- Email: [email protected]
Quantification of apolipoprotein D in human urine by zone immunoelectrophoresis assay: a methodological and clinical study
Journal of Biochemical and Biophysical Methods, 23 (1991) 315-327
315
© 1991 Elsevier Science Publishers B.V. All rights reserved 0165-022X/91/$03.50
JBBM 00915
Quantification of apolipoprotein D in human urine by zone immunoelectrophoresis assay: a methodological and clinical study Leif Holmquist and Olof Vesterberg Division of Medical Chemistry, National Institute of Occupational Health, Solna, Sweden (Received 15 July 1991) (Revised version received 22 August 1991) (Accepted 22 August 1991)
Summary A zone immunoelectrophoresis assay (ZIA) has been developed for the quantification of apolipoprotein D (apo D) in unconcentrated native human urine. A standard curve, linear between 1 and 8 mg apo D/I was obtained with ZIA. The relative coefficients of variation for this method were 5-9% ( n = 1 5 × 6 ) with a mean_+SD of 7_+1.4% and below 11% ( n = 6 × 1 5 ) for within-run and between-run reproducibility, respectively. Equal amounts of apo D in unconcentrated and diluted urines, in serum and of the purified protein produced the same zone migration distances indicating parallelism between the immunologic reactions of apo D in different sample matrixes. Storage experiments with normal urines demonstrated good stability of apo D in both acidic and alkalinized urine over at least 2 days at +5°C and during several days at -20°C to -40°C. Using ZIA, urine samples from 50 normal healthy men aged 23-65 years were analyzed for apo D. Mean and SD were: 2.8_+2.1 mg/l, 2.6_+ 1.8 ~ g / m i n and 0.24_+0.13 mg/mmol for concentration, rate of excretion and mass/creatinine concentration, respectively. Key words: Apolipoprotein D; Normal urine; Quantification; Zone immunoelectrophoresis
Introduction Recently we reported that normal human urine contains apolipoprotein D (apo D) as one of the major proteins which was quantified by rocket immunoelecAbbreviations: HDL, high density lipoproteins; Tricine, N-tris(hydroxymethyl)methylglycine; ZIA, zone immunoelectrophoresis assay; u, 1 unit of mass (1.66 x 10 -27 kg).
Correspondence: L. Holmquist, Division of Medical Chemistry, National Institute of Occupational Health, S-17184 Solna, Sweden.
316
trophoresis [1]. This low-molecular-mass apolipoprotein was originally discovered in plasma by McConathy and Alaupovic [2] and was found to be associated with its high density lipoprotein fraction (HDL; d = 1.063-1.21 kg/l). Although well characterized, the role and function of this apolipoprotein is unknown. However, its amino acid sequence shows homology to that of a~-microglobulin and retinolbinding protein [3] both of which are urinary proteins of diagnostic value [4,5]. The physico-chemical properties of plasma apo D, which has an apparent molecular mass of 33000 u (1 u -- 1.66 × 10 -27 kg) and a documented stability to acid and urea [2] make urinary apo D a potential complement to these proteins for the study of kidney dysfunction, especially with respect to the diagnosis of tubular proteinuria. To investigate this the quantification of urinary apo D was adapted to the zone immunoelectrophoresis assay (ZIA) as described by Vesterberg [6,7]. In this method, the lateral diffusion of the antigen occurring in rocket immunoelectrophoresis is inhibited by allowing the formation of the antigen-antibody complex to develop in glass capillaries; this results in narrow immunoprecipitates with sharp zone-fronts. The migration distances of such zones are linearly related to the concentration of antigen in the sample. This technique has the further advantages over rocket immunoelectrophoresis that large sample volumes can be used, making concentration steps unnecessary, and that the zone migration distances are linearly related to the amount of the antigen even if applied in different volumes over a wide range. This communication reports the details for the quantification of apo D in unconcentrated urine by zone immunoelectrophoresis assay and its application for the determination of this urinary protein in a large group of apparently healthy normal men.
Materials and Methods
Study population and samples Samples of urine, for the evaluation of method characteristics, were obtained by midstream collection from apparently healthy males or females 22-53 years old. A portion of each urine sample was immediately alkalinized with sodium bicarbonate. The remaining urine samples were promptly chilled to 0°C, centrifuged at 4000 × g for 10 min and added to solid sodium azide (0.1 g/l). Aliquots of all the urine samples were frozen and stored for stability tests. For evaluation of apo D concentration in urine of normal subjects, samples were collected in the mornings before noon from 50 non-fasting apparently healthy men 23-65 years old. None was on treatment with drugs or diet, nor had any acute or chronic disease. They were all feeling in good health. The donor was instructed first to void without collecting the urine and note the time of the clock. After 2 h he had to void in a pre-weighed clean polyethylene flask. Thereafter the flask was transported to the analytical laboratory within 15 min. The weight, density and pH of the urine was recorded and aliquots were centrifuged at 0°C without delay and
317 sodium azide was added as described above. The quantifications of apo D in the urines were started on non-frozen samples within 2 h of collection.
Materials Authentic apo D for primary immunological standards was isolated from normal human H D L prepared by ultracentrifugation of pooled plasma from apparently healthy subjects as described by McConathy and Alaupovic [2]. The isolated apo D was electrophoretically and immunologically characterized as previously reported [1], using inter alia a reference goat anti-human apo D, obtained as a gift from P. AIaupovic, Oklahoma City, U.S.A. Monospecific polyclonal antibodies against the purified apo D were raised in albino rabbits and were the same as used in an earlier report [1]. Stock solutions of purified apo D at a 10 mg/1 concentration for preparation of primary immunological standards were made in 0.04 M Tris-Tricine buffer (pH 8.6) containing 1% (w/v) bovine serum albumin. Only glass tubes and vessels were used for stock solutions and dilutions of apo D and serum. A reference serum pool for preparation of apo D secondary standards was obtained from fasting normolipidemic healthy subjects and was isolated and treated as previously described [8]. The apo D concentration of this reference serum was estimated to be 119 mg/1 using rocket immunoelectrophoresis [9]. Aliquots of the stock solution of apo D and the reference serum were stored under a nitrogen atmosphere in capped glass vials at - 70°C. After thawing, the individual solutions were carefully homogenized before use.
General methods Total protein was determined according to Lowry et al. [10] using purified bovine serum albumin as a standard (fraction V powder, Sigma, St. Louis, MO, U.S.A.). Urine creatinine concentrations were determined at the Karolinska Hospital, Solna, Sweden, using a Kodak Ectachem 7000 instrument and the kit, Creatinine Two Point Rate Test, with the enzyme creatinine amido hydrolase (Eastman Kodak, Rochester, NY, U.S.A.). Between-run and within-run reproducibility determinations were performed during one week on a neutral native urine containing sodium azide and stored at +5°C. In each ZIA experiment, performed to establish the stability of urinary apo D on storage, a sample of a reference serum (Human Serum Protein Calibrator; Dakopatts, Glostrup, Denmark) containing apo D was used as an internal standard to normalize the zone migration distances of the urine samples of days 1 and 2 to those of day 0. The day 0 urine samples were analyzed within 1 h of sampling.
