Production of Double Homozygous Agtr1a −/−Agtr1b −/− Mice.

Mice with targeted mutations of the AT_1A receptor gene (Agtr1a) have been produced in our laboratories as described previously (15). Several breeding strategies were used to produce double homozygous Agtr1a −/−Agtr1b −/− mice. First, Agtr1a−/− and Agtr1b −/− mice were crossed to produce double heterozygous Agtr1a +/−Agtr1b +/− animals that were then intercrossed. To increase the likelihood of obtaining double homozygotes, we also intercrossed Agtr1a −/−Agtr1b +/− or Agtr1a +/−Agtr1b −/− mice. Finally, once double homozygous mice were obtained, they were crossed with Agtr1a −/−Agtr1b +/− or Agtr1a +/−Agtr1b −/− mice.

RNA Isolation and Reverse Transcription—PCR (RT-PCR) Analysis of AT_1B Expression.

Adrenal glands were harvested from Agtr1b +/+ and Agtr1b −/− mice and RNA was extracted from the tissues by using a commercially available reagent (Tri-reagent; Sigma). Total RNA was reverse transcribed to cDNA by using poly-dT (22). A fragment of the AT_1B receptor cDNA was than amplified by PCR by using a sense primer derived from a portion from the Agtr1b gene that was deleted in the targeted gene (5′-ATACCGCTATGGAATACCAG-3′) and an antisense primer from the 3′-region of the Agtr1b gene (5′-TTGATAACCCTGCATGCGACC-3′). The diagnostic 700-bp fragment was identified on ethidium-stained gels. The presence and identity of the fragment were then confirmed by Southern blotting with a probe from the Agtr1b gene that was internal to the primer sequences.

Analysis of Growth and Development and Kidney Histology.

To determine the effects of the combined Agtr1a/Agtr1b gene disruptions on growth and development, we measured body weights along with wet weights of kidneys and heart in 4-month-old male wild-type and Agtr1a −/−Agtr1b −/− mice. The major organs from these animals were fixed in formalin, sectioned, and stained with hematoxylin and eosin. The histomorphology of these organs was then examined by light microscopy.

Assessment of Urinary Concentrating Capacity.

To examine the relative importance of the individual AT₁ receptors in urinary concentration, we measured urine osmolality in groups of wild-type (n = 9), Agtr1a +/+Agtr1b −/− (n = 4), Agtr1a −/−Agtr1b +/+ (n = 14), and Agtr1a −/−Agtr1b −/− (n = 5) mice. Urine osmolalities were measured while mice had free access to drinking water and following 12 hours of complete water deprivation. Urine was collected after bladder massage and urine osmolality was measured immediately with a vapor pressure osmometer (Wescor, Logan, UT).

Vasoconstrictor Effects of Angiotensin II.

In preparation for studies to assess the vascular effects of angiotensin II, the angiotensin converting enzyme (ACE) inhibitor enalapril was administered to groups of Agtr1a +/+Agtr1b +/+ (n = 4), Agtr1a +/+Agtr1b −/− (n = 4) and Agtr1a −/−Agtr1b −/− (n = 4) mice to inhibit endogenous production of angiotensin II (23). Thirty mg/kg per day of enalapril was administered in the drinking water for 5 days. On the day of study, animals were anesthetized with isoflurane and a flexible plastic catheter (0.020ID; Braintree Scientific) was placed in the carotid artery to monitor arterial pressure. A second catheter was placed in the jugular vein to infuse angiotensin II. Intraarterial blood pressure was recorded continuously through the carotid catheter with a pressure transducer and windaq acquisition and playback software (Dataq Instruments, Akron, OH).

Beginning with an equilibration period of 5 min, blood pressure was recorded continuously for the duration of the experiment at 50 recordings per sec. Data collected from each animal were compressed to five recordings per sec, and the mean arterial blood pressure (MAP) recordings from animals in each experimental group were integrated and averaged. Immediately after the equilibration period, mice received a bolus injection of 0.9% sodium chloride solution (NS) in a volume equal to 0.1% of their body weight. This is the volume used for each of the injections throughout the remainder of the experiment. At 10-min intervals thereafter, increasing doses of 0.1, 1.0, and 10 μg/kg of angiotensin II (Peninsula Laboratories) were injected intravenously. Ten minutes after the 10 μg/kg dose of angiotensin II, 10 μg/kg of epinephrine was administered intravenously and MAP was monitored for an additional 5 min.

Systolic Blood Pressure Measurements in Conscious Mice.

Systolic blood pressures were measured in groups of Agtr1a+/+Agtr1b+/+ (n = 8), Agtr1a +/+Agtr1b −/− (n = 10), and Agtr1a −/−Agtr1b −/− (n = 13) mice by using a computerized tail-cuff system (Visitech Systems, Cary, NC) as previously described (24). Before the study was initiated, mice were adapted to the apparatus for at least 5 days. Systolic blood pressures were then measured for 10 days.

