Clinical and Experimental Pharmacology & Physiology (1979) 6,611-624.

PROPERTIES OF INACTIVE RENIN IN HUMAN PLASMA R. P. Day* and B. J. Morris? Department of Physiology. University of California, San Francisco, California, U.S.A.

(Receiced 24 Decemher 1978; revision received 16 March 1979)

SUMMARY

1. ‘Inactive’ renin in human plasma can be revealed by pH 3.3- or coldmediated activation and in normal plasma represented 76% of the ‘total’ renin. 2. Pregnancy plasma contained considerably more ‘inactive’ renin and consisted of 932, of ‘total’ renin. 3. ‘Active’ renin in normal plasma had an apparent molecular weight of 43,000 compared with 60,000 for ‘active’ renin in pregnancy plasma by gel filtration. 4. ‘Inactive’ renin in pregnancy plasma also had an apparent molecular weight of 60,000, while in normal plasma there were two peaks of inactive renin at 62,000 and 46,000. 5. After affinity chromatography of a protein preparation from pregnancy plasma on Concanavalin A-Sepharose activation by pH 3.3 could no longer be produced, suggesting that the activating factor had been removed, as would occur if it were not a glycoprotein. When pepsinogen, in a concentration similar to that found in plasma, was added prior to dialysis to pH 3.3 activation was restored. 6. Ion-exchange chromatography demonstrated that at pH 8.4 ‘inactive’ renin bore slightly less negative charges than ‘active’ renin. 7. ‘Inactive’ renin in human plasma therefore appears to be a larger molecular weight species than the ‘active’ renin in normal plasma and is capable of activation during treatment to pH 3.3 or cold with no apparent alteration in size. The results suggest an important role of pepsin (after conversion from pepsinogen) in the activation of ‘inactive’ renin during dialysis at pH 3.3. Key words: pepsinogen, pepsin, gel filtration, lectin affinity chromatography, ionexchange chromatography, pregnancy. Present address: Department of Internal Medicine, RG-20, University of Washington, Seattle. Washington, USA. t Present address: Department of Physiology, The University of Sydney, Sydney, New South Wales. Australia. Correspondence: Dr Brian J. Morris, Department of Physiology, The University of Sydney, DF13. Sydney, N.S.W.2006, Australia. 0305-1870/79/1100-0611$02.00 0 1979 Blackwell Scientific Publications

61 1

612

R. P . Day and B. J . Morris INTRODUCTION

An ‘inactive’ form of renin has been observed in human plasma and amniotic fluid (Lumbers, 1971; Morris & Lumbers, 1972; Gibson, 1973; Osmond, Ross & Scaiff, 1973; Day & Luetscher, 1974, 1975; Day, Luetscher & Gonzales, 1975; Skinner el al., 1975a; Sealey & Laragh, 1975; Sealey et al., 1976; Derkx et al., 1976, 1978; Boyd, 1977; Cooper, Murray & Osmond, 1977; Leckie et al., 1977; Weinberger, Aoi & Grim, 1977) and in kidney (for a recent review see Reid, Morris & Ganong, 1978). ‘Inactive’ renin in plasma can be activated by treatment to pH 3.3 (Lumbers, 1971; Day & Luetscher, 1974; Skinner et al., 1975a; Derkx et al., 1976; Leckie et al., 1977; Weinberger et al., 1977), to cold (Osmond et al., 1973; Sealey & Laragh, 1975) and with pepsin and trypsin (Day & Luetscher, 1975; Cooper ef al., 1977). The molecular nature of ‘inactive’ renin in plasma has not been investigated thoroughly. Such studies are limited owing to the presence in plasma of relatively low concentrations of renin compared with other sources. Studies of the plasma of patients with the rare renin-secreting, Wilms’ tumour by Day & Luetscher (1975) showed that instead of ordinary renin, of apparent molecular weight 43,000, all of the renin was of apparent molecular weight 63,000 and that this ‘big’ renin could be activated by pH 3.3-, pepsin- or trypsin-treatment with no noticeable change in size. Kinetic studies excluded increased enzyme-substrate affinity as being responsible for the activation. ‘Big’ renin in human plasma appears to be similar to the renin in human amniotic fluid, which has an apparent molecular weight of 58,000 and can be activated by pH 3.3 or by proteases with no apparent change in size, despite extensive degradation of other proteins by pepsin (Morris & Lumbers, 1972; Morris, 1978a,b). ‘Big’ renin is also present in mouse plasma, has an apparent molecular weight of 70,000 and can be activated by pH 3.3 or by pepsin at pH 4.5 with no detectable change in molecular weight (Nielsen, Malling & Poulsen, 1978). In contrast to these findings, Boyd (1977) and Shulkes, Gibson & Skinner (l978), were unable to find ‘big’ (ca. 60,000 molecular weight) renin in normal human plasma and amniotic fluid. The ‘active’ renin present in plasma had an apparent molecular weight of 41,000-46,000 and treatment of plasma with pepsin at pH 3.3 resulted in new renin activity havingan apparent molecular weight in this range. To add to the confusion, Hseuh et al. (1978a,b) have found that both ‘big’ and normal sized plasma renin can be activated and that the proportion of each of these is dependent on the Na+ status of the individual. Forms of renin of molecular weight ca. 60,000 have also been observed in kidney and, in contrast to observations with plasma and amniotic fluid, some investigators have observed a reduction in molecular weight to ca. 40,000 upon activation (reviewed by Reid et al., 1978). In any case, there is a clear enzymatic and immunological relationship between all of these forms of renin (Morris, 1978b; Nielsen et al., 1978; Shulkes et al., 1978) and the present study set out to investigate further the molecular properties of ‘inactive’ renin in human plasma. A readily available source of ‘big’ plasma renin was discovered, namely pregnancy plasma, and was utilized in these studies. METHODS Materials Samples of blood were collected from normal male and female laboratory personnel on an unrestricted Na+ intake and from primigravidae of 26-30 weeks gestation. The

Properties of inactive renin in human plasma

613

collection was into 1/10 volume of 0.3 mol/l Na2 EDTA. Plasma was separated by centrifugation at 1000 g for 20 min at 4°C. Hog pepsinogen was Sigma Grade I and was stored at 4°C.