Quantification procedure The zone immonoelectrophoresis assay of Vesterberg [6,7] was adapted for quantification of apo D, using 1.3% (w/v) HSA agarose (FMC-Litex, Glostrup) in 0.04 M Tris-Tricine buffer (pH 8.6) containing 4% (w/v) polyethylene glycol 6000. The concentration of rabbit anti-apo-D (precipitated immunoglobulins) was 1 pA (44 /~g) per ml of agarose gel supplemented with 5 /.~l/ml of non-immune rabbit serum. Final dilutions of stock solutions of purified apo D and the reference serum
318
were made in Tris-Tricine buffer containing 1% (w/v) bovine serum albumin, which was added to prevent adsorption of the apolipoprotein to the vessel walls. Before application of the samples (10/~1) using a Hamilton syringe, their volumes were increased by 20% (v/v) by addition of a mixture (2 /~1) of 1 vol. 0.04 M Tris-Tricine buffer (pH 8.6), 2 vol. 80% (w/v) sucrose, and 0.2 vol. 0.5% (w/v) methylene blue in 0.04 M Tris-Tricine buffer (pH 8.6). This increased the density of the sample for safe application on the gel surface beneath the electrophoresis buffer. Urine samples or dilutions in 0.04 M Tris-Tricine buffer (pH 8.6) containing 1% (w/v) bovine serum albumin were applied in relevant volumes to give immunoprecipitates compatible with those of the apo D standards ranging 1-8 mg/1. Densities of the samples were increased as above. The quantification of urinary apo D in the study population of the 50 men was performed with ZIA using 5 different concentrations of the standard and, in addition, a sample of a control human serum containing apo D (Seronorm; Nycomed, Oslo, Norway) run simultaneously with the urine samples. Electrophoresis was run at 30 V and 2 mA for 18 h. Gels were treated as previously described [6,7] and stained in a solution of 1 g Coomassie Brilliant blue R in a mixture of concentrated acetic acid, ethanol and water (1 : 4.5 : 4.5, 200 ml) for 5-10 min. Destaining was performed in the same solution not containing the stain. A detailed description of measurement of the migration distances of the immunoprecipitate zones has been published earlier [7] and, if not otherwise stated, ZIA determinations of apo D concentrations were made in duplicate for samples and as singlets for each concentration of the standard series. Skewness and kurtosis of measured values were estimated using the program StatWiev 512 + (BrainPower Inc., Calabasas, CA, U.S.A.).
Results
A linear relationship between zone migration distances of the apo D-containing immunoprecipitates and unconcentrated urine sample volumes up to the largest tested 50-/~1 sample was obtained with the zone immunoelectrophoresis technique as shown in Fig. 1A. Experiments with dilutions of the urine resulted in zone migration distances that matched well with those expected of corresponding volumes of unconcentrated urine (Fig. 1B). Apo D isolated from H D L of plasma or present in serum also produced zone migration distances in the 1-8 mg/1 concentration range, using 10-pA samples, that were linearly related to the corresponding apolipoprotein concentrations (Fig. 2A). Within the experimental error of estimation the two different apo D standards produced the same curves (Fig. 2B). The zone migration distance values obtained with different amounts of purified apo D standard and urinary apo D were scattered along the regression line for the 10-80-ng range (Fig. 2C), indicating parallelism between the immunologic reactions in the different sample matrixes. Urinary and standard purified apo D gave immunoprecipitation zones of equal quality (Fig. 3). The coefficients of variation for the ZIA of apo D were 5 - 9 % (n = 15 x 6) with a mean + SD of 7 +_ 1.4% and below 11% (n = 6 x 15) for within-run and between-
319 3o
A E E c
20,
"o
E
10
o N
0
i
i,
10
20
i
i
I
30
40
50
60
Unconcentrated urine, pl
14
112 B
0
!
i
5
10
15
20
Urine, pI
Fig. 1. (A) Plot of zone migration distances (ram) against applied volumes (/zl) of unconcentrated native h u m a n urines obtained on ZIA. n , • : values for two different normal urine samples. (B) Plot of zone migration distances (mm) against applied volumes (/zl) of unconcentrated native, normal urine ([]) and of the same urine after dilution with Z I A buffer ( • ) . The diluted urine was applied as 50-/zl samples.
run, respectively, as estimated with 10-1 /xl samples of normal urine and a series of purified apo D standard solutions of concentrations 2, 4, 6 and 8 mg protein/1. Storage of native urine samples at + 5°C for 1 and 2 days and for several days at - 2 0 to - 4 0 ° C resulted in an average recovery of apolipoprotein D of 98%, 87% and 86% in relation to day 0, respectively, in a population of 20 samples, estimated as zone migration distances on ZIA as shown in Tables 1, 2, and 3. The acidic urines having pH values between 5.1 and 5.4 yielded essentially the same results on storage as the whole population of native urines (Tables 1, 2 and 3). In another population of 9 urine samples, storage for 1 and 2 days at + 5°C and for several days at - 4 0 ° C yielded an average recovery of apo D of 105%, 107% and 102%, respectively, as compared to freshly collected urines. Alkalinization of these native urines with sodium bicarbonate gave values for zone migration distances for apo D
320 30
E E t~o
20,
~5
10 E
=., o N 0 0
i
i
i
i
2
4
6
8
10
Apolipoprotein D, mg/I 14
B
12
°i g
I,q
0
10
~5
N 0
20
40
60
so
100
Apolipoprotein D, ng
Fig. 2. (A) Plot of zone migration distances (mm) against concentrations of purified apo D from H D L (mg/l), applied as 10-/zl samples. (For details see Materials and Methods). (B) Plot of zone migration distances (mm) against concentrations of purified apo D ( m g / l , O) and against samples of 2 different normal h u m a n sera of known apo D content, as determined by rocket immunoelectrophoresis (11) applied in amounts corresponding to the purified apo D. (For details see Materials and Methods). (C) Plot of zone migration distances (mm) against amounts of apo D (ng) for samples of 2 different normal h u m a n sera of known apo D content ([3) and for a normal h u m a n urine of 4.5 mg apo D / I (11), as determined by rocket immunoelectrophoresis. (For details see Materials and Methods).
9.6__+0.43 (4) 6.0_+0.25 (4) 13.3_+0.33 (5) 5.9 -+ 0.22 (5) 9.9_+0.41 (5) 10.7_+0.49 (5) 19.3+0.30(5) 10.8_+0.71 (4) 5.7_+0.42 (5) 7.2 _+0.82 (5) 9.6_+0.50 (5) 29.8_+ 1.76 (5) 30.7_+1.16(5) 6.6_+0.86 (5) 11.0_+0.26 (5) 2.6 _+0.49 (3) 11.5_+0.21 (2) 6.3 + 0.38 (3) 7.1 _+0.14 (2) 19.6 _+0.00 (2)
11.0_+0.28 (4) 5.9_+0.06 (4) 13.7__+0.45 (5) 5.9 -+ 0.38 (5) 7.9_+0.65 (5) 9.9_+0.41 (5) 18.3_+0.86(5) 9.2_+0.30(4) 3.7__+0.50 (5) 8.4 _+0.75 (5) 8.4_+0.69 (5) 31.4_+ 1.96 (5) 26.3_+0.79(5) 7.6_+0.37 (5) 8.2_+0.18 (5) 2.0 _+0.00 (3) 8.6-+0.19 (4) 4.9 __+0.00 (3) 6.0_+0.92 (2) 20.5 __+0.85 (2)
2 (mean_+ SD, (n))
S1
145 120 83 94 105 92 101 89 83 80 102 109 133 80 82 108 92 94 82 76
(%)
$2
69 73 69 82
83
114 118 85 94 84 85 96 76 53 93 89 115 114 92 61
(%)
8.0__+0.17 11.0_+0.57 19.0_+0.97 15.8_+0.49 3.7_+0.26 5.6 _+0.52 7.4_+0.00 25.6_+0.00 18.4_+1.31
(5/28) (5/35) (5/28) (5/15) (5/10) ( 5 / 9) (5/38) (5/25) (5/24)
(n/days)
Thawed after storage at - 4 0 ° C (mean -+ SD)
S1 and $2 represent recovery values for urines stored 1 and 2 days, respectively; $3 represents values for frozen and thawed urines.
