- Email: [email protected]
Angiotensin carboxypeptidase activity in urine from normal subjects and patients with kidney damage
Life Sciences, Vol. 48, pp. 1529-1535 Printed in the U.S.A.
Pergamon Press
ANGIOTENSIN CARBOXYPEPTIDASE ACTIVITY IN URINE FROM NORMAL SUBJECTS AND PATIENTS WITH K]DNEY DAMAGE James J. Miller*, David G. Changaris, and Robert S. Levy Departments of Biochemistry and Neurology and the Laboratory of Biological Psychiatry, University of Louisville School of Medicine, Louisville, KY 40292 (Received in final form February ii, 1991)
Summary Angiotensin carboxypeptidase (ACP) activity has been detected in urine samples from normal subjects and patients with hypertension and diabetes by determining the enzyme's ability to convert angiotensin I to des-Leu angiotensin I. Gel filtration chromatography of a concentrated urine sample indicated that about equal amounts of the enzyme exist as 100 kDa and 500 kDa molecular weight forms, respectively. This ACP activity co-eluted with activity that cleaved histidine from des-Leu angiotensin I to form angiotensin II and activity that cleaved tyrosine from benzyloxycarbonyl-glutamyl-tyrosine (ZGT). These results suggest that the urinary ACP activity is due to cathepsin A as we have reported previously for the porcine kidney enzyme. Analysis of sequential urine samples from a single individual over a 6-day period revealed as much as a 6-fold fluctuation in creaUnine-normalized ACP activity. Of five male healthy adult subjects, the creatininenormalized urinary ACP activity ranged from 1.7 to 3.7 mU/mL with a mean of 2.8 mU/mL. However, five male patients with renovascular hypertension had elevated levels of ACP activity with a mean of 11.6 mU/mL. Of five male patients with diabetic nephropathy, all had elevated ACP activity levels with a mean of 21.0 mU/mL. It is concluded that ACP activity in the urine is due to cathepsin A probably derived from kidney tissue, and that the release is increased in patients with kidney damage. We suggest that urinary ACP activity should be evaluated further for a possible relationship to renal hypertension and as a potentially early marker for diabetic nephropathy. We have been studying alternate pathways for the production of angiotensin II (A II) in the kidney that do not require the participation of angiotensin converting enzyme (ACE) (t). Recently, we have reported on the angiotensin carboxypepUdase (ACP) activity of porcine kidney cathepsin A by which the enzyme generates A II from angiotensin I (A I) by two sequential C-terminal cleavages (Fig. 1) (2). It was of interest, then, to determine whether this ACP activity of cathepsin A was detectable in human urine and to learn to what extent the activity of this lysosomal enzyme might increase in urine in those patients who have experienced tissue damage due to the onset of kidney disease. A preliminary report of some of the following data has been presented (3). * Fellow of the Glenmore Foundation. Reprint requests to Robert S. Levy, Department of Biochemistry, University of Louisville School of Medicine, Louisville, KY 40292 0024-3205/91 $3.00 +.00 Copyright (c) 1991 Pergamon Press plc
1530
Anglotensln Carboxypeptldase
in Urine
Vol. 48, No. 16, 1991
Asp---Arg-VaI-Tyr-Ile-His-Pro-Pt'~-His-Leu Angiotensin I
ACE
A S l o - A r g - V a l - Tyr - I l e - H i s - P r o - P h e - H i s des-Leu Angioter~lin I H
His Asp---,~g-Val- Tyr - Ile-His-I~'o-Vl"~
Angiotensin II
Fig. 1 Conversion of angiotensin I to angiotensln II by angiotensin converting enzyme (ACE) and angiotensin carboxypeptidase (ACP). Materials and Methods Materials Angiotensin I and Creatinine Coiorimetric Kit #555-A were obtained from Sigma Chemical Company, St. Louis, MO. Sephadex G-200 was obtained from Pharmacia Fine Chemicals, Piscataway, NJ. Samples Random urine samples were obtained from five male healthy adult subjects. The samples were chilled immediately and assayed for ACP activity on the same day. The refrigerated samples also were assayed for creatinine, sodium, and potassium within one week of collection. One volunteer collected random urine samples for a period of six days. Each of these was assayed for ACP activity and creatinine. For molecular weight determination the first 910 mL (approximating a 24 hr collection) of these latter samples was pooled and concentrated about 100-fold by ultrafiltration (Amicon PM-30; no activity was present in the effluent). The protein then was precipitated with 80% ammonium sulfate, resuspended in 0.1 M NaOAc, 0.1 M sucrose, 0.1 M KCI, 1 mM EDTA, pH 5.2, and dialyzed extensively against the same buffer. The molecular weight of the ACP in this concentrated sample was determined by gel filtration chromatography. Urine samples from diabetic and non-diabetic kidney patients were obtained from the Kidney Disease Program of the Department of Medicine. Hiqh Pressure Liquid Chromatography (HPLC) A Varian model 5020 high pressure liquid chromatograph equipped with a Micromeritics 725 Autosampler; a Varian UV-50 variable wavelength detector set at 210 nm; and a Vydac C18 reversed phase column (218TP104; 10 I~, 25 x 0.46 cm) were employed for HPLC analysis. The voltage output from the detector was converted to the digital form using the Nelson Analytical Chromatography Software Model 2600 in conjunction with the Nelson Analytical Interface Model 761.
