9568 Biochemistry, Vol. 48, No. 40. 2009
`19. l-ludder, B. N.. Morales, l. G., Stubna. A.. Mitnck, E.. Hcndrich.
`M. l’.. fad Lindahl. P. A. (2007) Electron paramagnetic resonance
`and Mossbauer spectroscopy of intact mitochondria from rcspiring
`Sacrharamyces t‘erevi.riae. J. Biol. Inorg. Chem. /2. l029-I053.
`. Lindahl. F., Morales. .l. (1. Miao. R.. and Holmes-Hampton. G.
`(2009)
`Isolation of .5'at‘(‘hI1I0n1_yt*¢'.s certviriae mitochondria for
`Mnssbaucr, EPR. and electronic absorption spectroscopic analyses.
`Method: Enzymol. 456. 267-285.
`. Schilkc, B.. Vols-inc. C.. Eeinert, H..and Craig. E. (1999) Evidence for
`a conserved system for iron metabolism in the mitochondria of
`Sacrharnmyces cerevisine. Proc. Natl. Acad. Sri. U.S.A. 96. lD20fr
`l02ll.
`. Kohll-«aw. G. B. (1988) lsopropylmalatc dehydratase from yeast.
`Method: Enzymal. 166, 423429.
`. Munujox. F., Collcanti..l.. Gonzalezsastre. F., and Gella. FJ. (1993)
`Assay of suocinatc—dehydrogenase activity by a v;olorimctric-contin-
`Lrous method using iodonitrotetrnzolium chloride as electron-accep
`tor. Anal. Biochem. 212. 506-509.
`. Tamarit, .l., lrazusta, V.. Morcno—Ccrrneno, A... and Ros, J. (2006)
`Colorimetric assay for the qunrttitation of iron in yeast. Anal.
`Biurhem. J51. l49- l5l.
`. Poiurkova, A. V., and Rehr. .l. 1. (I999) Multiple-scattering x-ray-
`absorption fine-structure Debyevwaller factor calculations. Phys.
`Rev. 859,948 957.
`. Plattner. H., and Schatz, G. U969) Promitochondria ofanaerohicully
`grown yeast. 3. Morphology. Bioz'hen:i:tr_v8. 339 343.
`. Kwnst. K. E., Lai. L. C.. Menda. N., James. D. T.. Aref, 5., and
`Burke, P. V. (2002) Genorrricanalyses of anaerobically induced genes
`in Sarcharanr)'ce.!
`cmrvisiae: Functional
`roles of Rex!
`and
`other factors in mediating the anoztic response. J. Baclelial.
`I84.
`250-265.
`. Paltauf, F., and Sclratz. G. (1969) Promilochondria ofanaerobically
`grown yeast. 2. Lipid composition. Bt'ot'hemirIr_v 8, 33$339.
`. Rosenfcld. E., and Beauvoit. B. (2003) Role of the non-respiratory
`pathways in the utilization of molecular oxygen by Sm'rharam)'¢'e.:
`rerevisiae. Year! 20. lll$—l I44.
`. Barres. M. H., Nobrega, F. 0.. and Tzagolofl. A. (2002) Mitochon-
`drial ferredoxiri is required for heme A synthesis in Sot'chammyre.t
`rerevixiae. J. Biol. Chem 277. 9997-10002.
`. Bultcnu, A. L.. Dancis. A.. Gnreil. M., Montague, l. .l.. Camadra.
`.l. M., and Lesuisse. E. (2007) Oxidative stress and protease dysfunc-
`tion in the yeast model of Friedrcich ataxia. Free Radical Biol. Merl.
`42. I561 l570.
`. l-lassett, R. F., Romeo, A. M., and Kosman, D. J. (1998) Regulation
`of high affinity iron uptake in the yeast Sacrharnmycer rer:w'.riue—
`Role of dioxygen and Fe(ll). .1. Biol. Chem. 273, 7628-7636.
`
`_
`
`. Jensen. L. T.. and Culuttu. V. C. (2002) Regulation ofsarcharamyces
`t-ereviiriue FET4 by oxygen and iron. J. Mal. Biol. 318. 25!-260.
`. waters. B. M., and Eide. D. l. (2002) Combinatorial control ofyeast
`FET4 gene expression by iron, zinc. and oxygen. .1. Biol. Chem. 277.
`33749-33757.
`. Cavaclini. P.. Biasiotto, (3.. Poll. M.. Levi. S.. Verardi. R., Lanella. l.,
`Derosas. M..lngrassit1. R.. Corrado. M., and Arosio, P. (2007) RNA
`silencing of the mitochondrial ABCE7 transporter in HeLa cells
`causes an iron-deficient phenotype with mitochondrial iron overload.
`Blood 109. 3552-3559.
`. Halliwell, B. (2009) The wanderings of a free radical. Free Radical
`Biol. Med. 46. 531-542.
`. lmlay. J. A.
