The 2Fe–2S clusters in rhombic form possess 2 sulfide ions and 2 irons, which coordinate four cysteinyl sulfhydryl side chains (Figure 1A). This rhombic-form cluster exhibits two oxidation states: the oxidized status with two Fe³⁺⁺, and the reduced status with one Fe³⁺ and one Fe²⁺. As mentioned, the mitochondrial ISC pathway is essential for ISC biogenesis. This pathway starts with delivering iron and sulfur ions to scaffold protein ISCU2. ISCU has two isoforms: the mitochondrial isoform ISCU2 and cytosolic and nuclear isoform ISCU1. Cysteine desulfurase NFS1, which interacts with ISD11 and ACP to form a stable complex provides sulfur. However, the iron source of ISC is unknown yet. Frataxin (FXN) and ferredoxin2 (FDX2) are also important for ISCs de novo assembly. The former is thought to regulate NFS1 activity (Fox et al., 2019), while the latter is proposed to donate electrons for reduction (Cai et al., 2017; Gervason et al., 2019).

The eukaryotic cysteine desulfurase NFS1 is a pyridoxal phosphate-dependent enzyme. Sulfur is transferred from cysteine and activated into a persulfide form, which can be used for ISC assembly. In Cory et al. (2017), the first investigation into the eukaryotic NFS1 crystal structure revealed some key features: First, NFS1 binding to its substrate cysteine relies on a pyridoxal phosphate (PLP) cofactor. Second, there is a metal-binding cysteine site located in the C-terminus of NFS1. The activated sulfur abstracted is transferred to this cysteine site. Third, similar to the prokaryotic version of the enzyme, NFS1 forms a dimer. In the field of the enzymatic cycle, the PLP cofactor mediated the interaction of NFS1 and substrate cysteine. Then, NFS1 conformational change results in closing the activity site of cysteine with substrate and proceeding a nucleophilic attach (Johnson et al., 2005). During this second step, additional eukaryotic-specific subunits (ISD11 and ACP) interact with NFS1 and form a stable complex. ISD11 is a small protein of the LYR (Leu-Tyr-Arg motif) family, and ACP, which is an acyl carrier protein, regulates fatty acid synthesis. The long chain fatty acid of ACP is inserted into the helical center of the ISD11 subunit (Pandey et al., 2012; Parent et al., 2015; Cory et al., 2017). Finally, the persulfide sulfur is moved from NFS1 to the scaffold protein ISCU2 for ISC assembly.

The transfer of persulfur to the ISCU2 scaffold seems to be mediated by frataxin (Brancaccio et al., 2014; Fox et al., 2015). Frataxin is the earliest identified as a positive modulatory factor for NFS1 and ISC assembly. Frataxin interacts with NFS1 and enhance NFS1 cysteine desulfurase activity. Human neurodegenerative disease Friedreich’s ataxia results from the deficient of frataxin. Deletion of the yeast orthologous gene, Yfh1, leads to excess iron accumulation in mitochondria (Campuzano et al., 1996; Babcock et al., 1997; Karthikeyan et al., 2002). Because of frataxin’s weak Fe(II)-binding ability, it might be a potential iron donor to scaffold protein ISCU2 (Adamec et al., 2000; Yoon and Cowan, 2003). Structural studies have revealed that neither iron nor ISU oligomerization is essential for the interaction between bacterial frataxin, CyaY, and the ISU complex. Prokaryotic frataxin directly interacts with the bacterial desulfurase, IscS, but not the scaffold, IscU (Prischi et al., 2010). There are two apparently different reports for frataxin functions. One report suggested that frataxin promotes the interaction between NFS1 and substrate (Pandey et al., 2013), while frataxin was reported to promote sulfide transfer from NFS1 to ISCU1 (Parent et al., 2015). Furthermore, frataxin enhances sulfide transfer to ISCU1, forming a 2Fe–2S product on the ISCU scaffold (Bridwell-Rabb et al., 2014).

