Human Renin-Dependent Hypertension in Rats Transgenic for Human Angiotensinogen
Abstract To examine the utility of rats transgenic for human angiotensinogen in the study of human renin-induced hypertension, we first developed assays to measure both the human and rat renin-angiotensin systems in these rats. We used human and mouse renin, transgenic human angiotensinogen, and the human renin inhibitor Ro 42-5892 to determine human- and rat-specific plasma angiotensinogen concentrations, renin activity, and renin concentration. The assays were validated with rat and human plasma mixed in known amounts and with plasma from rats transgenic for human renin. We then tested the human angiotensinogen–transgenic rats by infusing recombinant human renin over 10 days (50 ng/h, n=4) with osmotic minipumps. High human angiotensinogen transgene expression was found in the liver, brain, kidney, gastrointestinal tract, and aorta, whereas rat angiotensinogen gene expression was detected in the liver and brain. During human renin infusion, blood pressure increased to >200/150 mm Hg. Before infusion, human angiotensinogen was 100-fold greater than rat angiotensinogen (141±73 versus 1.2±0.16 μg angiotensin I/mL); the relation was not changed by renin infusion. Plasma renin activity increased 300-fold; human plasma renin concentration increased to very high levels (449±262 ng of angiotensin I per mL per hour), whereas rat plasma renin concentration decreased to undetectable levels. Thus, chronic human renin infusion resulted in severe hypertension with extreme plasma renin activity and plasma renin concentration. However, even at these levels, human angiotensinogen was not rate limiting and angiotensin II was not a significant stimulus for angiotensinogen production. We conclude that these transgenic rats represent a novel model of human renin-dependent hypertension.
The species specificity of the renin-angiotensinogen reaction makes the reaction rate of human renin with rAOGEN extremely slow relative to the cleavage of hAOGEN.1 Conversely, hAOGEN is cleaved to Ang I by human but not by rat renin, which makes the assessment of human renin inhibition impossible in rats. To study the effects of human renin in an animal model, we therefore developed a TGR that harbors the hAOGEN gene.2 We showed previously2 that a single acute injection of human renin into these TGRs increases their blood pressure compared with controls. We now examine the effects of chronic human renin infusion. To study human or rat renin and AOGEN in such animals, new specific measurement methods are necessary to validly and reliably determine renin-angiotensin system components in small plasma samples. We developed such methods and used them to characterize these TGRs as a model for human renin-induced chronic hypertension. Continuous telemetric blood pressure monitoring, together with species-specific determinations of rat and human renin-angiotensin system components, permitted us to explore the effects of chronic intravenous infusion of human renin in a rat model. This model will facilitate preclinical research on renin inhibitors.
Male Sprague-Dawley rats heterozygous for the complete human genomic angiotensinogen gene2 [TGR(hAOGEN)1623] and male Sprague-Dawley rats heterozygous for the complete human genomic renin gene [TGR(hRENIN)L10J] were used for all experiments. Rats transgenic for the human renin gene were generously provided by Prof K. Murakami, Tsukuba (Japan) University. The rats were housed and cared for as described elsewhere.3 All procedures were performed according to American Physiological Society guidelines with due approval of our institute.
In Vitro Enzyme-Kinetic Assays
This protocol was conducted to investigate whether hAOGEN and human renin concentrations can be measured validly and reliably by in vitro enzyme-kinetic methods in the presence of rat plasma. A pool of rat plasma was added to a pool of human plasma at stepwise increasing concentrations (0 to 100% by volume). Enzyme-kinetic incubations4 5 were performed to determine human and rat AOGEN concentrations and human- and rat-specific renin activities and concentrations in each plasma mixture. The human-specific renin inhibitor Ro 42-5892 (Hoffmann–La Roche) was used to distinguish human from rat renin.6 At 2.5×10−7 mol/L, this specific inhibitor does not interfere with rat renin but completely blocks the activity of human renin.