`(2006) lron-sulphur clusters and the problem with
`oxygen. Mal. Microbial. 59. 1073-1082.
`.MlihlenhotT. U.. Richhardt. N.. Gerber. L and Lill, R.
`(2002)
`Characterization of iron-sulfur protein assembly in isolated mito-
`chondria-A requirement for ATP. NADH, and reduced iron. J. Biol.
`Chem. 277, 29810-29816.
`. Rnuen, U.,Springer. A.. Weisheit. D.. Petrat. F., Korth. H. G.. deG:-oat.
`H.. and Sustrnann. R. (2007) Assessment of chelauahle mitochondrial
`iron by using mitochondrion-selective fluorescent iron indicators with
`dilferent iron-binding allinities. ChemBiaChem 8, 341-352.
`.Cambridgc
`Structural Database
`(http://www.ccdc.cam.ac.uk/
`products/csd/).
`.Wilkinson. E. C.. Dung. Y. H., and Que. L.
`(1994) Modeling
`hydrolysis at dinucleat iron centers. J. Am. Chem. Soc. 116. 8394-
`8395.
`.Lange,
`I-l.. Kispal. (i.. and Lill, R.
`(1999) Mechanism of iron
`transport to the site of heme synthesis inside yeast mitochondria.
`J. Biol. Chem. 274. i898‘)-H9996.
`. Lill, P... and Miihlenhoft‘. U. (2006) Iron-sulfur protein biogcncsis in
`eukaryotes: Components and mechanisms. Annu. Rev. Cell. Dev. Biol.
`22, 457-486.
`. Schueck. N. D., Woontner, M., and Koeller. D. M. (200l)The role of
`the mitochondrlon in cellular iron homeostasis. Milnchondriari 1,
`51-60.
`. Burke, M. A., and Ardehali. H. (2007) Mitochondrial ATP-binding
`cassette proteins. Trurtvl. Res. [50, 73-80.
`. Chen. S.. Sanchcz—l-‘crnandez. R.. Lyver, E. R.. Dancis, A., and Rea,
`P. A. (2007) Functional chut‘actcrization ofAtATM 1, AtATMZ, and
`AtATM3, a subfamily of Arabidopsis half-molecule ATP-binding
`cassette transporters implicated in iron homeostasis. J. Biol. Chem.
`282, 21561-21571.
`.Liu. W. T.. and Thorp. H. H. (1993) Bond valence sum analysis
`of metal-ligand bond lengths in mctallocnzyrrtes and model com-
`plexes 2. Refined distances and other enzymes. Irmrg. Chem. 32.
`4102-4105.
`
`Biochemistry 2009, 48, 9569-9581
`DOI:
`l0.l02l/bi90ll82w
`
`9569
`
`The Yeast Iron Regulatory Proteins Grx3/4 and Fra2 Form Heterodimeric Complexes
`Containing a [2Fe-2S] Cluster with Cysteinyl and Histidyl Ligationl
`
`Haoran Li,‘ Daphne T. Mapolelofl Nin N. Dingrafxsunil G. Naik," Nicholas S. Lees} Brian M. Hoffman,‘
`Pamela J. Riggs-Gelascof’ Boi Hanh Huynh,' Michael K. Johnson,” and Caryn E. Outten*'*
`‘Department ofChemistry and Biochemistry, Univerrily o/South Carolina. Columbia, South Carolina 29208,
`°Department of Chemistry and Cemerfar Melalloerzzyme Studies, University of Georgia, Athens, Georgia 30602. Department of
`Physics, Emory University, Atlanta, Georgia 30322, ‘Chemistry Department, Northwestern University, Evanston. 1ll1'n0Is60208.am1
`Departmerrt of Chemistry and Biochemistry, College of Charleston. Charleston, Soul/1 Carolina 29424
`Received July 10, 2009: Revised Manuscript Received A ugust 28, 2009
`
`ABSTRACT: The transcription of iron uptake and storage genes in Sacrharomycer cerevisiae is primarily
`regulated by the transcription factor Aftl. Nuclcocytoplasmic shuttling oi‘ Aftl
`is dependent upon
`mitochondrial Fe-S cluster biosymhesis via a signaling pathway that includes the cytosolic monothiol
`glutaredoxins (Grx3 and Grx4) and the BolA homologue Fra2. However, the interactions between these
`proteins and the iron—dependent mechanism by which they control Aftl
`localization are unclear. To
`reconstitute and characterize components of this signaling pathway in vitro, we have overexpressed yeast
`Fra2 and Grx3/4 in Escherichia ca/i. We have shown that coexpression of recombinant Fra2 with Grx3 or
`Grx4 allows purification of a stable [2Fe-2S]“ cluster-containing Fra2-Grx3 or Fra2—Grx4 heterodimerlc
`complex. Reconstitution of a [2Fe~2S] cluster on Grx3 or Grx4 without Fra2 produces a [2Fe—2S]-bridged
`homodimer. UV~—visible absorption and CD, resonance Raman. EPR, ENDOR, Méissbnuer, and EXAFS
`studies of [2Fe-ZS] Grx3/4 homodimers and the [2Fe-2S] Fra2—Grx3/4 heterodimers indicate that inclusion of
`Fra2 in the Grx3/4 Fe-S complex causes a change in the cluster stability and coordination environment.
