Radioimmunoassay of plasma renin activity

Radioimmunoassay of Plasma Renin Activity Jean E. Sealey and John H. Laragh Plasma renin activity is quantitated by measuring the rate of angiotensin generation during incubation of plasma renin with endogenous renin substrate. The angiotensin is quantitated by radioimmunoassay. Our studies indicate that the incubation step is best carried out in undiluted plasma at the pH optimum for renin (pH 5.7) in the presence of EDTA, neomycin, and DFP or PMSF. By using these conditions, incubation of Iow-renin samples can be prolonged for up to 18 hr, because angiotensinases and converting enzyme are completely inhibited. Accuracy is enhanced by prolongation of the incubation time, which results in more angiatensin and eliminates the need for blank subtraction. Incubation at pH 7.4 is disadvantageous because of lower rates of generation of angiotensin I, because of the inability to maintain pH constant without addition of strong buffer, and because the incubation step cannot be prolonged beyond 3 hr. Dilution of plasma is undesirable because it results in a slower reaction rate due to dilution of both enzyme and sub-

strate, and it is not possible to correct for the effect of substrate dilution. The radioimmunoassay of angiotensin I presents few unusual problems. The volume of plasma assayed should be 20/zl or less. If blank subtraction is used, antibodies should be screened to determine the extent to which they bind nonspecific substances in plasma, and then to ascertain whether the blank is entirely additive when angiotensin is added to it. Assay sensitivity is an important issue, since approximately 30% of patients with essential hypertension have subnormal plasma renin activity. In a study of three different commercial kits we found that many Iow-renin samples were undetectable and major fractions could not be discriminated with precision or consistency from normal renin samples. However, the incubation conditions can be easily modified, so that an 18-hr incubation can be utilized and then Iow-renin samples can be detected with the same degree of accuracy as those with normal plasma renin activity.

INTRODUCTION AND HISTORICAL PERSPECTIVE

ENIN WAS discovered in 1898 by Robert Tigerstedt and Per Bergman, who noted an increase in the blood pressure of anesthetized rabbits when saline extracts of kidney were injected. ~These observations remained in dispute until 1938, when three groups confirmed this earlier finding.2-4 Meanwhile, other studies had implicated the kidney in the hypertensive process? and in 1934 Goldblatt produced chronic experimental hypertension for the first time by placing a clip around the renal artery of an uninephrectomized dog. 6 In 1939 Page v and Braun-Menendez and co-workers 8 reported that renin was not by itself a pressor substance but acted in plasma as an enzyme to release a pressor peptide, angiotensin. Angiotensin is released from a circulating plasma globulin 9'm and has been shown by Skeggs and associates and by Peart's

R

From the Department of Medicine, Columbia Presbyterian Medical Center, New York. N. Y.

Jean E. Sealey, D.Se.: Research Associate in Medicine; John H. Laragh, M.D.: Professor of Clinical Medicine; Department of Medicine, Columbia Presbyterian Medical Center, New York, N. Y. Requests for reprints may be addressed to Dr. J. E. Sealey, Cardiovascular Center, New York Haspital-Cornell Medical Center, 525 East 68th Street, New York, N. Y. 10021. © 1975 by Grune & Stratton, Inc. Seminars in Nuclear Medicine, Vol. 5, No. 2 (April), 1975

189

190

SEALEY AND LARAGH

group to be present in plasma in two forms, a decapeptide and an octapeptide. Ij012Skeggs and his colleagues in 1956~3 isolated from horse plasma a converting enzyme that splits the decapeptide angiotensin I to angiotensin II. These workers were able to identify angiotensin II and the dipeptide histidyl-leucine as products of reaction. Measurement of the components of the renin system was not a simple task. Circulating levels of angiotensin are so minute that, in the early assays, excessive amounts of blood had to be collected, t4015 However, in 1963 Helmer and Judson 16 developed a bioassay method for measuring plasma renin which utilized the fact that renin reacts with plasma renin substrate to generate angiotensin. By incubating the plasma for a fixed period of time, prior to bioassay, sufficient angiotensin was generated to make detection and quantitation in a small volume of plasma possible. Most methods which have been developed for measurement of plasma renin are based on this principle and differ mainly in techniques used for inactivati6n of angiotensinases which destroy the angiotensin formed during incubation. Techniques for angiotensinase inhibition include chemical separation from renin, ~7 adsorption of the formed angiotensin onto a resin, 15 dialysis, ~8-2° inactivation by heat at pH 4.5, 20and addition of inhibitors. ~7,19a~-2s The use of plasma renin activity rather than angiotensin II measurements in assessing the physiologic role of the renin system has introduced some complications. Since the plasma concentration of renin substrate is normally rate limiting, the capacity of plasma to generate angiotensin is a function of both renin and substrate concentrations. Thus, methods for measurement of plasma renin may differ in whether they measure plasma renin concentration or plasma renin activity. In the first approach, a fixed amount of renin substrate (angiotensinogen) is added to the incubation medium so that the measurement reflects only differences in enzyme concentration. ~7.2°'24,29In the second approach, endogenous plasma angiotensinogen acts as the substrate in the incubation step, and the measurement reflects the net capacity of plasma to generate angiotensin. ~5'ts.2°-23,~s's Thus, the measurement is a reflection of a first-order reaction between the enzyme and its substrate and for this reason is referred to as plasma renin activity, and not plasma renin concentration. Most commercial kits for renin assays measure plasma renin activity. Plasma renin substrate can be measured by its capacity to yield angiotensin I during incubation in the presence of excess human renin, tga2 In addition, the capacity of plasma to generate angiotensin can be evaluated by measuring the response of plasma to the addition of a fixed amount of endogenous human renin. ~s,22 This latter measurement, which we sometimes refer to as substrate reactivity, allows analysis of the effects of activators and inhibitors and of differences in renin substrate concentration on the renin reaction. At first, a rat bioassay was used to quantitate the angiotensin generated during the incubation step. ~5-22.3° More recently, with the introduction of radioimmunoassay techniques by Yalow and Berson, 3~ the rat pressor bioassay has been replaced by radioimmunoassay of formed angiotensin I. 22-3°.32 In the following pages a description of the method that we have developed for measurement of plasma renin activity will be presented. ~.22,33,34 The renin

