Accumulation of Mn(II) in Deinococcus radiodurans Facilitates Gamma-Radiation Resistance

See allHide authors and affiliations

Science  05 Nov 2004:
Vol. 306, Issue 5698, pp. 1025-1028
DOI: 10.1126/science.1103185


Deinococcus radiodurans is extremely resistant to ionizing radiation. How this bacterium can grow under chronic γ radiation [50 grays (Gy) per hour] or recover from acute doses greater than 10 kGy is unknown. We show that D. radiodurans accumulates very high intracellular manganese and low iron levels compared with radiation-sensitive bacteria and that resistance exhibits a concentration-dependent response to manganous chloride [Mn(II)]. Among the most radiation-resistant bacterial groups reported, Deinococcus, Enterococcus, Lactobacillus, and cyanobacteria accumulate Mn(II). In contrast, Shewanella oneidensis and Pseudomonas putida have high iron but low intracellular manganese concentrations and are very sensitive. We propose that Mn(II) accumulation facilitates recovery from radiation injury.

Deinococcus radiodurans is a nonpathogenic, nonsporulating, obligate aerobic bacterium that typically grows in undefined rich medium (TGY) as clusters of two cells (diplococci) in the early stages of growth and as four cells (tetracocci) in the late stages (1) (Fig. 1A). D. radiodurans is extremely resistant to ionizing radiation (IR) and desiccation (2) and maintains four to eight genomic copies per cell (3). The repair of IR- and desiccation-induced DNA double-stranded breaks (DSBs) (2) is known to be mediated by homologous recombination (4), but no SOS response (5) (table S1) or nonhomologous end-joining of DSBs (6) is observed in this organism. Yet, the complexity of the genetic systems underlying DNA repair in D. radiodurans remains poorly defined (48), and three hypotheses have been proposed: (i) D. radiodurans uses conventional repair pathways with greater efficiency than other bacteria (46); (ii) there are repair functions encoded among its hypothetical genes (8); or (iii) repair is facilitated by its ringlike nucleoids (RNs) (9). These hypotheses emphasize repair of DNA damage caused by the direct effects of γ photons and the indirect effects of reactive oxygen species (ROS) induced during irradiation (10). When ROS exceed the capacity of endogenous scavengers to neutralize them, cells become vulnerable to damage, a condition referred to as oxidative stress (11).

Fig. 1.

TEM (12). (A) D. radiodurans grown in TGY, late-log phase (LLP). tet, tetracoccus; dip, diplococcus. (B) D. radiodurans grown in TGY, early-stationary phase (ESP). (C) D. radiodurans grown in DMM, ESP. (D) D. radiodurans, diplococcus, grown in TGY, ESP. (E) D. radiodurans, diplococcus, grown in DMM, ESP. (F and G) D. grandis, grown in TGY, ESP. Scale bars, 0.5 μm.

One DSB repair model for D. radiodurans (9) attributes the resistance phenotype to the presence of RNs (Fig. 1, B and D). D. radiodurans grows predominantly as diplococci in defined minimal medium (DMM) (12), even in the late stages of growth (13). We examined cells grown in DMM or TGY by transmission electron microscopy (TEM) to determine the prevalence of RNs (Fig. 1) and also tested cells for their resistance to IR (Fig. 2). The resistance of D. radiodurans cultures grown in TGY (Fig. 2) that contained cells that lacked RNs (Fig. 1A) was greater than the resistance of cultures grown in DMM (Fig. 2 and SOM Text) that contained cells with RNs (Fig. 1, C and E). Deinococcus grandis grows as single cells (14), is similarly resistant to IR in either TGY or DMM (Fig. 2), and rarely displayed RNs (Fig. 1, F and G). So far, neither genomic (5, 7) nor experimental (39) analyses unequivocally support any model that explains the radioresistance of Deinococcaceae.

Fig. 2.