`Taken together. our analytical, spectroscopic, and mutagenesis data indicate that Grx3/4 and Fra2 form a Fc-
`S-bridged heterodimeric complex with Fe ligands provided by the active site cysteine ofGrx3/4, glutathione,
`and a histidine residue. Overall, these results suggest that the ability ofthe Fra2-Grx3/4 complex to assemble a
`[2Fe-2S] cluster may act as a signal to control the iron regulon in response to cellular iron status in yeast.
`
`Maintenance of optimal iron levels inside the cell is critical for
`all eukaryotes and most prokaryotes, for iron is both essential
`and potentially toxic. As a protein cofactor, iron can bind directly
`to amino acids, forming mono- or dinuclear iron centers, or it can
`be incorporated with porphyrins or sulfide to form hcmc or Fe-S
`clusters, respectively. However, uncontrolled iron redox chem-
`istry can lead to formation of reactive oxygen species, causing
`extensive cellular and molecular damage that eventually leads to
`cell death (1). Therefore, cells must be able to sense iron levels and
`regulate iron homeostasis accordingly to maintain critical, non-
`toxic levels of this key nutrient.
`The expression of iron uptake and storage genes in the model
`eukaryote Saccharomyces cerevisiae is primarily controlled by the
`Fe-responsive transcription factor Aftl. Aftl is located in the
`cytosol under iron-replete conditions and moves to the nucleus
`
`llliis work was supported by the NIH (ESl3780. C.E.0.: GM62524.
`M.K.J,; GM47295, B.H.l-l.; HLl353l, B.M.H.; P20 RR0l646l, P..l.R.
`-G.) and by the Camille and Henry Drcyfus Foundation (Henry Dreyfus
`Teacher-Scholar Award to P.J.R.-G.). Both the National Synchrotron
`
`under iron—depleted conditions, where it activates genes involved
`in high-affinity ionic iron uptake. sidcrophore iron uptake. and
`vacuolar
`iron transport, known collectively as
`the iron
`regulon (2-7). Nuclcocytoplasmic shuttling of Aftl in yeast is
`regulated by mltocliondrial Fe-S cluster biosynthesis via a
`signaling pathway that includes the cytosolic monothiol glutar-
`edoxins (Grx3' and Grx4),
`the BolA homologue Fra2 (Fe
`repressor of activation-2), and the aminopeptidase P-like protein
`Fral (Fe repressor of activation-l) (8-1 1). When Fe-S cluster
`biosynthesis is active (i.e., under Fe replete conditions),
`this
`signaling pathway is proposed to induce multimerization ofAftl
`in an unknown manner to promote its export from the nu-
`cleus (12). A MCXCZ” motifis required for Aftl trtmslocation in
`response to iron. but the specific function of these residues is
`unclear (2, I l, 12). Under low iron conditions or upon disruption
`of mitochondrial Fe-S cluster biogenesis, the Fral/Fra2/Grx3/
`Grx4 pathway is shut offl allowing Aftl to accumulate in the
`nucleus and activate iron uptake systems. Despite the identifica-
`tion of multiple components in this signaling pathway,
`the
`
`

`
`9570 Biochemistry, Val. 48, Na. 40, 2009
`specific mechanism of iron—dependent regulation of Aftl locali-
`nation by Fral and Grx3/4 is a key gap in our understanding of
`intracellular iron metabolism.
`Yeast Grx3 and Grx4 are members of the monothiol glutar-
`edoxin (Grx) family, which is found in organisms ranging from
`bacteria to humans. Grid and Grx4 are highly homologous
`proteins that possess both an N-terminal thioredoxin~ (Trx-) like
`domain and a C—terminal Grx-like domain (13). Cytosolic Grit}
`and Grx4 perform redundant functions since deletion of each
`gene singly has little effect on iron regulation, while a grx3A
`gr.r4A double mutant exhibits constitutive expression of iron
`regulon genes (9, 10). The putative active site in the Grx~like
`domain of Grx3/4 has a highly conserved CGFS motif that is
`specifically required for interaction with M1! and regulation ofits
`activity (9). A third CGFS-type monothiol Grx in yeast (Grx5) is
`localized to mitochondria where it plays an essential but
`ill—defined role in mitochondrial Fe-S cluster biogenesis (14, 15).
`The human homologue of the yeast cytosolic monothiol Grxs
`(hGrx3, also known as PICOT for PKC—interacting cousin of
`Trx) was initially identified in T lymphocytes where it acts as a
`negative regulator of protein kinase C-9 via an unknown mechan-
`ism (I6). More recently, mammalian Grx3 was shown to inhibit
`cardiomyocyte hypertrophy (i.e.. thickening of the heart muscle)
`by binding to the muscle LEM protein (ML?) and blocking the
`stress-responsive. prohypertrophic calcineurin-nuclear factor of
`the activated T-txll (NFAT) signaling pathway (17).
