Heart and Circulatory Physiology

The elevated blood pressure of human GRK4γ A142V transgenic mice is not associated with increased ROS production

Zheng Wang, Ines Armando, Laureano D. Asico, Crisanto Escano, Xiaoyan Wang, Quansheng Lu, Robin A. Felder, Christine G. Schnackenberg, David R. Sibley, Gilbert M. Eisner, Pedro A. Jose


G protein-coupled receptor (GPCR) kinases (GRKs) regulate the sensitivity of GPCRs, including dopamine receptors. The GRK4 locus is linked to, and some of its polymorphisms are associated with, human essential hypertension. Transgenic mice overexpressing human (h) GRK4γ A142V on a mixed genetic background (C57BL/6J and SJL/J) have impaired renal D1-dopamine receptor (D1R) function and increased blood pressure. We now report that hGRK4γ A142V transgenic mice, in C57BL/6J background, are hypertensive and have higher blood pressures than hGRK4γ wild-type transgenic and nontransgenic mice. The hypertensive phenotype is stable because blood pressures in transgenic founders and F6 offspring are similarly increased. To determine whether the hypertension is associated with increased production of reactive oxygen species (ROS), we measured renal NADPH oxidase (Nox2 and Nox4) and heme oxygenase (HO-1 and HO-2) protein expressions and urinary excretion of 8-isoprostane and compared the effect of Tempol on blood pressure in hGRK4γ A142V transgenic mice and D5R knockout (D5−/−) mice in which hypertension is mediated by increased ROS. The expressions of Nox isoforms and HO-2 and the urinary excretion of 8-isoprostane were similar in hGRK4γ A142V transgenic mice and their controls. HO-1 expression was increased in hGRK4γ A142V relative to hGRK4γ wild-type transgenic mice. In contrast with the hypotensive effect of Tempol in D5−/− mice, it had no effect in hGRK4γ A142V transgenic mice. We conclude that the elevated blood pressure of hGRK4γ A142V transgenic mice is due mainly to the effect of hGRK4γ A142V transgene acting via D1R and increased ROS production is not a contributor.

  • hGRK4 polymorphism
  • hypertension
  • reactive oxygen species
  • G protein-coupled receptor kinases

the kidney participates in the long-term and day-to-day control of blood pressure by regulating the transport of electrolytes (e.g., sodium, chloride, and others) and secretion/degradation of vasoactive hormones/humoral agents (21, 32). Some agents promote vasoconstriction and induce sodium retention, like the renin-angiotensin-aldosterone system; others promote vasodilation and enhance sodium excretion, like atrial natriuretic peptide, dopamine, some eicosanoids, and nitric oxide (1, 5, 15, 23, 26). During states of sodium depletion, the renin-angiotensin system is activated to conserve sodium, whereas under conditions of sodium excess, renal dopamine production and the inhibitory effect of dopamine on sodium transport are enhanced (15, 23, 26). Increased activity of the renin angiotensin system and decreased activity of the renal dopaminergic system have been found in some forms of genetic hypertension (6, 15, 16, 26, 53). Thus the ability of dopamine and D1-like agonists to decrease renal tubular transport is impaired in genetic rodent hypertension and human essential hypertension (1, 15, 37). The impairment is not caused by changes in the coding region of the D1-dopamine receptor (D1R), or abnormalities in effector enzymes or proteins, but rather by an uncoupling of the D1R from its G protein/effector complex, similar to a state of desensitization (14, 15).1

G protein-coupled receptor kinases (GRKs) are important in regulating the sensitivity of G protein-coupled receptors to agonist stimulation (9, 17, 27) and have been implicated in genetic and secondary hypertension by participating in the desensitization of G protein-coupled receptors (14, 15). There are seven members of the GRK family, divided into three subfamilies: GRK1 and GRK7 belong to the rhodopsin kinase subfamily; GRK2 and GRK3 belong to the β-adrenergic receptor kinase subfamily; and GRK4, GRK5, and GRK6 belong to the GRK4 subfamily (9, 14, 15, 39, 40). The GRK4 gene locus, 4p16.3, is linked to human essential hypertension (2, 3, 10). Single-nucleotide polymorphisms of GRK4 are associated with essential hypertension (8, 14, 20, 47, 50, 61) and with increasing blood pressure with age (66). We have reported previously that transgenic mice overexpressing human (h) GRK4γ A142V have increased blood pressure and impaired natriuretic and diuretic but conserved systemic vasodilatory responses to D1-like receptor agonist stimulation (14). hGRK4γ A142V transgenic mice also have increased sensitivity to the hypertensinogenic effects of ANG II and an increased hypotensive response to ANG II type 1 (AT1) receptor blockade (57).