hAOGEN was cleaved by excess human renin and, in a separate assay, rAOGEN was cleaved by excess purified mouse submaxillary gland renin.7 Both human and mouse renin display high substrate specificity for either human or rat AOGEN simultaneously present in the artificial plasma mixtures.8 9 To determine rAOGEN concentrations, 50 μL plasma prediluted 1:100 in assay buffer was incubated together with 445 μL of 0.1 mol/L Tris-HCl buffer (pH 7.4) containing 0.025 mol/L Na2-EDTA, 1.0 g/L BSA (Sigma A-4378), and purified mouse submaxillary gland renin and with 5 μL PMSF (50 g/L ethanol) for 1 hour at 37°C. Mouse renin was capable of generating 95 μg·mL−1·h−1 of Ang I in this assay under conditions of substrate abundance. This was tested with an aliquot of rat plasma taken 48 hours after bilateral nephrectomy and added to the assay system to provide a final rAOGEN concentration of 2 nmol/mL. To determine hAOGEN concentrations, 25 μL of plasma prediluted 1:100 in assay buffer was incubated together with 470 μL of 0.15 mol/L citrate-phosphate buffer (pH 5.7) containing 0.025 mol/L Na2-EDTA, 1.0 g/L BSA, and 50 ng human recombinant renin and with 5 μL PMSF for 1 hour at 37°C. The Ang I generated was measured by direct radioimmunoassay and AOGEN concentrations were expressed as micrograms of Ang I per milliliter, based on an equimolar production of Ang I from AOGEN.
Plasma Renin Activity
Total PRA, as defined by the combined production of Ang I by both human and rat renins, was determined by incubating 100 μL of plasma together with 95 μL of Tris buffer (pH 7.4) and 5 μL PMSF for 1 hour at 37°C. Rat-specific PRA, as defined by the action of rat renin on rAOGEN, was determined by a similar assay performed in the presence of Ro 42-5892. Human-specific PRA, as defined by the action of human renin on hAOGEN, was calculated from the difference between total and rat-specific PRA.
Plasma Renin Concentration
To determine human PRC, 25 μL of each plasma sample was incubated together with 175 μL renin-free human plasma containing 4.7 nmol/mL AOGEN and providing excess substrate, 25 μL of 3.0 mol/L Tris-HCl buffer containing 0.2 mol/L Na2-EDTA,5 and 5 μL PMSF for 1 hour at pH 7.4 and 37°C. To obtain renin-free plasma, a pool of human plasma with high AOGEN levels was stripped from its renin and prorenin by twice incubating aliquots of 400 μL for 3 hours in assay tubes coated with an anti-human renin antibody that effectively retains both active renin and prorenin and is used in a renin measurement kit10 (No. 79986, Sanofi-Pasteur). Rat PRC was determined at pH 7.4 with a similar assay with 48-hour bilaterally nephrectomized rat plasma used as a source of excess rat substrate (4.1 nmol/mL AOGEN). These plasma pools were checked for absence of any residual Ang I generation activity. All renin determinations were repeated in the presence of Ro 42-5892 to control for species specificity during Ang I generation.
TGR as a Plasma Source for hAOGEN
We also investigated whether plasma from TGR(hAOGEN)1623 can be validly used as a source of hAOGEN during enzyme-kinetic determinations of human renin. A pool of 48-hour bilaterally nephrectomized TGR plasma without any residual renin activity containing 297.5 nmol/mL hAOGEN and 3.6 nmol/L rAOGEN was used. At first, 200 μL of increasing concentrations of TGR plasma (0.1% to 100% by volume diluted in assay buffer) was incubated together with 50 μL of 1.0 ng/mL human recombinant renin, 245 μL of citrate-phosphate buffer (pH 5.7), and 5 μL PMSF for 1 hour at 37°C. We then checked whether Ang I generation was a linear function of human renin concentrations when transgenic plasma concentrations were fixed. Fifty microliters of 0 to 1.0 ng/mL human recombinant renin was incubated together with 345 μL citrate-phosphate buffer (pH 5.7), 100 μL TGR plasma, and 5 μL PMSF for 1 hour at 37°C. Ang I generation rate was expressed as ng·mL−1·h−1.
Finally, this assay was verified with plasma from rats transgenic for human renin. Venous blood (2 mL) was obtained from six male TGRs by jugular vein puncture. Human PRC was measured with a 25-μL plasma-sample volume in the incubation assay discussed above and in parallel by the direct renin measurement kit. All incubations were repeated in the presence of Ro 42-5892.