Once the initial 2Fe–2S cluster has been generated by the ISC system in mitochondria, it is transferred to the glutaredoxin 5 (GLRX5) dimer with the help of HSC20 and HSPA9. This process requires energy, which is supplied by ATP hydrolysis carried out by HSPA9 (Dutkiewicz et al., 2003). The binding of the chaperones HSC20 and HSPA9 leads to the dissociation of the assembly complex consisting NFS1/ISD11/ISCU/ACP/FXN. Competition of FXN with HSPA9 for the LPPVK binding site on ISCU acts as a molecular switch between assembly and transfer complexes (Majewska et al., 2013; Manicki et al., 2014). Cluster–bound GLRX5 includes a 2Fe–2S cluster connecting a GLRX5 dimer. Each GLRX5 contributes one cysteine ligand to the 2Fe–2S cluster, while a second thiolate ligand coming from a GLRX5–bound glutathione stably binds GLRX5 (Banci et al., 2014). glrx5 deletion in yeast cells has dysfunctional phenotypes in both mitochondrial ISC de novo biogenesis and cytosolic ISC assembly (Muhlenhoff et al., 2003; Uzarska et al., 2013).

The cubane-type cluster comprises 4 iron and 4 sulfide ions coordinated to four sulfhydryl side chains (Figure 1), which can be subdivided into low- or high-potential clusters³⁹. The oxidation states of the low-potential clusters are the oxidized [2Fe³⁺, 2Fe²⁺] and the reduced [Fe³⁺, 3Fe²⁺] forms, while the oxidation states for the high potential clusters switch between the reduced [2Fe³⁺, 2Fe²⁺] and the oxidized [3Fe³⁺, Fe²⁺] forms. Hence, the two ferric-two ferrous state is shared by the two families. Within a cell, the most common clusters are 4Fe–4S clusters. The ISA system mediates the transformation of 2Fe–2S clusters into 4Fe–4S clusters in the mitochondrial matrix. Unlikely to mitochondrial ISC biogenesis system, the cytoplasm contains different machineries to assembly 4Fe–4S clusters. The biogenesis of cytoplasmic 4Fe–4S clusters rely on substrate exported from the matrix by ABCB7 (Lill et al., 2015).

In brief, ISCA1, ISCA2, and IBA57 are responsible for the generation of mitochondrial 4Fe–4S clusters. ISCA1 and ISCA2 interact with each other (Muhlenhoff et al., 2011; Beilschmidt et al., 2017), and structure study reveals that ISCA2, but not ISCA1, is able to bind IBA57 (Muhlenhoff et al., 2011; Beilschmidt et al., 2017; Gourdoupis et al., 2018). Two GLRX5-derived 2Fe–2S clusters are converted to a 4Fe–4S cluster on the ISCA1-ISCA2 complex. This process required the presence of IBA57 and the electron transfer chain NADPH-FDXR-FDX2 (Weiler et al., 2020). Finally, NFU1 promotes the 4Fe–4S cluster transfer from the ISCA1-ISCA2-IBA57 complex to apoproteins. However, one biochemical study indicated that only ISCA1, but neither ISCA2 nor IBA57, is needed for the maturation of the 4Fe–4S cluster in mouse skeletal muscle and in primary neurons (Beilschmidt et al., 2017).

ISA system is only responsible for mitochondrial 4Fe–4S clusters biogenesis. In the absence of the ISA complex, cells showed mitochondrial function defects with uncompromised cellular viability, since the sulfur-containing component required for cytoplasmic Fe/S biogenesis still can be exported from mitochondria.

