1. Intrinsic Physicochemical Properties of the Metal Cations Determine to a Great Extent the Metal Selectivity of the Binding Site
1.1. Competition between Ca2+ and Sr2+ in Calcium Receptors/Signaling Proteins
Strontium (Sr
2+), a member of the alkaline earth metal group of the periodic table, is employed in medicine as a diagnostic or therapeutic agent. Its radioactive isotopes
85Sr and
89Sr—with a half-life of ~65 and ~51 days, respectively—are used to follow the Ca
2+ kinetics and treat bone cancer sufferers
[1][2][14,15]. Another strontium formulation, strontium ranelate, containing stable, non-radioactive strontium isotopes, has been shown to exert beneficial effects in treating people with osteoporosis, mostly postmenopausal women
[1][3][4][5][6][14,16,17,18,19]. It has been shown that strontium exerts a dual beneficiary effect: it reduces the bone degradation and promotes bone anabolism
[1][3][4][5][6][14,16,17,18,19].
Strontium’s clinical applications stem from its ability to closely imitate biogenic calcium ions (Ca
2+) and follow Ca
2+-specific pathways involved in cell signaling and bone formation. Both Ca
2+ and Sr
2+ are spherical, doubly charged “hard” cations that strongly prefer “hard” oxygen-containing ligands (side chains of Asp
−/Glu
−, Ser, Asn/Gln and backbone peptide groups). Their ionic radii are similar: 1.0/1.06 Å for Ca
2+ and 1.18/1.21 Å for Sr
2+ in hexacoordinated/heptacoordinated complexes, respectively
[7][20]. The respective hydration free energies are also quite close: −359.7 kcal/mol for Ca
2+ and −329.8 kcal/mol for Sr
2+ [8][21]. In the human body, the two metals behave similarly, both exhibiting distinct bone-seeking properties
[1][14]. They possess flexible coordination/hydration spheres comprising 6 to 9 ligands
[9][10][22,23]. As a Ca
2+-mimetic species, Sr
2+ radio-isotopes preferentially accumulate at sites of increased osteogenesis, thus focusing the radiation exposure on the cancerous regions. By mimicking Ca
2+, non-radioactive Sr
2+ has been postulated to bind and activate the calcium-sensing receptor (CaSR), a representative of the G-protein coupled receptor (GPCR) family
[1][3][6][11][12][14,16,19,24,25]. CaSR activation triggers a cascade of signaling pathways promoting apoptosis of bone tissue-degrading cells (osteoclasts) and differentiation of bone-synthesizing cells (osteoblasts). In addition to CaSr, Sr
2+ can also compete with Ca
2+ in binding to proteins such as parvalbumin, alkaline phosphatase, calbindin, Ca
2+-sensitive ATPase, and calmodulin
[13][14][26,27].
The intimate mechanism of Sr
2+ activation of CaSR is, however, not fully understood. Several important questions remain: How Ca
2+/Sr
2+-selective are the metal binding sites of the activated CaSR? How efficiently could the “alien” Sr
2+ compete with the native Ca
2+ for binding to the receptor? What are the key determinants of the metal affinity/selectivity of CaSR in the activated state? To address these questions, a QM/PCM modeling study has been undertaken
[15][28] and the Gibbs free energies for the Ca
2+ → Sr
2+ exchange in different dielectric media have been evaluated, as shown in Equation (1):
In Equation (1), [Ca2+/Sr2+-CaSR] and [Ca2+/Sr2+-aq] represent the metal cation bound to receptor ligands inside the binding pockets and unbound outside the binding cavity (in bulk solvent), respectively. A positive free energy for Equation (1) implies a Ca2+-selective site, whereas a negative value suggests a Sr2+-selective one.
The Ca
2+/Sr
2+-loaded binding pockets of CaSR (Sites 1–4
[6][19]) have been modeled and their thermodynamic characteristics evaluated
[15][28]. Site 1 is situated in a loop region where backbone peptide groups of Ile81, Ser84, Leu87 and Leu88 orbit the metal cation. The metal center in Site 2 is directly coordinated by the side chain of Thr100 and indirectly (via a water molecule) by the side chain of Asn102. The calcium ion in Site 3 is bound in an outer-shell mode (via water molecules) to Ser302 and Ser303, whereas the side chain of Asp234 and backbone carbonyl groups of Glu231 and Gly557 coordinate the metal cation in Site 4 in an inner-shell fashion. The role of bound Ca
2+ ions in activating CaSR has been found to be mostly structural: they stabilize the active state by strengthening the homodimer interactions between membrane-proximal domains
[6][19].