PLASMA

RENIN

ACTIVITY

191

incubation procedure will be described first. The rationale behind each of the steps will be presented, and, whenever possible, reference will be made to differences in our method from other methods, especially those available in commercial kits. Then the radioimmunoassay steps will be considered, with special reference to simplification and quality-control procedures that are in operation in our laboratory. INCUBATION

FOR GENERATION

OF ANGIOTENSIN

I

Effect of Plasma Renin Substrate: Why Dilution Should Be A voided The concentration of plasma renin substrate in normal man is rate-limiting, 22,34and, as already noted, differences in its concentration affect the rate of angiotensin generation. Data of substrate concentrations in 19 normal subjects are presented on the abscissa of Fig. 1. The range of values is quite wide, from less than 1000 to greater than 2000 ng/ml, but in individual subjects it was found to be constant from day to day and did not fluctuate with wide changes in renin activity induced by sodium depletion or sodium loading, st The rate ofangiotensin generation induced in each plasma in response to addition of a fixed amount of renin is plotted on the ordinate of Fig. 1 (substrate reactivity). It.is apparent that the rate of angiotensin generation is dependent on the substrate concentration such that, when the substrate concentration is 1000 ng/ml, almost half the amount of angiotensin is generated than when the substrate concentration is 2000 ng/ml. However, as long as the incubation step is carried out in undiluted plasma, differences in substrate concentration do not present a problem since, for clinical studies, we are interested primarily in the capacity of plasma to generate angiotensin II. Plasma renin activity measurements appear to be a meaningful indicator of this since there is no evidence that converting enzyme activity is ever rate-limiting. Accordingly, dilution of plasma during incubation should be avoided. ~ A threefold dilution is recommended by the New England Nuclear kit. If plasma RELATIONSHIP

BETWEEN

RENIN SUBSTRATE AND SUBSTRATE

REACTIVITY

19 Normal Subjects 90

•

o,

80

Fig. 1. Direct relationship between plasma renin substrafe concentration and the rate of angiotensin genera. tion in response to addition of a fixed amount of human renin (substrate reactivity) in 19 normal subjects. The normal range of substrate concentration is from 1000 to 2200

ng/ml.

22

.0~ •

oo

ro •

•

•

6o •

SUBST~=T[ REACTIVITY 50

•

ng/ml/ht

o•

°o

•

•

e a

40 •

0•6

30

•

:•

20 io o

500 tOOO RENIN $UBSTRATE .g

1500

2000

ongiotensmgenerotedper ml

i

2500

192

SEALEY AND LARAGH

renin substrate were not rate limiting, dilution of plasma would lead merely to a predictable reduction in the amount of formed angiotensin, due to renin dilution, which then could be corrected for by a simple dilution factor. However, we have shown that reduction in the concentration of renin substrate leads to a reduction in the rate of angiotensin generation. ~ Since first-order kinetics apply, 50~o dilution of one plasma may lead to a 50~o reduction in the rate of angiotensin generation, but in other plasmas with higher substrate concentrations, only a 10~o reduction may occur. No simple formula can be used to correct for the reduction in angiotensin production due to substrate dilution; therefore, dilution should be avoided altogether.

Why Use pH 5.7? Subnormal plasma renin activity has been demonstrated in a large fraction of patients with hypertension. These patients include not only those with primary aldosteronism 35,36 and other uncommon forms of mineralocorticoid exce~s,37 but also about a quarter or more of all patients with essential hypertension28-4° In these latter patients, the renin value is the only consistent feature which distinguishes them from patients with other forms of essential hypertension. Methods of the greatest sensitivity are therefore required so that truly low values can be discriminated from those falling into the normal range. We decided to carry out the incubation step at the pH optimum since this would provide the maximum rate of angiotensin generation and thus maximum sensitivity. In addition, the use of the pH optimum also provides the best conditions for angiotensinase inhibition and for stability in the rate of angiotensin generation for up to 18 hr because the pH does not fluctuate during incubation. The pH optimum for the reaction of human renin with plasma renin substrate was found to be between pH 5.5 and 6.5 (Fig. 2), 22 confirming previous reports. =6,tS'2smThe rate ofangiotensin generation fell steeply below pH 5.5 and declined less steeply above 6.5. In contrast to pH 3.7, in studies performed at pH 7.4 the rate of angiotensin generation was almost halved. Moreover, we

PLASMA

RENIN A C T I V I T Y

pH OPTIMUM Mean

6.0 5.0

pH~

Incubolion

4.0 PRA

i

ng A't/rnl/hr 3.0 2.0 1,0

; M

0

I

4.5

!