Survival of strains exposed to acute IR (12). Open triangle, D. radiodurans (pregrown in TGY, plated on TGY) (LLP, Fig. 1A); solid circle, D. radiodurans (pregrown in DMM, plated on TGY) (ESP, Fig. 1, C and E); solid triangle, D. radiodurans [pregrown in DMM + 50 μM 2,2′-dipyridyl (Dp) and 50 μM deferoxamine mesylate (Ds) (Dp and Ds are Fe-chelators), plated on TGY] (ESP); open circle, D. grandis (pregrown in TGY, plated on TGY) (LLP); solid diamond, D. grandis (pregrown in DMM, plated on TGY) (ESP, Fig. 1, F and G); open diamond, E. coli (OD600 0.9) (pregrown in TGY, plated on TGY); solid square, S. oneidensis (OD600 0.9) (pregrown in TGY, plated on TGY).

We determined that growth of D. radiodurans in DMM is dependent on Mn(II) (Fig. 3A) but not on Fe, Co, or Mo (Fig. 3B). Moreover, growth and resistance of D. radiodurans was unaffected by Fe chelators (Fig. 2) (Fig. 3C). To further examine the dependence of D. radiodurans on Mn and Fe, we used 54Mn and 59Fe to assay accumulation relative to the bacterium Shewanella oneidensis (Fig. 3D) (15), and the total Mn and Fe contents of D. radiodurans and other bacteria were determined by using an inductively coupled plasma-mass spectrometry method (ICP-MS) (16). D. radiodurans accumulated substantially more Mn than S. oneidensis but less Fe (Table 1). Mn(II) is essential for the detoxification of ROS in most bacteria, principally as a cofactor for the Mn-dependent enzyme superoxide dismutase (Mn-SOD). However, Lactobacillus plantarum incorporates Mn(II) as a protectant in place of Mn-SOD (17); L. plantarum is an Fe-independent, radioresistant bacterium that exhibits energy-dependent Mn accumulation (1618). Mn transport in D. radiodurans is also energy dependent, because carbonyl cyanide 3-chlorophenylhydrazone inhibited 54Mn accumulation by >75% (Fig. 3D).

Fig. 3.

Role of transition metals in wild-type D. radiodurans (12). (A) Growth dependence on Mn(II) in DMM. (B) Dependence on transition metals in DMM containing 2.5 μM Mn(II), Fe(II), Co(II), Mo(II), or [Cd(II) (2.5 μM) + Mn(II) (2.5 μM)]. Trace, 0.2 μM each of Mo, Cu, Cr, Bo, Zn, and I. (C) Effect of Fe chelators (50 μM Dp and 50 μM Ds) on recovery from IR. (D) D. radiodurans accumulated 54Mn, but not 59Fe, in an energy-dependent manner. Inset circle for designation.

Table 1.

D10 survival values and intracellular Mn and Fe levels (12).

StrainsView inline Radiation dose yielding 10% CFU survival Desiccation dose yielding 10% CFU survivalView inline59Fe accumulation, atoms/cell 54Mn accumulation, atoms/cell Total Fe: ICP-MS/nmol Fe/mg protein Total Mn: ICP-MS/nmol Mn/mg protein Intracellular Mn/Fe (ICP-MS) concentration ratio
D. radiodurans View inline 16 kGy >30 days 2.7 × 103View inline 1.08 × 105View inline 1.49 (±0.39) 0.36 (±0.11) 0.24
Deinococcus geothermalis View inline 10 kGyView inline >30 days 7.7 × 103 (±1.1 × 103) 2.17 × 105 (±3.7 × 103) 1.7 (±0.39) 0.78 (±0.15) 0.46
Enterococcus faecium View inline 2.0 kGyView inline >30 days ND ND 6.3 (±2.8) 1.1 (±0.21) 0.17
E. coli View inline 0.7 kGy 8 days 7 × 105 (View inline) 3.8 × 104 (View inline) 2.72 (±0.63) 0.0197 (±0.0027) 0.0072
Pseudomonas putida View inline 0.25 kGyView inline 1 day ND ND 6.8 (±0.70) <0.001 <0.0001
S. oneidensis View inline 0.07 kGy <1 day 2.7 × 104View inline <5 × 102View inline 4.98 (±0.40) 0.0023 (±0.00005) 0.0005
D. radioduransView inline low-Mn DMM 10 kGy ND ND ND 2.1 0.7 0.33
D. radioduransView inline high-Mn DMM ND ND ND ND 1.6 3.9 2.5
  • View inline* Unless stated otherwise, strains were grown to OD600 0.9 in TGY medium, washed twice in 1 × PBS containing 1 mM EDTA, and subjected to ICP-MS (16).