Our radioimmunoassay methods for Ang I3 and hAOGEN11 as well as the techniques for RNase protection assay3 are described in detail elsewhere. The recovery rates for Ang I were >90% in our enzyme-kinetic assays. All plasma samples were anticoagulated with 6.25×10−6 mol Na2-EDTA per milliliter of blood.
Human Renin Infusion into hAOGEN TGR
TGR(hAOGEN)1623 were anesthetized by intra peritoneal injection of ketamine with xylazine (15 and 5 mg/kg body wt, respectively), and the right jugular vein was surgically exposed. A polyethylene catheter (0.61-mm diameter, 13-μL dead space) was inserted centrally into the vein, locally fixed, then subcutaneously routed to the interscapular region and exteriorized through a separate skin incision. An osmotic minipump (No. 2002, Alza Corp) with a discharge rate of 0.5 μL/h was connected to the catheter and implanted into a subcutaneous pouch. Incisions were closed by suture. Minipumps were filled with 100 μg/mL purified human recombinant renin diluted in sterile water containing 5 mg/L BSA. Recombinant human renin was a generous gift of Drs S. Mathews and W. Fischli, Hoffmann–La Roche AG, Basel, Switzerland. Control blood samples (0.6 mL) were drawn from the jugular vein catheter into syringes containing Na2-EDTA before the minipumps were placed. Further blood samples were drawn on day 2 and day 9 of the experiment by jugular vein puncture under short ether anesthesia.
Arterial blood pressure and heart rate were both continuously recorded with a radio telemetry system (Data Sciences International). Two weeks before the experiment, blood pressure sensors were implanted into the abdominal aorta of rats according to previously published surgical procedures.12 Blood pressure and heart rate were recorded on a computer at time intervals of 10 minutes from 5 days before until 12 days after minipump implantation.
Mean values with SDs or medians were calculated. We used Statview software on an Apple Macintosh computer to perform linear regression analysis and one-way ANOVA. A value of P<.05 was considered significant.
Human and Rat Renin-Angiotensin System Components
Fig 1⇓ shows results for the in vitro enzyme-kinetic determinations of human and rat AOGEN concentrations and human and rat plasma renin activity and concentration in our mixtures of human with rat plasma. The various parameters were successively determined in the same mixtures. In Fig 1⇓, the results are shown as a function of increasing human (0 to 100%) and decreasing rat plasma concentrations (100% to 0). Presence of linearity between expected (abscissa) and actually measured concentrations (ordinate) was used as a criterion for assay validity. As human plasma percentages increased (Fig 1A⇓), hAOGEN concentrations increased and rAOGEN concentrations decreased with close linearity maintained for both rat and human AOGEN determinations (r=.95, P<.001). Fig 1B⇓ shows the PRA determinations for the same plasma mixtures. As human plasma percentages increased, human PRA increased while rat PRA decreased. In the absence of excess AOGEN, these relations were not linear.
Results for PRC are shown in Fig 1C⇑ and 1D⇑. Rat PRC is shown in the absence and presence of human renin inhibitor (Fig 1C⇑). Rat PRC increased linearly (r=.97, P<.001) irrespective of the presence of the inhibitor. Fig 1D⇑ shows human PRC. Values have been corrected for residual rat PRA measured in the presence of Ro 42-5892. Human PRC increased in a linear fashion (r=.97, P<.01).
Fig 2A⇓ shows in vitro Ang I generation as a function of increasing hAOGEN concentrations from added nephrectomized TGR plasma while human renin concentration was constant. The Ang I generation tended toward a plateau at hAOGEN concentrations >30 nmol in the assay. Fig 2B⇓ conversely shows Ang I generation as a function of increasing human renin concentrations when 60 nmol transgenic hAOGEN from nephrectomized TGR plasma was present in the assay system. A linear relation was observed between actual (ordinate) and measured renin concentrations (abscissa) (r=.98, P<.001). The inset shows the lower end of the scale. At physiological human renin concentrations between 0 and 100 pg/mL, the same linear relation was observed.