To examine the effect of reducing angiotensin II levels on blood pressure, these same groups of mice were treated with enalapril in drinking water at a dose of 30 mg/kg per day for 10 days, during which their systolic blood pressures were measured.

AT_1B Receptor mRNA Expression in Mutant Mice.

To verify that the targeted mutation introduced into the Agtr1b locus resulted in inactivation of the gene, we examined expression of AT_1B receptor mRNA in adrenal glands by RT-PCR. As shown in Fig. ​Fig.2,2, AT_1B receptor mRNA could not be detected in Agtr1b −/− mice, confirming the efficacy of the gene targeting strategy.

Open in a separate window

Figure 2

Assessment of AT_1B mRNA expression by RT-PCR. Total RNA was isolated from adrenal glands of two wild-type and two Agtr1b −/− mice and each specimen was subjected to RT-PCR to amplify AT_1B receptor mRNA. In tissues from the wild-type mice (lanes 1 and 2), a PCR product of the expected size (700 bp) was detected on an ethidium bromide stained gel (A) and was confirmed to be AT_1B receptor mRNA by hybridization (B). This band was not detected in RNA from the Agtr1b −/− mice (lanes 3 and 4).

Production of Mice Lacking Both AT_1A and AT_1B Receptors.

Because the Agtr1a and Agtr1b genes are located on different chromosomes (4), the mutations could be combined through simple breeding. Agtr1a −/−mice were crossed with Agtr1b −/− mice to produce progeny that were heterozygous for both the Agtr1a and Agtr1b disruptions. These Agtr1a +/−Agtr1b +/− animals were then intercrossed to produce mice that were double homozygotes for the null Agtr1a and Agtr1b alleles. To increase the frequency of generating Agtr1a −/−Agtr1b −/− animals, we also crossed mice that were heterozygous for one disruption and homozygous for the other (e.g., Agtr1a +/−Agtr1b −/− × Agtr1a −/− Agtr1b +/−). Among 343 consecutive 21-day-old weanlings from these different crosses, there were significantly fewer Agtr1a −/−Agtr1b −/− pups than predicted by simple Mendelian inheritance (29 observed vs. 50 expected; P = 0.024 by χ² analysis). Thus, in these mixed breedings, the complete absence of AT₁ receptors is detrimental to survival, with an apparent perinatal mortality rate of approximately 40%. By comparison, the perinatal mortality rate of angiotensinogen-deficient (Agt −/−) weanlings in our colony is approximately 60% (13 observed vs. 37 expected; P = 0.0016 by χ² analysis).

Once double homozygous Agtr1a −/−Agtr1b −/− mice were obtained, some of these animals were crossed with Agtr1a −/−Agtr1b +/− or Agtr1a +/−Agtr1b −/− mice to further increase the frequency of producing Agtr1a −/Ågtr1b −/− animals. When the genotypes of pups from these litters were analyzed, the observed frequency of Agtr1a −/−Agtr1b −/− weanlings was virtually identical to predicted (22 observed vs. 23 expected; P = 0.9 by χ². Thus, the survival disadvantage associated with the absence of AT₁ receptors can be overcome by selective breeding. This observation suggests that this trait can be modified significantly by background genes that are present in the C57BL/6 × 129 mixed breedings. In mice lacking only AT_1A receptors, we have identified similar modifying effects of background genes (25).

Both male and female Agtr1a −/−Agtr1b −/− mice are fertile. The frequency of productive matings, gestation, delivery, and care of litters was not different between Agtr1a −/−Agtr1b −/− and wild-type females. Male Agtr1a −/−Agtr1b −/− mice produced litters of normal size. Our group (12) and others (10) have previously identified fertility problems in male ACE-deficient mice, but not in male angiotensinogen-deficient mice (11). In view of our current results, the defective fertility observed in ACE-deficient mice is not because of a lack of AT₁-receptor signaling and may be caused by the loss of unique functions of testicular ACE probably unrelated to angiotensin II production. For example, other potential ACE substrates (bradykinin, enkephalins, and substance P) can alter functions of sperm (26, 27).

Growth Is Impaired in Mice with Combined AT_1A/AT_1B Receptor Deficiency.