`Grxs are part of the Trx~fold superfamily and typically
`catalyze thiol-disulfide exchange reactions via monothiol or
`dithiol mechanisms ( 18). The dithiol mechanism requires two Cys
`in the active site (usually in a CPYC motif), with the N-terrninal
`Cys forming a mixed disulfide between Grx and the target
`protein, which is resolved by the second active site Cys. The
`monothiol mechanism requires only the N-terminal Cys in the
`active site and is used for reducing mixed disulfides between GSH
`and target proteins. Although the N-terminal Cys is conserved in
`CGFS-type monothiol Grx, they lack oxidoreductase activity
`when tested with standard Grx model substrates (18). Several
`members ofthe monothiol Grxs family, including yeast Grx3 and
`Grx4. were recently shown to fonn [2Fe-2S]~bridged homodimers
`with the CGFS active sites providing two Cys ligands (I9-21).
`interestingly. two glutathione molecules (GSH or y—glutamyl-
`cysteinylglycine) provide the other two cluster ligands. GSH is an
`abundant tripeptide that serves as the primary intracellular thiol
`redox buffer and a cofactor for glutaredoxins and other anti-
`oxidant enzymes (22). in the three published crystal structures of
`[2Fe-2S]-bridged dlthiol and monothiol Grx proteins, the GSH
`molecules are covalently linked to the cluster but held in place by
`noncovalent interactions with theGSH binding pocket in the Grx
`protein (21. 23, 24). GSH seems to play a role in yeast Grx3/4
`function since GSH binding residues in Grx3/4 are essential for
`regulation of Aftl activity (9). Based on previous studies. several
`possible functions for monothiol Grxs in iron metabolism have
`been proposed. Grits may act as scaffolds for Fe-S cluster
`
`Li et al.
`
`Article
`
`Escherichia call‘ as a protein that induces a round cell shape when
`overcxpressed (26). Although BolA-like proteins are found in
`a wide variety of prokaryotes and eukaryotes, their specific
`molecular function is unknown. Genome-wide yeast two-hybrid
`assays have identified a physical interaction between cytosolic
`monothiol Grxs and BolA-like proteins in both S. cerevisiae and
`Drasaphila mclanogaster (27, 28). In addition, proteomewide
`FLAG- and TAP-tag affinity purification studies in yeast and
`E. cali have shown that BolA-like proteins such as Fra2 copurify
`with monothiol Grxs (29-31). The physical interaction between
`yeast Cvrx3/Grx4 and the BolA-like protein Fra2 was recently
`confirmed by immunoprecipitation and split YFP tagging (8).
`Finally, comparative genomic analyses also predict a functional
`interaction between monothiol Grxs and BolA-like proteins
`since they are neighbors in several prokaryotic genomes (25).
`Furthermore. genes encoding the BolA and monothiol Grx
`proteins exhibit strong genome cooccurrence since almost all
`organisms that possess a BolA-like protein also have a CGFS
`monothiol Grx, while organisms that lack a BolA-like protein
`almost always lack a CGFS-type Grx (25, 32). Taken together,
`these data demonstrate a strong phylogenetic connection be-
`tween the two protein families. However,
`the nature of the
`interaction between CGFS-type Grxs and BolA-like proteins
`has not been previously determined,
`This study is aimed at characterizing the interaction between
`yeast Grx3/4 and Fra2 in vitro using biochemical, spectroscopic,
`and analytical techniques. We show here that both Grx3 and
`Grx4 form [2Fe-2S]—containing heterodimers with Fra2 that have
`similar Fe-S coordination environments. However, the UV-vi-
`sible absorption. CD.
`resonance Raman. EPR, ENDOR,
`Mossbauer, and EXAFS spectra of reconstituted [2Fe-2S]-
`bridged Grx3 or Grx4 homodimers are markedly different from
`[2Fe-ZS] Fra2-Grx3/4 heterodimers,
`indicating differences in
`cluster coordination. Furthermore, we have detennined that
`conserved residues required for Grx3 and Grx4 Fe signaling
`in vivo are also required for Fe-S complex formation with Fral
`in virro. This study thus provides new insight into the molecular
`details of intracellular Fe signaling and establishes the ubiquitous
`monothiol Grxs and BolA«like proteins as a novel type of Fe-S
`cluster binding regulatory complex.
`
`EXPERIMENTAL PROCEDURES
`Plasmid Construction. Construction of the yeast Grx3
`E. cali expression vector pET2la-Grx3 was described pre-
`viously (33). The ORF of yeast Grx4 were amplified from
`S. cerevisiae genomic DNA by PCR using the primers shown
`in Supporting Information Table l and cloned into the Ndel and
`EL-URI sites of pET2la (Novagen) to generate pET2la-Grx4.
