Identification of new genes contributing to the extreme radioresistance of Deinococcus radiodurans using a Tn5-based transposon mutant library.

Here, we have developed an extremely efficient in vivo Tn5-based mutagenesis procedure to construct a Deinococcus radiodurans insertion mutant library subsequently screened for sensitivity to genotoxic agents such as γ and UV radiations or mitomycin C. The genes inactivated in radiosensitive mutants belong to various functional categories, including DNA repair functions, stress responses, signal transduction, membrane transport, several metabolic pathways, and genes of unknown function. Interestingly, preliminary characterization of previously undescribed radiosensitive mutants suggests the contribution of cyclic di-AMP signaling in the recovery of D. radiodurans cells from genotoxic stresses, probably by modulating several pathways involved in the overall cell response. Our analyses also point out a new transcriptional regulator belonging to the GntR family, encoded by DR0265, and a predicted RNase belonging to the newly described Y family, both contributing to the extreme radioresistance of D. radiodurans. Altogether, this work has revealed new cell responses involved either directly or indirectly in repair of various cell damage and confirmed that D. radiodurans extreme radiation resistance is determined by a multiplicity of pathways acting as a complex network.

Introduction

The extremely radiation resistant organism, D. radiodurans has been extensively studied since several decades to elucidate the molecular mechanisms responsible for its exceptional ability to withstand lethal effects of various DNA-damaging agents, such as ionizing and UV radiation, toxic chemicals and desiccation (for recent reviews, see [1-3]).

Prevalent features playing a key role in this extreme radioresistance have been already described: (i) D. radiodurans possesses highly proficient DNA double strand break (DSB) repair mechanisms, as homologous recombination (HR) [4], Extended Synthesis-Dependent Strand Annealing (ESDSA) [5,6], and Single-Strand Annealing (SSA) [7,8], that enable D. radiodurans to accurately reassemble its genome from hundreds of DNA fragments produced by irradiation (ii) D. radiodurans has also evolved a combination of very efficient non-enzymatic and enzymatic antioxidant defenses which specifically protect proteins against oxidative damage (for reviews, see [3,9,10]) (iii) a highly condensed ring-like nucleoid may also facilitate genome reassembly [11], although this hypothesis is still controversial [9,12].

A rapid and efficient response is required for cell recovery from the various cellular damages induced by irradiation. A subset of Deinococcus genus-specific genes, ddrA, ddrB, ddrC, ddrD, ddrI, ddrO (for DNA damage response), and pprA (Pleiotropic protein promoting DNA repair), have been identified as strongly induced by exposure to ionizing radiation or desiccation [13]. A common radiation/dessication response motif (RDRM) was found upstream of many highly radiation-induced genes in D. radiodurans as well as upstream of their homologs in D. geothermalis and D. deserti, thus defining the RDR regulon [14]. However, the regulatory mechanisms underlying the response of D. radiodurans to radiation are still poorly understood. The protein IrrE (also referred to as PprI), was described as a general switch, up-regulating expression of various proteins in D. radiodurans or Deinococcus deserti [15-17]. DdrO was proposed to be the global transcriptional regulator of the RDR regulon [14], and was recently shown to be cleaved in an IrrE-dependent manner upon exposure to ionizing radiation [17].

Finally, it now appears that the extreme radiation resistance of D. radiodurans is also due to a combination of diverse metabolic and regulatory pathways, but the links making a comprehensive network from these various mechanisms is still missing and several factors acting in these pathways still remain to be discovered.

Large-scale mutant libraries remain an efficient method to identify individual proteins required for a complex biological response such as for radiation resistance. Early efforts were based upon classical MNNG (N-methyl-N’-nitro-N-nitrosoguanidine) mutagenesis technique but this approach identified only a few new D. radiodurans loci due to the difficulties to map point mutations [18,19]. The transcriptome and proteome approaches have identified differentially regulated genes after exposure to ionizing radiation and desiccation, [13,20-23], but did not allow identification of genes constitutively expressed for cell defense against genotoxic stresses.

Here, we describe a highly efficient D. radiodurans in vivo mutagenesis method based on a hyperactive version of the Tn5 transposition system, carried by a temperature-sensitive vector suitable for D. radiodurans [24] as transposon delivery system. This system combines mutations in the Tn5 transposase encoding gene (tnp) as well as in the ends of the transposon [25-28]. The use of this mutagenesis system provide large collections of mutants since insertions of this element into DNA are highly random and the in vivo transposition does not need host factors [29,30]. The D. radiodurans Tn5-based insertion library has been subsequently screened for sensitivity to γ- and UV rays, as well as after exposure to mitomycin C (MMC). The transposon insertion site was mapped on the genome by arbitrary PCR and sequencing for each of the 206 mutants sensitive to at least one of these DNA damaging agents, mainly γ-rays. This analysis and further determination of the mutant survival rates after γ-irradiation enabled us to identify 37 genes that significantly contribute to the extreme resistance of D. radiodurans to genotoxic stresses. These include genes involved in DNA repair, stress responses and various metabolic processes. We also performed an initial characterization of three previously undescribed radiosensitive mutants inactivated for loci DR0007, DR0265 and DR2462.

Results and Discussion

Construction of a Tn5-based transposon mutant library in D. radiodurans

To generate a collection of mutants in D. radiodurans, we have developed an in vivo Tn5-based mutagenesis system. For this purpose, we cloned a mini-Tn5 derived (Tn5-hph) transposable element and a mutant tnp gene encoding a hyperactive Tn5 transposase [26] into a conditionally replicating temperature-sensitive shuttle vector (repUTs), that was shown previously to be stably maintained at 28°C in D. radiodurans and rapidly lost at 37°C [24] (plasmid p13554, Fig 1A). The Tn5-hph mini-transposon was constructed by assembling in vitro a cassette conferring hygromycin resistance and two flanking optimized 19-bp transposase recognition sequences (Mosaic Ends; [28]) (see Material and Methods for details of the construction). The tnp gene encoding the transposase was placed under the control of the P promoter [31] and cloned outside the mobile element to obtain stable insertions upon the loss of the delivery vector.

Fig 1

Description of the transposition procedure used to create the D. radiodurans Tn5-insertion mutant library.

(A) The Tn5-based transposon delivery vector, p13554 is a derivative of the temperature sensitive plasmid p13841 [24]. The mini Tn5 (Tn5-hph) consists of a hygromycin resistance cassette (hph) as a selectable marker, flanked by the optimized 19-bp mosaic ends (ME) of Tn5 [28]. The hph gene is expressed under the control of the P promoter. The hyperactive Tn5 transposase [25,26,94] is cloned outside the mobile element to generate stable insertions and is expressed from the P promoter inducible with IPTG [31]. Ter116 is a D. radiodurans transcription terminator. (B) Flow chart of the procedure used to screen the D. radiodurans Tn5-insertion mutant library for sensitivity to DNA damaging agents.

