N6022 We found that activator drugs

We found that activator drugs decrease NOsGC protein levels in Sf9 cells. This decrease in protein levels is more pronounced for BAY 60–2770 compared to cinaciguat, and more obvious for α1/β1 compared to α2/β1 (see Fig. 2). This led us to hypothesize that the reduction in protein level correlates with the activity induced by the respective drug. Our idea was confirmed by expression of the inactive form α1D530A/β1 with BAY 60–2770, where no influence on protein level was detectable. Initially we believed that the reduction in GTP and a massive increase in cGMP lead to an unspecific reduction in any expressed protein. But an influence of activator drugs towards protein levels of the inactive, non-nucleotide binding mutant was also missing when it was co-expressed with the catalytically active NOsGC fusion construct β1α1 in the presence of BAY 60–2770 (see Fig. 3 B). At the same time, the catalytically active β1α1 fusion construct in this cytosolic fraction decreased in the presence of activator. While we show that activator drugs decrease NOsGC subunit levels depending on intact catalytic activity and nucleotide-binding ability of the expressed enzyme, Stasch et al. found that cinaciguat stabilizes the enzyme and increases β1 subunit level in porcine endothelial and smooth muscle N6022 [38]. We assume that both effects occur in parallel: direct stabilization of the enzyme by activator and destabilization by increased cGMP or decreased GTP level. Depending on the activator concentration and the concentration and activity of enzyme, stabilization or destabilization prevails. This balance probably depends on the experimental conditions, particularly cell type and incubation time. These differ between the earlier studies by Stasch et al. [38] and the current study (endogenous porcine NOsGC vs. overexpressed human NOsGC and 24 h vs. 72 h of incubation). The biphasic course of α2/β1 protein levels with cinaciguat (see Fig. 2C) supports the idea of two parallel effects. After a decrease in protein levels, they return to normal at concentrations 1 and 10 µM of cinaciguat. This (re-) stabilization effect is more obvious for the β1 subunit and has also been described for the β1 subunit in vivo [38]. Preparation of a heme-free activator-containing enzyme is of particular interest for crystallization studies. Despite highly motivated attempts in academia and industry, the crystal structure of full-length NOsGC remains unsolved to date [39], [40]. Crystallography requires a homogenous population of stable protein for effective crystal growth [41], [42]. As a hemoprotein, NOsGC is compromised by oxidation and heme loss, thereby producing heterogeneity. Binding of activator does not depend on redox state, leading to an advantage of heme reconstitution by activator. The conventional procedures of heme replacement by detergents or unspecific oxidants might produce an altered, possibly less stable protein. Furthermore, the removal of heme might not be complete, thereby again producing heterogeneity. The method we present in this paper leads to a completely heme-free protein, because heme is not removed but the insertion of heme is prevented, without any alteration of the native protein. Successful crystallization often requires high amounts of protein [43]. The problem of decreasing protein levels, resulting in low yields, can be overcome by working with the inactive NOsGC such as the α1D530A/β1 mutant used in the present paper. Previous studies have focused on NOsGC as a redox dependent drug target where ferric heme is replaced by sGC activators [38]. This seems to be particularly true for the mature enzyme, where a high percentage of reduced NOsGC with ferrous heme leads to weak activation, while a high percentage of oxidized NOsGC with ferric heme in tissue exposed to oxidative stress leads to strong activation. To remove ferrous heme by cinaciguat from the mature enzyme a concentration much higher than the therapeutic dose is needed [44]. Our data indicate that the activator drugs compete with heme independent of its redox state during protein biosynthesis and maturation. Expression with activator leads to a highly active protein population, indicating a high percentage of enzyme with incorporated activator. Furthermore, the complete heme displacement after expression with activator supports the idea that activator drugs are incorporated into enzyme during de novo synthesis.