To further validate this method, we tested plasma from six human renin TGRs. We compared the enzyme-kinetic assay described above with the commercially available kit for human renin. The results from these studies showed that the relation between Ang I generation and human renin concentration as measured directly was linear (r=.93, P<.01; data not shown). Furthermore, the relation between Ang I generation (y, ng·mL−1·h−1) and directly measured plasma renin concentration (x, pg/mL) was similar to our experiment (Fig 2B⇑) in which human recombinant renin was used as an independent renin source (y=0.15x−0.18 versus y=0.18x−0.47).
Tissue Expression and Plasma Levels of hAOGEN
Results from an RNase protection assay are shown in Fig 3⇓. One microgram of total RNA per organ was tested simultaneously for the presence of rAOGEN (upper band, 290 nucleotides) and hAOGEN mRNA (lower band, 132 nucleotides). rAOGEN was detected in normal (control) rat liver and in TGR liver, brain, and cerebellum. hAOGEN was found in all tested organs except normal rat liver. The transgene expression was relatively low in pancreas, spleen, and testes.
We measured plasma rAOGEN and hAOGEN in 41 male heterozygous hAOGEN TGRs. Rats were 5 months old and represented offspring from 3 homozygous male TGRs each mated with 3 outbred Sprague-Dawley female rats. All blood samples were drawn by jugular vein puncture on one occasion (ketamine/xylazine anesthesia) and tested together in one assay. rAOGEN levels in these animals were 1.17±0.16 μg Ang I/mL; the values were normally distributed. The values for hAOGEN were 141±98 μg Ang I/mL. The median was 99.1; the values were variable and not normally distributed. We also measured hAOGEN in these samples directly by radioimmunoassay (data not shown). The enzymatic measurement techniques (y, micrograms of Ang I per milliliter) and direct methods (x, milligrams per milliliter) gave similar results that were highly correlated (y=14.5x+54.8; r=.95, P<.001).
Human Renin Infusion in TGR
The effects of chronic human renin infusion (50 ng/h) in four hAOGEN TGRs are shown in the Table⇓. Systolic blood pressure increased by 90 mm Hg during the infusion and heart rate progressively increased by 65 beats per minute; body weight decreased. We measured rAOGEN, hAOGEN, PRA, rat PRC, and human PRC before the pumps were placed and after they had been operative for 2 and 9 days. PRA was determined at pH 7.4. Human PRC was determined at pH 5.7 as shown in Fig 2B⇑. The trend toward an increase of rAOGEN and a decrease in hAOGEN was not significant. Total PRA as well as human PRC increased in parallel with blood pressure, while rat PRC decreased.
The data show that human and rat AOGEN and human and rat PRA as well as PRC can all be measured simultaneously and specifically in <200 μL of rat plasma. Moreover, plasma from hAOGEN TGRs can be used to measure the enzymatic activity of human renin when an excess of homologous substrate is necessary. We found that TGRs exhibit heterogeneous values of hAOGEN but nevertheless regularly develop prompt and severe hypertension during human renin infusion.
However, observing simultaneously the dynamics of the human and rat renin-angiotensin systems simultaneously requires new and valid micromeasurement methods, which we have specifically developed for this purpose. We measured rAOGEN and hAOGEN in the same samples by relying on the specificities of human and rodent renins on their respective substrates. We further distinguished between rat and human PRA and PRC by means of the human renin inhibitor Ro 42-5892, and direct measurement methods were only used as a validation of our enzyme-kinetic techniques. We used a direct immunoradiometric assay of human renin in plasma that uses two monoclonal antibodies that do not recognize the same epitope. The assay has been successfully developed to distinguish plasma active from plasma total renin concentration.10 However, enzyme-kinetic PRA and PRC determinations are mandatory in our model because of various sample requirements and because of the lack of an established direct assay of rat renin.13 14
Gahnem et al1 mixed human and rat plasma in their study of human renin inhibition by rat plasma. They used approaches similar to ours and found that human renin can cleave rat substrate, albeit at an extremely slow rate. They also reported that rAOGEN is a weak competitive inhibitor of the hAOGEN–human renin interaction. Therefore, rAOGEN possibly binds and inhibits human renin in vivo while also influencing the results of our in vitro enzyme-kinetic assays. In our model, determination of the various plasma parameters serves different purposes. The in vitro measurement of total PRA aims at investigating the overall in vivo plasma Ang I generation rate to which transgenic animals will be exposed when human renin is infused. This value does not differentiate rat from human systems. The determination of rat- and human-specific PRA, as