Iron-sulfur proteins hiding in plain sight Tracey A Rouault


© 2015 Nature America, Inc. All rights reserved.

Recent studies suggest that iron-sulfur (Fe-S) proteins may be unexpectedly abundant and functionally diverse in mammalian cells, but their identification still remains difficult. The use of informatics along with traditional spectroscopic analyses could be key to discovering new Fe-S proteins and validating their functional roles.


first became interested in Fe-S proteins when my colleagues and I discovered that a post-transcriptional regulatory protein, IRP1, functioned either as a regulatory protein, without a Fe-S cluster, or as a cytosolic aconitase that interconverted citrate and isocitrate when it contained a [4Fe-4S] cluster (reviewed in refs. 1,2). At that time, inorganic chemists had shown that Fe-S clusters could be readily assembled under anaerobic conditions in vitro in organic solvents (reviewed in ref. 3), but little was known about how they were assembled in cells. In addition, only a few mammalian Fe-S proteins had been identified, and some believed that Fe-S proteins had been gradually replaced over the course of evolution by non–Fe-S proteins to protect processes from oxygen toxicity. One example was the enzyme fumarase, an Fe-S protein in bacteria but not in mammals4. Because the identification of Fe-S proteins requires special spectroscopic techniques and large amounts of protein, I thought it was possible that many mammalian Fe-S proteins might remain unrecognized. Indeed, new approaches have enabled investigators to focus on particular candidates using sensitive and unbiased methods, such as detection of interactions with known Fe-S donor proteins using mass spectrometry5,6, or informatics analysis to detect conserved motifs, as was used in expanding the recognition of members of the radical SAM family of proteins7. Recently, my colleagues and I described a tripeptide motif prevalent in Fe-S proteins that is directly involved in engaging the Fe-S transfer apparatus with recipient proteins8. Thus, for the first time, the power of bioinformatics may potentially be used to identify candidate Fe-S proteins before performing experimental work. The use of these new approaches should lower the 442

barrier to Fe-S discovery by reducing the heavy reliance on complex spectroscopy techniques. In this Commentary, I will discuss the unique properties of Fe-S clusters and the methods used in their discovery and characterization, citing examples of well-known and more recently discovered mammalian Fe-S proteins (Fig. 1). Moreover, I will discuss how the pace of Fe-S protein identification could further be accelerated using an informatics approach and sensitive techniques for detecting bound iron (Fig. 2). The recent progress creates the exciting possibility of elucidating the roles of these uncharacterized iron-sulfur proteins in the chemistry of numerous biological pathways. As one example, the potential involvement of Fe-S proteins in critical biochemical pathways might significantly change how experimental collaborations between bioinorganic chemists and biologists are designed and interpreted, and may alter our general understanding of mammalian biochemistry and energy homeostasis.

What are iron-sulfur clusters?

Iron-sulfur clusters are ancient cofactors composed of iron and sulfur in different and interchangeable stoichiometries, which are usually ligated to cysteines of associated proteins9. They confer properties to associated proteins that cannot be readily mimicked by other prosthetic groups, including the ability to accept and donate single electrons in electron transfer reactions, undergo oxidationreduction reactions, perform regulatory sensing, ligate specific substrate ligands and drive particular protein conformations. Although the first iron-sulfur protein was initially discovered in 1960 (ref. 10), in the mammalian respiratory chain complex II, or succinate dehydrogenase, many of the

known and highly characterized iron-sulfur proteins, including bacterial ferredoxins and nitrogenase, are bacterial11. The key to the discovery of these novel cofactors was the development of the spectroscopic technique known as electron paramagnetic resonance (EPR), in which a magnetic field is used to detect unpaired electrons, which are characteristically found in transition metals12. Mössbauer spectroscopy has also had an important role in studies of ironsulfur proteins, but this approach requires Fe57 to be incorporated into a protein of interest. The electromagnetic emission and absorption of g-rays is characteristic and recognizable when an atomic nucleus is excited by a specific radioactive source (in the Mössbauer effect)13. Many iron-sulfur clusters have a recognizable UV-visible absorption spectrum, but finding an absorption peak on UV-visible absorption spectroscopy does not constitute a definitive test for presence of an iron-sulfur cluster because many contaminants can absorb at 400 nm in the UV spectrum. In the past, the few Fe-S proteins discovered in mammals have sometimes been viewed as evolutionary relics from iron-sulfur–rich prokaryotes. However, numerous discoveries over the last decade suggest that the notion of a gradual loss of Fe-S cofactors in critical enzymes in mammals may not represent a real effect. In fact, the failure to detect iron-sulfur cofactors in numerous mammalian proteins may have reflected difficulties with overexpression and with failure to protect iron-sulfur proteins from oxidative destruction of their Fe-S cofactors before characterization. The initial EPR and Mössbauer studies revealed that ironsulfur clusters are sensitive to oxidation and usually need protection from oxidants during intracellular growth and subsequent purification. Careful regulation of oxygen


