The Periodic Table’s Impact on Bioinorganic Chemistry and Biology’s Selective Use of Metal Ions

Abstract

Despite the availability of a vast variety of metal ions in the periodic table, biology uses only a selective few metal ions. Most of the redox active metals used belong to the first row of transition metals in the periodic table and include Fe, Co, Ni, Mn and Cu. On the other hand, Ca, Zn and Mg are the most commonly used redox inactive metals in biology. In this chapter, we discuss the periodic table’s impact on bio-inorganic chemistry, by exploring reasons behind this selective choice of metals biology. A special focus is placed on the chemical and functional reasons why one metal ion is preferred over another one. We discuss the implications of metal choice in various biological processes including catalysis, electron transfer, redox sensing and signaling. We find that bioavailability of metal ions along with their redox potentials, coordination flexibility, valency and ligand affinity determine the specificity of metals for biological processes. Understanding the implications underlying the selective choice of metals of the periodic table in these biological processes can help design more efficient catalysts, more precise biosensors and more effective drugs.

1. Introduction

Metalloproteins are at the heart of numerous biological processes ranging from photosynthesis and respiration to natural product biosynthesis [1, 2]. For example, a manganese-calcium cluster is involved in the oxidation of water during photosynthesis [3] while an iron-copper center reduces oxygen to water during respiration [4]. Vertebrates utilize heme iron in myoglobin and hemoglobin for oxygen transport in their bodies [5], whereas some invertebrates like mollusks and arthropods utilize copper in hemocyanin for oxygen transport [6]. This choice of copper over iron has resulted in profound physiological differences between the two phyla. Mollusks, for example, have colorless blood due to the presence of Cu^I and on exposure to oxygen, Cu^I oxidizes to Cu^II explaining why mollusks bleed blue. The bioavailability of these metal ions (Fig. 1) have often been given as the reason for these choices in proteins and organisms [7].

Figure 1.

Periodic table of elements based on their abundance in the earth’s crust. The abundance values have been obtained from reference [11].

Iron is the fourth most abundant metal in the earth’s crust [8, 9] and is widely used in various redox processes in terrestrial animals e.g. oxygen transport (hemoglobin), electron transfer (cytochromes) and oxygen reduction during respiration (heme-copper oxidase). The level of iron in some surface sea waters, however, can be up to ten-fold lower than copper [10]. This higher bioavailability of copper (v/s iron) in aquatic life may explain the selective use of copper for oxygen transport in marine mollusks and arthropods. Other highly bioavailable redox active metal ions include titanium, vanadium, chromium, manganese, cobalt, nickel and manganese. While cobalt, nickel, vanadium and manganese often feature as essential cofactors in various biochemical redox processes, biological use of titanium and chromium remains unknown. Factors other than bioavailability, therefore, play a critical role in determining biology’s selective use of metal ions. Our increased biochemical and mechanistic knowledge of enzymatic processes has begun to unravel the reason behind the choice of metal ions (Table 1). In this chapter, we will provide an overview of the roles of common physiologically available metal ions in some important biological processes with special focus on the chemical and functional reason why one metal was preferred over another, and how replacing one metal ion with another would have an impact on the functional properties. Understanding the implications of selective choice of metals from the periodic table for these biological processes can help design more efficient catalysts, more precise biosensors and more effective drugs.

Table 1.

Metalloproteins involved in biological processes classified based on metal types. X, XH and XH₂ are substrates. The table has been adapted from reference [23].

