Reversible Cryoactivation of Recombinant Human Prorenin"

Cleavage of prorenin's prosegment causes irrevers- ible formation of renin. In contrast, renin activity is reversibly exposed when prorenin is acidified to pH 3.3. Nonetheless, acidification of plasma results in ir- reversible activation of prorenin, because endogenous proteases cleave the prosegment of acid-activated pro- renin. Chilling of plasma results in irreversible cryoactivation of prorenin. In this study we investigated whether cryoactivation of purified prorenin is reversible. The intrinsic renin activity of recombinant human prorenin was measured by an enzyme kinetic assay using partially purified human angiotensinogen as substrate. Results are expressed as a percent (mean f S.E.) of the maximal activity exposed after limited proteolysis by trypsin. The intrinsic renin activity of two pools (0.3 and 0.06 Goldblatt units/ml) was 1.5% f 0.3 and 1.2% 2 0.6 at 37 "C. Activity increased to 19% f 0.3 and 26% f 0.5 after incubation at 0 "C and to 5.4% k 0.5 and 2.1% f 1.2 at room temperature. Cryoactiva- tion did not occur in buffers containing more than 1 M NaCl. It took 8 min at 37 "C or 180 min at room temperature for cryoactivated prorenin to lose half of its intrinsic renin activity. It took 48 and 26 h, respec- tively, at 0 and subjected to ammonium sulfate precipitation. The proteins precipitating between 30 and 65% satu-ration were removed by centrifugation. The pellet was dialyzed and purified by ion exchange chromatography on DEAE-Sepharose. Angiotensinogen was eluted with a 0.07 M NaCl to 0.4 M NaCl gradient in 50 mM Tris-C1, pH 8.0, containing 3 mM Na2EDTA. The final substrate preparation was able to generate 20,000 ng of AI/ml during incubation with excess renin (17 GU/liter).


Reversible Cryoactivation of Recombinant Human Prorenin"
(Received for publication, October 21, 1991) Tina M. Pitarresi Cleavage of prorenin's prosegment causes irreversible formation of renin. In contrast, renin activity is reversibly exposed when prorenin is acidified to pH 3.3. Nonetheless, acidification of plasma results in irreversible activation of prorenin, because endogenous proteases cleave the prosegment of acid-activated prorenin. Chilling of plasma results in irreversible cryoactivation of prorenin. In this study we investigated whether cryoactivation of purified prorenin is reversible.
The intrinsic renin activity of recombinant human prorenin was measured by an enzyme kinetic assay using partially purified human angiotensinogen as substrate. Results are expressed as a percent (mean f S.E.) of the maximal activity exposed after limited proteolysis by trypsin. The intrinsic renin activity of two pools (0.3 and 0.06 Goldblatt units/ml) was 1.5% f 0.3 and 1.2% 2 0.6 at 37 "C. Activity increased to 19% f 0.3 and 26% f 0.5 after incubation at 0 "C and to 5.4% k 0.5 and 2.1% f 1.2 at room temperature. Cryoactivation did not occur in buffers containing more than 1 M NaCl. It took 8 min at 37 "C or 180 min at room temperature for cryoactivated prorenin to lose half of its intrinsic renin activity. It took 48 and 26 h, respectively, at 0 "C for the two pools of prorenin at 37 O C to regain half of their maximum intrinsic activity at 0 "C.
A direct immunoradiometric assay that detects active renin but not prorenin was able to detect cryoactivated prorenin.
These results show that human prorenin can be reversibly cryoactivated in buffers of low ionic strength and has greater intrinsic activity at room temperature than at 37 "C.
Renin, an aspartyl protease, is synthesized and stored in the juxtaglomerular cells of the kidney (1). Prorenin is the biosynthetic precursor of renin (2)(3)(4). Several characteristics make prorenin an unusual precursor. It circulates together with renin but at 10-fold higher concentrations (5)(6)(7). It is the major product of renin gene expression in the kidney. Renin is packaged into granules to be secreted on demand but prorenin is secreted constitutively as it is synthesized (8). The kidney is the only source of circulating renin; it is the predom-inant but not the only source of circulating prorenin. The eye, adrenal cortex, and female reproductive organs are examples of extra-renal sources of prorenin (9)(10)(11)(12)(13). Prorenin is not converted to renin in the circulation (14).
