The helicase DinG responds to stress due to DNA double strand breaks

Neisseria meningitidis (Nm) is a Gram-negative nasopharyngeal commensal that can cause septicaemia and meningitis. The neisserial DNA damage-inducible protein DinG is a helicase related to the mammalian helicases XPD and FANCJ. These helicases belong to superfamily 2, are ATP dependent and exert 5′ → 3′ directionality. To better understand the role of DinG in neisserial genome maintenance, the Nm DinG (DinGNm) enzymatic activities were assessed in vitro and phenotypical characterization of a dinG null mutant (NmΔdinG) was performed. Like its homologues, DinGNm possesses 5′ → 3′ directionality and prefers DNA substrates containing a 5′-overhang. ATPase activity of DinGNm is strictly DNA-dependent and DNA unwinding activity requires nucleoside triphosphate and divalent metal cations. DinGNm directly binds SSBNm with a Kd of 313 nM. Genotoxic stress analysis demonstrated that NmΔdinG was more sensitive to double-strand DNA breaks (DSB) induced by mitomycin C (MMC) than the Nm wildtype, defining the role of neisserial DinG in DSB repair. Notably, when NmΔdinG cells grown under MMC stress assessed by quantitative mass spectrometry, 134 proteins were shown to be differentially abundant (DA) compared to unstressed NmΔdinG cells. Among the DNA replication, repair and recombination proteins affected, polymerase III subunits and recombinational repair proteins RuvA, RuvB, RecB and RecD were significantly down regulated while TopA and SSB were upregulated under stress condition. Most of the other DA proteins detected are involved in metabolic functions. The present study shows that the helicase DinG is probably involved in regulating metabolic pathways as well as in genome maintenance.


Introduction
Neisseria meningitidis (Nm) is a Gram-negative bacterium that frequently colonizes the human nasopharynx in small children and adolescents. In the lack of bactericidal antibodies, Nm can enter the bloodstream and cross the blood-brain barrier leading to septicaemia and meningitis, respectively [1]. We are interested in how Nm cells survive on the oral mucosal surface, in the bloodstream and the meninges, where it is exposed to the host defence, including DNA damaging reactive oxygen and nitrogen species [2]. Therefore, we postulate that DNA repair pathways that promote genome stability play important roles in the survival of Nm under genotoxic stress. Many proteins involved in DNA repair pathways have multiple overlapping functions and crosstalk between these pathways has been identified [3]. However, the functions of many DNA repair proteins are still not clearly discerned. In this study, we aimed to understand the functional role of the neisserial helicase DinG in the genome maintenance of Nm under genotoxic stress.
Helicases are molecular motor proteins that unwind double stranded nucleic acids using the energy provided by ATP hydrolysis. By doing so, helicases facilitate various aspects of nucleic acid metabolism such as replication, repair, recombination, transcription, translation and splicing of RNA transcripts [4]. Helicases are classified into six superfamilies (SF1-6) based on the sequence identity among the conserved helicase motifs [5]. Helicases belonging to the SF1 and SF2 categories share a catalytic core with high structural similarity, but even within each SF the different enzymes exert distinct functions on diverse nucleic acid substrates [6].
The Escherichia coli DNA damage-inducible protein DinG (DinG Ec ) belongs to the SF2 helicases which translocates on single-stranded DNA (ssDNA) in the 5 0 !3 0 direction [7]. A recent report suggested that DinG Ec is involved in the dissolution of R-loops during replication restart following the collision of replication forks with the transcription unit [8]. It was also reported to be able to resolve intermolecular but not intramolecular G4 DNA [9]. DinG Ec is a structure-specific enzyme, related to human xeroderma pigmentosum group D (XPD), FANCJ, also known as BACH1, as well as to Saccharomyces cerevisiae Rad3 and Chl1, and Schizosaccharomyces pombe Rad15 [10]. XPD is a subunit of TFIIH, a large multiprotein complex that plays a dual role in transcription initiation and nucleotide excision repair [11]. Mutations in the human XPD helicase gene are found in patients with three inherited diseases: xeroderma pigmentosum (XP), Cockayne syndrome (CS) and trichothiodystrophy (TTD) [12]. Mutations in the gene encoding FANCJ predispose individuals to breast cancer, suggesting a tumour suppressor role for FANCJ [13][14][15]. The role of the helicase DinG in bacteria is not as well defined, and deletion or overexpression of dinG in E. coli results in poorly discernible phenotypes [7].
In order to define the biological role of DinG helicase in neisserial genome maintenance and bacterial survival, we analysed an Nm wildtype and a dinG null mutant, NmΔdinG. The fitness of these Nm strains under various forms of genotoxic stress was assessed and protein expression levels were compared. The gene encoding the Nm helicase DinG (DinG Nm ) was cloned and overexpressed, the recombinant DinG Nm protein was purified to homogeneity and its enzymatic activities were characterized. Taken together, the results show that DinG Nm responds to stress due to DNA double strand breaks and possibly is involved in metabolic pathway regulation.