	Heme Fe	Nonheme Fe	Cu	Other metals
Electron transfer	Cytochromes	Iron-Sulfur cluster	Cupredoxin
Monooxygenase X+O2+2e−+2H+ → XO + H2O	P450 monoxygenase Secondary amine monoxygenase NO synthase	Soluble MMO Pterin-dependent hydroxylase (Phe, Tyr and Trp)	Particulate MMO Tyrosinase Polysaccharide monoxygenase Dopamine β-hydroxylase Phenylalanine hydroxylase Peptidylglycine α-amidating monoxygenase
Dioxygenase X+O2→XO + H2O	Indoleamine 2,3-dioxygenase Tryptophan 2,3-dioxygenase Prostaglandin H synthetase	Lipoxygenase Catechol dioxygenase α-ketoglutarate dependent dioxygenase Arene dioxygenase TET dioxygenase	Quercetinase
Oxidases O2 + 4e− + 4H+ → 2H2O	Heme-copper oxidase (Cu)	Ribonucleotide reductase (Mn) Clavaminate synthase Isopenicillin N synthase	Laccase Ascorbate oxidase Phenoxazinone synthase Galactose oxidase Amine oxidase
Peroxidase and catalases H2O2 + XH2 → X + 2H2O 2H2O2 → O2 + 2H2O	Catalase Horseradish peroxidase Cytochrome c peroxidase Chloroperoxidase Lignin peroxidase Prostglandin H synthetase			Mn catalase V bromoperoxidase Mn peroxidase Co peroxidase
Superoxide dismutase 2O2− + 2H+ → O2 + H2O2		FeSOD	Cu-Zn SOD	Mn SOD
Hydrogenase H2 → 2H+ + 2e−		[Fe-Fe] hydrogenase [Fe] hydrogenase		[Ni-Fe] hydrogenase
Nitrogenase N2 + 6e− + 6H+ → 3NH3				[FeMoC] cluster [FeVC] cluster
Oxygen production 2H2O → O2 + 4e− + 4H+				[Mn4Ca] cluster in photosystem II
NO reduction 2NO + 2e− + 2H+ → N2O + H2O	NO reductases (nonheme Fe)	Flavodiiron NO reductase
Sulfite reductases SO32− + 4e− + 6H+ → H2S + 2H2O	SiR (iron-sulfur) SiR (copper)
Oxygen transport	Hemoglobin Myoglobin	Hemerythrin Myohemerythrin	Hemocyanin
Redox Sensing	H-NOX	HIF-1α hydroxylase		Calmodulin (Ca)

2. Use of different redox-active metal ions in oxidoreductive biocatalysis

Enzymes are amazing catalysts – they are fast, highly specific, green and can exhibit thousands of turnovers. For some of the challenging reactions, including oxidoreductive biocatalysis, nature recruits a variety of metal cofactors. Table 1 provides an overview of some of these well-characterized enzymes classified on the basis of reaction types, substrates and metal cofactors used. An obvious inference from the table is that Fe is utilized for almost all kinds of catalytic reactions ranging from C-H bond functionalization to O-O, H-H and N-N bond cleavages, with a few exceptions, such as O-O bond formation during water oxidation, in which Mn is used in preference. The reason for this exclusive choice of Mn may emanate from low oxidizing potential of this metal ion. It is easier to generate highly reactive Mn^v under physiological conditions compared with Fe^v, for instance. On the other hand, nature uses different metal ions for the same reaction type. For instance, superoxide dismutases from different life forms use Fe, Mn or Cu. Similarly, while sulfate reducing bacteria utilize a Ni-Fe cofactor for hydrogen oxidation, methanogenic archaea utilize only Fe cofactor for the same reaction [12]. Therefore, why are different metals used for the same reaction? Is it just bioavailability or can a more mechanistic, chemical and functional explanation can be provided? We discuss below a few case studies where biology’s selective use of metal ions for catalysis and preference of one metal over another from periodic table has been unravelled.

2.1. Copper in oxygen reduction v/s iron in nitric oxide reduction:

Heme-copper oxidases (HCOs) perform the four electron reduction of oxygen to water under aerobic conditions [13]. Nitric oxide reductases (NORs), on the other hand, perform two electron reduction of nitric oxide to nitrous oxide during anaerobic denitrification [14]. The similarity in their catalytic active sites (Fig. 2a–b), reaction mechanisms and the fact that the two enzymes exhibit significant common sequences and structural homology suggested that the enzymes probably evolved from the same ancestral protein [4]. Then, why was copper chosen for oxygen reduction and nonheme iron for nitric oxide reduction? This question eluded scientists for several decades and answering it was not straightforward as the two enzymes assembled and folded in the presence of their respective metal cofactors. Because these nonheme metal ions are tightly bound in the proteins, removing the nonheme ion from a well-folded, membrane-bound NOR and replace it with the other metal ion was very difficult and resulted in non-active enzymes. To understand the structural and functional reason for preference of Cu over Fe in oxygen reduction, we have designed and used myoglobin-based protein models of HCOs[15–20] that had a heme cofactor and an empty nonheme metal site wherein different nonheme metals such as Fe^II, Cu^I, Mn^II, Co^II and Zn^II could be incorporated [21, 22]. Oxygen reduction assays with different metal incorporated HCO models showed that while the Zn^II containing HCO model performed incomplete reduction of oxygen producing almost 50% redox active species, addition of redox active metals like Cu^II, Co^II, Mn^II or Fe^II reduced 90–98% of oxygen to water. These results suggested that the redox active nature of nonheme metal was crucial for complete four electron oxygen reduction in HCOs. In subsequent electrochemical, kinetics and vibrational spectroscopy-based studies, we explained the reason behind the specific choice of copper over iron in HCOs. The Cu was superior to Fe for two reasons. First, Cu exhibited ~ 100 mV higher reduction potential (E°′) value than Fe promoting faster electron transfer to the active site. Second, Cu^II had a d⁹ electron configuration as compared to d⁶ electrons of Fe^III, enabling a stronger dπ-pπ back bonding for Cu and weakening the O-O bond for its ready cleavage. These experiments showed how small molecule/protein-based enzyme models can help answer mechanistic questions relating to the selective use of metals for a particular reaction in biology.

