Determination of albumin in caprine serum

Research in Veterinary Science 1986, 41, 82-84

Determination of albumin in caprine serum M:

s. BAIN, Ministry of Agriculture,

Lincoln

Fisheries and Food, Veterinary Investigation Centre Riseholme ' ,

Standards

Dye-binding techniques using bromocresol green and bromocresol purple (BCP) were compared with cellulose acetate electrophoresis for the determination of caprine serum albumin. When a caprine standard was used the electrophoretic values showed little difference from those obtained with BCG. However, when a bovine standard or BCP were used, .there were differences. Bias was detected between the values determined with bovine and caprine albumin standards owing to the different absorbances of the dye-protein mixtures. Using BCG, a closer correlation was obtained between the electrophoretic and caprine standard values than between the electrophoretic and bovine standard values but with BCP the reverse was true although these differences were not significant. BCP was considered a less suitable dye than BCG for routine analysis because of low absorbance changes and the resulting poor precision. (BCG)

Bovine and caprine albumin were obtained as dry Kohn Fraction V powder (Sigma). Samples were dried carefully and a series of dilutions made in deionised water. Bovine and caprine albumin standards were used to produce working curves for the serum samples and to test precision within and between assays. The bromocresol green (BCG) assay was performed using the A-Gent kit (Abbott Laboratories) and based on the method of Rodkey (1965) as modified by Doumas et al (1971) using a 1:200 dilution of serum with reagent. A bromocresol purple (HCP) working solution containing HCP 0'0324 g litre-I, sodium acetate ,10 g litre:", acetic acid I· 5 ml litre", brij 35 (250 g htre- I), 1'0 ml litre "' and sodium azide 0'1 g litre-I at pH 4· 2 (BDH Chemicals) was used in a I :50 dilution of serum with reagent. Dye-binding assays were performed on an Abbott ABA 100 bichromatic discrete analyser. Total protein was determined using the biuret reaction according to the method of Gornall et al (1949), modified for use with the Abbott ABA 100 discrete analyser. Electrophoresis was performed using the Gelman system on cellulose acetate membranes (Gelman Sciences) which allows eight samples to be run per strip including the serum pool as a control. Three strips were electrophoresed in each tank at 200 volts for 20 minutes in a hippurate buffer (pH 8'6) which allows a separation into five protein fractions. The protein fractions were fixed and stained in a solution of Ponceau S (5 g litre - I) in 75 g litre -I trichloracetic acid in water. The fractions were quantitated in a Beckman R-112 densitometer at 550 nm with a Beckman Rl15 microprocessor attachment (Beckman RIIC). This integrates the curve automatically and calculates the absolute values from the total protein con.tent obtained as. described above. Intra-assay v.anance was determmed by using multiple applicatl~ns of the serum pool and inter-assay variance by using the serum pool in each strip. Because cellulose acetate was used as the support medium a transmittance densitometer could be used to quantitate the fractions; as only the albumin

OF all the proteins in blood plasma, one of the most important is .albumin. It has a role in maintaining plasma onconc pressure, it is a nutrient and it acts as a carrier of vitamins, hormones and minerals; it is also a major contributor to the binding and transport of drugs and toxic metabolites. The concentration of albumin in plasma or serum can be used to assess an animal's protein intake, its degree of hydration and its liver function, and it is also important in the eval~ation o.f ot,her plasma constituents, particularly calcium, which IS 50 per cent albumin-bound. The purpose of this study was to determine the cause of a discrepancy between the concentrations of caprine serum albumin determined by two different dye-binding techniques and to establish the effect of using bovine albumin controls and standards. Materials and methods

Samples Blood was obtained from clinically normal goats at monthly intervals. Serum was separated within six hours of sampling and stored at - 20°C until required. For the experimental comparisons serum from 40 animals was mixed and pooled. 82

Albumin in caprine serum fraction was required a separation into five fractions was considered to be adequate.

Statistical evaluation Data obtained by the different methods were analysed statistically to determine the precision of the assays. The relationships between the methods were determined by calculating the correlation coefficients and by using Student's t test. Results

83

O' 886 for the values derived from a standard curve obtained with caprine albumin standards (Fig la). Using the BCP reagent, the bovine standard values had a correlation of O' 835 and the caprine standard values a correlation of O' 820 (Fig 1b). All the correlations were highly significant (P
The precision of the electrophoretic assays was determined by repeated measurements on a serum pool, since it was not possible to determine the standards quantitatively. The densitometer is designed to evaluate the percentage contributions of the fractions and hence calculate their concentration from a determination of total protein. These Discussion measurements gave an intra-assay coefficient of Electrophoresis has been employed as a reference variation (cv) of O' 97 per cent and an inter-assay cv method by Allerhaud et al (1967), Hernandez et al of 6 per cent. (1967) and Webster et al (1974) for the measurement The biuret determination of total protein produced linear responses within the range 20 to 80 g litre- J. Bovine and caprine standards produced virtually Dye identical linear responses each with a cv of 2 to 3 per binding cent. The biuret inter-assay mean was used to produce 50 a standard curve from which the total protein concentration in an individual animal's serum was I obtained. These values were then used to calculate the ~ 40 Cl Caprine albumin concentration from the electrophoretic c: Y = 0·95X + , ·96 '" 0 30 measurements. ';; R = 0·886 ~ The absorbance of the bovine and caprine albumin E standards using the BCG reagent displayed a Bovine u 20 y = 0·75X + 2·63 0 difference in dye-binding, as shown by Keay and uc: R = 0·852 Doxey (1983), but both produced linear responses in 10 the range 20 to 60 g litre - J, with a cv of less than 5 per cent. The absorbance of the bovine and caprine albumin standards with the BCP reagent showed some loss of (b) BCP 50 linearity above 60 g litre - J, which was almost certainly due to the high reagent concentration used. I ~ 40 A linear response was obtained with a I:200 dilution ~Cl Caprine (as for BCG). However, even at the higher concentray = 0·91X + 1·34 '"c:0 30 tion, the colour change was very slight and the .~ R = 0·820 precision was much poorer than with BCG (cv up to IO E 20 Bovine per cent), particularly with low concentrations of uc: 0 y = t-t tx + 2·48 bovine albumin. u 10 R = 0·835 Despite the relatively high cv (6 per cent interassay) of the electrophoretic method and the additional variance of the biuret method for total protein 10 20 30 40 50 60 an attempt was made to test which dye gave the most Concentration g litre -1 Electrophoresis accurate determination of plasma albumin concentraFIG 1: Comparison of the dye-binding methods and electrophoresis tion. Using the BCG method a correlation of O:852 was for the determination of caprine serum albumin. la) Using BCG. obtained for the values, derived from a standard (bl Using BCP for bovine (el and caprine (0) standard values. Solid Curve obtained with bovine albumin standards, and line Y=X Q)