6.6__+0.38(4) 5.0_+0.00 (4) 16.1 __+0.13 (5) 6.3 __+0.23 (5) 9.4_+0.62 (5) 11.7__+0.45(5) 19.1_+0.26(5) 12.1_+0.63(4) 6.9_+0.80 (5) 9.0 _+0.72 (5) 9.4_+ 1.03 (5) 27.4_+0.43 (5) 23.0_+0.51 (5) 8.3_+ 1.04 (5) 13.4_+0.59 (5) 2.4 _+0.20 (3) 12.5__+0.71(2) 6.7 _+0.42 (3) 8.7_+0.14 (2) 25.5 _+0.71 (2)
7.0 5.3 5.3 6.0 6.9 5.3 6.1 5.4 5.4 5.2 6.9 7.2 5.4 5.4 5.3 6.5 5.8 5.3 6.7 5.1
1 (mean_+ SD, (n))
Time of storage at +5°C (days)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
pH
0 (mean_+ SD, (n))
-
No.
Sample
-
85 94 99 130 54 62 79 93 80
(%)
$3
Zone migration distances (mm_+ SD) produced by urinary apolipoprotein D on ZIA of native urine samples after different times of storage at + 5 ° C and 20 to 40°C
TABLE 1
Native
6.3
No.
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9
7.4
7.5
7.4
7.6
7.2
7.3
7.4
7.2
7.4
Adjusted
8.3 +_ 1.16 9.5_+0.77 9.3_+0.48 8.9_+0.21 3.0_+0.32 2.8_+0.42 12.3_+0.45 8.3_+0.33 11.0_+1.42 11.6_+0.48 8.3_+0.51 7.6_+0.27 9.5_+0.44 8.4_+0.58 10.0-+0.08 9.5 -+ 0.45 26.1 _+O.17 26.1 _+0.35
0 (mean _+SD) 10.1 -+ 0.48 10.3+0.42 9.7+0.35 9.1_+0.34 3.6_+0.24 2.8_+0.20 12.2_+0.22 9.3_+0.05 11.2_+0.20 11.1_+0.51 6.2_+0.70 5.1_+0.68 7.0-+0.50 6.0_+0.54 15.2-+0.11 13.1 -+ 1.00 26.6 + 0.98 24.8_+0.81
1 (mean _+SD)
Time of storage at + 5°C (days)
10.2 _+0.38 10.2_+0.23 8.9-+0.29 7.6+0.06 3.5-+0.46 2.9_+0.26 13.5_+0.45 11.9_+0.36 10.1+0.50 9.3_+0.44 5.9-+0.42 4.4_+0.45 9.5-+0.52 7.6_+0.47 15.1-+0.00 13.3 _+0.25 27.0 _+0.37 26.9_+0.42
2 (mean _+SD)
S1
122 108 104 102 120 100 99 112 102 96 74 67 73 71 152 138 102 95
(%)
$2
123 107 96 85 117 104 110 143 92 80 71 58 100 90 151 138 103 103
(%) 9.0 -+ 0.35 8.5_+0.17 8.9_+0.31 8.1-+0.27 4.9_+0.32 4.3-+0.35 12.4_+0.68 8.3_+0.39 7.0_+0.68 8.5_+0.52 7.6_+0.41 7.0_+0.51 11.0_+0.48 10.2-+0.35 11.1-+0.55 9.9 _+0.42 18.3 _+0.58 17.0_+0.99
(9) (9) (21) (21) (21) (21) (11) (11) (10) (10) (6) (6) (6) (6) (27) (27) (27) (27)
(days)
Thawed after storage at - 40°C (mean _+SD)
SI and $2 represent recovery values for urines stored 1 and 2 days, respectively. $3 represent values for frozen and thawed urines.
6.5
7.0
6.8
6.4
7.6
5.3
6.1
5.5
pH
Sample
108 89 96 91 163 209 101 100 64 73 92 92 116 121 111 104 70 65
(%)
$3
Zone migration distances (ram_+ SD, n = 6) produced by urinary apolipoprotein D on Z I A of native and alkalinized urine samples after different times of storage at + 5 and - 40°C
TABLE 2
ba
323
12345
6789101112
E ¢,)
+ Fig. 3. Electrophoretogram of ZIA demonstrating Coomassie-stained immunoprecipitation zones obtained with samples of purified apo D (20 to 80 ng, gel rod 1 to 5) and by a sample of unconcentrated normal human urine (20/zl, gel rod 6 to 12). For details see Materials and Methods.
that did not seem to differ from the values obtained from the corresponding native urine samples and recoveries were on average unaffected by the different conditions of storage as shown in Tables 1, 2 and 3. In order to test the practicability of the Z I A method and to obtain normal concentration values (reference population), the apo D concentrations in urine of
TABLE 3 Average recoveries of apolipoprotein D in native and alkalinized urines after different times of storage at + 5 and - 4 0 ° C (percentage_+ SD of zone migration distances in relation to corresponding values for freshly collected urine samples, calculated from the values of columns S1, $2 and $3 in Tables 1 and 2) Urine samples
Native Native Native, pH 5.1 - 5.4 Native Native Alkalinized Alkalinized
Number of samples
Condition of storage
20 9 11 9 9 9 9
+5 - 20 to - 40 +5 +5 - 40 +5 - 40
a Average time of storage.
( o C)
Time of storage (days) I 2 98 + 18.5
87 _+18.1
92+_18.1 105 _+24.3
85.+19.7 107 + 22.3
99 -+21.3
101 -+27.0
18 _+9.7 a
86_+22.1
102 _+28.8 115+55.5
324
o
I
2
1
3
4 5 6 Apo D in urine, mg/1
7
8
9
I
4 5 6 7 Excretion of apo D, ~Jg/min.
8
9
10
2
~
B 10.
8
g 8
6 4.
2, O. 0 7 .
1 .
.
2 .
.
3 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
C
8.
5.
8
3. 2,
'
nN ,
0
.
,
.1
.
,
.2
.
,
.
,
.
,
.
,
.
,
.3 .4 .5 .6 .7 Apo D in urine, mg/mmol creatinine
.
,
.8
.
,
.
,
.9
Fig. 4. (A) Frequency distribution of urinary apo D concentrations obtained for the group of 50 normal men. The values are arranged in 0.50-mg/t intervals each including the lowest value. The mean and the median are 2.8 and 2.4 mg/1, respectively. (B) Frequency distribution of the rate of excretion of apo D obtained for the group of 50 normal men. The values are arranged in 0.5-p~g/min intervals, each including the lowest value. The mean and median are 2.6 and 2.5/.tg/min, respectively. (C) Frequency distribution of mg of urinary apo D per mmol of creatinine obtained for the group of 50 normal men. The values are arranged in 0.0125-mg/mmol intervals and each includes the lowest value. The mean and median are 0.24 and 0.23 m g / m m o l , respectively.
