The ternary complex of Mn2+, synthetic decapeptide DP1 (DEHGTAVMLK), and orthophosphate is a superb antioxidant Authors: Hao Yang, Ajay Sharma, Michael J. Daly, Brian M. Hoffman Published: 2024 ABSTRACT SignificanceThis study reveals the mechanism for protein radioprotection by the superb Mn-based MDP antioxidant containing Pi and decapeptide DP1, designed through considerations of Deinococcus radiodurans low molecular-weight Mn-antioxidants: The exceptionally active antioxidant in MDP is shown to be a Mn2+ (Pi/DP1) ternary complex, which, protects enzymes from radiation doses far exceeding the capabilities of Mn2+-Pi complexes alone. Substitutionally labile calcium ions coaccumulate with Mn in Mn-dependent organisms, and as an illuminating parallel, 1H NMR of DP1 is used to measure its affinity for diamagnetic Ca2+, whose coordination chemistry parallels that of paramagnetic Mn2+. These findings could unlock new strategies for enhancing cellular oxidative-stress resistance, spur the development of radiation-inactivated whole-cell vaccines, and potentially lead to other medical advances. Superoxide radicals (O2·−) formed through by radiation exposure or metabolic processes can cause amino acid–specific oxidative damages to the cellular proteome (1), but O2·− does not damage DNA or RNA (2–4). The consequences of O2·−-mediated protein damage in irradiated cells are severe and include disruption of protein homeostasis and DNA repair malfunction, known as “death by protein damage” (5–7). As partner with the enzyme manganese superoxide dismutase (MnSOD) in the defense against O2·−, cells can accumulate up to millimolar concentrations of low-molecular weight (LMW) Mn2+ antioxidant complexes, a mixture of complexes with orthophosphate (Pi), metabolites, amino acids, peptides, nucleosides, carboxylic acids as ligands, which, act as radioprotective agents by catalytically scavenging O2·− (8–16). These intracellular Mn antioxidant complexes are simply denoted here as H-Mn, based on their electron paramagnetic resonance (EPR) characterization as exhibiting high symmetry (11, 13, 14). The understanding of the intracellular function of the antioxidant H-Mn complexes has led to their utilization as reagents that enable the development and production of radiation-inactivated whole-cell vaccines (17–19). Ionizing radiation had been explored as a strategy for vaccine production since the early days of vaccinology, but this approach was largely abandoned decades ago due to the oxidative destruction of critical surface epitopes during pathogen inactivation. Unfortunately, while the radiation-induced inactivation of a pathogen by destruction of its nucleic acids is central to such vaccine production, the attendant radiation damage to the structural proteins on the pathogen’s surface was found to compromise antigenic potency (17). More recently, however, it has been shown that when bacteria and viruses are suspended in a solution of Mn antioxidants and subjected to the sterilizing doses of gamma radiation required in vaccine production, these antioxidants selectively protect surface proteins from the reactive oxygen species (ROS) generated in the surrounding aqueous solution, while leaving DNA and RNA inside cells unprotected. This reopened the door to radiation-inactivated whole-cell vaccines, with their advantageous offering of multiple epitopes to the immune system, unlike single-epitope mRNA vaccines (17–23). The efforts to develop Mn-antioxidants for vaccine production (17) and for in vivo radiation protection (24) led to the design of a synthetic decapeptide (DP1: DEHGTAVMLK) to act as a ligand to Mn2+ in the vaccine preparation medium, with amino acid residues, (Fig. 1) chosen based on their abundance in the