- Email: [email protected]
Quantification of albumin in urine by a new method: zone immuno-electrophoresis assay (ZIA)
C/rrrr~~u (‘hlnllu ACYa. II3 (1981) 305-310 Elsevier/North-~iolland Biomedical Press
CCA
305
IWO
Quantification of albumin in urine by a new method: zone immuno-electrophoresis assay ( 214
Summary
Albumin in urine was determined by a new procedure called zone immunoelectrophoresis assay. A special apparatus was used with 20 vertical glass tubes (diameter 2 mm) which were filled up to 2 cm below their upper orifices with agarose gel containing antibodies against albumin. The samples were pipetted on top of each gel. Albumin was transported by electrophoresis into the agarose gels, in which zones of immunoprecipitates were formed. The amount of albumin in a sample was directly proportional to the distance from the gel surface to the front of the zone. The coefficient of variation was 8.8% within-day and 9.6% between days. The recovery of albumin added to urine:, was about 101%. The normal range for 40 adult persons was 2-23 mg/l with a mean and a median of 9.9 and 9.5 mg/l, respectively. The detection limit was well below the lowest value. The method was simple to use and the antiserum needed was only 0.5 ,ul per test.
Introduction Albuminuria, the most commonly used sign of impaired renal function, is generally evaluated by qualitative methods. However, these methods involve large errors [ 11. In addition to various classical forms of kidney disease the determination of albumin concentration (U-alb) is of interest in, e.g. pregnancy [2]. Even moderately increased concentrations may be of interest, e.g. to detect effects of some pharmaceuticals [3,4], or chemicals in occupational use [3%5]. Quantification of U-alb also gives valuable information in hypertonia [6] and diabetes [7,8]. In these studies analysis was usually by non-commercially available radioimmunoassay procedures [9]. Paper and SDS-electrophoresis not only require the urines to be concentrated about IOO-fold, but provide qualitative data only. Radial immunodiffusion also requires prior concentration, but this causes a variable loss of albumin. There is obviously a need for a simple method capable of determining U-alb at normal and slightly increased concentrations, without the need for prior concentration. Zone immuno-electrophoresis fulfills these requirements. Its advantages over some other methods has been described in some detail elsewhere [lo,1 I]. 0009.X9Xl/Sl/OOH-0000/$02.50
3: Elsevier/North-IIollaIld
Biomedical
Press
306
Materials and methods Zone immuno-electrophoresis was performed using a Quantiphor (Desaga, Heidelberg, F.R.G.) as described before [ 111. The antiserum used at 2 pi/ml gel was rabbit antihuman serum albumin from DAK0 (Copenhagen, Denmark). Standards of human serum albumin (HSA) from Kabi (Stockholm, Sweden) dissolved in the electrophoresis buffer containing 1 g/l of ovalbumin * (Sigma, U.S.A.). The final dilution of standards and urine samples was made in a similar solution but with 80 g/l of sucrose **. 50 ~1 of these solutions were pipetted on top of each gel. Pipettes were from Swelab (Stockholm, Sweden) with tips drawn out into capillaries (0.8 mm Ld.). The tubes were then filled with the buffer solution containing 5 mgjl of methylene blue (Sigma) using an injection needle with a bent tip and a 1 ml disposable injection syringe. Electrophoresis: 130 V, 12 mA for at least 3 h. or 30 V. 2 mA for about 16 h. The time periods are not very critical. Then the gel rods were taken out and laid side by side on a plastic plate (thickness 0.2 mm, Gel Bond %, Marine Colloids, ME, U.S.A.). The gels were then pressed and stained as described earlier [ 111. Samples
Morning urine spot samples were obtained from 40 healthy persons, 17 women and 23 men in the age range 20-61 years, whose work was mainly of a clerical nature. The samples were collected after a few hours in the office. Within 10 min sodium azide was added to give a final concentration of 0.1 g/l. The specimens were stored at 4°C. A “TS Meter” from American Optical Corp. (Buffalo, NY, U.S.A.) was used for measuring the urine density, p. U-alb concentration, in mg/l, was calculated using the formula U-alb X 24 X 10 ’ p-
1.00
Results A typical result after staining of the immunoprecipitates is shown in Fig. 1. A standard curve is shown in Fig. 2. When no egg albumin was added, with diluted standards and urine samples, the results were less reproducible. At low albumin concentrations the results were often lower than with egg albumin. We have also tried the addition of bovine serum albumin at 0.3 g/l. This regularly gave about 20% elevated values, probably due to some cross-reactivity of the antibodies. At an early stage in our studies we used a barbital buffer and often obtained lower than expected results with some urine samples only diluted 1 : 4 or less. We interpreted this as an effect of salts partly because the sample volumes are large in proportion to that of the gel. Some information about the influence of salt has recently been published [ 111. We could diminish this effect by using a buffer of high enough buffer capacity (cf. methods) for electrophoresis and dilution of samples. A typical assay of * Ovalbumin was used to prevent losses at low protein concentrations. ** Addition of ~uw.xe was used to increase the density of the sample solutions the tubes on top of the gels.
