Targeted disruption of Adamts16 gene in a rat genetic model of hypertension

Abstract

A disintegrin-like metalloproteinase with thrombospondin motifs–16 (Adamts16) is an important candidate gene for hypertension. The goal of the present study was to further assess the candidacy of Adamts16 by targeted disruption of this gene in a rat genetic model of hypertension. A rat model was generated by manipulating the genome of the Dahl Salt–sensitive (S) rat using zinc-finger nucleases, wherein the mutant rat had a 17 bp deletion in the first exon of Adamts16, introducing a stop codon in the transcript. Systolic blood pressure (BP) of the homozygous Adamts16^mutant rats was lower by 36 mmHg compared with the BP of the S rats. The Adamts16^mutant rats exhibited significantly lower aortic pulse wave velocity and vascular media thickness compared with S rats. Scanning electron and fluorescence microscopic studies indicated that the mechanosensory cilia of vascular endothelial cells from the Adamts16^mutant rats were longer than that of the S rats. Furthermore, Adamts16^mutant rats showed splitting and thickening of glomerular capillaries and had a longer survival rate, compared with the S rats. Taken together, these physiological observations functionally link Adamts16 to BP regulation and suggest the vasculature as the potential site of action of Adamts16 to lower BP.

Despite strong evidence that susceptibility or resistance to the development of hypertension is heritable, the identification of genetic variants that cause blood pressure (BP) to rise into a hypertensive state has remained difficult (1, 2). Classic genetic mapping and association studies in both humans and in rats point to several genetic elements as potential candidates causing hypertension (3, 4). Most of the prioritized candidate genes for hypertension await functional assessments.

Linkage analysis in the Quebec Family Study identified a quantitative trait locus (QTL) for systolic BP on human chromosome 5p15 (5). The corresponding comparative segment of human chromosome 5p15 on rat chromosome 1 is also linked to a BP QTL in rats (6). Improved resolutions of this locus in rats were obtained through iterative substitution mapping using strains differentially susceptible to the development of hypertension (6–10). A disintegrin-like metalloproteinase with thrombospondin motifs–16 (Adamts16), which was the only known gene with exonic variants within the highly resolved congenic interval, was prioritized as a candidate BP quantitative trait gene (QTG) (8). More importantly, following the congenic mapping study in rats, human allelic variants of Adamts16 were confirmed as being associated with BP in two independent cohorts, one of which was the Quebec Family Study (8). Taken together, all these studies point to Adamts16 as a prominent candidate locus linked to BP control across two species. However, due to the limitations of recombination frequencies, both the linkage and substitution mapping studies in rats cannot validate Adamts16 as the BP QTG because of the presence of other candidate variants within the linked or introgressed flanking genomic segments, respectively. Until recently, targeted gene functional studies were impossible in the rat due to technical limitations of site-targeted gene editing strategies. In 2009, a technique of zinc-finger nuclease (ZFN)–mediated targeted gene disruption was successfully described in the rat (11). The goal of the current study was to apply this targeted gene-disruption strategy to assess the functional link between Adamts16 and BP.

In this study we describe the creation of a rat Adamts16^mutant model using the ZFN method. The characterization of this model demonstrates the functional link between Adamts16 and BP. Further, the vasculature was identified as one of the potential target tissues affected by the disruption of the Adamts16 gene. In summary, this study applied a gene-targeting approach for direct physiological assessement of a candidate gene previously mapped to a high resolution using linkage and substitution mapping in rats.

Results

Targeted Mutation of Adamts16 Decreased BP.

