Intrarenal Angiotensin II Formation in Humans
Evidence From Renin Inhibition
Abstract The intrarenal production of angiotensin II (Ang II) as a local hormone, suggested by multiple lines of investigation, has been difficult to buttress with evidence of functional significance in humans. During studies designed to assess the renal vascular responses to the renin inhibitor enalkiren, an agent (like others in its class) with great substrate specificity, we noted in some subjects that the time course of the effect of enalkiren on renal plasma flow was not congruent with the time course of its influence on the renin-angiotensin system in the plasma compartment. We pursued this discrepancy in the current study of 18 healthy men and 9 men with essential hypertension, who each received one or more doses of enalkiren while on a fixed sodium diet. Plasma enalkiren and Ang II concentration and renal plasma flow were measured in each subject at intervals during and after discontinuation of the enalkiren infusion. Plasma enalkiren concentration fell progressively in each subject after administration was discontinued, the fall becoming evident 10 minutes after discontinuation without exception. In plasma samples obtained 90 minutes after the end of the infusion, drug levels were generally less than half of their peak value. Plasma Ang II concentration, at nadir levels by the end of the enalkiren administration, rose consistently during recovery. Renal plasma flow, in contrast, rose during infusion but did not begin to fall when enalkiren was discontinued. In 26 of 31 studies, renal plasma flow remained at peak level or even continued to rise; this discordance in the effects on plasma Ang II concentration and on renal plasma flow after discontinuation of enalkiren was highly significant (P<.0005). Sustained renal vascular activity of the renin inhibitor, in marked contrast to waning enalkiren concentration and activity in the plasma compartment, provides strong evidence for an action at the tissue level and for a biological influence of intrarenal Ang II formation in humans.
Multiple lines of evidence in animal models have suggested not only the intrarenal production of angiotensin II (Ang II) but also its importance as a local hormone.1 2 3 4 5 6 7 8 9 10 11 12 13 However, opportunities for extending these concepts to humans and assessing the functional significance of intrarenal Ang II formation in human disease have been limited.14
During studies designed to assess the renal vascular responses to the renin inhibitor enalkiren, an agent (like others in its class) with great substrate specificity, we noted in some subjects that the time course of the effect of enalkiren on renal plasma flow (RPF) was not congruent with the time course of its influence on the renin-angiotensin system in the plasma compartment.15 This observation led to the current study, in which we compare the effect of enalkiren on RPF with its plasma concentration and influence on plasma Ang II concentration in humans, studied when in balance during a restricted sodium intake. The findings provide evidence for an influence of enalkiren locally, presumably in the kidney, and thus for the primacy of intrarenal angiotensin formation as the determinant of renal vascular tone when the renin-angiotensin system is activated.
The subjects were 18 healthy men ranging in age from 20 to 49 years (mean±SEM, 35±2.1 years) and 9 men with essential hypertension (27 to 61; 47±3.7 years). Two of the healthy subjects and two of the hypertensive subjects were black. The remainder were white. All were free of cardiovascular, renal, and endocrine disease, except for blood pressure elevation in the hypertensive subjects, whose untreated blood pressures were between 90 and 110 mm Hg diastolic and less than 170 mm Hg systolic. All were within 20% of ideal body weight. The acute renal vascular response to enalkiren has been reported in these subjects.15 16
After an outpatient evaluation, during which secondary forms of hypertension were excluded by history, physical examination, and appropriate laboratory studies, hypertensive subjects were taken off all antihypertensive agents for at least 3 weeks before study. All subjects were studied during a 10-day admission to a metabolic ward, the Clinical Research Center of the Brigham and Women’s Hospital. Written informed consent was obtained from each subject, and the protocol was approved by the Human Subjects Committee of the institution.
All subjects were placed on low salt, constant isocaloric diets through the hospitalization, with 10 mmol daily sodium intake; this was followed by 2 to 3 days of high salt intake (200 mmol Na+) before the final study in four of the healthy subjects. Daily dietary potassium (100 mmol) and fluid intake (2500 mL) were constant. Twenty-four-hour urine samples were collected daily and analyzed for sodium, potassium, and creatinine. When 24-hour urinary sodium matched sodium intake (usually on day 5), the first study was initiated.
All of the subjects received enalkiren by intravenous infusion. Studies began at about 7 am. Subjects had been recumbent and fasting overnight and remained recumbent throughout the study. A 60-minute control period was used to establish basal effective RPF. Then blood samples were drawn for measurement of plasma Ang II concentration (by either radioimmunoassay or high-performance liquid chromatography [HPLC]). After blood samples were drawn, an infusion of enalkiren was superimposed for 90 minutes. Repeat hormonal measurements were taken at the end of the infusion and either 45 or 90 minutes later while the subjects remained recumbent. Each subject received one or more doses of enalkiren, ranging from 0.128 to 1 mg/kg, the dose range shown to be effective in inducing renal vasodilation.15 When multiple doses were administered, they were separated by a rest interval of 48 to 72 hours.
