Recommended method of the Japanese Association of Medical Technologists for the determination of protein in urine by HPLC

Clinica Chimica Acta 303 (2001) 95–104 www.elsevier.com / locate / clinchim

Recommended method of the Japanese Association of Medical Technologists for the determination of protein in urine by HPLC Susumu Iwata* Japanese Association of Medical Technologists, Ichigaya Hoso Building, 4 -1 -5 Chiyoda-ku, Kudan-kita, Tokyo 102 -0073, Japan Received 1 December 1998; received in revised form 1 August 2000; accepted 15 September 2000

Abstract A recommended method for the determination of total urinary proteins as a designated comparative method (DCM) is described. The method is based on gel permeation high performance liquid chromatography (HPLC). A total of 20 ml of urine diluted 5-fold with saline solution were applied to a column, eluted with 0.13 mol / l phosphate buffer (pH 7.0) at a flow rate of 1.0 ml / min at 358C, and the absorbance is monitored at 220 nm. The fraction eluted in the void volume was quantified for urinary protein, using bovine serum albumin (NIST SRM 927) as the standard. The linear range of the assay was 10–5000 mg / l, and the precision was 1.5% in two samples with concentrations of 1037 and 4338, respectively. The analytical recovery was 96–104%. Low molecular weight substances such as glucose, ascorbic acid and bilirubin did not interfere the assay. The assay was satisfactorily applied to a variety of proteins regardless of their nature or molecular weights (12.4 kDa) including b2-microglobulin and Bence Jones Proteins. Surveys in five clinical laboratories, employing different types of HPLC apparatus with the same column type, showed satisfactory measurement precision (CV 2.1%) for three urine samples with different protein concentrations.  2001 Elsevier Science B.V. All rights reserved. Keywords: Recommended method; Urine proteins; HPLC; Comparison method

1. Introduction Clinically, total protein excretion in urine is used widely in the diagnosis and follow-up of renal disorders, as an important indicator of glomerular and tubular lesions. Large differences in the value of total protein that are obtained with different measurement methods and reference materials are a matter of some concern. Of the methods used to measure total urine protein, there were several that were potentially useful as a designated comparative method (DCM) [1]: a turbidimetric method using trichloroacetic acid *Tel.: 181-33-2300-634; fax: 181-33-2211-296.

(TCA) [2], the biuret reaction assay [3], and dye methods using Coomassie brilliant blue [4] or pyogallol red-molybdenum [5]. However, each of these has limitations [1]. In 1975, Doetsh and Gadsden [6] proposed a selected method which measured the protein concentration by the reaction of copper ion with the protein. However, as a standard method this approach had serious drawbacks, including a complex procedure that requires two gel-filtration steps, poor precision, and incomplete elimination of interfering materials in the urine. In 1994, Ise et al. [7] proposed using high performance liquid chromatography (HPLC), with human serum albumin as the reference material.

0009-8981 / 01 / $ – see front matter  2001 Elsevier Science B.V. All rights reserved. PII: S0009-8981( 00 )00382-X

96

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

Based on these developments, the Standardization Committee of the JAMT established the Working Group on Standardization for the Determination of Protein in Urine (WGS-DPU) in 1994. The Group selected HPLC with ultraviolet (UV) detection as a DCM, with a gel permeation column that allowed separation of proteins with molecular weights both above and below 10 000. This assay is described herein.

2. Selection of assay method The assay method was selected with consideration of the following criteria: (1) Proteins can be separated from other low molecular weight components and be assayed specifically; (2) the differences in the detection sensitivity, depending on the kind of protein or the molecular weight, are small; (3) the assay is targeted to proteins with molecular weights of 10 000 or above; (4) the sensitivity is at least 10 mg / l; (5) the method has excellent precision; (6) the linearity is up to 5000 mg / l; and (7) the method can be used for the assessment of routine methods and the validation of secondary reference materials. Ultraviolet absorption methods, are classified into those based on absorption by aromatic amino acids (280 nm) [8] and peptide bonds (190–240 nm) [9]. The former methods produce widely varied results due to the differences in aromatic amino acid contents (tryptophan, tyrosine and phenylalanine), and have low sensitivity. The latter methods showed no marked difference in absorption according to the type and molecular weight (MW) of protein. However, these UV absorption methods can be applied only after isolation of proteins from other components. A dye-binding method was developed in the 1970s, of which, the TCA ponceau-S method [10] and Coomassie brilliant blue G-250 (CBB) method [4] are variations. A drawback of these methods is poor coloration with low-MW proteins and Bence Jones protein (BJP) due to the heterogeniety of these proteins [11]. However, these effects are smaller than those seen with the dye complex method. The CBB method was used in 11.2% of the laboratories in Japan, according to the 1999 Quality Control Survey