Renin assay Plasma renin activity (PRA) (Skinner, 1967) was measured by the rate of formation of angiotensin I (AI) from endogenous angiotensinogen during a 3 h incubation of plasma at 37°C with 10 mmol/l EDTA, 2 mmol/l phenylmethylsulphonyl fluoride (PMSF) and 5 mmol/l N-ethyl maleimide (NEM), pH 7.4; EDTA, PMSF and NEM were used to inhibit the activity of enzymes that degrade AI. After incubation samples were diluted with two volumes of 10 mmol/l citric acid, placed in boiling water for 5 min, centrifuged at 1000 g for 15 min, and A1 in the supernatant measured by an established radioimmunoassay technique (Stockigt, Collins & Biglieri, 1971). Plasma renin concentration (PRC) (Skinner, 1967) was measured by the rate of formation of A1 during incubation of one part sample to five parts nephrectomized sheep plasma prepared by the method of Skinner (1967) and containing 2200 pmol angiotensinogen/ml. This concentration of angiotensinogen gave zero-order reaction conditions with respect to substrate. The PRC assay used the same concentrations of angiotensinase inhibitors and protocol as for the PRA assay. Except where otherwise indicated all renin measurements were made using the PRC method of assay. PRC before activation of ‘inactive’ renin was termed ‘active PRC’, and after activation was termed ‘total PRC’ (Skinner et al., 1975a). ‘Inactive PRC’ =‘total PRC’ -‘active PRC’. Angiotensinogen assay Angiotensinogen concentration was measured from the number of moles of A1 that could be hydrolysed completely from angiotensinogen during incubation with human renin (0.01 Goldblatt Units/ml) at 37”C, pH 7.4 (Morris & Reid, 1978), since one mole of angiotensinogen yields one mole of AI. Detection of protein The distributions of protein eluting from columns was determined by measuring the absorbance of the fractions at 280 nm with a Gilford spectrophotometer, Model 2400-S. The method of Lowry et al. (1951) was used to quantify protein, with bovine serum albumin (Cohn Fraction V) as standard. Activation procedures Various procedures were used to activate ‘inactive’ renin. These were (1) dialysis of sample to pH 3.3 for 24 h at 4”C, followed by dialysis to pH 7.4 overnight, (2) incubation of samples at -4°C for 72 h and (3) dialysis to pH 3.3 as above, but with the addition of pepsinogen, 100 ng/ml. Buffers of pH 3.3 and pH 7.4 were prepared as described by Skinner (1967). GelJiltrat ion Samples of 0.5 ml of plasma were applied to a column of Sephadex G-100 (Pharmacia) (1.6 x 80 cm) equilibrated with 50 mmol/l sodium phosphate buffer, pH 7.4, 4°C. Fractions of 2 ml were collected with a Gilson Micro Fractionator. The flow rate of the column

