Skip to main content
European Commission logo print header

Regulation of iron-sulfur (Fe-S) cluster assembly in plastids and coordination with plant physiology

Final Report Summary - INTEGRREGULFESPLAST (Regulation of iron-sulfur (Fe-S) cluster assembly in plastids and coordination with plant physiology.)

• An executive summary (not exceeding 1 page).
Iron-sulfur (Fe-S) clusters are among the oldest and most versatile cofactors found in living organisms. Fe–S proteins are pivotal components of the electron transfer chains in the respiratory complexes of mitochondria and in the photosynthetic apparatus in the chloroplasts. Fe-S proteins also have catalytic roles in plastids, mitochondria and in the cytosol and nucleus.
To form their Fe-S clusters, plants have three assembly machineries which are localized in the plastids, mitochondria and cytoplasm, respectively. While the mitochondrial and the cytosolic systems are deeply studied in other prokaryotic and eukaryotic models, our knowledge regarding the green organism specific machinery located in the chloroplast is very elusive. Still, by sequence analogy, several components potentially involved in Fe-S biogenesis in the chloroplast have been isolated. They belong to the NFS, SUF and NFU protein families. Classic reverse genetic approaches are difficult because of the importance of the process these proteins are involved in. In vitro biochemical approaches are also challenging because of the requirement to keep conditions in which the Fe-S cluster is protected from aerobic damages. For all these reasons, progress in the understanding of the biology of Fe-S in the chloroplast requires novel experimental approaches.
The chloroplast Fe-S assembly system is a perfect model system to further decipher how Fe-S clusters are formed and how the system is regulated. There is a limited set of components known to be involved in Fe-S clusters biogenesis that serve to deliver these cofactors to about 30 known target proteins all within one organelle, the chloroplast. At the entry of the system is the cysteine desulfurase CpNifS and its associated SufE activator. The biochemistry of the CpNifS/SufE complex is well described and both proteins are essential. It was proposed that sulfur from SufE1 is transferred to a scaffold to initiate Fe-S cluster formation. Both Fe and S are then assembled as a cluster by a set of scaffold proteins and the cluster is delivered to the appropriate target protein.
Clearly several major questions remain, in relation to this system: How is the activity of this core machinery regulated? How does the plastidic system of Fe-S assembly respond to environmental stresses? Are Fe-S proteins a priority when Fe is limiting?
For this project we had the following objectives:
Obj. 1→ To identify proteins that interact with the CpNifS/SufE complex.

Obj. 2→ To analyze Fe-S cluster synthesis, in planta, with respect to environmental fluctuations. 
• A summary description of project context and objectives (not exceeding 4 pages).
Background:
Iron-sulfur (Fe-S) clusters are among the oldest and most versatile cofactors found in living organisms. They participate in electron transfer, catalysis, and regulatory processes. In plants 5 different cluster types exist: the ferredoxin type 2Fe-2S, the Rieske-type 2Fe-2S; the 4Fe-4S cluster, 3Fe-4S cluster and the 4Fe-4S-siroheme cluster. Differences in the ligands and in the protein environment of the Fe–S clusters allow a wide range of redox potentials. Fe–S proteins are therefore pivotal components of the electron transfer chains in the respiratory complexes of mitochondria and in the photosynthetic apparatus in the chloroplasts. However, Fe-S proteins also have catalytic roles in plastids, mitochondria and in the cytosol and nucleus.
Photosynthesis is the process that determines plant productivity and ultimately drives all life on earth. Its importance for biology, agriculture and the environment cannot be overstated and its efficiency is directly dependent on Fe-S clusters.Furthermore plant iron status is of importance for human health, since billions of people worldwide are suffering from iron deficiency, and most of the dietary iron in developing countries is derived from vegetarian food sources. Therefore a thorough understanding of how these Fe-S clusters are formed in plant chloroplasts and insight into factors that may affect cluster assembly, such as S and Fe deficiency and oxidative stress, is of utmost interest and importance.