Vol. 48, No. 16, 1991
Angiotensln Carboxypeptidase in Urine
1531
Enzyme Assays ACP activity was determined by measuring the conversion of A I to des-Leu anglotensin I (dL-A I). The reaction mixture contained 200 pL of buffer (0.1 M NaOAc, 0.5 M sucrose, 0.4 M KCI, 1 mM EDTA, pH 5.8) and 25 pL of 6.87 mM A I (0.687 mM in the assay mixture). After prewarming the assay mixture to 37°C, 25 pL of urine was added and incubated for 90 min. The reaction was terminated by the addition of 500 pL of 250 mM phosphoric acid. The resulting dLA I formed was measured by reversed phase HPLC on the C18 column. The mobile phase consisted of 50 mM sodium phosphate, pH 4.0, and acetonitrile (AcN) and was pumped at one mL/min. The gradient was started at 83% aqueous, and the AcN was increased by 2%/min for 7.5 min. Detection was by absorbance at 210 nm. Peak areas were compared against those for dL-A I standards. A typical HPLC assay is shown in Fig. 2. For assays of fractions from Sephadex G-200 chromatography, 10 I=L aliquots of each fraction were incubated with 250 i~L of 68.7 pM A I or 61.1 I~M dL-A I in the above buffer for 60 min or with 250 I~L of 3 mM benzyloxycarbonyl-glutamyl-tyrosine (ZGT - cathepsin A substrata) in the same buffer. The products, dL-A I, A II or benzyloxycarbonyl-glutamic acid (ZG), were quantitated by comparison of peak areas with those of standards. Activity was expressed as milliunits (mU) per mL where one unit equals one micromole of product per minute. Unless otherwise indicated, activity is normalized by dividing by the creatinine concentration in g/L. Molecular Wei.qht Determination Sephadex G-200 was equilibrated with 0.1 M NaOAc, 0.1 M NaCI, 0.1 M sucrose, 1 mM EDTA, pH 5.2, and packed into a column (63 x 2.6 cm; 334 mL). Ten mL of the concentrated urine (see above) was applied to the column, and 5 mL fractions were collected. Fractions were assayed for cathepsin A using ZGT as substrata and for ACP activity using both A I and dL-A I as substrates. Protein in the fractions was estimated by absorbance at 280 nm. The molecular weight of the enzyme was estimated by comparison of its elution volume from this column with the elution volumes of five proteins of known molecular weight: thyroglobulin, 670 kDa; immunoglobulin G, 158 kDa; bovine serum albumin, 67 kDa; ovalbumin, 44 kDa; and myogiobin, 17 kDa. Creatinine Determination Creatinine (g/L) was measured by a modification ofthe normal phase HPLC method of Patel and George (4). Ten I~L of urine or creatinine standard was diluted with 250 pL of water and 500 I~L of AcN was added. These mixtures then were centrifuged at 13,000 x g for 5 min. A 10 pL aliquot was injected onto a Varian silica column (SI-5; 5 I~, 30 x 0.4 cm). The mobile phase, 10/150/840 parts by volume of ammonium hydroxide, methanol, and AcN, was pumped at 2.5 mL/min. The effluent was monitored at 210 nm, CreaUnine had a retention time of 5.0 min and appeared to be well-resolved from other compounds. CreaUninein five urine samples determined by this method and by the Kodak Ektachem 400 Analyzer two slide method compared favorably (least squares regression: slope, 1.012; y-intercept, -0.0204; correlation coefficient, 0.9986). The Sigma Chemical Co. Creatinine I~t based on the Jaffa method as modified by Heineg~,rd and Tiderstr0m (5) was used for all patient samples. Other Assays Sodium and potassium were assayed on a Kodak Ektachem 400 Analyzer.
