Nutritional Immunity: S100 Proteins at the Host-Pathogen Interface*

The S100 family of EF-hand calcium (Ca2+)-binding proteins is essential for a wide range of cellular functions. During infection, certain S100 proteins act as damage-associated molecular patterns (DAMPs) and interact with pattern recognition receptors to modulate inflammatory responses. In addition, these inflammatory S100 proteins have potent antimicrobial properties and are essential components of the immune response to invading pathogens. In this review, we focus on S100 proteins that exhibit antimicrobial properties through the process of metal limitation, termed nutritional immunity, and discuss several recent advances in our understanding of S100 protein-mediated metal sequestration at the site of infection.


Structure and Metal Binding
The basic unit of EF-hand proteins is a helix-Ca 2ϩ binding loop-helix motif; these motifs are typically packed in pairs to form a stable globular four-helix bundle domain (8). Each S100 protein contains a distinctive S100-specific N-terminal EFhand motif and a C-terminal canonical EF-hand motif (Fig. 1). The fundamental structural unit of S100 proteins is a highly integrated antiparallel dimer (7); all S100 proteins form this structure as homodimers, and some will also heterodimerize. S100A8 and S100A9 are unique among all members of the family because they preferentially form a heterodimer (18), which is termed calprotectin based on its role in innate immunity. S100 proteins are also known to form higher order oligomers, usually mediated by high levels of Ca 2ϩ or Zn 2ϩ .
Like other EF proteins that function in signaling, the binding of Ca 2ϩ causes a conformational change, in this case within each S100 protein subunit (Fig. 1A) (8,9,19). Ca 2ϩ plays an important role in the functional duality of calprotectin. Inside cells, where the basal level of Ca 2ϩ is in the nanomolar range, calprotectin can serve as a sensor of Ca 2ϩ signals, which are associated with ϳ100-fold increase in Ca 2ϩ concentration into the micromolar range. This in turn results in the binding of ions, conformational change, and interaction with intracellular target proteins. Ca 2ϩ is also known to stimulate formation of higher order oligomers of S100 proteins, including S100A8/ S100A9 tetramers that have been suggested to play a role in some of calprotectin's activities (15, 20 -22). In the extracellular milieu, S100 proteins do not function as Ca 2ϩ sensors because Ca 2ϩ concentration is in the mM range and the proteins are perpetually Ca 2ϩ -bound. Secretion of S100 proteins therefore causes a change in the functional properties. Thus, it has been proposed that Ca 2ϩ may act as the molecular switch between the intracellular and extracellular functions of calprotectin (21). Because extracellular calcium is constant, the molecular switch can also be viewed as the secretion of the protein.
Regardless, extracellular function of calprotectin is governed by being Ca 2ϩ -bound. Importantly, Ca 2ϩ has been shown to stimulate the binding of transition metals in calprotectin, which as discussed below is essential for its role in host defense against pathogens (2). S100 proteins are distinguished from other EF-hand proteins by the presence of two transition metal binding sites at the dimer interface. In calprotectin, the first transition metal binding site (Site I) is capable of binding both zinc (Zn 2ϩ ) and manganese (Mn 2ϩ ) with high affinity (K d (Zn 2ϩ ) ϳ10 Ϫ9 M, K d (Mn 2ϩ ) ϳ10 Ϫ7 -10 Ϫ8 M), whereas the second binding site (Site II) binds only Zn 2ϩ with high affinity (K d (Zn 2ϩ ) ϳ10 Ϫ9 M) (12, 20 -22). An x-ray crystal structure, as well as spectroscopic and mutagenesis studies, revealed that the Zn 2ϩ binding in Site I involves His-17 and His-27 from S100A8 and His-91 and His-95 from S100A9, whereas chelation of Mn 2ϩ involves the same four residues along with two additional histidine residues from the C-terminal tail of the S100A9 subunit (Fig. 1A), which enable binding in the requisite octahedral geometry (12,21,22). Remarkably, a His 6 Mn 2ϩ binding site is unique to calprotectin and is not seen in any other Mn 2ϩ -binding protein. Site II chelates Zn 2ϩ with His-83 and His-87 from S100A8 and His-20 and Asp-30 from S100A9 ( Fig. 1A) (12,20,21), but lacks appropriately positioned additional ligands to enable high affinity binding of Mn 2ϩ .