325 TABLE 4 Some relevant statistical parameters for the three different frequency distributions of apo D values, shown in Fig. 4 Urinary apo D
Mean Median Skewness Kurtosis
Concentration (mg/l)
Rate of excretion (/zg/min)
Mass/creatinine (mg/mmol)
2.8 2.4 0.94 0.26
2.6 2.5 0.45 - 0.49
0.24 0.23 0.46 - 0.26
50 apparently healthy normal men were determined. The apo D mean_+ SD concentrations assayed with ZIA in fresh urine samples from this group of males was 2.8 + 2.1 mg/1 (range: 0.10-8.70 rag/l). The distribution of the values is shown in Fig. 4A which demonstrates a skewness to the right. The mean _+ SD rate of excretion of urinary apo D, calculated from concentration values for urine collected during 2 h was 2.6_+ 1.8 /zg/min (range: 0.23-6.07 Izg/min) for the 50 males. The frequency distribution of these values was close to Gaussian distribution (Fig. 4B). In Fig. 4C the frequency distribution of the values for the calculated ratios between mg of apo D and mmol of creatinine is shown. The mean _+ SD for these ratios were 0.24 mg/mmol and 0.13 mg/mmol (range: 0.027-0.56 mg/mmol), respectively. Some relevant statistical parameters for the 3 different frequency distributions are compiled in Table 4.
Discussion In a preliminary study describing apo D as a urinary protein it was quantified in a small population by rocket immunoelectrophoresis [1] following concentration of the urine samples. However, concentration procedures are laborious and might cause variable recoveries of the apolipoprotein. To evaluate the utility of the apo D concentration value in native uneoncentrated urine as a parameter in clinical investigations of kidney dysfunctions we have adapted its quantification to the ZIA method of Vesterberg [6,7]. With this method there was a linear relationship between apo D zone migration distances and undiluted native urine sample volumes (Fig. 1A), indicating that the composition of the native matrix has negligible influence on the formation of the immunoprecipitates in the agarose gel rods. This property is imperative when apo D-deficient urines are studied, as they necessitate the application of large sample volumes of urine. In addition, dilution of urine samples resulted in zone migration distance values indistinguishable from those produced by the corresponding volumes of undiluted urines (Fig. 1B). Dilution may be necessary when working with samples from patients and may in addition result in normalization of the sample matrix in urine samples in respect of
326 p H and salt composition. The observed parallelism between the zone migration distance values obtained with purified apo D and serum and urinary apo D (Fig. 2B) is in agreement with previous findings using rocket immunoelectrophoresis [1]. The present study also demonstrates a good stability of apo D in acidic and alkalinized urines for up to 2 days of storage at +5°C and for several days in frozen conditions (Tables 1, 2 and 3). This is in contrast to urinary a~-microglobulin, /3e-microglobulin and retinol-binding protein, all of which have been reported to decline in concentrations at acidic pH, making alkalinization of the urine necessary to retard their degradation [11]. Analysis of apo D with Z I A yielded essentially the same values for native and alkalinized urine samples (Tables 2 and 3). Thus apo D might also be accurately estimated in samples prepared for determination of the commonly used urinary proteins for studies of kidney function and disease. Many factors may affect proteins in urine samples, such as proteolytic enzymes, changes in the unique salt balance due to change in physiological temperature, denaturation due to acidic pH, losses caused by adsorption to storage vessel walls and the presence of bacteria. As any storage time of a urine sample increases the risk of a concentration value that differs from that of a fresh sample, the apo D concentrations of the urine samples collected from the 50 normal males were analyzed within 2 h. The mean concentration value of 2.8 mg/1 for urinary apo D of the 50 males using the Z I A method is in agreement with the previously reported mean of 1.4 mg/1 of the group of 9 normal males obtained with rocket immunoelectrophoresis and samples concentrated by ultrafiltration [1]. The frequency distribution of directly estimated apo D concentration values was positively skewed (Fig. 4A). However, the distribution of rate of excretion values plotted in 0.50 /xg/min-intervals for the examined population was narrow, between 0.27 and 6.07 p~g/min, and close to Gaussian distribution with a skewness value of only 0.45 and with relatively few observations in the tails as indicated by a negative kurtosis value of - 0 . 4 9 (Fig. 4B and Table 4). This narrow distribution might reflect the small concentration range of apo D, around 100 _+ 40 mg/1, in normal human plasma [12]. A similar symmetry of distribution was obtained with the calculated ratios between mass of apo D and mmol creatinine, demonstrating a still lower numerical value, - 0 . 2 6 for kurtosis (Fig. 4C and Table 4) than that obtained for the excretion rates. The results of this study demonstrate that the Z I A method seems to be a reliable reference method for quantification of urinary apo D. Its application for the evaluation of apo D as a potential marker for kidney and metabolic dysfunctions in different clinical study populations is currently under investigation.
Simplified description of the method and its application A zone immunoelectrophoresis assay of urinary apolipoprotein D has been developed. In this quantitative immunological technique a set of 20 glass capillaries filled with agarose gel containing antibodies against apolipoprotein D is used. The apolipoprotein produces zones of immunoprecipitates with migration distances directly propotional to its concentration. Using this method the apolipoprotein
327 D concentrations in both unconcentrated and diluted urine samples could be determined. The method yielded good agreement between apolipoprotein D quantifications performed in samples originating from urine, serum and purified apolipoprotein. The special advantages of this sensitive method are: • unconcentrated native urine samples can be analyzed in a wide volume range; • the same method can be used for both serum and urine samples; • linear standard curves are produced with low amounts of antiserum. This method will have applications in studies of kidney dysfunctions and in investigations of the unknown biological and physiological roles of apolipoprotein D.
Acknowledgements o
The authors are indebted to Birgit Akerlund and Helene Saranius for skilfull technical assistance.
References 1 Holmquist, L. (1990) Identification and quantification of apolipoprotein D in normal human urine. Electrophoresis 11, 93-94. 2 McConathy, W.J. and Alaupovic, P. (1973) Isolation and partial characterization of apolipoprotein D: a new protein moiety of the human plasma lipoprotein system. FEBS Lett. 37, 178-182. 3 Dryana, D., Fielding, C.,'McLean, J., Baer, B., Castro, G., Chen, E., Comstock, L., Henzel, W., Kohr, W., Rhee, L., Wion, K. and Lawn, R. (1986) Cloning and expression of human apolipoprotein D cDNA. J. Biol. Chem. 261, 16535-16539. 4 Yu, H., Yanagisawa, Y., Forbes, M.A., Cooper, E.H., Crockson, R.A. and MacLennan, I.C.N. (1983) Alpha-l-microglobulin: an indicator protein for renal tubular function. J. Clin. Pathol. 36; 253-259. 5 Bernard, A.M., Vyskocil, A.A., Mahieu, P. and Lauwerys, R. (1987) Assessment of urinary retinol-binding protein as an index of proximal tubular injury. Clin. Chem. 33, 775-779. 6 Vesterberg, O. (1980) Quantification of proteins with a new sensitive method: zone immunoelectrophoresis assay Z. Physiol. Chem. 361,617-624. 7 Vesterberg, O. (1983) Protein quantification with zone immunoelectrophoresis assay (ZIA). In: Tschesche, H. (Ed.), Modern methods in protein chemistry-review articles, pp. 187-206, Walter de Gruyter & Co., Berlin, New York. 8 Holmquist, L. and Vesterberg, O. (1986) Quantification of human serum apolipoprotein A-I by zone immunoelectrophoresis assay and a procedure for the preparation of an A-I standard. Clin. Chim. Acta 156, 131-144. 9 Weeke, B. (1973) General remarks on principles, equipment, reagents and procedures. In: Axelsen, N.H., Kr611, J. and Weeke, B. (Eds.), A manual of quantitative immunoelectrophoresis, pp. 15-46, Universitetsforlaget, Oslo. 10 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 266-275. 11 Donaldson, M.D.C., Chambers, R.E., Woolridge, M.W. and Wicher, J.T. (1989) Stability of alphal-microglobulin, betae-microglobulin and retinol binding protein in urine. Clin. Chim. Acta 179, 73-78. 12 Curry, M.D., McConathy, W.J. and Alaupovic, P. (1977) Quantitative determination of human apolipoprotein D by electroimmunoassay and radial immunodiffusion. Biochim. Biophys. Acta 491, 232-241.