614

R. P . Duy and B. J . Morris

was 0.1 ml/min. Void volume was measured with blue dextran and the column was calibrated with aldolase (molecular weight I58,000), human serum albumin (molecular weight 67,000), ovalbumin (molecular weight 45,000), and chymotrypsinogen (molecular weight 27,000). The variability in elution position of material was less than 2”/,. k c t i n qffi’nity chromatography A sample of 30 ml, prepared from plasma using (NH&S04 as described below and containing column buffer reagents, was applied to a 250 ml column of Concanavalin A-Sepharose (Sigma) kept at 4°C and equilibrated in a solution of 10 mmol/l NaCI, 10 mmol/l Na acetate, I mmol/l MnClz, 1 mmol/l CaClz and 0.01% merthiolate and titrated to pH 6.5 with acetic acid. Gradient elution was performed at 4°C using a reservoir containing 250 ml of column buffer alone and a reservoir containing 250 ml of column buffer + 0.1 mol/l a-methyl-D-mannoside and 0.1 mol/l 1-0-methyl-D-glucopyranoside. Bufferscontained in addition 0.2 mmol/l PMSF and 0.5 mmol/l NEM. The flow rate was 0.5 ml/min and fractions of approximately 6.5 ml were collected into tubes containing 0.3 ml of 200 mmol/l EDTA. After completion of the gradient run, 100 ml of 1 mol/l NaCl was applied to the column and additional fractions were collected. lon-exchange chromatography A sample from Concanavalin-A Sepharose chromatography (see below) was dialysed overnight at 4°C in column buffer (50 mmol/l Tris, pH 8.4, containing 30 mmol/l NaC1) and then 5 ml was applied to a column of DE52 Cellulose (Whatman). A gradient of 30-1 50 mmol/l NaCl was formed using a reservoir containing 150 ml of column buffer alone (30 mmol/l NaCI) and a reservoir containing 150 ml of column buffer with NaCl increased to 150 mmol/l. Buffers contained in addition 1 mmol/l EDTA, 0.2 mmol/l PMSF and 0.5 mmol/l NEM. The flow rate was 0.4 ml/min and fractions of 9 ml were collected. The gradient was revealed by measuring sodium in the fractions with a Flame photometer (Model 143, Instrumentation Laboratory, Inc., Lexington, MA, USA). Partial purijication of renin in pregnancy plasma A sample of 200 ml of pooled pregnancy plasma containing 30 mmol/l EDTA was mixed with PMSF to 2 mmol/l and NEM to 5 mmol/l. (NH4)2S04 was added to give a 1.5 mol/l solution, pH 7.6, and the solution was centrifuged at 7000 g for 15 min. The supernatant was mixed with additional (NH&S04 to give a 3.5 mol/l solution and centrifuged at 20,OOOg for 30 min. The resulting precipitate was resuspended in 10 mmol/l sodium phosphate buffer, pH 7.4, dialysed against the same buffer, then diluted to 90 ml with buffer containing EDTA, PMSF and NEM to give concentrations of 10, 2 and 5 mmoljl of these reagents, respectively. Chromatography was then performed with Concanavalin A-Sepharose, followed by DE52 Cellulose as described above, to increase the specific activity of ‘active’ and ‘inactive’ renin. Since the partial purification was performed at 4°C PMSF was included to inhibit possible activation by serine proteases at this temperature (Osmond et al., 1973; Osmond & Loh, 1978). Treatment of column fractions with pepsinogen ‘Inactive’ renin in fractions from the Concanavalin A-Sepharose column and the DE52 Cellulose column was revealed by treatment with pepsinogen at pH 3.3. Pepsinogen (0.1

Properties of inactive renin in human plasma

615

ml of a 600 ng/ml solution in water) was added to 0.5 ml of the combined renin-containing fractions from the Concanavalin A-Sepharose column (pH 6.5) and to 0.5 ml of alternate fractions from the DE52 Cellulose column (pH 8.4), giving a final concentration of pepsinogen of 100 ng/ml. The samples were then dialysed in pH 3.3 buffer at 4°C for 24 h and then to pH 7.4 before assay of renin. Buffers contained 10 mmol/l EDTA, 2 mmol/l PMSF and 5 mmol/l NEM. Portions of the Concanavalin A renin pool were also dialysed in pH 3.3 buffer with concentrations of pepsinogen of 0, 10, 30, 100 and 300 ng/ml. The rationale behind using pepsinogen + pH 3.3 is as follows: Pepsin is a potent activator of ‘inactive’ renin (Morris & Lumbers, 1972; Day & Luetscher, 1975; Morris, 1978a,b), but also hydrolyses A1 from angiotensinogen (Croxatto & Croxatto, 1942; Franze de Fernandez, Paladini & Delius, 1965), although the latter effect is quite small (at pH 4.0 and 6-4) for human angiotensinogen as compared to angiotensinogens of other species (Skinner et al., 1975b). Since (1) human angiotensinogen becomes destroyed below pH 4 (Skinner, 1967; Sealey et al., 1976; Derkx et al., 1978), (2) the autocatalytic conversion to pepsin from pepsinogen is not complete until below pH 4 (Herriott, 1938) and (3) the optimum pH for activation of ‘inactive’ renin is pH 3.3 (Lumbers, 1971; Day & Luetscher, 1975; Leckie e f al.. 1977; Weinberger et al., 1977; Shulkes et al., 1978), it was decided to mix samples with pepsinogen and then dialyse them to pH 3.3. This permitted the appearance of pepsin activity in the sample after denaturation of angiotensinogen. A more important practical reason for using pepsinogen is that it could be added at neutral pH at the commencement of dialysis, and become active at acid pH and then be denatured at neutrality. RESULTS ‘Inactive‘renin in normal and pregnancy plasma ‘Inactive’ renin was revealed using pH 3.3-treatment to produce activation (Table 1). Results for males and females were similar and when combined indicated that normal human plasma contained ‘inactive’ renin as 76% of the ‘total’ renin present. In pregnancy plasma ‘inactive’ renin comprised 93% of ‘total’ renin and the levels of both ‘active’ and Table 1. ‘Active’ renin and ‘total’ renin (‘active’+‘inactive’) in plasma from normal and pregnant subjects. Activation was produced by dialysis of samples to pH 3.3 before assay at pH 7.4 ‘Active’ ‘Total’ Plasma PRC PRC PRA angiotensinogen (pmol AI.h-’/ml) (pmol/ml) Normal subjects mean (n=7) s.e.m. Primigravidae (26-30 weeks, n = 16) a) Lateral recumbant mean s.e.m. b) Supine

mean s.e:m.

5.8 1.20

24.6 0.60 3.66 0.12

I260 153

11.0 0.97

162 17.1

0.15

3960 217

11.7 1.39

157 1.67 17.6 0.25

3890 180

1.40

R . P. Day and B. J . Morris

616

‘inactive’ renin were higher than in normal plasma. Assumption of the upright posture of primigravidae after resting in a lateral recumbent position for several hours did not alter the proportion or amount of ‘inactive’ renin (Table I).