General principles of biological Fe-S assembly:
Iron–sulfur (Fe–S) clusters are chemically simple and can be formed in vitro from sulfide and an iron salt and inserted “spontaneously” into a certain apo-protein such as ferredoxin if it has free reduced thiols (cysteines) to receive the cluster. However, most apo-proteins are inefficiently or not at all reconstituted in a test tube: in vivo all Fe-S proteins need protein cofactors to mature. The pioneering work of Dean and coworkers in nitrogen fixing bacteria and in Escherichia coli revealed a general framework for the assembly of Fe-S clusters in living cells. In vivo three steps can be distinguished: 1) mobilization of sulfur from cysteine and of Fe from as yet mostly undetermined sources, 2) assembly of a cluster on a scaffold and 3) transfer of the cluster from the scaffold to an apoprotein. Nitrogen fixing bacteria have a gene cluster (Nif operon) that encodes a specialized machinery dedicated to the synthesis of the Fe-S cofactors of nitrogen reductase. A second gene cluster termed the Isc operon mediates Fe-S assembly for other “housekeeping” Fe-S proteins in these bacteria and is also found in E. coli. In E. coli and several other bacteria another major gene cluster termed the suf operon encodes for a machinery for Fe-S assembly under sulfur limitation, iron starvation and oxidative stress conditions. All these bacterial gene clusters have in common that they include a gene encoding a cysteine desulfurase (involved in the release of S from cysteine) as well as genes encoding possible scaffolds on which clusters may be pre- formed before transfer to target apo-proteins. The need for different Fe-S cluster types probably necessitates the existence of multiple scaffolds or carriers of Fe-S intermediates and factors that can modulate the structure and redox state of clusters.
To form their Fe-S clusters, plants have three assembly machineries which are localized in the plastids, mitochondria and cytoplasm, respectively. The mitochondrial Fe-S assembly machinery in plants and other eukaryotes tresembles the bacterial Isc system, while the plastidial machinery shares several components with the bacterial suf systems. The Fe-S machinery in the cytosol of eukaryotes comprises a set of scaffolds and transfer proteins. This machinery seems to depend on the cysteine desulfurase of the mitochondria, at least in yeast and in plants.

Fe–S protein biogenesis in the plastid:
The plastids of plants harbor a complete Fe-S cluster assembly machinery which is probably optimized to function in an oxygen-saturated environment and to facilitate the biogenesis of numerous Fe-S proteins that function in photosynthetic electron transfer, nitrogen and sulfur assimilation, as well as chlorophyll metabolism. Other fundamental processes, such as protein import in plastids, have also been shown to depend on Fe–S clusters.

Plant plastids have inherited much of the Fe-S cluster assembly machinery from their prokaryotic cyanobacterial evolutionary ancestor, albeit that the genes all have transferred to the nucleus. A large number of components of the plastid Fe-S assembly machinery have now been identified. All the proteins that function in plastidial Fe-S assembly are nuclear-encoded and imported from the cytosol.

Research objectives:
Genomic approaches in the last decade have identified many new genes in model organisms. In the majority of cases either a biochemical function is proposed or a macroscopic phenotype of a mutant is described. It is now recognized that most proteins function in concert with other proteins, often forming complexes. However, very often the direct connections between biochemical functions, organization in protein assemblies and whole-plant phenotypes are not clear. This situation means that we still have only a rudimentary insight into many biological processes. To make the connections between genes, biochemical functions and whole-plant responses to environmental challenges is an exciting goal for the future of plant biology. I like to take a lead in this post-genomic challenge, using the chloroplast Fe-S assembly system as a perfect model system. There is a limited set of components that serve to supply Fe-S clusters to about 30 known target proteins all within one organelle, the chloroplast. Chloroplasts can be isolated intact and many of their biochemical functions are well-described.
While most of the components are identified, an overall understanding of how the chloroplastic Fe-S pathway is working as a whole is still far from clear. At the core of the system is the cysteine desulfurase CpNifS and its associated SufE1 activator. The biochemistry of the CpNifS/SufE1 complex is well described and both proteins are essential. It was proposed that sulfur from SufE1 is transferred to a scaffold to initiate Fe-S cluster formation. Clearly several major questions remain, in relation to this system: How do S and Fe reach the scaffolds in planta? How is the activity of this core machinery regulated? How does the plastidic system of Fe-S assembly respond to environmental stresses? 
In order to get insights into the functioning of the system in planta, future efforts have to be targeted at the identification of the in planta functions of the different proteins that are involved.
This is especially important for the possible scaffolds. As long as the in vivo involvement of potential scaffolds has not been established, the Fe-S assembly pathway is still highly hypothetic. It is likely that a certain redundancy may occur between different scaffolds and that some of the components may be required only in specific conditions. For this reason, another important challenge for the future in the field is the study of the regulation of the pathway, especially in relation to environmental stress. 
I propose to study the plastid Fe-S cluster assembly pathway, and exploit the CpNifS-SufE1 complex as a molecular entry point into the pathway. In planta protein interaction studies will reveal new candidates for the regulation of the assembly pathway and provide direct insight into the scaffolding process. The second part of the project will focus on the study of the regulation of the Fe-S cluster assembly with respect to the physiology of the plant, and will reinvestigate the in vivo function of the molecular components identified.

