Differential Regulation of Angiotensinogen Transcripts After Renin Infusion
To investigate angiotensinogen regulation in high-renin hypertension, we infused porcine renin intravenously at either a low (4 mU/kg per hour, n=6) or high (20 mU/kg per hour, n=9) dose into male Sprague-Dawley rats (225 to 250 g) for 5 days using osmotic minipumps. Control rats received 0.9% NaCl. In renin-infused rats, mean arterial pressure and plasma renin activity were significantly elevated. Both low- and high-renin infusions lowered plasma angiotensinogen levels. Plasma angiotensin II was elevated in rats given renin but reached statistical significance only at the higher dose. Angiotensinogen mRNA isolated from the liver, adrenal gland, kidney, and brain was measured by slot blot analysis. Both renin doses were associated with significant decreases in the levels of liver and hypothalamic angiotensinogen mRNA. In the medulla oblongata, angiotensinogen mRNA was reduced only by the higher renin dose. The lower dose increased angiotensinogen mRNA in the adrenal gland, and in kidney, angiotensinogen mRNA level was unchanged by renin infusion. Angiotensinogen mRNA visualized on Northern blots showed that the number of mRNA species in liver decreased from three in control rats to a single mRNA species after renin infusion. Tissue differences in the size of the major angiotensinogen mRNA species were also apparent. This, together with changes in the total hybridization signal of angiotensinogen mRNA in tissues, suggests that renin differentially affects the different angiotensinogen mRNA transcripts. Results of this study indicate that angiotensinogen gene expression is regulated not only by alterations in levels of circulating angiotensin II but also by other mechanisms, presently unidentified, that are activated by renin infusions.
The RAS participates in the maintenance of the homeostatic processes that regulate blood pressure, fluid and sodium balance, and hormonal levels.1 2 3 Although many studies have demonstrated that the RAS is activated in hypertensive animals,4 5 the relative importance of RAS components differs in the various animal models of hypertension. We showed that in the early phase of renin-dependent renal hypertension, an activated circulating RAS participates in the maintenance of high blood pressure, but in the later phase of renal hypertension, additional factors become more important.6 Results of this and other studies suggest that additional renin-dependent humoral factors might exert tissue-specific effects on RAS genes.
As the only known substrate for renin, Aogen represents a primary target for RAS regulation. The potential significance of this molecule in the hypertensive rat is underscored by a recent study with antisense oligonucleotides used to block translation of Aogen mRNA in the brains of spontaneously hypertensive rats, resulting in lowered blood pressure.7 Other studies support the concept that regulation of Aogen gene expression centrally affects the maintenance of hypertension.8 To begin to examine the mechanisms regulating Aogen expression both centrally and peripherally in high-renin hypertensive animals, we infused renin intravenously into rats for 5 days and analyzed Aogen mRNA in the brain and peripheral tissues.
Twenty-one male Sprague-Dawley rats (225 to 250 g) were obtained from Harlan Farms (Madison, Wis) and housed individually in a constant-temperature room (22°C) lighted for 12 hours. During the experiments, rats had free access to tap water and rat chow (Purina, Fetzer Brothers). Rats were divided into three groups: one group (n=6) received a 5-day intravenous infusion of vehicle (0.9% NaCl), and the two other groups received a continuous infusion of porcine renin (Sigma Chemical Co) at either a low (4 mU/kg per hour, n=6) or high (20 mU/kg per hour, n=9) dose. One unit of the renin preparation generates 100 μg Ang I (Sigma). Solutions were given via osmotic minipumps (model 2001, Alza Corp) calibrated to deliver the agents at a rate of 1 μL/h. Four days later, an arterial catheter (PE-50) was implanted into the abdominal aorta through a femoral artery with rats under light halothane anesthesia. Baseline measurements of MAP and heart rate were obtained 24 hours later. Arterial pressure was recorded with a solid strain transducer (MP-150, Micron Instruments Inc) on a multichannel polygraph (model 2000, Gould Instruments) in conscious, freely moving rats. Rats were killed by decapitation, and blood and tissues were collected. Tissue samples were snap-frozen on dry ice and stored at −80°C for RNA extraction. Blood samples were collected for determination of plasma Aogen, PRA, and plasma Ang II immunoreactivity (ir-Ang II). PRA, Aogen, and ir-Ang II were measured as described elsewhere.9 10 The mean residual volumes of vehicle or renin solutions in the osmotic minipump were 76±2 μL.
With rats under light halothane anesthesia, polyethylene tubing (PE-50, Clay Adams) was placed into a femoral vein for infusion of either renin or vehicle. The free end of the tubing was connected to an osmotic minipump implanted within the subcutaneous tissue at the level of the interscapular region. All surgical procedures were done under sterile conditions.