Similar to ISC biogenesis in mitochondria, the initial step of cytosolic and nuclear ISC synthesis is transient transfer a 4Fe–4S cluster to the cytosolic scaffold protein complex. This complex comprises two P-loop NTPases CFD1 and NBP35 (Roy et al., 2003; Hausmann et al., 2005; Netz et al., 2007; Stehling et al., 2018). CFD1 interact with NBP35 and form a heterotetrameric complex. This complex is able to coordinate two different types of 4Fe–4S clusters (Netz et al., 2012). One type of 4Fe–4S cluster can loosely bind to a conserved CX₂C motif that is located at the C-termini of CFD1 and NBP35. The second type of 4Fe–4S clusters bind at a ferredoxin-like CX₁₃CX₂CX₅C motif, which is located at the N terminus of NBP35. This motif is essential for NBP35 function. Interestingly, a pulse-chase experiment with ⁵⁵Fe labeled yeast cells revealed the different labilities of the two ISCs associated with the CFD1–NBP35 complex (Pallesen et al., 2013). The 4Fe–4S cluster that binds to the N terminus of NBP35 is more stable than the 4Fe–4S cluster that binds the C-terminus, which transfers the loose-binding 4Fe–4S cluster to target proteins.

Another feature of cytosolic and nuclear ISC biogenesis that is similar to mitochondrial ISCs biogenesis is the dependency on a supply of electrons (Webert et al., 2014). The electron transfer chain of the CIA system comprises NADPH, NDOR1, and the Fe/S protein CIAPIN1 (Netz et al., 2010; Banci et al., 2013). NDOR1 is a key member of the electron transfer chain. It contains NADPH-, FAD- and FMN- binding domains. Protein-protein interaction and high-throughput studies have demonstrated that CIAPIN1 physically interacts with NDOR1 (Vernis et al., 2009). Dre2 is the CIAPIN1 yeast homolog: it is a crucial component of the cytosolic and nuclear ISC biogenesis system. The synthetically lethal effect was observed when deletion Dre2 and mitochondrial iron importers (Mrs3 and Mrs4) (Zhang et al., 2008). Dre2 contains a conserved C-terminal Fe/S domain which is responsible for coordinating one 2Fe–2S or one 4Fe–4S cluster by cysteine residues. The N-terminal of Dre2 is a SAM methyl-transferase-like domain, the middle linker domain of Dre2 mediate the connection of the N-terminus and C-terminus (Zhang et al., 2008; Netz et al., 2010, 2014).

The second step of cytosolic ISC biogenesis is initialed by releasing the newly assembled 4Fe–4S cluster from the CFD1-NBP35 scaffold complex. Then, the 4Fe–4S cluster is inserted into targeted proteins (Balk et al., 2004, 2005; Song and Lee, 2008, 2011). Both the CIA targeting complex and the iron-only hydrogenase-like protein, IOP1, are essential for this step reaction. Nar1 is the yeast ortholog of IOP1. Structure study of Nar1 revealed that the four conserved Cys residues, which are in the C-terminal of Nar1 are responsible for binding ISCs. The CIA targeting complex comprises CIA1, CIA2B, and human ortholog for the yeast methyl methanesulfonate-sensitivity protein 19 (MMS19). And these components physically interact with a large number of target proteins in the cytoplasm and nucleus (Srinivasan et al., 2007; Weerapana et al., 2010; van Wietmarschen et al., 2012; Stehling et al., 2013; Kassube and Thoma, 2020). The cytosolic ISC biogenesis contains two stages. In yeast, inactivation of early stage CIA assembles complex leads to the immaturity of Fe/S protein Nar1, while defect of late-stage proteins CIA1, CIA2B, and MMS19 do not affect ISC insert to target proteins (Balk et al., 2004). Base on this study, the early and late stage of CIA system is connected by Nar1via an unknown mode of action (Stehling et al., 2013).

