- Email: [email protected]
Disproportional release of differently glycosylated forms of human renin by furosemide
Life Sciences, Vol. 47, pp. 1903-1913 Printed in the U.S.A.
Pergamon Press
DISPROPORTIONAL RELEASE OF DIFFERENTLY GLYCOSYLATED FORMS OF HUMAN RENIN BY FUROSEMIDE Masayukl Hosol, Shekel Kim, Fumlhlko Ikemoto, and KenJlro Yamamoto Department of Pharmacology, Osaka City Unlverslty Medical School, Abeno, Osaka 545, Japan
(Received in final form September 14, 1990) Summary Concanavalln A (con A) chromatography of human plasma revealed the presence of three differently glycosylated forms of active renln(AR) and prorenln(PR), including the con A unbound forms(AR-I and PR-I), the loosely-bound forms(AR-II and PR-II), and the tightly-bound forms(AR-III and PR-III). These three forms of AR and PR were observed in human renal extracts. Normal male volunteers were Intravenously given the diuretic furosemlde(20 mg), kept standing for one hr and the effect on each form of renln was examined. These treatments elevated the plasma concentrations of AR-I, II and III by 2.1 f 0.2, 2.6 * 0.5, and 6.3 f l.l-fold, respectlvely(n=lZ), thereby lndlcating that the increase in AR-III was slgnlflcantly larger than that in the other two forms(P< 0.01). This dlsproportlonal increase was accompanied by a slgnlflcant Increase in the relative percent of AR-III in plasma from 21.6 f 2.5 to 42.2 f 3.0 %(P
0024-3205190 $3.00 + -00 Copyright (c) 1990 Pergamon Press plc
1904
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
Using con A chromatography we have now obtalned evidence for the presence of differently glycosylated forms of active renln and prorenln in human plasma and the dlsproportlonal release of each form of renln by acute renln stlmulatlon with furosemlde. Materials and Methods Concanavalln A (con A)-Sepharose 4B was from Pharmacla Fine Chemlcals(Plscataway, NJ). Ethylenedlamlnetetraacetate(EDTA), 8hydroxyqulnollne, and aprotlnln were from Wako Pure Chemicals (Osaka, Japan). Benzamldlne HCl was from Tokyo Kasel Kogyo (Tokyo,Japan). Phenylmethylsulfonyl cr-methylglucoslde and fluoride (PMSF), dllsopropylfluorophosphate(DFP), a-methylmannoslde were from Sigma Chemlcal(St.Louls, MO). Bovine serum albumin (BSA) was from Selkagaku Kogyo(Tokyo, Japan). Pepstatln and H-77 were obtained from Peptlde Institute Inc.(Osaka,Japan). Furosemlde was from Hoechst(Tokyo, Japan). International Reference Standard Preparation of Human Renln 68/356 was from the National Institute of Blologlcal Standards and Controls (Holly Hill, London). Protein A-Bacterial Adsorbent was from BloMakor (Rehovot,Israel). Preparation of pepstatln and H-77 affinity columns. Pepstatln was coupled to amlnohexyl-Sepharose, as described by Murakaml and Inagaml(l0). An octapeptlde renln lnhlbltor H-77(11) was coupled to activated 6-amlnohexanolc acldSepharose, according to the method of McIntyre et al(12). Sublects. Twelve normal men, aged 22 to 31 yr, were studied. They were allowed a diet with unrestricted sodium. Informed consent for the test was obtanned from all these volunteers. Con A chromatography. Before the appllcatlon to a con A column, all samples of olasma. renal extracts and column elutlons were adlusted to pH 7.2 with 0.2 M Tr1.s or O.lN HCl and CaC12 was added at a final concentration of 1.0 mM. The samples were then applied to a con A-Sepharose column(0.76 x 2.2 cm ) equlllbrated with 50 mM Trls-HCl buffer (pH7.2) contalnlng 0.2 M NaCl, 4 mM benzamldlne HCl, 0.1 mM PMSF and 100 nM DFP, to separate renln, on the basis of The elutlon was the difference In carbohydrate structure, as reported(8,9). performed with each 10 ml of the equlllbratlon buffer, the buffer contalnlng 10 mM a-methylglucoslde and 200 mM a-methylmannoslde, in a stepwlse manner(8,9). Flow rate was 10 ml/h. Fractions of 1 ml were collected into plastic tubes to which had been added 100 nl of the equlllbratlon buffer contalnlng 0.5% BSA at a final concentration. All chromatographlc procedures were performed at 4'C. After the subJects had remalned Acute renln stlmulatlon with furosemlde. recumbent for 1 hr. a blood samnle(20 ml) was drawn from an antecubltal vein for measurement of each form of 'r-e&r. Then the dluretlc furosemlde (20 mg) was intravenously inJected and the subjects were kept standlng for one hr, after which another blood sample(20 ml) was taken for measurement of each form Blood samples were collected Into chilled tubes contalnlng Na2 of renin. -EDTA(lmg/ml blood), and the plasma was rapidly separated by centrlfugatlon at 3,000 rpm at 4'C for 20 mln and stored at -3O'C until use. Measurement of differently glycosylated forms of renln In plasma. To measure each form of active renln in the plasma, crude plasma from lndlvldual SubJects For measurement of each form of lnactlve renln, the active renlnwas used. In free plasma was prepared by using pepstatln-afflnlty chromatography(l0). brief, 1.5 ml plasma samples from lndlvldual SubJects, adJusted to pH 6.5 with 0.2 M Trls-acetate(pH 4.1), were applied to a pepstatln-amlnohexyl-Sepharose column(0.5 x 2 cm) equlllbrated with 10 mM sodium acetate buffer, pH 6.5, containing 2 mM EDTA, 0.4 mM PMSF, 0.4 mM 8-hydroxyqulnollne, 4 mM benzamldlne HCl and the column was washed with 0.8 ml of the equlllbratlon buffer. The col-
Vol. 47, No. 21, 1990
Glycosylation of Human Renin
1905
lected breakthrough fraction, which contalned more than 90% of the applied Inactive renln and no detectable active renin, was used as the active renin-free plasma. Crude plasma samples(0.5 ml) and the active renin-free plasma samples(l.O ml) were subJected to con A chromatography, as described above. The fractions contalnlng each form of active or lnactlve renln were pooled and dialyzed against 10 mM sodium phosphate buffer(pH 7.4), contalnlng 5 mM EDTA, 5 mM benfor 12hr at 4"C, the zamldlne HCl, O.lmM PMSF, O.lmM 8-hydroxyqulnollne recovery of which was 112.4 f 15.9 and 102.8 f 8.2 % for active renln and inactive renin, respectively. The dialyzed samples were lyophlllzed, the resulting powder was resolved with 1 ml of 50 mM sodium phosphate, pH 7.4, contalnlng 5mM EDTA, 5mM benzamldlne, 0.1 mM PMSF, 0.5% BSA and 0.1% sodium azlde, and were measured for renln activity. In preliminary experiments, we examined the effect of lyophlllzatlon on the renin activity, using pure human renal renln