Each tissue was weighed and immediately homogenized in guanidine isothiocyanate RNA isolation buffer.6 Total cellular RNA from each experimental sample was aliquoted in six serial dilutions (4 to 0.125 μg) and applied to a nitrocellulose membrane in a slot blot apparatus. Blots were hybridized overnight at 42°C with 1×106 cpm/mL α-32P–labeled cDNA probe and were washed as described.6 In hybridizations with γ-32P–labeled 18S rRNA oligonucleotide probe, the membranes were processed as described previously.6 Blots were exposed to XAR-5 x-ray film (Eastman Kodak Co) with two intensifying screens (Cronex, EI du Pont de Nemours & Co). Radioactive standards (ARC-146 carbon-14 standard, American Radiolabeled Chemicals Inc) were included for determination of relative optical densities, and autoradiographs were analyzed with Image software (version 1.28) (National Institutes of Health). Because the signals of the sixth dilution (0.125 μg) were sometimes too low for detection by this image-analysis system, we used five dilutions of total cellular RNA (4 to 0.25 μg) in the analysis. The quantity of cDNA probe hybridized was plotted as a function of 18S rRNA oligonucleotide hybridized, and the slope of the linear regression line was taken as the relative level of Aogen mRNA. Means of the slopes were calculated for each group, and statistical differences between groups were tested. For Northern blot analysis, total cellular RNA (20 μg) was separated by electrophoresis through 0.8% agarose gels and analyzed as described.8 Membranes were hybridized overnight at 42°C with 1×106 cpm/mL α-32P–labeled cDNA probe.
Preparation of 32P-Labeled Probes
Double-stranded DNA hybridization probes were labeled with [α-32P]dATP using a nick translation kit (Boehringer Mannheim GmbH). Rat Aogen cDNA was originally a gift from Dr K.R. Lynch, Charlottesville, Va. Specific activities of the probes were 2.2×108 to 3.0×108 disintegrations per minute/μg. Oligodeoxynucleotide complementary to rat 18S rRNA was labeled with [γ-32P]dATP using polynucleotide kinase (Bio-Rad Laboratories). Since rRNA is the predominant constituent in total cellular RNA and 18S rRNA represents a fixed proportion of rRNA, hybridization to the oligonucleotide was used as an estimate of loaded total cellular RNA.6 11
Data are expressed as mean±SE. Differences between groups were evaluated by one-way ANOVA followed by Duncan's multiple range test. Univariate correlation analysis was used for examination of relationships between mRNA levels and other parameters. The criterion for statistical significance was a value of P<.05.
Hemodynamic and Humoral Effects of Renin Infusions
Two doses of renin were chosen to achieve PRA levels comparable to those obtained in earlier studies in aortic-ligated6 or Ang II–infused8 rat models of hypertension. Both renin doses elevated MAP to a similar extent (Table⇓). Heart rate was unchanged among the groups. Body weights of rats did not change before and after infusions (data not shown). PRA increased in a dose-dependent fashion and was accompanied by an increase in plasma ir-Ang II; however, the Ang II increase was statistically significant only in the group exposed to the higher dose of infused renin. Plasma Aogen levels were decreased in both renin infusion groups.
Elevated arterial pressure was significantly correlated with the increase in PRA (r=.459, P<.05). In the same rats, plasma levels of Aogen were inversely correlated with MAP (r=−.540, P<.05). PRA correlated positively with plasma ir-Ang II (r=.730, P<.001) and negatively with plasma Aogen levels (r=−.459, P<.05).
Effects of Renin Infusions on Tissue Aogen mRNA Levels
Fig 1⇓ shows changes in Aogen mRNA levels in liver, hypothalamus, medulla oblongata, kidney, and adrenal glands. Aogen mRNA levels in the liver were significantly decreased by both low and high doses of renin compared with the vehicle group. In the hypothalamus, Aogen mRNA levels were also reduced by renin infusions and did not differ between low-dose and high-dose renin, as in the liver. In the medulla oblongata, Aogen mRNA levels decreased only in the high-dose renin group. In the kidney, Aogen mRNA levels were unchanged by renin infusions, and in adrenal glands, Aogen mRNA levels increased only with the lower dose of renin.
MAP correlated negatively with Aogen mRNA levels in hypothalamus (r=−.579, P<.01) and positively with Aogen mRNA levels in adrenal glands of rats infused with the lower dose of renin (r=.607, P<.01). PRA was negatively correlated with Aogen mRNA levels in hypothalamus (r=−.572, P<.01) and medulla oblongata (r=−.626, P<.01). Plasma Aogen levels were correlated positively with Aogen mRNA levels in liver (r=.426, P<.05) and hypothalamus (r=.569, P<.01) but negatively with adrenal Aogen mRNA levels of rats infused with the lower dose of renin (r=−.539, P<.05).