Figure 2.

Varied metal cofactors used in biological oxidoreductive catalysis. Structures of the catalytic domains of heme-copper oxidase [PDB: 3MK7] (a) and nitric oxide reductase [PDB: 3AYF] (b). Fe in orange and copper is in pink. Molecular structures of Fe-Fe hydrogenase (c) and Ni-Fe hydrogenase (d). At site marked X is where molecular hydrogen presumably binds [32]. Structures of the catalytic M clusters for Mo- (e) and V- (f) nitrogenases [33]. (g) Fe in Fe-SOD (in orange) and Mn in Mn-SOD (in purple) exhibit E⁰ʹ values of 0.2 – 0.3 V. Metal switching makes Fe and Mn-SODs inactive due to E⁰ʹ value modulation [34].

2.2. Nickel v/s iron in hydrogenases:

The [Fe-Fe] and [Ni-Fe] hydrogenases catalyze the formal interconversion between hydrogen and protons/electrons, possess characteristic non-protein ligands at their catalytic sites and thus share common mechanistic features (Fig. 2 c–d) [24]. Despite the similarities between the two hydrogenase systems, they clearly have distinct evolutionary origins and probably emerged from different selective pressures [12]. [Fe-Fe] hydrogenases are widely distributed in fermentative anaerobic microorganisms and probably evolved under selective pressure to couple hydrogen production to the recycling of electron carriers that accumulate during anaerobic metabolism. In contrast, many [Ni-Fe] hydrogenases catalyze hydrogen oxidation as part of energy metabolism and were likely key enzymes in early life [12]. Although the reversible combination of protons and electrons to generate hydrogen gas is the simplest of chemical reactions, the [Fe-Fe] and [Ni-Fe] hydrogenases have distinct mechanisms and differ in the fundamental chemistry associated with proton transfer and control of electron flow [25]. These differences are apparent at various steps in the reaction. To begin with, H₂ binds to both Ni and Fe as a bridging ligand in [Ni-Fe], while for [Fe-Fe] H₂ binding and activation takes place on a mononuclear Fe site (Fig. 2c–d). The Ni atom undergoes redox cycle in +1/+2/+3 oxidation states as H₂ is oxidized in [Ni-Fe] while Fe stays put in its +2 oxidation state. For [Fe-Fe], both Fe atoms transition between their +1/+2 oxidation states. Among all hydrogenases, [Fe-Fe] show an unsurpassed H₂ release activity but are intolerant to oxygen in the environment. [Ni-Fe], though, not as catalytically fast as [Fe-Fe], is quite tolerant to oxygen [12, 26]. A unifying feature of both [Ni-Fe] and [Fe-Fe] enzymes is their use of strong field, organometallic CO and CN ligands that maintain the metals in low-spin, low-valent states for efficient hydrogen activation. The two enzymes also use a conduit of [Fe-S] clusters along with conserved Arg, His and Glu residues in their secondary coordination sphere for efficient electron and proton transfer during the hydrogenase reaction.

2.3. Molybdenum v/s vanadium in nitrogenases:

Nitrogenase catalyzes the biological conversion of atmospheric dinitrogen to bioavailable ammonia [27]. The Mo and V dependent nitrogenases are two homologous members of this metalloenzyme family sharing a good degree of similarity in their primary sequences and cluster compositions (Fig. 2e–f) [28]. Electrons are transferred sequentially from ATP-dependent Fe-S cluster (Fe₄-S₄) to P-cluster (Fe₈S₇) to Fe-Mo or Fe-V containing M-clusters where N₂ reduction takes place [29]. Recent structural, spectroscopic and structural studies have established the structure of M-clusters as (Fe₇MS₉C), where M is Mo or V. Both Mo and V nitrogenases are capable of reducing the physiological substrate (i.e., N₂) and alternative substrates [e.g., proton (H⁺), acetylene (C₂H₂) and cyanide (CN⁻)]. The V nitrogenase, however, is a less efficient N₂ reduction catalyst at ambient temperatures, requiring more ATP and additional reducing equivalents. However, its activity is relatively unaffected by decreased temperatures. Furthermore, it has been shown that the vanadium nitrogenase exhibits a unique ability to promote the reduction of CO to short-chain hydrocarbons [28]. This reductive C–C bond coupling is akin to Fischer–Tropsch chemistry, but utilizes protons and electrons in place of H₂. The relative inability of the molybdenum nitrogenase to perform the same chemistry (CO is a reversible inhibitor of molybdenum nitrogenase) poses significant questions regarding the structural and/or electronic differences underlying the disparate reactivity of these two isozymes. X-ray spectroscopic studies on the two enzymes and their structural models [30, 31] have shown that the iron-heterometal bonds are weaker in V than Mo resulting in a more reduced iron, which may explain the differential reaction profiles of the two enzyme systems.