The modeling study reveals that the metal binding sites—although comprising a different number and type of protein ligands, overall structure and charge state—are all selective for Ca2+ over Sr2+. Thus, strontium is predicted to be unable to dislodge the cognate calcium from the respective metal centers. The four binding sites, regardless of their structural differences, exhibit almost equal metal selectivity.
Data analysis suggests that several factors—such as the number and type of protein ligands, charge state of the binding pocket and its solvent exposure—do not seem to play any significant role in governing the competition between Ca
2+ and Sr
2+ in CaSR. Rather, these are the intrinsic physicochemical properties of the two competing metal species that to a great extent orchestrate the process: Ca
2+ has higher charge density than Sr
2+ (0.40 vs. 0.27 e/Å
3, respectively) and is a better Lewis acid than its bulkier counterpart. As a result, Ca
2+ interacts more favorably with the protein ligands than Sr
2+, yielding higher absolute value interaction energies. Similar conclusions have been drawn for the competition between Ca
2+ and Sr
2+ in parvalbumin—a representative of the EF-hand family of proteins involved in calcium signaling
[14][27]. CD and EF heptacoordinated binding sites, comprising Asp
−, Glu
−, Ser side chains, water and backbone peptide ligands, have been predicted to be Ca
2+/Sr
2+ selective. The conclusions have been confirmed by experimental measurements
[14][27].
1.2. Fe2+ vs. Mg2+, Mn2+ and Zn2+ in Non-Heme Iron Proteins
Iron, a redox-active element with oxidation state alternating between +2 and +3 (and sometimes +4), plays a key role in a number of essential biological processes such as respiration, cell division, nitrogen fixation, oxygen transport, nucleotide synthesis, oxidant protection, gene regulation, and protein structure stabilization
[16][17][30,31]. In mononuclear non-heme iron proteins, the metal cation is usually coordinated to His and Asp
−/Glu
− side chains
[18][32]. Typical Fe
2+ binding site configuration is His
2(Asp
−/Glu
−)
1, designated as “2-His-1-carboxylate facial triad motif”
[19][33], which has been found in a large group of iron dioxygenases, hydrolases, and synthases
[19][20][21][22][23][24][33,34,35,36,37,38]. Other combinations between His and acidic residues exist as well: His
1(Asp
−/Glu
−)
2, His
2(Asp
−/Glu
−)
2 and His
3(Asp
−/Glu
−)
1 [18][32]. The coordination number of Fe
2+ varies between 5 and 6 with water or substrate molecules supplementing the coordination sphere. Inside the cell, Fe
2+ faces a competition from other biogenic metal species (i.e., Mg
2+, Mn
2+, and Zn
2+) for binding the protein. Although these metal species are characterized with the same charge and similar ionic radii (R
Fe2+ = 0.78 Å, R
Mg2+ = 0.72 Å, R
Zn2+ = 0.74 Å, and R
Mn2+ = 0.83 Å for hexacoordinated cations
[7][20]), they possess different ligand affinities (due mostly to varying charge accepting abilities), as reflected in the Irving−Williams series
[25][39]:
Magnesium and manganese ions, positioned at the far left-hand side of the series, have weaker ligand affinities than Fe2+. Zinc cations, on the other side, are much stronger binders than their Fe2+ counterparts and form, as a rule, more stable complexes. Several outstanding questions arise:
-
How does the Fe2+ binding site sequesters the “right” (native) cation from the cellular fluids and protect itself from attacks by other biogenic cations such as Mg2+, Mn2+, and Zn2+?
-
What kind of selectivity strategies do iron binding sites employ toward metal cations having different ligand affinities and cytosolic concentrations?
-
What are the key factors governing the metal selectivity in Fe2+ proteins?