Tissue-Specific Differential Effects of Renin on Aogen mRNA Lengths
After renin infusions, not only did the quantity of hepatic Aogen mRNA decrease, but also the number of Aogen mRNA species was reduced (Fig 2⇓, left). The longest of the three visible Aogen mRNAs was present only in RNA prepared from livers of control rats. In the rat Aogen gene, two alternate start sites have been identified at the 5′ end12 as well as four potential polyadenylation sites at the 3′ end.13 Alternative selection of any of these sites could account for differently sized Aogen mRNA transcripts. For example, the longest of the liver Aogen transcripts could represent the use of the polyadenylation signal 1839 nucleotides from the start of transcription,13 since the addition of a polyadenylate tail would bring the length of the mRNA to 1.9 to 2.0 kb. Both doses of renin eliminated this band. The smallest Aogen mRNA was the only species that remained after the higher-dose renin infusion. This band could reflect use of the polyadenylation signal 1784 nucleotides from the start of transcription, which Ohkubo et al13 determined was the most abundant form of Aogen mRNA present. In the medulla oblongata, the largest Aogen mRNA species was not present, and it was unclear whether the intermediate-sized Aogen mRNA was present or was differentially affected by the renin infusions (Fig 2⇓, middle). In RNA preparations from the adrenal gland, there was no apparent selective increase or decrease in Aogen mRNAs of different lengths, although the total signal increased after infusion of the lower renin dose. In the adrenal gland preparation, the intensity of largest and smallest Aogen mRNA bands was approximately equal (Fig 2⇓, right). Interestingly, in kidney, where Aogen mRNA levels did not change after renin infusion, only the largest of the three Aogen mRNA species was visible (Fig 2⇓, right). Multiple Aogen mRNA species and a change in number of Aogen mRNA species were best seen in the liver, although the number of Aogen mRNA species in the medulla oblongata may also have decreased to one.
High levels of circulating renin, produced in this study by infusion of exogenous renin, resulted in downregulation of steady-state levels of Aogen mRNA in liver, hypothalamus, and medulla oblongata but not in kidney and adrenal gland. Furthermore, the number of Aogen mRNA species also decreased. The number of Aogen transcripts varied among tissues, and the Aogen transcripts were not equally affected in each tissue, nor did all transcripts appear in all tissues. One explanation of these effects is that the renin infusions produced conditions that destabilized Aogen mRNA by an indirect mechanism, as yet unrecognized. In hepatocytes, Ang II is known to increase Aogen mRNA expression by stabilizing the mRNA.14 Renin might regulate Aogen mRNA by an analogous mechanism. Our renin infusion studies, although accompanied by increased levels of circulating Ang II at the higher dose of renin, must involve additional mechanisms that interfere with the classic patterns of Ang II stimulation of Aogen. Another possible explanation is that in adult rats, the Aogen gene is chronically stimulated by hormones and other factors and that in our study, the renin infusions resulted in removal of these stimulatory regulating factors. Absence of these factors would be reflected as decreases in Aogen mRNA levels. Then the renin infusions would have indirect effects on Aogen transcription by disrupting the normal mechanisms by which RAS components and other factors regulate Aogen gene expression.