The mutant library was constructed in a D. radiodurans host (GY10973) expressing the LacI repressor to repress expression of the transposase from the P promoter and avoid possible toxic effects previously reported in E. coli [32,33]. The isolation of Tn5 insertion mutants was performed in two steps: (i) the host strain was transformed with the transposase delivery vector and the transformants were selected at a permissive temperature (30°C) on TGY1X plates supplemented with both hygromycin and spectinomycin. (ii) Transformed cells were cultivated into TGY2X medium at 30°C to an A650 ≈ 0.1 (transposition step) and appropriate dilutions were plated on hygromycin plates at the non-permissive temperature (37°C) to simultaneously select for Tn5-hph insertion mutants and to cure the delivery vector (see Materials and Methods for details). The transposition frequency is about 1x10-2 (insertion mutants/viable cell), which is about 10-fold higher than observed in other bacterial species when using the Tn5 hyperactive transposase [29,30,34], thus demonstrating the efficiency of our method.

To test whether Tn5-hph insertions occurred randomly into the D. radiodurans genome, the insertion sites for a sample of 42 hygromycin-resistant mutants isolated at 37°C were localized by an arbitrary PCR procedure followed by DNA sequencing (see Materials and Methods). We found that each Tn5-hph insertion mapped to a unique location (data not shown), indicating that the insertion mutants of our library were independent and that the insertions occurred randomly throughout the genome of D. radiodurans.

Large-scale isolation of ionizing radiation sensitive mutants

A total of 6207 HygR single-gene insertion mutants were screened for sensitivity to genotoxic agents such as γ-rays, UV-rays, and MMC according to the flow chart depicted in Fig 1B. The procedure included purification steps (steps 2 and 5 in Fig 1B) followed by two consecutive screening steps, replica plating and semi-quantitative spot test (steps 3 and 4, Fig 1B). Due to the multi-genomic status of D. radiodurans (4 to 8 genomic copies per cell according to [35]), the purification steps are important to favour homogenotization of the mutated alleles. Mutants that appeared sensitive to at least one DNA damaging treatment after the screening steps were further analyzed by backcross assays (introduction of the insertion into the parental reference strain R1) to confirm the link between the mini-Tn5 insertion and their respective phenotype (steps 5 and 6, Fig 1B). A cohort of 206 mutants was mapped by arbitrary PCR followed by sequencing and analyzed for their homozygous/heterozygous status by diagnostic PCR (step 7, Fig 1B). Master screening data of backcrossed insertion mutants are given in S1 Table.

As shown in S1 Table, the insertions mapped largely within the coding region of genes except rare events for which the transposon was inserted in intergenic regions. Most of the insertions were found at only one position in a given gene, whereas for around 17% of the mutants, single insertions were found independently at different sites into the same gene (S1 Table). Except the mutants inactivated for DR0400 and DRB0002 that were found sensitive to only MMC, the majority of mutants were sensitive to ionizing radiation (IR) and many showed cross-sensitivity to MMC and/or UV. About half of the mutants were heterozygous. These include mutants with insertions in genes involved in essential processes such as DNA replication (dnaE, dnaN, holA), DNA supercoiling (topA, gyrB), translation and ribosome biogenesis or assembly. Interestingly, the mutant disrupted for DR2606 encoding the homolog of PriA, a key protein of the main pathway for reactivation of stalled replication fork in bacteria [36,37], is sensitive to γ-, UV-radiation, and MMC. This result suggests the existence of a PriA-dependent replication restart primosome in D. radiodurans involved in the restoration of an intact genome after irradiation. A classification of the inactivated genes in accordance with the Cluster of Orthologous Groups (COG) data base (S1 Fig) shows that the mutants affected for functions involved in DNA replication, recombination and repair represented only a fraction (8.2%) of the total number of mutants. Indeed, the great majority of mutants were affected in various metabolic pathways including energy, coenzyme, amino acid, nucleotide and lipid metabolism, cell envelope biogenesis, and posttranslational modification (S1 Fig and S1 Table).

To evaluate more precisely the contribution of these genes to radioresistance, we measured the survival of each insertion mutant exposed to γ-rays at high doses of 10 and 15 kGy since sensitivity of repair genes mutants (as i.e radA, polX, sbcCD) become apparent only in heavily irradiated cells (for review, see [2]). As shown in S1 Table, the majority of the mutants were only marginally sensitive with less than a log decline in survival after exposure to 15 kGy. Only 37 insertion mutants showed a significant decline in their survival rate (survival rate lower than 6% when exposed to a dose of 15 kGy) as compared to the wild-type strain (Table 1). This result might be partly explained by the different physiological conditions between the screening procedure and the survival assay, which were performed by spot test on TGY plates and in liquid, respectively. The subset of the most radiation sensitive mutants (Table 1) highlights genes involved in diverse DNA repair pathways such as nucleotide excision repair (uvrA-1, uvrB, and uvrC) with a survival rate lower than 1% at 15 kGy, and recombinational repair (recG, recN, and ruvA), all being also sensitive to MMC and at a lesser extent to UV. We also identified DRA0346 encoding PprA required for accurate cell division following repair of DNA DSB [38-40].

Table 1

Genes detected in our screen for which the corresponding Tn5 insertion mutants exhibit the highest gamma radiation sensitive phenotype.