1532
Anglotensin Carboxypeptldase in Urine
Vol. 48, No. 16, 1991
Results Sequential urine samples from one normal subject over a six-day period were assayed for ACP activity and creatinine. A typical HPLC tracing is shown in Fig. 2. Creatinine-normalized ACP activity varied from 1.0to 6.1 (mean 2.9) mU/mL (Fig. 3). No clear diurnal variation was apparent; however, activity did tend to be higher in the afternoon and evening. These data indicated that urine normally contains detectable ACP activity. 200 180
0 ~
140
0
® 120
(3
dL-AI
100
0
All
AI
I
I
I
I
I
I
I
I
I
I
l
I
I
1
2
3
4
5
6
7
8
9
10
11
12
13
Retention
Time
14
(min)
Fig. 2 A HPLC tracing of an aliquot of reaction mixture after incubation of urine with A I as described in the Methods section. The detector wavelength was set at 210 nm.
7
+u" 3
:
0. 2 .
i ::
<
i i
1.i
:
::: O
"
o
"
:::
•
2
3
4
5
6
7
Days
Fig. 3 Creatinine-normalized ACP activity was measured on each random urine sample from one normal subject over a six-day period. There was no clearly apparent diurnal variation.
Vol. 48, No. 16, 1991
Angiotensin Carboxypeptidase in Urine
1533
Fig. 4 shows gel filtration chromatography of a concentrated sample of urine from the same subject. The two peaks of ACP activity (conversion of A I to dL-A I and dL-A I to A II) co-elute with cathapsin A activity (conversion of ZGT to ZG). The elution positions of these two peaks of activity are consistent with molecular weights of approximately 100 kDa and 500 kDa, respectively. These properties are similar to those of porcine kidney cathepsin A (2). This molecular weight study of the enzyme suggests that the source of the enzyme is the kidney, since the enzyme molecule probably is too large to be filtered by the glomerulus. 12
0.50
-•10
I
D E8
,,I-'
0
/~
Substrate T
/1/
f~
\
<4
~
\
0.40
ZGT. mU/mL j_
AI.
A
des-leu-AI
0.30
0.20
n O < 2 0
0 eO <
O. 10
~ ":" : - / 10 20
0.00 30
40
50
60
70
80
F r a c t i o n Number Fig. 4
Sephadex G-200 chromatography of a concentrated sample of human urine. Enzyme activity in the fractions (5 mL) was measured using ZGT, A I, or dL-A I as substrates and quantitation of the product formed (ZG, dL-A I, or A II, respectively) by HPLC. Enzyme activity is expressed as mU/mL (multiplied by 4 for A I and by 20 for dL-A I). The protein concentration in the fractions was estimated by absorbance at 280 nm. Random urine samples were obtained for assay from five normal male subjects, fk,e male patients with renovascular hypertension, and five male patients with diabetic nephropathy. The distribution of creatinine-normalized urinary ACP activity values is shown in Fig. 5. ACP activity in normal subjects ranged from 1.7 to 3.7 mU/mL (mean 2.8 mU/mL). All the patients had elevations of ACP activity. In those with renovascular hypertension the values ranged from 5.2 to 22.2 mU/mL (mean 11.6 mU/mL) while in those with diabetic nephropathy the values ranged from 8.2 to 45.6 mU/mL (mean 21.0 mU/mL). Comparison between groups was performed using Student's t test for unpaired data (7). There were significant differences between the normal subjects and the hypertensive patients (P < 0.01) and between the normal subjects and the diabetic patients (P < 0.02). In addition, levels of sodium and potassium were measured in 20 of the urine samples from normal subjects. Table 1 shows the correlation coefrmlents determined by least squares linear regression analysis for ACP activity versus sodium (0.681), potassium (0.627), and creatinine (0.894). The higher correlation of ACP activity with creatinine than with sodium or potassium suggests that the major source of variability in ACP activity is due to urine concentration. The lower and similar correlations of ACP activity with sodium and potassium are most likely due to
1534
Angiotensln Carboxypeptidase in Urine
Vol. 48, No. 16, 1991
electrolytes with creatinine. These results suggest that the urinary ACP activity is not related to systemic A II levels where A II stimulation of aldosterone release would be expected to decrease urinary sodium and increase urinary potassium.