Report by the Japanese Association of Medical Technologists (JAMT). The dye complex method utilizes the binding of protein with a dye to form a complex. Methods based on this principle include the pyrogallol redmolybdenum (IV) complex (PR) method [5]. Similar to the dye binding method, the dye complex method has the advantages of less variation among proteins than observed with the turbidimetric method, no marked dye deposition on cuvettes or tubes as in the CBB method, easy application to automatic analyzers, and excellent precision. However, the poor coloration with low-MW and Bence Jones proteins (BJP) is a drawback [11]. Some 76.3% Japanese laboratories use the PR method (1999 JAMT QC). Consistent with the above, the dye binding and complex methods are recommended today as routine procedures, but it is difficult to use either of these as a DCM due to the above-mentioned problems. Ise et al. [7] reported a HPLC UV detection method that isolates urinary protein from other components according to molecular size, based on gel permeation chromatography. This method, adapted from the selected method of Doetsch and Gadsden [6], allows specific assay of proteins, because they are isolated from the other components. Moreover, the WGS-DPU has concluded from comparisons of protein isolation procedures using columns with different isolation mechanisms, that gel permeation provides the greatest isolation ability. However, of laboratories that perform urinary protein assays, only a limited number use this method, so we recommended that it be a DCP rather than a routine method. It can also be used for evaluation of the accuracy of routine methods and validation of secondary reference materials.

3. Assay principles The assay is performed after isolation of protein and other urinary components by gel permeation chromatography (GPC). Macromolecular proteins do not permeate into the gel and are eluted first from the column, after which other smaller molecules are eluted. Proteins are then detected according to absorption at 220 nm by peptide bonds. The protein

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

concentration is determined from the peak height in the chromatogram obtained using human serum albumin (HSA) as a reference material. HAS verified with NIST 70 g / l bovine serum albumin (BSA) standard reference materials (SRM) 927 by HPLC UV detection [7] was used as a secondary reference material.

4. Materials and methods

4.1. Apparatus The HPLC measurement system composed of pump, injector, column heater, UV detector and data processor. The injector used was either a loop type injector or an autoinjector (precision #1.0%). The temperature of the column heater was controlled to 60.1 at 358C. The bandwidth of the UV light detectors (diffraction gratings) was 8 nm, and the precision of the wavelength was 60.5 nm. A spectrophotometer with a noise level of 61.5310 25 A / s at 220 nm and a drift of 61310 23 A / h was used. The data processor must allow measurement of the peak height and confirmation of the positions of fractions. The column was an Asahipak GS-220H (Showa Denko Co., Tokyo, Japan). The column size was 7.6 mm ID3250 mm. The column had a minimum cut-off MW of 3000 and a matrix made of synthetic macromolecular hard gel. Other columns with comparable protein isolation capacity may be used.

4.2. Reagents In solution A, 20.28 g sodium dihydrogen phosphate dihydrate (Na 2 PO 4 22H 2 O) was dissolved in purified water to a total volume of 1 l. In solution B, 93.12 g disodium hydrogen phosphate, 12-hydrate (Na 2 HPO 4 212H 2 O) was dissolved to a total volume of 2 l. The eluent was obtained by adding solution A to 2 l solution B, and adjusting the pH to 7.0. The mixture was passed through a 0.22 mm membrane filter and degassed before use. A sample diluent solution of 8.5 g / l NaCl solution (saline solution) was stored in glass at room temperature. The primary reference solution was NIST SRM927 bovine solution diluted 100-fold with saline. The secondary

97

reference of 1000 mg / l HAS in saline solution (99% pure by agarose electrophoresis, fatty acid-free, globulin-free, Cat. No A3782, Sigma, USA) was certified by the present method using the primary reference solution (NIST SRM 927) as a reference.