In contrast to calprotectin, S100A7 and S100A12 function as homodimers (6,(23)(24)(25). S100A7 binds two Zn 2ϩ ions at symmetrically disposed sites across the dimer interface using residues His-86 and His-90 from one subunit and residues His-17 and Asp-24 from the other (Fig. 1B) (23). Binding of Zn 2ϩ is believed to stabilize the dimer and potentially mediate S100A7 function during infection (23). S100A12 homodimerization leads to the formation of two symmetrically disposed transition metal binding sites capable of binding Zn 2ϩ and Cu 2ϩ (24). The ions are ligated by His-15 and Asp-25 from one subunit and His-85 and His-89 from the other subunit of the ( Fig. 1C) (24). Interestingly, metal binding at these sites leads to substantial changes in the functional properties of S100A12. At the biochemical level, Zn 2ϩ binding stimulates a 1500-fold increase in Ca 2ϩ binding affinity (24). Furthermore, Zn 2ϩ and Cu 2ϩ binding promote formation of an S100A12 tetramer (24). As discussed below, this tetrameric form of S100A12 likely mediates important inflammatory functions during infection.

Expression and Regulation of S100 Proteins
To maximize the protective function of antimicrobial S100 proteins while simultaneously maintaining immune system homeostasis, expression of S100 proteins is tightly regulated (1,4). Calprotectin and S100A12 are primarily expressed in cells of myeloid origin, such as neutrophils, monocytes, and early macrophages (5,26,27). In neutrophils, calprotectin accounts for over 40% of the cytoplasmic fraction, highlighting its importance in the neutrophilic immune response (28,29). Expression of calprotectin and S100A12 can be induced in keratinocytes, endothelial cells, and epithelial cells during inflammation (27, 30 -34). Furthermore, various in vitro studies have shown that induction of macrophages with pro-inflammatory cytokines can lead to the expression and release of calprotectin and S100A12 (35)(36)(37). Secretion of S100 proteins including calprotectin and S100A12 is facilitated by: (a) active release through intact microtubule networks in a Golgi-independent pathway (38); (b) release during the formation of neutrophil extracellular traps (39); or (c) release through passive release during cell necrosis (40). At some sites of infection and inflammation, calprotectin concentrations exceed 1 mg/ml, suggesting massive expression and/or mobilization of this protein during infection (41). S100A7 is constitutively expressed in skin at relatively high levels, and expression is amplified in keratinocytes upon induction by pro-inflammatory cytokines IL-17 and IL-22 and bacterial products, such as flagellin (42). The differential expression profiles of S100 proteins allow for an immediate antimicrobial response upon infection in certain tissues, while limiting potentially detrimental inflammatory responses associated with each of these proteins. The capacity for S100 proteins to mediate inflammation and the potential link to chronic inflammatory disease will be discussed below.