• A description of the main S&T results/foregrounds CONFIDENTIAL (not exceeding 25 pages),
For the project, we had 2 main objectives :

• Objective 1: To identify proteins that interact with the CpNifS-SufE1 complex.

In the present model, SufE1 activates CpNifS activity by accepting sulfur from CpNifS. Thus, SufE1 (or the complex CpNifS-SufE1) should donate S to downstream scaffolds for Fe-S assembly. In the bacterial and mitochondrial Isc systems, the cysteine desulfurase IscS binds to the scaffold IscU. However, there is no IscU in plastids. Among the possible targets for interaction with the CpNifS-SufE1 complex are the potential scaffolds Nfu1-3, IscA, Hcf101, GrxS14-16 and the SufB/C/D complex. Such interactions have not yet been demonstrated in plants.
SufE1 has a C-terminal domain with similarity to BolA. This region is not required to activate CpNifS in vitro. We consider it likely that the BolA-domain of SufE1 regulates cysteine desulfurase activity, perhaps in response to demand for Fe-S assembly. Bioinformatic analysis suggests that glutaredoxins (Grx) could be involved in such regulation. Indeed, in yeast, the Bol-A like protein FRA2 interacts with a Grx and this interaction is important for the proper regulation of the Fe-responsive transcription factor AFT1.
A first aim of this project will be to identify proteins that interact with the CpNifS-SufE1 complex. Interacting proteins will be considered as putative regulatory proteins or as scaffold proteins.

• Objective 2: To investigate the Fe-S cluster assembly pathway and its regulation in planta.
Plants grow in fluctuating environments that require constant remodeling of metabolism in order to adjust their physiology. Fe and S are both crucial elements for plant physiology. Starvation for S or Fe dramatically affects plant photosynthesis and growth. In this regard it must be noted that APS reductase (APR; adenosine 5’-phosphosulfate reductase) and sulfite reductase, two key enzymes in the S assimilation pathway in plastids, are themselves Fe-S enzymes. The chloroplast is also a major site of reactive oxygen species (ROS) production in plant cells particularly under environmental stresses or high light intensities. Oxidative stress is thought to be deleterious for photosynthetic activity, mainly through its effect on Fe-S protein stability. Moreover, ROS scavenging systems in the chloroplast are highly dependent on Fe co-factors and S-derived glutathione. It is therefore to be expected that Fe-S cluster assembly will be regulated in coordination with the physiology of the plant, particularly in the chloroplast. I consider 3 conditions that should lead to the regulation of the system:
- modification in Fe and S supply
- increased demand for Fe-S cluster proteins
- increased Fe-S cluster instability because of oxidative stress, thus increasing the need for Fe-S repair
It is also possible that the activity of key components of the Fe-S assembly pathway will be regulated in response to the activity of other components of the pathway. For instance, we propose that Fe and S are incorporated as a Fe-S cluster onto a scaffold, then transferred to carrier proteins that donate the Fe-S cluster to the appropriate target proteins. In such a system, sensing steps are required in order to provide feedback to the early steps of the process, i.e. the entry of Fe and S in the pathway. Such a hypothesis implies that alterations in the pathway should affect the regulation of the system.