These questions have been addressed in a modeling study employing a combined DFT/PCM approach
[18][32].
Results obtained reveal the following trends: (i) Mg2+ and Mn2+ are not able to dislodge Fe2+ from the respective binding sites, as evidenced by the positive free energies of metal substitution in both the gas phase and condensed media. This finding comes as no surprise in view of the weaker ligand affinities of Mg2+ and Mn2+ cations relative to those of the Fe2+ cation. However, Mn2+—being closer in physicochemical properties to Fe2+ than Mg2+ to Fe2+—is a more potent iron contender than Mg2+ (less positive free energies for the Fe2+ → Mn2+ exchange than for the Fe2+ → Mg2+ substitution). (ii) The Fe2+ binding sites, however, are ill protected against attacks by the rival Zn2+ cations, which form more stable complexes and are able to displace Fe2+ from the respective metal centers (negative ΔG values for the Fe2+ → Zn2+). (iii) Solvation does not appear to be a key determinant of the metal selectivity in these systems, as it weakly affects the free energies of metal substitution and does not alter the trends observed in the gas phase.
The theoretical results imply that Mg2+ cannot successfully compete with Fe2+ in these binding sites (relatively high positive free energies evaluated for the Fe2+ → Mg2+ substitution). The major determinant of the high Fe2+/Mg2+ selectivity in these systems is the chemical nature of the contending metal species which confers on Fe2+ higher ligand affinity than on Mg2+.
Divalent manganese and iron cations are neighbors in the Irving−Williams series, exhibiting similar ligand affinities, ion radii (see above), coordination preferences (penta- or hexacoordinated first-shell ligand complexes), and cytosolic concentrations (in the micromolar range
[26][1]), thus appearing to be comparably strong contenders for protein binding sites. The calculations show, not surprisingly, that iron centers, although still preferably binding Fe
2+ (the latter being a better complexation agent than Mn
2+), are weakly selective for Fe
2+ over Mn
2+ and are vulnerable to Mn
2+ attacks. This is evidenced by positive, but low in absolute value, free energies (just a few kcal/mol) of the Fe
2+ → Mn
2+ exchange in both the gas phase and protein environment. The poor Fe
2+/Mn
2+ selectivity—supposedly resulting in the easily surmountable thermodynamic barrier for the Fe
2+ → Mn
2+ substitution—might, however, be advantageous for the cell metabolism and/or cell survival. Under conditions of Fe
2+ deprivation, the iron protein may sequester Mn
2+ cations from the surrounding fluids which, due to the close resemblance between the two metal species, might secure uninterrupted cell metabolism
[27][28][40,41].
The zinc cation, characterized by greater ligand affinity than Fe
2+, can outcompete the iron cation and displace it from its binding sites regardless of their composition, structure and solvent exposure (negative free energies for the Fe
2+ → Zn
2+ substitution in the entire series of complexes. The results are in line with a number of in vitro experiments showing that, indeed, Zn
2+ binds to the host protein with much greater affinity than Fe
2+ [29][30][31][32][33][42,43,44,45,46]. Note that, although the protein preferentially coordinates to Zn
2+ in vitro, it binds and is activated by Fe
2+ in vivo
[29][30][31][32][33][34][42,43,44,45,46,47]. Inside the living cell, since the protein alone is not able to repel the attacks by the rival Zn
2+, it is the cell machinery which, by tightly controlling the metal homeostasis and maintaining the free Zn
2+ concentration at very low levels (in the picomolar to femtomolar range
[26][1]), turns the balance in favor of Fe
2+ whose free cytosolic concentration is in the micromolar range
[26][1].
The theoretical study suggests that the inherent physicochemical properties of the contending metal species, reflected in the Irving−Williams series, are the major factor governing the metal selectivity in the non-heme iron centers. In addition, the free cytosolic concentration of the metal competitors, which correlates inversely with the Irving−Williams series, also affects the process of metal competition in vivo.
Notably, by using theoretical calculations, Kumar and Satpati have also found that the metal affinity of the wild-type and mutant CRISPR-associated protein 1 (with the first-shell E190, H254 and D268 residues lining the metal binding site) is in the order Ca
2+ < Mg
2+ < Mn
2+ which is in agreement with the Irving−Williams series
[35][48].
