Iron-sulfur cluster biogenesis, trafficking, and signaling: Roles for CGFS glutaredoxins and BolA proteins

Abstract

The synthesis and trafficking of iron-sulfur (Fe-S) clusters in both prokaryotes and eukaryotes requires coordination within an expanding network of proteins that function in the cytosol, nucleus, mitochondria, and chloroplasts in order to assemble and deliver these ancient and essential cofactors to a wide variety of Fe-S-dependent enzymes and proteins. This review focuses on the evolving roles of two ubiquitous classes of proteins that operate in this network: CGFS glutaredoxins and BolA proteins. Monothiol or CGFS glutaredoxins possess a Cys-Gly-Phe-Ser active site that coordinates an Fe-S cluster in a homodimeric complex. CGFS glutaredoxins also form [2Fe-2S]-bridged heterocomplexes with BolA proteins, which possess an invariant His and an additional His or Cys residue that serve as cluster ligands. Here we focus on recent discoveries in bacteria, fungi, humans, and plants that highlight the shared and distinct roles of CGFS glutaredoxins and BolA proteins in Fe-S cluster biogenesis, Fe-S cluster storage and trafficking, and Fe-S cluster signaling to transcriptional factors that control iron metabolism--.

Keywords: BolA; Glutaredoxin; Glutathione; Iron homeostasis; Iron regulation; Iron-sulfur cluster biogenesis.

Figure 1.

Structures of a typical class I dithiol Grx (A) and a class II monothiol Grx (B). A, NMR structures of E. coli Grx1 in the reduced form (PDB entry 1EGR), the disulfide oxidized form (PDB entry 1EGO), and the glutathionylated form (PDB entry 1GRX). NMR structure of E. coli Grx4 in the apo form (PDB entry 1YKA) and the X-ray crystal structure of the [2Fe-2S]-GSH₂-bridged homodimer form (PDB entry 2WCI). The variable-length loop dictating Grx function is located between ß1 (in dark blue) and the N-terminal active site Cys (Cys-11 in Grx1 and Cys 30 in Grx4) [7].

Figure 2.

Domain structures of Class II Grxs from bacteria (Escherichia coli), trypanosomes (Trypanosoma brucei), yeast (Saccharomyces cerevisiae), plants (Arabidopsis thaliana), and mammals (Homo sapiens). The conserved cysteines in the Trx and Grx domains are shown in red. Additional conserved residues are shown in black. In eukaryotes, mitochondrial CGFS Grxs are single domain while cytosolic versions are multi-domain. Plants additionally have both single and multi-domain CGFS Grxs located in the chloroplast. Please note that the active site motifs of both cytosolic and mitochondrial Class II Grxs from trypanosomes (CAYS and CGFT) deviate from the canonical CGFS motif [117].

Figure 3.

Diagrams depicting the subcellular localization of Grx and BolA homologs (dark red text) in bacteria, yeast, plant, and human cells.

Figure 4.

Structures of human apo-BOLA proteins, [2Fe-2S]-GLRX5 homodimer, and [2Fe-2S]-GLRX5-BOLA1/3 heterodimers. The invariant His in BolA proteins is shown as red text. (A) Human apo-BOLA2 modeled on NMR structure of mouse BOLA2 (PDB entry 1V9J) (left) and NMR structures of human apo-BOLA1 (PDB entry 5LCI) (center) and human apo-BOLA3 (PDB entry 2NCL) (right). (B) X-ray crystal structure of human [2Fe-2S]-GLRX5 homodimer (PDB entry 2WUL) and structural models of [2Fe-2S]-GLRX5-BOLA1 (center) and [2Fe-2S]-GLRX5-BOLA3 (right) heterodimers [35]. The GLRX5 monomers in the heterodimer structures in the middle and right were aligned with the left protomer of the [2Fe-2S]-GLRX5 homodimer using PyMol [118].

Figure 5.