Functional category	Mutant locus	survival to γ-rays (%)		Sensitivity to			Comments
		10 kGy	15 kGy	MMC a	UV b	H2O2 c
CONTROL STRAINS
wild-type control		85	38	R	R	R
ΔrecA		<10 –5	<10 –5	SS	SS
ΔddrB		1	0.1	S	S
ΔpolX		50	5	s	s
katA::Tn5		48	15.5	R	R	SS
DNA replication, recombination, and repair
Nucleotide excision repair
UvrA1	DR1771	5.7	0.3	SS	s	s	heterogenote
UvrB	DR2275	6.2	0.7	SS	R	R
UvrC	DR1354	5.0	0.03	S	s	R
Homologous recombination
RecG	DR1916	23.0	4.1	SS	S	R
RecN	DR1477	28.9	5.5	SS	S	R
RuvA	DR1274	18.8	5.4	SS	SS	s	heterogenote
Other mechanisms
DNA topoisomerase I	DR1374	12.8	1.9	s	s	R	heterogenote
PprA	DRA0346	0.4	0.1	SS	SS	R
PriA	DR2606	85	4.2	SS	S	R
Stress response and other regulatory functions
Radiation resistance
Global regulator IrrE	DR0167	1	0.01	SS	SS	S
Response regulator DrRRA	DR2418	32	1.5	SS	R	R
DNA-damage responsive membrane protein	DR2518	26.3	6	SS	SS	s
Putative transcription factors
HTH transcriptional regulator, GntR family (COG2188)	DR0265	19.3	0.6	S	S	S
HTH transcription factor, CAP family DdrI	DR0997	4.7	0.4	SS	SS	s	heterogenote
Translation, ribosomal structure and biogenesis
30S ribosomal protein S19, RpsS	DR0315	39.2	2.7	R	s	s	heterogenote
Ribonuclease P protein component (RnpA)	DR2151	45.4	0.3	R	R	R
Alanyl tRNA synthetase, AlaS	DR2300	23.2	2.7	s	s	s	heterogenote
Metabolic pathways
Energy production and conversion
Rieske like Fe-S protein (COG0723)	DR0342	51.2	4.6	s	s	R
Cytochrome C-Type biogenesis protein CycJ	(u) DR0347	4.5	1.2	R	R	R
Aromatic compound dioxygenase, ferredoxin (COG2146)	DR1950	24.5	0.6	S	R	R	heterogenote
Cytochrome C oxidase subunit III, CyoB	DR2620	39.1	5.3	R	R	NA d	heterogenotegrowth defect
Malic enzyme	(o) DRA0276	39.4	4.2	R	R	R
Carbohydrate metabolism
Fructokinase RbsK ortholog	DR1525	48.7	3.4	SS	S	s
Amino acid transport and metabolism
Glycine/serine hydroxymethyltransferase (GlyA)	DR0038	41	2.1	s	s	R
Metabolism of coenzymes, cofactors and vitamins
Pantothenate synthetase, PanC	DR1164	51.6	4.1	S	S	R
ApbE family protein (involved in thiamine biosynthesis)	DR1794	42.4	5.6	R	R	R	heterogenote
Ketopantoate hydroxymethyl transferase	DR2615	44.5	4.3	SS	S	R
Membrane transport
Signal peptidase I (COG0681)	DR1321	42.6	4.6	S	S	R
NRAMP family membrane transporter	DR1709	28.8	4.6	R	s	R	heterogenote
Poorly characterized or uncharacterized proteins
Uncharacterized conserved protein (COG1624)	DR0007	10.1	0.5	SS	s	R
Small nucleotidyltransferase- like protein (COG1669)	DR0679	28.3	1.4	s	s	S
Exopolyphosphatase-related protein (COG0618) or DHH superfamily hydrolase	DR0826	33.1	2.5	SS	s	s
Protein containing an N-terminal CDNR domain	DR1740	38.5	5.5	R	R	NAd	heterogenotegrowth defect
Uncharacterized secreted protein	DR2058	51.9	4.6	R	R	R
RNase Y family	DR2462	30.1	0.9	S	SS	R
Predicted membrane protein	DR2572	38.9	3.4	R	s	R
Uncharacterized conserved secreted enzyme	DRA0022	40.5	4.8	R	R	R	heterogenote
intergenic insertion	between DR0251 and DR0252	24.2	0.1	R	R	R
intergenic insertion	between DR2613 and DR2614	33.2	5.3	R	R	R

a, bSS, highly sensitive, survival <10-6; S, sensitive, survival comprised between 10-6 and 10-4; s, slightly sensitive, survival comprised between 10-4 and 10-2; R, resistant, survival >10-2.

a onto TGY plates supplemented with 40 ng/mL mitomycin C (see Materials and Methods).

b onto TGY plates exposed to 600 J m-2 UV-rays (see Materials and Methods).

c by using the disc inhibition assay as described in Materials and Methods and Fig 2 (for the inhibition diameters corresponding to the different levels of sensitivity).

d NA: Not applicable because of growth defect.

(o) gene is located in a putative operon (predicted by FGENESB (www.softberry.com).

(u) transposon inserted in the upstream region of the corresponding CDS.

Fig 2

Measurement of sensitivity to H2O2 stress by disc inhibition assay (procedure described in Materials and Methods).

(A) Phenotype of the wild-type (resistant), the katA mutant (highly sensitive), and the DR0265 mutant (middle sensitive) are shown. (B) The mutants are classified into four categories depending on the diameter (in mm) of growth inhibition area as indicated.

The presence of genes involved in DNA repair and stress response regulation among the genes listed in Table 1 was expected. However, several genes, such as the recA, recO and recR genes involved in DNA double strand break repair by ESDSA and homologous recombination [41], or more generally known for their involvement in radioresistance, were not found by our screening procedure. Previous work showed that a limited number of RecA molecules (2500 molecules per cell in place of 44000 RecA molecules per cell in the wild type) were sufficient to ensure the same survival as those of the wild type bacteria upon γ-irradiation [42]. Interestingly, we found a slightly radiosensitive recF mutant in our screening on plates and we showed that mutation was not homogenotized. The purification steps on selective medium to favor homogenotization of the insertion mutations were not sufficient to obtain homogenotes when the mutations conferred an important selective disadvantage (S1 Table). We verified by PCR the homogenotization status of the mutants and 100 over 206 were found heterozygous, some of them being essential genes involved in DNA replication, DNA supercoiling, translation or metabolism (see S1 Table). Our screening also did not uncover the non-essential ddrA and ddrB genes, known to be involved in protection of 3’ single-stranded DNA ends and in DNA DSB repair by single-strand annealing, respectively. Given that 14957 independent mutants are required to obtain a 99% chance of inactivating a particular gene in D. radiodurans ([43] and see Material and Methods), we did not expect to identify all the genes involved in radioresistance. Indeed, we can estimate that approximately 14% of the genes were not inactivated among the 6207 insertion mutants we screened (see Material and Methods for calculation details).

Our screening also highlights genes previously reported as key players in the regulation of the radiation/dessication response of D. radiodurans such as: (i) irrE encoding the positive regulator of the DNA damage response [15-17,44]; (ii) DR2418 encoding DrRRA, the response regulator of a two-component system responsible for transcriptional regulation of numerous genes related to stress response and DNA repair [45]; (iii) DR2518 encoding a DNA damage-inducible membrane protein kinase important for DNA DSB repair [46,47]. In addition, we also isolated the heterozygous insertion mutant ddrI (DR0997), a DNA damage response gene encoding a transcription factor of the CAP family whose relative expression was shown to be significantly reduced in the drRRA mutant under both normal conditions and γ-radiation stress [45].