so that these rcmaincd
in
307
o“0
_ 1 _ _
20
=
30
=
Fig 1. Typical results after ZIA and staining of the immunoprecipitates. A mm scale is shown to the left and the gel rod numbers above. Numbers l-6 are standard albumin concentrations 6.0. 6.0, 2.5. 1.25, 0.83, 0.50 mg/l, respectively. Numbers 7- IO are different urine samples.
a urine sample at different dilutions is shown in Figs. 3 and 4. It can be seen in Fig. 4 that good correlations were obtained. The reproducibility was quite good as shown in Table I. When human albumin was added to ten urine samples containing about 10 mg/l so as to obtain twice this concentration, the recovery was 101% t 10% SD. The deviations observed were mainly explainable by pipetting inaccuracies, which were multiplied by the dilution factors. The results obtained indicate good accuracy of the method and also that
;
~
;
b
~ “4
/ I
Fig. 2. Standard curve for quantification of human albumin by ZIA. The distances from the upper gel surfaces to the front of the immunoprecipitatcs are shown to the left. The regression coefficient was 0.998.
308
urine samples can be evaluated against a standard curve of albumin in buffer. Urine samples originally containing 2 mg/l of albumin and diluted 1 :4 gave a distance from the gel surface to the front of the precipitates of 3.5 -2 0.3 mm and were thus easy to measure. Even lower concentrations can he easily determined by either diluting the samples less or lowering the antilx~dy concentration in the gel. Samples with U-alb > 50 mg/l. which are easily recognized by test sticks or precipitation tests, should be run highly diluted in ZIA. The distance from the gel
TABLE
I
I’KEC’ISION
OF Z1.A IN
rHE
I>El~tKMINA1’10iX
Ot; AI.RUbIIN
IN
CJKIiXk
(mg
I)
Fig. 5. Frequency distribution of urinary albumin concentration after correction for density on the .x-axis. The number of observations is on the J.-axis. The mean value was 9.Y mg/l and the median 9.5.
surface to the front of the zones of immunoprecipitates seemed to be inversely proportional to the concentration of antiserum in the gel. This implies that at highly elevated levels of U-alb one can use a higher concentration of antibodies instead of diluting the samples to a greater extent. At antiserum concentrations above 7 mg/l the precipitates are visible in oblique light without staining. The U-alb concentrations for the 40 healthy persons were in the range 1.9923.3 mg/l and the results are illustrated in Fig. 5. There seemed to be no systematic difference between women and men nor any obvious age difference. As timed sampling, and especially for 24 h, has many practical difficulties, we find it more rational to compensate for differences in urine flow rate by adjusting the U-alb concentrations for density. This cannot be used for samples with high glucose concentrations, of course. In order to prevent degradation of the U-alb by proteolytic enzymes the samples should be stored at -20°C or at 4°C with sodium azide. Discussion Determination of U-alb by rocket electrophoresis showed a detection limit which was inadequate. Trials with larger sample volumes or undiluted specimens gave highly curved calibration lines with low accuracy and reproducibility. The explanation for these phenomena, and the advantages of using ZIA, have been described recently [ 111. The detection limit is so low with the latter method that it is possible to measure even very low normal concentrations of U-alb after dilution, by which means interference from salts is reduced. The results obtained indicate that the reliability of the new method is satisfactory. The concentrations of U-alb in healthy persons have been reported [12] to be in the range 2- 15 mg/l, which is in good agreement with our results on our rather
3 IO
limited population. It can be expected, however, that if less rigorous criteria are used for the selection of the “normal” population, the upper limit will be higher. This would be the case if, for example, individuals with orthostatic albuminuria or poorly controlled hypertension are included. Therefore, if possible, it is advantageous to use each individual as his own control. Advantages of ZIA over radioimmunoassay are the absence of radioactivity with its related precautions and regulations, and independence from expensive kits. The cost of equipment for ZIA is at least ten times less. Furthermore the equipment is less complex, which should mean fewer breakdowns and lower servicing costs. Acknowledgements Thanks are due to Prof. diligent technical assistance.