The Adamts16^mutant rats had a shorter (290 bp) PCR-amplified genomic DNA fragment compared with the Dahl Salt–sensitive (S) (307 bp) genomic fragment (Fig. 1A). The 17 bp deletion within exon 1 of Adamts16 of these PCR fragments was confirmed by sequencing (Fig. 1B). cDNA isolated from the Adamts16^mutant rats also reflected this 17 bp deletion (Fig. 1B). Both the homozygous and heterozygous Adamts16^mutant rats appeared healthy and gained weight similar to the S rats. BP of the homozygous Adamts16^mutant rats, as measured by the tail-cuff method, were significantly lower by 36 mmHg compared with the BP of the S rats (184 ± 1.16 versus 220 ± 3.64 mmHg, P < 0.001) (Fig. 2A). Similarly, the BP of the heterozygous Adamts16^mutant rats was also lower than the BP of the S rats by 14 mmHg (198 ± 4.94 versus 220 ± 3.64 mmHg, P < 0.01) (Fig. 2A). To further confirm these observations, rats were surgically implanted with radiotransmitters and BP was continuously recorded by telemetry. During all of the 5 d of observation, both systolic and diastolic pressure of both homozygous and heterozygous Adamts16 ^mutant rats were consistently lower than that of the S rats (Fig. 2 B and C).

Fig. 1.

Screening animals for ZFN-targeted mutation at the Adamts16 locus. (A) Representative agarose gel image of PCR-amplified genomic DNA from an intercross of heterozygous Adamts16^mutant carriers. The normal allele of the Dahl S rat is 307 bp. Rats with alternate allele size of 290 bp are representative mutants. M, males; F, females; 95745–95769, rat tag numbers. (B) Representative sequencing results from the 307 bp and 290 bp bands shown in panel A confirmed the 17 bp deletion in the genomic DNA. The same 17 bp deletion was confirmed by amplification and sequencing of cDNA from the homozygous mutants. TGA in red depicts the stop codon in exon 1.

Fig. 2.

BP measurements of homozygous Adamts16^mutant (n = 6), heterozygous Adamts16^mutant (n = 5), and S (n = 7) rats. Levels of statistical significance for all data were analyzed by ANOVA followed by Tukey’s test. (A) Mean systolic BP effect ± SEM by the tail-cuff method. (B and C) BP measures of the same animals monitored after surgical implantation of radiotelemetry transmitters. Data plotted are the recordings obtained once every 5 min continuously for 24 h and averaged for 4 h intervals; **P < 0.01, ***P < 0.001.

Decreased Pulse Wave Velocity and Media Thickness in Adamts16^mutant Rats.

In accordance with the changes in BP, the heart weight–body weight ratios of the Adamts16^mutant rats were lower than that of the S rats (Fig. 3A). When the animals were examined by echocardiography, the left ventricular relative wall thickness (RWT) and aortic pulse wave velocity (PWV) of both the homozygous and heterozygous Adamts16^mutant rats were lower than that of the S rats (Fig. 3 B and C). This was interesting because PWV is a measure of arterial stiffness (12). Elevated PWV and thereby increased arterial stiffness is an independent marker of cardiovascular risk in human hypertension (13). Therefore, we examined the structure of the abdominal aorta more closely by second harmonic generation microscopy. As seen in Fig. 4A, the media area of aortae from the homozygous Adamts16^mutant rats were lower than that of the control S rat.

Fig. 3.

Cardiac parameters of homozygous Adamts16^mutant (n = 6), heterozygous Adamts16^mutant (n = 5), and S (n = 7) rats. (A) Relative heart weight. (B) Left ventricular relative wall thickness. (C) Aortic pulse wave velocity. Left Ventricular (LV) RWT and aortic PWV were determined by echocardiography. Levels of statistical significance for all data were analyzed by ANOVA followed by Tukey’s test; *P < 0.05, ***P < 0.001.

Fig. 4.

Vascular features of homozygous Adamts16^mutant (n = 6) and S (n = 7) rats. (A) Quantitation of media area. (B) Images of vascular endothelial primary cilia. From left first and second panels, representative images of primary cilia from vascular endothelia observed by immunofluorescence microscopy. Arrows indicate green fluorescently labeled cilia. Cilia were stained with monoclonal acetylated α-tubulin (green), whereas the nuclei were stained with DAPI (blue). From left third panel, scanning electron micrographs of cilia in vascular endothelia; arrows point to cilia. (C) Quantitation of cilia length from scanning electron micrograph images. More than 200 cilia were counted in each strain; **P < 0.01, ***P < 0.001.

