Roles and maturation of iron–sulfur proteins in plastids

Roles and maturation of iron–sulfur proteins in plastids One reason why iron is an essential element for most organisms is its presence in prosthetic groups such as hemes or iron– sulfur (Fe–S) clusters, which are notably required for electron transfer reactions. As an organelle with an intense metabolism in plants, chloroplast relies on many Fe–S proteins. This includes those present in the electron transfer chain which will be, in fact, essential for most other metabolic processes occurring in chloroplasts, e.g., carbon fixation, nitrogen and sulfur assimilation, pigment, amino acid, and vitamin biosynthetic pathways to cite only a few examples. The maturation of these Fe–S proteins requires a complex and specific machinery named SUF (sulfur mobilisation). The assembly process can be split in two major steps, (1) the de novo assembly on scaffold proteins which requires ATP, iron and sulfur atoms, electrons, and thus the concerted action of several proteins forming early acting assembly complexes, and (2) the transfer of the pre- formed Fe–S cluster to client proteins using a set of late-acting maturation factors. Similar machineries, having in common these basic principles, are present in the cytosol and in mitochondria. This review focuses on the currently known molecular details concerning the assembly and roles of Fe–S proteins in plastids. Keywords Biogenesis · Iron–sulfur proteins · Plastids · Electron transfer · Photosynthesis Introduction numerous metabolic pathways occurring totally or partially in this organelle are directly or indirectly dependent on the Iron (Fe) and sulfur are critical elements for plant growth functioning of Fe–S proteins. This review is organized in and development. Sulfur is notably required for cysteine three parts, describing how Fe and sulfur species get reduced and methionine synthesis, and is present in a large number and imported in chloroplasts, how the various types of Fe–S of molecules, whereas Fe atoms are associated with many clusters are built from Fe and cysteine and incorporated into proteins as part of hemes, mono- or di-iron non-heme cent- the tenths of client proteins, and finally which chloroplastic ers, or iron–sulfur (Fe–S) clusters. Chloroplasts and plastids pathways/processes are dependent on these cofactors. in general, are highly demanding organelles for both ele- ments due notably to the presence of a translation machinery and of the photosynthetic electron transfer chain. Besides, Supply of iron and sulfur to plastids Iron transport The original version of this article was revised due to a retrospective Open Access order. Chloroplasts, where photosynthesis and heme synthe- Jonathan Przybyla-Toscano and Mélanie Roland have equally sis occur, represent the major subcellular Fe sink in plant contributed to the work. leaves [1]. Photosynthetic organisms uptake Fe from the soil using a sophisticated pumping system that differs between * Nicolas Rouhier nicolas.rouhier@univ-lorraine.fr Poaceae and dicotyledon species, which developed, respec- tively, either a phytosiderophore-dependent chelation-based Université de Lorraine, Interactions Arbres- strategy or a reduction-based strategy (see [2] for an over- Microorganismes, UMR1136, 54500 Vandoeuvre-lès-Nancy, view). In Arabidopsis thaliana, Fe is acquired in several France steps. By extruding protons via the H -ATPase AHA2 and Biochimie et Physiologie Moléculaire des Plantes, coumarins via the PDR9 ABC transporter, A. thaliana can CNRS/INRA/Université Montpellier 2, SupAgro Campus, 34060 Montpellier, France Vol.:(0123456789) 1 3 546 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 3+ solubilize and chelate F e forms by lowering the soil pH. overexpression lines for this gene show abnormal chloro- 3+ 2+ Then, the reduction of Fe to Fe is performed by the fer- plast development and perturbed iron homeostasis and avail- ric reductase-oxidase (FRO) family protein, FRO2, before ability [6, 7]. The loss-of-function mutants are dwarf and its uptake by the plasma membrane Fe transporter named chlorotic (even white), and they grow only heterotrophically. iron-regulated transporter 1 (IRT1) [2]. In the cytosol of root Moreover, they accumulate Fe into ferritins, the function of cells, Fe complexes are formed with organic acids (malate which is normally to protect this organelle from oxidative or citrate) or nicotianamine before being translocated to stress by sequestering Fe. The PIC1-overexpressing plants the shoots and unloaded in the cytosol of mesophyll cells suffer from oxidative stress and leaf chlorosis likely due to [3]. After this step, little is known concerning Fe acquisi- a Fe overload in chloroplasts. Although this permease is tion by chloroplasts, its subsequent storage, and delivery mentioned to be part of the translocase of the outer/inner to dedicated proteins and machineries. It is possible that chloroplast membrane (Tic–Toc) complex in other studies, 3+ a voltage-dependent transport system allows F e -citrate a Fe transport function is clear from the complementation complexes to pass the outer membrane of the plastid enve- of a yeast fer3fer4 mutant which is defective in Fe uptake, lope [4]. Once in the chloroplastic intermembrane space, leading to the conclusion that PIC1 may have a dual function 3+ 2+ FRO7 may reduce ferric (F e ) to ferrous iron (F e ) via [6]. Another putative Fe transporter, named NAP14 (non- its reductase activity [5]. Several transporters located in the intrinsic ABC protein 14), was identified from its homology inner membrane of the chloroplast envelope are candidates with the ABC transporter FutC belonging to the FutABC for Fe import into the stroma (Fig. 1). The first one is named iron uptake system in cyanobacteria [8]. As observed for permease in chloroplast 1 (PIC1) [6]. Both knock-out and pic1, a nap14 knock-out mutant accumulates Fe in shoots, PSI Target Apoproteins SIR FTR HCF101 SUFA1 NFU1 IBA57.2 NFU3 NFU2 3+ BOLA4 Fe Heme and GRXS14 GRXS16 2+ Fe FRO7 siroheme SUFB SUFD BOLA1 3+ biosynthesis Fe SUFC SUFC ATP ADP+FAD ADP FADH Iron-related proteins FH SUFB SUFD Cysteine desulfurase SUFC SUFC ATP SufE-like proteins Scaffold proteins alanine + SUFE1-3 Transfer proteins cysteine Target proteins NFS2 SUFE1-3 Fig. 1 Working model for iron uptake and maturation of Fe–S pro- associated with each protein function is indicated directly on the fig- teins by the SUF machinery in plastids of eukaryotic photosynthetic ure. The detailed description of the maturation process and the con- organisms. Besides the putative Fe transporters located at the mem- nections between the SUF proteins are described in the text. Except brane of the chloroplast envelope, which would serve for providing NFU2, NFU3, and HFC101, all maturation factors have been grouped the required Fe atoms to the SUF machinery, this scheme integrates in a blue circle in the absence of information concerning their precise the 17 putative SUF components. In the absence of stronger evidence function, but two-way arrows indicate that physical interactions have concerning the implication of frataxin, it is not integrated among SUF been observed between some proteins components and is represented by a dashed circle. The color code 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 547 exhibits abnormal chloroplast structures, and shows deregu- (SUF) machinery is autonomous, the cytosolic iron–sulfur lated levels of Fe homeostasis-related genes. However, in the assembly (CIA) machinery is dependent on the mitochon- absence of other Fut orthologs in A. thaliana, the question of drial iron–sulfur cluster (ISC) machinery as it relies on a whether NAP14 can work alone or in pair with other uniden- sulfur-containing compound generated in the first steps and tified partners remains open. A third candidate transporter exported from mitochondria by an ABC transporter. We for Fe uptake in chloroplasts is mitoferrin-like1 (MFL1) [9]. invite the readers interested in the ISC and CIA machineries However, although its gene expression is dependent on Fe in plants to refer to the following recent reviews [12, 13]. For supply and the protein is in principle located to the inner all these machineries and in particular the chloroplastic SUF membrane of the chloroplast envelope, the growth of knock- machinery, the biosynthesis and delivery of Fe–S clusters out mutants is only moderately affected. While all these pro - can be separated in two major steps: their de novo assembly teins seem to be involved in Fe homeostasis in chloroplasts, on scaffold proteins and their incorporation into final client further characterization is urgently needed to clarify their proteins. This second step may necessitate the exchange and exact function and respective importance. possibly conversion of Fe–S clusters between scaffold pro- teins and maturation factors including Fe–S cluster transfer Sulfate import and reduction in plastids proteins and targeting/recruiting factors. Repair mechanisms for the synthesis of cysteine, the sulfur donor may eventually account for the recycling of damaged Fe–S of Fe–S clusters clusters, which could be important in chloroplasts consider- ing the presence of reactive oxygen and nitrogen species, but Photosynthetic organisms use sulfate present in the soils as this will not be discussed further as information in plants is a primary source of sulfur. Sulfate is incorporated into the very scarce. roots through an active proton/sulfate co-transport system located at the plasma membrane [10]. Once in the xylem, The de novo Fe–S cluster assembly on scaffold sulfate is transported to the shoots, unloaded into the cytosol protein of mesophyll cells, and then transported in the chloroplasts for its ATP-dependent reductive assimilation into sulfide In chloroplasts, it seems now clear that the sole scaffold (see [10] for review). The involved transporters all along system is formed by the SUFBCD proteins (Fig. 1) [14]. these steps belong to the sulfate transporter (SULTR) family, The assembly of a Fe–S cluster on this scaffold complex which is composed of 12 members in A. thaliana that can be theoretically requires the concerted action of several proteins grouped into four classes. The SULTR3 class comprises the as it requires the polypeptide backbones, ATP, Fe, and sul- chloroplast-localized sulfate transporters [11]. The sulfide fur atoms and electrons. There are still many uncertainties generated by the ferredoxin (FDX)-dependent sulfite reduc- about the involved actors in plants and the molecular details. tase (SIR) will be used for cysteine biosynthesis by cysteine Thus, we will often make analogies to the Escherichia coli synthase, a complex of two enzymes, serine acetyltrans- SUF system, which has been better characterized. The best, ferase (SAT) that uses acetyl-coA to form O-acetylserine not to say the only, well-characterized actors in plants of (OAS) from serine and O-acetylserine-(thiol)-lyase (OAS- this assembly complex are proteins required for the produc- TL) which can substitute the acetyl moiety by sulfide to form tion and transfer of the required sulfur. The NFS2 protein cysteine. While the first steps of sulfate reduction into sulfide (formerly referred to as CpNifS) is a pyridoxal-l -phosphate are clearly restricted to the chloroplasts, cysteine synthesis (PLP)-dependent cysteine desulfurase, which catalyzes the can also occur in the cytosol and in mitochondria owing to extraction of the sulfur atoms from cysteine, producing a the ubiquitous expression of SAT and OAS-TL and exchange persulfide group on a catalytic cysteine with the concomitant of sulfide across organelle membranes [10]. release of an alanine (Fig. 1) [15]. As a class II cysteine des- ulfurase, similar to the bacterial SufS orthologs, the acces- sibility of the persulfide group is limited by the presence The biogenesis of Fe–S proteins of a β-hairpin near the catalytic cysteine [16, 17]. For this in chloroplasts by the SUF machinery reason, the transfer of sulfur atoms to the scaffold complex relies on an additional protein named SUFE. In A. thaliana, Several dozen of proteins containing Fe–S clusters are found there are three SUFE proteins (SUFE1-3) targeted to chlo- in various subcellular compartments in the model plant A. roplasts [18]. In addition to the SUFE domain, SUFE1 has a thaliana as in other plants. Accordingly, in plant cells, three C-terminal BOLA domain, the role of which is unknown but assembly machineries exist in plastids, in mitochondria, may preg fi ure a control by glutaredoxins (GRXs, see below) and in the cytosol, the latter being dedicated to the mat- and SUFE3 possesses a quinolinate synthase (NadA) domain uration of Fe–S proteins found both in the cytosol and in at the C-terminus, which is involved in NAD biosynthesis the nucleus. Whereas the chloroplastic sulfur mobilisation [18, 19]. As shown for the corresponding E. coli couple 1 3 548 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 [20], each SUFE protein enhances the cysteine desulfurase it is very likely that the SufBC D scaffold binds a [Fe S ] 2 4 4 activity of NFS2 by accepting the persulfide group on its cluster in vivo and considering the conservation between own catalytic cysteine, thus serving as a relay to the scaffold A. thaliana and E. coli sequences, we anticipate that this system [18, 19]. At the structural level, A. thaliana NFS2 is mechanism should also prevail for plant proteins. However, a dimeric protein with two distant active sites, which sug- we cannot completely rule out that SufBC D or other forms, gests that the functional NFS2-SUFE unit should be a het- such as the SufB C form detected with E. coli proteins 2 2 erotetramer [17]. In addition to the existence of additional [29], can bind other cluster types in some conditions. For domains in SUFE1 and SUFE3, the existence of three SUFE instance, transcriptomic data indicate that the SUFB, SUFC, isoforms may be also linked to their expression pattern as and SUFD genes may not be co-expressed in all organs and for instance SUFE2 is mostly expressed in flowers [18]. The cell types of A. thaliana. central role of these proteins has been validated by genetic At this stage of the assembly process, there are many studies, since the study of knock-out A. thaliana lines proved other crucial questions concerning the source of electrons 3+ 2+ that NFS2, SUFE1, SUFE3, and SUFBCD genes are essen- required for the reduction of Fe to Fe or of the persulfide 0 2− tial [14, 18, 21, 22]. The use of RNAi lines showed that (S ) to a sulfide (S ), the source of Fe, and the control of NFS2 and SUFBCD are required for the maturation of all its entry in the complex. In this respect, it is important to plastidial Fe–S proteins tested so far [14, 22]. note that the SufBC D complex was purified with a bound In E. coli as in A. thaliana, the scaffold complex is prob- reduced flavin-adenine dinucleotide (FADH ) molecule [29, ably composed by three subunits, SUFB, SUFC, and SUFD, 30]. While SufB alone can bind the flavin in vitro [30], SufD very likely in a 1:2:1 stoichiometry and will be referred to is also required in vivo [29]. It is currently believed that as SUFBC D (Fig. 1) [14, 20]. It seems that NFS2, SUFEs, this FADH provides the necessary reducing equivalents for 2 2 and SUFBCD do not form a large and stable complex as the reduction of ferric iron. Since FAD is released from the recently shown in the case of the mitochondrial ISC system complex upon oxidation, an external regeneration system in yeast and human [23, 24]. Indeed, some in vitro biochemi- is needed, which could be possibly an FDX or an NADPH- cal analyses using the bacterial SufS, SufE, and SufBCD dependent flavin reductase. enzymes indicated that SufS does not seem to make stable The mechanisms and actors involved in the delivery of interactions with SufBCD, unlike SufE whose presence is Fe for Fe–S cluster biosynthesis in plastids are completely absolutely required for an efficient Fe–S cluster reconsti- unknown. The Fe–Storage proteins, ferritins, have been tution in vitro on SufBCD [20, 25]. Besides, it has been excluded from Fe donor candidates, because an Arabidopsis shown that the presence of SufC, but not SufD, is required mutant (fer1-3-4) for the three ferritins found in leaves has for the transfer of the sulfur atoms bound to E. coli SufE to no apparent phenotype [31], while mutant plants modified SufB. Upon ATP binding, the SufC ATPase would induce for the expression of these early biogenesis factors are either structural changes on SufB and SufD that are necessary for lethal or at least strongly affected. Another candidate for Fe Fe–S cluster binding [26]. Some residues important for these delivery is a small acidic protein with iron-binding proper- interactions have been identified from the 3D structures and ties named frataxin. In the mitochondrial ISC machinery, validated by mutagenesis [26, 27]. Among the numerous frataxin controls iron entry in the assembly complex by acti- cysteines present in SufB, the primary sulfur acceptor would vating sulfide formation by the cysteine desulfurase [32, 33]. be the conserved Cys254 (E. coli numbering). This sulfur In this complex, frataxin can interact both with the cysteine atom would then be transferred to Cys405, one of the Fe–S desulfurase and the ISCU scaffold protein. Except for a few cluster ligands owing to the existence of a tunnel inside the organisms like Z. mays, there is usually a single gene cod- β-helix core domain of SufB [27]. The question of which ing for frataxin (FH) in plants. While frataxin was believed type of Fe–S clusters is bound to this complex has been for a long time to be exclusively located in mitochondria, it investigated in detail. It was shown that E. coli SufB alone was recently reported that A. thaliana FH (AtFH) and two can assemble both [Fe S ] and [Fe S ] clusters in vitro and isoforms from Z. mays may have a dual targeting into both 2 2 4 4 that a conversion from the [Fe S ]-loaded SufB form to a mitochondria and plastids [34, 35]. According to this pos- 4 4 stable [Fe S ]-loaded form is possible upon exposure to air sible chloroplastic localization, Arabidopsis FH-deficient 2 2 [25, 28]. However, based on the structure of an apoSufBCD plants show a decrease in the heme content [36]. Moreover, complex, it was proposed that a histidine of SufD may be they present a decrease in the total chlorophyll content, in a Fe–S cluster ligand [26]. Consistently, a mutated variant the levels of two plastidial FDXs and in nitrite reductase for this histidine lost the ability to assemble a Fe–S clus- (NIR, a siroheme-containing enzyme) activity which could ter in vivo, and both SufC and SufD were required for the explain the observed changes in the rate of the photosyn- in vivo maturation of SufB [29]. In this cellular context, E. thetic electron transport chain [35]. The impact on heme coli SufBC D complex mostly binds a [F e S ] cluster with content would be in good agreement with the described 2 4 4 some residual amount of linear [F e S ] clusters [29]. Hence, interaction between yeast frataxin and ferrochelatase, the 3 4 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 549 terminal enzyme of heme synthesis performing porphyrin plants overexpressing GRXS14 have a decreased chlorophyll metalation [37]. All these observations suggest an impair- content [44]. Considering that several enzymes involved in ment of the plastidial Fe–S cluster biosynthesis and/or of the chlorophyll catabolism require Fe–S clusters, this may con- heme or siroheme biosynthesis, although stronger and more stitute a first hint towards a role of GRXS14 in the matu- direct biochemical evidence is still required. ration of specific client proteins in this pathway. Counter - intuitive to this first observation, plants lacking GRXS14 Delivery and trafficking of preformed Fe–S clusters showed accelerated chlorophyll loss compared to wild-type by maturation factors plants when exposed to prolonged darkness, suggesting more complex connections [44]. A redundancy may exist between The preformed Fe–S cluster on the SUFBC D complex, be both plastidial GRXs, since a double mutant with about 20% it a [Fe S ] or a [Fe S ] cluster, has then to be correctly GRXS16 remaining exhibits a 20% biomass reduction in 2 2 4 4 targeted to client apoproteins. This requires several other standard conditions compared to wild-type plants. How- proteins referred to as maturation factors. Among these, one ever, this phenotype is not exacerbated under stress condi- could differentiate the so-called Fe–S cluster transfer/carrier tions. Overall, unlike the knock-out mutant of mitochondrial proteins (belonging to NFU, SUFA, GRX, and HCF101 fam- GRXS15 which is embryo-lethal [45], these results point ilies) from targeting factors (belonging to BOLA and IBA57 either to non-essential roles of these isoforms or to a redun- families) which, contrary to the proteins of the first group, dant function with the remaining GRXS16 level being suf- are not able to bind Fe–S clusters by themselves, although ficient to sustain an essential role similar to GRXS15. BOLAs do it in complex with GRXs [38, 39]. It is interest- Concerning BOLA proteins, several roles have been pro- ing to note that all proteins of these families have mitochon- posed, but only those connected to their participation in Fe drial counterparts in the ISC machinery, whereas the compo- metabolism have been really validated [46]. Their involve- nents forming the eukaryote-specific CIA machinery usually ment in Fe–S cluster biogenesis was demonstrated from belong to different protein families [ 12]. This analogy to the study of bol1/3 mutant in yeast and of human patients the mitochondrial system is the reason why some of these defective for the mitochondrial BOLA3. Both types of cells plastidial members, whose role in the maturation of Fe–S display protein lipoylation defects due to the incorrect matu- proteins in plastid has not been yet established, have been ration of lipoate synthase and a decrease in activity for some included in this section. The current model for these steps in other [Fe S ] proteins as aconitase and succinate dehydro- 4 4 the plant mitochondrial ISC machinery derives mainly from genase [47–49], whereas human patients also have defects studies conducted in yeast and human and can be summa- in the mitochondrial respiratory complexes I and III [49]. rized as follows [40]. A glutaredoxin (GRXS15 in plants) is Three isoforms with a BOLA domain are found in plant the primary transfer protein receiving a [Fe S ] cluster from chloroplasts. As already mentioned, the C-terminal region 2 2 ISCU proteins. This cluster can be either directly inserted of SUFE1 contains a BOLA domain. The two other iso- into [Fe S ]-recipient apoproteins or used to build [Fe S ] forms, BOLA1 and BOLA4, comprise a single domain. Both 2 2 4 4 clusters on a heterocomplex formed by ISCAs and possibly BOLA4 and SUFE1 could also be targeted to mitochondria IBA57. Some mechanistic and structural aspects of the clus- [21, 50]. Interactions between these plastidial BOLA pro- ter conversion from the [Fe S ]-loaded GLRX5 form to the teins and GRXS14 and GRXS16 have been demonstrated 2 2 [Fe S ]-loaded ISCA1-2 form have been recently delineated both in vitro and in planta [38, 50]. These proteins can in fact 4 4 using human proteins [41, 42]. Then, the insertion of the form both apo-heterodimers and holo-heterodimers bridging [Fe S ] clusters into client Fe–S proteins might be direct or a [Fe S ] cluster [39], as also demonstrated for bacterial, 4 4 2 2 facilitated by NFU and BOLA proteins that likely act in con- yeast, and mammalian isoforms [46]. In this respect, it is cert for the maturation of specific targets notably the lipoate interesting to note that adding BOLA to a GRX homodi- synthase or by IND1/INDH, a close HCF101 homolog, mer converts it to a more stable holo GRX-BOLA heter- which seems specific for the respiratory chain complex I. odimer. This interconversion might represent a regulatory The current genetic and biochemical evidence indicate mechanism either to shut down or activate some specific that this sequence of events should be very different for the pathways by favouring one target over another. At the struc- plastidial SUF machinery (Fig. 1). Although the two plastid- tural level, all BOLA isoforms have a similar well-conserved ial isoforms, named GRXS14 and GRXS16, have the ability fold [39]. Two subgroups can, however, be distinguished to bind the regular [F e S ] cluster in homodimer (or in heter- based on the length of the β1–β2 loop referred to as the 2 2 odimer with BOLA, see below) and to complement a yeast variable [C/H] loop, because it contains one of the ligands mutant for the mitochondrial Grx5 [43], strong genetic and provided by BOLA either a cysteine or a histidine, the sec- physiological evidence for a similar involvement in plants is ond ligand being a totally conserved histidine found in the still missing. Single mutants for each of these genes have no α3–β3 loop [39, 48]. Other cysteine ligands are provided phenotype when grown under standard conditions, whereas by a glutathione molecule and by the one present in the 1 3 550 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 conserved CGFS signature of the GRX partner, as in regular two orthologs found in A. thaliana, IBA57.1 and IBA57.2, GRX homodimers [51]. While there is no true ortholog of are, respectively, localized in mitochondria and plastids [61]. yeast Bol3 in plants, the observation that Bol3 might interact It is interesting to note that both isoforms can complement with Nfu1 rather than with Grx5 in yeast could point to a the growth defects of an E. coli ygfZ mutant observed on a different role in the late steps of the mitochondrial system minimal medium or upon oxidative stress [60]. This is the [47, 48]. Although single bol1 and bol3 mutants do not have only physiological information obtained so far for these plant phenotypes and the respective molecular roles of Bol1 and isoforms, since an Arabidopsis iba57.1 mutant is embryo- Bol3 are still unclear, a connection between Bol3 and Nfu1 lethal and an iba57.2 mutant has not been described. While is also evident from the quite similar phenotype of the bol1/3 the exact function of IBA57 is still unknown, it is important and nfu1 mutant cells [47]. to note that there is a conserved cysteine residue in a KGCY- In mitochondria, ISCA proteins are central for the matu- x-GQE-x3-R/K motif, which is almost the only conserved ration of [F e S ] proteins, presumably ensuring the con- motif in this protein family [62]. Moreover, consistent with 4 4 version of [Fe S ] centers into [Fe S ] centers. In bacteria, the structural similarity of IBA57 with folate-dependent 2 2 4 4 the different A-type isoforms (IscA, SufA, ErpA) are also enzymes [63], E. coli YgfZ can bind tetrahydrofolate [60]. required for the maturation of [F e S ] proteins, even though Another category of proteins strictly required for the mat- 4 4 in vitro studies demonstrated that Azotobacter vinelandii uration of [Fe S ] clusters is the NFU family that exists in 4 4 IscA, for example, can reversibly cycle between [Fe S ] all kingdoms. In mitochondria, the study of the yeast mutant 2 2 and [Fe S ] forms through electron reductive coupling or and several human patients indicates that NFU1 is required 4 4 oxidative cleavage [52]. Some biochemical redundancy for the maturation of lipoate synthase, which affects several seems to exist between them as demonstrated for the Fe–S ketoacid dehydrogenases dependent on lipoic acid, and for cluster assembly of IspG and IspH, two enzymes involved the maturation of complexes I, II or III depending on the in isoprenoid synthesis and also present in plant chloroplasts patients [49, 64]. A. thaliana encodes five NFU isoforms, [53, 54]. In plastids, the only representative of this family two (NFU4 and NFU5) should be targeted to mitochondria, should be SUFA1, also referred previously to as CpISCA and three (NFU1, NFU2, and NFU3) are localized in chloro- and ISCA-I [55, 56]. As an A-type carrier protein, SUFA1 plasts [65, 66]. All these proteins share an NFU domain pos- possesses the three characteristic conserved cysteines [54, sessing a CXXC motif necessary for the binding of a [F e S ] 4 4 55] that allow the binding of a [F e S ] center in a dimer as in a dimer [67]. Chloroplastic isoforms have an additional 2 2 observed upon in vitro Fe–S cluster reconstitution assays NFU domain in the C-terminal extremity which does not [55–57]. According to the ISC model, Fe–S cluster transfer have the cysteine residues, whereas mitochondrial isoforms experiments showed that GRXS14 can efficiently and uni - have an additional N-terminal domain of unknown function directionally transfer its [Fe S ] cluster to SUFA1; however, (Fig. 2) [65, 68]. Loss-of-function nfu2 and nfu3 mutants 2 2 there was no sign of a [F e S ] cluster formation [57]. Using have a dwarf phenotype with pale green leaves [69, 70]. 