5.0

!

5.5

!

6,0

f

6.5 pH

!

7.0

!

7.5

I

8.0

!

8,5

I

9.0

Fig. 2. pH optimum of human renin with human renin substrate. A pool of plasma was incubated at different pH levels for 3 hr in the presence of EDTA, DFP, and neomycin. Arrows represent the change in pH which occurred during incubation. The pH remains quite stable during incubation below pH 6.0 but can increase by almost one pH unit during incubation at alkaline pH.34

PLASMA RENIN ACTIVITY

193

found that the pH of plasma did not remain stable during incubation. In some samples it drifted high enough to almost stop the reaction. In this study of pH 7.4, buffers were not used to adjust the pH, on the assumption that the endogenous plasma buffers would hold the pH constant. This had previously been confirmed for pH 5.7. The gradual increase in pH observed during incubation of samples adjusted to pH 7.4 may have been due to slow elimination of carbon dioxide from plasma. The lack of change in pH below 6.0 may have been caused by prior elimination of CO2 by addition of acid to adjust the pH. Thus, the use of the pH optimum for incubation precludes the need for addition of strong buffer to hold the pH constant, and it leads to a twofold increase in the rate of angiotensin generation. In addition, if perchance the pH does change slightly during incubation, the rate of generation of angiotensin will not vary, since the fiat portion of the pH optimum curve is exploited. In this context it should be emphasized that incubation of unadjusted plasma should n e v e r be carried out. ~'4t~ Many kits recommend this. Plasma collected in EDTA and stored at - 2 0 " C prior to incubation can range in pH from 7.4 to 8.5 and, as illustrated in Fig. 2, a difference in pH of one unit can cause a twofold difference in the renin assay of samples with identical renin levels. In a recent abstract, Oparil, Koerner, and Haber 45 questioned the validity of incubating at'the pH optimum since they were unable to detect renin activity in plasma from nephrectomized patients during incubation at pH 7.4, but in samples incubated at pH 5.5 renin activity ranged from 0.12 to 0.30 ng/ml/hr (mean, 0.18). These authors suggested that this discrepancy was due to acid proteases in plasma which release angiotensin I only when acid pH is used for incubation. This group also reported in the same abstract that incubation of low-renin +samples at acid pH results in variable differences in the rate of angiotensin generation when compared to incubation at pH 7.4. Thus, in their study, 48~o of plasmas exhibited higher values at pH 5.5 and 46~o had lower values. It is difficult to ascribe these differences to the action of acid proteases, since according to this thesis one would expect that low-renin samples would consistently exhibit relatively higher rates of angiotensin generation at acid pH. To test these observations, we compared the relative rates of angiotensin generation at pH 5.7 and 7.4 in ten low-, nine normal-, and eight high-renin plasmas. The ratio of renin activity at pH 5.7 relative to pH 7.4 was 1.58 4- 0.33 (SD), 1.76 4- 0.32, and 1.69 4- 0.20, respectively, and none of the samples exhibited a higher value at pH 7.4. In other studies in nephrectomized man we found that, when renin activity was detectable at pH 5.7, it was also detectable at pH 7.4. However, prolonged incubation (18 hr) was necessary to detect renin in some very low samples. Perhaps the much lower sensitivity of a method which utilizes pH 7.4 and a short incubation time explains the erratic relationship between values detected at the two pH's reported by Oparil and coworkers in their study of low-renin samples. Poor sensitivity may also explain their lack of ability to detect renin in plasmas from nephrectomized man. The presence of renin in plasma from nephrectomized patients is not surprising since renin has been demonstrated in several tissues from nephrectomized animals, and there have been several reports o f renin activity in plasma of

194

SEALEY AND LARAGH

nephrectomized man. 46.47 In fact, in a previous paper, these same authors had been able to detect renin (0.15 n g / m l / h r ) in two of five plasmas from nephrectomized patients. 26 These reported assays were incubated at unadjusted pH, presumably greater than pH 7.4. Altogether, it is still our view that incubation at the pH optimum provides the most reliable and most sensitive conditions for generation o f angiotensin, and, as outlined in the next section, it also provides optimum conditions for inhibition of angiotensinases.

Inhibition of A ngiotensinases and Converting Enzyme The effectiveness o f E D T A as both a converting enzyme and angiotensinase inhibitor is well documented. 13,Is D F P (diisopropylfluorophosphate) is an effective adjunct to E D T A as an angiotensinase inhibitor, especially at acid pH. ~s BAL (dimercaprol) and 8-hydroxyquinoline have also been recommended for use as protective agents during incubation at pH 7.4. 26,30,48 However, in oui" studies, they were found to be less effective than D F P both at pH 5.7 and 7.4 (Table 1), u although other studies have questioned this. 49 In our studies, during 3 hr incubation at pH 5.7, a mean of 12.9 ng/ml angiotensin I were generated in the presence of BAL and 8-hydroxyquinoline, compared to 23.6 ng/ml using D F P as an angiotensinase inhibitor. During 18 hr incubation at pH 7.4, a mean value o f 3.0 ng/ml was found using BAL and 8-hydroxyquinoline, compared with 5.9 ng/ml in the presence o f DFP. However, we showed that neither o f these sets of inhibitors could effectively inhibit angiotensinase activity for 18 hr at pH 7.4. In contrast, at pH 5.7 in the presence of EDTA, DFP, and neomycin, linearity in the rate o f angiotensin generation can be demonstrated for up to 18 hr (Fig. 3). Thus, a long incubation time can only be carried out at acid pH, and we have shown that, for low-renin samples, a more prolonged incubation is necessary to discriminate them accurately from normal renin samples, u A recent report by Kodish and Katz 28 has drawn attention to the effectiveness of P M S F (paramethylsulfonyl fluoride) as an angiotensinase i n h i b i t o r y This substance appears to be an effective and less toxic substitute for DFP. We have tested the effectiveness of P M S F during 18-hr incubations and have demonstrated linearity in the rate o f angiotensin generation. Table 1. Comparison of the Effect of DFP or BAL and 8-hydroxyquinoline on Incubation for Generation of Angiotensin I pH 5.7

