperitoneal injections of 2 mg/40 g body wt tamoxifen (stock: 20 mg/mL corn oil) on days 15.5 and 16.5 of pregnancy, and embryos were collected at E18.5. For neonatal studies, we set up similar crosses, injected the progeny with 50 µL tamoxifen intragastrically (stock: 1 mg/mL corn oil) at P1, P2, and P3, and analyzed the animals at P5. For experiments in adult animals, 2 mg/20 g body wt tamoxifen was injected intraperitoneally (stock: 20 mg/mL corn oil). The number and timing of injections are detailed in results for each experiment and summarized in Table 1.

Table 1.

Summary of tamoxifen administration protocols used in the study

Figure	Injection Protocol	Sample Collection
Figure 2A	3 tam i.g. injections at P1, P2, and P3	2 days after last injection
Figure 2B	3 tam i.p. injections at 1 mo	2 days after last injection
Figure 2C	3 vehicle i.p. injections at 1 mo	2 days after last injection
Figure 2D	3 tam i.p. injections at 2 mo	3 days after last injection
Figure 3	3 tam i.p. injections at 2 mo	4 days after last injection
Figure 4	2 tam i.p. injections at pregnancy days 15.5 and 16.5	2 days after last injection
Figure 5	5 tam i.p. injections at 1–2 mo	3 days after last injection
Figures 6 and 7	5 tam i.p. injections at 2–4 mo	10–15 days after last injection
Figure 8	5 tam i.p. injections at 1 mo	5 mo after last injection
Supplemental Figure S1	3 tam or vehicle i.p. injections at 2 mo	3 wk after last injection
Supplemental Figure S2	3 vehicle i.p. injections at 1 mo	2 days after last injection

i.g., intragastric; i.p., intraperitoneal; tam, tamoxifen.

Physiological Challenge

To induce an increase in the number of renin-expressing cells in the kidney, mice were subjected to a low-salt diet (0.1% NaCl) and the angiotensin-converting enzyme inhibitor captopril (Sigma-Aldrich, St. Louis, MO) in the drinking water (0.5 mg/L) for 7–8 days before the harvesting of tissues. Control mice received a normal-salt diet (0.3% NaCl) and no captopril.

Metabolic Cage Experiments

Adult male and female mice aged 80 to 100 days were housed individually in metabolic cages designed for urine collection. After an acclimation period of 1 day in metabolic cages with normal chow diet and regular water, urine samples were collected every 24 h for 2 days, and drinking water and urine volume were recorded.

Electrolyte Measurements

Urinary creatinine, Na⁺, and K⁺ measurements were performed by the University of Virginia Hospital Clinical Laboratory. Urine osmolality (Osm) was determined using a vapor pressure osmometer (Vapro Model 5600, Wescor, Logan, UT), which was calibrated using the ELITechGroup OPTIMOLE Osmolality Standard (290 mmol/kg, ELITechGroup, Logan, UT).

Blood Pressure Measurement

Mean arterial, systolic, and diastolic blood pressures were measured in isoflurane-anesthetized animals over a 10-min period from a catheter inserted into the right carotid artery using a RX104A transducer coupled to a data acquisition system and AcqKnowledge software (Biopac Systems, Goleta, CA).

Plasma Renin

Animals were anesthetized by injecting tribromoethanol intraperitoneally (300 mg/kg). Blood was collected by cardiac puncture and placed into K₂EDTA plasma separator tubes (BD Microtainer, Becton Dickinson, Franklin Lakes, NJ) on ice. For mice subjected to blood pressure measurement, blood was collected from the right carotid artery under isoflurane anesthesia. Plasma specimens were obtained from blood after centrifugation at 375 g at 4°C for 20 min and stored at −80°C. The plasma renin concentration was quantified using an ELISA kit for mouse renin (#ELM-Renin1, Ray Biotech, Norcross, GA) as previously described (15). This assay uses an antibody specific for mouse Renin 1 coated on a 96-well plate. Renin 1 present in standards and samples was bound to the wells by the immobilized antibody. After washing, a biotinylated anti-mouse Renin 1 antibody was added followed by horseradish peroxidase (HRP)-conjugated streptavidin. After washing, a TMB substrate solution was added to the wells, allowing color development in proportion to the amount of Renin 1 bound. A stop solution changed the color from blue to yellow, and the intensity of the color was measured at 450 nm. The minimum detectable concentration of renin, defined as the analyte concentration resulting in an absorbance that is 2 SDs higher than that of the blank, is 6 pg/mL.

Histological and Immunohistochemical Analyses

Mice were anesthetized by injecting tribromoethanol intraperitoneally (300 mg/kg). Kidneys were removed, fixed overnight in Bouin’s solution at room temperature or 4% paraformaldehyde (PFA) at 4°C, and embedded in paraffin. Immersion fixation, as opposed to perfusion fixation, was performed so that RNA extraction for gene expression analysis could be performed for the same animal. Sections (5 µm) from Bouin’s-fixed, paraffin-embedded kidneys were stained with hematoxylin and eosin (MilliporeSigma, Burlington, MA) to examine kidney morphology. Immunostaining was performed as previously described (26). In brief, 5-µm sections of Bouin’s-fixed, paraffin-embedded kidneys were incubated overnight at 4°C with a rabbit polyclonal anti-mouse renin antibody (1:500) generated in our laboratory (26) or mouse anti-α-smooth muscle actin (α-SMA) monoclonal antibody (1:10,000; MilliporeSigma) and biotinylated secondary goat anti-rabbit IgG or horse anti-mouse IgG (1:200; Vector Laboratories, Newark, CA) for renin or α-SMA, respectively. Staining was amplified using the Vectastain ABC kit (Vector Laboratories) and developed with 3,3-diaminobenzidine (MilliporeSigma). The sections were counterstained with hematoxylin (MilliporeSigma), dehydrated, and mounted with Cytoseal XYL (Thermo Fisher Scientific, Waltham, MA).