4 4 recombinant proteins, it was shown in vitro that an apo-SufA By coupling chlorophyll fluorescence and P700 absorption from E. coli could promote the maturation of an apo-FDX measurements to western blot analyses, it was shown that from a [F e S]-loaded SufBC D scaffold, indicating that this phenotype is due to the impairment of photosystem I 4 4 2 SUFA proteins would directly interact with the scaffold but (PSI) architecture and activity which is explained by a defect also that it facilitates Fe–S cluster conversion. Nevertheless, in the maturation of the three [F e S ] clusters assembled in 4 4 knock-out mutants have no visible phenotype when grown the psaA, psaB, and psaC subunits. The only other nota- under standard conditions, indicating that the role of SUFA1 ble and robust molecular default observed is that the SIR is dispensable [55, 56]. Whether it is involved in the matu- level and activity are decreased in nfu2 [14, 65, 70, 71]. The ration of [F e S ] proteins, [Fe S ] proteins or both remains fact that a double nfu2-nfu3 mutant is lethal [69] indicates 2 2 4 4 thus to be determined. that both NFU isoforms should have partially overlapping It is getting clear that, in yeast and human mitochondria, functions. This raises also the question of their contribu- ISCA proteins interact with IBA57 (Iron–Sulphur cluster tion relatively to the high chlorophyll fluorescence 101 assembly factor for Biotin synthase- and Aconitase-like (HCF101) protein, a plastidial 51 kDa protein belonging to mitochondrial proteins with a mass of 57 kDa). They form the NTPase protein family (Fig. 2). The hcf101 Arabidop- a complex involved in the maturation of several [F e S ] pro- sis mutant plants have globally similar molecular defects, 4 4 teins including radical-S-adenosylmethionine (SAM) pro- although this is exacerbated as the strongest allele is lethal teins, homoaconitase, aconitase, biotin synthase, and lipoic at the seedling stage and the decrease in the amounts of PSI acid synthase [58, 59]. Depletion of the E. coli ortholog subunits is stronger, almost complete [72–74]. Besides, there YgfZ also affects some [Fe S ] proteins such as succinate is a decrease in the ferredoxin-thioredoxin reductase (FTR) 4 4 dehydrogenase, fumarase, dimethylsulfoxide reductase, and levels, another [F e S ] protein [72]. Overall, in accordance 4 4 MiaB, an enzyme involved in tRNA thiolation [58–60]. The with the capacity of Arabidopsis NFU2 and HCF101 to bind 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 551 NFS2 463 AA SUFB 557 AA SUFC 338 AA SUFD 475 AA 180 AA SUFA1 CXC IBA57.2 432 AA KGCY-X-GQE HCF101 532 AA DUF59 DUF971 NFU1 NFU2 231/235/236 AA CXXC NFU3 Domain of unknown funcƒon SUFE1 371 AA Monothiol glutaredoxin domain SUFE2 258 AA Endonuclease domain SUFE3 718 AA SUFE domain BOLA domain BOLA1 160/177 AA NFU domain BOLA4 Degenerated NFU domain GRXS14 173 AA CGFS P-loop NTPase moƒf GRXS16 293 AA YgfZ signature CGFS Fig. 2 Protein domain organization of SUF components. The with the length in amino acids of the Arabidopsis proteins indicated. domains (identified using pfam or the NCBI conserved domain tools) The Fe–S binding cysteine and histidine residues are represented in present in SUF components have been represented using the color yellow and black, respectively, while other conserved cysteines are in code defined on the figure. Except for the chloroplastic targeting orange, although their function is sometimes unclear if any sequence (light green boxes), the domains are represented at scale, [Fe S ] cluster in vitro [67, 73], this indicates that all these and two histidines, is found in the Rieske protein of the 4 4 proteins are required for the maturation of [F e S ] proteins, cytochrome b f complex. In the genome of eukaryotes and 4 4 6 particularly PSI subunits, and that HCF101 would act down- in some cyanobacteria, the Rieske protein is encoded by stream of NFU2 and NFU3 (Fig. 1). a single gene named photosynthetic electron transfer C In summary, there are currently ten putative maturation (petC), whereas in most cyanobacteria, there are additional factors in the SUF machinery for several dozens of plastidial isoforms whose physiological function is still uncertain client proteins. The role of some of these maturation factors [75]. The absence of the petC proteins is lethal in the early still awaits conr fi mation not to speak about their connections developmental stages both in A. thaliana and cyanobacte- and hierarchical organization. There is also an urgent need ria (Table 1) [75, 76]. Three low potential [Fe S ] clusters 4 4 to learn more about how specificity towards target proteins are attached to the thylakoid membrane but face the reduc- is achieved and about the molecular and structural aspects ing, stromal side of PSI, and function in series. The first, of these interactions. referred to as F , is associated with a PsaA–PsaB heter- odimer via cysteine residues, while the two others, named F and F , are bound to PsaC [77]. These clusters transfer A B Functional diversity among client Fe–S electrons to FDXs, small soluble proteins, which contain proteins in plastids a classical [Fe S ] cluster, e.g., rhombic cluster ligated by 2 2 four cysteines. The nuclear genome of algae and plants har- Fe–S clusters in the functioning and protection bours a variable number of FDX homologs, differentially of the photosynthetic electron transport chain expressed in plant organs or at different development stages or in response to different stimuli [78]. The A. thaliana Among other functions, Fe–S clusters have a crucial role genome contains four genes encoding four well-described in electron transfer reactions, and thus, several Fe–S pro- plastidial FDXs (Fd1 to Fd4) and at least two additional teins are found in the thylakoid membrane as part of the genes, referred to as FdC1 and FdC2, encoding proteins photosynthetic electron transport chain. A Rieske-type bearing C-terminal extensions and whose functions remain Fe–S cluster, i.e., a [Fe S ] cluster ligated by two cysteines elusive [79, 80]. In specific physiological situations such as 2 2 1 3 552 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 1 3 Table 1 Phenotypes of A. thaliana mutant lines for plastidial Fe–S proteins Short Name AGI number Cluster type Type of mutants Mutant phenotypes References DHAD At3g23940 [Fe S ] Knock-out Embryo-lethal Zhang et al. [149] 2 2 Knock-down Shorter root, hypersensitive to salt stress IPMI (LSU1) At4g13430 [Fe S ] Knock-down Pleiotropic growth abnormalities Sureshkumar et al. [152], Knill et al. [151] 4 4 DWARF27.1 At1g03055 [Fe S ] Knock-down Increase in axillary rosette branches Waters et al. [168] 4 4 DWARF27.2 At1g64680 [Fe S ] Not yet described 4 4 DWARF27.3 At4g01995 [Fe S ] Not yet described 4 4 ISPG At5g60600 [Fe S ] Knock-out Albino phenotype, proplastid growth and thylakoid Gutiérrez-Nava et al. [163] 4 4 membrane formation affected ISPH At4g34350 [Fe S ] Knock-out Albino phenotype, proplastid growth and thylakoid Gutiérrez-Nava et al. [163], Hsieh and Hsieh [165], 4 4 membrane formation affected Guevara-García et al. [164] THIC At2g296302x [Fe S ] Knock-down Lethal (development arrested at the cotyledon stage Raschke et al. [140], Kong et al. [142] 4 4 with chlorotic phenotype) NIR At2g15620 [Fe S ], siroheme X-ray mutagenesis Lethal in barley unless a nitrogen source is provided Duncanson et al. [120] 4 4 SIR At5g04590 [Fe S ], siroheme Knock-out Lethal Khan et al. [119] 4 4 Knock-down Early seedling lethal Khan et al. [119] SIRB At1g50170 [Fe S ] Knock-out Seedling lethal (post-germination arrest) Saha et al. [126] 2 2 ATase1 At2g16570 [Fe S ] Knock-out No phenotype Hung et al. [144] 4 4 ATase2 At4g34740 [Fe S ] X-ray mutagenesis Small and albino/pale reticulated leaves, cell division Kinsman and Pyke [171], Hung et al. [144], van den 4 4 affected Graaf et al. [146], Rosar et al. [172] ATase3 At4g38880 [Fe S ] Not yet described 4 4 APR1 At4g04610 [Fe S ] Not yet described 4 4 APR2 At1g62180 [Fe S ] Knock-out None but increased sensitivity to selenate tolerance Grant et al. [173] 4 4 APR3 At4g21990 [Fe S ] Not yet described 4 4 cLIP1 At5g084152x [Fe S ] Not yet described 4 4 GLT1 At5g53460 [Fe S ] Knock-out No phenotype but decreased chlorophyll content, Lancien et al. [123] 3 4 growth defect under low CO GLU1 At5g04140 [Fe S ] Knock-down Dwarf photorespiratory phenotype Somerville and Ogren [125], Coschigano et al. [122], 3 4 Lancien et al. [123] GLU2 At2g41220 [Fe S ] Knock-down No phenotype Potel et al. [124] 3 4 DjC17 At5g23240 [Fe S ] Knock-out Defective root hairs Petti et al. [91] 4 4 DjC18 At2g42750 [Fe S ] Not yet described 4 4 ndhI AtCg010902x [Fe S ] Not yet described 4 4 ndhK AtCg00430 [Fe S ] Not yet described 4 4 petC At4g03280rieske [Fe S ] Knock-out Seedling lethal Maiwald et al. [76] 2 2 psaA AtCg00350 [Fe S ] with psaB Not yet described 4 4 psaB AtCg00340 [Fe S ] with psaA Not yet described 4 4 psaC AtCg010602x [Fe S ] Not yet described 4 4 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 553 1 3 Table 1 (continued) Short Name AGI number Cluster type Type of mutants Mutant phenotypes References TIC55 At2g24820rieske [Fe S ] Knock-out No phenotype Boij et al. [106], Hauenstein et al. [104] 2 2 Knock-down No phenotype Tanaka et al. [174] PAO (ACD1) At3g44880rieske [Fe S ] Knock-out Age- and light-dependent cell death phenotype in Pružinská et al. [114] 2 2 leaves and flowers. Stay-green phenotype in the dark rieske [Fe S ] Knock-down Light-dependent lesion mimic phenotype, increased Greenberg and Ausubel [112], Yang et al. [175] 2 2 sensitivity to biotic and mechanic stresses PTC52 (ACD1-like) At4g25650rieske [Fe S ] Knock-out No phenotype Boij et al. [106] 2 2 CMO At4g29890rieske [Fe S ] Not yet described 2 2 CAO At1g44446rieske [Fe S ] X-ray mutagenesis Pale green phenotype with no Chl b, highly photo- Espineda et al. [103], Ramel et al. [107] 2 2 sensitive HCAR At1g046202x [Fe S ] Knock-out No phenotype, stay-green mutant upon dark exposure Meguro et al. [108] 4 4 NEET At5g51720 Neet-[Fe S ] Knock-down Late bolting, early senescence Nechushtai et al. [170] 2 2 SUFE3 At5g50210 [Fe S ] Knock-out Lethal Murthy et al. [18] 4 4 Fd1 At1g10960 [Fe S ] Knock-down Enhanced linear electron flow Hanke and Hase [72] 2 2 Fd2 At1g60950 [Fe S ] Knock-out Growth arrest and inactivation of photosynthesis Voss et al. [176] 2 2 Knock-down Lower biomass accumulation and retarded linear Hanke and Hase [78] electron flow Fd3 At2g27510 [Fe S ] Knock-down Photoinhibition, with a reduction in maximum PSII Hanke and Hase [78] 2 2 yield following dark adaptation Fd4 At5g10000 [Fe S ] Not yet described 2 2 FdC1 At1g32550 [Fe S ] Not yet described 2 2 FdC2 At4g14890 [Fe S ] EMS mutagenesis Yellow–green leaf phenotype in rice Li et al. [177]; Zhao et al. [178] 2 2 FTR At2g04700 [Fe S ] Knock-out Lethal Wang et al. [97] 4 4 Virus-induced silencing Chlorosis, abnormal chloroplast development Wang et al. [97] All described phenotypes come from studies performed with A. thaliana unless otherwise stated in the “mutant phenotype” column 554 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 environmental constraints, FDXs can recycle electrons to Based on the fact that HSP70B plays a role in the repair and the plastoquinone pool, contributing to the so-called cyclic protection of PSII (Photosystem II) from photoinhibition electron flow [81]. The major cyclic pathway is dependent [89], and together with the CDJ2 paralog in the biogenesis/ on the PGR5 (proton gradient regulation 5)/PGRL1 (PGR5- maintenance of thylakoid membranes [90], we could specu- like photosynthetic phenotype 1) proteins [82]. The other late that CDJ3-5 may have a similar role. However, this has involves the NAD(P)H dehydrogenase (NDH) complex. In not been addressed so far for Arabidopsis orthologs, DjC17 higher plants, it forms a large complex associated with PSI, and DjC18. There is no biochemical information on these which is composed of 11 plastid-encoded subunits, some proteins and genetic evidence has been obtained only for additional nuclear-encoded subunits, and auxiliary factors DjC17, the mutation of which results in an altered root hair [83]. Among these, the NDH-I and NDH-K subunits bind development and reduced hair length due to aberrant cortical two and one [F e S ] clusters, respectively [84]. While Arabi- cell division [91]. 4 4 dopsis knock-out mutants for NDH-I and NDH-K genes have not been characterized, tobacco knock-out mutants of ndh A multitude of ferredoxin‑dependent Fe–S proteins genes usually have no phenotype under standard conditions and pathways but are sensitive to environmental stresses [82]. In eukaryotic microalgae and cyanobacteria, an additional The formation of reducing equivalents pathway directly coupled to the photosynthetic electron transport chain and involving hydrogenases allows the pho- FDXs are soluble proteins positioned at a metabolic cross- toproduction of ATP at the expense of reductant synthesis road, controlling the electron flow necessary for CO fixa - in specific conditions such as the response to anaerobiosis tion, nitrogen, and sulfur assimilation but also chlorophyll or anoxia. Chlamydomonas reinhardtii contains two [FeFe]- metabolism to cite a few examples (Fig. 3). Their primary hydrogenases, namely HYDA1 and HYDA2, which will role is to transfer electrons to various acceptors in the produce molecular hydrogen H from protons by accepting stroma, in the thylakoids, and in the inner membrane, includ- electrons from FDXs. These HYDA contain a complex Fe–S ing a large variety of Fe–S proteins but also proteins contain- cluster at their active sites, the H-cluster that is essential for ing heme and non-heme iron centers and flavoproteins [92, catalytic activity [85]. It consists of a classic [F e S ] clus- 93]. Among the latter category, ferredoxin-NADP reductase 4 4 ter linked to a complex 2Fe sub-cluster [86]. Whereas the (FNR) may be the most important one, since it will drive [Fe S ] cluster is assembled by the regular SUF machinery, most of these electrons for the regeneration of NAPDH, 4 4 the sub-cluster requires specific maturation proteins, HYDE, which will then supply in particular the Calvin–Benson HYDF, and HYDG, for this assembly. The HYDE and cycle. It is worth nothing that a significant fraction of FNR HYDG gene products are radical-SAM enzymes, whereas is bound to the thylakoid membrane and it could partici- HYDF is a P-loop NTPase protein constituting a scaffold pate to the cyclic electron flow via the PGR5-dependent assembly platform. These proteins incorporate themselves pathway by interacting with PGLR1 and recruiting FDXs. [Fe S ] clusters that are required for their activity. Another enzyme crucial for carbon fixation and metabolism, 4 4 As oxygenic photosynthesis releases massive amounts in general, is the FTR. This key enzyme, which is almost of oxygen from water, reactive oxygen species are rou- uniquely found in photosynthetic organisms, catalyzes the tinely generated and damage some proteins in many physi- reduction of most thioredoxins (TRXs) found in plastids, ological conditions. Thus, several proteins are implicated thus indirectly participating to the regulation of all TRX- in the repair and protection of the photosystems and their dependent targets in a light-dependent manner [94]. FTR is antennae. One of these, photosystem II protein33 (PSB33), a heterodimer composed of a catalytic and a variable subunit is an integral membrane protein, which contributes to the [95]. In A. thaliana, there is a single gene for the catalytic maintenance of PSII-light-harvesting complex II (LHCII) subunit (FTRB) but two for the variable subunits (FTRA1 supercomplex organization in response to changing light and FTRA2). The function of the [F e S ] cluster found in the 4 4 levels [87]. Whereas the Arabidopsis protein is annotated catalytic subunit is to aid for the reduction of a redox-active as containing a Rieske-type Fe–S cluster by analogy to some disulfide, which reduces, in turn, the TRX disulfide [96]. bacterial counterparts, it does not have the Fe–S binding Given the numerous functions played by plastidial TRXs, it residues contrary to C. reinhardtii ortholog. Other Chla- is not surprising that an ftr knock-out mutant for the catalytic mydomonas proteins, referred to as CDJ3-5 for chloroplast- subunit is lethal. However, a virus-induced gene silencing targeted DnaJ-like proteins, might be important for PSII pro- (VIGS) approach led to plants exhibiting a sectored chlorotic tection. It was shown that CDJ3 and CDJ4 which interact leaf phenotype [97]. It could be observed that these plants with chloroplast ATP-bound HSP70B (heat-shock protein have (1) an abnormal chloroplast biogenesis, (2) a reduced 70B) and are located either in the stroma or attached to thy- photosynthetic performance as measured by the photochemi- lakoids, respectively, are able to bind a [Fe S ] cluster [88]. cal activities, the amount of assembled photosystems and 4 4 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 555 Nitrogen assimila�on Sulfur assimila�o n Carbon fixa� on Fa y acid CB GLU1 biosynthesis SIRB cycle Reducing GLT1 FTR TK FAB2 power supply Glycine betaine SIR NADPH siroheme NIR synthesis ? FAD5 FNR Phytochrome CMO APR1-3 synthesis HO1-4 HYDE HY1 HYDF NEET FDC1/2 FDX1-4 NAD SUFE3 HYDG synthesis psaC PSII repair/ PSB33 protec�on HYDA1-2 psaA psaB Iron signalling ? petC DjC17-18 NDH BCAA synthesis PSI PSII Cytb6/f VAL, ILE, LEU PAO RCCR HCAR IPP Chlorophyll IPMI LA CAO TK Plasto/phyloquinone HDR cLIP α-tocophérol MEP DHAD Chlorophyll Carotenoid pathway HDS catabolism GA-3P AHAS PDH DXS Thiamin DWARF pyruvate acetyl-coA 27.1-3 synthesis THIC ATase1-3 Strigolactone AIR PRA PRPP PTC52 TIC55 synthesis Fig. 3 Fe–S protein-dependent metabolic processes in plastids. Fe–S DjC, and HYDA1/2, there are several close isoforms which have not proteins are represented by dark red boxes. The light red boxes indi- been distinguished. The nomenclature used is the one of A. thaliana cate specificities found in algae either, because they do not exist in except for algal enzymes whose name is from C. reinhardtii. Abbre- terrestrial plants or in the case of PSB33, because only the algal iso- viations for all enzyme names can be found in the text. Other abbre- forms should incorporate a Fe–S cluster. Known FDX-dependent viations are: LA lipoic acid, BCAA br anched-chain amino acids, IPP enzymes have a red outline. Enzymes in green or outlined in orange isopentenyl diphosphate, GA-3P glyceraldehyde 3-phosphate, AIR use, respectively, thiamin or lipoic acid as cofactors. Note that PDH is 5-aminoimidazole ribonucleotide, PRA phospho-ribosylamine, PRPP dependent on both cofactors. For APR, DWARF, ATase, FDX, FDC, 5-phosphoribosyl-1-pyrophosphate CO assimilation rates, and (3) a defective PEP (plastid (pchlide)-dependent translocon component of 52  kDa RNA polymerase)-dependent plastid gene expression, very (PTC52) [102]. All possess a Rieske-type [Fe S ] cluster 2 2 likely because of FTR connection with TRX z [98]. Besides and a mononuclear iron-binding domain. While the five the redox regulation of carbon metabolism enzymes, other enzymes are dependent on FDX, HCAR, CAO, and PTC52 important functions of TRXs in chloroplasts are their par- are involved in chlorophyll synthesis, whereas PAO and ticipation to stress response by regenerating thiol-dependent TIC55 operate in its degradation. More precisely, CAO peroxidases and methionine sulfoxide reductases [99] and to and HCAR are part of the chlorophyll cycle, the process the chlorophyll metabolism by regulating several enzymes of interconversion between chlorophyll a and chlorophyll of the tetrapyrrole biosynthesis pathway [100]. b. The balance between chlorophyll a/b is important for both the stabilization and turnover of chlorophyll in the Beyond tetrapyrrole: chlorophyll and phytochromobilin light-harvesting complexes (LHCs) in diverse physiologi- cal situations, notably during greening and senescence Interestingly, several enzymes of the chlorophyll metabo- when LHCII is massively synthesized or degraded. CAO lism are FDX targets and/or possess Fe–S clusters. The is a thylakoid membrane-anchored enzyme catalyzing 7-hydroxymethyl chlorophyll a reductase (HCAR) is an the two steps of chlorophyll a-to-chlorophyll b oxidation enzyme binding two [Fe S ] clusters and an FAD [101]. [103]. PTC52 would catalyze an analogous oxidation, but 4 4 Besides there are four non-heme oxygenases, namely, using protochlorophyllide a as substrate. However, PTC52 pheophorbide a oxygenase (PAO), chlorophyll a oxyge- is localized at the envelope [104], suggesting that it may nase (CAO), translocon at the inner envelope membrane have another dispensable function, being part of a translo- of chloroplasts 55 (TIC55), and protochlorophyllide cation complex for the import of the protochlorophyllide 1 3 556 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 oxidoreductase A (PORA) precursor in plastids [105]. is encoded by a single gene (HY2), the heme oxygenase is Indeed, A. thaliana ptc52 knock-out lines have a growth encoded by four members in A. thaliana, HY1/HO1, and indistinguishable from wild-type plants (Table 1) [106]. HO2-4 [118]. On the contrary, an Arabidopsis mutant for CAO, named chlorina1, exhibits a pale green phenotype characterized Macronutrient assimilation: similarities in nitrogen by a chlorophyll b decrease [103] and is extremely sensi- and sulfur assimilation pathways tive to photooxidation due to the lack of chlorophyll–pro- tein antenna complexes in PSII and to an increased pro- The reductive assimilations of nitrogen and sulfur constitute duction of singlet oxygen [107]. HCAR catalyzes the two other chloroplastic metabolic processes, which rely on second half-reaction in chlorophyll b-to-chlorophyll a FDX-dependent Fe–S proteins. As already presented, sul- conversion, the first one being catalyzed by chlorophyll fate assimilation is extremely important, because it provides b reductases (CBR) [101]. While hcar mutants have no cysteine, which is the source of sulfur for many molecules phenotype under standard growth conditions, they exhibit but also the substrate of cysteine desulfurases and a protein a stay-green phenotype after transfer to darkness (Table 1) ligand in all plastidial Fe–S proteins known so far. Of the [108]. four enzymes/complexes, which allow forming cysteine from In higher plants, chlorophyll is broken down to colour- sulfate, two possess a Fe–S cluster. The second reaction, e.g, less linear tetrapyrroles in a series of reactions. One of the transformation of adenosine 5′ phosphosulfate (APS) to these steps, the porphyrin ring opening of pheophorbide a, sulfite, is catalyzed by adenosine 5′ -phosphosulfate reduc- is catalyzed by PAO. This step occurs in senescent leaves tases (APR). There are three isoforms in A. thaliana, APR1- and fruits, and requires FDXs and NADPH [109–111]. The 3, all localized in plastids. The enzymes are formed by two PAO proteins possess a C-terminal transmembrane domain domains, a reductase domain, that bears a [Fe S ] cluster, 4 4 for their binding to the thylakoid membrane [104]. The and a GRX domain at the C-terminus, that makes these PAO gene was identified by genetic studies and was initially enzymes glutathione-dependent [10]. The SIR catalyzes referred to as accelerated cell death 1 (ACD1) in Arabidopsis the next step, the six electron reduction of sulfite to sulfide. [112] or LLS1 (Lethal leaf spot 1) in maize [113]. Extinction This FDX-dependent enzyme incorporates a siroheme, e.g, of PAO in knock-out mutants or in antisense lines from dif- a heme whose iron atom is liganded by the thiolate ligand ferent plant species leads to a light-dependent premature cell of a [F e S ] cluster, which is crucial for its activity. There is 4 4 death phenotype, most likely due to cytotoxic effects of the a single, essential, SIR gene in Arabidopsis and the protein increased pheophorbide a [109, 112, 114]. Similar to hcar is found exclusively in plastids. A weak allele mutant with mutants, pao mutants have a stay-green phenotype in dark about 25% SIR activity is viable but has a strongly retarded [109, 114]. The product of the reaction catalyzed by PAO is growth, pointing to the extreme importance of this enzyme red chlorophyll catabolite, which is then reduced by a FDX- for plant development [119]. dependent red chlorophyll catabolite reductase (RCCR) to The assimilation of inorganic nitrogen (mostly in the yield the primary fluorescent chlorophyll catabolite (FCC), form of nitrate and ammonium) is another essential pro- pFCC [115]. From this primary phyllobilin, a large variety cess for plants taking place in part in plastids. Nitrate will of other phyllobilins is formed subsequently. Although a be reduced in two steps. The first one, catalyzed by nitrate tic55 mutant in Arabidopsis does not show any detectable reductase, gives nitrite, which is reduced to ammonia by a phenotype, TIC55, which is localized in the inner membrane FDX-dependent nitrite reductase (NIR). As the SIR enzyme, of the chloroplast envelope, is responsible for phyllobilin NIR binds a siroheme that is mandatory for the six elec- hydroxylation during senescence [104, 106]. This would be tron reduction of nitrite. This gene is also essential, since a the last step in this subcellular compartment, and in that mutant in barley does not grow in the absence of an external sense, TIC55 may contribute to chlorophyll catabolite export nitrogen source [120]. In the next steps, ammonia, including from plastids for their subsequent vacuolar detoxification. the part coming from the photorespiration process, is assimi- A closely related molecule to chlorophyll and incidentally lated via glutamine synthetase (GS) which catalyzes the con- heme is phytochromobilin (PfB), the chromophore usually densation of glutamate and ammonia into glutamine and via covalently bound to phytochromes of higher plants. All these glutamate synthase (GOGAT) which forms two molecules of molecules branch from protoporphyrin IX in the tetrapyr- glutamate from glutamine and 2-oxoglutarate. There is evi- role synthesis pathway. From the closed tetrapyrrole ring dence that NIR, GS, and GOGAT can form a complex within of heme, a heme oxygenase catalyzes the oxidative open- the chloroplast [121]. Plants possess two forms of chloro- ing of this chain to yield biliverdin IXa. This molecule is plastic GOGAT, which are dependent either on NADH or on then reduced into phytochromobilin by a PfB synthase. In FDX. All contain an FMN and a [Fe S ] cluster. In A. thali- 3 4 higher plants, both types of enzymes are soluble and depend ana, NADH-GOGAT is encoded by a unique gene (GLT1), on FDXs for their activity [116, 117]. While PfB synthase whereas two genes encode Fd-GOGAT (GLU1 and GLU2), 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 557 GLU1 is the predominant form in leaves [122]. Arabidopsis CMO-like gene in E. coli does not promote betaine synthesis mutants for GLU2 and GLT1 have no growth phenotype, [130]. This protein is unique to eukaryotic photosynthetic although a decrease in the chlorophyll content was measured organisms as it is not found in cyanobacteria, supporting in the glt1 mutant [123, 124]. An Arabidopsis mutant for a recent evolution of this enzyme. No Arabidopsis mutant GLU1 has a respiratory phenotype, i.e, a dwarf and chloro- has been characterized so far, but antisense CMO transgenic tic phenotype in air which is no longer visible under high sugar beet plants are susceptible to salt stress [131]. CO conditions [122, 123, 125]. Of importance for these To conclude on this part, it is important to note that most pathways, it is worth mentioning that sirohydrochlorin fer- of these enzymes are also expressed in plastids of non-pho- rochelatase (SIRB), the enzyme responsible for the last step tosynthetic tissues. In this context, FDXs are maintained of siroheme biosynthesis by inserting ferrous iron into the reduced by FNR and NADPH generated in the oxidative tetrapyrrole ring of sirohydrochlorin, is a [F e S ] enzyme pentose phosphate pathway, the reverse reaction compared 2 2 unlike bacterial orthologs. In this essential protein, the Fe–S to photosynthetic organs. A few other enzymes such as cluster is not mandatory for the enzymatic reaction, but it (1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate synthase might have a regulatory role [126]. (HDS), zeaxanthine epoxidase, and β-carotene 3 hydroxylase 1, 2) have been also described as FDX-dependent proteins, Fatty acid biosynthesis but they will be discussed in the next sections. However, several additional proteins or pathways are yet unidentified. The biosynthesis of fatty acids is another crucial pathway It is for instance worth mentioning that studies devoted to occurring in plastids, which depends directly or indirectly the isolation of FDX partners by proteomic approaches led to on Fe–S proteins. First, the acetyl-coenzyme A, that is the identification of novel putative targets at least in cyano- used as a building block for fatty acids, is generated by the bacteria and Chlamydomonas [132, 133]. In this respect, a plastidial pyruvate dehydrogenase (PDH) complex, its E2 pyruvate:ferredoxin oxidoreductase (PFO), found in many subunit being lipoylated and thus dependent on the Fe–S unicellular eukaryotes, decarboxylates pyruvate to acetyl- containing lipoate synthase (see below). After the synthe- coenzyme at the expense of FDXs [134]. The C. reinhardtii sis of saturated fatty acids, their conversion to unsaturated PFO possesses three distinct [Fe S ] clusters. It may also 4 4 forms, which are required for membrane fluidity, is catalyzed contribute to the light-independent H production by passing by fatty acid desaturases. Some of them