`E. cult‘ expression vectors for His—tagged Fral (pET2la-Fral-
`His.) and Fra2 (pET2la-Fra2-His6) were kindly provided by
`Jerry Kaplan (University of Utah) (8). The untagged Fra2
`expression vector pET2la~Fra2 was constructed by amplifying
`the Fra2 ORF without the His tag from pET2la-Fra2~His,<, and
`
`(QuikChange kit; Stratagene) using primers listed in Supporting
`Information Table 1. pET2la—Grx3(Al-121) was created by
`introducing an Ndel restriction site at position 122 in pET2la—
`Grx3 by site-directed mutagenesis, digesting the plasmid with
`Ndel to remove the coding sequence for amino acids I-121, and
`religating the plasmid. pET‘2la-Grx3(Al22-250) were created
`by introducing a stop codon and Hindlll site at position 122,
`digesting the plasmid with Hindlll to remove the coding sequence
`for amino acids 122-250, and religating the plasmid. The cDNA
`for human Grx2 (hGrx2) (Open Biosystcms) was PCR amplified
`without the mitochondrial targeting signal (amino acids 41-164)
`and subcloned into pET24cl (Novagcn) using the N001 and
`Ec0Rl restriction sites to make pE'I24d-hGrx2. The sequence
`integrity ofall plasmids was continued by double—stranded DNA
`sequencing (Environmental Genomics Facility. University of
`South Carolina School of Public Health).
`Protein Expression and Purifimtion. Recombinant Grid
`and Grx4 were both purified using the previously published
`protocol for Grid (33). We note that recombinant Grx3 was
`cloned from the second start site (encoding Met36) to the stop
`codon after determining that the first start site (encoding Metl)
`was not utilized in viva (N. Dingra and C. Outten, unpublished
`data). The Grx3 amino acid sequence numbering in this study
`thus starts with the second start site as Metl. Grx3 (or Grx4) was
`coexprcsscd with Fral and Fra2 by transforming pET2la—Grx3
`(or pET2la-Grx4) and pRSFDuet-l—Fral-l-Iisg/Fra2 into the
`E. cali strain BL2l (DE3). Generally, a l L LB culture was grown
`with shaking at 30 °C and induced with 1 mM isopropyl
`fi—1)—thiogalactoside (IPTG) at OD”, 0.6-0.8. The cells were
`collected by centrifugation 18 h after induction and resuspended
`in 50 mM Tris-MES. pH 8.0,
`followed by sonication
`and ccntrifugation to remove the cell debris. The cell—free
`extract was loaded onto a DEAE anion-exchange column (GE
`Healthcare) equilibrated with 50 mM Tris-MES, pH 8.0.
`The protein was eluted with a salt gradient. and the fractions
`containing Grx3 (or Grx4) and Fra2 were pooled and concen-
`trated to 2 mL. A fraction of Fra2 that was not bound to Grx3
`was also present in the DEAE flow-through and further purified
`as described below. (NH4)2SO4 was added to the Fra2-Grx3
`(or Fra2—Grx4) complex to a final concentration of l M, and
`the sample was loaded onto a phenyl-Sepharose column (GE
`Healthcare) equilibrated with 50 mM Tris-MES, pH 8.0,
`100 mM NaCl. and l M (NH..);SO4. The protein was then eluted
`with a decreasing (NH4)2S04 salt gradient, and the fractions
`containing Grx3 (or Grx4) and Fra2 were concentrated and
`loaded onto a HiLoad Superdex 75 gel filtration column (GE
`Healthcare) equilibrated with 50 mM Tris-MES, pH 8.0, and
`150 mM NaCl. The purest fractions ofthe Fra2—Grx3/4 complex
`as judged by SDS-PAGE were collected and concentrated to
`~250 ;4L with the addition of 5% glycerol and stored at -80 °C.
`Purification of Fra2-Grx3/4 was done aerobically: however, the
`procedure was completed in 1 day using degassed buffers to
`minimize loss of the Fe~S cluster.
`57]-"e-labeled samples ofthe Fra2-Grx3 or F1212-Grx4 complex
`
`9571
`Biochemistry, Vol. 48, Na. 40, 2009
`centrifugation l8 l1 after induction. Subsequent purification of
`l7Fe-labeled F1a2-Grx3 or Fra2-Grx4 utilized the same protocol
`described in the previous paragraph.
`For purification of Fra2 without Grx3 or GIX4, BL21(DE3)
`E. coli cells were transformed with pET2la-Fraz. Fra2 expressed
`in this strain resides in inclusion bodies, thus requiring unfolding
`and subsequent refolding to purify. However, uncomplexed Fra2
`could be purified from the DEAE column l'low—through from
`cells coexpressing Grx3 and F1112 as mentioned above. The Fraz-
`containing DEAE flow—through fractions were adjusted to pH
`6.0 and loaded onto an SP FF cation-exchange column
`(GE Healthcare) equilibrated with 50 mM MES—Na, pH 6.0.