50
40
30
<(
20 4,
< 10
%'.0 Normal
Hypertensive
Diabetic
Fig 5 The creatinine-normalized urinary ACP activity of five normal male subjects, five male patients with renovascular hypertension, and five male patients with diabetic nephropathy. The mean of urinary ACP activity in mU/mL per g/L of creaUnine were: normal subjects, 2.8 ; patients with renovascular hypertension, 11.6 ; and patients with diabetic nephropathy, 21.0. TABLE I Analyte Sodium Potassium Creatinine
r 0.681 0.627 0.894
Correlation coeff'¢ients between urinary ACP activity and urinary sodium, potassium, and creatinine. Correlation coefficients (r) were determined by least squares linear regression analysis. In each case, y is the ACP activity (not creatinine-normalized) and x is the analyte in the table. The data include values for ten urine samples from normal subjects.
Vol. 48, No. 16, 1991
AngiotenSsn Carboxypeptidase in Urine
1535
Discussion We have shown that ACP activity is present in human urine and has properties similar to porcine kidney cathepsin A. The lysosomal location and large size (100 kDa and 500 kDa) of this enzyme suggest that the source of the urinary ACP activity is the kidney itself. Over a period of several days, there is a relatively large variation in creatinine-normalized ACP activity (about 6fold). There does not appear to be a clear diurnal variation in urinary ACP activity, but the slightly higher values in afternoon specimens may be related to physical activity. More importantly, urinary ACP activity is elevated in patients with renovascular hypertension and in patients with diabetic nephropathy. These preliminary findings have suggested two lines of further investigation which are presently being pursued in our laboratory. The elevated levels of urinary ACP activity in diabetic nephropathy suggest that slight elevations may be detectable early in the course of the disease. We currently are comparing the utility of increased urinary ACP activity with that of slight elevations of urinary albumin for the purpose of identifying very early stage diabetic nephropathy (6). In addition, we are continuing our study of urinary ACP activity in hypertensive patients with the hope that elevated levels of this enzyme either may assist in the differential diagnosis of hypertension or be a prognostic indicator. Acknowledgements This study was supported by grants from the Humana Centers of Excellence (Humana Heart Institute International), the Glenmore Foundation, and the University of Louisville School of Medicine, Louisville, KY; the Veterans Administration (Merit Review Grants 86-001 and 88-001); and the National Institutes of Health (5 K08 NS01164-03). We are indebted to Ms. Nina W. Lesousky and Ms. Rebecca S. Sloan for their technical assistance. One of us (RSL) gratefully acknowledges the more than thirty years' of encouragement and counsel of William A. Blodgett, M.D. Re~rences 1. 2. 3. 4. 5. 6. 7.
D.G.CHANGARIS, J.J. MILLER, and R.S. LEVY, Biochem. Biophys. Res. Commun. 138, 573579 (1886). J.J. MILLER, D.G. CHANGARIS, and R.S. LEVY, Biochem. Biophys. Res. Commun. 154, 1122-1129 (1988). J.J. MILLER, D.G. CHANGARIS, and R.S. LEVY, FASEB J. 3, A995 (1989). C.P. PATEL and R.C. GEORGE, Clin. Chem. ~ 983 (1980). D. HEINEGARD and G. TIDERSTROM, Clin. Chim. Acta 4._33,305-310 (1973). H.-H. PARVING, B. OXENB®LL, P.Aa. SVENDSEN, J. SANDAHL CHRISTENSEN, and A.R. ANDERSEN, Acta Endocrinol. 100, 550-555 (1982). G.W. SNEDECOR and W. G. COCHRAN, Statistical Methods (7th Edition), pp. 54-59, Iowa State University Press, Ames, Iowa (1980).