DinG Nm is a conserved protein among N. meningitidis isolates
The 467 dinG variants available in PubMLST (PubMLST ID: NEIS0293) exhibit only minor variation, with the seven helicase motifs showing strong conservation (Fig 1). Among the 66 amino acid positions exhibiting variations located in the helicase motifs, only 9 are parsimonyinformative sites (S1 Fig). According to SNAP2 predictions, all variants at these nine sites impose a neutral functional effect. The sequence of DinG Nm from strain MC58 readily aligns to the sequence of DinG Ec identifying amino acid K72 in DinG Nm as equivalent to the amino acid K60 required for ATPase activity in DinG Ec [7]. The four conserved DinG cysteine residues numbered Cys-120, -194, -199 and -205 in E. coli [10], correspond to Cys-133, -209, -214 and -220 in Nm.
DinG Nm is an iron containing DNA dependent ATPase and the DinG NmK72A mutant protein is inactive The recombinant DinG Nm and DinG NmK72A (an ATPase inactive mutant) proteins were purified to homogeneity. We observed that both proteins exhibited a yellow-brown colour which is an attribute for Fe-S proteins such as XPD [17] and AddAB [18]. Using an iron chelation assay, we determined that DinG Nm contained four iron molecules per one protein molecule. The ATP hydrolysis activity of both proteins was tested in the presence of poly dT 100 as single stranded (ss) DNA cofactor. The ATP hydrolysis activity increased with increasing DinG Nm protein concentration, and nearly 100% of the ATP was hydrolysed in the presence of 400 nM DinG Nm under the conditions used (Fig 2). In the absence of ssDNA as a cofactor, DinG Nm did not show any ATPase activity, indicating that the enzyme depends on DNA for this activity A) Plotting of the similarity of DinG Nm sequences from PubMLST with sliding window sizes (w) ranging from 4 to 20 aa. Data were generated with plotcon (EMBOSS) and the average similarity for the whole sequence was set to 1. The helicase motive I, Ia, II, III, IV, V, and VI are under laid in orange colour. B) Similarity of the DinG Nm sequences taken from PubMLST plotted onto the predicted structure of DinG Nm from MC58 using default Consurf colouring with cysteins shown in black. The helicase motifs are shown as spheres and the rest of the protein is shown as cartoon.  (Fig 2A, lane 1). The mutant protein DinG NmK72A had no ATPase activity under any condition (Fig 2).

DinG Nm and DinG NmK72A bind to DNA
The binding affinities of DinG Nm and the DinG NmK72A were examined in the presence of ssDNA substrates of variable length. Both proteins bound ssDNA with similar affinity with an ssDNA of 40 nt length being sufficient for a complete shift under the conditions tested (Fig 3).
To further examine the DNA binding affinity of these proteins, a 100 nt oligomer substrate (poly dT 100 ) was incubated with increasing concentration of the proteins (Fig 4). The results showed that the ATPase mutant protein, DinG NmK72A retained its DNA binding affinity.
DinG Nm unwinds DNA in the 5 0 ! 3 0 direction To further investigate the helicase activity of DinG Nm , forked DNA substrates were used (S2 Table). Initial studies showed that more ssDNA product was generated from forked DNA  substrate (T1+B1) as the concentration of DinG Nm was increased up to 400 nM, while DinG NmK72A failed to unwind the DNA substrate even at higher concentrations ( Fig 5).
To determine the polarity of the DinG Nm -catalysed unwinding, forked DNA substrates with switched polarities were made (T8-3 0 -3 0 + B9 and T8+ B9-5 0 -5 0 ) such that the unpaired single stranded ends contain either only 5 0 or only 3 0 ends. As its E. coli homolog DinG Ec , DinG Nm also unwound only the substrate containing open 5 0 ends, showing that the unwinding activity has 5 0 ! 3 0 polarity ( Fig 6A). The helicase activity of DinG Nm was also shown to be dependent on the presence of Mg 2+ or Mn 2+ (Fig 6B). In addition, the unwinding activity of DinG Nm was observed only in the presence of ATP or dATP (Fig 6C), indicating that only the hydrolysis of ATP or dATP can be used by DinG Nm to deliver the energy needed for unwinding.