CIA targeting complex component CIA1 contains seven WD40-repeat domains. The structural analysis demonstrated that these seven WD40-repeats distribute around a central axis, which functions as binding region docking site of the CIA targeting complex (Srinivasan et al., 2007). Point mutation of CIA1 has revealed that the conserved, surface-exposed residue R127 is responsible for assembling other subunits of cytosolic Fe/S protein (Paul and Lill, 2015). The conserved Cys residue is important for CIA2 function, which is also conserved in eukaryotes (Weerapana et al., 2010; Luo et al., 2012; Stehling et al., 2013). Knockout human CIA2B or its ortholog Cia2 suppresses the Fe/S proteins maturation (Chen et al., 2012). MMS19 contains 4 HEAT repeats at N-terminal. As the largest component of CIA, MMS19 is associated with the multitude of biological processes. For example, impaired chromosome segregation, defective double-strand break repair via homologous recombination, and immature cytosolic and nuclear Fe/S proteins (Prakash and Prakash, 1977; Lauder et al., 1996; Kou et al., 2008; Ito et al., 2010). For a long time, it was difficult to associate these phenotypes of MMS19-deficient cells with one molecular function. Until known that MMS19 is involved in cytosolic ISC biogenesis, this problem was resolved (Gari et al., 2012; Stehling et al., 2012). As a major determinant of the CIA targeting complex, MMS19 interacts with numerous target proteins and promotes the insertion of ISCs into them, including key enzymes in DNA synthesis (POLD1, PRIM2), DNA repair [XPD, DNA2 (DNA replication helicase/nuclease 2)], and telomere length regulation (RTEL1). Deletion of these enzymes, respectively, phenocopied variant MMS19 depletion defects.

Helicases are a class of motor proteins that can unwind structured nucleic acids in an ATP-dependent manner. In this way, helicases can regulate many different processes that depend on strand separation during DNA metabolism (Lohman and Bjornson, 1996), including transcription, DNA replication, DNA repair, and telomere length regulation. Thus, helicases are important for genomic stability (Patel and Donmez, 2006; Brosh and Bohr, 2007; Lohman et al., 2008; Pyle, 2008). Helicases are classified into six super-families according to their primary amino acid sequences, and ISCs exist in numerous helicases.

In the helicase super-family 1, DNA2 is a multifunction enzyme not only involved in DNA replication, but also in double-strand DNA break (DSB) repair and telomere maintenance (Budd et al., 2005; Kang et al., 2010; Balakrishnan and Bambara, 2011). AddAB which contains a 4Fe–4S cluster, is a helicase-nuclease complex in bacteria. Eukaryotic helicase-nuclease DNA2 putative metal-binding motif was identified by sequence alignment with AddAB. Interestingly, structure studies revealed that four conserved Cys residues which are coordinated ISCs exist in the nuclease domain. This suggests that ISCs might function in stabilizing the nuclease domain conformation (Yeeles et al., 2009). Biochemical studies confirmed that yeast DNA2 coordinates ISCs by its conserved Cys residues (Pokharel and Campbell, 2012). ISC binding cysteine residues mutation results in nuclease activity and ATPase function defects in DNA2. However, the DNA binding ability of DNA2 is normal. Another biochemical study revealed that pro residue at position 504 of DNA2 is crucial to stabilize the ISC. These studies confirmed that the ISC regulates DNA2 nuclease and helicase activities by mediating conformational changes.

In the helicase super-family 2, an XPD homolog from Archaea was the first DNA repair helicase to be identified. A sequence alignment revealed that all the XPD helicase family members contain four highly conserved Cys residues. These conserved Cys residues which coordinate an ISC are crucial for 5′–3′ DNA helicases activity. The XPD helicase family comprises XPD and several related super-family 2 DNA helicases including DDX11/ChlR1 (DEAD/DEAH box helicase 11), RTEL1 (regulator of telomere elongation 1), and FANCJ (Fanconi anemia complementation group J). Many human diseases are linked to mutations in these three proteins (Table 3; White, 2009; Wu and Brosh, 2012).

XPD is a crucial subunit of the transcription initiation factor TFIIH, which is involved in NER and transcription (Compe and Egly, 2012). TFIIH comprises two major functional subcomplexes, a core complex (XPB, p8, p34, p44, p52, and p62), and a CAK (CDK–activating kinase) complex (cyclin H, CDK7, and MAT1). Helicase XPD is an important bridge between these two subcomplexes.