Reactive oxygen species (ROS) may also play an important pathophysiological role in the development and maintenance of hypertension (5, 25, 28, 33, 38, 45, 46, 49, 53, 54, 60, 62). ROS encompass a series of oxygen intermediates that include the free radical superoxide anion (O2), the nonradical hydrogen peroxide (H2O2), the highly reactive hydroxyl radical (·OH), and hypochlorous acid (42, 53, 60). Generation of ROS is increased in human essential and animal models of hypertension, and a variety of antioxidant treatments reduces blood pressure in genetic and experimentally induced models of hypertension (5, 13, 25, 28, 38, 45, 46, 49, 5254, 60, 62).

D1-like receptors are involved in the regulation of ROS production. Low concentrations of dopamine, acting at D1R and D5R, decrease ROS in renal tubular and vascular smooth muscle cells (59, 63) and brain cortical cells (36). We have previously reported that stimulation of the D5R results in an antioxidant response that is mediated by inhibition of NADPH oxidase activity, in the short term, and inhibition of NADPH oxidase expression, in the long term (62). Furthermore, hypertension in mice lacking the D5R is caused, in part, by increased systemic ROS production that is related to increased expression/activity of pro-oxidants and decreased expression/activity of antioxidants (62). In contrast, stimulation of AT1 receptors increases ROS production (53, 60). ANG II increases superoxide formation in aortic rings (28), and vascular smooth muscle (13, 33), neuronal (67), renal tubular (60) and afferent arteriolar cells (16).

The aim of the present study was to determine whether increased blood pressure in transgenic mice overexpressing hGRK4γ A142V is mediated by increased ROS production. To this end, we determined the renal expression of NADPH oxidase (Nox2 and Nox4), the major nonmitochondrial enzyme involved in the generation of ROS (60), as well as the expression of heme oxygenase (HO-1 and HO-2), an enzyme that has antioxidant activity (31, 41) . We measured the urinary excretion of 8-isoprostane as an index of oxidative stress (35). We also studied the blood pressure response of transgenic mice overexpressing hGRK4γ wild-type, hGRK4γ A142V, and nontransgenic mice to administration of the superoxide dismutase mimetic Tempol (38, 49). The responses were compared with those in D5−/− mice, in which the increase in blood pressure is mediated by increased ROS (62).


Transgenic Mice

Construction of the hGRK4γ wild-type and hGRK4γ A142V transgenes.

The full-length hGRK4γ wild-type cDNA was originally obtained from reverse transcription-PCR of mRNA of human kidney cortex and subcloned into a pTet-Off response plasmid (Clontech, Palo Alto, CA). The cDNA was removed and inserted into the pcDNA3.1 vector (Invitrogen, Carlsbad, CA) between the KpnI and BamHI restriction sites. The A142V polymorphism was produced by site-directed mutagenesis (Clontech). Both cDNAs including 3-bp 5′ untranslated region and 3-bp 3′ untranslated region were produced. Sequencing of the complete transgene and junction fragments was carried out to confirm that each transgene construct had a functional cytomegalovirus promoter, an initiation codon, and a bovine growth hormone polyadenylation signal (Georgetown University Macromolecular Analysis Shared Resources).

Generation of transgenic mice.

The transgene inserts were removed from the plasmids by digestion with Mfe I and Pvu II (New England Biolabs, Beverly, MA). The digestion reactions yielded a 2.7-kb cDNA product that included the cDNA, promoter, and poly-A tail. The concentrations of the cDNA fragments were adjusted to 0.1–0.2 μg/μl, and the cDNAs were resolved on 1% (wt/vol) low-melt preparative agarose gels and excised. Gel pieces containing the cDNAs were purified using Nucleospin columns (Clontech). The purified hGRK4γ wild-type and hGRK4γ A142V transgene constructs were quantified and microinjected into (C57BL/6J × SJL/J) F2 oocytes and surgically transferred into ovariduct of ICR mice pseudopregnant foster dams previously mated with vasectomized ICR males (14). Tail biopsies (0.5 cm) were obtained 5 wk after microinjection of the eggs (3-wk gestation and 2-wk postnatal) and analyzed for the presence of the transgenic constructs. All procedures involving the generation of transgenic mice were carried out in the University of Michigan Transgenic Core Facility according to the University of Michigan Committee on Use and Care of Animals approved protocols and Guide for the Care and Use of Laboratory Animals, published by the National Institutes of Health (NIH Publication No. 85-23, Revised 1996).