→ Iron Economy and Priorities in Fe-deficient Arabidopsis Leaves (Objective 2). CONFIDENTIAL
How plants do adapt to iron depletion has been investigated for many years providing a wealth of knowledge about how iron is taken up in roots, transported through the plant to its designated sink. What is unknown is whether or not plants have an iron economy system similar to what has been proposed for copper. We have established hydroponic growth conditions that allowed us to analyse plant responses to nutrition and environmental challenges. During the outgoing phase of the project, a lot of effort has been spent on the experimental set-up allowing us to analyse plant response to Fe depletion and Fe resupply. We aimed at analysing plant response to deficient but physiologically relevant conditions and also to avoid secondary effects on plants. We considered that developing a system in which plants are able to fully recover from the Fe depletion symptoms would keep us away from severe indirect effects. To control this, all experiments are monitored by chlorophyll fluorescence measurements and other phenotypic analyses.
The experimental setting is now functional and reproducible; we collected data and samples from a dozen of subsequent experiments combining various treatment combinations (Fe depletion then Fe resupply; only, or in combination with low S nutrition, Fe excess-, hydrogen peroxide- and paraquat-induced oxidative stress). Most valuable results were obtained when plants were subjected to the Fe depletion/Fe resupply treatment. We followed plants over 7 days of Fe depletion then 7 days of resupply treatment. We detailed the analysis with 7 time-points measurements and sample collections. The physiological and phenotypic analyses are nearly fully documented with plant biomass, developmental and molecular parameters, photosynthetic activity using chlorophyll fluorescence measurement, imaging systems, and redox state analysis of the photosystem I, chlorophyll content, respiration, gas exchange and, elemental analysis.
At the phenotypic level, loss of chlorophyll became visible after 4 days of Fe depletion and could be nearly fully recovered at the end of a week of Fe-resupply. Indeed, plant growth was reduced as measured by shoot biomass production, leaf area and, root length (not shown). We followed the abundance of iron-dependent proteins through one week of iron deprivation and then through one week of iron recovery.
Because Fe has a prominent function in the chloroplast and in electron transport, we extensively analyzed the photosynthetic performance of plant subjected to our Fe depletion/Fe resupply treatment. Our initial photosynthesis analyses were conducted using the combination of two chlorophyll fluorescence devices: a imaging system allowing whole plant level studies (Fluorcam) and a clamp-based device analysis of allowing more quantitative analyses (FMS). The photosynthesis parameter Fv/Fm illustrating defects in the photosystem II was only slightly affected, yet the difference was significant confirming that PSII is relatively unaffected by Fe depletion as already reported. However, the efficiency of the PSII was dramatically reduced. The non-photochemical quenching was also strongly reduced, especially when measured at very high light, when NPQ is physiologically relevant for plants. These data suggested that some important defects have occurred in the electron transport chain downstream of the photosystem II complex. To narrow down our search, we measured oxidation-reduction rates of P700 Our data revealed that Fe deprivation had only mild effects on the PSI which remained partially functional. Our data rather pointed out a defect upstream of PSI. This is not in agreement with several reports reporting the strongest effects on the PSI subunits. Our kinetic study allowed us to unravel that there is something upstream of PSI that is more vulnerable to Fe depletion than PSI itself.