The mutations of XPD gene are related to three genetic diseases: xeroderma pigmentosum (XP), Cockayne syndrome (CS), and trichothiodystrophy (TTD) (White, 2009; Wu and Brosh, 2012). All three disorders have similar characteristics, with patients’ skin being hypersensitive to sun exposure. This is due to the defect of the NER pathway (Lehmann, 2003). In 2006, biochemical and spectroscopic analyses indicated that XPD coordinates a 4Fe–4S cluster, which is a key determinant of XPD helicase activity. This finding significantly contributes to revealing the molecular differences of how mutations in a single gene result in different diseases (Stehling et al., 2012). Subsequently, structural analysis showed that the 4Fe–4S domain forms a channel with an arch domain, that can accommodate single-stranded DNA (ssDNA) (Fan et al., 2008; Liu et al., 2008; Wolski et al., 2008). Mutational analysis of conserved cysteine residues in the Fe/S domain of XPD indicated that an intact Fe/S domain is essential for helicase activity and/or stabilizing the protein structure (Liu et al., 2008; Pugh et al., 2008). For patients with XP, mutations in XPD primarily inhibit helicase activity without affecting the protein structure. Interestingly, all XP-causing mutations are conserved in archaeal XPD (Fan et al., 2008; Liu et al., 2008). However, most of the mutated residues in TTD are not conserved in the archaeal protein (Liu et al., 2008). In TTD patients, R112H exchange is the most common mutation. This amino acid substitution leads to loss of XPD helicase activity and deficiency of NER (Dubaele et al., 2003). Biochemical and structural studies demonstrated that this Arg residue is essential for the Fe/S domain. These findings underscore the structural role played by the ISC in helicase activity and highlight the close relationship between ISCs and DNA replication. In addition to these disease-associated mutations, other mutations of XPD destabilize the helicase structure and compromise interactions between the two TFIIH sub-complexes (Dubaele et al., 2003; Liu et al., 2008).

FANCJ, which was able to interact with the breast cancer C-terminal (BRCT) repeats of BRCA1, is another important member of the XPD helicase family (Cantor et al., 2001). FANCJ has been identified as the gene that is mutated in the J complementation group of Fanconi anemia (FA), a genome instability disorder with an elevated risk of developing cancer. FANCJ is known as an anti-oncogene because of its functions in DNA repair (Wu and Brosh, 2009). The substitution A349P in FANCJ is a common mutation seen in patients with FA (Wu et al., 2010). Although this alanine is not a conserved site in the XPD helicase family, the residue is near the fourth highly conserved cysteine residue in the ISC. Consistent with this, recombinant FANCJ-A349P protein was shown to decrease iron content and inhibit the separation of double-stranded DNA (dsDNA) (Wu et al., 2010). This finding indicates that, like XPD helicase, the catalytic activities of FANCJ critically rely on an intact Fe/S domain.

DDX11/CHLR is the third member of the XPD helicase family. The genetic disease Warsaw breakage syndrome (WABS) arises from a mutation in the human CHLR1 gene (van der Lelij et al., 2010). In S. cerevisiae, a mutation in chl1 causes chromosome loss and unusual mating phenotypes (Liras et al., 1978). Consistent with this finding, mutations in chl1 or CHLR1, the human homolog, show similar results (Skibbens, 2004; Parish et al., 2006). Unsurprisingly, patient-derived mutations also abolish helicase activity due to their perturbance of DNA binding and DNA-dependent ATPase activity (Wu et al., 2012).

A common feature of all DNA polymerases is that they are unable to initiate de novo synthesis of a DNA strand; they can only elongate an existing strand. Synthesis of a new strand can only begin from a primer with the 3′-OH end. Hence, a primase is required to catalyze the priming, form a primer, and initiate DNA replication. Primase in eukaryotic cells comprises two subunits, the catalytic PRIM1 subunit, and a large subunit PRIM2, both interacted with DNA polymerase-α (Frick and Richardson, 2001; Kang et al., 2010). Although only the PRIM1 subunit possesses catalytic activity, PRIM2 is also crucial for primase function (Zerbe and Kuchta, 2002). Spectroscopic analysis indicates that PRIM2 is able to bind a 4Fe–4S cluster, which is conserved from Archaea to eukaryotic cells (Weiner et al., 2007). Without this ISC, its enzymatic activity is compromised. High-resolution structural studies show that the conserved Lys314 in the C-terminal domain of human PRIM2 is supported by the 4Fe–4S cluster. This Lys314 mutant abolishes primer synthesis and DNA binding (Vaithiyalingam et al., 2010). This finding suggests that ISCs facilitate DNA binding via organizing the protein surface (Vaithiyalingam et al., 2010).

