Small-Molecule Mn Antioxidants in Caenorhabditis elegans and Deinococcus radiodurans Supplant MnSOD Enzymes during Aging and Irradiation

ABSTRACT

Denham Harman’s oxidative damage theory identifies superoxide (O₂^•−) radicals as central agents of aging and radiation injury, with Mn²⁺-dependent superoxide dismutase (MnSOD) as the principal O₂^•−-scavenger. However, in the radiation-resistant nematode Caenorhabditis elegans, the mitochondrial antioxidant enzyme MnSOD is dispensable for longevity, and in the model bacterium Deinococcus radiodurans, it is dispensable for radiation resistance. Many radiation-resistant organisms accumulate small-molecule Mn²⁺-antioxidant complexes well-known for their catalytic ability to scavenge O₂^•−, along with MnSOD, as exemplified by D. radiodurans. Here, we report experiments that relate the MnSOD and Mn-antioxidant content to aging and oxidative stress resistances and which indicate that C. elegans, like D. radiodurans, may rely on Mn-antioxidant complexes as the primary defense against reactive oxygen species (ROS). Wild-type and ΔMnSOD D. radiodurans and C. elegans were monitored for gamma radiation sensitivities over their life spans while gauging Mn²⁺-antioxidant content by electron paramagnetic resonance (EPR) spectroscopy, a powerful new approach to determining the in vivo Mn-antioxidant content of cells as they age. As with D. radiodurans, MnSOD is dispensable for radiation survivability in C. elegans, which hyperaccumulates Mn-antioxidants exceptionally protective of proteins. Unexpectedly, ΔMnSOD mutants of both the nematodes and bacteria exhibited increased gamma radiation survival compared to the wild-type. In contrast, the loss of MnSOD renders radiation-resistant bacteria sensitive to atmospheric oxygen during desiccation. Our results support the concept that the disparate responses to oxidative stress are explained by the accumulation of Mn-antioxidant complexes which protect, complement, and can even supplant MnSOD.

INTRODUCTION

In the current Denham Harman oxidative damage theory, antioxidant enzymes defend against reactive oxygen species (ROS) that contribute to aging and radiosensitivity, with Mn-dependent superoxide dismutase (MnSOD) being the central superoxide (O₂^•−) scavenger (1–3). Here, we report tests of the hypothesis that this theory must be broadened to incorporate an additional major defense mechanism, small-molecule Mn-antioxidant complexes (denoted H-Mn). To begin, MnSOD has been shown to be dispensable both for longevity and radiation resistance in several model prokaryotes, fungi, and simple animals (3–5). Building on these findings, we apply a new spectroscopic approach to measure cellular Mn antioxidant content and integrate our present findings with the previous studies. The result is a broadened picture of cellular defense against the oxidative stress caused by the diverse mechanisms of O₂^•− formation.

Focusing first on radiation resistance, the model bacterium Deinococcus radiodurans typically survives after 10 kGy of acute gamma radiation and grows luxuriantly under chronic ionizing radiation (60 Gy/h). However, MnSOD was shown to be dispensable for such radiation resistance in several prokaryotes, including D. radiodurans, and in some fungi (4–6).

There is a well-established linkage between radiation resistance and aging. It has been found that in all known cell types, the age at the time of radiation exposure is one of the principal factors controlling radiation-induced cell injury. Resistance typically increases until maturity but then declines through senescence until death (7–9). The short-lived nematode model organism Caenorhabditis elegans, which is extremely resistant to ionizing radiation, has been used to critically examine aging. Adults can survive 1,000 Gy with unaltered life spans (10). However, by manipulating expression of the C. elegans genes encoding its five SOD isoforms, it was found that the absence of MnSOD or CuZnSOD has little, if any, effect on the life span of C. elegans. As these results contradicted the view that SOD enzymes are the primary defense against ROS-induced aging, they led to the surprising conclusion that O₂^•− is not a major determinant of aging in this model organism (11).

However, cellular defenses against O₂^•− are more diverse than these considerations recognize. In desiccation- and radiation-resistant species of archaea, bacteria, and yeasts (12, 13) Mn is present not only as MnSOD, but also in a nonenzymatic form as low-molecular-weight H-Mn²⁺ complexes with accumulated metabolites (e.g., carboxylic acids, amino acids, peptides, nucleosides, orthophosphate) that exhibit catalytic ROS scavenging (5, 14–19) (Text S1). In particular, studies have revealed the exceptional efficacy of such small-molecule Mn antioxidants in protecting proteins against oxidative damage (20, 21). Indeed, rationally designed Deinococcus Mn²⁺-peptide antioxidants are now used as radioprotectants in the production of irradiated vaccines (22–25).

TEXT S1

What is the nature of Mn-peptide antioxidants? Text S1, DOCX file, 0.02 MB ^{(16.3KB, docx)}

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

As radiation and desiccation resistance and aging all are tied to the deleterious influences of O₂^•−, in this introduction we first briefly review aspects of the role of Mn in O₂^•− defense, both Mn in the enzymatic form as MnSOD and as small-molecule Mn-antioxidant complexes. We then extend this to the linkage between radiation resistance, desiccation resistance, and aging, and the consequences of the presence or absence of MnSOD and Mn antioxidants, which is a focus of the analytical approach described in this report.

Mn and resistance to superoxide. Mn²⁺ ions are unique among redox-active transition metals that accumulate in cells in that they are innocuous under conditions where other biologically active transition metals (in particular, Fe²⁺) tend to promote formation of ROS (26). Therefore, cells and their organelles can accumulate up to millimolar concentrations of Mn²⁺ ions in the form of Mn²⁺-metabolite complexes (Mn antioxidants), along with MnSOD (13, 16, 27). Both MnSOD and Mn antioxidants catalytically scavenge O₂^•− formed through metabolic processes or by radiation exposure, converting these ions to hydrogen peroxide (H₂O₂) (20, 28, 29) (Fig. 1). H₂O₂ is membrane-permeable and can escape the cell (20); in contrast, O₂^•− is not membrane-permeable and becomes trapped and accumulates in cells (30). Without these Mn-mediated means of eliminating O₂^•−, the O₂^•−radicals can directly react with certain amino acids and Fe-S groups of proteins or can further react to form hydroxyl (HO^•) radicals that cause indiscriminate proteome and genome damage, which in turn lead to disruption of protein homeostasis and DNA repair malfunction (20, 21, 31–34) (Fig. 1).

FIG 1.

Ionizing radiation-driven reactions related to dismutation of superoxide by MnSOD and Mn antioxidants, illustrating the two complementary catalysts of superoxide defense, MnSOD and Mn antioxidants (H-Mn).

The importance of Mn antioxidants extends beyond radiation survival. Deinococci are nonsporulating obligate aerobes and are further renowned for their desiccation resistance (12), a Mn-dependent metabolic trait associated with proteome protection under atmospheric O₂ (35). Whereas O₂^•− in desiccating cells is derived mostly from atmospheric O₂, O₂^•− in irradiated cells is also formed by radiolytic reduction of O₂ released through the intracellular decomposition of ionizing radiation-generated H₂O₂ as catalyzed by enzymes (e.g., catalase) (Fig. 1) (20, 26). Mounting evidence supports the hypothesis that ionizing radiation resistance is simply a secondary phenotype that accompanies resistance to other more commonly encountered environmental stressors, particularly the hyperoxic conditions during cycles of exposure to atmospheric O₂ during desiccation and subsequent rehydration (12, 35).

The present study. Both D. radiodurans and C. elegans are known for their extreme radiation resistance. However, MnSOD is found to be dispensable for radiation protection, a finding that contradicts the view of cellular defense against oxidative stress that implicates MnSOD as the dominant protection against ROS. Thus, these organisms are models for study to identify the roles of Mn antioxidants and MnSOD in protection against various sources of oxidative stress.

