Protein disulfides with peroxiredoxin IV and PDI peroxidases

Two distinct types of flavin-linked oxidases catalyze the generation of protein disulfide bonds in the secretory apparatus of mammals. The Ero1α and – β isoforms oxidize protein disulfide isomerases (PDI) leading to the net generation of disulfide bonds as shown in Figure A. Here, PDI is the immediate oxidant for unfolded reduced proteins.

prxdn4_ero_qsox

In contrast, the Quiescin-sulfhydryl oxidases (QSOX; Figure B) insert disulfide bonds directly into unfolded reduced proteins – PDI is only used to isomerize mis-pairings. Ero1 has been widely regarded as the preeminent catalyst driving oxidative protein folding in metazoans. However, this “Ero-centric” view has been questioned – by the very modest phenotype accompanying ablation of the two Ero1 isoforms in the mouse (Zito et al. [PubMed]). This influential paper has encouraged a more nuanced view of the enzymology of oxidative protein folding and a search for new pathways for disulfide bond generation.

In this note we highlight a newly recognized pathway involving peroxiredoxin IV (PRDX4). Peroxiredoxins catalyze the thiol-dependent reduction of hydrogen peroxide or alkylhydroperoxides:

and participate in hydrogen peroxide signaling and peroxide removal.

Tavender et al. [PubMed] had previously identified PRDX4 as an antioxidant enzyme – suggesting that it removes hydrogen peroxide formed by the sulfhydryl oxidases in the ER lumen. They have now found that reduced PDIs family members are oxidized by PRDX4 [PubMed] :

prxd4_pdi_correct

and so the hydrogen peroxide generated by sulfhydryl oxidases can be leveraged to generate a second protein disulfide bond (see above).

Here, two disulfide bonds would be formed for every oxygen molecule used – now without the potentially troublesome accumulation of hydrogen peroxide.

Employing a completely different approach, Ron and coworkers (Zito et al. [PubMed] and see a related commentary [PubMed]) went fishing with a mutant of PDI primed to capture bona-fide targets in mixed disulfide linkage. One of their catches was the same PRDX4. Zito et al. provide strong in vivo evidence for the importance of PRDX4 in Ero1-compromised cells.

What is the source of hydrogen peroxide driving disulfide generation with PRDX4 in that Ero1-deficient mouse? In addition to QSOX, both papers consider membrane-bound NADPH oxidases and the mitochondrion as potential peroxide donors for the ER.

These studies raise a number of fascinating questions. What is the topology of hydrogen peroxide generation by the NADPH oxidase (NOX4) – e.g. is peroxide generated on the luminal face of the ER? Exactly how membrane-permeable is hydrogen peroxide? What is the net peroxide output of mitochondria, and how effectively is this hydrogen peroxide delivered to the ER?

Finally, PRDX4 is a secreted protein [PubMed] like QSOX1. What are the extracellular roles of these proteins?

An addendum (March 29, 2011)

Ruddock and colleagues have added two further enzymes that can utilize hydrogen peroxide for disulfide generation in the ER. They had previously shown that hydrogen peroxide could directly serve to drive oxidative protein folding in vitro and demonstrated that PDI could accelerate this reaction [PubMed]. They have now identified two glutathione peroxidases (GPx7 and GPx8) as de facto PDI peroxidases [PubMed]. Using bimolecular fluorescence complementation, Ruddock and coworkers showed that both PDI-peroxidases transiently interact with Ero1-alpha in the ER of HeLa cells. They further provide in vitro evidence for stimulation of Ero1-alpha oxidase activity by these PDI-peroxidases [PubMed].

In aggregate, the studies of the Bulleid, Ron and Ruddock laboratories are changing the way we think about hydrogen peroxide in the ER. Hydrogen peroxide was formerly regarded as a problem for the ER and, consequently, for cells with a heavy output of disulfide-containing proteins. Now peroxide appears to be also part of the solution.