contain a di-iron electron to the hydrogenase [135]. center and are FDX-dependent proteins [127]. The FAB2 protein is a soluble stearoyl-ACP desaturase introducing the Biosynthesis of lipoic acid and thiamin cofactors first double bond into stearoyl-ACP between carbons 9 and and their dependent pathways 10 to produce oleoyl-ACP (18:1 Delta9-ACP). The FAD5 protein attached to the chloroplast envelope inner membrane Requirement of two atypical radical‑SAM enzymes catalyzes the earliest step of 16:0 desaturation initiating the very rapid 16:0–16:1–16:2–16:3 desaturation of monogalac- Beyond their role in electron transfer, Fe–S clusters are also tosyldiacylglycerol (MGDG), one of the four main classes important for enzyme catalysis, especially during the bio- of glycerolipids found in the photosynthetic membranes of synthesis of vitamin B1/thiamin and of lipoic acid. Whereas higher plant chloroplasts with the digalactosyldiacylglycerol thiamin is only synthesized in chloroplasts a lipoic acid bio- (DGDG), the phospholipid phosphatidylglycerol (PG), and synthesis pathway is present in both plastids and mitochon- the sulfolipid sulfoquinovosyldiacylglycerol (SQDG) [128]. dria. This is consistent with the existence of two distinct Other plastidial linoleate/oleate desaturases (FAD4, 6, 7, 8) genes encoding a mitochondrial (mLIP) and a chloroplastic and the numerous FAD5-like proteins may also be depend- (cLIP) lipoate synthase [136]. ent on FDX as they also probably contain di-iron centers. Lipoic acid is synthesized from octanoic acid and thus via the fatty acid biosynthesis pathway by the addition of two Other metabolic processes sulfur atoms into the octanoyl group bound to an acyl carrier protein (ACP) via a radical-SAM mechanism. This reaction Additional FDX-dependent proteins are present in chloro- is catalyzed by lipoic acid synthase [136]. It is important to plasts. Besides PAO, CAO, PTC52, and TIC55, the fifth note that lipoic acid is synthesized attached to proteins and non-heme oxygenase found in plants [102] is referred to as no free lipoic acid is produced. The E. coli Lip5 binds two choline monooxygenase (CMO), because it was found to [Fe S ] clusters [137]. One cluster, coordinated by cysteines 4 4 catalyze the oxidation of choline, the first step of glycine present in a Cx Cx C motif common to all classical radical- 3 2 betaine biosynthesis in spinach [129]. However, this might SAM enzymes, is required for the formation of the activated not be the sole or main function, since Arabidopsis does not adenosyl radical from SAM molecules. The second clus- produce glycine betaine and expression of the Arabidopsis ter, coordinated by a Cx Cx C motif specific of lipoic acid 4 5 1 3 558 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 synthases, was suggested to provide the sulfur atoms and A single lipoic acid‑dependent enzyme but several thus to be degraded at each turnover of the enzyme. The thiamin‑dependent enzymes in plastids presence of a Fe–S cluster has not yet been demonstrated in plant cLIP, but Arabidopsis cLIP possesses both cysteine In plastids, the only known lipoic acid-dependent enzyme motifs and is able to complement the E. coli lip5 mutant is PDH. A similar complex is found in plant mitochondria, [136]. Plants impaired in cLIP have not been characterized but it uses lipoic acid synthesized in this compartment, as yet, but Arabidopsis mutants for genes involved in the syn- does another citric acid cycle enzyme, the α-ketoglutarate thesis of lipoic acid in mitochondria are lethal [138]. dehydrogenase or 2-oxoglutarate dehydrogenase complex, Thiamin is made of pyrimidine and thiazole heterocy- but also two complexes involved in the amino acid metabo- cles, both being synthesized in the chloroplast. The syn- lism, the glycine cleavage complex, and the branched-chain thesis of the thiazole moiety involves a 4-methyl-5-b- oxoacid dehydrogenase (BCDH) complex. On the other hydroxyethylthiazole phosphate (HET-P) synthase (THI1) hand, there is a single pathway for the synthesis of ThDP forming an adenylated thiazole intermediate (ADT) at the which is used as a coenzyme by many enzymes of the pri- expense of nicotinamide adenine dinucleotide (NAD) and mary metabolism, notably involved in the catabolism of glycine, ADT, which is then hydrolyzed to HET-P [139]. sugars and amino acids, and found in the chloroplasts, mito- The pyrimidine heterocycle is derived from purine bio- chondria, and cytosol [139]. In plastids, besides the PDH synthesis. The first step in the synthesis of the pyrimidine complex, thiamin is also a cofactor for transketolase (TK) moiety is catalyzed by the 4-amino-2-methyl-5-hydroxy- of both the Calvin–Benson cycle and the non-oxidative pen- methylpyrimidine phosphate (HMP-P) synthase (THIC), a tose phosphate pathway, for 1-deoxy-d -xylulose 5-phosphate radical-SAM Fe–S enzyme that forms HMP-P from 5-ami- synthase (DXS) of the methylerythritol phosphate (MEP) noimidazole ribonucleotide (AIR) and SAM. In contrast to pathway and for acetohydroxy acid synthase (AHAS) of the canonical radical-SAM enzymes, all THIC proteins harbour branched-chain amino acid (BCAA) biosynthesis pathway a Cx Cx C motif involved in the binding of a [Fe S ] cluster [139]. 2 4 4 4 in their C-terminal part [140, 141]. Then, an HMP-P kinase/ thiamin monophosphate (ThMP) pyrophosphorylase (TH1) The central pyruvate dehydrogenase complex PDH cata- phosphorylates HMP-P to HMP-PP but also condenses the lyzes the decarboxylation of pyruvate into acetyl-coA that latter compound to HET-P to form ThMP. This ThMP is is used in particular for fatty acid synthesis as already men- transformed into the diphosphate form ThDP in the cyto- tioned [128]. It consists of three subunits, E1–E3, each sol through the action of two consecutive enzymes before requiring a different cofactor. Thiamin is bound to the pyru- being redistributed to mitochondria and plastids. THIC is vate dehydrogenase subunit (E1), whereas the lipoic acid encoded by a single essential gene in Arabidopsis [140]. is covalently attached to the dihydrolipoyl acyltransferase An Arabidopsis thic mutant is lethal at the cotyledon stage subunit (E2) and an FAD is bound to the dihydrolipoam- unless supplemented with thiamin [140, 142]. Another fam- ide dehydrogenase subunit (E3). The attached lipoyl moiety ily of plastidial Fe–S enzymes is linked indirectly to thia- functions as a carrier of reaction intermediates among the min biosynthesis, because they catalyze the first committed active sites of the components of the complex. The E3 subu- step of the de novo synthesis of purine in chloroplasts. The nit has a key regulatory role by reoxidizing the lipoamide glutamine phosporibosyl pyrophosphate amidotransferases cofactor and thus completing the catalytic cycle. The dis- (ATases, also known as GPAT) catalyze the amination of ruption of the gene encoding the E2 subunit of plastidial 5-phosphoribosyl-1-pyrophosphate (PRPP) to 5-phospho- PDH results in an early embryo-lethal phenotype in Arabi- ribosylamine (PRA) with the concomitant conversion of dopsis [147]. glutamine into glutamate [143]. After four additional steps, PRA is transformed into AIR, the THIC substrate. In Arabi- Branched‑chain amino acid biosynthesis The AHAS pro- dopsis, ATase is encoded by a family of three genes (ATase1 tein is involved in the first steps of BCAA biosynthesis. to ATase3) which are expressed in various tissues at dif- This is a heterodimer, composed of separate catalytic and ferent levels [144, 145]. Whereas E. coli ATase does not regulatory subunits, which catalyzes the conversion of two require a Fe–S cluster as cofactor, the human enzyme uses molecules of pyruvate into 2-acetolactate used for valine a [Fe S ] cluster. Based on the conservation of the involved and leucine synthesis or of one molecule of pyruvate and 4 4 cysteines, the three A. thaliana isoforms should also bind a one molecule of 2-oxobutanoate into 2-aceto-2-hydroxy- Fe–S cluster. Whereas Arabidopsis ATase1 mutant has no butyrate used for isoleucine synthesis [148]. Interestingly, growth phenotype mutants lacking ATase2 exhibit strong two Fe–S enzymes named dihydroxyacid dehydratase growth retardation with bleached leaves (Table 1) [144]. In (DHAD) and isopropylmalate isomerase (IPMI) are the latter mutant exhibiting a decreased capacity in chlo- also required for BCAA synthesis. DHAD catalyzes the roplast protein import, cells are smaller in size [144, 146]. penultimate step before the formation of isoleucine and 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 559 valine, e.g., the dehydration of 2,3-dihydroxy-3-isovaler- suppressed by expressing the eukaryotic mevalonate path- ate or 2,3-dihydroxy-3-methylvalerate to the 2-oxo acids way, which does not rely on Fe–S proteins [54, 155]. (3-methyl-2-oxobutanoate or 3-methyl-2-oxopentanoate). Besides its involvement in the Calvin–Benson cycle In Arabidopsis, there is a single essential gene for DHAD where it catalyzes the formation of ribose-5-phosphate and [149]. However, Arabidopsis mutants with intermedi- xylulose-5-phosphate from sedoheptulose-7-phosphate and ate DHAD levels obtained by an RNAi approach indeed glyceraldehyde-3-phosphate (GA-3P), TK operates in the have reduced amounts of BCAA in roots, which cause a opposite direction in the non-oxidative pentose phosphate short root phenotype [149]. The only biochemical charac- pathway, forming GA-3P, which is then condensed to pyru- terization performed so far has been done with an enzyme vate to form 1-deoxy-d -xylulose 5-phosphate (DXP), a reac- purified from spinach leaves. Unlike the E. coli enzyme, tion catalyzed by DXP synthase (DXS). Although poorly which incorporates a [Fe S ] center, the spinach enzyme characterized in plants, it was demonstrated that antisense 4 4 incorporates a [Fe S ] cluster required for activity [150]. tobacco plants with variable TK levels have a marked 2 2 Leucine biosynthesis requires an additional Fe–S enzyme shoot weight decrease [156]. For the most affected lines, for the late reactions. The isopropylmalate isomerase cata- a decrease in chlorophylls and carotenoids was measured lyzes the reversible conversion of 2-isopropylmalate into which is consistent with the importance of DXP for the 3-isopropylmalate. In plants, IPMI consists of a heterodi- MEP pathway. Surprisingly, overexpression of an A. thali- mer composed of a large (LSU) and a small (SSU) subunit ana chloroplastic TK in tobacco leads to chlorosis, which is encoded by one and three genes in A. thaliana, respec- annihilated by thiamin supplementation [157]. In A. thali- tively [151]. The genetic analyses demonstrated that A. ana, DXS is an essential gene [158]. In sense and antisense thaliana knock-down mutants for the large subunit, which Arabidopsis lines exhibiting altered levels of DXS, both binds a [Fe S ] center, display a severe delay in devel- the chlorophylls, carotenoids, tocopherols, gibberellin, and 4 4 opment [151, 152]. Concerning small subunits, the SSU1 abscisic acid contents and the growth and germination rate protein is required for viability, unlike SSU2 and SSU3, are slightly affected [159]. Finally, downstream of DXS, two which might be redundant, because A. thaliana mutants Fe–S proteins are required to form two key intermediates in have no phenotype [151, 153]. In addition to a role in leu- isoprenoid biosynthesis: isopentenyl diphosphate (IPP) and cine biosynthesis, IPMI is involved in the biosynthesis of dimethylallyl diphosphate (DMAPP). Both the 1-hydroxy- glucosinolates, sulfur-containing secondary metabolites, 2-methyl-2-(E)-butenyl 4-diphosphate synthase (HDS/ serving for defence reactions. This is consistent with the ISPG) and the 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphos- fact that an identical reaction type exists for the Met chain phate reductase (HDR/ISPH/LytB) bind a [Fe S ] cluster 4 4 elongation cycle for glucosinolate formation and that A. [160, 161], FDX being able to provide electrons to HDS thaliana mutant plants for the large subunit accumulate [162]. Plant mutants disrupted in ISPG or ISPH gene have a both Leu biosynthesis and Met chain elongation interme- severely impaired chloroplastic development that causes an diates [151]. albino phenotype [163–165]. At least two enzymes participating in the carotenoid Isoprenoid biosynthesis and its derived molecules Isopre- biosynthesis pathways are dependent on FDXs. With- noids are very diverse metabolites, central to plant devel- out describing all the steps, the β-carotene 3 hydroxylase opment. We have already discussed the biosynthesis of 1, 2 which contains a di-iron center catalyzes two suc- chlorophylls, which consist of a tetrapyrrole ring with an cessive steps, the transformation of all-trans β-carotene attached isoprenoid-derived phytol chain, but many other to β-cryptoxanthin and then to zeaxanthin. Then, the fla- isoprenoids are present in plastids such as α-tocopherol, voprotein zeaxanthin epoxidase catalyzes the conversion phylloquinone, plastoquinone, and carotenoids to cite of zeaxanthin to antheraxanthin and then to violaxanthin only the most important. Moreover, several plant hor- [166]. These four steps require oxygen and FDX as an elec- mones are derived from carotenoids. All isoprenoids are tron donor. Other proteins in this pathway might, in fact, derived from a prenyl diphosphate (prenyl-PP) precursor, be dependent on FDXs. For instance, there are several which is synthesized by two independent pathways, the cytochrome P450 monooxygenases participating in this cytosolic mevalonate (MVA) pathway, and the plastidial pathway (and other pathways) in plastids, whose electron 2-C-methyl-d -erythritol 4-phosphate (MEP) pathway donors/acceptors are yet unknown. [154]. The latter pathway is dependent on both Fe–S and Derived from carotenoids, strigolactones (SL) are plant thiamin (TK and DXS)-dependent enzymes. In E. coli, the hormones having diverse functions in plant growth and Fe–S proteins belonging to this MEP pathway are the only development. Their biosynthesis begins with the conver- one that are completely essential. Indeed, the lethality of sion of all-trans β-carotene to 9-cis-β-carotene, a reaction sufa-isca and erpa mutants observed under aerobiosis is performed by a β-carotene isomerase named DWARF27. This protein, found from algae to higher plants, and first 1 3 560 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 characterized in rice is a Fe–S enzyme [167]. The Arabidop- holoproteins and predictions of Fe–S proteins from the pro- sis genome encodes three orthologs. An Arabidopsis mutant tein primary sequences are often impossible, because there for one of these genes and a rice mutant have shoot branch- is no universal signature for identifying Fe–S cluster ligands. ing phenotypes, but it remains relatively weak compared to Acknowledgements The work of both laboratories is supported other mutants affected in SL biosynthesis [168]. by the Agence Nationale de la Recherche, Grant no. ANR-2013- BSV6-0002-01. The UMR 1136 is also supported by a Grant over- Are there other plastidial Fe–S proteins to discover? seen by the French National Research Agency (ANR) as part of the “Investissements d’Avenir” program (ANR-11-LABX-0002-01, Lab of Excellence ARBRE). This publication is based upon work from Some Fe–S proteins, such as NEET, have been recently iden- COST Action CA15133, supported by COST (European Cooperation tified or characterized in plants. Unlike mitoNEET, which in Science and Technology). The authors are grateful to Pr Jean-Pierre is bound to the outer membrane of mitochondria in animals Jacquot for its careful reading of the manuscript. owing to a membrane anchoring extension, the Arabidop- sis NEET protein is located exclusively in the chloroplast Open Access This article is distributed under the terms of the Crea- tive Commons Attribution 4.0 International License (http://creat iveco stroma [169]. As its vertebrate counterparts, Arabidopsis mmons .org/licen ses/by/4.0/), which permits use, duplication, adapta- NEET forms dimers; each monomer harbouring an atypi- tion, distribution and reproduction in any medium or format, as long cal [Fe S ] cluster coordinated by three Cys and one His 2 2 as you give appropriate credit to the original author(s) and the source, [170]. While obtaining knock-out plants may have been provide a link to the Creative Commons license and indicate if changes were made. hampered by the fact, it is an essential gene, Arabidopsis lines with reduced AtNEET transcript levels exhibit late greening, delayed bolting, and early senescence. Moreover, these plants accumulate ROS and have an altered sensitivity References to Fe levels, which led to the proposal that AtNEET likely plays a role in the regulation of Fe homeostasis [170]. From 1. Briat J-F, Ravet K, Arnaud N et al (2010) New insights into fer- its capacity to transfer its Fe–S cluster to a FDX in vitro ritin synthesis and function highlight a link between iron homeo- [170], it may be hypothesized that NEET could be part of the stasis and oxidative stress in plants. Ann Bot 105:811–822. https ://doi.org/10.1093/aob/mcp12 8 SUF machinery and facilitate the trafficking of Fe–S clusters 2. Brumbarova T, Bauer P, Ivanov R (2015) Molecular mechanisms towards certain client proteins. governing Arabidopsis iron uptake. Trends Plant Sci 20:124–133. Another reason why we expect to discover novel Fe–S https ://doi.org/10.1016/j.tplan ts.2014.11.004 proteins is that some proteins may be specific to photosyn- 3. Jeong J, Guerinot ML (2009) Homing in on iron homeostasis in plants. Trends Plant Sci 14:280–285. https ://doi.org/10.1016/j. thetic organisms because of their atypical structure organi- tplan ts.2009.02.006 zation or their involvement in specific plastidial functions. 4. Solti Á, Kovács K, Müller B et al (2016) Does a voltage-sen- An interesting example in this regard is SUFE3, a chimeric sitive outer envelope transport mechanism contributes to the protein formed by an SUFE domain fused to a quinolinate chloroplast iron uptake? Planta 244:1303–1313. https ://doi. org/10.1007/s0042 5-016-2586-3 synthase domain, NADA [18]. This enzyme, which carries a 5. Jeong J, Cohu C, Kerkeb L et al (2008) Chloroplast Fe(III) che- [Fe S ] cluster indispensable for its activity and thus crucial 4 4 late reductase activity is essential for seedling viability under for NAD biosynthesis, is the sole NADA representative of iron limiting conditions. PNAS 105:10619–10624. https ://doi. A. thaliana. The fact that the Fe–S cluster in SUFE3 can be org/10.1073/pnas.07083 67105 6. Duy D, Wanner G, Meda AR et al (2007) PIC1, an ancient per- reconstituted using its own SUFE domain in the presence mease in Arabidopsis chloroplasts, mediates iron transport. Plant of NFS2, cysteine, and ferrous iron may render this protein Cell 19:986–1006. https ://doi.org/10.1105/tpc.106.04740 7 independent on the SUFBCD scaffold complex [18]. 7. Duy D, Stübe R, Wanner G, Philippar K (2011) The chloro- Other Fe–S protein-dependent processes likely remain plast permease PIC1 regulates plant growth and development by directing homeostasis and transport of iron. Plant Physiol to be identified in plastids as in other subcellular compart - 155:1709–1722. https ://doi.org/10.1104/pp.110.17023 3 ments. For instance, the affinity purification strategy used 8. Shimoni-Shor E, Hassidim M, Yuval-Naeh N, Keren N (2010) for cyanobacterial and algal enzymes indicates that numer- Disruption of Nap14, a plastid-localized non-intrinsic ABC ous FDX-dependent processes await identification [132, protein in Arabidopsis thaliana results in the over-accumu- lation of transition metals and in aberrant chloroplast struc- 133]. The same is true in mitochondria where the roles and tures. Plant Cell Environ 33:1029–1038. https ://doi.org/10.111 partners of the two FDXs are unknown. On the other hand, 1/j.1365-3040.2010.02124 .x novel Fe–S proteins will be undoubtedly identified in the 9. Tarantino D, Morandini P, Ramirez L et al (2011) Identifica- future thanks to the dozens of annotated sequenced genomes tion of an Arabidopsis mitoferrinlike carrier protein involved in Fe metabolism. Plant Physiol Biochem 49:520–529. https ://doi. now available for model plants and to the ever larger collec- org/10.1016/j.plaph y.2011.02.003 tions of available Arabidopsis mutants. The reasons why it 10. Takahashi H, Kopriva S, Giordano M et al (2011) Sulfur assimi- is not trivial to isolate them is that the sensitivity of these lation in photosynthetic organisms: molecular functions and metallic cofactors to oxygen may hamper the isolation of regulations of transporters and assimilatory enzymes. Annu Rev 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 561 Plant Biol 62:157–184. https ://doi.org/10.1146/annur ev-arpla 27. Yuda E, Tanaka N, Fujishiro T et al (2017) Mapping the key resi- nt-04211 0-10392 1 dues of SufB and SufD essential for biosynthesis of iron–sulfur 11. Cao M-J, Wang Z, Wirtz M et al (2013) SULTR3;1 is a chloro- clusters. Sci Rep. https ://doi.org/10.1038/s4159 8-017-09846 -2 plast-localized sulfate transporter in Arabidopsis thaliana. Plant 28. Blanc B, Clémancey M, Latour J-M et al (2014) Molecular inves- J 73:607–616. https ://doi.org/10.1111/tpj.12059 tigation of iron–sulfur cluster assembly scaffolds under stress. 12. Couturier J, Touraine B, Briat J-F et al (2013) The iron–sulfur Biochemistry 53:7867–7869. https://doi.or g/10.1021/bi5012496 cluster assembly machineries in plants: current knowledge and 29. Saini A, Mapolelo DT, Chahal HK et al (2010) SufD and SufC open questions. Front Plant Sci 4:259. https ://doi.org/10.3389/ ATPase activity are required for iron acquisition during in vivo fpls.2013.00259 Fe–S cluster formation on SufB. Biochemistry 49:9402–9412. 13. Balk J, Schaedler TA (2014) Iron cofactor assembly in plants. https ://doi.org/10.1021/bi101 1546 Annu Rev Plant Biol 65:125–153. https://doi.or g/10.1146/annur 30. Wollers S, Layer G, Garcia-Serres R et al (2010) Iron–sulfur (Fe– ev-arpla nt-05021 3-03575 9 S) cluster assembly the SufBCD complex is a new type of Fe–S 14. Hu X, Kato Y, Sumida A et al (2017) The SUFBC2D complex scaffold with a flavin redox cofactor. J Biol Chem 285:23331– is required for the biogenesis of all major classes of plastid Fe–S 23341. https ://doi.org/10.1074/jbc.M110.12744 9 proteins. Plant J 90:235–248. https ://doi.org/10.1111/tpj.13483 31. Ravet K, Touraine B, Boucherez J et al (2009) Ferritins con- 15. Pilon-Smits EAH, Garifullina GF, Abdel-Ghany S et al (2002) trol interaction between iron homeostasis and oxidative stress in Characterization of a NifS-like chloroplast protein from Arabi- Arabidopsis. Plant J 57:400–412. https://doi.or g/10.1111/j.1365- dopsis. implications for its role in sulfur and selenium metabo-313X.2008.03698 .x lism. Plant Physiol 130:1309–1318. https ://doi.or g/10.1104/ 32. Parent A, Elduque X, Cornu D et al (2015) Mammalian frataxin pp.102.01028 0 directly enhances sulfur transfer of NFS1 persulfide to both ISCU 16. Léon S, Touraine B, Briat J-F, Lobréaux S (2002) The AtNFS2 and free thiols. Nat Commun 6:5686. https ://doi.org/10.1038/ gene from Arabidopsis thaliana encodes a NifS-like plastidial ncomm s6686 cysteine desulphurase. Biochem J 366:557–564. https ://doi. 33. Colin F, Martelli A, Clémancey M et  al (2013) Mammalian org/10.1042/BJ200 20322 frataxin controls sulfur production and iron entry during de novo 17. Roret T, Pégeot H, Couturier J et al (2014) X-ray structures of Fe4S4 cluster assembly. J Am Chem Soc 135:733–740. https :// Nfs2, the plastidial cysteine desulfurase from Arabidopsis thali-doi.org/10.1021/ja308 736e ana. Acta Crystallogr F Struct Biol Commun 70:1180–1185. 34. Buchensky C, Sánchez M, Carrillo M et al (2017) Identification https ://doi.org/10.1107/S2053 230X1 40170 26 of two frataxin isoforms in Zea mays: structural and functional 18. Murthy NMU, Ollagnier-de-Choudens S, Sanakis Y et al (2007) studies. Biochimie 140:34–47. https ://doi.org/10.1016/j.bioch Characterization of Arabidopsis thaliana SufE2 and SufE3 func- i.2017.06.011 tions in chloroplast iron–sulfur cluster assembly and NAD syn- 35. Turowski VR, Aknin C, Maliandi MV et  al (2015) Frataxin thesis. J Biol Chem 282:18254–18264. https ://doi.org/10.1074/ is localized to both the chloroplast and mitochondrion and is jbc.M7014 28200 involved in chloroplast Fe–S protein function in Arabidop- 19. Ye H, Abdel-Ghany SE, Anderson TD et al (2006) CpSufE acti- sis. PLoS One 10:e0141443. https ://doi.or g/10.1371/jour n vates the cysteine desulfurase CpNifS for chloroplastic Fe–S al.pone.01414 43 cluster formation. J Biol Chem 281:8958–8969. https ://doi. 36. Maliandi MV, Busi MV, Turowski VR et al (2011) The mito- org/10.1074/jbc.M5127 37200 chondrial protein frataxin is essential for heme biosynthe- 20. Outten FW, Wood MJ, Muñoz FM, Storz G (2003) The SufE sis in plants. FEBS J 278:470–481. h t t p s : / / d oi . o r g / 1 0. 1 1 1 protein and the SufBCD complex enhance SufS cysteine desul-1/j.1742-4658.2010.07968 .x furase activity as part of a sulfur transfer pathway for Fe–S clus- 37. Söderberg C, Gillam ME, Ahlgren E-C et al (2016) The structure ter assembly in Escherichia coli. J Biol Chem 278:45713–45719. of the complex between yeast frataxin and ferrochelatase: char- https ://doi.org/10.1074/jbc.M3080 04200 acterization and pre-steady state reaction of ferrous iron delivery 21. Xu XM, Møller SG (2006) AtSufE is an essential activator of and heme synthesis. J Biol Chem 291:11887–11898. https://doi. plastidic and mitochondrial desulfurases in Arabidopsis. EMBO org/10.1074/jbc.M115.70112 8 J 25:900–909. https ://doi.org/10.1038/sj.emboj .76009 68 38. Dhalleine T, Rouhier N, Couturier J (2014) Putative roles of 22. Van Hoewyk D, Abdel-Ghany SE, Cohu CM et al (2007) Chloro- glutaredoxin-BolA holo-heterodimers in plants. Plant Signal plast iron–sulfur cluster protein maturation requires the essential Behav 9:e28564. https ://doi.org/10.4161/psb.28564 cysteine desulfurase CpNifS. PNAS 104:5686–5691. https://doi. 39. Roret T, Tsan P, Couturier J et al (2014) Structural and spectro- org/10.1073/pnas.07007 74104 scopic insights into BolA-glutaredoxin complexes. J Biol Chem 23. Cory SA, Van Vranken JG, Brignole EJ et al (2017) Structure of 289:24588–24598. https ://doi.org/10.1074/jbc.M114.57270 1 human Fe–S assembly subcomplex reveals unexpected cysteine 40. Braymer JJ, Lill R (2017) Iron–sulfur cluster biogenesis and traf- desulfurase architecture and acyl-ACP-ISD11 interactions. PNAS ficking in mitochondria. J Biol Chem 292:12754–12763. https:// 114:E5325–E5334. https ://doi.org/10.1073/pnas.17028 49114 doi.org/10.1074/jbc.R117.78710 1 24. Boniecki MT, Freibert SA, Mühlenhoff U et al (2017) Structure 41. Brancaccio D, Gallo A, Piccioli M et al (2017) [4Fe–4S] cluster and functional dynamics of the mitochondrial Fe/S cluster syn- assembly in mitochondria and Its impairment by copper. J Am thesis complex. Nat Commun 8:1287. https ://doi.org/10.1038/ Chem Soc 139:719–730. https ://doi.org/10.1021/jacs.6b095 67 s4146 7-017-01497 -1 42. Brancaccio D, Gallo A, Mikolajczyk M et al (2014) Formation of 25. Layer G, Gaddam SA, Ayala-Castro CN et al (2007) SufE trans- [4Fe–4S] clusters in the mitochondrial iron–sulfur cluster assem- fers sulfur from SufS to SufB for iron–sulfur cluster assembly. bly machinery. J Am Chem Soc 136:16240–16250. https ://doi. J Biol Chem 282:13342–13350. https ://doi.or g/10.1074/jbc.org/10.1021/ja507 822j M6085 55200 43. Bandyopadhyay S, Gama F, Molina-Navarro MM et al (2008) 26. Hirabayashi K, Yuda E, Tanaka N et al (2015) Functional dynam- Chloroplast monothiol glutaredoxins as scaffold proteins for the ics revealed by the structure of the SufBCD complex, a novel assembly and delivery of [2Fe–2S] clusters. EMBO J 27:1122– ATP-binding cassette (ABC) protein that serves as a scaffold for 1133. https ://doi.org/10.1038/emboj .2008.50 44. Rey P, Becuwe N, Tourrette S, Rouhier N (2017) Involve- iron–sulfur cluster biogenesis. J Biol Chem 290:29717–29731. ment of Arabidopsis glutaredoxin S14 in the maintenance of https ://doi.org/10.1074/jbc.M115.68093 4 1 3 562 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 chlorophyll content. Plant Cell Environ 40:2319–2332. https :// iron–sulphur cluster metabolism. J Exp Bot 63:403–411. https:// doi.org/10.1111/pce.13036 doi.org/10.1093/jxb/err28 6 45. Moseler A, Aller I, Wagner S et al (2015) The mitochondrial 62. Hasnain G, Waller JC, Alvarez S et al (2012) Mutational analysis monothiol glutaredoxin S15 is essential for iron–sulfur protein of YgfZ, a folate-dependent protein implicated in iron/sulphur maturation in Arabidopsis thaliana. PNAS 112:13735–13740. cluster metabolism. FEMS Microbiol Lett 326:168–172. https:// https ://doi.org/10.1073/pnas.15108 35112 doi.org/10.1111/j.1574-6968.2011.02448 .x 46. Li H, Outten CE (2012) Monothiol CGFS glutaredoxins and 63. Teplyakov A, Obmolova G, Sarikaya E et  al (2004) Crystal BolA-like proteins: [2Fe–2S] binding partners in iron homeosta- structure of the YgfZ protein from Escherichia coli suggests sis. Biochemistry 51:4377–4389. https ://doi.org/10.1021/bi300 a folate-dependent regulatory role in one-carbon metabo- 393z lism. J Bacteriol 186:7134–7140. https ://doi.or g/10.1128/ 47. Melber A, Na U, Vashisht A et al (2016) Role of Nfu1 and Bol3 JB.186.21.7134-7140.2004 in iron–sulfur cluster transfer to mitochondrial clients. eLife. 