2.4. Iron v/s manganese in superoxide dismutases:

Superoxide dismutases (SODs) catalyze the disproportionation of superoxide to molecular oxygen and hydrogen peroxide, as part of the cells’ defense against oxidative stress and are found in all aerobic organisms [34]. Depending on the organism type, SODs have evolved to use different metals. The Cu- and Zn-containing SODs are generally eukaryotic, Ni-containing SODs are found in certain fungi and bacteria, and the Fe- or Mn-dependent SODs (Fe/MnSODs) are found predominantly in bacteria, some plant chloroplasts (FeSOD), and mitochondria (MnSOD) [35]. Specifically, Fe- and MnSODs share structural and amino acid sequence homology and are considered to constitute a single class of SODs. Some members of this class display significant activity with either Mn or Fe (cambialistic Fe/MnSODs), whereas others are active only with Fe (FeSODs) or Mn (MnSODs). Miller and coworkers have done considerable work in understanding the molecular basis of this metal choice [34]. They swapped the metal ions – e.g. incorporated Mn^II in FeSOD and Fe^II in MnSOD and found that such a switch of metal ions rendered the enzymes inactive. The relative inactivity was proposed to be due to differences in redox tuning applied by Fe-specific vs Mn-specific proteins, which would be appropriate for the native metal ion but not the other. The hexa-aquo ions exhibit 3+/2+ E°′ as 1.5 and 0.77 V for Mn and Fe, respectively making it easier to reduce Mn³⁺ over Fe³⁺ in water [36]. However, the two metals in the enzyme scaffold exhibit similar E°′ values of +220 mV for FeSOD and +290 mV for MnSOD, demonstrating that an E°′ value of 200–300 mV is optimal for SOD activity (Fig. 2g). Swapping the two metals resulted in significant changes in E°′ values: Fe in MnSOD exhibited an E°′ of −240 mV and Mn in FeSOD exhibited an E°′ value > +900 mV. The secondary coordination sphere and hydrogen bonding environment of the enzymes played important roles in maintaining E°′ values of the enzymes. Through hydrogen bond modulation of the active site and using Mn^II or Fe^II metal ions, the E°′ of FeSOD could differ by over 900 mV while retaining the overall structure. These studies revealed that the structure, coordination and secondary coordination sphere of an enzyme has been fine tuned by evolution to control the redox properties of a selected metal for its reactivity.

3. Use of heme/non-heme iron and copper in biological electron transfer processes:

Efficient transfer of electrons through long distances is crucial for many biological processes from capture and storage of light energy in photosynthesis to oxidoreductive catalytic reactions in respiration. Despite the availability of numerous redox metal ions in the periodic table, biology employs a surprisingly limited number of them for the biological electron transfer (ET) processes (Fig. 3) [37, 38]. Prominent members of redox centers involved in ET processes include iron-containing iron-sulfur (Fe-S) clusters, cytochromes, and copper-containing cuperedoxins. Together these centers cover the whole range (approximately 1.5 V) of E⁰ʹ in biology [37–39]. Recent reviews [39–50] have provided extensive coverage on classification and description of these redox centers. In this chapter, we will focus on the structural features that have been employed to fine-tune the chemical properties of these metal ions to make them the ideal choices to cover the entire range of biological redox potentials.

Figure 3.

(a) Metal cofactors in Fe-S clusters (Fe in orange, S in yellow), cupredoxin (Cu in blue) and cytochromes (Fe in orange). (b) Reduction potential range of redox centers in electron transfer processes. Fe-S proteins in grey, cytochromes in orange and cupredoxins in blue[38]