The main category of radiation sensitive mutants isolated in this study affects various metabolic functions. These results agree with recent data indicating that metabolic reprogramming plays a central role in the survival of organisms exposed to oxidative stress [48]. Mn ions play a major role in protecting proteins from oxidative damage as component of non-enzymatic metabolite complexes [49]. In accordance with the importance of a high Mn/Fe ratio for efficient recovery from irradiation injury, we isolated two mutants likely impaired in this ratio: (1) one heterozygous mutant for the gene DR1709, encoding a NRAMP family Mn(II) transporter [14], in which the intracellular concentration of Mn2+ might be lower than in wild type (2) the second mutant was inactivated for DR2106 encoding SufB (S1 Table), a member of the SUF system involved in both assembly and repair of oxidized oxygen-labile Fe-S clusters [50]; for a review, see [51]. Thus, alteration of SufB activity may increase intracellular free iron, promoting oxidative damage through the Fenton reaction [9,52]. SufB was also shown upregulated in D. geothermalis aerobically cultivated in low-Mn medium [53]. Likewise, inactivation of DR1321 encoding signal peptidase I protein might affect the release of exogenous amino acids and peptides. These products are important to protect cellular proteins against oxidative damage as components of manganese complexes and/or membrane proteins required for metabolite transport or cell wall integrity [49]. Disruption of gene DRA0276, encoding the malic enzyme that converts malate to oxaloacetate in the glyoxylate bypass and TCA cycle, might also contribute to lower the D. radiodurans capacity to deal with ROS generated by γ-irradiation. The malic enzyme has been shown to be involved in the NADH-to-NADPH conversion cycle used by Pseudomonas fluorescens to counteract oxidative stress [48].

A significant number of radiosensitive mutants were inactivated for genes involved in various metabolic functions such as energy production and conversion, amino acid transport and metabolism of coenzymes and cofactors (Table 1). In particular, inactivation of DR0342, DR0347, DR1950, DR2620 genes will greatly impair electron transfer through the respiratory chain as well as the associated ATP production. A failure to respond to the increased demand of energy during cell recovery could explain the IR sensitivity of these mutants. Likewise, protein synthesis which is essential for cell survival after irradiation [54] might be disturbed in another class of IR sensitive mutants such as DR0315, DR2151, and DR2300 affected in translation activity.

Given the γ-radiation-induced production of intracellular ROS, we assessed whether the observed radiosensitive phenotype was primarily due to ROS sensitivity or not, by testing 158 mutants for their sensitivity to hydrogen peroxide stress (Table 1 and S1 Table) as described in Material and Methods and Fig 2. The few mutants found sensitive to oxidative stress fall into two categories: (i) the expected one, including the katA mutant (Table 1 and Fig 2; [55,56]), the irrE mutant, previously shown to exhibit significantly reduced catalase activities [57], and the mutant inactivated for DR2518 encoding a DNA-damage sensor kinase (RqkA) involved in DNA repair and whose kinase activity is stimulated in vitro by the antioxidant pyrroloquinoline-quinine (PQQ) [46] (ii) the previously undescribed mutants including regulatory mutants inactivated for the transcription factors DR0265 (Fig 2) and DdrI (Table 1), and mutant for DR0679, encoding a putative small nucleotidyltransferase (Table 1). In addition, several less γ-rays sensitive mutants were found sensitive to hydrogen peroxide. These include the DR1131 mutant, inactivated for the hemZ gene involved in heme biosynthesis, the mutant for DR1207 of unknown function, the DR2374 mutant, inactivated for a ribonucleotide reductase of archeal type, and the mutant for DR2417m, encoding DncA, a novel essential β-CASP family nuclease contributing to the radiation resistance of D. radiodurans ([58] and S1 Table). The low fraction of oxidative stress mutants identified in our library may also be due to the non-viable status of the Tn5-insertions into the genes essential for oxidative stress recovery in response to high level of radiations.

This broad distribution of the inactivated genes among diverse functional categories confirms that multiple pathways participate to the extraordinary radiation resistance of D. radiodurans. In addition to known genes involved in repair of DNA damage or in regulation, we have identified new genes whose involvement in IR resistance was not previously reported or suspected. Inactivation of six of these genes, resulted in a highly γ-radiation sensitive phenotype (with a survival to γ-rays of <1% at 15 kGy; Table 1). Among these six genes, we targeted the two putative D. radiodurans loci DR0007 (encoding a di-adenylate cyclase homolog) and DR2462 (encoding an RNase Y homolog), as well as DR0265 (likely encoding a transcription factor of unknown function). A more complete functional characterization of knockout mutants of these three genes is described below.

Cyclic di-AMP (c-di-AMP) contributes to radioresistance of D. radiodurans

One of the radiosensitive mutants previously undescribed was inactivated for the DR0007 locus. This gene encodes a homolog of CdaA (37% amino acid identity with Bacillus subtilis YbbP, renamed CdaA), one of the three diadenylyl cyclase (DAC) enzymes found in B. subtilis, that catalyzes signalling nucleotide c-di-AMP synthesis [59]. The second gene, DR0008 encodes a homolog of CdaR, which stimulates the diadenylate cyclase activity of CdaA in B. subtilis [59]. Sequence analysis of this locus with FGENESB (www.softberry.com; [60]) suggests that DR0007, DR0008 and DR0009 are within a predicted operon. Accordingly, these three genes showed highly significant correlation of their expression pattern, based on transcriptome analysis [20].

To confirm the importance of these proteins for D. radiodurans recovery after irradiation, several single and double deletion mutants were constructed. Homogenotes of the ΔDR0007 mutant were easily obtained on selective medium and did not display any growth defect under standard cultivation conditions (same doubling time as those of the wild type R1 strain), indicating that DR0007 gene is not essential for cell viability under our culture conditions. The survival of ΔDR0007 was decreased by a factor of 7- and 21-fold after γ-irradiation at 15 and 20 kGy, respectively, when compared to the wild type strain. Ectopic expression of DR0007 gene alone did not fully restore the wild type radiation resistance (Fig 3A). Although we did not find insertions in DR0008 in our initial screening, we constructed a DR0008 deletion mutant to completely inactivate the gene. The ΔDR0008 mutant displayed the same γ-radiation sensitive phenotype as the ΔDR0007 mutant, and again ectopic expression of DR0008 did not restore the wild type phenotype (Fig 3A). The double mutant ΔDR0007 ΔDR0008 showed the same survival rate than those of the single mutants (Fig 3A). In contrast, ectopic expression of both of these genes in the double mutant was sufficient to restore the wild type radiation resistance (Fig 3A). These results suggest that DR0007 and DR0008 are functionally linked and that a coordinated expression of the two genes is required to fulfill their role in radioresistance. To examine whether c-di-AMP could complement the radiosensitive phenotype of ΔDR0007 and ΔDR0008, exogenous c-di-AMP was added immediately after exposure to γ-radiation, in the presence of polyamines to favor c-di-AMP uptake according to the procedure of Oppenheimer-Shaanan et al. [61]. Nevertheless, addition of c-di-AMP did not restore radiation resistance in the single and double mutants. This may be due to the use of polyamines which decrease cell survival even in the wild type (by a factor of 3.5 at 10 kGy and 4 at 15 kGy). Finally, deletion of DR0007, DR0008 or both only slightly sensitizes the cells to MMC (Fig 4A) and to UV light (Fig 5A). Deletion of DR0009 gene did not affect resistance to γ-irradiation, but we cannot exclude that DR0025 that shares 35% identity and 53% similarity with DR0009 may functionally complement the deletion of DR0009. This putative redundancy of function may explain why we did not find insertion mutants inactivated for DR0009 in our initial screening.