315
© 1991 Elsevier Science Publishers B.V. All rights reserved 0165-022X/91/$03.50
JBBM 00915
Quantification of apolipoprotein D in human urine by zone immunoelectrophoresis assay: a methodological and clinical study Leif Holmquist and Olof Vesterberg Division of Medical Chemistry, National Institute of Occupational Health, Solna, Sweden (Received 15 July 1991) (Revised version received 22 August 1991) (Accepted 22 August 1991)
Summary A zone immunoelectrophoresis assay (ZIA) has been developed for the quantification of apolipoprotein D (apo D) in unconcentrated native human urine. A standard curve, linear between 1 and 8 mg apo D/I was obtained with ZIA. The relative coefficients of variation for this method were 5-9% ( n = 1 5 × 6 ) with a mean_+SD of 7_+1.4% and below 11% ( n = 6 × 1 5 ) for within-run and between-run reproducibility, respectively. Equal amounts of apo D in unconcentrated and diluted urines, in serum and of the purified protein produced the same zone migration distances indicating parallelism between the immunologic reactions of apo D in different sample matrixes. Storage experiments with normal urines demonstrated good stability of apo D in both acidic and alkalinized urine over at least 2 days at +5°C and during several days at -20°C to -40°C. Using ZIA, urine samples from 50 normal healthy men aged 23-65 years were analyzed for apo D. Mean and SD were: 2.8_+2.1 mg/l, 2.6_+ 1.8 ~ g / m i n and 0.24_+0.13 mg/mmol for concentration, rate of excretion and mass/creatinine concentration, respectively. Key words: Apolipoprotein D; Normal urine; Quantification; Zone immunoelectrophoresis
Introduction Recently we reported that normal human urine contains apolipoprotein D (apo D) as one of the major proteins which was quantified by rocket immunoelecAbbreviations: HDL, high density lipoproteins; Tricine, N-tris(hydroxymethyl)methylglycine; ZIA, zone immunoelectrophoresis assay; u, 1 unit of mass (1.66 x 10 -27 kg).
Correspondence: L. Holmquist, Division of Medical Chemistry, National Institute of Occupational Health, S-17184 Solna, Sweden.
316
trophoresis [1]. This low-molecular-mass apolipoprotein was originally discovered in plasma by McConathy and Alaupovic [2] and was found to be associated with its high density lipoprotein fraction (HDL; d = 1.063-1.21 kg/l). Although well characterized, the role and function of this apolipoprotein is unknown. However, its amino acid sequence shows homology to that of a~-microglobulin and retinolbinding protein [3] both of which are urinary proteins of diagnostic value [4,5]. The physico-chemical properties of plasma apo D, which has an apparent molecular mass of 33000 u (1 u -- 1.66 × 10 -27 kg) and a documented stability to acid and urea [2] make urinary apo D a potential complement to these proteins for the study of kidney dysfunction, especially with respect to the diagnosis of tubular proteinuria. To investigate this the quantification of urinary apo D was adapted to the zone immunoelectrophoresis assay (ZIA) as described by Vesterberg [6,7]. In this method, the lateral diffusion of the antigen occurring in rocket immunoelectrophoresis is inhibited by allowing the formation of the antigen-antibody complex to develop in glass capillaries; this results in narrow immunoprecipitates with sharp zone-fronts. The migration distances of such zones are linearly related to the concentration of antigen in the sample. This technique has the further advantages over rocket immunoelectrophoresis that large sample volumes can be used, making concentration steps unnecessary, and that the zone migration distances are linearly related to the amount of the antigen even if applied in different volumes over a wide range. This communication reports the details for the quantification of apo D in unconcentrated urine by zone immunoelectrophoresis assay and its application for the determination of this urinary protein in a large group of apparently healthy normal men.
Materials and Methods
Study population and samples Samples of urine, for the evaluation of method characteristics, were obtained by midstream collection from apparently healthy males or females 22-53 years old. A portion of each urine sample was immediately alkalinized with sodium bicarbonate. The remaining urine samples were promptly chilled to 0°C, centrifuged at 4000 × g for 10 min and added to solid sodium azide (0.1 g/l). Aliquots of all the urine samples were frozen and stored for stability tests. For evaluation of apo D concentration in urine of normal subjects, samples were collected in the mornings before noon from 50 non-fasting apparently healthy men 23-65 years old. None was on treatment with drugs or diet, nor had any acute or chronic disease. They were all feeling in good health. The donor was instructed first to void without collecting the urine and note the time of the clock. After 2 h he had to void in a pre-weighed clean polyethylene flask. Thereafter the flask was transported to the analytical laboratory within 15 min. The weight, density and pH of the urine was recorded and aliquots were centrifuged at 0°C without delay and
317 sodium azide was added as described above. The quantifications of apo D in the urines were started on non-frozen samples within 2 h of collection.
Materials Authentic apo D for primary immunological standards was isolated from normal human H D L prepared by ultracentrifugation of pooled plasma from apparently healthy subjects as described by McConathy and Alaupovic [2]. The isolated apo D was electrophoretically and immunologically characterized as previously reported [1], using inter alia a reference goat anti-human apo D, obtained as a gift from P. AIaupovic, Oklahoma City, U.S.A. Monospecific polyclonal antibodies against the purified apo D were raised in albino rabbits and were the same as used in an earlier report [1]. Stock solutions of purified apo D at a 10 mg/1 concentration for preparation of primary immunological standards were made in 0.04 M Tris-Tricine buffer (pH 8.6) containing 1% (w/v) bovine serum albumin. Only glass tubes and vessels were used for stock solutions and dilutions of apo D and serum. A reference serum pool for preparation of apo D secondary standards was obtained from fasting normolipidemic healthy subjects and was isolated and treated as previously described [8]. The apo D concentration of this reference serum was estimated to be 119 mg/1 using rocket immunoelectrophoresis [9]. Aliquots of the stock solution of apo D and the reference serum were stored under a nitrogen atmosphere in capped glass vials at - 70°C. After thawing, the individual solutions were carefully homogenized before use.
General methods Total protein was determined according to Lowry et al. [10] using purified bovine serum albumin as a standard (fraction V powder, Sigma, St. Louis, MO, U.S.A.). Urine creatinine concentrations were determined at the Karolinska Hospital, Solna, Sweden, using a Kodak Ectachem 7000 instrument and the kit, Creatinine Two Point Rate Test, with the enzyme creatinine amido hydrolase (Eastman Kodak, Rochester, NY, U.S.A.). Between-run and within-run reproducibility determinations were performed during one week on a neutral native urine containing sodium azide and stored at +5°C. In each ZIA experiment, performed to establish the stability of urinary apo D on storage, a sample of a reference serum (Human Serum Protein Calibrator; Dakopatts, Glostrup, Denmark) containing apo D was used as an internal standard to normalize the zone migration distances of the urine samples of days 1 and 2 to those of day 0. The day 0 urine samples were analyzed within 1 h of sampling.