3.1. Iron-sulfur clusters:

Fe-S clusters are amongst the oldest metalloproteins on earth. The early oxygen-free reducing atmosphere, under which both sulfur and ferrous iron were abundant, has enabled the spontaneous assembly of these two elements into clusters [44, 51]. While Fe is the primary choice of metal in these clusters, spontaneous replacement of one of the Fe²⁺ atoms with Co²⁺ has been suggested under excess cobalt stress in E. coli [52, 53]. Protein scaffolds hold onto the Fe-S cluster by coordinating the Fe through cysteine or histidine ligands. The Fe-S clusters exhibit significant structural diversity ranging from 1Fe in rubredoxin, 2Fe-2S, 3Fe-4S and 4Fe-4S clusters in ferredoxins where the ligands are cysteines. One or more of the cysteine ligands can be replaced with histidine, resulting in 2Fe-2S found in rieske clusters and 4Fe-4S found in high potential iron-sulfur proteins (HiPIPs). Moreover, Fe-S clusters also exhibit varied redox transitions. 4Fe-4S clusters can undergo Fe^+3/+2 and Fe^+2/+1 transitions under physiological conditions. Other structural features fine-tuning the E⁰′ and functional properties of Fe-S clusters includes geometry of the cluster (e.g., distortion from tetrahedral geometry of iron), ligand types (e.g., cysteine vs. histidine), extent of cluster burial within the protein environment (the more buried the cluster, the higher E⁰′) and secondary coordination sphere (SCS) interactions. These structural, chemical and functional modulations make it possible for the Fe-S clusters to cover a wide biological redox range (from −680 mV to +488 mV), making them excellent choice to serve as electron transfer cofactors in a variety of biological processes [38].

Rubredoxin with its relatively simple tetracysteine ligated tetrahedral iron center, low molecular mass (6 kDa) and easy metal reconstitution was one of the first proteins on which metal substitution studies were conducted [54, 55]. The Fe^II in rubredoxin was switched to its periodic table neighbors, such as Co^II, Ni^II, Cu^I and Zn^II, as well as those away from it, such as Ga^II, Cd^II and Hg^II. The Co^II-, Ni^II- and Ga^II-substituted rubredoxins exhibited metal-binding sites with geometries similar to that of the Fe^II form, as did the Cd^II and Hg^II proteins, but with a significant expansion of the metal-sulfur bond lengths [56]. Recently, catalytic hydrogen evolution activity of Ni^II-substituted rubredoxin was investigated [57, 58] and it was shown that the proton-coupled electron transfer to the Ni center was essential for catalysis. The coordinating thiolate ligands served as protonation sites with reduction occurring at the Ni center. The rate-determining step was suggested to be intramolecular proton transfer via thiol inversion to generate a Ni^III–hydride species. Replacing the native metal ion of ET proteins such as rubredoxin with other metal ions, therefore, not only provides a method to model active sites of difficult to study enzymes (like [Ni-Fe] hydrogenase) but also generates novel catalysts for bioenergetics.

3.2. Cytochromes:

Cytochromes are proteins containing heme as a cofactor whose E⁰ʹs appear within the middle range (from −485 mV to +345 mV) of biological redox scale (Fig. 2). The extended conjugation of the porphyrin macrocycle renders these proteins with an intense red color.[59–61] Cytochromes are one of the best studied protein families due to their small size, high solubility, and thermodynamic stability along with very specific heme-dependent spectroscopic signatures. As of today, more than 70,000 cytochromes have been discovered and characterized.[61] All of these cytochromes utilize heme coordinated iron as the redox center. So, how are they able to tune their E⁰ʹ by > 800 mV? Structural and redox diversity in cytochromes arise from modifications in their porphyrin groups, and variations in axial coordinating amino acids and SCS interactions around them. For instance, His/Met-coordinating cytochromes exhibit higher E⁰′ values than bis-His due to lower sigma electron donating properties of the sulfur ligand in Met group. Similarly, the presence of electron-withdrawing formyl group in heme a and disruption of (4n+2)π conjugation in heme d increases E⁰ʹ values for these porphyrins as compared to heme b and c.

Numerous studies have focused on replacing the Fe^II in the heme group of cytochromes with Mn^II, Co^II, Ni^II and Zn^II to understand the functional implications of metal-based functional property changes [62–65]. When Fe^II in cytochrome c was replaced by Mn^II, Co^II and Ni^II, significant changes in the redox properties of the protein were observed. While native cytochrome c with Fe^3+/2+ exhibited an E⁰ʹ value of +250 mV, Mn^3+/2+ and Co^3+/2+ cytochromes exhibited significantly reduced E⁰ʹ values of +60 ± 40 and −140 ± 40 mV respectively [63, 65]. Ni^II and Zn^II cytochromes, on the other hand, were redox inactive under physiological conditions [62]. The metal-substituted cytochromes also exhibited enhanced reactivities to NO and distorted geometry. Mn^II, for instance, was out of porphyrin plane and ligated to only histidine ligand, unlike native Fe^II containing cytochrome, which is coordinated to both His and Met. Notably, swapping Fe^II with Zn^II in cytochrome c rendered the protein fluorescent such that FRET-based protein folding studies were possible [64]. Overall, metal substitution studies on cytochromes showed that while the protein fold and structural aspects of this class of proteins were optimized for ET processes in a broad redox range, swapping Fe with other metals generated novel functional properties.