commentary concentrations during growth should mimic physiological oxygen concentrations as closely as possible, and harvesting of proteins should be done anaerobically. Many 1960: Magnetic signals detected in EPR in mammalian SDH and xanthine oxidase (1969)

labs currently lack the proper tools, and as a result their experiments suffer as a result of the degradation of Fe-S clusters during preparation. However, proper handling of these proteins along with some recent technical advances have uncovered new biological functions for Fe-S clusters in DNA metabolism and ribosomal function and metabolism, which should stimulate further interest in identifying and characterizing new mammalian Fe-S proteins using the appropriate preparation conditions (Fig. 1).

Early progress 1980–1990s: EPR, Mössbauer and structural proof of mitochondrial aconitase cubane Fe-S cluster


© 2015 Nature America, Inc. All rights reserved.

1985: Stoichiometry of three Fe-S clusters of SDHB worked out 1992: EPR, Mössbauer and structural proof of cytosolic aconitase cubane Fe-S cluster

1997: cloning of NTHL1, human endonuclease III Fe-S protein

2004: Molybdenum cofactor 1A synthase vFe-S cluster identified 2005: Numerous radical SAM enzymes and human LIAS proteins contain Fe-S clusters 2006: Essential Fe-S domains found in XPD and FANCJ 2008: Fe-S clusters of ABCE1 characterized 2012: Fe-S proteins in DNA replication— DNA pol α, δ, ε, DNA primase, DNA2, glycosylases and base excision repair (presumed from homology)—and in RAD3 helicase family, including RTEL1, FANCJ, CHLR1

Figure 1 | A timeline that illustrates the increasing pace of mammalian Fe-S discovery. The first Fe-S cofactors were not found until 1960, lagging behind the discovery of other metal-containing prosthetic groups such as heme. Because Fe-S proteins tend to lack a distinctive visible color, and the UV-visible spectrum can be misleading, new techniques, including EPR and Mössbauer spectroscopies, were required to definitively detect magnetism and assign stoichiometries. In the last ten years, Fe-S proteins have been found at the nexus of ribosomal function and translation and in multiple DNA metabolism proteins. These discoveries were unexpected, and they raise the possibility that many more mammalian Fe-S proteins remain to be identified in central metabolic pathways.

Mitochondrial aconitase and succinate dehydrogenase were among the earliest mammalian Fe-S proteins identified. The [4Fe-4S] cluster of mammalian mitochondrial aconitase was characterized using EPR, Mössbauer and electronnuclear double resonance (ENDOR) spectroscopies, chemical analyses and biochemical experiments. These experiments established that the fourth iron of the cubane Fe-S cluster is involved in ligating the enzymatic substrates, citrate and isocitrate, whereas the other three Fe atoms are ligated by cysteines in the aconitase structure (reviewed in ref. 14). Succinate dehydrogenase subunit B contains three Fe-S clusters, and experiments using magnetic circular dichroism and EPR were crucial in deconvoluting the stoichiometries of these Fe-S clusters (reviewed in ref. 15). The enzymatic activity of both proteins depends on the presence of intact Fe-S clusters. For many years, biochemical assays of mitochondrial aconitase and succinate dehydrogenase activity were used to assess the integrity of Fe-S clusters in mammalian cells. Sensitive colorimetric assays on unpurified cell extracts could quantify their activities, which revealed important information about citric acid cycle function, mitochondrial integrity and oxidative stress.

New developments More recently, the preparation of crystal structures under anaerobic conditions detected the presence of two Fe-S clusters in the bacterial homolog of ABCE1, an ABC cassette protein that plays an apparently central role but incompletely defined role in mRNA translation and ribosomal function. The two [4Fe-4S] clusters are within close enough range to transfer electrons. The Fe-S–rich N terminus is closely attached through a hinge mechanism to the ATP binding cassette portion of a protein that is involved in ribosomal biogenesis, translation initiation and/or formation of translational initiation components16. The authors raised the possibility that ABCE1 represents a chemomechanical engine module in which