Quantification procedure The zone immonoelectrophoresis assay of Vesterberg [6,7] was adapted for quantification of apo D, using 1.3% (w/v) HSA agarose (FMC-Litex, Glostrup) in 0.04 M Tris-Tricine buffer (pH 8.6) containing 4% (w/v) polyethylene glycol 6000. The concentration of rabbit anti-apo-D (precipitated immunoglobulins) was 1 pA (44 /~g) per ml of agarose gel supplemented with 5 /.~l/ml of non-immune rabbit serum. Final dilutions of stock solutions of purified apo D and the reference serum
318
were made in Tris-Tricine buffer containing 1% (w/v) bovine serum albumin, which was added to prevent adsorption of the apolipoprotein to the vessel walls. Before application of the samples (10/~1) using a Hamilton syringe, their volumes were increased by 20% (v/v) by addition of a mixture (2 /~1) of 1 vol. 0.04 M Tris-Tricine buffer (pH 8.6), 2 vol. 80% (w/v) sucrose, and 0.2 vol. 0.5% (w/v) methylene blue in 0.04 M Tris-Tricine buffer (pH 8.6). This increased the density of the sample for safe application on the gel surface beneath the electrophoresis buffer. Urine samples or dilutions in 0.04 M Tris-Tricine buffer (pH 8.6) containing 1% (w/v) bovine serum albumin were applied in relevant volumes to give immunoprecipitates compatible with those of the apo D standards ranging 1-8 mg/1. Densities of the samples were increased as above. The quantification of urinary apo D in the study population of the 50 men was performed with ZIA using 5 different concentrations of the standard and, in addition, a sample of a control human serum containing apo D (Seronorm; Nycomed, Oslo, Norway) run simultaneously with the urine samples. Electrophoresis was run at 30 V and 2 mA for 18 h. Gels were treated as previously described [6,7] and stained in a solution of 1 g Coomassie Brilliant blue R in a mixture of concentrated acetic acid, ethanol and water (1 : 4.5 : 4.5, 200 ml) for 5-10 min. Destaining was performed in the same solution not containing the stain. A detailed description of measurement of the migration distances of the immunoprecipitate zones has been published earlier [7] and, if not otherwise stated, ZIA determinations of apo D concentrations were made in duplicate for samples and as singlets for each concentration of the standard series. Skewness and kurtosis of measured values were estimated using the program StatWiev 512 + (BrainPower Inc., Calabasas, CA, U.S.A.).
Results
A linear relationship between zone migration distances of the apo D-containing immunoprecipitates and unconcentrated urine sample volumes up to the largest tested 50-/~1 sample was obtained with the zone immunoelectrophoresis technique as shown in Fig. 1A. Experiments with dilutions of the urine resulted in zone migration distances that matched well with those expected of corresponding volumes of unconcentrated urine (Fig. 1B). Apo D isolated from H D L of plasma or present in serum also produced zone migration distances in the 1-8 mg/1 concentration range, using 10-pA samples, that were linearly related to the corresponding apolipoprotein concentrations (Fig. 2A). Within the experimental error of estimation the two different apo D standards produced the same curves (Fig. 2B). The zone migration distance values obtained with different amounts of purified apo D standard and urinary apo D were scattered along the regression line for the 10-80-ng range (Fig. 2C), indicating parallelism between the immunologic reactions in the different sample matrixes. Urinary and standard purified apo D gave immunoprecipitation zones of equal quality (Fig. 3). The coefficients of variation for the ZIA of apo D were 5 - 9 % (n = 15 x 6) with a mean + SD of 7 +_ 1.4% and below 11% (n = 6 x 15) for within-run and between-
319 3o
A E E c
20,
"o
E
10
o N
0
i
i,
10
20
i
i
I
30
40
50
60
Unconcentrated urine, pl
14
112 B
0
!
i
5
10
15
20
Urine, pI
Fig. 1. (A) Plot of zone migration distances (ram) against applied volumes (/zl) of unconcentrated native h u m a n urines obtained on ZIA. n , • : values for two different normal urine samples. (B) Plot of zone migration distances (mm) against applied volumes (/zl) of unconcentrated native, normal urine ([]) and of the same urine after dilution with Z I A buffer ( • ) . The diluted urine was applied as 50-/zl samples.
run, respectively, as estimated with 10-1 /xl samples of normal urine and a series of purified apo D standard solutions of concentrations 2, 4, 6 and 8 mg protein/1. Storage of native urine samples at + 5°C for 1 and 2 days and for several days at - 2 0 to - 4 0 ° C resulted in an average recovery of apolipoprotein D of 98%, 87% and 86% in relation to day 0, respectively, in a population of 20 samples, estimated as zone migration distances on ZIA as shown in Tables 1, 2, and 3. The acidic urines having pH values between 5.1 and 5.4 yielded essentially the same results on storage as the whole population of native urines (Tables 1, 2 and 3). In another population of 9 urine samples, storage for 1 and 2 days at + 5°C and for several days at - 4 0 ° C yielded an average recovery of apo D of 105%, 107% and 102%, respectively, as compared to freshly collected urines. Alkalinization of these native urines with sodium bicarbonate gave values for zone migration distances for apo D
320 30
E E t~o
20,
~5
10 E
=., o N 0 0
i
i
i
i
2
4
6
8
10
Apolipoprotein D, mg/I 14
B
12
°i g
I,q
0
10
~5
N 0
20
40
60
so
100
Apolipoprotein D, ng
Fig. 2. (A) Plot of zone migration distances (mm) against concentrations of purified apo D from H D L (mg/l), applied as 10-/zl samples. (For details see Materials and Methods). (B) Plot of zone migration distances (mm) against concentrations of purified apo D ( m g / l , O) and against samples of 2 different normal h u m a n sera of known apo D content, as determined by rocket immunoelectrophoresis (11) applied in amounts corresponding to the purified apo D. (For details see Materials and Methods). (C) Plot of zone migration distances (mm) against amounts of apo D (ng) for samples of 2 different normal h u m a n sera of known apo D content ([3) and for a normal h u m a n urine of 4.5 mg apo D / I (11), as determined by rocket immunoelectrophoresis. (For details see Materials and Methods).
9.6__+0.43 (4) 6.0_+0.25 (4) 13.3_+0.33 (5) 5.9 -+ 0.22 (5) 9.9_+0.41 (5) 10.7_+0.49 (5) 19.3+0.30(5) 10.8_+0.71 (4) 5.7_+0.42 (5) 7.2 _+0.82 (5) 9.6_+0.50 (5) 29.8_+ 1.76 (5) 30.7_+1.16(5) 6.6_+0.86 (5) 11.0_+0.26 (5) 2.6 _+0.49 (3) 11.5_+0.21 (2) 6.3 + 0.38 (3) 7.1 _+0.14 (2) 19.6 _+0.00 (2)
11.0_+0.28 (4) 5.9_+0.06 (4) 13.7__+0.45 (5) 5.9 -+ 0.38 (5) 7.9_+0.65 (5) 9.9_+0.41 (5) 18.3_+0.86(5) 9.2_+0.30(4) 3.7__+0.50 (5) 8.4 _+0.75 (5) 8.4_+0.69 (5) 31.4_+ 1.96 (5) 26.3_+0.79(5) 7.6_+0.37 (5) 8.2_+0.18 (5) 2.0 _+0.00 (3) 8.6-+0.19 (4) 4.9 __+0.00 (3) 6.0_+0.92 (2) 20.5 __+0.85 (2)
2 (mean_+ SD, (n))
S1
145 120 83 94 105 92 101 89 83 80 102 109 133 80 82 108 92 94 82 76
(%)
$2
69 73 69 82
83
114 118 85 94 84 85 96 76 53 93 89 115 114 92 61
(%)
8.0__+0.17 11.0_+0.57 19.0_+0.97 15.8_+0.49 3.7_+0.26 5.6 _+0.52 7.4_+0.00 25.6_+0.00 18.4_+1.31
(5/28) (5/35) (5/28) (5/15) (5/10) ( 5 / 9) (5/38) (5/25) (5/24)
(n/days)
Thawed after storage at - 4 0 ° C (mean -+ SD)
S1 and $2 represent recovery values for urines stored 1 and 2 days, respectively; $3 represents values for frozen and thawed urines.