DinG Nm directly interacts with SSB Nm
To determine whether DinG Nm interacts with SSB Nm , we employed ammonium sulphate coprecipitation and microscale thermophoresis (MST). The precipitation indicated that SSB Nm readily precipitates at 150 g/l ammonium sulphate while only about ½ of DinG Nm precipitates under this conditions (Fig 7).
The combination of DinG Nm together with SSB Nm resulted in the complete co-precipitation of DinG Nm and SSB Nm . The MST assays with the SSB proteins as labelled molecules and the DinG proteins as ligands confirmed the interaction. The binding between SSB Nm and DinG Nm had a K d value of 1.36 ±0.11 μM, and the binding between SSB NmΔ8C and DinG Nm had a K d

NmΔdinG cells are sensitive to DNA intrastrand crosslinking agents
The influence of genotoxic agents on Nm survival in the absence of the helicase DinG was examined by comparing Nm wildtype with NmΔdinG. The Nm wildtype and NmΔdinG were tested for sensitivity to UV irradiation, hydrogen peroxide (H 2 O 2 ), paraquat, methyl methanosulphonate (MMS) and mitomycin C (MMC). When exposing the cells to increasing doses of UV the mutant cells survived slightly but not significantly better (IC 50 = 40 J/m 2 ) than the wildtype (IC 50 = 20 J/m 2 ) ( Fig 9A). The NmΔdinG mutant was equally sensitive to oxidative stress (H 2 O 2 and paraquat) or alkylating stress (MMS) as the Nm wildtype ( Fig 9B). The stress inflicted by MMC resulted in a significant difference in survival rates between wildtype and mutant, indicating a key role for DinG in repair of DNA double strand breaks caused by MMC ( Fig 9B). This sensitivity of the mutant was confirmed in a similar experiment using bleomycin, another DNA double strand break causing agent (S6 Fig).

Response of N. meningitidis wildtype and NmΔdinG to DNA double strand break damage
Since the MMC induced genotoxic stress led to a significantly lower survival rate of NmΔdinG, we further investigated the protein expression profiles of the Nm wildtype and mutant constitutively and under MMC stress. Quantitative mass spectrometry-based protein profiling was employed to identify differentially abundant (DA) proteins. A good overlap of detected proteins was reached (S2 Fig). When disrupting dinG in Nm (NmΔdinG), 48 DA proteins were identified with 9 proteins more and 39 proteins less abundant when compared to the wildtype (S1 Table). Under normative condition, in the NmΔdinG the proteins MinC and NuoE were found to be less abundance than in wildtype, by 3.4 and 2.1 fold, respectively. Parallel to this a more than 3 fold increase was seen in the amount of the SOS response repressor LexA homologue NMB0556. Under stress condition this protein was less abundant in the NmΔdinG cells by 1.6 fold (S1 Table). The highest number of DA proteins was observed when comparing the MMC stressed NmΔdinG to the unstressed NmΔdinG cells. We found 82 more abundant and 52 less abundant proteins in MMC stressed NmΔdinG compared to the unstressed NmΔdinG. The proteins MinC and NuoE were found to be more abundant in stressed NmΔdinG by 4.2 and 2.2 fold, respectively (S7 Table). Interestingly, 70% (27out of the 39) of the more abundance proteins in unstressed NmΔdinG (column H in S1 Table) were found to be less abundant in MMC stressed NmΔdinG (column O in S1 Table). In addition, 89% (8 out of 9) of the less abundant proteins were more abundant (columns G and P in S1 Table). This shows the general opposite tendencies of the effects of dinG deletion and MMC stress on the mutant. These opposing tendencies were confirmed by Principle Component Analysis (PCA) (Fig 10). The PCA was done using R with the package FactoMineR [19,20].
In addition to the amino acid and carbon metabolism, including 2-oxocarboxylic acid metabolism, the proteins involved in DNA metabolism are one of the most affected categories in the stressed NmΔdinG. The changes for the DA proteins involved in DNA metabolism are shown in Fig 11A. The list of these proteins includes eight less abundant proteins, namely DnaG, DnaQ-2, DnaX, DnaE, RecB, RecD, RuvA and RuvB, and the two more abundant proteins TopA and SSB (S1 Table). The DA proteins involved in electron transfer and therefore in the maintenance of the redox state of the cell are shown in Fig 11B, with AniA being a central electron acceptor up-regulated under MMC stress condition.