Diagram depicting the ISC pathway in yeast mitochondria highlighting the central role of Grx5 and the proposed roles of Bol1/Bol3 proteins. A [2Fe-2S] cluster is assembled on the scaffold protein Isu1 by the early ISC machinery, which is then delivered to the intermediate carrier protein Grx5. The chaperone/co-chaperone complex Jac1/Ssq1 is proposed to assist in release of the cluster from Isu1. Grx5 may deliver [2Fe-2S] clusters to apo targets in the mitochondria and is also required for formation of an unidentified sulfur-containing species (X-S) that exported to the cytosol. X-S is used by the CIA machinery for nuclear/cytosolic Fe-S cluster assembly and is required to relay the cellular iron status to iron-responsive transcription factors in the nucleus. In addition, Grx5 interacts with and transfers [2Fe-2S] clusters to the late ISC machinery, a complex composed of Isa1, Isa2, and Iba57, which then synthesizes [4Fe-4S] clusters via coupling of two [2Fe-2S] clusters. Nfu1, Ind1, Bol1, and Bol3 serve as targeting factors that interact with Grx5 and/or the late ISC machinery to deliver [4Fe-4S] clusters to specific apo acceptor proteins.

Figure 6.

Models for iron-responsive transcriptional activation by Aft1/2 in S. cerevisiae (A) and iron-responsive transcriptional repression by Fep1 and Php4 in S. pombe (B). When iron levels are low in S. cerevisiae (A, top), Aft1 and Aft2 bind DNA and activate expression of iron acquisition genes. During conditions of iron sufficiency (A, bottom), Grx3 and Grx4 form Fe-S-bridged heterodimers with Bol2 and deliver [2Fe-2S] clusters to Aft1/2, which promotes their DNA dissociation, dimerization, and export to the cytosol. This change, in turn, leads to deactivation of Aft1/2-regulated genes. During iron deficiency in S. pombe (B, top), Php4 binds to Php2/3/5 which represses transcription of genes encoding proteins in non-essential, iron-utilizing pathways. There are conflicting reports whether or not Grx4 maintains an interaction with Php4 during iron deficiency. In either case, Grx4 does not interfere in Php4’s association with Php2/3/5. Fep1 dissociates from its target genes when iron is low, leading to the expression of iron acquisition genes. The dissociation of Fep1 from DNA is presumably triggered by [2Fe-2S] cluster transfer from Fep1 to the Grx4-Fra2(Bol2) heterodimer. When iron levels increase (B, bottom), Grx4 and Php4 bind a [2Fe-2S] cluster that promotes dissociation of Php4 from Php2/3/5 and its nuclear export, leading to derepression of the Php4 regulon. In parallel, Fep1 maintains an interaction with Grx4-Fra2 and binds [2Fe-2S] clusters that facilitate its DNA binding activity and repression of its target genes.

Figure 7.

Model for the roles of GLRX3 and BOLA2 in mammalian cytosolic iron trafficking. The cytosolic iron chaperone PCBP1 acquires Fe(II) from the labile iron pool and forms an Fe(II)-bridged complex with BOLA2 that requires GSH binding for stability. This complex is proposed to interact with apo-GLRX3 to form a GLRX3-BOLA2 heterocomplex. Each Grx domain in GLRX3 forms a [2Fe-2S]-binding complex with a BOLA2 monomer and GSH leading to formation of a [2Fe-2S]₂-bridged GLRX3-BOLA2₂-GSH₂ heterotrimer. The sulfur source required to form the [2Fe-2S] clusters is unclear, possibly provided by the cytosolic isoform of the cysteine desulfurase NFS1 or exported from the mitochondrial ISC pathway. Both the [2Fe-2S]₂-GLRX3-BOLA2₂ heterotrimer and the [2Fe-2S]₂-GLRX3₂ homodimer can deliver [2Fe-2S] clusters to the CIA Fe-S cluster assembly protein CIAPIN1 in vitro, although both GLRX3 and BOLA2 are required for Fe incorporation into CIAPIN1 in cell-based assays. [2Fe-2S]₂-GLRX3₂ may also deliver clusters to the CIA scaffold protein NUBP1, which requires reductive coupling of [2Fe-2S]²⁺ clusters to form [4Fe-4S]²⁺ clusters. CIAPIN1, its functional electron transfer partner NDOR1, and NUBP1 are all components of the CIA machinery that assembles and delivers [4Fe-4S] clusters to target proteins in the cytosol and nucleus.