6.6__+0.38(4) 5.0_+0.00 (4) 16.1 __+0.13 (5) 6.3 __+0.23 (5) 9.4_+0.62 (5) 11.7__+0.45(5) 19.1_+0.26(5) 12.1_+0.63(4) 6.9_+0.80 (5) 9.0 _+0.72 (5) 9.4_+ 1.03 (5) 27.4_+0.43 (5) 23.0_+0.51 (5) 8.3_+ 1.04 (5) 13.4_+0.59 (5) 2.4 _+0.20 (3) 12.5__+0.71(2) 6.7 _+0.42 (3) 8.7_+0.14 (2) 25.5 _+0.71 (2)
7.0 5.3 5.3 6.0 6.9 5.3 6.1 5.4 5.4 5.2 6.9 7.2 5.4 5.4 5.3 6.5 5.8 5.3 6.7 5.1
1 (mean_+ SD, (n))
Time of storage at +5°C (days)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
pH
0 (mean_+ SD, (n))
-
No.
Sample
-
85 94 99 130 54 62 79 93 80
(%)
$3
Zone migration distances (mm_+ SD) produced by urinary apolipoprotein D on ZIA of native urine samples after different times of storage at + 5 ° C and 20 to 40°C
TABLE 1
Native
6.3
No.
1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9
7.4
7.5
7.4
7.6
7.2
7.3
7.4
7.2
7.4
Adjusted
8.3 +_ 1.16 9.5_+0.77 9.3_+0.48 8.9_+0.21 3.0_+0.32 2.8_+0.42 12.3_+0.45 8.3_+0.33 11.0_+1.42 11.6_+0.48 8.3_+0.51 7.6_+0.27 9.5_+0.44 8.4_+0.58 10.0-+0.08 9.5 -+ 0.45 26.1 _+O.17 26.1 _+0.35
0 (mean _+SD) 10.1 -+ 0.48 10.3+0.42 9.7+0.35 9.1_+0.34 3.6_+0.24 2.8_+0.20 12.2_+0.22 9.3_+0.05 11.2_+0.20 11.1_+0.51 6.2_+0.70 5.1_+0.68 7.0-+0.50 6.0_+0.54 15.2-+0.11 13.1 -+ 1.00 26.6 + 0.98 24.8_+0.81
1 (mean _+SD)
Time of storage at + 5°C (days)
10.2 _+0.38 10.2_+0.23 8.9-+0.29 7.6+0.06 3.5-+0.46 2.9_+0.26 13.5_+0.45 11.9_+0.36 10.1+0.50 9.3_+0.44 5.9-+0.42 4.4_+0.45 9.5-+0.52 7.6_+0.47 15.1-+0.00 13.3 _+0.25 27.0 _+0.37 26.9_+0.42
2 (mean _+SD)
S1
122 108 104 102 120 100 99 112 102 96 74 67 73 71 152 138 102 95
(%)
$2
123 107 96 85 117 104 110 143 92 80 71 58 100 90 151 138 103 103
(%) 9.0 -+ 0.35 8.5_+0.17 8.9_+0.31 8.1-+0.27 4.9_+0.32 4.3-+0.35 12.4_+0.68 8.3_+0.39 7.0_+0.68 8.5_+0.52 7.6_+0.41 7.0_+0.51 11.0_+0.48 10.2-+0.35 11.1-+0.55 9.9 _+0.42 18.3 _+0.58 17.0_+0.99
(9) (9) (21) (21) (21) (21) (11) (11) (10) (10) (6) (6) (6) (6) (27) (27) (27) (27)
(days)
Thawed after storage at - 40°C (mean _+SD)
SI and $2 represent recovery values for urines stored 1 and 2 days, respectively. $3 represent values for frozen and thawed urines.
6.5
7.0
6.8
6.4
7.6
5.3
6.1
5.5
pH
Sample
108 89 96 91 163 209 101 100 64 73 92 92 116 121 111 104 70 65
(%)
$3
Zone migration distances (ram_+ SD, n = 6) produced by urinary apolipoprotein D on Z I A of native and alkalinized urine samples after different times of storage at + 5 and - 40°C
TABLE 2
ba
323
12345
6789101112
E ¢,)
+ Fig. 3. Electrophoretogram of ZIA demonstrating Coomassie-stained immunoprecipitation zones obtained with samples of purified apo D (20 to 80 ng, gel rod 1 to 5) and by a sample of unconcentrated normal human urine (20/zl, gel rod 6 to 12). For details see Materials and Methods.
that did not seem to differ from the values obtained from the corresponding native urine samples and recoveries were on average unaffected by the different conditions of storage as shown in Tables 1, 2 and 3. In order to test the practicability of the Z I A method and to obtain normal concentration values (reference population), the apo D concentrations in urine of
TABLE 3 Average recoveries of apolipoprotein D in native and alkalinized urines after different times of storage at + 5 and - 4 0 ° C (percentage_+ SD of zone migration distances in relation to corresponding values for freshly collected urine samples, calculated from the values of columns S1, $2 and $3 in Tables 1 and 2) Urine samples
Native Native Native, pH 5.1 - 5.4 Native Native Alkalinized Alkalinized
Number of samples
Condition of storage
20 9 11 9 9 9 9
+5 - 20 to - 40 +5 +5 - 40 +5 - 40
a Average time of storage.
( o C)
Time of storage (days) I 2 98 + 18.5
87 _+18.1
92+_18.1 105 _+24.3
85.+19.7 107 + 22.3
99 -+21.3
101 -+27.0
18 _+9.7 a
86_+22.1
102 _+28.8 115+55.5
324
o
I
2
1
3
4 5 6 Apo D in urine, mg/1
7
8
9
I
4 5 6 7 Excretion of apo D, ~Jg/min.
8
9
10
2
~
B 10.
8
g 8
6 4.
2, O. 0 7 .
1 .
.
2 .
.
3 .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
C
8.
5.
8
3. 2,
'
nN ,
0
.
,
.1
.
,
.2
.
,
.
,
.
,
.
,
.
,
.3 .4 .5 .6 .7 Apo D in urine, mg/mmol creatinine
.
,
.8
.
,
.
,
.9
Fig. 4. (A) Frequency distribution of urinary apo D concentrations obtained for the group of 50 normal men. The values are arranged in 0.50-mg/t intervals each including the lowest value. The mean and the median are 2.8 and 2.4 mg/1, respectively. (B) Frequency distribution of the rate of excretion of apo D obtained for the group of 50 normal men. The values are arranged in 0.5-p~g/min intervals, each including the lowest value. The mean and median are 2.6 and 2.5/.tg/min, respectively. (C) Frequency distribution of mg of urinary apo D per mmol of creatinine obtained for the group of 50 normal men. The values are arranged in 0.0125-mg/mmol intervals and each includes the lowest value. The mean and median are 0.24 and 0.23 m g / m m o l , respectively.
325 TABLE 4 Some relevant statistical parameters for the three different frequency distributions of apo D values, shown in Fig. 4 Urinary apo D
Mean Median Skewness Kurtosis
Concentration (mg/l)
Rate of excretion (/zg/min)
Mass/creatinine (mg/mmol)
2.8 2.4 0.94 0.26
2.6 2.5 0.45 - 0.49
0.24 0.23 0.46 - 0.26
50 apparently healthy normal men were determined. The apo D mean_+ SD concentrations assayed with ZIA in fresh urine samples from this group of males was 2.8 + 2.1 mg/1 (range: 0.10-8.70 rag/l). The distribution of the values is shown in Fig. 4A which demonstrates a skewness to the right. The mean _+ SD rate of excretion of urinary apo D, calculated from concentration values for urine collected during 2 h was 2.6_+ 1.8 /zg/min (range: 0.23-6.07 Izg/min) for the 50 males. The frequency distribution of these values was close to Gaussian distribution (Fig. 4B). In Fig. 4C the frequency distribution of the values for the calculated ratios between mg of apo D and mmol of creatinine is shown. The mean _+ SD for these ratios were 0.24 mg/mmol and 0.13 mg/mmol (range: 0.027-0.56 mg/mmol), respectively. Some relevant statistical parameters for the 3 different frequency distributions are compiled in Table 4.