Genotyping of transgenic founders.

The screening of transgenic mice was performed with two rounds of PCR. The first set of PCR primers (pcDNA, sense 5′-CGACTCACTATAGGGAGAC-3′; hGRK4, antisense 5′-ATGGTTCCCCTCTTAGGTAG-3′) with a thermal cycler profile of 95°C/3 min; 94°C/45 s, 53°C/45 s, 72°C/45 s, 30 cycles; 72°C/10 min, with a 4°C hold, generated a 530-bp product. The second primer set (sense, 5′ CGACTCACTATAGGGAGAC 3′ and antisense, 5′ CTTGATTCTTTGATCGACCTCCTCCC 3′) with a thermal cycle profile of 95°C/3 min; 94°C/45 s, 53°C/45 s, 72°C/95 s, 30 cycles; 72°C/10 min, with a 4°C hold, generated a 1,260-bp product. Only those mice that passed two rounds of PCR screenings were used as transgenic founders (Fig. 1). Transgenic founders were identified and weaned at 3 wk of age.

Fig. 1.

Identification of transgenic founders by PCR. Top: representative image of a gel of the first-round PCR screening for transgenic founders. Only one of fifteen mice was transgene positive. L, 1-kb ladder; lane 1 to lane 15, DNA samples from 15 puppies; lane 16, positive control. Bottom: representative image of a gel of the second-round of PCR screening for transgenic founders. Eight of thirteen mice that were transgene positive in the first-round screening were also positive in the second-round screening and were chosen as breeding founders. Lane 1 to 13, DNA of the first-round PCR positive samples; lane 14, positive control.

Breeding of transgenic mice.

Three to five founders of each transgene (wild-type transgenic founder nos. 312, 649, and 657, and A142V transgenic founder nos. 39, 50, 258, 263, and 284) were bred into C57BL/6J (Jackson Laboratory, Bar Harbor, ME). Transgenic positive offspring were identified with the same PCR method used to identify the transgenic founders. The transgene positive offspring were repeatedly backcrossed to different C57BL/6J mice up to F6 for both hGRK4γ wild-type and hGRK4γ A142V transgenes to obtain congenic strains (34). Analysis of transgene expression and phenotype were conducted in these offspring. Both hGRK4γ wild-type and hGRK4γ A142V were expressed in the kidney as determined by reversed transcription-PCR and immunoblotting (14). The ROS study utilized 6- to 13-mo-old (male and female) F5/F6 mice in C57BL/6J background. Nontransgenic littermates of hGRK4γ wild-type and hGRK4γ A142V transgenic mice were used as controls.

D5−/− Mice

The generation of D5−/− mice has been reported previously (22). In our studies, we used >F6 generation mice in a C57BL/6T (C57BL/6 mice from Taconic Farms, Germantown, NY) background (3–6 mo old, male and female) (>98% congenic).

Blood Pressure Determination Under Anesthesia

The animal studies were approved by the Georgetown University Animal Care and Use Committee. The mice were housed in a temperature-controlled facility with a 12:12-h light-dark cycle and fed with mouse chow and water ad libitum. The mice were anesthetized with pentobarbital sodium (50 mg/kg ip) and placed on a heated board to maintain rectal temperature at 37°C. A catheter (PE-50 heat-stretched to 180-μm tip) was then inserted into the right femoral artery and pushed toward the aorta to monitor blood pressure. Another catheter was inserted into the right jugular vein for fluid administration. After a 30- to 60-min stabilization period, blood pressures were monitored using Cardiomax II (Columbus Instruments, Columbus, OH) (14, 22). At the conclusion of the experiment, the mice were killed with an intravenous injection of pentobarbital (100 mg/kg).