Endothelial Cells from the Adamts16^mutant Rats Have Elongated Cilia.

To further evaluate phenotypic differences in the vasculature, we examined cilia within the endothelial cells of arteries. Cilia are microtubule-based, antenna-like organelles projecting from the apical surface of endothelial cells that are directly associated with collagen in the extracellular matrix (ECM). Immunohistochemical observation was suggestive of longer cilia within the endothelial cells of Adamts16^mutant rats compared with the S rats (Fig. 4B). This initial observation was further confirmed by scanning electron microscopic images of cilia (Fig. 4B). Cilia from the Adamts16^mutant rats were prominently elongated in the Adamts16^mutant rats compared with the cilia in the S rats (Fig. 4C).

Targeted Mutation in Adamts16 Increased Proteinuria of the S Rats.

Urinalysis was conducted to assess renal function of the Adamts16^mutant rats. Total protein content of the 24 h urine collected from Adamts16^mutant rats was significantly higher than that of the S rats (Fig. 5A), reminiscent of increased damage to kidney function in the Adamts16^mutant rats.

Fig. 5.

Renal parameters of experimental groups of animals. (A) Total 24 h urine protein was assessed in S (n = 15) and homozygous Adamts16^mutant rats (n = 12) as described in the Materials and Methods section. (B) Representative Jones silver-stained kidney sections. Yellow arrows in the top panel point to thickened arteries. White arrow in the bottom panel center image from the kidney of a homozygous Adamts16^mutant rat shows split glomerular capillary. (C) Quantitation of intrarenal artery thickness shown in the top panel of B; n = 3/group; ***P < 0.0001.

Adamts16^mutant Rats Exhibit Splitting and Thickened Glomeruli Capillaries.

To evaluate the histology of the kidneys, sections from S and Adamts16^mutant rats were examined (Fig. 5B). Jones silver staining showed thickened arteries in S rats, with extremely small lumen, whereas both homozygous and heterozygous Adamts16^mutant rats did not show thickening of the arteries (Fig. 5 B and C). Further, splitting of the glomeruli capillaries was observed in Adamts16^mutant rats (Fig. 5B). These results also point to vascular alterations within the kidneys of Adamts16^mutant rats.

Adamts16^mutant Rats Survive Longer than the S Rats.

Decreased BP and increased proteinuria were contrasting features of pathology of the Adamts16^mutant rats. To assess the outcome of these two dichotomously opposing phenotypes, the end points of survival of the rats were compared. The mean survival of the Adamts16^mutant rats was significantly (P < 0.05) longer than that of the S rats (Fig. 6).

Fig. 6.

Kaplan–Meir plot. Data from rats (n = 10 S rats and n = 10 Adamts16^mutant rats) in the survival study were plotted, and median survival of each group was calculated using the Graphpad prism software. Median survival of the Adamts16^mutant rats was greater than the survival of the S rats; *P < 0.05.

Discussion

Adamts16 is one of 19 members of the ADAMTS family of metalloproteinases (14). Known functions of the ADAMTS proteases include processing of procollagens and von Willebrand factor as well as cleavage of aggrecan, versican, brevican, and neurocan (14, 15). They have been demonstrated to have important roles in connective tissue organization, coagulation, inflammation, arthritis, angiogenesis, and cell migration (14). Previously, our positional cloning strategy in rats pointed to Adamts16 as one of only two positional candidate genes for BP within a highly resolved 804.6 kb region of the rat genome and demonstrated that variants of Adamts16 are associated with human essential hypertension (8). The results of the present study, using a targeted mutant of Adamts16, provide strong evidence to suggest that the function of this gene is linked to BP regulation. Further, the observation of lowering of hypertension in Adamts16^mutant rats coupled with large structural alterations in vessel architecture point to structural integrity of the vasculature as being tightly linked to the function of Adamts16 in BP regulation.