prepared by ammonium sulfate precipitation, pepstatln and H-77 affinity chromatographles, according to the method of Do et al(13). We found that after lyophlllzatlon the enzymatic actlvlty of pure renln was almost completely retalned(99.1 + 1.3%, n=5), thereby indicating that the loss of renin activity by lyophlllzatlon was negligible. Preparation of human renal active and inactive renlns. Human cadaver kidneys with no history of renal disease were obtained at autopsy, lmmedlately frozen and kept at -90°C until use. The renal cortlces(1 g) were homogenized with 2 ml of 50 mM Trls-acetate buffer, pH7.4, contalnlng 1mM EDTA, 0.2 mM PMSF, 4 mM benzamldlne, 100 uM DFP, 4 mM potassium tetrathlonate, 4 mM N-ethylmalelmlde, and 40,000 kalllkreln lnhlbltor units / 1 aprotlnln, centrifuged at 15,000 r.p.m. for 30 min. To separate active and lnactlve renlns, the supernatant(1 ml) was applied to an affinity column(0.5 X 2 cm) of H-77-Sepharose(l2), equlllbrated with the above buffer. Inactive renln was recovered in the breakthrough fraction, After washing the column with the equlllbratlon buffer, active renln was eluted with 0.1 M acetlc acld(pH 3.5) containing the above protease inhlbltors, and fractions containing active renln were lmmedlately adJusted to pH 7.2 with 0.2 M Trls. To search for the three forms of active and inactive renln in the kidney, each solution containing 75 mllllGoldblatt unlts(mGu) of active or inactive renin was subJected to con A chromatography, as described above. Measurement of renln activity. Renln actlvlty was measured as the rate of anglotensln I(A1) formation, as reported(14) In brief, the samples of plasma, renal extracts and chromatographlc fractions were incubated with nephrectomlzed sheep plasma(flna1 concentration of anglotenslnogen, 0.6 uM ) in O.lM sodium phosphate buffer(pH 7.6 ), containing 10 mM EDTA, 10 mM benzamldlne HCl, 1 mM 8-hydroxyqulnollne, 5 mM PMSF, 50 pM DFP at 37°C for 30 mln to 6h. During this lncubatlon period, the generation of AI was linear as a function of time. The generated AI was measured by radlolmmunoassay(l5). In the above assay system, 1 Gu of human standard renln produced 131 ug AI/hr. Activation of inactive renln. For measurement of total renln, inactive renln was activated with trypsln lmmoblllzed on CNBr activated Sepharose 4B, according to the method of Derkx et a1.(16). In brief, the samples of plasma and chromatographlc fractions were incubated with trypsln-Sepharose suspension in O.lM sodium phosphate buffer (pH 7.5) containing 76 mM NaCl, 1mM EDTA with gentle agitation at 4'C for 18hr. The final trypsln concentration was 0.5 and 0.25 mg/ml for plasma and chromatographlc fractions, respectively, the optimal condltlon for the actlvatlon of inactive renln noted in preliminary experiments. Following the lncubatlon, the samples were centrifuged and the renln actlvlty in the supernatant was measured, as described above. Inactive
1906
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
renin was calculated as the difference between total and active renin. Immunological identification of active renin as true renin and inactive renin as prorenin. To examine whether the renin activity was due to true renin, the sample(280 nl) containing 5 JJGU of each form of plasma active renin was incubated with 20 nl of rabbit specific anti-recombinant human renin antlserum(l7, 18) or pre-immunized serum as a control for 18 hr at 4'C. After the lncubatlon, the residual renln activity was measured, as described above. For the identification of inactive renln as prorenln, the sample(l80 nl) containing 7.5 pGu of each form of plasma inactive renin was incubated with 20 nl of rabbit specific antiserum raised against the 43-amino acid prosegment of human prorenln(l8) or pre-immunized serum as a control for 18 hr at 4'C. Following the incubation, 150 nl of Protein A-Bacterial Adsorbent, in an amount sufficient to precipitate the included immunoglobulin G, was added and another After incubation was carried out with gentle agitation at 4°C for 18 hr. centrifugation, the residual inactive renin in the supernatant was measured after activation with trypsln, as described above. Statistical analysis. Values are expressed as means f SE. Student's t-test the means of plasma active and inactive renin was used to compare concentrations, and the relative percent of each form of renin between before and after furosemide administration. Analysis of variance followed by the Tukey test was used to compare the means of relative proportion among the three forms of active or inactive renin, and the means of increase in plasma concentrations by furosemide administration among the three forms of active or inactive renin. Results Chromatography of human plasma on con A-Sepharose. As shown in Fig. 1, con A chromatography of plasma from a normal human volunteer standing for several hours showed that both active renin(AR) and inactive renin(PR) consisted of three forms, including the con A unbound forms(AR-I and PR-I), the con Aloosely bound forms(AR-II and PR-II) eluted with 10 mM a-methylglucoside,and the con A tightly-bound forms(AR-III and PR-III) eluted with 200 mM a-methylmannoside. Rechromatography of each form of the separated renin on con A-Sepharose revealed that more than 90% of the re-applied renin was eluted with the same buffer as previously eluted, for all forms, findings confirming that they indeed had a different affinity to con A. Furthermore, con A chromatography of the active renin-free plasma showed that there was no detectable active renln in any fraction from the column, a finding indicative of a negligible conversion of inactive to active renin during the chromatography. The AIImmunological identification of each form of renin as true renin. generating activity of plasma AR-I,11 and III was inhibited by prior incubation human renin antiserum(100, 100 and 98% with specific anti-recombinant inhibition, respectively; n=2), thereby indicating that these forms were true renin. When plasma PR-I, II and III(each 7.5 pGu) were incubated with antiprorenin antiserum followed by the precipitation of the immunocomplex with Protein A-Bacterial Adsorbent, no inactive renin was detected in the supernatant(n=2), showing that they all represented prorenin, findings in agreement with the observations by Higashimori et a1.(19).