Inflammatory Response and Regulation
In the extracellular matrix, S100 proteins can act as potent modulators of inflammation. Once released by cells, these proteins are classified as DAMPs because of their important role in regulating inflammatory responses (10). Extracellular S100 proteins can exhibit chemokine-and cytokine-like activity, initiate pro-and anti-inflammatory responses, and interact with pattern recognition receptors. Growing evidence suggests that S100-mediated inflammation is driven by endogenous interaction with pattern recognition receptors including RAGE and Toll-like receptors (10,11). It has been demonstrated that calprotectin is an endogenous agonist of TLR4 (43). Binding to TLR4 and several other components of the lipopolysaccharide complex initiates a signaling cascade that promotes inflammation, autoimmunity, and tumor development in an NF-B-dependent manner (43)(44)(45). Apart from TLR4, evidence also suggests that calprotectin, S100A7, and S100A12 each independently interact with RAGE. S100 protein activation of RAGE drives an NF-B-mediated pro-inflammatory response and recruitment of neutrophils, monocytes, and macrophages (11, 46 -48). Additionally, calprotectin has been suggested to activate pro-inflammatory cytokine production in monocytes and macrophages through NF-B and p38 MAPK pathways (47, FIGURE 1. Structures of S100 proteins involved in nutritional immunity. A, structure of Zn 2ϩ -bound S100A7 homodimer (23). B, structure of Cu 2ϩbound S100A12 homodimer (55). C, structure of S100A8/S100A9 calprotectin heterodimer bound to Zn 2ϩ in Site II and Mn 2ϩ in Site I (12,75). 49). During infection, calprotectin also acts as a chemotactic for neutrophils and can promote neutrophil adhesion at the site of infection in a RAGE-independent manner (50). Beyond their role in the pro-inflammatory response during infection, these three S100 proteins may also have important anti-inflammatory functions. For example, calprotectin appears to have the ability to scavenge reactive oxygen species (ROS) (51,52). It has been proposed that scavenging enables calprotectin to minimize collateral damage associated with neutrophil ROS (27). Furthermore, calprotectin inhibits growth or promotes apoptotic and autophagy-like death in several cell types including macrophages, lymphocytes, endothelial cells, and tumor cells (53,54). Taken together, these data suggest that the impact of the three S100 proteins on the immune response is complex and likely dependent on a combination of factors including the local concentration of metal ions, the distribution of immune cells, and the site of infection.
An important concept associated with S100 proteins is the regulatory role that metal binding plays in modulating structure and functions of these proteins. As noted above, Zn 2ϩ and Ca 2ϩ stimulate oligomerization of S100A12. Interestingly, oligomerization is believed to be required for RAGE binding and subsequent inflammation (55). Evidence also suggests that the RAGE-dependent chemo-attractant activities of S100A7 may be dependent on Zn 2ϩ binding (56). Similarly, Ca 2ϩ binding stimulates transition metal binding, which is essential for extracellular functions. Binding of Zn 2ϩ has also been suggested to mediate calprotectin's apoptosis-inducing activity (57). The ability to control function via concentration or localization of transition metal ions may allow for the immense adaptability and diversity of activities seen in many S100 proteins. Table 1 lists the specific association of S100A7, S100A8/ S100A9, and S100A12 with inflammatory diseases (Table 1), which is not surprising given their expression patterns and immunomodulatory effects during inflammation. Calprotectin and S100A12 are present at high concentrations in inflamed tissues that harbor neutrophil and monocyte infiltrate. Similarly, S100A7 is present at high abundance in inflamed skin. It has been suggested that release of these S100 proteins during inflammation is at least partially associated with Ca 2ϩ influx upon activation of monocytes (35). Calprotectin and S100A12 have been associated with various inflammatory diseases such as rheumatoid arthritis, psoriasis, and inflammatory bowel disease (10,32,58). Furthermore, these proteins are believed to play an important role in several cancers (59 -61). Similarly, TABLE 1 S100 proteins in nutrional immunity S100A7 is linked to several inflammatory skin diseases, including psoriasis and atopic dermatitis (42,62).

S100 Proteins Associated with Inflammatory Diseases
The mechanism by which the three antimicrobial S100 proteins mediate autoimmune, inflammatory, and pro-cancer activities is directly associated with their role as potent immune modulatory DAMPs. Through the activation of pro-inflammatory signaling cascades at the sites of disease, these proteins contribute to a positive feedback loop that produces overly active inflammatory responses and pro-tumor microenvironments. Due to their association with inflammation, these proteins have also been exploited as noninvasive biomarkers for several disorders, including inflammatory bowel disease, rheumatoid arthritis, and colon cancer (63)(64)(65). In addition, they represent logical targets for therapeutics that aim to minimize aberrant inflammation associated with disease.