Using the same gene targeting strategy described in our study, recently, one of the subunits of reduced nicotinamide adenine dincleotide phosphate (Nadph) oxidase, p^67Phox, was reported to play an important role in salt-sensitive hypertension (16). Unlike the known function of the gene product of p^67Phox in scavenging reactive oxygen species (16), the function of Adamts16 is not known. Therefore, to bridge the gap in knowledge between Adamts16 and BP regulation, it will now be important to identify whether Adamts16 is indeed a metalloproteinase, and if so, it will also be important to identify specific substrate/s and interacting partners of Adamts16 in the vasculature.

The S rat model of hypertension is also a model of renal disease, which is primarily characterized by proteinuria. Increased proteinuria observed in the Adamts16^mutant rats compared with S rats suggests that the function of Adamts16 may not be limited to the vasculature. Future renal studies will be required to determine the extent to which Adamts16 might be affecting BP by way of effects on glomerular filtration rate, renal hemodynamics, and renal handling of sodium.

The ZFN-based targeted gene editing strategy for Adamts16 constitutes an important advancement and addition to the traditional congenic mapping of Adamts16 as a gene regulating BP. However, because the ZFN-based gene editing strategy and the natural recombination-based congenic strategy are not directly comparable at the level of genome composition, to further reconcile our findings in this study with previous data obtained from S and S.LEW congenic rats (8), it will be necessary to combine the targeted disruption of the S genome at the Adamts16 locus with a targeted insertion of the LEW rat allele of Adamts16.

Materials and Methods

Animals.

All animal research protocols were preapproved by the University of Toledo’s Institutional Animal Care and Use Committee and conducted as per the Guide for the Care and Use of Animals of the National Research Council. The Dahl Salt–sensitive (S) rats were from the colony inbred at the University of Toledo.

Generation of a ZFN-Mediated Adamts16^mutant Rat.

ZFN construct pairs specific for the rat Adamts16 gene were designed, assembled, and validated by Sigma-Aldrich to target the first exon of Adamts16 (target sequence CCGCGGTTGCTTTGCGCTCTGGGTGCTGTTGCTGGCGCA; ZFN binds to each sequence shown in bold on opposite strands). mRNA encoding the Adamts16 ZFN pairs were diluted in RNase-free tris-EDTA buffer, pH 7.4 at a concentration of 2 ng/μL and injected into fertilized Dahl S rat eggs as described previously (17). Twenty rat egg donors produced 432 eggs, 137 were fertilized and microinjected, 116 eggs survived injection, and 106 eggs were transferred to pseudopregnant Sprague-Dawley females (Charles River Laboratory SAS SD rats), which gave birth to 37 rat pups. DNA was extracted (Wizard SV 96 Genomic DNA purification system, Promega) from tail tissue and was amplified using Adamts16 ZFN Forward (5′ctgcagtgataactccgatg 3′) and Adamts16 ZFN Reverse (5′aggacacctcagtaaaacgg 3′) primers. PCR products were analyzed through 2% (wt/vol) agarose gel followed by DNA sequencing using MWG operon sequencing service. The sequencing data were analyzed using Sequencher 4.10.1. Among the 37 pups born, three positive heterozygous founder males were identified. The founder male rats were backcrossed to the same littermate of nonfounder females. Multiple separate pairs of mutation-carrying progeny were then intercrossed to generate an F2 population that was used for phenotyping and breeding to homozygosity. The DNA sequencing of the homozygous Adamts16^mutant rats showed a 17 bp deletion of the sequence “gctctgggtgctgttgc” in exon 1 of the Adamts16 gene. Renal mRNA from all of the Adamts16^mutant homozygous animals reported in this study was extracted using TRIzol Reagent (Life Technologies), and cDNA was obtained by reverse transcription with SuperScript III (Invitrogen) using an Oligo dT primer. The cDNA was PCR amplified using Adamts16 exon–specific primers (sense 5′ TAGCGGCCGCCACCATGGAACCCCGCGGTTGCTTTG 3′ and antisense 5′ CCAAGGTGAAAGTTGTCATGGTGCAG 3′). The PCR products were sequenced and all of the homozygous Adamts16^mutant rats were confirmed to harbor the 17 base pair deletion and stop codon in the first exon of the Adamts16 transcript. Protein status of Adamts16 could not be assessed, as specific antibodies to Adamts16 were not available.