Prorenin may have a regulated function. Secretion of renal renin and prorenin is usually stimulated by the same stimuli, such as sodium depletion or converting enzyme inhibition, but ovarian prorenin secretion is regulated by luteinizing hormone and human chorionic gonadotropin (11). For example plasma prorenin increases 2-fold at ovulation and 10-fold during pregnancy (15). Uterine prorenin increases markedly during gestation, and the placenta synthesizes large amounts of prorenin which are secreted into the amniotic cavity.
In contrast to renin which causes increased blood pressure and reduced tissue perfusion via the vasoconstriction effects of angiotensin 11, increased prorenin is usually associated with low blood pressure (as in pregnancy) or with hyperperfusion (as in pregnancy (15) and diabetes mellitus (16)). This has given rise to speculation that prorenin has a counteractive role to that of renin (17). The concentrations of prorenin in the circulation and in reproductive tissues are much greater than that of renin giving it the potential, if activated, to produce much higher concentrations of angiotensin I1 than renin. High angiotensin I1 concentrations can cause vasodilation (18). Thus, the phenomenon of reversible activation of prorenin is interesting, since it opens the possibility that prorenin could be reversibly activated in vivo at its target site and could evoke properties of angiotensin I1 that were previously not thought to occur at physiological concentrations.
Prorenin is irreversibly activated in human plasma that has been acidified to pH 3.3 and returned to neutral pH (19,20) or that has been incubated at -4 "C (21). Irreversible acid activation and cryoactivation of prorenin do not occur in plasma that is devoid of factor XI1 or prekallikrein (22,23). Acid activation of purified prorenin is reversible (24,25). The phenomenon of cryoactivation has not yet been studied with purified reagents. The purpose of this study was to investigate the effects of temperatures from 37 to 0 "C and sodium chloride concentration on the intrinsic catalytic activity of purified recombinant human prorenin.

Reversible Cryoactivation
of Human Prorenin hamster ovary cells transfected with a vector containing cDNA for preprorenin and was purified by affinity chromatography (26). It was a gift from the Upjohn Company. Prorenin concentration was determined by standardization against human renal renin after trypsin activation (described below); 1 Goldblatt unit (GU) of renin generates 120 pg of angiotensin I/h with human renin substrate (angiotensinogen) at K,,,. A stock solution (determined to have a prorenin concentration after activation with trypsin of 115 GU/ml) in 50 mM Tris-C1, pH 7.5, containing 5 mM benzamidine hydrochloride, was stored in 5O-pl aliquots at -70 "C. Two prorenin pools (0.3 and 0.06 GU/ml) were prepared in 50 mM sodium phosphate buffer, pH 7.4, containing 0.1 M NaC1, 10 mM benzamidine hydrochloride, 3 mM Na,EDTA, 0.1% sodium azide, and 0.5% BSA. Both pools were equilibrated at 0 "C for 7 days and stored in 50-p1 aliquots at -70 "C. Two other sources of recombinant human prorenin, gifts from California Biotechnology, Inc. (Cal Bio) (27) and Pfizer, Inc. (28), were also diluted to 0.3 and 0.06 GU/ml and stored at -70 "C.
Renin Inhibitor-Hui Peptide 27, a gift from Dr. K. Y. Hui,20 nM was used to stop the action of renin on angiotensinogen (29).
Angiotensinogen-Angiotensinogen was partially purified from human plasma that had been heat-treated (50 "C for 30 min) to inactivate renin (30). Heat-denatured proteins were removed by centrifugation (10,000 X g for 30 min a t 4 "C). The supernatant was diluted 1:2 with 50 mM Tris-C1, pH 8.0, and subjected to ammonium sulfate precipitation. The proteins precipitating between 30 and 65% saturation were removed by centrifugation. The pellet was dialyzed and purified by ion exchange chromatography on DEAE-Sepharose. Angiotensinogen was eluted with a 0.07 M NaCl to 0.4 M NaCl gradient in 50 mM Tris-C1, pH 8.0, containing 3 mM Na2EDTA. The final substrate preparation was able to generate 20,000 ng of AI/ml during incubation with excess renin (17 GU/liter).