  • View inline Deinococcus-Thermus phylum (5).

  • View inline Gram-positive.

  • View inline§ Gram-negative.

  • View inline Grown in/recovered on DMM + 25 nM Mn(II).

  • View inline Grown in/recovered on DMM + 2.5 μM Mn(II).

  • View inline# See fig. S3.

  • View inline** See fig. S5.

  • View inline†† See Fig. 3D. For TGY, ICP-MS showed 204 (±78) nM Mn and 6085 (±1111) nM Fe. For DMM/DRM [without Mn(II) supplementation], ICP-AES (Atomic Emission Spectrometry) showed 5.6 (±2.1) nM Mn and 1799 (±2.8) nM Fe. CFU, colony-forming unit. ND, not determined.

  • Because D. radiodurans does not grow as a monococcus (1, 3), survival frequencies for a single-celled population cannot be determined experimentally (SOM Text). At 12 kGy, 17% of D. radiodurans cells are statistically calculated to survive (12). The IR doses that yield 17% survival of Escherichia coli and S. oneidensis cells are lower by factors of 20 and 200, respectively, than those for D. radiodurans (Fig. 2, inset). High levels of Mn(II) have been reported to be associated with D. radiodurans DNA (19) (fig. S1), but it is unlikely that Mn(II) protects DNA during irradiation itself, because the numbers of DSBs per Gy per genome for a given dose in D. radiodurans, E. coli, and S. oneidensis are very similar (fig. S2). However, the differences in resistance observed between these and other organisms reported here mirror the trend in their intracellular Mn/Fe concentration ratios (Table 1).

    ROS produced by IR or metabolism can kill cells (10, 11, 13). Hydroxyl radicals (HO·) are a primary product of the radiolysis of water, are extremely toxic (10), and in the presence of O2 can generate other ROS, including superoxide ions (Math) and hydrogen peroxide (H2O2). Probably the most important source of ROS in aerobic cells is the respiratory chain, which can give rise to high levels of Math and H2O2 (11). Irrespective of source, H2O2 is relatively stable and diffusible, but in the presence of free Fe(II) the Fenton and Haber-Weiss reactions decompose H2O2 to HO· (11). In contrast, Mn(II) is not known to participate in Fenton-type chemistry in vivo (17). The mechanism by which Mn(II) scavenges Math is not understood but requires considerably higher intracellular Mn(II) levels than those needed for Mn-SOD–mediated protection (17). ICP-MS analysis showed that the Mn content of D. radiodurans grown in high-Mn(II) conditions is 5.6 times as much as that in low-Mn(II) conditions (Table 1). In the presence or absence of 50 Gy/hour, D. radiodurans growth was similarly good on high-Mn DMM (Fig. 4A). However, growth was inhibited on low-Mn DMM under chronic radiation but not in the nonirradiated control (Fig. 4A). For cells unable to grow under 50 Gy/hour on low-Mn DMM, increasing the concentration of amino acids restored growth (Fig. 4A). Thus, oxidative stress produced by metabolism potentiates the lethal effects of radiation in D. radiodurans, where cells growing on DMM limited in amino acids and Mn(II) are overwhelmed by two sources of ROS, metabolism and irradiation.

    Fig. 4.

    Effect of γ radiation on Mn-accumulating bacteria (12). (A) Growth of D. radiodurans on DMM plates under 50 Gy/hour is dependent on Mn(II). Cells were pregrown in DMM with 50 nM Mn(II), 100 nM Mn(II), or 250 nM Mn(II). I, no irradiation control, [DMM 25 + nM Mn(II)]; II, [DMM 2.5 μM Mn(II)] + 50 Gy/hour; III, [DMM + 25 nM Mn(II)] + 50 Gy/hour; IV, [DRM (12) 25 nM + Mn(II)] + 50 Gy/hour. (B) Recovery of D. radiodurans from acute IR exhibits a concentration-dependent response to Mn(II) (fig. S3). I and III, no-Mn DRM; II and IV, DRM + 2.5 nM Mn(II). (C) Growth in genotoxic environments. I, control, TGY; II, TGY + 50 Gy/hour; III, TGY + Dp + Ds (each, 125 μM); IV, TGY + 50 Gy/hour + Dp + Ds (each, 125 μM). DR, D. radiodurans; DG, D. grandis; SO, S. oneidensis; EF, E. faecium; LP, L. plantarum; EC, E. coli.