DinG Nm does not affect transformation or replication in N. meningitidis
Nm is naturally competent for transformation throughout its lifecycle, and this feature plays a major role in genome maintenance in Neisseria spp. Therefore, we quantified the transformation efficiency in NmΔdinG. No significant difference between wildtype and NmΔdinG in transformation was detected (S3  (Ng) strain MS11 and the NgΔdinG mutant were used to determine the chromosome equivalents and the number of active replication forks per cell. For exponentially grown cultures, no difference in the number of active replication forks was detected, and the total DNA contents of the wildtype and mutant cells were similar (S4 and S5 Figs). Upon treatment with rifampicin (RIF) and cephalexin (CPX), the DNA content of the Ng MS11 wildtype and NgΔdinG cells did not differ significantly (S4C and S4D Fig).

Discussion
The bacterial helicase DinG is homologous to the archaeal and eukaryotic helicases XPD/Rad3 and FANCJ. They all belong to the SF2 helicase with an intrinsic 5 0 ! 3 0 helicase activity and a characteristic Fe-S binding domain [6]. The role of eukaryotic XPD in nucleotide excision repair is well recognized [21,22], but the role of the bacterial DinG and archaeal XPD in DNA repair is not well understood. In this study, we characterized the neisserial helicase DinG to understand its role in DNA repair, transformation and replication.
The Expression of dinG Ec was found to be induced by MMC, which induces DSBs due to interstrand cross-links in DNA [23]. The stress experiments presented here showed that loss of dinG in Nm severely reduces the survival of the cells when exposed to MMC. On the other hand, the survival of NmΔdinG cells was not significantly reduced due to stress inflicted by UV, H 2 O 2 , paraquat or MMS. We suggest that DinG Nm plays a role in maintaining genome integrity, especially when encountering severe forms of genotoxic stress, such as DSBs.
Our comparative quantitative proteomics revealed differences in the expression level of proteins in NmΔdinG compared to the Nm wildtype. Eighteen of the thirty nine proteins more abundant upon dinG deletion are involved in central housekeeping functions, such as amino acid, carbon and nucleic acid metabolisms. This suggests that there might be a general change in the regulation of the metabolism in response to deletion of dinG.
Opc, a phase variable outer membrane protein and an effective invasin for human endothelial cells, was down regulated when comparing the mutant with the wildtype. This is pointing to a random phase switch in the gene which happened during the cloning of the mutant and can also be attributed to the selection of an opacity negative colony during cloning. The generally lower expression of Opc in the mutant is therefore a none-stress related result [24][25][26].
MinC was found to be less abundant in the mutant compared to the wildtype by 3.4 fold. MinC is a cell-division inhibitor preventing the formation of the Z ring and is essential for proper cell division in cocci [27,28]. Two proteins involved in the electron transport chain (NuoE, CycP) were found to be more abundant in mutant than the wildtype. Only two DNA metabolic proteins, namely DnaG and RecD, were found to be more abundant in the mutant. The DnaG, a DNA primase, is involved in short strand RNA synthesis during DNA replication. RecD is a helicase in the enzyme complex RecBCD, which is involved in conjugal and transductional recombination, DNA repair, and degradation of foreign DNA [29]. Thereby, in the absence of DinG, RecD is the only DA protein found having a comparable and possibly compensatory activity.
Co-transcription of genes in an operon could lead to a symmetric differential expression of proteins. For three out of seven genes in operons (S1 Table) this symmetry was found [30]. Interestingly, this was not the case for minC and oxyR although OxyR negatively regulates the promoter in the minCDE-oxyR gene cluster [31]. OxyR, a regulator for the expression of catalase [32], is overexpressed in NmΔdinG under the MMC stress condition. Although an association of MMC with oxidative stress (OS) was shown in eukaryotic systems [33], a homologue to xanthine dehydrogenase, an enzyme responsible for oxygen radical generation [34,35], was not identified in Neisseria sp. This still leaves the possibility of another enzyme exerting this activity, possibly an oxidoreductase involved in the electron transport chain [36]. With the OS signal provoked by MMC the increase in OxyR would lead to an activation of the catalase gene (kat) and thereby an alleviation from the OS.
Cytochrome c 0 was less abundant in the NmΔdinG mutant while AniA, a nitrite reductase, was more abundant under MMC stress. Both proteins participate in the electron transport chain and are involved in the defence against nitrosative stress [37,38].
Three of the DA proteins (BfrA, BfrB, GshB) take part in the defence against oxidative stress and OxyR is a regulator of the oxidative stress response [39]. In addition many proteins involved in the electron transport, which ultimately feeds into the production of ROS, are differentially abundant (S1 Table). While iron is an important co-factor for proteins involved in the electron transport, only a few (4) proteins directly regulated by Fur were DA with NMB1395, an alcohol dehydrogenase, being 3.8 times more abundant in stressed NmΔdinG compared to unstressed NmΔdinG (S1 Table) [40]. Although no unilateral direction of up-or down-regulation is visible, all these changes may reflect the general adaption of the system to cope with ROS and also to reduce the overall oxidation state of the cell. The general change in regulation of genes when the Nm bacterium is under stress, like in the blood, comprise that the genes involved in energy metabolism are up-regulated and house-keeping genes are downregulated [41].