3.3. Cupredoxins:

Copper is the second most abundant transition metal in biological systems, next to iron. Aqueous Cu^2+/1+ transition happens at ca. 0.3 V higher potential than Fe^3+/2+ transition [36]. Copper-containing redox proteins, called cupredoxins, therefore, appear at the higher side of biological redox range (from +180 to +720 mV).[66–68] Two kinds of cupredoxins mediate ET – mononuclear type 1 (T1) and dinuclear Cu_A centers [69, 70]. The two centers share several common features - Both contain a Cu-thiolate bond, both are located in a cupredoxin fold and both are highly optimized for ET, displaying low reorganization energies and high ET rates. In fact, the T1 Cu proteins and Cu_A domains in heme–copper oxidases share the same cupredoxin fold, with three ligands of T1 Cu and four ligands of Cu_A residing in the so-called “ligand loop”. By careful design, it is possible to transplant the ligand loop of one protein into another, enabling interconversion between T1 copper and Cu_A [71, 72]. In terms of E⁰ʹ, Cu_A proteins exhibit a narrower range of E⁰ʹ of +200 mV to +310 mV than those of T1 Cu centers (from +180 to +790 mV). Therefore, how do the T1 Cu centers possessing similar coordination mode and ligand geometry modulate their E⁰ʹ values by > 600 mV? It turns out that secondary coordination sphere interactions through both hydrophobicity and hydrogen bond play a major role in redox tuning [73]. For instance, while most T1Cu centers contains a Met as an axial ligand, leucine is present at the same location in laccase and ceruloplasmin, which display the highest E⁰ʹ among T1Cu centers. In addition, replacing the Met with Leu has resulted in higher E⁰ʹ. In fact a linear relationship has been found between E⁰ʹ and hydrophobicity of the axial ligands. On the other hand, hydrophobicity alone cannot explain the whole wide range of E⁰ʹ of T1Cu centers. For example, while rusticyanin has a higher potential relative to other T1 copper proteins, it still contains Met as the axial ligand. By sequence comparison, it was established that there is a Ser in rusticyanin at the position corresponding to Asn that “zips” two ligand loops together. Asn had been proposed to decrease the E⁰ʹ by strengthening the hydrogen bonding interactions between two ligand-containing loops. Mutating Ser86 in rusticyanin to Asn established such a hydrogen bond and lowered the E⁰ʹ by 77 mV. On the other hand, changing Asn in azurin to Ser eliminated one hydrogen bond between two loops and resulted in a protein with a 131 mV higher E⁰ʹ. Unlike mutations on the copper ligands, mutations of residues in the SCS did not perturb the chemical and electronic characteristics of T1 center. Interestingly, the mutants exhibited up to 10-fold greater ET rate due to lower reorganization energy and greater flexibility of the designed copper center [74].

Cupredoxins act as “rack” and are able to ligate to and support both Cu^I and Cu^II that ordinarily prefer tetrahedral and linear geometries respectively. In fact, T1 Cu proteins like azurin are purified without a metal; the absence of a metal does not perturb its overall fold or structure. Due to this “rack” like property of azurin, transition metals like Fe^II, Co^II, Ni^II and Zn^II have been incorporated into this protein. The incorporation of different metals modulates the redox property of azurin – Ni^2+/1+ azurin (E° = −590 mV) exhibits approximately 1 V lower E° values than and Cu^2+/1+ azurin (E° = +310 mV) [75]. Interestingly, replacing Cu with Ni in multiple variants of Az nearly always results in lowering of E°ʹ by ~1V, making the E°ʹ of the engineered proteins additive and predictable. By using secondary coordination sphere modulated variants of Cu and Ni azurin, entire 2V of biological redox range was covered [76]. Apart from ET, Fe^II and Ni^II containing azurin have also been engineered for catalysis by removing their Met121 coordination that opens up the metal site for substrate binding. While replacing Cu^II in WT azurin with Fe^II resulted in an redox inactive protein, an Fe^II containing M121E mutant of azurin was redox active with an E⁰ʹ value of +320 mV [77, 78]. This Fe-azurin variant was further engineered as a superoxide reductase mimic by adding in a lysine that interacted and stabilized the binding of superoxide at the Fe center. Similarly, Ni^II containing azurin was engineered into acetyl coenzyme synthase through M121A mutation [75, 79]. This Ni-azurin variant could access three (Ni^I/Ni^II/Ni^III) distinct oxidation states and bound CO and −CH₃ groups with biologically relevant affinity. In summary, by selective choice of metals and rational design of protein ligands and secondary coordination sphere interactions, cupredoxins can be engineered to perform a wide variety of catalytic reactions.