Kidneys were homogenized in isolation medium (10 mM triethanolamine and 250 mM sucrose adjusted at pH 7.6) containing protease inhibitors, PMSF (100 μg/ml), and leupeptin (10 μg/ml). Homogenates were centrifuged at 1,000 g for 10 min at 4°C. The supernatant was removed, transferred to another tube, and centrifuged at 42,000 g for 30 min at 4°C. The pellet was resuspended in isolation solution and, after protein concentration measurement, mixed with Laemmli solution and kept frozen until assay. Samples were subjected to immunoblotting as previously reported (62). The proteins were electrophoresed and transferred onto nitrocellulose membranes. The membranes were probed with monoclonal anti-HO-1 and polyclonal anti-HO-2 (Stressgen, Victoria, BC, Canada), monoclonal anti-Nox2 (62) and polyclonal anti-Nox4 or polyclonal anti-β-actin antibodies (Sigma). The Nox4 affinity-purified (SulfoLink, Pierce) antibody used in these studies was raised in rabbit against the peptide KVPSRRTRRLLDKSKT, which is 100% homologous to both rat (NCBI accession no. NP-445976) and mouse [National Center for Biotechnology Information (NCBI) accession no. NP-056575] Nox4 and 93% homologous to human Nox4 (NCBI accession no. AAH40105) at sequence 88–103 of 578 amino acids. Two bands at 55 kDa and 60 kDa were detected in mouse renal cortex homogenates, probably representing Nox4 and its isoforms. The 60-kDa band was no longer visible, whereas the 55-kDa band was greatly diminished when the antibody was preadsorbed by its immunizing peptide, confirming the specificity of the Nox4 antibody (data not shown).

The blots were incubated with peroxidase-conjugated secondary antibody, the bands visualized by ECL reagents (Amersham, Arlington Heights, IL), and the density of the bands quantified by densitometry using Quantiscan (Ferguson, MO), as previously reported (62). The amount of protein transferred onto the membranes was verified by Ponceau-S staining and immunoblotting for β-actin.

Urine and Tissue Collection

For urine collection, the mice were housed for 24 h in metabolic cages with food and water ad libitum. After this period, the mice were anesthetized and blood pressure was measured as described above. After the animals were killed, the kidneys were harvested and immediately processed for immunoblotting. Urinary protein was determined by Bio-Rad protein assay (Hercules, CA), and 8-isoprostanes were determined by enzyme immunoassay (Cayman Chemicals, Ann Arbor, MI), following the manufacturer's protocol, and values were corrected for urinary creatinine.

Treatment With Tempol

In preliminary experiments, D5−/− mice and their D5+/+ littermate controls were given a bolus intravenous injection of two different doses of Tempol, 10 mg/kg or 30 mg/kg (58 μmol/kg and 174 μmol/kg, respectively) diluted in 40 μl saline. Blood pressure was continuously recorded for 30 min after Tempol injection. In D5−/− mice, the hypotensive effect of a 30 mg/kg dose of Tempol was not greater than that of 10 mg/kg (not shown). Thus, in further experiments, only the 10 mg/kg dose was used in all the mice studied. Transgenic mice overexpressing hGRK4γ wild-type and hGRK4γ A142V and nontransgenic mice were given a bolus infusion of Tempol at a dose of 10 mg/kg, and the blood pressure was monitored for 30 min after Tempol injection.

Statistical Analysis

The data are expressed as means ± SE. Significant differences among and within groups were determined by one-way repeated-measures and factorial ANOVA, Duncan's test for n > 2, and t-test for n = 2. A P < 0.05 was considered as significant.


Blood Pressure Under Pentobarbital Anesthesia

Blood pressure was measured under pentobarbital anesthesia in 37 hGRK4γ wild-type mice, 46 hGRK4γ A142V transgenic mice, and 7 nontransgenic littermates. hGRK4γ wild-type transgenic mice had the same basal systolic and diastolic blood pressure as their nontransgenic littermates (systolic, 102 ± 1 vs. 99 ± 1; diastolic, 72 ± 2 vs. 69 ± 1 mmHg). In contrast, both systolic and diastolic blood pressures of hGRK4γ A142V transgenic mice were higher than those of their nontransgenic littermates or hGRK4γ wild-type transgenic mice (systolic, 124 ± 2; diastolic, 92 ± 1 mmHg) (Fig. 2). The increased blood pressure in hGRK4γ A142V transgenic mice was independent of age and did not affect heart weight as a function of body weight (Table 1). The blood pressures of hGRK4γ wild-type transgenic mice remained in normal range even at 86 wk of age. In contrast, the hGRK4γ A142V transgenic mice had elevated blood pressure as young as 6.5 wk and did not increase any further with age (Table 1). There was no effect of sex on blood pressure in hGRK4γ wild-type or hGRK4γ A142V transgenic mice. Both systolic and diastolic blood pressures of hGRK4γ A142V transgenic mice [males (n = 25): systolic, 126 ± 2; diastolic, 94 ± 2; females (n = 21): systolic, 122 ± 4; diastolic, 91 ± 4 mmHg] were higher than those of hGRK4γ wild-type transgenic mice [males (n = 25): systolic, 98 ± 1; diastolic, 68 ± 1; females (n = 12): systolic, 100 ± 2; diastolic, 72 ± 2 mmHg] regardless of sex (Fig. 3A). The blood pressure phenotype was inherited because the blood pressures of the transgenic founders and their offspring (up to F6) were similar (Fig. 3B).