BP Measurements by Tail-Cuff and Radiotelemetry.

Each set of homozygous Adamts16^mutant (n = 6 males), heterozygous Adamts16^mutant (n = 5 males), and parental strain S rats (n = 7 males) were bred, housed, and studied concomitantly to minimize environmental effects. Age-matched rats were weaned at 30 d of age and given a low-salt diet (0.3% NaCl, Harlan Teklad). Systolic BP measurements were obtained on all rats at 14.5 wk of age by the tail-cuff method as described previously (8). The week after the tail-cuff BP measurements, radiotelemetry experiments and statistical analyses were conducted as previously reported (8). Rats were euthanized after obtaining BP measurements, and total body weights and organ weights were collected.

Echocardiography.

Left ventricular RWT of Adamts16 ^mutant (n = 6 males) and control S (n = 7 males) rat hearts was evaluated by echocardiography as described previously (18). To estimate aortic PWV, ECG and Doppler signals obtained from the carotid and the iliac arteries were recorded, and the path length (Lci, distance between the points of probe applanation on the carotid and iliac arteries) was measured using a tape measure. The time intervals between the R-wave of the ECG to the foot of the Doppler carotid and iliac waveforms were averaged over three cardiac cycles, and the pulse-transit time from the carotid-iliac (PTTci) was calculated by subtracting the mean R-carotid time interval from the mean R-iliac time interval. PWV of the abdominal aorta was then estimated using the following formula: PWV = Lci/PTTci.

Vascular Imaging.

Vessels were fixed in 10% neutral buffered formalin followed by paraffin processing, embedding, and sectioning as previously described (19). Sections were de-paraffinized, stained with DRAQ5 (Biostatus Limited) and imaged.

Confocal Microscopy of the ECM.

A Leica TCS SP5 laser scanning confocal microscope (Leica Microsystems) equipped with a ti-sapphire tunable multiphoton laser (Coherent) was used to image collagen and elastin of the vessel wall. Collagen is a very bright second harmonic generator (hyperpolarizability) due to the shell of the collagen fibril (intrinsic tissue structure) and can be imaged either ex vivo or in vivo (20, 21). Second harmonic generation (SHG) for collagen was optimally imaged using 860 nm excitation multiphoton [(MP) laser] for maximum efficiency and emission collection was in the range of 425–435 nm with a peak emission generated at 430 nm (20, 21). Elastin also exhibits an inherent auto fluorescence and was excited at a wavelength of 488 nm with an emission in the range of 500–575 nm. Concurrently, nuclei within the vessels were stained with DRAQ5 and imaged at 647/681 nm. Images were acquired in 1 μm z-stacks in a sequential manner at 10× and 20× magnification. Media area was calculated using Image J software by measuring the area between the internal and external lamina borders in 5 micron cross sections of the abdominal aorta.

Cilia Measurement and Analysis.

Primary cilia were observed both with immunofluorescence and scanning electron microscopy as described previously (22).

Renal Histology.

Kidneys from groups of rats (n = 3/strain) were fixed in 10% neutral buffered formalin. Glomeruli capillary thickness and splitting were determined using Jones silver-stained sections (23). Three sections per animal (n = 3/group) were used to determine the thickness of the glomeruli capillaries using Image Pro Plus, version 6.0.0.260.

Urinary Protein Excretion.

Urinary protein excretion (UPE) determination was done as previously described (24).

Survival Study.

Adamts16^mutant (n = 10 males) and S rats (n = 10 males) were weaned at 30 d of age and given a low-salt diet (0.3% NaCl, Harlan Teklad). The rats were switched to a high-salt-containing diet (2% NaCl, Harlan Teklad) on day 124 and continued on the 2% NaCl diet until their natural death.