4. Use of heme/non-heme iron and calcium for redox sensing and signaling:

Cells use diverse metalloproteins to sense their redox microenvironment and relay this signal downstream for metabolic decisions [80, 81]. A common mechanism of sensing in these metalloproteins is binding of redox molecules like CO, NO, O₂ and H₂O₂ to metal cofactors followed by subsequent conformational changes and protein-protein interactions [82, 83]. Genetic and proteomic analysis of microbiome has revealed a number of these sensing/signaling metalloproteins [84]. However, structural and mechanistic details of only few of these redox sensing/signaling metalloenzymes have been deciphered. Below, we describe a few of these metalloenzymes with a specific focus on their sensitivity and metal of choice.

4.1. Heme iron in H-NOXs for sensing nitric oxide:

Bacteria employ nonlethal, sub-micromolar NO concentrations as signaling agents to control bacterial communal behavior [85]. NO regulates biofilm formation and dispersal, motility, symbiosis, and quorum sensing [86]. Heme-nitric oxide/oxygen binding (H-NOX) domains function as sensors for this gaseous signaling agent [86]. HNOXs share high sequence homology with the heme-binding domain of mammalian NO receptor, soluble guanylate cyclase. Extensive structural and site-specific mutational studies on these proteins have revealed the reason behind their 1000-fold higher selectivity to NO over O₂ [87]. The NO sensitivity in HNOXs arise from severe distortion of heme cofactor from planarity. A conserved proline residue (P115 in Tt H-NOX) and a conserved hydrophobic residue (I5 in Tt H-NOX) contact neighboring heme pyrrole from opposite sides, causing a distinct high-energy kink in the porphyrin (Fig. 4a) [88]. Binding of NO ligand leads to dissociation of the axial His residue due to its strong trans effect. The axial His dissociation results in a five-coordinate heme-nitrosyl complex and subsequent relaxation of heme distortion (Fig. 4b). Unlike NO, O₂ is not a strong trans-directing ligand and is unable to relax the heme cofactor explaining strong selectivity of HNOXs for NO (over O₂). Heme conformational dynamics is effectively used by HNOXs for sensing slight modulations in cellular redox states.

Figure 4.

Biological metalloprotein sensors. (a) Catalytic site of Tt H-NOX showing the distorted heme [PDB: 4FDK] (b) Schematic showing binding of NO relaxes heme from its distorted form (c) Active site of prolyl HIF-1α hydroxylase showing its nonheme iron active site. [PDB: 6EYI] (d) Binding of calcium ions results in huge conformational changes in CaM protein structure [89]

4.2. Non-heme iron in HIF-1α hydroxylases for sensing low oxygen:

Mammalian cells utilize nonheme iron containing HIF-1α hydroxylases for sensing low cellular oxygen concentrations or hypoxia [90]. The HIF-1α hydroxylases belong to the 2-oxoglutarate/nonheme iron dependent-oxygenase family of enzymes that use a conserved two-histidine, one-carboxylate motif to coordinate Fe²⁺ at the catalytic site (Fig. 4c) [90]. These endogenous protein ligands form a ‘facial triad’ that occupies three of the six possible coordination sites in an octahedral coordination geometry. The remaining three coordination sites are occupied by two to three labile water molecules that are readily displaced by substrates [91]. The enzyme first binds to 2-oxoglutarate followed by protein substrate HIF-1α and finally oxygen for initiation of hydroxylation reaction. The hydroxylation of prolyl/aspargyl residues in HIF-1α facilitates large conformational changes in the protein backbone that eventually leads to its proteolytic degradation. Under hypoxia, nonheme iron is unable to bind O₂ and hydroxylate HIF-1α cofactor such that the HIF-1α signaling pathway is deactivated [92]. So, why is nonheme iron uniquely chosen for sensing low O₂ during HIF-1α signaling? The answer can be in low oxygen affinities of nonheme iron proteins like HIF-1α hydroxylases (K_m > 30 μM) that match well with hypoxic environment in mammalian cells ([O₂] < 50 μM).

