In engineering, control systems regulate the operation of devices and their processes using control loops. For a functioning control loop, you need to measure a process value which can be either below or above a target set point. The device or process then needs to be adjusted to attain the desired process value. In a similar manner, bacteria have evolved to control the intracellular concentrations of nutrients and metabolites, including iron, to pre-set conditions required for growth and/or maintenance. For this process, called homeostasis, bacteria produce sensors that measure the amount of intracellular iron, and a control system (or systems) that can change the expression of proteins and the functionality of enzymes in order to reach the target value. This control occurs at: transcriptional (DNA→ mRNA), post-transcriptional (stabilizing or degradation of mRNA), translational (mRNA→ protein) or post-translational level (protein degradation, modification, and allosteric interaction). These four levels allow very precise tuning and distribution of iron, depending on necessity and environmental conditions. Precise tuning is important because Fe is essential for life, while at the same time Fe²⁺ can generate toxic reactive oxygen species (ROS) with O₂, and Fe³⁺ is insoluble under neutral aqueous conditions.

In Staphylococcus aureus, Fur (ferric uptake regulator) is the major control system for iron (Figure 1). The Fur protein is homodimeric, with each monomer consisting of an N-terminal DNA binding domain and a C-terminal dimerization domain (Price and Boyd 2020). Between the two domains is a metal ion binding site, which was recently shown to bind a [2Fe-2S] cluster in E. coli (Fontenot et al., 2020). S. aureus Fur has been described as binding two separate Fe²⁺ ions in the hinge regions between the N- and C-terminal domains. E. coli and S. aureus Fur share 30% sequence identity and 49% sequence similarity, which includes three conserved Cysteines (Supplementary Material). The precise nature of the Fe-bound form of S. aureus Fur remains to be established. Upon dimerization, the DNA binding part of Fe-bound Fur is a transcriptional repressor of a range of genes related to iron homeostasis. It functions by binding to a so-called Fur-box upstream of the coding genes. These genes involve Fe transporters and many other genes, as discussed below. Some are established virulence factors, which means they are involved in disease processes. Related proteins, called Fur family proteins, with affinity to other metal ions or compounds have been discovered, such as Zur (zinc uptake regulator) for Zn²⁺ and PerR (peroxide operon regulator) which is a metal-dependent regulator for hydrogen peroxide. The Fe²⁺ and Mn²⁺ dependence of PerR highlights crosslinks between the different control systems (Horsburgh, Ingham, and Foster 2001).

Ferrous iron uptake by transporter FeoB is still poorly characterized, although recently inhibitors have been identified that may have important medical implications as novel antibiotics against MRSA and other multidrug resistant Gram-positive bacteria (Shin et al., 2021). Fur and PerR differentially regulate the S. aureus ferritin gene (FtnA), which encodes the Fe-storage protein ferritin (Morrissey et al., 2004). Ferritin can take up circa 4,000 Fe atoms in the form of a ferrihydrite mineral nanoparticle inside a 24-meric protein sphere (Honarmand Ebrahimi, Hagedoorn, and Hagen 2015). Iron storage by ferritin involves oxidation of Fe²⁺ to Fe³⁺ and concerted incorporation into a growing ferrihydrite mineral core. Upon reduction, by a mechanism that has not yet been established, Fe can also be released from ferritin as Fe²⁺.

Another Fur controlled Fe uptake system is the FepABC (Fe dependent peroxidase) transporter. The transporter has not been well characterized, but it has been implicated in Fe and possibly heme uptake. The transporter consists of FepA, a predicted membrane anchored lipoprotein that may act as an Fe (compound) binding protein, peroxidase FepB, and integral membrane protein FepC (Biswas et al., 2009). FepB can bind heme and protoporphyrin IX (heme without Fe) and has a low peroxidase activity (Turlin et al., 2013). FepB is a substrate of the Twin-Arginine Translocation pathway, which allows membrane translocation of fully folded cofactor bound proteins (Biswas et al., 2009). Heterologous expression of the S. aureus FepAB in E. coli allowed heme utilization in this organism (Turlin et al., 2013).

Siderophores are small, extracellular, peptide-derived compounds with a high affinity for Fe³⁺. S. aureus produces two siderophores: Staphyloferrin A (SA) and Staphyloferrin B (SB). SA is produced using the gene cluster sfa, and SB is produced using the sbn gene cluster (Marchetti et al., 2020). The gene clusters are both transcriptionally repressed by Fur. Recently, a heme sensitive regulator of siderophore production was identified: SbnI (Laakso et al., 2016; Verstraete et al., 2019). The gene product of SbnI is an enzyme producing a precursor to the siderophore. Furthermore, the protein can bind DNA and upon dimerization contains a heme binding domain. Heme transfer from IsdI to SnbI has been suggested to help control the production of siderophores, thereby shifting focus to heme utilization rather than free iron uptake. Fe³⁺ bound SA is taken up using the ABC transporter HtsABC (heme transport system). Despite the name, it is unclear if the Hts transporter is involved in heme uptake, and if so by which mechanism (Price and Boyd 2020). Hts transcription is regulated by Fur. SB is taken up by the ABC transporter SirABC (Grigg et al., 2010a).