Bacteria respond to DNA damage by mounting a coordinated cellular response, governed by the RecA and LexA proteins, called SOS response, by down-regulating more than 40 unlinked genes [42]. The inactivation of LexA by activated RecA-ssDNA filaments causes the expression of genes that increases DNA damage repair and tolerance [43]. The Nm-genome lacks the conserved LexA binding site (SOS box) [44]. However, recently a LexA orthologue (GenBank ID: NGO1427) has been identified in Ng and is found to be modulated by oxidation [45]. Here, the Nm orthologue (GenBank ID: NMB0556) was found to be more abundant (by 3 fold) in the unstressed NmΔdinG, but under MMC stress it was less abundant (by 1.6 fold). At present, it is not known how this orthologue regulates the SOS response under MMC stress in the absence of the canonical SOS box, but our results point to a significant impact.
Looking at the DA proteins detected when comparing stressed and unstressed NmΔdinG we found that two proteins, TopA and SSB, were more abundant in the stressed NmΔdinG (S1 Table). The function of topoisomerase I is required for efficient transcriptional activation of the recA and dinD1 promoters during the E. coli SOS response to trimethoprim or MMC [46]. Recently, it was also shown that topoisomerase I interacts with RecA with ATP as an enhancing factor [47]. SSB is known to bind ssDNA and interacts with more than 15 other proteins [48]. An in vitro study revealed that E. coli SSB was able to form a stable protein complex with DinG and stimulate its DNA helicase activity [49]. On the other hand a study in B. subtilis failed to see interactions between DinG and SSB in pulldown experiments [50]. Our analyses revealed that DinG Nm interacts with SSB Nm with a value that can be considered to be in the range between stable and transient interactions [51].
Eight DNA metabolic proteins were less abundant in MMC stressed NmΔdinG cells when compared to the unstressed mutant (S1 Table). Among them, the gene products of dnaE, dnaX and dnaQ-2 are subunits of the DNA polymerase III holoenzyme. An interaction network analysis in E. coli revealed that Pol III subunits and clamp loader co-purify together with TopA and PriA helicase [52]. Several studies have reported the direct binding of the E. coli SSB to the Pol III holoenzyme [53][54][55]. This shows an interaction network of proteins involved in diverse biological process and has been proposed to be important for stabilization of the replication machinery and for facilitating Okazaki fragment replication initiation [53][54][55]. Other less abundant DNA metabolic proteins in MMC stressed NmΔdinG cells were RuvA, RuvB, RecB, RecD and DnaG. RuvA and RuvB together with RuvC form a complex and play important roles in the homologous genetic recombination and the recombinational repair of damaged DNA [56]. In E. coli, the ruvA and ruvB genes are SOS inducible and regulated by LexA repressor [57]. DnaG and RecD are two DNA metabolic proteins that are inversely regulated in stressed and unstressed NmΔdinG. These observations reveal that, especially under stress, DinG is involved in the recombinational repair pathway as the lack of it leads to a deregulation of compounds of the DSB repair system.
Only a few studies tried to pinpoint the role of DinG helicase in DNA repair pathways, and suggested that its possible involvement in replication and recombinational repair [8,10,58]. The study by Boubakri et al. suggested that DinG Ec enables the replication fork to move along transcribed DNA by unwinding R-loops or displacing the transcribing RNA polymerase in conjunction with UvrD or Rep helicase [8]. DinG Nm might play a similar role or might help remove other structures that inhibit progression of DNA replication.
The DinG helicase and its homologues contain four conserved cysteine residues. These residues are ligands for a Fe-S cluster, which is redox active and DinG Ec is fully active only under oxidizing conditions [59]. We show here that DinG Nm also contains four iron molecules per protein molecule for a [4Fe-4S] cluster. It was recently suggested that helicases with an Fe-S binding domain might unwind non-conventional DNA or RNA structures or displace proteins from cognate DNA binding sites, thus relieving stalled DNA replication or transcription complexes [60]. This ability would make DinG Nm into an oxidation state dependent transcriptional regulator.
In summary, our data indicate an important role for DinG Nm in DSB repair. Nm cells lacking DinG helicase show, compared to the wildtype cells, the de-regulation of many proteins. Under MMC stress this de-regulation is inversed and over compensated. The regulation includes mainly proteins which are involved in energy, amino acid and nucleotide metabolism, and DNA replication and repair. Future studies on the functions and interactions of DinG Nm , also focusing on the redox state of the cells under stress, will reveal more insights into its cellular roles and thus show the part this helicase takes in DNA repair pathways preserving the integrity of the genome.