4.3. Samples The samples were centrifuged at 20003g or above for 10 min to separate the cellular components in urine. It is recommended that urine be used after filtration through a 0.22 mm membrane filter to eliminate impurities such as bacteria. HSA solution and urine samples diluted 5-fold with saline solution and 20 ml of the diluted sample were injected into the HPLC.

4.4. Secondary reference solution Currently, there is no accepted reference material for urinary total protein assays in Japan or any other country. Therefore, we determined the purity and concentration of the standards, testing method, state and ability of the primary reference material, as below. We selected HSA with a purity of 99% by agarose electrophoresis for use as the secondary reference material. The purity was confirmed by immunoelectrophoresis. HSA powder was used after being brought to a constant weight by drying at 808C. One gram of HSA was dissolved in 1000 mg / l sodium azide in 1 l. This solution was stable for 1 year under storage at 48C. It should be warmed to room temperature before use. The concentration of the HSA secondary reference solution was evaluated by the present method using the 100-fold dilution of NIST BSA SRM927 as the primary reference.

4.5. Assay The assays were performed at an eluent flow rate of 1.0 ml / min, a detection wavelength of 220 nm, and a column temperature of 358C. After the predetermined column temperature and flow rate of the eluent has been reached, the stability at the baseline of the chromatogram was confirmed. A diluted reference solution and a urine sample (20 ml) was injected. No more than 20 samples should be assayed in any one sample series. The reference solution is

98

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

assayed in duplicate at the beginning, middle and end of a series of samples. Quantification was determined using the peak height of the urine sample and HSA reference solution.

5. Results

5.1. Selection of eluent When the protein recovery ratio with a column was studied by changing the molar concentration of the buffers at pH 7.0 and 8.0 with samples of 1000 mg / l human albumin and g-globulin, the peak area was compared with and without the column. Fig. 1 shows that the molar concentration was 0.10 mol / l at pH 8.0 and 0.13 mol / l at pH 7.0, and that the recovery ratio was nearly 100% in both cases. A phosphate buffer concentration of 0.1 mol / l or above was needed to avoid adsorption of protein by the column. The absorption change of a protein by a change in buffer pH was evaluated from pH 5.0 to 8.0, 0.1 mol / l phosphate buffer using 1000 mg / l human albumin and g-globulin as a sample. The relative sensitivity of both proteins was nearly unchanged at pH 6.0–8.0, but decreased under pH 6.0 (Fig. 2). A pH of about 7.0 was reasonable based on considerations of the pKa of the phosphate buffer,

Fig. 2. Peak area change of various proteins by change in pH.

protein denaturation and durability of the column. On the basis of these results, we selected pH 7.0 and 0.13 mol / l phosphate buffer for the assays.

5.2. Selection of the measurement wavelength A total of 1000 mg / l human albumin, bovine albumin and g-globulin samples were compared using human albumin as a reference with pH 7.0 phosphate buffer at a detection wavelength of 200– 220 nm. The convergence of the relative absorption of various protein samples is shown in Table 1. The CVs were smallest at 2.4–3.5% at 200 nm regardless of the pH of the buffer. However, the wavelength was set at 220 nm in consideration of the large absorption by the buffer and the instability of the baseline at 200 or 210 nm, and the sensitivity of the assay.

5.3. Isolation and identification of protein

Fig. 1. Protein recovery test from the column at various pH levels.

Fig. 3 compares chromatograms obtained before and after ultrafiltration of the patient’s urine. Before ultrafiltration, proteins that had passed through the column without being adsorbed appeared first at about 4.3 min, and various low- and middle-MW urinary components are eluted thereafter. In the chromatogram of the same urine sample after ultrafil-

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

99

Table 1 Assay results for various proteins relative to human albumin at pH 7.0, and detection wavelengths a Wavelength (nm)

Human albumin (SIGMA)

Human g-globulin (SIGMA)

Bovine albumin (ARMOUR)

200 210 220

100 100 100

112 107 105

101 98 99

a The relative percent concentration of other 1000 mg / l protein solutions with the concentration of 1000 mg / l human albumin (fatty acid free) solution as 100%.

tration (10 000 Da cut-off), the protein fraction of 10 000 Da was eliminated. This peak was assumed to be protein, because various reference proteins eluted at this position.