In this report, we tracked the gamma radiation survival of D. radiodurans bacteria and C. elegans nematodes throughout their life span. In this effort, we controlled MnSOD content by mutation while probing Mn²⁺ speciation and coordination of Mn antioxidants by electron paramagnetic resonance (EPR) spectroscopy. Prior to the use of EPR, there was no means of determining Mn²⁺ antioxidant content in vivo. Standard analytical procedures that disrupt cells alter the chemical speciation of Mn²⁺, thus destroying information about the in vivo properties. In contrast, EPR methodologies, namely, EPR, ENDOR, and ESEEM spectroscopies, directly characterize the speciation of Mn²⁺ in vivo. We previously used EPR to show that radiation-resistant cell-types possess a high population of Mn antioxidant complexes and are readily distinguishable from radiation-sensitive cell types that lack Mn antioxidants (5, 36). We showed that radiation resistance correlates directly with cellular concentrations of Mn antioxidants (5, 15, 20, 26) and extended this here to include the model bacterium Lactobacillus plantarum, which does not encode SODs but hyperaccumulates Mn antioxidants (37–39).

This study enables us to consider paradoxes and puzzles in Denham Harman’s oxidative damage theory viewed from the perspective of three extremely radiation-resistant model organisms—C. elegans, D. radiodurans, and L. plantarum. We evaluated the variations in MnSOD and H-Mn antioxidant content with aging and oxidative stress resistances, and formulate an extended theory of oxidative stress defense in which small-molecule Mn antioxidants are complementary to MnSOD, and in some cases can supplant MnSOD, in O₂^•−-scavenging.

RESULTS

We begin by showing (in the following subsection) that cellular Mn antioxidant content in D. radiodurans can be manipulated by aging and is independent of the presence of MnSOD. Next, it is shown that the same is true for C. elegans. However, it is finally shown (in the section “D. radiodurans and L. plantarum become desiccation-sensitive upon loss of MnSOD”) that MnSOD is nonetheless required for defense against the oxidative stress of desiccation in bacteria, as it is for defense against chemical agents such as paraquat.

Mn antioxidant content and gamma radiation resistance decrease in unison with age in D. radiodurans, independent of the presence of MnSOD.

The chronological life span of D. radiodurans is measured here as the time a cell survives in a nonreplicating (stationary) state, with survival being defined as the ability to form a colony on fresh medium. The effect of aging on wild-type (WT) and ΔMnSOD cultures is characterized by measuring the decrease in radiation survival with time of the stationary-phase cultures (Fig. 2A and B), and in cellular Mn antioxidant content as determined by EPR (Fig. 2C and D). The experiments began by inoculating mid-exponential-phase WT (ATCC BAA-816) and sodA⁻ (ΔMnSOD) cells into separate Erlenmeyer flasks with 0.8 L liquid TGY (1% bactotryptone, 0.5% yeast extract, and 0.1% glucose). These cultures were followed during incubation with agitation at 32°C for up to 10 days without change of broth (Fig. 2A, B and F; see also replicate study Fig. S1A, D, and E).

FIG 2.

Gamma radiation survival and H-Mn content over the life span of D. radiodurans. (A) WT survival (CFU/mL) of young (day 1 [d1]), mature (d2), old (d6), and dying (d10) D. radiodurans (DR) cells exposed to 0, 10, and 15 kGy (⁶⁰Co) (replicate, Fig. S1D). (B) ΔMnSOD (sodA⁻) survival as in panel A (replicate, Fig. S1E). (C) Normalized EPR spectra of WT cells of panel A showing a relative increase with aging in the intensity of the broad wings, associated with L-Mn relative to the normalized intensity of the narrow response from H-Mn. (Inset) Partition of cellular Mn²⁺ into two pools, H-Mn (Mn antioxidants) and L-Mn (Mn-proteins) for day-10 cells (see Fig. S2, which shows the partition of cellular Mn²⁺ into H-Mn/L-Mn for all cell-types). (D) EPR spectra for sodA⁻ cells of panel B with aging. (Inset) Partition of cellular Mn²⁺ into H-Mn and L-Mn for day-10 cells. (E) Changes in fractions of cellular H-Mn and L-Mn complexes in WT and sodA⁻ with bacterial aging, as deduced from decompositions of EPR spectra (see Fig. S2). For details of the procedures for quantifying H-Mn and L-Mn contributions, see reference 5. (F) ΔMnSOD-induced growth lag in cultures (replicate, Fig. S1A).

FIG S1

Growth, H-Mn content, and gamma radiation survival over the lifespan of D. radiodurans (WT), and ΔMnSOD (sodA⁻). Download FIG S1, DOCX file, 0.3 MB ^{(267.3KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

FIG S2

Fraction of H-Mn and L-Mn and their contribution to Mn²⁺ cellular EPR spectra of D. radiodurans (Dr WT) and ΔMnSOD (Dr sodA⁻). Download FIG S2, DOCX file, 0.3 MB ^{(272.5KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

Over the 10-day study, we measured cell survival and gamma radiation resistance during aging, which are measures of replicative fitness, by use of the CFU assay. We show that young and mature WT and sodA⁻ cells are similarly resistant to 10 kGy and that they become progressively less resistant as they grow older (Fig. 2A and B). Unexpectedly, the sodA⁻ strain displayed even greater radiation resistance and longevity compared to the WT (Fig. 2B; see also Fig. S1D and E). This can be explained by delayed growth of the sodA⁻ mutant (Fig. 2F; see also Fig. S1A), which would conserve metabolites needed for radiation survival (40); however, other explanations are equally plausible (see Discussion).

In parallel, we employed EPR spectroscopy to measure the content of Mn antioxidants, which is a ruler of cellular radiation resistance (5). We collected 35-GHz continuous-wave (CW) absorption-display EPR spectra for WT and sodA⁻ D. radiodurans cells harvested at days 1, 2, 6, and 10 (Fig. 2C and D; see also Fig. S1B and C). We had previously shown that such absorption-display EPR spectra at X- or Q-bands, but not derivative-display spectra, reveal that cellular Mn²⁺ exists as two distinct pools of Mn²⁺ complexes, denoted H-Mn and L-Mn (5, 41). The H-Mn spectrum is centered around an ∼12-kG magnetic field and is dominated by a relatively narrow Mn²⁺ signal (∼1.5 kG half-max width) that displays a sextet pattern arising from hyperfine interactions with the ⁵⁵Mn (nuclear spin, I = 5/2) nucleus (hyperfine coupling, A ∼90 G) (5, 36, 41) and arises from the high-symmetry, cellular small-molecule Mn antioxidant complexes (H-Mn, for example; these include Mn²⁺ complexed with orthophosphate [Pi] and peptides). The H-Mn spectrum may “ride on” and be flanked by broad “wings” extending from low fields, ∼3 kG, to higher fields well beyond the magnet limit. These features are associated with a heterogeneous population of low-symmetry Mn²⁺-bound complexes and enzymes denoted L-Mn (41). The fractional contributions from H-Mn and L-Mn populations to the Mn²⁺ EPR spectra are shown in Fig. 2E, as calculated using the protocol described previously (5).

Comparison of the EPR spectra of young, mature, and old WT and ΔMnSOD D. radiodurans cells showed that life stage and radiation resistance track with the population of H-Mn complexes, but not with MnSOD content (Fig. 2C and D). Specifically, young (day 1 [d1]) D. radiodurans cells contain almost exclusively Mn²⁺ in antioxidant H-Mn complexes (Fig. 2E), characteristic of high radiation resistance. As cells senesce (d2 to d10), their EPR spectra progressively display increasingly intense wings derived from L-Mn complexes characteristic of radiation-sensitive organisms (5) (Fig. 2C and D; see also Fig. S1B and C), as the fraction of H-Mn decreases and that of L-Mn increases (Fig. 2E). Correspondingly, the radiation resistance decreases track with the fraction of H-Mn (Fig. 2A and B). Of particular importance, as noted above, radiation resistance is not merely undiminished by elimination of MnSOD in the sodA⁻ strain, but in fact, the resistance of the strain is somewhat greater than that of WT during aging.