1.3. Competition between Cr3+ and Fe3+, Fe2+, Mg2+ and Zn2+ in Chromodulin
Chromodulin (low molecular weight chromium-binding substance, LMWCr) is a 1.5 kDa oligopeptide that, in Cr
3+-loaded form, plays an essential role in the metabolism of carbohydrates and lipids by interfering with the insulin signaling pathways
[36][49]. It has been implicated in reducing the insulin resistance in type 2 diabetic patients
[37][38][50,51]. Although chromodulin’s primary and 3D structure have not yet been unraveled, it is known that chromodulin contains only four amino acid types in the ratio of Glu
−:Gly:Cys:Asp
− = 4:2:2:2
[39][52]. An indispensable integral part of the oligopeptide in its active (holo-) form are four chromium cations in the oxidation state of 3+, located in two metal binding sites containing three and one Cr
3+ ions (“3 + 1” mode of binding). Structural investigations on holo-chromodulin are not abundant and, to date, only limited information is available about the basic characteristics of the Cr
3+ binding sites. The seminal experimental study of Jacquamet et al. sheds light on the following aspects of the metal-occupied binding centers
[40][53]: (i) chromium does not alter its oxidation state upon binding and forms Cr
3+ complex with the host chromodulin; (ii) the four Cr
3+ cations are clustered into two separated centers containing 3 and 1 metals; (iii) metal cations are six-coordinated surrounded by oxygen-containing ligands arranged in a nearly octahedral fashion; (iv) cysteine side chains, oligomer end groups as well as water molecules appear not to be likely ligands for the metal, nor have sulfur bridges involving cysteines been identified; (v) Cr
3+ cations in the trinuclear center are organized in the form of a (not ideal) isosceles triangle with the shorter side intermetallic distance of ~2.79 Å and longer sides lengths of ~3.79 Å; (vi) the metal ion pair forming the shorter side of the triangle is “glued” by hydroxo (but not oxo) bridges, whereas Asp
−/Glu
− carboxylate bridges connect these Cr
3+ cations with the third, more distant Cr
3+.
Note that the paradigm of Cr
3+ binding to chromodulin is especially intriguing from both experimental and theoretical points of view, since LMWCr appears to be the only molecule of biochemical importance whose native metal cofactor is Cr
3+. This prompts several questions: Why chromium? What are the advantages of binding Cr
3+ over other cellular biogenic metal species (e.g., Fe
3+, Fe
2+, Mg
2+, Zn
2+)? What factors influence the metal cation competition in chromodulin? These questions have been addressed recently by modeling the holo-chromodulin binding sites (following closely the findings from the experimental structural studies mentioned above) and evaluating the free energies of metal competition between the native Cr
3+ and other biologically relevant metal species such as Fe
3+, Fe
2+, Mg
2+ and Zn
2+ [41][54]. A combination of density functional theory (DFT) calculations and polarizable continuum method (PCM) computations has been employed.
Guidelines derived from the experiment (see above) have been used to model the structures of the mono- and trinuclear Cr
3+ metal centers. The basic characteristics of the optimized metal complexes are in agreement with the available experimental data: (i) the Cr
3+ cations are six-coordinated with oxygen-containing ligands (acetates, backbone amide groups and hydroxyls) orbiting the metals in an octahedral fashion; (ii) neither water nor sulfur-containing ligands coordinate the metal cations; (iii) in the trinuclear sites, the metal cations form an isosceles triangle with two Cr
3+ from the shorter side being connected by hydroxo bridges, whereas the third Cr
3+ is linked to them by acetate bridges and a hydroxo-bridge. Note that the calculated geometrical parameters of the triangle are in good agreement with those evaluated experimentally
[40][53]: R
Cr–Cr shorter side (Calc) = 2.79–2.80 Å and R
Cr–Cr shorter side (Exp) = ~2.79 Å; R
Cr–Cr longer side (Calc) = 3.58–3.65 Å and R
Cr–Cr longer side (Exp) = ~3.79 Å.