Fig 3

The D. radiodurans mutants deleted for DR0007, DR0008 or both (A), DR0265 (B) and DR2462 (C) show increased sensitivity to γ-irradiation.

Bacteria were exposed to γ-irradiation at doses indicated on the abscissa. Symbols: (A) wild-type (blue squares), ΔDR0007 (red closed triangles), ΔDR0008 (green closed circles), ΔDR0007ΔDR0008 (yellow closed diamonds), ΔDR0007/p11559-DR0007 + (red open triangles), ΔDR0008/p11559-DR0008 + (green open circles), ΔDR0007 ΔDR0008/p11559-DR0007 + DR0008 + (yellow open diamonds). (B) wild-type (blue squares), ΔDR0265 (purple closed triangles), ΔDR0265/p11520-DR0265 + (pink open triangles). (C) wild-type (blue squares), ΔDR2462 (brown closed circles), ΔDR2462/p11520-DR2462 + (ochre open circles).

Fig 4

Deletion of DR0007, DR0008 or both (A), DR0265 (B), or DR2462 (C) sensitizes D. radiodurans to MMC.

Bacteria were grown in TGY2X liquid medium to A650 = 1, serially diluted and dilutions were spotted onto TGY agar plates supplemented or not with MMC at the indicated doses, and supplemented with spectinomycin for strains harboring derivatives of p11559 or p11520 plasmids.

Fig 5

ΔDR0265 and ΔDR2462 are sensitive to UV but not ΔDR0007 and ΔDR0008.

Bacteria were grown in TGY2X liquid medium to A650 = 1, serially diluted and dilutions were spotted onto TGY agar plates subsequently exposed to UV-irradiation at the indicated doses.

To determine whether the increased sensitivity to ionizing radiation of the ΔDR0007 mutant was due to an altered DNA repair capacity after irradiation, we measured by pulse-field gel electrophoresis the kinetics of DNA DSB repair in the ΔDR0007 mutant exposed to 3.8 kGy γ-irradiation. The ΔDR0007 mutant reassembles an intact genome as fast as the wild type strain after exposure to γ-irradiation, indicating that its radiosensitivity is unrelated to a defect in DNA DSB repair.

Interestingly, identification, in our screening for radiosensitive mutants, of a putative c-di-AMP synthesizing enzyme, the DR0007 protein, and its potential positive regulator, the DR0008 protein suggests the contribution of c-di-AMP signalling in the extreme radiation resistance of D. radiodurans. c-di-AMP is a second messenger recently discovered in bacteria, and involved in the control of diverse cellular pathways (for reviews, see [62-64]) such as regulation of fatty acid synthesis [65], response to cell wall stress [66-68], regulation of potassium transport [63,69]. In addition, the common riboswitch class ydaO has been recently identified as receptors for c-di-AMP to control different biological processes [70]. As most of the bacterial species, D. radiodurans encodes a single DAC (diadenylate cyclase) domain-containing protein, the DR0007 product, while B. subtilis harbors three specialized c-di-AMP synthases: DisA (DNA Integrity Scanning protein) which is involved in coupling DNA integrity with progression of sporulation [61,71], CdaA and CdaS (which is sporulation-specific). In contrast to DisA, the DR0007 protein does not contain any DNA binding domain. Although we showed that both DR0007 and DR0008 proteins are required for radioresistance, the specific set of receptor and effector proteins of the c-di-AMP signaling system in D. radiodurans remain to be identified.

Identification of a new transcription factor, DR0265 involved in extreme radioresistance of D. radiodurans

Our screen also identified a new radiosensitive mutant inactivated for DR0265, encoding a putative transcription factor belonging to the GntR family. This protein contains the C-terminal effector-binding domain UTRA (UbiC transcription Regulator Associated domain) of the HutC subfamily [72]. While the repertoire of HTH-containing proteins identified in D. radiodurans reflects the diversity of prokaryotic transcriptional regulators [73], only two members of the HutC subfamily (DR0265 and DRA0211) are encoded by D. radiodurans (versus 24 HutC-like regulators in Streptomyces coelicolor, 6 in E. coli and 7 in B. subtilis).

Compilation of palindromic cis-acting elements recognized by regulators of the GntR family has identified the sequence 5’GT-N(1)-TA-N(1)-AC 3’ as the cis-element consensus for the HutC subfamily [72]. Members of this subfamily control various biological processes (antibiotic production, sensing of nutritional status, growth, proliferation and development). In Pseudomonads, Klebsiella and Brucella, HutC acts as a transcriptional repressor of the Hut system responsible for histidine utilization (for review, see [74]). In D. radiodurans, the Hut operon (DRA0151–DRA0147) is preceded by the RDRM sequence, and therefore may belong to the predicted radiation response regulon [14]. Nevertheless, it is unlikely that transcriptional regulation of the hut operon of D. radiodurans is mediated by a HutC-type factor for the following reasons: (i) the consensus sequences for HutC subfamily are not found upstream of the D. radiodurans hut operon. (ii) the organization of the D. radiodurans hut operon shows similarity to those of Corynebacterium resistens with an adjacent gene encoding a transcription regulator of the IclR family (DRA0152), and this IclR factor has been shown to activate the hut operon in C. resistens [75], suggesting a similar regulation of the Hut system in both organisms.

Deletion of DR0265 sensitizes cells to γ-irradiation, as shown by the 7- and 44-fold decrease of ΔDR0265 survival at doses of 15 and 20 kGy, respectively, when compared to those of the wild type strain (Fig 3B). Trans expression of the DR0265 gene under control of its own promoter restored the γ-ray resistance of the mutant strain to the wild type level (Fig 3B), confirming that the radiosensitivity of ΔDR0265 was solely due to the absence of the DR0265 protein. ΔDR0265 bacteria were also sensitive to UV light (Fig 5B), but only slightly sensitive to MMC (Fig 4B). As previously observed with ΔDR0007, the kinetics of DNA DSB repair of ΔDR0265 mutant shows no delay in restoration of intact genomic DNA after γ-irradiation compared to the wild type. Interestingly, ΔDR0265 bacteria were shown to be among the rare mutants found sensitive to hydrogen peroxide (Table 1, Fig 2), suggesting that DR0265 may be involved in the response to ROS but the targets of this putative regulator remain to be discovered.