bacterium Deinococcus radiodurans (2, 7). This was then formulated as a solution of orthophosphate (Pi) (pH 7.4), DP1, and Mn2+ (MnCl2) in proportions 25 mM/3 mM/1 mM, denoted MDP (2). The efficacy of MDP in providing radiation protection against exceptionally high gamma-radiation doses (>60 kGy) led to the development of radiation-inactivated whole-cell vaccines whose antigenic epitopes were preserved at the sterilizing radiation doses required for inactivation (2); the gamma radiation sterilization dose for viruses can exceed 40 kGy (17, 18, 20). MDP has the remarkable ability to protect the structure and function of large enzymes, such as the dodecamer glutamine synthetase, at gamma radiation doses exceeding 60 kGy, as exemplified by studies by Berlett and Levine (25), while others observed the effectiveness of MDP in protecting antigenic structures of viruses and bacteria during sterilizing exposure to gamma-rays (18, 19) and ultraviolet C radiation (21). Structure of DP1 and 600 MHz 1H NMR spectra. The residues that bind Mn2+, circled in blue, are identified by the broadening of their residues into undetectability by the addition of substoichiometric concentration of Mn2+ (0.5 mM) to DP1 (3 mM) in Tris buffer, pH 7.0. This report presents a detailed study of the nature of the labile Mn2+ species in MDP that confers radioprotection. To begin we use 1D 1H NMR, ENDOR, and ESEEM measurements to characterize the mode of DP1 binding to Mn2+, then measure the thermodynamic binding affinity of DP1 for the substitutionally labile Mn2+ ion through the use of isothermal titration calorimetry (ITC). In parallel, we use changes in the Mn2+ solution EPR spectrum to determine the individual binding affinities of both DP1 and of Pi for Mn2+, and further monitor Mn2+ ligation during a titration in which Pi and DP1 are added in the fixed ratio, [Pi]/[DP1] = 25/3, matching that in MDP. As Ca2+ is often coaccumulated with Mn2+, but at fivefold higher concentrations for the Deinococcus bacteria, Bacillus spores, and mitochondria (26–28), in an illuminating parallel, 1H NMR of DP1 is used to characterize the binding of DP1 to the substitutionally labile, but diamagnetic, Ca2+, whose binding affinity we confirm to parallel that of paramagnetic Mn2+ (29). Unexpectedly, we find that affinity of DP1 for Mn2+ is far too low to form the radioprotective Mn2+ complex of MDP and that in MDP it is the ternary Mn2+[Pi/DP1] complex that acts as a superbly radioprotective antioxidant complex. Experimental Methods ITC. ITC was performed using a MicroCal Itc200 calorimeter. For the titration experiment, the sample cell contains 240 μL of a 2 mM DP1 solution in 30 mM Tris buffer at pH 7.3. The solution of 200 mM Manganese chloride was then injected into the sample cell in 1 μL aliquots every 5 mins. Data were fit in Origin 7.0 using the one site model to obtain enthalpies (ΔH), entropies (ΔS), and binding constants (Ka). X-band Continuous Wave (CW) EPR. CW X-band EPR experiments were performed on a modified Varian E-4 spectrometer with a liquid nitrogen finger-dewar at 293 K. To measure the binding affinity, a series of 200 μL sample solutions was prepared to achieve a 1 mM Manganese chloride with varied concentration of DP1 and phosphate in Tris buffer at pH 7.3. 