Methods
Time Course of Cryoactiuation-0.3 and 0.06 GU/ml pools of prorenin were preincubated for either 1 h at 37 "C or 3 h at room temperature and then placed in a 0 "C ice bath for 5 days. Samples were removed and frozen a t 6-h intervals for the first 24 h and then at 24-h intervals for up to 5 days. Additional samples were removed in duplicate at 0 h, 2 days, and 5 days; one was incubated for 1 h at 37 "C to determine if cryoactivation was reversible, the other was trypsin-activated (see below) to determine if total renin concentration remained stable. Activity was determined by generation of angiotensin I (see below).
Time Course of Reversal of Cryoactiuation-Reversal of cryoactivation was studied at room temperature and at 37 "C. Samples of the 0.3 and 0.06 GU/ml pools which had been incubated for 5 days at 0 "C were rapidly thawed and then incubated a t room temperature or in a 37 "C water bath. Samples incubated at room temperature were frozen at 0, 15, 30, and 60 min and then at hourly intervals for 8 h. Samples incubated at 37 "C were frozen at 5-min intervals for the first 20 min and then at 60 and 120 min. Additional samples were taken, in duplicate, at 0, 2, and 8 h which were either re-incubated at 0 "C for 4 days to determine if cryoactivation could still occur or trypsin-activated to determine if total renin concentration remained stable.
Effect of Sodium Chloride Concentration on Cryoactiuation-The time course of cryoactivation was examined using the buffer in which the prorenin pools were prepared (described above) to which was added 0-3 M sodium chloride. The final concentrations of sodium and chloride in the buffers was measured using ion-sensitive electrodes (Astra 8, Beckman).
Trypsin Actiuation of Prorenin-Total renin activity was determined by activation of 0.3 ml of sample with 3 1 1 of 10 mg/ml TPCKtreated trypsin (final concentration 100 pg/ml) for 1 h at room temperature. The reaction was stopped by the addition of 3 pl of 20 mg/ml soybean trypsin inhibitor (final concentration 200 pg/ml) and incubation for 15 min at room temperature (31).

Measurement of Renin Activity by Generation of Angiotensin I-
Renin-like activity was assessed by incubating samples for 5 min at 37 "C, pH 7.4, in the presence of 3 mM Na,EDTA, 3 mM phenylmethanesulfonyl fluoride, and 1330 ng of AI equivalents/ml (K,) human an equal volume of 20 nM renin inhibitor a t 0 "C. The short incubation angiotensinogen. The reaction was stopped abruptly by addition of time was chosen to minimize the reversal of cryoactivation of prorenin. Trypsin-activated samples were incubated for 1 h at 37 "C to generate AI. Angiotensin I was quantitated by radioimmunoassay (32). Results were expressed as nanograms AI/ml/h and then calculated as a percent of the maximal activity determined by limited proteolysis with trypsin.
Effect of Cryoactiuated Prorenin on a Direct Renin Radioimmunoassay-A direct immunoradiometric assay (IRMA) for active renin (Renin radioimmunoassay, Institut Pasteur, Paris) is unable to detect prorenin in plasma when the assay is performed at room temperature. To determine if prorenin was detectable after cryoactivation, the IRMA was also run at 4 "C. Before assay 0.3 GU/ml prorenin samples were cryoactivated at 0 "C for 7 days followed by (a) no further treatment, (b) re-equilibration at room temperature for 24 h, (c) reequilibration at 37 "C for 1 h, or (d) trypsin activation to determine the total enzyme present. Samples were diluted using the diluent supplied with the kit.