As gauged by EPR, Mn-antioxidant content decreases in unison with age in C. elegans, independent of the presence of MnSOD.

Similar to D. radiodurans, the life span of C. elegans is reported to be unaffected or even somewhat increased in ΔMnSOD mutants (11, 42, 43) (Fig. 2A and B). As we now show in Fig. 3, C. elegans nematodes also display the two important molecular hallmarks seen for radiation-resistant microorganisms: (i) an extremely high H-Mn content and near-absence of L-Mn as measured by Mn²⁺ EPR, in contrast to radiation-sensitive organisms (5) (Fig. 3A), and (ii) as correspondingly measured analytically, a high cellular concentration of small-molecule O₂^•−-scavenging Mn antioxidants that protect proteins (15, 20, 22–24, 44, 45) (Fig. 3F and G).

FIG 3.

Mn antioxidant content of C. elegans and their small-molecule extracts. (A) Normalized EPR spectra of radiation-resistant young (d2) WT (N2) and ΔSOD mutant nematodes (CE) (see inset), dominated by H-Mn, plus the radiation-resistant L. plantarum (LP), compared to the two naturally radiosensitive species, E. coli K-12 (EC) and S. cerevisiae (SC) EXF-6218, which show the strong wings associated with L-Mn in young (d1) cells. (B) Normalized EPR spectra of WT nematodes across their life span; d1 (young), d5 (mature), d13 (mature), d26 (old, sluggish, and nonresponsive), and d33 (dead) displaying a relative increase in intensity of the broad wings, associated with L-Mn compared to the intensity of the narrow response from H-Mn with aging. (C) EPR spectra of WT nematodes (d2, d7, d18). (Inset) Partition of cellular Mn²⁺ into H-Mn and L-Mn pools. (D) EPR spectra of sod-2,3 (ΔMnSOD) nematodes (d2, d7, d18). (Inset) Partition of cellular Mn²⁺ into H-Mn and L-Mn. (E) Changes in fractions of cellular H-Mn and L-Mn complexes in WT nematodes with aging, as deduced from decompositions of EPR spectra (panel B). For details of the procedures for quantifying H-Mn and L-Mn contributions, see reference 5. (F) CEU (C. elegans ultrafiltrate) of WT (d2) nematodes protects irradiated proteins. CEU was mixed with purified E. coli proteins and exposed to the indicated doses of gamma rays (kGy). Irradiated proteins were then separated by polyacrylamide gel electrophoresis, and protein oxidation was visualized by Western blotting carbonyl analysis (Oxyblot), which reveals the presence (black) or absence (no signal) of protein oxidation. CEU was compared to D. radiodurans ultrafiltrate (DRU) of WT (d2) bacteria. Noncontiguous portions of the same blot are separated by white lines. (G) CEU of WT (d2) nematodes preserves the activity of an irradiated enzyme. BamHI was irradiated in potassium phosphate buffer (PiB) or in CEU or DRU and then incubated with lambda phage DNA and subjected to agarose gel electrophoresis. Noncontiguous portions of the same agarose gel are separated by white lines.

In relation to item (i), above, Fig. 3A shows the absorption-display EPR spectra of young (d2) C. elegans WT (N2) and sod deletion mutants (inset); sod-1,2,3,4,5 (ΔSOD) is a quintuple mutant with complete loss of MnSOD and Cu/ZnSOD activities; sod-2,3 is a double mutant deficient in MnSOD enzymes (ΔMnSOD); sod-2 and sod-3 are single ΔMnSOD mutants (43, 46). Included for comparison are the EPR spectra of the radiation-resistant bacterium L. plantarum (5, 37, 38, 47) and two radiation-sensitive species, the bacterium Escherichia coli K-12 and the yeast Saccharomyces cerevisiae EXF-6218 (5). The EPR spectra of all the young nematodes (WT and the four sod knockout mutants; see inset of Fig. 3A) show that the Mn²⁺ is almost exclusively present as H-Mn, with a negligible contribution from L-Mn, and likewise for L. plantarum. This contrasts with the EPR spectra of radiation-sensitive E. coli K-12 and S. cerevisiae EXF-6218 (mature, healthy cells), which are dominated by a large pool of L-Mn that generates the intense wings of the spectra, with only a small pool of H-Mn.

The first important result from Fig. 3 is that EPR shows that C. elegans hyperaccumulates H-Mn, with H-Mn content undiminished for all SOD knockout strains compared to WT (Fig. 3A). We next compared aging WT and ΔMnSOD nematode populations and focused on their Mn antioxidant content before and after the onset of senescence, as characterized by increasingly sluggish, nonmotile, and unresponsive worm behaviors, which are measures of metabolic fitness. The nematodes were monitored for survival over 3 to 4 weeks, displaying the following stages: young worms (d1 to d3), mature (middle-aged) worms (d5 to d13), old worms (d15 to d21), and dead worms (>d26). Analogous to aging D. radiodurans bacteria (Fig. 2C), the normalized EPR spectra for WT nematodes exhibited an age-dependent loss of H-Mn antioxidant content over the life span (Fig. 3B); EPR spectra show that young and mature worms (inset) contain only H-Mn, whereas H-Mn decreases as L-Mn increases in older worms (Fig. 3E). As a follow-up to this primary study (Fig. 3B), we grew WT and ΔMnSOD in parallel and harvested worms at d2 (young), d7 (mature), and d18 (old), and collected their EPR spectra (Fig. 3C and D). Consistent with D. radiodurans, the young and mature worms displayed very high H-Mn antioxidant levels, with a loss in H-Mn antioxidant content and an increase in L-Mn in the older worms (Fig. 3C and D; see also Fig. S3). Moreover, at the various stages of aging, the H-Mn population is unchanged in ΔMnSOD C. elegans compared to WT (Fig. 3C and D; see also Fig. S3), as is the case with D. radiodurans.

FIG S3

Changes in fractions of cellular H-Mn and L-Mn complexes in WT and sod-2,3 C. elegans with aging. Download FIG S3, DOCX file, 0.3 MB ^{(324.5KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

In relation to item (ii), above, the small-molecule antioxidant capacity of C. elegans was directly confirmed using aqueous-phase extracts of WT (d2) homogenates, which were first subjected to ultracentrifugation and then to ultrafiltration. Ultracentrifugation removed proteins and other macromolecules and peptides greater than 1 kDa. As a precaution, the ultracentrifuged supernatants were also ultrafiltrated to remove any contaminating molecules (>3 kDa) released from the pellets during collection of the supernatants.

When WT C. elegans ultrafiltrates were mixed with proteins purified from E. coli and exposed to gamma radiation, high levels of protein protection were indicated by Western blotting carbonyl analysis (Fig. 3F); carbonyl groups are the most widely used marker of irreversible protein oxidation (15, 20, 31, 48–51). Side-by-side comparisons showed that C. elegans ultrafiltrate (CEU) was even more radioprotective than D. radiodurans ultrafiltrate (DRU), which is highly enriched in small-molecule Mn complexes (15). At 4,000 Gy of gamma rays, for proteins mixed with PiB alone, proteins were heavily oxidized (carbonylated) as visualized through blacking in the Western blots. In stark contrast, when the E. coli proteins were irradiated in the presence of CEU, there was no significant protein carbonylation (Fig. 3F).