Geljltrarion of normal plasma ‘Active’ renin in the plasma of a non-pregnant adult female eluted from Sephadex G-100 in a region corresponding to an apparent molecular weight of 43,000 (Fig. I ) . Dialysis of fractions to pH 3.3 increased the renin levels in the fractions (Fig. 1) and to a degree similar to that observed when the whole plasma was treated to pH 3.3 (uiz. a 4.4-fold increase in renin in plasma compared with a 3.6-fold increase in chromatography fractions). Recoveries of renin from gel filtration were 5&70%. Much of the new activity appeared in a region of lower elution volume. By subtracting the level of ‘active’ renin in each fraction from the total present after acid activation, the elution profile of the new activity (‘inactive’ renin) was revealed. This indicated a major peak corresponding to an apparent molecular weight of 62,000 and a minor peak at molecular weight 46,000. In another experiment values at 43,000 and 61,000 were obtained for ‘active’ and ‘inactive’ renin, respectively. Since molecular weight of ‘inactive’ renin was obtained indirectly the values obtained are obviously less accurate than values obtained directly from an elution

0.4

-E

c

> 0.3 f CI

9

-i .-c

z 0.2

0.1

0 0

’

I

I

1.0

1.2

1

I

PO.(

1.4 1.6 1.8 Relative elution volume (Ve/Vo)

Fig. 1. Gel filtration of normal human plasma on Sephadex G-100. ‘Active’ renin in untreated fractions (m) was measured by the rate of generation of angiotensin I from sheep angiotensinogen at pH 7.4. ‘Inactive’ renin was revealed by treating fractions to pH 3.3 for 24 hat 4°C before assay at pH 7.4 (A). The elution profile of ‘inactive’ renin was derived (A ) by subtracting ‘active’ renin from renin values obtained after pH 3.3 treatment. This gave a major peak close to the peak of plasma albumin.

Properties of inactive renin in human plasma

617

peak. However, ‘inactive’ renin can only be detected subsequent to its activation so that such limitations in accuracy must be considered in relation to values for molecular weight of ‘inactive’ renin. Moreover, any variability in recovery between different fractions would influence the result. It is worth noting that the last protein peak from the column is albumin (molecular weight 67,000) which acts as an internal marker for molecular weight determinations. Ge1.filtrationof pregnancy plasma ‘Active’ renin in pregnancy plasma eluted from Sephadex G-100 in a region corresponding to an apparent molecular weight of 60,000 (Fig. 2a) and thus appeared to differ in size from ‘active’ renin in normal plasma. It should be noted, however, that a form of molecular weight ca. 43,000 may have been present in similar concentrations as in normal plasma, but would have been masked by the high concentrations of the 60,000 molecular weight form present in pregnancy plasma. Treatment of column fractions to cold elicited activation (Fig. 2a) and the resulting peak of renin corresponded in position to an

Relofive elution volume (Ve/Vo)

Fig. 2. Gel filtration of human pregnancy plasma on Sephadex G-100. (a) plasma run on column and then fractions treated to cold (-4°C. 72 h) to reveal ‘inactive’ renin (A), untreated (B). (b) Plasma treated to cold before chromatography(e). (c) Plasma treated to pH 3.3 before chromatography (0). Protein is indicated by the absorbance at 280 nm.Note the position of the albumin peak (molecular weight 67,000).

apparent molecular weight of 60,000. Similar results were obtained in experiments where activation was elicited by treatment to pH 3.3. Angiotensinogen eluted from the column as a single sharp peak corresponding to an apparent molecular weight of 80,000. When ‘inactive’ renin in pregnancy plasma was activated by cold or pH 3.3 before the sample was applied to the column, the renin eluted from the column in a region corresponding to an apparent molecular weight of 59,000 in each case (Fig. 2b and 2c), indicating no noticeable change in size upon activation. Lecrin affinity chromatography Concanavalin A-Sepharose affinity chromatography was used as a step in the partial purification of ‘active’ and ‘inactive’ renin in pregnancy plasma. A 1.5-3.5 mol/l

R.P. Day and B. J . Morris

618

- 30 0 Start grodient of j i 0-200 rnmol/L carbohydrates

End grodient and start I m m o l / l NaCl

--

000

E

-

-20-

a

500

-e

e ‘t 2

B

E \

C 0 rn

.-E

c

-I0

e

0

P 4

Volume of eluate (rnl) Fig. 3. Affinity chromatography of a 1 5 - 3 3 mol/l (NH4)2S04 fraction of pregnancy plasma on Concanavalin A-Sepharose. Elution was performed with a linear gradient of 0-200 mmol/l a-methyl-D-mannoside and 1-0-methyl-a-o-glucopyranoside, followed by 1 mol/l NaCI. ‘Active’ renin in untreated fractions is shown. Treatment to pH 3.3 was no longer able to produce activation. Protein is also indicated.