GFP expression after Cre recombination was visualized in frozen sections. Tissues were fixed in 4% PFA for 1 h at 4°C. After washing, the samples were incubated in 30% sucrose overnight at 4°C and frozen in OCT (Thermo Fisher Scientific). The frozen blocks were sectioned at 12 μm thickness and mounted in phosphate-buffered saline (PBS) for visualization.

In Situ Hybridization

To generate the probe for Akr1b7 in situ hybridization, we synthesized a 450-bp DNA fragment by PCR using cDNA from WT C57BL/6 mouse kidneys and primers 5′- AATTAACCCTCACTAAAGGGTGACCAACCAGATTGAGAGC-3′ and 5′- TAATACGACTCACTATAGGGCAGTATTCCTCGTGGAAAGGAT-3′ containing a 3′ T3 promoter and 5′ T7 promoter sequences, respectively. Digoxigenin (DIG)-labeled RNA sense and antisense probes were generated by in vitro transcription using DIG RNA Labeling Mix and T3 or T7 polymerases (MilliporeSigma). In situ hybridization was performed as previously described (15). In brief, 7 μm 4% PFA-fixed, paraffin-embedded kidney sections were deparaffinized, rehydrated, and postfixed with 4% PFA, followed by acetylation (0.375% acetic anhydride) and permeabilization with proteinase K (10 μg/mL). After preincubation with hybridization buffer (500 ng/mL in hybridization buffer of 50% formamide, 5× SSC, 50 μg/mL yeast transfer RNA, 1% SDS, and 50 μg/mL heparin), sections were incubated with the DIG-labeled sense or antisense riboprobes at 55°C overnight. The sections were washed and then incubated with anti-digoxigenin-alkaline phosphatase antibody (1:4,000, MilliporeSigma) overnight at 4°C. After washing, sections were treated with levamisole and incubated with BM Purple (MilliporeSigma) until the signals were visible. Sections were postfixed with 0.2% glutaraldehyde + 4% PFA and mounted with Glycergel Mounting Medium (Agilent Technologies, Santa Clara, CA). The antisense probe generated with T7 polymerase showed the specific signal. The sense probe generated with T3 polymerase showed no signal.

Microscopy

Tissue sections were visualized using a Zeiss Imager M2 microscope equipped with the AxioCam 305 color and AxioCam 506 mono cameras and ApoTome.2 (Zeiss, Oberkochen, Germany).

Pseudoquantitative Assessment

We determined the juxtaglomerular area index (JGAi) as the number of yellow fluorescent protein (YFP) or tdTomato-positive JG areas ÷ total number of glomeruli × 100 in 12 randomly selected ×20 cortex images for each animal under basal physiological conditions. To quantify the renin immunostaining in kidney sections, we used the magic wand tool of Photoshop software to automatically select positive areas in 12 randomly selected ×20 cortex images. The measurements were normalized for the total cortex area in each image and expressed as percentages. All the image analysis parameters were kept constant among different samples.

Tissue Clearing and Whole Mount Immunostaining

For the tissue clearing and whole mount immunostaining of Akr1b7^CreERt2;R26R^tdTomato mouse kidneys, we used the updated clear, unobstructed brain/body imaging cocktails and computational analysis (CUBIC) protocol (27). The mice were anesthetized with tribromoethanol (300 mg/kg) and perfused with 20 mL of PBS and 30 mL of 4% PFA via the left ventricle of the heart. Kidneys were removed, divided into two sections each, and fixed overnight in 4% PFA. All subsequent steps were performed with gentle shaking. The samples were washed with PBS for 6 h followed by immersion in CUBIC-L (10 wt% N-butyldiethanolamine and 10 wt% Triton X-100) at 45°C for 5 days. After the samples were washed with PBS for 6 h, they were placed in blocking buffer (PBS, 1.0% BSA, and 0.01% sodium azide) overnight at room temperature. Then the samples were immersed in immunostaining buffer (PBS, 0.5% Triton X-100, 0.25% BSA, and 0.01% sodium azide) containing 1:100 diluted anti-actin, α-smooth muscle (Acta2)-fluorescein (FITC) antibody (F3777, Sigma-Aldrich) for 6 days at room temperature. After washing again with PBS for 6 h, the samples were immersed in CUBIC-R+ [T3741 (45 wt% 2,3-dimethyl-1-phenyl-5-pryrazolone, 30 wt% nicotinamide, and 5 wt% N-butyldiethanolamine), Tokyo Chemical Industry, Tokyo, Japan] diluted 1:1 in water overnight at room temperaure. The samples were then immersed in CUBIC-R+ at room temperature for 2 days.

Macroscopic whole mount images were acquired with a light-sheet fluorescence (LSF) microscope (ZEISS Lightsheet7, Zeiss). The samples were imaged in mounting solution (RI:1.520) (M3294, Tokyo Chemical Industry) with Clr Plan-Neofluar 20x/1.0 Corr nd = 1.53 detection optics. The voxel resolution was as follows: x = 0.656 µm, y = 0.656 µm, z = 1.0 µm (zoom ×0.36); x = 0.169 µm, y = 0.169 µm, z = 0.560 µm (zoom ×1.4, 2 × 2 tiles). The FITC signals of α-SMA-expressing cells were measured by excitation with 488 nm lasers. The TdTomato signals were measured by excitation with 561 nm lasers.

Three-dimensionally (3-D) rendered images were visualized and edited with Imaris software (V. 10.0.0; Bitplane, Belfast, UK). Raw image files obtained from the Zeiss LSF microscope (.czi) were converted into Imaris files (.ims) using Imaris File Converter 10.0.0. Tilescan images were stitched using Imaris Stitcher 10.0.0. The image processing by Imaris software was performed as previously described (28). The reconstituted 3-D images were cropped to a region of interest using the crop function. The snapshot and animation functions were used to capture images and videos, respectively.