64. Navarro-Sastre A, Tort F, Stehling O et al (2011) A fatal mito- https ://doi.org/10.7554/elife .15991 chondrial disease is associated with defective NFU1 function 48. Uzarska MA, Nasta V, Weiler BD et al (2016) Mitochondrial in the maturation of a subset of mitochondrial Fe–S proteins. Bol1 and Bol3 function as assembly factors for specific iron– Am J Hum Genet 89:656–667. https ://doi.or g/10.1016/j. sulfur proteins. eLife. https ://doi.org/10.7554/elife .16673 ajhg.2011.10.005 49. Cameron JM, Janer A, Levandovskiy V et al (2011) Mutations 65. Léon S, Touraine B, Ribot C et al (2003) Iron–sulphur cluster in iron–sulfur cluster scaffold genes NFU1 and BOLA3 cause assembly in plants: distinct NFU proteins in mitochondria and a fatal deficiency of multiple respiratory chain and 2-oxoacid plastids from Arabidopsis thaliana. Biochem J 371:823–830. dehydrogenase enzymes. Am J Hum Genet 89:486–495. https ://https ://doi.org/10.1042/bj200 21946 doi.org/10.1016/j.ajhg.2011.08.011 66. Yabe T, Morimoto K, Kikuchi S et al (2004) The Arabidop- 50. Couturier J, Wu H-C, Dhalleine T et al (2014) Monothiol glutar- sis chloroplastic NifU-like protein CnfU, which can act as an edoxin–BolA interactions: redox control of Arabidopsis thaliana iron–sulfur cluster scaffold protein, is required for biogenesis of BolA2 and SufE1. Mol Plant 7:187–205. https://doi.or g/10.1093/ ferredoxin and photosystem I. Plant Cell 16:993–1007. https :// mp/sst15 6doi.org/10.1105/tpc.02051 1 51. Couturier J, Przybyla-Toscano J, Roret T et al (2015) The roles 67. Gao H, Subramanian S, Couturier J et al (2013) Arabidopsis of glutaredoxins ligating Fe–S clusters: sensing, transfer or repair thaliana Nfu2 accommodates [2Fe–2S] or [4Fe–4S] clusters and functions? Biochim Biophys Acta Mol Cell Res 1853:1513– is competent for in vitro maturation of chloroplast [2Fe–2S] and 1527. https ://doi.org/10.1016/j.bbamc r.2014.09.018 [4Fe–4S] cluster-containing proteins. Biochemistry 52:6633– 52. Mapolelo DT, Zhang B, Naik SG et al (2012) Spectroscopic and 6645. https ://doi.org/10.1021/bi400 7622 functional characterization of iron–sulfur cluster-bound forms 68. Py B, Gerez C, Angelini S et al (2012) Molecular organization, of Azotobacter vinelandii NifIscA. Biochemistry 51:8071–8084. biochemical function, cellular role and evolution of NfuA, an https ://doi.org/10.1021/bi300 6658 atypical Fe–S carrier. Mol Microbiol 86:155–171. https ://doi. 53. Tanaka N, Kanazawa M, Tonosaki K et al (2015) Novel features org/10.1111/j.1365-2958.2012.08181 .x of the ISC machinery revealed by characterization of Escheri- 69. Nath K, O’Donnell JP, Lu Y (2017) Chloroplastic iron–sulfur chia coli mutants that survive without iron–sulfur clusters. Mol scaffold protein NFU3 is essential to overall plant fitness. Plant Microbiol. https ://doi.org/10.1111/mmi.13271 Signal Behav 12:e1282023. h tt ps : // d oi .o r g/ 10 .1 08 0 /1 55 92 54. Vinella D, Brochier-Armanet C, Loiseau L et al (2009) Iron–324.2017.12820 23 sulfur (Fe/S) protein biogenesis: phylogenomic and genetic 70. Nath K, Wessendorf RL, Lu Y (2016) A nitrogen-fixing subunit studies of A-type carriers. PLoS Genet 5:e1000497. https ://doi. essential for accumulating 4Fe–4S-containing photosystem I core org/10.1371/journ al.pgen.10004 97 proteins. Plant Physiol 172:2459–2470. https://doi.or g/10.1104/ 55. Abdel-Ghany SE, Ye H, Garifullina GF et al (2005) Iron–sul- pp.16.01564 fur cluster biogenesis in chloroplasts. Involvement of the scaf- 71. Touraine B, Boutin J-P, Marion-Poll A et al (2004) Nfu2: a scaf- fold protein CpIscA. Plant Physiol 138:161–172. https ://doi. fold protein required for [4Fe–4S] and ferredoxin iron-sulphur org/10.1104/pp.104.05860 2 cluster assembly in Arabidopsis chloroplasts. Plant J 40:101–111. 56. Yabe T, Nakai M (2006) Arabidopsis AtIscA-I is affected by defi-https ://doi.org/10.1111/j.1365-313X.2004.02189 .x ciency of Fe–S cluster biosynthetic scaffold AtCnfU-V. Biochem 72. Lezhneva L, Amann K, Meurer J (2004) The universally con- Biophys Res Commun 340:1047–1052. https://doi.or g/10.1016/j. served HCF101 protein is involved in assembly of [4Fe–4S]- bbrc.2005.12.104 cluster-containing complexes in Arabidopsis thaliana 57. Mapolelo DT, Zhang B, Randeniya S et al (2013) Monothiol glu- chloroplasts. Plant J 37:174–185. https: //doi.org/10.1046/j.1365- taredoxins and A-type proteins: partners in Fe–S cluster traffick -313X.2003.01952 .x ing. Dalton Trans 42:3107. https ://doi.org/10.1039/c2dt3 2263c 73. Schwenkert S, Netz DJA, Frazzon J et al (2010) Chloroplast 58. Gelling C, Dawes IW, Richhardt N et al (2008) Mitochondrial HCF101 is a scaffold protein for [4Fe–4S] cluster assembly. Iba57p is required for Fe/S cluster formation on aconitase and Biochem J 425:207–218. https ://doi.org/10.1042/BJ200 91290 activation of radical SAM enzymes. Mol Cell Biol 28:1851– 74. Stöckel J, Oelmüller R (2004) A novel protein for photosys- 1861. https ://doi.org/10.1128/MCB.01963 -07 tem I biogenesis. J Biol Chem 279:10243–10251. https ://doi. 59. Sheftel AD, Wilbrecht C, Stehling O et al (2012) The human org/10.1074/jbc.M3092 46200 mitochondrial ISCA1, ISCA2, and IBA57 proteins are required 75. Schneider D, Berry S, Volkmer T et al (2004) PetC1 is the major for [4Fe–4S] protein maturation. Mol Biol Cell 23:1157–1166. Rieske iron–sulfur protein in the cytochrome b6f complex of https ://doi.org/10.1091/mbc.E11-09-0772 Synechocystis sp. PCC 6803. J Biol Chem 279:39383–39388. 60. Waller JC, Alvarez S, Naponelli V et al (2010) A role for tet-https ://doi.org/10.1074/jbc.M4062 88200 rahydrofolates in the metabolism of iron–sulfur clusters in 76. Maiwald D, Dietzmann A, Jahns P et al (2003) Knock-out of the all domains of life. PNAS 107:10412–10417. h t t p s : / / d o i . genes coding for the Rieske protein and the ATP-synthase delta- org/10.1073/pnas.09115 86107 subunit of Arabidopsis. Effects on photosynthesis, thylakoid pro- tein composition, and nuclear chloroplast gene expression. Plant 61. Waller JC, Ellens KW, Alvarez S et al (2012) Mitochondrial and Physiol 133:191–202 plastidial COG0354 proteins have folate-dependent functions in 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 563 77. Golbeck JH (2003) The binding of cofactors to photosystem I 94. Chibani K, Wingsle G, Jacquot J-P et al (2009) Comparative analyzed by spectroscopic and mutagenic methods. Annu Rev genomic study of the thioredoxin family in photosynthetic Biophys Biomol Struct 32:237–256. https ://doi.or g/10.1146/ organisms with emphasis on Populus trichocarpa. Mol Plant annur ev.bioph ys.32.11060 1.14235 6 2:308–322. https ://doi.org/10.1093/mp/ssn07 6 78. Hanke GT, Hase T (2008) Variable photosynthetic roles of two 95. Jacquot J-P, Eklund H, Rouhier N, Schürmann P (2009) Struc- leaf-type ferredoxins in arabidopsis, as revealed by RNA interfer- tural and evolutionary aspects of thioredoxin reductases in ence. Photochem Photobiol 84:1302–1309. https: //doi.org/10.11 photosynthetic organisms. Trends Plant Sci 14:336–343. https 11/j.1751-1097.2008.00411 .x://doi.org/10.1016/j.tplan ts.2009.03.005 79. Voss I, Goss T, Murozuka E et al (2011) FdC1, a novel ferredoxin 96. Dai S, Glauser DA, Bourquin F et al (2007) Structural snap- protein capable of alternative electron partitioning, increases in shots along the reaction pathway of ferredoxin–thioredoxin conditions of acceptor limitation at photosystem I. J Biol Chem reductase. Nature 448:92–96 286:50–59. https ://doi.org/10.1074/jbc.M110.16156 2 97. Wang P, Liu J, Liu B et  al (2014) Ferredoxin:thioredoxin 80. Hanke GT, Kimata-Ariga Y, Taniguchi I, Hase T (2004) A post reductase is required for proper chloroplast development genomic characterization of Arabidopsis ferredoxins. Plant Phys- and is involved in the regulation of plastid gene expression iol 134:255–264. https ://doi.org/10.1104/pp.103.03275 5 in Arabidopsis thaliana. Mol Plant 7:1586–1590. https ://doi. 81. Rumeau D, Peltier G, Cournac L (2007) Chlororespiration and org/10.1093/mp/ssu06 9 cyclic electron flow around PSI during photosynthesis and plant 98. Arsova B, Hoja U, Wimmelbacher M et al (2010) Plastidial stress response. Plant Cell Environ 30:1041–1051. https ://doi. thioredoxin z interacts with two fructokinase-like proteins in org/10.1111/j.1365-3040.2007.01675 .x a thiol-dependent manner: evidence for an essential role in 82. Yamori W, Shikanai T (2016) Physiological functions of cyclic chloroplast development in Arabidopsis and Nicotiana bentha- electron transport around photosystem I in sustaining photosyn- miana. Plant Cell 22:1498–1515. https ://doi.or g/10.1105/ thesis and plant growth. Annu Rev Plant Biol 67:81–106. https tpc.109.07100 1 ://doi.org/10.1146/annur ev-arpla nt-04301 5-11200 2 99. Vieira Dos Santos C, Rey P (2006) Plant thioredoxins are 83. Suorsa M, Sirpiö S, Aro E-M (2009) Towards characterization key actors in the oxidative stress response. Trends Plant Sci of the chloroplast NAD(P)H dehydrogenase complex. Mol Plant 11:329–334. https ://doi.org/10.1016/j.tplan ts.2006.05.005 2:1127–1140. https ://doi.org/10.1093/mp/ssp05 2 100. Brzezowski P, Richter AS, Grimm B (2015) Regulation and 84. Peng L, Yamamoto H, Shikanai T (2011) Structure and bio- function of tetrapyrrole biosynthesis in plants and algae. Bio- genesis of the chloroplast NAD(P)H dehydrogenase complex. chim Biophys Acta Bioenerget 1847:968–985. https ://doi. Biochim Biophys Acta Bioenerget 1807:945–953. https ://doi.org/10.1016/j.bbabi o.2015.05.007 org/10.1016/j.bbabi o.2010.10.015 101. Wang X, Liu L (2016) Crystal structure and catalytic mecha- 85. Peters JW, Broderick JB (2012) Emerging paradigms for complex nism of 7-hydroxymethyl chlorophyll a reductase. J Biol Chem iron–sulfur cofactor assembly and insertion. Annu Rev Biochem 291:13349–13359. https ://doi.org/10.1074/jbc.M116.72034 2 81:429–450. https ://doi.org/10.1146/annur ev-bioch em-05261 102. Gray J, Wardzala E, Yang M et al (2004) A small family of 0-09491 1 LLS1-related non-heme oxygenases in plants with an origin 86. Sawyer A, Bai Y, Lu Y et al (2017) Compartmentalisation of amongst oxygenic photosynthesizers. Plant Mol Biol 54:39–54. [FeFe]-hydrogenase maturation in Chlamydomonas reinhardtii. https ://doi.org/10.1023/B:PLAN.00000 28766 .61559 .4c Plant J 90:1134–1143. https ://doi.org/10.1111/tpj.13535 103. Espineda CE, Linford AS, Devine D, Brusslan JA (1999) The 87. Fristedt R, Herdean A, Blaby-Haas CE et al (2015) PHOTO- AtCAO gene, encoding chlorophyll a oxygenase, is required SYSTEM II PROTEIN33, a protein conserved in the plastid line- for chlorophyll b synthesis in Arabidopsis thaliana. PNAS age, is associated with the chloroplast thylakoid membrane and 96:10507–10511. https ://doi.org/10.1073/pnas.96.18.10507 provides stability to photosystem II supercomplexes in Arabi- 104. Hauenstein M, Christ B, Das A et al (2016) A role for TIC55 dopsis. Plant Physiol 167:481–492. https ://doi.or g/10.1104/ as a hydroxylase of phyllobilins, the products of chlorophyll pp.114.25333 6 breakdown during plant senescence. Plant Cell 28:2510–2527. 88. Dorn KV, Willmund F, Schwarz C et al (2010) Chloroplast DnaJ-https ://doi.org/10.1105/tpc.16.00630 like proteins 3 and 4 (CDJ3/4) from Chlamydomonas reinhardtii 105. Reinbothe S, Quigley F, Gray J et al (2004) Identification of contain redox-active Fe–S clusters and interact with stromal plastid envelope proteins required for import of protochloro- HSP70B. Biochem J 427:205–215. https ://doi.or g/10.1042/ phyllide oxidoreductase A into the chloroplast of barley. PNAS BJ200 91412 101:2197–2202. https ://doi.org/10.1073/pnas.03072 84101 89. Schroda M, Vallon O, Wollman FA, Beck CF (1999) A chloro- 106. Boij P, Patel R, Garcia C et al (2009) In vivo studies on the plast-targeted heat shock protein 70 (HSP70) contributes to the roles of Tic55-related proteins in chloroplast protein import photoprotection and repair of photosystem II during and after in Arabidopsis thaliana. Mol Plant 2:1397–1409. https ://doi. photoinhibition. Plant Cell 11:1165–1178org/10.1093/mp/ssp07 9 90. Liu C, Willmund F, Golecki JR et al (2007) The chloroplast 107. Ramel F, Ksas B, Akkari E et al (2013) Light-induced accli- HSP70B-CDJ2-CGE1 chaperones catalyse assembly and disas- mation of the Arabidopsis chlorina1 mutant to singlet oxy- sembly of VIPP1 oligomers in Chlamydomonas. Plant J 50:265– gen. Plant Cell 25:1445–1462. https ://d oi.or g/10.1105/ 277. https ://doi.org/10.1111/j.1365-313X.2007.03047 .xtpc.113.10982 7 91. Petti C, Nair M, DeBolt S (2014) The involvement of J-protein 108. Meguro M, Ito H, Takabayashi A et al (2011) Identification of AtDjC17 in root development in Arabidopsis. Front Plant Sci the 7-hydroxymethyl chlorophyll a reductase of the chlorophyll 5:532. https ://doi.org/10.3389/fpls.2014.00532 cycle in Arabidopsis. Plant Cell 23:3442–3453. https ://doi. 92. Hanke G, Mulo P (2013) Plant type ferredoxins and ferredoxin-org/10.1105/tpc.111.08971 4 dependent metabolism. Plant Cell Environ 36:1071–1084. https 109. Pruzinská A, Tanner G, Anders I et al (2003) Chlorophyll break- ://doi.org/10.1111/pce.12046 down: pheophorbide a oxygenase is a Rieske-type iron–sulfur 93. Terauchi AM, Lu S-F, Zaffagnini M et  al (2009) Pattern of protein, encoded by the accelerated cell death 1 gene. PNAS expression and substrate specificity of chloroplast ferredoxins 100:15259–15264. https ://doi.org/10.1073/pnas.20365 71100 110. Hörtensteiner S, Wüthrich KL, Matile P et al (1998) The key from Chlamydomonas reinhardtii. J Biol Chem 284:25867– step in chlorophyll breakdown in higher plants cleavage of 25878. https ://doi.org/10.1074/jbc.M109.02362 2 1 3 564 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 pheophorbide a macrocycle by a monooxygenase. J Biol Chem castor seed and its relationship to other di-iron proteins. EMBO 273:15335–15339. https ://doi.org/10.1074/jbc.273.25.15335 J 15:4081–4092 111. Hörtensteiner S, Kräutler B (2011) Chlorophyll breakdown in 128. Li-Beisson Y, Shorrosh B, Beisson F et  al (2013) Acyl- higher plants. Biochim Biophys Acta Bioenerget 1807:977–988. lipid metabolism. Arabidopsis Book 11:e0161. https ://doi. https ://doi.org/10.1016/j.bbabi o.2010.12.007 org/10.1199/tab.0161 112. Greenberg JT, Ausubel FM (1993) Arabidopsis mutants com- 129. Rathinasabapathi B, Burnet M, Russell BL et al (1997) Choline promised for the control of cellular damage during pathogenesis monooxygenase, an unusual iron–sulfur enzyme catalyzing the and aging. Plant J 4:327–341. https ://doi.org/10.1046/j.1365- first step of glycine betaine synthesis in plants: prosthetic group 313X.1993.04020 327.x characterization and cDNA cloning. PNAS 94:3454–3458 113. Gray J, Close PS, Briggs SP, Johal GS (1997) A novel suppressor 130. Hibino T, Waditee R, Araki E et al (2002) Functional charac- of cell death in plants encoded by the Lls1 gene of maize. Cell terization of choline monooxygenase, an enzyme for betaine 89:25–31. https ://doi.org/10.1016/S0092 -8674(00)80179 -8 synthesis in plants. J Biol Chem 277:41352–41360. https://doi. 114. Pružinská A, Tanner G, Aubry S et al (2005) Chlorophyll break-org/10.1074/jbc.M2059 65200 down in senescent Arabidopsis leaves. Characterization of chlo- 131. Yamada N, Takahashi H, Kitou K et  al (2015) Suppressed rophyll catabolites and of chlorophyll catabolic enzymes involved expression of choline monooxygenase in sugar beet on the in the degreening reaction. Plant Physiol 139:52–63. https://doi. accumulation of glycine betaine. Plant Physiol Biochem org/10.1104/pp.105.06587 0 96:217–221. https ://doi.org/10.1016/j.plaph y.2015.06.014 115. Rodoni S, Vicentini F, Schellenberg M et al (1997) Partial puri- 132. Hanke GT, Satomi Y, Shinmura K et  al (2011) A screen fication and characterization of red chlorophyll catabolite reduc- for potential ferredoxin electron transfer partners uncovers tase, a stroma protein involved in chlorophyll breakdown. Plant new, redox dependent interactions. Biochim Biophys Acta Physiol 115:677–682 1814:366–374. https ://doi.org/10.1016/j.bbapa p.2010.09.011 116. Muramoto T, Tsurui N, Terry MJ et al (2002) Expression and 133. Peden EA, Boehm M, Mulder DW et al (2013) Identification biochemical properties of a ferredoxin-dependent heme oxyge- of global ferredoxin interaction networks in Chlamydomonas nase required for phytochrome chromophore synthesis. Plant reinhardtii. J Biol Chem 288:35192–35209. https ://doi. Physiol 130:1958–1966. https ://doi.org/10.1104/pp.00812 8org/10.1074/jbc.M113.48372 7 117. Kohchi T, Mukougawa K, Frankenberg N et al (2001) The Arabi- 134. van Lis R, Baffert C, Couté Y et  al (2013) C h l a - dopsis HY2 gene encodes phytochromobilin synthase, a ferre- mydomonas reinhardtii chloroplasts contain a homodimeric doxin-dependent biliverdin reductase. Plant Cell 13:425–436 pyruvate:ferredoxin oxidoreductase that functions with FDX1. 118. Davis SJ, Bhoo SH, Durski AM et al (2001) The heme-oxygenase Plant Physiol 161:57–71. https: //doi.org/10.1104/pp.112.20818 family required for phytochrome chromophore biosynthesis is 1 necessary for proper photomorphogenesis in higher plants. Plant 135. Noth J, Krawietz D, Hemschemeier A, Happe T (2013) Physiol 126:656–669. https ://doi.org/10.1104/pp.126.2.656 Pyruvate:ferredoxin oxidoreductase is coupled to light-independ- 119. Khan MS, Haas FH, Samami AA et  al (2010) Sulfite reduc- ent hydrogen production in Chlamydomonas reinhardtii. J Biol tase defines a newly discovered bottleneck for assimilatory Chem 288:4368–4377. https://doi.or g/10.1074/jbc.M112.42998 sulfate reduction and is essential for growth and development 5 in Arabidopsis thaliana. Plant Cell 22:1216–1231. https ://doi. 136. Yasuno R, Wada H (2002) The biosynthetic pathway for lipoic org/10.1105/tpc.110.07408 8 acid is present in plastids and mitochondria in Arabidopsis thali- 120. Duncanson E, Gilkes AF, Kirk DW et al (1993) nir1, a condi- ana 1. FEBS Lett 517:110–114. https ://doi.org/10.1016/S0014 tional-lethal mutation in barley causing a defect in nitrite reduc--5793(02)02589 -9 tion. Mol Gen Genet 236:275–282. https ://doi.or g/10.1007/ 137. Cicchillo RM, Lee K-H, Baleanu-Gogonea C et al (2004) Escher- BF002 77123 ichia coli lipoyl synthase binds two distinct [4Fe–4S] clusters 121. Kimata-Ariga Y, Hase T (2014) Multiple complexes of nitro- per polypeptide. Biochemistry 43:11770–11781. https ://doi. gen assimilatory enzymes in spinach chloroplasts: possible org/10.1021/bi048 8505 mechanisms for the regulation of enzyme function. PLoS One 138. Ewald R, Hoffmann C, Florian A et al (2014) Lipoate-protein 9:e108965. https ://doi.org/10.1371/journ al.pone.01089 65 ligase and octanoyltransferase are essential for protein lipoyla- 122. Coschigano KT, Melo-Oliveira R, Lim J, Coruzzi GM (1998) tion in mitochondria of Arabidopsis. Plant Physiol 165:978–990. Arabidopsis gls mutants and distinct Fd-GOGAT genes. Impli-https ://doi.org/10.1104/pp.114.23831 1 cations for photorespiration and primary nitrogen assimilation. 139. Goyer A (2017) Thiamin biofortification of crops. Curr Opin Plant Cell 10:741–752 Biotechnol 44:1–7. https: //doi.org/10.1016/j.copbio .2016.09.005 123. Lancien M, Martin M, Hsieh M-H et al (2002) Arabidopsis glt1-T 140. Raschke M, Bürkle L, Müller N et al (2007) Vitamin B1 bio- mutant defines a role for NADH-GOGAT in the non-photorespi- synthesis in plants requires the essential iron sulfur cluster pro- ratory ammonium assimilatory pathway. Plant J 29:347–358 tein, THIC. PNAS 104:19637–19642. https ://doi.org/10.1073/ 124. Potel F, Valadier M-H, Ferrario-Méry S et al (2009) Assimilation pnas.07095 97104 of excess ammonium into amino acids and nitrogen translocation 141. Fenwick MK, Mehta AP, Zhang Y et al (2015) Non-canonical in Arabidopsis thaliana—roles of glutamate synthases and car- active site architecture of the radical SAM thiamin pyrimidine bamoylphosphate synthetase in leaves. FEBS J 276:4061–4076. synthase. Nat Commun 6:6480. https ://doi.org/10.1038/ncomm https ://doi.org/10.1111/j.1742-4658.2009.07114 .x s7480 125. Somerville CR, Ogren WL (1980) Inhibition of photosynthesis 142. Kong D, Zhu Y, Wu H et al (2008) AtTHIC, a gene involved in in Arabidopsis mutants lacking leaf glutamate synthase activity. thiamine biosynthesis in Arabidopsis thaliana. Cell Res 18:566– Nature 286:257–259. https ://doi.org/10.1038/28625 7a0 576. https ://doi.org/10.1038/cr.2008.35 126. Saha K, Webb ME, Rigby SEJ et al (2012) Characterization of 143. Zrenner R, Stitt M, Sonnewald U, Boldt R (2006) Pyrimidine the evolutionarily conserved iron–sulfur cluster of sirohydro- and purine biosynthesis and degradation in plants. Annu Rev chlorin ferrochelatase from Arabidopsis thaliana. Biochem J Plant Biol 57:805–836. h ttp s : //do i.o rg/1 0.1 146/ an nur ev .a r pl a 444:227–237. https ://doi.org/10.1042/BJ201 11993 nt.57.03290 5.10542 1 144. Hung W-F, Chen L-J, Boldt R et  al (2004) Characteriza- 127. Lindqvist Y, Huang W, Schneider G, Shanklin J (1996) Crystal tion of Arabidopsis glutamine phosphoribosyl pyrophosphate structure of delta9 stearoyl-acyl carrier protein desaturase from 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 565 2+ amidotransferase-deficient mutants. Plant Physiol 135:1314– spectroscopy identifies a [4Fe–4S] center with unusual coordi- 1323. https ://doi.org/10.1104/pp.104.04095 6 nation sphere in the LytB protein. J Am Chem Soc 131:13184– 145. Ito T, Shiraishi H, Okada K, Shimura Y (1994) Two amidophos- 13185. https ://doi.org/10.1021/ja901 2408 phoribosyltransferase genes of Arabidopsis thaliana expressed 162. Seemann M, Tse Sum Bui B, Wolff M et al (2006) Isoprenoid in different organs. Plant Mol Biol 26:529–533. https ://doi. biosynthesis in plant chloroplasts via the MEP pathway: direct org/10.1007/BF000 39565 thylakoid/ferredoxin-dependent photoreduction of GcpE/IspG. 146. van der Graaff E, Hooykaas P, Lein W et al (2004) Molecular FEBS Lett 580:1547–1552. https ://doi.or g/10.1016/j.febsl analysis of “de novo” purine biosynthesis in solanaceous species et.2006.01.082 and in Arabidopsis thaliana. Front Biosci 9:1803–1816 163. de la Luz Gutiérrez-Nava M, Gillmor CS, Jiménez LF et al (2004) 147. Lin M, Behal R, Oliver DJ (2003) Disruption of plE2, the gene Chloroplast biogenesis genes act cell and noncell autonomously for the E2 subunit of the plastid pyruvate dehydrogenase com- in early chloroplast development. Plant Physiol 135:471–482. plex, in Arabidopsis causes an early embryo lethal phenotype. https ://doi.org/10.1104/pp.103.03699 6 Plant Mol Biol 52:865–872 164. Guevara-García A, San Román C, Arroyo A et al (2005) Charac- 148. Binder S (2010) Branched-chain amino acid metabolism in terization of the Arabidopsis clb6 mutant illustrates the impor- Arabidopsis thaliana. Arabidopsis Book 8:e0137. https ://doi. tance of posttranscriptional regulation of the methyl-d -eryth- org/10.1199/tab.0137 ritol 4-phosphate pathway. Plant Cell 17:628–643. https ://doi. 149. Zhang C, Pang Q, Jiang L et al (2015) Dihydroxyacid dehy-org/10.1105/tpc.104.02886 0 dratase is important for gametophyte development and disruption 165. Hsieh M-H, Goodman HM (2005) The Arabidopsis IspH causes increased susceptibility to salinity stress in Arabidopsis. homolog is involved in the plastid nonmevalonate pathway of J Exp Bot 66:879–888. https ://doi.org/10.1093/jxb/eru44 9 isoprenoid biosynthesis. Plant Physiol 138:641–653. https://doi. 150. Flint DH, Emptage MH (1988) Dihydroxy acid dehydratase from org/10.1104/pp.104.05873 5 spinach contains a [2Fe–2S] cluster. J Biol Chem 263:3558–3564 166. Bouvier F, d’Harlingue A, Hugueney P et al (1996) Xanthophyll 151. Knill T, Reichelt M, Paetz C et  al (2009) Arabidopsis thali- biosynthesis. Cloning, expression, functional reconstitution, and ana encodes a bacterial-type heterodimeric isopropylmalate regulation of beta-cyclohexenyl carotenoid epoxidase from pep- isomerase involved in both Leu biosynthesis and the Met chain per (Capsicum annuum). J Biol Chem 271:28861–28867 elongation pathway of glucosinolate formation. Plant Mol Biol 167. Lin H, Wang R, Qian Q et al (2009) DWARF27, an iron-con- 71:227–239. https ://doi.org/10.1007/s1110 3-009-9519-5 taining protein required for the biosynthesis of strigolactones, 152. Sureshkumar S, Todesco M, Schneeberger K et  al (2009) A regulates rice tiller bud outgrowth. Plant Cell 21:1512–1525. genetic defect caused by a triplet repeat expansion in Arabidopsis https ://doi.org/10.1105/tpc.109.06598 7 thaliana. Science 323:1060–1063. https://doi.or g/10.1126/scien 168. Waters MT, Brewer PB, Bussell JD et al (2012) The Arabidop- ce.11640 14 sis ortholog of rice DWARF27 acts upstream of MAX1 in the 153. He Y, Mawhinney TP, Preuss ML et al (2009) A redox-active control of plant development by strigolactones. Plant Physiol isopropylmalate dehydrogenase functions in the biosynthesis of 159:1073–1085. https ://doi.org/10.1104/pp.112.19625 3 glucosinolates and leucine in Arabidopsis. Plant J 60:679–690. 169. Su L-W, Chang SH, Li M-Y et al (2013) Purification and bio- https ://doi.org/10.1111/j.1365-313X.2009.03990 .x chemical characterization of Arabidopsis At-NEET, an ancient 154. Vranová E, Coman D, Gruissem W (2013) Network analysis of iron–sulfur protein, reveals a conserved cleavage motif for the MVA and MEP pathways for isoprenoid synthesis. Annu Rev subcellular localization. Plant Sci 213:46–54. ht t ps : / /d oi . Plant Biol 64:665–700. https ://doi.org/10.1146/annur ev-arpla org/10.1016/j.plant sci.2013.09.001 nt-05031 2-12011 6 170. Nechushtai R, Conlan AR, Harir Y et al (2012) Characteriza- 155. Loiseau L, Gerez C, Bekker M et al (2007) ErpA, an iron–sulfur tion of Arabidopsis NEET reveals an ancient role for NEET pro- (Fe–S) protein of the A-type essential for respiratory metabo- teins in iron metabolism. Plant Cell 24:2139–2154. https ://doi. lism in Escherichia coli. PNAS 104:13626–13631. https ://doi.org/10.1105/tpc.112.09763 4 org/10.1073/pnas.07058 29104 171. Kinsman EA, Pyke KA (1998) Bundle sheath cells and cell-spe- 156. Henkes S, Sonnewald U, Badur R et al (2001) A small decrease cific plastid development in Arabidopsis leaves. Development of plastid transketolase activity in antisense tobacco transfor- 125:1815–1822 mants has dramatic effects on photosynthesis and phenylpropa- 172. Rosar C, Kanonenberg K, Nanda AM et al (2012) The leaf reticu- noid metabolism. Plant Cell 13:535–551 late mutant dov1 is impaired in the first step of purine metabo- 157. Khozaei M, Fisk S, Lawson T et al (2015) Overexpression of lism. Mol Plant 5:1227–1241. https ://doi.org/10.1093/mp/sss04 plastid transketolase in tobacco results in a thiamine auxotrophic 5 phenotype. Plant Cell 27:432–447. https ://doi.or g/10.1105/ 173. Grant K, Carey NM, Mendoza M et al (2011) Adenosine 5’-phos- tpc.114.13101 1 phosulfate reductase (APR2) mutation in Arabidopsis implicates 158. Budziszewski GJ, Lewis SP, Glover LW et al (2001) Arabidop- glutathione deficiency in selenate toxicity. Biochem J 438:325– sis genes essential for seedling viability: isolation of insertional 335. https ://doi.org/10.1042/BJ201 10025 mutants and molecular cloning. Genetics 159:1765–1778 174. Tanaka R, Hirashima M, Satoh S, Tanaka A (2003) The Arabi- 159. Estévez JM, Cantero A, Reindl A et al (2001) 1-Deoxy-d -xylu- dopsis-accelerated cell death gene ACD1 is involved in oxy- lose-5-phosphate synthase, a limiting enzyme for plastidic iso- genation of pheophorbide a: inhibition of the pheophorbide a prenoid biosynthesis in plants. J Biol Chem 276:22901–22909. oxygenase activity does not lead to the “Stay-Green” phenotype https ://doi.org/10.1074/jbc.M1008 54200 in Arabidopsis. Plant Cell Physiol 44:1266–1274. https ://doi. 160. Seemann M, Wegner P, Schünemann V et al (2005) Isoprenoid org/10.1093/pcp/pcg17 2 biosynthesis in chloroplasts via the methylerythritol phosphate 175. Yang M, Wardzala E, Johal GS, Gray J (2004) The wound- pathway: the (E)-4-hydroxy-3-methylbut-2-enyl diphosphate syn- inducible Lls1 gene from maize is an orthologue of the Arabi- thase (GcpE) from Arabidopsis thaliana is a [4Fe–4S] protein. dopsis Acd1 gene, and the LLS1 protein is present in non- J Biol Inorg Chem 10:131–137. https ://doi.org/10.1007/s0077 photosynthetic tissues. Plant Mol Biol 54:175–191. https ://doi. 5-004-0619-zorg/10.1023/B:PLAN.00000 28789 .51807 .6a 161. Seemann M, Janthawornpong K, Schweizer J et al (2009) Iso- 176. Voss I, Koelmann M, Wojtera J et al (2008) Knockout of major prenoid biosynthesis via the MEP pathway: in vivo Mössbauer leaf ferredoxin reveals new redox-regulatory adaptations in 1 3 566 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 Arabidopsis thaliana. Physiol Plant 133:584–598. https ://doi. 178. Zhao J, Qiu Z, Ruan B et al (2015) Functional inactivation of org/10.1111/j.1399-3054.2008.01112 .x putative photosynthetic electron acceptor Ferredoxin C2 (FdC2) 177. Li C, Hu Y, Huang R et al (2015) Mutation of FdC2 gene encod- induces delayed heading date and decreased photosynthetic rate ing a ferredoxin-like protein with C-terminal extension causes in rice. PLoS ONE 10:e0143361. https ://doi.org/10.1371/journ yellow-green leaf phenotype in rice. Plant Sci 238:127–134. https al.pone.01433 61 ://doi.org/10.1016/j.plant sci.2015.06.010 1 3 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png JBIC Journal of Biological Inorganic Chemistry Springer Journals
Free
22 pages

Loading next page...
 
/lp/springer_journal/roles-and-maturation-of-iron-sulfur-proteins-in-plastids-wsV6WwxVYI
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Biochemistry, general; Microbiology
ISSN
0949-8257
eISSN
1432-1327
D.O.I.
