A selective protein sensor for heparin detection

Abstract

No clinical assays for the direct detection of heparin in blood exist. To create a heparin sensor, the hyaluronan (HA)-binding domain (HABD) of a protein that binds heparin and HA was engineered. GST fusion proteins containing one to three HABD modules were cloned, expressed, and purified. The affinities of each construct for heparin and for HA were determined by a competitive enzyme-linked immunosorbent assay using immobilized HA or heparin. Each of the constructs showed modest affinity for immobilized HA. However, heparin was 100-fold more potent than HA as a competing ligand. With immobilized heparin, affinity increased as the HABD copy number increased. The three-copy construct, GST-HB3, detected unfractionated free heparin (UFH) as low as 39 ng/ml (equivalent to approximately 0.1 U/ml) with a signal-to-noise ratio of 5.6. GST-HB3 also showed 100-fold selectivity for heparin in preference to other glycosaminoglycans. The plot of log K d vs log [Na +] showed 2.5 ionic interactions per heparin–HB3 interaction. GST-HB3 showed a linear detection of both UFH (15 kDa) and low-molecular-weight heparin (LMWH; 6 kDa) added to human plasma. For UFH, the range examined was 78 to over 2000 ng/ml (equivalent to 0.2 to 5.0 U/ml). For LMWH, the useful range was 312 to over 2000 ng/ml. The coefficient of variance for the assay was <9% for six serial heparin dilutions and <12% for three plasma samples. In clinical use, GST-HB3 could accurately measure therapeutic heparin levels in plasma (0.2 to 2 U/ml).