Automated turbidimetric benzalkonium chloride method for measurement of protein in urine and cerebrospinal fluid.
Albumin fraction V (bovine), disodium molybdate x 2 [H.sub.2]O, NaOH, sodium azide, sodium benzoate, potassium oxalate, and succinic acid were obtained from Merck; BTC, Coomassie Brilliant Blue G 250, EDTA, and PYR were from Sigma; BC was from Serva; and phosphoric acid was from Riedel de Haen. A commercially available PYR reagent (M-TP) and protein calibrators were obtained from Beckman Coulter. A globulin solution containing mostly IgG and IgM and no albumin was obtained from the Refik Saydam Center of Hygiene, Turkey.
We developed a new method that uses the same principle as the method of Iwata and Nishikaze (3) but with BC instead of BTC. BC is a detergent and causes protein denaturation in an alkaline medium. The NaOH/EDTA reagent (reagent 1) contained 33 mmol/L EDTA and 0.5 mol/L NaOH. The BC solution (5 g/L; reagent 2) was prepared from a 50 g/L aqueous stock solution of BC. Calibrators containing 50, 100, 200, 500, and 1000 mg/L protein were prepared from a Beckman serum calibrator (Multical). In a manual procedure, we mixed 2.0 mL of NaOH/EDTA with 200 [micro]L of sample and immediately added 0.5 mL of BC reagent. After a 10-min incubation, we measured the absorbance at 450 nm. We then adapted our method to a Synchrone LX 20 Pro analyzer, using the following conditions: reagent 1 volume, 200 [micro]L; reagent 2 volume, 50 [micro]L; sample volume, 20 [micro]L; primary wavelength, 470 nm; secondary wavelength, 700 nm; blank reading (sample blank), 48-72; reaction time, 10 min; reaction reading, 700-748; calibrators, 50, 100, 200, 500, and 1000 mg/L.
We compared the performance characteristics of our method with those of the Bradford CBB assay (7), the PYR assay (14), and the BTC assay (3). We adapted the PYR and BTC methods to the same analyzer and used the Beckman Coulter M-TP assay, which is also based on the PYR-molybdate method, for comparison. It has previously been demonstrated that the PYR-molybdate method is analytically and clinically comparable to the reference biuret method (15). We measured the total protein concentrations in 100 urine and 88 CSF samples simultaneously by each of the methods. Urine samples were assayed on the day of collection after centrifugation, and CSF samples were stored at -20[degrees]C before analysis. Deming regression analysis, the Pearson correlation coefficient between the methods, and all other statistical calculations were performed with the Microsoft Excel Analyze-it program. The study was approved by the Local Ethics Committee of our hospital.
Samples containing 100-5000 mg/L protein were prepared by adding known amounts of protein calibrators to a urine pool. The PYR assay was linear up to 3000 mg/L; the BC, BTC, and M-TP assays were linear up to 1500 mg/L; and the CBB assay was linear up to 1000 mg/L. Serial dilutions of a sample containing 200 mg/L protein were prepared (range, 6.25-200 mg/L), and a protein concentration that produced an absorbance +3 SD above that for a zero calibrator (n = 21 replicates) was accepted as the detection limit of the method. The detection limits were 50 mg/L for the CBB, PYR, and M-TP methods and 25 mg/L for the BTC and BC methods.
The results of our imprecision studies (n = 21) and for recovery of protein added to urine and CSF are shown in Table 1. Albumin and globulin solutions at the same concentration were prepared, and absorbance ratios were calculated (Table 1). We measured the total protein concentrations of 100 urine and 88 CSF samples in duplicate by the five methods simultaneously and averaged the duplicates for each method. Deming regression and Pearson correlation analyses were performed. We also adapted the nephelometric BC method described by Shephard and Whiting (17) to the Beckman Coulter Immage nephelometer and determined its performance characteristics separately. The nephelometric BC method was linear at protein concentrations of 50-1000 mg/L. Within-run CVs were 5.7%, 5.1%, and 3.6% for samples with mean protein concentrations of 127, 497, and 844 mg/L, respectively. Between-run CVs for the same samples were 12%, 5.7%, and 5.9%, respectively.