RESULTS
Temperature Dependence of the Intrinsic Activity of Human Prorenin-The level of intrinsic renin-like activity of recombinant human prorenin was dependent on the temperature at which the prorenin was equilibrated (Table I). After equilibration at 0 "C the 0.3 and 0.06 GU/ml prorenin pools were, respectively, 19% f 0.3 and 26% +. 0.3 of maximal activity.
The difference in activity between the two pools was statistically significant ( p < 0.05 by Student's t test). After 8 h at room temperature the levels of activity were 5.4% f 0.5 and 2.1% f 1.2 of maximal (not significant). After 2 h at 37 "C the prorenin pools exhibited the lowest level of activity, 1.5% +-0.3 and 1.2% f 0.6 of maximal (not significant). Prorenin equilibrated at room temperature or 37 "C always exhibited an increase in activity when re-equilibrated to 0 "C. There was no permanent change in the activity of the pools that would suggest that the prorenin had been irreversibly converted to renin. There was no change in the level of activity measured after trypsin activation that would suggest any change in the total amount of prorenin present.
Time Course of Cryoactiuation-Recombinant human prorenin that had been pre-equilibrated at 37 "C ( Fig. 1) or at room temperature ( Fig. 2) exhibited increased catalytic activity when incubated at 0 "C. This increase in activity was completely reversible at 37 "C. Extrapolation analysis (33) revealed differences between the two different concentrations of prorenin. In samples pre-equilibrated at 37 "C, maximum activity was calculated to be 21% for the 0.3 GU/ml pool and 35% for the 0.06 GU/ml pool. Similarly, the maximal activity was 22 and 38%, respectively, for the two pools pre-equilibrated at room temperature. tlIz to maximal activity was different between the two pools. When pre-equilibrated at 37 "C, it took 26 h for the 0.3 GU/ml pool and 48 h for the 0.06 GU/ml pool. When preequilibrated at room temperature these times were 23 and 58 h, respectively. In fact, equilibrium was not reached in the current studies which lasted a maximum of 5 days (120 h). Samples taken at 0, 48, and 120 h decreased in activity   when they were incubated for 3 h at room temperature but the activity did not return to base line, because 3 h a t room temperature was not long enough for the prorenin to reach a new equilibrium (Fig. 4 indicates that at least 8 h were needed t o reach equilibration at room temperature).
Both sources of prorenin showed reversal of cryoactivation, but activity did not return Pfizer prorenins cryoactivated to a lesser degree then the 0.06 GU/ml pool. Cryoactivation was reversed at 37 "C.
Time Course of Reversal of Cryoactiuation-Inactivation of cryoactivated prorenin occurred quickly (Fig. 4). The half-life (tip ) is defined as the time needed for the activity to decrease to one-half of its initial level. The tl,P was 8 min for both pools a t 37 "C (Fig. 4, upper panels) and 120 min for both pools at room temperature (Fig. 4, lower panels). Following reversal of cryoactivation it was possible to recryoactivate both pools.

-
tion-Cryoactivation was completely blocked by high sodium chloride concentration (Fig. 5 ) . All samples were diluted to a sodium chloride concentration of 0.1 M prior to assay for renin activity. The degree of cryoactivation was reduced in buffers containing more than 0.5 M NaCl and did not occur in buffers with more than 1 M sodium chloride. Effect of Cryoactivated Prorenin on a Direct Renin Radwimmunoassay-Cryoactivated prorenin was detected by a di- Two pools of human recombinant prorenin were equilibrated for 5 days at 0 "C and then incubated a t either 37 "C for up 2 h (upperpanels) or at room temperature for up to 8 h. Activity is expressed (mean f S.E.) as a percentage of the maximal activity generated by limited proteolysis by trypsin. Maximal activity in the 0.3 GU/ml pool was 3325 * 89 ng AI/ml/h. Maximal activity in the 0.06 GU/ml pool was 654 f 13 ng AI/ml/h. Filled circles indicate the activity in samples incubated at 37 "C or a t room temperature. Open circles indicate the activity in samples that were subsequently cryoactivated for 4 days at 0 "C. rect IRMA developed to detect active renin but not