The experiment does not provide information about the nature of the non-bridging oxygen-containing ligands. Therefore, several metal centers have been constructed comprising various combinations of non-bridging acetates and backbone amides while, at the same time, maintaining the triangle structure with the respective OH− and acetate bridges.
The respective mononuclear binding sites have been modeled in agreement with the composition of their partner trinuclear structures: since the number of carboxylic residues in the host oligopeptide is 6 (2Asp− and 4Glu−, see above), the number of acetates in the mononuclear constructs has been adjusted to that in the trinuclear center so that the total number of carboxylates in the two binding sites (trinuclear and mononuclear) sums up to 6.
The data presented reveals that the trivalent Fe
3+ cannot outcompete Cr
3+ in either mononuclear or trinuclear metal centers, as demonstrated by positive ΔGs ranging from 2 to 7 kcal/mol for the former and between 14 and 20 kcal/mol for the latter. Evidently, the trinuclear chromium center is more resistant to Fe
3+ attack than its mononuclear counterpart (higher positive ΔGs for the trinuclear structure compared to those for the mononuclear binding site). These findings are not unexpected in view of the higher affinity of Cr
3+ to oxygen-containing ligands stabilizing the chromic complexes to a greater extent than the respective ferric structures. As seen from the data collected (bottom two rows), the energies of formation of chromic-single ligand complexes (ligand = species building metal binding sites in chromodulin, i.e., acetate, backbone amide and hydroxyl) are lower (more favorable) than their Fe
3+ counterparts. Note, however, that the trend reverses for sulfur- (CH
3S
−) and nitrogen-containing (imidazole) amino-acid residues, which preferably bind Fe
3+ over Cr
3+. Importantly, such ligands, which would promote Fe
3+ over Cr
3+ selectivity, do not participate in metal binding in chromodulin, as they are either absent from the amino acid sequence (His) or, even though part of the oligopeptide buildup (Cys), are, apparently, far from the metal binding center
[42][55].
Since the Lewis acidity/complexation power of divalent metals is lower than those of the trivalent cations, it is expected that M2+ metals would be weaker competitors to Cr3+. Indeed, Cr3+ binding sites are very well protected against attacks from divalent biogenic metals: ΔGs of the Cr3+ → M2+ (M = Fe, Mg and Zn) substitution vary between 127 and 165 kcal/mol in mononuclear complexes, and between 251 and 330 kcal/mol in the trinuclear constructs. The inherent physicochemical characteristics of the rival metal species emerge as the key factor controlling the metal competition in LMWCr.
2. Metal Coordination Number Is an Important Determinant of the Metal Selectivity
Sodium (Na
+) is an indispensable allosteric regulator in a number of signal-transducing proteins, such as neurotransmitter transporters and G-protein coupled receptors (GPCRs). These have been recognized as drug targets for psychiatric disorders
[43][44][57,58] and addictive behavior
[45][59]. In the holo-protein, sodium cation(s) are usually penta- or hexacoordinated and predominantly bind to oxygen-containing ligands such as Asp
−/Glu
− and Ser/Thr side chains, backbone peptide groups or water
[46][60]. The competition between Na
+ and Li
+ (non-biogenic metal cation known for its beneficial therapeutic effect on patients with mental disorders) in model sodium-binding sites have been studied by a combined DFT/PCM approach
[46][60], and key determinants controlling the selectivity for Na
+ over Li
+ in sodium proteins have been elucidated.
Two types of sodium binding sites have been modeled: flexible ones that allow for ligand rearrangement upon Na+ → Li+ exchange; and rigid binding pockets which preserve the original ligand arrangements in the “mother” sodium complex during metal substitution. The results reveal that the coordination number of the competing metals is an important factor in the selectivity process: Li+, when allowed to adopt its preferred tetrahedral ligand arrangement (decreasing its coordination number from 6 to 4), outcompetes the six-coordinated Na+ in the entire dielectric range. On the other hand, in rigid binding sites, competitiveness of the lithium cation decreases (positive free energies in protein environment; blue numbers in parentheses) as it is forced to adopt the unfavorable octahedral ligand surrounding of the native sodium: thus, ligand repulsion between the six bulky ligands around the small Li+ cation attenuates its efficiency of binding. Note that the rigid binding sites preserve the original, relatively large, binding cavity optimized to fit the size of the bulkier Na+ but not the smaller Li+, which additionally decreases the strength of the interactions between Li+ and ligands lining the pore.
