A comparison of methods for the immunoassay of serum apolipoprotein B in man.

Three assays for serum apolipoprotein B, radioimmunoassay, automated immunoprecipitation and rocket immunoelectrophoresis were compared. The antiserum used was raised against lipoprotein of density 1.040-1.053 g/ml (lipoprotein B). Each of the methods had a high degree of specificity when tested against potential interfering substances. The lowest levels of apolipoprotein B could be measured with radioimmunoassay but this sensitivity entailed the use of high dilutions of serum and resulted in a lower precision. Concentration response curves of lipoprotein B solution and serum were parallel for radioimmunoassay and automated immunoprecipitation but were not for rocket immunoelectrophoresis. Serum apolipoprotein B could be assayed by immunoelectrophoresis, however, if serum calibrated against the protein concentration of lipoprotein B solution by either of the other two methods was used as a secondary standard. Such a secondary serum standard also proved advantageous for all the methods because of the relative stability of the apolipoprotein B content of serum as compared to aqueous solutions of lipoprotein B. The mean apolipoprotein B concentration in 29 normolipaemic subjects aged 20-30 years was 0.84 +/- 0.12 g/1 (mean +/- S.D.) by radioimmunoassay, 0.85 +/- 0.11 g/1 by automated immunoprecipitation and 0.88 +/- 0.11 g/1 by rocket immunoelectrophoresis. The correlation between apolipoprotein B levels measured by the three methods was good. The ratio of serum cholesterol to serum apolipoprotein B was unaffected by individual differences in serum cholesterol or triglyceride levels. There was no significant difference between fasting and post-prandial serum apolipoprotein B concentrations. Radioimmunoassay is particularly suited to the measurement of low levels of apolipoprotein B, automated immunoprecipitation for large numbers of samples and rocket immunoelectrophoresis, when care is taken in the interpretation of results, for small numbers of samples.