In addition, ISCs serve as a major determinant for regulation through their physical interactions with other proteins involved in DNA replication, the DNA damage response, stalled replication fork, and telomere maintenance (Weiner et al., 2007). The N- and C-terminal domains of PRIM2 folded together and are connected by a flexible 18-residue linker (Baranovskiy et al., 2018). Crystal structure studies reviewed that there are three metal-binding sites in the DNA primase, a Zn²⁺-binding site, a PRIM1 catalytic site which coordinates two Mg²⁺ (or Mn²⁺) ions, and a 4Fe–4S binding site in PRIM2. Furthermore, PRIM2 has four conserved Cys residues: Cys287, Cys367, Cys384, and Cys424, which are important for coordinating ISC. Point mutation of these Cys residues cause instability of both PRIM1 and PRIM2. The unstable structure of PRIM1 and PRIM2 lead to dysfunction of DNA polymerase-α primase complex and stalled replication fork (Liu and Huang, 2015). In fact, even a single point mutation of the conserved Cys residues is sufficient to reduce the activities of DNA primase and DNA polymerase. This result indicates that ISCs have an important role in enzyme functions⁽¹²⁾.

In eukaryotes, four types of class B family DNA polymerase complexes mediate replication and replication-associated genome maintenance. During normal replication, DNA polymerases (Pol) α, δ, and ε are responsible for replication fork extension. While the fourth polymerase, Polζ, is required for DNA synthesis at damaged sites (Johansson and Macneill, 2010). These polymerases are comprised of catalytic, regulatory, and accessory subunits (Burgers et al., 2001). Biochemical and structural studies demonstrate that there are two metal-binding motifs with conserved cysteine (CysA and CysB) located at Pol α, Pol δ, and Pol ε C-terminal catalytic subunits. At first, it was reported that these two metal-binding motifs coordinate Zn²⁺ ions (Evanics et al., 2003; Klinge et al., 2009). And they are essential for the stability of replisome. However, synthetically lethal effects are observed in yeast containing a single point mutant in the Pol3 CysB motif with essential components (DRE2, NBP35, and TAH18) of CIA complex (Chanet and Heude, 2003). Furthermore, pulse-chase ⁵⁵Fe experiment, UV–Vis, and electron paramagnetic resonance (EPR) spectroscopic studies proved that the CysB motifs of all B-family DNA polymerases coordinate ISCs rather than Zn²⁺ (Netz et al., 2011; Suwa et al., 2015; Ter Beek et al., 2019). Overexpression S. cerevisiae Polδ subunit Pol31 enhances the ability of binding ISCs (Sanchez Garcia et al., 2004). Consist with this study, Polζ catalytic subunit Rev3 also coordinates the 4Fe–4S cluster in CysB. And the 4Fe–4S cluster is crucial for stabilizing the polymerase complex (Baranovskiy et al., 2018). Together, these findings suggest that the proper activities of DNA polymerases require 4Fe–4S cluster coordination. In addition to CysB, CysA is also important for the interaction between PCNA with Polδ on DNA. PCNA is a major determinant for regulating DNA replication and cell cycle. Notably, the DNA polymerase and exonuclease of Polδ were regulated by coordinated ISC (Jozwiakowski et al., 2019). In addition to class B-family polymerase complex, biochemical studies revealed that D-family polymerases also coordinate ISCs in their CysB motif. Furthermore, ISCs are important for polymerase complex formation. Point mutation of conserved Cys residue in Pol3 results in the reduction of coordinated ISC and disassociation with Polδ subunits Pol31 and Pol32. Moreover, Pol3 and the Fe/S biosynthetic genes are synthetic lethal, indicate that ISC is an essential cofactor for DNA polymerase to regulate its structure and functions (Chanet and Heude, 2003).