H. Rilbe
for his comments
and to Mrs. N. Elms for
References 1 Gyure, W.L. (1977) Comparison of several methods for semiquantitative determination of urinary protein. Clin. Chem. 23, 876-879 2 Studd, J.W. and Wood, S. (1976) Serum and urinary proteins in pregnancy. In: Obstetrics and Gynecology Annual. Vol. 5 (Wynn. R.M., ed.). pp. 103- 123 3 Maher, J.F. (I 970) Nephrotoxicity of drugs and chemicals. Pharmacology for Physicians 4, I - 5 4 Baiter, P., Muehrcke, R.C., Morris, A.M.. Moles, J.B. and Lawrence. A.G. (1976) Chronic toxic nephropathies-Diagnosis and management. Ann. Clin. Lab. Sci. 6. 306-311 5 Ehrenreich, T. (1977) Renal disease from exposure to solvents. Ann. Clin. Lab. Sci. 7, 6- 16 6 Parving, H.-H., Jensen, H.AE., Mogensen. C.E. and Evrin, P.-E. (1974) Increased urinary albumin excretion rate in benign essential hypertension. Lancet. June 15. 1190- 1192 7 Mogensen, C.E. and Vittinghus, E. (1975) Urinary albumin excretion during exercise in juvenile diabetes. Stand. J. Clin. Lab. Invest. 35, 295-300 8 Viberti, G.C., Jarrett, R.J., McCartney, M. and Keen, H. (1978) Increased glomerular permeability to albumin induced by exercise in diabetic subjects. Diabetologia 14, 293-300 9 Keen, H. and Chlouverakis, C. (1963) An immunoassay method for urinary albumin at low concentrations. Lancet, Nov. 2, 913-914 IO Vesterberg. 0. (1980) New sensitive method for quantitative determination of proteins by zone immunoelectrophoresis (ZI). Fresenius Z. Anal. Chem. 30 I, 134- 135 I I Vesterberg, 0. (1980) Quantification of proteins with a new’ sensitive methods--Zone immunoelectrophoresis assay. Hoppe-Seyler’s Z. Physiol. Chem. 36 I, 6 17-624 12 Sayegh, H.A. and Jarrett, R.J. (1979) Oral glucose-tolerance tests and the diagnosis of diabetes: Results of a prospective study based on the Whitehall survey. Lancet. Sept. I, 431-433
CCA
305
IWO
Quantification of albumin in urine by a new method: zone immuno-electrophoresis assay ( 214
Summary
Albumin in urine was determined by a new procedure called zone immunoelectrophoresis assay. A special apparatus was used with 20 vertical glass tubes (diameter 2 mm) which were filled up to 2 cm below their upper orifices with agarose gel containing antibodies against albumin. The samples were pipetted on top of each gel. Albumin was transported by electrophoresis into the agarose gels, in which zones of immunoprecipitates were formed. The amount of albumin in a sample was directly proportional to the distance from the gel surface to the front of the zone. The coefficient of variation was 8.8% within-day and 9.6% between days. The recovery of albumin added to urine:, was about 101%. The normal range for 40 adult persons was 2-23 mg/l with a mean and a median of 9.9 and 9.5 mg/l, respectively. The detection limit was well below the lowest value. The method was simple to use and the antiserum needed was only 0.5 ,ul per test.