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
200
IOmM
1907
mM
a-MM
(r-MG
0.6
0.3
0
0
10
20
Fraction
30
number
FIG. 1 Con A chromatography of active renln and prorenln in plasma By concanavalln A normal volunteer standing for several hours. chromatography, human active renln and prorenln(lnactlve renln) could be separated into three forms, lncludzng the con A unbound forms( I; AR-I and PR-I), the con A loosely-bound forms(I1; AR-II and PR-II), and the con A tlghtlyeluted with 10 mM a-methylglucoslde(a-MG) AR-III and PR-III ) eluted with 200 mM bound forms (III; a-methylmannoslde(n-MM).
Effects of furosemlde admlnlstratlon on plasma concentrations of each form of renln. Table I shows that furosemlde admlnlstratlon then 1 hr of standlng increased plasma active renln concentration by about 4-fold, while the increase in plasma prorenln concentration was sllght(n=12).
TABLE I Effect of Furosemlde Treatment on Plasma Renln
Recumbency Furosemlde
Active Renln (ng AI/h/ml)
Prorenln (ng AI/h/ml)
2.1 k 0.3 7.9 f 1.6"
21.8 f 2.6 29.1 + 2.6"
Values are means f SE(n=12).
AI, anglotensln I. * P
1908
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
As shown In Fig. 2, the acute renln stlmulatlon increased the plasma concentrations of AR-I, II and III by 2.1 f 0.2, 2.6 + 0.5, and 6.3 f l.l-fold, respectively, thereby lndlcatlng that the increase In AR-III was slgnlflcantly larger than that in the other two forms(P < 0.01 ). As shown in Fig. 3, this dlsproportlonal increase was accompanied by a slgnlflcant increase in the relative percent of AR-III in plasma from 21.6 + 2.5 to 42.2 f 3.0 %(P< 0.05) and a decrease In that of AR-I from 40.0 k 4.3 to 28.8 f 2.1 %(P
I-
*
FI
I -
I -
, -
N.S. -I-
I-
I Actwe
renm
Prorenln
FIG. 2 Increase In each form of plasma active renln and prorenln by The value of each form of active renln furosemlde admlnlstratlon. and prorenln In plasma before furosemlde admlnlstratlon was regarded as 1. Each bar represents mean f SE. P
Con A chromatography of human renal renln. Fig. 4 shows that the three forms of AR and PR In the plasma were also present in the human renal extracts. The relative percent of AR-I, II and III was 8.2 k4.3, 15.7 f 2.1, and 74.0 f 5.5 %, respectively, and that of PR-I, II and III was 17.5 f 5.4, 34.9 f 1.6 and 46.8 f 6.3 %, respectlvely(n=3). Thus, the major form was the con A tlghtlybound form for both AR and PR.
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
1909
Discussion Renin is a heterogeneous enzyme with respect to the molecular weight, isoelectric point, and enzymatic activity(20, 21). The present work revealed the heterogeneity of human renin, with respect to glycosylation.
(*I
80
P 6o > ._ z z 40 ._ 5 a &? 20
I
Recumbency
Ea
Furosemide
pi005
PC005
q
AR- I
AR- II
AR-III
PC005
@)
80 r
-r
.r” > 6o
PC005
3
Y
E 4o ._ C
dc
PR- I
PR- II
PR- III
FIG. 3 Relative percent of each form of plasma active renin(A) and The prorenin(B) before and after furosemide administration(n=lZ). value of each form was expressed as the percentage of total active Recovery of active renin or prorenin recovered from the column. renin and prorenin from the con A column was 70-80X,in all samples. Each value represents mean f SE.
Using highly purified
251-labeled rat renal renin, we demonstrated that
1910
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
circulating renln 1s mainly taken up by the liver(4-7) and that the carbohydrate portion of renln is essential for the hepatic uptake of circulating renin(7). Using a con A column with a high affinity for the high-mannose and hybrid types of glycoprotein, and a low affinity for the biantennary complex type of glycoprotein, but no apparent affinity for the tri and tetraantennary types of glycoprotein(22, 23), we found that three differently glycosylated forms of renin, including the con A unbound, loosely-bound and tightly-bound forms, are released from the rat kidney into the blood circulation(8,9) and they have a different half life in the plasma(g). Chronic stimulation of renin with 2 wk of sodium depletion and captopril treatment led to the preferential release of the con A unbound form of renin, with the longest half life, which contributes to the increase in plasma renin concentrations(9). Thus, in rats, the use of con A chromatography facilitated demonstration of the importance of glycosylation for renln secretion as well as the regulation of concentrations of renin in the plasma. However, the structure and physiological role of the carbohydrate portion of human renin remained to be elucidated.
200mM ru-MM
1 OmM (Y-MG
30
40
20
Fraction FIG.
number 4
Con A chromatography of active renin and prorenin in human renal extracts. Each form of active renin and prorenin was present in the I, AR-I and PR-I; II, AR-II and PR-II; III, ARrenal cortex. III and PR-III.
In the present study, con A chromatography of human plasma showed that both active renin and prorenin consisted of three forms, including the con A unbound form, the loosely-bound form and the tightly-bound form(Fig. l), thereby indicating the heterogeneity of glycosylation of human renin, as is the case with As shown in Fig. 4, con A chromatography of human renal exrat renin(8,9).