the Fe-S domains work in concert with the ABC domain, perhaps using electron transfer to energize its activity. Fe-S clusters have also been discovered in the DNA processing machinery, including enzymes involved in DNA replication, repair, recombination and primase activity17. Proteins required for DNA replication, including DNA polymerases a, d, e, DNA primase and DNA2, glycosylases, and base excision repair proteins were presumed to be Fe-S proteins from homology studies. Recently, an alternative approach for detecting Fe-S proteins through their physical interactions with the Fe-S assembly complex was developed. Mass spectrometry studies identified proteins that interacted with MMS19, a protein that when mutated in yeast causes DNA repair abnormalities5,6. MMS19 was found to interact directly with cytosolic proteins known to be involved in Fe-S transfer (CIAO1 and NARFL)6 as well as with known and previously unrecognized Fe-S proteins involved in DNA metabolism such as DNA polymerase d and the RAD3 helicase. Further studies using EPR confirmed that members of the RAD3 helicase family, including RTEL1, FANCJ, CHLR1 (ref. 5,6), three yeast replicative polymerases18 and human DNA primase17, are Fe-S proteins. The loss of an iron-sulfur cluster in RAD3 helicase results in a failure to unwind DNA19. Finally, a Fe-S cluster was found in the structure of the bacterial homolog of the DNA recombination and repair enzyme DNA2, and its positioning might provide a ‘wedge’ that facilitates separation of DNA strands20. The reasons that Fe-S chemistry is used in DNA metabolism are not yet clear, and different functions have been suggested, including the sensing of DNA damage21 and the maintenance of electronic charge that can be dispersed as needed, in a manner analogous to battery function22. Oxidationreduction reactions of such Fe-S clusters in proteins that separate DNA strands could potentially drive conformational changes that facilitate ‘wedge’ activity. Because reduced Fe-S proteins are more energy rich, their oxidation may also be linked to replication and maintenance of DNA integrity, so that cells can use their energy supplies to maintain vital baseline functions.

Bioinformatics approaches The overall success of using mass spectrometric techniques to detect potential Fe-S proteins through interactions with the Fe-S transfer complex provides a systematic approach that will complement confirmation through traditional methods. An alternative method using the Fe-S transfer system as 443

commentary ID conserved LYR motif in candidate protein Overexpress tagged protein in bacteria with ISC operon Detect presence of iron using ICP-MS Complexity or difficulty?

UV spectroscopy

Positive for iron Scale up for spectroscopic techniques

Mössbauer spectroscopy

Electron paramagnetic resonance (EPR)

+ for Fe-S


© 2015 Nature America, Inc. All rights reserved.


+ for Fe-S


3,400 3,600

Field (Gauss)


–4 –3 –2 –1




Velocity (mm/s)



Figure 2 | A flow chart for an approach to the identification of new Fe-S proteins. Informatics could be used to screen mammalian proteins for a signature motif, LYR, which would enhance the likelihood of a recipient protein receiving nascent Fe-S clusters from the ISCU scaffold protein. Upon identification of an iteration of the LYR motif (step 1), a candidate protein could be overexpressed in bacteria, along with the ISC operon to increase the likelihood of Fe-S cluster insertion (step 2). The purified tagged proteins could then be analyzed by ICP-MS (step 3), which can detect iron at concentrations of 50 parts per billion and is about 10,000 times more sensitive than EPR. Mössbauer spectroscopy is less sensitive than EPR and requires Fe57 to be incorporated into the protein of interest, raising the expense and difficulty of the experiment (step 4). A typical spectrum for a [4Fe-4S] cluster is depicted for EPR and Mössbauer techniques.

a discovery tool was to identify particular motifs that might increase the possibility of a protein receiving an iron-sulfur cluster. Interestingly, recent work from my own group suggests that bioinformatics approaches could be an important aid in identifying Fe-S proteins. Studies on the Fe-S subunit of succinate dehydrogenase (SDHB) revealed that the presence of a tripeptide motif, LYR, was critical for the acquisition of Fe-S clusters. Homology searches allowed us to expand our characterization of the LYR motif to include tripeptides that contain a small hydrophobic residue in position 1 of the motif, followed by a large hydrophobic (Y or F) in position 2 and a positively charged residue (R or K) in position 3, now defined as a signature that could facilitate identification of Fe-S proteins23. Though this motif might have not been expected to provide discriminatory capability, similar objections could have been raised against using the CxxxCxxC motif to expand the radical SAM family of proteins7. LYR motif was found to be critical for binding to the cochaperone HSC20, which enables transfer of nascent Fe-S clusters from the scaffold ISCU to recipient proteins8,23. Several 444

other LYR motif–containing proteins were identified as HSC20-binding proteins in the screen in which SDHB was discovered, including glutamyl-prolyl-tRNA synthetase (EPRS) and cullin 5 (ref. 8), suggesting that these proteins, which are important for protein synthesis and cellular growth pathways24, may be unrecognized Fe-S proteins.