pH 7.4

BAL etc.

None

DFP

BAL etc.

None

3-hr incubation: normal and hlgh-renln samples (n = 10) Angiotensln I* (ng/ml) 23.6 12.9 (6.8) (3.6) % AI found at pH 5.7 with DFP 55

11.4 (2.9) 48

10.9 (3.0) 46

10.2 (3.0) 43

10.3 (2.7) 44

18-hr incubation: low-renln samples (n = 5) Angiotensin I* (ng/ml) 15.8 (2.1) % AI found at pH 5.7 with DFP

1.3 (0.6) 8

5.9 (1.4) 37

3.0 (0.5) 19

2.6 (0.7) 17

DFP

2.8 (0.9) 18

*Anglotensln I generated during incubation. The 3-h~"incubation was carried out using different sampTes from those used in the 18-hr incubation. Figures in parenthesis represent the standard error of the mean.

PLASMA RENIN ACTIVITY

195 COMPARISON OF 3 AND 18 HOUR INCUBATIONS SAMPLES WITH LOW PLASMA RENIN ACTIVITY

1.8 1.6

I • •r1.04 r ( N"3z )

(J~

O~, q

1.4 1,2 pRA i~g AZ/ml/hr 1.0 /BHRIHCMBATIO,V

0.8 o°

•

0,6 0.4

Fig. 3. No difference was found in the hourly rate of angiotensin generation when samples were incubated for 3 or 18 hr at pH 5.7 in the presence of EDTA, DFP, and neomycin. ~4

O,Z 0

g~ o'., o!~ ot~ ,'.o ,!,

PRA ng AZ/ml/ht

,'., ,I~ :.~

3HR INCUBATION

Most of the kits recommend the use of EDTA together with BAL and 8hydroxyquinoline as inhibitors of angiotensinases and converting enzyme. This approach may be effective during short incubation times at alkaline pH. However, using these inhibitors, it is not possible to demonstrate linearity in angiotensin production for more than 3 hr. Under the conditions recommended by most kits, the amounts of angiotensin generated in samples with low plasma renin activity are below the lower limit of detectibility of the kit radioimmunoassay (Fig. 4). Therefore, utilizing the incubation conditions recommended by the kits, they are only useful for measuring samples with normal or high plasma renin activity. ESTIMATION OF PLASMA RENIN ACTIVITY USING 3 COMMERCIAL KITS I PRA SAMPLES 30

•H/gh • Normal

•

A Low

20

+

+

PLASMA RENIN ACTIVITY 1,0

nglAT/ml/hr &

0.1

of

Below Lower Limil Oetecl3bility

SEALEY ET AL

NEW ENGLAND NUCLEAR

SCHWARZ /MANN

SQUIBB

Fig. 4. Plasma renin assays in normal, low-, and high-renin samples. Comparison of our own method with those recommended by three commercial kits. Most values for Iow-renin samples were undetectable using the kits. The absolute amounts of angietensin generated in Iow-renin scimples using an 18-hr incubation (Sealey) were almost identical to the yield in normal renin samples incubated for 3 hr. Thus, by using a prolonged incubation time, normal and Iow-renin samples can be evaluated with equal accuracy.

SEALEY AND LARAGH

196 Table 2. Incubation for Generation of Angiotentin I Combine

2 ml plasma containing 0.003 M EDTA* f"~25/a110% neomycin sulfate

1 2 drops (40 pl) DFP (1/20 dilution in isopropyl alcahol)t Adjust pH to 5.7~ with 1 N, 0.5 N, or 0.1 N HCI. Divide into three portions. Incubate two portions for 3 hr and one for 18 hr at 37°C. Freeze. Assay a 3-hr aliquot. If renin < 1.0 ng/ml/hr, reassay 18-hr sample. Twenty or mare samples can be incubated as one batch. *Blood collected into potassium EDTA vacutainers. "J'ToxlclWear disposable gloves (or add 25 .ul of 5% PMSF in ethanol). CAcceptable range, 5.6--5.7.