We also tested the ability of CEU to protect the activity of the restriction endonuclease BamHI, a radiation-sensitive enzyme inactivated in vitro by ∼150 Gy (15). Whereas diluted DRU protected the in vitro activity of BamHI to 4,000 Gy, an equivalent amount of CEU protected the enzyme to 8,000 Gy (Fig. 3G). Thus, C. elegans cells have an even greater antioxidant capacity than the most radioresistant bacteria, D. radiodurans (20).

ΔSOD (sod-1,2,3,4,5) C. elegans outlives WT under intense gamma irradiation. Longevity in this study represents the ability of nematodes to maintain healthy (active, motile, responsive) behaviors under high-level chronic radiation, in physiological conditions where nematodes are metabolically active but cannot proliferate, as fecundity is the earliest major function lost to irradiation. Reproductive sterilization is reported at doses as low as 500 Gy (10), and consistently, we did not see proliferation of nematodes growing under chronic irradiation. Taking the major endpoint as survival of individual animals, we studied the effect on the activity and life span of groups of 40 adult nematodes on strain OP50 plates incubated under 34.75 Gy/h over 8 days (Fig. 4A).

FIG 4.

Differential effects of MnSOD on radiation and desiccation tolerance. (A) Comparison of Kaplan-Meier survival curves for irradiated WT (N2) (red) and ΔSOD (sod-1,2,3,4,5) (blue) C. elegans strains. Ten gravid adult worms per group were inoculated on plates (NGP) with OP50 lawns (4 plates/group) and monitored for survival and scored for activity over 8 days under 34.75 Gy/h (¹³⁷Cs). Shaded regions represent 95% Cis; the P value based on regression analysis described in the main text was 0.00101. This result was similar in two independent experiments. (B) Desiccation resistance of WT and ΔMnSOD D. radiodurans and WT L. plantarum. Cells were grown to the early stationary phase, desiccated at room temperature in a hermetically sealed chamber containing Drierite (CaSO₄), and tested for viability by CFU analysis after the indicated number of weeks. CFU time points were in triplicate; standard deviations are shown.

The Kaplan-Meier survival curves for chronically irradiated WT and ΔSOD C. elegans strains are shown (Fig. 4A; see also Table S1). Differences in survival between these two strains show that the ΔSOD nematodes actually exhibited increased radiation fitness and a longer life span compared to the WT growing under intense gamma radiation. The results of statistical analyses of the data used a parametric survival regression model based on the Weibull distribution (Table S2). This analysis implies that the absence of SOD (coded by the SOD variable) in fact significantly reduced the radiation death hazard (P value = 0.00101). We note that the ΔSOD worms had better radiation survival than the wild-type, yet the initial levels of H-Mn in the C. elegans strains were the same (Fig. 3A). While this may suggest that something other than H-Mn is responsible, the distribution of Mn antioxidants in C. elegans cells is expected to change with the loss of MnSOD enzymes in mitochondria (see Discussion).

TABLE S1

Survival of WT (+) SOD and ΔSOD (sod-1,2,3,4,5) (–) nematode groups under chronic gamma irradiation (34.75 Gy/h). Download Table S1, DOCX file, 0.01 MB ^{(14.6KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

TABLE S2

Best-fit parameter values for the survival analysis of irradiated C. elegans strains. Download Table S2, DOCX file, 0.01 MB ^{(13.4KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

D. radiodurans and L. plantarum become desiccation-sensitive upon loss of MnSOD.

Both desiccation-induced hyperoxic conditions and ionizing radiation produce O₂^•−, whose damage to cell functions can result in mutations and death. For example, the aerobic dehydration of D. radiodurans for 6 weeks causes approximately 60 DNA double-strand breaks (DSBs) per genome, the same as caused by exposure to 6 kGy of gamma rays (12). Although such findings appear to support the idea that the etiologic ROS underlying ionizing radiation and desiccation toxicity are the same, it is of significance that the primary location of O₂^•− generation in cells is different in the two cases. In drying cells, O₂^•− is the product of direct contact with atmospheric O₂, and thus O₂^•− forms near the cell surface; in irradiated cells, O₂^•−, is derived mostly from H₂O₂ produced by water radiolysis and is formed throughout the cell (26). We thus tested whether MnSOD, which is dispensable in D. radiodurans resistance to radiation-induced ROS, also is dispensable for resistance to desiccation, which is a more commonly encountered environmental oxidative stressor. We further complemented this by examining the desiccation resistance of L. plantarum, a radiation-resistant bacterium (4, 47) (Fig. S4) that also accumulates Mn antioxidants but does not encode enzymatic SODs (4, 37, 38).

FIG S4

Gamma radiation survival curves for WT L. plantarum, which is naturally deficient in SOD enzymes. Download FIG S4, DOCX file, 0.06 MB ^{(65.9KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

We report the first demonstration that the sodA⁻ D. radiodurans mutant is rendered desiccation-sensitive. Fig. 4B shows the desiccation survival curves for WT and sodA⁻ D. radiodurans over 15 weeks under ambient atmospheric conditions. The sodA⁻ mutant is in fact far more sensitive to desiccation than the WT; it survived 1,000× less well after desiccation for 10 weeks. Correspondingly, L. plantarum, which is devoid of SODs, is even more desiccation-sensitive (Fig. 4B). Thus, Mn antioxidants do not substitute for MnSOD in slowly desiccating bacterial cells under atmospheric conditions, namely, 21% O₂.

Finally, we confirmed by phenotypic analyses that the sodA⁻ D. radiodurans mutant remains resistant to chronic gamma irradiation and ultraviolet C (UVC) irradiation (Fig. 5A and B). However, the mutant is rendered hypersensitive to paraquat (methyl viologen) (Fig. 5C), a powerful redox-active O₂^•−-generating herbicide (52).

FIG 5.

Oxidative stress resistance phenotypes of D. radiodurans (DR) WT, sodA⁻, and katA⁻ (catalase-deficient) under aerobic growth conditions. (A) Chronic gamma irradiation (CIR) (¹³⁷Cs). Proliferation of bacteria under 57 Gy/h for 6 days. In this panel, the dilutions shown are on a log₁₀ scale and represent order of magnitude changes in initial cell concentration. (B) UVC irradiation (254 nm). Clonogenic survival after an acute UVC dose of 1 kJ/m². Dilutions are as in panel A. (C) Paraquat. Hypersensitivity of sodA⁻ colony biofilm growth to paraquat applied to the central disc, which contained 20 μL 12.5 mM methyl viologen dichloride.

DISCUSSION

The spectroscopic measure of cellular Mn antioxidant (H-Mn) content is currently the strongest known biological indicator of cellular radiation resistance (5). This remarkable association was first demonstrated in archaea, bacteria, and yeasts grown to the early stationary phase when cells are the most resistant (5). However, it was unknown whether or not the H-Mn content of a given organism is fixed or varies over its life span; neither was it known if EPR is sufficiently sensitive to detect life stage-dependent changes in radioresistance. We experimentally manipulated the cellular H-Mn content of cells by aging; EPR spectra collected over a 10-day aging study of D. radiodurans correspondingly show progressive loss of H-Mn antioxidants as the medium becomes depleted (Fig. 2C and D), as is quantified in Fig. 2E, and this tracks with the loss in radiation resistance (Fig. 2A and B). This result is similar to H-Mn responses observed over the life span of C. elegans nematodes (Fig. 3).

