Protein C inhibitor (PCI): Possible role as a retinoid-binding and delivering protein

Serpins (serine protease inhibitors) are a family of closely related glycoproteins that inactivate serine proteases by forming enzymatically inactive, stable 1:1 complexes. They are involved in the regulation of the activity of the coagulation system, the fibrinolytic system, and other intra- and extravascular protease systems. Besides these inhibitors of serine proteases the serpin family also includes members that have lost protease inhibitory activity and have acquired other biological functions, such as the hormone precursor angiotensinogen or the hormone binding proteins corticosteroid-binding globulin (CBG) and thyroxin-binding globulin (TBG). So far it is completely unknown, whether or not inhibitory members of the serpin family could also have additional non-inhibitory function. In the previous project (P12308-Gen) we have obtained data that suggest for the first time that an inhibitory serpin (protein C inhibitor, PCI) specifically binds a hormone (retinoic acid) and can enhance the biological activity of that hormone in a tissue culture model. We and others have furthermore shown that PCI is highly expressed in many tissues requiring retinoids for differentiation and maturation. It therefore seems to be important to evaluate a possible role of PCI within the retinoic acid signaling pathway. In this project we are planning to analyze in detail the interaction of PCI with retinoids as well as to investigate the biological impact of such an interaction. We will determine affinities of different retinoids for the different forms (native, complexes, cleaved) of PCI and will analyze the effect of retinoids on the inhibitory activity and enzyme specificity of PCI. We are also planning to analyze the interaction of retinoids with PCI present in biological fluids as compared to their reaction with other retinoid-binding proteins. Furthermore we will determine the binding site(s) for retinoids within the PCI molecule and study the effect of PCI on the stability of retinoids as well as on their uptake by target cells. Results obtained should therefore clarify to which extent PCI might participate in the delivery of and/or the signaling by retinoids. They might furthermore represent the basis for new directions of research not only in the field of serpins but also in the field of signal transduction by retinoids.

We analyzed the interaction of a protease inhibitor (protein C inhibitor) with retinoids (derivatives of vitamin A). Results obtained strongly suggest that the analyzed inhibitor could have an additional function (besides inhibiting the activity of proteases), namely in the transport and delivery of retinoids to cells. The serpin (serine protease inhibitor) family includes inhibitors of serine proteases and non-inhibitory members with other biological functions. Non-inhibitory members of the serpin family are the hormone precursor angiotensinogen or the hormone binding proteins CBG (corticosteroid-binding globulin) and TBG (thyroxin- binding globulin). So far it is unknown to which extent inhibitory serpins might have additional non-inhibitory functions. We therefore analyzed binding of hydrophobic hormones to inhibitory serpins. These studies, in which we used steroid hormones and retinoic acid, revealed binding of retinoic acid to protein C inhibitor (PCI), while none of the other serpins used (antithrombin III, heparin cofactor II, plasminogen activator inhibitor-1) bound to any of the hormones analyzed (aldosterone, cortisol, estradiol, progesterone, testosterone, retinoic acid). Also PCI did not bind to any of the steroid hormones tested. We therefore analyzed the interaction of PCI with retinoids in more detail. We were able to calculate a K d of 2.43M for the interaction of PCI with all- trans-retinoic acid. The number of retinoid-binding sites within the PCI molecule was close to one. Furthermore, in purified systems as well as in biological fluids retinoic acid was mainly found in association with a high M r form of PCI, possibly a polymer. In order to determine the binding site(s) for retinoids within the PCI molecule, we mutated the PCI molecule in distinct parts. In addition to a mutant that lacks the complete H-helix, which is supposed to be the main heparin- binding site, we now made mutants, in which single amino acids have been replaced. The amino acids selected have either been described to be important for heparin binding, or are located within a sequence (beta-barrel) that is highly homologous to the hormone-binding sites within the TBG-molecule. These mutants are currently being used to analyze their interaction with retinoids. We also analyzed to which extent PCI can bind other lipids. So far we have data indicating that PCI binds arachidonic acid with an affinity that is ~10fold higher as compared to the affinity of PCI for retinoic acid. The number of binding sites per molecule of PCI seems to be ~0.1, suggesting that either part of the binding sites are already occupied by endogenous ligands, or that PCI has to form a polymer prior to binding one molecule of arachidonic acid. These questions are currently being analyzed.