Blood pressure during each infusion was recorded by an automatic recording device (Dinamap, Critikon Inc) at 5-minute intervals, and the electrocardiogram was monitored continuously.
Renal Clearance Studies
Para-aminohippurate (PAH) (Merck, Sharp & Dohme) clearance was assessed after a metabolic balance had been achieved in subjects on each diet. An intravenous catheter was placed in each of the subject’s arms, one for infusion and the other for blood sampling. A control blood sample was obtained, and a loading dose of PAH (8 mg/kg) was given. A constant infusion of PAH was initiated immediately at a rate of 12 mg/min with a pump (IMED Corp). This infusion rate achieves a plasma PAH concentration in the middle of the range in which tubular secretion dominates excretion. At this plasma level of PAH, clearance is independent of plasma concentration and, when corrected for individual body surface area, represents about 90% of RPF. PAH clearance was calculated from the plasma levels and infusion rates for each subject.17 Plasma samples reflecting the control clearance were obtained 60 minutes after the start of the PAH infusion, when a steady state had been achieved, and at 45-minute intervals thereafter.
Blood samples were collected on ice at the start of the infusion, the end of the infusion, and at follow-up either 45 or 90 minutes later and were spun immediately; the plasma was frozen until the time of assay. Plasma Ang II concentration was measured in all of the subjects. In 13 subjects, HPLC was used to separate Ang II from fragments,16 and in the remaining subjects, immunoreactive Ang II was measured by radioimmunoassay as follows.18
One milliliter of 8 mol/L urea with 10 mmol triethanolamine (TEA) per liter was added to individual 1-mL plasma samples as they slowly thawed. The final concentration, 4 mol/L urea, is sufficient to block any synthetic or degradative pathway of the renin-angiotensin system. For assessment of the influence of 8 mol/L urea on the stability of Ang II, tritiated Ang II was exposed to plasma and remained intact (100±3%), without addition of angiotensin-converting enzyme inhibitors. Similarly, the specific activity of Ang II remained constant, indicating that hypertonic urea limited Ang II formation. Thawed and mixed plasma samples were extracted on 3-mL, 500-mg C18 Sep-Pak cartridges (Waters Chromatography Division, Millipore) with the use of a vacuum extractor device (VacElut SPS 24, Analytichem International). The cartridges were first activated with 2 mL of 0.1% TEA, followed by 2 mL of 80% methanol with 0.1% TEA, and finally with 1 mL of 0.1% TEA. After extraction of the plasma samples, the cartridges were rinsed with 1 mL saline. Absorbed angiotensins were eluted from the cartridges with 4 mL of 80% methanol. Fifty microliters of 10% glycerol was added to each extract, and the samples were dried overnight in a SpeedVac dryer (Savant). The recovery associated with the extraction procedure, assessed with tritiated Ang II in five samples, was 75±5%.
After reconstitution of 550 μL sample solvent (10 mmol/L sodium acetate, 10 mmol/L TEA, 5% methanol, 0.15 mol/L Na2HPO4, 10 mL assay buffer per 500 mL), 500-μL aliquots were injected into a 5-μm 0.4×12.5-cm C18 column (Pharmacia LKB). The HPLC apparatus included an LKB 2150 pump, 2152 controller, and 2211 fraction collector. Solution A contained 30% methanol, 10 mmol/L TEA, and 10 mmol/L sodium acetate titrated to pH 6.2. Solution B was similar but contained 80% methanol. The elution times of the angiotensin peptides were assessed by injecting 4-nmol aliquots of each peptide and reading the absorbance with an LKB 2151 variable-wavelength monitor tuned to 214 nm and connected to an IBM XT computer loaded with chromatochart software (IMI). After the proper gradient was set and the SD of the elution times of each peptide was determined, the system was washed for several days until baseline peptide levels, determined by radioimmunoassay, were low. Fractions were collected every minute over 72 minutes in polypropylene test tubes containing 50 μL of 10% glycerol and 150 μL of 50% assay buffer (0.05 mol/L K2HPO4, 0.003 mol/L EDTA, 0.02% sodium azide, 0.01% Triton X-100 [Serva]) and then dried overnight in a SpeedVac concentrator.