Vol. 47, No. 21, 1990
Glycosylation of Human Renin
1911
tracts revealed the presence of all these forms of active renin and prorenin in the kidney. These results, taken together with the fact that the kidney is the predominant source of circulating active renin(24, 25), indicate that these multiple forms of active renln are synthesized within the kidney and secreted into the blood circulation. On the other hand, it is unclear whether all forms of prorenin are derived from the kidney, because of the existence of extrarenal sources of circulating prorenin(24, 25). followed by one hr of As shown in Fig. 2, furosemide administratzon standing, the method widely used to stimulate renin release(26, 27, 28), disproportionally elevated each form of plasma active renin. The increase in plasma AR-111(6.3-fold) was significantly larger than that in AR-1(2.1-fold) and AR-11(2.6-fold)(P
1912
Glycosylationof Human Renin
Vol. 47, No. 21, 1990
for the gift of anti-recombinant human renln antiserum and anti-prorenln prosegmentantlserum,and M. Ohara for pertinent comments, References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21.
22. 23. 24. 25. 26. 27. 28. 29. 30.
I.A. REID, HypertensionResearch, F.M.RADZIALOWSKI.(ed),lOl-137,Dekker, New York (1983) C.M.WEST, Mol.Cell.Blochem.72: 3-20(1986) G.ASHWELL, and J.HARFORD, AnG.Rev.Blochem. 51: 531-554(1982) S.KIM, H.IWAO, N.NAKAMURA, F.IKEMOTO and K.YAMAMOTO, Am.J. Physlol. 252: E136-El46 (1987) S.KIM, N.NAKAMURA, H.IWAO, F.IKEMOTO and K.YAMAMOTO J.Hlstochem,Cytochem.35: 549-557(1987) S. KIM, H.IWAO, N.NAGURA, F.IKEMOTO, and K.YAMAMOTO,Am. J.Physlo1.253: E621-E628(1987) S.KIM, M.HIRUMA, F.IKEMOTO and K.YAMAMOTO, Am.J.Physlol. 255: E642E651(1988) S.KIM, M.HOSOI, M.HIRUMA, F.IKEMOTO, and K.YAMAMOTO, Clln.Exp.Hypertens. AlO(6): 1203-1211(1988) S.KIM, M.HOSOI, M.HIRUMA, F.IKEMOTO. and K.YAMAMOTO, Am.J.Physlol. 256: E798-E804(1989) K.MURAKAMI,and T.INAGAMI, Blochem.Blophys.Res.Commum. 2: 757-763 (1975) M.SZELKE, B.J.LECKIE, M.TREE, A.BROWN, J.GRANT, A.HALLETT, A.HUGHES, D.M.JONES, and A.F.LEVER,Hypertension.4(Supp12):59-69(1982) G.D.MCINTYRE, B.LECKIE, A.HALLETT, and M.SZELKE B1ochem.J. -211: 519522(1983) Y.S.DO, T.SHINAGAWA, H.TAM, T.INAGAMI, and W.A.HSUEH, J.Blol.Chem. -262: 1037-1043(1987) K.TAKAORI, T.TAKEDA, H.YAMADA, K.MIURA, T.YUKIMURA, F.IKEMOTO, and K.YAMAMOTO,Jpn.Clrc.J.52: 1346-1350(1988) E.HABER, T.KOERNERT L.B.PAGE, B.KLIMAN, and A.PURNODE, J.Clln.Endocrlnol.Metab. 29: 1349-1355(1969) F.H.M.DERKX, L.TAN-TJIONG, G.J.WENTING, F.BOOMSMA, A.J.M.VELD, and M.A.D.H.SCHALEKAMP,Hypertension5: 244-256 (1983) M.HIRUMA, S.KIM, F.IKEMOTO, K.MURAKAMI and K.YAMAMOTO, Hypertension-12: 317-323 (1988) S.KIM, M.HOSOI, F.IKEMOTO, K.MURAKAMI, Y.ISHIZUKA,and K.YAMAMOTO, Am.J.Physlol.258: E451-E458(1990) K.HIGASHIMORI,K.MIZUNO, S.NAKAJO, F.H.BPEHN, P.A.MARCOTTE, D.A. EGAN, W.H.HOLLEMEN, C.HEUSSER, A.M.POISONER, and T.INAGAMI, J.Blol.Chem. -264: 14662-14667(1989) J-J.CHANG, M.KISARAGI, H.OKAMOTO, and T.INAGAMI, Hypertension. 3: 509515(1981) P.CORVOL, F.X.GALEN, C.DEVAUX, and J.MENARD, Hypertension, J.GENEST, O.KUCHEL, P.HAMET, M.CANTIN(eds), 184-194, McGraw-Hill Book Company, New York(1983) R.D.CUMMINGS,and S.KORNFELD,J.Blol.Chem.-257: 11235-11240(1982) S.OGATA, T.MURAMATU,and A.KOBATA. J.Blochem. 78: 687-696(1975) W.A.HSUEH, E.J.CARLSON,and V.J.DZAU, J.Clln.IGest. 71: 506-517(1983) J.E.SEALEY and S.RUBAITLl, Am.J.Hypertens._?:358-366(s89) N.M.KAPLAN, D.C.KEM, O.B.HOLLAND, N.J.KRAMER, J.HIGGINS, and C.GOMEZSANCHEZ, Ann.Inter.Med.84: 639-645(1976) M.BRYER-ASH, E.B.FRAZE, and J.A.LUETSCHER, J.Clln.Endocrlnol.Metab.-66: 454-458(1988) J.HIGAKI, T.OGIHARA, and Y.KUMAHARA,Clin.Sci. 75: 669-672(1988) F.X.GALEN, C.DEVAUX, A.M.HOUOT, J.MENARD: P.CORVOL, M.T.CORVOL, M.C.GUBLER, F.MOUNIER,and J.P.CAMILLERI,J.Clln.Invest.73:1144-1155(1984) M.SCHAMBELAN, E.B.TOFFELMIRE, K.SLATER, P.CORVOL, J.MENARD, az J.Clln.Invest.83: 679-687(1989)
Vol. 47, No. 21, 1990
Glycosylationof Human Renin
1913
31. M.PAUL, N.NAKAMLJRA,R.E.PRATI',and V.J.DZAU, J.Hypertens. 6(suppl 4): S487-S489(1988) 32. H.HORI, T.YOSHINO, Y.ISHIZUKA, T.YAMAUCHI, and K.MURAKAMI. FEB. -232: 391394(1988) 33. E.D.GREEN, I.BOIME, and J.U.BAENZINGER.Mol.Cell.Blochem.72: 81-lOO(1986) J.ROBBINS, A.PALEAROTTI, 34. L.BARTALENA, F.MARTINOand PINCHERA, Endocrinology.-112: 1479-1482(1987)
Pergamon Press
DISPROPORTIONAL RELEASE OF DIFFERENTLY GLYCOSYLATED FORMS OF HUMAN RENIN BY FUROSEMIDE Masayukl Hosol, Shekel Kim, Fumlhlko Ikemoto, and KenJlro Yamamoto Department of Pharmacology, Osaka City Unlverslty Medical School, Abeno, Osaka 545, Japan
(Received in final form September 14, 1990) Summary Concanavalln A (con A) chromatography of human plasma revealed the presence of three differently glycosylated forms of active renln(AR) and prorenln(PR), including the con A unbound forms(AR-I and PR-I), the loosely-bound forms(AR-II and PR-II), and the tightly-bound forms(AR-III and PR-III). These three forms of AR and PR were observed in human renal extracts. Normal male volunteers were Intravenously given the diuretic furosemlde(20 mg), kept standing for one hr and the effect on each form of renln was examined. These treatments elevated the plasma concentrations of AR-I, II and III by 2.1 f 0.2, 2.6 * 0.5, and 6.3 f l.l-fold, respectlvely(n=lZ), thereby lndlcating that the increase in AR-III was slgnlflcantly larger than that in the other two forms(P< 0.01). This dlsproportlonal increase was accompanied by a slgnlflcant Increase in the relative percent of AR-III in plasma from 21.6 f 2.5 to 42.2 f 3.0 %(P
0024-3205190 $3.00 + -00 Copyright (c) 1990 Pergamon Press plc
1904