(NH&S04 fraction of plasma was applied to the column. ‘Active’renin was bound to the Concanavalin A-Sepharose; half was eluted with a gradient of G200 mmol/l carbohydrate an4 the remainder with 1 mol/l NaCl (Fig. 3). The latter is sufficient to remove all remaining glycoproteins from the column and is performed to regenerate such columns. The Concanavalin A-Sepharose step produced a fourteen-fold increase in specific activity of ‘active’ renin over the 1.5-3.5 mol/l (NH&S04 cut of pregnancy plasma (Table 2). Fractions from the column were dialysed to pH 3.3, but activation Table 2. Partial purification of renin activity in pregnancy plasma, showing increases in specific activity of ‘active’ renin, ‘total’ renin (‘active’ ‘inactive’)and angiotensinogen

+

‘Total’ renin Angiotensinogen Protein ‘Active’renin (pmolA1.h-’/mg (pmol AI.h-’/mg (pmol/mg) (mg/ml) Protein) Protein) a. Initial plasma b. ( N H W 0 4 I .5-3.5 mol/l c. Concanavalin A-Sepharose and lyophilized d. DE52 Cellulose and lyophilized Overall increase in specific activity

66 69 (59%)

59 72

141

20

(53%)

2.37 2.4t (57%) 6.33 (37%) 17,6$ (48%)

30

8

26

0.16 0.22 (76%*) 1.58 ( 102%)

4.85

* Recoveries from step to step are shown in parentheses. t Activation produced by pH 3.3 treatment.

(60%) 17218 (63%)

1Activation produced by pepsinogen +pH 3.3. 4 Fractions eluting between 275 and 310 ml from the DE52 Cellulose column.

3.5

619

Properties of inactive renin in human plasma

could no longer be produced. This would have occurred if ( I ) ‘inactive’ renin did not bind to the column or (2) an activating factor which acts at pH 3.3 to activate ‘inactive’ renin had been separated from the fractions containing ‘inactive’ renin. The fractions containing renin activity were pooled and lyophilized for later ion-exchange chromatography. But, before the latter was performed it was decided to test for the presence of ‘inactive’ renin in the pooled ‘active’ renin fractions as follows. Treatment with pepsinogen Treatment of the pooled renin-containing fractions from the Concanavalin A-Sepharose column to pH 3.3 alone could not produce activation. However, it was discovered that if pepsinogen, 100 ng/ml, was added before dialysis to pH 3.3 then activation could be produced (Table 2). The effects of different concentrations of pepsinogen were then studied and it was found that the degree of activation was proportional to the concentration of pepsinogen present between 0 and 100 ng pepsinogen/ml (Fig. 4). With 300 ng pepsinogen/ml the level of renin was only 20% higher than that observed with 100 ng pepsinogen/ml. Ion-exchange chromatography The pooled renin-containing fractions from Concanavalin A-Sepharose were run on DE52 Cellulose. ‘Active’ renin eluted between 75 and 100 mmol/l NaCl, with a major peak at 88 mmol/l NaCl and several minor peaks, the significance of which is not known (Fig. 5). Dialysis of fractions to pH 3.3+pepsinogen resulted in a 12.7-fold increase in renin. The new activity had a major peak at 72 mmol/l NaCl and minor peaksat 60,88 and 90 mmol/l NaCI. Angiotensinogen eluted at 100-1 15 mmol/l NaCl and was thus clearly separated from the renin activities. The amount of ‘active’ renin eluted was 8% of the ‘total’ renin measured after treatment of the fractions with pepsinogen, indicating 92% ‘inactive’ renin. This proportion of inactive renin is similar to that measured in the

Lu

pH7.4

I 0

I

I

I

100

200

300

+ Pepsinogen (ng/ml),pH 3.3 Fig. 4. Effect of treatment of the pooled renin-containing fractions from the Concanavalin A-Sepharose column with pepsinogen (0-300 ng/ml) added before dialysis to pH 3.3 (24 h, 4°C).

620

R. P. Day and B. J . Morris

Volume of eluate (mi) Fig. 5. Ion-exchangechromatography of the pooled renin-containingfractions (solid lines) from ConcanavalinA-Sepharoseon DE52 Cellulose.‘Active’renin in untreated fractions(m) consisted of several peaks. ‘Inactive’ renin was revealed by dialysing fractions to pH 3.3 after adding pepsinogen (A), 100 ng/ml. The position of angiotensinogen (A) is also shown and was clearly separated from the rain activities. Protein (0) and sodium ( x ) are also indicated.

original plasma sample (93%) shown in Table 2. Renin-containing fractions from the DE52 Cellulosecolumn when pooled, lyophilized, reconstituted in water and treated with pepsinogen +pH 3.3 gave only a 3.6-fold increase in renin (Table 2), a value similar to the four-fold rise seen when the pooled Concanavalin A renin fractions were treated with pepsinogen +pH 3.3 (Table 2). The reason for this reduced degree of activation is not apparent. Nevertheless, the present experiments show that ‘inactive’ renin and ‘active’ renin can be recovered in similar proportions at the final (DE52 Cellulose) step of the purification when calculations are performed on column fractions before pooling. The inability to activate ‘inactive’renin in Concanavalin A-Sepharose fractions using pH 3.3 treatment alone may be a result of removal of an activating factor whose activity can be reproduced by adding pepsinogen before dialysis to pH 3.3. The clear separation of ‘inactive’ renin from angiotensinogen is good evidence that angiotensinogen does not contribute to the phenomenon of ‘activation’and that pepsin (formed from pepsinogen during dialysis to pH 3.3) does not hydrolyse human angiotensinogen to give A1 in amounts sufficient to influence the results obtained. DISCUSSION The present study has examined ‘active’ and ‘inactive’renin in human plasma, using pH 3.3- or cold-treatment to induce activation. In normal human plasma ‘active’renin had an apparent molecular weight of 43,000 and pH 3.3-treatment of column fractions resulted in the appearance of new activity, the molecular weight of which was determined by an indirect subtraction method with obvious limitations in accuracy; the new activity corresponded to peaks of molecular weight of 62,000 and 46,000. Thus ‘inactive’renin is mostly larger than ‘active’ renin in normal plasma. Since ‘full’ activation could be produced after gel filtration the fractions presumably contained sufficient activating enzyme for this purpose. Pertinent to the discussion later the molecular weight of pepsinogen is 43,000 and would therefore co-elute with ‘active’ renin. In pregnancy plasma the amount and proportion of ‘inactive’ renin was much higher and it had an