As small-molecule H-Mn antioxidants globally protect the proteome of cells (20, 21), this most likely includes protection of MnSOD; the radiation-stable H-Mn “defender” protects the radiation-sensitive one (26). The correlation of age with decrease in H-Mn content might be thought to suggest that as the cells age and the pool of metabolites that bind to Mn²⁺ to form H-Mn antioxidants gets consumed, the cells eventually rely solely on enzymatic SOD defenses. This view would then predict that old cells of ΔMnSOD D. radiodurans and C. elegans are more radiation-sensitive than WT. However, our results contradict this idea and even appear to pose a further puzzle; we find that the ΔMnSOD D. radiodurans mutant lacking any SOD activity (52) consistently displayed not merely undiminished, but indeed modestly increased radiation fitness (Fig. 2A and B; see also Fig. S1D and E). Caenorhabditis elegans, known to be highly radiation-resistant, had been viewed as presenting this same puzzle, in that mutants devoid of SODs have undiminished life spans compared to WT and, as we now show, even under intense gamma irradiation (Fig. 4A).

In discussing the correlation of MnSOD and Mn antioxidant content to aging and oxidative stress resistances, we integrate them with those of previous studies, which leads us to a broadened picture of cellular defense against oxidative stress. EPR spectroscopy has resolved the apparent puzzle presented by these two organisms by showing that this MnSOD-independent resistance in both is associated with the hyperaccumulation of radioprotective H-Mn (Fig. 2C and D; Fig. 3C and D), as confirmed analytically (Fig. 3F and G). Our data thus show that radiation fitness and life span can be decoupled from enzymatic-MnSOD defenses in D. radiodurans and C. elegans; our findings imply that H-Mn antioxidants supplant at least some enzymatic SOD functions in Deinococcus bacteria as well as in C. elegans mitochondria, as visualized in the scheme of Fig. 6.

FIG 6.

Role of H-Mn antioxidants and MnSOD enzymes in radiation- and desiccation-resistant organisms. (A) D. radiodurans; radiation and aging (left) (Fig. 5A). In irradiated cells, O₂^•− is derived mostly from H₂O₂ produced by water radiolysis and is formed throughout the cytoplasm. O₂^•− is membrane impermeable and becomes trapped in cells. Desiccation (right) (Fig. 4B). During desiccation, O₂^•− is the product of metabolism in cells in direct contact with atmospheric O₂, and thus O₂^•− forms near the cell surface. H-Mn antioxidants accumulate in young cells and catalyze the same O₂^•− dismutation reaction as MnSOD, yielding H₂O₂. Catalase (KatA) is periplasmic. (B) C. elegans; aging (Fig. 3B to D and Fig. 4A). The essential O₂^•−-scavenging features of H-Mn antioxidants and MnSOD in mitochondria are the same as for D. radiodurans, except that mitochondria release H₂O₂ into the cytoplasm; MnSOD is resident in the matrix and therefore distal to cell surfaces most exposed to O₂, and loss of MnSOD adds Mn²⁺ ions to the pool of Mn antioxidants in the matrix.

Of course, the enzyme MnSOD does play a critical role in O₂^•− scavenging and resistance to extreme oxidative stress other than radiation insult in these two organisms. For example, MnSOD deletion mutants render D. radiodurans and C. elegans hypersensitive to paraquat (43, 46, 52–54) (Fig. 5C), and we show here that ΔMnSOD D. radiodurans is desiccation-sensitive (Fig. 4B). This hitherto unknown contrast between the importance of MnSOD in protection against the hyperoxic conditions of desiccation (and paraquat), and yet the absence of deleterious effects on radiation resistance and aging from the elimination of MnSOD in ΔMnSOD mutants, implies that Mn antioxidants and SOD enzymes serve different roles in antioxidant defense (Fig. 6A).

This conclusion is reinforced by considering the bacterium L. plantarum, which is unusual in that it does not encode enzymatic O₂^•− scavengers, but hyperaccumulates Mn antioxidants (Fig. 3A) and, as a result, is radiation-resistant (4, 5, 13, 37, 38, 47) (Fig. S4). And again, with L. plantarum, as with D. radiodurans, the absence of an enzymatic O₂^•− defense accompanies sensitivity to desiccation (Fig. 4B). The need for MnSOD during desiccation of both bacteria studied here suggests a fundamental role of this enzyme is to defend nutrient-restricted or starving cells that are directly exposed to the high atmospheric levels of O₂, a common combination (12, 44) (Fig. 6A). The distribution of SOD enzymes in a cell also shapes our considerations. For example, MnSOD is resident in the cytoplasm and associated with membranes of Deinococcus cells (55, 56), whereas in C. elegans cells, MnSOD is localized within the mitochondrial matrix; it is Cu/ZnSOD that is resident in the cytoplasm and implicated in desiccation-survival of dauer larvae, a developmentally arrested form induced by starvation (57–60).

An extended theory of oxidative stress. The observation that ΔMnSOD strains of D. radiodurans and C. elegans show both an undiminished life span and undiminished radiation resistance contradicts the current oxidative damage theory, which identifies MnSOD as the preeminent defense against the O₂^•− generated during irradiation as well as aging. Our results show that this paradox is removed if the defense against oxidative stress is extended to include H-Mn as complementary to MnSOD (Text S2).

TEXT S2

Previously published Mn metrics for D. radiodurans. Download Text S2, DOCX file, 0.01 MB ^{(14.4KB, docx)} .

This is a work of the U.S. Government and is not subject to copyright protection in the United States. Foreign copyrights may apply.

The extended oxidative stress theory clearly explains the absence of deleterious effects on D. radiodurans radiation resistance caused by the absence of MnSOD in young and mature ΔMnSOD cells. We attribute the puzzling modest increase in life span and radiation survival of aging ΔMnSOD bacteria to the lag in growth (Fig. 2F). This growth delay would certainly conserve nutrients that facilitate radiation survival and longevity (40). For C. elegans, mutants lacking SODs also have undiminished life spans compared to those of the WT (42, 43), even under intense gamma irradiation (Fig. 4A). Once again, this can be explained through protection by Mn antioxidants. We note that mitochondria are reported to accumulate 10 to 20 mmol/L Mn²⁺, which is 10 to 20 times more Mn than accumulated overall in D. radiodurans and in the range of Mn concentrations in L. plantarum (20, 37, 38, 61, 62). In both L. plantarum and C. elegans, EPR shows that, essentially, all this Mn²⁺ is present in the radioprotective H-Mn form (Fig. 3A), which shows that both cytosolic Mn²⁺ and Mn²⁺ in mitochondria in fact all exist as H-Mn, explaining the similar responses of the nematodes and the bacteria (Fig. 6). In L. plantarum bacteria, the accumulated Mn²⁺ antioxidants (H-Mn) bind carboxylic acids (e.g., lactate) to form Mn antioxidants (27) (Fig. 3A).

Our finding that Mn antioxidants hyperaccumulated in C. elegans are exceptionally protective of proteins (Fig. 3F and G) and that H-Mn content drops precipitously in the last third of the life span (Fig. 3E) together explain reports that the steady-state level of oxidized proteins increases exponentially during the last third of the life span in animals (31). However, a more detailed understanding of why the loss of MnSOD in C. elegans actually appears to facilitate survival under intense irradiation (Fig. 4A) would benefit from a determination of the metrics of Mn distribution and the composition of the pool of Mn²⁺-binding metabolites in its mitochondria. In our model, MnSOD competes with metabolites for Mn²⁺, and the loss of Mn-enzymes increases the cellular budget of free Mn²⁺, which leads to enhanced radiation survival in the worms (Fig. 6B, ΔMnSOD). The model may even explain reports by others that the presence of MnSOD can actually amplify radiosensitivity in bacteria (63). However, until there is a genetic route to control Mn antioxidant content, we cannot rule out the possibility that an as yet unidentified factor may be the causal agent governing age-dependent decreases in radiation resistance.