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
Using con A chromatography we have now obtalned evidence for the presence of differently glycosylated forms of active renln and prorenln in human plasma and the dlsproportlonal release of each form of renln by acute renln stlmulatlon with furosemlde. Materials and Methods Concanavalln A (con A)-Sepharose 4B was from Pharmacla Fine Chemlcals(Plscataway, NJ). Ethylenedlamlnetetraacetate(EDTA), 8hydroxyqulnollne, and aprotlnln were from Wako Pure Chemicals (Osaka, Japan). Benzamldlne HCl was from Tokyo Kasel Kogyo (Tokyo,Japan). Phenylmethylsulfonyl cr-methylglucoslde and fluoride (PMSF), dllsopropylfluorophosphate(DFP), a-methylmannoslde were from Sigma Chemlcal(St.Louls, MO). Bovine serum albumin (BSA) was from Selkagaku Kogyo(Tokyo, Japan). Pepstatln and H-77 were obtained from Peptlde Institute Inc.(Osaka,Japan). Furosemlde was from Hoechst(Tokyo, Japan). International Reference Standard Preparation of Human Renln 68/356 was from the National Institute of Blologlcal Standards and Controls (Holly Hill, London). Protein A-Bacterial Adsorbent was from BloMakor (Rehovot,Israel). Preparation of pepstatln and H-77 affinity columns. Pepstatln was coupled to amlnohexyl-Sepharose, as described by Murakaml and Inagaml(l0). An octapeptlde renln lnhlbltor H-77(11) was coupled to activated 6-amlnohexanolc acldSepharose, according to the method of McIntyre et al(12). Sublects. Twelve normal men, aged 22 to 31 yr, were studied. They were allowed a diet with unrestricted sodium. Informed consent for the test was obtanned from all these volunteers. Con A chromatography. Before the appllcatlon to a con A column, all samples of olasma. renal extracts and column elutlons were adlusted to pH 7.2 with 0.2 M Tr1.s or O.lN HCl and CaC12 was added at a final concentration of 1.0 mM. The samples were then applied to a con A-Sepharose column(0.76 x 2.2 cm ) equlllbrated with 50 mM Trls-HCl buffer (pH7.2) contalnlng 0.2 M NaCl, 4 mM benzamldlne HCl, 0.1 mM PMSF and 100 nM DFP, to separate renln, on the basis of The elutlon was the difference In carbohydrate structure, as reported(8,9). performed with each 10 ml of the equlllbratlon buffer, the buffer contalnlng 10 mM a-methylglucoslde and 200 mM a-methylmannoslde, in a stepwlse manner(8,9). Flow rate was 10 ml/h. Fractions of 1 ml were collected into plastic tubes to which had been added 100 nl of the equlllbratlon buffer contalnlng 0.5% BSA at a final concentration. All chromatographlc procedures were performed at 4'C. After the subJects had remalned Acute renln stlmulatlon with furosemlde. recumbent for 1 hr. a blood samnle(20 ml) was drawn from an antecubltal vein for measurement of each form of 'r-e&r. Then the dluretlc furosemlde (20 mg) was intravenously inJected and the subjects were kept standlng for one hr, after which another blood sample(20 ml) was taken for measurement of each form Blood samples were collected Into chilled tubes contalnlng Na2 of renin. -EDTA(lmg/ml blood), and the plasma was rapidly separated by centrlfugatlon at 3,000 rpm at 4'C for 20 mln and stored at -3O'C until use. Measurement of differently glycosylated forms of renln In plasma. To measure each form of active renln in the plasma, crude plasma from lndlvldual SubJects For measurement of each form of lnactlve renln, the active renlnwas used. In free plasma was prepared by using pepstatln-afflnlty chromatography(l0). brief, 1.5 ml plasma samples from lndlvldual SubJects, adJusted to pH 6.5 with 0.2 M Trls-acetate(pH 4.1), were applied to a pepstatln-amlnohexyl-Sepharose column(0.5 x 2 cm) equlllbrated with 10 mM sodium acetate buffer, pH 6.5, containing 2 mM EDTA, 0.4 mM PMSF, 0.4 mM 8-hydroxyqulnollne, 4 mM benzamldlne HCl and the column was washed with 0.8 ml of the equlllbratlon buffer. The col-
Vol. 47, No. 21, 1990
Glycosylation of Human Renin
1905
lected breakthrough fraction, which contalned more than 90% of the applied Inactive renln and no detectable active renin, was used as the active renin-free plasma. Crude plasma samples(0.5 ml) and the active renin-free plasma samples(l.O ml) were subJected to con A chromatography, as described above. The fractions contalnlng each form of active or lnactlve renln were pooled and dialyzed against 10 mM sodium phosphate buffer(pH 7.4), contalnlng 5 mM EDTA, 5 mM benfor 12hr at 4"C, the zamldlne HCl, O.lmM PMSF, O.lmM 8-hydroxyqulnollne recovery of which was 112.4 f 15.9 and 102.8 f 8.2 % for active renln and inactive renin, respectively. The dialyzed samples were lyophlllzed, the resulting powder was resolved with 1 ml of 50 mM sodium phosphate, pH 7.4, contalnlng 5mM EDTA, 5mM benzamldlne, 0.1 mM PMSF, 0.5% BSA and 0.1% sodium azlde, and were measured for renln activity. In preliminary experiments, we examined the effect of lyophlllzatlon on the renin activity, using pure human renal renln prepared by ammonium sulfate precipitation, pepstatln and H-77 affinity chromatographles, according to the method of Do et al(13). We found that after lyophlllzatlon the enzymatic actlvlty of pure renln was almost completely retalned(99.1 + 1.3%, n=5), thereby indicating that the loss of renin activity by lyophlllzatlon was negligible. Preparation of human renal active and inactive renlns. Human cadaver kidneys with no history of renal disease were obtained at autopsy, lmmedlately frozen and kept at -90°C until use. The renal cortlces(1 g) were homogenized with 2 ml of 50 mM Trls-acetate buffer, pH7.4, contalnlng 1mM EDTA, 0.2 mM PMSF, 4 mM benzamldlne, 100 uM DFP, 4 mM potassium tetrathlonate, 4 mM N-ethylmalelmlde, and 40,000 