10.1007/s00775-018-1532-1
Publisher site
See Article on Publisher Site

Abstract

One reason why iron is an essential element for most organisms is its presence in prosthetic groups such as hemes or iron– sulfur (Fe–S) clusters, which are notably required for electron transfer reactions. As an organelle with an intense metabolism in plants, chloroplast relies on many Fe–S proteins. This includes those present in the electron transfer chain which will be, in fact, essential for most other metabolic processes occurring in chloroplasts, e.g., carbon fixation, nitrogen and sulfur assimilation, pigment, amino acid, and vitamin biosynthetic pathways to cite only a few examples. The maturation of these Fe–S proteins requires a complex and specific machinery named SUF (sulfur mobilisation). The assembly process can be split in two major steps, (1) the de novo assembly on scaffold proteins which requires ATP, iron and sulfur atoms, electrons, and thus the concerted action of several proteins forming early acting assembly complexes, and (2) the transfer of the pre- formed Fe–S cluster to client proteins using a set of late-acting maturation factors. Similar machineries, having in common these basic principles, are present in the cytosol and in mitochondria. This review focuses on the currently known molecular details concerning the assembly and roles of Fe–S proteins in plastids. Keywords Biogenesis · Iron–sulfur proteins · Plastids · Electron transfer · Photosynthesis Introduction numerous metabolic pathways occurring totally or partially in this organelle are directly or indirectly dependent on the Iron (Fe) and sulfur are critical elements for plant growth functioning of Fe–S proteins. This review is organized in and development. Sulfur is notably required for cysteine three parts, describing how Fe and sulfur species get reduced and methionine synthesis, and is present in a large number and imported in chloroplasts, how the various types of Fe–S of molecules, whereas Fe atoms are associated with many clusters are built from Fe and cysteine and incorporated into proteins as part of hemes, mono- or di-iron non-heme cent- the tenths of client proteins, and finally which chloroplastic ers, or iron–sulfur (Fe–S) clusters. Chloroplasts and plastids pathways/processes are dependent on these cofactors. in general, are highly demanding organelles for both ele- ments due notably to the presence of a translation machinery and of the photosynthetic electron transfer chain. Besides, Supply of iron and sulfur to plastids Iron transport The original version of this article was revised due to a retrospective Open Access order. Chloroplasts, where photosynthesis and heme synthe- Jonathan Przybyla-Toscano and Mélanie Roland have equally sis occur, represent the major subcellular Fe sink in plant contributed to the work. leaves [1]. Photosynthetic organisms uptake Fe from the soil using a sophisticated pumping system that differs between * Nicolas Rouhier nicolas.rouhier@univ-lorraine.fr Poaceae and dicotyledon species, which developed, respec- tively, either a phytosiderophore-dependent chelation-based Université de Lorraine, Interactions Arbres- strategy or a reduction-based strategy (see [2] for an over- Microorganismes, UMR1136, 54500 Vandoeuvre-lès-Nancy, view). In Arabidopsis thaliana, Fe is acquired in several France steps. By extruding protons via the H -ATPase AHA2 and Biochimie et Physiologie Moléculaire des Plantes, coumarins via the PDR9 ABC transporter, A. thaliana can CNRS/INRA/Université Montpellier 2, SupAgro Campus, 34060 Montpellier, France Vol.:(0123456789) 1 3 546 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 3+ solubilize and chelate F e forms by lowering the soil pH. overexpression lines for this gene show abnormal chloro- 3+ 2+ Then, the reduction of Fe to Fe is performed by the fer- plast development and perturbed iron homeostasis and avail- ric reductase-oxidase (FRO) family protein, FRO2, before ability [6, 7]. The loss-of-function mutants are dwarf and its uptake by the plasma membrane Fe transporter named chlorotic (even white), and they grow only heterotrophically. iron-regulated transporter 1 (IRT1) [2]. In the cytosol of root Moreover, they accumulate Fe into ferritins, the function of cells, Fe complexes are formed with organic acids (malate which is normally to protect this organelle from oxidative or citrate) or nicotianamine before being translocated to stress by sequestering Fe. The PIC1-overexpressing plants the shoots and unloaded in the cytosol of mesophyll cells suffer from oxidative stress and leaf chlorosis likely due to [3]. After this step, little is known concerning Fe acquisi- a Fe overload in chloroplasts. Although this permease is tion by chloroplasts, its subsequent storage, and delivery mentioned to be part of the translocase of the outer/inner to dedicated proteins and machineries. It is possible that chloroplast membrane (Tic–Toc) complex in other studies, 3+ a voltage-dependent transport system allows F e -citrate a Fe transport function is clear from the complementation complexes to pass the outer membrane of the plastid enve- of a yeast fer3fer4 mutant which is defective in Fe uptake, lope [4]. Once in the chloroplastic intermembrane space, leading to the conclusion that PIC1 may have a dual function 3+ 2+ FRO7 may reduce ferric (F e ) to ferrous iron (F e ) via [6]. Another putative Fe transporter, named NAP14 (non- its reductase activity [5]. Several transporters located in the intrinsic ABC protein 14), was identified from its homology inner membrane of the chloroplast envelope are candidates with the ABC transporter FutC belonging to the FutABC for Fe import into the stroma (Fig. 1). The first one is named iron uptake system in cyanobacteria [8]. As observed for permease in chloroplast 1 (PIC1) [6]. Both knock-out and pic1, a nap14 knock-out mutant accumulates Fe in shoots, PSI Target Apoproteins SIR FTR HCF101 SUFA1 NFU1 IBA57.2 NFU3 NFU2 3+ BOLA4 Fe Heme and GRXS14 GRXS16 2+ Fe FRO7 siroheme SUFB SUFD BOLA1 3+ biosynthesis Fe SUFC SUFC ATP ADP+FAD ADP FADH Iron-related proteins FH SUFB SUFD Cysteine desulfurase SUFC SUFC ATP SufE-like proteins Scaffold proteins alanine + SUFE1-3 Transfer proteins cysteine Target proteins NFS2 SUFE1-3 Fig. 1 Working model for iron uptake and maturation of Fe–S pro- associated with each protein function is indicated directly on the fig- teins by the SUF machinery in plastids of eukaryotic photosynthetic ure. The detailed description of the maturation process and the con- organisms. Besides the putative Fe transporters located at the mem- nections between the SUF proteins are described in the text. Except brane of the chloroplast envelope, which would serve for providing NFU2, NFU3, and HFC101, all maturation factors have been grouped the required Fe atoms to the SUF machinery, this scheme integrates in a blue circle in the absence of information concerning their precise the 17 putative SUF components. In the absence of stronger evidence function, but two-way arrows indicate that physical interactions have concerning the implication of frataxin, it is not integrated among SUF been observed between some proteins components and is represented by a dashed circle. The color code 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 547 exhibits abnormal chloroplast structures, and shows deregu- (SUF) machinery is autonomous, the cytosolic iron–sulfur lated levels of Fe homeostasis-related genes. However, in the assembly (CIA) machinery is dependent on the mitochon- absence of other Fut orthologs in A. thaliana, the question of drial iron–sulfur cluster (ISC) machinery as it relies on a whether NAP14 can work alone or in pair with other uniden- sulfur-containing compound generated in the first steps and tified partners remains open. A third candidate transporter exported from mitochondria by an ABC transporter. We for Fe uptake in chloroplasts is mitoferrin-like1 (MFL1) [9]. invite the readers interested in the ISC and CIA machineries However, although its gene expression is dependent on Fe in plants to refer to the following recent reviews [12, 13]. For supply and the protein is in principle located to the inner all these machineries and in particular the chloroplastic SUF membrane of the chloroplast envelope, the growth of knock- machinery, the biosynthesis and delivery of Fe–S clusters out mutants is only moderately affected. While all these pro - can be separated in two major steps: their de novo assembly teins seem to be involved in Fe homeostasis in chloroplasts, on scaffold proteins and their incorporation into final client further characterization is urgently needed to clarify their proteins. This second step may necessitate the exchange and exact function and respective importance. possibly conversion of Fe–S clusters between scaffold pro- teins and maturation factors including Fe–S cluster transfer Sulfate import and reduction in plastids proteins and targeting/recruiting factors. Repair mechanisms for the synthesis of cysteine, the sulfur donor may eventually account for the recycling of damaged Fe–S of Fe–S clusters clusters, which could be important in chloroplasts consider- ing the presence of reactive oxygen and nitrogen species, but Photosynthetic organisms use sulfate present in the soils as this will not be discussed further as information in plants is a primary source of sulfur. Sulfate is incorporated into the very scarce. roots through an active proton/sulfate co-transport system located at the plasma membrane [10]. Once in the xylem, The de novo Fe–S cluster assembly on scaffold sulfate is transported to the shoots, unloaded into the cytosol protein of mesophyll cells, and then transported in the chloroplasts for its ATP-dependent reductive assimilation into sulfide In chloroplasts, it seems now clear that the sole scaffold (see [10] for review). The involved transporters all along system is formed by the SUFBCD proteins (Fig. 1) [14]. these steps belong to the sulfate transporter (SULTR) family, The assembly of a Fe–S cluster on this scaffold complex which is composed of 12 members in A. thaliana that can be theoretically requires the concerted action of several proteins grouped into four classes. The SULTR3 class comprises the as it requires the polypeptide backbones, ATP, Fe, and sul- chloroplast-localized sulfate transporters [11]. The sulfide fur atoms and electrons. There are still many uncertainties generated by the ferredoxin (FDX)-dependent sulfite reduc- about the involved actors in plants and the molecular details. tase (SIR) will be used for cysteine biosynthesis by cysteine Thus, we will often make analogies to the Escherichia coli synthase, a complex of two enzymes, serine acetyltrans- SUF system, which has been better characterized. The best, ferase (SAT) that uses acetyl-coA to form O-acetylserine not to say the only, well-characterized actors in plants of (OAS) from serine and O-acetylserine-(thiol)-lyase (OAS- this assembly complex are proteins required for the produc- TL) which can substitute the acetyl moiety by sulfide to form tion and transfer of the required sulfur. The NFS2 protein cysteine. While the first steps of sulfate reduction into sulfide (formerly referred to as CpNifS) is a pyridoxal-l -phosphate are clearly restricted to the chloroplasts, cysteine synthesis (PLP)-dependent cysteine desulfurase, which catalyzes the can also occur in the cytosol and in mitochondria owing to extraction of the sulfur atoms from cysteine, producing a the ubiquitous expression of SAT and OAS-TL and exchange persulfide group on a catalytic cysteine with the concomitant of sulfide across organelle membranes [10]. release of an alanine (Fig. 1) [15]. As a class II cysteine des- ulfurase, similar to the bacterial SufS orthologs, the acces- sibility of the persulfide group is limited by the presence The biogenesis of Fe–S proteins of a β-hairpin near the catalytic cysteine [16, 17]. For this in chloroplasts by the SUF machinery reason, the transfer of sulfur atoms to the scaffold complex relies on an additional protein named SUFE. In A. thaliana, Several dozen of proteins containing Fe–S clusters are found there are three SUFE proteins (SUFE1-3) targeted to chlo- in various subcellular compartments in the model plant A. roplasts [18]. In addition to the SUFE domain, SUFE1 has a thaliana as in other plants. Accordingly, in plant cells, three C-terminal BOLA domain, the role of which is unknown but assembly machineries exist in plastids, in mitochondria, may preg fi ure a control by glutaredoxins (GRXs, see below) and in the cytosol, the latter being dedicated to the mat- and SUFE3 possesses a quinolinate synthase (NadA) domain uration of Fe–S proteins found both in the cytosol and in at the C-terminus, which is involved in NAD biosynthesis the nucleus. Whereas the chloroplastic sulfur mobilisation [18, 19]. As shown for the corresponding E. coli couple 1 3 548 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 [20], each SUFE protein enhances the cysteine desulfurase it is very likely that the SufBC D scaffold binds a [Fe S ] 2 4 4 activity of NFS2 by accepting the persulfide group on its cluster in vivo and considering the conservation between own catalytic cysteine, thus serving as a relay to the scaffold A. thaliana and E. coli sequences, we anticipate that this system [18, 19]. At the structural level, A. thaliana NFS2 is mechanism should also prevail for plant proteins. However, a dimeric protein with two distant active sites, which sug- we cannot completely rule out that SufBC D or other forms, gests that the functional NFS2-SUFE unit should be a het- such as the SufB C form detected with E. coli proteins 2 2 erotetramer [17]. In addition to the existence of additional [29], can bind other cluster types in some conditions. For domains in SUFE1 and SUFE3, the existence of three SUFE instance, transcriptomic data indicate that the SUFB, SUFC, isoforms may be also linked to their expression pattern as and SUFD genes may not be co-expressed in all organs and for instance SUFE2 is mostly expressed in flowers [18]. The cell types of A. thaliana. central role of these proteins has been validated by genetic At this stage of the assembly process, there are many studies, since the study of knock-out A. thaliana lines proved other crucial questions concerning the source of electrons 3+ 2+ that NFS2, SUFE1, SUFE3, and SUFBCD genes are essen- required for the reduction of Fe to Fe or of the persulfide 0 2− tial [14, 18, 21, 22]. The use of RNAi lines showed that (S ) to a sulfide (S ), the source of Fe, and the control of NFS2 and SUFBCD are required for the maturation of all its entry in the complex. In this respect, it is important to plastidial Fe–S proteins tested so far [14, 22]. note that the SufBC D complex was purified with a bound In E. coli as in A. thaliana, the scaffold complex is prob- reduced flavin-adenine dinucleotide (FADH ) molecule [29, ably composed by three subunits, SUFB, SUFC, and SUFD, 30]. While SufB alone can bind the flavin in vitro [30], SufD very likely in a 1:2:1 stoichiometry and will be referred to is also required in vivo [29]. It is currently believed that as SUFBC D (Fig. 1) [14, 20]. It seems that NFS2, SUFEs, this FADH provides the necessary reducing equivalents for 2 2 and SUFBCD do not form a large and stable complex as the reduction of ferric iron. Since FAD is released from the recently shown in the case of the mitochondrial ISC system complex upon oxidation, an external regeneration system in yeast and human [23, 24]. Indeed, some in vitro biochemi- is needed, which could be possibly an FDX or an NADPH- cal analyses using the bacterial SufS, SufE, and SufBCD dependent flavin reductase. enzymes indicated that SufS does not seem to make stable The mechanisms and actors involved in the delivery of interactions with SufBCD, unlike SufE whose presence is Fe for Fe–S cluster biosynthesis in plastids are completely absolutely required for an efficient Fe–S cluster reconsti- unknown. The Fe–Storage proteins, ferritins, have been tution in vitro on SufBCD [20, 25]. Besides, it has been excluded from Fe donor candidates, because an Arabidopsis shown that the presence of SufC, but not SufD, is required mutant (fer1-3-4) for the three ferritins found in leaves has for the transfer of the sulfur atoms bound to E. coli SufE to no apparent phenotype [31], while mutant plants modified SufB. Upon ATP binding, the SufC ATPase would induce for the expression of these early biogenesis factors are either structural changes on SufB and SufD that are necessary for lethal or at least strongly affected. Another candidate for Fe Fe–S cluster binding [26]. Some residues important for these delivery is a small acidic protein with iron-binding proper- interactions have been identified from the 3D structures and ties named frataxin. In the mitochondrial ISC machinery, validated by mutagenesis [26, 27]. Among the numerous frataxin controls iron entry in the assembly complex by acti- cysteines present in SufB, the primary sulfur acceptor would vating sulfide formation by the cysteine desulfurase [32, 33]. be the conserved Cys254 (E. coli numbering). This sulfur In this complex, frataxin can interact both with the cysteine atom would then be transferred to Cys405, one of the Fe–S desulfurase and the ISCU scaffold protein. Except for a few cluster ligands owing to the existence of a tunnel inside the organisms like Z. mays, there is usually a single gene cod- β-helix core domain of SufB [27]. The question of which ing for frataxin (FH) in plants. While frataxin was believed type of Fe–S clusters is bound to this complex has been for a long time to be exclusively located in mitochondria, it investigated in detail. It was shown that E. coli SufB alone was recently reported that A. thaliana FH (AtFH) and two can assemble both [Fe S ] and [Fe S ] clusters in vitro and isoforms from Z. mays may have a dual targeting into both 2 2 4 4 that a conversion from the [Fe S ]-loaded SufB form to a mitochondria and plastids [34, 35]. According to this pos- 4 4 stable [Fe S ]-loaded form is possible upon exposure to air sible chloroplastic localization, Arabidopsis FH-deficient 2 2 [25, 28]. However, based on the structure of an apoSufBCD plants show a decrease in the heme content [36]. Moreover, complex, it was proposed that a histidine of SufD may be they present a decrease in the total chlorophyll content, in a Fe–S cluster ligand [26]. Consistently, a mutated variant the levels of two plastidial FDXs and in nitrite reductase for this histidine lost the ability to assemble a Fe–S clus- (NIR, a siroheme-containing enzyme) activity which could ter in vivo, and both SufC and SufD were required for the explain the observed changes in the rate of the photosyn- in vivo maturation of SufB [29]. In this cellular context, E. thetic electron transport chain [35]. The impact on heme coli SufBC D complex mostly binds a [F e S ] cluster with content would be in good agreement with the described 2 4 4 some residual amount of linear [F e S ] clusters [29]. Hence, interaction between yeast frataxin and ferrochelatase, the 3 4 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 549 terminal enzyme of heme synthesis performing porphyrin plants overexpressing GRXS14 have a decreased chlorophyll metalation [37]. All these observations suggest an impair- content [44]. Considering that several enzymes involved in ment of the plastidial Fe–S cluster biosynthesis and/or of the chlorophyll catabolism require Fe–S clusters, this may con- heme or siroheme biosynthesis, although stronger and more stitute a first hint towards a role of GRXS14 in the matu- direct biochemical evidence is still required. ration of specific client proteins in this pathway. Counter - intuitive to this first observation, plants lacking GRXS14 Delivery and trafficking of preformed Fe–S clusters showed accelerated chlorophyll loss compared to wild-type by maturation factors plants when exposed to prolonged darkness, suggesting more complex connections [44]. A redundancy may exist between The preformed Fe–S cluster on the SUFBC D complex, be both plastidial GRXs, since a double mutant with about 20% it a [Fe S ] or a [Fe S ] cluster, has then to be correctly GRXS16 remaining exhibits a 20% biomass reduction in 2 2 4 4 targeted to client apoproteins. This requires several other standard conditions compared to wild-type plants. How- proteins referred to as maturation factors. Among these, one ever, this phenotype is not exacerbated under stress condi- could differentiate the so-called Fe–S cluster transfer/carrier tions. Overall, unlike the knock-out mutant of mitochondrial proteins (belonging to NFU, SUFA, GRX, and HCF101 fam- GRXS15 which is embryo-lethal [45], these results point ilies) from targeting factors (belonging to BOLA and IBA57 either to non-essential roles of these isoforms or to a redun- families) which, contrary to the proteins of the first group, dant function with the remaining GRXS16 level being suf- are not able to bind Fe–S clusters by themselves, although ficient to sustain an essential role similar to GRXS15. BOLAs do it in complex with GRXs [38, 39]. It is interest- Concerning BOLA proteins, several roles have been pro- ing to note that all proteins of these families have mitochon- posed, but only those connected to their participation in Fe drial counterparts in the ISC machinery, whereas the compo- metabolism have been really validated [46]. Their involve- nents forming the eukaryote-specific CIA machinery usually ment in Fe–S cluster biogenesis was demonstrated from belong to different protein families [ 12]. This analogy to the study of bol1/3 mutant in yeast and of human patients the mitochondrial system is the reason why some of these defective for the mitochondrial BOLA3. Both types of cells plastidial members, whose role in the maturation of Fe–S display protein lipoylation defects due to the incorrect matu- proteins in plastid has not been yet established, have been ration of lipoate synthase and a decrease in activity for some included in this section. The current model for these steps in other [Fe S ] proteins as aconitase and succinate dehydro- 4 4 the plant mitochondrial ISC machinery derives mainly from genase [47–49], whereas human patients also have defects studies conducted in yeast and human and can be summa- in the mitochondrial respiratory complexes I and III [49]. rized as follows [40]. A glutaredoxin (GRXS15 in plants) is Three isoforms with a BOLA domain are found in plant the primary transfer protein receiving a [Fe S ] cluster from chloroplasts. As already mentioned, the C-terminal region 2 2 ISCU proteins. This cluster can be either directly inserted of SUFE1 contains a BOLA domain. The two other iso- into [Fe S ]-recipient apoproteins or used to build [Fe S ] forms, BOLA1 and BOLA4, comprise a single domain. Both 2 2 4 4 clusters on a heterocomplex formed by ISCAs and possibly BOLA4 and SUFE1 could also be targeted to mitochondria IBA57. Some mechanistic and structural aspects of the clus- [21, 50]. Interactions between these plastidial BOLA pro- ter conversion from the [Fe S ]-loaded GLRX5 form to the teins and GRXS14 and GRXS16 have been demonstrated 2 2 [Fe S ]-loaded ISCA1-2 form have been recently delineated both in vitro and in planta [38, 50]. These proteins can in fact 4 4 using human proteins [41, 42]. Then, the insertion of the form both apo-heterodimers and holo-heterodimers bridging [Fe S ] clusters into client Fe–S proteins might be direct or a [Fe S ] cluster [39], as also demonstrated for bacterial, 4 4 2 2 facilitated by NFU and BOLA proteins that likely act in con- yeast, and mammalian isoforms [46]. In this respect, it is cert for the maturation of specific targets notably the lipoate interesting to note that adding BOLA to a GRX homodi- synthase or by IND1/INDH, a close HCF101 homolog, mer converts it to a more stable holo GRX-BOLA heter- which seems specific for the respiratory chain complex I. odimer. This interconversion might represent a regulatory The current genetic and biochemical evidence indicate mechanism either to shut down or activate some specific that this sequence of events should be very different for the pathways by favouring one target over another. At the struc- plastidial SUF machinery (Fig. 1). Although the two plastid- tural level, all BOLA isoforms have a similar well-conserved ial isoforms, named GRXS14 and GRXS16, have the ability fold [39]. Two subgroups can, however, be distinguished to bind the regular [F e S ] cluster in homodimer (or in heter- based on the length of the β1–β2 loop referred to as the 2 2 odimer with BOLA, see below) and to complement a yeast variable [C/H] loop, because it contains one of the ligands mutant for the mitochondrial Grx5 [43], strong genetic and provided by BOLA either a cysteine or a histidine, the sec- physiological evidence for a similar involvement in plants is ond ligand being a totally conserved histidine found in the still missing. Single mutants for each of these genes have no α3–β3 loop [39, 48]. Other cysteine ligands are provided phenotype when grown under standard conditions, whereas by a glutathione molecule and by the one present in the 1 3 550 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 conserved CGFS signature of the GRX partner, as in regular two orthologs found in A. thaliana, IBA57.1 and IBA57.2, GRX homodimers [51]. While there is no true ortholog of are, respectively, localized in mitochondria and plastids [61]. yeast Bol3 in plants, the observation that Bol3 might interact It is interesting to note that both isoforms can complement with Nfu1 rather than with Grx5 in yeast could point to a the growth defects of an E. coli ygfZ mutant observed on a different role in the late steps of the mitochondrial system minimal medium or upon oxidative stress [60]. This is the [47, 48]. Although single bol1 and bol3 mutants do not have only physiological information obtained so far for these plant phenotypes and the respective molecular roles of Bol1 and isoforms, since an Arabidopsis iba57.1 mutant is embryo- Bol3 are still unclear, a connection between Bol3 and Nfu1 lethal and an iba57.2 mutant has not been described. While is also evident from the quite similar phenotype of the bol1/3 the exact function of IBA57 is still unknown, it is important and nfu1 mutant cells [47]. to note that there is a conserved cysteine residue in a KGCY- In mitochondria, ISCA proteins are central for the matu- x-GQE-x3-R/K motif, which is almost the only conserved ration of [F e S ] proteins, presumably ensuring the con- motif in this protein family [62]. Moreover, consistent with 4 4 version of [Fe S ] centers into [Fe S ] centers. In bacteria, the structural similarity of IBA57 with folate-dependent 2 2 4 4 the different A-type isoforms (IscA, SufA, ErpA) are also enzymes [63], E. coli YgfZ can bind tetrahydrofolate [60]. required for the maturation of [F e S ] proteins, even though Another category of proteins strictly required for the mat- 4 4 in vitro studies demonstrated that Azotobacter vinelandii uration of [Fe S ] clusters is the NFU family that exists in 4 4 IscA, for example, can reversibly cycle between [Fe S ] all kingdoms. In mitochondria, the study of the yeast mutant 2 2 and [Fe S ] forms through electron reductive coupling or and several human patients indicates that NFU1 is required 4 4 oxidative cleavage [52]. Some biochemical redundancy for the maturation of lipoate synthase, which affects several seems to exist between them as demonstrated for the Fe–S ketoacid dehydrogenases dependent on lipoic acid, and for cluster assembly of IspG and IspH, two enzymes involved the maturation of complexes I, II or III depending on the in isoprenoid synthesis and also present in plant chloroplasts patients [49, 64]. A. thaliana encodes five NFU isoforms, [53, 54]. In plastids, the only representative of this family two (NFU4 and NFU5) should be targeted to mitochondria, should be SUFA1, also referred previously to as CpISCA and three (NFU1, NFU2, and NFU3) are localized in chloro- and ISCA-I [55, 56]. As an A-type carrier protein, SUFA1 plasts [65, 66]. All these proteins share an NFU domain pos- possesses the three characteristic conserved cysteines [54, sessing a CXXC motif necessary for the binding of a [F e S ] 4 4 55] that allow the binding of a [F e S ] center in a dimer as in a dimer [67]. Chloroplastic isoforms have an additional 2 2 observed upon in vitro Fe–S cluster reconstitution assays NFU domain in the C-terminal extremity which does not [55–57]. According to the ISC model, Fe–S cluster transfer have the cysteine residues, whereas mitochondrial isoforms experiments showed that GRXS14 can efficiently and uni - have an additional N-terminal domain of unknown function directionally transfer its [Fe S ] cluster to SUFA1; however, (Fig. 2) [65, 68]. Loss-of-function nfu2 and nfu3 mutants 2 2 there was no sign of a [F e S ] cluster formation [57]. Using have a dwarf phenotype with pale green leaves [69, 70]. 