5.4. Performance of assay methods Within-run imprecision was evaluated with three different concentrations of urine samples assayed five times each. The means and CVs for 364, 452, 926, 1496 and 4477 mg / l were 1.5, 1.0, 1.0, 0.4 and 0.1%, respectively. Between-run imprecision was evaluated with two stabilized pooled urine samples. The means and CVs were 1037, 4338 mg / l and 1.20%, 1.23% (n520), respectively. Linearity was obtained up to the 5000 mg / l concentration, when

the HSA solution (10 g / l) diluted by the saline solution was measured. The analytical recovery obtained adding HSA solution which corresponded to 10% volume of the patient urine. It was 96–104% as the result which obtained the analytical recovery by respectively adding HSA of 900 mg / l and 2040 mg / dl on six patients urine (90–1070 mg / l) and four patients urine (230–610 mg / l). The effects of 5000 mg / l ascorbic acid, 20 000 mg / l glucose, 4000 mg / l creatinine, 500 mg / l creatine, 500 mg / l bilirubin and 900 mg / l hemoglobin on the assay were evaluated. The interference ratio of these components, excluding hemoglobin, was 24.8–5.8%. The effects of 100 units / ml heparin (heterogeneous structures of 4000–400 000 Da) and 50 mg / ml dextran (40 000–70 000 Da) on the assay results were evaluated. The interfering effects of heparin

Fig. 3. Comparison of chromatograms of patient’s urine before and after ultrafiltration. Ultrafiltration membrane: Ultrafree C3LGC (allows materials with a molecular weight of ,10 000 to pass).

100

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

and dextran were also negligible at these concentrations.

5.6. Comparisons of elution time and assay results according to molecular weight

5.5. Comparison of sensitivity of various assay methods

We investigated the elution time of various MW markers, as shown in Table 2. These markers were dissolved to a concentration of 1000 mg / l with saline, and substances of up to 12 400 Da, such as cytochrome C, were eluted at 4.3–4.4 min. The smaller molecular weight substances such as uric acid in urine were separated from proteins of 12 400 Da. All proteins 12 300 Da were eluted at 4.3 min, and the total protein in urine could be eluted with the same elution time.

Fig. 4 compares the sensitivity of various assay methods to each 1000 mg / l human albumin (globulin-free), a 1 -acid glycoprotein, transferrin and human g-globulin, using 1000 mg / l human albumin (fatty acid-free, purity of 99% by agarose electrophoresis) as a reference. Four assay methods, the PR method with pyrogallol red-molybdate dye (Micro TP-AR, Wako Pure Chemical Industries, Osaka, Japan), the CBB method with Coomassie brilliant blue dye (Tonein TP, Otsuka Assay, Tokushima, Japan), the benzethonium chloride method, and the TCA-biuret method, were compared. The measured value of 1000 mg / l human albumin (fatty acid-free, purity of 99% by agarose electrophoresis) by each measuring method was assumed 100%, and relative concentration (%) of each protein was obtained. The sensitivity of all methods to human albumin (globulin-free) was close to 100%. However, sensitivity to a 1 -acid glycoprotein was 24–76%, transferrin 83–114% and human g-globulin 70–105%. The smallest variation in the sensitivity among proteins was obtained with the present method.

5.7. Reactivity of urine of patients with multiple myeloma vs. various methods The relationships of the results of the HPLC assay and other assay methods were evaluated using urine from eight patients with multiple myeloma (Table 3). With the value obtained by the present method taken as 100%, the reactivity ratio of the PR method was 41–72% and the CBB method 50–79%; the results of these widely-used routine methods were all lower than the values obtained by the present method. However, the reactivity of the present method to the k and l types of BJP was not different than that of the routine methods.

Fig. 4. Comparison of detection sensitivity of proteins by various assay methods.