50 μL of the mixture is transferred to quartz capillary tube for the X-band EPR measurements. The changes in amplitude of the 55Mn2+ g-2 EPR sextet in the presence of added ligand were used to determine the fraction of ligand-bound Mn2+. The normalized fraction of bound Mn2+ is then plotted against the known concentration of DP1 and phosphate for the isothermal curve fitting and extracting binding constant Ka. 600 MHz NMR. 1D 1H NMR spectra were collected on 3 mM of DP1 in Tris buffer containing ~10% D2O, which were recorded at 290 K on a Bruker Neo system operating at 600 MHz and equipped with QCI-F cryoprobe to enhance 1H sensitivity. The presaturation pulse was typically 1 μs, centered at the water frequency. EPR/ENDOR/ESEEM. Measurements are described in SI Appendix. Materials. DP1 was synthesized as described (1, 2). Other reagents were purchased from Sigma-Aldrich. Results NMR and DPI Binding Sites for Mn2+. Previously, an extensive 1H and 13C NMR study of DP1 identified and assigned the resonances seen in spectra of DP1, while the addition of substoichiometric Mn2+ was used to identify residues of DP1 that bind to Mn (30). We here performed 1D 1H NMR experiments and tracked the paramagnetic broadening of 1H peaks upon substoichiometric addition of the high-spin (S = 5/2) Mn2+ (Fig. 1, and see below). As reported (30), peaks from six of the DP1 amino acids do not change significantly with the substoichiometric addition of Mn2+, while 1H peaks of the first three amino acids (Asp1, Glu2, and His3) broaden and disappear quickly, Fig. 1, indicating that these three residues of DP1coordinate to Mn2+, while ESEEM measurements presented in SI Appendix demonstrate that 14N of the histidyl imidazole indeed is bound. The NMR measurements further demonstrate that in fluid solution, the Mn2+ ligands undergo rapid exchange. Namely, Mn2+ is a substitutionally labile ion, with ligand-Mn2+ complexes, is in rapid equilibrium with unbound Mn2+ (29, 31). Thus, it was to be expected that the weakly bound DP1 (see below) binds to Mn2+ in multiple, rapidly interconverting conformations, and this is confirmed below; it is further seen that the same is true for Mn2+ complexes formed in the presence of both DP1 and Pi. ITC and EPR Measurements of Mn2+ Affinity for DP1 and Pi. What is the affinity of DP1 for Mn2+: Does DP1 bind Mn2+ tightly or weakly? As will be seen, different spectroscopic methods of measuring ligand coordination to Mn2+ have their individual strengths and weaknesses, so we first present a direct, thermodynamic measurement of binding of DP1 to Mn2+, then show the value of EPR in monitoring binding of DP1 and of Pi to Mn2+. As an illuminating parallel, NMR of DP1 is used to measure the affinity of DP1 for the diamagnetic Ca2+, which in general has similar binding behavior to that of the paramagnetic Mn2+, the two forming substitutionally labile ligand complexes with similar affinities. With this foundation, we discuss why the “apparent” binding affinity obtained by using NMR of a peptide to monitor its binding to a paramagnetic metal ion such as Mn2+ (or Cu2+) is orders of magnitude greater than the actual thermodynamic affinity (30). A series of ITC (defined above) experiments were performed on a MicroCal VP-Isothermal Titration Calorimeter instrument, with aliquots of Mn2+ added to DP1 in noncoordinating Tris Buffer (32), 30 mM, pH 7.3; Fig. 2, upper presents a representative baseline-corrected raw dataset. In such an experiment, the fraction of DP1 bound in the Mn2+(DP1) complex at titration point i, f([Mi]), equals the ratio of the heat liberated during the titration up to point i, and the total heat liberated during a full titration, which is proportional to the molar binding enthalpy. Analysis of the data with the Origin 5.0 software package from MicroCal showed that it was described by the binding equilibrium,[1]Mn2++DP1=Mn2+(DP1). ITC titration, aliquots of Mn2+ were added to DP1(2 mM) in 30 mM Tris Buffer, pH 7.3. (Upper) a representative baseline-corrected raw dataset, (Lower) fitted binding isotherm (Eq. 1) with, ∆H(7.3) = −5.5(2) kCal/M, association constant, Ka(pH 7.3) = 98(6) M−1 n = 1. with a fit to the associated isotherm (Fig. 2, Lower) yielding