Samples were reconstituted in 50 μL of a 50% assay buffer and 2.5 mg/mL radioimmunoassay-grade bovine serum albumin (Sigma Chemical Co) containing 125I–Ang II (DuPont–New England Nuclear) and 100 μL of assay buffer containing Ang II antibody (Arnel Inc) and were incubated for 48 hours at 4°C. Two hundred microliters of donkey anti-rabbit magnetic separation reagent (Amersham International) was added, and samples were placed into magnetic test tube holders 15 minutes later. The trays were emptied after 10 minutes, washed with 750 μL of buffer (0.1% gelatin, 0.01% Triton X-100, 0.05 mol/L NaCl, 0.10 mol/L MgCl2, 0.02% sodium azide), and again emptied after 10 minutes. Tracer counts were recorded on a Micromedic 4/200 automatic gamma counter for 3 minutes per tube. Counts were converted to femtomoles of Ang II using standard curves and plotted with an ria-aid software package (RMA Inc). peakfit software (Jandel Scientific) was used to analyze the area under each peak and to calculate total Ang II per sample. Recovery in the HPLC system was 94±3%. The results have not been corrected for recovery. This method produced a lower limit of detection for angiotensins of 0.2 fmol per tube. Buffer blank was 0.1 to 0.2 fmol per tube, and plasma blank was similarly 0.1 to 0.2 fmol per tube. During HPLC separation, the Ang II distributed in 5 of the 80 tubes used, with approximately 90% distributed in 3 central tubes. Making correction for the losses associated with processing described in the recovery experiments above, the sensitivity for detection of Ang II in plasma of this assay is 0.6 to 0.7 fmol/mL.
Plasma enalkiren levels were measured by HPLC in nine subjects receiving enalkiren (Abbott Laboratories) doses of 128 μg/kg per 90 minutes or higher at points throughout the infusion and into recovery. Briefly, enalkiren and the internal standard (IS Abbott-75247) were extracted from a 1-mL aliquot of plasma (sample or standard) into a mixture of ethyl acetate and hexane (7:3). After evaporation of the ethyl acetate/hexane layer under nitrogen, the residue was reconstituted with mobile phase and transferred to microvials. A 100-μL volume was injected into the C18 column (Beckman Ultrasphere ODS 7.5 cm×4.6 mm ID, 3-μm particle size), where enalkiren and internal standard were separated from endogenous substances. The UV absorbance (model 484, Waters) was monitored at 205 nm. The mobile phase was a mixture of methanol, acetonitrile, and 0.01 mol/L tetramethylammonium perchlorate in 0.05% trifluoroacetic acid (pH 3.0). The relationship between concentration and peak height ratio (enalkiren to internal standard) was linear within the range of 0 to 3000 ng/mL.
Interassay coefficients of variation, as estimated from the results of the analyses of quality control samples, ranged from 5.2% to 6%. The calibration curves were linear, with a mean regression correlation coefficient of .994. The limit of detection was 10 ng/mL.
Far smaller doses of a renin inhibitor are needed to make plasma renin activity unmeasurable than to induce a biological response (influence on blood pressure, the kidney, the adrenal, or plasma Ang II concentration). All data were obtained from enalkiren doses above biological threshold for each analysis. For RPF, doses of 128 μg/kg and higher and for Ang II, doses of 256 μg/kg and higher were used.15 Although we entertained no prior hypotheses concerning racial differences in renal vascular responsiveness, the responses of whites and blacks appeared to be quite different during analysis. We therefore describe results for the whites and blacks separately, in both the text and figures. The overall conclusion of this study—that the time course of the renal vascular response is discordant with events in the vascular compartment—was not influenced by this decision.
Group means are presented with the SEM as the index of dispersion. A χ2 analysis was used to determine differences between RPF responses at 45 and 90 minutes and between postinfusion plasma Ang II and RPF levels. The difference between black and white postinfusion RPF changes was estimated by the Wilcoxon rank sum test. We used the average response in each subject. The alpha level for significance was .05 or less.
Plasma enalkiren concentration, as anticipated, fell consistently and continuously in every subject after the enalkiren infusion was discontinued (Fig 1⇓). In plasma samples obtained 90 minutes after the end of the infusion, drug levels were generally less than half of their peak value.
Plasma Ang II concentration in the whites had fallen to nadir levels by the end of the enalkiren infusion. After enalkiren administration was discontinued, there was a consistent rise in plasma Ang II concentration in samples obtained 45 and 90 minutes later (Fig 2⇓). This rise was apparent with either assay method for plasma Ang II, with the expected distribution of HPLC-measured levels at a lower range (<10 fmol/mL), because this assay measures authentic Ang II, having separated other metabolites discernible as immunoreactive Ang II.