If H-Mn is indeed complementing MnSOD, as proposed here, then Mn²⁺ ion supplementation alone should facilitate oxidative survival, and this in fact has been observed in several studies; Mn²⁺ added to D. radiodurans and E. coli facilitates radiation survival (4, 13, 15), and Mn²⁺ addition to a short-lived, paraquat-sensitive C. elegans mutant (mev-1) accelerates development, enhances stress resistance, and rescues the life span (64). Likewise, supplementation of mice with MnCl₂ alone markedly increases ionizing radiation survival (65). Collectively, these observations can now be understood as resulting from the increased amounts of cellular Mn²⁺ in the form of radioprotective H-Mn complexes (Fig. 6A and B).

To summarize, the present results for WT and ΔMnSOD of D. radiodurans and C. elegans, and for the radioresistant L. plantarum, which does not express SODs, show that radioresistance is not supported by MnSOD. This suggests that Mn antioxidants have comparable or greater efficacy in this role, with wider cellular distribution than MnSOD. Finally, that H-Mn cannot protect cells from atmospheric O₂ during desiccation (Fig. 4B) is implicit in the nature of this stress; starving cells cannot accumulate the metabolites that coordinate to Mn²⁺ to form Mn antioxidants, which leaves MnSOD as the only defense against elevated O₂^•− in cells that are hyperoxic because they are exposed directly to the atmosphere (Fig. 6A). The extended oxidative stress defense theory further explains the radiosensitivity of the E. coli K-12 and S. cerevisiae EXF-6218 (5) (Fig. 3A). In those cells, radioprotective H-Mn is only a small minority of the nonenzymatic Mn²⁺ and thus can do little to defend against a sudden buildup of cytoplasmic O₂^•−, whereas in D. radiodurans cells, antioxidant H-Mn is the major form of Mn²⁺ and thus the major defense against radiation.

Looking forward, EPR spectroscopic methods now offer a novel path to bridging the gap between in vitro and in vivo approaches to further testing and, as needed, further improving the modified oxidative stress defense mechanism. One might further consider an even more dramatic possible application for the use of EPR to monitor Mn²⁺ speciation in vivo. Viewing C. elegans nematodes (Fig. 3) as surrogates for human tissues, the present results suggest a possible use of EPR to examine cancer biopsies samples. An EPR determination of the amount of antioxidant Mn in a biopsy sample, and thus the radiation resistance of an individual tumor, might well assist in optimizing a radiation/chemotherapeutic regime for specifically attacking that tumor (66).

Conclusions. We experimentally manipulated the cellular H-Mn antioxidant content by aging viable cultures of D. radiodurans (Fig. 2A to E) and C. elegans (Fig. 3 and 4A), and showed that MnSOD is expendable for the radiation survivability and longevity of the organisms. However, the well established role of MnSOD in defense against oxidative stress is nonetheless exemplified by the fact that its absence makes the D. radiodurans ΔMnSOD mutant sensitive to the exposure to atmospheric O₂ attendant to desiccation and to the O₂^•−-generating paraquat, as well as delaying growth compared to the WT (Fig. 2F, 4B, and 5C). These complementary roles of H-Mn antioxidants and MnSOD against various forms of oxidative stress are also displayed by L. plantarum, which does not encode SODs but hyperaccumulates H-Mn antioxidants and thus is strongly radiation-resistant but sensitive to desiccation (Fig. 4B; see also Fig. S4). Such findings require extension of Denham Harman’s theory to include accumulated H-Mn antioxidants as partners and complements to MnSOD (1, 2, 5).

The extended theory of oxidative stress defense posits that MnSOD and H-Mn antioxidants play complementary roles in defense against oxidative stress caused by O₂^•−. This formulation solves the apparent puzzle presented by observations in model organisms that radiation resistance and longevity are independent of MnSOD, while defense against other modes of oxidative stress are not (3, 56, 67–69).

MATERIALS AND METHODS

Bacteria.

The WT bacterial strains used were as follows: Deinococcus radiodurans (USUHS) (ATCC BAA-816), Lactobacillus plantarum (ATCC 14917), and Escherichia coli (strain K-12 MG1655). The ΔMnSOD D. radiodurans MD885 is isogenic with ATCC BAA-816 and was derived by transfer of the disrupted sodA locus (DR1279) of KKW7003 by kanamycin selection (52); the catalase A (katA⁻) mutant is KKW7003 (52); E. coli mutant strain OP50 is a uracil auxotroph; OP50-1 is also streptomycin-resistant.

Yeasts.

The WT Saccharomyces cerevisiae strain EXF-6218 was chosen from a collection of fungi previously gauged for gamma radiation resistance. Strain EXF-6218 is a diploid and one of the most radiation-sensitive S. cerevisiae strains reported (5, 35).

Nematodes.

The WT Caenorhabditis elegans strain N2 was originally obtained from David Eisenmann, University of Maryland Baltimore County and was used as a control strain in the nematode studies. The following ΔSOD mutant strains were compared to N2: MQ1766 is a quintuple sod-1,2,3,4,5 [sod2(ok1030) I; sod5(tm1146) sod1(tm783) II; sod4(gk101) III; sod3(tm760) X] mutant with complete loss of MnSOD and Cu/ZnSOD activities; GA480 is a sod-2,3 [sod2(gk257) I; sod3(tm760) X] double mutant deficient in MnSOD enzymes (ΔMnSOD); GA184 sod-2 [sod2(gk257) I], RB1072 sod-2 [sod2(ok1030) I], and GA186 sod-3 [sod-3(tm760) X] are single ΔMnSOD mutants. All mutants were purchased from Caenorhabditis Genetics Center (CGC), University of Minnesota.

D. radiodurans growth, gamma irradiation, and samples for EPR.

WT D. radiodurans (ATTC BAA-816) was inoculated into defined liquid TGY rich medium (1% bactotryptone, 0.5% yeast extract, and 0.1% glucose). ΔMnSOD D. radiodurans (MD885) was inoculated into TGY supplemented with 25 μg/mL kanamycin (TGY-Km25). Both cultures were incubated with orbital shaking at 200 rpm at 32°C up to an optical density at 600 nm (OD₆₀₀) of ∼0.9, which corresponds to the mid-exponential phase of growth. Mid-exponential-phase WT and sodA⁻ D. radiodurans cells were then inoculated into two 2-L flasks with 0.8 L TGY or TGY-Km25. This corresponded to an inoculation of approximately 10⁵ CFU/mL, followed by incubation with aeration (200 rpm orbital shaking) at 32°C for 10 days without change of broth.

Mid-exponential-phase (young) actively dividing cells were harvested when a culture reached OD₆₀₀ 0.9, defined as day 1 (d1). After an additional 24 h (d2), early-stationary-phase (mature), metabolically active but nondividing cells were harvested from liquid medium; late-stationary-phase (old) cells were harvested at 144 h (d6), and mostly replication-defective cells on d10. The D. radiodurans WT and ΔMnSOD cells thus isolated at each stage were irradiated acutely at 8.8 kGy/h (⁶⁰Co) on ice, and age-dependent survival was determined by CFU assay after exposure to 0, 10, and 15 kGy (⁶⁰Co) (70, 71). Samples were serially diluted using TGY medium, and the dilutions were plated on TGY agar plates (100 μL per plate). CFU were counted after 1 week at 32°C (Fig. 2A, B, and F; see also Fig. S1A).

For EPR analysis, 30-mL cultures (1 × 10⁸ to 3 × 10⁸ CFU/mL) were harvested by centrifugation, and cells were washed twice with MilliQ water from “Banstead Nano Pure Diamond” [Thermo Scientific] and resuspended in 0.3 mL 20% Glycerol (vol/vol) MilliQ water. EPR tubes (quartz, inner diameter 2.1 mm, length 65 to 70 mm; Wilmad-LabGlass) were filled with ∼80 μL of the concentrated cell suspensions and stored frozen at −80°C for EPR analyses.

