Oxidative Protein Folding: Nature's Knotty Challenge

Introduction

Every day, each nucleated cell in our bodies must correctly fold thousands of protein molecules. If a protein has n cysteine residues and p disulfide bonds, the number of possible fully folded disulfide-bonded isomers (i) is given by Equation (1): \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}i = \frac { n! } { ( n - 2p)!p!2^p } \tag { 1 } \end{align*} \end{document}

Thus, for example, a protein with 18 cysteines and 9 disulfide bonds can theoretically form 34,459,425 different disulfide-bonded isomers, only one of which imparts correct protein function (Fig. 1). Fortunately, oxidative protein folding (OPF) is an iterative process with consistent results [even in the absence of protein chaperones (1, 4)]—because proper OPF is as important to health and survival as proper protein expression. A protein molecule with an uncorrected or incomplete set of correct disulfide bonds is a detriment to cellular health, more so than if that singular molecule had never been expressed. Unfolded proteins can contribute to inflammatory phenomena that underlie many of the chronic diseases which plague society. Hence, the need to understand OPF goes hand in hand with efforts to develop new therapies and, in some cases, improved markers of disease. To accomplish these tasks, we should elucidate the fundamental biochemical pathways that govern proper folding. Moreover, tools are needed to analyze the products of these pathways. Six of the Forum Review Articles in this Forum examine the latest discoveries in the biochemistry of OPF and describe their relationships to human disease as well as the footholds they provide for improving the expression of recombinant proteins for pharmaceutical and industrial applications. The remaining two articles review some of the latest mass spectrometry-based bioanalytical approaches for the direct study of disulfide-bond and higher-order protein structures.

OPEN IN VIEWER

FIG. 1.

The core problem of oxidative protein folding. Both in vitro and in vivo oxidative folding systems generate disulfides and then rearrange them as necessary until the correct structure is arrived at. This generally takes place through three competing pathways. The red pathway represents reduction of disulfides that are resistant to attack or reform very rapidly. These are likely to be stable disulfides. Shown in green is the pathway of reducing equivalent-catalyzed intramolecular rearrangement. This pathway tends to replace more reactive disulfides with less reactive ones. The blue pathway represents partial to full reduction followed by complete rearrangement of the disulfides. Whether redox buffers such as DTT or folding chaperones such as PDI are employed (red discs), the basic strategy is to gently but continually assault the existing disulfide structures with reducing equivalents until a structure forms that is resistant to further reduction by this onslaught (3, 4). As the number of cystines increases, the number of possible disulfide structures skyrockets according to Equation (1) (plotted within inset). Cystines with redox potentials that make them only marginally resistant to reduction can serve as allosteric disulfide bonds (2). Figure adapted with permission from Ref. (3). PDI, protein disulfide isomerase. To see this illustration in color, the reader is referred to the web version of this article at www.liebertpub.com/ars

Fundamentals and Critical Molecular Players (Lu and Holmgren)

Lu and Holmgren discuss OPF and the thioredoxin superfamily. They begin with a listing of the large number of proteins that contain the thioredoxin structure followed by a detailed discussion of the physical chemistry of thiol-disulfide redox reactions. The protein disulfide isomerase (PDI) family, proteins with multiple thioredoxin domains, is examined in the context of oxidizing equivalent transfer pathways. Other redox molecules and proteins that interact with PDI are reviewed, such as glutathione and Ero1 in endoplasmic reticulum (ER) and the cytosol. In addition to the PDI family and homologues, quiescin sulfhydryl oxidase and thiol dependent peroxidases that produce hydrogen peroxide as a result of their redox activity are highlighted due to their thioredoxin domains. Inactivation of PDI and ER stresses are briefly discussed in the context of S-nitrosylation and S-glutathionylation of PDI. OPF in bacteria is examined and is of considerable interest, as many recombinant proteins are produced in Escherichia coli. In particular, the Dsb family of proteins is discussed in the context of thioredoxin domain structure, individual active cysteines, and redox flow of electrons during protein folding. Finally, thioredoxin and glutaredoxin systems are compared along with their client oxidoreductase substrates and functions/locations in bacterial and mammalian cells. This review underscores the importance of thioredoxin and the many proteins with thioredoxin domains that are key to correct protein folding in client proteins with two or more cysteines.