`Fra2 was eluted with a salt gradient, concentrated, and loaded
`onto a HiLoad Superdex 75 gel
`filtration column (GE
`Healthcare) equilibrated with 50 mM Tris-MES, pH 8.0. and
`ISO mM NaCl. The purest fractions of Fraz were collected.
`concentrated to ~500 )rL with the addition of 5% glycerol, and
`stored at -80 °C. The yield of uncomplexed Fra2 from the
`DEAE flow-through was highest when Fraz was coexpressed
`with Grx3(Cl76S) (see Results).
`Recombinant hGrx2 was overexpressed in the E. calf BL2l-
`(DE3) strain and grown at 37 °C with shaking until ODW, = 0,6.
`The cultures were cooled to 20 °C, and 1 mM IPTG was added to
`induce hGrx2 expression. After overnight growth (~18 h) at
`room temperature, cells were harvested by centrifugation and
`stored at -80 °C. The cell pellet was subjected to three freeze-
`thaw cycles, and soluble protein was extracted with 50 mM Tris»
`lICl,
`131'] 8.0. The protein was precipitated with 25-60%
`(Nl—I4);SO4 and the pellet resuspended in 50 mM Tris-I'lCl, plrl
`8.0, and subsequently loaded on a desalting column followed by a
`DEAE column (GE Healthcare) both equilibrated with 50 mM
`Tris-HCL pH 8.0. The majority of hGrx2 did not bind to the
`DEAE column and was collected in the flow—through. These
`fractions were concentrated and loaded onto a HiLoad Superdex
`75 gel filtration column (GE Healthcare) equilibrated with
`50 mM Tris-l-ICl. pH 8.0, and 150 mM NaCl. [2Fe-2S] hGrx2
`elutes as a dimer, while apo hGrx2 elutes as a monomer as
`previously reported (34).
`In Vitro Fe-S Cluster Rez‘or1.r2‘1'ruf1'ar1 an A110 GrX3 and
`Grx4. Reeonstitution of an Fe-S cluster on ape Grx3, 1 mM in
`I00 mM Tris-HCl buffer at pH 7.8, was carried under anaerobic
`conditions (0; < 5 ppm) in a glovebox (Vacuum Atmospheres,
`Hawthorne, CA). The reaction mixture involved 2 mM GSH, a
`16-fold excess of ferrous ammonium sulfate (57Fe—labeled for
`Miissbauer samples) and L»cysteine, and catalytic amounts of
`Azotobacter vinelandii NifS (6.27,uM) and was incubated at room
`temperature for 50 min. Reagents in excess were removed
`anaerobically by loading onto a High-Trap Q-Scpharosc column
`(GE Healthcare) inside the glovebox. Elution was achieved using
`a NaCl gradient with cluster-bound Grx3 eluting between 0.60
`and 0.70 M NaCl. Samples were pooled together as a single
`fraction before concentrating and desalting using Amicon ultra-
`filtration with a YMl0 membrane. The same protocol was
`followed for reconstituting an Fe-S cluster on apt) Grx4.
`
`

`
`9572 Biochemistry. Vol. 48, N0. 40. 2009
`
`5,5'-dithiobis(2-nitrobcnzoic acid)-GSSG reductase cycling
`assay as previously described (38).
`Analytical and Spectroscopic Methods. Analytical gel
`filtration analyses were performed on a Superdex 75 10/300
`GL column (GE Healthcare) equilibrated with 50 mM
`Tris-MES, pH 8.0. and 150 mM NaCl and calibrated with
`the low molecular weight
`gel
`filtration calibration kit
`(GE Healthcare). The buffer was bubbled with N2 overnight
`and degassed to minimize dissolved 02 levels. Elution profiles
`were recorded at 280 nm with a flow rate of 0.5 mL/min.
`UV-visible absorption spectra were recorded using a Beckman
`DU-800 or Shimadzu UV-310] spectrophotometer, and CD
`spectra were recorded on identical samples using a lasco J-715
`spectropolarimeter. Mass spectrometry analysis of purified pro-
`teins was determined using a Bruker UltraFlex MALDI~TOF/
`TOF mass spectrometer. A saturated solution ofsinapinic acid in
`50% acetonitrile and 0. l% nifluoroacetic acid was used as the
`matrix, and myoglobin and ubiquitin were the calibration
`standards. Resonancx: Raman spectra were recorded as pre-
`viously described (39), using an Instruments SA Ramanor
`Ul00O spectrometer coupled with a Coherent Sabre argon ion
`laser, with 20;¢L frozen droplets of l.5~2.6 rnM sample mounted
`on the cold finger of an Air Products Displex Model CSA-202E
`closed cycle refrigerator. X-band (--9.6 GI-I2) EPR spectra were
`recorded using a ESP-300D spectrometer (Bruker, Billerica,
`MA). muipped with an ESR 900 helium flow cryostat (Oxford
`Instruments, Concord. MA). and quantified under nonsaturating
`conditions by double integration against a 1.0 mM CuEDTA
`standard. Q-band (~35 GHz) CW ENDOR experiments were
`carried out at 2 K using a spectrometer described previously (40).