prorenin (Table 11). The IRMA normally requires samples to be incubated at room temperature for a total of 5 h. Since this may cause reversal of cryoactivation, we also performed the IRMA at 4 "C. Immunoreactive renin detected in the 4 "C IRMA was greatest in the cryoactivated sample (12% f 1. binant human prorenin is inhibited by high concentrations of sodium chloride. A 0.3 GU/ml pool of recombinant human prorenin was incubated at 0 "C for 0 , 6 , and 18 h in the presence of increasing concentrations of sodium chloride in a 50 mM sodium phosphate 10 buffer, pH 7.4, containing 10 mM benzamidine HC1,3 mM Na2EDTA, 0.1% sodium azide, and 0.5% BSA. The sodium concentration (measured with a sodiumsensitive electrode) is indicated on the ordinate. The intrinsic renin activity is expressed (mean f S.E.) as a percentage 5 of the maximal activity generated by limited proteolysis by trypsin, maximal activity was 3911 f 42 ng AI/ml/h. The inset graph shows the difference in the degree of activation after 18 and 6 h at 0 "C at the various sodium concentrations. RT, room temperature. Measured after trypsin activation; 244 ng/ml = 3660 ng/ml/h, 203 ng/ml = 3045 ng/ml/h. room temperature (4.3% & 0.3 of maximal). Lower levels of immunoreactive renin were detected in the IRMA carried out at room temperature.
Technical Problems-An early difficulty with these studies was caused by the fact that the purified recombinant prorenin was susceptible to irreversible activation, perhaps because of inadvertent contamination of buffers with proteases. In some experiments the activity of the cryoactivated samples did not return to base line after a 37 "C incubation (Fig. 3). This problem was solved when buffers were made with high grade crystalline BSA heat-treated and sterile-filtered.
The renin assay is an enzyme kinetic assay that is carried out at 37 "C, and because cryoactivation of prorenin is reversed a t 37 "C, we chose a 5-min incubation time at 37 "C for generation of angiotensin I to minimize the exposure of prorenin to the elevated temperature. Nonetheless since the t1r2 of reversal at 37 "C is only 8 min we must have slightly underestimated the maximum degree of cryoactivation that actually occurred.
For consistency we pre-equilibrated all pools of prorenin together at 0 "C. They were then frozen. Samples were thawed rapidly in front of a fan and equilibrated to room temperature or 37 "C as needed. This transition to and from the frozen state resulted in some variability of the base-line estimation of activity. This is illustrated in Fig. 4 where the initial level of cryoactivated prorenin is actually lower than in the samples that were re-cryoactivated after exposure to 37 "C and room temperature. It is also illustrated in Figs. 1-3 where the baseline values of the samples before cryoactivation are actually higher than those in the samples used to test reversal of cryoactivation.

DISCUSSION
In this study we showed that human prorenin can be reversibly cryoactivated. After 5 days at 0 "C, prorenin had close to 20% of the maximal activity formed after cleavage of the prosegment by trypsin. At room temperature 3.5% of maximal activity was measured, whereas only 1.4% was detected at 37 "C. The intrinsic renin activity of prorenin could be repeatedly changed by altering the temperature. There was no increase in activity over that measured at 37 "C when prorenin was incubated at 0 "C in the presence of 3.6 M sodium. Antibodies directed against renin's active site that do not recognize prorenin were able to detect cryoactivated prorenin.
The recombinant human prorenin used in this study had a small degree of residual renin-like activity at 37 "C and in the presence of high concentration of sodium chloride. This preparation contains a small proportion of truncated prorenin (26). Amino acids at positions 9-14 are reported to be needed to keep prorenin inactive (26). We found previously that the source of recombinant human prorenin synthesized by Carilli et al. (27) had undetectable basal activity after incubation at 37 "C (14). This same preparation had 5 1 0 % activity when tested several months later in this study, perhaps because of bacterial contamination of the buffers.