Pergamon Press
ANGIOTENSIN CARBOXYPEPTIDASE ACTIVITY IN URINE FROM NORMAL SUBJECTS AND PATIENTS WITH K]DNEY DAMAGE James J. Miller*, David G. Changaris, and Robert S. Levy Departments of Biochemistry and Neurology and the Laboratory of Biological Psychiatry, University of Louisville School of Medicine, Louisville, KY 40292 (Received in final form February ii, 1991)
Summary Angiotensin carboxypeptidase (ACP) activity has been detected in urine samples from normal subjects and patients with hypertension and diabetes by determining the enzyme's ability to convert angiotensin I to des-Leu angiotensin I. Gel filtration chromatography of a concentrated urine sample indicated that about equal amounts of the enzyme exist as 100 kDa and 500 kDa molecular weight forms, respectively. This ACP activity co-eluted with activity that cleaved histidine from des-Leu angiotensin I to form angiotensin II and activity that cleaved tyrosine from benzyloxycarbonyl-glutamyl-tyrosine (ZGT). These results suggest that the urinary ACP activity is due to cathepsin A as we have reported previously for the porcine kidney enzyme. Analysis of sequential urine samples from a single individual over a 6-day period revealed as much as a 6-fold fluctuation in creaUnine-normalized ACP activity. Of five male healthy adult subjects, the creatininenormalized urinary ACP activity ranged from 1.7 to 3.7 mU/mL with a mean of 2.8 mU/mL. However, five male patients with renovascular hypertension had elevated levels of ACP activity with a mean of 11.6 mU/mL. Of five male patients with diabetic nephropathy, all had elevated ACP activity levels with a mean of 21.0 mU/mL. It is concluded that ACP activity in the urine is due to cathepsin A probably derived from kidney tissue, and that the release is increased in patients with kidney damage. We suggest that urinary ACP activity should be evaluated further for a possible relationship to renal hypertension and as a potentially early marker for diabetic nephropathy. We have been studying alternate pathways for the production of angiotensin II (A II) in the kidney that do not require the participation of angiotensin converting enzyme (ACE) (t). Recently, we have reported on the angiotensin carboxypepUdase (ACP) activity of porcine kidney cathepsin A by which the enzyme generates A II from angiotensin I (A I) by two sequential C-terminal cleavages (Fig. 1) (2). It was of interest, then, to determine whether this ACP activity of cathepsin A was detectable in human urine and to learn to what extent the activity of this lysosomal enzyme might increase in urine in those patients who have experienced tissue damage due to the onset of kidney disease. A preliminary report of some of the following data has been presented (3). * Fellow of the Glenmore Foundation. Reprint requests to Robert S. Levy, Department of Biochemistry, University of Louisville School of Medicine, Louisville, KY 40292 0024-3205/91 $3.00 +.00 Copyright (c) 1991 Pergamon Press plc
1530
Anglotensln Carboxypeptldase
in Urine
Vol. 48, No. 16, 1991
Asp---Arg-VaI-Tyr-Ile-His-Pro-Pt'~-His-Leu Angiotensin I
ACE
A S l o - A r g - V a l - Tyr - I l e - H i s - P r o - P h e - H i s des-Leu Angioter~lin I H
His Asp---,~g-Val- Tyr - Ile-His-I~'o-Vl"~
Angiotensin II
Fig. 1 Conversion of angiotensin I to angiotensln II by angiotensin converting enzyme (ACE) and angiotensin carboxypeptidase (ACP). Materials and Methods Materials Angiotensin I and Creatinine Coiorimetric Kit #555-A were obtained from Sigma Chemical Company, St. Louis, MO. Sephadex G-200 was obtained from Pharmacia Fine Chemicals, Piscataway, NJ. Samples Random urine samples were obtained from five male healthy adult subjects. The samples were chilled immediately and assayed for ACP activity on the same day. The refrigerated samples also were assayed for creatinine, sodium, and potassium within one week of collection. One volunteer collected random urine samples for a period of six days. Each of these was assayed for ACP activity and creatinine. For molecular weight determination the first 910 mL (approximating a 24 hr collection) of these latter samples was pooled and concentrated about 100-fold by ultrafiltration (Amicon PM-30; no activity was present in the effluent). The protein then was precipitated with 80% ammonium sulfate, resuspended in 0.1 M NaOAc, 0.1 M sucrose, 0.1 M KCI, 1 mM EDTA, pH 5.2, and dialyzed extensively against the same buffer. The molecular weight of the ACP in this concentrated sample was determined by gel filtration chromatography. Urine samples from diabetic and non-diabetic kidney patients were obtained from the Kidney Disease Program of the Department of Medicine. Hiqh Pressure Liquid Chromatography (HPLC) A Varian model 5020 high pressure liquid chromatograph equipped with a Micromeritics 725 Autosampler; a Varian UV-50 variable wavelength detector set at 210 nm; and a Vydac C18 reversed phase column (218TP104; 10 I~, 25 x 0.46 cm) were employed for HPLC analysis. The voltage output from the detector was converted to the digital form using the Nelson Analytical Chromatography Software Model 2600 in conjunction with the Nelson Analytical Interface Model 761.