Properties qf inuctiile renin in lzuman plusmu

62 1

apparent molecular weight of 60,000. Of interest was that the ‘active’ renin in pregnancy plasma also had an apparent molecular weight of 60,000. The amount of ‘active’ 60,000 molecular weight renin was sufficiently high to mask any 43,000 molecular weight ‘active’ renin that may have been present in addition in pregnancy plasma. The findings are consistent with our previous findings of a ‘big’ renin, in Wilms tumour plasma and human amniotic fluid (Day & Luetscher, 1975; Morris, 1978b), but at variance with Boyd (1977) and Shulkes er al. (1978) who, using AcA 44 Ultrogel and Sephadex G-75 superfine, respectively, found activation at ca. 46,000 molecular weight. We are currently investigating the influence of gel filtration media in these results. ‘Active’ renin in a 1.5-2.3 mol/l (NH&SOd preparation of proteins from pregnancy plasma bound to Concanavalin A-Sepharose. Thus plasma renin is a glycoprotein.* Others have found that renin from human kidneys also binds to Concanavalin A-Sepharose. It too can be eluted as two peaks, one after the addition of 5 mmol/l mannoside and the other after addition of 500 mmol/l mannoside to the column (Printz & Dworschack, 1977). These investigators also studied renal renin from rabbits and found that ‘acid’ activation (pH 2.5, 1 h, 22°C) before chromatography gave an elution profile of renin similar to that of untreated sample. Activation after chromatography was not tested in their studies. In the present work, treatment of fractions, from Concanavalin A-Sepharose chromatography, at pH 3.3 was no longer able to reveal ‘inactive’ renin. One interpretation of this finding is that a factor other than H is needed to produce activation and that this factor does not bind to Concanavalin A. When pepsinogen was added during dialysis of eluate in pH 3.3 buffer, activation was again produced. ‘Inactive’ renin from the Concanavalin A-Sepharose column when chromatographed on a DE52 Cellulose column could again be activated by treating fractions to pepsinogen + pH 3.3 and at the concentration used, 100 ng pepsinogen/ml, the degree of activation observed in Fig. 5 was similar to that observed after treatment of the original sample of plasma to pH 3.3 alone. Since human pepsinogen is not a glycoprotein and the concentration of human serum group I pepsinogens is 50-1 75 ng/ml (mean f s.d. = 1 1 1 f41) (Samloff, Liebman & Panitch, 1975) the present results support the possibility that pepsinogen, upon conversion to pepsin, could be responsible for at least a major portion of the activation of ‘inactive’ renin observed during dialysis of plasma to pH 3.3. In previous studies the pH profile of activation and the existence of an activating factor that behaved like an enzyme and was resistant to pH 1.5 suggested that the activation of ‘inactive’ renin in human amniotic fluid is due to an action of pepsin subsequent to its conversion from pepsinogen by H + and autocatalytic mechanisms (Morris & Lumbers, 1972). The inability of inhibitors of serine- and sulphydryl-proteases and of chelating agents to prevent activation by pH 3.3-treatment lends further support to a role of a pepsin-like (carboxyl) protease. The elution of ‘inactive’ renin from DE52 Cellulose during partial purification of pregnancy plasma suggests that it is more positively charged than ‘active’ renin. A similar pattern has recently been reported by Shulkes et al. (1978) who studied normal male plasma using DEAE-Sephadex. Both species consisted of several peaks, suggesting that forms of ‘active’ and ‘inactive’ renin exist with minor differences in charge. Forms of renin differing in charge have been commonly observed and may reflect differences in carbo* The apparent molecular weight of ‘big’ renin in plasma may be an overestimate as glycoproteins are often +

retarded on Sephadex, although ‘big’ renin in amniotic fluid is not a glycoprotein and has a molecular weight, estimated using Sephadex, of 59.000 (Morris, 1978b).