3. Adjacent Metal Cation May Reverse the Metal Selectivity in Binuclear/Trinuclear Metal Binding Sites
Nickel-containing enzymes are of vital importance for a number of plants and primitive organisms, such as archaea, bacteria, fungi and low-trophic level marine eucaryota
[47][48][61,62], where they fulfill various tasks ranging from energy generation to detoxification, oxidative stress protection and virulence
[49][63]. To date, nine nickel-dependent biocatalysts, subdivided into non-redox (urease, glyoxylase I, acireductone dioxygenase and lactate racemase) and redox enzymes (CO-dehydrogenase, acetyl-CoA synthase, [NiFe]-hydrogenase, methyl-SCoM reductase, and Ni-superoxide dismutase), have been identified and characterized
[50][64]. The structure and composition of their metal centers are quite diverse, varying from mononuclear nickel binding sites to homo-binuclear and hetero-binuclear constructs
[48][49][50][51][62,63,64,65]. Cysteine is the predominant amino acid residue in redox enzymes, whereas histidines and aspartates/glutamates are the ligands of choice in the non-redox metaloproteins. Nickel’s valence state in the latter is 2+, while it alternates between 1+, 2+ and 3+ in the former.
Nickel enzymes can be deactivated/inhibited by other metal cations such as Zn
2+ (in
Escherichia coli glyoxalase I)
[52][66] and Ag
+ (in urease)
[53][54][55][67,68,69]. The inhibition by Ag
+ of urease (a pathogen in humans) is of high significance as it may have strong implications for pharmacology and medicine. The mechanism of silver antibacterial action in urease, however, is still not completely understood. Although the prevailing hypothesis postulates that Ag
+ binds to some sulfur containing amino acid ligands (not nickel-bound) at the periphery of the active site which disrupts the enzyme structure
[53][67], the substitution of Ni
2+ cations from the metal center by Ag
+ cannot be excluded
[54][55][68,69]. Therefore, it is of special interest to determine to what extent the Ni
2+ binding sites in urease (and other nickel enzymatic binding sites) are predisposed to Ni
2+ → Ag
+ substitution. Of note, information on the factors controlling the competition between the cognate Ni
2+ and abiogenic (“alien”) Ag
+ in biological systems is critically lacking.
To fill in the gap, the rivalry between Ni
2+ and Ag
+ in nickel enzymes has been studied by a combined DFT/PCM approach and key determinants of the process have been revealed
[56][70].
Glyoxalase I utilizes intracellular thiols to convert cytotoxic ketoaldehydes, such as methylglyoxal, into nontoxic D-hydroxy acids
[50][64]. The enzyme contains a mononuclear nickel center where two histidine and two glutamate amino acid residues along with two water molecules surround the metal in an octahedral fashion. Upon Ni
2+ → Ag
+ substitution, the binding site undergoes quite drastic structural changes resulting in a distorted tetrahedral silver complex with an acetate (model for the Glu
- side chain) and a water ligand transferred to the second coordination layer of the metal. The free energy calculations suggest that the nickel active center is well protected against Ag
+ attack and that the “alien” Ag
+ cannot dislodge the native Ni
2+ from its binding site as evidenced by highly positive ΔG
εs spanning the entire dielectric region. It is of note that the gas-phase exchange reaction, which is entirely dominated by electronic effects (being in favor of the divalent Ni
2+), is characterized by quite high positive ΔG
1s. These are attenuated in the condensed phase, where solvation effects favor to a greater extent the monovalent Ag
+, yielding smaller (although still positive) free energies of metal exchange. This is in line with findings from a similar investigation
[56][70] (not shown here), which demonstrates that increasing the number of charged/polar ligands, surrounding the metal (2 methylimidazoles and 2 acetates representing His and Glu
- amino acid side chains, respectively) and donating more charge to divalent Ni
2+ than to its monovalent contender, increases the competitiveness of Ni
2+ over Ag
+.
