Figure 5

Full Model, signal/response curves for varying parameter values. Steady state concentrations of free FIH (blue), and of unhydroxylated FIH-sequestering ARs (grey, left hand panels), as well as of total HIFα (black), non-CAD-hydroxylated HIFα (green) and CAD-hydroxylated HIFα (red, right hand panels) are shown as functions of oxygen. Corresponding pairs of curves are marked at their intersection. A. The amount of ARs (A _tot ) determines the oxygen threshold for FIH-release and HIFα CAD-hydroxylation. The larger the number of ARs, the higher the oxygen-tension at which FIH is released (left hand panel). Competing ARs also create an oxygen-threshold for CAD-hydroxylation, above which non-CAD-hydroxylated HIFα drops much more sharply with increasing oxygen levels than in the case where there is no competition (right hand panel, green solid curves). In addition, the FIH/AR-interaction restricts the presence of CAD-hydroxylated HIFα to moderate hypoxia and decreases its peak values (right hand panel, red dotted curves). In contrast to B and C, here FIH is assumed to not bind to hydroxylated ARs at all. B. Binding of FIH to hydroxylated ARD proteins has to be weak for efficient FIH-release. γ, the FIH affinity for hydroxylated relative to unhydroxylated ARs, must be small for both efficient FIH-release (left hand panel) and efficient HIFα CAD-hydroxylation (right hand panel). C. ARD proteins must be stable compared to the basal stability of HIFα for efficient HIFα CAD-hydroxylation to occur. Large ε indicates stable ARD proteins, which allows FIH-release to occur at lower oxygen levels (left hand panel). Large ε increases extent and sharpness of HIFα CAD-hydroxylation and shifts the peak value to more severe hypoxia (right hand panel).