Two antioxidants are better than one

New evidence suggests that the cellular oxygen-sensing hypoxia-inducible factor (HIF) pathway may be protected by a double buffer of cellular antioxidant defense. Key players in the oxygen-dependent regulation of this pathway are the prolyl hydroxylase domain–containing enzymes (PHDs) that catalyze the prolyl-4-hydroxylation of HIFα, dependent on the presence of oxygen, 2-oxoglutarate, and iron in the ferrous (Fe²⁺) form. Vitamin C is also required as a cofactor, possibly to maintain the catalytic iron center in its functional Fe²⁺ state, although both the mechanism and the in vivo requirement are not absolutely clear.

In this issue of Blood, Nytko and colleagues¹  address this by reporting that vitamin C–deprived Gulo^−/− knockout mice show normal HIF-dependent gene expression. Furthermore, they found that glutathione, which prevents oxidative damage to important cellular components, might substitute for the vitamin C requirement of the PHDs.

The PHD enzymes have emerged as the main oxygen-dependent regulators of HIF, the master transcriptional activator of the body's response to hypoxia. In humans, 3 PHD isoforms(1-3) and factor inhibiting HIF (FIH) regulate HIF protein levels and transactivation function, respectively.²  They are members of an extended family of dioxygenases that are dependent on nonheme ferrous iron and 2-oxoglutarate and are present in most, if not all, forms of life.³  In humans they are involved in many biologic processes ranging from DNA repair to collagen biosynthesis.⁴  A feature of many (but not all) of these dioxygenases is that activity can be stimulated by ascorbate.

In their elegant study Nytko and colleagues note that in Gulo^−/− mice that lack the gene encoding L-gulono-1, 4-lactone-oxidase, a key enzyme for vitamin C synthesis, both the steady-state and hypoxic induction of the HIF pathway appeared normal. In addition, Gulo^−/− mice showed similar circulating erythropoietin (Epo) levels in groups of animals whose diet was either supplemented or not supplemented with vitamin C. The latter group of animals had plasma levels of vitamin C below the detection limit of the assay. This suggests that the ability of the PHDs to hydroxylate HIF-2α, the prevalent regulator of circulating EPO levels, may not be affected by vitamin C status. Overall, these data imply that PHDs are still active and are not affected by the absence of vitamin C in this model. A similar finding was previously described for the collagen prolyl hydroxylases in these Gulo^−/− mice.⁵ 

One possibility is that PHD activity is maintained by an alternative cellular reducing agent that can substitute for vitamin C. Reduced glutathione (L-γ-glutamyl-L-cysteinyl-glycine; GSH) acts as the major cellular antioxidant defense protecting against damage from free radicals and reactive oxygen species. GSH was tested in vitro, and consistent with previously reported findings⁶  it was found to stimulate hydroxylation in reactions involving all 3 PHD isoforms. GSH, normally present in cells in the millimolar range, was also able to reduce HIFα protein levels and endogenous target gene expression in cells exposed to cobalt chloride.

In vitro, treatment with cobalt chloride resulted in direct oxidation of PHD2 protein that was attenuated by GSH. To further explore the mechanism of the reductive action of GSH, the authors investigated a mutant form of PHD2 with a Cys201Ser mutation. This variant showed increased basal hydroxylation rates and was less susceptible to oxidative damage, suggesting that this cysteine residue, which in PHD2 is surface accessible, could be a target for some of the protective effects of GSH.

These data provide an interesting insight into the potential regulation of PHDs by 2 reducing agents, vitamin C and GSH (see figure). This cofactor requirement has traditionally been linked to the maintenance of iron in the Fe²⁺ form but (for GSH at least) can also be extended to a role in protection against debilitating protein oxidations in the PHD2 protein. It suggests that the PHD enzymes may be protected by multiple (and partially redundant) cellular antioxidant defenses. Because oxidative stress is important in many physiologic responses it may be important to shield the HIF system for maximal operation.