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

101

Table 2 Comparison of assay results and elution time of various molecular weight markers a Molecular weight marker

Molecular weight

DNP-L-Alanine Cytochrome C Ribonuclease C Myoglobin Chymotrypsinogen A Albumin (egg) Albumin (bovine) Aldolase (rabbit) Catalase (bovine) Ferritin (horse) a

Present method

255 12 400 13 700 17 800 25 000 45 000 67 000 160 000 240 000 450 000

Pyrogaroll red method

CBB method

Elution time (min)

(mg / l)

(mg / l)

(mg / l)

– 4.5 4.5 4.4 4.3 4.3 4.3 4.3 4.3 4.3

–

50 3410 1580 7460 6180 490 1220 1010 1310 380

50 1250 130 1560 3610 710 1280 700 760 190

5189 484 8650 11 350 1344 1518 1183 1022 453

Molecular weight marker: Dalton Standard MS-II.

Table 3 Comparison of the results of assay of patients’ urine with multiple myoloma by various methods a Protein type

Albumin content

Present method

Progaroll red method

CBB method

BJP-l BJP-k, IgG-k BJP-k RJP-l BJP-l BJP-l

76 41 55 43 15 1

1004 181 2246 4311 1882 3665

582 130 1585 1750 1230 2195

791 119 1392 2312 1030 1844

a Units: mg / l. Albumin was assayed by the immuno-turbidimetric method.

5.8. Comparison of the detection of various patients urine by various assay methods Urine samples from 13 patients were measured by the present method; PR method and CBB method are shown in Table 4. The albumin and globulin ratio (A / G) was determined by the total protein concentration using the present method and the albumin concentration by an immune turbidimetric method. The total protein value with the PR method and CBB method was smaller than with the present method with lower A / G ratio (A / G,1.0) samples. However, the total protein value obtained with the two dye

Table 4 Comparison of detection of low-molecular weight proteins by various methods a Patient disease

A/G ratio

Present method

Progaroll red method

CRB method

Chronic renal failure Chronic nephritis After kidney transplantation Chronic glomerulonephritis Acute renal failure Gestosis Metastatic adrenal gland tumor Chronic nephritis After kidney transplantation Collagen disease Diabetic nephropathy Kidney dysfunction Chronic renal failure

1.7 0.9 0.7 1.8 2.0 2.8 0.1 1.6 0.7 1.5 0.8 0.6 0.5

3262 6324 373 225 802 1776 402 5756 478 1810 3333 637 522

3122 6322 305 224 752 1785 353 5575 392 1683 3133 541 360

3082 6273 421 239 801 1691 386 5331 362 1704 2966 473 385

a

Units: mg / l.

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

102

method was close to the value with the present method with high A / G ratio (A / G,1.0) samples. With dye method for globulin rich-samples with a poor sensitivity, the reaction of the dye method to globulin was poor.

(casual urine) of 125 healthy males (15.5–41.2 years) and 115 healthy females (16.3–42.0 years) was 38– 74 mg / l and 31–56 mg / l, respectively, (36–66 mg / l for all 224 subjects, 15.8–41.5 years). A difference was observed between males and females, with the males having higher values.

5.9. Correlations between routine methods Urine samples of 45 patients were assayed to examine the correlation of the results of the present method with those of routine methods. The correlation coefficient between the CBB method ( y), PR method ( y) and present method (x) were r50.993, y50.91x220 and r50.993, y50.96x233, respectively. The two routine methods showed greater negative proportional and constant systematic errors than did the present method.

5.10. Inter-laboratory precision Table 5 shows the results of a mini-survey conducted by the WGS-DPU, in which three urine samples from a patient with a protein concentration of 387.0–2791.0 mg / l were assayed using the same column (Asahipak GS-220H) on different measurement instruments (Shimadzu, Toso, Waters) at five laboratories. The urinary protein concentrations were 387.0, 1090.0, and 2791.0 mg / l, and the CV values (%) at the laboratories were 1.8, 2.1 and 1.0, respectively.

5.11. Reference interval The reference interval obtained from the urine

Table 5 Results of a small interlaboratory survey at five laboratories a Laboratories

Sample 1

Sample 2

Sample 3

A B C D E

382 399 387 384 383

1113 1079 1117 1067 1074

2751 2777 2792 2822 2813

Mean S.D. CV(%)

387.0 7.0 1.8

1090.0 23.3 2.1

2791.0 28.5 1.0

a

Units: mg / l.