RNA Isolation and Real-Time RT-PCR Analysis

Quantitative real-time PCR (qPCR) analysis was performed in samples from kidney cortices. Total RNA was extracted using TRIzol reagent (Thermo Fisher Scientific) and RNeasy Mini Kit (Qiagen, Germantown, MD). Reverse transcription was performed using oligo(dT) primers and M-MLV reverse transcriptase (Promega, Madison, WI) at 42°C for 1 h according to the manufacturer’s instructions. qPCR was performed using SYBR Green I (Thermo Fisher Scientific) in a CFX Connect system (Bio-Rad Laboratories, Hercules, CA). PCR was performed using the following primers: Ren1, forward: 5′- ACAGTATCCCAACAGGAGAGACAAG-3′ and reverse: 5′- GCACCCAGGACCCAGACA-3′; Akr1b7, forward: 5′- CCTGTTGGATGCAAGGACTGA-3′ and reverse: 5′- CCTGCATTGCCAGATGGTAG-3′; and Rps14, forward: 5′- CAGGACCAAGACCCCTGGA-3′ and reverse: 5′- ATCTTCATCCCAGAGCGAGC-3′. Ren1 mRNA expression was normalized to Rps14 expression, and the changes in expression were determined by the ΔΔCt method and reported as relative expression compared with control mice.

Statistical Analysis

Data are presented as means ± SD. Statistical analysis was performed using GraphPad Prism 10 software (GraphPad Software, San Diego, CA). Comparisons between two groups were performed by two-tailed unpaired Student’s t test. Comparisons between more than two groups were performed using Welch’s ANOVA due to unequal variances between groups. Post hoc comparisons between each group’s mean were evaluated using Dunnett’s T3 multiple comparison test. P values of <0.05 were considered statistically significant.

RESULTS

Akr1b7^CreERT2 Mice Exhibit Normal Kidney Structure and Expression Levels of Akr1b7 Than Wild-Type Mice

To achieve temporal control of gene expression in renin cells, we generated Akr1b7^CreERT2 mice by inserting a P2A-CreERT2 knockin cassette immediately upstream of the TGA stop codon of the mouse Akr1b7 gene followed by the endogenous 3′-UTR using a homologous recombination approach. Figure 1A shows a schematic diagram of the mouse Akr1b7 gene, the targeting vector, and the final knockin allele. We confirmed the correct insertion and retention of the Akr1b7^CreERT2 knockin allele by PCR in ESCs (Fig. 1B) and the presence of a single insertion of the targeted allele (Fig. 1C). To determine whether this genetic manipulation had any effect on the endogenous locus, we performed in situ hybridization for Akr1b7 in the kidney of mice carrying the Akr1b7^CreERT2 allele. We observed no differences in the abundance or localization of Akr1b7 mRNA (Supplemental Fig. S1). In addition, Akr1b7^CreERT2 mice exhibited normal kidney morphology (Supplemental Fig. S2) and function (Supplemental Table 1).

Akr1b7^CreERT2 Mice Display Specific Cre Recombinase Activity in Renin Cells

We first looked at the pattern of Cre-dependent recombination during development and in the adult using Akr1b7^CreERT2;R26R^mTmG reporter mice where Akr1b7-expressing cells become labeled with GFP upon Cre recombinase activation with tamoxifen and all nonrecombined cells express tdTomato. Figure 2 shows the GFP and tdTomato expression in frozen kidney sections during postnatal life (P5–P60). Neonates received three intragastric injections of 50 µg tamoxifen at P1, P2, and P3 and were analyzed at P5. We observed the GFP signal along the arterioles, in the glomerular mesangium, and in JG cells (Fig. 2A). Adult mice (P30–P90) were injected with tamoxifen for 3 consecutive days and analyzed 2 days after the last injection. GFP signal was mainly confined to JG cells (Fig. 2B) and some arterioles (Supplemental Fig. S3, A and B). Akr1b7 is expressed in a few organs outside the kidney, most prominently in the cortex of the adrenal gland (29). At P30, we observed strong GFP expression in the cortex of adrenal glands where the signal was restricted to the zona fasciculata (Supplemental Fig. S3C) as reported for the endogenous Akr1b7 protein. Mice injected with corn oil (vehicle) did not show any GFP signal in the kidney (Fig. 2C) or the adrenal cortex (Supplemental Fig. S3B), indicating that Cre recombination did not occur in the absence of tamoxifen. Next, we examined the pattern of Cre-dependent recombination in conditions of physiological stress. Adult mice (P60–P120) were treated with captopril and low-salt diet for 8 days to induce the endocrine transformation of renin lineage cells, given tamoxifen on days 3, 4, and 5 of the experiment, and harvested 3 days after the last injection. As expected for captopril + low-salt diet-treated animals, we observed high GFP expression in JG cells, along the afferent arterioles, and the intraglomerular mesangium (Fig. 2D).

Figure 2.

Pattern of Cre-dependent recombination during postnatal life in Akr1b7^CreERT2 reporter mice. GFP and tdTomato expression in frozen sections of the kidneys from tamoxifen (A, B, and D)- and corn oil (vehicle) (C)-treated Akr1b7^CreERT2;R26R^mTmGmice. In this reporter mouse, Akr1b7-expressing cells become labeled with GFP upon Cre recombinase activation with tamoxifen. Nonrecombined cells express tdTomato. Left panels show GFP+ cells and right panels show GFP+ recombinant cells and tdTomato+ nonrecombinant cells. A: P5 neonates after administration of intragastric injections of tamoxifen at P1, P2, and P3. We observed GFP signal along the vasculature (yellow arrows) and in the intraglomerular mesangium (blue arrows) and juxtaglomerular (JG) areas (white arrows). n = 5. B: in adult mice, GFP signal was mainly confined to JG areas (white arrows). Shown is a kidney from a 30-day-old male mouse after administration of 3 consecutive tamoxifen injections and analyzed 2 days after the last injection. n = 3 females and 4 males; age: 30–90 days. C: no GFP signal was observed in the kidneys of mice injected with corn oil, indicating that Cre recombination did not occur in the absence of tamoxifen. Shown is a kidney from a male mouse. n = 2 females and 4 males; age: 30–90 days. D: 2-mo-old mouse treated with captopril and a low-salt diet for 8 days to induce the endocrine transformation of renin lineage cells. Tamoxifen was injected on days 3, 4, and 5 of the experiment and harvested 3 days after the last injection. GFP expression extended along the afferent arterioles (yellow arrows) and the intraglomerular mesangium (blue arrows). n = 3 females and 3 males; age: 60–120 days.

