Prolactin Heterogeneity: A Limitation on the Evaluation of Results from Prolactin Assays due to Differences in Immunoassays and the Different Bioactivities of Prolactin Forms

Abstract
Prolactin exists in biological fluids in several molecular forms. This raises two questions: (1) whether the assay of prolactin by immunotechniques is valid and reliable and (2) whether the different forms have different physiological roles, which might be exploited to improve diagnostic accuracy and data interpretation by the use of appropriate methods. To investigate these questions, prolactin from human amniotic fluid was separated, by concanavalin A-Sepharose affinity chromatography, into bound, retarded and unbound fractions (bound prolactin fraction, retarded prolactin fraction, unbound prolactin fraction), which were characterized by electrophoresis, immunoblotting and glycan detection blot. Virtually no contamination was found in the bound prolactin fraction, and the unbound prolactin fraction and retarded prolactin fraction were 74-83% pure according to densitometry of the electrophoretic and immunoblot patterns. High variability was found among the individual patterns. Glycan detection in the blotted fractions revealed that the bound prolactin fraction bands corresponding to M(r)25,000-29,000 were weakly glycolysated, whereas the bands of M(r)60,000-64,000 were significantly glycan-positive. Immunoreactive bands of unbound prolactin fraction and retarded prolactin fraction also stained positively for glycans. Using two commercial prolactin kits, the bound prolactin fraction forms were virtually undetectable. To demonstrate that the prolactin forms may depend on the hypothalamic state, two behaviourly different breeds of cattle were used as an animal model for studying hypothalamic activities. The number of immunoreactive bands, representing the prolactin forms, and the change of the forms in response to thyroliberin differed strikingly among the groups. The bioactivity of the forms was examined in bovine granulosa, oviductal, endometrial and spleen cells, and in murine splenocytes, the latter being activated by concanavalin A or allogeneically to create in vitro conditions that may have relevance for situations in vivo. The rate of incorporation of [3H]thymidine in murine splenocytes was dose-dependently enhanced only by bound prolactin fraction. The increase was abolished by purified anti-prolactin antiserum. However, the standard prolactin from the kits inhibited the proliferation even in low dose (1.25 microgram/l) and the inhibition was abolished in part by bound prolactin fraction. Thymidine incorporation into the bovine cells was significantly increased by low concentrations (2 micrograms/l) of unbound prolactin fraction and retarded prolactin fraction. Oviduct epithelial cells and splenocytes were stimulated by unbound prolactin fraction but not by retarded prolactin fraction in a dose of 16 micrograms/l. Thymidine incorporation into granulosa cells was inhibited by retarded prolactin fraction (16 micrograms/l) but not by unbound prolactin fraction.(ABSTRACT TRUNCATED AT 400 WORDS)