Siderophores play a very important role in Staphylococcus biofilm formation to ensure Fe availability (Johnson et al., 2005; Oliveira et al., 2021). Fe chelators that compete with siderophores can disturb biofilm formation and may therefore be of medical importance (Richter et al., 2017; Coraça-Huber et al., 2018).

Heme obtained from red blood cells is a major source of iron for Staphylococcus aureus during infections (Skaar et al., 2004). There is evidence that S. aureus has evolved a specificity towards human hemoglobin versus other mammalian orthologs that is unique among pathogenic bacteria (Pishchany et al., 2010). However, a high level of intracellular heme is dangerous due to its potential to form reactive oxygen species (ROS). S. aureus uses the two-component regulator HssRS (Figure 1) (Heme sensing two-component system) to control the intracellular level of free heme (Price and Boyd 2020; Stauff, Torres, and Skaar 2007; Stauff and Skaar 2009). HssS is a transmembrane protein, which responds to the heme level by an unknown mechanism. Upon activation, HssS acts as a histine kinase to phosphorylate the histidine of HssR, thereby activating the protein as a transcriptional activator of the heme efflux transporter HrtAB (Heme regulator transporter efflux pump). Whether the precise compound that is expelled by HrtAB is heme or a heme metabolite is unknown (Price and Boyd 2020). Additional targets of HssR have not been identified to date.

The cell wall of S. aureus contains a unique system to acquire heme, which is called Isd (Iron-regulated surface determinant system) (Skaar and Schneewind 2004; Grigg et al., 2010b; Mazmanian et al., 2003). Similar systems are present in other Gram-positive pathogenic bacteria, such as Listeria monocystogenes and Clostridium tetani. This system takes up heme from human (host) hemoproteins. The Isd system involves nine different proteins, of which four are bound to the cell wall: IsdA, IsdB, IsdC, and IsdH. Two proteins, IsdE and IsdF, constitute an ABC-transporter for the heme cofactor. IsdD is a transmembrane protein of unknown function. And the final two proteins, IsdI and IsdG, are soluble intracellular heme degrading enzymes. The outer cell wall proteins IsdB and IsdH bind free heme, methemoglobin and hemoglobin-haptoglobin complexes from the host. The cell wall proteins IsdC and IsdA are involved in heme transport through the 15–30 nm thick cell wall to the ligand binding component of the ABC-transporter IsdE. After translocation of the heme to the cytoplasm, the cofactor is degraded by the heme degrading enzymes IsdI and IsdG. These enzymes are distantly related to well-characterized heme oxygenases and have been found to release Fe from the cofactor, yet the precise reaction mechanism remains to be solved (Grigg et al., 2010b). Fur regulates the expression of the genes for IsdA, IsdB, IsdC and IsdH. The gene for the enzyme sortase B (SrtB) is part of the same transcriptional unit as IsdC, and therefore also regulated by Fur. Sortase B is involved in the cell wall anchoring of the Isd components.

Upon infection, the host starts the immune response. Macrophages are activated by interaction with S. aureus via Toll-like receptors (TLRs) (Pidwill et al., 2021). This starts signal transduction cascades which include mechanisms to limit the availability of iron in blood (Pandur et al., 2021). Central in the regulation of these processes is the hormone hepcidin. Hepcidin interacts with the host Fe efflux protein ferroportin, thereby limiting Fe export from macrophages (Theurl et al., 2008). Interestingly the same exposed S. aureus lipoproteins that support Fe acquisition, for example, via the Isd system, are also recognized by the TLRs, thereby evoking inflammation responses in the host (Schmaler et al., 2009; Sheldon and Heinrichs 2012). One of the cellular host responses involved is the endocytosis and degradation of erythrocytes by the macrophages in a process called erythrophagocytosis (Knutson et al., 2005). The Fe retained by the host cells is also put to good use, as Fe²⁺ enabled production of ROS is used to kill bacteria taken up by these cells (Rosen et al., 1995; Haschka, Hoffmann, and Weiss 2021). Interestingly. ROS also induce antibiotic resistance in S. aureus, indicating a negative side-effect of our innate immune response (Rowe et al., 2020).

In the context of sepsis, S. aureus is capable of lysing erythrocytes by secreting hemolytic toxins to free hemoglobin and obtain it through the Isd system (Torres et al., 2010). The S. aureus heme-oxygenases IsdG and IsdI have been shown to be important for full virulence with heme as the primary iron source (Reniere and Skaar 2008). Host heme oxygenase 1 (HO1) catalyzes the rate-limiting step in heme degradation, producing biliverdin, Fe²⁺ and CO (carbon monoxide) (Singh et al., 2018). CO can act as a messenger in various protective cascades. Links between HO1 and protective effects on S. aureus infection have been shown (MacGarvey et al., 2012; Gahlot et al., 2017).