kalllkreln lnhlbltor units / 1 aprotlnln, centrifuged at 15,000 r.p.m. for 30 min. To separate active and lnactlve renlns, the supernatant(1 ml) was applied to an affinity column(0.5 X 2 cm) of H-77-Sepharose(l2), equlllbrated with the above buffer. Inactive renln was recovered in the breakthrough fraction, After washing the column with the equlllbratlon buffer, active renln was eluted with 0.1 M acetlc acld(pH 3.5) containing the above protease inhlbltors, and fractions containing active renln were lmmedlately adJusted to pH 7.2 with 0.2 M Trls. To search for the three forms of active and inactive renln in the kidney, each solution containing 75 mllllGoldblatt unlts(mGu) of active or inactive renin was subJected to con A chromatography, as described above. Measurement of renln activity. Renln actlvlty was measured as the rate of anglotensln I(A1) formation, as reported(14) In brief, the samples of plasma, renal extracts and chromatographlc fractions were incubated with nephrectomlzed sheep plasma(flna1 concentration of anglotenslnogen, 0.6 uM ) in O.lM sodium phosphate buffer(pH 7.6 ), containing 10 mM EDTA, 10 mM benzamldlne HCl, 1 mM 8-hydroxyqulnollne, 5 mM PMSF, 50 pM DFP at 37°C for 30 mln to 6h. During this lncubatlon period, the generation of AI was linear as a function of time. The generated AI was measured by radlolmmunoassay(l5). In the above assay system, 1 Gu of human standard renln produced 131 ug AI/hr. Activation of inactive renln. For measurement of total renln, inactive renln was activated with trypsln lmmoblllzed on CNBr activated Sepharose 4B, according to the method of Derkx et a1.(16). In brief, the samples of plasma and chromatographlc fractions were incubated with trypsln-Sepharose suspension in O.lM sodium phosphate buffer (pH 7.5) containing 76 mM NaCl, 1mM EDTA with gentle agitation at 4'C for 18hr. The final trypsln concentration was 0.5 and 0.25 mg/ml for plasma and chromatographlc fractions, respectively, the optimal condltlon for the actlvatlon of inactive renln noted in preliminary experiments. Following the lncubatlon, the samples were centrifuged and the renln actlvlty in the supernatant was measured, as described above. Inactive
1906
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
renin was calculated as the difference between total and active renin. Immunological identification of active renin as true renin and inactive renin as prorenin. To examine whether the renin activity was due to true renin, the sample(280 nl) containing 5 JJGU of each form of plasma active renin was incubated with 20 nl of rabbit specific anti-recombinant human renin antlserum(l7, 18) or pre-immunized serum as a control for 18 hr at 4'C. After the lncubatlon, the residual renln activity was measured, as described above. For the identification of inactive renln as prorenln, the sample(l80 nl) containing 7.5 pGu of each form of plasma inactive renin was incubated with 20 nl of rabbit specific antiserum raised against the 43-amino acid prosegment of human prorenln(l8) or pre-immunized serum as a control for 18 hr at 4'C. Following the incubation, 150 nl of Protein A-Bacterial Adsorbent, in an amount sufficient to precipitate the included immunoglobulin G, was added and another After incubation was carried out with gentle agitation at 4°C for 18 hr. centrifugation, the residual inactive renin in the supernatant was measured after activation with trypsln, as described above. Statistical analysis. Values are expressed as means f SE. Student's t-test the means of plasma active and inactive renin was used to compare concentrations, and the relative percent of each form of renin between before and after furosemide administration. Analysis of variance followed by the Tukey test was used to compare the means of relative proportion among the three forms of active or inactive renin, and the means of increase in plasma concentrations by furosemide administration among the three forms of active or inactive renin. Results Chromatography of human plasma on con A-Sepharose. As shown in Fig. 1, con A chromatography of plasma from a normal human volunteer standing for several hours showed that both active renin(AR) and inactive renin(PR) consisted of three forms, including the con A unbound forms(AR-I and PR-I), the con Aloosely bound forms(AR-II and PR-II) eluted with 10 mM a-methylglucoside,and the con A tightly-bound forms(AR-III and PR-III) eluted with 200 mM a-methylmannoside. Rechromatography of each form of the separated renin on con A-Sepharose revealed that more than 90% of the re-applied renin was eluted with the same buffer as previously eluted, for all forms, findings confirming that they indeed had a different affinity to con A. Furthermore, con A chromatography of the active renin-free plasma showed that there was no detectable active renln in any fraction from the column, a finding indicative of a negligible conversion of inactive to active renin during the chromatography. The AIImmunological identification of each form of renin as true renin. generating activity of plasma AR-I,11 and III was inhibited by prior incubation human renin antiserum(100, 100 and 98% with specific anti-recombinant inhibition, respectively; n=2), thereby indicating that these forms were true renin. When plasma PR-I, II and III(each 7.5 pGu) were incubated with antiprorenin antiserum followed by the precipitation of the immunocomplex with Protein A-Bacterial Adsorbent, no inactive renin was detected in the supernatant(n=2), showing that they all represented prorenin, findings in agreement with the observations by Higashimori et a1.(19).
