Evaluation of free psa isoforms, psa complex formation, and specificity of anti‐psa antibodies by hplc and page‐immunoblotting techniques

Abstract

Both high performance liquid chromatographic (HPLC) and polyacrylamide gel electrophoresis-immunoblotting (PAGE-immunoblotting) procedures have been established for the study of isoforms of free prostate-specific antigen (PSA) and the complex formation between free PSA and protease inhibitors, and for the evaluation of the specificities of various anti-PSA antibodies. We found multiple isoforms of free PSA on PAGE, which were all capable of forming complexes with protease inhibitors. The same isoform pattern can be produced from the original seminal fluid. The PSA isoforms differ from each other most likely in charge because they could be converted to one band on SDS-PAGE and to a single peak by gel filtration chromatography. We found it difficult to form large quantities of PSA complex when mixing free PSA from seminal fluid with protease inhibitors, regardless of whether the free PSA or the protease inhibitors were in excess. Except for the PSA-ACT complex, which was consistently detectable by both HPLC and PAGE-immunoblotting techniques after incubation, these two procedures disagreed in their detection of PSA-A2M and PSA-AT complexes. The PSA-A2M complex was usually observable by immunoblotting techniques but barely detectable on HPLC, whereas PSA-AT was totally invisible by immunoblotting but appeared as a peak in the HPLC elution profile. Mixing free PSA with serum clearly resulted in both PSA-ACT and PSA-A2M complexes. However, more PSA-ACT than PSA-A2M was formed; the result was also confirmed by using 125I-PSA for mixing. PSA could be separated into active and inactive PSA by DEAE Sepharose chromatography with a 14-fold difference in protease activity. The difference in enzymatic activity apparently had no effect on complex formation. All the anti-PSA antibodies examined in this study reacted with PSA isoforms, PSA-ACT and PSA-A2M complexes. We conclude that it would be almost impossible to establish an assay to measure all forms of PSA in the serum and to expect to produce precise and accurate PSA values.