Fig. 2.

Mice blood pressures measured under pentobarbital sodium anesthesia. Data are means ± SE. Both systolic and diastolic blood pressures were higher in hGRK4γ A142V transgenic mice (n = 46) than in nontransgenic littermates (NT; n = 7) and hGRK4γ wild-type transgenic mice (n = 37) (*P < 0.05, one-way factorial ANOVA, Duncan's test).

Fig. 3.

Mice blood pressures measured under pentobarbital anesthesia. A: systolic and diastolic blood pressures in hGRK4γ wild-type (male, n = 25; female, n = 12) and hGRK4γ A142V (male, n = 25; female, n = 21) transgenic mice grouped by sex. B: comparison of systolic and diastolic blood pressures of transgenic founders (hGRK4γ wild type, n = 22; hGRK4γ A142V, n = 12) and their offspring (hGRK4γ wild type, n = 15; hGRK4γ A142V, n = 34).

View this table:
Table 1.

Blood pressure and body weight of anesthetized transgenic mice

Renal Expression of Nox and HO isoforms

Both Nox2 and Nox4 were expressed in kidneys of nontransgenic, hGRK4γ wild-type, and hGRK4γ A142V transgenic mice. There were no significant differences in the renal expression of both Nox isoforms among the groups. The absolute difference of Nox2 between nontransgenic littermate and hGRK4γ A142V transgenic mice appeared greater than the difference between nontransgenic littermate and hGRK4γ wild-type transgenic mice (Fig. 4A). Both HO isoforms, HO-1 and HO-2, were also expressed in kidneys of the three groups of mice. There were no significant differences among the three groups in the renal expression of HO-2. However, renal HO-1 protein was increased in hGRK4γ A142V relative to hGRK4γ wild-type transgenic mice (P < 0.05, ANOVA, Duncan's test) (Fig. 4B).

Fig. 4.

Expression of NADPH oxidase and heme oxygenase (HO) isoforms in kidney membranes from hGRK4γ A142V and hGRK4γ wild-type transgenic and NT mice. Immunoblotting results are expressed as relative density units (DU) and normalized by β-actin. A: expression of Nox2 (55–60 kDa) and Nox4 (60 kDa). Inset: one blot (n = 7 in NT and hGRK4γ A142V; n = 6 in hGRK4γ wild type). B: expression of HO-1 (35kDa) and HO-2 (40kDa). Inset: one blot (n = 7 in NT and hGRK4γ A142V; n = 6 in hGRK4γ wild type. *P < 0.05, hGRK4γ A142V vs. hGRK4γ wild-type, one-way factorial ANOVA, Duncan's test).

Urinary Excretion of 8-Isoprostane

There were no significant differences in urinary excretion of 8-isoprostane among the three groups although the levels of urinary 8-isoprostane tended to be lower in hGRK4γ A142V transgenic mice relative to hGRK4γ wild-type and nontransgenic littermates (Fig. 5A).

Fig. 5.

Urinary levels of 8-isoprostane and protein. A: urinary excretion of 8-isoprostane (an index of oxidative stress) in hGRK4γ A142V and hGRK4γ wild-type transgenic and nontransgenic mice. 8-Isoprostane was determined by enzyme immunoassay and normalized by urinary excretion of creatinine (n = 7 in NT and hGRK4γ A142V; n = 6 in hGRK4γ wild type). B: urinary protein was determined by Bio-Rad protein assay and normalized by urinary excretion of creatinine (n = 7 in NT and hGRK4γ A142V; n = 6 in hGRK4γ wild type; 3/group of D5R mice).