Involvement of a putative RNase Y in D. radiodurans extreme radioresistance

Another gene identified as important for radioresistance, DR2462, encodes a homolog of the RNase Y recently discovered in B. subtilis [76-78], as a key endoribonuclease for mRNA turnover, with an important general role in synthesis of components involved in DNA replication, iron metabolism and the cell envelope and cell wall [79,80].

To confirm the role of the DR2462 protein in the D. radiodurans radioresistance, we have constructed a ΔDR2462 mutant. Homogenotes of the ΔDR2462 mutant were easily obtained. Bacteria devoid of the DR2462 protein were sensitive to γ-irradiation as shown by the strong decrease of their survival rate compared to those of the wild type (38-fold and 3021-fold decrease at 15 and 20 kGy, respectively) (Fig 3C). They were also moderately sensitive to UV light (Fig 5C) while they showed a wild type resistance to MMC (Fig 4C). To prove that the absence of the DR2462 protein is solely responsible for the γ-ray sensitivity, the DR2462 gene including its natural promoter and its putative transcription terminator was cloned into plasmid p11520 and the resulting plasmid p13563 was used to transform ΔDR2462 bacteria. The resulting transformants recovered the typical wild type survival following exposure to γ-rays, even at a massive dose of 20 KGy, confirming the requirement of DR2462 for the radiation resistance (Fig 3C).

The kinetics of DNA DSB repair in the ΔDR2462 mutant exposed to 3.8 kGy γ-irradiation was measured by pulse-field gel electrophoresis. Cells devoid of DR2462 protein showed a delay shorter than one hour in the restoration of an intact genome after irradiation compared to the wild type bacteria (Fig 6B) but it seems that replication did not restart immediately after reconstitution of the genome (Fig 6B), a delay that was also observed for cell division restart (Fig 6A).

Fig 6

The ΔDR2462 bacteria show an increased delay in cell division and in reconstitution of genomic DNA after γ-irradiation.

A. Growth delay after irradiation. Wild type (blue squares) and ΔDR2462 (brown circles) bacteria were exposed (filled symbols) or not (open symbols) to γ-irradiation at a dose of 3.8 kGy, diluted in TGY2X to an A650 of 0.3 and incubated at 30°C. At different times after irradiation, aliquots were taken to measure the number of viable cells per mL. B and C. Kinetics of restoration of genomic DNA. Bacteria were treated as in (A). DNA agarose plugs were prepared at the indicated post-irradiation times and digested with NotI prior to analyses by PFGE. B: wild type; C: ΔDR2462.

In B. subtilis, depletion of RNase Y (YmdA) resulted in an aberrant distribution of cell lengths, with a few unusually longer cells and many short, almost spherical, cells reminiscent of minicells [81,82]. It was proposed that this phenotype might be related to increased concentration of dnaA transcript [80], since overexpression of DnaA has been linked to aberrant changes in cell shape [83]. D. radiodurans mutant bacteria devoid of DR2462 protein grew normally and did not show dramatic alterations of their morphologies when they were observed by microscopy after nucleoid and membrane staining, except that they are slightly smaller than wild type cells (Fig 7), suggesting a more modest role of RNase Y in D. radiodurans cell viability than those played by RNase Y in B. subtilis.

Fig 7

Cell morphology in D. radiodurans deleted for DR2462 observed by microscopy.

D. radiodurans cells from the wild type strain (top panels) and from ΔDR2462 strain (bottom panels) were grown to OD650 = 0.3. Left panels: Nomarski interference contrast (DIC). Middle panels: membrane staining (FM4-64). Right panels: DNA staining (DAPI). All pictures are at the same scale (bar = 2 μm).

Conclusion

As indicated by [84], broad genetic screens to identify all processes contributing to radiation resistance are very difficult to perform in Deinococcus, due to its multi-genomic status. For these reasons, they decided to identify genes involved in radioresistance in an E. coli strain exhibiting levels of radiation resistance approaching that of D. radiodurans [85]. They used a Transposon Directed Insertion Sequencing (traDIS) strategy to locate the transposon insertion and to compare the frequency of each insertion within an un-treated population and a population subjected to repeated exposures to ionizing radiation. They identified 46 candidate genes that appear to have a significant role in survival after exposure to ionizing radiation [84]. Here, using a more classical strategy taking into account the multi-genomic status of D. radiodurans, we identified 37 genes and two intergenic sequences highlighting the importance of DNA repair, stress response, translation but also energy production, carbohydrate metabolism, membrane transport and poorly characterized or uncharacterized proteins in Deinococcus radioresistance. Among the 46 genes identified as involved in E. coli radioresistance by Byrne et al ([84]), we found 9 D. radiodurans homologs involved in DNA metabolism (recN, recG, recR, recF, uvrA, uvrB, uvrC, topA) or with uncharacterized functions (DR1167 homolog of the E. coli yabA). We also found an insertion located in the intergenic region between DR2614 and DR2613 (the latter encoding a homolog of the E. coli yab1 gene). Interestingly, we found the key genes involved in regulation of the DNA damage response, most being specific to the Deinococcaceae. We more extensively characterized three mutants not found in E. coli as contributing to radioresistance, impaired in a putative transcriptional regulator, a putative protein of the RNase Y family and a putative protein proposed to be involved in c-di-AMP synthesis, all being never described to date to participate to Deinococcus radioresistance. All the genes identified in E. coli [84] and, here in D. radiodurans, as participating to radioresistance reinforce the idea that a complex network involving efficient DNA repair, protein protection against oxidation, and tightly coordinated combination of diverse metabolic and regulatory pathways, is the key of bacterial radioresistance.

Materials and Methods

Bacterial strains, media and growth conditions

Bacterial strains are listed in S2 Table. E. coli strain DH5α was the general cloning host and strain SCS110 was used to propagate plasmids prior to introduction into D. radiodurans via transformation [86]. To produce p13554 in E. coli, we used XL1Blue that expresses lacI on the F’ plasmid to avoid the toxicity due to overproduction of Tn5 transposase as described previously [32]. All D. radiodurans strains were derivatives of strain R1 ATCC 13939. TGY2X liquid medium and TGY plates [87] were used for D. radiodurans growth and Luria-Bertani (LB) broth for E. coli. Media were supplemented with the appropriate antibiotics used at the following concentrations: spectinomycin 40 μg mL-1 for E. coli and 75 μg mL-1 for D. radiodurans; hygromycin 50 μg mL-1 for E. coli and 50 to 100 μg mL-1 for D. radiodurans, kanamycin 25 μg mL-1 for E. coli and 6 μg mL-1 for D. radiodurans. When necessary, expression of Tn5 transposase was induced by adding 1 mM IPTG in media. Transformation of D. radiodurans with genomic DNA, PCR products, or plasmid DNA was performed as previously described [87].