Introduction Albuminuria, the most commonly used sign of impaired renal function, is generally evaluated by qualitative methods. However, these methods involve large errors [ 11. In addition to various classical forms of kidney disease the determination of albumin concentration (U-alb) is of interest in, e.g. pregnancy [2]. Even moderately increased concentrations may be of interest, e.g. to detect effects of some pharmaceuticals [3,4], or chemicals in occupational use [3%5]. Quantification of U-alb also gives valuable information in hypertonia [6] and diabetes [7,8]. In these studies analysis was usually by non-commercially available radioimmunoassay procedures [9]. Paper and SDS-electrophoresis not only require the urines to be concentrated about IOO-fold, but provide qualitative data only. Radial immunodiffusion also requires prior concentration, but this causes a variable loss of albumin. There is obviously a need for a simple method capable of determining U-alb at normal and slightly increased concentrations, without the need for prior concentration. Zone immuno-electrophoresis fulfills these requirements. Its advantages over some other methods has been described in some detail elsewhere [lo,1 I]. 0009.X9Xl/Sl/OOH-0000/$02.50
3: Elsevier/North-IIollaIld
Biomedical
Press
306
Materials and methods Zone immuno-electrophoresis was performed using a Quantiphor (Desaga, Heidelberg, F.R.G.) as described before [ 111. The antiserum used at 2 pi/ml gel was rabbit antihuman serum albumin from DAK0 (Copenhagen, Denmark). Standards of human serum albumin (HSA) from Kabi (Stockholm, Sweden) dissolved in the electrophoresis buffer containing 1 g/l of ovalbumin * (Sigma, U.S.A.). The final dilution of standards and urine samples was made in a similar solution but with 80 g/l of sucrose **. 50 ~1 of these solutions were pipetted on top of each gel. Pipettes were from Swelab (Stockholm, Sweden) with tips drawn out into capillaries (0.8 mm Ld.). The tubes were then filled with the buffer solution containing 5 mgjl of methylene blue (Sigma) using an injection needle with a bent tip and a 1 ml disposable injection syringe. Electrophoresis: 130 V, 12 mA for at least 3 h. or 30 V. 2 mA for about 16 h. The time periods are not very critical. Then the gel rods were taken out and laid side by side on a plastic plate (thickness 0.2 mm, Gel Bond %, Marine Colloids, ME, U.S.A.). The gels were then pressed and stained as described earlier [ 111. Samples
Morning urine spot samples were obtained from 40 healthy persons, 17 women and 23 men in the age range 20-61 years, whose work was mainly of a clerical nature. The samples were collected after a few hours in the office. Within 10 min sodium azide was added to give a final concentration of 0.1 g/l. The specimens were stored at 4°C. A “TS Meter” from American Optical Corp. (Buffalo, NY, U.S.A.) was used for measuring the urine density, p. U-alb concentration, in mg/l, was calculated using the formula U-alb X 24 X 10 ’ p-
1.00
Results A typical result after staining of the immunoprecipitates is shown in Fig. 1. A standard curve is shown in Fig. 2. When no egg albumin was added, with diluted standards and urine samples, the results were less reproducible. At low albumin concentrations the results were often lower than with egg albumin. We have also tried the addition of bovine serum albumin at 0.3 g/l. This regularly gave about 20% elevated values, probably due to some cross-reactivity of the antibodies. At an early stage in our studies we used a barbital buffer and often obtained lower than expected results with some urine samples only diluted 1 : 4 or less. We interpreted this as an effect of salts partly because the sample volumes are large in proportion to that of the gel. Some information about the influence of salt has recently been published [ 111. We could diminish this effect by using a buffer of high enough buffer capacity (cf. methods) for electrophoresis and dilution of samples. A typical assay of * Ovalbumin was used to prevent losses at low protein concentrations. ** Addition of ~uw.xe was used to increase the density of the sample solutions the tubes on top of the gels.
so that these rcmaincd
in
307
o“0
_ 1 _ _
20
=
30
=
Fig 1. Typical results after ZIA and staining of the immunoprecipitates. A mm scale is shown to the left and the gel rod numbers above. Numbers l-6 are standard albumin concentrations 6.0. 6.0, 2.5. 1.25, 0.83, 0.50 mg/l, respectively. Numbers 7- IO are different urine samples.
a urine sample at different dilutions is shown in Figs. 3 and 4. It can be seen in Fig. 4 that good correlations were obtained. The reproducibility was quite good as shown in Table I. When human albumin was added to ten urine samples containing about 10 mg/l so as to obtain twice this concentration, the recovery was 101% t 10% SD. The deviations observed were mainly explainable by pipetting inaccuracies, which were multiplied by the dilution factors. The results obtained indicate good accuracy of the method and also that
;
~
;
b
~ “4
/ I
Fig. 2. Standard curve for quantification of human albumin by ZIA. The distances from the upper gel surfaces to the front of the immunoprecipitatcs are shown to the left. The regression coefficient was 0.998.
308
urine samples can be evaluated against a standard curve of albumin in buffer. Urine samples originally containing 2 mg/l of albumin and diluted 1 :4 gave a distance from the gel surface to the front of the precipitates of 3.5 -2 0.3 mm and were thus easy to measure. Even lower concentrations can he easily determined by either diluting the samples less or lowering the antilx~dy concentration in the gel. Samples with U-alb > 50 mg/l. which are easily recognized by test sticks or precipitation tests, should be run highly diluted in ZIA. The distance from the gel
TABLE
I
I’KEC’ISION
OF Z1.A IN
rHE
I>El~tKMINA1’10iX
Ot; AI.RUbIIN
IN
CJKIiXk
(mg
I)
Fig. 5. Frequency distribution of urinary albumin concentration after correction for density on the .x-axis. The number of observations is on the J.-axis. The mean value was 9.Y mg/l and the median 9.5.