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
200
IOmM
1907
mM
a-MM
(r-MG
0.6
0.3
0
0
10
20
Fraction
30
number
FIG. 1 Con A chromatography of active renln and prorenln in plasma By concanavalln A normal volunteer standing for several hours. chromatography, human active renln and prorenln(lnactlve renln) could be separated into three forms, lncludzng the con A unbound forms( I; AR-I and PR-I), the con A loosely-bound forms(I1; AR-II and PR-II), and the con A tlghtlyeluted with 10 mM a-methylglucoslde(a-MG) AR-III and PR-III ) eluted with 200 mM bound forms (III; a-methylmannoslde(n-MM).
Effects of furosemlde admlnlstratlon on plasma concentrations of each form of renln. Table I shows that furosemlde admlnlstratlon then 1 hr of standlng increased plasma active renln concentration by about 4-fold, while the increase in plasma prorenln concentration was sllght(n=12).
TABLE I Effect of Furosemlde Treatment on Plasma Renln
Recumbency Furosemlde
Active Renln (ng AI/h/ml)
Prorenln (ng AI/h/ml)
2.1 k 0.3 7.9 f 1.6"
21.8 f 2.6 29.1 + 2.6"
Values are means f SE(n=12).
AI, anglotensln I. * P
1908
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
As shown In Fig. 2, the acute renln stlmulatlon increased the plasma concentrations of AR-I, II and III by 2.1 f 0.2, 2.6 + 0.5, and 6.3 f l.l-fold, respectively, thereby lndlcatlng that the increase In AR-III was slgnlflcantly larger than that in the other two forms(P < 0.01 ). As shown in Fig. 3, this dlsproportlonal increase was accompanied by a slgnlflcant increase in the relative percent of AR-III in plasma from 21.6 + 2.5 to 42.2 f 3.0 %(P< 0.05) and a decrease In that of AR-I from 40.0 k 4.3 to 28.8 f 2.1 %(P
I-
*
FI
I -
I -
, -
N.S. -I-
I-
I Actwe
renm
Prorenln
FIG. 2 Increase In each form of plasma active renln and prorenln by The value of each form of active renln furosemlde admlnlstratlon. and prorenln In plasma before furosemlde admlnlstratlon was regarded as 1. Each bar represents mean f SE. P
Con A chromatography of human renal renln. Fig. 4 shows that the three forms of AR and PR In the plasma were also present in the human renal extracts. The relative percent of AR-I, II and III was 8.2 k4.3, 15.7 f 2.1, and 74.0 f 5.5 %, respectively, and that of PR-I, II and III was 17.5 f 5.4, 34.9 f 1.6 and 46.8 f 6.3 %, respectlvely(n=3). Thus, the major form was the con A tlghtlybound form for both AR and PR.
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
1909
Discussion Renin is a heterogeneous enzyme with respect to the molecular weight, isoelectric point, and enzymatic activity(20, 21). The present work revealed the heterogeneity of human renin, with respect to glycosylation.
(*I
80
P 6o > ._ z z 40 ._ 5 a &? 20
I
Recumbency
Ea
Furosemide
pi005
PC005
q
AR- I
AR- II
AR-III
PC005
@)
80 r
-r
.r” > 6o
PC005
3
Y
E 4o ._ C
dc
PR- I
PR- II
PR- III
FIG. 3 Relative percent of each form of plasma active renin(A) and The prorenin(B) before and after furosemide administration(n=lZ). value of each form was expressed as the percentage of total active Recovery of active renin or prorenin recovered from the column. renin and prorenin from the con A column was 70-80X,in all samples. Each value represents mean f SE.
Using highly purified
251-labeled rat renal renin, we demonstrated that
1910
Glycosylation of Human Renin
Vol. 47, No. 21, 1990
circulating renln 1s mainly taken up by the liver(4-7) and that the carbohydrate portion of renln is essential for the hepatic uptake of circulating renin(7). Using a con A column with a high affinity for the high-mannose and hybrid types of glycoprotein, and a low affinity for the biantennary complex type of glycoprotein, but no apparent affinity for the tri and tetraantennary types of glycoprotein(22, 23), we found that three differently glycosylated forms of renin, including the con A unbound, loosely-bound and tightly-bound forms, are released from the rat kidney into the blood circulation(8,9) and they have a different half life in the plasma(g). Chronic stimulation of renin with 2 wk of sodium depletion and captopril treatment led to the preferential release of the con A unbound form of renin, with the longest half life, which contributes to the increase in plasma renin concentrations(9). Thus, in rats, the use of con A chromatography facilitated demonstration of the importance of glycosylation for renln secretion as well as the regulation of concentrations of renin in the plasma. However, the structure and physiological role of the carbohydrate portion of human renin remained to be elucidated.
200mM ru-MM
1 OmM (Y-MG
30
40
20
Fraction FIG.
number 4
Con A chromatography of active renin and prorenin in human renal extracts. Each form of active renin and prorenin was present in the I, AR-I and PR-I; II, AR-II and PR-II; III, ARrenal cortex. III and PR-III.
In the present study, con A chromatography of human plasma showed that both active renin and prorenin consisted of three forms, including the con A unbound form, the loosely-bound form and the tightly-bound form(Fig. l), thereby indicating the heterogeneity of glycosylation of human renin, as is the case with As shown in Fig. 4, con A chromatography of human renal exrat renin(8,9).
Vol. 47, No. 21, 1990
Glycosylation of Human Renin
1911
tracts revealed the presence of all these forms of active renin and prorenin in the kidney. These results, taken together with the fact that the kidney is the predominant source of circulating active renin(24, 25), indicate that these multiple forms of active renln are synthesized within the kidney and secreted into the blood circulation. On the other hand, it is unclear whether all forms of prorenin are derived from the kidney, because of the existence of extrarenal sources of circulating prorenin(24, 25). followed by one hr of As shown in Fig. 2, furosemide administratzon standing, the method widely used to stimulate renin release(26, 27, 28), disproportionally elevated each form of plasma active renin. The increase in plasma AR-111(6.3-fold) was significantly larger than that in AR-1(2.1-fold) and AR-11(2.6-fold)(P
1912
Glycosylationof Human Renin
Vol. 47, No. 21, 1990
for the gift of anti-recombinant human renln antiserum and anti-prorenln prosegmentantlserum,and M. Ohara for pertinent comments, References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.