Urinary Proteins

There were no differences in urinary protein concentrations among hGRK4γ A142V and hGRK4γ wild-type transgenic mice and nontransgenic littermates (Fig. 5B). However, the urinary protein concentration in D5−/− was higher than in D5+/+, hGRK4γ A142V and hGRK4γ wild-type transgenic mice and their nontransgenic littermates (P < 0.05, ANOVA, Duncan's test) (Fig. 5B).

Effect of Tempol

Bolus intravenous injection of a 10 mg/kg dose of Tempol had no significant effect on mean arterial blood pressure (MAP) in nontransgenic, hGRK4γ wild-type or hGRK4γ A142V transgenic mice (Fig. 6A). There was no significant percent change from control in any of the three groups during the period of study (Fig. 6B). One hGRK4γ A142V transgenic mouse had a very high blood pressure and was omitted from the statistical analysis after performing the Grubb's Outlier test. Even in this mouse, Tempol had no hypotensive effect. In contrast, Tempol significantly decreased MAP in D5−/− mice. The maximal effect (11.3% decrease in MAP) was reached 5 min after injection. The effect lasted until the end of the experiment. The hypotensive effect of Tempol was not observed in D5+/+ mice, their littermate controls (Fig. 6C). The baseline mean arterial pressures of D5−/− mice used in the Tempol study were higher than those observed in D5+/+ littermates (D5−/−, 110 ± 1.8 mmHg, vs. D5+/+, 86 ± 1.2 mmHg. P < 0.001, t-test), in agreement with our previous reports (22, 62).

Fig. 6.

Effect of Tempol (10 mg/kg; 58 μmol/kg) on mean arterial blood pressure (MAP) in anesthetized mice. A: MAP after Tempol administration in NT littermates (n = 5), hGRK4γ wild-type (n = 3), or hGRK4γ A142V transgenic mice (n = 4). Tempol was administered at time 0. B: percent change of MAP after Tempol administration in NT littermates (n = 5), hGRK4γ wild-type (n = 3), and hGRK4γ A142V transgenic mice (n = 4). C: percent change of MAP after Tempol administration (time 0) in D5R-deficient mice (D5/: n = 7) and their littermate controls (D5+/+: n = 7) (*P < 0.05 vs. D5+/+, t-test). [The baseline MAPs were higher in D5/ (110 ± 1.8 mmHg) than in D5+/+ mice (86 ± 1.2 mmHg) (P < 0.001, t-test)].


The GRK4 gene fulfills the criteria needed to qualify gene or genes as causal of complex disease, suggested by Glazier et al. (18). Thus the GRK4 gene locus 4p16.3 (3) is linked to hypertension (2, 10). Single-nucleotide polymorphisms of GRK4, by themselves, or with polymorphisms of angiotensin-converting enzyme, are associated with essential hypertension in several populations (Africans, Caucasians, Chinese, and Japanese) (8, 20, 47, 50, 61, 66). Of all the gene variations shown to be associated with essential hypertension, only the hGRK4γ A142V gene variant has been shown to produce hypertension in transgenic mice fed a normal NaCl diet (14).

There are four isoforms of human GRK4 (α, β, γ, and δ) (40). While there are also at least five GRK4 isoforms in rodents (A, B, C, D, and E) (48, 55), only the GRK4α in humans and GRK4A in rats are homologous. There is no equivalent of hGRK4γ in rodents. Expression of hGRK4γ in Chinese hamster ovary or hGRK4α in human embryonic kidney cells constitutively impairs the ability of D1 receptors to increase cAMP production (14, 44). The GRK4 gene is expressed in renal proximal tubule and thick ascending limb of Henle in rodent and human kidney (14, 48) and renal arteries in rats (48). Inhibition of GRK4 expression with antisense oligonucleotides normalizes D1R malfunction in renal proximal tubule cells from hypertensive humans (14, 15). Renal cortical GRK4 expression is increased in spontaneously hypertensive rats (SHRs) relative to their normotensive Wistar-Kyoto (WKY) controls, and chronic inhibition of GRK4 expression in the renal cortex lowers blood pressure in SHRs but not in WKY rats (48).