To better understand the spatial relationship between Cre-dependent recombination patterns and the renal vascular tree, we performed tissue clearing-based 3-D imaging using Akr1b7^CreERT2;R26R^tdTomato reporter mice and immunostaining for Acta2. Three-month-old control mice treated with captopril and a low-salt diet for 7 days received tamoxifen on days 1, 2, and 3 of the experiment. We visualized the morphology of the peripheral renal arterial tree, including afferent arterioles, and the pattern of Cre-dependent recombination in 3-D (Supplemental Videos S1 and S2). 3-D visualization clearly demonstrated that the tdTomato signal was restricted to the JG areas at the tips of afferent arterioles under basal conditions following tamoxifen injections. Under physiological challenge, tdTomato signal was more intense, and the labeled JG areas appeared enlarged compared with controls (Fig. 3B). Furthermore, tdTomato signal was clearly extended along the afferent arterioles in a characteristic ring-like striped pattern of recruited renin cells (Fig. 3B). In summary, the 3-D imaging experiments confirmed that Cre-dependent recombination correctly reflected the expression pattern of Akr1b7 and expected distribution of renin cells. Moreover, 3-D imaging provided a more comprehensive view of the kidney vasculature and distribution of renin cells under basal conditions and physiological stress in the intact tissue compared with classical 2-D sections.

Figure 3.

Tissue clearing and three-dimensional (3-D) imaging of kidneys from Akr1b7^CreERT2;R26R^tdTomato mice under normal conditions and physiological stress. Shown are Imaris software (Bitplane) reconstructed high-magnification 3-D images of kidneys from 3-mo-old mice under basal conditions (A) and after administration of captopril and low-salt diet for 7 days to induce the endocrine transformation of renin lineage cells (B). Mice received 3 consecutive intraperitoneal injections of tamoxifen (2 mg/20 g body wt) on days 1, 2, and 3 of the experiment. Red fluorescence indicates Akr1b7^CreERT2 expression upon tamoxifen administration, and FITC shows immunofluorescence for Acta2. Under basal conditions, tdTomato signal was clearly localized to the tip of the arterioles in the juxtaglomerular (JG) areas (white arrowheads). Under physiological stress, tdTomato signal was much brighter, and the labeled JG areas were enlarged compared with controls. In addition, the tdTomato signal extended into the arterioles as expected in recruited mice (white arrowheads). See also Supplemental Videos S1 and S2. AA, afferent arteriole; G, glomerulus. n = 1 male and 1 female.

Next, we investigated the pattern of Cre-dependent recombination during embryonic life using Akr1b7^CreERT2;R26R^mTmG reporter mice. Within the kidney, we observed GFP expression primarily along the vasculature, in the glomerular mesangium, and in JG areas of mature glomeruli (Fig. 4, A and B). In addition, we found GFP expression in the zona fasciculata of the embryonic adrenal cortex (Fig. 4, A and C). These results are consistent with the pattern of expression of Akr1b7 and renin at E18.5.

Figure 4.

Pattern of Cre-dependent recombination during embryonic life in Akr1b7^CreERT2 reporter mice. GFP and tdTomato expression in frozen sections of the kidneys and adrenal glands from tamoxifen-treated Akr1b7^CreERt2;R26R^mTmG E18.5 embryos is shown. Female mice received 2 consecutive intraperitoneal injections of 2 mg/40 g body wt tamoxifen on days 15.5 and 16.5 of pregnancy. A: low-magnification image of the kidney and adrenal gland. B and C: high-magnification images of the adrenal gland and kidney, respectively, of the areas indicated in A. GFP was expressed in the zona fasciculata of the embryonic adrenal cortex. In the kidney, we observed GFP signal along the arteries and arterioles (yellow arrows) in the glomerular mesangium (blue arrows) and in JG areas (white arrows) of mature glomeruli. n = 5.

In all, our results indicate that, in Akr1b7^CreERT2 mice, Cre-dependent recombination occurs in the same areas reported for endogenous Akr1b7 during development and in the adult under basal conditions and physiological stress. In addition, Cre activity is tightly regulated by tamoxifen.

Cre and Renin Highly Colocalize in the Kidney of Akr1b7^CreERt2 Mice in the Basal State and Under Physiological Stress

Next, we asked whether reporters for Akr1b7^CreERT2 and renin colocalize in the kidney of these mice. To answer this question, we generated Akr1b7^CreERT2;Ren1^cYFP;R26R^tdTomato reporter mice where cells actively expressing renin are labeled with YFP, as previously shown (24), and cells undergoing Cre-dependent recombination are labeled with tdTomato upon tamoxifen administration. We found that tdTomato and renin (YFP) were highly colocalized under basal conditions in JG cells and under physiological stress in JG cells, along the arterioles, and in the intraglomerular mesangium (Fig. 5A). These observations strongly support that the reporters in our model follow exactly the same pattern of expression as endogenous Akr1b7 and renin. The JGAi, defined as the number of renin positive JG areas ÷ total number of glomeruli × 100, is an established method to quantitate and compare levels of renin expression in mice (30, 31). To evaluate the degree of colocalization of tdTomato and renin (YFP) in Akr1b7^CreERT2 mice, we compared the JGAi using each reporter independently. We found no significant differences between the JGAi for renin (YFP)-positive cells or Cre (tdTomato)-positive cells (Fig. 5B). In addition, all YFP (renin)-positive cells were positive for tdTomato, suggesting that Akr1b7^CreERT2 mice exhibit a high rate of recombination.