distinguished by Ro 42-5892, gives the separate human and rat renin contributions to total PRA. This information remains relevant for the in vivo situation, even if human renin would be inhibited slightly by the presence of rAOGEN. With an objective to determine actual PRCs, the use of excess homologous substrate in the PRC assay will therefore minimize the issue of possible interactions of rat substrate with human renin. We verified the linearity of our human PRC assay that uses transgenic plasma through comparison with direct immunoradiometric methods. Under our assay conditions, any influence of rat renin substrate appeared to be negligible. As for human AOGEN, plasma concentrations were 100× the concentration of rAOGEN in TGR. Thus, in our enzyme-kinetic determinations, an error introduced through possible cleavage of rAOGEN by human renin was in the magnitude of <1%. Furthermore, we also validated this method by a direct radioimmunoassay. Finally, during human renin infusion, human PRC levels were so high compared with rat renin that any error that resulted from residual rat renin activity would hardly influence the precision of our human PRC determinations. Our radioimmunoassay for Ang I has been standardized with the international standard preparation of Ang I (71/328). The fact that mathematical regression lines comparing direct with indirect measurement methods did not always pass through zero was probably an effect of dilution errors or variations in the actual protein content of the AOGEN and renin preparations used as standards.
Preparation of renin-free human plasma for PRC determinations is time-consuming. Enzyme-kinetic assays that use hAOGEN from TGR plasma may therefore also facilitate laboratory investigation of plasma renin in humans. We pragmatically looked for assay conditions that provide linearity of renin determinations with a minimum dependency on variations of transgenic hAOGEN concentrations. These criteria were achieved with 30 nmol hAOGEN in the assay, equivalent to a concentration of 12 nmol/mL. In this context, we tested Ang I generation by adding increasing amounts of transgenic nephrectomized plasma to the assay system, which increased its plasma protein concentration by a factor of 100. Plasma protein content, ionic strength, ion composition, the rAOGEN, and a variety of other physicochemical factors, however, are well known to influence the human renin substrate interaction in an enzyme-kinetic assay. This would explain why in an experimental system with purified renin and angiotensinogen, others found a plateau in Ang I generation at a much lower hAOGEN concentration than we did.9 With our methods, reliable results can be expected when assay conditions are kept constant, internal controls are used, and a pool of nephrectomized plasma is prepared from a large number of TGRs tested for hAOGEN content before use.
The nongaussian distribution of plasma hAOGEN levels in TGRs remains an intriguing observation. Our preliminary data (not shown) from a protection assay study indicate that plasma hAOGEN levels would correlate with hAOGEN mRNA expression in the liver. Thus, in these TGRs, the genetic background, site of transgene insertion, number of active transgene copies, or factors specifically affecting mRNA transcription are obviously important factors that control transgenic plasma protein levels rather than, for example, exogenous or environmental influences.
We and other investigators2 15 have previously examined the effects of human renin infusion into normal rats and found little effect on blood pressure. The presence of an acceptable substrate, hAOGEN, in TGRs made them very susceptible to human renin. The rats thus have two pathways for Ang II generation, namely, their own and that engendered by the transgene and the human renin infusion. We found that after a 9-day human renin infusion, mean 24-hour blood pressure increased to values >200/150 mm Hg. In plasma, human PRC increased to very high levels of ≈300 ng · Ang I · mL−1·h−1, equivalent to 2 ng/mL of human renin. At day 2, the values were much lower, with a large interindividual variability compared with day 9, probably because the pumps required unequal times to overcome dead space. The decrease of rat renin in plasma indicates that the normal Ang II-dependent negative feedback on rat renin secretion by the kidneys was still effective. In this small group of rats, plasma hAOGEN values barely decreased (P=NS), suggesting that hAOGEN was certainly not rate limiting in our study. Global PRA increased 300-fold, which indicated that the TGRs were indeed a model of high human renin hypertension. Gahnem et al16 found that rAOGEN values were inversely correlated with PRC in rats given losartan, an Ang II receptor blocker. In our study, hAOGEN was 100-fold higher than the rat values and was present to such an excess that even the extremely high PRA values did not result in an hAOGEN rate-limiting system.