`Mossbauer spectra were recorded by using the previously de-
`scribed instrumentation (41). The zero velocity of the spectra
`refers to the centroid of a room temperature spectrum of a
`metallic Fe foil. Analysis of the Mdssbauer data was performed
`with the WMOSS program (SEE Co.).
`X-ray absorption spectra were measured at
`the National
`Synchrotron Light Source (NSLS) beamline X3-B and at Stan-
`ford Synchrotron Radiation Laboratory (SSRL) beamlines 9-3
`and 7-3. Concentrated samples of the Fraz-Grx3/4 heterodimer
`were mixed with glycerol (final concentration 30% v/v glycerol
`and 1—2 mM Fe) and were frozen in a Lucite sample cell covered
`with Kapton tape. Spectra were collected at cryogenic tempera-
`tures using a helium displex cryostat at NSLS (30 K) or with an
`Oxford Instruments continuous flow liquid helium crysostat at
`SSRL (10 K). Beamlines were equipped with double crystal
`monochrornators with Si[l ll] (NSLS) or Si[220] (SSRL) crystals.
`Hannonic rejection mirrors were used. and spectra were collected
`under fully tuned conditions. Canberra solid-state germanium
`detectors (l3 element (NSLS) or 30 element (SSRL)) were used to
`detect the iron Kot fluorescence. Data points were collected every
`5 eV in the proedge region. every 0.25 eV in the edge region.
`and every 005 k in the EXAFS region. Scans were typically
`collected for 45 min, with data acquisition time increasing to 5 s
`
`Li et al.
`
`scans were initiated at fresh locations of the sample surface. At
`this size. six fresh sample spots were monitored per sample.
`Replicate samples were measured with equivalent results.
`Data were processed using the Mac OS 10 version of EX-
`AFSPAK, a suite of data analysis programs available on the
`SSRL Web site. This program was interfaced to Fel‘l‘7.2 to
`generate the theoretical scattering models used in the fits. The
`value of A539 was fixed to a value of -10 eV and the scale factor
`was fixed at 0.9, values that yield the correct crystallographically
`determined bond lengths and coordination numbers for iron
`model complexes Bond length. R. and the Debye-Waller factor
`(oz) were varied freely in each fit. Coordination numbers were
`incremented in fractional steps to refine the optimal number
`based on goodness oftit. Fit quality wasjudged using a modified
`F value, F’, that adjusts for the number ofvariables used in a fit.
`All fits reported here are to unfiltered data over a k range of 1-14
`A". Both single scattering and multiple scattering models were
`used to lit the data. In the latter, parameters for an imidazole ring
`scatterer were used to model histidine ligands. For these fits, the
`Fe-N;,.,;,, distance and o2 values were floating freely. and the
`other atoms in the imidazole ring were linked to the refined value
`F€“Namia-
`
`RESULTS
`
`Fra2 and Grx3/4 Copurify as a Heterodimeric Complex.
`To characterize the interactions between Fra2, Grx3, and Grx4,
`the individual proteins were initially expressed and purified
`separately for in vitro analysis. Soluble recombinant Grx3 and
`Grx4 were easily extracted from E. coli yielding the apo,
`monomeric
`forms upon aerobic purification (Figure
`1,
`Table 1). However. recombinant Fraz was largely found in
`inclusion bodies and partially degraded upon purification
`(Figure l). The molecular mass of the purified proteins (with
`the first Met removed in each case) was confirmed by MALDI-
`TOF mass spectrometry (Table 1). Since Fra2 interacts with Fral
`and Grx3/4 in vivo (8), we tested whether cocxprcssion of the
`interacting proteins improved the solubility of Fra2. Fral, Fra2,
`and Grid were cocxpressed using an E. coli strain cotrans-
`formed with one plasmid expressing Fral and H32 and another
`expressing Grx3. After collecting the induced cells. we immedi-
`ately noted that the cell pellet was a reddish brown color that was
`not observed when the proteins were expressed separately.
`SDS~PAGE analysis showed that Fraz and Grid were both
`expressed at high levels, while the Fral protein was not visibly
`detectable(cxpccted mass = 86.0 kDa with His tag) (Figure 1A).
`Furthermore. Fra2 and Grid were found to copurify as a reddish
`brown-colored complex with a higher apparent molecular mass
`than the individual proteins as determined by analytical gel
`filtration (Figure 1B. Table 1). Similar results were obtained
`upon coexpression of Fra2 with Grx4. The reddish brown Fra2-
`Grx complex was also purified upon expression of Fra2-Grx3
`or Fra2-Grx4 without Fral.