4 4 recombinant proteins, it was shown in vitro that an apo-SufA By coupling chlorophyll fluorescence and P700 absorption from E. coli could promote the maturation of an apo-FDX measurements to western blot analyses, it was shown that from a [F e S]-loaded SufBC D scaffold, indicating that this phenotype is due to the impairment of photosystem I 4 4 2 SUFA proteins would directly interact with the scaffold but (PSI) architecture and activity which is explained by a defect also that it facilitates Fe–S cluster conversion. Nevertheless, in the maturation of the three [F e S ] clusters assembled in 4 4 knock-out mutants have no visible phenotype when grown the psaA, psaB, and psaC subunits. The only other nota- under standard conditions, indicating that the role of SUFA1 ble and robust molecular default observed is that the SIR is dispensable [55, 56]. Whether it is involved in the matu- level and activity are decreased in nfu2 [14, 65, 70, 71]. The ration of [F e S ] proteins, [Fe S ] proteins or both remains fact that a double nfu2-nfu3 mutant is lethal [69] indicates 2 2 4 4 thus to be determined. that both NFU isoforms should have partially overlapping It is getting clear that, in yeast and human mitochondria, functions. This raises also the question of their contribu- ISCA proteins interact with IBA57 (Iron–Sulphur cluster tion relatively to the high chlorophyll fluorescence 101 assembly factor for Biotin synthase- and Aconitase-like (HCF101) protein, a plastidial 51 kDa protein belonging to mitochondrial proteins with a mass of 57 kDa). They form the NTPase protein family (Fig. 2). The hcf101 Arabidop- a complex involved in the maturation of several [F e S ] pro- sis mutant plants have globally similar molecular defects, 4 4 teins including radical-S-adenosylmethionine (SAM) pro- although this is exacerbated as the strongest allele is lethal teins, homoaconitase, aconitase, biotin synthase, and lipoic at the seedling stage and the decrease in the amounts of PSI acid synthase [58, 59]. Depletion of the E. coli ortholog subunits is stronger, almost complete [72–74]. Besides, there YgfZ also affects some [Fe S ] proteins such as succinate is a decrease in the ferredoxin-thioredoxin reductase (FTR) 4 4 dehydrogenase, fumarase, dimethylsulfoxide reductase, and levels, another [F e S ] protein [72]. Overall, in accordance 4 4 MiaB, an enzyme involved in tRNA thiolation [58–60]. The with the capacity of Arabidopsis NFU2 and HCF101 to bind 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 551 NFS2 463 AA SUFB 557 AA SUFC 338 AA SUFD 475 AA 180 AA SUFA1 CXC IBA57.2 432 AA KGCY-X-GQE HCF101 532 AA DUF59 DUF971 NFU1 NFU2 231/235/236 AA CXXC NFU3 Domain of unknown funcƒon SUFE1 371 AA Monothiol glutaredoxin domain SUFE2 258 AA Endonuclease domain SUFE3 718 AA SUFE domain BOLA domain BOLA1 160/177 AA NFU domain BOLA4 Degenerated NFU domain GRXS14 173 AA CGFS P-loop NTPase moƒf GRXS16 293 AA YgfZ signature CGFS Fig. 2 Protein domain organization of SUF components. The with the length in amino acids of the Arabidopsis proteins indicated. domains (identified using pfam or the NCBI conserved domain tools) The Fe–S binding cysteine and histidine residues are represented in present in SUF components have been represented using the color yellow and black, respectively, while other conserved cysteines are in code defined on the figure. Except for the chloroplastic targeting orange, although their function is sometimes unclear if any sequence (light green boxes), the domains are represented at scale, [Fe S ] cluster in vitro [67, 73], this indicates that all these and two histidines, is found in the Rieske protein of the 4 4 proteins are required for the maturation of [F e S ] proteins, cytochrome b f complex. In the genome of eukaryotes and 4 4 6 particularly PSI subunits, and that HCF101 would act down- in some cyanobacteria, the Rieske protein is encoded by stream of NFU2 and NFU3 (Fig. 1). a single gene named photosynthetic electron transfer C In summary, there are currently ten putative maturation (petC), whereas in most cyanobacteria, there are additional factors in the SUF machinery for several dozens of plastidial isoforms whose physiological function is still uncertain client proteins. The role of some of these maturation factors [75]. The absence of the petC proteins is lethal in the early still awaits conr fi mation not to speak about their connections developmental stages both in A. thaliana and cyanobacte- and hierarchical organization. There is also an urgent need ria (Table 1) [75, 76]. Three low potential [Fe S ] clusters 4 4 to learn more about how specificity towards target proteins are attached to the thylakoid membrane but face the reduc- is achieved and about the molecular and structural aspects ing, stromal side of PSI, and function in series. The first, of these interactions. referred to as F , is associated with a PsaA–PsaB heter- odimer via cysteine residues, while the two others, named F and F , are bound to PsaC [77]. These clusters transfer A B Functional diversity among client Fe–S electrons to FDXs, small soluble proteins, which contain proteins in plastids a classical [Fe S ] cluster, e.g., rhombic cluster ligated by 2 2 four cysteines. The nuclear genome of algae and plants har- Fe–S clusters in the functioning and protection bours a variable number of FDX homologs, differentially of the photosynthetic electron transport chain expressed in plant organs or at different development stages or in response to different stimuli [78]. The A. thaliana Among other functions, Fe–S clusters have a crucial role genome contains four genes encoding four well-described in electron transfer reactions, and thus, several Fe–S pro- plastidial FDXs (Fd1 to Fd4) and at least two additional teins are found in the thylakoid membrane as part of the genes, referred to as FdC1 and FdC2, encoding proteins photosynthetic electron transport chain. A Rieske-type bearing C-terminal extensions and whose functions remain Fe–S cluster, i.e., a [Fe S ] cluster ligated by two cysteines elusive [79, 80]. In specific physiological situations such as 2 2 1 3 552 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 1 3 Table 1 Phenotypes of A. thaliana mutant lines for plastidial Fe–S proteins Short Name AGI number Cluster type Type of mutants Mutant phenotypes References DHAD At3g23940 [Fe S ] Knock-out Embryo-lethal Zhang et al. [149] 2 2 Knock-down Shorter root, hypersensitive to salt stress IPMI (LSU1) At4g13430 [Fe S ] Knock-down Pleiotropic growth abnormalities Sureshkumar et al. [152], Knill et al. [151] 4 4 DWARF27.1 At1g03055 [Fe S ] Knock-down Increase in axillary rosette branches Waters et al. [168] 4 4 DWARF27.2 At1g64680 [Fe S ] Not yet described 4 4 DWARF27.3 At4g01995 [Fe S ] Not yet described 4 4 ISPG At5g60600 [Fe S ] Knock-out Albino phenotype, proplastid growth and thylakoid Gutiérrez-Nava et al. [163] 4 4 membrane formation affected ISPH At4g34350 [Fe S ] Knock-out Albino phenotype, proplastid growth and thylakoid Gutiérrez-Nava et al. [163], Hsieh and Hsieh [165], 4 4 membrane formation affected Guevara-García et al. [164] THIC At2g296302x [Fe S ] Knock-down Lethal (development arrested at the cotyledon stage Raschke et al. [140], Kong et al. [142] 4 4 with chlorotic phenotype) NIR At2g15620 [Fe S ], siroheme X-ray mutagenesis Lethal in barley unless a nitrogen source is provided Duncanson et al. [120] 4 4 SIR At5g04590 [Fe S ], siroheme Knock-out Lethal Khan et al. [119] 4 4 Knock-down Early seedling lethal Khan et al. [119] SIRB At1g50170 [Fe S ] Knock-out Seedling lethal (post-germination arrest) Saha et al. [126] 2 2 ATase1 At2g16570 [Fe S ] Knock-out No phenotype Hung et al. [144] 4 4 ATase2 At4g34740 [Fe S ] X-ray mutagenesis Small and albino/pale reticulated leaves, cell division Kinsman and Pyke [171], Hung et al. [144], van den 4 4 affected Graaf et al. [146], Rosar et al. [172] ATase3 At4g38880 [Fe S ] Not yet described 4 4 APR1 At4g04610 [Fe S ] Not yet described 4 4 APR2 At1g62180 [Fe S ] Knock-out None but increased sensitivity to selenate tolerance Grant et al. [173] 4 4 APR3 At4g21990 [Fe S ] Not yet described 4 4 cLIP1 At5g084152x [Fe S ] Not yet described 4 4 GLT1 At5g53460 [Fe S ] Knock-out No phenotype but decreased chlorophyll content, Lancien et al. [123] 3 4 growth defect under low CO GLU1 At5g04140 [Fe S ] Knock-down Dwarf photorespiratory phenotype Somerville and Ogren [125], Coschigano et al. [122], 3 4 Lancien et al. [123] GLU2 At2g41220 [Fe S ] Knock-down No phenotype Potel et al. [124] 3 4 DjC17 At5g23240 [Fe S ] Knock-out Defective root hairs Petti et al. [91] 4 4 DjC18 At2g42750 [Fe S ] Not yet described 4 4 ndhI AtCg010902x [Fe S ] Not yet described 4 4 ndhK AtCg00430 [Fe S ] Not yet described 4 4 petC At4g03280rieske [Fe S ] Knock-out Seedling lethal Maiwald et al. [76] 2 2 psaA AtCg00350 [Fe S ] with psaB Not yet described 4 4 psaB AtCg00340 [Fe S ] with psaA Not yet described 4 4 psaC AtCg010602x [Fe S ] Not yet described 4 4 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 553 1 3 Table 1 (continued) Short Name AGI number Cluster type Type of mutants Mutant phenotypes References TIC55 At2g24820rieske [Fe S ] Knock-out No phenotype Boij et al. [106], Hauenstein et al. [104] 2 2 Knock-down No phenotype Tanaka et al. [174] PAO (ACD1) At3g44880rieske [Fe S ] Knock-out Age- and light-dependent cell death phenotype in Pružinská et al. [114] 2 2 leaves and flowers. Stay-green phenotype in the dark rieske [Fe S ] Knock-down Light-dependent lesion mimic phenotype, increased Greenberg and Ausubel [112], Yang et al. [175] 2 2 sensitivity to biotic and mechanic stresses PTC52 (ACD1-like) At4g25650rieske [Fe S ] Knock-out No phenotype Boij et al. [106] 2 2 CMO At4g29890rieske [Fe S ] Not yet described 2 2 CAO At1g44446rieske [Fe S ] X-ray mutagenesis Pale green phenotype with no Chl b, highly photo- Espineda et al. [103], Ramel et al. [107] 2 2 sensitive HCAR At1g046202x [Fe S ] Knock-out No phenotype, stay-green mutant upon dark exposure Meguro et al. [108] 4 4 NEET At5g51720 Neet-[Fe S ] Knock-down Late bolting, early senescence Nechushtai et al. [170] 2 2 SUFE3 At5g50210 [Fe S ] Knock-out Lethal Murthy et al. [18] 4 4 Fd1 At1g10960 [Fe S ] Knock-down Enhanced linear electron flow Hanke and Hase [72] 2 2 Fd2 At1g60950 [Fe S ] Knock-out Growth arrest and inactivation of photosynthesis Voss et al. [176] 2 2 Knock-down Lower biomass accumulation and retarded linear Hanke and Hase [78] electron flow Fd3 At2g27510 [Fe S ] Knock-down Photoinhibition, with a reduction in maximum PSII Hanke and Hase [78] 2 2 yield following dark adaptation Fd4 At5g10000 [Fe S ] Not yet described 2 2 FdC1 At1g32550 [Fe S ] Not yet described 2 2 FdC2 At4g14890 [Fe S ] EMS mutagenesis Yellow–green leaf phenotype in rice Li et al. [177]; Zhao et al. [178] 2 2 FTR At2g04700 [Fe S ] Knock-out Lethal Wang et al. [97] 4 4 Virus-induced silencing Chlorosis, abnormal chloroplast development Wang et al. [97] All described phenotypes come from studies performed with A. thaliana unless otherwise stated in the “mutant phenotype” column 554 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 environmental constraints, FDXs can recycle electrons to Based on the fact that HSP70B plays a role in the repair and the plastoquinone pool, contributing to the so-called cyclic protection of PSII (Photosystem II) from photoinhibition electron flow [81]. The major cyclic pathway is dependent [89], and together with the CDJ2 paralog in the biogenesis/ on the PGR5 (proton gradient regulation 5)/PGRL1 (PGR5- maintenance of thylakoid membranes [90], we could specu- like photosynthetic phenotype 1) proteins [82]. The other late that CDJ3-5 may have a similar role. However, this has involves the NAD(P)H dehydrogenase (NDH) complex. In not been addressed so far for Arabidopsis orthologs, DjC17 higher plants, it forms a large complex associated with PSI, and DjC18. There is no biochemical information on these which is composed of 11 plastid-encoded subunits, some proteins and genetic evidence has been obtained only for additional nuclear-encoded subunits, and auxiliary factors DjC17, the mutation of which results in an altered root hair [83]. Among these, the NDH-I and NDH-K subunits bind development and reduced hair length due to aberrant cortical two and one [F e S ] clusters, respectively [84]. While Arabi- cell division [91]. 4 4 dopsis knock-out mutants for NDH-I and NDH-K genes have not been characterized, tobacco knock-out mutants of ndh A multitude of ferredoxin‑dependent Fe–S proteins genes usually have no phenotype under standard conditions and pathways but are sensitive to environmental stresses [82]. In eukaryotic microalgae and cyanobacteria, an additional The formation of reducing equivalents pathway directly coupled to the photosynthetic electron transport chain and involving hydrogenases allows the pho- FDXs are soluble proteins positioned at a metabolic cross- toproduction of ATP at the expense of reductant synthesis road, controlling the electron flow necessary for CO fixa - in specific conditions such as the response to anaerobiosis tion, nitrogen, and sulfur assimilation but also chlorophyll or anoxia. Chlamydomonas reinhardtii contains two [FeFe]- metabolism to cite a few examples (Fig. 3). Their primary hydrogenases, namely HYDA1 and HYDA2, which will role is to transfer electrons to various acceptors in the produce molecular hydrogen H from protons by accepting stroma, in the thylakoids, and in the inner membrane, includ- electrons from FDXs. These HYDA contain a complex Fe–S ing a large variety of Fe–S proteins but also proteins contain- cluster at their active sites, the H-cluster that is essential for ing heme and non-heme iron centers and flavoproteins [92, catalytic activity [85]. It consists of a classic [F e S ] clus- 93]. Among the latter category, ferredoxin-NADP reductase 4 4 ter linked to a complex 2Fe sub-cluster [86]. Whereas the (FNR) may be the most important one, since it will drive [Fe S ] cluster is assembled by the regular SUF machinery, most of these electrons for the regeneration of NAPDH, 4 4 the sub-cluster requires specific maturation proteins, HYDE, which will then supply in particular the Calvin–Benson HYDF, and HYDG, for this assembly. The HYDE and cycle. It is worth nothing that a significant fraction of FNR HYDG gene products are radical-SAM enzymes, whereas is bound to the thylakoid membrane and it could partici- HYDF is a P-loop NTPase protein constituting a scaffold pate to the cyclic electron flow via the PGR5-dependent assembly platform. These proteins incorporate themselves pathway by interacting with PGLR1 and recruiting FDXs. [Fe S ] clusters that are required for their activity. Another enzyme crucial for carbon fixation and metabolism, 4 4 As oxygenic photosynthesis releases massive amounts in general, is the FTR. This key enzyme, which is almost of oxygen from water, reactive oxygen species are rou- uniquely found in photosynthetic organisms, catalyzes the tinely generated and damage some proteins in many physi- reduction of most thioredoxins (TRXs) found in plastids, ological conditions. Thus, several proteins are implicated thus indirectly participating to the regulation of all TRX- in the repair and protection of the photosystems and their dependent targets in a light-dependent manner [94]. FTR is antennae. One of these, photosystem II protein33 (PSB33), a heterodimer composed of a catalytic and a variable subunit is an integral membrane protein, which contributes to the [95]. In A. thaliana, there is a single gene for the catalytic maintenance of PSII-light-harvesting complex II (LHCII) subunit (FTRB) but two for the variable subunits (FTRA1 supercomplex organization in response to changing light and FTRA2). The function of the [F e S ] cluster found in the 4 4 levels [87]. Whereas the Arabidopsis protein is annotated catalytic subunit is to aid for the reduction of a redox-active as containing a Rieske-type Fe–S cluster by analogy to some disulfide, which reduces, in turn, the TRX disulfide [96]. bacterial counterparts, it does not have the Fe–S binding Given the numerous functions played by plastidial TRXs, it residues contrary to C. reinhardtii ortholog. Other Chla- is not surprising that an ftr knock-out mutant for the catalytic mydomonas proteins, referred to as CDJ3-5 for chloroplast- subunit is lethal. However, a virus-induced gene silencing targeted DnaJ-like proteins, might be important for PSII pro- (VIGS) approach led to plants exhibiting a sectored chlorotic tection. It was shown that CDJ3 and CDJ4 which interact leaf phenotype [97]. It could be observed that these plants with chloroplast ATP-bound HSP70B (heat-shock protein have (1) an abnormal chloroplast biogenesis, (2) a reduced 70B) and are located either in the stroma or attached to thy- photosynthetic performance as measured by the photochemi- lakoids, respectively, are able to bind a [Fe S ] cluster [88]. cal activities, the amount of assembled photosystems and 4 4 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 555 Nitrogen assimila�on Sulfur assimila�o n Carbon fixa� on Fa y acid CB GLU1 biosynthesis SIRB cycle Reducing GLT1 FTR TK FAB2 power supply Glycine betaine SIR NADPH siroheme NIR synthesis ? FAD5 FNR Phytochrome CMO APR1-3 synthesis HO1-4 HYDE HY1 HYDF NEET FDC1/2 FDX1-4 NAD SUFE3 HYDG synthesis psaC PSII repair/ PSB33 protec�on HYDA1-2 psaA psaB Iron signalling ? petC DjC17-18 NDH BCAA synthesis PSI PSII Cytb6/f VAL, ILE, LEU PAO RCCR HCAR IPP Chlorophyll IPMI LA CAO TK Plasto/phyloquinone HDR cLIP α-tocophérol MEP DHAD Chlorophyll Carotenoid pathway HDS catabolism GA-3P AHAS PDH DXS Thiamin DWARF pyruvate acetyl-coA 27.1-3 synthesis THIC ATase1-3 Strigolactone AIR PRA PRPP PTC52 TIC55 synthesis Fig. 3 Fe–S protein-dependent metabolic processes in plastids. Fe–S DjC, and HYDA1/2, there are several close isoforms which have not proteins are represented by dark red boxes. The light red boxes indi- been distinguished. The nomenclature used is the one of A. thaliana cate specificities found in algae either, because they do not exist in except for algal enzymes whose name is from C. reinhardtii. Abbre- terrestrial plants or in the case of PSB33, because only the algal iso- viations for all enzyme names can be found in the text. Other abbre- forms should incorporate a Fe–S cluster. Known FDX-dependent viations are: LA lipoic acid, BCAA br anched-chain amino acids, IPP enzymes have a red outline. Enzymes in green or outlined in orange isopentenyl diphosphate, GA-3P glyceraldehyde 3-phosphate, AIR use, respectively, thiamin or lipoic acid as cofactors. Note that PDH is 5-aminoimidazole ribonucleotide, PRA phospho-ribosylamine, PRPP dependent on both cofactors. For APR, DWARF, ATase, FDX, FDC, 5-phosphoribosyl-1-pyrophosphate CO assimilation rates, and (3) a defective PEP (plastid (pchlide)-dependent translocon component of 52  kDa RNA polymerase)-dependent plastid gene expression, very (PTC52) [102]. All possess a Rieske-type [Fe S ] cluster 2 2 likely because of FTR connection with TRX z [98]. Besides and a mononuclear iron-binding domain. While the five the redox regulation of carbon metabolism enzymes, other enzymes are dependent on FDX, HCAR, CAO, and PTC52 important functions of TRXs in chloroplasts are their par- are involved in chlorophyll synthesis, whereas PAO and ticipation to stress response by regenerating thiol-dependent TIC55 operate in its degradation. More precisely, CAO peroxidases and methionine sulfoxide reductases [99] and to and HCAR are part of the chlorophyll cycle, the process the chlorophyll metabolism by regulating several enzymes of interconversion between chlorophyll a and chlorophyll of the tetrapyrrole biosynthesis pathway [100]. b. The balance between chlorophyll a/b is important for both the stabilization and turnover of chlorophyll in the Beyond tetrapyrrole: chlorophyll and phytochromobilin light-harvesting complexes (LHCs) in diverse physiologi- cal situations, notably during greening and senescence Interestingly, several enzymes of the chlorophyll metabo- when LHCII is massively synthesized or degraded. CAO lism are FDX targets and/or possess Fe–S clusters. The is a thylakoid membrane-anchored enzyme catalyzing 7-hydroxymethyl chlorophyll a reductase (HCAR) is an the two steps of chlorophyll a-to-chlorophyll b oxidation enzyme binding two [Fe S ] clusters and an FAD [101]. [103]. PTC52 would catalyze an analogous oxidation, but 4 4 Besides there are four non-heme oxygenases, namely, using protochlorophyllide a as substrate. However, PTC52 pheophorbide a oxygenase (PAO), chlorophyll a oxyge- is localized at the envelope [104], suggesting that it may nase (CAO), translocon at the inner envelope membrane have another dispensable function, being part of a translo- of chloroplasts 55 (TIC55), and protochlorophyllide cation complex for the import of the protochlorophyllide 1 3 556 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 oxidoreductase A (PORA) precursor in plastids [105]. is encoded by a single gene (HY2), the heme oxygenase is Indeed, A. thaliana ptc52 knock-out lines have a growth encoded by four members in A. thaliana, HY1/HO1, and indistinguishable from wild-type plants (Table 1) [106]. HO2-4 [118]. On the contrary, an Arabidopsis mutant for CAO, named chlorina1, exhibits a pale green phenotype characterized Macronutrient assimilation: similarities in nitrogen by a chlorophyll b decrease [103] and is extremely sensi- and sulfur assimilation pathways tive to photooxidation due to the lack of chlorophyll–pro- tein antenna complexes in PSII and to an increased pro- The reductive assimilations of nitrogen and sulfur constitute duction of singlet oxygen [107]. HCAR catalyzes the two other chloroplastic metabolic processes, which rely on second half-reaction in chlorophyll b-to-chlorophyll a FDX-dependent Fe–S proteins. As already presented, sul- conversion, the first one being catalyzed by chlorophyll fate assimilation is extremely important, because it provides b reductases (CBR) [101]. While hcar mutants have no cysteine, which is the source of sulfur for many molecules phenotype under standard growth conditions, they exhibit but also the substrate of cysteine desulfurases and a protein a stay-green phenotype after transfer to darkness (Table 1) ligand in all plastidial Fe–S proteins known so far. Of the [108]. four enzymes/complexes, which allow forming cysteine from In higher plants, chlorophyll is broken down to colour- sulfate, two possess a Fe–S cluster. The second reaction, e.g, less linear tetrapyrroles in a series of reactions. One of the transformation of adenosine 5′ phosphosulfate (APS) to these steps, the porphyrin ring opening of pheophorbide a, sulfite, is catalyzed by adenosine 5′ -phosphosulfate reduc- is catalyzed by PAO. This step occurs in senescent leaves tases (APR). There are three isoforms in A. thaliana, APR1- and fruits, and requires FDXs and NADPH [109–111]. The 3, all localized in plastids. The enzymes are formed by two PAO proteins possess a C-terminal transmembrane domain domains, a reductase domain, that bears a [Fe S ] cluster, 4 4 for their binding to the thylakoid membrane [104]. The and a GRX domain at the C-terminus, that makes these PAO gene was identified by genetic studies and was initially enzymes glutathione-dependent [10]. The SIR catalyzes referred to as accelerated cell death 1 (ACD1) in Arabidopsis the next step, the six electron reduction of sulfite to sulfide. [112] or LLS1 (Lethal leaf spot 1) in maize [113]. Extinction This FDX-dependent enzyme incorporates a siroheme, e.g, of PAO in knock-out mutants or in antisense lines from dif- a heme whose iron atom is liganded by the thiolate ligand ferent plant species leads to a light-dependent premature cell of a [F e S ] cluster, which is crucial for its activity. There is 4 4 death phenotype, most likely due to cytotoxic effects of the a single, essential, SIR gene in Arabidopsis and the protein increased pheophorbide a [109, 112, 114]. Similar to hcar is found exclusively in plastids. A weak allele mutant with mutants, pao mutants have a stay-green phenotype in dark about 25% SIR activity is viable but has a strongly retarded [109, 114]. The product of the reaction catalyzed by PAO is growth, pointing to the extreme importance of this enzyme red chlorophyll catabolite, which is then reduced by a FDX- for plant development [119]. dependent red chlorophyll catabolite reductase (RCCR) to The assimilation of inorganic nitrogen (mostly in the yield the primary fluorescent chlorophyll catabolite (FCC), form of nitrate and ammonium) is another essential pro- pFCC [115]. From this primary phyllobilin, a large variety cess for plants taking place in part in plastids. Nitrate will of other phyllobilins is formed subsequently. Although a be reduced in two steps. The first one, catalyzed by nitrate tic55 mutant in Arabidopsis does not show any detectable reductase, gives nitrite, which is reduced to ammonia by a phenotype, TIC55, which is localized in the inner membrane FDX-dependent nitrite reductase (NIR). As the SIR enzyme, of the chloroplast envelope, is responsible for phyllobilin NIR binds a siroheme that is mandatory for the six elec- hydroxylation during senescence [104, 106]. This would be tron reduction of nitrite. This gene is also essential, since a the last step in this subcellular compartment, and in that mutant in barley does not grow in the absence of an external sense, TIC55 may contribute to chlorophyll catabolite export nitrogen source [120]. In the next steps, ammonia, including from plastids for their subsequent vacuolar detoxification. the part coming from the photorespiration process, is assimi- A closely related molecule to chlorophyll and incidentally lated via glutamine synthetase (GS) which catalyzes the con- heme is phytochromobilin (PfB), the chromophore usually densation of glutamate and ammonia into glutamine and via covalently bound to phytochromes of higher plants. All these glutamate synthase (GOGAT) which forms two molecules of molecules branch from protoporphyrin IX in the tetrapyr- glutamate from glutamine and 2-oxoglutarate. There is evi- role synthesis pathway. From the closed tetrapyrrole ring dence that NIR, GS, and GOGAT can form a complex within of heme, a heme oxygenase catalyzes the oxidative open- the chloroplast [121]. Plants possess two forms of chloro- ing of this chain to yield biliverdin IXa. This molecule is plastic GOGAT, which are dependent either on NADH or on then reduced into phytochromobilin by a PfB synthase. In FDX. All contain an FMN and a [Fe S ] cluster. In A. thali- 3 4 higher plants, both types of enzymes are soluble and depend ana, NADH-GOGAT is encoded by a unique gene (GLT1), on FDXs for their activity [116, 117]. While PfB synthase whereas two genes encode Fd-GOGAT (GLU1 and GLU2), 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 557 GLU1 is the predominant form in leaves [122]. Arabidopsis CMO-like gene in E. coli does not promote betaine synthesis mutants for GLU2 and GLT1 have no growth phenotype, [130]. This protein is unique to eukaryotic photosynthetic although a decrease in the chlorophyll content was measured organisms as it is not found in cyanobacteria, supporting in the glt1 mutant [123, 124]. An Arabidopsis mutant for a recent evolution of this enzyme. No Arabidopsis mutant GLU1 has a respiratory phenotype, i.e, a dwarf and chloro- has been characterized so far, but antisense CMO transgenic tic phenotype in air which is no longer visible under high sugar beet plants are susceptible to salt stress [131]. CO conditions [122, 123, 125]. Of importance for these To conclude on this part, it is important to note that most pathways, it is worth mentioning that sirohydrochlorin fer- of these enzymes are also expressed in plastids of non-pho- rochelatase (SIRB), the enzyme responsible for the last step tosynthetic tissues. In this context, FDXs are maintained of siroheme biosynthesis by inserting ferrous iron into the reduced by FNR and NADPH generated in the oxidative tetrapyrrole ring of sirohydrochlorin, is a [F e S ] enzyme pentose phosphate pathway, the reverse reaction compared 2 2 unlike bacterial orthologs. In this essential protein, the Fe–S to photosynthetic organs. A few other enzymes such as cluster is not mandatory for the enzymatic reaction, but it (1-hydroxy-2-methyl-2-(E)-butenyl 4-diphosphate synthase might have a regulatory role [126]. (HDS), zeaxanthine epoxidase, and β-carotene 3 hydroxylase 1, 2) have been also described as FDX-dependent proteins, Fatty acid biosynthesis but they will be discussed in the next sections. However, several additional proteins or pathways are yet unidentified. The biosynthesis of fatty acids is another crucial pathway It is for instance worth mentioning that studies devoted to occurring in plastids, which depends directly or indirectly the isolation of FDX partners by proteomic approaches led to on Fe–S proteins. First, the acetyl-coenzyme A, that is the identification of novel putative targets at least in cyano- used as a building block for fatty acids, is generated by the bacteria and Chlamydomonas [132, 133]. In this respect, a plastidial pyruvate dehydrogenase (PDH) complex, its E2 pyruvate:ferredoxin oxidoreductase (PFO), found in many subunit being lipoylated and thus dependent on the Fe–S unicellular eukaryotes, decarboxylates pyruvate to acetyl- containing lipoate synthase (see below). After the synthe- coenzyme at the expense of FDXs [134]. The C. reinhardtii sis of saturated fatty acids, their conversion to unsaturated PFO possesses three distinct [Fe S ] clusters. It may also 4 4 forms, which are required for membrane fluidity, is catalyzed contribute