as the enthalpy of binding, ∆H(7.3) = −5.5(2) kCal/M, and the association constant, Ka(pH 7.3) = 98(6) M−1. Analogous “inverse” titrations were carried out in which aliquots of DP1 (Fig. 3A) or Pi (Fig. 3B) were added to a fixed concentration of Mn2+ in Tris buffer (pH 7.3) at ambient temperature, and the formation of the Mn2+–DP1 and Mn2+–Pi complexes was measured through changes in the amplitude of the Mn2+ fluid–solution X-Band EPR spectrum. Pi is a common intracellular ligand of Mn2+, the Mn2+-Pi complex is a major component of the antioxidant H-Mn pool of metabolite complexes of Mn2+, and Pi is a component of the MDP solution that offers an attractive medium for generating vaccines to radiation-resistant pathogens. Importantly, the MDP (Mn–DP1–Pi) mixture also serves as an analog of the natural pool of Mn antioxidants in Deinococcus bacteria––because a standard MDP solution has the same concentrations of the corresponding pool of inorganic and organic metabolites in Deinococcus bacteria, with the ratio, Mn/DP/Pi = 1 mM/3 mM/25 mM (2). Titration curves (pH 7.3) of ~1 mM Mn2+ in Tris 30 mM buffer, pH 7.3. Fractional decrease in EPR amplitude (1 − h/h0) during titration of the Mn2+ with: (A) DP1 (B) Pi, (C) [Pi]:[DP1] = 8:1 overlaid with fit to isotherm Eq. 2. Inset: Room temperature solution state X-band EPR spectra recorded using quartz capillaries. The room temperature solution state X-band EPR spectrum of aqueous Mn2+ displays a sharp six-line spectrum characteristic of the electron-nuclear hyperfine coupling to 55Mn (I = 5/2). During titrations of the substitutionally labile Mn2+ with both ligands, the resulting ligand-Mn2+ complex is in rapid equilibrium with unbound Mn2+, and thus at any point in a titration, the EPR spectrum of high spin (S = 5/2) Mn2+ is an average of those for the free and bound Mn2+. As seen in the insets to Fig. 3 A and B, the sextet homogeneously broadens and decreases in amplitude as the average population of complex increases during the titration (Inset). Fig. 3 A and B themselves plot the magnitude of changes in EPR signal amplitude/intensity during the titrations. When binding is weak for such dynamically formed complexes ligand (L) complexes, namely when Ka/[L]tot << 1, the binding isotherm associated with the reaction of Eq. 1 takes the form of Eq. 2,[2]fL=KaL1+KaL, where f[(L)] is the fraction of ligand-bound Mn2+ at total ligand concentration, [L], as measured by the fractional change in the EPR amplitude at ligand concentration [L], h(L),[3]fL=h0-hL/h0-h∞, where h([0]) and h([∞]), respectively, are the EPR amplitudes for free Mn2+ and for the complex. The fit of this isotherm for the binding of DP1 to Mn2+ yields the binding constant, Ka(pH 7.3) = 40(3) M−1 (Fig. 3A), comparable in magnitude to that obtained by ITC, confirming the utility of the EPR approach. Thus, given that limitations in the available amount of DP1 precluded a closer approach to the titration endpoint, one may confidently say that the thermodynamic and spectroscopic techniques agree that the affinity of DP1 for Mn2+ is low, with Ka ~ 40 to 100 M−1 in the vicinity of pH 7.3. Correspondingly, in the EPR titration of Mn2+ in Tris buffer by Pi, pH 7.3, the plotted changes in intensity of the EPR spectra taken across the titration are fit by a binding isotherm with Ka(pH 7) = 385 (28) (Fig. 3B). Thus, for discussion purposes, we may say that Pi has a roughly 10-fold stronger affinity for Mn2+ than does DP1. To complete the binding measurements, Fig. 3C. shows the results of a “titration” in which the titration points contained 1 mM Mn2+ and increasing concentrations of Pi/DP1 in the concentration ratio of MDP, namely Pi/DP1 = 25/3, with the results plotted in terms of the Pi