Fe-S proteins and metabolism? Given the large number of cytosolic and mitochondrial Fe-S proteins identified so far, it is possible to envision an extensive network of Fe-S proteins that disperses and stores energy throughout the cell. My own experience working on a protein that senses iron status and regulates expression of multiple mRNAs, IRP1, sensitized me to recognize the potential of Fe-S proteins to use their Fe-S clusters in sensing and regulation1. Having Fe-S clusters embedded in central elements of the proteome would allow proteins involved in protein synthesis, DNA metabolism and growth pathways to sense the reducing environment of the cell. Each Fe-S protein might have a different oxidation potential, allowing pathways to

sense a broad range of redox conditions. The reducing potential of a cell, as captured by NADH, NADPH, glutathione, and potentially a pool of iron-sulfur proteins, might represent a potent source of energy currency analogous to the ATP energy pool. I think it is possible that the activities of many Fe-S proteins are regulated by the oxidation status of the cell and may regulate the growth and division of cells. Protein redox activity is commonly thought of in terms of disulfides versus free cysteines, but what if an entire group of Fe-S proteins is designed to sense and modulate the reduction potential of the cell, reducing growth as the cell becomes progressively more oxidized? In metabolic pathways, the information from each Fe-S protein could change sensing from an on-off switch to a rheostatic form of sensing in which some proteins of the pathway diminish or heighten activity of an overall process, whereas others would remain unaffected. I propose that some central metabolic pathways involved in sensing energy status may contain unrecognized Fe-S proteins that are regulated by redox conditions (Fig. 3).

Workflow for Fe-S protein discovery How can we accelerate discovery of mammalian Fe-S proteins and insights into their function? I believe that the use of informatics to identify proteins containing LYR motifs could greatly facilitate new discovery of Fe-S proteins. Before this past year, there were no motifs identified that could characterize potential Fe-S proteins. Because both EPR and Mössbauer spectroscopy require large amounts of protein, more sensitive techniques are needed for initial identification of potential Fe-S proteins. I outline an algorithm (Fig. 2) to potentially enhance discovery by first identifying highly conserved LYR motifs in mammalian proteins and then validating that they are Fe-S recipients with a sensitive technique that quantitatively detects iron and sulfur atoms. In order to obtain sufficient amounts of protein, a tagged version of a candidate protein will need to be overexpressed in bacteria that are transformed with the ISC (bacterial ironsulfur cluster) operon to optimize insertion of Fe-S clusters11. The protein could then be immunoprecipitated under anaerobic conditions to protect the presumed Fe-S cluster and analyzed using inductively coupled plasma mass spectrometry (ICPMS), a type of mass spectrometry that is capable of detecting iron and sulfur at concentrations as low as 50 parts per billion15,25. ICP-MS requires much smaller amounts of sample protein for analysis than the other major spectrometry techniques.


Citric acid cycle and respiratory chain activity

tRNA synthetases (EPRS, etc.)

Energy sensing pathways?

Multiple DNA metabolism proteins Translational initiation and elongation


© 2015 Nature America, Inc. All rights reserved.

Ribosomal biogenesis and function



Figure 3 | Pathways in which Fe-S proteins may be critical. It has long been known that the Fe-S enzymes mitochondrial aconitase and succinate dehydrogenase are critical to citric acid cycle activity and that respiratory chain complexes I–III depend on Fe-S clusters for function. The last decade has revealed that Fe-S clusters are required for ribosomal function and translation (ABCE1) and for numerous DNA metabolism proteins. Fe-S clusters would be ideal cofactors in central energy-sensing pathways, where they could both store energy in the form of reducing potential and also sense whether cellular energy stores are replete.

Moreover, ICP-MS can sometimes give accurate stoichiometry for metal-binding proteins, and discovery of four iron atoms per protein enhances the likelihood that an Fe-S cluster is present. Positive results from informatics and ICP-MS could provide an impetus for researchers to scale up and perform EPR and Mössbauer analyses, which would then verify the presence of a true Fe-S cluster (Fig. 2). An advantage of this workflow scheme is that a relatively small commitment of resources and time is required to perform the early steps. Detection of bound iron by ICP-MS might incentivize researchers to overcome experimental impediments such as failure to detect a signal on EPR, which can sometimes be overcome by oxidizing or reducing the proteins of interest to allow signal detection.