Hypercalprotectinemia is an extremely rare inflammatory disorder that is associated with extraordinarily high levels of calprotectin. The few patients who have been described with this disorder have abnormally high levels of Zn 2ϩ in their tissues and develop aberrant systemic inflammatory responses (66 -69). Chronic inflammation leads to symptoms such as dermal ulcers, folliculitis, and anemia. To this point, the mechanisms that drive this rare disease are largely unknown; however, it is postulated that calprotectin catabolism may be defective in these individuals (66). Furthermore, it is possible that calprotectin release into the extracellular matrix is dysregulated during hypercalprotectinemia, leading to increased release of calprotectin and a subsequent hyperactive systemic inflammatory response.

Nutritional Immunity
During infection, invading bacterial pathogens require access to essential transition metals to colonize the host and cause disease (70). To combat this, the host exploits the pathogen's need for nutrient metals by producing factors that limit metal availability and starve pathogens in a process termed nutritional immunity. The S100 protein calprotectin is the best studied of these nutritional immunity factors. Calprotectin has broad-spectrum antimicrobial activity based on its ability to sequester Zn 2ϩ and Mn 2ϩ at the site of infection. Several recent studies have shown that it inhibits growth of numerous important human pathogens in vitro including Staphylococcus aureus, Acinetobacter baumannii, Candida albicans, and Helicobacter pylori (13,14,16,39). In each case, calprotectin-mediated inhibition is reversed by ablating calprotectin metal binding activity through mutagenesis or by complementing with excess Zn 2ϩ or Mn 2ϩ . Utilizing S100A9 Ϫ/Ϫ mice, which are calprotectin-deficient, several studies have shown that calprotectin is important for protecting against bacterial infection in murine models of A. baumannii pneumonia, H. pylori gastric infection, S. aureus systemic infection, and C. albicans subcutaneous and pulmonary infections (12-14, 16, 39). These in vivo studies demonstrate the importance of calprotectin-mediated nutritional immunity during infection.
There have been few studies to establish the microbial processes that are impacted by calprotectin metal limitation. Interestingly, calprotectin-mediated sequestration of Mn 2ϩ at the site of infection increases S. aureus susceptibility to neutrophil killing by superoxide through the inhibition of the staphylococcal Mn 2ϩ -dependent superoxide dismutase defense system. This is one example of how calprotectin-induced metal restriction impacts bacterial metabolism (16). These findings suggest that at the host-pathogen interface, calprotectin functions by outcompeting or stripping metals from pathogen metalloproteins, rendering them inactive and weakening their defenses to the host immune response. In some cases, pathogens have developed the ability to compete with calprotectin-mediated metal starvation in the gut. Salmonella serovar Typhimurium expresses a high affinity Zn 2ϩ transporter (ZnuABC) to acquire Zn 2ϩ during infection (17). This strategy confers resistance to calprotectin-mediated metal chelation and allows S. Typhimurium to thrive under conditions of inflammation in the gut and outcompete the resident commensal bacteria (17).
As noted above, the binding of Ca 2ϩ has an effect on calprotectin's affinity for Zn 2ϩ and Mn 2ϩ , and therefore its function (21,71). Inside cells, the concentrations of free Zn 2ϩ and Mn 2ϩ are held so low that even during Ca 2ϩ signaling events, it is highly unlikely that calprotectin or other S100 proteins are important for the homeostasis of these important transition metals. In contrast, the higher levels of Ca 2ϩ in the extracellular space maximize the affinity for Zn 2ϩ and Mn 2ϩ . This ensures activation of the antibacterial activity that allows for calprotectin to outcompete high affinity bacterial metal transporters for Zn 2ϩ and Mn 2ϩ at the site of infection (16,21).