Summary of Proceduresfor Incubation of Renin with Its Substrate (Table 2) Since renin is measured by its capacity to form angiotensin I during a fixed incubation time, generation of angiotensin I prior to incubation must be inhibited. For this reason, the plasma should be chilled immediately after collection. In addition, conversion of angiotensin I to angiotensin II or to breakdown products during the incubation must be inhibited. Three factors operate to destroy angiotensin I: converting enzyme, angiotensinases, and bacteria. These must all be inhibited during the incubation step. EDTA is added to plasma during collection of the blood. In addition, the concentration used as anticoagulant (0.003 M) also inhibits converting enzyme and, to some extent, angiotensinases.49 DFP or PMSF is added to the plasma to inhibit remaining angiotensinases, and addition of neomycin sulfate helps to retard bacterial growth.22 Prior to incubation, the plasma is adjusted to the pH optimum 5.7 with hydrochloric acid. The sample is then divided into three aliquots, two of which are incubated for 3 hr, and the third for 18 hr (overnight). If the samples are not chilled immediately after collection and continuously thereafter, a 4* C incubation 34 should also be included. The sample incubated for 3 hr is assayed first. If the renin value is found to be less than 1 ng/ml/hr, the sample incubated for 18 hr is then assayed and this result is reported. 33 In summary, using only minor modifications to the components provided by the kits, it is possible to incubate plasma so that low-renin samples can be discriminated accurately from those with normal plasma renin activity. It should be stressed, however, that samples should not be stored for more than 2 months prior to measuring renin since we have recently demonstrated that plasma renin activity increases during storage, perhaps because renin is split from an inactive "prorenin" during storage at -20* C. 5° This presents a problem, especially in low-renin samples, since some samples from low-renin patients, collected during sodium deprivation, have abnormally large amounts of "prorenin" that converts to active renin during storage. ~° RADIOIMMUNOASSAY

OF A N G I O T E N S I N

I

In this section we will describe some details of our method and accent possible pitfalls to keep in mind with regard to the radioimmunoassay step) ~ In

PLASMA RENIN ACTIVITY

197

general, we found that the radioimmunoassay components provided by kits were quite good, but no systematic study was carried out. Tris buffer is widely used in the radioimmunoassay step. Protein is added to the buffer to inhibit adsorption of angiotensin onto glassware and plastic containers, and to increase the stability of the diluted antibody. Considerable adsorption to plastic has been observed when the protein concentration is reduced below 0.4~. We also add neomycin and phenylmercuric acetate to retard bacterial growth. 33 A unique feature of our radioimmunoassay is to mix the radioactive angiotensin with antibody less than I hr prior to use. This step greatly simplifies the radioimmunoassay and reduces inaccuracies caused by multiple pipettings. In our assay, mI-angiotensin, (approximately 250,000 cpm) is added to 200 ml oftris buffer and mixed thoroughly. Ten milliliters of mixture is removed prior to addition of antibody for measurement of nonspecific binding of angiotensin to proteins in the buffer solution. Prior to radioimmunoassay, we dilute the plasma samples in tris buffer if the renin value is expected to be above the normal range. However, for most samples no predilution is necessary. Since angiotensin I loses activity when stored in dilute solution, we prepare the angiotensin I standards daily from a stock solution of 10 t~g/ml. One-tenthmolar tris buffer containing 0.6~ protein and bacteriostatie agents is used as diluent. Each month, the 10 ~g/ml stock solution is prepared from a 50 t~g/ml solution which is stored at - 2 0 " C in boiled distilled water containing 0.2~ neomycin sulfate. The 50 t~g/ml solution is stored in siliconized glass test tubes, the others in plastic. Some kits provide the angiotensin I standard in dilute solution. When we checked commercial kits, the diluted standard provided by Squibb was identical t o both our own and an undiluted standard in the New England Nuclear kit. However, the diluted standard provided by Schwartz-Mann appeared to have lost activity. 34 We include ten tubes in duplicate in the standard curve, i.e., seven pairs of tubes with from 10 to 400 pg angiotensin I, as well as tubes in duplicate for measuring the total counts, the nonspecifie binding of angiotensin to proteins in the buffer solution, and the per cent binding of radioactive angiotensin when no added cold angiotensin I is present (Bo). Ten and twenty microliters of approximately 20 incubated plasmas and of two standard plasma samples are assayed with each set. The standard plasma samples are included for quality control purposes and are prepared by adding a known amount of angiotensin to charcoal-extracted plasma. A pool of plasma is mixed with 1~o charcoal which is then rigorously removed by filtration and centrifugation. The plasma is diluted twofold with tris buffer, and 1/100 volume of 10~o EDTA is added. The pool is divided into two portions to which 3 and 15 ng/ml of angiotensin are added. This provides 30, 60, 150, and 300 pg angiotensin in 10- and 20-t~l aliquots, thus spanning the range of the standard curve. Angiotensin bound to antibody is separated from free angiotensin by adsorption of the free onto dextran-coated charcoal. Because of the tendency of char-

198

SEALEY AND LARAGH

coal to slowly adsorb a small amount of bound angiotensin, the duplicates of the standard curve are divided so that charcoal is added to one-half at the beginning and to the other half at the end of the series. Blank Subtraction