We decided to further investigate this using immunoblots and probing for photosynthesis related proteins. Pivotal for the project progress was also the early development of analytic and molecular tools. The study of the expression of proteins involved in Fe-S biogenesis as well as target proteins is of prime importance to the project. We planed to first analyze gene expression at the protein level because we consider it is the more direct approach to unravel most relevant regulatory networks. We developed, collected, and optimized (by immunopurification) antibodies raised against Fe-S assembly and target proteins, mainly the chloroplastic ones. Progress we made led us to also gathered antibodies detecting other component of the photosystem complex (non-Fe proteins), light harvesting complex proteins (LHC) and chlorophyll synthesis. We have now a set of antibody large enough (around 50 antibodies) to extensively characterize the molecular remodelling in the chloroplast when exposed to fluctuating Fe availability. Antibodies for cytosolic and mitochondrial proteins involved in Fe-dependent or Fe-independent processes allowed us comparative studies.
The very large set of immunoblots we performed on the whole kinetic of Fe-depletion and Fe-resupply revealed that most affected are proteins involved in Fe-S cluster biogenesis, photosynthesis (light reactions), sulfur assimilation as well as some proteins involved in the protection against oxidative stress. By contrast, only slight effect was observed for proteins involved in carbon (dark reactions) and nitrogen assimilation. The mitochondrial Fe-proteins (Aconitase Aco, Alternative oxidase AOX…) we tested were not as affected as are the chloroplastic, giving support for a prioritization of Fe for mitochondrial activities when it becomes scarce.
Regarding the photosynthetic proteins, it was reported that PSI was the main target of iron deficiency in the chloroplast followed by Cytb6/f and that PSII being not very affected. Indeed, PSI is the largest iron sink followed by the Cytb6/f complex in the electron transport chain. As expected, we observed a significant decreased abundance of PSI subunits, especially PsaB. However this decrease was not as dramatic as is the one of the Cytb6/f complex. The Rieske protein (PetC) and the Cytf were strongly and quickly affected, and became barely detectable after only 4 days following Fe depletion. The lack of Cytb6/f function is supported by the decrease in electrons flowing through the electron transport chain and a strong defect upstream of the PSI. Whereas Cytb6/f and to a lesser extent the PSI are affected by Fe depletion, the electron carrier between these two complexes through plastocyanin, PC, a copper protein was not affected.
Other proteins are strongly and quickly decreased in abundance after four days of Fe depletion treatment. Most of proteins involved in Fe-S biogenesis (NFU2 and 3, SufB, HCF101, GrxS14 and S16) are dramatically affected. Interestingly, SufE and CpNifS are not affected. Since these are not Fe-binding proteins, their specific function in the early S assembly may explain this observation. Notably, NFU1 does not follow the trend observed for NFU2 and NFU3 because it also appears unaffected by the treatment. Two other proteins dramatically affected are the APS reductase (APR) and the sulfite reductase (SiR) both crucial for early steps of S incorporation into Fe-S clusters. Interestingly, these proteins are themselves Fe-S proteins. Our physiological approach also combined an elemental analysis revealed that S ions specifically accumulates in response to the Fe-treatment. Therefore our integrative approach gives support for a major link between Fe and S homeostasis.
We aimed at determining a hierarchy of the effects observed on the protein levels in response to the Fe treatment. Therefore using dilution series on western blots and signal quantification using ImageJ software, we quantitatively assessed the effect for the most affected proteins.
This hierarchical classification represents precious information regarding the regulation of the system. The next question is whether the disappearance of these proteins is mediated by changes in RNA levels or solely due to protein turnover due to the lack of the Fe co-factor. Nevertheless, this approach may unravel key upstream components of Fe sensing and signaling in plants, which may act at transcriptional or post transcriptional level. This part definitively constituted a very strong integrative baseline for the project and this experimental set up allowed us the completion of all other objectives.