Figure 5.

Cre and Renin highly colocalize in the kidney of Akr1b7^CreERt2 mice in the basal state and under physiological stress. A: YFP and tdTomato expression in frozen sections of kidneys from tamoxifen-treated Akr1b7^CreERt2/+;Ren1^cYFP;R26R^tdTomato/+ reporter mice. Tamoxifen (2 mg/20 g body wt) was administered intraperitoneally on 3 consecutive days, and mice were studied 3 days after the last injection. A group of mice were treated with captopril (0.5 mg/mL H₂O) + low salt (0.1%) diet for 8 days, and tamoxifen was administered on days 3, 4, and 5 of recruitment. Cre (tdTomato) and Renin (YFP) highly localized to the same areas: under basal conditions (left) in juxtaglomerular (JG) cells (green arrows) and under physiological stress (right) in JG cells (green arrows), along the arterioles (yellow arrows), and in the intraglomerular mesangium (blue arrows). All YFP (renin)-positive cells were positive for tdTomato. Basal state: n = 3 females and 2 males; physiological stress: n = 1 female and 1 male. Age: 35–65 days old. B: juxtaglomerular area index (JGAi) calculated using YFP⁺ cells or tdTomato⁺ cells in the basal state. We found no significant differences between the JGAi for renin (YFP)-positive cells and Cre (tdTomato)-positive cells. Unpaired Student’s t test P > 0.05. n = 3 females and 2 males.

These results clearly show that in Akr1b7^CreERT2;Ren1^cYFP;R26R^tdTomato mice, the reporters for Cre and renin are highly colocalized, confirming the renin cell specificity and high rate of recombination of our model.

Renin Can Be Efficiently Deleted in the Kidney of Adult Akr1b7^CreERT2 Mice

To study the effect of renin deletion in the adult, we used the Akr1b7^CreERT2 model to generate Akr1b7^CreERT2;Ren1^cFl/−;R26^mTmG mice. In this mouse, the Ren1^c floxed allele can be deleted, and Akr1b7^CreERT2 expressing cells can be simultaneously labeled with GFP upon tamoxifen administration in a time-controlled manner.

First, we conducted a short-term deletion study to compare tamoxifen-treated Akr1b7^CreERT2/+;Ren1^cFl/−;R26^mTmG/+ (Renin cKO^GFP) with Akr1b7^CreERT2/+;Ren1^c+/−;R26^mTmG/+ controls. Renin hemizygous mice are normal and therefore can be used as controls in this study (9). Adult mice received tamoxifen intraperitoneally on 5 consecutive days and were analyzed 10–15 days after the last injection. A group of mice was treated with captopril and low-salt diet for the last 8 days of the experiment to induce recruitment of renin cells. We observed a marked decrease in renin immunostaining in the kidneys of Renin cKO^GFP mutant mice compared with controls under basal conditions and physiological stress (Fig. 6A). Semiquantitative analysis of renin-positive immunostaining areas showed a significant decrease in areas of renin immunostaining in mutant mice compared with control animals under basal conditions (control-basal: 0.159 ± 0.040%, n = 8; mutant-basal: 0.040 ± 0.018, n = 9; P < 0.001) (Fig. 6B). Similarly, renin immunostaining under recruitment conditions was significantly lower in Renin cKO^GFP mutant mice compared with controls (control-recruited: 0.441 ± 0.139%, n = 6; mutant-recruited: 0.069 ± 0.040, n = 5; P < 0.01) (Fig. 6B). Under physiological stress, control mice exhibited significantly higher levels of renin expression compared with basal conditions (control-basal: 0.159 ± 0.040%, n = 8; control-recruited: 0.441 ± 0.139%, n = 6; P < 0.05) (Fig. 6B). The renin levels in mutant recruited mice were not significantly different from those in nonrecruited animals (mutant-basal: 0.040 ± 0.018, n = 9; mutant-recruited: 0.069 ± 0.040, n = 5, P > 0.05) (Fig. 6B). In addition to the changes at the protein level, we observed a significant decrease in renin mRNA measured by qPCR in Renin cKO^GFP mutant mouse kidney cortices compared with controls in both the basal state (control-basal: 0.92 ± 0.26, n = 8; mutant-basal: 0.20 ± 0.1; n = 8; P < 0.001) and under captopril + low-salt diet (control-recruited: 5.27 ± 1.65, n = 5; mutant-recruited: 0.49 ± 0.21, n = 5; P < 0.05) (Fig. 7A). As expected, we found that under physiological stress, control mice exhibited significantly higher levels of renin mRNA compared with basal conditions (control-basal: 0.92 ± 0.26, n = 8; control-recruited: 5.27 ± 1.65, n = 5; P < 0.05). Renin mRNA levels in mutant recruited mice were not significantly different than in nonrecruited animals (mutant-basal: 0.20 ± 0.1, n = 8; mutant-recruited: 0.49 ± 0.21, n = 5; P = 0.146). In addition, we found that circulating renin levels followed a similar trend: control-basal: 53,714 ± 27,908, n = 7, versus mutant-basal: 12,466 ± 7,035, n = 9, P < 0.05; control-recruited: 309,954 ± 122,023, n = 5, versus mutant-recruited: 44,375 ± 12,955, n = 5, P < 0.05; control-basal: 53,714 ± 27,908, n = 7, versus control-recruited: 309,954 ± 122,023, n = 5, P < 0.05; and mutant-basal: 12,466 ± 7,035, n = 9, versus mutant-recruited: 44,375 ± 12,955, n = 5, P < 0.05 (Fig. 7B). In line with our renin mRNA and protein results, we found that renin mutant mice exhibited a significant decrease in arterial blood pressure under basal conditions compared with controls: control-basal: 82.7 ± 2.6; n = 7, versus mutant-basal: 68.9 ± 5.3, n = 9; P < 0.005 (Fig. 7C).