In addition to calprotectin, accumulating evidence supports a potential role for S100A7 in limiting bacterial infection through metal limitation (42,72), although there are differences in the effect on different organisms. At low doses, purified S100A7 from keratinocytes has antimicrobial activity against Escherichia coli, but at higher concentrations, S100A7 exhibits killing activity against Pseudomonas aeruginosa and S. aureus (42). The antimicrobial activity of S100A7 is dependent on the limitation of Zn 2ϩ , as complementation with excess Zn 2ϩ ablates bactericidal activity (42). Additionally, an alternative mechanism for antimicrobial activity has been proposed, by which S100A7 directly adheres to and reduces survival of E. coli and potentially other pathogens found on the epidermis (73). Through cross-linking to bacterial components, S100A7 may reduce survival and act as a physical barrier during infection. Further studies are needed to explore this phenomenon.
Human S100A12 possesses antimicrobial activity against several parasites (5,72). However, it has been difficult to study in the context of infection because it is not present in mice. The mechanism of antimicrobial activity is unclear, but it is tempting to speculate that S100A12-mediated metal limitation plays a key role during infection. This hypothesis is supported by the antimicrobial properties observed with calcitermin, a protein homologous to the C terminus of S100A12. Calcitermin exhibits Zn 2ϩ -dependent antimicrobial activity against P. aeruginosa, C. albicans, and E. coli (74). It has also been suggested that Cu 2ϩ bound to S100A12 actively produces superoxide, which could potentially be antiparasitic (72). Given the significant gaps in our knowledge regarding the role of S100 proteins in host defense against infection, it is clear that further research is needed to characterize whether S100A12 contributes to nutritional immunity. Furthermore, additional studies are required to define the mechanisms by which metal limitation impacts invading organisms. We believe that developing a greater understanding of the influence of metal limitation on an array of organisms will lay the groundwork for development of novel therapeutics to target critical pathogens.

Concluding Remarks and Future Challenges
Evidence strongly supports an important role for S100A7, calprotectin, and S100A12 as antimicrobial proteins that protect against infection (Fig. 2). However, these S100 proteins can also have a negative impact on the host by amplifying aberrant pro-inflammatory responses and potentiating disease (Fig. 2). It is clear that we are just beginning to understand the importance of S100 proteins and the roles they play in both inflammation and nutritional immunity. A detailed analysis of the mechanisms that regulate S100 protein function during infection and inflammation will improve our understanding of how immune homeostasis is maintained during health and disrupted during S100 protein-associated disease.
Critical insights have been obtained on the impact of S100 protein-mediated metal limitation at sites of infection. However, the impact of metal loading on S100 protein immunoregulatory effects at the site of infection remains largely unstudied. With the development of techniques such as CRISPR-CAS9 genome editing, we anticipate an accelerating pace of discovery. Interrogation of the metal binding sites of S100 proteins and performing in vivo characterization at the site of infection will provide critical new insights. It is anticipated that obtaining better understanding of these complex processes will allow for the development of therapeutics directed to S100 proteins and their targets during aberrant inflammatory disease states, while maintaining their essential role in nutritional immunity to invading pathogens.
Consideration must be given to the specific environment within the host where S100 proteins are interacting with the both the immune system and the pathogen. During infection, different tissues have dramatically altered metal concentrations, cell types, and stresses. S100 proteins may be involved in controlling transition metal distribution and homeostasis in infected tissue. Understanding how S100 proteins differentially function at diverse sites of infection and during infections by different organisms is essential for elucidating their role in nutritional immunity and inflammation. Key questions still remain on the pathogen side of the host-pathogen interface. Little is known about the targets of antimicrobial metal limitation during infection. Elucidation of these pathogen factors could lead to the development of novel drug targets that focus on the fundamental nutrient requirements of pathogens. Continued work studying S100 protein biology in these contexts is expected to lead to significant advances in our understanding of infection, autoimmunity, and cancer.