4.3. Intracellular calcium sensing by calmodulin:

Rapid changes in cytosolic Ca²⁺ concentration is responsible for a wide number of cellular responses, including muscle contraction and neuronal firing [93]. At rest, the cytosolic Ca²⁺ concentration is maintained at ~100 nM, but can increase to more than 100 μM, when Ca²⁺ channels open in the plasma membrane [93]. Detection of this steep change in Ca²⁺ concentration depends on Ca²⁺ binding protein, calmodulin (CaM). The structure of CaM reflects its refined Ca²⁺ sensing abilities. Calmodulin is a 16.7 kDa protein consisting of two lobes connected by a flexible and unstructured or α-helical linker. Each lobe has two EF-hands, which can each coordinate one Ca²⁺ ion (Fig. 4d) [94]. The C-terminal lobe of CaM binds Ca²⁺ with six times higher affinity (K_D = 2.5 μM) than the N-terminal lobe (K_D = 16 μM), allowing CaM to sense Ca²⁺ across a wide concentration range. Hydrophobic patches on the inside of each lobe recognize binding motifs on interaction partners, and thereby facilitate CaM binding and target regulation. Ca²⁺ binding to CaM and its’ binding to target proteins allosterically affect the affinity of each other, and this protein-protein interaction specifically modulates the conformation of CaM. In this way, the small CaM protein displays a range of binding and regulation properties [95, 96]. So, why was Ca²⁺ uniquely chosen for signaling events over other redox inactive bioavailable metals like Mg²⁺. The answer could be in its flexible coordination geometry [97]. Calcium is readily accepted by sites of irregular geometry that will, for instance, not accommodate Mg²⁺. The coordination flexibility of calcium (the coordination number is usually 6–8, but up to 12 is possible), its variable bond length/angle are at sharp variance with those of Mg²⁺, which, because of its smaller size (0.65 as compared with 0.99Å for Ca²⁺) and much lower polarizability, requires a fixed octahedral geometry with six coordinating ligands and minimal bond length variability.

Recently, a highly selective lanthanide (Ln^III) binding protein analogous to CaM has been discovered in a methylotrophic bacteria and called lanmodulin [98]. Similar to CaM, lanmodulin possesses four metal-binding EF hand motifs. In contrast, however, lanmodulin undergoes a large conformational change from a largely disordered state to a compact, ordered state in response to picomolar concentrations of all Ln^III (Ln = La–Lu, Y), whereas it only responds to Ca^II at near-millimolar concentrations. This exciting discovery provides insights into how biology selectively recognizes low-abundance Ln^III over higher-abundance Ca^II. These selective metal sensing and capturing enzymes can be used in biotechnologies for detecting, sequestering, and separating of lanthanide metals.

5. Use of redox inactive zinc and magnesium in biology

Zinc and magnesium are the most commonly utilized metal cofactors comprising ~16% and ~9% of all enzymes, respectively [1]. The two metal ions either participate directly in catalysis as Lewis acids or are important for maintaining protein structure and stability. The ionic radii of the two metal ions are similar. However, switching the two metals often results in activity loss. For instance, all enzymes utilizing ATP use Mg²⁺ as a co-substrate to activate the terminal phosphoryl group for its transfer. This role of Mg²⁺ cannot be carried out by Zn²⁺.This is probably because Mg²⁺ generally binds to oxygen ligands while Zn²⁺ prefers softer nitrogen and sulfur as ligands especially if the coordination number is low [99].

6. Summary

Nature has selected metal ions from the periodic table for specific functions. Redox inactive Zn²⁺ and Mg²⁺ are used for structural stability or as Lewis acids for catalysis. Redox inactive Ca²⁺, on the other hand, is used for signaling and communication. Nonheme iron is used for sensing low oxygen while heme iron is used for sensing NO. Copper is used for ET at high redox potential ranges, heme iron is used for ET at mid-redox ranges while Fe-S clusters are used for ET at low redox ranges. In regard to oxidoreductive reactions, Heme-Cu is used for oxygen reduction while Heme-Fe is used NO reduction. [Mn₄Ca] heterocluster is exclusively used for water oxidation while [Fe₇MoS₉C] is used for nitrogen reduction. Understanding why one metal is preferred over another is crucial for our fundamental understanding of these biological processes. The best way to perform these studies is swapping out one metal with another, which is not trivial due to high affinity of proteins for selective metal ions. Small molecule model complexes or enzyme models of these enzymes where the metal site is endogenously created can help elucidate the reason behind metal ion selectivity for different biological processes.[16, 20, 100, 101] Study of why one metal is used over another for certain reactions or biological processes will help design improved catalysts, sensors and metal-based pharmaceutical drugs.

Abbreviations:

HCO

Heme-Copper Oxidase

NOR

Nitric Oxide Reductase

NO

Nitric oxide

E

Reduction Potential

SOD

Superoxide Dismutase

SCS

Secondary coordination sphere

ET

Electron transfer

CaM

Calmodulin

Ln

Lanthanide