20. 21.
22. 23. 24. 25. 26. 27. 28. 29. 30.
I.A. REID, HypertensionResearch, F.M.RADZIALOWSKI.(ed),lOl-137,Dekker, New York (1983) C.M.WEST, Mol.Cell.Blochem.72: 3-20(1986) G.ASHWELL, and J.HARFORD, AnG.Rev.Blochem. 51: 531-554(1982) S.KIM, H.IWAO, N.NAKAMURA, F.IKEMOTO and K.YAMAMOTO, Am.J. Physlol. 252: E136-El46 (1987) S.KIM, N.NAKAMURA, H.IWAO, F.IKEMOTO and K.YAMAMOTO J.Hlstochem,Cytochem.35: 549-557(1987) S. KIM, H.IWAO, N.NAGURA, F.IKEMOTO, and K.YAMAMOTO,Am. J.Physlo1.253: E621-E628(1987) S.KIM, M.HIRUMA, F.IKEMOTO and K.YAMAMOTO, Am.J.Physlol. 255: E642E651(1988) S.KIM, M.HOSOI, M.HIRUMA, F.IKEMOTO, and K.YAMAMOTO, Clln.Exp.Hypertens. AlO(6): 1203-1211(1988) S.KIM, M.HOSOI, M.HIRUMA, F.IKEMOTO. and K.YAMAMOTO, Am.J.Physlol. 256: E798-E804(1989) K.MURAKAMI,and T.INAGAMI, Blochem.Blophys.Res.Commum. 2: 757-763 (1975) M.SZELKE, B.J.LECKIE, M.TREE, A.BROWN, J.GRANT, A.HALLETT, A.HUGHES, D.M.JONES, and A.F.LEVER,Hypertension.4(Supp12):59-69(1982) G.D.MCINTYRE, B.LECKIE, A.HALLETT, and M.SZELKE B1ochem.J. -211: 519522(1983) Y.S.DO, T.SHINAGAWA, H.TAM, T.INAGAMI, and W.A.HSUEH, J.Blol.Chem. -262: 1037-1043(1987) K.TAKAORI, T.TAKEDA, H.YAMADA, K.MIURA, T.YUKIMURA, F.IKEMOTO, and K.YAMAMOTO,Jpn.Clrc.J.52: 1346-1350(1988) E.HABER, T.KOERNERT L.B.PAGE, B.KLIMAN, and A.PURNODE, J.Clln.Endocrlnol.Metab. 29: 1349-1355(1969) F.H.M.DERKX, L.TAN-TJIONG, G.J.WENTING, F.BOOMSMA, A.J.M.VELD, and M.A.D.H.SCHALEKAMP,Hypertension5: 244-256 (1983) M.HIRUMA, S.KIM, F.IKEMOTO, K.MURAKAMI and K.YAMAMOTO, Hypertension-12: 317-323 (1988) S.KIM, M.HOSOI, F.IKEMOTO, K.MURAKAMI, Y.ISHIZUKA,and K.YAMAMOTO, Am.J.Physlol.258: E451-E458(1990) K.HIGASHIMORI,K.MIZUNO, S.NAKAJO, F.H.BPEHN, P.A.MARCOTTE, D.A. EGAN, W.H.HOLLEMEN, C.HEUSSER, A.M.POISONER, and T.INAGAMI, J.Blol.Chem. -264: 14662-14667(1989) J-J.CHANG, M.KISARAGI, H.OKAMOTO, and T.INAGAMI, Hypertension. 3: 509515(1981) P.CORVOL, F.X.GALEN, C.DEVAUX, and J.MENARD, Hypertension, J.GENEST, O.KUCHEL, P.HAMET, M.CANTIN(eds), 184-194, McGraw-Hill Book Company, New York(1983) R.D.CUMMINGS,and S.KORNFELD,J.Blol.Chem.-257: 11235-11240(1982) S.OGATA, T.MURAMATU,and A.KOBATA. J.Blochem. 78: 687-696(1975) W.A.HSUEH, E.J.CARLSON,and V.J.DZAU, J.Clln.IGest. 71: 506-517(1983) J.E.SEALEY and S.RUBAITLl, Am.J.Hypertens._?:358-366(s89) N.M.KAPLAN, D.C.KEM, O.B.HOLLAND, N.J.KRAMER, J.HIGGINS, and C.GOMEZSANCHEZ, Ann.Inter.Med.84: 639-645(1976) M.BRYER-ASH, E.B.FRAZE, and J.A.LUETSCHER, J.Clln.Endocrlnol.Metab.-66: 454-458(1988) J.HIGAKI, T.OGIHARA, and Y.KUMAHARA,Clin.Sci. 75: 669-672(1988) F.X.GALEN, C.DEVAUX, A.M.HOUOT, J.MENARD: P.CORVOL, M.T.CORVOL, M.C.GUBLER, F.MOUNIER,and J.P.CAMILLERI,J.Clln.Invest.73:1144-1155(1984) M.SCHAMBELAN, E.B.TOFFELMIRE, K.SLATER, P.CORVOL, J.MENARD, az J.Clln.Invest.83: 679-687(1989)
Vol. 47, No. 21, 1990
Glycosylationof Human Renin
1913
31. M.PAUL, N.NAKAMLJRA,R.E.PRATI',and V.J.DZAU, J.Hypertens. 6(suppl 4): S487-S489(1988) 32. H.HORI, T.YOSHINO, Y.ISHIZUKA, T.YAMAUCHI, and K.MURAKAMI. FEB. -232: 391394(1988) 33. E.D.GREEN, I.BOIME, and J.U.BAENZINGER.Mol.Cell.Blochem.72: 81-lOO(1986) J.ROBBINS, A.PALEAROTTI, 34. L.BARTALENA, F.MARTINOand PINCHERA, Endocrinology.-112: 1479-1482(1987)