Establishment and maintenance of C. elegans colonies on NGM agar plates.

Colonies of the following C. elegans strains were established and maintained on nematode growth medium (NGM) agar and fed OP50 E. coli. Strain Bristol N2 (WT) (a gift from David Eisenmann, UMBC, Baltimore, MD) and SOD mutants (see above, “Nematodes”) were provided by the CGC, funded by the NIH Office of Research Infrastructure Programs (P40OD010440). Briefly, ∼1,000 worms were maintained at ambient temperature on a 100-mm petri dish containing NGM medium (1.7% agar [wt/vol], 50 mM NaCl, 0.25% peptone [wt/vol], 1 mM CaCl₂, 5 μg/mL cholesterol, 25 mM KH₂PO₄, and 1 mM MgSO₄) and seeded with a lawn of E. coli OP50 as the food source. As the OP50 lawns were depleted, an ∼1-cm² section of agar (containing worms) was incised from the agar plate and inverted onto a fresh lawn of OP50, thus giving the worms sufficient nutrients to continue developing and reproducing.

Establishment and maintenance of C. elegans culture in axenic medium.

C. elegans cultured on NGM medium with OP50 was allowed to reach a density of ∼85% gravid adult worms (3 to 5 days). They were rinsed from the plate in sterile distilled water (dH₂O), and pelleted at 10,000 × g for 30 s; the supernatant was discarded. The worms were repeatedly rinsed and pelleted as above. C. elegans eggs were isolated from the gravid adults by incubating the worm pellet in freshly prepared 20% alkaline hypochlorite solution (0.25 M NaOH, 20% bleach) for 3 to 5 min with occasional vortexing. The suspension was monitored for dissociation of adult worms and egg release. The eggs were pelleted (10,000 × g for 30 s), rinsed 3 times with sterile dH₂O, and resuspended in 3 mL of fresh sterile dH₂O. The egg suspension was transferred to a clean 15-mL conical tube and incubated overnight at ambient temperature with gentle shaking, allowing for egg hatch and developmental arrest of larval stage 1 (L1) worms.

The newly hatched L1 larvae were pelleted (800 × g for 3 min), the supernatant was discarded, and the worms were resuspended in 10 mL of axenic medium. The suspension was transferred to a T25 polystyrene culture flask and incubated at ambient temperature with gentle shaking. Cultures were expanded to 30 mL of medium in T75 flasks and subcultured weekly to maintain a viable worm concentration of ∼20,000 worms/mL. The axenic medium used in this study is a modified version of that previously described (72, 73). A 2× stock of custom-ordered C. elegans maintenance medium (CeMM; MediaTech, Inc.) was diluted to a 1× working stock and supplemented with 20 μM hemin chloride, 10% ultrapasteurized (UHT) skim milk, 5 μg/mL cholesterol, 100 μg/mL tetracycline, and 100 units/mL of penicillin-streptomycin.

C. elegans sample preparation for EPR analysis.

The EPR study of C. elegans populations began by inoculating ∼3,000 L1 worms (WT and sod mutants) into 30 mL of fresh axenic liquid medium (ALM), yielding an optimal density of 20,000 worms/mL in 4 to 5 days at ambient temperature (74). In that time, the worms grew from ∼200 μm in length into ∼1-mm adults. We collected EPR spectra for intact WT and ΔSOD worms over their life span. Under our growth conditions, C. elegans lived 18 to 26 days, displaying the following stages: young worms (d1 to d3), mature (middle-aged) worms (d5 to d13), old worms (d15 to d21), and unresponsive worms (d26 to d33). One milliliter of axenically cultured C. elegans was collected at the indicated times and harvested by centrifugation (16,000 × g, 30 s). Worms were washed twice with 1 mL of MilliQ water, pelleted, and resuspended in 200 μL 20% glycerol in MilliQ water (vol/vol). Two EPR tubes were filled with ∼80 μL of the worm suspension. All EPR samples were stored at −80°C and transferred to liquid nitrogen prior to their being shipped to Northwestern University for EPR analysis.

EPR spectroscopy.

35-GHz continuous-wave (CW) EPR spectra were recorded using a lab-built EPR spectrometer (75). Absorption-display EPR spectra of frozen cells were collected in the “rapid passage” mode at 2 K as previously described (5, 76) (experimental conditions: medium wave [MW] frequency, 34.9 GHz; MW power, 1 mW; temperature, 2 K’ modulation amplitude, 2 G; time constant, 64 ms; scan rate, 1 kG/min). Data acquisition was done using a home-written program in LabVIEW. EPR spectra were simulated using the software EasySpin (5.2.30). For details of the procedures for determining the fractional contributions of H-Mn and L-Mn to the Mn²⁺ EPR spectra, see reference 5.

Chronic gamma irradiation of C. elegans.

C. elegans strains Bristol N2 (WT) and MQ1766 (sod-1,2,3,4,5) were cultured on NGM plates seeded with OP50 bacteria and prepped as described above for egg isolation and L1 larvae synchronization (see above, “Establishment and Maintenance of C. elegans Culture in Axenic Medium”). The newly hatched L1 larvae were resuspended in 200 μL dH₂O, transferred to fresh NGM plates seeded with OP50 bacteria (NGM/OP50 plates), and maintained at ambient temperature until they reached the young adult stage (N2, ∼4 days of incubation; ΔSOD mutants ∼6 days of incubation). Then, 80 young adult worms of each strain were transferred to fresh 60-mm NGM/OP50 plates at 10 worms per plate.

Four plates of each strain were exposed to the ¹³⁷Cs source, which generated a dose rate of 34.74 Gy/h and a temperature range of 23 to 26°C, for 8 days. Four control plates per strain were maintained for 8 days in the dark at a comparable temperature. Worms were monitored daily for mobility and developmental progression.

C. elegans and D. radiodurans protein-free small-molecule cell extract (ultrafiltrate) preparation.

C. elegans WT strain Bristol N2 was grown on NGM/OP50-1 plates for egg isolation and L1 larvae synchronization as described above. Synchronized day 2 L4-stage worms were washed with M9 minimal medium (6 mg/mL of K₂HPO₄ and 3 mg/mL of KH₂PO₄, 1 mg/mL of NH₄Cl, and 0.5 g of NaCl, 1 mM MgSO₄) off the NGM plates and agitated on a rocker for 1 h to allow full digestion of OP50-1 E. coli in worm intestines. The worm sample was washed in a cycle of additions of M9 and removal of M9 after centrifugation at 2,000 × g for 1 min each, with the final washing in 50 mM phosphate buffer, pH 7.4, and concentrated by centrifugation at 2,000 × g for 1 min. Supernatant was removed and worms were resuspended in lysis buffer (50 mM phosphate buffer, pH 7.4, 10 mM KCl, 1.5 mM MgCl₂, 1 mM DTT, and 2× EDTA-free protease inhibitor cocktail), twice the volume of the worm pellet based on Stiernagle 2006. The worms were then lysed using sonication and multiple freeze-thaw cycles in liquid nitrogen and spun at 1,000 × g for 10 min to remove nuclei and worm debris. The supernatants of the worms from the 1,000 × g spin were then centrifuged at 20,000 × g for 30 min to generate the soluble fraction. The soluble protein concentration was determined using protein assay kit (Bio-Rad), and the supernatant was adjusted to 20 mg/mL protein by dilution with MilliQ water. The supernatants were then ultracentrifuged at 190,000 × g (69 h, 4°C), and the ultracentrifuged supernatants were subjected to filtration through 3-kDa Centriplus centrifugal filter devices (YM-3) (Millipore Corporation, Bedford, MA, USA) at room temperature for 3 to 4 h. The ultrafiltrates were then concentrated 5× in a SpeedVac concentrator (Savant, GMI, Inc., Ramsey, MN, USA) at room temperature, boiled for 30 min, aliquoted, and stored at −80°C. D. radiodurans ultrafiltrate was prepared similarly (5, 15). Briefly, D. radiodurans cells were grown as a batch culture (in 1.5 L TGY medium) to an optical density of 0.9 (mid-exponential phase) determined at 600 nm. Harvested cells were washed twice in MilliQ water, centrifuged at 3,000 × g for 15 min at 4°C, and broken open in MilliQ water by passage through a One-Shot cell disrupter (38,000 lb/in²). Upon lysis, the crude D. radiodurans cell extract was centrifuged at 12,000 × g (1 h, 4°C), yielding an aqueous phase which contained water-soluble proteins and small molecules and a pellet of insoluble cell debris (15). The soluble protein concentration was determined, followed by ultracentrifugation and ultrafiltration. The ultrafiltrate was then concentrated 5 times at room temperature, boiled for 30 min, aliquoted, and stored at −80°C.