In contrast, RPF did not begin to fall when the enalkiren infusion ended (Fig 3⇓). In 26 of the 31 studies performed in white subjects, RPF either continued at the peak level achieved or continued to rise after the end of enalkiren administration. Indeed, the rise in RPF at 90 minutes after discontinuation (median, 29; mean, 30.1±9.7 mL/min per 1.73 m2; paired t test, P=.0078) significantly exceeded the rise at 45 minutes (median, 0; mean, 3.9±15.9 mL/min per 1.73 m2; P<.05). The discordance in the time course of plasma Ang II concentration and RPF after discontinuation of the renin inhibitor administration was significant (χ2=21.5, P<.0005). These results were not influenced by the subjects’ dietary sodium content. Furthermore, we have described a separate placebo infusion in 18 of these subjects.16 During the crucial time interval after discontinuation of the infusion, RPF was very stable.
Plasma Ang II concentration in the blacks, as in the whites, had fallen to its lowest level by the end of the enalkiren infusion and rose after the enalkiren infusion had been discontinued in all nine studies performed (Fig 4⇓). Moreover, the rise in RPF during enalkiren infusion in blacks (94±35 mL/min per 1.73 m2) was essentially identical to the response in whites (89±17.5 mL/min per 1.73 m2). In contrast, RPF measurements in the blacks revealed a pattern different from that in whites after discontinuation of drug. RPF fell in six of nine studies in blacks (Fig 4⇓). This unanticipated racial difference in postinfusion RPF changes was statistically significant (Wilcoxon rank sum test, P≅.036).
The discordance between renal hemodynamics and plasma Ang II was not influenced by the division of study groups into whites and blacks. The difference in time course of plasma Ang II and RPF was highly significant even with all subjects combined (χ2=24.7, P<.0005). Similarly, there were no differences seen when either the hypertensive or normotensive subjects were examined separately.
We have described a dramatic discordance of biological responses to the renin inhibitor enalkiren. The anticipated evidence of a waning plasma concentration and activity of the agent in the plasma compartment after discontinuation of the infusion stands in contrast to its continued renal vascular effects. The sustained influence of the agent on RPF in the face of a rising plasma Ang II concentration suggests that enalkiren exerted its main influence outside of the plasma compartment. We hypothesize that enalkiren-induced interruption of local, intrarenal formation of Ang II in whites is responsible.
Local intrarenal Ang II formation was demonstrated by Admiraal et al14 from studies involving infusions of labeled Ang I and Ang II in humans. Their experiments provide unambiguous evidence for the source of local Ang II production, given that Ang II is recovered from selective renal venous catheters in greater amounts than expected, based on the inflowing arterial renin-angiotensin system. Our study supplies functional evidence for the importance of this extra-plasma compartment for the renal circulation, and thus the two study techniques are complementary and might fruitfully be applied together in the future.
Our findings in blacks were unanticipated and involve only a small number of observations. Despite the absence of a prior hypothesis, we undertook a separate analysis to examine preliminary but intriguing evidence for an underlying racial difference in a local renin-angiotensin system. The rise in RPF during enalkiren administration was essentially identical in blacks and whites, and an identical fall in plasma Ang II was documented. Thus, the contribution of Ang II to the maintenance of renal vascular tone based on circulating Ang II was probably identical in the two groups. As drug levels fell and plasma Ang II rose after enalkiren discontinuation in blacks, RPF fell toward baseline, as anticipated if circulating Ang II were the dominant influence. The possibility exists that intrarenal Ang II in blacks might be regulated differently or have different functional significance than in whites. Indeed, previous studies have documented racial differences in the renal circulation. Levy et al19 reported significantly lower renal blood flow in black hypertensive patients than in whites, and their findings were confirmed by Frohlich et al,20 who showed lower renal blood flow and increased renal vascular resistance in blacks. Although speculative, our results might eventually lead to a better understanding of well-known clinical profiles of black hypertension, including a relatively high frequency of low-renin hypertension, lesser antihypertensive efficacy of angiotensin-converting enzyme inhibition, and a predilection to developing end-stage renal disease as a complication of diabetes or hypertension.
This study is limited in not providing definitive evidence that the site of extra-plasma Ang II is in fact the kidney. The association is probable but not proven. Furthermore, results of this study depend on measurement of PAH clearance and are therefore potentially limited by errors inherent in clearance techniques, especially during non–steady-state conditions. However, given the reliability of the assay method, the capacity of the method to measure the dilator response during enalkiren administration, and the known pharmacokinetics, the time interval assessed, up to 90 minutes, was adequate to have documented a fall, had it occurred. Strong internal evidence for the continued renal vascular action of enalkiren, even after its effects in the plasma were dissipating, lies in the statistically greater postinfusion rise in RPF at 90 minutes compared with 45 minutes in whites. This difference is consistent with the continued activity or accumulation of enalkiren in the kidney at a site that is perhaps relatively inaccessible.