Vol. 48, No. 16, 1991
Angiotensln Carboxypeptidase in Urine
1531
Enzyme Assays ACP activity was determined by measuring the conversion of A I to des-Leu anglotensin I (dL-A I). The reaction mixture contained 200 pL of buffer (0.1 M NaOAc, 0.5 M sucrose, 0.4 M KCI, 1 mM EDTA, pH 5.8) and 25 pL of 6.87 mM A I (0.687 mM in the assay mixture). After prewarming the assay mixture to 37°C, 25 pL of urine was added and incubated for 90 min. The reaction was terminated by the addition of 500 pL of 250 mM phosphoric acid. The resulting dLA I formed was measured by reversed phase HPLC on the C18 column. The mobile phase consisted of 50 mM sodium phosphate, pH 4.0, and acetonitrile (AcN) and was pumped at one mL/min. The gradient was started at 83% aqueous, and the AcN was increased by 2%/min for 7.5 min. Detection was by absorbance at 210 nm. Peak areas were compared against those for dL-A I standards. A typical HPLC assay is shown in Fig. 2. For assays of fractions from Sephadex G-200 chromatography, 10 I=L aliquots of each fraction were incubated with 250 i~L of 68.7 pM A I or 61.1 I~M dL-A I in the above buffer for 60 min or with 250 I~L of 3 mM benzyloxycarbonyl-glutamyl-tyrosine (ZGT - cathepsin A substrata) in the same buffer. The products, dL-A I, A II or benzyloxycarbonyl-glutamic acid (ZG), were quantitated by comparison of peak areas with those of standards. Activity was expressed as milliunits (mU) per mL where one unit equals one micromole of product per minute. Unless otherwise indicated, activity is normalized by dividing by the creatinine concentration in g/L. Molecular Wei.qht Determination Sephadex G-200 was equilibrated with 0.1 M NaOAc, 0.1 M NaCI, 0.1 M sucrose, 1 mM EDTA, pH 5.2, and packed into a column (63 x 2.6 cm; 334 mL). Ten mL of the concentrated urine (see above) was applied to the column, and 5 mL fractions were collected. Fractions were assayed for cathepsin A using ZGT as substrata and for ACP activity using both A I and dL-A I as substrates. Protein in the fractions was estimated by absorbance at 280 nm. The molecular weight of the enzyme was estimated by comparison of its elution volume from this column with the elution volumes of five proteins of known molecular weight: thyroglobulin, 670 kDa; immunoglobulin G, 158 kDa; bovine serum albumin, 67 kDa; ovalbumin, 44 kDa; and myogiobin, 17 kDa. Creatinine Determination Creatinine (g/L) was measured by a modification ofthe normal phase HPLC method of Patel and George (4). Ten I~L of urine or creatinine standard was diluted with 250 pL of water and 500 I~L of AcN was added. These mixtures then were centrifuged at 13,000 x g for 5 min. A 10 pL aliquot was injected onto a Varian silica column (SI-5; 5 I~, 30 x 0.4 cm). The mobile phase, 10/150/840 parts by volume of ammonium hydroxide, methanol, and AcN, was pumped at 2.5 mL/min. The effluent was monitored at 210 nm, CreaUnine had a retention time of 5.0 min and appeared to be well-resolved from other compounds. CreaUninein five urine samples determined by this method and by the Kodak Ektachem 400 Analyzer two slide method compared favorably (least squares regression: slope, 1.012; y-intercept, -0.0204; correlation coefficient, 0.9986). The Sigma Chemical Co. Creatinine I~t based on the Jaffa method as modified by Heineg~,rd and Tiderstr0m (5) was used for all patient samples. Other Assays Sodium and potassium were assayed on a Kodak Ektachem 400 Analyzer.