The ability to prolong the incubation time can eliminate a problem inherent to radioimmunoassay. Most antibodies to angiotensin I cross-react to some extent with nonspecific substances in plasma, 52 and this interference can result in falsely high values unless a blank is subtracted. Also, a small amount of angiotensin I always circulates in plasma. By prolonging the incubation time, it is possible to generate so much angiotensin that the blank can be ignored. 33 This has the additional advantage of reducing the number of samples for assay by half. However, it should be stressed that the blank can only be ignored if the samples are chilled continuously prior to assay, are not stored for more than 2 months, and samples with PRA less than 1 ng/ml/hr are incubated for 18 hr. We have recently observed a characteristic of the blank that brings into question the validity of blank subtraction when using antibodies that have a high nonspecific cross-reactivity. In Fig. 5, lower plot, mean data are presented from five unincubated samples assayed using four different antibodies. 66.6 pg of angiotensin were added to these unincubated plasmas, and the samples were assayed again (upper plot). The mean blank ranged from 10 to 34 pg per 20 pl plasma depending on which antibody was used. The amount of angiotensin detected after addition of 66.6 pg was 80 pg with the antibody with the lowest blank (No. 1) and 83 pg using antibody No. 4, which had the highest blank. Thus, when the blank was subtracted, the recovery of angiotensin I using antibody No. 1 was within the expected range, but the recovery using antibody BLANK SUBTRACTION .....

IS IT VALID.~ Ioo

Unlncubo/ed Somple Plus 6 6 . 6 ~ A Z

80

70 60

~

!

g

~

p9 ANGLO 1" 50 IN 20Jul PLASMA

e

n

sin

Recovery

(A-a)

40

Unlncuboted Somple

30

(Blonk)

(a;

20 I0

:

~,,,o

~

./

...,,o.,~~~ ''(J'' o,," ° ' '

I

NO: I

I

I

2 3 ANTIBODY SOURCE

I

4

Fig. 5. Mean angiotensin I assays are presented far Rye unlncubated plasmas (B) using four different antibodies. (A) represents repeat assays after addition of 66.6 pg angiotensin I, and (A-B) is the calculated recovery. Using antibodies Nes. 1-3, the blank is additive and, when subtracted, results in a good recovery of angiotensin h However, using antibody No. 4, the added angiotensin I appears to displace the blank and, in this case, blank subtraction leads to a falsely law value. Therefore, antibodies should be characterized if blank subtraction is used, and those with the characteristics of No. 4 should be discarded.

PLASMA RENIN ACTIVITY

199

No. 4 was low (75~o). In contrast, after blank subtraction using antibody No. 3 the recovery was 107~o (i.e., within the expected range) despite the fact that the blank with this antibody was also quite high (21 pg). Thus, blank subtraction may sometimes be invalid (antibody No. 4). However, as with antibody No. 3, the nonspecific blank may sometimes be additive, and, in this circumstance, blank subtraction is valid. This observation reemphasizes the importance of prolonging the incubation time.until the blank, whether due to angiotensin I or t o a nonspecific substance, can be ignored. It also emphasizes the importance of working with a characterized source of antibodies.

Volume of Plasma Added to Radioiramunoassay Many methods and most commercial kits recommend addition of 50 ~1 or more of plasma to 1 ml of radioimmunoassay mixture.26'4~'42 In our method we assay a maximum volume of 20 ~1. This small volume was chosen because, when it was increased to 50 tzl, a lower value for plasma renin activity was obtained. Thus, mean values for plasma renin activity for ten samples fell from 0.68 ng/ml/hr (10-/~1 aliquot) to 0.50 ng/ml/hr (50-t~l aliquot) (p < 0.01). 34

Troubleshooting Radioimmunoassay involves many steps, and when variability occurs it can be ascribed to many different problems. A useful guide to changes in the radioimmunoassay is the nonspecific binding. If the duplicate counts of these tubes are variable, this may be due to erratic addition of charcoal. The addition of charcoal for separation of bound from free is the source of many errors. The charcoal should be stirred vigorously, and every effort should be made to add i t as quickly and consistently as possible. If the nonspecifie binding gradually and consistently increases with each set (it should be 2yo-3~o of the total counts), this is a sign that the radioactive angiotensin is deteriorating. Also, when changing to a new batch of charcoal, the nonspecific binding may increase or fall. Different batches of charcoal may have different activity, and the optimum amount should be calculated for each batch. If too much is added, the adsorption of antibody-bound angiotensin will increase to unacceptable levels, and if too little is used erratic results will be achieved. Another guide to the adequacy of the label is the variability of the standards at the extremes of the standard curve. As the label ages, the sensitivity of the standard curve decreases, and the lowest point often fails to fall on the straight line of a logit plot. 33 Also, if the amount of antibody used is too much or too little, the points at the extremes of the standard curve will not fall on the straight line. Unknown plasma samples for which the 10- and/or 20-t~l aliquots fall outside the range of the standard curve should be repeated. As indicated above, the standards used in the radioimmunoassay are prepared fresh daily. This can lead to pipetting errors, and because of this it is especially important that the two standard plasma samples are run with each set. If these do not fall within a predetermined range, the set should be discarded and the assays repeated.

SEALEY AND LARAGH

200

Table 3. Some Problems with the Procedures Recommended by Commercial Kits Questionable Recommendations

Problem

Unadjusted pH Dilution of plasma I- or 3-hr incubation

PRA differences can be due to differences in pH, not renin. No simple correction factor for dilution of renin substrate possible. Insufiqc;ent anglotensin generated in Iow-renin samples to avoid blank subtraction. Leads to underestimation of anglotensin I. Leads to inaccuracies. (1) Cannot incubate more than 3 hr because anglotensinases not completely inhibited. (2) Need strong buffers to hold pH constant. Renin, renin substrate, and angiotensin I destruction.