Acireductone dioxygenase is a mononuclear nickel enzyme involved in the methionine salvage pathway. In the process, methylthioadenosin is transformed into acireductone which, consequently, is converted to formate, carbon monoxide, and methylthiobutyric acid
[50][64]. The nickel binding site comprises three histidines, one glutamate and two water molecules octahedrally arranged around the Ni
2+ cation. Upon Ni
2+ → Ag
+ exchange, the complex isomerizes to a four-coordinated structure with a water molecule and methylimidazole ligand relegated to the metal second coordination sphere. The positive ΔG
εs evaluated for the metal substitution reaction suggest that Ag
+ cannot outcompete the native Ni
2+ in this system (due mainly to the presence of strong charge-donating ligands in the binding cavity) and that the acireductone dioxygenase binding site is reliably shielded from “alien” monocationic attack.
Urease, a nickel-dependent enzyme, catalyzes the hydrolytic decomposition of urea producing ammonia and carbamic acid which, subsequently, decomposes into another molecule of ammonia and carbonic acid
[49][63]. The enzyme has been recognized as a virulence factor in several pathogenic (mostly antibiotic-resistant) bacteria, which makes it a plausible target for antibacterial therapy
[53][54][55][67,68,69]. Urease comprises a homo-binuclear active center where the protein donates four histidines (two to each metal), an aspartate (in binding site 2) and a bridging carbamylated lysine to the metal cations.
Incorporating Ag+ in the binding sites, as expected, alters their structure (the silver cation prefers smaller coordination numbers and longer metal-ligand bond distances), although, as shown, in different fashion. Binding site 1 preserves its pentacoordinated structure but alters its shape from almost regular squire pyramidal construct (with the native Ni2+) to a distorted squire pyramidal one (with the “alien” Ag+). Moreover, the Ag1-ligand coordinative bonds are considerably elongated in comparison with the respective Ni1-ligand counterparts: the mean of Ag1-ligand bond distance is 2.488 Å, whereas that of the Ni1-ligand bond distance is 2.068 Å. The binding site 2 structure changes more dramatically upon Ni2+ → Ag+ exchange: it transforms from a hexacoordinated (nearly octahedral) complex to a semi-squire planar complex with two water molecules transferred to the second coordination layer of Ag+.
Thermodynamic evaluations reveal that the first Ag+ → Ni2+ exchange in buried binding pockets (ε = 4) in either site is favorable, and characterized with negative ΔG4s (−4.2 kcal mol−1 for binding site 1, and −2.7 kcal mol−1 for binding site 2. left-hand side). Substituting the second Ni2+ cation with Ag+, however, is thermodynamically unfavorable, as Gibbs free energies for the entire dielectric region are positive. Notably, the calculations suggest that only one mole of metal cations is exchanged during the process.
Why is the Ag+ → Ni2+ substitution in binuclear urease more favorable than that in the mononuclear glyoxalase I and acireductone dioxygenase (see above)? Why is the competitiveness of Ni2+ over Ag+ in bimetallic center compromised? This is mainly due to the presence of a second neighboring metal atom in the active center which, with its positive charge, attenuates the strength of the charge transfer through the bridging carboxylate to the other metal. Indeed, instead of coordinating to the “pure” strong charge-donating anionic carbamylated lysine (represented by CH3CH2NHCOO−), the Ni12+ cation, in fact, binds to the metal-bound cationic [CH3CH2NHCOO−-Ni22+]+ ligand characterized with a poorer charge-donating power that significantly attenuates the strength of the interaction between the Ni12+ and [CH3CH2NHCOO-Ni22+]+. Note that, in [CH3CH2NHCOO-Ni22+]+, the Ni22+ cation withdraws electron charge density toward itself, thus reducing the amount of charge that the metal-bound carbamylated lysine can donate to the adjacent Ni12+. As a result of the decreased charge-donating strength of the metal-bound carbamylated lysine, the Ni12+–ligand interactions are attenuated to a greater extent than those of the Ag1+–ligand interactions which, as expected, results in reduced Ni12+/Ag1+ competitiveness. The same considerations hold for the Ni22+/Ag2+ exchange as well.