In neonatal rats, AT₁ receptor antagonists impair somatic and renal growth (28). However, in a previous study, we found that mice lacking only the AT_1A receptor grow and develop normally (29). To determine whether the combined absence of AT_1A and AT_1B receptors adversely affects growth, we compared body, heart, and kidney weights between 4-month-old male wild-type and double homozygous Agtr1a −/−Agtr1b −/− mice. The surviving Agtr1a −/−Agtr1b −/− mice appeared healthy and could not be distinguished from wild-type littermates based on their external appearance. However, their body weights were significantly lower than wild-type mice (25.7 ± 1.4 vs. 29.6 ± 0.6g; P = 0.05). In addition, both kidney weights (0.25 ± 0.02 vs. 0.32 ± 0.02 g; P = 0.03) and heart weights (0.09 ± 0.01 vs. 0.13 ± 0.01 g; P = 0.02) were also significantly reduced in the AT₁-deficient mice compared with wild-type controls. Because the ratios of heart weight to body weight (0.37 ± 0.02% vs. 0.45 ± 0.04%; P = 0.1) and kidney weight to body weight (0.96 ± 0.06% vs. 1.08 ± 0.06%; P = 0.2) are similar in Agtr1a −/−Agtr1b −/− and wild-type control mice, the combined absence of AT_1A and AT_1B receptors appears to cause a generalized impairment of growth.

Abnormalities in Kidney Structure in Double Homozygous Agtr1a −/−Agtr1b −/− Mice.

Agt −/− or Ace −/− mice that are unable to synthesize angiotensin II develop structural abnormalities in the kidney. These include thickening of arterial walls, focal areas of interstitial fibrosis and tubular atrophy, and atrophy of the renal papilla (10, 11, 13). The absence of any single known angiotensin II receptor, AT_1A, AT_1B, or AT₂, does not completely recapitulate this phenotype (15–20). To determine whether the pathogenesis of these renal phenotypes requires the simultaneous absence of signaling through both AT_1A and AT_1B receptors, we examined renal histomorphology in Agtr1a −/−Agtr1b −/− mice and compared them to wild-type controls. As depicted in Fig. ​Fig.33B, Agtr1a −/−Agtr1b −/− mice develop atrophy of the inner medulla that is easily detected by light microscopy. These mice also develop severe thickening of arterial walls within the kidney (Fig. ​(Fig.33D) that resembles the vascular abnormalities seen in Agt −/− and Ace −/− mice (10–13). We also found focal areas of cortical interstitial inflammation, fibrosis, and tubular atrophy in the Agtr1a −/−Agtr1b −/− mice (Fig. ​(Fig.33F), similar to lesions that we observed in Agt −/−mice (11). These findings are in general agreement with those of Tsuchida et al., who recently reported that combined disruption of both AT₁ receptor genes produces a severe kidney phenotype (30). There were no abnormalities noted in the histology of hearts and adrenal glands of Agtr1a −/−Agtr1b −/− mice. In particular, there was no evidence of arterial or arteriolar thickening at these sites.

Open in a separate window

Figure 3

Kidney histomorphology of wild-type and Agtr1a −/−Agtr1b −/− mice. (A and B): Axial sections from wild-type (A) and Agtr1a−/−Agtr1b −/− mice (B). (×5.) There is marked atrophy of the papilla in the AT₁-deficient kidney compared with the control. (C and D): Cross sections of representative small arteries in kidneys from wild-type (C) and Agtr1a −/−Agtr1b −/− (D) mice. The media of the vessels from the mutant animals was markedly thickened and many of the vessels were surrounded by inflammatory cells as shown. (×150.) (E and F): Sections of renal cortex from wild-type (E) and Agtr1a−/−Agtr1b −/− (F) mice. In the AT₁-deficient mice there were focal areas with tubular dropout and inflammatory cell infiltration. (×100.)

These data suggest that both AT₁ receptor isoforms contribute to the regulation of vascular structures in the kidney and to the maintenance of the renal papilla. Furthermore, as with the growth responses, the absence of a single AT₁ receptor isoform can be compensated by the other. However, the AT_1A receptors have a greater capacity for compensation than the AT_1B receptors, perhaps reflecting their relative levels of expression in kidney (15, 31). Accordingly, renal morphology is completely normal in the Agtr1b −/− mice that possess the full complement of AT_1A receptors. In contrast, subtle abnormalities can be detected in kidneys of AT_1A-deficient mice. By using magnetic resonance microscopy, modest reductions in the size of the renal papilla are revealed in Agtr1a −/− mice that were not detected by routine pathological evaluation [Oliverio et al., unpublished work]. In addition, minimal but significant thickening of renal interlobular arteries has been detected in Agtr1a −/− mice by using a careful morphometric analysis (18). Nonetheless, the difference between the modest abnormalities seen in Agtr1a −/− mice and the florid pathology seen in double homozygous Agtr1a −/−Agtr1b −/− mice must be due to compensation by the AT_1B receptor. This level of compensation seems remarkable because our previous studies have shown that AT_1B receptor binding is virtually undetectable in kidneys of Agtr1a −/− mice (15). Alternatively, it has been hypothesized that atrophy of the papilla may not depend directly on expression of AT₁ receptors in the kidney, but instead is related to loss of AT₁ receptor functions in the urinary collecting system (32).