We have previously suggested the hAOGEN TGR as a potential model for studying renin-angiotensin–related effects at the tissue level.3 We used an RNase protection assay to localize and compare the expression of human and rat AOGEN in various TGR organs. rAOGEN was readily detectable in liver, brain, and cerebellum. In contrast, transgene expression was more marked in these tissues and also present in many other organs where ordinary AOGEN mRNA expression has been detectable only under more sensitive conditions.17 18 19 20 21 The hAOGEN expression in these rats is surely exaggerated compared with humans, and the amounts of human renin we infused were likely far beyond those necessary for physiological or pathological effects. Furthermore, the reconstitution of the human renin substrate interaction in the rat surely makes it susceptible to influences that do not normally act in humans. Various other aspects of this model also remain to be investigated for a more complete understanding of its pathophysiological characteristics. For example, the regulation of transgene expression, dose-response relationships, and the dynamics of the plasma renin-angiotensin system still need to be clarified. Also, the number of animals studied was too small to make conclusions concerning reproducibility and reliability of results. Nevertheless, we believe our model, coupled with the new assay systems we have developed, will have considerable utility in examining the human renin-angiotensin system and its potential pathological effects under controlled experimental conditions.
Selected Abbreviations and Acronyms
|BSA||=||bovine serum albumin|
|PRA||=||plasma renin activity|
|PRC||=||plasma renin concentration|
|TGR(hAOGEN)1623||=||male Sprague-Dawley rats heterozygous for the complete human genomic angiotensinogen gene|
|TGR(hRENIN)L10J||=||male Sprague-Dawley rats heterozygous for the complete human genomic renin gene|
This study was supported by a grant-in-aid from Hoffmann–La Roche AG, Basel, Switzerland. We thank Drs Fischli and Mathews (Hoffmann–La Roche) for providing us with human recombinant renin and Dr K. Murakami (Tsukuba University, Japan) for allowing us to use his human renin TGR strain. We also wish to thank I. Strauss, C. Lipka, A. Müller, and M. Somnitz for their expert technical assistance during the experiments. Dr Ménard is a 1994 Alexander V. Humboldt scholar.
Gahnem F, Sealey JE, Atlas SA, Laragh JH. Inhibition of human renin by rat plasma: rat angiotensinogen is a competitive inhibitor of the human renin substrate interaction. Am J Hypertens. 1992;5:495-501.
Ganten D, Wagner J, Zeh K, Bader M, Michel JB, Paul M, Zimmermann F, Ruf P, Hilgenfeldt U, Ganten U, Kaling M, Bachmann S, Fukamizu A, Mullins JJ, Murakami K. Species specificity of renin kinetics in transgenic rats harboring the human renin and angiotensinogen genes. Proc Natl Acad Sci U S A. 1992;89:7806-7810.
Müller D, Hilgers K, Bohlender J, Lippoldt A, Wagner J, Fischli W, Ganten D, Mann JFE, Luft FC. Effects of human renin in the vasculature of rats transgenic for human angiotensinogen. Hypertension. 1995;26:272-278.
Fischli W, Clozel JP, Amrani KE, Wostl W, Neidhard W, Stadler H, Branca Q. Ro 42-5892 is a potent orally active renin inhibitor in primates. Hypertension. 1991;18:22-31.
Suzuki F, Nakamura Y, Nagata Y, Ohsawa T, Murakami K. A rapid and large scale isolation of renin from mouse submaxillary gland by pepstatin-aminohexyl-agarose affinity chromatography. J Biochem. 1981;89:1107-1112.
Ménard J, Guyene TT, Corvol P, Pau B, Simon D, Roncucci R. Direct immunometric assay of active renin in human plasma. J Hypertens. 1985;3(suppl 3):S275-S278.
Kim S, Iwao H, Nakamura N, Ikemoto F, Yamamoto K. Fate of circulating renin in conscious rats. Am J Physiol. 1987;252:E136-E146.
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.
Campbell DJ, Habener JF. Angiotensinogen gene is expressed and differentially regulated in multiple tissues of the rat. J Clin Invest. 1986;78:31-39.
Paul M, Wagner J, Dzau VJ. Gene expression of the renin angiotensin system in human tissues: quantitative analysis by the polymerase chain reaction. J Clin Invest. 1993;91:2058-2064.