6. Discussion In selecting a comparative method for the measurement of total urine protein, the Working Group searched for those methods that met all seven of the criteria we established for any proposed comparative method. We selected the present method in which low MW components of urine other than protein were separated out using a gel permeation column, and all high MW proteins were eluted out at the same time. The column selected, an Asahipak GS220H, eluted proteins above 10 000 Da at the same time. From measurements of protein MW markers used to measure MW with this column (Table 2), it was found that nine types of protein, with molecular weights ranging from 12 400 (cytochrome c) to 450 000 Da, all eluted at approximately the same time. Because the recovery ratio with this column was thought to decline due to the adsorption of protein to the base material in the column, we found that by raising the molar concentration of the phosphate buffer to 0.13 mol / l, almost 100% of the main protein components in urine, albumin and g-globulin was be recovered (Fig. 1). The biuret method [12] and UV spectrophotometric method [13] are known to exhibit small difference depending on the type of protein. Due to the low color sensitivity reported in studies of the biuret method, however, this method cannot be considered sufficient for the detection of urine protein. Mayer et al. [9] reported little difference in sensitivity and the detection of different types of protein with low wavelengths, so we conducted a comparison at 200, 210 and 220 nm. Using the 0.13 mol / l phosphate buffer selected for this method, the difference in detection of various types of protein was smallest at 200 nm, consistent with previous reports [9]. However, the absorption of the buffer itself was also large at 200 nm, which affected the linearity. Considering both the differences in de-

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

tection of various proteins and the linearity, we selected a wavelength of 220 nm. The result was that, with urine diluted 5-fold in physiological saline, superior quantification characteristics were obtained, including linearity up to 5000 mg / l, and a minimum detection sensitivity of 10 mg / l. The degree of variation in detecting samples having the same concentrations of various proteins such as albumin and globulin was 6% (Table 1). This was probably because the wavelength of 220 nm was chosen near the absorption wavelength of the peptide bonds. Proteins undergo a structural change due to pH, and changes in the molar absorption coefficient [14]. Using a pH 5.0–8.0 phosphate buffer, we investigated the changes in relative absorbance of various proteins, and found that at 220 nm and pH 6.0–8.0, the absorbance of various proteins was the same. In the present study, a pH of 7.0 was selected. From an examination of the conditions for the measurement, the quantitative characteristics of the method ultimately selected fulfilled the requirements for a comparative method with regard to precision, linearity, and specificity. There is no international standard material for urine protein. NIST SRM927 is the standard material for bovine albumin, but not for human albumin. Though the certified value of the albumin of SRM470 is shown, total amount of protein including the other protein has not been demonstrated. Given this situation, we decided to use the diluted primary standard material of NIST SRM927 as a more practicable standard material. As a secondary standard material that can be used in routine tests, we certified the secondary standard with the present method using NIST SRM927 as a primary standard. Using a commercial human serum albumin sample with a purity of 99%, we confirmed the manufacturer’s stated purity by electrophoresis and immunologic methods. From the present experimental results, it was found that proteins .12 000 Da were eluted at the same time and detected, and that low MW components of urine could be completely separated in the column. Therefore, it was possible to quantify total protein in the urine. With the present detection wavelength, the specific absorbancies of proteins were not used, so it is possible that errors will occur in specimens in which the protein is bound with a

103

drug or other substance. However, because there was a high molar concentration of 0.13 mol / l in the buffer, it is assumed that agents bound to protein by ion bonds will separate to become free low molecular weight components. This is assumed because the drugs released in the urine will be mainly composed of metabolites and aqueous conjugated drugs. Measuring the urine of patients with various conditions using the present method and the routine dye methods and the albumin and globulin ratio (A / G ratio), the differences in the present method and dye methods can be seen for patient urine with a A / G below 1 (Table 4). In Table 3, the difference between the two methods is conspicuous in the urine of myeloma patients. The reason is that abnormal immunoglobulins such as BJP cannot be sufficiently detected using the dye methods. The present method, however, can properly detect then abnormal proteins. Moreover, the results using different methods to measure specimens with the same concentrations of various protein samples in urine (Fig. 4) show that the present method has superior detection sensitivity in all protein specimens. An investigation of the degree of variability between five laboratories showed excellent accuracy, with a CV% of about 2%, indicating its practicality and usefulness. Thus, the method developed in the present study would seem to be useful as a means to evaluate the routine methods used to quantify urine protein.