We have reported that hGRK4γ A142V transgenic founders had elevated blood pressure relative to hGRK4γ wild-type transgenic founders under anesthesia (14) and that hGRK4γ A142V transgenic offspring had elevated blood pressure relative to hGRK4γ wild-type transgenic offspring in conscious unanesthetized states (57). Interestingly, in mice, GRK4 is located on chromosome 5, near the D5MIT75 marker (39); the mouse GRK4 has V142 as its wild-type allele, but there is only one form of GRK4 in mice (39), which is equivalent to hGRK4α. Since the mouse GRK4 sequence is only 77% identical to human GKR4 (39), the A142V in different GRK4 proteins of humans or mice may play different roles in the protein structure and function. In this report, we show that the inheritance of hGRK4γ A142V (697 C to T) in transgenic mice was stable. The blood pressure of hGRK4γ A142V mice was higher than that found in hGRK4γ wild-type transgenic or nontransgenic littermates, whereas there was no difference between hGRK4γ wild-type transgenic and nontransgenic littermates. The elevated blood pressure was noted as early as 6.5 wk of age and as old as 2 yr and unrelated to sex. The normal blood pressure of nontransgenic littermates and the hGRK4γ wild-type transgenic mice eliminated the possibility of genetic background differences (besides hGRK4γ A142V) as a cause for the blood pressure difference. The large number of transgenic founders, a total of 22 hGRK4γ wild-type and 12 hGRK4γ A142V founders, suggested that the blood pressure difference between the hGRK4γ wild-type transgenic and hGRK4γ A142V transgenic mice was unlikely to be due to the transgene integration sites, flanking genes, or copy number (56).

As mentioned above, oxidative stress in the kidney appears to play a key role in the development and maintenance of some forms of hypertension, and increased ROS or ROS-dependent products in renal tissue, mediated by increased production or decreased scavenging or metabolism, have been demonstrated in kidneys of experimentally induced or genetically hypertensive animals (5, 13, 25, 28, 38, 45, 46, 49, 5254, 59, 60, 62, 65). We have used several different approaches to assess the contribution of increased ROS in the increased blood pressure of hGRK4γ A142V transgenic mice. We first determined the expression of Nox2 and Nox4, the two main Nox isoforms expressed in the kidney, since one of the main mechanisms underlying an increase in ROS production is a chronic increase in the expression and abundance of these components subunits (60, 62). Both isoforms were present in kidneys of hGRK4γ A142V, hGRK4γ wild-type transgenic and nontransgenic mice, and the levels of expression were similar in all three groups. The slight and nonsignificant tendency to lower levels of Nox2 expression in hGRK4γ A142V relative to nontransgenic littermates and hGRK4γ wild-type mice would suggest, if anything, decreased rather than increased ROS production. This finding is in contrast with that reported in D5−/− mice, which are hypertensive and have increased renal expression of both Nox isoforms (62).

A measure of oxidative stress in vivo is the quantification of 8-isoprostanes, the prostaglandin-like compounds that are the result of the free-radical catalyzed peroxidation of arachidonic acid. The renal excretion of 8-isoprostane is increased in several experimental models of hypertension (60), including mice, and has been used as a parameter of oxidative stress (35). In agreement with the studies on NADPH oxidase expression in the kidney, the excretion of 8-isoprostane in hGRK4γ A142V mice was not statistically different from that in the other two groups of mice, although it tended to lower levels. D5−/− mice, which are hypertensive because of increased ROS production, had increased protein excretion and cardiac hypertrophy (22). However, hGRK4γ A142V transgenic mice that were also hypertensive had normal urinary protein and heart weight. Proteinuria was not apparent, and heart weights were not increased in the hypertension induced by ANG II (58); heart weight was also not increased in hypertensive p47phox −/− mice (19). Although the hypertension in the latter mice was not associated with oxidative stress, the hypertension in glutathione-depleted mice was associated with oxidative stress but not with proteinuria or cardiac hypertrophy (11). Thus oxidative stress may not always be associated with proteinuria or cardiac hypertrophy (7, 11, 24).

To determine directly a role of increased oxidative stress in the high blood pressure of hGRK4γ A142V transgenic mice, we determined the effects on blood pressure of acute administration of the superoxide dismutase mimetic Tempol. Acute and chronic administration of Tempol has been shown to be effective in decreasing blood pressure in animals with oxidative stress-dependent hypertension (49, 60). Tempol did not lower blood pressure in hGRK4γ A142V mice; however, the same dose was effective in decreasing both systolic and diastolic blood pressure in D5−/− mice. We have already reported that D5−/− mice have increased expression of Nox2 and Nox4 and increased NADPH oxidase activity in kidneys and brain and increased plasma thiobarbituric acid-reactive substances, a marker of oxidative stress (62). We have also shown that blood pressure in these animals was normalized after treatment with apocynin, an NADPH oxidase inhibitor (62). Taken together, the biochemical and physiological data support the notion that increased ROS production is not the cause of the high blood pressure in the hGRK4γ A142V mice.