surface to the front of the zones of immunoprecipitates seemed to be inversely proportional to the concentration of antiserum in the gel. This implies that at highly elevated levels of U-alb one can use a higher concentration of antibodies instead of diluting the samples to a greater extent. At antiserum concentrations above 7 mg/l the precipitates are visible in oblique light without staining. The U-alb concentrations for the 40 healthy persons were in the range 1.9923.3 mg/l and the results are illustrated in Fig. 5. There seemed to be no systematic difference between women and men nor any obvious age difference. As timed sampling, and especially for 24 h, has many practical difficulties, we find it more rational to compensate for differences in urine flow rate by adjusting the U-alb concentrations for density. This cannot be used for samples with high glucose concentrations, of course. In order to prevent degradation of the U-alb by proteolytic enzymes the samples should be stored at -20°C or at 4°C with sodium azide. Discussion Determination of U-alb by rocket electrophoresis showed a detection limit which was inadequate. Trials with larger sample volumes or undiluted specimens gave highly curved calibration lines with low accuracy and reproducibility. The explanation for these phenomena, and the advantages of using ZIA, have been described recently [ 111. The detection limit is so low with the latter method that it is possible to measure even very low normal concentrations of U-alb after dilution, by which means interference from salts is reduced. The results obtained indicate that the reliability of the new method is satisfactory. The concentrations of U-alb in healthy persons have been reported [12] to be in the range 2- 15 mg/l, which is in good agreement with our results on our rather
3 IO
limited population. It can be expected, however, that if less rigorous criteria are used for the selection of the “normal” population, the upper limit will be higher. This would be the case if, for example, individuals with orthostatic albuminuria or poorly controlled hypertension are included. Therefore, if possible, it is advantageous to use each individual as his own control. Advantages of ZIA over radioimmunoassay are the absence of radioactivity with its related precautions and regulations, and independence from expensive kits. The cost of equipment for ZIA is at least ten times less. Furthermore the equipment is less complex, which should mean fewer breakdowns and lower servicing costs. Acknowledgements Thanks are due to Prof. diligent technical assistance.
H. Rilbe
for his comments
and to Mrs. N. Elms for
References 1 Gyure, W.L. (1977) Comparison of several methods for semiquantitative determination of urinary protein. Clin. Chem. 23, 876-879 2 Studd, J.W. and Wood, S. (1976) Serum and urinary proteins in pregnancy. In: Obstetrics and Gynecology Annual. Vol. 5 (Wynn. R.M., ed.). pp. 103- 123 3 Maher, J.F. (I 970) Nephrotoxicity of drugs and chemicals. Pharmacology for Physicians 4, I - 5 4 Baiter, P., Muehrcke, R.C., Morris, A.M.. Moles, J.B. and Lawrence. A.G. (1976) Chronic toxic nephropathies-Diagnosis and management. Ann. Clin. Lab. Sci. 6. 306-311 5 Ehrenreich, T. (1977) Renal disease from exposure to solvents. Ann. Clin. Lab. Sci. 7, 6- 16 6 Parving, H.-H., Jensen, H.AE., Mogensen. C.E. and Evrin, P.-E. (1974) Increased urinary albumin excretion rate in benign essential hypertension. Lancet. June 15. 1190- 1192 7 Mogensen, C.E. and Vittinghus, E. (1975) Urinary albumin excretion during exercise in juvenile diabetes. Stand. J. Clin. Lab. Invest. 35, 295-300 8 Viberti, G.C., Jarrett, R.J., McCartney, M. and Keen, H. (1978) Increased glomerular permeability to albumin induced by exercise in diabetic subjects. Diabetologia 14, 293-300 9 Keen, H. and Chlouverakis, C. (1963) An immunoassay method for urinary albumin at low concentrations. Lancet, Nov. 2, 913-914 IO Vesterberg. 0. (1980) New sensitive method for quantitative determination of proteins by zone immunoelectrophoresis (ZI). Fresenius Z. Anal. Chem. 30 I, 134- 135 I I Vesterberg, 0. (1980) Quantification of proteins with a new’ sensitive methods--Zone immunoelectrophoresis assay. Hoppe-Seyler’s Z. Physiol. Chem. 36 I, 6 17-624 12 Sayegh, H.A. and Jarrett, R.J. (1979) Oral glucose-tolerance tests and the diagnosis of diabetes: Results of a prospective study based on the Whitehall survey. Lancet. Sept. I, 431-433