Future perspectives

Given the surprise that large classes of nuclear and cytosolic proteins are newly

recognized Fe-S proteins, it is possible that there may be many more Fe-S proteins than are presently recognized. If these proteins are indeed widespread throughout mammalian biochemical pathways, it may open up a whole new perspective on chemical biology. It is possible that these proteins could function as energy reservoirs and sensors that regulate activity of metabolic pathways. In addition, their stored energy could be used in multiple energy-requiring reactions, such as separation of DNA strands, and in multiple biosynthetic reactions, including protein synthesis (Fig. 3). Determining the role of Fe-S clusters in mechanism of action will require productive collaborations between chemical biologists and biochemists over a wide range of processes. Despite all of these recent advances, I have long wished that approaches that don’t require large amounts of holoprotein could be developed to enable researchers to


identify candidate Fe-S proteins. I hope that chemical biologists can improve the existing techniques and methodologies, or that more sensitive and definitive techniques can be developed to lower the barrier to Fe-S protein discovery. Tracey A. Rouault is at the Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda, Maryland, USA. e-mail: [email protected] References 1. Rouault, T.A. Nat. Chem. Biol. 2, 406–414 (2006). 2. Beinert, H., Kennedy, M.C. & Stout, C.D. Chem. Rev. 96, 2335– 2374 (1996). 3. Venkateswara Rao, P. & Holm, R.H. Chem. Rev. 104, 527–559 (2004). 4. Flint, D.H., Tuminello, J.F. & Emptage, M.H. J. Biol. Chem. 268, 22369–22376 (1993). 5. Gari, K. et al. Science 337, 243–245 (2012). 6. Stehling, O. et al. Science 337, 195–199 (2012). 7. Frey, P.A., Hegeman, A.D. & Ruzicka, F.J. Crit. Rev. Biochem. Mol. Biol. 43, 63–88 (2008). 8. Maio, N. et al. Cell Metab. 19, 445–457 (2014). 9. Beinert, H., Holm, R.H. & Munck, E. Science 277, 653–659 (1997). 10. Beinert, H. & Sands, R.H. Biochem. Biophys. Res. Commun. 3, 41–46 (1960). 11. Johnson, D.C., Dean, D.R., Smith, A.D. & Johnson, M.K. Annu. Rev. Biochem. 74, 247–281 (2005). 12. Münck, E., Surerus, K.K. & Hendrich, M.P. Methods Enzymol. 227, 463–479 (1993). 13. Petasis, D.T. & Hendrich, M.P. in Iron-Sulfur Clusters in Chemistry and Biology (ed. Rouault, T.) 21–48 (deGruyter, 2014). 14. Beinert, H. & Kennedy, M.C. Eur. J. Biochem. 186, 5–15 (1989). 15. Singer, T.P. & Johnson, M.K. FEBS Lett. 190, 189–198 (1985). 16. Karcher, A., Schele, A. & Hopfner, K.P. J. Biol. Chem. 283, 7962– 7971 (2008). 17. White, M.F. & Dillingham, M.S. Curr. Opin. Struct. Biol. 22, 94–100 (2012). 18. Netz, D.J. et al. Nat. Chem. Biol. 8, 125–132 (2012). 19. Weiner, B.E. et al. J. Biol. Chem. 282, 33444–33451 (2007). 20. Saikrishnan, K. et al. EMBO J. 31, 1568–1578 (2012). 21. Grodick, M.A., Segal, H.M., Zwang, T.J. & Barton, J.K. J. Am. Chem. Soc. 136, 6470–6478 (2014). 22. Wu, X. et al. Proc. Natl. Acad. Sci. USA 102, 14058–14062 (2005). 23. Rouault, T.A. Nat. Rev. Mol. Cell Biol. 16, 45–55 (2015). 24. Bulatov, E. et al. J. Biol. Chem. 290, 4178–4191 (2015). 25. Holmes-Hampton, G.P., Tong, W.H. & Rouault, T.A. Methods Enzymol. 547, 275–307 (2014).

Acknowledgments I thank G. Holmes-Hampton and W.H. Tong for helpful discussions and G. Holmes-Hampton for generating the tracings in Figure 2.

Competing financial interests The author declares no competing financial interests.


Iron-sulfur proteins hiding in plain sight.

Iron-sulfur proteins hiding in plain sight. - PDF Download Free
771KB Sizes 0 Downloads 11 Views