`indicating that Fral was not
`
`Grxii
`+Fra2
`GIX3
`Ft'a2
`3‘ldNllPNlIPNl|P
`
`Biochemistry, Vol. 48, No. 40, 2009
`
`9573
`
`-Fa-S
`-—-Fe-S
`
`619:3
`Fro2
`«G06
`
`6
`
`10
`
`12
`volume (mL)
`
`14
`
`15
`
`FIGURE l: Fra2 and Grx3/4 copurify as a complex. (A) SDS—PAGE
`analysis of purified Fra2 and Grx3. Fra2 and Grx3 wereindividually
`expressed or coexpressed. NI = noninduced cells, I = induced cells,
`and P = purified protein. The 12- 13 kDa band below Fra2 in the
`purified Fra2 and Grx3-Fra2 lanes is a degradation product of Fru2.
`The bands shown with an asterisk in the last lane were identified
`by MALDI-TOF as the E. mli translational elongation factors
`TufA (43.4 kDa) and FusA (77.6 kDa). (B) Gel filtration chromato-
`grams of apo (upper chrornatogram) and Fe-S forms (bottom
`chromntogram) of Fral and Grx3 (0.5 ,ug loaded).
`
`Table 1: Molecular Mass Analysis of Fraz. Grid. and Grx4 Complexes“
`theoretical
`sample
`gel filtration
`MALDI
`
`13971 (monomer)
`13971. l2824"
`19500
`apo Fraz
`28130 (monomer)
`28120
`37000
`npo Grx]
`27493 (monomer)
`27492
`37200
`apo Grx4
`42101 (heterodimer)
`ND"
`52600
`[2Fe-ZS] Grx}-Fra2
`41464 (heterodirner)
`ND
`52300
`[2Fe-ZS] Grx4-FmZ
`56260 (homodimer)
`ND
`71500
`[2F=~2S] Grx3
`54968 (homodimer)
`ND
`65700
`[2Fc-2S] Grx4
`“All masses are shown in Da. "This second peak is attributed to the Fr-a2
`proteolytie fragment shown in Figure 1A. ‘ND, not determined.
`
`Grx3/4 does not involve a covalent,
`bond.
`Fra2-Grx3/4 Birtdsa [2Fe-25]‘, ‘L Cluster. The UV—visible
`absorption spectra of both Fra2-Grx complexes are nearly
`
`intermolecular disulfide
`
`500 600 700
`300 400
`Wavelength (nm)
`
`800
`
`FIGURE 2: UV—visible absorption spectra of as-purified forms of
`Fe-5 cluster-bound Fral-Grx3 (black line) and Fraz-Grx4 (gray
`line). c values are based on Fra2—Grx heterodimer concentrations.
`
`Table 2: Fe. 52’. and GSH Measurements in Purified Fe-S Proteins”
`sample
`Fe
`52'
`GSH
`Fe:S:GSH
`1.7:l:0.2
`0.8:l:0.l
`l:l:0.5
`[2Fe-2S] Fraz-Grx}
`1.7 i 0.2
`1.3 :l: 0.1
`0.6 i0.l
`1
`9:0.4
`[2Fc-2S] Fru2—Grx4
`1.4 :l: 0.1
`[ZI-‘e-2S] hGrx2
`l.9:l:0.l
`2.0i0.l
`1.510.]
`l:l:0.8
`‘Molar values are reported per dimer. Data are the average of three
`independent samples.
`
`and ~0.7 [2Fe~2S]“ clusters per Fra2-Grx4 heterodimer in
`aerobically purified samples. Based on the analytical data and
`the Mossbauer data for [2Fc-2S] Fra2-Grx3. which indicate that
`97% of the Fe is in the form of [2Fc-2S]2+ clusters (see below).
`the UV—visible absorption data indicate on extinction coefficicnl
`at 393 nm = 8.3 :: 1.0 mM"‘ cm" on a per cluster busis for the
`[2Fe-2S]“ center in the Fra2-Grx3/4heterodimer. which is at the
`middle of the range reported for the dominant hand in the
`390-430 nm region from [2Fe-2S]2 ‘ clusters (42). As mentioned
`above, monothio Grxs have been shown to form [2Fc-2S]-
`bridged homodimers (19-21); however. a [2Fe-ZS]-bridged het-
`crodimer involving Grx and another protein has not previously
`been reported.
`To test whether Fe-S binding is required for the Fra2-Grx}f4
`interaction in vitro,
`the purified Fra2-Grx3 hctcrodimer was
`treated with a 20-fold excess of EDTA and DTT anaerobically
`for l h at 4 °C and then loaded onto an analytical gel filtration
`column. The proteins lost
`the reddish brown color but still
`coeluted as a heterodimeric complex. Measurement of Fe and
`acid-labile sulfide in the eluted apo heterodimer confirmed loss of
`the cluster (:03 Fe/heterodimer, 50.5 S2‘/heterodimer), sug-
`gesting that the Fra2-Grx3 interaction is not absolutely metal-
`dependent. This result is consistent with in viva coimmunopreci—
`