Figure 6.

Renin immunostaining of kidney sections from adult control and Akr1b7^CreERT2/+; Ren1^cFl/−; R26R^mTmG (Renin cKO^GFP) mice under basal conditions and physiological stress. A: results showing the normal renin distribution (brown, yellow arrows) at the entrance of the glomeruli in control kidneys under basal conditions, whereas a marked decrease in renin distribution and storage was seen in the kidneys of Renin cKO^GFP mice. In response to captopril + low-salt diet, control mice showed renin signal along the arterioles (red arrows) and in the mesangium (green arrow). The response of Renin cKO^GFP mice was greatly reduced with no visible extension into arterioles and mesangium. B: semiquantitative assessment of areas of renin-positive immunostaining. We observed a significant decrease in renin immunostaining in mutant mice compared with control animals under basal conditions (control-basal: n = 7 females and 1 male; mutant-basal: n = 6 females and 3 males; ***P < 0.001). Similarly, renin immunostaining under recruitment conditions was significantly lower in mutant mice compared with controls (control-recruited: n = 2 females and 4 males; mutant-recruited: n = 2 females and 3 males; **P < 0.01). Under physiological stress, control mice exhibited significantly higher levels of renin expression compared with basal conditions (control-basal vs. control-recruited; *P < 0.05). Although renin levels in mutant-recruited mice were higher than in nonrecruited animals, the difference was not significant (mutant-basal vs. mutant-recruited, P > 0.05). Analysis was by Welch ANOVA followed by Dunnett’s T3 multiple comparisons test. Age of mice: 60–120 days in all groups.

Figure 7.

Renin mRNA expression, plasma renin levels, and blood pressure in Renin cKO mutant mice. Tamoxifen-treated Akr1b7^CreERT2/+;Ren1^cFl/−;R26^mTmG/+ (Renin cKO^GFP) mice were compared with Akr1b7^CreERT2/+;Ren1^c+/−; R26R^mTmG control mice under basal conditions and physiological stress. Adult (2- to 4-mo-old) mice received tamoxifen (2 mg/20 g body wt ip) on 5 consecutive days and were analyzed after 10–15 days. A group of mice was treated with captopril (0.5 mg/mL) + low-salt (0.1%) diet for the last 8 days of the experiment to induce recruitment of renin cells. A: renin mRNA levels in kidney cortices. We observed a significant decrease in renin mRNA levels in Renin cKO^GFP mutant mice compared with controls in both the basal state (control-basal: n = 7 females and 1 male; mutant-basal: n = 4 females and 4 males; ***P < 0.001) and under physiological threat (control-recruited: n = 2 females and 3 males; mutant-recruited: n = 2 females and 3 males; *P < 0.05). Under physiological stress, control mice exhibited significantly higher levels of renin mRNA compared with basal conditions (control-basal: vs. control-recruited; *P < 0.05). Renin mRNA levels in mutant recruited mice were not significantly higher than in nonrecruited animals (mutant-basal vs. mutant-recruited; ns, nonsignificant; P = 0.146). B: plasma renin. Circulating renin levels followed a similar trend: control-basal: n = 6 females and 1 male vs. mutant-basal: n = 6 females and 3 males; *P < 0.05; control-recruited: n = 2 females and 3 males vs. mutant-recruited: n = 2 females and 3 males; *P < 0.05; control-basal vs. control-recruited: *P < 0.05; mutant-basal vs. mutant-recruited: *P < 0.05. Analysis was by Welch ANOVA followed by Dunnett’s T3 multiple comparisons test (A and B). C: arterial blood pressure. Renin mutant mice exhibited a significant decrease in arterial blood pressure under basal conditions compared with controls: control-basal: n = 6 females and 1 male vs. mutant-basal: n = 6 females and 3 males; ***P < 0.005. Analysis was by a two-tailed unpaired Student’s t test. MBP, mean blood pressure.

These results indicate that renin can be efficiently deleted in adult mice using our Akr1b7^CreERT2 conditional model.

Mice Develop Concentric Vascular Hypertrophy After Akr1b7^CreERT2-Mediated Renin Deletion in the Adult

Next, we sought to use the Akr1b7^CreERT2 model to determine the effects of long-term deletion of renin in the adult mouse, specifically whether the mice develop concentric vascular hypertrophy upon renin deletion. In these experiments, we administered tamoxifen (2 mg/20 g body wt) intraperitonially on 5 consecutive days to Akr1b7^CreERT2/+;Ren1^cFl/−;R26^mTmG/+ (Renin cKO^GFP) and Akr1b7^CreERT2/+;Ren1^c+/−;R26^mTmG/+ controls at 1 mo of age on 5 consecutive days and studied the mice at 6 mo. Immunostaining for Acta2 revealed thicker arteries and arterioles in Renin cKO^GFP kidneys (Fig. 8A). This phenotype is similar to those observed with mutations of the RAS genes or chronic treatment with RAS inhibitors. In addition, renin cells could be tracked by the expression of GFP even when they were negative for renin by immunostaining (Fig. 8B), allowing the tracking and isolation of these abnormal cells.

Figure 8.

Long-term effect of renin deletion in the adult using the conditional Akr1b7^CreERT2 model. Tamoxifen (2 mg/20 g body wt) was administered intraperitoneal to Akr1b7^CreERT2/+;Ren1^cFl/−;R26^mTmG/+ (Renin cKO^GFP) and Akr1b7^CreERT2/+;Ren1^c+/−;R26^mTmG/+ controls at 1 mo of age for 5 consecutive days, and the mice were studied at 6 mo of age. A, top: immunostaining for Acta2 of kidney sections from tamoxifen-treated Renin cKO^GFP and control mice. Renin cKO^GFP kidneys showed thicker intrarenal arterial and arteriolar (red arrows) walls than controls. A, bottom: renin immunostaining of adjacent sections. Control mice showed the normal renin distribution (brown, yellow arrows) at the entrance of the glomeruli in control kidneys under basal conditions, whereas Renin cKO^GFP exhibited a marked decrease in renin immunostaining. B: GFP and tdTomato expression in frozen sections of kidneys from the same animals shown in A. Renin cells can be tracked by the expression of GFP even when they are unable to express renin, allowing the tracking and isolation of these abnormal cells. Control, n = 1 female and 2 males. Renin cKO^GFP, n = 5 females and 2 males.