Compartmentalized Design and Function (Herrmann and Riemer)

The review by Herrmann and Riemer details current, yet rapidly developing knowledge on the process of OPF in the mitochondrial intermembrane space (IMS) and compares it with the better understood systems in the ER and bacterial periplasm. For the purposes of comparison, these latter systems are briefly summarized. The detailed discussion of IMS protein folding begins with a comprehensive list of the currently known IMS proteome and then provides an overview of the two major mechanisms for protein translocation into the IMS: bipartite presequences and Mia40-mediated import. Mia40, in particular, takes center stage in IMS protein folding with roles as protein import mediator, oxidoreductase, and chaperone. All said, OPF in the three compartments follows the general paradigm in which the specificity of an oxidoreductase is combined with the raw oxidation power of a sulfhydryl oxidase. However, key differences exist such as the coupling of disulfide formation in the IMS with both protein translocation and cellular respiration. The former is entirely unique to the IMS, and the latter results in the reduction of O₂ to H₂O rather than H₂O₂ as in the ER, which has adapted numerous mechanisms for dealing with and even harnessing this dangerous byproduct. Several major aspects of OPF in the IMS remain only partially understood, including how mitochondria refold disulfide-scrambled proteins, how the IMS senses an accumulation of unfolded proteins, or how stressed mitochondria communicate with other organelles. Constructive comparisons of the IMS, ER, and bacterial periplasm such as carried out by Herrmann and Riemer in their review should help fill in these missing gaps in our knowledge of protein folding in the mitochondrial IMS.

When It All Goes Wrong (Cao and Kaufman)

Cao and Kaufman astutely point out that protein folding is the most error prone step in gene expression. They go on to provide a thorough summary of the cellular response to the accumulation of unfolded and misfolded proteins in the ER—a process known as the unfolded protein response. Though OPF in the ER usually produces H₂O₂ as a byproduct, the ER seems to have relatively limited enzymatic protection against reactive oxygen species (ROS), potentially making it susceptible to oxidative stress during times when protein folding loads are increased. Cao and Kaufman emphasize the fact that ER stress and oxidative stress can work together in a feed-forward mutual amplification cycle which can, ultimately, activate pro-apoptotic pathways, resulting in cell death. As described in detail, this phenomenon plays important roles in numerous chronic human diseases, including cancer, metabolic disorders, immune disease, and neurodegenerative disease. (The latter is not covered in detail, as it was recently reviewed elsewhere.) For the latter three disease classes, it would appear that a therapeutic strategy targeting both ER and oxidative stress may be the most effective. Ultimately, elucidating the mechanisms by which these two stresses cross-talk with one another in different cell types and disease models represents a significant goal in the treatment of chronic disease.

Uncleared Wreckage (Singh Group)

Protein misfolding to the point of aggregation and precipitation occurs in several neurological diseases. The mechanism(s) behind these process(es) remain incompletely understood. In Creutzfeldt-Jakob disease (sCJD) and in Parkinson's disease (PD), the proteins that aggregate, PrP^C (which on misfolding becomes the pathogenic and infectious PrP^Sc) and α-synuclein, play substantial roles in maintaining cellular iron homeostasis. Moreover, redox-active ROS-producing iron co-accumulates with these protein aggregates—but whether this is a cause or effect of protein precipitation remains controversial. Singh et al. point to the recent finding that the cerebrospinal fluid of sCJD and PD patients reveals the loss of brain iron homeostasis (well before end-stage disease) as a common feature associated with these diseases and then describe the normal function of these proteins in cellular iron metabolism and the processes that lead to iron imbalance in sCJD and PD. With the picture that emerges from recent findings, Singh et al. present a striking case that the roles of iron homeostasis and iron redox chemistry deserve increased attention for their potential roles in the etiology of sCJD and PD.

Renegade Machinery (Lake and Faigel)

Lake and Faigel review the role of Quiescin sulfhydryl oxidase 1 (QSOX1), a flavin adenine dinucleotide (FAD)-binding enzyme, in cancer. In particular, QSOX1 over-expression appears to be a driver of tumor cell invasion and important in the extracellular matrix of tumor cells and stromal fibroblasts. The authors provide a history of the discovery and early characterization of QSOX1, detailing how it is different from other sulfhydryl oxidases involved in protein folding and that it is the only known sulfhydryl oxidase with both disulfide-generating and disulfide-transferring properties. The authors then provide insights for which CxxC domains are involved and important for the enzymatic activity of QSOX1, based on the crystal structure of the enzyme. It is considered that PDI and QSOX1 collaborate in redox-mediated folding of proteins, but interestingly, neither is a substrate of the other. Faigel and Lake provide clear evidence that QSOX1 is over-expressed in a wide range of solid tumors and lymphoid malignancies, while not expressed in adjacent nonmalignant tissues tested. They then discuss the tumor biology of QSOX1, which likely involves matrix metalloproteinases and invasion through basement membranes. Finally, Lake and Faigel discuss potential consequences of QSOX1 over-expression. These include tumor-promoting effects of hydrogen peroxide (ROS) toward epithelial-to-mesenchymal transition, upregulation of oncogenes, and increased angiogenesis. QSOX1 may prove to be a viable anti-neoplastic target.