1532
Anglotensin Carboxypeptldase in Urine
Vol. 48, No. 16, 1991
Results Sequential urine samples from one normal subject over a six-day period were assayed for ACP activity and creatinine. A typical HPLC tracing is shown in Fig. 2. Creatinine-normalized ACP activity varied from 1.0to 6.1 (mean 2.9) mU/mL (Fig. 3). No clear diurnal variation was apparent; however, activity did tend to be higher in the afternoon and evening. These data indicated that urine normally contains detectable ACP activity. 200 180
0 ~
140
0
® 120
(3
dL-AI
100
0
All
AI
I
I
I
I
I
I
I
I
I
I
l
I
I
1
2
3
4
5
6
7
8
9
10
11
12
13
Retention
Time
14
(min)
Fig. 2 A HPLC tracing of an aliquot of reaction mixture after incubation of urine with A I as described in the Methods section. The detector wavelength was set at 210 nm.
7
+u" 3
:
0. 2 .
i ::
<
i i
1.i
:
::: O
"
o
"
:::
•
2
3
4
5
6
7
Days
Fig. 3 Creatinine-normalized ACP activity was measured on each random urine sample from one normal subject over a six-day period. There was no clearly apparent diurnal variation.
Vol. 48, No. 16, 1991
Angiotensin Carboxypeptidase in Urine
1533
Fig. 4 shows gel filtration chromatography of a concentrated sample of urine from the same subject. The two peaks of ACP activity (conversion of A I to dL-A I and dL-A I to A II) co-elute with cathapsin A activity (conversion of ZGT to ZG). The elution positions of these two peaks of activity are consistent with molecular weights of approximately 100 kDa and 500 kDa, respectively. These properties are similar to those of porcine kidney cathepsin A (2). This molecular weight study of the enzyme suggests that the source of the enzyme is the kidney, since the enzyme molecule probably is too large to be filtered by the glomerulus. 12
0.50
-•10
I
D E8
,,I-'
0
/~
Substrate T
/1/
f~
\
<4
~
\
0.40
ZGT. mU/mL j_
AI.
A
des-leu-AI
0.30
0.20
n O < 2 0
0 eO <
O. 10
~ ":" : - / 10 20
0.00 30
40
50
60
70
80
F r a c t i o n Number Fig. 4
Sephadex G-200 chromatography of a concentrated sample of human urine. Enzyme activity in the fractions (5 mL) was measured using ZGT, A I, or dL-A I as substrates and quantitation of the product formed (ZG, dL-A I, or A II, respectively) by HPLC. Enzyme activity is expressed as mU/mL (multiplied by 4 for A I and by 20 for dL-A I). The protein concentration in the fractions was estimated by absorbance at 280 nm. Random urine samples were obtained for assay from five normal male subjects, fk,e male patients with renovascular hypertension, and five male patients with diabetic nephropathy. The distribution of creatinine-normalized urinary ACP activity values is shown in Fig. 5. ACP activity in normal subjects ranged from 1.7 to 3.7 mU/mL (mean 2.8 mU/mL). All the patients had elevations of ACP activity. In those with renovascular hypertension the values ranged from 5.2 to 22.2 mU/mL (mean 11.6 mU/mL) while in those with diabetic nephropathy the values ranged from 8.2 to 45.6 mU/mL (mean 21.0 mU/mL). Comparison between groups was performed using Student's t test for unpaired data (7). There were significant differences between the normal subjects and the hypertensive patients (P < 0.01) and between the normal subjects and the diabetic patients (P < 0.02). In addition, levels of sodium and potassium were measured in 20 of the urine samples from normal subjects. Table 1 shows the correlation coefrmlents determined by least squares linear regression analysis for ACP activity versus sodium (0.681), potassium (0.627), and creatinine (0.894). The higher correlation of ACP activity with creatinine than with sodium or potassium suggests that the major source of variability in ACP activity is due to urine concentration. The lower and similar correlations of ACP activity with sodium and potassium are most likely due to
1534
Angiotensln Carboxypeptidase in Urine
Vol. 48, No. 16, 1991
electrolytes with creatinine. These results suggest that the urinary ACP activity is not related to systemic A II levels where A II stimulation of aldosterone release would be expected to decrease urinary sodium and increase urinary potassium.