Discussion In a preliminary study describing apo D as a urinary protein it was quantified in a small population by rocket immunoelectrophoresis [1] following concentration of the urine samples. However, concentration procedures are laborious and might cause variable recoveries of the apolipoprotein. To evaluate the utility of the apo D concentration value in native uneoncentrated urine as a parameter in clinical investigations of kidney dysfunctions we have adapted its quantification to the ZIA method of Vesterberg [6,7]. With this method there was a linear relationship between apo D zone migration distances and undiluted native urine sample volumes (Fig. 1A), indicating that the composition of the native matrix has negligible influence on the formation of the immunoprecipitates in the agarose gel rods. This property is imperative when apo D-deficient urines are studied, as they necessitate the application of large sample volumes of urine. In addition, dilution of urine samples resulted in zone migration distance values indistinguishable from those produced by the corresponding volumes of undiluted urines (Fig. 1B). Dilution may be necessary when working with samples from patients and may in addition result in normalization of the sample matrix in urine samples in respect of
326 p H and salt composition. The observed parallelism between the zone migration distance values obtained with purified apo D and serum and urinary apo D (Fig. 2B) is in agreement with previous findings using rocket immunoelectrophoresis [1]. The present study also demonstrates a good stability of apo D in acidic and alkalinized urines for up to 2 days of storage at +5°C and for several days in frozen conditions (Tables 1, 2 and 3). This is in contrast to urinary a~-microglobulin, /3e-microglobulin and retinol-binding protein, all of which have been reported to decline in concentrations at acidic pH, making alkalinization of the urine necessary to retard their degradation [11]. Analysis of apo D with Z I A yielded essentially the same values for native and alkalinized urine samples (Tables 2 and 3). Thus apo D might also be accurately estimated in samples prepared for determination of the commonly used urinary proteins for studies of kidney function and disease. Many factors may affect proteins in urine samples, such as proteolytic enzymes, changes in the unique salt balance due to change in physiological temperature, denaturation due to acidic pH, losses caused by adsorption to storage vessel walls and the presence of bacteria. As any storage time of a urine sample increases the risk of a concentration value that differs from that of a fresh sample, the apo D concentrations of the urine samples collected from the 50 normal males were analyzed within 2 h. The mean concentration value of 2.8 mg/1 for urinary apo D of the 50 males using the Z I A method is in agreement with the previously reported mean of 1.4 mg/1 of the group of 9 normal males obtained with rocket immunoelectrophoresis and samples concentrated by ultrafiltration [1]. The frequency distribution of directly estimated apo D concentration values was positively skewed (Fig. 4A). However, the distribution of rate of excretion values plotted in 0.50 /xg/min-intervals for the examined population was narrow, between 0.27 and 6.07 p~g/min, and close to Gaussian distribution with a skewness value of only 0.45 and with relatively few observations in the tails as indicated by a negative kurtosis value of - 0 . 4 9 (Fig. 4B and Table 4). This narrow distribution might reflect the small concentration range of apo D, around 100 _+ 40 mg/1, in normal human plasma [12]. A similar symmetry of distribution was obtained with the calculated ratios between mass of apo D and mmol creatinine, demonstrating a still lower numerical value, - 0 . 2 6 for kurtosis (Fig. 4C and Table 4) than that obtained for the excretion rates. The results of this study demonstrate that the Z I A method seems to be a reliable reference method for quantification of urinary apo D. Its application for the evaluation of apo D as a potential marker for kidney and metabolic dysfunctions in different clinical study populations is currently under investigation.
Simplified description of the method and its application A zone immunoelectrophoresis assay of urinary apolipoprotein D has been developed. In this quantitative immunological technique a set of 20 glass capillaries filled with agarose gel containing antibodies against apolipoprotein D is used. The apolipoprotein produces zones of immunoprecipitates with migration distances directly propotional to its concentration. Using this method the apolipoprotein
327 D concentrations in both unconcentrated and diluted urine samples could be determined. The method yielded good agreement between apolipoprotein D quantifications performed in samples originating from urine, serum and purified apolipoprotein. The special advantages of this sensitive method are: • unconcentrated native urine samples can be analyzed in a wide volume range; • the same method can be used for both serum and urine samples; • linear standard curves are produced with low amounts of antiserum. This method will have applications in studies of kidney dysfunctions and in investigations of the unknown biological and physiological roles of apolipoprotein D.
Acknowledgements o
The authors are indebted to Birgit Akerlund and Helene Saranius for skilfull technical assistance.
References 1 Holmquist, L. (1990) Identification and quantification of apolipoprotein D in normal human urine. Electrophoresis 11, 93-94. 2 McConathy, W.J. and Alaupovic, P. (1973) Isolation and partial characterization of apolipoprotein D: a new protein moiety of the human plasma lipoprotein system. FEBS Lett. 37, 178-182. 3 Dryana, D., Fielding, C.,'McLean, J., Baer, B., Castro, G., Chen, E., Comstock, L., Henzel, W., Kohr, W., Rhee, L., Wion, K. and Lawn, R. (1986) Cloning and expression of human apolipoprotein D cDNA. J. Biol. Chem. 261, 16535-16539. 4 Yu, H., Yanagisawa, Y., Forbes, M.A., Cooper, E.H., Crockson, R.A. and MacLennan, I.C.N. (1983) Alpha-l-microglobulin: an indicator protein for renal tubular function. J. Clin. Pathol. 36; 253-259. 5 Bernard, A.M., Vyskocil, A.A., Mahieu, P. and Lauwerys, R. (1987) Assessment of urinary retinol-binding protein as an index of proximal tubular injury. Clin. Chem. 33, 775-779. 6 Vesterberg, O. (1980) Quantification of proteins with a new sensitive method: zone immunoelectrophoresis assay Z. Physiol. Chem. 361,617-624. 7 Vesterberg, O. (1983) Protein quantification with zone immunoelectrophoresis assay (ZIA). In: Tschesche, H. (Ed.), Modern methods in protein chemistry-review articles, pp. 187-206, Walter de Gruyter & Co., Berlin, New York. 8 Holmquist, L. and Vesterberg, O. (1986) Quantification of human serum apolipoprotein A-I by zone immunoelectrophoresis assay and a procedure for the preparation of an A-I standard. Clin. Chim. Acta 156, 131-144. 9 Weeke, B. (1973) General remarks on principles, equipment, reagents and procedures. In: Axelsen, N.H., Kr611, J. and Weeke, B. (Eds.), A manual of quantitative immunoelectrophoresis, pp. 15-46, Universitetsforlaget, Oslo. 10 Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193, 266-275. 11 Donaldson, M.D.C., Chambers, R.E., Woolridge, M.W. and Wicher, J.T. (1989) Stability of alphal-microglobulin, betae-microglobulin and retinol binding protein in urine. Clin. Chim. Acta 179, 73-78. 12 Curry, M.D., McConathy, W.J. and Alaupovic, P. (1977) Quantitative determination of human apolipoprotein D by electroimmunoassay and radial immunodiffusion. Biochim. Biophys. Acta 491, 232-241.