AT₁ Receptors and Urinary Concentration.

Among its many biological actions, the RAS regulates water metabolism, in part through diverse effects on urinary concentrating mechanisms (33, 34). Whereas a range of abnormalities in water homeostasis has been described in mice lacking angiotensin II or AT_1A receptors, impairment of urinary concentrating capacity is a feature common to all of these mutants (10, 35–37). To clarify the role of AT_1A and AT_1B receptors in these processes, we measured urinary concentrating capacities of wild-type, Agtr1a −/−, Agtr1b −/−, and Agtr1a −/−Agtr1b −/− mice. As shown in Fig. ​Fig.4,4, after 12 hours of water deprivation, urine osmolalities in Agtr1b −/− mice increased to the same extent as in wild-type controls. In contrast, Agtr1a −/− mice have a defect in urinary concentration that is clearly apparent after 12 hours of water deprivation. Furthermore, the combined absence of both AT_1A and AT_1B receptors causes a urinary concentrating defect that is even more severe. Their urine osmolality remains virtually unchanged after 12 hours of water deprivation. Whereas this severe defect in the Agtr1a −/−Agtr1b −/− mice is not unexpected given their inner medullary atrophy (see Fig. ​Fig.33B), the incremental differences in maximal urine osmolality between the three mutant groups further illustrate the relative abilities of the AT_1A and AT_1B receptor isoforms to compensate for one another.

Open in a separate window

Figure 4

Urine osmolality following 12 hours of water deprivation. The experimental groups are represented as follows: open bar is Agtr1a−/−Agtr1b −/− group, gray bar is theAgtr1b⁻ −/− group, hatched bar is the Agtr1a⁻ −/−group, and the black bar is the wild-type group. (P < 0.0003 vs. wild-type or Agtr1b −/−; ^†P < 0.0001 vs. wild-type or Agtr1b −/−, P = 0.004 vs. Agtr1a −/−.

Vasoconstrictor Actions of Angiotensin II.

The vasoconstrictor actions of angiotensin II are mediated by AT₁ receptors (2). We have previously shown that the AT_1A receptor mediates a substantial portion of the pressor effects of angiotensin II based on the marked attenuation of these responses in Agtr1a −/− mice (15). However, when endogenous angiotensin II production is suppressed, a modest pressor and renal vasoconstrictor effect of angiotensin II can be detected in AT_1A-deficient mice. This response is blocked by AT₁ receptor antagonists, but not by sympatholytic agents, suggesting that it is a direct effect of AT_1B receptors (23, 38).

To directly test this hypothesis and to clarify the role of other angiotensin receptors in the vasoconstrictor actions of angiotensin II, we examined pressor responses to angiotensin II in wild-type, Agtr1b −/− and Agtr1a −/−Agtr1b −/− mice. As shown in Fig. ​Fig.5,5, angiotensin II caused similar dose-proportional increases in mean arterial pressure in the wild-type and Agtr1b −/− mice. In contrast, angiotensin II had no discernable effect on MAP in the double homozygous Agtr1a −/−Agtr1b −/− mice. To determine whether responses to other pressor agents might also be impaired in these animals, each group received an infusion of epinephrine after the angiotensin II infusions were completed. As shown in Fig. ​Fig.5,5, epinephrine caused vasoconstriction with significant and equivalent increases in MAP in all of the groups, including the Agtr1a −/−Agtr1b −/− mice. These studies verify the importance of AT₁ receptors in the vasoconstrictor effects of angiotensin II. In addition, they confirm the conclusions from our previous study that the vasoconstrictor actions of angiotensin II in Agtr1a −/− mice are mediated by AT_1B receptors (23). Finally, the absence of any hemodynamic response to angiotensin II in Agtr1a −/−Agtr1b −/− mice suggests that there are no other angiotensin II receptors with significant vasoactive functions.

Open in a separate window

Figure 5

Hemodynamic effects of angiotensin II in enalapril-pretreated, anesthetized wild-type, Agtr1b −/−, and Agtr1a −/− Agtr1b −/− mice. The changes in MAP at 120 seconds after injection of vehicle, 0.1, 1.0, and 10 μg/kg of angiotensin II, or 10 μg/kg epinephrine are shown. The experimental groups are represented as follows: open bars are the Agtr1a −/−Agtr1b −/− group, gray bars are the Agtr1b −/− group, and black bars are the wild-type group (P < 0.0003 vs. wild-type or Agtr1b−/− groups).

The authors thank Norma B. Turner for secretarial and administrative help and Kamie Snow for expert technical assistance. We also thank Drs. William Arendshorst and Paul Klotman for their critical review of the manuscript. These studies were supported by the National Institutes of Health (HL58554, HL49277, DK38108, and GM20069) and the Research Service of the Department of Veterans Affairs.