100/~1 plasma in RIA Blank subtraction pH 7.4

No bocteriostatic agents

DISCUSSION

The method for measuring plasma renin activity can be considered to comprise two entirely separate steps: (1) incubation to generate angiotensin I and. (2) radioimmunoassay of the formed angiotensin. Many kits are available for measurement of plasma renin activity, and the major problem with these commercial kits is in the recommendations for carrying out the preliminary incubation step. In general, the kit standards, ~2SI-angiotensin and antibodies, appear to be quite good, and, since these are the major components of the kit, it is quite simple to modify the preliminary incubation step by purchasing readily available reagents. The reasons for modifying the incubation step are presented in Table 3. All of these problems can be overcome by following the incubation directions outlined in Table 2. Accordingly, if commercial kits are used, we recommend the following: (1) The incubation step should be carried out at pH 5.7 in the presence of DFP (or PMSF) and neomycin, as well as EDTA. (2) For samples with low plasma renin activity (less than 1 ng/ml/hr) an aliquot incubated for 18 hr should be assayed. (3) Dilution of plasma should be minimal during pH adjustment. (4) Angiotensin I generation should be prolonged until the necessity for blank subtraction is eliminated. (5) A maximum volume of 20 #1 plasma should be assayed. (6) The same source of antibody should be used consistently so that variability in antibody characteristics does not become a problem. (7) For clinical laboratories which have little control over the temperature of the blood after collection (it should be chilled), assay of the blank should not be discontinued since high blanks will expose those samples which have not been kept chilled. (8) Strict quality control is essential in any radioimmunoassay. The same two standard plasma samples should be run with each set, and if they vary by more than ± 15~, the set should be discarded and then repeated. REFERENCES

1. Tigerstedt R, Bergman PG: Niere und Kreislauf. Skand Arch Physiol 8:223-271, 1898 2. Pickering GW, Prinzmetal M: Some observations on renin, presser substance contained in normal kidney, together with method for its biological assay. Clin Sci 3:211-227, 1938

3. Landis EM, Montgomery H, Sparkman D: Effects of presser drugs and of saline kidney extracts on blood pressure and skin temperature. J Clin Invest 17:189-206, 1938 4. Hessel G: Uber Renin. Kiln Wochenschr 17:843-850, 1938

PLASMA RENIN ACTIVITY

5. Volhard F, Fahr T: Die brightsche Nierenkranheit: Klinik, Pathologie und Atlas. Berlin, Springer, 1914 6. Goldblatt H J, Lynch RF, Hanzal RF et al.: Studies on experimental hypertension. Production of persistent elevation of systolic blood pressure by means of renal ischemia. J Exp Med 59:347-379, 1934 7. Page IH: On the nature of the pressor action of renin. J Exp Med 70:521-542, 1939 8. Braun-Menendez E, Fasciolo JC, Leloir LF ct al.: La substancia hipertensora de la sangre del rinon isquemiado. Rev Soc Argent Biol 15:420-430, 1939 9. Page IH, Helmet OM, Plentl AA et al.: Suggested change in the designation of "renin activator" (hypertensinogen) to "renin substrate" (a2 globulin). Science 98:153-155, 1943 10. Braun-Menendez E: Pharmacology of renin and angiotensin. Pharmacol Rev 8:25-32, 1956 11. Skeggs LT Jr, Marsh WH, Kahn JR et al.: The existence of two forms of hypertensin. J Exp Med 99:275-282, 1954 12. Elliott DF, Peart WS: Amino acid sequence in a hypertensin. Biochem J 65(No 2): 246-254, 1957 13. Skeggs LT Jr, Kahn JR, Shumway NP: The preparation and function of the hypertensin converting enzyme. J Exp Med 103:295305, 1956 14. Paladini AC, Braun-Menendez E, Del Frade IS et al.: The estimation of angiotensin in blood. J Lab Clin Med 53:264-271, 1959 15. Boueher R, Veyrat R, DeChamplain J et al.: New procedures for measurement of h u m a n p l a s m a angiotensin and renin activity levels. Can Med Assoc J 90:194-201, 1964 16. Helmer OM, Judson WE: The quantitative determination of renin in the plasma of patients with arterial hypertension. Circ Res 27:1050-1060, 1963 17. Lever AF, Robertson JIS, Tree M: The estimation of renin in plasma by an enzyme kinetic technique. Biochem J 91:346-372, 1964 18. Pickens PT, Bumpus F, Merlin F et al.: Measurement of renin activity in human plasma. Circ Res 17:438-448, 1965 19. Gould AB, Skeggs LT, Kahn JR: The presence of renin activity in blood vessel wails. J Exp Med 119:389-399, 1964 20. Skinner SL: Improved assay methods for renin concentration and activity in human plasma. Methods using selective denaturation of renin substrate. Circ Res 20:391-402, 1967 21. Newton MA, Laragh JH: Effect of