DISCUSSION

In this study, we describe the first knockin inducible Cre recombinase mouse model for gene targeting specifically in renin cells. This model consists of a knockin P2A-CreERT2 cassette inserted immediately upstream of the TGA stop codon of the mouse Akr1b7 gene. The Akr1b7^CreERT2 model has significant advantages over existing renin cell Cre lines. First, by following a knockin experimental approach, we avoided potential problems associated with transgenic lines that are prone to spurious and/or ectopic gene expression. There is one reported tamoxifen-inducible line where Cre is driven by a Ren1^c promoter as a BAC transgene, but, to our knowledge, it has not yet been fully characterized (16). In addition, a BAC transgene-based Tet-on inducible triple-transgenic LacZ line has been reported (17). This model is better characterized and effective at knocking out genes in renin lineage cells; however, potential changes in gene expression due to insertion effects and/or copy number of the transgene cannot be ruled out. Second, by using an inducible model, we were able to circumvent the confounding developmental effects introduced by noninducible Cre recombinase lines when the effects of gene deletion are studied in the adult animal. Third, by selecting the independent marker of renin cells, Akr1b7, we ensured that renin expression is not affected by the incorporation of the CreERT2 cassette at the renin locus. In the kidney, Akr1b7 is expressed exclusively in renin cells. Akr1b7 is coexpressed with renin during development and in adults under different physiological and pathological conditions (21, 22). Although Akr1b7 and renin are regulated by common factors (22), the expression of one gene does not influence the expression of the other (22, 32). As a result, Akr1b7 can be used as a marker for cells that are programmed for the renin phenotype even in the absence of renin (22). Unlike renin, apart from the adrenal gland, Akr1b7 is expressed at relatively low levels only in a few organs outside the kidney (33).

The placement of the P2A-CreERT2 cassette at the Akr1b7 locus had no adverse effects on the mice. Both heterozygous and homozygous Akr1b7^CreERT2 mice were viable, fertile, normal in size, and did not display any gross physical or behavioral abnormalities. The function of Akr1b7 remains elusive. Akr1b7 is dispensable for mouse normal development and reproduction (34). Moreover, despite its specific localization in renin cells (21, 32), Akr1b7 does not affect renin expression, localization, and release under normal conditions and in response to physiological stress (32). For these reasons, we were unable to evaluate the functional integrity of Akr1b7 in Akr1b7^CreERT2 mice. Nevertheless, we observed no difference in the levels and localization of Akr1b7 mRNA by in situ hybridization in the kidney of mice carrying the Akr1b7^CreERT2 allele, indicating that the genetic manipulation did not affect the expression of endogenous Akr1b7.

Since the P2A-CreERT2 cassette was inserted upstream of the TGA of the endogenous Akr1b7 locus, it is expected to be regulated by the same elements that control Akr1b7 expression. Our results show that the fluorescence signal from heterozygous Akr1b7^CreERT2/+;R26R^mTmG/+ and Akr1b7^CreERT2/+;R26R^tdTomato/+ reporter mice clearly aligned with the pattern of endogenous Akr1b7 protein expression during embryonic life and in adults under basal conditions when homeostasis was compromised (21, 22). Notably, within the kidney cortex, Akr1b7^CreERT2 mice exhibited Cre recombinase activity specifically in the JG areas. The embryonic and early postnatal Akr1b7^CreERT2 expression pattern, encompassing arteries, arterioles, intraglomerular mesangium, and JG areas, underscores the importance of the Akr1b7^CreERT2 model as a unique tool to perform inducible gene manipulation in renin cells during development.

In addition to being cell specific, an optimal inducible model must be tightly regulated. Since CreERT2 is constitutively expressed, recombination at LoxP sites might happen even in the absence of tamoxifen. Our results using Akr1b7^CreERT2/+;R26R^mTmG/+ mice showed absence of recombination when tamoxifen was not given to the animals, indicating that the system is tightly regulated based on this particular reporter. In addition, in adult animals treated with tamoxifen, the GFP signal was confined strictly to JG cells with no expression in other portions of the vasculature, which would have resembled the embryonic pattern, thus indicating that recombination did not occur before induction with tamoxifen.

Akr1b7 is expressed in a few organs outside the kidney and, in most cases, at very low levels (33). However, Akr1b7 is highly expressed in the murine adrenal gland where it is confined to the zona fasciculata of the cortex (29, 34, 35). The cells in this area of the adrenal cortex synthesize glucocorticoids and androgens, and it has been suggested that Akr1b7 may be involved in the detoxification of harmful aldehydes produced during the synthesis of steroids (35). In the adrenal gland of tamoxifen-induced Akr1b7^CreERT2/+;R26R^mTmG/+ reporter mice, we found high levels of GFP expression exclusively in the zona fasciculata of the cortex. This finding suggests that the Akr1b7^CreERT2 model may be a valuable resource for lineage tracing and gene targeting in the glucocorticoid- and androgen-producing cells of the adrenal gland.