to the light-independent H production by passing by fatty acid desaturases. Some of them contain a di-iron electron to the hydrogenase [135]. center and are FDX-dependent proteins [127]. The FAB2 protein is a soluble stearoyl-ACP desaturase introducing the Biosynthesis of lipoic acid and thiamin cofactors first double bond into stearoyl-ACP between carbons 9 and and their dependent pathways 10 to produce oleoyl-ACP (18:1 Delta9-ACP). The FAD5 protein attached to the chloroplast envelope inner membrane Requirement of two atypical radical‑SAM enzymes catalyzes the earliest step of 16:0 desaturation initiating the very rapid 16:0–16:1–16:2–16:3 desaturation of monogalac- Beyond their role in electron transfer, Fe–S clusters are also tosyldiacylglycerol (MGDG), one of the four main classes important for enzyme catalysis, especially during the bio- of glycerolipids found in the photosynthetic membranes of synthesis of vitamin B1/thiamin and of lipoic acid. Whereas higher plant chloroplasts with the digalactosyldiacylglycerol thiamin is only synthesized in chloroplasts a lipoic acid bio- (DGDG), the phospholipid phosphatidylglycerol (PG), and synthesis pathway is present in both plastids and mitochon- the sulfolipid sulfoquinovosyldiacylglycerol (SQDG) [128]. dria. This is consistent with the existence of two distinct Other plastidial linoleate/oleate desaturases (FAD4, 6, 7, 8) genes encoding a mitochondrial (mLIP) and a chloroplastic and the numerous FAD5-like proteins may also be depend- (cLIP) lipoate synthase [136]. ent on FDX as they also probably contain di-iron centers. Lipoic acid is synthesized from octanoic acid and thus via the fatty acid biosynthesis pathway by the addition of two Other metabolic processes sulfur atoms into the octanoyl group bound to an acyl carrier protein (ACP) via a radical-SAM mechanism. This reaction Additional FDX-dependent proteins are present in chloro- is catalyzed by lipoic acid synthase [136]. It is important to plasts. Besides PAO, CAO, PTC52, and TIC55, the fifth note that lipoic acid is synthesized attached to proteins and non-heme oxygenase found in plants [102] is referred to as no free lipoic acid is produced. The E. coli Lip5 binds two choline monooxygenase (CMO), because it was found to [Fe S ] clusters [137]. One cluster, coordinated by cysteines 4 4 catalyze the oxidation of choline, the first step of glycine present in a Cx Cx C motif common to all classical radical- 3 2 betaine biosynthesis in spinach [129]. However, this might SAM enzymes, is required for the formation of the activated not be the sole or main function, since Arabidopsis does not adenosyl radical from SAM molecules. The second clus- produce glycine betaine and expression of the Arabidopsis ter, coordinated by a Cx Cx C motif specific of lipoic acid 4 5 1 3 558 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 synthases, was suggested to provide the sulfur atoms and A single lipoic acid‑dependent enzyme but several thus to be degraded at each turnover of the enzyme. The thiamin‑dependent enzymes in plastids presence of a Fe–S cluster has not yet been demonstrated in plant cLIP, but Arabidopsis cLIP possesses both cysteine In plastids, the only known lipoic acid-dependent enzyme motifs and is able to complement the E. coli lip5 mutant is PDH. A similar complex is found in plant mitochondria, [136]. Plants impaired in cLIP have not been characterized but it uses lipoic acid synthesized in this compartment, as yet, but Arabidopsis mutants for genes involved in the syn- does another citric acid cycle enzyme, the α-ketoglutarate thesis of lipoic acid in mitochondria are lethal [138]. dehydrogenase or 2-oxoglutarate dehydrogenase complex, Thiamin is made of pyrimidine and thiazole heterocy- but also two complexes involved in the amino acid metabo- cles, both being synthesized in the chloroplast. The syn- lism, the glycine cleavage complex, and the branched-chain thesis of the thiazole moiety involves a 4-methyl-5-b- oxoacid dehydrogenase (BCDH) complex. On the other hydroxyethylthiazole phosphate (HET-P) synthase (THI1) hand, there is a single pathway for the synthesis of ThDP forming an adenylated thiazole intermediate (ADT) at the which is used as a coenzyme by many enzymes of the pri- expense of nicotinamide adenine dinucleotide (NAD) and mary metabolism, notably involved in the catabolism of glycine, ADT, which is then hydrolyzed to HET-P [139]. sugars and amino acids, and found in the chloroplasts, mito- The pyrimidine heterocycle is derived from purine bio- chondria, and cytosol [139]. In plastids, besides the PDH synthesis. The first step in the synthesis of the pyrimidine complex, thiamin is also a cofactor for transketolase (TK) moiety is catalyzed by the 4-amino-2-methyl-5-hydroxy- of both the Calvin–Benson cycle and the non-oxidative pen- methylpyrimidine phosphate (HMP-P) synthase (THIC), a tose phosphate pathway, for 1-deoxy-d -xylulose 5-phosphate radical-SAM Fe–S enzyme that forms HMP-P from 5-ami- synthase (DXS) of the methylerythritol phosphate (MEP) noimidazole ribonucleotide (AIR) and SAM. In contrast to pathway and for acetohydroxy acid synthase (AHAS) of the canonical radical-SAM enzymes, all THIC proteins harbour branched-chain amino acid (BCAA) biosynthesis pathway a Cx Cx C motif involved in the binding of a [Fe S ] cluster [139]. 2 4 4 4 in their C-terminal part [140, 141]. Then, an HMP-P kinase/ thiamin monophosphate (ThMP) pyrophosphorylase (TH1) The central pyruvate dehydrogenase complex PDH cata- phosphorylates HMP-P to HMP-PP but also condenses the lyzes the decarboxylation of pyruvate into acetyl-coA that latter compound to HET-P to form ThMP. This ThMP is is used in particular for fatty acid synthesis as already men- transformed into the diphosphate form ThDP in the cyto- tioned [128]. It consists of three subunits, E1–E3, each sol through the action of two consecutive enzymes before requiring a different cofactor. Thiamin is bound to the pyru- being redistributed to mitochondria and plastids. THIC is vate dehydrogenase subunit (E1), whereas the lipoic acid encoded by a single essential gene in Arabidopsis [140]. is covalently attached to the dihydrolipoyl acyltransferase An Arabidopsis thic mutant is lethal at the cotyledon stage subunit (E2) and an FAD is bound to the dihydrolipoam- unless supplemented with thiamin [140, 142]. Another fam- ide dehydrogenase subunit (E3). The attached lipoyl moiety ily of plastidial Fe–S enzymes is linked indirectly to thia- functions as a carrier of reaction intermediates among the min biosynthesis, because they catalyze the first committed active sites of the components of the complex. The E3 subu- step of the de novo synthesis of purine in chloroplasts. The nit has a key regulatory role by reoxidizing the lipoamide glutamine phosporibosyl pyrophosphate amidotransferases cofactor and thus completing the catalytic cycle. The dis- (ATases, also known as GPAT) catalyze the amination of ruption of the gene encoding the E2 subunit of plastidial 5-phosphoribosyl-1-pyrophosphate (PRPP) to 5-phospho- PDH results in an early embryo-lethal phenotype in Arabi- ribosylamine (PRA) with the concomitant conversion of dopsis [147]. glutamine into glutamate [143]. After four additional steps, PRA is transformed into AIR, the THIC substrate. In Arabi- Branched‑chain amino acid biosynthesis The AHAS pro- dopsis, ATase is encoded by a family of three genes (ATase1 tein is involved in the first steps of BCAA biosynthesis. to ATase3) which are expressed in various tissues at dif- This is a heterodimer, composed of separate catalytic and ferent levels [144, 145]. Whereas E. coli ATase does not regulatory subunits, which catalyzes the conversion of two require a Fe–S cluster as cofactor, the human enzyme uses molecules of pyruvate into 2-acetolactate used for valine a [Fe S ] cluster. Based on the conservation of the involved and leucine synthesis or of one molecule of pyruvate and 4 4 cysteines, the three A. thaliana isoforms should also bind a one molecule of 2-oxobutanoate into 2-aceto-2-hydroxy- Fe–S cluster. Whereas Arabidopsis ATase1 mutant has no butyrate used for isoleucine synthesis [148]. Interestingly, growth phenotype mutants lacking ATase2 exhibit strong two Fe–S enzymes named dihydroxyacid dehydratase growth retardation with bleached leaves (Table 1) [144]. In (DHAD) and isopropylmalate isomerase (IPMI) are the latter mutant exhibiting a decreased capacity in chlo- also required for BCAA synthesis. DHAD catalyzes the roplast protein import, cells are smaller in size [144, 146]. penultimate step before the formation of isoleucine and 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 559 valine, e.g., the dehydration of 2,3-dihydroxy-3-isovaler- suppressed by expressing the eukaryotic mevalonate path- ate or 2,3-dihydroxy-3-methylvalerate to the 2-oxo acids way, which does not rely on Fe–S proteins [54, 155]. (3-methyl-2-oxobutanoate or 3-methyl-2-oxopentanoate). Besides its involvement in the Calvin–Benson cycle In Arabidopsis, there is a single essential gene for DHAD where it catalyzes the formation of ribose-5-phosphate and [149]. However, Arabidopsis mutants with intermedi- xylulose-5-phosphate from sedoheptulose-7-phosphate and ate DHAD levels obtained by an RNAi approach indeed glyceraldehyde-3-phosphate (GA-3P), TK operates in the have reduced amounts of BCAA in roots, which cause a opposite direction in the non-oxidative pentose phosphate short root phenotype [149]. The only biochemical charac- pathway, forming GA-3P, which is then condensed to pyru- terization performed so far has been done with an enzyme vate to form 1-deoxy-d -xylulose 5-phosphate (DXP), a reac- purified from spinach leaves. Unlike the E. coli enzyme, tion catalyzed by DXP synthase (DXS). Although poorly which incorporates a [Fe S ] center, the spinach enzyme characterized in plants, it was demonstrated that antisense 4 4 incorporates a [Fe S ] cluster required for activity [150]. tobacco plants with variable TK levels have a marked 2 2 Leucine biosynthesis requires an additional Fe–S enzyme shoot weight decrease [156]. For the most affected lines, for the late reactions. The isopropylmalate isomerase cata- a decrease in chlorophylls and carotenoids was measured lyzes the reversible conversion of 2-isopropylmalate into which is consistent with the importance of DXP for the 3-isopropylmalate. In plants, IPMI consists of a heterodi- MEP pathway. Surprisingly, overexpression of an A. thali- mer composed of a large (LSU) and a small (SSU) subunit ana chloroplastic TK in tobacco leads to chlorosis, which is encoded by one and three genes in A. thaliana, respec- annihilated by thiamin supplementation [157]. In A. thali- tively [151]. The genetic analyses demonstrated that A. ana, DXS is an essential gene [158]. In sense and antisense thaliana knock-down mutants for the large subunit, which Arabidopsis lines exhibiting altered levels of DXS, both binds a [Fe S ] center, display a severe delay in devel- the chlorophylls, carotenoids, tocopherols, gibberellin, and 4 4 opment [151, 152]. Concerning small subunits, the SSU1 abscisic acid contents and the growth and germination rate protein is required for viability, unlike SSU2 and SSU3, are slightly affected [159]. Finally, downstream of DXS, two which might be redundant, because A. thaliana mutants Fe–S proteins are required to form two key intermediates in have no phenotype [151, 153]. In addition to a role in leu- isoprenoid biosynthesis: isopentenyl diphosphate (IPP) and cine biosynthesis, IPMI is involved in the biosynthesis of dimethylallyl diphosphate (DMAPP). Both the 1-hydroxy- glucosinolates, sulfur-containing secondary metabolites, 2-methyl-2-(E)-butenyl 4-diphosphate synthase (HDS/ serving for defence reactions. This is consistent with the ISPG) and the 1-hydroxy-2-methyl-2-(E)-butenyl 4-diphos- fact that an identical reaction type exists for the Met chain phate reductase (HDR/ISPH/LytB) bind a [Fe S ] cluster 4 4 elongation cycle for glucosinolate formation and that A. [160, 161], FDX being able to provide electrons to HDS thaliana mutant plants for the large subunit accumulate [162]. Plant mutants disrupted in ISPG or ISPH gene have a both Leu biosynthesis and Met chain elongation interme- severely impaired chloroplastic development that causes an diates [151]. albino phenotype [163–165]. At least two enzymes participating in the carotenoid Isoprenoid biosynthesis and its derived molecules Isopre- biosynthesis pathways are dependent on FDXs. With- noids are very diverse metabolites, central to plant devel- out describing all the steps, the β-carotene 3 hydroxylase opment. We have already discussed the biosynthesis of 1, 2 which contains a di-iron center catalyzes two suc- chlorophylls, which consist of a tetrapyrrole ring with an cessive steps, the transformation of all-trans β-carotene attached isoprenoid-derived phytol chain, but many other to β-cryptoxanthin and then to zeaxanthin. Then, the fla- isoprenoids are present in plastids such as α-tocopherol, voprotein zeaxanthin epoxidase catalyzes the conversion phylloquinone, plastoquinone, and carotenoids to cite of zeaxanthin to antheraxanthin and then to violaxanthin only the most important. Moreover, several plant hor- [166]. These four steps require oxygen and FDX as an elec- mones are derived from carotenoids. All isoprenoids are tron donor. Other proteins in this pathway might, in fact, derived from a prenyl diphosphate (prenyl-PP) precursor, be dependent on FDXs. For instance, there are several which is synthesized by two independent pathways, the cytochrome P450 monooxygenases participating in this cytosolic mevalonate (MVA) pathway, and the plastidial pathway (and other pathways) in plastids, whose electron 2-C-methyl-d -erythritol 4-phosphate (MEP) pathway donors/acceptors are yet unknown. [154]. The latter pathway is dependent on both Fe–S and Derived from carotenoids, strigolactones (SL) are plant thiamin (TK and DXS)-dependent enzymes. In E. coli, the hormones having diverse functions in plant growth and Fe–S proteins belonging to this MEP pathway are the only development. Their biosynthesis begins with the conver- one that are completely essential. Indeed, the lethality of sion of all-trans β-carotene to 9-cis-β-carotene, a reaction sufa-isca and erpa mutants observed under aerobiosis is performed by a β-carotene isomerase named DWARF27. This protein, found from algae to higher plants, and first 1 3 560 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 characterized in rice is a Fe–S enzyme [167]. The Arabidop- holoproteins and predictions of Fe–S proteins from the pro- sis genome encodes three orthologs. An Arabidopsis mutant tein primary sequences are often impossible, because there for one of these genes and a rice mutant have shoot branch- is no universal signature for identifying Fe–S cluster ligands. ing phenotypes, but it remains relatively weak compared to Acknowledgements The work of both laboratories is supported other mutants affected in SL biosynthesis [168]. by the Agence Nationale de la Recherche, Grant no. ANR-2013- BSV6-0002-01. The UMR 1136 is also supported by a Grant over- Are there other plastidial Fe–S proteins to discover? seen by the French National Research Agency (ANR) as part of the “Investissements d’Avenir” program (ANR-11-LABX-0002-01, Lab of Excellence ARBRE). This publication is based upon work from Some Fe–S proteins, such as NEET, have been recently iden- COST Action CA15133, supported by COST (European Cooperation tified or characterized in plants. Unlike mitoNEET, which in Science and Technology). The authors are grateful to Pr Jean-Pierre is bound to the outer membrane of mitochondria in animals Jacquot for its careful reading of the manuscript. owing to a membrane anchoring extension, the Arabidop- sis NEET protein is located exclusively in the chloroplast Open Access This article is distributed under the terms of the Crea- tive Commons Attribution 4.0 International License (http://creat iveco stroma [169]. As its vertebrate counterparts, Arabidopsis mmons .org/licen ses/by/4.0/), which permits use, duplication, adapta- NEET forms dimers; each monomer harbouring an atypi- tion, distribution and reproduction in any medium or format, as long cal [Fe S ] cluster coordinated by three Cys and one His 2 2 as you give appropriate credit to the original author(s) and the source, [170]. While obtaining knock-out plants may have been provide a link to the Creative Commons license and indicate if changes were made. hampered by the fact, it is an essential gene, Arabidopsis lines with reduced AtNEET transcript levels exhibit late greening, delayed bolting, and early senescence. Moreover, these plants accumulate ROS and have an altered sensitivity References to Fe levels, which led to the proposal that AtNEET likely plays a role in the regulation of Fe homeostasis [170]. From 1. Briat J-F, Ravet K, Arnaud N et al (2010) New insights into fer- its capacity to transfer its Fe–S cluster to a FDX in vitro ritin synthesis and function highlight a link between iron homeo- [170], it may be hypothesized that NEET could be part of the stasis and oxidative stress in plants. Ann Bot 105:811–822. https ://doi.org/10.1093/aob/mcp12 8 SUF machinery and facilitate the trafficking of Fe–S clusters 2. Brumbarova T, Bauer P, Ivanov R (2015) Molecular mechanisms towards certain client proteins. governing Arabidopsis iron uptake. Trends Plant Sci 20:124–133. Another reason why we expect to discover novel Fe–S https ://doi.org/10.1016/j.tplan ts.2014.11.004 proteins is that some proteins may be specific to photosyn- 3. Jeong J, Guerinot ML (2009) Homing in on iron homeostasis in plants. Trends Plant Sci 14:280–285. https ://doi.org/10.1016/j. thetic organisms because of their atypical structure organi- tplan ts.2009.02.006 zation or their involvement in specific plastidial functions. 4. Solti Á, Kovács K, Müller B et al (2016) Does a voltage-sen- An interesting example in this regard is SUFE3, a chimeric sitive outer envelope transport mechanism contributes to the protein formed by an SUFE domain fused to a quinolinate chloroplast iron uptake? Planta 244:1303–1313. https ://doi. org/10.1007/s0042 5-016-2586-3 synthase domain, NADA [18]. This enzyme, which carries a 5. Jeong J, Cohu C, Kerkeb L et al (2008) Chloroplast Fe(III) che- [Fe S ] cluster indispensable for its activity and thus crucial 4 4 late reductase activity is essential for seedling viability under for NAD biosynthesis, is the sole NADA representative of iron limiting conditions. PNAS 105:10619–10624. https ://doi. A. thaliana. The fact that the Fe–S cluster in SUFE3 can be org/10.1073/pnas.07083 67105 6. Duy D, Wanner G, Meda AR et al (2007) PIC1, an ancient per- reconstituted using its own SUFE domain in the presence mease in Arabidopsis chloroplasts, mediates iron transport. Plant of NFS2, cysteine, and ferrous iron may render this protein Cell 19:986–1006. https ://doi.org/10.1105/tpc.106.04740 7 independent on the SUFBCD scaffold complex [18]. 7. Duy D, Stübe R, Wanner G, Philippar K (2011) The chloro- Other Fe–S protein-dependent processes likely remain plast permease PIC1 regulates plant growth and development by directing homeostasis and transport of iron. Plant Physiol to be identified in plastids as in other subcellular compart - 155:1709–1722. https ://doi.org/10.1104/pp.110.17023 3 ments. For instance, the affinity purification strategy used 8. Shimoni-Shor E, Hassidim M, Yuval-Naeh N, Keren N (2010) for cyanobacterial and algal enzymes indicates that numer- Disruption of Nap14, a plastid-localized non-intrinsic ABC ous FDX-dependent processes await identification [132, protein in Arabidopsis thaliana results in the over-accumu- lation of transition metals and in aberrant chloroplast struc- 133]. The same is true in mitochondria where the roles and tures. Plant Cell Environ 33:1029–1038. https ://doi.org/10.111 partners of the two FDXs are unknown. On the other hand, 1/j.1365-3040.2010.02124 .x novel Fe–S proteins will be undoubtedly identified in the 9. Tarantino D, Morandini P, Ramirez L et al (2011) Identifica- future thanks to the dozens of annotated sequenced genomes tion of an Arabidopsis mitoferrinlike carrier protein involved in Fe metabolism. Plant Physiol Biochem 49:520–529. https ://doi. now available for model plants and to the ever larger collec- org/10.1016/j.plaph y.2011.02.003 tions of available Arabidopsis mutants. The reasons why it 10. Takahashi H, Kopriva S, Giordano M et al (2011) Sulfur assimi- is not trivial to isolate them is that the sensitivity of these lation in photosynthetic organisms: molecular functions and metallic cofactors to oxygen may hamper the isolation of regulations of transporters and assimilatory enzymes. Annu Rev 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 561 Plant Biol 62:157–184. https ://doi.org/10.1146/annur ev-arpla 27. Yuda E, Tanaka N, Fujishiro T et al (2017) Mapping the key resi- nt-04211 0-10392 1 dues of SufB and SufD essential for biosynthesis of iron–sulfur 11. Cao M-J, Wang Z, Wirtz M et al (2013) SULTR3;1 is a chloro- clusters. Sci Rep. https ://doi.org/10.1038/s4159 8-017-09846 -2 plast-localized sulfate transporter in Arabidopsis thaliana. Plant 28. Blanc B, Clémancey M, Latour J-M et al (2014) Molecular inves- J 73:607–616. https ://doi.org/10.1111/tpj.12059 tigation of iron–sulfur cluster assembly scaffolds under stress. 12. Couturier J, Touraine B, Briat J-F et al (2013) The iron–sulfur Biochemistry 53:7867–7869. https://doi.or g/10.1021/bi5012496 cluster assembly machineries in plants: current knowledge and 29. Saini A, Mapolelo DT, Chahal HK et al (2010) SufD and SufC open questions. Front Plant Sci 4:259. https ://doi.org/10.3389/ ATPase activity are required for iron acquisition during in vivo fpls.2013.00259 Fe–S cluster formation on SufB. Biochemistry 49:9402–9412. 13. Balk J, Schaedler TA (2014) Iron cofactor assembly in plants. https ://doi.org/10.1021/bi101 1546 Annu Rev Plant Biol 65:125–153. https://doi.or g/10.1146/annur 30. Wollers S, Layer G, Garcia-Serres R et al (2010) Iron–sulfur (Fe– ev-arpla nt-05021 3-03575 9 S) cluster assembly the SufBCD complex is a new type of Fe–S 14. Hu X, Kato Y, Sumida A et al (2017) The SUFBC2D complex scaffold with a flavin redox cofactor. J Biol Chem 285:23331– is required for the biogenesis of all major classes of plastid Fe–S 23341. https ://doi.org/10.1074/jbc.M110.12744 9 proteins. Plant J 90:235–248. https ://doi.org/10.1111/tpj.13483 31. Ravet K, Touraine B, Boucherez J et al (2009) Ferritins con- 15. Pilon-Smits EAH, Garifullina GF, Abdel-Ghany S et al (2002) trol interaction between iron homeostasis and oxidative stress in Characterization of a NifS-like chloroplast protein from Arabi- Arabidopsis. Plant J 57:400–412. https://doi.or g/10.1111/j.1365- dopsis. implications for its role in sulfur and selenium metabo-313X.2008.03698 .x lism. Plant Physiol 130:1309–1318. https ://doi.or g/10.1104/ 32. Parent A, Elduque X, Cornu D et al (2015) Mammalian frataxin pp.102.01028 0 directly enhances sulfur transfer of NFS1 persulfide to both ISCU 16. Léon S, Touraine B, Briat J-F, Lobréaux S (2002) The AtNFS2 and free thiols. Nat Commun 6:5686. https ://doi.org/10.1038/ gene from Arabidopsis thaliana encodes a NifS-like plastidial ncomm s6686 cysteine desulphurase. Biochem J 366:557–564. https ://doi. 33. Colin F, Martelli A, Clémancey M et  al (2013) Mammalian org/10.1042/BJ200 20322 frataxin controls sulfur production and iron entry during de novo 17. Roret T, Pégeot H, Couturier J et al (2014) X-ray structures of Fe4S4 cluster assembly. J Am Chem Soc 135:733–740. https :// Nfs2, the plastidial cysteine desulfurase from Arabidopsis thali-doi.org/10.1021/ja308 736e ana. Acta Crystallogr F Struct Biol Commun 70:1180–1185. 34. Buchensky C, Sánchez M, Carrillo M et al (2017) Identification https ://doi.org/10.1107/S2053 230X1 40170 26 of two frataxin isoforms in Zea mays: structural and functional 18. Murthy NMU, Ollagnier-de-Choudens S, Sanakis Y et al (2007) studies. Biochimie 140:34–47. https ://doi.org/10.1016/j.bioch Characterization of Arabidopsis thaliana SufE2 and SufE3 func- i.2017.06.011 tions in chloroplast iron–sulfur cluster assembly and NAD syn- 35. Turowski VR, Aknin C, Maliandi MV et  al (2015) Frataxin thesis. J Biol Chem 282:18254–18264. https ://doi.org/10.1074/ is localized to both the chloroplast and mitochondrion and is jbc.M7014 28200 involved in chloroplast Fe–S protein function in Arabidop- 19. Ye H, Abdel-Ghany SE, Anderson TD et al (2006) CpSufE acti- sis. PLoS One 10:e0141443. https ://doi.or g/10.1371/jour n vates the cysteine desulfurase CpNifS for chloroplastic Fe–S al.pone.01414 43 cluster formation. J Biol Chem 281:8958–8969. https ://doi. 36. Maliandi MV, Busi MV, Turowski VR et al (2011) The mito- org/10.1074/jbc.M5127 37200 chondrial protein frataxin is essential for heme biosynthe- 20. Outten FW, Wood MJ, Muñoz FM, Storz G (2003) The SufE sis in plants. FEBS J 278:470–481. h t t p s : / / d oi . o r g / 1 0. 1 1 1 protein and the SufBCD complex enhance SufS cysteine desul-1/j.1742-4658.2010.07968 .x furase activity as part of a sulfur transfer pathway for Fe–S clus- 37. Söderberg C, Gillam ME, Ahlgren E-C et al (2016) The structure ter assembly in Escherichia coli. J Biol Chem 278:45713–45719. of the complex between yeast frataxin and ferrochelatase: char- https ://doi.org/10.1074/jbc.M3080 04200 acterization and pre-steady state reaction of ferrous iron delivery 21. Xu XM, Møller SG (2006) AtSufE is an essential activator of and heme synthesis. J Biol Chem 291:11887–11898. https://doi. plastidic and mitochondrial desulfurases in Arabidopsis. EMBO org/10.1074/jbc.M115.70112 8 J 25:900–909. https ://doi.org/10.1038/sj.emboj .76009 68 38. Dhalleine T, Rouhier N, Couturier J (2014) Putative roles of 22. Van Hoewyk D, Abdel-Ghany SE, Cohu CM et al (2007) Chloro- glutaredoxin-BolA holo-heterodimers in plants. Plant Signal plast iron–sulfur cluster protein maturation requires the essential Behav 9:e28564. https ://doi.org/10.4161/psb.28564 cysteine desulfurase CpNifS. PNAS 104:5686–5691. https://doi. 39. Roret T, Tsan P, Couturier J et al (2014) Structural and spectro- org/10.1073/pnas.07007 74104 scopic insights into BolA-glutaredoxin complexes. J Biol Chem 23. Cory SA, Van Vranken JG, Brignole EJ et al (2017) Structure of 289:24588–24598. https ://doi.org/10.1074/jbc.M114.57270 1 human Fe–S assembly subcomplex reveals unexpected cysteine 40. Braymer JJ, Lill R (2017) Iron–sulfur cluster biogenesis and traf- desulfurase architecture and acyl-ACP-ISD11 interactions. PNAS ficking in mitochondria. J Biol Chem 292:12754–12763. https:// 114:E5325–E5334. https ://doi.org/10.1073/pnas.17028 49114 doi.org/10.1074/jbc.R117.78710 1 24. Boniecki MT, Freibert SA, Mühlenhoff U et al (2017) Structure 41. Brancaccio D, Gallo A, Piccioli M et al (2017) [4Fe–4S] cluster and functional dynamics of the mitochondrial Fe/S cluster syn- assembly in mitochondria and Its impairment by copper. J Am thesis complex. Nat Commun 8:1287. https ://doi.org/10.1038/ Chem Soc 139:719–730. https ://doi.org/10.1021/jacs.6b095 67 s4146 7-017-01497 -1 42. Brancaccio D, Gallo A, Mikolajczyk M et al (2014) Formation of 25. Layer G, Gaddam SA, Ayala-Castro CN et al (2007) SufE trans- [4Fe–4S] clusters in the mitochondrial iron–sulfur cluster assem- fers sulfur from SufS to SufB for iron–sulfur cluster assembly. bly machinery. J Am Chem Soc 136:16240–16250. https ://doi. J Biol Chem 282:13342–13350. https ://doi.or g/10.1074/jbc.org/10.1021/ja507 822j M6085 55200 43. Bandyopadhyay S, Gama F, Molina-Navarro MM et al (2008) 26. Hirabayashi K, Yuda E, Tanaka N et al (2015) Functional dynam- Chloroplast monothiol glutaredoxins as scaffold proteins for the ics revealed by the structure of the SufBCD complex, a novel assembly and delivery of [2Fe–2S] clusters. EMBO J 27:1122– ATP-binding cassette (ABC) protein that serves as a scaffold for 1133. https ://doi.org/10.1038/emboj .2008.50 44. Rey P, Becuwe N, Tourrette S, Rouhier N (2017) Involve- iron–sulfur cluster biogenesis. J Biol Chem 290:29717–29731. ment of Arabidopsis glutaredoxin S14 in the maintenance of https ://doi.org/10.1074/jbc.M115.68093 4 1 3 562 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 chlorophyll content. Plant Cell Environ 40:2319–2332. https :// iron–sulphur cluster metabolism. J Exp Bot 63:403–411. https:// doi.org/10.1111/pce.13036 doi.org/10.1093/jxb/err28 6 45. Moseler A, Aller I, Wagner S et al (2015) The mitochondrial 62. Hasnain G, Waller JC, Alvarez S et al (2012) Mutational analysis monothiol glutaredoxin S15 is essential for iron–sulfur protein of YgfZ, a folate-dependent protein implicated in iron/sulphur maturation in Arabidopsis thaliana. PNAS 112:13735–13740. cluster metabolism. FEMS Microbiol Lett 326:168–172. https:// https ://doi.org/10.1073/pnas.15108 35112 doi.org/10.1111/j.1574-6968.2011.02448 .x 46. Li H, Outten CE (2012) Monothiol CGFS glutaredoxins and 63. Teplyakov A, Obmolova G, Sarikaya E et  al (2004) Crystal BolA-like proteins: [2Fe–2S] binding partners in iron homeosta- structure of the YgfZ protein from Escherichia coli suggests sis. Biochemistry 51:4377–4389. https ://doi.org/10.1021/bi300 a folate-dependent regulatory role in one-carbon metabo- 393z lism. J Bacteriol 186:7134–7140. https ://doi.or g/10.1128/ 47. Melber A, Na U, Vashisht A et al (2016) Role of Nfu1 and Bol3 JB.186.21.7134-7140.2004 in iron–sulfur cluster transfer to mitochondrial clients. eLife. 