Bioinformatics analysis
Sequence data for dinG was obtained from the PubMLST database [61] and from NCBI [17]. The dinG Nm nucleotide sequences were searched for occurrences of the DNA uptake sequence (DUS) and single nucleotide polymorphisms, and in the deduced DinG Nm amino acid sequence structural helicase motifs were identified by alignment with the sequence for DinG Ec . Sequence conservation was calculated using plotcon from the EMBOSS package [62] and visualized on the protein structure using ConSurf [63]. The multiple sequence alignment grid profile was extracted using UGENE [64] and the prediction of functional effects of sequence variants was done with SNAP2 [65].

Bacterial strains and growth conditions
Neisseria strains were grown on GC agar plates or in liquid GC medium supplemented with IsoVitaleX at 37˚C and 5% CO 2 . When required, kanamycin at a final concentration of 100 μg/ml was added. E. coli strains were grown at 37˚C on LB plates or in LB medium containing kanamycin (50 μg/ml). The bacterial strains and plasmids employed in this study are listed in Table 1.
Cloning of Neisseria meningitidis SSB, SSB NmΔ8C , dinG Nm and dinG NmK72A The genes were PCR amplified from genomic DNA isolated from N. meningitidis MC58 using the appropriate primers (S2 Table). The PCR product was digested with the restriction enzymes XbaI and SalI and ligated into a identically digested pET-28b(+) vector (Novagen). The resulting plasmid pSB13 coding for a protein with a C-terminal 6xHis-tag was transformed into E. coli NiCo21 (DE3) (NEB) containing the plasmid pLysS (Novagen). The point mutation in the ATP binding motif (K72A) was created in pSB13 using site-directed mutagenesis with the primers SF177 and SF178 resulting into pSAF92. Cloning of SSB Nm has been described earlier [74]. Primers SF275 and SF276 were used to amplify the vector pSAF104 using the vector pEH1 as a template to obtain the ssb NmΔ8C construct.
The sequences of all constructs were verified using appropriate sequencing primers.

