IGF-I binding to the IGF-I receptor is affected by contaminants in commercial BSA: The contaminants are proteins with IGF-I binding properties

ObjectiveTo determine whether different albumins have an effect on IGF-I binding assays.MethodsWe have studied the effect of five different albumins in plate antibody capture binding assay. For IGF–IR studies the IGF–IR specific antibody 24-31 was used and for IR/IGF–IR hybrid receptors the IR specific antibody 83-7 was used. Binding to IGF–IR was studied by displacement of 125I-IGF-I with IGF-I in the absence or presence of 0.1%, 0.5% or 1% (w/v) albumin. The IR/IGF–IR hybrid receptors were studied in the presence of 0.5% (w/v) of HSA A-1887 and BSA A-7888 and with IGF-I or insulin displacement of 125I-IGF-I. The albumins used were purchased from Sigma–Aldrich. Two batches of albumins from each catalog number were tested. The albumins were: HSA A-1887, BSA A-4503, BSA A-6003, BSA A-7030, and BSA A-7888. Contaminants in the albumins were characterized as proteins with IGF-I binding properties by cross-linking to 125I-IGF-I and SDS-page analysis.ResultsBSA A-4503, A-7030 and A-7888 from Sigma–Aldrich contain proteins with IGF-I binding properties. These contaminants increased the determined EC50 for displacement of 125I-IGF-I from IGF–IR up to 40-fold in a BSA dependent manner. The presence of BSA-7888 in binding experiments increased the determined EC50 for IR/IGF–IR hybrid receptors 8–16-fold.ConclusionsWhen IGF-I is characterized with respect to the effect on living cells and on binding to potential receptors unspecific binding to surfaces is often prevented by the addition of albumin in the assay. Here we report that when binding to the classical IGF–IR and IR/IGF–IR hybrid receptors are studied the measured EC50 values can be albumin dependent if it is contaminated with proteins with IGF-I binding properties. The free IGF-I concentration will be lower than estimated. Thus, the contaminated BSA preparations result in artifacts leading to misinterpretations and underestimation of the effect of IGF-I. Our results provide one possible explanation as to why different laboratories report different EC50 values for IGF-I.