→ APR is a newly identify protein that interacts with the CpNifS/SufE complex and may link Fe and S homeostasis (Objectives 1 and 2). CONFIDENTIAL
The outgoing phase of the project allowed us to identify APR as a new partner for SufE. In vitro experiments first focused on the potential binding of APR to SufE suggested by previous studies in the host laboratory. Such protein interaction is of high interest because it would link a protein involved in sulfur metabolism (APR) to Fe-S cluster biogenesis (SufE). Therefore it would constitute an entry point in the regulation analysis. Co-elution, native gels and gel-filtration were the methods by which we first analyzed and identified the APR-SufE interaction. Although we gather encouraging preliminary results suggesting that APR proteins can interact with SufE when we separately expressed APR and SufE proteins using pET28 vectors and performed co-elution assays, an issue turned out to be the instability of some proteins during purification (APR2 mainly and APR1 to a lesser extent). This issue has led to difficulties in reproducibility, therefore as proposed in the project we decided to co-express proteins in E. coli using the DUET-vector system. We then quickly recover possible complexes by one-step affinity chromatography. This strategy seems to work better and allowed us to confirm SufE-APR interaction albeit the interaction is weak.
Such protein interaction is of high interest because it links a key protein involved in sulfur metabolism to Fe-S cluster biogenesis. SufE1 protein is composed of 2 domains, an E-domain conserved among all SufE proteins and an additional BolA-like domain. APR contains a GRX domain. In order to test this, we produced deletions and site mutagenesis versions of SufE1 and APR1, targeting Bol-A and Grx domains, respectively. The mutagenesis combined with biochemical assays revealed that the SufE1-APR interaction likely occurs through the Bol-A domain of SufE and the Grx domain of APR (not shown).
Grx are usually classified as monothiols (one Cys) or dithiols (two Cys) proteins, based on the active site motif (CGFS or CxxC) although recent comparative genomics analysis has shown a greater variety of Grx isoforms bringing the total to six classes. According to the TAIR website, APR1 and the two other isoforms in Arabidopsis have dithiol motifs. Therefore, we looked for other similar dithiols and monothiol Grx proteins in the Arabidopsis genome expressed in plastids or mitochondria, such as GrxS12, GrxS14, GrxS15, GrxS16 and GrxC5. GrxS12, a monothiol (CSYS), is localized to the chloroplast making it a good candidate for a potential interacting partner within the larger protein complex containing CpNifS and SufE1. GrxS12 is also in class I with APR1, having only one Grx domain motif. GrxC5 is another plant Grx that belongs to the dithiol Grx, which is localized to the chloroplast. However when co-expressed with SufE1 using the Duet system, we could no detect any evidence of co-elution of these proteins with SufE1 (not shown). GrxS14, GrxS15 and GrxS16 all belong to monothiol Grx (CGFS Class). GrxS14 and S16 are predicted to be localized to the chloroplast, while the GrxS15 would be localized to the mitochondria. Based on preliminary biochemical assays, the only Grx that showed some evidence of a slight interaction is GrxS15 (not shown). All together these results suggest that the interaction of SufE1 and APR proteins in the chloroplast is rather specific, since none of the tested plastidial Grx show any interaction. However, since SufE1 is proposed to be dual-localized in the plastid and in the mitochondria, the interaction between SufE1 and GrxS15 may be biologically relevant. More work will be required to address that question.
Such BolA-GRX domain interaction has been documented in yeast and plays important function in Fe sensing. A SufE-APR interaction would suggest a control of Fe-S cluster biosynthesis by Fe and/or S nutrition. Our physiological and molecular approaches revealed that APR proteins very quickly disappear during Fe starvation. This observation would lend support for a Fe-dependent modulation of the cysteine desulfurase activity. Pilot experiments revealed that NifS activity in the chloroplast is also modulated by Fe supply, but importantly negatively correlates with the abundance of APR proteins. Interestingly, Fe does not affect the abundance of NifS and SufE proteins. Therefore, the observed modulation of CpNifS activity has to be mediated by post-translational modifications, potentially through the observed protein-protein interaction. Such a model would be supported by the various results we collected suggesting a strong impact on S assimilation.
To test a potential in vivo effect of APR proteins on the cysteine desulfurase activity in the chloroplast, we initiated the phenotypic analysis of APR KO mutants. Preliminary in vitro culture experiments revealed that depletion of APR proteins (through deletion of APR1 or APR2, the 2 most expressed APR in plants) results in an increased sensitivity to Fe deficiency (when ferrozine, an iron chelator, is added to t he media). While both mutants were undistinguishable from the wild type when grown under control conditions, both exhibited shorter roots and yellow leaves when Fe was limiting. More detailed analysis regarding the cysteine desulfurase activity in the apr mutants will now be required to firmly link APR proteins to Fe homeostasis and more specifically to the regulation of cysteine desulfurase activity when Fe is limiting. Another strategy to test the in vivo function of the APR-SufE interaction is the use of truncated versions of SufE. Several transgenic lines expressing only the E-domain or only the BolA-domain in the WT background as well as in the sufe1 KO background are now available for further investigation. It has been reported that the constitutive overexpression of SufE full-length is deleterious for the plant. Indeed, in a context where APR binds SufE to modulate the cysteine desulfurase activity, the overproduction of SufE might disturb the regulatory network. Therefore, as previously mentioned, we complemented the lethal SufE KO with constructs expressing only the conserved E-domain or only the BolA-domain in transgenic plants under the control of a 35S promoter. These lines, when compared to the SufE-FL also driven by the 35S promoter, will help us understanding why the overexpression of SufE has a negative effect. The construct expressing only the BolA domain is also investigated in the WT background in order to test any potential dominant negative effect of the BolA-domain. Most of these lines have been validated for expression of the transgene. Molecular and physiological analysis is initiated. Alternative approach such as in vitro assays testing the effect of APR on the capacity of SufE to activate the cysteine desulfurase CpNifS will be conducted using the different truncated proteins we expressed in E.coli.
All together, our results allowed us to build up a working model where APR interaction with SufE would constitute a molecular checkpoint coordinating S assimilation and Fe-S cluster biogenesis. This hypothetical model still needs to be challenged using the various genetic tools we isolated.