622

R . P. Day and B. J . Morris

hydrate attachments (Skeggs et al., 1967; Printz & Dworschack, 1977). In human amniotic fluid ‘inactive’ renin also appears to be less negatively charged than the ‘active’ renin present (Morris, 1978b). It was reported recently that renin values after activation by dialysis of plasma to pH 3.3 (and then back to pH 7.4) were 2.5 times higher than renin values after activation by exposure to -4°C (Atlas, Sealey & Laragh, 1978).Our findings have been similar (1.2-2.9 times higher; unpublished results with normal plasma). A two-step mechanism has been suggested to explain activation by pH 3.3/7.4 treatment: ( I ) acid protease activates ‘inactive’ renin at pH 3.3 and (2) at pH 3.3 inhibitors are removed, thus unmasking neutral protease which activates ‘inactive’ renin during dialysis to pH 7.4. (Atlas et al., 1978). However, at the pH used for assay (5.7) other proteases may generate A1 from angiotensinogen (Oparil, Koerner & Haber, 1974) so that the degree of contribution of neutral proteases to acid activation should be examined more closely. This did not occur in our experiments since the pH of renin assay was 7.4. We have found that after chromatography of pregnancy plasma on Sephadex G- 100, activation produced by pH 3.3/7.4 treatment of fractions was similar to that produced by exposure to cold. Furthermore, pepsinogen at physiological concentrations at pH 3.3 (4°C) could produce full activation and these samples had been treated with PMSF (see below). Although it appears that pepsinogen/pepsin has an important role in activation induced by treatment to pH 3.3, the mechanism of activation that occurs during treatment to cold is different. Cold-induced activation can be inhibited by DFP (Atlas, Sealey & Laragh, 1977, 1978; Osmond & Loh, 1978), trasylol (Osmond & Loh, 1978) or PMSF (Day & Morris, unpublished results). Inhibition by PMSF is greater with 10 mmol/l than 1 mmol/l of the compound and is slightly greater for plasma than serum. A role of serine protease(s) which become(s) activated at low temperature is therefore apparent. However, pH 3.3 and -4°C are decidedly ‘non-physiological’ conditions. Trypsin can activate ‘inactive’renin at pH 7.4 (Morris & Lumbers, 1972; Day & Luetscher, 1975; Cooper et al., 1977), but trypsin would not occur in plasma except in pancreatic disease states. We have found that the serine proteases kallikrein and fibrinolysin can activate inactive renin (Morris & Day, 1978, 1979). With prior acidification of plasma, which may denature endogenous kallikrein inhibitors, ca. 1.3 x M kallikrein is required for 50% activation (at 25°C) (calculated from Sealey et al., 1978). Although kallikrein has been reported to be ten times more potent than trypsin (Sealey et al., 1978), on a molar basis, pepsin at pH 3.3 would appear to be more potent still; at 4°C the concentration of pepsinM at pH M and of pepsin is ca. 3 x ‘ogen’ required for 50% activation is ca. 5 x 4.8 (22°C) (Morris, 1978a), which is a pH far from pepsin’s pH optimum. The capacity of endogenous serine proteases to elicit activation of ‘inactive’ renin in human plasma at pH 7.4 has recently been demonstrated in experiments in which puff adder venom was used to destroy their endogenous inhibitor proteins (Lawrence & Morris, 1979). ACKNOWLEDGMENTS

We thank Barbara Moffat for technical assistance. Dr Day was supported by a stipend from the Skaggs Foundation. Dr Morris was supported sequentially by a C. J. Martin Research Fellowship from the National Health and Medical Research Council of Australia, an Advanced Fellowship from the California Heart Association and a stipend from the Skaggs Foundation. This work was funded by the Skaggs Foundation.

Properties ojinactive renin in human plasma

623

REFERENCES Atlas, S.A., Sealey, J.E. & Laragh. J.H. (1977) Dependency of acid and cryoactivation of inactive plasma renin on prior activation of a neutral serine protease. Kidney International, 12,495 (Abstract). Atlas, S.A., Sealey, J.E. & Laragh, J.H. (1978) ‘Acid’- and ‘cryo’-activated inactive plasma renin. Similarity of their changes during 8-blockade. Evidence that neutral protease(s) participate in both activation procedures. Circulation Research, 43, Suppl. I, 1128-1 133. Boyd, G.W. (1977) An inactive higher-molecular-weight renin in normal subjects and hypertensive patients. Lanccv. i, 215-218. J Cooper, R.M., Murray, G.E. & Osmond, D.H. (1977) Trypsin-induced activation of renin precursor in plasma of normal and anephric man. Circulation Research, 40, Suppl. 1, I 1 7 1-1 179. Croxatto, H. & Croxatto, R. (1942) ‘Pepsitensin’-a hypertensin like substance produced by peptic digestion of proteins. Science. 95, 101-102. Day, R.P. & Luetscher, J.A. (1974) Big renin: a possible prohormone in kidney and plasma of a patient with Wilms’ tumor. Journal of Clinical Endocrinology and Metabolism. 38,923-926. Day, R.P. & Luetscher, J.A. (1975) Biochemical properties of big renin extracted from human plasma. Journalof Clinical Endocrinology and Metabolism. 40, 1085-1093. Day, R.P., Luetscher, J.A. & Gonzales, C.M. (1975) Occurrence of big renin in human plasma, amniotic fluid and kidney extracts. Journal of Clinical Endocrinology and Metabolism. 40, 1078-1084. Derkx, F.H.M., Wenting, G.J., Man in’t Veld, A.J., Van Cool, J.M.G., Verhoekn, R.P. & Schalekamp, M.A.D.H. (1976) Inactive renin in human plasma. Luncet, ii, 496499. Derkx, F.H.M., Wenting, G.J., Man in’t Veld, A.J., Verhoeven, R.P. & Schalekamp, M.A.D.H. (1978) Control of enzymatically inactive renin in man under various pathological conditions. Implications for the interpretation of renin measurements in peripheral and renal venous plasma. Clinical Science and Molecular Medicine, 54, 529-538. Franze de Fernandez, M.T., Paladini, A.C. & Delius, A.E. ( 1 965) Isolation and identification of pepsitensin. The Biochemical Journal. 97,540-546. Gibson, R. (1973) Properties of renin in human amniotic fluid and foetal membranes. MSc Thesis, University of Melbourne, Australia. Herriott, R.M. (1938) Isolation. crystallization, and properties of serine pepsinogen. Journal of General 21, 501 -540. Ph~~.siology. Hseuh, W.A., Luetscher, J.A., Carlson. E.J. & Grislis, G. (l978a) Big renin in plasma of healthy subjects on high sodium intake. Lancet. i, 1281-1284. Hseuh, W.A., Luetscher, J.A., Carlson, E., Grislis, G., Elbaum, D. & Chavarri, M. (1978b) Comparison of cold and acid activation of big renin and of inactive renin in normal plasma. Journal of Clinical Endocrinology and Merabolism, 47, 792-799. Lawrence, C.H. & Morns, B.J. (1979) Activation of inactive renin in human plasma by PuffAdder venom, IRCS Medicul Science. 7 , 208. Leckie, B.J., McConnell, A., Grant, J., Morton, J.J., Tree, M. 81Brown, J.J. (1977) An inactive renin in human plasma. Circulation Research, 40, Suppl. I. 146151. Lowry, O.H., Rosebrough, N.J., Farr, A.L. & Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry. 193,265-275. Lumbers, E.R. (1971) Activation of renin in human amniotic fluid by low pH. Enzymologia. 40,329-336. Morris, B.J. (1978a) Activation of human inactive (‘pro-’) renin by cathepsin D and pepsin. Journalof Clinical Endocrinology and Metabolism. 46, 153-1 57. Morris, B.J. (1978b) Properties of the activation by pepsin of inactive renin in human amniotic fluid. Biochimica et Biophysica Acta. 527,86-97. Morris, B.J. & Day, R.P. (1978) Activation of inactive renin by kallikrein. IRCS Medical Science, 6,348. Morris, B.J. & Day, R.P. (1979) Activation of inactive renin by fibrinolysin, IRCS Medical Science, 7, 188. Morris, B.J. & Lumbers, E.R. (1972) The activation of renin in human amniotic fluid by proteolytic enzymes. Biochimica et Biophysica Acta, 289,385-391. Morris, B.J. & Reid, I.A. (1978) The distribution of angiotensinogen in dog brain studied by cell fractionation. Endocrinology, 103,492-500. Nielsen, A.H., Malling, C. & Poulsen, K. (1978) Characteristics and conversion of high molecular weight forms of renin in plasma and their incomplete activation by the current acid treatment. Biochimica et Biophysicu Acra, 534,246-257.