Acknowledgements This work was funded by the Working Group on Standardization for the Determination of Protein in Urine and cooperation was received from the Japanese Association of Medical Technology. The members of Working Group on Standardization for the Determination of Protein in Urine, Japanese Association of Medical Technology. Chairman: Masaaki Ochi (Ehime University Hospital), author: Susumu Osawa (Chiba University Hospital), Toshiaki Shibou (Municipal Asahikawa Hospital), Nobuhiro Sakuma (Tohoku University Hospital), Isamu Shimada (Gichi Medical School Hospital), Masahiro Daimon (Omiya Medical Center, Gichi Medical School Hospital), Keiko Ise (Chiba University Hospital),

104

S. Iwata / Clinica Chimica Acta 303 (2001) 95 – 104

Kiyoko Kanamori (Tokyo Medical and Dental University Hospital), Masashi Chiba (Tokyo Metropolitan Hachioji Children’s Hospital), Akitomo Naito (Kinki University Hospital), Nobuya Sakamoto (Kumamoto Regional Medical Center). The WGS-DPU, JAMT, was established with the co-operation of the following: Chizuru Aoki (Yoshida Hospital), Shingi Fujino (Omiya Medical Center, Gichi Medical School Hospital), Masanori Seimiya (Chiba University Hospital), Shiro ligima (Kyoritsu Pharmaceutical University), Kengi Tokunaga, Tatsuya Nishimiya (Ehime University Hospital), Junko Kawano, Miyuki Miyazaki (Kumamoto Regional Medical Center), Kuniaki Tokuda, Masako Shiogiri (Wako Pure Chemical Industries Ltd.), Kazumi Fumihiko, Asahi Hidetoshi (Eiken Chemical Ltd.).

References [1] Hohnadel DC, Koller A. Urine protein, total. In: Pesce AJ, Kaplan LA, editors, Methods in clinical chemistry, St. Louis: Mosby, 1987, pp. 35–45. [2] Shahangian S, Brown PI, Ash OK. Turbidimetric measurement of total urinary proteins: a revised method. Am J Clin Pathol 1984;81:651–4. [3] Savory J, Pu PH, Sunderman Jr. FW. A biuret method for determination of protein in normal urine. Clin Chem 1968;14:1160–71.

[4] McIntosh JC. Application of a dye-binding method for the determination of protein in urine and cerebrospinal fluid. Clin Chem 1977;23:1939–40. [5] Watanabe N, Kamei S, Ohkubo A et al. Urinary protein as measured with a pyrogallol red-molybdate complex, manually and in a Hitachi 726 automated analyzer. Clin Chem 1986;32:1551–4. [6] Doetsch K, Gadsden RH. Determination of urinary total protein by use of gel filtration and a modified biuret method. Clin Chem 1975;21:778–81. [7] Ise K, Seimiya M, Osawa S. Development of a reference method for total protein in urine by high performance liquid chromatography. Igaku Kensa 1994;43:994–1002. [8] Layne E. Spectrophotometric and turbidimetric methods for measuring proteins. In: Methods in enzymology, New York: Academic Press, 1980, pp. 447–52. [9] Mayer MM, Miller JA. Photometric analysis of proteins and peptides at 191–194 mm. Anal Biochem 1970;36:91–100. [10] Meola JM, Vargas MA, Brown HH. Simple procedure for measuring total protein in urine. Clin Chem 1977;23:975–7. [11] Ochi M, Tokunaga K, Nishiyama T et al. Urinary protein measurement by HPLC for reference method. [Jap] Shikoku J Clin Chem 1996;12:143–54. [12] Gornall AG. Determination of serum proteins by means of the biuret reaction. J Biol Chem 1949;177:751–66. [13] Waddell WJ, Hill C. A simple ultraviolet spectrophotometric method for the determination of protein. J Lab Clin Met 1956;48:311–4. [14] Rosenheck K, Doty P. The far ultraviolet absorption spectra of polypeptide and protein solutions and their dependence on conformation. Proc Natl Acad Sci USA 1961;47:1775–85.