Taking Advantage: An Engineering Perspective (Mattanovich Group)

Recombinant proteins are absolutely indispensable to both basic research and innumerable industrial applications. However, as anyone who has struggled at the bench to produce an uncooperative protein knows, one recipe for production of a particular protein does not fit all. Often, protein expression difficulties are rooted in protein folding bottlenecks. Delic et al. provide an in-depth review of biological engineering approaches that are being adopted to overcome folding-related problems in both bacterial and eukaryotic expression systems. In some cases, protein expression logjams can be cleared by co-overexpression of chaperones or other folding-related proteins. In other cases, pathways that divert the desired protein to be misfolded, mistargeted, or degraded should be eliminated. Approaches that optimize the expression of intracellular and secreted proteins in bacteria and eukaryotes are discussed, with an emphasis on facilitating translocation to and proper folding in the bacterial periplasm and the eukaryotic ER. Major post-ER bottlenecks, including vesicular transport to the Golgi, proper Ca²⁺ homeostasis, and mis-sorting and proteolytic degradation and engineering approaches to solve them, are also described. This thorough review concludes with a discussion of techniques for novel target identification and an emphasis on the need to manipulate multiple pathways simultaneously for synergistic and optimally efficient recombinant protein expression.

Unraveling the Knots (Borges and Sherma)

Cysteines serve as the critical, focal-point residues of OPF. As such, their redox status—be static or dynamic—is critical to establishing and maintaining proper protein activity. The review by Borges and Sherma focuses on mass spectrometry-based techniques for tracking cysteine sulfhydryls in proteins and for mapping disulfide structures of folded proteins. As illustrated with various examples, the unique chemistry of sulfhydryls provides an invaluable molecular handle by which to pinpoint and track individual protein thiols vis-à-vis a large number of different probes that can serve an incredibly flexible number of research purposes. Cysteines involved in disulfide bonds are the most evolutionarily conserved amino acids—a fact that testifies to the importance of OPF in protein activity. However, mere knowledge of amino-acid sequence does not provide information about the disulfide connectivities that, by providing proper protein shape, play a foundational role in defining protein function. As described in the second half of this review, the low sample requirements of mass spectrometry have led to the development of numerous unique approaches for determining disulfide bond connectivities—many of which serve complementary purposes under the variable and unique circumstances that can be encountered with different proteins. Combined with innovative software algorithms, modern mass spectrometry-based approaches are verging on making the complex task of disulfide structure elucidation both routine and, in the not-too-distant future, automated.

Mapping the Dance (Wilson Group)

In general, this Forum focuses on protein folding from a biological or biochemical perspective. However, at a deeper level, protein folding should be understood as a problem to be addressed at the intersection of physical and analytical chemistry. To understand transient higher-energy species and intermediate folding structures, analytical approaches are required that enable us to capture detailed information from them during their fleeting existence. Hydroxyl radical-based oxidative labeling followed by mass spectrometric analysis has emerged as a powerful approach to map solvent exposed protein sites. Liuni et al. begin their review by providing an overview of alternative structure-dependent labeling techniques, including hydrogen-deuterium exchange and chemical cross-linking. They then move on to an in-depth discussion of the features of the oxidative-labeling method, theoretical fundamentals and practical principles behind the most common technical approaches to generating hydroxyl radicals, and caveats and constraints that should be heeded throughout the course of these sensitive experiments. The second half of the review focuses on applications. Examples overviewed that showcase the power and flexibility of this now well-established approach include the assessment of activity-linked conformational changes in proteins, determination of protein complex interfaces, quantitative analysis of secondary structure stability, ultra-fast kinetic protein folding experiments, simultaneous study of folding and dimerization, identification of metal binding domains, and structural characterization of integral membrane proteins.

Summary

OPF is a requirement for cell viability and is as important as protein expression itself, because without the ability to provide the correct disulfide structure, proteins cannot perform complex molecular tasks at the cellular level. Defective and dysregulated disulfide-bonding enzyme systems adversely affect eukaryotic cells at the levels of energy production in mitochondria and protein folding in the ER. These defects result in oxidative stress in the ER, aggregated proteins, potential cell death, and considerable pathology in metabolic, neurological, inflammatory, and neoplastic diseases. OPF is particularly important in diseases that manifest with protein aggregation problems such as sCJD, Alzheimer's disease, and PD. There is still much to learn about OPF and how it contributes to or can prevent disease when it functions correctly in cells. Basic studies of disulfide-bond formation in recombinant protein production will provide clues for which accessory molecules and redox conditions are required for proper OPF. The tools to dissect OPF are being developed and should provide the means to determine structural information in the absence of X-ray crystal structure. Finally, as biochemists unravel the mechanisms of how OPF occurs under various conditions, biologists will be better able to understand why tissues and living organisms behave normally or abnormally in health and disease.