concentration. As can be seen, the formation of complexed Mn2+ is significantly enhanced when DP1 is added to a solution of Pi. NMR Measurements of DP1 Binding to Substitutionally Labile Ca2+. As discussed below and in SI Appendix, changes in the 1H NMR spectrum of DP1 upon additions of Mn2+ give information about the nature of DP1 binding to Mn2+ in the presence and absence of Pi but cannot be used to determine affinity constants. As illustrated in Fig. 1, by a concentration ratio of Mn2+/DP1 ~1/10, corresponding to an average population of Mn2+-bound DP1 of only a few percent according to the binding affinity thermodynamically determined above, the paramagnetic broadening of resonances-associated binding residues of DP1 completely quenches their NMR signal. To provide insights into the coordination geometry and dynamics of binding of DP1 to the substitutionally labile, paramagnetic Mn2+, we instead performed corresponding NMR titrations of DP1 by the substitutionally labile, but diamagnetic, Ca2+ ion, whose ionic radius corresponds to that of Mn2+ (29, 31). Mn2+ is well established as a biological analog for Ca2+ (33), and as we now show, Ca2+ binds to DP1 with a roughly comparable affinity as Mn2+. This correspondence is of further interest because in vitro radiation studies on the O2·− scavenging capacity of Mn-antioxidants show that radioprotection of enzymes is enhanced by excess Ca2+ when Mn2+ is present, but limited (34), even though Ca2+ is redox inactive and unable to scavenge ROS. After presenting these results, we next present measurements of radioprotection by Mn2+ complexes, then discuss the analysis of NMR titrations by paramagnetic Mn2+ itself. Fig. 4 A–C respectively, show the changes in a portion of the DP1 NMR spectrum (His3-σ2, Asp1-β and Glu2-γ) during progressive additions of Ca2+. During a titration of DP1 by the substitutionally labile Ca2+, bound and free DP1 are in rapid equilibrium, and as a result the DP1 signals are an average of those for the two pools of DP1. Thus, as shown in Fig. 4, the proton resonances homogeneously shift and decrease in amplitude with increased Ca2+ binding. Fig. 4D shows the fit of the Ca2+ binding isotherm to the (exact) equation, rather than isotherms of the form Eq. 2, for precision in comparisons:[4]Δσ=(Δσmax)∗f([Ca2+]),f([Ca2+])={[Dp1]+[Ca]+1/Ka-√[([Dp1]+[Ca]+1/Ka)2-4∗[Dp1]∗[Ca]]}/2[Dp1]. 600 MHz 1H NMR of titrating Ca2+ to DP1. Plot of change in chemical shift (Δδ(Hz) of His3-σ2 protons (A), Asp1-β protons (B), and Glu2-γ protons (C), during titration in which aliquots of Ca2+ (0 to 15 equiv) were added to DP1 (4.5 mM) in Tris Buffer (23 mM), pH 7.6, 10% D2O. 1H NMR spectra showing decrease in chemical shift with increasing [Ca2+] of a portion of the DP1 spectrum during the titration. The decrease is plotted against the concentration of Ca2+. (D) Ka is obtained by fit to Eq. 4. where f([Ca2+]) is the fractional change in the chemical shift at concentration [Ca2+]. As expected, the binding constants as measured for the three residues identified in Fig. 1, are of comparable magnitude, Ka(His3) = 111(20)M−1, Ka(Asp1) = 67(17)M−1 and Ka(Glu2) = 46(8) M−1, with all being comparable to that for binding to Mn2+ as measured using ITC (Fig. 1) and Mn2+ EPR (Fig. 3). However, the variation among the constants for the three as seen in the Ca2+ titration does provide an indication that this complex exhibits multiple rapidly interconverting conformers with differing Ca2+ coordination. Radiation Protection by Mn2+ Complexes: Implications for Binding of Pi and DP1 to Mn2+. The level of radioresistance conferred by the MDP complex was previously determined by testing the activity limits of enzymes exposed to gamma radiation in the presence or absence of MDP (1 