Construction of deletion mutants in D. radiodurans

To construct each mutant deleted for a given gene, the locus of interest was replaced with the appropriate antibiotic resistance cassette (either the HygR resistance cassette expressed from the P promoter or the TetR resistance cartridge expressed from the PgroEL promoter) using the tripartite ligation method [88]. The deletion mutants generated in this way and used in this study are: ΔpprAΩhph (for use as a control mutant in treatment with DNA damaging agents), ΔDR0007Ωhph, ΔDR0008Ωhph, double mutant ΔDR0007–DR0008Ωhph, ΔDR0009Ωhph, ΔDR0265Ωhph, ΔoxyRΩhph, ΔoxyR2ΩtetA (for using as control mutants in screening for sensitivity to H2O2) and ΔDR2462Ωhph. The double mutant ΔoxyRΩhph ΔoxyR2ΩtetA was constructed by transforming the ΔoxyR2ΩtetA single mutant with genomic DNA from the ΔoxyRΩhph mutant. See S3 Table for oligonucleotides used for strain construction. The genetic structure and the purity of the resulting mutant strains were verified by PCR. Oligonucleotides used for diagnostic PCR and sequencing are available upon request.

DNA manipulations

Plasmid DNA was extracted from E. coli using the NucleoSpin Plasmid miniprep kit (Macherey-Nagel). D. radiodurans chromosomal DNA was isolated as described previously [89]. PCR reactions were carried out with Phusion DNA Polymerase (Thermo Scientific) to amplify fragments subsequently used for cloning or with GoTaq Flexi DNA Polymerase (Promega) for all other applications. PCR products were purified using the NucleoSpin Gel and PCR Clean-up kit (Macherey-Nagel). To analyze the homozygous/heterozygous status of the mutants, diagnostic PCR were performed using appropriate pairs of primers encompassing the Tn5 insertion site. Oligonucleotides used for all these diagnostic PCR are available upon request.

Construction of Tn5 delivery vector, p13554

The lacI gene was amplified by PCR using primers ForLacI and RevLacIBg and the plasmid pTRC99A (Pharmacia, see also S2 Table) as template. The PCR fragment was digested by BglII, and ligated into the BamHI site of p13841, resulting in the plasmid p13537 expressing lacI in the reverse orientation to P. The tnp gene encoding the hyperactive Tn5 transposase was amplified by PCR with primers Tnp5UP and Tnp5Dra and plasmid pWH1891 ([90]; see also S2 Table) as template. After cleavage with NdeI and DraI, the PCR fragment was ligated into plasmid p13537, generating plasmid p13545, expressing the transposase under the control of P. The mini-Tn5-HygR transposon was then amplified by PCR with primers MEUpBst and MEDnBgl and plasmid p12625 as template. After digestion with Bst1107I and BglII, the PCR fragment containing ME (mosaic ends derived from Tn5) was inserted into p13545 to generate plasmid p13547.

To use the Tn5-delivery system into D. radiodurans GY10973 strain expressing an additional chromosomal copy of lacI under the control of P promoter, we have deleted lacI from p13547 as follows: p13547 was digested with DraI and PscI and the 10531-bp fragment was ligated to a linker containing the DraI and PscI ends and an internal SalI site, to generate plasmid p13554 (see S3 Table for oligonucleotides used for p13554 construction and Fig 1).

Plasmids used for complementation analyses

Please see S2 and S3 Tables for details of construction of plasmids p14726, p14729, p14731, p14728, and p13567 (used for complementation analyses of mutants ΔDR0007Ωhph, ΔDR0008Ωhph, ΔDR0007ΔDR0008Ωhph and ΔDR0009Ωhph respectively), p13564 (for complementation analyses of mutant ΔDR0265Ωhph) and p13563 (for complementation analyses of mutant ΔDR2462Ωhph).

Mapping Tn5 insertions sites into D. radiodurans genome

Insertion mutants were mapped by the arbitrary-primed (AP)-PCR procedure [91,92], using GoTaq Flexi DNA polymerase (Promega). The first PCR round was performed in a final volume of 50μL with 1 μL genomic DNA from single HygR colonies as template. The arbitrary primer (ARB1c) was paired either with a primer specific for the 5’ end of the mini-Tn5 (Tn5-212) or with a primer specific for the 3’ end of the mini-Tn5 (Tn5-991), both at a final concentration of 0.8 μM and PCR was performed as follows: 2 min 95°C, 6 cycles of 45 s 95°C, 45 s 30°C, 1 min 30 s 72°C; 30 cycles of 45 s 95°C, 45 s 45°C, 2 min 72°C; and finally 72°C for 5 min. The second round was performed in a final volume of 50 μL with 5 μL of the purified PCR product from round 1 as template. A second arbitrary primer (ARB3) was paired either with the Tn5-166 primer (5’ end of Tn5) or with the Tn5-1055 primer (3’ end of Tn5), each at a final concentration of 0.8 μM and PCR was performed as follows: 2 min 95°C, 30 cycles (45 s 95°C, 45 s 52°C, 2 min 72°C); 72°C for 5 min. The products of this PCR were directly sequenced with the SeqRE primer (5’ end of Tn5) or with the EB89 primer (3’ end of Tn5) by Cogenics (Meylan, France). Oligonucleotides used are listed in S3 Table.

Procedure for in vivo transposition of mini-Tn5-hph into D. radiodurans

We have used the D. radiodurans GY10973 strain expressing a chromosomal copy of lacI under the control of the P promoter as a recipient, to ensure a non-toxic level of the transposase during the mutagenesis process. The isolation of Tn5 insertion mutants was performed in two steps as follows. First, the D. radiodurans GY10973 strain was transformed with the Tn5-hph delivery plasmid p13554 and transformants were selected at permissive temperature (30°C) on TGY agar supplemented with both hygromycin and spectinomycin. Ten independent HygR SpcR colonies were then pooled and resuspended into 200 μl of TGY2X. 15 μl of this resuspension was inoculated into 3 ml TGY2X supplemented with spectinomycin and incubated at 30°C with shaking to an A650 ~ 0.1 (transposition step). This culture was finally serially diluted and plated onto both TGY agar without antibiotic to quantitate the viable cells and onto TGY agar supplemented with hygromycin and incubated at a non-permissive temperature (37°C) to simultaneously select for Tn5-hph insertion mutants and cure the delivery plasmid.