The host restricts the availability of iron in its different forms further by producing the hemoglobin binding protein haptoglobin, the heme binding protein hemopexin, the free Fe³⁺ binding proteins transferrin and lactoferrin, and the free Fe²⁺ binding protein calprotectin (Nakashige et al., 2015; Marchetti et al., 2020). Haptoglobin binding to hemoglobin inhibits uptake of heme by the Isd system of S. aureus, although the protein still binds to IsdH (Mikkelsen, Runager, and Andersen 2020). Calprotectin (CP) was originally identified to be involved in Mn²⁺ limitation by the host, but was more recently found to bind Fe²⁺ efficiently in the presence of Ca²⁺ (Nakashige et al., 2015). It has been demonstrated that CP induces Fe starvation in S. aureus cultures (Obisesan, Zygiel, and Nolan 2021; Zygiel et al., 2021). The ability of S. aureus to efficiently incorporate heme affords protection against CP induced Fe starvation (Zygiel et al., 2021). In the preceding sections we have described the mechanisms through which S. aureus controls intracellular Fe levels in response to iron sources in the human host. However, these control systems may be insufficient when Fe availability is strongly reduced. Pathogenic bacteria such as S. aureus have evolved several strategies to tackle the metal restrictions imposed through nutritional immunity: 1) hijacking host proteins, 2) replacing metal cofactors, and 3) replacing functions by non-metal dependent enzymes. We will discuss examples of each strategy below.

S. aureus cannot use hemopexin as a heme source using the Isd system. However, it has been reported to take up iron from host transferrin using a transferrin receptor. The nature of the transferrin receptor of S. aureus is convoluted in literature. This cell-wall associated protein was first identified as a functional glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Modun, Morrissey, and Williams 2000). However, this was shown to be incorrect, and the protein was identified as staphylococcal transferrin-binding protein StbA (Taylor and Heinrichs 2002). Later, it was shown that StbA is the same protein as IsdA, part of the Isd system for heme uptake described above (Clarke, Wiltshire, and Foster 2004; Maresso and Schneewind 2006). The ongoing tug of war for Fe between host transferrin and bacterial transferrin receptors has caused rapid evolutionary development of the involved proteins (Barber and Elde 2014). The presence of a transferrin receptor indicates that S. aureus can take up Fe from transferrin, at least to some extent, although the major Fe source from the host is heme.

A well-established example of replacing metal cofactors by S. aureus are the Mn-dependent superoxide dismutases (SODs). Neutrophils, and other host immune cells, can induce oxidative bursts as a defensive strategy against S. aureus (Rigby and DeLeo 2012). This process generates high levels of damaging ROS, including superoxide (Forrester et al., 2018; Jakubczyk et al., 2020). The expression of SODs is one way in which S. aureus can combat ROS. S. aureus has two superoxide dismutases, SodA and SodM. SodA can incorporate only Mn as metal cofactor while SodM can use either Fe or Mn depending on the conditions (Garcia et al., 2017). As part of the host immune response, the neutrophil protein CP sequesters trace metals, including Mn. The action of CP disturbs the correct metalation of SodA (Kehl-Fie et al., 2013). However, S. aureus encodes an additional Mn-dependent SOD, SodM, which can substitute its metal cofactor for Fe under Mn-limiting conditions (Garcia et al., 2017; Treffon et al., 2020). In this way, S. aureus can retain sufficient SOD activity despite CP activity and maintain virulence. Small non-coding regulatory RNA molecule RsaC (co-transcribed with Mn transporter MntABC) represses the translation of the SodA coding mRNA under Mn limiting conditions. So, if there is a shortage of Mn (Mn uptake by MntABC needed), the Mn SOD is suppressed in favor of the Fe containing SOD. A clinically relevant example of a highly oxidative stressful environment with strong CP presence are the airways of cystic fibrosis patients, where S. aureus can cause persistent infections for years. Investigation of gene expression demonstrated significantly elevated SodM expression levels in clinical isolates compared to laboratory strains (Treffon et al., 2020).

An alternative strategy of S. aureus to respond to nutritional immunity is to use a protein variant that lacks a metal cofactor altogether. The consumption of glucose through glycolysis is a process fundamental to many life forms. For bacteria, some of the enzymes involved are Mn-dependent. Yet even under Mn-limited conditions, S. aureus was shown to prefer glucose as main carbon source despite its burden on cellular Mn demand (Radin et al., 2019b). It has recently been demonstrated that S. aureus can express a Mn-independent variant of phosphoglycerate mutase to maintain glucose consumption under Mn-stress (Radin et al., 2019a). The discovery of metal-independent variants is not unique to S. aureus and may indicate a broader pattern among bacteria.