In support of this alternative, plasma Ang II levels were not always informative indicators of Aogen gene expression in our study. It is possible that the proportion of other angiotensin peptides in plasma did increase; we did not measure this in the present study. The high levels of circulating renin might stimulate alternate processing pathways for angiotensin peptides. Also, angiotensin peptides such as Ang-(2-8), Ang IV, or Ang-(1-7)10 might themselves have inhibitory effects on tissue levels of Aogen mRNA. With high renin and low plasma Aogen levels, one would anticipate that circulating levels of Ang II would be substantially increased. Plasma levels of Ang II did not increase as much as expected, although plasma Aogen levels were decreased significantly. In a recent study in stroke-prone spontaneously hypertensive and Dahl salt-sensitive rats on prolonged high salt diets, high plasma renin levels also were associated with low plasma Aogen; the low Aogen levels appeared to result from Aogen depletion by high levels of plasma renin.15
PRA showed significant negative correlations with Aogen mRNA levels in hypothalamus and medulla oblongata, and plasma Aogen levels were positively correlated with Aogen mRNA levels in liver and hypothalamus. Several possible mechanisms exist by which renin might inhibit Aogen mRNA expression in tissues. High levels of active renin in plasma might suppress the tissue RAS by negative-feedback inhibition of the production of prorenin.16 Another possibility is that circulating renin itself may have inhibitory effects on Aogen mRNA levels in liver and brain. Several investigators have suggested direct actions of renin and/or prorenin.16 17 18
Our finding of three differently sized classes of Aogen mRNA in the liver was consistent both with reports describing at least four different Aogen mRNA species, transcribed from a single gene, that differ in the lengths of their 3′ untranslated regions in Aogen mRNA in the liver, brain, and kidney13 and with Ben-Ari et al,12 who described alternative 5′ start sites for Aogen transcription. There are no other reports of preferential utilization of one polyadenylation signal over another or selection of one start site over the other in transcription of the Aogen gene. A mechanism involving alternative use of signals within the Aogen gene could be reflected in differences in levels of the various Aogen transcripts in different tissues. With the renin infusions, we suggest that conditions were produced that led to selective loss of the larger Aogen mRNA transcripts in liver and perhaps in the medulla oblongata. Four polyadenylation sites have been identified in the Aogen gene and four sizes of Aogen mRNA species measured.13 With evidence that in addition to the normally used start site, either of two alternative transcription start sites can be induced by glucocorticoids,12 there are multiple possible combinations of start/stop signals that could be used to produce different sizes of Aogen mRNA species. In the present study, we found three Aogen mRNA species, and in contrast to the findings reported earlier,13 we detected differences in the relative intensities of Aogen mRNA bands in different tissues. Although the significance of these size differences is unclear, alternative use of polyadenylation sites or transcription start sites represents a potential means of regulating gene expression.12 13 19 20
In summary, by stressing the RAS with administration of exogenous renin, we produced conditions that reduced Aogen mRNA levels both centrally and peripherally. We attribute the renin effects on Aogen gene expression to indirect actions that selectively affect multiple Aogen transcripts in a differential manner.
Selected Abbreviations and Acronyms
|MAP||=||mean arterial pressure|
|PRA||=||plasma renin activity|
This work was supported in part by grant HL-6835 from the National Institutes of Health.
- Received November 3, 1995.
- Revision received January 12, 1996.
- Accepted May 12, 1996.
Ferrario CM, Milsted A, Santos RAS. Renin, converting enzyme and angiotensin peptides: a review. In: Keane WF, ed. A Focus on the Clinical Effects of a Long-Acting ACE Inhibitor/Hypertension. New York, NY: Raven Press Publishers; 1990:1-20.
Dzau VJ, Paul M, Nakamura N, Pratt RE, Ingelfinger JR. Role of molecular biology in hypertension research. Hypertension. 1989;13:731-740.
Gomez RA, Chevalier RL, Carey RM, Peach MJ. Molecular biology of the renin-angiotensin system. Kidney Int. 1990;38:S-18-S-23.
Harrap SB. A developmental genetic mechanism involving angiotensin in spontaneously hypertensive rats. Clin Exp Pharmacol Physiol. 1992;19(suppl 19):19-22.
Nishimura M, Milsted A, Block CH, Brosnihan KB, Ferrario CM. Tissue renin-angiotensin systems in renal hypertension. Hypertension. 1992;20:158-167.
Kohara K, Brosnihan KB, Ferrario CM, Milsted A. Peripheral and central angiotensin II regulates expression of genes of the renin-angiotensin system. Am J Physiol. 1992;262:E651-E657.
Kohara K, Brosnihan KB, Chappell MC, Khosla MC, Ferrario CM. Angiotensin-(1-7): a member of circulating angiotensin peptides. Hypertension. 1991;17:131-138.
Chan YL, Gutell R, Noller HF, Wool IG. The nucleotide sequence of a rat 18S ribosomal ribonucleic acid gene and a proposal for the secondary structure of 18S ribosomal nucleic acid. J Biol Chem. 1984;259:224-230.
Ben-Ari ET, Lynch KR, Garrison JC. Glucocorticoids induce the accumulation of novel angiotensinogen gene transcripts. J Biol Chem. 1989;264:13074-13079.
Ohkubo H, Nakayama K, Tanaka T, Nakanishi S. Tissue distribution of rat angiotensinogen mRNA and structural analysis of its heterogeneity. J Biol Chem. 1986;261:319-323.
Klett C, Bader M, Ganten D, Hackenthal E. Mechanism by which angiotensin II stabilizes messenger RNA for angiotensinogen. Hypertension. 1994;23(suppl I):I-120-I-125.
Hsueh WA, Baxter JD. Human prorenin. Hypertension. 1991;17:469-477.
Herrmann HC, Dzau VJ. The feedback regulation of angiotensinogen production by components of the renin-angiotensin system. Circ Res. 1983;52:328-334.