Pre- and postirradiation protein carbonylation detection.

Protein carbonylation was detected as described previously (15) with small modifications. Soluble protein fractions purified from an E. coli (K-12 MG1655) homogenate were used as the substrate for carbonyl analyses. Proteins were prepared from E. coli grown in TGY to OD₆₀₀ 0.9. Cells were broken open by passage through a One-Shot cell disrupter (38,000 lb/in²). Insoluble proteins and cell debris were removed by centrifugation at 31,000 × g (1 h, 4°C). The supernatant containing soluble E. coli proteins was subjected to ultracentrifugation (190,000 × g [69 h, 4°C]), and the protein pellet was dissolved in MilliQ water and quantified for concentration using the protein assay kit (Bio-Rad). The proteins were stored at −80°C as 44-mg/mL stocks in 150-μL aliquots, which were used once and then discarded. Before irradiation, a stock of E. coli proteins was diluted in MilliQ H₂O and concentrated D. radiodurans (DR) or C. elegans (CE) ultrafiltrate preparations, yielding mixtures with 4.4 mg/mL of E. coli protein in ultrafiltrate (50%). For a given E. coli protein-ultrafiltrate sample (44 μL), ⁶⁰Co irradiations were kept on ice under aerobic conditions. At the indicated gamma radiation doses, 10-μL samples were removed and stored on ice until the time course of irradiation was completed. The carbonyl groups in the E. coli protein samples were identified by the OxyBlot protein oxidation detection kit (Millipore, Sigma). Then, 10 μg of E. coli protein after DNPH (2,4-dinitrophenylhydrazine) reaction neutralization was loaded on denatured polyacrylamide protein gel for carbonyl Western blotting.

Pre- and postirradiation activity of BamHI.

Pre- and postirradiation activity of BamHI was determined as described previously (15, 20) with some modifications. BamHI (5,000 U/μL) (without bovine serum albumin [BSA]) (New England Biolabs, Ipswich, MA, USA) was diluted in D. radiodurans or in C. elegans ultrafiltrate to 0.54 U/μL. Then, 200 μL of the BamHI mixture was gamma-irradiated aerobically on ice. Following irradiation, 20 μL of each radiation-treated BamHI sample was assayed for residual endonuclease activity in separate reaction mixtures (final volume, 30 μL) containing 125 ng lambda phage DNA, 50 mM NaCl, 10 mM Tris-HCl (pH 7.9), 10 mM MgCl₂, and 1 mM dithiothreitol (New England Biolabs; buffer 3). BamHI/lambda DNA mixtures were incubated for 1.25 h at 37°C, followed by agarose (0.8%) gel electrophoresis.

Gamma-irradiations of protein samples.

E. coli soluble proteins mixed with ultrafiltrate samples and BamHI-ultrafiltrate samples were irradiated in 0.5-mL Eppendorf tubes on wet ice with ⁶⁰Co.

Growth of D. radiodurans under high-level chronic gamma irradiation (¹³⁷Cs).

To evaluate the growth of D. radiodurans WT, sodA⁻ (MD885), and katA⁻ (KKW7003) strains under continuous gamma irradiation, cells were pregrown to an OD₆₀₀ of 0.9 (∼10⁸ CFU/mL) in liquid TGY medium at 32°C with shaking at 200 rpm. Cells were serially diluted (10⁻¹ to 10⁻⁵) in liquid TGY, and 5 μL of each culture was spotted onto two separate TGY plates. All plates were incubated at 28 to 30°C and sealed with Parafilm “M” laboratory film, one set exposed to 57 Gy/h within a ¹³⁷Cs irradiator and one set in the absence of radiation (outside the irradiator). The images were taken after 6 days of incubation under 57 Gy/h.

UVC irradiation.

WT, sodA⁻ (MD885), and katA⁻ (KKW7003) D. radiodurans were grown in TGY or TGY-Km25 medium to the early stationary phase. Cells were serially diluted (10⁻¹ to 10⁻⁵) in liquid TGY, and 5 μL of each culture was spotted onto two separate TGY plates. Opened petri plates were exposed to UVC (254 nm) for 18 min. Irradiance on plate surfaces was 1 W/m², which gives radiant exposure of 1 J/m²/sec. The radiant exposure accumulated in 18 min was 1,080 J/m².

Desiccation.

WT and sodA⁻ D. radiodurans cells were grown in TGY or TGY-Km25 medium to the early stationary phase (1 × 10⁸ CFU/mL), transferred to Costar Stripwell 96-well microplates (Corning; catalog no. 9102), 25 μL per well. The microplates were transferred to desiccation chambers with Drierite and stored at room temperature. For survival monitoring, the well contents were resuspended in 250 μL of medium (dilution factor of 10⁻¹), and the samples were diluted serially using TGY medium. The dilutions were plated on TGY agar plates (100 μL of sample per plate). The plates were then incubated at 32°C for up to 1 week. For L. plantarum, desiccation survival assays were on cells in MRS medium at 37°C for 18 h.

Paraquat.

Overnight cultures of D. radiodurans WT (in TGY medium) and sodA⁻ mutant MD885 (in TGY-Km25) in the volume of 100 μL were spread as colony biofilms (lawns). Sterile paper disks saturated with 20 μL of 12.5 mM paraquat (methyl viologen) dissolved in MilliQ H₂O were then placed on the middle of the plates, followed by incubation at 32°C for 6 days.

Statistics.

For Fig. 4A, the data were compiled on irradiated WT (SOD+) (N2 strain) and ΔSOD (sod-1,2,3,4,5) (MQ1766 strain) C. elegans with survival as a function of time under chronic gamma irradiation at 34.74 Gy/h. All worms that became static (immobile) or were lost to observation were considered dead, and those that retained some mobility or twitching were considered alive. Based on these assumptions, survival analysis for WT and SOD– worms was performed using the survival, survminer, and ggplot2 packages in R 4.0.3 software.

Kaplan-Meier survival curves were constructed with 95% confidence intervals (95% CIs) for each worm group. The curves could not be reliably compared by log rank test because the survival curves visibly crossed. For the same reason, proportional hazard modeling (Cox regression) was also unreliable because the Chi² test for the validity of the proportional hazard assumption (using the cox.zph command in R) returned a small P value of 0.00019.

To address these issues, a more flexible parametric survival modeling approach (mixed effect excess hazard models) using the mexhaz R package was implemented. In this method, the death hazard function was modeled using the Weibull distribution with one of the two parameters allowed to differ between the SOD+ and SOD– worm groups. This approach could handle the situation where Kaplan-Meier survival curves for different groups crossed each other and the hazard functions were not proportional.