→ The plastidial SufA protein is important for plant to cope with Fe deficiency and may be directly involve in the maintenance of the Fe-sensitive cytochrome b6/f complex (Objective 2). CONFIDENTIAL

The study of the in vivo function of hypothetical scaffold/carrier protein such as the SUF, NFU and NFS was also reinvestigated. Several T-DNA KO mutants have been isolated. However, because the constitutive absence of essential proteins may lead to seedling lethal plant, as does a few of the T-DNA insertion mutants, we designed inducible artificial RNA interference constructs for conditional suppression of SufB (SufBCD complex), NFU2, NFU3, NFU2 and NFU3, SUFE1, or HCF101 proteins. We used a system allowing the production in planta of artificial microRNAs (amiRNA) and designed two different amiRNA for each targeted protein. Given the 6 different targeted proteins, we designed 12 constructs in total. These constructs have been sequenced and used for plant transformation. We isolated several mono-insertional and homozygous transformants. The host lab in France will now focus on the final characterization of these lines. The selection of these lines has taken more time than expected because I decided at the mid-term milestone to investigate the function of Fe-scaffold/carrier proteins such as Nfu2, Nfu3 and SufA using viable T-DNA KO lines. Nethertheless, the artificial amiR lines might be used to confirm the results obtained with the T-DNA mutants, but more importantly will substitute T-DNA mutants for the study of essential components of the machinery.
We used the Fe deficiency/Fe resupply strategy to analyse the different scaffold mutants we had. It confirmed that nfu2 and nfu3 have clear phenotype, even on sufficient Fe supply, with reduced growth and photosynthetic defects (not shown). The host laboratory in France is studying the phenotypic characterization of these mutants since already a couple of years; therefore in the context of that project, I primarily focused on trying to unravel SufA1 function in the chloroplast. SufA importance in shoots is debated since several works could not report any phenotype associated with the mutation. SufA protein is mainly expressed only in green tissues, with a higher abundance in organs with active photosynthesis. As previously mentioned, SufA1 protein very quickly disappears upon Fe depletion, as does most of the protein involved in Fe-S biosynthesis. Our Fe depletion/Fe resupply experiment allow us to reveal that sufa mutant (2 independent alleles) cannot recover from Fe-deficiency in the same extent as do all other genotypes, including the WT. Sufa mutant plants were more affected by Fe depletion and could not recover as well as did the WT and several scaffold mutants we tested. A function for SufA protein in response to Fe deficiency could make sense in regard to the observed strong decrease in SufA protein levels during Fe depletion. Such function for SufA-like proteins during Fe deficiency has been proposed in some prokaryotes. This result suggests that SufA1 protein might be important for plants to cope with Fe deficiency, or important to enable recovery after Fe is back to the media.
We extended our study of the sufa mutants using another experimental set up allowing us to modulate Fe supply. We ended up growing plants on soil, complemented with calcium oxide, which increases the pH in the soil. Because Fe is poorly bioavailable for plants in alkaline soil, this represents a convenient way to mimic a Fe starvation. Indeed, using alkaline soil we confirmed that sufa mutants (2 independent alleles) are more sensitive to Fe deficiency. Although the mutant was undistinguishable from the WT in control soil (acidic, pH 6, CaO 0%), it appears more affected over the time (dwarf, chlorotic) when the pH of the soil increases. Sufa mutant growth was more affected and chlorosis was more pronounced than the WT for a given CaO concentration in the soil. Soil pH ranged from 6 (0% CaO) to 8 (0.8% CaO). The strongest symptoms were indeed observed at highest pH, where growth of the mutant eventually is arrested. Importantly, although CaO treatment is commonly used in soil to mimic a Fe deficiency condition, increasing pH does not only affect Fe availability. Therefore we tested if the phenotype would be abolished when an external Fe-EDDHA chelate is applied. Feeding plants with a Fe-EDDHA solution (0.5g.l-1) allowed us to make sure the observed defects are indeed due to Fe and not to a general effect on the availability of several nutrients. When Fe-EDDHA was applied, sufa remained as green as the WT did regardless of the CaO concentration.
To get further insights into the molecular mechanism underlying the defects observed in sufa mutant when grown in Fe limiting conditions, we analysed its photosynthetic performances. Several photosynthetic parameters confirmed that electron transport is more impaired than the WT in alkaline soils.
Further molecular analyses using the antibody collection allowed us to identify a strong and specific defect in the abundance of the Cytb6/f complex (. The [2Fe-2S]-Rieske protein PetC appeared to be the most affected, followed by the Cytf. Since we already identified the Cytb6/f complex as one of the most sensitive complex to Fe deficiency, a major question that remains is whether SufA has a specific function in the synthesis/ maintenance of Rieske protein of that complex or, SufA has a more global function during Fe deficiency. Our preliminary data would favor a specific function related to the maturation of the Rieske protein since other Fe proteins shown to be affected by Fe starvation, such as SufB, are not altered in sufa mutants.