50
40
30
<(
20 4,
< 10
%'.0 Normal
Hypertensive
Diabetic
Fig 5 The creatinine-normalized urinary ACP activity of five normal male subjects, five male patients with renovascular hypertension, and five male patients with diabetic nephropathy. The mean of urinary ACP activity in mU/mL per g/L of creaUnine were: normal subjects, 2.8 ; patients with renovascular hypertension, 11.6 ; and patients with diabetic nephropathy, 21.0. TABLE I Analyte Sodium Potassium Creatinine
r 0.681 0.627 0.894
Correlation coeff'¢ients between urinary ACP activity and urinary sodium, potassium, and creatinine. Correlation coefficients (r) were determined by least squares linear regression analysis. In each case, y is the ACP activity (not creatinine-normalized) and x is the analyte in the table. The data include values for ten urine samples from normal subjects.
Vol. 48, No. 16, 1991
AngiotenSsn Carboxypeptidase in Urine
1535
Discussion We have shown that ACP activity is present in human urine and has properties similar to porcine kidney cathepsin A. The lysosomal location and large size (100 kDa and 500 kDa) of this enzyme suggest that the source of the urinary ACP activity is the kidney itself. Over a period of several days, there is a relatively large variation in creatinine-normalized ACP activity (about 6fold). There does not appear to be a clear diurnal variation in urinary ACP activity, but the slightly higher values in afternoon specimens may be related to physical activity. More importantly, urinary ACP activity is elevated in patients with renovascular hypertension and in patients with diabetic nephropathy. These preliminary findings have suggested two lines of further investigation which are presently being pursued in our laboratory. The elevated levels of urinary ACP activity in diabetic nephropathy suggest that slight elevations may be detectable early in the course of the disease. We currently are comparing the utility of increased urinary ACP activity with that of slight elevations of urinary albumin for the purpose of identifying very early stage diabetic nephropathy (6). In addition, we are continuing our study of urinary ACP activity in hypertensive patients with the hope that elevated levels of this enzyme either may assist in the differential diagnosis of hypertension or be a prognostic indicator. Acknowledgements This study was supported by grants from the Humana Centers of Excellence (Humana Heart Institute International), the Glenmore Foundation, and the University of Louisville School of Medicine, Louisville, KY; the Veterans Administration (Merit Review Grants 86-001 and 88-001); and the National Institutes of Health (5 K08 NS01164-03). We are indebted to Ms. Nina W. Lesousky and Ms. Rebecca S. Sloan for their technical assistance. One of us (RSL) gratefully acknowledges the more than thirty years' of encouragement and counsel of William A. Blodgett, M.D. Re~rences 1. 2. 3. 4. 5. 6. 7.
D.G.CHANGARIS, J.J. MILLER, and R.S. LEVY, Biochem. Biophys. Res. Commun. 138, 573579 (1886). J.J. MILLER, D.G. CHANGARIS, and R.S. LEVY, Biochem. Biophys. Res. Commun. 154, 1122-1129 (1988). J.J. MILLER, D.G. CHANGARIS, and R.S. LEVY, FASEB J. 3, A995 (1989). C.P. PATEL and R.C. GEORGE, Clin. Chem. ~ 983 (1980). D. HEINEGARD and G. TIDERSTROM, Clin. Chim. Acta 4._33,305-310 (1973). H.-H. PARVING, B. OXENB®LL, P.Aa. SVENDSEN, J. SANDAHL CHRISTENSEN, and A.R. ANDERSEN, Acta Endocrinol. 100, 550-555 (1982). G.W. SNEDECOR and W. G. COCHRAN, Statistical Methods (7th Edition), pp. 54-59, Iowa State University Press, Ames, Iowa (1980).