Over-expression and purification of recombinant proteins
E. coli NiCo21(DE3) cells carrying the expression plasmid (Table 1) were grown in LB medium containing 0.5 M sorbitol, 2.5 mM betaine, 1% glucose, 50 μg/ml kanamycin and 20μg/ml chloramphenicol at 37˚C until OD 600nm % 0.4, then the temperature was reduced to 18˚C. 1 mM isopropyl β-D-thiogalactopyranoside (IPTG) was added and the culture incubated overnight. Cells were harvested, resuspended in lysis buffer (10 mM immidazole, 50 mM NaH 2 PO 4, 300 mM NaCl, 20 mM β-mercaptoethanol, Complete Protease Inhibitor (Roche), 2 mM MgCl 2 , Benzonase (Novagen)) and disrupted by sonication. Cleared lysate was loaded onto a Ni-NTA column (Qiagen) and the column was washed according to the manufacturer's instructions. Bound proteins were eluted with buffer containing 250 mM imidazole. The fractions containing the recombinant protein were dialysed overnight against dialysis buffer Stress response of the helicase DinG (20 mM Tris-HCl pH 7.5, 100 mM NaCl, 2 mM dithiothreitol (DTT), and subjected to further purification by ion exchange chromatography using a Resource Q column (GE Healthcare) equilibrated with a buffer containing 20 mM Tris pH 7.5, 100 mM NaCl, 2 mM DTT. Proteins were eluted with an increasing ionic strength gradient up to 1 M NaCl. Fractions with pure protein were dialyzed against 20 mM Tris pH 7.5, 600 mM NaCl, 2 mM DTT, pooled and concentrated by ultrafiltration (Amicon). The protein was stored in 20 mM Tris-HCl pH 7.5, 300 mM NaCl, 10% glycerol at -80˚C. The over-expression and purification of SSB Nm and SSB NmΔ8C were done as described earlier [74].
ATPase, DNA binding and unwinding assays All DNA oligonucleotides used in this study were purchased from Operon Biotechnologies, Inc., and the sequences were adopted from previous studies [75]. DNA substrates were prepared as described elsewhere [76]. The oligonucleotides used in this study are given in S2

Ammonium sulfate co-precipitation
Co-precipitation experiments were performed as described before [77]. Briefly, 20 μM of DinG Nm was incubated with 80 μM SSB Nm in 20 μl co-precipitation buffer (10 mM Tris-HCl, pH 7.2, 300 mM NaCl, 10% [v/v] glycerol) on ice for 15 min. 10 μl of ammonium sulphate solution (450 g/l) was added, the reaction incubated on ice for 15 min, and then centrifuged for 1 min at 18,000×g. The pellets were washed three times with 50 μl co-precipitation buffer including 150 g/l ammonium sulphate and then dissolved in 50 μl 1× NuPAGE LDS sample buffer (Thermo Fisher Scientific). 10% polyacrylamide gels were loaded with 10 μl sample/ well, run at 15 V/cm and stained with Coomassie brilliant blue.

Microscale thermophoresis
Microscale thermophoresis (MST), a method for measuring molecule interaction, is described extensively elsewhere [78]. Labelling of SSB Nm was carried out following the manufacturers' instructions using the Monolith NT Protein Labeling Kit RED-NHS (NanoTemper Technologies GmbH) resulting in a degree of labelling (DOL) of 0.7. Different concentrations of DinG Nm where incubated with 20.7 nM SSB Nm in 20 mM HEPES buffer (pH 7.5) containing 300 mM NaCl, 0.05% Tween 20, 0.1% Pluronic F-127, 0.1% PEG 8000 and 2 mM DTT. Samples were immediately loaded into Premium coated capillaries (NanoTemper Technologies GmbH) and measured at 22˚C and 20% MST power in a Monolith NT.115 series instrument (NanoTemper Technologies GmbH).