mM MnCl2, 3 mM DP1, 25 mM Pi buffer, pH 7.4), as well as its individual components. The activity limit is defined as the dose of γ-rays which abolishes an enzyme’s function in vitro (SI Appendix, Fig. S3). There are three characteristics of the radioresistance conferred by Mn2+ and its complexes that we consider in light of the binding measurements reported here.1)The activities of the restriction endonuclease BamHI (2), the metabolic enzyme glutamine synthetase (2, 25), and the DNA repair enzyme T4 DNA Ligase (24), in Pi buffer solution (pH 7.4) is abolished by radiation at ~1.5 kGy.2)Mn2+ in a Pi solution at a near-neutral pH as buffered by the Pi (pH 7.4) forms a Pi complex with modest affinity (11, 35), as reaffirmed above (Fig. 3). This complex acts as an effective SOD antioxidant (8–12), in particular protecting the enzymes at gamma radiation doses up to 10 kGy (2, 24, 25).3)Addition of DP1 to a Pi solution of Mn2+ at the ratio, DP1/Pi = 3 mM/25 mM = 1/8.33, as in MDP, strongly enhances the effectiveness of the Mn2+ as an antioxidant, protecting the enzyme to >60 kGy (SI Appendix, Fig. S3) (2, 24, 25). 1)The activities of the restriction endonuclease BamHI (2), the metabolic enzyme glutamine synthetase (2, 25), and the DNA repair enzyme T4 DNA Ligase (24), in Pi buffer solution (pH 7.4) is abolished by radiation at ~1.5 kGy. The activities of the restriction endonuclease BamHI (2), the metabolic enzyme glutamine synthetase (2, 25), and the DNA repair enzyme T4 DNA Ligase (24), in Pi buffer solution (pH 7.4) is abolished by radiation at ~1.5 kGy. 2)Mn2+ in a Pi solution at a near-neutral pH as buffered by the Pi (pH 7.4) forms a Pi complex with modest affinity (11, 35), as reaffirmed above (Fig. 3). This complex acts as an effective SOD antioxidant (8–12), in particular protecting the enzymes at gamma radiation doses up to 10 kGy (2, 24, 25). Mn2+ in a Pi solution at a near-neutral pH as buffered by the Pi (pH 7.4) forms a Pi complex with modest affinity (11, 35), as reaffirmed above (Fig. 3). This complex acts as an effective SOD antioxidant (8–12), in particular protecting the enzymes at gamma radiation doses up to 10 kGy (2, 24, 25). 3)Addition of DP1 to a Pi solution of Mn2+ at the ratio, DP1/Pi = 3 mM/25 mM = 1/8.33, as in MDP, strongly enhances the effectiveness of the Mn2+ as an antioxidant, protecting the enzyme to >60 kGy (SI Appendix, Fig. S3) (2, 24, 25). Addition of DP1 to a Pi solution of Mn2+ at the ratio, DP1/Pi = 3 mM/25 mM = 1/8.33, as in MDP, strongly enhances the effectiveness of the Mn2+ as an antioxidant, protecting the enzyme to >60 kGy (SI Appendix, Fig. S3) (2, 24, 25). The obvious implication of the above is that although DP1 by itself binds weakly to Mn2+, DP1 and Pi jointly bind to Mn2+ to form a “ternary” Mn2+(Pi) (DP1) complex that is a superb antioxidant and that it is this complex that enables the MDP to protect solution enzyme far better than the Mn2+Pi complex alone. In the next section, we establish this to be the case. DP1 NMR in the Presence of Mn2+ and Pi Shows Formation of a Ternary Complex. The enhanced formation of complexed Mn2+ in the presence of DP1 and Pi vs that in Pi or DP1 alone (Fig. 3) supports this conclusion, but to definitively test, it we performed additional experiments in which the NMR spectrum of 2.2 mM DP1 was monitored: i) in the presence of increasing amounts of Pi; ii) with several concentrations of Mn2+; iii) upon independently varying concentrations of Mn2+ and Pi, Fig. 5. 