Q)

M. S. Bain

84

s:

60

I

~

E

50

c

.,

.o

l!! 40

~c

o

~ 30 l;;

"C C

ffi

OJ

20

c

.~


10

10 20 30 40 50 60 Caprine standard concentration (g litre-I) FIG 2: Caprine serum albumin as calculated from bovine and caprine albumin standards. Comparison of the BCG and BCP dye-binding techniques

of human albumin and by Keay and Doxey (1983) and Gentry and Lumsden (1978) for the measurement of albumin in animal species. All these authors used total proteins, determined by the biuret method, for evaluating the electrophoretic patterns and this method is independent of species variation. BCG was introduced by Rodkey in 1965 and is now well established. The main disadvantage with this dye is that it is not entirely specific for albumin. Webster (1977) showed that the binding of BCG to albumin occurred in less than 30 seconds but that with time BCG would also bind to the serum globulin, increasing the colour density. Gustafsson (1976) showed that there was a marked difference between albumin values determined by BCG at one and 60 minutes, especially when the albumin concentration was low. The binding of BCG to globulin fractions is no doubt one of the factors contributing to the demonstrated variance, especially as several animals had high globulin concentrations. Despite the Abbott analyser being designed to give a five minute incubation for albumin determination, the benefits of using an automated technique outweigh the slight overestimate of the albumin concentration obtained. Attempts to improve the accuracy of albumin determinations in human medicine resulted in the use of IlCP as a dye-binding reagent. IlCP was shown to bind specifically to albumin, and Pinnel and Northam (1978) demonstrated that there was close agreement

between measurements with BCP and an electroimmunoassay procedure. However, Duggan and Duggan (1982) showed that the colour developed by bovine serum albumin with BCP was only 26 per cent that of an equivalent amount of human albumin. The poorer precision thus obtained makes BCP unsuitable for routine analysis. The bias observed between the two dye-binding methods is a reflection of the different absorbances of the dye protein complexes for bovine and caprine albumin. The effects of this difference are displayed in Fig 2; for BCG the regression coefficient was x O' 83 and for BCP x l- 22. This study has shown that although there are no significant differences between the individual albumin values for each technique employed, there are differences in the calculated concentrations dependent on the method. Cellulose acetate electrophoresis is accurate but time consuming. Its use should, however, be considered where low levels of albumin or high levels of globulin are present. If, when estimating caprine serum album, bovine albumin is used as a standard, a correction factor must be employed to give a 'true' value. Using the BCG reagent, a factor of x l : 20 is required. The equivalent factor for BCP is xO·81 but, as was shown, this dye is not suitable for routine use. To obtain consistently accurate results, a species standard would be the most satisfactory. However, if non-bovine samples are assayed only occasionally and they can be incorporated into an analytical run, then such a correction factor would be of value. References ALLERHAUD, J., GISLER, L., McCARRITT, J. & CORUJO, M. (1967) Clinical Chemistry 13, 701 DOUMAS, B. T., WATSON, W. A. & BIGGS, H. G. (1971) Clinica Chimica Acta 31,87-96 DUGGAN, J. & DUGGAN, P. F. (1982) Clinical Chemistry 28, 1407-1408 GENTRY, P. A. & LUMSDEN, J. H. (1978) Veterinary Clinical Pathology 7, 12-15 GORNALL, A. G., IlARDAWILL, C. J. & DAVID, M. M. (1949) Journal of Biological Chemistry 177,751 GUSTAFSSON, J. E. C. (1976) Clinical Chemistry 22, 616-622 HERNANDEZ, 0., MURRAY, L. & DOUMAS, B. (1967) Clinical Chemistry 13, 701 KEAY, G. & DOXEY, D. L. (1983) Research in Veterinary Science 35,58-60 PINNELL, A. E. & NORTHAM, B. E. (1983) Clinical Chemistry 24,80-86 RODKEY, F. L. (1965) Clinical Chemistry 11,478-487 WEBSTER, D., BIGNELL, A. H. C. & ATTWOOD, E. C. (1974) Clinica Chlmica Acta 53,101-108 WEBSTER, D. (1977) Clinical Chemistry 23, 663-665

Accepted August /6, /985