Methods for determining total protein in urine and CSF samples have different advantages and disadvantages. Biuret-gel-filtration or biuret-precipitation methods are accurate, but they are time-consuming and impractical for routine use. Dye-binding methods using CBB and PYR are rapid, simple, and suitable for automation (15,18,19) but tend to give different responses to different types of proteins (9-13,18). The BTC assay can also be automated (4). In our study, the PYR method was linear up to 3000 mg/L and seemed to be more advantageous than the other methods, but albumin-globulin response studies showed that globulins produced 50% lower responses than did albumin; adding sodium dodecyl sulfate to the reagent decreased this difference to 20%, but this modification also decreased the upper limit linearity of the method to 1500 mg/L. The M-TP, BTC, and BC methods had also 20% lower responses to globulins, whereas the CBB method had a 30% difference, although sodium dodecyl sulfate was added as suggested (20). In the imprecision studies, the BC method had the lowest within- and between-run CVs. In between-run imprecision studies, the BC method had a significantly lower CV than the BTC method, and this was thought to be related to reagent stability. This finding is particularly important because the BTC and BC methods are based on the same principle. The BC method is also more advantageous than the BTC method in a manual procedure because the BC method has a 10-min incubation time at room temperature, whereas the BTC method requires incubation for 40-60 min (3,21) at room temperature.
For urinary total protein, the 150-200 mg/L concentration range is important because this range is used for some clinical decisions. In this concentration range, the BC assay measured 100% of the expected protein concentration. Marshall and Williams (22) compared the Sigma PYR and CBB assays in their study and obtained the equation: CBB = 0.685PYR + 17 mg/L. They proposed using urinary protein as the calibrator and found that it increased the agreement between the methods (slope, 0.972; y-intercept, -16 mg/L). We also found a 25% difference between the CBB and M-TP methods in urine samples but good agreement in CSF samples. Our results showed that, for urine samples, the results obtained with the CBB method do not agree with those obtained with the M-TP, BC, BTC, or PYR methods. The comparison studies showed that BC correlates better with the M-TP method than do the BTC and CBB methods for urine and CSF samples, based on the slope, intercept, and correlation (r) values.
BC is a widely used disinfectant and is much less expensive than the other reagents, especially the dyes CBB and PYR. It also inhibits bacterial contamination, and this can be a reason for the better reagent stability. Shephard and Whiting (17) compared four protein precipitants--BC, BTC, trichloroacetic acid, and sulfosalicylic acid for nephelometry--and found that BC had good precision and agreed most closely with the trichloroacetic acid-Ponceau S dye-binding method. In our hands, the nephelometric method of Shephard and Whiting was linear 50-1000 mg/L protein, similar to the measuring range (78-1250 mg/L) reported by those authors (17). In imprecision studies, the precision and linearity of our method were better than those of the nephelometric method. Nephelometers are also expensive and are not used in many in small- and medium-sized clinical laboratories.
In conclusion, we present a new turbidimetric method that correlates better with the M-TP method than do the BTC and CBB methods; is more precise than the M-TP, CBB, PYR, and BTC methods; has better precision and recovery in the critical decision concentration range; has satisfactory albumin-globulin response and analytical range; uses less expensive and accessible chemicals; and is suitable for automation.
(1.) Shahanigan S, Brown PI, Ash K0. Turbidimetric measurement of total urinary proteins: a revised method. Am J Clin Pathol 1984;81:651-4.
(2.) Henry RJ, Sobel C, Seglove M. Turbidimetric determination of proteins with sulfosalicylic and trichloroacetic acids. Proc Soc Exp Biol Med 1956;92: 748-51.
(3.) Iwata J, Nishikaze O. New micro-turbidi metric method for determination of protein in cerebrospinal fluid and urine. Clin Chem 1979;25:1317-9.
(4.) McDowell TL. Benzethonium chloride method for proteins adapted to centrifugal analysis. Clin Chem 1985;31:864-6.
(5.) Savory J, Pu PH, Sunderman FW. A biuret method for determination of protein in normal urine. Clin Chem 1968;14:1160-71.
(6.) Doetsch K, Gadsden RH. Determination of total urinary protein, combining Lowry sensitivity and biuret specificity. Clin Chem 1973;19:1170-8.
(7.) Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 1976;72:248-54.
(8.) Lott JA, Stephan VA, Pritchard KA. Evaluation of the Coomassie Brilliant Blue G-250 method for urinary protein. Clin Chem 1983;29:1946-50.
(9.) Read S, Northcote D. Minimization of variation in the response to different proteins of Coomassie Blue G dye binding assay for protein. Anal Biochem 1981;116:53-64.
(10.) Duhamel R, Meezan E, Brendel K. The addition of SIDS to the Bradford dye-binding protein assay, a modification with increase sensitivity to collagen. J Biochem Biophys Methods 1981;5:67-74.
(11.) Macart M, Gerbaut L. An improvement of the Coomassie Blue dye binding method allowing an equal sensitivity to various proteins: application to cerebrospinal fluid. Clin Chim Acta 1982;122:93-101.
(12.) Macart M, Gerbaut L. Evaluation of an improved Coomassie dye binding method for urinary protein assay. Clin Chim Acta 1984;141:77-84.