Calculation of the number of predicted inactivated ORFs in the Tn5 insertion mutant library

Given that Tn5 insertion is random, the number of independent single-gene insertions in D. radiodurans required to inactivate 99% of the genes can be calculated using the following formula [43]: P = 1 – (1 - [x/g])n, where P = probability of finding one transposon insertion within a given gene (0.99), x = average length of a Deinococcal gene (1011 bp), g = D. radiodurans genome size (3284156 bp), and n = number of independent insertion mutants. The number of genes inactivated among the 6207 insertion mutants analysed in our screening was estimated, based on Poisson’s law: P (k) = (λk/k!)e-λ, where k = number of insertions within a given gene and λ = the average number of insertions per gene. Given that the D. radiodurans genome encodes 3,195 predicted protein-encoding genes [73], one can estimate λ = 6,207/3,195 = 1.9427 insertions per gene. Therefore, the probability of having no insertions within a given gene is P(0) = (λ0/0!)e-λ = e-λ = 0.1433, indicating that approximately 14% of the genes were not inactivated among the 6207 mutants.

Library screening for sensitivity to DNA damaging agents (γ, UV-irradiation and MMC treatment)

First, the HygR SpcS insertion mutants selected at non-permissive temperature (37°C) were homogenotized (for non-essential genes) as follows: the colonies were arrayed to 96-well microtiter plates containing 100 μL of TGY2X broth supplemented with hygromycin (50 μg/mL) per well, grown statically at 37°C for one day, followed by three serial replica always on hygromycin supplemented TGY agar plates. The purified individual mutant clones (homogenotized or partially homogenotized) were replica plated on TGY-agar plates containing hygromycin and exposed either to either γ-rays (at a dose of 7500 Gy) or UV-rays (at a dose of 600 J m-2) or onto TGY plates supplemented with mitomycin C (at 30 ng/mL). To confirm their radiosensitive phenotype, the candidates arising from the first screening were then grown overnight in TGY2X supplemented with hygromycin and 5 μl of undiluted or 1/10th diluted culture were spotted on plates subsequently γ-irradiated (7500 Gy) or UV-irradiated (600 J m-2) or onto plates supplemented with MMC. To validate our screening and homogenotization procedure, we have included in our screen a control mutant by transformation of the tester strain GY10973 with in vitro engineered pprA gene inactivated by insertion of the same hygromycin cassette. Lastly, to ensure that the observed radiosensitive phenotypes are genetically linked to Tn5 insertion, all candidates sensitive to DNA damaging agents isolated in the GY10973 strain were further confirmed by backcross to the wild type D. radiodurans R1 strain (see Fig 1B for screening procedure).

Screening of D. radiodurans mutants for sensitivity to H2O2 stress

All mutants sensitive to γ-rays and further confirmed by backcross were analysed for their sensitivity to hydrogen peroxide (Sigma-Aldrich) by using the disc inhibition assay as follows. 1 mL of exponential phase cultures (A650 ~ 0.2) was spread-plated onto TGY plates, and sterilized 6-mm-diameter filter paper discs (Dominique Dutscher) were placed on the agar surface. Then, 10 μL of various concentrations (50, 100, 120, 150 and 200 mM) of H2O2 (freshly diluted in 10 mM potassium phosphate buffer) was spotted onto each disc. After incubation at 30°C for 2 days, the diameters of the growth inhibition zones were measured. The mutants were classified into four categories depending on the diameter of growth inhibition area as described in Fig 2.

Survival of D. radiodurans to different DNA damaging agents

For all experiments, the colonies were counted after 3–4 days incubation at 30°C.

Gamma irradiation. Bacteria were grown in TGY2X or in TGY2X media supplemented with spectinomycin when they contained a plasmid or with 10 mM IPTG and spectinomycin when they contained plasmids p11559 or derivatives to an A650 ≈ 1. The cultures were concentrated six times in TGY2X and irradiated on ice with a 137Cs irradiation system (Institut Curie, Orsay, France). Following irradiation, diluted samples were plated on TGY plates, or TGY plates supplemented with spectinomycin when bacteria contained plasmid p11559 or derivatives and incubated at 30°C before the colonies were counted.

UV irradiation. Bacterial cultures at an A650 = 1 were serially diluted in TGY and plated on TGY agar plates (or TGY plates supplemented with spectinomycin when bacteria contained plasmid p11559 or derivatives), subsequently exposed to UV light at a dose rate of 3.5 J/m2/s and incubated at 30°C.

MMC treatment. Bacterial cultures at an A650 = 1 were serially diluted in TGY and plated on TGY plates (or TGY plates supplemented with spectinomycin when bacteria contained plasmid p11559 or derivatives) supplemented with MMC at final concentrations of 40 and 60 ng mL-1 and incubated at 30°C.

Kinetics of DNA repair measured by pulse-field gel electrophoresis

Non-irradiated or irradiated (3,800 Gy) cultures were diluted in TGY2X to an A650 = 0.3 and incubated at 30°C. At different post-irradiation recovery times, samples (5 ml) were harvested to prepare DNA plugs as described [93]. The agarose embedded DNA plugs were digested for 16 h at 37°C with 60 units of NotI restriction enzyme. After digestion, the plugs were subjected to pulsed field gel electrophoresis for 28 hours at 10°C using a CHEF MAPPER electrophoresis system (Biorad) with the following conditions: 5.5 V / cm, linear pulse of 40 s, and a switching angle of 120° (- 60° to + 60°).

Fluorescence microscopy

The cells were fixed by adding toluene (3% final concentration) to culture aliquots, and kept at 4°C. DNA and cell membranes were stained with DAPI (40 μg/ml) and FM4-64 (50μg/ml), respectively, as previously described [24] and spotted on a thin layer of TGY2X agarose 1% for microscopy observation. The stained cells were observed using a Leica DMIRE2 microscope with a 100X objective and the appropriate fluorescence filters.

Functional categories of Tn5-inserted genes in mutants sensitive to DNA damaging agents isolated by screening onto TGY agar plates.

(PDF)

Click here for additional data file.

Master screening data of backross Tn5 insertion mutants for sensitivity to γ- and UV rays, MMC or hydrogen peroxide (H2O2).

(PDF)

Click here for additional data file.

Bacterial strains and plasmids and corresponding references.

(PDF)

Click here for additional data file.

Overview of primers used for strains construction, cloning, mutagenesis, and AP-PCR experiments.

(PDF)

Click here for additional data file.