    The intracellular concentration ratios of Mn and Fe in D. radiodurans grown in TGY or low-Mn DMM are similar (0.2 to 0.3), and no significant differences in cell-survival (12) were observed on either medium for acutely irradiated cells (D10, ∼10 to 12 kGy) (Table 1). D. radiodurans can grow in defined rich medium (DRM) without Mn(II) supplementation (no-Mn DRM) (12), and the effect of Mn(II) depletion on recovery was tested (Fig. 4B). For D. radiodurans grown in no-Mn DRM, the intracellular Mn/Fe concentration ratio was 0.04, and the D10 cell-survival value on no-Mn DRM was ≤2.5 kGy (fig. S3), quantitatively similar to the IR resistance of several highly sensitive D. radiodurans DNA repair mutants (e.g., DNA polymerase I) (20). Mn-SOD is not critical to survival following acute irradiation (21), and we have shown that Mn-SOD is not needed for growth under 50 Gy/hour (fig. S4). In this context, we note that D. radiodurans has one of the highest catalase activities reported for any bacteria (22), and this would serve to remove H2O2 generated by nonenzymic Mn(II)-based dismutation of Math (17).

    In general, most of the radiation-resistant microorganisms reported have been Gram positive and the most sensitive have been Gram negative (23). One exception is the Gram-negative cyanobacterium Chroococcidiopsis, which is extremely resistant (D10, 5 kGy) (24). We believe it is noteworthy that the most resistant non–spore-forming bacteria reported belong to the deinococci (5), cyanobacteria (25), enterococci (23), and lactobacilli (18). These groups share traits including Mn accumulation (Table 1) (17, 25), high intracellular Mn/Fe concentration ratios (Table 1) (16, 25), resistance to desiccation (fig. S5) (24), and growth in the presence of Fe chelators and/or 50 Gy/hour (Fig. 4C) (16). An exception to this paradigm is Neisseria gonorrhoeae, which accumulates Mn(II) but is sensitive to IR (D10, 0.125 kGy) (26). However, N. gonorrhoeae also has a high Fe requirement; the intracellular Mn/Fe concentration ratio of N. gonorrhoeae is 0.004 (27).

    DNA repair systems identified in D. radiodurans appear less complex and diverse than those reported for E. coli, S. oneidensis, or P. putida (tables S1 and S2). Regarding the three hypotheses presented above, the high intracellular Mn/Fe concentration ratio of D. radiodurans might underlie the efficiency of its repair pathways by protecting cells from ROS generated during recovery. Our findings do not preclude the existence of novel DNA repair genes, but few have been identified to date (5, 8), and they do not support a role of RNs in resistance (9). For Fe-rich, Mn-poor cells, death at low doses might not be caused by DNA damage inflicted during irradiation. For example, 90% of S. oneidensis cells do not survive 0.07 kGy, a dose that induces <1 DSB/genome (fig. S2) (Table 1). Instead, S. oneidensis might be primed for Fenton-type chemistry by the release of Fe(II) during irradiation (28) but unable to recalibrate enzymic defense systems (table S3) in time to counter sudden increases in Math-related ROS during recovery (11). In contrast, accumulated Mn(II) would be unaffected by radiation and functionally poised to act against increases in Math; similar arguments can be made to explain resistance to desiccation (2) (fig. S5) (Table 1).

    We have demonstrated a critical role for the accumulation of Mn(II) in D. radiodurans in a mechanism toward surviving IR that is independent of Mn-SOD. The existence of high intracellular Mn/Fe concentration ratios in phylogenetically distant, radiation-resistant bacteria but not in sensitive cells supports the idea that Mn(II) accumulation (with low Fe) might be a wide-spread strategy that facilitates survival. Taken collectively, our results indicate that Mn(II) transport and regulation systems (figs. S6 and S7) are potential new targets to control recovery from radiation injury.

    Supporting Online Material

    Materials and Methods

    SOM Text

    Figs. S1 to S7

    Tables S1 to S3

    References and Notes

    References and Notes

    Stay Connected to Science

    Navigate This Article