Iron chelation assay
The amount of iron present in purified protein was determined by using bathophenanthroline as described earlier [17]. Briefly, 50 μl of a known amount of protein was denatured using 15 μl of 38% HCl at 100˚C for 15 min, and centrifuged to pellet insoluble material. The control reaction contained only protein storage buffer. The supernatant was collected, neutralized with 650 μl, 0.5 M Tris-HCl (pH 8.5) and freshly prepared 50 μl of 5% ascorbic acid and 200 μl of 0.1% [w/v] bathophenanthroline disulfonic acid disodium salt (Sigma) were added. After incubation at room temperature for 1 h, absorbance was measured at 535 nm. The iron concentration was calculated by employing a standard curve set up with ferrous ammonium sulfate hexa hydrate of known concentration (Sigma). The assay was performed three times.

Construction of the NmΔdinG mutant
To generate the NmΔdinG mutant, a vector was constructed to be used in transformation for replacing dinG with the marker gene aph. The vector backbone was amplified by PCR with the primers JEE61 and JEE62 with vector pJet1.2 as template. The aph was amplified with the primer 8184OHA_AphEcoRI_REV and 8186DUS_AphNheI_FOR from pUP6. Sequences adjacent to dinG were amplified with the primer pairs SF179/SF180 and SF181/SF182 from genomic DNA from MC58 (Table 1). Gibson assembly was performed to combine the sequences and the product was transformed into E. coli ER2566. Positive clones were selected for by growth on agar plates containing 50 μg/ml kanamycin and 100 μg/ml ampicillin. The correctness of the final plasmid pSAF93B was confirmed by restriction digest and sequencing ( Table 1). The plasmid pSAF93B was transformed into the Neisseria strains MC58, M1080, M400, MS11 and N400 by natural transformation using kanamycin resistance for the selection. Positive clones were checked by PCR for proper insertion of the aph.

DNA damage sensitivity assays
Nm cells from overnight plate culture were suspended in liquid GC medium to OD 660 % 0.3, and diluted 10 times in CO 2 saturated GC medium containing IsoVitaleX. The cells were allowed to grow for two hours at 37˚C with tumbling. Then the cells were treated separately with the following chemicals; 100 μM H 2 O 2 [3], 0.5 mM paraquat [3], 10 mM MMS [3] or 10 μg/l MMC [79]. After additional growth for 1 h with tumbling at 37˚C, a tenfold serial dilutions were prepared in PBS and 50 μl aliquots of the 10 −5 and 10 −6 dilutions were spread out on GC agar plates. To test sensitivity to ultraviolet radiation, 50 μl aliquots of the 10 −5 and 10 −6 dilutions of untreated cells were spread on GC agar plates, irradiated at UV intensities of 0-80 J/m 2 by using a CL-1000 Ultraviolet cross linker (Upland America). Finally, the plates were incubated overnight at 37˚C with 5% CO 2 for 12 to 18 h. Colonies were counted and survival rates were calculated as the ratio of the number of colony forming units (CFU) from treated to non-treated cells. The MMC treated samples were also subjected to quantitative mass spectrometry analyses.

Proteomics analyses
Peptide characterization and quantitation were performed by electrospray-based high resolution mass spectrometry using a Q-Exactive instrument (Thermo Scientific). A detail description of sample pre-treatment, preparation and mass spectrometry analysis can be found in the S1 Text.

Quantitative transformation assay
Quantitative transformation was performed essentially as previously described [12,52]. Briefly, N. meningitidis cells were grown on GC plates overnight at 37˚C and suspended in CO 2 saturated GC medium containing IsoVitaleX and 7 mM MgCl 2 . A 500 μl of cell suspension was mixed with 5 μl of plasmid DNA (pOHA-D1, 100 ng/μl) and incubated at 37˚C for 15 min without agitation. In order to degrade extracellular DNA, benzonase (25 U/ml) was added and incubated for 10 min. Then 4.5 mL of pre-warmed CO 2 saturated GC medium was added and the culture incubated at 37˚C with tumbling for 4.5 h. Of each sample, 50 μl aliquots were spread on GC agar plates containing 8 μg/ml erythromycin and 100 μl of the 10 −5 and 10 −6 samples diluted in PBS were spread on plain GC agar plates. Following overnight incubation at 37˚C and 5% CO 2 colonies were counted. Transformation frequencies were calculated as the number of antibiotic-resistant colony forming units (CFU) per total CFU. The experiments were repeated at least three times.  Table. Oligonucleotides. (PDF) S1 Text. Materials and methods.

Supporting information
(PDF)