We used double reporter mice, Akr1b7^CreERT2;Ren1^cYFP;R26R^tdTomato, to evaluate the degree of colocalization of renin and Akr1b7^CreERT2 and the rate of recombination after tamoxifen induction. The expression of YFP driven by the Ren1^c-YFP transgene during development, in adult life, and in response to physiological stress (i.e., recruitment) precisely follows the pattern of renin expression visualized by immunostaining (24). Our results show that all YFP (renin)-expressing cells were positive for tdTomato (Akr1b7^CreERT2), strongly suggesting that Akr1b7^CreERT2 confers not only renin cell-specific Cre activity but also a high rate of recombination after tamoxifen induction. This conclusion is also supported by the lack of significant difference between the JGAi determined using YFP⁺ cells and tdTomato⁺ cells. Our observation that both reporters were highly colocalized to JG cells, along the arterioles, and in the intraglomerular mesangium under conditions of physiological stress suggests that Akr1b7^CreERT2 model can also be used for the genetic manipulation of cells during reacquisition of the renin phenotype.

The activation of reporter alleles using Cre-LoxP systems does not indicate that the recombination necessarily will take place at other floxed loci in the cell, a process known as nonparallel recombination (36). It has been reported that the rate of Cre-mediated recombination depends on several factors, including the distance between the LoxP sites, the chromosomal location, and epigenetic context of the floxed alleles (34 and references therein). To investigate the potential use of the Akr1b7^CreERT2 model for genetic manipulation of genes beyond well-established reporter floxed alleles, we generated Akr1b7^CreERT2/+;Ren1^cFl/+;R26^mTmG/+ mice to target renin in the adult animal. Compared with controls, renin mutant animals exhibited significantly lower expression of renin in the kidney at both mRNA and protein levels, with only a few renin-positive cells detectable by immunostaining. In addition, mutant mice exhibited significantly lower levels of circulating renin and low blood pressure. These results demonstrate that renin can be efficiently deleted in the adult using our Akr1b7^CreERT2 conditional model. Furthermore, the response of renin mutant mice to a physiological challenge, i.e., captopril + low-salt diet, was severely compromised. When homeostasis is threatened (e.g., dehydration, hemorrhage, hypoxemia, and/or deletion/inhibition of RAS genes), Akr1b7 is upregulated, along with renin, in recruited arteriolar smooth muscle cells, mesangial cells, and interstitial cells (22), The lack of renin expression in these cells is probably the result of further renin deletion as Akr1b7 gets activated during recruitment in renin-mutant mice.

Our group and others have shown that experimental or spontaneous mutations of any of the RAS genes or long-term treatment with inhibitors of the RAS in mammals, including humans, lead to the development of concentric arterial and arteriolar hypertrophy (7–15). We found that mice with ablation of renin cells using diphtheria toxin (37) or conditional deletion of integrin β1 (15) do not develop kidney vascular hypertrophy, indicating that renin cells per se are responsible for the vessel thickening. Furthermore, we demonstrated that vascular hypertrophy in renin-deficient mice is not the result of the proliferation of renin cells (38) but instead is due to their transformation from an endocrine to an invasive matrix-secretory phenotype accompanied by the inward accumulation of smooth muscle cells (15). These observations indicate that renin cells are not proliferative, and they increase their numbers by phenotypic switching and cell transformation. Nevertheless, the underlying mechanisms leading to this vascular pathology remain unknown. In this study, we show that deletion of renin during adulthood leads to the development of concentric vascular hypertrophy. This morphological phenotype is similar to the one observed in animals with deletion of renin globally or constitutively in renin cells (9, 13). However, the latter show additional renal abnormalities, including papillary atrophy, underdeveloped medulla, hydronephrosis, interstitial fibrosis, focal glomerulosclerosis, and perivascular infiltration of mononuclear cells. These morphological alterations can be the result of developmental changes caused by the absence of renin expression throughout the life of the animals and may alter the vascular phenotype. Our inducible model offers the possibility to study the vascular hypertrophy phenotype independently of these confounding factors. In addition, our model allows the labeling of the abnormal vascular cells for isolation, which would enable the study of the molecular pathways and cellular interactions underlying the observed vascular hypertrophy. Further studies with these animals will help elucidate the specific role of renin deficiency in the development of concentric vascular hypertrophy. In addition, unraveling the signaling pathways triggered by renin deficiency may uncover novel therapeutic targets for conditions associated with vascular hypertrophy.

In conclusion, the Akr1b7^CreERT2 mouse showed high levels of tamoxifen-induced recombination that is renin cell specific and tightly regulated. The Akr1b7^CreERT2 model constitutes a powerful tool for the temporal and spatial control of target genes and for lineage tracing during development and in adult mice.

DATA AVAILABILITY

Data will be made available upon reasonable request.

SUPPLEMENTAL DATA

Supplemental Figs. S1–S3, Supplemental Table S1, and Supplemental Videos S1 and S2: https://doi.org/10.6084/m9.figshare.26090023.v2.

GRANTS

This work was funded by National Institutes of Health Grants P50DK096373 and R01DK116718 (to R.A.G.), R01HL148044 (to M.L.S.S.-L.), and P01HL084207 (to C.D.S.).

DISCLOSURES

Maria Luisa S. Sequeira-Lopez is an associate editor of the American Journal of Physiology-Renal Physiology but was not involved and did not have access to information regarding the peer-review process or final decision of this article. An alternative editor oversaw the peer-review and decision-making process for this article. None of the other authors has any conflicts of interest, financial or otherwise, to disclose.

AUTHOR CONTRIBUTIONS

S.M., C.D.S., M.L.S.S.-L., and R.A.G. conceived and designed research; S.M., M.Y., L.F.d.A., and H.Y. performed experiments; S.M., M.Y., L.F.d.A., J.P.S., and H.Y. analyzed data; S.M., M.Y., L.F.d.A., J.P.S., H.Y., M.L.S.S.-L. and R.A.G. interpreted results of experiments; S.M., M.Y., L.F.d.A., and J.P.S. prepared figures; S.M., M.Y., and L.F.d.A. drafted manuscript; S.M., M.Y., L.F.d.A., J.P.S., H.Y., C.D.S., M.L.S.S.-L., and R.A.G. edited and revised manuscript; S.M., M.Y., L.F.d.A., J.P.S., H.Y., C.D.S., M.L.S.S.-L., and R.A.G. approved final version of manuscript.