201

corticotropin on aldosterone excretion and plasma renin in normal subjects, in essential hypertension and in primary aldosteronism. J Clin Endocrinol Metab 28:1006-1022, 1968 22. Sealey JE, Gerten-Banes J, Laragh JH: The renin system: Variations in man measured by radioimmunoassay or bioassay. Kidney lnt 1:240-253, 1972 23. Boyd GW, Adamson AR, Fitz AE et al.: Radioimmunoassay determination of plasma. renin activity. Lancet 1:213-218, 1969 24. Stockigt JR, Collins RD, Biglieri E: Determination of plasma renin concentration by angiotensin I immunoassay. Circ Res 28, 29 (suppl II): 175-189, 1971 25. Cohen EL, Grim CE, Conn JW et al.: Accurate and rapid measurement of plasma renin activity by radioimmunoassay. Results in normal and hypertensive people. J Lab Clin Med 77:1025-1038, 1971 26. Haber E, Koerner T, Page LB et al.: Application of a radioimmunoassay for angiotensin 1 to the physiologic measurements of plasma renin activity in normal human subjects. J Clin Endocrinol Metab 29:1349-1355, 1969 27. Goodfriend TL: Radioimmunoassay of angiotensins and renin activity, in Berson SA, Yalow RS (eds): Methods in Investigative and Diagnostic Endocrinology, vol. 2B. Amsterdam, North Holland-American Elsevier, 1973, pp 1158-1168 28. Kodish ME, Katz FH: Plasma renin concentration: comparison of angiotensinase inhibitors and correlation with plasma renin activity and aldosterone. J Lab Clin Med 83: 705-715, 1974 29. Giese J, Jorgensen M, Nielson MD et al.: Plasma renin concentration measured by use of radioimmunoassay for angiotensin I. Scand J Clin Lab Invest 26:355-367, 1970 30. Chervu LR, Lory M, Liang T et al." Determination of plasma renin activity by radioimmunoassay: comparison of results from two commercial kits. J Nucl Med 13:806-810, 1972 31. Yalow RS, Berson SA: Immunoassay of endogenous plasma insulin in man. J Ciin Invest 39:1157, 1960 32. Poulsen K: Kinetics of the renin system. The basis for determination of the different components of the system. Scand J Clin Lab Invest 31 (suppl 132): 1-86, 1973 33. Sealey JE, Laragh JH, Gerten-Banes J et al.: The measurement of plasma renin activity in man, in Laragh JH (ed): Hypertension

202

Manual. New York, Yorke Medical Books, 1974, pp 621-640 34. Sealey JE, Laragh JH: Searching out low renin patients: Limitations of some commonly used methods. Am J Med 55:303-314, 1973 35. Conn JW: Plasma renin activity in primary aldosteronism. Importance in differential diagnosis and in research of essential hypertension. JAMA 190:222-225, 1964 36. Baer L, Sommers SC, Krakoff LR et al.: Pseudo-primary aldosteronism: an entity distinct from true primary aldosteronism. Cite Res 26, 27 (suppl I): 203-216, 1970 37. Melby JC, Dale SL, Wilson TE: 18hydroxy-deoxycorticosterone in human hypertension. Circ Res 28, 29 (suppl 11):143-152, 1971 38. Laragh JH, Baer L, Brunner HR et al." Renin, angiotensin, and aldosterone system in pathogenesis and management of hypertensive vascular disease. Am J Med 52:633-652, 1972 39. Brunner HR, Laragh JH, Baer L e t al.: Essential hypertension: renin and aldosterone, heart attack and stroke. N Engl J Med 286:441449, 1972 40. Brunner HR, Sealey JE, Laragh JH: Renin subgroups in essential hypertension: further analysis of their pathophysiological and eipdemiologieal characteristics. Cire Res 32, 33 (suppl 1): 99-115, 1973 41. McDonald JM, Fischer GA: pH dependence of the renin reaction: a modification of commercial renin activity radioimmunoassay kits. Am J Clin Pharmacol 59:858-868, 1973 42. Viol GW, Keane PM, Speed JF et al.: Assay of plasma renin activity using commer-

SEALEY AND LARAGH

cially available reagents. Clin Biochem 5:251265, 1972 43. Osmond DH, McFadzean PA, Scaiff KD: Renin activity by radioimmunoassay: the plasma incubation step. Clin Biochem 7:52-63, 1974 44. Fleming LW, Stewart WK, Hutchinson F: Incubation conditions during the radioimmunoassay of plasma renin activity. Clin Chim Acta 50:215-222, 1974 45. Oparil S, Koerner T, Haber E: Modifications in the radioimmunoassay for plasma renin activity: implications for low renin state. Clin Res 22:345A, 1974 46. McKenzie JK, Montgomerie JZ: Reninlike activity in the plasma of anephric men. Nature 223:1156, 1969 47. Capelli JP, Wesson LG, Aponte GE et al.: Characterization and source of a rehinlike enzyme in anephric human. J Clin Endocrinol Metab 28:221-230, 1968 48. Fleming LW, Hutchinson F, Stewart WK: Radioimmunoassay of plasma-renin activity. Lancet 1:1392-1393, 1973 49. Oparil S, Koerner T, Haber E: Effects of pH and enzyme inhibitors on apparent generation of angiotensin I in human plasma. Circulation 49, 50(suppl 111):32, 1974 50. Sealey JE, Laragh JH: Prorenin in human plasma? Methodological and physiological implications. Circ Res (suppl) {in press) 51. Yalow RS: Radioimmunoassay: practices and pitfalls. Cite Res 32(suppl I):116-128, 1973 52. Page LB, Dessaulles E, Lagg S e t al.: Interference with immunoassays of angiotensin I and II by proteins in human plasma. Clin Chim Acta 34:55-60, 1971