(13.) Perini JM, Mizon C, Dehon B, Fialdes P, Lefebvre I, Mizon JPR. Urinary protein determination using Coomassie Brilliant Blue in the presence of sodium dodecyl sulphate. Clin Chim Acta 1984;143:321-8.
(14.) Fujita Y, Mori I, Kitano S. Color reaction between pyrogallol red-molybdenum (IV) complex and protein. Bunseki Kagaku 1983;32:E379-86.
(15.) Watanabe N, Kamei S, Ohkubo A, Yamanaka M, Ohsawa S, Makino K, 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.
(16.) Pesce MA, Strande CS. A new micromethod for determination of protein in cerebrospinal fluid and urine. Clin Chem 1973;19:1265-7.
(17.) Shephard MDS, Whiting JW. Nephelometric determination of total protein in cerebrospinal fluid and urine using benzalkonium chloride as precipitation reagent. Ann Clin Biochem 1992;29:411-7.
(18.) Heick HMC, Begin-Heick N, Acharya C, Mohammed A. Automated determination of urine and cerebrospinal fluid proteins with Coomassie Brilliant Blue and the Abbott ABA-100. Clin Biochem 1980;13:81-3.
(19.) Sano K, Kanamori K, Shiba A, Nakao M. Automated assay of urinary protein using Coomassie Brilliant Blue G-250. Anal Biochem 1981;113:197-210.
(20.) Lim CW, Chisnall WN, Stokes YM, Pratt R, Crooke MJ. Effects of sodium dodecylsulphate, dye concentration and paraprotein on Coomassie Blue dye binding assays for protein in urine. Clin Biochem 1988;21:277-81.
(21.) Flachaire E, Damour 0, Bienvenu J, Aouiti T, Later R. Assessment of the benzethonium chloride method for routine determination of protein in cerebrospinal fluid and urine. Clin Chem 1983;29:343-5.
(22.) Marshall T, Williams KM. Total protein determination in urine: Elimination of a differential response between the Coomassie Blue and Pyrogallol Red protein dye binding assays. Clin Chem 2000;46:392-8.
Fatma Meric Yalmaz, Nermin Celebi, and Dogan Yucel * (Ministry of Health, Ankara Hospital, Clinical Biochemistry Laboratory, Ankara 06340, Turkey; * author for correspondence: fax 90-312-3621857, e-mail email@example.com)
Table 1. Performance characteristics and comparison of the methods. A. Imprecision, recovery, and globulin albumin response studies Within-run CV, % Between-day CV, % Assay Low Medium High Low Medium High CBB 5.46 2.06 1.74 15 10.7 8.24 BTC 2.85 1.28 0.79 16 12.7 4.8 M-TP 3.3 1.62 1.3 13.1 8.5 6.2 PYR 2.89 1.8 1.52 12.1 6.3 2.4 BC 2.63 0.91 0.73 8.6 5.6 4.8 Response, % Mean (SD) G/A (a) ratio Assay Low Medium High (n = 11) CBB 120 91 110 0.7 (0.04) BTC 111 97 107 0.8 (0.03) M-TP 89 80 104 0.8 (0.02) PYR 106 90 109 0.5 (0.01) BC 100 90 101 0.8 (0.02) B. Deming regression analysis (b) Slope (95% CI) Intercept (95% Correlation (r) CI), [micro]g/L Urine CBB 0.78 (0.73-0.83) 77 (47-107) 0.947 BTC 1.18 (1.12-1.24) 45 (9.3-80) 0.966 PYR 0.96 (0.93-0.98) -2.5 (-17-12) 0.990 BC 1.08 (1.05-1.12) 8.4 (-13-30) 0.985 CSF CBB 0.99 (0.94-1.04) -69 (-181-43) 0.975 BTC 0.98 (0.95-1.01) 2.8 (-74-80) 0.988 PYR 1.01 (0.99-1.03) 24 (-29-77) 0.994 BC 0.99 (0.95-1.01) -44 (-119-30) 0.989 (a) G/A, Globulin-to-albumin response (reactivity) ratio; CI, confidence interval. (b) M-TP was used as the comparison method.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Technical Briefs|
|Author:||Yilmaz, Fatma Meric; Celebi, Nermin; Yucel, Dogan|
|Date:||Aug 1, 2004|
|Previous Article:||Direct nonisotopic assay of 3-methylglutaconyl-coa hydratase in cultured human skin fibroblasts to specifically identify patients with...|
|Next Article:||Serial analysis of plasma proteomic signatures in pediatric patients with severe acute respiratory syndrome and correlation with viral load.|