600 MHz 1H NMR spectra. (A) 1H NMR region between 6.8 and 8.5 ppm of 3 mM DP1 (black) is initially titrated into 25 mM Pi (red), and the (DP1+Pi) complex is sequentially titrated into 250 μM and 500 μM Mn2+ (blue); (B) the accentuated 1H NMR region of His3-σ2 protons of 3 mM DP1(black) is into 500 μM Mn (pink), and the (DP1+Mn) complex is sequentially titrated into 7.6 mM, 22 mM, and 35 mM Pi (blue). Titration of DP1 with Pi to 25 mM differentially influences elements of the DP1 spectrum in a variety of ways, as illustrated in the behavior of the amide and histidine aromatic protons (∊1and σ-2) Fig. 5A. Thus, amide proton doublet at 7.75 and 7.95 are unaffected as Pi is added; the peaks between ~ 8.1 and 8.4 ppm broaden slightly without shifting; whereas, the sharp peak His3-σ2 proton at 7.05 ppm shifts without broadening, as is similarly observed for other binding residues including Asp1 β protons between 2.6 and 2.8 ppm, and Glu3 γ protons between 2.0 and 2.2 ppm (SI Appendix, Fig. S1). The similarly sharp peaks of His ∊1proton at 8.0 ppm likewise shift without broadening, but the shift is greater than His3-σ2 proton, allowing the observation that it approaches a “saturating” shift by ~25 mM Pi. Such observations indicate that addition of Pi to 25 mM causes essentially complete formation of a dynamic DP1/Pi complex. Subsequent addition of Mn2+ to 250 uM and 500 μM then progressively induces paramagnetic broadening of those peaks without further change in chemical shift, Fig. 5A. In complementary experiments, we monitored the 1H NMR spectra of DP1, which accentuates the His3-σ2 protons response as we added Mn2+ and Pi to DP1 in the “opposite” order. First adding substoichiometric (500 μM) of substitutionally labile Mn2+ to DP1 differentially induces paramagnetic broadening of all those resonances from the coordinating residues, without causing additional changes in the chemical shift of any Figs. 1 and 5B. In particular, as shown in Figs. 1 and 5B, this broadens into the inobservability the signal from His3-σ2. The subsequent addition of up to 35 mM Pi to a solution of DP1 plus 500 μM Mn2+ then narrows the peaks perturbed by the Mn2+ binding, in particular resurrecting the now broadened and shifted signal from His3-σ2 (Fig. 5B). This results in the same shifted, broadened peaks seen upon addition of Mn2+ to the Pi/DP1 complex, 1Figs. 1 and 5A and SI Appendix, Fig. S2: In short, the chemical shifts and breadth of all peaks at 25 mM Pi and 500 μM Mn2+ are essentially the same, independent of the order of addition, with integrated areas unchanged from those of DP1 itself. These observations jointly establish that Mn2+ in MDP solution indeed forms the proposed substitutionally labile “ternary” complex, Mn[DP1;Pi]. Binding Equilibria and Mn2+ Speciation in the Presence of Pi and DP1. As a means of understanding the Mn2+ speciation in the presence of both DP1 and Pi, we consider a binding model in which Mn2+ can bind a DP1 or a Pi with the measured affinity constants, KaD ≈ 0.04 mM−1 and KaP ≈ 0.4 mM−1, respectively (Fig. 3 A and B) but also can form the ternary complex with its own phenomenological affinity constants for binding the Pi and the DP1, or—equivalently—for binding the Pi/DP1 complex; the equations for the Mn2+ speciation within such a model are presented in SI Appendix. Fig. 6 presents the Mn2+ speciation calculated, for the titration of Mn2+ by a titrant that contains Pi and DP1 in the ratio, 25 mM/3 mM, the ratio in MDP and in the experiment of Fig. 3C. Total fraction of ligand-bound Mn and fractions individual species during the titration are calculated with the binding isotherms presented in SI Appendix. In addition to employing the measured Ka’s for Pi and DP1 given above, the computation uses an equilibrium constant for the formation of the ternary complex from Pi and DP1, KaT = 0.5 mM−2, chosen to model the enhanced total complex formation by Mn2+ during the “MDP titration” of Fig. 3C (KaL ≈ 0.7 mM−1). The isotherm that describes the