The level of ROS depends not only on their generation but also on their metabolism by antioxidant enzymes. Heme oxygenase, of which there are at least two isoforms (HO-1 and HO-2), is an important antioxidant enzyme that degrades heme, a pro-oxidant, and generates biliverdin, which has antioxidant activity against O2 (4, 12, 25, 41, 46, 51, 60). The acute and chronic administration of inducers of HO-1 normalizes blood pressure in SHRs (25, 41, 46). Renal HO-1 (but not HO-2) protein was increased in hGRK4γ A142V, relative to hGRK4γ wild-type mice, supporting the notion that ROS production is not increased and may be decreased in hGRK4γ A142V transgenic mice. In the kidney, HO-1 is found in proximal and distal tubules (12, 41), and HO-2 is present in the medullary thick ascending limb and preglomerular arterioles (12, 41). As a heat shock protein, HO-1 has multiple protective functions, including protecting thick ascending limb cells from ANG II-evoked oxidative injury (41). Because hGRK4γ A142V transgenic mice have decreased D1-like receptor function (14) but increased AT1 receptor function (57), an increase in ROS production may have been expected. We suggest that this did not occur because of the increase in HO-1 expression. Indeed, Wesseling et al. (58) have suggested that the resistance to oxidative stress following ANG II infusion in mice is due to increased HO-1 expression. The ability of medullary HO-1 to induce a natriuresis (29) could have alleviated the impaired natriuretic effect of D1R in hGRK4γ A142V transgenic mice and prevented an even higher blood pressure (29). HO-1 expression and activity were decreased in D5−/− mice, and D5R increased HO-1 activity and expression (30). D5R is probably not regulated by GRK4 because D5R action is not impaired in SHRs, which have increased GRK4 expression and activity (48, 64). It is possible that other G protein-coupled receptors, including D5R (30) and D2R (4), may have prevented any increase in ROS production in hGRK4γ A142V transgenic mice.

There are several instances in which hypertension is not caused by increased ROS production. Norepinephrine-induced hypertension in rats was not associated with O2 production (43). The hypertension in mice caused by a chronic infusion of ANG II or inhibition of nitric oxide synthesis with nitro-l-arginine was not associated with increased urinary thiobarbituric acid reactive substances, an index of oxidative stress (11, 58). Grote et al. (19) reported that p47phox−/− mice had increased blood pressure, plasma renin, and renal angiotensin-converting enzyme activities but not oxygen radical formation (19). The hypertension was ameliorated by AT1 receptor blockade but not by Tempol (19), findings similar to the hGRK4γ A142V transgenic mice (current report and Ref. 57). There is, however, a difference in the blood pressure phenotype in p47phox−/− mice and hGRK4γ A142V transgenic mice; blood pressure elevation was no longer apparent after 60 wk of age in p47phox−/− mice, whereas in hGRK4γ A142V transgenic mice, the blood pressure continued to be elevated at 86 to 104 wk of age.

In summary, we report that overexpression of hGRK4γ A142V in mice increases blood pressure. To date hGRK4γ is the only gene the locus of which is linked to and the variants of which are associated with human essential hypertension and the expression of which in mice produces hypertension. We also report that hypertension in hGRK4γ A142V mice is not dependent on increased ROS production, possibly because of compensatory mechanisms that increased antioxidant activity, e.g., HO-1. The cause of the increase in HO-1 activity is not readily apparent but could be due to increased expression or activity of other G protein-coupled receptors. Although D1R function is decreased, D5R and D2R function may not have been affected in these animals and may have been able to effectively downregulate ROS production.


This work was supported by National Institutes of Health Grants PG-00127–2004.R1, HL-23081, DK-39308, HL-68686, DK-52612, and HL-074940.


The authors express thanks to Dr. Jean E. Robillard for help in generating the transgenic mice.


  • 1 This paper was presented at the 9th Cardiovascular-Kidney Interactions in Health and Disease Meeting at Amelia Island Plantation, Florida, on May 26–29, 2006.

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


View Abstract