We were able to extract meaningful results from this study because the pharmacological interruption used was intravenous, so the peak was predictable. It is much less likely that a discordance between events occurring within and outside the plasma compartment would be distinguishable if an oral agent were used to interrupt the renin-angiotensin system. Extended studies of the time course of renal vascular changes with intravenous angiotensin-converting enzyme or renin inhibition might allow even better elucidation of this extra-plasma, Ang II–dependent renal vascular tone and may prove useful for exploring the contribution of intrarenal Ang II formation in health and disease.
This work was supported in part by National Institutes of Health grants 5 T32 HL-07609 and 5 M01 RR 02635, the Canadian Medical Research Council Fellowship Award, and Abbott Pharmaceutical Research Laboratories. Dr Naomi Fisher was supported through a Clinical Associate Physician Award from the National Institutes of Health. It is a pleasure to acknowledge the nursing support of Charlene Malarick, RN, the research support of Diane Passan, RT, and the administrative support of Diana Capone in the preparation and submission of this manuscript.
Reprint requests to Norman K. Hollenberg, MD, PhD, Brigham and Women’s Hospital, 75 Francis St, Boston, MA 02115.
- Received April 6, 1994.
- Revision received May 4, 1994.
- Accepted January 16, 1995.
Johnston CI. Renin-angiotensin system: a dual tissue and hormonal system for cardiovascular control. J Hypertens Suppl. 1992;10:13-26.
Campbell DJ. Circulating and tissue angiotensins. J Clin Invest. 1987;79:1-6.
Inagami T, Mizuno K, Naruse K, Okamura T, Kawamura M. Intracellular formation and release of angiotensins from juxtaglomerular cells. Kidney Int Suppl. 1990;38:33-37.
Hunt MK, Ramos SP, Geary KM, Norling LL, Peach MJ, Gomes RA, Carey RM. Colocalization and release of angiotensin and renin in renal cortical cells. Am J Physiol. 1992;263:363-373.
Siragy HM, Howell NL, Peach MJ, Carey RM. Combined intrarenal blockade of the renin-angiotensin system in the conscious dog. Am J Physiol. 1990;258:F522-F529.
Fox J, Guan S, Hymel AA, Navar LG. Dietary Na and ACE inhibition effects of renal tissue angiotensin I and II and ACE activity in rats. Am J Physiol. 1992;262:F902-F909.
Campbell DJ, Lawrence AC, Towrie A, Kladis A, Valentijn AJ. Differential regulation of angiotensin peptide levels in plasma and kidney of the rat. Hypertension. 1991;18:763-773.
Reams G, Villarreal D, Bauer JH. Intrarenal metabolism of angiotensin II. Am J Physiol. 1990;258:F1510-F1515.
Danser AHJ, Koning MMG, Admiraal PJJ, Sassen LMA, Derkx FHM, Verdouw PD, Schalekamp MADH. Production of angiotensins I and II at tissue sites in intact pigs. Am J Physiol. 1992;263:H429-H437.
Seikaly MG, Arant BS, Seney FD. Endogenous angiotensin concentrations in specific intrarenal fluid compartments in the rat. J Clin Invest. 1990;86:1352-1357.
Admiraal PJJ, Danser AHJ, Jong MS, Pietermann H, Derkx FHM, Schalekamp MADH. Regional angiotensin II production in essential hypertension and renal artery stenosis. Hypertension. 1993;21:173-184.
Cordero P, Fisher ND, Moore TJ, Gleason R, Williams GH, Hollenberg NK. Renal and endocrine responses to a renin inhibitor, enalkiren, in normal humans. Hypertension. 1991;17:510-516.
Fisher NDL, Allan D, Kifor I, Gaboury CL, Williams GH, Moore TJ, Hollenberg NK. Responses to converting enzyme and renin inhibition: role of angiotensin II in humans. Hypertension. 1994;23:44-51.
Shoback DM, Williams GH, Moore TJ, Dluhy RG, Podolsky S, Hollenberg NK. Defect in the sodium-modulated tissue responsiveness to angiotensin II in essential hypertension. J Clin Invest. 1983;72:2115-2124.
Levy SB, Talner LB, Coel MN, Holle R, Stone RA. Renal vasculature in essential hypertension: racial differences. Ann Intern Med. 1978;88:12-16.