titration with Pi alone is replotted for comparison. As shown in the figure, in the MDP solution itself, corresponding to 25 mM Pi on the graph, roughly 80% of the Mn2+ exists as the Mn2+(Pi/DP1) ternary complex. Plots of the Mn2+ species indicated during a titration in which [Pi]/[DP1] = 25/3, calculated as described in SI Appendix and plotted as a function of [Pi]. Exploring the Speciation of Mn2+ in Pi, DP1, and MDP Solutions by 2 K, 35 GHz EPR/ESEEM/ENDOR. We have made extensive use of 2 K, 35 GHz EPR, ENDOR, and ESEEM to characterize intracellular Mn2+ speciation (11, 14, 36, 37), and have applied these techniques (38) in the present study to characterize Mn2+ speciation in MDP. Here, we summarize the results presented in full in SI Appendix. i) When Mn2+ is frozen in Tris buffer with 3 mM DP1, 14N ESEEM shows that ~20% of the Mn2+ has bound 14N of the DP1 histidyl imidazole, versus ~5% in ambient-temperature fluid solution (see above). Thus, as noted above, the 2 K ESEEM and ambient 1H NMR agree on histidyl ligation by DP1, while the ESEEM unsurprisingly indicating that binding is exothermic and somewhat enhanced during freezing. ii) As reported, a Mn2+ ion frozen in 25 mM Pi solution binds a Pi, accompanied by loss of bound H2O, as shown by a strong 31P ENDOR signal accompanied by decreased 1H ENDOR intensity. iii) Mn2+ in the frozen MDP solution shows 31P and 1H ENDOR responses unchanged from those in Pi alone and thus also binds a Pi. Surprisingly, the 14N ESEEM shows only weak modulation from the histidyl imidazole of DP1, corresponding to at most a few percent occupancy, SI Appendix, Fig. S4, although Fig. 6 shows that a majority of the Mn2+ is present as the ternary complex. A plausible explanation for this finding is that the ternary complex does not form through independent binding by Mn2+ of Pi and a DP1 histidine and that DP1 binds differently in the presence of the bound Pi of the ternary complex, and indeed, it may well be that DP1 binds the interacting Pi/DP1 species that forms in solution (see above). In either case, the ternary complex does not involve extensive coordination of the histidyl imidazole (for further discussion, SI Appendix). Conclusions Unexpectedly, the ITC and EPR measurements show that the binding affinity of DP1 for Mn2+ is far too low to effectively form a radioprotective antioxidant complex with Mn2+ alone. 1H NMR measurements of DP1 during a titration with Mn2+ gives information about the residues that bind to Mn2+ but are not a useful means of measuring its affinity for Mn2+ because of strong paramagnetic broadening of the signals (39, 40). Overall, we show that: i) Pi and DP1 form a dynamic binary complex; ii) DP1 binds weakly to Mn2+; iii) DP1 and Pi jointly bind to Mn2+ to form a ternary Mn2+[Pi/DP1] complex with a greater thermodynamic binding constant than that of Pi alone; and iv) it is this complex that acts as the superb antioxidant that protects the DNA repair enzyme T4 DNA ligase up to a dose of >60 kGy, a far higher dose than the >10 kGy enabled by the Mn2+–Pi antioxidant complex alone (SI Appendix, Fig. S3). These results will help guide the optimization of whole-cell, radiation-inactivated vaccines against bacterial pathogens including Staphylococcus spp. (17), Acinetobacter spp. (19), Chlamydia spp. (23), and Borrelia spp. (22). The present approach to understanding ligand binding to Mn2+ binding will enable further optimization of the radioprotective properties of MDP-type agents, could unlock new strategies for enhancing cellular resistance to oxidative stress, and improve the efficacy of radioprotective agents, potentially leading to advances in medical treatments and biotechnology applications. Supplementary Material Appendix 01 (PDF)