Cryoactivation of prorenin occurred slowly over days. Inactivation of prorenin, by comparison, occurred quickly (8 min at 37 "C and 180 min at room temperature). The difference in the time course of activation versus inactivation may be due to several factors. First, the greater kinetic energy at higher temperatures would favor a more rapid change in threedimensional structure. Second, based on pepsinogen as a model, activation of prorenin may occur in two stages (34). The prosegment of pepsinogen sits in, or near, the active cleft

Reversible Cryoactivation
of Human Prorenin of the enzyme (35). When activated by acid pH, pepsinogen's prosegment moves away from the body of the protein and then the N-terminal amino acid residues of pepsin move to become part of a six-strand anti-parallel / 3 sheet (34, 36)) common to renin and other aspartyl proteases. Inactivation may occur by reversal of either of these two steps and thereby could be faster. The prosegments of aspartic protease zymogens are rich in basic amino acids (37, 38), which may interact with carboxylate groups of the enzyme, holding the prosegment in the active site (38). At acid pH protonation of the carboxylate groups releases the prosegment of pepsin (37,38). At neutral pH a lysine residue at position 36 of the prosegment of pepsin, a highly conserved region of aspartic proteases, is hydrogenbonded to 1 of the aspartic acid residues in the active site and is stabilized by other hydrogen bonds and dipole interactions (38). Amino acids P9-13 of pepsinogen (corresponding to amino acids P9-14 of prorenin) form a hydrophobic core that fits into the substrate binding site (38). Hydrophobic interactions, a major force in protein folding, are disrupted by low temperatures which perhaps explains why prorenin can be cryoactivated (39). High ionic strength strengthens hydrophobic interactions, and this may explain the lack of cryoactivation of human prorenin in the presence of 3 M sodium chloride.
The direct IRMA for human renin uses a monoclonal antibody directed against renin's active site. We found that cryoactivated prorenin is detected in this assay, suggesting that the prosegment has moved away from the active site, making it accessible to the antibody. Movement of either the N-terminal amino acid residues of renin or of the prosegment could result in prorenin that is inactive yet not in its native three-dimensional state. This improperly folded prorenin may retain its antigenic site and bind to the monoclonal antibody used to detect renin. The current studies do not support that concept, since the direct IRMA detected a lower proportion of the cryoactivated prorenin than the enzyme kinetic assay, perhaps because the assay was carried out at 4 "C rather than 0 "C.
Cryoactivation of prorenin was first discovered in human plasma (40, 41). Human plasma stored cold but not frozen will develop higher renin levels with time. This increase is irreversible, because plasma serine protease inhibitors are inactivated in the cold allowing activation of the enzymes of the coagulation and fibrinolytic pathways (22). Both kallikrein and plasmin are capable of cleaving the prosegment of prorenin (5). It is likely that cryoactivation initially is reversible in plasma. As the prosegment moves away from the active site, however, prorenin may be more susceptible to enzymatic cleavage, as is pepsinogen when its prosegment shifts away from its active site (36). We have shown that reversibly acidactivated prorenin is more susceptible to cleavage by kallikrein than inactive prorenin.' Demonstrating that prorenin can have intrinsic activity at neutral pH suggests that prorenin could have intrinsic renin activity in uiuo under certain circumstances. For example, the prosegment may preferentially bind to other proteins rather than to the main body of the renin molecule. Such an interaction might lead to exposure of prorenin's active site at localized target sites. This could allow angiotensin to be formed locally.
Elevated plasma prorenin has been reported to predict microvascular complications in Type I diabetes mellitus (42). Whether the elevated prorenin levels of diabetes are just markers of disease or whether they play an active role in the disease remains to be established. The fact that prorenin can T. M. Pitarresi, S. A. Atlas, and J. E. Sealey, unpublished data. have intrinsic catalytic activity in vitro makes it possible that prorenin may have the capacity to behave like renin in vivo without cleavage of the prosegment. This could localize angiotensin production to a particular target site and might allow the local accumulation of angiotensin I1 in higher concentrations than that produced by circulating renin. High concentrations of angiotensin I1 can cause vasodilation, not vasoconstriction, which may explain the association of prorenin with increased blood flow and vasodilation (17). The reversibility of activation of prorenin may be the key to understanding its role in uiuo.