→ The Fe deficiency/ Fe resupply system as a promising strategy to uncover new targets for Fe-S scaffold proteins: the plastidial [2Fe-2S] HCAR protein as a potential direct target for NFU2 (Objective 2). CONFIDENTIAL

The experimental setting we set up appeared as an excellent tool to revisit the function of hypothetical components of Fe-S biogenesis. Because several Fe-S target proteins were decreased upon Fe depletion, and importantly they were resynthesized after Fe resupply, it gave us the opportunity to screen for proteins which would not recover in absence of one of the assembly components. These would likely allow us to identify specific targets for the corresponding scaffold protein.
Therefore, we analysed the de novo Fe-S protein synthesis in the WT as well as in various T-DNA mutants deleted in Fe-S scaffold/carriers. As a proof of concept, we used nfu2 mutant since it is already characterized for some defects in specific PSI’s Fe-S subunits. In addition to the validation of proteins known to be affected in nfu2 background (such as PSI’s Fe-S components), we unravel a new protein, which cannot be resynthesized in absence of NFU2. This candidate protein cross-reacts with the anti-HCAR antibody. HCAR is a plastidial Fe-S protein involved in chlorophyll metabolism. However the identified protein exhibits a larger size than expected when run into a SDS-PAGE gel. Extensive work is actually on-going to firmly identifying this protein. We have now biological material to test if this protein is indeed HCAR (recombinant HCAR protein, plant thylakoid extracts form WT and hcar mutant). If it is not HCAR, then alternative strategies will be developed to identify the nature of this protein, possibly through proteomic.
This promising strategy to uncover targets has been extended to other scaffold mutants we gathered. We concomitantly tested nfu3 mutant (which also exhibit growth defects), mutants in the two plastidial glutaredoxins GrxS14 and S16, mutant in the SufA1 protein as well as mutant in the two most important APR proteins (APR1 and APR2). The sample collection still needs to be analysed.


• The potential impact (including the socio-economic impact and the wider societal implications of the project so far) and the main dissemination activities and exploitation of results (not exceeding 10 pages).

This project is unique because it really links protein functions to whole plant physiology and vice versa. I focus on the model system of Fe-S cluster assembly in plastids, with special emphasis on the regulation of early steps of the process, when an Fe-S cluster is first assembled through the cysteine desulfurase activity. This work will advance understanding of an important understudied aspect of the Fe utilization: its regulation. Moreover, we revealed that key components may serve as points of cross-talk between Fe-S cluster synthesis and nutrient availability of Fe and S. In addition, we may have uncover a specific function for a Fe-S carrier protein important for photosynthesis when Fe is limiting. Thus, this system forms an exciting and challenging model to study in the context of the physiology of the whole plant.
Plants are sessile and must adjust their biochemistry to all possible environmental challenges. Photosynthesis is pivotal to biomass production and it is important that we understand its limiting factors in view of the pressures imposed by environmental impacts and global change. As C, N and S assimilation in plants all depend on Fe-S proteins, progress here is likely to be of importance for a number of research areas. In the particular case of the BolA domain interactions, the project will also address the question of its conservation as a general regulatory feature in living organisms. The relevance of these studies is not only fundamental, but should be very applicable. Insight into Fe-S formation in plastids will facilitate the breeding of plants with enhanced Fe levels, enhanced photosynthesis and other Fe-dependent processes such as S assimilation. Such crops with higher productivity and nutritional value would be of extreme significance in the future.
This work will lead to new perspectives on how the environment controls development in plants and has broad applications for agriculture and human health through impacts on nutrition and the environment.


• The address of the project public website, if applicable as well as relevant contact details.
Fellow (present contact details):
Dr Karl RAVET, Research Scientist II, Biology Department, Colorado State University, Fort Collins, 200 West Lake Street, Campus Delivery 1878, 80523-1878 CO, United States.
kravet@lamar.colostate.edu