624

R . P. Day and B. J . Morris

Oparil, S., Koerner, T.J. & Haber, E. (1974) Etrects of pH and enzyme inhibitors on apparent generation of angiotensin I in human plasma. Journal of Clinical Endocrinology and Metabolism, 39.965968. Osmond, D.H. & Loh, A.Y. (1978) Protease as endogenous activator of inactive renin. Lancer, i, 1313 (letter). Osmond, D.H.. Ross, L.J. & Scaiff, K.D. (1973) Increased renin activity after cold storage of human plasma. Canadian Journal of Physiology and Pharmacology, 51,705-708. Printz, M.P. & Dworschack, R.T. (1977) Evidence for the glycoprotein nature of kidney renin. Biochimica et Biophysica Acta. 494, 162-171. Reid, LA., Morris, B.J. & Ganong, W.F. (1978) The renin-angiotensin system. Annual Review of Physiology, 40, 377-4 10. Samloff, I.M., Liebman, W.M. & Panitch, N.M. (1975) Serum group I pepsinogens by radioimmunoassay in control subjects and patients with peptic ulcer. Gastroenterology. 69,83-90. Sealey, J.E., Atlas, S.A., Laragh, J.H., Om, N.B. & Ryan, J.W. (1978) Human urinary kallikrein converts inactive to active renin: a possible physiological activator of renin. Nature, 275, 144146. Sealey, J.E. & Lnragh, J.H. (1975) 'Prorenin' in human plasma? Methodological and physiological implications. Circularion Research, 36,37, Suppl. I, 110-116. Sealey, J.E., Moon, C., Laragh, J.H. & Alderman, M. (1976) Plasma prorenin: cryoactivation and relationship to renin substrate in normal subjects. American Journal of Medicine, 61, 731-738. Shulkes, A.A., Gibson, R.R. & Skinner, S.L. (1978) The nature of inactive renin in human plasma and amniotic fluid. Clinical Science and Molecular Medicine, 55,41-50. Skeggs, L.T., Lentz, K.E., Kahn, J.R. & Hwhstrasser, H. (1967) Studies on the preparation and properties of renin. Circulation Research, 20,21, Suppl. II,91-100. Skinner, S.L. (1967) Improved assay methods for renin concentration and activity in human plasma. Circularion Research, 20,391402. Skinner, S.L., Cran, E.J., Gibson, R., Taylor, R., Walters, W.A.W. & Catt, K.J. (1975a) Angiotensin I and 11, active and inactive renin, renin substrate, renin activity, and angiotensinase in human liquor amnii and plasma. American Journal of Obstetrics and Gynecology, 121,626-630. Skinner, S.L., DUM, J.R., Mazzetti, J., Campbell, D.J. & Fidge, N.H. (1975b) Purification, properties and kinetics of sheep and human renin substrates. Australian Journal oJ Experimental Biology and Medical Science, 53,77-88. Stockigt, J.R., Collins,R.D. & BigPeri, E.G. (1971) Determination of plasma renin concentration by angiotensin I immunoassay. Diagnostic import of precise measurement of subnormal renin in hyperaldosteronism. Circulation Research, 28,29, Suppl. 11,11175-11191. Weinberger, M., Aoi W. & Grim, C. (1977) Dynamic responses of active and inactive renin in normal and hypertensive humans. Circulation Research, 41, Suppl. II,I121-II25.