64. Navarro-Sastre A, Tort F, Stehling O et al (2011) A fatal mito- https ://doi.org/10.7554/elife .15991 chondrial disease is associated with defective NFU1 function 48. Uzarska MA, Nasta V, Weiler BD et al (2016) Mitochondrial in the maturation of a subset of mitochondrial Fe–S proteins. Bol1 and Bol3 function as assembly factors for specific iron– Am J Hum Genet 89:656–667. https ://doi.or g/10.1016/j. sulfur proteins. eLife. https ://doi.org/10.7554/elife .16673 ajhg.2011.10.005 49. Cameron JM, Janer A, Levandovskiy V et al (2011) Mutations 65. Léon S, Touraine B, Ribot C et al (2003) Iron–sulphur cluster in iron–sulfur cluster scaffold genes NFU1 and BOLA3 cause assembly in plants: distinct NFU proteins in mitochondria and a fatal deficiency of multiple respiratory chain and 2-oxoacid plastids from Arabidopsis thaliana. Biochem J 371:823–830. dehydrogenase enzymes. Am J Hum Genet 89:486–495. https ://https ://doi.org/10.1042/bj200 21946 doi.org/10.1016/j.ajhg.2011.08.011 66. Yabe T, Morimoto K, Kikuchi S et al (2004) The Arabidop- 50. Couturier J, Wu H-C, Dhalleine T et al (2014) Monothiol glutar- sis chloroplastic NifU-like protein CnfU, which can act as an edoxin–BolA interactions: redox control of Arabidopsis thaliana iron–sulfur cluster scaffold protein, is required for biogenesis of BolA2 and SufE1. Mol Plant 7:187–205. https://doi.or g/10.1093/ ferredoxin and photosystem I. Plant Cell 16:993–1007. https :// mp/sst15 6doi.org/10.1105/tpc.02051 1 51. Couturier J, Przybyla-Toscano J, Roret T et al (2015) The roles 67. Gao H, Subramanian S, Couturier J et al (2013) Arabidopsis of glutaredoxins ligating Fe–S clusters: sensing, transfer or repair thaliana Nfu2 accommodates [2Fe–2S] or [4Fe–4S] clusters and functions? Biochim Biophys Acta Mol Cell Res 1853:1513– is competent for in vitro maturation of chloroplast [2Fe–2S] and 1527. https ://doi.org/10.1016/j.bbamc r.2014.09.018 [4Fe–4S] cluster-containing proteins. Biochemistry 52:6633– 52. Mapolelo DT, Zhang B, Naik SG et al (2012) Spectroscopic and 6645. https ://doi.org/10.1021/bi400 7622 functional characterization of iron–sulfur cluster-bound forms 68. Py B, Gerez C, Angelini S et al (2012) Molecular organization, of Azotobacter vinelandii NifIscA. Biochemistry 51:8071–8084. biochemical function, cellular role and evolution of NfuA, an https ://doi.org/10.1021/bi300 6658 atypical Fe–S carrier. Mol Microbiol 86:155–171. https ://doi. 53. Tanaka N, Kanazawa M, Tonosaki K et al (2015) Novel features org/10.1111/j.1365-2958.2012.08181 .x of the ISC machinery revealed by characterization of Escheri- 69. Nath K, O’Donnell JP, Lu Y (2017) Chloroplastic iron–sulfur chia coli mutants that survive without iron–sulfur clusters. Mol scaffold protein NFU3 is essential to overall plant fitness. Plant Microbiol. https ://doi.org/10.1111/mmi.13271 Signal Behav 12:e1282023. h tt ps : // d oi .o r g/ 10 .1 08 0 /1 55 92 54. Vinella D, Brochier-Armanet C, Loiseau L et al (2009) Iron–324.2017.12820 23 sulfur (Fe/S) protein biogenesis: phylogenomic and genetic 70. Nath K, Wessendorf RL, Lu Y (2016) A nitrogen-fixing subunit studies of A-type carriers. PLoS Genet 5:e1000497. https ://doi. essential for accumulating 4Fe–4S-containing photosystem I core org/10.1371/journ al.pgen.10004 97 proteins. Plant Physiol 172:2459–2470. https://doi.or g/10.1104/ 55. Abdel-Ghany SE, Ye H, Garifullina GF et al (2005) Iron–sul- pp.16.01564 fur cluster biogenesis in chloroplasts. Involvement of the scaf- 71. Touraine B, Boutin J-P, Marion-Poll A et al (2004) Nfu2: a scaf- fold protein CpIscA. Plant Physiol 138:161–172. https ://doi. fold protein required for [4Fe–4S] and ferredoxin iron-sulphur org/10.1104/pp.104.05860 2 cluster assembly in Arabidopsis chloroplasts. Plant J 40:101–111. 56. Yabe T, Nakai M (2006) Arabidopsis AtIscA-I is affected by defi-https ://doi.org/10.1111/j.1365-313X.2004.02189 .x ciency of Fe–S cluster biosynthetic scaffold AtCnfU-V. Biochem 72. Lezhneva L, Amann K, Meurer J (2004) The universally con- Biophys Res Commun 340:1047–1052. https://doi.or g/10.1016/j. served HCF101 protein is involved in assembly of [4Fe–4S]- bbrc.2005.12.104 cluster-containing complexes in Arabidopsis thaliana 57. Mapolelo DT, Zhang B, Randeniya S et al (2013) Monothiol glu- chloroplasts. Plant J 37:174–185. https: //doi.org/10.1046/j.1365- taredoxins and A-type proteins: partners in Fe–S cluster traffick -313X.2003.01952 .x ing. Dalton Trans 42:3107. https ://doi.org/10.1039/c2dt3 2263c 73. Schwenkert S, Netz DJA, Frazzon J et al (2010) Chloroplast 58. Gelling C, Dawes IW, Richhardt N et al (2008) Mitochondrial HCF101 is a scaffold protein for [4Fe–4S] cluster assembly. Iba57p is required for Fe/S cluster formation on aconitase and Biochem J 425:207–218. https ://doi.org/10.1042/BJ200 91290 activation of radical SAM enzymes. Mol Cell Biol 28:1851– 74. Stöckel J, Oelmüller R (2004) A novel protein for photosys- 1861. https ://doi.org/10.1128/MCB.01963 -07 tem I biogenesis. J Biol Chem 279:10243–10251. https ://doi. 59. Sheftel AD, Wilbrecht C, Stehling O et al (2012) The human org/10.1074/jbc.M3092 46200 mitochondrial ISCA1, ISCA2, and IBA57 proteins are required 75. Schneider D, Berry S, Volkmer T et al (2004) PetC1 is the major for [4Fe–4S] protein maturation. Mol Biol Cell 23:1157–1166. Rieske iron–sulfur protein in the cytochrome b6f complex of https ://doi.org/10.1091/mbc.E11-09-0772 Synechocystis sp. PCC 6803. J Biol Chem 279:39383–39388. 60. Waller JC, Alvarez S, Naponelli V et al (2010) A role for tet-https ://doi.org/10.1074/jbc.M4062 88200 rahydrofolates in the metabolism of iron–sulfur clusters in 76. Maiwald D, Dietzmann A, Jahns P et al (2003) Knock-out of the all domains of life. PNAS 107:10412–10417. h t t p s : / / d o i . genes coding for the Rieske protein and the ATP-synthase delta- org/10.1073/pnas.09115 86107 subunit of Arabidopsis. Effects on photosynthesis, thylakoid pro- tein composition, and nuclear chloroplast gene expression. Plant 61. Waller JC, Ellens KW, Alvarez S et al (2012) Mitochondrial and Physiol 133:191–202 plastidial COG0354 proteins have folate-dependent functions in 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 563 77. Golbeck JH (2003) The binding of cofactors to photosystem I 94. Chibani K, Wingsle G, Jacquot J-P et al (2009) Comparative analyzed by spectroscopic and mutagenic methods. Annu Rev genomic study of the thioredoxin family in photosynthetic Biophys Biomol Struct 32:237–256. https ://doi.or g/10.1146/ organisms with emphasis on Populus trichocarpa. Mol Plant annur ev.bioph ys.32.11060 1.14235 6 2:308–322. https ://doi.org/10.1093/mp/ssn07 6 78. Hanke GT, Hase T (2008) Variable photosynthetic roles of two 95. Jacquot J-P, Eklund H, Rouhier N, Schürmann P (2009) Struc- leaf-type ferredoxins in arabidopsis, as revealed by RNA interfer- tural and evolutionary aspects of thioredoxin reductases in ence. Photochem Photobiol 84:1302–1309. https: //doi.org/10.11 photosynthetic organisms. Trends Plant Sci 14:336–343. https 11/j.1751-1097.2008.00411 .x://doi.org/10.1016/j.tplan ts.2009.03.005 79. Voss I, Goss T, Murozuka E et al (2011) FdC1, a novel ferredoxin 96. Dai S, Glauser DA, Bourquin F et al (2007) Structural snap- protein capable of alternative electron partitioning, increases in shots along the reaction pathway of ferredoxin–thioredoxin conditions of acceptor limitation at photosystem I. J Biol Chem reductase. Nature 448:92–96 286:50–59. https ://doi.org/10.1074/jbc.M110.16156 2 97. Wang P, Liu J, Liu B et  al (2014) Ferredoxin:thioredoxin 80. Hanke GT, Kimata-Ariga Y, Taniguchi I, Hase T (2004) A post reductase is required for proper chloroplast development genomic characterization of Arabidopsis ferredoxins. Plant Phys- and is involved in the regulation of plastid gene expression iol 134:255–264. https ://doi.org/10.1104/pp.103.03275 5 in Arabidopsis thaliana. Mol Plant 7:1586–1590. https ://doi. 81. Rumeau D, Peltier G, Cournac L (2007) Chlororespiration and org/10.1093/mp/ssu06 9 cyclic electron flow around PSI during photosynthesis and plant 98. Arsova B, Hoja U, Wimmelbacher M et al (2010) Plastidial stress response. Plant Cell Environ 30:1041–1051. https ://doi. thioredoxin z interacts with two fructokinase-like proteins in org/10.1111/j.1365-3040.2007.01675 .x a thiol-dependent manner: evidence for an essential role in 82. Yamori W, Shikanai T (2016) Physiological functions of cyclic chloroplast development in Arabidopsis and Nicotiana bentha- electron transport around photosystem I in sustaining photosyn- miana. Plant Cell 22:1498–1515. https ://doi.or g/10.1105/ thesis and plant growth. Annu Rev Plant Biol 67:81–106. https tpc.109.07100 1 ://doi.org/10.1146/annur ev-arpla nt-04301 5-11200 2 99. Vieira Dos Santos C, Rey P (2006) Plant thioredoxins are 83. Suorsa M, Sirpiö S, Aro E-M (2009) Towards characterization key actors in the oxidative stress response. Trends Plant Sci of the chloroplast NAD(P)H dehydrogenase complex. Mol Plant 11:329–334. https ://doi.org/10.1016/j.tplan ts.2006.05.005 2:1127–1140. https ://doi.org/10.1093/mp/ssp05 2 100. Brzezowski P, Richter AS, Grimm B (2015) Regulation and 84. Peng L, Yamamoto H, Shikanai T (2011) Structure and bio- function of tetrapyrrole biosynthesis in plants and algae. Bio- genesis of the chloroplast NAD(P)H dehydrogenase complex. chim Biophys Acta Bioenerget 1847:968–985. https ://doi. Biochim Biophys Acta Bioenerget 1807:945–953. https ://doi.org/10.1016/j.bbabi o.2015.05.007 org/10.1016/j.bbabi o.2010.10.015 101. Wang X, Liu L (2016) Crystal structure and catalytic mecha- 85. Peters JW, Broderick JB (2012) Emerging paradigms for complex nism of 7-hydroxymethyl chlorophyll a reductase. J Biol Chem iron–sulfur cofactor assembly and insertion. Annu Rev Biochem 291:13349–13359. https ://doi.org/10.1074/jbc.M116.72034 2 81:429–450. https ://doi.org/10.1146/annur ev-bioch em-05261 102. Gray J, Wardzala E, Yang M et al (2004) A small family of 0-09491 1 LLS1-related non-heme oxygenases in plants with an origin 86. Sawyer A, Bai Y, Lu Y et al (2017) Compartmentalisation of amongst oxygenic photosynthesizers. Plant Mol Biol 54:39–54. [FeFe]-hydrogenase maturation in Chlamydomonas reinhardtii. https ://doi.org/10.1023/B:PLAN.00000 28766 .61559 .4c Plant J 90:1134–1143. https ://doi.org/10.1111/tpj.13535 103. Espineda CE, Linford AS, Devine D, Brusslan JA (1999) The 87. Fristedt R, Herdean A, Blaby-Haas CE et al (2015) PHOTO- AtCAO gene, encoding chlorophyll a oxygenase, is required SYSTEM II PROTEIN33, a protein conserved in the plastid line- for chlorophyll b synthesis in Arabidopsis thaliana. PNAS age, is associated with the chloroplast thylakoid membrane and 96:10507–10511. https ://doi.org/10.1073/pnas.96.18.10507 provides stability to photosystem II supercomplexes in Arabi- 104. Hauenstein M, Christ B, Das A et al (2016) A role for TIC55 dopsis. Plant Physiol 167:481–492. https ://doi.or g/10.1104/ as a hydroxylase of phyllobilins, the products of chlorophyll pp.114.25333 6 breakdown during plant senescence. Plant Cell 28:2510–2527. 88. Dorn KV, Willmund F, Schwarz C et al (2010) Chloroplast DnaJ-https ://doi.org/10.1105/tpc.16.00630 like proteins 3 and 4 (CDJ3/4) from Chlamydomonas reinhardtii 105. Reinbothe S, Quigley F, Gray J et al (2004) Identification of contain redox-active Fe–S clusters and interact with stromal plastid envelope proteins required for import of protochloro- HSP70B. Biochem J 427:205–215. https ://doi.or g/10.1042/ phyllide oxidoreductase A into the chloroplast of barley. PNAS BJ200 91412 101:2197–2202. https ://doi.org/10.1073/pnas.03072 84101 89. Schroda M, Vallon O, Wollman FA, Beck CF (1999) A chloro- 106. Boij P, Patel R, Garcia C et al (2009) In vivo studies on the plast-targeted heat shock protein 70 (HSP70) contributes to the roles of Tic55-related proteins in chloroplast protein import photoprotection and repair of photosystem II during and after in Arabidopsis thaliana. Mol Plant 2:1397–1409. https ://doi. photoinhibition. Plant Cell 11:1165–1178org/10.1093/mp/ssp07 9 90. Liu C, Willmund F, Golecki JR et al (2007) The chloroplast 107. Ramel F, Ksas B, Akkari E et al (2013) Light-induced accli- HSP70B-CDJ2-CGE1 chaperones catalyse assembly and disas- mation of the Arabidopsis chlorina1 mutant to singlet oxy- sembly of VIPP1 oligomers in Chlamydomonas. Plant J 50:265– gen. Plant Cell 25:1445–1462. https ://d oi.or g/10.1105/ 277. https ://doi.org/10.1111/j.1365-313X.2007.03047 .xtpc.113.10982 7 91. Petti C, Nair M, DeBolt S (2014) The involvement of J-protein 108. Meguro M, Ito H, Takabayashi A et al (2011) Identification of AtDjC17 in root development in Arabidopsis. Front Plant Sci the 7-hydroxymethyl chlorophyll a reductase of the chlorophyll 5:532. https ://doi.org/10.3389/fpls.2014.00532 cycle in Arabidopsis. Plant Cell 23:3442–3453. https ://doi. 92. Hanke G, Mulo P (2013) Plant type ferredoxins and ferredoxin-org/10.1105/tpc.111.08971 4 dependent metabolism. Plant Cell Environ 36:1071–1084. https 109. Pruzinská A, Tanner G, Anders I et al (2003) Chlorophyll break- ://doi.org/10.1111/pce.12046 down: pheophorbide a oxygenase is a Rieske-type iron–sulfur 93. Terauchi AM, Lu S-F, Zaffagnini M et  al (2009) Pattern of protein, encoded by the accelerated cell death 1 gene. PNAS expression and substrate specificity of chloroplast ferredoxins 100:15259–15264. https ://doi.org/10.1073/pnas.20365 71100 110. Hörtensteiner S, Wüthrich KL, Matile P et al (1998) The key from Chlamydomonas reinhardtii. J Biol Chem 284:25867– step in chlorophyll breakdown in higher plants cleavage of 25878. https ://doi.org/10.1074/jbc.M109.02362 2 1 3 564 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 pheophorbide a macrocycle by a monooxygenase. J Biol Chem castor seed and its relationship to other di-iron proteins. EMBO 273:15335–15339. https ://doi.org/10.1074/jbc.273.25.15335 J 15:4081–4092 111. Hörtensteiner S, Kräutler B (2011) Chlorophyll breakdown in 128. Li-Beisson Y, Shorrosh B, Beisson F et  al (2013) Acyl- higher plants. Biochim Biophys Acta Bioenerget 1807:977–988. lipid metabolism. Arabidopsis Book 11:e0161. https ://doi. https ://doi.org/10.1016/j.bbabi o.2010.12.007 org/10.1199/tab.0161 112. Greenberg JT, Ausubel FM (1993) Arabidopsis mutants com- 129. Rathinasabapathi B, Burnet M, Russell BL et al (1997) Choline promised for the control of cellular damage during pathogenesis monooxygenase, an unusual iron–sulfur enzyme catalyzing the and aging. Plant J 4:327–341. https ://doi.org/10.1046/j.1365- first step of glycine betaine synthesis in plants: prosthetic group 313X.1993.04020 327.x characterization and cDNA cloning. PNAS 94:3454–3458 113. Gray J, Close PS, Briggs SP, Johal GS (1997) A novel suppressor 130. Hibino T, Waditee R, Araki E et al (2002) Functional charac- of cell death in plants encoded by the Lls1 gene of maize. Cell terization of choline monooxygenase, an enzyme for betaine 89:25–31. https ://doi.org/10.1016/S0092 -8674(00)80179 -8 synthesis in plants. J Biol Chem 277:41352–41360. https://doi. 114. Pružinská A, Tanner G, Aubry S et al (2005) Chlorophyll break-org/10.1074/jbc.M2059 65200 down in senescent Arabidopsis leaves. Characterization of chlo- 131. Yamada N, Takahashi H, Kitou K et  al (2015) Suppressed rophyll catabolites and of chlorophyll catabolic enzymes involved expression of choline monooxygenase in sugar beet on the in the degreening reaction. Plant Physiol 139:52–63. https://doi. accumulation of glycine betaine. Plant Physiol Biochem org/10.1104/pp.105.06587 0 96:217–221. https ://doi.org/10.1016/j.plaph y.2015.06.014 115. Rodoni S, Vicentini F, Schellenberg M et al (1997) Partial puri- 132. Hanke GT, Satomi Y, Shinmura K et  al (2011) A screen fication and characterization of red chlorophyll catabolite reduc- for potential ferredoxin electron transfer partners uncovers tase, a stroma protein involved in chlorophyll breakdown. Plant new, redox dependent interactions. Biochim Biophys Acta Physiol 115:677–682 1814:366–374. https ://doi.org/10.1016/j.bbapa p.2010.09.011 116. Muramoto T, Tsurui N, Terry MJ et al (2002) Expression and 133. Peden EA, Boehm M, Mulder DW et al (2013) Identification biochemical properties of a ferredoxin-dependent heme oxyge- of global ferredoxin interaction networks in Chlamydomonas nase required for phytochrome chromophore synthesis. Plant reinhardtii. J Biol Chem 288:35192–35209. https ://doi. Physiol 130:1958–1966. https ://doi.org/10.1104/pp.00812 8org/10.1074/jbc.M113.48372 7 117. Kohchi T, Mukougawa K, Frankenberg N et al (2001) The Arabi- 134. van Lis R, Baffert C, Couté Y et  al (2013) C h l a - dopsis HY2 gene encodes phytochromobilin synthase, a ferre- mydomonas reinhardtii chloroplasts contain a homodimeric doxin-dependent biliverdin reductase. Plant Cell 13:425–436 pyruvate:ferredoxin oxidoreductase that functions with FDX1. 118. Davis SJ, Bhoo SH, Durski AM et al (2001) The heme-oxygenase Plant Physiol 161:57–71. https: //doi.org/10.1104/pp.112.20818 family required for phytochrome chromophore biosynthesis is 1 necessary for proper photomorphogenesis in higher plants. Plant 135. Noth J, Krawietz D, Hemschemeier A, Happe T (2013) Physiol 126:656–669. https ://doi.org/10.1104/pp.126.2.656 Pyruvate:ferredoxin oxidoreductase is coupled to light-independ- 119. Khan MS, Haas FH, Samami AA et  al (2010) Sulfite reduc- ent hydrogen production in Chlamydomonas reinhardtii. J Biol tase defines a newly discovered bottleneck for assimilatory Chem 288:4368–4377. https://doi.or g/10.1074/jbc.M112.42998 sulfate reduction and is essential for growth and development 5 in Arabidopsis thaliana. Plant Cell 22:1216–1231. https ://doi. 136. Yasuno R, Wada H (2002) The biosynthetic pathway for lipoic org/10.1105/tpc.110.07408 8 acid is present in plastids and mitochondria in Arabidopsis thali- 120. Duncanson E, Gilkes AF, Kirk DW et al (1993) nir1, a condi- ana 1. FEBS Lett 517:110–114. https ://doi.org/10.1016/S0014 tional-lethal mutation in barley causing a defect in nitrite reduc--5793(02)02589 -9 tion. Mol Gen Genet 236:275–282. https ://doi.or g/10.1007/ 137. Cicchillo RM, Lee K-H, Baleanu-Gogonea C et al (2004) Escher- BF002 77123 ichia coli lipoyl synthase binds two distinct [4Fe–4S] clusters 121. Kimata-Ariga Y, Hase T (2014) Multiple complexes of nitro- per polypeptide. Biochemistry 43:11770–11781. https ://doi. gen assimilatory enzymes in spinach chloroplasts: possible org/10.1021/bi048 8505 mechanisms for the regulation of enzyme function. PLoS One 138. Ewald R, Hoffmann C, Florian A et al (2014) Lipoate-protein 9:e108965. https ://doi.org/10.1371/journ al.pone.01089 65 ligase and octanoyltransferase are essential for protein lipoyla- 122. Coschigano KT, Melo-Oliveira R, Lim J, Coruzzi GM (1998) tion in mitochondria of Arabidopsis. Plant Physiol 165:978–990. Arabidopsis gls mutants and distinct Fd-GOGAT genes. Impli-https ://doi.org/10.1104/pp.114.23831 1 cations for photorespiration and primary nitrogen assimilation. 139. Goyer A (2017) Thiamin biofortification of crops. Curr Opin Plant Cell 10:741–752 Biotechnol 44:1–7. https: //doi.org/10.1016/j.copbio .2016.09.005 123. Lancien M, Martin M, Hsieh M-H et al (2002) Arabidopsis glt1-T 140. Raschke M, Bürkle L, Müller N et al (2007) Vitamin B1 bio- mutant defines a role for NADH-GOGAT in the non-photorespi- synthesis in plants requires the essential iron sulfur cluster pro- ratory ammonium assimilatory pathway. Plant J 29:347–358 tein, THIC. PNAS 104:19637–19642. https ://doi.org/10.1073/ 124. Potel F, Valadier M-H, Ferrario-Méry S et al (2009) Assimilation pnas.07095 97104 of excess ammonium into amino acids and nitrogen translocation 141. Fenwick MK, Mehta AP, Zhang Y et al (2015) Non-canonical in Arabidopsis thaliana—roles of glutamate synthases and car- active site architecture of the radical SAM thiamin pyrimidine bamoylphosphate synthetase in leaves. FEBS J 276:4061–4076. synthase. Nat Commun 6:6480. https ://doi.org/10.1038/ncomm https ://doi.org/10.1111/j.1742-4658.2009.07114 .x s7480 125. Somerville CR, Ogren WL (1980) Inhibition of photosynthesis 142. Kong D, Zhu Y, Wu H et al (2008) AtTHIC, a gene involved in in Arabidopsis mutants lacking leaf glutamate synthase activity. thiamine biosynthesis in Arabidopsis thaliana. Cell Res 18:566– Nature 286:257–259. https ://doi.org/10.1038/28625 7a0 576. https ://doi.org/10.1038/cr.2008.35 126. Saha K, Webb ME, Rigby SEJ et al (2012) Characterization of 143. Zrenner R, Stitt M, Sonnewald U, Boldt R (2006) Pyrimidine the evolutionarily conserved iron–sulfur cluster of sirohydro- and purine biosynthesis and degradation in plants. Annu Rev chlorin ferrochelatase from Arabidopsis thaliana. Biochem J Plant Biol 57:805–836. h ttp s : //do i.o rg/1 0.1 146/ an nur ev .a r pl a 444:227–237. https ://doi.org/10.1042/BJ201 11993 nt.57.03290 5.10542 1 144. Hung W-F, Chen L-J, Boldt R et  al (2004) Characteriza- 127. Lindqvist Y, Huang W, Schneider G, Shanklin J (1996) Crystal tion of Arabidopsis glutamine phosphoribosyl pyrophosphate structure of delta9 stearoyl-acyl carrier protein desaturase from 1 3 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 565 2+ amidotransferase-deficient mutants. Plant Physiol 135:1314– spectroscopy identifies a [4Fe–4S] center with unusual coordi- 1323. https ://doi.org/10.1104/pp.104.04095 6 nation sphere in the LytB protein. J Am Chem Soc 131:13184– 145. Ito T, Shiraishi H, Okada K, Shimura Y (1994) Two amidophos- 13185. https ://doi.org/10.1021/ja901 2408 phoribosyltransferase genes of Arabidopsis thaliana expressed 162. Seemann M, Tse Sum Bui B, Wolff M et al (2006) Isoprenoid in different organs. Plant Mol Biol 26:529–533. https ://doi. biosynthesis in plant chloroplasts via the MEP pathway: direct org/10.1007/BF000 39565 thylakoid/ferredoxin-dependent photoreduction of GcpE/IspG. 146. van der Graaff E, Hooykaas P, Lein W et al (2004) Molecular FEBS Lett 580:1547–1552. https ://doi.or g/10.1016/j.febsl analysis of “de novo” purine biosynthesis in solanaceous species et.2006.01.082 and in Arabidopsis thaliana. Front Biosci 9:1803–1816 163. de la Luz Gutiérrez-Nava M, Gillmor CS, Jiménez LF et al (2004) 147. Lin M, Behal R, Oliver DJ (2003) Disruption of plE2, the gene Chloroplast biogenesis genes act cell and noncell autonomously for the E2 subunit of the plastid pyruvate dehydrogenase com- in early chloroplast development. Plant Physiol 135:471–482. plex, in Arabidopsis causes an early embryo lethal phenotype. https ://doi.org/10.1104/pp.103.03699 6 Plant Mol Biol 52:865–872 164. Guevara-García A, San Román C, Arroyo A et al (2005) Charac- 148. Binder S (2010) Branched-chain amino acid metabolism in terization of the Arabidopsis clb6 mutant illustrates the impor- Arabidopsis thaliana. Arabidopsis Book 8:e0137. https ://doi. tance of posttranscriptional regulation of the methyl-d -eryth- org/10.1199/tab.0137 ritol 4-phosphate pathway. Plant Cell 17:628–643. https ://doi. 149. Zhang C, Pang Q, Jiang L et al (2015) Dihydroxyacid dehy-org/10.1105/tpc.104.02886 0 dratase is important for gametophyte development and disruption 165. Hsieh M-H, Goodman HM (2005) The Arabidopsis IspH causes increased susceptibility to salinity stress in Arabidopsis. homolog is involved in the plastid nonmevalonate pathway of J Exp Bot 66:879–888. https ://doi.org/10.1093/jxb/eru44 9 isoprenoid biosynthesis. Plant Physiol 138:641–653. https://doi. 150. Flint DH, Emptage MH (1988) Dihydroxy acid dehydratase from org/10.1104/pp.104.05873 5 spinach contains a [2Fe–2S] cluster. J Biol Chem 263:3558–3564 166. Bouvier F, d’Harlingue A, Hugueney P et al (1996) Xanthophyll 151. Knill T, Reichelt M, Paetz C et  al (2009) Arabidopsis thali- biosynthesis. Cloning, expression, functional reconstitution, and ana encodes a bacterial-type heterodimeric isopropylmalate regulation of beta-cyclohexenyl carotenoid epoxidase from pep- isomerase involved in both Leu biosynthesis and the Met chain per (Capsicum annuum). J Biol Chem 271:28861–28867 elongation pathway of glucosinolate formation. Plant Mol Biol 167. Lin H, Wang R, Qian Q et al (2009) DWARF27, an iron-con- 71:227–239. https ://doi.org/10.1007/s1110 3-009-9519-5 taining protein required for the biosynthesis of strigolactones, 152. Sureshkumar S, Todesco M, Schneeberger K et  al (2009) A regulates rice tiller bud outgrowth. Plant Cell 21:1512–1525. genetic defect caused by a triplet repeat expansion in Arabidopsis https ://doi.org/10.1105/tpc.109.06598 7 thaliana. Science 323:1060–1063. https://doi.or g/10.1126/scien 168. Waters MT, Brewer PB, Bussell JD et al (2012) The Arabidop- ce.11640 14 sis ortholog of rice DWARF27 acts upstream of MAX1 in the 153. He Y, Mawhinney TP, Preuss ML et al (2009) A redox-active control of plant development by strigolactones. Plant Physiol isopropylmalate dehydrogenase functions in the biosynthesis of 159:1073–1085. https ://doi.org/10.1104/pp.112.19625 3 glucosinolates and leucine in Arabidopsis. Plant J 60:679–690. 169. Su L-W, Chang SH, Li M-Y et al (2013) Purification and bio- https ://doi.org/10.1111/j.1365-313X.2009.03990 .x chemical characterization of Arabidopsis At-NEET, an ancient 154. Vranová E, Coman D, Gruissem W (2013) Network analysis of iron–sulfur protein, reveals a conserved cleavage motif for the MVA and MEP pathways for isoprenoid synthesis. Annu Rev subcellular localization. Plant Sci 213:46–54. ht t ps : / /d oi . Plant Biol 64:665–700. https ://doi.org/10.1146/annur ev-arpla org/10.1016/j.plant sci.2013.09.001 nt-05031 2-12011 6 170. Nechushtai R, Conlan AR, Harir Y et al (2012) Characteriza- 155. Loiseau L, Gerez C, Bekker M et al (2007) ErpA, an iron–sulfur tion of Arabidopsis NEET reveals an ancient role for NEET pro- (Fe–S) protein of the A-type essential for respiratory metabo- teins in iron metabolism. Plant Cell 24:2139–2154. https ://doi. lism in Escherichia coli. PNAS 104:13626–13631. https ://doi.org/10.1105/tpc.112.09763 4 org/10.1073/pnas.07058 29104 171. Kinsman EA, Pyke KA (1998) Bundle sheath cells and cell-spe- 156. Henkes S, Sonnewald U, Badur R et al (2001) A small decrease cific plastid development in Arabidopsis leaves. Development of plastid transketolase activity in antisense tobacco transfor- 125:1815–1822 mants has dramatic effects on photosynthesis and phenylpropa- 172. Rosar C, Kanonenberg K, Nanda AM et al (2012) The leaf reticu- noid metabolism. Plant Cell 13:535–551 late mutant dov1 is impaired in the first step of purine metabo- 157. Khozaei M, Fisk S, Lawson T et al (2015) Overexpression of lism. Mol Plant 5:1227–1241. https ://doi.org/10.1093/mp/sss04 plastid transketolase in tobacco results in a thiamine auxotrophic 5 phenotype. Plant Cell 27:432–447. https ://doi.or g/10.1105/ 173. Grant K, Carey NM, Mendoza M et al (2011) Adenosine 5’-phos- tpc.114.13101 1 phosulfate reductase (APR2) mutation in Arabidopsis implicates 158. Budziszewski GJ, Lewis SP, Glover LW et al (2001) Arabidop- glutathione deficiency in selenate toxicity. Biochem J 438:325– sis genes essential for seedling viability: isolation of insertional 335. https ://doi.org/10.1042/BJ201 10025 mutants and molecular cloning. Genetics 159:1765–1778 174. Tanaka R, Hirashima M, Satoh S, Tanaka A (2003) The Arabi- 159. Estévez JM, Cantero A, Reindl A et al (2001) 1-Deoxy-d -xylu- dopsis-accelerated cell death gene ACD1 is involved in oxy- lose-5-phosphate synthase, a limiting enzyme for plastidic iso- genation of pheophorbide a: inhibition of the pheophorbide a prenoid biosynthesis in plants. J Biol Chem 276:22901–22909. oxygenase activity does not lead to the “Stay-Green” phenotype https ://doi.org/10.1074/jbc.M1008 54200 in Arabidopsis. Plant Cell Physiol 44:1266–1274. https ://doi. 160. Seemann M, Wegner P, Schünemann V et al (2005) Isoprenoid org/10.1093/pcp/pcg17 2 biosynthesis in chloroplasts via the methylerythritol phosphate 175. Yang M, Wardzala E, Johal GS, Gray J (2004) The wound- pathway: the (E)-4-hydroxy-3-methylbut-2-enyl diphosphate syn- inducible Lls1 gene from maize is an orthologue of the Arabi- thase (GcpE) from Arabidopsis thaliana is a [4Fe–4S] protein. dopsis Acd1 gene, and the LLS1 protein is present in non- J Biol Inorg Chem 10:131–137. https ://doi.org/10.1007/s0077 photosynthetic tissues. Plant Mol Biol 54:175–191. https ://doi. 5-004-0619-zorg/10.1023/B:PLAN.00000 28789 .51807 .6a 161. Seemann M, Janthawornpong K, Schweizer J et al (2009) Iso- 176. Voss I, Koelmann M, Wojtera J et al (2008) Knockout of major prenoid biosynthesis via the MEP pathway: in vivo Mössbauer leaf ferredoxin reveals new redox-regulatory adaptations in 1 3 566 JBIC Journal of Biological Inorganic Chemistry (2018) 23:545–566 Arabidopsis thaliana. Physiol Plant 133:584–598. https ://doi. 178. Zhao J, Qiu Z, Ruan B et al (2015) Functional inactivation of org/10.1111/j.1399-3054.2008.01112 .x putative photosynthetic electron acceptor Ferredoxin C2 (FdC2) 177. Li C, Hu Y, Huang R et al (2015) Mutation of FdC2 gene encod- induces delayed heading date and decreased photosynthetic rate ing a ferredoxin-like protein with C-terminal extension causes in rice. PLoS ONE 10:e0143361. https ://doi.org/10.1371/journ yellow-green leaf phenotype in rice. Plant Sci 238:127–134. https al.pone.01433 61 ://doi.org/10.1016/j.plant sci.2015.06.010 1 3

Journal

JBIC Journal of Biological Inorganic ChemistrySpringer Journals

Published: Jan 18, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off